Datasets:

gabrielaltay
/

pmcoa

Dataset card Data Studio Files Files and versions Community

Split (2)

train · 4.94M rows

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 Eur J Pediatr Eur J Pediatr European Journal of Pediatrics 0340-6199 1432-1076 Springer Berlin Heidelberg Berlin/Heidelberg 4718 10.1007/s00431-022-04718-y Research Group A streptococcal disease in paediatric inpatients: a European perspective Boeddha Navin P. 12 Atkins Lucy 3 de Groot Ronald 4 Driessen Gertjan 15 Hazelzet Jan 6 Zenz Werner 7 Carrol Enitan D. 89 Anderson Suzanne T. 10 Martinon-Torres Federico 11 Agyeman Philipp K. A. 12 Galassini Rachel 13 Herberg Jethro 13 Levin Michael 13 Schlapbach Luregn J. 14 Emonts Marieke [email protected] 31516 1 grid.416135.4 0000 0004 0649 0805 Department of Pediatrics, Erasmus MC-Sophia Children’s Hospital, Rotterdam, the Netherlands 2 grid.416213.3 0000 0004 0460 0556 Department of Pediatrics, Maasstad Hospital, Rotterdam, the Netherlands 3 grid.419334.8 0000 0004 0641 3236 Paediatric Immunology, Infectious Diseases & Allergy Dept., Great North Children’s Hospital, Newcastle Upon Tyne Hospitals NHS Foundation Trust, RVI, Clinical Resources Building, Queen Victoria Road, Newcastle Upon Tyne, NE1 4LP UK 4 grid.461760.2 0000 0004 0580 1253 Division of Pediatric Infectious Diseases and Immunology and Laboratory of Infectious Diseases, Department of Pediatrics, Radboud Institute of Molecular Life Sciences, Radboudumc, Nijmegen, the Netherlands 5 grid.412966.e 0000 0004 0480 1382 Department of Paediatrics, Maastricht University Medical Center, Maastricht, the Netherlands 6 grid.5645.2 000000040459992X Department of Public Health, Erasmus MC, Rotterdam, the Netherlands 7 grid.11598.34 0000 0000 8988 2476 Department of General Pediatrics, Medical University of Graz, Graz, Austria 8 grid.10025.36 0000 0004 1936 8470 Institute of Infection, Veterinary and Ecological Sciences Global Health, University of Liverpool, Liverpool, UK 9 grid.417858.7 0000 0004 0421 1374 Alder Hey Children’s NHS Foundation Trust, Liverpool, UK 10 grid.415063.5 0000 0004 0606 294X Medical Research Council Unit The Gambia at LSHTM, Fajara, The Gambia 11 Translational Pediatrics and Infectious Diseases Section, Pediatrics Department, Santiago de Compostela, Spain 12 grid.411656.1 0000 0004 0479 0855 Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland 13 grid.7445.2 0000 0001 2113 8111 Section of Paediatrics Division of Infectious Disease, Imperial College of London, London, UK 14 grid.412341.1 0000 0001 0726 4330 Neonatal and Pediatric Intensive Care Unit, University Children`s Hospital Zürich and Children`s Research Center, Zurich, Switzerland 15 grid.1006.7 0000 0001 0462 7212 Translational and Clinical Research Institute, Newcastle University, Newcastle Upon Tyne, UK 16 grid.454379.8 NIHR Newcastle Biomedical Research Centre Based at Newcastle Upon Tyne Hospitals NHS Trust and Newcastle University, Newcastle Upon Tyne, UK Communicated by Daniele De Luca 30 11 2022 110 15 10 2022 13 11 2022 14 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. Group A streptococcal (GAS) disease shows increasing incidence worldwide. We characterised children admitted with GAS infection to European hospitals and studied risk factors for severity and disability. This is a prospective, multicentre, cohort study (embedded in EUCLIDS and the Swiss Pediatric Sepsis Study) including 320 children, aged 1 month to 18 years, admitted with GAS infection to 41 hospitals in 6 European countries from 2012 to 2016. Demographic, clinical, microbiological and outcome data were collected. A total of 195 (61%) patients had sepsis. Two hundred thirty-six (74%) patients had GAS detected from a normally sterile site. The most common infection sites were the lower respiratory tract (LRTI) (22%), skin and soft tissue (SSTI) (23%) and bone and joint (19%). Compared to patients not admitted to PICU, patients admitted to PICU more commonly had LRTI (39 vs 8%), infection without a focus (22 vs 8%) and intracranial infection (9 vs 3%); less commonly had SSTI and bone and joint infections (p < 0.001); and were younger (median 40 (IQR 21–83) vs 56 (IQR 36–85) months, p = 0.01). Six PICU patients (2%) died. Sequelae at discharge from hospital were largely limited to patients admitted to PICU (29 vs 3%, p < 0.001; 12% overall) and included neurodisability, amputation, skin grafts, hearing loss and need for surgery. More patients were recruited in winter and spring (p < 0.001). Conclusion: In an era of observed marked reduction in vaccine-preventable infections, GAS infection requiring hospital admission is still associated with significant severe disease in younger children, and short- and long-term morbidity. Further advances are required in the prevention and early recognition of GAS disease. What is Known: • Despite temporal and geographical variability, there is an increase of incidence of infection with group A streptococci. However, data on the epidemiology of group A streptococcal infections in European children is limited. What is New: • In a large, prospective cohort of children with community-acquired bacterial infection requiring hospitalisation in Europe, GAS was the most frequent pathogen, with 12% disability at discharge, and 2% mortality in patients with GAS infection. • In children with GAS sepsis, IVIG was used in only 4.6% of patients and clindamycin in 29% of patients. Supplementary Information The online version contains supplementary material available at 10.1007/s00431-022-04718-y. Keywords Streptococcus pyogenes Child Hospital Outcome http://doi.org/10.13039/501100004963 Seventh Framework Programme EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 EUCLIDS Grant Agreement no. 279185 Swiss National Science Foundation342730_153158/1 342730_153158/1 Swiss Society of Intensive CareBangerter FoundationVinetum and Borer FoundationFoundation for the Health of Children and Adolescents ==== Body pmcIntroduction Group A streptococcal (GAS) infection is characterised by a wide variety of phenotypes, from upper respiratory tract and focal skin and soft tissue infections to necrotising fasciitis and streptococcal toxic shock syndrome (STSS), as well as peri-infectious phenomena such as rheumatic fever and post-streptococcal glomerulonephritis [1]. Extensive strain diversity—there are more than 200 emm-types—is presumed to contribute to the diversity of observed clinical syndromes [2–4]. In Western Europe, GAS was a leading cause of child death until the mid-twentieth century [5]. Although the incidence fell rapidly during the mid-twentieth century, it began rising again in the 1980s, and was estimated to be 3–4/100,000 in Northern Europe by the early 2000s [6]. Surges continue in Europe [7, 8], and further afield, for instance, South Korea [9, 10] and Utah, where a 2010 study found a rate of up to 14.1/100,000 cases in children [6, 11]. Severity of illness also seems to be increasing [12–14]. Mean case fatality rates in affluent countries remain relatively low at 8–16% [5]. However, the mortality of iGAS can rise rapidly up to 60–70% with any delay in antibiotics and interventions against toxins such as clindamycin and intravenous immunoglobulin [15–17]. Epidemiological studies capturing the full spectrum of GAS disease are challenging. Many cases never present to health care providers, or are treated in the community without microbiological diagnostics (tonsillitis, cellulitis). Studies have shown that microbiologically proven paediatric iGAS occurs most commonly with bacteraemia, soft tissue infections, STSS or necrotizing fasciitis [18], and that risk factors include other children in the household, preceding coryzal illness and varicella infection [19]. However, it has been difficult to study a wider population of children with severe bacterial infections where GAS is the likely cause but not isolated from a sterile site. In this large study, we aimed to characterise children admitted with GAS infection to European hospitals and study risk factors for severity and disability. We gathered information about possible risk factors, clinical presentation, progress and outcomes for children presenting with suspected severe bacterial infection for whom the cause was proven or probable GAS. Methods Consortium and study sites This study used data from the European Childhood Life-threatening Infectious Disease Study (EUCLIDS) and the Swiss Paediatric Sepsis Study (SPSS) [20, 21]. EUCLIDS is a prospective, multicentre, cohort study aimed to identify genes and pathways determining susceptibility and severity of life-threatening bacterial infections. The network included 185 predominantly academic hospitals from 8 European countries. Details of EUCLIDS inclusion and exclusion criteria, as well as clinical definitions, have been published elsewhere [20]. In short, for this sub-cohort, children with suspected severe bacterial infection were recruited prospectively from 1 July 2012 to 31 December 2016, as early as possible in admission and before culture results became available. The SPSS is a prospective, national, observational, multicentre, cohort study investigating blood culture–proven sepsis in children under 17 years of age from all 10 major children’s hospitals in Switzerland from 1 January 2012 to 31 December 2015 [21]. In brief, children with blood culture–proven sepsis meeting the criteria for systemic inflammatory response syndrome (SIRS), as defined by the 2005 paediatric consensus definition [22] at the time of blood culture sampling, were included. Details of the study design and the study protocol have been published elsewhere [21]. Inclusion criteria The combined EUCLIDS-SPSS database was used to identify all children for whom GAS was the most likely cause of their illness. Children were included if they met any of the following criteria:Proven GAS: GAS grown from a normally sterile site or positive by pathogen-specific PCR (blood, CSF, joint fluid, pleural fluid, peritoneal fluid, tissue, urine, intra-operative pus or internal swab). GAS detection by PCR was performed according to local accredited hospital or specialised molecular microbiology laboratories. Probable GAS: Clinical symptoms consistent with GAS disease, NO other causative organism identified AND at least one of the following:GAS grown from a potential carriage site (throat, naso-pharynx, eye surface, ear, endo-tracheal tube, broncho-alveolar lavage, skin) Antistreptolysin O titre (ASOT) ≥ 300 IU/L [23] Local rapid streptococcal antigen test (RST) from pharyngeal sample positive. Cases were excluded if they had been enrolled retrospectively to avoid selection bias, or were from the non-European EUCLIDS sites. In addition, patients in whom other potential causative pathogens were detected from sterile or non-sterile site cultures were excluded. Sepsis, severe sepsis and septic shock were defined according to Goldstein criteria, and focal infection was used for patients with an organ system identifiable febrile illness not matching sepsis according to Goldstein criteria [22]. Clinical data collection Data on demographics, clinical presentation, underlying disease, exposure to varicella-zoster virus (VZV), smoking, recent surgery, illness severity, management, microbiological results and outcome were collected prospectively. Exposure to VZV or smoking was not available for the Swiss patients. Underlying diseases at admission to hospital were classified using the Paediatric Complex Chronic Conditions classification system [24]. Illness severity in PICU patients was measured by the Paediatric Risk of Mortality score (EUCLIDS patients only) [25] and the Paediatric Index of Mortality-2 (PIM2) [26]. Lactate values were obtained on PICU admission only, concomitant with PIM2 data collection. Outcome included mortality, disability, PICU-free days and length of hospital stay. Disability was defined as a Pediatric Overall Performance Category (POPC) score greater than 1 [27], need for skin graft, amputation, hearing loss, neurodisability or need for surgery. The POPC score was determined either by direct observation or by chart review and ranges from 1 to 6, varying from (1) good overall performance to (6) brain death [27] (Supplementary Table 1 for description of categories). PICU-free days (days alive and free from the need for intensive care) were censored at day 28. In patients who died, PICU-free days were considered zero. Patients were grouped as no focus (primary bloodstream infection and sepsis without a known source) versus patients with a clinical focus of infection. All data were collected in web-based case report forms. Monthly telephone conferences, biannual meetings, clinical protocols including case definitions, data audits and monitoring ensured uniform procedures amongst study sites. Statistical analysis Categorical variables were presented as counts (percentages). The chi-square or Fisher’s exact test was used to compare frequency distributions between two categorical variables. Continuous variables were presented as median (interquartile range (IQR)) for non-parametric data. ANOVA, Kruskal–Wallis, Student’s t, or Mann–Whitney U tests were used to test differences between groups, as appropriate. Statistical analysis was performed with IBM® SPSS version 24 (Armonk, USA). A p value of less than 0.05 was considered statistically significant. Results During the study period, 346/4025 (9%) of the children prospectively enrolled at any of the participating hospitals had proven/probable GAS disease. Other commonly identified pathogens were Neisseria meningitidis, Staphylococcus aureus and Streptococcus pneumoniae in about 8% of patients each [20]. Twenty-two GAS patients were excluded because they had been recruited retrospectively as part of the genetics study, and 4 because they were from non-European EUCLIDS sites (Fig. 1), leaving 320 patients for analyses.Fig. 1 Study design. CONSORT flow chart including and excluding patients Demographics and clinical spectrum A total of 161 (50%) were male, with median age 47 (IQR 27–84) months. 47 (15%) presented without a clinical focus of infection. In children with a focus of infection, bone/joint, soft tissue or respiratory tract infection was the predominant presentation. One or more underlying conditions were present in 117 (37%) patients of which the most common were recent VZV (n = 21, 6.6%), congenital or genetic conditions (n = 18, 5.6%) and eczema (n = 11, 3.4%). Further details are presented in Table 1.Table 1 Characteristics of children admitted with GAS infection All patients (n = 320) No PICU admission (n = 172) PICU admission (n = 148) p Sex (male, n, %) 161 (50%) 86 (50%) 75 (51%) 0.9 Age (months) (IQR) 47 [27–84] 57 [36–85] 40 [21–83] 0.01 Time interval onset symptoms to hospital admission (n = 251, days) 3.0 [1.8–6.0] 3.5 [1.8–6] 3.0 [1.8–5.7] 0.67 Immunizations up-to-date (n = 222) 211 (95%) 103/110 (94%) 108/112 (96%) 0.06 Number of underlying conditions None 203 (63%) 107 (62%) 96 (65%) 0.11 1 77 (24%) 48 (28%) 29 (20%) ≥ 2 40 (13%) 17 (10%) 23 (16%) Underlying conditions Congenital or genetic defect 18 (5.6%) 4 (2.3%) 14 (9.5%) 0.006 Prematurity 10 (3.1%) 5 (2.9%) 5 (3.4%) 0.11 Immunodeficiency 6 (1.9%) 1 (0.6%) 5 (3.4%) 0.19 Cardiac condition 5 (1.6%) 2 (1.2%) 3 (2.0%) 0.47 Epilepsy 5 (1.6%) 0 5 (3.4%) 0.01 Respiratory 9 (2.8%) 6 (3.5%) 3 (2.0%) 0.12 Haematological 1 (0.3%) 0 1 (0.7%) 0.28 Oncological 1 (0.3%) 1 (0.6%) 0 0.35 Inflammatory 2 (0.6%) 1 (0.6%) 1 (0.7%) 0.36 Liver disease 1 (0.3%) 0 1 (0.7%) 0.28 Renal disease 0 0 0 NA Metabolic disease 2 (0.6%) 1 (0.6%) 1 (0.7%) 0.36 Recent surgery 3 (0.9%) 1 (0.6%) 2 (1.4%) 0.48 Eczema/dermatitis 11 (3.4%) 6 (3.5%) 5 (3.4%) 0.96 Recent chickenpox 21 (6.6%) 10 (5.8%) 11 (7.4%) 0.09 Primary infection site < 0.001 None 47 (15%) 14 (8%) 33 (22%) Lower respiratory tract 71 (22%) 14 (8%) 57 (39%)* Skin/Soft tissue 73 (23%) 50 (29%) 23 (16%) Bone/joint 60 (19%) 52 (30%) 8 (5%) Upper respiratory tract 46 (14%) 34 (20%) 12 (8%) Intracranial 18 (6%) 5 (3%) 13 (9%) Peritoneal 3 (1%) 1 (1%) 2 (1%) Renal 2 (1%) 2 (1%) 0 Microbiology 0.11 Sterile site positive culture or PCRa 236 (74%) 120 (70%) 116 (78%) Blood 145 (61%) 80 (67%) 65 (56%) CSF 5 (2%) 2 (2%) 3 (3%) Joint fluid 18 (8%) 17 (14%) 1 (1%) Pleural fluid 37 (16%) 3 (3%) 34 (30%) Peritoneal fluid 1 (0.4%) 0 1 (1%) Tissue 4 (2%) 1 (1%) 3 (3%) Urine 1 (0.4%) 1 (1%) 0 Abscess/pus 26 (11%) 15 (13%) 11 (10%) Intra-operative swab 13 (6%) 6 (5%) 7 (6%) GAS clinical syndrome, AND 84 (26%) 52 (30%) 32 (22%) Potential carriage site positive 68 (21%) 40 (23%) 28 (19%) Elevated ASOT 7 (2%) 4 (2%) 3 (2%) Pharyngeal RST positive 9 (3%) 8 (5%) 1 (1%) Inflammatory markers Max CRP (n = 256, mg/L) 185 (80–286) 119 (62–228) 256 (149–328) < 0.001 Hospital length of stay (n = 318, days) 11 [6-18] 8 [4-13] 17 [11-26] < 0.001 PICU-free days at day 28 (n = 318) 28 (23–28) 28 23 [18-25] < 0.001 ASOT antistreptolysin O titre, CRP C-reactive protein, CSF cerebrospinal fluid, GAS group A streptococcal, PCR polymerase chain reaction, PICU paediatric intensive care unit, RST rapid streptococcal antigen test *34/57 ((69%) patients in PICU and 4/14 (29%) patients not requiring PICU for lower respiratory tract infection had pleural empyema aBreakdown exceeds 100% as GAS could have been identified from multiple sources per patient Characteristics of PICU cases and risk factors for severe disease A total of 148 (46%) children were admitted to PICU. One hundred five (71%) patients in PICU required invasive ventilation, with a median (IQR) of 5 (3–8) days (n = 92). Eighty-eight (59%) PICU patients required inotropes, with a median (IQR) of 3 (2–4.3) days (n = 74). None of the patients required extracorporeal membrane oxygenation. The median (IQR) PRISM score was 14 (8–20, n = 72) and PIM2 6.1% (1.4–11.5, n = 88) predicted death rate. Median lactate on PICU admission was 1.5 mmol/L (IQR 0.9–2.4, n = 90). The number of PICU-free days at day 28 in the PICU group was 23 days (IQR 18–25) (Table 1). The proportion of patients admitted to PICU differed between countries (p = 0.026; Supplementary Fig. 1). Patients admitted to PICU, compared to those not admitted to PICU, were younger (40 [21–83] vs 57 [36–85] months, respectively, p = 0.01), had more frequent epilepsy (p = 0.01) and congenital and genetic defects (p = 0.005) as underlying conditions, and more often had LRTI, intracranial infection and infection without a focus, whereas SSTI and bone and joint infections were more common in patients not requiring PICU (p < 0.001) (Table 1). Also, the maximum CRP in PICU patients was significantly higher and hospital admission was associated with a two-fold increase in duration. Time from onset of symptoms to admission to hospital did not differ between PICU and non-PICU patients. Eczema (n = 11) and recent VZV infection (n = 21) were not associated with PICU admission. Also, there was no difference in exposure to smoking between PICU and non-PICU patients (20 (12%) and 29 (20%), respectively, p = 0.13). Forty-two children (13%) had a severe infection (defined as a clinical syndrome suspected for severe invasive bacterial disease such as septicaemia, toxic shock syndrome, pneumonia, empyema, meningitis, osteomyelitis and septic arthritis) in the previous medical history, but this was not related to increased risk of PICU admission. In fact, those with a severe infection in the past were less often admitted to PICU (7.4 vs 18%, p = 0.02). This could not be explained by a difference in onset of symptoms until admission (p = 1.0). Microbiology The diagnosis of GAS infection was based on a positive culture and/or PCR from a normally sterile site in 236 (74%) patients, of which 145 (61%) in blood and 37 (16%) in pleural fluid were the most common sites (Table 1). For the remaining patients, GAS clinical syndrome was determined by the local team based on positive culture and/or PCR from a potential carriage site (n = 68, 21%), elevated ASOT (n = 7, 2%) and a positive pharyngeal RST (n = 9, 3%). Streptococcal titres were measured with a median of 16 days (IQR 10–28) after onset of symptoms and 8 days (IQR 5–12) after admission to the hospital. There was no difference in the means of GAS identification between non-PICU and PICU patients (p = 0.11). Seasonality A seasonal pattern was noted with more GAS-infected patients recruited in the winter and spring months (n = 223 (70%), December–May) compared to the rest of the year (n = 97 (30%), June–November, p < 0.001; Fig. 2).Fig. 2 Numbers of patients with GAS disease per month (2012–2016 combined). Accumulated numbers of patients with GAS disease over the 5-year study period per month of presentation. Total n = 320 Sepsis versus focal infection A total of 195 (61%) patients had sepsis, of which 47 (24%) without a focus, and in 125 (39%) patients disease was limited to a focal infection. Patients with sepsis tended to be younger (median 44 months) compared to patients with focal infection (median 57 months, p = 0.07). Sex distribution was not significantly different for patients with sepsis or focal infection. Patients with sepsis relatively more often had LRTI (25.1 vs 17.6%) as focus of infection compared to the other foci (p < 0.001). Sepsis was associated with a higher CRP than focal infection (median 228 (IQR 114–303) vs 111 (IQR 53–211); p < 0.001). The proportion of patients with sepsis was higher in those requiring PICU (n = 116, 78%) than in those not requiring PICU admission (n = 79, 46%, p < 0.001) (Table 2). All but one (73/74 (99%)) patient with septic shock or toxic shock were admitted to PICU. Patients with sepsis more often had GAS identified from a normally sterile site than those with focal infection (n = 168, 86% vs n = 68, 54%; p < 0.001).Table 2 Sepsis in iGAS infection Sepsis severity No PICU, N = 172 PICU, N = 148 p None 93 (54%) 32 (22%) p < 0.001 Sepsis 74 (43%) 34 (23%) Severe sepsis 4 (2%) 9 (6%) Septic shock 1 (1%) 58 (40%) Toxic shock syndrome 15 (10%) iGAS invasive group A streptococcal, PICU paediatric intensive care unit With regard to adjunctive treatment of patients with sepsis, intravenous immunoglobulin (IVIG) was administered in 9/195 (4.6%) patients and administration did not differ between countries (7/86 (8.1%) patients with sepsis from the UK and 2/39 (5.1%) patients with sepsis from the Netherlands; p = 0.32). However, clindamycin was prescribed in 57/195 (29%) sepsis patients and prescription rate was different between countries (39/86 (45%) from the UK, 8/39 (21%) from the Netherlands, 4/7 (57%) from Spain, 4/56 (7%) from Switzerland, 1/3 (33%) from Austria and 1/4 (25%) from Germany; p < 0.001). Outcome Six children died, reflecting a crude mortality of 2% (Table 3). Overall, 231 (72%) children survived without disability and 39 (12%) with disability, including 23/168 (14%) children who did not have an underlying condition at hospital admission, i.e. previously healthy children. For the remaining 44 (14%) patients, information on disability was not classified. The majority of patients where disability was not classified were transferred back to their local hospital for ongoing care, from which point no reliable judgement could be made regarding full recovery. The proportion of survivors without disability was lower for those admitted to PICU (57%, p < 0.001). Age was not associated with outcome (disability vs no disability). Skin graft and need for surgery were seen in patients with sepsis and focal infection, but other complications, such as death, amputation and neurodisability, were observed at discharge in patients with sepsis only, and limited to those admitted to PICU (Fig. 3).Table 3 Outcome of patients with GAS disease All patients (n = 320) No PICU admission (n = 172) PICU admission (n = 148) p Died 6 (2%) 0 6 (4%) < 0.001 Survived with disability 39 (12%) 5 (3%) 34 (23%) Mild overall disability 20 (6%) 2 (1%) 18 (12%) Moderate overall disability 5 (2%) 0 5 (3%) Severe overall disability 3 (1%) 0 3 (2%) Amputation 4 (1%) 0 4 (3%) Skin graft 9 (3%) 0 9 (6%) Amputation and skin graft 2 (1%) 0 2 (1%) Need for surgery 10 (3%) 4 (2%) 6 (4%) Neurodisability 2 (1%) 0 2 (1%) Survived without disability 231 (72%) 146 (85%) 85 (57%) Unknown 44 (14%) 21 (12%) 23 (16%) GAS group A streptococcal, PICU paediatric intensive care unit Fig. 3 Outcome of GAS disease in PICU and ward patients. Relative outcomes of patients with GAS per admission category (PICU or non-PICU) Discussion In this European cohort, GAS is a significant cause of probable or confirmed severe bacterial infection, with a significant burden of mortality and persistent morbidity. Risk factors for PICU admission were lower age, LRTI, intracranial infection and infection without a focus. The need for PICU admission did not seem to be related to delayed presentation. The proportion of patients admitted to PICU differed between countries. A survey amongst participating centres showed that criteria for PICU admission and availability of resources differed between centres. In some centres, non-invasive ventilation (e.g. CPAP) was only supported in PICU, whilst in others it could be offered on a paediatric ward or high dependency unit (unpublished). In addition, except for Switzerland, the participating centres were not representing the entire population, and selection bias may contribute to differences. Data from this study originates from two separate cohorts: EUCLIDS and SPSS. In EUCLIDS, recruitment of patients took place on admission and largely before the causative pathogen was known. Only patients with suspected severe bacterial infection admitted to the hospital were recruited, which means those with milder infections not requiring admission were not included. Also, due to the nature of the study, it is not clear exactly what proportion of overall eligible children was recruited for the study. In SPSS, only children with blood culture–proven sepsis were recruited, meaning that children with GAS disease and negative blood culture were omitted. Therefore, data from our study could be an underestimation of the true impact of GAS disease in Europe and should be interpreted with caution. The overall mortality (2%) was comparable to other studies on bacteraemic children including all patients in hospital [21, 28], but was lower than most previously reported mortality rates in patients admitted to PICU [5]. When assessing severity, considering mortality alone risks underestimating the true impact of iGAS. Whilst for most patients full recovery at hospital discharge was noted, significant sequelae were identified in 12% overall, increasing to 23% for those admitted to PICU. As patients were not followed after hospital discharge, no information is available on potential resolution of some of the sequelae and longer term morbidity and functional outcome related to iGAS disease. In addition, no data on baseline POPC scores were available, which means pre-existing comorbidity could not be taken into account assessing the difference in functioning pre- and post-infection. Overall, 74% of patients had GAS identified from a sterile site. For 26% of patients, a positive potential carriage site, a rapid antigen test or raised ASOT was the only method of microbiological GAS confirmation. We acknowledge these diagnostic methods as a limitation of our study. However, whilst analysis of proven GAS infections might be the gold standard, it is recognised that GAS cannot be cultured in all patients. By including probable cases, we better reflected the actual demographics of GAS disease. We reduced the risk of including non-GAS cases by excluding patients in whom other potential causative pathogens were detected from sterile or non-sterile site cultures. Only patients in whom diagnostics exclusively identified GAS were included. As not all GAS isolates were kept, we were unable to obtain their M types to assess potential association with phenotype and severity. The variability in these proteins, associated with diversity in disease phenotypes, makes development of a generic GAS vaccine challenging [2, 29, 30]. Future research focussing on bacterial phenotypes related to LRTI, intracranial infection and sepsis might help prioritise vaccine development, in order to prevent most severe disease. Interestingly, only a few patients with recent VZV infection were identified. Whilst it is well-known that VZV increases vulnerability to iGAS infection [16], our study confirms that GAS infection predominantly occurs in individuals with no obvious risk factors. It has to be acknowledged that for patients recruited in Switzerland, unfortunately, VZV and other exposures were not recorded. Despite the fact that this study was not purposely designed to study epidemiology, a seasonal variation was noted. Patients were recruited to the study early during the admission when the cause of infection was not yet known, limiting recruitment bias. Most patients with GAS infection were recruited in winter and spring, with a clear reduction in the summer months, whilst recruitment took place year-round. Seasonal increase has also been noted in Australian children, where iGAS coincides with the influenza season [31]. An increased incidence over the winter months has also been reported in Hong Kong, South Korea, the USA, Iceland and other European countries [6, 10, 32–36]. Conclusion Our study showed that LRTI, intracranial infection and infection without a focus more commonly resulted in severe GAS disease requiring admission to PICU. PICU admission for GAS infection was associated with worse outcomes with regard to mortality and disability. With increasing incidence of iGAS disease worldwide, and increased morbidity and mortality in those requiring PICU, future research should focus on prevention of iGAS infection. Vaccination development should target iGAS serotypes associated with severe disease requiring PICU admission. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 46 KB) Acknowledgements A list of authors and their affiliations appears in the Supplementary Information. Authors’ contributions FMT, JH, EDC, ME, RdG, WZ and ML designed the study and obtained funding. JH, NPB, PA, LJS, GJD, STA, ME, LJS and EDC assisted with recruitment of patients, data collection and sample collection. ME and NPB did the statistical analyses. LA and RG provided database and informatics support. LA, ME, NPB, LJS and GJD wrote the first draft of the manuscript. FMT, JAH, RdG, WZ, EDC, STA, FMT, PA, RG, JH and ML contributed to writing of the manuscript. All authors approved the final manuscript. Funding This work was supported by the European Seventh Framework Programme for Research and Technological Development (FP7) under EUCLIDS Grant Agreement no. 279185. The Swiss Pediatric Sepsis Study was funded by grants from the Swiss National Science Foundation (342730_153158/1), the Swiss Society of Intensive Care, the Bangerter Foundation, the Vinetum and Borer Foundation and the Foundation for the Health of Children and Adolescents. Data Availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate This study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. The study protocol was approved by at least one ethical review board in every country (Coordinating Center Research Ethics Committee reference: 11/LO/1982). Written informed consent was obtained from parents or legal guardians. In the Swiss study, consent was obtained for collection of research blood, but waiver of consent for collection of anonymized epidemiological data was approved. Competing interests The authors declare no competing interests. Disclaimer The funders were not involved in the design of the study, collection, analysis, interpretation of data or in writing the manuscript. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Navin P. Boeddha and Lucy Atkins contributed equally. ==== Refs References 1. Dietrich ML Steele RW Group A Streptococcus Pediatr Rev 2018 39 8 379 391 10.1542/pir.2017-0207 30068739 2. Shulman ST Tanz RR Dale JB Steer AC Smeesters PR Added value of the emm-cluster typing system to analyze group A Streptococcus epidemiology in high-income settings Clin Infect Dis 2014 59 11 1651 1652 10.1093/cid/ciu649 25115872 3. Davies MR McIntyre L Mutreja A Lacey JA Lees JA Towers RJ Atlas of group A streptococcal vaccine candidates compiled using large-scale comparative genomics Nat Genet 2019 51 6 1035 1043 10.1038/s41588-019-0417-8 31133745 4. Sanderson-Smith M De Oliveira DM Guglielmini J McMillan DJ Vu T Holien JK A systematic and functional classification of Streptococcus pyogenes that serves as a new tool for molecular typing and vaccine development J Infect Dis 2014 210 8 1325 1338 10.1093/infdis/jiu260 24799598 5. Steer AC Lamagni T Curtis N Carapetis JR Invasive group a streptococcal disease: epidemiology, pathogenesis and management Drugs 2012 72 9 1213 1227 10.2165/11634180-000000000-00000 22686614 6. Lamagni TL Darenberg J Luca-Harari B Siljander T Efstratiou A Henriques-Normark B Epidemiology of severe Streptococcus pyogenes disease in Europe J Clin Microbiol 2008 46 7 2359 2367 10.1128/JCM.00422-08 18463210 7. Plainvert C Loubinoux J Bidet P Doloy A Touak G Dmytruk N Epidemiology of Streptococcus pyogenes invasive diseases in France (2007–2011) Arch Pediatr 2014 21 Suppl 2 S62 S68 10.1016/S0929-693X(14)72262-6 25456682 8. Scaber J, Saeed S, Ihekweazu C, Efstratiou A, McCarthy N, O’Moore E (2011) Group A streptococcal infections during the seasonal influenza outbreak 2010/11 in South East England. Euro Surveill 16(5) 9. Filleron A Jeziorski E Michon AL Rodiere M Marchandin H Current insights in invasive group A streptococcal infections in pediatrics Eur J Pediatr 2012 171 11 1589 1598 10.1007/s00431-012-1694-8 22367328 10. Park DW Kim SH Park JW Kim MJ Cho SJ Park HJ Incidence and characteristics of scarlet fever, South Korea, 2008–2015 Emerg Infect Dis 2017 23 4 658 661 10.3201/eid2304.160773 28322696 11. Stockmann C Ampofo K Hersh AL Blaschke AJ Kendall BA Korgenski K Evolving epidemiologic characteristics of invasive group a streptococcal disease in Utah, 2002–2010 Clin Infect Dis 2012 55 4 479 487 10.1093/cid/cis422 22534148 12. Lithgow A Duke T Steer A Smeesters PR Severe group A streptococcal infections in a paediatric intensive care unit J Paediatr Child Health 2014 50 9 687 692 10.1111/jpc.12601 24909187 13. Nasser W Beres SB Olsen RJ Dean MA Rice KA Long SW Evolutionary pathway to increased virulence and epidemic group A Streptococcus disease derived from 3,615 genome sequences Proc Natl Acad Sci USA 2014 111 17 E1768 E1776 10.1073/pnas.1403138111 24733896 14. Al-Shahib A Underwood A Afshar B Turner CE Lamagni T Sriskandan S Emergence of a novel lineage containing a prophage in emm/M3 group A Streptococcus associated with upsurge in invasive disease in the UK Microb Genom 2016 2 6 e000059 10.1099/mgen.0.000059 28348855 15. Carapetis JR Jacoby P Carville K Ang SJ Curtis N Andrews R Effectiveness of clindamycin and intravenous immunoglobulin, and risk of disease in contacts, in invasive group a streptococcal infections Clin Infect Dis 2014 59 3 358 365 10.1093/cid/ciu304 24785239 16. Laupland KB, Davies HD, Low DE, Schwartz B, Green K, McGeer A (2000) Invasive group A streptococcal disease in children and association with varicella-zoster virus infection. Ontario Group A Streptococcal Study Group. Pediatrics 105(5):E60 17. Whitehead BD Smith HV Nourse C Invasive group A streptococcal disease in children in Queensland Epidemiol Infect 2011 139 4 623 628 10.1017/S0950268810001378 20609283 18. Zachariadou L Stathi A Tassios PT Pangalis A Legakis NJ Papaparaskevas J Differences in the epidemiology between paediatric and adult invasive Streptococcus pyogenes infections Epidemiol Infect 2014 142 3 512 519 10.1017/S0950268813001386 23746128 19. Factor SH Levine OS Harrison LH Farley MM McGeer A Skoff T Risk factors for pediatric invasive group A streptococcal disease Emerg Infect Dis 2005 11 7 1062 1066 10.3201/eid1107.040900 16022781 20. Martinon-Torres F Salas A Rivero-Calle I Cebey-Lopez M Pardo-Seco J Herberg JA Life-threatening infections in children in Europe (the EUCLIDS Project): a prospective cohort study Lancet Child Adolesc Health 2018 2 6 404 414 10.1016/S2352-4642(18)30113-5 30169282 21. Agyeman PKA Schlapbach LJ Giannoni E Stocker M Posfay-Barbe KM Heininger U Epidemiology of blood culture-proven bacterial sepsis in children in Switzerland: a population-based cohort study Lancet Child Adolesc Health 2017 1 2 124 133 10.1016/S2352-4642(17)30010-X 30169202 22. Goldstein B, Giroir B, Randolph A, International Consensus Conference on Pediatric S (2005) International pediatric sepsis consensus conference: definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med 6(1):2–8. 10.1097/01.PCC.0000149131.72248.E6. 23. Pagana KD, Pagana TJ, Pagana TN (2019) Mosby’s diagnostic & laboratory test reference. 14 ed. Elsevier 24. Feudtner C Feinstein JA Zhong W Hall M Dai D Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and complex medical technology dependence and transplantation BMC Pediatr 2014 14 199 10.1186/1471-2431-14-199 25102958 25. Pollack MM Ruttimann UE Getson PR Pediatric risk of mortality (PRISM) score Crit Care Med 1988 16 11 1110 1116 10.1097/00003246-198811000-00006 3048900 26. Slater A, Shann F, Pearson G, Paediatric Index of Mortality Study G (2003) PIM2: a revised version of the Paediatric Index of Mortality. Intensive Care Med 29(2):278–85. 10.1007/s00134-002-1601-2 27. Fiser DH Assessing the outcome of pediatric intensive care J Pediatr 1992 121 1 68 74 10.1016/s0022-3476(05)82544-2 1625096 28. Odetola FO Gebremariam A Freed GL Patient and hospital correlates of clinical outcomes and resource utilization in severe pediatric sepsis Pediatrics 2007 119 3 487 494 10.1542/peds.2006-2353 17332201 29. Steer AC Law I Matatolu L Beall BW Carapetis JR Global emm type distribution of group A streptococci: systematic review and implications for vaccine development Lancet Infect Dis 2009 9 10 611 616 10.1016/S1473-3099(09)70178-1 19778763 30. Safar A Lennon D Stewart J Trenholme A Drinkovic D Peat B Invasive group A streptococcal infection and vaccine implications, Auckland New Zealand Emerg Infect Dis 2011 17 6 983 989 10.3201/eid/1706.100804 21749758 31. Oliver J Thielemans E McMinn A Baker C Britton PN Clark JE Invasive group A Streptococcus disease in Australian children: 2016 to 2018 - a descriptive cohort study BMC Public Health 2019 19 1 1750 10.1186/s12889-019-8085-2 31888568 32. Lee CF Cowling BJ Lau EHY Epidemiology of reemerging scarlet fever, Hong Kong, 2005–2015 Emerg Infect Dis 2017 23 10 1707 1710 10.3201/eid2310.161456 28930009 33. Smeesters PR Laho D Beall B Steer AC Van Beneden CA Seasonal, geographic, and temporal trends of emm clusters associated with invasive group A streptococcal infections in US multistate surveillance Clin Infect Dis 2017 64 5 694 695 10.1093/cid/ciw807 28184410 34. Nelson GE Pondo T Toews KA Farley MM Lindegren ML Lynfield R Epidemiology of invasive group A streptococcal infections in the United States, 2005–2012 Clin Infect Dis 2016 63 4 478 486 10.1093/cid/ciw248 27105747 35. O’Loughlin RE Roberson A Cieslak PR Lynfield R Gershman K Craig A The epidemiology of invasive group A streptococcal infection and potential vaccine implications: United States, 2000–2004 Clin Infect Dis 2007 45 7 853 862 10.1086/521264 17806049 36. Olafsdottir LB Erlendsdottir H Melo-Cristino J Weinberger DM Ramirez M Kristinsson KG Invasive infections due to Streptococcus pyogenes: seasonal variation of severity and clinical characteristics, Iceland, 1975 to 2012 Euro Surveill 2014 19 17 5 14 10.2807/1560-7917.ES2014.19.17.20784 24821122	36449079	PMC9709363	NO-CC CODE	2022-12-01 23:23:04	no	Eur J Pediatr. 2022 Nov 30;:1-10	utf-8	Eur J Pediatr	2,022	10.1007/s00431-022-04718-y	oa_other
==== Front Eur J Nutr Eur J Nutr European Journal of Nutrition 1436-6207 1436-6215 Springer Berlin Heidelberg Berlin/Heidelberg 3058 10.1007/s00394-022-03058-9 Original Contribution Characterizing fluid intake and physical activity in university students within the United States during the COVID-19 pandemic http://orcid.org/0000-0001-7372-8455 Adams William M. [email protected] 123 http://orcid.org/0000-0002-7031-6057 Zaplatosch Mitchell E. 3 Glenn Shaylynn E. 4 http://orcid.org/0000-0002-8720-5206 Butts Cory L. 4 http://orcid.org/0000-0001-7555-2166 Scarneo-Miller Samantha E. 5 1 grid.430643.6 0000 0004 0582 7955 Division of Sports Medicine, United States Olympic & Paralympic Committee, 1 Olympic Plaza, Colorado Springs, CO 80909 USA 2 United States Coalition for the Prevention of Illness and Injury in Sport, Colorado Springs, CO USA 3 grid.266860.c 0000 0001 0671 255X Hydration, Environment, and Thermal Stress Lab, Department of Kinesiology, University of North Carolina at Greensboro, Greensboro, NC USA 4 grid.268072.9 0000 0001 2224 125X Department of Exercise and Nutrition Sciences, Weber State University, Ogden, UT USA 5 grid.268154.c 0000 0001 2156 6140 Division of Athletic Training, School of Medicine, West Virginia University, Morgantown, WV USA 30 11 2022 120 6 3 2022 16 11 2022 © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany 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 This study determined fluid intake and physical activity behaviors among college students during the COVID-19 pandemic. Methods College students (n = 1014; females, 75.6%) completed an online survey during the Spring 2020 academic semester following the initial global response to the COVID-19 pandemic. Academic standing, habitation situation, and University/College responses to COVID-19 were collected. Participants completed the Godin Leisure-Time Exercise Questionnaire and a 15-item Beverage Questionnaire (BEVQ-15) to determine physical activity level and fluid intake behaviors, respectively. Results Females (1920 ± 960 mL) consumed significantly less fluid than males (2400 ± 1270 mL, p < 0.001). Living off-campus (p < 0.01) and living with a spouse/partner (p < 0.01) was associated with increased consumption of alcoholic beverages. 88.7% of participants reported being at least moderately active; however, Black/African American and Asian participants were more likely to be less active than their Caucasian/White counterparts (p < 0.05). Participants reporting no change in habitation in response to COVID-19 had a higher fluid intake (p = 0.002); however, the plain water consumption remained consistent (p = 0.116). While there was no effect of habitation or suspension of classes on physical activity levels (p > 0.05), greater self-reported physical activity was associated with greater fluid intake (std. β = 0.091, p = 0.003). Conclusions Fluid intake among college students during the initial response to the COVID-19 pandemic approximated current daily fluid intake recommendations. Associations between COVID-19-related disruptions (i.e., suspension of classes and changes in habitation) and increased alcohol intake are concerning and may suggest the need for the development of targeted strategies and programming to attenuate the execution of negative health-related behaviors in college students. Supplementary Information The online version contains supplementary material available at 10.1007/s00394-022-03058-9. Keywords Water Hydration Alcohol Exercise Emerging adults ==== Body pmcIntroduction Existing evidence [1–13] examining the impact of COVID-19 on health-related behaviors have reported that the associated lockdown measures have had a significant impact on behaviors, such as dietary intake and physical activity. Specifically, individuals tended to consume greater quantities of food or report increased eating frequency [3, 5, 6, 12, 13], consumption of greater quantities of alcohol [5, 12, 13], and increased consumption of either sweet foods or sugar-sweetened beverages [2–4]. These unhealthy behaviors were associated with those who also reported increased levels of stress, anxiety, or psychological or emotional disturbances [6, 10, 14, 15], whereas some individuals reported improvements in health-related behaviors during this time [1, 7, 10]. Health-related behaviors, such as engagement in physical activity and total fluid intake (i.e., hydration status), and their association with human health and performance outcomes have been extensively studied within the scientific literature. These data overwhelmingly support the associations of physical inactivity and low daily fluid intake, specifically total water intake, on increased risks of obesity [16–19], cardiovascular disease [20–22], and diabetes [20, 23–25]. Of particular interest regarding the impact of COVID-19 on health-related behaviors are emerging adults (18–29 years) enrolled in college/university, herein referred to as college students. In this stage of life, most college students are living away from home and are developing independent health-related behaviors for the first time. Evidence shows that college students have a decline in physical activity engagement and associated increase in sedentary behaviors compared to their previous behaviors [26, 27]. Furthermore, poor dietary intake (e.g., increased consumption of foods that are high in fat, sodium, and sugar, decreased consumption of fruits and vegetables, and evidence suggesting that college students are underhydrated) contribute to the increased risk profile for long-term health outcomes [26, 28–32]. While many universities/colleges have begun to integrate initiatives to improve student health and wellness, a college student’s environmental context (e.g., food, physical, and social environment) may impact the development and maintenance of health-related behaviors [26, 28, 33–36]. However, with the changes Colleges and Universities had to make in response to COVID-19, these initiatives were likely unavailable to assist college students in improving health-related behaviors during the latter-part of the Spring 2020 academic semester. Physical activity [37–39] and hydration status [39–41] has been shown to positively impact immune function, potentially being useful in the prevention and mitigation of adverse disease outcomes, such as COVID-19. In particular, exercise-induced immunomodulation may improve the innate and adaptive immune responses to the COVID-19 virus [42]. It is also hypothesized that underhydration may increase the risk of COVID-19 infection via increased angiotensin-converting enzyme 2 receptors in the lung, resulting in increased capillary permeability in the lungs that promotes an environment that reduces the capacity for fluid transport out of the lung [43]. Although prior evidence has examined changes in health-related behaviors in college students throughout the college/university experience, no known literature that has examined the impact of a global pandemic (i.e., COVID-19) on levels of physical activity and fluid intake in this population. Furthermore, with the disruptions of in-person classes, and other associated university services (on-campus living, food services, health and wellness programs and recreation facility access) it is prudent that inquiry into these behaviors is ascertained to inform improved policy. Therefore, the primary aim of this study was to characterize fluid intake and physical activity behaviors in college students during the initial response to the COVID-19 pandemic. As a secondary aim, this study also sought to determine the impact of disruptions in normal living (i.e., suspension of in-person classes and/or a change in living situation [habitation]) on fluid intake and physical activity behaviors. Specifically, we included total fluid intake, exercise participation, and alcohol consumption as outcome variables. Based on the reported and predicted impacts of the COVID-19 pandemic on global mental health [44–46], we predicted that the COVID-19-specific disruptions in a college student’s class structure and habitation status would impact health behaviors, such as fluid intake (including types of fluids consumed) and physical activity. Materials and methods For this study, college students enrolled in college/university in the United States during the Spring 2020 academic semester were recruited to complete an online survey (Qualtrics, Provo, UT, USA) investigating their daily fluid intake and physical activity behaviors during the initial response (March–May 2020) to the COVID-19 global pandemic. University email listservs, social media posts, and convenience sampling were used to recruit eligible participants. Given the inability to decipher the total number of individuals who received the link or saw the social media posts, we are unable to calculate a valid response rate. However, the initial dissemination of the survey using email listservs at the primary author’s institution was sent to 8000 students that were enrolled in classes for the Spring 2020 academic semester. Consent was implied by participants clicking on the “I agree to participate” button at the bottom of the first information page and completing the study. This study was ruled exempt from review by the Institutional Review Board at the University of North Carolina at Greensboro (#20–0442). Survey The survey (Supplemental File) had three main parts: (1) participant characteristics and demographics, (2) fluid intake behaviors and (3) physical activity behaviors. Participant characteristics and demographics. Participants indicated their sex, race and ethnicity, academic classification (i.e., Freshman, Sophomore, Junior, Senior, Graduate Student), habitation during the 2019–2020 academic year, change in habitation as a result of COVID-19, whether or not their university suspended in-person classes and the state in which the participant’s college or university was located. Habitual Fluid Intake. To assess fluid intake behaviors participants completed a validated 15-item Beverage Questionnaire (BEVQ-15) [47]. This questionnaire, used to determine one’s typical beverage intake behaviors during the previous 30 days (one month), was not altered for this study and has shown sufficient construct validity and reproducibility to capture participant fluid intake against a 4-day dietary recall and the BEVQ-2 survey [47]. Participants were instructed to indicate the types of beverages they consumed on average during the previous month. Beverage types included water, beer, wine, hard liquor, coffee, energy drinks, milk, low fat milk, nut milk, fruit juice, sweetened juice beverages, soft drinks, diet soft drinks, sweet tea, and any additional beverages not included in the survey (“other”). For each beverage type, participants selected an option for how often (“never or less than 1 per week”, “1 time per week”, “2–3 times per week”, “4–6 times per week”, “1 time per day”, “2 times per day”, and “3 + times per day”) and how much each time (“less than 6 fl oz [3/4 cup]”, “8 fl oz [1 cup]”, “12 fl oz [1.5 cups]”, “16 fl oz [2 cups]”, “20 fl oz [2.5 cups]”, “ > 20 fl oz”). Fluid intake volume for each beverage type and total fluid intake were calculated based on methods previously established [47]. Physical Activity. Participants were instructed to self-report their levels of physical activity during the 7 days prior to completing the survey. To quantify amount of physical activity, the Godin Leisure-Time Exercise Questionnaire was used [48, 49]. This self-reported physical activity questionnaire provides a valid and reliable assessment of individuals who are classified as ‘active’, ‘moderately active’, and ‘insufficiently active’ [50] and is based on current physical activity recommendations by the American College of Sports Medicine [51]. Participants are asked to report the frequency (times per week) of strenuous, moderate, and mild/light exercise performed for a minimum of 15 consecutive minutes during the previous 7-day period. The frequency was multiplied by a metabolic equivalent (MET; strenuous = 9, moderate = 5, and mild/light = 3) with these arbitrary units summed to give a classification of activity. A score ≥24 was interpreted as active, 14–23 as moderately active, and < 14 as insufficiently active/sedentary. This questionnaire demonstrates sufficient reliability and concurrent validity [48]. Statistical analyses All data analyses were conducted using R statistical software [52]. For the primary aim, robust multiple regression models were run using the “rlm” function in R to assess whether academic classification, habitation, sex, and race/ethnicity were predictors of fluid intake (total and beverage type volume) using iterated re-weighted least squares. This analysis followed this equation for volume of each type of beverage consumed (Y):1 Y=β1Academic Classification+β2Habitation+β3Sex+β4Race+β5Ethnicity+∫ A multinomial logistic regression was run with age, academic classification, habitation, sex, and race/ethnicity as predictors of self-reported physical activity level category (“active”, “moderately active”, or “insufficiently active”):2 lnPPA=activePPA=insufficiently active=β10Academic Classification+β11Habitation+β12Sex+β13Race+β14Ethnicity+∫1 3 lnPPA=moderately activePPA=insufficiently active=β20Academic Classification+β21Habitation+β22Sex+β23Race+β24Ethnicity+∫2 For the second aim, path analysis models were created using the “Lavaan” package in R to explore the impact of the suspension of university classes, as mediated by a change in habitation and moderated by academic classification, on total fluid intake and physical activity classification. The model was adjusted for race/ethnicity and sex (Fig. 1). Significance was set at a-priori at p < 0.05 for all analyses. The indirect effect (moderated mediation) for each of these models was calculated as the product of the regression coefficient for the effect of a suspension of in-person classes on a change in habitation moderated by different levels of academic classification, with the effect of a change in habitation on the outcome variable of interest. The total effect of each model was calculated as the addition of the indirect effect and the direct effect (effect of suspension of in-person classes on the outcome variable without mediation or moderation). Additional models following this format were created with specific beverages of interest as outcome variables, specifically total fluid, alcohol, and water. Figure 1 displays the model with Total Fluid (mL) as the outcome variable. The same model was used with plain water, beer, wine, and hard liquor as outcome variables (see supplementary figures). A separate model was run with physical activity classification as a moderator of Total Fluid intake (Fig. 2). Such path analysis models have the advantage of examining both the direct influence of one independent variable on the dependent variable, as well as chains of influence, whereby the independent variable may or may not influence an intermediate variable which affects the outcome of interest (indirect effect)[53]. Inclusion of the moderator academic classification allowed us to assess whether the strength of the relationship between a suspension of in-person university classes and total fluid intake and physical activity levels varied across academic school year.Fig. 1 Moderated Mediation model examining the influence of a Suspension of In Person Classes (SIP) as Mediated by Change in Habitation and moderated by Academic Classification (AC) on Total fluid Intake. Solid lines represent estimated parameter for the predicted standardized change in the exogenous variable for every one unit change in the predictor variable, when holding other variables constant. For example, a 1 standard deviation change in sex (i.e., females) predicted a – 0.189 standard deviation decrease in total fluid intake, when controlling for SIP, Race/Ethnicity, Habitation, and the moderating effect of SIP and AC. SIP coded as “No suspension of in-person classes” = 0, “Suspension of in-person classes” = 1. Race/Ethnicity coded as “White/Caucasian” = 0, “Non-white” = 1. Sex coded as “Male” = 0. “Female” = 1. Habitation coded as “No living change” = 0, “Living Change” = 1. AC used as a continuous moderator with ‘0’ representing the lowest academic classification “Freshman”. Dashed lines represent covariance between variables. Dashed circular arrows represent variance of exogenous variables. Solid circular arrows represent error variance for endogenous variables Fig. 2 Moderated Mediation examining the influence of a Suspension of In Person Classes (SIP) as Mediated by Change in Habitation and moderated by Physical Activity (PA) on Total Fluid Intake, covaried by Sex and Race/Ethnicity. Solid lines represent estimated parameter for the predicted standardized change in the exogenous variable for every one unit change in the predictor variable, when holding other variables constant.. SIP coded as “No suspension of in-person classes” = 0, “Suspension of in-person classes” = 1. Race/Ethnicity coded as “White/Caucasian” = 0, “Non-white” = 1. Sex coded as “Male” = 0. “Female” = 1. Habitation coded as “No living change” = 0, “Living Change” = 1. PA was a continuous moderating variable of physical activity score based on the Godin questionnaire. Dashed lines represent covariance between variables. Dashed circular arrows represent variance of exogenous variables. Solid circular arrows represent error variance for endogenous variables Results A total of 1156 participants began the survey, from which 141 were excluded based on limited completion (< 80% of questions completed), leaving 1014 for analysis (Table 1). Self-reported fluid intake across each race and sex are displayed in Table 2. The study population primarily identified as being female [n = 766 (75.6%) vs n = 241 (23.8%) male]. Participants self-reported primarily consuming water (n = , 47.9%), followed by sweet beverages (16.9%), caffeinated beverages (12.1%), milk or milk alternatives (6.5%), and alcohol (5.6%). The level of self-reported physical activity is depicted in Table 3. Among the responses included for analysis, 75.7% of participants were classified as “active”, 13% as “moderately active”, and 11% as “insufficiently active/sedentary”.Table 1 Respondent demographic information Overall (N = 1014) Sex Female 766 (75.5%) Male 241 (23.8%) Prefer not to answer 6 (0.6%) Missing 1 (0.1%) Race American Indian 26 (2.6%) Asian 54 (5.3%) Black or Black/African American 168 (16.6%) Native Hawaiian 1 (0.1%) Other 32 (3.2%) White or Caucasian 724 (71.4%) Missing 9 (0.9%) Ethnicity Hispanic or Latino or Spanish Origin 98 (9.7%) Not Hispanic or Latino or Spanish Origin 907 (89.4%) Missing 9 (0.9%) Academic Classification Freshman 59 (5.8%) Graduate student 289 (28.5%) Junior 264 (26.0%) Senior 245 (24.2%) Sophomore 139 (13.7%) Missing 18 (1.8%) Self-reported physical activity level Sedentary 113 (11.1%) Moderately active 133 (13.1%) Active 768 (75.7%) Habitation (Start of 2019–2020 Academic Year) Off campus 388 (38.3%) On campus 253 (25.0%) Other 13 (1.3%) Parent/guardian 182 (17.9%) Spouse/partner 177 (17.5%) Missing 1 (0.1%) Change in habitation No 587 (57.9%) Yes 426 (42.0%) Missing 1 (0.1%) New habitation location (As a result of COVID-19) Off campus 15 (1.5%) On campus 7 (0.7%) Parent/guardian 358 (35.3%) Spouse/partner 21 (2.1%) Missing 613 (60.5%) In-person Classes Suspended? No 12 (1.2%) Yes 1000 (98.6%) Missing 2 (0.2%) Table 2 Self-reported daily fluid intake by race/ethnicity and sex Black/African American American Indian Asian White/Caucasian Native Hawaiian or Pacific Islander Other Hispanic or Latino or Spanish v Overall Female (N = 137) Male (N = 31) Female (N = 20) Male (N = 6) Female (N = 35) Male (N = 18) Female (N = 544) Male (N = 175) Female (N = 1) Female (N = 25) Male (N = 7) Female (N = 82) Male (N = 16) Female (N = 766) Male (N = 241) Total Fluid (mL) Mean (SD) 2000 (1040) 2330 (1640) 1790 (1020) 3370 (2120) 1810 (833) 1730 (790) 1840 (955) 2220 (1770) 1530 (NA) 1580 (908) 2170 (795) 1840 (955) 2220 (1770) 1920 (960) 2400 (1270) Median [Min, Max] 1850 [467, 7100] 1800 [532, 7140] 1520 [526, 4330] 2690 [1750, 7540] 1820 [390, 3790] 1700 [532, 3620] 1690 [272, 5750] 1460 [544, 6600] 1530 [1530, 1530] 1320 [272, 4470] 2310 [1080, 3220] 1690 [272, 5750] 1460 [544, 6600] 1770 [118, 7980] 2170 [532, 7540] Water (mL) Mean (SD) 1010 (550) 1050 (561) 793 (576) 942 (694) 1010 (448) 1060 (493) 936 (568) 736 (504) 1060 (NA) 880 (567) 936 (532) 936 (568) 736 (504) 963 (549) 1010 (519) Median [Min, Max] 1060 [0, 1860] 1060 [189, 1860] 710 [94.6, 1770] 1060 [142, 1770] 1060 [331, 1860] 1060 [94.6, 1860] 828 [0, 1860] 665 [142, 1860] 1060 [1060, 1060] 946 [124, 1860] 710 [166, 1770] 828 [0, 1860] 665 [142, 1860] 946 [0, 1860] 1060 [47.3, 1860] Beer (mL) Mean (SD) 23.7 (82.3) 37.5 (99.9) 49.7 (78.7) 291 (418) 16.6 (37.9) 5.59 (13.8) 47.8 (168) 236 (307) 142 (NA) 14.7 (39.4) 79.4 (98.7) 47.8 (168) 236 (307) 39.6 (95.2) 116 (260) Median [Min, Max] 0 [0, 591] 0 [0, 473] 0 [0, 248] 124 [0, 1060] 0 [0, 142] 0 [0, 47.3] 0 [0, 1240] 237 [0, 1180] 142 [142, 142] 0 [0, 189] 23.7 [0, 248] 0 [0, 1240] 237 [0, 1180] 0 [0, 1240] 0 [0, 1860] Missing 16 (19.5%) 3 (18.8%) 16 (19.5%) 3 (18.8%) Wine (mL) Mean (SD) 46.0 (127) 24.5 (72.8) 42.5 (58.9) 240 (403) 15.9 (35.4) 18.9 (51.3) 36.6 (79.2) 39.8 (97.7) 59.2 (NA) 28.6 (52.1) 27.2 (39.1) 36.6 (79.2) 39.8 (97.7) 41.4 (100) 34.4 (101) Median [Min, Max] 0 [0, 1180] 0 [0, 355] 0 [0, 189] 62.1 [0, 946] 0 [0, 142] 0 [0, 189] 0 [0, 473] 0 [0, 331] 59.2 [59.2, 59.2] 0 [0, 189] 17.7 [0, 94.6] 0 [0, 473] 0 [0, 331] 0 [0, 1180] 0 [0, 946] Missing 16 (11.7%) 3 (9.7%) 4 (20.0%) 1 (16.7%) 9 (25.7%) 2 (11.1%) 13 (15.9%) 5 (31.2%) 0 (0%) 6 (24.0%) 2 (28.6%) 13 (15.9%) 5 (31.2%) 114 (14.9%) 35 (14.5%) Hard Liquor (mL) Mean (SD) 23.3 (78.1) 30.1 (69.8) 21.7 (49.7) 126 (286) 12.8 (44.7) 12.8 (30.5) 11.5 (34.9) 131 (357) 0 (NA) 8.28 (29.6) 23.7 (45.4) 11.5 (34.9) 131 (357) 16.4 (51.9) 34.9 (108) Median [Min, Max] 0 [0, 710] 0 [0, 248] 0 [0, 189] 11.8 [0, 710] 0 [0, 237] 0 [0, 94.6] 0 [0, 177] 8.87 [0, 1240] 0 [0, 0] 0 [0, 142] 0 [0, 124] 0 [0, 177] 8.87 [0, 1240] 0 [0, 710] 0 [0, 1240] Missing 18 (22.0%) 4 (25.0%) 18 (22.0%) 4 (25.0%) Coffee (mL) Mean (SD) 124 (252) 55.5 (175) 139 (229) 447 (717) 151 (217) 125 (226) 158 (195) 203 (247) 166 (NA) 115 (149) 279 (311) 158 (195) 203 (247) 179 (272) 278 (409) Median [Min, Max] 17.7 [0, 1770] 0 [0, 946] 17.7 [0, 710] 172 [0, 1860] 47.3 [0, 946] 23.7 [0, 710] 94.6 [0, 946] 94.6 [0, 710] 166 [166, 166] 23.7 [0, 532] 237 [0, 710] 94.6 [0, 946] 94.6 [0, 710] 71.0 [0, 1770] 94.6 [0, 1860] Missing 11 (13.4%) 3 (18.8%) 11 (13.4%) 3 (18.8%) Energy Drinks (mL) Mean (SD) 26.5 (119) 118 (255) 3.55 (10.9) 29.6 (39.4) 7.44 (29.1) 29.9 (111) 34.1 (93.2) 183 (314) 0 (NA) 50.6 (124) 65.9 (86.4) 34.1 (93.2) 183 (314) 29.3 (100) 86.2 (213) Median [Min, Max] 0 [0, 1060] 0 [0, 1240] 0 [0, 35.5] 11.8 [0, 94.6] 0 [0, 166] 0 [0, 473] 0 [0, 473] 41.4 [0, 1180] 0 [0, 0] 0 [0, 473] 35.5 [0, 237] 0 [0, 473] 41.4 [0, 1180] 0 [0, 1240] 0 [0, 1860] Missing 18 (22.0%) 2 (12.5%) 18 (22.0%) 2 (12.5%) Milk (mL) Mean (SD) 66.1 (178) 109 (256) 143 (208) 280 (232) 108 (155) 69.7 (90.4) 80.1 (152) 97.8 (100) 0 (NA) 96.5 (235) 35.5 (54.2) 80.1 (152) 97.8 (100) 71.9 (161) 141 (278) Median [Min, Max] 0 [0, 1770] 23.7 [0, 1420] 0 [0, 710] 243 [0, 710] 23.7 [0, 532] 11.8 [0, 237] 23.7 [0, 1180] 94.6 [0, 331] 0 [0, 0] 17.7 [0, 1180] 0 [0, 142] 23.7 [0, 1180] 94.6 [0, 331] 0 [0, 1770] 23.7 [0, 1860] Missing 2 (2.4%) 3 (18.8%) 2 (2.4%) 3 (18.8%) Low Fat Milk (mL) Mean (SD) 24.2 (96.2) 20.8 (36.1) 16.6 (36.1) 15.8 (38.6) 72.7 (132) 33.8 (77.1) 0 [0, 1060] 0 [0, 710] 35.5 (NA) 9.23 (26.3) 55.8 (100) 26.5 (57.4) 113 (339) 29.9 (87.3) 36.9 (116) Median [Min, Max] 0 [0, 946] 0 [0, 94.6] 0 [0, 124] 0 [0, 94.6] 0 [0, 473] 0 [0, 237] 43.8 (131) 40.6 (72.5) 35.5 [35.5, 35.5] 0 [0, 94.6] 0 [0, 248] 0 [0, 237] 0 [0, 1180] 0 [0, 1060] 0 [0, 1180] Missing 9 (11.0%) 4 (25.0%) Nut Milk (mL) Mean (SD) 61.0 (159) 34.9 (63.5) 54.7 (79.6) 56.2 (98.1) 89.4 (221) 0.986 (4.18) 0 [0, 1420] 0 [0, 473] 0 (NA) 53.5 (109) 67.6 (94.1) 75.4 (132) 70.1 (160) 47.9 (119) 33.3 (85.2) Median [Min, Max] 0 [0, 1180] 0 [0, 248] 11.8 [0, 237] 8.87 [0, 248] 0 [0, 1180] 0 [0, 17.7] 0 [0, 0] 0 [0, 473] 0 [0, 237] 0 [0, 710] 0 [0, 591] 0 [0, 1180] 0 [0, 710] Missing 10 (12.2%) 2 (12.5%) Fruit Juice (mL) Mean (SD) 210 (355) 227 (417) 91.7 (219) 84.8 (133) 55.3 (89.6) 69.0 (167) 37.6 (150) 41.6 (109) 23.7 (NA) 109 (207) 66.8 (83.0) 108 (226) 69.3 (96.6) 77.6 (204) 69.4 (179) Median [Min, Max] 94.6 [0, 1860] 47.3 [0, 1420] 0 [0, 946] 17.7 [0, 331] 0 [0, 355] 11.8 [0, 710] 0 [0, 1770] 0 [0, 710] 23.7 [23.7, 23.7] 0 [0, 710] 17.7 [0, 189] 17.7 [0, 1420] 11.8 [0, 331] 0 [0, 1860] 0 [0, 1420] Missing 3 (3.7%) 2 (12.5%) Sweetened Juice Beverages (mL) Mean (SD) 109 (262) 177 (356) 31.9 (81.4) 79.9 (164) 39.4 (106) 57.8 (118) 101 (221) 165 (306) 0 (NA) 57.5 (147) 62.5 (70.6) 80.7 (250) 65.1 (105) 51.1 (174) 62.5 (169) Median [Min, Max] 17.7 [0, 1860] 23.7 [0, 1420] 0 [0, 355] 20.7 [0, 414] 0 [0, 532] 0 [0, 473] 0 [0, 1860] 35.5 [0, 1770] 0 [0, 0] 0 [0, 710] 59.2 [0, 189] 0 [0, 1770] 23.7 [0, 355] 0 [0, 1860] 0 [0, 1420] Missing 8 (9.8%) 4 (25.0%) Soft Drinks (mL) Mean (SD) 77.2 (222) 90.4 (234) 102 (133) 302 (401) 14.5 (33.4) 79.9 (121) 81.1 (242) 64.6 (191) 35.5 (NA) 36.0 (82.6) 55.8 (122) 80.0 (201) 217 (372) 90.1 (212) 148 (285) Median [Min, Max] 0 [0, 1860] 0 [0, 946] 41.4 [0, 473] 142 [0, 1060] 0 [0, 94.6] 29.6 [0, 473] 0 [0, 1860] 0 [0, 1420] 35.5 [35.5, 35.5] 0 [0, 355] 0 [0, 331] 0 [0, 1180] 53.2 [0, 1180] 0 [0, 1860] 23.7 [0, 1770] Missing 10 (12.2%) 2 (12.5%) Diet Drinks (mL) Mean (SD) 32.3 (125) 132 (453) 23.7 (80.9) 246 (424) 28.4 (100) 44.7 (117) 1.62 (5.93) 2.22 (7.03) 0 (NA) 16.6 (56.1) 123 (175) 77.2 (314) 101 (153) 65.9 (214) 77.9 (243) Median [Min, Max] 0 [0, 946] 0 [0, 1860] 0 [0, 355] 29.6 [0, 1060] 0 [0, 473] 0 [0, 473] 0 [0, 63.0] 0 [0, 60.0] 0 [0, 0] 0 [0, 248] 35.5 [0, 473] 0 [0, 1860] 8.87 [0, 473] 0 [0, 1860] 0 [0, 1860] Missing 102 (18.8%) 27 (15.4%) 15 (18.3%) 4 (25.0%) Sweet Tea (mL) 2.22 (6.73) 2.30 (7.53) 1.53 (2.59) 3.06 (6.51) 1.19 (3.27) 0.686 (2.12) NA (NA) 0.981 (2.74) 1.52 (2.78) 1.69 (5.83) 2.16 (6.76) Mean (SD) 0 [0, 63.0] 0 [0, 40.0] 0 [0, 8.00] 0 [0, 14.7] 0 [0, 16.0] 0 [0, 8.00] NA [NA, NA] 0 [0, 11.2] 0 [0, 6.40] 1.17 (5.35) 2.20 (3.87) 0 [0, 63.0] 0 [0, 60.0] Median [Min, Max] 17 (12.4%) 2 (6.5%) 0 (0%) 1 (16.7%) 8 (22.9%) 4 (22.2%) 1 (100%) 4 (16.0%) 2 (28.6%) 0 [0, 42.0] 0 [0, 11.2] 132 (17.2%) 38 (15.8%) Missing 15 (18.3%) 4 (25.0%) Table 3 Level of self-reported physical activity by race/ethnicity and sex Insufficiently active (N = 113) Moderately active (N = 133) Active (N = 768) Overall (N = 1014) Sex Female 90 (79.6%) 105 (78.9%) 571 (74.3%) 766 (75.5%) Male 22 (19.5%) 25 (18.8%) 194 (25.3%) 241 (23.8%) Prefer not to answer 1 (0.9%) 2 (1.5%) 3 (0.4%) 6 (0.6%) Missing 0 (0%) 1 (0.8%) 0 (0%) 1 (0.1%) Race/Ethnicity Not Hispanic or Latino or Spanish Origin 96 (85.0%) 120 (90.2%) 691 (90.0%) 907 (89.4%) American Indian 2 (1.8%) 4 (3.0%) 20 (2.6%) 26 (2.6%) Asian 9 (8.0%) 8 (6.0%) 37 (4.8%) 54 (5.3%) Black/African American 30 (26.5%) 26 (19.5%) 112 (14.6%) 168 (16.6%) Other 6 (5.3%) 5 (3.8%) 21 (2.7%) 32 (3.2%) White or Caucasian 65 (57.5%) 86 (64.7%) 573 (74.6%) 724 (71.4%) Native Hawaiian 0 (0%) 0 (0%) 1 (0.1%) 1 (0.1%) Missing 1 (0.9%) 4 (3.0%) 4 (0.5%) 9 (0.9%) Hispanic or Latino or Spanish Origin 15 (13.3%) 12 (9.0%) 71 (9.2%) 98 (9.7%) Missing 2 (1.8%) 1 (0.8%) 6 (0.8%) 9 (0.9%) Fluid intake behaviors Multiple regression models incorporating sex, race/ethnicity, academic classification, habitation, and exercise participation as predictors of self-reported total fluid intake and a type of fluids consumed are displayed in Table 4. Females self-reported consuming significantly less daily total fluid, as well as less beer, hard liquor, coffee, energy drinks, milk, sweetened juice beverages, and soft drinks compared to males (p < 0.001), but consumed more nut milk (p < 0.05). Compared to participants identifying as White/Caucasian, there were no significant differences in total fluid intake between races (p > 0.05). Black/African American participants consumed less beer, coffee, soft drinks, and diet soft drinks but more nut milk, fruit juice, and sweetened juice beverages (p < 0.05). Asian participants consumed less beer, energy drinks and soft drinks (p < 0.05). American Indian participants consumed more wine, milk, nut milk, soft drinks, sweet tea, and other beverages (p < 0.05). Native Hawaiians consumed less beer and low-fat milk (p < 0.05). Participants of Hispanic/Latino/Spanish origin consumed more nut milk and fruit juice but less soft drinks and sweet tea (p < 0.05).Table 4 Self-reported fluid intake predicted by sex, race/ethnicity, academic classification, and habitation. Sex coded as 0 = “Male”, 1 = “Female” Dependent variable: Total Fluid (mL) Water (mL) Beer (mL) Wine (mL) Hard Liquor (mL) Coffee (mL) Energy Drinks (mL) Milk (mL) Low Fat Milk (mL) Nut Milk (mL) Fruit Juice (mL) Sweetened Juice Beverages (mL) Soft Drinks (mL) Diet Soft Drinks (mL) Sweet Tea (mL) Other (mL) Sex – 326.019* (68.180) – 1.124 (44.795) – 11.523* (2.211) 2.523 (1.838) – 3.549*** (0.745) – 35.028* (15.229) – 0.128* (0.017) – 27.330* (5.645) 0.003 (0.003) 2.350* (1.034) – 3.841 (2.495) – 2.667* (1.123) – 24.553* (5.970) – 0.004 (0.009) – 0.005 (0.005) 0.0001 (0.0002) Race/Ethnicity African American 67.417 (79.850) 92.116 (52.462) – 10.260* (2.589) 1.281 (2.125) – 0.973 (0.873) – 72.349*** (17.835) – 0.007 (0.019) – 5.596 (6.611) – 0.002 (0.003) 2.664* (1.211) 61.859* (2.922) 10.020* (1.315) – 17.296* (6.992) – 0.022* (0.010) 0.008 (0.006) – 0.00003 (0.0002) Asian – 188.830 (127.551) 93.897 (83.802) – 11.346** (4.136) – 3.368 (3.564) – 2.022 (1.395) – 53.599 (28.489) – 0.068* (0.031) 7.810 (10.561) 0.009 (0.006) – 0.823 (1.934) 6.320 (4.668) 2.667 (2.100) – 30.694** (11.169) – 0.026 (0.016) – 0.007 (0.009) 0.00003 (0.0003) American Indian – 27.268 (175.966) – 156.867 (115.611) 5.272 (5.705) 10.766* (4.845) 1.292 (1.924) – 24.415 (39.303) – 0.034 (0.043) 70.432* (14.570) – 0.003 (0.008) 7.168 (2.668) 6.667 (6.439) 3.640 (2.898) 31.872* (15.409) – 0.001 (0.022) 0.026* (0.012) 0.013* (0.0005) Native Hawaiian – 370.999 (880.280) 89.468 (578.352) 119.289* (28.542) 39.122 (21.844) – 5.796 (9.624) – 18.886 (196.617) – 0.037 (0.213) – 27.786 (72.886) 35.473*** (0.038) – 6.433 (13.346) 14.320 (32.212) – 2.662 (14.496) – 7.647 (77.085) – 0.054 (0.111) – 0.022 (0.062) – 0.0005 (0.002) Other Race – 240.203 (180.338) – 8.227 (118.484) – 4.119 (5.847) 1.960 (5.089) – 0.469 (1.972) 5.674 (40.280) 0.009 (0.044) – 15.729 (14.932) – 0.005 (0.008) 1.362 (2.734) 3.868 (6.599) 5.612 (2.970) – 31.349* (15.792) 0.017 (0.023) 0.007 (0.013) 0.0002 (0.0005) Hispanic/Latino/Spanish Origin 9.314 (108.256) – 28.738 (71.125) – 0.500 (3.510) 0.982 (2.933) – 1.697 (1.184) – 28.353 (24.180) 0.015 (0.026) 15.746 (8.964) 0.002 (0.005) 3.780* (1.641) 8.013* (3.961) 0.075 (1.783) – 7.137 (9.480) – 0.028* (0.014) – 0.017* (0.008) – 0.0001 (0.0003) Academic Classification 50.783 (27.482) 38.403* (18.056) 2.837 (0.891) 1.912 (0.734) 0.200 (0.300) 20.247*** (6.138) – 0.013* (0.007) – 2.073 (2.276) – 0.001 (0.001) 0.152 (0.417) – 2.008* (1.006) – 2.175* (0.453) – 7.091 (2.407) 0.005 (0.003) – 0.005* (0.002) 0.00000 (0.0001) Habitation Living off Campus 199.662* (81.531) 85.790 (53.567) 5.877* (2.644) 9.164* (2.193) 2.995* (0.891) 30.279 (18.211) 0.034 (0.020) – 5.631 (6.751) 0.008* (0.004) 1.057 (1.236) 1.908 (2.984) 1.208 (1.343) 4.314 (7.140) – 0.002 (0.010) – 0.002 (0.006) 0.0001 (0.0002) Living with spouse/partner 288.597** (98.965) 31.915 (65.021) 3.359 (3.209) 5.952* (2.651) 1.852 (1.082) 76.232*** (22.105) 0.008 (0.024) – 17.612* (8.194) 0.005 (0.004) 0.802 (1.500) – 0.184 (3.621) – 0.227 (1.630) – 0.782 (8.666) 0.019 (0.012) – 0.003 (0.007) 0.0004 (0.0003) Living with parent/guardian 135.440 (89.491) 89.886 (58.796) – 0.894 (2.902) 0.467 (2.398) – 0.226 (0.978) – 4.971 (19.988) 0.049* (0.022) 5.814 (7.410) 0.005 (0.004) – 1.370 (1.357) 5.358 (3.275) 0.914 (1.474) 12.911 (7.837) 0.005 (0.011) – 0.003 (0.006) – 0.0002 (0.0002) Other living arrangements – 120.392 (254.216) – 179.918 (167.022) – 8.099 (8.243) – 1.319 (6.834) – 2.627 (2.779) 61.022 (56.781) – 0.047 (0.062) 16.866 (21.049) 0.004 (0.011) – 1.547 (3.854) 1.771 (9.303) – 4.001 (4.186) 55.669* (22.261) 0.022 (0.032) – 0.021 (0.018) – 0.0005 (0.001) Total Exercise 3.122* (0.722) 2.422* (0.475) – 0.006 (0.023) – 0.016 (0.019) 0.009 (0.008) 0.232 (0.161) 0.001* (0.0002) 0.202* (0.060) 0.00005 (0.00003) 0.025* (0.011) – 0.007 (0.026) – 0.011 (0.012) – 0.196** (0.063) 0.00003 (0.0001) – 0.0001* (0.0001) 0.00001 (0.00000) Constant 1,745.315* (108.357) 678.799* (71.191) 17.106* (3.513) 1.090 (2.990) 5.334* (1.185) 102.697* (24.202) 0.161* (0.026) 64.182* (8.972) 0.009* (0.005) 1.827 (1.643) 19.478* (3.965) 13.085* (1.784) 96.456* (9.489) 0.039 (0.014) 0.051*** (0.008) 0.0002 (0.0003) Observations 972 972 972 827 972 972 972 972 972 972 972 972 972 972 972 972 Residual Std. Error 882.251 (df = 958) 625.964 (df = 958) 25.168 (df = 958) 20.472 (df = 813) 8.250 (df = 958) 195.806 (df = 958) 0.179 (df = 958) 66.622 (df = 958) 0.037 (df = 958) 10.723 (df = 958) 23.142 (df = 958) 11.451 (df = 958) 72.704 (df = 958) 0.110 (df = 958) 0.061 (df = 958) 0.004 (df = 958) Race variables with 0 = “White/Caucasian” as baseline Habitation with 0 = “On Campus” as baseline Note: p < 0.05, p < 0.01, p < 0.001 When examining associations of fluid intake by academic classification and habitation, those of higher academic classification (i.e., graduate student > senior > junior > sophomore > freshman) had increased water (p < 0.05), beer (p < 0.05), coffee (p < 0.001), and wine consumption (p < 0.001), but less energy drink (p < 0.05), fruit juice (p < 0.05), sweetened juice beverage (p < 0.05), soft drink (p < 0.05) and sweet tea consumption (p < 0.05). Living off campus was associated with greater total fluid intake (p < 0.05), including more beer (p < 0.01), wine (p < 0.001), hard liquor (p < 0.01), and low-fat milk consumption (p < 0.05) compared to living on campus. Living with a spouse/partner was associated with increased total fluid intake (p < 0.01), including more wine (p < 0.01), and coffee consumption (p < 0.01), but less milk consumption (p < 0.05) compared to living on campus. Living with a parent(s)/guardian(s) was associated with increased energy drink consumption (p < 0.05) compared to living on campus. Greater exercise participation predicted greater total fluid intake (p < 0.001), including more water (p < 0.001), energy drink (p < 0.001), milk (p < 0.001), and nut milk (p < 0.05) intake but less soft drink (p < 0.01) and sweet tea (p < 0.05) intake. Physical activity behaviors Multinomial logistic regression models incorporating sex, race, academic classification, and habitation as predictors of self-reported physical activity level are displayed in Table 5. Black/African American participants were less likely to be active than Caucasian/White respondents (p < 0.001). Native Hawaiian participants were less likely to be moderately active (p < 0.001), but more likely to be active (p < 0.001) than Caucasian/White respondents. Asian participants were less likely to be classified as active (p < 0.05) compared to Caucasian/White respondents. Participants living off-campus were more likely to be moderately active (p = 0.05) and active (p < 0.01) compared to those living on campus.Table 5 Self-reported physical activity predicted by sex, race/ethnicity, academic classification, and habitation Dependent variable Moderately active Active Sex 0.071 (0.340) -0.340 (0.263) Race/ethnicity Black/African American – 0.461 (0.325) – 0.908* (0.253) Asian – 0.416 (0.523) – 0.837* (0.405) American Indian 0.321 (0.889) 0.017 (0.761) Native Hawaiian – 5.687* (0.000) 9.633* (0.00000) Other race – 0.245 (0.660) – 0.707 (0.521) Academic classification 0.061 (0.128) – 0.013 (0.101) Habitation Living off Campus 0.759* (0.380) 0.932 (0.308) Living with spouse/partner – 0.222 (0.445) 0.100 (0.336) Living with parent/guardian – 0.066 (0.403) 0.239 (0.302) Other living arrangements 0.492 (0.967) -0.116 (0.836) Constant – 0.132 (0.459) 2.116* (0.354) Akaike Inf. Crit 1,415.428 1,415.428 Sex coded as 0 = “Male”, 1 = “Female”. Race variables with 0 = “White/Caucasian” as baseline. Living arrangement variables with 0 = “On Campus” as baseline Note: p < 0.05, p < 0.01, *p < 0.001 Moderated mediation analyses Moderated mediation analyses were conducted with “suspension of in-person classes” as a predictor of total fluid intake, mediated by whether students experienced a change in habitation, and moderated by academic classification (Table 6). This model included Sex and Race/Ethnicity (where 0 = non-Hispanic White, 1 = non-White) as covariates for the prediction of total fluid intake (Fig. 1). The chi-square test of model fit showed this model was not a perfect fit (chi-square = 8556.186, df = 11, p = 0.000). However, the comparative fit index (CFI) and Tucker–Lewis Index (TLI) both suggested a good model fit (CFI = 1.000, TLI = 1.001). This fit was also supported by an RMSEA of 0.000 (90% CI 0.000–0.065) and SRMR of 0.001, both of which suggest a good model fit. Thus, this model was considered acceptable for further interpretation. Both the total and indirect effects were not statistically significant (indirect effect, p = 0.489; total effect, p = 0.182). However, participants who did not experience a change in habitation reported greater total fluid intake than those who experienced a change in habitation, when controlling for sex, race/ethnicity, and academic standing (p = 0.002). Individuals of higher academic classification (grad student > senior > junior > sophomore > freshman) were less likely to experience a change in habitation due to COVID-19 disruptions (p < 0.001).Table 6 Moderated mediation model results with total fluid consumption as outcome variable. Race/ethnicity coded as “White” = 0, “Non-white” = 1. Sex coded as “Male” = 0. “Female” = 1 Variables β SE z p value ci.lower ci.upper β (std) Total fluid–habitation 207.82 66.12 3.14 0.00 76.92 336.28 0.10 Total fluid–suspension of in person classes – 329.30 243.86 – 1.35 0.18 – 835.99 122.97 – 0.03 Total fluid–sex – 435.23 78.95 – 5.51 0.00 – 595.46 – 288.75 – 0.19 Total fluid–race/ethnicity – 32.60 77.88 – 0.42 0.68 – 182.41 120.31 – 0.01 Habitation–suspension of in person classes 0.07 0.30 0.24 0.81 – 0.14 0.93 0.02 Habitation–academic classification 0.16 0.01 14.44 0.00 0.14 0.18 0.39 Habitation–suspension of in person classes: academic classification 0.08 0.10 0.86 0.39 – 0.18 0.19 0.05 Habitation–sex – 0.08 0.03 – 2.64 0.01 – 0.14 – 0.02 – 0.08 Habitation– race/ethnicity – 0.07 0.03 – 2.08 0.04 – 0.13 – 0.00 – 0.06 Indirect1: = (a1 + a3)b1 32.48 47.33 0.69 0.49 0.07 197.34 0.01 Total: = c1 + (a1a3)b1 – 328.04 245.08 – 1.34 0.18 – 833.53 128.41 – 0.03 Separate moderated mediation analyses were conducted with different categories of alcohol (Beer, Wine, Hard Liquor, see: “Supplementary Figures”) as outcome variables, when controlling for sex and race/ethnicity. The model for Beer (Fig. S1) produced good global fit (chi-square = 1.070, p = 0.301; CFI = 1.000; TLI = 0.999; RMSEA = 0.008 (90% CI 0.000–0.086); SRMR = 0.003). This model revealed a non-significant indirect (p = 0.666) and total effect (p = 0.916) of suspension of in-person classes and change in habitation on beer consumption. However, there was a significant increase in Beer consumption among those who did not experience a change in habitation (β = 0.104, p < 0.001) compared to those who did experience a habitation change (Table 7).Table 7 Moderated mediation model results with Beer consumption as outcome variable. Race/ethnicity coded as “White” = 0, “Non-white” = 1 Variables β SE z p value ci.lower ci.upper β (std) Beer –habitation 35.90 11.27 3.18 0.00 15.78 60.34 0.10 Beer–suspension of in person classes – 5.96 56.76 – 0.11 0.92 – 77.47 145.44 – 0.00 Beer–sex – 70.89 16.92 – 4.19 0.00 – 108.52 – 41.49 – 0.20 Beer–race/ethnicity – 33.09 10.49 – 3.15 0.00 – 53.16 – 12.01 – 0.09 Beer–academic classification 5.51 4.96 1.11 0.27 -4.78 14.65 0.04 Habitation –suspension of in person classes – 0.01 0.31 – 0.02 0.98 – 0.20 0.88 – 0.00 Habitation –academic classification 0.15 0.01 12.11 0.00 0.13 0.17 0.37 Habitation –suspension of in person classes: academic classification 0.11 0.10 1.09 0.27 – 0.16 0.21 0.07 Habitation –sex – 0.08 0.03 – 2.31 0.02 – 0.14 – 0.01 – 0.07 Indirect1: = (a1 + a3)b1 3.56 8.25 0.43 0.67 – 1.28 31.88 0.01 Total: = c1 + (a1a3)b1 – 5.99 56.78 – 0.11 0.92 – 78.25 145.61 – 0.00 Sex coded as “Male” = 0. “Female” = 1 Global fit indices for the model with wine consumption as an outcome variable suggest a good model fit (chi-square = 575.320, p = 0.000; CFI = 1.000; TLI = 1.019; RMSEA = 0.000 (90% CI 0.000–0.034); SRMR = 0.000). There was a significant total effect of a suspension of in-person classes on wine consumption (Table 8, Fig. S2) when controlling for sex and race/ethnicity, driven by the significant reduction in wine consumption among participants who did not experience a suspension of in person classes (direct effect) (total effect, p = 0.035; direct effect, p – 0.033). However, there was no significant indirect effect (mediation and moderation effect) of the suspension of in-person classes on wine consumption (p = 0.727).Table 8 Moderated mediation model results with wine consumption as outcome variable. Race/ethnicity coded as “White” = 0, “Non-white” = 1 Variables β SE z p value ci.lower ci.upper β (std) Wine –habitation 11.37 8.15 1.39 0.16 – 5.30 26.64 0.06 Wine –suspension of in person classes – 26.56 12.48 – 2.13 0.03 – 49.43 – 0.48 – 0.03 Wine –sex 9.01 7.42 1.22 0.22 – 6.51 22.67 0.04 Wine –race/ethnicity 7.25 8.84 0.82 0.41 – 8.11 27.05 0.03 Wine –academic classification 8.79 3.02 2.91 0.00 3.51 15.27 0.11 Habitation –suspension of in person classes – 0.01 0.31 – 0.02 0.98 – 0.19 0.89 – 0.00 Habitation –academic classification 0.15 0.01 12.16 0.00 0.13 0.18 0.37 Habitation –suspension of in person classes: academic classification 0.11 0.10 1.10 0.27 – 0.16 0.21 0.07 Habitation –sex – 0.08 0.03 – 2.53 0.01 – 0.15 – 0.02 – 0.08 Indirect1: = (a1 + a3)b1 1.15 3.30 0.35 0.73 – 0.57 14.59 0.00 Total:c1 + (a1a3)b1 – 26.56 12.58 – 2.11 0.03 – 49.88 – 0.18 – 0.03 Sex coded as “Male” = 0. “Female” = 1 With hard liquor consumption as an outcome variable, global fit indices suggested a good model fit (chi-square = 0.072, p = 0.789; CFI = 1.000; TLI = 1.029; RMSEA = 0.000 (90% CI 0.000–0.055); SRMR = 0.001). Students who did not experience a suspension of in person classes self-reported consuming less Hard Liquor (total effect, p = 0.005; direct effect, p = 0.005), driven by the significant increase in consumption among those who did not experience a living change (p = 0.010) (Table 9, Fig. S3).Table 9 Moderated mediation model results with hard liquor consumption as outcome variable. Race/ethnicity coded as “White” = 0, “Non-white” = 1 Variables β SE z p value ci.lower ci.upper β (std) Hard liquor –habitation 14.05 5.44 2.58 0.01 4.30 25.81 0.09 Hard liquor –suspension of in person classes – 23.85 8.45 – 2.82 0.00 – 39.40 – 5.43 – 0.04 Hard liquor –sex – 13.89 7.92 – 1.75 0.08 – 30.95 0.10 – 0.08 Hard liquor –race/ethnicity 7.66 6.68 1.15 0.25 – 3.93 22.45 0.04 Hard liquor –academic classification 0.35 2.21 0.16 0.87 – 4.26 4.41 0.01 Habitation –suspension of in person classes – 0.01 0.31 – 0.02 0.98 – 0.19 0.90 – 0.00 Habitation –academic classification 0.15 0.01 11.94 0.00 0.13 0.18 0.37 Habitation –suspension of in person classes: academic classification 0.11 0.10 1.07 0.28 – 0.17 0.21 0.07 Habitation –sex – 0.07 0.03 – 2.13 0.03 – 0.14 – 0.01 – 0.07 Indirect1:(a1 + a3)b1 1.38 3.40 0.41 0.68 – 0.50 13.38 0.01 Total: = c1 + (a1a3)b1 – 23.87 8.56 – 2.79 0.01 – 39.60 – 4.88 – 0.04 Sex coded as “Male” = 0. “Female” = 1 By contrast, the model showed no significant total effect (p = 0.116), indirect effect (p = 0.747), or direct effect (p = 0.116) of a suspension of in person classes on plain Water consumption, when controlling for sex, race/ethnicity, and academic standing (Table 10, Fig. S4). Global fit indices for this model suggested a good fit (chi-square = – 0.746, p = 0.746; CFI = 1.000; TLI = 1.002; RMSEA = 0.000 (90% CI 0.000–0.059); SRMR = 0.001). The moderated mediation model for self-reported physical activity classification produced a good global fit (chi-square = 0.025, p = 0.873; CFI = 1.000; TLI = 1.053; RMSEA = 0.000 (90% CI 0.000–0.044); SRMR = 0.000) (Table 11), there was no significant indirect effect (p = 0.934), direct effect (p = 0.129) or total effect (p = 0.131) of a suspension of in person classes on activity level. However, non-White participants reported lower activity levels (p < 0.001), when controlling for a suspension of in person classes, habitation, sex, and academic classification. Females also reported lower physical activity levels (p = 0.032), independent of habitation, race/ethnicity, or a suspension of in person classes.Table 10 Moderated mediation model results with water consumption as outcome variable. Race/ethnicity coded as “White” = 0, “Non-white” = 1 Variables β SE z p value ci.lower ci.upper β (std) Water –habitation 27.54 38.86 0.71 0.48 – 49.02 102.85 0.02 Water –suspension of in person classes – 238.36 151.64 – 1.57 0.12 – 535.40 60.32 -0.05 Water –sex – 29.14 35.83 – 0.81 0.42 – 99.68 39.59 – 0.02 Water –race/ethnicity 27.36 38.48 0.71 0.48 – 47.38 103.49 0.02 Water –academic classification 29.11 15.59 1.87 0.06 – 1.77 59.87 0.06 Habitation –suspension of in person classes 0.07 0.30 0.24 0.81 – 0.14 0.93 0.02 Habitation –academic classification 0.16 0.01 14.32 0.00 0.14 0.18 0.39 Habitation –suspension of in person classes: academic classification 0.08 0.10 0.86 0.39 – 0.18 0.19 0.05 Habitation –sex – 0.08 0.03 – 2.64 0.01 – 0.14 – 0.02 – 0.08 Indirect1: = (a1 + a3)b1 4.30 13.33 0.32 0.75 – 8.51 55.33 0.00 Total: = c1 + (a1a3)b1 – 238.19 151.70 – 1.57 0.12 – 535.35 60.46 – 0.05 Sex coded as “Male” = 0. “Female” = 1 Table 11 Moderated mediation model results with physical activity classification as outcome variable. Race/ethnicity coded as “White” = 0, “Non-white” = 1 Variables β SE z p value ci.lower ci.upper β (std) Physical activity –habitation – 0.02 0.05 – 0.43 0.67 -0.11 0.07 – 0.02 Physical activity –suspension of in person classes 0.26 0.17 1.52 0.13 – 0.14 0.50 0.04 Physical activity –sex – 0.10 0.05 – 2.14 0.03 – 0.19 – 0.01 – 0.07 Physical activity –race/ethnicity – 0.19 0.05 – 3.70 0.00 – 0.30 – 0.09 – 0.13 Physical activity –academic classification 0.02 0.02 0.85 0.39 -0.02 0.05 0.03 Habitation –suspension of in person classes 0.07 0.30 0.24 0.81 -0.15 0.92 0.02 Habitation –academic classification 0.16 0.01 14.20 0.00 0.14 0.18 0.39 Habitation –suspension of in person classes: academic classification 0.08 0.10 0.86 0.39 – 0.18 0.19 0.05 Habitation –sex – 0.08 0.03 – 2.66 0.01 – 0.14 – 0.02 – 0.08 Indirect1: = a1b1 – 0.00 0.02 – 0.08 0.93 – 0.06 0.02 – 0.00 Total: = c1 + a1b1 0.26 0.17 1.51 0.13 – 0.14 0.50 0.04 Sex coded as “Male” = 0. “Female” = 1 To explore whether the effects of change in habitation on total fluid intake were influenced by indirectly by changes in exercise habits, a moderated mediation model was conducted with the suspension of in-person classes as a predictor of Total Fluid intake, moderated by Physical Activity level (where ‘0’ = “Sedentary”, ‘1’ = “moderately active”, ‘2’ = “active”), mediated by change in habitation (‘0’ = “No living change”, ‘1’ = “Living change”), and covaried by Sex, Race/Ethnicity (Table 12, Fig. S5). Global fit indices for this model suggested a good fit (chi-square = 0.002, p = 0.961; CFI = 1.000; TLI = 1.001; RMSEA = 0.000 (90% CI = 0.000–0.000); SRMR = 0.000). This model produced a significant indirect effect of a suspension of in-person classes on total fluid intake, when passing through a change in habitation, when controlling for sex, race/ethnicity, and physical activity level (std. β = 0.011, p = 0.006). However, the total effect (including the direct effect of suspension of in-person classes on total fluid intake) was not significant (p = 0.271), suggesting these effects were primarily due to a change in habitation, where a change in living situation led to increased total fluid intake (std. β = 0.10, p < 0.001). Greater physical activity classification was associated with increased fluid consumption (std. β = 0.091, p = 0.003), when controlling for sex and race/ethnicity.Table 12 Moderated mediation model results with total fluid as outcome variable when moderated by physical activity. Race/ethnicity coded as “White” = 0, “Non-white” = 1 Variables β SE z p value ci.lower ci.upper β (std) Total fluid –habitation 212.69 65.15 3.26 0.00 87.26 341.97 0.10 Total fluid –suspension of in person classes – 366.71 237.70 – 1.54 0.12 – 853.10 69.46 – 0.04 Total fluid –sex – 406.28 78.66 – 5.16 0.00 – 567.91 – 259.97 – 0.18 Total fluid –race/ethnicity – 11.51 75.90 – 0.15 0.88 – 158.02 144.02 – 0.00 Total fluid –physical activity 142.10 47.79 2.97 0.00 46.96 233.75 0.09 Habitation –suspension of in person classes 0.48 0.09 5.27 0.00 0.30 0.68 0.11 Habitation –physical activity – 0.01 0.02 – 0.28 0.78 – 0.05 0.04 – 0.01 Habitation –suspension of in person classes: physical activity – 0.06 0.07 – 0.99 0.32 – 0.21 0.05 – 0.04 Habitation –sex – 0.10 0.03 – 3.19 0.00 – 0.17 – 0.04 -0.10 Sex coded as “Male” = 0. “Female” = 1 Discussion This cross-sectional study sought to characterize fluid intake and physical activity habits of a diverse sample of college students during the initial response (Spring 2020 academic semester) to the COVID-19 pandemic. Furthermore, we explored the impact of disruptions in normal living on fluid intake and physical activity behaviors. To our knowledge, this is the first study that has focused specifically on fluid intake and physical activity behaviors in this population and how mid-academic semester changes in living situations in response to college/university-driven policies to combat the spread of COVID-19 impacted these health-related behaviors. Given the evidence supporting the health benefits associated with physical activity and maintaining an appropriate level of hydration through adequate water intake, understanding the impact of a global pandemic on these health-related behaviors in a college population that may be developing their independent health behaviors is vital for the development of evidence driven policy. The ensuing discussion compares the findings of our study to existing literature with a specific focus on fluid intake, physical activity, and alcohol consumption. Total fluid intake Overall, fluid intake among the college students (females, 1920 ± 960 mL; males, 2400 ± 1270 mL) in this study approximated the adequate intake recommendations established by the European Food Safety Authority (EFSA, 2.0 L/d and 2.5 L/d for females and males, respectively) [54]; however, only 39.6% of females and 38.6% of males in our study self-reported meeting or exceeding these recommendations. Furthermore, when comparing self-reported fluid intake to the adequate intake recommendations by the Institutes of Medicine (IOM, 2.7 L/d and 3.7 L/d for females and males, respectively) [55], only 16.3% of females and 13.3% of males in the current study met this threshold. Previous data depict the prevalence of underhydration among population-based data; however, these data support that females are more likely to consume adequate volumes of fluid than males [18, 56]. Our data show an approximately equal percentage of males and females meeting or exceeding daily fluid intake recommendations, albeit greater than 60–85% of participants in our study failed to meet daily fluid intake requirements as established by EFSA and IOM, respectively. While we are unable to determine the reasons for the low prevalence of both males and females in meeting daily fluid intake recommendations among our participants compared to previously published data sets, we speculate that this may be one or combination of: (1) the impact of the societal restrictions (i.e., closure of university campuses, stay-at-home orders, mask requirements, etc.) due to COVID-19 may have influenced fluid intake differently among males and females given the timing (May 2020) of the initial dissemination of the survey, (2) since recruitment of participants was isolated to only those that were currently enrolled in college/university during the Spring 2020 academic semester, access and availability of fluids may have influenced one’s ability to meet daily fluid intake recommendations, and (3) differences in our sample compared to previous literature based on race/ethnicity and the known differences in fluid intake behaviors across populations. We also cannot discount the potential influence of our sample’s pre-pandemic fluid intake behaviors in that our sample may have been habitually under consuming fluids. However, our previous work disputes the low prevalence of college students meeting fluid intake recommendations [57, 58]. In addition to females exhibiting a lower total fluid intake than males, our study also found that females consumed less total fluid, beer, hard liquor, coffee, energy drinks, milk, and soft drinks than their male counterparts. While the reduced total fluid intake in females compared to males could be explained by differences in body composition and fluid needs between males and females, we are unable to provide a thorough explanation on the reduced intake of the specific beverage types discussed above. Our findings also showed that higher academic standing was associated with greater total fluid consumption, specifically coffee, in agreement with a previous survey of caffeine habits among college students [59]. Furthermore, our data suggest a greater total fluid intake among participants who maintained the same habitation during COVID-19. Upon subsequent analysis, this increased fluid intake may be attributed to greater alcohol intake among those who stayed in their current living situation (see below). Greater physical activity participation predicted increased total fluid consumption, independent of whether participants experienced a change in living situation, suggesting most participants attempt to compensate for increased fluid losses associated with physical activity. Physical activity Participants generally met the physical activity guidelines based on their self-reported level of physical activity, with 76% classified as “active” and 13.1% classified as “moderately active”. Black/African American, Asian, and Native Hawaiian participants were less likely to report being active compared to Caucasian peers. This is consistent with previous literature identifying lower physical activity among Black/African American female college students [60]. Those who lived off campus tended to report being more “moderately active” and “active” compared to those living on campus, which may be the result of the closure of university recreation facilities during this time. A similar study examining physical activity behaviors pre- and post-pandemic found students who were initially the most active tended to be disproportionately impacted (i.e., decreased physical activity the most in those who were most physical active) by the COVID-19 pandemic in terms of physical activity participation [61]. Perhaps students who lived on campus were accustomed to using their campus recreation facilities and were less likely to find alternative solutions in the event of such facility closures. Interestingly, one study observed an increase in both physical activity and sedentary time among college students during the pandemic [62]. However, we did not capture sedentary behaviors in the present study. By contrast, some studies have shown a decrease in physical activity behaviors, particularly in countries with stricter confinement guidelines meant to reduce the spread of COVID-19[63]. Other research examining COVID-19 and physical activity has noted substantial changes in resistance training routines with individuals reporting resistance training being less enjoyable despite similar or lower engagement in resistance training routines [64]. Adequate levels of physical activity may promote resilience against COVID-19 symptoms through exercise-induced immunomodulation [65], thus, strategies to maintain activity levels in the event of such disruptions in the future is important both for acute disease prevention, as well as the well-established impact of physical activity on chronic disease risk. Alcohol consumption Disruptions imposed by the COVID-19 pandemic (i.e., suspension of classes and changes in habitation) were associated with increased alcohol consumption. In particular, our findings show that beer intake was higher among participants who did not report a change in habitation. However, students of higher academic standing also tended to consume more fluid, which may have impacted these results, since higher academic students were more likely to be of legal age to consume alcohol and typically tend to live in off-campus housing, which may not have required a change of living compared to students who may have lived on campus. In the moderated mediation models, students who experienced a suspension of in-person classes reported consuming more hard liquor and wine and those who did not experience a change in living tended to consume more beer. The joint effects (total effect) of suspension of in-person university classes and a change in living situation predicted lower wine consumption and lower hard liquor consumption in those who did not experience such disruptions. However, these joint effects were not observed for beer consumption. The transition to online learning, particularly mid-semester as occurred for many during the COVID-19 pandemic, comes with its own set of psychological stressors [66]. It is plausible that changes in the learning environment, combined with anxiety regarding the state of the pandemic, may have contributed to this consumption, which has previously been identified as a growing area of concern for the public even prior to the pandemic [67]. Similarly, Robinson et al. found 36% of UK adults reported consuming either “a little more” (17%), “more” (12%), or “a lot more” (7%), alcohol after lockdown [5]. Another study in Poland found an increase in the frequency of alcohol consumption among adults before and during lockdown [13]. Of note, mean alcohol consumption was relatively modest for both males and females overall (male = 181 ± 379 mL, female = 91.2 ± 171), well below Dietary Guideline recommendations of 2 standard drinks per day for males, and 1 standard drink per day for females [68]. This was expected, given the wide age range of participants, but under-reporting was also a possibility among this population. The maximum reported alcohol consumption in the present study was 2720 mL for males and 1240 mL for females, well beyond current daily intake recommendations. From our study it is unclear whether individuals near the higher end of the alcohol consumption range drank this much before the start of the COVID-19 pandemic. However, one study [69] examined reported alcohol intake in university students the week before compared to the week after school closure, finding a significant increase in consumption coinciding with symptoms of depression and anxiety. In the same study, those with greater perceived social support tended to consume less alcohol. Unfortunately, our survey did not capture whether participants lived alone, which could then influence alcohol consumption through a reduction in perceived social support for these adverse psychological outcomes. In addition, given the cross-sectional nature of the survey, we cannot say for certain whether the higher total fluid and alcohol intake among those who did not experience a change in habitation was more attributed to this factor or that they are of older academic standing and thus able to purchase alcohol themselves. Strengths and limitations There were several strengths of this study. The cross-sectional design of this study allowed us to capture a sample population representative of the racial and ethnic breakdown within the United States. Given the known racial and ethnic differences in both fluid intake [32, 70] and physical activity [71, 72], the representative sample in this study permits improved generalizability of the results. The timing of this study (May 2020) also allowed us to gather fluid intake and physical activity behaviors following the societal restrictions put in place due to COVID-19. Since the initial responses and restrictions to COVID-19 varied across the United States from a timing perspective, ascertaining these behaviors at the conclusion of the Spring 2020 Academic Semester guarantee that the self-reported fluid intake and physical activity behaviors occurred during the implementation of the restrictions. This study was not without limitations, however. As a cross-sectional survey design, we are not able to fully capture the impact of COVID-19 and associated disruptions on fluid intake and physical activity behaviors and we cannot be certain that the observed behaviors were present before the COVID-19 pandemic. This study also did not consider the effects mask-wearing may have on fluid intake. Future studies should identify individuals whose workplace or school required/did not require masks throughout the pandemic and whether this influenced fluid consumption. Climate, more importantly, students relocating to geographic locations that were different from their institution of enrollment was not captured. Given that some existing literature [73–75] suggests that ambient temperatures may influence daily fluid intake, students participating in the current study that relocated to geographically different domiciles from their University/College may have altered their fluid intake prior to the investigator’s assessment of these behaviors. Participant personal income level, nor household income level, was not assessed in the present study. Individuals with lower personal or household income may have reduced access to certain beverage categories. While personal income of college students may fall within a defined threshold of “low income,” household income (i.e., income from their parents/guardians) and the financial support provided directly to the participants may vary considerably. This may be reflective in that 42% of students participating in this study reported that they relocated during the pandemic. Further study should explore fluid intake habits among this population with regard for both individual income levels and any parental support received during this timeframe. Conclusions Reported fluid intake among college students during the initial response to the COVID-19 pandemic approximated current daily fluid intake recommendations on average; however, only 40% of participants (both male and female) met or exceeded fluid intake recommendation guidelines established by the European Food Safety Authority. Significant associations between COVID-19-related disruptions and increased alcohol intake are concerning and suggest the need for the development of targeted strategies and programming to attenuate the execution of negative health-related behaviors in emerging adults enrolled in university. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file1 (TIFF 1660 KB) Fig. S1 Moderated Mediation examining the influence of a Suspension of In Person Classes (SIP) as Mediated by Change in Habitation and moderated by Academic classification (AC) on Beer consumption, covaried by Sex and Race/Ethnicity. Dashed lines represent covariance between variables. Solid lines represent estimated parameter for the predicted standardized change in the exogenous variable for every one unit change in the predictor variable, when holding other variables constant. Dashed circular arrows represent variance of exogenous variables. Solid circular arrows represent error variance for endogenous variables Supplementary file2 (TIFF 1660 KB) Fig. S2 Moderated Mediation examining the influence of a Suspension of In Person Classes (SIP) as Mediated by Change in Habitation and moderated by Academic classification (AC) on Wine consumption, covaried by Sex and Race/Ethnicity. Solid lines represent estimated parameter for the predicted standardized change in the exogenous variable for every one unit change in the predictor variable, when holding other variables constant. Dashed lines represent covariance between variables. Dashed circular arrows represent variance of exogenous variables. Solid circular arrows represent error variance for endogenous variables Supplementary file3 (TIFF 1660 KB) Fig. S3 Moderated Mediation examining the influence of a Suspension of In Person Classes (SIP) as Mediated by Change in Habitation and moderated by Academic classification (AC) on Hard Liquor consumption, covaried by Sex and Race/Ethnicity. Solid lines represent estimated parameter for the predicted standardized change in the exogenous variable for every one unit change in the predictor variable, when holding other variables constant. Dashed lines represent covariance between variables. Dashed circular arrows represent variance of exogenous variables. Solid circular arrows represent error variance for endogenous variables Supplementary file4 (TIFF 1660 KB) Fig. S4 Moderated Mediation examining the influence of a Suspension of In Person Classes (SIP) as Mediated by Living Change and moderated by Academic classification (AC) on Water consumption, covaried by Sex and Race/Ethnicity. Solid lines represent estimated parameter for the predicted standardized change in the exogenous variable for every one unit change in the predictor variable, when holding other variables constant. Dashed lines represent covariance between variables. Dashed circular arrows represent variance of exogenous variables. Solid circular arrows represent error variance for endogenous variables Supplementary file5 (TIFF 1660 KB) Fig. S5 Moderated Mediation examining the influence of a Suspension of In Person Classes (SIP) as Mediated by Living Change and moderated by Academic classification (AC) on self-reported physical activity level, covaried by Sex and Race/Ethnicity. Physical activity coded as ‘0’ = “Sedentary”, ‘1’= “moderately active”, ‘2’ = “active”. Solid lines represent estimated parameter for the predicted standardized change in the exogenous variable for every one unit change in the predictor variable, when holding other variables constant. Dashed lines represent covariance between variables. Dashed circular arrows represent variance of exogenous variables. Solid circular arrows represent error variance for endogenous variables Supplementary file6 (PDF 399 KB) Survey tool that was used for data collection. Acknowledgements The authors would like to thank all participants who took the time to complete this survey. There are no funding sources to report for supporting this project. Data availability Data is available upon reasonable request. Declarations Conflict of interest None. Ethical statement This work is the authors own and not that of the United States Olympic & Paralympic Committee, or any of its members or affiliates. WMA has received industry-funded grants from QCK LLC, Statim Technologies LLC, and Techguard LLC. He has also received consulting fees/travel costs from Clif Bar & Company, Gatorade, Samsung, BSX Athletics, and Danone Research Nutricia. ==== Refs References 1. Rodríguez-Pérez C Molina-Montes E Verardo V Changes in dietary behaviours during the COVID-19 outbreak confinement in the Spanish COVIDiet study Nutrients 2020 12 1730 10.3390/nu12061730 32531892 2. Ruiz-Roso MB de Carvalho PP Mantilla-Escalante DC Covid-19 confinement and changes of adolescent’s dietary trends in Italy, Spain, Chile, Colombia and Brazil Nutrients 2020 12 1807 10.3390/nu12061807 32560550 3. Scarmozzino F Visioli F Covid-19 and the subsequent lockdown modified dietary habits of almost half the population in an Italian sample Foods 2020 9 675 10.3390/foods9050675 32466106 4. Pietrobelli A Pecoraro L Ferruzzi A Effects of COVID-19 lockdown on lifestyle behaviors in children with obesity living in Verona, Italy: a longitudinal study Obesity 2020 28 1382 1385 10.1002/oby.22861 32352652 5. Robinson E Boyland E Chisholm A Obesity, eating behavior and physical activity during COVID-19 lockdown: a study of UK adults Appetite 2021 156 104853 10.1016/j.appet.2020.104853 33038479 6. Ingram J Maciejewski G Hand CJ Changes in diet, sleep, and physical activity are associated with differences in negative mood during COVID-19 lockdown Front Psychol 2020 11 588604 10.3389/fpsyg.2020.588604 32982903 7. Pérez-Rodrigo C Gianzo Citores M Hervás Bárbara G Patterns of change in dietary habits and physical activity during lockdown in Spain due to the COVID-19 pandemic Nutrients 2021 10.3390/nu13020300 33494314 8. Castañeda-Babarro A Arbillaga-Etxarri A Gutiérrez-Santamaría B Coca A Physical activity change during COVID-19 confinement Int J Environ Res Public Health 2020 10.3390/ijerph17186878 32967091 9. Martínez-de-Quel Ó Suárez-Iglesias D López-Flores M Pérez CA Physical activity, dietary habits and sleep quality before and during COVID-19 lockdown: a longitudinal study Appetite 2021 158 105019 10.1016/j.appet.2020.105019 33161046 10. Cheval B Sivaramakrishnan H Maltagliati S Relationships between changes in self-reported physical activity, sedentary behaviour and health during the coronavirus (COVID-19) pandemic in France and Switzerland J Sports Sci 2020 10.1080/02640414.2020.1841396 33118469 11. Qin F Song Y Nassis GP Physical activity, screen time, and emotional well-being during the 2019 novel coronavirus outbreak in China Int J Environ Res Public Health 2020 10.3390/ijerph17145170 33121139 12. Robinson E Gillespie S Jones A Weight-related lifestyle behaviours and the COVID-19 crisis: An online survey study of UK adults during social lockdown Obes Sci Pract 2020 6 735 740 10.1002/osp4.442 33354349 13. Błaszczyk-Bębenek E Jagielski P Bolesławska I Nutrition Behaviors in polish adults before and during COVID-19 lockdown Nutrients 2020 10.3390/nu12103084 33050404 14. Brooks SK Webster RK Smith LE The psychological impact of quarantine and how to reduce it: rapid review of the evidence The Lancet 2020 395 912 920 10.1016/S0140-6736(20)30460-8 15. Leng G Adan RAH Belot M The determinants of food choice Proc Nutr Soc 2017 76 316 327 10.1017/S002966511600286X 27903310 16. Bel-Serrat S Ojeda-Rodríguez A Heinen MM Clustering of multiple energy balance-related behaviors in school children and its association with overweight and obesity-WHO European childhood obesity surveillance initiative (COSI 2015–2017) Nutrients 2019 10.3390/nu11030511 30818859 17. Catenacci VA Ostendorf DM Pan Z The impact of timing of exercise initiation on weight loss: an 18-month randomized clinical trial Obes Silver Spring Md 2019 10.1002/oby.22624 18. Chang T Ravi N Plegue MA Inadequate hydration, BMI, and obesity among US adults: NHANES 2009–2012 Ann Fam Med 2016 14 320 324 10.1370/afm.1951 27401419 19. Stookey JD Kavouras SΑ Suh H Lang F Underhydration is associated with obesity, chronic diseases, and death within 3 to 6 years in the U.S. population aged 51–70 years Nutrients 2020 12 905 10.3390/nu12040905 32224908 20. Chudyk A Petrella RJ Effects of exercise on cardiovascular risk factors in type 2 diabetes: a meta-analysis Diabetes Care 2011 34 1228 1237 10.2337/dc10-1881 21525503 21. D’hooge R, Hellinckx T, Van Laethem C, Influence of combined aerobic and resistance training on metabolic control, cardiovascular fitness and quality of life in adolescents with type 1 diabetes: a randomized controlled trial Clin Rehabil 2011 25 349 359 10.1177/0269215510386254 21112904 22. Daniels D Angiotensin II (de)sensitization: Fluid intake studies with implications for cardiovascular control Physiol Behav 2016 162 141 146 10.1016/j.physbeh.2016.01.020 26801390 23. Melander O Vasopressin, from regulator to disease predictor for diabetes and cardiometabolic risk Ann Nutr Metab 2016 68 Suppl 2 24 28 10.1159/000446201 27299865 24. Melander O Vasopressin: novel roles for a new hormone - Emerging therapies in cardiometabolic and renal diseases J Intern Med 2017 282 281 283 10.1111/joim.12656 28929632 25. Carroll HA Davis MG Papadaki A Higher plain water intake is associated with lower type 2 diabetes risk: a cross-sectional study in humans Nutr Res 2015 35 865 872 10.1016/j.nutres.2015.06.015 26255759 26. Small M Bailey-Davis L Morgan N Maggs J Changes in eating and physical activity behaviors across seven semesters of college: living on or off campus matters Health Educ Behav Off Publ Soc Public Health Educ 2013 40 435 441 10.1177/1090198112467801 27. Deforche B Van Dyck D Deliens T De Bourdeaudhuij I Changes in weight, physical activity, sedentary behaviour and dietary intake during the transition to higher education: a prospective study Int J Behav Nutr Phys Act 2015 12 16 10.1186/s12966-015-0173-9 25881147 28. Brunt A Rhee Y Zhong L Differences in dietary patterns among college students according to body mass index J Am Coll Health 2008 56 629 634 10.3200/JACH.56.6.629-634 18477517 29. Shin N Hyun W Lee H A study on dietary habits, health related lifestyle, blood cadmium and lead levels of college students Nutr Res Pract 2012 6 340 348 10.4162/nrp.2012.6.4.340 22977689 30. Nelson MC Lust K Story M Ehlinger E Alcohol use, eating patterns, and weight behaviors in a University population Am J Health Behav 2009 33 227 237 10.5993/AJHB.33.3.1 19063644 31. Demory-Luce D Morales M Nicklas T Changes in food group consumption patterns from childhood to young adulthood: the Bogalusa Heart Study J Am Diet Assoc 2004 104 1684 1691 10.1016/j.jada.2004.07.026 15499355 32. Adams WM Hevel DJ Maher JP McGuirt JT Racial and sex differences in 24 hour urinary hydration markers among male and female emerging adults: a pilot study Nutrients 2020 12 1068 10.3390/nu12041068 32290616 33. Liao Y Intille SS Dunton GF Using ecological momentary assessment to understand where and with whom adults’ physical and sedentary activity occur Int J Behav Med 2015 22 51 61 10.1007/s12529-014-9400-z 24639067 34. Dunton GF Whalen CK Jamner LD Floro JN Mapping the social and physical contexts of physical activity across adolescence using ecological momentary assessment Ann Behav Med 2007 34 144 153 10.1007/BF02872669 17927553 35. Higgs S Ruddock H Meiselman HL Social Influences on Eating Handbook of Eating and Drinking: Interdisciplinary Perspectives 2020 Cham Springer International Publishing 277 291 36. Robinson E Blissett J Higgs S Social influences on eating: implications for nutritional interventions Nutr Res Rev 2013 26 166 176 10.1017/S0954422413000127 24103526 37. Fragala MS Kraemer WJ Denegar CR Neuroendocrine-immune interactions and responses to exercise Sports Med Auckl NZ 2011 41 621 639 10.2165/11590430-000000000-00000 38. Freidenreich DJ Volek JS Immune responses to resistance exercise Exerc Immunol Rev 2012 18 8 41 22876721 39. Gleeson M Nieman DC Pedersen BK Exercise, nutrition and immune function J Sports Sci 2004 22 115 125 10.1080/0264041031000140590 14971437 40. Bishop NC Blannin AK Walsh NP Nutritional aspects of immunosuppression in athletes Sports Med Auckl NZ 1999 28 151 176 10.2165/00007256-199928030-00002 41. Fortes MB Diment BC Di Felice U Walsh NP Dehydration decreases saliva antimicrobial proteins important for mucosal immunity Appl Physiol Nutr Metab Physiol Appliquée Nutr Métabolisme 2012 37 850 859 10.1139/h2012-054 42. Leandro CG Ferreira E Silva WT Lima-Silva AE Covid-19 and Exercise-Induced Immunomodulation NeuroImmunoModulation 2020 10.1159/000508951 32506067 43. Stookey JD Allu PKR Chabas D Hypotheses about sub-optimal hydration in the weeks before coronavirus disease (COVID-19) as a risk factor for dying from COVID-19 Med Hypotheses 2020 144 110237 10.1016/j.mehy.2020.110237 33254543 44. Torales J O’Higgins M Castaldelli-Maia JM Ventriglio A The outbreak of COVID-19 coronavirus and its impact on global mental health Int J Soc Psychiatry 2020 66 317 320 10.1177/0020764020915212 32233719 45. Pfefferbaum B North CS Mental health and the Covid-19 pandemic N Engl J Med 2020 383 510 512 10.1056/NEJMp2008017 32283003 46. Copeland WE McGinnis E Bai Y Impact of COVID-19 pandemic on college student mental health and wellness J Am Acad Child Adolesc Psychiatry 2021 60 134 141.e2 10.1016/j.jaac.2020.08.466 33091568 47. Fausnacht AG Myers EA Hess EL Update of the BEVQ-15, a beverage intake questionnaire for habitual beverage intake for adults: determining comparative validity and reproducibility J Hum Nutr Diet Off J Br Diet Assoc 2020 10.1111/jhn.12749 48. Godin G Shephard RJ A simple method to assess exercise behavior in the community Can J Appl Sport Sci J Can Sci Appl Au Sport 1985 10 141 146 49. Godin G (2011) The Godin-Shephard Leisure-time physical activity questionnaire. Health Fit J Can 4:18–22. 10.14288/hfjc.v4i1.82 50. Amireault S Godin G The Godin-Shephard Leisure-time physical activity questionnaire: validity evidence supporting its use for classifying healthy adults into active and insufficiently active categories Percept Mot Skills 2015 120 604 622 10.2466/03.27.PMS.120v19x7 25799030 51. Garber CE Blissmer B Deschenes MR American College of Sports Medicine position stand. Quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: guidance for prescribing exercise Med Sci Sports Exerc 2011 43 1334 1359 10.1249/MSS.0b013e318213fefb 21694556 52. R Core Team (2013) R: A Language and Environment for Statistical Computing 53. Streiner DL Finding our way: an introduction to path analysis Can J Psychiatry 2005 50 115 122 10.1177/070674370505000207 15807228 54. Agostoni CV Bresson JL Fairweather-tait S Scientific opinion on dietary reference values for water EFSA J 2010 10.2903/j.efsa.2010.1459 55. IOM Dietary reference intakes for water, potassium, sodium, chloride, and sulfate. Washington, DC 56. Guelinckx I Ferreira-Pêgo C Moreno LA Intake of water and different beverages in adults across 13 countries Eur J Nutr 2015 54 Suppl 2 45 55 10.1007/s00394-015-0952-8 26072214 57. Sims JNL Holland JJ Anderson T Adams WM Daily fluid intake behaviors and associated health effects among Australian and United States populations Front Sports Act Living 2022 10.3389/fspor.2022.898720 35755610 58. Zaplatosch ME Bechke EE Goldenstein SJ The influence Of menstrual cycle phase on fluid intake and urinary hydration markers: 1726 Med Sci Sports Exerc 2022 54 419 420 10.1249/01.mss.0000880340.18876.f2 59. Bertasi RAO, Humeda Y, Bertasi TGO, et al Caffeine intake and mental health in college students. Cureus 13:e14313. 10.7759/cureus.14313 60. McArthur LH Raedeke TD Race and sex differences in college student physical activity correlates Am J Health Behav 2009 33 80 90 10.5993/ajhb.33.1.8 18844523 61. Barkley JE Lepp A Glickman E The acute effects of the COVID-19 pandemic on physical activity and sedentary behavior in University students and employees Int J Exerc Sci 2020 13 1326 1339 33042377 62. Romero-Blanco C Rodríguez-Almagro J Onieva-Zafra MD Physical activity and sedentary lifestyle in University students: changes during confinement due to the COVID-19 pandemic Int J Environ Res Public Health 2020 17 E6567 10.3390/ijerph17186567 63. Castañeda-Babarro A Arbillaga-Etxarri A Gutiérrez-Santamaría B Coca A Physical activity change during COVID-19 confinement Int J Environ Res Public Health 2020 17 E6878 10.3390/ijerph17186878 64. Steele J Androulakis-Korakakis P Carlson L The impact of coronavirus (COVID-19) related public-health measures on training behaviours of individuals previously participating in resistance training: a cross-sectional survey study Sports Med Auckl NZ 2021 51 1561 1580 10.1007/s40279-021-01438-5 65. Woods JA Hutchinson NT Powers SK The COVID-19 pandemic and physical activity Sports Med Health Sci 2020 2 55 64 10.1016/j.smhs.2020.05.006 34189484 66. Fawaz M Samaha A E-learning: depression, anxiety, and stress symptomatology among Lebanese university students during COVID-19 quarantine Nurs Forum (Auckl) 2021 56 52 57 10.1111/nuf.12521 67. Sugarman DE Greenfield SF Alcohol and COVID-19: How do we respond to this growing public health crisis? J Gen Intern Med 2021 36 214 215 10.1007/s11606-020-06321-z 33105004 68. 2015–2020 Dietary Guidelines \| health.gov. https://health.gov/our-work/food-nutrition/2015-2020-dietary-guidelines/guidelines/. Accessed 12 Jul 2020 69. Lechner WV Laurene KR Patel S Changes in alcohol use as a function of psychological distress and social support following COVID-19 related University closings Addict Behav 2020 110 106527 10.1016/j.addbeh.2020.106527 32679435 70. Kenney EL Long MW Cradock AL Gortmaker SL Prevalence of inadequate hydration among US children and disparities by gender and race/ethnicity: national health and nutrition examination survey, 2009–2012 Am J Public Health 2015 105 e113 118 10.2105/AJPH.2015.302572 26066941 71. Sohn EK Porch T Hill S Thorpe RJ Geography, race/ethnicity, and physical activity among men in the United States Am J Mens Health 2017 11 1019 1027 10.1177/1557988316689498 28147893 72. Watson KB Whitfield G Chen TJ Trends in aerobic and muscle-strengthening physical activity by race/ethnicity across income levels among US adults, 1998–2018 J Phys Act Health 2021 18 S45 S52 10.1123/jpah.2021-0260 34465650 73. Malisova O Bountziouka V Panagiotakos DΒ Evaluation of seasonality on total water intake, water loss and water balance in the general population in Greece J Hum Nutr Diet Off J Br Diet Assoc 2013 26 Suppl 1 90 96 10.1111/jhn.12077 74. Westerterp KR Plasqui G Goris AHC Water loss as a function of energy intake, physical activity and season Br J Nutr 2005 93 199 203 10.1079/bjn20041310 15788113 75. Heller KE Sohn W Burt BA Eklund SA Water consumption in the United States in 1994–96 and implications for water fluoridation policy J Public Health Dent 1999 59 3 11 10.1111/j.1752-7325.1999.tb03228.x 11396041	36449091	PMC9709366	NO-CC CODE	2022-12-01 23:23:04	no	Eur J Nutr. 2022 Nov 30;:1-20	utf-8	Eur J Nutr	2,022	10.1007/s00394-022-03058-9	oa_other
==== Front Educ Inf Technol (Dordr) Educ Inf Technol (Dordr) Education and Information Technologies 1360-2357 1573-7608 Springer US New York 11468 10.1007/s10639-022-11468-9 Article Post hoc identification of student groups: Combining user modeling with cluster analysis http://orcid.org/0000-0002-4367-9629 Balaban Igor [email protected] Filipović Danijel [email protected] http://orcid.org/0000-0002-6061-1896 Zlatović Miran [email protected] grid.4808.4 0000 0001 0657 4636 Faculty of Organization and Informatics, University of Zagreb, Pavlinska 2, 42 000 Varaždin, Croatia 30 11 2022 126 8 6 2022 14 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 aims to discover groups of students enrolled in the emergency remote teaching online course based on the various course-related data collected throughout the first year of COVID-19 pandemic. Research was conducted among 222 students enrolled in the course “Business Informatics” at the Faculty of Organization and Informatics of the University of Zagreb in the academic year 2020/2021. Overlays were used to model students’ success on the various quizzes and exams within the course. The k-means clustering was employed to classify students into groups, based on combination of students’ overlay values, frequency of accessing course lessons and the final grades. Three distinct clusters (i.e., students’ groups) were discovered and explained in the given context. The identified groups of students can be used for future adaptations of the online course design in order to improve the retention and their final grades. Keywords Emergency remote teaching Student activity Overlay model Clustering http://dx.doi.org/10.13039/501100004488 Hrvatska Zaklada za Znanost IP-2020-02-5071 ==== Body pmcIntroduction The COVID-19 pandemic has accelerated the digitalization in higher education institutions (HEI) and has acted as a change driver in teaching and learning practice. Emergency remote teaching (ERT) that arose as a new phenomenon (Hodges et al., 2020) is described as an immediate change of conventional teaching practice through the application of online tools. Teachers extensively relied on the institutional virtual learning platforms and videoconferencing systems which became foundational to the education experience, highlighting the transfer from onsite teaching practice into online (Bond et al., 2021; Lowenthal et al., 2020). More research is needed on the topic since the ERT presents both technological and pedagogical change. Many HEIs did not have time to properly design and adapt their courses for online mode since the shift “… has been an abrupt one due to unprecedented lockdown imposed to manage the COVID-19“ (Muthuprasad et al., 2021). Although the advanced adaptive online education techniques and systems could help to improve student retention rates (Smaili et al., 2021), the hard reality of many institutions which were forced to shift to emergency online teaching is that they do not have the access to or cannot use the adaptive online education platforms. In situations where the curricula were transformed to online form within the traditional non-adaptive paradigm (i.e., basic Learning Management Systems (LMS) usage), it is not possible to continuously adapt the learning process to the individual needs of students. It seems to be a reasonable alternative that educators can use the data from the LMS systems collected during the initial years of crisis to perform post hoc user analysis and identify different groups of students in their courses. Consequently, they can refine course content and structure to offer adaptations in future in order to improve students’ retention rate. The various forms of resource usage analysis and user modeling have been used for more than a decade as a foundation for redesigning online courses (i.e., in Camacho et al., 2009; Mazza & Botturi, 2007). Learning Management Systems (LMS) can thus present a valuable tool to help teachers to analyze student work and activities in online courses. Many resources describe how logs from such systems can be used to research the relation between the time spent online and students’ final grades (Ryabov, 2012), to research the impact of online learning activities on learning outcomes in blended courses (Nguyen, 2017) or to explore the relationship between the time students spent on the course website and the assessment performance (Korkofingas & Macri, 2013). This research aims to classify students enrolled in the online course during ERT period, based on the post hoc analysis of the available data about the students’ activities throughout the entire semester and on their mastery of knowledge domains. To achieve this, the final course grades were correlated with the results of two types of activities: (i) the knowledge displayed on various online tests during the semester (many formative and summative online tests within the course - multiple flash exams and self-assessments, two midterm and the final exam) and (ii) the frequency of reading the provided lessons (i.e. students’ activity). Such identification of groups of students should allow teachers to adapt the structure of teaching and learning activities in the course, aiming to improve the students’ retention rate and their final grades. In this paper, we define the term ‘retention’ as the number of students who continue their studies by re-enrolling from one academic year to the next, i.e. we are referring to the students’ retention rate or academic retention. The term ‘retention’ should therefore not be interpreted as knowledge retention or memory retention. Considerations towards improving academic retention are valid, especially at the beginning of a study programme. Some studies have shown that students at the beginning of their academic career have problems with assessing the workload brought by certain courses, and that failure/not finding the right way at the beginning of their studies can negatively affect the students’ retention rate. According to Otrel-Cass et al. (2009), “… findings suggest that during their first year science students need to be reassured that they are valued, and that their education is taken very seriously by the institution and their lecturers. Student commentary suggests this can be achieved by personalizing lectures, ensuring personal contact with lecturers and monitoring how students are coping with the challenges and stresses that affect workload issues and subsequently their academic progress.”. This finding is relevant for the course which is subject of this research, since our course is one of the foundational courses within the curriculum and it takes place in the first semester of a three-year undergraduate professional study program. Theoretical background According to Kebritchi et al. (2017), it is a great challenge to develop online courses which not only cover the curricular aspect, but also succeed in engaging the students and preparing the instructors for transition to online teaching and learning. Bignoux and Sund (2018) have identified major differences between the online environments and traditional classrooms regarding the student’s satisfaction, motivation and interaction. Other studies have shown that the effectiveness of online classes depends not only on the use of advanced technology, but also on the quality of course content structure, instructors’ preparedness and the quality of instruction (Sun & Chen, 2016; Gilbert, 2015). Using LMS data to analyze students’ activity LMS systems record vast amounts of logs containing various details about students’ activities during course duration (i.e., what activities students have accessed, when, for how long, etc.). Teachers should be, therefore, empowered to collect, analyze and interpret the data collected in the LMS with learning analytics (LA) tools to improve students’ learning (see for example Rapanta et al., 2021). The necessity of LA usage is further highlighted by Ferguson (2012) who noted the importance of using the data about students and their contexts to understand and optimize the learning environments, as well as the learning process. For example, Ryabov (2012) reports that the overall logged time within LMS (i.e., time spent online) had a positive influence on the final grade. Literature review by Nguyen (2017) reports positive correlations between readings of various contents (i.e. pages viewed, read discussions, discussion posts made) and the learning outcomes. Wei et al. (2015) impact analysis of the activity within LMS systems on the academic performance has shown that the results of various online assignments (including online tests/exams), as well as the overall access time (i.e., overall number of logins, number of posts, time spent to read various documents) had significant effect on the learning performance. Approaches to identification of students’ groups As noted in the previous section, one of the important research streams in the analysis of student data in LMS is the user modeling, especially in the context of adaptive education to explain the student groups and thus to better adapt the course learning design (i.e., in Corrin et al., 2017). User modeling is defined as a process of obtaining a user model as “… a source of information, which contains assumptions about those aspects of a user that might be relevant for behaviour of information adaptation.” (Schreck, 2003) In education, such approach includes overlay student models, stereotypes, perturbation models, machine learning techniques, constraint-based models, fuzzy models, Bayesian networks and ontology-based models. Hybrid student models in which researchers combine various modeling techniques have also been recorded. However, not every approach is suited to model every characteristic of a student. Chrysafiadi and Virvou (2013) review reveals that a student’s mastery of knowledge is predominantly modeled with overlays (20% of all research) and stereotypes (14.44% of all research). The overlay model assumes that the student has incomplete but correct knowledge of the domain, i.e., the model of student’s knowledge is a subset of the domain model. The domain model represents the expert-level knowledge of the domain and it is compared with the model of student’s knowledge. The difference between them is believed to originate from the student’s lack of skills and knowledge. Therefore, the main objective of the instruction process is to eliminate these differences as much as possible. Essentially, the domain must be decomposed into groups of smaller, interconnected elements (knowledge topics) and, therefore, the individual student’s overlay model consists of a set of masteries (student’s recorded levels of knowledge) of those elements. Masteries can be expressed as Boolean value (true/false – i.e., student does or does not possess the knowledge about an element) or more refined qualitative (i.e., good/average/scales) or quantitative measures (i.e., probability level that student has the knowledge). (Martins et al., 2008; Nguyen & Do, 2008; Brusilovsky & Millán, 2007; Bontcheva & Wilks, 2005). More modern approaches also include various data mining and classification techniques (i.e., decision trees) and clustering techniques (Francis & Babu, 2019) that enable real-time grouping. Research objectives and methodology This research aims to discover possible groups of students enrolled in an online course1 based on post hoc activity data collected from LMS logs (frequency of accessing course lessons, quiz scores) and students’ final grades. Several studies have shown that the post hoc analysis of LMS data combined with clustering techniques can be used to derive student groups, i.e., by analyzing self-assessment scores (Watson et al., 2017) or by analyzing students’ activities including login records and content reads (Tseng et al., 2016). Research objectives In this research the students’ knowledge and the overlay model are used to calculate students’ masteries post hoc, when all the lectures in a course are finished. Calculation of masteries is based on the students’ scores obtained within a number of course online quizzes (including informal self-assessments and formal midterms/exams). Clustering is used to classify our students into typical groups based on their masteries from overlay model, activity logs data and final exam grades. The research was conducted during an emergency lockdown period, when many pedagogical and technological aspects of teaching were pushed forward by institutions (e.g., asynchronous and synchronous teaching and learning, prescribed LMS and video streaming platforms, etc.). Although the transition to online teaching was mostly ad-hoc for many courses, LMS platforms enabled systematic recording of the activities of the learners. With respect to the research aim and the above-mentioned context, the following main research question has been formulated: What types of student groups can be detected based on calculated students’ masteries post hoc, activity logs and final exam grades? For that purpose, several associated research objectives were set: To calculate students’ masteries post hoc using an overlay model. To classify students into groups based on the overlay model, activity logs and final grades using clustering. To describe these student groups, with implications for improving the course structure and activities. Research process The research process has been divided into three phases: Course preparation. Data collection and preprocessing. Data analysis. In the first phase, the online course was prepared according to the principles of programmed learning in LMS Moodle. The variety of online resources and knowledge assessments were implemented along with conditional activities to create a clear learning path for students. More information about the used programmed learning principles, conditional activities and the course in general is given in the subsequent Section 4. Within the data collection and preprocessing phase, raw data about various activities of students have been gathered from different sources. The dataset consisted of three parts: (1) the overlay model of students’ domain knowledge, (2) students’ frequency of viewing course material – students’ activities, and (3) students’ final course grades. A special database was designed to accumulate the collected data and provide a more structured and organized data source to generate the desired dataset. Python programming language, SQLAlchemy database toolkit and a SQLite database were used to implement and work with the database. The data for the overlay model was collected from the Moodle LMS by exporting the questions and students’ responses from all quiz activities (midterms, flash-tests, self-assessment quizzes, etc.). Students’ activities were collected from course activity logs on Moodle LMS. Students’ final course grades were collected from two data sources: (1) from our own automatic grading sheet implemented in Microsoft Excel, and (2) from the national HE information system (ISVU) exported as an Excel sheet. Grades from (1) were the grades students achieved during the semester. Grades from (2) were the grades students achieved during multiple post-semester examinations. In the end, a pivot table was created as the main dataset for analysis, where the columns Student and Center were used as indices, column Property name was used to generate the column headers, and Property value was used to display the pivot table values. The structure was as follows: Student ‒ student’s full name. Center ‒ the center (cities where our course is taught, more details in subsequent Section 4) student belonged to. Student’s self-assessment scores for the knowledge domains (more details on knowledge domains can also be found in subsequent Section 4): Information systems (IS) Memory unit (MU) Basic computer principles (BCP) Computer software (CS) Information system security (ISS) Central unit of a computer (CUC) Input/Output unit (IOU) Lessons 1.1–1.5 (L1.1-L1.5), Lessons 2.1–2.18 (L2.1-L2.18), Lessons 3.1–3.4 (L3.1-L3.4), Lessons 5.1–5.2 (L5.1-L5.2) ‒ student’s activity in lessons (reading materials for all knowledge domains, more details on lessons in subsequent Section 4). Grade ‒ student’s final grade. The presented dataset was used to: Perform the correlation analysis between the main components within the dataset (i.e., overlay model values, students’ activities and final course grades). Analyze the entire dataset by means of clustering, using the Silhouette method to determine the optimal number of clusters and k-means clustering algorithm to form the actual clusters. Apply descriptive statistics to determine the meaning of each cluster, i.e., what kind of student group each cluster may represent. Having a full dataset available, the first step within the data analysis phase was to analyze the correlations between the main components of the data set (i.e., overlay model values, students’ activities and final course grades). Afterwards, the entire dataset was analyzed by means of clustering, using the Silhouette method to determine the optimal number of clusters and k-means clustering algorithm to form the actual clusters. Having the clusters formed, descriptive statistics was used to determine the meaning of each cluster, i.e., what kind of student group each cluster may represent. In order to assess the homogeneity of grade distribution across academic years and centers, a log-linear model was analyzed using generalized linear regression with Poisson distribution and the natural logarithm link function. Course description The course “Business Informatics” is taught in the first semester of a three-year undergraduate professional study program at the Faculty of Organization and Informatics of the University of Zagreb. The study program is transdisciplinary, including the fields of informatics and economics. The syllabus of “Business Informatics” covers several major units of content – overview of the information systems and their business applications, deeper insight into computer hardware and software (basic elements of information systems) and the basics of information systems’ security. Under normal circumstances “Business Informatics” is a blended course held at four different locations in Croatia - at the main Faculty location in Varaždin (hereafter, Main Center - MC) and at three additional dislocated study centers in other towns in Croatia (hereafter, Dislocated centers - DCs). However, during the past five years the course was also piloted as an online course in DCs and with that respect, the materials as well as the methods were prepared for online teaching and learning. Therefore, during the COVID-19, the course “Business Informatics” was carried out as a fully online course, for MC and DCs according to the methodology used in previous years for DCs. Onsite classes were replaced by synchronous2 and asynchronous online activities. All asynchronous activities were offered as digital contents within LMS (reading materials, quizzes, self-assessments), supplementing the synchronously delivered online lectures. Students could have worked at those activities at their own pace, without strictly enforced schedules or deadlines. During the academic year 2020/2021, the course was enrolled by 222 students in total, out of which 101 were enrolled in MC and 121 in DCs. Course design Synchronous and asynchronous parts of the course were administered in LMS Moodle and its learning design aspects were organized according to the basic principles of programmed learning paths, as summarized by Seel (2012): learning contents are broken down into smaller pieces of content, which are immediately followed by one or more comprehension questions. student receives immediate feedback about the correctness of the answers. if the answers are correct, students can proceed to the next piece of content. if the answers are not correct, students are required to revise the content and answer the questions again. The asynchronous materials for online classes were prepared for every part of the syllabus. Students are supposed to use the asynchronous programmed learning path of the course activities to relearn already taught course topics at their own pace (i.e., topics covered during synchronous online lectures), or to use it as a primary source of learning about topics which were planned for asynchronous self-studying. The technical part of the programmed learning path was facilitated by the built-in Moodle features, primarily by Lesson activities. Lessons in Moodle allow combining the content pages (text + multimedia) and questions with feedback into a single unit which was found to be ideal for creating basic building blocks of the programmed instruction within the “Business Informatics” course. According to Britain (2004), the learning workflow originating from a learning design enables the teachers to create more structured teaching activities, thus leading to more effective learning. Both Britain (2004) and Dohn (2010) stress that the inclusion of learning designs into a course puts the learning activities in focus and provides a framework for deep reflection during the course design process. Every major unit of content was organized as the domain of knowledge in the following way (a programmed learning path): Unit topics were broken down into a sequence of lessons. Each lesson consisted of several pages and finished with several questions that all must be answered correctly. The immediate feedback was given after each question and if the answer was not correct, students could retry the question or return to the content pages within the lesson. Access to the following lesson was allowed only when all the questions of a previous lesson had been answered correctly (i.e., 100% completed previous lesson). Every major unit (domain of knowledge) ended with a final self-assessment quiz which covered all the lessons within the unit and assessed the domain knowledge. Access to the self-assessment was possible only when a student achieved 100% completion rate within all lessons in that unit. The number of attempts was unlimited. Figure 1 additionally shows the sequential flow of the conditional activities within any major unit/domain knowledge of content. Conditional activities (technical features of the Moodle LMS) were used to control students’ flow through the elements of each domain of knowledge. Access to most of the resources in Moodle (lessons and quizzes included) can be based on students’ individual results achieved on any other LMS resource within the course (i.e., another lesson, quiz, assignment, forum participation, etc.). These conditional activities have been used to prevent students from accessing further elements within the domain of knowledge, until they had shown the required mastery of the previous elements. The precondition for accessing the following lesson was achieving the required percentage of correct answers to the control questions at the end of a previous lesson, while the access to the final self-assessment quiz was conditioned by achieving required success in all lessons within the domain. Fig. 1 Lessons flow within major units of content, followed by final self-assessment All major units of knowledge, organized as shown in Fig. 1, were available in LMS to provide the asynchronous portion of the classes. Students from all centers (MC and DCs) had equal access options to them. For completing the final self-assessment for a domain knowledge with at least 75% success, a student has been awarded a badge, which proved that a certain level of mastery within that unit was achieved. Ideally, each student would earn a badge for every major unit in course. In total, 7 major units or domains of knowledge are covered in the course, as shown in Fig. 2: Information Systems, Information Systems Security, Basic Computer Principles, Central Unit of a Computer, Memory Unit, Input/Output Unit, and Computer Software. The mappings between domains of knowledge and lessons are also highlighted. Named circles at the beginnings of the two chains in Fig. 2 represent domains of knowledge which are mutually independent, and students may choose to begin the asynchronous portion of studying from either of these. Subsequent circles within the chains represent domains which depend upon previous domains and should be studied after the required level of mastery in previous domains is acquired. For example, to advance to the “Memory Unit”, students must first study “Basic Computer Principles” and then “Central Unit of a Computer”. Fig. 2 Dependencies between the main domains of knowledge (units of learning contents) Formal knowledge assessments Besides the mentioned self-assessments for domain knowledge, the formal online midterms and the final exam had to be taken. For the midterms (during semester) and the final exam (after the semester has ended) the formal online tests were prepared within the same Moodle course (new tests for each midterm/exam period). The grades from all those formal tests were indirectly included in research as part of students’ final grades. Results Dataset analysis started with the calculation of correlation coefficients to understand the relationships between the main components of the dataset (overlay model, students’ activity, and final grade), i.e., how much effect do variables from one component have on the variables in other components. Next, the cluster analysis was conducted on the dataset. Before cluster analysis, the Silhouette method was employed to determine the optimal number of clusters for the clustering process. Secondly, after dataset clustering, a total number of students per cluster and distribution of final grades per cluster were counted as a form of introductory insight into the clustering result. Finally, descriptive statistics were applied to analyze the characteristics of individual clusters, i.e., to determine what kind of student group each cluster represents. Correlation analysis between the main components of the dataset The correlations have been analyzed in programming language Python using the pandas data analysis library and visualized with the seaborn data visualization library as a heatmap. Firstly, the Shapiro-Wilk test was applied on all variables to check the dataset for normality (Yap & Sim, 2011). Since none of the variables were normally distributed, the Spearman correlation was used due to the several characteristics that are of interest for this analysis (Schober et al., 2018): it is useful for non-normally distributed data. it uses rank of values of the variables with calculated correlations, instead of the actual values and. can be used for ordinal data. For (1), Shapiro-Wilk test was already used to determine that none of the variables have normally distributed data. For (2) and (3), since the values of all the variables (except for Grade) are in range between 0 and 1, the actual values are converted into ordinal ranks. Table 1 shows the conversion information, as well as the description of what a rank means for the type of variable. Variable Grade is already considered to have ranked values. Table 1 Conversion information for the Spearman correlation Actual value Ranked value Description for the overlay model Description for the students’ activity 0.0–0.2 1 Lowest knowledge of the domain Lowest activity on lesson 0.2–0.4 2 Low knowledge of the domain Low activity on lesson 0.4–0.6 3 Moderate knowledge of the domain Moderate activity on lesson 0.6–0.8 4 High knowledge of the domain High activity on lesson 0.8–1.0 5 Highest knowledge of the domain Highest activity on lesson Figure 3 shows the Spearman correlation of the dataset as a heatmap. It can be noted that there is a segregation between the subsets of data. While the variables from the overlay model (self-assessments IS, BCP, CUC, MU, IOU, CS and ISS) clearly have a medium-to-large effect on each other (from 0.5 to 0.8), they mostly display a small effect (0.2 to 0.4) on the variables that form the student activity (reading of lessons, L1.1 until L5.2). Likewise, the variables that form student activity have a larger effect on each other (from 0.4 to 0.8) and mostly a small effect on the variables of the overlay model (from 0.2 to 0.4). It can be concluded that, for the majority of the students, accessing lessons does not contribute greatly to their understanding of the domain knowledge. Fig. 3 Spearman correlation of the dataset Furthermore, it is important to note the relationship between the overlay model, the student activity variables and the Grade variable. While the overlay model variables have a medium-to-large effect on the grade (from 0.4 to 0.6), the effect of the student activity on the grade is low (from 0 to 0.3). Based on these findings it can be concluded that the level of knowledge affects the student grade but the frequency of reading the lessons doesn’t significantly affect the increase in their final grade. Cluster analysis of the dataset The cluster analysis was performed by application of the k-means clustering algorithm on the dataset followed by the content analysis of the clusters. The k-means algorithm is a non-hierarchical clustering algorithm that can be used for various tasks and analyses in various data-related fields. As reviewed by Dutt et al. (2017) this algorithm is frequently used in the field of educational data mining. However, before application of the k-means algorithm, the Silhouette method was used to determine the optimal number of clusters (i.e., in Rousseeuw, 1987; Chiang & Mirkin, 2010; De Amorim & Hennig, 2015). For determining the optimal number of clusters and clustering, the data in the student activity subset of variables have been scaled due to the extremely high or low values for some students. The dataset has been clustered nine times, each time with a different number of clusters: starting with two and ending with ten clusters. Then, using the scikit learn library for the Python programming language, specifically the metrics module, the Average Silhouette Width (ASW) was calculated for each clustering process. Finally, the ASW scores were plotted to visually determine the optimal number of clusters. The Silhouette method plot shown in Fig. 4 was used to detect the optimal number of clusters. The highest score of 0.44 was given to the clustering with three clusters, which was selected as the optimal number of clusters for clustering analysis. The three clusters were labeled C1, C2, and C3 and were further analyzed. Fig. 4 Silhouette method plot Final grades per cluster The frequency of each individual grade was counted for each cluster and displayed in Table 2. Grades3 range between 1, representing the lowest grade (meaning students failed the course) and 5 representing the highest. Table 2 Final grades for each cluster Grade Clusters Total C1 C2 C3 1 1 125 22 148 2 37 0 7 44 3 23 0 4 27 4 2 0 1 3 5 0 0 0 0 No. of students 63 125 34 222 No. of dropoutsa 2 21 3 26 aStudents who left the study program within 9 months after the course ended Cluster C1 consists mostly of students with the passing grades except for one student that failed the course. In comparison, C2 includes only the students that have failed the course. Finally, C3 contains a smaller number of both, students that passed and the students that failed the course. Regarding the student dropouts, cluster C2 includes the highest number of dropouts (21 dropouts which is 80.77% of the total number of dropouts in the course). The characteristics of C3 will be elaborated after further analysis in the following sections. Standard deviations per cluster In this section, the standard deviation for each cluster is calculated and displayed within the charts in Fig. 5. Figure 5a shows the standard deviations of the overlay model part of the dataset and Fig. 5b shows the standard deviation of the students’ activity. In these cases, the values for the overlay model and the values of the students’ activities are scaled to range between 0 and 1, so the calculated standard deviations are in the same range. Fig. 5 a Standard deviation for the overlay model for each cluster. b Standard deviation for students’ activities for each cluster Cluster C1 consists mostly of students earning their passing grades (grades 2, 3, and 4), with one exception, and with the standard deviations fluctuating around 0.2 in the overlay model. The standard deviations of students’ activities fluctuate around the 0.1 value which could point to the fact that the clustering algorithm was more dependent on students’ activities than domain knowledge. The deviations not being closer to zero could also be attributed to not all students in the cluster making the same effort, but still earning the passing grade. A student that earned 2 as the final grade probably didn’t fully understand the domain knowledge and/or wasn’t very active in studying the course materials. On the contrary, a student that earned 4 as a final grade probably understood the domain knowledge to a greater degree and/or was actively studying through the lessons. In the case of cluster C2, the standard deviation for the overlay model for the first 4 knowledge domains is above 0.2 (with one domain above 0.3), while for the remaining 3 domains the values are around 0.2 or lower. This is the cluster that contains only the students who failed the course. Low standard deviations in the students’ activities (mostly below 0.1) suggest that the clustering algorithm depended mostly on students’ lower activities and failing grade. Standard deviations for the cluster C3 are higher than in the other two clusters which can be explained by more varied final grades and the self-assessment results of these students. Additional analysis will show that this cluster also contains the most active students, which in conjunction with varying final grades (ranging from 1 to 4) could also explain the higher standard deviation in the overlay model. Mean and median comparison per cluster In this section, the mean and median values were calculated for each attribute of each cluster and displayed in two charts: (1) for the overlay model part of the data (see Fig. 6), and (2) for the students’ activities part (see Fig. 7). The levels of domain knowledge and the levels of activity are presented using the ranks from Table 1. Fig. 6 Mean and median values of the overlay model for all three clusters Fig. 7 Mean and median values of students’ activities for all three clusters In the overlay model of the cluster C1, the means and medians show moderate levels of knowledge (0.4 to 0.6 range) for all domains except for one, having low level of knowledge (0.2 to 0.4 range). The students’ activities in this cluster falls into the lowest activities range (0.0 to 0.2) for all lessons but one. However, in comparison with the C2, C1 is still a more active cluster. For the cluster C2, means and medians of the overlay model show the lowest level of knowledge for all domains (all below 0.1). Medians for all domains are zero, indicating that at least half of all the students in this cluster have the lowest level of knowledge of the domains (0.0 to 0.2 range). This cluster also shows the lowest students’ activities levels, consistently being below 0.1. Starting with the lesson L1.3 median values are zero, indicating that from that lesson onwards at least half the students in the cluster C2 stopped accessing course materials. There are notable inconsistencies in mean and median values for the C3 cluster’s overlay model. Means range from the lowest levels of the domain knowledge (slightly below 0.2) to the moderate levels (0.4 to 0.6). Median values are zero for two domains (see Fig. 6), indicating that at least half of the students in C3 have shown the lowest level of knowledge of these domains (0.0 to 0.2 range). Regarding the students’ activities in C3 cluster, both means and medians are relatively equally distributed between low (0.2 to 0.4) and moderate (0.4 to 0.6) levels. Figure 7 clearly indicates that the cluster C3 is the most active of all three clusters. These data support the initial assumption that the cluster C3 contains mostly the students who try to compensate for the difficulties in understanding the domain knowledge by the increased activity (i.e. more frequent reading of lessons). Additional analyses of the clusters Additional analyses were made on the clustered dataset to examine the per-cluster distribution of students from all centers (MC and DCs) and the frequency of taking self-assessment quizzes by clusters and knowledge domains. Distribution of students from study centers by clusters Table 3 shows the distribution of students from the study centers by clusters. Table 3 Distribution of students from each study center by clusters Cluster MC DC Total (clusters) C1 56 7 63 C2 31 94 125 C3 14 20 34 Total (centers) 101 121 222 From the perspective of the clusters, 88.89% of the students (56 out of 63) in the cluster C1 are enrolled in the MC. Most of the students in the cluster C2 (75.2%, 94 out of 125) belong to the dislocated centers (DCs). The most equal per-center distribution is noted for cluster C3 - approx. 42% of students (14 out of 34) belong to MC and approx. 58% (20 out of 34) belong to DCs. The analysis from the perspective of the centers gives better insights, particularly when it comes to identifying the problematic centers. Students from the DCs are placed predominantly in the clusters C2 (94 out of 121 students, 78%) and C3 (20 out of 121 students, 16%). Students from the MC are mostly distributed in clusters C1 and C2, with cluster C1 having the dominant position (56 out of 101 students, 55% in C1 vs. 31 out of 101 students, 31% in C2). According to this data, it can be concluded that the DCs are more problematic centers, having the majority of students (78%) in the cluster C2, representing the students that failed the course. The MC has approximately 55%/45% split between the successful students (in C1) and the students that either failed the course (in C2) or had difficulties in their studies (in C3), which could indicate there is still room for improvement in course design and more active monitoring of such students. Frequencies of taking self-assessment quizzes by clusters Figure 8 shows the frequencies of taking self-assessments, which are sorted in order defined by the course curriculum. Only 137 out of 222 students in the dataset have been taking the self-assessment quizzes. As mentioned in the Subsection 4.1, the number of attempts to solve the self-assessments was not limited. Fig. 8 Frequency of taking self-assessment quizzes by clusters The Information systems self-assessment quiz (the first self-assessment quiz chronologically available in the course) shows the highest frequency of attempts. Also, it is the only quiz with more than 100 attempts of solving in all clusters. This could be attributed to different reasons: students starting the course with high hopes, interest and/or confidence, students having difficulties with understanding the domain knowledge and taking the quiz countlessly until they get the answers correct, etc. A significant decrease in number of attempts for the next 3 self-assessment quizzes and even greater decrease for the last 3 was recorded. This could be attributed to either students losing interest in the self-assessment quizzes (especially in the cluster C2) or to students having less difficulties with understanding the topics. From the perspective of the clusters, we can see that the cluster C1 (which includes the students that, with one exception, have all received passing grades and are considered to have no difficulties with understanding the domain knowledge) has the second highest frequency of taking self-assessment quizzes. The lowest frequencies of taking the self-assessments are noted in the cluster C2 (students that failed the course and had low activity in accessing course materials). The cluster C3, with the most active students having various but mostly failing grades, also has the highest frequencies in taking the self-assessment quizzes. These frequencies reinforce the idea that the cluster C3 includes the students that have difficulties with understanding domain knowledge and spend more time studying in order to compensate. Discussion The dataset that was used in the analysis was constructed from students’ overlay model, their frequency of reading materials in Moodle LMS and the earned final grade. Next, a clustering technique was used to generate clusters which were further analyzed to determine student groups. After determining the optimal number of clusters with the Silhouette method, three clusters were formed from the dataset and identified as the following student groups: Cluster C1 contains 63 students, out of which 62 have passed the course. Cluster standard deviations show a low variance in data when it comes to students’ activities within the LMS. The mean and median values for the students’ activities report they had a low to medium level of activity within the course. Overall, C1 seems to contain somewhat active students with better understanding of the domain knowledge than the students in other clusters. Cluster C2 contains 125 students, all of which have failed the course. Standard deviations also show a low variance in data regarding students’ activities in LMS. Cluster’s mean and median values show that at least half of the students had the lowest level of domain knowledge, as well as not being active in the course. Overall, C2 seems to contain the students that have failed the course, had low or no understanding of the domain, and were either slightly active or not active at all. Cluster C3 was the most intriguing for further analysis. It contains 34 students, out of which 12 have passed and 22 have failed the course. Its standard deviations show a higher variance in comparison to C1 and C2. Results show that at least half of the students had the lowest level of knowledge of two domains (Basics of computer operations and Computer software), a low level of knowledge for three domains (Central processing, Memory and I/O units) and moderate level of knowledge for two domains (Informations systems and Information systems security). The mean and median values place them in the moderate level of activity. Overall, C3 seems to include students that were very active on the course (this seems to be their primary characteristic), but a higher percentage of them failing the course, and having difficulties understanding most of the domains or having difficulties with studying. These characteristics could mean that the C3 consists of students that are struggling with the course. Regarding the grade, they either barely got the passing grade or have the great potential of getting it. When comparing activities of students in cluster C3 with the activities of students in other two clusters, we can observe the following differences, which support the conclusion about students in C3 having difficulties with understanding knowledge domains or with the studying process: Figure 7 shows that the students from C3 had the highest number of recorded access to each of the 29 lessons in LMS. When compared with C1 (students with mostly positive grades, see Table 2), the cluster C3 containing almost 50% less students than C1 (34 from C3 vs. 64 from C1, see Table 3) has achieved ranks which are at least twice as high as ranks from C1. This indicates that the students from C3 have been reading all the lessons more often than students from C1 and definitely more often than students from C2 (inactive and failing students, having median activity rank equal to 0 for 27 of the 29 lessons). The frequencies of taking self-assessment quizzes at the end of domains (Fig. 8) show that the students from C3 have generated the highest number of attempts to solve each of the 7 self-assessments. For the first quiz (on Information systems) it is evident that 34 students from C3 have accumulated approx. 175 attempts, averaging at approx. 5.1 attempts per student, while 64 students from C1 have accumulated approx. 150 attempts, averaging at approx. 2.3 attempts per student. Similar pattern can be observed for the other 6 quizzes. We can interpret more frequent attempts at solving these quizzes in cluster C3 as an indicator of insecurity in their knowledge, i.e. needing more attempts to achieve the positive effects of repeated quizzing on long-term retention of knowledge (Larsen et al., 2015). When these frequency-based findings for C3 are combined with the scores achieved in those quizzes (see Fig. 6), we can see that the scores of C3 are consistently one rank lower than C1 scores for 6 out of 7 quizzes. This supports the conclusion that students in C3 have difficulties in understanding most of the topics in the course. One of the future improvements regarding the course design, especially in the context of the cluster C3 including active students who struggle with understanding would be to embed the instant feedback within online quizzes, as suggested by Jia and Zhang (2019) in order to help students with poor academic performance in the final exam. Fig. 9 Mosaic plot of the number of students by grade, academic year, and center (MC = main center, DC = dislocated center) In order to put the results of clustering in a wider context, the distribution of grades by the center and academic year was analyzed (Fig. 9). The course takes place in the winter semester, thus the academic year 2020/21 was the first ERT year. There was variation in grade distribution among the three academic years, but the ERT year was within the pre-COVID variation in grade distribution. There was also a consistent difference in grade distribution between the centers, with more students failing in the dislocated centers. This difference was more pronounced in the ERT year. Analysis of the log-linear model with student numbers as the dependent variable, and grade, year, and center as independent variables, showed that three-way interaction was statistically significant (likelihood ratio χ2=13.74,df=4,p=0.0082). Specifically, in 2020/21 the main center had more students with grades 2 (β=1.526,z=2.862,p=0.0042), and 3–5 (β=1.408,z=2.019,p=0.044) in comparison to the baseline year 2018/19, dislocated center, and grade 1, after accounting for independent effects of grade, year, center, and their pairwise interactions. Therefore, it is evident that the number of failing final grades was high even before COVID-19 (esp. in DCs) and that it was not a reflection of the ERT and the COVID-19 context. It is also evident that students from the MC have consistently shown better results than students from the DCs. The locality can also be excluded as one of the factors for allocating students from DCs primarily to the cluster C2 as students in that cluster come from all three DCs (being geographically quite distant too), as well as from the MC. The high number of failing students within the cluster C2, which mostly includes the students from DCs, may also be explained by the insufficient individual communication with the teachers. The students from DCs mostly use asynchronous materials since the majority of them are part-time students attending or participating in the synchronous activities very rarely. Furthermore, most of them haven’t used the available means of consultations with the teachers (e-mail, online synchronous consultations, and forums available in LMS). Rienties and Toetenel (2016) state that the time students spend on communication activities is one of the major predictors for academic retention suggesting that the learning outcomes should be aligned with well-designed communication activities. Therefore, for the students in the cluster C2, as well as for the failing students in the cluster C3, a series of consultations should be organized with the teacher throughout the semester to address their issues at the individual level. Although the engagement level (i.e., activity recorded in LMS) can be used as a predictor of students’ academic performance (Moubayed et al., 2018), it does not necessarily mean that the students with the best performance will be the most active students (i.e., those that will read the learning materials most often). The comparison of the activity levels between clusters C1 and C3 shows that less-to-moderate successful students (C3) have been more active in LMS than the most successful students (C1), i.e. they were reading the materials more often. These results support the findings of Marques et al. (2018), who had a similar observation that the most successful students were not the ones with the highest average access scores within the e-learning platform. Likewise, we can conclude that students with the best final results are more confident in their domain knowledge, resulting in less frequent access to the course materials provided in LMS. Cluster analysis results from Subsection 5.1.1 suggest that students have been using assessments more consistently than accessing the lessons. This is in line with the observations from Manwaring et al. (2017), stating that students’ perception of the importance of an activity has a strong positive effect on students’ engagement level, both cognitive and emotional. In line with this study, we can conclude that higher engagement in self-assessments gave our students better perception of learning and improvement. Correlations observed in Subsection 5.1 would also suggest that students’ activities (i.e., accessing the materials) did not contribute greatly to their understanding of knowledge domains, i.e. accessing materials does not significantly affect the increase of the knowledge because it does not imply that they have read and understood the materials. Although this is in contrast with findings from other studies (Orji & Vassileva, 2020; Nguyen, 2017; etc.), in our particular case it may be partially explained by two preliminary remarks: The very technical nature of materials - materials can be printed-out from the LMS so students may have been reading them outside the LMS too. The programmed learning paths of the course - realistically, students had to go through the lessons only once to unlock the access to self-assessments. So, once they gained the access to self-assessments, which they arguably also perceive as more beneficial for their learning, they lost the interest to re-read the materials. Especially if they also had printouts for “easier” learning. Decision to rely strongly on various assessment activities in the learning design used in this course is supported by Nguyen et al. (2017) and Lei et al. (2018), showing a significant relation between assessment activities and students’ success rates, as well as the fact that both the learning design and the computer-based assessments affect students’ online learning. Findings from Orji and Vassileva (2020) also suggest that the academic performance, as well as their final grade, can be predicted by students’ engagement (i.e., activity recorded in LMS), assessment and assignment scores. Limitations and future work Study limitations There were several constraints which have to be taken into account when trying to generalise the results of this study. As we have already mentioned, the study included a large number of students (200+) and therefore the results may not apply equally to situations where the ERT was conducted with smaller groups. It is especially seen in the communication activities between teachers and students, i.e. teachers cannot regularly communicate with every student to offer help or guidance. Another specific factor was the nature of self-assessment quizzes used in course. Since all the quizzes were automatically evaluated by LMS without teacher’s intervention, all the questions were designed to enable the automatic grading (single/multi choice, matching, etc.) and the essay-type questions were not used. This limitation can also be linked to the large student groups - manual grading of essays would overburden the teachers, especially considering that all quizzes had an unlimited number of attempts. The nature of the course itself may also be a differentiating factor since the course used in this research is mainly theoretical and the results may not be generalised to more practical courses. Finally, since this study is the first stage of the planned research (see the possibilities for future research outlined above), only a limited set of criteria directly available in LMS and in the Faculty’s Information System was measured - activity in reading lessons, results of self-assessment quizzes and the final grades. Other sets of criteria which could additionally explain students’ behaviour were not included in this study (students’ attitudes, learning styles and strategies, ICT literacy, studying conditions, etc.). Future research This study is focused on the so-called “specific domain of information” according to Anouar Tadlaoui et al. (2016) which involves the level of knowledge and other specific information about the learner such as records of learning activities and records of evaluation. Since students enter this course in the first year of the undergraduate study programme, and since this is a rather complex and demanding 5 ECTS course, it usually results in a large number of failures in the end. Due to the fact that the high dropout rate can discourage students in further studying, there is a need to detect the students at risk at the early stages of their education and to assist them to master this course. Therefore, it was decided to focus on the data that show students’ performance in the course since improving student retention rate is a vital objective for this particular group of students and it is tightly connected with the success rate of the study programme. However, according to the same source (Anouar Tadlaoui et al., 2016), user models can be enriched with the so-called “independent domain information” which involves users’ goals, attitudes, motivation, background, experience and preferences. With this in mind, the next step in this research should include the latter data to better understand how different students’ characteristics influence their learning and their final success. As the next step, a course satisfaction survey will be prepared with the socio-demographics questions that will enable further research on the higher failure rate of the DCs students. Even without using enriched data sets, there are other possibilities for future research. For example, current cluster C3 (very active students, but struggling with studies and failing to achieve passing grade) would benefit from applying predictive analytics in order to enable early detection of such students and offering them more teacher guidance. Predictive analytics could also be used with current cluster C2 (inactive and failing students), again to enable their early detection and to incentivize their activity within the course. Conclusion This research follows contemporary innovations in education and uses different techniques of learning analytics (LA) to identify possible improvements in the course design and teaching strategies, as it is highlighted within the most recent research. It complements the field by determining actual student groups post hoc, based on the analysis of available data about students’ activities throughout the entire semester and modeling students’ knowledge. This is especially important in the current situation where emergency online teaching facilitated the shift to the online environment and a lot of data is recorded about students and their behavior that could be used to improve learning design and teaching strategies. In this case, final grades were correlated with the results of two types of activities: (i) knowledge displayed on various online tests during the semester (many formative and summative online tests within the course - multiple flash exams and self-assessments, two midterms and the final exam) and (ii) frequency of accessing the course materials (i.e., students’ activity). Three different clusters of user groups were detected: C1 including students that earned passing grade, that were moderately active during semester and better understood the domain; C2 including the students that have failed the course, had low or no understanding of the domain and were mostly not active; and C3 including very active students but many of them having difficulty with understanding the domain or having difficulties with their studies. Finally, the research showed that the clustering of students can enable teachers to re-think about the design and teaching strategies used in the course to increase students’ retention rate and their final grades. However, it was shown that in many cases the programmed learning path with well-balanced self-assessment can lead to successful mastery of the course and that many students can be guided automatically throughout the course using LMS thus helping teachers to better monitor students in large classes. Authors’ contributions Study conception, design and methodology received the most contribution from Igor Balaban and minor contribution from Miran Zlatović. Data collection, curation, analysis and visualization were performed by Danijel Filipović. Writing of the first draft of the manuscript, as well as all the intermediary versions and the final version was the collaborative effort of all authors. All authors read and approved the final manuscript. Funding This work was co-financed by the Croatian Science Foundation project IP-2020-02-5071. Data availability The datasets generated and/or analyzed during the current study are not publicly available due to GDPR restrictions i.e., them containing information that could compromise research participants’ privacy but are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare that they have no competing interests. Abbreviations ASW Average silhouette width BCP knowledge domain Basic computer principles CS knowledge domain Computer software CUC knowledge domain Central unit of a computer DC Dislocated study center ERT Emergency remote teaching HEI Higher education institution IOU knowledge domain Input/Output unit IS knowledge domain Information systems ISS knowledge domain Information system security LA Learning analytics LMS Learning management system MC Main center MU knowledge domain Memory unit 1 Course is held fully online as part of the emergency remote teaching process. Under normal circumstances it would be held as a blended course. 2 Classes were live-streamed using video conference tools, according to the formal schedule. 3 In Croatia the following official grade scale applies to elementary school, high school and university students: 1-insufficient/failed, 2-sufficient, 3-good, 4-very good, 5-excellent. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References Anouar Tadlaoui M Souhaib A Khaldi M Carvalho R Learner modeling in adaptive educational systems: a comparative study International Journal of Modern Education and Computer Science 2016 8 1 10 10.5815/ijmecs.2016.03.01 Bignoux S Sund KJ Tutoring executives online: what drives perceived quality? Behaviour & Information Technology 2018 37 703 713 10.1080/0144929X.2018.1474254 Bond, M., Bedenlier, S., Marín, V., & Händel, M. (2021). Emergency remote teaching in higher education: mapping the first global online semester. 18. 10.1186/s41239-021-00282-x Bontcheva K Wilks Y Tailoring automatically generated hypertext User Modeling and User-Adapted Interaction 2005 15 135 168 10.1007/s11257-004-5637-6 Britain, S. (2004). A review of learning design:Concept, specifications and tools. Brusilovsky, P., & Millán, E. (2007). User models for adaptive hypermedia and adaptive educational systems. 4321. 10.1007/978-3-540-72079-9_1 Camacho, D., Pulido, E., R-Moreno, M., Carro, R., Ortigosa, A., & Bravo, J. (2009). Automatic course redesign: Global vs. individual adaptation. International Journal of Engineering Education, 25. Chiang M Mirkin B Intelligent choice of the number of clusters in K-Means clustering: an experimental study with different cluster spreads Journal of Classification 2010 27 3 40 10.1007/s00357-010-9049-5 Chrysafiadi K Virvou M Student modeling approaches: a literature review for the last decade Expert Systems with Applications 2013 40 4715 4729 10.1016/j.eswa.2013.02.007 Corrin, L., de Barba, P. G., & Bakharia, A. (2017). Using learning analytics to explore help-seeking learner profiles in MOOCs. Proceedings of the Seventh International Learning Analytics &Amp Knowledge Conference, 424–428. 10.1145/3027385.3027448 de Amorim RC Hennig C Recovering the number of clusters in data sets with noise features using feature rescaling factors Information Sciences 2015 324 126 145 10.1016/j.ins.2015.06.039 Dohn, N. B. (2010). Teaching with wikis and blogs: Potentials and pitfalls. Proceedings of the 7th International conference on networked learning, 142–150. https://www.lancaster.ac.uk/fss/organisations/netlc/past/nlc2010/abstracts/PDFs/Dohn.pdf. Accessed 2 June 2022. Dutt A Ismail MA Herawan T A systematic review on educational data mining IEEE Access: Practical Innovations, Open Solutions 2017 5 15991 16005 10.1109/ACCESS.2017.2654247 Ferguson R Learning analytics: drivers, developments and challenges International Journal of Technology Enhanced Learning 2012 4 304 317 10.1504/IJTEL.2012.051816 Francis B Sasidhar Babu D Predicting academic performance of students using a hybrid data mining approach Journal of Medical Systems 2019 43 162 10.1007/s10916-019-1295-4 31037484 Gilbert, B. (2015). Online learning revealing the benefits and challenges. Hodges, C., Moore, S., Lockee, B., Trust, T., & Bond, M. (2020). The difference between emergency remote teaching and online learning. Educational Review. https://hdl.handle.net/10919/104648 Jia, J., & Zhang, J. (2019). The analysis of online learning behavior of the students with poor academic performance in mathematics and individual help strategies (pp. 205–215). 10.1007/978-3-030-21562-0_17 Kebritchi M Lipschuetz A Santiague L Issues and challenges for teaching successful online courses in higher education: a literature review Journal of Educational Technology Systems 2017 46 4 29 10.1177/0047239516661713 Korkofingas, C., & Macri, J. (2013). Does time spent online have an influence on student performance? Evidence for a large business studies class. Journal of University Teaching and Learning Practice, 10(2), 1–13. Larsen, D. P., Butler, A. C., Aung, W. Y., Corboy, J. R., Friedman, D. I., & Sperling, M. R. (2015). The effects of test-enhanced learning on long-term retention in AAN annual meeting courses. Neurology, 84(7), 748–754. 10.1212/WNL.0000000000001264 Lei H Cui Y Zhou W Relationships between student engagement and academic achievement: a meta-analysis Social Behavior and Personality: an International Journal 2018 46 517 528 10.2224/sbp.7054 Lowenthal P Borup J West R Archambault L Thinking beyond zoom: using asynchronous video to maintain connection and engagement during the COVID-19 pandemic Journal of Technology and Teacher Education 2020 28 383 391 Manwaring KC Larsen R Graham CR Henrie CR Halverson LR Investigating student engagement in blended learning settings using experience sampling and structural equation modeling The Internet and Higher Education 2017 35 21 33 10.1016/j.iheduc.2017.06.002 Marques, B., Villate, J., & de Vaz, C. (2018). Student activity analytics in an e-learning platfom: Anticipating potential failing students. Journal of Information Systems Engineering & Management, 3. 10.20897/jisem.201812 Martins C Faria L de Carvalho V Carrapatoso E User modeling in adaptive Hypermedia Educational Systems Educational Technology & Society 2008 11 194 207 Mazza, R., & Botturi, L. (2007). Monitoring an online course with the GISMO tool: A case study. Journal of Interactive Learning Research, 18(2), 251–265. Moubayed, A., Injadat, M., Shami, A., & Lutfiyya, H. (2018). Relationship between student engagement and performance in e-learning environment using association rules. 2018 IEEE World Engineering Education Conference (EDUNINE), 1–6. 10.1109/EDUNINE.2018.8451005 Muthuprasad T Aiswarya S Aditya KS Jha GK Students’ perception and preference for online education in India during COVID – 19 pandemic Social Sciences & Humanities Open 2021 3 100101 10.1016/j.ssaho.2020.100101 34173507 Nguyen L Do P Learner model in adaptive learning World Academy of Science Engineering and Technology 2008 45 395 400 Nguyen Q Rienties B Toetenel L Ferguson R Whitelock D Examining the designs of computer-based assessment and its impact on student engagement, satisfaction, and pass rates Computers in Human Behavior 2017 76 703 714 10.1016/j.chb.2017.03.028 Nguyen VA The impact of online learning activities on student learning outcome in blended learning course Journal of Information & Knowledge Management 2017 16 1750040 10.1142/S021964921750040X Orji, F., & Vassileva, J. (2020). Using machine learning to explore the relation between student engagement and student performance. 2020 24th International Conference Information Visualisation (IV), 480–485. Otrel-Cass K Cowie B Campbell A What determines perseverance in studying science? Journal of Institutional Research 2009 14 2 30 44 Rapanta, C., Botturi, L., Goodyear, P., Guàrdia, L., & Koole, M. (2021). Balancing technology, pedagogy and the new normal: Post-pandemic challenges for higher education. Postdigital Science and Education, 3, 715–742. 10.1007/s42438-021-00249-1 Rienties B Toetenel L The impact of learning design on student behaviour, satisfaction and performance: a cross-institutional comparison across 151 modules Computers in Human Behavior 2016 60 333 341 10.1016/j.chb.2016.02.074 Rousseeuw PJ Silhouettes: a graphical aid to the interpretation and validation of cluster analysis Journal of Computational and Applied Mathematics 1987 20 53 65 10.1016/0377-0427(87)90125-7 Ryabov I The effect of time online on grades in online sociology courses MERLOT Journal of Online Learning and Teaching 2012 8 13 23 Schober P Boer C Schwarte LA Correlation coefficients: appropriate use and interpretation Anesthesia & Analgesia 2018 126 1763 1768 10.1213/ANE.0000000000002864 29481436 Schreck J User modeling. In: security and privacy in user modeling 2003 Springer Seel, N. (2012). Programmed learning. In N. M. Seel (Ed.), Encyclopedia of the sciences of learning (p. 2706). Springer US. 10.1007/978-1-4419-1428-6_671 Smaili, E. M., Sraidi, S., Azzouzi, S., & Charaf, M. E. H. (2021). Towards sustainable e-learning systems using an adaptive learning approach. Emerging Trends in ICT for Sustainable Development (pp. 365–372). Springer. Sun, A., & Chen, X. (2016). Online education and its effective practice: A research review. Journal of Information Technology Education, 15. Tseng, S. F., Tsao, Y. W., Yu, L. C., Chan, C. L., & Lai, K. (2016). Who will pass? Analyzing learner behaviors in MOOCs. Research and Practice in Technology Enhanced Learning, 11. 10.1186/s41039-016-0033-5 Watson SL Watson WR Yu JH Alamri H Mueller C Learner profiles of attitudinal learning in a MOOC: an explanatory sequential mixed methods study Computers & Education 2017 114 274 285 10.1016/j.compedu.2017.07.005 Wei HC Peng H Chou C Can more interactivity improve learning achievement in an online course? Effects of college students’ perception and actual use of a course-management system on their learning achievement Computers & Education 2015 83 10 21 10.1016/j.compedu.2014.12.013 Yap BW Sim CH Comparisons of various types of normality tests Journal of Statistical Computation and Simulation 2011 81 2141 2155 10.1080/00949655.2010.520163	36465418	PMC9709367	NO-CC CODE	2022-12-01 23:23:04	no	Educ Inf Technol (Dordr). 2022 Nov 30;:1-26	utf-8	Educ Inf Technol (Dordr)	2,022	10.1007/s10639-022-11468-9	oa_other
==== Front Intern Emerg Med Intern Emerg Med Internal and Emergency Medicine 1828-0447 1970-9366 Springer International Publishing Cham 3160 10.1007/s11739-022-03160-0 Em - Commentary PEGALUS and other patient predictive scores of COVID-19 patients http://orcid.org/0000-0001-7599-8459 Conners Gregory P. [email protected] grid.411023.5 0000 0000 9159 4457 Departments of Pediatrics, Emergency Medicine, and Public Health and Preventive Medicine, Norton College of Medicine, State University of New York Upstate Medical University, Syracuse, NY USA 30 11 2022 13 29 8 2022 20 11 2022 © The Author(s), under exclusive licence to Società Italiana di Medicina Interna (SIMI) 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 COVID-19 Clinical prediction scores Point of care ultrasound Electronic medical record ==== Body pmcWe physicians are taught to look for patterns in our patients. Mnemonics and other memory aids make it easier. Even more effective are clinical scores, combining data like patient demographics, medical, surgical and family histories, signs, symptoms, medication use, physical findings, and data from laboratory and imaging testing into a single score, ideally one that is simple to calculate and use, and hopefully has some useful predictive power. Of course, such scores work best when they have a memorable name! From the first minute of a patient’s life, for example, we have his or her Apgar score [1], and it continues from there. Clinical scoring systems can provide valuable guidance for resource use such as hospitalization, prophylaxis, treatment, and additional testing, and for prognosis. Health service researchers and biostatisticians have helped create robust methods for derivation and validation of such clinical scoring systems. COVID-19 has caused the first pandemic in the age of the electronic medical record (EMR). Huge amounts of data of all sorts have become readily available to help us understand the effects and predict the course of COVID-19 infection. Researchers are, naturally, looking for ways to turn potentially complex statistical models into clinical scoring systems that are at once powerful yet simple enough to allow them to be quickly and readily used in a busy clinical environment, such as an Emergency Department, perhaps while donning and doffing PPE and also assessing a variety of other patients. Many examples have already been developed in clinical centers all around the world, including: the “EXAM” score to predict future oxygen requirements of COVID-19 patients [2]; the “COVID-GRAM” score to predict development of critical illness [3]; the “JRSS” score to risk-stratify Emergency Department patients [4]; the “COVID-IRS” score to predict risk of mechanical ventilation [5]; and the “PAINT” [6], “CANPT” [7], and an Iranian score to predict severe COVID-19 illness [8], among many others. Generalizability, both geographically and over time, is an additional challenge; Wyants et al. reported that COVID-19 prediction models “are at high risk of bias, raising concern that their predictions could be unreliable when applied in daily practice.” [9]. Soto-Mota et al. have noted that the predictive power of COVID-19 mortality scores tend to decay over time, sometimes to the point where they are no better than clinical gestalt [10]. Of course, any scoring system for COVID-19 will be most accurately predictive in the context of the clinical situation when it was derived, potentially requiring revision in response to such changes as new immunization technologies and strategies, and to evolving COVID-19 virus strains. The most recent foray into the realm of COVID-19 scoring systems is the “PEGALUS” score of Borio et al. [11]. Developed in Italy, this scoring system is intended to predict death and/or orotracheal intubation within 30 days among Emergency Department patients, based on the study of 230 COVID-19-positive patients, 21.5% of whom went on to either or both of these endpoints. Although not specifically stated as an exclusion, the study does not appear to have included children. The scoring system includes points for: age 65 or more (yes/no); PO2/Fi O2 ratio (in one of four groups); PCO2 < 35 mmHg (yes/no); duration of symptoms < 7 days (yes/no); and visual Lung Ultrasound Score (LUSS) pattern (in one of four groups). The LUSS pattern component is based on bedside point of care ultrasound (POCUS) in the Emergency Department, and is a refinement of several earlier LUSS studies [12–15]. It is itself a novel scoring system, using protocolized POCUS scanning of twelve lung regions, with a score of 0–3 points being assigned to each region, based on severity of disease. Each region’s points are then totaled, and patients with point totals within specified ranges are allocated to LUSS pattern 1, 2, 3, or 4; patterns are then converted to points on the PEGALUS score. Total PEGALUS scores range from 0 to a possible maximum of 21.5, with worse scores corresponding to worse prognoses. PEGALUS scores < 7 were strongly associated with good outcomes, while those > 11 were strongly associated with adverse outcomes. Thus, PEGALUS identified both high-risk and low-risk patient populations, as well as a mid-range group. The authors suggest that patients with PEGALUS scores < 7 may be safely managed at home, those with scores of 7–11 undergo additional evaluation, and those with scores > 11 be rapidly hospitalized, perhaps in an intensive care unit setting. The PEGALUS scoring system has pros and cons. It uses data that are readily obtained in an Emergency Department setting: demographics, duration of symptoms, COVID-19 testing, blood gas results, and findings from POCUS, the use of which is rapidly becoming common in the Emergency Department setting. The area under the curve (AOC) for the Receiver-Operator Characteristic (ROC) curve, for the PEGALUS score was strong, at 0.866. Readers practicing in Italy may be especially interested in using a score derived in the same country. And, of course, the very name PEGALUS is inspiring! However, the process of using POCUS to determine LUSS for each of twelve regions, totaling the scores to convert them into a LUSS pattern group, then assigning PEGALUS points based on that pattern group, is somewhat cumbersome. While a patient’s COVID-19 vaccination status may alter the importance of clinic-demographic variables included in PEGALUS, vaccination history is not considered in the scoring system. Although the system has good predictive power as retrospectively derived, specifically for adult patients, there has not been a corresponding prospective or validation study in another patient population to test its generalizability. The intersection of the age of the EMR with the COVID-19 pandemic has given us unprecedented opportunities to rapidly develop prediction models specific for infection by this single virus. The PEGALUS score may represent a genuine advance in Emergency Department-based care of the adult COVID-19 patient. Further, as the authors point out, the PEGALUS score may also work well in understanding other forms of interstitial pneumonia [11]. The PEGALUS score must, however, undergo additional validation study before being widely adopted. As COVID-19 infection and its management evolve with changing immunization science, changing predominance of viral strains, improved viral detection and supportive care methods, and changing societal factors, we will need to continually re-test and refine this and all other COVID-19-specific predictive patient scoring methods. Perhaps there will be a PEGALUS-2 score in the future! What we learn through development of clinical scoring systems will certainly inform us about creating and maintaining patient predictive scores during the data-rich, high stakes pandemics of the future. Declarations Conflict of interest The author declares that he has no conflict of interest. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References 1. Ehrenstein V Association of apgar scores with death and neurologic disability Clin Epidemiol 2009 1 August 9 45 53 10.2147/clep.s4782 20865086 2. Dayan I Roth HR Zhong A Federated learning for predicting clinical outcomes in patients with COVID-19 Nat Med 2021 27 15 September 2021 1735 1743 10.1038/s41591-021-01506-.3 34526699 3. Liang W Liang H Ou L Development and validation of a clinical risk score to predict the occurrence of critical illness in hospitalized patients with COVID-19 JAMA Intern Med 2020 180 8 1081 1089 10.1001/jamainternmed.2020.2033 32396163 4. Peterson J, Jhala D (2022) Practical risk scoring system for predicting severity of COVID-19 disease. Clin Pathol Jan 7;15:2632010X211068427. doi: 10.1177/2632010X211068427. eCollection 2022 Jan-Dec. 5. Garcia-Gordillo JA Camiro-Zúñiga A Aguilar-Soto M Cuenca D Cadena-Fernández A Khouri LS Rayek JN Mercado M ARMII Study Group COVID-IRS: A novel predictive score for risk of invasive mechanical ventilation in patients with COVID-19 PLoS One 2021 16 4 e0248357 10.1371/journal.pone.0248357 33819261 6. Wang M, Wu D, Liu CH, et al. (2022) Predicting progression to severe COVID-19 using the PAINT score BMC Infect Dis 2022 May 26;22:498. 10.1186/s12879-022-07466-4. 7. Chen Y, Zhou X, Yan H, Huang H, LI S, Jiang Z, Shao J, Meng Z (2021) CANPT score: a tool to predict severe COVID-19 on admission. Frontiers in Medicine 2021 Feb18;8:608107. https://www.frontiersin.org/articles/10.3389/fmed.2021.608107/full.10.3389/fmed.2021.608107. 8. Aghajani MH, Sistanizad M, Pourhoseingholi A, Asadpoordezaki Z, Taherpour N. Development of a scoring system for the prediction of in-hospital mortality among COVID-19 patients (2021) Clinical Epidemiology and Global Health 6 October 2021, 12:100871. 10.1016/j.cegh.2021.100871. 9. 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 14 April 2020 m1328 10.1136/bmj.m1328 32265220 10. Soto-Mota A Marfil-Garza BA Castiello-de Obeso S Martinez Rodriguez EJ Carrillo Vazquez DA Tadeo-Espinoza H Guerrero Cabrera JP Dardon-Fierro FE Escobar-Valderrama JM Alanis-Mendizabal J Gutierrez-Mejia J Prospective predictive performance comparison between clinical gestalt and validated COVID-19 mortality scores J Investig Med 2022 70 2 415 420 10.1136/jim-2021-002037 34620707 11. Borio G, Tentori S, Farolfi F, et al. (2022) PEGALUS: predictivity of elderly age, arterial gas analysis, and lung ultrasound. a new prognostic score for COVID-19 patients in the emergency department—an observational prospective study. Intern Emerg Med 17(8); 2357-2365 10.1007/s11739-022-03047-0. 12. Lichter Y Topilsky Y Taieb P Banai A Hochstadt A Merdler I Gal Oz A Vine J Goren O Cohen B Sapir O Granot Y Mann T Friedman S Angel Y Adi N Laufer-Perl M Ingbir M Arbel Y Matot I Szekely Y (2020) Lung ultrasound predicts clinical course and outcomes in COVID-19 patients Intensive Care Med 2020 46 10 1873 1883 10.1007/s00134-020-06212-1 32860069 13. Ji L Cao C Gao Y Prognostic value of bedside lung ultrasound score in patients with COVID-19 Crit Care 2020 24 1 700 10.1186/s13054-020-03416-1 33353548 14. Manivel V Lesnewski A CLUE: COVID-19 lung ultrasound in emergency department Emerg Med Australas 2020 32 4 694 696 10.1111/1742-6723.13546 32386264 15. Via G et al (2012) Lung ultrasound in the ICU: from diagnostic instrument to respiratory monitoring tool. Minerva Anesthesiol 78(11): 1282-1296	36449259	PMC9709369	NO-CC CODE	2022-12-01 23:23:04	no	Intern Emerg Med. 2022 Nov 30;:1-3	utf-8	Intern Emerg Med	2,022	10.1007/s11739-022-03160-0	oa_other
==== Front Geriatr Nurs Geriatr Nurs Geriatric Nursing (New York, N.y.) 0197-4572 1528-3984 Elsevier Inc. S0197-4572(22)00255-5 10.1016/j.gerinurse.2022.11.005 Article Impact of using a centralized matching process on nursing home staffing Zarei Hamid Reza MS a Bart Yakov PhD b⁎ Ergun Ozlem PhD c a PhD Candidate, Department of Mechanical & Industrial Engineering, Northeastern University, Boston, MA, USA b Associate Professor of Marketing, D'Amore-McKim School of Business, Northeastern University, Boston, MA, USA c COE Distinguished Professor, Department of Mechanical & Industrial Engineering, Northeastern University, Boston, MA, USA ⁎ Corresponding author. 30 11 2022 January-February 2023 30 11 2022 49 8993 21 9 2022 8 11 2022 9 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. Objective To examine the effectiveness of adopting a novel centralized matching process for reducing staff shortages in Massachusetts nursing homes during the COVID-19 pandemic. Methods This study involved several datasets and 216 Massachusetts nursing homes that used a novel online portal to enter demand for nursing staff from May 2020 to April 2021. Results There were significant associations between the staff-to-resident ratio and demand entries lagged by three and four weeks, and no significant associations between the staff-to-resident ratio and demand entries lagged by one and two weeks. In contrast, we found significant associations between the staff-to-resident ratio and the number of generated staff matches lagged by one, two and three weeks, with larger impacts overall. Conclusion This study shows how adopting a centralized matching process may expedite and increase improvement in the staff-to-resident ratio in nursing homes, compared with the setup in which nursing homes need to seek nurses on their own. Keywords Nursing homes Staff-to-resident ratio Centralized matching process COVID-19 Staff shortages Resource allocation ==== Body pmcIntroduction Staffing shortages have plagued US nursing homes for decades.1 , 2 Barriers such as lower wages compared to alternative options, staff workload, and stressful duties make hiring and keeping workers complicated for nursing homes.3, 4, 5 This problem has worsened with the COVID-19 pandemic onset.6 According to a survey, 94% of nursing homes faced staffing shortages during the COVID-19 pandemic,7 caused by nurses testing positive for COVID-19,8 staff quitting,7 lack of childcare for staff,1 and overall pandemic burnout and fatigue.1 In addition to nationwide efforts to reduce these shortages,9 many state governments have implemented various local policies for the same goal. For instance, Michigan offered bonus payments to recently hired staff, created a rapid response team, and paid for staffing services; Utah invested in recruitment efforts and volunteer programs; Washington created a new tool for requesting volunteer staff, California subsidized travel costs for volunteer staff commuting across the state, Georgia allowed nursing homes to hire temporary (not certified) nursing assistants; Delaware provided nursing aid training for unemployed workers; Indiana created a list of available staff for nursing homes facing shortages; Wisconsin gave free training programs to volunteers.10 Such a wide variety of potential solutions deployed by different states reflects not only differences in policies across states, but also our lack of understanding of how effective these solutions are.10 This paper aims to comprehensively evaluate the effectiveness of reducing staff shortages in nursing homes for one of such solutions, implemented by the Commonwealth of Massachusetts during COVID-19. The core idea was to design and run a centralized process for matching demand for and supply of nursing staff, based on prior successful implementations of similar processes in various other labor markets.11, 12, 13, 14 The solution, supervised by Commonwealth's Executive Office of Elder Affairs and executed in collaboration with local researchers (experts from the Executive Office of Health and Human Services, Northeastern University, and University of Massachusetts Chan Medical School), involved creating an online portal to enable the dynamic and real-time collection of nursing staff demand and supply data, and using this data to optimize matching staff to nursing homes in need algorithmically.15 Nursing homes looking to hire nursing staff and nurses looking for work could register and interact with the portal. Nursing homes could use the portal to report their current needs for each nursing staff position in real time; nurses could note their current availability to start working. Throughout the paper, staff positions refer to three nursing positions which are certified nursing assistant (CNA), licensed practical nurse (LPN), and registered nurse (RN). After the Commonwealth's Executive Office of Elder Affairs set policies and criteria of matching that considered the temporal and spatial aspects of staff and nursing homes (e.g., the closeness of staff to nursing homes, staff available date and demand urgency), the centralized matching process generated matches that were then shared with the matched nursing homes. Nursing homes then could contact, interview, and hire the suggested matches to reduce their nursing staff shortage. Fig. 1 shows the roles of each stakeholder in the centralized matching process.15 Fig. 1 Overview of the centralized matching process. Fig 1Source/Notes: Zarei et al., 2022. The study hypothesis was that implementing the centralized matching process is strongly associated with increases in staff-to-resident ratios of nursing homes after controlling for the weekly numbers of resident deaths due to COVID-19, weekly numbers of COVID-19 positive cases among staff, and weekly numbers of resident admissions. We further hypothesized that these increases occurred faster, when compared with the setup in which nursing homes would need to seek nurses on their own. To test these hypotheses, we examined the centralized matching process effectiveness by combining the proprietary portal data with publicly available longitudinal datasets. Methods Design We designed this descriptive study to investigate the performance of the centralized matching process. Sample and procedure Our dataset combines weekly data (for 45 weeks, from the week ending on May 31, 2020, until the week ending on April 4, 2021) for nursing homes in Massachusetts from the following four sources:1. We used the Center of Medicare and Medicaid Services (CMS) COVID-19 nursing home database16 for the number of cases and deaths among residents, shortage of different resources such as PPE and staff, COVID-19 testing, and weekly resident admissions. 2. We used the CMS Payroll Based Journal (PBJ) nurse staffing database17 for the daily hours of nurse staffing and the number of residents in nursing homes. We have transformed these daily values to weekly ones for each nursing home by averaging nurse (RN, LPN, and CNA) staffing hours and the number of residents over the seven days comprising the corresponding week. 3. We used the portal database for weekly demand entries of nursing homes for different nurse positions and the weekly numbers of matched nurse staff to nursing homes. This novel data source, a central and crucial element of this study, was made available to us in real time during the study period (the data was used to optimize matching staff to nursing homes in need algorithmically, on a weekly basis). This real-time data availability allowed the central matching process to operate without any data lag. 4. We used LTCFocus (2019)18 (available from Brown University) database for some of the general characteristics of nursing homes and residents in Table 1 .Table 1 General characteristics of participating nursing homes in Massachusetts. Table 1 Participating nursing homes Mean Min Max SD Median Proportion of all admissions during the calendar year 2019 that were from an acute care hospital (n = 208, unit: percentage) 86.92 41.46 98.45 9.67 89.49 Acuity Index (a measure of the care needed by a nursing home's residents. (n = 213, range is from 0 to 28. 0 indicates completely independent and 28 completely dependent) 12.19 8.566 18.835 1.01 12.2 Average resident age (n = 213, unit: years) 81.37 25.74 91.29 7.21 82.97 Availability of an Alzheimer's disease Special Care Unit (SCU) (n = 213) Yes = 46 No = 167 Median Length of Stay (n = 209, unit: days) 30.45 14 120 23.31 22 Proportion of residents present on the 1st Thursday in April, 2019 who had a body mass index (BMI) of 35 or higher. (n = 188, unit: percentage) 25.75 0 60 6.96 25 Proportion of residents present on the 1st Thursday in April, 2019 who have congestive heart failure. (n = 169, unit: percentage) 23.33 0 46.67 7.38 22.99 Proportion of residents admitted during the calendar year (2019) with Alzheimer's disease or related dementia. (n = 208, unit: percentage) 27.79 0 98.18 13.86 24.95 Proportion of residents admitted during the calendar year who were low care, according to the broad definition (n = 68, unit: percentage) 3.48 0 42.42 7.19 0 Proportion of residents whose primary support is Medicaid (n = 213, unit: percentage) 63.44 0 100 18.59 64.94 Proportion of residents whose primary support is Medicare (n = 213, unit: percentage) 11.08 0 75 9.06 9.89 Proportion of long-stay residents with ADL decline (n = 203, unit: percentage) 14.46 0 45.45 6.66 14.29 Proportion of high-risk long-stay residents with a pressure ulcer (n = 203, unit: percentage) 5.61 0 17.14 3.51 5.48 For profit (n = 213) Yes = 150 No = 63 Overall Rating (n=214, ranges from 1 to 5 where 5 is highest) 3.27 1 5 1.37 4 Long-Stay QM Rating (n=214, ranges from 1 to 5 where 5 is highest) 3.53 1 5 1.29 4 Short-Stay QM Rating (n=200, ranges from 1 to 5 where 5 is highest) 3.67 1 5 1.32 4 Number of certified beds (n = 214) 119 26 333 45.46 120 NOTES: Among the nursing homes in Massachusetts that got matched to a worker at least once during the study time window (N=216); for each metric, we also list the number of nursing homes with available data for that metric. After removing 83 (0.9%) nursing-home-week records with missing values in either the CMS COVID-19 nursing home or CMS PBJ nurse staffing databases for 216 nursing homes in Massachusetts that participated in the centralized matching process (i.e., got matched to a worker at least once during the study time window), our resulting dataset had 9,637 nursing-home-week records. Except for the portal data, all other datasets used in our study are publicly available. We received an exemption from the IRB at Northeastern University for collecting the portal data. We kept all personal information associated with nursing homes confidential. Outcome variable We operationalized our outcome variable as a weekly staff-to-resident ratio (S/R ratio) for each nursing home; such ratio is frequently used for measuring staffing levels in nursing homes.19 We calculated it by dividing inferred weekly nurse staffing hours by the inferred number of residents in the same week for each nursing home, using the CMS PBJ daily nurse staffing database.17 Other nursing home characteristics We controlled for other nursing home characteristics that could affect the S/R ratio, including the weekly numbers of resident deaths due to COVID-19, weekly numbers of COVID-19 positive cases among staff, and weekly numbers of resident admissions (number of residents admitted or readmitted after being previously hospitalized and treated for COVID-19). Data analysis The performance of the centralized matching process is quantified by how much it impacts the number of nursing staff hired through that process. The most direct way to examine that would have been to collect accurate feedback from the nursing homes on how many additional nursing staff they hired as a result of receiving staff matches through this process. However, while the operational team has tried implementing several different approaches focused on collecting such feedback (such as calling the nursing homes after not receiving feedback and providing easier interfaces for nursing homes to provide accurate feedback on matches), none of these approaches were successful. Because we cannot observe that metric of the centralized matching process performance directly, we estimate it by comparing the results of two regression models: one that accounts for the nursing homes receiving weekly matches provided by the process, and another that approximates the setup in which nursing homes would need to seek nurses on their own. Specifically, we examined the relative effectiveness of the centralized matching process by comparing impacts of the weekly matches that the process provided for nursing homes with impacts of the weekly demand entries of nursing homes alone (approximating the setup in which nursing homes would need to seek nurses on their own), using two nursing-home-week fixedeffects regression models. In both models, we used lagged (by 1-4 weeks) values of the number of demand entries (Model 1) and the number of matches (Model 2) to estimate and compare their impacts on the S/R ratio of nursing homes. Results Out of 373 nursing homes that submitted data to the CMS,16 216 (58%) participated in the centralized matching process. These nursing homes vary on quality, as measured by CMS overall ratings19 (Min = 1, Max = 5, Mean = 3.27, Median = 3.5, SD = 1.37), and capacity, as measured by the number of beds (Min = 26, Max = 333, Mean = 119, Median = 120, SD = 45.46). Table 1 presents descriptive statistics for general characteristics of the participating nursing homes (available in the CMS20 and LTCFocus18 datasets), and Table 2 presents descriptive statistics for characteristics of the participating nursing homes that change weekly, at the nursing-home-week level.Table 2 Focal characteristics of participating nursing homes in Massachusetts at the weekly level. Table 2 Participating nursing homes Mean Min Max SD Median Number of daily residents 87.81 4.57 292.57 36.32 86.14 S/R ratio 3.64 0.04 13.51 0.86 3.52 Residents COVID-19 cases 0.46 0 122 2.37 0 Staff COVID-19 cases 0.48 0 39 1.49 0 Residents COVID-19 deaths 0.12 0 19 0.60 0 Demand quantities 3.98 0 225 12.09 0 Number of matched staff 2.83 0 116 6.40 0 NOTES: N=216, number of nursing homes in Massachusetts that got matched to a worker at least once during the study time window. During the study time window, the participating nursing homes have entered a demand for 38,522 nursing staff positions and received 10,581 matches. Performance of the centralized matching process Table 3 presents the results of our two nursing-home-week fixed effects regression models. We find a significant impact of the number of needed staff (demand entries) on the S/R ratio after three and four weeks, but not after one or two weeks (Model 1). In contrast, we find a significant impact of the number of matches driven by centralized matching process on the S/R ratio after one, two and three weeks (Model 2). Moreover, the effects sizes in Model 1 are smaller than in Model 2.Table 3 Associations of demand and number of matches with staff to resident ratio. Table 3Model 1 Model 2 Variable Lag Coeff. P value Variable Lag Coeff. P value Number of needed staff 1 -0.0003 (-0.0016, 0.0010) 0.574 Number of matches 1 0.0021* (-0.0004, 0.0046) 0.083 2 0.0006 (-0.0007, 0.0018) 0.148 2 0.0021 (-0.0004, 0.0046) 0.022 3 0.0012* (-0.0001, 0.0024) 0.004 3 0.0026 (0.0001, 0.0051) 0.034 4 0.0014 (0.0002, 0.0025) 0.02 4 0.0042 (0.0017, 0.0067) 0.165 Residents weekly COVID-19 deaths - 0.1019* (0.0860, 0.118) <0.001 Residents weekly COVID-19 deaths - 0.1016* (0.0857, 0.118) <0.001 Staff weekly COVID-19 cases - 0.0058 (0.0005, 0.0121) 0.239 Staff weekly COVID-19 cases - 0.0061 (-0.0002, 0.0123) 0.223 Residents weekly admissions COVID-19 - -0.0106 (-0.0157, -0.0055) 0.014 Residents weekly admissions COVID-19 - -0.0106 (-0.0156, -0.0055) 0.014 NOTES: Model 1 estimates the association between weekly staff to resident ratio and weekly number of staff that nursing homes in Massachusetts needed. Model 2 estimates the association between weekly staff to resident ratio and weekly number of staff that the central matching process provided for nursing homes in Massachusetts. Time lags are in weeks. We used nursing-home-week fixedeffects regression models. Coefficients show changes in daily staff to resident ratio by 1 unit change of each variable. p<0.1, p<0.05, **p<0.01. Discussion The number of nursing matches driven by centralized matching process was strongly and significantly associated with the S/R ratio of nursing homes when controlling for different measures of nursing homes. For instance, for a nursing home with 100 residents, each suggested match in a given week is associated with an increase of 0.21 staffing hours per day in the next week, 0.21 in two weeks, and 0.26 three weeks later. Overall, we found that the centralized matching process increased the improvement of the S/R ratio of nursing homes and helped them address the staff shortage issue faster. Our results suggest that developing and using a centralized platform for matching a severely limited number of available workers to nursing homes during public emergencies, such as the COVID-19 pandemic, can help nursing homes improve their S/R ratio. Furthermore, implementing a centralized matching process enables policymakers to observe real-time supply and demand levels at the nursing home level and promptly make necessary policy adjustments (such as modifying local hiring incentives). However, these benefits of having a centralized matching process in place could materialize only if the process is developed through a close collaboration of nursing homes, policymakers and the centralized matching process development team. Therefore, it is essential to design such a process prior to public health emergencies and assign teams to launch the process and communicate with staff and nursing homes. Moreover, since it is critical to access accurate real-time data for a robust centralized matching process performance, local policymakers may want to provide a secure, accessible, and incentive-compatible mechanism for nursing homes to share their demands for staff and hiring decisions. Finally, while in our study the nurses have not received any incentives connected with their online portal enrollment and participating in the centralized matching process, we hope future research will examine how various incentive schemes may interact with the process effectiveness. Study limitations Given that the CMS PBJ daily nurse staffing and the CMS COVID-19 nursing home databases have data on nursing homes only, we have restricted our analysis accordingly (while more LTCFs such as assisted living centers and rest homes in Massachusetts have also used the centralized matching process). Also, since the CMS COVID-19 nursing home database has weekly data only starting from the week ending on May 31, 2020, we have adjusted the starting date for our statistical analyses accordingly. Conclusion and implications This study shows how adopting a centralized matching process may expedite and increase improvement in the staff-to-resident ratio in nursing homes, compared with the setup in which nursing homes need to seek nurses on their own. Public policymakers may consider developing a similar process prior to public health emergencies and assigning teams to launch, run and monitor the process during the emergencies. Funding This material is based upon work partially supported by the National Science Foundation RAPID Grant No. 2038421. Acknowledgments We would like to thank Elizabeth Chen, Ph.D. MBA MPH, Secretary of Elder Affairs, Executive Office of Elder Affairs, MA, Patricia Yu, Ph.D. LCSW, Senior Director of Healthcare Workforce Policy, Executive Office of Health and Human Services, MA, and Leanne Winchester, MSc RN DCS, School of Nursing, University of Massachusetts Chan Medical School – Commonwealth Medicine, MA for their collaboration on establishing the centralized matching process. ==== Refs References 1 Abbasi J. “Abandoned” nursing homes continue to face critical supply and staff shortages as COVID-19 toll has mounted JAMA 324 2 2020 123 32525535 2 Quinton Sophie Staffing Nursing Homes was Hard Before the Pandemic. Now it’s Even Tougher 2020 Stateline, an initiative of The Pew Charitable Trusts https://pew.org/3bCwvwU 3 Harrington C Swan JH. Nursing home staffing, turnover, and case mix Med Care Res Rev 60 3 2003 366 392 12971234 4 Harrington C. Time to ensure sufficient nursing home staffing and eliminate inequities in care Gerontol Geriatr Med 7 3 2021 1 5 5 Lapane K Hughes C. Considering the employee point of view: perceptions of job satisfaction and stress among nursing staff in nursing homes J Am Med Dir Assoc 8 1 2007 8 13 17210497 6 Kirkham C Lesser B Special report: pandemic exposes systemic staffing problems at U.S. nursing homes Reuters 2020 https://www.reuters.com/article/us-health-coronavirus-nursinghomes-speci-idUSKBN23H1L9 7 AHCA, NCAL. 94% of all nursing homes still facing staffing shortages, new survey shows [Internet]. McKnight's Long-Term Care News. 2021. Available from:https://www.mcknights.com/news/94-of-all-nursing-homes-still-facing-staffing-shortages-new-survey-shows/ 8 Xu H Intrator O Bowblis JR. Shortages of staff in nursing homes during the COVID-19 pandemic: what are the driving factors? J Am Med Dir Assoc 21 10 2020 1371 1377 32981663 9 FACT SHEET: Protecting Seniors by Improving Safety and Quality of Care in the Nation's Nursing Homes [Internet]. The White House. 2022. Available from:https://www.whitehouse.gov/briefing-room/statements-releases/2022/02/28/fact-sheet-protecting-seniors-and-people-with-disabilities-by-improving-safety-and-quality-of-care-in-the-nations-nursing-homes/ 10 CMS. Toolkit on State Actions to Mitigate COVID-19 Prevalence in Nursing Homes. 2021. 11 Roth AE. The evolution of the labor market for medical interns and residents: a case study in game theory J Pol Econ 92 6 1984 991 1016 12 Roth AE, Sotomayor M. Two-sided matching. Handbook of Game Theory with Economic Applications. 1:485–541. 13 Abdulkadiroğlu A Sönmez T. School choice: a mechanism design approach Am Econ Rev 93 3 2003 729 747 14 Hitsch GJ Hortaçsu A Ariely D. Matching and sorting in online dating Am Econ Rev 100 1 2010 130 163 15 Zarei HR, GhanbarpourMamaghani M, Ergun O, Yu P, Winchester L, Chen E. Matching Medical Staff to Long Term Care Facilities to Respond to COVID-19 Outbreak. Submitted Rev Available Req. 2022 16 Centers for Medicare and Medicaid Services COVID-19 Nursing Home Data 2021 Centers for Medicare & Medicaid Services (CMS) Available from https://data.cms.gov/covid-19/covid-19-nursing-home-data 17 Centers for Medicare and Medicaid Services Payroll Based Journal Daily Nurse Staffing 2021 Centers for Medicare & Medicaid Services (CMS) Available from https://data.cms.gov/quality-of-care/payroll-based-journal-daily-nurse-staffing 18 Brown University School of Public Health. LTCfocus. Available from: https://ltcfocus.org. Accessed 9 July 2022. 19 McGarry BE Gandhi AD Grabowski DC Barnett ML. Larger nursing home staff size linked to higher number Of COVID-19 cases in 2020: study examines the relationship between staff size and COVID-19 cases in nursing homes and skilled nursing facilities Health Aff 40 8 2021 1261 1269 20 Centers for Medicare & Medicaid Services Provider Information 2021 Centers for Medicare & Medicaid Services (CMS) Available from https://data.cms.gov/provider-data/dataset/4pq5-n9py	36462228	PMC9709370	NO-CC CODE	2022-12-01 23:23:04	no	Geriatr Nurs. 2023 Nov 30 January-February; 49:89-93	utf-8	Geriatr Nurs	2,022	10.1016/j.gerinurse.2022.11.005	oa_other
==== Front Chem Heterocycl Compd (N Y) Chem Heterocycl Compd (N Y) Chemistry of Heterocyclic Compounds 0009-3122 1573-8353 Springer US New York 3132 10.1007/s10593-022-03132-4 Article The synthesis of ortho-stilbazoles (2-styrylpyridines) (microreview) Sorokin Saveliy P. [email protected] Ershov Oleg V. grid.411669.d 0000 0001 0664 3937 Chuvash State University named after I. Ulyanov, 15 Moskovsky Ave., Cheboksary, 428015 Russia 30 11 2022 13 27 7 2022 24 10 2022 © 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 microreview compiles methods for the synthesis of ortho-stilbazoles (2-styrylpyridines) described in the literature in 2017–2022. Depending on the synthons from which the target structure is formed, four main synthetic approaches can be distinguished: coupling reactions, Wittig reactions, condensation reactions, and pyridine ring formation reactions. ==== Body pmc Introduction Organic molecules based on the framework of orthostilbazole are widely employed in medicinal chemistry1 and materials chemistry.2 Such heterocycles exhibit high antioxidant,1a,b antitumor,1c and anti-inflammatory activity.1d The possibility of inhibiting the replication of the SARS-CoV-2 coronavirus,1e VEGFR-2 kinase,1f and Mur ligase1g has also been noted. Conjugated molecules based on ortho-stilbazole are used as chemosensors for the determination of Cr2O72– and MnO4–,2a CN–,2b F–, and AcO–2c anions, Hg2+,2d Zn2+,2e 2,4,6-trinitrophenol2f cations, as fluorescent probes and labels,2g as candidate materials for photonic devices and optical switches.2h Coupling reactions The coupling reactions are one of the principal methods for the synthesis of ortho-stilbazoles.1g,3 The Heck reaction based on the use of styrene and 2-bromopyridine can be provided as a classical example. A feature of the transformation is the use of ionic liquids.3a The [PAIM][NTf2]/[BMIM][X] (X = PF6, BF4) system also demonstrated its efficacy in promoting the original tandem Wittig–Suzuki reaction.3a Both transformations involving ionic liquids proceed stereoselectively with the formation of the E-isomers. Saveliy P. Sorokin was born in 1999 in Alikovo, Chuvashia, Russia. He recieved the BSc degree at the Chuvash State University named after I. Ulyanov in 2020, and the MSc degree in 2022. He currently works as a laboratory assistant in the Department of organic and pharmaceutical chemistry. His research interests include cyanosubstituted pyridines and pyridones, donoracceptor chromophores, design of heterocyclic fluorescent molecules. Oleg V. Ershov was born in 1975 in Petrovka, Odessa Oblast, Ukraine. He graduated from the Chuvash State University named after I. Ulyanov in 1997. He received the PhD degree in Chemistry in 2000. Currently, he serves as assistant professor at the Department of organic and pharmaceutical chemistry at the same University. His scientific interests include chemistry of polynitriles, heterocyclic fluorescent and biologically active compounds, donor-acceptor chromophores. Coupling reactions (continued) The elusive Z-isomer of ortho-stilbazole can be obtained by a one-pot synthesis.3b For this, the sequence of the Sonogashira coupling in the presence of a cobalt-palladium catalyst and the semihydrogenation of alkynes with borazane is employed. The original rhodium(III)-catalyzed Suzuki reaction of 1,3-dienes with arylboronic acids is accompanied by elimination of the styrene fragment and proceeds with excellent chemoselectivity.3c The conversion is carried out in the presence of silver oxide to regenerate the catalyst and return it to the catalytic cycle. The Wittig reaction The classical Wittig reaction remains one of the most accessible and versatile tool in the synthesis of styrylpyridines.1c,2f,4 One example of the application of the Wittig reaction in the synthesis of styrylpyridines is the reaction of benzaldehyde derivatives with 2-picolylphosphonium chloride in the presence of NaH.4a A method of direct Wittig olefination of alcohols was proposed. The reaction was carried out under aerobic conditions using air as an inexpensive and clean oxidizing agent. ortho-Stilbazoles were formed as a mixture of stereoisomers with a significant predominance of the E-isomer.4b A modified Wittig reaction, the Horner–Wadsworth–Emmons reaction, is also used for the synthesis of derivatives of ortho-stilbazole. In this case, the transformations proceed with the participation of pyridine-2-carbaldehyde and benzylphosphonate in the presence of a base.4c Condensation reactions Condensation reactions are one of the simplest approaches to the synthesis of ortho-stilbazoles.1a,b,5 For example, reactions between α-picolines and benzaldehyde derivatives are carried out in acetic anhydride.5a An innovative synthesis of 2-styrylpyridines from benzylamines and ortho-picolines in DMSO medium with the addition of HCl and iodine as an oxidizing agent was shown.5b These transformations were also carried out using methylpyridinium salts in n-butanol1b or a n-butanol–toluene mixture.1a Demethylation of salts of the compounds was carried out by heating to 210°С with anhydrous pyridinium chloride. In this case, the time of the condensation step was significantly reduced. Pyridine ring formation reactions Heterocyclization6 or trans-heterocyclization7 reactions are also widely employed for the preparation of orthostilbazole derivatives. A method of synthesis from N-vinyl-α,β-unsaturated nitrones in the presence of an iron catalyst and molecular sieves via cleavage of the N+–O– bond was presented.6a It was shown that oximes, close structural analogs of nitrones, can be used instead by reacting them with methyl acrylates.6b In this case, palladium(II) acetate with the ligand (2,6-bis-(adamantan-1-yl)oxypyridine) and a silver salt were used as the catalyst and as an oxidizing agent, respectively. An example of a trans-heterocyclization reaction is the conversion of isoxazoles.7a The copper catalyst initiates ring opening by cleaving the N–O bond. DMSO serves as a one-carbon synthon generating an active methylene group in the course of the reaction, which leads to the formation of two C–C bonds during the formation of the pyridine ring. The reaction of 1,2,4-triazine derivatives with norbornadiene proceeds according to the mechanism of the aza-Diels–Alder reaction.7b The transformations result in the elimination of a nitrogen molecule and formation of the pyridine ring at the expense of two carbon atoms of the diene. The study was supported by the Russian Science Foundation grant No. 22-13-00157, https://rscf.ru/project/22-13-00157. Translated from Khimiya Geterotsiklicheskikh Soedinenii, 2022, 58(11), 582–584 ==== Refs References 1. (a) Semenov, A. V.; Balakireva, O. I.; Tarasova, I. V.; Semenova, E. V.; Zulfugarov, P. K. Med. Chem. Res. 2020, 29, 1590. (b) Semenov, A. V.; Balakireva, O. I.; Tarasova, I. V.; Burtasov, A. A.; Semenova, E. V.; Petrov, P. S.; Minaeva, O. V.; Pyataev, N. A. Med. Chem. Res. 2018, 27, 1298. (c) Pugachev, M. V.; Pavelyev, R. S.; Nguyen, T. N. T.; Gabbasova, R. R.; Bulatov, T. M.; Iksanova, A. G.; Aljondi, B.; Bondar, O. V.; Grishaev, D. Yu., Yamaleeva, Z. R.; Kataeva, O. N.; Nikishova, T. V.; Balakin, K. V.; Shtyrlin, Y. G. Bioorg. Med. Chem. 2021, 30, 115957. (d) Chen, G.; Shan, W.; Wu, Y.; Ren, L.; Dong, J.; Ji, Z. Chem. Pharm. Bull. 2005, 53, 1587. (e) Li, Y.-Q.; Li, Z.-L.; Zhao, W.-J.; Wen, R.-X.; Meng, Q.-W.; Zeng, Y. Eur. J. Med. Chem. 2006, 41, 1084. (f) Sun, W.; Fang, S.; Yan, H. Med. Chem. Commun. 2018, 9, 1054. (g) Hrast, M.; Frlan, R.; Knez, D.; Zdovc, I.; Barreteau, H.; Gobec, S. Bioorg. Med. Chem. Lett. 2021, 40, 127966. 2. (a) Zhang, X.-D.; Zhao, Y.; Chen, K.; Jiang, Y.-F.; Sun, W.-Y. Chem.–Asian J. 2019, 14, 3620. (b) Guan, R.; Chen, H.; Cao, F.; Cao, D.; Deng, Y. Inorg. Chem. Commun. 2013, 38, 112. (c) Xie, P.; Guo F.; Zhang, D.; Zhang, L. Chin. J. Chem. 2011, 29, 1975. (d) Zho, H.; Sun, L.; Chen, W.; Tian, G.; Chen, Y.; Li, Y.; Su, J. Tetrahedron 2016, 72, 2300. (e) Gabr, M. T.; Pigge, F. C. Dalton Trans. 2016, 45, 14039. (f) Zhang, X.-D.; Hua, J.-A.; Guo, J.-H.; Zhao, Y.; Sun, W.-Y. J. Mater. Chem. C 2018, 6, 12623. (g) Wang, M.-Q.; Ren, G.-Y.; Zhao, S.; Lian, G.-C.; Chen, T.-T.; Ci, Y.; Li, H.-Y. Spectrochim. Acta A: Mol. Biomol. Spectrosc. 2018, 199, 441. (h) Senthil, K.; Kalainathan, S.; Kumar, A. R.; Aravindan, P. G. RSC Adv. 2014, 4, 56112. 3. (a) Savanur, H. M.; Kalkhambkar, R. G.; Laali, K .K. Appl. Catal. A: Gen. 2017, 543, 150. (b) Clauss, R.; Baweja, S.; Gelman, D.; Hey-Hawkins, E. Dalton Trans. 2022, 51, 1344. (c) Tan, G.; Das, M.; Maisuls, I.; Strassert, C. A.; Glorius, F. Angew. Chem., Int. Ed. 2021, 60, 15650. 4. (a) Tian, J.-J.; Yang, Z.-Y.; Liang, X.-S.; Liu, N.; Hu, C.-Y.; Tu, X.-S.; Li, X.; Wang, X.-C. Angew. Chem., Int. Ed. 2020, 59, 18452. (b) Li, Q.-Q.; Shah, Z.; Qu, J.-P.; Kang, Y.-B. J. Org. Chem. 2018, 83, 296. (c) Cao, C.; Zeng, Z.; Cao, C. J. Phys. Org. Chem. 2022, 35(4), e4319. 5. (a) Nguyen, T. B.; Nguyen, T. M.; Retailleau, P. Chem.–Eur. J. 2020, 26, 4682. (b) Sharma, R.; Abdullaha, M.; Bharate, S. B. J. Org. Chem. 2017, 82, 9786. 6. (a) Chen, C.-H.; Wu, Q.-Y.; Wei, C.; Liang, C.; Su, G.-F.; Mo, D.-L. Green Chem. 2018, 20, 2722. (b) Yamada, T.; Hashimoto, Y.; Tanaka, K., III; Morita, N.; Tamura, O. Org. Lett. 2021, 23, 1659. 7. (a) Kumar, P.; Kapur, M. Org. Lett. 2020, 22, 5855. (b) Khasanov, A. F.; Kopchuk, D. S.; Nikonov, I. L.; Taniya, O. S.; Kovalev, I. S.; Zyryanov, G. V.; Rusinov, V. L.; Chupakhin, O. N. Russ. Chem. Bull. 2021, 70, 999.	36467773	PMC9709373	NO-CC CODE	2022-12-08 23:16:03	no	Chem Heterocycl Compd (N Y). 2022 Nov 30; 58(11):582-584	utf-8	Chem Heterocycl Compd (N Y)	2,022	10.1007/s10593-022-03132-4	oa_other
==== Front Appl Intell (Dordr) Appl Intell (Dordr) Applied Intelligence (Dordrecht, Netherlands) 0924-669X 1573-7497 Springer US New York 4278 10.1007/s10489-022-04278-6 Article Front-end deep learning web apps development and deployment: a review http://orcid.org/0000-0002-7730-5465 Goh Hock-Ann [email protected] 1 Ho Chin-Kuan [email protected] 2 Abas Fazly Salleh [email protected] 1 1 grid.411865.f 0000 0000 8610 6308 Faculty of Engineering and Technology, Multimedia University, Jalan Ayer Keroh Lama, Bukit Beruang, 75450 Melaka Malaysia 2 grid.444468.e 0000 0004 6004 5032 Asia Pacific University of Technology and Innovation, Jalan Teknologi 5, Technology Park Malaysia, 57000 Kuala Lumpur, Malaysia 30 11 2022 123 17 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. Machine learning and deep learning models are commonly developed using programming languages such as Python, C++, or R and deployed as web apps delivered from a back-end server or as mobile apps installed from an app store. However, recently front-end technologies and JavaScript libraries, such as TensorFlow.js, have been introduced to make machine learning more accessible to researchers and end-users. Using JavaScript, TensorFlow.js can define, train, and run new or existing, pre-trained machine learning models entirely in the browser from the client-side, which improves the user experience through interaction while preserving privacy. Deep learning models deployed on front-end browsers must be small, have fast inference, and ideally be interactive in real-time. Therefore, the emphasis on development and deployment is different. This paper aims to review the development and deployment of these deep-learning web apps to raise awareness of the recent advancements and encourage more researchers to take advantage of this technology for their own work. First, the rationale behind the deployment stack (front-end, JavaScript, and TensorFlow.js) is discussed. Then, the development approach for obtaining deep learning models that are optimized and suitable for front-end deployment is then described. The article also provides current web applications divided into seven categories to show deep learning potential on the front end. These include web apps for deep learning playground, pose detection and gesture tracking, music and art creation, expression detection and facial recognition, video segmentation, image and signal analysis, healthcare diagnosis, recognition, and identification. Keywords Deep learning web apps TensorFlow.js Front-end deep learning Browser-based deep learning Client-side deep learning ==== Body pmcIntroduction Since the early 2010s, the discipline of deep learning has made astounding progress, resolving previously unsolved issues and generating intriguing new possibilities. Deep learning development is fueled by algorithmic breakthroughs enabled by a stack of open source tools, hardware advancements, greater availability of labeled data, and algorithmic advancements [17]. Deep learning (DL) models can be deployed on various platforms for real-world use after thorough validation and testing. One approach is to host the DL model on a server or in the cloud. This method allows developers to launch DL-enabled services by contacting API endpoints. Frameworks (for example, TF Serving1) and cloud platforms (for example, Google Cloud Machine Learning Engine2) may aid with this deployment. Apart from online cloud frameworks, the model can also be served on a smaller scale, using specialized back-end technologies such as Django and Flask. However, since these deployments need a cloud service or a dedicated computer to execute the model, these back-end deployment methods are prohibitively expensive. The introduction of deep learning libraries that specifically cater to client-side deployment has changed the landscape [78]. Previously, it was believed that computation-intensive deep learning models could only be effectively run on platforms with GPU support and that they could not be successfully deployed on browsers or mobile platforms owing to the client-side platforms’ limited processing power. However, this has changed. With the front-end option becoming available, the deployment process has shifted its attention to adapting deep learning models to the front-end platform. According to a study on the challenges of deep learning deployment [21], the query about deployment has grown in popularity, based on relevant postings from 2015 to 2019 on Stack Overflow, a developer-focused Q&A website. The questions about front-end deployments only surfaced in 2018 and remain relatively few. By 2019, the volume of users and queries about front-end deployment was only a quarter of back-end and mobile deployment. The researcher concurred with Ma et al. [78] that deep learning in browsers is still in its infancy. According to the study, it takes approximately 3 times longer to answer deployment questions than non-deployment questions (404.9 min vs. only 145.8 min). These findings show that issues surrounding the deployment of deep learning software are difficult to address, partially confirming a prior result in [4] that model deployment is the most challenging step of the machine learning life cycle. While deployments are the most time-consuming to resolve, browser-based or front-end deployments seem to have faster median response times. The median response time for server/cloud, mobile, and browser-related queries, respectively, was 428.5, 405.2, and 180.9 min [21]. These findings suggest that the front end may be an excellent choice for deep learning deployments. This paper aims to discuss and review the deployment stack for front-end deep learning and the development process that specifically targets this deployment. No other study has explicitly addressed and reviewed deep learning web apps development and deployment for the front end. In [29], however, a review of deep learning on mobile devices has been provided. In [109], a series of 52 review papers on deep learning are presented. The papers are grouped into topics, such as ‘computer vision, forecasting, image processing, adversarial cases, autonomous vehicles, natural language processing, recommender systems, and big data analytics’. These review papers focused on particular topics and not on model deployment. To close the gap, this paper reviews the development and deployment of deep learning models on the front-end, client-side browser. Due to the length of content, Table 1 presents the paper’s organization and section summary, which helps the reader’s navigation and highlights key points. Table 1 Paper organization, descriptions and highlights Section 1. Introduction Section 1 introduces front-end deep learning deployment as a viable alternative to back-end deployment. It also outlines the objective and highlights the content of the paper Section 2. Front-end deep learning deployment stack Section 2 discusses why the ideal technology stack for front-end deep learning deployment includes browsers (front end), JavaScript, and TensorFlow.js. It also includes the constraints and considerations to be taken when adopting the stack Section 2.1 Why not deploy deep learning on the client-side browser? Section 2.2 Why is JavaScript the language of choice? Section 2.3 Why is TensorFlow.js ideal for front-end deployment? Section 3. Front-end deep learning development approach Section 3 covers the most commonly used approaches to build deep learning models suitable for front-end deployment, namely reusing, customizing, converting, and optimizing. Constraints and considerations for each approach are also highlighted here. Section 3.1 Reusing pre-trained TensorFlow.js models Section 3.2 Customizing existing models using transfer learning Section 3.3 Converting Python models into TensorFlow.js models Section 3.4 Optimizing the model prior to deployment Section 4. Front-end deep learning web apps Section 4 provides samples of research work that leverages front-end deep learning web apps for development and deploying their work. These web apps are grouped into seven research categories Section 4.1 Deep-learning playground apps Section 4.2 Body pose detection and gesture tracking apps Section 4.3 Music and art creation apps Section 4.4 Facial expression detection and recognition apps Section 4.5 Video, image, and signal analysis apps Section 4.6 Scientific computation, diagnosis, and optimization apps Section 4.7 Classification and detection apps Section 5. Conclusions Section 5 wraps up the paper with a summary of the key points presented in the paper Front-end deep learning deployment stack Machine learning was previously restricted to only experts in research laboratories; today, it is widely accessible. Now, it is much simpler for machine learning experts and others to get started by leveraging popular front-end technologies. This section goes through the reasoning for developing real-world deep learning apps using the browser, JavaScript, and TensorFlow.js as the front-end stack. These technological stacks are built on top of one another to form an ecosystem for front-end deep learning deployment. Why not deploy deep learning on the client-side browser? Once trained, the machine-learning model should be deployed to generate predictions on actual data. It is meaningless to train a model if it will not be used [17]. Based on the benefits mentioned later, the front-end may be a viable deployment option. The conventional approach is deploying a deep learning model on the server backend and accessing it through complicated API calls sent via HTTP requests using JavaScript. Because of their semantics and requirements, correct and efficient use of machine learning APIs is challenging. Accuracy-performance tradeoffs must also be carefully considered. In [123], a research study of 360 apps that use Google or AWS cloud-based ML APIs was performed. It was found that 70% of these applications include API misuses that impair the functional, performance, or economic quality of the program. The mobile platform is also a potential deployment platform. However, creating mobile AI apps that are cross-platform compatible is not simple. It is incredibly challenging to develop and maintain versions for iOS and Android, as well as test and submit the mobile apps for distribution on an app store [78]. A taxonomy of 23 types of fault symptoms is given in [22], showing that deploying deep learning models on mobile devices covers a wide variety of problems. However, with advancements in front-end development, researchers now have the option of deploying on the front-end, which is much easier [21], thus migrating away from back-end data centers and toward clients. Front-end deep learning applications offer the following advantages: wide reach, reduced server costs, timely response, and privacy [17, 111]. Wide reach The web environment is the most commonly used deployment platform for applications, with a growing number of customers expecting to use machine learning web applications while surfing the web [16]. Reduced server cost Server cost is often an issue in deep learning deployments. Accelerating deep-learning models often requires GPU acceleration. Models implemented on Google Cloud or Amazon Web Services virtual machines with CUDA GPUs may be costly. The cost increases with traffic, as does scalability and server architectural complexity. A simplified deployment stack using the front-end lowers server expenses and developer concerns. For example, in [13], the front-end deployment method was selected over cloud deployment to save server costs. Timely response Running from the client-side is essential for low latency applications, such as real-time audio, image, and video data. Consider what happens if an image needs to be sent to the server for inference. With client-side inference, the data and processing remain on the device, reducing latency and connection issues. Due to the proximity of data, the front-end browser is the perfect platform for interaction and visualization while working with deep learning applications. In [19], for example, a real-time and interactive online experience is provided by creating models on the front-end, avoiding the size and latency limitations associated with conventional neural audio synthesis models. Data privacy Data privacy is a must for certain applications, such as health-related deep learning models inferred from medical data. Using just the client-side to execute the model effectively addresses this issue since no data is sent or kept outside the user’s device, guaranteeing the privacy of sensitive or personal data. For example, FreddieMeter3 is an artificial intelligence-powered singing challenge that determines how well a singer’s voice (timbre, pitch, and melody) resembles Freddie Mercury’s while protecting user privacy. Constraints and considerations The deployment of deep learning models to mobile devices and browsers has been discovered to have compatibility and dependability problems [45], with variations in prediction accuracy and performance between models trained on PC platforms and those deployed to mobile devices and browsers. To bridge the gap, software design tools for development and deployment on the target platform are suggested. The security of the deep learning model is one major drawback of using the browser to generate predictions. When the model is in the user’s possession, it may be saved and used for unintended purposes. A secure way of distribution would be to keep the model on the cloud or servers and offer services through APIs. Many production use cases, for example, require that the model be securely sent to the client so that it cannot be copied and used on other websites. This is especially true if the model has financial value. While traditional cybersecurity methods can safeguard deep learning models on back-end servers, deep learning models on mobile devices and browsers pose new security challenges. An adversarial attack on neural networks may cause misclassifications, and the deep learning model’s hidden states and outputs may be exploited to reconstruct user inputs, possibly compromising user privacy [59]. Thus, additional concerns regarding protecting models against attacks are needed to securely and discreetly deploy deep learning models. Why is JavaScript the language of choice? Python is the preferred language for most machine learning projects. JavaScript, however, may be more practical for browser-based machine learning web applications because of its ease of using front-end components. Although alternative front-end programming languages such as Typescript, CoffeeScript, and JQuery exist, these languages are compiled into JavaScript to be executed on the browser. The JavaScript engine exists in all browsers and is the one responsible for creating a responsive, interactive environment. The benefits of developing deep learning applications with JavaScript include universality, user interaction, no installation, and direct input access. Universality JavaScript has been the most widely used open-source programming language for nine years4 and is often considered a universal language for browsers [67]. It has a thriving ecosystem with broad client and server applications and has grown rapidly. Unlike backend web development, which uses Python, Java, C#, or PHP, frontend web development uses JavaScript. JavaScript allows users to execute code on almost any device, virtually unchanged while providing an intrinsically linked environment with direct access to different resources. Unlike native mobile apps, web apps may be deployed, regardless of the underlying hardware or operating system. User interaction JavaScript enhances the user experience by adding responsive, interactive components to web pages. It can update webpage content, animate graphics, and manage multimedia. This kind of user engagement is critical in deep learning applications. For example, in a semantic image search web app,5 an INCEPTIONV3 convolutional neural network is used to enhance user interactions. When a picture is chosen, the neural network examines the content of all the photos in the dataset and displays the top 15 most similar images. No installation The simplicity with which JavaScript code and static resources may be shared with the help of a Content Delivery Network (CDN) and performed without installation is one of the main advantages of the JavaScript ecosystem [17]. This feature eliminates potentially tedious and error-prone installation processes and potentially dangerous access control when installing new software. This is in contrast to the Python environment, where installation becomes difficult due to dependencies. Developers may also make use of a large variety of JavaScript APIs, such as the WebGL API. For example, in [43], DeepFrag,6 a 3D CNN model, has been deployed as a web app to allow broader usage by researchers to conduct optimization experiments in their browsers. This is done without submitting potentially proprietary structures to the server or installing any additional software. Another example is the Cell Profiler Analyst Web [9], which highlights that no installation or software update is necessary to use the tool. Direct input access Web browsers offer an extensive set of tools for managing and displaying text, audio, and visual data. These data, produced and made accessible to the client, may be used directly with permission in the browser by deep learning models with the user’s consent. Aside from these data sources, modern web browsers may access sensor data, providing exciting possibilities for many deep learning applications. For example, in [77], face recognition on the front end allowed developers to run models on the client-side, reducing latency and server load. Constraints and considerations Deploying deep learning models on browsers presents developers with unique programming difficulties, such as converting the models to the formats required by the target platforms. The key to deploying deep learning models is the interplay between deep learning knowledge and web development. As a result, deploying deep learning models requires developers to understand both disciplines, making this a complex process. Because machine learning development environments are incompatible with deployment environments, transferring them requires much time and effort to accommodate the new environment. The reason is that deep learning frameworks are often developed on powerful machines with GPUs, which speeds up training but may hinder inference during deployment on low-end devices. Web applications written in JavaScript are often challenging to create for developers that prefer Python, which may introduce additional challenges to overcome. To make matters worse, only a few publicly accessible deep learning packages and built-in functions for JavaScript are available to choose from compared to Python. This lack of library support may add to the complexity and lengthen development time. Why is TensorFlow.js ideal for front-end deployment? TensorFlow.js is a JavaScript library for creating machine learning models compatible with TensorFlow Python APIs [111], allowing it to benefit from its strengths. However, unlike TensorFlow Python APIs, TensorFlow.js may be easily shared and performed on any platform without installation, making it more powerful. TensorFlow.js also provides high-level modeling APIs that support all Keras layers [106], making it easy to deploy pre-trained Keras or TensorFlow models into the browser and run them with TensorFlow.js. Hardware acceleration Apart from offering the flexibility required for web-based machine learning applications, TensorFlow.js also performs well. Since the TensorFlow.js library is linked with the acceleration mechanism offered by current web browsers, the TensorFlow.js platform enables high-performance machine learning and numerical computing in JavaScript [17]. TensorFlow.js’s initial hardware acceleration method is WebGL [111]. Modern web browsers have WebGL APIs, which were initially intended to accelerate the rendering of 2D and 3D visuals in web pages. However, it has since been repurposed for parallel numerical computing in neural networks in TensorFlow.js [111]. This implies that the integrated graphics card may speed deep learning operations, eliminating the requirement for dedicated graphics cards, such as those from NVIDIA, needed by native deep learning frameworks. While WebGL-based acceleration is not on a par with native GPU acceleration, it nevertheless results in orders of magnitude speedups compared to a CPU-only backend. Tapping into these performance improvements enables real-time inference for complex tasks such as PoseNet extraction of body posture. TensorFlow.js launched the WebAssembly backend (WASM) in 2020.7 WASM is a cross-browser, portable assembly, and binary format optimized for the web that enables near-native code execution. Although WebGL is quicker than WASM for most models, WASM may outperform WebGL for small models because of the fixed overhead costs associated with WebGL shader execution. Thus, when models are tiny or when low-end devices lack WebGL support or have less capable GPUs, WASM is an excellent option used in place of JavaScript’s vanilla CPU and WebGL-accelerated backend. TensorFlow.js assigns a priority to each backend and automatically picks the one best supported in an environment. Presently, WebGL takes precedence over WASM and subsequently the vanilla JS backend.8 An explicit call is needed to use a particular backend. Apart from hardware acceleration through WebGL and WebAssembly, WebGPU is still in its experimental phase as of June 2022.9 Robust ecosystem The TensorFlow.js ecosystem supports all major processes [17], such as training and inference, serialization, deserialization, and conversion of models. It also supports a wide variety of environments [17, 106], including browsers, browser extensions, web servers (Node.js), cloud servers, mobile apps (ReactNative), desktop apps (Electron.js), platforms for app plugins, and single-board computers. It also has built-in functionality for data input and a visualization API [106]. For example in [95], a browser extension for online news credibility assessment tools10 was developed using a collection of interactive visualizations that explain the reasoning behind the automated credibility evaluation to the user. As another example, in [11], a neural network was developed and exported to the Arduino Nano Module for geometric object identification by their acoustic signatures, which attempts to replicate the way dolphins and bats perceive by using sonic waves. Alternative library for front-end deep learning Several JavaScript-based deep learning frameworks enable deep learning in browsers. In [78], deep learning in browsers was studied to evaluate how well these frameworks perform, but the work does not consider recent TensorFlow.js speed gains. They assessed seven JavaScript-based deep learning frameworks to see how far browsers have supported deep learning tasks and compared the performance of these frameworks on various deep learning tasks. They found ConvNetJS11 excels in both training and inference. TensorFlow.js is recommended in place of ConvNetJS, which is no longer being developed, because of its feature set and performance. Libraries such as ConvNetJS, Keras.js,12 and Mind13 are no longer supported, while WebDNN14 only supports inference tasks. For training tasks, Brain.js15 supports DNN, RNN, LSTM, and GRU, while Synaptic16 supports first-order or even second-order RNN. In [94], a series of deep learning frameworks tailored to the browser environment, such as TensorFlow.js, Brain.js, Keras.js, and ConvNet, are reviewed and compared for object detection. For fast inference with acceleration support, frameworks like TensorFlow.js and WebDNN were recommended. Integration of low-level programming (such as mathematical operations and data manipulation) with higher-level programming (such as deep learning model development, training, and execution in the browser) makes TensorFlow.js a more resilient framework. According to npmtrends.com, which monitors package download counts over time, TensorFlow.js is well ahead of other ML frameworks in terms of popularity. The other frameworks (Brain.js, ConvNet.js, synaptic, and WebDNN), including the Onnx.js and Neataptic.js libraries, have not yet reached 10% of TensorFlow.js’s weekly download rate of 75 thousand.17 Available high-level API extensions Several deep learning high-level libraries are built based on TensorFlow.js, which allows them to take advantage of its extensive library and hardware-accelerated inference. These libraries are highlighted in Table 2. The table shows the functionality and use cases of these libraries. Table 2 High-level libraries that are built based on TensorFlow.js Library Functionality and use cases ml5.js[a] This library offers artists, creative programmers, and students with high-level access to machine learning techniques and models. Examples of pre-trained models and use cases include object detection [94, 100, 132], detecting human postures [15, 80], generating text, drawing pictures, creating music, pitch recognition, and common English language word associations magenta.js[b] Using the pre-trained Magenta models in the browser, the API can generate music and art. It contains VAE and RNN models for musical note-based models, sketch drawing models (including SketchRNN), and image style transfer models (including Arbitrary Style Transfer). Examples of use cases include [19, 33, 35, 53, 102, 106] face-api.js[c] This is a high-level API for face detection, face landmark detection, face recognition, facial expression recognition, age estimation, and gender recognition. Examples of applications include [8, 30, 38, 46, 49, 60, 71, 75, 77] handtrack.js[d] This real-time hand detection library frames hand tracking as an object detection problem and predicts bounding boxes for the position of hands in an image using a trained convolutional neural network. Example works include [58] and [120] machinelearn.js[e] This is a library for machine learning algorithms, similar to scikit-learn from Python. An example can be found in [106] Danfo.js[f] Inspired by Pandas, it provides a high-performance, intuitive, and easy-to-use API for manipulating and processing structured data such as arrays, JSON Objects and Tensors a https://ml5js.org b https://magenta.tensorflow.org/ c https://github.com/justadudewhohacks/face-api.js/ d https://github.com/victordibia/handtrack.js/ e https://github.com/machinelearnjs/machinelearnjs/ f https://danfo.jsdata.org Constraints and considerations Although TensorFlow.js is a practical and accessible deep learning framework, it has certain drawbacks. TensorFlow.js is not a framework aimed at resource-constrained IoT and embedded devices [26]. TensorFlowLite, another library from the TensorFlow framework, is more appropriate. In TensorFlow.js, memory management is essential and tensors that are no longer required must be manually cleaned up by using the tidy() and dispose() functions [101, 111]. If they are not explicitly removed, the tensors will continue to use memory, resulting in memory leaks. The JavaScript engine in browser execution environment is also restricted and single-threaded. Thus, computationally intensive tasks may cause the UI to stall [111]. The asynchronous function should be used for computationally demanding operations [101]. These environmental issues only apply to the TensorFlow.js environment and must be considered. Front-end deep learning development approach Researchers often face new challenges and constraints when deploying deep learning models in the browser because these models may not have been explicitly developed for client-side execution. Furthermore, the high-performance back-end server with hardware acceleration such as the GPU differs from the front-end environment in the browser with limited resources. Because of these environmental differences, the emphasis has shifted to adapting models for effective deployment and improved user experience. When a model is deployed on the front-end, the user experience is a crucial aspect to consider. As a result, much thought goes into making models that are compact and can be executed quickly [17]. In fact, model loading performance exceeds inference task performance because the process of loading and warming up the deep learning model takes longer than performing the inference job [78]. For example, if the object detection model size is larger than 10MB, it will take a long time to load, slowing down the website considerably. Creating models from the ground up, particularly for large and complicated models, is not advisable for frontend deep learning. Because of the limitations of the browser environment, browser models must be small, have fast inference, ideally in real-time, and be as easy to train as possible [17]. However, decreasing the model’s size may cause reduced accuracy, but most models will work adequately well in a browser [45]. Generally, when accuracy is weighed against user experience, a minor loss of precision is acceptable, since the user experience is emphasized. Researchers often choose one of many options when confronted with implementing a deep learning challenge in the browser. The most straightforward and most commonly used approach is to use pre-trained TensorFlow.js models that are ready to deploy. A model that has previously solved a similar issue may also be reused by retraining the model using transfer learning to adapt it for its particular application. However, if an existing model is built in Python, conversion to a TensorFlow.js model is required. Before deployment, it is also critical to test and optimize the model. These points are addressed in more detail in the following sections. Reusing pre-trained TensorFlow.js models The creation of machine learning models, which are at the heart of AI software development, is not a simple task. Significant technical skills and resources are required to build, train, and deploy modern deep learning models. Special abilities in reading and comprehending professional AI literature are needed to apply deep learning algorithms. Training machine learning models also takes considerable resources. For instance, models that perform complicated tasks like image classification, object recognition, or text embedding need extensive calculations. The tasks also take a long time to train on large-scale datasets, using a significant amount of computing resources. The complexity of creating machine learning models drives the effective reuse of machine learning models. To help with model discovery and reuse, public machine learning package repositories that collect pre-trained models may be employed. Once suitable models are found, these models may also serve as a testing ground for researchers to see whether a concept is viable before delving further and developing a model of their own. Using models from the TensorFlow.js library TensorFlow.js includes a collection of Google’s pre-trained models. These pre-trained models, as shown in Table 3, are ready to execute tasks like object identification, picture segmentation, voice recognition, and text toxicity categorization. The models may be used directly or customized using transfer learning. TensorFlow.js pre-trained models may be classified according to whether they are object-utility models, face and pose models, or text and language models. Table 3 TensorFlow.js pre-trained models TensorFlow.js pre-trained models[a] and use cases Object-utility models mobilenet Trained using the ImageNet database, this image classification model has 1.2 million training images and 1,000 object classes. It is applied in [1, 42, 73, 89, 106, 129] coco-ssd This object detection model is trained on the COCO dataset and can classify items into 80 distinct categories deeplab This model can perform semantic segmentation, which is the process of comprehending a picture at the pixel level and then labeling each pixel by generating a two-dimensional tensor with class labels knn-classifier This is a tool for building a K-Nearest Neighbors classifier, commonly used with activations from another model, such as with transfer learning. Application of this model can be found in [73] Face and pose models blazeface This is a face detection model based on Single Shot Detector architecture that can detect one or more faces inside a photograph taken with a smartphone camera HandPose This lightweight machine learning pipeline is composed of two models: a palm detector and one that tracks the fingers on the hand-skeleton. It predicts 21 3D hand keypoints for each identified hand. The applications of this model can be found in [42, 116] Pose-detection This is a unified pose detection toolkit that makes use of one of three models (MoveNet, BlazePose, and PoseNet) to identify atypical postures and rapid body movements in real-time. MoveNet is a lightning-fast and highly accurate model that detects the body’s 17 key points; BlazePose can detect 33 keypoints, and PoseNet can detect multiple poses, each of which includes 17 key points. Works on Pose-detection can be found in [20, 32, 81, 85, 92, 96, 96, 98, 106, 116, 124] BodyPix This is a body segmentation model that segments 24 body components from a background image or video in real-time, and it also works for multiple people. Its design is based on either MobileNetV1, a smaller but less precise model, or ResNet50, a bigger but more precise model. Use cases on BodyPix can be found in [91], and [41] Text and language models speech-commands This is a speech commands recognition model used to categorize audio clips from the dataset of voice commands. It can recognize spoken instructions composed of basic isolated English words from a limited vocabulary universal-sentence-encode This model can compress text into a 512-dimensional embedding that can be used for natural language processing tasks like textual similarity and sentiment classification. Applications of this model can be found in [42, 101] toxicity This is a model trained using the civil comments dataset, which includes over 2 million comments, that can classify text as ‘Very poisonous’ to ‘Very healthy’. Use case for this model can be found in [42, 73] qna This is a BERT[b] natural language question-answering model using the SQuAD 2.0 dataset[c] that responds to queries about the content of a particular text excerpt a https://github.com/tensorflow/tfjs-models/tree/master/ b BERT, or Bidirectional Encoder Representations from Transformers, is a technique for pre-training language representations that achieves state-of-the-art performance on a broad range of Natural Language Processing tasks c The Stanford Question Answering Dataset, or SQuAD, is a reading comprehension dataset comprised of Wikipedia articles and a collection of question-answer pairings for each article Using models from online model repositories TensorFlow Hub18 is a massively scalable, open repository and library for reusable machine learning algorithms. The TensorFlow Hub included trained machine learning models that were ready for fine-tuning and could be deployed anywhere. The TensorFlow Hub provides a central location for searching and discovering hundreds of trained, ready-to-deploy models. With a few lines of code, the most recent trained models, such as BERT and Faster R-CNN, can be downloaded and reused. There are thousands of models available, with an increasing number in each of the four input domains: text, image, video, and audio. Apart from filtering by domain, models may be filtered by formats: TensorFlow.js, TFLite, coral, TF1, and TF2. Models may also be filtered by architectural type, including BERT, EfficientDet, Inception, MobileNet, ResNet, Transformer, and VGG-style. Another repository that is linked to their framework includes the PyTorch Hub,19 which presently includes packages that may be accessed through the PyTorch framework’s APIs. Other such repositories include Microsoft cognitive toolkit,20 Caffe/Caffe2 model zoo,21 and MXNet model zoo.22 Two repositories that are not connected to any specific framework are Model Zoo23 and ModelHub.24 Model Zoo curates and hosts a repository of open-source deep learning code and pre-trained models for various platforms and applications. The repositories also offer filtering capabilities to assist users in locating the models they need. TensorFlow, Keras, PyTorch, and Caffe are all included in this framework package. ModelHub is another self-contained deep learning model repository. It is a crowdsourced platform for scientific research that highlights current developments in deep learning applications and aims to encourage reproducible science. The following repositories focus on certain fields: SRZoo25 in [24] is a centralized repository for super-resolution tasks that collects super-resolution models in TensorFlow. ARBML26 in [6] showcased a series of TensorFlow.js NLP models trained for Arabic, which support the NLP pipeline development. EZ-MMLA toolkit27 in [51] was created as a website that makes it simple to access machine learning algorithms in TensorFlow.js for collecting a series of multimodal data streams. Constraints and considerations Using an open-source pre-trained model may be very simple and effortless. However, some models do not disclose how they were created, the dataset on which they were trained, or even the method employed. These “black box” models can be a problem, as they can lead to legal liability [42] when an explanation of why the model made a certain prediction is required. Customizing existing models using transfer learning Although TensorFlow.js is not suggested for intensive training, it is suitable for small-scale interactive learning and experimenting. By integrating available pre-trained models into an appropriate use case, the development process may be significantly accelerated. Transfer learning is a method that enables us to apply previously learned models to our unique use cases. It is repurposing a trained model for a second similar job. Transfer learning enables the combination of pre-trained models with customized training data. This implies that by simply adding custom samples, the functionality of a model may be leveraged without having to recreate everything from the start. Most projects use transfer learning to achieve the following recurring benefits: obtaining a solution with little data, getting a solution quicker, and reusing a time-tested model structure [101]. Transfer learning has the major advantage of using less training data to develop an effective model for new classes. Rather than performing time-consuming relearning, previously acquired features, weights, or biases may be transferred to another situation. Modern, state-of-the-art models often include millions of parameters and train slowly. Transfer learning simplifies this training process by reusing a model learned on one task for a second related task. For instance, if an image classification model has been trained on thousands of pictures, rather than starting from scratch, fresh, unique image samples may be merged with the pre-trained model to produce a new image classifier using transfer learning. This feature enables the user to quickly and easily create a more personalized classifier. Transfer learning has been applied in a variety of use cases for front-end deep learning. For example, in [118], a convolutional neural network (CNN) computer vision model was trained on approximately 10,000 pictures (5,500 snails and 5,100 cercariae). Because the image dataset was small, transfer learning was used by using seven pre-trained CNN models, where InceptionResNetV2 was found to be the best pre-trained model. After developing and training the CNN model, the best performing Keras model was converted to TensorFlow.js to be deployed. Other examples of work that applies transfer learning include [89] which did transfer learning on InceptionV3, Resnet50, MobileNet, and [32] on PoseNet. Transfer learning is accomplished by retraining selected parts of previously trained models [101]. This may be done by substituting new layers for the last layers of the pre-trained model and training the new, much smaller model on top of the original truncated model using freshly tailored data. For example, in [1], transfer learning is accomplished using TensorFlow’s MobileNet. The MobileNet architecture’s last completely linked layer has been deleted. To categorize the dataset, the higher-level characteristics are input into machine learning classifiers such as the Logistic Regression model, the Support Vector Machine model, and the Random Forest model. Because of the variety of model implementation types, there are also different ways to access them for transfer learning purposes. For instance, the downloaded bundle from TensorFlow Hub does not include the whole model but rather truncates it into a feature vector output that may be connected to exploit the model’s learned features. Fortunately, TensorFlow’s environment is sufficiently flexible to support transfer learning. Understanding the model and adequate planning on reusing the underlying model is required to implement this correctly. Constraints and considerations Two primary considerations must be addressed before initiating transfer learning. First, it is essential to validate the data’s quality. If the data used in training is of poor quality, the result of the training will be worthless [73]. For this to work, the model properties gained in the first task must also be transferrable. In short, the features should be suitable for both the first and second tasks. When the machine-learning challenge is unique, transfer learning is not suitable, and developing a model from scratch is best. However, in other instances, the problem is generic for which pre-trained models exist that either precisely fit the need or can fulfill the requirements with minimal modification. Sharing and repurposing deep learning models and resources is expected to continue to increase in popularity [17]. As modern deep-learning models are becoming more solid and broader, reusing pre-trained models for direct inference or transfer learning is becoming more practical and cost-efficient. Converting Python models into TensorFlow.js models While there are many open-source pre-trained models accessible online, most of these models are trained and available in TensorFlow and Keras Python formats, compared to TensorFlow.js open-source pre-trained models. These pre-trained Python models can actually be converted to TensorFlow.js for front-end deployment. For example, in [127], the model is trained in Python using the TensorFlow library and then transformed into a layer model using TensorFlow.js. The TensorFlow.js framework provides a converter tool28 that enables direct conversion of models trained using Python frameworks, such as TensorFlow and Keras, allowing for direct inference and transfer learning in web pages [73]. TensorFlow.js supports the following model formats: TensorFlow SavedModel, Keras model, and TensorFlow Hub module [106]. When the TensorFlow.js tools converted the model, they produced a JSON file containing the model’s information and a binary file holding the weights and biases. While conversion is workable for TensorFlow SavedModel and Keras models, the conversion will fail if the model contains operations that are not supported by TensorFlow.js and modification of the original model needs to be done. The full list of TensorFlow.js operations is available here.29 For example, CamaLeon [30] was trained in Keras before being converted to TensorFlow.js for browser-based inference. After many optimization rounds, the model was reported to have 485,817 parameters and a weight size of just 2MB when converted to TensorFlow.js. Other examples of conversions from Keras for usage in web applications are reported in [6, 88, 99]. In addition, models created using PyTorch may be used, but the models must go through an extra conversion before they can be used for inference with TensorFlow.js. It must first be converted to Keras or Onnx format before being converted to TensorFlow.js. Again, changes may be required if any operations are not supported because of differences in library support across models. For example, in [108], the conversion process was accomplished via the conversion of PyTorch to the Onnx standard and then to TensorFlow and TensorFlow.js. Constraints and considerations Converting models poses a challenge when the model is not compatible with the TensorFlow.js library and changes to the API calls need to be made. In [45], compatibility and reliability issues were highlighted, as well as accuracy loss in certain instances when moving and quantizing a deep learning model from one platform to another. They suggested implementing platform-agnostic deep learning solutions, particularly for mobile and browser platforms, to tackle the problem. Optimizing the model prior to deployment Before deploying TensorFlow.js models to production, testing and model optimization are strongly recommended. Optimizing the download and inference speeds is critical for the client-side deployment of TensorFlow.js models to succeed [17]. The TensorFlow.js converter supports both optimization methods:30 graph-model conversion to improve inference speed, and post-training weight quantization to reduce the model size. Inference speed is optimized via the use of graph optimization, which simplifies computation graphs and decreases the amount of computation needed for model inference. Model size is optimized through weight quantization, which reduces the model weight to a lower size. Weight quantification may not result in a significant decrease in forecast accuracy. In most instances, this phase has a minimal effect on total model correctness [45]. However, if accuracy falls, repeatedly omitting or adding tensors from quantization may help identify the tensors that cause the model’s accuracy to decrease after their values have been quantized. As a result, it is critical to ensure that the model keeps an appropriate level of accuracy following quantization. Optimization has been used successfully in several use cases. As an example, the original ssd_mobilenet_v2 _cocomodel COCO-SSD object identification model is 187.8 MB in size. In comparison with the original model, the TensorFlow.js version of the model is very lightweight and optimized for browser execution. The TensorFlow.js lite_mobilenet_v2 model is less than 1MB and has the quickest inference performance. As another example, in [53], by using post-training weight quantization, the downloaded weights are compressed, resulting in a 400KB payload size with no discernible loss of quality. By using dilated depth-wise separable convolutions and fusing procedures, the researchers decreased the model’s run-time for one harmonization from 40s to 2s. Research on model optimization Because models must be both compact and powerful, how to construct a neural network with the lowest possible size while still completing tasks with an accuracy equal to that of a larger neural network, is a prominent subject of research in deep learning. A method for expanding and describing neural net information in a model based on hierarchical choices is suggested in [40]. In addition, to identify performance bottlenecks and aid system design, a thorough characterization of the results for a selection of paradigmatic deep learning workloads was provided in [44]. To produce a smaller model and low-latency inference, context-aware pruning was introduced in [55] that takes into consideration latency, network state, and computational capability of the mobile device. A binary convolutional neural network was later introduced in [56]. The updated approach reduces ‘model size by 16x to 29x’ and reduces ‘end-to-end latency by 3x to 60x’ compared to existing methods. A progressive transfer framework for deep learning models is introduced in [76] by using the principle of transferring large image files over the web. The framework enables a deep learning model to be divided and progressively transferred in stages. Gradually, each part added builds a better model on the user device. Constraints and considerations While model optimization is beneficial, the best approach to guarantee that the model works effectively is to design it from the start with resource restrictions. This means avoiding excessively complicated designs and, where feasible, reducing the number of parameters or weights. A possible fallback alternative if the front end does not perform well would be to use back-end services. For example, in [53], support for Tensor Processing Units (TPU) on Google Cloud was also introduced, besides front-end inferences for preparation of large-scale deployment. Here, a speed test is run to see whether the user’s device can run the model in the browser. Otherwise, the requests were routed to distant servers. Front-end deep learning web apps Web browsers have drawn the interest of AI researchers because they offer a cross-platform computing target for deep learning on the client’s side. Privacy, accessibility, and low latency interactions are just a few advantages of using web browsers for deployment. The browser’s access to components, including the web camera, microphone, and accelerometer, allows for simple integration of deep learning models and sensor data. Because of this connectivity, user data may be kept on the device while maintaining user privacy, allowing personalized deep learning apps in fields like medicine and education. Deep learning improves many existing solutions while also bringing new practical applications. Sometimes, an entirely new area of application is introduced. A few applications of TensorFlow.js are available in TensorFlow.js’s gallery of applications.31 This section focuses on research that leverages front-end deep learning web apps. It explores the question of what front-end deep learning can do by providing a glimpse of the work that has been deployed on the front end. Instead of delving deeply into any one field, the section concentrates on how working on the front end may help solve user problems. The collected works have 18 different fields, and they are grouped into 7 categories based on the purpose of the web apps. To show how researchers construct front-end deep learning web applications, the section also provides links to resources, such as the website where it is deployed or the GitHub repository where the source code is located. To help navigate the broad list of diverse topics, Fig. 1 has been designed to outline and highlight the major groupings of applications that will be covered. For example, if gesture tracking is of interest, following the section number (Section 4.2) will lead to examples of what has been done and how web apps are used to achieve gesture tracking. Fig. 1 Examples of front-end deep learning web apps Deep-learning playground apps Education playground Teachable Machine,32 in [18], is a web app that enables non-programmers to train their machine learning classification models using videos, pictures, or sounds from their devices. To build a model, it uses transfer learning to uncover the trends and patterns within images or sound samples. The trained model can also be downloaded locally, eliminating the need to save and save huge files, datasets, or models on the cloud. Teachable Machine offers expandable panels for hyperparameter tweaking and model assessment visualizations for customers who desire greater control over model training. Two other attempts have been made to create online AI learning applications comparable to Google’s Teachable Machine. In [107], the aim is to develop a web-based learning application33 to assist novices in comprehending deep learning processes and resolving real-world issues. In [90], the aim is to create a web app for an open artificial intelligence platform that helps preprocess input data, trains artificial neural networks, and sets up future actions based on inference findings. For children’s education, an interactive online explanation using picture recognition is suggested in [100]. When prompted by the website’s instructional video, the user will take a picture of the item, and the website will identify it using a deep learning algorithm created using ml5.js. Visualization playground GAN Lab,34 in [66], is an interactive visual learning and experimentation app for Generative Adversarial Networks (GANs). Users may train GANs interactively and visually examine the model training process to grasp how GANs function during training and inference. In a follow-up study, an in-person observational study was conducted to investigate how GAN Lab is used and what users gain from it. Design considerations and difficulties associated with interactive teaching systems for deep learning have also been highlighted in [65]. The same research group presented CNN Explainer,35 in [126] and [125], an interactive visualization application that helps explore convolutional neural networks. This app imports the pre-trained Tiny VGG model and uses TensorFlow.js to calculate results in real-time. They also presented a review of visualization and visual analytics in deep learning research, using a human-centered, interrogative approach [52]. Anomagram36 is another interactive visualization app for studying deep learning models, and it looks at how an autoencoder model may be used for anomaly detection. The ECG5000 dataset was used for interactive training and testing, in which the autoencoder model predicts whether an ECG signal sample is normal or abnormal. Two other studies that introduced visualization for learning deep learning include TensorSpace,37 a framework for visually representing pre-trained neural network models in three dimensions, and LEGION,38 in [28], a graphical analytic app that enables users to compare and choose regression models that have been created either via hyperparameter tweaking or feature engineering. Development playground Milo39 in [97] is a web-based visual programming environment for data science. It provides an abstraction for language-specific implementations of machine learning principles using graphical blocks (Blocky40) and the creation of interactive visualizations. Similarly, DeepScratch,41 in [5], is a new Scratch programming language extension that adds strong language features to aid in the development and exploration of deep learning models. Likewise, Marcelle42 is another toolkit that enables interactive machine learning development by composing or customizing component-based machine learning workflows based on the Tensorflow.js library [39]. To employ gamification to facilitate deep learning, a series of tasks43 was presented in [105]. These tasks assess the human-interpretability of generative model representations, in which users change model parameters interactively to recreate target instances. Similarly, in [64], students were given 30 ‘half-baked’ artificial intelligence projects44 in the Snap! block language programming system to explore and improve. Body pose detection and gesture tracking apps Rehabilitation and monitoring To facilitate fall risk assessment and rehabilitation monitoring for the elderly, a web app was introduced in [98]. A similar home care monitoring system was reported in [20], where it can assess the quality of in-home postural alterations, such as posture conversion, body movement, and positional changes. In a similar work, a fall detector is proposed in [7] using convolutional neural networks and recurrent neural networks. The suggested technique is appropriate for real-time detection of a single individual’s fall in a controlled setting, even on low-end computers. Most recently, in [25], PoseNet was used for in-home rehabilitation where its skeletal tracking was used to identify and monitor patients’ angular motions as they conduct rehabilitation activities in front of a webcam. After patients have completed their rehabilitation activities, doctors may examine and assess the deviation rate of their angular motions across various days to find out their rate of recovery. Physical activity coaching In [104], web-based video annotations were presented that support multimodal annotations and can be used in a variety of scenarios, including dance rehearsals. Pose estimation is used to detect a human skeleton in video frames, giving the user an app for locating potential annotations. To make it an interactive tool, voice integration was later added by using the ML5 speech and sound classifier to provide human-computer interaction [103]. In another study, the PoseNet machine learning model from ml5.js was used to create an exercise tracking system that uses two cameras to track and assess the difficulty of an exercise while providing feedback [96]. According to them, a single RGB camera may miss certain essential information, and incorrect body movement during workouts may also be overlooked. A self-monitoring and coaching system for online squat fitness is also designed in [124]. In [32], a Digital Coaching System uses PoseNet to provide real-time feedback on exercise performance to the trainee by comparing keypoints from two video feeds, one from the trainer (using ResNet50) and one from the trainee (using MobileNet). The trainer’s video feed is pre-recorded and processed with a pre-trained pose estimation model to generate keypoints that will be compared to the keypoints from the trainee’s live video feed. An assessment of pre-trained CNN models for player detection and motion analysis in pre-recorded squash games was carried out in [15]. One of the ml5.js-based algorithms demonstrated the fastest inference by using a 28-layer deep MobileNetV1 architecture, with the least depth of the CNNs evaluated but the lowest accuracy across all threshold stages in all videos. To estimate human pose and identify anatomical points, a computer vision-based mobile tool is proposed in [82] for assessing human posture. Gesture tracking HeadbangZ,45 in [81], is a web-based game that shows a 3D gesture-based input using deep posture estimation and user interaction. Likewise, Scroobly46 and PoseAnimator47 map the user’s actual motion to their animations using Facemesh and PoseNet machine learning models. When the user moves, the machine learning algorithm changes the animation shown on the screen. TensorFlow.js pose estimation was also used in the development of Learn2Sign [92], which provides feedback for learning signed languages. In contrast, a real-time translation of American Sign Language (ASL) into text is available in [86]. Music and art creation apps Music creation Magenta.js [102] is an open-source toolkit with a simple JavaScript API that abstracts away technical complexities, allowing application developers to build new interfaces for generative models. While Magenta.js’s goal is far broader than music, the suite’s first package, @magenta/music, contains several state-of-the-art music-generating models. Trained on Magenta’s MusicVAE’s latent space, MidiMe,48 in [33], is a compact Variational Autoencoder (VAE) that enables artists to summarize the musical characteristics of the input and create new customized samples based on them. Again powered by Magenta.js, Bach Doodle,49 in [53], attempts to make music creation more accessible to amateurs and professionals alike. Users may compose their tune and have it harmonized in baroque-style counterpoint composition using an easy sheet music interface supported by a machine learning model capable of filling in arbitrarily incomplete scores. In addition, Tone Transfer,50 in [19], is an interactive online app that allows users to convert any audio input into various musical instruments using differentiable digital signal processing (DDSP) models. In a similar work, Rhythm VAE51 [122] is a system for exploring musical rhythms’ latent spaces. It uses minimal training data, which allows for quick customization and exploration by individual users. Likewise, JazzICat,52 in [72], a deep neural network LSTM model, shows how it may simulate a jazz improvisation scenario involving a human soloist and an artificial accompaniment in a real-time environment. In addition, Essentia.js,53 in [27], is a collection of publicly available pre-trained TensorFlow.js models for music-related tasks such as signal analysis, signal processing, and feature extraction. Art creation Using the results of picture recognition in Sketcher,54 a convolutional neural network model that recognizes drawings was implemented in [36], to offer human-like complimenting feedback. In contrast, a web app for collaborative sketching is presented in [35]. It leverages Magenta.js’ sketch-rnn to be both cooperative and adaptable to the actions of its human partner. During collaborations, this app may suggest logical extensions of incomplete drawings. Finally, also following earlier work, this time using Sketch-rnn and Sketch-pix2seq, an image-to-sequence VAE model is constructed to guide the user via an interactive GUI environment using conditionally produced lines [127]. Facial expression detection and recognition apps Expression detection A comparison study of two open-source emotion recognition software packages was performed under a range of lighting and distance conditions, in which they found that Face-api.js outperforms CLMTrackr with 64% vs. 28% average accuracy [8]. An intelligent environment capable of detecting students’ learning-related emotions was introduced in [99], allowing different interventions to be carried out automatically in response to students’ emotional states. In addition, an online report and dashboard were also suggested in [46] that offer instructors an assessment of students’ involvement and emotional states while using an online tutoring system. Likewise, a head pose app that predicts students’ real-time head direction and reacts to potential disengagement was also proposed in [128]. The app also includes a facial expression dashboard that detects students’ emotional states and delivers this information to instructors to help instructors assess student development and identify students who need more help. To assist with emotional healthcare monitoring during mental treatment, a video analysis app on facial expression and emotion visualization was created in [48] and [49]. Similarly, a WebRTC-based real-time video-conferencing application that can detect facial emotions by reading the participants’ facial expressions was presented in [34]. Interestingly, the relationship between human mental fatigue level as measured by bio-signals (i.e., eye blink data) and plant health was studied in [68]. Face-api.js is also used to conduct neuromarketing research by identifying respondents’ emotions after seeing advertisements [38]. Facial recognition A 68-point facial landmark predictor that is under 200kb in size and capable of matching the predicted 68-point landmarks with conventional face recognition landmarks has been presented in [77]. In addition, to prevent unauthorized access, a system was created that uses facial recognition and landmark detection techniques to perform identity verification and liveness detection tasks [75]. Similarly, a smart identification system is featured in [2] for video conferencing applications based on face-api.js, and a distributed hash table using blockchain technology. Video, image and signal analysis apps Video segmentation To transform the backdrop of the distant peer’s webcam feed into one that matches the receiver’s, a browser-based application in [30] uses UNet and TensorFlow.js to perform real-time machine vision. In a similar work, TensorFlow.js’s BodyPix model is used to segment the backdrop and the individual for enhancing remote user presence. This is done by replacing the distant user’s backdrop with a real-time acquired picture of the region behind the receiving side [41]. A real-time painting algorithm is proposed in [12] that can remove unwanted human objects from the video by performing segmentation on the video frame by frame using BodyPix and TensorFlow.js. Instead of altering the backdrop, Invisibility Cloak55 is a real-time person removal system that runs in the browser. To predict the areas and landmarks of the faces for real-time face augmentation, a pre-trained FaceMesh model was used in [115] to overlay various augmented reality filters and effects over the identified face regions. Interestingly, rather than adding or removing video regions of the face and body, a face-touching web application has been developed in [91]. By using BodyPix 2.0, the application accepts real-time input from the built-in camera and identifies face-touching using the intersection of the hands and facial areas, delivering a warning notice to the user. Image processing Front-end deep learning web apps have also been used to provide a centralized tool that enables individuals working in disparate places and industries to interact. An online Integrated Fingerprint Image system was created with web-based tools that can view, edit, apply pattern recognition, analyze, and interact with pictures in real-time [54]. The system can extract characteristics and discriminate between normal and abnormal regions of the image, while allowing for comments. In addition, preprocessing of the input picture has also been shown in [62] to significantly affect the quantity of input data that must be sent to the cloud service, which subsequently improves server-side processing. Cell Profiler Analyst Web (CPAW)56 in [9] allows users to examine image-based data and categorize complex biological traits via an interactive user interface, allowing for better accessibility, faster setup, and a simple workflow. During the active learning phase, cells are fetched to be trained and categorized by the machine learning classifier into their appropriate class using TensorFlow.js. MedSeg57 is another clinical web app that allows simple volume segmentation of organs, tissue and pathologies for radiological images. The segmentation of the images can be done manually or automatically using deep learning models. Signal analysis A virtual laboratory for EEG data analysis that enables data analysis, pre-processing, and model development was built with TensorFlow.js in [3]. Using real-time sensor measurements of rib-cage movement, a separate clinical study in [108] found that an LSTM network model can predict physical effort by comparing how much air a person breathes in a minute to their perceived workout intensity. In [70], using captured audio data during class observation, deep learning models were built to classify classroom activities. The CNN classifier model outperformed the KNN and Random Forest models in this study. Using vibrational signals, a real-time classification model for rolling bearing diagnostics based on MobileNet was built in [129] and tested on a range of fault types. In the study, the improved ReLU is superior to the standard ReLU activation function. Scientific computation, diagnosis and optimization apps Scientific computation MLitB [79] is the first prototype machine learning framework written completely in JavaScript without Tensorflow.js. It can conduct large-scale distributed computing with diverse classes of devices using Web browsers. JSDoop58 is a high-performance volunteer-based web computing library introduced in [84] that splits a problem into tasks and distributes the computation using various queues. TensorFlow.js is used as a proof-of-concept to train a recurrent neural network that produces or predicts the following letter of an input text. In a follow-up study [83], a federated learning system that is dynamic and adaptable has been implemented and evaluated to train common machine learning models. The system has been evaluated with up to 24 desktops working together via web browsers to train common machine learning models, while allowing users to join or leave the computation and keeping personal data stored locally. For exploratory data analysis of high-dimensional data, a T-distributed Stochastic Neighbor Embedding (t-SNE) algorithm59 is proposed in [93]. Their technique decreases the computational cost by orders of magnitude while maintaining or improving the accuracy of previous methods. In contrast, LatentMap [61], a GAN-based method, was developed to analyze the latent space of density maps. The method can be applied to any density map in spatiotemporal visualization. It may speed up front-end loading, complete or anticipate stream data information, and visualize missing data. DeepFrag is a deep learning application presented in [43] for lead optimization, an important step in earlystage drug development, which involves making chemical modifications to a small-molecule ligand to improve features such as binding affinity. In another study to help predict drug responsiveness, a deep learning model60 was used on unlabeled cell culture images in [23]. Transfer learning was applied using the MobileNetV2 architecture and converted to TensorFlow.js format before deployment in the browser. Healthcare diagnosis For healthcare diagnosis, front-end deep learning web apps have been used for a wide variety of use cases. ImJoy,61 in [88], is a versatile and open-source app that enables the broad reuse of deep learning methods and plugins for interactive image analysis and genomics. For instance, the Skin-Lesion-Analyzer plugin uses a deep convolutional network in TensorFlow.js to categorize images of skin into seven distinct kinds of (possibly malignant) skin lesions. Similarly, a web application was created in [113] that allows users to submit images and analyze them for skin melanoma lesions using a convolutional neural network previously trained on Google Teachable Machine. In [31], a classification of skin cancer lesions was presented using two different implementations, a basic CNN model and a transfer learning model implemented using ResNet50 pre-trained with ImageNet. The transfer learning model was found to give higher accuracy. To train the model for COVID-19 Chest X-ray Diagnosis, CustomVision from Microsoft Azure Cognitive Services was used and exported to TensorFlow.js in [14]. Similarly, to detect and monitor abnormal shapes of the skull in children, the Google Inception V3 model was trained using a transfer learning approach and converted to TensorFlow.js before deployment [117]. In [121], an ear infection classifier62 was implemented and trained with published otoscopic images, from which transfer learning on MobileNetV2 was found to outperform Inception-V3, ResNet-50, and InceptionResnet-V2. In [57], the viability of Google’s Teachable Machine was assessed against the diagnosis of tooth-marked tongue, a condition in which tooth traces develop on the tongue. In [110], to assess public perceptions of physical distance, social media content posted during the COVID-19 epidemic was categorized using gated recurrent unit GRU-based recurrent neural network models. In [13], an online platform that enables users’ annotations, promotes active learning, and offers model inference calculated immediately in the web browser is presented. For the case study, breast cancer tissue microarray images and COVID-19 computed tomography images were used. TensorFlow.js was used to deploy the model, which was originally built using Google Cloud AutoML. Lastly, in [37], a remote healthcare cyber-physical systems (CPS) application that employs TensorFlow.js to allow machine learning algorithms to work on collected data in a variety of applications, such as image recognition and audio classification, has been proposed. Bandwidth optimization Video streaming and cloud gaming services can be improved by reducing delays or latency between user input and feedback. An adaptive video streaming solution was suggested in [10], which includes bandwidth prediction and model auto-selection methods tailored especially for low-latency live streaming. In [114], Deep Reinforcement Learning-based algorithms were used for low-latency live video streaming adaptation that could choose the video rate at each request. A networking application was used to show how domain knowledge in networking can enhance the robustness of the systems for dynamic adaptive video streaming [131]. To compensate for the high latency of online video games, artificial neural networks have been introduced in [50] to predict the player’s next move and implement it in the game before receiving the actual user input. This effectively reduces the time required to process the feedback loop between player and game, thus improving user performance and user experience. Classification and detection apps Image classification Image classification has a wide range of applications. Here, the works are arranged chronologically. For image forensic analysis, a Deep Convolutional Neural Network with transfer learning was suggested in [1], where the model can accurately classify various classes of camera models using the Open-Image dataset. To reduce the computing load associated with image processing, a web application that enables users to evaluate Image Visual Quality (IVQ) and conduct image enhancement by comparing it to IVQ prediction was developed using neural networks and TensorFlow.js in [119]. A low-cost augmented reality system for pre-school books was described in [74], which uses convolutional neural networks (CNN) models to recognize pages and select which 3D graphics to display. By producing unique 3D views of pages under a range of lighting conditions and camera navigations on each page, the method claimed to require less time and effort to generate datasets. To automate the encoding of early seventeenth-century music prints, an online Optical Music Recognition (OMR63) system was created in [112]. The system can process pictures of written music and classify these instances using supervised learning with convolutional neural networks and TensorFlow.js. To analyze unsteady gas flows, a neural network was built to classify the shadowgraph picture dataset and identify images with shock waves [132]. A separate CNN model was also developed to perform the regression job and describe the location of shock waves. To classify medically essential snails and their parasitic equivalents, CNN models were employed in image identification tasks in [118], and the CNN’s classification performance was comparable to that of expert parasitologists. To support post-earthquake damage inspection and investigation, the CNN model was trained with a transfer learning approach using 1780 manually labeled images of structural damage in [87]. Of the six classic pre-trained models, MobileNet’s fine-tuned model proves to be superior to the others and is further developed as a web-based application. Object detection A dynamic object classification technique was developed and validated using low-quality webcam images in [63]. They examined several designs and discovered that although well-known baseline architectures such as Xception, DenseNet, and ResNet perform well on high-quality ImageNet datasets, they fall short on low-quality image datasets. MobileNet, however, works better with low-quality pictures. Similarly, a garbage classification system that can dynamically self-learn using real-time data has been created in [130] to assist consumers in classifying domestic waste correctly. An object-based sound feedback system was developed in [85] that uses the SSD-MobileNetV2 object detection algorithm to assist visually impaired individuals in comprehending their environment via sound production. In [89], for Indian sign language recognition, transfer learning was used where pre-trained MobileNet model output features were fed to a KNN classifier to identify actions and predict respective words. To distinguish chosen invasive plants from comparable non-invasive species, the TensorFlow.js’ MobileNets model was retrained using pictures of invasive plants and their characteristics in [69]. This is done in a variety of light situations and stages of the plant’s life cycle. Likewise, preliminary detection and sexing of cricket species was performed using a model generated by Google Teachable Machine that uses transfer learning on MobileNet [47]. Similarly, Plant AI64 is a web app that identifies diseases in plants. Conclusions The most challenging aspect of the deep learning process is deployment. However, front-end deployment is more straightforward than back-end server, and mobile app deployment [21]. In addition, front-end deployment works best when the model has to work interactively and access resources on the front end without installation while preserving data privacy and offering real-time inference speed. Although in-browser deep learning is still in its infancy, progress in front-end technologies has helped drive the development of deep learning web apps. This is especially true for TensorFlow.js and JavaScript in the browser, which have become a viable front-end stack for deep learning research and deployment. However, the front-end stack poses concerns on model security, IP issues, and potential attacks, all of which must be considered before deployment. Further studies on web security are needed to tackle these concerns. During front-end web development, it is also necessary to consider releasing used resources using garbage collection, calling, and releasing the CPU thread to improve performance and prevent blocking. The challenge is that front-end deep learning deployment needs both deep learning knowledge and web development (software engineering) know-how. Development for the front-end Deep learning differs from the back end. The model must be small, have fast inference, and user experience must be considered. A typical starting point for front-end development and deployment is to reuse an existing model from one of the many repositories available. However, background research on the model is recommended to prevent the black-box model issue. In addition, transfer learning may be used to extend or customize an existing model. However, compatibility between the current training task and the new task is needed so that the previous task can be transferable to the new task. If an existing model is converted, compatibility issues must be addressed by rewriting certain functions. Finally, optimization of inference speed and model size is critical before deployment to maximize user experience. To understand better how it could be used in a broad range of disciplines, the article also discusses reported efforts that use deep learning on the front end while using the principles summarized here. The article explores how web apps are used as playgrounds, with a focus on artificial intelligence learning and the creation of music and art. To improve human wellbeing, front-end deep learning provides interesting human-centered apps such as pose detection, gesture tracking, expression detection and facial recognition. In addition, the paper examines how deep learning web apps may assist in video, image and signal processing, scientific computation, healthcare diagnosis, bandwidth optimization, image classification and object detection in the real world. Finally, this paper provides many links and citations to examples and codebases, which may assist individuals interested in front-end deep learning in overcoming entrance hurdles. 1 https://www.tensorflow.org/tfx/guide/serving 2 https://cloud.google.com/gcp 3 https://experiments.withgoogle.com/freddiemeter 4 https://insights.stackoverflow.com/survey/2021 5 https://convnetplayground.fastforwardlabs.com/ 6 http://git.durrantlab.com/jdurrant/deepfrag-app 7 https://blog.tensorflow.org/2020/03/introducing-webassembly-backend-for-tensorflow-js.htmlhttps://blog.tensorflow.org/2020/03/introducing-webassembly-backend-for-tensorflow-js.html 8 https://www.tensorflow.org/js/guide/platform_environment 9 https://github.com/tensorflow/tfjs/tree/master/tfjs-backend-webgpu 10 https://github.com/piotrmp/credibilator 11 https://github.com/karpathy/convnetjs 12 https://github.com/transcranial/keras-js 13 https://github.com/stevenmiller888/mind 14 https://mil-tokyo.github.io/webdnn/ 15 https://brain.js.org/ 16 https://github.com/cazala/synaptic 17 https://www.npmtrends.com/@tensorflow/tfjs-vs-brain.js-vs-convnetjs-vs-onnxjs-vs-synaptic-vs-webdnn-vs-neataptic 18 https://tfhub.dev/ 19 https://pytorch.org/hub/ 20 https://www.microsoft.com/en-us/cognitive-toolkit/features/model-gallery/ 21 https://github.com/caffe2/models/ 22 https://mxnet.apache.org 23 https://modelzoo.co/ 24 http://modelhub.ai/ 25 https://github.com/idearibosome/srzoo 26 https://github.com/ARBML/ARBML 27 https://mmla.gse.harvard.edu/ 28 https://www.tensorflow.org/js/guide/conversion 29 https://github.com/tensorflow/tfjs/blob/master/tfjs-converter/docs/supported_ops.md 30 https://github.com/tensorflow/tfjs/tree/master/tfjs-converter 31 https://github.com/tensorflow/tfjs/blob/master/GALLERY.md 32 https://teachablemachine.withgoogle.com/ 33 http://ai.uol.de 34 https://github.com/poloclub/ganlab/ 35 https://github.com/poloclub/cnn-explainer 36 https://anomagram.fastforwardlabs.com/ 37 https://tensorspace.org/ 38 https://gtvalab.github.io/projects/legion.html 39 https://miloide.github.io/ 40 https://developers.google.com/blockly/ 41 https://github.com/Noufst/DeepScratch 42 https://github.com/marcellejs/marcelle 43 https://hreps.s3.amazonaws.com/quiz/manifest.html 44 https://ecraft2learn.github.io/ai/ 45 https://github.com/dermotte/headbangz 46 https://experiments.withgoogle.com/scroobly 47 https://github.com/yemount/pose-animator/ 48 https://midi-me.glitch.me/ 49 https://magenta.tensorflow.org/coconet 50 https://magenta.tensorflow.org/ddsp 51 https://github.com/vigliensoni/R-VAE-JS 52 https://github.com/kosmasK/JazzICat 53 https://mtg.github.io/essentia.js/ 54 https://zaidalyafeai.github.io/sketcher/ 55 https://github.com/jasonmayes/Real-Time-Person-Removal 56 https://mpsych.github.io/CellProfilerAnalystWeb/ 57 https://www.medseg.ai/ 58 https://github.com/jsdoop/ 59 https://github.com/tensorflow/tfjs-tsne 60 https://bioanalysis-79545.web.app/main/images/a549 61 https://imjoy.io/ 62 https://headneckml.com/tympanic.html 63 https://github.com/jjstoessel/IntelliOMR_2020_prototype_release 64 https://github.com/Rishit-dagli/Greenathon-Plant-AI Chin-Kuan Ho and Fazly Salleh Abas contributed equally to this work. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References 1. Al Banna MH, Ali Haider M, Al Nahian MJ et al (2019) Camera model identification using deep CNN and transfer learning approach. In: 2019 international conference on robotics, electrical and signal processing techniques (ICREST). IEEE, Dhaka, pp 626–630. 10.1109/ICREST.2019.8644194 2. Alizadeh M Andersson K Schelén O DHT- and blockchain-based smart identification for video conferencing Blockchain: Res Appl 2022 3 2 100,066 10.1016/j.bcra.2022.100066 3. Alphonse J Diwakar S Deploying a web-based electroencephalography data analysis virtual laboratory Procedia Comput Sci 2020 171 2420 2425 10.1016/j.procs.2020.04.261 4. Alshangiti M, Sapkota H, Murukannaiah PK et al (2019) Why is developing machine learning applications challenging? a study on stack overflow posts. In: 2019 ACM/IEEE international symposium on empirical software engineering and measurement (ESEM). IEEE, Porto de Galinhas, pp 1–11. 10.1109/ESEM.2019.8870187 5. Alturayeif N, Alturaief N, Alhathloul Z (2020) DeepScratch: scratch programming language extension for deep learning education. IJACSA 11(7). 10.14569/IJACSA.2020.0110777 6. Alyafeai Z, Al-Shaibani M (2020) ARBML: democritizing arabic natural language processing tools. In: Proceedings of second workshop for NLP open source software (NLP-OSS). Association for Computational Linguistics, Online, pp 8–13. 10.18653/v1/2020.nlposs-1.2 7. Apicella A, Snidaro L (2021) Deep neural networks for real-time remote fall detection. In: Del Bimbo A, Cucchiara R, Sclaroff S et al (eds) Pattern recognition. ICPR international workshops and challenges. Springer, Cham, pp 188–201. Lecture notes in computer science. 10.1007/978-3-030-68790-8_16 8. Aranha RV, Casaes AB, Nunes FLS (2020) Influence of environmental conditions in the performance of open-source software for facial expression recognition. In: Proceedings of the 19th Brazilian symposium on human factors in computing systems. ACM, Diamantina, pp 1–10. 10.1145/3424953.3426630 9. Baidak B, Hussain Y, Kelminson E et al (2021) CellProfiler analyst web (CPAW) - exploration, analysis, and classification of biological images on the web. In: 2021 IEEE visualization conference (VIS). IEEE, New Orleans, pp 131–135. 10.1109/VIS49827.2021.9623317 10. Bentaleb A Begen AC Harous S Data-driven bandwidth prediction models and automated model selection for low latency IEEE Trans Multimed 2021 23 2588 2601 10.1109/TMM.2020.3013387 11. Bertemes-Filho P Gandolphi de Almeida MP Acquisition and recognition of ultrasonic signatures using multi-layer neural network IJBSBE 2020 6 3 70 73 10.15406/ijbsbe.2020.06.00190 12. Bharathi Kannan B, Daniel A, Pandey DK et al (2021) Real-time person removal from video. In: Prateek M, Singh TP, Choudhury T et al (eds) Proceedings of international conference on machine intelligence and data science applications. Springer, Singapore, pp 295–298. Algorithms for intelligent systems. 10.1007/978-981-33-4087-9_26 13. Bhawsar PS Abubakar M Schmidt M Browser-based data annotation, active learning, and real-time distribution of artificial intelligence models: from tumor tissue microarrays to COVID-19 radiology J Pathol Inform 2021 12 1 38 10.4103/jpi.jpi_100_20 34760334 14. Borkowski A Using artificial intelligence for COVID-19 chest X-ray diagnosis Fed Pract 2020 37 9 398 404 10.12788/fp.0045 33029064 15. Brumann C Kukuk M Reinsberger C Evaluation of open-source and pre-trained deep convolutional neural networks suitable for player detection and motion analysis in squash Sensors 2021 21 13 4550 10.3390/s21134550 34283127 16. Cai CJ, Guo PJ (2019) Software developers learning machine learning: motivations, hurdles, and desires. In: 2019 IEEE symposium on visual languages and human-centric computing (VL/HCC). IEEE, Memphis, pp 25–34. 10.1109/VLHCC.2019.8818751 17. Cai S Bileschi S Nielsen ED Deep learning with JavaScript: neural networks in Tensorflow 2020 Shelter Island Js Manning Publications Co. 18. Carney M, Webster B, Alvarado I et al (2020) Teachable machine: approachable web-based tool for exploring machine learning classification. In: Extended abstracts of the 2020 CHI conference on human factors in computing systems. ACM, Honolulu, pp 1–8. 10.1145/3334480.3382839 19. Carney M, Li C, Toh E et al (2021) Tone transfer: in-browser interactive neural audio synthesis. In: Joint proceedings of the ACM IUI 2021 workshops, vol 2903. CEUR Workshop Proceedings, College Station 20. Chen S Saiki S Nakamura M Nonintrusive fine-grained home care monitoring: characterizing quality of in-home postural changes using bone-based human sensing Sensors 2020 20 20 5894 10.3390/s20205894 33081059 21. Chen Z, Cao Y, Liu Y et al (2020b) A comprehensive study on challenges in deploying deep learning based software. In: Proceedings of the 28th ACM joint meeting on european software engineering conference and symposium on the foundations of software engineering. ACM, Virtual Event USA, pp 750–762. 10.1145/3368089.3409759 22. Chen Z, Yao H, Lou Y et al (2021) An empirical study on deployment faults of deep learning based mobile applications. In: 2021 IEEE/ACM 43rd international conference on software engineering (ICSE). IEEE, Madrid, pp 674–685. 10.1109/ICSE43902.2021.00068 23. Cho K Choi ES Kim JH Numerical learning of deep features from drug-exposed cell images to calculate IC50 without staining Sci Rep 2022 12 1 6610 10.1038/s41598-022-10643-9 35459284 24. Choi JH, Kim JH, Lee JS (2020) Srzoo: an integrated repository for super-resolution using deep learning. In: ICASSP 2020–2020 IEEE international conference on acoustics, speech and signal processing (ICASSP). IEEE, Barcelona, pp 2508–2512. 10.1109/ICASSP40776.2020.9054533 25. Chua J Ong LY Leow MC Telehealth using PoseNet-based system for in-home rehabilitation Future Internet 2021 13 7 173 10.3390/fi13070173 26. Cornetta G Touhafi A Design and evaluation of a new machine learning framework for IoT and embedded devices Electronics 2021 10 5 600 10.3390/electronics10050600 27. Correya A Marcos-Fernández J Joglar-Ongay L Audio and music analysis on the web using essentia.js Trans Int Soc Music Inf Retr 2021 4 1 167 181 28. Das S, Endert A (2020) LEGION: visually compare modeling techniques for regression. In: 2020 visualization in data science (VDS). IEEE, Salt Lake City, pp 12–21. 10.1109/VDS51726.2020.00006 29. Deng Y (2019) Deep learning on mobile devices: a review. In: Mobile multimedia/image processing, security, and applications 2019, vol 10993. International Society for Optics and Photonics, Maryland, p 109930A. 10.1117/12.2518469 30. Denoue L, Carter S, Kim C (2019) CamaLeon: smart camera for conferencing in the wild. In: Proceedings of the 27th acm international conference on multimedia. ACM, Nice France, pp 1038–1040. 10.1145/3343031.3350583 31. Devarapalli DJ, Mavilla VSD, Karri SPR et al (2021) Classification of skin cancer lesions using deep neural networks and transfer learning. In: Saini HS, Sayal R, Govardhan A et al (eds) Innovations in computer science and engineering. Springer, Singapore, pp 259–268. Lecture notes in networks and systems. 10.1007/978-981-33-4543-0_28 32. Díaz RG Laamarti F El Saddik A DTCoach: your digital twin coach on the edge during COVID-19 and beyond IEEE Instrum Meas Mag 2021 24 6 22 28 10.1109/MIM.2021.9513635 33. Dinculescu M, Engel J, Roberts A (2019) MidiMe: personalizing a MusicVAE model with user data. In: Workshop on machine learning for creativity and design. NeurIPS, Vancouver 34. Eltenahy SAM (2021) Facial recognition and emotional expressions over video conferencing based on web real time communication and artificial intelligence. In: Hassanien AE, Darwish A, Abd El-Kader SM et al (eds) Enabling machine learning applications in data science. Springer, Singapore, pp 29–37. Algorithms for intelligent systems. 10.1007/978-981-33-6129-4_3 35. Fan JE, Dinculescu M, Ha D (2019) Collabdraw: an environment for collaborative sketching with an artificial agent. In: Proceedings of the 2019 on creativity and cognition. ACM, San Diego, pp 556–561. 10.1145/3325480.3326578 36. Fang Z, Paliyawan P, Thawonmas R et al (2019) Towards an angry-birds-like game system for promoting mental well-being of players using art-therapy-embedded procedural content generation. In: 2019 IEEE 8th global conference on consumer electronics (GCCE). IEEE, Osaka, pp 947–948. 10.1109/GCCE46687.2019.9015247 37. Fiaidhi J Mohammed S Virtual care for cyber– physical systems (VH_CPS): NODE-RED, community of practice and thick data analytics ecosystem Comput Commun 2021 170 84 94 10.1016/j.comcom.2021.01.029 38. Filipovic F, Despotovic-Zrakic M, Radenkovic B et al (2019) An application of artificial intelligence for detecting emotions in neuromarketing. In: 2019 international conference on artificial intelligence: applications and innovations (IC-AIAI). IEEE, Belgrade, pp 49–494. 10.1109/IC-AIAI48757.2019.00016 39. Françoise J, Caramiaux B, Sanchez T (2021) Marcelle: composing interactive machine learning workflows and interfaces. In: The 34th annual ACM symposium on user interface software and technology, UIST ’21. Association for Computing Machinery, New York, pp 39–53. 10.1145/3472749.3474734 40. Frosst N, Hinton G (2017) Distilling a neural network into a soft decision tree. In: Besold TR, Kutz O (eds) Proceedings of the first international workshop on comprehensibility and explanation in AI and ML 2017, CEUR workshop proceedings, vol 2071. CEUR, Bari 41. Furuya Y, Takashio K (2020) Telepresence robot blended with a real landscape and its impact on user experiences. In: 2020 29th IEEE international conference on robot and human interactive communication (RO-MAN). IEEE, Naples, pp 406–411. 10.1109/RO-MAN47096.2020.9223346 42. Gerard C Practical machine learning in JavaScript: TensorFlow.js for web developers 2021 Berkeley Apress 43. Green H Durrant JD DeepFrag: an open-source browser app for deep-learning lead optimization J Chem Inf Model 2021 61 6 2523 2529 10.1021/acs.jcim.1c00103 34029094 44. Guignard M, Schild M, Bederián CS et al (2018) Performance characterization of state-of-the-art deep learning workloads on an ibm” minsky” platform. In: Proceedings of the 51st Hawaii international conference on system sciences 45. Guo Q, Chen S, Xie X et al (2019) An empirical study towards characterizing deep learning development and deployment across different frameworks and platforms. In: 2019 34th IEEE/ACM international conference on automated software engineering (ASE). IEEE, San Diego, pp 810–822. 10.1109/ASE.2019.00080 46. Gupta A, Menon N, Lee W et al (2021) Affective teacher tools: affective class report card and dashboard. In: Roll I, McNamara D, Sosnovsky S et al (eds) Artificial intelligence in education, vol 12748. Springer International Publishing, Cham, pp 178–189. 10.1007/978-3-030-78292-4_15 47. Gupta YM, Homchan S (2021) Short communication: Insect detection using a machine learning model. Nusantara Biosci 13(1). 10.13057/nusbiosci/n130110 48. Hadjar H, Lange J, Vu B et al (2020) Video-based automated emotional monitoring in mental health care supported by a generic patient data management system. In: Gigliotta O, Ponticorvo M (eds) Proceedings of the second symposium on psychology-based technologies, CEUR workshop proceedings vol 2730. CEUR Workshop Proceedings, Naples 49. Hadjar H, Reis T, Bornschlegl MX et al (2021) Recognition and visualization of facial expression and emotion in healthcare. In: Reis T, Bornschlegl MX, Angelini M et al (eds) Advanced visual interfaces. Supporting artificial intelligence and big data applications, vol 12585. Springer, Cham, pp 109–124. 10.1007/978-3-030-68007-7_7 50. Halbhuber D Henze N Schwind V Increasing player performance and game experience in high latency systems Proc ACM Hum-Comput Interact 2021 5 CHI PLAY 1 20 10.1145/3474710 51. Hassan J, Leong J, Schneider B (2021) Multimodal data collection made easy: the EZ-MMLA toolkit: a data collection website that provides educators and researchers with easy access to multimodal data streams. In: LAK21: 11th international learning analytics and knowledge conference. ACM, Irvine, pp 579–585. 10.1145/3448139.3448201 52. Hohman F Kahng M Pienta R Visual analytics in deep learning: an interrogative survey for the next frontiers IEEE Trans Visual Comput Graphics 2019 25 8 2674 2693 10.1109/TVCG.2018.2843369 53. Huang A, Hawthorne C, Roberts A et al (2019a) Bach doodle: approachable music composition with machine learning at scale. In: Proceedings of the 20th international society for music information retrieval conference (ISMIR), pp 793–800 54. Huang CY Liu L Chen YL An online integrated fingerprint image system IJMLC 2019 9 1 51 56 10.18178/ijmlc.2019.9.1.764 55. Huang Y, Qiao X, Tang J et al (2020) DeepAdapter: a collaborative deep learning framework for the mobile web using context-aware network pruning. In: IEEE INFOCOM 2020 - IEEE conference on computer communications. IEEE, Toronto, pp 834–843. 10.1109/INFOCOM41043.2020.9155379 56. Huang Y, Qiao X, Ren P et al (2021) A lightweight collaborative deep neural network for the mobile web in edge cloud. IEEE Trans Mobile Comput:1–1. 10.1109/TMC.2020.3043051 57. Jeong H Feasibility study of google’s teachable machine in diagnosis of tooth-marked tongue J Dent Hyg Sci 2020 20 4 206 212 10.17135/jdhs.2020.20.4.206 58. Ionescu TB Adaptive simplex architecture for safe, real-time robot path planning Sensors 2021 21 8 2589 10.3390/s21082589 33917089 59. Isakov M, Gadepally V, Gettings KM et al (2019) Survey of attacks and defenses on edge-deployed neural networks. In: 2019 IEEE high performance extreme computing conference (HPEC). IEEE, Waltham, pp 1–8. 10.1109/HPEC.2019.8916519 60. Jayaswal R, Dixit M (2020) comparative analysis of human face recognition by traditional methods and deep learning in real-time environment. In: 2020 IEEE 9th international conference on communication systems and network technologies (CSNT). IEEE, Gwalior, pp 66–71. 10.1109/CSNT48778.2020.9115779 61. Jiang S Li C Wang L LatentMap: effective auto-encoding of density maps for spatiotemporal data visualizations Graphics and Visual Computing 2021 4 200,019 10.1016/j.gvc.2021.200019 62. Juranek L, Stastny J, Skorpil V et al (2019) Acceleration of server-side image processing by client-side pre-processing in web application environment. In: 2019 42nd international conference on telecommunications and signal processing (TSP). IEEE, Budapest, pp 127–130. 10.1109/TSP.2019.8768889 63. Kabir MM, Ohi AQ, Rahman MS et al (2020) An evolution of CNN object classifiers on low-resolution images. In: 2020 IEEE 17th international conference on smart communities: improving quality of life using ICT, IoT and AI (HONET). IEEE, Charlotte, pp 209–213. 10.1109/HONET50430.2020.9322661 64. Kahn K Winters N Learning by enhancing half-baked AI projects Künstl Intell 2021 35 2 201 205 10.1007/s13218-021-00732-8 65. Kahng M, Chau DHP (2020) How does visualization help people learn deep learning? evaluating GAN lab with observational study and log analysis. In: 2020 IEEE visualization conference (VIS). IEEE, Salt Lake City, pp 266–270. 10.1109/VIS47514.2020.00060 66. Kahng M Thorat N Chau DHP et al GAN Lab: understanding complex deep generative models using interactive visual experimentation IEEE Trans Visual Comput Graphics 2019 25 1 310 320 10.1109/TVCG.2018.2864500 67. Kanber B Hands-on machine learning with JavaScript: solve complex computational web problems using machine learning 2018 Birmingham Packt Publishing 68. Kanda M, Kunze K (2021) Tranquillity at home: designing plant-mediated interaction for fatigue assessment. In: Augmented humans conference, vol 2021. ACM, Rovaniemi, pp 292–294. 10.1145/3458709.3458978 69. Kasthurirathna D, Lokuge K, Mendis R et al (2020) Invasive plant detection and management platform. In: 2020 IEEE international conference on environment and electrical engineering and 2020 IEEE industrial and commercial power systems europe (EEEIC/I&CPS Europe). IEEE, Madrid, pp 1–6. 10.1109/EEEIC/ICPSEurope49358.2020.9160590 70. Khan MS (2020) Using convolutional neural networks for smart classroom observation. In: 2020 international conference on artificial intelligence in information and communication (ICAIIC). IEEE, Fukuoka, pp 608–612. 10.1109/ICAIIC48513.2020.9065260 71. Klym H (2020) Face detection using an implementation running in a web browser. In: 2020 IEEE 21st international conference on computational problems of electrical engineering (CPEE). IEEE, Pińczów, pp 1–4. 10.1109/CPEE50798.2020.9238754 72. Kritsis K Kylafi T Kaliakatsos-Papakostas M On the adaptability of recurrent neural networks for real-time jazz improvisation accompaniment Front Artif Intell 2021 3 508,727 10.3389/frai.2020.508727 73. Laborde G Learning TensorFlow.js. 2021 Sebastopol O’Reilly Media Inc. 74. Le H, Nguyen M, Nguyen Q et al (2020) Automatic data generation for deep learning model training of image classification used for augmented reality on pre-school books. In: 2020 international conference on multimedia analysis and pattern recognition (MAPR). IEEE, Ha Noi, pp 1–5. 10.1109/MAPR49794.2020.9237760 75. Lee DJ, Pan TY, Hu MC (2020) Design of identity recognition and liveness detection system for mobile phones. In: 2020 Indo – Taiwan 2nd international conference on computing, analytics and networks (Indo-Taiwan ICAN). IEEE, Rajpura, pp 113–118. 10.1109/Indo-TaiwanICAN48429.2020.9181332 76. Lee Y, Yun S, Kim Y et al (2021) Progressive transmission and inference of deep learning models. In: 2021 20th IEEE international conference on machine learning and applications (ICMLA). IEEE, Pasadena, pp 271–277. 10.1109/ICMLA52953.2021.00049 77. Li C (2019) Web front-end realtime face recognition based on TFJS. In: 2019 12th international congress on image and signal processing, biomedical engineering and informatics (CISP-BMEI). IEEE, Suzhou, pp 1–5. 10.1109/CISP-BMEI48845.2019.8965963 78. Ma Y, Xiang D, Zheng S et al (2019) Moving deep learning into web browser: how far can we go?. In: The world wide web conference, WWW ’19. ACM, New York, pp 1234–1244. 10.1145/3308558.3313639 79. Meeds E Hendriks R Al Faraby S MLitB: machine learning in the browser PeerJ Comput Sci 2015 1 e11 10.7717/peerj-cs.11 80. Milkes Espinosa S, Graves J, Towery J (2021) What the flock?: fostering collaborative active breaks for online education. In: Extended abstracts of the 2021 CHI conference on human factors in computing systems, vol 499. ACM, New York, pp 1–6 81. Moll P, Leibetseder A, Kletz S et al (2019) Alternative inputs for games and AR/VR applications: deep headbanging on the web. In: Proceedings of the 10th ACM multimedia systems conference. ACM, Amherst Massachusetts, pp 320–323. 10.1145/3304109.3323832 82. Moreira R Fialho R Teles AS A computer vision-based mobile tool for assessing human posture: a validation study Comput Methods Programs Biomed 2022 214 106,565 10.1016/j.cmpb.2021.106565 83. Morell JÁ Alba E Dynamic and adaptive fault-tolerant asynchronous federated learning using volunteer edge devices Futur Gener Comput Syst 2022 133 53 67 10.1016/j.future.2022.02.024 84. Morell JA Camero A Alba E JSDoop and TensorFlow.js: volunteer distributed web browser-based neural network training IEEE Access 2019 7 158,671 158,684 10.1109/ACCESS.2019.2950287 85. Nguyen H, Nguyen M, Nguyen Q et al (2020) Web-based object detection and sound feedback system for visually impaired people. In: 2020 international conference on multimedia analysis and pattern recognition (MAPR). IEEE, Ha Noi, pp 1–6. 10.1109/MAPR49794.2020.9237770 86. Njazi S (2021) Veritas: a sign language-to-text translator using machine learning and computer vision. In: 2021 the 4th international conference on computational intelligence and intelligent systems. ACM, Tokyo, pp 55–60. 10.1145/3507623.3507633 87. Ogunjinmi PD Park SS Kim B Rapid post-earthquake structural damage assessment using convolutional neural networks and transfer learning Sensors 2022 22 9 3471 10.3390/s22093471 35591163 88. Ouyang W Mueller F Hjelmare M ImJoy: an open-source computational platform for the deep learning era Nat Methods 2019 16 12 1199 1200 10.1038/s41592-019-0627-0 31780825 89. Ozarkar S, Chetwani R, Devare S et al (2020) AI for accessibility: virtual assistant for hearing impaired. In: 2020 11th international conference on computing, communication and networking technologies (ICCCNT). IEEE, Kharagpur, pp 1–7. 10.1109/ICCCNT49239.2020.9225392 90. Park HJ Lee K Implementation of an open artificial intelligence platform based on web and tensorflow J Inf Commun Converg Eng 2020 18 3 176 182 91. Patel S, Madhani H, Garg S et al (2021) An ai-based solution to reduce undesired face-touching as a precautionary measure for COVID-19. In: Chaubey N, Parikh S, Amin K (eds) Computing science, communication and security. Springer International Publishing, Cham, pp 30–45. Communications in computer and information science. 10.1007/978-3-030-76776-1_3 92. Paudyal P, Lee J, Kamzin A et al (2019) Learn2Sign: explainable ai for sign language learning. In: 2019 joint ACM IUI workshops, ACMIUI-WS, 2019, vol 2327. CEUR-WS, Los Angeles, p 7 93. Pezzotti N Thijssen J Mordvintsev A et al GPGPU linear complexity t-SNE optimization IEEE Trans Visual Comput Graphics 2020 26 1 1172 1181 10.1109/TVCG.2019.2934307 94. Pournaras X (2020) Deep learning on the web: state-of-the-art object detection using web-based client-side frameworks. In: 2020 11th international conference on information, intelligence, systems and applications (IISA). IEEE, Piraeus, pp 1–8. 10.1109/IISA50023.2020.9284358 95. Przybyła PSoto AJ When classification accuracy is not enough: explaining news credibility assessment Inf Process Manag 2021 58 5 102,653 10.1016/j.ipm.2021.102653 96. Ranasinghe I, Dantu R, Albert MV et al (2021) Cyber-Physiotherapy: rehabilitation to training. In: 2021 IFIP/IEEE international symposium on integrated network management (IM). IEEE, Bordeaux, pp 1054–1057 97. Rao A, Bihani A (2018) Milo: a visual programming environment for data science education. In: 2018 IEEE symposium on visual languages and human-centric computing (VL/HCC). IEEE, Lisbon, pp 211–215. 10.1109/VLHCC.2018.8506504 98. Rick SR, Bhaskaran S, Sun Y et al (2019) NeuroPose: geriatric rehabilitation in the home using a webcam and pose estimation. In: Proceedings of the 24th international conference on intelligent user interfaces: companion. ACM, Marina del Ray, pp 105–106. 10.1145/3308557.3308682 99. Ríos Félix JM, Zatarain Cabada R, Barrón Estrada ML et al (2020) An intelligent learning environment for computational thinking. CyS 24(3). 10.13053/cys-24-3-3480 100. Risal MF, Sukaridhoto S (2019) Web explainer for children’s education with image recognition based on deep learning. In: 2019 international electronics symposium (IES). IEEE, Surabaya, pp 406–410. 10.1109/ELECSYM.2019.8901627 101. Rivera JDDS Practical TensorFlow.js: deep learning in web app development 2020 Berkeley Apress 102. Roberts A, Hawthorne C, Simon I (2018) Magenta.js: a JavaScript API for augmenting creativity with deep learning. In: Joint workshop on machine learning for music (ICML) 103. Rodrigues R (2022) Interactive intelligent tools for creative processes using multimodal information. In: 27th international conference on intelligent user interfaces. ACM, Helsinki, pp 134–137. 10.1145/3490100.3516479 104. Rodrigues R, Madeira RN, Correia N et al (2019) Multimodal web based video annotator with real-time human pose estimation. In: Yin H, Camacho D, Tino P et al (eds) Intelligent data engineering and automated learning – IDEAL 2019. Springer International Publishing, Cham, pp 23–30. Lecture notes in computer science. 10.1007/978-3-030-33617-2_3 105. Ross A, Chen N, Hang EZ et al (2021) Evaluating the interpretability of generative models by interactive reconstruction. In: Proceedings of the 2021 CHI conference on human factors in computing systems. ACM, Yokohama, pp 1–15. 10.1145/3411764.3445296 106. Sasaki K Hands-on machine learning with TensorFlow. Js: a guide to building ml applications integrated with web technology using the TensorFlow.Js library 2019 Birmingham Packt Publishing 107. Schultze S, Gruenefeld U, Boll S (2020) Demystifying deep learning: a learning application for beginners to gain practical experience. In: Hansen C, Nürnberger A, Preim B (eds) Mensch und computer 2020 - workshopband. Gesellschaft für Informatik e.V., Bonn. 10.18420/muc2020-ws111-334 108. Sen S, Bernabé P, Husom EJB (2020) DeepVentilation: learning to predict physical effort from breathing. In: Proceedings of the twenty-ninth international joint conference on artificial intelligence. International Joint Conferences on Artificial Intelligence Organization, Yokohama, pp 5231–5233. 10.24963/ijcai.2020/753 109. Sengupta S Basak S Saikia P A review of deep learning with special emphasis on architectures, applications and recent trends Knowl-Based Syst 2020 194 105,596 10.1016/j.knosys.2020.105596 110. Sesagiri Raamkumar A Tan SG Wee HL Use of health belief model–based deep learning classifiers for COVID-19 social media content to examine public perceptions of physical distancing: model development and case study JMIR Public Health Surveill 2020 6 3 e20,493 10.2196/20493 111. Smilkov D, Thorat N, Assogba Y et al (2019) TensorFlow.Js: machine learning for the web and beyond. In: Proceedings of the 2nd SysML conference, Palo Alto 112. Stoessel J, Collins D (2020) Using optical music recognition to encode 17th-century music prints: the canonic works of Paolo Agostini (c.1583–1629) as a test case. In: 7th international conference on digital libraries for musicology. ACM, Montréal, pp 1–9. 10.1145/3424911.3425517 113. Stratulat-Diaconu A Cocu A Classifying skin moles using convolutional neural networks. The annals of “Dunarea de Jos” university of Galati Fascicle IX, Metallurgy Mater Sci 2020 43 2 9 13 114. Sun L, Zong T, Wang S et al (2021) Towards optimal low-latency live video streaming. IEEEACM Trans Netw:1–12. 10.1109/TNET.2021.3087625 115. Sun TR (2020) FaceAUG: a cross-platform application for real-time face augmentation in web browser. In: 2020 IEEE international conference on artificial intelligence and virtual reality (AIVR). IEEE, Utrecht, pp 290–293. 10.1109/AIVR50618.2020.00058 116. Suryadevara NK Beginning machine learning in the browser: quick-start guide to gait analysis with JavaScript and TensorFlow.js 2021 Berkeley Apress 117. Tabatabaei SAH Fischer P Wattendorf S Automatic detection and monitoring of abnormal skull shape in children with deformational plagiocephaly using deep learning Sci Rep 2021 11 1 17,970 10.1038/s41598-021-96821-7 33420128 118. Tallam K Liu ZYC Chamberlin AJ Identification of snails and schistosoma of medical importance via convolutional neural networks: a proof-of-concept application for human schistosomiasis Front Public Health 2021 9 900 10.3389/fpubh.2021.642895 119. Tsekhmystro R, Oliinyk V, Proskura G et al (2020) Web assembled benchmark for image visual quality assesment, prediction and improvement. In: 2020 IEEE 15th international conference on advanced trends in radioelectronics, telecommunications and computer engineering (TCSET). IEEE, Lviv-Slavske, pp 791–795. 10.1109/TCSET49122.2020.235543 120. Tsuji M Kubo H Jayasuriya S Touch sensing for a projected screen using slope disparity gating IEEE Access 2021 9 106,005 106,013 10.1109/ACCESS.2021.3099901 121. Tsutsumi K Goshtasbi K Risbud A A web-based deep learning model for automated diagnosis of otoscopic images Otol Neurotol 2021 42 9 e1382 10.1097/MAO.0000000000003210 34191783 122. Vigliensoni G, McCallum L, Fiebrink R (2020) Creating latent spaces for modern music genre rhythms using minimal training data. In: International conference on computational creativity (ICCC). Goldsmiths University of London, Coimbra 123. Wan C, Liu S, Hoffmann H et al (2021) Are machine learning cloud APIs used correctly?. In: 2021 IEEE/ACM 43rd international conference on software engineering (ICSE). IEEE, Madrid, pp 125–137. 10.1109/ICSE43902.2021.00024 124. Wang T Kamon M Okada S Design and evaluation of an online squat fitness system: lessons learned during the early COVID-19 pandemic in japan Front Digit Health 2021 3 55 10.3389/fdgth.2021.679630 125. Wang ZJ, Turko R, Shaikh O, et al (2020) CNN 101: interactive visual learning for convolutional neural networks. In: Extended abstracts of the 2020 CHI conference on human factors in computing systems, CHI EA ’20. ACM, New York, pp 1–7. 10.1145/3334480.3382899 126. Wang ZJ Turko R Shaikh O et al CNN explainer: learning convolutional neural networks with interactive visualization IEEE Trans Visual Comput Graphics 2021 27 2 1396 1406 10.1109/TVCG.2020.3030418 127. Wu SJ, Lin PS, Huang PC et al (2020) Variational-autoencoder-based environment for interactive sketch tutoring aiming for kids. In: 2020 2nd international workshop on artificial intelligence and education. ACM, Montreal, pp 12–17. 10.1145/3447490.3447493 128. Yu H, Gupta A, Lee W, et al (2021) Measuring and integrating facial expressions and head pose as indicators of engagement and affect in tutoring systems. In: Sottilare RA, Schwarz J (eds) Adaptive instructional systems. Adaptation strategies and methods. Springer International Publishing, Cham, pp 219–233. Lecture notes in computer science. 10.1007/978-3-030-77873-6_16 129. Yu W Lv P An end-to-end intelligent fault diagnosis application for rolling bearing based on MobileNet IEEE Access 2021 9 41,925 41,933 10.1109/ACCESS.2021.3065195 130. Zhaojie D, Chenjie Z, Jiajie W et al (2020). In: 2020 15th international conference on computer science & education (ICCSE). IEEE, Delft, pp 349–352. 10.1109/ICCSE49874.2020.9201690 131. Zheng Y, Chen H, Duan Q et al (2021) Leveraging domain knowledge for robust deep reinforcement learning in networking. In: IEEE INFOCOM 2021 - IEEE conference on computer communications. IEEE, BC, pp 1–10. 10.1109/INFOCOM42981.2021.9488863 132. Znamenskaya I, Doroshchenko I, Tatarenkova D (2020) Edge detection and machine learning approach to identify flow struc tures on schlieren and shadowgraph images. In: Proceedings of the 30th international conference on computer graphics and machine vision (GraphiCon 2020). CEUR Workshop Proceedings, Saint Petersburg, pp 15–1 to 15–14. 10.51130/graphicon-2020-2-3-15	36466774	PMC9709375	NO-CC CODE	2022-12-01 23:23:04	no	Appl Intell (Dordr). 2022 Nov 30;:1-23	utf-8	Appl Intell (Dordr)	2,022	10.1007/s10489-022-04278-6	oa_other
==== Front Rev Ind Organ Rev Ind Organ Review of Industrial Organization 0889-938X 1573-7160 Springer US New York 9890 10.1007/s11151-022-09890-x Article Recent Developments at DG Competition: 2021/2022 Buettner Thomas [email protected] 1 Coublucq Daniel [email protected] 2 Kotzeva Rossitza [email protected] 3 Sauri-Romero Lluis [email protected] 4 Régibeau Pierre [email protected] 5 1 grid.270680.b European Commission, MADO 17/025, 1049 Brussels, Belgium 2 grid.270680.b European Commission, BRU-J-79 06/208, 1049 Brussels, Belgium 3 grid.270680.b European Commission, MADO 17/006, 1049 Brussels, Belgium 4 grid.270680.b European Commission, MADO 17/042, 1049 Brussels, Belgium 5 grid.270680.b European Commission, MADO 17/026, 1049 Brussels, Belgium 30 11 2022 139 5 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. The Directorate General for Competition at the European Commission enforces competition law in the areas of antitrust, merger control, and State aid. After providing a general presentation of the role of the Chief Competition Economist’s team, this article surveys some of the main developments at the Directorate General for Competition over 2021/2022. In particular, the article reviews the new antitrust “Vertical Block Exemption Regulation” and “Vertical Guidelines”, the new “Guidelines on State aid for climate, environmental protection, and energy”, and the Veolia/Suez merger. Keywords Competition policy Antitrust Mergers State aid ==== Body pmc1. Introduction: The CET in 2021-2022 The Chief Competition Economist’s team (CET) is a group of about 30 economists who provide advice to the Commissioner (Executive Vice-President Margrethe Vestager) and to the Director General (Olivier Guersent) of DG Competition. This advice concerns ongoing cases, revisions of practices and guidelines, as well as broader policy issues (e.g., green policies, industrial policies, digital sector regulation). The CET does not just express an opinion on cases. Often some of its members are embedded in the case teams. This is generally the rule for mergers and has become much more common in antitrust and State aid—especially on the most relevant and complex cases. In addition, given the sheer number of State aid cases, the CET’s involvement in some of these cases is limited to performing specialised tasks and to vouching for the economic coherence of the analysis. The Fig. 1 below describes the allocation of resources across tasks over the last few years.Fig. 1 Workload across antitrust, merger and State aid (percentage of working hours) Overall, while merger control used to account for a large part of our time, the dedication of resources has been increasingly more balanced. Recently, a similar amount of resources have been devoted to antitrust, merger control, and State aid control. In the following sections, we summarize some of the main developments in our work over the last year: In antitrust, we outline the main changes in the new version of the Vertical Block Exemption Regulation (VBER) and the corresponding Vertical Guidelines. (Sect. 2). While it was a very busy year for State aid, we feel that the regulation of State aid that aims at achieving environmental objectives is worth special attention. Accordingly, with respect to the “Green Deal”, we discuss the new “Guidelines on State aid for climate, environmental protection, and energy” (CEEAG) (Sect. 3). Finally, Sect. 4 is devoted to a merger review—the Veolia/Suez case—where a catchment area analysis contributed to the assessment of the impact of the merger on local markets and bidding data were used to assess the closeness of competition between the merging parties and other suppliers. 2. Antitrust Introduction Between January 2021 and October 2022, the Commission took several interesting decisions in the antitrust area (outside cartels and merger control). We have already reported about the Aspen and Video Games cases in last year’s agency update (see Baltzopoulos et al., 2021).1 In addition, in Insurance Ireland, the Commission concluded that Insurance Ireland had breached EU antitrust rules by restricting competition in the Irish motor vehicle insurance market.2 The Commission considered that Insurance Ireland arbitrarily delayed—or in practice denied—access to its Insurance Link information exchange system, which contains information that is important in order to be active in the motor vehicle insurance market in Ireland. The Commission considered that restricted access to Insurance Link placed rival companies at a competitive disadvantage in comparison to incumbent companies that had access to the information exchange. This acted as a barrier to entry that reduced the possibility of more competitive prices and a greater choice of suppliers for consumers who seek motor vehicle insurance. In Czech Network Sharing, the Commission was concerned that a network-sharing agreement among several mobile telecommunications operators in the Czech Republic was designed in a way that would reduce infrastructure competition between them3: E.g., unilateral deployments and upgrades were charged by one party to the other in a way that reduced operators' individual incentives to invest. Moreover, the scope of the information that was exchanged among the parties went beyond what was necessary and included information that reduced the companies' incentives to compete. The parties remedied these issues through a broad set of commitments that addressed the Commission’s concerns. There were also many interesting developments on the policy front: The Commission adopted a new Vertical Block Exemption Regulation (VBER) that was accompanied by new Vertical Guidelines, which will be discussed in detail below.4 Moreover, the Commission published the draft for a revised Horizontal Block Exemption Regulation (HBER) accompanied by draft Horizontal Guidelines for public consultation.5 The Commission also adopted Guidelines on collective agreements by solo self-employed people and a revised Informal Guidance Notice.6 Moreover, the Commission launched a public consultation and evaluation of Regulation 1/2003 (the procedural framework that underpins the enforcement of EU antitrust rules).7 Finally, the Digital Markets Act (DMA) was adopted.8 There were also a significant number of important court judgments by the Court of Justice of the European Union in the antitrust area. In Intel (a referral back from the Court of Justice), the General Court found that the Commission’s original decision against Intel had not proven to the requisite legal standard that its loyalty rebates were likely to generate anticompetitive effects.9 This case is now on appeal at the Court of Justice again, as it concerns important questions with regard to the standard of proof in unilateral conduct cases—including the use of economic evidence. Also in Qualcomm Exclusivity, the General Court annulled a prior Commission decision about an alleged abuse of dominance.10 The decision was overturned both for procedural and substantive reasons—notably due to a failure to take into account all of the relevant factual circumstances in the Commission’s analysis of whether the payments were capable of having anticompetitive effects.11 On the other hand, the Commission prevailed in Google Shopping, where the General Court largely confirmed the Commission’s earlier decision.12 The Commission had argued that Google had engaged in self-preferencing of its own comparison shopping service over competing services.13 The Commission also prevailed in Google Android, where the General Court also largely confirmed the Commission’s earlier decision.14 In that case, the Commission had argued that Google had imposed unlawful restrictions on manufacturers of Android mobile devices and mobile network operators so as to consolidate the dominant position of its search engine.15 The Vertical Block Exemption Regulation and Vertical Guidelines On 10 May 2022, the Commission adopted a new Vertical Block Exemption Regulation (VBER)16 and new Guidelines on vertical restraints (VGL).17The VBER, which entered into force on 1 June 2022, and the VGL replaced their respective previous versions of 2010. The adoption concluded an extensive review process, which began in October 2018. The new VBER and VGL will remain in force for a period of 12 years. The rationale for the VBER and its structure remain generally unchanged: In contrast to restrictions in horizontal agreements between undertakings, restrictions in agreements between undertakings that are in a vertical relationship with one another—in particular restrictions in contracts between a supplier of goods or services and a buyer of such goods or services—are likely to lead to efficiencies that generally outweigh the possible negative effects of the restrictions on competition. For example, when a supplier restricts competition between its distributors in order to prevent them from free riding on each other’s sales efforts, this restriction may have an overall pro-competitive effect, as it may incentivise the distributors to invest more in sales efforts—such as pre-sales advice, customer service etc. The VBER creates a safe harbour for vertical agreements by companies without market power. It does so by exempting certain agreements between undertakings from the application of Art 101(1) TFEU (VBER Art 2). This provides legal certainty to undertakings with regard to the legality of their vertical agreements—without the need for a detailed case-specific assessment. However, this exemption is subject to important conditions: First, the VBER applies only where each of the undertakings that are party to the agreement has a market share not exceeding 30% of the relevant market: on the supply side and on the demand side of the market for the contracted goods or services (VBER Art 3).18 Hence, the VBER does not apply to the agreements of undertakings with significant market power. Second, vertical agreements that contain restrictions that are considered to have a particularly severe impact on competition—“hardcore restrictions”—fall outside of the block exemption in their entirety (VBER Art 4). The Commission applies a rebuttable presumption that such agreements are unlikely to give rise to efficiencies that fulfil the conditions of the exception provided by Art 101(3) TFEU. Third, when vertical agreements include any of the “excluded restrictions” set out in VBER Art 5, only the restraint in question falls outside the safe harbour19 and has to be assessed individually, with no presumption as to whether or not the restraint restricts competition within the meaning of Art 101(1) or fulfils the conditions of Art 101(3) TFEU. Finally, the Commission or the competition authorities of the Member States may withdraw the benefit of the VBER in individual cases where they find that one or more of the conditions of Art 101(3) are not met (VBER Art 6). The VGL provide detailed guidance on the application of the VBER and on the assessment of vertical restraints in individual cases that are outside the scope of the VBER. By design, the VBER and VGL have a limited duration and hence need to be reviewed periodically. The objectives of the latest review were to: (i) adjust the safe harbour in order to eliminate false positives and false negatives under the previous VBER, as identified via enforcement experience and through several stakeholder consultations; (ii) provide up-to-date rules and guidance to help businesses self-assess the compliance of their vertical agreements with Article 101 TFEU; and (iii) ensure a harmonised application of the antitrust rules that relate to vertical agreements across the EU, as well as to simplify and improve the clarity of the rules and guidance. Based on developments in the case law, case practice, and feedback from stakeholders, the new VBER contains several important changes: For example, changes to the rules on exclusive and selective distribution—as well as more detailed guidance on when online sales amount to active or passive selling—address “false negatives”.20 This gives firms greater flexibility in the design of their distribution systems and their use of vertical restraints to achieve efficiencies, without the need for a case-by-case assessment. Specifically, in the context of exclusive distribution systems,21 the VBER now allows the supplier to appoint up to five distributors for each exclusive territory/customer group and to require its buyers to pass on the restriction of active sales to their immediate customers. With regard to selective distribution systems,22 the supplier may now restrict all of its buyers—including those located outside the territory in which the selective distribution system is operated—from reselling the contract goods or services to unauthorised resellers that are located in that territory. An example of changes that are aimed at eliminating “false positives”23 concerns dual distribution, where a supplier sells to end customers both directly and through independent distributors. Evidence that was collected during the review process indicated that dual distribution has become more prevalent and may raise horizontal concerns—in particular related to information exchange. Therefore, while the dual distribution exemption is maintained and even extended to cover more levels of the supply chain, information exchange in dual distribution is exempted only where it is: (i) directly related to the implementation of the vertical agreement; and (ii) necessary to improve the production or distribution of the contract goods or services. Perhaps the most notable changes of approach are linked to the significant growth in e-commerce over the last decade and to the emergence or growth of new players, such as online platforms. At the time of the last review in 2010, online sales were less prevalent than they are today. The policy approach that was taken at the time was therefore aimed at supporting the growth of the online channel, which was expected to bring benefits to consumers and promote market integration. The previous VBER and VGL hence took a strict approach towards restrictions of online sales. For example, under the previous VBER, essentially all online sales were considered “passive sales”24 with the result that restrictions of such sales were generally considered as hardcore restrictions. Today, online sales have evolved into a well-functioning sales channel that no longer needs special protection relative to offline sales channels. Quite the opposite: The decline of traditional high street stores—and physical stores more generally—is now a greater concern. The Covid-19 pandemic has further exacerbated the shift from offline to online sales. As a result, the policy focus has now shifted from a need to promote online sales to a need to maintain an effective balance between online sales and offline sales through brick-and-mortar stores. Moreover, the growth of large online platforms has increased concerns about the exercise of market power by such platforms. The remainder of this section therefore focuses on changes in the new VBER and VGL in three important areas that relate to e-commerce: the assessment of online sales restrictions; the approach towards online platforms; and the treatment of parity clauses. Assessment of Online Sales Restrictions Under the new VBER, online sales restrictions that have as their object “the prevention of the effective use of the internet” (VBER Art 4(e)) constitute hardcore restrictions. As set out in the VGL, this will generally be the case for restrictions that: (i) require sales to be made only in a physical store; (ii) ban the use of the supplier’s brand on the distributor’s website; (iii) require the buyer to block website access to customers located outside the buyer’s territory or to re-route them; (iv) require the buyer to reject payments that are made with the use of foreign credit cards; (v) require the buyer to make a certain share of its total sales offline; or (vi) ban the use of entire online advertising channels: e.g., price-comparison websites or search engine advertising. By contrast, restrictions of online sales that fall short of a prevention of the effective use of the internet remain block-exempted. This includes, for example, restrictions that impose: (i) quality requirements; (ii) marketplace bans; (iii) restrictions on online advertising, including restrictions on the use of particular providers (except where the restriction amounts to a de facto ban on the use of the entire advertising channel); (iv) a requirement to operate an offline store; or (v) a minimum absolute volume of offline sales. This general framework applies irrespective of the type of distribution system that is operated by the supplier. Moreover, agreements must be assessed as a whole, as the combined use of several online sales restrictions—which individually may not prevent the effective use of the internet—may de facto achieve that result and hence amount to a hardcore restriction when they are used together. The VGL contain extensive guidance with regard to the application of this framework—in particular as regards the assessment of restrictions on the use of online marketplaces and price comparison websites (VGL Sects. 8.2.3 and 8.2.4). Beyond this general framework, and in light of the development of online sales, the approach towards a number of specific online sales restrictions has also been relaxed: In particular, a supplier may now, within the safe harbour, charge different wholesale prices to the same buyer depending on whether goods are to be sold online or offline—provided that such “dual pricing” does not have the object of preventing the effective use of the internet or restricting sales to particular territories or customers. Similarly, where a supplier operates a selective distribution system, the criteria that it imposes with respect to online sales no longer need to be “equivalent” to those that it imposes for offline sales—provided again, that the criteria do not indirectly have the object of preventing the effective use of the internet. Moreover, in the context of exclusive distribution, certain online sales are now considered “active sales” into an exclusive territory, and hence can be subject to restrictions: for example, if they originate from a website with a domain extension that is linked to the exclusive territory of another distributor, or if the website is in a language other than those commonly used in the distributor’s own territory.25 Compared to the previous VBER, these changes allow suppliers greater flexibility to apply restrictions of online sales, while still providing limiting principles that indicate when such restrictions will amount to hardcore restrictions. Treatment of Online Platforms Although the intermediation of sales through platforms is not a new phenomenon, the rapid growth of online platforms—such as online marketplaces or price comparison websites—over the past decade has raised the question of how the rules on vertical agreements should be applied to intermediation services that take place in an online context, and whether online platforms need to be subject to a specific set of rules. To tackle this issue, the VBER relies on the notion of “online intermediation services” (OIS), which are defined as information society services that facilitate direct transactions between undertakings that rely on such services or between such undertakings and final consumers (VBER Art 1(1)(e)).26 The direct consequence of falling within this OIS definition is that the platform will be categorised as a supplier of such OIS services and not as a buyer of the products that it intermediates. Whether or not the platform falls within the VBER’s market share threshold hence depends on the market share that is held by the platform in the relevant market for the provision of OIS—and not in the market for the intermediated goods or services. Another important consequence is that the VBER’s list of hardcore restrictions applies to restrictions that are imposed by the platform on buyers of its OIS—the sellers that use the platform to sell their products; but the hardcore list does not apply to restrictions that are imposed on the platform by those undertakings. For example, pursuant to Article 4(a) of the VBER, the safe harbour will not apply to an agreement under which a provider of OIS imposes a fixed or minimum sale price for a transaction that it facilitates. By contrast, restrictions that are imposed on a provider of OIS by buyers of the online intermediation services that relate to the price at which, the territories to which, or the customers to whom the intermediated products may be sold are block-exempted—since the provider of OIS does not buy or sell the intermediated products. For many platforms—including marketplaces and price comparison websites—this increases legal certainty as to how the rules of the VBER apply to their agreements. Platforms that do not meet the new OIS definition may be categorised as suppliers or buyers—depending on the particular facts of the case—as under the previous VBER. Another prevalent feature of many online platforms is that they also sell goods and services via the platform on their own behalf: “hybrid platforms”. In view of the competitive relationship between such hybrid platforms and the sellers that use them, the relative lack of enforcement experience more generally in relation to such hybrid platforms, and the fact that the Commission is only empowered to block-exempt agreements for which it has sufficient certainty that they generally meet the conditions of Article 101(3) TFEU, the Commission exercised a policy choice to exclude certain hybrid platform agreements from the safe harbour (VBER Art 2(6)). For the individual assessment of such platform agreements outside the VBER, any vertical restraints will be assessed using the approach that is set out in the VGL, while horizontal aspects—including possible collusive effects—will be assessed under the guidelines on horizontal co-operation agreements.27 At the same time, the VGL provide comfort to smaller hybrid platforms by making clear that their agreements will not be an enforcement priority if they do not contain by object restrictions28 and if the platform does not hold market power. This approach preserves resources for the scrutiny of agreements that are concluded by larger hybrid platforms while not hindering the emergence and growth of smaller platforms via a “hybrid platform” business model. Parity Clauses Another particularly hotly debated issue in recent years has been the use of parity clauses by online platforms. Such clauses typically take the form of platforms obliging sellers that rely on their OIS to not sell their goods or services at a lower price or better conditions: (i) via the seller’s own direct channels, such as their own website (“narrow” parity clauses); and/or (ii) via other indirect channels, such as other online platforms (“across-platform” parity clauses). In the previous VBER, all parity clauses fell within the safe harbour of the VBER (subject to the market-share thresholds). This included “retail parity clauses”: parity clauses that relate to the conditions under which goods/services can be sold to final consumers, as well as parity clauses that relate to sales between professional suppliers and buyers: e.g., “most-favoured customer” (MFC) clauses. At the same time, recent enforcement experience—in particular, cases by EU National Competition Authorities in which the VBER market share thresholds were exceeded, namely in the hotel booking sector29 —indicates that retail parity clauses cannot be assumed to generally meet the conditions of Article 101(3) TFEU. There is also a growing economic literature which emphasizes that (retail) parity clauses can have pro- as well as anti-competitive effects. Unfortunately, the economic literature and case experience to date remains relatively thin and typically deals only with the use of retail parity clauses by relatively large platforms: cases that do not fall under the VBER due to the market share thresholds. The economic effects of retail parity clauses are likely to depend on many factors, including, for example: the intensity of competition between platforms; the importance of the direct and indirect sales channels; the degree of free riding across channels; consumer search patterns; whether sellers are able to refuse parity clauses; whether sellers need to be present on all platforms; etc. However, the specific mechanisms and conditions under which retail parity clauses are likely to lead to pro- or anti-competitive effects are not yet sufficiently well understood. It is hence difficult to draw general conclusions with regard to the effects of retail parity clauses. More research and case experience is needed. The above considerations are reflected in the changes that were made in the new VBER with respect to retail parity clauses: First, across-platform retail parity clauses are now classified as excluded restrictions (VBER Art. 5), with the consequence that this type of parity clause (or measures that achieve the same result) will be subject to an individual assessment. This reflects the concern that across-platform parity clauses tend to eliminate the incentives for platforms to compete on the commissions or fees that they charge for their OIS, because reducing the commission rate is not attractive for the platform unless it results in lower retail prices for the goods or services that are offered via the platform relative to the prices that are offered via rival platforms. In this sense, across-platform retail parity clauses are more severe restrictions than are narrow retail parity clauses. Second, all other types of parity clause—including narrow retail parity clauses and non-retail parity clauses—remain block-exempted, which ensures continuity in this regard with the previous VBER. One reason for maintaining the block exemption for narrow retail parity clauses is that the risk of free riding on a platform’s investments via the direct channel appears greater than free riding between platforms, because consumer search costs for the direct channel are generally low and because suppliers do not need to pay OIS fees for sales on the direct channel. Third, as the conditions under which the use of narrow retail parity obligations by multiple platforms will lead to negative cumulative effects are not yet well understood, the VBER (Art 6(1)) and VGL explicitly warn that the benefit of the block exemption may be withdrawn in individual cases in concentrated platform markets where several platforms use narrow retail parity obligations. Overall, the approach to retail parity clauses strikes a reasonable balance between continuity and avoiding the risk of exempting anti-competitive agreements. State Aid control Introduction Between January 2021 and July 2022, the Commission continue to adopt a very large number of decisions in the area of State aid control. Most of those decisions concluded that the actions were compatible with the Commission’s criteria for justifiable actions or did not actually involve aid. During this period, the impact of the Covid-19 pandemic continued to be significant. Individual cases and schemes have been approved under the “temporary framework” (TF) that was adopted in March 2020 and amended several times thereafter. In addition, as part of the EU response to the crisis and its evolution, the “recovery and resilience facility” (RRF) for all Member States has been approved and its implementation has started, which has led to an increasing number of State aid cases that are financed through RRF funds. The CET has been closely involved in the assessment of TF and RRF cases. The aviation sector has been particularly affected since the onset of the Covid crisis. During this period the CET has contributed significantly to a number of cases in the field. Such cases have been under: the TF (e.g., Air France, Berlin airport); Article 107(2)(b) TFEU (damage compensation); the rescue and restructuring guidelines (such as TAP); or the market conform assessment of the establishment of the new company of ITA. In response to the Russian military aggression to Ukraine, the Commission adopted a “temporary crisis framework” (TCF), which includes in particular measures for: liquidity support; solvency support; as well as support for continued production by firms that are exposed to large increases in energy costs. The CET has been closely involved in the design and application of the TCF. In parallel to the numerous cases and policy initiatives related to the Covid crisis and to the Ukraine crisis, there has also been significant work on State aid guidelines. The Commission finalised and adopted the new “climate, environmental, and energy aid guidelines” (CEEAG) in January 2022 (which will be discussed in more detail in the remainder of this Section). The Commission’s work on the CEEAG has been in parallel with the continuous extensive work on schemes and individual cases in the energy sector—especially to support investment in renewable energy sources and improved energy efficiency. Revised guidelines on State aid that promote risk finance investments, “important projects of common European interest” (IPCEI), and regional aid have also been adopted. The Commission has continued working on the assessment of IPCEI—in particular on the development and adoption of hydrogen-based technologies. The CET has contributed to the assessment of these projects, especially by: identifying market failures that require State aid; reviewing the funding gaps of the projects; and assessing potential distortions of competition. Securing supply chains is a priority on the agenda of the Commission. One such important area concerns the Commission proposal of the EU Chips Act in February 2022. The CET has been working on the framework of assessment of open strategic autonomy under State aid rules and has been contributing to specific cases in the field of semiconductors. The European “Green Deal” Climate change and environmental degradation pose an existential threat to the European and global economy. To overcome these challenges, the EU has launched a wide-ranging set of policies to make the EU’s economy more resource-efficient; the goal is for no net emissions of greenhouse gases by 2050 and the development of a strategy for economic growth that is decoupled from ever-increasing resource use. The European Climate Law30 also sets the intermediate target of reducing net greenhouse gas emissions by at least 55% by 2030, compared to 1990 levels. Climate neutrality by 2050 means achieving net zero greenhouse gas emissions for EU countries as a whole, mainly by: cutting emissions; investing in green technologies; and protecting the natural environment. The law aims to ensure that all EU policies contribute to this goal and that all sectors of the economy and society play their part. The EU Institutions and Member States are bound to take the necessary measures at EU and national level to meet the targets, taking into account the importance of promoting fairness and solidarity among Member States. Key areas of intervention include:Cleaning the energy system by significantly reducing greenhouse gas emissions: requiring higher shares of renewable energy and greater energy efficiency. Decarbonising the industrial base by adopting cleaner technologies and products, particularly in certain sectors, including: energy, transport, construction, and energy-intensive users. Transitioning towards greener mobility, promoting the growth of the market for zero- and low-emissions vehicles and extending carbon pricing to road, air, and maritime transport. Improving the energy efficiency of buildings: both residential and commercial. The rationale for State Aid Achieving these ambitious targets will require: significant behavioural changes; transformation of many economic activities to meet new standards; and significant investment in innovation, technology adoption, adaptation of industrial processes and products, as well as the training of workforces. Markets are unlikely to trigger the necessary changes and deliver efficient outcomes in a timely manner by themselves due to the various market failures related to climate and environmental protection:Negative externalities arise when pollution is not adequately priced. Firms acting in their own private interest may therefore have insufficient incentives to take the negative externalities that arise from their economic activity into account—either when they choose a particular technology or when they decide on their output level. Positive externalities may occur, for instance, in the case of investments in eco-innovation, system stability, new and innovative renewable technologies, and innovative demand-response measures or—in the case of energy infrastructures or security of electricity supply—measures that benefit a wider number of firms and consumers. Asymmetric information can exist in relation to the likely returns and risks of climate and environmental projects. The current context of rapid technological and regulatory change creates uncertainty that may be distinctly perceived by different market players, which exacerbates the problem of asymmetric information. For instance, entrepreneurs, investors and lenders may have different access and ability to assess information with regard to the long-term prospects of investments in nascent technologies. Coordination failures may prevent the development of a project or its effective design due to: diverging interests and incentives among investors (“split incentives”; the costs of contracting or of liability insurance arrangements; uncertainty about the collaborative outcome; and network effects. Coordination failures may also stem from the need to reach a certain critical mass before it is commercially attractive to start a project—which may be a particularly relevant aspect in (cross-border) infrastructure projects. In the presence of such climate and environmental market failures, public intervention may improve the efficiency of market outcomes through: regulation, for instance establishing stricter standards; taxation, for instance by increasing its private cost for polluters; as well as subsidies that incentivise more environmentally sustainable conducts. Given the various forms of public intervention to address climate and environmental market failures, it is important that their design take into account the interaction and coherence between the various coexisting interventions. This is particularly relevant in relation to the EU emissions trading system (ETS), which is the cornerstone of the EU’s policy to reduce greenhouse gas emissions. The ETS contributes to internalising the cost of carbon emissions according to the “polluter pays principle” by establishing a market for carbon emission rights. While the system applies to a wide range of sectors and the overall quota of carbon emission rights has decreased progressively since its introduction in 2005, the system still does not cover many economic activities.31 Certain polluters are exempted from carbon emission costs in the ETS, and certain electricity-intensive consumers are compensated for the higher cost of their electricity consumption due to the carbon emission costs of electricity generation. This means that while the ETS contributes to the internalisation of carbon emission costs by market participants to a significant extent, there remain significant unaddressed residual externalities. Climate and environmental regulation, taxation, and State aid can therefore be seen as necessary complements to the ETS in correcting the climate and environmental market failures and achieving the decarbonisation targets that are set out in the European Green Deal. New Guidelines on State Aid for Climate, Environmental Protection, and Energy (CEEAG) State aid control in the climate, environmental, and energy fields aims precisely at ensuring that State aid measures are well-designed, in combination with coexisting climate and environmental regulations and policies, to address specific residual market failures and to enable competitive markets to produce more sustainable outcomes. The principles of necessity, incentive effect, proportionality, and avoidance of undue distortions to competition in the compatibility assessment of climate and environmental State aid are key—from an economic perspective—to ensure efficient intervention in the market. In January 2022 the Commission formally adopted the new Guidelines on State aid for climate, environmental protection, and energy (CEEAG). The new rules adapt the core principles of State aid control to the specificities of climate, environmental, and energy State aid in the very challenging context that is defined by the European Green Deal. The new rules aim at providing an economic and legal framework so as to enable the necessary public support to reach the European Green Deal targets in a cost-effective manner that minimises distortions to competition. The CEEAG set the general principles of assessment and include a broad section on aid for the reduction in greenhouse gases emissions, as well as specific sections on aid for energy efficiency buildings, clean mobility, infrastructure, circular economy, pollution reduction, and the protection and restoration of biodiversity, as well as measures to ensure the security of energy supply. State Aid to Address Residual Market Failures According to the CEEAG, State aid measures with a climate or environmental objective should address a well-defined objective in the form of a residual market failure that is not yet addressed by existing climate and environmental policy instruments.32 For instance, one could envisage granting a subsidy to a firm conditional on the adoption of a production technology with lower carbon emissions. If the adoption of such a technology is efficient from a climate or environmental perspective—given the reduction in carbon emissions that it can deliver relative to its additional costs—one should assess whether the existing climate and environmental policy instruments in place do not provide sufficient incentives for the firm to adopt the technology without the need of State aid support. These existing instruments may for instance include the ETS,33 which might not apply to the firm in question, or regulatory standards, which may fall short of what would be efficient to incentivise a climate or environmentally efficient behaviour from that firm. In this example, the limits of existing climate and environmental policy instruments to trigger the efficient adoption of more sustainable technologies by the firm would leave a residual market failure unaddressed, which would justify the need for the State aid measure. State aid should be used only if it is more likely to address efficiently the residual market failure than would alternative possible measures, such as climate and environmental standards or taxation.34 On the one hand, climate and environmental State aid measures should comply to the extent possible with the ‘polluter pays’ principle, by which the social cost of negative climate and environmental externalities are internalised by the market participant whose behaviour generates the negative external effect.35 On the other hand, even when State aid may be the most appropriate measure to address a residual market failure, the aid needs to take the most appropriate and least distortive form.36 For instance, if an environmentally beneficial investment is not taking place because the uncertainties about its prospects make access to private financing difficult, providing a public guarantee on a private loan may be less costly and less distortive than a grant. Incentivising Behavioural Change A climate or environmental State aid measure needs to incentivise the firm that receives support to change its behaviour in a way that is more sustainable for climate or from an environmental perspective37: The climate or environmental State aid measure has an incentive effect if it triggers a conduct with lower climate or environmental impact than the one that the firm would have pursued in the absence of State aid. Establishing the incentive effect of a State aid measure requires establishing credible factual and counterfactual scenarios, which are based on reliable evidence.38 While this does not necessarily entail a detailed quantification of the financial prospects in the factual and counterfactual scenarios, it does require establishing with sufficient certitude that the firms would likely act as described in the counterfactual scenario: complying with any climate and environmental regulations, but in the absence of the proposed State aid measure. Achieving Climate Targets in a Cost-Effective Way The comparison of the factual and counterfactual scenarios allows not just establishing whether the State aid measure actually incentivises a more sustainable conduct, but it also provides the starting point to quantify the minimum amount of State aid that is needed to trigger the change of conduct. This minimum necessary amount of State aid corresponds to the net extra cost to the firm of following the conduct in the factual scenario instead of the conduct in the counterfactual scenario. It is the “funding gap”: the difference between the net present values of the factual and counterfactual scenarios, excluding State aid, for the firm.39 The funding gap is at the same time the minimum amount of State aid that is needed for the measure to have an incentive effect on the firm and the maximum amount of State aid that is proportionate to the attainment of the climate or environmental objective of the State aid measure. Any amount of State aid in excess of the funding gap would not be justified by the need to trigger the desired change in the firm’s conduct and therefore would entail a degree of overcompensation. Amounts of State aid that are significantly below the funding gap may call into question the incentive effect of the measure, which suggests either a possible overestimation of the funding gap or the need for additional subsequent financial support to complete the project in the factual scenario. The direct assessment of a funding gap requires the identification of all relevant expected cash flows of the firm in the factual and counterfactual scenarios; hence this entails an economic and financial assessment that is based on detailed information and assumptions. The assumptions and parameters of the funding gap assessment normally include (amongst others): expected revenues and costs; capital expenditures; the expected economic life of relevant assets; the estimated weighted average cost of capital (WACC); and reasonable terminal values (for instance, based on Gordon growth model or other well-established asset valuation approaches). Importantly, whenever the projects relate to intermediate inputs of production processes, profit contributions may need to be allocated and synergies with complementary activities of the beneficiary may need to be incorporated into the analysis. Ultimately, the exercise intends to provide reliable economic estimates of the net present values for the firm in the factual and counterfactual scenarios, so as to quantify the resulting funding gap.40 Tendering, if properly designed to ensure effectively competitive bidding, provides a mechanism to reveal the funding gap of the marginal bidder and to select the firms that can more efficiently contribute to the attainment of the climate or environmental objective of the State aid measure, and thereby reduce the need for detailed information and assumptions ex ante. Therefore, whenever State aid is allocated through effectively competitive tenders, the CEEAG do not require a detailed direct assessment of the funding gap.41 In such cases, the proportionality assessment focuses on the proper design of the tender and the economic context in which it will take place, so that effective competition in the tender will ensure that the State aid that is allocated is the minimum necessary, and thereby reveals the underlying funding gap of the beneficiaries.42 With regard to the main design features of competitive tenders, the CEEAG prescribe that:“the bidding process is competitive, namely: it is open, clear, transparent and non-discriminatory, based on objective criteria, defined ex ante in accordance with the objective of the measure and minimising the risk of strategic bidding”; “the criteria are published sufficiently far in advance of the deadline for submitting applications to enable effective competition”; “the budget or volume related to the bidding process is a binding constraint in that it can be expected that not all bidders will receive aid, the expected number of bidders is sufficient to ensure effective competition, and the design of undersubscribed bidding processes during the implementation of a scheme is corrected to restore effective competition in the subsequent bidding processes or, failing that, as soon as appropriate”; “ex post adjustments to the bidding process outcome (such as subsequent negotiations on bid results or rationing) are avoided as they may undermine the efficiency of the process’s outcome”. These general requirements attempt to identify the basic conditions that are required to ensure the competitiveness of a tender and, therefore, the presumed proportionality of its outcome. Starting from these necessary conditions, the tendering design of each State aid measure requires an individualised assessment that takes into account the economic context of implementation. If a tender is not effectively competitive, it does not reveal the minimum amount of State aid that is required and consequently proportionality cannot be presumed. Administrative procedures that are structured formally as tenders—but that are not effectively competitive and do not lead to identifying the minimum State aid required—cannot be presumed to deliver proportional State aid. This is particularly relevant, for instance, whenever the risks of undersubscription emerge. Tenders that are expected by bidders to be undersubscribed are not competitive. When bidders are able to anticipate that the volume tendered is unrealistically high and that all eligible bidders will be successful, then they have an incentive to align their bids at the maximum level accepted by the auctioneer—the price cap—instead of bidding competitively. Undersubscribed tenders then become merely a formal procedure for administrative pricing at the level of bid caps, and are no longer effectively competitive tenders. This does not mean that a tender may not be occasionally undersubscribed and at the same time effectively competitive. To the extent that there is a significant degree of uncertainty about the effectively available supply at a given point in time, an occasionally undersubscribed tender may not require immediate corrective measures on its design. Corrections in the design are only needed when undersubscription is repeated or expected. Corrections may include, for instance, broadening the eligibility criteria to increase the potential number of bidders or adjusting tendered volumes downwards ex ante to reflect potential supply. These considerations are particularly relevant in view of ensuring a cost-effective path towards decarbonisation. Undersubscription of tenders increase the cost of State aid measures in support of decarbonisation without increasing the actual volumes that are contracted. Often, undersubscription signals the existence of exogenous constraints on the supply side that limit participation to the tenders in the short term. A cost-effective path towards decarbonisation therefore may require a combination of short-term measures to unlock potential supply with an ambitious increase of tendered volumes in the mid and long term—all of this while preserving the tenders’ competitiveness along the way. The CEEAG build on all of these facts, and require the correction of tender design so as to restore effective competition as soon as possible when undersubscription is identified—while remaining flexible about the precise ways of ensuring the competitiveness of tenders. Either through competitive tendering or by assessing directly the funding gap, the proportionality of climate and environmental State aid measures is key for ensuring their cost-effectiveness and the overall financial feasibility of climate and environmental State aid policies. Using Technologies Efficiently and Avoiding Undue Distortions to Competition Climate and environmental objectives can be reached efficiently only if two additional conditions are fulfilled: First, as is required in the CEEAG, possible interactions of a State aid measure with other instruments of climate and environmental policy that contribute to the same or related climate or environmental objective must be considered. This is especially important for interactions with the ETS. Second, one must avoid any unnecessary distortions to competition that—by undermining the efficient functioning of markets—would impair the efficacy of complementary climate and environmental policies. The proportionality of the State aid, especially when achieved through competitive tendering, provides a good starting point for minimising distortions to competition as it leads to the selection of the most efficient contributors to the achievement of the climate or environmental objective, and minimises the amount of State aid that is disbursed. However, by its very nature, any State aid measure will generate or threaten to generate some distortions of competition as the aid reinforces the competitive position of the beneficiaries.43 In essence, climate and environmental State aid interventions may also distort price signals and market structure, which will consolidate market power by incumbent firms and inefficiently alter the process of entry and exit into the market. These risks can be particularly relevant in energy and industrial sectors that: are large carbon emitters; will undergo radical processes of decarbonisation over the next years; are highly concentrated; and exhibit significant barriers to entry and exit. Risks of undesired distributive impacts in society and of subsidy races between Member States may also arise. Ultimately, the compatibility of a climate or environmental State aid measure with the EU State aid rules will require a balancing of its net contribution to the climate or environmental objective pursued against the competition distortions and welfare losses it may simultaneously cause. The CEEAG identify in a non-exhaustive way a number of possible distortions that would undermine competition and the efficiency of market outcomes:Undermining market rewards to the most efficient, innovative producers as well as undermining incentives for the least efficient producers to improve, restructure, or exit the market. This may also result in inefficient barriers to the entry of more efficient or innovative potential competitors. In the long term, such distortions may stifle innovation, efficiency, and the adoption of cleaner technologies.44 Strengthening or maintaining substantial market power of the beneficiary. Even where aid does not strengthen substantial market power directly, it may do so indirectly: by discouraging the expansion of existing competitors or inducing their exit or discouraging the entry of new competitors.45 Affecting trade and location choice. Those distortions can arise across countries—either when undertakings compete across borders or when they consider different locations for investment. Aid that is aimed at preserving economic activity in one region or attracting it away from other regions within the internal market may displace activities or investments from one region into another without any positive net climate or environmental impact.46 A way of mitigating the distortions to competition of a State aid measure is to ensure that all potential market participants that can contribute to the achievement of its climate or environmental objective compete on their relative merits to be selected for State aid support.47 This principle has often been referred to as “technology neutrality” of State aid measures: The design of the measure should not be arbitrarily biased in favour of certain beneficiaries. For instance, in principle, tenders in support of decarbonisation of electricity generation should be open to all low-carbon generation technologies to ensure that more cost-efficient approaches would be selected. However, technology neutrality does not mean negating objective differences across beneficiaries. For example, there are significant differences across electricity generation technologies in terms of costs, intermittency, or operating times. Even within the same technology, significant differences can exist depending on the location of the units, their size, or their access to the grid. These objective differences may justify technology specific designs or other restrictions in eligibility criteria. The CEEAG acknowledge this and allow for decarbonisation State aid measures that target specific technologies and economic activities, if reasons that are based on objective considerations are provided—in particular related to efficiency and cost heterogeneity.48 The choice between broader and narrower tender designs depends largely on informational asymmetries and uncertainties about the relative merits of various technologies. An auctioneer that has limited knowledge about the private characteristics of the various technologies—especially on current and future costs—may want to use a broader tender design so that the relative efficiency of the various bidders is revealed through their bids. The cost of learning from a broader tender takes the form of higher infra-marginal rents. The more accurate is the knowledge that the auctioneer has about the features of distinct technologies, the better positioned it is to design separate tenders that reflect their specific cost structure and reduce infra-marginal rents. At the same time, by separating tenders the auctioneer sets the relative quantities to be procured of each technology, which may not be trivial from a dynamic perspective. As Fabra and Montero (forthcoming) put it, “the comparison of a technology neutral versus a technology specific approach is faced with a fundamental trade-off. By allowing quantities to adjust to cost shocks, the technology neutral approach achieves cost efficiency at the cost of leaving high rents with infra-marginal producers. In contrast, the technology specific approach sacrifices cost efficiency in order to reduce those rents. In doing so, it also exploits the benefits that accrue from the (possibly, imperfect) substitutability across technologies. Therefore, whether one approach dominates over the other depends on the specifics of each case.” Hybrid approaches with minimum technology quotas can also be envisaged and may in certain circumstances offer a way out from this trade-off. The neutrality of a tender also depends on the criteria to select the beneficiaries. While the amount of aid that is needed per unit of carbon emission abated should be a main decision factor, some of the other dimensions that were mentioned above might also be valuable to the auctioneer. The CEEAG allows such complementary criteria to be given some limited weight.49 Importantly, all of these considerations about tender design are relevant for State aid and competition distortions to be the minimum that is necessary—as well as to ensure an efficient use of the available technologies. Phasing Out Fossil Fuels Decarbonising the economy will require not just significant investment in non-carbon-emitting technologies, but also the phasing-out of carbon-emitting ones. The CEEAG provide two complementary approaches to speed up this transition: limiting the granting of State aid to support new investment in fossil fuels, and allowing the granting of State aid to accelerate the decommissioning of fossil fuel plants. The CEEAG consider that measures that incentivise new investments in energy or industrial production based on the most polluting fossil fuels—such as coal, diesel, lignite, oil, peat, and oil shale—increase the negative climate and environmental externalities in the market. They will not be considered to have any positive climate or environmental effect—given the incompatibility of these fuels with the climate targets. Similarly, measures that incentivise new investments in energy or industrial production based on natural gas may reduce greenhouse gas emissions and other pollutants in the short term but aggravate negative climate and environmental externalities in the longer term, compared to alternative investments. For investments in natural gas to be seen as having positive climate or environmental effects, Member States must explain how they will ensure that the investment contributes to achieving the Union’s 2030 climate target and 2050 climate neutrality target. In particular, the Member States must explain how a lock in of this gas-fired energy generation or gas-fired production equipment will be avoided.50 This approach is consistent with the view of natural gas as a transitional source of energy in the context of the European Green Deal. While the shift away from coal, peat, and oil shale activities is largely driven by regulation, market forces such as the effects of carbon prices and competition from renewable sources with lower costs can also be important. Member States may decide to accelerate this market-driven transition by prohibiting the generation of power that is based on these fuels as of a certain date. This prohibition can create situations in which profitable coal, peat, and oil shale activities have to close before the end of their economic lifetime and can hence result in forgone profit. Member States may wish to grant compensation to facilitate the green transition.51 Moreover, the closure of uncompetitive coal, peat, and oil shale activities can generate significant social, climate, and environmental costs at the level of the power plants and the mining operations. Member States may decide to cover such exceptional costs to mitigate the social and regional consequences of the closure.52 Preserving Industrial Competitiveness Through the Energy Transition The European Green Deal requires that Member States put in place ambitious decarbonisation policies with budgets that will continue to be financed at least partly through levies on electricity consumption. For economic sectors that are particularly exposed to international trade and heavily reliant on electricity for their value creation, the obligation to pay the full amount of such levies can heighten the risk of moving to locations where climate and environmental policies are absent or less ambitious. As decarbonisation is a worldwide concern, such relocation would entail both a loss of local employment and a worsening of global climate and environmental conditions. To mitigate those risks the CEEAG foresee the possibility of granting reductions from such levies for firms that could be particularly affected.53 The basic limiting principle for the granting of reductions from levies is the “cost-reflectiveness principle”. Levies that reflect part of the cost of providing electricity expose final customers to the true costs of supplying the secure electricity that they consume, which contributes to efficient demand response. In order to preserve the cost reflectiveness of electricity bills, exemptions from network charges or from charges that finance capacity are not allowed. Member States may instead grant reductions from levies on electricity consumption that do not directly reflect part of the cost of providing secure electricity—such as levies that finance climate or environmental policy objectives or other social policy objectives.54 The CEEAG limit the exemptions from levies on electricity consumption to firms that may be at risk of delocalisation—which is identified at the sectoral level, based on the electro-intensity of the sector and its openness to international trade.55 Concluding Remarks: The Challenges Ahead Since the adoption of the CEEAG in January 2022, energy markets have been affected by exceptional events: Russia’s invasion of Ukraine; interruptions in gas supplies; and an escalation in energy prices. In response to these events, already in March 2022 the European Commission proposed the outline of a plan to make Europe independent from Russian fossil fuels before 2030 and adopted a Temporary Crisis Framework to enable Member States to use the flexibility that was foreseen under State aid rules to support the European economy. In May 2022 the European Commission presented the REPowerEU Plan in response to the hardships and global energy market disruption that were caused by Russia's invasion of Ukraine. The Commission’s plan aims to reduce the EU's dependence on fossil fuels and to accelerate the energy transition that was already foreseen to attain the targets of the European Green Deal. The plan includes actions in five main areas:Reducing energy consumption by enhancing long-term energy efficiency measures and triggering behavioural changes in energy consumption; Diversifying supplies by securing higher levels of LNG imports, optimising infrastructure use, pooling demand, and supporting international suppliers; Accelerating the rollout of renewables and the development of renewable hydrogen technologies and infrastructure; Reducing fossil fuel consumption in industry and transport through energy savings, efficiency enhancement, fuel substitution, electrification, and an enhanced uptake of renewable hydrogen, biogas, and biomethane; Enabling additional investment and reorienting the existing national Recovery and Resilience Plans to align them with the priorities in the REPowerEU plan. The priorities and lines of action that were set out in the REPowerEU Plan will require an even greater effort in terms of public and private investments in line with the European Green Deal: The targets are more ambitious, and the timeline to achieve them is shorter. In September 2022 the Commission proposed emergency interventions in Europe's energy markets to limit the impact of disruptions in gas supplies and increases in energy prices. These interventions included exceptional electricity demand reduction measures and measures to claw-back infra-marginal rents of electricity generation so as to redistribute them amongst consumers, as well as solidarity measures that tax the profits of oil and gas companies. These measures were added to previously agreed measures on filling gas storage and reducing gas demand in preparation for the upcoming winter. The Commission continues its work to improve liquidity for market operators, bring down the price of gas, and reform the electricity market design for the longer term. A reflection was also launched by the Commission to study options to improve the long-term functioning of the electricity market. These include: market-based instruments to protect consumers against price volatility; measures to enhance demand-response and promote individual self-consumption schemes; appropriate investment signals; a more transparent market surveillance; and possible adjustments to the electricity market design. While all of these elements depict a challenging context for the implementation of the CEEAG, their design responds precisely to the needs for State aid that are posed by the European Green Deal, and are reinforced by the current energy crisis. The CEEAG—together with the Temporary Crisis Framework—aim to provide the guidance to enable the necessary State aid support for the economy today and over the coming years. Mergers Main Developments Between January 2021 and September 2022, 673 merger investigations were concluded at DG Competition. The vast majority (523 cases) were unconditional clearances under a simplified procedure. 13 cases were abandoned in phase I. Of the remaining cases, 122 were concluded during a (non‑simplified) phase I investigation and 15 were concluded during a phase II investigation.56 Of the (non‑simplified) phase I cases, 110 were cleared unconditionally, and 12 could be cleared in phase I subject to commitments. The phase II investigations resulted in 6 clearances that were subject to commitments,57 no unconditional clearance, 2 prohibitions,58 and 6 abandoned transactions.59 Therefore, 5.8% of cases were not cleared unconditionally during this period. The CET was involved in all second‑phase investigations as well as in many complex first‑phase investigations. Analyses by members of the CET included, for instance: bidding analyses; quantitative market delineation (such as catchment-area analysis); as well as conceptual contributions to the construction and testing of sound theories of harm. As was mentioned in last year’s RIO paper on DG Competition developments,60 as a new policy initiative the Commission welcomes Member State referrals of mergers that are below the EU’s revenue thresholds with significant anti-competitive potential in the context of Article 22 of the Merger Regulation.61 The first case that was investigated by the Commission in the context of this initiative was the Illumina/Grail merger.,6263 The case involved the vertical integration of Illumina—the unrivalled supplier of “next generation sequencing” (NGS) systems for genetic and genomic analysis—with GRAIL: a customer of Illumina that used NGS systems to develop cancer detection tests. The Commission prohibited the merger as it would have stifled innovation, and reduced choice in the emerging market for blood-based early cancer detection tests through vertical foreclosure. Illumina has indicated that it will appeal the Commission’s decision.64 Veolia/Suez Merger On 14 December 2021, the Commission cleared a merger between Veolia and Suez, subject to conditions.65 Veolia and Suez are leaders in the water treatment and waste management sectors, where the two companies offer a wide range of services to municipal and industrial customers, including in particular:For the water sector: the provision of services that relate to the design and construction of water treatment facilities, the operation and maintenance of these water treatment facilities, the supply of water treatment chemicals, and the provision of mobile water solutions; For the waste sector: the provision of services that relate to the collection and treatment of non-hazardous waste, regulated waste (such as electrical equipment, medical waste), and hazardous waste.66 The Commission found that the Transaction, as initially notified, created significant horizontal overlaps and would lead to competition concerns in various markets, notably: (i) municipal water management in France; (ii) industrial water management in France; (iii) mobile water services in the EEA; (iv) the collection and treatment of non-hazardous and regulated waste in France; and (v) the treatment of hazardous waste in France. With regard to water management, the Commission considered national markets: Market shares were particularly high in France, with a combined market share of the two merging companies of 80–90% for the production and distribution of municipal water; 70–80% for the collection and treatment of used water; and 50–60% for industrial water management. With regard to waste management, the Commission found that the local dimension was an important element for the assessment of the proposed Transaction for non-hazardous waste, regulated waste, and hazardous waste in France:For non-hazardous waste, the Commission assessed the market structure for each French department (including also neighbouring departments) where a treatment site of the merging parties was located.67 We considered a catchment area of 200 km around each site for both the incineration of non-hazardous waste and the landfilling of non-hazardous waste.68 For regulated waste, the Commission assessed the market structure at the regional level for the collection and the treatment of electrical equipment, at the departmental level and regional level for the treatment of medical waste, and at departmental level and regional level for the collection of furnishing waste. For both the landfilling and incineration of hazardous waste, the Commission considered local markets with a 300 km catchment area around each site of the merging parties. In addition, the Commission also used a customer-centric approach, by assessing the accessibility of sites. This customer-centric approach complemented the site-centric approach and provided additional insights for the assessment of the proposed Transaction. The CET was involved in the assessment of local markets, supporting the case for the catchment-area analyses that are more standard—centered on each site of the merging parties—and designing the customer-centric catchment areas analysis for hazardous waste (Sect. 4.2.1). In the various markets for water management and waste management, customers typically select the services providers through tenders. The Commission relied on the analysis of tender data in its competitive assessment to measure the competitive interactions between the merging parties that would be lost post-Transaction (Sect. 4.2.2). The next two sections focus on the main economic contributions to the case: (i) the catchment area analyses that were carried out for hazardous waste in France (Sect. 4.2.1); and (ii) the analysis of tender data (Sect. 4.2.2). Site-Centric and Customer-Centric Catchment Area Analyses Description of the available data and the different approaches for catchment-area analyses The Commission identified competition concerns for the treatment of hazardous waste in France through landfilling and incineration.69 The evidence indicated that clients have a strong preference for sending their dangerous waste over a relatively short distance. In particular, the Commission analysed data on the volumes of dangerous waste that were collected by each site of the merging parties and their main competitors and calculated the distance that corresponded to 70% of the volume collected, as well as for 80% and for 90%.70 The Commission considered that the distance that corresponded to the 80% volume-threshold was the most appropriate since it was also confirmed by feedback that was received from the market investigation; this threshold also removed volumes that were sent over longer distance due to exceptional circumstances, and made it possible to capture the competitive interactions between the merging parties and their competitors. This led to catchment areas of 300 km radius for both the landfilling and the incineration of hazardous waste. In practice, a circle of 300 km radius is drawn around each site of the merging parties, and market shares are calculated by taking into consideration the total volume that is treated by each site included in the catchment area (see Fig. 2). Since data on the total volume that is treated by a site is generally available for both the merging parties and their competitors, the site-centric approach is relatively straightforward to implement.Fig. 2 Illustration of site-centric market shares In its assessment, the Commission also considered a catchment-area analysis that centered on customers and was based on the actual shipments of hazardous waste that were sent by customers to the sites of the different competitors.71 The combination of the two approaches provided important insights for the design of the remedy in order for the merging parties to divest the appropriate sites to resolve competition concerns at a local level. The customer-centric approach was possible in this case due to the quality of the data available. Calculating market shares around a specific customer (and based on actual shipments from this specific customer to the sites of the merging parties and their competitors) requires data on all waste flows that were received by each site of the merging parties and their competitors from this specific customer. This approach raises practical issues in terms of data, which explain why in practice this is less often used than is the site-centric approach:First, this customer-centric approach requires data from the merging parties and their competitors on the waste that is sent by each of their customers to each of their sites. While it is generally feasible to obtain this (granular) data from the merging parties, getting such data from third-parties is more difficult as it involves significant work from third-parties.72 Moreover, making data consistent across the merging parties and third-parties is also necessary to proceed with a market reconstruction: The main task is to calculate a reliable market size that captures all of the alternatives that are available for each specific customer. This data-management task is often very time-consuming: For example, for the incineration of hazardous waste in France, the Commission considered 11 competitors in the customer-centric analysis; Second, while this approach is not too hard to implement with only a few customers, it quickly becomes impractical when customers become more numerous and more dispersed. Fortunately in this case, the shipment of hazardous waste in France is regulated, and each shipment needs to be declared to the public administration. A database can therefore be made available with the customer names anonymised73:As the French department where the customer is located is available in the database, it is possible to aggregate all customers at a department level. Since departments in France are sufficiently granular (see Figs. 3 and 4), this allows the pooling of customers that face similar competitive conditions. For a specific department, the data identify the receiving site for each shipment. We can therefore construct waste flows for both the merging parties and the third-party competitors, which can be used to calculate market shares based on a market size that captures all alternatives that are available for customers that are located in a specific department. Fig. 3 Customer-centric approach for the landfilling of hazardous waste in France. Source: extract from the European Commission decision in Case M.9969 – Veolia/Suez (Figures 2, 3) Fig. 4 Customer-centric approach for the incineration of hazardous waste in France. Source: extract from the European Commission decision in Case M.9969 – Veolia/Suez (Figures 5, 9) The Catchment Area Analysis for the Landfilling of Dangerous Waste in France The site-centric market shares for the landfilling of dangerous waste are reported in Table 1; these data show that the proposed Transaction would create:A quasi-monopoly in the South of France, around a site of Suez (Bellegarde) with a combined share in the range of 90-100% and a site of Veolia (Occitanis) with a combined share in the range of 90-100%; A quasi-monopoly in the East of France, with a significant increment around one particular site of Suez (Laimont) with a combined share of 90-100%; A quasi-monopoly in the Paris area, with a combined market share in the range of 90-100% around a site of Veolia (Guitrancourt) and a site of Suez (Villeparisis); A high combined market share in the North- of France, in the range of 60-70% around a site of Veolia (Solicendre). Table 1 Site-centric market shares for the landfilling of hazardous waste in France Group Site Region Suez’ market share Veolia’s market share Combined market share Séché’s market share Suez Bellegarde Occitanie [70–80] % [20–30] % [90–100] % [0–5] % Suez Laimont Grand Est [60–70] % [30–40] % [90–100] % [0–5] % Suez Jeandelaincourt Grand Est [90–100] % [0–5] % [90–100] % [0–5] % Suez Villeparisis Ile-de-France [40–50] % [50-60] % [90–100] % [0–5] % Suez Vaivre-Et-Montoille Bourgogne-Franche-Comté [90–100] % [0-5] % [90–100] % [0–5] % Suez Champteussé- Sur-Baconne Pays de la Loire [5–10] % [20-30] % [30–40] % [60–70] % Suez Drambon Bourgogne-Franche-Comté [90–100] % [0-5] % [90–100] % [0–5] % Veolia Guitrancourt Ile-de-France [40–50] % [50-60] % [90–100] % [0–5] % Veolia Occitanis Occitanie [70–80] % [20-30] % [90–100] % [0–5] % Veolia SIRA Auvergne-Rhône-Alpes [90–100] % [0-5] % [90–100] % [0–5] % Veolia SERAF Normandie [20–30] % [30-40] % [60–70] % [30–40] % Veolia SOLICENDRE Normandie [20–30] % [30-40] % [60–70] % [30–40] % Veolia SOLTTOP Pays de la Loire [10–20] % [10-20] % [20–30] % [70–80] % Source: extract from the European Commission decision in Case M.9969—Veolia/Suez (Table 29). The market share of the merged entity is highlighted in bold However, in the West of France, the combined market share of the merged entity would be lower, in the range of 30–40% around a site of Suez (Champteussé-Sur-Baconne) and 20–30% around a site of Veolia (Solitop), with the presence of a strong competitor (Séché). As was described above, for each flow of waste that is sent to a site in France, the data identify the department of origin of the customer that sent the waste for treatment. We can therefore adopt a customer-centric approach, where for each department, one can calculate the market shares of the merging parties and their competitors based on the actual volumes that were sent by customers. This approach has several advantages compared to a site-centric approach:The market shares are based on the sites that are effectively used by the merging parties; and The market shares do not depend on any specific radius since actual volumes are used. Figure 3 reports, for each department in France, the combined market share of the merging parties and the increment. This customer-centric analysis confirmed our previous findings as to the likely consequences of the merger based on site-centric market shares:A quasi-monopoly in the South of France, with a significant increment above 10% in several departments (for example, Aude, Hérault, Haute-Garonne, Tarn-et-Garonne, Lot, Cantal), and even above 40% in three departments (Tarn, Aveyron, Gard); A quasi-monopoly in the Paris area, with a significant increment in Paris area and in the North of France; In addition, a customer-centric approach revealed additional evidence for the West of France, where customers used much more Suez (which owns one site in the West of France) compared to what was suggested by site-centric market shares (Table 1 above). In particular, the proposed Transaction led to a significant increment for several departments in the West of France (for example, above 20% in Gironde, Charente-Maritime, Deux-Sèvre, Vendée, Maine-et-Loire, Sarthe, and above 10% in Loire-Atlantique, Orne, Calvados, Manche). The catchment area analysis for the incineration of dangerous waste in France The corresponding market shares (site-centric approach) for the incineration of hazardous waste in France are reported in Table 2; these data show that the proposed Transaction would lead to combined market shares above 50% for seven of the eight Suez sites.Table 2 Site-centric market shares for the incineration of hazardous waste in France Group Site Region Suez’ market share Veolia’s market share Combined market share Séché’s market share Market share of others Suez Givors Auvergne-Rhône-Alpes [30–40] % [10–20] % [50–60] % [30–40] % [10–20] % Suez Airvault Nouvelle-Aquitaine [10–20] % [60–70] % [80-90] % [0–5] % [10–20] % Suez Frontignan Occitanie [30–40] % [20–30] % [50–60] % [30-40] % [5–10] % Suez Amnéville Grand-Est [5–10] % [0–5] % [5–10] % [10–20] % [70–80] % Suez Barlin Hauts-De-France [5–10] % [50–60] % [60–70]% [0–5] % [30–40] % Suez Oriolles Nouvelle-Aquitaine [10–20] % [60–70] % [80–90] % [0–5] % [10–20] % Suez Roussillon Auvergne-Rhône-Alpes [30-40] % [10–20] % [50–60] % [20–30] % [10–20] % Suez Le Pont-De-Claix Auvergne-Rhône-Alpes [30–40] % [10–20] % [50–60] % [30–40] % [10–20] % Veolia Limay Ile-de-France [5–10] % [50–60] % [60–70] % [0–5] % [30–40] % Veolia SIAP Nouvelle-Aquitaine [10–20] % [60–70] % [80-90] % [0–5] % [10–20] % Veolia SEDIBEX Normandie [5–10] % [50–60] % [60–70] % [0–5] % [30–40] % Veolia SMTB Nouvelle-Aquitaine [10–20] % [80–90] % [90–100] % [0–5] % [5–10] % Veolia SOLAMAT-MEREX/Fos-sur-Mer Prorence-Alpes-Côte d'Azur [40–50] % [20–30] % [60–70] % [20–30] % [5–10] % Veolia SOLAMAT-MEREX/Rognac Provence-Alpes-Côte d'Azur [40–50] % [20–30] % [60–70] % [20–30] % [5–10] % Veolia SOTRENOR Hauts-de-France [5–10] % [50–60] % [60–70] % [0–5] % [30-40] % Source: extract from the European Commission decision in Case M.9969—Veolia/Suez (Tableau 35). The market share of the merged entity is highlighted in bold The customer-centric approach, which considers the sites that are effectively used by customers for the incineration of hazardous waste, also showed a high degree of competitive interactions between the sites of the merging parties. For each department, the Commission calculated the share of volume that was sent to each site of the suppliers that were active in the incineration of hazardous waste.74 In order to identify the sites of the merging parties that compete against each other for the customers in a given department, Fig. 4 reports the combined market share of the merging parties and the increment. The customer-centric approach showed that several sites of the merging parties were competing head-to-head:In the South-East of France (for example, Bouches-du-Rhône, Drome, Gard, Alpes-de-Haute-Provence, Vaucluse), customers used mainly the sites of the merging parties, representing more than 40% of the hazardous waste incinerated (and up to 80%-90% in some departments). A similar finding applies to other departments in the South of France (Aveyron, Haute-Garonne, Hautes-Pyrénées, Tarn), in the South-West of France (Charente-Maritime, Corrèze, Dordogne, Gironde, and Landes), in the West of France (Vienne, Loire-Atlantique, Loiret, Maine-et-Loire), and in the North of France (Aisne, Nord, Paris, Seine-et-Marne, Yvelines, Val-d’Oise, Oise, Calvados, Hauts-de-Seine). Conclusion of the Catchment-Area Analysis The Commission complemented the commonly used site-centric approach with a customer-centric approach that was based on the actual volumes that were collected by the sites of the merging parties and their competitors. Using a customer-centric approach based on actual volumes allowed an improvement of the analysis along several dimensions:The customer-centric approach is based on the alternatives that were effectively used by customers (instead of a standardised area defined by a specific radius); the customer-centric analysis is not sensitive to the choice of an arbitrary catchment radius; Considering the sites that were effectively used by customers allows us to identify the sites of the merging parties that are in head-to-head competition. It helps us assess the effects of the proposed Transaction at a local level and decide which sites of the merging parties had to be divested to alleviate competition concerns.75 The Bidding Data Analysis The Commission also carried out an assessment of the competitive interactions between the different suppliers that participated in tenders in both the water management sector and the waste management sector. The Commission conducted a number of empirical analyses of the bidding data76:Conditional participation analysis. The Commission analysed how often the merging parties participated against each other in tenders, relative to other suppliers. Conditional loss analysis. The Commission analysed how often the merging parties lost to each other. In its assessment, the Commission considered both analyses jointly: On the one hand, the participation analysis should be interpreted with the loss analysis, since a participant in a tender is credible only if it has a material probability to win a tender. A supplier that often participates but never wins would not be considered to be a credible alternative. On the other hand, a loss analysis is often characterised by a smaller number of observations: It considers only the tenders that were lost by one of the merging parties, while a participation analysis considers all tenders where a party participates. In markets where the sample size is too small for the loss analysis, the Commission puts more weight on the participation analysis. Finally, the Commission performed two additional kinds of analyses that were related to the market structure:Pre-Transaction: by considering the average renewal rate in tenders for Veolia and Suez,77 the average number of participants, and the number of tenders with one participant, two participants, three participants, etc.; Post-Transaction: by considering the proportion of tenders where the number of participants decrease from two to one participant, from three to two participants, etc. In particular, while a low renewal rate indicates that customers can easily switch suppliers and is a sign of a competitive landscape, a high renewal rate is consistent with a certain inertia in the market and a limited possibility for customers to switch suppliers: for example, due to important switching costs. A high renewal rate is also consistent with the existence of barriers to entry. Last, the Commission carried out the above analyses in terms of the number of tenders and also in terms of the values of tenders. Given the significant heterogeneity in the values of tenders and the difficulties for small suppliers to participate in important tenders (due to technical requirements and financial requirements), the Commission has put more weight on the analyses in terms of values.78 For several markets where the Commission raised competition concerns,79 the analysis of tender data showed that: (i) the merging parties were particularly close competitors, where Suez was participating often against Veolia (and vice versa), and Veolia was losing a high proportion of tenders to Suez (and vice versa)80; and (ii) the market structure was already concentrated pre-Transaction with a significant proportion of tenders with few participants, which would be even more concentrated post-Transaction.81 The analysis of tenders was also used as evidence to clear some markets—for example, municipal water in Czech Republic and in Spain; incineration of hazardous waste in Spain; chemical treatment of hazardous waste in Belgium—where the analysis of tender data showed that the merging parties were not particularly close competitors and that several credible alternatives would remain post-Transaction. Outcome of the Case/Remedy After its investigation, the Commission concluded that the proposed Transaction raised serious doubts in various markets in water and waste management sectors. The Commission ultimately approved the merger, subject to compliance with a comprehensive commitments package that was offered by Veolia, including: (i) the divestment of almost all of Suez’s activities in the municipal water market in France; (ii) the divestment of almost all of Veolia’s activities in the mobile water market in the EEA; (iii) the divestment of the vast majority of Veolia’s activities in the French segment of the industrial water management market; (iv) the divestment of almost all of Suez’s activities in the non-hazardous and regulated waste management markets in France; and (v) the divestment of part of Veolia’s and Suez’s landfilling activities of hazardous waste and all of Suez’s activities for the incineration and chemical treatment of hazardous waste in France. Conclusion Between January 2021 and September 2022, the CET contributed to a number of policy projects and cases of the Directorate General for Competition at the European Commission, across antitrust, State aid control and merger control. The dedication of resources at the CET has been increasingly more balanced and a similar amount of resources have been devoted to antitrust, State aid control, and merger control over the first six months of 2022. On the antitrust policy front, the Commission adopted a new Vertical Block Exemption Regulation (VBER), accompanied by new Vertical Guidelines. In the field of State aid control, the exercise of reviewing sectoral guidelines continued, notably leading to the adoption of new Guidelines on State aid for climate, environmental protection, and energy (CEEAG). The work of the CET on merger cases included the economic assessment of the Veolia/Suez merger, approved in Phase II subject to commitments. These are just few examples of the numerous work stream to which the CET has contributed over the past months. Acknowledgements The views that are expressed in this paper are those of the authors and do not necessarily reflect the views of DG Competition or the European Commission. We would like to thank Guillaume Debarbat, Philipp Dimakopoulos, Szabolcs Lorincz, Paul Bridgeland, Johannes Holzwarth, and Hans Zenger for helpful comments. Author contributions All authors contributed equally to this manuscript. Declarations Competing interests The authors declare no competing interests. 1 Case AT.40394 Aspen (Commission Decision of 10 February 2021); Case AT.40413 Focus Home, Case AT.40414 Koch Media, Case AT.40420 ZeniMax, Case AT.40422 Bandai Namco, Case AT.40424 Capcom (Commission Decisions of 20 January 2021). 2 Case AT.40511 Insurance Ireland: Insurance claims database and conditions of access (Commission decision of 30 June 2022). 3 Case AT.40305 Network sharing—Czech Republic (Commission decision of 11 July 2022). 4 See the press release at https://ec.europa.eu/commission/presscorner/detail/en/IP_22_2844. 5 See the public consultation at https://competition-policy.ec.europa.eu/public-consultations/2022-hbers_en 6 See, respectively, the press release at https://ec.europa.eu/commission/presscorner/detail/en/ip_22_5796 and the public consultation at https://competition-policy.ec.europa.eu/public-consultations/2022-informal-guidance-notice_en. 7 See the press release at https://ec.europa.eu/commission/presscorner/detail/en/IP_22_4194. 8 See the detailed description in last year’s agency update (Baltzopoulos et al., 2021). 9 Case T-286/09 RENV Intel Corporation v Commission (General Court judgment of 26 January 2022). 10 Case T-235/18 Qualcomm v Commission (Qualcomm—Exclusivity Payments) (General Court judgment of 15 June 2022). 11 The Commission did not appeal the General Court’s judgment in this case, which makes it a final decision. 12 Case T-612/17 Google and Alphabet v Commission (Google Shopping) (General Court judgment of 10 November 2021). 13 See Amelio et al. (2018) for a description of the case. 14 Case T-604/18 Google and Alphabet v Commission (Google Android) (General Court judgment of 14 September 2022). 15 See Kotzeva et al. (2019) for a description of the case. 16 Commission Regulation (EU) 2022/720 of 10 May 2022 on the application of Article 101(3) of the Treaty on the Functioning of the European Union to categories of vertical agreements and concerted practices. OJ L 134, 11.5.2022, 4–13. 17 Commission Notice: Guidelines on vertical restraints. OJ C 248, 30.6.2022, 1–85. 18 Section 5 of the Guidelines on vertical restraints addresses the definition of the relevant markets and the calculation of market shares, by reference to the Market Definition Notice. 19 Provided that the rest of the agreement is capable of being applied without the restraint in question. 20 False negatives refer to vertical agreements and restrictions which were previously not covered by the VBER but for which it can be assumed with sufficient certainty that they generally fulfill, under certain conditions, the requirements for an exemption pursuant to Article 101(3) of the Treaty. 21 Exclusive distribution systems are distribution systems where the supplier allocates a territory or group of customers exclusively to certain buyers (distributors) and restricts all of its other buyers from actively selling into the exclusive territory or to the exclusive customer group. 22 Selective distribution systems are distribution systems where a supplier selects its distributors in a certain territory on the basis of specified criteria. 23 False positives concern vertical agreements and restrictions that were previously covered by the safe harbour of the VBER but for which it cannot be assumed with sufficient certainty that they are generally on balance efficiency-enhancing and, thus, fulfill the conditions for an exemption pursuant to Article 101(3) of the Treaty. 24 “Passive sales” are sales made in response to unsolicited requests from customers, whereas “active sales” are sales that result from a firm’s actively targeting customers. The distinction is important in the context of exclusive distribution, where restrictions of active sales into the exclusive territory of another distributor or to a customer group exclusively allocated to another distributor are block exempted, while restrictions of ‘passive’ sales are considered as hardcore restrictions. While the 2010 Vertical Guidelines (para 53) provided some guidance on the distinction between active and passive online sales, in practice online sales seem to have been perceived and treated as passive sales; see Expert report on the review of the Vertical Block Exemption Regulation: Active sales restrictions in different distribution models and combinations of distribution models (2021, pp. 11–12), https://competition-policy.ec.europa.eu/system/files/2021-06/kd0821131enn_VBER_active_sales.pdf. 25 An exception is made for English, which is commonly used throughout the EU. 26 The definition of online intermediation services is based on the definition used in the Platform-to-business Regulation (Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services. OJ L 186, 11.7.2019, 57–79). 27 Communication from the Commission: Guidelines on the applicability of Article 101 of the Treaty on the Functioning of the European Union to horizontal co-operation agreements Text with EEA relevance. OJ C 11, 14.1.2011, 1–72. 28 Restrictions of competition "by object" are those that by their very nature have the potential to restrict competition; see Commission Staff Working Document: Guidance on restrictions of competition "by object" for the purpose of defining which agreements may benefit from the De Minimis Notice (2014). Available at https://ec.europa.eu/competition/antitrust/legislation/de_minimis_notice_annex.pdf.. 29 See, for example, the Swedish competition authority’s Booking.com decision of 15/04/2015. 30 https://ec.europa.eu/clima/eu-action/european-green-deal/european-climate-law_en 31 The sectors that are currently covered by the EU ETS account for around 41% of the EU's total emissions. The Commission proposed the extension to additional sectors—notably transport—over the coming years, progressively as of 2025. 32 “The Commission will consider that aid is necessary if the Member State demonstrates that it effectively targets residual market failures, also taking into account any other policies and measures already in place to address some of the market failures identified.” CEEAG, Paragraph 36. 33 “(…) Other policies and measures may already be in place to address some of the identified market failures. Examples include sectorial regulation, mandatory Union pollution standards, supply obligations, pricing mechanisms such as the Union’s Emissions Trading System (ETS) and carbon taxes. Additional measures, including State aid, may only be directed at residual market failures, that is to say those that remain unaddressed by such other policies and measures. (…)” CEEAG, Paragraph 35. 34 “State aid is not the only policy instrument available to Member States to promote increased levels of environmental protection or to ensure an efficient internal energy market. There may be other, more appropriate instruments available (…)” CEEAG, Paragraph 40. 35 “Compliance with the ‘polluter pays’ principle through environmental legislation aims at ensuring that a market failure linked to negative externalities will be rectified. Therefore, State aid is not an appropriate instrument and cannot be granted insofar as the beneficiary of the aid could be held liable for the pollution under existing Union or national law.” CEEAG, Paragraph 42. 36 “(…) the Member State is required to demonstrate why other potentially less distortive forms of aid are less appropriate, such as: repayable advances as compared to direct grants; tax credits as compared to tax reductions; or forms of aid that are based on financial instruments, such as debt as compared to equity instruments, including, for example, low-interest loans or interest rebates, State guarantees, or an alternative provision of financing on favourable terms.” CEEAG, Paragraph 44. 37 “Aid can be considered as facilitating an economic activity only if it has an incentive effect. An incentive effect occurs when the aid induces the beneficiary to change its behaviour, to engage in additional economic activity or in more environmentally-friendly economic activity, which it would not carry out without the aid or would carry out in a restricted or different manner.” CEEAG, Paragraph 26. 38 “Proving an incentive effect entails the identification of the factual scenario and the likely counterfactual scenario in the absence of aid (39). The Commission will examine this based on the quantification referred to in Sect. 3.2.1.3.” CEEAG, Paragraph 28. 39 “As a general principle, aid will be considered as limited to the minimum needed for carrying out the aided project or activity if the aid corresponds to the net extra cost (‘funding gap’) necessary to meet the objective of the aid measure, compared to the counterfactual scenario in the absence of aid. The net extra cost is determined by the difference between the economic revenues and costs (including the investment and operation) of the aided project and those of the alternative project which the aid beneficiary would credibly carry out in the absence of aid.” CEEAG, Paragraph 48. 40 “Where the aid is not granted under a competitive bidding process, the net extra cost must be determined by comparing the profitability of the factual and counterfactual scenarios. To determine the funding gap in such cases, the Member State must submit a quantification, for the factual scenario and a credible counterfactual scenario, of all main costs and revenues, the estimated weighted average cost of capital (WACC) of the beneficiaries to discount future cash flows, as well as the net present value (NPV) for the factual and counterfactual scenarios, over the lifetime of the project. The Commission will verify whether this counterfactual is realistic (45). The Member State must provide reasons for the assumptions used for each aspect of the quantification, and explain and justify any methodologies applied. The typical net extra cost can be estimated as the difference between the NPV for the factual scenario and for the counterfactual scenario over the lifetime of the reference project.” CEEAG, Paragraph 51. 41 Competitive tendering can be an efficient tool in the presence of informational asymmetries—especially when the auctioneer has limited information about the bidders’ costs. If the auctioneer had perfect information about current and future costs of all potential bidders, it could use this information to discriminate across bidders to extract all surplus. However, often the auctioneer has limited information, and competitive tendering can help reveal the overall total minimum amount of aid that is needed to conduct the projects—at the cost of allowing infra-marginal rents to be captured by infra-marginal bidders. 42 “A detailed assessment of the net extra cost will not be required if the aid amounts are determined through a competitive bidding process, because it provides a reliable estimate of the minimum aid required by potential beneficiaries (…)” CEEAG, Paragraph 49. 43 CEEAG, Paragraph 64. 44 CEEAG, Paragraph 67. 45 CEEAG, Paragraph 68. 46 CEEAG, Paragraph 69. 47 “The Commission considers that schemes open to a broader range of potential beneficiaries have or are likely to have a more limited distortive effect on competition than support targeted at a limited number of specific beneficiaries only, in particular where the scope of the aid measure includes all competitors willing to deliver the same service, product or benefit.” CEEAG, Paragraph 66. 48 “(…) the Member State should give reasons for measures which do not include all technologies and projects that are in competition—for example all projects operating in the electricity market, or all undertakings producing substitutable products and which are technically capable of contributing efficiently to greenhouse gas emissions reductions. These reasons should be based on objective considerations linked, for example, to efficiency or costs or other relevant circumstances (…)” CEEAG, Paragraph 95. 49 “The selection criteria used for ranking bids and, ultimately, for allocating the aid in the competitive bidding process should as a general rule put the contribution to the main objectives of the measure in direct or indirect relation with the aid amount requested by the applicant. This may be expressed, for example, in terms of aid per unit of environmental protection or aid per unit of energy. It may also be appropriate to include other selection criteria that are not directly or indirectly related to the main objectives of the measure. In such cases, these other criteria must account for not more than 30% of the weighting of all the selection criteria. The Member State must provide reasons for the proposed approach and ensure it is appropriate to the objectives pursued.” CEEAG, Paragraph 50. 50 CEEAG, Paragraph 128 and 129. 51 CEEAG, Paragraph 423 and 424. 52 CEEAG, Paragraph 436. 53 CEEAG, Paragraph 399 and 400. 54 CEEAG, Paragraph 403. 55 CEEAG, Paragraph 405. 56 Mergers must be notified to the European Commission if the annual turnover of the combined business exceeds certain thresholds in terms of global and European sales. Notification triggers a 25-working-day phase I investigation. In the majority of cases, this follows a simplified procedure. If the transaction does not raise serious doubts with respect to its compatibility with the common market at the end of phase I, the Commission issues an unconditional clearance decision. If concerns exist but are addressed in a clear‑cut manner by remedies that have been proposed by the parties, the transaction can be cleared conditionally in phase I. Otherwise, the Commission will start a 90‑working‑day phase II investigation. At the end of phase II, the transaction is either cleared (conditionally or unconditionally) or prohibited; the latter occurs if the Commission finds that the transaction would lead to a significant impediment of effective competition even after taking into account any commitments that have been proposed by the parties. Details on the European Union merger regulation are available at https://ec.europa.eu/competition/mergers/procedures_en.html. Detailed statistics on the number of merger notifications and decisions are available at https://ec.europa.eu/competition/mergers/statistics.pdf. 57 Case M.9564 LSEG/Refinitiv Business (Commission decision of 13 January 2021); Case M.9820 Danfoss/Eaton Hydraulics (Commission decision of 18 March 2021); Case M.9569 EssilorLuxottica/Grandvision (Commission decision of 23 Mar 2021); Case M.9829 Aon/Willis Towers Watson (Commission decision of 9 Jul 2021); Case M.10262 Meta (formerly Facebook)/Kustomer (Commission decision of 27 January 2022); Case M.10078 Cargotec/Konecranes (Commission decision of 24 February 2022). 58 Case M.9343 HHI/DSME (Commission decision of 13 January 2022); Case M.10188 Illumina/Grail (Commission decision of 6 September 2022). 59 Case M.9162 Fincantieri/Chantiers de l’Atlantique (withdrawn 27 January 2021); Case M.9489 Air Canada/Transat (withdrawn 2 April 2021); Case M.9637 IAG/Air Europa (withdrawn 12 December 2021); Case M.9987 NVIDIA/ARM (withdrawn 8 February 2022); Case M.10319 Greiner/Recticel (withdrawn 28 February 2022); Case M.9938 Kingspan/Trimo (withdrawn 21 April 2022). 60 Baltzopoulos et al. (2021). 61 Commission Guidance on the application of the referral mechanism set out in Article 22 of the Merger Regulation to certain categories of cases, 26 March 2021, available at: https://ec.europa.eu/competition/consultations/2021_merger_control/guidance_article_22_referrals.pdf. 62 Case M.10188 Illumina/Grail (Commission decision of 6 September 2022). Press release available at: https://ec.europa.eu/commission/presscorner/detail/en/IP_22_5364. 63 In a parallel litigation, the General Court of the EU upheld the Commission's referral decisions of 19 April 2021, thereby confirming the Commission's jurisdiction to examine the impact of the transaction. See the GC’s judgement of 13 July 2022 available at: https://curia.europa.eu/jcms/upload/docs/application/pdf/2022-07/cp220123en.pdf. 64 Press release available at: https://www.illumina.com/company/news-center/press-releases/2022/1ef95365-0ca9-4726-a683-37124b1116b5.html. 65 Case M.9969 Veolia/Suez. 66 Non-hazardous waste includes wastes without any negative impact on people or on the environment. Regulated waste includes (for example) electrical equipment and medical waste. Hazardous waste is waste of industrial origin whose toxicity requires specific treatment techniques, equipment and expertise. In addition, hazardous waste is subject to specific control procedures and regulations that distinguish it from ordinary and regulated waste. 67 The department is one of the three administrative divisions of France, between the administrative regions (less granular) and the municipalities (more granular). France includes 96 departments in metropolitan France, and 5 overseas departments. 68 Catchment areas were centered on the site of the merging parties with a radius of 200 km. 69 The landfilling of hazardous waste is a method of treatment for storing hazardous waste underground at specially equipped sites. The incineration of hazardous waste is a method of treatment consisting of burning hazardous waste. It requires reaching temperatures that are generally higher than those required for the incineration of non-hazardous waste. 70 The data that were provided by the merging parties were extracted from the GEREP database, which is published every year by the French administration (« Ministère de la Transition Ecologique»). The French administration collects in particular data on hazardous waste from classified facilities each year. 71 Catchment areas can be drawn around customers, around suppliers, and/or within the intersection of suppliers’ catchment areas, depending on the specificities of the case. While it is often preferable to assess competitive conditions at each customer location, it may not be possible to draw catchment areas around customer locations—for example, because customers are many and dispersed or because there is no information on the location of customers of competitors. For practical purposes, one may then draw catchment areas around supplier locations. 72 See for example the Case M.7408 Cargill/ADM, where the Commission carried out a market reconstruction that was based on third-party data in order to implement a customer-centric approach. 73 See footnote 74 for further details on the data used. 74 Namely: Veolia, Suez, Séché, Calcia, Eqiom, Flamme, Holcim-Larfarge, Inertam, Kereneos, Lhoist, and Vicat. 75 For the remedy assessment, see Sect. 9.5 of the European Commission decision in Case M.9969—Veolia/Suez. 76 See also for example: Case M.9829 Aon/Willis Towers Watson, Case M.9779 Alstom/Bombardier Transportation, Case M.8677 Siemens/Alstom, Case M.7278 General Electric/Alstom, Case M.6851 Baxter International/Gambro, Case M.4747 IBM/Telelogic. 77 The renewal rate is defined as the proportion of tenders that were won by Veolia (respectively Suez) amongst all tenders where Veolia (respectively Suez) was the incumbent. For example, this analysis showed that in the market for the production and distribution of municipal water in France, the renewal rate of Veolia is in the range of 90–100%. 78 See also the Commission’s decisions in Case M. 9829 Aon/Willis Towers Watson and Case M.8677 Siemens/Alstom. 79 For example: the market for the management of municipal water in France; the market for the design and construction of water treatment facilities in France; the market for industrial water management in the EEA; the market for mobile water in the EEA; the market for the collection of household waste in France; the market for the incineration of non-hazardous waste in France; the market for the landfilling of non-hazardous waste in France; the market for the treatment of regulated waste in France; the market for the landfilling of hazardous waste in France; the market for the incineration of hazardous waste in France. 80 The Commission noted as well that participation rates and loss rates between the merging parties were higher in value, which indicates that the merging parties were even closer competitors for high-value tenders. 81 For example, post-Transaction, 70–80% of the tenders would have been characterised by a monopoly or a duopoly in the market for the treatment of used water in France. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References Amelio A Buettner T Hariton C Koltay G Papandropoulos P Sapi G Valletti T Zenger H Recent Developments at DG competition: 2017/2018 Review of Industrial Organization 2018 53 653 679 10.1007/s11151-018-9671-7 30546197 Baltzopoulos A Karlinger L Magos D Régibeau P Vareda J Recent developments at DG competition: 2020/2021 Review of Industrial Organization 2021 59 567 598 10.1007/s11151-021-09847-6 34840422 Commission notice Guidelines on vertical restraints OJ C 2022 248 6 1 85 Communication from the Commission Guidelines on the applicability of Article 101 of the treaty on the functioning of the European Union to horizontal co-operation agreements Text with EEA relevance OJ C 2011 11 1 1 72 Commission Staff Working Document: Guidance on restrictions of competition "by object" for the purpose of defining which agreements may benefit from the De Minimis Notice (2014). Available at https://ec.europa.eu/competition/antitrust/legislation/de_minimis_notice_annex.pdf. EU Regulation 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services OJ L 2019 186 7 57 79 Expert report on the review of the vertical block exemption regulation: Active sales restrictions in different distribution models and combinations of distribution models (2021). Available at https://competition-policy.ec.europa.eu/system/files/2021-06/kd0821131enn_VBER_active_sales.pdf. Fabra, N., & Montero, J-P. Technology neutral vs. technology specific procurement. The Economic Journal (in press). Kotzeva R Kovo D Lorincz S Sapi G Sauri L Valletti T Recent developments at DG competition: 2018/2019 Review of Industrial Organization 2019 55 551 578 10.1007/s11151-019-09739-w Regulation Commission (EU), 2022/720 of 10 May 2022 on the application of Article 101(3) of the Treaty on the Functioning of the European Union to categories of vertical agreements and concerted practices OJ L 2022 134 5 4 13	36466379	PMC9709377	NO-CC CODE	2022-12-04 23:14:54	no	Rev Ind Organ. 2022 Nov 30; 61(4):449-487	utf-8	Rev Ind Organ	2,022	10.1007/s11151-022-09890-x	oa_other
==== Front Educ Inf Technol (Dordr) Educ Inf Technol (Dordr) Education and Information Technologies 1360-2357 1573-7608 Springer US New York 11400 10.1007/s10639-022-11400-1 Article Analyzing the outcomes of a robotics workshop on the self-efficacy, familiarity, and content knowledge of participants and examining their designs for end-of-year robotics contests Mallik Abhidipta Liu Dongdong http://orcid.org/0000-0001-5994-256X Kapila Vikram [email protected] grid.137628.9 0000 0004 1936 8753 Department of Mechanical and Aerospace Engineering, NYU Tandon School of Engineering, Brooklyn, NY USA 30 11 2022 140 1 7 2022 4 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. Rapid advances in science and engineering, and pervasive adoption of resulting technological products, are influencing every aspect of human living and fueling a growing demand for a workforce that is adequately prepared for the emerging occupations in STEM fields. Educating students for success in the modern technology-rich workplace requires teachers who have the knowledge, comfort, capability, and training to adopt and integrate new technologies for classroom teaching and learning. Thus, to prepare high school teachers for incorporating robotics in their students’ education and promoting their understanding of engineering concepts and technology applications, a four-week long robotics workshop was designed and conducted annually for three summers. Examination of changes in the workshop participants’ levels of robotics self-efficacy, familiarity, and content knowledge, as well as analysis of outcomes of robotics capstone projects and end-of-year contests, is suggestive of study findings being promising for education researchers and professional development providers interested in leveraging the potential of robotics in STEM education. Keywords Content knowledge Contest Entrepreneurship Familiarity High school Professional development Robotics Self-efficacy STEM http://dx.doi.org/10.13039/501100008982 National Science Foundation ITEST DRL: 1614085 Mallik Abhidipta ==== Body pmcIntroduction The advancements in scientific discovery, engineering innovation, and technology adoption are shaping all human endeavors including education, work, and leisure. Sustaining this ongoing societal transformation requires an adequately educated STEM workforce. According to the World Economic Forum (2018), one consequence of the ongoing technological advances will be the emergence of new job roles, e.g., Robotics Specialists and Engineers. Preparing students for success in a technology-rich workplace (Varier et al., 2017) requires teachers with knowledge, comfort, and capacity to adopt new technologies effectively for classroom practice (Buabeng-Andoh, 2012; Office of Educational Technology, 2017). A careful incorporation of advanced technologies in developing and implementing curricula and methodologies for STEM education can fulfill the goal of producing a technologically trained and globally competitive U.S. workforce for the 21st century innovation economy (Office of Educational Technology, 2017; Varier et al., 2017). With the U.S. being eliminated from the list of the top 10 countries for scientific innovation (Jamrisko and Lu, 2018), educating and training a strong STEM-prepared U.S. workforce has become urgent. Thus, government, education, and corporate sectors are all focused on creating innovative STEM teaching and learning opportunities for students at all levels. In fact, the Next Generation Science Standards (NGSS) (NGSS Lead States, 2013) emphasize the need for integrating engineering design and practices within the K-12 curriculum. To impart engaging and interactive STEM learning experiences to students—with personally meaningful and motivating contexts—teachers are increasingly incorporating educational robots (Barker et al., 2012) in classrooms. Recent studies show improvement in student learning with the use of robots (Chin et al., 2014). Students are fascinated with (Vollmer et al., 2011) and appreciate the opportunity to learn with robots, and allowing students to solve real-life problems can enhance their STEM interest (Riojas et al., 2011). The use of robots in teaching embeds opportunities for students to experience kinesthetic learning, i.e., learning by interacting with the physical world (Syofyan and Siwi, 2018), which can engage and retain their attention and improve their understanding of the concepts of a lesson. Realizing the full promise of robotics in STEM learning necessitates consideration of teacher preparation. Teachers often report low self-efficacy (Bandura, 1982) in their ability to teach science, which in turn affects students’ science learning (Palmer et al., 2015). Schina et al. (2021b) point to low self-efficacy of teachers in digital technologies and their classroom adoption. Prior research has emphasized the importance of developing familiarity with robotics (Schina et al., 2021a) and the need for robotics training for teachers (Vollmer et al., 2011). While teachers indicate that embedding design, engineering, and technology in the classroom is important, they exhibit low levels of familiarity (VandenBos, 2007) in these domains (Hsu et al., 2010). Lack of teacher training is a key challenge preventing the adoption of robotics in K-12 STEM education (Mataric et al., 2007). Teachers’ low level of familiarity with technology can cause a suboptimal integration of robotics in classrooms that can negatively affect student learning. Many prior studies on educational robotics in K-12 STEM education solely focused either on students or on teachers. Moreover, prior studies have not considered a synergistic combination of robotics and entrepreneurship under the project-based learning (PBL) framework (Barron et al., 1998; Larmer and Mergendoller, 2010). Finally, often teachers may underestimate the potential of students’ ability to succeed in robotics-based learning. This paper is focused on addressing these gaps in the literature. Specifically, collaborative workshops involving teachers and students can reveal to teachers the promise of robotics and entrepreneurship in promoting student learning as well as the ability of students to understand and practice advanced robotics-related concepts. In this spirit, as seen below, we performed comparisons of learning outcomes of students vs. teachers. To bridge the aforementioned gaps in prior research, we designed a four-week long robotics workshop and conducted it annually for three years, each summer. The workshop simultaneously provided professional development (PD) to high school teachers and educational enrichment to their students within a PBL framework focused on robotics design and entrepreneurship. The workshop curriculum, summarized in subsection 3.1.1, included essential elements required to initiate a novice in building and exploring robotics devices. The workshop was expected to provide several benefits to the teachers and students. First, the workshop participants were expected to gain familiarity with and understanding of the robotics fundamentals. Second, the teachers were to be supported during the academic year (AY) to integrate robotics curriculum and projects in classrooms and after-school programs for their students. Third, the students were encouraged to participate in an end-of-year robotics contest. Since the teachers from the workshop are expected to continue providing robotics education and the PD community may seek to upgrade and conduct similar workshops, it is critical to analyze and share the workshop results and help improve future offerings. While being familiar with a disciplinary domain (e.g., robotics) can catalyze one’s participation in further learning about that discipline, cultivating self-efficacy in the discipline can influence one’s learning ability through their persistence in learning tasks. Accordingly, self-efficacy in and familiarity with robotics are relevant factors to gauge the ability and preparation of teachers to effectively deliver and promote robotics learning to students. Thus, in addition to measuring changes in the content knowledge of participants, it is essential to assess changes in their self-efficacy and familiarity as a result of participating in the robotics workshop. Accordingly, using two survey instruments and a technical quiz, we investigate whether the workshop contributes to any changes in participants’ robotics self-efficacy, familiarity, and content knowledge. To examine the impact of AY activities, we consider design artifacts submitted for the end-of-year robotics contests. As evidenced through the examination of the pre- and post-workshop survey and technical quiz responses, for both teacher and student participants, numerous statistically significant changes with medium to very large effect sizes were seen in their levels of robotics self-efficacy, familiarity, and content knowledge. Following the summer workshops, through teachers’ incorporation of robotics learning in formal classrooms, clubs, or after-school programs, many students were introduced to the fundamentals and applications of robotics via hands-on experiential learning. Moreover, each year, student teams participated in robotics contests that integrated entrepreneurial explorations and technology creation while intentionally being guided to make meaningful contributions to their communities and society. The results of observations, discussions, reflections, and reviews of students’ robotics designs, under the summer capstone projects and end-of-year contests, indicate that the students gained an appreciation for the real-world relevance of fundamental science, math, and robotics content knowledge and they had ample opportunities to learn, experience, and hone 21st century skills. The results of this study suggest that robotics education programs can produce significant learning gains among participants, a finding that is promising for education researchers and professional development providers interested in leveraging the potential of robotics in STEM education. The rest of the paper is organized as follows. Section 2 reviews relevant prior literature to give details of various topics related to this work. Section 3 discusses the structure of the program, including the summer workshops, AY activities, and annual robotics contests. Section 4 enumerates the research questions that are used to examine the efficacy of the workshop and AY follow-up. Section 5 details the research methods, including the instruments and data analysis, used to address the research questions of Section 4. Section 6 provides the pre- and post-workshop survey and quiz results for the participants as well as the highlights of annual robotics contests. Section 7 provides a discussion of the study results, including its limitations and implications. Finally, concluding remarks are drawn in Section 8. Literature review The theoretical framework of this work is based on PBL, robotics, entrepreneurship, STEM, social cognitive career theory (SCCT), and teacher PD, and various interactions among them. A brief review of these topics follows. Project-based learning PBL is a framework that is widely used in engineering education as it provides a pedagogical approach to facilitate the process of learning new knowledge and gaining a deep understanding through experiential explorations while solving practical problems. Within the PBL framework, to learn about a topic, participants wrestle with challenging problems by posing questions, discussing concepts, anticipating outcomes, formulating investigations, gathering and analyzing data, synthesizing findings, communicating with peers, refining questions, and building artifacts (Blumenfeld et al., 1991). As observed by Barron et al. (1998), PBL embeds learning opportunities through projects that offer ample scope to readily apply concepts and topics learned in classrooms to real-world situations. For example, Sadler et al. (2000) employed cooperative design projects to engage middle school students in constructing physical prototypes for six varied engineering challenges while helping illustrate links between the design challenges and science concepts. PBL has been deemed as a highly effective approach to learn the process of engineering design through the experience of design as active participants (Dym et al., 2005). Moreover, engineering design projects are known to embed myriad opportunities for students to practice and enhance their science and mathematics knowledge (Akins and Burghardt, 2006). Teachers play a vital role in the implementation of PBL in a classroom environment (Kokotsaki et al., 2016). These roles include assistant, emotional support, task manager, reviewer, facilitator, and information collector (Sabouri et al., 2020). Finally, according to Larmer and Mergendoller (2010), PBL consists of seven major steps (see Fig. 1).Fig. 1 Seven essentials for project-based learning according to Larmer and Mergendoller (2010) Robotics and PBL As an interdisciplinary field, robotics requires foundational knowledge from several disciplines. Formulating robotics-based learning in the context of a real-world project constitutes student-centered learning that can spark students’ intellectual curiosity (Endo et al., 2013), e.g., about the underlying content knowledge, while being engaged in hands-on learning. With robotics, students learn both content and thinking strategies through problem-solving (Barron et al., 1998; Blumenfeld et al., 1991) contextualized in a design project entailing more than one solution (Hmelo-Silver, 2004). By promoting knowledge synthesis from multiple domains (Dym et al., 2005), PBL can foster students’ higher-order cognitive skills (Muldoon et al., 2013). Robotics PBL interventions can increase STEM knowledge in secondary school students (Petre and Price, 2004). Educational robot kits (e.g., LEGO (Perdue, 2007), BoeBot (Lindsay, 2010), and VEX (Innovation First Inc, 2007)) offer myriad components that support students’ creativity and allow the use of PBL. For example, robotics sensors can allow students to connect a textbook formula to a tangible measurement they have generated. Robotics PBL activities provide compelling learning experiences and impart critical academic, life, and professional skills to students (Center for Youth and Communities, 2010). Entrepreneurship and PBL Today’s students effortlessly interact with technological artifacts (Orsak et al., 2004), yet they often lack an understanding of the engineering and business methods underlying these products (Richards et al., 2002). Combining product and business development experience with robotics can impact students’ STEM abilities and career awareness. Exposing students to the world of technology creation, combined with communication and business skills, can allow them to envision concrete applications of STEM knowledge and skills. PBL integrated with entrepreneurial training addresses a combination of fear of failure, lack of confidence, creativity, flexibility, innovation, and communication skills (Beary, 2013; Nair et al., 2017). Entrepreneurship education programs (EEP) increase multiple markers of education and career success, e.g., self-efficacy, need for achievement, risk-taking propensity, persistence, and being proactive (Beary, 2013; Huber et al., 2014). Exposing students to EEP engenders positive intentions and attitudes that influence their perceptions of desirability and feasibility towards entrepreneurial careers (Cioffi et al., 2014; Krueger, 1993; Pruett, 2012). The Network for Teaching Entrepreneurship (NFTE) has conducted EEP for K-12 students nationally. A study by Nakkula et al. (2003) documents that students showed increased interest in attending college and increased time spent independently reading after NFTE program participation. In a longitudinal study of ≈1,300 former NFTE students, compared to national averages, NFTE program participation was found to be associated with higher rates of high school graduation, STEM degree attainment, self-employment, and higher average income (Beary, 2013). These participants, as a whole, also created hundreds of jobs for others. PBL alone is not sufficient for preparing today’s students to create technology companies of future that produce thousands of well-paying jobs. While PBL may allow students to envision and build innovative products, lacking entrepreneurial strategies their products may not reach customers. Thus, the integration of entrepreneurship into the curriculum is paramount. Robotics and entrepreneurship Robotics, while still a nascent technology, constitutes a platform for innovation to address niche problems with potential to develop into successful businesses. As labor costs rise abroad (Martin, 2012) and use of robots in advanced manufacturing becomes feasible (Javaid et al., 2021), robots have the potential to transform the U.S. manufacturing industry. The decreasing cost of electronics and hardware components, along with advancements in additive manufacturing, promises to lower the cost of robot hardware, making robots economical for diverse industries (Martin, 2012). Students’ fascination with robots can be used to stimulate their interest not only in STEM disciplines but also as a safe context to experience the entrepreneurial process. Robot design activities can embed a comprehensive experience in system integration and product development that can be applied, through the EEP module, to diverse real-world projects. Through PBL experiences, students can practice and hone their creativity, inventiveness, and entrepreneurial skills as well as envision and gain an appreciation of the pathway from STEM education to careers. Robotics and STEM learning Since many STEM principles are inherently incorporated into performing simple tasks with a robot, robotics can illustrate connections between STEM fields and practical applications of classroom learning. For example, students can connect classroom science and math topics to the robotics concepts of travel distance, drive mechanism, and sensor operation. With robotics, students apply tools used by STEM practitioners to visualize and practice STEM concepts that they otherwise find difficult to comprehend (Faisal et al., 2012). Participation in after-school robotics contests is often the only venue where students can explore advanced tools used by engineers and experience engineering design as a model for problem-solving (Benitti, 2012). Such learning opportunities can be made available to students by preparing teachers to adapt a yearlong course in their schools. The NGSS integrate engineering design in science standards, promote science learning in an engineering context, and connect applications of math to science learning. Transition of robotics activities from an after-school setting into the classroom can support this vision of NGSS. Social cognitive career theory SCCT (Lent et al., 2002) provides an empirically based framework wherein cognitive, affective, and contextual factors (e.g., self-efficacy beliefs, outcome expectations, interests, choice goals, and social supports/barriers) mediate students’ selection of academic majors and careers. SCCT offers interest-, choice-, and performance-models, with complex interplay between cognitive and social factors, to reveal the career development pathways from initial interest formation, to activity choice, and finally performing and persisting in a chosen discipline. SCCT studies have revealed cognitive and contextual factors, and their interconnections, that bear on disciplinary selection, action, and persistence across racial and gender groups (Inda et al., 2013; Lent et al., 2011; Wang, 2013). Thus, it is critical to i) promote early STEM interest formation among underrepresented minority (URM) and female students by using robotics to stimulate situational interest (Subramaniam, 2009); ii) build self-efficacy beliefs by offering social support via interactions with well-prepared teachers, graduate students, and researchers; and iii) enable students to envision diverse academic and career options in STEM fields. In Chemers et al. (2011), science self-efficacy and science identity were deemed strong predictors of commitment to undergraduate and graduate STEM studies among 665 URM students. Hands-on experiences with robotics and entrepreneurship can ignite students’ STEM interests, yield gains in their STEM self-efficacy, and promote their STEM self-identity, increasing STEM career likelihood. Self-efficacy Bandura (1982) characterized self-efficacy as one’s belief in their ability to “organize and execute courses of action required to deal with prospective situations […]”. Self-efficacy considers one’s assessment of their ability to execute planned actions to succeed in completing the task they have undertaken (Palmer et al., 2015). It is determined from ratings individuals assign to themselves to complete tasks along the four dimensions of confidence, motivation, outcome expectancy, and anxiousness (Carberry et al., 2010). In robotics education, robotics design is integral to both learning and development of robots. Similar to engineering design (Carberry et al., 2010), to meet robotics design requirements, one must consider individual components, subassemblies, and integration of subsystems. One’s self-efficacy in engineering design can impact their learning of engineering itself (Carberry et al., 2010). Familiarity Familiarity refers to a form of remembering an event, individual, or object that renders a sense of awareness without a specific recall from memory (VandenBos, 2007). That is, familiarity suggests knowledge about something. Cognitive fluency, prototypicality, and habit all contribute to being familiar with something (Mullin, 2015). Robotics-based teacher professional development Educators are increasingly seeking to adopt robotics to teach STEM concepts across all education levels (Brophy et al., 2008; Cejka and Rogers, 2005; Erwin et al., 2000; Verner et al., 2007). Yet, the potential of robotics remains largely untapped in the school curriculum (Norton et al., 2007), for example as a stand-alone course in high schools. One reason preventing the adoption of robotics in K-12 education is that teachers often find it difficult to connect robotics activities to curriculum outcomes (Norton et al., 2007). The sustainability of robotics activities in K-12 education also depends on the quality of teacher PD. The limited knowledge of teachers in engineering and robotics affects student learning and creates a workforce that is unprepared for STEM occupations. The need for and challenge of robotics PD programs are well documented in the literature (Mataric et al., 2007; Stolkin et al., 2007; Vollmer et al., 2011). Only through teacher PD and comfort with the material and technology can there be curriculum implementation, a multiplier effect, and sustainability (Mataric et al., 2007; Vollmer et al., 2011). Teacher PD programs require significant upgrading to embed engineering content in the teaching of science and math (Kelley and Wicklein, 2009; National Research Council, 2001). To enhance teacher effectiveness, PD programs should support participants in acquiring a conceptual understanding of disciplinary fundamentals and deep knowledge of curricular content, gaining fluency in pedagogical content knowledge, and mastering rapidly evolving tools to promote hands-on learning using contemporary technology (Shulman, 1986; Valdez et al., 1999). Ideally, PD programs need to develop teachers’ capacity to connect STEM learning with real-world problems that are personally meaningful to students. Moreover, PD should support transfer of training by immersing participants in content knowledge, allow modeling and practice of desired skills (e.g., problem-solving and active-learning), promote collective participation, and last for sufficient duration to handle the cognitive demands of new learning (Guskey and Yoon, 2009; Loucks-Horsley et al., 2009). Program structure Informed by the literature review, the main objectives of our program were to allow the participants to learn fundamentals of robotics and entrepreneurship, and augment this learning through hands-on PBL for increasing their robotics self-efficacy, familiarity, and content knowledge. We conducted a robotics workshop in summer for three consecutive years. Each year, 10 to 18 teachers and 22 to 36 students from 8 to 12 urban high schools attended the workshop. Each year’s workshop included two weeks of guided training and two weeks of collaborative robotics capstone projects. Mentored by an engineering and a science education faculty, each year, four to six undergraduate and graduate engineering students served as workshop facilitators, delivering lectures and supervising hands-on learning. Teachers were expected to deliver the lessons from the workshop to their students in classrooms or after-school programs during the AY. Thus, each teacher was asked to bring two students to the workshop who were to serve as classroom supports during the AY, especially during the hands-on learning sessions. A typical high school class may have 20 to 30 students, requiring three individuals (one teacher and two students) to lead, support, and manage hands-on robotics sessions. For example, a teacher’s student partners from the summer workshop could support them in: distributing and tracking robot components to the class; examining and approving electrical connections of robotics creations of students; and ensuring adherence to various safety precautions for the robot. Throughout the summer workshops, there was ample evidence that often the students absorbed robotics fundamentals and interacted with sensors, motors, electronics, and microcontroller components with a greater ease than the teachers. Thus, these students became a valuable classroom resource for the teachers who were teaching robotics for the first time. At the end of each AY, the schools participated in a robotics-focused entrepreneurship contest. While the workshop was conducted for three years, the AY follow-up was completed for only two years. Due to COVID-19 pandemic, the AY follow-up and end-of-year contest were canceled in March 2020. Thus, the paper presents results from three years of workshop and two years of end-of-year robotics contest. Summer workshop Guided training The guided training included twenty half-day sessions, each having two parts. The first hour-long part provided short introductions to the scientific, mathematical, and engineering concepts underpinning the session’s focus area. The second three-hour-long part augmented learning through hands-on explorations. The guided training curriculum included the following aspects, all with relevance to robotics: i) mechanical (a refresher on physics of motion and introduction to drive mechanics); ii) electro-mechanical (electrical and electronic components and circuitry, sensors, and motors); iii) computing (Arduino microcontroller and programming (Margolis, 2011)); and iv) applications (introduction, challenge, and capstone). See Fig. 2 and Mallik et al. (2018) for further details of the robotics curriculum. Corresponding to each concept lesson, for experiential learning, the participants performed hands-on activities in teams of two teachers and four to five students. They were provided worksheets containing the principles of session’s concepts, some elementary exploration activities, and instructions for hands-on learning. Experimental demonstrations and activities that followed the introductory lectures were supposed to strengthen participants’ foundations in robotics and stimulate their interest in the workshop. Figure 3a and b show some participants during a lecture session and performing hands-on learning, respectively. Beginning with introductory lessons on robotics fundamentals (Martin, 2000), the participants gradually learned and practiced the major concepts of mobile robots (Siegwart et al., 2011). By day eight, they made their robot follow a black line on the ground and their robot gripper open/close under user or program control. On the last two days, a technology management faculty delivered an entrepreneurship module wherein the participants learned about business planning, business model canvas (Osterwalder and Pigneur, 2010), market analysis, product-market matrix (Ansoff, 1957), Porter’s 5 forces (Porter, 1989), technology S-curve (Schilling and Esmundo, 2009), product development process, and raising capital. They also learned about business incubators, managing intellectual property, and social entrepreneurship. Next, the participants engaged in experiential learning to create and present pitches for their proposed new ventures to the facilitators and peers.Fig. 2 Robotics content addressed in the workshop Fig. 3 Participants attending guided training sessions Incorporation of entrepreneurship can make robotics designs personally relevant and meaningful for participants, specially during capstone projects and robotics contests. Familiarity with and knowledge of entrepreneurship can help students conceive and prototype products with potential market demands. Performing robotics learning within an entrepreneurship context can also support various steps of PBL, e.g., robotics designs that are purpose-driven and have personal relevance to students; activities that hone 21st century skills; and engagement of students in public demonstrations of their creations; among others. In this manner, the adoption of entrepreneurship in the curriculum offers a unique mechanism to engage students in robotics experiences through the PBL framework. Robotics capstone projects During the capstone project, teacher-student teams performed engineering design, prototyping, and testing to develop the mechanical structure, drive mechanism, sensor and drive electronics, computer interface, and microcontroller program, which were integrated to produce robots for competing on a mock-up game field inspired from real-world scenarios. Each year, the project staff created and offered different socially relevant robotics challenges. For example, in year three (Y3), the participants developed robots to move along a layout of streets, pick up trash bins from houses along the streets, sort the trash, and deliver it to a sorting facility based on the type of recyclables. Figure 4 shows the mock street view resembling the grid-like streets. Trash bins containing recyclable garbage were modeled using cups placed randomly at the end of the branch in front of houses. The types of trash in the bins were different reflecting real-world scenarios. If a trash bin was present near the house, the robot needed to pick it up and deliver it to the sorting facility classifying the trash bins. The robot needed to identify three different types of recyclable garbage and the method for identifying the garbage type was left open-ended. The robot could traverse the arena by line following using various sensor combinations to make a turn, move forward, or stop.Fig. 4 Capstone project for year three workshop Alignment of workshop activities with PBL Each year’s capstone project statement was prepared with the seven-step PBL process of Fig. 1 in mind. For example, automating the familiar activity of garbage pickup trucks working in their neighborhoods sought to build on the prior knowledge of participants for engaging their interests. Open-ended aspects of the project encouraged participants to reflect on the scenario and produce creative solutions. For example, participants created varied algorithms for robots to follow lines on the street grid and made use of bin color or bin material to distinguish the trash type. The project engaged students to use advanced robotics technology while learning, experiencing, and honing 21st century skills by collaborating with peers, communicating design concepts to teachers, and critically thinking to defend their designs in front of the facilitators. The guided learning and capstone projects aroused participants’ intellectual curiosity and allowed them to answer their own questions through experiential learning. They explored applications of theoretical learning, revised their understanding of learned concepts, refined their hands-on skills for robot operation, prototyped design and programming ideas, assessed their experimental implementations, and drew conclusions. As participants learned that their initial designs usually did not produce the best products, they were encouraged to follow a systematic process to iteratively refine their design with feedback from peers and facilitators. Finally, they presented their product to the facilitators and others, received feedback on its quality and performance, and reflected on their experiences. The participating teachers engaged in self- and student-assessments as well. Thus, under the PBL umbrella, the participants gained knowledge, skills, experiences, and pride as they successfully worked on various challenges. Academic year activities Some teachers introduced a robotics elective course, which met for two to five days a week, while others conducted after-school robotics programs. All teachers used robotics kits from the workshop and some procured additional robots. They adapted the workshop curriculum for their students’ needs. The students who attended the workshop supported their teachers during the hands-on learning sessions. Each year, four to five graduate engineering facilitators weekly visited schools to observe and support classroom robotics activities. The facilitators and teachers also collaborated to create worksheets, conduct hands-on activities, debug circuits and codes, and assess student work. Each cohort of teachers and facilitators met four times during the AY, along with the project faculty, to share their progress, challenges, and suggested solutions. Most teachers sought to complete the classroom component of their robotics elective by February to allow students to devote the remaining academic calendar to prepare for the end-of-year robotics contest. Table 1 summarizes the demographic information of the students engaged by the teachers at their schools.Table 1 Demographic information of the students in school Year Total Male Female Underrepresented minority Y1 241 120 121 127 Y2 214 82 132 190 Y3 163 80 83 116 Total 618 282 336 433 Robotics contest As evidenced above, PBL offers a compelling framework to learn about and experience robotics, while the adoption of entrepreneurship engenders personal relevance. Integrating entrepreneurship experiences with technology education can expose learners to the world of technology creation while promoting the development of STEM knowledge and skills with a view toward identifying and solving societal challenges. Thus, under the entrepreneurship umbrella, as students identify problems of societal relevance that can be addressed through robotics, they find personal meaning in their robotics explorations. This approach engenders a need to know robotics and exposes various STEM concepts that students need to understand. The contest creates an opportunity for students to take ownership of their learning for being successful in the contest. Thus, through PBL, students envision and build innovative products using robotics knowledge and employ entrepreneurial strategies to help their products reach customers. In this vein, the participants were challenged to identify a home-related problem in Y1 and for a school-related problem in Y2. Since developing participants’ entrepreneurship skills was an integral element of this effort, they had to characterize their solution for its market potential in the form of a presentation to a team of judges. Description of contest requirements Student teams from each school had the opportunity to develop several prototypes, however, each school could showcase only one prototype in the contest. The budget to purchase additional robot prototype material was limited to $20. Each team was allowed three to five students to represent their school. The contest prize consisted of paid internships at local start-ups. Winners were selected based on: best overall pitch–first prize, best robotics engineering design pitch–second prize, and best entrepreneurship pitch–third prize. Identifying and selecting a problem Teams were advised to identify problems through surveys and analyze opportunities for robotics solutions to make a positive impact. Using suggestions of peers, family, and neighbors students selected appealing options. Next, they developed a problem statement, suggested solution strategies, and selected one deemed the best. The user surveys formed an important factor for the contest, thus the students needed to conduct them diligently. They also needed to explain the iterations involved in coming up with the solution. They focused on selecting a problem that they deemed feasible and interesting to address with an innovative robotics solution. Designing and building Each team had to design, build, program, and test a robotics device to address the identified problem. Their design had to be a mechanically stable solution with a practical application and people having an interest to use it. Teams had to submit and present a video demonstration of their design. Marketing After identifying a problem and building a prototype, each team had to explore concepts of marketing, customer base, and pricing for creating a marketing plan for their product, a key aspect of entrepreneurship. Moreover, they had to research requirements of budget, materials, resources, and advertising strategy. As a final step, they needed to tie together all their learning to develop a pitch that would persuade people to buy their product. Research questions The purpose of this study is to examine the efficacy of the four-week long workshops and AY follow-up on the abilities of teachers and students to participate in and perform robotics-based learning activities. Thus, the study analyzes the workshop outcomes on participants’ robotics self-efficacy, familiarity, and content knowledge. Moreover, it examines their design solutions for the robotics contests. The following two research questions are considered. What is the extent of change in participants’ robotics self-efficacy, familiarity, and content knowledge after they participate in the summer workshop? How successful were the participants in designing, communicating, demonstrating, and pitching their products after participating in the workshop and AY follow-up? Research methods Participants Over three years, the workshop was attended by 44 teachers of whom 43 teachers responded fully to various data collection instruments. The demographic information for the 44 teacher participants is summarized in Table 2. A total of 96 students attended the workshop over three years of whom 81 students responded fully to the data collection instruments. The demographic information reported by 91 students is summarized in Table 3.Table 2 Demographic and discipline information of the teacherss Year No. Gender Race Subject Male Female Other White African American Hispanic Asian Other/Mixed Math Science Other Y1 18 10 7 1 8 4 1 2 3 6 12 0 Y2 16 9 7 0 6 3 4 0 3 8 7 1 Y3 10 5 5 0 3 2 2 3 0 5 5 0 Total 44 24 19 1 17 9 7 5 6 19 24 1 Table 3 Demographic information of the students Year No. Gender Race Male Female Other White African American Hispanic Asian Other/Mixed Y1 33 18 15 0 6 9 9 5 4 Y2 36 24 12 0 7 8 7 7 7 Y3 22 14 7 1 3 0 5 7 7 Total 91 56 34 1 16 17 21 19 18 Instruments Since the teachers were expected to deliver robotics education to students during the AY, the workshop integrated instruction and hands-on exploration in robotics and entrepreneurship to increase their self-efficacy, familiarity, and content knowledge. It was paramount to assess participants’ self-efficacy perceptions to establish any possible benefits engendered through the workshop. Moreover, the results of such a study can be used to enhance the content, structure, and organization of the workshop’s future offerings. Even as teachers’ attitudes and beliefs about the significance of robotics and technology are important factors in PD programs (Vollmer et al., 2011), self-efficacy is not the only factor that impacts student learning. As indicated previously, familiarity suggests knowledge about something. Moreover, teachers’ knowledge of the disciplinary content is correlated to their students’ learning gains (Chaney, 1995; Hill et al., 2005). Thus, we also included a familiarity survey and a technical quiz. In summary, the participant responses were obtained for the robotics self-efficacy, familiarity, and content knowledge instruments. While the self-efficacy survey of this study was adapted from Carberry et al. (2010), the familiarity survey and robotics content quiz were self-designed. Sample items from the self-efficacy survey, familiarity survey, and quiz are shown in Fig. 5. Finally, a self-designed rubric, detailed below, was utilized to assess the entries for robotics contests. See below and Mallik et al. (2018) for further details.Fig. 5 Sample items from the self-efficacy survey, familiarity survey, and quiz Self-efficacy The self-efficacy survey included four dimensions to obtain respondent ratings, on a scale of 0 to 100, for their perceived confidence, motivation, success expectation, and anxiety for several items of project-based robotics design (e.g., conducting robotics design, identifying a robotics need, researching a robotics design need, developing robotics design solutions, selecting the best possible robotics design, constructing a robotics prototype, evaluating and testing a robotics design, communicating a robotics design, and redesigning a robot). While the confidence, motivation, and success expectation scores are positively directed, the anxiety-related questions are negatively directed. The anxiety responses are reverse coded by subtracting them from 100, making all four dimensions positively directed. Familiarity with robotics components and concepts Since teachers’ disciplinary familiarity impacts student learning, we administered a survey to determine pre/post-workshop changes in respondents’ level of familiarity with the common robotics concepts. Based on the workshop curriculum, familiarity survey had four sections: mechanical, electrical/electronic, sensor/actuator/microcontroller, and programming. It included 21 items on a five-point Likert scale and respondents were asked to rate their level of familiarity with each item (0 = not familiar, 5 = very familiar). Robotics content knowledge quiz As teacher content knowledge is an important factor for student learning, we designed a quiz to assess respondents’ knowledge of robotics. The quiz had 30 questions, organized in four sections, to probe respondents’ understanding of the concepts delivered during the workshop. The drive mechanism section had five items on drive train components, differential drive, types of drive mechanisms, and types of motion. The electronics section had 10 items on electrical and electronic components. The actuation/manipulation/localization section had seven items on the terminology, components, and operations related to robots. Finally, the gears/motors section had eight items on scalar, vector, motion, energy, torque, and speed concepts. Performance of individual respondents in the pre/post-workshop quiz gives an effective measurement of their ability to sustain learned knowledge. From the average performance of all the respondents in the quiz, we can gauge the effectiveness of the PD program. Robotics contest rubric The robotics contest served as an important data source that included students’: user survey results, robotics artifacts highlighting engineering designs and creativity, a marketing pitch, as well as demonstrations and presentations. Each year, during the contest, teams presented their work to three judges with experiences in entrepreneurship, start-ups, and robotics. Each team analyzed its solution for market potential and presented a pitch by combining all their learning. They explained how they came up with user surveys and analyzed their results. They described design iterations, challenges, and successes in developing prototypes. The judges sought to understand the iterations involved in coming up with the solution. The teams had to explain why they chose the solution, literature research, and the innovation in the product. The judges used the rubric of Appendix A to select the winners. Data collection and analysis methods All survey and quiz data were collected from the participants anonymously using Qualtrics. To examine whether the workshop was effective in increasing participants’ robotics self-efficacy, familiarity, and content knowledge, their responses were obtained at the start and end of the workshop. To examine any statistically significant differences between the means of the pre/post-tests, paired t-tests were performed. The choice of t-tests is guided by our interest in finding the difference between two measurements for the same subject (Elliott and Woodward, 2007), specifically, statistically significant mean differences between the pre/post-test responses of the surveys and quiz. Next, we examined any variations in results for each year. Moreover, we compared the teacher vs. student outcomes and analyzed their correlation. Finally, to assess the quality of student entries in the contest, the judges used the rubrics of Appendix A. Results Teacher results The results summarized in Table 4 show that the teachers had significant pre/post-test gains in robotics self-efficacy, familiarity, and content knowledge. The t-test values in Table 4 show statistical significance for all cases except for self-efficacy and technical quiz scores of Y3 and we reject the null hypothesis for all cases except these two. The Cohen’s d values (Cohen, 1988) in Table 4 indicate a medium effect size of the treatment on teachers’ self-efficacy and a large to very large effect size for all other cases.Table 4 Survey and quiz results for teachers Instrument Scale Pre-test average Post-test average n t p Cohen’s d Year 1 Self-efficacy 100 58.97 68.63 18 2.66 <0.025 0.73 (medium) Robotics familiarity 5 1.74 3.37 18 4.73 <0.01 1.57 (very large) Technical quiz 30 14.39 18.17 18 2.69 <0.025 0.99 (large) Year 2 Self-efficacy 100 51.15 65.35 15 3.47 <0.01 0.797 (medium) Robotics familiarity 5 1.82 3.47 15 7.24 <0.01 1.59 (very large) Technical quiz 30 13.53 18.33 15 5.18 <0.01 1.10 (very large) Year 3 Self-efficacy 100 63.19 72.17 10 1.79 n.s. – Robotics familiarity 5 1.99 3.44 10 3.49 <0.01 1.16 (very large) Technical quiz 30 14.4 18.6 10 2.02 n.s. – Cumulative for three years Self-efficacy 100 57.22 68.31 43 4.70 <0.01 0.68 (medium) Robotics familiarity 5 1.83 3.42 43 8.53 <0.01 1.5 (very large) Technical quiz 30 14.09 18.33 43 5.26 <0.01 0.998 (large) Teacher self-efficacy survey results Figure 6 shows that the confidence dimension had the largest improvement each year, stemming from a low confidence level at the start of teachers’ foray into robotics. However, tasting success in building and programming robots boosted their confidence level. Figure 6 shows that the motivation dimension was higher than other dimensions in pre/post-tests for all three years. This is expected as the teachers self-selected to apply to this program. However, in Y1, the motivation dimension dropped slightly in the post-test (Fig. 6). This may have resulted from our failure to integrate motivational talks providing verbal affirmation of teachers’ progress—a strategy that was included in the early part of the workshop. Teachers’ motivation levels were high when such motivational talks were enacted. Yet, as they handled challenging capstone projects, perhaps their motivation declined and including motivational talks may have boosted their morale. Thus, in Y2, we embedded verbal affirmation throughout the workshop to ensure that the participants did not feel demotivated. Figure 6 shows a slight improvement in teacher motivation in pre/post-tests in Y2. Comparing pre/post-test scores for success in Fig. 6, considerable improvement is seen. The Y2 pre-test shows a drop in the success dimension compared to the Y1 pre-test. However, this drop was overcome in the Y2 post-test. Keeping the lower pre-test success score of Y2 in mind, in Y3 we emphasized to participants that they need not worry excessively about their success in robotics activities since we had planned scaffolds to support them. At each stage of learning, facilitators were available to help the participants overcome any obstacles and participants simply needed to stay focused on each day’s instruction and guidance. As a result, success scores in pre/post-tests improved in Y3. For each year, Fig. 6 shows a small improvement for the anxiety dimension of pre/post-tests. We posit that as most teachers were working with robotics for the first time, they were anxious about their ability to employ robotics in classroom.Fig. 6 Teacher self-efficacy section wise average score for each year Teacher familiarity survey results Table 4 shows that the pre-test familiarity score in Y1 was lower in comparison to Y2 and Y3. Moreover, even as each year’s post-test familiarity score was higher than the corresponding pre-test familiarity score, Y2 and Y3 post-test scores were also higher than Y1 post-test score. Nonetheless, the increase in pre/post-tests showed the largest percentage improvement in Y1 as seen in Table 6. With the pre-test score being highest for Y3, the resulting pre/post-tests improvement was only 72.86% (Table 6). Figure 7 reveals that in the pre-test, the teachers had particularly low scores on two sections: sensor/actuator/microcontroller and programming and they exhibited a significant improvement in these sections in the post-test.Fig. 7 Teacher familiarity section wise average score for each year Teacher technical quiz results As seen from Fig. 8, teachers demonstrated the best performance and the most improvement in the drive mechanism section of pre/post-test. There was a small drop in pre/post-tests average in gears/motors section in Y3. The results in Table 6 show an improvement of 30.09% for the teachers for all the three years combined in the technical quiz.Fig. 8 Teacher quiz section wise average score for each year Table 6 shows that among the three areas considered (two surveys and one quiz), the familiarity aspect showed maximum percentage improvement for each year and for all three years combined with 86.89% improvement. Student results Similar to the analysis of results for teacher gains, we analyzed the results for student gains. We had matching pre/post-test data for 81 students and these are summarized in Table 5 and Figs 9, 10 and 11. The t-test values in Table 5 show statistical significance for all cases except for Y3 self-efficacy scores and we reject the null hypothesis for all cases except this one. The Cohen’s d values in Table 5 indicate medium to very large effect sizes for all treatments.Table 5 Survey and quiz results for students Instrument Scale Pre-test average Post-test average n t p Cohen’s d Year 1 Self-efficacy 100 63.92 73.24 26 2.40 <0.025 0.65 (medium) Robotics familiarity 5 2.46 3.61 26 5.01 <0.01 1.27 (very large) Technical quiz 30 10.42 13.62 26 5.46 <0.01 1.04 (very large) Year 2 Self-efficacy 100 58.43 68.67 33 4.82 <0.01 0.8 (large) Robotics familiarity 5 1.63 3.43 33 11.65 <0.01 1.90 (very large) Technical quiz 30 10.52 15.64 33 4.11 <0.01 1.04 (very large) Year 3 Self-efficacy 100 67.09 71.72 22 1.24 n.s. – Robotics familiarity 5 1.84 3.49 22 6.51 <0.01 1.67 (very large) Technical quiz 30 11.82 15.82 22 3.88 <0.01 0.86 (large) Cumulative for three years Self-efficacy 100 62.55 70.96 81 4.63 <0.01 0.59 (medium) Robotics familiarity 5 1.95 3.5 81 11.59 <0.01 1.6 (very large) Technical quiz 30 10.84 15.04 81 6.91 <0.01 0.96 (large) Fig. 9 Student self-efficacy section wise average score for each year Fig. 10 Student familiarity section wise average score for each year Fig. 11 Student quiz section wise average score for each year Student self-efficacy survey results Figure 9 shows an improvement in scores for all dimensions of self-efficacy for all three years. Similar to the case of teachers, the largest improvement in pre/post-tests is seen in the confidence dimension for students in all three years. However, unlike teachers, student pre/post-test scores do not show any decline in the motivation scores. The pre-test success score of students in Y2 is seen to be lower than in Y1, which is similar to what we found for teachers. This issue was addressed by ensuring that the participants had access to facilitators whenever they faced problems, and this improved their Y2 post-test success scores. Finally, the students had the lowest score in pre/post-tests for the anxiety dimension. Next, note that the students generally scored better than the teachers on all dimensions of self-efficacy in pre/post-tests for all three years. Even though in Table 5 the post-test self-efficacy scores for the students are generally higher than those of the teachers (see Table 4), since the students also had generally higher pre-test self-efficacy scores, the percentage improvement results for self-efficacy are seen to be better for teachers in Table 6.Table 6 Percentage improvement in the results Teacher Student Instrument Y1 Y2 Y3 Combined Y1 Y2 Y3 Combined Self-efficacy 16.38 27.76 14.21 19.38 14.58 17.53 6.9 13.45 Robotics familiarity 93.68 90.66 72.86 86.89 46.75 110.43 89.67 79.49 Technical quiz 26.27 35.48 29.17 30.09 30.71 48.67 33.84 38.74 Student familiarity survey results In Table 5 we see that the familiarity score of students in Y1 pre-test was high (2.46) compared to other years and also compared to teachers’ pre-test familiarity score. Thus, even with the largest post-test student familiarity score in Y1 (3.61), the first year data produced the smallest percentage improvement (46.75%) (Table 6). From Fig. 10, we see that student responses produced a trend similar to that of teacher responses. Specifically, in the pre-test familiarity survey, they indicated lower levels of familiarity with the topics of sensor/actuator/microcontroller and programming and then demonstrated a significant improvement on these topics in the post-test responses. Since the students already had some ideas about the mechanical, electrical, and electronics topics, possibly from high school physics courses, they responded with higher levels of familiarity for these topics. The remaining topics were new to them and they seemed to have gained good levels of familiarity with them. Since the student responses to the familiarity pre-test for Y2 were quite low, while their Y2 post-test responses were similar to Y3, we find that in Table 6 the Y2 familiarity showed the largest percentage improvement of all cases 110.43%. Student technical quiz results The technical quiz results in Fig. 11 show the largest improvement in the actuation/manipulation/localization section and relatively minor improvement in the gear/motor section. Overall, for all three years combined, the technical quiz results show an improvement of 38.74% (Table 6). Finally, using the data in Tables 4 and 5, correlation analyses (Liero and Zwanzig, 2012) for the scores of teachers vs. students are performed. Their average scores for self-efficacy, familiarity, and robotics quiz are found to be highly correlated with the correlation coefficients of 0.977, 0.944, and 0.957, respectively. Robotics contest results Seven teams participated in the contest in spring 2018 and eight teams in spring 2019. A brief description of the first-ranked project for each year is provided below as an illustration of the quality of student work. Pupbot: Pet owners often avoid overnight travel because they cannot leave their pets home alone. According to user survey results, pet owners want a solution that can take care of and give company to pets in the absence of owners. This team created a complete set of robot solutions called Pupbot (see Fig. 12a) to feed, watch, and play with the pet. They mounted a camera on a mobile robot for remote surveillance of the pet, used an ultrasonic sensor to control a food dispenser, and installed a peristaltic pump to control a water dispenser. Moreover, they endowed their mobile robot with an arm mechanism to throw a ball to engage the pet in play. During the contest presentation, the team demonstrated a thorough mastery of robotics concepts, especially, the knowledge of various sensors, and successfully put their learning into practice to address a customer need. They did an extensive literature search and clearly articulated the survey analysis in the presentation. They exhibited excellent collaboration during their prototype demonstration. Overall, the team provided a compelling and informative presentation with careful considerations of entrepreneurial and business strategies.Fig. 12 Winning entries of the robotics contests Detec Tech: Based on the surveys conducted at their high school, this team recognized a need to report students’ unauthorized presence in school hallways. In response, Detec Tech robot (see Fig. 12b) was developed as a hallway monitor to determine if a student wandering in the hallways left a classroom without a pass. The robot is instrumented with a timer that sets off upon spotting a student outside classrooms. The student can turn off the robot using their class pass within a one-minute time window. If the one-minute timer finishes before being turned off, the robot reports to the main office to alert the staff. This team showed novelty in combining machine vision technology with the knowledge gained through the robotics curriculum. In addition to gaining a deep understanding of STEM concepts such as mechanical design, sensor technology, and programming, they successfully employed impressive entrepreneurship concepts by proposing marketing strategies, such as various pricing models for different customers, to increase their customer base. Their pitch video successfully emphasized the need for their proposed solution and how it can provide students and teachers with a good study environment. Finally, their presentation showcased their excellent communications abilities. The entries to and presentations for the contest were assessed using the rubrics of Appendix A. The five categories of the rubrics, viz., robotics engineering, innovation/creativity, entrepreneurship, video/presentation, and written assignments can also be utilized to assess participants’ robotics self-efficacy, familiarity, and content knowledge. Specifically, an indirect measure of participants’ self-efficacy is available from the observations of judges concerning: the confidence that contest participants show during their presentation; the motivation that they demonstrate in persisting to solve a challenging real-world problem; and their expressions of anxiety during the presentation and prototype demonstration. The judges also had opportunities to gauge participants’ robotics familiarity and content knowledge through detailed technical probing in the robotics engineering section of the rubrics, e.g., concerning the programming sophistication, sensors technology, and actuation knowledge, among others. Based on the overall quality of team submissions, confidential scores assigned by the judges, and the project team’s observations, on average, the student participants were deemed to be highly successful across all aspects of the robotics contest. Discussion Robotics is increasingly proving to be a powerful tool (Chambers et al., 2007; Vollmer et al., 2011) for transforming STEM teaching and learning. Educational robotics can be used for hands-on activities that effectively impart engaging STEM learning experiences (Chambers et al., 2007). The potential of a robotics based curriculum to increase the achievement scores of students in an after-school program has been examined in Barker and Ansorge (2007), however this prior research did not consider formal in-school educational programming. An educational robot based learning system for in-school teaching of students has shown promise to influence motivational factors among students (Chin et al., 2014). To meet the demands of tomorrow’s workforce needs, it is imperative that teachers adopt robotics technologies to impart STEM knowledge to their students (Mataric et al., 2007; Schina et al., 2021a; Vollmer et al., 2011). Effective incorporation of robotics in curriculum, teaching, and learning requires that teachers develop comfort and familiarity with and knowledge of robotics. Efficient PD can enhance technology (e.g., robotics) knowledge of teachers and promote successful integration of technology in STEM curriculum (Dana et al., 2001). However, there is limited prior research on the development and examination of PD that permits teachers to create curricula materials and hone classroom practices for robotics-based teaching and learning. While You et al. (2021) recently developed and examined a robotics-based PD program, it included only teachers as summer program participants and sought to embed robotics in middle school science and math classrooms. Regarding the role of entrepreneurship in K-12 education, the development of entrepreneurship knowledge among primary school students has been shown to yield a robust positive effect on their non-cognitive skills (Huber et al., 2014). Moreover, the impact of entrepreneurship workshops, conducted with support from college students, on the entrepreneurship self-efficacy of middle and high school students has been examined (Cioffi et al., 2014). Yet, prior research has not considered professional development of teachers and educational enrichment of students in programs that synergistically combine robotics and entrepreneurship under the PBL framework. Thus, we developed a program to enhance teachers’ capacity to engage students in robotics through PBL (Larmer and Mergendoller, 2010). We brainstormed to design the PD curriculum and structure and iteratively revised it using participant feedback. We had clearly defined goals of formulating robotics activities under PBL that would help participants gain theoretical knowledge, thinking strategies, and entrepreneurship skills. Participants received numerous opportunities for professionalization and partnership by interacting with engineering students and faculty. Since effective PD can support the transfer of training (Loucks-Horsley et al., 2009), we purposely assessed and supported teachers to enhance their self-efficacy, familiarity, and content knowledge. Self-efficacy Individuals with low self-efficacy in a discipline do not: believe that they can learn the content of that discipline, trust their abilities to teach the discipline, and have confidence in their students’ ability to learn the discipline (Palmer et al., 2015). Self-efficacy encompasses nuanced issues, thus we considered participants’ self-efficacy in learning robotics concepts by embedding throughout the project the following four components of self-efficacy (Bandura, 1977). i) Performance accomplishment is related to an individual experiencing success in assigned tasks (Bandura, 1977). During guided training, capstone experience, and end-of-year contest, the participants conducted experiential learning activities assembling, programming, and operating robotics devices for real-world situations to credibly perform assigned tasks. Throughout the project, they were prompted to answer questions and explain their hands-on learning activities. Moreover, they took a technical quiz in robotics before and after the workshop. Project activities gradually and purposefully increased in difficulty from low, to moderate, and finally to high. Many participants were undertaking robotics activities for the first time but with practice, and by following instructions, they became adept in performing them. As they experienced success, their interests and confidence increased. Moreover, scaffolds such as assembly diagrams, circuit schematics, programs, and one-on-one tutorials were slowly faded as participants progressed in their abilities. Facilitators’ instrumental assistance played a major role in participants’ performance accomplishment. In this manner, the participants were offered multiple opportunities to taste success, which contributed to their sense of performance accomplishment (Bandura, 1977). ii) Vicarious learning, also known as modeling, occurs when one learns by observing successful peers perform tasks (Bandura, 1977). Participants with varied educational backgrounds and levels of expertise worked in teams for hands-on learning. As team members experienced difficulty in construction, circuitry, or programming, they received support from others who were strong in the corresponding areas. The teams who successfully completed their robotics design tasks were encouraged to help the teams that were lagging. The facilitators encouraged the participants to observe other teams and learn from one another. Observing peers successfully completing challenging robotics tasks allowed the participants to envision being persistent and successful themselves (Bandura, 1977). Finally, facilitators’ career-related role modeling contributed to participants’ vicarious learning. iii) Verbal persuasion is related to the verbal affirmation that an individual can master or complete the assigned task (Bandura, 1977). In the workshop, a common and effective way to address this was by incorporating motivational talks. Whenever some participants felt demotivated, the facilitators delivered such talks or played compelling videos. The participants received answers to their queries by interacting with the facilitators. It was observed that participants’ motivation improved when videos of motivational talks were played. Moreover, verbal encouragement from the facilitators positively impacted participants’ engagement in the workshop activities. iv) Emotional arousal is related to the state of increased physiological activity (Bandura, 1977). It can be caused by the anxiety of not being able to complete a given task, lowering confidence and negatively affecting one’s performance. Most workshop participants were working with robots for the first time, so it was natural for them to feel nervous, especially during the initial days. Some participants were afraid of damaging components and others were concerned about not understanding concepts or not being able to complete assigned tasks. With each day, as their familiarity increased they experienced a cascade of successes and the negative influence of emotional arousal was lowered. The facilitators observed that when a participant accidentally damaged any components, they became demotivated and hesitant to continue the work. The facilitators encouraged them to follow safety measures and to learn from their mistakes. If a participant continued to experience difficulty after several days of the workshop, the facilitators engaged in one-on-one interactions with them to identify and address the source of their challenges. During their project demonstrations, the participants were able to work on anxiety related to public speaking. Over-excitement of working with novel tools and techniques can also contribute to the emotional arousal. Emotional support of the facilitators was critical in the management of participant emotions. Familiarity Each year, the familiarity survey score showed pre/post-test improvement in averages for teachers and students in all its four sections: mechanical, electrical/electronic, sensor/actuator/microcontroller, and programming. To be an effective robotics engineer, it is important to be knowledgeable in all the four sections. For the robotics design projects, in each team, group members who were good at working with the mechanical, electrical, or programming aspects, initially focused on working on the corresponding design elements. However, as each member of a team needed to know all basic aspects of their robotics design, they engaged in learning from their peers. While students had been previously exposed to the basic concepts of mechanical, electrical, and electronics disciplines in their physics courses, for most students the principles of sensing and programming were challenging and it took them time to grasp these concepts. This was true for the teachers also who had math or science backgrounds and were working with sensors, microcontrollers, and coding for the first time. Having observed this pattern and to prevent overwhelming the participants, we followed a slow but steady pace in introducing relevant topics. Specifically, analyzing the pre-test responses and identifying the low scores permitted us to adapt the PD schedule, not only in the first year but also in the subsequent years. Thus, in the post-test responses, participants scored on average above 3 out of 5 for all sections of the familiarity survey. Content knowledge The technical quiz had four sections: drive mechanism, electronics, actuation/manipulation/localization, and gears/motors. A robotics engineer must have sufficient fundamental knowledge and practical experience in all four sections. It is seen from the pre/post-tests that the teacher respondents performed the best and showed the largest improvement in the drive mechanism section. The student respondents also scored the most on the drive mechanism section in the pre/post-tests but they showed the largest improvement in the actuation/manipulation/localization section. There was a drop in pre/post-test average in the gears/motors section in the third year, both for teachers and students. Several questions in this section entailed mathematical calculations and ongoing interactions with the participants revealed that some of them made computational mistakes, despite having a conceptual understanding, causing the drop in the post-test scores. Since in the pre-test students had low scores in the electronic section, the workshop facilitators were vigilant in checking students’ circuit connections to prevent damage to circuit components. In this manner, during the experiential learning sessions, the facilitators were able to address student misunderstandings about electrical connections. The quiz results showed significant improvement in pre/post-test for most sections and an overall improvement. Robotics contest The increased self-efficacy, familiarity, and content knowledge of the teachers in the post-test illustrates that they have the potential to influence student learning in school. In fact, through the end-of-year contest, the facilitators found that the students had learned about robotics, electronics, coding, and entrepreneurship concepts. They demonstrated their learning and skills through real-world projects in the end-of-year contest. They were provided the opportunity to represent their class and winning teams won bragging rights for their school. Moreover, some students from the winning teams won paid internship awards at local start-up businesses. This created awareness about the importance and prospects of robotics education. It also provided encouragement to other students of the school to take part in robotics activities. Students who get exposed to engineering design, entrepreneurship, and internship at an early stage in life have the potential to develop a keen entrepreneurial mindset and make a lasting impact on human life. Strategies for NGSS alignment As stated earlier, NGSS highlight the need for integrating engineering design and practices within the K-12 curriculum. This workshop introduced robotics design principles to the participants. The NGSS also draw effective connections with the Common Core State Standards for Math. Many concepts introduced in the workshop involved mathematical understanding and calculations, e.g., binary to decimal conversion and vice versa, calculations involving gear ratios, sensor calibration, and motor turning angles. Use of PBL and a design canvas (Kline et al., 2017) enabled consideration of the following eight-step engineering design process (Massachusetts Department of Education, 2006). i) Identifying a problem: e.g., delivering coffee in a coffee shop, garbage disposal, etc., using robotics. ii) Researching a problem: e.g., through online search seeking current solutions, understanding their workings, and identifying their limitations. iii) Developing solutions: brainstorming with team members to identify various ways of solving the problem using the tools available at the team’s disposal. iv) Selecting solutions: selecting the best possible solution after analyzing the pros and cons of all solutions. v) Prototyping: developing a physical prototype representing the selected solution by integrating the mechanical, electrical, electronic, and computing subsystems. vi) Testing/evaluation: validating the solution prototype by running different tests under varying conditions. vii) Communicating solutions: giving demonstration, making presentation, and preparing a report. viii) Redesigning: revising and reiterating the prototype solution by analyzing the feedback from peers and facilitators. These steps of the engineering design process are consistent with the science and engineering practices (SEPs) of NGSS that entail “asking questions and defining problems”, “analyzing and interpreting data”, and “engaging in argument from evidence”, among others. Starting in Y2, the workshop participants were introduced to the design canvas of Kline et al. (2017), which assists in data collection and analysis while giving prominence to the business aspects of the problem under consideration. The design canvas (Kline et al., 2017) consists of nine elements of an organization’s building blocks, viz., stakeholders, actors, features, interactions, modes, inputs/outputs, functions, components, and designs, and the interactions between them. The design canvas helped the participants systematically advance through the various stages of product design and development. It revealed to them areas requiring greater attention and allowed them to consider feedback from diverse stakeholders, thus improving both the design process and design itself. It allowed them to explore possible solutions through literature search and brainstorming. They recognized and considered tradeoffs among the different solutions. In Y2, the design canvas was used in a treatment vs. control group setting, revealing its efficacy as a tool through which participants can learn better, have better project results, and gain valuable knowledge (Mallik et al., 2019). Design canvas also served as an assessment tool for the facilitators to gauge participant performance in projects and to refine the pedagogical and support strategies. Limitations This study has several limitations. First, the sample of teachers and students who participated in the study and responded to the surveys is small. Second, the responses to the surveys are self-reported by the participants and thus may be limited in their reliability and validity (Fan et al., 2006; West, 2014). Third, the study provides at best the examination of a project implementation in a specific setting, limiting its potential concerning representativeness or generalizability. Some of these limitations may be addressed by conducting summer workshops at scale with additional time, resources, and participants and utilizing multiple ways of assessing participants’ robotics self-efficacy, familiarity, and content knowledge with validated instruments. Finally, using gender-balanced sampling, with a suitable sample number, can improve the representativeness and generalizability of the results (Schina et al., 2021a). Implications Being one of the few studies to outline the challenges and successes of implementing a robotics curriculum with high school teachers and students, this study has implications for STEM educators and PD providers. The study employed and examined varied strategies that can enhance participants’ robotics self-efficacy, familiarity, and content knowledge and student outcomes, as seen through the end-of-year robotics contest. The results suggest that a well-designed and executed robotics education program has the potential to produce significant gains in robotics self-efficacy, familiarity, and content knowledge among participants despite their low performance on these measures when they begin pursuing robotics explorations. Systematic attention to the four components of self-efficacy, careful examination of pre-test responses, and proactive solicitation of participant feedback all contributed to the iterative refinement of workshop curriculum, structure, and activities. The results of the study illustrate that teachers can successfully embed robotics learning, under the PBL framework, at their schools. Finally, combining robotics with entrepreneurial explorations allowed the students to partake in personally meaningful experiential learning with potential for societal benefits. Future directions While this work measured gains in robotics self-efficacy, familiarity, and content knowledge through participant responses to pre/post-tests, future work may include observational methods and assessment rubrics as part of capstone projects to obtain richer and robust measures and analyses. In this work, the student and teacher participants had diverse characteristics (backgrounds, interests, and commitments). A future study may recruit participants with some common characteristics or examine assessment results by duly considering participant characteristics. Future studies can examine how robotics electives and after-school programs can be designed and implemented to explicitly incorporate the disciplinary core ideas and crosscutting concepts of NGSS. By organizing robotics education workshops in an ongoing manner, there will be numerous opportunities to analyze, learn, and improve their varied aspects and mechanics. Research can also examine how changes in the content and organization of curriculum impact participant learning. Further studies can analyze the quality of classroom implementation of teachers’ learning and student performance. Conclusion There is a growing interest in developing a workforce that is well prepared in STEM disciplines. Equipping students to be successful in the technology-rich modern workplace calls for educators who have pedagogical skills, content knowledge, and technical capacity to integrate new technologies in their curriculum and instruction. Thus, teachers are increasingly seeking to adopt robotics as a technological tool for teaching STEM subjects. The workshop discussed in this paper provided PD to teachers within a PBL framework with a focus on robotics design and entrepreneurship. It sought to impart to teachers and students the ability to effectively work with robotics technology and entrepreneurship in a classroom learning context. By enhancing the robotics self-efficacy, familiarity, and content knowledge, the workshop demonstrated its potential to support teachers overcome barriers that make it difficult to incorporate new technology-rich content in the classroom. The workshop of this paper integrated strategies to serve as gateways for students’ entrepreneurship explorations. Various assessment tools were used to measure participants’ learning gains and the assessment results can be used to characterize the efficacy of the PD itself. We also examined the outcomes of participants’ AY activities through the end-of-year contests. The descriptions of workshop content, hands-on activities, assessments, and AY follow-up may be useful to the PD community that may seek to upgrade and conduct similar workshops. The robotics content learned during the workshop was tailored and adapted by the teachers for their classroom use and the teachers shared various classroom resources with one another. Each school was visited by one facilitator for observing the class and assisting the teacher in lesson delivery. The students learned a significant amount of content about robotics while gaining practical and theoretical engineering knowledge. This observation was validated through the end-of-year contest where student teams presented high quality, competitive projects. A: Rubics: For each aspect, place a check mark in the column most appropriate Team Name: __________________________ Categories Multipliers X1=1 X2=2 X3=3 X4=4 Did no meet requirement Needs improvement Satisfactory Excelled Robotics engineering Design (aesthetics, robustness) Programming sophistication (logic behind the code, complexity, and approach taken) Sensors and human interface (use of sensors and LCD, type, applications) Actuation (type of motors, mechanisms used to move parts around) Functionality: Does the robot work? Total Innovation/Creativity Solution: Was the solution based on a household problem? Creative problem-solving approaches taken Total Entrepreneurship Marketing (based on survey questions, social media, research, advertising, robot name) Scalability (feasibility in real world) Feasibility of robot price Total Video/Presentation Presentation (pitch, video, clear, articulation of product) Were the result of the survey mentioned? Teamwork (were all team members involved in some aspect? Total Written Assignments One page typed (12pt font) document answering the questions: (1) What knowledge in robotics did you have before starting this course in robotics? (2) What did you learn by making your robot and participating in this competition? Survey: Type your survey questions and answers from your interviews Total (count and report checkmarks in each column) Overall Total Acknowledgements This work is supported in part by the National Science Foundation grant ITEST DRL: 1614085. The authors thank the high school teachers and their students for their participation in this study. The first author additionally thanks his colleague Dr. Hye Sun You for her time and effort in reviewing the early drafts of the manuscript and providing valuable comments. Data availability The datasets generated and analyzed during this study are available from the corresponding author on reasonable request. Declarations Ethical standard All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee (Inst. IRB-FY2016-492) and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent Informed consent was obtained from all individual participants included in the study. Conflicts of interest The authors declare that they have no conflict of interest. This paper is dedicated to the memory of Dr. Abhidipta Mallik, a great mentor to high school teacher and student participants of this study, who passed away untimely. ==== Refs References Akins, L., Burghardt, D. (2006). Work in progress: Improving K-12 mathematics understanding with engineering design projects. In: Proceedings IEEE Annual Conference on Frontiers in Education (pp. 13–14). Ansoff HI Strategies for diversification Harvard Business Review 1957 35 5 113 124 Bandura A Self-efficacy: Toward a unifying theory of behavioral change Psychological Review 1977 84 2 191 215 10.1037/0033-295X.84.2.191 847061 Bandura A Self-efficacy mechanism in human agency American Psychologist 1982 37 2 122 147 10.1037/0003-066X.37.2.122 Barker BS Ansorge J Robotics as means to increase achievement scores in an informal learning environment Journal of Research on Technology in Education 2007 39 3 229 243 10.1080/15391523.2007.10782481 Barker BS Nugent G Grandgenett N Adamchuk VI Robots in K-12 Education: A New Technology for Learning 2012 Hershey, PA IGI Global Barron BJS Schwartz DL Vye NJ Moore A Petrosino A Zech L Bransford JD Doing with understanding: Lessons from research on problem-and project-based learning Journal of the Learning Sciences 1998 7 3–4 271 311 Beary, V. E. (2013). The NFTE difference: Examining the impact of entrepreneurship education. https://www.foroige.ie/sites/default/files/nfte_difference_final_report_2013.pdf. Accessed 6 Dec 2020. Benitti FBV Exploring the educational potential of robotics in schools: A systematic review Computers and Education 2012 58 3 978 988 10.1016/j.compedu.2011.10.006 Blumenfeld PC Soloway E Marx RW Krajcik JS Guzdial M Palincsar A Motivating project-based learning: Sustaining the doing, supporting the learning Educational Psychologist 1991 26 3–4 369 398 10.1080/00461520.1991.9653139 Brophy S Klein S Portsmore M Rogers C Advancing engineering education in P-12 classrooms Journal of Engineering Education 2008 97 3 369 387 10.1002/j.2168-9830.2008.tb00985.x Buabeng-Andoh C Factors influencing teachers’ adoption and integration of information and communication technology into teaching: A review of the titerature International Journal of Education and Development using ICT 2012 8 1 136 155 Carberry AR Lee H Ohland MW Measuring engineering design self-efficacy Journal of Engineering Education 2010 99 1 71 79 10.1002/j.2168-9830.2010.tb01043.x Cejka, E., Rogers, C. (2005). Inservice teachers and the engineering design process. In: ASEE Annual Conference and Exposition. https://peer.asee.org/14552. Accessed 6 Dec 2020. Center for Youth and Communities (2010). Evaluation of the FIRST LEGO League. Brandeis University. https://www.usfirst.org/aboutus/content.aspx?id=46. Accessed 6 Dec 2020. Chambers JM Carbonaro M Rex M Grove S Scaffolding knowledge construction through robotic technology: A middle school case study Electronic Journal for the Integration of Technology in Education 2007 6 55 70 Chaney, B. (1995). Student outcomes and the professional preparation of eighth-grade teachers in science and mathematics. NSF/NELS:88 Teacher transcript analysis. ERIC. https://files.eric.ed.gov/fulltext/ED389530.pdf. Accessed 6 Dec 2020. Chemers MM Zurbriggen EL Syed M Goza BK Bearman S The role of efficacy and identity in science career commitment among underrepresented minority students Journal of Social Issues 2011 67 3 469 491 10.1111/j.1540-4560.2011.01710.x Chin K-Y Hong Z-W Chen Y-L Impact of using an educational robot-based learning system on students’ motivation in elementary education IEEE Transactions on Learning Technologies 2014 7 4 333 345 10.1109/TLT.2014.2346756 Cioffi, N., Kulturel-Konak, S., Konak, A. (2014). “Anything is possible”–Teaching entrepreneurship in an interactive K-12 workshop. In: IEEE Integrated STEM Education Conference. https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=6891013. Accessed 6 Dec 2020. Cohen J Statistical Power Analysis for the Behavioral Sciences 1988 USA Lawrence Erlbaum Associates Dana TM Zembal-Saul C Munford D Tsur C Friedrichsen PM Learning to teach with technology model: Implementation in secondary science teacher education Journal of Computers in Mathematics and Science Teaching 2001 20 4 377 394 Dym CL Agogino AM Eris O Frey DD Leifer LJ Engineering design thinking, teaching, and learning Journal of Engineering Education 2005 94 1 103 120 10.1002/j.2168-9830.2005.tb00832.x Elliott AC Woodward WA Statistical Analysis Quick Reference Guidebook: With SPSS Examples 2007 Thousand Oaks, CA SAGE Publications Endo, G., Yamada, H., Aoki, T., Hirose, S. (2013). Development of biologically inspired educational robots based on gliding locomotion. In: IEEE/RSJ International Conference on Intelligent Robots and Systems (pp. 3291–3296). Erwin B Cyr M Rogers C LEGO engineer and robolab: Teaching engineering with LabView from kindergarten to graduate school International Journal of Engineering Education 2000 16 3 181 192 Faisal, A., Kapila, V., Iskander, M. G. (2012). Using robotics to promote learning in elementary grades. In: ASEE Annual Conference and Exposition. https://peer.asee.org/22196. Accessed 6 Dec 2020. Fan X Miller BC Park K-E Winward BW Christensen M Grotevant HD Tai RH An exploratory study about inaccuracy and invalidity in adolescent self-report surveys Field Methods 2006 18 3 223 244 10.1177/152822X06289161 Guskey TR Yoon KS What works in professional development? Phi Delta Kappan 2009 90 7 495 500 10.1177/003172170909000709 Hill HC Rowan B Ball DL Effects of teachers’ mathematical knowledge for teaching on student achievement American Educational Research Journal 2005 42 2 371 406 10.3102/00028312042002371 Hmelo-Silver CE Problem-based learning: What and how do students learn? Educational Psychology Review 2004 16 3 235 266 10.1023/B:EDPR.0000034022.16470.f3 Hsu, M., Cardella, M., Purzer, S., Diaz, N. M. (2010). Elementary teachers’ perceptions of engineering and familiarity with design, engineering and technology: Perspectives from a national population. In: ASEE Annual Conference and Exposition. https://peer.asee.org/16286. Accessed 6 Dec 2020. Huber LR Sloof R Van Praag M The effect of early entrepreneurship education: Evidence from a randomized field experiment European Economic Review 2014 72 76 97 10.1016/j.euroecorev.2014.09.002 Inda M Rodríguez C Peña JV Gender differences in applying social cognitive career theory in engineering students Journal of Vocational Behavior 2013 83 3 346 355 10.1016/j.jvb.2013.06.010 Innovation First Inc. (2007). Vex robotics. https://www.vexrobotics.com. Accessed 6 Dec 2020. Jamrisko, M., Lu, W. (2018). The U.S. drops out of the top 10 in innovation ranking. Bloomberg. https://www.bloomberg.com/news/articles/2018-01-22/south-korea-tops-global-innovation-ranking-again-as-u-s-falls. Accessed 13 June 2021. Javaid M Haleem A Singh RP Suman R Substantial capabilities of robotics in enhancing Industry 4.0 implementation Cognitive Robotics 2021 1 58 75 10.1016/j.cogr.2021.06.001 Kelley TR Wicklein RC Teacher challenges to implement engineering design in secondary technology education Journal of Industrial Teacher Education 2009 46 3 34 50 Kline, W., Schindel, W., Tranquillo, J., Bernal, A., Hixson, C. (2017). Development of a design canvas with application to first-year and capstone design courses. In: ASEE Annual Conference and Exposition. https://peer.asee.org/28159. Accessed 6 Dec 2020. Kokotsaki D Menzies V Wiggins A Project-based learning: A review of the literature Improving Schools 2016 19 3 267 277 10.1177/1365480216659733 Krueger N The impact of prior entrepreneurial exposure on perceptions of new venture feasibility and desirability Entrepreneurship Theory and Practice 1993 18 1 5 21 10.1177/104225879301800101 Larmer J Mergendoller JR Seven essentials for project-based learning Educational Leadership 2010 68 1 34 37 Lent, R. W., Brown, S. D., Hackett, G. (2002). Social cognitive career theory. In: D. Brown (ed.), Career Choice and Development (pp. 255–311). San Francisco, CA: Jossey-Bass, 4th edition. Lent RW Lopez FG Sheu H-B Lopez AM Jr Social cognitive predictors of the interests and choices of computing majors: Applicability to underrepresented students Journal of Vocational Behavior 2011 78 2 184 192 10.1016/j.jvb.2010.10.006 Liero H Zwanzig S Introduction to the Theory and Statistical Inference 2012 Boca Raton, FL CRC Press Lindsay, A. (2010). Robotics with the Boe-Bot, Ver. 3.0. Parallax, Inc. https://www.parallax.com/package/robotics-with-the-boe-bot-kit-downloads/. Accessed 6 Dec 2020. Loucks-Horsley S Stiles KE Mundry S Love N Hewson PW Designing Professional Development for Teachers of Science and Mathematics 2009 Thousand Oaks, CA Corwin press Mallik, A., Rajguru, S. B., Kapila, V. (2018). Fundamental: Analyzing the effects of a robotics training workshop on the self-efficacy of high school teachers. In: ASEE Annual Conference and Exposition. https://peer.asee.org/30550. Accessed 6 Dec 2020. Mallik, A., Rajguru, S. B., Kapila, V. (2019). Use of a design canvas in a robotics workshop and analysis of its efficacy (Fundamental). In: ASEE Annual Conference and Exposition. https://peer.asee.org/33489. Accessed 6 Dec 2020. Margolis M Arduino Cookbook: Recipes to Begin, Expand, and Enhance Your Projects 2011 Sebastopol, CA O’Reilly Media Inc Martin, A. (2012). State of entrepreneurship in robotics. New Venturist. http://newventurist.com/2012/06/state-of-entrepreneurship-in-robotics/. Accessed 6 Dec 2020. Martin FG Robotic Explorations: A Hands-on Introduction to Engineering 2000 Upper Saddle River, NJ Prentice Hall Massachusetts Department of Education (2006). Massachusetts Science and Technology/Engineering Curriculum Framework. Massachusetts Department of Education, Malden, MA. https://www.doe.mass.edu/frameworks/scitech/1006.doc. Accessed 6 Dec 2020. Mataric, M. J., Koenig, N. P., and Feil-Seifer, D. (2007). Materials for enabling hands-on robotics and STEM education. In: AAAI Spring Symposium: Semantic Scientific Knowledge Integration, (pp. 99–102). Muldoon, J., Phamduy, P. T., Le Grand, R., Kapila, V., and Iskander, M. G. (2013). Connecting cognitive domains of Bloom’s taxonomy and robotics to promote learning in K-12 environment. In: ASEE Annual Conference and Exposition. https://peer.asee.org/19343. Accessed 6 Dec 2020. Mullin, S. (2015). The science of familiarity: How to increase conversions by being completely unoriginal. CXL. https://cxl.com/blog/science-of-familiarity/. Accessed 2 May 2022. Nair, P., Huang, J., Jackson, J., and Cox-Petersen, A. (2017). Combining STEM and business entrepreneurship for sustaining STEM-readiness. In: IEEE Integrated STEM Education Conference (ISEC), (pp. 76–78). Nakkula, M., Pineda, C., Dray, A., Lutyens, M. (2003). Expanded explorations into the psychology of entrepreneurship: Findings from the 2001–2002 Study of NFTE in two Boston Public High-schools. Working Paper, Harvard University Graduate School of Education. National Research Council Educating Teachers of Science, Mathematics, and Technology: New Practices for the New Millennium 2001 Washington, DC National Academies Press NGSS Lead States Next Generation Science Standards: For States, by States 2013 Washington, DC National Academies Press Norton SJ McRobbie CJ Ginns IS Problem solving in a middle school robotics design classroom Research in Science Education 2007 37 3 261 277 10.1007/s11165-006-9025-6 Office of Educational Technology (2017). Reimagining the Role of Technology in Education: 2017 National Education Technology Plan Update. U.S. Department of Education. https://tech.ed.gov/files/2017/01/NETP17.pdf. Accessed 6 Dec 2020. Orsak GC Munson DC Weil A Conner M Rummel D High-tech engineering for high school: It’s time! IEEE Signal Processing Magazine 2004 21 1 103 108 10.1109/MSP.2004.1267053 Osterwalder A Pigneur Y Business Model Generation: A Handbook for Visionaries, Game Changers, and Challengers 2010 Hoboken, NJ Wiley Palmer D Dixon J Archer J Changes in science teaching self-efficacy among primary teacher education students Australian Journal of Teacher Education 2015 40 12 27 40 10.14221/ajte.2015v40n12.3 Perdue DJ The Unofficial LEGO Mindstorms NXT Inventor’s Guide 2007 San Francisco, CA No Starch Press Petre M Price B Using robotics to motivate ‘back door’ learning Education and Information Technologies 2004 9 2 147 158 10.1023/B:EAIT.0000027927.78380.60 Porter, M. E., et al. (1989). How competitive forces shape strategy. In D. Asch et al. (Ed.), Readings in Strategic Management (pp. 133–143). Macmillan Publishers Limited. Pruett M Entrepreneurship education: Workshops and entrepreneurial intentions Journal of Education for Business 2012 87 2 94 101 10.1080/08832323.2011.573594 Richards, L. G., Laufer, G., Humphrey, J. A. C. (2002). Teaching engineering in the middle schools: Virginia middle schools engineering education initiative. In: ASEE/IEEE Annual Frontiers in Education Conference (Vol. 1, pp. T1C–6–T1C–11). Riojas M Lysecky S Rozenblit J Educational technologies for precollege engineering education IEEE Transactions on Learning Technologies 2011 5 1 20 37 10.1109/TLT.2011.16 Sabouri, P., Ghosh, S., Mallik, A., Kapila, V. (2020). The formation and dynamics of teacher roles in a teacher-student groupwork during a robotic project (Fundamental). In: ASEE Annual Conference and Exposition. https://peer.asee.org/35323. Accessed 6 Dec 2020. Sadler PM Coyle HP Schwartz M Engineering competitions in the middle school classroom: Key elements in developing effective design challenges The Journal of the Learning Sciences 2000 9 3 299 327 10.1207/S15327809JLS0903_3 Schilling MA Esmundo M Technology s-curves in renewable energy alternatives: Analysis and implications for industry and government Energy Policy 2009 37 5 1767 1781 10.1016/j.enpol.2009.01.004 Schina D Esteve-González V Usart M An overview of teacher training programs in educational robotics: Characteristics, best practices and recommendations Education and Information Technologies 2021 26 3 2831 2852 10.1007/s10639-020-10377-z Schina D Valls-Bautista C Borrull-Riera A Usart M Esteve-González V An associational study: Preschool teachers’ acceptance and self-efficacy towards educational robotics in a pre-service teacher training program International Journal of Educational Technology in Higher Education 2021 18 1 1 20 10.1186/s41239-021-00264-z Shulman LS Those who understand: Knowledge growth in teaching Educational Researcher 1986 15 2 4 14 10.3102/0013189X015002004 Siegwart R Nourbakhsh IR Scaramuzza D Introduction to Autonomous Mobile Robots 2011 Cambridge, MA MIT press Stolkin, R., Hotaling, L., Sheryll, R., Sheppard, K., Chassapis, C., McGrath, E. (2007). A paradigm for vertically integrated curriculum innovation–How curricula were developed for undergraduate, middle and high school students using underwater robotics. In: International Conference of Engineering Education, (pp. 5–10). Subramaniam P Motivational effects of interest on student engagement and learning in physical education: A review International Journal of Physical Education 2009 46 2 11 19 Syofyan, R., Siwi, M. K. (2018). The impact of visual, auditory, and kinesthetic learning styles on economics education teaching. In: International Conference on Economics Education, Economics, Business and Management, Accounting, and Enterpreneurship (PICEEBA) (Vol. 57, pp. 642–649). Valdez G McNabb M Foertsch M Anderson M Hawkes M Raack L Computer-based Technology and Learning: Evolving uses and Expectations 1999 Oak Brook, IL North Central Regional Educational Laboratory VandenBos, G. R. (2007). APA Dictionary of Psychology. American Psychological Association (p 410). Varier D Dumke EK Abrams LM Conklin SB Barnes JS Hoover NR Potential of one-to-one technologies in the classroom: Teachers and students weigh in Educational Technology Research and Development 2017 65 4 967 992 10.1007/s11423-017-9509-2 Verner IM Ahlgren DJ Miller DP Robotics olympiads: A new means to integrate theory and practice in robotics Computers in Education Journal 2007 17 4 11 21 Vollmer, U., Jeschke, S., Burr, B., Knipping, L., Scheurich, J., Wilke, M. (2011). Teachers need robotics-training, too. In: Automation, Communication and Cybernetics in Science and Engineering 2009/2010 (pp. 359–364). Springer. Wang X Why students choose STEM majors: Motivation, high school learning, and postsecondary context of support American Educational Research Journal 2013 50 5 1081 1121 10.3102/0002831213488622 West, M. R. (2014). The limitations of self-report measures of non-cognitive skills. Brookings. https://www.brookings.edu/research/the-limitations-of-self-report-measures-of-non-cognitive-skills/. Accessed 2 May 2022 World Economic Forum (2018). The Future of Jobs Report. Center for the New Economy and Society. https://www.weforum.org/reports/the-future-of-jobs-report-2018. Accessed 6 Dec 2020. You HS Chacko SM Kapila V Examining the effectiveness of a professional development program: Integration of educational robotics into science and mathematics curricula Journal of Science Education and Technology 2021 30 4 567 581 10.1007/s10956-021-09903-6	36465423	PMC9709379	NO-CC CODE	2022-12-01 23:23:04	no	Educ Inf Technol (Dordr). 2022 Nov 30;:1-40	utf-8	Educ Inf Technol (Dordr)	2,022	10.1007/s10639-022-11400-1	oa_other
==== Front J Acad Ethics J Acad Ethics Journal of Academic Ethics 1570-1727 1572-8544 Springer Netherlands Dordrecht 9463 10.1007/s10805-022-09463-3 Article Undergraduate Ethics Education in Paramedicine in Australia http://orcid.org/0000-0003-1866-7084 Shearer Kirsty [email protected] 1 http://orcid.org/0000-0002-5553-5825 Thomas Matthew 2 http://orcid.org/0000-0001-5677-9496 Signal Tania 1 http://orcid.org/0000-0003-3621-7976 Townsend Ruth 3 Stepanov Nikola 4 1 grid.1023.0 0000 0001 2193 0854 CQUniversity Australia, Rockhampton, QLD Australia 2 grid.1023.0 0000 0001 2193 0854 CQUniversity Australia, Adelaide, South Australia Australia 3 Paramedicine Council of NSW, Sydney, NSW Australia 4 grid.1011.1 0000 0004 0474 1797 Division of Tropical Health and Medicine, James Cook University, Douglas, QLD Australia 30 11 2022 116 7 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. In Australia, paramedics are obliged to practice ethically. Graduates of baccalaureate degrees in paramedicine should therefore possess a common grounding in ethics to meet the professional capabilities expected of registered paramedics. However, there is a lack of clarity regarding ethics education for paramedicine students, including what is taught, how it is taught, and how it is assessed. This paper explores ethics education for paramedicine students in Australia, how it aligns with current professional expectations, and how it may be enhanced. Point-in-time data regarding ethics education was collected from websites of fifteen Australian universities offering undergraduate baccalaureate degrees in paramedicine. Data collection was supported by consultation with academics from several institutions. Content analysis was utilised to categorise and analyse data to explore similarities and differences in curricula. Similarities included approaches to learning and teaching and the use of case-based learning, with variability found across teaching staff profiles and content areas. Findings suggest it is time for collaboration to develop a model ethics curriculum for paramedicine students in Australia. Keywords Ethics education Paramedic Paramedic ethics Paramedicine curriculum ==== Body pmcIntroduction Paramedics in Australia play a crucial role in delivering advanced pre- and out-of-hospital care to the community, from acute critical care to palliative support at the end of life. Diversity is not only experienced through their clinical caseload, but also across cultural contexts and in geographically disparate settings, from metropolitan locations through to regional and remote sites (Hartley, 2012; O'Meara & Duthie, 2018). The cases attended by paramedics can present a range of ethical dilemmas, including but not limited to refusal of service, the protection of vulnerable persons, resuscitation decisions, and challenges involving advance care directives (Adams et al., 1992; Heilicser et al., 1996; Moore, 2020; Moritz et al., 2020; Nordby & Nøhr, 2012; Shearer et al., 2021). In the past, paramedics have articulated the need for improved ethics education to assist them in dealing with the ethical problems faced (Heilicser et al., 1996, p. 242). There have been few studies addressing this need; however, recent research has revealed that paramedics rely more on experience than education to resolve ethical dilemmas (Shearer et al., 2021, p. 336). This is problematic as students of paramedicine have limited opportunities to gain clinical experience underscoring the need to rely on education to provide the theoretical knowledge and skills underpinning good ethical decision making for paramedics. In December 2018, paramedicine became a regulated profession in Australia under the National Registration and Accreditation Scheme, established by Australian governments (Commonwealth of Australia, 2022) and regulated by the Paramedicine Board of Australia. The Board is supported by the administrative arm of the scheme, the Australian Health Practitioner Regulation Agency (‘AHPRA’) (Australian Health Practitioner Regulation Agency, 2018; Paramedicine Board of Australia, 2021a). Among other roles, the Paramedicine Board functions to develop standards, codes, and guidelines for the profession, approve accreditation standards, and subsequently accredit approved programs of study (Paramedicine Board of Australia, 2021a). Three documents published by the Paramedicine Board of Australia guide the expectations of paramedics. Two of these documents, the ‘Code of Conduct’ (2018) and ‘Professional capabilities for registered paramedics’ (2021), focus on the expectations of the individual paramedic; while the ‘Accreditation Standards: Paramedicine’ (2020) outline the expectations on education providers to develop ‘work-ready’ graduates capable of meeting professional requirements (Paramedicine Board of Australia, 2018, 2020, 2021b). The accreditation standards require courses to have learning outcomes that address all the professional capabilities for paramedics (Paramedicine Board of Australia, 2020, p.16), including Domain 1 of the professional capabilities document: ‘The professional and ethical practitioner’ (Paramedicine Board of Australia, 2021b, p. 3–4). Australian universities offering degrees in paramedicine are currently aligning their courses with the new accreditation standards. This realignment and impending accreditation of undergraduate courses under the new scheme presents a timely opportunity to explore ethics education for paramedic students in Australia, to identify how it could be enhanced to ensure graduates are adequately prepared to practice professionally and in accordance with the requisite standards. This study, focusing on undergraduate baccalaureate students of paramedicine in Australia, seeks to document the design and delivery of ethics education and consider whether it meets the current and future needs of students seeking to develop practical ethical aptitude. As the teaching of ethics in paramedicine remains a relatively young field, the medical literature will be drawn on to inform an understanding of the findings, as independent practice is a common feature of both disciplines, and the challenges of practice are shared. Teaching ethics within medical education is well-developed and provides insights that may assist in elevating the teaching of ethics to students of paramedicine, ensuring graduates are prepared for professional registration and practice. Of note when discussing ethics education is the often observed overlap between ethics and law. As Olick (2001) comments, it is not infrequent that clinicians are required to consider both the ethical and legal dimensions of challenges presented in patient care. In both medicine (and paramedicine as evidenced below), some institutions teach ethics and law separately, whereas others prefer teaching them together. It is beyond the scope of this paper to explore this aspect in any detail, though it may be an important topic for future research. Materials and Methods During the point-in-time data collection period in January/February 2021, a search for university courses offering paramedicine was conducted using the peak paramedic industry group, the Council of Ambulance Authorities (2022) website, and cross-referenced with accredited programs of study information from the Australian Health Practitioner Regulation Agency (AHPRA) Paramedicine Board of Australia website. Fifteen universities offering pre-registration undergraduate baccalaureate degrees in paramedicine were identified and included in this study. This incorporated courses that were dual degrees (for example, Bachelor of Nursing/Bachelor of Paramedicine), conversion courses (for example, Bachelor of Paramedic Practice [Conversion Pathway]), and Honours courses. Pre-registration refers to a comprehensive undergraduate degree with a full complement of units leading to registration as a paramedic. Postgraduate course unit offerings were excluded, as the intent was to focus on foundational ethics education and assess like-for-like courses. Publicly available course information was obtained from relevant university websites, and the course structure was purposefully searched for subjects/units pertaining to ethics. Where no dedicated ethics subject/unit was easily identified, other subject/units within the course were searched for relevant ethics content. All subjects/units that contained content related to ethics were then subjected to analysis using a data extraction framework. Core details were recorded according to the data extraction framework, including subject/unit code and title; study mode; the year of study in which the subject/unit was offered (for example, 1st, 2nd or third year); the unit overview or aims; learning outcomes; syllabus; learning and teaching strategies; assessment strategy and rationale and any resource information, such as utilised textbooks (where available). A qualitative study design employing document analysis using a systematic process for sourcing, selecting, evaluating, and synthesising data from printed and electronic documents was used to analyse data from the fifteen university websites (Bowen, 2009). Data were transferred from each website verbatim into an Excel spreadsheet. Due to the diversity of data format and content available across university websites, the design framework was expanded to ensure data were captured relevant to each category. As a result, there are some gaps in the data where the information was not available on a particular university website. Some of these gaps were able to be filled via follow-up Zoom discussions with key academic staff involved in course delivery. Each university was contacted with the offer to add detail to the data collected from the websites. Initial contact was via email, with non-responses followed up twice more to ensure maximum opportunity to participate in the discussions. Nine of the fifteen universities engaged in this opportunity. This information was entered into a separate section of the Excel spreadsheet to ensure clarity of the data source. Transcript documents were created from the discussions, and qualitative content analysis, a ‘…systematic and objective means to make valid inferences from verbal, visual, or written data’, was used to analyse data from the transcripts (Downe-Wamboldt, 1992, p. 314; Elo & Kyngäs, 2008). Data extracted from websites was then triangulated with transcripts from the discussions with academic staff to enable corroboration and validation (Bowen, 2009). Evaluated data were then integrated, categorised, and thematically analysed by the first author to formulate results. Please note that the first author was appointed as an Accreditation Assessor with the Paramedicine Board of Australia during the data collection phase of this study. This appointment was declared at consultations with academics after the commencement date in January 2021. The first author did not participate in any accreditation assessments during the data collection phase of this study. Results Teaching Staff The data were not available for academics allocated teaching of the units at four of the fifteen universities. For the remaining eleven universities, there was a mix of academic profiles. Some ethics units were taught purely by paramedics, while others involved a combination of paramedics and lawyers (whether the same individual or two (or more) separate academics). Other academic profiles included a lawyer, a registered nurse/lawyer, an epidemiologist, a nurse/philosopher, or a legal practitioner/health professional (clinical counsellor). Stand-alone Or Embedded Ethics Content? Nine of the fifteen universities offered discrete units listing ethics and law in the unit title. Three universities offered ethics content in ‘foundations of paramedicine’ units, while the remaining three offered ethics content in combination with professionalism, policy or society and culture. Teaching Format There was no consistency as to when the ethics content was taught across university courses, with scheduling occurring across the first, second, and third year of courses. Delivery also varied, however, some of the variations occurred due to the impact of COVID-19 and the need for some universities to move to an online teaching mode where they usually would offer units face-to-face. Of the fifteen universities, seven offered their unit in a mixed-mode, which encompasses ‘blended’ learning, or a combination of face-to-face and online learning. Five universities offered their ethics units face-to-face only, though one of these universities transitioned to online learning due to COVID-19. Three universities offered their units online only. Textbooks Eleven of the fifteen universities offered information on prescribed or recommended textbooks via their respective websites or through discussion with academics. No specific data were available from four universities, although an academic from one of these remaining four universities indicated that they use references from various texts, rather than one specifically, with a preference for journal articles to blend the research and best contemporary evidence. Supplementary readings were also used by other universities to support learning. There are nine specific law and/or ethics textbooks used by universities, listed in Table 1. Of those nine, five covered ethics and law, with four of these five being paramedic specific and at least three of these engaging the expertise of a paramedic ethicist. Of the remaining four texts, two were ethics specific, and two were law specific, though neither of these four were paramedic specific.Table 1 Textbooks Unit Overviews Unit overviews expanded upon and reflected the subject titles (see Table 2). Again, there was considerable variance between the detail provided within the unit overviews, with some offering a brief synopsis of the unit content while others were more expansive.Table 2 Unit overviews Decision-making Foundations of clinical decision-making, including tools & frameworks Intuitive/analytical/safe/reasoned Paramedic knowledge/skills/attributes Introduction to paramedicine (role, administration, policy, context) Introductory paramedic skills Difference between wellness & illness Approach to patient-centred care Therapeutic, social & reasoning skills Emotional intelligence, physical & mental preparedness, resilience, self-care Context Healthcare system Regulation, registration, national law Philosophy Introduction to philosophical principles & reflection A philosophical exploration of practical moral problems Ethics Exploration of key issues in metaethics Theories & concepts in bioethics Principles, guidelines & how ethical decisions are made Ethical issues in the prehospital setting Law Statutory & common law components of paramedic practice Legal principles, legislation, actions, process & cases Introduction to the Australian legal system Legal aspects/responsibilities/obligations/context for paramedic practice Ethics & law Ethical (bioethical) & legal theoretical principles/values & how they inform practice Ethical & legal awareness, issues, choices & challenges Ethical & legal implications of professional conduct Conformity to legal doctrines & ethical standards Client autonomy & self-determination, client rights Professionalism Professional practice, professional responsibility & interpersonal communication Theoretical & practical aspects of health communication Medico-legal concepts of professionalism Codes of conduct, competency standards & professional registration Professionalism issues & challenges Culture & specific groups Cultural awareness, sensitivity, & safety for individuals & communities Caring for vulnerable populations (the aged, children, disabled, Indigenous, those who lack capacity) Unit overviews identified broad areas covered within units related to ethics, including decision-making (clinical, intuitive, analytical, safe & reasoned); a wide range of paramedic knowledge skills and attributes, for example, approach to patient-centred care, the role of the paramedic, introductory skills, and physical and mental preparedness; and references to context, such as the healthcare system, regulatory structures, registration, and policy. Learning and teaching approaches, both online and face-to-face, were articulated in some unit overviews, with the use of case studies, critical reflection and discussion featuring heavily. There was a strong emphasis on ethics and law across unit overviews, focusing on theories, principles, choices, and challenges. Legal principles, legislation and the legal system were particularly showcased. Topics concerning professionalism were included in some unit overviews, covering areas such as professional practice, communication, conduct, standards, and professional responsibility. Other areas less frequently covered include philosophy, cultural awareness and safety, and caring for vulnerable populations. Learning Outcomes Available data were collated regarding learning outcomes for each identified unit. Learning outcomes were available for all but one of the universities. The data were tabulated with learning outcomes transcribed verbatim, then grouped according to the subject area (refer to Fig. 1).Fig. 1 Learning outcomes Learning outcomes relating to law only were identified as such. Where there was generalised mention of law and ethics, learning outcomes were placed within that grouping. Any specific mention of legal or ethical principles or ethical decision-making were identified and grouped separately. Paramedic specific areas included a broad range of paramedic-specific outcomes, including but not limited to emotional/social skills and paramedic wellbeing, patient assessment and history taking, basic procedures and skills in paramedicine, manual handling, and communication, including medical terminology. Content Areas For two universities, no unit content or syllabus data were available from their respective websites or via Zoom consults. Only limited data were available from respective university websites at a further four universities, again with no additional context available via Zoom consults. The remaining nine university data sets regarding content offer greater depth, with information from their websites supported by information elicited from Zoom consults with academics (refer to Fig. 2 and Table 3).Fig. 2 Content areas. OHS Occupational health and safety, EEA Emergency Examination Authority, AHPRA Australian Health Practitioner Regulation Agency Table 3 Content per university units Yellow shading syllabus available from website, grey shading data supported by Zoom consult with academic, blank column no data available There was a broad range of content areas across the data set. Common content included professionalism and related topics and ethics, including applied ethics. Legal and ethical issues related to vulnerable populations, such as Indigenous health, child protection, mandatory reporting, elder abuse, guardianship, sexual assault, developmental disability, women, adolescents and children, were also common. Other popular content included privacy and confidentiality; and consent and refusal. Less commonly covered content included medico-legal issues of restraint, ethics of expert testimony, and biomedical research. Approaches to Learning and Teaching Data regarding approaches to learning and teaching was available from twelve of the fifteen universities. Information was extracted from unit overviews and learning and teaching strategies from university websites, supported by data from Zoom consults with academics where available. Analysed results were then tabulated (see Figs. 3 and 4).Fig. 3 Pedagogical tools Fig. 4 Pedagogical approaches Pedagogical tools were examined and identified. The use of workbooks or modules featured, with several universities utilising these tools as pre-work for either face-to-face or online tutorials. There were consistent themes within pedagogical approaches, including encouraging engaged discussion, group work and collaborative learning, and a strong focus on the practical application of knowledge. Consistent with these approaches was an emphasis on case-based education across universities, with many utilising case scenarios as an applied approach to considering principles, concepts, theories, and legislation. Less common strategies included the analysis of arguments and challenges, problem-solving and communication strategies. Assessment The types of assessments utilised across paramedic science ethics curricula are shown in Fig. 5. No data were available for two of the universities. Of the available information, it is evident that a broad range of assessments are conducted at various points throughout the curricula. The most common assessment type was written assessment, followed by online tests or quizzes. Examination was the third most utilised form of assessment. A mix of question types was used in the tests/quizzes and examinations, including multiple-choice and short answer questions. Other methods of assessment included case studies, posters, and group presentations. During a Zoom consult, one university academic articulated that they use a continuous assessment process in the form of written assessments at the end of each module rather than one end of unit examination.Fig. 5 Assessment types used Discussion The introduction of professional registration for paramedics in Australia in late 2018, and the publication more recently of documents outlining the obligation for paramedics to practice ethically, suggests that it is timely to critically reflect on the formal ethics education provided in undergraduate paramedic education in Australia. Teaching Staff As an initial point of discussion, it is important to explore who should be teaching ethics in paramedicine. This study found heterogeneity among academics engaged in teaching ethics-related units to paramedicine students at Australian universities with respect to their backgrounds and fields of expertise. While further research may elicit academics’ formal qualifications, there is no current consensus about those qualifications. Braunack‐Mayer et al. (2001) identified the lack of standardised qualifications for teaching ethics in medical schools as a challenge in medicine. Stipulating specific qualifications can prove difficult however the literature indicates academics with an appropriate ethics background, such as a higher degree with an emphasis on ethics, was highly desirable (Braunack‐Mayer et al., 2001). At the very least, academics engaged in teaching ethics should possess a good understanding of moral philosophy and familiarity with and confidence in the clinical environment (Braunack‐Mayer et al., 2001; Giubilini et al., 2016), emanating from professional experience. A range of teaching profiles were endorsed by the academics consulted in this study. As an example, one academic voiced that the combination of an expert in law and ethics and a paramedic for contextualisation was the perfect combination of teaching team for such subjects. This academic noted that while paramedics have a good working knowledge, we are not experts in the law. As such, we should engage people to teach it that are, although a paramedic should be present to contextualise learning from praxis, which would meet the accreditation requirement for academics to hold the relevant qualifications and experience (Paramedicine Board of Australia, 2020). Previous research in the medical field, for example, by Lehmann et al. (2004), found that Deans of U.S. and Canadian medical schools expressed that being a physician was insufficient to be an effective teacher of professional ethics. Similarly, Giubilini et al. (2016) asserted that an ethics expert and a clinical expert would be a more effective way to deliver ethics education sessions. While it is acknowledged teachers of ethics require specialist training (Lehmann et al., 2004) to meet the need for vertical and horizontal integration of ethics throughout the paramedicine curriculum, Stirrat (2010) argues that teaching should be a shared obligation, not only the responsibility of dedicated academics. As such, Stirrat (2010) advocates support for improving the knowledge of all academics regarding ethics and law. Therefore, key stakeholders ought to be engaged to explore best practice regarding academic staff involved in teaching ethics. Stand-alone or Embedded Content/Teaching Format While most (n = 9) of the fifteen universities offering paramedicine have discrete units addressing ethics and law within the curriculum, others have elected to integrate the study of ethics with other subject areas such as professionalism or culture. Accompanying this variance was a disparity in views expressed by academics as to when ethics should be taught. Some felt students needed at least some foundational knowledge prior to any clinical placement, while others felt it important for students to develop an understanding of the profession gained through clinical exposure to support their learning of ethics in paramedicine. Perhaps the answer lies in the overall program design. ‘The Accreditation Standards: Paramedicine’ specifically articulate the need for ‘vertical and horizontal integration of theoretical concepts and practical application throughout the program’ (Paramedicine Board of Australia, 2020, p. 16). The integration of ethics education within current paramedicine curricula was not obvious to the authors of this study. How best to integrate ethics education vertically and horizontally to reflect ethics being intrinsic to paramedic practice is one area that needs further consideration and was raised in discussions with academics as a point of reflection and future action. Paramedicine can learn about improved integration of ethics education from the experience gained in medicine. Teachers of medical ethics and law in U.K. medical schools published a consensus statement, which asserts that: ‘ethics and law teaching should be features of the whole curriculum, should begin early and be reinforced throughout the course’ (Ashcroft et al., 1998, p. 191). Similarly, a 2004 survey of U.S. and Canadian medical schools found that many of their respondents articulated the need for ‘horizontal and longitudinal integration’ of ethics curricula across each year of a medical degree (Lehmann et al., 2004). In 2001 a working group on behalf of teachers of ethics and law in Australian and New Zealand medical schools followed the U.K.’s lead and published a position statement on an integrated ethics curriculum (Braunack‐Mayer et al., 2001). Findings from a recent study by Torda and Mangos (2020) suggests that ethics education is now integrated and interspersed with other content in Australasian medical schools. The medical research demonstrates that a more integrated approach better endows students with core foundational knowledge and experience as they enter the profession and gain further insights from praxis. Paramedic practice involves the integration of clinical and non-clinical skills (including law and ethics). As such, an approach to enhancing ethics education in paramedicine may be to introduce a foundational unit on ethics and law early in the curriculum. This unit would introduce key concepts taught by subject matter (law and ethics) experts who would work alongside a paramedic to provide context. Ethical and legal principles and concepts could then be better incorporated throughout other subjects over the remaining curriculum, and thus more accurately reflect practice which integrates both clinical and non-clinical considerations and decision-making. It must be noted, however, that one of the risks of integrated ethics content is that it becomes diminished if not reinforced as a key aspect of the curriculum (Ashcroft et al., 1998; DuBois & Burkemper, 2002). As a profession, paramedicine could gain from medicine’s experience by engaging key stakeholders to develop a consensus statement on vertical and horizontal integration of ethics within paramedicine curricula. Textbooks The utilisation of paramedic specific textbooks across universities teaching undergraduate paramedicine in Australia is fairly homogenous, with strong engagement with ethics and law texts written by subject matter experts. This differs substantially to international findings, such as those articulated by DuBois and Burkemper (2002, p. 434) of 58 syllabi explored in US medical schools consisting of 1,191 distinct readings, with only eight of these homogenous across six schools. Similarly, a broad and generic list of ethics curriculum resources for emergency medicine graduate medical education was put forth by Marco et al. (2011). Paramedic specific textbooks are particularly important when reflecting on the heterogeneity of academics involved in teaching ethics. Paramedic academics lacking expertise in ethics are likely to be dependent on textbooks to help inform their teaching and serve as a particularly helpful reference given the unique challenges faced in praxis. Content Areas The breadth of content areas found in this study reflects the various approaches taken by universities to fit ethics-related content into a crowded paramedicine curriculum. The problem with the varied approach is that graduates from different universities potentially possess vastly different underpinning knowledge and understanding of how to resolve ethical challenges in practice. Australian and international medical schools addressed this consistency issue by developing position statements on ethics core curriculum (Ashcroft et al., 1998; Braunack‐Mayer et al., 2001). These statements outline the core content, including knowledge, skills, and attitudes, as well as recommended teaching methods and assessment for ethics education in medical schools (Ashcroft et al., 1998; Braunack‐Mayer et al., 2001). Similarly, in Australia, the law discipline has 11 compulsory core units as the minimum academic study requirement for admission to legal practice in Australia, of which legal ethics is one (Victorian Legal Admissions Board, 2021). Paramedicine could consider a similar solution. A national consensus for ethics curricula could consist of agreed core content areas and a range of electives. Electives would be an essential consideration as elements of the ethics curriculum must remain dynamic to accommodate changes in contemporary practice and to ensure students can develop skills in addressing novel questions raised by current events, such as the impact of the COVID-19 pandemic, as one example (Maguire et al., 2020). Approaches to Learning and Teaching (Pedagogical Tools/Approaches) A broad range of pedagogical tools and approaches were found to be used by academics involved in this study. These included the use of modules and workbooks, engaged discussion, group work and collaborative learning. Irrespective of the delivery platform, the use of case studies was common and justified by participants as a way to enable students to practically engage in cases and discussion. The early utilisation of case studies in ethics education in paramedicine is a positive, given the evidence that they are an effective form of teaching (Brooks & Bell, 2017; Lehmann et al., 2004). Medical students struggled with applying knowledge of ethics to clinical cases as much ethics education focused on knowledge acquisition rather than the process of ethical analysis and reasoning (Myser et al., 1995). It was argued that programs should reorient themselves more towards questions that challenge students to extend their current ethical thinking and apply it to the situations faced in ordinary clinical practice (Scher & Kozlowska, 2018, p. 114). Later studies of undergraduate medical programs, such as a scoping review by Souza and Vaswani (2020), found simulation using problem-based learning and case studies featured heavily as teaching methods. Engaging students in small groups utilising these approaches was also effective (Torda & Mangos, 2020). Approaches to learning and teaching ethics and law should be “attuned to the students’ needs appropriate both to their particular stage of training and to relevant specialty-specific ethical issues”(Stirrat, 2010, p.156). This suggests that learning should be scaffolded to become more complex and challenging as students’ progress in their studies, further emphasising the need for vertical and horizontal integration of ethics in the curriculum. Within paramedicine, the current use of case-based scenarios could be used as a foundation for a greater focus on ethics learning outcomes in scaffolded clinical placements to provide students with a practical opportunity to experience, observe, and reflect upon ethical challenges handled in praxis (Brooks & Bell, 2017). This would allow students to develop transferable skills to enable real-world decision-making. Assessment Within this study cohort, written assessment, online tests or quizzes, and examinations were the most utilised assessment approaches. Other methods of assessment included case studies, posters, and group presentations. While written assessment, online quizzes and examinations certainly have their place in assessment strategies, this could be balanced with more applied methods assessing ethical and legal competencies, demonstrated in practical assessments such as OSCEs and simulated case studies. In their study of 125 U.S. and 16 Canadian medical schools, Lehmann et al. (2004) reported that just over half of the medical schools assessed students’ ability to reason morally; in addition, one third formally evaluated students’ performance in ethically difficult simulated situations, such as delivering bad news or discussing do not resuscitate orders. It seems remiss in practically orientated professions such as medicine and paramedicine that more applied strategies are not well utilised. Other suggestions include student reflections after the course, simulated patient interactions, and patient evaluation of the students (Souza & Vaswani, 2020). Students are more likely to integrate clinical ethical reasoning skills into their learning when they understand the practical application of those skills in the management of cases will be assessed (Myser et al., 1995). Furthermore, the assessment of competence in ethics reinforces the importance of ethical analysis and reasoning relative to other discipline areas within the curriculum (Myser et al., 1995; Savulescu et al., 1999). The authors suggest that the assessment strategy should be scaffolded, increasing experiential complexity as students’ progress with their studies. Hence high-fidelity case scenarios using trained actors followed by real-time feedback and student reflection on the case may be a good approach (Marco et al., 2011), particularly if feedback is offered from the situational perspective of both the patient and professional, which Higgs (1987) considers essential. Approaches to scenarios could be adapted for both face-to-face and online learning environments. Limitations All available data from university websites were collected during a short period and taken at face value, supported by consultation with academics where available. As not all universities participated in the discussion activity, there is an inevitable difference in the range and depth of information collected across the fifteen institutions providing undergraduate baccalaureate degrees in paramedicine. This was particularly the case when unit titles did not explicitly mention ethics. Where unit titles were overt, no further exploration of other units occurred – alternate sources of ethics education were only sought where specific units of study were not self-evident. As a result, there may well be introductory sources of ethics education in addition to overt units of study. As previously noted, compared to medicine, the teaching of ethics is still a relatively young field in paramedicine. As such, the medical literature regarding ethics education has been referenced in this paper due to a paucity of relevant literature in the paramedic discipline. It must be noted that a full comparison of the two professions has not been explored in this paper. Perhaps associated with endeavours to achieve consensus on what should be delivered in paramedic ethics education, should be an accompanying consideration of how medicine and paramedicine differ, and how paramedicine could engage in further research to build a stronger foundation of literature from which our own profession can draw. Conclusion Graduates of baccalaureate degrees in paramedicine should possess a common grounding in ethics to meet the professional capabilities expected of registered paramedics in Australia. Findings in this study reveal differences across university courses, with a wide variety of content, methods of delivery, and assessment resulting in a diverse overall design of ethics curricula suggesting that this common ground is not yet present. To date, there has been limited opportunity to engage academics to discuss and achieve consensus about the elements spoken of in this study. However, the evolution of paramedicine as a profession, the new regulatory environment, and the current focus on educational adjustment to meet accreditation requirements serves as the perfect juncture to explore ethics education in the paramedicine curriculum in Australia. Ideally, an ethics curriculum should be integrated horizontally and vertically across the years of the paramedicine program to reflect the experience of praxis and the inherent nature of ethical challenges, experiences and dilemmas encountered by paramedics. Future endeavours should identify barriers to integrated ethics curriculum and seek to address the need for consensus on a responsive model curriculum that prepares graduates to engage holistically in the profession they are about to enter. Acknowledgements The first author would like to acknowledge the Australian Government Research Training Program, which has provided a tuition offset. Author Contributions Kirsty Shearer, Matthew Thomas, Tania Signal, and Nikola Stepanov contributed to the study conception and design. Material preparation, data collection and analysis were performed by Kirsty Shearer. The first draft was written by Kirsty Shearer and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Data Availability The datasets generated and analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics Approval The study was approved by the HREC of CQUniversity Australia (0000022092). Academics participating in discussions were emailed an information sheet and consent form, with consent to participate verbally confirmed prior to any discussion. Competing Interests Kirsty Shearer was appointed as a paid Accreditation Assessor with the Paramedicine Board of Australia during the data collection phase of this study. Matthew Thomas, Tania Signal, Ruth Townsend and Nikola Stepanov have no competing interests to declare. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References Adams J Arnold R Siminoff L Wolfson A Ethical conflicts in the prehospital setting Annals of Emergency Medicine 1992 21 10 1259 1265 10.1016/S0196-0644(05)81759-7 1416311 Ashcroft R Baron D Benatar S Bewley S Boyd K Caddick J Campbell A Cattan A Clayden G Day A Dlugolecka M Dickenson D Doyal L Draper H Farsides B von Fragstein M Fulford K Gillon R Goodman D de Zulueta P Teaching medical ethics and law within medical education: A model for the UK core curriculum Journal of Medical Ethics 1998 24 3 188 192 10.1136/jme.24.3.188 9650114 Australian Health Practitioner Regulation Agency. (2018). Paramedics: Welcome to the profession. Retrieved November 29, 2021, from https://www.ahpra.gov.au/News/2018-11-30-Paramedics-Welcome-to-the-National-Scheme.aspx Bowen GA Document analysis as a qualitative research method Qualitative Research Journal 2009 9 2 27 40 10.3316/QRJ0902027 Braunack-Mayer AJ Gillam LH Vance EF Gillett GR Kerridge IH McPhee J Saul P Smith DE Wellsmore HM Koczwara B Rogers WA Stoffell BF McNeill PM Newell CJ Parker MH Walton M Whitehall JS An ethics core curriculum for Australasian medical schools Medical Journal of Australia 2001 175 4 205 210 10.5694/j.1326-5377.2001.tb143097.x 11587281 Brooks L Bell D Teaching, learning and assessment of medical ethics at the UK medical schools Journal of Medical Ethics 2017 43 9 606 612 10.1136/medethics-2015-103189 27974470 Commonwealth of Australia. (2022). National Registration and Accreditation Scheme. Retrieved January 25, 2022, from https://www.health.gov.au/initiatives-and-programs/national-registration-and-accreditation-scheme Council of Ambulance Authorities. (2022). Where can I study to become a paramedic? Retrieved January 25, 2022, from https://www.caa.net.au/ Downe-Wamboldt B Content analysis: Method, applications, and issues Health Care for Women International 1992 13 3 313 321 10.1080/07399339209516006 1399871 DuBois, J., & Burkemper, J. (2002). Ethics education in U.S. medical schools: A study of syllabi. Academic Medicine, 77(5), 432–437. Retrieved February 17, 2022, from https://journals.lww.com/academicmedicine/Fulltext/2002/05000/Ethics_Education_in_U_S__Medical_Schools__A_Study.19.aspx Elo S Kyngäs H The qualitative content analysis process Journal of Advanced Nursing 2008 62 1 107 115 10.1111/j.1365-2648.2007.04569.x 18352969 Giubilini A Milnes S Savulescu J The medical ethics curriculum in medical schools: Present and future Journal of Clinical Ethics 2016 27 2 129 145 27333063 Hartley PR Paramedic practice and the cultural and religious needs of pre-hospital patients in Victoria 2012 Victoria University Heilicser B Stocking C Siegler M Ethical dilemmas in emergency medical services: The perspective of the emergency medical technician Annals of Emergency Medicine 1996 27 2 239 243 10.1016/S0196-0644(96)70330-X 8629761 Higgs R CABGs and KINGS: Relevance and realism in the teaching of clinical ethics in Camberwell Journal of Medical Ethics 1987 13 3 157 159 10.1136/jme.13.3.157 3669047 Lehmann LS Kasoff WS Koch P Federman DD A survey of medical ethics education at U.S. and Canadian medical schools Academic Medicine 2004 79 7 682 689 10.1097/00001888-200407000-00015 15234922 Maguire, B., O’Neill, B., Shearer, K., McKeown, J., Phelps, S., Gerard, D., Handal, K., & Maniscalco, P. (2020). The ethics of PPE and EMS in the COVID-19 era. Journal of Emergency Medical Services. Retrieved December 15, 2021, from https://www.jems.com/exclusives/ethics-of-ppe-and-ems-in-the-covid-19-era/ Marco CA Lu DW Stettner E Sokolove PE Ufberg JW Noeller TP Ethics curriculum for emergency medicine graduate medical education Journal of Emergency Medicine 2011 40 5 550 556 10.1016/j.jemermed.2010.05.076 20888722 Moore E Respecting an autonomous decision to refuse life-saving treatment: a case study Journal of Paramedic Practice 2020 12 8 304 309 10.12968/jpar.2020.12.8.304 Moritz D Ebbs P Carver H Paramedic ethics, capacity and the treatment of vulnerable patients Journal of Paramedic Practice 2020 12 12 1 7 10.12968/jpar.2020.12.12.CPD1 Myser C Kerridge IH Mitchell KR Teaching clinical ethics as a professional skill: Bridging the gap between knowledge about ethics and its use in clinical practice Journal of Medical Ethics 1995 21 2 97 103 10.1136/jme.21.2.97 7608948 Nordby H Nøhr Ø The ethics of resuscitation: How do paramedics experience ethical dilemmas when faced with cancer patients with cardiac arrest? Prehospital and Disaster Medicine 2012 27 1 64 70 10.1017/S1049023X1200026X 22591932 O'Meara P Duthie S Paramedicine in Australia and New Zealand: A comparative overview The Australian Journal of Rural Health 2018 26 5 363 368 10.1111/ajr.12464 30303284 Olick RS It's ethical, but is it legal? Teaching ethics and law in the medical school curriculum The Anatomical Record 2001 265 1 5 9 10.1002/ar.1035 11241205 Paramedicine Board of Australia. (2018). Code of Conduct for registered health practitioners (interim), June 2018. Retrieved November 30, 2021, from https://www.paramedicineboard.gov.au/Professional-standards/Codes-guidelines-and-policies.aspx Paramedicine Board of Australia. (2020). Accreditation Standards: Paramedicine. Retrieved November 30, 2021, from https://www.paramedicineboard.gov.au/Accreditation/Accreditation-publications-and-resources.aspx Paramedicine Board of Australia. (2021a). About. Retrieved November 29, 2021, from https://www.paramedicineboard.gov.au/About.aspx Paramedicine Board of Australia. (2021b). Professional capabilities for registered paramedics. Retrieved November 30, 2021, from https://www.paramedicineboard.gov.au/Professional-standards/Professional-capabilities-for-registered-paramedics.aspx Savulescu J Crisp R Fulford KW Hope T Evaluating ethics competence in medical education Journal of Medical Ethics 1999 25 5 367 374 10.1136/jme.25.5.367 10536759 Scher S Kozlowska K Rethinking Health Care Ethics 2018 Palgrave Macmillan Shearer K Thomas M Signal T Perceptions of ethical dilemmas in Australian paramedicine Journal of Paramedic Practice 2021 13 8 332 342 10.12968/jpar.2021.13.8.332 Souza A Vaswani V Diversity in approach to teaching and assessing ethics education for medical undergraduates: A scoping review Annals of Medicine and Surgery 2020 56 178 185 10.1016/j.amsu.2020.06.028 32642060 Stirrat G Teaching and learning medical ethics and law in UK medical schools Clinical Ethics 2010 5 3 156 158 10.1258/ce.2010.010029 Torda A Mangos J Medical ethics education in Australian and New Zealand (ANZ) medical schools: a mixed methods study to review how medical ethics is taught in ANZ medical programs International Journal of Ethics Education 2020 5 2 211 224 10.1007/s40889-020-00097-w Victorian Legal Admissions Board. (2021). Academic Qualifications and Training. Retrieved December 1, 2021, from https://www.lawadmissions.vic.gov.au/qualifications-and-training/academic	36466716	PMC9709382	NO-CC CODE	2022-12-01 23:23:04	no	J Acad Ethics. 2022 Nov 30;:1-16	utf-8	J Acad Ethics	2,022	10.1007/s10805-022-09463-3	oa_other
==== Front Lancet Planet Health Lancet Planet Health The Lancet. Planetary Health 2542-5196 The Author(s). Published by Elsevier Ltd. S2542-5196(21)00076-0 10.1016/S2542-5196(21)00076-0 Viewpoint The blueprint of disaster: COVID-19, the Flint water crisis, and unequal ecological impacts Ezell Jerel M PhD ab* Griswold Delilah MS c Chase Elizabeth C MS d Carver Evan PhD e a Africana Studies and Research Center, Cornell University, Ithaca, NY, USA b Cornell Center for Health Equity, Cornell University, Ithaca, NY, USA c Development Sociology, Cornell University, Ithaca, NY, USA d Department of Biostatistics, University of Michigan, Ann Arbor, MI, USA e Program on the Global Environment, University of Chicago, Chicago, IL, USA * Correspondence to: Dr Jerel M Ezell, Africana Studies and Research Center, Cornell University, Ithaca 14850, NY, USA 5 5 2021 5 2021 5 5 2021 5 5 e309e315 © 2021 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY-NC-ND 4.0 license 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. COVID-19 is unique in the scope of its effects on morbidity and mortality. However, the factors contributing to its disparate racial, ethnic, and socioeconomic effects are part of an expansive and continuous history of oppressive social policy and marginalising geopolitics. This history is characterised by institutionally generated spatial inequalities forged through processes of residential segregation and neglectful urban planning. In the USA, aspects of COVID-19's manifestation closely mirror elements of the build-up and response to the Flint crisis, Michigan's racially and class-contoured water crisis that began in 2014, and to other prominent environmental injustice cases, such as the 1995 Chicago (IL, USA) heatwave that severely affected the city's south and west sides, predominantly inhabited by Black people. Each case shares common macrosocial and spatial characteristics and is instructive in showing how civic trust suffers in the aftermath of public health disasters, becoming especially degenerative among historically and spatially marginalised populations. Offering a commentary on the sociogeographical dynamics that gave rise to these crises and this institutional distrust, we discuss how COVID-19 has both inherited and augmented patterns of spatial inequality. We conclude by outlining particular steps that can be taken to prevent and reduce spatial inequalities generated by COVID-19, and by discussing the preliminary steps to restore trust between historically disenfranchised communities and the public officials and institutions tasked with responding to COVID-19. ==== Body pmcThe blueprint of disaster: unequal ecological effects Information slowly trickled in: data passively hinting at an emergent public health crisis and at a rupture in civic governance. Government and health officials were furtive as they became increasingly assured of the scope of the crisis. However, in public they remained silent and demurring. The signals then became undeniable: various facets of the population's morbidity were growing well beyond conventional epidemiological surveillance estimates. What was forthcoming was an unavoidable surge in adverse health outcomes, far beyond what the average US citizen would reasonably imagine, or accept, in their country. This event took place in Flint, Michigan, in 2015, less than 1 year after the city's water source was switched, from Lake Huron to the Flint River. This fateful transition was part of a broad austerity movement spearheaded by an unelected emergency manager, installed by Michigan's then-governor to manage finances in the segregated, socioeconomically disadvantaged city. More than 1 year since the emergence of COVID-19, an uncanny geopolitical enlargement of the racial, ethnic, economic, and spatial inequities seen in the Flint water crisis is taking place because public health disasters that wield disproportionate effects on historically disadvantaged populations share the same macrosocial and ecological blueprint. In this context, the enormity of the current global health catastrophe is impossible to ignore, even for those most insulated from the effects of the pandemic.1, 2 In addition to a more generalised recognition of the gravity of COVID-19, as a seminal cross-cultural event, there is an increasingly broad awareness of the striking racial, ethnic, and spatial inequities that the pandemic acts upon and exacerbates.3 For marginalised communities that have long endured dysfunction both in the health-care system and in their built environment,4 the ongoing pandemic is another situation where a seemingly natural calamity is exacerbated by historically embedded systems of sociospatial inequity. In this Viewpoint, we present the Flint water crisis as a salient and instructive case study and forewarning of how COVID-19 further entrenches existing social inequalities but is lived as bodily and ecological erasure as a result of politicised, unsystematic institutional responses to its broader causes and consequences. In the immediate aftermath of the city's water source switch to the Flint River, lead, an invisible, odourless, and tasteless neurotoxin, was released from the city's antiquated and undertreated water delivery infrastructure into the water supply.5 The surge in the concentration of lead occurred alongside increases in the concentrations of other heavy metals, carcinogenic trihalomethane compounds, and pneumonia-inducing legionella bacteria.6 Well before the contamination was publicly acknowledged by government officials, the broader Flint public, roused by a nascent citizen science movement, discovered that the water supply was contaminated.7 Still, officials and clinicians in the city deflected and negletected their responsibilities, presenting the water quality woes as transient consequences of the water source switch, and residents' emergent health issues as aberrations or casually indeterminate.8 Mona Hanna-Attisha, a paediatrician in Flint, provides a telling account in her book, What the Eyes Don't See,9 describing an epidemic of clinical neglect as resident reports of water contamination-related symptomatology began to pour in: “I don't know if I'd ever felt so stunned and disillusioned and sad all at once. What had I done? Where had I been? Baby after baby had come into our clinic. We gave the same advice. ‘It's fine. Yes, it's fine. Drink the water. Of course it's okay.’ Flint had low breastfeeding rates for a number of reasons. Powdered formula is the norm. To make formula from the powder, you have to mix it with tap water. Meanwhile, for older kids, because of our emphasis on healthy living and lowering sugar in their diets, we are always recommending fewer juices and soda. And more water.” Abundant media and academic comments on the Flint water crisis have acknowledged the structural and strategic racism and associated environmental injustice in action in the crisis.8, 10, 11, 12, 13 Similarly, as the COVID-19 pandemic continues, data are quickly emerging about the gross disparities between historically marginalised populations and their neighbourhoods: people from racial and minority ethnic groups face a heightened risk of exposure to COVID-19 because of social factors rooted in structural inequality and overt discrimination that feed directly into social determinants of health. This reality is illustrated by the fact that a disproportionate number of people from racial and minority ethnic groups have been infected and killed by COVID-19.14, 15 Specific factors that contribute to this morbidity and mortality trend include the higher likelihood that people from racial and minority ethnic groups have inadequate access to housing,16, 17 predisposing health conditions, make up a higher relative proportion of essential workers, or have jobs that cannot be done remotely.18 Moreover, and directly related to the points we stress here, long-standing distrust in both the government and public health institutions further limits their engagement with health-care deployment strategies.19 We describe these factors as a process of reproducing marginalisation, as illustrated by the connections between the distrust that resulted from the Flint water crisis and the scepticism around COVID-19 mitigation measures and vaccination. The prospective consequences of these factors are shown by the outsized devastation of COVID-19 in Black communities; a study in selected US states and cities found that 34% of COVID-19 deaths were among non-Hispanic Black people, although they compose only 12% of the total population of the USA.20 Spatial epidemiology shows how a common genealogy of oppression and distrust consistently reproduces these inequities. Overlapping and embedded crises The blueprint of disaster is fundamentally mediated by fluid and historical processes of social and economic stratification, which are connected to, and frequently motivated by, attitudes to race.21 For the past 4 years, we have conducted community-based, mixed-methods research with lifelong residents and local professional stakeholders in Flint, including government officials and clinicians, to understand the causes, consequences, and recovery needs of the crisis. This research, part of the Flint Engagement Community Project, has focused on identifying and contextualising the attitudes, beliefs, and health outcomes of Flint residents living in the city during, and in the years after, the water crisis. Lessons on national, state, and local government ineptitude derived from the Flint Engagement Community Project foretell many of the public health weaknesses that typify the COVID-19 response. That Flint was unprepared to deal with its water crisis and bore the consequences in population morbidity is obvious. That the city could have been prepared to address what lay ahead, however, is a more fertile debate, pertinent also to COVID-19 as social and health sciences researchers look for so-called best practices from the communities and countries that were more successful in protecting their citizens from the dire extremes of the pandemic. COVID-19's most vivid parallel with the Flint water crisis is its accentuation of exceptional inadequacy of morbidity prevention and management in the US health-care system—especially when compared with countries with similar financial means—which ignored warning signs, was underprepared in terms of resources and public communication (ie, across and between federal, state, and local government, and to the communities), and buckled under the pressure of multiple infection surges. Shortcomings in epidemiological oversight and public health reporting will invariably lead to population health crises, which will almost always exact their most serious effects on the most socioeconomically vulnerable individuals in the population, who are unable to access health promotion resources. What is most deplorable is that this chasm is not natural, but the result of a specific form of moral and technical disinvestment in public health for racially minoritised communities. As COVID-19 spread, WHO issued clear directives to countries on how to confront the pandemic.22 These measures, such as closing schools and workplaces, were easier to implement in wealthier nations, but were also based on an assumption that the areas that heeded this advice would ultimately be able to deploy a functioning health-care system to manage the reduced number of cases—namely, one in which the entire population had adequate access to screening and health care. Although the specific structures of health-care systems worldwide vary, the general lesson of COVID-19 is that mitigation of the crisis will be more effective in areas where public instead of private provisioning is the norm.23, 24 The recurrent failure of the USA to provide such a system limits the remedial capacity of any present or near-future turn towards purposeful investment in approaches to reducing health disparities. This limitation exists, in part, because trust in the US health-care system has been vigorously eroded,25, 26, 27 in racial and minority ethnic groups more than in any other populations.28, 29 The pandemic has shown that there are not only material barriers to care, but also racial, ethnic, and political difficulties in the uptake of mitigation protocols, associated with factors ranging from structural racism to health efficacy, or confidence in one's ability to pursue a healthy life, and health literacy.28, 30 COVID-19 will stand as a particular flash point, as the ruptures of trust happen not just between the public and the government and health-care institutions (eg, local hospitals, public health officials, the US National Institutes of Health, and the Centers for Disease Control and Prevention) but, crucially, within communities, as neighbours engage in fractious debates on the nature of the COVID-19 threat in public spaces and online.31, 32 The ubiquity of this discord suggests that mixed adoption of COVID-19 mitigation strategies is reflective of other forms of social and bodily resistance related to political philosophy and stigma and blame. This observation builds on research emphasising the spatial dynamics of differentiation in this erosion of trust.33, 34 Such a situation has, of course, happened before. Over 20 years ago, much of the hypersegregated metropolitan Chicago (IL, USA) area was affected by a heatwave; roughly 750 people died, most of whom were individuals of low income and older than 65 years, living in under-resourced communities in the city's economically despondent south and west sides, affected by commercial decline and neighbourhood disinvestment.35, 36 During this time, public officials clashed with these community members on their neighbourhoods' poor resilience and dereliction in looking out for one another.37, 38 COVID-19 has wielded similarly stark, disproportionate effects in these same Chicago communities39, 40 while state and local politicians hesitated on shutdown restrictions and, at times, seemingly pivoted towards a focus on a concomitant increase in crime in the city.41, 42, 43 Historically produced intersections of race, sustainability, and urban planning matter. In Chicago's sobering case of environmental health inequity during its mid-1990s heatwave, this intersection was that of the so-called heat islands, scarce tree shading, and commercial exodus (businesses had begun leaving the city's south and west sides in droves in the 1980s);35 in New Orleans (LA, USA), it was a porous and overextended patchwork of levees that bore the brunt of Hurricane Katrina;44 and in Flint, it was a flagging, undertreated water infrastructure system and a dilapidated housing stock.6 With the COVID-19 pandemic, the amalgamation of such lapses in municipal and city planning and design have left parts of the global population to endure a range of unmitigated vulnerabilities produced through the built environment, thus exponentially more exposed to environmental health and infectious disease threats than others.45, 46 Many of the comorbidities directly linked to increased hospitalisation and mortality due to COVID-19 (eg, asthma or chronic obstructive pulmonary disease, diabetes, obesity, cardiac disease, and cancer) are closely associated with built environment factors such as air, water, and soil pollution, unavailability of green spaces and of walkable places, and diminished access to healthy food (so-called food deserts), all of which are traits found far more frequently in low-income, non-White neighbourhoods.47, 48, 49 This mosaic of environmental injustice is attributable to a history of racially motivated community disinvestment and residential segregation and redlining (discriminatory real estate practices engineered to preclude racially minoritised and minority ethnic communities from living in neighbourhoods predominantly inhabited by White people),50 factors that the Flint residents we interviewed were implicitly familiar with. Other inherently unhealthy urban spaces are the shameful legacy of urban renewal, which disproportionately targeted and fragmented low-income, minority ethnic neighbourhoods and that, ironically, replaced dense walkable districts and civic centres with features such as freeways, in the name of public health and economic progress.51, 52, 53 According to the fundamental causes theory, the privilege of accumulated resources leads to the ability to avoid some harms.54 As COVID-19 has shown, “key resources such as knowledge, money, power, prestige, and beneficial social connections can be used no matter what the risk and protective factors are in a given circumstance”.55 Much of this enactment of social and political capital is relational in nature—for example, the influx of tourists from high-income countries led to pronounced surges in COVID-19 cases in minority ethnic enclaves in Asia, Africa, and South America.56, 57, 58 Arline Geronimus' work on the weathering hypothesis59 provides a similar call to action, signalling the potential for COVID-19-related chronic mental illness and trauma for doubly disenfranchised groups, such as Black women and mothers who bear an unequal burden in domestic duties.59, 60 When considering the effects on children, long-term reciprocal disadvantages emerging from COVID-19's sprawling effects on education (eg, remote learning requiring child care and technology that might be difficult to purchase or access) are also a cause for concern.61, 62 Preparing for future outbreaks in marginalised spaces We now turn to observations from our work in Flint to provide some recommendations on how to prepare for further COVID-19-related disparities. Importantly, although access to screening was touted as a panacea for the mitigation for water contamination-related morbidity in Flint and is being prioritised globally for COVID-19 mitigation, screening does not fundamentally address built-in risks that racially minoritised communities frequently face. Although focusing on the reduction of individual-level barriers to screening is important—including by addressing poor health literacy and restricted access to transportation—a fully realised assessment of disasters such as COVID-19, the Flint water crisis, and other similar potential public health calamities requires continuous attention to macrosocial forces. Beyond the failure to consistently and clearly articulate and implement physical distancing guidelines and maintain medical equipment reserves during the incipient stages of the spread of COVID-19 in China,63, 64 many other miscommunications took place during the virus' initial spread in the USA. The variety of misstatements and semantic lapses from governmental and public health leaders were most centrally around how severe COVID-19 was, or could be, and around who should access screening and whether insurance was needed to do so. As we learned in Flint, because of issues caused by little access to information, low health literacy, and low general literacy, many residents were unsure of the health risks posed by consumption of contaminated water, and whether and how to find out if their water was contaminated. Furthermore, residents were confused about the utility of blood lead screenings and whether they were free (which they were, at least for a certain period and at certain clinical outposts in the city), which deterred residents from getting medical screenings. Other residents whom we have engaged held deeply fatalistic perspectives about their health, believing that a positive blood lead screening was a deleterious and irreversible outcome, or were unsure of what to do because of contradictory information that suggested that continued water consumption was safe, but that it may need to be boiled or filtered first. Confusion and disinformation in Flint, like that regarding the origins and severity of COVID-19,65 require ongoing contestation, particularly from local trusted stakeholders—such as faith-based institutions, schools, and small businesses—to generate and sustain the community trust and adherence needed to implement effective mitigation strategies. In Flint, we frequently observed inconsistencies in how the government and health-care providers described both risk and risk mitigation, as officials misrepresented and hesitated in disclosing the potential consequences of exposure to contaminated water, on who needed to get tested for blood lead concentrations and for contaminant sequelae, and on who needed to have their water infrastructure replaced by the city, the centrepiece of the government's response effort.7, 66 Officials also commonly shifted and closed bottled water distribution sites, and altered their positions on issues such as whether or when it was necessary to boil tap water or to use tap water filters to remove contaminants. Therefore, unsurprisingly, many of the measures taken by the government in response to the Flint water crisis were rejected by residents, who were either consummately dubious about or unfamiliar with the potential benefits of these resources, unaware these resources existed, or simply did not have the means to access them. The fact that this reality contributed to increased distrust of local government and of public health-care institutions among Flint residents relates to reproducing marginalisation. Even the most substantial and visible solution implemented by the government in response to Flint's crisis—replacing lead water lines throughout the city—has done little to assuage the internalised suspicions of Flint residents, as had the switch-back to Flint's precrisis water source and the citywide provisioning of tap water filters. Evidence suggests that residents continue to avoid consuming the city's tap water in favour of bottled water,67 although by even the most rigorous accounts and those published by the US Environmental Protection Agency, Flint's water has been consistently safe to consume since at least late 2017.68 The residents' persistent avoidance of the city's tap water highlights how, in addition to the common social determinants of health, there is an additional psychological barrier of eroded trust that must be remediated in communities who have felt chronic and historic neglect from public health institutions, even if future investments are made at the level of services and resource provisioning. This nuanced form of social vulnerability has clear historical antecedents. Little investment in public health infrastructures (in terms of services, resources, and the establishment of transparent, culturally sensitive public communication standards) reproduces racial and social inequities, contributing to the inability of these institutions to effectively mitigate bodily and psychological risk for individuals, or invest in the necessary work of cultivating community trust. In the case of COVID-19, existing social vulnerability and distrust is heightened by previous experiences of institutional and health-care disenfranchisement. For example, questions about the effectiveness of face masks and vaccines further contribute to an insufficient use of mitigation strategies, and to a general sense that political and public health officials are misguided or dubious about their own proposed strategies (eg, consider the high-profile cases of politicians not wearing masks or attending gatherings that are not physically distanced).69, 70 As the Tuskegee syphilis study71 and the broader legacy of racialised medical experimentation in high-income countries poignantly affirm,71, 72 breaches such as these have both short-term and long-term consequences. In conclusion, the interchangeability of the very notion of crisis should be recognised; scale and impact are relative and, thus, subjective. Now, more than 7 years after the start of the Flint water crisis, the episode has largely faded from the public's, and from scholars', gaze. Nevertheless, with the benefit of hindsight, there are many lessons to be taken from the situation. Rectifying the wrongs of institutionally generated environmental trauma, to the extent that can be done, requires a reorganisation of systems and structures that have helped to generate and sustain frequent racial and other forms of social inequity in the USA. A complete description of acts necessary to do so is beyond the scope of this Viewpoint. However, a focus on where this effort can begin might be instructive. First and foremost, public health officials should make dense and spatially vulnerable communities—often racially minoritised communities—explicit priorities in terms of allocation of mitigation resources, including masks, screenings, treatment, and other forms of material and social support. Recognising that the effects of COVID-19 have followed classic lines of social and economic stratification, the most intuitive approach to prevent, disrupt, and ultimately undo the pattern of disaster is concentrating resources in communities where stratification is most pronounced. Simultaneously, public health officials should strive for consistency in communicating the utility and means of accessing these resources, insofar as scientific predictability allows, and enrol trusted community stakeholders as the primary messengers. Although much attention has been paid to the physical consequences of COVID-19, commensurate attention must be paid also to the potential impact of COVID-19 on mental health and trauma. Evidence has consistently shown that, after large-scale crises, survivors are prone to have severe depression, anxiety, and trauma.73, 74 Because socioenvironmental distress is already pronounced in many racially minoritised and minority ethnic populations,75 COVID-19 will probably intensify embedded existing mental health disparities. As we found by speaking with Flint residents, the effect of COVID-19 will not just be on the surface, manifesting on the body of individuals, but will come to reshape their mental health profiles.76 Indeed, the effects will remain fixed in the psyches of many of those who were directly affected and felt that they were targeted or preyed upon, through symptomatology that they observe in themselves and their loved ones; and for those who are more indirectly affected through ongoing social isolation, job and education loss, and other factors. In addition to buffering local economic and educational opportunities (an upstream approach), officials should work to ensure that mental health resources are broadly accessible in these communities as part of the recovery efforts. Furthermore, there is also a pivotal and perhaps time-limited opportunity to tend to previous environmental injustices and show concern for the overall wellness of specific populations and the health of their environments. Engaging racially minoritised and minority ethnic communities in the environmental remediation process, greening, and other measures to improve local ecosystem services and neighbourhood liveability—without immediately triggering gentrification and displacement—will be crucial to promote a public health preparedness ethos, to rebuild trust, and to combat the underlying health factors that predispose many people in these communities to deleterious contacts with infectious diseases and with public and environmental health crises more broadly. A reckoning with the low uptake of preventive public health measures and sentiments of social and bodily vulnerability is likely to be most effectively achieved through efforts that empower individuals and restore a sense and possibility of individual agency, community ownership and capacity, and institutional earnestness and goodwill. Declaration of interests We declare no competing interests. Acknowledgments We thank the participants of the Flint Community Engagement Project for their intentionality and vast intellectual contributions to the research underlying this Viewpoint. Contributors All authors contributed to the conception, writing, and editing of this Viewpoint. JME did the data collection and community outreach for the underlying research in Flint that is discussed here. ==== Refs References 1 Rouf K Wainwright T Linking health justice, social justice, and climate justice Lancet Planet Health 4 2020 e131 e132 32353288 2 Ruger JP The injustice of COVID-19: we need a moral constitution for our planet's health Lancet Planet Health 4 2020 e264 e265 32681893 3 Blundell R Costa Dias M Joyce R Xu X COVID-19 and inequalities Fisc Stud 41 2020 291 319 32836542 4 Frumkin H Health, equity, and the built environment Environ Health Perspect 113 2005 A290 A291 15866747 5 Pieper KJ Tang M Edwards MA Flint water crisis caused by interrupted corrosion control: investigating “ground zero” home Environ Sci Technol 51 2017 2007 2014 28145123 6 Masten SJ Davies SH Mcelmurry SP Flint water crisis: what happened and why? J Am Water Works Assoc 108 2016 22 34 28316336 7 Peplow M The Flint water crisis: how citizen scientists exposed poisonous politics Nature 559 2018 180 181 8 Ahlness E Flint fights back: environmental justice and democracy in the Flint water crisis Electron Green J 1 2019 1 3 9 Hanna-Attisha M What the eyes don't see: a story of crisis, resistance, and hope in an American city 2019 One World New York, NY 10 Butler LJ Scammell MK Benson EB The Flint, Michigan, water crisis: a case study in regulatory failure and environmental injustice Environ Justice 9 2016 93 97 11 Campbell C Greenberg R Mankikar D Ross RD A case study of environmental injustice: the failure in Flint Int J Environ Res Public Health 13 2016 951 27690065 12 Hammer PJ The Flint water crisis, the Karegnondi water authority and strategic–structural racism Crit Sociol 45 2019 103 119 13 Pulido L Flint, environmental racism, and racial capitalism Capital Nat Social 24 2016 1 16 14 Lewis NM Friedrichs M Wagstaff S Disparities in COVID-19 incidence, hospitalizations, and testing, by area-level deprivation—Utah, March 3–July 9, 2020 MMWR Morb Mortal Wkly Rep 69 2020 1369 1373 32970656 15 Millett GA Jones AT Benkeser D Assessing differential impacts of COVID-19 on black communities Ann Epidemiol 47 2020 37 44 32419766 16 Okoh AK Sossou C Dangayach NS Coronavirus disease 19 in minority populations of Newark, New Jersey Int J Equity Health 19 2020 93 32522191 17 Rodriguez-Lonebear D Barceló NE Akee R Carroll SR American Indian reservations and COVID-19: correlates of early infection rates in the pandemic J Public Health Manag Pract 26 2020 371 377 32433389 18 Hawkins D Differential occupational risk for COVID-19 and other infection exposure according to race and ethnicity Am J Ind Med 63 2020 817 820 32539166 19 Institute of Medicine (US) Committee on Assuring the Health of the Public in the 21st Century The future of the public's health in the 21st century 2003 National Academy Press Washington DC 20 Holmes L Jr Enwere M Williams J Black–White risk differentials in COVID-19 (SARS-COV-2) transmission, mortality and case fatality in the United States: translational epidemiologic perspective and challenges Int J Environ Res Public Health 17 2020 4322 21 Neely B Samura M Social geographies of race: connecting race and space Ethn Racial Stud 34 2011 1933 1952 22 Sohrabi C Alsafi Z O'Neill N World Health Organization declares global emergency: a review of the 2019 novel coronavirus (COVID-19) Int J Surg 76 2020 71 76 32112977 23 Armocida B Formenti B Palestra F Ussai S Missoni E COVID-19: universal health coverage now more than ever J Glob Health 10 2020 010350 24 Blumenthal D Fowler EJ Abrams M Collins SR Covid-19—implications for the health care system N Engl J Med 383 2020 1483 1488 32706956 25 Han Q Zheng B Cristea M Trust in government and its associations with health behaviour and prosocial behaviour during the COVID-19 pandemic PsyArXiv 2020 published online June 29. 10.31234/osf.io/p5gns (preprint). 26 Jaiswal J LoSchiavo C Perlman DC Disinformation, misinformation and inequality-driven mistrust in the time of COVID-19: lessons unlearned from AIDS denialism AIDS Behav 24 2020 2776 2780 32440972 27 Sibley CG Greaves LM Satherley N Effects of the COVID-19 pandemic and nationwide lockdown on trust, attitudes toward government, and well-being Am Psychol 75 2020 618 630 32496074 28 Phiri P Delanerolle G Al-Sudani A Rathod S COVID-19 and Black, Asian, and Minority Ethnic communities: a complex relationship without just cause JMIR Public Health Surveill 7 2021 e22581 29 Smith Jervelund S Eikemo TA The double burden of COVID-19 Scand J Public Health 49 2021 1 4 33528311 30 Davlantes E Tippins A Espinosa C Mitigating SARS-CoV-2 transmission in Hispanic and Latino communities-Prince William Health District, Virginia, June 2020 J Racial Ethn Health Disparities 2021 published online Feb 4. 10.1007/S40615-021-00968-Y 31 May T Anti-vaxxers, politicization of science, and the need for trust in pandemic response J Health Commun 25 2020 761 763 33345732 32 Rothgerber H Wilson T Whaley D Politicizing the COVID-19 pandemic: ideological differences in adherence to social distancing PsyArXiv 2020 published online April 22. 10.31234/osf.io/k23cv (preprint). 33 Elgar FJ Stefaniak A Wohl MJA The trouble with trust: time-series analysis of social capital, income inequality, and COVID-19 deaths in 84 countries Soc Sci Med 263 2020 113365 34 Krishnan L Michelle Ogunwole S Cooper LA Historical insights on coronavirus disease 2019 (COVID-19), the 1918 influenza pandemic, and racial disparities: illuminating a path forward Ann Intern Med 173 2020 474 481 32501754 35 Browning CR Wallace D Feinberg SL Cagney KA Neighborhood social processes, physical conditions, and disaster-related mortality: the case of the 1995 Chicago heat wave Am Sociol Rev 71 2006 661 678 36 Klinenberg E Heat wave: a social autopsy of disaster in Chicago 2015 University of Chicago Press Chicago, IL 37 Klinenberg E Denaturalizing disaster: a social autopsy of the 1995 Chicago heat wave Theory Soc 28 1999 239 295 38 Klinenberg E Blaming the victims: hearsay, labeling, and the hazards of quick-hit disaster ethnography Am Sociol Rev 71 2006 689 698 39 Kim SJ Bostwick W Social vulnerability and racial inequality in COVID-19 deaths in Chicago Health Educ Behav 47 2020 509 513 32436405 40 Maroko AR Nash D Pavilonis BT COVID-19 and inequity: a comparative spatial analysis of New York City and Chicago hot spots J Urban Health 97 2020 461 470 32691212 41 Campedelli GM Favarin S Aziani A Piquero AR Disentangling community-level changes in crime trends during the COVID-19 pandemic in Chicago Crime Sci 9 2020 21 33134029 42 Husain N Reyes C Before data showed Chicago blacks dying at higher rates, communities of color knew recovery from COVID-19 would be slow https://www.chicagotribune.com/coronavirus/ct-coronavirus-chicago-health-disparities-data-20200410-rf7lmmvgurfwxpxiatebsozwsu-story.html April 21, 2020 43 Taylor K-Y The black plague https://www.newyorker.com/news/our-columnists/the-black-plague April 16, 2020 44 Leavitt WM Kiefer JJ Infrastructure interdependency and the creation of a normal disaster: the case of hurricane Katrina and the city of New Orleans Public Works Manag Policy 10 2006 306 314 45 Dietz L Horve PF Coil DA Fretz M Eisen JA Van Den Wymelenberg K 2019 novel coronavirus (COVID-19) pandemic: built environment considerations to reduce transmission mSystems 5 2020 e00245 e00320 32265315 46 Fears R Gillett W Haines A Norton M Ter Meulen V Post-pandemic recovery: use of scientific advice to achieve social equity, planetary health, and economic benefits Lancet Planet Health 4 2020 e383 e384 32918882 47 Cutts BB Darby KJ Boone CG Brewis A City structure, obesity, and environmental justice: an integrated analysis of physical and social barriers to walkable streets and park access Soc Sci Med 69 2009 1314 1322 19751959 48 Wolch JR Byrne J Newell JP Urban green space, public health, and environmental justice: the challenge of making cities ‘just green enough’ Landsc Urban Plan 125 2014 234 244 49 Rundle A Neckerman KM Freeman L Neighborhood food environment and walkability predict obesity in New York City Environ Health Perspect 117 2009 442 447 19337520 50 Rothstein R The color of law: a forgotten history of how our government segregated America 2017 Liveright Publishing New York, NY 51 Highsmith AR Demolition means progress: Flint, Michigan, and the fate of the American metropolis 2015 University of Chicago Press Chicago, IL 52 Rast J Regime building, institution building: urban renewal policy in Chicago, 1946–1962 J Urban Aff 31 2009 173 194 53 Teaford JC Urban renewal and its aftermath Hous Policy Debate 11 2000 443 465 54 Link BG Phelan J Social conditions as fundamental causes of disease J Health Soc Behav 35 1995 80 94 55 Phelan JC Link BG Tehranifar P Social conditions as fundamental causes of health inequalities: theory, evidence, and policy implications J Health Soc Behav 51 suppl 2010 S28 S40 20943581 56 Gilbert M Pullano G Pinotti F Preparedness and vulnerability of African countries against importations of COVID-19: a modelling study Lancet 395 2020 871 877 32087820 57 Marinelli NP Albuquerque LPA Sousa IDB Batista FMA Mascarenhas MDM Rodrigues MTP Evolution of indicators and service capacity at the beginning of the COVID-19 epidemic in Northeast Brazil, 2020 Epidemiol Serv Saude 29 2020 e2020226 58 Tremblay-Huet S COVID-19 leads to a new context for the “right to tourism”: a reset of tourists' perspectives on space appropriation is needed Tour Geogr 22 2020 720 723 59 Geronimus AT The weathering hypothesis and the health of African-American women and infants: evidence and speculations Ethn Dis 2 1992 207 221 1467758 60 Simons RL Lei M-K Klopack E Beach SRH Gibbons FX Philibert RA The effects of social adversity, discrimination, and health risk behaviors on the accelerated aging of African Americans: further support for the weathering hypothesis Soc Sci Med 2020 published online July 7. 10.1016/j.socscimed.2020.113169 61 Azevedo JP Hasan A Goldemberg D Iqbal SA Geven K Simulating the potential impacts of covid-19 school closures on schooling and learning outcomes: a set of global estimates 2020 The World Bank https://pubdocs.worldbank.org/en/798061592482682799/covid-and-education-June17-r6.pdf 62 Van Lancker W Parolin Z COVID-19, school closures, and child poverty: a social crisis in the making Lancet Public Health 5 2020 e243 e244 32275858 63 Alvarez FE Argente D Lippi F A simple planning problem for COVID-19 lockdown April, 2020 National Bureau of Economic Research https://www.nber.org/papers/w26981 64 Sharifi A Khavarian-Garmsir AR The COVID-19 pandemic: impacts on cities and major lessons for urban planning, design, and management Sci Total Environ 749 2020 142391 65 Brzezinski A Kecht V Van Dijcke D Wright AL Belief in science influences physical distancing in response to COVID-19 lockdown policies. Working paper 2020–56 April, 2020 University of Chicago Becker Friedman Institute for Economics https://repec.bfi.uchicago.edu/RePEc/pdfs/BFI_WP_202056.pdf 66 Carey MC Lichtenwalter J “Flint can't get in the hearing”: the language of urban pathology in coverage of an American public health crisis J Commun Inq 44 2020 26 47 67 Kruger DJ Cupal S Franzen SP Toxic trauma: household water quality experiences predict posttraumatic stress disorder symptoms during the Flint, Michigan, water crisis J Community Psychol 45 2017 957 962 68 Hughes S Flint, Michigan, and the politics of safe drinking water in the United States Perspect Polit 2020 published online July 13. 10.1017/S153759272000136X 69 Peeples L What the data say about wearing face masks Nature 586 2020 186 189 33024333 70 Plohl N Musil B Modeling compliance with COVID-19 prevention guidelines: the critical role of trust in science Psychol Health Med 26 2021 1 12 71 Washington HA Medical apartheid: the dark history of medical experimentation on Black Americans from colonial times to the present 2006 Doubleday Books New York, NY 72 Davis D-A Obstetric racism: the racial politics of pregnancy, labor, and birthing Med Anthropol 38 2019 560 573 30521376 73 Bromet EJ Mental health consequences of the Chernobyl disaster J Radiol Prot 32 2012 N71 N75 22394694 74 Gill DA Secondary trauma or secondary disaster? Insights from Hurricane Katrina Sociol Spectr 27 2007 613 632 75 Matthews SA Yang T-C Exploring the role of the built and social neighborhood environment in moderating stress and health Ann Behav Med 39 2010 170 183 20300905 76 Ezell JM Chase EC A population-based assessment of physical symptoms and mental health outcomes among adults following the Flint water crisis J Urban Health 2021 published online March 31. 10.1007/S11524-021-00525-2	33964240	PMC9709384	NO-CC CODE	2022-12-01 23:23:05	no	Lancet Planet Health. 2021 May 5; 5(5):e309-e315	utf-8	Lancet Planet Health	2,021	10.1016/S2542-5196(21)00076-0	oa_other
==== Front Int J Nurs Stud Int J Nurs Stud International Journal of Nursing Studies 0020-7489 1873-491X Elsevier Ltd. S0020-7489(22)00218-8 10.1016/j.ijnurstu.2022.104389 104389 Article Mandatory COVID-19 vaccination for healthcare workers: A discussion paper Maneze Della abcd⁎ Salamonson Yenna abd Grollman Maxwell e Montayre Jed bd Ramjan Lucie bd a University of Wollongong, School of Nursing, Wollongong, NSW, Australia b Western Sydney University, School of Nursing and Midwifery, Australia c South Western Sydney Local Health District, Multicultural Health Service, Australia d Australian Centre for Integration of Oral Health (ACIOH), Ingham Institute for Applied Medical Research, Liverpool, NSW 2170, Australia e University of California, Los Angeles, Institute for Society and Genetics, Los Angeles, CA, USA ⁎ Corresponding author at: University of Wollongong, School of Nursing, Australia. 9 11 2022 2 2023 9 11 2022 138 104389104389 25 5 2022 31 10 2022 1 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 devastating effects of COVID-19 sparked debates among professionals in the fields of health, law, and bioethics regarding policies on mandatory vaccination for healthcare workers. Suboptimal vaccine uptake among healthcare workers had been implicated in the increased risk of nosocomial spread of COVID infection and absenteeism among healthcare workers, impacting the quality of patient care. However, mandatory vaccine policies were also seen to encroach on the autonomy of healthcare workers. Aims and objectives To synthesise the arguments for and against mandatory vaccination for healthcare workers (HCWs) and its long-term impact on the healthcare workforce, through an analysis of texts and opinions of professionals from different fields of study. Methods This is a systematic review of opinions published in peer-reviewed journals. After initial search in Cochrane and JBI systematic review databases to ensure no previous review had been done, five databases were searched (PsychInfo, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Medline and Scopus). Inclusion criteria were: 1) focused on COVID-19; 2) healthcare workers specific; 3) specific to mandatory vaccination; 4) opinion piece with an identified author; and 5) in English. Exclusion: 1) focus on other vaccine preventable diseases, not COVID-19 and 2) discussion on mandatory vaccination not-specific to healthcare workers. The Joanna Briggs Critical Appraisal tool for Text and Opinions was used to assess quality. Data were synthesised in the summary table. Results The review included 28 opinion and viewpoint articles. Of these, 12 (43 %) adopted a pro-mandatory vaccination stance, 13 (46 %) were neutral or had presented arguments from both sides of the debate and only three (11 %) were against. The overall arguments among those who were pro-, neutral and anti-mandatory COVID-19 vaccination were underpinned by ethical, moral and legal principles of such a mandate on a vulnerable healthcare workforce. This review highlighted the polarised opinions concerning choices, human rights, professional responsibilities and personal risks (i.e. health risks, losing a job) with the introduction of vaccination mandate. However, the articles found in this review discussed mandatory vaccination of healthcare workers in the USA, Europe and Australia only. Conclusion The review underscores the need to balance the rights of the public to safe and quality care with the rights and moral obligations of healthcare workers during a public health emergency. This can be achieved when policies and mandates are guided by reliable scientific evidence which are flexible in considering legal and ethical dilemmas. Tweetable abstract To mandate or not to mandate COVID-19 vaccination for healthcare workers: A synthesis of published opinions in health, law, and bioethics. Keywords Policy Mandatory vaccine Care workers Corona virus disease SARS-CoV-2 Vaccination Ethics Legal ==== Body pmcWhat is already known • Mandatory vaccination policies may be implemented in cases of vaccine hesitancy among healthcare workers. • Healthcare workers are at higher risk of acquiring nosocomial SARS-CoV-2 infection and transmitting it to vulnerable and immunocompromised individuals. • Mandatory vaccination for healthcare workers has been previously implemented to prevent spread of vaccine preventable infectious diseases. What this paper adds • Arguments in favour of mandatory vaccination for healthcare workers prioritised two key tenets based on fiduciary duty and ethical responsibilities. • Arguments against mandatory vaccination for healthcare workers focused on the violation of rights to autonomy, legal grounds for exemptions, and compensation for inadvertent harm. • A collaborative and participatory review of mandatory vaccination policies is recommended to ensure the right balance between autonomy and public safety. 1 Background World Health Organization (WHO) surveillance data reported that since the start of the pandemic in January 2020, there had been approximately 613,410,796 confirmed cases of COVID-19 infection and 6,518,749 COVID-19 related deaths globally (to September 2022) (World Health Organization, 2021; American Library Association, 2022). Of these, healthcare workers (HCWs) represented an estimated 80,000 to 180,000 (with a population-based estimate of 115,493), recorded from January 2020 to May 2021 to World Health Organization (2021). A more recent review of cross-sectional studies showed that the global pooled prevalence of COVID-19 infection among HCWs was calculated to be 7 % (using the antibody method) and 11 % (using the PCR method) (Dzinamarira et al., 2021). The COVID-19 pandemic has highlighted the hazardous nature of working in healthcare not only for HCWs and their families, but also for the patients that they care for. Before the emergence of variants, on average, each infected person could transmit the infection to 2.2 people (Li et al., 2020), but a more recent study showed increased transmissibility of the infection by 43 % to 82 % (Davies et al., 2021). Considering that many HCWs are in close contact with vulnerable and immunocompromised patients, the risk of transmitting the disease and worsening their health is high. In fact nosocomial transmission of COVID-19 infection in early 2020 was estimated to be 15 % in a London hospital with a high case fatality rate of 36 % (Rickman et al., 2021). Whilst vaccine development has been fast tracked and made available, there are reports of vaccine hesitancy and poor vaccine uptake among HCWs (Al-Amer et al., 2022) sparking debate regarding mandatory vaccination for HCWs. This is because voluntary vaccination has not achieved the target vaccination rates, prompting some governments to introduce mandatory vaccination policies for the protection of public health (Cole and Swendiman, 2015; Flood et al., 2021). The notion of instituting mandatory vaccination is not new nor unique. The historical roots can be traced back to the Ottoman Empire in the 1700s (Evered and Evered, 2020), with the introduction of variolation and later vaccination by Jenner in 1796 (Stewart and Devlin, 2006). Individuals were compelled to receive vaccination in the mid-1800s to the mid-1900s in many countries with subsequent eradication of smallpox declared by WHO in 1980 (Esparza et al., 2018). More recently, compulsory vaccination was also instituted for HCWs against influenza after strong recommendations achieved no higher than 50 % vaccination uptake among HCWs (Moran et al., 2019). Vaccination is recognised as an important component in the fight against pandemic level infection. In the current COVID-19 pandemic, unvaccinated individuals are reported to be 17 times more likely to be hospitalised (Havers et al., 2022) and 11 times more likely to die from COVID-19 infection compared to those who are vaccinated (Dyer, 2021). Whilst mandatory vaccination is not new, professionals in the fields of health, bioethics and the legal system continue to express differing opinions about the issue of mandating COVID-19 vaccination for HCWs. Since the declaration of the COVID-19 pandemic, a range of perspectives, opinions and discussion papers have been published in peer-reviewed journals on this issue. Hence, it is timely to synthesise the views on mandatory vaccination for HCWs from different perspectives to enable people to make an informed judgement based on the evidence provided by professionals on both sides of the issue (McArthur et al., 2015). Hence, the aim of this paper is to understand mandatory vaccination for HCWs through a synthesis and critical analysis of different opinions from multiple sectors. Specifically, it seeks to answer the question: What are the moral, legal and ethical arguments regarding mandatory vaccination for healthcare workers? 2 Methods The discussion for or against mandatory vaccination for healthcare workers was based on the integrated opinions of professionals in the field of health, bioethics and law. A systematic search of literature was undertaken in the following databases: PsychInfo, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Medline and Scopus from January 2020 to November 2021. Search terms (Table 1 ) included healthcare workers and variations to the term including health professionals, health staff, health workers, care workers and specific health workers such as nurses, physicians and allied health professionals. This was then combined with COVID-19 or corona virus, vaccine or vaccination and the term mandatory including similar terms such as compulsory, mandate and requirement or required. In addition, forward (those who cited the included articles) and backward (those that were cited by authors of the included articles) searches were conducted to ensure that all relevant articles were included. Articles were included if: 1) the focus was only on COVID-19 or corona virus; 2) specific to healthcare workers; 3) centred on mandatory vaccination; 4) an opinion piece, editorial or commentary with an identified author; and 5) written in English. News articles or surveys and those that discussed other vaccine preventable diseases other than COVID-19 were excluded. Opinion pieces about COVID-19 mandatory vaccination for non-healthcare workers or the general public were likewise excluded. After full text review, backward and forward searches were carried out through Google Scholar searching to ensure comprehensiveness. All authors (DM, MG, YS, JM, LR) summarised each included article and compiled this in a table (Table 2 ).Table 1 Search strategy. Table 1 Search terms Population Healthcare workers, health professionals, health staff, care workers and specific healthcare workers (nurses, doctors, physicians, allied health professionals), hospital staff, health staff Interest Mandatory vaccination, compulsory vaccination, vaccination requirement Context COVID-19, corona virus, COVID-19 (MESH) OR SARS-CoV-2 (MESH) OR COVID-19 Vaccines (MESH) OR vaccines (MESH) OR vaccination (MESH) OR immunization (MESH) OR Immunization programs (MESH) or COVID-19 OR COVID OR "COVID-19 pandemic" OR "Severe Acute Respiratory Syndrome Coronavirus 2" OR "2019 NCOV" OR SARS-CoV-2 Table 2 Summary table. Table 2No First author, month & year of publication; country of study; type of publication Credentials & qualifications of first author Position on mandatory vaccine Key points \| key questions Findings \| considerations \| recommendations Comments 1 Ayukekbong, 2021 Canada Published in: Canadian Journal of Infection Control, editorial Summer (June–August) 2021 BMLS, MSc, PhD, CIC, Editor-in-Chief Pro mandatory vaccine especially among HCWs o Refers to previous bioethical analysis justifying public health coercion when necessary o Compares mandatory vaccination to other covid policies like masking and contact tracing that infringe on human rights o Take measures to boost confidence in vaccine o Require vaccine unless legitimate excuse o Explore options to minimize risk of unvaccinated to others if necessary o Nuanced bioethical look into the topic o Carefully articulates and weighs other side of the argument 2 Baker, N., 2021 USA Published in: Alabama Nurse, opinion piece Aug, Sept, Oct 2021 DNP, CRNP, GS-C, CNE, FAANP, University of Alabama at Birmingham School of Nursing Neutral o Legality of mandatory vaccinations for HCWs discussed o Federal and local legislation as well as union bargaining could play a role o Religious and medical exemptions would likely be necessary o Requiring vaccination for only specific departments could be less tenuous o They argue debate will not end quickly in US o Legal troubles and exemptions are one worry o They don't directly state their opinion on mandatory vaccination, so it may not fit the criteria? 3 Bowen, 2020 USA Published in: Clinical Chimica Acta; letter to the editor Aug 2020 PhD, MHA, MLT., Clinical Professor of Pathology, Stanford Medicine Yes, but only after an extensive review and careful consideration o Refers to deontological, utilitarian, and bioethical theories and principles to analyse ethics of mandate o Carefully weighs benefits, risks, and issues of concern o Argues decision should be dependent on how effective vaccine is and what the risks of side effects are o Develops meticulous criteria for how an ethical and effective mandatory vaccine policy can be created o Argues for equitable distribution of risk and benefit and vaccine injury compensation program o Piece published before vaccine safety and efficacy known, so provides a unique and interesting perspective 4 Bradfield, 2021 Australia Published in: Clinical Ethics, opinion piece May 2021 Medical Practitioner, Health Ethics and PhD Candidate, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Victoria, Australia Pro conditional vaccination (a form of mandatory vaccination) o Vigorously defends mandatory vaccination for HCWs from a bioethical perspective o Nuanced argument for a conditional vaccination requirement where covid-vaccine resistant HCWs are first moved to non-clinical roles/given paid/unpaid leave for period before being terminated o Concludes patient would be more harmed by not having a mandate than HCWs would be harmed with a mandate o Recommends conditional vaccine compromise o Acknowledges potential drawbacks of proposal including potential for higher morbidity and death o Argument for conditional mandatory vaccination for HCWs (intermediate approach) is fairly unique o Structure and analysis of article suggest it was written before safety/efficacy data known? Would their argument still be the same now? 5 Dean, 2021 UK Published in: Nursing Management, opinion piece June 2021 Freelance health writer Neutral o Discusses overview of proposed policy as well as precedence for it abroad and domestically within England o Brings up significant opposition to mandatory COVID-19 vaccine policy among nurses o Legality of policy also discussed o Although benefits are discussed, article suggests policy would face pushback from a legal and acceptance perspective o Article doesn't state position on mandatory vaccine policy for HCWs, so it may not fit the criteria? 6 Emmanuel, 2021 USA Published in: Annals of Internal Medicine, opinion piece July 2021 MD, PhD; Department of Medical Ethics and Health Policy. University of Pennsylvania Pro-mandatory vaccination Key points: Mandatory vaccination is an ethical issue. It covers two aspects:o General ethical duty of HCPs to protect others when threat from vaccine harm to oneself is very minimal. o HCPs have ethical and professional responsibility to protect others, by ensuring the health and wellbeing to all population groups. Vaccination mandate is not new to HCPs, but an extension of existing practices and policies during the COVID pandemic. Consideration in mandating COVID vaccination: Mandating vaccination is an ethical obligation that employers should fulfill in healthcare and long-term care settings. A joint statement was created on mandatory vaccination endorsed and supported by 88 organisations. 7 Flood, 2021 Canada Published in: Canadian Medical Association Journal, opinion piece February 2021 SJD; MD, Faculty of Law Common Law Section, and Department of Medicine (Wilson), School of Epidemiology and Public Health, University of Ottawa Neutral Key points: Mandatory vaccination comes with several legal implications when implemented among HCPs. The legal issues in implementing vaccine mandate sit in the national government and the labour organisations (employers). A vaccine mandate will be challenged by extant laws and policies and legal conditions created toward its implementation may vary as some private organisations have their own vaccination policies for healthcare workers. Consideration in mandating COVID vaccination: Mandating vaccinations for HCPs will require strong scientific evidence of the inadequacy of other measures such as PPEs (mask), and a strong justification that vaccination is the ultimate solution. A vaccination mandate should have the exemptions (those who cannot be vaccinated and bonafide religious reasons). There is also a need for a policy to protect those who might be harmed by mandatory vaccination (no-fault vaccination compensation). Mandatory vaccination as a legal issue will need to be justified in the existing laws using scientific evidence of vaccine benefits. Issues and legal processes from previous mandates such as compulsory influenza vaccination among HCPs provide insights on the implementation of mandatory vaccinations for COVID-19. 8 Frati, 2021 Italy Published in: Vaccines, opinion piece August 2021 Department of Anatomical, Histological, Forensic and Orthopaedical Sciences, Sapienza University of Rome, Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS) Neuromed, Department of Clinical and Experimental Medicine, University of Foggia, Foggia, Italy Pro mandatory vaccine o Discusses legality of compulsory COVID-19 vaccines for HCWs in Europe's legal framework o Explores reasons for HCW hesitancy o Argues benefits of increasing vaccine uptake both for health of HCWs and patients o Mandatory COVID-19 vaccine is ethically and legally justifiable in Italy Structure and goals of the piece were a little more difficult for me to decipher in this one versus other ones 9 & 10 Glasper, 2021 UK Published: Health Matters; July 2021 British Journal of Nursing, July 2021 Emeritus Professor of Nursing, University of Southampton PhD, BA, RSCN, RGN, ONC, DN, CertEd, RNT (Professor of Child Health Care, School of Nursing & Midwifery, University of Southampton, UK) Neutral Key points: Mandatory COVID-19 vaccination among HCPs as proposed in the UK aimed at protecting the public and the healthcare workers. Vaccination mandate might lead to unintended consequences as staff resignation and further mistrust to the health system to those who are vaccine hesitant. Consideration in mandating COVID vaccination: Support from peak bodies and professional organisations is important in the implementation of vaccine mandate. In the UK, the main peak body for nurses - RCN is not supportive of mandatory vaccination. Mandatory vaccination for HCPs in aged care homes, aimed at protecting the most vulnerable (older people) is still a huge issue and might cause healthcare workers in aged care to leave, which causes further strain to the aged care sector during a pandemic. 11 Green, 2021 UK Published in: BMJ (letters) July 2021 Honorary Professor of International Health, Sheffield Hallam University, Sheffield, UK Pro-mandatory vaccination. o Does mandatory vaccination apply to only “frontline care workers”? Should mandatory vaccination be introduced in the NHS, it should include all NHS workers (wider team) who share responsibility for care (nurses, doctors, leaders, managers, human resources, administration etc.) – an issue of equity o At the time of this letter, compulsory vaccination introduced for care home staff in England – unclear as to whether only “frontline staff” o Mandatory vaccination was being considered for NHS workers – unclear as to whether only “frontline staff” o Shared common obligation for all who work in the NHS and care homes to be fully vaccinated to defeat COVID-19, shared burden and potential morale boost Quoting Negash Ali, “You have a deeper connection with people who you have shared experiences with and shared pain.” 12 Gur-Arie, 2021 USA Published in BMJ Global Health January 2021 PhD, Berman Institute of Bioethics, Johns Hopkins, Oxford-Johns Hopkins Global Infectious Disease Ethics Collective Yes, but only as a last resort o Argues least restrictive option that can achieve the most good should always be prioritized o That is if a less restrictive measure than mandatory vaccination for HCWs can work as good as that policy it should be used instead o Risk of harm to the patient should be factored in, but so should ability for PPE/other intervention alternatives to mandatory vaccine as well o As a matter of principle the most restrictive/intrusive policy/intervention should only be used as a last resort o Mandatory vaccination for HCWs fits those criteria o Only ethically justifiable if alternatives are explored and deemed unsatisfactory 13 Hayes, 2021 UK Published in BMJ 8 July 2021 Professor, Head of Law School, University of Kent Against mandatory vaccination o Mandating vaccination was seen as coercive and impinged on the civil liberties of its citizens (“Liberty of Non-Vaccination” principle in UK Law since 1898) o Vaccination was not a “panacea for safety” o Care home workers who are unvaccinated against COVID-19 face loss of job o By law, care homes can only permit vaccinated staff, visitors into the premises o Wales and Scotland have rejected mandatory vaccination for care home workers but uptake was high (85–96%) and “virtually all” care homes staff o Wales and Scotland have registration for care home workers which England does not have. This enabled Wales and Scotland to monitor care home workers, provide training and professional development o Civil liberty is a necessary element of public health Authors contend that increasing stricter sanitation and infection control measures in care homes and funding for PPEs, in addition to providing access to vaccination, training, paid leave for vaccination and decent wages are more effective measures Mandatory vaccination in care homes is “unnecessary, disproportionate and misguided” 14 Hughes, 2021 USA Published in: Current Medical Research and Opinion (commentary) April, 2021 Department of Psychiatry, College of Medicine and Life Sciences, University of Toledo, OH, USA Neutral–ethical analysis, although commentary suggests pro-mandatory vaccination may be necessary for healthcare workers o Addresses questions of fair prioritisation and sub-prioritisation among various groups Ethical considerations for mandatory COVID-19 vaccination among healthcare staff o Challenges in fair prioritisation among vulnerable groups, like minorities and determining “essential workers” o Legally healthcare systems can impose mandatory licensed vaccination, with medical and religious exemptions. Precedent set with influenza vaccine. o Ethical justification: harm reductions including a covid infection, spread and death, duty to make healthcare environments safe vs risks o Autonomy of choice can be overridden; therefore voluntary vaccination should be desirable (determine reasons for vaccine hesitancy) Healthcare systems could partially mandate for higher exposure risk personnel and those who work with vulnerable populations 15 Kates, 2021 USA Published in: Open Forum Infectious Diseases (perspectives) March 2021 MD, Division of Allergy & Infectious Diseases, University of Washington, Seattle, Washington Vaccine and Infectious Diseases Division, Fred Hutchinson Cancer Research Centre Neutral–ethical analysis, although commentary suggests pro-mandatory vaccination may be necessary for healthcare workers Should health care institutions mandate SARS-CoV-2 vaccination for staff? o Healthcare workers prioritised for vaccination but hesitancy is limiting uptake o Hesitancy = novelty and mRNA-based mechanism, accelerated development, side-effects, sceptisim: low transmission and seriousness, safety and utility o Pro-mandates: favour balance of harms and benefits for individuals and communities (beneficence vs maleficence) plus moral imperative of health workers and organisations to provide a safe environment, reduce transmission and role model behaviour. Precedent set with influenza vaccine. o Against mandates: violation of personal autonomy, vaccine still ‘experimental’ (limited evidence), alternate strategies to prevent transmission, mandate may lead to redeployment and shortages, financial burdens and relationship breakdown between employer/employee. Minority groups may be more affected by mandates. o Interventions to promote vaccine acceptance (education, peer champions, modest incentives) Mandates may be ethical for health care workers in some circumstances when licensed (i.e. high risk populations) 16 Kevat, 2021 Australia Published in: The Medical Journal of Australia, pre-print 22 April 2021 MD, Endocrinologist Pro-mandatory vaccination. It also outlined restriction the types of law that can be made by the Federal Parliament, defined within the Australian Constitution. Public health legislation is primarily the responsibility of Territories and States governments. o Under the Australian law, is mandatory vaccination of healthcare workers permissible? o At the time of this paper publication, the aged care workforce had yet to be mandated to be vaccinated o Pay compensation secondary to effects of vaccination as a condition of employment has been considered by New South Wales government o Freedom of belief may undermine compulsory vaccination for healthcare workers; however, this could be argued as the least restrictive method to achieve public health outcome o Alternative approaches should be made available for HCPs who are unable or refuse vaccination o Considering discussions by the Fair Work Commission, dismissal for refusing vaccination is unlikely to be considered unreasonable, unless for medical reasons. o Healthcare workers are at high risk for COVID-19, as such, mandatory vaccination may be lawful and reasonable, except for medical exemptions. 17 Khunti, 2021 UK Published in: Journal of the Royal Society of Medicine (commentary) May 2021 CBE, FRCGP, FRCP, MD, PhD, FMedSci University of Leicester, UK Chair of the SAGE Ethnicity sub-panel, Professor of Primary Care Diabetes & Vascular Medicine Against mandatory COVID-19 vaccination, that is, to avoid mandate, until all concerns of healthcare workers were addressed. o Issue of concern raised in this commentary: low uptake of COVID-19 vaccine among healthcare workers in UK and USA due to perceived vaccine hesitancy o Reasons for vaccine hesitancy: a) skepticism; b) low perceived risk; c) fears of adverse effects; d) cultural and religious beliefs; and e) mistrust of healthcare system. o Reasons for mandatory vaccination considerations also discussed o More acceptable alternatives than mandatory vaccination: o Perceived support of family and friends was associated with increased vaccine uptake o Supportive workplace policies (example, onsite vaccination services), reduced barriers to vaccination uptake o Mandatory COVID-19 vaccination can be viewed as discriminatory, leading to stigmatization, further eroding of trust and widening inequalities 18 Klompas, 2021 USA Published in: Annals of Internal Medicine Ideas and opinions September 2021 MD, MPH, Department of Population Medicine, Harvard Medical School & Brigham and Women's Hospital, Boston, USA For mandatory COVID-19 vaccination of healthcare workers. o Justifications of why HCWs should be vaccinated o Mandatory vaccination is a common policy for healthcare workers, example mandating influenza vaccination o Morbidity and mortality of COVID-19 far exceed influenza o COVID-19 risks the lives of essential workers o Hospitals are a common site for SARS-CoV-2 transmission o Healthcare workers should protect their patients, and the onus is also on them o COVID-19 vaccines are more effective than influenza vaccines o SARS-CoV-2 is more disruptive to workforce continuity and hospital operations compared to influenza o SARS-CoV-2 vaccines are safe o Reluctance of many organisation to mandate this vaccine while under emergency use authorization, perhaps concerned about legal challenges o With full approval by the US Food and Drug Administration, it was argued that adopting of mandatory vaccination policies should ensue. 19 Leask et al 2021 Australia Published in: Medical Journal of Australia 13 September 2021 Professor of Nursing University of Sydney Neutral - may be justified for HCWs. The mandate for vaccination for HCWs may be justified depending on the situation where employees are at high risk of getting COVID or infecting vulnerable groups of people as part of the work o There should be a pre-requisite for mandatory vaccines, these requirements should consider the following: legal mandate at state level; burden of disease; availability of safe vaccines; transmission reduction due to vaccination; vaccine supply and accessibility. o For mandatory vaccination, there were also procedural recommendations such as:Supporting those who might not be able to access vaccination or removing all barriers, such as providing on-site workplace vaccination and allowing staff to ask questions about vaccines (addressing health literacy issues). o Support plan for the mandate (i.e. allowing medical exemptions and policies in consultation with the concern groups (HCWs and peak bodies) o HCWs should be supported in the mandating process and mandatory vaccines for HCWs should be considered in the right context. 20 Mittelman, 2021 USA Published in: BMJ August 2021 Patient with rare kidney disease three times post kidney transplant patient Patient advocate, patient editor at The BMJ, a global PCORI (Patient-Centered Outcomes Research Institute) ambassador; patient adviser Pro Vaccine should be mandated for health staff o Patients especially those who are immunocompromised must be able to trust their healthcare and support workers; to be protected “by and from staff” o Employment in the healthcare sectors should be conditional on having the vaccine. Soft mandates like weekly testing in lieu of vaccination mandate are not as effective. o Healthcare workers must support staff to have vaccination through paid time off o Exemptions should only be considered for those who are medically ineligible. o Those who are medically ineligible to have vaccination should not be assigned to direct patient care especially those who are at high risk o Views of patient regarding mandatory vaccine for healthcare workers 21 Osbourne, 2021 UK Published in the British Journal of Nursing January 2021 Third year Nursing student, De Monfort University, Leicester, UK Ambivalent/neutral Mandatory vaccine for HCWs may be beneficial for patients and colleagues in the clinical setting but unethical o Explored ethical framework of mandatory vaccine for nurses and other HCWs, through the lens of beneficence, non-maleficence, autonomy and justice o Beneficence – evidence provided of the benefits of vaccination is not significant enough to override the right to choose o Non-maleficence – mandatory vaccination deprives HCWs' right to consent and right to refuse treatment, may cause anxiety and stress o Autonomy – right to decide based on evidence and knowledge o Justice – right to make personal decisions without interference from the government; however, as mandated by law, right to choose can be restricted by government in the interest of public health concerns o Opting out of vaccination would restore autonomy of HCWs making them feel empowered o Looking at the issue through the lens of beneficence, non-maleficence, autonomy and justice oversimplifies the ethical dilemma o Utilitarianism or “the greatest good for the greatest number of people” may be relevant to mandatory vaccine however because vaccine is largely experimental, it may cause harm to HCWs when they are most needed o HCWs, particularly nurses are influential in the decision to vaccinate and therefore appropriate and accurate evidence-based information must be provided o High uptake of COVID 19 vaccination among HCWs is essential for a mass COVID19 vaccination to be successful 22 Palmer, 2021 United Kingdom Published in: British Journal of Healthcare Assistants August 2021 RN, freelance writer Neutral Argued that a key reason for resistance against mandatory COVID-19 vaccination could be that healthcare professionals were feeling resentful and mistrust the Government, and perhaps, a pay increase of larger than 1% may help. o Commentary on the Bradfield paper, of proposed intervention ladder for COVID-19 vaccination policies for healthcare workers o Also a commentary on the Royal College of Nursing's position on mandatory vaccination, and the UK Government decision o Key point of this policy: unvaccinated healthcare workers who are not exempted due to medical reason would be temporarily redeployed (and eventually suspended) if continued to be unvaccinated, to reduce risk to patients and colleagues o Overly coercive regulation undermined the goodwill of frontline healthcare workers, and foster resistance, resentment, and mistrust o Therefore, proposing principles of least restrictive alternative, the intervention ladder for COVID-19 vaccination policies for healthcare workers [Bradfield and Giubilini, 2021] paper:o Education campaigns or professional development activities to encourage vaccine uptake of frontline healthcare workers o Unvaccinated healthcare workers to sign statement to explain why refusal, and low rates of vaccination groups were highlighted o Restriction of employee privileges to unvaccinated employees o Redeployment to non-clinical duties, or working from home o Suspension of employment, conditions imposed on professional registration o Fines or imprisonment, termination of employment and cancellation of professional registration o Forcible vaccination using chemical or physical restraint, if required o View from UK Royal College of Nursing: No mandatory vaccination, but increase vaccination access, promote discussion of concerns in a supportive environment o Government decision: Introduction of mandatory vaccination in care homes on 16 June 2021 23 Parker, 2021 UK Published in: BMJ August 2021 B.Ed (Hons), MA, PhD Ethox Centre, Nuffield Department of Population Health, University of Oxford One author for mandatory vaccination: Parker:o Vaccine should be mandated for health and care staff Three authors (Bedford, Ussher and Stead) againsto Mandatory vaccine is unnecessary and counterproductive For mandatory vaccination: o From a moral standpoint, vaccination poses low risk for HCWs, if their unvaccinated status poses a risk to patients then they are obliged to accept vaccination. o “Duty of easy rescue” COVID-19 vaccine has been shown to have positive impact on patient safety and has low risk for adverse effects which establish the duty of healthcare professionals to protect patients Against mandatory vaccination o Mandating vaccination impinge on ethical issues of freedom of choice o Mandating vaccine is not necessary and not effective in increasing vaccination uptake o Vaccine hesitancy should be addressed to improve vaccine uptake (improving access, providing evidence-based information, being open and non-judgmental to questions and concerns) o Mandating vaccines increases resistance and mistrust o Some care workers prefer to leave the job rather than support mandatory vaccination which will cause staff shortages o Mandating vaccine negatively affects the morale of health staff who are already pressured during the pandemic response. o Unvaccinated employee be redeployed to areas where they pose no threat to patient safety. Risk assessment should be undertaken o Current vaccination rate of healthcare staff in the UK is already high (90% among NHS staff and 87% for staff in older adult care homes). Introducing coercive policies will be counterproductive o Duty of care vs ethics of freedom of choice 24 Shemtob et al., 2021 UK Published in BMJ; Policymakers and Healthcare Workers Editorial August 2021 PhD; Department of Primary Care and Public Health, School of Public Health, Imperial College London, London, UK Against o Risk assessment and redeployment of unvaccinated workers could adequately mitigate risk o UK doesn't have history of mandatory vaccination requirement to work like other countries o Policy could lead to a staffing shortage o Policy could disproportionately impact minority communities and workers o Policy could increase distrust toward government and institutions o Assessing the risk unvaccinated workers pose to patient population and redeploying workers to roles that aren't patient facing is a better solution to unvaccinated HCWs than a mandatory policy o Well supported argument against mandatory vaccination policy 25 Stokel-Walker, 2021 UK Published in: BMJ Feature June 2021 Freelance journalist News and features journalist, The Times & Sunday Times, The Economist, Bloomberg, the BBC and WIRED, specialising in digital culture and YouTube. Not stated Key points: o Examined decisions on mandatory vaccination globally o Discussed causes of vaccine hesitancy in different parts of the world, including religious reasons, lack of trust in the government and in the vaccine and manufacturers. o The threat of loss of income can both be an incentive to take action and a disincentive. o Incentive can be in the form of paid leave for those who need to self-isolate; and disincentive for those who feel threatened with job loss should they test positive o Pressure from management may reduce likelihood of getting vaccinated o HCWs need to be allowed a forum to ask questions in a safe supportive environment o Accurate information should be provided to those who are vaccine hesitant for religious reasons o One reason for vaccine hesitancy can be a lack of trust in vaccine manufacturers o Vaccine hesitancy will go down with time as more gets vaccinated 26 Talbot, 2021 USA Published in: JAMA Viewpoint June 2021 MD, MPH Professor of Medicine, Dept of Medicine, Division of Infectious Diseases, Chief Hospital Epidemiologist, Vanderbilt University Medical Center Pro-mandatory vaccine for HCPs Key points for mandating vaccination: o HCPs higher risk of infection many with severe outcomes but mitigated by use of PPEs o Asymptomatic infection is high among HCPs, high risk of transmission especially to immunocompromised and vulnerable patients o COVID-19 vaccines safe and highly effective, reduced asymptomatic infection and transmission Consideration in mandating COVID vaccination: o Unclear whether vaccines approved by Emergency Use Authorization (EUA) can legally be mandated o Allowances should be made for individuals who cannot be vaccinated o Alternative approaches should be made available for HCPs who are unable or refuse vaccination o HCPs should not inadvertently spread contagious infections to their patients and other HCPs 27 Visagie, 2021 USA Published in: Belitung Nursing Journal (letter to editor) August 2021 DNP, MSN, PHN, RN Executive Director -The Doctors for Global Health Institute Pro-mandatory vaccination. o Healthcare professionals have an ethical responsibility to health and well-being of society and the vulnerable in the community o Vaccination is integral to duty of care and code of conduct (particularly nurses) o Health care professional organisations are calling for employers to mandate vaccination among workforce (medical, nursing, pharmacy, public health, home care, hospice etc.) o Ethical responsibility among health care workers to abide by code and uphold ethical principles and maintain community health and safety o All healthcare professionals have an ethical responsibility to be vaccinated in the fight against COVID-19. 28 White, et al 2021 USA Published in JACC: Cardiovascular Interventions Editorial September 2021 White, C. M.D Professor of Medicine & Chairman of Cardiology, Ochsner Medical Centre, new Orleans, Louisiana. Editor JACC, Cardiology Interventions For mandatory vaccine for HCWs o Using the Prospect Theory by Tversky and Kahneman, people assess their potential to lose or gain, asymmetrically with a focus on aversion of loss. o People who are vaccine hesitant magnify the risks of vaccine side effects and minimise the vaccine benefit (belief that their risk or consequence of infection is less than what it truly is) o A novel approach considered by hospitals is reframing loss aversion of those who are vaccine hesitant by increasing burden for the unvaccinated. o For example, for the unvaccinated, mandating discomfort of measures to reduce spread such as N95 mask instead of surgical masks for those who are vaccinated will magnify the benefits of vaccination o HCWs have a clear duty to do no harm to patient. Potential of an asymptomatic COVID-19-infected caregiver transmitting the infection to vulnerable patients and co-workers have been documented and therefore mandating vaccination is an obvious step to control the pandemic. o Increasing the burden of being unvaccinated for HCWs (for example weekly testing, using N95 and extra protection will increase perception of benefits of vaccination and reduce loss aversion (loss of freedom or autonomy). 2.1 Data synthesis and analysis Data relevant to the research question were extracted (DM, MG, YS) which included authors and their qualifications, journal, source country, date published and position on the issue of mandatory vaccines including the relevant points considered by the authors. Content analyses using an inductive approach were carried out and discussed iteratively by all the authors until consensus was achieved. 3 Overview of the included articles The search identified 1413 articles from four databases (PsychInfo, CINAHL, Medline and Scopus). After removal of duplicates, 870 titles and abstracts were screened independently by two authors (DM and MG) of which 53 full text articles were examined further. Of these, 21 articles were considered eligible for inclusion with a further seven articles added from forward and backward searches of included articles, yielding a total of 28 articles for the review. Of the 28 included articles, only one was written in 2020 (Bowen, 2020), six (21 %) were published in the first four months (January to April) of 2021 (Osbourne and Clark, 2021; Flood et al., 2021; Hughes et al., 2021; Gur-Arie et al., 2021; Kates et al., 2021; Kevat et al., 2021), 18 (64 %) in the next four months (May to August 2021) (Ayukekbong, 2021; Visagie, 2021; Talbot, 2021; Stokel-Walker, 2021; Parker et al., 2021; Palmer, 2021; Mittelman, 2021; Shemtob et al., 2021; Khunti et al., 2021; Hayes and Pollock, 2021; Green, 2021; Klompas et al., 2021; Glasper, 2021b; Glasper, 2021a; Emanuel and Skorton, 2021; Bradfield and Giubilini, 2021; Dean, 2021; Frati et al., 2021) and a further three (11 %) (Baker and Blakely, 2021; Leask et al., 2021; White et al., 2021) from September to November 2021 when the search was completed. Articles were published as discussion papers, brief summaries, viewpoints or feature articles (64 %), editorials (25 %) or letters to the editor (11 %). Authors of these articles were from five countries, mostly from the USA (n = 11) (Baker and Blakely, 2021; Bowen, 2020; White et al., 2021; Visagie, 2021; Talbot, 2021; Mittelman, 2021; Klompas et al., 2021; Kates et al., 2021; Hughes et al., 2021; Gur-Arie et al., 2021; Emanuel and Skorton, 2021); and the UK (n = 11) (Dean, 2021; Stokel-Walker, 2021; Parker et al., 2021; Osbourne and Clark, 2021; Shemtob et al., 2021; Khunti et al., 2021; Green, 2021; Glasper, 2021b; Glasper, 2021a; Hayes and Pollock, 2021; Palmer, 2021); with three from Australia (Bradfield and Giubilini, 2021; Leask et al., 2021; Kevat et al., 2021); two from Canada (Ayukekbong, 2021; Flood et al., 2021), and one from Italy (Frati et al., 2021). The lead author of 20 articles (71 %) had health-related credentials (doctors, nurses or allied health specialty), three (11 %) were freelance journalists, two (7 %) had qualifications in law, and two (7 %) had expertise in medical ethics and interestingly, one (4 %) was a patient advocate. 4 Pro-mandatory, neutral and anti-mandatory vaccination for HCWs Of the 28 articles included, 12 (43 %) (Ayukekbong, 2021; Bradfield and Giubilini, 2021; Emanuel and Skorton, 2021; Frati et al., 2021; Green, 2021; Gur-Arie et al., 2021; Kevat et al., 2021; Klompas et al., 2021; Mittelman, 2021; Talbot, 2021; Visagie, 2021; White et al., 2021) took a pro-mandatory vaccination stance, 13 (46 %) (Baker and Blakely, 2021; Dean, 2021; Flood et al., 2021; Glasper, 2021b; Glasper, 2021a; Hughes et al., 2021; Kates et al., 2021; Leask et al., 2021; Palmer, 2021; Parker et al., 2021; Stokel-Walker, 2021; Osbourne and Clark, 2021; Bowen, 2020) were neutral or presented both sides of the debate, and three (11 %) were against (Hayes and Pollock, 2021; Khunti et al., 2021; Shemtob et al., 2021). Seven (58 %) of the twelve authors who were in favour of mandatory vaccination were from the USA (Emanuel and Skorton, 2021; Gur-Arie et al., 2021; Klompas et al., 2021; Mittelman, 2021; Talbot, 2021; Visagie, 2021; White et al., 2021) whilst a similar number, 7 of 13 (54 %) of those who were neutral or presented both sides of the issue were from the UK (Dean, 2021; Glasper, 2021b; Glasper, 2021a; Palmer, 2021; Parker et al., 2021; Stokel-Walker, 2021; Osbourne and Clark, 2021). All authors who were against mandatory vaccination for HCWs were from the UK (Hayes and Pollock, 2021, Khunti et al., 2021, Shemtob et al., 2021). The only patient advocate (Mittelman, 2021) was from the USA who supported vaccination mandate for HCWs. Among those with medical or health related credentials, nine of 12 authors (75 %) were pro-vaccination mandate (Ayukekbong, 2021; Bradfield and Giubilini, 2021; Emanuel and Skorton, 2021; Frati et al., 2021; Green, 2021; Gur-Arie et al., 2021; Kevat et al., 2021; Klompas et al., 2021; Mittelman, 2021; Talbot, 2021; Visagie, 2021; White et al., 2021), eight (67 %) were neutral (Baker and Blakely, 2021; Glasper, 2021b; Glasper, 2021a; Kates et al., 2021; Leask et al., 2021; Osbourne and Clark, 2021; Bowen, 2020) and two (17 %) were against (Khunti et al., 2021; Shemtob et al., 2021). The two authors with law qualifications were either neutral (Flood et al., 2021) or against (Hayes and Pollock, 2021) mandatory vaccination. Two of those with medical ethics credentials were pro-mandatory vaccination (Bradfield and Giubilini, 2021; Emanuel and Skorton, 2021) and one was neutral (Parker et al., 2021). A summary of the characteristics of these articles is in Table 3 .Table 3 Characteristics of pro-, neutral and anti- mandatory vaccination articles for HCWs. Table 3Parameters Pro (n = 12) Neutral (n = 13) Anti (n = 3) Country ◦ Australia 2 (17 %) 1 (8 %) 0 ◦ Canada 1 (8 %) 1 (8 %) 0 ◦ Italy 1 (8 %) 0 0 ◦ United Kingdom 1 (8 %) 7 (54 %) 3 (100 %) ◦ United States 7 (58 %) 4 (31 %) 0 Credentials ◦ Medical/health-related 9 (75 %) 8 (62 %) 2 (67 %) ◦ Medical ethics 2 (17 %) 1 (8 %) 0 ◦ Health writer/journalist 0 3 (23 %) 0 ◦ Legal/law 0 1 (8 %) 1 (33 %) ◦ Patient advocate 1 (8 %) 0 0 Date of publication ◦ Mid-year, 2020 (n = 1) 0 1 (8 %) 0 ◦ Jan–April 2021 (n = 6) 4 (33 %) 2 (15 %) 0 ◦ May–August 2021 (n = 18) 9 (75 %) 6 (46 %) 3 (100 %) ◦ Sep–Dec 2021 (n = 3) 1 (8 %) 2 (15 %) 0 4.1 Pro-mandatory vaccination: the arguments A number of authors cited ethical and legal arguments in favour of mandatory vaccination, including those with neutral views. The three ethical principles of justice, beneficence and non-maleficence were discussed as the primary justification for mandating vaccination for HCWs. 4.1.1 Ethical considerations (justice, beneficence and non-maleficence) Arguments in favour of mandatory vaccination for healthcare workers prioritised two key professional responsibilities based on fiduciary duty: to uphold the bioethical and utilitarian principles of beneficence and non-maleficence. Many authors spoke of the expectation HCWs had to act in the best interests of the patient (beneficence) and their duty within codes of conduct to minimise the risk on the public (Ayukekbong, 2021; Emanuel and Skorton, 2021; Klompas et al., 2021; Parker et al., 2021; Visagie, 2021; Osbourne and Clark, 2021) and themselves (Frati et al., 2021; Klompas et al., 2021) through vaccination. Both Emanuel and Skorton (2021) and Parker et al. (2021) added that vaccination had a limited burden of risk to HCWs themselves hence there was a moral imperative. Others discussed making judgements based on vaccine efficacy (Bowen, 2020) and implementing least restrictive measures first but did not rule out mandatory vaccination (Gur-Arie et al., 2021; Kevat et al., 2021), although alternative options should be offered to those who refuse or had medical exemptions (Kevat et al., 2021; Mittelman, 2021; Talbot, 2021) including leave or redeployment to non-clinical roles or areas (Bradfield and Giubilini, 2021; Mittelman, 2021) prior to termination (Bradfield and Giubilini, 2021). Similarly, the importance of not doing harm (non-maleficence) and weighing the risks against benefits was reported by authors in favour of mandatory vaccination. Without mandates, patients were seen to be at risk of harm (Bradfield and Giubilini, 2021). Mittelman (2021), the patient advocate, explained that HCWs had an obligation to protect people who were vulnerable and immunocompromised from any harm and their employment should be conditional on vaccination. HCWs might be asymptomatic but were seen to be vectors for high transmission to people who were vulnerable so should be vaccinated to fulfil the duty of ‘do no harm’ (Talbot, 2021; White et al., 2021). Furthermore, Hughes et al. (2021), when arguing for mandatory vaccination, stated that healthcare facilities and HCWs should be held accountable in providing a safe environment for patient care. Overall the benefits were seen to outweigh the risks, and the vaccine was considered more effective than influenza vaccines (Klompas et al., 2021) and therefore generally safe and effective for HCWs (Talbot, 2021). Three authors discussed justice and fair and equitable distribution of risks and benefits (Bowen, 2020; Green, 2021; Osbourne and Clark, 2021), with Green (2021) maintaining that all staff in the UK (not only frontline) in the National Health Service (NHS) and care homes should be vaccinated for equity. 4.1.2 Legal rights of patients vs legal rights of HCWs (right to autonomy) Legislation regarding the legal rights of HCWs and the occupational health and safety legislation governing employers in different countries varied. For example, Canadian legislation required employers to ensure safety and protect employees from occupational hazards (Ayukekbong, 2021) which placed mandatory vaccination as fulfilling this decree. The legality of the employers' requirement for vaccination from their employees must be evaluated based on “reasonableness” and challenged under the Canadian Charter of Rights and Freedom specifically for employees with medical or cultural reasons to refuse vaccination (Flood et al., 2021). In Australia, the Biosecurity Act 2015, required employers to implement appropriate measures to prevent and reduce the spread of diseases which could include mandating effective vaccination (Kevat et al., 2021). The Amendments to the Vaccination Act of 1898 and 1907 in the UK, in the wake of protests against compulsory smallpox vaccination, legally recognised the rights of those who were “honestly opposed to vaccination” (Hayes and Pollock, 2021). Whilst mandating vaccination was seen as a violation of the rights of choice and autonomy of HCWs, caregiving that might potentially put patients at risk, was also seen as a violation of the patient's rights to safe care (Ayukekbong, 2021). Flood et al. (2021), on the other hand, discussed a case ruling which found that mandatory policy encroached upon the rights of a person, stating that “forced medical treatment (flu vaccination in this case) is an assault if there is no consent” (p. E219). In the United States, Baker and Blakely (2021) argued that whilst private employers might enforce mandatory vaccinations, legal exemptions should be granted on religious and disability-related reasons, however, should unvaccinated employees threaten the health of others, then termination could be considered. On the other hand, Klompas et al. (2021) cited the efficacy of the vaccine and impending approval (at the time of review) of the United States Food and Drug Administration and precedent US court ruling in favour of healthcare organisations, as the legal basis for mandating COVID-19 vaccination for healthcare workers. 5 Anti-mandatory vaccination: the arguments Three authors expressed views against mandatory vaccines for HCWs (Hayes and Pollock, 2021; Khunti et al., 2021; Shemtob et al., 2021). Authors who had neutral views also presented negative sides of the argument but balanced it with a discussion on the perceived benefits. The anti-mandatory vaccination arguments centred on the ethics of mandating vaccination, contending that: i) the benefits of vaccination were not significant enough to override the right of healthcare workers to choose; ii) mandating vaccination was discriminatory and might cause stigmatisation, isolation and mistrust in those who refused to be vaccinated; and iii) mandating vaccination deprived healthcare workers of the right of choice (Osbourne and Clark, 2021; Shemtob et al., 2021; Khunti et al., 2021; Kates et al., 2021; Gur-Arie et al., 2021). Hayes & Pollock (2021) believed that mandating vaccination for HCW was unnecessary as the vaccination rates among HCW in the US were already high. Further, like other medical interventions, vaccination should be offered as a choice, providing recipients with accurate information on which to base their decision (Kates et al., 2021; Shemtob et al., 2021). Mandatory vaccination was argued to take away HCWs' right to choose treatment preferences including vaccination (Hughes et al., 2021). Parker et al. (2021) and Palmer (2021) expressed the opinions that mandating vaccination for HCWs was an easy solution to a more complex issue of vaccine hesitancy and did not adequately address the underlying problem. Other strategies were proposed by some authors (Parker et al., 2021; Palmer, 2021; Leask et al., 2021; Kates et al., 2021; Gur-Arie et al., 2021) including use of masks, stringent testing and improving policies in relation to a stepwise approach to vaccination strategies, including redeployment to non-clinical areas. Valid and legitimate reasons for vaccination exemptions, such as medical conditions and religious beliefs, were also discussed (Ayukekbong, 2021; Baker and Blakely, 2021). Medical conditions included those conditions in which the person believed that their risk of serious illness or death would be increased should they receive the vaccination (Baker and Blakely, 2021; Mittelman, 2021). Therefore, based on their non-vaccinated status, some authors proposed that these HCWs should not be assigned to provide care to patients who are at high risk (Kates et al., 2021; Shemtob et al., 2021). Whilst this could be an option for health services to explore, a concern raised was redeployment of essential non-vaccinated care workers to non-clinical roles will further decimate the HCW workforce in critical areas of healthcare such as aged care and emergency (Glasper, 2021b; Kevat et al., 2021; Parker et al., 2021; Shemtob et al., 2021). In addition, pressure from employers for staff to get vaccinated may push essential health staff to leave employment or be dismissed (Palmer, 2021; Parker et al., 2021). 6 Recommendations from articles reviewed Authors of the articles reviewed discussed four core recommendations. First, public health authorities should consider less intrusive public health strategies before implementing a mandatory vaccination policy (Bradfield and Giubilini, 2021; Palmer, 2021), including stricter monitoring of infection control measures, adequate supplies of personal protective equipment and resources and appropriate COVID-19 testing. Secondly, mandatory vaccination policies should be implemented equitably, consistent with broader public health safety measures (Flood et al., 2021; Hughes et al., 2021; Kates et al., 2021; Klompas et al., 2021), ensuring that education on vaccine safety was provided as well as, timely and transparent information on vaccine effectiveness. Thirdly, these policies should clearly communicate any medical-based or religious exemptions (Baker and Blakely, 2021; Kevat et al., 2021; Leask et al., 2021), reasons for redeployment to non-clinical roles (Palmer, 2021; Parker et al., 2021; Shemtob et al., 2021) and compensation to healthcare staff who developed adverse symptoms post COVID-19 vaccination (Bowen, 2020; Flood et al., 2021). Lastly, logistical and financial barriers to vaccination must not be overlooked, and any mandatory vaccination policy should be accompanied by support for HCWs to get paid time off to receive the vaccination and to have access to on-site vaccination options (Mittelman, 2021; Hayes and Pollock, 2021; Leask et al., 2021; Khunti et al., 2021). 7 To mandate or not to mandate This review has identified that there are polarised opinions on mandatory vaccination for HCWs that centre on weighing the risks versus benefits, ethical and legal responsibilities as well as rights for personal autonomy. Those who expressed views in favour of mandatory vaccination for HCWs primarily cited ethical and legal rights and duties to protect themselves and the public, justifying the safety profile of the vaccine. There is evidence that COVID-19 is serious, with high morbidity and mortality rates (Piroth et al., 2021; Zylke and Bauchner, 2020; Chang et al., 2022), notwithstanding the latest reports of long-covid symptoms (Evans et al., 2022; Brown et al., 2022). Yet understandably there will be ‘vaccine refusal’ and ‘vaccine hesitancy’ (Wiysonge et al., 2022; Al-Amer et al., 2022), a term defined by healthcare professionals, journalists, and policymakers, as the research evidence is limited with this being a relatively new viral infection with novel vaccines (Karafillakis et al., 2022). Furthermore, those presenting arguments against a mandate emphasised that mandatory vaccination for HCWs impinged on and violated the human right of personal autonomy, culminating in fear and mistrust of health authorities, stigma, and isolation. Making legislation and laws more stringent does not address the issue of hesitancy (Drew, 2019), although education and authoritative and reliable information can make an impact (Al-Amer et al., 2022). The complexity for healthcare workers is that they are often duty-bound in their professions to uphold bioethical and utilitarian principles as part of professional and ethical codes of practice (Brenna and Das, 2021). For instance, the Hippocratic oath crafted 2500 years ago sets out the historical and philosophical underpinning of the role of the medical profession, in that medicine should be practised to benefit the sick and protect patients against harm (Hajar, 2017). A contemporary companion to this oath is the World Medical Association's Declaration of Geneva which outlines the values and principles of being a doctor and includes statements, among many others, such as: “I solemnly pledge to consecrate my life to the service of humanity; will practise my profession with conscience and dignity; The health of my patient will be my first consideration” (World Medical Association, 2018). Similarly the nursing profession is bound by codes and professional standards with the founder, Florence Nightingale declaring the first overriding principle to “do the sick no harm” (Nightingale, 1863). Whilst healthcare professionals are bound by a duty to these principles, the dilemma has been that professional organisations have not voiced clear and consistent messages on their position on this issue. The expectations for those who work in these professions and the health industry have also fluctuated since the emergence of COVID-19. Regular handwashing and donning personal protective equipment were mitigation measures implemented in the early stages to limit the spread of the virus (McCarthy et al., 2020), yet as new evidence has emerged and cases have increased, there has been a push for redeployment, reduced hours and even unemployment (job losses) for the unvaccinated; “no jab, no job” (Tobin, 2021). Professional bodies for healthcare workers play a role in ensuring that members are adequately supported in making decisions whilst tackling compliance to mandatory vaccination as a condition of their employment, albeit professional organisations have had their own challenges in taking a clear and united stance on the issue of mandatory vaccination. For example, the American Nurses Association (American Nurses Association, 2021) and Australian College of Nursing (Australian College of Nursing, 2021) in their position statement strongly recommended and supported the mandate that all registered nurses must be vaccinated against COVID-19. However, in the United Kingdom, the Royal College of Nursing promotes that nurses should be vaccinated to protect themselves and their patients but was cautious about supporting mandatory vaccination in a view that nurses are a diverse group, which might cause division rather than increased uptake of vaccines among healthcare workers (Royal College of Nursing, 2022a; Royal College of Nursing, 2022b). More recently the International Council of Nurses released a statement highlighting nurses' professional responsibility in making decisions to be vaccinated (Australian College of Nursing, 2021). It may be the case therefore that the variations in opinions on mandatory vaccination by professional bodies representing healthcare professions, may have some influence on the current and ongoing divergent discourse related to vaccine mandates. Nevertheless, the COVID-19 pandemic itself has been the catalyst for a flow on effect that has impacted healthcare workers and healthcare systems. There has undoubtedly been a financial impact to services with workforce shortages due to increased sick leave and quarantine of staff during periods of the pandemic (Newham and Hewison, 2021), with concerns for redeployment and redundancy adding further pressure on a depleted and exhausted workforce (Royal College of Nursing, 2022b). The consequent workforce shortages have overburdened health systems and staff during the pandemic, with high bed occupancy rates, increased need for intensive care medical units and lack of personal protective equipment (Sen-Crowe et al., 2021) leading to high levels of HCW burnout, fatigue and stress (Tracy et al., 2020; Yamane et al., 2022) which all ultimately impact on the quality of patient care. Despite the diverse opinions and differing standpoints on vaccination noted in this review, a considered and thoughtful review of policies, focusing on prioritising the best interests of healthcare workers and the general public whilst maintaining some flexibility is recommended. Noni MacDonald, a founding member of WHO's Global Advisory Committee on Vaccine Safety advises that “governments should frame their policy-making decisions around two questions: ‘What problem are you trying to fix? And is a mandate the way to fix it?’” (Drew, 2019). For that reason, whether opinion is for, against or undecided on a vaccine mandate, ultimately HCWs should consider employing means at all costs to protect their health and those in their care and continue to ‘do no harm’. 8 Strengths and limitations The opinions of professionals from different areas of specialty including a patient advocate were analysed and discussed. However, the continuing and dynamic waxing and waning of the COVID-19 pandemic, including the differences in vaccination policies and pandemic response measures undertaken by governments, posed challenges in synthesising opinions on mandatory vaccination for HCWs. It is important to note that the opinion articles published between January 2020 and November 2021 were reviewed in this discussion paper and considering the lag time between completion of the manuscript and publication of included articles, new data may have emerged which could have influenced the opinions of authors in this review. Restricting the search to articles published in English, whilst necessary due to funding constraints, is a serious limitation of the discussion on mandatory vaccination for HCWs considering that many non-English speaking countries have experienced a high number of cases extracting more critical demands from healthcare workers. It is recommended that a review of opinions in countries whose primary language is not English be undertaken to provide a deeper insight into other cultural dimensions that may have influenced opinions regarding mandating vaccination for HCWs, as a multi-country Asia-Pacific study (Chew et al., 2021) identified that more than 95 % of HCWs self-reported a willingness to vaccinate. A strength of the review was that opinions on mandatory vaccination were diverse, with opinions from health and legal professionals as well as a healthcare user. However, the fast and evolving science behind the COVID-19 pandemic and the shifting opinions of professionals in the field warrant ongoing review, to ensure that the pulse of public opinion is considered in health policy planning and implementation. 9 Conclusion Mandating COVID-19 vaccinations for healthcare workers is a complex issue. A careful and collaborative, participatory review by key stakeholders regarding mandatory vaccination policies is needed. This should ensure a balance between the rights of healthcare workers for autonomy and the rights of the public to safe, quality healthcare during a pandemic. Healthcare workers' acceptance or resistance to mandates can be influenced by policies that are based on solid scientific evidence but at the same time, these need to be flexible to consider ethical dilemmas and personal rights. As Hayes & Pollock (2021) argued, “vaccination is not a panacea for safety.” Further, professional bodies have a role to play in ensuring that members have adequate access to reliable information resources and support, as well as expert vaccination guidelines that they can refer to when required. Support for vulnerable HCWs is essential and should be part of a short- and long-term pandemic response. Funding No external funding. CRediT authorship contribution statement Della Maneze: Conceptualization, Methodology, Data curation, Validation, Formal analysis, Writing – original draft, Writing – review & editing, Supervision, Project administration. Yenna Salamonson: Conceptualization, Methodology, Data curation, Validation, Formal analysis, Writing – original draft, Writing – review & editing, Supervision. Maxwell Grollman: Conceptualization, Methodology, Data curation, Validation, Formal analysis, Writing – original draft, Writing – review & editing, Project administration. Jed Montayre: Conceptualization, Methodology, Data curation, Validation, Formal analysis, Writing – original draft, Writing – review & editing. Lucie Ramjan: Conceptualization, Methodology, Data curation, Validation, Formal analysis, Writing – original draft, Writing – review & editing, Supervision. Declaration of Competing Interest None. Appendix A Supplementary data The following are the supplementary data related to this article.Supplementary table Image 1 Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijnurstu.2022.104389. ==== Refs References Al-Amer R. Maneze D. Everett B. Montayre J. Villarosa A.R. Dwekat E. Salamonson Y. COVID-19 vaccination intention in the first year of the pandemic: a systematic review J. Clin. Nurs. 31 2022 62 86 10.1111/jocn.15951 34227179 American Library Association COVID-19 Coronavirus pandemic [online] Available 2022 American Library Association Geneva https://www.wma.net/what-we-do/medical-ethics/declaration-of-geneva/ American Nurses Association Immunizations [online] Available: 2021 American Nurses Association Maryland https://www.nursingworld.org/practice-policy/nursing-excellence/official-position-statements/id/immunizations/ Australian College of Nursing Australian College of Nursing supports the requirement for nurses to be vaccinated [Online] Available: 2021 Australian College of Nursing Deakin https://www.acn.edu.au/media-release/australian-college-of-nursing-supports-the-requirement-for-nurses-to-be-vaccinated Ayukekbong J. Public health and human rights during a pandemic: an unresolved dilemma concerning mandatory vaccination against COVID-19 for healthcare workers Can. J. Infect. Control 36 2021 74 76 Baker N. Blakely K. A brief summary on pitfalls, policy, and practice for a mandatory vaccination plan Alabama Nurse 48 3 2021 p15 Bowen R.A.R. Ethical and organizational considerations for mandatory COVID-19 vaccination of health care workers: a clinical laboratorian's perspective Clin. Chim. Acta 510 2020 421 422 10.1016/j.cca.2020.08.003 32771485 Bradfield O.M. Giubilini A. Spoonful of honey or a gallon of vinegar? A conditional COVID-19 vaccination policy for front-line healthcare workers J. Med. Ethics 47 2021 467 472 10.1136/medethics-2020-107175 33975928 Brenna C.T. Das S. The divided principle of justice: ethical decision-making at surge capacity Am. J. Bioeth. 21 2021 37 39 10.1080/15265161.2021.1940358 Brown K. Yahyouche A. Haroon S. Camaradou J. Turner G. Long COVID and self-management Lancet (London, England) 399 2022 355 10.1016/S0140-6736(21)02798-7 Chang D. Chang X. He Y. Tan K.J.K. The determinants of COVID-19 morbidity and mortality across countries Sci. Rep. 12 2022 1 17 10.1038/s41598-022-09783-9 34992227 Chew N.W. Cheong C. Kong G. Phua K. Ngiam J.N. Tan B.Y. Wang B. Hao F. Tan W. Han X. An Asia-Pacific study on healthcare workers’ perceptions of, and willingness to receive, the COVID-19 vaccination Int. J. Infect. Dis. 106 2021 52 60 10.1016/j.ijid.2021.03.069 33781902 Cole J.P. Swendiman K.S. Mandatory Vaccinations: Precedent and Current Laws. Current Politics and Economics of the United States, Canada and Mexico 17 2015 255 Davies N.G. Abbott S. Barnard R.C. Jarvis C.I. Kucharski A.J. Munday J.D. Pearson C.A. Russell T.W. Tully D.C. Washburne A.D. Estimated transmissibility and impact of SARS-CoV-2 lineage B. 1.1. 7 in England Science 2021 372 34437095 Dean E. Could vaccination become compulsory for nurses? COVID-19 has highlighted the issue of mandatory vaccination for healthcare professionals Nurs. Older People 2021 10 11 10.7748/nop.33.3.10.s4 Drew L. The case for mandatory vaccination Nature 575 2019 S58 10.1038/d41586-019-03642-w 31776503 Dyer O. Covid-19: unvaccinated face 11 times risk of death from delta variant, CDC data show Br. Med. J. 2021 374 10.1136/bmj.n2282 Dzinamarira T. Murewanhema G. Mhango M. Iradukunda P.G. Chitungo I. Mashora M. Makanda P. Atwine J. Chimene M. Mbunge E. Chingombe I. Masuka G. Nkambule S.J. Ngara B. COVID-19 prevalence among healthcare workers: a systematic review and meta-analysis Int. J. Environ. Res. Public Health 19 2021 146 10.3390/ijerph19010146 35010412 Emanuel E.J. Skorton D.J. Mandating COVID-19 vaccination for health care workers Ann. Intern. Med. 30 2021 30 10.7326/M21-3150 Esparza J. Nitsche A. Damaso C.R. Beyond the myths: novel findings for old paradigms in the history of the smallpox vaccine PLoS Pathog. 14 2018 e1007082 10.1371/journal.ppat.1007082 Evans R.A. Leavy O.C. Richardson M. Elneima O. Mccauley H. Shikotra A. Singapuri A. Sereno M. Saunders R.M. Harris V.C. Clinical characteristics with inflammation profiling of long COVID and association with 1-year recovery following hospitalisation in the UK: a prospective observational study Lancet Respir. Med. 2022 10.1016/S2213-2600(22)00127-8 Evered E.Ö. Evered K.T. Mandating immunity in the Ottoman Empire: a history of public health education and compulsory vaccination Heliyon 6 2020 e05488 10.1016/j.heliyon.2020.e05488 Flood C.M. Thomas B. Wilson K. Mandatory vaccination for health care workers: an analysis of law and policy Can. Med. Assoc. J. 193 2021 E217 E220 10.1503/cmaj.202755 33468521 Frati P. La Russa R. Di Fazio N. Del Fante Z. Delogu G. Fineschi V. Compulsory vaccination for healthcare workers in Italy for the prevention of SARS-CoV-2 infection Vaccines 9 2021 966 10.3390/vaccines9090966 34579203 Glasper A. Mandatory COVID-19 vaccination for frontline care home staff soon to become law Brit. J. Healthc. Assist. 15 2021 296 299 10.12968/bjon.2021.30.13.828 Glasper A. Should SARS-CoV-2 vaccination for all frontline healthcare staff be compulsory? Br. J. Nurs. 30 2021 828 829 10.12968/bjha.2021.15.6.296 34251852 Green S.T. Recognising the value of "shared pain"-mandatory covid-19 vaccination should include all NHS workers Br. Med. J. 374 2021 n1668 10.1136/bmj.n1668 Gur-Arie R. Jamrozik E. Kingori P. No jab, no job? Ethical issues in mandatory COVID-19 vaccination of healthcare personnel Br. Med. J. Glob. Health 6 2021 e004877 10.1136/bmjgh-2020-004877 Hajar R. The physician's oath: historical perspectives Heart Views 18 2017 154 10.4103/HEARTVIEWS.HEARTVIEWS_131_17 29326783 Havers F.P. Pham H. Taylor C.A. Whitaker M. Patel K. Anglin O. Kambhampati A.K. Milucky J. Zell E. Chai S.J. COVID-19-associated hospitalizations among vaccinated and unvaccinated adults ≥ 18 years–COVID-NET, 13 states, January 1–July 24, 2021 JAMA Intern. Med. 182 2022 1071 1081 10.1001/jamainternmed.2022.4299 36074486 Hayes L. Pollock A. Mandatory covid-19 vaccination for care home workers unnecessary, disproportionate, and misguided BMJ 374 2021 n1684 10.1136/bmj.n1684 Hughes K. Gogineni V. Lewis C. Deshpande A. Considerations for fair prioritization of COVID-19 vaccine and its mandate among healthcare personnel Curr. Med. Res. Opin. 37 2021 907 909 10.1080/03007995.2021.1908245 33760673 Karafillakis E. Van Damme P. Hendrickx G. Larson H.J. COVID-19 in Europe: new challenges for addressing vaccine hesitancy Lancet 399 2022 699 701 10.1016/S0140-6736(22)00150-7 35123665 Kates O.S. Diekema D.S. Blumberg E.A. Should health care institutions mandate SARS-CoV-2 vaccination for staff? Open Forum Infect. Dis. 8 2021 ofab155 10.1093/ofid/ofab155 Kevat D.A. Panaccio D.C. Pang S.C. Dean J.M. Farmer C.C. Mahar P.D. Medico-legal considerations of mandatory COVID-19 vaccination for high risk workers Med. J. Aust. 215 2021 22 24 10.5694/mja2.51128 e1 34117640 Khunti K. Kamal A. Pareek M. Griffiths A. Should vaccination for healthcare workers be mandatory? J. R. Soc. Med. 114 2021 235 236 10.1177/01410768211013525 34028294 Klompas M. Pearson M. Morris C. The case for mandating COVID-19 vaccines for health care workers Ann. Intern. Med. OC1 2021 10.7326/M21-2366 Leask J. Seale H. Williams J.H. Kaufman J. Wiley K. Mahimbo A. Clark K.K. Danchin M.H. Attwell K. Policy considerations for mandatory COVID-19 vaccination from the Collaboration on Social Science in Immunisation Med. J. Aust. 2021 10.5694/mja2.51269 Li Q. Guan X. Wu P. Wang X. Zhou L. Tong Y. Ren R. Leung K.S. Lau E.H. Wong J.Y. Early transmission dynamics in Wuhan, China, of novel coronavirus–infected pneumonia N. Engl. J. Med. 2020 10.1056/NEJMoa2001316 Mcarthur A. Klugárová J. Yan H. Florescu S. Innovations in the systematic review of text and opinion JBI Evid. Implement. 13 2015 188 195 10.1097/XEB.0000000000000060 Mccarthy R. Gino B. D'entremont P. Barari A. Renouf T.S. The importance of personal protective equipment design and donning and doffing technique in mitigating infectious disease spread: a technical report Cureus 12 2020 10.7759/cureus.12084 Mittelman M. Patient commentary: protect patients like me-make covid vaccines mandatory for all eligible staff in care settings BMJ 374 2021 n1921 10.1136/bmj.n1921 Moran A. Agaliotis M. Seale H. The views of key stakeholders around mandatory influenza vaccination of hospital and aged care staff: examining the current climate in Australia Vaccine 37 2019 705 710 10.1016/j.vaccine.2018.12.029 30626529 Newham R. Hewison A. Covid-19, ethical nursing management and codes of conduct: an analysis Nurs. Ethics 28 2021 82 90 10.1177/0969733020988316 33472524 Nightingale F. Notes on Nursing: What it is, and What it is Not 1863 Lippincott Williams & Wilkins London Osbourne R.M. Clark S.J. Should the SARS-CoV-2 vaccine be mandatory for nurses? An ethical debate Br. J. Nurs. 30 2021 116 121 10.12968/bjon.2021.30.2.116 33529104 Palmer S.J. Ethics of enforcing the vaccine on healthcare staff Brit. J. Healthc. Assist. 15 2021 354 358 10.12968/bjha.2021.15.7.354 Parker M. Bedford H. Ussher M. Stead M. Should covid vaccination be mandatory for health and care staff? BMJ 374 2021 n1903 10.1136/bmj.n1903 Piroth L. Cottenet J. Mariet A.-S. Bonniaud P. Blot M. Tubert-Bitter P. Quantin C. Comparison of the characteristics, morbidity, and mortality of COVID-19 and seasonal influenza: a nationwide, population-based retrospective cohort study Lancet Respir. Med. 9 2021 251 259 10.1016/S2213-2600(20)30527-0 33341155 Rickman H.M. Rampling T. Shaw K. Martinez-Garcia G. Hail L. Coen P. Shahmanesh M. Shin G.Y. Nastouli E. Houlihan C.F. Nosocomial transmission of coronavirus disease 2019: a retrospective study of 66 hospital-acquired cases in a London teaching hospital Clin. Infect. Dis. 72 2021 690 693 10.1093/cid/ciaa816 32562422 Royal College of Nursing COVID-19, mandatory vaccination and vaccination as a condition of employment [online] Available: 2022 Royal College of Nursing London https://www.rcn.org.uk/Get-Help/RCN-advice/covid-19-and-vaccination Royal College of Nursing RCN position on mandating vaccination for health and social care staff [Online] Available: 2022 Royal College of Nursing London https://www.rcn.org.uk/About-us/Our-Influencing-work/Position-statements/rcn-position-on-mandating-vaccination-for-health-and-social-care-staff Sen-Crowe B. Sutherland M. Mckenney M. Elkbuli A. A closer look into global hospital beds capacity and resource shortages during the COVID-19 pandemic J. Surg. Res. 260 2021 56 63 10.1016/j.jss.2020.11.062 33321393 Shemtob L. Ferris M. Asanati K. Majeed A. Vaccinating healthcare workers against covid-19 Br. Med. J. 374 2021 10.1136/bmj.n1975 Stewart A.J. Devlin P.M. The history of the smallpox vaccine J. Infect. 52 2006 329 334 10.1016/j.jinf.2005.07.021 16176833 Stokel-Walker C. Covid-19: the countries that have mandatory vaccination for health workers Br. Med. J. 373 2021 10.1136/bmj.n1645 Talbot T.R. COVID-19 vaccination of health care personnel as a condition of employment: a logical addition to institutional safety programs J. Am. Med. Assoc. 326 2021 23 24 10.1001/jama.2021.8901 Tobin A. No jab, no job: mandating workplace COVID-19 vaccinations in the current legal landscape [Online] Available: 2021 HG Lawyers Brisbane https://www.hopgoodganim.com.au/page/knowledge-centre/blog/no-jab-no-job-covid-workplace-vaccine-mandates-2021 Tracy D.K. Tarn M. Eldridge R. Cooke J. Calder J.D. Greenberg N. What should be done to support the mental health of healthcare staff treating COVID-19 patients? Br. J. Psychiatry 217 2020 537 539 10.1192/bjp.2020.109 32423523 Visagie N. The war on COVID-19 and vaccination mandates: ethical code of conduct Belitung Nurs. J. 2021 10.33546/bnj.1768 White C.J. Samady H. Moliterno D.J. The Case for Mandatory COVID-19 Vaccination of Health Care Workers 2021 American College of Cardiology Foundation Washington DC Wiysonge C.S. Ndwandwe D. Ryan J. Jaca A. Batouré O. Anya B.-P.M. Cooper S. Vaccine hesitancy in the era of COVID-19: could lessons from the past help in divining the future? Hum. Vaccin. Immunother. 18 2022 1 3 10.1080/21645515.2021.1893062 World Health Organization The Impact of COVID-19 on Health and Care Workers: A Closer Look at Deaths 2021 World Health Organization World Medical Association WMA declaration of Geneva [online]. France Available: https://www.wma.net/policies-post/wma-declaration-of-geneva/ 2018 [Accessed] Yamane D. Zarabian K. Devine K. Benjenk I. Farrar K. Park O.L. Kim J. Davison D. Heinz E. Hospital-based healthcare worker perceptions of personal risk related to COVID-19: one year follow-up J. Am. Board Fam. Med. 35 2022 284 294 10.3122/jabfm.2022.02.210272 35379716 Zylke J.W. Bauchner H. Mortality and morbidity: the measure of a pandemic J. Am. Med. Assoc. 324 2020 458 459 10.1001/jama.2020.11761	36462385	PMC9709452	NO-CC CODE	2022-12-01 23:23:06	no	Int J Nurs Stud. 2023 Feb 9; 138:104389	utf-8	Int J Nurs Stud	2,022	10.1016/j.ijnurstu.2022.104389	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)00144-8 10.1016/j.ijregi.2022.11.011 Article High anti-SARS-CoV-2 seroprevalence among unvaccinated mother-child pairs from a rural setting of north-eastern Tanzania during the second wave of COVID-19 Msemo Omari Abdul 1 Pérez-Alós Laura 2 Minja Daniel T.R. 1 Hansen Cecilie Bo 2 Gesase Samwel 1 Mtove George 1 Mbwana Joyce 1 Larsen Victoria Marie Linderod 2 Bøgestad Emilie Caroline Skuladottir 2 Grunnet Louise Groth 3 Christensen Dirk Lund 4 Bygbjerg Ib Christian 4 Burgner David 56 Schmiegelow Christentze 78 Garred Peter 2# Hjort Line 89#⁎ 1 National Institute for Medical Research, Tanga Center, Tanga, Tanzania 2 Laboratory of Molecular Medicine, Department of Clinical Immunology, Section 7631, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark 3 Steno Diabetes Center Copenhagen, Herlev, Denmark 4 Global Health Section, Department of Public Health, University of Copenhagen, Denmark 5 Murdoch Children´s Research Institute, Melbourne, Victoria, Australia 6 Department of Pediatrics, Melbourne University, Melbourne, Victoria, Australia 7 Centre for Medical Parasitology, Department of Immunology and Microbiology, University of Copenhagen and Department of Infectious Diseases, Copenhagen University Hospital, Denmark 8 Department of Obstetrics, Copenhagen University Hospital, Copenhagen, Denmark 9 Novo Nordisk Foundation Center for Basic Metabolic Research, Metabolic Epigenetics Group, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark ⁎ Corresponding author: Dr. Line Hjort, University of Copenhagen, Denmark AND Department of Obstetrics, Novo Nordisk Foundation Center for Basic Metabolic Research, Metabolic Epigenetics Group, Faculty of Health and Medical Sciences, N/A, Denmark # Shared last authorship 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. Background The reported infection rates, and the burden of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) in low- and middle-income countries, including sub-Saharan Africa, are relatively low compared to Europe and America, partly due to limited testing capabilities. Unlike many countries, in Tanzania, neither mass screening nor restrictive measures such as lockdowns have been implemented to date. The prevalence of SARS-CoV-2 infection in rural mainland Tanzania is largely unknown. Methods Between April and October 2021, we conducted a cross-sectional study to assess anti-SARS-CoV-2 seroprevalence among mother-child pairs (n=634 children, n=518 mothers) in a rural setting of north-eastern Tanzania. Findings We found a very high prevalence of anti-SARS-CoV-2 antibody titres with seroprevalence rates ranging from 29% among mothers and 40% among children, with a dynamic peak in seropositivity incidence at the end of July/early in August being revealed. Significant differences in age, socioeconomic status and body composition were associated with seropositivity in mothers and children. No significant associations were observed between seropositivity and comorbidities, including anaemia, diabetes, malaria, and HIV. Interpretations The SARS-CoV-2 transmission in a rural region of Tanzania during 2021 was high, indicating a much higher infection rate in rural Tanzania compared to that reported in the UK and USA during the same period. Ongoing immune surveillance may be vital to monitoring the burden of viral infection in rural settings without access to molecular genotyping where a load of communicable diseases may mask COVID-19. Surveillance could be implemented in tandem with the intensification of vaccination strategies. Keywords COVID-19, SARS-CoV-2, Tanzania, Low-and-Middle-Income-Countries, rural seroprevalence, COMORBIDITIES, children, Mothers, antibodies, future strategies Abbreviations FOETALforNCD, Foetal exposure and Epidemiological Transition: the role of Anaemia in early Life for Non-Communicable Diseases in later life NCDs, non-communicable diseases Hb, haemoglobin LMIC, low- and middle-income countries LBW, low birthweight KDH, Korogwe district hospital BP, blood pressure hs-CRP, high sensitivity C-reactive protein GA, gestational age STOPPAM, Strategies TO Prevent Pregnancy Associated Malaria A. U, Arbitrary units ==== Body pmcINTRODUCTION Apart from South Africa (Cohen et al. 2022), the epidemiology of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) in Africa is not well described and infection rates, and disease burden of Coronavirus Disease 2019 (COVID-19) are poorly understood. In many low- and middle-income countries (LMICs), including Tanzania, diagnostic and epidemiological surveillance systems are sub-optimal (Obande et al. 2021). Thus, it is plausible that transmission burden and dynamics have been underestimated, and the impact on vulnerable populations is largely unknown. Therefore, for effective control and management of COVID-19, there is an immediate need for reliable data on SARS-CoV-2 infection and transmission dynamics of the disease in LMICs. While previous reports indicated less clinical severity of COVID-19 in LMICs (Simeni Njonnou et al. 2021; Johns Hopkins University 2020), a recent systematic review has illustrated that the risk of infection and most significant impact of the pandemic occurred in poorer countries (Levin et al. 2022). In sub-Saharan African (SSA) countries, COVID-19 presents a serious health threat in terms of morbidity and mortality and has severe economic impact (Tripathi et al. 2021). In Tanzania, more than 46% of the population lives below 2 USD per day, and COVID-19 has adversely affected all economic sectors in the country (Tripathi et al. 2021; The World Bank 2020). The pandemic continues to inflict a heavy toll on the already fragmented health care system (USAID 2021). During the very early phases of the pandemic, the Tanzanian government initiated interventions to curtail SARS-CoV-2 transmission including suspending large gatherings and introducing facemasks. However, the majority of such efforts were short-lived due to logistical challenges. By May 2020, no SARS-CoV-2 data from Tanzania were relayed to the Centers for Disease Control and Prevention (CDC) (Mfinanga et al. 2021). Furthermore, neither mass screening, contact tracing, nor restrictive measures such as lockdowns have been implemented in Tanzania to date. However, following the swearing-in of President Samia Suluhu Hassan in March 2021, COVID-19 preventive strategies were revitalised by forming a National COVID-19 Taskforce. As of July 27 2022, 36,886 confirmed cases of COVID-19, including 841 deaths from Tanzania, have officially been reported (WHO 2022). Notwithstanding, there are no reliable estimates on the burden of SARS-CoV-2 infection in Tanzania, with an estimated population size of over sixty million (WHO 2022). Serological detection of specific anti-SARS-CoV-2 antibodies is a valuable tool for understanding infection dynamics, and epidemiology of SARS-CoV-2 (Hansen et al. 2021). Few population-based anti-SARS-CoV-2 serological surveys have been done in the SSA (Sagara et al. 2021; Salako et al. 2021; Uyoga et al. 2021; Mandolo et al. 2021). The most extensive are urban studies from South Africa (n=2757)11, and Zambia (n=4258) (Mulenga et al. 2021). In both studies, the estimated number of SARS-CoV-2 cases using the presence of anti-SARS-CoV-2 antibodies was much higher than the number of reported COVID-19 cases using other surveillance platforms (Shanaube et al. 2022; Cohen et al. 2022). No estimate exists on the severity of the COVID-19 pandemic, and the number of infections in rural Tanzania. We therefore investigated anti-SARS-CoV-2 seroprevalence in an unvaccinated population cohort of mothers and children in a rural Tanzanian setting from April to October 2021, including antibody neutralising capacities and potential socioeconomic and clinical factors influencing the seropositivity. METHODOLOGY Study design This population-based, cross-sectional study was conducted in Korogwe and Handeni Districts, Tanga Region, north-eastern Tanzania, and covered 47 rural villages and 7 peri-urban townships located in Korogwe village. Here, we define peri-urban as a village/township that has urban capacities including shops and schools, but retain rural characteristics such as substantial reliance on agricultural production. The cohort included children aged 5-12 years and their mothers aged 21-56 years, who previously participated in two longitudinal pregnancies studies, entitled “Strategies TO Prevent Pregnancy Associated Malaria” (STOPPAM) (Schmiegelow et al. 2012) and “Foetal exposure and epidemiological transition: the role of anaemia in early life for non-communicable diseases in later life” (FOETALforNCD) (Hjort et al. 2019). From 26th April to 27th October 2021, when the children were between age 5-6 years (FOETALforNCD) and 11-12 years old (STOPPAM), the child and mother pairs were invited to participate in a follow-up health examination study (the PONA2 study). Trained field workers made door-to-door visits to identify participants who previously participated in the STOPPAM and/or FOETALforNCD study. The inclusion criteria were as follows: 1) Mothers must have participated in the STOPPAM and/or FOETALforNCD study, 2) Data on gestational age and/or birth weight should be available, 3) The child should be enrolled in the PONA2 study examination, for the mother to be enrolled (mothers could not participate without child). If eligible, detailed information of the study was provided, and those who provided informed consent to participate (parent(s) or legal guardian consented on behalf of the children) were enrolled in PONA2. To prevent viral transmission during data collection participants presenting with current, common COVID-19 symptoms, had their enrolment postponed until they were symptom free. Ethical considerations The study received ethical approval from the Tanzania Medical Research Coordinating Committee (NIMR/HQ/R.8a/Vol. IX/3503, August 26 2020 and amendment number NIMR/HQ/R.8a/Vol. 1/970, November 12 2021). Ethical approval for COVID-19 research of the present cohort was not approved before data collection was finished. Hence, we could not include COVID-19 testing and disease symptoms reporting as part of the clinical examination. All study procedures were performed according to good clinical and laboratory practices and the Declaration of Helsinki. All participants were treated according to the Tanzanian national guidelines and the project assisted all participants in obtaining the best local medical care available if disease was diagnosed. The data sharing followed the Tanzanian Medical Research Coordinating Committee ethical guidelines. Sociodemographic and clinical information At enrolment, sociodemographic data, including profession, education, housing, main source of drinking water, and other lifestyle factors, were collected using structured interviews. Previous medical history of both mothers and children was documented. Body composition (body weight, body fat percentage, fat mass, and muscle mass) was estimated by bioimpedance analysis methodology, using the Tanita DC-400M (TANITA). Blood pressure was measured in seated position on the right arm (Omron Health Care Europe). Height was measured with a stadiometer (precision 1mm). Mid-upper arm circumference (MUAC) was measured on the upper right arm at the midpoint of the acromion process and the tip of the olecranon (precision 1mm). Skinfold thickness was measured using the Harpenden skinfold calliper (BATY International). Waist circumference was measured just above the iliac crest in the horizontal plane and hip circumference was measured at the point yielding the maximum circumference over the buttocks, both using a standard measuring tape to the nearest 1mm. All anthropometric measures were collected using standard operating procedures. Finally, left and right hand grip strength was measured for all mothers, and children aged 11-12 years using SAEHAN hand grip dynamometer (SAEHAN, Korea). Blood sample collection and processing Venous blood was collected from both children and mothers in EDTA-coated, plain and heparinized vacutainer tubes, and transported at 2-8 ⁰C to the National Institute for Medical Research (NIMR) Korogwe Research Laboratory for further processing and laboratory analyses. Plasma was separated within two hours of collection in a refrigerated centrifuge. All samples were stored at -80 ⁰C and later shipped on dry ice to Rigshospitalet, Denmark for detection of anti-SARS-CoV-2 antibodies by ELISA. High-sensitivity C-reactive protein (hsCRP) was measured at the Department of Clinical Biochemistry, Copenhagen University Hospital, using the Cobas 8000 c702 system (ROCHE diagnostics). At the study site, point-of-care diagnostics were performed on the venous blood including estimation of haemoglobin (Hb) (Sysmex KX-21N), fasting blood glucose (HemoCue 301), and malaria rapid diagnostic test (mRDT) (ParaHIT or CareStart Malaria Pf). For participants with a positive mRDT, the parasitaemia was confirmed by two independent expert microscopists. HIV infection was tested by the Determine HIV-1/2 test kit (Alere ltd) and seropositive cases were confirmed using Unigold test kit (Trinity Biotech Plc). All tests were done according to the manufacturer's instructions. Detection of total anti-SARS-CoV-2 antibodies Detection of total antibodies against SARS-CoV-2 spike (S) protein and receptor binding domain (RBD) as a proxy of previous infection in non-vaccinated individuals was performed using an in-house sandwich ELISA (S-ELISA) as described previously(Hansen et al. 2021), with minor modifications. Briefly, Nunc MaxiSorp 96-well plates (Thermo Fisher Scientific) were coated with 0.5 µg/mL of RBD overnight at 4°C. Plates were blocked for 1 hour with PBS + 0.05% Tween (Merck) (PBS-T) followed by the addition of plasma samples at 1:2 dilution in PBS-T and incubated for 1 hour. A solution containing 0.5 µg/mL RBD biotinylated in PBS-T was applied and incubated for 1 hour. A solution of 1:16000 HRP-conjugated high-sensitivity streptavidin (Pierce) in PBS-T was applied and incubated for 1 hour. TMB-One (Kem-En-Tec) was used as a substrate and plates were revealed for 5 min. The reaction was stopped by using 0.3M H2SO4 and optical density (O.D.) was measured using a Synergy H.T. absorbance reader (Biotek Instruments) at 450-630 nm. A recombinant-human IgG monoclonal antibody against protein S/RBD (Genscript) diluted 1:2000 in PBS-T was used as a positive control. A pool of uninfected/unvaccinated normal human serum was used as a negative control. Plates were washed three times in PBS-T between steps, and incubations were performed while shaking at room temperature. The assay positivity threshold was set to O.D.> 0.1585. Quantitative determination of anti-SARS-CoV-2 antibody isotypes Samples that were positive on the S-ELISA were further analysed for the quantitative determination of IgG, IgM and IgA levels using an in-house ELISA-based assay as described elsewhere (Hansen et al. 2021; Pérez-Alós et al. 2022). ACE-2/RBD antibody inhibition measurement To study the ability of the measured antibodies to prevent the binding of the protein S RBD to the host receptor angiotensin-converting enzyme-2 (ACE-2), a previously described in-house ELISA-based assay was used (Bayarri-Olmos et al. 2021). A pool containing plasma from vaccinated individuals (previously quantified into I.U./mL using The Working Reagent for anti-SARS-CoV-2 immunoglobulin 21/234, NIBSC) diluted 1:50 in PBS-T was used as a positive control. The assay positivity threshold was set to 420IU/mL. Statistics Assessment of antibody isotypes and neutralization levels was performed using GraphPad version 9.3.1 (GraphPad Software). IgG, IgM, IgA, and neutralization levels were interpolated using the non-linear regression four-parameter curve fitting. IgG, IgM, and IgA results were given in A.U./mL, where the highest concentration of the calibrator was given a value of 200AU/mL. Neutralization results were given in I.U./mL, where the highest concentration of the calibrator was given a value of 520 IU/mL. For univariate analyses, normally distributed data are presented as mean±SD and compared by Student's t-test, whereas non-normally distributed data are presented as median and interquartile range (IQR), and compared by the Mann-Whitney U test. Proportions of categorical data are compared using Fisher´s Exact test. Potential differences in socioeconomic and clinical characteristics between sero-negative and -positive cases were examined by separate analyses between the STOPPAM and PONA groups, due to the apparent age differences between these participants. RESULTS Anti-SARS-CoV-2 seroprevalence in 5-12-year-old children and mothers in rural north-eastern Tanzania In total, 634 children and 513 mothers had a fasting venous blood sample collected and were included in the SARS-CoV-2 antibody study (n=1147). Of note, 94 of the 634 children participated in the study without their mothers since the mothers were not available due to either work, travelling, or being deceased. Among the 634 children, 45 were siblings, of which there were 13 twin pairs. One mother participated with 3 children, of which two were twins and one was a younger sibling. Hence, 21 sibling pairs and one sibling group of three children were included. Among all participants (n=1147), the overall anti-SARS-CoV-2 seroprevalence was 37.1%, with children having a seroprevalence of 39.0% (n=634), and mothers of 29.0% (n=513) (Table 1 ). Stratifying the children by age groups and anti-SARS-CoV-2 seroprevalence, the seroprevalence was significantly higher among the 11-12 year-old children (43.3%) compared to the 5-6 year-old children (33.0%) (Fisher´s exact test, p=0.008). Female children were more likely to be seropositive than males in the 11-12-year-old group, (50.5% vs. 35.9%, Fisher´s exact test, p=0.006).Table 1 Anti-SARS-CoV-2 Seroprevalence of total antibodies against SARS-CoV-2 protein receptor-binding domain in two mother-child cohorts, from Korogwe District, Tanga Region, Tanzania, from April 26 to October 27, 2021 Table 1Children SARS-CoV-2 negative SARS-CoV-2 positive % positive PONA1 cohort, 5-6-years-old (n=267) 179 88 33.0% STOPPAM cohort, 11-12-years-old (n=367) 208 159 43.3% In total (n=634) 387 247 39.0% Mothers PONA1 cohort (mean age 34.5 years, n=222) 151 71 32.0% STOPPAM cohort (mean age 39.9 years, n=291) 213 78 26.8% In total (n=513) 364 149 29.0% Of the 513 mother-child pairs included, the majority had concordant seroprevalence (positive vs. negative), with 253 (49.3%) pairs being negative, and 89 (17.4%) pairs being positive. However, in 111 (21.6%) mother-child pairs, the mother was seronegative and the child seropositive; conversely, in 60 (11.7%) pairs, the mother was seropositive and the child seronegative for anti-SARS-CoV-2 antibodies. Of the sibling pairs/groups, 14 pairs had concordant seroprevalence, and eight with divergent seroprevalence (negative vs. positive). The majority (96%) of the mother-child pairs enrolled were living in the same households. Date-specific and demographical effects on anti-SARS-CoV-2 seroprevalence As shown in Fig. 1 , a clear rise in anti-SARS-CoV-2 seropositivity occurred both among children (Fig. 1 A) and mothers (Fig. 1 B), by the end of July/early August 2021.Figure 1 Anti-SARS-CoV-2 seroprevalence of total Ig against SARS-CoV-2 RBD of the spike protein in A) children age 5-12 years, and B) mothers aged 21-59 years. Number of individuals examined each date was on average seven children and six mothers. Exact numbers of individuals by date are shown in Supplementary Figure 1. Figure 1 Since the participants lived in 54 smaller villages (47 rural and 7 peri-urban located in the periphery of Korogwe township), we examined the numbers of positive vs. negative cases across the different villages. The anti-SARS-CoV-2 seroprevalence was heterogeneous across the villages, ranging from 0% in Kitifu and Kwamdulu, to 70% in Michungwani. As highlighted in Fig. 2 , the villages located along the main road between north and south through the study area, appeared to have higher sero-positivity rates than the villages in the periphery of the main roads. The average anti-SARS-CoV-2 seroprevalence positivity in the 47 villages characterized as rural were 40% and 27.5%, for children and mothers, respectively, and hence higher than in the 7 villages characterized as peri-urban with a seropositivity of 25.5% and 21.5% of children and mothers, respectively.Figure 2 Overview of the villages included in the study. The Korogwe town consist of seven smaller city areas, not shown on the map and includes Kilole, Magundi, Majengo, Manundu, Masuguru, Mtonga, and Old Korogwe. Figure 2 IgG, IgM, IgA antibody levels and neutralising antibodies Levels of IgM, IgA and IgG and neutralising antibodies against SARS-CoV-2 were measured in plasma samples from the 396 individuals that were identified as positive for total SARS-CoV-2 Ig. IgG was the most abundant isotype compared to IgM and IgA (Fig. 3 A, Supplementary Fig. 2). The mothers had higher levels of antibodies and neutralising antibody titres than the children (p≤0.01) did (Fig. 3 , Supplementary Fig. 2).Figure 3 Quantitative determination of IgG antibody levels and neutralising antibodies against RBD in anti- SARS-CoV-2 seropositive individuals. A) IgG levels, represented in log (A.U./mL), in mothers aged 21-59 years and children age 5-12 years. B) Levels of neutralising antibodies, represented in log (I.U./mL), measured in mothers aged 21-59 years and children age 5-12 years. Horizontal dashed line represents assay positivity threshold. A p-value <0.05 was considered statistically significant. ** p<0.01 by Mann-Whitney U test. Figure 3 Corresponding to the rise in seropositive cases experienced by the end of July/early August 2021 (Fig. 1), the levels of IgG antibodies against SARS-CoV-2 were also higher in both children and mothers during this period of time (Supplementary Fig. 3). Association between anti-SARS-CoV-2 seropositivity and socioeconomic and clinical characteristics, and comorbidities Among the older mothers, with a mean age of ∼40 years (STOPPAM), but not among the younger mothers with a mean age of ∼34.5 years (FOETALforNCD), several significant differences were observed between the seropositive versus seronegative women (Table 2 ). Specifically, we observed that seropositive women were more likely to rely on a public domestic water source, such as a river, compared to the seronegative women who were more likely to have tap/well water available (p=0.04) (Table 2). There were no differences in ethnicity, rural vs. peri-urban village, education level, profession, or monthly household income between the seropositive and seronegative mothers (Table 2). A higher weight and body fat percentage, more muscle mass, larger MUAC and skinfold thickness, and larger right hand grip strength were observed among the older seropositive mothers compared to the seronegative cases, (p≤0.04) (Table 2).Table 2 Clinical characterization of the mothers Table 2PONA1 mothers STOPPAM mothers SARS-CoV-2 Negative (n=151) SARS-CoV-2 Positive (n=71) p-value SARS-CoV-2 Negative (n=213) SARS-CoV-2 Positive (n=78) p-value Age (years) 34.3 ±7.0 34.7 ±6.4 0.69 39.8 ±6.1 40.3 ±7.0 0.58 Ethnicity (n) Sambaa 61 (40.4%) 24 (33.8%) 0.59 105 (49.3%) 41 (52.6%) Zigua 50 (33.1%) 30 (42.3%) 47 (22.1%) 15 (19.2%) Pare 7 (4.6%) 5 (7.0%) 0 (0.0%) 0 (0.0%) 0.99 Bondei 5 (3.3%) 2 (2.8%) 7 (3.3%) 2 (2.6%) Others 28 (18.5%) 10 (14.1%) 53 (24.9%) 20 (25.7%) Main source of water (n) Tap 100 (66.2%) 57 (80.3%) 0.07 147 (69.0%) 46 (59.0%) Well 15 (9.9%) 6 (8.5%) 39 (18.3%) 12 (15.4%) 0.04 River 36 (23.8%) 8 (11.3%) 26 (12.2%) 20 (25.6) Rural vs. peri-urban village (%) Rural (95.4%) Rural (97.2%) 0.72 Rural (82.2%) Rural (82.1%) 1.00 Educational level None 17 (11.3%) 9 (12.7%) 22 (10.3%) 3 (3.9%) Partial primary 19 (12.6%) 9 (12.7%) 0.35 39 (18.3%) 14 (18.0%) 0.14 Complete primary 104 (69.9%) 52 (73.2%) 138 (64.8%) 51 (65.4%) Secondary or higher 11 (7.3%) 1 (1.4%) 14 (6.6%) 10 (12.8%) Profession Housewife 12 (8.0%) 4 (5.6%) 19 (8.9%) 7 (9.0%) Farmer/Livestock keeper 117 (77.5%) 49 (69.0%) 143 (67.1%) 49 (62.8%) Service 6 (4.0%) 2 (2.8%) 0.13 10 (4.7%) 5 (6.4%) 0.66 Business/Professional 14 (9.3%) 16 (22.5%) 38 (17.8%) 16 (20.5%) Other 2 (1.3%) 0 (0.0%) 3 (1.4%) 1 (1.3%) Household monthly income (TZS)* 150.000 (70.000,200.000) 120.000 (60.000,200.000) 0.78 100.000 (50.000,200.000) 150.000 (100.000,300.000) 0.18 Height (cm) 155.6 ±5.8 154.9 ±6.3 0.42 156.1 ±5.9 157.0 ±5.5 0.26 Weight (kg) 58.4 (50.3,70.9) 58.4 (50.7,64.9) 0.62 56.8 (49.8,68.2) 59.3 (54.1,73.1) 0.03 BMI (kg/m2) 24.0 (20.8,29.3) 24.4 (20.7,27.7) 0.88 23.2 (20.5,28.2) 25.2 (21.3,29.2) 0.051 Skinfold (mm) 20.0 (12.2,30.0) 21.8 (15.9,28.3) 0.40 19.8 (13.7,28.1) 21.2 (16.9,31.4) 0.04 MUAC (cm) 30.4 ±5.1 30.2 ±4.3 0.77 29.8 ±4.4 31.1 ±4.8 0.03 Total body fat (%) 31.9 ±7.7 32.4 ±6.8 0.67 32.0 ±7.5 34.2 ±6.8 0.02 Muscle mass (kg) 38.4 ±5.4 37.7 ±5.2 0.34 37.8 ±4.9 39.5 ±5.7 0.02 Right hand grip strength (unit) 61.9 ±10.8 59.5 ±9.5 0.11 59.0 ±10.6 62.5 ±10.7 0.01 Left hand grip strength (unit) 58.6 ±11.3 57.5 ±10.8 0.50 56.9 ±9.8 58.1 ±9.8 0.38 Diastolic blood pressure (mmHg) 78.6 ±12.5 77.6 ±12.3 0.60 81.8 ±11.3 80.6 ±11.8 0.42 Systolic blood pressure (mmHg) 121.0 (113.7,130.5) 118.0 (109.7,128.5) 0.13 123.5 (114.5,134.0) 123.0 (113.5,137.3) 0.99 Fasting hsCRP (pmol/L) 1.55 (0.64,3.70) 1.48 (0.59,2.87) 0.46 1.29 (0.59,3.14) 1.36 (0.71,4.26) 0.22 Haemoglobin (g/dL) 12.5 ±1.6 12.5 ±1.4 0.94 12.7 ±1.8 12.7 ±1.3 0.93 Anaemia (n) 9 (6.0%) 3 (4.2%) 0.76 10 (4.7%) 2 (2.6%) 0.52 Diabetes (n) 15 (9.9%) 6 (8.5%) 0.81 17 (8.0%) 3 (3.9%) 0.30 Malaria (n) 3 (2.0%) 0 (0.0%) 0.55 4 (1.9%) 1 (1.3%) 1.00 HIV (n) 5 (3.3%) 4 (5.6%) 0.47 7 (3.3%) 3 (3.9%) 0.43 ⁎ Household monthly income was selfreported. Data is presented as mean (±SD), median (IQR), or percentage. Among the older children, we observed that seropositive children were slightly older, more likely to live in a rural compared to peri-urban village, and to use a river as main water source, compared to the seronegative children (p≤0.02). Furthermore, the seropositive children had a larger handgrip strength, similar to what was observed for the seropositive mothers (p=0.03). Finally, children of the Sambaa ethnicity were more likely to be seropositive (p=0.01). Importantly, we found no associations between seropositivity and other comorbidities, including anaemia, diabetes, malaria and HIV for neither the mothers nor the children (Table 3 ).Table 3 Clinical characterization of the children Table 3PONA1 children (age 5-6 years) STOPPAM children (age 11-12 years) SARS-CoV-2 Negative (n=179) SARS-CoV-2 Positive (n=88) p-value SARS-CoV-2 Negative (n=208) SARS-CoV-2 Positive (n=159) p-value Age (years) 5.4 ±0.4 5.5 ±0.4 0.16 11.6 ±0.4 11.8 ±0.4 0.001 Sex (female) 85 (47.5%) 46 (52.3%) 0.52 92 (44.2%) 94 (59.1%) 0.006 Ethnicity (n) Sambaa 72 (40.2%) 6 (40.9%) 84 (44.4%) 84 (52.8%) Zigua 3 (29.6%) 20 (22.7%) 40 (19.2%) 37 (23.3%) Pare 10 (5.6%) 7 (8.0%) 0.52 12 (5.8%) 9 (5.7%) 0.01 Bondei 6 (3.4%) 6 (6.8%) 6 (2.9%) 2 (1.3%) Others 38 (21.2%) 19 (21.6%) 66 (31.7%) 27 (17.0%) Main source of water (n) Tap 126 (70.4%) 64 (72.7%) 147 (70.7) 100 (62.9) Well 17 (9.5%) 6 (6.8%) 0.83 38 (18.3) 21 (13.2) 0.003 River 36 (20.1%) 18 (20.5%) 22 (10.6) 38 (23.9) Rural vs. peri-urban village (%) Rural (94.4%) Rural (98.9%) 0.11 Rural (77.9%) Rural (87.4%) 0.02 Household monthly income (TZS)* 145.000 (70.000,200.000) 150.000 (70.000,300.000) 0.47 100.000 (50.000,200.000) 155.000 (100.000,300.000) 0.06 Primary school attendance N.A. NA - 206 (99.0%) 158 (99.4%) 1.00 Height (cm) 106.9 ±5.4 107.8 ±5.0 0.20 138.2 ±7.0 139.5 ±7.3 0.10 Weight (kg) 16.4 ±2.1 16.8 ±2.4 0.20 29.9 (26.9,32.9) 30.2 (27.2,33.9) 0.32 BMI (kg/m2) 14.3 ±1.2 14.4 ±1.4 0.62 15.4 (14.7,16.7) 15.7 (14.6,16.7) 0.39 Skinfold (mm) 8.2 ±1.9 8.9 ±3.9 0.11 8.1 (6.5,10.3) 8.5 (7.0,10.9) 0.14 MUAC (cm) 16.0 ±1.1 16.1 ±1.3 0.23 19.1 (17.9,20.3) 19.4 (17.9,20.6) 0.50 Total body fat (%) 17.2 ±3.4 17.9 ±3.7 0.20 12.9 (10.7,16.1) 13.5 (11.0,18.0) 0.16 Muscle mass (kg) 12.7 ±1.6 12.9 ±1.6 0.35 24.5 ±3.3 25.1 ±3.9 0.12 Right hand grip strength (unit) N.A. NA - 33.3 ±7.3 35.0 ±7.6 0.03 Left hand grip strength (unit) N.A. NA - 31.4 ±6.9 33.0 ±7.4 0.03 Diastolic blood pressure (mmHg) 66.3 ±9.2 66.2 ±9.6 0.94 70.3 ±8.2 70.4 ±8.1 0.90 Systolic blood pressure (mmHg) 101.9 ±10.2 100.5 ±9.9 0.29 113.4 ±9.5 114.0 ±9.3 0.50 Fasting hsCRP (pmol/L) 0.72 (0.30,1.65) 0.59 (0.30,1.73) 0.63 0.58 (0.30,1.66) 0.48 (0.30,1.38) 0.33 Haemoglobin (g/dL) 12.1 ±1.3 11.9 ±1.1 0.34 12.8 ±1.3 12.8 ±1.3 0.96 Anaemia (n) 5 (2.8%) 7 (8.0%) 0.07 2 (1.0%) 5 (3.1%) 0.25 Malaria (n) 15 (8.4%) 6 (6.8%) 0.81 13 (6.3%) 16 (10.1%) 0.24 HIV (n) 0 (0.0%) 0 (0.0%) - 0 (0.0%) 1 (0.6%) 0.43 ⁎ Household monthly income was selfreported. Data is presented as mean (±SD), median (IQR), or percentage. DISCUSSION This is the first epidemiological study of the anti-SARS-CoV-2 seroprevalence in the Tanzanian mainland. Between April and October 2021, we found a very high anti-SARS-CoV-2 seroprevalence with seropositivity rates ranging from 29-40% in relatively young adult women and 5-12-year-old children. We found important epidemiological differences in seropositivity rates, including increased with female sex (in children), with increasing age, adiposity, and handgrip strength and with lower socioeconomic status. We were able to document dynamic peaks in viral infections through anti-SARS-CoV-2 seroprevalence measurements, and our data support that this may be an important tool when assessing previous SARS-CoV-2 infections in rural and non-tested/unvaccinated settings, and is highly suitable for large-scale surveillance for SARS-CoV2 antibodies (Nuccetelli et al. 2022). High SARS-CoV-2 seroprevalence in rural, north-eastern Tanzania Tanzania reported its first laboratory-confirmed COVID-19 case on March 15 2020 (Tarimo and Wu 2020), and by July 27 2022, 36,886 confirmed COVID-19 cases have been made in Tanzania, according to WHO (WHO 2022). With a population of 59.7 million people (WHO 2022), this corresponds to an overall percentage of 0.06%, when conservatively assuming that all confirmed SARS-CoV-2 cases were Tanzanian residents, and none were diagnosed more than once. In an epidemiological setting, seroprevalence estimates are useful for understanding the burden of asymptomatic or subclinical infections that would not otherwise be detected. Based on the average anti-SARS-CoV-2 seroprevalence of 37.1% in the present study, the anti-SARS-CoV-2 seropositivity rate is hence more than 700 times higher than the reported cases when considering the results from the rural north-eastern region to be projectable for the entire country. The seroprevalence is in line with results from the same time period (Jan-Aug 2021), in unvaccinated individuals of other SSA countries, including Mali, with seropositivity of 58.5% (n=2672) (Sagara et al. 2021), Nigeria with seropositivity of 42% (n=802) (Chechet et al. 2022), and in Zambia, where seropositivity increased from ∼13.5% in December 2020 to 35% in March 2021 (n=2977) (Shanaube et al. 2022). A recent similar study that was conducted in Zanzibar isles has also shown a seroprevalence of almost 60% (Salum et al. 2022). Finally, a malariometric and SARS-CoV-2-19 survey from July 2021, conducted in two villages from the same study area as the present cohort, was very recently published (Lyimo et al. 2022). One of these two villages, Mkokola, was also included in our study, and in comparsion, Lyimo et al. found a seropositivity rate of 32.5% (76/234), where we in the present study found a rate of 15.4% (16/104). The seroprevalence observed in our study and other SSA countries are substantially higher than that reported in population-wide studies from U.K. and the U.S.; One large US study (n>2mio.), also conducted during the second wave, estimated approximately 18% of the entire U.S. population to have infection-induced positive seroprevalence of SARS-CoV-2-19 (Jones et al. 2022). However, it is worth noting that contrary to other SSA studies (Uyoga et al. 2021; Chechet et al. 2022; Sagara et al. 2021), our study was conducted in a rural setting with little demographic movement. Therefore, in larger cities and urban regions with more opportunity for transmission, the anti-SARS-CoV-2 seroprevalence is likely to be much higher than in rural areas. In addition, our population is young compared to other large-scale SSA studies where older age groups and men were also included (Cohen et al. 2022; Shanaube et al. 2022). Surprisingly, the anti-SARS-CoV-2 seroprevalence was higher among children than in mothers (39% versus 29%). This is in contrast to previous studies from same time period in South Africa, Mali and Zambia showing that seroprevalence increased with age and was generally higher in adults than children (Simeni Njonnou et al. 2021; Sagara et al. 2021; Cohen et al. 2022; Mulenga et al. 2021). In addition, the anti-SARS-CoV-2 seroprevalence in children in our study was much higher than that reported in paediatric populations elsewhere. Specifically, a study conducted in South Africa after the second wave reported a seroprevalence of 18% in children less than 5 years of age (Cohen et al. 2022). Similarly, lower seroprevalence rates have been reported in children in Zambia (4.0%) (Mulenga et al. 2021), Denmark (3.2%) (Espenhain et al. 2021), Germany (2-10%) (Sorg et al. 2022), and Canada (5.8%) (Zinszer et al. 2021). However, recently a large study from South Africa published results from the omicron wave, and showed that in both rural and urban areas, the largest increase in seroprevalence in this later time period was seen in children of 13-18 years of age compared to both younger and older age groups (Kleynhans et al. 2022). There are still limited data on the SARS-CoV-2 seroprevalence among the paediatric population in rural African settings. To the best of our knowledge, our study is the most extensive analysis showing individual-level paediatric SARS-CoV-2 infection in SSA, with detailed clinical and socioeconomic data available. It remains unknown why a higher seroprevalence among children is found in the rural setting of our study. However, it can be speculated to involve socioeconomic status and access to clean water, perhaps in combination with different genetic backgrounds, that may increase infection vulnerability. Current approach to SARS-CoV-2 testing in Tanzania is missing most infections The high anti-SARS-CoV-2 seropositivity in children and mothers reported here, may have important implications for this silent burden of disease and transmission dynamics in our study area, where the majority of the population was, and may still be, ignorant about the disease. The uptake of COVID-19 vaccination in Tanzania is suspected to have remained low, but there are no good estimates. It is also important to note that the first SARS-CoV-2 vaccines reached Tanzania by end of July 2021, and at that time, only healthcare workers and elderly Tanzanian citizens were prioritized for vaccination. None of the participants in our study were offered SARS-CoV-2 vaccination while our data collection was ongoing. Furthermore, when the first batch of COVID-19 vaccines was received in the country, only specific groups including healthcare workers who were working in the frontline were given the priority. Our findings align with a recent meta-analysis of socioecological, biophysical, and public health interventions from 47 African countries, which estimated that the African region has had a similar number of COVID-19 infections compared to the rest of the world (Cabore et al. 2022). The higher than expected anti-SARS-CoV-2 seropositivity rate for mothers and children in our study is likely to be partly attributable to continued viral transmission over the study period in this rural and peri-urban population, where no effective preventive strategies were in place (Patterson 2022). Indeed, we found a progressive increase in seropositivity from April to August 2021 during the sample collection, indicative that active viral transmission from an unfolding COVID-19 wave was occurring during this period in rural north-eastern Tanzania. However, the majority of the anti- SARS-CoV-2 seropositive cases from this time period in our study may be speculated to have been asymptomatic since participants presenting with any of the common COVID-19 symptoms had their enrolment and clinical examination postponed preventing viral transmission during our data collection. Asymptomatic individuals have been shown to transmit SARS-CoV-2 at similar levels to symptomatic ones (Cohen et al. 2022). Therefore, there is a need for surveillance of hospitalisations, comorbidities, and scale-up of representative seroprevalence studies, as core response strategies. This continued viral transmission in rural areas in a context of low vaccine coverage creates room for the emergence of viral variants of concern, suggesting that interventions targeting asymptomatic individuals, including increased vaccine coverage to reach children, is needed (Shiri et al. 2021). Being female and at school-age is associated with a higher risk of being SAR-CoV-2 seropositive We found a higher prevalence of SARS-CoV-2 seropositivity among older children (11-12 years) compared to their younger counterparts (5-6 years). The older children comprise majority of school-age children and at the time of data collection, children were attending school. Hence, the higher proportion of SARS-CoV-2 seropositivity among older children may reflect age-dependent susceptibility to infection and differences in contact patterns between younger children (Davies et al. 2020). The transmission might have predominantly spread from one child to another as they interact with each other with less protection in place. Indeed, we observed trends towards an increased seropositivity rate from April-June 2021 when schools were open, dipping again in July as schools were closed for the summer holidays. This is in line with a recent epidemiological study from Kenya, which found that reopening schools between the second and third waves led to only a minor increase in viral transmission and that socioeconomic and urban-rural population structures were critical determinants of the disease transmission (Brand et al. 2021). Other factors such as socioeconomic status, occupation of children and parents, and weather patterns might as well have influenced transmission dynamics; but these were beyond the scope of the current analysis. The impact of sex on SARS-CoV-2 infection is somewhat inconsistent and there is no general agreement about the impact of sex on antibody generation and prognosis in SARS-CoV-2 infection (Ho et al. 2022; Poustchi et al. 2021). While several studies involving the adult population have reported higher anti-SARS-CoV-2 antibody titres in women (Poustchi et al. 2021; Grzelak et al. 2021), other studies have reported equivalent levels in both sexes (Luo et al. 2021). We found that SARS-CoV-2 infection was higher in female compared to male children. Previous studies suggest that the immune response to most pathogens is lower in men compared to women (Takahashi et al. 2020). Therefore, an increased capacity to mount greater magnitudes of immune responses against the infection in females as compared to males may partly explain the difference between sexes in our study, but this was not investigated. Other factors such as cultural practices like different gender-specific roles may render more female children to interact with each other more than males, for instance, during in-doors household work and sharing beds and bednets. Clinical and socioeconomic differences between the anti- SARS-CoV-2 seropositive and seronegative cases The understanding of the COVID-19 pathophysiology and how SARS-CoV-2 transmits and affects LMICs and vulnerable populations with low socioeconomic status levels is still poor, which is a cause for concern. Not only due to the high and probably underestimated SARS-CoV-2 transmission rates in these regions (Obande et al. 2021; Cabore et al. 2022), but also since such populations often live under a double burden of infectious diseases and increasing risks of non-communicable diseases, including hypertension, diabetes, and obesity (Owino 2019). Noticeable is the lack of associations between seropositivity and comorbidities in this cohort, including HIV, malaria, anaemia and diabetes, which is not immediately easy to explain. This may be due to the young age of the participants, or simply sample size limitations, but could also reflect underlying important biological processes. Nevertheless, this should be examined in future investigations, including men and elderly individuals. However, we found that participants who did not have a private water source but relied on drinking water from the river were more likely to have had a previous SARS-CoV-2 infection. Further, we found that the older SARS-CoV-2 seropositive mothers had higher weight and body percentage, body fat mass, more muscle mass, larger MUAC and skinfold thickness, and larger right-hand grip strength compared to the seronegative cases. In addition, seropositive children also tended to be larger with more adipocity than the seronegative children. Available data show that obesity is a major risk factor for hospitalization and mortality related to COVID-19 (Wu and McGoogan 2020; Frasca et al. 2021), which is in line with our results showing increased adiposity among previously infected individuals. It can also be speculated that obesity may be a risk factor for disease and/or antibody acquisition too, which could explain why our results are not indicative of any association between sero-positivity and other underlying conditions such as HIV and diabetes. Limitations of the study Data collection of the crossectional study lasted from late April till late October, and due to the limited knowledge of viral transmission in both Tanzania and neighbouring countries in this time period, it remains unclear when specific waves of pandemic transmissions were occurring in the country and region. This means that it is possible that mothers and children who participated in April may have seroconverted before October. Additionally, the present study involved only participants from a specific cohort and geographic area, hence their clinical and socioeconomic characteristics might differ from the general Tanzanian population. Furthermore, it is a limitation to this study that neither male adults nor elderly were included. Therefore, our results should be interpreted with caution. Conclusion Our study shows that SARS-CoV-2 transmission in Tanzania is greater than previously reported, even in rural areas. Furthermore, dynamic peaks in viral transmission can be documented through seroprevalence, which could be an important epidemiological tool when assessing previous SARS-CoV-2 infection in rural and non-tested/unvaccinated settings without access to rapid molecular diagnostics. Finally, and importantly, we show differences in age, socioeconomic status and body composition between mothers and children who previously had SARS-CoV-2 infection. In this setting, continued surveillance is vital to monitor new circulating viral variants that significantly pose a great burden among vulnerable individuals, as well as to further investigate comorbidities and COVID-19 pathophysiology in populations like the one studied here. This should be implemented in tandem with the intensification of vaccination strategies. Contributions O.A.M, D.T.R.M., S.G., G.M., L.G.G., D.L.C., I.C.B., D.B., C.S., and L.H. designed the PONA2 follow-up, and O.A.M., L.P-A., D.T.R.M., C.B.H., P.G. and L.H. designed the COVID-19 study. O.A.M., D.T.R.M., S.G., G.M., J.M., C.S., and L.H. performed the clinical study, and LP-A., V.M.L.L., and E.C.S.B. performed the COVID-19 antibody measurements. O.A M., L.P-A., C.B.H. and L.H. analyzed the data, and O.A.M., L.P-A., D.T.R.M., C.B.H., D.B., C.S., P.G. and L.H. interpreted the results of the measurements. O.A.M. and L.H. wrote the manuscript, with contributions from all the co-authors who critically revised the manuscript and had access to the final version. Data access The datasets obtained and analysed in the current study are available from the last authors on reasonable request. REFERENCES Bayarri-Olmos, Rafael, Manja Idorn, Anne Rosbjerg, Laura Pérez-Alós, Cecilie Bo Hansen, Laust Bruun Johnsen, Charlotte Helgstrand, et al. 2021. “SARS-CoV-2 Neutralizing Antibody Responses towards Full-Length Spike Protein and the Receptor-Binding Domain.” Journal of Immunology (Baltimore, Md. : 1950) 207 (3): 878–87. https://doi.org/10.4049/jimmunol.2100272. Brand, Samuel P C, John Ojal, Rabia Aziza, Vincent Were, Emelda A Okiro, Ivy K Kombe, Caroline Mburu, et al. 2021. “COVID-19 Transmission Dynamics Underlying Epidemic Waves in Kenya.” Science (New York, N.Y.) 374 (6570): 989–94. https://doi.org/10.1126/science.abk0414. Cabore, Joseph Waogodo, Humphrey Cyprian Karamagi, Hillary Kipchumba Kipruto, Joseph Kyalo Mungatu, James Avoka Asamani, Benson Droti, Regina Titi-Ofei, et al. 2022. “COVID-19 in the 47 Countries of the WHO African Region: A Modelling Analysis of Past Trends and Future Patterns.” The Lancet. Global Health 10 (8): e1099–1114. https://doi.org/10.1016/S2214-109X(22)00233-9. Chechet, Gloria D, Jacob K P Kwaga, Joseph Yahaya, Harry Noyes, Annette MacLeod, and Walt E Adamson. 2022. “SARS-CoV-2 Seroprevalence at Urban and Rural Sites in Kaduna State, Nigeria, during October/November 2021, Immediately Prior to Detection of the Omicron Variant.” International Journal of Epidemiology, June. https://doi.org/10.1093/ije/dyac141. Cohen, Cheryl, Jackie Kleynhans, Anne von Gottberg, Meredith L McMorrow, Nicole Wolter, Jinal N Bhiman, Jocelyn Moyes, et al. 2022. “SARS-CoV-2 Incidence, Transmission, and Reinfection in a Rural and an Urban Setting: Results of the PHIRST-C Cohort Study, South Africa, 2020-21.” The Lancet. Infectious Diseases 22 (6): 821–34. https://doi.org/10.1016/S1473-3099(22)00069-X. Davies, Nicholas G, Petra Klepac, Yang Liu, Kiesha Prem, Mark Jit, CMMID COVID-19 working group, and Rosalind M Eggo. 2020. “Age-Dependent Effects in the Transmission and Control of COVID-19 Epidemics.” Nature Medicine 26 (8): 1205–11. https://doi.org/10.1038/s41591-020-0962-9. Espenhain, Laura, Siri Tribler, Charlotte Sværke Jørgensen, Christian Holm Hansen, Ute Wolff Sönksen, and Steen Ethelberg. 2021. “Prevalence of SARS-CoV-2 Antibodies in Denmark: Nationwide, Population-Based Seroepidemiological Study.” European Journal of Epidemiology 36 (7): 715–25. https://doi.org/10.1007/s10654-021-00796-8. Frasca, Daniela, Lisa Reidy, Carolyn Cray, Alain Diaz, Maria Romero, Kristin Kahl, and Bonnie B Blomberg. 2021. “Influence of Obesity on Serum Levels of SARS-CoV-2-Specific Antibodies in COVID-19 Patients.” PloS One 16 (3): e0245424. https://doi.org/10.1371/journal.pone.0245424. Grzelak, Ludivine, Aurélie Velay, Yoann Madec, Floriane Gallais, Isabelle Staropoli, Catherine Schmidt-Mutter, Marie-Josée Wendling, et al. 2021. “Sex Differences in the Evolution of Neutralizing Antibodies to Severe Acute Respiratory Syndrome Coronavirus 2.” The Journal of Infectious Diseases 224 (6): 983–88. https://doi.org/10.1093/infdis/jiab127. Hansen, Cecilie Bo, Ida Jarlhelt, Laura Pérez-Alós, Lone Hummelshøj Landsy, Mette Loftager, Anne Rosbjerg, Charlotte Helgstrand, et al. 2021. “SARS-CoV-2 Antibody Responses Are Correlated to Disease Severity in COVID-19 Convalescent Individuals.” Journal of Immunology (Baltimore, Md. : 1950) 206 (1): 109–17. https://doi.org/10.4049/jimmunol.2000898. Hjort, Line, Sofie Lykke Møller, Daniel Minja, Omari Msemo, Birgitte Bruun Nielsen, Dirk Lund Christensen, Thor Theander, et al. 2019. “FOETAL for NCD-FOetal Exposure and Epidemiological Transitions: The Role of Anaemia in Early Life for Non-Communicable Diseases in Later Life: A Prospective Preconception Study in Rural Tanzania.” BMJ Open 9 (5): e024861. https://doi.org/10.1136/bmjopen-2018-024861. Ho, Jim Q, Mohammad Reza Sepand, Banafsheh Bigdelou, Tala Shekarian, Rahim Esfandyarpour, Prashant Chauhan, Vahid Serpooshan, Lalit K Beura, Gregor Hutter, and Steven Zanganeh. 2022. “The Immune Response to COVID-19: Does Sex Matter?” Immunology 166 (4): 429–43. https://doi.org/10.1111/imm.13487. Johns Hopkins University. 2020. “Understanding the COVID-19 Pandemic: In Insights from Johns Hopkins University Experts.” 2020. https://coronavirus.jhu.edu/. Jones, Jefferson M, Jean D Opsomer, Mars Stone, Tina Benoit, Robyn A Ferg, Susan L Stramer, and Michael P Busch. 2022. “Updated US Infection- and Vaccine-Induced SARS-CoV-2 Seroprevalence Estimates Based on Blood Donations, July 2020-December 2021.” JAMA 328 (3): 298–301. https://doi.org/10.1001/jama.2022.9745. Kleynhans, Jackie, Stefano Tempia, Nicole Wolter, Anne von Gottberg, Jinal N Bhiman, Amelia Buys, Jocelyn Moyes, et al. 2022. “SARS-CoV-2 Seroprevalence after Third Wave of Infections, South Africa.” Emerging Infectious Diseases 28 (5): 1055–58. https://doi.org/10.3201/eid2805.220278. Levin, Andrew T, Nana Owusu-Boaitey, Sierra Pugh, Bailey K Fosdick, Anthony B Zwi, Anup Malani, Satej Soman, et al. 2022. “Assessing the Burden of COVID-19 in Developing Countries: Systematic Review, Meta-Analysis and Public Policy Implications.” BMJ Global Health 7 (5). https://doi.org/10.1136/bmjgh-2022-008477. Luo, Chunhua, Min Liu, Qianyuan Li, Xiaoling Zheng, Wen Ai, Feng Gong, Jinhong Fan, Shaowei Liu, Xi Wang, and Jun Luo. 2021. “Dynamic Changes and Prevalence of SARS-CoV-2 IgG/IgM Antibodies: Analysis of Multiple Factors.” International Journal of Infectious Diseases : IJID : Official Publication of the International Society for Infectious Diseases 108 (July): 57–62. https://doi.org/10.1016/j.ijid.2021.04.078. Lyimo, Eric, Cyrielle Fougeroux, Anangisye Malabeja, Joyce Mbwana, Paul M Hayuma, Edwin Liheluka, Louise Turner, et al. 2022. “Seroprevalence of SARS-CoV-2 Antibodies among Children and Adolescents Recruited in a Malariometric Survey in North-Eastern Tanzania July 2021.” BMC Infectious Diseases 22 (1): 846. https://doi.org/10.1186/s12879-022-07820-6. Mandolo, Jonathan, Jacquline Msefula, Marc Y R Henrion, Comfort Brown, Brewster Moyo, Aubrey Samon, Thandeka Moyo-Gwete, et al. 2021. “SARS-CoV-2 Exposure in Malawian Blood Donors: An Analysis of Seroprevalence and Variant Dynamics between January 2020 and July 2021.” BMC Medicine 19 (1): 303. https://doi.org/10.1186/s12916-021-02187-y. Mfinanga, Sayoki G, Nicholaus P Mnyambwa, Daniel T Minja, Nyanda Elias Ntinginya, Esther Ngadaya, Julie Makani, and Abel N Makubi. 2021. “Tanzania's Position on the COVID-19 Pandemic.” Lancet (London, England) 397 (10284): 1542–43. https://doi.org/10.1016/S0140-6736(21)00678-4. Mulenga, Lloyd B, Jonas Z Hines, Sombo Fwoloshi, Lameck Chirwa, Mpanji Siwingwa, Samuel Yingst, Adam Wolkon, et al. 2021. “Prevalence of SARS-CoV-2 in Six Districts in Zambia in July, 2020: A Cross-Sectional Cluster Sample Survey.” The Lancet. Global Health 9 (6): e773–81. https://doi.org/10.1016/S2214-109X(21)00053-X. Nuccetelli, Marzia, Massimo Pieri, Francesca Gisone, and Sergio Bernardini. 2022. “Combined Anti-SARS-CoV-2 IgA, IgG, and IgM Detection as a Better Strategy to Prevent Second Infection Spreading Waves.” Immunological Investigations 51 (2): 233–45. https://doi.org/10.1080/08820139.2020.1823407. Obande, Godwin Attah, Ahmad Ibrahim Bagudo, Suharni Mohamad, Zakuan Zainy Deris, Azian Harun, Chan Yean Yean, Ismail Aziah, and Kirnpal Kaur Banga Singh. 2021. “Current State of COVID-19 Pandemic in Africa: Lessons for Today and the Future.” International Journal of Environmental Research and Public Health 18 (19). https://doi.org/10.3390/ijerph18199968. Owino, Victor O. 2019. “Challenges and Opportunities to Tackle the Rising Prevalence of Diet-Related Non-Communicable Diseases in Africa.” The Proceedings of the Nutrition Society 78 (4): 506–12. https://doi.org/10.1017/S0029665118002823. Patterson, Amy S. 2022. “The Tanzanian State Response to COVID-19 (2022).” Pérez-Alós, Laura, Jose Juan Almagro Armenteros, Johannes Roth Madsen, Cecilie Bo Hansen, Ida Jarlhelt, Sebastian Rask Hamm, Line Dam Heftdal, et al. 2022. “Modeling of Waning Immunity after SARS-CoV-2 Vaccination and Influencing Factors.” Nature Communications 13 (1): 1614. https://doi.org/10.1038/s41467-022-29225-4. Poustchi, Hossein, Maryam Darvishian, Zahra Mohammadi, Amaneh Shayanrad, Alireza Delavari, Ayad Bahadorimonfared, Saeid Eslami, et al. 2021. “SARS-CoV-2 Antibody Seroprevalence in the General Population and High-Risk Occupational Groups across 18 Cities in Iran: A Population-Based Cross-Sectional Study.” The Lancet. Infectious Diseases 21 (4): 473–81. https://doi.org/10.1016/S1473-3099(20)30858-6. Sagara, Issaka, John Woodford, Mamady Kone, Mahamadoun Hamady Assadou, Abdoulaye Katile, Oumar Attaher, Amatigue Zeguime, et al. 2021. “Rapidly Increasing SARS-CoV-2 Seroprevalence and Limited Clinical Disease in Three Malian Communities: A Prospective Cohort Study.” MedRxiv : The Preprint Server for Health Sciences, April. https://doi.org/10.1101/2021.04.26.21256016. Salako, A O, O S Amoo, O O Odubela, K A Osuolale, A B James, D A Oladele, A Z Musa, et al. 2021. “Prevalence and Clinical Characteristics of Coronavirus Disease 2019 Seen at a Testing Centre in Lagos Nigeria.” West African Journal of Medicine 38 (1): 54–58. Salum, Salum Seif, Mohammed Ali Sheikh, Antje Hebestreit, and Sørge Kelm. 2022. “Anti SARS-CoV-2 Seroprevalence in Zanzibar in 2021 before the Omicron Wave.” IJID Regions (Online) 4 (September): 120–22. https://doi.org/10.1016/j.ijregi.2022.06.007. Schmiegelow, Christentze, Thomas Scheike, Mayke Oesterholt, Daniel Minja, Caroline Pehrson, Pamela Magistrado, Martha Lemnge, et al. 2012. “Development of a Fetal Weight Chart Using Serial Trans-Abdominal Ultrasound in an East African Population: A Longitudinal Observational Study.” PloS One 7 (9): e44773. https://doi.org/10.1371/journal.pone.0044773. Shanaube, K, A Schaap, E Klinkenberg, S Floyd, J Bwalya, M Cheeba, P de Haas, et al. 2022. “SARS-CoV-2 Seroprevalence and Associated Risk Factors in Periurban Zambia: A Population-Based Study.” International Journal of Infectious Diseases : IJID : Official Publication of the International Society for Infectious Diseases 118 (May): 256–63. https://doi.org/10.1016/j.ijid.2022.03.021. Shiri, Tinevimbo, Marc Evans, Carla A Talarico, Angharad R Morgan, Maaz Mussad, Philip O Buck, Phil McEwan, and William David Strain. 2021. “Vaccinating Adolescents and Children Significantly Reduces COVID-19 Morbidity and Mortality across All Ages: A Population-Based Modeling Study Using the UK as an Example.” Vaccines 9 (10). https://doi.org/10.3390/vaccines9101180. Simeni Njonnou, Sylvain Raoul, Nadia Christelle Noumedem Anangmo, Fernando Kemta Lekpa, Diomede Noukeu Njinkui, Dominique Enyama, Christian Ngongang Ouankou, Eric Vounsia Balti, Esther Astrid Mbono Samba Eloumba, Jean Roger Moulion Tapouh, and Simeon Pierre Choukem. 2021. “The COVID-19 Prevalence among Children: Hypotheses for Low Infection Rate and Few Severe Forms among This Age Group in Sub-Saharan Africa.” Interdisciplinary Perspectives on Infectious Diseases 2021: 4258414. https://doi.org/10.1155/2021/4258414. Sorg, Anna-Lisa, Leon Bergfeld, Marietta Jank, Victor Corman, Ilia Semmler, Anna Goertz, Andreas Beyerlein, et al. 2022. “Cross-Sectional Seroprevalence Surveys of SARS-CoV-2 Antibodies in Children in Germany, June 2020 to May 2021.” Nature Communications 13 (1): 3128. https://doi.org/10.1038/s41467-022-30482-6. Takahashi, Takehiro, Mallory K Ellingson, Patrick Wong, Benjamin Israelow, Carolina Lucas, Jon Klein, Julio Silva, et al. 2020. “Sex Differences in Immune Responses That Underlie COVID-19 Disease Outcomes.” Nature 588 (7837): 315–20. https://doi.org/10.1038/s41586-020-2700-3. Tarimo, Clifford Silver, and Jian Wu. 2020. “The First Confirmed Case of COVID-19 in Tanzania: Recommendations Based on Lesson Learned from China.” Tropical Medicine and Health 48: 25. https://doi.org/10.1186/s41182-020-00214-x. The World Bank. 2020. “Tanzania Economic Update : Addressing the Impact of COVID-19.” Tripathi, Hemant G., Harriet E. Smith, Steven M. Sait, Susannah M. Sallu, Stephen Whitfield, Astrid Jankielsohn, William E. Kunin, Ndumiso Mazibuko, and Bonani Nyhodo. 2021. “Impacts of COVID-19 on Diverse Farm Systems in Tanzania and South Africa.” Sustainability 13 (17): 9863. https://doi.org/10.3390/su13179863. USAID. 2021. “Tanzania Overview 2021.” 2021. https://www.usaid.gov/tanzania/newsroom/fact-sheets. Uyoga, Sophie, Ifedayo M O Adetifa, Henry K Karanja, James Nyagwange, James Tuju, Perpetual Wanjiku, Rashid Aman, et al. 2021. “Seroprevalence of Anti-SARS-CoV-2 IgG Antibodies in Kenyan Blood Donors.” Science (New York, N.Y.) 371 (6524): 79–82. https://doi.org/10.1126/science.abe1916. WHO, The World health Organization. 2022. “WHO Coronavirus (COVID-19) Dashboard.” 2022. https://covid19.who.int/. Wu, Zunyou, and Jennifer M McGoogan. 2020. “Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for Disease Control and Prevention.” JAMA 323 (13): 1239–42. https://doi.org/10.1001/jama.2020.2648. Zinszer, Kate, Britt McKinnon, Noémie Bourque, Laura Pierce, Adrien Saucier, Alexandra Otis, Islem Cheriet, et al. 2021. “Seroprevalence of SARS-CoV-2 Antibodies Among Children in School and Day Care in Montreal, Canada.” JAMA Network Open 4 (11): e2135975. https://doi.org/10.1001/jamanetworkopen.2021.35975. Appendix Supplementary materials Image, application 1 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. Funding The PONA2 study was funded by the “Læge Sofus Carl Emil Friis og Hustru Olga Doris Friis” Foundation, and the Augustinus Foundation. L.H. is funded by the Danish Diabetes Academy supported by the Novo Nordisk Foundation, and the Danish Diabetes Association (Diabetesforeningen). P.G. is supported by funds from the Carlsberg Foundation (CF20-0045), the Novo Nordisk Foundation (NFF205A0063505 and NNF20SA0064201) and The Svend Andersen Research Foundation (SARF2021). C.S. is funded by the Danish Independent Research Fund: Clinician Scientist Positions, Medical Sciences. D.B. is supported by a National Health and Medical Research Council (Australia) Investigator Grant. D.L.C. is supported by grants from the Danish International Development Agency (DANIDA No. 17-03-KU and DANIDA No. 19-M06-KU). The Novo Nordisk Foundation Center for Basic Metabolic Research is an independent research center at the University of Copenhagen, partially funded by an unrestricted donation from the Novo Nordisk Foundation (NNF18CC0034900). Acknowledgements First and foremost, we greatly appreciate the participation of all the women and the children in the study, and all committed health care workers assisting in the care for the women and their children. We are also thankful to the PONA2 study team members in Tanzania including Jaqueline Kichungo, Neema Malle, Regina Malugu, Celina Mzava, Eva Rimoy, Mohamed Mapondela, Rashid Mtumba, Walter Maranga, Emmanuel Kessy, Zeno Manjulungu, Gerson Maro, Francis Mkongo, Simba Athumani, Claud Tesha, and Humphrey Mathew, for their tremendous work efforts in recruitment, collecting data, and generating the biobank and database. We are thankful for the collaborations with the NIMR-Tanga Centre, Joint Malaria Programme, and the administration at Korogwe District hospital and all government staff working within KDH and satellite dispensaries. Finally, the authors would like to thank Mads Engelhardt Knudsen and Sif Kaas Nielsen from the Department of Clinical Immunology at Rigshospitalet for their excellent technical assistance. Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ijregi.2022.11.011.	36466214	PMC9709504	NO-CC CODE	2022-12-16 23:20:06	no	IJID Reg. 2023 Mar 28; 6:48-57	utf-8	IJID Reg	2,022	10.1016/j.ijregi.2022.11.011	oa_other
==== Front Eur Neuropsychopharmacol Eur Neuropsychopharmacol European Neuropsychopharmacology 0924-977X 1873-7862 Elsevier B.V. and ECNP. S0924-977X(22)00891-4 10.1016/j.euroneuro.2022.11.001 Corrigendum Corrigendum to ‘Suicide-related thoughts and behavior and suicide death trends during the COVID-19 in the general population of Catalonia, Spain’ [European Neuropsychopharmacology 56 (2021) 4–12] Pérez V. abc Elices M. bc# Vilagut G. bd Vieta E. ce Blanch J. ef Laborda-Serrano E. g Prat B. f Colom F. abc Palao D. h Alonso J. bdi a Institut de Neuropsiquiatria i Addiccions (INAD), Parc de Salut Mar, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain b Institut Hospital del Mar d'Investigacions Mèdiques, (IMIM), Parce de Salut Mar, Barcelona, Spain c Centro de Investigación Biomédica en Red de Salud Mental, CIBERSAM, Madrid, Spain d CIBER Epidemiología y Salud Pública (CIBERESP), Madrid, Spain e Hospital Clinic, Institute of Neuroscience, University of Barcelona, IDIBAPS, Barcelona, Spain f Master Plan on Mental Health and Addictions, Ministry of Health, Catalan Government, Spain g Grupo Regional de Atención a la Víctima (RPMN), Mossos d'esquadra h Department of Mental Health, Parc Taulí-University Hospital; Unitat Mixta de Neurociència Traslacional I3PT, Department of Psychiatry and Forensic Medicine, Autonomous University of Barcelona; CIBERSAM, Spain i Department of Experimental and Health Sciences, Pompeu Fabra University, Barcelona, Spain # Correspondence author. 30 11 2022 1 2023 30 11 2022 66 6666 © 2022 Elsevier B.V. and ECNP. All rights reserved. 2022 Elsevier B.V. and European College of Neuropsychopharmacology 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 would like to apologise for any inconvenience caused. Corrigendum to Role of the Funding Source: The authors regret that the printed version of the above article omitted the following important funding source: “This project was supported by Instituto de Salud Carlos III, Ministerio de Ciencia e Innovación/FEDER (grant ISCIII/FEDER PI17/00521)” which should have been included.	36462446	PMC9709522	NO-CC CODE	2022-12-01 23:23:07	no	Eur Neuropsychopharmacol. 2023 Jan 30; 66:66	utf-8	Eur Neuropsychopharmacol	2,022	10.1016/j.euroneuro.2022.11.001	oa_other
==== Front Int Immunopharmacol Int Immunopharmacol International Immunopharmacology 1567-5769 1878-1705 Elsevier B.V. S1567-5769(22)00963-8 10.1016/j.intimp.2022.109478 109478 Article IFIH1/IRF1/STAT1 promotes sepsis associated inflammatory lung injury via activating macrophage M1 polarization Wang Ailing ab Kang Xueli a Wang Jing a Zhang Shi ac⁎ a Department of Pulmonary and Critical Care Medicine, Central Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, China b Department of Ultrasound, Central Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, China c Jiangsu Provincial Key Laboratory of Critical Care Medicine, Department of Critical Care Medicine, Zhongda Hospital, School of Medicine, Southeast University, Nanjing, China ⁎ Corresponding author. 30 11 2022 1 2023 30 11 2022 114 109478109478 24 9 2022 12 11 2022 16 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. Background A growing body of research has shown that the phenotypic change in macrophages from M0 to M1 is essential for the start of the inflammatory process in septic acute respiratory distress syndrome (ARDS). Potential treatment targets might be identified with more knowledge of the molecular regulation of M1 macrophages in septic ARDS. Methods A multi-microarray interrelated analysis of high-throughput experiments from ARDS patients and macrophage polarization was conducted to identify the hub genes associated with macrophage M1 polarization and septic ARDS. Lipopolysaccharide (LPS) and Poly (I:C) were utilized to stimulate bone marrow-derived macrophages (BMDMs) for M1-polarized macrophage model construction. Knock down of the hub genes on BMDMs via shRNAs was used to screen the genes regulating macrophage M1 polarization in vitro. The cecal ligation and puncture (CLP) mouse model was constructed in knockout (KO) mice and wild-type (WT) mice to explore whether the screened genes regulate macrophage M1 polarization in septic ARDS in vivo. ChIP-seq and further experiments on BMDMs were performed to investigate the molecular mechanism. Results The bioinformatics analysis of gene expression profiles from a clinical cohort of 26 ARDS patients and macrophage polarization found that the 5 hub genes (IFIH1, IRF1, STAT1, IFIT3, GBP1) may have a synergistic effect on macrophage M1 polarization in septic ARDS. Further in vivo investigations indicated that IFIH1, STAT1 and IRF1 contribute to macrophage M1 polarization. The histological evaluation and immunohistochemistry of the lungs from the IRF1-/- and WT mice indicated that knockout of IRF1 markedly alleviated CLP-induced lung injury and M1-polarized infiltration. Moreover, the molecular mechanism investigations indicated that knockdown of IFIH1 markedly promoted IRF1 translocation into the nucleus. Knockout of IRF1 significantly decreases the expression of STAT1. ChIP-seq and PCR further confirmed that IRF1, as a transcription factor of STAT1, binds to the promoter region of STAT1. Conclusion IRF1 was identified as the key molecule that regulates macrophage M1polarization and septic ARDS development in vivo and in vitro. Moreover, as the adaptor in response to infection mimics irritants, IFIH1 promotes IRF1 (transcription factor) translocation into the nucleus to initiate STAT1 transcription. Keywords ARDS macrophage M1 polarization IFIH1 IRF1 STAT1 Abbreviations ARDS, acute respiratory distress syndrome LPS, Lipopolysaccharide BMDMs, bone marrow derived macrophages KO, knocking out CLP, cecal ligation perforation WT, wide-type ChIP, chromatin immunoprecipitation seq, sequence ICU, intensive care unit FDR, false discovery rate SD, standard deviation ANOVA, one-way analysis of variance IFIH1, IRF1, Interferon-Induced Helicase C Domain-Containing Protein 1 IRF3, Interferon-Induced Helicase C Domain-Containing Protein 3 STAT1, signal transducer and activator of transcription 1 IFIT3, interferon-induced protein with tetratricopeptide repeats 3 GBP1, guanylate binding protein 1 S, supplementary GEO, Gene Expression Omnibus WGCNA, Weighted correlation network analysis MEs, module eigengenes PPI, Protein-protein interaction IRF1-/-, IRF1 knockout mice IL-6, Interleukin-6 CCL2, Chemokine ligand 2 IL-1, Interleukin-1 WB, Western blot GAPDH, glyceraldehyde-3-phosphate dehydrogenase ELISA, Enzyme-linked immunosorbent assay RT-PCR, Reverse Transcription-Polymerase Chain Reaction IFN, Interferon TF, transcription factor TLR4, toll like receptor 4 ERK, extracellular regulated protein kinases NFκB, nuclear factor kappa-B ==== Body pmc1 Background A common and severe respiratory disease, acute respiratory distress syndrome (ARDS), accounts for 10 % of all intensive care unit (ICU) hospitalizations and has a 40 % mortality rate [1]. The current COVID-19 pandemic has significantly increased the mortality rate of ARDS in comorbid cases, reaching 93 % [2]. A total of 77.5 % of individuals with ARDS also have sepsis or an infection [3]. The onset of ARDS activates immune cells, leading to uncontrolled and persistent sepsis-induced multi-organ dysfunction [4]. A better understanding of the sepsis-induced pathophysiological mechanism that causes ARDS may allow the development of new treatment options [5]. Sepsis-induced inflammatory lung damage begins with the polarization of macrophages to the M1 phenotype [6], [7]. The molecular control of macrophage polarization during the development of ARDS remains largely unknown. Our previous studies [8] showed that 5 hub genes (interferon-induced helicase C domain-containing protein 1, IFIH1; interferon-induced helicase C domain-containing protein 1, IRF1; signal transducer and activator of transcription 1,STAT1; interferon-induced protein with tetratricopeptide repeats 3, IFIT3; and guanylate binding protein 1, GBP1) may exert synergistic effects in macrophage M1 polarization and in the progression of sepsis-induced ARDS. Previous studies [9], [10], [11], [12] have confirmed that IFIH1 and STAT1 regulate macrophage M1 polarization and sepsis-induced ARDS. However, it remains unknown whether IRF1, GBP1 and IFIT3 are involved in macrophage M1 polarization and sepsis-induced ARDS. IRF1, a member of the interferon transcription factor family, coordinates with IRF3 to control the transcription of interferon, which act as antiviral agents. Recent studies have demonstrated that IRF1 simultaneously regulates inflammation, but the precise mechanism remains unknown. Moreover, our earlier research verified that IFIH1 controls macrophage M1 polarization and fosters inflammation by regulating the translocation of IRF3 into the nucleus. According to our bioinformatics analysis, IFIH1, IRF1, STAT1, GBP1, and IFIT3 may be involved in a molecular process that cooperatively controls macrophage M1 polarization. However, elucidation of the detailed mechanisms requires further experimental exploration and validation. To investigate these mechanisms, we utilized shRNA to knock down the expression of IRF1, IFIT3 and GBP1 to evaluate whether these genes are involved in macrophage M1 polarization. Knockout and rescue experiments were conducted for validation both in vivo and in vitro. In addition, molecular experiments were performed to uncover the mechanism linking these hub genes. 2 Methods 2.1 ARDS patients Sepsis and ARDS were defined according to the Sepsis 3.0 criteria and 2012 Berlin definition [13], [14]. Sepsis-induced ARDS was named septic ARDS. The revised Helsinki Declaration was followed when conducting the current study. The Critical Care Centre at Zhongda Hospital's biological specimen bank provided the samples from the individuals who were included. The Institutional Ethics Committee of Zhongda Hospital approved and oversaw the establishment of this specimen bank in 2017. (Registration number: 2017ZDSYLL105). Figure Supplementary (S) 1 displayed the Ethical documents. As is descript in our previous study, to screen hub genes involved in ARDS and M1-polarized macrophages, we conducted interrelated bioinformatics analysis of mRNA matrix from 26 septic ARDS patients and gene expression profiles of macrophages. The mRNA matrix from septic ARDS patients was built from the biological specimen bank of the Critical Care, Zhongda Hospital. These gene expression profiles of macrophages were screened from the public Gene Expression Omnibus (GEO) database. Whole blood was drawn from 26 ARDS patients to evaluate all gene expression patterns using the Human mRNA Microarray V4.0 (Arraystar) chip in order to identify potential candidate genes linked to the severity of ARDS. At the time of ICU triage for this trial, critically ill patients admitted via the emergency department were included. If a patient satisfied the requirements for having ARDS within 24 h of enrolling in the trial, they were considered to have the condition. The following conditions qualified as exclusions: any prior history of cancer, immune or hematological disorders, or treatments including chemotherapeutic drugs or steroids within six months of hospitalization. Within 24 h after being admitted to the ICU, whole blood was collected for the isolation of RNA. Weighted correlation network analysis (WGCNA) was carried out on all gene expression profiles from the 26 ARDS patients to explore potential genes related to the severity of ARDS [15]. Based on co-expression associations, the WGCNA R program was used to categorize all expressed genes in microarrays into different module eigengenes (MEs). The severity of ARDS was then compared with the MEs using Spearman's correlation that had been adjusted for clusters. For further investigation, the genes from modules that had the highest correlation coefficient and the P value < 0.05 with the severity of ARDS were chosen. 2.2 Screen macrophage M1 polarization related genes We performed a secondary analysis on the GEO public database (GSE46903, human alveolar macrophages from 125 volunteers) to identify significantly differentially expressed genes between M1 macrophages and M0 and M2 macrophages (fold difference > 2; false discovery rate, FDR < 0.05), in order to investigate the genes potentially involved in macrophage M1 polarization [16]. We used the edgeR R package to filter for differentially expressed molecules based on RNA sequencing data and the Limma R tool to find genes that were differently expressed based on mRNA microarray data [17], [18]. The method for differentially expressed analysis uses moderated t-tests to construct an empirical Bayesian technique to assess changes in gene expression. 2.3 Screen hub genes A Venn diagram was also drawn to show the overlap between the genes linked to ARDS severity and macrophage M1 in order to find the possible genes shared by both ARDS and these cells. We used connection degree analysis to perform hub gene analysis to identify the critical genes (number of neighbours). First, we used the STRING website (https://string-db.org/cgi/input.pl) to plot the protein–protein interaction (PPI) network [19]. Second, R was used to determine the overall connection degree of each network node in order to identify the genes with the greatest connectivity degrees. Hub genes were defined as those located above the first connection degree inflection point [20]. 2.4 Sepsis induced ARDS model construction Animal research in the current study were evaluated and authorized by the Jinan Central Hospital's Animal Care and Use Committee. The Ethical approval document was shown in Figure S2. The laboratory animal facility donated C57BL/6 mice (male, 6–8 weeks old) for use in this investigation (Jinan, China). There were 3 groups, and each group contained 6 mice. CLP induced septic ARDS: A septic ARDS model was created using the cecal ligation perforation (CLP) method. The mice underwent lower abdomen hair removal, intraperitoneal administration of 50 mg/kg pentobarbital, then cleaning with 75 percent ethanol. An incision is made along the midline of the abdomen to expose the cecum and prevent vascular damage. A 22-gauge needle was used to puncture the cecum, which was then ligated with silk suture 1 cm from its apex. Fecal material was then squeezed out of the puncture. The abdominal incision was then two-layered, 4–0 silk sutured closed. Finally, to help the mouse recover from anesthesia, its back was gently put on a warm blanket. CLP-Sham mice:The sham group's cecum was neither ligated or perforated. Finally, to help the mouse recover from anesthesia, its back was gently put on a warm blanket. 2.5 IRF1 -/- mice The IRF1 knockout (KO) mice (IRF1-/-) were purchased from the Gempharmatech Biosciences (China). The primers were used to test the knockout of IRF1. 5′- GTCCTTGACCTAAGCCCCAT −3′ (Forward), 5′-GCCAGACTCGGGATAAAACTAC G-3′ (Reverse), fragment size of 386 bp; Purified amplified products underwent DNA sequencing analysis. The mice utilized for breeding and the following investigations were homozygous IRF1-/- animals. From mouse tail tissue, genomic DNA was isolated, and PCR (Polymerase Chain Reaction) was used to pinpoint particular bands. BMDM were obtained from littermate control and IRF1-/- mice, and portion of the BMDM from IRF1-/- mice were transfected with IRF1 over-expressed plasmid for rescue experiment. Western blot were then used to confirm IRF1 expression in macrophages of each group. (For verification findings, see Figure S5). 2.6 Bone marrow isolation and Macrophage culture Bone marrow-derived macrophages (BMDMs) were generated largely in accordance with other reports. BM cells were taken from the femur and tibia's medullary cavities on an incredibly clean bench. The erythrocytes were lysed using lysing buffer (BD Pharm LyseTM, USA), washed three times in phosphate-buffered saline (PBS), and then cultivated for seven days at 37 °C in a humidified 5 percent CO2 sterile incubator in fresh DMEM with 10 percent FBS and 20 ng/ml M−CSF. F4/80, which is used to identify BMDMs, was found using flow cytometry. 2.7 Cell culture and reagent treatment The BMDMs were cultured in Dulbecco's modified medium (DMEM; Wisent Biotechnology, Nanjing, China) containing 10 % foetal bovine serum (FBS; Coring, Australia), 100 IU/ml penicillin, and 100 g/ml Poly(I:C) have undergone strong demonstrations showing that they are the activators of RIG-I and IFIH1. Lipopolysaccharide (LPS) has also been demonstrated time and time again to be the traditional activator for M1 macrophage polarization. Inflammatory cell models for ARDS in animals are frequently created using bacterial lipopolysaccharides, which have been solidly established as important pathogenic components of the disease. In the current investigation, Poly(I:C) and LPS were utilized as M1 macrophage polarization stimulators taking these aspects into account. According to our prior research, the concentrations of LPS (500 ng/mL) and Poly(I:C) (50 ng/mL) were determined. 2.8 Overexpression of IRF1 The full-length coding sequence of IRF1 (NM 001164477.1, 2931 bp) was first amplified by PCR. The primer: Xhol-IRF1-F: ATACTCGAGCGATGCCAATCACTCGAATGCGGATGA. Notl-IRF1-R: ATAGCGGCCGCTCATCCGCATTCGAGTGATTGGCAT. The empty vector served as treatment controls. The entire production process was examined in our previous work. The IRF1 over-expression vector and an empty vector were transfected into several BMDMs. Western blots were used to evaluate how over-expression affected the results. 2.9 Knocking down IFIH1/IRF1/STAT1/GBP1/IFIT3 Three different sequences were developed specifically for mouse IFIH1, IRF1, STAT1, GBP1, and IFIT3, according to GeneChem Co., ltd. (https://www.genechem.com.cn; Table S1-S5). BMDMs were transfected using lentivirus supernatant (infection multiplicity = 50). Three days after transfection, we utilized a western blot to evaluate which shRNA was most successful in down-regulating specific genes. The sequences of the most efficient shRNAs are finally shown in bold italics in Table S1–S5. The control sequence was designated as shCtrl and was TTCTCCGAACGTGTGTCACGTT. The knock down effectiveness was evaluated by Western blot, as seen in Figure S4. 2.10 Evaluation of lung histopathology The right upper lobe was sectioned sagittally into five 5-meter thick sections after being preserved in paraffin. The sections were stained with hematoxylin and eosin. Edema, alveolar and interstitial inflammation and bleeding, atelectasis, necrosis, and the development of the hyaline membrane were all rated on a scale from 0 to 4. The sum of the scores was used to calculate the extent of lung injury, as was previously mentioned. The ratio of lung wet weight to body weight (LWW/BW) in each group was determined to reflect the severity of the pulmonary vascular permeability and pulmonary oedema. 2.11 Flow cytometry Cultured BMDMs suspension was re-suspended in PBS, incubated for 15 min with FcR blocking reagent, and then incubated for 15 min with an APC-conjugated anti-mouse CD86 (1:200) or FITC-conjugated anti-mouse F4/80 (1:200) antibody, as instructed by the manufacturer, for phenotypic analysis of cell surface marker expression. Macrophages were recognized by the F4/80 macrophage marker, whereas M1 macrophages were recognized by the CD86 M1 macrophage marker. The level of CD86 expression was calculated using fluorescence intensity. The labeled cells were washed twice, resuspended in cold buffer, and then the data were analyzed using flow cytometry (ACEA NovoCyte, China) and FlowJo software version X. (Tree Star, USA). CD86: Abcam, Mouse anti-Rat mAb #ab218757. F4/80: Abcam, Mouse anti-Rat mAb #ab60343. 2.12 Immunohistochemical staining Slices of lung were moistened and deparaffinized. Slides were blocked with 5 percent goat serum for an hour after antigen extraction at a high temperature. The sections were incubated with primary antibodies overnight at 4 °C after blocking (diluted 1:100). The primary antibodies (Abcam, Mouse anti-Rat mAb #ab218757) were directed against CD86. After three PBS rinses, the slices were then subjected to a 1:50 dilution of a biotinylated secondary antibody. Hematoxylin was used to counterstain the reaction products after diaminobenzidine (DAB, China) incubation. The positive areas were quantified using Image J. All of the photographs were taken using a light microscope at a high magnification (4 0 0). (Olympus). 2.13 ELISAa Interleukin-6 (IL-6), chemokine ligand 2 (CCL2), and interleukin-1β (IL-1β) concentrations in macrophage culture supernatant and bronchoalveolar lavage fluid (BALF) were measured were measured by Enzyme-linked immunosorbent assay (ELISA). IL-6: Mouse IL-6 ELISA Kit (ab222503), Abcam (England). CCL2: Mouse ELISA Kit (MJE00B), R&D Systems (The United States). IL-1β: Mouse IL-1 beta ELISA Kit (ab197742), Abcam (England). 2.14 Western blot (WB) analysis Proteins were separated using sodium dodecyl sulfate–polyacrylamide gel electrophoresis and then transferred to PVDF membranes. The membranes were treated with primary antibodies overnight at 4 °C against IFIH1 (1:1000), iNOS (1:1000), IRF1 (1:1000), STAT1 (1:1000), IFIT3 (1:1000), GBP1 (1:1000), GAPDH (1:1000), and -Tubulin (1:1000). The secondary antibody was applied to the membranes and left on them for an hour at room temperature. To visualize immunoblots, enhanced chemiluminescence was utilized (ECL; Thermo Scientific). To make the expression levels of the entire cell extract normal, the expression levels of -Tubulin were employed. iNOS: Abcam, rabbit mAb #ab178945. β-Tubulin: Abcam, rabbit mAb #ab108342. IRF1: Abcam, rabbit mAb #ab245338. STAT1: Abcam, rabbit mAb #ab92506. IFIH1: Cell Signaling Technology, rabbit mAb #5321. IFIT3: Invitrogen, rabbit pAb # PA5-22230. GBP1: Abcam, rabbit mAb # ab119236. GAPDH: Abcam, rabbit mAb # ab181602. 2.15 PCRa TRIzol was used to extract the total RNA from lung tissue samples or cells. For reverse transcription of RNA, Takara, Japan's Prime ScriptTM Trimester Mix was used. According to the manufacturer's recommendations, Reverse Transcription-Polymerase Chain Reaction (RT-PCR) was carried out using a Step One Plus RT-PCR equipment (Life Technologies, USA) and SYBR Premix Ex TaqTM11 (Takara, Japan). Reverse transcription of RNA was carried out using Trimester Mix (Takara, Japan). According to the manufacturer's recommendations, RT-PCR was carried out using a Step One Plus RT-PCR equipment (Life Technologies, USA) and SYBR Premix Ex TaqTM11 (Takara, Japan). The genomic DNA fragments from transgenic mice's tail snips were identified using agarose gel electrophoresis. With the use of the EcoRV restriction enzyme and 1.0 percent agarose gel, DNA samples taken from the transgenic and parental lines were broken down. After that, a flask containing 0.20 g of agarose and 20 ml of 1X Tris-acetate-EDTA buffer was cooked in a microwave for 5 min at 100 °C. Thermo Fisher Scientific, Inc.) was then added, and 2 l of ethidium bromide was poured onto a taped plate with casting combs. Then, 2 µl Once separation was accomplished, samples of mouse tail DNA or macrophage DNA were added to the 5X agarose gel and subjected to electrophoresis at 120 mA for 40 min at 25 °C. The Bio-Rad gel imager was used to see the DNA fragments. 2.16 ChIP-seq and ChIP-PCR ChIP-seq (chromatin immunoprecipitation sequence) was utilized to investigate whether IRF1 binds to promoter region of STAT1, and ChIP-PCR was conducted to validate the results analyzed by CHIP-seq. By centrifuging BMDMs, crude nuclear pellets were extracted, resuspended in lysis buffer, and incubated on ice for ten minutes. To create chromatin fragments of around 200–400 bp in length, the chromatin was sonicated at 4 °C using a Bioruptor 300 at the maximum setting for fifteen 1-minute cycles of 30 s on and 30 s off. The soluble chromatin was precleared using protein A agarose beads and diluted 1:10 with dilution buffer. Antibodies against IRF1 were incubated overnight at 4 °C with the precleared supernatant. After washing, 100 mM NaCl was added, and the immunoprecipitated material was allowed to sit at 65 °C for 12 h before being eluted. The enriched genomic DNA was extracted using phenol, chloroform and isoamyl alcohol, followed by ethanol precipitation, after protein and RNA had been removed. ChIP-seq and ChIP-PCR were used to analyze the immunoprecipitated DNA after it had been dissolved in water. The following were the ChIP-PCR primers: GCAGTGAGTGAGTGAGAG (Forward). AGTGAGAACGGCAGGATA (Reverse). 2.17 Statistical analysis In vivo and in vitro experiments in each group used six and three mice, respectively. Data are expressed as the mean and standard deviation (SD) of repeated trials. One-way analysis of variance (ANOVA) was used to compare among groups. The cut-off value for statistical significance was set at P < 0.05. In these multi-group comparisons, the Tukey’s style post-hoc pairwise test were utilized to analysis the high-throughput experiments and a false discovery rate (FDR)-adjusted P < 0.05 was set as the cut-off [18], [21]. These high-throughput experiments included ChIP-seq data and differentially expressed analysis on mRNA microarray data of macrophage polarization (GSE46903). 3 Results 3.1 Three hub gene molecules may play a synergistic role in macrophage M1 polarization The interrelated analysis was performed on high-throughput experimental data from peripheral blood samples of 26 ARDS patients (clinical information shown in Table S6) and in vitro human alveolar macrophages from 125 volunteers (GEO public database, GSE46903). The analysis indicated the involvement of the hub genes STAT1, IFIH1, GBP1, IFIT3, and IRF1 (Figure S3), consistent with the results of our previous studies [8]. Prior research has established that STAT1 and IFIH1 are essential for macrophage M1 polarization and inflammation in the lung [8], [9], [10], [11], [12], [13], [22], [23], [24], [25]. To further verify whether GBP1, IFIT3, and IRF1 influence macrophage M1 polarization, shRNA sequences for each gene were added to create transfected BMDMs with GBP1, IFIT3, and IRF1 knockdown. These cells were then subjected to LPS induction for 24 h. The shRNA sequence was able to silence GBP1, IFIT3, and IRF1in BMDMs, as demonstrated in FigureS4. In contrast to control treatment (control shRNA transfection), Western blot analysis showed that silencing IRF1 significantly reduced LPS-induced iNOS production in BMDMs (P < 0.05, Fig. 1 ). However, GBP1 and IFIT3 knockdown did not influence LPS-induced iNOS expression in BMDMs compared with controls (P > 0.05, Fig. 1). In addition, the vector transfection of shRNAs and controls did not influence iNOS expression.Fig. 1 IFIH1, IRF1 and STAT1 were identified as regulators of macrophage M1 polarization in ARDS. Three mice were included in each group. (A/B/C) In BMDMs, transfected BMDMs with shGBP1, shIFIT3, and shIRF1, as well as control BMDMs with shRNA, the expression of iNOS was detected by Western blotting. After IRF1 silencing, LPS-induced iNOS expression in BMDMs was significantly reduced, according to the quantitative analysis of western blotting (P < 0.05). GBP1 and IFIT3 knockdown had no effect on iNOS expression in BMDMs (P > 0.05). Additionally, neither the vector transfection of shIFIH1 nor the control shRNA had an impact on the expression of iNOS in BMDMs. The statistical analysis was from three independent experiments, and the bar indicates the SD values. These results suggest that IFIH1, IRF1 and STAT1 may contribute to macrophage M1 polarization synergistically. 3.2 IRF1 contributes to macrophage M1 polarization To provide solid proof that IRF1 contributes to macrophage M1 polarization, IRF1 deletion, rescue, and over-expression experiments were conducted. Western blots revealed that LPS-induced iNOS expression in BMDMs from IRF1-/- mice was considerably reduced compared to that in BMDMs from WT mice (P < 0.05, Fig. 2 A). Flow cytometry revealed that CD86 expression was significantly lower after IRF1 knockout (P < 0.05, Fig. 2B). Furthermore, vector transfection had no effect on iNOS or CD86 expression (P > 0.05). ELISA further indicated that the inflammatory cytokine levels of IL1β, IL6, and CCL2 were significantly decreased in culture medium after deleting IRF1 (P < 0.05, Fig. 2C/D/E).Fig. 2 IRF1 contributes to macrophage M1 polarization. Three mice were included in each group. (A) Western blotting was performed to determine the expression of iNOS on BMDMs with IRF1 knocking out, rescuing and over-expressing. GAPDH was used as the standard for verifying equivalent loading (n = 3 for each group). (B) Flow cytometric analysis was performed to analyze the MFI of CD86 on each group of BMDMs (n = 3 for each group). (C) ELISA detected the expression of IL1β, IL6 and CCL2 in culture medium from each BMDMs (n = 3 for each group). The statistical analysis was from three independent experiments, and the bar indicates the SD values. Rescue experiments showed that BMDMs from the IRF1-/- mice were transfected with IRF1 high expression plasmid, and the expression of iNOS and CD86 increased again after LPS stimulation. In the above rescue experiment, the pro-inflammatory factors IL-1β, IL-6 and CCL2 also showed similar changes (Fig. 2). Compared with control blank vectors, the M1-polarized markers (iNOS, CD86, L-1β, IL-6 and CCL2) were markedly increased after over-expression of IRF1in BMDMs from the WT mice (P < 0.05, Fig. 2). These results validated that IRF1 contributes to LPS-induced macrophage M1 polarization. 3.3 IRF1 plays critical roles in ARDS development We established a CLP-induced septic ARDS model and knocked out the IRF1 gene in mice to further confirm the involvement of IRF1 in the development of ARDS. We next evaluated the effect of IRF1 on lung injury. We used a histological evaluation of the lungs to confirm the effect of IRF1 on CLP-induced lung injury in mice. Compared with the sham group, acute lung injury caused the CLP model; that is, the pathological specimens showed extensive thickening of the alveolar wall and obvious infiltration of inflammatory cells (Fig. 3 A). The lung injury scores in the CLP-IRF1-/- group were significantly lower than those in the CLP-WT group, suggesting that IRF1 could alleviate LPS-induced septic ARDS(P < 0.01) (Fig. 3C).Fig. 3 IRF1 knockout alleviated sepsis induced inflammatory lung injury in mice. There were 3 groups, and each group contained 6 mice. (A) Representative images of lung sections stained with H&E from CLP induced ARDS mice at 24 h, the magnification of microscopic images is × 200 (n = 6 for each group). (B) Representative histopathologic and IHC image of lung tissues from CLP induced ARDS mice at 24 h, the magnification of microscopic images is × 400 (n = 6 for each group). (C) Lung injury scores were estimated by the method of Mikawa (n = 6 for each group). (D) Comparison of the lung wet weight to body weight ratio (LWW/BW) in different groups at 24 h (n = 6 for each group). (E) The percentage of M1-polarized macrophage is assessed through CD86 positive cells to the total cells (n = 6 for each group). (F) Comparison of the pro-inflammatory cytokine IL-1β in alveolar lavage fluid (BALF) by ELISA (n = 6 for each group). (G) Comparison of the pro-inflammatory cytokine IL-6 in BALF by ELISA (n = 6 for each group). (H) Comparison of the pro-inflammatory cytokine CCL2 in BALF by ELISA (n = 6 for each group). The LWW/BW response to pulmonary oedema is related to the severity of lung injury. In the CLP-induced ARDS mice, the LWW/BW was significantly higher than that in the control group at 24 h (P < 0.05) (Fig. 3D). After IRF1 knockout, LWW/BW was significantly reduced in the mice with “CLP-induced ARDS” (P < 0.05) (Fig. 3D), suggesting that regulation of low IRF1 expression can effectively inhibit lung injury and alleviate ARDS. Previous studies have shown that macrophage M1 polarization contributes to particulate matter (PM)-induced lung injury, while inhibition of M1 polarization can alleviate acute lung injury [18]. Therefore, we evaluated the effect of IRF1 knockout on the number of lung macrophages on M1 polarization in ARDS by immunohistochemistry. Compared with the sham group, the proportion of CD86+ macrophages in the lungs of CLP-induced ARDS mice was significantly increased (Fig. 3B) (P < 0.05), indicating that M1-polarized macrophages aggregated into the lung during pulmonary ARDS. After IRF1 knockout, the proportion of CD86+ macrophages in the lungs of the mice were significantly reduced again (Fig. 3B) (P < 0.05). In this study, the levels of IL-1β, IL-6 and CCL2 in the BALF of CLP-induced ARDS mice were significantly higher than those in the BALF of control mice (P < 0.05).Knockout of IRF1inhibited the production of the pro-inflammatory cytokines IL-1β, IL-6 and CCL2 (Fig. 3 F-H) in the BALF of mice with CLP-induced ARDS (P < 0.05). These results indicate that high expression of AKDND22 is the critical link in the ARDS inflammatory storm and exacerbation of lung injury. These results further suggest that IRF1 promotes M1-polarized macrophage lung aggregation in CLP-induced ARDS. 3.4 IFIH1 modulates macrophage M1 polarization depending on IRF1 translocation into the nucleus IFIH1, STAT1 and IRF1 have been validated to be involved in macrophage M1 polarization and septic ARDS in previous studies and the current study. The PPI and WGCNA analyses suggested that these three molecules may synergistically regulate macrophage M1 polarization. To investigate this possibility, we first knocked down IFIH1 expression to evaluate whether IFIH1 could affect IRF1 and STAT1 (transcription factors) translocation into the nucleus. Western blot analysis of the cell nucleus indicated that IFIH1 knockdown had no effect on STAT1 translocation into the cell nucleus (Fig. 4 , P > 0.05). Compared with the control treatment (shCtrl), Western blot in the cell nucleus revealed that both Poly (I:C)-induced and LPS-induced expression of IRF1 in BMDM nuclei was significantly reduced after IFIH1 knockdown (Fig. 4, P < 0.05). In addition, compared with the control treatment, IFIH1 knockdown had no effect on STAT1 and IRF1 protein expression in BMDMs (Fig. 4, P < 0.05).Fig. 4 In response to both Poly(I:C)-induced and LPS-induced circumstances, IFIH1 knockdown inhibited IRF1 translocation into the nucleus. Three mice were included in each group. (A) IRF1 and STAT1 expression in the nucleus as well as total IRF1 and STAT1 levels were examined using Western blotting in each set of BMDMs (n = 3 for each group). (B) IFIH1 knockdown had no effect on STAT1 translocation into the cell nucleus, according to quantitative measurement of STAT1 in the cell nucleus, (P > 0.05, n = 3 for each group). (C) Comparing the effects of IFIH1 silencing to the control treatment, quantitative examination of IRF1 in the cell nucleus revealed that both Poly(I:C)-induced and LPS-induced expression of IRF1 in BMDMs nuclei was significantly reduced (P < 0.05, n = 3 for each group). (D/E) Compared with the control treatment, IFIH1 knockdown had no effect on STAT1 and IRF1 protein expression in BMDMs (P < 0.05, n = 3 for each group). The vector transfection of shIFIH1 and control shRNA had no effect on the expression of IRF1 and STAT1 in BMDMs' nucleus. The statistical analysis was from three independent experiments, and the bar indicates the SD values. These results indicate that IFIH1 is an adaptor and signal transduction molecule that activates IRF1 translocation into the nucleus. 3.5 IRF1 is the transcription factor of STAT1 IRF1 and STAT1 are transcription factors that regulate molecular expression. To further explore the molecular mechanism linking IFIH1, STAT1 and IRF1, we knocked down STAT1 expression to assess whether STAT1 influences IFIH1 and IRF1 expression. Then, we utilized BMDMs from IRF1-/-mice to evaluate whether IRF1 regulates STAT1 and IFIH1 expression. Western blotting indicated that STAT1 knockdown did not influence the expression of IRF1 or IFIH1 (Fig. 5 A, P > 0.05). Western blotting indicated that IRF1 knockout markedly decreased STAT1 expression (Fig. 5B, P < 0.05) but did not influence the expression of IFIH1 (P > 0.05).Fig. 5 IRF1 was the transcription factor of STAT1. Three mice were included in each group. (A) Western blotting was performed to detect the expression of IRF1 and IFIH1 after STAT1 knocking down. The results indicated that STAT1 knockdown did not influence the expression of IRF1 and IFIH1 (P > 0.05, n = 3 for each group). (B) Western blotting was performed to detect the expression of STAT1 and IFIH1 after IRF1 knocking out. The results indicated that IRF1 knockout markedly decrease STAT1 expression (P < 0.05), but did not influence the expression of IFIH1 (P > 0.05, n = 3 for each group). (C) CHIP-seq indicated that IRF1 binds to the promoter region of STAT1, compared with input (FDR < 0.05, n = 3 for each group). (D) CHIP-PCR further validated that IRF1 to the promoter region of STAT1(P > 0.05, n = 3 for each group). ChIP-seq further found that IRF1 binds to the promoter region of STAT1 (Fig. 5C, FDR < 0.05) compared with the input group. The binding region was located on Chr1 start=“52118977″ stop=”52119911″; the sequence and length are shown in Table S7. ChIP-PCR further confirmed that IRF1 binds to the promoter region of STAT1 compared with the IgG group. These results indicate that IRF1 is the transcription factor (TF) of STAT1. Fig. 6 illuminates the molecular mechanism of this study.Fig. 6 The molecular mechanism link of IFIH1/IRF1/STAT1 induced macrophage M1 polarization. IFIH1 is activated by infection, which triggers the nuclear translocation of the transcription factor IRF1. IRF1 then attaches to the STAT1 promoter region to initiate the transcription of STAT1 and polarization of macrophage M1. 4 Discussion Sepsis-induced ARDS is a lung injury condition caused by dysregulation of the systemic inflammatory host response to infections [25], [26], [27], [28]. Macrophage M1 polarization fuels the inflammatory process [29]. We conducted an interrelated bioinformatics analysis and found that IFIH1, IRF1 and STAT1 are all associated with sepsis-induced ARDS as well as macrophage M1polarization.Previous studies have confirmed that IFIH1 and STAT1 regulate macrophage M1 polarization and promote the development of inflammatory diseases, including sepsis-induced ARDS. The current study further indicated that IRF1 contributes to M1-macrophage polarization and the development of sepsis-induced ARDS both in vivo and in vitro. Furthermore, molecular experiments revealed the underlying mechanisms of this link: infection activates IFIH1, which triggers the nuclear translocation of the transcription factor IRF1. IRF1 then attaches to the STAT1 promoter region to initiate the transcription of STAT1 and macrophage M1 polarization. IFIH1 is a cytosolic receptor responsible for binding viral RNA and activating IFN regulatory factor 3 (IRF3),resulting in the induction of inflammatory and antiviral genes [30], [31], [32].Accumulating evidence of the pro-inflammatory function of IFIH1, together with updated studies, suggests that LPS and RNA mimics can both trigger IFIH1 and IRF1 activation. Our previous study and a study by Stone et al. identified IFIH1 as a regulator of macrophage M1 polarization, contributing to the development of sepsis-induced ARDS. Li et al. indicated that IFIH1 mediates the activation and upregulation of STAT1 [32]. Our present study further revealed the underlying mechanism; namely, IFIH1 facilitates the translocation of IRF1, the upstream transcription factor of STAT1, into the nucleus, leading to upregulation of STAT1 expression. IRF1 is an important antiviral molecule and is implicated in HIV susceptibility and pathogenesis. Specifically, after infection with simian immunodeficiency virus, IFN-(α,β)-producing plasmatoid dendritic cells, macrophages, and large increases in IFN-γ expression were found in cervical vaginal tissues[33]. These findings are especially interesting because IRF1 is a critical TF in the IFN pathway. Zhu et al. and Guo et al. found that miR-19a-3p and miR-130b-3p regulate macrophage M1 polarization through the suppression of the STAT1/IRF1 pathway in the RAW264.7 cell line [23], [24]. However, in these studies, the impact of IRF1 on macrophage M1 polarization and ARDS was not convincingly demonstrated within vitro or in vivo experiments, nor were they validated in primary cells (particularly gene knockout mice). The present study included IRF1 deletion, rescue, and over-expression to provide solid evidence that IRF1 is an essential molecule for macrophage M1 polarization, which promotes inflammatory lung damage in CLP-induced sepsis-induced ARDS. ChIP and further experiment indicate that IRF1 is the TF of STAT1 and thus increases STAT1 expression. TFs that regulate macrophage M1 polarization, such as STAT1, are robustly expressed in classically activated macrophages and regulate the macrophage inflammatory response [33]. The polarization of macrophages to the M1 phenotype depends on the activation of the TLR4 or IFN-γ pathways, which in turn activate the ERK, NFκB, and STAT1 pathways [34]. Specifically, IFN-α is known to initiate signal transducers and STAT1 signalling, which upregulates the gene expression of interferon regulatory factor-1, a key transcription factor necessary for the cytotoxicity of NK cells, as well as the expression of the effector molecules Fas-L and perforin[35]. Previous studies revealed that IFN-I triggers STAT1 activation. However, the current study indicates that the IFIH1/IRF1 pathway enhances STAT1 upregulation because IRF1 is the TF of STAT1, synergizing with IFN-I-induced STAT1 activation. 5 Conclusion IRF1 has been identified as the essential molecule that controls the polarization of M1 macrophages and the onset of septic ARDS both in vivo and in vitro. Additionally, IFIH1 promotes IRF1 (transcription factor) translocation into the nucleus to initiate STAT1 transcription as the adapter in response to infection mimic irritants. Declarations. Ethics approval and consent to participate The revised Helsinki Declaration was followed when conducting the current study. The Critical Care Centre at Zhongda Hospital's biological specimen bank provided the samples from the individuals who were included. The Institutional Ethics Committee of Zhongda Hospital approved and oversaw the establishment of this specimen bank in 2017. (number: 2017ZDSYLL105). Figure S1 displayed the Ethical documents. Animal research in the current study were evaluated and authorized by the Jinan Central Hospital's Animal Care and Use Committee, shown in Figure S2. Consent for publication. Not applicable. Availability of data and materials. The high-through dataset was down loaded from the GEO public database (GSE46903). Xue J, Schmidt SV, Sander J, Draffehn A, Krebs W, et al. Transcriptome-based network analysis reveals a spectrum model of human macrophage activation.2014. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc = GSE46903. Further information and requests for resources and reagents should be directed to and will be fulfilled by the first author, Shi Zhang, E-mail: [email protected]. Funding Supported in part by grants from.National Natural Science Foundation of China (Grant No: 8220081418); National Natural Science Foundation of Shandong (Grant No: ZR2022QH332); Jinan Science and Technology Bureau's Clinical Technology Innovation Program (Grant No: 202134058); Scientific Research Start-up Funds for talent introduction of Jinan Central hospital (Grant No: YJRC2021010); 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 Supplementary data 2 Data availability The authors do not have permission to share data. Acknowledgements We would like to thank all the editors, reviewers and scholars due to their contribution on this study Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.intimp.2022.109478. ==== Refs References 1 Bellani G. Laffey J.G. Pham T. Fan E. Brochard L. Esteban A. Epidemiology, patterns of care, and mortality for patients with acuterespiratory distress syndrome in intensive care units in 50 countries JAMA 23 319 2016 Feb 698 710 2 Zhou F. Yu T. Du R. Fan G. Liu Y. Liu Z. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study Lancet. 395 10229 2020 Mar 28 1054 1062 32171076 3 Liu L. Yang Y. Gao Z. Li M. Mu X. Ma X. Practice of diagnosis and management of acute respiratory distress syndrome in mainland China: a cross-sectional study J. Thorac. Dis. 10 9 2018 Sep 5394 5404 30416787 4 Lin S. Wu H. Wang C. Xiao Z. Xu F. Regulatory T Cells and Acute Lung Injury: Cytokines, Uncontrolled Inflammation, and Therapeutic Implications Front Immunol. 9 9 2018 Jul 1545 30038616 5 Meyer N.J. Gattinoni L. Calfee C.S. Acute respiratory distress syndrome Lancet. 398 10300 2021 Aug 14 622 637 34217425 6 Garnier M. Gibelin A. Mailleux A.A. Leçon V. Hurtado-Nedelec M. Laschet J. Macrophage polarization favors epithelial repair during acute respiratory distress syndrome Crit. Care Med. 46 7 2018 Jul e692 e701 29649066 7 Zhang S. Wu Z. Chang W. Liu F. Xie J. Yang Y. Qiu H. Classification of patients with sepsis according to immune cell characteristics: a bioinformatic analysis of two cohort studies Front Med (Lausanne). 3 7 2020 Dec 598652 8 Zhang S. Chu C. Wu Z. Liu F. Xie J. Yang Y. Qiu H. IFIH1 Contributes to M1 Macrophage Polarization in ARDS Front Immunol. 14 11 2021 Jan 580838 9 Ahmad S. Mu X. Yang F. Greenwald E. Park J.W. Jacob E. Breaching Self-Tolerance to Alu Duplex RNA Underlies MDA5-Mediated Inflammation Cell. 172 4 2018 Feb 8 797 810.e13 29395326 10 Stone A.E.L. Green R. Wilkins C. Hemann E.A. Gale M. Jr. RIG-I-like receptors direct inflammatory macrophage polarization against West Nile virus infection Nat. Commun. 10 1 2019 Aug 13 3649 31409781 11 Gan ZS, Wang QQ, Li JH, Wang XL, Wang YZ, Du HH. Iron Reduces M1 Macrophage Polarization in RAW264.7 Macrophages Associated with Inhibition of STAT1. Mediators Inflamm. 2017;2017:8570818. Epub 2017 Feb 13. 12 Liu Y. Liu Z. Tang H. Shen Y. Gong Z. Xie N. The N6-methyladenosine (m6A)-forming enzyme METTL3 facilitates M1 macrophage polarization through the methylation of STAT1 mRNA Am. J. Physiol. Cell Physiol. 317 4 2019 Oct 1 C762 C775 31365297 13 Guo FM, Qiu HB. Definition and dignosis of sepsis 3.0. Zhonghua Nei Ke Za Zhi. 2016 Jun;55(6):420-2. 14 Thompson B.T. Chambers R.C. Liu K.D. Acute respiratory distress syndrome N Engl J Med 377 19 2017 1904 1905 15 Langfelder P. Horvath S. WGCNA: an R package for weighted correlation network analysis BMC Bioinformatics. 29 9 2008 Dec 559 16 Xue J. Schmidt S.V. Sander J. Draffehn A. Krebs W. Quester I. Transcriptome-based network analysis reveals a spectrum model of human macrophage activation Immunity. 40 2 2014 Feb 20 274 288 24530056 17 Robinson M.D. McCarthy D.J. Smyth G.K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data Bioinformatics. 26 1 2010 Jan 1 139 140 19910308 18 Ritchie M.E. Phipson B. Wu D. Hu Y. Law C.W. Shi W. limma powers differential expression analyses for RNA-sequencing and microarray studies Nucleic Acids Res. 43 7 2015 Apr 20 e47 25605792 19 Stelzl U. Worm U. Lalowski M. Haenig C. Brembeck F.H. Goehler H. A human protein-protein interaction network: a resource for annotating the proteome Cell. 122 6 2005 Sep 23 957 968 16169070 20 Lai X. Schmitz U. Gupta S.K. Bhattacharya A. Kunz M. Wolkenhauer O. Computational analysis of target hub gene repression regulated by multiple and cooperative miRNAs Nucleic Acids Res. 40 18 2012 8818 8834 22798498 21 Cui Z. Liu Y. Zhang J. Qiu X. Super-delta2: an enhanced differential expression analysis procedure for multi-group comparisons of RNA-seq Data Bioinformatics. 2021 Mar 8:btab155 22 Wang J. Li H. Xue B. Deng R. Huang X. Xu Y. IRF1 Promotes the innate immune response to viral infection by enhancing the activation of IRF3 J Virol. 94 22 2020 Oct 27 e01231 e10320 32878885 23 Zhu X. Guo Q. Zou J. Wang B. Zhang Z. Wei R. MiR-19a-3p Suppresses M1 Macrophage Polarization by Inhibiting STAT1/IRF1 Pathway Front Pharmacol. 4 12 2021 May 614044 24 Guo Q. Zhu X. Wei R. Zhao L. Zhang Z. Yin X. miR-130b-3p regulates M1 macrophage polarization via targeting IRF1 J. Cell Physiol. 236 3 2021 Mar 2008 2022 32853398 25 Liu F. Xie J. Zhang X. Wu Z. Zhang S. Xue M. Overexpressing TGF-β1 in mesenchymal stem cells attenuates organ dysfunction during CLP-induced septic mice by reducing macrophage-driven inflammation Stem cell res. therapy 11 1 2020 378 26 Guillen-Guio B. Lorenzo-Salazar J.M. Ma S.F. Hou P.C. Hernandez-Beeftink T. Corrales A. Sepsis-associated acute respiratory distress syndrome in individuals of European ancestry: a genome-wide association study Lancet Respir. Med. 8 3 2020 Mar 258 266 31982041 27 Zhang S. Chang W. Xie J. Wu Z. Yang Y. Qiu H. The efficacy, safety, and optimal regimen of corticosteroids in sepsis: a bayesian network meta-analysis Crit. Care Explor. 2 4 2020 Apr 29 e0094 32426736 28 Zhang S. Liu F. Wu Z. Xie J. Yang Y. Qiu H. Contribution of m6A subtype classification on heterogeneity of sepsis Ann Transl Med. 8 6 2020 Mar 306 32355750 29 Jiao Y. Zhang T. Zhang C. Ji H. Tong X. Xia R. Exosomal miR-30d-5p of neutrophils induces M1 macrophage polarization and primes macrophage pyroptosis in sepsis-related acute lung injury Crit. Care. 25 1 2021 Oct 12 356 34641966 30 Lee W. Lee S.H. Kim M. Moon J.S. Kim G.W. Jung H.G. Vibrio vulnificus quorum-sensing molecule cyclo (Phe-Pro) inhibits RIG-I-mediated antiviral innate immunity Nat. Commun. 9 2018 1606 29686409 31 Jaeger M, van der Lee R, Cheng SC, Johnson MD, Kumar V, Ng A, et al. The RIG-I-like helicase receptor MDA5 (IFIH1) is involved in the host defense against Candida infections. Eur J Clin Microbiol Infect Dis (2015) 2015-05-0134(5):963–74. 32 Li Y. Yu P. Qu C. Li P. Li Y. Ma Z. MDA5 against enteric viruses through induction of interferon-like response partially via the JAK-STAT cascade Antiviral Res. 176 2020 Apr 104743 10.1016/j.antiviral.2020.104743 Epub 2020 Feb 10 PMID: 32057771 33 Roy S. Schmeier S. Arner E. Alam T. Parihar S.P. Ozturk M. Redefining the transcriptional regulatory dynamics of classically and alternatively activated macrophages by deep CAGE transcriptomics Nucleic Acids Res. 43 14 2015 Aug 18 6969 6982 26117544 34 Ma G. Pan P.Y. Eisenstein S. Divino C.M. Lowell C.A. Takai T. Paired immunoglobin-like receptor-B regulates the suppressive function and fate of myeloid-derived suppressor cells Immunity. 34 3 2011 Mar 25 385 395 21376641 35 Ahlenstiel G, Titerence RH, Koh C, Edlich B, Feld JJ, Rotman Y, et al. Natural killer cells are polarized toward cytotoxicity in chronic hepatitis C in an interferon-alfa-dependent manner. Gastroenterology. 2010 Jan;138(1):325-35.e1-2.	36462334	PMC9709523	NO-CC CODE	2022-12-01 23:23:07	no	Int Immunopharmacol. 2023 Jan 30; 114:109478	utf-8	Int Immunopharmacol	2,022	10.1016/j.intimp.2022.109478	oa_other
==== Front Cir Esp Cir Esp Cirugia Espanola 0009-739X 1578-147X AEC. Published by Elsevier España, S.L.U. S0009-739X(20)30402-4 10.1016/j.ciresp.2020.12.007 Scientific Letter Surgical education during pandemic times: How the virtual world can help us in real life? The Hernia U experience Formación quirúrgica en tiempos de pandemia: ¿ cómo puede ayudarnos el mundo virtual en la vida real? La experiencia de Hernia UMalcher Flavio a Lima Diego Laurentino b⁎ Cavazzola Leandro Totti c CL Lima Raquel Nogueira d Davila Eduardo Parra e Morales-Conde Salvador f a Director Abdominal Wall Program, Department of Surgery, Montefiore Medical Center and Assistant Professor, Albert Einstein College of Medicine, Bronx, NY, USA b Department of Surgery, Montefiore Medical Center, Bronx, NY, USA c Department of Surgery, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS, Brazil d Pernambuco Health College, Recife, PE, Brazil e Director of Hernia and Abdominal Wall Reconstruction, Good Samaritan Medical Center-TENET Health, West Palm Beach, FL, USA f Chief of Innovation in Minimally Invasive Surgery of the University Hospital Virgen del Rocio and Head of the General and Digestive Surgery Unit of Hospital Quironsalud Sagrado Corazon, Sevilla, Spain ⁎ Corresponding author. 16 1 2021 4 2021 16 1 2021 99 4 315316 © 2020 AEC. Published by Elsevier España, S.L.U. All rights reserved. 2020 AEC 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 pmcDear Editor, The World Health Organization declared the Corona Virus Disease (COVID-19) a pandemic on March 11th. As this unprecedent situation deteriorated all universities have stopped all types of lectures and changed to online methods.1, 2 The first activities suspended were medical students’ clinical clerkships and elective opportunities.3, 4 Surgical residents faced a great challenge during the pandemic. Elective surgeries for benign diseases were postponed and the urgent or emergency surgeries were performed by more experienced surgeons. This reduced the number of learning opportunities for surgical residents.1 During the global crisis, social media platforms (Twitter, Facebook, WhatsApp) and also applications for videoconferencing (Google Hangouts, Skype, Zoom, WebEx) have played a prominent role.1 Surgery Departments were able to implement activities such as journal clubs, case discussions and lectures to provide adequate surgical education to medical students, residents and attendings. Surgical societies tried to mitigate the scientific damage imposed by the pandemic. Many surgical meetings were postponed or cancelled this year due to the pandemic. Others had online editions at a low cost or free of charges. The same phenomenon happened with hands-on courses, lectures and social meetings. Hernia U steps up The Hernia U team realized the need to strengthen surgical education online during this difficult time. Hernia U (www.herniau.com) was created in 2016 by hernia specialists with the objective to expand the abdominal wall surgery (AWS) educational landscape and make it available for surgeons who wanted to revisit their hernia education. It is an online platform where surgeons can register with no cost and subscribe for different activities: Hernia A to Z Fundamentals, Advanced, Live surgeries, and Hernia U library which is a depository for high quality lectures, cases and the Hernia U podcast. Some courses (Hernia A to Z Fundamentals and Advanced) are available in other languages: Chinese, Spanish, French or Arabic. Currently, more than 15,000 professionals from 157 countries have already participated in one of the courses available in the platform since its beginning. A new modality of live lectures was created after March 11th. Not just exclusivelyour team perceived the need to strengthen surgical education online, but also surgeons around the globe. In 2019, 1709 new surgeons registered in the platform from March 11th to the end of September. In 2020, this number increased more than 3 times: 5523 new surgeons registered in the same period. In the last 6 months, Hernia U has broadcasted 28 live events (21 lectures and 7 surgeries) against 7 live surgeries in the same period last year. (Table 1 ) There were not just hernia-related topics. Specialists also discussed what surgeons should know during the pandemic (lecture in English and Spanish) and the importance of social media. Furthermore, Hernia U has joined forces with the Americas Hernia Society (AHS), the European Hernia Society (EHS) and the International Hernia Collaboration (IHC) from Facebook to proportionate scientific activities. Hernia U didn’t just offer more activities, but they offered activities that surgeons were eager to participate in Table 1.Table 1 Hernia U attendants from March 11th to October in 2019 and 2020. Table 1Hernia U 2019 2020 p-Value New Surgeons registered (March 11-oct.) 1709 5523 Live Surgeries (March 11-oct.) 7 7 Live Surgeries surgeon attendance (March 11-oct.)a 803 1202 Logins/surgery (March 11-oct.) 114.7 195.6 0.021 For statistical analysis, IBM SPSS Statistics for Windows, version 26 (IBM Corp., Armonk, N.Y., USA). a Total number of different surgeons that attended the events. Despite the pandemic, Hernia U team managed to have the same number of live surgeries as in the same period of 2019. To our surprise, the number of surgeons who attended these live cases was higher (1202 versus 803) and the logins per surgery was significantly higher: 195.6 versus 114.7 logins/surgery (p = 0.021). It is important to note the 1202 was the number of different surgeons who watched the live surgeries, but to calculate the audience (logins/surgery), the same surgeon could be counted more times, depending on how many events he was present. We are living in a time where social distancing can mitigate the spread of the disease. This cannot stop us from sharing knowledge and experiences which will improve healthcare worldwide. We have the tools that allow us to make it available to everyone and Hernia U is part of this arsenal. Funding There was no funding for this research project. Conflicts of interest DL Lima and RNCL Lima have no conflict of interests. Drs DL Lima, RNCL Lima, SMConde reports consulting fees from BD Bard, Medtronic, Ethicon, Stryker, Storz, Olympus, Baxter, Dipro, BBraum, Gore, outside the submitted work. EPDavila discloses consulting fees from Bard Davol, outside the submitted work. Dr. Malcher discloses consulting fees from BD & Medtronic, outside the submitted work. Dr. Cavazzola discloses consulting fees from Strattner and Intuitive, outside the submitted work. Acknowledgments We’d like to thank Jennifer Petrie for helping with the manuscript. ==== Refs References 1 Dedeilia A. Sotiropoulos M.G. Hanrahan J.G. Janga D. Dedeilias P. Sideris M. Medical and surgical education challenges and innovations in the COVID-19 era: a systematic review In Vivo 34 Suppl. 2020 1603 1611 32503818 2 Mian A. Khan S. Medical education during pandemics: a UK perspective BMC Med 18 2020 100 32268900 3 Gallo G. Trompetto M. The effects of COVID-19 on academic activities and surgical education in Italy J Invest Surg 33 2020 687 689 32249660 4 Calhoun K.E. Yale L.A. Whipple M.E. Allen S.M. Wood D.E. Tatum R.P. The impact of COVID-19 on medical student surgical education: Implementing extreme pandemic response measures in a widely distributed surgical clerkship experience Am J Surg 220 2020 44 47 32389331	36465488	PMC9709581	NO-CC CODE	2022-12-01 23:23:07	no	Cir Esp. 2021 Apr 16; 99(4):315-316	utf-8	Cir Esp	2,021	10.1016/j.ciresp.2020.12.007	oa_other
==== Front Am J Infect Control Am J Infect Control American Journal of Infection Control 0196-6553 1527-3296 Mosby S0196-6553(22)00818-5 10.1016/j.ajic.2022.11.017 Erratum Erratum 30 11 2022 30 11 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 pmcIn the article “Novel Case of Candida auris in the Veterans Health Administration and in the state of South Carolina.” by Lucy Austin, et al. in the November issue of the American Journal of Infection Control (2022;50(11):1258-62) The author names were incorrectly listed as surname followed by given name. The correct author names are Lucy Austin MSN, RN, CIC; Paula Guild MN, RN, CIC; Christine Rovinski MSN, APRN; and Jailan Osman MD, MBChB, FCAP. This correction has been made to the online version of the article.	36462955	PMC9709597	NO-CC CODE	2022-12-01 23:23:07	no	Am J Infect Control. 2022 Nov 30; doi: 10.1016/j.ajic.2022.11.017	utf-8	Am J Infect Control	2,022	10.1016/j.ajic.2022.11.017	oa_other
==== Front Heart Lung Circ Heart Lung Circ Heart, Lung & Circulation 1443-9506 1444-2892 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier B.V. S1443-9506(22)01138-6 10.1016/j.hlc.2022.10.008 Original Article Clinical Value and Mechanism of Long Non-Coding RNA UCA1 in Acute Respiratory Distress Syndrome Induced by Cardiopulmonary Bypass Chen Yongliang MM a Xue Jing MM b∗ Fang Daguang MM a Tian Xuefei MM a a Department of Cardiac Surgery, Affiliated Hospital of Chengde Medical University, Chengde, Hebei, China b School of Basic Medicine, Chengde Medical University, Chengde, Hebei, China ∗ Corresponding author at: School of Basic Medicine, Chengde Medical University, Anyuan Road, Shuangqiao District, Chengde, 067000 Hebei, China. 30 11 2022 30 11 2022 21 6 2022 12 10 2022 13 10 2022 © 2022 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier B.V. All rights reserved. 2022 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ) 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. Aim Long non-coding RNA (lncRNA) can be used as a biological marker for the diagnosis and treatment of various diseases. The study aimed to detect changes in the expression of lncRNA for urothelial carcinoma associated 1 (UCA1) in patients with cardiopulmonary bypass (CPB)-induced acute respiratory distress syndrome (ARDS). Clinical values and cell function in ARDS were explored. Method In total, 195 patients without CPB-induced ARDS were included in the control group, and 85 patients with ARDS were included in the ARDS group. Serum UCA1 levels were measured by quantitative real-time polymerase chain reaction. A549 was used for the cell experiments by establishing oxygen–glucose deprivation/reperfusion (OGD/R) cell models, and the cell viability and apoptosis were tested. The concentration of inflammatory factors was tested by an enzyme-linked immunosorbent assay. A luciferase reporting assay was applied for target gene analysis. Results Quantitative real-time polymerase chain reaction revealed a gradual increase in serum UCA1 in both control and ARDS cases, and patients with ARDS had higher levels of UCA1 than those in the control group. Serum UCA1 was positively correlated with serum tumour necrosis factor-α and interleukin-6 concentration in patients with ARDS. UCA1 had the ability to distinguish patients with ARDS from those without it. UCA1 inhibition protected against lung injury and inhibited cell inflammation in vitro. MicroRNA (miR-182-5p) was downregulated in OGD/R-induced cell models and sponged by UCA1. Conclusions Elevated expression of UCA1 may be associated with the occurrence of ARDS after CPB surgery. The regulatory role of UCA1 in ARDS might be related to inflammation and downregulated miR-182-5p in alveolar epithelial cells. Keywords UCA1/miR-182-5p CPB ARDS Inflammation Alveolar epithelial cell ==== Body pmcIntroduction In recent years, the wide application of cardiopulmonary bypass (CPB) has greatly improved the success rate of cardiac surgery. Although the postoperative monitoring technology for CPB is improving continuously, organ damage caused by CPB is still common and cannot be completely avoided [1]. The incidence of lung injury after CPB is as high as 12%–50%, and most of these patients require ventilator treatment, or even develop acute respiratory distress syndrome (ARDS) [2]. Once CPB patients develop ARDS, the fatality rate is 15%–68% [3]. In addition, ARDS will lead to prolonged hospitalisation and increased economic burden. Therefore, it is necessary to find more optimised CPB-related diagnostic indicators for ARDS. Long non-coding RNAs (lncRNAs) are a class of transcripts of over 200 nucleotides in length. LncRNAs have a variety of biological functions and play a role in regulating the expression of specific genes [4]. Differential expression of lncRNAs has been widely identified in many human diseases [5]. Notably, studies on lncRNA in ischaemia–reperfusion tissue injury have been widely reported, including H19, metastasis associated lung adenocarcinoma transcript 1 (MALAT1), and nuclear paraspeckle assembly transcript 1 (NEAT1) [[6], [7], [8]]. High levels of urothelial carcinoma associated 1 (UCA1) were detected in pulmonary artery smooth muscle cells under hypoxia, which is further involved in promoting the development of pulmonary arterial hypertension [9]. UCA1 is also highly expressed in septic rats, which further aggravates lung inflammation and apoptosis, leading to pneumonia [10]. In lipopolysaccharide (LPS)-treated WI-38 cells, upregulation of UCA1 is also reported to aggravate LPS-induced cell apoptosis and the inflammatory response, ultimately affecting the development of pneumonia [11]. However, the role of UCA1 in CPB-induced ARDS has not been examined. The competitive endogenous RNA (ceRNA) networks hypothesis has been proposed in a variety of diseases, indicating that lncRNA can serve as a microRNA (miRNA) sponge or decoy to modulate the expression of miRNA targets [12]. Previously, miR-182-5p has been identified to be a target gene of UCA1 in several diseases [13,14]. miR-182-5p is a member of the miR-183/96/182 gene cluster [15]. In recent years, miR-182-5p has been reported to be abnormally expressed in inflammatory tissues and acts as an important regulatory factor in the inflammatory response. For example, miR-182-5p can inhibit the transcription of proinflammatory genes driven by tumour necrosis factor (TNF)-α [16,17]. More importantly, miR-182-5p may affect the proliferation, inflammation, and apoptosis of immune response cells related to the therapeutic effect of bone marrow mesenchymal stem cells in acute lung injury [18]. In addition, miR-182-5p is detected to be highly expressed in lung tissues and the bronchoalveolar lavage fluid of LPS-induced acute lung injury mice; its inhibitory effect on the release of inflammatory factors has also been confirmed in vitro [19]. It is also widely reported to be involved in the long disease, such as acute lung injury and lung cancer [20,21]. Therefore, in this study, we preliminarily explored the relationship between miR-182-5p and UCA1 in ARDS. In this study, the expression of UCA1 in patients with CPB-induced ARDS was detected, and its association with the clinical data was evaluated. In addition, in light of the ceRNA between lncRNA and miRNAs, the underlying mechanism of UCA1 in ARDS with the involvement of miR-182-5p was further elucidated. Materials and Methods Study Subjects A total of 280 patients who underwent CPB surgery at the Affiliated Hospital of Chengde Medical University were included in the study and divided into two groups according to the incidence of postoperative ARDS. Altogether, 195 patients without ARDS were included in the control group, and 85 patients with ARDS were included in the ARDS group. Patients with ARDS were diagnosed using the Berlin Definition criteria, which include bilateral lung infiltrates detected on chest radiographs, a pulmonary capillary pressure of ≤18 mmHg, and a partial pressure of O2/fraction of inspired O2 ≤ 200 mmHg [22]. The retrospective nature of this study makes it difficult to differentiate between ARDS, transfusion-related acute lung injury (TRALI), and transfusion-associated cardiac overload (TACO). We took all possible measures to exclude patients with TRALI and TACO from the ARDS group. We used expert panel criteria for the adjudication of non-acute post-transfusion hypoxaemia with bilateral pulmonary infiltrates to differentiate between TRALI, TACO, and ARDS, or other aetiologies such as pneumonia, aspiration, congestive heart failure, or diffuse pulmonary haemorrhage by chart review. Initial screening was done by two authors followed by adjudication by the senior author to minimise the risk of inaccuracy. The inclusion criteria of patients in this study were as follows: (1) underwent CPB surgical treatment, and received general anaesthesia under endotracheal intubation; (2) all patients in the observation group were diagnosed with ARDS; (3) no complicated diseases such as chest trauma, endocrine disease, liver and kidney dysfunction, and so on; (4) patients with a history of lung infection and chronic lung disease were excluded; (5) had no history of intravenous antibiotics prior to surgery for any active infections; (6) patients and their families volunteered to participate and signed an informed consent form; and (7) the experiment was approved by the Affiliated Hospital of Chengde Medical University Ethics Committee. Serum Sample Collection Five millilitres of peripheral venous blood was collected at four time points after thoracotomy, including before CPB (T0, after thoracotomy but before CPB), 4 hours after CPB (T1), 8 hours after CPB (T2), and 16 hours after CPB (T3). Serum was collected after centrifugation at 1,800 g for 10 minutes. Cell Culture and Transfection The human lung adenocarcinoma cell line A549 was purchased from the cell bank of the Chinese Academy of Sciences, Shanghai, and cultured in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum in a culture environment of 5% CO2 at 37°C. Logarithmic-phase cells were selected for subsequent experiments. To mimic ARDS, cells were cultured under anaerobic conditions in 85% N2 for 4 hours, following 24 hours of normal conditions [23]. Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, MA, USA) was used for the cell transfection, according to the manufacturer's instructions. Sequences of small interfering RNA of UCA1 (si-UCA1) and its negative control (si-NC), miR-182-5p mimic or inhibitor, and their negative control (mimic-NC or inhibitor-NC) were constructed and synthesised by Gene Pharma (Shanghai, China). Quantitative Real-Time Polymerase Chain Reaction Quantitative real-time polymerase chain reaction (qRT-PCR) was done to detect RNA levels. Firstly, total RNA was extracted from serum samples or cells using Trizol (Invitrogen, Carlsbad, CA, USA), and transcribed into cDNA using FastKing gDNA Dispelling RT SuperMix (Tiangen, Beijing, China) for UCA1 or a miRcute Plus miRNA First-Strand cDNA Kit (Tiangen) for miR-182-5p. Secondly, qRT-PCR was performed using a SuperReal PreMix Plus (SYBR Green) kit (Tiangen) for UCA1 or a miRcute Plus miRNA qPCR Kit (SYBR Green) (Tiangen) for miR-182-5p, according to the kit instruction leaflet. RNA levels were calculated via the 2–ΔΔCt method. Cell Counting Kit-8 Cell counting kit-8 (CCK-8) was performed to evaluate cell viability. Cell suspensions of each group were inoculated in 96-well plates at a density of 1,000 cells per well. CCK-8 (10 μL) was added to each well at 0, 24, 48, and 72 hours after culture, and culture was continued for 4 hours. The optical density (OD) of each group of cells at 450 nm was measured and calculated with a microplate meter. Flow Cytometry Assay A flow cytometry assay was performed to assess cell apoptosis. Cells during the logarithmic growth were inoculated in 96-well plates. After different treatments, cells were collected and washed twice. Then, 5 μL Annexin V and propidium iodide (PI) were added into the cells with gentle blending. Cell apoptosis was detected by flow cytometry after 15 minutes of incubation at room temperature. Enzyme-Linked Immunosorbent Assay An enzyme-linked immunosorbent assay (ELISA) was done to detect the concentrations of TNF-α and IL-6 in both serum and cell supernatant. The ELISA kits for TNF-α and IL-6 were purchased from tBlue Gene Biotech (Shanghai, China). Luciferase Reporting Assay LncBase Predicted v.2 (Sun Yat-sen University, Guangzhou, China) predicted the binding sites between lncRNA UCA1 and miR-182-5p, which was verified with a luciferase reporting assay. The wild type (wt) or mutant type (mut) seed region in UCA1 was synthesised by Gene Pharma (Shanghai, China), and cloned into the luciferase reporter vector psiCHECK-2 (Promega, Madison, WI, USA). The vector was then transfected into A549 cells together with miR-182-5p mimic or inhibitor using lipofectamine 2000. After 48 hours of culture, luciferase activity was tested under a microplate reader (Molecular Devices, San Jose, CA, USA). Renilla luciferase was applied for normalisation. Statistical Analysis Statistical analysis was performed using SPSS 17.0 (IBM, Armonk, NY, USA). Measurement data were expressed as mean and standard deviation (SD), and compared for the difference between groups using a one-way analysis of variance. A chi square test was performed to analyse differences in the categorical variables. Correlation analysis was done using Pearson’s correlation analysis. Diagnostic ability was assessed by establishing a receiver operating characteristic (ROC) curve. A p-value <0.05 was considered to be statistically significant. Results Basic Clinical Information of the Study Population A total of 280 patients who underwent CPB surgery were collected; of these, 85 developed ARDS after surgery. All patients were placed in either the control group or the ARDS group, based on the onset of ARDS; clinical information was compared between the two groups. As shown in Table 1 , the two groups were age- and sex-matched, and there was no significant difference in body mass index, indicating that the two groups were comparable. In addition, other conditions, including C reactive protein levels, reasons for surgery, preoperative cardiac function grading, operative category, and aorta blocking time, were also compared, and no significant differences were detected (Table 1). However, patients in the ARDS group had a long bypass time than those without ARDS (p<0.001).Table 1 Basic clinical information of the study population. Controls (n=195) Patients with ARDS (n=85) P Sex 0.996 Male 94 (48.2) 41 (48.2) Female 101 (51.8) 44 (51.8) Age (yr) 57.08±11.01 59.08±9.55 0.146 Weight (kg) 64.66±6.00 65.59±7.54 0.274 Height (m) 1.70±0.07 1.67±0.79 0.905 BMI (kg/m2) 23.21±1.86 23.55±2.14 0.181 CRP 14.59±4.13 15.41±4.07 0.130 Causes of surgery 0.920 Rheumatic heart disease with valvular involvement 115 (59.0) 48 (56.5) Coronary heart disease 74 (38.0) 34 (40.0) Dissecting aneurysms 6 (3.1) 3 (3.5) NYHA class 0.697 II 19 (9.7) 6 (7.1) III 173 (88.7) 77 (90.6) IV 3 (1.5) 2 (2.3) Operative category 0.996 CABG 12 (6.1) 6 (7.1) Heart valve replacement 141 (72.3) 70 (82.3) Great vessels surgery 34 (17.4) 8 (9.4) Others 8 (4.1) 1 (1.2) CPB intraoperative situation Bypass time (min) 99.08±28.97 112.99±29.31 <0.001 Aorta blocking time (min) 67.31±26.69 73.19±26.15 0.090 Data are provided as n (%) or mean ± standard deviation. Abbreviations: ARDS, acute respiratory distress syndrome; BMI, body mass index; CRP, C-reactive protein; NYHA, New York Heart Association; CABG, coronary artery bypass grafting; CPB, cardiopulmonary bypass. Long Non-Coding RNA Urothelial Carcinoma Associated 1 Levels in the Serum Patients with Acute Respiratory Distress Syndrome Patients After Cardiopulmonary Bypass As shown in Figure 1 , qRT-PCR revealed a gradual increase in serum UCA1 in both control and ARDS cases. However, at any time point before and after surgery, the serum UCA1 levels in the ARDS group were statistically significantly higher than in the control group (p<0.001).Figure 1 Long non-coding RNA urothelial carcinoma associated 1 (UCA1) levels in the serum of patients with acute respiratory distress syndrome (ARDS) after cardiopulmonary bypass. ∗p<0.05, ∗∗p<0.01, and ∗∗∗p<0.001 vs the control group. Correlation of Serum Urothelial Carcinoma Associated 1 with Inflammatory Cytokines Tumour necrosis factor-α and IL-6 concentrations were detected in the serum of study patients, to evaluate the inflammatory response. The ELISA demonstrated an increasing trend for TNF-α and IL-6 release in the control and ARDS groups, in a time-dependent manner (Figure 2 A, B). Patients with ARDS had higher levels of TNF-α and IL-6 than those in the control group (Figure 2A, B). In addition, Pearson’s correlation analysis was done for the correlation analysis between UCA1 and TNF-α and IL-6 in the serum of patients with ARDS 8 hours after CPB. Serum UCA1 was positively correlated with serum TNF-α (r=0.723, p<0.001; Figure 2C) and IL-6 concentration (r=0.800, p<0.001; Figure 2D).Figure 2 Correlation of serum urothelial carcinoma associated 1 (UCA1) with inflammatory cytokines. Patients with acute respiratory distress syndrome (ARDS) had increased levels of (A) tumour necrosis factor (TNF)-α and (B) interleukin (IL)-6 vs the control group. Serum UCA1 was positively correlated with serum (C) TNF-α and (D) IL-6 concentration. ∗p<0.05, ∗∗p<0.01, and ∗∗∗p<0.001. Diagnostic Value of Serum Urothelial Carcinoma Associated 1 Using serum UCA1 level as the test variable and the occurrence of ARDS as the state variable, a ROC curve was drawn to evaluate the diagnostic value of UCA1 for ARDS. As shown in Figure 3 , serum UCA1 can identify patients with ARDS from those who underwent CPB with the area under the curve of 0.934. The biggest Youden index was obtained at cut-off value of 1.211, with a sensitivity of 76.5% and a specificity of 96.4%.Figure 3 Serum urothelial carcinoma associated 1 can identify patients with acute respiratory distress syndrome from those who underwent cardiopulmonary bypass with the area under the curve (AUC) of 0.934. Urothelial Carcinoma Associated 1 Inhibition Protects Against Lung Injury and Inhibits Cell Inflammation in vitro OGD/R-induced cell models were established for in vitro experiments. Similarly to the results observed in patients with ARDS, high levels of UCA1 were also observed in the cell models, which was reversed by si-UCA1 transfection (Figure 4 A). Cell counting kit-8 (CCK-8) demonstrated the cell viability inhibition induced by OGD/R treatments, and si-UCA1 rejuvenated cell viability (Figure 4B). In addition, OGD/R-induced cell apoptosis was recovered by UCA1 block down (Figure 4C). Cell inflammation was also evaluated. As shown in Figure 4(D), OGD/R promoted the release of both TNF-α and IL-6, and these effects were cancelled out by si-UCA1. The results demonstrated that si-UCA1 inhibited OGD/R-induced cell apoptosis and the release of TNF-α, but it did not recover to the levels of untreated cells.Figure 4 Urothelial carcinoma associated 1 (UCA1) inhibition protects against lung injury and inhibits cell inflammation in vitro. (A) Transfection with small-interfering UCA1 RNA (si-UCA1) inhibits the increased trend of UCA1. (B) si-UCA1 rejuvenated the cell viability inhibition induced by OGD/R. (C) OGD/R-induced cell apoptosis was also recovered by UCA1 block down. (D) si-UCA1 inhibits tumour necrosis factor (TNF)-α and interleukin (IL)-6 release. ∗p<0.05, ∗∗∗p<0.001. Abbreviation: si-NC, negative control. Urothelial Carcinoma Associated 1 May Function Through Sponging miR-182-5p The binding sites between UCA1 and miR-182-5p were predicted by LncBase Predicted v.2 (Figure 5 A). In cells transfected with wt-UCA1, the luciferase assay demonstrated weakened cell luciferase activity, which co-transfected with the miR-182-5p mimic, while stronger luciferase activity was detected in cells co-transfected with miR-182-5p inhibitor (Figure 5B). However, levels of miR-182-5p did not influence the luciferase activity of cells transfected with mut-UCA1. In addition, the qRT-PCR revealed the downregulation of miR-182-5p in OGD/R-induced cell models, and miR-182-5p levels were elevated after UCA1 downregulation (Figure 5C).Figure 5 Urothelial carcinoma associated 1 (UCA1) functions through sponging miR-182-5p. (A) Binding sites between UCA1 and miR-182-5p. (B) MiR-182-5p mimics weakened cell luciferase activity. (C) miR-182-5p was downregulated in OGD/R-induced cell models, and elevated after small-interfering UCA1 RNA (si-UCA1) transfection. ∗∗∗p<0.001. Abbreviations: si-NC, negative control; wt, wild type; mut, mutant. Discussion With the rapid development of surgical technology, CPB has become increasingly mature. At present, CPB is widely used in clinical practice [24]. However, according to investigation and statistics, the fatality rate of CPB is still as high as 1.7%–3%, and most patients died from respiratory failure, especially ARDS [25,26]. Acute respiratory distress syndrome caused by CPB is a special form of ARDS. Domestic studies have reported that the incidence of ARDS after CPB is as high as 2.5%, leading to a related mortality rate of about 68% [3]. The duration of bypass is known to increase postoperative complications, including ARDS. The current baseline clinical information demonstrated that patients with ARDS had a longer duration of bypass than those without ARDS, revealing its potential association with the occurrence of ARDS. Existing studies indicate that using biomarkers can improve the early diagnosis and treatment of ARDS [27]. It can delay or even prevent the progression of high-risk patients to ARDS, which is the key to reducing the incidence and mortality of ARDS, and improving its diagnosis and treatment level [28]. Biomarkers can stratify disease risk under specific conditions, which are of great significance in disease prediction and treatment. As the most widely studied, lncRNA has been proven in recent years to be associated with the pathogenesis of many diseases [29]. Existing studies have shown that lncRNA plays a broad regulatory role in epigenetic and other aspects, and increasing evidence has shown that lncRNA can be used as a biological marker for the diagnosis and treatment of diseases [30]. In the present study, serum UCA1 levels gradually increased in patients after receiving CPB, and patients with ARDS had high levels of UCA1 in comparison to non-ARDS patients at the same time point. In addition, the ROC curve demonstrated the diagnostic ability of serum UCA1 for ARDS. However the retrospective nature of our study made it difficult to collect information of the ARDS stage, as well as the association of serum UCA1 with disease stage. This was a limitation of our study, and should be considered in future research. In addition, a significantly long duration of bypass was detected in patients with ARDS, and the causal relationship between bypass time and UCA1 levels needs further verification. The main causes of CPB-induced ARDS include ischaemia–reperfusion-induced injury and activation of systemic inflammatory response [31]. Previous studies have confirmed that the levels of TNF-α and IL-6 are increased in CPB-induced lung injury or ARDS [32]. The level of inflammation may be an important indicator in evaluating the occurrence and development of ARDS [33]. The current ELISA assay consistently demonstrated an increasing trend for TNF-α and IL-6 release in both control and ARDS groups in a time-dependent manner, and patients with ARDS had increased levels of TNF-α and IL-6 vs the control group. The findings reflect the close association between a severe inflammatory response and the development of ARDS. In addition, many studies have used TNF-α and IL-6 levels as biological indicators for diagnosis, prediction, and evaluation of lung injury or ARDS [34]. The Pearson correlation analysis in this study demonstrated a close correlation between UCA1 and TNF-α and IL-6 in the serum of patients with ARDS 8 hours after CPB. It is known that the number and function of alveolar epithelial cells are the functional and structural basis of the lung, and also play a role in the repair of lung injury [35]. In the current study, OGD/R-induced A549 cell models were established for in vitro experiments. Similar to the results observed in patients with ARDS, high levels of UCA1 were also observed in cell models. Besides, cell transfection was performed to regulate UCA1 levels in the cell models. The gain- and loss-of-function experiments demonstrated that UCA1 knockdown protected against OGD/R-induced cell apoptosis and inflammatory response. The findings led us to conclude that UCA1 is an important factor in the development of ARDS. UCA1 has been reported to regulate the occurrence and development of human diseases by acting as a molecular sponge of miRNA [36]. Previous evidence and the present findings have highlighted the target relationship between UCA1 and miR-182-5p. The present cell experiments showed the downregulation of miR-182-5p in OGD/R-induced cell models. Consistently, miR-182-5p has been widely reported to be involved in lung disease, such as acute lung injury and lung cancer [20,21]. We speculated that miR-182-5p might be related to the regulatory role of UCA1 in the development of ARDS. In conclusion, patients who received CPB had increased levels of UCA1, and high levels of UCA1 were associated with the occurrence of postoperative ARDS. The regulatory role of UCA1 in ARDS might be related to inflammation and downregulated miR-182-5p in alveolar epithelial cells. Conflicts of Interest There are no conflicts of interest to disclose. Funding Sources This study was funded by Hebei Province Medical Science Research Youth Science and Technology Project (No. 20200182), S&T Program of Chengde (No. 202006A048）and Youth Science Foundation of Hebei Education Department (No. QN2019131). ==== Refs References 1 Ji X. Zhou P. Zhong L. Xu A. Tsang A.C.O. Chan P.K.L. Smart surgical catheter for C-reactive protein sensing based on an imperceptible organic transistor Adv Sci (Weinh) 5 2018 1701053 2 Liu X. Chen Q. Shi S. Shi Z. Lin R. Tan L. Plasma sRAGE enables prediction of acute lung injury after cardiac surgery in children Crit Care 16 2012 R91 22616947 3 Bashar F.R. Vahedian-Azimi A. Farzanegan B. Goharani R. Shojaei S. Hatamian S. Comparison of non-invasive to invasive oxygenation ratios for diagnosing acute respiratory distress syndrome following coronary artery bypass graft surgery: a prospective derivation-validation cohort study J Cardiothorac Surg 13 2018 123 30482210 4 Cai J. Sun H. Zheng B. Xie M. Xu C. Zhang G. Curcumin attenuates lncRNA H19 induced epithelial mesenchymal transition in tamoxifen resistant breast cancer cells Mol Med Rep 2021 23 5 Wu L. Deng L. Hong H. Peng C. Zhang X. Chen Z. Comparison of long noncoding RNA expression profiles in human dental follicle cells and human periodontal ligament cells Mol Med Rep 20 2019 939 950 31173189 6 Wan P. Su W. Zhang Y. Li Z. Deng C. Li J. LncRNA H19 initiates microglial pyroptosis and neuronal death in retinal ischemia/reperfusion injury Cell Death Differ 27 2020 176 191 31127201 7 Wang H. Zheng X. Jin J. Zheng L. Guan T. Huo Y. LncRNA MALAT1 silencing protects against cerebral ischemia-reperfusion injury through miR-145 to regulate AQP4 J Biomed Sci 27 2020 40 32138732 8 Meng X.Z. Jiang W.W. Guo H.W. Li M. Zhao M.M. Yu F.J. lncRNA NEAT1 regulates cerebral ischemia/reperfusion injury through miR-1306-5p J Biol Regul Homeost Agents 35 2021 239 244 33508927 9 Zhu T.T. Sun R.L. Yin Y.L. Quan J.P. Song P. Xu J. Long noncoding RNA UCA1 promotes the proliferation of hypoxic human pulmonary artery smooth muscle cells Pflugers Arch 471 2019 347 355 30353369 10 Zhang X. Tang X. Pan L. Li Y. Li J. Li C. Elevated lncRNA-UCA1 upregulates EZH2 to promote inflammatory response in sepsis-induced pneumonia via inhibiting HOXA1 Carcinogenesis 43 2022 371 381 35018436 11 Zhao Y.J. Chen Y.E. Zhang H.J. Gu X. LncRNA UCA1 remits LPS-engendered inflammatory damage through deactivation of miR-499b-5p/TLR4 axis IUBMB Life 73 2021 463 473 33368965 12 Miao Y.R. Liu W. Zhang Q. Guo A.Y. lncRNASNP2: an updated database of functional SNPs and mutations in human and mouse lncRNAs Nucleic Acids Res 46 2018 D276 D280 29077939 13 Cheng M. Wang Q. Chen L. Zhao D. Tang J. Xu J. LncRNA UCA1/miR-182-5p/MGMT axis modulates glioma cell sensitivity to temozolomide through MGMT-related DNA damage pathways Hum Pathol 123 2022 59 73 35219686 14 Wang W. Hu W. Wang Y. An Y. Song L. Shang P. Long non-coding RNA UCA1 promotes malignant phenotypes of renal cancer cells by modulating the miR-182-5p/DLL4 axis as a ceRNA Mol Cancer 19 2020 18 31996265 15 Amorim M. Lobo J. Fontes-Sousa M. Estevao-Pereira H. Salta S. Lopes P. Predictive and prognostic value of selected microRNAs in luminal breast cancer Front Genet 10 2019 815 31572437 16 Stittrich A.B. Haftmann C. Sgouroudis E. Kuhl A.A. Hegazy A.N. Panse I. The microRNA miR-182 is induced by IL-2 and promotes clonal expansion of activated helper T lymphocytes Nat Immunol 11 2010 1057 1062 20935646 17 Miller C.H. Smith S.M. Elguindy M. Zhang T. Xiang J.Z. Hu X. RBP-J-regulated miR-182 promotes TNF-alpha-induced osteoclastogenesis J Immunol 196 2016 4977 4986 27183593 18 Park J. Jeong S. Park K. Yang K. Shin S. Expression profile of microRNAs following bone marrow-derived mesenchymal stem cell treatment in lipopolysaccharide-induced acute lung injury Exp Ther Med 15 2018 5495 5502 29904430 19 Zhu M. Li Y. Sun K. MicroRNA-182-5p inhibits inflammation in LPS-treated RAW264.7 cells by mediating the TLR4/NF-kappaB signaling pathway Int J Clin Exp Pathol 11 2018 5725 5734 31949658 20 Lv S. Qu X. Qu Y. Wang Y. LncRNA NEAT1 knockdown alleviates lipopolysaccharide-induced acute lung injury by modulation of miR-182-5p/WISP1 axis Biochem Genet 59 2021 1631 1647 34046810 21 Yang W. Yin Y. Bi L. Wang Y. Yao J. Xu L. MiR-182-5p promotes the metastasis and epithelial-mesenchymal transition in non-small cell lung cancer by targeting EPAS1 J Cancer 12 2021 7120 7129 34729113 22 Force A.D.T. Ranieri V.M. Rubenfeld G.D. Thompson B.T. Ferguson N.D. Caldwell E. Acute respiratory distress syndrome: the Berlin Definition JAMA 307 2012 2526 2533 22797452 23 Yang K. Gao B. Wei W. Li Z. Pan L. Zhang J. Changed profile of microRNAs in acute lung injury induced by cardio-pulmonary bypass and its mechanism involved with SIRT1 Int J Clin Exp Pathol 8 2015 1104 1115 25972997 24 Hessel E.A. 2nd What's new in cardiopulmonary bypass J Cardiothorac Vasc Anesth 33 2019 2296 2326 30928282 25 Ho K.M. Tan J.A. Benefits and risks of maintaining normothermia during cardiopulmonary bypass in adult cardiac surgery: a systematic review Cardiovasc Ther 29 2011 260 279 20041882 26 Lin Y. Chen M. Peng Y. Chen Q. Li S. Chen L. Feeding intolerance and risk of poor outcome in patients undergoing cardiopulmonary bypass surgery Br J Nutr 126 2021 1340 1346 33468265 27 Wang Z.F. Yang Y.M. Fan H. Diagnostic value of miR-155 for acute lung injury/acute respiratory distress syndrome in patients with sepsis J Int Med Res 48 2020 300060520943070 28 Dong H. Li J. Lv Y. Zhou Y. Wang G. Hu S. Comparative analysis of the alveolar macrophage proteome in ALI/ARDS patients between the exudative phase and recovery phase BMC Immunol 14 2013 25 23773529 29 Han S. Liang Y. Li Y. Du W. Lncident: a tool for rapid identification of long noncoding RNAs utilizing sequence intrinsic composition and open reading frame information Int J Genomics 2016 2016 9185496 30 Schmitz S.U. Grote P. Herrmann B.G. Mechanisms of long noncoding RNA function in development and disease Cell Mol Life Sci 73 2016 2491 2509 27007508 31 Zheng X.M. Yang Z. Yang G.L. Huang Y. Peng J.R. Wu M.J. Lung injury after cardiopulmonary bypass: alternative treatment prospects World J Clin Cases 10 2022 753 761 35127892 32 Wang S.T. Bao C. He Y. Tian X. Yang Y. Zhang T. Hydrogen gas (XEN) inhalation ameliorates airway inflammation in asthma and COPD patients QJM 113 2020 870 875 32407476 33 Thiel M. Chouker A. Ohta A. Jackson E. Caldwell C. Smith P. Oxygenation inhibits the physiological tissue-protecting mechanism and thereby exacerbates acute inflammatory lung injury PLoS Biol 3 2005 e174 15857155 34 Hu J. Liu L. Zeng X. Wang K. Wang H. Zeng Z. Prognostic value of angiopoietin-like 4 in patients with acute respiratory distress syndrome Shock 56 2021 403 411 33900712 35 Meng P.Z. Liu J. Hu P.S. Tong F. Protective effect of dexmedetomidine on endotoxin-induced acute lung injury in rats Med Sci Monit 24 2018 4869 4875 30006502 36 Wang C.J. Zhu C.C. Xu J. Wang M. Zhao W.Y. Liu Q. The lncRNA UCA1 promotes proliferation, migration, immune escape and inhibits apoptosis in gastric cancer by sponging anti-tumor miRNAs Mol Cancer 18 2019 115 31272462	36463076	PMC9709611	NO-CC CODE	2022-12-01 23:23:08	no	Heart Lung Circ. 2022 Nov 30; doi: 10.1016/j.hlc.2022.10.008	utf-8	Heart Lung Circ	2,022	10.1016/j.hlc.2022.10.008	oa_other
==== Front Eur J Intern Med Eur J Intern Med European Journal of Internal Medicine 0953-6205 1879-0828 The Author(s). Published by Elsevier B.V. on behalf of European Federation of Internal Medicine. S0953-6205(22)00419-8 10.1016/j.ejim.2022.11.029 Review Article Reshaping care in the aftermath of the pandemic. Implications for cardiology health systems Jordan-Rios Antonio a Nuzzi Vincenzo b Bromage Daniel I a McDonagh Theresa a Sinagra Gianfranco b Cannata Antonio ab⁎ a Department of Cardiovascular Science, Faculty of Life Science and Medicine, King's College London, 125 Coldharbour lane, London SE5 9RS, UK b Cardiothoracovascular Department, Azienda Sanitaria Universitaria Integrata Giuliano Isontina (ASUGI), University of Trieste, Trieste, Italy ⁎ Corresponding author at: Department of Cardiovascular Science, Faculty of Life Science and Medicine, King's College London, 125 Coldharbour lane, London SE5 9RS, UK. 30 11 2022 30 11 2022 19 8 2022 12 10 2022 23 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. In the last two years, the COVID-19 pandemic has undeniably changed everyday life and significantly reshaped the healthcare systems. Besides the direct effect on daily care leading to significant excess mortality, several collateral damages have been observed during the pandemic. The impact of the pandemic led to staff shortages, disrupted education, worse healthcare professional well-being, and a lack of proper clinical training and research. In this review we highlight the results of these important changes and how can the healthcare systems can adapt to prevent unprecedented events in case of future catastrophes. ==== Body pmc1 Introduction Since the COVID-19 pandemic commenced more than two years ago, it has taken the lives of more than six million people around the globe [1]. In parallel with high mortality rates, health care delivery has been severely disrupted. A pandemic such as this has shown how vulnerable and unequal global health care systems are. It has also been accompanied by significant economic and social consequences [2,3]. Now it is understood that while limiting the rate spread of virus is important, the need to build strong and resilient health structures, that can offer essential services without delay, is also mandatory. In this review, we highlight the many disrupted areas in cardiovascular medicine and describe their impact on quality of care and mortality. Starting from the excess mortality occurred during the pandemic, the review will focus on the impact on different areas of health system, mainly staff and first contact personnel, the impact of the pandemic on the generation of scientific evidence (clinical trials) and on the effects on patient control and adherence to treatment. Finally, we review the implementation of different technological and logistical strategies to contain a possible future event. 1.1 Excess mortality Health systems across the globe have seen: (1) a decline in CV hospitalisations (mainly acute presentations), (2) the carrying out of fewer diagnostic and interventional procedures, and (3) fewer outpatient and community consultations [4,5]. Therefore, patients have presented later and sicker. All of these factors have resulted in more than a 5% excess in CV mortality (Fig. 1 ) [5].Fig. 1 Factors contributing to the excess mortality during the COVID-19 pandemic. Fig 1 The effect of COVID-19 on patients with cardiac comorbidities and the impact of acute cardiovascular events in patients affected by respiratory failure due to COVID-19 is well established [6], [7], [8], [9], [10]. However, another important issue that emerged during the pandemic was the increased cardiovascular mortality due to an unprecedented disruption of health systems [11,12]. For cardiovascular death, an up to 3-fold excess mortality was identified, especially during the first lockdown, which impacted most severely upon patients with existing cardiovascular disease, and this was maintained through 2021 [13]. The impact in Europe has been estimated at 31,000 to 62,000 excess deaths with a relative risk of 1.5 to 2.013. In the United States of America (USA), excess mortality caused by COVID-19 was observed specifically in the subgroup of patients with a history of ischaemic and hypertensive heart disease [14]. During the COVID-19 pandemic, there was a significant decrease in acute admissions for cardiovascular diseases across all European countries. Specifically, hospitalizations decreased by 31% for acute coronary syndromes, 34% for acute heart failure, and 32.3% for arrhythmias [15]. Only pulmonary embolism admission and out of hospital cardiac arrests were more common during the COVID-19 outbreak [5,16]. When compared to 2019, those admitted to the emergency department had a much higher mortality risk (4 times higher death risk) during the COVID-19 outbreak, which was not directly related to COVID-19 [15]. This generalised increase in cardiovascular mortality is considered to be due to several factors: firstly, a general reluctance on the part of patients to go to hospitals for fear of becoming infected; secondly, a reduction in health care staff availability either because of the high rate of infection or the assignment of staff to attend areas dedicated to COVID and thirdly, the cancellation of follow-up consultations, as well as necessary investigations or procedures [11,17]. Not only excess mortality, but also the incidence of out-of-hospital cardiac arrest was a notable problem. In an analysis of the Lombardia Cardiac Arrest Registry (CARe), Baldi et al. reported 362 cases of out-of-hospital cardiac arrest during 2020 compared to 229 cases during the same period in 2019 (58% increase) [16]. The incidence of out-of-hospital cardiac arrest at home, as well as unwitnessed cardiac arrest was also higher (7.3% and 11.3% higher, respectively). The magnitude of the disruption to health systems was evident as in the same study, the average arrival time for emergency services was 3 min slower compared to 2019, and the proportion of patients receiving cardiopulmonary resuscitation was 15.6% lower than in the year prior to the pandemic [16]. 1.2 Staff shortage and mental health One of the immediate effects of the coronavirus pandemic was the absence of staff and emotional damage within the workforce, with the consequences now being palpable [18]. The absence of staff (either through contagion or emotional stress) and the lack of sufficient supply of personal protective equipment contributed to the increased rate of infection, death and spread of the disease [19]. Elements such as access to personal protective equipment, health and safety protocols, and adequate rest and recovery periods determined how effectively health workers treated patients [20]. Although there are specific protocols to mitigate medical staff shortages [21], the situation in each country was highly variable. In the case of the Italian health system, students in their final years were allowed to join hospitals to help (Calabria decree). In the case of the USA, medical students in advanced stages were permitted to graduate earlier to assist in patient care [22]. In some other hospitals, medical specialities unrelated to critical respiratory management were even requested to provide continuous emergency care [21]. Incorporating diverse specialities into these patients' care helped obtain favourable outcomes, but with the knock-on effect of shortages in other routine services and staff burn-out [20,23,24]. These manifestations indicate that staffing shortages were severe and that there is no perfect solution to resolve them. Moreover, hospitals across the USA, Singapore and the UK implemented various solutions to staff shortages. For example, staff were divided into different non-contact groups with alternating work weeks, which reduced the risk of exposure and contact with each other and with the patient, while at the same time providing a reasonable interval of time to rest [[25], [26], [27]] as well as reducing the number of screening tests. Today, in the age of vaccination, hospitals around the globe and national health services will need to develop innovative systems that incorporate hybrid models capable of delivering care without depleting their own human and material resources. Together with staff shortages, the maintenance of strong mental health was crucial. Mental health of staff during the pandemic was critically influenced by the unprecedented scenario. According to a survey conducted by the British health service (NHS), burn-out and staff shortage impacted the mental well-being of health care professionals. Almost 40% of those reported some form of emotional exhaustion [24]. Hence, medical staff shortages and loss of emotional well-being may impact upon the quality of services provided, as the proportion of staff who feel that they provide a quality service is lower compared to 2020 (68% vs 74%) due to some degree of emotional exhaustion from lack of sufficient support in performing routine tasks [28,29]. Furthermore, The high rate of infection among health workers, who witnessed an increased number of deaths and becoming potential carriers of the virus for others helped to perpetuate a sense of anxiety, despair and uncertainty about the future. Understanding how the changes occurred during the pandemic have influenced healthcare professionals’ mental health may help in preventing future burnouts and promote a better care. 1.3 Collateral damage to medical education in the pandemic It is estimated that students from more than 2400 medical schools worldwide were directly or indirectly affected in the initial months of the pandemic [30,31]. Unlike other sciences, the teaching of medicine is delivered, in a hands-on fashion, during tutorials, in direct contact with patients and in a face-to-face manner. In cardiology, the traditional two-way teaching model (consultant-student, i.e., ward-rounds and clinics) and (patient-student: bedside and clinical appointments) meant that the normal learning flow was halted and required the implementation of technologies such as videoconferencing. Despite this, fellows were forced to temporarily put aside their cardiology studies to learn about microbiology, critical care and palliative medicine, which presented a significant intellectual burden in addition to the already strenuous frontline duty [32]. On average, it is estimated that more than 95% of cardiology fellows were affected by a substantial change in their training programs [[31], [32], [33]] causing a significant impact on their training, mainly regarding clinical activities. A suspension of clinical rotations and face-to-face sessions were paralleled by a decrease in the number of procedures fellows needed to develop a particular skill [[33], [34], [35]]. Those training in subspecialties such as echocardiography, interventional cardiology and electrophysiology [31] had difficulties obtaining the number of cases or procedures considered sufficient to achieve competencies, in part because of an environment without full supervision by senior physicians and by the number of patients reduced to the minimum necessary within the hospital [36,37]. Two different studies conducted among fellows in various interventional cardiology departments in New York City reported a moderate to severe impact on more than 90% of respondents. It highlights that the damage from the pandemic impacted in the short term, particularly in those areas of expertise that require achieving a competency-based on performance of a specific number of procedures [28,29]. It is estimated that for a 1-year training program, suspension or postponement of procedures by only 1 to 3 months reduces a fellow's volume of experience by 10–25% [37]. As a solution, extensions to the number of days required to obtain these competencies were offered. In some places in the USA, the academic program director was allowed to decide on the competency of the fellow [22]. However, we should note that already proven educational strategies such as medical simulator training was implemented fully due to the face-to-face restrictions that were in operation (23). Tools like telemedicine and home-based work compensated for this collateral damage. In some institutions, fellows contacted patients from home (telemedicine) and remotely interpreted studies such as echocardiograms [17,32]. As is the case with the emergence of new challenges, there were some unique opportunities to review topics not often used in standard practice. Thrombolysis for acute coronary syndromes (dosage, indications, complications) [33] or discussion on concepts of end-of-life ethics and mechanical ventilation strategies. A particular phenomenon happened within the research field among cardiology fellows. On the one hand, a very low proportion of them expressed the desire to pursue an academic career (<1%) [31] but paradoxically, it was during the first two waves of COVID that research activities became relevant as it was a suitable time for the fellows to write manuscripts, do statistical analyses and to discuss results and hypotheses [17,32]. During the pandemic, the way of spreading knowledge turned to videoconference and webinars platforms that largely replaced face-to-face sessions and scientific meetings. There are two schemes for delivering knowledge remotely: synchronous and asynchronous. In the former, participants are connected live, while the latter allows the recording of academic material and makes it available online for consultation with the flexibility of time and space (OnDemand) [38]. A recent meta-analysis demonstrated higher overall satisfaction with online vs traditional teaching (mean difference 0.60, 95% CI 0.38 to 0.83; p<0.001) [39]. However, the population studied consisted of medical students, not cardiology fellows. It is likely that satisfaction would be lower fot the latter because clinical and hands-on interaction plays such a crucial role in cardiology training. Webinars became one of the leading educational tools [32,33]. Some of their advantages are: accessibility from anywhere in the world, without the need to travel, lower costs compared to a face-to-face meeting, and they can be recorded for later consultation by attendees [35]. It is also possible to obtain a more significant number of attendees, as even those isolated by the virus can attend. Interestingly, e-learning, at least in the experience of 1 centre, allowed fellows to venture into areas other than cardiology (pulmonology, critical care medicine and anaesthesiology) [37]. Not least, the mental health of the fellows was also affected in more than 2/3rds [31], mainly due to the loss of the barrier between work/study and home. The teaching model shifted to a 100% digital format. With it, the structure of a timetable and a work schedule was lost; it demanded being connected all the time, which brought severe burn-out problems and emotional breakdown [37]. The professional frustration towards performing roles not typically practised by cardiology fellows (elements of critical care medicine, infection control, advanced airway management, disaster medicine) was notorious. In several centres, cardiovascular care units were transformed into COVID wards [37]. The lag in completing the curriculum brought anxiety and uncertainty about the fellows' future (post-graduate offers). Of note, more than half felt this lag could not be made up in the following years of their training [36]. The COVID-19 disruption in medical training highlighted the dissatisfaction of many fellows with the current curricula. They report being insufficiently prepared with the existing programmes [31]; this may provide an opportunity to develop a hybrid curriculum toward a competency-based system [34] where even the fellows participate in the design of the new academic structure. Research activities should play a central role in the training of cardiologists, as the generation of novel ideas and projects that advance the science of cardiology depends on it. To this end, it is crucial to structure mixed programs that, using digital tools, can create sufficient competence for fellows and relieve them of unnecessary activities so that they can use that time for research. In other words, it is necessary to consider whether aspects such as the duration and structure of the program itself are sufficient. 2 Clinical trials Cardiology clinical research has gone through radical changes in the last two years [40]. From a clinical point of view, many cardiologists put their efforts into the fight against the pandemic in COVID-19 units, consequently reducing the possibility of devoting time to clinical research. On the other hand, several trials were abandoned in favour of studies assessing specific therapies for patients with COVID-19 respiratory failure [41]. Furthermore, many activities such as patient screening, randomisation, follow-up visits, follow-up blood tests and event adjudications were frequently performed remotely. The missing in-person visits may affect the quality of the evaluation. Indeed, only patient-reported information can be collected on the phone. At the same time, clinical signs of ongoing cardiovascular problems, such as congestion and hypoperfusion, cannot be precisely detected without a focused clinical and instrumental assessment [42]. Moreover, the documented reduction in the admission rates for acute cardiovascular diseases may underpower any trial due to a reduced endpoint rate compared to the planned rate [43]. Stunning evidence emerged from the GUIDE-HF trial's results, which aimed to assess the utility of remote monitoring of pulmonary artery pressure in patients with chronic heart failure. The enrolment was concluded in December 2020. Overall, the trial result was negative, as the incidence of the primary HF outcome was not different between the two groups. However, a dramatic reduction in HF hospitalisations was noted during the pandemic, and a pre-specified analysis including only the pre-pandemic period showed a significantly lower incidence of the primary outcome in the arm treated with the investigational device [44]. These results might be explained both by a global reduction in the admissions for HF and by a change in the behaviour of the patients (reduced causes of acute decompensated HF such as respiratory tract infections, healthier lifestyle, more precise self-adjustment of daily diuretic dose, etc.) [45,46]. The consequences of all these deficiencies in conducting clinical trials in cardiology might dramatically impact the development of new therapies. Indeed, on the one hand, the interruption of ongoing trials, alongside the economic loss, prevents investigational therapies from being adequately studied. On the other hand, negative trials due to COVID-19 effects on the outcomes assessment might lead to incorrect conclusions [47]. Hence, specific solutions were suggested to avoid underpowering of clinical trials in cardiology due to the COVID-19 pandemic: Firstly, more accurate remote assessment of potential endpoints (home weight, self-reported symptoms, devices for arrhythmias, home blood pressure monitoring) [48]. Secondly, meticulous pursuit of the potential endpoint by the investigators, even though it might involve more work. Thirdly, statistical adaptation may be useful to face the possible underestimation/underrepresentation of clinical endpoints. Another study hit by the pandemic was the AFFIRM AHF trial, from the analysis of which we can draw some solutions. In order to analyse the impact that the loss of patients to follow-up could have, a sensitivity analysis had to be performed prior to COVID-19 [49]. Other measures, such as electronic data capture and home visits by clinical trial conductors (CROs), help minimise interruptions to follow-up as used in the REVIVED trial [50]. It is essential to mention that these statistical strategies, along with others such as sub-analysis studies, only partially help to resolve trial problems during pandemics. Behind GUIDE-HF trial, many other trials were also disrupted such as PARADISE-MI, IAMI and Dal-GenE. Important statements released by EMA, FDA and NIH during the pandemic should serve as a guide to interpret and conduct them. Estimating the decisive impact of the pandemic on the results of a multicentre study is very challenging as the decrease in admission rates observed globally may dilute the ability to observe treatment-specific differences. COVID-19 also impacted poor-quality research. In a relevant analysis by Glasziou, it is clear that within all the clinical trials registered for COVID-19, there were essential flaws in the design, sample size, presence of a control group and multicentre scope; likewise, the measured relays were not homogeneous. The distribution and reach of the preprints led to misinformation to the public with erroneous conclusions, such as in the case of hydroxychloroquine [51]. The pandemic also exacerbated unnecessary duplication of studies. For example, a large number of clinical trials were simultaneously registered for hydroxychloroquine that turned out to have no clinical benefit. This highlighted a limited global infrastructure for communication and collaboration in hypothesis generation that, in turn, led to duplication and wasted content, at least during the first year of the pandemic. 2.1 Risk control in patients In addition to the excess mortality and its undeniable impact on health systems, the pandemic has also affected human lifestyles, particularly in the most vulnerable patients who, unable to continue receiving the same level of care as before, experience declines in their health and long-term prognosis [52]. Sedentary individuals have twice the relative risk of developing coronary events as physically active individuals [53]. Moderate exercise is recommended on most days of the week for a minimum of 150 min per week for both primary and secondary prevention [54,55]. The globally mandated isolation measures and social distancing contributed to people staying at home and adopting a more sedentary lifestyle than before. Gyms and sports facilities closed their doors. There is diverse evidence documenting a significant decrease in physical activity during major confinements; on average, from March to April 2020, there was a 45.2% decrease in physical activity reported in papers from the UK, USA, Australia and Poland [56]. This trend appears to be replicated in paediatric and adolescent populations (5 to 13 years of age), with information coming from parental reports [57]. Finally, a global study collected data from 455,404 patients on their step count and found a 27.3% decrease in mean daily steps 30 days after lockdown initiation [58]. The pandemic led to changes in dietary habits [59] and an increase in junk food consumption [60]. These changes are worrying given the risks associated with physical inactivity, unhealthy eating and consequent weight gain. We still don't know the long-term outcomes. However, the excess mortality rate recorded during the second year of the pandemic (2021) may partly be explained by the vulnerability of patients as they increase their cardiovascular risk factors and experience disruption in the provision of services to control them. It is also important to note that excess mortality may be related to the negative impact of lockdown. It has been shown that there is a substantial psychological impact of confinement on lifestyle-related risk factors. Not only that, highly stressful and demoralising events represent a risk factor for acute coronary syndromes [61]. Therefore, we must consider for future epidemics that lockdown must be carefully managed. 2.2 The importance of continuity The coronavirus pandemic event demonstrated the need to create more flexible health systems to continue delivering services despite severe disruptions. Detaining or deferring patient care has serious consequences: increased in-hospital mortality and increased mortality at home and in nursing homes, particularly for conditions such as ST-segment elevation myocardial infarction and heart failure [4,11,62,63]. Those patients who did make it to the hospital presented with more severe disease [4,64]. This needs to be countered with confusion-avoidance messages from authorities emphasising that patients should not defer care in case of alarming symptoms. Notable efforts such as the European Society of Cardiology (ESC) campaign “You can't pause a heart” seek to balance staying at home with prompt transfer to a hospital in case of an emergency [65]. The degree of disruption caused by the pandemic on the healthcare system was never anticipated, and its consequences are only beginning to be visible. Deferring interventional procedures and failing to conduct necessary clinical trials leaves patients with significant residual risk and contributes to excess mortality [66,67]. Conversely, strategies such as e-health and the adaptation of new protocols in imaging studies and interventional procedures helped maintain continuity in healthcare systems [4]. This crisis served as a reminder of the value of primary healthcare in order to manage unanticipated spikes in demand and ensure everyone receives continuity of care. A successful response from the health system is facilitated by strong primary health care, organised in multidisciplinary teams with innovative roles for health professionals, integrated with community health services, outfitted with digital technology, and working with well-designed incentives [68]. A robust primary care setting allows patients to receive medical attention without attending busy hospitals and helps maintain continuity of care while reducing pressure on the entire health system [69]. Part of the strategies included in strengthening the primary care setting is: establishing teams with solid links to the communities, expanding home-based programmes, allowing digital tools to communicate primary with secondary levels and effectively sharing clinical information inside the country [69]. Making healthcare systems more robust against future public health emergencies, preserving the innovations developed and implemented during the pandemic is necessary. Indeed, after the crisis' acute phase, there will be a massive wave of common chronic diseases that will cause death and disability, with cardiometabolic disorders riding the peak of the wave [70]. Primary care should be strengthened to facilitate access to health services during a pandemic, especially for vulnerable patients such as those with chronic diseases [70]. Also, primary care needs to be appreciated as the cornerstone of health systems. More novel resources must reach primary care to ensure medical follow-up during rough times. Remote telemonitoring stands as a novel approach that might help not only in keeping with uninterrupted health access but also in the interplay between primary and secondary settings. Nevertheless, we also need clinical trials to support the evidence that remote appointments are equally beneficial to close in-person follow-up 2.3 Telemedicine and e-health The digital transformation that has taken place in cardiology has helped to provide a continuum of care in cardiovascular health. Incorporating telemedicine into outpatient settings with a particular focus on risk factor management and implementing rehabilitation activities within the patient's home demonstrated improved patient adherence with outcomes similar to in-person visits [[71], [72], [73]]. In addition, remote cardiac care proved to be convenient for both patients and physicians, offering essential advantages such as reduced travel and waiting times, avoidance of unnecessary transfers, reduced costs, less crowding of physicians and patients at hospital centres and flexible schedules (24/7), as well as allowing for proper medication reconciliation [74]. Paradoxically, the pandemic also revealed a possible overuse of invasive procedures that requires a more thoughtful approach in the future [75,68]. The conditions that were most easily managed and followed up with telemedicine were: heart failure, atrial fibrillation, and ischaemic heart disease [76]. Since 2010, a study has shown that between 36% and 63% of myocardial infarctions could be prevented by reducing risk factors alone [77]. In pandemic terms, risk factor control has never been more critical, as cardiovascular disease and its risk factors are associated with unfavourable outcomes in COVID-19 infection [78]. Moreover, this care modality was as satisfactory and comfortable for the patient and the cardiologist as face-to-face consultations [79]. Concerning heart failure, Xu et al.'s group demonstrated that patients who received either in-person or remote follow-up after hospitalisation had a lower 30-day risk of readmission, giving telemedicine an essential place as a cost-effective and valuable tool, especially in times of crisis in the health system [80]. Telemedicine has been used even before the pandemic in the field of arrhythmias and atrial fibrillation. During the pandemic, wearables and portable electrocardiography devices allowed physicians to detect abnormal rhythms in real-time with the possibility of early intervention and improved prognosis [81,82]. There are significant barriers that need to be considered when implementing an e-health strategy. Access to technology is not equal in all parts of the world, and some demographic groups are more disadvantaged than others. Elderly patients and those living in rural areas with inconsistent internet access [83] would be virtually excluded from receiving care; furthermore, communities with poor health services are also, in many cases, those with limited internet access, which, far from benefiting, would contribute to further widening the inequality gap [84]. Therefore, for telemedicine implementation to grow, new mechanisms must emerge to enable universal and intuitive access for doctors and patients. 2.4 Adapting new protocols Several protocols were developed around the world that focused on both more judicious selection of which patients would benefit most from the study or procedure and, at the same time, designed rapid acquisition protocols without compromising the quality of the study. The ORACLE protocol made it possible to obtain haemodynamic and pulmonary information through a rapid ultrasonographic sequence without compromising the quality of the measurements or the integrity of the healthcare personnel performing it in less than 25 min [85]. Appropriate patient selection using stricter criteria allowed for prioritised studies with an impact on clinical management decision [86], making it an effective strategy in periods of shortage of medical staff and protective equipment. In interventional cardiology, stricter criteria were also applied, based on the hospital occupancy rate during the peak of the pandemic. Only STEMI patients were admitted to the catheterisation laboratory and low-risk NSTEMIs were postponed [87]. On the other hand, the EAPCI gave a prominent role to thrombolysis as the strategy of choice in scenarios where immediate access to the catheterisation suite was not possible, due to lack of staff or bed saturation, and favoured ventriculography to assess ventricular function in all patients without prior echocardiography [88]. Furthermore, an impressive reduction in the number of elective procedures in cardiology departments was observed during the pandemic [89,90]. Similarly, a striking reduction in the number of electrophysiology procedures was observed during the pandemic waves [91]. Priority was given to those patients at higher risk of fatal events e.g. with complete heart block, generator change for pacemaker-dependent patients and recurrent life-threatening arrhythmias requiring ablation, at the expense of a reduced amount of procedures targeting stable patients [92]. Alongside all these features, all the healthcare workers must attempt to reduce the risk of contracting the infection both from the patient and from other colleagues. For this reason, appropriate personal protective equipment should be regularly worn, including eye protection [17]. 2.5 Future direction 2.5.1 Prolonged COVID-19 and cardiovascular sequelae According to the World Health Organisation (WHO), it is recognised that after the acute phase of infection, some subjects experience persistent symptoms not explained by any other reason; in fact, there is a code in the ICD for this new condition (UO9) [93]. There is consensus that patients who persist with cardiovascular symptoms 12 weeks post-infection may have prolonged COVID without a precise end date. Current studies show that after one year of diagnosis, only 22.9% of patients are symptom-free and that a slow recovery is related to the severity of the disease [94]. The long-term consequences of COVID-19 are not entirely clear. A previous study demonstrated that 42% of the patients had subtle right ventricular dysfunction after clinical recovery from COVID-19 [95]. Furthermore, a 12-month follow-up demonstrated an increased risk for major cardiovascular outcomes: stroke, arrhythmias, coronary heart disease and thromboembolism (HR: 1.52, 1.84, 1.75, 2.93, respectively) regardless of the prior existence of other risk factors [96]. Given the significant number of infected people worldwide (more than 355 million) and based on cardiovascular risk factors prevalence, the potential number of people affected in the long term may be enormous. This highlights that strategies should be designed to monitor long-term cardiovascular outcomes and prepare governments and health systems to cope with increased numbers of patients with cardiovascular disease, increased economic burdens and possibly a change in the life expectancy of the world's population [97,98]. 2.5.2 Towards new health systems The damage to the health systems caused by COVID-19 has been major. Pandemics such as the coronavirus are unlikely to be the last we will face. During the pandemic, unnecessary or excessive patient follow-up was identified [75] and also, for the first time, visits to emergency departments for non-urgent (non-relevant) conditions reached their lowest level [99], allowing a unique window of opportunity for systematic changes focused on adopting lower-cost strategies with greater reach [100]. For the most part, funding for health systems has been insufficient and keeps them in a perpetual state of vulnerability to future challenges. Although many countries have subsidised COVID-19 testing and treatment, better strategies must be implemented to ensure that people do not fall into poverty due to high out-of-pocket health expenditures [100]. The resilience of a health system lies mainly in ensuring health workers' physical, mental and economic protection [101]. Incorporating technology into the health system with regularity strengthens the healthcare environment, improves major health outcomes, and takes healthcare to another level: precision medicine [102]. The future involves hybrid environments capable of offering e-health under a more holistic concept: telemedicine and remote liaison with patients, continuous remote monitoring through wearables, adding artificial intelligence to routine decision making [103,69] as well as predicting major cardiovascular events with simple tools such as the electrocardiogram or biomarkers as old as the human voice itself (Fig. 2 ) [104].Fig. 2 Healthcare systems adaptations throughout the pandemic and future directions. Fig 2 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 COVID-19 Data repository by the center for systems science and engineering (CSSE) at Johns Hopkins University. 2 Abi Khalil C. Omar O.M. Al Suwaidi J. Taheri S. Aspirin use and cardiovascular outcome in patients with type 2 diabetes mellitus and heart failure: a population-based cohort study J Am Heart Assoc 7 21 2018 e010033 10.1161/JAHA.118.010033 3 Pan K.Y. Kok A.A.L. Eikelenboom M. The mental health impact of the COVID-19 pandemic on people with and without depressive, anxiety, or obsessive-compulsive disorders: a longitudinal study of three Dutch case-control cohorts Lancet Psychiatry 8 2 2021 121 129 10.1016/S2215-0366(20)30491-0 33306975 4 Nadarajah R. Gale C.P. Collateral cardiovascular damage during the COVID-19 pandemic Nat Rev Cardiol 19 2 2022 81 82 10.1038/s41569-021-00661-x 34857956 5 Cannata A. Bromage D.I. McDonagh T.A. The collateral cardiovascular damage of COVID-19: only history will reveal the depth of the iceberg Eur Heart J 42 15 2021 1524 1527 10.1093/eurheartj/ehab097 33624020 6 Iorio A. Lombardi C.M. Specchia C. Combined role of troponin and natriuretic peptides measurements in patients with Covid-19 (from the Cardio-COVID-Italy multicenter study) Am J Cardiol 167 2022 125 132 10.1016/j.amjcard.2021.11.054 35063263 7 Tomasoni D. Inciardi R.M. Lombardi C.M. Impact of heart failure on the clinical course and outcomes of patients hospitalized for COVID-19. Results of the Cardio-COVID-Italy multicentre study Eur J Heart Fail 22 12 2020 2238 2247 10.1002/ejhf.2052 33179839 8 Ameri P. Inciardi R.M. Di Pasquale M. Pulmonary embolism in patients with COVID-19: characteristics and outcomes in the Cardio-COVID Italy multicenter study Clin Res Cardiol 110 7 2021 1020 1028 10.1007/s00392-020-01766-y 33141251 9 Paris S. Inciardi R.M. Lombardi C.M. Implications of atrial fibrillation on the clinical course and outcomes of hospitalized COVID-19 patients: results of the Cardio-COVID-Italy multicentre study Europace 23 10 2021 1603 1611 10.1093/europace/euab146 34297833 10 Nuzzi V. Merlo M. Specchia C. The prognostic value of serial troponin measurements in patients admitted for COVID-19 ESC Heart Fail 8 5 2021 3504 3511 10.1002/ehf2.13462 34236135 11 Cannata A. Watson S.A. Daniel A. Impact of the COVID-19 pandemic on in-hospital mortality in cardiovascular disease: a meta-analysis Eur J Prev Cardiol 29 8 2022 1266 1274 10.1093/eurjpc/zwab119 34297822 12 Madjid M. Safavi-Naeini P. Solomon S.D. Vardeny O. Potential effects of coronaviruses on the cardiovascular system: a review JAMA Cardiol 5 7 2020 831 840 10.1001/jamacardio.2020.1286 32219363 13 Banerjee A. Chen S. Pasea L. Excess deaths in people with cardiovascular diseases during the COVID-19 pandemic Eur J Prev Cardiol 28 14 2021 1599 1609 10.1093/eurjpc/zwaa155 33611594 14 Wadhera R.K. Shen C. Gondi S. Chen S. Kazi D.S. Yeh R.W. Cardiovascular deaths during the COVID-19 pandemic in the United States J Am Coll Cardiol 77 2 2021 159 169 10.1016/j.jacc.2020.10.055 33446309 15 Sokolski M. Gajewski P. Zymliński R. Impact of coronavirus disease 2019 (COVID-19) outbreak on acute admissions at the emergency and cardiology departments across Europe Am J Med 134 4 2021 482 489 10.1016/j.amjmed.2020.08.043 33010226 16 Baldi E. Sechi G.M. Mare C. Canevari F. Out-of-hospital cardiac arrest during the Covid-19 outbreak in Italy N Engl J Med 383 5 2020 496 498 10.1056/NEJMc2010418 32348640 17 Welt F.G.P. Shah P.B. Aronow H.D. Catheterization laboratory considerations during the coronavirus (COVID-19) pandemic: from the ACC's interventional council and SCAI J Am Coll Cardiol 75 18 2020 2372 2375 10.1016/j.jacc.2020.03.021 32199938 18 Kinman G. Teoh K. Harriss A. Supporting the well-being of healthcare workers during and after COVID-19 Occup Med (Lond) 70 5 2020 294 296 10.1093/occmed/kqaa096 32428225 19 Bressan S. Buonsenso D. Farrugia R. Preparedness and response to pediatric COVID-19 in European emergency departments: a survey of the REPEM and PERUKI networks Ann Emerg Med 76 6 2020 788 800 10.1016/j.annemergmed.2020.05.018 32419713 20 Buonsenso D. De Rose C. Pierantoni L. Doctors' shortage in adults COVID-19 units: a call for pediatricians Eur J Pediatr 180 7 2021 2315 2318 10.1007/s00431-021-03995-3 33594541 21 (https://www.cdc.gov/coronavirus/2019-ncov/hcp/mitigating-staff-shortages.html) last access 28/11/2022. 22 Accreditation. Council for graduate medical education. Stage 2: increase clinical demands guidance. (https://acgme.org/COVID-19/Stage-2-Increased-Clinical-Demands-Guidance) last access 28/11/2022. 23 Guan W.J. Ni Z.Y. Hu Y. Clinical characteristics of coronavirus disease 2019 in China N Engl J Med 382 18 2020 1708 1720 10.1056/NEJMoa2002032 32109013 24 Deakin M. NHS workforce shortages and staff burnout are taking a toll BMJ 377 2022 o945 10.1136/bmj.o945 35410885 25 Mascha E.J. Schober P. Schefold J.C. Stueber F. Luedi M.M. Staffing with disease-based epidemiologic indices may reduce shortage of intensive care unit staff during the COVID-19 pandemic Anesth Analg 131 1 2020 24 30 10.1213/ANE.0000000000004849 32343514 26 Willan J. King A.J. Jeffery K. Bienz N. Challenges for NHS hospitals during covid-19 epidemic BMJ 368 2020 m1117 10.1136/bmj.m1117 32198166 27 Wong J. Goh Q.Y. Tan Z. Preparing for a COVID-19 pandemic: a review of operating room outbreak response measures in a large tertiary hospital in Singapore Can J Anaesth 67 6 2020 732 745 10.1007/s12630-020-01620-9 32162212 28 Gupta T. Nazif T.M. Vahl T.P. Impact of the COVID-19 pandemic on interventional cardiology fellowship training in the New York metropolitan area: a perspective from the United States epicenter Catheter Cardiovasc Interv 97 2 2021 201 205 10.1002/ccd.28977 32415916 29 Shah S. Castro-Dominguez Y. Gupta T. Impact of the COVID-19 pandemic on interventional cardiology training in the United States Catheter Cardiovasc Interv 96 5 2020 997 1005 10.1002/ccd.29198 32767717 30 Nicola M. Alsafi Z. Sohrabi C. The socio-economic implications of the coronavirus pandemic (COVID-19): A review Int J Surg 78 2020 185 193 10.1016/j.ijsu.2020.04.018 32305533 31 Strangio A. Leo I. Spaccarotella C.A.M. Effects of the COVID-19 pandemic on the formation of fellows in training in cardiology J Cardiovasc Med (Hagerstown) 22 9 2021 711 715 10.2459/JCM.0000000000001185 34009835 32 Narula N. Singh H.S. Cardiology practice and training post-COVID-19: achieving “normalcy” after disruption J Am Coll Cardiol 76 4 2020 476 479 10.1016/j.jacc.2020.06.036 32703519 33 Kadavath S. Hawwas D. Strobel A. How the COVID-19 pandemic has affected cardiology fellow training Am J Cardiol 151 2021 114 117 10.1016/j.amjcard.2021.03.052 34052015 34 Goldhamer M.E.J. Pusic M.V. Co J.P.T. Weinstein D.F. Can covid catalyze an educational transformation? Competency-based advancement in a crisis N Engl J Med 383 11 2020 1003 1005 10.1056/NEJMp2018570 32543795 35 Kailey B.S. Dev D. Sado D.M. Luther V. CardioWebinar: the evolution of digital education during the COVID-19 pandemic Heart 107 24 2021 2004 2005 10.1136/heartjnl-2021-319815 34675039 36 Rao P. Diamond J. Korjian S. The impact of the COVID-19 pandemic on cardiovascular fellows-in-training: a national survey J Am Coll Cardiol 76 7 2020 871 875 10.1016/j.jacc.2020.06.027 32561407 37 DeFilippis E.M. Stefanescu Schmidt A.C. Reza N. Adapting the educational environment for cardiovascular fellows-in-training during the COVID-19 pandemic J Am Coll Cardiol 75 20 2020 2630 2634 10.1016/j.jacc.2020.04.013 32304798 38 Papapanou M. Routsi E. Tsamakis K. Medical education challenges and innovations during COVID-19 pandemic Postgrad Med J 98 1159 2022 321 327 10.1136/postgradmedj-2021-140032 33782202 39 He L. Yang N. Xu L. Synchronous distance education vs traditional education for health science students: a systematic review and meta-analysis Med Educ 55 3 2021 293 308 10.1111/medu.14364 32881047 40 Park J.J.H. Mogg R. Smith G.E. How COVID-19 has fundamentally changed clinical research in global health Lancet Glob Health 9 5 2021 e711 e720 10.1016/S2214-109X(20)30542-8 33865476 41 Thornton J. Clinical trials suspended in UK to prioritise covid-19 studies and free up staff BMJ 368 2020 m1172 10.1136/bmj.m1172 32205354 42 Pellicori P. Shah P. Cuthbert J. Prevalence, pattern and clinical relevance of ultrasound indices of congestion in outpatients with heart failure Eur J Heart Fail 21 7 2019 904 916 10.1002/ejhf.1383 30666769 43 Psotka M.A. Abraham W.T. Fiuzat M. Conduct of clinical trials in the era of COVID-19: JACC scientific expert panel J Am Coll Cardiol 76 20 2020 2368 2378 10.1016/j.jacc.2020.09.544 33183511 44 Lindenfeld J. Zile M.R. Desai A.S. Haemodynamic-guided management of heart failure (GUIDE-HF): a randomised controlled trial Lancet 398 10304 2021 991 1001 10.1016/S0140-6736(21)01754-2 34461042 45 Bromage D.I. Cannata A. Rind I.A. The impact of COVID-19 on heart failure hospitalization and management: report from a Heart Failure Unit in London during the peak of the pandemic Eur J Heart Fail 2020 10.1002/ejhf.1925 46 Cannata A. Bromage D.I. Rind I.A. Temporal trends in decompensated heart failure and outcomes during COVID-19: a multisite report from heart failure referral centres in London Eur J Heart Fail 2020 10.1002/ejhf.1986 47 Anker S.D. Butler J. Khan M.S. Conducting clinical trials in heart failure during (and after) the COVID-19 pandemic: an Expert Consensus Position Paper from the Heart Failure Association (HFA) of the European Society of Cardiology (ESC) Eur Heart J 41 22 2020 2109 2117 10.1093/eurheartj/ehaa461 32498081 48 Perez M.V. Mahaffey K.W. Hedlin H. Large-scale assessment of a smartwatch to identify atrial fibrillation N Engl J Med 381 20 2019 1909 1917 10.1056/NEJMoa1901183 31722151 49 Ponikowski P. Kirwan B.A. Anker S.D. McDonagh T. Ferric carboxymaltose for iron deficiency at discharge after acute heart failure: a multicentre, double-blind, randomised, controlled trial Lancet 396 10266 2020 1895 1904 10.1016/S0140-6736(20)32339-4 33197395 50 Perera D. Clayton T. O'Kane P.D. Greenwood J.P. REVIVED-BCIS2 investigators. percutaneous revascularization for ischemic left ventricular dysfunction N Engl J Med 2022 10.1056/NEJMoa2206606 51 Glasziou P.P. Sanders S. Hoffmann T. Waste in covid-19 research BMJ 369 2020 1847 10.1136/bmj.m1847 52 Mattioli A.V. Sciomer S. Cocchi C. Maffei S. Gallina S. Quarantine during COVID-19 outbreak: changes in diet and physical activity increase the risk of cardiovascular disease Nutr Metab Cardiovasc Dis 30 9 2020 1409 1417 10.1016/j.numecd.2020.05.020 32571612 53 Sans S. Mediterranean diet, active lifestyle and cardiovascular disease: a recipe for immortality? Eur J Prev Cardiol 25 11 2018 1182 1185 10.1177/2047487318785745 29966434 54 Pelliccia A. Sharma S. Gati S. 2020 ESC guidelines on sports cardiology and exercise in patients with cardiovascular disease Eur Heart J 42 1 2021 17 96 10.1093/eurheartj/ehaa605 32860412 55 Fabris E. Sinagra G. Physical activity in older people: better late than never, but better early than late Heart 108 5 2022 328 329 10.1136/heartjnl-2021-320462 35165167 56 Pina A. Castelletti S. COVID-19 and cardiovascular disease: a global perspective Curr Cardiol Rep 23 10 2021 135 10.1007/s11886-021-01566-4 34410538 57 Dunton G.F. Do B. Wang S.D. Early effects of the COVID-19 pandemic on physical activity and sedentary behavior in children living in the U.S BMC Public Health 20 1 2020 1351 10.1186/s12889-020-09429-3 32887592 58 Tison G.H. Avram R. Kuhar P. Worldwide effect of COVID-19 on physical activity: a descriptive study Ann Intern Med 173 9 2020 767 770 10.7326/M20-2665 32598162 59 Zupo R. Castellana F. Sardone R. Preliminary trajectories in dietary behaviors during the COVID-19 pandemic: a public health call to action to face obesity Int J Environ Res Public Health 17 19 2020 10.3390/ijerph17197073 60 Ammar A. Brach M. Trabelsi K. Effects of COVID-19 home confinement on eating behaviour and physical activity: results of the ECLB-COVID19 international online survey Nutrients 12 6 2020 10.3390/nu12061583 61 Brooks S.K. Webster R.K. Smith L.E. Woodland L. The psychological impact of quarantine and how to reduce it: rapid review of the evidence Lancet 2020 912 920 10.1016/S0140-6736(20)30460-8 32112714 62 Cannata A. Bromage D.I. McDonagh T.A. COVID-19 and heart failure: the dark side of the moon Eur J Heart Fail 24 6 2022 1129 1131 10.1002/ejhf.2518 35481841 63 De Rosa S. Spaccarotella C. Basso C. Reduction of hospitalizations for myocardial infarction in Italy in the COVID-19 era Eur Heart J 41 22 2020 2083 2088 10.1093/eurheartj/ehaa409 32412631 64 Kiss P. Carcel C. Hockham C. Peters S.A.E. The impact of the COVID-19 pandemic on the care and management of patients with acute cardiovascular disease: a systematic review Eur Heart J Qual Care Clin Outcomes 7 1 2021 18 27 10.1093/ehjqcco/qcaa084 33151274 65 Cardiology ESo. You can't pause a heart. (https://www.cantpauseaheart.org/) last access 28/11/2022. 66 Moreno R. Diez J.L. Diarte J.A. Consequences of canceling elective invasive cardiac procedures during Covid-19 outbreak Catheter Cardiovasc Interv 97 5 2021 927 937 10.1002/ccd.29433 33336506 67 Prachand V.N. Milner R. Angelos P. Medically necessary, time-sensitive procedures: scoring system to ethically and efficiently manage resource scarcity and provider risk during the COVID-19 pandemic J Am Coll Surg 231 2 2020 281 288 10.1016/j.jamcollsurg.2020.04.011 32278725 68 Cannata A. Bromage D.I. McDonagh T. Cardiology after COVID-19: quo vademus? Eur Heart J Qual Care Clin Outcomes 6 3 2020 208 209 10.1093/ehjqcco/qcaa042 32379892 69 The Future of Cardiology A Paper Produced by the British Cardiovascular Society Working Group on The Future of Cardiology August 2020 2020 British Cardiovascular Society https://www.britishcardiovascularsociety.org/__data/assets/pdf_file/0010/21142/BCS-Future-of-Cardiology-17-Aug-2020.pdf 70 Califf R.M. Avoiding the coming tsunami of common, chronic disease: what the lessons of the COVID-19 pandemic can teach us Circulation 143 19 2021 1831 1834 10.1161/CIRCULATIONAHA.121.053461 33820441 71 Green J.R. Smith J. Teale E. Use of the confusion assessment method in multicentre delirium trials: training and standardisation BMC Geriatr 19 1 2019 107 10.1186/s12877-019-1129-8 30991945 72 Green B.B. Cook A.J. Ralston J.D. Effectiveness of home blood pressure monitoring, Web communication, and pharmacist care on hypertension control: a randomized controlled trial JAMA 299 24 2008 2857 2867 10.1001/jama.299.24.2857 18577730 73 Dalal H.M. Doherty P. McDonagh S.T. Paul K. Taylor R.S. Virtual and in-person cardiac rehabilitation BMJ 373 2021 n1270 10.1136/bmj.n1270 34083376 74 Yuan N. Pevnick J.M. Botting P.G. Patient use and clinical practice patterns of remote cardiology clinic visits in the era of COVID-19 JAMA Netw Open 4 4 2021 e214157 10.1001/jamanetworkopen.2021.4157 75 Berwick D.M. Choices for the “new normal” JAMA 323 21 2020 2125 2126 10.1001/jama.2020.6949 32364589 76 Mishra K. Edwards B. Cardiac outpatient care in a digital age: remote cardiology clinic visits in the era of COVID-19 Curr Cardiol Rep 24 1 2022 1 6 10.1007/s11886-021-01618-9 35029784 77 Kahn R. Robertson R.M. Smith R. Eddy D. The impact of prevention on reducing the burden of cardiovascular disease Circulation 118 5 2008 576 585 10.1161/CIRCULATIONAHA.108.190186 18606915 78 Bae S. Kim S.R. Kim M.N. Shim W.J. Park S.M. Impact of cardiovascular disease and risk factors on fatal outcomes in patients with COVID-19 according to age: a systematic review and meta-analysis Heart 107 5 2021 373 380 10.1136/heartjnl-2020-317901 33334865 79 Poppas A. Rumsfeld J.S. Wessler J.D. Telehealth is having a moment: will it last? J Am Coll Cardiol 75 23 2020 2989 2991 10.1016/j.jacc.2020.05.002 32475633 80 Xu H. Granger B.B. Drake C.D. Peterson E.D. Dupre M.E. Effectiveness of telemedicine visits in reducing 30-day readmissions among patients with heart failure during the COVID-19 pandemic J Am Heart Assoc 11 7 2022 e023935 10.1161/JAHA.121.023935 81 Varma N. Marrouche N.F. Aguinaga L. HRS/EHRA/APHRS/LAHRS/ACC/AHA worldwide practice update for telehealth and arrhythmia monitoring during and after a pandemic Europace 23 2 2021 313 10.1093/europace/euaa187 32526011 82 Linz D. Pluymaekers N. Hendriks J.M. TeleCheck-AF for COVID-19 Eur Heart J 41 21 2020 1954 1955 10.1093/eurheartj/ehaa404 32379309 83 Eberly L.A. Khatana S.A.M. Nathan A.S. Telemedicine outpatient cardiovascular care during the COVID-19 pandemic: bridging or opening the digital divide? Circulation 142 5 2020 510 512 10.1161/CIRCULATIONAHA.120.048185 32510987 84 Hong Y.A. Cho J. Has the digital health divide widened? Trends of health-related internet use among older adults From 2003 to 2011 J Gerontol B Psychol Sci Soc Sci 72 5 2017 856 863 10.1093/geronb/gbw100 27558403 85 Garcia-Cruz E. Manzur-Sandoval D. Rascon-Sabido R. Critical care ultrasonography during COVID-19 pandemic: the ORACLE protocol Echocardiography 37 9 2020 1353 1361 10.1111/echo.14837 32862474 86 Kirkpatrick J.N. Mitchell C. Taub C. Kort S. Hung J. Swaminathan M. ASE statement on protection of patients and echocardiography service providers during the 2019 novel coronavirus outbreak: endorsed by the American college of cardiology J Am Soc Echocardiogr 33 6 2020 648 653 10.1016/j.echo.2020.04.001 32503700 87 Rodriguez-Leor O. Cid-Alvarez B. Perez de Prado A. Impact of COVID-19 on ST-segment elevation myocardial infarction care. The Spanish experience Rev Esp Cardiol (Engl Ed) 73 12 2020 994 1002 10.1016/j.rec.2020.08.002 32917566 88 Chieffo A. Stefanini G.G. Price S. EAPCI position statement on invasive management of acute coronary syndromes during the COVID-19 pandemic EuroIntervention 16 3 2020 233 246 10.4244/EIJY20M05_01 32404302 89 Szerlip M. Anwaruddin S. Aronow H.D. Considerations for cardiac catheterization laboratory procedures during the COVID-19 pandemic perspectives from the Society for Cardiovascular Angiography and Interventions Emerging Leader Mentorship (SCAI ELM) members and graduates Catheter Cardiovasc Interv 96 3 2020 586 597 10.1002/ccd.28887 32212409 90 Waxman S. Garg A. Torre S. Prioritizing elective cardiovascular procedures during the COVID-19 pandemic: the cardiovascular medically necessary, time-sensitive procedure scorecard Catheter Cardiovasc Interv 96 6 2020 E602 E607 10.1002/ccd.29093 32588955 91 Pothineni N.V.K. Santangeli P. Deo R. Marchlinski F.E. Hyman M.C. COVID-19 and electrophysiology procedures-review, reset, reboot!!! J Interv Card Electrophysiol 59 2 2020 303 305 10.1007/s10840-020-00871-2 32929603 92 Lakkireddy D.R. Chung M.K. Deering T.F. Guidance for rebooting electrophysiology through the COVID-19 pandemic from the heart rhythm society and the American heart association electrocardiography and arrhythmias committee of the council on clinical cardiology Circ Arrhythm Electrophysiol 13 7 2020 e008999 10.1161/CIRCEP.120.008999 93 Di Toro A. Bozzani A. Tavazzi G. Long COVID: long-term effects? Eur Heart J Suppl 23 Suppl E 2021 E1 E5 10.1093/eurheartj/suab080 34650349 94 Seessle J. Waterboer T. Hippchen T. Persistent symptoms in adult patients 1 year after coronavirus disease 2019 (COVID-19): a prospective cohort study Clin Infect Dis 74 7 2022 1191 1198 10.1093/cid/ciab611 34223884 95 Nuzzi V. Castrichini M. Collini V. Impaired right ventricular longitudinal strain without pulmonary hypertension in patients who have recovered from COVID-19 Circ Cardiovasc Imaging 14 4 2021 e012166 10.1161/CIRCIMAGING.120.012166 96 Xie Y. Xu E. Bowe B. Al-Aly Z. Long-term cardiovascular outcomes of COVID-19 Nat Med 28 3 2022 583 590 10.1038/s41591-022-01689-3 35132265 97 Alwan N.A. The road to addressing long Covid Science 373 6554 2021 491 493 10.1126/science.abg7113 34326224 98 Briggs A. Vassall A. Count the cost of disability caused by COVID-19 Nature 593 7860 2021 502 505 10.1038/d41586-021-01392-2 34040204 99 Scaramuzza A. Tagliaferri F. Bonetti L. Changing admission patterns in paediatric emergency departments during the COVID-19 pandemic Arch Dis Child 105 7 2020 704 706 10.1136/archdischild-2020-319397 100 Moynihan R. Sanders S. Michaleff Z.A. Impact of COVID-19 pandemic on utilisation of healthcare services: a systematic review BMJ Open 11 3 2021 e045343 10.1136/bmjopen-2020-045343 101 Haldane V. De Foo C. Abdalla S.M. Health systems resilience in managing the COVID-19 pandemic: lessons from 28 countries Nat Med 27 6 2021 964 980 10.1038/s41591-021-01381-y 34002090 102 Chaudhry S.P. Stewart G.C. Advanced heart failure: prevalence, natural history, and prognosis Heart Fail Clin 12 3 2016 323 333 10.1016/j.hfc.2016.03.001 27371510 103 Yasmin F. Shah S.M.I. Naeem A. Artificial intelligence in the diagnosis and detection of heart failure: the past, present, and future Rev Cardiovasc Med 22 4 2021 1095 1113 10.31083/j.rcm2204121 34957756 104 Sara J.D.S. Maor E. Orbelo D. Gulati R. Lerman L.O. Lerman A. Noninvasive voice biomarker is associated with incident coronary artery disease events at follow-up Mayo Clin Proc 97 5 2022 835 846 10.1016/j.mayocp.2021.10.024 35341593	36462964	PMC9709614	NO-CC CODE	2022-12-01 23:23:08	no	Eur J Intern Med. 2022 Nov 30; doi: 10.1016/j.ejim.2022.11.029	utf-8	Eur J Intern Med	2,022	10.1016/j.ejim.2022.11.029	oa_other
==== Front Radiat Phys Chem Oxf Engl 1993 Radiat Phys Chem Oxf Engl 1993 Radiation Physics and Chemistry 0969-806X 0969-806X Elsevier Ltd. S0969-806X(22)00741-1 10.1016/j.radphyschem.2022.110678 110678 Article Simulation study to assess the effectiveness of gamma radiation for inactivation of viruses on food packaging material Tripathi Jyoti a Saxena Sudhanshu a Gautam Satyendra ab∗ a Food Technology Division, Bhabha Atomic Research Centre, Mumbai, 400085, India b Homi Bhabha National Institute, Anushaktinagar, Mumbai, 400094, India ∗ Corresponding author. Food Technology Division, Bhabha Atomic Research Centre, Trombay, Mumbai, 400085, Maharashtra, India. 30 11 2022 30 11 2022 11067816 3 2022 25 11 2022 26 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. The recent COVID-19 pandemic spread across the globe has raised the concern about the possible transmission of viruses through food packaging material during domestic and international trade. Therefore, mitigation strategies are needed to address these safety issues. Preliminary in-silico study showed that interactions between food packaging material and viral surface proteins were possibly hydrophobic in nature with most favourable interaction having a binding free energy of −5.24 kcal/mol. Since these interactions can cause viruses to adsorb on the food packets and get transmitted during supply chain, it is necessary to inactivate the viruses. In this context, efficacy of gamma irradiation in inactivating the viruses on the food packaging material was assessed. For this simulation study P1 (virulent) bacteriophage of E. coli was used as a model system. Gamma irradiation of food packets at an absorbed dose >8 kGy was found to completely inactivate the infectivity of P1(virulent) bacteriophage when co-cultured with E. coli host and assayed for viral plaque formation. Reduction in infectivity of P1(vir) phage was more prominent at ambient temperature(25 ± 2 °C) as compared to cold temperature(6 ± 2 °C) when assayed after storage (one week). Gamma irradiation (2 kGy) completely inactivated the virus particles on food packets when stored for 1 week at both the above temperatures. It is thus proposed that gamma irradiation (2 kGy) can possibly be integrated as a final treatment of the packaged food products to rule out the possibility of viral transmission. However, the efficacy of radiation processing against different pathogenic viruses needs to be determined prior to actual commercial deployment. Keywords Gamma radiation Packaging material P1(virulent) bacteriophage Infectivity ==== Body pmc1 Introduction The COVID-19 pandemic caused by SARS-CoV-2, a novel coronavirus, has spread worldwide in almost every country. This has adversely impacted the different socio-economic domains of the life across the nations negatively affecting their economies. The food supply sector has also been significantly affected during this pandemic. As per the World Health Organization (WHO) and Centre for Disease Control (CDC), the primary source of disease chain propagation is human-to-human contact through respiratory droplets. However, the likelihood of virus transmission through other infected material, including food and packaging surfaces cannot be completely ruled out (Han et al., 2021). Also, there are limited scientific reports on the duration for which SARS-CoV-2 can remain viable on food-packaging surfaces. Occurrence and detection of SARS-Co-V-2 in frozen foods apparently indicate that viral contamination and food borne transmission may present a systemic risk in the ongoing pandemic. Since the beginning of July 2020, at least nine incidences of food contamination have been reported across China where occurrence of SARS-CoV-2 was documented on imported foods, predominantly on the packaging materials (Han et al., 2021). Furthermore, Chinese Centre for Disease Control and Prevention also pointed that imported sea-foods contaminated with the virus were the more probable cause of the re-emerged outbreak (China Daily, 2020; Global Times, 2020). New Zealand had also revealed viral transmission through an import receiving facility (Fisher et al., 2020). Besides, during April–May 2020, the CDC identified significant (16,233) cases of COVID-19, which includes 86 deaths, among the workers at meat and poultry processing facilities located in 23 states (CDC 2020). The possibility of virus transmission through frozen food is possible as the virus has been reported to be retained in frozen flesh foods for up to 21 days (Chin et al., 2020). Due to the limitation of proper scientific evidences, the WHO has recommended further investigation of frozen foods and packaging as a potential source of transmission. Though different quality checks are implemented in all the stages of the food processing and supply chain, utmost safety considerations are required at the consumer level as this serves as the prominent source of infection. There are different processing modalities that are implemented during food processing. However, gamma irradiation being an eco-friendly, cold physical process is being used in more than 60 countries across the world for various socio-economic applications such as assurance of food safety, security, to overcome quarantine barrier of international trade, sterilization of medical devices, inactivation of bacteria and viruses in animal sera and bio-therapeutics (Jinia et al., 2020). Therefore, in the present simulation study, P1(virulent) phage was employed as a model system for assessing the effectiveness of gamma radiation technology for disinfecting food packaging material, as all the viruseshave a similar structure comprising of viral capsid protein encapsulating the nucleic acid. Moreover, an in-silico study to understand the nature of interactions between polyethylene (commonly used as food packets) and viral surface proteins was also performed. 2 Material and methods 2.1 In-silico interaction of viral surface proteins with polyethylene The protein structure file (.pdb file) of some of theviral surface proteins like SARS-CoV-2 spike glycoprotein (PDB ID: 6VXX), major capsid protein P1 of the phage phi6 (PDB ID: 4K7H), capsid protein P1 of the phage phi8 (PDB ID: 4BTP), were obtained from protein data bank. The 3D structure data file (.sdf file) of polyethylene was obtained from https://www.chemtube3d.com. The interaction of the protein with polyethylene was predicted using Autodock tools version 1.5.6 (Morris et al., 2009; Cosconati et al., 2010), with the Lamarckian genetic algorithm (Morris et al.,1998). For preparation of protein input files, all water molecules were removed, polar hydrogen were added and molecule's energy was minimised. The grid box covering whole protein structure for docking was centered at x = 197.505, y = 223.533, z = 207.441 points forSARS-CoV-2 spike glycoprotein, x = 87.416, y = 17.821, z = 28.335points for major capsid protein P1 of the phage phi6 and x = −27.158, y = 371.387, z = 417.904 points for capsid protein P1 of the phage phi8. Docking for 50 number of runs was carried out using Lamarckian Genetic Algorithm (LGA), and all other parameters set to default. The results were evaluated using inter molecular interactions and binding free energy (ΔGbind) in the lowest energy cluster with maximum cluster size. The interaction was visualized using Pymol (Seeliger and de Groot, 2010) and were calculated using Ligplot (Laskowski and Swindells, 2011). 2.2 Phage lysate preparation and enumeration of plaque forming unit The model system used for the study was P1 bacteriophage which infects Escherichia coli (E. coli MG1655: F − λ - ilvG - rfb-50 rph −1) bacteria and cause cell lysis forming plaques. The phage lysate was prepared as per the method described earlier. Briefly, exponentially (O.D.600nm ∼ 0.3) growing E. coli MG1655 culture (5 ml), was co-incubated with P1 phage stock (200 μl), in Luria broth medium supplemented with 0.2% glucose and 25 mM of calcium chloride. The incubation was performed under shaking condition at 37 °C. After 6 h, the culture lysate was centrifuged and the cell free supernatant containing viral particles was collected. The plaque forming units (PFU) were determined using phage titration, where different dilutions of the lysate were added to 100  μl culture of E. coli cells and mixed with Top-agar (0.8% agar in nutrient broth supplemented with 0.2% glucose and 25 mM calcium chloride) and poured onto the pre-set bottom agar plates. These plates were kept at 37 °C overnight and the PFU enumeration was performed the following day. 2.3 Gamma radiation treatment An aliquot (100 μl) of the viral lysate was spread onto 4 cm2 area on the packaging material (low density polyethylene; 200 gauge) under sterile conditions and dried for around 3–4 h at room temperature. This was later subjected to gamma radiation treatment in a cobalt-60 gamma chamber at Food Technology Division, Bhabha Atomic Research Centre, Mumbai, India. The dosimetry was performed using standard ceric-cerous sulphate dosimeter where the absorbed dose was measured by potentiometry (American Society for Testing and Materials, 1993). The ceric-cerous sulphate dosimeters (9 nos.) were affixed on to a plastic plate in three rows and columns at equal spacing and three such plates were placed across the chamber. The absorbed dose was determined from potential difference using the dose calibration data. The ratio of the maximum (D max) to the minimum dose (D min) in the irradiation chamber was calculated as the dose uniformity ratio (Dmax/Dmin), which was found to be 1.25. The dose of irradiation was used in a range of 2–10 kGy. After the radiation treatment individual samples were reconstituted and dilution plating was performed as described above. 2.4 Simulation for phage inactivation from the food packet by irradiation The simulation was performed wherein a food commodity (∼150 g turmeric) was packed in LDPE packets. The P1 phage lysate stock was spread onto the package and dried as described above. These packets were stored at ambient (25 ± 2 °C) as well as cold temperature (6 ± 2 °C) for one week, simulating the time period required during transport. At the end of one week, the food packets were subjected to different doses (2–10 kGy) of gamma radiation. At the end, infectivity of the reconstituted phage was evaluated as per the method described above. All the experiments were conducted in three independent sets having three replicates each. 3 Result and discussion Countries like USA, China, Germany, Brazil, Japan, France, India, are among the largest importers as well as exporters of food and agricultural commodities. The import-export flows of food commodities could pose threat of viral transmission during international/domestic trade along with the food commodities. A preliminary study was performed to understand the kind of interactions between viral surface protein and food packaging material. Since polyethylene is among the most commonly used food packaging material, its interaction with surface proteins of some of the RNA viruses was analysed through in-silico studies. The binding energy (ΔGbind) obtained after docking the interacting molecule with the protein, provides the binding affinity between the two molecules. The degree of binding between the protein and interacting molecule is termed as binding affinity. Since the free energy of a favourable system is always negative as it is the energy released when two molecules interact, therefore a more negative binding energy denotes high binding affinity. It was observed that the major interactions between the polyethylene (commonly used as food packets) and viral surface proteins are possibly hydrophobic in nature (Fig. 1 ). The interaction of the major capsid protein P1 of the bacteriophage phi6 with polyethylene showed a binding energy of −5.24 kcal/mol, while that of SARS-CoV-2 spike protein with polyethylene was found to be −3.43 kcal/mol. The capsid protein P1 of the bacteriophage phi8 showed a binding energy of −2.78 kcal/mol. The interactions were found to be weak in nature but favourable to let the viruses adsorb on the food packets. Although these weak interactions can be disintegrated by soaps and detergents, but since the idea is inconceivable for food packets, alternate technological intervention like gamma radiation is necessary for effective control of viruses. Further, the efficacy of gamma irradiation was evaluated for phage inactivation from the surface of the food packaging material.Fig. 1 Result of the interaction of viral surface proteins with polyethylene molecules. (a) Major capsid protein P1 of the phage phi6 (A chain) interacting with polyethylene molecule; (b) capsid protein P1 of the bacteriophage phi8 (A chain) with polyethylene molecule and (c)SARS-CoV-2 spike glycoprotein with polyethylene molecule. The green cartoon represents the viral surface proteins, while the cyan colour sphere molecule denotes the polyethylene as visualized by Pymol software. The interacting protein residues have been depicted using LigPlot+ software where the red lashes indicate the hydrophobic interaction of polyethylene molecule with the protein residues of the viral surface protein. Fig. 1 3.1 Determination of gamma irradiation dose required for inactivation of P1 phage The complete loss of P1 phage infectivity was achieved by the application of gamma radiation at a dose of 8 kGy (Fig. 2 ). The phage lysate showed a titre of 7.08 ± 0.02 log10 PFU/ml. The phages upon adsorption onto the food packaging material showed a titre value of6.08 ± 0.02 log10 PFU/ml. The dose required to reduce phage titre by one log (D 10 value) was estimated based upon phage infectivity inactivation kinetics. The D 10value was found to be ∼1 kGy after adsorption at the total titre of 7.08 ± 0.02 log10 PFU/ml in suspension(Fig. 3 A). Most of the coronaviruses have been reported to have a D 10 value of less than 2 kGy. The virus sensitivity to irradiation also depends on the medium of sample suspension. Previous studies on independent samples of MS2 bacteriophages in water, autoclaved raw sewage, and a tryptone solution were exposed to gamma rays which showed that the dose required for a 1 log reduction, was 0.5 kGy in water, 1.0 kGy in autoclaved raw sewage, and 1.2 kGy in 1% tryptone (Jebri et al., 2013). A more recent study by Leung et al. (2020), showed that radiation dose of 10 kGy was required to completely inactivate 106.5 TCID50/ml of SARS-CoV-2, where TCID50 is the median tissue culture infectious dose and the value corresponds to 6.5 log10units. The inactivation of viruses by gamma radiation can be attributed to the destruction of genetic material either directly by radiolytic cleavage or cross-linking, or indirectly leading to radiolysis of water molecules generating hydroxyl radicals (·OH), which cause irreparable damage to the nucleic acids and proteins (Hume et al., 2016). In this study the P1(vir)phage particles were adsorbed onto the surface of the low density polyethylene packaging material, making them more sensitive to environmental conditions and radiation stress. However, if the phage particles remain suspended in Luria Broth medium (containing 0.2% glucose and 25 mM of calcium chloride), a much higher dose was required for its complete inactivation (Fig. 3B), indicating that viruses in protein rich medium have a protective mechanism which dampens the effects of gamma irradiation. Some studies have reported that proteins in solution have a negative impact on gamma radiation induced degradation, by acting as a scavenger (Hume et al., 2016). These scavenger molecules quench hydroxyl radicals thereby preventing the destruction of viral nucleic acids. Previous studies recommended 20 kGy radiation dose as safe for inactivation of viruses like coronaviruses. A more recent study has shown that even at a dose of 10 kGy the SARS-CoV was completely inactivated (Feldmann et al., 2019). These studies were performed with the virus suspended in cell culture medium (like Dulbecco's modified Eagle's medium [DMEM] with 10% fetal bovine serum [FBS]), which is rich in protein and therefore required a higher dose of inactivation.Fig. 2 Figure shows the plaque forming ability of P1 phage on (E. coli),(A) Control (non-irradiated samples):E. coli cells infected with P1 phage showing plaques(marked by arrows) at dilution 10−1, (B) Irradiated at 8 kGy: No plaques observed even in undiluted samples. Fig. 2 Fig. 3 Figure showing the loss in infectivity of the P1 phage virus as observed by reduction in phage titre upon irradiation (A) Viral phage particles dried onto the surface of food package material, (B) Viral particles suspended in liquid medium. Different letters (a–d) indicate significant differences in the mean value at p < 0.05. Fig. 3 3.2 Low dose of gamma irradiation required for inactivation of phage from food packets On the infected food packets, the viral infectivity reduced on its own at cold temperature from 6.08 ± 0.02 log10 PFU/ml to 2.57 ± 0.05 log10 PFU/ml, after one week. Reduction in infectivity of P1 phage was more prominent at ambient temperature and this reduced to 1.7 ± 0.1 log10 PFU/ml during one-week time period. Furthermore, upon gamma irradiation (2 kGy), no plaques were observed indicating that dose of 2 kGy or above is able to reduce the infectivity of the P1(vir) phage completely (Fig. 4 ).Fig. 4 Figure showing the inactivation of P1 phage during the simulation study at cold temperature after one-week storage (A) Control (non-irradiated): few plaques were observed in undiluted samples; (B) Irradiated at 2 kGy: no plaques observed in undiluted sample. Fig. 4 An earlier report too has indicated that human coronaviruses can remain infectious on inanimate surfaces like metal, glass or plastic under room temperature for up to 9 days (Kampf et al., 2020). Besides, previous studies have shown that at relatively higher temperatures including 30 °C or more the duration of persistence of virus is shorter (Kampf et al., 2020). This is in accordance with our observation, and therefore only 2 kGy gamma radiation dose was found to be effective in inactivating the P1 phage at the initial titre level of 7.08 ± 0.02 log10 PFU/ml. Benefits of the lower dose of irradiation required to inactivate phage is that it can be applied to a variety of food commodities including cereals, fresh fruits and vegetables, seafood and meat (fresh and frozen), the major food commodities traded internationally. 4 Conclusion The current simulation study ascertained the effectiveness of gamma irradiation (2 kGy) in inactivating P1(vir) phage from the food packets after storage of one week. This irradiation dose offers a practical feasibility for irradiation of broad range food commodities and may provide a solution to the concern of viral contamination through trade during the current scenario. Author Stataement Jyoti Tripathi: Conceptualization, Writing - original draft, Writing-review & editing, Investigation, Formal analysis, Validation, Data curation. Sudhanshu Saxena: Conceptualization, Writing - original draft, Writing-review & editing, Investigation, Formal analysis. Satyendra Gautam: Conceptualization, Writing - original draft, Writing - review & editing, Supervision. 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 Data will be made available on request. Acknowledgement The sole funding agency is the Government of India and no external source of fund or study sponsor is involved. ==== Refs References American Society for Testing and Materials Standard practice for use of a ceric cerous dosimetry system. E1205-93 Annual Book of Standards vol. 12 1993 AmericanSociety for Testing and Materials West Conshohoken, PA 02 Centers for Disease Control and Prevention (CDC) Update: COVID-19 Among Workers in Meat and Poultry Processing Facilities-United States 2020 April-May 2020 https://www.cdc.gov/mmwr/volumes/69/wr/mm6927e2.htm Chin A.W.H. Chu J.T.S. Perera M.R.A. Hui K.P.Y. Yen H. Chan M.C.W. Peiris M. Poon L.L.M. Stability of SARS-CoV-2 in different environmental conditions The Lancet Microb. 1 1 2020 10.1016/S2666-5247(20)30003-3 Cosconati S. Forli S. Perryman A.L. Harris R. Goodsell D.S. Olson A.J. Virtual screening with AutoDock: theory and practice Expet Opin. Drug Discov. 5 2010 597 607 China Daily Experts see similarity in Beijing, Dalian outbreaks https://www.chinadaily.com.cn/a/202007/30/WS5f228da3a31083481725d32c.html 2020 Feldmann F. Shupert W.L. Haddock E. Twardoski B. Feldmann H. Gamma irradiation as an effective method for inactivation of emerging viral pathogens Am. J. Trop. Med. Hyg. 100 5 2019 1275 1277 10.4269/ajtmh.18-0937 30860018 Fisher D. Reilly A. Zheng A.K.E. Cook A.R. Anderson D.E. Seeding of outbreaks of COVID-19 by contaminated fresh and frozen food BioRxiv. 2020-08 2020 Global Times COVID-19 Outbreaks in Wuhan, Beijing and Dalin Share Certain Similarities: China's Top Epidemiologist 2020 https://www.globaltimes.cn/content/1196130.shtml Han J. Zhang X. He S. Jia P. Can the coronavirus disease be transmitted from food? A review of evidence, risks, policies and knowledge gaps Environ. Chem. Lett. 19 1 2021 5 16 10.1007/s10311-020-01101-x 33024427 Hume A.J. Ames J. Rennick L.J. Duprex W.P. Marzi A. Tonkiss J. Mühlberger E. Inactivation of RNA viruses by gamma irradiation: a study on mitigating factors Viruses 8 7 2016 204 10.3390/v8070204 27455307 Jebri S. Hmaied F. Jofre J. Yahya M. Mendez J. Barkallah I. Hamdi M. Effect of gamma irradiation on bacteriophages used as viral indicators Water Res. 47 11 2013 3673 3678 10.1016/j.watres.2013.04.036 23726703 Jinia A.J. Sunbul N.B. Meert C.A. Miller C.A. Clarke S.D. Kearfott K.J. Matuszak M.M. Pozzi S.A. Review of sterilization techniques for medical and personal protective equipment contaminated with SARS-CoV-2 IEEE Access 8 2020 111347 111354 10.1109/ACCESS.2020.3002886 34192107 Kampf G. Todt D. Pfaender S. Steinmann E. Persistence of coronaviruses on inanimate surfaces and their inactivation with biocidal agents J. Hosp. Infect. 104 3 2020 246 251 10.1016/j.jhin.2020.01.022 32035997 Laskowski R.A. Swindells M.B. LigPlot+: multiple ligand-protein interaction diagrams for drug discovery J. Chem. Inf. Model. 51 10 2011 2778 2786 10.1021/ci200227u.Epub2011Oct5 21919503 Leung A. Tran K. Audet J. Lavineway S. Bastien N. Krishnan J. In vitro inactivation of SARS-CoV-2 using gamma radiation Appl. Biosaf. 25 3 2020 157 160 36035758 Morris G.M. Goodsell D.S. Halliday R.S. Huey R. Hart W.E. Belew R.K. Olson A.J. Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function J. Comput. Chem. 19 14 1998 1639 1662 Morris G.M. Huey R. Lindstrom W. Sanner M.F. Belew R.K. Goodsell D.S. Olson A.J. Autodock4 and AutoDockTools4: automated docking with selective receptor flexibility J. Comput. Chem. 30 2009 2785 2791 19399780 Seeliger D. de Groot B.L. Ligand docking and binding site analysis with PyMOL and Autodock/Vina J. Comput. Aided Mol. Des. 24 2010 417 422 20401516	36466007	PMC9709647	NO-CC CODE	2022-12-03 23:16:11	no	Radiat Phys Chem Oxf Engl 1993. 2023 Mar 30; 204:110678	utf-8	Radiat Phys Chem Oxf Engl 1993	2,022	10.1016/j.radphyschem.2022.110678	oa_other
==== Front J Mol Struct J Mol Struct Journal of Molecular Structure 0022-2860 1872-8014 Elsevier B.V. S0022-2860(22)02335-3 10.1016/j.molstruc.2022.134690 134690 Article Discovery of Novel Thioquinazoline-N-aryl-acetamide/N-arylacetohydrazide Hybrids as Anti-SARS-CoV-2 Agents: Synthesis, in vitro Biological Evaluation, and Molecular Docking Studies Abdel-Mohsen Heba T. ⁎a Omar Mohamed A. a Kutkat Omnia b Kerdawy Ahmed M. El c Osman Alaa A. d GabAllah Mohamed b Mostafa Ahmed ⁎b Ali Mohamed A. b Diwani Hoda I. El a a Department of Chemistry of Natural and Microbial Products, Pharmaceutical and Drug Industries Research Institute, National Research Centre, El‐Buhouth St., Dokki, P.O. Box 12622, Cairo, Egypt b Center of Scientific Excellence for Influenza Viruses, National Research Centre, Giza 12622, Egypt c Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Cairo University, Kasr El-Aini Street, Cairo, P.O. Box 11562, Egypt d Department of Pharmaceutical Chemistry, School of Pharmacy, NewGiza University (NGU), NewGiza, km 22 Cairo–Alexandria Desert Road, Cairo, Egypt ⁎ Corresponding authors: 30 11 2022 30 11 2022 13469012 8 2022 10 11 2022 29 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. In the current investigation, two novel series of (tetrahydro)thioquinazoline-N-arylacetamides and (tetrahydro)thioquinazoline-N-arylacetohydrazides were designed, synthesized and investigated for their antiviral activity against SARS-CoV-2. The thioquinazoline-N-arylacetamide 17g as well as the tetrahydrothioquinazoline-N-arylacetohydrazides 18c and 18f showed potent antiviral activity with IC50 of 21.4, 38.45 and 26.4 µM, respectively. In addition, 18c and 18f demonstrated potential selectivity toward the SARS-CoV-2 over the host cells with SI of 10.67 and 16.04, respectively. Further evaluation of the mechanism of action of the three derivatives 17g, 18c, and 18f displayed that they can inhibit the virus at the adsorption as well as at the replication stages, in addition to their virucidal properties. In addition, 17g, 18c, and 18f demonstrated satisfactory physicochemical properties as well as drug-likeness properties to be further optimized for the discovery of novel antiviral agents. The docking simulation predicted the binding pattern of the target compounds rationalizing their differential activity based on their hydrophobic interaction and fitting in the hydrophobic S2 subsite of the binding site Graphical abstract Image, graphical abstract ==== Body pmc1 Introduction Recently, coronavirus disease 2019 (COVID-19) has been identified as a global pandemic disease that affects the survival of population all over the world [1]. COVID-19 is a respiratory disease that causes upper and lower respiratory tract infection which can be further progressed further into respiratory failure by complex mechanisms and may end up with premature mortality [1, 2]. It was reported that COVID-19 is caused by a novel zoonotic member of betacoronaviruses called severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) which is a single stranded RNA virus of the Coronaviridae family [3, 4]. The huge problem, which is currently facing the world, is the remarkable ability of SARS-CoV-2 to mutate over a short period of time [5]. Hence, a plethora of studies have been performed to get some insightful knowledge about SARS-CoV-2 virus [6]. Successfully, some information has been reported including its molecular structure, life cycle, and its interactions with the host cells. This enabled the development various vaccines with different mechanisms of action to be used by humans [7]. Up till now, the effective and specific antiviral agents for the treatment of SARS-CoV-2 infection are limited or rare [8]. Hence, numerous investigations have been carried out to identify new targets to control this pathogen without affecting the host cells. Recently, some structural elements that can act as potential therapeutic targets have been recognized. Spike glycoprotein was identified as a promising target that is present on the virus surface, and it is responsible for the virus binding to the host cell [9]. RNA-dependent RNA polymerase (RdRp) is another attractive target participating in the replication of RNA from an RNA template [10]. Meanwhile, the 3-C-like protease (3CLpro or Mpro) and papain-like protease (PLpro) were pointed out to be the most important targets for the design of promising antiviral agents against SARS-CoV-2 as they play a key role in the life cycle of SARS-CoV-2 virus with no homologues proteins in human cells [11], [12], [13], [14]. Hence, inhibition of Mpro and/or PLpro can result in a selective antiviral activity without side effects on humans [12]. In this context, quinazolines are regarded as one of the most interesting scaffolds for designing antiviral agents [15, 16]. Quinazolines were reported to possess potential antiviral activity against diverse types of RNA viruses including respiratory syncytial virus (RSV) [17], influenza A virus (IAV) [18] as well as hepatitis C virus (HCV) [19]. For instance, Zhang and co-workers [18] reported the design of a series of 2,4-disubstituted quinazoline derivatives incorporating S-acetamide and NH-acetamide moieties at position 4. Compound I, a representative of this series, was found to have a potent antiviral activity against IAV. Moreover, Hwu et al., [19] described the potent antiviral activity of a class of quinazoline-coumarin conjugates, for example compound II showed potent antiviral activity against hepatitis C virus and chikungunya virus. Meanwhile, Lee et al. [20] reported the synthesis of a series of 4-anilino-6-aminoquinazolines as anti-MERS-CoV inhibitors, compound III, a representative example of the synthesized series, showed IC50 = 0.157 µM, CC50 = 3.59 µM and SI = 25. Recently, some studies reported the activity of quinazoline derivatives against SARS‑CoV‑2 [21, 22]. For example, Zhao et al. [21] reported the promising activity of a series of quinolone and quinazoline derivatives in inhibiting RNA synthesis driven by SARS-CoV-2 RdRp. For instance, compounds IV and V revealed 58.43% and 58.83% inhibition on SARS-CoV-2 RdRp at 10 µM concentration. Additionally, Rothan and Teoh [22] reported the expected interesting potential of quinazoline derivatives, for example compound VI, in inhibiting SARS‑CoV‑2 Main Protease (Mpro) in a high throughput virtual screening campaign (Figure 1 ).Figure 1 Examples of antiviral agents I-VI incorporating quinazoline scaffold Figure 1: On the other hand, several studies reported the promising antiviral properties of various heterocycles substituted with N-aryl-2-(thio)acetamide moieties. For instance, RDEA806 (VII) was reported to possess a potential HIV-1 reverse transcriptase (RT) inhibitory activity against wild-type (WT) as well as some non-nucleoside reverse transcriptase inhibitor (NNRTI) resistant viruses [23]. Moreover, Zhan and co-workers [24] reported the synthesis and the interesting antiviral properties of novel 1,2,4-triazin-6-yl-thioacetamide derivatives as potent HIV-1 NNRTIs. Compound VIII was an example of this series displaying nanomolar activity against HIV-1(WT) as well as a moderate potency against the double mutant strain RES056. Furthermore, Zhang et al. [25] reported the discovery of some indol-3-yl-thio-N-phenyl-acetamides with potent antiviral activity. For instance, compound IX revealed a dual potency against RSV and IAV [25]. Moreover, Yu et al. [ 26 ] reported the potent anti-influenza activity of different pyrimidines substituted at 2 position with a N-aryl-2-(thio)acetamide moiety. For example, compound X demonstrated a broad activity against IAV and IBV. In addition, Zhan and co-workers [27] reported the anti-influenza properties of some thiazolyl-N-aryl-2-(thio)acetamides, for example, the derivative XI displayed a potent activity on influenza A/H1N1 virus (Figure 2 ).Figure 2 Structures of antiviral agents VII-XI incorporating N-aryl-2-(thio)acetamide moiety Figure 2: Rationale design of thioquinazoline-N-aryl-acetamide/N-arylacetohydrazide hybrids Encouraged by the previous findings, and since most of the antiviral drugs in the clinical use were not designed specifically for SARS-CoV-2, we were interested in the current study in designing a new series of anti-SARS-CoV-2 through the application of the molecular hybridization strategy between the quinazoline ring and the N-aryl-2-(thio)acetamide moiety to design a novel series of (tetrahydro)thioquinazoline-N-arylacetamide hybrids XII (Figure 3 ). For further structure-activity relationship investigation, further elongation of the N-aryl-2-(thio)acetamido linkage was carried out by its replacement with a 2-(thio)acetohydrazide linker to afford scaffold XIII (Figure 3). Derivatives from the designed scaffolds were synthesized utilizing the conventional methods of organic chemistry. Subsequently, the synthesized compounds were assayed for their antiviral activity against SARS-CoV-2. Promising hits were further examined for their expected mode of antiviral activity. Additionally, their physicochemical as well as pharmacokinetic properties were predicted using the SwissADME free webtool. Concurrently, in silico molecular docking studies were performed on the most potent derivatives in the SARS‑CoV‑2 Main Protease (Mpro) binding site to study their binding mode in order to rationalize their promising antiviral activity.Figure 3 The design strategy of the target (tetrahydro)thioquinazoline-N-arylacetamides XII and (tetrahydro)thioquinazoline-N-arylacetohydarzides XIII as anti-SARS-CoV-2 agents Figure 3: 2 Results and discussion 2.1 Chemistry Initially, the starting quinazoline 3 was synthesized by the reaction of 1 with CS2 (2) in the presence of KOH under reflux [28, 29]. Meanwhile, the starting material 6 was synthesized by the condensation of 4 with thiourea (5) under basic conditions [30, 31] (Scheme 1 ).Scheme 1 Synthesis of the starting quinazolines 3 and 6 Scheme 1: For the synthesis of the target compounds 11a-c and 12a-c; 2-aminobenzoic acid (7a), 3-aminobenzoic acid (7b), and 4-aminobenozic acid (7c) were reacted with chloroacetyl chloride (8) in DMF to afford the intermediates 9a-c, respectively, which were subsequently reacted with the starting materials 3 and 6 under basic conditions to afford the target compounds 11a-c and 12a-c, respectively, in excellent yields (Scheme 2 ).Scheme 2 Synthesis of the target compounds 11 and 12 Scheme 2: For the synthesis of the target quinazolines 17 and 18, different benzoic acid derivatives 13 were first esterified to give the corresponding esters 14 which were subsequently reacted with hydrazine hydrate to afford the corresponding acid hydrazides 15 (Scheme 3 ). Further reaction of 15 with chloroacetyl chloride (8) was performed to give the intermediates 16 which were reacted with the thioquinazolines 3 and 6 to yield the target compounds 17 and 18, respectively, in excellent yields.Scheme 3 Synthesis of the target compounds 17 and 18 Scheme 3: 2.2 Biological evaluation All the target compounds 11a-c, 12a-c, 17a-g and 18a-g were evaluated for their antiviral activity against NRC-03-nhCoV as well as for their cytotoxic activity on Vero-E6 cells employing MTT assay [32]. The CC50 (concentration necessary for 50% growth inhibition of normal cell line compared to the control experiment) of the target compounds on Vero-E6 cells and IC50 (concentration necessary for 50% reduction of virus-induced cytopathic effect (CPE) compared to the virus control experiment) of the target compounds against NRC-03-nhCoV virus in Vero-E6 cells are presented in table 1 . The selectivity indices which are the ratio of CC50 relative to IC50 of the tested compounds were calculated and depicted in table 1. In addition, figure 4 presents the inhibition curves of the most potent compounds 17g, 18c, and 18f (For the rest of the inhibition curves see the SI).Table 1 IC50 against NRC-03-nhCoV, CC50 on Vero-E6, and the selectivity index (SI) [CC50/IC50] of the target compounds Table 1Image, table 1 Compound ID R IC50a (μM) CC50b (μM) SIc 11a 2-COOH 104 450.5 4.3 11b 3-COOH 79.7 394 4.9 11c 4-COOH 79.2 442.6 5.6 12a 2-COOH 173.6 431.4 2.5 12b 3-COOH 531.3 524.3 0.98 12c 4-COOH 530.4 333.5 0.63 17a H 317.5 186.4 0.59 17b 4-Me 126.7 271.4 2.14 17c 4-NO2 121.1 316.1 2.61 17d 2-OMe 258.8 101.2 0.4 17e 2-Cl 457.8 114.2 0.24 17f 4-Cl 74.3 319 4.3 17g 2-Br 21.4 184.9 8.64 18a H 140.2 235 1.68 18b 4-Me 113.7 503.8 4.43 18c 4-NO2 38.45 410.3 10.67 18d 2-OMe 107 315.5 2.95 18e 2-Cl 459.8 524.5 1.14 18f 4-Cl 26.4 423.6 16.04 18g 2-Br 93.4 488.8 5.23 aIC50 (half maximal inhibitory concentration); bCC50 (half maximal cytotoxic concentration); cSI (Selectivity index) Figure 4 Dose-inhibition curves of 17g, 18c, and 18f against NRC-03-nhCoV [33] and Vero-E6 cells; IC50 and CC50 values were calculated using nonlinear regression analysis of GraphPad Prism software (version 5.01) by plotting log inhibitor versus normalized response (variable slope). Figure 4 The presented study showed promising activity in comparison to the results reported in the literature [20, 21]. From the IC50 results presented in table 1, it is obvious that the synthesized thioquinazoline and tetrahydrothioquinazolines displayed moderate to potent antiviral activity against NRC-03-nhCoV. Compounds 17g, 18c and 18f showed the most potent inhibitory activity with IC50 of 21.4, 38.45 and 26.4 µM, respectively, on NRC-03-nhCoV. Close investigation of the antiviral results revealed that the thioquinazoline-N-aryl-acetamide hybrids 11a-c demonstrated moderate antiviral properties with an IC50 range of 79.2 to 104 µM. Compounds 11b and 11c incorporating 3 and 4 carboxylic groups, respectively, showed higher potencies in comparison to the 2-carboxylic acid derivative 11a with IC50 of 79.7, 79.2, and 104 µM, respectively, as well as higher SIs. Replacing the thioquinazoline moiety in series 11a-c with a tertahydrothioquinazoline moiety in series 12a-c decreased the antiviral activity (IC50 range of 173.6 to 531.3 µM). Compound 12a, with a carboxylic acid moiety at the 2 position, showed an IC50 of 173.6 µM and a 2.5-fold higher selectivity toward the virus over the host cells (CC50 = 431.4 µM). Shifting the COOH moiety to the 3 and 4 positions in 12b and 12c, respectively, resulted in more than a 2-fold decrease in the antiviral potency with no preferential selectivity (SI = 0.98 and 0.63, respectively). Replacing the thio-N-aryl-acetamide moiety of series 11a-c with thioacetohydrazide moiety in series 17a-g, resulted in a different pattern of potency as well as specificity. The thioquinazoline-N-aryl-acetohydrazide derivative 17a with an unsubstituted terminal phenyl moiety displayed a weak antiviral activity (IC50 = 317.5 µM) and high cytotoxic activity (CC50 = 186.4 µM, SI = 0.59). Introduction of 4-Me and 4-NO2 groups at the terminal phenyl in 17b and 17c increased the antiviral properties with IC50 of 126.7 and 121.1 µM, respectively and SI of 2.14 and 2.61 µM, respectively. Moreover, the 2-OMe 17d and 2-chloro 17e derivatives demonstrated a decrease in the antiviral activity with IC50 of 258.8 and 457.8 µM, respectively, with concomitant high cytotoxic (CC50 = 101.2 and 114.2 µM, respectivley). Introduction of 4-chloro group at the terminal phenyl group in 17f resulted in a more than 4-fold increase in the antiviral activity (IC50 = 74.3 µM) in comparison to 17a (IC50 = 317.5 µM) with reduced cytotoxic activity (CC50 = 319 µM and SI = 4.30). Furthermore, the 2-bromo congener 17g demonstrated the highest antiviral activity (IC50 = 21.4 µM) with 8.64 times higher selectivity towards the virus over the host cells (CC50 =184.9 µM). Hydrogenation of the fused phenyl group of the thioquinazoline ring in series 17a-g to afford the tetrahydrothioquinazoline series 18a-g increased the antiviral activity in all cases with the exception of compounds 18e and 18g. Compound 18a with unsubstituted phenyl group displayed a moderate antiviral activity (IC50 = 140.2 µM) with a low SI of 1.68 (CC50 = 235 µM). Introduction of 4-NO2 and 4-Cl groups at the terminal phenyl moiety to yield 18c and 18f, respectively, resulted in a more than three-fold increase in potency in comparison to 18a with IC50 of 38.45 and 26.4 µM, respectively, in addition to a more than 10-fold higher selectivity toward the virus over the host cell (SI = 10.67 and 16.04, respectively). On the other hand, introduction of 4-Me, 2-OMe, 2-Br groups at the terminal phenyl moiety to give 18b, 18d, and 18g showed a decrease in the antiviral activity with IC50 = 93.40 to 113.7µM. Slight decrease in the antiviral activity was found for 18e (IC50 = 459.8 µM) in comparison to 17e (IC50 = 457.8 µM) and more than four-fold reduction was obsereved for 2-bromo derivative 18g (IC50 = 93.40 µM) in copmarison to 17g (IC50 = 21.4 µM). In summary, although compound 17g displayed the highest antiviral activity with IC50 of 21.4 µM, it was less favored because of its cytotoxic properties (CC50 = 184.9 µM and SI = 8.64). On the contrary, 18f and 18c were regarded to be the most promising derivatives in terms of their potency (IC50 = 26.4 and 38.45 µM) as well as their selectively and safety (SI = 16.04 and 10.67). 2.3 Mechanism of Anti-SARS-CoV-2 Activity For further investigation of the mechanism of virus inhibition of the most promising derivatives 17g, 18c and 18f, a plaque infectivity reduction assay was performed [34]. This assay studied whether the promising derivatives affected the virus at the adsorption stage and/or replication stage and/or due to their direct virucidal effect. The inhibition results are depicted in table 2 . Interestingly, the three derivatives were found to have multiple inhibitory effects to different extents in the three stages. Nevertheless, the replication stage was found to be the most affected following treatment with the tested compounds (40-50% viral inhibition at 125 µM and 38.9-40% viral inhibition at 62.5 µM).Table 2 Mechanisms of action of 17g, 18c and 18f against SARS-CoV-2 Table 2Compound ID Conc (µM) Mode of action Virus inhibition% Adsorption Replication Virucidal 17g 125 29.6 ± 8.3 50.0 ± 3.3 47.8 ± 5.1 62.5 26.7 ± 6.65 45.1 ± 1.71 37.3 ± 6.43 31.25 16.7 ± 3.35 38.2 ± 4.29 25.6 ± 5.1 18c 125 44.4 ± 3.87 43.3 ± 3.35 38.9 ± 2.31 62.5 28.9 ± 10.18 38.9 ± 1.91 33.6 ± 6.03 31.25 20.0 ± 3.3 30.0 ± 3.3 22.2 ± 5.1 18f 125 20.0 ± 3.3 40.0 ± 10 53.3 ± 3.35 62.5 13.3 ± 3.35 40.0 ± 3.3 34.4 ± 5.1 31.25 6.7 ± 6.65 31.6 ± 4.27 15.6 ± 13.89 As can be seen in table 2, the thioquinazoline-N-aryl acetamide derivative 17g showed at 125 µM a moderate inhibition of the virus at the adsorption stage 29.6%, whereas it showed a potent inhibition at the replication stage (50%) as well as a potent virucidal effect (47.8%). The tetrahydrothioquinazoline-N-arylacetohydrazide 18c displayed moderate inhibitory activity on the three tested mechanisms. At 125 µM, it revealed inhibition% of 44.4 and 43.3% by applying adsorption and replication mechanisms, respectively, and a 38.9% virucidal effect was noticed. Moreover, at 125 µM, compound 18f was found to have 53% virucidal activity as well as a 40% inhibitory effect on virus replication and only 20% inhibition of the viral adsorption mechanism. 2.4 Molecular Modeling Protease enzyme plays a critical role in viral protein maturation by cleaning proproteins after their translation into the host cell cytosol. As a result, viral proteases are considered potential antiviral drug targets [35]. The inhibition of a viral protease can reduce the assembly of mature viral particles. Therefore, SARS-CoV-2 main protease (Mpro) could be a plausible target for the newly synthesized compounds, especially with its reported pyrimidinedione inhibitors which are analogues to our designed antiviral derivatives [36], thus, molecular docking simulations have been carried out to study the binding pattern of the target compounds 11a-c, 12a-c, 17a-g, and 18a-g in the active site of SARS-CoV-2 main protease (Mpro). For this end, the X-ray crystallographic structure of SARS-CoV-2 main protease (Mpro) co-crystalized with a pyrimidine-2,4-dione inhibitor (YD1) (PDB ID: 7LTJ) was retrieved from the protein data bank (https://www.rcsb.org/)[36]. The molecular docking protocol was first validated by self-docking of the co-crystallized ligand (YD1) in the vicinity of the enzyme active site. The self-docking step reproduced the co-crystalized ligand pose efficiently with a docking score (S) of −13.26 kcal/mol and a root mean square deviation (RMSD) of 1.538 Å. Moreover, the docking protocol reproduced all the key interactions with the active site amino acids (Figure 5, Figure 6 ). Using the validated molecular docking protocol, the target compounds 11a-c, 12a-c, 17a-g, and 18a-g were docked in the SARS-CoV-2 Mpro active site.Figure 5 2D interaction diagram of the co-crystalized inhibitor YD1 in Mpro active site (PDB ID: 7LTJ) Figure 5: Figure 6 2D diagram (A) and 3D representation (B) of the superimposition of the co-crystallized (red) and the docking pose (green) of YD1 in the active site of Mpro (PDB ID: 7LTJ). Figure 6: Generally, the target compounds showed a common binding pattern in the target enzyme (Mpro) active site accommodated in subsites S1 and S2 (figures 7 -9 ). The (tetrahydro)quinazolinone moiety occupies YD1 uracil binding subsite S1 driven by polar contacts interacting by its carbonyl group with the key amino acid His163 through hydrogen bonding. On the other side, the substituted phenyl group occupies the largely hydrophobic S2 subsite occupied by YD1 dichlorophenyl moiety. The target compounds’ substituted phenyl group is stabilized through π-π stacking with the imidazole side chain of the key catalytic amino acid His41. The substituted phenyl group is sandwiched between the side chains of the amino acids His41 and Gln189. The linker in-between interacts through multiple H-bond interactions with the surrounding key amino acids Asn142, Gly143, Cys145, and His164. In compounds 11a-c and 12a-c, the peripheral carboxylate moiety on the distal phenyl group is involved in extra hydrogen bond interactions with the amino acids His41, Cys44, Met49, Pro52, and/or Met165 (For further details see supporting materials).Figure 7 2D diagram (A) and 3D representation (B) of compound 17g showing its interaction in Mpro active site (PDB ID: 7LTJ). Figure 7 Figure 8 2D diagram (A) and 3D representation (B) of compound 18c showing its interaction in Mpro active site (PDB ID: 7LTJ). Figure 8 Figure 9 2D diagram (A) and 3D representation (B) of compound 18f showing its interaction in Mpro active site (PDB ID: 7LTJ). Figure 9 Table 3 shows the docking score of the target compounds and the co-crystalized ligand (YD1). The newly synthesized compounds show a predicted docking score range of −13.99 to −11.73 kcal/mol, whereas the co-crystalized ligand YD1 showed a predicted docking score of −13.26 kcal/mol. As for the most promising compounds 17g, 18c, and 18f, compound 18c showed the most negative docking score −13.99 kcal/mol more negative than that of YD1. Whereas compounds 17g and 18f exhibited comparable docking score (−12.45 and −12.07 kcal/mol, respectively) which is less negative than that of the co-crystalized ligand (YD1) (−13.26 kcal/mol) indicating their less predicted binding affinity than YD1. Compounds 17d, 18a, 17a, and 17e showed the least negative predicted docking scores (−11.54, −11.47, −11.38, and −11.37 kcal/mol) which agree with their poor experimental activity (258.8, 140.2, 317.5, and 457.8 μM).Table 3 Docking energy scores (S) in kcal/mol for the target compounds and the co-crystalized compound (YD1) in Mpro active site (PDB ID: 7LTJ). Table 3Compound Energy score (S) kcal/mol IC50 (μM) 11a −11.60 104 11b −11.58 79.7 11c −11.92 79.2 12a −11.98 173.6 12b −11.96 531.3 12c −12.39 530.4 17a −11.38 317.5 17b −12.16 126.7 17c −12.99 121.1 17d −11.54 258.8 17e −11.37 457.8 17f −11.98 74.3 17g −12.45 21.4 18a −11.47 140.2 18b −13.36 113.7 18c −13.99 38.45 18d −11.97 107 18e −11.78 459.8 18f −12.07 26.4 18g −13.18 93.4 YD1 −13.26 4.2 [36] The predicted binding pattern of the target compounds could rationalize their differential activity based on their hydrophobic interaction and fitting in the hydrophobic S2 subsite of the binding site. Compounds 11a-c and 12a-c showed a relatively less predicted binding affinity due to their ionic polar carboxylate substitution on their distal phenyl group which decreases the probable hydrophobic interactions with the hydrophobic S2 subsite, however, this loss of proper hydrophobic interactions is somehow compensated by the carboxylate involvement in multiple extra hydrogen bond interactions with the amino acids His41, Cys44, Met49, Pro52, and/or Met165 like in case of compound 11b. On the other hand, compounds 17a-g and 18a-g show better predicted binding affinity because of their hydrophobic substitutions on the distal phenyl group and the longer thioacetohydrazide linker which makes the (substituted)phenyl group better fitted in the hydrophobic S2 subsite. Compounds achieving higher hydrophobic interaction with S2 subsite and better fit of their (substituted)phenyl group show strong predicted binding affinity as reflected in their promising experimental activity e.g., compounds 17g, 18c, and 18f with o-bromo, p-nitro, and p-chloro substitution, respectively. Alternatively, compounds with unsubstituted distal phenyl ring 17a and 18a show less predicted binding affinity relative to their substituted congeners due to their less possible hydrophobic interactions with S2 subsite. Compounds with less hydrophobic substituents (OMe) e.g., compounds 17d or that are not well fitted in the S2 subsite e.g., compounds 17e and 18e show weaker predicted binding affinity which is reflected in their weak experimental activity (IC50 = 258.8, 457.8, and 459.8 μM). 2.5 Estimation of physicochemical, pharmacokinetic and ADME properties Encouraged by the promising antiviral properties of 17g, 18c and 18f, they were further selected to predict for their physicochemical and ADME properties using SwissADME free web tool [37]. Table 4 presents some selected results of 17g, 18c and 18g.Table 4 Selected calculated physicochemical and pharmacokinetic properties of 17g, 18c and 18f from SwissADME free webtool [37]. Table 4Product 17g 18c 18f MW 433.28 403.41 392.86 Rotatable bonds 7 8 7 H-bond acceptors 4 6 4 H-bond donors 3 3 3 MR 102.11 103.16 99.35 TPSA 129.25 175.07 129.25 Log P 1.77 1.60 2.62 GI absorption High Low High BBB permeant No No No P-gp substrate No Yes Yes Analysis of the obtained results (Table 4) revealed that the target thioquinazoline-N-aryl acetamide 17g and tetrahydrothioquinazoline-N-aryl acetohydrazides 18c and 18f express acceptable levels of physicochemical properties. The molecular weight of 17g, 18c and 18f is spanning between 392.86 to 433.28 g/mol. They incorporate acceptable numbers of hydrogen bond donors (less than 5) and acceptors (less than 10). Additionally, the topological polar surface area (TPSA) for 17g and 18f is 129.25 Å, while it is slightly high in case of the 18c TPSA = 175.07 Å and the ilogP (octanol–water partition coefficient) is ranging from 1.60 to 2.62 [38]. The target (tetrahydro)thioquinazoline-N-arylacetamides 17g and 18f are predicted to be well absorbed from the GIT while 18c is predicted to have low GIT absorption. The three target compounds 17g, 18c and 18f have no predicted ability to penetrate the blood brain barrier which decreases their probable adverse effect at the central level. Compounds 18c and 18f were found to substrates for P-glycoprotein (P-gp), which plays a significant role in the removal of strange substances outside the cells. Meanwhile, 17g is not a substrate for P-glycoprotein (P-gp) [39]. Figure 10, presents the bioavailability radar chart of compounds 17g, 18c and 18f provided by the SwissADME web tool [37]. Generally, the bioavailability radar presented by SwissADME demonstrates a pink-colored area that identifies the optimum space of six physicochemical parameters for oral bioavailability. These six properties are size, polarity, lipophilicity, solubility, flexibility, and saturation. Close investigation of the presented radar charts showed that compound 18f occupies the ideal space of the six physicochemical properties for oral bioavailability, whereas the derivatives 17g and 18c are nearly fully located in the pink area, only the degree of unsaturation slightly deviates from the optimum in 17g while the polarity is slightly deviates from the ideal for 18c (Figure 10).Fig. 10 Bioavailability radar plot from SwissADME online web tool for 17g, 18c and 18f. Fig 10 Moreover, it is attractive that 17g and 18f follow all rules of drug-likeness, they do not violate Lipinski's rule [40], Veber rule [41], Ghose-filter [42], Egan [43] or Muegge´s filter [44]. Because of the high TPSA of compound 18c, it satisfies only Lipinski's rule. In addition, it worth pointing out that the three derivatives do not incorporate in their structures Pan Assay Interference (PAINS) fragments [45]. Hence, in addition to the promising antiviral activity of the target compounds, their promising drug-likeness parameters suggests their potential to be subjected for future optimization for the discovery of chemotherapeutic agents. 3 Conclusion The presented study involves the design, synthesis, and anti-SARS-CoV-2 activity evaluation of two novel series of (tetrahydro)thioquinazoline-N-arylacetamides and (tetrahydro)thioquinazoline-N-arylacetohydrazides. The thioquinazoline-N-arylacetamide 17g beside the tetrahydrothioquinazoline-N-arylacetohydrazides 18c and 18f showed potent antiviral activity with IC50 = 21.4, 38.45, and 26.4 µM, respectively. The derivatives 18c and 18f exhibited SI = 10.67 and 16.04, respectively, towards the virus over the host cells. In addition, the three derivatives 17g, 18c, and 18f showed promising virucidal properties beside their ability to inhibit the virus at the adsorption as well as at the replication stages. Moreover, besides their promising antiviral activity, compounds 17g, 18c and 18f displayed acceptable physicochemical and pharmacokinetic properties for further optimization as antiviral agents. 4 Experimental 4.1 Chemistry 4.1.1 General remarks Chemicals along with solvents used for chemical reaction were obtained from commercial companies. Follow up of the reactions were carried out using analytical thin layer chromatography (TLC). Uncorrected Melting points were recorded on a Stuart SMP30 melting point apparatus. Elemental analyses of the synthesized hybrids were recorded in the micro analytical labs, National Research Centre, Cairo, Egypt. IR spectra (4000–400 cm−1) were recorded on Jasco FT/IR 300 E Fourier transform infrared spectrophotometer. 1H NMR as well as 13C NMR spectra were measured DMSO-d 6 as a solvent at 500 (125) MHz and 400 (100) MHz on Bruker instruments. 4.1.2 General procedure I for the synthesis of the target compounds 11a-c, 12a-c, 17a-g and 18a-g A mixture of 3 or 6 and anhydrous K2CO3 was stirred for 30 min at room temperature then 3a-c or 16a-g was added and the mixture was heated under reflux for 3 h. The reaction mixture was then cooled to room temperature poured on ice and neutralized with few drops of 2 N HCl and the precipitated product was filtered, dried and purified by crystallization from MeOH; DCM 1:1 mixture to give the corresponding target product 11a-c, 12a-c, 17a-g, 18a-g in analytical pure form. 2-(2-((4-Oxo-3,4-dihydroquinazolin-2-yl)thio)acetamido)benzoic acid (11a) Off white powder; yield = 95%; mp 250-252 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 4.15 (s, 2H), 7.13 (t, 3 J = 7.5 Hz, 1H), 7.38 (t, 3 J = 7.5 Hz, 1H), 7.46 (d, 3 J = 8.0 Hz, 1H), 7.56 (t, 3 J = 7.5 Hz, 1H), 7.69 (t, 3 J = 7.5 Hz, 1H), 7.95 (d, 3 J = 7.5 Hz, 1H), 8.00 (d, 3 J = 7.5 Hz, 1H), 8.49 (d, 3 J = 8.5 Hz, 1H), 11.77 (br., 1H), 12.47 (br., 1H), 12.76 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 35.08, 115.85, 116.70, 119.94, 122.92, 124.32, 125.79, 126.02, 131.13, 133.97, 134.56, 140.47, 148.22, 154.62, 161.17, 166.65, 169.34 ppm; Anal. Calcd for C17H13N3O4S: C, 57.46; H, 3.69; N, 11.82. Found: C, 57.73; H, 3.93; N, 11.65. 3-(2-((4-Oxo-3,4-dihydroquinazolin-2-yl)thio)acetamido)benzoic acid (11b) Off white powder; yield = 87%; mp 252-254 °C; 1H-NMR (400 MHz; DMSO-d 6) δ H 4.19 (s, 2H), 7.40 (d, 3 J = 7.2 Hz, 1H), 7.43 (t, 3 J = 9.2 Hz, 1H), 7.63 (d, 3 J = 7.6 Hz, 1H), 7.72 (dt, 3 J = 7.8 Hz, 4 J = 1.2 Hz, 1H), 7.80 (dd, 3 J = 8.6 Hz, 4 J = 1.2 Hz, 1H), 8.02 (d, 3 J = 8.0 Hz, 4 J = 1.2 Hz, 1H), 8.25 (s, 1H), 10.55 (s, 1H), 12.68 (br., 2H), 12.95 ppm (br., 1H); 13C-NMR (100 MHz; DMSO-d 6) δ C 35.13, 115.86, 116.21, 119.91, 123.27, 124.28, 125.76, 126.10, 129.12, 131.39, 134.67, 139.18, 148.21, 155.26, 161.14, 166.24, 167.11 ppm; Anal. Calcd for C17H13N3O4S: C, 57.46; H, 3.69; N, 11.82. Found: C, 57.15; H, 3.87; N, 12.03. 4-(2-((4-Oxo-3,4-dihydroquinazolin-2-yl)thio)acetamido)benzoic acid (11c) Pale yellow powder; yield = 83%; mp 249-251 °C; 1H-NMR (400 MHz; DMSO-d 6) δ H 4.11 (s, 2H), 7.29-7.40 (m, 2H), 7.68-7.69 (m, 2H), 7.88-7.89 (m, 2H), 7.99-8.00 (m, 2H), 11.06 (br., 2H), 12.45 ppm (br., 1H); Anal. Calcd for C17H13N3O4S: C, 57.46; H, 3.69; N, 11.82. Found: C, 57.64; H, 3.90; N, 11.66. 2-(2-((4-Oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetamido)benzoic acid (12a) White powder; yield = 90%; mp 227-229 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 1.58-1.60 (m, 4H), 2.14-2.24 (m, 2H), 2.35-2.36 (m, 2H), 4.01 (s, 2H), 7.14-7.15 (m, 1H), 7.56-7.57 (m, 1H), 7.95-7.96 (m, 1H), 8.47 (d, 3 J = 6.5 Hz, 1H), 11.61 (s, 1H), 12.71 ppm (br., 2H); 13C-NMR (125 MHz; DMSO-d 6) δ C 21.35, 21.38, 21.67, 30.86, 34.97, 116.59, 116.67, 119.97, 122.88, 131.08, 133.96, 140.44, 155.90, 162.72, 166.71, 169.23 ppm; Anal. Calcd for C17H17N3O4S: C, 56.81; H, 4.77; N, 11.69. Found: C, 56.57; H, 4.45; N, 11.48. 3-(2-((4-Oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetamido)benzoic acid (12b) White powder; yield = 81%; mp 263-265 °C; 1H-NMR (400 MHz; DMSO-d 6) δ H 1.61-1.62 (m, 4H), 2.24-2.25 (m, 2H), 2.39-2.40 (m, 2H), 4.04 (s, 2H), 7.41 (t, 3 J = 8.0 Hz, 1H), 7.62 (d, 3 J = 7.6 Hz, 1H), 7.76 (d, 3 J = 8.0 Hz, 1H), 8.20 (s, 1H), 10.45 (s, 1H), 12.53 ppm (br., 1H); 13C-NMR (100 MHz; DMSO-d 6) δ C 21.40, 21.72, 30.83, 34.95, 116.53, 119.92, 123.14, 124.24, 129.03, 131.72, 139.08, 156.49, 159.42, 162.79, 166.28, 167.28 ppm; Anal. Calcd for C17H17N3O4S: C, 56.81; H, 4.77; N, 11.69. Found: C, 56.62; H, 4.95; N, 11.47. 4-(2-((4-Oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetamido)benzoic acid (12c) White powder; yield = 87%; mp 237-239 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 1.61-1.63 (m, 4H), 2.24-2.25 (m, 2H), 2.37-2.38 (m, 2H), 4.04 (s, 2H), 7.66 (d, 3 J = 8.0 Hz, 2H), 7.88 (d, 3 J = 8.0 Hz, 2H), 10.65 (s, 1H), 12.58 ppm (br., 2H); 13C-NMR (125 MHz; DMSO-d 6) δ C 21.95, 22.25, 31.30, 35.58, 116.98, 118.94, 126.38, 130.93, 143.27, 157.23, 163.56, 167.15, 167.71, 167.81 ppm; Anal. Calcd for C17H17N3O4S: C, 56.81; H, 4.77; N, 11.69. Found: C, 56.98; H, 4.53; N, 11.81. N'-(2-((4-Oxo-3,4-dihydroquinazolin-2-yl)thio)acetyl)benzohydrazide (17a) White powder; yield = 85%; mp 249-251 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 4.11 (s, 2H), 7.40-7.43 (m, 1H), 7.47-7.50 (m, 2H), 7.54-7.60 (m, 2H), 7.75-7.76 (m, 1H), 7.84-7.85 (m, 2H), 8.01-8.03 (m, 1H), 10.37 (br., 1H), 10.49 (br., 1H), 12.62 ppm (br., 1H); Anal. Calcd for C17H14N4O3S: C, 57.62; H, 3.98; N, 15.81. Found: C, 57.32; H, 4.15; N, 15.54. 4-Methyl-N'-(2-((4-oxo-3,4-dihydroquinazolin-2-yl)thio)acetyl)benzohydrazide (17b) Pale yellow powder; yield = 91%; mp 243-245 °C;1H-NMR (500 MHz; DMSO-d 6) δ H 2.34 (s, 3H), 4.11 (s, 2H), 7.27-7.28 (m, 2H), 7.42-7.43 (m, 2H), 7.60-7.61 (m, 1H), 7.76-7.77 (m, 2H), 8.02-8.04 (m, 1H), 10.31 (s, IH), 10.41 (s, 1H), 12.69 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 20.95, 32.10, 119.97, 125.75, 125.96, 126.20, 127.46, 128.92, 129.54, 134.56, 141.81, 148.26, 154.78, 161.13, 165.25, 166.56 ppm; Anal. Calcd for C18H16N4O3S: C, 58.68; H, 4.38; N, 15.21. Found: C, 58.45; H, 4.68; N, 15.57. 4-Nitro-N'-(2-((4-oxo-3,4-dihydroquinazolin-2-yl)thio)acetyl)benzohydrazide (17c) Pale yellow powder; yield = 96%; mp 256-258 °C;1H-NMR (500 MHz; DMSO-d 6) δ H 4.13 (s, 2H), 7.39-7.44 (m, 1H), 7.60 (t, 3 J = 6.5 Hz, 1H), 7.76-7.78 (m, 1H), 8.02 (t, 3 J = 6.5 Hz, 1H), 8.07-8.09 (m, 2H), 8.33-8.34 (m, 2H), 10.50 (s, IH), 10.87 (s, 1H), 12.68 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 32.05, 119.93, 123.64, 125.74, 125.97, 128.99, 134.53, 137.96, 148.21, 149.37, 154.87, 161.20, 163.78, 166.50 ppm; Anal. Calcd for C17H13N5O5S: C, 51.13; H, 3.28; N, 17.54. Found: C, 51.46; H, 3.55; N, 17.16. 2-Methoxy-N'-(2-((4-oxo-3,4-dihydroquinazolin-2-yl)thio)acetyl)benzohydrazide (17d) White solid; yield = 85%; mp 224-226 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 3.86 (s, 3H), 4.10 (s, 2H), 7.05 (t, 3 J = 7.5 Hz, 1H), 7.14 (d, 3 J = 8.5 Hz, 1H), 7.42 (t, 3 J = 7.5 Hz, 1H), 7.50 (dt, 3 J = 8.0 Hz, 4 J = 1.5 Hz, 1H), 7.58 (d, 3 J = 8.0 Hz, 1H), 7.71 (dd, 3 J = 7.8 Hz, 4 J = 1.5 Hz, 1H), 7.76 (dt, 3 J = 7.5 Hz, 4 J = 1.5 Hz, 1H), 8.03 (dd, 3 J = 7.5 Hz, 4 J = 1.0 Hz, 1H), 10.08 (s, 1H), 10.65 (br., 1H), 12.71 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 31.97, 55.93, 112.11, 119.94, 120.55, 121.14, 125.70, 125.97, 130.39, 132.80, 134.52, 148.15, 155.07, 156.99, 161.28, 163.39, 165.52 ppm; Anal. Calcd for C18H16N4O4S: C, 56.24; H, 4.20; N, 14.58. Found: C, 56.49; H, 4.53; N, 14.81. 2-Chloro-N'-(2-((4-oxo-3,4-dihydroquinazolin-2-yl)thio)acetyl)benzohydrazide (17e) White solid; yield = 74%; mp 246-248 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 4.09 (s, 2H), 7.40-7.43 (m, 2H), 7.45-7.49 (m, 2H), 7.52 (d, 3 J = 7.5 Hz, 1H), 7.59 (d, 3 J = 8.0 Hz, 1H), 7.75 (t, 3 J = 7.5 Hz, 1H), 8.02 (d, 3 J = 7.5 Hz, 1H), 10.47 (s, 1H), 10.50 (br., 1H), 12.70 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 32.04, 119.95, 125.74, 125.94, 126.21, 127.07, 129.28, 129.78, 130.39, 131.42, 134.54, 140.40, 148.25, 154.77, 161.14, 165.05, 166.22 ppm; Anal. Calcd for C17H13ClN4O3S: C, 52.51; H, 3.37; N, 14.41. Found: C, 52.21; H, 3.66; N, 14.74. 4-Chloro-N'-(2-((4-oxo-3,4-dihydroquinazolin-2-yl)thio)acetyl)benzohydrazide (17f) White solid; yield = 86%; mp 257-259 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 4.11 (s, 2H), 7.42 (t, 3 J = 8.0 Hz, 1H), 7.57 (d, 3 J = 8.5 Hz, 2H), 7.60 (d, 3 J = 8.0 Hz, 1H), 7.77 (t, 3 J = 8.5 Hz, 1H), 7.87 (d, 3 J = 8.5 Hz, 2H), 8.03 (d, 3 J = 7.5 Hz, 1H), 10.39 (s, IH), 10.60 (s, 1H), 12.70 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 32.08, 119.94, 125.76, 125.97, 126.21, 128.57, 129.38, 131.08, 134.57, 136.70, 148.26, 154.75, 161.13, 164.37, 166.56 ppm; Anal. Calcd for C17H13ClN4O3S: C, 52.51; H, 3.37; N, 14.41. Found: C, 52.77; H, 3.05; N, 14.73. 2-Bromo-N'-(2-((4-oxo-3,4-dihydroquinazolin-2-yl)thio)acetyl)benzohydrazide (17g) White solid; yield = 93%; mp 233-235 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 4.10 (s, 2H), 7.41-7.42 (m, 4H), 7.58 (t, 3 J = 8.0 Hz, 1H), 7.66 (t, 3 J = 8.0 Hz, 1H), 7.73-7.75 (m, 1H), 8.01 (t, 3 J = 8.0 Hz, 1H), 10.46 (s, 1H), 10.50 (br., 1H), 12.69 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 32.03, 119.27, 119.94, 125.73, 125.93, 126.20, 127.52, 129.30, 131.52, 132.91, 134.53, 136.56, 148.24, 154.78, 161.11, 165.85, 166.19 ppm; Anal. Calcd for C17H13BrN4O3S: C, 47.13; H, 3.02; N, 12.93. Found: C, 47.38; H, 3.32; N, 12.66. N'-(2-((4-oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetyl)benzohydrazide (18a) White powder; yield = 90%; mp 246-248 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 1.60-1.65 (m, 4H), 2.24-2.25 (m, 2H), 2.49 (ov., 2H), 3.96 (s, 2H), 7.44-7.47 (m, 2H), 7.54 (t, 3 J = 7.5 Hz, 1H), 7.83 (d, 3 J = 7.5 Hz, 2H), 10.23 (s, IH), 10.44 (s, 1H), 12.52 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 21.42, 21.77, 30.99, 31.92, 116.68, 127.43, 128.41, 131.81, 132.34, 156.28, 160.44, 162.73, 165.36, 166.66 ppm; Anal. Calcd for C17H18N4O3S: C, 56.97; H, 5.06; N, 15.63. Found: C, 56.67; H, 5.29; N, 15.92. 4-Methyl-N'-(2-((4-oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetyl)benzohydrazide (18b) Off white powder; yield = 91%; mp 245-247 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 1.61-1.66 (m, 4H), 2.24-2.25 (br., 2H), 2.32 (s, 3H), 2.49 (br. ov, 2H), 3.95 (s, 2H), 7.26 (d, 3 J = 7.0 Hz, 2H), 7.74 (d, 3 J = 6.5 Hz, 2H), 10.19 (s, 1H), 10.35 (s, 1H), 12.52 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 20.95, 21.41, 21.78, 30.96, 31.92, 116.84, 127.44, 128.91, 129.52, 141.79, 155.53, 160.59, 162.58, 165.21, 166.62 ppm; Anal. Calcd for C18H20N4O3S: C, 58.05; H, 5.41; N, 15.04. Found: C, 58.27; H, 5.23; N, 15.38. 4-Nitro-N'-(2-((4-oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetyl)benzohydrazide (18c) Pale yellow powder; yield = 92%; mp 251-253 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 1.61-1.66 (m, 4H), 2.25 (br., 2H), 2.49 (br., 2H), 3.96 (s, 2H), 8.05 (d, 3 J = 8.5 Hz, 2H), 8.31 (d, 3 J = 8.5 Hz, 2H), 10.35 (s, 1H), 10.84 (br., 1H), 12.44 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 21.43, 21.78, 30.98, 31.88, 116.60, 123.64, 129.01, 137.97, 149.38, 156.12, 161.02, 163.80, 166.59 ppm; Anal. Calcd for C17H17N5O5S: C, 50.61; H, 4.25; N, 17.36. Found: C, 50.89; H, 4.05; N, 17.63. 2-Methoxy-N'-(2-((4-oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetyl)benzohydrazide (18d) Off white powder; yield = 93%; mp 212-214 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 1.60-1.66 (m, 4H), 2.25-2.26 (m, 2H), 2.49 (ov. br., 2H), 3.85 (s, 3H), 3.95 (s, 2H), 7.04 (t, 3 J = 7.5 Hz, 1H), 7.14 (d, 3 J = 8.5 Hz, 1H), 7.49 (t, 3 J = 7.0 Hz, 1H), 7.69 (d, 3 J = 7.5 Hz, 1H), 10.04 (s, 1H), 10.52 (br., 1H), 12.57 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 21.41, 21.78, 30.97, 31.79, 55.95, 112.12, 116.53, 120.57, 121.12, 130.16, 130.41, 132.83, 157.01, 160.14, 162.47, 163.36, 165.59 ppm; Anal. Calcd for C18H20N4O4S: C, 55.66; H, 5.19; N, 14.42. Found: C, 55.45; H, 5.47; N, 14.08. 2-Chloro-N'-(2-((4-oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetyl)benzohydrazide (18e) Off white powder; yield = 95%; mp 238-240 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H1.61-1.65 (m, 4H), 2.26 (br. ov, 2H), 2.49 (br., 2H), 3.94 (s, 2H), 7.42-7.50 (m, 4H), 10.38 (s, 1H), 10.41 (s, 1H), 12.54 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) 21.40, 21.77, 30.92, 31.85, 116.59, 127.05, 127.99, 129.27, 129.77, 130.38, 131.41, 134.45, 160.36, 162.50, 165.01, 166.28 ppm; Anal. Calcd for C17H17ClN4O3S: C, 51.97; H, 4.36; N, 14.26. Found: C, 51.73; H, 4.62; N, 14.54. 4-Chloro-N'-(2-((4-oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetyl)benzohydrazide (18f) white powder; yield = 95%; mp 252-254 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H1.61-1.66 (m, 4H), 2.19-2.26 (br., 2H), 2.34-2.49 (m, 2H), 3.95 (s, 2H), 7.54 (d, 3 J = 8.0 Hz, 2H), 7.85 (d, 3 J = 8.5 Hz, 2H), 10.23 (s, 1H), 10.52 (s, 1H), 12.49 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 21.40, 21.77, 30.95, 31.88, 116.34, 126.43, 128.54, 129.35, 131.06, 136.68, 160.13, 162.32, 164.31, 166.60 ppm; Anal. Calcd for C17H17ClN4O3S: C, 51.97; H, 4.36; N, 14.26. Found: C, 51.77; H, 4.72; N, 14.48. 2-Bromo-N'-(2-((4-oxo-3,4,5,6,7,8-hexahydroquinazolin-2-yl)thio)acetyl)benzohydrazide (18g) White powder; yield = 94%; mp 248-250 °C; 1H-NMR (500 MHz; DMSO-d 6) δ H 1.61-1.65 (m, 4H), 2.25-2.26 (m, 2H), 2.49 (br. ov, 2H), 3.94 (s, 2H), 7.38-7.45 (m, 3H), 7.65-7.67 (m, 1H), 10.39 (s, 2H), 12.55 ppm (br., 1H); 13C-NMR (125 MHz; DMSO-d 6) δ C 21.43, 21.80, 30.96, 31.88, 119.30, 127.57, 129.34, 131.58, 132.95, 132.96, 136.59, 156.27, 160.34, 162.55, 165.91, 166.35 ppm; Anal. Calcd for C17H17BrN4O3S: C, 46.69; H, 3.92; N, 12.81. Found: C, 46.45; H, 3.78; N, 12.57. 4.2 Biological Evaluation 4.2.1 In vitro bioassay of cytotoxicity and antiviral activity 4.2.1.1 MTT cytotoxicity assay The cytotoxic activity of the synthesized compounds were determined employing MTT assay as previously described [32] 4.2.1.2 Inhibitory concentration 50 (IC50) determination The values of IC50 for the target quinazolines were determined as reported [46]. 4.2.1.3 Mechanism of Action(s) To investigate whether the most potent candidates affect the (a) viral adsorption, (b) viral replication, or (c) has a virucidal effect, the plaque infectivity reduction assay was performed according to the reported procedure [46]. 4.3 Molecular Modeling Molecular docking studies were carried out using Molecular Operating Environment (MOE, 2020.0901) software. All minimizations were performed with MOE until an RMSD gradient of 0.05 kcal∙mol−1Å−1 with MMFF94x force field and the partial charges were automatically calculated. The X-ray crystallographic structure of SARS-CoV-2 main protease (Mpro) co-crystalized with a pyrimidine-2,4-dione inhibitor (YD1) (PDB ID: 7LTJ) was downloaded from the protein data bank [36]. Water molecules and ligands which are not involved in the binding were first removed. Next, the protein structure was prepared for the molecular docking study using QuickPrep protocol in MOE with the default options. YD1 exhibits several binding interactions with Mpro active site with the amino acids Asn142, Gly143, Cys145, His163, Met165, Glu166, and Asp187 either directly or through water mediated interactions (Figure 5). To perform the molecular docking study, the co-crystalized ligand (YD1) was used to define the active site and Triangle Matcher placement method and London dG scoring function were used. The molecular docking protocol was first validated by self-docking of the co-crystallized ligand (YD1) in the vicinity of the enzyme active site. The self-docking step reproduced the co-crystalized ligand pose efficiently with docking score (S) of −13.26 kcal/mol and a root mean square deviation (RMSD) of 1.538 Å, moreover, the docking protocol reproduced all the key interactions with the active site amino acids indicating the suitability of the adopted molecular docking protocol for the intended molecular docking study (Figure 6). 4.4 Estimation of physicochemical, pharmacokinetic and ADME properties The open SwissADME web tool available from the Swiss Institute of Bioinformatics (SIB) was used for the calculation of the physicochemical descriptors as well as to predict the ADME parameters, and pharmacokinetic properties of the most potent compounds [37]. The compounds’ structures were drawn on the web user interface, converted to SMILES notations, then submitted to the online server for calculation. Author statement Heba T. Abdel-Mohsen: Participated in the development of the idea of the project, organic synthesis of the target compounds, structure elucidation of the target compounds, analysis of the biological results, writing, revising and finalizing the manuscript. Mohamed A. Omar: Participated in organic synthesis of the target compounds and revising the interpretation of NMR results. Omnia Kutkat: Participated in analysis of the synthesized compounds for their antiviral activity. Ahmed M. El Kerdawy: Run the computational studies to the target compounds, analysed the obtained results and revised the manuscript. Alaa A. Osman: Participated in the in silico analyses of the target compounds. Mohamed GabAllah: Participated in analysis of the synthesized compounds for their antiviral activity. Ahmed Mostafa: Participated in the development of the idea of the project, analysis of the synthesized compounds for their antiviral activity and revising the biological part of the manuscript. Mohamed A. Ali: Participated in the development of the idea of the project and revising the biological part of the manuscript. Hoda I. El Diwani: Participated in the development of the idea of the project and revising the manuscript. Conflict of interest The authors have no conflict of interest to declare. Appendix Supplementary materials Image, application 1 Data Availability Data will be made available on request. Acknowledgments Special thanks to National Research Centre (Egypt) for their fund through the project ID 12060119. Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.molstruc.2022.134690. ==== Refs References 1 Chen N. Zhou M. Dong X. Qu J. Gong F. Han Y. Qiu Y. Wang J. Liu Y. Wei Y. Xia J. Yu T. Zhang X. Zhang L. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study Lancet (London, England) 395 10223 2020 507 513 32007143 2 Chen T. Wu D. Chen H. Yan W. Yang D. Chen G. Ma K. Xu D. Yu H. Wang H. Wang T. Guo W. Chen J. Ding C. Zhang X. Huang J. Han M. Li S. Luo X. Zhao J. Ning Q. Clinical characteristics of 113 deceased patients with coronavirus disease 2019: retrospective study BMJ 368 2020 m1091 32217556 3 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 33116300 4 Rothan H.A. Byrareddy S.N. The epidemiology and pathogenesis of coronavirus disease (COVID-19) outbreak J Autoimmun 109 2020 102433 5 Harvey W.T. Carabelli A.M. Jackson B. Gupta R.K. Thomson E.C. Harrison E.M. Ludden C. Reeve R. Rambaut A. Consortium C.-G.U. Peacock S.J. Robertson D.L. SARS-CoV-2 variants, spike mutations and immune escape Nat Rev Microbiol 19 7 2021 409 424 34075212 6 Saied E.M. El-Maradny Y.A. Osman A.A. Darwish A.M.G. Abo Nahas H.H. Niedbala G. Piekutowska M. Abdel-Rahman M.A. Balbool B.A. Abdel-Azeem A.M. A Comprehensive Review about the Molecular Structure of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2): Insights into Natural Products against COVID-19 Pharmaceutics 13 11 2021 7 Zhao J. Zhao S. Ou J. Zhang J. Lan W. Guan W. Wu X. Yan Y. Zhao W. Wu J. Chodosh J. Zhang Q. COVID-19: Coronavirus Vaccine Development Updates Front Immunol 11 2020 602256 8 Artese A. Svicher V. Costa G. Salpini R. Di Maio V.C. Alkhatib M. Ambrosio F.A. Santoro M.M. Assaraf Y.G. Alcaro S. Ceccherini-Silberstein F. Current status of antivirals and druggable targets of SARS CoV-2 and other human pathogenic coronaviruses Drug Resist Updat 53 2020 100721 9 Walls A.C. Park Y.J. Tortorici M.A. Wall A. McGuire A.T. Veesler D. Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein Cell 181 2 2020 281 292 e6 32155444 10 Elfiky A.A. SARS-CoV-2 RNA dependent RNA polymerase (RdRp) targeting: an in silico perspective J Biomol Struct Dyn 39 9 2021 3204 3212 32338164 11 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 32272481 12 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 33531496 13 Huff S. Kummetha I.R. Tiwari S.K. Huante M.B. Clark A.E. Wang S. Bray W. Smith D. Carlin A.F. Endsley M. Rana T.M. Discovery and Mechanism of SARS-CoV-2 Main Protease Inhibitors Journal of medicinal chemistry 2021 14 Cannalire R. Cerchia C. Beccari A.R. Di Leva F.S. Summa V. Targeting SARS-CoV-2 Proteases and Polymerase for COVID-19 Treatment: State of the Art and Future Opportunities Journal of medicinal chemistry 2020 15 Bansal R. Malhotra A. Therapeutic progression of quinazolines as targeted chemotherapeutic agents Eur J Med Chem 211 2021 113016 16 Das R. Mehta D.K. Dhanawat M. Bestowal of Quinazoline Scaffold in Anticancer Drug Discovery Anticancer Agents Med Chem 21 11 2021 1350 1368 32593282 17 Matharu D.S. Flaherty D.P. Simpson D.S. Schroeder C.E. Chung D. Yan D. Noah J.W. Jonsson C.B. White E.L. Aube J. Plemper R.K. Severson W.E. Golden J.E. Optimization of potent and selective quinazolinediones: inhibitors of respiratory syncytial virus that block RNA-dependent RNA-polymerase complex activity Journal of medicinal chemistry 57 24 2014 10314 10328 25399509 18 Zhang G. Wang M. Zhao J. Wang Y. Zhu M. Wang J. Cen S. Wang Y. Design, synthesis and in vitro anti-influenza A virus evaluation of novel quinazoline derivatives containing S-acetamide and NH-acetamide moieties at C-4 Eur J Med Chem 206 2020 112706 19 Hwu J.R. Kapoor M. Gupta N.K. Tsay S.C. Huang W.C. Tan K.T. Hu Y.C. Lyssen P. Neyts J. Synthesis and antiviral activities of quinazolinamine-coumarin conjugates toward chikungunya and hepatitis C viruses Eur J Med Chem 232 2022 114164 20 Lee J.Y. Shin Y.S. Lee J. Kwon S. Jin Y.H. Jang M.S. Kim S. Song J.H. Kim H.R. Park C.M. Identification of 4-anilino-6-aminoquinazoline derivatives as potential MERS-CoV inhibitors Bioorg Med Chem Lett 30 20 2020 127472 21 Zhao J. Zhang Y. Wang M. Liu Q. Lei X. Wu M. Guo S. Yi D. Li Q. Ma L. Liu Z. Guo F. Wang J. Li X. Wang Y. Cen S. Quinoline and Quinazoline Derivatives Inhibit Viral RNA Synthesis by SARS-CoV-2 RdRp ACS Infect Dis 7 6 2021 1535 1544 34038639 22 Rothan H.A. Teoh T.C. Cell-Based High-Throughput Screening Protocol for Discovering Antiviral Inhibitors Against SARS-COV-2 Main Protease (3CLpro) Mol Biotechnol 63 3 2021 240 248 33464543 23 Moyle G. Boffito M. Stoehr A. Rieger A. Shen Z. Manhard K. Sheedy B. Hingorani V. Raney A. Nguyen M. Nguyen T. Ong V. Yeh L.T. Quart B. Phase 2a randomized controlled trial of short-term activity, safety, and pharmacokinetics of a novel nonnucleoside reverse transcriptase inhibitor, RDEA806, in HIV-1-positive, antiretroviral-naive subjects Antimicrob Agents Chemother 54 8 2010 3170 3178 20498326 24 Zhan P. Li X. Li Z. Chen X. Tian Y. Chen W. Liu X. Pannecouque C. De Clercq E. Structure-based bioisosterism design, synthesis and biological evaluation of novel 1,2,4-triazin-6-ylthioacetamides as potent HIV-1 NNRTIs Bioorg Med Chem Lett 22 23 2012 7155 7162 23084898 25 Zhang G.N. Li Q. Zhao J. Zhang X. Xu Z. Wang Y. Fu Y. Shan Q. Zheng Y. Wang J. Zhu M. Li Z. Cen S. He J. Wang Y. Design and synthesis of 2-((1H-indol-3-yl)thio)-N-phenyl-acetamides as novel dual inhibitors of respiratory syncytial virus and influenza virus A Eur J Med Chem 186 2020 111861 26 Yu M. Liu A. Du G. Naesens L. Vanderlinden E. De Clercq E. Liu X. Discovery of dihydro-alkyloxy-benzyl-oxopyrimidines as promising anti-influenza virus agents Chem Biol Drug Des 78 4 2011 596 602 21752202 27 Zhan P. Wang L. Liu H. Chen X. Li X. Jiang X. Zhang Q. Liu X. Pannecouque C. Naesens L. De Clercq E. Liu A. Du G. Arylazolyl(azinyl)thioacetanilide. Part 9: Synthesis and biological investigation of thiazolylthioacetamides derivatives as a novel class of potential antiviral agents Arch Pharm Res 35 6 2012 975 986 22870806 28 Chou S.-Y. Yin W.-K. Chung Y.-S. Chang L.-S. Liu C.-W. Chen S.-F. Shih K.-S. Kilogram-Scale Synthesis of a Highly Selective α 1-Adrenoceptor Antagonist (DL-028A) Org. Process Res. Dev. 6 3 2002 273 278 29 Abdel-Mohsen H.T. Omar M.A. Petreni A. Supuran C.T. Novel 2-substituted thioquinazoline-benzenesulfonamide derivatives as carbonic anhydrase inhibitors with potential anticancer activity Arch Pharm (Weinheim) 2022 e2200180 30 Abdel-Mohsen H.T. Conrad J. Harms K. Nohr D. Beifuss U. Laccase-catalyzed green synthesis and cytotoxic activity of novel pyrimidobenzothiazoles and catechol thioethers RSC Adv 7 28 2017 17427 17441 31 Abdel-Mohsen H.T. Petreni A. Supuran C.T. Investigation of the carbonic anhydrase inhibitory activity of benzenesulfonamides incorporating substituted fused-pyrimidine tails Arch Pharm (Weinheim) 355 11 2022 e2200274 32 Mosmann T. Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays Journal of Immunological Methods 65 1 1983 55 63 6606682 33 Kandeil A. Mostafa A. El-Shesheny R. Shehata M. Roshdy W.H. Ahmed S.S. Gomaa M. Taweel A.E. Kayed A.E. Mahmoud S.H. Moatasim Y. Kutkat O. Kamel M.N. Mahrous N. Sayes M.E. Guindy N.M.E. Naguib A. Ali M.A. Coding-Complete Genome Sequences of Two SARS-CoV-2 Isolates from Egypt Microbiol Resour Announc 9 22 2020 34 Mostafa A. Kandeil A. Y A.M.M.E. Kutkat O. Moatasim Y. Rashad A.A. Shehata M. Gomaa M.R. Mahrous N. Mahmoud S.H. GabAllah M. Abbas H. Taweel A.E. Kayed A.E. Kamel M.N. Sayes M.E. Mahmoud D.B. El-Shesheny R. Kayali G. Ali M.A. FDA-Approved Drugs with Potent In Vitro Antiviral Activity against Severe Acute Respiratory Syndrome Coronavirus 2 Pharmaceuticals (Basel) 13 12 2020 35 Mengist H.M. Dilnessa T. Jin T. Structural Basis of Potential Inhibitors Targeting SARS-CoV-2 Main Protease Front Chem 9 2021 622898 36 Clyde A. Galanie S. Kneller D.W. Ma H. Babuji Y. Blaiszik B. Brace A. Brettin T. Chard K. Chard R. Coates L. Foster I. Hauner D. Kertesz V. Kumar N. Lee H. Li Z. Merzky A. Schmidt J.G. Tan L. Titov M. Trifan A. Turilli M. Van Dam H. Chennubhotla S.C. Jha S. Kovalevsky A. Ramanathan A. Head M.S. Stevens R. High-Throughput Virtual Screening and Validation of a SARS-CoV-2 Main Protease Noncovalent Inhibitor Journal of chemical information and modeling 62 1 2022 116 128 34793155 37 Daina A. Michielin O. Zoete V. SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules Scientific reports 7 2017 42717 28256516 38 Daina A. Michielin O. Zoete V. iLOGP: a simple, robust, and efficient description of n-octanol/water partition coefficient for drug design using the GB/SA approach Journal of chemical information and modeling 54 12 2014 3284 3301 25382374 39 Amin M.L. P-glycoprotein Inhibition for Optimal Drug Delivery Drug target insights 7 2013 27 34 24023511 40 Lipinski C.A. Lombardo F. Dominy B.W. Feeney P.J. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings Adv Drug Deliv Rev 46 1-3 2001 3 26 11259830 41 Veber D.F. Johnson S.R. Cheng H.Y. Smith B.R. Ward K.W. Kopple K.D. Molecular properties that influence the oral bioavailability of drug candidates Journal of medicinal chemistry 45 12 2002 2615 2623 12036371 42 Ghose A.K. Viswanadhan V.N. Wendoloski J.J. A knowledge-based approach in designing combinatorial or medicinal chemistry libraries for drug discovery. 1. A qualitative and quantitative characterization of known drug databases J Comb Chem 1 1 1999 55 68 10746014 43 Egan W.J. Merz K.M. Jr. Baldwin J.J. Prediction of drug absorption using multivariate statistics Journal of medicinal chemistry 43 21 2000 3867 3877 11052792 44 Muegge I. Heald S.L. Brittelli D. Simple selection criteria for drug-like chemical matter Journal of medicinal chemistry 44 12 2001 1841 1846 11384230 45 Baell J.B. Holloway G.A. New substructure filters for removal of pan assay interference compounds (PAINS) from screening libraries and for their exclusion in bioassays Journal of medicinal chemistry 53 7 2010 2719 2740 20131845 46 Mostafa A. Kandeil A. Elshaier Y.A.M.M. Kutkat O. Moatasim Y. Rashad A.A. Shehata M. Gomaa M.R. Mahrous N. Mahmoud S.H. GabAllah M. Abbas H. Taweel A.E. Kayed A.E. Kamel M.N. Sayes M.E. Mahmoud D.B. El-Shesheny R. Kayali G. Ali M.A. FDA-Approved Drugs with Potent In Vitro Antiviral Activity against Severe Acute Respiratory Syndrome Coronavirus 2 Pharmaceuticals 13 12 2020 443 33291642	36465802	PMC9709698	NO-CC CODE	2022-12-15 23:18:05	no	J Mol Struct. 2023 Mar 15; 1276:134690	utf-8	J Mol Struct	2,022	10.1016/j.molstruc.2022.134690	oa_other
==== Front Ann Intern Med Ann Intern Med aim Annals of Internal Medicine 0003-4819 1539-3704 American College of Physicians 36442063 10.7326/M22-2116 aim-olf-M222116 Original Research 9197Clinical trials3122457COVID-19362Hospitalizations9333Intubation3124810Mortality8693OxygenearlyCurrently Online FirstcoronavirusCoronavirus Disease 2019 (COVID-19)hospitalHospital MedicinerctRandomized-Controlled Trialpoc-eligiblePOC EligibleTemporal Improvements in COVID-19 Outcomes for Hospitalized Adults: A Post Hoc Observational Study of Remdesivir Group Participants in the Adaptive COVID-19 Treatment Trial Temporal Improvements in COVID-19 Outcomes for Hospitalized Adults Potter Gail E. PhD https://orcid.org/0000-0001-6667-6992 Bonnett Tyler MS https://orcid.org/0000-0001-5545-6433 Rubenstein Kevin MS https://orcid.org/0000-0003-3516-5932 Lindholm David A. MD https://orcid.org/0000-0001-5428-7404 Rapaka Rekha R. MD, PhD Doernberg Sarah B. MD, MAS https://orcid.org/0000-0003-1727-6014 Lye David C. MBBS https://orcid.org/0000-0003-0324-0205 Mularski Richard A. MD, MSHS, MCR https://orcid.org/0000-0001-6979-2542 Hynes Noreen A. MD, MPH https://orcid.org/0000-0001-6898-0621 Kline Susan MD, MPH Paules Catharine I. MD https://orcid.org/0000-0002-9064-6678 Wolfe Cameron R. MBBS, MPH https://orcid.org/0000-0002-5365-5030 Frank Maria G. MD https://orcid.org/0000-0002-0420-0574 Rouphael Nadine G. MD, MSc Deye Gregory A. MD https://orcid.org/0000-0001-5527-4541 Sweeney Daniel A. MD https://orcid.org/0000-0002-5398-3528 Colombo Rhonda E. MD, MHS Davey Richard T. MD https://orcid.org/0000-0002-3992-5524 Mehta Aneesh K. MD https://orcid.org/0000-0002-6552-9162 Whitaker Jennifer A. MD, MS https://orcid.org/0000-0002-7727-6086 Castro Jose G. MD https://orcid.org/0000-0002-8642-1724 Amin Alpesh N. MD, MBA https://orcid.org/0000-0002-9790-0245 Colombo Christopher J. MD, MA https://orcid.org/0000-0001-5499-3368 Levine Corri B. PhD, MS, MPH https://orcid.org/0000-0003-0405-5191 Jain Mamta K. MD, MPH https://orcid.org/0000-0003-2292-6544 Maves Ryan C. MD https://orcid.org/0000-0001-6234-6160 Marconi Vincent C. MD https://orcid.org/0000-0001-8409-4689 Grossberg Robert MD https://orcid.org/0000-0001-6322-8393 Hozayen Sameh MD, MSc https://orcid.org/0000-0002-4582-6530 Burgess Timothy H. MD, MPH https://orcid.org/0000-0003-1247-8370 Atmar Robert L. MD https://orcid.org/0000-0001-9989-6772 Ganesan Anuradha MBBS, MPH Gomez Carlos A. MD https://orcid.org/0000-0001-5486-5710 Benson Constance A. MD Lopez de Castilla Diego MD, MPH Ahuja Neera MD George Sarah L. MD Nayak Seema U. MD Cohen Stuart H. MD Lalani Tahaniyat MBBS, MHS Short William R. MD, MPH https://orcid.org/0000-0003-4225-8336 Erdmann Nathaniel MD, PhD Tomashek Kay M. MD, MPH, DTM https://orcid.org/0000-0001-7741-7072 * Tebas Pablo MD https://orcid.org/0000-0001-5345-7942 * Biostatistics Research Branch, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, Maryland (G.E.P.) Clinical Monitoring Research Program Directorate, Frederick National Laboratory for Cancer Research, Frederick, Maryland (T.B., K.R.) Department of Medicine, Uniformed Services University of the Health Sciences, Bethesda, Maryland, and Brooke Army Medical Center, Joint Base San Antonio-Fort Sam Houston, Texas (D.A.L.) University of Maryland Center for Vaccine Development and Global Health, University of Maryland School of Medicine, Baltimore, Maryland (R.R.R.) Department of Medicine, Division of Infectious Diseases, University of California, San Francisco, San Francisco, California (S.B.D.) National Centre for Infectious Diseases, Tan Tock Seng Hospital, Yong Loo Lin School of Medicine, and Lee Kong Chian School of Medicine, Singapore (D.C.L.) Department of Pulmonary and Critical Care Medicine, Northwest Permanente PC, and Kaiser Permanente Northwest Center for Health Research, Portland, Oregon (R.A.M.) Division of Infectious Diseases, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland (N.A.H.) Department of Medicine, Division of Infectious Diseases and International Medicine, University of Minnesota Medical School, Minneapolis, Minnesota (S.K.) Division of Infectious Diseases, Penn State Health Milton S. Hershey Medical Center, Hershey, Pennsylvania (C.I.P.) Division of Infectious Diseases, Department of Medicine, Duke University Medical Center, Durham, North Carolina (C.R.W.) Department of Medicine, Denver Health Hospital Authority, Denver, Colorado, and University of Colorado School of Medicine, Aurora, Colorado (M.G.F.) Hope Clinic, Emory Vaccine Center, Infectious Diseases Division, Atlanta, Georgia (N.G.R.) Division of Microbiology and Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland (G.A.D., S.U.N., K.M.T.) Division of Pulmonary, Critical Care and Sleep Medicine, Department of Medicine, University of California San Diego, San Diego, California (D.A.S.) Madigan Army Medical Center, Tacoma, Washington, Infectious Diseases Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland, and The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., Bethesda, Maryland (R.E.C.) National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland (R.T.D.) Division of Infectious Diseases, Emory University School of Medicine, and National Emerging Special Pathogens Training and Education Center, Atlanta, Georgia (A.K.M.) Departments of Molecular Virology and Microbiology and Medicine, Section of Infectious Diseases, Baylor College of Medicine, Houston, Texas (J.A.W.) Division of Infectious Diseases, University of Miami, Miami, Florida (J.G.C.) Department of Medicine, University of California, Irvine, Irvine, California (A.N.A.) Madigan Army Medical Center, Tacoma, Washington, and Department of Medicine, Uniformed Services University of the Health Sciences, Bethesda, Maryland (C.J.C.) Department of Internal Medicine, Division of Infectious Disease, University of Texas Medical Branch, Galveston, Texas (C.B.L.) Department of Internal Medicine, Division of Infectious Disease and Geographic Medicine, UT Southwestern Medical Center, and Parkland Health & Hospital System, Dallas, Texas (M.K.J.) Wake Forest University School of Medicine, Winston-Salem, North Carolina, and Infectious Diseases Clinical Research Program, Uniformed Services University of the Health Sciences, Bethesda, Maryland (R.C.M.) Emory University School of Medicine, Rollins School of Public Health, and Atlanta Veterans Affairs Medical Center, Atlanta, Georgia (V.C.M.) Department of Medicine, Division of Infectious Diseases, Montefiore Medical Center/Albert Einstein College of Medicine, Bronx, New York (R.G.) Department of Medicine, Division of Hospital Medicine, University of Minnesota, Minneapolis, Minnesota (S.H.) Infectious Disease Clinical Research Program, Department of Preventive Medicine and Biostatistics, Uniformed Services University of the Health Sciences, Bethesda, Maryland (T.H.B.) Department of Medicine, Baylor College of Medicine, Houston, Texas (R.L.A.) Infectious Disease Clinical Research Program, Department of Preventive Medicine and Biostatistics, Uniformed Services University of the Health Sciences, The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., and Walter Reed National Military Medical Center, Bethesda, Maryland (A.G.) Department of Internal Medicine, Division of Infectious Diseases, University of Nebraska Medical Center, Omaha, Nebraska (C.A.G.) Division of Infectious Diseases and Global Public Health, University of California San Diego, San Diego, California (C.A.B.) Division of Infectious Diseases, Evergreen Health Medical Center, Kirkland, Washington (D.L.) Department of Internal Medicine, Stanford University Medical Center, Palo Alto, California (N.A.) Saint Louis University and St. Louis VA Medical Center, Saint Louis, Missouri (S.L.G.) Division of Infectious Diseases, University of California, Davis, Sacramento, California (S.H.C.) Naval Medical Center Portsmouth, Portsmouth, Virginia, Infectious Disease Clinical Research Program, Department of Preventive Medicine and Biostatistics, Uniformed Services University of the Health Sciences, Bethesda, Maryland, and The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., Bethesda, Maryland (T.L.) Department of Medicine, Division of Infectious Diseases, University of Pennsylvania, Philadelphia, Pennsylvania (W.R.S.) Division of Infectious Diseases, Department of Medicine, School of Medicine, University of Alabama at Birmingham, Birmingham, Alabama (N.E.) Division of Infectious Diseases/Clinical Trials Unit, University of Pennsylvania, Philadelphia, Pennsylvania (P.T.). Disclaimer: The content of this article does not necessarily reflect the views or policies of the U.S. Department of Health and Human Services; the Uniformed Services University of the Health Sciences; the Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc.; the Departments of the Army, Navy, or Air Force; the Defense Health Agency; the Department of Defense; or the Department of Veterans Affairs, nor does any mention of trade names, commercial products, or organizations imply endorsement by the U.S. government. Acknowledgment: The authors thank Lori Dodd for the initial idea for this manuscript and Alyssa La Regina for outstanding administrative support. This work utilized the computational resources of the National Institutes of Health (NIH) high-performance computing Biowulf cluster (http://hpc.nih.gov). Financial Support: This work was supported with funds from the NIAID Division of Intramural Research and from the National Cancer Institute of the NIH under contract no. 75N91019D00024. The analysis used data from ACTT-1 (14), ACTT-2 (16), ACTT-3 (15), and ACTT-4 (17). The ACTT trials were sponsored and primarily funded by the NIAID of the NIH, Bethesda, Maryland. These trials have been funded in part with federal funds from NIAID and the National Cancer Institute of the NIH under contract HHSN261200800001E 75N910D00024, task order number 75N91019F00130/75N91020F00010, and by the Department of Defense, Defense Health Program. These trials have been supported in part by the NIAID of the NIH under award numbers UM1AI148684, UM1AI148576, UM1AI148573, UM1AI148575, UM1AI148452, UM1AI148685, UM1AI148450, and UM1AI148689. These trials have also been funded in part by the governments of Denmark, Japan, Mexico, and Singapore. The trial site in South Korea received funding from the Seoul National University Hospital. Support for the London International Coordinating Centre was also provided by the United Kingdom Medical Research Council (MRC_UU_12023/23). Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M22-2116. Data Sharing Statement: The following data will be made available with publication: deidentified participant data and data dictionary (https://accessclinicaldata.niaid.nih.gov). The following supporting documents will be made available with publication: statistical/analytical code (https://github.com/gepotter/actt1234). To access the data, a data access request (DAR) is required to be submitted to NIAID by the requester using the electronic DAR form as part of the process for requesting access found on the AccessClinicalData@NIAID data platform. The DAR will be reviewed by the NIAID Clinical Trials Data Access Committee. Upon approval of the DAR by NIAID and before accessing the data set, the primary requester and their institutional official will be notified and will be required to agree to and sign an NIAID Data Use Agreement. The agreement outlines the terms of use of the data and can be found on the AccessClinicalData@NIAID data platform. Corresponding Author: Gail E. Potter, PhD, National Institutes of Health, 5601 Fisher's Lane, MSC 9820, Rockville, MD 20892-9820; e-mail, [email protected]. Author Contributions: Conception and design: G.E. Potter, T. Bonnett, D.C. Lye, R.A. Mularski, N.A. Hynes, C.I. Paules, C.R. Wolfe, R.E. Colombo, J.A. Whitaker, C.J. Colombo, V.C. Marconi, R. Grossberg, S. Hozayen, C.A. Benson, S.U. Nayak, N. Erdmann, K.M. Tomashek, P. Tebas. Analysis and interpretation of the data: G.E. Potter, T. Bonnett, K. Rubenstein, D.A. Lindholm, R.R. Rapaka, S.B. Doernberg, D.C. Lye, R.A. Mularski, N.A. Hynes, S. Kline, C.R. Wolfe, N.G. Rouphael, G.A. Deye, D.A. Sweeney, R.E. Colombo, R.T. Davey, J.A. Whitaker, J.G. Castro, A.N. Amin, C.B. Levine, M.K. Jain, R.C. Maves, V.C. Marconi, S. Hozayen, T.H. Burgess, R.L. Atmar, C.A. Benson, N. Ahuja, S.L. George, S.U. Nayak, S.H. Cohen, N. Erdmann, K.M. Tomashek, P. Tebas. Drafting of the article: G.E. Potter, R.A. Mularski, N.A. Hynes, C.R. Wolfe, G.A. Deye, D.A. Sweeney, R.E. Colombo, J.G. Castro, A.N. Amin, C.J. Colombo, C.B. Levine, R.C. Maves, V.C. Marconi, S. Hozayen, R.L. Atmar, C.A. Benson, D. Lopez de Castilla, S.L. George, S.U. Nayak, N. Erdmann, K.M. Tomashek, P. Tebas. Critical revision for important intellectual content: G.E. Potter, T. Bonnett, K. Rubenstein, D.A. Lindholm, R.R. Rapaka, S.B. Doernberg, D.C. Lye, R.A. Mularski, N.A. Hynes, S. Kline, C.I. Paules, C.R. Wolfe, M.G. Frank, N.G. Rouphael, G.A. Deye, D.A. Sweeney, R.E. Colombo, A.K. Mehta, J.A. Whitaker, J.G. Castro, A.N. Amin, C.J. Colombo, C.B. Levine, R.C. Maves, R. Grossberg, S. Hozayen, R.L. Atmar, A. Ganesan, C.A. Gomez, C.A. Benson, S.L. George, S.U. Nayak, T. Lalani, N. Erdmann, K.M. Tomashek, P. Tebas. Final approval of the article: G.E. Potter, T. Bonnett, K. Rubenstein, D.A. Lindholm, R.R. Rapaka, S.B. Doernberg, D.C. Lye, R.A. Mularski, N.A. Hynes, S. Kline, C.I. Paules, C.R. Wolfe, M.G. Frank, N.G. Rouphael, G.A. Deye, D.A. Sweeney, R.E. Colombo, R.T. Davey, A.K. Mehta, J.A. Whitaker, J.G. Castro, A.N. Amin, C.J. Colombo, C.B. Levine, M.K. Jain, R.C. Maves, V.C. Marconi, R. Grossberg, S. Hozayen, T.H. Burgess, R.L. Atmar, A. Ganesan, C.A. Gomez, C.A. Benson, D. Lopez de Castilla, N. Ahuja, S.L. George, S.U. Nayak, S.H. Cohen, T. Lalani, W.R. Short, N. Erdmann, K.M. Tomashek, P. Tebas. Provision of study materials or patients: D.A. Lindholm, R.R. Rapaka, D.C. Lye, R.A. Mularski, N.A. Hynes, S. Kline, M.G. Frank, N.G. Rouphael, R.E. Colombo, R.T. Davey, J.A. Whitaker, C.J. Colombo, R.C. Maves, R. Grossberg, R.L. Atmar, A. Ganesan, C.A. Benson, N. Ahuja, S.H. Cohen, T. Lalani, P. Tebas. Statistical expertise: G.E. Potter, K. Rubenstein. Obtaining of funding: D.C. Lye. Administrative, technical, or logistic support: D.A. Lindholm, N.G. Rouphael, R.L. Atmar. Collection and assembly of data: D.A. Lindholm, R.R. Rapaka, S.B. Doernberg, D.C. Lye, R.A. Mularski, S. Kline, N.G. Rouphael, R.E. Colombo, R.T. Davey, A.K. Mehta, J.A. Whitaker, A.N. Amin, R.C. Maves, A. Ganesan, C.A. Benson, D. Lopez de Castilla, S.U. Nayak, N. Erdmann, K.M. Tomashek, P. Tebas. 29 11 2022 29 11 2022 M22-21162022 American College of Physicians This article is made available via the PMC Open Access Subset for unrestricted re-use for research, analyses, and text and data mining through PubMed Central. Acknowledgement of the original source shall include a notice similar to the following: "© 2020 American College of Physicians. Some rights reserved. This work permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited." 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 standard of care (SOC) for COVID-19 evolved rapidly during 2020 and 2021, but its cumulative effect over time is unclear. In this post hoc analysis of a series of phase 3 trials that evaluated COVID-19 therapeutics from February 2020 through May 2021, the authors sought to evaluate whether recovery and mortality improved as the SOC evolved. Visual Abstract. Temporal Improvements in COVID-19 Outcomes for Hospitalized Adults. The standard of care (SOC) for COVID-19 evolved rapidly during 2020 and 2021, but its cumulative effect over time is unclear. In this post hoc analysis of a series of phase 3 trials that evaluated COVID-19 therapeutics from February 2020 through May 2021, the authors sought to evaluate whether recovery and mortality improved as the SOC evolved. Background: The COVID-19 standard of care (SOC) evolved rapidly during 2020 and 2021, but its cumulative effect over time is unclear. Objective: To evaluate whether recovery and mortality improved as SOC evolved, using data from ACTT (Adaptive COVID-19 Treatment Trial). Design: ACTT is a series of phase 3, randomized, double-blind, placebo-controlled trials that evaluated COVID-19 therapeutics from February 2020 through May 2021. ACTT-1 compared remdesivir plus SOC to placebo plus SOC, and in ACTT-2 and ACTT-3, remdesivir plus SOC was the control group. This post hoc analysis compared recovery and mortality between these comparable sequential cohorts of patients who received remdesivir plus SOC, adjusting for baseline characteristics with propensity score weighting. The analysis was repeated for participants in ACTT-3 and ACTT-4 who received remdesivir plus dexamethasone plus SOC. Trends in SOC that could explain outcome improvements were analyzed. (ClinicalTrials.gov: NCT04280705 [ACTT-1], NCT04401579 [ACTT-2], NCT04492475 [ACTT-3], and NCT04640168 [ACTT-4]) Setting: 94 hospitals in 10 countries (86% U.S. participants). Participants: Adults hospitalized with COVID-19. Intervention: SOC. Measurements: 28-day mortality and recovery. Results: Although outcomes were better in ACTT-2 than in ACTT-1, adjusted hazard ratios (HRs) were close to 1 (HR for recovery, 1.04 [95% CI, 0.92 to 1.17]; HR for mortality, 0.90 [CI, 0.56 to 1.40]). Comparable patients were less likely to be intubated in ACTT-2 than in ACTT-1 (odds ratio, 0.75 [CI, 0.53 to 0.97]), and hydroxychloroquine use decreased. Outcomes improved from ACTT-2 to ACTT-3 (HR for recovery, 1.43 [CI, 1.24 to 1.64]; HR for mortality, 0.45 [CI, 0.21 to 0.97]). Potential explanatory factors (SOC trends, case surges, and variant trends) were similar between ACTT-2 and ACTT-3, except for increased dexamethasone use (11% to 77%). Outcomes were similar in ACTT-3 and ACTT-4. Antibiotic use decreased gradually across all stages. Limitation: Unmeasured confounding. Conclusion: Changes in patient composition explained improved outcomes from ACTT-1 to ACTT-2 but not from ACTT-2 to ACTT-3, suggesting improved SOC. These results support excluding nonconcurrent controls from analysis of platform trials in rapidly changing therapeutic areas. Primary Funding Source: National Institute of Allergy and Infectious Diseases. ==== Body pmcThe COVID-19 pandemic has caused more than 6 million deaths and half a billion cases globally (1). The standard of care (SOC) for patients hospitalized with COVID-19 has evolved rapidly during the pandemic and includes changes in oxygenation practices; airway management; use of prone positioning; anticoagulation practices; and use of antivirals, corticosteroids, and other immunomodulators (2–7). Some of these interventions were implemented after efficacy was determined by large clinical trials, whereas others were based on results of observational cohort studies or extrapolation from other disease states. These interventions have affected the morbidity and mortality of patients with COVID-19, but it is difficult to quantify their cumulative effect as the pandemic progresses. In the United States, the in-hospital mortality rate for patients with COVID-19 peaked in March and April 2020, then decreased from approximately 20% to 10% by June 2020 (8). Recovery rates also improved during this period (9, 10). In-hospital mortality is associated with patient characteristics and hospital factors (10, 11), but adjustment for these factors explains only some of the decrease in the mortality rate over time (10–13). Trends in COVID-19 SOC, including medication use and oxygen supplementation, may further explain the reduction in mortality rates (9). We analyzed clinical outcome data from 3 sequential cohorts of hospitalized patients in the first 3 stages of ACTT (Adaptive COVID-19 Treatment Trial), which was conducted in multiple countries, with 86% of the participants from the United States (Supplement Table 1) (14–17). The first 3 stages of ACTT each included a remdesivir group; it was the treatment group in ACTT-1 and the control group in ACTT-2 and ACTT-3 (Figure 1). Instead of comparing treatment groups within each stage, we analyzed the 3 remdesivir-only groups from these 3 stages. We compared recovery and mortality between remdesivir cohorts, using trial stage as a proxy for SOC given during that period. Propensity scores were used to balance cohorts on baseline characteristics. Figure 1. U.S. hospitalization rates (8) and treatment milestones (A) and time trends of concomitant medication use (B) in ACTT-1, ACTT-2, and ACTT-3. Treatment groups and periods of enrollment (dark colors) and follow-up after the last enrolled participant (light colors) are shown at the top of the figure. The colors of the bars correspond to the treatment groups. Treatment recommendations are from the National Institutes of Health COVID-19 Treatment Guidelines Panel (2). ACTT = Adaptive COVID-19 Treatment Trial; Bari = baricitinib; CQ = chloroquine; DEX = dexamethasone; EAP = Expanded Access Program; EUA = emergency use authorization; FDA = U.S. Food and Drug Administration; HCQ = hydroxychloroquine; RDV = remdesivir. ACTT-4 did not include a remdesivir monotherapy group. A secondary analysis applied the same methods to compare outcomes between participants in the remdesivir-plus-dexamethasone group in ACTT-4 versus remdesivir recipients in ACTT-3 who also received dexamethasone. Methods Data The ACTT trials were sequential, double-blind, randomized, placebo-controlled trials that evaluated novel investigational therapeutics for the treatment of adults hospitalized with COVID-19. Figure 1 shows enrollment and follow-up periods for ACTT-1, ACTT-2, and ACTT-3, whose consecutive remdesivir groups make up our primary analysis population. These stages enrolled approximately 500 people per group (Supplement Figure 1) and completed follow-up before COVID-19 vaccination began. ACTT-4 compared remdesivir plus baricitinib to remdesivir plus dexamethasone and was included in a secondary analysis (Supplement Figure 2). The ACTT protocol is available at Annals.org. Data collected on demographic characteristics, laboratory parameters, baseline disease severity, comorbidities, and concomitant medication use were similar for the 4 stages of ACTT, with additional data added as understanding of COVID-19 evolved and based on the interventions studied in each stage. For example, beginning with ACTT-2, additional data were collected on baseline risk factors, such as history of deep venous thrombosis, pulmonary embolism, and coagulopathy, and more specific data were collected on dexamethasone use for COVID-19 before enrollment. Clinical assessments were performed daily from day 1 through day 29 while patients were hospitalized and at follow-up visits on days 15, 22, and 29 for those who were discharged from the hospital. Disease severity was measured with an 8-category ordinal scale (Supplement Table 2) and the National Early Warning Score (NEWS) (Supplement Table 3) (18). Categories of ordinal scale for hospitalized patients are based on oxygen delivery method (Figure 2), and a person's value on the scale is their “ordinal score” (OS). Our outcomes are 28-day mortality and 28-day recovery, defined as the day of discharge or the first day of continued hospitalization without a requirement for supplemental oxygen or medical care. Figure 2. Oxygen delivery system/OS by day and phase (ACTT-1, ACTT-2, and ACTT-3) for enrolled remdesivir recipients. Panel A shows weekly enrollments by baseline oxygen delivery system, which corresponds to the OS of disease severity for hospitalized patients. Panel B shows proportions rather than counts to facilitate comparison of OS distributions between stages. Panels A and B show baseline distributions, whereas panel C includes all observations for each participant from enrollment until day 28 or discharge. ACTT = Adaptive COVID-19 Treatment Trial; ECMO = extracorporeal membrane oxygenation; NIPPV = noninvasive positive pressure ventilation; OS = ordinal score. Statistical Analysis Our primary analysis included participants assigned to receive remdesivir plus SOC in ACTT-1, ACTT-2, and ACTT-3 (Figure 1). We used Cox regression to estimate recovery and mortality rates, with trial stage as a categorical predictor variable representing SOC given during that stage. Because we analyzed only remdesivir group participants, the data are effectively observational, as the randomization does not relate to our scientific question. Therefore, we analyzed the “as-treated” population and used propensity score weighting to balance baseline characteristics between cohorts. We followed STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) reporting guidelines for observational cohort studies (19). Although ACTT-4 did not include a remdesivir monotherapy group, a secondary analysis compared outcomes between those in the remdesivir-plus-dexamethasone group in ACTT-4 with the 77% of remdesivir recipients in ACTT-3 who also received dexamethasone as part of SOC. The platform trial was designed to keep inclusion criteria relatively constant across stages, but exclusion criteria could be modified for study product safety considerations, including changes to laboratory thresholds and concomitant medication use. Modifications to stage-specific exclusion criteria were minor, and the proportions of patients who were excluded due to stage-specific exclusion criteria were 1.9% for ACTT-1, 1.6% for ACTT-2, 1.5% for ACTT-3, and 2.7% for ACTT-4 (Supplement Tables 4 and 5). Modifications to inclusion criteria were minor except for those relating to baseline OS (Supplement Table 5): ACTT-3 excluded patients with a baseline OS of 7 and also excluded those with a baseline OS of 6 after a mid-trial review by a data safety monitoring board (Figure 2). ACTT-4 included only participants with a baseline OS of 5 or 6. Causal inference with propensity scores requires all participants to have a nonzero probability of falling into any cohort (20). To ensure this, we omitted participants with a baseline OS of 7 when comparing ACTT-2 with ACTT-3 and included only those with an OS of 5 or 6 when comparing ACTT-3 with ACTT-4. We also excluded participants with chronic liver disease (an exclusion criterion for ACTT-3) from comparisons involving ACTT-3. Sensitivity analyses were planned to enforce identical laboratory exclusion criteria across stages (Supplement Table 5) and to include only sites participating in both stages for a given comparison. Covariates were selected for inclusion in the propensity score model on the basis of clinical judgment and are summarized in Table 1. As measures of baseline disease severity, we included the NEWS and a modified 4C Mortality Score, a validated score that ranges from 0 to 21 and is based on age, sex, number of comorbidities, respiratory rate, peripheral oxygen saturation, level of consciousness, urea level, and C-reactive protein (CRP) level (Supplement Tables 3 and 6) (21–30). Because urea level was not recorded systematically in ACTT, our modified scale omits the urea point contributions (+1 for urea level of 7 to 14 mmol/L and +3 for urea level >14 mmol/L), but we included estimated glomerular filtration rate (eGFR) as a separate variable in the propensity score model to account for differences in kidney function. Although most covariates in the propensity score model were missing few data (0% to 2% of participants [Supplement Table 7]), CRP level was missing for 10% of participants (21% in ACTT-1, 3% in ACTT-2, 3% in ACTT-3, and 2% in ACTT-4). Multiple imputation was performed for missing CRP values. The Supplement provides further details on the statistical analysis. Table 1. Baseline Characteristics of Analyzed Remdesivir Recipients in ACTT-1 and ACTT-2 Evolving clinical understanding and external epidemiologic data suggest that intubation practices became more conservative between ACTT-1 and ACTT-2 (31, 32), but precise intubation protocols were not documented for ACTT sites. To estimate a change in intubation practice between ACTT-1 and ACTT-2, we fit a logistic regression model with baseline intubation as the outcome and trial stage as the predictor. The analysis included only patients with a baseline OS of 6 or 7 because we believe that few participants with an OS of 4 or 5 in ACTT-2 would have been intubated at baseline if they had instead contracted COVID-19 during ACTT-1 (Supplement Figure 3). Propensity scores were used to adjust for baseline characteristics. We modeled baseline intubations rather than time to intubation to prevent confounding from temporal improvements in SOC causing fewer patients to progress to severe illness in ACTT-2 than in ACTT-1. A small amount of confounding could have been introduced by care received between hospitalization and enrollment, so a sensitivity analysis repeated the model for only those enrolled within 2 days of hospitalization. Institutional Review Board Approval The trial protocol was approved by the institutional review board at each site or by a centralized institutional review board as appropriate. Role of the Funding Source The ACTT protocols were designed and written by the ACTT investigators and the study sponsor (National Institute of Allergy and Infectious Diseases [NIAID]), with input from the manufacturers of remdesivir (Gilead), baricitinib (Eli Lilly), and interferon-β1a (EMD Serono). Principal investigators and site staff gathered the data, which were analyzed by statisticians at the statistical and data coordinating center (The Emmes Company) and NIAID. The funder (NIAID) participated in the writing of the manuscript and the decision to submit the manuscript for publication. Results Figure 1 shows time trends in use of concomitant medications by remdesivir recipients in ACTT. Enrollment and follow-up periods are distinguished visually because periods with no new enrollments include more participants later in their disease trajectory who tended to have less severe disease. Hydroxychloroquine was used by about 40% of participants each month during ACTT-1, but very little was used in ACTT-2 and ACTT-3. Use of antibiotics in general and azithromycin in particular decreased across the 3 stages. Steroids were not recommended initially because of concerns about worsening viral replication from prior studies of severe acute respiratory syndrome (33). Use of corticosteroids, which were allowed for indications other than COVID-19, was about 20% in ACTT-1 and ACTT-2 and surged to about 70% in ACTT-3. Anticoagulant use (both prophylactic and therapeutic) was high during all 3 stages. Use of antivirals other than remdesivir was generally under 10% and decreased over time. Table 1 shows unweighted and propensity score–weighted statistical summaries for remdesivir recipients analyzed in ACTT-1 and ACTT-2. Several variables, including age, race, OS distribution, modified 4C Mortality Score, and interval between symptom onset and enrollment, suggest that the ACTT-2 cohort was at lower risk for poor outcomes than the ACTT-1 cohort. ACTT-2 had substantially more Hispanic patients, whose risk profile may differ from that in non-Hispanic patients (34), than ACTT-1 (52% vs. 26%). A “Love plot” (Supplement Figure 4) (35) shows excellent balance after weighting. Figure 3 shows propensity score–weighted and unweighted survival curves with hazard ratio (HR) estimates and CIs. The unadjusted recovery rate for ACTT-2 participants was 1.23 times higher than for ACTT-1 participants (95% CI, 1.06 to 1.40). However, the adjusted HR was close to 1 (1.04 [CI, 0.92 to 1.17]), indicating that remdesivir recipients with comparable characteristics had similar recovery rates between ACTT-1 and ACTT-2. Similarly, although the unweighted 28-day mortality rate was estimated to be 0.69 times lower in ACTT-2 than ACTT-1 (CI, 0.45 to 1.04), the weighted HR was closer to the null value of 1 (0.90 [CI, 0.56 to 1.40]). Figure 3. Propensity score–weighted and unweighted survival curves for recovery and mortality in ACTT-1 and ACTT-2. Note that the y-axis scale differs between panels. ACTT = Adaptive COVID-19 Treatment Trial; HR = hazard ratio. Figure 2 summarizes oxygen delivery systems by day and trial stage for remdesivir participants. For those with a baseline OS of 6 or 7, the unadjusted odds of intubation in ACTT-2 were 0.31 times lower than in ACTT-1 (CI, 0.20 to 0.48). The adjusted odds ratio was 0.75 (CI, 0.53 to 0.97), suggesting a change in practice between stages. Sensitivity analysis results were similar (Supplement Figure 5 and Supplement Table 9). Table 2 shows unweighted and weighted statistical summaries for remdesivir recipients in ACTT-2 and ACTT-3. Although several variables (including race, ethnicity, OS, NEWS, and CRP level) suggest that the ACTT-3 population was at lower risk for poor outcomes, the ACTT-3 population was older (mean age, 59 vs. 55 years). The 2 cohorts were well balanced after weighting (Supplement Figure 6). Figure 4 shows survival curves with HR estimates. The adjusted HR for recovery indicates that the recovery rate was 1.43 times higher during the ACTT-3 period than the ACTT-2 period for people with comparable characteristics (CI, 1.24 to 1.64). The 28-day mortality rate in ACTT-3 was 0.45 times that in ACTT-2 for people with comparable characteristics (CI, 0.21 to 0.97). Figure 4. Propensity score–weighted and unweighted survival curves for recovery and mortality for the comparison between ACTT-2 and ACTT-3. This analysis excludes patients with chronic liver disease and a baseline ordinal score of 7, as these were exclusion criteria for ACTT-3. Note that the y-axis scale differs between panels. ACTT = Adaptive COVID-19 Treatment Trial; HR = hazard ratio. Table 2. Baseline Characteristics of Remdesivir Recipients in ACTT-2 and ACTT-3, From an Analysis That Excluded Patients With Chronic Liver Disease and a Baseline Ordinal Score of 7 (Exclusion Criteria for ACTT-3) Supplement Figure 7 and Supplement Table 10 show results from 3 preplanned sensitivity analyses that 1) applied the same laboratory-based eligibility criteria across study stages, 2) included only sites that enrolled participants in both stages being compared, and 3) analyzed only patients with complete baseline data. Results of these analyses are similar to the main results. Outcomes for recipients of remdesivir plus dexamethasone between ACTT-3 and ACTT-4 were similar (Supplement Figure 8). Recovery rates were nearly identical between stages for this group: The unweighted HR was 0.97 (CI, 0.84 to 1.11), and the weighted HR was 1.02 (CI, 0.89 to 1.19). The mortality rate was higher in ACTT-4 than in ACTT-3 (weighted HR, 1.75 [CI, 0.84 to 3.78]), but the CI was wide and these findings did not provide strong evidence of a difference in mortality rates between stages. Discussion This study compared 28-day recovery and mortality between comparable cohorts of hospitalized adults with COVID-19 who participated in 4 sequential stages of ACTT spanning February 2020 to May 2021. Although our unadjusted HR estimates describe differences in outcomes between trial stages, the propensity score–weighted HRs account for changes in patient composition over time and represent a “stage effect” attributed to the SOC received during different stages. Remdesivir recipients in ACTT-2 (spanning May to July 2020) recovered faster and had numerically better mortality outcomes than those in ACTT-1 (February to May 2020). Observed SOC changes included a dramatic decrease in hydroxychloroquine use between ACTT-1 and ACTT-2 and a gradual decrease in empirical antibiotic use. We also found that the odds of baseline intubation in ACTT-2 were 25% lower than for comparable ACTT-1 participants. However, we did not find evidence that these changes affected 28-day recovery or mortality: Our adjusted HR estimates were close to 1, indicating that the better outcomes in ACTT-2 were due to differences in patient composition rather than improved SOC. When comparing ACTT-3 (August to December 2020) with ACTT-2, the adjusted analyses showed improved 28-day recovery and mortality, suggesting improved SOC. Given that 77% of remdesivir recipients in ACTT-3 but only 11% in ACTT-2 received dexamethasone as SOC and the RECOVERY trial found a mortality benefit from dexamethasone, dexamethasone use is likely a key contributor to these improvements, although this observational analysis cannot confirm causality (36, 37). Another change in SOC between these stages was a gradual decrease in the use of antibiotics (including azithromycin). The large mortality reduction of 0.45 must be interpreted in the context of its wide CI (0.21 to 0.97), which is consistent with more moderate reductions. A detailed examination suggests low risk of bias from differences in stage-specific laboratory eligibility criteria (Supplement Table 11). We did not find evidence for improved outcomes between ACTT-3 and ACTT-4 (December 2020 to May 2021) among people receiving remdesivir plus dexamethasone. This may be because concomitant medication use stayed fairly constant between these periods. The use of combination immunomodulatory therapy (dexamethasone plus a Janus kinase inhibitor or an interleukin-6 inhibitor) in patients with the most severe disease did not become part of SOC until March 2021 with the COV-BARRIER and RECOVERY baricitinib and tocilizumab trials (6, 7). Hospitalization rates were higher during ACTT-4 (Supplement Figure 2), and COVID-19 surges can stress the hospital system, increasing mortality (38), which could explain the numerically worse mortality outcomes in this stage. The appearance of the B.1.1.7 (Alpha) variant during ACTT-4 could also explain part of this mortality difference (39) (Supplement Figure 9). ACTT-4 spanned the initial availability of vaccines, but only 25 (0.5%) analyzed ACTT-4 participants were vaccinated, so bias from vaccination is probably low. This study illustrates several issues related to the inclusion of nonconcurrent controls in analysis of data from platform trials (40–42). Even with similar eligibility criteria across stages, participant composition changed enough over time to substantially affect clinical outcomes. Early in a pandemic of a novel pathogen, patients may be more hesitant to present to a hospital, so those who do may tend to have more severe disease. Similarly, patients may be more hesitant to enroll in a trial of a novel therapeutic unless they are experiencing severe disease. We also found that SOC changed enough over time to substantially affect clinical outcomes for comparable patients. These analyses also highlight issues related to outcome definitions. Although time to intubation and time to intubation or death have been used as outcome measures in clinical trials (43, 44), we found that intubation practices became more conservative over time. This could make outcomes appear worse for nonconcurrent controls, thus exaggerating the treatment effect. When contemporaneous controls are used, the temporal shift in the relationship between the outcome and the underlying disease severity affects treated and control participants simultaneously and equally. Time to recovery is also not a purely objective outcome. ACTT participants were defined as having “recovered” when they were discharged from the hospital or if they remained hospitalized without requiring supplemental oxygen or ongoing medical care. This definition is a proxy for a certain point in the underlying, unobserved, actual disease trajectory. The relationship between this proxy and the actual disease trajectory can vary. As the pandemic progressed, clinicians may have discharged patients earlier in their recovery trajectory, which could artificially inflate a treatment effect in a comparison with nonconcurrent controls. It could also explain part of the improvement in recovery outcomes between ACTT-2 and ACTT-3. Mortality is potentially more objective, although this can depend on varying protocols for withdrawal of care. Furthermore, low death rates mean that very large trials are needed for adequate power to detect a mortality reduction. In addition, treatments that reduce objectively measured symptoms are beneficial even if they do not affect mortality, and objective outcomes are needed to test such treatments. This study has limitations. First, valid propensity score inference requires inclusion of all confounders in the analysis. Although we included many important confounders, some potential confounders were not measured. In particular, urea level was not recorded in ACTT, and this variable contributes either 1 or 3 points to the 21-point 4C Mortality Score. Our modified scale omitting urea level is therefore less predictive than the complete scale, although we included eGFR as a separate variable to account for differences in kidney function. D-dimer, interleukin-6, and interleukin-10 were not collected across ACTT stages. Another limitation is the restriction of mortality comparisons to a 28-day interval. It is possible that SOC changes between ACTT-2 and ACTT-3 delayed deaths past the 28-day time point without reducing the overall in-hospital mortality rate. However, more people in ACTT-2 were still intubated on day 28 than in ACTT-3 (4% vs. 1% weighted [Supplement Table 12]), making a reversal of the mortality HR after day 28 unlikely. Improved outcomes may have been influenced by unmeasured time-dependent factors, such as greater clinician bedside experience, although this may be more applicable to the comparison of ACTT-1 versus ACTT-2. COVID-19 surges can increase mortality, but hospitalization rates were higher in ACTT-3 than ACTT-2, which would tend to attenuate the estimated mortality improvement from SOC (average weekly rates were 8.5 and 6.3 per 100 000 persons in ACTT-3 and ACTT-2, respectively [Figure 1]). Circulating variants were generally unidentified during 2020 (Supplement Figure 9), and their effect on mortality was probably small. This trial mostly enrolled patients at academic research sites; results are generalizable to similar hospitals whose populations resemble the ACTT patient population. This study analyzed time trends of mortality and recovery for comparable cohorts of hospitalized COVID-19 clinical trial participants and described evolving SOC practices. We found that intubation practice became more conservative between the period from February to May 2020 and the period from May to July 2020 and that improvements in recovery and mortality between these intervals were explained by differences in cohort composition. This contrasts with other studies examining U.S. in-hospital mortality, which found that mortality improvements persisted after patient characteristics were accounted for (10–13). Our mortality rates were lower: The unadjusted rate decreased from 11% to 7%, whereas other studies found a decrease from approximately 20% to 10% (10–13). The difference could be because the hospitals participating in ACTT may have more resources and experience and/or a different learning curve than other hospitals, followed protocol-specified care requirements, or drew a different patient population. We found improved recovery and mortality outcomes between the period from May to July 2020 and the period from August to December 2020, likely due to increased dexamethasone use. We did not find evidence for improvements in recovery or mortality between the period from August to December 2020 and the period from December 2020 to May 2021, possibly because SOC did not change substantially (the use of combination immunomodulatory treatment was implemented later). These findings support the exclusion of nonconcurrent controls when analyzing data from platform trials, particularly for COVID-19 treatments and vaccines. Supplementary Material Click here for additional data file. Click here for additional data file. This article was published at Annals.org on 29 November 2022. * Drs. Tomashek and Tebas contributed equally to this work. ==== Refs References 1. Johns Hopkins University. Johns Hopkins University Center for Systems Science and Engineering COVID-19 Dashboard. Accessed at https://coronavirus.jhu.edu on 17 June 2022. 2. Kuriakose S , Singh K , Pau AK , et al; NIH COVID-19 Treatment Guidelines Panel. Developing treatment guidelines during a pandemic health crisis: lessons learned from COVID-19. Ann Intern Med. 2021;174 :1151-8. [PMID: ] doi:10.7326/M21-1647 34125574 3. Ghelichkhani P , Esmaeili M . Prone position in management of COVID-19 patients; a commentary. Arch Acad Emerg Med. 2020;8 :e48. [PMID: ]32309812 4. Lawler PR , Goligher EC , Berger JS , et al; ATTACC Investigators. Therapeutic anticoagulation with heparin in noncritically ill patients with Covid-19. N Engl J Med. 2021;385 :790-802. [PMID: ] doi:10.1056/NEJMoa2105911 34351721 5. Gordon AC , Mouncey PR , Al-Beidh F , et al; REMAP-CAP Investigators. Interleukin-6 receptor antagonists in critically ill patients with Covid-19. N Engl J Med. 2021;384 :1491-1502. [PMID: ] doi:10.1056/NEJMoa2100433 33631065 6. Marconi VC , Ramanan AV , de Bono S , et al; COV-BARRIER Study Group. Efficacy and safety of baricitinib for the treatment of hospitalised adults with COVID-19 (COV-BARRIER): a randomised, double-blind, parallel-group, placebo-controlled phase 3 trial. Lancet Respir Med. 2021;9 :1407-18. [PMID: ] doi:10.1016/S2213-2600(21)00331-3 34480861 7. RECOVERY Collaborative Group. Tocilizumab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial. Lancet. 2021;397 :1637-45. [PMID: ] doi:10.1016/S0140-6736(21)00676-0 33933206 8. Centers for Disease Control and Prevention. COVID Data Tracker. Accessed at https://covid.cdc.gov/covid-data-tracker/#hospitalizations-severity on 10 June 2022. 9. Garg S , Patel K , Pham H , et al. Clinical trends among U.S. adults hospitalized with COVID-19, March to December 2020: a cross-sectional study. Ann Intern Med. 2021;174 :1409-19. [PMID: ] doi:10.7326/M21-1991 34370517 10. Roth GA , Emmons-Bell S , Alger HM , et al. Trends in patient characteristics and COVID-19 in-hospital mortality in the United States during the COVID-19 pandemic. JAMA Netw Open. 2021;4 :e218828. [PMID: ] doi:10.1001/jamanetworkopen.2021.8828 33938933 11. Asch DA , Sheils NE , Islam MN , et al. Variation in US hospital mortality rates for patients admitted with COVID-19 during the first 6 months of the pandemic. JAMA Intern Med. 2021;181 :471-8. [PMID: ] doi:10.1001/jamainternmed.2020.8193 33351068 12. Horwitz LI , Jones SA , Cerfolio RJ , et al. Trends in COVID-19 risk-adjusted mortality rates. J Hosp Med. 2021;16 :90-92. [PMID: ] doi:10.12788/jhm.3552 33147129 13. Moon RC , Mackey RH , Cao Z , et al. Is coronavirus disease 2019 (COVID-19) less deadly now? Trends in in-hospital mortality among hospitalized COVID-19 patients in the United States. Clin Infect Dis. 2022;74 :2238-42. [PMID: ] doi:10.1093/cid/ciab830 34534276 14. Beigel JH , Tomashek KM , Dodd LE , et al; ACTT-1 Study Group Members. Remdesivir for the treatment of Covid-19 - final report. N Engl J Med. 2020;383 :1813-26. [PMID: ] doi:10.1056/NEJMoa2007764 32445440 15. Kalil AC , Mehta AK , Patterson TF , et al; ACTT-3 study group members. Efficacy of interferon beta-1a plus remdesivir compared with remdesivir alone in hospitalised adults with COVID-19: a double-bind, randomised, placebo-controlled, phase 3 trial. Lancet Respir Med. 2021;9 :1365-76. [PMID: ] doi:10.1016/S2213-2600(21)00384-2 34672949 16. Kalil AC , Patterson TF , Mehta AK , et al; ACTT-2 Study Group Members. Baricitinib plus remdesivir for hospitalized adults with Covid-19. N Engl J Med. 2021;384 :795-807. [PMID: ] doi:10.1056/NEJMoa2031994 33306283 17. Wolfe CR , Tomashek KM , Patterson TF , et al; ACTT-4 Study Group. Baricitinib versus dexamethasone for adults hospitalised with COVID-19 (ACTT-4): a randomised, double-blind, double placebo-controlled trial. Lancet Respir Med. 2022;10 :888-99. [PMID: ] doi:10.1016/S2213-2600(22)00088-1 35617986 18. Smith GB , Prytherch DR , Meredith P , et al. The ability of the National Early Warning Score (NEWS) to discriminate patients at risk of early cardiac arrest, unanticipated intensive care unit admission, and death. Resuscitation. 2013;84 :465-70. [PMID: ] doi:10.1016/j.resuscitation.2012.12.016 23295778 19. von Elm E , Altman DG , Egger M , et al; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147 :573-7. [PMID: ]17938396 20. Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983;70:41-55. doi:10.1093/biomet/70.1.41 21. Wynants L , Van Calster B , Collins GS , et al. Prediction models for diagnosis and prognosis of Covid-19: systematic review and critical appraisal. BMJ. 2020;369 :m1328. [PMID: ] doi:10.1136/bmj.m1328 32265220 22. Miller JL , Tada M , Goto M , et al. Prediction models for severe manifestations and mortality due to COVID-19: a systematic review. Acad Emerg Med. 2022;29 :206-16. [PMID: ] doi:10.1111/acem.14447 35064988 23. van Dam PMEL, Zelis N, van Kuijk SMJ, et al. Performance of prediction models for short-term outcome in COVID-19 patients in the emergency department: a retrospective study. Ann Med. 2021;53 :402-9. [PMID: ] doi:10.1080/07853890.2021.1891453 33629918 24. Knight SR , Ho A , Pius R , et al; ISARIC4C investigators. Risk stratification of patients admitted to hospital with Covid-19 using the ISARIC WHO Clinical Characterisation Protocol: development and validation of the 4C Mortality Score. BMJ. 2020;370 :m3339. [PMID: ] doi:10.1136/bmj.m3339 32907855 25. Yildiz H , Castanares-Zapatero D , Hannesse C , et al. Prospective validation and comparison of COVID-GRAM, NEWS2, 4C Mortality Score, CURB-65 for the prediction of critical illness in COVID-19 patients [Letter]. Infect Dis (Lond). 2021;53 :640-2. [PMID: ] doi:10.1080/23744235.2021.1896777 33691577 26. Jones A , Pitre T , Junek M , et al; COREG Investigators. External validation of the 4C Mortality Score among COVID-19 patients admitted to hospital in Ontario, Canada: a retrospective study. Sci Rep. 2021;11 :18638. [PMID: ] doi:10.1038/s41598-021-97332-1 34545103 27. Ocho K , Hagiya H , Hasegawa K , et al. Clinical utility of 4C mortality scores among Japanese COVID-19 patients: a multicenter study. J Clin Med. 2022;11 . [PMID: ] doi:10.3390/jcm11030821 35160272 28. Citu C , Gorun F , Motoc A , et al. Evaluation and comparison of the predictive value of 4C Mortality Score, NEWS, and CURB-65 in poor outcomes in COVID-19 patients: a retrospective study from a single center in Romania. Diagnostics (Basel). 2022;12 . [PMID: ] doi:10.3390/diagnostics12030703 35328256 29. Wirth A , Goetschi A , Held U , et al. External validation of the modified 4C Deterioration Model and 4C Mortality Score for COVID-19 patients in a Swiss tertiary hospital. Diagnostics (Basel). 2022;12 . [PMID: ] doi:10.3390/diagnostics12051129 35626285 30. Colombo CJ , Colombo RE , Maves RC , et al. Performance analysis of the National Early Warning Score and Modified Early Warning Score in the Adaptive COVID-19 Treatment Trial cohort. Crit Care Explor. 2021;3 :e0474. [PMID: ] doi:10.1097/CCE.0000000000000474 34278310 31. Sweeney DA , Malhotra A . Coronavirus disease 2019 respiratory failure: what is the best supportive care for patients who require ICU admission. Curr Opin Crit Care. 2021;27 :462-7. [PMID: ] doi:10.1097/MCC.0000000000000863 34310373 32. Doidge JC , Gould DW , Ferrando-Vivas P , et al. Trends in intensive care for patients with COVID-19 in England, Wales, and Northern Ireland. Am J Respir Crit Care Med. 2021;203 :565-74. [PMID: ] doi:10.1164/rccm.202008-3212OC 33306946 33. Hui DS . Systemic corticosteroid therapy may delay viral clearance in patients with Middle East respiratory syndrome coronavirus infection [Editorial]. Am J Respir Crit Care Med. 2018;197 :700-1. [PMID: ] doi:10.1164/rccm.201712-2371ED 29227752 34. Qeadan F , VanSant-Webb E , Tingey B , et al. Racial disparities in COVID-19 outcomes exist despite comparable Elixhauser comorbidity indices between Blacks, Hispanics, Native Americans, and Whites. Sci Rep. 2021;11 :8738. [PMID: ] doi:10.1038/s41598-021-88308-2 33888833 35. Ahmed A , Husain A , Love TE , et al. Heart failure, chronic diuretic use, and increase in mortality and hospitalization: an observational study using propensity score methods. Eur Heart J. 2006;27 :1431-9. [PMID: ]16709595 36. The RECOVERY Collaborative Group. Dexamethasone in hospitalized patients with Covid-19 — preliminary report. N Engl J Med. 2020. doi:10.1056/NEJMoa2021436 37. Horby P , Lim WS , Emberson JR , et al; RECOVERY Collaborative Group. Dexamethasone in hospitalized patients with Covid-19. N Engl J Med. 2021;384 :693-704. [PMID: ] doi:10.1056/NEJMoa2021436 32678530 38. Kadri SS , Sun J , Lawandi A , et al. Association between caseload surge and COVID-19 survival in 558 U.S. hospitals, March to August 2020. Ann Intern Med. 2021;174 :1240-51. [PMID: ] doi:10.7326/M21-1213 34224257 39. Davies NG , Jarvis CI , Edmunds WJ , et al; CMMID COVID-19 Working Group. Increased mortality in community-tested cases of SARS-CoV-2 lineage B.1.1.7. Nature. 2021;593 :270-4. [PMID: ] doi:10.1038/s41586-021-03426-1 33723411 40. Lee KM , Wason J . Including non-concurrent control patients in the analysis of platform trials: is it worth it. BMC Med Res Methodol. 2020;20 :165. [PMID: ] doi:10.1186/s12874-020-01043-6 32580702 41. Cohen DR , Todd S , Gregory WM , et al. Adding a treatment arm to an ongoing clinical trial: a review of methodology and practice. Trials. 2015;16 :179. [PMID: ] doi:10.1186/s13063-015-0697-y 25897686 42. Dodd LE , Freidlin B , Korn EL . Platform trials - beware the noncomparable control group [Letter]. N Engl J Med. 2021;384 :1572-3. [PMID: ] doi:10.1056/NEJMc2102446 33882210 43. WHO Working Group on the Clinical Characterisation and Management of COVID-19 infection. A minimal common outcome measure set for COVID-19 clinical research. Lancet Infect Dis. 2020;20 :e192-e197. [PMID: ] doi:10.1016/S1473-3099(20)30483-7 32539990 44. Bégin P , Callum J , Jamula E , et al; CONCOR-1 Study Group. Convalescent plasma for hospitalized patients with COVID-19: an open-label, randomized controlled trial. Nat Med. 2021;27 :2012-24. [PMID: ] doi:10.1038/s41591-021-01488-2 34504336	36442063	PMC9709721	NO-CC CODE	2022-12-03 23:19:51	no	Ann Intern Med. 2022 Nov 29;:M22-2116	utf-8	Ann Intern Med	2,022	10.7326/M22-2116	oa_other
==== Front Ann Intern Med Ann Intern Med aim Annals of Internal Medicine 0003-4819 1539-3704 American College of Physicians 36442068 10.7326/P22-0024 aim-olf-P220024 Summaries for Patients 3389Antiviral therapy3122457COVID-192882Drug therapy2237Outpatients2892Systematic reviewsearlyCurrently Online FirstcoronavirusCoronavirus Disease 2019 (COVID-19) Summaries for Patients are a service provided by Annals to help patients better understand the complicated and often mystifying language of modern medicine.Summaries for Patients are presented for informational purposes only. These summaries are not a substitute for advice from your own medical provider. If you have questions about this material, or need medical advice about your own health or situation, please contact your physician. The summaries may be reproduced for not-for-profit educational purposes only. Any other uses must be approved by the American College of Physicians.Summary for Patients: Outpatient Treatment of Confirmed COVID-19 From: Sommer I, Dobrescu A, Ledinger D, et al. Outpatient treatment of confirmed COVID-19. A living, rapid review for the American College of Physicians. Ann Intern Med. 29 November 2022. [Epub ahead of print]. doi:10.7326/M22-2202 29 11 2022 29 11 2022 P22-00242022 American College of Physicians This article is made available via the PMC Open Access Subset for unrestricted re-use for research, analyses, and text and data mining through PubMed Central. Acknowledgement of the original source shall include a notice similar to the following: "© 2020 American College of Physicians. Some rights reserved. This work permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited." 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. ==== Body pmcWhat is the problem and what is known about it so far? Various pharmacologic therapies, including antiviral drugs, corticosteroids, and other repurposed medications, have emerged as possible treatment options for outpatients with COVID-19. Why did the researchers do this particular study? The aim of this living, rapid review was to systematically collate and assess the evidence on the benefits and harms of COVID-19 treatments of interest to support the American College of Physicians in developing practice points on the use of COVID-19 treatments in adult outpatients. Who was studied? Across 26 studies, the number of participants in studies ranged from 18 to 5607. The median ages varied from 26 to 77 years. Trials were conducted in the United States, Canada, Argentina, Brazil, Colombia, Spain, Italy, the Middle East, or multiple countries. All studies were conducted before the Omicron variant became the dominant strain. In all studies, SARS-CoV-2 infection status was confirmed by a diagnostic test. How was the study done? The researchers searched the Epistemonikos COVID-19 L·OVE Platform up to 4 April 2022 and the COVID-NMA initiative website, with a surveillance search on 17 August 2022. Two trained reviewers independently screened studies against predefined eligibility criteria. One reviewer abstracted data, assessed risk of bias, and assessed overall strength of the evidence, with verification by a second reviewer. Statistical methods were used to combine information across studies. What did the researchers find? Three antiviral medications and 3 monoclonal antibodies were effective, although all studies were conducted before the emergence of the current Omicron variant. Nirmatrelvir–ritonavir probably reduced hospitalizations and death. Molnupiravir may reduce death. Remdesivir may improve recovery. Casirivimab–imdevimab reduced time to recovery and probably reduced hospitalizations. Regdanvimab probably improved recovery. Sotrovimab may reduce hospitalizations. Lopinavir–ritonavir and azithromycin may have increased harms, and hydroxychloroquine may result in lower recovery rates. Other treatments had insufficient evidence or no benefit compared with placebo. What were the limitations of the study? All studies were conducted before the emergence of the currently dominant Omicron variant. What are the implications of the study? Three antiviral medications may improve outcomes for outpatients with mild to moderate COVID-19, although the effect may be different with the current Omicron variant. There is no or insufficient evidence for a benefit with other antiparasitic, antibiotic, or immunosuppressive medications. The review will be updated as new information becomes available. This article was published at Annals.org on 29 November 2022.	36442068	PMC9709727	NO-CC CODE	2022-12-03 23:19:51	no	Ann Intern Med. 2022 Nov 29;:P22-0024	utf-8	Ann Intern Med	2,022	10.7326/P22-0024	oa_other
==== Front Ann Intern Med Ann Intern Med aim Annals of Internal Medicine 0003-4819 1539-3704 American College of Physicians 36442056 10.7326/M22-2202 aim-olf-M222202 Reviews 3389Antiviral therapy3122457COVID-192882Drug therapy2237Outpatients2892Systematic reviewsearlyCurrently Online FirstcoronavirusCoronavirus Disease 2019 (COVID-19)poc-eligiblePOC EligibleOutpatient Treatment of Confirmed COVID-19 A Living, Rapid Review for the American College of Physicians Living, Rapid Review on Outpatient Treatment of Confirmed COVID-19 Sommer Isolde MSc, MPH, PhD https://orcid.org/0000-0003-3592-1507 Dobrescu Andreea MD, PhD Ledinger Dominic MSc, MPH https://orcid.org/0000-0002-2031-5398 Moser Isabel MD Thaler Kylie MD, MPH Persad Emma MD https://orcid.org/0000-0003-2719-3685 Fangmeyer Martin MScN https://orcid.org/0000-0002-2514-9325 Emprechtinger Robert MSc https://orcid.org/0000-0003-3114-9812 Klerings Irma https://orcid.org/0000-0001-6644-9845 Gartlehner Gerald MD, MPH https://orcid.org/0000-0001-5531-3678 Cochrane Austria, Department for Evidence-based Medicine and Evaluation, University for Continuing Education Krems (Danube University Krems), Krems, Austria (I.S., A.D., D.L., I.M., K.T., E.P., M.F., I.K.) Faculty of Health and Medicine, University for Continuing Education Krems (Danube University Krems), Krems, Austria (R.E.) Cochrane Austria, Department for Evidence-based Medicine and Evaluation, University for Continuing Education Krems (Danube University Krems), Krems, Austria, and RTI International, Research Triangle Park, North Carolina (G.G.) Acknowledgment: The authors thank the COVID-NMA initiative for providing access to its data and allowing use of them. The authors also thank Manuela Müllner and Petra Wellemsen for project administration and Jennifer Yost from the ACP team for her support. Financial Support: By the ACP. Disclosures: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M22-2202. Reproducible Research Statement: Study protocol: Available from Dr. Sommer (e-mail, [email protected]). Statistical code and data set: Available from Mr. Emprechtinger (e-mail, [email protected]). Corresponding Author: Isolde Sommer, MSc, MPH, PhD, Cochrane Austria, Department for Evidence-based Medicine and Evaluation, University for Continuing Education Krems, Dr. Karl Dorrek Straße 30, 3500 Krems, Austria; e-mail, [email protected]. Author Contributions: Conception and design: I. Sommer, G. Gartlehner. Analysis and interpretation of the data: I. Sommer, A. Dobrescu, D. Ledinger, I. Moser, K. Thaler, E. Persad, M. Fangmeyer, R. Emprechtinger, G. Gartlehner. Drafting of the article: I. Sommer, A. Dobrescu, D. Ledinger, R. Emprechtinger. Critical revision for important intellectual content: I. Sommer, A. Dobrescu, D. Ledinger, K. Thaler, E. Persad, I. Klerings, G. Gartlehner. Final approval of the article: I. Sommer, A. Dobrescu, D. Ledinger, I. Moser, K. Thaler, E. Persad, M. Fangmeyer, R. Emprechtinger, I. Klerings, G. Gartlehner. Statistical expertise: R. Emprechtinger, G. Gartlehner. Obtaining of funding: G. Gartlehner. Administrative, technical, or logistic support: A. Dobrescu. Collection and assembly of data: I. Sommer, A. Dobrescu, D. Ledinger, I. Moser, K. Thaler, E. Persad, M. Fangmeyer, I. Klerings. 29 11 2022 29 11 2022 M22-22022022 American College of Physicians This article is made available via the PMC Open Access Subset for unrestricted re-use for research, analyses, and text and data mining through PubMed Central. Acknowledgement of the original source shall include a notice similar to the following: "© 2020 American College of Physicians. Some rights reserved. This work permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited." 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. Clinicians and patients want to know the benefits and harms of outpatient treatment options for SARS-CoV-2 infection. This living, rapid review assesses the benefits and harms of different COVID-19 treatments in the outpatient setting. Evidence for 4 antiviral medications, 3 monoclonal antibodies, 4 antibiotic or antiparasitic medications, convalescent plasma, corticosteroids, and fluvoxamine is summarized in this article. Background: Clinicians and patients want to know the benefits and harms of outpatient treatment options for SARS-CoV-2 infection. Purpose: To assess the benefits and harms of 12 different COVID-19 treatments in the outpatient setting. Data Sources: Epistemonikos COVID-19 L·OVE Platform, searched on 4 April 2022. Study Selection: Two reviewers independently screened abstracts and full texts against a priori–defined criteria. Randomized controlled trials (RCTs) that compared COVID-19 treatments in adult outpatients with confirmed SARS-CoV-2 infection were included. Data Extraction: One reviewer extracted data and assessed risk of bias and certainty of evidence (COE). A second reviewer verified data abstraction and assessments. Data Synthesis: The 26 included studies collected data before the emergence of the Omicron variant. Nirmatrelvir–ritonavir and casirivimab–imdevimab probably reduced hospitalizations (1% vs. 6% [1 RCT] and 1% vs. 4% [1 RCT], respectively; moderate COE). Nirmatrelvir–ritonavir probably reduced all-cause mortality (0% vs. 1% [1 RCT]; moderate COE), and regdanvimab probably improved recovery (87% vs. 72% [1 RCT]; moderate COE). Casirivimab–imdevimab reduced time to recovery by a median difference of 4 days (10 vs. 14 median days [1 RCT]; high COE). Molnupiravir may reduce all-cause mortality, sotrovimab may reduce hospitalization, and remdesivir may improve recovery (low COE). Lopinavir–ritonavir and azithromycin may have increased harms, and hydroxychloroquine may result in lower recovery rates (low COE). Other treatments had insufficient evidence or no statistical difference in efficacy and safety versus placebo. Limitation: Many outcomes had few events and small samples. Conclusion: Some antiviral medications and monoclonal antibodies may improve outcomes for outpatients with mild to moderate COVID-19. However, the generalizability of the findings to the currently dominant Omicron variant is limited. Primary Funding Source: American College of Physicians. (PROSPERO: CRD42022323440) ==== Body pmcIn the United States, COVID-19 has resulted in more than 1 million deaths (1) and led to a decrease in life expectancy of 1.87 years (2). Various pharmacologic therapies, including antiviral drugs, corticosteroids, and other repurposed medications, have emerged as treatment options for outpatients with COVID-19. Several reviews have systematically assessed the efficacy and safety of these therapies (3–10). However, given the pace of the pandemic and the emerging evidence, without regular updates these reviews quickly become outdated. In addition, most included both inpatient and outpatient management and focused only on 1 specific COVID-19 treatment. The aim of this living, rapid review was to systematically collate and assess the evidence regarding the benefits and harms of COVID-19 treatments of interest to support the American College of Physicians (ACP) Scientific Medical Policy Committee (SMPC) in developing practice points on the use of COVID-19 treatments in adult outpatients. Methods We conducted this living, rapid review in accordance with the Cochrane Rapid Reviews Methods Group guidance (11). We registered our protocol in PROSPERO (CRD42022323440) and made no changes to it. Throughout this review, we adhered to the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement (12). Our methods differed from those of a systematic review in the following ways: We searched only 1 electronic database (the Epistemonikos COVID-19 L·OVE Platform [13]), and single reviewers extracted data and rated risk of bias and certainty of evidence (COE); a second, senior investigator verified data abstraction and assessments. We plan to conduct monthly surveillance searches over a period of 1 year for new randomized controlled trials (RCTs). The study eligibility criteria might be revised if the treatments of interest change. The methodological approach will remain the same. The SMPC is planning to maintain this topic as living, rapid practice points with literature surveillance and periodic updating of the living, rapid review and SMPC practice points. Details of the practice points' living process, including signals for updating and retirement, can be found in ACP's methods articles (14). Research Questions and Eligibility Criteria We addressed the following key questions (KQs): KQ: What are the benefits and harms of COVID-19 treatments in symptomatic and asymptomatic adult patients with a confirmed SARS-CoV-2 infection in the outpatient setting? KQa: Do the benefits and harms vary by patient characteristics (age, gender, or comorbid conditions), type of SARS-CoV-2 variant, immunity status (prior SARS-CoV-2 infection, vaccination status, or time since infection or vaccination), symptom duration, or disease severity? We considered RCTs that included adult outpatients with a confirmed diagnosis of SARS-CoV-2 infection and were published in English. Treatments of interest included antiviral drugs, neutralizing monoclonal antibodies, antibiotic or antiparasitic drugs, convalescent plasma, corticosteroids, and fluvoxamine. Comparators were placebo to determine treatment efficacy or standard of care if no placebo-controlled trials were available, which was not the case for any of the treatments of interest. The ACP SMPC selected all-cause mortality, COVID-19–specific mortality, recovery, time to recovery, hospitalization due to COVID-19, and incidences of serious or any adverse events as critical outcomes for decision making. Supplement Table 1 presents the a priori–specified inclusion and exclusion criteria. Data Sources and Searches An experienced information specialist (I.K.) searched Epistemonikos COVID-19 L·OVE, a free-access repository and classification platform for COVID-19 evidence (13), up to 4 April 2022 (Supplement Table 2). In addition, we searched the COVID-NMA initiative website, a living evidence database of COVID-19 trials (15). On 17 August 2022, a surveillance search was conducted to identify studies to be included in periodic updating of the living, rapid review and SMPC practice points. Study Selection Two trained reviewers (from among I.S., A.D., D.L., I.M., E.P., K.T., and G.G.) independently screened titles, abstracts, and relevant full-text articles against predefined eligibility criteria using DistillerSR (Evidence Partners). Conflicts were resolved by discussion or by consulting a third reviewer. All results were tracked in an EndNote 20 database (Clarivate). Data Extraction and Quality Assessment One reviewer (I.S., A.D., D.L., I.M., E.P., or K.T.) abstracted characteristics of the study populations, settings, interventions, comparators, methods, and results from each included study. A second reviewer (I.S., A.D., D.L., I.M., E.P., or K.T.) checked all data abstractions for completeness and accuracy. A single investigator assessed the risk of bias of the included RCTs using the Cochrane Risk of Bias Tool 2.0 (16). We validated the ratings against the risk-of-bias assessments provided by COVID-NMA, which had applied the same tool (15). If the ratings differed, we involved a second investigator to resolve the discrepancy. For trials that were not included in the COVID-NMA database, we dually assessed the risk of bias. Supplement Figure 1 presents the risk-of-bias assessments. Data Synthesis and Analysis If we found 2 or more similar studies for a comparison of interest, we conducted meta-analyses. We chose the Bayesian random-effects model because it allows us to update the analyses without concern for P value inflation (17, 18). We conducted all analyses with R, version 4.1.3 (19), using the bayesmeta (20) and metafor (21) packages. We chose noninformative priors for both the treatment effect (mean, 0; SD, 4) and the heterogeneity (half-normal with a scale of 0.5). The results were calculated as risk ratios (RRs) and presented as forest plots. We determined the appropriateness of a meta-analysis by assessing the clinical and methodological heterogeneity following established guidance (22). Although we used an intention-to-treat-analysis for data we pooled in a meta-analysis, we relied on the data as reported in the individual studies for the narrative summary. When possible, we conducted sensitivity analysis to explore potential sources of heterogeneity. Although we had planned to perform subgroup analyses, we were unable to identify enough studies to do so. Certainty of Evidence We graded the COE on the basis of the guidance established by the GRADE (Grading of Recommendations Assessment, Development and Evaluation) Working Group (23). A single investigator assessed the COE for each key outcome, and a second senior investigator checked this for plausibility and consistency. Role of the Funding Source This living, rapid review was funded by ACP, which assisted in the development of the KQs and study inclusion criteria and selection of the outcomes of interest. The ACP was not involved in data collection, analysis, or manuscript preparation. Results The searches yielded 679 references, from which we included 26 RCTs (24–49). Figure 1 shows the study selection process. Supplement Tables 3 to 5 list eligible preprints, ongoing studies, and other excluded studies with the reasons for exclusion. Figure 1. PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) flowchart. Figure 1. PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) flowchart. Study and Participant Characteristics The number of participants in the included studies ranged from 18 to 5607. The median ages of participants varied from 26 to 77 years, and the proportion of females varied between 1% and 72%. Trials were conducted in the United States (31, 36, 40, 41, 43, 46), Canada (27, 45), Argentina (24, 47), Brazil (30, 32, 39, 49), Colombia (26), Spain (29, 37), Italy (48), the Middle East (28), or multiple countries (25, 33–35, 38, 42, 44). Out of 26 trials, 16 were funded with industry involvement (24, 25, 27, 31, 33–35, 37, 38, 41–46, 48). Among studies reporting vaccination status as an eligibility criterion, 11 studies (44%) (25, 28, 31, 33–37, 39, 44, 45) excluded vaccinated participants, and 4 studies (12%) included them (42, 43, 46, 49). Five studies excluded participants who had previously been diagnosed with COVID-19 (35, 37, 42, 43, 46); 1 study included them only if they had not been hospitalized or treated (44). All studies were conducted before the Omicron variant became the dominant strain. Participants were symptomatic across studies except in 1 study that included both symptomatic and asymptomatic participants (41). Ten studies (24, 26, 30, 32, 34, 37, 39, 40, 42, 43) provided data on disease severity; in 6 of them, participants had only mild symptoms (24, 26, 30, 32, 39, 40, 42). In all studies, the SARS-CoV-2 infection status was confirmed by a diagnostic test, usually a reverse transcriptase polymerase chain reaction test; 7 studies (26, 34, 37, 38, 41, 46, 49) also accepted antigen tests. Supplement Table 6 presents the characteristics and results of the included studies; Supplement Table 7 lists the definitions of “recovery” that were used in the included studies. We rated 9 studies as having low risk of bias (31, 33, 34, 37–39, 45–47), 16 as having some bias concerns (24–30, 32, 35, 36, 40–44, 49), and 1 as having high risk of bias (48). The risk-of-bias ratings of 8 studies differed from those in the COVID-NMA database (15) and required the involvement of a second reviewer. We dually assessed the risk of bias of 2 studies (42, 49) that were not available in the COVID-NMA database. Risk of bias included possible reporting bias, unclear blinding, lack of information on randomization and allocation concealment, or lack of an intention-to-treat analysis (Supplement Figure 1). Efficacy and Risk for Harms of COVID-19 Treatments Overall, only nirmatrelvir–ritonavir, casirivimab–imdevimab, and sotrovimab reduced hospitalizations due to COVID-19 compared with placebo (Figure 2). Lopinavir–ritonavir and azithromycin led to higher incidence of adverse events than placebo (Figure 3). Molnupiravir and nirmatrelvir–ritonavir reduced all-cause mortality (Supplement Figure 2). Nirmatrelvir–ritonavir, remdesivir, casirivimab–imdevimab, and sotrovimab reduced the incidence of serious adverse events (Supplement Figure 3). Figure 2. Summary plot of hospitalization due to COVID-19. The risk ratios were self-calculated. CI = confidence interval; COE = certainty of evidence; CrI = credible interval. Figure 2. Summary plot of hospitalization due to COVID-19. The risk ratios were self-calculated. CI = confidence interval; COE = certainty of evidence; CrI = credible interval. Figure 3. Summary plot of incidence of adverse events. The risk ratios were self-calculated. CI = confidence interval; COE = certainty of evidence; CrI = credible interval. Figure 3. Summary plot of incidence of adverse events. The risk ratios were self-calculated. CI = confidence interval; COE = certainty of evidence; CrI = credible interval. The Table summarizes results and COE ratings for each treatment versus placebo. Supplement Figures 4 to 19 display meta-analyses, and Supplement Table 8 presents summary-of-findings tables. Table. Results and COE Ratings for Each Treatment Versus Placebo Table. Results and COE Ratings for Each Treatment Versus Placebo Antiviral Drugs Lopinavir–Ritonavir One RCT (n = 471; some risk of bias) assessed 800 mg of lopinavir and 200 mg of ritonavir at the first 2 intakes, followed by 400 mg of lopinavir and 100 mg of ritonavir for the next 9 days, compared with placebo (30). Lopinavir–ritonavir may have no effect on hospitalization due to COVID-19 (5.6% vs. 4.8%; hazard ratio, 1.16 [95% confidence interval {CI}, 0.53 to 2.56]; low COE) but may increase the incidence of adverse events (39.7% vs. 20.9%; RR, 1.90 [CI, 1.40 to 2.57]; low COE). Although larger, the difference in serious adverse events between lopinavir–ritonavir and placebo was not statistically significant (8.6% vs. 5.5%; RR, 1.58 [CI, 0.79 to 3.16]; low COE). The evidence for all-cause mortality was insufficient to draw conclusions. Molnupiravir Two RCTs (n = 1637; low risk of bias) assessed molnupiravir, 800 mg (31, 33) or 200 to 800 mg (31), compared with placebo. The MOVe-OUT study (33) reported a reduction in all-cause mortality (which corresponded to COVID-19–related mortality as all deaths were due to COVID-19) (<0.1% vs. 1.3%; RR, 0.11 [CI, 0.01 to 0.86]; low COE) with molnupiravir and no effect on hospitalization due to COVID-19 (6.2% vs. 7.9%; RR, 0.79 [CI, 0.54 to 1.16]; low COE). Molnupiravir at doses of 200, 400, or 800 mg probably results in similar recovery (48.4% vs. 48.3%; odds ratio, 1.04 [CI, 0.84 to 1.29]; 1 RCT; moderate COE) (31, 33) and time to recovery (median, 5.5 to 9.0 vs. 8.5 days; 1 RCT; low COE) compared with placebo (31, 33). The proportion of participants affected by serious or any adverse events in the 2 studies did not differ statistically between groups (serious adverse events: 6.1% vs. 8.7%; RR, 0.77 [95% credible interval {CrI}, 0.32 to 2.03]; low COE [Supplement Figure 4]; any adverse events: 30.1% vs. 32.0%; RR, 0.96 [CrI, 0.55 to 1.73]; moderate COE [Supplement Figure 5]). Nirmatrelvir–Ritonavir One RCT (n = 2246; some risk of bias) assessed nirmatrelvir–ritonavir (300 and 100 mg) every 12 hours for 5 days compared with placebo (35). Nirmatrelvir–ritonavir probably reduced all-cause mortality (0% vs. 1.1%; RR, 0.04 [CI, 0.002 to 0.68]; moderate COE) and hospitalization due to COVID-19 for patients with 5 or fewer days of symptoms (0.7% vs. 6.2%; RR, 0.12 [CI, 0.06 to 0.26]; moderate COE). The incidence of any adverse events did not statistically differ compared with placebo (22.6% vs. 23.9%; RR, 0.95 [CI, 0.82 to 1.10]; high COE). Remdesivir One RCT (n = 584; some risk of bias) assessed remdesivir, 200 mg on day 1 and 100 mg on days 2 and 3, compared with placebo (44). Remdesivir may improve recovery between days 1 and 14 (36.1% vs. 20.0%; rate ratio, 1.92 [CI, 1.26 to 2.94]; low COE). There was no statistical difference in incidence of any adverse events (42.3% vs. 46.3%; RR, 0.91 [CI, 0.76 to 1.10]; moderate COE). Evidence was insufficient to draw conclusions about other outcomes. Monoclonal Neutralizing Antibodies We identified studies for 3 out of 5 monoclonal neutralizing antibodies approved by the U.S. Food and Drug Administration or the European Medicines Agency at the date of our search (4 April 2022). Casirivimab–Imdevimab One RCT (n = 4057; some risk of bias) assessed casirivimab–imdevimab, 1200 to 8000 mg, compared with placebo (25). Casirivimab–imdevimab reduced time to recovery (10 vs. 14 median days; high COE) and probably decreased hospitalizations due to COVID-19 (1.3% vs. 4.4%; RR, 0.30 [CI, 0.20 to 0.45]; moderate COE). Evidence was insufficient to draw conclusions about other outcomes. Regdanvimab Two RCTs (n = 345; 1 with low risk of bias and 1 with some risk of bias) assessed regdanvimab, 20 to 80 mg/kg of body weight, compared with placebo (34, 42). Although 1 study (n = 250) found that regdanvimab probably improved recovery (86.6% vs. 71.7%; RR, 1.21 [CI, 1.05 to 1.38]; moderate COE) (34), together the studies did not find a statistically significant effect on time to recovery (5.5 to 9.0 vs. 8.0 to 8.5 median days; low COE). The results for hospitalization due to COVID-19 (4.4% vs. 8.7%; RR, 0.51 [CI, 0.21 to 1.26]; 1 RCT; low COE) (34) and incidence of adverse events (29.4% vs. 30.7%; RR, 0.97 [CrI, 0.44 to 2.58]; 2 RCTs; low COE) (Supplement Figure 6) also did not differ statistically between groups. Evidence was insufficient to draw conclusions about any of the other outcomes. Sotrovimab One RCT (n = 1057; low risk of bias) assessed sotrovimab, 500 mg, compared with placebo (38). Sotrovimab may reduce hospitalization due to COVID-19 (1.1% vs. 5.7%; RR, 0.20 [CI, 0.08 to 0.48]; low COE) and resulted in no statistical difference in incidence of adverse events (21.8% vs. 23.4%; RR, 0.93 [CI, 0.74 to 1.17]; moderate COE). Evidence was insufficient to draw conclusions about other outcomes. Antibiotic or Antiparasitic Drugs Azithromycin One RCT (n = 263; some risk of bias) assessed azithromycin in a single 1.2-g dose compared with placebo (41). Azithromycin may have no effect on recovery at day 14 (50.4% vs. 50.0%; RR, 1.02 [CI, 0.91 to 1.13]; low COE) and may increase the incidence of adverse events (56.6% vs. 26.4%; RR, 2.14 [CI, 1.42 to 3.23]; low COE). Evidence was insufficient to draw conclusions about other outcomes. Chloroquine or Hydroxychloroquine Three RCTs (n = 893; some risk of bias) assessed hydroxychloroquine, 800 mg on day 1 followed by 400 mg/d for 5 days then 600 mg/d for 9 days, compared with placebo (27, 28, 30). Hydroxychloroquine may reduce the likelihood of recovery (60.9% vs. 78.4%; RR, 0.78 [CI, 0.62 to 0.97]; 1 RCT; low COE), but the median time to recovery (14 vs. 12 days; low COE) did not differ statistically between the treatment groups after 30 days (27). Hydroxychloroquine may not reduce risk for hospitalization due to COVID-19 (3.0% vs. 3.6%; RR, 0.90 [CrI, 0.37 to 2.21]; 3 RCTs; low COE) (Supplement Figure 7). Hydroxychloroquine may not result in any statistical difference in serious adverse events (2.2% vs. 2.9%; RR, 1.02 [CrI, 0.36 to 2.96]; 3 RCTs; low COE) (Supplement Figure 8) or any adverse events (22.2% vs. 20.9%; RR, 1.06 [CI, 0.74 to 1.53]; 1 RCT; low COE) (30). Evidence was insufficient to draw conclusions about other outcomes. Ivermectin Five RCTs (n = 2452; 4 with some risk of bias and 1 with high risk of bias) compared ivermectin, 200 to 1200 mcg/kg in a single dose or for 2 to 5 days, with placebo (26, 29, 47–49). Ivermectin may not have any statistically significant benefit on all-cause mortality (2.0% vs. 2.3%; RR, 0.89 [CrI, 0.42 to 1.91]; low COE) (Supplement Figure 9), recovery (68.2% vs. 65.6%; RR, 1.04 [CrI, 0.61 to 1.72]; moderate COE) (Supplement Figure 10), or hospitalization due to COVID-19 (8.1% vs. 9.9%; RR, 0.81 [CI, 0.49 to 1.34]; low COE) (Supplement Figure 11). A sensitivity analysis without the study that had high risk of bias found similar results for reduced hospitalization (8.1% vs. 10.2%; RR, 0.78 [CrI, 0.46 to 1.28]). There was no statistical difference in incidence of adverse events (27.7% vs. 31.8%; RR, 0.89 [CrI, 0.67 to 1.16]; moderate COE) (Supplement Figure 12). Evidence was insufficient to draw conclusions about other outcomes. Nitazoxanide Two RCTs (n = 1567; some risk of bias) assessed nitazoxanide, 1200 or 500 mg/d, compared with placebo (32, 43). Nitazoxanide resulted in no statistical difference in recovery (69.3% vs. 73.7%; RR, 0.94 [CI, 0.83 to 1.07]; moderate COE) (32), median number of days to sustained clinical recovery (13.3 [IQR, 6.3 to 21] vs. 12.4 [IQR, 7.2 to 21]; P = 0.88; moderate COE) (43), or hospitalization due to COVID-19 (0.7% vs. 1.1%; RR, 0.69 [CrI, 0.19 to 2.5]; low COE) (Supplement Figure 13). There were also no statistical differences in the incidence of serious adverse events (0.3% vs. 1.1%; RR, 0.33 [CrI, 0.07 to 1.56]; low COE) (Supplement Figure 14) or any adverse events (14.2% vs. 19.3%; RR, 0.79 [CrI, 0.38 to 1.62]; moderate COE) (Supplement Figure 15). Evidence was insufficient to draw conclusions about other outcomes. Convalescent Plasma Four RCTs (n = 2272; 2 with low risk of bias and 2 with some risk of bias) assessed convalescent plasma, 250 to 300 mL in a single dose, compared with placebo (24, 36, 37, 46). Convalescent plasma may have no statistical effect on all-cause mortality (0.6% vs. 0.9%; RR, 0.68 [CrI, 0.20 to 2.34]; 4 RCTs; low COE) (Supplement Figure 16), incidence of serious adverse events (1.1% vs. 1.1%; RR, 1.09 [CrI, 0.38 to 3.78]; 4 RCTs; low COE) (Supplement Figure 17), or time to symptom resolution (12 vs. 12 median days; hazard ratio, 1.05 [CI, 0.85 to 1.30]; 1 RCT; low COE) (46). Evidence was insufficient to draw conclusions about other outcomes. Other Drugs Corticosteroids One RCT (n = 215; low risk of bias) assessed ciclesonide, 1200 mcg inhaled twice daily or 200 mcg intranasally per day, compared with placebo (45). Ciclesonide may result in no statistically significant difference for recovery (65.7% vs. 58.2%; RR, 1.13 [CI, 0.91 to 1.40]; low COE), incidence of serious adverse events (6.6% vs. 4.9%; RR, 1.36 [CI, 0.45 to 4.15]; low COE), or incidence of any adverse events (21.9% vs. 15.3%; RR, 1.43 [CI, 0.79 to 2.58]; low COE). Evidence was insufficient to draw conclusions about other outcomes. Fluvoxamine Two trials (n = 1649; 1 with low risk of bias and 1 with some risk of bias) assessed fluvoxamine, 100 mg/d, compared with placebo (39, 40). Fluvoxamine may have no statistically significant effect on all-cause mortality (2.1% vs. 3.0%; RR, 0.71 [CrI, 0.24 to 2.10]; 2 RCTs; low COE) (Supplement Figure 18), hospitalization due to COVID-19 (9.1% vs. 12.2%; RR, 0.71 [CrI, 0.22 to 1.70]; 2 RCTs; low COE) (Supplement Figure 19), or any adverse events (15.0% vs. 15.3%; RR, 0.98 [CI, 0.46 to 2.09]; 1 RCT; low COE). Evidence was insufficient to draw conclusions about other outcomes. Subgroup Analysis One fluvoxamine trial (39) found no statistically significant interaction for the effect of age, sex, time from symptom onset, and comorbidities for hospitalization or extended emergency department visit due to COVID-19. Several other trials reported comparisons of the study groups in population subsets but without testing for interaction. Most confirmed the overall result (30, 32, 35, 41, 43, 44, 46). Two studies reported an increased or decreased risk for hospitalization due to COVID-19 or recovery for certain subgroups despite the overall effect showing no difference between the groups (33, 34) (Supplement Table 9). Surveillance The first surveillance search yielded 6 eligible RCTs (50–55). The studies compared molnupiravir (51), ivermectin (50–52, 54), fluvoxamine (55), and the monoclonal neutralizing antibodies tixagevimab–cilgavimab (53) with placebo (Supplement Table 10). The study on tixagevimab–cilgavimab (53) reported a reduction in COVID-19–related deaths or progression to severe disease (4% vs. 10%; RR, 0.43 [CI, 0.25 to 0.75]) and an increase in any adverse events (29% vs. 36%; RR, 0.81 [CI, 0.67 to 0.98]). It was conducted before the emergence of the Omicron variant. The remaining 5 studies reported no beneficial or harmful effects for outcomes of interest (50–52, 54, 55). Discussion This living, rapid review on 12 COVID-19 outpatient treatments, which included 26 RCTs conducted before dominance of the current Omicron variant, found that nirmatrelvir–ritonavir and the monoclonal antibodies casirivimab–imdevimab and regdanvimab had the strongest evidence for benefit in outpatients with COVID-19, with reduced hospitalizations, reduced all-cause mortality, or both. Molnupiravir may also reduce all-cause mortality and remdesivir may improve recovery, but evidence is less certain. However, these findings must be interpreted with caution because all studies were conducted before the dominance of the current Omicron variant. Specifically, a preprint article of the unblinded PANORAMIC (Platform Adaptive trial of NOvel antiviRals for eArly treatMent of covid-19 In the Community) trial (n = 25 783), which was conducted in the United Kingdom during dominance of the Omicron variant, reports no difference for hospitalization, mortality, or serious adverse events but improved early sustained recovery and time to first reported recovery between molnupiravir plus usual care and usual care (56). However, as a preprint article that has not yet been subject to peer review, this study did not meet the inclusion criteria for our surveillance. Several in vitro studies have found that the monoclonal antibodies that were found to be effective in our review (casirivimab–imdevimab, regdanvimab, and sotrovimab) are ineffective against the Omicron subvariant BA.5 (57–59). Because Omicron and its subvariants have become the dominant strains in the United States during 2022 (60), the U.S. Food and Drug Administration has revoked authorization for casirivimab–imdevimab (in January 2022) and sotrovimab (in May 2022) (61, 62). Regdanvimab was never approved in the United States. The antivirals remdesivir, molnupiravir, and nirmatrelvir–ritonavir have been shown to retain susceptibility to Omicron subvariants, including BA.5, similar to that for the ancestral strain (63). Despite retaining neutralizing activities, the absolute effect of antivirals to prevent hospitalization and death might be lower due to the reduced overall severity of the Omicron variant compared with previous variants (64). Current evidence does not support the efficacy of convalescent plasma and several drugs that were repurposed for use in outpatients with COVID-19, such as ivermectin, lopinavir–ritonavir, azithromycin, chloroquine or hydroxychloroquine, nitazoxanide, inhaled or intranasal corticosteroids, and fluvoxamine. Lopinavir–ritonavir and azithromycin may even have harmful effects, and hydroxychloroquine may lead to lower recovery rates. We did not identify any results related to COVID-19 rebound, a phenomenon in which patients develop symptoms of COVID-19 after taking the drug (65). However, because rebound has also been observed in untreated persons with COVID-19 (66), clinical trials are needed to understand the effects of antivirals on rebound. Our results are largely consistent with findings from other reviews, which were conducted in mixed populations of inpatients and outpatients and used standard of care as a comparison in addition to placebo. In line with our review, Cochrane reviews found beneficial effects for monoclonal antibodies (3) and nirmatrelvir–ritonavir (10) and no beneficial effects for chloroquine or hydroxychloroquine (5), convalescent plasma (9), ivermectin (6), or azithromycin (4). Other reviews showed that remdesivir increased recovery and reduced time to recovery and serious adverse events but also increased adverse events in hospitalized patients (67), and that fluvoxamine led to fewer hospitalizations in outpatients (68) when, unlike in this review, unpublished data were included. This living, rapid review considered many aspects not evaluated in previous reviews. One of its strengths is its comprehensive assessment of the benefits and harms of 12 COVID-19 treatments of interest. Another strength of this study is its focus on placebo-controlled trials, which is the most rigorous study design for evaluating treatment efficacy because it ensures assay sensitivity (the ability to distinguish between effective and ineffective treatments) (69). This review also has limitations. Although we restricted the literature search to only 1 database, evaluations of the Epistemonikos COVID-19 L·OVE Platform database have shown that it provides a comprehensive compilation of COVID-19 treatments, containing nearly all cited studies (70). To prevent missing relevant studies, we double-checked our list of included studies with that of the COVID-NMA database (15). Another limitation of our review is the lack of sufficient data for some outcomes. Included studies provided very low rates of hospitalization and mortality and low power in a population with mild to moderate disease severity. Insufficient data also precluded the exploration of heterogeneity across studies (71). The reported subgroup analyses were predominantly limited to exploratory or post hoc analyses and relied on small sample sizes. Although these analyses are useful for generating new hypotheses, recommendations for clinical practice should rely on prespecified subgroup analyses (72). Finally, the greatest limitation is that included studies were conducted before the Omicron variant became dominant and lacked information on vaccination or prior infection status, which reduces the generalizability of the findings. In conclusion, some antivirals and some monoclonal antibodies may improve recovery and reduce the risk for hospitalization in outpatients with mild to moderate COVID-19 from previous variants of SARS-CoV-2. However, the benefits of these therapies, particularly monoclonal antibodies, may be limited against the currently dominant Omicron variants. Supplementary Material Click here for additional data file. This article was published at Annals.org on 29 November 2022. ==== Refs References 1. Mathieu E, Ritchie H, Rodés-Guirao L, et al. Coronavirus (COVID-19) Deaths. Our World In Data. Accessed at https://ourworldindata.org/covid-deaths?country=[APPROX]USA on 18 May 2022. 2. Woolf SH , Masters RK , Aron LY . Changes in life expectancy between 2019 and 2020 in the US and 21 peer countries. JAMA Netw Open. 2022;5 :e227067. [PMID: ] doi:10.1001/jamanetworkopen.2022.7067 35416991 3. Kreuzberger N , Hirsch C , Chai KL , et al. SARS-CoV-2-neutralising monoclonal antibodies for treatment of COVID-19. Cochrane Database Syst Rev. 2021;9 :CD013825. [PMID: ] doi:10.1002/14651858.CD013825.pub2 34473343 4. Popp M , Stegemann M , Riemer M , et al. Antibiotics for the treatment of COVID-19. Cochrane Database Syst Rev. 2021;10 :CD015025. [PMID: ] doi:10.1002/14651858.CD015025 34679203 5. Singh B , Ryan H , Kredo T , et al. Chloroquine or hydroxychloroquine for prevention and treatment of COVID-19. Cochrane Database Syst Rev. 2021;2 :CD013587. [PMID: ] doi:10.1002/14651858.CD013587.pub2 33624299 6. Popp M , Reis S , Schießer S , et al. Ivermectin for preventing and treating COVID-19. Cochrane Database Syst Rev. 2022;6 :CD015017. [PMID: ] doi:10.1002/14651858.CD015017.pub3 35726131 7. Griesel M , Wagner C , Mikolajewska A , et al. Inhaled corticosteroids for the treatment of COVID-19. Cochrane Database Syst Rev. 2022;3 :CD015125. [PMID: ] doi:10.1002/14651858.CD015125 35262185 8. Wagner C , Griesel M , Mikolajewska A , et al. Systemic corticosteroids for the treatment of COVID-19. Cochrane Database Syst Rev. 2021;8 :CD014963. [PMID: ] doi:10.1002/14651858.CD014963 34396514 9. Piechotta V , Iannizzi C , Chai KL , et al. Convalescent plasma or hyperimmune immunoglobulin for people with COVID-19: a living systematic review. Cochrane Database Syst Rev. 2021;5 :CD013600. [PMID: ] doi:10.1002/14651858.CD013600.pub4 34013969 10. Reis S , Metzendorf MI , Kuehn R , et al. Nirmatrelvir combined with ritonavir for preventing and treating COVID-19. Cochrane Database Syst Rev. 2022;9 :CD015395. [PMID: ] doi:10.1002/14651858.CD015395.pub2 36126225 11. Garritty C , Gartlehner G , Nussbaumer-Streit B , et al. Cochrane Rapid Reviews Methods Group offers evidence-informed guidance to conduct rapid reviews. J Clin Epidemiol. 2021;130 :13-22. [PMID: ] doi:10.1016/j.jclinepi.2020.10.007 33068715 12. Page MJ , Moher D , Bossuyt PM , et al. PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic reviews. BMJ. 2021;372 :n160. [PMID: ] doi:10.1136/bmj.n160 33781993 13. Epistemonikos. L·OVE Platform. Accessed at https://app.iloveevidence.com/topics on 21 September 2022. 14. Qaseem A , Yost J , Forciea MA , et al; Scientific Medical Policy Committee of the American College of Physicians. The development of living, rapid practice points: summary of methods from the Scientific Medical Policy Committee of the American College of Physicians. Ann Intern Med. 2021;174 :1126-32. [PMID: ] doi:10.7326/M20-7641 34029483 15. Boutron I , Chaimani A , Meerpohl JJ , et al; COVID-NMA Consortium. The COVID-NMA project: building an evidence ecosystem for the COVID-19 pandemic [Editorial]. Ann Intern Med. 2020;173 :1015-7. [PMID: ] doi:10.7326/M20-5261 32931326 16. Sterne JAC , Savovic J , Page MJ , et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366 :l4898. [PMID: ] doi:10.1136/bmj.l4898 31462531 17. Kruschke JK , Liddell TM . The Bayesian New Statistics: Hypothesis testing, estimation, meta-analysis, and power analysis from a Bayesian perspective. Psychon Bull Rev. 2018;25 :178-206. [PMID: ] doi:10.3758/s13423-016-1221-4 28176294 18. Gronau QF, Heck DW, Berkhout SW, et al. A primer on Bayesian model-averaged meta-analysis. Adv Methods Pract Psychol Sci. 2021;4:1-19. doi:10.1177/25152459211031256 19. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing; 2022. 20. Röver C. Bayesian random-effects meta-analysis using the bayesmeta R package. J Stat Softw. 2020;93:1-51. doi:10.18637/jss.v093.i06 21. Viechtbauer W. Conducting meta-analyses in R with the metafor package. J Stat Softw. 2010;36:1-48. doi:10.18637/jss.v036.i03 22. West SL , Gartlehner G , Mansfield AJ , et al. Comparative Effectiveness Review Methods: Clinical Heterogeneity. Report no. 10-EHC070-EF. Agency for Healthcare Research and Quality. 2010. [PMID: ]21433337 23. Balshem H , Helfand M , Schünemann HJ , et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol. 2011;64 :401-6. [PMID: ] doi:10.1016/j.jclinepi.2010.07.015 21208779 24. Libster R , Pérez Marc G , Wappner D , et al; Fundación INFANT–COVID-19 Group. Early high-titer plasma therapy to prevent severe Covid-19 in older adults. N Engl J Med. 2021;384 :610-8. [PMID: ] doi:10.1056/NEJMoa2033700 33406353 25. Weinreich DM , Sivapalasingam S , Norton T , et al; Trial Investigators. REGEN-COV antibody combination and outcomes in outpatients with Covid-19. N Engl J Med. 2021;385 :e81. [PMID: ] doi:10.1056/NEJMoa2108163 34587383 26. López-Medina E , López P , Hurtado IC , et al. Effect of ivermectin on time to resolution of symptoms among adults with mild COVID-19: a randomized clinical trial. JAMA. 2021;325 :1426-35. [PMID: ] doi:10.1001/jama.2021.3071 33662102 27. Schwartz I , Boesen ME , Cerchiaro G , et al; ALBERTA HOPE COVID-19 Collaborators. Assessing the efficacy and safety of hydroxychloroquine as outpatient treatment of COVID-19: a randomized controlled trial. CMAJ Open. 2021;9 :E693-E702. [PMID: ] doi:10.9778/cmajo.20210069 34145052 28. Omrani AS , Pathan SA , Thomas SA , et al. Randomized double-blinded placebo-controlled trial of hydroxychloroquine with or without azithromycin for virologic cure of non-severe Covid-19. EClinicalMedicine. 2020;29 :100645. [PMID: ] doi:10.1016/j.eclinm.2020.100645 33251500 29. Chaccour C , Casellas A , Blanco-Di Matteo A , et al. The effect of early treatment with ivermectin on viral load, symptoms and humoral response in patients with non-severe COVID-19: a pilot, double-blind, placebo-controlled, randomized clinical trial. EClinicalMedicine. 2021;32 :100720. [PMID: ] doi:10.1016/j.eclinm.2020.100720 33495752 30. Reis G , Moreira Silva EADS , Medeiros Silva DC , et al; TOGETHER Investigators. Effect of early treatment with hydroxychloroquine or lopinavir and ritonavir on risk of hospitalization among patients with COVID-19: the TOGETHER randomized clinical trial. JAMA Netw Open. 2021;4 :e216468. [PMID: ] doi:10.1001/jamanetworkopen.2021.6468 33885775 31. Fischer WA 2nd, Eron JJ Jr, Holman W, et al.. A phase 2a clinical trial of molnupiravir in patients with COVID-19 shows accelerated SARS-CoV-2 RNA clearance and elimination of infectious virus. Sci Transl Med. 2022;14 :eabl7430. [PMID: ] doi:10.1126/scitranslmed.abl7430 34941423 32. Rocco PRM , Silva PL , Cruz FF , et al; SARITA-2 investigators. Early use of nitazoxanide in mild COVID-19 disease: randomised, placebo-controlled trial. Eur Respir J. 2021;58 . [PMID: ] doi:10.1183/13993003.03725-2020 33361100 33. Jayk Bernal A , Gomes da Silva MM , Musungaie DB , et al; MOVe-OUT Study Group. Molnupiravir for oral treatment of Covid-19 in nonhospitalized patients. N Engl J Med. 2022;386 :509-20. [PMID: ] doi:10.1056/NEJMoa2116044 34914868 34. Streinu-Cercel A , Sandulescu O , Preotescu LL , et al. Efficacy and safety of regdanvimab (CT-P59): a phase 2/3 randomized, double-blind, placebo-controlled trial in outpatients with mild-to-moderate coronavirus disease 2019. Open Forum Infect Dis. 2022;9 :ofac053. [PMID: ] doi:10.1093/ofid/ofac053 35295819 35. Hammond J , Leister-Tebbe H , Gardner A , et al; EPIC-HR Investigators. Oral nirmatrelvir for high-risk, nonhospitalized adults with Covid-19. N Engl J Med. 2022;386 :1397-1408. [PMID: ] doi:10.1056/NEJMoa2118542 35172054 36. Korley FK , Durkalski-Mauldin V , Yeatts SD , et al; SIREN-C3PO Investigators. Early convalescent plasma for high-risk outpatients with Covid-19. N Engl J Med. 2021;385 :1951-60. [PMID: ] doi:10.1056/NEJMoa2103784 34407339 37. Alemany A , Millat-Martinez P , Corbacho-Monné M , et al; CONV-ERT Group. High-titre methylene blue-treated convalescent plasma as an early treatment for outpatients with COVID-19: a randomised, placebo-controlled trial. Lancet Respir Med. 2022;10 :278-88. [PMID: ] doi:10.1016/S2213-2600(21)00545-2 35150610 38. Gupta A , Gonzalez-Rojas Y , Juarez E , et al; COMET-ICE Investigators. Effect of sotrovimab on hospitalization or death among high-risk patients with mild to moderate COVID-19: a randomized clinical trial. JAMA. 2022;327 :1236-46. [PMID: ] doi:10.1001/jama.2022.2832 35285853 39. Reis G , Dos Santos Moreira-Silva EA , Silva DCM , et al; TOGETHER investigators. Effect of early treatment with fluvoxamine on risk of emergency care and hospitalisation among patients with COVID-19: the TOGETHER randomised, platform clinical trial. Lancet Glob Health. 2022;10 :e42-e51. [PMID: ] doi:10.1016/S2214-109X(21)00448-4 34717820 40. Lenze EJ , Mattar C , Zorumski CF , et al. Fluvoxamine vs placebo and clinical deterioration in outpatients with symptomatic COVID-19: a randomized clinical trial. JAMA. 2020;324 :2292-2300. [PMID: ] doi:10.1001/jama.2020.22760 33180097 41. Oldenburg CE , Pinsky BA , Brogdon J , et al. Effect of oral azithromycin vs placebo on COVID-19 symptoms in outpatients with SARS-CoV-2 infection: a randomized clinical trial. JAMA. 2021;326 :490-8. [PMID: ] doi:10.1001/jama.2021.11517 34269813 42. Kim JY , Jang YR , Hong JH , et al. Safety, virologic efficacy, and pharmacokinetics of CT-P59, a neutralizing monoclonal antibody against SARS-CoV-2 spike receptor-binding protein: two randomized, placebo-controlled, phase I studies in healthy individuals and patients with mild SARS-CoV-2 infection. Clin Ther. 2021;43 :1706-27. [PMID: ] doi:10.1016/j.clinthera.2021.08.009 34551869 43. Rossignol JF , Bardin MC , Fulgencio J , et al. A randomized double-blind placebo-controlled clinical trial of nitazoxanide for treatment of mild or moderate COVID-19. EClinicalMedicine. 2022;45 :101310. [PMID: ] doi:10.1016/j.eclinm.2022.101310 35237748 44. Gottlieb RL , Vaca CE , Paredes R , et al; GS-US-540-9012 (PINETREE) Investigators. Early remdesivir to prevent progression to severe Covid-19 in outpatients. N Engl J Med. 2022;386 :305-15. [PMID: ] doi:10.1056/NEJMoa2116846 34937145 45. Ezer N , Belga S , Daneman N , et al. Inhaled and intranasal ciclesonide for the treatment of Covid-19 in adult outpatients: CONTAIN phase II randomised controlled trial. BMJ. 2021;375 :e068060. [PMID: ] doi:10.1136/bmj-2021-068060 34728476 46. Sullivan DJ , Gebo KA , Shoham S , et al. Early outpatient treatment for Covid-19 with convalescent plasma. N Engl J Med. 2022;386 :1700-11. [PMID: ] doi:10.1056/NEJMoa2119657 35353960 47. Vallejos J , Zoni R , Bangher M , et al. Ivermectin to prevent hospitalizations in patients with COVID-19 (IVERCOR-COVID19) a randomized, double-blind, placebo-controlled trial. BMC Infect Dis. 2021;21 :635. [PMID: ] doi:10.1186/s12879-021-06348-5 34215210 48. Buonfrate D , Chesini F , Martini D , et al. High-dose ivermectin for early treatment of COVID-19 (COVER study): a randomised, double-blind, multicentre, phase II, dose-finding, proof-of-concept clinical trial. Int J Antimicrob Agents. 2022;59 :106516. [PMID: ] doi:10.1016/j.ijantimicag.2021.106516 34999239 49. Reis G , Silva EASM , Silva DCM , et al; TOGETHER Investigators. Effect of early treatment with ivermectin among patients with Covid-19. N Engl J Med. 2022;386 :1721-31. [PMID: ] doi:10.1056/NEJMoa2115869 35353979 50. Biber A , Harmelin G , Lev D , et al. The effect of ivermectin on the viral load and culture viability in early treatment of nonhospitalized patients with mild COVID-19 - a double-blind, randomized placebo-controlled trial. Int J Infect Dis. 2022;122 :733-40. [PMID: ] doi:10.1016/j.ijid.2022.07.003 35811080 51. Caraco Y, Crofoot GE, Moncada PA, et al. Phase 2/3 trial of molnupiravir for treatment of Covid-19 in nonhospitalized adults. NEJM Evidence. 2022;1. doi:10.1056/EVIDoa2100043 52. Mirahmadizadeh A , Semati A , Heiran A , et al. Efficacy of single-dose and double-dose ivermectin early treatment in preventing progression to hospitalization in mild COVID-19: a multi-arm, parallel-group randomized, double-blind, placebo-controlled trial. Respirology. 2022;27 :758-66. [PMID: ] doi:10.1111/resp.14318 35738778 53. Montgomery H , Hobbs FDR , Padilla F , et al; TACKLE study group. Efficacy and safety of intramuscular administration of tixagevimab-cilgavimab for early outpatient treatment of COVID-19 (TACKLE): a phase 3, randomised, double-blind, placebo-controlled trial. Lancet Respir Med. 2022;10 :985-96. [PMID: ] doi:10.1016/S2213-2600(22)00180-1 35688164 54. Rezai MS , Ahangarkani F , Hill A , et al. Non-effectiveness of ivermectin on inpatients and outpatients with COVID-19; results of two randomized, double-blinded, placebo-controlled clinical trials. Front Med (Lausanne). 2022;9 :919708. [PMID: ] doi:10.3389/fmed.2022.919708 35783616 55. Seo H , Kim H , Bae S , et al. Fluvoxamine treatment of patients with symptomatic COVID-19 in a community treatment center: a preliminary result of randomized controlled trial. Infect Chemother. 2022;54 :102-13. [PMID: ] doi:10.3947/ic.2021.0142 35384422 56. Butler CC, Hobbs FDR, Gbinigie OA, et al. Molnupiravir plus usual care versus usual care alone as early treatment for adults with COVID-19 at increased risk of adverse outcomes (PANORAMIC): preliminary analysis from the United Kingdom randomised, controlled open-label, platform adaptive trial. SSRN. Preprint posted online 17 October 2022. doi:10.2139/ssrn.4237902 57. Arora P , Kempf A , Nehlmeier I , et al. Augmented neutralisation resistance of emerging Omicron subvariants BA.2.12.1, BA.4, and BA.5 [Letter]. Lancet Infect Dis. 2022;22 :1117-8. [PMID: ] doi:10.1016/S1473-3099(22)00422-4 35777385 58. Wang Q , Guo Y , Iketani S , et al. Antibody evasion by SARS-CoV-2 Omicron subvariants BA.2.12.1, BA.4 and BA.5. Nature. 2022;608 :603-8. [PMID: ] doi:10.1038/s41586-022-05053-w 35790190 59. Tuekprakhon A , Nutalai R , Dijokaite-Guraliuc A , et al; OPTIC Consortium. Antibody escape of SARS-CoV-2 Omicron BA.4 and BA.5 from vaccine and BA.1 serum. Cell. 2022;185 :2422-2433.e13. [PMID: ] doi:10.1016/j.cell.2022.06.005 35772405 60. Global Change Data Lab. Our World in Data. Accessed at https://ourworldindata.org/grapher/covid-cases-omicron?tab=chart&country=[APPROX]USA on 13 July 2022. 61. Cavazzoni P. Coronavirus (COVID-19) Update: FDA Limits Use of Certain Monoclonal Antibodies to Treat COVID-19 Due to the Omicron Variant. U.S. Food and Drug Administration; 2022. Accessed at www.fda.gov/news-events/press-announcements/coronavirus-covid-19-update-fda-limits-use-certain-monoclonal-antibodies-treat-covid-19-due-omicron on 12 June 2022. 62. U.S. Food and Drug Administration. FDA updates Sotrovimab emergency use authorization. Accessed at www.fda.gov/drugs/drug-safety-and-availability/fda-updates-sotrovimab-emergency-use-authorization on 12 June 2022. 63. Takashita E , Yamayoshi S , Simon V , et al. Efficacy of antibodies and antiviral drugs against Omicron BA.2.12.1, BA.4, and BA.5 subvariants [Letter]. N Engl J Med. 2022;387 :468-70. [PMID: ] doi:10.1056/NEJMc2207519 35857646 64. Nyberg T , Ferguson NM , Nash SG , et al; COVID-19 Genomics UK (COG-UK) consortium. 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. 2022;399 :1303-12. [PMID: ] doi:10.1016/S0140-6736(22)00462-7 35305296 65. Centers for Disease Control and Prevention. COVID-19 Rebound After Paxlovid Treatment. 24 May 2022. Accessed at https://emergency.cdc.gov/han/2022/pdf/CDC_HAN_467.pdf on 20 September 2022. 66. Deo R , Choudhary MC , Moser C , et al; ACTIV-2/A5401 Study Team. Viral and symptom rebound in untreated COVID-19 infection [Preprint]. med. Rxiv. 2022. [PMID: ] doi:10.1101/2022.08.01.22278278 35982660 67. Kaka AS , MacDonald R , Linskens EJ , et al. Major update 2: remdesivir for adults with COVID-19: a living systematic review and meta-analysis for the American College of Physicians practice points. Ann Intern Med. 2022;175 :701-9. [PMID: ] doi:10.7326/M21-4784 35226522 68. Lee TC , Vigod S , Bortolussi-Courval É , et al. Fluvoxamine for outpatient management of COVID-19 to prevent hospitalization: a systematic review and meta-analysis. JAMA Netw Open. 2022;5 :e226269. [PMID: ] doi:10.1001/jamanetworkopen.2022.6269 35385087 69. Castro M . Placebo versus best-available-therapy control group in clinical trials for pharmacologic therapies: which is better. Proc Am Thorac Soc. 2007;4 :570-3. [PMID: ]17878471 70. Pierre O , Riveros C , Charpy S , et al. Secondary electronic sources demonstrated very good sensitivity for identifying studies evaluating interventions for COVID-19. J Clin Epidemiol. 2022;141 :46-53. [PMID: ] doi:10.1016/j.jclinepi.2021.09.022 34555426 71. Deeks JJ, Higgins JPT, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses. In: Higgins JPT, Thomas J, Chandler J, et al, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 6.3. The Cochrane Collaboration; February 2022. Accessed at www.training.cochrane.org/handbook on 10 November 2022. 72. Li M , Lou F , Fan H . SARS-CoV-2 variant Omicron: currently the most complete “escapee” from neutralization by antibodies and vaccines. Signal Transduct Target Ther. 2022;7 :28. [PMID: ] doi:10.1038/s41392-022-00880-9 35091532	36442056	PMC9709728	NO-CC CODE	2022-12-03 23:19:51	no	Ann Intern Med. 2022 Nov 29;:M22-2202	utf-8	Ann Intern Med	2,022	10.7326/M22-2202	oa_other
==== Front Dtsch Z Akupunkt Deutsche Zeitschrift Fu¨r Akupunktur 0415-6412 1439-4359 Springer Medizin Heidelberg 521 10.1007/s42212-022-00521-w Editorial Rätselhafte Phänomene – Hilflosigkeit im System Mysterious phenomena – helplessness in the systemBachmann Jürgen [email protected] Schmerzmedizin – Orthopädie – Translationale Medizin, Sünsbruch 16, 45527 Hattingen, Deutschland 30 11 2022 2022 65 4 217217 12 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. issue-copyright-statement© The Author(s), under exclusive licence to Springer Medizin Verlag GmbH, ein Teil von Springer Nature 2022 ==== Body pmcLiebe LeserInnen und KollegInnen, vor Ihnen liegt die letzte Ausgabe des 1. Coronajahrganges – dies war der einleitende Satz meines Editorials vor 2 Jahren. Richtig müsste es jetzt also lauten: Vor Ihnen liegt die letzte Ausgabe des 3. Coronajahrganges. Wir haben Ihnen innerhalb dieser neuen Zeitrechnung immer wieder, so aktuell wie für Periodika möglich, Aspekte der Akupunktur und chinesischen Medizin im Zusammenhang mit der pandemischen Erkrankung aufgezeigt. Dies vor allem durch unsere Rubrik News unter der Ägide von A. Wiebrecht, dem ich an dieser herausgehobenen Stelle meinen herzlichsten Dank bekunden und tiefen Respekt zollen möchte. Dies ist leider nicht ohne Anlass, da er diese Aufgabe nach vielen Jahren des Engagements nicht mehr fortführen möchte. Daher mein Ruf in die Leser:innenschaft: Wo versteckt sich seine Nachfolgerin? Angesichts der hohen Zahl von Publikationen in unserem Interessenfeld – siehe den notorischen Studienzähler auf unserer Titelseite – eine wahrhaft wesentliche Aufgabenstellung! Seinerzeit, Ende 2020, hatte ich Ihnen angekündigt, dass wir ergänzend zu den News rund um die Akupunktur die einschlägigen Nachrichten zu den Möglichkeiten der chinesischen Medizin und den Erfahrungen im Mutterland des Virus als eigenständige Rubrik unserer News gefasst haben, verbunden mit der Hoffnung, dass sich die Notwendigkeit einer solchen Rubrik alsbald verlieren möge. Der Verlauf war anders, die Verläufe waren anders. Titelte die Ärztezeitung am 31.08.2022: „Das rätselhafte Phänomen Long-COVID“, gefolgt von der Schlagzeile „Patienten mit Long-COVID haben Anzeichen einer Autoimmunerkrankung“ am 22.09.2022, setzte die Presseagentur Gesundheit am Folgetag, dem 23.09.2022 mit der Meldung nach: „ME/CFS: Hilflosigkeit im System“. Daher sehen wir ähnlich wie die Kolleg:innen in der amerikanischen Zeitschrift Medical Acupuncture in ihrer Ausgabe vom Juni 2022 die Notwendigkeit, dem protrahierten Verlauf der Covid-19-Erkrankung, dort unter der Überschrift „Long Haulers“, in der vorliegenden Ausgabe einen Schwerpunkt zu widmen. Das Projekt wurde eine Chefsache, der Unterzeichner schritt ans Werk. Erfreulicherweise ist es nicht nur gelungen, interessante erste Erfahrungen und instruktive Fallberichte zu den verschiedenen Methoden der chinesischen Medizin zusammenzutragen, sondern auch die Grenzen eines Klassifikationsansatzes nach Krankheitsentitäten aufzuzeigen. Vor dem Hintergrund der vielfältigen und unspezifischen Symptome und daraus resultierend bislang nur unscharfen Definitionen erscheint eine am leiblichen Erleben und den Phänomenen orientierte Medizin wie die chinesische Medizin zumindest im derzeitigen Stadium der Erkenntnis besonders geeignet, handlungsleitende Einordnungen zu entwickeln. Allerdings: Angesichts der ausufernden Vielgestaltigkeit des Krankheitsbildes, der gedanklichen Ausgangspunkte, der konzeptionellen Einordnungen und der therapeutischen Lösungsansätze bestand die Berechtigung auf ungewöhnliche Wege. Die Kolleg:innen aus den USA setzten auf erweiterten Einsatz: ein Editorial, 2 Gast-Editorials sowie das Editorial eines Senior Editor. Die deutsche Zeitschrift für Akupunktur beschritt einen schmaleren und doch anspruchsvollen Handlungspfad: eine Einführung aus integrativer psychiatrischer Sicht mit dem Gast-Editorial von T. Heise. C. Chiu teilt anhand ihrer Darstellung über einige Fälle zur Anwendung des Lasers Erstaunliches mit, das manchem gestandenen Trigger-Akupunkteur lieb gewordene Lehrsätze erschüttern könnte. Auch die Rubrik „Wissenschaft für die Kitteltasche“ hat diesmal Pause Unser Ex-Chefredakteur und Krimiautor T. Ots verzichtet diesmal auf seine Kolumne, die Erkenntnisse zur Krise haben ebenso wenig wie unser Projekt „Die andere Seite“ relevanten Nachhall in der Leserschaft gefunden. Auch die Rubrik „Wissenschaft für die Kitteltasche“ hat diesmal Pause, der „Journal Club“ aber nicht. J. Fleckenstein stellt Ihnen eine Bayes-Netzwerk-Analyse zu einer zentralen Akupunkturindikation im deutschen Gesundheitswesen vor, dem chronischen unspezifischen Lenden-Becken-Hüft-Schmerz, vulgo: Rückenschmerz. Der Erkenntnisgewinn liegt vor allem darin, eine solche Analyse kennenzulernen. In diesem Sinne wünsche ich Ihnen eine stimulierende Lektüre. Falls diese überschwellig ausfallen sollte: Leserbriefe per E‑Mail sind ab sofort portofrei! Mit herzlichen kollegialen Grüßen Ihr Jürgen Bachmann Interessenkonflikt J. Bachmann gibt an, dass kein Interessenkonflikt besteht.	0	PMC9709730	NO-CC CODE	2022-12-01 23:23:08	no	Dtsch Z Akupunkt. 2022 Nov 30; 65(4):217	utf-8	null	null	null	oa_other
==== Front International Journal of Ethics Education International Journal of Ethics Education 2363-9997 2364-0006 Springer International Publishing Cham 159 10.1007/s40889-022-00159-1 Article Cultivating character in female student leaders: Case of a leadership program of an NGO in the Philippines http://orcid.org/0000-0002-2770-0248 Contreras Eunice [email protected] grid.443220.3 0000 0001 2170 0223 School of Education and Human Development, University of Asia and the Pacific, Pearl Drive, Ortigas Center, 1605 Pasig City, Philippines 30 11 2022 119 1 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. How can students’ character formation be supported such that their youthful energy can become a force for the good? There is burgeoning research on how universities can help form people of character (Brooks et al. in International Journal of Ethics Education 4(2):167–182, 2019; Lamb et al. in Journal of Moral Education 1–23, 2021b). Nongovernmental organizations can also play a role. This article explores how a leadership program for students in the Philippines cultivates character using as theoretical framework the 7 strategies of character development presented in Lamb et al. in Journal of Character Education 17(1):81–108, 2021a. Upon examination of how the strategies are integrated, recommendations for improvement of the program design are given. Keywords Character Leadership program Virtue NGO Female student leaders ==== Body pmcBackground How can young students’ character formation be supported such that their youthful energy can become force for the good? There is burgeoning research on how universities can play an important role in forming people of character through extra-curricular programs and course interventions (Brooks et al. 2019; Lamb et al. 2021b). Nevertheless, universities are not the only venue for character education. Education is not limited to formal settings such as degree-granting academic institutions. Education can also take place in non-formal settings to meet particular needs (e.g., professional development courses) and informal settings such as in the family and the workplace (Altarejos and Naval 2011). Van Noy et al. (2016) discuss two types of informal learning in their review of literature—(1) organized and (2) everyday informal learning. This article focuses on an organized informal learning program offered by a non-profit nongovernmental organization (NGO) in the Philippines, the Kalinangan Youth Foundation, Inc. (KALFI). The learning program, called KALFI LEAD, is a leadership program for young women that aims for holistic character development. Virtues of character and leadership Leadership has been associated with virtues of character. In various speeches of elected officials in the Philippines, leadership is always connected with virtuous qualities. For example, the late Senator Miriam Defensor Santiago, in a speech for Filipino students participating in the Youth Forum on Student Leadership and Nationalism in 2015, said that leaders, for their wide reach, must “influence for the better, not for the worst.” She emphatically underlined that “leaders are not born, [rather] it takes practice to develop the qualities of a leader” and indicated the following characteristics championed by good student leaders: integrity—which includes trustworthiness, honesty and consistency—compassion for others, initiative, and vision. Senator Panfilo “Ping” Lacson, a candidate for president in the 2022 National Elections, said in December 2021 that courage, integrity, and loyalty—virtues he learned during his studies at the Philippine Military Academy—are needed to lead the country in the face of many challenges. Speaking at the 2018 Young Professionals Summit, Vice President Leni Robredo—also a candidate for president in the 2022 National Elections—described the good leader as one who inspires, empowers, and listens (Rey 2018). She added that leadership is not about instilling fear in the people but about inspiring trust and confidence, practicing empathy and collaborative leadership. Even former President Rodrigo Duterte, whose administration’s “war on drugs” has allegedly committed human rights violations, called on the Filipino people on National Heroes Day in 2017 to imitate the good qualities of Filipino heroes being commemorated: “we will harness the same virtues [of our heroes] as we continue to fight against lawlessness, criminality, and poverty that hinder us from achieving our full potential.” The literature on leadership also includes virtues of character pertinent to good and effective leadership (see Cameron 2011; Gini and Green 2013; Hackett and Wang 2012; Newstead et al. 2021). For example, Hackett and Wang (2012) reviewed the literature in seven leadership theories, namely 1) moral leadership, 2) ethical leadership, 3) servant leadership, 4) spiritual leadership, 5) charismatic leadership, 6) transformational leadership, and 7) visionary leadership, and—from the 59 virtues or character traits mentioned—found nine virtues common across these approaches: caring, courage, honesty, integrity, justice, prudence, responsibility, temperance, and trustworthiness. Hackett and Wang (2012) further identified six virtues as cardinal leader virtues that are foundational, universal, and interrelated: courage, temperance, justice, prudence, humanity, and truthfulness. Focusing on the virtuous leadership theory, Wang and Hackett (2016) distinguished it from other leadership theories with moral or ethical features defining it as a “leader–follower relationship wherein a leader's situational appropriate expression of virtues triggers follower perceptions of leader virtuousness, worthy of emulation” (p. 326). The emphasis is on the virtues of character—“‘good’ leader traits” (p. 329)—expressed in virtuous behavior and sought intentionally as a good in itself. Recently, Brooks (2021) made a case for the role of character in leadership in the twenty-first century. He claimed that to appropriately respond to contemporary challenges, leadership competencies must be developed but the cultivation of leaders’ character virtues cannot be neglected. Brooks (2021) argued that the current societal challenges have their origin or are due, in part, to lack of virtuous character in leaders: “greed, hubris, dishonesty, dogmatism, close-mindedness, and presumption—to name only some prominent leadership vices—have played a part in the cause and ongoing impact of such events as the global financial crisis, turmoil in the Middle East, the COVID pandemic, and climate change” (p. 40). He also indicated that virtues serve as a bridge between aspiration and action thus ensuring a stable and wise integration of values in concrete behavior. He referred to the virtues “act[ing] as the intellectual and moral muscles that enable values to live” (p. 40). Lastly, Brooks (2021) considered that a variety of virtues like empathy, compassion and service (classified as pro-social virtues), open-mindedness, judgment and practical wisdom (classified as intellectual virtues), and hope are needed given the “connectedness, complexity, and precarity” of the current environment facing leaders (p. 40). Cultivating virtues of character Developing people as leaders involves the cultivation of virtues of character. Virtue, according to Aristotle (ca. 350 B.C.E./2004), “is a state involving rational choice, consisting in a mean relative to us and determined by reason—the reason, that is, by reference to which the practically wise person would determine it” (1107a 1–5). Virtues are generally considered as excellent qualities of one’s character—ἀρετή, virtue in Greek, basically means excellence or perfection—that allow the person to think, feel, and act “at the right time, about the right things, towards the right people, for the right end, and in the right way” (Aristotle ca. 350 B.C.E./2004, 1106b21-23). According to Aristotle, a virtue of character is acquired by carrying out actions like those of persons already possessing a given virtue (Pakaluk 2005). Although this gives character educators some hints as how to support character formation—by modelling themselves the virtues they would like their students to cultivate—other strategies culled from diverse fields of research are also valuable. Lamb et al. (2021a) synthesized the research in philosophy, psychology, and education and presented seven strategies for cultivating virtue within an Aristotelian framework: (1) habituation through practice, (2) reflection on personal experience, (3) engagement with virtuous exemplars, (4) dialogue that increases virtue literacy, (5) awareness of situational variables, (6) moral reminders, and (7) friendships of mutual accountability. For Aristotle, virtue, a stable disposition, is born from repeated actions which have an effect on the agent (ca. 350 B.C.E./2004, 1050b) hence, the strategy of habituation through practice. However, virtue is not mere automatism, a simple repetition of acts devoid of consciousness of the end and its choice, rather, it includes knowledge of the end and the right judgment of practical wisdom (Rodriguez Luño 1988) thus the pertinence of reflection on personal experience. The person of good character or the wise person (φρόνιμος) is the model or measure for virtue, according to Aristotle (ca. 350 B.C.E./2004, 1107a 1–5, 1113a 32–34), thus the need for engagement with virtuous exemplars. The practice of virtues would require knowledge of particular virtues and their application in different situations thus the relevance of dialogue that increases virtue literacy. In addition, given that actions are a fruit of a confluence of factors including situational variables and biases, it is necessary to increase awareness of situational variables that may inhibit or promote the practice of virtue, thereby allowing the agent to address them appropriately. As mentioned, the practice of virtue requires constancy thus consistent moral reminders that prompt persons about their moral commitments are fitting. Lastly, the cultivation of friendships of mutual accountability is vital for a person is a social being (Aristotle ca. 350 B.C.E./2004, 1169b 18) and one’s character is not formed in isolation. For Aristotle (ca. 350 B.C.E./2004), friendship is a relationship characterized by reciprocal goodwill where each person desires what is good for the other, i.e., to become more virtuous (1156a 1–5). Purpose of the study The seven aforementioned strategies provide educators interested in supporting the youth’s character development a concrete course of action; however, effective virtue cultivation would require a concentrated and intentional effort which can be addressed by holistic programs of character development. One such program is the Oxford Global Leadership Initiative (GLI), a voluntary, extracurricular leadership program for postgraduate students in the University of Oxford established in 2014. Brooks et al. (2019) reported the considerable potential of programs integrating leadership and character development like GLI in influencing the emerging leaders’ lives and their immediate context, together with the promise of long-term social impact. Given this encouraging picture, this article explores KALFI LEAD, a leadership program that claims to promote holistic character development in its participants. Since such integrated leadership and character development programs have a great potential for the individual and the society, it is necessary to inquire whether the KALFI LEAD program design incorporates research-based strategies for character development. Using documentary analysis of the so-called analytic induction approach (Patton 2002), I examined the documents provided by KALFI LEAD program staff (i.e., handbook and the curriculum), and those publicly available on the internet (i.e., website and YouTube videos) applying Lamb et al.’s (2021a) Aristotelian model of character development as theoretical framework. In other words, I analyzed the KALFI LEAD documents with the seven strategies of character cultivation and their definitions in mind, discerning whether a particular strategy is incorporated and in what manner. Lamb et al.’s (2021a) description of each strategy was supported by specific and practical examples on its application in a holistic leadership and character development program thus lending itself excellently to educators seeking to design or improve programs of character development. This study builds upon the contribution of Lamb et al. (2021a) to the theory and practice of character education by applying the framework they developed as guide in improving KALFI LEAD. Although the character development strategies Lamb et al. (2021a) presented was applied in a university setting through GLI, the strategies themselves can be used in character formation programs for diverse age groups. In fact, Lamb et al.’s (2021a) discussion of each strategy utilized literature dealing with varying age groups. Excepting the differing target age group, KALFI LEAD is akin to GLI in that it is also a voluntary leadership program and aims to supplement the curricular offerings of formal educational institutions. I begin with a brief description of the NGO called KALFI to facilitate an understanding of the context in which the leadership program is delivered. I then proceed with the analysis proper where I first describe each of the KALFI LEAD program design components and then examine in what ways are the seven strategies for cultivating virtue (Lamb et al., 2021a) integrated. The KALFI LEAD program is a holistic leadership program with six main components. It is the whole program, and not simply one component, that aims to cultivate leadership and character in the participants. For each component, I specify what strategies for virtue cultivation are incorporated. I also cite examples of how the strategies are translated in real-life, with the aim of giving program designers and other educators inspiration on how the Aristotelian framework could be concretely integrated into their own leadership programs. Furthermore, I intersperse the analysis with testimonials from the KALFI LEAD participants providing glimpses into the impact of the program on their lives. I end the analysis by giving several recommendations for improvement of the program design. Program description Kalinangan1 Youth Foundation, Inc. (KALFI) Kalinangan Youth Foundation, Inc. (KALFI) is a private non-stock, non-profit foundation established in 1983 with the mission of providing young Filipino women personalized mentoring and meaningful youth development programs, envisioning them as “empowered young women with integrity, achieving their personal best and contributing positively to society” (KALFI 2019). KALFI’s various activities are offered in the study centers it established or assists in eight different locations in the country. The centers are located near schools and universities making its activities for holistic formation accessible to the youth. KALFI LEAD: Raising a new wave of women leaders KALFI started a leadership program in 2011 called KALFI LEAD with the aim of forming young women as exemplary servant leaders through holistic character development (KALFI 2022c). LEAD stands for Leadership, Excellence, Accountability, and Discipline and the four-year program aims to inculcate these virtues in its participants. KALFI LEAD is offered to a select number of female students. Senior high school and college students apply and undergo a selection process based on academic qualifications, past leadership experiences, leadership essays, and recommendation forms. As a monitoring tool for fulfillment of requirements, and hence retention, the leadership program implemented a points system where participants gain a number of points for required, additional, and optional activities and must meet a minimum of 500 points at the end of each semester. The program’s motto, “in order to be useful, serve,” shows KALFI LEAD’s emphasis on servant leadership. The program seeks to develop in the participants an “attitude of and ability to SERVE, i.e., self-giving” (KALFI 2021). This conception of leadership as service to others recalls Robert K. Greenleaf’s theory of servant leadership—a “philosophy, embedded in a set of behaviors and practices that place a primary emphasis on the well-being of those being served” (Greenleaf Center for Servant Leadership 2021). The program’s focus on instilling social responsibility in its participants is due to KALFI LEAD being inspired by the social teachings of the Catholic Church that foregrounds the dignity of the human person and the preferential option for the poor and vulnerable, among others. Although inspired by Christian ideals and offers optional spiritual formation activities like retreats, the KALFI LEAD program is open to students, volunteers, and staff of all creeds. Cognizant that virtuous qualities are necessary for a person to lead through service, KALFI LEAD provides character formation through the different components of the leadership program. The program is “designed to assist students’ growth in virtue, which enables them to struggle and attain the good that they seek” (KALFI 2022a). This recognition of the role of virtues of character in leadership brings to mind Brooks’s (2021) understanding of virtues acting as the muscles that enable values to be brought to life in stable concrete behavior. KALFI LEAD program design components KALFI LEAD cultivates leadership and virtues of character in its participants through a holistic program. The analysis of the program documents reveals six main components: (1) leadership development curriculum, (2) outreach or service projects, (3) participation in orientation events, teambuilding, and conferences, (4) on-the-job training, (5) leadership project, and (6) mentoring. All components work in tandem to cultivate leadership and character. The discussion is presented in tabulated form at the end of the section, followed by recommendations for the improvement of the program design. Leadership development curriculum (1) Conscious that leaders are not born, KALFI LEAD developed a leadership development curriculum that intends to provide participants with the theoretical knowledge necessary to lead effectively. The curriculum is delivered through various modules under nine themes grouped under four blocks as indicated in Table 1 that are spread throughout the four-year program. Resource speakers are invited to deliver each module through dynamic workshops or talks.Table 1 Leadership Curriculum from 2019–2022a as indicated in the KALFI Handbook and Planner (2021) Block 1: Defining for purpose Block 2: Designing for impact Block 3: Developing capability Block 4: Driving performance Theme 1: Knowing the self Theme 4: Impact through leadership: best practices Theme 6: Marketing and public relations Theme 9: Positive change in society Theme 2: Knowing what society needs Theme 5: Impact through strategic excellence Theme 7: Finance Theme 3: Knowing what is right Theme 8: Human Resources aThe leadership curriculum in the previous years had 4 different themes: 1) Introduction to leadership: Who am I?, 2) Leading with integrity: Am I good? Am I whole?, 3) Lead to serve: Do I live for others?, and 4) Vision: Where will I lead my people? The curriculum from Academic Year 2022 onwards still follows the four Blocks but the nine themes have been replaced by concrete topics that may be offered. One example of a module delivered in 2021 under the theme “Knowing what society needs” is “Step Up: A KALFI LEAD Module on Serving the Poor.” The three guest speakers invited were involved in social initiatives. One speaker, a management staff of one KALFI study center, drew from her economic and theological background to discuss the principle of preferential option for the poor. Another speaker was an official of a school established by an NGO that offers technical-vocational education to underprivileged young Filipino women who talked about the social impact of their school. The last speaker was the executive director of another NGO that provides poverty alleviation interventions in various sites in the Philippines. After the resource speakers, six older KALFI LEADers were invited to present their own initiatives for their communities to serve as inspiration for the new participants. For example, one second year university student spoke about her volunteer work at a local NGO offering weekly character building sessions for teachers, students, and parents. Another KALFI LEADer, a senior high school student, presented the social initiative she started called Escentia PH that supports the women of a local village who became unemployed due to the Covid-19 pandemic by marketing and selling their homemade organic soap products. The leadership development curriculum of the KALFI LEAD program incorporates various strategies of character development provided by Lamb et al. (2021a). Given that each module is delivered by resource speakers invited for their expertise and leadership experience characterized by service and virtue, participants are given the opportunity to engage with exemplars that serve as models to admire, emulate, and learn from. When asked what impact did KALFI LEAD have on her, one student said “To be honest, what I like most about KALFI LEAD are those Sunday2 classes where they introduce us to professors3 that are highly esteemed. We get to learn a lot from them which are outside the four walls of the classroom, the things which you can use in real life and for that I am really grateful” (Kalfi Leader 2017). Additionally, given that the delivery of the modules are spread out throughout the four-year program, the participants are provided regular moral reminders of their commitment to both servant and virtuous leadership. The leadership modules also serve to increase the participants’ virtue literacy. Revealing the impact of the KALFI LEAD modules, a KALFI LEADer shared that “the lessons I learn from KALFI LEAD online4 modules and talks also widen my perspective about social topics and youth values. All the formation I get from KALFI helps me improve myself every day and remain grounded at the same time” (KALFI LEAD 2021). Outreach or service projects (2) With the goal of fostering an attitude of service and a stable commitment to serve society, KALFI LEADers join an outreach activity organized in the study center they are assigned to. They dedicate time and energies to giving weekly academic tutorials or catechism classes to children from underprivileged neighborhoods. Another outreach activity that students can choose to join are regular visits to the sick in hospitals or in their homes or visits to poor communities. Students are also asked to spearhead a service project of their own choice and design. As a leadership program, KALFI LEAD is unique in that the participants are “offer[ed] constant opportunities to participate in socio-civic activities and extend help to local communities” (KALFI 2022d). The many opportunities offered to KALFI LEADers to engage in service projects integrates the strategy of habituation through practice, the first strategy presented by Lamb et al. (2021a). Recognizing that leadership marked by service and virtue cannot be learned from the leadership modules alone, the KALFI LEAD program stresses upon the participants’ stable and active involvement in outreach activities and other service projects held annually or monthly. Students have remarked on how beneficial these outreach activities have been: one student said, “one of the main things that KALFI LEAD has encouraged over the past year was doing visits [to the sick and the poor]. At first I didn’t realize how important they were but then I think that these visits were incredibly essential to help me get a better picture of what the world is really like” and she continued, “it really pushed me out of my comfort zone because you never really know how generous you can be until you have given up everything” (Kalfi Leader 2017). Her remarks reveal that her participation in the outreach activities have helped her see the real needs of the community, which is important for the cultivation of empathy (Anderson and Konrath 2011), and helped her grow in generosity through the commitment to a regular outreach activity. Attesting to how the practice of service and virtue serves to cultivate such in the person, another student declared that “we are not only helping the beneficiaries but we are also helped in the process” (Kalfi Leader 2017). For one LEADer, the weekly dedication to outreach “taught [her] that service is a commitment” (Banilad Study Center A Project of KALFI 2018). Participation in orientation events, teambuilding, and conferences (3) At the start of the program, new participants take part in an orientation organized by older KALFI LEADers. Participants also take part in a teambuilding called Anchored Leadership with their cohort each year as well as in an inter-cohort retreat-seminar, Leap to Lead. These events allow the KALFI LEADers to forge friendships with their peers. This component of the KALFI LEAD program exemplifies the following strategies of character development provided by Lamb et al. (2021a): habituation through practice and friendships of mutual accountability. The annual teambuilding is a venue where the LEADers can practice leadership in an informal setting through collaboration with one another. For KALFI LEAD, leadership is not reduced to possession of power based on an institutional position but is characterized by collaboration, service, and virtue. Additionally, these orientation and teambuilding activities provide an optimum space for the development of friendships among the participants. The teambuilding is usually held for two days and the opportunities for sharing meals and other moments outside the formal program allow the participants to converse and deepen in their knowledge of one another thus creating a sense of community within each cohort of LEADers. One participant appreciated this aspect: “It’s very overwhelming to be surrounded by women with the same values who have voices wanting to be heard as we continue to empower one another” (Banilad Study Center A Project of KALFI 2018). Another expressed her gratitude for the friendships that were made: “Words can’t express how thankful I am towards KALFI. As I look back and remember all the memories we have made together, I would say that I will be forever thankful and grateful to KALFI because of the amazing experiences I have shared with the wonderful people I met (mentors and unexpected fellow girls who also became my friends)” (KALFI LEAD 2021). KALFI LEADers also participate in an annual conference. The KALFI LEAD Summit is a student leadership conference organized by fourth-year KALFI LEADers for their peers. The Summit convenes high school and college students and current KALFI LEADers to talks by successful KALFI LEAD alumnae and other prestigious professionals tackling a particular theme that hopes to broaden the students’ perspectives and inspire novel and relevant social initiative ideas. The Summit also serves to commemorate the KALFI LEAD program’s recent milestones and other accomplishments. KALFI LEADers also prepare an exhibit of their current projects or experiences in the program and answer guests’ questions, creating a space for social networking and friendships with other like-minded youth leaders. To date, there have been eight KALFI LEAD Summits5—the first one, held in 2014, was organized by the first KALFI LEAD cohort. The Summit incorporates the following strategies of character development provided by Lamb et al. (2021a): engagement with virtuous exemplars, dialogue that increases virtue literacy, awareness of situational variables, and friendships of mutual accountability. Resource speakers invited for the Summit are prestigious professionals and KALFI alumnae who speak on a given theme that serves to widen the participants’ perspective. Their sharing of their own experiences in leading in their own milieu serves to make leadership within the reach of the participants and be a model for them to emulate. The Summit also incorporates dialogue that increases virtue literacy and awareness of situational variables through the open forum where KALFI LEADers ask the speakers further questions regarding what they share that are, oftentimes, questions about how they faced a particular situation and what steps they took to overcome a certain challenge. Lastly, given the time provided for interaction between the participants during the exhibits, there is an opportunity for friendships of mutual accountability to develop. Other participants pose questions to the poster presenters in an atmosphere of trust and desire to learn or help the other improve. On-the-job training (4) KALFI LEAD participants are fielded out to different partner institutions for on-the-job training (OJT) which lasts for 12 days or 96 hours throughout one academic year. Although due to the Covid-19 pandemic, KALFI LEADers since 2020 have not had their OJT, past partner institutions included vocational training schools for the Home Leadership program; a Philippine government agency for Media Leadership; an NGO for Servant Leadership; a restaurant for Business Leadership; and the KALFI offices for the Administrative Leadership program. This component of the KALFI LEAD program exemplifies the first strategy of character development provided by Lamb et al. (2021a), habituation through practice. Through the OJT, KALFI LEADers are given the opportunity to practice leadership in diverse settings—home, government agency, NGO, and industry. The OJT is also an opportunity for the KALFI LEADers to engage with exemplars—their OJT work supervisors and other colleagues—whom they can learn from, emulate or, in the case that they discern actions not in keeping with virtuous practice, refrain from imitating. Their OJT workplace could also be conducive in fostering awareness of situational variables for the KALFI LEADers may learn from the workplace how a certain virtue is practiced and what variables may affect, positively or negatively, the effort to practice virtue within the daily operations of an organization. Leadership project (5) KALFI LEADers are given the opportunity to apply what they have learned from the leadership modules by spearheading various projects throughout the four-year program. For example, they organize the orientation for the new cohorts as well as the KALFI LEAD Summit in their last year on the program. They also design and manage their own social initiatives to benefit selected beneficiaries. They work on the projects first as a cohort, then in small groups, dyads, and as an individual. This component of the KALFI LEAD program incorporates habituation through practice, the first strategy of character development provided by Lamb et al. (2021a). By actually collaborating with other KALFI LEADers in group projects, the students are given the opportunity to practice leadership within their given role in a project which may not necessarily be as the overall-in-charge. Such arrangements bring to the fore that leadership is not about occupying a position but about leading characterized by collaboration, service, and virtue. Mentoring (6) Mentoring is a major component of the KALFI LEAD program and is claimed to be the secret of the leadership program’s success. The mentoring program gives KALFI LEADers an opportunity to gain from the mentor’s “practical wisdom, sincere guidance, undivided attention, and friendship” (KALFI 2022c). It also seeks to support the participants in learning from and finding meaning in their journey into young adulthood marked by significant changes and transitions. According to KALFI (2022b), “helping someone learn how to learn is the fundamental process and the primary purpose of mentoring.” KALFI LEADers are paired with experienced professionals who serve as their mentors. These mentors are volunteers of KALFI who receive training for basic mentoring skills (e.g. presence, active listening, communication, receiving and giving feedback, building resiliency). To safeguard the culture of excellence, the mentees are encouraged to give feedback about the mentor to the KALFI LEAD program heads. Mentors and mentees meet at least twice a month for a mentoring session where the mentee’s holistic development—socio-emotional, cognitive, and identity—is the focus. At the start of the KALFI LEAD program, both mentors and mentees fill out a mentoring agreement where they settle on the frequency of the mentoring sessions, the duration for each session, and the general topics to be covered. The mentoring session is intended to be a reflective space for the mentee where the mentee can “reflect on and learn from key incidents and think through their own solutions” (KALFI 2022b). The mentor is tasked to create an atmosphere of openness and trust that promotes a healthy dialogue. Through the mentor’s questions, the KALFI LEADer is enabled to “reflect on her motivation for wanting to excel in her chosen field and contribute to society in a significant way” (KALFI 2022b). In mentoring, “the mentee gains from the experience and knowledge of the mentor who uses such knowledge to hone the student towards developing her talents and possibilities to the full” (KALFI 2021). Mentors bring up the topics tackled in the leadership modules and support the mentee in her own goal-setting. Mentoring becomes a venue where theories are concretized so as to be lived by the student. The mentoring conversation highlights the cultivation of virtues like “moral integrity, perseverance, magnanimity, and humility” because KALFI LEAD considers that people “cannot take the lead without developing themselves into complete individuals” (KALFI 2022c). The mentee is helped to both appreciate and exercise the virtues cognizant that “servant leaders should practice what they preach” and “demonstrate these admirable traits and lead by example” (KALFI 2022d). The mentors also model the target LEAD virtues and share their own experiences and effort in living them. The mutual sharing lays the ground for a meaningful friendship—appropriate to the mentoring relationship—between mentor and KALFI LEADer. For KALFI LEAD, “mentoring and friendship are keys to developing character” (KALFI 2022b). Although not strictly a mentoring relationship, KALFI LEADers also have frequent interactions with the KALFI LEAD program manager, sector heads (of Student Affairs, Mentoring, Events and LEAD Curriculum), and the LEAD coordinator of their assigned study center with whom they can consult matters and ask advice from. The program staff themselves model the target virtues. This component of the KALFI LEAD program incorporates various strategies presented by Lamb et al. (2021a). Habituation through practice and engagement with virtuous exemplars are incorporated through mentors and program officers modeling the target virtues of the LEAD program. Additionally, reflection on personal experience is incorporated in the one-on-one mentoring session where the LEADers are asked to reflect upon their circumstances and come up with their own action plans. A KALFI LEADer commented that “my mentors are always willing to listen to my plans and give their guidance, suggestions, and support” (KALFI LEAD 2021). Such mentoring sessions also incorporate dialogue that encourages virtue literacy for they are structured not as a lecture where the LEADer simply absorbs all that the mentor talks about but as a two-way conversation where mentors actively listen to the LEADer’s concerns while, at the same time, giving the necessary advice and encouragement that the LEADer can clarify and adapt or adopt to her life. One KALFI LEADer shares her experience: “To my mentor, I open up about my struggles and everything that I need to overcome and I think that it really helps because she’s there to say it’s okay, try again” (Kalfi Leader 2017). The mentor’s sharing of her own experiences is a concrete manifestation of how the program tries to raise awareness of situational variables that affect virtuous behavior. Through the sharing of someone older and more experienced, the LEADers are made aware of the situational variables that can inhibit the carrying out of virtuous behavior or practices in the workplace etc. The twice a month mentoring session also serves as a moral reminder of the LEADer’s commitment to leadership characterized by service and virtue. Lastly, although a professional tone is maintained in the dealings between mentor and mentee and between the program officers and the participants, friendships of mutual accountability are developed—the mentors and program officers strive to help the participants improve and thus practice correction through giving of regular feedback and strive to develop the students’ sense of responsibility. The participants themselves are asked to practice correction and help the mentors and program officers by giving their observations if the culture of excellence is being preserved or not. The one-on-one mentoring is a venue for valuable learnings and cultivation of friendships. One participant shared “through mentoring, I get advice that is tailor-fit to my needs as a student, as a leader, and as a person. Not only that, I also get to bond with my mentor through our regular chikahans [catching up] in the mentoring sessions” (KALFI LEAD 2021). The preceding analysis is presented in tabulated form below. Based on the data from the documentary analysis, KALFI LEAD does incorporate all seven strategies of character development. The first column of Table 2 lists the seven strategies by Lamb et al. (2021a) and the second column specifies in what way KALFI LEAD integrates them.Table 2 Analysis of KALFI LEAD’s integration of the 7 strategies of virtue cultivation (Lamb et al., 2021a) Seven strategies of Character Development (Lamb et al., 2021a) How KALFI LEAD integrates them (1) habituation through practice participation in regular outreach or service projects participation in annual teambuilding where participants practice leadership in an informal setting on-the-job training (OJT) which lasts for 12 days or 96 hours throughout 1 academic year in diverse settings—home, government agency, NGO, and industry participants designing and managing their own leadership project mentors and program staff model the target virtues (2) reflection on personal experience one-on-one mentoring session where the LEADers are asked to reflect upon their circumstances and come up with their own action plans (3) engagement with virtuous exemplars invited resource speakers for the leadership modules and the annual leadership conference serve as models to admire, emulate, and learn from mentors and program officers model the target virtues OJT as an opportunity to engage with exemplars (e.g. OJT work supervisors and other colleagues) to learn from, emulate or, if the case may be, refrain from imitating (4) dialogue that increases virtue literacy open forum where KALFI LEADers ask invited speakers further questions mentoring sessions as a two-way conversation where mentors actively listen to the LEADer’s concerns while, at the same time, giving the necessary advice and encouragement that the LEADer can clarify and adapt or adopt to her life (5) awareness of situational variables open forum where speakers are asked further questions about their personal experience to overcome a certain challenge mentor’s sharing of her own experiences during mentoring experience in the OJT workplace reveals how a certain virtue is practiced and what variables may affect the effort to practice virtue within an organization (6) moral reminders whole KALFI LEAD program lasts four years delivery of leadership modules is spread throughout four years mentoring occurs twice a month (7) friendships of mutual accountability belonging to a community of youth leaders with shared values informal moments during teambuilding and outreach provide an optimum space for the development of friendships conversations between poster presenters and participants during the exhibit LEAD mentors and program officers give the LEADers regular feedback while LEADers are also asked for their observations on the mentors and officers Recommendations for the improvement of the program design Granted, KALFI LEAD incorporates all seven strategies of character development, however, the integration of each strategy can be improved. Based on the data from the analysis and the consulted literature on leadership and character education programs, I give several specific recommendations for program incorporation. A recapitulation of the recommendations are presented in tabular form at the end of the section. Based on the analysis of KALFI LEAD’s program design, many of the strategies of character development converges in the mentoring component. However, the strategies of habituation of virtues through practice and moral reminders can be integrated further. The mentor can encourage the KALFI LEADer to identify a virtue she would like to cultivate and commit herself to concrete behavior manifesting that virtue over a period of time akin to the ‘Plan for Character Development’ being used in the course, ‘Commencing Character: How Should We Live,’ of Wake Forest University (WFU) in the United States. Taking inspiration from a practice of Benjamin Franklin, in that course, “students choose a target virtue, write a conceptual analysis of the virtue and its related vices, apply a plan to habituate the target virtue over a two-week period, and then write a reflection on their experience” (Lamb et al. 2021b, p. 5). To serve as a ‘moral reminder,’ the mentor will guide the KALFI LEADer in each mentoring session to take stock of the progress from the past weeks and encourage her to consistently and continuously implement her plans for the cultivation of the target virtue. The above recommended practices would allow the intentional cultivation of virtues of character, essential in an Aristotelian conception of character formation. Such deliberateness is necessary for good intentions of living the virtues alone, without a plan for their cultivation, are fragile. To integrate reflection on personal experience and dialogue that encourages virtue literacy, the delivery of a pertinent leadership module as well as the teambuilding can be enhanced to also include a segment where KALFI LEADers can reflect and identify the virtues of character indispensable for good leadership based on their personal experience either through a group discussion or an individual written reflection (Lamb et al. 2021a). Such reflection is necessary for the development of practical wisdom, without which seemingly virtuous behavior would simply be an automatism, and not full virtuous action (Arthur et al. 2017). To integrate fostering awareness of situational variables that partly influence a person’s living out of virtues, a leadership module can incorporate a discussion of the toxic cultures or practices that may be present in the LEADers’ future professions, systemic inequalities and injustices in the Philippines society, and possible biases or prejudices that the LEADers may unconsciously hold. Behavior is not completely independent from the situation such that a person, for example, considered to be generous by her friends and family may, given an unfamiliar or extraordinary situation like being in a deserted street with a group of people of dubious character, may pass up the chance to help a needy person for fear of her safety. The disclosure of these external and internal variables—from those in the workplace or in wider society to personal tendencies and biases—that may influence virtuous behavior is valuable as a forewarning to guide one’s behavior appropriately. The current KALFI LEAD program gives prominence to engagement with virtuous exemplars by inviting resource speakers and selecting mentors who are not only competent in their fields but also are regarded as people of character. Another practice that could be adopted is to expose the KALFI LEADers to readings or videos about historic or contemporary exemplars, either autobiographies or third-person narratives. KALFI LEAD may find inspiration from the Oxford GLI that provides readings that “include an excerpt from a biography of Nelson Mandela; analyses of historical or contemporary leaders in politics, business, law, and medicine; and letters and personal narratives from influential thinkers and leaders” (Lamb et al. 2021a, p. 89). These readings or videos can be about Filipino historical and contemporary leaders in order to increase its relevance for the participants nevertheless, they can also feature leaders of other nations and cultures to develop participants’ universal perspective. KALFI LEAD may also find inspiration from the ‘Commencing Character’ course of WFU that requires a ‘Profile in Character,’ where students are asked to “identify a personal exemplar, interview that person about their life and character, and then write a 4–5-page profile of the exemplar’s character and its impact on the student’s life” (Lamb et al. 2021b, p. 5). Such a practice would generate an understanding of a more ordinary and everyday type of virtuous exemplarity and leadership which the LEADers could emulate in their own lives. Although the cohort-based design and various program features already encourage friendships among the KALFI LEADers themselves as well as with the program managers and mentors, friendships of mutual accountability can be further integrated through inclusion of a lecture or readings or videos on friendship—on what friendship is all about, its different kinds, and exemplars of friendships—and subsequent discussion of the topic among the KALFI LEADers and individually with their mentors. Given that not everyone may understand friendship as one that facilitates the cultivation of virtue, an exploration of Aristotle’s conception of authentic friendships is necessary. This is especially important in our times where a friend request sent and accepted already constitutes friendship and where friendships based on utility and pleasure or egocentrism are common. Lastly, and more importantly, the KALFI LEAD program can revisit its target virtues and elaborate on how each virtue is conceived and how might they be manifested in practice. As Berkowitz (2022) stated, starting with the end in mind is necessary if one is to design an effective character development intervention. Clearly, KALFI LEAD stresses on developing the participants’ attitude and ability to serve based on the conception of leadership as servant leadership and the program design is logically aligned with that outcome principally through its emphasis on participation in and organization of regular service projects. However, KALFI LEAD also claims itself to be a holistic character development program for virtues are regarded as indispensable if the LEADers are to really “attain the good that they seek” (KALFI 2022a). This is in consonance with Brooks (2021) conception of virtues acting as the “muscles” that enable the desire to serve to be brought to life in stable concrete behavior. Unfortunately, based on the documents analyzed, the KALFI LEAD program is not clear on the meaning of the virtues adopted as target virtues. There is a need to clarify the meaning of each target virtue within the context of a leadership development program so that program staff, mentors, and the participants themselves know exactly the desired outcomes.6 The first step to a robust conceptualization of each virtue is the classification of each target virtue as described below. Character education programs typically aim at cultivating a combination of moral, intellectual, civic, and performance virtues (McGrath et al. 2021; Shields 2011; Jubilee Centre for Character and Virtues 2017). The target virtues KALFI LEAD aims to cultivate in its participants are leadership, excellence, accountability, and discipline and they can be classified under moral and performance virtues, and, what Havard (2021) calls, leadership virtues. Accountability is a synonym for responsibility and discipline and these are readily understood as virtues. In fact, Character Counts! (2022) identifies responsibility as one of the six pillars of character while Lickona and Davidson (2005) include responsibility and discipline in their list of eight strengths of character. Both responsibility and discipline are considered as performance character strengths (Character.org 2022). Performance character strengths or performance virtues, according to the Jubilee Centre for Character and Virtues (2017), are “character traits that have an instrumental value in enabling the intellectual, moral and civic virtues” (p. 5). Leadership, on the other hand, can be classified under the moral virtues. The VIA Institute on Character (2022) conceptualizes leadership as a strength within the virtue of justice which is classically considered as one of the moral virtues.7 Lastly, excellence, intimately related with magnanimity, is one of the two virtues Havard (2021) considers as constitutive of the essence of leadership, the other being humility. The above recommendations are summarized on Table 3. The first column lists the seven strategies of character development and the second column contains the recommendations for the improvement of KALFI LEAD’s program design.Table 3 Recommendations for the improvement of KALFI LEAD's program design Seven strategies of Character Development (Lamb et al., 2021a) Recommendations for the improvement of KALFI LEAD’s program design (1) habituation through practice Bimonthly mentoring to incorporate a ‘Plan for Character Development’ where the participant is asked to identify a virtue she would like to cultivate and commit herself to concrete behavior over a period of time (2) reflection on personal experience An appropriate leadership module and teambuilding to be enhanced with a segment for reflection on the virtues indispensable for good leadership based on participants’ personal experiences either through a group discussion or an individual written reflection (3) engagement with virtuous exemplars Expose participants to readings or videos about Filipino and international historic or contemporary exemplars, either autobiographies or third-person narratives (4) dialogue that increases virtue literacy Same as recommendation for strategy #2 (5) awareness of situational variables A leadership module to incorporate a discussion of the toxic cultures or practices that may be present in the participants’ future professions, systemic inequalities and injustices in the Philippines society, and possible biases or prejudices that the LEADers may unconsciously hold (6) moral reminders Continuous and consistent implementation of ‘Plan for Character Development’ (7) friendships of mutual accountability Strengthen the participants’ understanding of friendship through the inclusion of a lecture, or readings or videos on friendship and subsequent discussion of the topic among the participants and individually with their mentors Clarify outcomes. It is recommended that program staff conceptualize what each KALFI LEAD target virtue (i.e., leadership, excellence, accountability, and discipline) means and how might they be manifested in practice Conclusion The KALFI LEAD program was designed to give female youth leaders relevant leadership training, instill in them a disposition of service particularly to the disadvantaged, and provide them with holistic character formation so as to “enable them to struggle and attain the good that they seek” (KALFI 2022a). Given such objectives and the concern for program effectiveness, it was necessary to confirm whether the KALFI LEAD program design incorporates research-based strategies for the cultivation of virtues of character necessary for leaders and thereafter, provide recommendations for the improvement of the program design. The program does incorporate the seven strategies of character development as presented in Lamb et al. (2021a) but there is room for improvement as can be seen in the given recommendations. More importantly, it is recommended that program staff, together with the mentors, conceptualize what each KALFI LEAD target virtue means and how might they be manifested in practice. Such a clarification on the outcomes is necessary in the efforts of program design improvement, implementation, and assessment. Given the great potential of holistic programs of leadership and character development for influencing emerging leaders’ lives and their communities (Brooks et al. 2019), programs similar to KALFI LEAD can be analyzed to determine whether strategies for leadership and character development are effectively incorporated in their program designs. The Aristotelian framework of character development (Lamb et al. 2021a), however, will have to be applied with attention to particular sociocultural and educational contexts of the institution and/or program under study. In the case of KALFI LEAD, it would be objectionable to indiscriminately adapt the practices of the Oxford GLI, the program that first embodied the framework, given the difference in sociocultural and educational contexts of both programs. Nevertheless, such framework for virtue cultivation was valuable in determining whether KALFI LEAD is designed in such a way that it is able to achieve what it claims to do. The next course of action would be to assess KALFI LEAD’s impact in the lives of its alumnae and their communities. This article has hopefully provided educators interested in the character development of the youth with ideas of how the Aristotelian framework can be concretely integrated into their own leadership programs. Acknowledgements I would like to thank Eula Marie D. Mangaoang for helpful comments on earlier drafts and the reviewers for their insightful suggestions. Declarations Conflict of interest None to declare. 1 The word “kalinangan” suggests the “laborious task of cultivating rough terrain” (KALFI 2019). 2 Prior to the Covid-19 pandemic, LEAD modules were given in person on Sundays. 3 Past resource speakers also included professors in Philippine universities. 4 Due to the Covid-19 pandemic, LEAD modules were delivered online through synchronous sessions. 5 The past KALFI LEAD Summits are as follows: Youth and Response Ability: The Essentiality of Initiative in 2014, Youth Empowering Society: Providing Transformative Solutions in 2015, Synergy: Individual Action Towards Conscious Co-Action in 2016, Kinaadman (Knowledge): Knowledge and Passion in Action for Service in 2017, Adhikain (Ambition): Fusing People, Purpose and Passion for Life to Impact Society in 2018, Likha: Adbokasiya sa Pagkakaisa (Creation: Advocacy for Solidarity) in 2019, Yugto: Ang Pagbangon ng Sambayanang Pilipino (Stage: The Rise of the Filipino People) in 2020, and MULAT: Pagtuklas sa Katotohanan ng Makabagong Mamamayan (CONSCIOUS: Discovering the Truth by the Modern Citizen) in 2021. 6 The description of the target virtues in the KALFI Handbook and Planner (2021) is not sufficient and robust enough to give program directors, facilitators, mentors, and participants an understanding of what the target virtue means and how it is manifested. For example, the target virtue of excellence is simply identified with the requirement of good academic standing. The identification of good academic results with excellence is suspect because a person may actually attain good grades in school through dishonest means. 7 Philosophers who have conceived of justice as a moral virtue are Plato, Aristotle, Augustine, and Thomas Aquinas, among others. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References Altarejos, F., and C. Naval. 2011. Filosofía de la educación (3rd ed.). Pamplona: EUNSA. Anderson, P., and S. Konrath. 2011. Why should we care? Psychology Today. Retrieved May 2, 2022, from https://www.psychologytoday.com/intl/blog/the-empathy-gap/201108/why-should-we-care Aristotle Nicomachean Ethics (R. Crisp, Trans.) 2004 Cambridge Cambridge University Press Arthur J Kristjánsson K Harrison T Sanderse W Wright D Teaching Character and Virtue in Schools 2017 London and New York Taylor & Francis Group Banilad Study Center A Project of KALFI (Director). 2018. KALFI LEAD CEBU 2018 AVP. Retrieved April 28, 2022, from https://www.youtube.com/watch?v=5tKWKuD_YOc Berkowitz MW Implementing and Assessing Evidence-Based Character Education Journal of Education 2022 202 2 191 197 10.1177/00220574211026908 Brooks E A New Generation of Wise Thinkers and Good Leaders The Journal of Character & Leadership Development 2021 8 2 32 44 Brooks E Brant J Lamb M How can universities cultivate leaders of character? Insights from a leadership and character development program at the University of Oxford International Journal of Ethics Education 2019 4 2 167 182 10.1007/s40889-019-00075-x Cameron K Responsible Leadership as Virtuous Leadership Journal of Business Ethics 2011 98 1 25 35 10.1007/s10551-011-1023-6 Character Counts! 2022. The six pillars of character. Retrieved June 1, 2022, from https://charactercounts.org/character-counts-overview/six-pillars/ Character.org. 2022. Character & social-emotional development (CSED) national guidelines. Retrieved June 1, 2022, from https://character.org/wp-content/uploads/2022/04/CSED-Natl-Guidelines-2022.pdf Gini A Green RM 10 Virtues of Outstanding Leaders: Leadership and Character 2013 Chichester, West Sussex John Wiley & Sons Greenleaf Center for Servant Leadership. 2021. What is servant leadership? Retrieved April 26, 2022, from https://www.greenleaf.org/what-is-servant-leadership/ Hackett RD Wang G Virtues and leadership: An integrating conceptual framework founded in Aristotelian and Confucian perspectives on virtues Management Decision 2012 50 5 868 899 10.1108/00251741211227564 Havard A Virtuous Leadership 2021 2 Ohio Scepter Publishers KALFI LEAD. 2021. KALFI Mulat exhibit. Retrieved April 28, 2022, from https://sites.google.com/view/kalfimulatexhibit/home Kalfi Leader (Director). 2017. The KALFI LEAD Philippine youth leadership program (case study). Retrieved April 28, 2022, from https://www.youtube.com/watch?v=6GMzJEQfr_U Kalinangan Youth Foundation, Inc. 2019. Kalinangan Youth Foundation, Inc. Home page. KALFI. Retrieved April 25, 2022, from http://kalfi.org/ Kalinangan Youth Foundation, Inc. 2021. Leaders on the good: KALFI LEAD handbook and planner. Kalinangan Youth Foundation, Inc. Kalinangan Youth Foundation, Inc. 2022a. KALFI LEAD Home page. Retrieved April 26, 2022, from https://kalfilead.org/ Kalinangan Youth Foundation, Inc. 2022b. Mentoring program. Retrieved April 28, 2022, from https://kalfilead.org/program-features/mentoring-program/ Kalinangan Youth Foundation, Inc. 2022c. Program features. KALFI LEAD. Retrieved April 13, 2022, from https://kalfilead.org/program-features/ Kalinangan Youth Foundation, Inc. 2022d. Youth leadership training program. Retrieved April 27, 2022, from https://kalfilead.org/program-features/leadership-training/ Lacson, P. 2021. Press release—PMA virtues my guide in navigating PH’s challenges. Senate of the Philippines. Retrieved April 20, 2022, from http://legacy.senate.gov.ph/press_release/2021/1218_lacson3.asp Lamb M Brant J Brooks E How Is Virtue Cultivated? Seven Strategies for Postgraduate Character Development Journal of Character Education 2021 17 1 81 108 Lamb, M., E. M. Dykhuis, S. E. Mendonça, and E. Jayawickreme. 2021b. Commencing character: A case study of character development in college. Journal of Moral Education, 1–23. 10.1080/03057240.2021.1953451 Lickona, M., and T. Davidson. 2005. Smart and good high schools: Integrating excellence and ethics for success in school, work and beyond. Center for the 4th and 5th Rs (Respect & Responsibility)/Washington, D.C.: Character Education Partnership. McGrath, R. E., H. Han, M. Brown and P. Meindl. 2021. What does character education mean to character education experts? A prototype analysis of expert opinions. Journal of Moral Education, 1–19. 10.1080/03057240.2020.1862073 Newstead T Dawkins S Macklin R Martin A We don’t need more leaders – We need more good leaders. Advancing a virtues-based approach to leader(ship) development The Leadership Quarterly 2021 32 5 101312 10.1016/j.leaqua.2019.101312 Pakaluk M Aristotle’s Nicomachean Ethics: An Introduction Cambridge University Press 2005 10.1017/CBO9780511802041 Patton MQ Qualitative Research & Evaluation Methods 2002 Thousand Oaks, London and New Delhi SAGE rappler.com. 2017. Duterte: Harness Filipino heroes’ virtues to fight crime, poverty. Rappler. Retrieved April 20, 2022, from https://www.rappler.com/nation/180219-duterte-national-heroes-day-message-2017-virtues-fight-lawlessness-criminality-poverty/ Rey, A. 2018. Robredo: “A good leader inspires, empowers, listens.” Rappler. Retrieved April 21, 2022, from https://www.rappler.com/nation/207871-robredo-good-leader-inspires-empowers-listens/ Rodriguez Luño, A. 1988. Il concetto di abito elettivo nella definizione delle relazioni tra le virtù morali e la libertà. In Scelta etica (pp. 148–162). Ares. Santiago, M. D. 2015. Press release—The virtues of student leadership and patriotism. Senate of the Philippines. Retrieved April 20, 2022, from https://legacy.senate.gov.ph/press_release/2015/1214_santiago2.asp Shields DL Character as the Aim of Education Phi Delta Kappan 2011 92 8 48 53 10.1177/003172171109200810 Jubilee Centre for Character and Virtues. 2017. A framework for character education in schools. Retrieved June 1, 2022, from https://www.jubileecentre.ac.uk/userfiles/jubileecentre/pdf/Research%20Reports/The_Good_Teacher_Understanding_Virtues_in_Practice.pdf Van Noy, M., , H. James and C. Bedley. 2016. Reconceptualizing learning: A review of the literature on informal learning. 10.7282/00000129 VIA Institute on Character. 2022. Leadership. Retrieved June 1, 2022, from https://www.viacharacter.org/character-strengths/leadership Wang G Hackett RD Conceptualization and Measurement of Virtuous Leadership: Doing Well by Doing Good Journal of Business Ethics 2016 137 2 321 345 10.1007/s10551-015-2560-1	0	PMC9709731	NO-CC CODE	2022-12-01 23:23:08	no	International Journal of Ethics Education. 2022 Nov 30;:1-19	utf-8	null	null	null	oa_other
==== Front Indian Econ Rev Indian Econ Rev Indian Economic Review 0019-4670 2520-1778 Springer India New Delhi 142 10.1007/s41775-022-00142-z Book Review P. G. Babu (ed.): economic policy in COVID-19 times Orient BlackSwan, 2022; ISBN: 9789354422904 Rangarajan C. [email protected] grid.448768.1 0000 0004 1772 7660 Madras School of Economics, Chennai, India 30 11 2022 13 10 11 2022 © The Author(s) under exclusive licence to Editorial Office, Indian Economic Review 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 pmcThe world including India has passed through a difficult time in the last 3 years. In recent memory, this is the first economic crisis trigged by a non-economic factor—a pandemic called COVID-19. It was the Spanish Flu more than hundred years ago that resulted in such a severe devastation. Without cure and vaccination (initially), the impact of the pandemic has been two fold. First was the human tragedy in terms of death and the second was the economic impact in terms of the loss of output caused by the restrictions such as lockdowns imposed to contain the spread of the pandemic. There have been several studies and surveys undertaken to measure and analyze the economic impact of COVID-19. The book under review titled “Economic Policy in Covid-19 Times” edited by P. G. Babu is one such study. But it is unique in that it focuses on the actions taken and their impact in the context of the pandemic in Tamil Nadu. The book contains 18 articles written primarily by the faculty of Madras Institute of Development Studies (MIDS) with each article concentrating on one aspect of the Tamil Nadu economy such as macroeconomic scene, environmental issues, the agricultural sector including agricultural marketing, issues of migrant and informal labour, the tourism sector etc. All the articles also deal with the medium-term and long-term concerns in each of the sectors studied. The book taken as a whole provides a comprehensive picture of the Tamil Nadu economy, its strengths and weaknesses and also outlines the possible strategy for accelerating growth in various sectors. The macroeconomic scene is elaborated by several articles. One point stressed by several writers is that the Indian economy at the start of the COVID-19 episode was not in good shape. We now know that the rate of growth of the Indian economy in 2019–20 was as low as 3.7%. The lockdowns imposed from time to time in 2020–21 had a severe impact on the economy because of the lack of mobility of people and goods and services. The GDP fell by 6.6 percent in 2020–21. In fact, Tamil Nadu is one of the very few states that had a positive growth of State Domestic Product in that year. Besides output loss, another major concern was the dip in labour participation rate. Yet another aspect of the labour problem was the exodus of migrant labour particularly in a state like Tamil Nadu. In fact, the mass exodus of migrant labour from all over Tamil Nadu showed how dependent Tamil Nadu was on “imported” labour. The articles by Shesadri Banerjee and P. G. Babu, Vikas Kumar and Poonam Singh talk of policy options. One major issue was how to contain the recessionary effect while trying to minimize human tragedy. The dilemma was posed as the conflict between livelihood and life. Some balance had to be struck which turned out to be difficult. This also depended upon existing social infrastructure particularly in hospital facilities. The horror through which migrants passed also exposed the need to provide reasonable facilities to migrant labour. The role of fiscal policy in the situation has also been discussed in two chapters. In fact, the rise in government expenditure at a time when revenues were falling resulted in higher fiscal deficits which could be sustained only by liquidity expansion by RBI. The fiscal deficit of the Tamil Nadu government also went up. It is the liquidity expansion which has come home to roost. The flare up in inflation that not only India but also USA and UK are facing now is a direct consequence of the liquidity explosion. The Indian economy grew fast enough in 2021–22 to compensate the loss of income in the previous year. But we are where we were two years ago. The expectation that 2022–23 will be a normal year was shattered by the Russia–Ukraine war. The COVID-19 episode not only underlined the need of how to prepare ourselves to meet a health emergency but also focus on some fundamental issues with respect to various sectors of the economy. L. Venkatachalam in his article “Reforms in the Agriculture Sector in the Post-Covid-19 Era” talks about market reforms including contract farming, emphasis on more value added products, faster procurement and flexible crop loans. Sivasubramaniyan’s article focuses on tank and canal irrigation in Tamil Nadu. In the agricultural sector, there is an interesting article by P. G. Babu, A. Ganesh Kumar and Chandan Kumar on the Controversial Farm Acts which though withdrawn contain according to the authors some relevant ideas for reforms. In the chapter ‘Environmental Reforms in Tamil Nadu after Covid-19’, P. Durairasu and L. Venkatachalam address the need for proper ‘green accounting’ and also advocate a move away from command and control system. Ajit Menon and Maarten Bavinck in their article besides talking about the measures taken by the government to provide relief to small and medium fishermen spell out medium term policy measures to promote the fisheries sector. C. Veeramani and P. G. Babu in their article argue for “Assemble in Tamil Nadu for the World”. It raises the question—why not make in Tamil Nadu for the world. In any case, we need not restrict ourselves to follow only one path. The choice must depend upon the good or commodity chosen. M. Vijayabaskar in his article talks about the garment industry and offers several recommendations for improving the quality and productivity of the industry. A.R. Venkatachalapathy examines the conditions of the Tamil publishing industry, an area which is rarely looked into. The tourism sector is one that had been badly hit by COVID-19. The problems of this industry are looked at by Krishanu Pradhan. The migration trends and hardships of migrant workers is a subject of analysis by K. Jafar and A. Kalaiyarasan. The plight of the domestic workers in general and in particular at the time of COVID-19 is the subject of analysis by S. Anandhi and E. Deepa. P. G. Babu and Chandan Kumar in their article on infrastructure development draw attention to the problems that arise in the PPP model. Karen Coelho and A. Srivathsan in their article look at the possibilities and difficulties in affordable housing. In a pandemic situation decentralized medical attention is critical. In this context, what the role of Panchayati Raj institutions can be is the subject matter of the article by Kripa Ananthpur. Gayathri Balagopal and M. Vijayabaskar in their article discuss social protection systems at each stage of the life cycle and then examine systems which cut across the life cycle like health and sanitation and food security. The food security system includes programmes of various types introduced by central and state governments. Umanath Malaiarasan analyzes the food security issues during COVID-19 and suggests various policy measures to get over difficulties. All the articles were perhaps written before the middle of 2021. That was the first year after the outbreak of COVID-19 when the economy suffered. In the next year, the economy recovered. But the human tragedy was high because of the relaxation of controls. The virus was also severe and it exposed the inadequacies in our health infrastructure. What we need is a companion volume to the present study by MIDS, depicting what happened in the following year. Some question whether we had been overly restrictive in the first year. These people argue that had we been less severe with lockdowns, the output loss in the first year would have been less. This is wisdom by hindsight! At the time of outbreak of COVID-19, no one including the medical profession was fully aware of the nature of the pandemic and how to control it. We could not gamble with human life. We have learnt much however at great cost on what needs to be done on our health infrastructure. As far as growth is concerned, we have lost two years. The loss of output measured from the trend line is even higher. The book under review is useful because almost every article contains suggestions for accelerating growth in the medium term. Tamil Nadu is well ahead of many states in terms of physical and social infrastructure. The goal of a trillion dollar economy is achievable as far as Tamil Nadu is concerned. It will take a minimum of 10 or 11 years and would require a real rate of growth of 8–9%. Fiscal stability is also important to achieve this goal. Tamil Nadu should get back to maintaining a fiscal deficit of 3% or less of State Domestic Product. The book does not address this issue directly. On the whole, a timely book with a lot of insights on Tamil Nadu economy. Funding The author did not receive support from any organization for the submitted work. Declarations Conflict of interest The author has no conflicts of interest to declare that are relevant to the content of this article. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.	0	PMC9709733	NO-CC CODE	2022-12-01 23:23:08	no	Indian Econ Rev. 2022 Nov 30;:1-3	utf-8	Indian Econ Rev	2,022	10.1007/s41775-022-00142-z	oa_other
==== Front Anaesthesiologie Anaesthesiologie Die Anaesthesiologie 2731-6858 2731-6866 Springer Medizin Heidelberg 36449054 1230 10.1007/s00101-022-01230-8 CME Extrakorporale Membranoxygenierung und Hämodynamik Die Therapie ist nicht nur des Herzens Freund Extracorporeal membrane oxygenation and hemodynamicsTherapy is not only a friend of the heart Haas Annika [email protected] 1 Busjahn Christoph 1 Crede David 1 Kilger Erich 2 Reuter Daniel A. 1 1 grid.413108.f 0000 0000 9737 0454 Klinik und Poliklinik für Anästhesiologie, Intensivmedizin und Schmerztherapie, Universitätsmedizin Rostock, Schillingallee 35, 18057 Rostock, Deutschland 2 grid.5252.0 0000 0004 1936 973X Klinik für Anästhesiologie, Ludwig-Maximilians-Universität München, München, Deutschland Wissenschaftliche Leitung T. Fuchs-Buder, Nancy A.R. Heller, Augsburg M. Rehm, München M. Weigand, Heidelberg A. Zarbock, Münster 30 11 2022 115 8 11 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. Extrakorporale Unterstützungssysteme von Herz und Lunge werden bei kardialem, pulmonalem oder auch kardiopulmonalem Versagen eingesetzt. Jedoch verhält sich weder die rein lungenunterstützende venovenöse extrakorporale Membranoxygenierung (vv-ECMO) noch die venoarterielle (va‑)ECMO hämodynamisch „inert“ gegenüber dem patienteneigenen Herz-Kreislauf-System. Der Erfolg einer ECMO-Therapie hängt entscheidend vom Monitoring vor und während der Durchführung sowie dem pathophysiologischen Verständnis für die hämodynamischen Veränderungen, die während der Therapie auftreten, ab. Der vorliegende Beitrag erläutert gezielt diese „Begleiterscheinungen“ und diskutiert fundamentale Aspekte der Herz-Kreislauf-Physiologie und ihr spezifisches Zusammenspiel mit einer ECMO-Therapie. Extracorporeal support systems for the heart and lungs are employed for cardiac, pulmonary and also cardiopulmonary failure; however, neither the pure lung support by venovenous extracorporeal membrane oxygenation (vvECMO) nor the venoarterial (va) ECMO behave in a hemodynamically inert manner with respect to the patientʼs own cardiovascular system. The success of ECMO treatment is decisively dependent on monitoring before and during the execution and the pathophysiological understanding of the hemodynamic changes that occur during treatment. This article explicitly elucidates these “concomitant phenomena” and discusses fundamental aspects of cardiovascular physiology and the specific interplay with ECMO treatment. Schlüsselwörter Herz Lunge Oxygenierung Decarboxylierung Pathophysiologie Keywords Heart Lungs Oxygenation Decarboxylation Pathophysiology ==== Body pmcLernziele Nach der Lektüre dieses Beitragskönnen Sie die wesentlichen Aspekte der Herz-Kreislauf-Physiologie und ihr spezifisches Zusammenspiel mit der extrakorporalen Membranoxygenierung (ECMO) benennen. sind Sie in der Lage, die hämodynamischen „Begleiterscheinungen“ einer ECMO-Therapie einzuordnen. können Sie die technische Durchführung der verschiedenen ECMO-Konfigurationen beschreiben. wissen Sie, welche Monitoring-Verfahren zu welchem Zeitpunkt eingesetzt werden sollten. Einleitung Extrakorporale UnterstützungssystemeExtrakorporale Unterstützungssysteme von Herz und Lunge werden bei kardialem, pulmonalem oder auch kardiopulmonalem Versagen implantiert. Sie geben dem behandelnden Team Zeit, um die Diagnostik und Therapie der zugrunde liegenden Erkrankung einzuleiten, bzw. Zeit, um weitreichendere Entscheidungen zu treffen („bridge to recovery“ oder als „bridge to decision“ bzw. „bridge to transplant“ und „bridge to destination“). Die erste Generation extrakorporaler Unterstützungssysteme für Herz und Lunge wurde in den 1970er-Jahren implantiert – damals als venoarterielle extrakorporale Membranoxygenierung (va-ECMO). Es dauerte aber mehrere Jahrzehnte, um Erfahrungen in dieser Technologie zu sammeln und das Verfahren so zu verbessern, dass der Einsatz mit einem gebesserten Outcome für die Patienten verbunden war [1]. Gerade in den letzten Jahren, und insbesondere im Rahmen der COVID-19-Pamdemie, hat die Anwendung der ECMO-Therapie, sowohl als venovenöse (vv-) als auch als va-Konfiguration, erheblich zugenommen. Jedoch stoßen auch diese Therapiemaßnahmen – gerade aufgrund der kardiovaskulären Grunderkrankungen und der akuten Ausgangssituation der Patienten häufig an ihre Grenzen [2, 3]. Auch können, insbesondere aufgrund von unterschiedlichen Kanülierungsstellen und Kanülenarten die durch das extrakorporale System erzeugten hämodynamischen Effektehämodynamischen Effekte stark variieren. Deshalb ist es bereits zur Indikationsstellung und zum Zeitpunkt der Implantation eines kardialen Unterstützungssystems, aber auch für den weiteren klinischen Verlauf, entscheidend, diese hämodynamischen Effekte einer ECMO-Therapie auf den Patientenkreislauf genau zu kennen und zu verstehen [4]. Eine ECMO-Therapie verhält sich bei Weitem nicht hämodynamisch „inert“ gegenüber dem patienteneigenen Herz-Kreislauf-System; dies gilt sowohl für die rein lungenunterstützende vv-ECMO als auch für die va-ECMO. Es ist nicht Ziel dieser Arbeit, einen vollständigen Überblick über Indikationen, Kontraindikationen und technisches Vorgehen der ECMO-Therapie zu geben; vielmehr sollen gezielt die hämodynamischen „Begleiterscheinungen“ einer ECMO-Therapie erläutert und diskutiert werden. Hierzu ist es unabdingbar, sich mit fundamentalen Aspekten der Herz-Kreislauf-Physiologie und ihrem spezifischen Zusammenspiel mit einer ECMO-Therapie auseinanderzusetzen. Venovenöse extrakorporale Membranoxygenierung Besteht im schweren LungenversagenLungenversagen eine therapierefraktäre Hypoxämie oder Hyperkapnie mit Acidose, kann die Implantation einer vv-ECMO erwogen werden. Die vv-ECMO drainiert venöses Blut, übernimmt den GasaustauschGasaustausch und pumpt das oxygenierte und decarboxylierte Blut zurück in den venösen Kreislauf. Das heißt, dass die vv-ECMO eine reine Oxygenierungs- und Decarboxylierungsunterstützung ist. Ein Einfluss auf die Hämodynamik ist nicht primäres Wirkungsziel. Jedoch bestehen auch bei der reinen vv-ECMO wichtige Wechselwirkungen mit der Herz-Kreislauf-Funktion, die beachtet werden müssen. Technische Durchführung Die vv-ECMO besteht aus einer zu- und einer rückführenden Kanüle, einer Zentrifugalpumpe, einem Oxygenator und der Steuereinheit. Das Blut kann entweder über die V. cava superior oder die V. cava inferior drainiert werden und sollte möglichst über die jeweils andere Hohlvene rückgeführt werden. Platzierung und Größe der abführenden Kanüle entscheiden über den BlutflussBlutfluss , der über die ECMO generiert werden kann. Gemäß dem Hagen-Poiseuille-GesetzHagen-Poiseuille-Gesetz (V˙=π×r4×Δp8×η×l) sollte die Kanüle so kurz, aber so großkalibrig wie möglich sein. Die klassische Variante der Anlage umfasst die Punktion der rechtsseitigen V. femoralis, um die abführende Kanüle (21–25 Fr) in der V. cava inferior zu platzieren, und die Punktion der rechtsseitigen V. jugularis interna (Kanüle 17–21 Fr), um das Blut in die V. cava superior zurückzuführen [5]. Die Spitzen der beiden Kanülen sollten etwa 20 cm auseinander liegen, um eine RezirkulationRezirkulation des oxygenierten Blutes so gering wie möglich zu halten (s. Phänomen der Rezirkulation). Bei der vv-ECMO besteht auch die Möglichkeit, zu- und abführende Kanüle als DoppellumenkanüleDoppellumenkanüle (23–31 Fr) in ein Gefäß einzuführen. Diese Doppellumenkanüle wird über die rechte V. jugularis interna so platziert, dass sie das Blut aus der V. cava superior drainiert und das oxygenierte Blut weiter distal direkt in den rechten Vorhof in Richtung Trikuspidalklappe reinfundiert [6]. Die Frage, welcher Patient welche Art der KanülierungKanülierung erhält, sollte immer individuell, nach jeweiliger patientenspezifischer Risiko-Nutzen-Abwägung entschieden werden. Generell gilt, dass der Fluss über die Kanülen bei jeder Kanülierungsvariante der vv-ECMO auch von der patienteneigenen Kreislauffunktion abhängt und regelmäßig in der Zusammenschau mit der Makrohämodynamik überwacht werden muss (s. unten, [6]). Hämodynamik Lungenversagen Im akuten Lungenversagen kann eine pulmonale Hypertensionpulmonale Hypertension aufgrund von Mikrothromben, von arteriellem Remodeling und einer pulmonalen Vasokonstriktion entstehen, wobei die pulmonale Vasokonstriktionpulmonale Vasokonstriktion (hypoxische pulmonale Vasokonstriktion, HPV) durch die Trias Hypoxie, Inflammation und Acidose hervorgerufen wird. Die pulmonale Hypertension erhöht die rechtskardiale Nachlastrechtskardiale Nachlast , die weiter durch die ÜberdruckbeatmungÜberdruckbeatmung gesteigert wird: Durch diesen erhöhten Widerstand im pulmonalvaskulären Stromgebiet nimmt der Blutfluss weiter ab, was zur Ausweitung der West-Zone 1 (Druck in Alveole > Druck in Arteriole > Druck in Venole) und klinisch einer weiteren Verschlechterung der Oxygenierung führt [7]. Gleichzeitig nehmen die Druck- und konsekutiv die Volumenbelastung des rechten Ventrikels zu und beeinträchtigen seine Funktion zunehmend. Besonders ein erhöhter „driving pressure“„driving pressure“ (Inspirationsdruck − „positive endexpiratory pressure“ [PEEP]), aber auch ein hoher PEEP > 15cm H2O, ist direkt mit einer Verschlechterung der systolischen rechtsventrikulären Funktion assoziiert [8, 9]. Der dünnwandige rechte Ventrikel ist – im Vergleich zum linken – hochgradig vorlastabhängig und nachlastsensibel; das heißt: Wenn die Vorlast durch verminderten venösen Rückfluss sinkt und insbesondere gleichzeitig die kardiale Nachlast steigt, wächst die Gefahr der Dekompensation des rechten Ventrikels signifikant. Deshalb können sowohl das Krankheitsbild des akuten Lungenversagens als auch seine Therapie mithilfe der Beatmung rasch zum akuten Cor pulmonaleCor pulmonale und zum Rechtsherzversagen führen [10]. Merke Das Lungenversagen, aber auch die Therapie mit Beatmung, kann ein Rechtsherzversagen bedingen. Unter laufender vv-ECMO Die vv-ECMO verbessert die Oxygenierung und Decarboxylierung und erlaubt damit die Reduktion der Beatmungsinvasivität mit entsprechender Verringerung ventilationsassoziierter Lungenschäden [11]. Durch den Einsatz der vv-ECMO ergeben sich jedoch auch wichtige Effekte auf bzw. Interaktionen mit der Hämodynamik: Einfluss auf die rechtsventrikuläre Nachlast. Wie oben beschrieben, gehen eine durch die vv-ECMO schlagartig verbesserte Oxygenierung und Decarboxylierung mit Abnahmen der pulmonalen Vasokonstriktion und der rechtsventrikulären Nachlast sowie einer Verbesserung des pulmonalen Blutflusses einher. Die vv-ECMO stellt eine Therapiemaßnahme des akuten Cor pulmonale bzw. eines sekundären Rechtsherzversagenssekundären Rechtsherzversagens dar [10, 12, 13]. Generell gilt: Den besten therapeutischen Effekt erzielt die vv-ECMO, wenn die größtmögliche Menge unzureichend gesättigten venösen Blutes drainiert, oxygeniert und decarboxyliert wird. Die hämodynamische Effizienz der vv-ECMO wird über die Größe der abführenden Kanüle, den Blutfluss, der darüber generiert werden kann, und die PräoxygenatorsättigungPräoxygenatorsättigung definiert [6]. Phänomen der Rezirkulation. Generell besteht bei einer vv-ECMO immer das Risiko der „Rezirkulation“. Das bedeutet, dass oxygeniertes Blut aus der zuführenden Kanüle direkt über die abführende Kanüle drainiert wird und nicht in den Patientenkreislauf gelangt. Dadurch wird die Effizienz der vv-ECMO deutlich reduziert, und die patienteneigene SauerstoffsättigungSauerstoffsättigung sinkt. Das Ausmaß der Rezirkulation hängt von der Position der Kanülen, dem Fluss der vv-ECMO sowie dem Volumenstatus und der kardialen Funktion des Patienten ab [13]. Sind die Kanülen der vv-ECMO zu nah aneinander platziert, steigt das Risiko der Rezirkulation; dies gilt ebenso bei zunehmendem Fluss über die vv-ECMO. Steigen der intrathorakale oder der intraabdominelle Druck des Patienten an, kann es durch den verminderten venösen Rückfluss zum Ansaugen der Drainagekanülen kommen, es kann aber auch anteilig mehr reinfundiertes Blut über die ECMO drainiert werden, sodass auch die Rezirkulation zunimmt [13]. Dies geschieht auch bei anhaltend hohem pulmonalarteriellem Widerstand und abnehmendem Herzzeitvolumen (HZV): Das rechtsventrikuläre enddiastolische Volumen steigt, und das Blut wird zunehmend über die ECMO drainiert. Die klinische Relevanz der Rezirkulation korreliert mit dem Ausmaß der effektiven Oxygenierungsstörung des Patienten. Je abhängiger dieser von der vv-ECMO ist, desto relevanter ist der Grad der Rezirkulation. Aktuell gibt es noch keine, im klinischen Alltag praktikable Methodik, um die Rezirkulation genau zu quantifizieren [13]. Es empfiehlt sich, den Verlauf der peripheren arteriellen Sauerstoffsättigung und der Präoxygenatorsättigung engmaschig gemeinsam zu beurteilen: Sinkt die arterielle Sauerstoffsättigung, während die Präoxygenatorsättigung steigt, ist es wahrscheinlich, dass die Rezirkulation zunimmt. Merke Eine Rezirkulation bei vv-ECMO sollte vermieden werden, da diese die Effizienz der ECMO reduziert. Monitoring Unter einer vv-ECMO-Therapie sollte die Hämodynamik kontinuierlich überwacht werden. Dazu gehört grundlegend die invasive und kontinuierliche arterielle Blutdruckmessungarterielle Blutdruckmessung ; dies ermöglicht auch regelmäßige BlutgasanalysenBlutgasanalysen (BGA). Auch eine kontinuierliche Messung des zentralen Venendruckszentralen Venendrucks (ZVD) mit Fokus auf dem ZVD-Trend ist für das Erkennen akuter rechtsventrikulärer Komplikationen klar zu empfehlen. Thermodilutionsverfahren, sowohl zur Bestimmung von HerzzeitvolumenHerzzeitvolumen oder VorlastvoluminaVorlastvolumina , gelten als unter einer vv-ECMO nicht verwendbar [14]. Die arterielle Pulskonturanalyse zur Messung des HZV ist zwar prinzipiell durch die vv-ECMO nicht affektiert, allerdings lässt sie sich nicht durch Thermodilution kalibrieren. Einen besonderen Stellenwert nimmt, auch wenn sie kein Monitoring-Verfahren ist, die wiederholte EchokardiographieEchokardiographie ein. Dies gilt insbesondere zur Beurteilung der rechtsventrikulären Funktion. Sonographisch ist die Größe des rechten Ventrikels ein guter Verlaufsparameter: Ein Verhältnis der rechtsventrikulären enddiastolische Fläche zur linksventrikulären enddiastolischen Fläche („right ventricular enddiastolic area/left ventricular enddiastolic area“, RVEDA/LVEDA) von 0,6–1 ist Zeichen einer moderaten rechtsventrikulären Dilatationrechtsventrikulären Dilatation , ein RVEDA/LVEDA > 1 Zeichen einer schweren Dilatation. Außer der rechtsventrikulären sollte ebenso die linksventrikuläre Funktion überwacht werden. Auch zur Therapieinitiierung des extrakorporalen Unterstützungsverfahrens ist die Echokardiographie mit Bestimmung der kardialen Funktionkardialen Funktion unerlässlich: Bei Patienten mit erhaltener linksventrikulären Funktion und einer Verschlechterung der rechtsventrikulären Funktion, die alleinig auf dem Lungenversagen beruht, ist eine vv-ECMO von Benefit [15]. Im primären Rechtsherzversagenprimären Rechtsherzversagen , oder wenn der linke Ventrikel ebenfalls betroffen ist, ist die va-ECMO das bevorzugte extrakorporale Therapieverfahren. Unter einer ECMO-Therapie sollte das HZV sonographisch ermittelt werden – die Pulskonturanalyse oder Thermodilutionsverfahren stellen im Rahmen einer ECMO-Therapie keine genauen Messverfahren dar [7, 16]. Rechtskardiale Dekompensation Trotz verbesserter Oxygenierung kann es während einer vv-ECMO-Therapie zum Kreislaufversagen des Patienten kommen. Besonders kritisch ist auch hier die rechtsventrikuläre Funktion zu beobachten. Nimmt das enddiastolische rechtsventrikuläre Volumen weiter zu, kann dies ein Pumpversagen des rechten Ventrikels mit konsekutiver venöser Kongestionvenöser Kongestion auslösen. Wichtige konservative Therapiemöglichkeiten sind: Optimierung der Beatmung mit Reduktion des PEEP und möglichst geringen Tidalvolumina, intravasale Volumenrestriktion sowie eine differenzierte Katecholamin- und Vasopressorentherapie. Der Einsatz von Inodilatoren und inhalativem Stickstoffmonoxid (NO) bzw. inhalativem Prostaglandin gilt zwar als Rescue-Therapie unter vv-ECMO, erfolgt klinisch aber teilweise schon früher [17]. Sind all diese Möglichkeiten ausgeschöpft, sollte das Herz durch die Insertion einer zusätzlichen arteriellen ECMO-Kanüle entlastet werden [12]. Der reine Wechsel zur va-ECMO kann jedoch zum Nord-Süd-SyndromNord-Süd-Syndrom , auch Harlekin-Syndrom genannt, führen (s. Abschn. „Venoarterielle extrakorporale Membranoxygenierung“): Das oxygenierte Blut wird über die Femoralarterie zurückgeführt und erreicht hauptsächlich die untere Extremität; die lebenswichtigen Organe (Herz, Gehirn) werden weiterhin mit schlecht oxygeniertem Blut aus der Pulmonalvene perfundiert [18]. Es zeigte sich, dass bei Anwendung der vv-ECMO im Fall des Acute Respiratory Distress Syndrome (ARDS) mit anhaltendem kardiogenen Schock bzw. Rechtsherzversagen die venoarteriovenöse Konfigurationvenoarteriovenöse Konfiguration (vav-ECMO) von Vorteil ist: Das venöse Blut wird der V. cava inferior entnommen, oxygeniertes Blut wird sowohl über die Femoralarterie als auch über die V. cava superior zurückgeführt. Somit erreicht gut oxygeniertes Blut auch im akuten Lungenversagen den linken Ventrikel. Die Risiken von va- und vav-ECMO sind vergleichbar, da das größte Risiko durch die arterielle Kanüle verursacht wird [19]. Diese HybridmodelleHybridmodelle mit insgesamt 3 ECMO-Kanülen sind zwar noch keine routinemäßig genutzte Verfahren, konnten aber in einigen Zentren, in denen sie bereits genutzt werden, gute Ergebnisse erzielen [20]. Venoarterielle extrakorporale Membranoxygenierung Die va-ECMO wird im Gegensatz zur vv-ECMO genutzt, um entweder die Pumpfunktion des Herzens oder auch die kombinierte Funktion von Herz und Lunge, d. h. Pumpfunktion und Oxygenierung/Decarboxylierung, zu ersetzen. Die va-ECMO kann also den Kreislauf ohne aktive Herzfunktion aufrechterhalten – kann in ihrer Funktionsweise aber durch eine mögliche residuelle patienteneigene Herzfunktion beeinflusst werden. Im Umkehrschluss ist dies von noch größerer Relevanz: Die va-ECMO beeinflusst die körpereigene Herzfunktion z. T. erheblich. Sie ist also hämodynamisch nicht inert, sondern kann beim „versagenden Herzen“ oder auch bei anderen lebenswichtigen Organen Entlastung, aber auch aktiv Belastung bzw. Schaden verursachen. Um diese Mechanismen besser zu verstehen, ist es entscheidend, sich vorab differenziert mit der kardialen Physiologie auseinanderzusetzen. Grundlagen der linksventrikulären Mechanik Die HerzaktionHerzaktion kann in die Phase der Anspannung vor dem Auswurf (isovolumetrische Kontraktion), die Phase der Austreibung (isotonische Kontraktion), die Phase der Erschlaffung (isovolumetrische Relaxation) und die Phase der Füllung unterteilt werden. Die ersten beiden gehören zur Systole, die letzten beiden zur Diastole. Diese Phasen können als „Arbeitsdiagramm“„Arbeitsdiagramm“ des Herzens im Druck-Volumen-DiagrammDruck-Volumen-Diagramm des linken Ventrikels abgebildet werden. Hierbei wird das intraventrikuläre Volumen auf der x‑Achse gegen den intraventrikulären Druck auf der y‑Achse aufgetragen. Die Eckpunkte des Diagramms kennzeichnen die jeweiligen Endpunkte der oben genannten Phasen ([21]; Abb. 1). Größe und Verlauf des Arbeitsdiagramms werden durch die beiden Kennlinien, die endsystolische Druck-Volumen-Kurveendsystolische Druck-Volumen-Kurve (ESPVR) und die enddiastolische Druck-Volumen-Kurveenddiastolische Druck-Volumen-Kurve (EDPVR) – auch als Ruhedehnungskurve bezeichnet, gebildet. Die ESPVR verläuft annähernd linear; die Steigung entspricht der endsystolischen Elastanceendsystolischen Elastance (Ees). Eine Linie, gezogen vom enddiastolischen Volumen (EDV) auf der x‑Achse zum endsystolischen Druck (pes) auf der ESPVR, reflektiert die Nachlastlinie – ihre Steigung ist die effektive arterielle Elastancearterielle Elastance (Ea). Das EDV steht für die kardiale Vorlast. Die Ea ist das Maß für die kardiale Nachlast und zeigt die hämodynamischen Eigenschaften des Gefäßsystems, gegen die der linke Ventrikel kontrahiert. Je steiler diese Linie verläuft, desto größer die Nachlast, also der Widerstand, gegen den der Ventrikel anarbeiten muss (Ea = „Nachlastlinie“; Abb. 2). Der pes, also der Schnittpunkt der „Nachlastlinie“ mit der ESPVR, korreliert mit dem arteriellen Mitteldruckarteriellen Mitteldruck („mean arterial pressure“, MAP), der in den peripheren Arterien gemessen werden kann (MAP ≈ 0,9 ⋅ pes, [22]). Damit hängt die Fläche des Arbeitsdiagrammes von den beiden Determinanten, der kardialen Vorlast, also dem enddiastolischen Volumen und der kardialen Nachlast, der Elastance bzw. dem arteriellen Gefäßwiderstand ab. Je größer die Fläche, desto größer das SchlagvolumenSchlagvolumen , aber auch desto größer die Druck-Volumen-Arbeit und der myokardiale Sauerstoffverbrauch. Merke Die Fläche des Druck-Volumen-Diagramms entspricht der Herzarbeit und korreliert mit dem myokardialen Sauerstoffverbrauch. Der Verlauf der EDPVR oder Ruhedehnungskurve ist nicht linear und beschreibt die passiven mechanischen Eigenschaften des erschlafften Herzmuskels in der Diastole. Dieser Kurvenverlauf hängt von der SteifigkeitSteifigkeit des linken Ventrikels ab (Steifigkeit = ∆p/∆V [linksventrikuläre Druckänderung/Volumenänderung]); bei einem definierten enddiastolischen Druck gilt: Je steifer der linke Ventrikel ist, desto steiler ist die Ruhedehnungskurve – und desto kleiner ist das enddiastolische Volumen, da sich der Ventrikel schlechter füllt. Der ESPVR-Verlauf ist linear und hängt im Wesentlichen von der linksventrikulären Kontraktilitätlinksventrikulären Kontraktilität ab. Sinkt diese, flacht die Kurve ab. Zumeist geht eine Abnahme der Kontraktilität mit einer gleichzeitigen Zunahme des endsystolischen Volumens einher, sodass die ESPVR sowohl abflacht als auch nach rechts verschoben wird ([21]; Abb. 2). Damit wird deutlich, welche klinischen Zustände zu einer veränderten Fläche des kardialen Arbeitsdiagrammes sowie der Arbeitsbelastung und des Sauerstoffbedarfs des linken Ventrikels führen [22].Volumenbelastung – Vorlast steigt (Abb. 3): Durch eine isolierte Zunahme der Vorlast steigt das enddiastolische Volumen, und der Punkt auf der Ruhedehnungskurve verschiebt sich nach rechts – die Fläche des Arbeitsdiagramms wird größer, und es resultiert ein erhöhtes Schlagvolumen. Druckbelastung – Nachlast steigt (Abb. 4): Der endsystolische Druck steigt aufgrund einer steiler werdenden Nachlastlinie; damit wird das Arbeitsdiagramm höher und schmaler; das Schlagvolumen sinkt. Bei erhaltener diastolischer Funktion verschiebt sich das Arbeitsdiagramm nach rechts, und das initiale Schlagvolumen wird wiederhergestellt. Abnahme der kardialen Kontraktilität – systolische Dysfunktion: Die Steigung der ESPVR flacht ab. Damit sinken auch der pes und sein peripheres Korrelat, der MAP. Das kardiale Schlagvolumen sinkt. Diastolische Dysfunktion: Eine diastolische Dysfunktion führt zu einer Zunahme der linksventrikulären Steifigkeit [23]. Es resultiert ein steilerer Anstieg der Ruhedehnungskurve, sodass der enddiastolische Druck bei einem definierten enddiastolischen Volumen deutlich zunimmt [24]. Um den Einfluss der va-ECMO auf die kardiale Funktion zu veranschaulichen, sollen diese Überlegungen von einem klinischen Fall, der an die Geschichte einer realen Patientin angelehnt ist, begleitet werden: Fallbeispiel. Eine 72-jährige Patientin wird vom Rettungsdienst mit seit 3 Tagen persistierendem Thoraxschmerz und Kurzatmigkeit in die zentrale Notaufnahme gebracht. Die Patientin ist hypotonisch, mit einem Blutdruck von 100/65 mm Hg, tachykard, und mithilfe der Pulsoxymetrie wird eine Sauerstoffsättigung von 90 % gemessen. Das EKG zeigt eine diffuse ST-Strecken-Senkung; laborchemisch werden eine Troponin-T(TnT)-Konzentration von 2,5 µg/l (Norm: < 0,4 µg/l) und eine „Brain-natriuretic-peptide“(BNP)-Konzentration von 837 ng/l (Norm: < 100 ng/l) festgestellt [25]. Die transthorakale Echokardiographie (TTE) zeigt Wandbewegungsstörungen des linken Ventrikels mit einer hochgradigen Mitralklappeninsuffizienz und Prolaps des posterioren Mitralklappensegels sowie einer linksventrikulären Ejektionsfraktion (LVEF) von 45 %. Der linksventrikuläre enddiastolische Diameter (LVEDD) ist mit 65 mm erhöht. Die rechtsventrikuläre Funktion ist ebenfalls beeinträchtigt, mit einer „tricuspid anular plane systolic excursion“ (TAPSE) von 14 mm. In der Koronarangiographie wird eine akute Läsion des R. circumflexus der linken Koronararterie gesehen. Unter der initialen konservativen Therapie mit zusätzlicher Implantation einer intraaortalen Ballonpumpe (IABP) zur Nachlastsenkung des linken Ventrikels verschlechtert sich der klinische Zustand der Patientin mit eintretendem kardiogenem Schock und beginnendem Multiorganversagen rapide. Aufgrund des hohen operativen Risikos für eine Bypass-Operation mit Mitralklappenrekonstruktion wird ein interdisziplinärer Heart-Team-Beschluss zur Implantation einer va-ECMO gefällt. Technische Durchführung Die arterielle und venöse Kanülierung erfolgen – außerhalb des Herz-OP, peripher [26]. Der am meisten genutzte Zugangsweg ist die V. und A. femoralis; alternativ können die abführende Kanüleabführende Kanüle über die V. jugularis und die zuführende Kanülezuführende Kanüle über die A. axillaris, A. subclavia oder den Truncus brachiocephalicus implantiert werden, wobei diese arteriellen Kanülen oft einen chirurgischen Zugangsweg benötigen [27]. Zentrale KanülierungenZentrale Kanülierungen des rechten Vorhofs und der Aorta ascendens erfolgen üblicherweise nach kardiochirurgischen Eingriffen. Das desoxygenierte Blut wird immer mithilfe einer ZentrifugalpumpeZentrifugalpumpe in den OxygenatorOxygenator transportiert, dort erwärmt und oxygeniert bzw. decarboxyliert. Das arterialisierte Blut wird dem Patienten über die arterielle Kanüle in die Aorta zurückgeführt. Mit den normalerweise verwendeten venösen Kanülen mit einem Außendurchmesser von 21–28 Fr bzw. den arteriellen Kanülen mit einen Außendurchmesser von 17–21 Fr können Blutflüsse von 2,5–7 l/min erreicht werden [22]. Fortsetzung des Fallbeispiels. Die Patientin erhält eine va-ECMO, mit üblicher peripherer Kanülierung über die Femoralvene und -arterie. Unter einem Blutfluss von 4 l/min bessert sich die Schocksymptomatik der Patientin rasch, und die Blutdruckamplitude in der rechten Radialarterie steigt. Jedoch zeigt die periphere Sättigung an der rechten Hand weiterhin deutlich erniedrigte Werte um 84 %; auch die BGA zeigt Zeichen einer Hypoxämie. Hämodynamik Anhand der Kanülierungsorte ist ersichtlich, dass das Blut unter va-ECMO nicht seinen normalen physiologischen Weg fließt, sondern das arterialisierte Blut auch retrograd strömen muss. Dies ist besonders bedeutsam, wenn die rückführende Kanüle in der A. femoralis platziert wird [17]. Dieses arterialisierte ECMO-Blut trifft in der Aorta auf natives linksventrikuläres Blut, sofern der Patient noch ein eigenes HZV generiert. Je nachdem, wie groß der Anteil des linksventrikulären Auswurfs und wie hoch der ECMO-Fluss ist, liegt dieser Punkt der „Wasserscheide“„Wasserscheide“ in der proximalen oder distalen Aorta thoracalis ([28, 29]; Abb. 5). Besteht neben einer kardialen Schädigung auch eine pulmonale Funktionsstörung (z. B. ein Lungenödem), kann es bei peripherer va-ECMO mit residuellem linksventrikulären Auswurf zur Hypoxämie der oberen Körperhälfte, dem Nord-Süd-Syndrom, kommen. Die vital wichtigen Gefäße der Perfusion des Kopfes und des Herzens (Karotiden, Koronararterien) werden mit schlechter – oftmals unzureichend – oxygeniertem Blut perfundiert als der Rest des Körpers (Abb. 6). Damit besteht trotz laufender ECMO das Risiko der zerebralen Hypoxiezerebralen Hypoxie und des hypoxischen kardialen Pumpversagens. Deshalb ist es wichtig, den Sauerstoffgehalt des Blutes, das Gehirn und Herz perfundiert, mithilfe von BGA der rechten A. radialis zu überprüfen, um regionale Hypoxien zu vermeiden. Auch eine kontinuierliche zerebrale Nah-Infrarot-SpektroskopieNah-Infrarot-Spektroskopie (NIRS) kann im Verlauf regionale zerebrale Ischämien aufzeigen [17]. Auf technische Fehler im oder am Oxygenator, die eine Rückfuhr von desoxygeniertem Blut zur Folge haben können, soll im vorliegenden Beitrag nicht eingegangen werden. Merke Bei femoraler Kanülierung im Rahmen der va-ECMO und beeinträchtigter Lungenfunktion kann es zu zerebralen und kardialen Hypoxien kommen. Tritt dennoch ein Harlekin-Syndrom auf, sollte durch konservative Maßnahmen versucht werden, die Lungenfunktion zu verbessern bzw. den linksventrikulären Auswurf nicht artifiziell zu steigern – der Einsatz von InotropikaInotropika unter va-ECMO-Therapie ist generell nicht zwingend indiziert, im Fall eines Harlekin-Syndroms sollten sie aber reduziert werden. Kann auch dadurch die Oxygenierung des kardial ausgeworfenen Blutes nicht verbessert werden, besteht die Möglichkeit, die va-ECMO in eine vav-ECMO zu konvertieren [12, 20]. Durch die zusätzliche venöse Kanüle (in die V. jugularis oder V. femoralis) wird dem rechten Herzen bereits oxygeniertes Blut zugeführt, das über den linken Ventrikel ausgeworfen werden kann. In einer 2021 von Blandino Ortiz et al. publizierten Multizenterstudie wurde gezeigt, dass die Dreifachkanülierung eine effiziente Therapieoption ist, um eine persistierende Hypoxämie unter va-ECMO zu behandeln – es fehlen aktuell jedoch noch Aussagen zur Mortalität [20]. Fortsetzung des Fallbeispiels. Die bettseitig durchgeführte TTE und die Thoraxröntgenuntersuchung bestätigen den Verdacht des persistierenden Lungenödems, wobei sich der linke Ventrikel unter der va-ECMO-Therapie bereits etwas erholt hat und seine Kontraktilität zunimmt. Die Diuretikadosen werden erhöht, und die unter ECMO-Therapie weiterhin in geringer Dosis verabreichten Inotropika werden weiter reduziert. Die Beatmung wird durch Optimierung von PEEP und inspiratorischer Sauerstofffraktion (FIO2) sowie Steigerung des Atemminutenvolumens angepasst. Darunter bessert sich langsam die Sauerstoffsättigung an der rechten Hand, und es kann auf die Implantation einer weiteren venösen Kanüle verzichtet werden. „Netto-Hämodynamik“ Wird eine va-ECMO im kardiogenen Schockkardiogenen Schock , charakterisiert durch linksventrikuläres Pumpversagen mit erhöhtem linksventrikulärem enddiastolischem Druck („left ventricular enddiastolic pressure“, LVEDP), niedrigem MAP, niedrigem Schlagvolumen und niedriger Ejektionsfraktion implantiert, erhöht die va-ECMO zunächst immer die linksventrikuläre Nachlastlinksventrikuläre Nachlast . Diese Nachlaststeigerung tritt sowohl bei zentraler als auch bei peripherer Kanülierung auf. (Die linksventrikuläre Nachlast wird zwar immer durch eine va-ECMO gesteigert, dieser Effekt ist jedoch viel deutlicher bei peripherer als bei zentraler Kanülierung ausgeprägt.) Eine BlutflusssteigerungBlutflusssteigerung der ECMO von ca. 1,5 auf 4,5 l/min erhöht weiter die Nachlast, da die Kontraktilität des Herzens durch die ECMO nicht gesteigert wird [22]. Das heißt: Da das Herz gegen einen zunehmenden Widerstand auswerfen muss, steigen das LVEDV und der LVEDP, wohingegen das Schlagvolumen abnimmt. Das Arbeitsdiagramm des Herzens wird enger und schmäler und verschiebt sich nach rechts, wobei die Ruhedehnungskurve immer steiler wird. Das bedeutet, dass eine geringe Zunahme des LVEDV eine starke Zunahme des LVEDP zur Folge hat (Abb. 7). Damit wird die Fläche des Arbeitsdiagrammes größer, was einen gesteigerten myokardialen Sauerstoffverbrauch anzeigt. Merke Vor allem die periphere va-ECMO steigert die linksventrikuläre Nachlast. Eine Erhöhung der Nachlast ist mit einem erhöhten myokardialen Sauerstoffverbrauch verbunden. Blutfluss und -druck, die sich nach einer ECMO-Implantation ausbilden und die immer im Zusammenhang mit den jeweiligen (patho‑)physiologischen, patienteneigenen Faktoren stehen, können als „Netto-Hämodynamik der ECMO“ bezeichnet werden [22]. Parameter wie eine chronische Herzinsuffizienz (links- oder rechtsventrikulär), die Regenerationsfähigkeit des linken Ventrikels, der pulmonalvaskuläre Widerstand, der systemische Widerstand sowie modulierende medikamentöse und metabolische Faktoren müssen immer in Verbindung mit dem implantierten Herzunterstützungssystem gesehen werden. Die Netto-Hämodynamik kann sich also im kardiogenen Schock unter va-ECMO auch deutlich verschlechtern: Liegt eine linksventrikuläre Dysfunktionlinksventrikuläre Dysfunktion zugrunde, kann es sein, dass der linke Ventrikel gegen die zunehmende Nachlast kein Blut mehr auswerfen kann – die Aortenklappe bleibt geschlossen. Wenn jedoch der venöse Rückfluss zum Herzen die venöse Drainage der ECMO übersteigt, füllt sich der linke Ventrikel zunehmend mit ggf. unzureichend oxygeniertem Blut. Zum venösen pulmonalen Rückfluss kommen venöses Blut aus den Bronchialvenen und den Vv. Thebesii sowie evtl. eine aortale Regurgitationaortale Regurgitation bei insuffizienter Aortenklappe hinzu [30]. Das heißt, dass LVEDV und LVEDP weiter steigen. Auch der pulmonalkapilläre Verschlussdruckpulmonalkapilläre Verschlussdruck („pulmonary capillary wedge pressure“ [PCWP], Wedge-Druck) steigt; ein LungenödemLungenödem entsteht, was den pulmonalen Gasaustausch weiter verschlechtert. Durch die sich so entwickelnde linksventrikuläre Distension nimmt die linksventrikuläre Wandspannung weiter zu, was die subendokardiale KoronarperfusionKoronarperfusion verschlechtern und damit die linksventrikuläre Funktion zum Erliegen bringen kann [31, 32, 33]. Merke Die Netto-Hämodynamik unter ECMO ist das Zusammenspiel von ECMO-Blutfluss und patienteneigenen kardiovaskulären Faktoren. Hierbei wird auch ersichtlich, dass besonders die Koronarperfusion im kardiogenen Schock unter einer ECMO-Therapie gefährdet sein kann: Wenn kein Vorwärtsfluss aus dem linken Ventrikel mehr existiert, hängt sie alleinig vom retrograden Fluss über die ECMO ab. Sie kann aber dennoch durch die oben beschriebene erhöhte Wandspannung des linken Ventrikels beeinträchtigt sein. Oder aber, bei weiter bestehendem antegradem Fluss über die Aortenklappe kann es sein, dass der linke Ventrikel unzureichend oxygeniertes Blut auswirft und damit die Koronarien perfundiert, was ebenfalls zur myokardialen Hypoxämie und verschlechterter Mikrozirkulation führt. Deshalb sollte der Pumpenfluss mit dem daraus resultierenden Perfusionsdruck schon bei der ECMO-Anlage und Therapieinitiierung so adjustiert werden, dass im besten Fall weiterhin eine pulsatile Herzfunktion mit antegradem Auswurfantegradem Auswurf besteht. Ein pulsatiler Auswurfpulsatiler Auswurf ist erreicht, sobald ein Pulsdruck ≥ 15 mm Hg besteht [34]. So werden eine intrakavitäre Stase und eine myokardiale Schädigung durch linksventrikuläre Distension verhindert [17]. Um bei laufender ECMO ein AnsaugphänomenAnsaugphänomen , das sowohl bei vv- als auch bei va-ECMO auftreten kann, zu verhindern, ist ein adäquates intravasales Volumenmanagementintravasales Volumenmanagement nötig. Bei Hypovolämie saugen sich die Kanülen an den Gefäßwänden an, was sich meist als „Zittern“ oder „Wackeln“ des venösen Gefäßschlauchs bemerkbar macht. Das kann durch die Applikation kristalloider Flüssigkeit behoben werden, wobei darauf geachtet werden sollte, so viel Flüssigkeit wie nötig, aber so wenig Flüssigkeit wie möglich zu applizieren, um ein Lungenödem zu vermeiden [35]. Fortsetzung des Fallbeispiels. Die Oxygenierung bleibt unter der Diuretikagabe stabil. Eine zielgerichtete Volumengabe sichert die notwendige Vorlast, um ein Ansaugphänomen zu verhindern. Dennoch flacht die arterielle Blutdruckkurve am Folgetag zunehmend ab. Das TTE zeigt zwar eine gute rechtskardiale Funktion mit ausreichender Füllung, aber eine Zunahme des LVEDV. Zur besseren Überwachung wird ein Pulmonalarterienkatheter (PAK) gelegt. Monitoring Ziel eines adäquaten Monitorings ist zu erkennen, ob der gesamte Blutfluss (erzeugt durch das extrakorporale Unterstützungssystem + residuelles HZV des Patienten) zusammen mit der intravaskulären Füllung und dem Gefäßwiderstand des Patienten eine adäquate OrganperfusionOrganperfusion ermöglicht. Das erweiterte Monitoring sollte eine invasive Blutdruckmessung, eine ZVD-Messung, die regelmäßigen Bestimmungen der zentral- oder gemischtvenösen Sauerstoffsättigung, der Diurese sowie der Lactatkonzentration beinhalten; auch das endtidale CO2 (etCO2) sollte kontinuierlich gemessen werden [17]. Außerdem sind BGA aus der rechten oberen Extremität nötig, um Rückschlüsse auf die Oxygenierung von Herz und Gehirn ziehen zu können [36]. Zudem wird aus dem oben Beschriebenen deutlich, dass die va-ECMO-Therapie vorlastabhängig und nachlastsensibel ist und besonderes Augenmerk auf den Volumenstatus, die linksventrikuläre Entlastung und das linksventrikuläre Inotropie-Nachlast-Verhältnislinksventrikuläre Inotropie-Nachlast-Verhältnis geworfen werden muss [17]. Um dies suffizient zu evaluieren, sind regelmäßige echokardiographische Untersuchungenechokardiographische Untersuchungen nötig. So können Aussagen über die Vorlast, anhand von Größe des rechten Vorhofs und rechten Ventrikels, über die linksventrikuläre Vorlast durch den LVEDD und den Schweregrad einer Mitralinsuffizienz sowie über die linksventrikuläre Inotropie durch das Öffnungsverhalten der Aortenklappe und die linksventrikuläre Kontraktion getroffen werden. Wie häufig dieser Point-of-care-UltraschallPoint-of-care-Ultraschall (POCUS) angewandt werden sollte, ist aktuell Gegenstand vieler Diskussionen – aber als Verlaufskontrolle zum Ausschluss oder zur Therapie der linksventrikulären Distension und zum ECMO-Weaning ist er unerlässlich [36]. Dennoch hat auch traditionell der PulmonalarterienkatheterPulmonalarterienkatheter mit regelmäßigen Messungen des PCWP seinen Stellenwert als Verlaufsparameter bezüglich des frühen Erkennens einer linksventrikulären Distension [37]. Zerebrale NIRS und NIRS an den Extremitäten können ebenfalls als Verlaufsparameter dienen, um lokale Oxygenierungsstörungen oder eine verminderte Perfusion der kanülierten Extremität zu erkennen [36]. Linksventrikuläre Distension und Therapieoptionen Pathophysiologische Bedeutung Wie oben beschrieben, kann der retrograde va-ECMO-Fluss mit Steigerung der linksventrikulären Nachlast bei geschwächter linksventrikulärer Pumpleistung zu einer Zunahme des LVEDV und zur weiteren Schwächung des linken Ventrikels führen, das in das Vollbild der linksventrikulären Distension münden kann [22]. Solch eine linksventrikuläre Überladung bei Patienten unter va-ECMO hat eine Prävalenz von bis zu 70 % [31, 38]. Besonders unter „High-flow“-va-ECMO mit einem Blutfluss > 4 l/min bei Patienten mit massiv reduzierter oder erloschener linksventrikulärer Pumpfunktion ist das Risiko der linksventrikulären Distension signifikant erhöht [38, 39]. Eine in dieser Situation gute rechtsventrikuläre Funktion kann die Situation weiter verschlechtern, da kontinuierlich Blut aus dem rechten Ventrikel auf physiologischem Wege in das linksventrikuläre System gepumpt und nicht vollständig über die ECMO drainiert wird [40]. Es ist deshalb wichtig, regelmäßig echokardiographisch das linksventrikuläre Volumen, das Öffnungsverhalten der Aortenklappe und den Fluss über die Aortenklappe (antegrad oder retrograd) sowie eine mögliche linksventrikuläre Stase (in der Echokardiographie als „Nebel“) zu evaluieren [36]. Die linksventrikuläre Überladung schwächt nicht nur die kardiale Erholung unter der ECMO, sondern resultiert auch in einem schlechteren Langzeitergebnis [41]. Als möglicher Therapieansatz muss, je nach klinischem Bild, zwischen einer reinen Nachlastsenkungreinen Nachlastsenkung oder einer linksventrikulären Entlastunglinksventrikulären Entlastung unterschieden werden. Bei beiden stehen sowohl medikamentöse als auch interventionelle mechanische Möglichkeiten zur Verfügung. Jedoch ist es aktuell immer noch schwierig vorauszusehen, welcher Patient unter welcher Therapie am meisten profitieren wird; konservativ vs. interventionell; und wenn interventionell: perkutan oder chirurgisch? Besonders die chirurgische linksventrikuläre Entlastung erhöht das Komplikations- und BlutungsrisikoBlutungsrisiko signifikant und sollte sorgfältig evaluiert werden [32]. Dennoch ist es so, dass bei einer linksventrikulären Distension die interventionelle Dekompressioninterventionelle Dekompression des linken Ventrikels die Überlebenswahrscheinlichkeit der Patienten steigert. Merke Eine regelmäßige Echokardiographie soll das Öffnungsverhalten der Aortenklappe und den Blutfluss darüber zeigen, um eine linksventrikuläre Distension frühzeitig zu erkennen. Konservative Möglichkeiten In milden und moderaten Fällen der linksventrikulären Distension tragen konservative vorlast- und v. a. nachlastsenkende Maßnahmen dazu bei, die Situation zu verbessern. Eine reine Nachlastsenkung lässt sich meist gut durch AntihypertensivaAntihypertensiva erzeugen. Befindet sich der Patient mit moderater Distension im oder kurz vor dem Weaning von der va-ECMO, kann ein Therapieversuch mit Inodilatoren wie Dobutamin, Milrinon oder Levosimendan gestartet werden. Allerdings ist immer Vorsicht geboten, da alle Inotropika den kardialen Sauerstoffverbrauch und das Risiko für supraventrikuläre Tachykardiensupraventrikuläre Tachykardien steigern [31]. Zur bestmöglichen kardialen Funktion sollten normwertige Blutgase und ein ausgeglichener Säure-Basen-Haushalt vorliegen. Bei beatmeten Patienten wird ein PEEP > 8–10 cm H2O empfohlen, um die linksventrikuläre Mechanik zu unterstützen und ein Lungenödem zu vermeiden oder zu mildern [31]. Interventionelle Therapieoptionen Zu den interventionellen Möglichkeiten, den linken Ventrikel zu entlasten, zählen die atriale Septostomieatriale Septostomie , die chirurgische und die perkutane linksventrikuläre Entlastung über einen Vent-Katheter, die IABP und die linksventrikuläre Mikroaxialpumpe. Die atriale Septostomie muss als individueller Heilversuch angesehen werden; dieser kann zwar durch den künstlich geschaffenen Links-rechts-Shunt gute Ergebnisse erzielen, geht aber mit einer hohen Komplikationsrate um 10 % einher [30, 42, 43, 44]. Auch die chirurgische linksventrikuläre Entlastung liefert durch den großen Venting-KatheterVenting-Katheter , der direkt postoperativ platziert werden kann, eine gute linksventrikuläre Entlastung, hat aber ebenfalls ein hohes Komplikationsrisiko [38]. Komplikationsärmer ist der perkutan platzierte Links-Vent, der den PCWP signifikant senken kann [45]. Aufgrund ihrer guten Nachlastsenkung kann die intraaortale Ballonpumpeintraaortale Ballonpumpe auch bei va-ECMO-Patienten zur Entlastung des linken Ventrikels implantiert werden. Obwohl die IABP bei va-ECMO seit vielen Jahren ein oft angewandtes Verfahren zur linksventrikulären Entlastung ist, wird der klinische Nutzen der Kombination va-ECMO + IABP kritisch diskutiert [32]. Insgesamt war der linkventrikuläre Entlastungseffekt der IABP in mehreren Studien eher gering [39, 46]. Dennoch zeigten Grandin et al. in ihrer kürzlich veröffentlichten großen Multizenterstudie das bessere Risikoprofil bei ECMO + IABP im Vergleich zu ECMO + Impella bei gleicher Letalität [46]. Eine weniger invasive Methode, zusätzlich venöses Blut zu drainieren und damit den Rückfluss zum linken Herzen zu reduzieren, ist die Implantation einer zweiten venösen ECMO-Kanüle und die Umstellung auf eine venovenoarterielle Konfigurationvenovenoarterielle Konfiguration (vva-ECMO). Dieses Vorgehen reduziert den rechtsatrialen Druck deutlich und sollte infolgedessen auch den LVEDP absenken [47]. Es ist allerdings anzumerken, dass dieses Verfahren bisher nur an einem kleinen hochselektiven Patientenkollektiv angewandt wurde und vergleichende Daten besonders zur Risikostratifizierung fehlen. Linksventrikuläre Mikroaxialpumpe Ein weiterer invasiver, perkutaner Weg, den linken Ventrikel zu entlasten, ist die Implantation einer linksventrikulären Mikroaxialpumpe. Sie ist weniger invasiv als chirurgische linksventrikuläre Entlastungsmethoden, zeigt aber dennoch gute Ergebnisse [32]: Durch den kontinuierlichen Blutfluss, der vom linken Ventrikel in Richtung Aorta generiert wird, sistiert die klassische isovolumetrische Kontraktion, und die Druck-Volumen-Kurve wird vom Recht- zum Dreieck. Beachte: Viel wesentlicher ist aber, dass sie sich durch den gesenkten LVEDP und das gesenkte LVEDV wieder nach links verschiebt. Somit wird die linksventrikuläre Vorlast reduziert, was die Gefahr eines möglichen Lungenödems reduziert [48]. Als sekundärer Effekt können durch die Verbesserung der Mikrozirkulation eine Zunahme der linksventrikulären Kontraktilität und eine Abnahme des totalen peripheren Widerstands („total peripheral resistance“, TPR) beobachtet werden, sodass der Effekt der linksventrikulären Entladung weiter zunimmt – die Steigung der ESPVR und damit die Fläche des Dreiecks nehmen zu. Trotz einer erhöhten Komplikationsrate bei va-ECMO + Impella betrachten die Autoren mehrerer, u. a. multizentrischer Studien diese Kombination als sinnvolle Möglichkeit, das Herz im kardiogenen Schock zu entlasten [39, 49, 50, 51, 52]. Abschluss des Fallbeispiels. Die pharmakologische Therapie erzielt trotz suffizienter Nachlastsenkung keine Besserung der linksventrikulären Distension, sodass die Entscheidung zur zusätzlichen Impella-Implantation fällt. Darunter bessern sich die Hämodynamik und der klinische Zustand der Patientin deutlich. Nach 7 Tagen „ECpella“-Therapie kann die Patientin eine kombinierte Mitralklappenrekonstrunktion und Aortokoronare-Venen-Bypass(ACVB)-Operation erhalten. Fazit für die Praxis Weder die venovenöse (vv-) noch die venoarterielle extrakorporale Membranoxygenierung (va-ECMO) sind hämodynamisch „inert“. Der Erfolg der Therapie mit einem kardiozirkulatorischen Unterstützungssystem hängt entscheidend vom Monitoring vor und während der ECMO-Therapie und dem pathophysiologischen Verständnis für die hämodynamischen Veränderungen, die während der Therapie auftreten, ab. Unter einer vv-ECMO-Therapie sollten besonders das Ausmaß der Rezirkulation und die Belastung des rechten Ventrikels beurteilt werden. Je größer der Anteil der Rezirkulation, desto ineffektiver die vv-ECMO-Therapie. Die rechtskardiale Funktion, die zumeist durch das zugrunde liegende Lungenversagen beeinträchtigt ist, kann sich unter einer vv-ECMO verbessern oder auch weiter verschlechtern. Unter einer va-ECMO-Therapie konkurrieren bei Kanülierung über die Femoralarterie die Flüsse in der Aorta descendens, und die linksventrikuläre (LV-)Nachlast steigt signifikant. Darunter muss die LV-Funktion kritisch evaluiert werden. Ziel ist es, einen antegraden Fluss über die Aorta zu erhalten. Sollte der linke Ventrikel unter der erhöhten Nachlast dekompensieren und sich das Vollbild der LV-Distension zeigen, stehen konservative pharmakologische und interventionelle therapeutische Methoden zur Verfügung, um die Nachlast wieder zu senken und einen antegraden Fluss über die Aortenklappe wiederherzustellen. CME-Fragebogen Welcher Aspekt zählt nicht zu den typischen Gründen einer Extrakorporalen-Membranoxygenierung(ECMO)-Implantation? „Bridge to recovery“ „Bridge to decision“ „Bridge to transplant“ „Bridge to improvement“ „Bridge to destination“ Eine Patientin der Intensivstation mit schwerem „acute respiratory distress syndrome“ (ARDS) und zunehmender Hyperkapnie soll eine venovenöse extrakorporale Membranoxygenierung (vv-ECMO) erhalten. Welcher Zugangsweg wird am ehesten gewählt? Abführende Kanüle in die V. femoralis dextra, zuführende Kanüle in die V. jugularis interna sinistra Abführende Kanüle in die V. jugularis externa dextra, zuführende Kanüle in die V. jugularis interna sinistra Abführende Kanüle in die V. femoralis sinistra, zuführende Kanüle in die V. femoralis dextra Abführende Kanüle in die V. femoralis sinistra, zuführende Kanüle in die V. jugularis interna sinistra Abführende Kanüle in die V. femoralis dextra, zuführende Kanüle in die V. jugularis interna dextra Was ist bei der Kanülierung für eine Extrakorporale-Membranoxygenierung(ECMO)-Therapie zu beachten? Bei einer venovenösen (vv-)ECMO soll der Abstand zwischen der abführenden und der zuführenden Kanüle innerhalb der V. cava möglichst klein sein. Je größer der Durchmesser und je kürzer eine Kanüle ist, desto höher sind die möglichen Blutflüsse. Die Nutzung einer Doppellumenkanüle ist sowohl bei der venovenösen (vv-)ECMO als auch der venoarteriellen (va-)ECMO sinnvoll. Um die negativen Auswirkung auf die Hämodynamik bei einer venoarteriellen (va-)ECMO zu reduzieren, sollte die abführende Kanüle immer in die linke V. femoralis gelegt werden. Für eine venoarteriovenöse ECMO müssen beide venösen Kanülierungen zwingend in den Femoralvenen erfolgen. Was stellt keinen sinnvollen Therapieansatz einer linksventrikulären (LV-)Distension unter extrakorporaler Membranoxygenierung (ECMO) dar? Verwendung von Inodilatoren Umstellung auf venoarteriovenöse ECMO Perkutane linksventrikuläre Entlastung Intraaortale Ballonpumpe Linksventrikuläre Mikroaxialpumpe Welchen Einfluss kann die venovenöse extrakorporale Membranoxygenierung (vv-ECMO) mit gutem Fluss über die zu- und abführende Kanüle auf die patienteneigene Hämodynamik haben? Abnahme der rechtsventrikulären Nachlast Zunahme der rechtskardialen Vorlast Rechtsventrikuläre Stase bei deutlicher Volumenüberladung Steigerung des linksventrikulären Widerstands Zunahme der aortalen Regurgitation bei Aortenklappeninsuffizienz Was ist im Druck-Volumen-Diagramm des linken Ventrikels ein Maß für die kardiale Nachlast? Größe der Fläche des Arbeitsdiagramms Steigung der enddiastolischen Druck-Volumen-Kurve Enddiastolisches Volumen Arterielle Elastance Steigung der Ruhedehnungskurve Ein Patient unter venoarterieller extrakorporaler Membranoxygenierung (va-ECMO) entwickelt ein Harlekin-Syndrom. Was ist in diesem Fall zu tun? Der ECMO-Fluss sollte so weit reduziert werden, dass ein pulsatiler Auswurf durch den linken Ventrikel generiert werden kann. Echokardiographisch sollte das Öffnungsverhalten der Aortenklappe kontrolliert werden – eine aortale Regurgitation ist wahrscheinlich. Verbesserung der pulmonalen Perfusion; Blut, das der rechte Ventrikel auswirft, sollte unbedingt auch oxygeniert werden, bevor es den linken Ventrikel erreicht. Es sollten interventionelle Therapieoptionen eingeleitet werden; ein konservatives Vorgehen ist zumeist obsolet. Es sollte die Implantation einer linksventrikulären Mikroaxialpumpe durchgeführt werden. Was bezeichnet man als „Netto-Hämodynamik“ unter extrakorporaler Membranoxygenierung (ECMO)? Die isolierte Leistung von nur einem extrakorporalen Unterstützungssystem (typischerweise ECMO-Fluss) Die Leistung aller implantierten kardialen Unterstützungssysteme (z. B. ECMO + linksventrikulären Mikroaxialpumpe) Die hämodynamische Leistung, die ein Patient direkt vor einer ECMO-Implantation bei Indikationsstellung erbrachte Die Leistung, die eine ECMO-Pumpe am Modell mit maximaler Effizienz generieren kann Die hämodynamische Leistung, die durch ein Zusammenspiel von ECMO-Fluss und nativer kardiovaskulärer Funktion entsteht Ein Patient im kardiogenen Schock erhält eine venoarterielle extrakorporale Membranoxygenierung (va-)ECMO. Welches Monitoring ist dazu sinnvoll? Point-of-care-Ultraschall: Auskunft über die linksventrikuläre Entlastung Absoluter Wert des zentralen Venendrucks (ZVD): Auskunft über den Volumenbedarf Kontinuierliche Pulskonturanalyse: Bestimmung des Herzzeitvolumens (HZV) Nichtinvasive Blutdruckmessung: Bestimmung des Perfusionsdrucks Renale Nah-Infrarot-Spektroskopie: Bestimmung der Nierenperfusion Das transthorakale Echokardiogramm (TTE) eines Patienten mit venoarterieller extrakorporaler Membranoxygenierung (va-ECMO) und beginnender linksventrikulärer Distension zeigt typischerweise folgenden Befund: D-Zeichen der beiden Ventrikel Erhöhten Druck im Continuous-Wave(CW)-Dopplersonogramm über der Trikuspidalklappe Linksventrikuläre Stase („Nebel“) Flussbeschleunigung über die Aortenklappe Hypovolämischer rechter Ventrikel Einhaltung ethischer Richtlinien Interessenkonflikt Gemäß den Richtlinien des Springer Medizin Verlags werden Autoren und Wissenschaftliche Leitung im Rahmen der Manuskripterstellung und Manuskriptfreigabe aufgefordert, eine vollständige Erklärung zu ihren finanziellen und nichtfinanziellen Interessen abzugeben. Autoren A. M. Haas: A. Finanzielle Interessen: A. M. Haas gibt an, dass kein finanzieller Interessenkonflikt besteht. – B. Nichtfinanzielle Interessen: Anästhesistin an der Universitätsmedizin Rostock, an der Klinik für Anästhesiologie, Intensivmedizin und Schmerztherpie \| Kein Mitglied in relevanten Gesellschaften oder Vereinigungen. C. Busjahn: A. Finanzielle Interessen: C. Busjahn gibt an, dass kein finanzieller Interessenkonflikt besteht. – B. Nichtfinanzielle Interessen: Klinische oberärztliche Leitung, Klinik und Poliklinik für Anästhesiologie und Intensivtherapie \| Universitätsmedizin Rostock, Schillingallee 35, D-18057 Rostock. D. Crede: A. Finanzielle Interessen: D. Crede gibt an, dass kein finanzieller Interessenkonflikt besteht. – B. Nichtfinanzielle Interessen: Oberarzt der Klinik und Poliklinik für Anästhesiologie, Intensivmedizin und Schmerztherapie \| Universitätsmedizin Rostock, Schillingallee 35, D-18057 Rostock \| Mitgliedschaften: DGAI, BDA, DIVI, DESA. E. Kilger: A. Finanzielle Interessen: E. Kilger gibt an, dass kein finanzieller Interessenkonflikt besteht. – B. Nichtfinanzielle Interessen: Leitender Oberarzt – Standort Augustinum, Klinik für Anaesthesiologie, Klinikum der Universität München, Marchioninistr. 15, 81377 München. D. Reuter: A. Finanzielle Interessen: Forschungsunterstützung durch Pulsion Medical Systems, Sentec AG, Edwards Lifesciences. – Wissenschaftliche Beratungstätigkeiten für Pulsion Medical Systems, Getinge, Edwards Lifesciences. – B. Nichtfinanzielle Interessen: Direktor der Klinik und Poliklinik für Anästhesiologie und Intensivtherapie, Universitätsmedizin Rostock \| Schillingallee 35, D-18057 Rostock. Wissenschaftliche Leitung Die vollständige Erklärung zum Interessenkonflikt der Wissenschaftlichen Leitung finden Sie am Kurs der zertifizierten Fortbildung auf www.springermedizin.de/cme. Der Verlag erklärt, dass für die Publikation dieser CME-Fortbildung keine Sponsorengelder an den Verlag fließen. 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. Brodie D The evolution of extracorporeal membrane oxygenation for adult respiratory failure Ann Am Thorac Soc 2018 15 S57 S60 10.1513/AnnalsATS.201705-386KV 29461889 2. ELSO annual report \| ECMO \| Extracorporeal membrane oxygenation. https://www.elso.org/AboutUs/AnnualReport.aspx. Zugegriffen: 3. März 2022 3. Feldhaus D Brodie D Lemaitre P Sonett J Agerstrand C The evolution of the use of extracorporeal membrane oxygenation in respiratory failure Membranes 2021 11 491 10.3390/membranes11070491 34208906 4. Barbaro RP Odetola FO Kidwell KM Association of hospital-level volume of extracorporeal membrane oxygenation cases and mortality. Analysis of the extracorporeal life support organization registry Am J Respir Crit Care Med 2015 191 894 901 10.1164/rccm.201409-1634OC 25695688 5. Müller T Lubnow M Philipp A Pfeifer M Maier LS Extracorporeal pulmonary support procedures in intensive care medicine 2014 Internist 2014 55 1296 1305 10.1007/s00108-014-3506-x 25260398 6. Lindholm JA Cannulation for veno-venous extracorporeal membrane oxygenation J Thorac Dis 2018 10 S606 S612 10.21037/jtd.2018.03.101 29732177 7. Vieillard-Baron A Matthay M Teboul JL Experts’ opinion on management of hemodynamics in ARDS patients: focus on the effects of mechanical ventilation Intensive Care Med 2016 42 739 749 10.1007/s00134-016-4326-3 27038480 8. Mekontso Dessap A Boissier F Charron C Acute cor pulmonale during protective ventilation for acute respiratory distress syndrome: prevalence, predictors, and clinical impact Intensive Care Med 2016 42 862 870 10.1007/s00134-015-4141-2 26650055 9. Orde SR Behfar A Stalboerger PG Barros-Gomes S Kane GC Oh JK Effect of positive end-expiratory pressure on porcine right ventricle function assessed by speckle tracking echocardiography BMC Anesthesiol 2015 15 49 10.1186/s12871-015-0028-6 25873786 10. Vogel DJ Fabbri A Falvo A Assessment of right ventricular function with CT and echocardiography in patients with severe acute respiratory distress syndrome on extracorporeal membrane oxygenation Crit Care Explor 2021 3 e0345 10.1097/CCE.0000000000000345 33634265 11. Del Sorbo L Cypel M Fan E Extracorporeal life support for adults with severe acute respiratory failure Lancet Respir Med 2014 2 154 164 10.1016/S2213-2600(13)70197-8 24503270 12. Grant C Richards JB Frakes M Cohen J Wilcox SR ECMO and right ventricular failure: review of the literature J Intensive Care Med 2021 36 352 360 10.1177/0885066619900503 31964208 13. Abrams D Bacchetta M Brodie D Recirculation in venovenous extracorporeal membrane oxygenation ASAIO J 2015 61 115 121 10.1097/MAT.0000000000000179 25423117 14. Minini A Raes M Taccone FS Malbrain MLNG Transpulmonary thermodilution during extracorporeal organ support (ECOS): is it worth it? A brief commentary on the effects of the extracorporeal circuit on TPTD-derived parameters J Clin Monit Comput 2021 35 681 687 10.1007/s10877-021-00699-9 33891251 15. Reis Miranda D van Thiel R Brodie D Bakker J Right ventricular unloading after initiation of venovenous extracorporeal membrane oxygenation Am J Respir Crit Care Med 2015 191 346 348 10.1164/rccm.201408-1404LE 25635492 16. Juhl-Olsen P Smith SH Grejs AM Jørgensen MRS Bhavsar R Vistisen ST Automated echocardiography for measuring and tracking cardiac output after cardiac surgery: a validation study J Clin Monit Comput 2020 34 913 922 10.1007/s10877-019-00413-w 31677135 17. https://www.awmf.org/uploads/tx_szleitlinien/011-021l_S3_Einsatz-der-extrakorporalen-Zirkulation-ECLS-ECMO-bei-Herz-Kreislaufversagen_2021-02.pdf. Zugegriffen: 28. Juli 2021 18. Bunge JJH Caliskan K Gommers D Miranda DR Right ventricular dysfunction during acute respiratory distress syndrome and veno-venous extracorporeal membrane oxygenation J Thorac Dis 2018 10 S674 S682 10.21037/jtd.2017.10.75 29732186 19. Ius F Sommer W Tudorache I Veno-veno-arterial extracorporeal membrane oxygenation for respiratory failure with severe haemodynamic impairment: technique and early outcomes Interact CardioVasc Thorac Surg 2015 20 761 767 10.1093/icvts/ivv035 25736272 20. Blandino Ortiz A Belliato M Broman LM Early findings after implementation of veno-arteriovenous ECMO: a multicenter European experience Membranes 2021 11 81 10.3390/membranes11020081 33499236 21. Pape H-C Klinke R Brenner B Physiologie 2014 7 Stuttgart Thieme 22. Burkhoff D Sayer G Doshi D Uriel N Hemodynamics of mechanical circulatory support J Am Coll Cardiol 2015 66 2663 2674 10.1016/j.jacc.2015.10.017 26670067 23. Weber KT Are myocardial fibrosis and diastolic dysfunction reversible in hypertensive heart disease? Congest Heart Fail 2005 11 322 324 10.1111/j.1527-5299.2005.04479.x 16330908 24. van Heerebeek L Hamdani N Handoko ML Diastolic stiffness of the failing diabetic heart: importance of fibrosis, advanced glycation end products, and myocyte resting tension Circulation 2008 117 43 51 10.1161/CIRCULATIONAHA.107.728550 18071071 25. Luchner A Birner C BNP und NT-proBNP: Zwei kardiale Marker werden „erwachsen“. Dtsch Arztebl Laufs U 2016 10.3238/PersKardio.2016.10.14.02 26. Danial P Hajage D Nguyen LS Percutaneous versus surgical femoro-femoral veno-arterial ECMO: a propensity score matched study Intensive Care Med 2018 44 2153 2161 10.1007/s00134-018-5442-z 30430207 27. Raffa GM Kowalewski M Brodie D meta-analysis of peripheral or central extracorporeal membrane oxygenation in postcardiotomy and non-postcardiotomy shock Ann Thorac Surg 2019 107 311 321 10.1016/j.athoracsur.2018.05.063 29959943 28. David S Napp LC Kühn C Hoeper MM Extracorporeal membrane oxygenation: Principles and medical indications Internist 2016 57 856 863 10.1007/s00108-016-0102-2 27411792 29. Wong JK Smith TN Pitcher HT Hirose H Cavarocchi NC Cerebral and lower limb near-infrared spectroscopy in adults on extracorporeal membrane oxygenation Artif Organs 2012 36 659 667 10.1111/j.1525-1594.2012.01496.x 22817780 30. Ricarte Bratti JP Cavayas YA Noly PE Serri K Lamarche Y Modalities of left ventricle decompression during VA-ECMO therapy Membranes 2021 11 3 209 10.3390/membranes11030209 33809568 31. Belohlavek J Hunziker P Donker DW Left ventricular unloading and the role of ECpella Eur Heart J Suppl 2021 23 A27 A34 10.1093/eurheartj/suab006 33815012 32. Meani P Gelsomino S Natour E Modalities and effects of left ventricle unloading on extracorporeal life support: a review of the current literature Eur J Heart Fail 2017 19 Suppl 2 84 91 10.1002/ejhf.850 28470925 33. Kawashima D Gojo S Nishimura T Left ventricular mechanical support with Impella provides more ventricular unloading in heart failure than extracorporeal membrane oxygenation Asaio J 2011 57 169 176 10.1097/MAT.0b013e31820e121c 21317769 34. Undar A Frazier OH Fraser CD Defining pulsatile perfusion: quantification in terms of energy equivalent pressure Artif Organs 1999 23 712 716 10.1046/j.1525-1594.1999.06409.x 10463494 35. Communications E AINS – Anästhesiologie · Intensivmedizin · Notfallmedizin · Schmerztherapie. https://eref.thieme.de/ejournals/1439-1074_2020_03#/10.1055-a-0853-4013. Zugegriffen: 23. Juni 2022 36. Krishnan S Schmidt GA Hemodynamic monitoring in the extracorporeal membrane oxygenation patient Curr Opin Crit Care 2019 25 285 291 10.1097/MCC.0000000000000602 30865613 37. Rao P Khalpey Z Smith R Burkhoff D Kociol RD Venoarterial extracorporeal membrane oxygenation for cardiogenic shock and cardiac arrest Circ Heart Fail 2018 11 e004905 10.1161/CIRCHEARTFAILURE.118.004905 30354364 38. Donker DW Brodie D Henriques JPS Broomé M Left ventricular unloading during veno-arterial ECMO: a review of percutaneous and surgical unloading interventions Perfusion 2019 34 98 105 10.1177/0267659118794112 30112975 39. Donker DW Brodie D Henriques JPS Broomé M Left ventricular unloading during veno-arterial ECMO: a simulation study ASAIO J 2019 65 11 20 10.1097/MAT.0000000000000755 29517515 40. Donker DW Sallisalmi M Broomé M Right-left ventricular interaction in left-sided heart failure with and without venoarterial extracorporeal membrane oxygenation support—A simulation study ASAIO J 2021 67 297 305 10.1097/MAT.0000000000001242 33627604 41. Mirabel M Luyt C-E Leprince P Outcomes, long-term quality of life, and psychologic assessment of fulminant myocarditis patients rescued by mechanical circulatory support Crit Care Med 2011 39 1029 1035 10.1097/CCM.0b013e31820ead45 21336134 42. Seib PM Faulkner SC Erickson CC Blade and balloon atrial septostomy for left heart decompression in patients with severe ventricular dysfunction on extracorporeal membrane oxygenation Cathet Cardiovasc Intervent 1999 46 179 186 10.1002/(SICI)1522-726X(199902)46:2<179::AID-CCD13>3.0.CO;2-W 43. Alhussein M Osten M Horlick E Percutaneous left atrial decompression in adults with refractory cardiogenic shock supported with veno-arterial extracorporeal membrane oxygenation J Card Surg 2017 32 396 401 10.1111/jocs.13146 28497496 44. Baruteau A-E Barnetche T Morin L Percutaneous balloon atrial septostomy on top of venoarterial extracorporeal membrane oxygenation results in safe and effective left heart decompression Eur Heart J 2018 7 70 79 45. Alkhouli M Narins CR Lehoux J Knight PA Waits B Ling FS Percutaneous decompression of the left ventricle in cardiogenic shock patients on venoarterial extracorporeal membrane oxygenation J Card Surg 2016 31 177 182 10.1111/jocs.12696 26809382 46. Grandin EW Nunez JI Willar B Mechanical left ventricular unloading in patients undergoing venoarterial extracorporeal membrane oxygenation J Am Coll Cardiol 2022 79 1239 1250 10.1016/j.jacc.2022.01.032 35361346 47. Napp LC Kühn C Hoeper MM Cannulation strategies for percutaneous extracorporeal membrane oxygenation in adults Clin Res Cardiol 2016 105 283 296 10.1007/s00392-015-0941-1 26608160 48. Meani P Mlcek M Kowalewski M Transaortic or pulmonary artery drainage for left ventricular unloading in venoarterial extracorporeal life support: a porcine cardiogenic shock model Semin Thorac Cardiovasc Surg 2020 33 3 724 732 10.1053/j.semtcvs.2020.11.001 33171234 49. Pappalardo F Schulte C Pieri M Concomitant implantation of Impella® on top of veno-arterial extracorporeal membrane oxygenation may improve survival of patients with cardiogenic shock Eur J Heart Fail 2017 19 404 412 10.1002/ejhf.668 27709750 50. Grajeda Silvestri ER Pino JE Donath E Torres P Chait R Ghumman W Impella to unload the left ventricle in patients undergoing venoarterial extracorporeal membrane oxygenation for cardiogenic shock: A systematic review and meta-analysis J Card Surg 2020 35 1237 1242 10.1111/jocs.14560 32531130 51. Schrage B Becher PM Bernhardt A Left ventricular unloading is associated with lower mortality in patients with cardiogenic shock treated with venoarterial extracorporeal membrane oxygenation: results from an international, multicenter cohort study Circulation 2020 142 2095 2106 10.1161/CIRCULATIONAHA.120.048792 33032450 52. Patel SM Lipinski J Al-Kindi SG Simultaneous venoarterial extracorporeal membrane oxygenation and percutaneous left ventricular decompression therapy with impella is associated with improved outcomes in refractory cardiogenic shock ASAIO J 2019 65 21 28 10.1097/MAT.0000000000000767 29489461	36449054	PMC9709734	NO-CC CODE	2022-12-08 23:16:03	no	Anaesthesiologie. 2022 Nov 30; 71(12):967-982	utf-8	Anaesthesiologie	2,022	10.1007/s00101-022-01230-8	oa_other
==== Front Development (Rome) Development (Rome) Development (Society for International Development) 1011-6370 1461-7072 Palgrave Macmillan UK London 357 10.1057/s41301-022-00357-w Who’s Who Who’s Who 30 11 2022 15 © Society for International Development 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 pmcDereje Alemayehu is currently the Executive Coordinator of the Global Alliance for Tax Justice. He has an MA in Development Studies and PhD in Economics (magna cum laude) from Free University Berlin. From 1987 to 1998 he worked as a Lecturer at the Free University Berlin for development studies and African political economy. He worked in the NGO sector as country director in Burkina Faso, Tanzania and Kenya. He served as Founding Chair of Tax Justice Network Africa from 2007 to 2014. Prior to taking the role of Executive Coordinator, he served as Chair of the Global Alliance for Tax Justice from 2014 to 2018. He has published two books on Africa’s development challenges and has authored several articles and book chapters on development policy, the role of the state in development, governance, accountability, tax and development, and illicit financial flows. Contact details: Email: [email protected] Nathalie Beghin Holding a BA in Economics from the Free University of Brussels (ULB) and an MA and a Ph.D. in Social Policy from the University of Brasília (UnB), Nathalie Beghin has been Policy Coordinator for Brazil’s Institute for Socioeconomic Studies (Inesc) for ten years. In 2022, she was elected co-chair of Latindadd–Latin American Network for Economic and Social Justice. She worked for Brazil’s Ministry of Health and at the Institute for Applied Economic Research (Ipea) on, among other issues, hunger and food and nutrition insecurity, poverty, racial equality, and social engagement. She was a policy advisor for Oxfam International in Brazil and, later on, headed the Oxfam office in Brazil. She was a member of President Dilma Rousseff’s team that designed and implemented the Brasil sem Miséria (Brazil without Poverty) national plan. Contact details: Email: [email protected]; [email protected] Juan Pablo Bohoslavsky is a researcher at the National Council for Scientific and Technical Research of Argentina (CONICET) at the National University of Río Negro (CIEDIS). Between 2014 and 2020 he was United Nations Independent Expert on Debt and Human Rights. He also worked as Sovereign Debt Expert at UNCTAD. He holds a Ph.D. in Law. He has published books and articles on finance and human rights. Contact details: Email: [email protected] Francisco J. Cantamutto PhD in Social Science Research (Latin American Faculty of Social Sciences, FLACSO México). Associate Researcher at the National Council of Scientific and Technical Research (CONICET), based in the Southern Institute of Economic and Social Research (IIESS), National University of the South (UNS). Bahía Blanca, Argentina. Assistant Professor, Department of Economics, UNS. Specialized in political economy of development, with emphasis on Latin America and sovereign debt issues. He is a member of the Sociedad de Economía Crítica, which aims to a pluralistic approach to economics, and is part of its journal’s editorial board, Cuadernos de Economía Crítica. Contact details: Email: [email protected] Daniel Chavez is a Uruguayan-Dutch social anthropologist working for the Transnational Institute (TNI, an Amsterdam-based think tank), where he co-runs the Knowledge Hub Unit. He is also a Senior Research Associate at the Department of Anthropology and Development Studies of the University of Johannesburg, in South Africa. His research interests focus on public policy, climate politics and social justice issues. He holds a MA and a PhD in Development Studies from the International Institute of Social Studies of Erasmus University Rotterdam (ISS). His most recent books are Repensar lo Público.: Estado, Sociedad y Servicios Básicos en América Latina (2019) and Public Water and COVID-19: Dark Clouds and Silver Linings (2021), both co-edited with David McDonald and Susan Spronk. Contact details: Email: [email protected] Laura Clérico is Law Professor for Constitutional and Human Rights Law, at the University of Buenos Aires and Honorary Professor at the University of Erlangen Nuremberg in Germany. She is also Independent Researcher at the National Scientific and Technical Research Council of Argentina (CONICET). She has written extensively on inequality and human rights and has recently served as an expert witness before the Inter-American Court in cases about discrimination based on gender and sexual orientation. Contact details: Email: [email protected] Mr. Danish is Programme Officer of the Sustainable Development and Climate Change Programme of the South Centre. He is a qualified lawyer from India where he obtained a B.A. LL.B. (Honours) degree and a Diploma in International Trade and Business Law. He also holds a LL.M. in International Law from the Graduate Institute of International and Development Studies in Geneva. His work includes a wide variety of issues of high interest to developing countries, including reform of the international investment regime, business and human rights, global governance, the right to development and the promotion of South-South cooperation among others. Contact details: Email: [email protected] Armando De Negri Filho, brazilian/italian medical doctor, born in Porto Alegre – Brazil in 1962. Specialist in Emergency Medicine with Masters’ Degree in Epidemiology; Public Health / Global Health and Health Diplomacy; and Clinical Management. Doctor in Sciences / Preventive Medicine, Health Systems and Policies. Worked as public health officer and advisor in Brazil, Venezuela, Colombia, Paraguay, Peru and Guatemala, advising governments and social movements. Former General Coordinator of the Latin American Association of Social Medicine (ALAMES), Vice President of the Brazilian Center for Health Studies (CEBES), General Coordinator of the Brazilian Network for Cooperation in Emergencies (RBCE), Coordinator of the Executive Committee of the World Social Forum on Health and Social Security (WSFHSS). Former Member of the Experts Mechanism of the Right of Development (EMRTD) of the United Nations Human Rights Council. Senior Researcher of the Center for International Relationships in Health (CRIS) of the FIOCRUZ Foundation /Brazil. International Specialist Advisor on Health Systems and Services of the Pan American Organization/World Health Organization in Mexico. Contact details: Email: [email protected] Nicoletta Dentico is a journalist and a human rights advocate, with a long experience in disarmament issues, international cooperation and global health. In 1993 she started the Coalition to Ban Landmines in Italy and coordinated it until early 2000, when she took over as Director-General of Médecins Sans Frontières (MSF). In 2004 she joined the MSF Access to Essential Medicines Campaign and then Drugs for Neglected Diseases Initiative (DNDi) in Geneva. She worked as WHO consultant in Geneva and at EMRO in Cairo. From 2011 to 2014 she coordinated the Democratizing Global Health Coalition on the WHO Reform (DGH). She has published extensively on issues related to access to medicines and the right to health. From 2013 to 2019 she served as board member in the Italian Ethical Bank (Banca Popolare Etica). She currently directs the global health programme at Society for International Development (SID). Contact details: Email: [email protected] Evelina Fokina is a Political Science student who is currently obtaining her Bachelor’s Diploma at The University of Rome, Tor Vergata in the Global Governance course in Rome, Italy. Throughout her studies, she has obtained her Overseas experience at the Seoul National University in Seoul, South Korea (Aug 2022–Jan 2023) majoring in Political Science and International Relations. Before coming to Italy, she studied in Gothenburg, Sweden in order to obtain International Baccalaureate Diploma. In regards to high school education, she completed her studies in Saint Petersburg (Scool №27), and Petrozavodsk (Lyseum №1), and graduated with honors. Contact details: Email: [email protected] Wendy Harcourt is Professor of Gender, Diversity and Sustainable Development at the International Institute of Social Studies (ISS) of the Erasmus University Rotterdam in The Hague. She joined ISS in November 2011 after 23 years working at the Society for International Development as Editor of Development and programme director. She is series editor of the Palgrave Gender, Development and Social Change and the ISS-Routledge Series on Gender, Development and Sexuality. She has published widely on critical development, post development, body politics and feminist political ecology. Her latest edited book Feminist Methodologies: Experiments, Collaborations and Reflections (2022) published by Palgrave can be downloaded for free at: https://link.springer.com/book/10.1007/978-3-030-82654-3. Contact details: Email: [email protected] Yuefen Li is Senior Advisor on South-South Cooperation and Development Finance. Contact details: Email: [email protected] Ronald Mangani is an Associate Professor of Economics at the University of Malawi. He served as Secretary to the Treasury of the Government of Malawi from 2014 to 2017. He is currently Chairman of Old Mutual Malawi, and Director of First Capital Bank Malawi. Previously, he also served as director of the Reserve Bank of Malawi and six other enterprises and sat on the Monetary Policy Committee of the RBM. He is a founding member of the Economics Association of Malawi, a network member of the African Economic Research Consortium, and part of the global faculty of the Trade Policy Training Centre in Africa. Contact details: Email: [email protected] Chiara Mariotti is Senior Policy and Advocacy Officer on development finance at Eurodad. She joined Eurodad in 2020 to lead work on IFI’s undue influence on developing countries’ policy and fiscal space. Before joining Eurodad, she was Inequality Policy Manager for Oxfam GB, leading their policy and global advocacy work on inequality. Previously, she was a researcher at the Overseas Development Institute, focusing on policy solutions to chronic poverty and combining field, qualitative and quantitative research. She holds a PhD in Economics from the School of Oriental and African Studies (SOAS) University of London and has taught heterodox and political economy in British universities. Contact details: Email: [email protected] Dante Maschio Graduate in Environmental Sciences and MSC in Renewable Energies and Sustainability. Member and activist of Engineering Without Borders (Associació Internacionald'Enginyeria Sense Fronteres-ESF) since 2015. Secretariat member of the European Water Movement. He is also a member and activist in other social movements related to water justice such as Water is Life platform (Aigua és Vida) and Barcelona local water group (Moviment per l'Aigua Pública i Democràtica a l'AMB). He currently works in ESF campaign "Water a right not a commodity" focused on basic services, human and environmental rights. Contact details: Email: [email protected] Bhumika Muchhala is a political economist and senior policy advisor at the Third World Network and PhD candidate in political economy at The New School. She has 20 years of experience in analysis, advocacy, campaigning and public education on the international financial system, sustainable development, feminist economics, climate justice, colonial histories, and decolonial theory and praxis. Bhumika is engaged in advocacy in the UN intergovernmental negotiations and conferences, the International Monetary Fund, and the G20, and has collaborated with developing country governments, international civil society and social movements and academics. In the context of the UN, she has contributed to the outcome and content of the UN conference on the global financial crisis, the Sustainable Development Goals and Financing for Development conference between 2009 and 2016. She has been an expert consultant with various UN agencies, including the Office of the High Commissioner for Human Rights, UN Women, and UN Department of Economic and Social Affairs. Her doctoral research focuses on examining structural inequalities in the global political economy through an intersectional approach between decolonial theory and dependency theories. She has published articles in journals such as World Development, Third World Quarterly, and Nature, and teaches graduate seminars in ‘Global Political Economy of Development’ and ‘Global Environmental Policies and Politics.’ She is of Indian origin, grew up in Indonesia and is now based in New York. Contact details: Email: [email protected] Arthur Muliro is the Deputy Managing Director at the Society for International Development (SID). He has been at the forefront of SID’s work in East Africa and on Scenarios and is also an Associate Editor of Development. Contact details: Email: [email protected] Adebayo Olukoshi is Distinguished Professor at the Wits School of Governance in Johannesburg, South Africa. He has previously served as Director for Africa and West Asia at International IDEA, Director of the UN African Institute for Economic Development and Planning, and Executive Secretary of the African Social Science Council, CODESRIA. His research focus is on the interface of governance and development in a global comparative context. Contact details: Email: [email protected] Oliver Pahnecke is a part-time PhD candidate at Middlesex University, London, where he researches sovereign debt and human rights under the supervision of Prof. William Schabas. He works as a rule of law expert for international missions and as a consultant for technical co-operation and finance. Contact details: Email: [email protected] Quim Perez He has studied Humanities and Economics. Activist for the Human Right to Water and Sanitation and the defense of rivers for 25 years. Co-founder of Aigua és Vida, Red Agua Pública, Movimiento Europeo del Agua, Alianza Contra la Pobreza Energética, Ecologistas en Acción Confederal and Catalonia. He is currently an activist of Aigua és Vida. Platform of Catalan organizations working for the Human Right to Water and Sanitation and for the good ecological status of water bodies. Contact details: Email: [email protected] Míriam Planas MSC in Chemical Engineering with a specialization in water treatment and sustainability. Member and activist of Engineering Without Borders (Associació Internacional d'Enginyeria Sense Fronteres-ESF) since 2005, where she has worked since 2016 in the Campaign "Water a right not a commodity". Spokesperson of the platform "Aigua és Vida (AEV)", a platform which brings together environmental and HRTWS movement to guarantee good state of water ecosystems and HRTWS in Catalonia, and for the citizen initiative of water promoted in 2018 in Barcelona. Member of the Red Agua Pública and the European Movement for Water. She is part of the executive and consultant commission of the "Associació de Municipis i Entitats per l'Aigua Pública" and of board of the Global Water Partnership Alliance (GWOPA) that depends UN-Habitat. Contact details: Email: [email protected] María José Romero is Policy and Advocacy Manager on development finance at Eurodad. She joined Eurodad in 2012 and her role involves research, advocacy and monitoring policy developments at the global and European levels. Before joining Eurodad she worked at the secretariat of the Latin American Network for Economic and Social Justice (Latindadd) on tax justice and development finance. While in Uruguay, her home country, she was for five years Coordinator of the IFIs Latin American Monitor project at the Third World Institute (ITeM), focusing on policy analysis on development finance. She is a PhD candidate in International Development at SOAS University of London and holds a Masters’ Degree in political science from the University of the Republic of Uruguay. Contact details: Email: [email protected] Crystal Simeoni is a Pan African feminist activist working on macro level economic issues. She currently serves as the Director of Nawi—Afrifem Macroeconomics Collective (The Nawi Collective). In her role as director, Crystal curates the work of the collective towards contributing to building a feminist community in Africa of individuals and organisations working on influencing, analysing, deconstructing and reconstructing macroeconomic policies, narratives. The collective also works on reimagining alternatives through an intersectional Pan African feminist lens. Before this she was head of Advocacy with a focus on Economic Justice at FEMNET and was the Policy Lead of the Tax and the International Financial Architecture at TJN-A before that. She is currently an Atlantic Fellow for Social and Economic Equity at the London School of Economics. Contact details: Email: [email protected] Lavinia Steinfort is a political geographer and activist. As a researcher at the Transnational Institute (TNI) she is working on public alternatives such as (re)municipalisation of public services and just transitions towards energy democracy. Lavinia co-implemented the mPOWER project that facilitated peer learning among more than one hundred European energy transition municipalities and coordinates international Energy Democracy alliance. Lavinia co-edited The Future is Public: Towards Democratic Ownership of Public Services, Public Finance for the Future We Want and One Treaty to rule them all about the Energy Charter Treaty. Lavinia also published the article Communal Performativity—A Seed for Change on the solidarity of Thessaloniki’s social movements in the diverse fights against neoliberalism (Antipode, 2017). Contact details: Email: [email protected] Daniel Uribe is Lead Programme Officer of the Sustainable Development and Climate Change (SDCC) Programme of the South Centre. He has been extensively engaged in the research of the implications of International Investment Agreements in sustainable development for developing countries, and on the issues of business and human rights; he has also conducted research on South-South and Triangular Cooperation for the achievement of the Sustainable Development Goals. Mr. Uribe is a lawyer from Ecuador where he obtained a Bachelor of Laws LL.B. from the Pontifical Catholic University of Ecuador and holds a Masters’ Degree (LL.M.) in International Law (Cum Laude) from the Graduate Institute of Geneva. During his thirteen years of professional career, Mr. Uribe has earned experience working for the Republic of Ecuador, intergovernmental organizations, and civil society institutions. This has allowed him to understand public policy concerns from diverse perspectives, as an advantage to forge strategic synergies to achieve common objectives towards sustainable development and international cooperation. Contact details: Email: [email protected] Ella Weber is an Undergraduate in Global Governance at the Tor Vergata University of Rome, Italy. She is majoring in Political Science, Law and History and currently pursuing her Overseas semester in Taichung, Taiwan to expand her education outside of the European boundaries. During her internship at the Society for International Development (SID) in Rome, she organized at Tor Vergata, together with co-author Evelina Fokina and the SID team, an intergenerational dialogue around the concept of development which has been transposed into the article found in this issue of the Development Journal—her first publication. Contact details: Email: [email protected]. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.	36467546	PMC9709735	NO-CC CODE	2022-12-01 23:23:08	no	Development (Rome). 2022 Nov 30;:1-5	utf-8	Development (Rome)	2,022	10.1057/s41301-022-00357-w	oa_other
==== Front Pediatr Nephrol Pediatr Nephrol Pediatric Nephrology (Berlin, Germany) 0931-041X 1432-198X Springer Berlin Heidelberg Berlin/Heidelberg 36449102 5783 10.1007/s00467-022-05783-z Original Article Is influenza vaccination associated with nephrotic syndrome relapse in children? A multicenter prospective study http://orcid.org/0000-0003-3863-2743 Ishimori Shingo [email protected] 1 Horinouchi Tomoko 2 Fujimura Junya 3 Yamamura Tomohiko 2 Matsunoshita Natsuki 4 Kamiyoshi Naohiro 5 Sato Mai 6 Ogura Masao 6 Kamei Koichi 6 Ishikura Kenji 7 Iijima Kazumoto 89 Nozu Kandai 2 1 grid.416862.f Department of Pediatrics, Takatsuki General Hospital, 1‑3‑13 Kosobe‑cho, Takatsuki, 569‑1192 Japan 2 grid.31432.37 0000 0001 1092 3077 Department of Pediatrics, Kobe University Graduate School of Medicine, 7‑5‑1 Kusunoki‑cho, Chuo‑ku, Kobe 650‑0017 Japan 3 Department of Pediatrics, Kakogawa Central City Hospital, 439 Honmachi, Kakogawa‑cho, Kakogawa 675‑8611 Japan 4 Department of Pediatrics, Kita-Harima Medical Center, 926‑250 Ichiba‑cho, Ono, 675‑1392 Japan 5 grid.414105.5 0000 0004 0569 0928 Department of Pediatrics, Himeji Red Cross Hospital, 1‑12‑1 Shimoteno, Himeji, 670‑8540 Japan 6 grid.63906.3a 0000 0004 0377 2305 Division of Nephrology and Rheumatology, National Center for Child Health and Development, 2-10-1 Okura, Setagaya-ku, Tokyo, 1578535 Japan 7 grid.410786.c 0000 0000 9206 2938 Department of Pediatrics, Kitasato University School of Medicine, 1-15-1 Kitazato, Minami-ku, Sagamihara, 2520374 Japan 8 grid.415413.6 0000 0000 9074 6789 Hyogo Prefectural Kobe Children’s Hospital, 1‑6‑7 Minatojima‑minamimachi, Chuo‑ku, Kobe 650‑0047 Japan 9 grid.31432.37 0000 0001 1092 3077 Department of Advanced Pediatric Medicine, Kobe University Graduate School of Medicine, 1‑6‑7 Minatojima‑minamimachi, Chuo‑ku, Kobe 650‑0047 Japan 30 11 2022 110 4 5 2022 17 9 2022 10 10 2022 © The Author(s), under exclusive licence to International Pediatric Nephrology Association 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. Background Prospective research of children receiving heterogeneous vaccines has shown that immunization is not associated with pediatric idiopathic nephrotic syndrome (NS) relapses. However, prospective data concentrating only on influenza (flu) virus vaccines are not available. Methods This multicenter prospective study was conducted in children with NS who received inactivated flu vaccines from June 2017 to July 2018. The day of flu vaccination was defined as day 0, and the period between prevaccination and postvaccination days was defined as − X to + Y (period from day − 180 to 0 as the precontrolled period). The primary outcome was the NS relapse rate from day 0 to + 30 as a direct association with vaccination compared with those in the precontrolled period. Exacerbation was defined as children experiencing more NS relapses after vaccination compared with those in the precontrolled period, or children starting any new immunosuppressants due to NS relapse after vaccination. Results Sixty-three children were included. Relapse rates were not significantly different between the precontrolled period and 0 to + 30 periods (0.38 vs. 0.19 times/person-year, p = 0.95). Although the exacerbation rate during the 0 to + 180 period in children without NS relapse in the precontrolled period was very low (4/54 [7.4 %]), children with at least one NS relapse in the precontrolled period showed a remarkable increase in the rate (4/9 [44.4%]; p = 0.01). Conclusions Flu vaccination did not significantly precipitate the direct relapse of NS in children. However, it might increase the disease activity in children with at least one NS relapse within a half year before vaccination. Graphical abstract A higher resolution version of the Graphical abstract is available as Supplementary information Supplementary Information The online version contains supplementary material available at 10.1007/s00467-022-05783-z. Keywords Adverse event Exacerbation Idiopathic nephrotic syndrome Influenza virus vaccination Relapse ==== Body pmcIntroduction Children with idiopathic nephrotic syndrome (NS) are in an immunocompromised state and require prolonged steroid therapy and immunosuppressive agents to prevent NS relapse [1, 2]. This immune suppression contributes to the potential high risk for upper respiratory viral infections and influenza (flu) virus. Guidelines suggest that these children should receive inactivated vaccines to reduce the risk of serious infections, and the use of live attenuated vaccines is not recommended [3]. Several reports have indicated that immunizations may precipitate or induce relapses of NS as an immunogenic stimulus [4–9]. T-cell–mediated podocyte injury has been hypothesized as the pathology underlying the elicitation of the immune response by vaccination [10]. Relapse of NS associated with a monovalent whole-virion inactivated flu virus vaccine during a pandemic influenza season has been described [11, 12]. Because there were no available data focusing on the relative risk of NS relapse related to the flu vaccine, we first conducted a retrospective study targeting vaccinated children with NS in a single institution [13]. In comparing the NS relapse rate between the prevaccination and postvaccination periods, we found that the flu vaccine was not associated with a higher risk of NS relapse. We subsequently conducted a nationwide retrospective cohort study including 306 children with idiopathic NS to investigate both vaccinated and unvaccinated patients at a number of facilities [14]. The results of the multivariate analysis showed that children who received flu vaccination had significantly fewer flu infections and NS relapses than those who did not receive the flu vaccination. Moreover, children receiving flu vaccination had a significantly lower risk of NS relapse during the postvaccination period (risk ratio: 0.31, 95% confidence interval: 0.17–0.56) than during the prevaccination period. Recently, a prospective study was published that evaluated the association between NS relapse and plural vaccination not limited to the flu vaccine. The authors of that research reported that the administration of vaccines developed by purified proteins was not associated with a higher risk of NS relapse in children with steroid-dependent NS (SDNS) [15]. However, the limitations of that research were the relatively small sample size, as only 19 vaccinated children were included, and that the vaccinated group was composed of a heterogeneous population receiving flu, diphtheria, tetanus, and acellular pertussis vaccine combined with inactivated polio virus vaccine. Here, we report a multicenter prospective study of children with NS who received inactivated subunit-antigen flu vaccination and evaluate the occurrence of NS relapses related to flu vaccinations. Materials and methods Patients We conducted a multicenter prospective study of patients with NS who received inactivated subunit-antigen flu vaccines from June 2017 to July 2018 at the following six hospitals: Takatsuki General Hospital, Kakogawa Central City Hospital, Kobe University Hospital, Kita-Harima Medical Center, Himeji Red Cross Hospital, and National Center for Child Health and Development. Patients were eligible if they were newly diagnosed with idiopathic NS between the ages of 6 months and 15 years old and if their last relapse was steroid-sensitive NS (SSNS), regardless of history of steroid-resistant NS (SRNS). Children with congenital NS or NS secondary to nephritis were excluded. We prospectively collected the following data: sex, age at first manifestation of NS, age at first inactivated subunit-antigen flu vaccination, history of SRNS, last NS type (frequently relapsing NS (FRNS)/SDNS or others), last creatinine (Cr) estimated glomerular filtration rate before administration of first vaccination, duration between onset of NS and flu vaccination, duration between last relapse of NS and flu vaccination (excluded children who have never had any relapse from onset of NS), total number of NS relapses during the period between onset of NS and flu vaccination, use of immunosuppressants at the day − 180 preceding flu vaccination (cyclosporine, mycophenolate mofetil, mizoribine, cyclophosphamide, tacrolimus, and rituximab (RTX)), use of immunosuppressants at first vaccination, use of RTX at first vaccination, use of prednisolone (PSL) at flu vaccination, total number of NS relapses during study period, total number of flu vaccinations, and total number of flu infections. We requested the participant to fill out a questionnaire that included the date of side effects within 1 week of flu vaccination (fever, local redness, swelling, pain, and feeling of heat) and fever, symptoms of various infections, and other events. Definitions The definitions of the general condition in pediatric NS used in the present study were in accordance with the clinical guidelines issued by the Japanese Society for Pediatric Nephrology [16, 17]. Idiopathic NS in children was defined as hypoalbuminemia (serum albumin levels ≤ 2.5 g/dL) and severe proteinuria (≥ 40 mg/h/m2 in pooled nighttime urine or an early morning urine protein Cr ratio > 2.0 g/g Cr). Complete remission was defined as a urine protein creatinine ratio < 0.2 g/g Cr or ≤ − protein on dipstick testing of early morning urine for 3 consecutive days. SSNS was defined as complete remission in less than 4 weeks after initiation of daily PSL therapy. Relapse of NS was defined as ≥ 3 + protein on dipstick testing of early morning urine for 3 consecutive days. FRNS was defined as two or more relapses within 6 months of the initial response or more than four relapses within any 12-month period. SDNS was defined as two consecutive relapses during PSL tapering or < 14 days after discontinuation of PSL therapy. SRNS was defined as the absence of complete remission after ≥ 4 weeks of daily PSL therapy. In our institutions, all patients with NS administer urine protein check on dipstick testing of every first-morning urine at home. In addition, if the results of their dipstick testing meet the criteria of NS relapse, they are instructed to contact their institution. The day of administration of inactivated subunit-antigen flu vaccination was defined as day 0. The periods within 1 month following and 6 months preceding flu vaccination were designated + 30 to − 180. Based on the background that the side effects of inactivated vaccines generally occur within 1 month of administration, we defined the period between 0 to + 30 as the direct association period involving vaccination. To define the period − 180 to 0 as the prevaccination period (precontrolled period), we examine the NS relapse rates during the period 0 to + 180. Children older than 13 years generally receive a flu vaccination once per year, and those younger than 13 years receive a flu vaccination twice per year. In children who received two vaccinations in the same year, we defined “first flu vaccine” as only their first vaccination of the two. The rate of NS relapse was defined as the number of relapses one person experienced within 1 year. The duration of RTX therapy was defined as the period from the day of RTX administration to the day of B-cell recovery (CD19 + B-cell count of ≥ 1% of total lymphocytes). Exacerbation of NS was defined following two patterns. The first pattern consisted of children with greater numbers of NS relapses during the period 0 to + 180 than the number of NS relapses occurring during the precontrolled period. The other pattern was children starting any new immunosuppressants because they experienced a relapse in NS during the period 0 to + 180. Study design The primary outcome was the rate of NS relapse occurring within the direct association period involving vaccination during the period 0 to + 30 compared with the rate during the precontrolled period in all patients. We examined and subdivided this outcome into each group as follows: (1) in patients receiving oral glucocorticoids at the time of flu vaccination or not, (2) in patients with at least one NS relapse during the precontrolled period (unstable group) or those without NS relapse during the precontrolled period (stable group), (3) in patients receiving immunosuppressants at the time of flu vaccination or not, and (4) in patients with side effects within 1 week of flu vaccination or not. Secondary outcomes were the rate of NS relapse during the period 0 to + 180 compared with the rate during the precontrolled period, the rate of patients infected with the flu virus, and the rate of patients with severe flu virus infection. In addition, we evaluated the clinical characteristics and percentage of children with an NS exacerbation and compared the results between patients in the unstable group and those in the stable group. Regimens Our NS treatment strategy is based on the modified protocol of the International Study of Kidney Diseases in Children, as shown in the Japanese pediatric idiopathic NS guideline [1, 17–19]. The initial 8-week treatment protocol was 60 mg/m2/day PSL (maximum daily dosage 60 mg) for 4 weeks, followed by 40 mg/m2/day (maximum daily dosage 40 mg) on alternate days for 4 weeks. Our policy was to administer inactivated subunit-antigen flu vaccine to consenting children with NS, except during a relapse of NS and during any period of PSL therapy of ≥ 2 mg/kg/day. Statistical analysis We performed all analyses using JMP version 11.0 (SAS institute Japan Ltd., Tokyo, Japan). The Wilcoxon rank-sum test was used to evaluate the association between categorical values and continuous values, and Fisher’s exact test was used for two categorical values. All data were expressed as median value + interquartile range or number (percentage). The NS relapse rate was calculated using the person-year method. A p value of < 0.05 was considered statistically significant. The calculated sample size was 40 children. Because we assumed that the median number of events was greater than two times per year in each patient, we required approximately 80 events (NS relapse) in our analysis. Our institutions included both children diagnosed as FRNS/SDNS and those diagnosed as non-FRNS/SDNS. Children with FRNS or SDNS experience two or more NS relapses within 6 months. If the ratio of the FRNS/SDNS and non-FRNS/SDNS groups was 1:1, the estimated median number of events was two times per year in each patient cumulatively. In each of our institutions, the number of patients with NS for < 15 years ranges from 5 to 80 (median: 15–20 children with NS/institution). If approximately half of the children with NS in our institutions receive the flu vaccine, the estimated number in six institutions would be 45–60 (7.5–10 children × 6 institutions = 45–60). Moreover, we assumed that 10–20 children would be omitted because of not gaining consent. Hence, to satisfy these assumptions, we calculated the sample size as 40 children. Results Clinical characteristics In the present study, we included 79 children who received the flu vaccination. We excluded eight children who were infected with the flu virus before receiving the flu vaccine during the study period, and eight children with a record that was inadequate for analysis. Thus, we included a total of 63 children (24 boys and 39 girls) in this study (Fig. 1). Table 1 summarizes the patients’ characteristics. The median age at first manifestation of NS and age at flu vaccination were 3.4 (2.3–7.5) and 10.8 (7.7–14.3) years, respectively. Only one patient (1.6%) was under 1 year old, and 11 children (17.4%) were teenage at the first manifestation of NS. The remaining 51 children (81.0%) were between 1 and 10 years old at the first manifestation of NS. Eight children (12.7%) had a history of SRNS. The duration between the last relapse of NS and flu vaccination in the unstable group was significantly shorter than those in the stable group (0.1 vs. 1.3 years; p = 0.002). The proportion of children who changed immunosuppressants during the precontrolled period in the unstable group was slightly higher than those in the stable group but without reaching significance. The numbers of children with adverse events within 1 week of flu vaccination were 22 (34.9%). During the period after flu vaccination from day 0 to + 180, there were 16 flu infections and 16 children with flu infections. No patients experienced a severe infection of the flu virus. In 16 children with flu infection in this study, there were two NS relapses associated directly with flu infection. Among them, the first patient showed no NS relapse during the precontrolled period. However, the patient relapsed a day after the flu infection from day 0 to + 180 (more than 1 month after flu vaccination). The second patient with a relapse of NS during the precontrolled period showed two NS relapses from day 0 to + 180. In this case, one relapse occurred 4 days after flu infection (more than 1 month after flu vaccination). Therefore, NS relapses within 4 days after flu infection might be associated directly with infection. We considered these events as confounders and excluded them from the analysis.Fig. 1 Patient flow chart. Of 79 children who received the flu vaccination, a total of 63 children were included in this study Table 1 Comparison of patient clinical characteristics between the unstable group and the stable group All (n = 63) NS relapse during the period between day –180 and 0 p value Unstable group (n = 9) Stable group (n = 54) Background Age at first manifestation of NS (years) 3.4 (2.3–7.5) 2.9 (1.9–7.5) 3.5 (2.3–7.8) 0.43 Age group of NS, n (%) under than 1 year old between 1 and 10 years old over than 10 years old 1 (1.6) 51 (81.0) 11 (17.4) 0 (0.0) 9 (100) 0 (0.0) 1 (1.8) 42 (77.8) 11 (20.4) Age at first flu vaccination (years) 10.8 (7.7–14.3) 9.8 (8.0–13.1) 11.0 (7.4–14.9) 0.34 Boy:girl, n (%) 39:24 (61.9:38.1) 7:2 (77.8:22.2) 32:22 (59.3:40.7) 0.46 Past history of flu vaccination, n (%)* 44 (81.5) 7 (87.5) 37 (80.4) 1.0 Past history of steroid-resistant NS, n (%) 8 (12.7) 1 (11.1) 7 (13.0) 1.0 Frequent relapsing or steroid-dependent NS as last NS type, n (%) 21 (33.3) 2 (22.2) 19 (35.2) 0.71 The last Cr-eGFR before administration of first vaccination (ml/min/1.73 m2)† 120.1 ± 18.7 113.2 ± 6.6 121.8 ± 3.2 0.25 Duration between onset of NS and flu vaccination (years) 5.4 (3.4–8.7) 5.3 (4.7–7.9) 5.7 (3.0–9.1) 0.73 Duration between last relapse of NS and flu vaccination (years) 0.8 (0.5–2.4) 0.1 (0.1–0.3) 1.3 (0.6–0.2.8) 0.002 Total number of NS relapse during the period between onset of NS and flu vaccination 8 (4–11) 10 (6.5–10.5) 7 (3–11) 0.77 The state of being on immunosuppressants at the day –180 before flu vaccinations 40 (63.5) 6 (66.7) 34 (63.0) 1.0 The change of immunosuppressants during the period between day –180 and 0 5 (7.9) 2 (22.2) 3 (5.6) 0.16 The state of being on glucocorticoid at flu vaccinations 6 (9.5) 5 (55.6) 1 (1.9) 0.0001 The state of being on immunosuppressants at flu vaccinations 42 (66.7) 7 (77.8) 35 (64.8) 0.71 being on cyclosporine 18 (28.6) 5 (55.6) 13 (24.1) 0.10 being on mycophenolate mofetil 18 (28.6) 2 (22.2) 16 (29.6) 1.0 being on mizoribine 8 (12.7) 0 (0.0) 8 (14.8) 0.59 being on tacrolimus 2 (3.2) 1 (11.1) 1 (1.9) 0.27 being on rituximab 2 (3.2) 1 (11.1) 1 (1.9) 0.27 Total number of flu infected patient, n (%) 16 (25.4) 3 (33.3) 13 (24.1) 0.68 Total number of flu infection, (times/person-year) 0.51 ± 0.88 0.67 ± 0.29 0.48 ± 0.12 0.56 Total number of patients who have the local side effects within a week of flu vaccination, n (%) 22 (34.9) 3 (33.3) 19 (35.2) 1.0 Local redness 14 (22.2) 2 (22.2) 12 (22.2) 1.0 Local swelling 15 (23.8) 2 (22.2) 13 (24.1) 0.71 Local pain 13 (20.6) 1 (11.1) 2 (22.2) 0.67 Local feeling of heat or fever 13 (20.6) 2 (22.2) 11 (20.4) 1.0 NS nephrotic syndrome, Flu influenza virus, Cr-eGFR creatinine-estimated glomerular filtration rate *Evaluated 54 children with data for past history of influenza vaccination †Evaluated 41 children with data of the last Cr-eGFR before administration of first vaccination The total number of NS relapses during the whole study period was 30 in 13 children. In comparison between the children with at least one relapse and those with no relapse, there was no significant difference in terms of NS (history of SRNS, last NS type of FRNS, or SDNS), age at first flu vaccination, age at the first manifestation of NS, history of flu vaccination, and type of immunosuppressants (cyclosporine, mycophenolate mofetil, mizoribine, cyclophosphamide, tacrolimus, and RTX). There were 12 NS relapses and 9 children who experienced NS relapse before flu vaccination during the study period. We evaluated and compared the patients’ clinical characteristics between the unstable group (n = 9) and the stable group (n = 54; Table 1). A significantly higher proportion of children was receiving glucocorticoids during the time of flu vaccination in the unstable group as compared with the stable group. There were no other significant differences between the two groups in other clinical characteristics, including background, receiving immunosuppressants, total number of flu-infected children, or total number of children experiencing a side effect within 1 week of flu vaccination. We observed 37 episodes in 23 children, which included symptoms of infectious diseases (for example, fever, coughing, and diarrhea) outside of flu infection during the study period. Of these, 15 episodes of infections occurred during the precontrolled period, and the remaining 22 occurred during the period between day 0 and + 180. Further, there were two NS relapses associated directly with infections other than flu infection. Among these, the first patient had a relapse 2 days after acute enterocolitis during the precontrolled period (more than 1 month before flu vaccination). In the second patient, a relapse occurred 3 days after upper respiratory tract infection during the precontrolled period (more than 1 month before flu vaccination). Although we considered that NS relapses 2 or 3 days after the appearance of any infectious symptoms might be associated directly with infection, relapse after flu vaccination was not seriously affected by these those occurring during the precontrolled period (more than 1 month before flu vaccination). We observed eight vaccinations in six children who received any of the following vaccines: Varicella vaccine, Mumps vaccine, Measles–Rubella mixed vaccine, and Pneumococcal vaccine, before flu vaccination during the precontrolled period, and five times in four children who received vaccinations other than flu vaccine (Varicella vaccine, Diphtheria-Tetanus mixed vaccine, Japanese encephalitis vaccine, Mixed Measles, and Rubella vaccine) after flu vaccination during the period from day 0 to + 180. Moreover, none of the above 10 children experienced NS relapse after 13 heterogeneous vaccinations during the study period. Therefore, we considered that other vaccines did not affect either flu vaccinations or NS relapse in these 10 children. The remaining 53 children received no other vaccines during the study period. Comparison of the NS relapse rate between the precontrolled period and the period after flu vaccination from day 0 to + 30 as a direct association with vaccination During the period after flu vaccination from day 0 to + 180, there were 18 occurrences of NS relapse and 11 children experiencing NS relapse. Figure 2 shows that among all 63 patients, the relapse rate was not significantly different between the precontrolled period and the period 0 to + 30 (0.38 vs. 0.19 times/person-year, p = 0.95) as the primary outcome. The additional analysis subdivided the primary outcome into each group as follows. The number of patients receiving glucocorticoids at the time of flu vaccination was too small (n = 6) to allow a comparison of the relapse rate between the prevaccination and postvaccination periods in this group. In patients not receiving glucocorticoids at the time of flu vaccination (n = 57), the relapse rate was 0 times/person-year during days 0 to + 30 (p = 0.94) as compared with 0.14 times/person-year during the precontrolled period (Supplementary Figure 1). Among the patients in the unstable group (n = 9) and in the stable group (n = 54), there was no significant difference in the relapse rate in each period during the precontrolled period between patients who were receiving immunosuppressants at the time of flu vaccination (n = 42) or not (n = 21) and in patients with adverse events within 1 week of flu vaccination (n = 22) or not (n = 41) as compared with those in each group during the 0 to + 30 period. The difference in the relapse rate in each period during the precontrolled period was not significant (Supplementary Figures 2–4).Fig. 2 Comparison of the relapse rate between the precontrolled period and the 0 to + 30 period in all 63 children. ns not significant Comparison of exacerbation rates during the precontrolled period between the unstable and stable groups As shown in Fig. 3a, the rate of children with an exacerbation of NS during the period 0 to + 180 in the unstable group (4/9 [44.4%]) was significantly higher than that in the stable group (4/54 [7.4%]; p = 0.01). The details of the four children in the unstable group were as follows: two children with a relapse of NS during the precontrolled period experienced three relapses during the period 0 to + 180 following initiating mizoribine therapy, and two children with two relapses of NS during the precontrolled period experienced two relapses during the period 0 to + 180 following initiating cyclosporine therapy. The details of the four children in the stable group were as follows: three patients without a relapse of NS during the precontrolled period experienced a relapse during the period 0 to + 180, and one patient without a relapse during the precontrolled period experienced two NS relapses during the period 0 to + 180. We performed an analysis of the comparison of the relapse rate between the precontrolled period and the day 0 to + 180 period in all 63 children (Fig. 3b) and subdivided results into the unstable and stable groups (Fig. 3c in 9 children in the unstable group and Fig. 3d in 54 children in the stable group). As shown in Fig. 3b–d), the relapse rates from day 0 to + 180 did not increase compared with those during the precontrolled period both in all children and in subdivided groups.Fig. 3 Comparison of the rates of children with exacerbation of nephrotic syndrome during the 0 to + 180 period between the unstable group and the stable group (a). Comparison of the relapse rate between the precontrolled period and the 0 to + 180 period (b-d). a The rates of children with an exacerbation of nephrotic syndrome during the 0 to + 180 period in the unstable group (4/9 [44.4%]) was significantly higher than those in the stable group (4/54 [7.4%]; p = 0.01): b in 63 children, ns not significant, c in 9 children in the unstable group, ns not significant, d in 54 children in the stable group, ns not significant Discussion In this current prospective cohort study, we indicate that flu vaccination does not significantly precipitate the direct relapse of NS in children. However, it might increase the disease activity in children with at least one NS relapse within a half year before vaccination. Children who did not have an NS relapse during the precontrolled period also experienced no relapse during the postvaccination 0 to + 30 period. In addition, the percentage of patients with an exacerbation of NS relapse during the 0 to + 180 period was significantly higher in the unstable group than in the stable group, although the relapse rate from day 0 to + 180 did not increase compared with those during the precontrolled period in the unstable group. To the best of our knowledge, this is the first prospective study to evaluate the relationship between NS relapses and only flu vaccination (i.e., without other vaccinations). In this prospective study, we compared relapse rates during the postvaccination 0 to + 30 period with those during the precontrolled period and found no significant difference between the two periods. The immunogenic pathogenesis by which vaccination itself can ignite NS relapse was speculated to be T-cell activation [10]; however, the in-depth mechanism has not been elucidated. Recently, two studies have reported a minimal change disease following the Coronavirus 2019 vaccination [20, 21], and the authors speculate that the mechanism was T-cell–mediated podocyte injury in the immunologic response to the vaccine. A review of reports with minimal change disease after vaccination described a range between 4 days and 4 months from the time of vaccine receipt to the onset of clinical symptoms [22]. From another point of view, adverse events associated with various vaccines can be observed within 1 month from vaccine administration [6, 23]. From these background studies, we set the period within 1 month after receiving flu vaccination as having a direct association with flu vaccination. Our results suggest that, because the NS relapse rates were not significantly different between the precontrolled period and the 0 to + 30 period, children with NS can receive the inactivated flu vaccine without fear of potential NS relapses as a directly associated adverse effect. In our study, no child without NS relapse in the 6 months prior to flu vaccination experienced a relapse in NS within 1 month after receiving the flu vaccine. This finding might be useful for determining the timing of flu vaccine administration in children with NS, as it provides information regarding the substantial risk of NS relapse due to vaccination. In general, any vaccination can have several adverse effects. One such effect is an allergic reaction characterized by fever, local redness, swelling, pain, and feeling of heat resulting from an immunologic response [24]. Although the infection by upper respiratory viruses and presentation of clinical symptoms such as general fever in children with NS can elicit a relapse in NS [25], there have been no reports on the association between the allergic reaction of immunization as a side effect and the relapse of NS. In the present study, we found that among patients who experienced a side effect within 1 week of flu vaccination, the relapse rate during the postvaccination 0 to + 30 period was not significantly different from that during the precontrolled period. Even if an allergic reaction to vaccines is caused by an immunologic mechanism, there might be no association between the allergic immune response and NS relapse. In this study, there was a possibility that the administration of flu vaccines might have exacerbated the relapse of NS in children who had experienced more than one NS relapse within 6 months before immunization. Macdonald et al. prospectively evaluated upper respiratory infections in 32 children with NS and found an association with the exacerbation of NS [25]. They defined an unstable state of NS as an initial diagnosis of NS or relapse in the year before the study, and the stable state of NS as no relapses in the year before the study. Fifteen of 19 patients (79%) with unstable NS experienced an exacerbation of NS compared with only 4 of 13 patients (31%) with stable NS (p < 0.001). In our previous report on children with NS, taking glucocorticoids at the time of flu vaccination was a significant risk factor for NS relapse [14]. We argue that children receiving steroids at the time of flu vaccination indicates it occurred soon after the last NS relapse. Their results might indicate that the immunologic latent state in children with NS was unstable during the period immediately after the initial diagnosis of NS or the last relapse, and immunologic stimulation involving vaccinations might exacerbate the immunologic state after the relapse of NS. Our prospective evaluation showed that an exacerbation of NS relapse may depend on the most recent timing of relapse according to the rates of children with an exacerbation of NS; however, the relapse rate from day 0 to + 180 did not increase compared with those during the precontrolled period in the unstable group. Despite the fact that there was a gap as described above, we speculated that the direct relapse rate of NS was more important than the exacerbation rate 30 days after vaccination, and the exacerbation rate as the rate of children with an exacerbation of NS was more important than the relapse rate of NS from day 0 to + 180 after vaccination. This study has several limitations. The first is the relatively small number of NS relapses that occurred during the study period after the patients received the flu vaccine. However, we believe that this problem is acceptable, as our sample size of children with NS was the largest among all prospective studies evaluating the relationship between various vaccinations and NS relapse, and pediatric idiopathic NS is a rare disease. The second limitation is that the study population in the present study included only children receiving flu vaccines, and we could not compare the relative rate of NS relapse in vaccinated children with those in nonvaccinated children. A further study is required in which the primary outcome is the relative rate of NS relapse as compared between vaccinated and nonvaccinated patients. Third, we noted several potential confounders of NS relapse, including various infections and vaccinations, outside of flu infection and flu vaccination. We cannot exclude these confounders completely because the study period was between − 180 days before flu vaccination and + 30 days after flu vaccination, which covered fall to winter when several infections become prevalent. Moreover, we extracted the precise interval between each infectious event and various vaccinations and the timing of NS relapse. We considered that NS relapses within 1 month (particularly within 1 week) after infectious events or several vaccinations might be associated directly with these events. We evaluated these factors and their association with NS relapse. However, we observed no NS relapse within 1 month after infectious events or vaccinations, excepting flu infection and flu vaccination, from day + 0 to + 180. We have interpreted these confounders by evaluating our study results. Fourth, the statement on significantly increased relapse rates between the unstable and stable groups is not supported convincingly as the group sizes are very different (9 vs. 54). Finally, a selection bias might have been present, as a high proportion of patients with severe disease is frequently referred to our institutions. In conclusion, our prospective study indicates that flu vaccination did not significantly precipitate the direct relapse of NS in children. However, it might increase the disease activity in children with at least one NS relapse within a half year before vaccination. Supplementary Information Below is the link to the electronic supplementary material.Graphical Abstract (PPTX 151 KB) Supplementary file2 (DOCX 526 KB) Acknowledgements The authors thank Enago (www.enago.jp) for editing a draft of this manuscript. Author contribution S.I. prepared the manuscript. T.H., J.F., T.Y., N.M., N.K., and K.K. collected the clinical data. M.S., M.O., K.Is., K.Ii., and K.N. revised the article. All the authors have read and approved the final manuscript. Data Availability Data from this study can be obtained from the corresponding authors at reasonable request. Declarations Ethics approval All procedures performed in this study involving human participants were in accordance with the ethical standards set by the Ethics Board at Kakogawa Central City Hospital in which the study was conducted (approval number 29–12) and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Consent to participate Before enrollment, written informed consent was obtained from all patients and their parents for this study in accordance with guidelines. Competing interests K.K. has received research funding from the Terumo Foundation for Life Sciences and Arts, Public Foundation of Vaccination Research Center, and Taiju Life Social Welfare Foundation; donations from Chugai Pharmaceutical Co. Ltd., Astellas Pharma Inc., Ono Pharmaceutical Co., Teijin Pharma Ltd. Shionogi Co. Ltd., and Otsuka Pharmaceutical Co. Ltd.; and lecture fees from Tanabe Mitsubishi Pharma, Baxter Ltd., and Zenyaku Kogyo Co. Ltd. K. Ii. reports consultancy agreements with Zenyaku Kogyo Co. Ltd., Ono Pharmaceutical Co. Ltd., Kyowa Kirin Co. Ltd., JCR Pharmaceuticals Co. Ltd., Takeda Pharmaceutical Co. Ltd. and Sanofi K.K.; research funding from Zenyaku Kogyo Co. Ltd., Astellas Pharma Inc., Air Water Medical Inc., Otsuka Pharmaceutical Co. Ltd., Mochida Pharmaceutical Co. Ltd., Eisai Co. Ltd., Shionogi Co. Ltd., JCR Pharmaceuticals Co. Ltd. and Nihon Pharmaceutical Co. Ltd.; honoraria from Chugai Pharmaceutical Co. Ltd., Zenyaku Kogyo Co. Ltd., Kyowa Kirin Co. Ltd., Integrated Development Associates Co. Ltd., Astellas Pharma Inc. and Shionogi Co. Ltd.; a patent application on the development of antisense nucleotides for exon skipping therapy in Alport syndrome with Daiichi Sankyo Co. Ltd.; scientific advisor or membership as a member of editorial board of Pediatric Nephrology and Clinical Journal of the American Society of Nephrology. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References 1. Tarshish P Tobin JN Bernstein J Edelmann CM Jr Prognostic significance of the early course of minimal change nephrotic syndrome: report of the International Study of Kidney Disease in Children J Am Soc Nephrol 1997 8 769 776 10.1681/ASN.V85769 9176846 2. Sinha A Hari P Sharma PK Gulati A Disease course in steroid sensitive nephrotic syndrome Indian Pediatr 2012 49 881 887 10.1007/s13312-012-0220-4 22791676 3. Kidney Disease Improving Global Outcomes (KDIGO): The KDIGO practice guideline on glomerulonephritis: reading between the (guide) lines-application to the individual patient. Available at. https://kdigo.org/guidelines/gd/. Accessed 1 January 2021 4. Szajner-Milart I Zajaczkowska M Zinkiewicz Z Borzecka H Efficacy of vaccination against viral hepatitis type B in children with the nephrotic syndrome Ann univ Mariae Curie Sklodowska Med 2003 58 402 408 15315023 5. Furth SL Arbus GS Hogg R Tarver J Varicella vaccination in children with nephrotic syndrome: a report of the Southwest Pediatric Nephrology Study Group J Pediatr 2003 142 145 148 10.1067/mpd.2003.37 12584535 6. Alpay H Yildiz N Onar A Temizer H Varicella vaccination in children with steroid-sensitive nephrotic syndrome Pediatr Nephrol 2002 17 181 183 10.1007/s00467-001-0789-7 11956856 7. Abeyagunawardena AS Goldblatt D Andrews N Trompeter RS Risk of relapse after meningococcal C conjugate vaccine in nephrotic syndrome Lancet 2003 362 449 450 10.1016/S0140-6736(03)14072-X 12927434 8. Yildiz N Sever L Kasapcopur O Cullu F Hepatitis B virus vaccination in children with steroid sensitive nephrotic syndrome: immunogenicity and safety? Vaccine 2013 31 3309 3312 10.1016/j.vaccine.2013.05.004 23684838 9. Liakou CD Askiti V Mitsioni A Stefanidis CJ Safety and immunogenicity of booster immunization with 7-valent pneumococcal conjugate vaccine in children with idiopathic nephrotic syndrome Vaccine 2014 32 1394 1397 10.1016/j.vaccine.2013.11.106 24486348 10. Vivarelli M Massella L Ruggiello B Emma F Minimal change disease C J Am Soc Nephrol 2017 12 332 345 10.2215/CJN.05000516 11. Fernandes P Jorge S Lopes JA Relapse of nephrotic syndrome following the use of 2009 pandemic influenza A (H1N1) vaccine Am J Kidney Dis 2010 56 185 186 10.1053/j.ajkd.2010.04.011 20620684 12. Kim SR Lee SB Kim IY Lee DW Relapse of minimal change disease following infection with the 2009 pandemic influenza (H1N1) virus Clin Exp Nephrol 2012 16 329 332 10.1007/s10157-011-0562-6 22116504 13. Ishimori S Kamei K Ando T Yoshikawa T Influenza virus vaccination in children with nephrotic syndrome: insignificant risk of relapse Clin Exp Nephrol 2020 24 1069 1076 10.1007/s10157-020-01930-8 32720203 14. Ishimori S Ando T Kikunaga K Terano C Influenza virus vaccination in pediatric nephrotic syndrome significantly reduces rate of relapse and virus infection as assessed in a nationwide survey Sci Rep 2021 11 23305 10.1038/s41598-021-02644-x 34857817 15. Angeletti A Bruschi M Bianchin S Bonato I Vaccines and disease relapses in children with nephrotic syndrome Clin J Am Soc Nephrol 2021 16 937 938 10.2215/CJN.01890221 34117084 16. Ishikura K Matsumoto S Sako M Tsuruga K Clinical practice guideline for pediatric idiopathic nephrotic syndrome 2013: medical therapy Clin Exp Nephrol 2015 19 6 33 10.1007/s10157-014-1030-x 25653046 17. Kaku Y Ohtsuka Y Komatsu Y Ohta T Clinical practice guideline for pediatric idiopathic nephrotic syndrome 2013: general therapy Clin Exp Nephrol 2015 19 34 53 10.1007/s10157-014-1031-9 25653047 18. van Husen M Kemper MJ New therapies in steroid-sensitive and steroid-resistant idiopathic nephrotic syndrome Pediatr Nephrol 2011 26 881 892 10.1007/s00467-010-1717-5 21229269 19. Yoshikawa N Nakanishi K Sako M Oba MS A multicenter randomized trial indicates initial prednisolone treatment for childhood nephrotic syndrome for two months is not inferior to six-month treatment Kidney Int 2015 87 225 232 10.1038/ki.2014.260 25054775 20. Leclerc S Royal V Lamarche C Laurin LP Minimal change disease with severe acute kidney injury following the Oxford-AstraZeneca COVID-19 vaccine: a case report Am J Kidney Dis 2021 78 607 610 10.1053/j.ajkd.2021.06.008 34242687 21. Lebedev L Sapojnikov M Wechsler A Varadi-Levi R Minimal change following the Pfizer-BioNTech COVID-19 vaccine Am J Kidney Dis 2021 78 142 145 10.1053/j.ajkd.2021.03.010 33839200 22. Clajus C Spiegel J Brocker V Chatzikyrkou C Minimal change nephrotic syndrome in an 82 year old patient following a tetanus-diphteria-poliomyelitis-vaccination BMC Nephrol 2009 10 21 10.1186/1471-2369-10-21 19656382 23. Long B Bridwell R Gottlieb M Thrombosis with thrombocytopenia syndrome associated with COVID-19 vaccines Am J Emerg Med 2021 49 58 61 10.1016/j.ajem.2021.05.054 34062319 24. Johansson SGO Bieber T Dahl R Friedmann PS Revised nomenclature for allergy for global use: report of the Nomenclature Review Committee of the World Allergy Organization, October 2003 J Allergy Clin Immunol 2004 113 832 836 10.1016/j.jaci.2003.12.591 15131563 25. MacDonald NE Wolfish N Mclaine P Phipps P Role of respiratory viruses in exacerbations of primary nephrotic syndrome J Pediatr 1986 108 378 382 10.1016/S0022-3476(86)80876-9 3005537	36449102	PMC9709736	NO-CC CODE	2022-12-01 23:23:39	no	Pediatr Nephrol. 2022 Nov 30;:1-10	utf-8	Pediatr Nephrol	2,022	10.1007/s00467-022-05783-z	oa_other
==== Front Neotrop Entomol Neotrop Entomol Neotropical Entomology 1519-566X 1678-8052 Springer International Publishing Cham 36449176 1005 10.1007/s13744-022-01005-1 Biological Control in Latin America Conservation Biological Control as an Important Tool in the Neotropical Region http://orcid.org/0000-0002-8273-0140 Vargas German [email protected] 1 Rivera-Pedroza Leonardo F. [email protected] 1 García Luis F. [email protected] 2 Jahnke Simone Mundstock [email protected] 3 1 Colombian Sugarcane Research Center (Cenicaña), San Antonio de los Caballeros, Vía Cali-Florida Km 26, Valle del Cauca, Colombia 2 grid.11630.35 0000000121657640 Northeastern Regional University Center, University of the Republic, Rivera, Uruguay 3 grid.8532.c 0000 0001 2200 7498 Postgraduate Program in Plant Science, Faculty of Agronomy, Federal University of Rio Grande Do Sul (UFRGS), Phytosanitary Dept, Porto Alegre, Rio Grande Do Sul Brazil Edited by Rogerio Biaggioni Lopes. 30 11 2022 118 1 6 2022 8 11 2022 © Sociedade Entomológica do Brasil 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 history and recent developments of conservation biological control (CBC) in the context of industrialized and small-scale agriculture are discussed from theoretical framework available in the Neotropical region. A historical perspective is presented in terms of the transition of the way pests have been controlled since ancestral times, while some of these techniques persist in some areas cultivated on a small-scale agriculture. The context of industrialized agriculture sets the stage for the transition from chemical pesticides promoted in the green revolution to the more modern concept of IPM and finds in conservation biological an important strategy in relation to more sustainable pest management options meeting new consumer demands for cleaner products and services. However, it also noted that conservation, considered within a more integrative approach, establishes its foundations on an overall increase in floral biodiversity, that is, transversal to both small-scale and industrialized areas. In the latter case, we present examples where industrialized agriculture is implementing valuable efforts in the direction of conservation and new technologies are envisioned within more sustainable plant production systems and organizational commitment having that conservation biological control has become instrumental to environmental management plans. In addition, a metanalysis on the principal organisms associated with conservation efforts is presented. Here, we found that hymenopteran parasitoids resulted in the most studied group, followed by predators, where arachnids constitute a well-represented group, while predatory vertebrates are neglected in terms of reports on CBC. Our final remarks describe new avenues of research needed and highlight the need of cooperation networks to propose research, public outreach, and adoption as strategic to educate costumers and participants on the importance of conservation as main tool in sustainable pest management. Supplementary Information The online version contains supplementary material available at 10.1007/s13744-022-01005-1. Keywords Small-scale agriculture Large-scale agriculture Meta-analysis Entomophagous Natural enemies ==== Body pmcThe history and recent developments of the conservation biological control From the green revolution to pest management programs focused on biological control Diseases and pests in the agricultural sector cause different types of yield losses depending on the crop and the causing agent. Therefore, farmers are continuously searching for efficient and affordable alternatives of control, chemical insecticides, or chemical pesticides currently stand out as the main alternative for agricultural pest control, although are not necessarily the most sustainable (Bernardes et al. 2015; Daam et al. 2019; Nicholls 2008). The group of substances known as pesticides pertains to compounds used as insecticides, fungicides, herbicides, rodenticides, molluscicides, and nematicides (Bernardes et al. 2015). The use of pesticides has been considerably higher in the twenty-first century. The exponential growth of the human population in the eighteenth century and the increase of agricultural production, also known as the green revolution of the twentieth century, in which technological improvements were produced that even led to doubling the yield of some cereals for mass consumption (Bernardes et al. 2015; Khush 1995), were based on products of non-specific toxicity for massive use (Martínez-Centeno and Huerta 2018). During this period, changes such as the increased use of pesticides, inorganics fertilizers, and water were implemented in agricultural production, favoring the operation or large-scale intensive monoculture systems (Khush 1995; Romero 2012). Chemical pesticides have been intensively used to prevent crop yield losses and increase efficiency in pest management (Mahmood et al. 2016). Its worldwide pesticide has increased its production at a rate of about 11% per year, from 0.2 million tons in the 1950s to more than 5 million tons by 2000 (Carvalho 2017). However, since the last century, several countries have warned of adverse effects on humans and non-target organisms, mainly caused by improper use or accidents (Rosival 1985; Carvalho 2017). And although efforts have been made to evaluate the effect on aquatic ecosystems, there is little information on the impact on the structure and function of terrestrial ecosystems (Daam et al. 2019), while there is also no full guarantee on its efficiency. Even so, the increase in the use of pesticides has not been adequately followed by studies that verify their environmental effects in tropical countries, and therefore, the requirements for their use are not available, are unclear, or are implemented and applied in an inadequate way, whereas in North America, Europe, and other countries such as Japan and Australia, there is a well-established legal framework (Daam et al. 2019). On the contrary, the frequent outbreaks of secondary pests after applications and the increasing prevalence of resistant pests have pointed the risks of dependence on these substances in relation to the environment, biodiversity, and human health (Bernardes et al. 2015; Boedeker et al. 2020; Bruinsma 2003; Del Puerto Rodríguez et al. 2014; Mahmood et al. 2016; Ruberson et al. 1998; WHO/OPS 1993). In addition, agrochemicals, mechanization, and irrigation operations, which are at the heart of industrial agriculture, are highly dependent on fossil fuels, are increasingly scarce, and, consequently, are more expensive (Altieri and Nicholls 2012). In perspective, the future of pesticides seems to tend towards less harmful chemical groups capable of selectively acting on specific targets while sparing on other components of the agroecosystem (Bruinsma 2003; Matthews 2015; Mahmood et al. 2016). During the last years, in different regions of the world, the focus of agricultural production, addressing environmental and socioeconomical impacts, has been reconsidered towards a more ancestral form of pest management (i.e., traditional), evolving within the framework of new scientific and technological knowledge. It is perceived as necessary to reduce risks and consider other more sustainable and more effective production modes (Tilman et al. 2002). New practices in agriculture include small changes such as increase diversity of plants associated with the crop and chemical pesticide elimination or avoiding herbicides to help increase biodiverse populations that include natural enemies (NE), such as parasitoids, predators, pathogens, and antagonists (Letourneau et al. 2011; Nicholls 2008; Peñalver-Cruz et al. 2019). This approach does not eliminate rational use of pesticides, resistant crop varieties, and ecological pest control (Bruinsma 2003), but adds natural products and systems as an alternative to chemical pesticides (Chirinos et al. 2020; Souto et al. 2021; Muhammad et al. 2022). From the integrated pest management (IPM) and ecological pest management (EPM) perspectives, to sustain productivity with minimum adverse effect on the environment (especially on pollinators and other beneficial fauna) (Egan et al. 2020), biological control has the potential of becoming the main strategy for pest management in intensive agriculture (UNODC 2010). This approach emerged for prevention, monitoring, and rapid identification of the pest(s) and disease(s) and their effective control by a coordinated use of all accessible technologies. For this reason, through a systematic knowledge on crop biology and the biology of their pests and diseases, it uses its NE to suppress them (UNODC 2010; Nicholls 2008; Owen et al., 2015). The biological control concept has been developed primarily by entomologists and is normally associated to the use of living NE to control pests, either through importation of exotic NE against either exotic or native pests (i.e., classical biological control), augmentation of already existing NE, and conservation of such existing NE (Ehler 1998). Conservation biological control (CBC) is a strategy that seeks to modify the environment biodiversity to provide habitat and resources that protect and enhance NE to reduce pest’s effects in crop systems (DeBach 1964; Moses 1993). Such modification may be directed at mitigating harmful conditions or enhancing favorable ones (Landis et al. 2000). The core of the strategy is the biotic interactions between pests and their wild NE, whether they are predators, parasitoids, or entomopathogens. The premise is that if habitat loss and environmental changes associated with intensive agricultural practices are compensated, NE are conserved and therefore pest population would be controlled. The possibility of incrementing effective beneficial arthropods populations will depend on the availability of food, refuge, and other resources with in and around the crop field (Huffaker and Messenger 1976). Therefore, CBC is a sustainable tool compatible with agroecological, organic, and other forms of agriculture based on the increase of natural agroecological processes and the elimination or reduction of chemical agents (Nicholls 2008), which have long overshadowed the importance of NE in pest management. It has been used to reverse harmful effects of intensive practices in agriculture, such as disturbance associated with extensive use of pesticides (Paredes et al. 2013), tillage, burning, and other agronomic interventions (Barbosa 1998). It differs from other biological control approaches in the way that is does not rely on massive releases of organisms, but in promoting a set of environment interventions, so biodiversity and crop can thrive (Barbosa 1998). However, even though pesticides and NE have often been viewed as incompatible, in some cases, there have been programs that develop pesticides that have minimal environmental impact and exhibit greater selectivity to NE (Ruberson et al. 1998), especially in agroecosystems highly dependent on chemical control (Torres and Bueno 2018). However, multiple field experiments have shown that pesticide applications do not significantly reduce pest density but do affect NE in most cases (Janssen and van Rijn 2021). Therefore, its main trend points to the reduction of pesticides from the management program to improve the abundance of NE in the crop, limiting pest infestations. In general, CBC does not exercise pest control directly; instead, it promotes the abundance and diversity of NE which are already present in the agroecosystem (Paredes et al. 2013), so it becomes a preventive measure within an integrative pest management system and an integrative crop production approach (Begg et al. 2017; Díaz et al. 2018). In this regard, NE populations’ responses to conservation strategies are not always consistent with pest suppression or the increase of crop yields, so it has been relegated and rarely used in commercial crop production (Begg et al. 2017) . The general scenario of conservation biological control in the Neotropical region The Neotropical region is the most biodiverse ecoregion and highly productive in terms of agricultural outputs. Therefore, it is an important scenario for biological control programs, since part of this great diversity are NE of different kinds (Smith and Bellotti 1996). In this region, CBC programs are strategic to conserve biodiversity in landscapes dominated by agriculture, since most of their practices involve the conservation of refuges for associated fauna providing protection against environmental threats (e.g., high temperatures, drought, and climate change), as well as biological (e.g., pest outbreaks) and anthropic threats (e.g., habitat loss and hunting) (Selwood and Zimmer 2020). Although conservation of NE is probably the oldest form of biological control (Barbosa 1998), in the Neotropical region, the situation of CBC seems paradoxical as it has been widely practiced by traditional farmers, but it is the least recognized (i.e., documented) and least financially supported, at least at the level of small-scale farming. In addition, it tends to be displaced by chemical insecticides. However, in some Latin American contexts, traditional agriculture is distinguished by the diversity it establishes in agricultural designs, in the diversity of plant species and genotypes, thus suffering less frequent pest attacks than those observed in monocultures (Trujillo 1992; Altieri and Nicholls 2012). Many of the practices carried out in traditional agriculture that are observed in the Neotropical region could be incorporated into modern designs of agricultural production. But various factors limit its implementation, for example, the strong focus on classical biological control and/or augmentation, with special emphasis on the use of parasitoids (Colmenarez et al. 2018), which is the most widespread and credible practice. In addition, many of the successful cases of CBC are not documented, since it is a daily practice of traditional farmers (Trujillo 1992), which limits producers to have more access to implementation methods and to understand their benefits. The latter does not contribute to a better understanding of the response of NE on pests (Begg et al. 2017) and confronted by short-term expectations, limits its practice in traditional small-scale crops. Extensive deforestation in the late twentieth century threatens the diversity of the Neotropics (Rull 2011). The reduction of species could deteriorate the processes of the ecosystem while the increase in diversity leads to more stability (Elton 1958). Agricultural and forestry activities seem to indicate that systems poor in species are not very stable; therefore, agricultural practices that conserve functional biodiversity are relevant. CBC models that include natural vegetation associated with crops allow the presence of a high diversity of organisms and functional groups that can act as pest regulators in the crop, while additionally being a form of conservation (Rivera-Pedroza et al. 2019). Understanding the ecosystem guidelines of the CBC and overcoming immediate expectations to control pests could expand the use of this practice as a fundamental tool of IPM programs in the Neotropics. In this sense, CBC requires an articulation to integrated crop management practices, and a more holistic approach considering landscape implications on different scenarios and the environmental effects of establishing or sustaining any agricultural activity. In this regard, if we consider farming in general as an agribusiness activity, all regulations and modern administrative methodologies apply in relation to formulating agricultural projects by ensuring the well-being of workers, final costumers, and the environment in general. Even though the latter could be attributed mainly to the socioeconomical context of industrial agriculture, Environmental Management Plans are benefiting from the concepts and implementation of CBC in relation to the mitigation of potentially adverse environmental impacts of the different agricultural projects. Of course, CBC techniques differ greatly between geographic areas and between extensive agricultural management or in peasant agriculture, especially because, in Latin America, what is considered extensive or small-area agriculture also differs from one country to another. In this review, we consider industrialized agriculture as those areas in which most of the labor is hired and with intensive use of agricultural machinery, high technology, and industrialization of the entire production process, including the final commercialization of the product. On the other hand, we consider small-scale agriculture, those areas depending on family labor, or peasant agriculture, with the use of traditional technology, often associated with community groups. Here, we are discussing the potential of CBC under small and industrialized scale agriculture and indicating which are the principal organisms associated with conservation in the Neotropical region. Conservation biological control in small-scale agriculture in the Neotropical region Ancestral ways of dealing with pests If we look for historical references of agricultural production after European colonization in small rural properties, or even in gardens in urban areas in Latin America, we will find records of environmental architecture seeking to associate food, medicinal and ornamental plants, and animals for human consumption. In the historical context, the concepts of vegetable garden, backyard, orchard, flower garden, and leisure area were mixed (Carneiro and Bertruy 2009). This type of environmental management, registered in different Latin American countries, sought firstly, to increase productivity and food diversification in small areas. Indirectly, these spaces maximized the ecological services provided by an associated biota that obviously included biological pest control (Chapling-Kramer and Kremer 2012). Empirical experiments, carried out by different ethnic groups and passed on orally to several generations, produced different spatial arrangements for subsistence purposes. The knowledge of native peoples was incorporated, adding many elements of native flora to the diet and, consequently, to vegetable gardens, orchards, and cultivated gardens (Melo and Melo 2015). Examples of these spaces can still be found in peasant agricultural communities, presenting a great diversity of agricultural systems, widely associated with the vernacular landscape of each place. The term “vernacular landscape” was proposed by Jackson (1984) and refers to landscapes resulting from successive interactions between local communities and their natural environment, including the genetic variety of plants and animals available and presenting a primary connection with functionality, of where the transformation of the local material takes place (Petry 2014; Frank and Yamaki 2018). Thus, although each region has particularities associated with native flora and fauna and the ethnic elements present, the basic structure related to functional diversity is always present, as in the area used as an example in Fig. 1.Fig. 1 Sketch representing an agroecological farm with family labor with small production plots and a mosaic of habitats (cultivated and uncultivated) as well as refuge areas (hedges, forests) in southern Brazil When agroecology studies began to emerge in the Americas, in the late 1970s and 1980s, pioneered by researchers such as Altieri, Gliessman, Ana Primavesi, and Lutzenberg among others, these models of agriculture started to be rescued and studied (Lutzenberger 1978; Primavesi 1984a, b; Altieri 1989; Gliessman 1998). The beneficial effects of diversity and the importance of plant architecture in the agroecosystem started to be accounted for. The hypothesis connecting ecological mechanisms for pest suppression via habitat diversification was solidified by Root (1973) and was one of the contributions that most influenced the discipline of CBC through habitat manipulation. Thus, in the Neotropical region, the bases for CBC were first established in small rural properties, with family and peasant labor. Many aspects of the ancient agricultural management could be considered CBC, but those management practices are rarely identified as such. One of these ancient practices includes the increment of plant diversity through intercropping (Thrupp 2000). For example, agroforestry systems containing dozens of plant species are common in South America. Likewise, since prehispanic times, it has been quite common to intercrop maize with other plant species (especially leguminous species) (Thrupp 2000). An integrative production system Studies related to CBC in small farms, which encompass different production systems, are much more recent (Wyckhuys and O’Neil 2007). This kind of research has increased recently in some developing countries such as Brazil and Cuba (Wyckhuys et al. 2013), but is rather lacking in others. These works have been looking to explore our own biodiversity and understand how the spatial arrangements and genetic variety of plants influence the ecological services provided by beneficial organisms, especially NE (Wyckhuys et al. 2013; Peñalver-Cruz et al. 2019). Our focus in this review will be on arthropods with entomophagous habits, NE of insects that attack cultivated plants and their effects on the management of these pests. Let us use as a basis the sketch shown in Fig. 1. This sketch was based on a 12-ha rural property, Sítio Aroeira, located in the municipality of Muitos Capões, in Rio Grande do Sul, in the extreme south of Brazil. In this area, there is a large part covered with native forest that surrounds and intermediates the property. The property is family-based, certified for organic production, and maintains ecologically based cultivation systems (agroecology) with different managements in the area. It is possible to observe a great diversity of crops intermingled with natural environments such as forest, prairie, and wetlands (Fig. 1). Understanding how these environments increase the presence and permanence of NE and other beneficial insects in agricultural areas is one of the main focuses of studies in CBC. Crop diversification and implementation of additional resources for NE and the impact of this on pest populations and crop productivity is another major focus of research. According to Heimpel and Mills (2017), the two most recorded classes of habitat manipulation are “improving michohabitats for NE” and “resource supplementation.” The species richness of NE in an environment, by itself, may not be sufficient to suppress pest populations (Letourneau et al. 2009), being necessary to know the NE associated with pests, since the species composition is fundamental (Alhadidi et al. 2018). Works that seek to identify potential NE of agricultural pests are developed in different countries in Latin America and may constitute a first step towards the implementation of CBC. An example of this is the work carried out by the PROINPA Foundation in Bolivia whose project seeks technological strategies for the sustainable management of quinoa in the Bolivian Altiplan. These surveys, however, are often restricted to technical bulletins or project reports (PROINPA 2013). Surveys of predators and parasitoids have been carried out throughout Latin America for pests belonging to different orders, such as Lepidoptera (Gómez-Jiménez 2018; Vargas 2018 – Colombia; Ribeiro et al. 2015 – Uruguay), Hemiptera (Pavis et al. 2003 – Guadaloupe; Ribeiro and Castiglioni 2008 – Uruguay; Gaimari, et al. 2012 – Colombia), and Diptera (Ovruski et al. 2000 – Latin America and Southern United States; García-Cancino et al. 2015 – Mexico; Hernandez-Mahecha et al. 2018 – Colombia) in addition to other herbivorous insects. Also in Paraguay, surveys of NE have been carried out in small areas, on crops such as peanuts, sesame, beans, and corn (Cabral-Antúnez, et al. 2020). More detailed descriptions of NE surveys are in an additional section of this document. In Fig. 1, we can also see several small forest areas, called “legal reserve” by Brazilian Law 12.651/12 (Brasil 2012), interspersed with crop areas. These wild vegetation areas can serve as a repository of NE, for both parasitoids (da Silva et al. 2016, 2019) and predators (Ferreira et al. 2014; Medeiros et al. 2018) of insect crop pests. Studies evaluating the effect of wild vegetation areas on the composition of beneficial arthropod fauna, and especially NE, are carried out as one of the first assumptions for the CBC. For example, on small coffee farms in Colombia, Armbrecht and Gallego (2007) evaluated the diversity of predatory ants and the predation rate in plantations shaded by native forest compared to plantations in the sun. Comparison of the parasitoid fauna (da Silva et al. 2016) and predators (Ferreira et al. 2014) between an area of preserved native vegetation and organic rice alongside showed that around 40% of the entomophagous species are shared. Both studies indicate the importance of native vegetation for the biological control of conservation at the site. Adaime et al. (2018) also show the importance of fruit trees that occur naturally in preserved areas for the maintenance of parasitoid of Anastrepha spp. (fruit fly) (Tephritidae) in the Brazilian Amazon. In agroforestry systems in Costa Rica, plants with alternative hosts of hymenopteran parasitoids of borers and other sap-sucking pests in pecan nut crops also have been identified (Mexzón 2001), showing that plant diversity in an agroecosystem can influence the abundance of predators (Barbosa and Wratten 1998), as well as parasitoids, for which plants act as mediators of chemical communication, refuge, and food resource (Barbosa and Benrey 1998). The evaluation of the effect of non-target plants in fallow areas, in windbreaks or simply on the edges of crops and their management in the CBC, has been the subject of many studies. Predators of the Chrysopidae and Coccinellidae families, for example, circulate between pastoral systems and areas of adjacent managed vegetation in Uruguay and do pest control on aphids and grazing mites (Ribeiro 2010). In organic management properties in Brazil, with production of at least 16 types of species of vegetable, a great richness of predators and herbivores in non-cropped habitats was recorded. Furthermore, all the habitats share species of natural NE throughout the year, indicating that species could disperse among habitats (Togni et al. 2016, 2019). Amaral et al. (2013) had already evaluated the role of non-cultivable weeds for the maintenance of aphidophagous predators in tropical agroecosystems associated with chilli pepper crops. The authors suggest that the management of specific weed species can provide an optimal strategy for the conservation of beneficial insects that use non-predatory foods. Non-crop habitats were also investigated in Chile and the authors showed that some parasitoids such as Aphelinus mali (Haldeman) (Hymenoptera: Aphelinidae) precociously colonize apple orchards maximizing the control of Eriosoma lanigerum (Hausmann) (Hemiptera: Aphididae) near edges with the presence of wild plants (Peñalver-Cruz et al. 2020). Floral resources close to or between cultivated areas (Fig. 1) are a common characteristic among small properties and rural labor throughout Latin America. The importance of flowers in attracting and maintaining NE, especially parasitoids, is widely studied (Pfiffner and Wyss 2004). The introduction of floral resources, in the already known flower strips, or in different arrangements, associated with crops, and their effects on the diversity of NE and on pest control have been the focus of many studies in Latin America. In Cuba, this focus has been explored and studied, with works that contribute to the knowledge of potential botanical species for the maintenance and reservoir of NE and other beneficial insects, especially in urban agriculture (Matienzo et al. 2007; Ceballos et al. 2009). In Colombia, the performance of Trichogramma atopovirilia Oatman and Platner (Hymenoptera: Trichogrammatidae) was studied according to the presence of different flower species and found that the presence of Trifolium pratense L. (Fabales: Fabaceae) was the most adequate to optimize the parasitoid in the field (Díaz et al. 2012). Floral biodiversity on small-scale farms The temporal and spatial variation of floral communities marginal to agroecosystems and the spatial and seasonal diversity of associated insects were evaluated in vegetable gardens in Cordoba, Argentina (Rojas-Rodrigues et al. 2019). This work shows that the abundance of insects in different plants increased significantly with the number of samplings where floral species were present. In Brazil, a study showed that Alisso, Lobularia maritima (L.) (Brassicales: Brassicaceae) strips between rows with beds of Brassica oleracea (L.) (Brassicales: Brassicaceae), contributed to increase the abundance of generalist predators which translated into a significant reduction of collards pests, especially aphids (Ribeiro and Gontijo 2016). Haro et al. (2018) also showed that there is a variation in the richness and abundance of specialist and generalist NE in different periods of development of non-target plants such as Tagetes erecta L. (Asterales: Asteraceae) in crops. The flowering period favored greater complexity in the food web, increasing the functional stability of the community in the agroecosystem. In addition to the specific composition of the plants, the strips of wild plants must be integrated into the culture in a practical way for producers and distributed in space and time so that they favor NE (Jahnke and da Silva 2021). Thus, there are suggestions for arranging wildflowers and plants between crops, such as those made by Aguiar-Menezes and Silva (2011), in a technical bulletin, aimed at farmers. Crops in consortium and agroforestry systems have also been the target of studies related to CBC, especially in peasant agriculture with little tech. In vegetable production, there is a record of consortia of sweet pepper, Capsicum annuum L. (Solanales: Solanaceae), associated with basil, Ocimum basilicum L. (Lamiales: Lamiaceae), and marigold, Tagetes erecta L. (Asterales: Asteraceae) that significantly increased the presence of parasitoids compared to single pepper crop (Souza et al. 2018). In a perennial plant production system, the beneficial effects of the associated diversity become even more explicit, since a brief ecological succession is allowed in the area, making the agricultural system to be situated in an associative phase concerning the community (Gliessman 2001). At this stage of development, more adapted species manage to establish themselves and occupy important niches for the system’s functionality (Odum 1988). In this way, systems served as a source of NE that can colonize horticultural crops when herbivores are present. Consequently, NE can establish a numerical response to herbivore abundance (Harterreiten-Souza et al. 2014). This effect was observed by evaluating the action of different plant covers on the arthropod community in apple orchards in Argentina (Fernández et al. 2008) or by evaluating the structure of the insect community concerning the integration of plant crops in agroforestry in Brazil (Harterreiten-Souza 2014). Also, Hoshino et al. (2018) found a lower infestation of the coffee leaf miner (CLM) Leucoptera coffeella (Guérin-Mèneville and Perrottet) (Lepidoptera: Lyonetiidae) in an organic coffee crop intercropped with pigeon pea plants, Cajanus cajan (L. Millsp.) (Fabales: Fabaceae), and greater predation by wasps in the presence of Leucaena plants, Leucaena leucocephala L. (Fabales: Fabaceae). More recently, studies related to CBC have focused on population dynamics and spatial distribution of NE between areas. Bidirectional movement of aphid parasitoids between cultivated and uncultivated plants was documented in Argentina (Zumoffen et al. 2017). In this work, variables that interfere in this dynamic were described, indicating that the abundance of alternative host aphids is decisive, thus inferring that the natural vegetation has an important role in pest control. The evaluation of quantitative indices showed that trophic networks between legume-aphid-parasitoids and entomopathogens vary in different plants cultivated in Uruguay, but the abundance of hosts and fungi is determinant for the structure of communities (Silva 2016). Finally, the understanding of chemical interactions involved in ecological processes related to CBC has been gaining ground. Herbivory-induced plant volatiles (HIPV) play an important role in tritrophic relationships in most ecosystems and have been studied with different approaches (Ponzio et al. 2013; Becker et al. 2015). This focus was given in the work of Togni et al. (2016) who explored the mechanisms involved in the attractiveness of aphidophagous Coccinelids to Coriandrum sativum L. (Apiales: Apieaceae). Also in Brazil, Ulhoa et al. (2020) described that rice plants produce different volatiles in response to damage caused by two species of stink bugs, bringing to the plant protection to conspecific herbivores. However, parasitoids can recognize and respond positively to both stimuli. In Mexico, semi-field studies carried out in tents showed that teosinte, Z. mays spp. Parviglumis (Poales: Poaceae) produce volatiles in the presence of Spodoptera frugiperda (J.E. Smith) (Lepidoptera: Noctuidae) which attract parasitoids (Lange et al. 2018). The authors commented however that the discussion about the adaptive function of HPIVs was not completely clarified. Although the number of studies on different aspects associated with CBC has been growing, especially in small areas, many interactions are still not well understood. In agroecosystems where several components of the landscape, as in the examples we present, act as refuges (e.g., in diversified agricultural ecosystems), resource availability can vary among components or patches, allowing NE to move from one patch to another, between others. Furthermore, the specificities of each system, related to abiotic and biotic environmental issues, in addition to the entire cultural and ethnic framework, are involved. Thus, these studies must be expanded precisely in the socio-economic and environmental conditions of the Neotropical region as a whole, to seek applicable solutions for the construction of environmentally responsible and economically viable agriculture. Conservation biological control in industrialized agriculture in the Neotropical region New trends in final costumers’ behavior drive the change It is possible to suggest that no other context brings more simplification of the landscape structure than industrialized agriculture, where large-scale, intensive production, often involving rutinary use of inorganic fertilizers and chemical pesticides, and synchronous monocultures result in an obvious obstacle for either planned or incidental plant diversity, this in comparison to small-scale agriculture (Rusch et al. 2014; Michaud 2018). In addition, these monotonous areas could also be facing additional pest pressures, this by the massive growing of vulnerable crop cultivars that promotes higher dependency on chemical control, driving the system in the selection for more persistent and/or insecticide-resistant herbivore pests. The latter challenges efforts on implementing biological control programs and more even the idea of CBC. However, it has been also argued that the implementation of agronomic practices is more prevalent at structuring insect communities than the landscape-level complexity (Rusch et al. 2014), so recent changes in certain agronomic practices, such as pest-resistant varieties, that require less use of pesticides; the implementation of soil conservation practices such as reduced till or not-tillage practices; and the development of more selective insecticides have the potential of benefiting the implementation of biological control (Michaud 2018). The context of public awareness of more sustainable production of goods and services has promoted the use of certification of good practices in agriculture, meaning that industrial production of food is also heading into the direction of bringing to their consumers’ “cleaner” products. The Neotropical region, composed mainly by developing countries with economies heavily relying on exports, is facing the challenge of adapting production systems to consumer demands, so more cost-effective ways of pest control are implemented. Therefore, CBC appears as a strategic opportunity for agribusiness companies seeking both, integrated pest management and social marketing of their brands. Here, we reviewed some cases of large-scale agriculture where CBC has become an option in the Neotropics, and whose adaptations via habitat manipulation and changes in agronomic practices have the potential to influence larger areas, including those under small-scale agriculture. Industrialized agriculture examples on efforts towards CBC in the Neotropics Palm oil could be often associated with the destruction of natural habitats and the reduction of flora and fauna populations (Vijay et al. 2016); however, these are not necessarily the characteristics of the whole industry. The Round Table for Sustainable Palm Oil Supply Chain Certification (RSPO) has opened the door for initiatives aiming oil palm production under following environmental and social standards, this by defining agronomic recommendations towards the preservation of ecosystem services, all the latter under the scope of guidance documents containing the respective Environmental Management Plans. In this regard, Mexson and Chinchilla (2003) discussing pest management alternatives for oil palm in Costa Rica mentioned that pest’s pressure in oil palm changes with crop age. In early crop ages, the abundance of other plant resources could be the reason to a higher NE’s abundance and a reduction on pest insects. However, as the crop progresses in age, plant diversity is reduced, and number of pest insects increases. The latter authors described 30 broad leaf weed species attractive to beneficial insects, suggesting the need to preserve them in corridors nearby the cultivated palms, so biological control service can be provided to the crop. In Colombia, according to Bustillo-Pardey (2014), the areas grown with palm oil were experiencing an increase in the early 2010s, achieving approximately 470,000 hectares. Several pests have been reported ranging from leaf feeders, sap-sucking in foliage and fruit, stem borers, among others. However, under these adverse circumstances, the goal has been to establish CBC, taking advantage of a great arrange of beneficial insects in the palm oil system (Calvache 1995). Palm oil besides being cultivated in large scale also represents a perennial crop that is expected to last at least 25 years, resembling forestal exploitations, where regulation of pests requires a landscape approach in relation to the extent of areas to manage, and here is where CBC is playing a critical role. A Colombian cooperative business structure denominated Fedepalma through their self-financed research institute Cenipalma, has proposed and engaged in a sectorial project–denominated Biodiverse Palm Landscape (Paisaje Palmero Biodiverso) (Fedepalma 2016) with the idea of formulating basic plans to preserve and stimulate the conservation of biodiversity in the plantations and the implementation of agroecological practices. In this regard, the growth of plants that present extra-floral nectaries around the cultivated areas has been oriented to stimulate biodiversity and beneficial insects. Important economic efforts were reported by Aldana et al. (2004) in relation to the establishment of nurseries to grow 13 species of plants where a total of 15 parasitoid species were reported, with increases in the level of parasitism in pests of economic importance such as Stenoma cecropia Meyrick (Lepidoptera: Elachistidae) (Aldana et al. 1997) a leaf feeder that has demonstrated a great number of NE in the palm oil agroecosystem (Sendoya-Corrales and Bustillo-Pardey 2016). Some extensive areas of coffee are planted in the neotropics, where also many small-scale coffee farms are in the neighboring areas. Here, the idea of CBC it has been also implemented, not necessarily with the purpose of controlling pests, but to favor diversity on agroforestry approaches, where coffee agroforestry has proven to be important at sustaining biodiversity and ecosystem services, and where intensification of crop production, usually associated to unshaded coffee, has risen concerns in terms of loss of diversity (Perfecto et al. 1996; Philpott et al. 2008). Aside from the discussion on the replacement of traditional coffee agroforestry for more intensive production systems and its effects on meeting a growing demand for coffee, shaded coffee has demonstrated multiple attributes in the sustainability of the production in the long term, such as soil conservation, reduction of plant physiological stress, and pest regulation (Haggar et al. 2021). In this regard, Armbrecht and Gallego (2007) found in Colombia that a more diverse soil dwelling community of ants removed significantly more coffee berry borer (CBB) individuals, Hypothenemus hampei Ferrari (Coleoptera: Curculionidae), in shaded coffee than in unshaded lots. Similar studies in Mexico have demonstrated that shaded coffee agroecosystems at harboring a more diverse ant community regulate infestations (Larsen and Philpott 2010; Morris et al. 2015), regulation that is also associated with interspecific interactions among the more dominant ant species in those communities as Newson et al. (2021) demonstrated in agroforestry coffee productions in Puerto Rico. It is noteworthy that the beneficial role of ants on the CBB biocontrol has been demonstrated even during coffee postharvest process (Velez-Hoyos et al. 2006). In addition, studies in Jamaica demonstrated the beneficial role of shaded coffee, in this case associated with the pest reduction via biocontrol services provided by birds (Johnson et al. 2009). Vegetational diversification has also provided additional benefits in coffee pest management; Amaral et al. (2018) found that an increase on plant diversity in an organic coffee production was associated with an increase in predation of the coffee leaf miner Leucoptera coffeella (Guerin-Meneville) (Lepidoptera: Lyonetiidae) by wasps. The latter was also demonstrated in a coffee intercropping system, with different plant species, including avocado (Persea americana Mill), where plots intercropped with avocado showed a larger overall diversity of social wasps, usually associated with predation in coffee and other agroecosystems (Tomazella et al. 2018). In a systematic review of the available biological control of the CBB, Escobar-Ramirez et al. (2019) found substantial evidence of successful cases in coffee agroforestry, where fungi, ants, parasitic wasps, birds, and nematodes can provide successful CBB biocontrol. However, landscape-scale studies are almost missing, indicating the need of more studies on CBC, also explaining a disagreement between coffee growers and researchers on the question if sustaining shaded crops reduce the incidence of main pests and diseases (Constantino et al. 2021; Jezeer and Verweij 2015). In this regard, there is no doubt that sustaining traditionally agroforestry systems would compensate production over biodiversity preservation if environmentally aware consumers are willing to accept premium prices for shade and/or socio-environmental certifications on a agroecosystem that contains as much diversity as forest habitats (Hardt et al. 2015; Perfecto et al. 1996, 2005). Row crops among all extensive crops could be considered those in which large, monospecific, and dense patches of land are dedicated to one or very few cultivars of the same crop species, affecting more severely the complexity of the landscape. However, as mentioned earlier, changes in the way crop rows are cultivated, and conceived from a landscape perspective, have opened the chances for CBC. In the case of Argentina, soybean cultivation has been expanding dramatically up to the point that by the beginning of the 2010s, soybean was using more than half of the cultivated land in the country (Aizen et al. 2009), and the latter at the expense of other crops and non-cultivated areas (Gonzalez et al. 2017). Gonzalez et al. (2017) demonstrated that fragments of natural vegetation were important in providing CBC against the stink bug Dichelops furcatus (F.) (Hemiptera: Pentatomidae) in soybean, whose populations were lowest when closer to surrounding forests. The latter authors also concluded that forest amount and landscape scale are more important than proximity to the forest, with the purpose of providing CBC against the stink bugs in soybean. Following the same idea of forest remnants and its effects on agroecosystem diversity, in Colombia, Rivera-Pedroza et al. (2019) found that species richness of ants and birds was decreasing from vegetation strips in sugarcane, with predatory functional groups with important implications for biological control services on key sugarcane pests, indicating that conserving natural vegetation strips is important in promoting CBC in this agroecosystem. In this regard, the sugarcane stem borers Diatraea spp. are considered the most economically important pest in Colombia, where biological control via releases of egg and larval parasitoids are main tools in pest management. However, one of the most important larval parasitoids is the wild tachinid Genea jaynesi (Aldrich) (Diptera: Tachinidae) (Sarmiento-Naizaque et al. 2021; Vargas et al. 2018), whose efforts to mass rearing have been futile, suggesting the need to focus on conservation of vegetation strips to increase G. jaynesi biocontrol services. New technologies on industrialized agriculture and CBC It is understood that practices that enhance both the crop and the surrounding environment by habitat management and/or cultural practices benefit CBC. On the other hand, the use of insecticides, either chemical or biological, is recognized as detrimental to non-target organisms and among them NE. However, Torres and Bueno (2018) discussing two major crop commodities in Brazil, soybean and cotton, which are highly dependent on chemical control of pests, suggest that NE and selective insecticides could be effectively combined to manage pest populations. They also discuss that more focus is needed in understanding the interaction of insecticide compounds and NE, and different avenues of interactions either direct or indirect exposure, considering in addition the chances of underlying physiological selectivity that can be enhanced by continuous exposure of some NE to pesticides. In this regard, there are some examples in predators such as green lacewings and lady beetles exhibiting resistance to several modes of action (Abbas et al. 2014; Costa et al. 2018). The above mentioned in a perspective that chemical selectivity, spraying techniques, and IPM principles of adopting economic thresholds applications can promote CBC. It could be argued that the idea of CBC does not necessarily imply organic production or agroecological approaches. A scenario of industrialized agriculture where insecticides and genetically modified plants, among other new technologies, is articulated in an IPM approach could be leaving room, and even facilitating CBC, without necessarily meaning that these systems are transitioning towards an organic system. In Brazil, major row crops, composed of Bt transformed plants such as soybean, cotton, corn, and lately sugarcane, are widely planted across extensive areas. In this regard, Luz et al. (2018) found that NE of Helicoverpa armigera (Hubner) (Lepidoptera: Noctuidae) were causing up to 41% of larval parasitism on refuge areas of cotton not subject to insecticide applications. The latter point out the structured refuge areas of these Bt crops as source of NE and promote CBC in these genetically modified agroecosystems. Row crops can also become more prone to CBC when transitioning to a more environmentally friendly organic farming, and to a school denominated biodynamic farming, more closely related to traditional farming, where the idea, beyond adding organic materials in the system (e.g., nutrients, insecticides, and fungicides), is stimulating and regulating the nutrients and energy cycles, thereby improving soil and crop quality (Robusti et al. 2020), and additionally more traditional CBC. Brazil is considered the greatest supplier of soybean in the world, and more conventional farmers are making the switch from conventional agriculture to organic/biodynamic soybean, where higher organic trading prices can cover higher costs and provide extra profitability besides environmental revenues in general (Robusti et al. 2020). Principal organisms associated with conservation in the Neotropical region Several countries in the Neotropical region are considered megadiverse zones in the world for several taxa (Morrone 2014; Rodrigues et al. 2003). This trend is observed in arthropods, where several countries in the Neotropical region are considered megadiverse for some groups such as coleopterans or lepidopterans (García-Robledo et al. 2020). The megadiversity present in the Neotropical region has worked to relevant purposes such as defining protected areas (De Carvalho et al. 2017), while some plant compounds have strongly contributed to the development of pharmaceutics and medicine (Desmarchelier 2010). However, other potential applied uses of local diversity in other field like the biological control remain poorly understood (Souza et al. 2019). As mentioned previously, CBC promotes the utilization of local beneficial fauna by different practices, increasing the diversity and abundance of groups like predators and parasitoids. Considering the megadiversity of Neotropical region, the implementation of CBC could be a practice widely used (Souza et al. 2019); however, it has received little attention when compared to other types of biological control such as the classical or the augmentation. One of the possible causes preventing the implementation of CBC programs could be the lack of knowledge regarding the biology and ecology of native NE. Therefore, it is necessary to provide updated information regarding to the use and current knowledge of entomophagous arthropods and its use in CBC in the region. Given the necessity of exploring the local diversity, the aim of this section was to evaluate the current knowledge of the use of entomophagous NE (predators and parasitoids) in the Neotropical region, focusing on Central and South America between 2010 and 2022. To do this, we used two databases, namely Scopus and Scielo. In the first selected database, we focused on the empirical papers published by Central and South American countries. Here, we filtered the search by selecting the production from the countries included in the Neotropical region and the time lapse selected. The algorithm search words are displayed in supplementary material (Appendices 1 to 4). In the case of the Scielo library, we filtered the search to the timespan selected. Once the dataset was obtained, we filtered results not related to our search parameter (other topics), those which were made in other countries than those belonging to the Neotropical region, reviews, and book chapter. Given the database compatibility and the higher number of papers when compared to the other database, data coming from Scopus was analyzed with the software Bibliometrix (Aria and Cuccurullo 2017). Data was analyzed considering the number of papers published per year, most productive countries, and the cooperative working network between authors. In a second stage, we combined the papers extracted from both database and classified them according to the functional group (predators or parasitoids) and taxonomic category in the case of predators, considering that most studies on parasitoids were focused on hymenopterans. Bibliographic production about CBC in Neotropical countries When analyzing the bibliometric production from the Scopus database, we found that the bibliographic production showed a slight but oscillating increase in the evaluated period (Fig. 2). Interestingly, the year 2021 recorded the highest productivity in the middle of the COVID-19 pandemic. When evaluating the production per country, we found that Brazil published the highest number of papers, followed by Argentina and Chile, while other countries showed a similar productivity. Interestingly, the USA showed a high productivity too, despite being excluded in our analysis. We also found a high cooperation between Brazil and European countries as well as with the USA, and between Argentina and Colombia and the USA. Some other connections are observed between some South American countries and other regions such as Africa and Oceania (Fig. 3).Fig. 2 Annual scientific production retrieved from the database Scopus between 2010 and 2022 about predators and parasitoids from the Neotropical region and their role as conservative biological control agents. Plot was generated using the R package bibliometrix (see text for details) Fig. 3 Country production and collaboration map. Darkest colors indicate a higher number of published papers, while red lines indicate scientific cooperation between different countries in the Neotropics. Plot was generated using the R package bibliometrix (see text for details) Bibliographic production in relation to different functional and taxonomic groups of NE in the Neotropical region When evaluating the scientific productivity in the different functional groups, we found that most of the studies included predators as main target followed by parasitoids, while in a lesser degree, some studies included in the category “mix” included both predators and parasitoids (Fig. 4A). When analyzing predators in detail, we found that studies focused on predatory insects were by far more common than the remaining groups, which included arachnids and vertebrates. In the category “mix,” we included studies that comprised both arthropods and vertebrates.Fig. 4 Number of published papers according to A functional groups and B taxonomic group of predators in the Neotropics. The category mix includes studies which included more than one functional or taxonomic group belonging to a different category. Plot was generated using the R package bibliometrix (see text for details. Original data can be requested to the author Luis Fernando Garcia) Parasitoids are widely used in biological control given their high specificity towards certain pest groups and host-killing behavior (Wang et at. 2019). In our analysis, we found that several studies were focused on hymenopterans belonging to different families, while some others included tachinid flies. In the case of parasitoids, studies included both diversity and applied biological control studies (see Appendix 1), attacking pests as important as fruit flies, or protecting crops such as coffee. Interestingly, new studies in poorly known group such as tachinids shed light on the importance of exploring the local diversity, especially when considering the relevance of tachinid flies as biological control agents (Grenier 1988). In the case of predators, most studies focused on some relatively specialist (i.e., aphidophagous) such as chrysopids, coccinelids, and syrphids, that have been traditionally studied (New 1975; Obrycki et al. 2009; Dunn et al. 2020). In contrast, studies on generalist predators are more scarce, although some groups such as ants and vespid wasps have been studied as native NE of relevant pests, including Spodoptera frugiperda and Hypothenemus hampei (Armbrecht and Gallego 2007; Montefusco et al. 2017). In the particular case of ants, its use in CBC programs can be controversial since some groups can offer some protection to local pests (Carabalí-Banguero et al. 2013). According to our analysis, arachnids (after arthropods as a whole) were the second more studied group. Acari were the most studied arachnid group followed by spiders. Studies about Acari and CBC in our analyses were focused on direct evidence of pest control, as well as habitat manipulation. In the case of spiders, studies were focused on diversity and abundance analyses in agroecosystems, as well as observations of feeding behavior against some local pests. Although the studies focused on spiders are comparatively low, the fact that this group is being considered in CBC in the Neotropics is important. Recent studies show that despite their generalist habits, spiders are effective biological control agents because of their dominance on crops as well as their direct and indirect effects against pest populations (Michalko et al. 2019), turning them into a promising group for future CBC programs in the Neotropical region. Vertebrates have been often neglected as CBC agents; however, recent evidence shows that several groups such as amphibians, birds, and mammals provide an important pest control service in crops (Riccucci and Lanza 2014; Khatiwada et al. 2016; Garcia et al. 2020). In our analysis, we found that in the Neotropics, most studies have been focused on pest control provided by birds and mammals. In the case of birds, their role as biological control agents has been evaluated mainly against insect pests, but also against rodents which can be both agricultural and sanitary pests. The habitat complexity has shown to have a positive effect on birds when controlling pests in the Neotropics (Olmos-Moya et al. 2022). In the case of mammals, most of the studies were focused on insectivorous bats that have an important role as CBC agents in Neotropical crops; however, recent evidence has shown that other groups such as armadillos might play an important role when controlling pests which are hard to eradicate such as Acromyrmex ants (Hymenoptera: Formicidae) (Elizalde and Superina 2019). Concluding remarks From a historical perspective and looking into the future, CBC arises as an alternative to conventional pest management practices (e.g., chemical pesticides), with the goal to decrease negative impacts on the ecosystem. However, as its implementation on pest management practices does not necessarily provide rapid and full regulation of pests, there must be an articulation to integrated and ecological pest management programs. In the case of small-scale agriculture, the use of CBC techniques in Latin America is highly associated with the practice of Agroecology, and the diversification of the landscape on these rural areas is linked to the genetic resources available in each region. In the case of the industrialized agriculture, there has been progress on its adoption, usually arguing its benefits but lacking scientific documentation. Agribusiness companies and private research institutes would surely be stimulated by pressure from the market to implement less environmentally harsh practices; however, the latter could not have the extent to stimulate farmers associations, aside from those already environmental conscious, and private research institutes to promote rigorous documentation and information on the matter, beyond internal reports or institutional magazines, indicating the need for further collaboration and partnerships between industry and the academia to promote better understanding and adoption of CBC. A general review of the CBC cases in the Neotropical region allows the conclusion of great need of appropriate documentation of cases; a great deal of information referenced in this review has come from extension manuals and internal publications, not often associated with scientific periodical publications. The latter agrees to Luliano and Gratto (2020) when suggested a bias on the CBC research towards the developed world, with few exceptions available on the tropics (Wyckhuys et al. 2013) and specifically in the Neotropical region (Peñalver-Cruz et al. 2019). Efforts directed toward the documentation of the examples and analysis of the CBC in the region will surely benefit the understanding of the dynamics of the CBC under these conditions, complementing the body of knowledge already obtained in the temperate regions. In relation to the more prevalent fauna associated to CBC in the Neotropical region, although our analysis is limited to two databases, it provides some interesting trends. For example, although the scientific production has shown an oscillating pattern in the evaluated period, we recorded an increase between the last 2 years, suggesting the local increase on documentation on entomophagous arthropods. As mentioned previously, many countries from the Neotropical region do not report research on the evaluated topic or when it is reported results from cooperation with North American or European countries. Given the local needs and further prospects, it would be important to increase cooperation between South Central American and the Caribbean. Although studies regarding NE focus on predators, only some groups, mainly specialists, have been evaluated. In complement, further research should focus in new native and promising predators such as ants, wasps, spiders, and vertebrates. The same trend therefore should be applied to local parasitoids, which are also promising but poorly known such as tachinid flies, representing a challenge in general for biological control researchers and practitioners when it comes to have on taxonomic identification, which at the same time highlights the great need of more efforts on biodiversity surveys and taxonomic studies to help in the general scope of CBC impact. Additionally, more efforts are necessary on determining the type of modifications in the agricultural landscapes to enhance beneficial fauna populations, aiming the determination of requirements on natural vegetation type (e.g., quantity and quality) and analyzing the effects of implementing multiple crops (intercrops, mixed crops, crops in strips, relay, green manures, among others), especially in industrialized crops. Overall, contributions and advances in CBC in the Neotropical region will be linked to the increase in local research and development in those countries where no research in this field has been made. In addition, given the local needs of increasing knowledge of CBC in Latin American and Caribbean countries, which are highly dependent on agriculture, stronger efforts should focus on increase local research between different countries from this region. It could be argued that research on CBC in North America and Europe has a long history and many documented advances in comparison to the Neotropical region where there is a lack of studies related to agroecosystem management seeking pest suppression. The latter has promoted the discussion within the Neotropical Regional Section of the International Organization for Biological Control (IOBC-NTRS) about the need to integrate researchers and promote advances in this area. As a new development and an important new avenue of advancing the study of CBC in the region, the working group on Conservation Biological Control in the Neotropical Region (WGCBC) was created and launched in 2022 with the purpose of enhancing the production of knowledge of the influence of native or implanted biodiversity and agricultural practices on habitat management that promote biological control in addition to other ecosystem services in agricultural areas. According to the general guidelines of the IOBC organization, working groups should allow IOBC members (and others) to focus on specific topics of interest in the discipline to provide scientific progress and a sense of community among working group members. Thus, the conformation of this new group has as a main objective to evaluate and disseminate knowledge about research and results on CBC in the Neotropical region, from which this review is considered an initial step in this direction. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (DOCX 23.6 KB) Supplementary file2 (DOCX 30.9 KB) Supplementary file3 (DOCX 12.3 KB) Supplementary file4 (DOCX 12.2 KB) Author contributions General overview and proposal: GV; Conceptualization: GV, LFR, LFG and SMJ; Writing—original draft preparation: GV, LFR, LFG and SMJ; Analysis on bibliometric production: LFG; Writing—review and editing: GV, LFR, LFG and SMJ. 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 Abbas N Mansoor MM Shad SA Pathan AK Waheed A Ejaz M Razaq M Zulfiqar Fitness cost and realized heritability of resistance to spinosad in Chrysoperla carnea (Neuroptera: Chrysopidae) Bull Entomol Res 2014 104 707 715 10.1017/S0007485314000522 25033090 Adaime R Lima AL Sousa MSM Controle biológico conservativo de moscas-das-frutas na Amazônia brasileira Innov Agron 2018 64 47 59 Aguiar-Menezes EL Silva AC Plantas atrativas para inimigos naturais e sua contribuição no controle biológico de pragas agrícolas 2011 Rio de Janeiro Embrapa Agrobiologia 60 Aizen MA Garibaldi LA Dondo M Expansión de la soja y diversidad de la agricultura argentina Ecol Austral 2009 19 45 54 Aldana JA Calvache H Escobar B Castro HB Las plantas arvenses benéficas dentro de un programa de manejo integrado de Stenoma cecropia meyrick, en palma de aceite Palmas 1997 18 11 21 Aldana JA Calvache H Daza CA Alternativas para siembra de plantas nectariferas Palmas 2004 25 194 204 Alhadidi SN Griffin JN Fowler MS Natural enemy composition rather than richness determines pest suppression Biocontrol 2018 63 575 584 10.1007/s10526-018-9870-z Altieri M Agroecologia: bases científicas para uma agricultura sustentável 1989 AS-PTA Expressão Popular 400 Altieri M, Nicholls C (2012) Agroecología: única esperanza para la soberanía alimentaria y la resiliencia socioecológica. Una contribución a las discusiones de Rio+20 sobre temas en la interface del hambre, la agricultura, y la justicia ambiental y social. SOCLA, junio 2012. p 21 Amaral DS Venzon M Duarte MVA Sousa FF Pallini A Hardwood JD Non-crop vegetation associated with chili pepper agroecosystems promote the abundance and survival of aphid predators Biol Control 2013 64 338 346 10.1016/j.biocontrol.2012.12.006 Amaral DS Venzon M Pallini A Lima PC DeSouza O A diversificação da vegetação reduz o ataque do bicho-mineiro-do-cafeeiro Leucoptera coffeella (Guérin-mèneville) (Lepidoptera: Lyonetiidae)? Neotrop Entomol 2018 39 543 548 10.1590/S1519-566X2010000400012 Aria M Cuccurullo C bibliometrix: an R-tool for comprehensive science mapping analysis J Informetr 2017 11 959 975 10.1016/J.JOI.2017.08.007 Armbrecht I Gallego MC Testing ant predation on the coffee berry borer in shaded and sun coffee plantations in Colombia Entomol Exp Appl 2007 124 261 267 10.1111/j.1570-7458.2007.00574.x Barbosa P Conservation biological control 1998 Londres, UK Academic press Barbosa P Benrey B Barbosa P The influence of plants on insect parasitoids: implications for conservation biological control Conservation Biological Control 1998 Academic Press 55 82 Barbosa P Wratten SD Barbosa P Influence of plants on invertebrate predators: implications to conservation biological control Conservation Biological Control 1998 Academic Press 83 100 Becker C Desneux N Monticelli L Fernandez X Michel T Lavoir AV Effects of abiotic factors on HIPV-mediated interactions between plants and parasitoids Biomed Res Int 2015 2015 342982 10.1155/2015/342982 26788501 Begg GS Cook SM Dye R Ferrante M Franck P Lavigne C Lövei GL Mansion-Vaquie A Pell JK Petit S A functional overview of conservation biological control Crop Prot 2017 97 145 158 10.1016/j.cropro.2016.11.008 Bernardes MFF, Pazin M, Pereira LC, Dorta DJ (2015) Impact of pesticides on environmental and human health. In Toxicology studies—cells, drugs and environment, IntechOpen: London, UK, pp. 195–233 Boedeker W Watts M Clausing P Marquez E The global distribution of acute unintentional pesticide poisoning: estimations based on a systematic review BMC Publ Health 2020 2020 1875 10.1186/s12889-020-09939-0 Brasil. Lei nº 12.651, de 25 de maio de 2012. Código Florestal Brasileiro. Brasilia, Distrito Federal, 2012. Available at: https://www.planalto.gov.br/ccivil_03/_ato2011-2014/2012/lei/l12651.htm. Accessed 23 Nov 2022 Bruinsma J (2003) World agriculture: towards 2015/2030 — an FAO perspective. Earthscan, London and FAO, Rome Bustillo-Pardey AE (2014) Manejo de insectos-plaga de la palma de aceite con énfasis en el control biológico y su relación con el cambio climático. Palmas:66–77 Cabral-Antúnez CC, Romero GR, López VAG (2020) Biological control in Paraguay. In: Van Lenteren J, Bueno VHP, Luna MG, Colmenarez YC. Biological control in Latin America and in Caribbean: its rich history and bright future. CAB International. pp 354–368 Calvache H Manejo integrado de plagas de la palma de aceite Revista Palmas 1995 16 255 264 Carabalí-Banguero DJ Wyckhuys KAG Montoya-Lerma J Do additional sugar sources affect the degree of attendance of Dysmicoccus brevipes by the fire ant Solenopsis geminata? Entomol Exp Appl 2013 148 65 73 10.1111/EEA.12076 Carneiro AR, SÁ, Bertruy RP (orgs.) (2009) Jardins Históricos Brasileiros e Mexicanos. Recife: Ed. UFPE, 2009. Carvalho FP Pesticides, environment, and food safety Food Energy Secur 2017 6 48 60 10.1002/fes3.108 Ceballos M Martínez M Duarte DLA Baños LHL Sánchez A Asociación áfidos-parasitoides em cultivos hortículas Rev Prot Veg 2009 24 180 183 Chapling-Kramer R Kremer C Pest control experiments show benefits of complexity at landscape and local scales Ecol Appl 2012 22 1936 1948 10.1890/11-1844.1 23210310 Chirinos DT Castro R Cun J Castro J Peñarrieta S Solis L Geraud F Los insecticidas y el control de plagas agrícolas: la magnitud de su uso en cultivos de algunas provincias de Ecuador Cienc Tecnol Agrop 2020 21 e1276 Colmenarez YC, Corniani N, Jahnke SM, Sampaio MV, Vásquez C (2018) Use of parasitoids as a biocontrol agent in the neotropical region: challenges and potential. In: IntechOpen (ed.), Hymenopteran wasps - the parasitoids. London, United Kingdom: IntechOpen. pp 1–23. 10.5772/intechopen.80720 Constantino LM Rendon JR Cuesta G Medina RD Benavides P Dinámica poblacional, dispersión y colonización de la broca del café Hypothenemus hampei en Colombia Cenicafe 2021 72 23 43 Costa PMG Torres JB Rondelli VM Lira R Field-evolved resistance to λ-cyhalothrin in the lady beetle Eriopis connexa Bull Entomol Res 2018 108 380 387 10.1017/S0007485317000888 28920566 Da Silva GS Jahnke SM Ferreira MLG Hymenoptera parasitoids in protected area of Atlantic Forest biomes and organic rice field: 2 compared assemblages Rev Colomb Entomol 2016 42 110 117 10.25100/socolen.v42i2.6680 Da Silva GS Jahnke SM Johnson N Riparian forest fragments in rice fields under different management: differences on hymenopteran parasitoids diversity Braz J Biol 2019 79 1 11 29590249 Daam MA Chelinho S Niemeyer JC Owojori OJ De Silva P Sousa JP van Gestel C Römbke J Environmental risk assessment of pesticides in tropical terrestrial ecosystems: Test procedures, current status and future perspectives Ecotoxicol Environ Saf 2019 181 534 547 10.1016/j.ecoenv.2019.06.038 31234068 De Carvalho DL, Sousa-Neves T, Cerqueira PV, et al. (2017) Delimiting priority areas for the conservation of endemic and threatened Neotropical birds using a niche-based gap analysis. PLoS One 1210.1371/JOURNAL.PONE.0171838 Del Puerto-Rodríguez AM, Suárez-Tamayo S, Palacio-Estrada DE (2014) Efectos de los plaguicidas sobre el ambiente y la salud. Rev Cub Hig Epidemiol 52(3) Desmarchelier C Neotropics and natural ingredients for pharmaceuticals: why isn’t South American biodiversity on the crest of the wave? Phyther Res 2010 24 791 799 10.1002/PTR.3114 Díaz MF Ramírez A Poveda K Efficiency of different egg parasitoids and increased floral diversity for the biological control of noctuid pests Biol Control 2012 60 182 191 10.1016/j.biocontrol.2011.11.001 Díaz S Pascual U Stenseke M Martín-López Assessing nature’s contributions to people Science 2018 359 270 272 10.1126/science.aap8826 29348221 Dunn L Lequerica M Reid CR Latty T Dual ecosystem services of syrphid flies (Diptera: Syrphidae): pollinators and biological control agents Pest Manag Sci 2020 76 1973 1979 10.1002/PS.5807 32115861 Egan P Dicks LV Hokkanen H Stenberg JA Delivering integrated pest and pollinator management (IPPM) Trends Plant Sci 2020 25 577 589 10.1016/j.tplants.2020.01.006 32407697 Ehler L (1998) Conservation biological control: past, present, and future. In: Conservation biological control. Academic Press. pp. 1–8 Elizalde L Superina M Complementary effects of different predators of leaf-cutting ants: Implications for biological control Biol Control 2019 128 111 117 10.1016/J.BIOCONTROL.2018.09.015 Escobar-Ramirez S Grass I Armbrecht I Tscharntke T Biological control of the coffee berry borer: main natural enemies, control success, and landscape influence Biol Control 2019 136 103992 10.1016/j.biocontrol.2019.05.011 Fedepalma (2016) Paisaje Palmero Biodiverso. Proyecto GEF/BID. Unidad Coordinadora del Proyecto. Available at: https://web.fedepalma.org/sites/default/files/files/2016-05%20Si%cc%81ntesis%20Proyecto%20GEF%20(1).pdf. Accessed 19 Nov 2022 Fernández DE Cichón LI Sánchez EE Garrido SA Cecilia GC Effect of different cover crops on the presence of arthropods in an organic apple (Malus domestica Borkh) Orchard J Sustainable Agric 2008 32 197 211 10.1080/10440040802170624 Ferreira MLG Jahnke SM Morais RM Da Silva GSD Diversidad de insectos depredadores en área orizícola orgánica y de conservación en Viamão, RS, Brasil Rev Colomb Entomol 2014 40 120 128 Frank BJR, Yamaki HA (2018) paisagem vernacular segundo perspectivas de Sauer, Hoskins E Jackson. Caminhos de Geografia Uberlândia - MG v. 19, n. 65 Março/2018 p. 245–256 Gaimari SD Quintero EM Kondo T First report of Syneura cocciphila (Conquillett, 1895) (Diptera: Phoridae), as a predator of the fluted scale Crypticerya multicicatrices Jondo and Uhruh, 2009 (Hemiptera: Monophlebidae) Boletín Del Museo De Entomología De La Universidad Del Valle 2012 13 26 28 Garcia K Olimpi EM Karp DS Gonthier DJ The good, the bad, and the risky: can birds be incorporated as biological control agents into integrated pest management programs? J Integr Pest Manag 2020 11 11 12 10.1093/JIPM/PMAA009 García-Cancino MD Gonzáles-Cabrera J Sánches- M-C Gonzáles JA Arredondo-Bernal HC Parasitoids of Drosophila suzukii (Matsumura) (Diptera: Drosophilidae) em Colima, México Sw Entomol 2015 40 855 858 García-Robledo C Kuprewicz EK Baer CS The Erwin equation of biodiversity: from little steps to quantum leaps in the discovery of tropical insect diversity Biotropica 2020 52 590 597 10.1111/BTP.12811 Gliessman SR Agroecology: ecological processes in sustainable agriculture 1998 Chelsca, Michigan Ann Arbor Press Gliessman SR (2001) Agroecologia: processos ecológicos em agricultura sustentável. 2ªedição. Porto Alegre: Ed. UFRGS, 653 p Gómez-Jiménez MI (2018) Parasitoides como controladores de Erinnyis ello. In: Cotes, A. M. (ed.) Control Biologico de fitopatógenso, insectos y ácaros. Volumen 1. Agentes de control biológico. Editorial Agrosavia, Bogotá, Colombia, pp 507–509 González E Salvo A Valladares G Arthropod communities and biological control in soybean fields: Forest cover at landscape scale is more influential than forest proximity Agric Ecosyst Environ 2017 239 359 367 10.1016/j.agee.2017.02.002 Grenier S (1988) Applied biological control with Tachinid flies (Diptera, Tachinidae): a review Anzeiger Für Schädlingskunde, Pflanzenschutz, Umweltschutz 1988 613 61 49 56 10.1007/BF01906254 Haggar J Casanoves F Cerda R Cerretelli S Gonzalez-Mollinedo LG Shade and agronomic intensification in coffee agroforestry systems: trade-off or synergy? Front Sustainable Food Sys 2021 10.3389/fsufs.2021.645958 Hardt E Borgomeo E dos Santos RF Pinto LFG Metzger P Sparovek G Does certification improve biodiversity conservation in Brazilian coffee farms? For Ecol Manage 2015 357 181 194 10.1016/j.foreco.2015.08.021 Harterreiten-Souza ES Togni PHB Pires CSS Sujii ER The role of integrating agroforestry and vegetable planting in structuring communities of herbivorous insects and their natural enemies in the Neotropical region Agroforest Syst 2014 88 205 219 10.1007/s10457-013-9666-1 HeimpeL GE, Mills NJ (2017) Biological control: ecology and applications. Cambridge. 380p Hernandez-Mahecha LM Manzano MR Guzmán YC Buhl PN Parasitoids of Prodiplosis longifilia Gagné (Diptera: Cecidomyiidae) and other Cecidomyiidae species in Colombia Acta Agron 2018 67 184 191 10.15446/acag.v67n1.62712 Hoshino AT Bortolotto OC Hata FT Ventura MU Menezes-Júnior AO Effect of pigeon pea intercropping or shading with leucaena plants on the occurrence of the coffee leaf miner and on its predation by wasps in organic coffee plantings Ciência Rural, Santa Maria 2018 48 e20160863 10.1590/0103-8478cr20160863 Huffaker CB Messenger PS Theory and practice of biological control 1976 Nueva York Academic Press Jackson JB Discovering the vernacular landscape 1984 New Haven Yale University Press Jahnke SM Da Silva GS Challenges in the applied use of parasitoids to control agricultural pests Awasthi LP 2021 Biopesticides in Organic Farming Recent Advances; CRC Press 261 266 Janssen A van Rijn P Pesticides do not significantly reduce arthropod pest densities in the presence of natural enemies Ecol Let 2021 24 2010 2024 10.1111/ele.13819 34160871 Jezeer RE, Verweij PA (2015) Café en sistema agroforestal – doble dividendo para la biodiversidad y los pequeños agricultores en Perú. Hivos, The Hague, The Netherlands Johnson MD Kellerman JL Stercho AM Pest reduction services by birds in shade and sun coffee in Jamaica Anim Conserv 2009 13 140 147 10.1111/j.1469-1795.2009.00310.x Khatiwada JR Ghimire S Paudel Khatiwada S Frogs as potential biological control agents in the rice fields of Chitwan Nepal Agric Ecosyst Environ 2016 230 307 314 10.1016/J.AGEE.2016.06.025 Khush GS Modern varieties—their real contribution to food supply and equity GeoJournal 1995 35 275 284 10.1007/BF00989135 Landis DA Wratten SD Gurr GM Habitat management to conserve natural enemies of arthropod pests in agriculture Annu Rev Entomol 2000 45 175 201 10.1146/annurev.ento.45.1.175 10761575 Lange ES Farnier K Degen T Gaudillat B Aguilar-Romero R Bahena-Juárez F Oyama K Turlings T Parasitic wasps can reduce mortality of teosinte plants infested with fall armyworm: support for a defensive function of herbivore-induced plant volatiles Front Ecol Evol 2018 6 55 10.3389/fevo.2018.00055 Larsen A Philpott SM Twig-nesting ants: the hidden predators of the coffee berry borer in Chiapas, Mexico Biotropica 2010 42 342 347 10.1111/j.1744-7429.2009.00603.x Letourneau DK Jedlicka JA Bothwell SG Moreno CR Effects of natural enemy biodiversity on the suppression of arthropod herbivores in terrestrial ecosystems Annu Rev Ecol Evol Syst 2009 40 573 592 10.1146/annurev.ecolsys.110308.120320 Letourneau DK Armbrecht I Rivera BS Does plant diversity benefit agroecosystems? A synthetic review Ecol Appl 2011 21 9 21 10.1890/09-2026.1 21516884 Luliano B, Gratton C (2020) Temporal resource (dis)continuity for conservation biological control: from field to landscape scales. Front Sustainable Food Syst 4: Article127 Lutzenberger JA (1978) Fim do Futuro? Manifesto Ecológico Brasileiro Ed. Movimento Luz PMC Paula-Moraes SV López JMP Pujol-Luz JR Penteado-Dias AM Specht A Diniz IV Parasitoid associated with of Helicoverpa armigera in refuge areas of cotton, in Western Bahia Brazil Cienc Rural 2018 48 e20170250 Mahmood I Imadi SR Shazadi K Gul A Hakeem KR Hakeem K Akhtar M Abdullah S Effects of pesticides on environment Plant, soil and microbes 2016 Cham Springer 253 269 Martínez-Centeno AL Huerta K La revolución verde Rev Iberoam Bioecon Cambio Clim 2018 4 8 10.5377/ribcc.v4i8.6717 Matienzo Y Rijo E Milán O Torres N Larrinaga J Massó E Contribución al conocimiento de espécies botânicas com potencialidade para el fomanto de reservorios de insectos benéficos em la agricultura urbana Fitosanidad 2007 11 128 129 Matthews G (2015) The future of pesticides. In: Matthews G (2015). Pesticides: health, safety and the environment. Second Edition. John Wiley & Sons. pp.225–252 Medeiros HR Hoshino AT Ribeiro MC Morale MN Martello FO Non-crop habitats modulate alpha and beta diversity of flower flies (Diptera, Syrphidae) in Brazilian agricultural landscapes Biodiversity Conserv 2018 27 1309 1326 10.1007/s10531-017-1495-5 Melo PCT, Melo AMT (2015) Olericultura brasileira: do descobrimento ao século 21. revista da APh. 119:22–27 Mexson RG, Chinchilla CM (2003) Especies vegetales atrayentes de la entornofauna benéfica en plantaciones de palma de aceite (Elaeis guineensis Jacq.) en Costa Rica. Palmas:33–57 Mexzón R (2001) Reconocimiento de las especies vegetales em las que se refugiam o laimentam las avispas parasitoides de los taladradores (Lepidopetar: Tortricidae) y de los chinches Fitófagos (hemiptera: Pentatomidae) de la nuez de la macadamia em Costa Rica. Proyecto de investigación Número 813-AO-146 Centro de Investigaciones em Protección de cultivos (CIPROC), Universidad de Costa Rica, San José, Costa Rica Michalko R Pekár S Dul’a M, Entling MH, Global patterns in the biocontrol efficacy of spiders: a meta-analysis Glob Ecol Biogeogr 2019 28 1366 1378 10.1111/geb.12927 Michaud JP Challenges to conservation biological control on the High Plains: 150 years of evolutionary rescue Biol Control 2018 125 65 73 10.1016/j.biocontrol.2018.07.001 Montefusco M Gomes FB Somavilla A Krug C Polistes canadensis (Linnaeus, 1758) (Vespidae: Polistinae) in the Western Amazon: a potential biological control agent Sociobiology. 2017 64 477 483 10.13102/SOCIOBIOLOGY.V64I4.1936 Morris JR Vandermeer J Perfecto I A keystone ant species provides robust biological control of the coffee berry borer under varying pest densities PLoS ONE 2015 10 1 15 10.1371/journal.pone.0142850 Morrone JJ Biogeographical regionalisation of the Neotropical region Zootaxa 2014 3782 1 110 10.11646/ZOOTAXA.3782.1.1 24871951 Moses M (1993) Pesticides. In: Paul, M. (Ed). Occupational and environmental reproductive hazards: a guide for clinicians. Baltimore: Williams & Wilkins New TR The biology of Chrysopidae and Hemerobiidae (Neuroptera), with reference to their usage as biocontrol agents: a review Trans R Entomol Soc London 1975 127 115 140 10.1111/J.1365-2311.1975.TB00561.X Newson J Vandermeer J Perfecto I Differential effects of ants as biological control of the coffee berry borer in Puerto Rico Biol Control 2021 160 104666 10.1016/j.biocontrol.2021.104666 Nicholls C Control biológico de insectos: un enfoque agroecológico 2008 Medellín, Colombia Editorial Universidad de Antioquia 282 Obrycki JJ Harwood JD Kring TJ O’Neil RJ Aphidophagy by Coccinellidae: application of biological control in agroecosystems Biol Control 2009 51 244 254 10.1016/J.BIOCONTROL.2009.05.009 Odum E (1988) Ecologia. Guanabara Koogan, 434p Olmos-Moya N Díaz-Siefer P Pozo RA The use of cavity-nesting wild birds as agents of biological control in vineyards of Central Chile Agric Ecosyst Environ 2022 334 107975 10.1016/J.AGEE.2022.107975 Ovruski S Aluja M Sivisnki J Wharton R Hymenopteran parasitoids on fruit infesting Tephitidae (Diptera) in Latin America and the southern Unided States: diversity, distribution, taxonomic status and their use in fruit fly biological control Integr Pest Manage Rev 2000 5 81 107 10.1023/A:1009652431251 Owen MDK Beckie HJ Leeson JY Norsworthy JK Steckel LE Integrated pest management and weed management in the United States and Canada Pest Manage Sci 2015 71 357 376 10.1002/ps.3928 Paredes D Campos M Cayuela L El control biológico de plagas de artrópodos por conservación: técnicas y estado del arte Ecosistemas 2013 22 56 61 Pavis C Huc JA Delvare G Boissot N Diversity of the parasitoids of Bemisia tabaci B-biotipe (Hemiptera: Aleyrodidae) in Guadeloupe Island (West indies) Environ Entomol 2003 32 608 613 10.1603/0046-225X-32.3.608 Peñalver-Cruz A Alvarez-Baca JK Alfaro-Tapia A Gontijo L Lavandero B Manipulation of agricultural habitats to improve conservation biological control in South America Neotrop Entomol 2019 48 875 898 10.1007/s13744-019-00725-1 31713220 Peñalver-Cruz A Alvarez D Lavandero B Do hedgerows infuence the natural biological control of woolly apple aphids in orchards? J Pest Sci 2020 93 219 234 10.1007/s10340-019-01153-1 Perfecto I, Rice RA, Greenberg R, Van der Voort ME (1996) Shade coffee: a disappearing refuge for diversity. Bioscience 46:598–608 Perfecto I Vandermeer J Mas A Soto-Pinto L Biodiversity, yield, and shade coffee certification Ecol Econ 2005 54 435 446 10.1016/j.ecolecon.2004.10.009 Petry C (2014) Paisagens e Paisagismo: Do apreciar ao fazer e usufruir January 2014 Edition: 1Publisher: Editora UPFISBN: 978–85–7515–867–8 Pfiffner L, Wyss E (2004) Use of wildflower strips to enhance natural enemies of agricultural pests. In: Gurr GM, Wratten SD, Altieri M (eds.). Ecological Engineering for Pest Management: Advances in Habitat Manipulation for Arthropods. CSIRO Publishing. 256p Philpott SM Arendt WJ Armbrecht I Biodiversity loss in Latin American coffee landscapes: review of the evidence on ants, birds, and trees Conserv Biol 2008 22 1093 1105 10.1111/j.1523-1739.2008.01029.x 18759777 Ponzio C Gols R Pieterse CMJ Dicke M Ecological and phytohormonal aspects of plant volatile emission in response to single and dual infestations with herbivores and phytopathogens Functional Ecol 2013 27 587 598 10.1111/1365-2435.12035 Primavesi A (1984a) Manejo ecologico del suelo, (1984a) (Parte 1/2). Agricultura sostenible. Agricultura y medio ambiente". Scribd. Retrieved 11 April 2020 Primavesi, A. (1984b). Manejo ecológico del suelo: la agricultura en regiones tropicales. 5ta edición. Editorial Ateneo. pp 475–484 Proinpa (2013) Informe Anual 2012–2013 del Proyecto: Desarrollo y validación participative de las innovaciones tecnológicas que mejoren las estratégias para manejo sostenible del sistema centrado em quinua em el Altiplano boliviano. Fundación McKnight, La Paz, Bolivia Ribeiro A Castiglioni E Characterization of the populations of natural enemies of Piezodorus guildinii (Westwood) (Hemiptera: Pentatomidae) Agrociência (uruguay) 2008 12 48 56 Ribeiro AL Gontijo LM Alyssum flowers promote biological control of collard pests Biocontrol 2016 10.1007/s10526-016-9783-7 Ribeiro A Silva H Castiglioni E Bartaburu S Martínez J Control natural de Crocidosema (Epinotoa) aporema (Walsingham) (Lepidopetra: Tortricidae) por parasitoides y hongos entomopatógenos em Lotus corniculatus y Glycine max Agrociencia (uruguay) 2015 19 36 41 Ribeiro A (2010) Prospección de agentes para el control natural de plagas en sistemas agrícolas pastoriles. In: Altier, N.; Rebuffo M, Cabrera K (eds) Altier N. Enfermedades y Plagas en Pasturas, INIA Serie Técnica, 183: 05–110 Riccucci M Lanza B Bats and insect pest control: a review Vespertilio 2014 17 161 169 Rivera-Pedroza LF Escobar F Philpott SM Armbrecht I The role of natural vegetation strips in sugarcane monocultures: ant and bird functional diversity responses Agric Ecosyst Environ 2019 284 106603 10.1016/j.agee.2019.106603 Robusti E Mazeto VA Ventura MU Soares D Menezes A Soybean crop profitability: biodynamic vs conventional farming in a 7-yr case study in Brazil Renew Agric Food Syst 2020 35 336 341 10.1017/S1742170518000613 Rodrigues ASL, Andelman SJ, Bakarr MI et al (2003) Advances in Applied Biodiversity Science: Global Gap Analysis: towards a representative network of protected areas. Adv Appl Biodivers Sci Glob Gap Anal: Towar a Represent Netw Prot areas 6–98. 10.1896/978-1-934151-14-3.6 Rojas-Rodriguez J, Rossetti MR, Videla M (2019) Importancia de las flores en bordes de vegetación espontánea para la comunidad de insectos en huertas agroecológicas de Córdoba, Argentina Rev. FCA UNCUYO. 51:249–259. Available at: https://revistas.uncu.edu.ar/ojs3/index.php/RFCA/article/view/2449/1767 [Accessed: 18 May 2022] Romero J (2012) La revolución verde. AGROECOLOGÍA UTN. Ibarra. Available at: http:// agroecologiautn.blogspot.com/ [Accessed: 18 May 2022] Root RB Organization of a plant-artthropod association in simple and diverse habitats: the fauna of collards (Brassica oleracea) Ecol Monogr 1973 43 95 124 10.2307/1942161 Rosival L Pesticides Scand J Work Environ Health 1985 11 3 189 197 10.5271/sjweh.2235 4035321 Ruberson J, Nemoto H, Hirose Y (1998) Pesticides and conservation of natural enemies in pest management. Conservation Biological Control. San Diego, Calif. (USA): Academic Press, p. 207–220 Rull, V (2011) Neotropical biodiversity: timing and potential drivers. Trends Ecol Evol. 1–6 Rusch A Birkhofer K Bommarco R Smith H Ekbom B Management intensity at field and landscape levels affects the structure of generalist predator communities Oecologia 2014 175 971 983 10.1007/s00442-014-2949-z 24810324 Sarmiento-Naizaque ZX Sarmiento CE Barreto-Triana N Parasitoides, Braconidae (Hymenoptera) y Tachinidae (Diptera) de barrenadores, Crambidae y Coleophoridae (Lepidoptera) de caña de azúcar para la producción de panela en Colombia Rev Colomb Entomol 2021 47 e10558 10.25100/socolen.v47i2.10558 Sendoya-Corrales CA Bustillo AE Enemigos naturales de Stenoma cecropia (Lepidoptera: Elachistidae) en palma de aceite, en el suroccidente de Colombia Rev Colomb Entomol 2016 42 146 154 10.25100/socolen.v42i2.6685 Silva H Quantitative description of a trophic network of thre levels: legumes-aphids-parasitoids and enthomopathogens 2016 Faculdad de Agronomia, Montevideo, Uruguay Msc theisis Universidad de la República Smith L, Bellotti AC (1996) Successful biocontrol projects with emphasis on the Neotropics. Available at: http://web.entomology. cornell.edu/shelton/cornell-biocontrolconf/talks/bellotti.html [Accessed: May 30, 2022] Souto AL Sylvestre M Tölke ED Tavares JF Barbosa-Filho JM Cebrián-Torrejón G Plant-derived pesticides as an alternative to pest management and sustainable agricultural production: prospects, applications and challenges Molecules (basel, Switzerland) 2021 26 4835 10.3390/molecules26164835 34443421 Souza IL, Tomazella VB, Santos AJN, Moraes T, Silveira LCP (2018) Parasitoids diversity in organic sweet pepper (Capsicum annuum) associated with basil (Ocimum basilicum) and marigold (Tagetes erecta). Braz J Biol 79 (4) 10.1590/1519-6984.185417 Souza B, Vázquez, LL, Marucci RC (2019) Natural enemies of insect pests in neotropical agroecosystems: biological control and functional biodiversity. Natural Enemies of Insect Pests in Neotropical Agroecosystems 1–53310.1007/978-3-030-24733-1 Thrupp LA Linking agricultural biodiversity and food security: the valuable role of agrobiodiversity for sustainable agriculture Internat Affairs 2000 76 283 297 10.1111/1468-2346.00133 Tilman D Cassman KG Matson PA Naylor R Polasky S Agricultural sustainability and intensive production practices Nature 2002 418 671 677 10.1038/nature01014 12167873 Togni PHB Venzon M Muniz CA Martins EF Pallini A Sujii ER Mechanisms underlying the innate attraction of an aphidophagous coccinellid to coriander plants: implications for conservation biological control Biol Control 2016 92 77 84 10.1016/j.biocontrol.2015.10.002 Togni PHB Venzon M Souza LM Dynamics of predatory and herbivorous insects at the farm scale: the role of cropped and noncropped habitats Agric for Entomol 2019 10.1111/afe.12337 Tomazella VB Jacques GC Lira AC Silveira LCP Visitation of social wasps in Arabica coffee crop (Coffea arabica L.) intercropped with different tree species Sociobiology 2018 65 299 304 10.13102/sociobiology.v65i2.1397 Torres JB Bueno A Conservation biological control using selective insecticides – a valuable tool for IPM Biol Control 2018 126 53 64 10.1016/j.biocontrol.2018.07.012 Trujillo AJ Control biológico por conservación: enfoque relegado. Perspectiva de su desarrollo en Latinoamérica Ceiba 1992 33 17 26 Ulhoa LA Barrigossi JAF Borges M Laumann RA Blassioli-Moraes MC Differential Induction of Volatiles in Rice Plants by Two Stink Bug Species Influence Behaviour of Conspecifics and Their Natural Enemy Telenomus Podisi Entomol Exp Appl 2020 68 76 90 10.1111/eea.12869 UNODC - Oficina de las Naciones Unidas contra la Droga y el Delito. 2010. Problemática ambiental y la utilización de agroquímicos en la producción de coca. Informe analítico. Viena, Austria. Available at: https://www.unodc.org/documents/peruandecuador/Informes/Informes-Analiticos/Informe_Analitico_Agroquimicos.pdf. (Accessed: 29 May 2022) Vargas G Lastra LA Ramírez GD Solís AM The Diatraea complex (Lepidoptera: Crambidae) in Colombia’s Cauca River valley: making a case for the geographically localized approach Neotrop Entomol 2018 47 395 402 10.1007/s13744-017-0555-6 28905324 Vargas G (2018) Los barrenadores del tallo Diatraea y su control biológico mediante parasitoides de huevos y larvas. In: Cotes AM (ed.) Control Biologico de fitopatógenso, insectos y ácaros. Volumen 1. Agentes de control biológico. Editorial Agrosavia, Bogotá, Colombia, pp 513–518 Vélez-Hoyos M Bustillo-Pardey AE Posada FJ Depredación de Hypothenemus hampei por hormigas, durante el secado solar del café Cenicafé 2006 57 198 207 Vijay V Pimm SL Jenkins CN Smith S The impacts of oil palm on recent deforestation and biodiversity loss PlosOne 2016 10.1371/journal.pone.0159668 Wang Z Liu Y Shi M Parasitoid wasps as effective biological control agents J Integr Agric 2019 18 705 715 10.1016/S2095-3119(18)62078-7 Wyckhuys KAG O’neil RJ Agro-ecological knowledge and its relationship to farmers’ pest management decision making in rural Honduras Agric Human Values 2007 24 307 321 10.1007/s10460-007-9068-y Wyckhuys KAG Lu Y Morales H Vazquez LL Legaspi JC Eliopoulos PA Hernandez LM Current status and potential of conservation biological control for agriculture in the developing world Biol Control 2013 65 152 167 10.1016/J.BIOCONTROL.2012.11.010 Zumoffen L Signorini M Salvo A Bidirectional movement of aphid parasitoids (Braconidae: Aphidiinae) between crops and non-crop plants in agroecosystems of central Argentina Appl Entomol Zool 2017 53 1 9 10.1007/s13355-017-0520-1	36449176	PMC9709742	NO-CC CODE	2022-12-01 23:23:39	no	Neotrop Entomol. 2022 Nov 30;:1-18	utf-8	Neotrop Entomol	2,022	10.1007/s13744-022-01005-1	oa_other
==== Front J Knowl Econ Journal of the Knowledge Economy 1868-7865 1868-7873 Springer US New York 1043 10.1007/s13132-022-01043-5 Article Information and Knowledge Management, Intellectual Capital, and Sustainable Growth in Networked Small and Medium Enterprises http://orcid.org/0000-0003-4466-7133 Jordão Ricardo Vinícius Dias [email protected] 123 http://orcid.org/0000-0002-3513-3220 Novas Jorge Casas [email protected] 4 1 Graduate Program in Business Administration at FPL Educational, Pedro Leopoldo, Brazil 2 The Center for Advanced Studies in Management and Economics, Évora, Portugal 3 Swiss Management Center, Zug, Switzerland 4 grid.8389.a 0000 0000 9310 6111 Management Department and CEFAGE-UE, University of Évora, Évora, Portugal 30 11 2022 133 7 7 2021 14 9 2022 © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 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. This paper aims to analyze the strategic effects of the association of small and medium-sized enterprises (SMEs) in knowledge networks (k-networks) on their information and knowledge management (IKM) and intellectual capital (IC). Taking as the object of study two innovative and successful Brazilian experiences, a descriptive survey with top-level executives (managing partners, presidents, and executive directors — henceforth CEOs) was held. Based on theories of management and economics, the findings revealed (i) that the network formation process (encompassing culture, context, incentives for information and knowledge sharing and especially strategy) is an important factor to explain IKM (for creating, systematizing and sharing data, information and knowledge) and IC in its three dimensions (human, relational and structural capital), promoting long-term sustainable growth (perceived by improvements in innovation, competitiveness and corporate results) for SMEs and their networks — a very relevant issue, but whose theoretical and managerial understanding is very incipient in international literature, especially in emerging economies. Keywords Information and knowledge management Intellectual capital Networks Small and medium-sized enterprises (SMEs) Sustainable growth Innovation Strategy and competitiveness JEL Classification Codes: O32 O34 L14 L26 M00 http://dx.doi.org/10.13039/501100001871 fundação para a ciência e a tecnologia UIDB/04007/2020 ==== Body pmcIntroduction In the knowledge-based economy, successful implementation of sustainable strategies in small and medium-sized enterprises (SMEs) involves intelligent information and knowledge management (IKM) practices and intellectual capital (IC) — which are critical to overcoming resource shortage and technological and managerial difficulties in SMEs (cf. Balestrin et al., 2008; Muhammad et al., 2011; Romiti & Sarti, 2011; Lin & Chen, 2016; Verbano & Crema, 2016; Jordão & Novas, 2017; Daňa et al., 2020; Zakery & Saremi, 2021). Recent studies such as Brassell and Boschmans (2019), under the seal of the Organisation for Economic Co-operation and Development (OECD), highlight the challenges faced by SMEs alongside the relevance of IC as the factor responsible for their survival, growth and development. Through cooperative relationships and networks, SMEs strategically manage to surpass their difficulties and combine competencies synergistically (i.e. knowledge, skills and attitudes) (Muhammad et al., 2011), creating, exploring or implementing new business opportunities (Romiti & Sarti, 2011), considered fundamental issues for their innovation and to create and sustain their corporate growth (Laperche, 2021). When SMEs use inter-organizational networks to carry out joint actions enhanced by significant use of information and knowledge, they can become knowledge networks (k-networks), in line with Moslehi et al. (2014). According to Munier (2021), this type of network is viewed as a locus where knowledge is continuously built, managed, combined, tested and selected between all participants. In such networks, knowledge is the core aspect, both in terms of evolution, competition, nature and also strategic elements. According to Jordão and Novas (2017), k-networks have been identified in the international literature as central to the generation of innovation, knowledge and IC for SMEs, constituting a strategic alternative to prevail in the market, helping them not only to survive but to grow and develop over time. However, despite the great importance attributed in the academic and business spheres to IKM and IC in SMEs and networks (e.g. Balestrin et al., 2008; Daňa et al., 2020; Jordão et al., 2020), the challenge of understanding more thoroughly these relationships between IKM, IC and SMEs k-networks and their contributions to sustainable growth (perceived by improvements in innovation, competitiveness, and corporate results) remains to be analysed in-depth, as highlighted by Verbano and Crema (2016), Jordão and Novas (2017), Martínez-Costa et al. (2019) and Agostini et al. (2020), mainly in emerging economies like Brazil. So, the research problem can be summarized in the following question: How do the intervening factors of k-networks can foster IKM and IC to enable SMEs to grow and develop? Recognizing this research gap, the objective of this paper is to analyze the strategic effects of the association of SMEs in k-networks on their IKM and IC. Taking as the object of study two innovative and successful Brazilian experiences, a descriptive survey with top-level executives (CEOs) was carried out in an attempt to analyze such issues, also helping to understand the resulting effects on the sustainable growth of such companies. The research justification comes from its social contributions, following Jordão et al. (2014). In this sense, besides the theoretical importance of the theme listed by Durst and Edvardsson (2012), Khalique et al. (2015), Jordão and Novas (2017) and Jordão et al. (2020), several international organizations such as the United Nations (UN) and the OECD draw attention to the key role of SMEs in economic and social development, particularly on job creation and income generation, also emphasizing the need to create forms and strategies for sustainable growth for these companies (OECD, 2010; UN, 2017). From a theoretical perspective, this paper contributes to a greater understanding of the topic, examining the relationship between the four analysed variables: k-networks, IKM, IC and SMEs — an issue whose understanding is very incipient in the international literature and almost non-existent in emerging economies, offering several paths for future research and business, economic and legal improvement (Cf. Balestrin et al., 2008; Jordão & Novas, 2017; Mertins et al., 2010). The originality of the study and the methodological proposal developed and applied must also be emphasized. As economical and managerial contributions, this study highlights how SMEs can and should use k-networks as a deliberate innovative strategy to overcome organizational and managerial difficulties through IKM practices and IC, generating sustainable growth for them (Cf. Munier, 2021). This study also shows how SMEs can use k-networks in practice to improve their IKM and IC, providing a benchmarking opportunity for companies and networks in similar situations. This paper is structured in 5 sections, including this introduction. “Information and Knowledge Management, Intellectual Capital and Sustainable Growth in SMEs K-networks” presents the theoretical support platform. “The Research Protocol and Methodological Process” describes the research methodology. “Results” presents and analyses the results of the research. “Deepening of the Analysis and Discussion of the Results” discusses the results. “Conclusions” contains the final remarks in line with the initial research goal. Information and Knowledge Management, Intellectual Capital and Sustainable Growth in SME K-networks Inter-organizational networks are agents of strategic relevance in current competitive dynamics, principally when formed among SMEs (Belso-Martinez & Diez-Vial, 2018; Jardon & Martos, 2012; Jordão & Novas, 2017; Khalique et al., 2015; Mariotti, 2011; Pöyhönen & Smedlund, 2004; Romiti & Sarti, 2011). Such firms are characterized by organizational flexibility and a shortage of resources — which makes it difficult for SMEs to compete with larger companies (Marques Júnior et al., 2020). On the other hand, when SMEs are organized in k-networks, they are better able to implement continuous improvement systems and put their innovation processes into practice (Chaston & Mangles, 2000). According to Hilmersson and Hilmersson (2021), the information, knowledge and temporal aspects of innovation processes in SMEs are influenced by the company’s network behaviour. In the perspective of inter-organizational networks and K-networks, IKM corresponds to the basic supporting level of competitive advantage, relating fundamentally to the processes, structures, systems and technologies aimed at the creation, use, systematization and sharing of data, information and knowledge. Prior research (e.g. Jordão, 2015; Jordão et al., 2020; Khalique et al., 2015; Lin & Chen, 2016) has stressed the role of the business strategy, trust, cooperation, context and organizational culture in the formation of successful networks, concluding they are critical for IKM practices and IC creation, being at once a source of growth, resulting in innovation, competitiveness and performance for SMEs. Jordão and Novas (2017) proposed a conceptual model to support future research regarding the contribution of five main intervening factors of inter-organizational relationships (network formation processes, organizational context, strategy, organizational culture and stimuli for knowledge sharing) on IKM (involving the creation, systematization, and sharing of data, information and knowledge) and IC (encompassing human, relational and structural capitals) in SMEs’ k-networks, providing organizational, innovative and competitive benefits for these companies. Such relationships are summarized in Fig. 1. Detailed information about the items comprising each intervening factor is presented in the Appendix (see Tables 7 and 8).Fig. 1 Analytical research model. Source: based on Jordão and Novas (2017: 683) The first intervening factor relates to the k-network formation process. The international literature is truly eloquent in terms of how the formation of networks supports the continued operation of SMEs and the achievement of competitive advantages being not only a survival strategy but also a way of growth for the networked companies, allowing the emergence of opportunities that would not be accessible acting in isolation (Cf. Jardon & Martos, 2012; Laperche, 2021; Romiti & Sarti, 2011). As Jordão and Novas (2017) pointed out, the formation of SMEs k-networks intensifies the sharing of resources, business opportunities, information and knowledge, also facilitating the transfer of explicit knowledge by creating a shared language, information systems and a common culture, with methods understood and used by network participants. Besides, the transfer of tacit knowledge takes place in an unstructured way, through informal conversations, values or traditions, occurring more easily through the knowledge networks formed. The second intervening factor in the model is the organizational context. The context refers to the specific environment in which the network is created and developed and the relationships within it. According to Caldas and Cândido (2013) and Zach and Hill (2017), the networked environment can stimulate the knowledge conversion dynamics because of its characteristics, such as cooperation, trust and social interaction, which collaborate to create and expand ideas, information and knowledge, in addition to enhancing the generation of innovation. Jordão (2015) has emphasized that to promote synergy and complementarities of knowledge into the network, it is necessary to create an environment in which trust and cooperation are at least superior to the opportunism and competition usually present in traditional economic relations. Due to its effect on IKM and IC in SMEs networks, strategy is also defined as the third intervening factor. According to Castells (1996), one of the most important elements for a successful management strategy may be to position the company within a network, which combines resources, capabilities and external factors in an attempt to generate competitive advantages and achieve objectives. In the view of Laperche (2021), a company’s formation and protection of knowledge strategy can provide crucial elements to understand the genesis of innovation. The development of collaborative strategies with other firms and institutions has considered central to maintaining and increasing the company’s innovation strength. According to Massaro et al. (2014), many SMEs still do not know how to establish analyses and strategies related to the exploitation of informational and cognitive resources, as well as with the use of different elements of the IC to improve organizational effectiveness, so networks can work as an instrument to fill these gaps. The fourth intervening factor is organizational culture. Although organizational culture is known to produce effects on IKM and on the elements that make up the IC (Cf. Martins & Solé, 2013), many companies find it difficult to implement and develop a culture of cooperation that leads to the acquisition of new data, information and knowledge and/or the enhancement of existing data, information and knowledge. According to Mason et al. (2008), a more sustainable approach to IKM consists of establishing an organizational culture based on networked learning that allows the development of common interests and the sharing of information and knowledge, especially if there is an orientation towards trust, cooperation and knowledge sharing by the people and organizations that make up the network. Stimuli also correspond to the fifth intervening factor as stimuli of knowledge sharing positively affect IKM and IC in networked SMEs. An essential aspect, at this level, is to understand whether stimuli are needed for the share of information and knowledge to potentiate the process of improving performance, establishing innovation and generating value (Bolade, 2021; Pöyhönen & Smedlund, 2004; Verbano & Crema, 2016), especially in SME networks consolidated as k-networks. On the other hand, the organizational network is formed as a result of interactions between actors, starting with the sharing of information, which is an important element that encourages the construction of knowledge (Jordão & Novas, 2017). Accordingly, the organization of the network structure must be thought of to promote an intense interrelation between individuals and between these individuals and their context. In SME networks, Jardon and Martos (2012) postulate the need to look at IC in a dynamic perspective, meaning to understand that the share of IC of people — human capital — creates structural capital (IC within the organization), and this structural capital creates relational capital (IC with the environment). These authors argue that it is this circular relationship that is responsible for the basic process of knowledge generation and dissemination. Therefore, it can be considered that, in the context of SMEs networks, there is a deep and intense relationship between IKM with the IC, where one of these aspects positively affects the other and vice-versa — these are the factors that underwent the intervention. Belso-Martinez and Diez-Vial (2018) examined how the evolution of k-networks and the strategic choices of companies affect innovation. The results indicated that the degree of involvement with these networks tends to increase the innovative capacity of companies over time. Moreover, as Balle et al. (2019) pointed out, contexts of cooperation are fertile ground for knowledge sharing. Likewise, SMEs need to adopt specific strategies for an effective exploration of IKM and IC in an attempt to generate competitive advantages and achieve business objectives, seeking to institutionalize people’s know-how in culture, routines and processes, stimulating individual and organizational learning processes. Finally, Xu et al. (2020) found that IC has a significant positive effect on the sustainable growth and performance of Chinese agricultural companies, suggesting that companies should invest in the development of human capital and effective IKM tools as a means to accumulate the necessary IC to grow and allow adaptation in environments of high changes. All in all, the set of previous empirical studies revealed that the process of transforming inter-organizational networks into k-networks, the organizational context, the strategy, the culture and the incentives for sharing information and knowledge are the central aspects to increase IKM practices and expand the IC, resulting in performance, innovation, sustainability and competitiveness — aspects that can help to analyze organizational growth in networked SMEs. The Research Protocol and Methodological Process Taking as the object of study two innovative and successful Brazilian experiences, an explanatory and descriptive survey was carried out (Cooper & Schindler, 2006), aiming to collect data and information based on the perceptions of top-level executives (CEOs), following Richardson (2013). This investigation strategy was useful to test a theoretical model (cf. Figure 1) describing, simultaneously, the characteristics and relationships of this phenomenon in its context. In an attempt to obtain a high level of accuracy, before the fieldwork, the state-of-the-art on the subject was mapped, involving the last 60 years — 1962 to 2021 (based on information from Web of Science and Scopus), covering four key elements (or variables): IKM, IC, k-networks (or inter-organizational networks) and SMEs. The analysis of these studies and the inferences derived formed the basis of theoretical support for the research, taking the model presented in Fig. 1 as the basis for analyses. Among the various sources of evidence in a study of a quantitative and qualitative nature (cf. Jick, 1979; Valkokari & Helander, 2007), questionnaires were applied. The questionnaires contained closed (using a 7-point Likert-type scale) and open questions — with variables extracted from the international literature (see Appendix, Tables 7 and 8). The closed questions aimed for greater comparability between firms, while open questions sought to broaden the understanding of the problem and with freedom for further explanations, observing the need for qualitative methodologies to better explore the peculiarities of IC in the target population, as prescribed by Henry (2013). The data-collecting instrument was pre-tested to find out ambiguous or unclear questions. The independent (corresponded to the intervening factors) and dependent (corresponded to the factors that underwent the intervention) variables were presented in the previous section and are summarized in Fig. 1. The criteria used to select the networks (cases) were (i) choosing networks that were already consolidated; (ii) indications that participant firms used formal or informal mechanisms for sharing data, information and knowledge, generating IC and (iii) access to high-level information — selection by typicality, according to Cooper and Schindler (2006). The analysis focused on two SMEs’ k-networks — which, according to Munier (2021), are the most efficient type of network for the knowledge creation process. One of these networks received technical support and management training from the Brazilian Service of Support for Micro and Small Enterprises (SEBRAE) in the State of Espírito Santo (ES), directed towards an international business network (hereafter IBN), in line with Zakery and Saremi (2021). The other network — a pharmacy network (hereafter PN) — is formed of SMEs grouped to face up to the enormous competitive pressure, in line with Brassell and Boschmans (2019). The characterization of each of the networks is presented in Table 1.Table 1 Characterization of IBN and PN networks Description IBN network PN network Birth 2008 1999 Number of companies 45 companies but only 28 can integrate the sample — of which 25 (89.28%) have participated in the study 10 companies — all have participated in the study Dimension (based on sales volume) 1 small-sized company (about US$2 M) 24 medium-sized companies (> 3,7 M USD and < 21.2 M USD 1 big-sized company (> 21.2 M USD) 4 small-sized companies (< US$1.5 M) 6 medium-sized companies (> 3,7 M USD and < 21.2 M USD Dimension (based on number of employees) 1 small-sized company (about 72 employees) 24 medium-sized companies (> 100 and < 499 employees) 1 big-sized company (> 500 employees) 1 micro-sized company (< 10 employees) 9 medium-sized companies (> 50 and < 99 employees) Degree of formalization Formal (contractual) Informal (trust relationships) Governance structure Coordination of SEBRAE with the network Coordination by network members Typology Inter-organisational network of SMEs transformed into knowledge network Inter-organisational network of SMEs transformed into knowledge network Characteristics Formed to meet the needs of export and internationalization, medium power asymmetry, high reciprocity, increasing efficiency, performance and dynamism, high stability and medium (increasing) legitimacy Formed to meet the needs of buying medicines and coping with the high pressures of the pharmacy and drugstore market, low power asymmetry, high reciprocity, high efficiency and good performance and dynamism, medium stability and high legitimacy (albeit informal) Partners type SEBRAE, universities, suppliers, customers, unions and governments Consulting companies, suppliers, customers and city hall Source: own elaboration, based on the research results Of the 45 companies belonging to the first network, the initial survey indicated that only 28 could participate in the research (integrating the sample), but at the time of data collection, three CEOs of these companies were not present, thus making a total of 25 companies (making up 89.28%). In the second network, all 10 firms (100%), accounting for 27 branches, participated in the research. Thus, respondents from 1 to 25 refer to IBN, and respondents from 26 to 35 refer to PN. Names will be omitted according to the confidentiality agreement. The units of observation consist of the main executive (CEO) of each firm, following Richardson (2013)’s orientation. The quantitative information was subject to descriptive and multivariate statistical techniques (using SPSS and Minitab® statistical software), including principal component analysis (PCA) and correlation analysis associated with applying simple and multiple linear regressions. A correlation analysis was performed to identify relationships between variables, followed by a simple linear regression analysis — allowing at confirming the results of the correlation analysis. A multiple linear regression analysis was also performed to understand the effects of the strategic association of companies in k-networks on the IKM and IC in SMEs. Each of the networks was then analysed separately, and then comparative analysis of the results of both networks was carried out. The main sample was segregated into two subsamples to understand internal consistency and the different peculiarities of the two networks. Together, the carried-out analyses will allow a deeper understanding of the relationships investigated in this study — considering quantitative and qualitative information. In this sense, the analysis of open questions (see Appendix, Table 8), allowed a better understanding of the phenomenon and its relations observing the effects from the association of SMEs in k-networks on IKM and IC, and the resulting influence on the corporate results such as organizational performance, innovation, sustainability and competitiveness (cf. Jordão & Almeida, 2017) — aspects considered to understand their organizational growth. Throughout data analysis, there was an alternation between induction and deduction (Eisenhardt, 1989; George & Bennett, 2005) with the latter predominating over the former, as the knowledge arising from the research derived from pre-established theoretical constructions and results. Seeking to increase the study’s internal consistency, the information gathered from the various sources of evidence was joined (triangulation process) (Jick, 1979). Therefore, whenever possible, the information from one source was compared with that from the others to confirm and validate it. Results This investigation focused on two SMEs’ k-networks: IBN and PN. IBN has been formed by SEBRAE to increase exportations, improve the volume and the quality of the businesses in SMEs and make them more competitive. The central premise of SEBRAE was the SMEs’ transformation to absorb the culture of internationalization and networking, through courses, visits and consultancies — which also increased the absorption of tacit and explicit knowledge. On the other hand, the SMEs that are part of the PN, instead of participating in a project that coordinated and supported their association, spontaneously grouped in an attempt to survive to a restructuring process in the pharmaceutical industry in Brazil, helping them become more competitive in front of the big competitors that dominate the market in which they operate. Although the initial idea of these companies was to make joint purchases, they gradually realized the benefits of being associated with and sharing data, information and knowledge. Thus, both the IBN and PN networks fall under the concept of k-networks by Moslehi et al. (2014). K-networks can be considered as a type of organizational innovation, according to the taxonomy proposed by OECD (2010), being a fundamental strategy to accelerate the pace of SME innovations (cf. Hilmersson & Hilmersson, 2021). To analyze empirically the role of k-networks on the IKM and IC in SMEs, the research model was tested. Internal consistency of the questionnaires, constructs and the research model were assessed using Cronbach’s alpha. A value of 0.886 was obtained, which denotes good internal consistency (Cf. Hair et al., 1998). Descriptive Statistics of the Two Networks Together and Constructs Validity Table 2 shows the aggregate result of the two networks together. Analysing the data structure, questions 3, 4, 5, 9, 10, 12 and 19 were seen to have a degree of significance greater than 5% (sig > 0,05), indicating they could be redundant by not differentiating respondents’ opinions. This uniformity of opinions may have been caused by the interaction between actors and sharing of data, information and knowledge, in line with the findings of Caldas and Cândido (2013), Khalique et al. (2015) and Vale et al. (2016), as the results served as an indicator of network consistency and maturity, revealing the systematization of new and old knowledge in a network asset.Table 2 Descriptive statistics and the results of principal component analysis are presented in the last two columns Intervening factors and items comprising each factor Descriptive statistics Principal components analysis (PCA) results M SD Sig. test Process 1 4.60 1.36 0.015 0.621 4 5.86 1.19 0.122 22 4.91 1.15 0.000 Context 16 4.57 1.63 0.001 0.644 17 5.80 2.43 0.050 20 5.54 1.36 0.001 Strategy 9 5.51 1.61 0.177 0.607 10 5.77 1.35 0.131 11 5.89 0.99 0.050 12 5.26 1.34 0.065 Culture 5 5.40 1.52 0.457 0.668 6 5.69 1.39 0.004 7 5.11 1.02 0.000 8 5.49 1.01 0.006 Stimuli 18 5.51 1.54 0.047 0.780 21 5.43 1.24 0.032 23 5.06 1.59 0.000 IKM 2 4.11 1.61 0.000 0.400 3 5.57 0.92 0.263 13 5.06 1.47 0.017 19 5.31 1.47 0.066 IC 14 4.91 1.36 0.035 0.533 15 4.80 1.45 0.013 24 5.29 1.20 0.000 Source: own elaboration, based on the research results Initial analyses cover basic estimates of the variables by group, through descriptive statistics (Hair et al., 1998), as presented in Table 2 in accordance with Appendix (see Tables 7). The results indicated that respondents agree with the questions to a greater or lesser extent, except for question 2, which revealed a feeling of almost indifference, and for questions 1 and 16, which showed low agreement and significant variability of understanding among respondents. In addition, to deepen the analysis of the constructs, their relationships and robustness, PCA was used (see Table 2), under Agostini et al. (2017). The PCA results revealed that the constructs forming the intervening factors present suitable reliability (except for Strategy, which was considered suitable by achieving more than 57% of explanation, but within the acceptable limit of 5%). The constructs subject to the intervention were not considered suitable — being a little below that minimum value. Seeking to maintain the logical integrity of the constructs in line with the theoretical model chosen, it was not possible to improve the reliability of the scale by excluding indicators, and so qualitative indicators were used to a perfect understanding of the phenomenon. That decision was also justified as the software used in this analysis (Minitab®) accepted the formation of the constructs with the variables suggested by the literature, confirming there was sufficient similarity among the questions. This approach is sufficiently sensitive to capture the complexity inherent to IKM and IC in SME networks according to Richardson (2013) and Agostini et al. (2017). The third and final component in assessing the validity of constructs is known as the nomological chain and aims to contrast the data obtained with the assumptions found in the literature by analysing the research model. This was achieved by complementing previous analyses with Spearman’s correlation analysis between the constructs (described in Table 3), combined with the use of simple and multiple linear regressions (described in Tables 4 and 5), evaluating to what extent the constructs explain IKM and IC in SME k-networks.Table 3 Results of Spearman’s correlation analysis (IBN and PN) Results of the Construct Correlation Analysis Relationship of the Constructs Based on PCA Constructs Process Context Strategy Culture Stimuli IKM IC Process 1 - - - - - - Context 0,363* 1 - - - - - Strategy 0,115 0,215 1 - - - - Culture 0,356* 0,019 -0,327 1 - - - Stimuli 0,382* 0,385* -0,102 0,623** 1 - - IKM 0,378* 0,573** 0,420* 0,107 0,433** 1 - IC 0,275 0,370* 0,448 0,222 0,483 0,427* 1 Source: Own elaboration, based on the research results Correlation is significant at the 0.05 level (2-tailed) or 0.01 level (2-tailed) Table 4 Results of simple linear regression analysis (IBN and PN) Relationship of variables trough simple linear regression IKM 0.157 0.285 0.350 0.015 0.164* – IC 0.059 0.190 0.233 0.05 0.141* 0.355** Regression equation IKM IKM = 5.680 + 0.4065 Process IKM = 6.075 + 0.4078 Context IKM = 4.377 + 0.4887 Strategy IKM = 8.093 + 0.1047 Culture IKM = 6.293 + 0.3120 Stimuli – IC IC = 6.302 + 0.2454 Process IC = 5.928 + 0.3270 Context IC = 4.567 + 0.3917 Strategy IC = 7.787 + 0.0604 Culture IC = 5.789 + 0.2842 Stimuli IC = 3.046 + 0.5847 IKM Source: own elaboration, based on the research results Correlation is significant at the 0.05 level (2-tailed) or 0.01 level (2-tailed) Table 5 Results of multiple linear regression analysis (IBN and PN) Relationship of variables through multiple linear regression IKM predictor Model summary ANOVA Coefficients Collinearity statistics Waste test Anderson–Darling Constant R2 Sig Non-standardized Tolerance VIF Cook distance Test Strategy 0.350* 0.000 0.438 0.982 1.018 0.000 (min) 0.204 (max) sig. level 0.05 Test p-value 0.318 AD 0.414 Context 0.000 0.354 0.982 1.018 Strategy and context 0.560 0.000 Dependent variable: IKM Regression equation: IKM = 2.196 + 0.438 strategy + 0.354 context KM predictors (constant): strategy (a), context (b), strategy and context (c) IC predictor Model summary ANOVA Coefficients Collinearity statistics Waste test Anderson–Darling Constant R2 Sig Non-standardized Tolerance VIF Cook distance Test Strategy 0.406 0.000 0.837 0.854 1.170 0.000 (min) 0.820 (max) Sig. level 0.05 Test p-value 0.243 AD 0.462 Culture 0.000 0.556 0.854 1.170 Strategy and culture 0.670** 0.000 Dependent variable: IC Regression equation: IC = − 5.830 + 0.837 strategy + 0.556 culture IC predictors (constant): strategy (a), culture (b), strategy and culture (c) Source: own elaboration, based on the research results Correlation is significant at the 0.05 level (2-tailed) or *0.01 level (2-tailed) Correlation Analysis A correlation analysis was performed to identify relationships between variables (see Table 3). Results of Table 3 show that Process, Context, Strategy and Stimuli are correlated to IKM, while Context, Strategy and Stimuli are correlated with IC. Moreover, several indirect effects on IKM and IC were also observed through correlations between these four constructs. A positive correlation between IKM and IC was also found. However, these results need to be interpreted in the light of the literature to reveal the causal relationships between the variables and support the inferences necessary to build a chain of evidence about the subject under study. In this sense, the results extend the findings of Pöyhönen and Smedlund (2004), López-Sáez et al. (2010) and Caldas and Cândido (2013), because most respondents understood that the network led to the creation of a shared language, facilitating the firms’ access to, and use of the information and knowledge built in IBN and PN. It was found that partnerships supply some technical and managerial help, encouraging the retention of information and knowledge and that the SMEs benefit from the exchange of information, understanding that the relations between the different companies in IBN and PN facilitate the IKM process, partially confirming the results of Valkokari and Helander (2007) and Jardon and Martos (2012). It was also perceived that IBN and PN provide participant firms with competitive differentials, generating the competencies (knowledge, skills and attitudes) necessary to develop their business, thereby complementing Jardon and Martos (2012), Lin and Chen (2016) and Cerchione and Esposito (2017). Inversely, culture does not produce significant effects on IKM or IC, contradicting the previous findings of Cegarra-Navarro et al. (2011) and Martins and Solé (2013). Finally, although the IKM and IC constructs were not very statistically robust, the literature and the software used to analyze the groups of variables that make up both constructs were largely supportive of such constructs. In both cases, the qualitative findings, analysed in the light of the international literature, showed that there is a strong link between IKM and IC, corroborating Jardon and Martos (2012) and Novas et al. (2017). Indeed, those questions concerning the sensitivity of the IKM and IC constructs could be better explained from the observations of Henry (2013), for whom IC involves tacit knowledge, and as such is a hidden dimension of knowledge which can often be ‘non-verbal’, or in some cases ‘unable to be verbalized’, being also intuitive and unarticulated. Because of this, it is more difficult to observe, capture and structure — which consequently hinders its management. Simple and Multiple Linear Regression Analysis The analysis of the simple linear regression, presented in Table 4, confirms the results of the correlation analysis, as Process, Context, Strategy and Stimuli can explain IKM, while Context, Strategy and Stimuli can explain IC. In an attempt to deepen the understanding of the effects of the k-networks on the IKM and IC on these SMEs, linear regressions were analysed, performing a simultaneous test of the relationships of all the constructs presented in Table 5. The results revealed that only Strategy and Context, in that order, are predictive constructs of IKM, and that only Strategy and Culture, in that order, are predictors able to explain the behaviour of IC in the SME networks under analysis. Analysis of Table 5, based on the multiple linear regressions, reveals that the Strategy construct, alone, can predict 35% of IKM, and when combined with Context, almost 56%, based on the variance analysis (ANOVA). Collinearity and multicollinearity tests showed that the tolerance levels of the explanatory model of IKM are completely suitable. A value of 0.982 was obtained for the first (values over 0.10 and closer to 1.00 are required) and a variance inflation factor (VIF) of 1.018 for the second (values under 5.00 and closer to 1.00 are required). Analysis of the multiple regression equation (IKM = 2.196 + 0.438 Strategy + 0.354 Context) indicates that a one-unit increase in Strategy and Context constructs can justify an increase of up to 0.438 and 0.354 units of IKM, respectively. To prevent the model from supplying potentially misleading or incorrect results, the Cook distance parameters observed (minimum 0.000 and maximum 0.204) were shown to be suitable. Finally, taking the Anderson–Darling (AD) test as a basis to test the normality of residuals, where the p-value of this test must be greater than the level of significance chosen (in this case 0.05), completely valid results were obtained (AD 0.414 > p-value of the test 0.318 > 0.05), indicating that the errors behaved like a normal variable. Similarly, this analysis revealed that the strategy construct was able to explain, alone, the variability of IC by slightly over 40%, and by 67% when combined with the Culture construct, with extremely high statistical significance. The collinearity tests of the explanatory model of IC were also completely suitable — 0.854 — as were the results of the VIF — 1170. Together, these findings revealed there were no problems in the multiple regression models of IKM and IC. Analysis of the multiple regression equation (IC = − 5.830 + 0.837 Strategy + 0.556 Culture) shows the potential increase in IC derived from the one-unit increase in Strategy and Culture constructs. Finally, the parameters observed for the Cook distance, (minimum 0.000 and maximum 0.820) and AD (AD 0.462 > p-value of the test 0.243 > 0.05), were shown to be perfectly suitable. As a whole, three constructs (Strategy, Context and Culture) are seen to be explanatory factors of IKM or IC, with only the Strategy construct producing direct effects on both simultaneously. The results of analysing Tables 2, 3 and 4 had already revealed the constructs Context, Strategy and Stimuli as having direct effects on IKM and IC through the correlations between these aspects. The Process construct was found to influence IKM directly. Only the Culture construct did not have a direct impact on these aspects. The joint test of the constructs revealed that only the Strategy and Context constructs, in that order, are predictors of the behaviour of IKM and that only Strategy and Culture, in that order, explain the behaviour of IC in the SME networks analysed. This shows the preponderance of Strategy as the main factor in explaining the phenomenon studied. Nevertheless, a joint analysis of the results of the simple and multiple linear regressions reveals that all constructs have direct or indirect effects on IKM or IC. Furthermore, the IKM and IC constructs were found to influence each other. Therefore, there is alignment between the theoretical assumptions established in the premises of the model proposed by Jordão and Novas (2017) and the empirical results observed in the research described in this paper. Deepening of the Analysis and Discussion of the Results Based on the analysis of the answers to closed and open questions (cf. Appendix, Tables 7 and 8), in each of the networks, we sought to understand the influence of the association of SMEs in k-networks, as well as the effects of this process. The statistical results indicate that the vast majority of SMEs in the two networks realized that the network constitution was relevant to improve the sharing of data, information and knowledge; expand communication and generate new skills. The analysis of the data of Table 6 reveals that the network formation process was successful since it allowed the creation of a shared language and the transfer of information and knowledge (explicit) in both networks, even though the results may have remained slightly below expectations, especially at IBN. Even so, the results are in line with the ideas of Nonaka and Takeuchi (1995) about ‘Ba’ as a space for the dissemination of knowledge, which are now extended to the context of SME networks.Table 6 Descriptive statistics of the two analysed networks (IBN and PN) Intervening factors and items comprising each factor IBN PN M SD Mo Md M SD Mo Md Process 1 4.44 1.39 5 5 5.00 1.25 5 5 4 5.80 1.32 7 6 6.00 0.82 6 6 22 4.88 1.01 5 5 5.00 1.49 6 6 Context 16 4.60 1,55 5 5 4.50 1.90 5 5 17 7.00 0 7 7 2.80 2.90 1 1 20 5.40 1.41 6 6 5.90 1.20 7 6 Strategy 9 6.08 1.22 6 6 4.10 1.66 5 4.5 10 6.08 1.22 6 6 5.00 1.41 6 5 11 5.96 0.93 6 6 5.70 1.16 6 6 12 5.36 1.44 6 6 5.00 1.05 6 5 Culture 5 4.80 1.38 5 5 6.90 0.32 7 7 6 5.40 1.53 6 6 6.40 0.52 6 6 7 5.08 1.08 6 5 5.20 0.92 5 5 8 5.12 0.83 5 5 6.40 0.84 7 7 Stimuli 18 5.28 1.37 5 5 6.10 1.85 7 7 21 5.20 0.87 5 5 6.00 1.83 7 6.5 23 4.92 1.50 6 5 5.40 1.84 6 6 IKM 2 4.56 1.23 5 5 3.30 1.89 3 3 3 5.72 0.68 6 6 5.20 1.32 5 5 13 5.08 1.08 5 5 5.00 2.26 5 5.5 19 5.08 1.29 5 5 5.90 1.79 6 6 IC 14 4.88 1.27 6 5 5.00 1.63 5 5 15 5.16 1.28 5 5 3.90 1.52 5 4 24 5.16 0.69 5 5 5.60 2.01 7 6.5 Source: own elaboration, based on the research results M mean, SD standard deviation, Mo mode, Md median Further details of the research results, bringing together descriptive and multivariate statistics (in particular the analysis of correlations and simple and multiple linear regressions), whose synthesis is presented in Fig. 2, reveal that all constructs are directly related either with IKM or IC, in addition to indirect relations with other constructs. The analysis of Fig. 2 reveals that direct statistical relationships were perceived between the Process construct and IKM, between the Culture construct and IC, and between the Context, Strategy and Stimulus constructs with both IKM and the IC.Fig. 2 Analysis of the effects of k-networks on IKM and IC. Source: own elaboration, based on the research results. Blue arrow: result of the Spearman correlation. Black arrow: result of the multiple linear regression Initially, some questions composing the constructs still have less agreement, and this could lead to questioning whether the formation of IBN and PN would be, or not, an impacting factor for IKM or IC. However, the set of qualitative results strongly suggests that without these networks and without SMEs wishing and striving to share information and knowledge, k-networks would not emerge — which ended up having a positive impact on IKM and indirectly on SMEs’ IC. The qualitative results indicate that the network was central in the process of competence development needed for the business of both networks, although with greater uniformity in PN than in IBN — in which a group of IBN CEOs considers that the process has not reached a satisfactory level. This can be explained because some IBN SMEs were not satisfied with the process of forming the network and the benefits derived from it. These results seem to suggest that forming a network with companies of different sizes, in different sectors, or with different levels of maturity (in the same workgroup) can generate dissatisfaction among more mature companies due to the slow learning process of beginners. Therefore, it is confirmed that the formation of the network was positively impacting for IKM but not for IC, with a tendency to have facilitated the creation and transfer of explicit and tacit knowledge, in line with the conclusions of Moslehi et al. (2014). In a similar and slightly more homogeneous way than the Process, the statistical results indicate that, in general, most CEOs agreed with the statements that explain the Context in which the k-networks are constituted. The importance of knowing the business process of other companies and taking advantage of the exchange of information from this knowledge as well as of the relationship between the different companies and their stakeholders was revealed, especially in the PN, with the creation and sharing of data, information and knowledge being necessary issues, considering the qualitative results, corroborating the findings of López-Sáez et al. (2010), Caldas and Cândido (2013) and Suárez (2013). Even so, quantitative results indicated that the contextual aspects were less important than expected in understanding the formation of k-networks. This also seems to suggest that some SMEs of both networks were not yet able to fully develop what the international literature calls Ba (Cf. Nonaka et al., 2000) considering open questions’ answers. In general, these Ba promoted by the network were considered a factor that stimulated IKM and IC, due to the greater flow of data, information and shared knowledge, as well as the feasibility of joint actions with partner companies, confirming the assumptions of Mertins et al. (2010), Santos-Rodrigues et al. (2012) Martins and Solé (2013) and Leal-Millán et al. (2016). The statistical results still indicate that the respondents agreed with the statements that explain the analysed strategic factors, revealing the need to know and use their information, knowledge and skills as strengths to overcome their weaknesses and generate new business opportunities. However, IBN companies’ CEOs seemed much more aware of the skills and competencies needed to carry out its processes. The qualitative results also indicate that the networks generated competitive differentials for the participating companies, providing the knowledge needed for the development of their businesses. Triangulated results confirm that the strategy (formal and informal) affected IKM and IC, revealing that networks have been used as a deliberate strategy to help SMEs assess their strengths, increase the opportunities presented in the environment and minimize the existing risks, going beyond the findings of Cegarra-Navarro et al. (2011), Mertins and Orth (2011), Khalique et al. (2015) and Marques Júnior et al. (2020), as this resulted in the strengthening of the IC’s component elements. These results also show that the vast majority of companies realized that the knowledge of the skills and necessary competencies in the performance of their processes can turn into corporate results, innovation and growth, confirming and expanding the classical ideas of Prahalad and Hamel (1990) and Montgomery and Porter (1991) that skills and competencies must be carefully developed to create and maintain competitive advantages. As with contextual aspects, cultural elements seemed less important than previously thought. A possible explanation would be because as the k-networks consolidated, they stimulated shared forms of behaviour, routines, beliefs and values (cf. Cegarra-Navarro et al., 2011; Mason et al., 2008). This condition was central to stimulating IKM practices of SMEs, allowing these companies to form coalitions, reshape their own culture based on their needs and those of the group, develop common interests, stimulating the creation of information centres and knowledge sharing, in addition to trying to improve the economic viability of enterprises through innovation, in line with the findings of Martínez-Costa et al. (2019). The findings reveal that the members of the network have promoted interactions that took place outside the formally planned moments in the network. CEOs of both networks, especially PN, agree that similar cultures, languages and shared experiences facilitated the dissemination of data, information and knowledge within the network. They also agree on the existence of a culture of passing practical knowledge from other companies to the network in the form of theoretical knowledge, which is then converted into practical knowledge when used by the companies in the network (however, this cultural element is a little more rooted in PN than in IBN). Finally, they agree that the network helped in forming a differentiated way of acting or in creating the group’s values. Despite this evidence, the relationship between organizational culture and IKM was only indirectly confirmed. Thus, considering qualitative results, it appears that cultural and social aspects helped to form and expand IC elements (i.e. human, structural and relational capitals), being indirectly important for practices related to information and organizational knowledge. Both networks demonstrated the role and relevance of stimuli for sharing data, information and knowledge within the network. The CEOs perceived network environment, intellectual honesty (i.e. people are authentic and make it clear what they know, or do not know, about the acquired experiences), authenticity in relationships and incentives to share knowledge and experiences as relevant factors for improvements in IKM and IC, aligning with López-Sáez et al. (2010), Mariotti (2011), Muhammad et al. (2011), Caldas and Cândido (2013), Verbano and Crema (2016) and Bolade (2021). The set of results also indicated that the respondents generally agreed with the statements that explain IKM in networked SMEs, pointing out their relevance in the document, creation and sharing of information (although more intensely in PN), knowledge and know-how; that discipline, efficiency and incentive are fundamental to systematize this knowledge and that the existence of tools and access to information are factors that catalyse both processes. Taken together, the analysis of the results indicates that the configuration of the k-network seems to have provided the IBN and PN SMEs with favourable conditions to create and expand organizational knowledge, a true strategic knowledge community and the emergence of some types of Ba. Such networks seem to have favoured effective interaction between people, groups and organizations, stimulating the sharing of skills, experiences, emotions, information and knowledge through face-to-face communication and generating an intense climate for the sharing of tacit knowledge, in line with the findings of Richardson (2013) and Agostini et al. (2017), understanding k-networks as fundamental for IKM and learning and innovative processes. Likewise, the results indicated that the respondents, in general, agreed with the statements that explain the elements of the IC. Quantitative and qualitative results indicate that the vast majority of companies in the two networks developed a physical and social environment (structural capital) to create knowledge after the network formation process. There was a high investment and incentive in terms of personal and professional training and qualification in each of the companies in the networks (although more widely and intensely in the IBN network), resulting in human capital, corroborating Xu et al. (2020). Even so, and more expressively among PN CEOs, it was noticed that the relationships between members (relational capital) stimulated the creation and sharing of information and knowledge between the companies in both networks, implying learning and/or innovation processes in accordance to Jordão (2015). The result set allows confirming the stated deep and intense relationship between IKM and IC, going beyond the findings of Henry (2013), Vale et al. (2016) and Agostini et al. (2017), since the IC has a direct and indirect positive effect on IKM in both networks, and vice versa, even though there seems to be a circular relationship between IKM and IC, confirming the premises of Novas et al. (2017). The qualitative findings also extend the observations of Korbi and Chouki (2017), because it seems that the interaction between individual knowledge at the SME level has expanded in a knowledge spiral, dynamically rising from a low to a high ontological level, up to reach and consolidate at the level of the inter-organizational network. Besides, it was noticed that the k-networks generated new knowledge, high growth and sustainable competitive advantages for the participating companies. This seems to be related to greater performance and innovations in SMEs. Aiming to deeper the analysis, data was triangulated to better understand how SMEs have used (or not) their IKM and IC to survive and to catalyse and sustain their organizational growth, going beyond the findings of Jordão et al. (2022). In this case, the qualitative findings were fundamental to understand the peculiarities of the IKM and IC in the studied k-networks, corroborating Valkokari and Helander (2007) and Henry (2013), since the IC involves tacit knowledge (which comprises ‘non-verbal’ or ‘non-verbalizable’ knowledge, being intuitive and disjointed). In both cases, it was realized that IC is central to SMEs’ results and sustainable growth, as can be seen in a CEO’s response.The idea of transferring our data, information and knowledge throughout the network is rather difficult to accept at the beginning, as the firms that today are partners were our competitors and that encourages individualism and each one keeping information and knowledge for their own business. However, we realized at the beginning that it would be vital to share knowledge within the network if we wanted to be competitive, innovate, survive and grow. I think the results are good and that’s the greatest incentive (respondent 31) [our underlining]. These findings not only corroborate the results of Mertins and Orth (2011), Brassell and Boschmans (2019) and Xu et al. (2020) since the IC proved to be essential for survival, accelerated growth and the development of SMEs but also amplify previous findings since this aspect was perceived at both companies and the network’s levels. The set of quantitative and qualitative findings revealed that the association of SMEs in k-networks significantly stimulated IKM and IC, besides promoting long-term sustainable growth (Cf. Gupta et al., 2019; Jordão & Almeida, 2017), perceived by improvements in innovation, competitiveness, performance and value creation, for these companies and their networks. The set of answers to the open questions revealed that in the PN, there was a percentage growth of SMEs not only well above the national and world GDP, but even higher than the most successful companies in the sector. Since the formation of the network, the founders have worked to achieve structured growth (in part through acquisitions), continuous improvement and the perpetuation of the brand, doubling the number of stores, quintupling revenue and tripling the number of customers served in about 10 years. The increase in average profitability, however, has been above 25% per year, reaching 50% in some periods (regardless of Brazilian or global crises such as Covid-19). The most expressive growth was in the non-medication category: which includes hygiene, perfumery and cosmetics items. These results are due to the greater professionalization of management and the intensive use of IKM — that helped them to generate greater sales volume, at the same time in implementing lean cost management, obtaining greater operational and administrative efficiency, resulting in better performance. The CEOs of the PN confirmed that:Forming and constituting the network stimulated the process of creating, schematizing, and dividing information and knowledge among the firms, giving us much greater and more sustainable growth than that observed in the competition between independent firms (respondent 35) Compared to firms in the sector that are not in the network, our results indicate high growth, reaching more than 60% in all these aspects: improvements in qualifications, experience, creativity, knowledge, skills, innovation capacity, and the task development of firm members (respondent 26) [our underlining]. At IBN, the k-network provided a realistic view of the market, to overcome the lack of resources and to achieve systematic growth in the market based on the vision, mission, values and principles developed in this network. In addition to the benefits observed in the PN (including revenue, profitability and customer growth), it was also possible to see IBN developing a set of knowledge that helped SMEs to grow and strengthen and competitively explore new business opportunities in international markets — which was a very significant contribution, considering the size of the companies, going behind the findings of Mertins and Orth (2011), Jardon and Martos (2012), Khalique et al. (2015), Rafique et al. (2018), Balle et al. (2019), Martínez-Costa et al. (2019), Agostini et al. (2020), Xu et al. (2020) and particularly Zakery and Saremi (2021) — who observed that international businesses play a significant role in the growth and survival of SMEs. The CEOs of the IBN confirmed that:Before there was little focus, difficulty in distributing tasks, according to people's competencies. There was a lack of a role defined by partners for collaborators. As well as the vision, the network helped SMEs to grow a lot, producing innovations in systems, structures, results, and behaviors, improving processes, productivity, responsibility, and team spirit (respondent 5). I understand that the knowledge acquired from the network not only help us to survive but greatly facilitated periods of high growth and sustainable development for all SMEs over the years, particularly due to the development of factors linked to the firm’s IC, including our major element: the relations, cooperation and trust we have with each other (respondent 21) [our underlining]. These results are in line with Mertins et al. (2010). In both networks, incremental (gradual improvements) and radical changes (revolutionary or drastic) were noticed — which boosted the competitiveness and performance of companies. It was found that the development of innovation in SMEs depends dramatically on their IC management and on their ability to enter collaborative and dynamic networks in open business environments. The findings made it possible to verify four different types of innovation: in products, processes, organizational and marketing, in line with the OECD model (2010). Although innovation had initially been thought of as a consequence of the growth of companies, it was noticed that innovation had a relevant role in the competitiveness, survival and growth of SMEs. More than that, the research results go beyond the conclusions of Hilmersson and Hilmersson (2021) because the association of these companies in networks and their behaviour not only constitute, de per si, a form of organizational innovation (Cf. OECD, 2010), besides affecting the temporal aspects of the innovation processes but also catalysed several innovations in processes, structures, results and behaviours derived directly from the expansion of the IC and knowledge flows in SMEs and their networks. That is one of the most surprising findings of the research. In aggregate, the analysis of Fig. 2 confirms the validity of the model proposed by Jordão and Novas (2017). Overall, the triangulation of results revealed that networks are capable of creating, absorbing, applying and disseminating data, information, knowledge and expertise, as well as successful practices and tools within SMEs and their networks, in line with Munier (2021). Moreover, the results indicate that the information and knowledge can be shared with all members having the network, and this can expand the stocks of IC, generating significant and profound contributions to economic and managerial practice. Many institutions that deal with SMEs recognize these factors and call attention to the necessary care and effort so that they could retain, systematize and share their knowledge and create new ones to achieve continuous improvements in their products, services and processes, in addition to achieving corporate sustainable growth and value generation, in accordance to Jardon and Martos (2012), Xu et al. (2020) and Zakery and Saremi (2021). In general, this process was stimulated by k-networks, directly and indirectly, contributing to improvements in IKM and IC. In this sense, the results now observed can also serve as competitive benchmarking for other companies and networks in similar situations. These issues gain special relevance in a competitive context in which authors (e.g. Brassell & Boschmans, 2019; Durst & Edvardsson, 2012; Jordão & Novas, 2017; Jordão et al., 2020; Khalique et al., 2015; Xu et al., 2020; Zakery & Saremi, 2021) and international organizations (e.g. UN and OECD) draw attention to the role of SMEs for the economies, governments and societies, emphasizing the growing relevance of IC as the main factor responsible for the survival, growth and development of these companies. As practical contributions, these results help to understand how SMEs can use the k-networks as a deliberate and innovative strategy to overcome their technical, organizational and managerial difficulties, using networks to improve IKM practices and to expand IC — generating corporate sustainable growth and development (Cf. Gupta et al., 2019; Jordão & Almeida, 2017), considering the improvements in innovativeness, performance and other organizational and competitive benefits observed in both networks. From a theoretical point of view, the results help fill a research gap and shed light on the understanding of the relationships between k-networks, SMEs, IKM and IC, contributing to the expansion of knowledge on the topic, in line with the assumptions of Balestrin et al. (2008), Mertins et al. (2010) and Jordão and Novas (2017) — a great relevance aspect, but whose understanding is still incipient in international literature and almost non-existent in emerging markets, offering several paths for future investigation. In this sense, the study’s originality and its approach and contribution are highlighted. This study advances the theory of economics and management by helping in the real understanding of the integration of these four elements, besides presenting, testing and validating variables that support the model proposed by Jordão and Novas (2017) in a valuable context — considering that Brazil is one of the world’s leading developing economies. So, the results of this research should and can be used to guide future studies in the form of cases or large-scale tests, as well as help to create, test or refine theories on the subject. The findings of this study seek to advance the research flows that examine how the idiosyncratic resources resulting from the involvement of SMEs in k-networks can contribute to the success and high growth of this type of company. This paper also extends the literature on IC, IKM and corporate growth, noting the direct effects of the association of SMEs in a network on the first two factors and the indirect effect on the latter. It is relevant to mention that most of the previous research on SMEs and networks has studied companies from developed countries and ignored emerging economies. Likewise, in line with observations of Fallatah (2021), most studies on knowledge creation have focused on developed countries, where knowledge is more likely to be created, somewhat neglecting developing and emerging countries. Taken together, the findings reveal that the transformation of SME inter-organizational networks into k-networks seems to have brought significant benefits concerning processes, people, and market relations, helping to create a shared language with methods understood and used by network partners. At both IBN and PN, this seems to have been fundamental for the co-evolution of members (SMEs), promoting greater engagement and commitment, improving the learning capacity and ways for networked SMEs to catalyse and sustain periods of high growth, as well as suggesting an increase in the means of access to information and knowledge and its creation, transmission, absorption and use, in addition to improvements in performance, competitiveness and innovation — which seem to be directly or indirectly linked to the greater creation of IC within these networks. Conclusions The international literature stresses that in the knowledge-based economy context, SMEs find it difficult to exploit the potential of IKM alone, needing to act in networks to overcome their limitations and reach their goals. So, such enterprises establish inter-organizational relationships, through which they create, obtain, systematize and share information and knowledge and produce innovations. Strangely and paradoxically, even considering the great relevance of the subject, the challenge of understanding the effect of SMEs’ k-networks on IKM and IC has yet to be overcome, especially in emerging economies. Recognizing and exploring this research gap, this paper aims to analyze the strategic effects of the association of SMEs in k-networks on their IKM and IC. The results of the study indicate that the processes of creating, systematizing and sharing data, information and knowledge in SME k-networks strengthen business projects and processes and the generation of IC in all its dimensions: human, structural and relational. In those networks, IC was heightened by practicing IKM and its underlying meanings in the scope of the SMEs and the networks themselves. Both networks were found to be successful, which is dependent on their actors’ ability to create, employ and circulate information and knowledge and generate IC, promoting long-term sustainable growth and producing innovations in processes, structures, systems, results and behaviours in addition to gains in competitiveness and corporate results. The findings revealed that the network formation process alone did not imply improvements in IKM or greater IC. The views of network formation were generally positive. However, work methodologies; partnerships; ways of leading groups; systematization of data, information and knowledge and the SME organization were matters showing a greater need for improvement. The findings indicate that it is not enough for SMEs to be similar or situated together. There must be structures and established contexts, strategies, cultures and stimuli for the sharing of information and knowledge among the firms participating in such networks. The interactions and relationships both inside and outside the network were seen to be important catalysts of the process of sharing data, information and knowledge by SMEs. Furthermore, contextual and cultural aspects in isolation were found to be less important than indicated in the literature. However, when combined with other aspects (such as strategy or stimuli), effects were produced on IKM and/or IC. The set of results led to the conclusion that context was less important in PN than in IBN, exactly the opposite of organizational culture. Within contextual aspects, the importance of creating a Ba stood out. This seems to help companies and people to absorb the tacit and explicit knowledge acquired and/or built in the process, with culture and inter-organizational relations being relevant factors in this regard. These networks led to the creation of an own informational and cognitive base and own network culture to orientate and direct inter and intra-organizational relationships in the SMEs. Similarly, the networks generated a community of interaction at an ontologically higher level than exists in firms — going beyond the limits of inter-organizational networks. Here, the results suggest that IBN and PN formed a meta-organization characterized by a complex system providing the infrastructure or resources to help the other members. Strategic factors were essential for successful undertakings, generating improvements in SMEs’ business performance, especially in innovation and competitiveness. All CEOs considered this was strategically important to create, develop and apply data, information and knowledge, generating higher and more innovative performance. Findings also revealed that mistrust was overcome by the commitment and intellectual honesty between network participants. These stimuli for good relations and greater trust inside and outside the network not only reduced uncertainties but also helped in exploiting new markets, products and technology, as well as encouraging learning and knowledge between people and SMEs, promoting long-term sustainable growth, perceived by improvements in innovation, competitiveness and corporate results for SMEs and their networks. Finally, it remains to be said that the strategy and stimuli implied a learning process and innovation in both networks, resulting in sustainability, competitiveness and business success. As a whole, findings revealed that the constructs of Context, Strategy, Culture and Stimuli were the four producing the most significant effects on IKM and/or IC, with the strategy being the most fundamental element in this process. Taken together, the results of the research confirmed the model’s conformity and internal consistency, in quantitative and qualitative terms, showing it to be valid and consistent. The main limitation of the research is related to a low generalization due to the small size of the sample. This leads to suggesting large-scale studies about SME k-networks to corroborate the findings of this study, to refine the model’s results and to expand theoretical understanding of the topic. However, before going on to such studies, more case studies of a qualitative nature, either an individual or comparative, are necessary to refine the model and for a better perception of the particularities of the phenomenon. All in all, it is hoped that, besides making contributions to theory by increasing understanding of k-networks’ effects on IKM and IC, these conclusions can also contribute to economic, legal and management practice, improving understanding of the importance and role of the strategic association of companies in k-networks for SMEs’ survival, growth and development, as well as providing owners and managers of SMEs in similar situations with a competitive benchmarking process. Appendix Table 7 The first column of the table presents the five intervening factors and the items comprising each factor; the second column presents the theoretical support of each item Intervening factors and items comprising each factor Theoretical support Process 1 – Network formation promoted the creation and sharing of a language and methods among participants. e.g. Nonaka and Takeuchi (1995); Jordão and Novas (2017) 4 – Network formation was fundamental in developing the necessary competences for your business. e.g. Jardon and Martos (2012); Mariotti (2011); Romiti and Sarti (2011); Jordão and Novas (2017) 22 – Network formation allowed the firm to access and uses the information and knowledge constructed in the network. e.g. Balestrin et al. (2008); Mariotti (2011); Suaréz (2013); Jordão and Novas (2017); Balle et al. (2019) Context 16 – SMEs know the business process of other firms and benefit from that knowledge. e.g. Romiti and Sarti (2011); Jordão and Novas (2017); Zach and Hill (2017) 17 – Network partnerships provide technical and management support, stimulating the creation and retention of information and knowledge. e.g. Mertins et al. (2010); Caldas and Cândido (2013); Jordão and Novas (2017) 20 – The relations between the different firms in the network facilitate information exchange and the knowledge creation process. e.g. Powell (1998); Caldas and Cândido (2013); López-Sáez et al. (2010); Suaréz (2013); Zach and Hill (2017) Strategy 9 – There is a high level of consensus regarding the firm’s strengths and weaknesses. e.g. Andrews (1971); Jordão and Novas (2017) 10 – The firm is well aware of its opportunities and threats. e.g. Andrews (1971); Durst and Ferenhof (2014) 11 – The firm is well aware of the competences necessary to carry out its processes. e.g. Montgomery and Porter (1991); González-Loureiro and Dorrego (2012); Jardon and Martos (2012); Khalique et al. (2015); Jordão and Novas (2017) 12 – The network allowed competitive differentials, generating information and knowledge for the development of SMEs’ business. e.g. Chaston and Mangles (2000); Romiti and Sarti (2011); Jardon and Martos (2012); Martins and Solé (2013); Massaro et al. (2014) Culture 5 – Network members have interactions besides those occurring formally in the network environment, facilitating the sharing of experiences. e.g. Polanyi (1966); Jordão and Novas (2017) 6 – Similar cultures, languages and shared experiences help to create and spread knowledge within the network. e.g. Cegarra-Navarro et al. (2011); Hofstede (1982); Jardon and Martos (2012); Jordão et al. (2014); López-Sáez et al. (2010); Martins and Solé (2013); Suaréz (2013) 7 – The existing culture stimulates the share, transformation and application of other SMEs’ theoretical and practical knowledge. e.g. Davenport and Klahr (1998); Jardon and Martos (2012); Jordão and Novas (2017) 8 – The network helped in forming a differentiated way of acting or in creating the group’s own values. e.g. Mason et al. (2008); López-Sáez et al. (2010) Stimuli 18 – Network members make it clear what they know and also what they do not know about acquired experiences. e.g. Ebers and Jarillo (1998); Valkokari and Helander (2007) 21 – The network made it possible for firms to share data, information and knowledge. e.g. Powell (1998); López-Sáez et al. (2010); Mariotti (2011); Verbano and Crema (2016); Balle et al. (2019); Xu et al. (2020) 23 – There are incentives for individual knowledge to be shared with other network members. e.g. Jarillo (1988); Jardon and Martos (2012); Verbano and Crema (2016) IKM 2 – There is discipline, efficiency and an incentive for the documentation of information, knowledge and know-how developed in the networks. e.g. Mariotti (2011); Jordão (2015); Jordão and Novas (2017) 3 – The network was able to create information and knowledge that could be absorbed and applied by the firms. e.g. Mariotti (2011); Jordão (2015); Jordão and Novas (2017); Balle et al. (2019) 13 – There are formal tools that help to spread knowledge and successful practices within the network. e.g. Prahalad and Hamel (1990); Jordão (2015); Jordão and Novas (2017); Xu et al. (2020) 19 – Information is shared and accessed by all elements of the network. e.g. López-Sáez et al. (2010) IC 14 – The network association led to the development of a physical and social environment (structural capital) for the creation of knowledge. e.g. López-Sáez et al. (2010); Nonaka et al. (2000); Balle et al. (2019) 15 – There is investment in, and incentives for the training of network members, resulting in human capital. e.g. Chaston and Mangles (2000); Muhammad et al. (2011); Novas et al. (2017); Xu et al. (2020) 24 – The relationships between network members (relational capital) stimulated the creation and share of information and knowledge, learning and/or innovation. e.g. Bengtsson and Kock (2000); Bhatt (2001); González-Loureiro and Dorrego (2012); Mason et al. (2008); Massaro et al. (2011); Mertins and Orth (2011); Pöyhönen and Smedlund (2004); Richardson (2013); Suaréz (2013); Belso-Martinez and Diez-Vial (2018) Table 8 Open questions used to deepen the understanding of the relationships between SME k-networks, IKM and IC, and theoretical support of each item I Open Please describe (i) the situation of your company before, during and after its integration into the network, highlighting, (ii) the indicators used to assess the generated knowledge, (iii) the results obtained and (iv) the critical factors of the information and knowledge exchange process. Grandori and Soda (1995); Mariotti (2011); Caldas and Cândido (2013); Martins and Solé (2013) II Open Did the formation and constitution of the network stimulate a process of creation, systematization and/or sharing of data, information and knowledge between companies? Muhammad et al. (2011); Mariotti (2011) III Open Did the knowledge acquired with the network organisation facilitate the development of factors related to the company's relational capital, such as: brand development, customer and supplier relations, company image, and ways of doing business and/or dealing with distribution channels? Mason et al. (2008); Mertins et al. (2010); Mertins and Orth (2011); Mariotti (2011); Muhammad et al. (2011); Novas et al. (2017); Jardon and Martos (2012); Martins and Solé (2013); Massaro et al. (2014) IV Open Did the knowledge acquired through the network facilitate the development of factors related to the company's human capital, such as: improvements in qualification, experience, creativity, knowledge, skills, capacity for innovation and/or to develop tasks of the company's members? Mason et al. (2008); Mariotti (2011); Muhammad et al. (2011); Novas et al. (2017); Jardon and Martos (2012); Martins and Solé (2013); Massaro et al. (2014) V Open Did the knowledge acquired through networking facilitate the development of factors linked to the company's structural capital, such as: improvements in structure, design, technologies, methodologies, employed processes, business systems and/or information systems? Mason et al. (2008); Mariotti (2011); Muhammad et al. (2011); Novas et al. (2017); Jardon and Martos (2012); Henry (2013); Martins and Solé (2013); Massaro et al. (2014) Author Contribution Among the individual contributions to the article provided by Professor Ricardo Vinícius Dias Jordão, the conceptualization, formal analysis, financing acquisition, research development, the design of the methodology, project administration, data validation and writing (draft, proofreading and editing). Professor Jorge Casas Novas helped with the literature review, conceptualization, formal analysis, financing acquisition, research development, refinement of the methodology and writing — proofreading and editing. Funding The authors received financial support from National Funds of the FCT – Portuguese Foundation for Science and Technology within the project grant ‘UIDB/04007/2020’. Declarations Consent for Publication The manuscript has not been submitted for publication nor has been published in whole or in part elsewhere, and is not under consideration for publication in another journal at the time of submission. I attest to the fact that all authors have read the manuscript, confirm the validity and legitimacy of the data and its interpretation, and agree to its submission to The Journal of the Knowledge Economy, and that, if accepted, it will not be published elsewhere without the written consent of the copyright holder. Competing Interests 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 Agostini L Nosella A Filippini R Does intellectual capital allow improving innovation performance? A quantitative analysis in the SME context Journal of Intellectual Capital 2017 18 2 400 418 10.1108/JIC-05-2016-0056 Agostini L Nosella A Sarala R Spender J-C Wegner D Tracing the evolution of the literature of knowledge management in inter-organisational contexts: A bibliometric analysis Journal of Knowledge Management 2020 24 2 463 490 10.1108/JKM-07-2019-0382 Andrews, K. R. (1971). The concept of corporate strategy. Dow Jones Irwin. Balestrin A Vargas L Fayard P Knowledge creation in small-firm network Journal of Knowledge Management 2008 12 2 94 106 10.1108/13673270810859541 Balle A Steffen M Curado C Oliveira M Interorganisational knowledge sharing in a science and technology park: The use of knowledge sharing mechanisms Journal of Knowledge Management 2019 23 10 2016 2038 10.1108/JKM-05-2018-0328 Belso-Martinez J Diez-Vial I Firm’s strategic choices and network knowledge dynamics: How do they affect innovation? Journal of Knowledge Management 2018 22 1 1 20 10.1108/JKM-12-2016-0524 Bengtsson M Kock S “Coopetition” in business networks – to cooperate and compete simultaneously Industrial Marketing Management 2000 29 1 411 426 10.1016/S0019-8501(99)00067-X Bhatt GD Knowledge management in organizations: examining the interaction between technologies, techniques, and people Journal of Knowledge Management 2001 5 1 68 75 10.1108/13673270110384419 Bolade S A complementary perspective of knowledge resources Journal of the Knowledge Economy, Ahead-of-Print. 2021 10.1007/s13132-021-00743-8 Brassell, M., & Boschmans, K. (2019). Fostering the use of intangibles to strengthen SME access to finance. OECD SME and Entrepreneurship Papers 12. OECD Publishing, Paris. 10.1787/729bf864-en Caldas, P., & Cândido, G. (2013). Inter-organisational knowledge conversion and innovative capacity in cooperative networks. Journal of Technology Management & Innovation, 8(S.1), 104–114. 10.4067/S0718-27242013000300009 Castells, M. (1996). The rise of the network society. The Information Age. Economy, Society, and Culture (Vol. I). Blackwell, Oxford. Cegarra-Navarro J Sánchez-Vidal M Cegarra-Leiva D Balancing exploration and exploitation of knowledge through an unlearning context: An empirical investigation in SMEs Management Decision 2011 49 7 1099 1119 10.1108/00251741111151163 Cerchione R Esposito E Using knowledge management systems: A taxonomy of SME strategies International Journal of Information Management 2017 37 1 1551 1562 10.1016/j.ijinfomgt.2016.10.007 Chaston I Mangles T Business networks: Assisting knowledge management and competence acquisition within UK manufacturing firms Journal of Small Business and Enterprise Development 2000 7 2 160 170 10.1108/EUM0000000006837 Cooper D Schindler P Business research methods 2006 McGraw Hill-Irwin Daňa J Caputo F Ráček Y Complex network analysis for knowledge management and organizational intelligence Journal of the Knowledge Economy 2020 11 2 405 424 10.1007/s13132-018-0553-x Davenport T Klahr P Managing customer support knowledge California Management Review 1998 54 3 195 208 10.2307/41165950 Durst S Edvardsson I Knowledge management in SMEs: A literature review Journal of Knowledge Management 2012 16 6 879 903 10.1108/13673271211276173 Durst S Ferenhof HA Knowledge leakages and ways to reduce them in small and medium-sized enterprises (SMEs) Information 2014 5 1 440 450 10.3390/info5030440 Ebers M Jarrillo JC The construction, forms, and consequences of industry networks International Studies of Management and Organization 1998 27 4 3 21 10.1080/00208825.1997.11656716 Eisenhardt, K. (1989). Building theories from case study research. Academy of Management Review, 14(4), 532–550. 258557. Fallatah M Innovating in the desert: A network perspective on knowledge creation in developing countries Journal of the Knowledge Economy 2021 12 3 1533 1551 10.1007/s13132-021-00755-4 George A Bennett G Case studies and theory development in the social sciences 2005 MIT Press González-Loureiro M Dorrego PF Intellectual capital and system of innovation: What really matters at innovative SMEs Intangible Capital 2012 8 2 239 274 10.3926/ic.273 Grandori A Soda G Inter-firm networks: antecedents, mechanisms and forms Organization Studies 1995 16 2 183 214 10.1177/017084069501600201 Gupta V Rose L Jordão RVD Guest editorial - healthy organizations: Insights from the Latin American research Management Research 2019 17 2 118 126 10.1108/MRJIAM-06-2019-914 Hair J Black W Babin B Anderson R Multivariate data analysis 1998 Prentice Hall Henry L Intellectual capital in a recession: Evidence from UK SMEs Journal of Intellectual Capital 2013 14 1 84 101 10.1108/14691931311289039 Hilmersson FP Hilmersson M Networking to accelerate the pace of SME innovations Journal of Innovation & Knowledge 2021 6 1 43 49 10.1016/j.jik.2020.10.001 Hofstede, G. (1982). Culture’s consequences: comparing values, behaviors, institutions, and organizations. Sage Publications. Jardon C Martos M Intellectual capital as competitive advantage in emerging clusters in Latin America Journal of Intellectual Capital 2012 13 4 462 481 10.1108/14691931211276098 Jarillo JC On strategic networks Strategic Management Journal 1988 9 1 31 41 10.1002/smj.4250090104 Jick T Mixing qualitative and quantitative methods: Triangulation in action Administrative Science Quarterly 1979 24 1 602 610 10.2307/2392366 Jordão RVD Almeida VR Performance measurement, intellectual capital e financial sustainability Journal of Intellectual Capital 2017 18 3 101 132 10.1108/JIC-11-2016-0115 Jordão RVD Knowledge and information management practices in small and medium-sized enterprises organized in cooperative networks: A multi case comparative study in the Brazilian industry Information Science Perspectives 2015 20 3 178 199 10.1590/1981-5344/1737 Jordão RVD Novas J Knowledge management and intellectual capital in networks of small- and medium-sized enterprises Journal of Intellectual Capital 2017 18 3 667 692 10.1108/JIC-11-2016-0120 Jordão RVD Almeida VR Novas JC Intellectual capital, sustainable economic and financial performance and value creation in emerging markets: The case of Brazil The Bottom Line. 2022 35 1 1 22 10.1108/BL-11-2021-0103 Jordão RVD Novas J Gupta V The role of knowledge-based networks in the intellectual and organisational performance of small and medium-sized enterprises Kybernetes 2020 49 1 116 140 10.1108/K-04-2019-0301 Jordão RVD Souza A Avelar E Organisational culture and post-acquisition changes in management control systems: An analysis of a successful Brazilian case Journal of Business Research 2014 67 4 542 549 10.1016/j.jbusres.2013.11.011 Khalique M Bontis N Shaari J Isa A Intellectual capital in small and medium enterprises in Pakistan Journal of Intellectual Capital 2015 16 1 224 238 10.1108/JIC-01-2014-0014 Korbi F Chouki M Knowledge transfer in international asymmetric alliances: The key role of translation, artifacts, and proximity Journal of Knowledge Management 2017 21 5 1272 1291 10.1108/JKM-11-2016-0501 Laperche B Large firms’ knowledge capital and innovation networks Journal of the Knowledge Economy 2021 12 1 183 200 10.1007/s13132-016-0391-7 Leal-Millán A Roldán J Leal-Rodríguez A Ortega-Gutiérrez J IT and relationship learning in networks as drivers of green innovation and customer capital: Evidence from the automobile sector Journal of Knowledge Management 2016 20 3 444 464 10.1108/JKM-05-2015-0203 Lin Y Chen T How does strategic orientation influence intellectual capital through value creating activities? Business Research Review 2016 2 1 13 32 10.1177/1094670508314285 López-Sáez P Navas-López J Martín-de-Castro G Cruz-González J External knowledge acquisition processes in knowledge-intensive clusters Journal of Knowledge Management 2010 14 5 690 707 10.1108/13673271011074845 Mariotti F Knowledge mediation and overlapping in interfirm networks Journal of Knowledge Management 2011 15 6 875 889 10.1108/13673271111179262 Marques Júnior E Gobbo J Fukunaga F Cerchione R Centobell P Use of knowledge management systems: Analysis of the strategies of Brazilian small and medium enterprises Journal of Knowledge Management 2020 24 2 369 394 10.1108/JKM-06-2019-0334 Martínez-Costa M Jiménez-Jiménez D Rabeh H The effect of organisational learning on interorganisational collaborations in innovation: An empirical study in SMEs Knowledge Management Research & Practice 2019 17 2 137 150 10.1080/14778238.2018.1538601 Martins B Solé F Roles-purpose-and-culture misalignments: A setback to bottom-up SME clusters Journal of Knowledge Management 2013 17 4 598 616 10.1108/JKM-03-2013-0122 Mason C Castleman T Parker C Communities of enterprise: Developing regional SMEs in the knowledge economy Journal of Enterprise Information Management 2008 21 6 571 584 10.1108/17410390810911186 Massaro, M., Bardy, R., & Zanin, F. (2011). Innovation strategy and management control: The link between knowledge management and management control systems. In Proceedings of the 3rd European Conference on Intellectual Capital. University of Nicosia, Cyprus. Massaro M Pitts M Zanin F Bardy R Knowledge sharing, control mechanisms and intellectual liabilities in knowledge-intensive firms Electronic Journal of Knowledge Management 2014 12 2 117 127 Mertins, K., & Orth, R. (2011). Integrating intellectual capital and sustainability management: perspectives for the internal management and external reporting in small and medium sized enterprises. In: Proceedings of the 3rd European Conference on Intellectual Capital, Nicosia, Cyprus, 11–12 April 2011, pp. 527–536. Nicosia (Cyprus): University of Nicosia. Mertins K Will M Meyer C Analysing and enhancing IC in business networks: Results from a recent study Electronic Journal of Knowledge Management 2010 8 2 245 252 Montgomery CA Porter M Strategy: Seeking and securing competitive advantage 1991 Harvard Business School Press Moslehi, A., Linger, H., & Tanner, K. (2014). Diversity of knowledge in patent co-authorship networks – case studies in the Victorian biotechnology industry. VINE: The journal of information and knowledge management systems, 44(4), 496–518. 10.1108/VINE-05-2014-0032 Muhammad, Y., Abdul, M., Iftikhar, A., & Naila, T. (2011). Structuring intellectual capital as an element of virtual organisation in the SME clusters. In: 3rd International Conference on Advanced Management Science, IPEDR, Vol. 19, IACSIT Press, Singapore. Munier F Knowledge-based network: The key is the solution of dilemmas Journal of the Knowledge Economy 2021 12 1 279 292 10.1007/s13132-016-0420-6 Nonaka I Takeuchi H The knowledge-creating company 1995 Oxford University Press, Oxford Nonaka I Toyama R Konno N SECI, Ba and leadership: A unified model of dynamic knowledge creation Long Range Planning 2000 33 1 5 34 10.1016/S0024-6301(99)00115-6 Novas J Alves M Sousa A The role of management accounting systems in the development of intellectual capital Journal of Intellectual Capital 2017 18 2 286 315 10.1108/JIC-08-2016-0079 Organisation for Economic Co-operation and Development (OECD). (2010). A new OECD project: new sources of growth: intangible assets. Available at: http://www.oecd.org/dataoecd/60/40/46349020.pdf. (Accessed on February 8th 2021). Polanyi M The tacit dimension 1966 Doubleday and Co Powell W Learning from collaboration: Knowledge and networks in the biotechnology and pharmaceutical industries California Management Review 1998 40 3 228 240 10.2307/41165952 Pöyhönen A Smedlund A Assessing intellectual capital creation in regional clusters Journal of Intellectual Capital 2004 5 3 351 365 10.1108/14691930410550345 Prahalad C Hamel G The core competence of the corporation Harvard Business Review 1990 68 3 79 91 10.1007/978-3-662-41482-8_46 Rafique M Hameed S Agha M Impact of knowledge sharing, learning adaptability and organisational commitment on absorptive capacity in pharmaceutical firms based in Pakistan Journal of Knowledge Management 2018 22 1 44 56 10.1108/JKM-04-2017-0132 Richardson C Knowledge-sharing through social interaction in a policy-driven industrial cluster Journal of Entrepreneurship and Public Policy 2013 2 2 160 177 10.1108/JEPP-08-2011-0010 Romiti, A., & Sarti, D. (2011). Governance of networks of small enterprises: a knowledge perspective - some case studies in the mechanical industry in Italy. In: Proceedings of the 3rd European Conference on Intellectual Capital, University of Nicosia, Cyprus. Santos-Rodrigues, H., González-Loureiro, M., & Figueroa-Dorrego, P. (2012). System of innovation and innovative SMEs, a model for measuring the intellectual capital of SMEs. Paper presented at the 4th European Conference on Intellectual Capital, Helsinki, Finland. Suárez M Espacios transnacionales de conocimiento a través de la formaci_on de redes en nanotecnología Journal of Technology Management and Innovation 2013 8 304 310 10.4067/S0718-27242013000300055 United Nations. (2017). Micro e pequenas empresas são a maior fonte de emprego da América Latina e Caribe, afirma OIT. Available at: https://nacoesunidas.org/micro-e-pequenas-empresas-sao-a-maior-fonte-de-emprego-da-america-latina-e-caribe-afirma-oit/. (Accessed on February 8th 2021). Vale J Branco M Ribeiro J Individual intellectual capital versus collective intellectual capital in a meta-organisation Journal of Intellectual Capital 2016 17 2 279 297 10.1108/JIC-05-2015-0044 Valkokari K Helander N Knowledge management in different types of strategic SME networks Management Research News 2007 30 8 597 608 10.1108/01409170710773724 Verbano C Crema M Linking technology innovation strategy, intellectual capital and technology innovation performance in manufacturing SMEs Technology Analysis & Strategic Management 2016 28 5 524 540 10.1080/09537325.2015.1117066 Xu, X., Chen, H., & Zhang, R. (2020). The impact of intellectual capital efficiency on corporate sustainable growth-evidence from smart agriculture in China. Agriculture, 10(6), 199, 1–15. 10.3390/agriculture10060199 Zach F Hill T Network, knowledge and relationship impacts on innovation in tourism destinations Tourism Management 2017 62 1 196 207 10.1016/j.tourman.2017.04.001 Zakery A Saremi M Knowledge and intellectual capital in internationalizing SMEs, case study in technology-based health companies Journal of Intellectual Capital 2021 22 2 219 242 10.1108/JIC-02-2020-0048	0	PMC9709746	NO-CC CODE	2022-12-01 23:23:09	no	J Knowl Econ. 2022 Nov 30;:1-33	utf-8	null	null	null	oa_other
==== Front Environ Sci Pollut Res Int Environ Sci Pollut Res Int Environmental Science and Pollution Research International 0944-1344 1614-7499 Springer Berlin Heidelberg Berlin/Heidelberg 36449241 24276 10.1007/s11356-022-24276-y Research Article Reassessing the embodied carbon emissions in China’s foreign trade: a new perspective from the export routes based on the global value chain Zhang Boya [email protected] http://orcid.org/0000-0003-0342-7934 Ning Yadong [email protected] Bai Shukuan [email protected] grid.30055.33 0000 0000 9247 7930 Key Laboratory of Ocean Energy Utilization and Energy Conservation of the Ministry of Education, School of Energy and Power Engineering, Dalian University of Technology, Dalian, 116024 China Responsible Editor: Ilhan Ozturk 30 11 2022 122 16 6 2022 14 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. A new route perspective measure based on the global value chain (GVC) distinguishes the embodied emission transfer destinations while concurrently considering the transfer process and GVC embedding position. This study applies this measure to reassess the characteristics of China’s foreign trade embodied emissions and their impacts on global emissions. The results show that: (1) China’s domestic embodied emission exports are mainly concentrated in manufacturing and dominated by final goods exports and simple GVC routes. China primarily imports foreign emissions via simple GVC routes. (2) The embodied emission intensity of China’s exports is much higher than that of its imports, and China’s expansion in imports indirectly promoted global emission reduction. (3) China’s foreign trade increases global emissions with a waning trend, while GVC-related trade reduces global emissions. Additionally, it is feasible to reduce global emissions by adjusting China’s bilateral trade structures with different countries. We conclude that China’s GVC-related trade has increased, but exports through complex GVC tend to be the resource-input type. We emphasize that China needs to actively participate in globalization and upgrade its GVC to step off the low-end locking predicament in global production and cope with the multiple pressures of global and domestic emission reduction and stable development. Keywords Embodied carbon emission Global value chain International trade Emission transfer route China http://dx.doi.org/10.13039/501100001809 National Natural Science Foundation of China 71873021 Ning Yadong ==== Body pmcIntroduction Climate change is closely related to human survival and social development, and combating climate change is an important global issue. Massive and growing carbon emissions resulting from global economic growth and industrialization are the leading causes of climate change. The environment is a non-exclusive public good; thus, combating climate change requires all countries to participate in action (Liu and Zhao 2021), making carbon emissions an important international political and economic issue in recent decades. Some developed countries tend to import carbon-intensive products from countries with weak environmental regulations to reduce their domestic emissions (Antweiler et al. 2001; Copeland and Taylor 1994). International trade has become a reallocation mechanism for pollution emissions (Duan et al. 2021). When developing countries relatively fall behind in terms of productivity in carbon-intensive industries, global emissions may increase due to international trade, resulting in carbon leakage. Many studies on international trade and embodied emissions have demonstrated that emissions embodied in international trade (EEIT) can lead to carbon leakage (Lin and Sun 2010; Mongelli et al. 2006). Better management measures for EEIT are needed to ensure global emission reduction and combat climate change. The accurate measurement of EEIT is a fundamental prerequisite. However, with the rapid growth of globalization and the deepening of global labor division, intermediate trade gradually occupies a dominant position in international trade. In particular, some intermediate products may cross borders back and forth, thus making it difficult to calculate the relevant EEIT. Therefore, it is necessary to pay attention to both direct trade partners and all third parties involved in the entire trade process. Comprehensive identification of the initial sources and final absorptions helps depict EEIT transfer routes. Such analyses are more conducive to clarifying the transfer process of emissions and play a vital role in formulating effective emission reduction policies and management programs and reducing carbon leakage. Moreover, facing the irreversible trends of globalization, the growth of international trade is inevitable, and all countries and regions are committed to enhancing their position in the global production chain. Carbon emissions are closely related to production and value chains. In the context of global emission reduction, the improvement of the value chain should consider both the economic benefits and the impact on global emissions. A detailed analysis of EEIT transfer routes and the impacts of foreign trade on global emissions is conducive to identifying the direction of value chain upgrades for emission reduction. However, existing research on EEIT focuses solely on direct trade parties and ignores other participants in the whole trade process. For example, some have analyzed the EEIT at a national level, such as Brazil (Schaeffer and De Sá 1996), Australia (Lenzen 1998), Spain (Sánchez-Chóliz and Duarte 2004), Italy (Mongelli et al. 2006), and the UK (Wiedmann et al. 2010). Some studies have been conducted at the international level (Ahmad and Wyckoff 2003; Chen and Chen 2011; Ding et al. 2018; Peters and Hertwich 2008; Tian et al. 2015). These studies regarded the total emissions embodied in trade between two direct trade partners as the transfer of embodied emissions. However, they did not consider the complex sources and whereabouts of the trade process and ignored the possible problems of double-counting in the calculation, which may lead to misunderstandings of EEIT. Especially for China, which has large amounts of processing trade, this kind of problem may be more pronounced. Given the characteristics of China as a major country in the global economy, trading, and carbon emissions, the issue of China’s EEIT has attracted widespread attention. In addition, among existing studies on EEIT, research on China’s EEIT accounts for a large proportion, covering various types and methods of research on embodied emissions. China is the largest carbon emitter and foreign trading country. According to the WTO data, China’s foreign trade value reached 5.36 trillion USD in 2019 (WTO 2021), accounting for 10.72% of the total global trade value. The IEA data shows that China’s carbon emissions reached 9919.1 Mt (IEA 2021), accounting for 29.50% of global emissions. China’s emissions have been the focus of global emission reduction. From the perspective of emission reduction, internally, the Chinese government has put forward targets for carbon peaking and carbon neutrality; carbon emission reduction is an essential direction for China’s future development. Externally, China is facing tremendous pressure and a problematic situation of global emission reduction. Therefore, the measurement and assessment of China’s EEIT are essential for the emission reduction of both China and the world. In recent decades, especially after China acceded to the WTO, studying China’s EEIT has become a hot topic in related fields. Many studies have been conducted on China’s EEIT (Dietzenbacher et al. 2012; Lin and Sun 2010; Pan et al. 2008; Shui and Harriss 2006; Su and Ang 2014; Su and Thomson 2016; Weber et al. 2008). However, owing to the limitations of technology and methods, there are more or fewer deficiencies in early related studies. In addition, with the rapid development of international trade and the continuous improvement of the international division of labor, the global value chain (GVC) trade pattern characterized by the repeated cross borders of intermediate goods has come into being. The calculation of EEIT based on gross value statistics in previous studies could no longer accurately reflect the actual situation of EEIT transfer. With the advent of trade-in-value-added accounting and the development of GVC theory, it is possible to completely decompose the value-added in international trade according to sources, destinations, and transfer routes, which also provides new measures for EEIT. The calculation of GVC started with the study by Hummels et al. They first defined a narrow concept of vertical specialization and proposed a quantitative index of systematic measurement, which could measure GVC (Hummels et al. 2001). Since then, the methodology has been continuously developed based on aspects of the accounting framework (Koopman et al. 2008, 2010, 2014; Los et al. 2016; Los and Timmer 2018; Wang et al. 2014), as well as measurement indicators (Daudin et al. 2011; Dietzenbacher et al. 2005; Johnson and Noguera 2012; Stehrer 2012; Wang et al. 2017). Wang et al. compared the trade-in value-added accounting method with the gross value accounting system from gross exports and decomposed total exports into 16 value-added components and double-counting items, thus realizing the complete decomposition of gross exports (Wang et al. 2014). Today, the decomposition framework for GVC accounting has been completed. The gross exports are further decomposed into 17 components to consider the final destination of the exports better, and this application calculates carbon transfer more accurately (Fei et al. 2020; Zhang et al. 2021a; Bai et al. 2022). In recent years, some studies have traced the EEIT by combining the multi-region input–output (MRIO) model and value-added accounting. Zhao et al. investigated China’s EEIT based on the work of Hummels et al. They pointed out that the rise in China’s EEIT was mainly driven by the increase in re-exported emissions in China’s imported intermediate inputs (Hummels et al. 2001; Zhao et al. 2014). Xu et al. recalculated China’s EEIT based on the model proposed by Koopman et al. and found that traditional trade statistics overestimated China’s emissions in trade (Koopman et al. 2014; Xu et al. 2017). Meng et al. proposed a unified framework to trace a country’s embodied emissions in GVC (Meng et al. 2018). However, they did not conduct a detailed analysis of the EEIT. Dai et al. followed Los and Timmer’s framework and calculated the embodied emissions in China-US trade (Dai et al. 2021; Los and Timmer 2018). Xiong and Wu also analyzed the economic benefits and environmental costs of China-US trade following Wang’s framework (Xiong and Wu 2021). Both studies highlight the imbalance in bilateral trade between China and the USA. Li et al. weighed China’s EEIT and value-added and pointed out that GVC-related emissions exacerbate China’s economic-environmental imbalance (Li et al. 2022). Some studies have also examined the impact of the GVC location on carbon emissions. Zhang et al. adopted a regression analysis of export embodied emissions and the GVC participation index (calculated as the foreign component in gross exports) and concluded that production globalization makes China’s exports cleaner (Zhang et al. 2020). Liu and Zhao adopted the GVC participation (measured by production length) to represent the degree of GVC embedment and conducted a regression analysis on GVC participation and emission intensity (Liu and Zhao 2021). Chen et al. simultaneously estimated these two indicators’ impacts on China’s EEIT (Chen et al. 2022). These two indicators are the leading indicators for measuring the degree of GVC embedment. Nevertheless, the former focuses on foreign inputs in export production and ignores the impact of the embedding position. For example, when a country is engaged in the final assembly stage of the global production of machinery and equipment, its participation in GVC is high in terms of the foreign component of the exports, even though the country is downstream in the production chain. The GVC participation index considers the entire production process (including domestic and foreign stages). Wang et al. claimed that the greater the number of downstream production stages, the more upstream side of the production chain the input is (Wang et al. 2017). To focus on the foreign production stage after the export of products, that is, considering the exporter’s position in the GVC of the subsequent production chain, this study puts forward a new perspective on export routes to analyze embodied emissions exports. Summarizing the existing literature, the theories and methods of EEIT accounting have made significant progress in recent decades. However, with the development of the GVC theory and technology and the enrichment of databases related to global trade and carbon emissions, several considerations can be expanded in the following aspects. First, most previous studies adopted measures based on gross-value accounting and only focused on direct trade partners. Trade-in value-added accounting and the GVC-based measure can avoid the double-counting problem in the previous method and consider participants in the whole trade process. Second, previous studies mainly concentrated on the value of gross emissions in exports and imports and ignored the analysis of the process and transfer routes. The transfer routes of export embodied emissions are closely related to the location in the GVC, and the route-based study on EEIT reflects the impact of GVC on embodied emissions. Third, discussions about the impacts of GVC on EEIT in previous studies were mainly focused on the emitting country, and econometric analysis was primarily used to obtain qualitative results. Since carbon emission reduction is a global task, the impacts of EEIT and GVC on global emissions also need more attention. In addition, quantitative results are also significant for global emission reduction, and quantitative results are difficult to obtain using only econometric analysis. Given this, the main contributions of this paper are as follows. First, trade-in value-added accounting and the gross export decomposition framework were adopted in this paper. Taking China as an empirical case, we accurately identified the final destinations of China’s embodied emission exports and the sources of China’s embodied emission imports. Second, our methodology divided routes according to the border crossings of intermediate inputs in exports and separated out the exports based on the gross export decomposition framework. This measure can distinguish the embodied emission transfer destinations and concurrently consider the transfer process and GVC embedding position. From this route-based perspective, this study reassessed China’s EEIT, focusing on the domestic emissions transfer process after exporting. Third, the balance of avoided emissions (BAE) was adopted and improved in combination with our GVC-based route analysis to assess the impact of China’s foreign trade on global carbon emissions. The remainder of this paper is organized as follows: the “Methodology and data” section provides the methodology and data sources. The “Results and discussion” section presents the basic results of China’s EEIT characteristics and further discussions about the impacts of China’s EEIT on global emissions. The “Conclusion and policy implications” section presents the conclusions and policy implications. Methodology and data Methodology Embodied emissions are most commonly measured using multi-region input–output (MRIO). Appendix 1 provides the basic framework in Table 6 and calculations. A country’s exports can be completely decomposed into 16 components according to Wang et al.’s decomposition framework (Wang et al. 2014). This decomposition framework contributes significantly to GVC theory and the trade-in value-added accounting system. Following Wang’s decomposition framework, Fei et al., Zhang et al., and Bai et al. decomposed a country’s bilateral exports into 17 terms (Fei et al. 2020; Zhang et al. 2021a; Bai et al. 2022). Zhang et al. decomposed the energy use embodied in a country’s exports into 13 terms based on the export decomposition framework. This study follows these works and puts forward a new model to decompose a country’s bilateral export embodied emissions into eight items while concentrating on the domestic emissions embodied in exports and absorbed abroad. A schematic plot is provided in Fig. 1.Fig. 1 Contents of gross export embodied emissions and emission export routes As shown in Fig. 1, gross export embodied emissions consist of foreign and domestic emissions according to the emitting sources. All foreign emissions are absorbed by direct importers (FEMsr). Domestic emissions comprise two parts: emissions that are absorbed abroad (DEM1sr to DEM6sr) and emissions that return home (RDEMsr).1 Domestic emissions that are absorbed abroad (DEM) consist of six parts, according to the final destinations and transfer routes:DEM1: Domestic emissions embodied in final exports and absorbed by the direct importer.1 DEM1sr=FsBssT#Ysr where the superscripts refer to the exporter (s) and direct importer (r). Fs is the direct emission coefficient matrix of country s, and Bss is the submatrix of the global Leontief inverse matrix of country s. Superscript “T” refers to the transpose of matrix. “#” is defined as an element-wise matrix multiplication operation (When a matrix is multiplied by a column vector, each row of the matrix is multiplied by the corresponding row of the vector). Ysr is the final use of country r produced in country s.DEM2: Domestic emissions that are embodied in intermediate exports and are absorbed by direct importers.where Lss is the local Leontief inverse matrix of country s. Asr in the direct input coefficients matrix of country s to country r, representing the direct input of country s to produce a unit of final use of country r. Brr is the submatrix of the global Leontief inverse matrix of country r. Yrr is domestic final use of country r.2 DEM2sr=FsLssT#AsrBrrYrr DEM3: Domestic emissions embodied in intermediate exports to the direct importer are used to produce exports to the third party and are finally absorbed by the direct importer.where t represents the third party and “∑” refers to the sum of all third parties involved in the trade.3 DEM3sr=FsLssT#Asr∑r≠s,rGBrtYtr DEM4: Domestic emissions embodied in intermediate exports to direct importers to produce final exports to third parties.where u refers to the fourth party involved in the trade.4 DEM4sr-t=FsLssT#AsrBrrYrtt≠s,r 5 DEM4sr=∑t≠s,rGDEM4sr-t DEM5: Domestic emissions embodied in intermediate exports to direct importers to produce intermediate exports to third parties for their domestic final consumption production.where u refers to the fourth party involved in the trade.6 DEM5sr-t=FsLssT#AsrBrtYttt≠s,r 7 DEM5sr=∑t≠s,rGDEM5sr-t DEM6: Domestic emissions embodied in intermediate exports to direct importers to produce intermediate exports to third parties for their export production (except for upstream suppliers).where u refers to the fourth party involved in the trade.8 DEM6sr-t-u=FsLssT#Asr∑t≠s,rGBrtYtuu≠s,r,t 9 DEM6sr=∑u≠s,r,tGDEM6sr-t-u The gross value of domestic emissions that are finally absorbed abroad can be calculated as:10 DEMsr=DEM1sr+DEM2sr+DEM3sr⏟absorbedindirecttradepartner+DEM4sr+DEM5sr+DEM6sr⏟absorbedinindirecttradepartner According to Eq. (10), the gross DEM can be divided into two parts (absorbed in the direct trade partner, i.e., r in this model, and absorbed in indirect trade partners, i.e., t and u in this model). To distinguish the actual final destination and calculate accurate emission transfer, we calculated the embodied emissions absorbed by all indirect trading partners separately according to the final destination, referring to Eqs. (4), (6), and (8). This measure solves the ambiguity of the final destinations caused by taking the gross value of all indirect partners as one item in previous studies. So far, we have decomposed the total emissions embodied in a country’s bilateral exports (exports from s to r in the model). According to Eqs. (1)–(10), the gross emissions embodied in exports are initially emitted from different sources and finally absorbed in various destinations. Calculations based on traditional gross value accounting commonly generate the sum of FEM, RDEM, and DEM, which fail to distinguish the sources and destinations and thus cannot represent the real embodied emission export. Only DEM is the emission initially emitted domestically and is finally absorbed abroad, depicting the actual emission export. And Meng claimed that this index is the only emission trade measure consistently associated with bilateral gross trade flows (Meng et al. 2018). Given that GVC refers to the participation of two or more countries in production, the significance of GVC lies in the cross-border flows of intermediate products. Unlike the commonly used evaluation indicators of a country’s position in GVC (i.e., GVC participation index and production length), this study focuses on the foreign downstream production stages of exported goods. Therefore, we divided the DEM transfer process into five routes according to the final absorption place and the border-crossing number of intermediate products. When the number of border crossings of the intermediate products was zero, we defined the trade route as a “onefold value chain route.” When intermediates cross the border once, we defined the trade route as a “simple GVC route.” When the border crossing number is two or more, we defined the trade route as a “complex GVC route.” Specifications are listed in Table 1. As Wang et al. claimed, the greater the number of downstream production stages, the more upstream side of the production chain the input is (Wang et al. 2017). We can infer that the more complex GVC export routes (routes 3 and 5) mean that the exporter stands at a higher position in the global production chain.Table 1 Definition of routes and corresponding calculations Route Value chain type Final destination Calculation Intermediate cross-border times Specifics Route 1 Onefold value chain Direct DEM1 0 s→Fr Route 2 Simple GVC Direct DEM2 1 s→Ir Route 3 Complex GVC Direct DEM3 2 s→Ir→It→Fr Route 4 Simple GVC Indirect DEM4 1 s→Ir→Ft Route 5 Complex GVC Indirect DEM5 + DEM6 2 or more s→Ir→Its→Ir→It→F/Iu In the column of final destination, “Direct” and “Indirect” refer to direct trade partner (r in the model) and indirect trade partners (t and u in the model). “F” and “I” refer to final and intermediate exports In theory, the statistics in the opposite direction of exports are imports. The methodology of this study is based on the decomposition framework of a country’s total exports; thus, we can analyze the source and destinations of each sector’s total export embodied emissions in a country. However, for the importer, we can only obtain the results of the emissions embodied in the importer’s industry-wide imports from a specific sector of the exporters. Therefore, in this study, all the analyses on import embodied emissions were conducted at the national level and cannot be carried out at the sectoral level. Econometric analysis is commonly used to distinguish the relationship between trade and emissions when weighing the impacts of trade on emissions in previous studies. Zhang et al. and Ostic et al. used this method to investigate the impacts of international trade and foreign direct investments on carbon emissions in OPEC and belt and road countries and obtained reliable results (Zhang et al. 2021b; Ostic et al. 2022). However, the results obtained are qualitative. To obtain quantitative results, we adopted the balance of avoided emissions (BAE) to examine the impact of a country’s foreign trade on global emissions. According to López et al., in a 2-country model, BAE is calculated using the export embodied emissions minus emissions avoided by imports, as in the following equation (López et al., 2013a):11 BAE1-2=EEX-EAM=ε1E1+ε2E2-ε1M1+ε1M2=ε1E1-ε1M1⏟11.1+ε2M2-ε1M2⏟11.2 where BAE1−2 refers to the BAE induced by bilateral trade between the two countries, and ε1 and ε2 refer to each country’s direct emissions coefficients. E1, E2, M1, and M2 represent the exports and imports of the two countries, respectively. This model was subsequently extended to multi-region models and GVC analyses (López et al. 2013b; Zhang et al. 2017). This study gives the equations of BAE induced by the foreign trade between countries r and s based on the combination of previous studies and proceeding route-based decompositions, specifically as follows:12 BAEs-r=EEXsr+EEXrs-EAMsr+EAMrs=EEXsr-EAMsr⏟BAEs-rs+EEXrs-EAMrs⏟BAEs-rr where EEXsr and EEXrs are the DEMs of the two countries, respectively. EAMsr and EAMrs correspond to emissions avoided by the imports of these two countries. BAEs−r(s) is the BAE induced by foreign trade of country s, while BAEs−r(r) is the BAE caused by country r, and the details are as follows:12.1 BAEs-rs=EEXsr-EAMsr=FsBssT#Ysr-Yrs⏟Route1+FsLssT#AsrBrrYrr-ArsBssYss⏟Route2+FsLssT#Asr∑t≠s,rGBrtYtr-Ars∑t≠s,rGBstYst⏟Route3+FsLssT#AsrBrr∑t≠s,rGYrt-ArsBss∑t≠s,rGYst⏟Route4+FsLssT#Asr∑t≠s,rGBrtYtt-Ars∑t≠s,rGBstYtt+FsLssT#Asr∑t≠s,rG∑u≠s,r,tGBrtYtu-Ars∑t≠s,rG∑u≠s,r,tGBstYtu⏟Route5 12.2 BAEs-rr=EEXrs-EAMrs=FrBrrT#Yrs-Ysr⏟Route1+FrLrrT#ArsBssYss-AsrBrrYrr⏟Route2+FrLrrT#Ars∑t≠s,rGBstYts-Asr∑t≠s,rGBrtYtr⏟Route3+FrLrrT#ArsBss∑t≠s,rGYst-AsrBrr∑t≠s,rGYrt⏟Route4+FrLrrT#Ars∑t≠s,rGBstYtt-Asr∑t≠s,rGBrtYttFrLrrT#Ars∑t≠s,rG∑u≠s,r,tGBstYtu-Asr∑t≠s,rG∑u≠s,r,tGBrtYtu⏟Route5 As Eqs. (12.1) and (12.2) show, the BAE induced by each country is driven by the direct emission intensity (Fs and Fr), production technology matrix (Lss, Bss, Lrr, and Brr), and the trade imbalance between the two countries. In conclusion, the BAE induced by each country is affected by its domestic production technology and foreign trade structure. Data The MRIO tables were derived from the World Input–Output Database (Timmer et al. 2015). The world input–output tables released in 2016 cover 43 economies and the “rest of the world region” (RW) for the years 2000 to 2014 (specifics refer to Table 7 in Appendix 3). The corresponding carbon emission data were obtained from the Joint Research Center of the European Commission (Corsatea et al. 2019). This study decomposes China’s gross export embodied emissions to other countries and regions covered in the world input–output table for the period 2000–2014. To analyze the situation at the sector level, we merged the 56 sectors in the world input–output tables into eight industries according to the industry-intensive category, as shown in Table 2. The details are listed in Table 8 in Appendix 3.Table 2 Sector industry-intensive type and abbreviations Abbreviations Sectors Abbreviations Sectors PRS Primary and natural resources sectors CIS Capital-intensive service CIM Capital-intensive manufacturing LIS Labor-intensive service LIM Labor-intensive manufacturing KIS Knowledge-intensive service KIM Knowledge-intensive manufacturing HEPO Health/education/public service/other services Results and discussion Characters of China’s DEM at the sectoral level Figure 2 shows the gross value of China’s DEM (domestic emissions absorbed abroad) and its sector structure. During the study period, China’s total DEM increased from 662.16Mt in 2000 to 2088.21Mt in 2014, peaking at 2192.55Mt in 2007. The DEM dominates China’s gross export embodied emissions which account for over 90%, accompanied by a slight decline with some fluctuation (referring to Fig. 7 in Appendix 2). The PRS and HEPO shares were the smallest and declined continuously. This decline was mainly due to the restructuring of China’s export product mix, with the share of total exports declining in these two sectors. Manufacturing accounts for over 70% of the total DEM. KIM accounted for the most significant and growing share and nearly half of the total DEM in the late study period. The CIM ranked second and experienced a decrease-to-increase trend, and the LIM followed a fluctuating downward trend. The services (except for HEPO) had an ever-changing share of approximately 20%. CIS was the most significant, followed by LIS, and KIS accounted for only a small percentage. The share of each sector analyzed here is mainly determined by China’s export structure and the domestic carbon emission intensity of each sector’s products.Fig. 2 Sector structure of domestic emissions absorbed abroad. Notes: PRS in the figure refers to primary resource sectors. CIM, LIM, and KIM refer to capital-intensive, labor-intensive, and knowledge-intensive manufacturing sectors, respectively. CIS, LIS, and KIS refer to capital-intensive, labor-intensive, and knowledge-intensive service sectors, respectively. HEPO refers to the health, education, public, and other service sectors The dominant position of KIM is primarily due to China’s export structure since KIM includes the manufacturing of electrical equipment and products, machinery equipment and products, and chemical industries, all of which account for a large proportion of China’s exports. CIM mainly includes manufacturing food and tobacco products, paper products, coke and refined petroleum products, rubber and plastic products, non-metallic mineral products, basic metals, and fabricated metal products. These sectors only accounted for a small proportion of China’s exports, except for basic metal and metal products. Thus, the high rank of CIM was mainly driven by the high emission intensity of these sectors. Although LIM held a higher proportion of total exports than CIM, its share was lower than that of CIM, indicating that the domestic emission intensity of LIM is lower than that of CIM. This difference is determined by the total emission intensity of each sector and its domestic emission share. The total emission intensity is relevant to the product’s characteristics; for example, the light industrial products included in the LIM have low emission intensity. The sector’s domestic emission share is closely related to its position in the value chain, i.e., the installation of machinery and equipment sectors included in the LIM. This type of processing trade is located in the downstream stage of the value chain, with a small proportion of domestic emissions, causing the LIM sectors to contribute less to the overall DEM. The analysis thus far shows that GVC-based export embodied emission studies are different from previous gross value studies. One of the advantages is the accurate identification of the composition of total export embodied emissions. As mentioned in the introduction, the GVC-based research framework can also track the transfer routes of embodied emissions in detail, which is the focus of this study. We can analyze the route-based results from two aspects: on the one hand, the five routes in this study represent five specific transfer modes; on the other hand, the combination of different routes can represent the result with different emphasis. For example, according to the trade pattern, route 1 represents the final goods trade, and routes 2–5 represent the intermediate goods trade. In terms of absorption destinations, routes 1–3 represent absorption by direct trade partners, while routes 4–5 represent absorption by indirect trade partners. According to the complexity of the value chain of the production process, route 1 is pure domestic production; routes 2 and 4 are simple GVC routes; routes 3 and 5 are complex GVC routes. Figure 3 shows the export route structure of each sector’s DEM. As reported in the figure, the overall route structure of China appears to be relatively stable, with about half in route 1, followed by routes 2, 5, and 4. Route 3 ranked the last with a small portion. In general, China exports domestic emissions in the form of final goods, different from the global trade form, which is mainly driven by intermediate exports. However, at the sector level, the route structure of each sector varied greatly and changed over time. Overall, the ranking order of route structures of LIM, KIM, and HEPO (except 2005) was the same as that of the national level; route 1 accounted for a large part, followed by routes 2, 5, 4, and 3. However, the sequences of CIM, CIS, LIS, and KIS differed, with route 2 accounting for the most significant proportion, followed by routes 1, 5, 4, and 3. The route structure of the PRS changed the most during the study period. Since routes 1 and 2 accounted for the vast majority, the route structures of sectors were mainly shown in the difference between routes 1 and 2. This is primarily determined by the model of China’s export production. As for KIM and LIM, China mainly participates in their export production in the form of processing exports, so final exports dominate in these sectors. While in other sectors, China’s role is often as an initial resource provider. Thus, intermediate exports account for more in these sectors.Fig. 3 Export route structure of sectoral export embodied domestic emissions. Notes: Graph A, B, C, and D in the figure, respectively, refer to 2000, 2005, 2010, and 2014. Gross in each graph refers to the route structure of China’s total export embodied domestic emissions From the perspective of time-series changes, the leading structural transformation occurred between routes 1 and 2, and the proportion changes of other routes were relatively insignificant. Direct trade partners absorbed over 80% of the embodied domestic emissions for all sectors from route combinations. LIM and KIM exported more domestic emissions through final trade, with a slight decrease in the proportion. This was because China adopted extensive economic growth driven by processing exports and adjusted the export structure in the late study period. Simultaneously, other sectors preferred intermediate trade (a time-series result of the final trade share for each sector refers to Fig. 8 in Appendix 4). Except for LIM and KIM, which mainly produce exports purely domestically, most sectors export domestic embodied emissions through a simple GVC route. The complex GVC route accounted for approximately 10% of the total value. Characters of China’s embodied emission exports and imports at the route level All the above analyses at the sectoral level are the results of China’s export embodied emissions. We can simultaneously analyze the emissions embodied in exports and imports only at the national level. Figure 4 shows the route structures of China’s embodied domestic emissions exports and embodied foreign emission imports. The chart with the red line represents the net exports of embodied emissions; that is, the difference between domestic emissions embodied in exports absorbed abroad and foreign emissions embodied in imports absorbed domestically. The line chart shares the left coordinate axis with the accumulation chart of domestic embodied carbon export through each route. The significant imbalance between China’s domestic embodied emission exports and foreign embodied emission imports stands out in the figure. As we can see in the line chart, the net export value of embodied emissions (i.e., domestic emission exports minus foreign emissions imports) is consistent with the total domestic emissions exported by China. It showed a rapid upward trend before 2007, a rapid decline and a rapid rebound before and after the global financial crisis, and a slow decline after 2011. The total foreign emissions imports showed an unbroken and slow upward trend throughout the study period. The vast surplus between embodied emission exports and imports proves China’s role as a net exporter of embodied emissions. It is also in line with the fact that China exports high-carbon-intensive products and imports low-carbon-intensive products. The imbalance of the total value between embodied emission exports and imports is not the only focus; there are significant differences in the route structure between emission exports and imports. Route 1 accounted for a dominant position for embodied domestic emission exports, with a share of over 50% of the gross value (except in 2014, with a slight difference). This was mainly because of the large proportion of processing trade in China. Route 2 was ranked second, accounting for over 30%. Routes 5 and 4 held similar proportions, approximately 5–7%, and the former had a slightly higher share. Route 3 accounted for only a small proportion. As for foreign emission imports, route 2 dominated the route structure, with a share of approximately 60%. Route 1 accounted for approximately 20%, with a general trend from upward to downward. Route 4 accounted for approximately 5–6%, while route 5 accounted for twice.Fig. 4 Route-based domestic embodied emission export and foreign embodied emission import It should be noted that the concept of the net export of embodied emissions in this study is different from the net exports in previous studies. In previous studies, net export embodied emissions were usually based on gross value accounting and equaled the total value of export embodied emissions minus import embodied emissions. In this study, the net exports of embodied emissions emphasize their source and final consumption destinations. Embodied emissions can clearly be defined as exports or imports only when the source and final sink differ. It is not comprehensive to analyze China’s foreign trade embodied emissions only from the quantitative characteristics, and emission intensity is a beneficial index for measuring the relationship between economic benefits and carbon emissions. This study uses export embodied domestic emissions absorbed abroad and export embodied domestic value-added to calculate the domestic embodied emission intensity. This embodied emission intensity can be interpreted as the domestic emission required to create a unit of domestic value-added benefit through exports. Similarly, this study also calculated the import embodied foreign emission intensity. Figure 5 shows the results of the domestic embodied emission intensity based on the routes. Since sector-level analysis can only be applied to exports, we provide the results of China’s export embodied domestic emission intensity by sector, referring to Fig. 9 in Appendix 4.Fig. 5 Import and export embodied emission intensity of each route. Notes: the insert on the top right corner is the value of 2014, and its unit is also kg/dollar. Taking the export embodied domestic emission intensity of route 1 as an example, the value is 0.99 kg/dollar, indicating that China needs to emit 0.99 kg of domestic emissions to meet foreign consumption to create 1 dollar of domestic value-added through export in route 1 mode From the overall level, two results stand out in the figure: China’s export embodied domestic emission intensity was much higher than the import embodied foreign emission intensity; China’s export embodied domestic emission intensity dropped sharply. The significant difference between imports and exports indicates that China has effectively replaced some domestic production of products that may generate more emissions if produced domestically through imports. It follows that China’s aggressive expansion of imports in recent years has indirectly contributed to global emission reduction. For both exports and imports, the intensities of route 1 were the lowest among the five routes. The sequences of the other four routes differed between exports and imports. For export, the line charts in green and blue overlapped most of the time, indicating that the export embodied domestic emission intensities of routes 2 and 3 were almost the same during the study period. The intensity of route 5 was also practically the same as that of routes 2 and 3 after 2004; however, it was apparently higher than the intensity of route 4 throughout the study period. Route 2 was the highest for imports during the entire study period, followed by route 3. The intensity of route 5 was higher than that of route 4 before 2007, and the trend reversed after that. In theory, after the intermediate products have been exported, their subsequent production and consumption processes abroad do not affect the value of domestic carbon emissions and the value-added embodied in them. In other words, the differences among the domestic embodied emission intensities of routes 2 to 5 are mainly driven by the export structure of intermediate products in different export routes. Different export routes indicate intermediate goods’ positions in the production chain when they are exported. The more complex the GVC of the export route, the more subsequent production processes, and the closer the exporters are to the upstream of the production chain. Suppose the upstream production process belongs to R&D products with high value-added and low emissions. In that case, its embodied domestic emission intensity should be very low. However, it is evident from Fig. 5 that the embodied domestic emission intensity of China’s complex value chain exports, namely routes 3 and 5, is not the lowest among the four routes. However, the intensity of route 5 is significantly higher than that of route 4, indicating that the upstream intermediate products of China’s exports tend to be more resource-input with higher emissions than R&D products. The more complex the GVC of import routes, the lower the intensity, indicating that China’s imports have more R&D products. Further discussion on global impacts The empirical results provide the characteristics of China’s foreign trade embodied emissions based on GVC accounting and route analysis. This section further discusses the impact of China’s foreign trade on global emissions. Figure 6 shows the BAE results of China’s foreign trade through five routes. The figure quantifies the impact of China’s foreign trade through five routes on global emissions. The results show that China’s foreign trade consistently increased global emissions during the study period, and the impact peaked in 2007. From a route perspective, the impact of each route varied. Trade through route 1 increased the largest volume of global emissions, dominating the trend of the total impact, and the value of route 1 was larger than the gross value. Considering route 1 represents the final trade, while other routes represent intermediate trade, this result shows that China’s final trade increased global emissions, while intermediate trade reduced global emissions. Among the four intermediate trade routes, routes 2 and 4, which we defined as simple GVC routes, decreased the global emissions for most of the study period. In contrast, routes 3 and 5, complex GVC routes, increased global emissions all the time. The global emissions increased by routes 3 and 5 were less than those reduced by routes 2 and 4, i.e., GVC-related trade of China promoted global emission reduction. This result proves to some extent that globalization promotes global emission reduction.Fig. 6 Route-based impact of China’s foreign trade on global emissions. Notes: The bar graphs in the figure show the BAE of China’s trade through each route, indicating the impacts on global emissions, for an increase in positive value and a decrease in negative value. The red line chart in the figure shows the gross value of BAE in these five routes Table 3 shows the route-based BAE induced by China and its trading partners. As mentioned in the “Methodology” section, the BAE induced by China is primarily influenced by China’s domestic production technology and foreign trade structures. Generally, the BAE caused by each trade partner has an opposite sign (i.e., a positive or negative value) in bilateral trade because a country’s production technology (i.e., carbon emissions per unit of output) is usually positive. Therefore, the positive and negative situations of the BAE caused by each trade partner depend on the positive and negative status of the surplus of the country’s exports and imports. As shown in Table 3, China’s trade through routes 1, 3, and 5 increased global emissions, while routes 2 and 4 decreased global emissions. This result indicates that China had a trade surplus through routes 1, 3, and 5 and a deficit in routes 2 and 4. Combined with the results in Fig. 6, it can be seen that the positive and negative status of the total impact of China’s trade on global emissions is consistent with the positive and negative relationship of the BAE caused by China. This result indicates that under the condition that the values of the imbalance between exports and imports are equal and the plus-minus signs will be reversed. China’s domestic production intensity is higher than its import production intensity, and this information is consistent with Fig. 5.Table 3 Route-based BAE induced by China and its trade partners (unit: Mt) Route 1 Route 2 Route 3 Route 4 Route 5 Gross CHN TP CHN TP CHN TP CHN TP CHN TP CHN TP 2000 185.29 − 48.72 − 84.04 37.19 1.58 − 0.41 − 5.11 2.64 15.63 − 4.41 113.35 − 13.71 2001 175.77 − 50.25 − 80.55 36.93 1.56 − 0.44 − 3.68 2.06 15.64 − 4.90 108.73 − 16.60 2002 211.62 − 61.34 − 75.39 35.13 1.74 − 0.51 − 7.83 3.00 16.13 − 5.58 146.27 − 29.30 2003 279.36 − 74.87 − 85.56 40.77 1.93 − 0.53 − 27.63 8.93 11.92 − 3.83 180.00 − 29.54 2004 398.82 − 99.66 − 77.08 37.56 2.60 − 0.71 − 49.85 13.80 9.36 − 3.28 283.85 − 52.29 2005 578.47 − 129.50 − 50.14 29.86 3.16 − 0.81 − 72.31 19.09 7.86 − 3.06 467.04 − 84.42 2006 700.60 − 162.87 5.34 14.88 4.10 − 1.08 − 72.97 19.24 16.40 − 5.53 653.47 − 135.35 2007 840.75 − 214.48 − 2.25 9.45 4.76 − 1.34 − 79.23 20.65 16.50 − 7.09 780.53 − 192.81 2008 772.42 − 219.83 − 3.63 11.41 4.75 − 1.50 − 51.50 15.26 31.17 − 11.78 753.21 − 206.44 2009 616.35 − 188.44 − 116.64 47.85 3.10 − 1.03 − 36.80 11.85 25.66 − 9.41 491.67 − 139.17 2010 668.77 − 220.07 − 129.70 53.89 3.49 − 1.20 − 46.27 15.73 24.59 − 9.09 520.89 − 160.75 2011 643.89 − 222.43 − 183.06 75.82 3.98 − 1.41 − 34.14 12.68 37.03 − 13.01 467.69 − 148.36 2012 652.60 − 237.10 − 185.23 82.75 4.33 − 1.63 − 26.50 11.41 44.20 − 15.79 489.40 − 160.36 2013 652.25 − 249.66 − 209.58 93.40 4.24 − 1.60 − 20.76 9.60 43.62 − 16.07 469.76 − 164.33 2014 623.33 − 264.78 − 91.64 55.28 4.66 − 1.93 − 4.70 4.05 53.55 − 21.64 585.20 − 229.01 This table shows the BAE induced by China and its trade partner through each route. Route 1 CHN refers to the BAE generated by China’s trade through route 1, and Gross CHN is the total BAE of China through five routes. “TP” in the table means BAE of China’s trade partners It is generally believed that countries in the upper position of GVC tend to export products with high value-added and low emissions. However, China’s complex GVC (in routes 3 and 5) export embodied emission intensities were not significantly lower than other routes, as shown in Fig. 5, inconsistent with this inference. At the same time, China’s imports are broadly consistent with the inference. Given this, China’s move to enlarge its imports in recent years has positively impacted global emissions reduction. In terms of exports, China’s technological advances and optimizations in its foreign trade structure in recent years have allowed it to reduce global emissions through complex GVC in its bilateral trade with some countries. Table 4 shows the BAE results for China’s bilateral trade with other countries and regions from 2000 to 2014. As shown in the table, China’s bilateral trade with Australia, Brazil, and Korea has decreased in most years during the study period. In particular, China-Korea trade contributed to global emission reduction throughout the study period, with a cumulative reduction of 634.86Mt. Bilateral trade with Canada, the European Union, Mexico, and the USA had consistently increasing global emissions, with a cumulative increase of 4045.79Mt, accounting for over 85% of the total cumulative increase in global emissions caused by China’s foreign trade during the study period.Table 4 BAE induced by China’s bilateral trade with each trade partner (unit: Mt) AUS BRA CAN EU IND JPN KOR MEX RUS RW USA Gross 2000 − 0.11 − 0.11 5.34 27.72 − 0.21 16.24 − 19.17 3.35 5.76 − 17.00 77.84 99.64 2001 − 1.47 − 0.42 4.78 18.62 0.48 17.82 − 16.19 4.13 5.73 − 13.35 72.01 92.14 2002 − 0.59 − 1.08 8.23 18.57 0.05 10.53 − 20.81 6.24 5.66 − 8.12 98.30 116.97 2003 0.08 − 3.20 12.73 28.39 − 1.20 8.63 − 34.19 7.37 4.98 − 3.95 130.82 150.47 2004 2.71 − 1.83 18.59 50.16 − 0.32 9.55 − 51.88 12.12 3.04 15.65 173.78 231.56 2005 1.49 − 1.98 26.40 97.33 0.28 25.87 − 56.67 14.07 4.85 32.05 238.93 382.62 2006 2.11 0.13 30.80 125.09 3.47 15.07 − 49.60 21.65 5.52 104.63 259.26 518.12 2007 0.83 1.20 31.54 139.74 7.83 8.16 − 42.78 21.95 10.95 163.87 244.41 587.71 2008 − 2.81 5.12 27.93 137.58 4.27 3.68 − 37.56 20.21 10.01 181.37 196.97 546.76 2009 − 3.32 − 1.24 22.25 78.14 4.34 − 4.71 − 51.70 16.61 3.46 130.42 158.25 352.49 2010 − 10.26 2.67 24.87 93.59 4.31 − 14.30 − 55.16 21.49 5.79 114.52 172.64 360.14 2011 − 11.46 − 0.64 24.32 73.58 6.40 2.74 − 54.05 22.80 4.90 92.72 158.03 319.33 2012 − 5.93 0.77 24.50 53.24 4.46 20.49 − 50.76 21.36 4.97 96.11 159.81 329.03 2013 − 19.04 4.53 22.74 73.34 1.30 22.24 − 48.13 20.83 6.89 74.90 145.82 305.43 2014 − 9.57 7.37 24.09 51.41 − 0.29 20.92 − 46.22 21.14 3.23 136.10 148.00 356.19 However, there are also differences in the impact on global emissions among the routes within each bilateral trade. Table 5 presents the route-based results for China’s bilateral trade with each trade partner in 2014. The results suggest that even if bilateral trade increases global emissions overall, there may still be ways to reduce global emissions through some routes. For example, bilateral trade with Brazil and Japan through global value chain routes reduced global emissions despite the overall increase in global emissions. China-Australia bilateral trade had a decrease in global emissions, while its trade through route 1 (final goods trade) had the opposite impact when compared to other routes. Moreover, when taken as a whole, these bilateral trades fall into four categories (except for region RW in the table). Specifically, bilateral trade with the USA, Mexico, the European Union, and Canada increased global emissions through all routes, and these are the top four trade partners contributing to global emissions. Bilateral trade with Australia, Brazil, and Japan through route 1 increased global emissions and decreased emissions through other routes. Bilateral trade with India and Korea showed the opposite results. China-Russia bilateral trade differs from other bilateral trades, and only trade through route 5 could reduce global emissions. Overall, China’s export trade did not break away from the negative impact on global emissions reduction during the study period; however, these results demonstrate the feasibility of reducing global emissions by adjusting the structure of China’s bilateral trade with different countries.Table 5 Route-based BAE of China’s bilateral trade for each trade partner in 2014 (unit: Mt) AUS BRA CAN EU IND JPN KOR MEX RUS RW USA Route 1 10.73 9.39 12.76 25.92 − 1.10 35.67 − 15.59 8.64 2.89 163.80 105.44 Route 2 − 14.45 − 0.17 7.17 15.59 0.60 − 12.76 − 30.06 5.53 0.30 − 45.68 37.57 Route 3 − 0.07 − 0.03 0.02 0.25 0.00 − 0.09 − 0.12 0.01 0.00 2.14 0.61 Route 4 − 3.46 − 1.24 1.40 4.18 0.11 − 1.80 − 2.13 4.06 0.07 − 2.69 0.85 Route 5 − 2.32 − 0.58 2.74 5.47 0.10 − 0.11 1.68 2.91 − 0.03 18.53 3.53 Gross − 9.57 7.37 24.09 51.41 − 0.29 20.92 − 46.22 21.14 3.23 136.10 148.00 In previous studies, GVC participation, measured by the proportion of foreign components in total exports, was used to calculate the extent of China’s participation in the global division of labor. The results generally indicate that China’s position in the GVC was climbing. Their regression analyses of GVC participation and China’s per capita emissions or export embodied emissions concluded that China’s GVC participation reduces emissions (Wang et al. 2019; Zhang et al. 2020). Participation in the GVC assuredly enables a country to benefit from the technology spillover effect and improve domestic technology. Moreover, the fierce competition of the international division can, in turn, force domestic innovation and enhance the country’s production efficiency. However, participating in global production also stimulates growth. It drives more emissions, especially when a country lacks comparative advantages in high-end technology production stages. China participated in GVC in processing trade from the beginning, engaging in processing manufacturing and other low value-added production stages. At the same time, developed countries occupied the R&D stages, critical components, special materials, and high value-added production stages. This kind of GVC embedding may form the “low-end locking” in the value chain, which is not conducive to China’s position climbing in the global division of labor. In addition, from the perspective of emission reduction, “low-end locking” hinders the improvement of China’s emission reduction technology, harming global emission reduction. To assess the impact of China’s foreign trade on exports embodied emissions and even global emissions only from the perspective of participation has neglected the impact of China’s positions in the production process. According to the results, on the whole, China’s export production has not gotten rid of the “low-end locking” dilemma during the study period. The breadth of participation in GVC (namely, the GVC-related trade in total international trade) is increasing for China. However, the depth of GVC participation (the leading role in GVC) still needs to be enhanced. Enhancing the competitiveness of high-end industries, especially the export of technology-intensive services, can improve China’s position in the GVC and promote global emission reduction. In recent years, international trade patterns have changed frequently. The development of the Internet, the Internet of things, and the digital economy have stimulated the GVC to evolve constantly. The impact of COVID-19 on the global economy has accelerated the restructuring of the GVC. China must take an active part in globalization through the reconstruction of the GVC to cope with the multiple pressures of domestic and international emission reduction and stable economic development. Conclusion and policy implications Based on the GVC decomposition model of total exports, this study proposes a route division method based on the number of border-crossings of intermediate goods in total exports. The various routes can distinguish the complexity and length of the value chain of products after being exported. From the perspective of export routes, this study analyzes the characteristics of carbon emissions embodied in China’s foreign trade from 2000 to 2014. It further discusses the impact of China’s foreign trade on global emissions and the challenges China faces, and we finally draw the following conclusions.During the study period, China’s total export embodied domestic emissions experienced a trend of rapid rise–temporary decline–rebound–steady, and its share in gross export embodied emissions went slightly down at the end. From the sector-level perspective, embodied domestic emission exports are mainly sourced from manufacturing sectors. Capital-intensive and labor-intensive services are the primary sources in the service industry. The composition of export routes varied greatly among sectors. Still, the final goods trade (route 1) and simple GVC-related trade (route 2) occupied most. The route structure changes over time in each sector also mainly occurred between routes 1 and 2. Complex GVC routes (routes 3 and 5) accounted for a relatively small proportion and did not change significantly during the study period. Route 1 accounted for most of China’s domestic embodied emission exports. Route 2 accounted for a large proportion. Foreign embodied emissions imports were dominated by route 2. Route 1 was the lowest among the five routes in China’s imports and exports in terms of intensity. For the remaining four GVC routes, import trade generally conforms to the traditional perception that the more complex the value chain, the lower the embodied emission intensity. However, this perception did not match China’s exports since China’s complex GVC exports mainly tended to be a resource-input type. China’s expansion of imports during the late study period significantly contributed to reducing global emissions. China’s foreign trade consistently increased global emissions during the study period, and its impact peaked in 2007. Trade through route 1 increased the largest volume of global emissions, dominating the trend of the total impact. Complex GVC trade (routes 3 and 5) has also increased global emissions all the time. Simple GVC trade (routes 2 and 4) decreased global emissions for most of the study period. GVC-related trade generally decreased global emissions, proving that globalization can promote global emission reduction to some extent. The feasibility of reducing global emissions exists by adjusting the structure of China’s bilateral trade with different countries. China’s foreign trade structure, particularly its export trade, contributes to global emissions growth, albeit at a diminishing rate. It is urgent to change China’s “low-end locking” dilemma in GVC and its role as a resource provider in complex GVC exports. The impact of the new international trade pattern, digital economy, and COVID-19 has stimulated the restructuring of the GVC. Key insights arising from the findings of this study are that China should actively participate in globalization by upgrading its GVC and promoting its position in global production in response to the multiple pressures it faces (domestic carbon–neutral targets, international action on climate change, and stable economic growth). Implications that the policymakers should consider are discussed below. First, manufacturing and capital-intensive services have always been the critical sectors for China’s embodied carbon exports and will remain so for the foreseeable future. It is urgent to decrease the emission intensity of manufacturing from the perspective of technological advancement and energy structure and efficiency. In addition, improving the structure of export products and transforming China’s role as a resource provider in export trade are also the focuses of China’s international trade transformation. Second, as a major processing factor in global trade, China’s embodied carbon export routes are dominated by final goods trade and simple GVC trade. China embeds at the lower end of the GVC and urgently needs to strengthen the role of high-end industries in global production. Since the technological competitiveness gap between China and developed countries is gradually narrowing, China should strengthen basic scientific research and promote independent innovation to enhance overall technological competitiveness in globalization and promote its position in the low-end value chain of global production. Besides, in the long run, China should adhere to supply-side reform and continue to improve its domestic supply chain network. While promoting domestic industrial upgrading, China should also improve the investment environment of domestic enterprises and attract high-quality production factors to converge in China. Finally, China’s globalization can promote global emission reduction, but for China itself, domestic emissions increased. Combined with the new development pattern of “dual circulation” in the future, China should firmly embed itself in the global division of labor and simultaneously lead low-emission-intensive value chains. China should unite all countries to adhere to the in-depth development of globalization and fully use domestic and international markets, optimizing the allocation of global resources and each country’s comparative advantages to reduce domestic emissions while promoting global emission reduction. Appendix 1 Multi-region input–output (MRIO) is most commonly used to measure embodied emissions. A fundamental MRIO framework is provided in Table 6: Table 6 A basic framework of a multi-region input–output table Intermediate use Final use Gross output Country Sector Country s ⋯ Country G Country s ⋯ Country G 1 ⋯ N 1 ⋯ N Intermediate input Country s 1 Zijss ⋯ ZijsG Yiss ⋯ YisG Xis ⋮ N ⋮ ⋯ ⋱ ⋯ ⋮ ⋱ ⋮ ⋮ Country G 1 ZijGs ⋯ ZijGG YiGs ⋯ Yi XiG ⋮ N Value-added Vjs Gross Input Xjs The subscripts i and j in the table denote the sector number, and i, j = 1, …, N According to the basic input–output model, all gross output of country s must be used as either intermediate goods or final goods at home or abroad:13 Xs=AssXs+Yss+∑r≠sGAsrXr+∑r≠sGYsr where Xs refers to gross output of country s. Asr is the direct input coefficient matrix, and each of its elements (aij, where i and j refer to the sectors 1 … N) equals the corresponding intermediate consumption (zij) divided gross input (xj), aij = zij/ xj. Ysr denotes the final use in country r of goods produced in country s. Stated in matrix form as:14 Xs⋮XG=Ass⋯AsG⋮⋱⋮AGs⋯AGGXs⋮XG+Yss+∑r≠sGYsr⋮YGs+∑r≠sGYGr With rearrangement, we can obtain the following:15 Xs⋮XG=I-Ass⋯-AsG⋮⋱⋮-AGs⋯I-AGG-1Yss+∑r≠sGYsr⋮YGs+∑r≠sG-1YGr=Bss⋯BsG⋮⋱⋮BGs⋯BGGYs⋮YG where Bsr is the total requirements matrix, which gives the total requirement to produce a unit of gross output of country r needed from country s. Ys is the gross final goods produced in country s, composed by domestic use Yss and abroad use ∑r≠sGYsr. Direct emission coefficient matrix F=Fs⋯0⋮⋱⋮0⋯FG, where each submatrix is the diagonal matrix of a country’s direct emission coefficient and Fs=f1s⋯0⋮⋱⋮0⋯fNs. The elements of Fs are denoted by fjs=emjs/xjs, where fjs is the direct emission coefficient of sector j in country s, and emjs is the direct carbon emission of sector j in country s. Then, we can obtain the gross direct emission vector as16 EMs⋮EMG=Fs⋯0⋮⋱⋮0⋯FGBss⋯BsG⋮⋱⋮BGs⋯BGGYs⋮YG Appendix 2 According to the decomposed framework, we can calculate the foreign emissions embodied in exports from country s to country r as:17 FEMsr=FrBrsT#Ysr⏟F1+FrBrsT#AsrLrrYrr⏟F2+∑t≠s,rGFtBtsT#Ysr⏟F3+∑t≠s,rGFtBtsT#AsrLrrYrr⏟F4 where FEMsr is the total emissions embodied in the exports of country s to country r. Fr is the direct emission coefficient matrix of country r. Brs represents the submatrix of the global Leontief inverse matrix, referring to the total requirements from country r to produce a unit of total output in country s. Ysr is the final use of country r produced in country s. Asr is the direct input matrix of country s to country r, referring to the direct input of country s to produce a unit of the total output of country r. Lrr is the local Leontief inverse matrix of country r. Yrr is the domestic final use of country r. Ft is similar to Fr, and Bts is similar to Brs. The superscript T refers to the transpose of the matrix. In the equation, (F1) is the final export embodied emissions from country r (i.e., direct trade partner); (F2) is the intermediate export embodied emissions from country r; (F3) is the final export embodied emissions from all of country t (i.e., all indirect trade partners, consisting of the remaining countries and regions except for countries s and r); and (F4) is the intermediate export embodied emissions from all country t. Domestic emissions which finally return home can be expressed as follows:18 RDEMsr=FsLssT#AsrBrrYrs⏟RDEM1+FsLssT#AsrBrsYss⏟RDEM2+FsLssT#Asr∑t≠s,rGBrtYts⏟RDEM3 where RDEMsr is the gross domestic emissions embodied in exports and finally returns home. In the equation, (RDEM1) is the emission that returns home through final imports from country r; (RDEM2) is the emissions that return home through intermediate imports from country r; and (RDEM3) is the emissions that return home through imports from all other indirect partners. Figure Fig. 7 China’s gross export embodied emissions and composition. Note: DEM in this figure represents domestic emissions that are finally absorbed abroad, RDEM refers to domestic emissions that finally return home, and FEM is foreign emissions embodied in China’s gross exports 7 As this study emphasizes, not all of the carbon emissions embodied in China’s exports originate in China, nor are they all absorbed abroad. This study first distinguishes the sources and final destinations of all the carbon emissions embodied in China’s gross exports. The results are presented in Fig. 7. During the study period, China’s total export embodied emissions rose before the global financial crisis in 2008, especially after 2001, when China joined the WTO. After a brief dip in 2008 and 2009, gross value rebounded in the following 2 years to a slight increase in the pre-crisis level in 2011. This rate continued to decline slowly during the later study period. DEM dominated China’s export embodied emissions in terms of composition, accounting for over 90% of the total value, accompanied by a slight decline with fluctuation. FEM and RDEM accounted for a small proportion but generally showed a slow-growth trend during the study period. The rising trend of the RDEM and FEM is in line with the deepening global division of labor. The emergence and increase of RDEM indicate that an increasing number of products in China is outsourcing part of the production stages. The transformation process of the global production mode from outsourcing all production to part production stages is the process of production fragmentation and the formation of GVC. Under the irreversible trend of globalization, it is foreseeable that the proportions of RDEM and FEM will continue to increase in the future. Appendix 3 Tables Table 7 Countries and regions covered in this paper Country/region Acronym Country/region Acronym Country/region Acronym Australia AUS The European Union EU Mexico MEX Brazil BRA India IND Russia RUS Canada CAN Japan JPN The USA USA China CHN Korea KOR Rest of the countries and regions covered in WIOT RW 7 and Table 8 Sector industry-intensive type corresponds to WIOT sector number Sectors Corresponding sector number in WIOT Sectors Corresponding sector number in WIOT Primary and natural resources sectors (PRS) 1 ~ 4 Capital-intensive service (CIS) 24, 25, 31 ~ 35, 37 ~ 39, 44 Capital-intensive manufacturing (CIM) 5, 8 ~ 10, 13 ~ 16 Labor-intensive service (LIS) 27 ~ 30, 36, 55 Labor-intensive manufacturing (LIM) 6, 7, 22, 23 Knowledge-intensive service (KIS) 26, 40 ~ 43, 45 ~ 50 Knowledge-intensive manufacturing (KIM) 11, 12, 17 ~ 21 Health/education/public service/other services (HEPO) 51 ~ 54, 56 8 Appendix 4 Figures Fig. 8 Share of final trade embodied emissions. Note: “Gross” in the figure refers to the national level 8 and Fig. 9 Sectoral export embodied domestic emission intensity. Note: “Gross” in the figure refers to the national level 9 Acknowledgements We thank anonymous reviewers for their helpful comments and suggestions. Author contribution Boya Zhang: conceptualization, methodology, writing — original draft. Shukuan Bai: conceptualization, writing — review and editing. Yadong Ning: conceptualization, writing — review and editing, supervision, funding acquisition. Funding This study was supported by the National Natural Science Foundation of China (Grant number:71873021). Data availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. 1 This study focuses on the domestic emissions absorbed abroad. The calculations and results of foreign emissions embodied in exports (FEM) and emissions embodied in exports and finally returned home (RDEM) are provided in Appendix 2. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References Ahmad N, Wyckoff A (2003) Carbon dioxide emissions embodied in international trade of goods, OECD Science, Technology and Industry Working Papers, 2003/15, OECD Publishing, Paris. 10.1787/421482436815 Antweiler W Copeland BR Taylor MS Is free trade good for the environment? Am Econ Rev 2001 91 877 908 10.1257/aer.91.4.877 Bai SK Ning YD Zhang BY Estimating the environmental and employment impacts of China’s value-added trade from the perspective of value chain routes Environ Sci Pollut Res 2022 48 73414 73443 10.1007/s11356-022-20575-6 Chen ZM Chen GQ Embodied carbon dioxide emission at supra-national scale: a coalition analysis for G7, BRIC, and the rest of the world Energ Policy 2011 39 2899 2909 10.1016/j.enpol.2011.02.068 Chen HX Zhang CX Yin KD The impact of global value chain embedding on carbon emissions embodied in China’s exports Front Environ Sci 2022 10 950869 10.3389/fenvs.2022.950869 Copeland BR Taylor MS North-South trade and the environment Q J Econ 1994 109 755 787 10.2307/2118421 Corsatea T, Lindner S, Arto I, Roman M, Rueda Cantuche J, Velazquez Afonso A, De Amores Hernandez A, Neuwahl F (2019) World input-output database environmental accounts: update 2000-2016. Publications Office, 2019. https://data.europa.eu/doi/10.2760/024036. Accessed 26 Nov 2022 Dai F Yang J Guo H Sun H Tracing CO2 emissions in China-US trade: a global value chain perspective Sci Total Environ 2021 775 145701 10.1016/j.scitotenv.2021.145701 Daudin G Rifflart C Schweisguth D Who produces for whom in the world economy? Can J Econ 2011 44 1403 1437 10.1111/j.1540-5982.2011.01679.x Dietzenbacher E Romero Luna I Bosma NS Using average propagation lengths to identify production chains in the Andalusian economy Estud Econ Apl 2005 23 405 422 Dietzenbacher E Pei J Yang C Trade, production fragmentation, and China’s carbon dioxide emissions J Environ Econ Manag 2012 64 88 101 10.1016/j.jeem.2011.12.003 Ding T Ning Y Zhang Y The contribution of China’s bilateral trade to global carbon emissions in the context of globalization Struct Change Econ 2018 46 78 88 10.1016/j.strueco.2018.04.004 Duan Y Ji T Yu T Reassessing pollution haven effect in global value chains J Clean Prod 2021 284 124705 10.1016/j.jclepro.2020.124705 Fei R Pan A Wu X Xie Q How GVC division affects embodied carbon emissions in China’s exports? Environ Sci Pollut Res Int 2020 27 36605 36620 10.1007/s11356-020-09298-8 32564310 Hummels D Ishii J Yi K The nature and growth of vertical specialization in world trade J Int Econ 2001 54 75 96 10.1016/S0022-1996(00)00093-3 IEA (2021) Data and statistics. https://www.iea.org/data-and-statistics/databrowser?country=WORLD&fuel=CO2%20emissions&indicator=TotCO2 [Accessed November 20, 2021] Johnson RC Noguera G Accounting for intermediates: production sharing and trade in value added J Int Econ 2012 86 224 236 10.1016/j.jinteco.2011.10.003 Koopman R, Wang Z, Wei S (2008) How much of Chinese exports is really made in China? Assessing domestic value-added when processing trade is pervasive, NBER Working Paper Series. Working paper 14109. http://www.nber.org/papers/w14109. Accessed 26 Nov 2022 Koopman R, Powers W, Wang Z, Wei S (2010) Give credit where credit is due: tracing value added in global production chains, NBER Working Paper Series. Working paper 16426. http://www.nber.org/papers/w16426. Accessed 26 Nov 2022 Koopman R Wang Z Wei S Tracing value-added and double counting in gross exports Am Econ Rev 2014 104 459 494 10.1257/aer.104.2.459 Lenzen M Primary energy and greenhouse gases embodied in Australian final consumption: an input-output analysis Energ Policy 1998 26 495 506 10.1016/S0301-4215(98)00012-3 Li QP Wu SM Li ST Weighing China’s embodied CO2 emissions and value added under global value chains: trends, characteristics, and paths J Environ Manage 2022 316 115302 10.1016/j.jenvman.2022.115302 35597213 Lin B Sun C Evaluating carbon dioxide emissions in international trade of China Energ Policy 2010 38 613 621 10.1016/j.enpol.2009.10.014 Liu C Zhao G Can global value chain participation affect embodied carbon emission intensity? J Clean Prod 2021 287 125069 10.1016/j.jclepro.2020.125069 López LA Arce G Zafrilla JE Parcelling virtual carbon in the pollution haven hypothesis Energ Econ 2013 39 177 186 10.1016/j.eneco.2013.05.006 López LA, Arce G, Kronenberg T (2013b) Pollution haven hypothesis in emissions embodied in world trade: the relevance of global value chains. The Wealth of Nations in a Globalizing World, Workshop EU FP7 WIOD Project, July, 18th–19th, 2013b. https://www.rug.nl/ggdc/docs/session8_arce_paper.pdf. Accessed 26 Nov 2022 Los B, Timmer MP (2018) Measuring bilateral exports of value added: a unified framework. National Bureauof Economic Research Working Paper 24896. 10.3386/w24896 Los B Timmer MP de Vries GJ Tracing value-added and double counting in gross exports: comment Am Econ Rev 2016 106 1958 1966 10.1257/aer.20140883 Meng B Peters GP Wang Z Li M Tracing CO2 emissions in global value chains Energy Econ 2018 73 24 42 10.1016/j.eneco.2018.05.013 Mongelli I Tassielli G Notarnicola B Global warming agreements, international trade and energy/carbon embodiments: an input-output approach to the Italian case Energ Policy 2006 34 88 100 10.1016/j.enpol.2004.06.004 Ostic D Twnm AK Agyemang AO Boahen HA Assessing the impact of oil and gas trading, foreign direct investment infows, and economic growth on carbon emission for OPEC member countries Environ Sci Pollut Res 2022 29 43089 43101 10.1007/s11356-021-18156-0 Pan J Phillips J Chen Y China’s balance of emissions embodied in trade: approaches to measurement and allocating international responsibility Oxf Rev Econ Pol 2008 24 354 376 10.1093/oxrep/grn016 Peters GP Hertwich EG CO2 embodied in international trade with implications for global climate policy Environ Sci Technol 2008 42 1401 1407 10.1021/es072023k 18441780 Sánchez-Chóliz J Duarte R CO2 emissions embodied in international trade: evidence for Spain Energ Policy 2004 32 1999 2005 10.1016/S0301-4215(03)00199-X Schaeffer R De Sá AL The embodiment of carbon associated with Brazilian imports and exports Energ Convers Manag 1996 37 955 960 10.1016/0196-8904(95)00283-9 Shui B Harriss RC The role of CO2 embodiment in US-China trade Energ Policy 2006 34 4063 4068 10.1016/j.enpol.2005.09.010 Stehrer R (2012) Trade in value added and the value added in trade. WIIW Working Paper 81. Available at: https://wiiw.ac.at/trade-in-value-added-and-the-valued-added-in-trade-dlp-2620.pdf. Accessed 26 Nov 2022 Su B Ang BW Input-output analysis of CO2 emissions embodied in trade: a multi-region model for China Appl Energ 2014 114 377 384 10.1016/j.apenergy.2013.09.036 Su B Thomson E China’s carbon emissions embodied in (normal and processing) exports and their driving forces, 2006–2012 Energ Econ 2016 59 414 422 10.1016/j.eneco.2016.09.006 Tian J Liao H Wang C Spatial-temporal variations of embodied carbon emission in global trade flows: 41 economies and 35 sectors Nat Hazards 2015 78 1125 1144 10.1007/s11069-015-1761-3 Timmer MP Dietzenbacher E Los B Stehrer R de Vries GJ An illustrated user guide to the world input–output database: the case of global automotive production Rev Int Econ 2015 23 575 605 10.1111/roie.12178 Wang Z, Wei S, Zhu K (2014) Quantifying international production sharing at the bilateral and sector level. NBER Working paper 19677. 10.3386/w19677 Wang Z, Wei S, Yu X, Zhu K (2017) Characterizing global value chains: production length and upstreamness. NBER Working Paper Series. Working Paper 23261. http://www.nber.org/papers/w23261. Accessed 26 Nov 2022 Wang J Wan G Wang C Participation in GVCs and CO2 emissions Energ Econ 2019 84 104561 10.1016/j.eneco.2019.104561 Weber CL Peters GP Guan D Hubacek K The contribution of Chinese exports to climate change Energ Policy 2008 36 3572 3577 10.1016/j.enpol.2008.06.009 Wiedmann T Wood R Minx JC Lenzen M Guan D Harris R A carbon footprint time series of the UK-results from a multi-region input–output model Econ Syst Res 2010 22 19 42 10.1080/09535311003612591 WTO (2021) WTO data. https://www.wto.org/english/res_e/statis_e/statis_e.htm [Accessed November 20, 2021] Xiong Y Wu S Real economic benefits and environmental costs accounting of China-US trade J Environ Manage 2021 279 111390 10.1016/j.jenvman.2020.111390 33213992 Xu X Mu M Wang Q Recalculating CO2 emissions from the perspective of value-added trade: an input–output analysis of China’s trade data Energ Policy 2017 107 158 166 10.1016/j.enpol.2017.04.026 Zhang Z Zhu K Hewings GJD A multi-regional input-output analysis of the pollution haven hypothesis from the perspective of global production fragmentation Energ Econ 2017 64 13 23 10.1016/j.eneco.2017.03.007 Zhang Z Meng J Zheng H Zhu K Du H Guan D Production globalization makes China’s exports cleaner One Earth 2020 2 468 478 10.1016/j.oneear.2020.04.014 Zhang B, Bai S, Ning Y (2021a) Embodied energy in export flows along global value chain: a case study of China’s export trade. Front Energy Res 9:1–25. 10.3389/fenrg.2021.649163 Zhang JJ Twum AK Agyemang AO Edziah BK Ayamba EC Empirical study on the impact of international trade and foreign direct investment on carbon emission for belt and road countries Energ Report 2021 7 7591 7600 10.1016/j.egyr.2021.09.122 Zhao Y Zhang Z Wang S Wang S CO2 emissions embodied in China’s foreign trade: an investigation from the perspective of global vertical specialization China World Econ 2014 22 102 120 10.1111/j.1749-124X.2014.12077.x	36449241	PMC9709747	NO-CC CODE	2022-12-01 23:23:39	no	Environ Sci Pollut Res Int. 2022 Nov 30;:1-22	utf-8	Environ Sci Pollut Res Int	2,022	10.1007/s11356-022-24276-y	oa_other
==== Front Dtsch Z Akupunkt Deutsche Zeitschrift Fu¨r Akupunktur 0415-6412 1439-4359 Springer Medizin Heidelberg 522 10.1007/s42212-022-00522-9 Einführung zum Schwerpunktthema Bringt uns die COVID-19-Erkrankung mit ihren Folgen die Chance einer Neubewertung unseres medizinischen Denkens? Does COVID-19 and its consequences give us the opportunity to reassess our medical thinking?Heise Thomas [email protected] 12PD Dr. med. Dr. phil. Thomas Heise ist Kulturwissenschaftler wie auch praktischer Arzt, Psychiater sowie Psychotherapeut und arbeitet nach Chefarzttätigkeit in eigener Praxis mit einem ganzheitlichen Ansatz inklusive Naturheilkundeverfahren und Traditionelle Chinesische Medizin (TCM). Er ist Gründer und Herausgeber der Fachbuchreihe „Das transkulturelle Psychoforum“ mit 20 Bänden. Autor von "Chinas Medizin bei uns", "Qigong in der VR China: Entwicklung, Theorie und Praxis und "Kulturen der Menschheit: Woher und Wohin? Transdisziplinäre Perspektiven unserer Vergangenheit". Prakt. Arzt, Naturheilverfahren, Akupunktur; FA für Psychiatrie, Psychotherapie; Ed. „Das transkulturelle Psychoforum“; Sinologe; Vicechair WPA Section „Conflict Management & Resolution“. 1 grid.10423.34 0000 0000 9529 9877 Abteilung für Sozialpsychiatrie, Medizinische Hochschule Hannover, Hannover, Deutschland 2 Institute for Holistic Health Counselling, Zürcherstr. 21, 8245 Feuerthalen, Schweiz 30 11 2022 2022 65 4 221222 12 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. issue-copyright-statement© The Author(s), under exclusive licence to Springer Medizin Verlag GmbH, ein Teil von Springer Nature 2022 ==== Body pmcAls Erstes möchte ich die Redaktion loben, sich diesem wichtigen und brisanten Thema zu widmen. Brisant deswegen, weil man sich letztlich auch fragen muss, ob es eine Möglichkeit gegeben hätte, einen Teil dieser Fälle zu vermeiden. Und zweitens ist den Autoren zu gratulieren für ihre qualitativ hochwertigen Beiträge. Hätte es eine Möglichkeit gegeben, einen Teil dieser Fälle zu vermeiden? Da wäre als Erstes J. Dietzel, die sich nicht nur mit der unklaren Nomenklatur und Definitionsfrage auseinandersetzt, sondern den Finger in die Wunde legt, wenn sie auf ähnliche, nicht ausreichend gelöste oder verkannte, parallele Fragestellungen, beispielsweise auf das Chronic-Fatigue-Syndrom, verweist: Myalgische Enzephalomyelitis 1969/Chronisches-Fatigue-Syndrom 1988 (ME/CFS) sind nicht nur neuroimmunologische, sondern vielmehr Multisystemerkrankungen, deren Ursachen und Auslöser weiter und intensiver erforscht werden sollten. Dies umso mehr, als viele Fachgebiete davon betroffen sind und die volkswirtschaftlichen Kosten noch nicht gänzlich erfasst wurden. Eine Vielfalt meist unspezifischer Symptome prägt das Bild wie auch bei der Multiple-Chemikalien-Sensitivität (MCS) oder der elektromagnetischen Hypersensitivität (EHS; [1]). Bei Letzteren stellen neuere Zusammenfassungen bisheriger Forschungen [2] auch hinsichtlich einer kritiklosen Überelektromagnetisierung unserer Umwelt (5G)1 ebenso unangenehme Fragen wie bei der immer noch nicht geklärten Herkunft des Corona-Virus [3]. Dieses beinhaltete, z. B. bei künstlicher Herkunft, weitere Unvorhersehbarkeiten: neben nicht hinreichend geklärten Interaktionen der mRNA-Impfstoffe mit einem neuen Virus und seinen Mutationen schließlich auch die im vorliegenden Schwerpunktheft aufgegriffene Frage, wie wir die möglichen Langzeitfolgen am besten angehen. Langzeitfolgen, welche eventuell hätten vermieden werden können, wenn ein vorhandenes Behandlungsprotokoll2, ab dem 4.–5. Tag ohne klinische Besserung anzuwenden, besser verbreitet worden wäre. Dieses Protokoll sieht neben Ivermectin auch die Gabe von Vitaminen und Mineralien vor. Ivermectin ist ein Wurmmittel, welches in Onchozerchiasis-befallenen Ländern Afrikas mit entsprechender Dosierung dort millionenfach täglich Anwendung findet, mit der Folge wesentlich besserer Corona-Verläufe als in Nachbarländern ohne diese Krankheit und ohne die Anwendung dieses Mittels. Ein in Deutschland angewandtes Protokoll, entwickelt nach den Forschungen zu Ivermectin aus Afrika [4], ergänzt um immunstärkende Substanzen und Therapien. Darüber noch deutlich hinausgehend berichtet uns A. Beer seine wertvollen Erfahrungen aus dem gesamten naturheilkundlichen Spektrum. Der Artikel von U. Franke zur Neuraltherapie referiert noch einmal die möglichen immunologischen Abläufe und Symptome sowie deren Bearbeitung durch entsprechende Interventionen. Diese sollten nach Franke bezüglich Injektionsstellen und Techniken symptombezogen eingesetzt und vorher sorgfältig ausgetestet werden. C. Thede sieht in seiner Behandlung von Long- bzw. Post-COVID-Syndromen mit chinesischer Arzneimitteltherapie ebenso die schon erwähnten Ähnlichkeiten mit der ME/CFS-Symptomatik, aber auch mit Epstein-Barr-Virus-Infektionen, die auch vielfacettige und sehr protrahierte sowie schwere Verläufe nehmen können. Disharmonien der Shaoyang-Meridiane (3E, Gb) behandelt er dabei mit Variationen des „kleinen Bupleurum-Dekokts“ (Xiao Chaihu Tang). Erfolge erfährt er besonders bei respiratorischen Symptomen, den sensorisch-mentalen Defiziten wie auch der chronischen Müdigkeit. 3 Fallbeispiele illustrieren dies eindrücklich. C. Lazar lässt uns an dem ersten Fall von Long-COVID teilhaben, den sie mit Akupunktur angegangen ist. Wie so oft ergibt sich ein Durchbruch erst durch begleitende Änderungen im Verhalten. F. Lozano schließlich zeigt uns auf, wie auf dem Boden der chinesischen Medizin sowohl für die Akupunktur wie auch für die chinesische Arzneimitteltherapie differenzierte Behandlungsansätze verfolgt werden können. Post-Corona-Patienten mit und ohne Langzeitsymptomen klagen über Ängste sowie Depressionen In meiner Praxis für Psychotherapie werden allgemeinärztliche, naturheilkundliche und TCM-Ansätze (Akupunktur, Farblichtakupunktur mit dem Raymedy-System, Qigong und Taiji Quan) holistisch integriert. So kann ich aus eigener kasuistischer Erfahrung hinzufügen, dass ja Post-Corona-Patienten mit und ohne Langzeitsymptomen über Ängste sowie Depressionen klagen, die mit therapeutischem Einsatz von Qigong- sowie Taiji-Quan-Übungen gut zu verbessern sind, abgesehen davon, dass sie bekanntermaßen auch für das Immunsystem von Segen sind [5]. Vom naturheilkundlichen Aspekt her noch zu ergänzen wäre die sehr hilfreiche Gabe von Zink hochdosiert (60 mg) sowie Vitamin K2 und D3. Insgesamt bietet die chinesische Medizin in ihrer phänomenologischen Ausrichtung nun tatsächlich die Möglichkeit, einigen versäumten Aspekten nachzugehen, welche die klassifikatorische „Schubkasten“-Medizin nicht beantworten kann [6]. So können wir uns in eine holistische Richtung orientieren. Dafür müsste aber auch einigen unangenehmen Fragen nachgegangen werden. Interessenkonflikt T. Heise gibt an, dass kein Interessenkonflikt besteht. 1 Der 5G-Appell, inzwischen unterzeichnet von mehr als 200 Wissenschaftlern, kann hier heruntergeladen werden: https://www.diagnose-funk.org/publikationen/artikel/detail?newsid=1220. 2 www.flccc.net/covid-19-protocols. ==== Refs Literatur 1. Firstenberg A Die Welt unter Strom. Eine Geschichte der Elektrizität und ihrer übersehenen Gesundheitsgefahren 2021 Kandern Narayana, Unimedica 2. Heise T Unsere Gesundheit und das System 2022 Norderstedt BoD Im Druck 3. Heise T Corona: Das Syndrom. Über Grenzen, vergeben und schützen 2022 Norderstedt BoD 4. Guerrero R Bravo L Muñoz E Grillo Ardila E Guerrero E COVID-19: the Ivermectin African enigma Colomb Med (Cali) 2020 51 4 e-2014613 10.25100/cm.v51i4.4613 33795896 5. Heise T Qigong in der VR China: Entwicklung, Theorie und Praxis 1999 Berlin VWB 6. Schilling F Long-Covid & Post-Vac. Erkennen – Verstehen – Behandeln 2022 Tredition	0	PMC9709748	NO-CC CODE	2022-12-01 23:23:09	no	Dtsch Z Akupunkt. 2022 Nov 30; 65(4):221-222	utf-8	null	null	null	oa_other
==== Front Heart Vessels Heart Vessels Heart and Vessels 0910-8327 1615-2573 Springer Japan Tokyo 36449044 2203 10.1007/s00380-022-02203-y Short Communication Relationship between the spread of COVID-19, social frailty, and depressive symptoms in patients with heart failure Shakuta Saki [email protected] 1 Yamashita Masashi [email protected] 12 http://orcid.org/0000-0002-7706-9511 Kamiya Kentaro [email protected] 13 Hamazaki Nobuaki [email protected] 4 Ueno Kensuke [email protected] 1 Nozaki Kohei [email protected] 4 Uchida Shota [email protected] 15 Noda Takumi [email protected] 1 Maekawa Emi [email protected] 6 Yamaoka-Tojo Minako [email protected] 13 Matsunaga Atsuhiko [email protected] 13 Ako Junya [email protected] 6 1 grid.410786.c 0000 0000 9206 2938 Department of Rehabilitation Sciences, Graduate School of Medical Sciences, Kitasato University, Sagamihara, Japan 2 Division of Research, ARCE Inc., Sagamihara, Japan 3 grid.410786.c 0000 0000 9206 2938 Department of Rehabilitation, School of Allied Health Sciences, Kitasato University, Sagamihara, Japan 4 grid.508505.d 0000 0000 9274 2490 Department of Rehabilitation, Kitasato University Hospital, Sagamihara, Japan 5 grid.54432.34 0000 0001 0860 6072 Research Fellow of Japan Society for the Promotion of Science, Tokyo, Japan 6 grid.410786.c 0000 0000 9206 2938 Department of Cardiovascular Medicine, Kitasato University School of Medicine, Sagamihara, Japan 30 11 2022 15 31 5 2022 10 11 2022 © Springer Japan KK, 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. In community-dwelling older people, coronavirus disease 2019 (COVID-19) has been reported to be associated with the development of frailty and depressive symptoms. We aimed to investigate whether the spread of COVID-19 is associated with the development of frailty in patients with heart failure (HF). The presence of the multi-domain of frailty in 257 patients with HF was assessed at hospital discharge. The spread of COVID-19 was significantly associated with the development of social frailty and depressive symptoms. Evaluation of these symptoms during hospitalization would support disease management and understanding of their social and psychological conditions. Keywords Heart failure Coronavirus disease 2019 Social frailty Depressive symptoms http://dx.doi.org/10.13039/501100001691 Japan Society for the Promotion of Science 21H03309 20J10290 Yamashita Masashi Kamiya Kentaro ==== Body pmcIntroduction Coronavirus disease 2019 (COVID-19) is still endemic, and policies to prevent its spread affect people. Although the situation has improved, people have still refrained from going out unnecessarily and have maintained social distancing due to the spread of COVID-19. These new lifestyle approaches have affected people physically, psychologically, and socially [1]. Previous studies on community-dwelling older people have reported that decreased social participation due to the spread of COVID-19 was associated with physical inactivity and depressive symptoms [2, 3]. Patients with heart failure (HF) are more likely to have social frailty, physical frailty, cognitive impairment, and depressive symptoms [4, 5], and an overlap of these conditions leads to adverse events [6]. Therefore, multi-domain assessment and understanding of the conditions of patients with HF are important for disease management. The spread of COVID-19 is a predicted risk factor for these events, but its impact on patients with HF has not been comprehensively investigated thus far. Examining the adverse events associated with social policy toward the spread of COVID-19 to date may help in the development of strategies for the management of patients with HF in the event of a future flare-up of infection. Therefore, we investigated whether the spread of COVID-19 is associated with the development of the multi-domain of frailty in patients with HF. Methods Study population and procedures This study is a sub-analysis of a prospective study conducted at Kitasato University Hospital on patients hospitalized with cardiovascular diseases between September 2018 and December 2020. Except for the period from January 2020 to March 2020, which was determined to be a period of indeterminate exposure, we included patients who had independent activities of daily living before being hospitalized at Kitasato University Hospital with a diagnosis of HF between September 2018 and December 2020 and patients in whom data on multi-domain frailty were available. We identified patients with HF who presented with acute or worsening HF symptoms. HF diagnosis was performed by an experienced cardiologist based on the Framingham criteria. We excluded patients with a history of COVID-19, patients who were admitted to a nursing home, patients with severe HF (left ventricular ejection fraction [LVEF] of < 10%), and patients who received new treatments such as angiotensin receptor-neprilysin inhibitors, sodium-glucose cotransporter 2 inhibitors, and MitraClips. An overview of the comprehensive study protocol was in accordance with the tenets of the Declaration of Helsinki, approved by the Ethics Committee of Kitasato University Hospital (B18-083), and published in a publicly available University Hospital Information Network (UMIN000038373). Assessment of symptoms We assessed the presence of social frailty, physical frailty, cognitive impairment, and depressive symptoms in these patients at hospital discharge. Definitions Social frailty was defined as ≥ 2 positive responses to Makizako’s five items: living alone, going out less frequently compared with last year, visiting friends sometimes, feeling helpful to friends or family, and talking with someone every day [7]. Physical frailty was defined as ≥ 3 in the Fried phenotype model, which consists of five items: slowness (walking speed), weakness (grip strength), weight loss, fatigue, and physical inactivity [8]. Cognitive impairment was defined as ≤ 2 on Mini-Cog, a combination of a 3-item recall test and a clock drawing test [9]. Depressive symptoms were defined as ≥ 3 on the Patient Health Questionnaire-2, which includes questions about the frequency of feelings of apathy and hopelessness in the past 2 weeks [10]. Statistical analyses We divided the patients into two groups: before and after the spread of COVID-19. Patient characteristics were compared between the groups using the χ2 test or Mann-Whitney U test. Multivariate logistic regression analyses were used to examine the impact of the spread of COVID-19 on the development of the multi-domain of frailty in patients with HF. In the multivariate analyses, we selected variables that were considered to be related to the development of frailty and severity of HF in patients with HF. The variables used in the multivariate analyses were as follows: before and after the spread of COVID-19; age; sex; body mass index; LVEF; New York Heart Association functional classification at the time of hospitalization; blood and biochemical test indices at the time of hospitalization (log B-type natriuretic peptide, estimated glomerular filtration rate, and hemoglobin); and history of hypertension, diabetes, atrial fibrillation, chronic obstructive pulmonary disease, prior HF, or prior myocardial infarction. Statistical significance was considered at p < 0.05 and was analyzed using JMP (version 15.1; SAS Institute Inc., Cary, NC, USA). Results During the patient recruitment period, 272 patients with HF were hospitalized. After exclusion, 257 patients were enrolled in this study. The median patient age was 73 years, and 66.9% (172/257) were male. The prevalence of physical frailty, social frailty, cognitive impairment, and depressive symptoms in all cases were 26.5% (68/257), 65.4% (168/257), 20.2% (52/257), and 22.6% (58/257), respectively. There were no significant differences between before and after the spread of COVID-19 except for the presence of social frailty and depressive symptoms (Table 1). In multivariate logistic regression analyses, the spread of COVID-19 was significantly associated with the development of social frailty (odds ratio [OR], 2.52; 95% confidence interval [CI], 1.17–5.41) and depressive symptoms (OR, 2.23; 95% CI, 1.01–4.92) but not with the development of physical frailty and cognitive impairment (Table 2).Table 1 Patient characteristics Before the spread of COVID-19 (n = 182) After the spread of COVID-19 (n = 75) P value Age [years] 73 [62–80] 71 [58–79] 0.276 Male, n (%) 119 (65.4) 53 (70.7) 0.467 Body mass index [kg/m2] 23.5 [21.2–26.2] 24.0 [21.4–27.0] 0.550 Left ventricular ejection fraction [%] 41.0 [30.0–59.2] 41.0 [27.0–60.0] 0.655 NYHA classification ≥ III, n (%) 152 (94.4) 64 (95.5) 1.000 Medications, n (%) ACE inhibitor 78 (42.9) 34 (45.3) 0.782 ARB 80 (44.0) 30 (40.0) 0.582 Beta-blocker 155 (85.2) 59 (78.7) 0.204 MRA 103 (56.6) 36.0 (48.0) 0.218 Comorbidities, n (%) Hypertension 99 (54.4) 46 (61.3) 0.335 Diabetes mellitus 64 (35.2) 25 (33.3) 0.886 Dyslipidemia 52 (28.6) 24 (32.0) 0.652 Atrial fibrillation 80 (44.0) 28 (37.3) 0.404 COPD 14 (7.7) 2 (2.7) 0.163 Dementia 2 (1.1) 0 (0.0) 1.000 Prior heart failure 78 (42.9) 26 (34.7) 0.264 Prior myocardial infarction 18 (9.9) 10 (13.3) 0.509 Laboratory data B-type natriuretic peptide [pg/mL] 824.8 [444.3–1527.9] 747.2 [357.9–1323.4] 0.239 C-reactive protein [mg/dL] 0.5 [0.2–1.4] 0.3 [0.1–1.1] 0.063 eGFR [mL/min/1.73 m2] 43.0 [29.0–54.3] 43.0 [29.0–54.0] 0.902 Hemoglobin [g/dL] 12.4 [10.7–13.7] 12.8 [10.6–14.8] 0.243 White blood cells [× 103 μL] 6.9 [5.5–9.0] 7.3 [5.8–9.5] 0.450 Physical frailty, n (%) 44 (24.2) 24 (32.0) 0.215 Social frailty, n (%) 111 (61.0) 57 (76.0) 0.022 Cognitive impairment, n (%) 36 (19.8) 16 (21.3) 0.865 Depressive symptoms, n (%) 34 (18.7) 24 (32.0) 0.032 Values are presented as median [interquartile range] or number (%) ACE inhibitor angiotensin-converting enzyme inhibitor; ARB angiotensin II receptor blocker; COPD chronic obstructive pulmonary disease; COVID-19 coronavirus disease 2019; eGFR estimated glomerular filtration rate; MRA mineralocorticoid receptor antagonist; NYHA classification New York Heart Association classification Table 2 Multivariate logistic regression analyses of the association between the spread of coronavirus disease 2019 (COVID-19) and the multi-domain of frailty in patients with heart failure Odds ratio 95% confidence interval P value [Outcome: physical frailty] After the spread of COVID-19 (vs before) 1.72 0.79–3.76 0.175 [Outcome: social frailty] After the spread of COVID-19 (vs before) 2.52 1.17–5.41 0.018 [Outcome: cognitive impairment] After the spread of COVID-19 (vs before) 1.51 0.61–3.73 0.374 [Outcome: depressive symptoms] After the spread of COVID-19 (vs before) 2.23 1.01–4.92 0.047 Discussion The results of this study indicate that a series of policies, including refraining from going out, accompanied by the spread of COVID-19, were significantly associated with the development of social frailty and depressive symptoms in patients with HF. These findings provide important information for the support of patients with HF if disasters such as the spread of COVID-19 occur again in the future. Because social frailty and depressive symptoms in patients with HF are related to all-cause death, readmission for HF, and reduced quality of life, [5, 11] multi-domain assessment and understanding of the conditions of patients with HF are important for disease management [6]. To the best of our knowledge, only a few studies have examined the impact of the spread of COVID-19 in patients with HF. However, some previous studies conducted on community-dwelling older people have shown that the decline in social participation due to the spread of COVID-19 affects depressive symptoms and reduces the patient's physical activity and social isolation [1–3]. The results of the previous studies support our study's findings, which suggest that the spread of COVID-19 is a risk factor for the development of social frailty and depressive symptoms in patients with HF. Policies to prevent the spread of COVID-19 have limited the social participation of older people such as closing community organizations and restricting family visits [12]. Social isolation has been reported to be associated with the development of depressive symptoms among community-dwelling older people [13, 14]. It is possible that decreased social participation due to the spread of COVID-19 [12] is associated with the development of depressive symptoms in this study. Based on our results and the results of previous studies, it is important to evaluate the social frailty and depressive symptoms of patients with HF during the spread of COVID-19 from an early stage. To reduce loneliness, depressive symptoms, lack of social support, and physical inactivity in patients with HF, multifaceted interventions according to multidisciplinary guidelines and cooperation with relatives are necessary. Older patients with HF whose outpatient cardiac rehabilitation was interrupted due to the spread of COVID-19 have been reported to have worse frailty after the spread of COVID-19 compared to before [15]. Therefore, it remains essential to consider effective and safe intervention methods using online medical care and tele rehabilitation to maintain social distance as a preventive approach against the impact of the spread of infection. Further research focusing on interventions for social and psychological conditions would support the current study findings. This study had several limitations. First, due to the small number of patients in this study, we were unable to sufficiently verify whether there were any differences in patients with HF between preserved, mildly reduced, or reduced ejection fraction due to the impact of the spread of COVID-19. Therefore, further investigation is needed to determine whether there is an interaction between the association of COVID-19 and frailty incident in different types of HF. Secondly, since this was a cross-sectional study, we were unable to examine long-term cardiovascular outcomes. Therefore, long-term follow-ups of patients with HF are needed to investigate whether the spread of COVID-19 affects long-term cardiovascular outcomes. Conclusions The spread of COVID-19 is associated with the development of social frailty and depressive symptoms in patients with HF. Evaluation of social frailty and depressive symptoms during hospitalization would support disease management and understanding of the patient’s social and psychological conditions that are specific to the spread of COVID-19. Author contributions All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by SS, MY, KU, SU, and TN. The first draft of the manuscript was written by SS and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Funding This study was supported by a grant from the Japan Society for the Promotion of Science (JSPS) KAKENHI, Grant Numbers 21H03309 and 20J10290. Data availability The data underlying this article cannot be shared publicly due to the privacy of individuals that participated in the study. The data will be shared on reasonable request with the corresponding author. Declarations Conflict of interest The authors declare that there is no conflict of interest. Ethical approval This study was performed in accordance with the tenets of the Declaration of Helsinki and was approved by the Ethics Committee of Kitasato University Hospital (B18-083) and published in a publicly available University Hospital Information Network (UMIN000038373). Consent to participate Informed consent was obtained from all individual participants included in the study. Consent to publish The participant has consented to the submission of the paper to the journal. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References 1. Sepúlveda-Loyola W Rodríguez-Sánchez I Pérez-Rodríguez P Ganz F Torralba R Oliveira DV Rodríguez-Mañas L Impact of social isolation due to COVID-19 on health in older people: mental and physical effects and recommendations J Nutr Health Aging 2020 24 9 938 947 10.1007/s12603-020-1500-7 33155618 2. Ettman CK Abdalla SM Cohen GH Sampson L Vivier PM Galea S Prevalence of depression symptoms in US adults before and during the COVID-19 pandemic JAMA Netw Open 2020 3 9 e2019686 10.1001/jamanetworkopen.2020.19686 32876685 3. Yamada M Kimura Y Ishiyama D Otobe Y Suzuki M Koyama S Kikuchi T Kusumi H Arai H The influence of the COVID-19 pandemic on physical activity and new incidence of frailty among initially non-frail older adults in Japan: a follow-up online survey J Nutr Health Aging 2021 25 6 751 756 10.1007/s12603-021-1634-2 34179929 4. Gorodeski EZ Goyal P Hummel SL Krishnaswami A Goodlin SJ Hart LL Forman DE Wenger NK Kirkpatrick JN Alexander KP Geriatric Cardiology Section Leadership Council, ACoC Domain management approach to heart failure in the geriatric patient: present and future J Am Coll Cardiol 2018 71 17 1921 1936 10.1016/j.jacc.2018.02.059 29699619 5. Jujo K Kagiyama N Saito K Kamiya K Saito H Ogasahara Y Maekawa E Konishi M Kitai T Iwata K Wada H Kasai T Nagamatsu H Ozawa T Izawa K Yamamoto S Aizawa N Yonezawa R Oka K Makizako H Momomura SI Matsue Y Impact of social frailty in hospitalized elderly patients with heart failure: a FRAGILE-HF registry subanalysis J Am Heart Assoc 2021 10 17 e019954 10.1161/JAHA.120.019954 34472374 6. Matsue Y Kamiya K Saito H Saito K Ogasahara Y Maekawa E Konishi M Kitai T Iwata K Jujo K Wada H Kasai T Nagamatsu H Ozawa T Izawa K Yamamoto S Aizawa N Yonezawa R Oka K Momomura SI Kagiyama N Prevalence and prognostic impact of the coexistence of multiple frailty domains in elderly patients with heart failure: the FRAGILE-HF cohort study Eur J Heart Fail 2020 22 11 2112 2119 10.1002/ejhf.1926 32500539 7. Makizako H Shimada H Tsutsumimoto K Lee S Doi T Nakakubo S Hotta R Suzuki T Social frailty in community-dwelling older adults as a risk factor for disability J Am Med Dir Assoc 2015 16 11 1003.e7 11 10.1016/j.jamda.2015.08.023 26482055 8. Fried LP Tangen CM Walston J Newman AB Hirsch C Gottdiener J Seeman T Tracy R Kop WJ Burke G McBurnie MA Frailty in older adults: evidence for a phenotype J Gerontol A Biol Sci Med Sci 2001 56 3 M146 156 10.1093/gerona/56.3.M146 11253156 9. Saito H Yamashita M Endo Y Mizukami A Yoshioka K Hashimoto T Koseki S Shimode Y Kitai T Maekawa E Kasai T Kamiya K Matsue Y Cognitive impairment measured by Mini-Cog provides additive prognostic information in elderly patients with heart failure J Cardiol 2020 76 4 350 356 10.1016/j.jjcc.2020.06.016 32624300 10. Kroenke K Spitzer RL Williams JB The patient health questionnaire-2: validity of a two-item depression screener Med Care 2003 41 11 1284 1292 10.1097/01.MLR.0000093487.78664.3C 14583691 11. Sbolli M Fiuzat M Cani D O'Connor CM Depression and heart failure: the lonely comorbidity Eur J Heart Fail 2020 22 11 2007 2017 10.1002/ejhf.1865 32468714 12. Cudjoe TKM Kotwal AA “Social distancing” amid a crisis in social isolation and loneliness J Am Geriatr Soc 2020 68 6 e27 e29 10.1111/jgs.16527 32359072 13. Sakurai R Kawai H Suzuki H Kim H Watanabe Y Hirano H Ihara K Obuchi S Fujiwara Y Poor social network, not living alone, is associated with incidence of adverse health outcomes in older adults J Am Med Dir Assoc 2019 20 11 1438 2144 10.1016/j.jamda.2019.02.021 31000349 14. Tsutsumimoto K Doi T Makizako H Hotta R Nakakubo S Kim M Kurita S Suzuki T Shimada H Social frailty has a stronger impact on the onset of depressive symptoms than physical frailty or cognitive impairment: a 4 year follow-up longitudinal cohort study J Am Med Dir Assoc 2018 19 6 504 510 10.1016/j.jamda.2018.02.008 29703687 15. Kato M Ono S Seko H Tsukamoto T Kurita Y Kubo A Omote T Omote S Trajectories of frailty, physical function, and physical activity levels in elderly patients with heart failure: impacts of interruption and resumption of outpatient cardiac rehabilitation due to COVID-19 Int J Rehabil Res 2021 44 3 200 204 10.1097/MRR.0000000000000473 34034289	36449044	PMC9709749	NO-CC CODE	2022-12-01 23:23:39	no	Heart Vessels. 2022 Nov 30;:1-5	utf-8	Heart Vessels	2,022	10.1007/s00380-022-02203-y	oa_other
==== Front Pediatr Nephrol Pediatr Nephrol Pediatric Nephrology (Berlin, Germany) 0931-041X 1432-198X Springer Berlin Heidelberg Berlin/Heidelberg 36449100 5828 10.1007/s00467-022-05828-3 Original Article Outcomes and perception of cloud-based remote patient monitoring in children receiving automated peritoneal dialysis: a prospective study http://orcid.org/0000-0002-1463-6005 Chan Eugene Yu-hin [email protected] Liu Mei-shan Or Po-chu Ma Alison Lap-tak Paediatric Nephrology Centre, Hong Kong Children’s Hospital, Kowloon City, Hong Kong SAR 30 11 2022 18 19 9 2022 25 10 2022 14 11 2022 © The Author(s), under exclusive licence to International Pediatric Nephrology Association 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. Background Remote patient monitoring (RPM) for automated peritoneal dialysis (APD) may improve clinical outcomes. Paediatric data, however, remain extremely scarce. Methods We conducted a prospective observational study of children (0–18 years) receiving APD with cloud-based RPM over two 24-week periods (pre- and post-RPM). Primary outcomes were unplanned hospitalizations and fluid management. Children receiving APD without RPM (non-RPM) were included as control. Results Seven patients (6 females) receiving APD were enrolled in the RPM programme at 11.3 years (IQR 2.6–17.1). Main indications for RPM included history of fluid overload (n = 3) and non-adherence (n = 2). Ten children were included in the non-RPM group (6 females; 16.9 years, IQR 12.8–17.6). Four patients (57.1%, 95% CI 22.5–100%) experienced fewer unplanned hospitalizations and 5 patients (71.4%, 95% CI 34.1–100%) had shorter hospital stays during the post-RPM period. The hospitalization rates and length of stay were reduced by 45% and 42%, respectively. The higher hospitalization rates among the RPM group, compared to the non-RPM group, were no longer observed following implementation of RPM. There was a significant increase in ultrafiltration (565.6 ± 248.7 vs. 501.7 ± 286.6 ml/day, p = 0.03) and reduction in systolic blood pressure (114.1 ± 12.6 vs. 119.9 ± 11.19 mmHg, p = 0.02) during the post-RPM period. All patients demonstrated satisfactory adherence. Although quality of life (PedsQL 3.0 ESRD module) was not different pre- and post-RPM, all patients agreed in the questionnaires that the use of RPM improved their quality of life and sense of security. Conclusions In conclusion, RPM in children receiving APD is associated with fewer and shorter unplanned hospitalizations, improved fluid management and favourable adherence to PD. Graphical abstract A higher resolution version of the Graphical abstract is available as Supplementary information Supplementary information The online version contains supplementary material available at 10.1007/s00467-022-05828-3. Keywords Claria Remote patient monitoring Peritoneal dialysis Children ==== Body pmcIntroduction Peritoneal dialysis (PD) is an important modality of kidney replacement therapy, especially among young children [1–3]. Hong Kong adopts a PD-first policy and 70% of children and young people are offered automated PD (APD) as initial modality [2]. While PD as a home therapy facilitates normal schooling, it is associated with hospitalizations due to mechanical or infectious complications [4]. There are also concerns with non-adherence, as well as stress and poor quality of life for patients and their families [5, 6]. APD systems equipped with cloud-based, remote patient monitoring (RPM) have been developed to enhance patient-centred care. This system allows active monitoring of PD-related parameters, such as blood pressure and ultrafiltration, and remote adjustment of PD prescription. Limited trials involving adult patients generally showed improved patient satisfaction, but the benefits in clinical outcomes and reduction of medical visits remained controversial [7, 8]. In the only available paediatric series, Bakkaloglu et al. focused on the perceptions of RPM from families and health care providers [9]. The aim of this study was to provide prospective data on important clinical outcome measures, such as hospitalization rates and blood pressure control, and evaluate patient quality of life among children receiving APD with RPM. Methods We conducted a prospective observational study of children (0–18 years) receiving APD, who were enrolled in a pilot RPM programme, at the Paediatric Nephrology Centre, Hong Kong Children’s Hospital, Hong Kong. The Paediatric Nephrology Centre is the territory-wide, designated site providing chronic kidney replacement therapy (KRT) in Hong Kong. Due to limited APD machines with cloud-based RPM function available, patients were recruited into the pilot RPM programme according to specific indications such as history of fluid overload and treatment non-adherence. The programme was led by a dedicated team comprised of a paediatric nephrologist and two dialysis nurses, and was launched on 1 September 2021. The study was conducted over two 24-week periods, before (March 2021 to August 2021) and after (September 2021 to February 2022) implementation of the RPM programme (pre- and post-RPM). Patients who received standard APD without RPM (non-RPM group) during the same study period were also included as control. All patients who did not complete the observation periods were excluded. Data pertaining to clinical and PD-related parameters, alarms, details of RPM and interventions, and quality of life of patients and caregivers were collected and evaluated. Questionnaires were given to all caregivers or adolescent patients (> 12 years) who received PD following the completion of the 6-month post-RPM period (Table 4). In addition, a questionnaire on perception of the RPM programme was distributed to the dialysis team 1 year after programme implementation (Table 5). Homechoice Claria® cycler equipped with the cloud-based Sharesource® platform (Baxter Healthcare, Deerfield, IL) was used in all patients. Through the Sharesource® platform, anonymized data (e.g. ultrafiltration volume, dwell time, alarms) and self-entered information (e.g. body weight and blood pressure) were accessible from the hospital. PD prescription could also be changed remotely. Dry weight and target blood pressure were set according to clinical and bioimpedance evaluation. Flag rules were set as follows (yellow flag and red flag, respectively): lost dwell time (45 min and 75 min), lost therapy volume (none and 10%), initial drain variance (none and 50%), adjusted peritoneal volume (none and 2) and events during treatment (5 and 10). Patient data were screened by the dialysis nurse at least twice a week. Detailed reviews were performed in patients with abnormal findings. Alternatively, patients might contact the team if they experienced problems. A case conference with the nephrologist was held once weekly during the first 2 weeks, bi-weekly during the third to twelfth week and monthly thereafter following programme launch. Dialysis interventions were communicated through phone calls and text messages by kidney nurses and the PD programme was adjusted online directly by the nephrologist. Primary outcomes were unplanned hospitalization rates related to PD and fluid management, including daily ultrafiltration volume and blood pressure control. Secondary outcomes included the frequency of clinic visits, reviews and interventions performed through RPM, PD-related anthropometric and laboratory parameters, treatment compliance, and quality of life, which was assessed by the PedsQL 3.0 ESRD module pre- and post-RPM [10]. The study was performed according to the Declaration of Helsinki and was approved by the institutional review board of the Hong Kong Children’s Hospital, Hospital Authority, Hong Kong (HKCH-REC-2020–006). IBM SPSS statistics version 26 software was used for all statistical analysis. Data were expressed in mean, standard deviation, median, number, percentage and 95% confidence intervals (95% CI) when appropriate. The mean difference of the continuous variables was analysed by paired t-test. A p-value less than 0.05 in two tails was treated as significant. Results At the beginning of the pre-RPM observation period, 8 out of 26 patients receiving APD were enrolled in the RPM programme. One patient received a kidney transplant 4 days after programme initiation and was excluded from the analysis. Seven patients (6 females; 6 Chinese and 1 Pakistani) were included. The median age at initiation of KRT and RPM enrolment was 9.7 years (IQR 2.3–14.9) and 11.3 years (IQR 2.6–17.1), respectively. The time on dialysis at programme enrolment was 17.2 months (IQR 7.6–27.2). Indications for RPM were history of significant fluid overload (n = 3), non-adherence (n = 2), repeated unplanned hospitalizations (n = 1) and social reason due to language barrier (n = 1). Ten patients who underwent APD without RPM (non-RPM group) during the pre- and post-RPM periods were included for analysis. Patients in the non-RPM group were older in age, on dialysis for a longer duration and had a lower prevalence of development delay. Details of the baseline demographics are presented in Table 1.Table 1 Baseline demographics for children receiving automated peritoneal dialysis with or without remote patient monitoring during the study period (March 2021 to February 2022) All (n = 17)a RPM (n = 7) Non-RPM (n = 10) Female 12 (71) 6 (86) 6 (60) Ethnicity Chinese 14 (82) 6 (86) 8 (80) Pakistani 2 (12) 1 (14) 1 (10) White 1 (6) 0 (0) 1 (10) Age at enrolment, years 14.9 (7.8–17.4) 11.3 (2.6–17.1) 16.9 (12.8–17.6) Causes of kidney failure Hereditary 7 (41) 3b (43) 4d (40) CAKUT 4 (24) 1 (14) 3 (30) Miscellaneous 4 (24) 1c (14) 3e (30) Glomerulopathy 2 (12) 2 (29) 0 (0) Age at KRT initiation, years 13.2 (5.9–15.5) 9.7 (2.3–14.9) 14.3 (8.7–16.01) Dialysis vintage, m 27.2 (8.2–41.4) 17.2 (7.6–27.2) 34.6 (8.8–46.7) Delayed development 7 (41) 4 (57) 3 (30) Data expressed as number (&), or median (interquartile range), as appropriate CAKUT, congential anomalies of kidney and urinary tract; KRT, kidney replacement therapy aReasons for exclusion from the analysis (RPM group, one patient due to transplant; Non-RPM group, n = 8; transplant, n = 4; Transition, n = 2; migration, n = 1; death, n = 1) bCongenital nephrotic syndrome due to NPHS1 variant, genetic podocytopathy due to PLCE1 variant and nephronophthisis cIschaemic nephropathy following major cardiac surgery for congential heart disease dRenal cystic disease, nephronophthisis, autosomal recessive polycystic kidney disease and autosomal dominant tubulointerstitial disease eAtypical HUS, mitochrondial disease and transplant-associated thrombotic microangiopathy Patient reviews and interventions Over the 24-week post-RPM period, there were 72 regular screenings performed by nurses and 11 case conferences were held. There were 202 patient episodes of detailed reviews. Total time spent and phone consults performed were 222.9 min and 8.1 phone consultations per patient/24 weeks, respectively. There were 265 yellow flag and 65 red flag alarms. The leading reasons for red flag alarms were initial drain variance (n = 23, 35.4%) and lost therapy time (n = 17, 26.2%) (Fig. 1A). A total of 90 interventions were recorded (Fig. 1B), including change in PD prescription (n = 31, 34.4%), alarm settings (n = 10, 11.1%), medications (n = 10, 11.1%), dry weight (n = 7, 7.8%), PD machine (n = 4, 4.4%) and advice on patient positioning (n = 9, 10%) and adherence of medication and fluid restriction (n = 19, 21.1%). While there were 5 changes in PD prescription made at clinic during both pre- and post-RPM periods, 31 remote adjustments were performed through RPM. Compared to the pre-RPM period, there was a significant increase in PD regimen adjustments during the post-RPM period (0.7 to 5.1 adjustments per patient/24 weeks, p = 0.05).Fig. 1 A Reasons for red flag alarms during remote patient monitoring. B Interventions performed through remote patient monitoring programme All children performed PD daily and did not miss PD treatment during the post-RPM period. However, one adolescent experienced the lost therapy time alarm for 13 episodes. The patient admitted that he decided for early termination of PD due to issues with schooling. With repeated counselling and support, the patient became adherent to PD treatments with no further issues. Patient outcomes Compared to the pre-RPM period, 4 patients (57.1%, 95% CI 22.5–100%) experienced fewer unplanned hospitalizations and 5 patients (71.4%, 95% CI 34.1–100%) had shorter hospital stays. The overall unplanned hospitalization rate and length of stay were reduced by 45% and 42%, respectively (Table 2). While the RPM group had a higher unplanned hospitalization rate than the non-RPM group at baseline (1 vs. 0.27 episode per patient/24 weeks, p = 0.03), following the implementation of RPM the hospitalization rates became similar between the two groups (0.55 vs. 0.27 episode per patient/24 weeks) (Table 3).Table 2 Outcomes of patients receiving automated peritoneal dialysis with remote patient monitoring Pre-RPM Post-RPM Total no. of unplanned hospitalizations 7 4 Causes of hospitalization Fluid retention 2 (28.6) 2 (50.0) Leakage 1 (14.3) 0 (0.0) Wet contamination 1 (14.3) 0 (0.0) Drainage pain 2 (28.6) 0 (0.0) Hypercalcemia 1 (14.3) 0 (0.0) Malnutrition 0 (0.0) 1 (25.0) Non-specific abdominal pain 0 (0.0) 1 (25.0) Unplanned hospitalization rates, episode per patient/24-week 1.0 0.55 Unplanned hospitalization days, days per patient/24-week 9.43 5.43 Ultrafiltration volume, ml per day 501.7 ± 286.6* 565.6 ± 248.7* Systolic blood pressure, mmHg 119.9 ± 11.2 114.1 ± 12.6 Diastolic blood pressure, mmHg 79.3 ± 7.4 75.4 ± 10.8 No. of anti-hypertensive medications 1.9 ± 1.0 2.0 ± 0.6 Dialysis adequacy (n = 6) Kt/V (total) 2.4 ± 0.7 2.3 ± 0.6 Kt/V (dialysate) 2.1 ± 0.7 2.1 ± 0.6 Kt/V (urine) 0.4 ± 0.6 0.3 ± 0.4 QOL (parent) (n = 7) 61.2 ± 12.2 57.3 ± 13.1 QOL (patient) (n = 4) 72.1 ± 6.7 62.5 ± 12.3 p = 0.03; p = 0.02 QOL, quality of life measure by PedsQL 3.0 ESRD module; RPM, remote patient monitoring Table 3 Outcomes during pre-RPM period (March 2021 to August 2021) between patients who were and were not enrolled into RPM Pre-RPM period Post-RPM period RPM Non-RPM RPM Non-RPM Total No. of unplanned hospitalizations 7 3 4 3 Causes of hospitalization Fluid retention/hypertension 2 (28.6) 3 (100) 2 (50) 2 (66) Leakage 1 (14.3) 0 (0) (0) (0) 0 (0) Wet contamination 1 (14.3) 0 (0) (0) (0) 0 (0) Drainage pain 2 (28.6) 0 (0) (0) (0) 0 (0) Hypercalcemia 1 (14.3) 0 (0) (0) (0) 0 (0) Malnutrition 0 (0) 0 (0) (0) (25) 0 (0) Non-specific abdominal pain 0 (0) 0 (0) (0) (25) 0 (0) Hyponatraemia 0 (0) 0 (0) (0) (0) 1 (33) Unplanned hospitalization rates, episode per patient/24-week 1 0.27* 0.55 0.27 Unplanned hospitalization days, days per patient/24-week 9.43 2.36 5.43 1.91 Systolic blood pressure, mmHg 119.9 ± 11.2 121.1 ± 11.4 114.1 ± 12.6 123.4 ± 10.9 Diastolic blood pressure, mmHg 79.3 ± 7.4 78 ± 8.68 75.4 ± 10.8 79.4 ± 8.21 *p = 0.03 Data presented as number (%), mean ± SD, as appropriate RPM, remote patient monitoring There was a significant increase in daily ultrafiltration (565.6 ± 248.7 vs. 501.7 ± 286.6 ml/day, p = 0.03) and reduction in systolic blood pressure (114.1 ± 12.6 vs. 119.9 ± 11.19 mmHg, p = 0.02) during the post-RPM period. There was also a trend of reduction in diastolic blood pressure (75.4 ± 10.8 vs. 79.3 ± 7.4 mmHg, p = 0.09). The number of anti-hypertensives prescribed was not different during the two periods. There was no difference between the number of clinic visits (7.4 ± 3.0 vs. 7.1 ± 2.2 visits per patient/24 weeks), as well as Kt/V, haemoglobin, calcium, phosphate and parathyroid hormone levels during the two periods. Details on the primary and secondary outcomes are presented in Table 2. None of the patients on RPM required conversion to haemodialysis. Change of RPM over time Compared to the first 3 months following the implementation of the RPM programme, the average time spent for reviews was significantly reduced by 41.1% (46.8 to 27.5 min per patient month, p < 0.001) during the second half of the post-RPM period. The number of red alarms per patient was also reduced by 36.8% (1.9 to 1.2 episodes per patient, p = 0.22). The rate of PD regimen adjustment was the same at 0.7 adjustments per patient month. Quality of life and perceptions Although the quality of life measured by the PedsQL 3.0 ESRD module was not different between the pre- and post-RPM periods, all patients agreed in the questionnaires that the use of RPM was beneficial, which improved their quality of life, sense of security, adherence and reduced PD-related clinic visits, admissions and related complications (Table 4). A follow-up questionnaire distributed to the dialysis team also demonstrated positive perceptions towards patient adherence, engagement, disease understanding and fluid management. Despite an improved efficiency in data interpretation with time, a significant proportion of dialysis team members (> 80%) expressed concerns about the burden of workload which might have affected their regular clinical duties. Nonetheless, most members (83%) agreed that the programme should be extended to all patients on APD with additional manpower support, preferably with an improved nurse-to-patient ratio of 1:5 (Table 5).Table 4 Results from post-RPM implementation questionnaires distributed to patients and carers who performed automated peritoneal dialysis at homea Questions Strongly disagree Disagree Neutral Agree Strongly agree I feel safer to perform peritoneal dialysis at home under frequent monitoring by health care workers 0 0 0 1 (17%) 5 (83%) I feel less burden with shared responsibility of home treatment 0 0 0 0 6 (100%) I feel my/my child’s quality of life has improved with the remote patient monitoring program 0 0 0 2 (33%) 4 (67%) I have become more adherent to the peritoneal dialysis treatment 0 0 0 0 6 (100%) I feel that that there are less unplanned/ unexpected peritoneal dialysis-related admissions with remote patient monitoring 0 0 0 0 6 (100%) I perceive that the number of peritoneal dialysis-related complications are either reduced or prevented 0 0 0 1 (17%) 5 (83%) I feel that there are less medical visits with remote patient monitoring program 0 0 0 2 (33%) 4 (67%) I feel that the number of phone calls or interventions received from the health care team is reasonable 0 0 0 1 (17%) 5 (83%) I think the remote patient monitoring program is beneficial to children on peritoneal dialysis 0 0 0 0 6 (100%) Data are expressed in number (%) aA total of 7 questionnaires were distributed to patients/caregivers who performed peritoneal dialysis following the post-RPM period. Six responses (86%) were received (4 caregivers and 2 patients) Table 5 Perception of the dialysis team towards RPM 1-year after programme implementationa Questions Strongly disagree Disagree Neutral Agree Strongly agree The burden of frequent patient reviews is reasonable 0 0 0 5 (83%) 1 (17%) I often need to spend additional work time on patient reviews 0 0 1 (17%) 4 (67%) 1 (17%) My routine clinical duty is not affected by the remote patient monitoring programme 0 1 (17%) 0 3 (50%) 2 (33%) The program should be extended to all children receiving peritoneal dialysis 0 0 1 (17%) 1 (17%) 4 (67%) Additional manpower is needed for expanding the remote patient monitoring service 0 0 0 1 (17%) 5 (83%) The overall efficiency in data analysis improves with gaining experience on the Sharesource® system 0 0 0 1 (17%) 5 (83%) More frequent reviews are required during the initial phase of enrolment 0 1 (17%) 0 1 (17%) 4 (67%) The efficiency in individual patient review often improves with time 0 0 0 5 (83%) 1 (17%) Patient/family are more adhered to dialysis treatments 0 0 0 4 (67%) 2 (33%) The dialysis team is more engaged with the patient/family 0 0 0 2 (33%) 4 (67%) The program helps patients to reach target body weight and improve blood pressure control 0 0 0 4 (67%) 2 (33%) The program improves patient/ family understanding of their disease condition and management 0 0 0 3 (50%) 3 (50%) The joint reviews with physician and technicians on Sharesource® system are helpful and beneficial 0 0 0 1 (17%) 5 (83%) In your opinion, what is the optimal nurse to patient ratio is? 1:5 (100%), 1:10 (0%), 1:15 (0%), 1:20 (0%) aA total of 6 questionnaires were distributed to 3 dialysis nurses, 2 paediatric nephrologists and 1 technical supporting staff for the cloud-based Sharesource®platform Cost savings related to reduction in hospitalizations We estimated the total medical expenditure related to unplanned hospitalizations for management of complications arising from PD. The average nominal cost for hospitalization in public hospital was USD 653.8/day. The reduced bed days in our cohort was 8.0 days/patient year, resulting in a cost saving of USD 5230.4/patient year. Discussion In this prospective study, children on APD benefited from RPM with fewer unplanned hospitalizations and shorter hospital stays. In addition, fluid management improved with increased daily ultrafiltration and lower systolic blood pressure. There was satisfactory treatment adherence and favourable perception to the RPM programme. Over time, our data demonstrated an improved PD performance as evidenced by fewer number of alarms, and consequently the time required for RPM service also was significantly reduced. Hospitalizations are common among children receiving PD [2, 4]. The main objective of RPM is to identify and resolve problems early, in order to reduce PD-related complications and avoid unnecessary admissions. Our data showed that 57% and 71% children had fewer unplanned admissions and shorter hospital stays, respectively. This may be attributable to the intensification of patient monitoring, frequent communications and timely interventions. In accordance with previous reports [11], there was enhanced blood pressure control following counselling on fluid restriction and proactive PD adjustments optimizing ultrafiltration. This postulation is supported by the fact that a similar number of anti-hypertensive medications were prescribed between the pre- and post-RPM periods. Furthermore, all our patients demonstrated excellent adherence to their PD treatments following the implementation of RPM. Non-compliance is well known to be prevalent among PD patients [12, 13], and may lead to higher rates of peritonitis, hospitalizations and mortality. Specifically, non-adherence was correctly identified in an adolescent and favourable outcome was observed following appropriate counselling. Minimizing hospital visits is important to facilitate schooling, and to reduce infection risk especially during the COVID-19 pandemic [11]. Indeed, questionnaires distributed to patients and dialysis teams showed that RPM was associated with better patient engagement, adherence and disease awareness. The success of RPM relies much on dedicated nurses who have spent hours on patient reviews, communications and interventions. Fortunately, the hours required for RPM tend to decrease over time because patients’ clinical status often improves. There were also fewer red flag alarms (reduced by 37%) during the second half of the post-RPM period. Smoother procedures may lead to better quality of sleep for patients at night [9]. Importantly, frequent patient reviews pose a significant burden on the dialysis team. Appropriate manpower provision and facilitation should be offered to ensure sustainability and quality of the RPM programme. The use of RPM is associated with improved quality of life in children and adults [8, 9]. A recent survey showed that 90% of patients/caregivers felt safe under RPM [9], although the scales on quality of life did not reflect an improvement in our cohort. There are a few explanations. First, the number of patients was too small to detect any significant difference. Second, initial frequent communications with patients/caregivers might have raised the awareness of their poor physical health and induced stress. However, as shown in our questionnaires, all patients perceived a better quality of life and sense of security. We believe that the benefit of improved quality of life would be more apparent with an extended observational period on RPM. Although we have a small sample size, our study provides important prospective outcome data with minimal missing data. However, there is potential selection bias as the most complicated patients were enrolled in the RPM programme since only limited PD machines in our unit had RPM function. Second, while there was a reduction in hospitalizations following implementation of RPM, improvement in certain hospitalizations, such as wet contamination, might not be truly measurable by the RPM programme. In conclusion, cloud-based RPM in children receiving APD is associated with fewer and shorter unplanned hospital visits, improved fluid management and excellent adherence to PD. The programme also improves patient engagement and disease awareness, and may potentially save medical expenditures due to fewer hospitalizations. Further well-designed studies with larger paediatric cohorts are required to evaluate patient-centred outcomes. Supplementary information Below is the link to the electronic supplementary material.Graphical Abstract (PPTX 99 KB) Acknowledgements We would love to thank Ms. Ivy Yu from Baxter Healthcare who provided staff training and technical support on the use of the Sharesource® platform. We would also love to thank occupational therapist Ms. Phoebe Chan for performing the quality of life assessment. 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. Chan Y Ma AL Tong P Lai W Tse NK Chronic peritoneal dialysis in Chinese infants and children younger than two years Hong Kong Med J 2016 22 365 371 27313274 2. Chan EYH Yap DYH Wong WHS Ho TW Tong PC Lai WM Chan TM Ma ALT Demographics and long-term outcomes of children with end-stage kidney disease: a 20-year territory-wide study Nephrology (Carlton) 2022 27 171 180 10.1111/nep.14007 34837272 3. Chan EY Borzych-Duzalka D Alparslan C Harvey E Munarriz RL Runowski D Vidal E Coccia PA Jankauskiene A Principi I Colostomy in children on chronic peritoneal dialysis Pediatr Nephrol 2020 35 119 126 10.1007/s00467-019-04372-x 31673828 4. Neu AM Sander A Borzych-Dużałka D Watson AR Vallés PG Ha IS Patel H Askenazi D Bałasz-Chmielewska I Lauronen J Comorbidities in chronic pediatric peritoneal dialysis patients: a report of the International Pediatric Peritoneal Dialysis Network Perit Dial Int 2012 32 410 418 10.3747/pdi.2012.00124 22859841 5. Chua AN Warady BA Adherence of pediatric patients to automated peritoneal dialysis Pediatr Nephrol 2011 26 789 793 10.1007/s00467-011-1792-2 21350797 6. Lai W-M Quality of life in children with end-stage renal disease: does treatment modality matter? Perit Dial Int 2009 29 190 191 10.1177/089686080902902S38 7. Jung H-Y Jeon Y Kim YS Kim DK Lee JP Yang CW Ko EJ Ryu D-R Kang S-W Park JT Outcomes of remote patient monitoring for automated peritoneal dialysis: a randomized controlled trial Nephron 2021 145 702 710 10.1159/000518364 34515160 8. Uchiyama K Morimoto K Washida N Kusahana E Nakayama T Itoh T Kasai T Wakino S Itoh H Effects of a remote patient monitoring system for patients on automated peritoneal dialysis: a randomized crossover controlled trial Int Urol Nephrol 2022 54 2673 2681 10.1007/s11255-022-03178-5 35362819 9. Uzun Kenan B Demircioglu Kilic B Akbalık Kara M Taktak A Karabay Bayazit A Yuruk Yildirim ZN Delibas A Aytac MB Conkar S Kaya Aksoy G Evaluation of the Claria sharesource system from the perspectives of patient/caregiver, physician, and nurse in children undergoing automated peritoneal dialysis Pediatr Nephrol 2022 14 1 7 10. Goldstein SL Graham N Warady BA Seikaly M McDonald R Burwinkle TM Limbers CA Varni JW Measuring health-related quality of life in children with ESRD: performance of the generic and ESRD-specific instrument of the Pediatric Quality of Life Inventory (PedsQL) Am J Kidney Dis 2008 51 285 297 10.1053/j.ajkd.2007.09.021 18215706 11. Bunch A Ardila F Castaño R Quiñonez S Corzo L Through the storm: automated peritoneal dialysis with remote patient monitoring during COVID-19 pandemic Blood Purif 2021 50 279 282 10.1159/000511407 33161396 12. Griva K Lai AY Lim HA Yu Z Foo MWY Newman SP Non-adherence in patients on peritoneal dialysis: a systematic review PLoS ONE 2014 9 e89001 10.1371/journal.pone.0089001 24586478 13. Bernardini J Nagy M Piraino B Pattern of noncompliance with dialysis exchanges in peritoneal dialysis patients Am J Kidney Dis 2000 35 1104 1110 10.1016/S0272-6386(00)70047-3 10845824	36449100	PMC9709751	NO-CC CODE	2022-12-01 23:23:39	no	Pediatr Nephrol. 2022 Nov 30;:1-8	utf-8	Pediatr Nephrol	2,022	10.1007/s00467-022-05828-3	oa_other
==== Front Ann Oper Res Ann Oper Res Annals of Operations Research 0254-5330 1572-9338 Springer US New York 5091 10.1007/s10479-022-05091-7 Original Research Product availability and stockpiling in times of pandemic: causes of supply chain disruptions and preventive measures in retailing http://orcid.org/0000-0003-4117-3330 Ovezmyradov Berdymyrat [email protected] grid.445897.3 0000 0004 0609 5872 Department of Transportation and Logistics, Transport and Telecommunication Institute, Lomonosova Iela 1, Riga, 1019 Latvia 30 11 2022 133 17 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. The coronavirus pandemic in 2020 brought global supply chain disruptions for retailers responding to the increased demand of consumers for popular merchandise. There is a need to adapt the existing supply chain models to describe the disruptions and offer the potential measures that businesses and governments can take to minimize adverse effects from a retail logistics perspective. This research analyses the possible reasons for supply and demand disruptions using a mathematical model of a retail supply chain with uncertain lead times and stochastic demand of strategic consumers. The established concepts of supply chain management are applied for the model analysis: multi-period inventory policies, bullwhip effect, and strategic consumers. The impact of the pandemic outbreaks in the model is two-fold: increased lead-time uncertainty affects supply, while consumer stockpiling affects demand. Consumers' rational hoarding and irrational panic buying significantly increase retailers' costs due to higher safety stock and demand variability. The bullwhip effect further exacerbates the disruption. The research contributes to the recent literature on business response to supply chain disruptions by developing a model where both retailers and consumers decide on the order quantity and reorder point during a pandemic outbreak. Buying limits, continuous inventory review, government rationing, substitutability, and omnichannel fulfillment are the measures that can limit the damage of supply chain disruptions from stockpiling during the pandemic. Effective communication and price and availability guarantees can mitigate the negative impact of panic buying. Keywords COVID-19 Supply chain disruption Omnichannel retail Bullwhip effect Inventory management Simulation http://dx.doi.org/10.13039/100010665 H2020 Marie Skłodowska-Curie Actions 870647 ==== Body pmcIntroduction Outbreaks have long been known to cause supply chain disruptions, but their severity was considered low relative to other disruption causes. COVID-19 changed that perception. The recent coronavirus pandemic has drawn the attention of policymakers towards how globalized supply chains of retailers can cope with consumers' increased demand for popular merchandise. When the outbreak of COVID-19 (hereinafter, the pandemic) happened at the beginning of 2020, even major retailers with sophisticated supply chain management were unable to cope with a surge in consumers' demand for products such as toilet paper, disinfectants, and certain foods. From a supply chain perspective, the highly publicized shortages resulted from disruptive changes in both supply and demand. Notably, the pandemic temporarily disrupted the supply of products transported from distant locations after governments imposed restrictions on business operations and travel. Yet, the disruptive force of consumer reaction to pandemics could seemingly exceed the effect of supply disruption. In many countries, there was a public perception that consumer stockpiling was a more prominent cause of shortage than supply chain disruptions (Ipsos, 2020). An essential research question is what impact the pandemic has on the supply chain when both the demand-side and the supply-side disruptions occur. Answering the question should be supported by facts and analytical modeling. The limited literature on modeling the business impact of the pandemic mainly focuses on supply-side disruptions, while the consumer-related demand disruptions received relatively little attention. This research addresses the urgent need for new investigations of pandemic effects on businesses by contributing to the modeling research in retail supply chain disruption that integrates stockpiling consumers. This research aims to identify the underlying reasons behind the supply chain disruptions caused by the pandemic and corresponding measures to manage post-pandemic operations. In this paper, the literature and model analysis results are discussed from both the supply and demand viewpoints, but the main focus is on the demand side of retail disruption. This paper makes a contribution by incorporating the perspective of stockpiling and demand substituting consumers in a novel model of supply chain disruption. Though firms actively adapt supply chains and governments plan to ease lockdown measures further, the negative impact of the aforementioned disruptive factors is likely to remain strong for a long time after 2020 as countries are still recovering from the pandemic outbreaks and shortages of certain goods. The findings of the presented model analysis thus can be relevant because waves of the pandemic expected to occur in the future are likely to cause recurring supply chain shocks. This research employs a mathematical model of the retail supply chain and stockpiling consumers to explain the disruptions during the epidemic and discuss the preventive measures. In the main model, the retailer uses reorder point and order-up-to policies in multi-period inventory problems with stochastic consumer demand. The retailer adjusts ordering according to the random demand of consumers who minimize their inventory and shopping costs. Consumers consider increased shortage per unit cost during the pandemic, which could motivate hoarding based on rational expectations in case of temporal supply chain disruptions. Furthermore, strategic consumer behavior could be irrational in panic buying and related to consumption's combined psychological and psychological utility. The business objective of the retailer is the achievement of the target inventory service level. A separate subsection discusses alternative inventory policies. In the extension of the main model, the retailer considers stockout-based substitution and corresponding risk-pooling reducing the bullwhip effect. In the second model extension, the forward-looking behavior of strategic consumers is taken into account. Thus the problem of inventory management during the pandemic is investigated from the perspective of three influential models in supply chain management: newsvendor-based multi-period models of stochastic demand, risk pooling, and strategic consumers. Results indicate that a combination of increasing uncertainties in the supply and demand sides of the supply chain leads to a surge in demand after the pandemic. Demand substitution reduces the harmful effect of uncertainty. Furthermore, making a strategic consumer buy quantities that maximize only their physiological utility, desirably once in lead-time between replenishment, could effectively minimize the damage of stockpiling. Specifically, buying limits and other preventive measures are discussed as managerial implications. The next section presents the literature review. The following two sections present the analysis of the main model and its two extensions. Numerical examples and managerial implications are discussed in the following two sections. The final section summarizes the results and suggests future research. Review of literature on supply chain disruptions related to pandemic This section covers several streams of literature relevant to the study: supply chain risks, risk pooling in the bullwhip effect, hoarding, panic buying, and disruptions during the COVID-19 pandemic. Supply chain risks Pandemic has long been considered low risk in supply chain management with low probability and seemingly controllable mitigation (Manuj & Mentzer, 2008). Disruptions caused by the most recent pandemic make researchers rethink the risks for global supply chains. The resulting losses for retailers could amount to $700 million in the US alone from March to April 2020 (Thomas, 2020). Despite the now evident scale of pandemic impact on supply chains, there was limited analytical research on the effects of disasters on supply chains, even less on pandemic-related disruptions in retail. However, a substantial number of publications currently exist on the topic of supply chain disruptions and resilience, and the recent papers include analysis of pandemic effects (Katsaliaki, 2021). One of the few papers on supply disruptions caused by epidemics before the pandemic in 2020 was the review by Dasaklis et al. (2012) that focused on medical supply. Another important work by Rodrigue (2016) mainly focused on transport freight relations to the pandemic. Despite the lack of analytical research on supply chain disruptions from outbreaks, results of studies modeling disasters, in general, could apply to the case of a pandemic. The ripple effect occurring when a disruption cascades downstream and impacts the performance of the entire supply chain is relevant for pandemic times; therefore, findings of related research are relevant to the issue of business responses to COVID-19 (Dolgui & Ivanov, 2021). Operational risks due to a pandemic in global supply chains to be considered in conjunction with supply risks can be summarized as follows (Rodrigue, 2016):The early phases of a pandemic in the modern high-speed transportation systems facilitate the spreading outbreak at the global level; In the later phases, economic activities are disrupted without continuous deliveries of resources as critical supply chains can shut down; The velocity of highly efficient global transport could lead to the paradox of the faster outbreak spread at the worldwide level relative to the local level; Modern food distribution relies on low levels of perishable goods for every day and stable demand, as supermarkets typically have only several days of supply for dairy, produce, and meat, while for packaged food (pasta, canned goods, etc.), the supply is one to two weeks. Several recent studies addressed the problem of supply chain risk and disaster relief to control the spread of COVID-19. The ripple effect in supply chains during the pandemic was simulated and visualized using the system dynamics approach (Ghadge et al., 2021). Critical facilities such as warehouses for storage of emergency supplies can be located to satisfy the varying demand caused by pandemics with the aid of a two-phase optimization framework based on the Lagrangian relaxation approach (Liu, 2021). Logistics service providers managed to stay resilient during the COVID-19 outbreak, an external shock of high impact and low probability (Herold et al., 2021). Organic and mechanistic management control enabled the management of the COVID-19 crisis (Passetti et al., 2021). Supply chain 4.0 concepts became even more relevant for resilient post-COVID-19 supply chains (Frederico, 2021). Both supply risk sources and supply network recoverability are important for supply resilience (Lorentz et al., 2021). Striking a balance between being lean and resilient became one of the most important managerial implications already during the early stages of the pandemic (Raassens et al., 2021). COVID-19 has managerial implications for alertness in scenarios of huge disruptions: there is a trade-off involving the supply chain efficiency and resources orchestration to support the resilience (Queiroz et al., 2022). In this paper, the operational risk factors of production and transport are included in the research model as an external parameter reflected in the lead-time variability, which retailers cannot control. Such an assumption simplifies decision-making in supply chain analytic practice as is common in real business. In fact, variance and related measures were widely used to analyze supply chain risk that can be highly relevant for the research agenda during the COVI-19 pandemic (Choi, 2020a). The main model described further assumes the standard deviation for calculating the retailer's safety stock increases during the pandemic. Bullwhip effect The well-studied bullwhip effect in supply chain management is particularly relevant for the extension of the main model in this research. The effect involves both the supply-side and demand side of the disruptive effect: demand signal processing, rationing, batching and price variations (Lee et al., 1997). The bullwhip effect of the epidemic from higher demand variability constitutes a big problem for retailers' supply chains. Numerous authors theoretically demonstrated negative consequences of the bullwhip effect (notably Chen et al., 2000; Lee et al., 1997; Metters, 1997). Highly variable orders and inventory implies additional costs for all supply chain partners: sudden surges in demand lead to rising production and storage and labor expenses even when long-term sales remain constant. On the other hand, few researchers, including Cachon et al. (2007) and Sucky (2009), found limited empirical support for the negative bullwhip effect in several sectors: the variability does not necessarily increase or decrease at upstream stages. The pandemic modeling results in this study will be mainly discussed from the perspective of the bullwhip effect at the business-to-customer (B2C) level rather than the business-to-business (B2B) level as was typical for previous research. Indeed, the pandemic could become a more significant problem for supply of the popular products only if the following happens on a larger scale: borders closure, shortage of drivers, ports closure, production or warehouse employees getting sick, export bans, and immigration restrictions on harvesting workers (Terazono & Evans, 2020). As already mentioned, demand-side disruptions for retailers could be a threat comparable to supply disruptions during the pandemic. Available statistics demonstrated substantially higher store traffic at major retailers at the beginning of the pandemic in the US: from up to 30% at Walmart to almost 100% at Costco (Placer Labs, 2020). However, this surge was quickly followed by a sharp decline (about 50%) in the following weeks. Though foot traffic is an imprecise measure of consumer demand due to an unidentified number of shoppers switching stores or postponing purchases before the outbreak, the variability increased during the pandemic. Following theoretical predictions in the models of Chen et al. (2000) and similar authors, the sudden change in downstream consumer demand could provoke a substantial bullwhip effect in upstream supply chain unless business partners closely coordinate their actions. The bullwhip effect can be driven by the ripple effect brought by disruptions due to COVID-19 and the corresponding impact on supply chain performance and changes in its structure (Ivanov & Dolgui, 2021). Simulations suggest that two-stage supply chains can be more vulnerable than three-stage supply chains during the pandemic disruption, but they show better effects at the recovery stage (Rozhkov et al., 2022). The bullwhip effect in this research is discussed from the perspective of consumer stockpiling and demand substitution in a novel model of supply chain disruption. Hoarding behavior Consumers could stockpile and impulse buy worrying about the availability of essential products during the pandemic (Anas et al., 2022; Satish et al., 2021). Early empirical and marketing research into consumer hoarding resulting from gasoline and toilet paper shortages dates back to the 1970s (Stiff et al., 1975). Observation of such behavior took part even earlier: during World War II, American consumers hoarded clothing in scare buying, and the federal government collected fines from retailers busting price ceilings (Mower & Pedersen, 2018). Stockpiling could thus happen in product categories other than grocery and hygiene, though empirical evidence is limited. Perceived scarcity of fast fashion could accelerate in-store hoarding (Byun et al., 2012). Shou et al. (2013) provided theoretical support for the conjecture that risk-averse consumers are likely to stockpile low-price products with low consumer holding costs (implying the expected spoilage cost was low), and quota policy could be beneficial for retailer's profit. Indeed, Table 1 supports such characteristics of products in high demand during the pandemic with affordable prices and long storage time (hereinafter, popular products).Table 1 Buying limits at selected retailers (Ziady, 2020; Daoud, 2020; Jumrisko, 2020; Salaverria, 2020; Repko, 2021) Retailer Country and year Products Limit per customer Sainsbury UK 2020 Toilet paper, soap, long-life milk 2 Tesco UK 2020 All products 3 Boots UK 2020 Hand sanitizers 2 Cole Australia 2020 Mince, pasta, flour, dry rice, paper towels, paper tissues, and handsanitizers 2 Woolworths US, Australia 2020 Packaged goods 2 REWE Group (including REWE and Penny) Germany 2020 Long-life foods, canned goods, and drugstore items Decision up to store manager Wal-Mart US 2020 Items in unusually high demand Decision up to store manager Aldi UK 2020 One unit of toilet paper Two units for dried pasta, flour, rice, paper towels, tissues, hand sanitizer NTUC FairPrice Singapore 2020 Four packs of paper products, two bags of rice, four bundles of instant noodles, $36 worth of vegetables Retailers* Philippines 2020 Disinfectant alcohol, hand, sanitizers, face masks, toilet paper, local canned, sardines, instant noodles, bath soap, milk, instant coffee in sachets, mineral water and bread (limits not specified) COSTCO US 2021 Toilet, paper, bottled water and cleaning supplies (limits not specified) Limits had been imposed on the country’s manufacturers and retailers that later asked the Department of Trade and Industry to remove the limits Even without a pandemic, retailers may use buying limits aiming at high, but not excessively high purchase quantities by means of two approaches: offer quantity limits (for example, allowing the price deal a maximum of two times) and unit quantity limits (for example, restricting to a maximum purchase of two units of the discounted product). The associated risk is that consumers misunderstand the limits, which leads to purchasing fewer units when one of the two restrictions is imposed on a multiple unit price promotion (Carlson, 2021). There appeared to be fewer reports of buying limits in 2021 after being increasingly announced in 2020 soon after the pandemic outbreak. Certain limits were again introduced on specific items across the US at the beginning of 2022 due to the issues related to the omicron variant of the coronavirus, weather, the supply chain struggles and labor shortages (Tyko, 2022). While the shortage of merchandise had been the main factor in 2020, the purchase limits in 2021 could be primarily driven by delays in deliveries despite the available supply of the merchandise (Repko, 2021). This research focuses on popular products during the pandemic, though other categories were negatively affected too. Apparel sales declined by more than half during March 2020, and already troubled American department stores were hit worst (Howland, 2020). Store traffic sharply decreased by almost 80% at major consumer electronics stores during the pandemic in the US in March (Placer Labs, 2020). However, electronics sales in Russia increased by about 20% over the same period (Maтoвникoв et al., 2020). This research does not consider product categories with fashion-like and perishable characteristics and non-essential items such as alcohol that could experience unusual demand during the pandemic. Not all stockpiling can be attributed to immediate consumer demand. Rational consumers could shop more often and buy less per trip when price variability is high (Ho et al., 1998). The impact of resellers speculating on popular products is hard to measure. An outrageous example of stockpiling that got caught in Australia could be the tip of the iceberg: a group of shoppers bought $10 000 worth of the popular items, failed to sell them online, and then tried to get a refund from a supermarket (Siebert, 2020). Retailers need to isolate the impact of the pandemic on sales from the resale, pricing, and other numerous factors that could affect aggregated consumer demand. This research assumes that demand increases are derived directly from consumption at fixed prices, not from resellers. Thus the impact of the coronavirus could be very different depending on location and product category. Furthermore, the benefits of imposing buying limits on popular products for retailers' service levels are discussed. Such purchase regulations might eliminate hoarding, but they exacerbate supply shortages due to firms reducing orders and production. A mixed approach combining price and purchase regulation can thus mitigate the shortages when capacity becomes insufficient at the beginning of a pandemic (Li and Dong, 2021). Panic buying Most empirical studies on the subject of panic buying involve consumer response to natural disasters. Japan, for instance, is a suitable country to study buying behavior in conditions of living within areas prone to earthquakes, typhoons, landslides, and tsunamis. Consumers in the Tokyo area exhibited higher levels of panic buying without apparent reason, particularly in households with many family members and a middle-aged or older homemaker (Masahiro and Koichiro, 2014). Various factors in the food supply chain such as resilience: emergency planning, staff training, food supply backup, food suppliers, infrastructure, location, service providers, and insurance could define the level of organizational preparedness for panic buying (Hecht et al., 2019). Panic buying during a pandemic is still an understudied topic. Misinformation was a serious contributing factor in the panic ensuing from the COVID-19 outbreak (Elavarasan & Pugazhendhi, 2020). Certain officials and retailers in 2020 called customers to refrain from buying unusually higher amounts of products compared to regular consumption before the pandemic, and there were even instances of public and media shaming that blamed consumers for the shortage of certain products (Daoud, 2020; Jumrisko, 2020; Taylor et al., 2020). Business, government, and media efforts to convince consumers about sufficient availability and warn about the damaging effects of panic buying for a society seemingly had limited success. In response, several supermarkets, particularly in the UK, moved toward rationing popular food and other supplies, facing increased demand during the pandemic (Table 1). Furthermore, businesses expanded shelf space, counters, and logistics capacity for popular products during the pandemic. Online retailers were not immune to panic buying either: Amazon prioritized sales of medical items for hospitals, and delivery of non-essential items (primarily by third-party sellers) was delayed (Rey, 2020). Several UK online retailers suggested temporary limits and introduced virtual queues for the most popular items (Ziady, 2020). Importantly for this research, a distinction has to be made between the changes in the underlying demand of consumers during the pandemic. Limited empirical evidence reveals critical differences in how retailers coped with panic buying of various products during the COVID-19 outbreak (Taylor et al., 2020). First, the total consumption of toilet paper as per actual use cannot realistically increase, but consumers still purchased (estimated) 40% more by remaining at home and using fewer public facilities. Despite enough domestic supply for total consumption of toilet paper, the challenging transition from commercial to retail channels contributed to the sense of widespread shortages. Unlike toilet paper, the actual use of products such as spaghetti, flour, sugar, and dry yeast increased significantly with the changing home consumption patterns. How well retailers were able to respond depends on the presence of multichannel suppliers (insufficient responses in case of competition between foodservice and retail for pasta) and whether products could be stocked at supermarkets by suppliers (proving to be resilient in case of drinks and snacks). Models of COVID-19 impact on global supply chain Numbers of infections are an important part of the shortage function describing changes in the inventory management model illustrated in this paper within the simulation section. Mathematical models simulating the disease spread were widely taken into account in policymaking at different levels of responding to the pandemic in 2020 (Adam, 2020). Alternative plans were suggested to reliably contain the pandemic while mitigating economic consequences (Baveja, 2020). Already early evidence on SARS-CoV-2 responsible for the COVID-19 outbreak pointed out the characteristics capable of disrupting activities of a wide range of organizations: the infection grew exponentially, justifying the typical responses to limit the disease such as isolation, quarantine, lockdown, social distancing, screening, and testing (Kaplan, 2020; Tsiligianni et al., 2022). Unlike the previous epidemics, the COVID-19 pandemic was difficult to model and control due to its long incubation period resulting in various measures determined by outdated data (Alvarez & Kreinovich, 2020). The traditional mitigation techniques of the past pandemics were not capable of containing the COVID-19 (Abideen, 2020). For policymakers, the COVID-19 characteristics also meant there was a time to be risk-averse and a time for risk-taking during the contagion or recovery phases of the pandemic (Van Oorschot et al., 2022). Overall, operations research can be applied to address the ripple effect at five pandemic stages as per the WHO classification: anticipation, early detection, containment, control and mitigation; and elimination (Ivanov & Dolgui, 2021). Earliest papers on the economic effects of the pandemic since 2020 were published soon after the outbreak had started. The number and speed of publications on COVID-19 by social scientists increased to the extent that concerns were raised about maintaining scientific rigor (Fowler, 2020). Among the vast number of papers already available on COVID-19, this subsection focuses on mathematical models related to supply chain disruptions. Various well-known and emerging theories were offered to help researchers build knowledge about the COVID-19 effects on supply chains (Craighead, 2020). The complex structure of global supply chains magnifies losses due to COVID-19, and pandemic control measures such as lockdowns require coordinated efforts and support across countries (Guan et al., 2020). Each industry might require unique practical approaches to minimize disruptions caused by the pandemic. For instance, a decision support system based on specialist medical knowledge and fuzzy inference can aid demand management in the healthcare supply chain (Govindan et al., 2020). Consumers' worry about COVID-19 might lead to the failure of the static service operation so that new "bring-service-near-your-home" operations can help save the service businesses (Choi, 2020b). Simulation experiments demonstrated the timing of the closing and opening of the facilities in a multi-echelon supply chain determined the COVID-19 impact rather than disruption duration or the epidemic propagation speed (Ivanov, 2020). Reducing risks in the post-COVID-19 supply chain should balance global sourcing with local sourcing and adopt multiple sources—management needs to focus not only on costs but also on resilience (Remko, 2020). Game theory and numerical examples were used to model supply chain network disruptions in terms of workforce shortages that became a critical issue as consequences of illnesses, death, travel, and other restrictions during the pandemic (Nagurney, 2021a). The pandemic devastated global economic growth due to its impact the consumption behavior (Ajmal et al., 2021). Internationally, the performance of particular private firms during the pandemic depended on financing sources, industry sectors and location (Golubeva, 2021). Predictive analytics for policymakers forecasting COVID-19 growth rates and consumer demand involved time-series, Google trends, epidemiological, and machine-learning models based on deep-learning, nearest neighbors and clustering (Nikolopoulos, 2021). Delasay et al. (2021) linked retailers’ operational changes in response to COVID-19 to the customers’ shopping behavior in a model of the delivery and curbside pickup. Researchers pointed to the challenges of competition and price pressure all being reinforced in the post-COVID period for omnichannel retail as a high-transparency context (Salvetti et al., 2022). Likewise, the relationships between the responses of omnichannel retailers and consumers to the pandemic are the focus of this research; however, the presented model incorporates a wider range of variables related to inventory policies, risk pooling, and strategic consumers. This literature review reveals a considerable number of empirical and modeling studies related to the COVID-19 effects on business as of 2022 (some of the publications are discussed further in the Discussion section of this paper). However, few academic papers integrated both supply and demand sides of the supply chain disruptions due to the pandemic. Furthermore, to the best of the author's knowledge, no detailed investigation was dedicated to demand substitution and interactions between omnichannel retailers and stockpiling consumers during the disruptions. This research fills the respective gaps. Model of retail supply chain with consumer hoarding At the most basic level of the supply chain, the following aspects of operations problems can be identified in retailers' inventory management during the pandemic: product availability, product variety, and avoidance of the bullwhip effect due to stockpiling. Within the existing literature presented in the previous section, hoarding and panic buying are often given separately as causes of disruptions. Table 2 presents the notations used in this paper.Table 2 Notations and symbols p Unit retail price of retailer w Unit wholesale price of retailer v Unit reduced price of retailer Q Order quantity of retailer R Reorder point level in RQ inventory policy of retailer O Fixed ordering cost of retailer H Unit holding cost of retailer M Mean of total demand observed by retailer J Standard deviation of demand observed by retailer L Mean lead-time in periods of retailer U Standard deviation of lead-time of delivery to retailer t Phase of current period B Safety stock in units of inventory of retailer Z Safety factor to determine safety stock level of retailer CRQ Retailer's expected total cost per period Ccons Individual consumer's expected total cost of consumption per cycle β Portion of brand-switching customers γ Portion of store-switching customers f(x) Probability density function (PDF) of demand F(x) Cumulative distribution function (CDF) of demand N Total number of sales periods in retailer's planning ζ Number of risk-averse consumers μ Individual consumer's mean consumption rate in units per period σ Individual consumer's standard deviation of consumption rate in units per period u Reservation price of strategic consumers ψ Percentage of consumer's purchase cost per unit of carried inventory over planning horizon χ Consumer;s fixed shopping cost Έ Consumer;s expected number of shortages per period η Consumer's individual shortage cost per unit δ Discount of future consumption by strategic consumers r Consumers' perceived probability of getting a product in the future at clearance price Ltransport Delivery lead-time in sS periods of retailer Lreview Inventory review periods in sS policy of retailer CV Coefficients of variation of B2B orders Θ Bullwhip effect as ratio of the coefficients of variation of orders, upstream CV to downstream CV ρ Coefficient of correlation of demands ʎ Retailer's cost of unit increase in bullwhip effect n ∈{a,b,..,i} Product brands k ∈ {1,2,..,i} Retail stores Ʒ New daily cases of COVID-19 infections Ƹ Average number of cases of COVID-19 infections as of May 2020 in simulation RQ Notation for inventory policy with reorder point and fixed order with continuous review sS Notation for inventory policy with periodic review and minimum/maximum levels I Initial inventory in each period with sS policy of retailer s Minimum level of inventory with sS policy of retailer S Maximum up-to level of inventory with sS policy of retailer π Expected profit of retailer This paper defines consumer stockpiling as purchasing above the average demand due to hoarding and panic buying behavior. Hoarding is distinguished from panic buying in the presented model as a type of stockpiling based on rational arguments for consumers to buy more than necessary for satisfying their physiological needs in a certain period given the actual circumstances of holding goods at home and ongoing product availability. Panic buying is defined in this paper as stockpiling based on less rational motives due to unlikely events that could happen shortly, such as extreme shortages or soaring prices. In the following subsection, the problem of availability due to hoarding is addressed in the main model, followed by a subsection on alternative inventory policies. The next section presents two extensions of the main model that address the issues of the bullwhip effect and panic buying. Model setting and assumptions The simple supply chain in this research includes a retailer that faces the stochastic demand of consumers for a popular product. The product is sold at a regular fixed price per unit, p, which does not change from period to period, even after the pandemic starts. The retailer buys the product at a fixed wholesale price, w, from external suppliers. Since the inventory can be transferred from period to period, this is a multi-period inventory problem for the retailer. The retailer has a reorder point and fixed order policy (hereinafter, RQ) in which it chooses two key decision variables that provide a reasonable target service level of inventory: order quantity, Q, and reorder point, R, that triggers the placement of a new order once the reorder point level of stock is reached. The inventory review system should be continuous, which is made possible through retailing Point-of-Sale systems. O is the retailer's fixed ordering cost that includes ordering and shipping and handling costs per order. H is holding cost per unit of inventory during one period. M is the mean demand in units per period, while N is the total number of sales periods, the retailer plans in its operations (planning horizon). Usually, N is assumed to be twelve months in an annual plan. J is the standard deviation of demand during one period. L is mean lead-time (in periods), and this is the time between the placement of an order by the buyer and the delivery of this order by the supplier. U is the standard deviation of lead-time. The choice of target service level (rather than profit maximization) and other variables in the model is supported by a simulation based on canned food data and generalizable to many firms facing supply chain disruptions due to COVI-19 (Dohmen et al., 2021). There are three distinct periods in this model denoted: 1st, pre-pandemic (before March 2020); 2nd, the start of pandemic (during March 2020); 3rd, after the pandemic start (after March 2020). As discussed later, preliminary empirical evidence supports such approximation. The following consumer behavior model is a simplified adaptation of the model described by Ho et al., 1998. Consumers' purchasing policy is analogous to the RQ policy of retailers, except that the objective is cost-minimization instead of target service level. Consequently, the consumer decides on optimal purchase quantity and reorder point that triggers store visits. To model consumer behavior, the following assumptions have to be made for tractability. Replenishment is instantaneous, so lead-time is negligible. External competition and store switching are not present in the main model. Consumers make only planned purchases buying a product after they run out of it at home. Consumers are homogeneous in the main model concerning the parameters mentioned above. There are ζ risk-averse consumers in the local market that periodically visit the retailer to satisfy their needs. A consumer consumes a product at a stochastic rate with a mean μ and standard deviation of σ units per period, which does not change during and even after the pandemic. Certain variability σ of the demand from each consumer is because of moderate uncertainty in consumption rate and timing of store arrival throughout a given period. The planning horizon of each consumer in which the consumption and shopping expenses are minimized lies within one replenishment cycle. Thus the total consumer demand that the retailer has to meet for each period is a sum of all individual consumer demands, so M = μ ζ. Therefore, the variability of consumer demand in the main model is a consequence of variable consumption and random store visits in each period, J = ζ σ. Each time a consumer visits a store to purchase the product, a fixed shopping cost, χ, is incurred, consisting of the travel cost of a trip and transaction cost of time spent inside a store. Costs per unit of inventory carried by consumers at household each period are defined as ψp, where ψ is the percentage of purchase cost over the planning horizon. This inventory cost percentage is assumed to be proportional to the individual time value of money and space the product occupies at household storage, and it is inverse proportionate to the average storage period of a consumed product at home before expiry. The retailer's local store serves a fixed number of consumers who reside in proximity so that their shopping and inventory costs remain constant. The consumer is willing to tolerate a temporal shortage as long as it contributes to the long-term objective of cost minimization. A consumer thus incurs individual shortage cost per unit, η, in each period the product is missing from consumption in the household. Έ is the expected number of shortages per period, which is based on retail inventory's service level in previous periods. The shortage cost comprises a constant value of physical disutility from non-consumption and psychological regret of not having something in stock, which strongly increases during the pandemic due to anxiety about future availability and pricing. Toilet paper is a widely publicized example of such popular products providing comfort to people in times of pandemic uncertainty (the total consumption itself is not likely to show a significant increase, as the literature review discussed). When the pandemic starts, consumer adjusts purchase quantity based on a rational belief about changing shortage cost due to heightened regret experienced after stockout. It is reasonably assumed no lockdown or other restrictions during the pandemic hampers the ability of consumers to visit the local store. The retailer's supply chain is assumed to be capable of restoring an acceptable inventory level after a certain period of adjustment. Therefore, after the initial uncertainty about product availability, the consumers' shortage cost decreases but still does not go back to pre-pandemic levels as consumers adjust their beliefs based on the observed retailing situation. Admittedly, consumers' disutility due to shortage could be a function of various factors such as beliefs proportional to the infected population curve. It would likely be a continuous one resembling the Bass Model diffusion equation as a function of time. However, the shortage parameter is assumed to be fixed in the main model for simplicity of exposition. Thus η2≥η3≥η1. This assumption is reasonable and critical for the main model. Analysis of data at the beginning of the pandemic suggested the surge in consumer demand for the popular demand currently could be a one-time event with declining oscillation. However, researchers were aware the second wave of the pandemic could start as early as autumn 2020, repeating the earlier stockpiling shock (Placer Labs, 2020). Though average monthly sales of the popular products increased up to 20% over the comparable figures for previous periods, online sales could increase as much as 300% during a week relative to the last week (BCG, 2020; Nielsen, 2020). It is those unusual hikes in day-to-day purchases that should attract the interest in modeling panic buying. The concern about shortages seems to be the main driver of hoarding behavior among consumers. Figures 1 and 2 appear to show the empirical support for the assumptions made about the retailer and consumers in the main model.Fig. 1 Percentage change in sales of the popular products after the start of the pandemic in Italy (Source: BCG, 2020) Fig. 2 The actual change in online sales of FMCG products after the start of the pandemic in Russia (Source: Nielsen, 2020) To isolate studied effects, additional simplifying assumptions are made about the retailer and consumer. The business objective of the retailer is the target service level of inventory. Such an objective is widespread due to its ease of implementation and beneficial for achieving a particular market share and customer satisfaction. Backorders are possible at unit shortage costs. Lead time to deliver a product from suppliers is positive and variable. RQ is a continuous review policy. The choice of RQ for the popular products, as indicated in Table 1, seems justified given the relatively low holding cost of such products as opposed to order-up-to and myopic newsvendor-based inventory policies that are more suitable for perishable and fashion products which have high spoilage and obsolescence rates. Suppliers impose capacity limitations. There is no variable purchasing cost for supplies, and pricing is fixed as markup. Costing parameters in the model do not change during the pandemic, which seems to be consistent with most popular products except for cases of severe disruptions, such as in the case of masks. Though it is assumed that the retailer does not change the price in the main model, a model extension later discusses the implications of the belief in future changes in pricing. One positive aspect of retail immediately after the pandemic in major countries was that prices did not seem to change significantly for most popular products (Office for National Statistics UK, 2020). There was, however, a long-term increase in global consumer price inflation due to the supply chain crisis as a result of increased maritime transport costs; it still remained below ten percentage points as of 2022 data (Grynspan, 2022). In this study, supply disruptions are incorporated into the variability of lead-time with a normal distribution. An alternative way to model disruptions in supply disruption literature is to present them as a random process governed by the probability distribution of the number of consecutive periods with disrupted supply (Schmitt et al., 2015). The model of uncertain lead-time in this research is widely described in supply chain textbooks and easier to use. The pandemic leads to an increase in lead-time variability after the start of the pandemic. Other parameters are assumed to remain constant. Notation for all periods before the pandemic (pre-pandemic phase) is t = 1; for the first period right after the start of the pandemic outbreak (pandemic-start phase), it is t = 2; and for all the periods after (after-pandemic phase), it is t = 3. The duration of phase 1 in the model is assumed to be long enough for the surge in consumer demand to occur after the pandemic outbreak before deciding on adjusting inventory policy. Analysis of demand disruptions due to stockpiling To calculate the reorder point in RQ, safety stock and order size should be defined first. With variable lead times, safety stock to achieve the desired service level1 B=ZL·J2+M2·U2 Z is a safety factor that depends on the inventory service level (probability of satisfying demand during lead-time). Safety stock could be negative (retailer would hold less inventory than average demand) if the service level is extremely low, which is excluded as an unrealistic case for real businesses. Supply disruptions in production and transportation due to the COVID-19 effect of increasing lead-time variability implies U2 ≥ U3 ≥ U1. The previous sections illustrated how lead times for certain popular products could go back closer to their normal pre-pandemic levels after an initial period of adjustment. Unfortunately, data as of 2022 for the years following the outbreak show the general tendency towards lengthening the lead-times: containers typically spent 20% more time in the system for door-to-door trade, with ships and trailers stuck in congested ports (Grynspan, 2022). From the expression above, the impact of simultaneous supply disruption and a sudden increase in demand can cause a "perfect storm" for retailers after lead-time variability becomes significant due to pandemic outbreaks. The scale of such disruptions for countries with high levels of offshoring could be enormous, as in the case of US businesses that source anywhere from 3% (Nordstrom) to 60% (Best-Buy) from China (Thomas, 2020). The typical approach to finding Q in RQ policy is to use the well-known EOQ model, which gives a reasonably good approximation of the optimal order quantity for practical use (Hillier & Lieberman, 2004). The optimal purchase quantity minimizing total relevant cost in EOQ with the deterministic assumption is conveniently derived with the first-order condition as the objective function is convex.2 Qrq=2OMNH The use of EOQ is nearly optimal in minimizing the retailer's long-term holding and ordering costs, CRQ. Then it is straightforward to calculate reorder point with the target service level:3 RRQ=LM+B Here, the reorder point is expressed as a sum of average demand during the lead time plus safety stock to ensure the target service level of inventory to protect against uncertainty. It is not difficult to see why maintaining the pre-pandemic levels of availability of the popular products for retailers is a challenging task: with soaring demand for items such as toilet paper, reorder points would become inadequate. To predict how consumer demand could change, a separate model has to be considered. A rational consumer minimizes the expected total cost of consumption per cycle by balancing inventory and shopping costs in the selection of purchase quantity and reorder point before each store visit is:4 Ccons=‵EημQcons+ψpQcons2+Rcons-μ+χμQcons The first term in the expression is expected shortage cost; the second term is expected inventory cost; the last term is expected shopping cost. The optimal decisions can be defined as follows.5 Qcons=2μχ+‵Eηψp 6 FRcons=ημημ+μψp F(Rcons) is the cumulative distribution function of individual consumer demand. The exact solution for both decision variables can be found using an iterative procedure, but an analogous approximation of EOQ planned shortages provides a reasonable heuristic for fast calculation. The result is equivalent to RQ model extensions with planned shortages and cost minimization targets. Such similarity is not surprising given analogous parameters and results in related models of consumer behavior that this study builds upon (Ho et al., 1998). When consumers' perceived shortage cost η increases with ψp remaining constant, consumers adjust purchase quantities increasing both Qcons and Rcons in (5) and (6). Retailer then increases Rrq in (3) since M = μ ζ increases proportionally to η, and also Rrq increases in J. Opposite direction of change after the initial pandemic period with decreasing shortage cost is derived similarly. Given all the input parameters and assumptions, the following Proposition summarizes the effects of hoarding on the supply chain. Proposition 1 Consumer demand and retailer's reorder points for a popular product increase immediately after the start of the pandemic and then decrease but do not go back to the pre-pandemic level. Hoarding unnecessarily expands inventory and costs in the supply chain for both retailers and consumers even when the actual consumption rate does not show a significant increase. A couple of interesting implications of Proposition 1 should be discussed here. First, soaring consumer purchases of the popular products might render the pre-pandemic service level of inventory infeasible. At the same time, low levels of availability could hardly be acceptable for the senior management and local communities. Then the retailer could consider a scenario where each store introduces a buying limit for a certain number of substitutable products per customer. It is assumed that the business can enforce the limit long-term, implying either disciplined consumers or perfect tracking of each purchase over the entire period. If the limit can be effectively set per consumer, then order quantities for both the retailer and consumers effectively remain constant at their pre-pandemic levels. Then neither the retailer nor consumers have the motivation to change the corresponding inventory policy and shopping behavior after the pandemic. Second, if there was a longer lead-time, as in the case of consumers' shift to online shopping due to concerns about availability or social distancing, the reorder point would increase further and lead to considerable shortages at e-commerce facilities. An example of Amazon struggling to deliver on the promise of quick delivery during the pandemic even to its Prime customers is a good illustration of such a situation. When η increases at the beginning of the pandemic with ψp remaining constant, consumer adjusts purchases increasing both Qcons and Rcons. The retailer then reactively increases RRQ since M increases proportionally to η, and, besides, RRQ increases in J. Both the retailer and consumers make those adjustments based on the most recent market signals and allow a high likelihood of sustained long-term changes. When η decreases, another adjustment has to be made. This implies suboptimal purchasing and ordering policies that lead to Ccons2≥Ccons1;CRQ2≥CRQ,1. Thus the costs of pandemic disruption can be summarized by Proposition 2 as follows: Proposition 2 Change of buying and ordering policies during the pandemic is costly for consumers and retailers, correspondingly. Indeed, dramatic developments in the global supply chain management happening in the three years after the pandemic outbreak illustrate the scale of some of the changes stated by the aforementioned propositions: retail supply chain disruptions provoked the widespread shift from the just-in-time inventory to just-in-case inventory build-up (Shih, 2022). It should be noted that buying patterns for popular products could differ from the general consumption trends. In the U.S., consumers’ spending initially had decreased relative to its pre-pandemic levels but showed a gradual increase afterwards (Elmassah et al., 2022). The lead-time variability will remain a serious issue as the pandemic does not appear to reverse the trend of the global sourcing any time soon (Koerber & Schiele, 2021). Alternative inventory policies and stockpiling Outside RQ with a continuous review, a wide range of inventory policies can be derived from a periodic review approach with maximum and minimum levels. sS policy is a periodic multi-period type that can be used by a retailer as a popular alternative to RQ policy described in the main model. When a significant fixed setup is present for each order and delivery schedules together with inventory review are periodic, sS policy can be preferable to RQ Total lead-time in sS would comprise of separate delivery time, Ltransport, plus review period, Lreview. Hence safety stock is larger in sS due to extra allowance for review. The optimal inventory policy with sS is to bring the inventory level up to S if the inventory falls below s level, and order nothing otherwise. It can be defined as follows:Qss=S-IifI<s0ifI≥s where I is the initial inventory at the beginning of each period, s is the minimum level triggering order placement, and S is the maximum up-to level derived in a manner similar to RQ with target service level objective, S=ML+ZJ. Retailer adjusts the policy in a range 0 < s < S. With sS policy, the analysis gets more involved as there is no straightforward method for determining optimal s level, though the safety stock calculation is similar to the one in RQ policy. If s is made equal to S, then the inventory policy is implemented in the same manner as in the order-up-to model, which can be considered as a simplified case of sS widely used in retailing practice. When order setup cost, O, gets insignificant, order-up-to is a widespread policy in retail known for its convenience of use with the periodic review. Myopic newsvendor policy can be considered a further simplification of an order-up-to policy when inventory cannot be transferred between periods (I = 0) due to expiry, maintaining freshness, and similar concerns. Overall, sS and its derived policies are widely used alternatives to RQ, and they could be more relevant for perishable products with high spoilage rates rather than popular products. The responsiveness of sS and similar policies is lower than with RQ because of the additional time, Lreview, required for replenishment between periodic reviews. When the pandemic starts, a retailer without continuous inventory review will be slower to adjust S to maintain the target service level of fast-moving inventory. Considering the characteristics of the discussed inventory policies, the following Proposition can be formulated: Proposition 3. Retailers' selection of sS, order-up-to, and myopic newsvendor policies, in comparison to RQ, leads to: (i) lower availability of the popular products; (ii) higher shortage costs in a setting of increased stockout penalty. sS appears to be a less suitable policy for retailers than RQ during the pandemic due to a higher shortage rate with more time needed to adjust inventory. Still, sS would remain an appropriate policy in many instances, mainly when a continuous review is challenging or long enough review periods are mandated by scheduling systems. Risk pooling and panic buying As extensions of the already presented main model, the following two subsections further discuss how various costs related to the bullwhip effect and panic buying have to be incurred by the businesses and consumers in addition to those outlined in Proposition 2. Mitigating bullwhip effect through substitutability and risk pooling In addition to the cost of a suboptimal inventory policy due to hoarding, retailers could incur additional expenses due to higher fluctuations in the end demand. The bullwhip effect can be defined as the ratio of the coefficients of variation of orders, upstream CV to downstream CV (the alternative measure is the ratio of variances). In the case of the simple retailer-consumer supply chain in the main model: Θ = CVRQ/CVcons. The extra costs of order setup and irregular delivery associated with increased variability due to the bullwhip effect are assumed to linearly increase in the ratio of CV: Bbullwhip = ʎ Θ. The value of the ʎ coefficient is challenging to quantify, as previous studies show, and it is outside of the scope of this research. It is enough to assume the bullwhip effect costs are likely to be amplified during the pandemic as variability shock reverberates at the upper stages of the supply chain among suppliers. Limited empirical evidence suggests the bullwhip effect before the pandemic has been moderate or non-existent for many popular products at the upstream stages of retail supply chains across various industries (Cachon et al., 2007). It is assumed for the focus of this analysis on demand signals that suppliers do not practice variable pricing and rationing, two essential causes of bullwhip (Lee et al., 1997). It is evident from Proposition 1 how stockpiling would increase the bullwhip effect when the variability at the lowest downstream level of end consumers hikes. In a sense, the negative effect of consumer stockpiling is similar to a combination of demand signal processing and rationing with anticipation of shortages in B2B (Lee et al., 1997). When the upstream stages in the supply chain network allow sufficient substitutability between suppliers, the harmful bullwhip effect could be reduced due to a sort of risk pooling (Sucky, 2009). One known implication of risk pooling is that stockout-based substitution of consumers facilitates inventory pooling across products and locations (Yang & Schrage, 2009). Risk pooling can help reduce safety stock (Eppen, 1979). Having substitutable products and alternative sites is more likely to reduce total inventory in a system when the following conditions are met (hereinafter, positive pooling conditions): (1) unit overstock cost is sufficiently high relative to the unit shortage cost; (2) positive skewness of the demand distribution is not strong; (3) the pooling effect is medium to high (Gerchak & Mossman, 1992; Yang & Schrage, 2009). Such conditions in a retail hold in a wide range of settings as availability policies demand low shortages. As the levels of substitution of demand increase, the demand pooling starts approaching the full pooling effect. In this model extension, there are k stores and n substitutable products with horizontally differentiated quality (the results of the subsequent analysis would be qualitatively similar for vertically differentiated albeit with less direct effect). Substitutability here could mean different brands of the same popular product horizontally or vertically differentiated (for instance, national and store brands of coke). But it could also mean different product categories that could partially substitute one another in consumption (for instance, toilet paper and paper towels). In the case of stockout, γ portion of consumers chooses to switch to another store. Alternatively, β portion of consumers substitutes product brands or categories. It is assumed for tractability that all stores and products are the same in terms of inventory policy, costing, and substitution parameters. The demand is a random variable, while lead-time is constant and equal to one in this model extension. Demands across stores and brands are correlated with coefficient ρ. When there is no stockout-based substitution, the total safety stock in the retailing chain is determined as:7 Btotal=Z∑1kJj In the case of demand substitution (store switching), the constant portion of switching consumers is assumed to be uniformly distributed among all locations. Then the total variance of retailing chain can be separated into pooled and non-pooled parts, so the total safety stock with demand substitution can be defined as follows:8 Bsub=Z∑1kJi21-γj+∑1kJi2γj2+2∑1≤i≤j≤kkγiγjCovMi,Mj As γ → 0, (8) becomes equal to (7). As γ → 1, the variance of total demand in (8) represents a sum of correlated random variables with the same variance and mean. From (7), total variance is non-increasing in γ. When the total system variance decreases, the retailer needs less safety stock to keep the same target inventory level. The ratio in expression for B decreases in the variance of the downstream stage. Consequently, both B and Q (or S) are non-increasing in γ. Results of risk pooling with product variety will be analogous if many substitutable product brands/categories are used instead of stores (β > 0). Based on the above analysis, the effect of demand substitution on inventory and variability of orders can be summarized in Proposition 4 as follows: Proposition 4 Assuming positive pooling conditions, safety stock and bullwhip effect in the supply chain are non-increasing in the level of store and brand switching. In a sense, the pooling of locations and products has different implications from classic risk-pooling models: while the number of facilities should desirably be reduced in standard settings, stockout-based substitution favors maintaining a sufficient number of locations and variety for consumers to switch between them. Risk pooling has favorable implications for competition and product variety as switching between shopping locations and brands can reduce the negative consequence of consumer hoarding. Demand substitution thus makes it easier for consumers to switch between products, stores, and even competing retail chains (further increasing β and γ). It helps reduce retailers' inventory costs while contributing to societal benefits with higher availability of popular products during the pandemic. The risk pooling effect would be most substantial with negatively correlated demands, which is unlikely during the general change in consumer demand during the pandemic. It has other limitations. The risk pooling would be less effective with fashion-like (myopic newsvendor) settings when a very high shortage or spoilage costs prevent inventory transfer to subsequent periods, limiting pooling opportunities. Queues inside stores and unnecessary visits to locations with insufficient stock should be minimized due to social distancing. The retail policy should restrict unnecessary switching and return visits while exploiting the positive effects of risk pooling. Though brand and store switching would positively address the average needs of consumers' physiological utility, they could also amplify panic buying behavior discussed in the following subsection due to the massive number of buyers changing stores. Negative impact of panic buying strategic consumers In this extension of the main model addressing panic buying, the behavior of strategic consumers follows standard models of forward-looking behavior in retail (Aviv & Pazgal, 2008; Su & Zhang, 2008; Cachon and Swinney, 2011; Swinney, 2011). In this original model setting previously applied to fashion products, there are two prices: full regular, p, and a changed (usually reduced) price, v. Consumers decide whether to buy now or later given the price difference, reservation price (equivalent to consumption utility or maximum acceptable price), u, a discount of future consumption, r, a belief about the probability of the product being in-stock later, δ. Strategic consumers have a choice of buying now and getting a consumption surplus of u1—p, or waiting to buy later but getting uncertain utility of future consumption, δ r (u2 − v). In the classic equilibrium with rational expectations widely applied in related supply chain models, a retailer chooses price and inventory levels maximizing the expected profit, given that homogeneous consumers purchase the product at the regular price, (q, p*) = argmax q,p π (q, p). The alternative equilibrium where consumers wait to purchase a product at a reduced price is not feasible under the typical pricing path. Unlike in fashion supply chains, the effect of strategic consumers in pandemic times is opposite to the usual impact of waiting for future bargains. While highly relevant for fashion retailers, the model of forward-looking behavior has been insignificant for grocers that less frequently discount prices for food and other popular products, while consumers exhibit a low discount of postponed consumption except for limited cases of items with early expiry. The perception of future shortages and price hikes experienced by panic buyers makes the model of strategic consumers relevant to retailers during the epidemic. Unlike in the fashion business with significant future discounts in both price and consumption value, the situation gets flipped here: consumers anticipate future price increases due to shortages. Therefore, the implications of the strategic consumer presence would be different for the popular products during the pandemic. However, in this research, motivating the consumer to postpone purchase is preferable because the retailer is less concerned about being profitable but instead focuses on keeping stable sales targets. In this sense, the focus of retailers on sales is beneficial for society during the pandemic. The consumer reservation price can include two components, psychological, upsy, and physiological, uphy, utilities of consumption so that the total is: u = upsy + uphy. Setting reasonable revenue targets instead of profitability matches the maximization of physiological utility among consumers. First, consider pre-pandemic periods where the consumers' reservation prices, u, are uniformly distributed across time, while availability, r, and pricing, p = v, do not change from period to period. In such a scenario, the consumers will be split more or less uniformly between each period when a discount of future consumption, δ, is negligible (equal to one), which is likely to hold for the popular products. While uphy does not change during the pandemic, upsy is likely to strongly increase even though it is not essential for the actual health and living needs of consumers. Perceived shortages imply r2 < r1. During the pandemic, higher price expectations for the popular products would likely become prevalent, v > p. The current utility satisfying u1 − p must then be minimum or even negative for any strategic consumers to not buy now because of the spreading panic during the pandemic, even though the immediate consumption might not be preferable to future one. Since this is a less likely outcome, most purchases will be made in the current period instead of postponing purchases. The following Proposition summarizes the effect of panic buying. Proposition 5 Panic buying of strategic consumers based on prevalent beliefs about availability and pricing during the pandemic leads to an increase in the current period purchases at the expense of purchases in the future period. Coupled with increasing price expectations, the changing utility of consumption could explain widespread panic buying, which is an irrational behavior as opposed to hoarding discussed previously. Indeed, the distorted risk perception during the pandemic could trigger panic behavior among consumers (Elavarasan & Pugazhendhi, 2020). Thus retailers must somehow assuage panic buying behavior by guaranteeing that the regular price and availability remain constant. Therefore, in the preferable equilibrium during the pandemic, the retailer encourages strategic consumers to believe that pricing would stay the same or only change downward. The problem here is that rational consumers have all the reasons to believe in non-decreasing the availability of the popular products at times of supply chain disruptions on a global scale. Furthermore, rational belief will be about increasing prices because even if businesses or governments impose limits, shortages can lead to buying at higher prices from resellers. The most effective remedy in such a situation still seems to be the buying limits. By adjusting maximum purchase for more even distribution of consumption in each period at which all consumers with the varying reservation prices are guaranteed to purchase a product at a fixed price, panic buying behavior could become irrelevant. It is acknowledged in this as well as previous models of strategic consumers that assumptions about consumer utility and equilibrium conditions are somewhat restrictive to apply in price and inventory optimization for retailers directly. Nevertheless, the modified model of forward-looking behavior of panic buyers could be valuable for a better understanding of the motives behind the panic buying aspect of consumer stockpiling. Numerical examples of heterogeneous consumers and demand for the popular products If the assumptions made in previous sections on homogeneous consumers with uniform reservation prices and consumption rates are relaxed. In that case, the findings of the main model do not change qualitatively, as numerical illustrations in the next section show. In this section, the theoretical and empirical justifications for parameter choice in the subsequent simulations should be discussed first. Consumers are far from being homogeneous in response to the pandemic. For instance, only 5% of Russian consumers admitted stockpiling food during the pandemic in March 2020; but 18% indicated stockpiling hygiene products (Nielsen, 2020). Senior consumers in Russia had stocked 126 consumption days of the popular products compared to 108 days for the average household (Eпaнчинцeв, 2020). The share of consumers stockpiling food ranged from 6% in Japan to 42% in China (Ipsos, 2020). The distribution of individual shortage cost that influences consumer demand in the simulation is assumed to follow gamma probability. Shift in demand uncertainty due to pandemics can be governed by the gamma distribution. The ability of gamma distribution to exclude negative values and approximate a wide range of demand patterns, including the most common normal and exponentially distributed ones, with varying shape and scale parameters is a significant advantage for inventory control (Burgin, 1975). Another reason for choosing gamma probability for modeling demand during a pandemic is its suitable density function with positive skewness. Its demand probability density can have a heavy right tail, which better represents a shift in consumer valuation than conventional symmetric distribution. Figure 3 describes a change in the shape and scale parameters of the gamma distribution.Fig. 3 Distribution of consumer shortage cost before (left) and after (right) pandemic In the following simulation illustrative of the main results in this research, the retailer owns a store serving one hundred local consumers. Lead-time is one week, L = 7, and its standard deviation is assumed to be zero before the pandemic, and afterward increasing function of global supply chain disruptions proportional to new coronavirus cases worldwide up to one week later: U = JƷ/Ƹ, where Ʒ is new daily actual cases and Ƹ is the stable average of new cases in May 2020 (Fig. 4). The target service level of inventory in RQ policy is F(Q) = 95%. Other retail parameters are the order setup cost O = 10 and unit holding cost per period H = 0.2. As for consumers, inventory unit cost ψp = 0.1; shopping cost χ = 4, and unit shortage cost, which is an increasing function of the local, new cases (only in Europe) and assumed to be: η = 10 JƷ/Ƹ. The consumption rate, μ = 1, is one unit daily with standard deviation σ = 0.1, for all consumers. Due to random customer arrivals, the final demand from customers for the retailer's inventory policy would still appear stochastic following an approximated distribution with aggregated mean M and standard deviation J. Retailer updates inventory policy based on ten-period moving average. There is no capacity limitation in supply, so the simulation does not apply to the case of masks and similar popular products missing on shelves during the pandemic. The modeling time framework from February 30th to May 30th of 2020 is divided into three phases of equal duration into periods, each representing the distinct stages of disruption in the model. It should be noted here that all the parameters and functions in this simulation are arbitrary to a certain extent but still represent one possible scenario among many other realistic or potential ones.Fig. 4 Simulation result for retailer's order quantity, reorder point, and inventory level before and after the pandemic relative to new coronavirus cases (nuber of infections taken from European Centre for Disease Prevention and Control 2020) From Fig. 4, the reorder point increases considerably at the start of the pandemic, while the order size remains fixed. Due to lead-time lag, the inventory hike is observed later after the reorder-point adjustment: the highest risk of shortages and plummeting fill rate can be anticipated at this time point. Figure 5 summarizes the general change in supply chain performance. As consumers stock more at the beginning of the pandemic, but their consumption rates do not change, sales decrease after reaching a peak in the second period. While consumers experience only a slight increase in cost, the retailer incurs higher costs due to soaring safety stock. Actual costs could be higher due to the increasing bullwhip effect.Fig. 5 Comparison of sales, costs, bullwhip effect, and days of supply (for consumer consumption) between periods Limited empirical evidence indirectly supports the simulated patterns of changing sales and inventory: volatility of both spaghetti sales and corresponding stockouts at US supermarkets sharply increased after the COVID-19 pandemic started, but the general trend had been the highest increase immediately following the outbreak, which gradually decreased later towards the pre-pandemic levels (Taylor et al., 2020). Other cases (Figs. 1 and 2) exist that suggest similar patterns. Discussion This section separately discusses the implications of the research for retail businesses and regulators. Implications for businesses and society This research suggests restrictions on the number of a product in high demand that each individual can receive during the pandemic helps reduce the stockpiling effects on retail. One possible problem with buying limits could be customers immediately returning to a store to buy over the buying limit. This problem could also exacerbate the social distancing situation if a majority of consumers responded that way. A simple solution could be to track such consumers using their credit/debit and loyalty cards, but this would raise privacy issues and potential discrimination against cash-only buyers. Retailers with appropriate capabilities such as Amazon could bypass such limitations by using information technology such as AI to identify customers returning to the same store to buy over the buying limit. Even without face recognition and other AI applications, the modern omnichannel retailers could require customers only to buy the most popular products using a buy-online-and-pickup-in-store option in advance before a store visit. This increasingly popular method of omnichannel fulfillment could also help minimize in-store time while supporting social distancing as many retailers maintain separate counters for in-store pickup. Furthermore, websites and shopping apps of omnichannel retailers or independent aggregators could help prevent unnecessary customer visits by showing real-time product availability. In theory, omnichannel retailers could switch to contactless delivery of the popular products from stores during the epidemic contributing to the societal goal of social distancing. Unfortunately, as the example of even the leading firms such as Amazon shows, the logistics capacity needs substantial expansion to meet consumer demand during the pandemic. Another underestimated benefit of omnichannel fulfillment is increased demand substitution, as previously revealed by Ovezmyradov and Kurata (2019). Product substitution during panic buying periods can increase profitability and customer satisfaction (Tsao et al., 2019). The extension of the main model on the bullwhip effect in this research indicates brand and store switching allows reducing the harmful effects of variability during the pandemic. Real business cases exist that illustrate how a limited product variety (single brand of dry yeast available by a grocer) or difficulty of switching (commercial and consumer brands of toilet paper) exacerbates the issue of availability during the pandemic; in contrast, interchangeable products (beer, soda, and snack food) across channels allow quick response to supply chain disruption (Taylor et al., 2020). The omnichannel approach was already becoming more relevant in the post-pandemic world, where consumers are likely to keep the online shopping behaviors adopted after the outbreak (Denise, 2020). This research provides extra findings supporting investment in omnichannel fulfillment enabling mobile-responsive sites, "buy online pick up in store" services, and consistent digital experience across channels. Switching from periodic to continuous review suggested in this research could be facilitated by rapidly developing technology such as item-level RFID tracking and AI. Closer coordination with suppliers could be required. Supply chain coordination becomes particularly important in conditions of supply-side disruptions. Cost-sharing, two-part tariff, revenue sharing, quantity discount, wholesale price, and other types of contracts are examples of supply chain coordination tools extensively analyzed in the past and recent literature (Hendalianpour et al., 2020; Liu et al., 2020; Qian, 2020). Such agreements ensuring stable pricing between suppliers and retailers could reduce the harmful effect of pricing affecting panic buying as described in the extension of the main model incorporating strategic consumers. The main model analysis suggests a sharp increase in consumer demand at the earliest pandemic stages, so it could be tempting for retailers to invest in additional capacity. While previous studies on the bullwhip effect almost exclusively focused on supply-side variability, the current crisis is distinguished by increased demand variability of unprecedented scale, which reverberates back to upstream stages of the supply chain. Overcapacity is wasteful not only for business but also for the economy and society. Changes to the planning process and time horizon could be more effective than increasing capacity (Dohmen, 2022). In the context of the findings of this research, businesses should be careful about excessive long-term capacity in anticipation that the current surge in consumer demand is sustainable with most high-demand items. This will help avoid the costs of excess capacity and underutilization in the long term. Implications for regulators and business-government cooperation This section discusses further implications of the research analysis for policymakers. Politicians might compare the epidemic to warfare and consider extraordinary measures. One long forgotten measure that governments could consider for items of extreme shortage and importance, such as masks at the early stages of the pandemic, is introducing a ration stamp or card. Such rationing was widely used in the UK and US during and immediately after World War II. In other major countries such as India and the former USSR, ration cards were widely used long after the war. Public and private funding of the promising new technology should address the priority areas as follows. Mobile applications. blockchain, 3D printing, artificial intelligence, digitalization, and related technology have become increasingly important for supply chain resilience and insights during the recent disruptions due to the pandemic (Elavarasan & Pugazhendhi, 2020; Belhadi et al., 2021; Gupta et al., 2021; Queiroz et al., 2021; Kronblad & Pregmark, 2021; Ye et al., 2022). Innovative concepts were proposed in the last-mile delivery to tackle disruptions in the supply of essential items, such as the truck-drone systems in the areas of severe infections, dysfunctioning warehouses, labour and truck driver shortages (Singh et al., 2021). Although users likely have not yet built a clear attitude towards autonomous delivery vehicles, the new technology could become essential in influencing lead times during the pandemic (Kapser, 2021).There is theoretical and empirical evidence of the importance of supply chain visibility supported by information systems during the pandemic (Yang et al., 2021). The numerical example in the previous section used an arbitrary function of disruptions and shortage costs with respect to the number of infections. Recent studies explored empirical relationships between disease transmission, social distancing and community mobility utilizing data: modeling, deep learning, historical pandemic data and mobility control (Chen et al., 2021). However, published common decision support tools and dashboards focusing on individual effects of a non-pharmaceutical intervention on health and the economy lacked visualizing the multi-criteria challenge (Tolk et al., 2021). The use of advanced technology should not be limited to retail. Early warning systems, possibly involving big data, should help governments identify early signs of the pandemic so retailers could be informed to take corrective actions proactively. Clear labeling and announcements of buying limits and availability are essential. Firms having overstock should be ready to indicate it to potential buyers and utilize transshipments effectively to compensate for shortages in other locations. The effectiveness of marketing campaigns to educate consumers and raise awareness of stockpiling harm cannot be overestimated. For instance, targeted ads could inform potential panic buyers via social networking that the maximum amount of house hoarding is limited by storage capacity and natural consumption. This approach could narrow the gap between real physiological and perceived psychological utilities for popular products, such as toilet paper. Again, governments could assume a leading role in facilitating such communication if businesses do not deliver products due to competition, cost, and other concerns. Notably, supermarkets, logistics providers, and suppliers were allowed to coordinate by competition watchdog in Australia in order to ensure supply at a fair price (Siebert, 2020). Whether governments should play a more prominent role in control over consumer stockpiling is debatable yet. Particular care should be taken in regulating a purchase quantity limit in view of its pitfall (Carlson, 2021). Relative retail price regulation not exceeding a fixed ratio of the wholesale price can, at least in theory, be more effective than absolute price regulation, as it provides a production boost effect for the supply chain without stimulating consumers’ hoarding (Li and Dong, 2021). Furthermore, more policy support could be provided to many companies being in the early stages of digital technologies, with low levels of data and a gap between digitization and implementation to improve supply chain performance in the COVID-19 crisis (Ye et al., 2022). While the results of this study on preventive measures could be applicable for most products facing stockpiling during the pandemic, there seems to be no immediate fix for disruption in the supply of special goods suddenly becoming scarce such as products of hygiene (masks and ventilators) before the global supply chain adjusts. The situation with the masks was unique due to the simultaneous effect of soaring demand from medical institutions and skyrocketing consumer demand at a time when the bulk of masks have traditionally been supplied from remote locations in global supply chains, mainly China. This is where governments and businesses should cooperate to find a long-term solution Governments could provide incentives for local production of the product now deemed strategic after the outbreak. Meanwhile, retailers could introduce new innovative solutions such as multiple-use masks and matching disinfection tools. Overall, timely communication of the following product data between businesses and consumers and regulators could play a crucial role in preventing negative consequences of stockpiling and unnecessary store traffic during a pandemic: availability, location, quality, substitution alternatives, and delivery methods. Implications for relief supply chain, and humanitarian logistics The importance and insufficiency of studies on supply chain disaster relief management and developing scales for resilience to supply chain disruptions have been highlighted during the COVID-19 disruption to operations experienced in a crisis-as-a-process context (Pournader et al., 2020). How suppliers and manufacturers together cope with the disruptions due to the pandemic inevitably influences the interconnected retailing, global sourcing, and relief supply chains. JD.com is a relevant example of a leading online retailer in China successfully delivering emergency supplies to individuals and hospitals in the heavily affected regions during the pandemic in collaboration with hundreds of partners across and beyond supply chains (Shen & Sun, 2021). Demand substitution is relevant for B2B settings too, and the following discussion mainly focuses on the implications of the study for partial risk pooling across upstream stages of disrupted supply chains. Supply chain disruption studies before 2020 primarily focused on the consequences of natural and humanitarian disasters for businesses' supplies. This interest was spurred by catastrophic events such as earthquakes that disrupted the supply of critical parts in the lean supply chains of modern car makers. In response, global leaders in the efficient manufacturing, such as Toyota, the pioneer of lean manufacturing, which traditionally preferred working closely with a few key suppliers, took steps to diversify their supply chains (Shih, 2022). There are other empirical and modeling studies demonstrating the substantial costs of overly relying on just-in-time or the primary supplier during disruptions (Sanci, 2021). The same case for redundancies in the supply chain received extra support in the theoretical and empirical literature on substitutability. Intertwined supply networks as the entirety of interconnected supply chains can help ensure resilience during coronavirus outbreaks (Ivanov & Dolgui, 2020). When supply can be disrupted by stochastic and especially deterministic demand, a decentralized system design with more facilities holding inventory could be preferable for firms (Schmitt et al., 2015). The diversity of suppliers, locations, and inventory contributes to a favourable redundancy that helps cope with post shocks in food systems (Hecht et al., 2019). There is a case study available on how substituting sourcing helped an equipment manufacturer as one of the main adaptation strategies to navigate the COVID-19 outbreak (Ivanov, 2021). Many manufacturers during the pandemic quickly switched to producing medical supplies (Elavarasan & Pugazhendhi, 2020). Globally, substitution supported the case for switching to wider use of domestic industries for capacity availability in addition to imports of essential products (Corominas, 2021). In fact, the perceived risk of COVID-19 could drive consumers to embrace locally produced food (Palau-Saumell et al., 2021). More broadly, substitution could play an important role in the process modularity of international humanitarian organizations speeding up the emergency order validation relative to the regular order validation (Salah et al., 2022). High product diversity together with the central position in the supply chain helped improve the operational resilience of some companies during the COVID-19 pandemic (Li et al., 2022). However, the revealed benefits of substitution do not always imply increasing product variety or capacity. In fact, a prioritization with a lowered number of SKUs to reduce changeover times could be a major part of reconfiguration by a non-perishable foods manufacturer in response to the pandemic (Dohmen, 2022). Meanwhile, surplus inventory and capacity in groceries can be utilized with the cooperation between retailers and food banks during COVID-19 crisis (Penco et al., 2021). To summarize, partial risk pooling to some extent favours higher capacity and a variety of substitutable products across all supply chain stages for relief supply during the pandemic. Alternative approaches and research The subsection of this research is related to socio-psychological and other effects of the pandemic on consumers and retailers. The measures discussed so far are by no means exhaustive that have or could be proven to be effective. It should be noted that, for instance, recent psychological studies suggest altruistic personality curbs consumer stockpiling (Johnson, 2020). Consequently, social shaming of excessive stockpiling and reselling could be a potent incentive to act in the interest of society. Another example is a sharply rising price for each extra unit purchase of the popular product. Social media such as Twitter can also become an attractive tool for disaster management, with problems and solutions shared in real-time (Kumar et al., 2021; Singh et al., 2019). Furthermore, machine learning and tweets were used for capturing consumers’ COVID-19 sentiments (Schlegelmilch et al., 2022). Attitudes towards online shopping, virtual meetings, and other teleactivities have to be studied in the context of communication issues during COVID-19 (Mouratidis & Peters, 2022). Consumer responses to COVID-19 restrictions involved complex psychological factors as evidenced by the “No Vax” and “No Green Pass” movements (Matarazzo & Diamantopoulos, 2022). Retailers can implement corporate citizenship campaigns thus reducing the consumers’ fear of COVID-19and hence its negative effects such as panic buying (Arachchi et al., 2022). Meanwhile, COVID-19 crisis could have increased the demand for reliable information alongside a significant use of data provided by public organizations (Dreisiebner et al., 2021). The need for transparent interdisciplinary studies of the COVID-19 has been driven by the concerns about an exclusive focus on one criterion resulting in new problems for others. For instance, while biomedical professionals could suggest the shutdown of facilities to minimize the contact rate, social science could raise concerns about a panic reaction with fears of attending life-saving services, and economists warned about associated financial issues (Tolk et al., 2021). Human resource management is another area of logistics within inventory control and transportation crucial for incorporation in different COVID-19 scenarios (Nagurney, 2021b). There are numerous other areas of retail supply chain management relevant for research (Schleper et al., 2021). Acknowledging the potential benefits of more nuanced approaches to stockpiling, this research suggests that simple scientifically-based measures could be practical and easier to implement in retailing practice. At the same time, just like in other areas of social science, exaggerated results of rushed research should not lead to misleading claims during the pandemic (Fowler, 2020). While the current focus is on the pandemic, other socio-economic areas of research risk being neglected (Walker et al., 2022). Generally, more interdisciplinary research and empirical support are required to tackle the unprecedented challenges to ensure product availability during the pandemic. Some of the future work is discussed at the end of the next section. Summary The pandemic of COVID-19 in 2020 brought the urgent need to adjust not only in sectors such as health and education but also in business practices. Importantly inventory management became an urgent priority for retailers as massive shortages of essential products affected consumers and society. This research builds upon the established supply chain management theory models to analyze the consumer stockpiling during the pandemic and suggest preventive measures. The model includes a retailer and consumers that both make decisions based upon their reorder points with order quantity. Lead-time variability is assumed to reflect the supply side disruption of retail. However, the most significant impact comes from demand-side disruption from consumer stockpiling exhibiting itself in two forms: rational hoarding and speculative panic buying. Consumers first increase purchases and later re-adjust closer to the pre-pandemic level. In response, the retailer raises the reorder point. Higher safety stock and bullwhip effect resulting from the pandemic increases costs for consumers and especially retailers. However, the inventory levels return to the levels close to the pre-pandemic normal after the phase of an initial surge. A continuous inventory review policy becomes more relevant than periodic review policies. Demand substitution due to brand and store switching allows reducing the harmful (bullwhip) effect of increased variability in the supply chain. Table 3 presents a summary of potential measures against the causes of disruptions discussed in this paper.Table 3 Causes and preventive measures in supply chain disruptions caused by stockpiling during the pandemic Causes Factors Countermeasures Implementation in practice Hoarding Rational expectations of consumers about decreasing availability Switch from periodic to continuous review of inventory Shift from sS and similar inventory policies to reorder-point-fixed-order policy using RFID and other technologies Rationing by government Ration stamps/cards for most scarce and critical items Higher substitution Omnichannel fulfillment, in-store pickup Buying limits on each consumer purchase Loyalty or membership card to control buying limits Panic buying Irrational or speculative belief of consumers about lower availability and higher price in the future Strongly increasing psychological utility of consumption Consumer education and timely communication Awareness campaigns and effective communication by using mass media/social networks Price guarantee of non-increase or decrease By government and business Availability guarantee Can be implemented only with sufficient inventory Future work to extend this research should involve more empirical support. This exploratory research used simple models with numerous assumptions and a limited number of numerical examples so that more analytical results would be desirable. The model of a retailer in this research assumes either RQ or sS inventory policies being prevalent in retailing practice. The wide use of EOQ quantity in RQ policy provides a good approximation for optimal order size convenient to implement in practice. Furthermore, the retailer is assumed to follow a sales target objective instead of profit maximization. The assumptions mentioned above in this research are reasonable in retailing practice considering the difficulty of profit maximization methods that are primarily based on the newsvendor model. However, an extension of the study to alternative business objectives and inventory policies could be valuable for generalizing results. Finally, finding a solution for optimal buying limits could be beneficial for practitioners. Acknowledgements The author is thankful to two anonymous reviewers for their invaluable comments that helped to improve the manuscript immensely. This research has received funding from the European Union's Horizon 2020 Marie Skłodowska-Curie Research and Innovation Staff Exchange (MSCA RISE) Program under the grant agreement No. 870647. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References Abideen AZ Mohamad FB Hassan MR Mitigation strategies to fight the COVID-19 pandemic—present, future and beyond Journal of Health Research 2020 34 547 562 10.1108/JHR-04-2020-0109 Adam D Special report: the simulations driving the world's response to COVID-19 Nature 2020 580 7802 316 319 10.1038/d41586-020-01003-6 32242115 Ajmal MM Khan M Shad MK The global economic cost of coronavirus pandemic: current and future implications Public Administration and Policy 2021 24 290 305 10.1108/PAP-10-2021-0054 Alvarez K Kreinovich V How can econometrics help fight the COVID-19 pandemic? Asian Journal of Economics and Banking 2020 4 29 36 10.1108/AJEB-07-2020-0027 Anas M Khan MN Rahman O Uddin SF Why consumers behaved impulsively during COVID-19 pandemic? South Asian Journal of Marketing 2022 3 7 20 10.1108/SAJM-03-2021-0040 Arachchi HDM Weerasiri RS Mendis T Impact of perceived corporate citizenship on purchase intention: across the fear of COVID-19 during the COVID-19 pandemic South Asian Journal of Marketing 2022 3 38 59 10.1108/SAJM-10-2021-0117 Aviv Y Pazgal A Optimal pricing of seasonal products in the presence of forward-looking consumers Manufacturing & Service Operations Management 2008 10 3 339 359 10.1287/msom.1070.0183 Baveja A Kapoor A Melamed B Stopping Covid-19: a pandemic-management service value chain approach Annals of Operations Research 2020 289 173 184 10.1007/s10479-020-03635-3 32421089 BCG Consumer spending tracker 2020 Illinois, United States IRI Global Headquarters Belhadi A Mani V Kamble SS Khan SAR Verma S Artificial intelligence-driven innovation for enhancing supply chain resilience and performance under the effect of supply chain dynamism: an empirical investigation Annals of Operations Research 2021 10.1007/s10479-021-03956-x Burgin TA The gamma distribution and inventory control Journal of the Operational Research Society 1975 26 3 507 525 10.1057/jors.1975.110 Byun SE Sternquist B Here today, gone tomorrow: Consumer reactions to perceived limited availability Journal of Marketing Theory and Practice 2012 20 2 223 234 10.2753/MTP1069-6679200207 Cachon GP Randall T Schmidt GM In search of the bullwhip effect Manufacturing & Service Operations Management 2007 9 4 457 479 10.1287/msom.1060.0149 Carlson JP A pitfall of using offer quantity limits The International Review of Retail, Distribution and Consumer Research 2021 31 3 358 374 10.1080/09593969.2020.1864657 Chen F Drezner Z Ryan JK Simchi-Levi D Quantifying the bullwhip effect in a simple supply chain: the impact of forecasting, lead times, and information Management Science 2000 46 3 436 443 10.1287/mnsc.46.3.436.12069 Chen K Pun CS Wong HY Efficient social distancing during the COVID-19 pandemic: Integrating economic and public health considerations European Journal of Operational Research 2021 304 84 98 10.1016/j.ejor.2021.11.012 34785855 Choi TM Risk analysis in logistics systems: a research agenda during and after the COVID-19 pandemic Transportation Research Part E Logistics and Transportation Review 2020 145 102190 10.1016/j.tre.2020.102190 Choi TM Innovative "bring-service-near-your-home" operations under corona-virus (COVID-19/SARS-CoV-2) outbreak: can logistics become the messiah? Transportation Research Part e: Logistics and Transportation Review 2020 140 101961 10.1016/j.tre.2020.101961 32346356 Corominas A A model for designing a procurement-inventory system as a defence against a recurring epidemic International Journal of Production Research 2021 10.1080/00207543.2021.1919779 Craighead CW Ketchen DJ Jr Darby JL Pandemics and supply chain management research: toward a theoretical toolbox Decision Sciences 2020 51 4 838 866 10.1111/deci.12468 34234384 Daoud, E., Mae, L., & Wadelto, B. (2020). Coles, Woolworths, Aldi: Grocery restriction limits on toilet paper, rice, pasta, milk and antibacterial wipes. 7News. Dasaklis TK Pappis CP Rachaniotis NP Epidemics control and logistics operations: a review International Journal of Production Economics 2012 139 2 393 410 10.1016/j.ijpe.2012.05.023 Delasay M Jain A Kumar S Impacts of the COVID-19 Pandemic on Grocery Retail Operations: An Analytical Model Production and Operations Management 2021 31 2237 2255 10.1111/poms.13717 Denise, L. (2020). The pandemic is rewriting the rules of retail. Harvard Business Review. Dohmen AE Merrick JR Saunders LW Stank TTP Goldsby TJ When preemptive risk mitigation is insufficient: the effectiveness of continuity and resilience techniques during COVID-19 Production and Operations Management 2022 10.1111/poms.13677 Dolgui A Ivanov D Ripple effect and supply chain disruption management: new trends and research directions International Journal of Production Research 2021 59 102 109 10.1080/00207543.2021.1840148 Dreisiebner S März S Mandl T Information behavior during the Covid-19 crisis in German-speaking countries Journal of Documentation 2021 78 160 175 10.1108/JD-12-2020-0217 Elavarasan RM Pugazhendhi R Restructured society and environment: a review on potential technological strategies to control the COVID-19 pandemic Science of the Total Environment 2020 725 138858 10.1016/j.scitotenv.2020.138858 32336562 Elmassah S Bacheer S Hassanein E US consumers' confidence and responses to COVID-19 shock Review of Economics and Political Science 2022 10.1108/REPS-10-2021-0098 Eppen GD Effects of centralization on expected costs in a multi-location newsboy problem Management Science 1979 25 5 498 501 10.1287/mnsc.25.5.498 European Centre for Disease Prevention and Control (2020). COVID-19 Coronavirus data. EU Open Data Portal 2020. Eпaнчинцeв, E. (2020). Aнaлитики пoдcчитaли, нa cкoлькo днeй poccиянe зaпacлиcь пpoдyктaми. PИA Hoвocти / X5 Retail Group. Fowler, A. (2020). Curing Coronavirus isn't a Job for Social Scientists. Bloomberg. Frederico, G. F. (2021). Towards a supply chain 4.0 on the post-COVID-19 pandemic: a conceptual and strategic discussion for more resilient supply chains. Rajagiri Management Journal. Gerchak Y Mossman D On the effect of demand randomness on inventories and costs Operations Research 1992 40 4 804 807 10.1287/opre.40.4.804 Ghadge A Er M Ivanov D Chaudhuri A Visualisation of ripple effect in supply chains under long-term, simultaneous disruptions: a system dynamics approach International Journal of Production Research 2021 10.1080/00207543.2021.1987547 Golubeva, O. (2021). Firms’ performance during the COVID-19 outbreak: international evidence from 13 countries. Corporate Governance: The International Journal of Business in Society. Govindan K Mina H Alavi B A decision support system for demand management in healthcare supply chains considering the epidemic outbreaks: a case study of coronavirus disease 2019 (COVID-19) Transportation Research Part e: Logistics and Transportation Review 2020 138 101967 10.1016/j.tre.2020.101967 32382249 Grynspan, R. (2022). Here's how we can resolve the global supply chain crisis. UNCTAD. Guan D Wang D Hallegatte S Davis SJ Huo J Li S Gong P Global supply-chain effects of COVID-19 control measures Nature Human Behaviour 2020 4 6 577 587 10.1038/s41562-020-0896-8 Gupta M Shoja A Mikalef P Toward the understanding of national culture in the success of non-pharmaceutical technological interventions in mitigating COVID-19 pandemic Annals of Operations Research 2021 10.1007/s10479-021-03962-z Hecht AA Biehl E Barnett DJ Neff RA Urban food supply chain resilience for crises threatening food security: a qualitative study Journal of the Academy of Nutrition and Dietetics 2019 119 2 211 224 10.1016/j.jand.2018.09.001 30527912 Hendalianpour, A., Hamzehlou, M., Feylizadeh, M. R., Xie, N., & Shakerizadeh, M. H. (2020). Coordination and competition in two-echelon supply chain using grey revenue-sharing contracts. Grey Systems: Theory and Application. Herold DM Nowicka K Pluta-Zaremba A Kummer S COVID-19 and the pursuit of supply chain resilience: reactions and “lessons learned” from logistics service providers (LSPs) Supply Chain Management: An International Journal 2021 26 702 714 10.1108/SCM-09-2020-0439 Hillier FS Lieberman GJ Introduction to Operations Research 2004 New York McGrawHill Ho TH Tang CS Bell DR Rational shopping behavior and the option value of variable pricing Management Science 1998 44 12-part-2 S145 S160 10.1287/mnsc.44.12.S145 Howland, D. (2020). Apparel retailers are in trouble. Retail Dive. Ipsos (2020). Tracking coronavirus. Results from a multi-country poll. Ivanov D Predicting the impacts of epidemic outbreaks on global supply chains: a simulation-based analysis on the coronavirus outbreak (COVID-19/SARS-CoV-2) case Transportation Research Part E: Logistics and Transportation Review 2020 136 101922 10.1016/j.tre.2020.101922 32288597 Ivanov D Supply chain viability and the COVID-19 pandemic: a conceptual and formal generalisation of four major adaptation strategies International Journal of Production Research 2021 59 12 3535 3552 10.1080/00207543.2021.1890852 Ivanov D Dolgui A Viability of intertwined supply networks: extending the supply chain resilience angles towards survivability. A position paper motivated by COVID-19 outbreak International Journal of Production Research 2020 58 10 2904 2915 10.1080/00207543.2020.1750727 Ivanov D Dolgui A OR-methods for coping with the ripple effect in supply chains during COVID-19 pandemic: managerial insights and research implications International Journal of Production Economics 2021 232 107921 10.1016/j.ijpe.2020.107921 32952301 Johnson, M. (2020) Who Hoards? The Personality of Stockpiling Behavior. Psychology Today. Jumrisko, M. (2020). Singapore Grocery Chain Starts Limiting How Much People Can Buy. Bloomberg. Kaplan EH OM Forum—COVID-19 scratch models to support local decisions Manufacturing & Service Operations Management 2020 22 4 645 655 10.1287/msom.2020.0891 Kapser S Abdelrahman M Bernecker T Autonomous delivery vehicles to fight the spread of Covid-19–How do men and women differ in their acceptance? Transportation Research Part A: Policy and Practice 2021 148 183 198 33776251 Katsaliaki K Galetsi P Kumar S Supply chain disruptions and resilience: a major review and future research agenda Annals of Operations Research 2021 10.1007/s10479-020-03912-1 Koerber T Schiele H Is COVID-19 a turning point in stopping global sourcing? Differentiating between declining continental and increasing transcontinental sourcing Journal of Global Operations and Strategic Sourcing 2021 10.1108/JGOSS-02-2021-0018 Kronblad C Pregmark JE Responding to the COVID-19 crisis: the rapid turn toward digital business models Journal of Science and Technology Policy Management 2021 10.1108/JSTPM-10-2020-0155 Kumar S Xu C Ghildayal N Chandra C Yang M Social media effectiveness as a humanitarian response to mitigate influenza epidemic and COVID-19 pandemic Annals of Operations Research 2021 10.1007/s10479-021-03955-y Placer Labs (2020). COVID-19 Retail Impact Tracker. Accessed on 13.04.2020 https://www.placer.ai/covid-19/. Lee HL Padmanabhan V Whang S Information distortion in a supply chain: the bullwhip effect Management Science 1997 43 4 546 558 10.1287/mnsc.43.4.546 Li D Dong C Government regulations to mitigate the shortage of life-saving goods in the face of a pandemic European Journal of Operational Research 2021 301 942 955 10.1016/j.ejor.2021.11.042 Li Y Wang X Gong T Wang H Breaking out of the pandemic: How can firms match internal competence with external resources to shape operational resilience? Journal of Operations Management 2022 10.1002/joom.1176 Liu K Liu C Xiang X Tian Z Testing facility location and dynamic capacity planning for pandemics with demand uncertainty European Journal of Operational Research 2021 304 150 168 10.1016/j.ejor.2021.11.028 34848916 Liu Y Wang DD Xu Q A supply chain coordination mechanism with suppliers' effort performance level and fairness concern Journal of Retailing and Consumer Services 2020 53 101950 10.1016/j.jretconser.2019.101950 Lorentz H Laari S Meehan J Eßig M Henke M An attention-based view of supply disruption risk management: balancing biased attentional processing for improved resilience in the COVID-19 context International Journal of Operations & Production Management 2021 41 152 177 10.1108/IJOPM-06-2021-0381 Manuj I Mentzer JT Global supply chain risk management strategies International Journal of Physical Distribution & Logistics Management 2008 38 192 223 10.1108/09600030810866986 Matarazzo M Diamantopoulos A Applying reactance theory to study consumer responses to COVID restrictions: a note on model specification International Marketing Review 2022 10.1108/IMR-12-2021-0370 Maтoвникoв, M., Кopжeнeвcкий, & H., Кaмpoтoв M. (2020). Oпepaтивнaя oцeнкa пoтpeбитeльcкoй aктивнocти poccиян. Лaбopaтopия CбepДaнныe. Metters R Quantifying the bullwhip effect in supply chains Journal of Operations Management 1997 15 2 89 100 10.1016/S0272-6963(96)00098-8 Mouratidis K Peters S COVID-19 impact on teleactivities: role of built environment and implications for mobility Transportation Research Part A: Policy and Practice 2022 158 251 270 35291720 Mower JM Pedersen EL Manufacturer, retailer and consumer misbehaviour in the United States during the Second World War Fashion, Style & Popular Culture 2018 5 1 41 57 10.1386/fspc.5.1.41_1 Nagurney A Optimization of supply chain networks with inclusion of labor: applications to COVID-19 pandemic disruptions International Journal of Production Economics 2021 235 108080 10.1016/j.ijpe.2021.108080 Nagurney A Supply chain game theory network modeling under labor constraints: applications to the Covid-19 pandemic European Journal of Operational Research 2021 293 3 880 891 10.1016/j.ejor.2020.12.054 33519049 Nielsen (2020). FMCG and Retail. Office for National Statistics UK (2020). Coronavirus, the U.K. economy and society, faster indicators. Ovezmyradov B Kurata H Effects of customer response to fashion product stockout on holding costs, order sizes, and profitability in omnichannel retailing International Transactions in Operational Research 2019 26 1 200 222 10.1111/itor.12511 Palau-Saumell R Matute J Derqui B Meyer JH The impact of the perceived risk of COVID-19 on consumers' attitude and behavior toward locally produced food British Food Journal 2021 123 281 10.1108/BFJ-04-2021-0380 Passetti EE Battaglia M Bianchi L Annesi N Coping with the COVID-19 pandemic: the technical, moral and facilitating role of management control Accounting, Auditing & Accountability Journal 2021 34 1430 1444 10.1108/AAAJ-08-2020-4839 Penco L Ciacci A Benevolo C Torre T Open social innovation for surplus food recovery and aid during COVID-19 crisis: the case of Fondazione Banco Alimentare Onlus British Food Journal 2021 10.1108/BFJ-02-2021-0116 Pournader M Kach A Talluri S A review of the existing and emerging topics in the supply chain risk management literature Decision Sciences 2020 51 4 867 919 10.1111/deci.12470 34234385 Queiroz MM Wamba SF Jabbour CJC Machado MC Supply chain resilience in the UK during the coronavirus pandemic: a resource orchestration perspective International Journal of Production Economics 2022 245 108405 10.1016/j.ijpe.2021.108405 35002082 Raassens N Haans H Mullick S Surviving the hectic early phase of the COVID-19 pandemic: a qualitative study to the supply chain strategies of food service firms in times of a crisis The International Journal of Logistics Management 2021 10.1108/IJLM-01-2021-0013 Remko VH Research opportunities for a more resilient post-COVID-19 supply chain–closing the gap between research findings and industry practice International Journal of Operations & Production Management 2020 40 4 341 355 10.1108/IJOPM-03-2020-0165 Repko M Costco brings back purchase limits on toilet paper, cleaning supplies and More 2021 New Jersey CNBC Rey, J. (2020). Amazon is banning the sale of N95 and surgical masks to the general public. Vox. Rodrigue JP The Geography of Transport Systems 2016 Taylor & Francis Rozhkov M Ivanov D Blackhurst J Nair A Adapting supply chain operations in anticipation of and during the COVID-19 pandemic Omega 2022 110 102635 10.1016/j.omega.2022.102635 35291412 Salaverria L Manufacturers, retailers seek lifting of purchase limits 2020 Makati Philippine Daily Inquirer Salvietti G Ziliani C Teller C Ieva M Ranfagni S Omnichannel retailing and post-pandemic recovery: building a research agenda International Journal of Retail & Distribution Management 2022 50 1156 1181 10.1108/IJRDM-10-2021-0485 Sanci E Daskin MS Hong YC Roesch S Zhang D Mitigation strategies against supply disruption risk: a case study at the Ford Motor Company International Journal of Production Research 2021 10.1080/00207543.2021.1975058 Satish K Venkatesh A Manivannan ASR Covid-19 is driving fear and greed in consumer behaviour and purchase pattern South Asian Journal of Marketing 2021 2 113 129 10.1108/SAJM-03-2021-0028 Schlegelmilch BB Sharma K Garg S Employing machine learning for capturing COVID-19 consumer sentiments from six countries: a methodological illustration International Marketing Review 2022 10.1108/IMR-06-2021-0194 Schleper MC Gold S Trautrims A Baldock D Pandemic-induced knowledge gaps in operations and supply chain management: COVID-19’s impacts on retailing International Journal of Operations & Production Management 2021 41 193 205 10.1108/IJOPM-12-2020-0837 Schmitt AJ Sun SA Snyder LV Shen ZJM Centralization versus decentralization: risk pooling, risk diversification, and supply chain disruptions Omega 2015 52 201 212 10.1016/j.omega.2014.06.002 Shen ZM Sun Y Strengthening supply chain resilience during COVID-19: a case study of JD. com Journal of Operations Management 2021 10.1002/joom.1161 Shih, W. (2022) From Just-In-Time to Just-In-Case: Is Excess and Obsolete Next? Forbes. Shou, B., Xiong, H., & Shen, X. M. (2013). Consumer panic buying and quota policy under supply disruptions. Working paper. Hong Kong: City University of Hong Kong. Siebert, B. (2020). Coronavirus hoarder tries to return $10,000 worth of goods to Adelaide supermarket. ABC Radio. Singh JP Dwivedi YK Rana NP Kumar A Kapoor KK Event classification and location prediction from tweets during disasters Annals of Operations Research 2019 283 1 737 757 10.1007/s10479-017-2522-3 Singh S Kumar R Panchal R Tiwari MK Impact of COVID-19 on logistics systems and disruptions in food supply chain International Journal of Production Research 2021 59 7 1993 2008 10.1080/00207543.2020.1792000 Stiff, R., Johnson, K., & Tourk, K. A. (1975). Scarcity and hoarding: economic and social explanations and marketing implications. ACR North American Advances. Su X Zhang F Strategic customer behavior, commitment, and supply chain performance Management Science 2008 54 10 1759 1773 10.1287/mnsc.1080.0886 Sucky E The bullwhip effect in supply chains—an overestimated problem? International Journal of Production Economics 2009 118 1 311 322 10.1016/j.ijpe.2008.08.035 Swinney R Selling to strategic consumers when product value is uncertain: the value of matching supply and demand Management Science 2011 57 10 1737 1751 10.1287/mnsc.1110.1360 Taylor, D., Pritchard, A., Duhan, D., & Mishra, S. (2020). What's behind the empty grocery shelves. Supply Chain Management Review, 1–4. Terazono, E., & Evans, J. (2020). How coronavirus is affecting pasta's complex supply chain. Financial Times. Thomas, L. (2020). Coronavirus wreaks havoc on retail supply chains globally, even as China's factories come back online. CNBC. Tolk A Harper A Mustafee N Hybrid models as transdisciplinary research enablers European Journal of Operational Research 2021 291 3 1075 1090 10.1016/j.ejor.2020.10.010 33078041 Tsao YC Raj P Raj PV Product substitution with customer segmentation under panic buying behavior Scientia Iranica 2019 10.24200/sci.2019.5099.1093 Tsiligianni C Tsiligiannis A Tsiliyannis C A stochastic inventory model of COVID-19 and robust, real-time identification of carriers at large and infection rate via asymptotic laws European Journal of Operational Research 2022 304 42 56 10.1016/j.ejor.2021.12.037 35035055 Tyko, K. (2022). Grocery stores still have empty shelves amid supply chain disruptions,omicron and winter storms. USA Today. Van Oorschot KE Van Wassenhove LN Jahre M Collaboration–competition dilemma in flattening the COVID-19 curve Production and Operations Management 2022 10.1111/poms.13709 Walker J Brewster C Fontinha R Haak-Saheem W Benigni S Lamperti F Ribaudo D The unintended consequences of the pandemic on non-pandemic research activities Research Policy 2022 51 1 104369 10.1016/j.respol.2021.104369 34565926 Yang H Schrage L Conditions that cause risk pooling to increase inventory European Journal of Operational Research 2009 192 3 837 851 10.1016/j.ejor.2007.10.064 Yang J Xie H Yu G Liu M Antecedents and consequences of supply chain risk management capabilities: an investigation in the post-coronavirus crisis International Journal of Production Research 2021 59 5 1573 1585 10.1080/00207543.2020.1856958 Ye F Liu K Li L Lai KH Zhan Y Kumar A Digital supply chain management in the COVID-19 crisis: an asset orchestration perspective International Journal of Production Economics 2022 245 108396 10.1016/j.ijpe.2021.108396 34931109 Ziady, H. (2020). Panic buying is forcing supermarkets to ration food and other supplies. CNN Business.	36467007	PMC9709757	NO-CC CODE	2022-12-01 23:23:09	no	Ann Oper Res. 2022 Nov 30;:1-33	utf-8	Ann Oper Res	2,022	10.1007/s10479-022-05091-7	oa_other
==== Front Environ Syst Decis Environ Syst Decis Environment Systems & Decisions 2194-5403 2194-5411 Springer US New York 9886 10.1007/s10669-022-09886-8 Perspective ESG risks and corporate survival Cohen Gil [email protected] Western Galilee Academic College, Acre, Israel 30 11 2022 16 23 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 research is the first attempt to examine the impact of corporate sustainability risks factors on its financial stability. By using S&P500 stocks data from 2019 to 2021 and calculating Altman’s Z-score, we examined the influence of ESG (Environmental, Social, and Corporate Governance) risks score on the company survival chances. We documented diminishing total ESG scores of S&P500 stocks in recent years pointing out that companies pay attention to sustainability issues and invest resources to reduce them. We documented that Altman’s Z-score is negatively influenced by E and S and not by G. These findings are very important since they prove for the first time that high environmental and social risks may reduce corporates’ financial stability and rise their default risks incurring default costs. Moreover, high sensitivity of Altman’s Z-score changes to S changes was found especially for relatively smaller firms. The result of this study emphasizes the importance of sustainability risk and especially social risk to a firm’s survival chances and therefore mitigating those risks can dramatically improve corporates’ financial stability. Keywords Sustainability Environmental Social Corporate governance Corporate survival Altman Z Western Galilee College ==== Body pmcIntroduction Environmental (E), Social (S), and corporate Governance (G) risks are becoming more and more important to corporates around the globe. The corporate concerns of ESG risks reflect the growing importance investors and the general public attribute to sustainability issues. Firms look for ways to mitigate those risks knowing that they will have to commit a lot of resources to ensure a successful process. These investments can weaken the firm’s financial stability because of their scope and nature; however, they can do the opposite through the positive corporate image among investors, customers, and the entire business circles. In this research, we investigate the S&P500 firm’s ESG risk’s impact on the firm’s survival chances measured by Altman’s Z-score. Moreover, we investigate the relative importance of ESG risks to the corporate’s survival chances compared to external economic factors such as GDP growth and U.S. government 10 years bonds yields. Our analysis has established that from the tree ESG risks factors only E and S risks have a negative impact on the firm’s financial stability. S risks have been found to have the strongest influence upon the companies Altman’s Z-score emphasizes that corporate social fairness to its workers, customers, and other business partners is conceived as a major factor in the company’s survival chances, and therefore, corporations must invest resources to ensure they mitigate those risks. Literature Review Sustainability ingredients have become crucial to corporate long-term success (Eccles et al., 2012; Ortiz-de-Mandojana and Bansal, 2016), and it has been increasingly studied in the academic literature in recent decades. However, there is a gap in the literature about the link between sustainability and a firm’s survival chances and bankruptcy cost that are associated with default risks. The aim of the following research is to bridge that gap of knowledge by examining the direct impact of ESG risks factors on Altman’s Z-score that measures the financial stability of firms. Moreover, we will try to determine whether sustainability issues are more important in recent years to corporation survival than major economic factors. Corporate sustainability generally refers to the integration of financial profitability, environmental protection, and social responsibility into organizations’ mission declarations and every applied to everyday activities (Elkington, 1997; Lo, 2010; Schaltegger et al. 2013). It is also defined as meeting the needs of a firm’s direct and indirect stakeholders such as shareholders, employees, clients, pressure groups, communities, etc., without compromising its ability to meet the needs of future stakeholders as well. Although many researchers have focused on corporate sustainability models and efforts, not many have tried to document the effect of sustainability risks on the financial markets. Cohen (2021) has found that E risks negatively affect excess return over the Nasdaq100 index in some years and that S risks are the most influential factor negatively related to excess returns. Nizam et al. (2019) have studied the impact of sustainability on banks’ performance and concluded that financial performance and social and environmental performance are related, evidence for the banking sector remains limited and inconclusive. Garcia et al. (2019) have found that larger companies have higher levels of performance. They also found that companies in sensitive industries present superior environmental performance even when controlling for size and country. Lee et al. (2013) examined whether portfolios comprising high‐ranked Corporate Social Performance (CSP) firms out/underperform portfolios comprised of low‐ranked CSP firms for a U.S. sample of firms covering the period from 1998 till 2007. Their results are consistent with the “no‐linkage” hypothesis, which argues that no significant difference in the risk‐adjusted performance is expected between high‐ and low‐ranked CSP‐formed portfolios. Unlike those important papers, we do not try to expose the linkage between ESG risk to market performances but rather to the firm’s survival chances and financial stability by linking sustainability risks factors to Altman (1968) who developed a model for bankruptcy prediction using multiple discriminant analysis called Altman’s Z-score model. Altman defined his Z-score model as a statistical measure to predict company financial failure, calculated as a linear combination of four or five common financial ratios, weighted by coefficients. At a later work (Altman 2000) he established two models: Model A Z-score for manufacturing companies closed, and Model B Z-score for non-manufacturing companies. Many researchers have followed his pioneered work and tested the ability of the Z models to forecast the business failure of companies from different sectors and economic regimes. Since the introduction of Altman’s models’ researchers has tested the model’s ability to detect corporate financial default. Hayes et al. (2010) had analyzed the construction of Z-score model by applying it to a sample of 17 U.S. firms from the retail industry, the study revealed that the model correctly predicts bankruptcy at a level of 94%. Mamo (2011) predicted financial distress on 43 banks and concluded that Altman’s model is correct in 80% of the cases in the financial sector. Other researchers have focused on bankruptcy costs that arise when the firm experiences financial distress and potential default. During a default process, a firm exhibits direct and indirect costs that can assume a considerable value of the firm (Altman, 1984; Opler and Titman, 1994). Andrade and Kaplan (1998) estimate financial distress costs to be 10–23% of the firm’s value while Glover (2016) estimates the figure to be 25%. Those costs come from different sources such as lower than market assets liquidation prices and business difficulties incurred by business partners. Since costs of default are high, a firm must invest efforts to avoid potential financial distress including taking care of its ESG risks. Methodologies and results This research is based on S&P500 stocks data from 2019 until 2021. For each stock, we calculated yearly Altman’s Z-score1 and collected ESG risks factors.2 Our data contains the years of Covid-19 world pandemic that have risen awareness to the globalization negative effects and imposed new challenges on many business aspects such as transportation of goods and on sight employment. The pandemic has also risen people’s awareness of the need to preserve the environment for future generations. Table 1 summarizes descriptive statistics of all the data (2019–2021) while Fig. 1 presents year-by-year averages.Table 1 Descriptive statistics data 2019–2021 E S G ESG Altman’s Z Average 5.67 11.05 7.34 24.06 3.71 St. Dev 5.38 4.04 2.15 7.10 3.00 Max 19.09 20.87 13.13 39.35 16.7 Min 0.02 1.93 3.26 10.78 − 1.33 Eenvironmental, S social, G corporate governance, ESG the total sustainability risk factor Altman’s Z = financial stability score, Z > 2.6—safe zone, 1.1 < Z < 2.6- gray area zone, Z < 1.1 distress zone Fig. 1 Risks and Altman’s Z-scores for S&P500 stocks ESG Table 1 and Fig. 1 show that the total ESG risks have been diminishing from 2019 until 2021. That phenomenon is consistent with the G and S risks. However, E risks of the S&P500 stocks demonstrate inconsistency in its trend, rising from 5.13 in 2019 to 6.08 in 2020 and dropping to 5.64 in 2021. S risks have been found to be higher than E and G risks for all the examined years. The highest standard deviation of the ESG factors is associated with E (5.38) pointing out that S&P500 stocks are relatively different in their environmental sensitivity and behavior while they act relatively similarly on corporate governance issues. Altman’s Z also varies dramatically in a wide range from highly financially stable firms to firms with real bankruptcy potential in the near future. Figure 1 also demonstrates that the average Altman’s Z-score has dropped from 3.71 in 2019 to 3.47 in 2020 and then risen to 4.18 in 2021. Lioui and Tarelli (2022) compared two dominant methodologies for the construction of an ESG factor: the time-series and cross-sectional approaches. Differences in ESG rating and exposure to other firm characteristics imply an ex ante expected return spread between the two factors. They documented strong variation of the alpha factor in the time series which is negatively related to media attention to ESG. We followed Lioui and Tarelli model of pure factors creation using ESG rating and Altman’s Z as stock’s returns explanatory variables. First, we measured the stand-alone survival factor on the stocks returns (Eq. 1), and then we added to the regression the ESG explanatory variables (Eq. 2).1 Ri=αi+β1Zi, 2 Ri=αi+β1Zi+β2ESGi. The result of Eq. 1 is as follows:Ri=8.77+2.71Z, t: stat2.693.97, R2=0.22,F=15.8. Equation 1 shows significant positive impact of Altman Z-score on stocks return. The βi which measures the impact of Z on the stocks return, is 2.71. Table 2 summarizes the results of Eq. 2.Table 2 The Impact of Altman Z and ESG Factors on the Stock’s Returns Model αi β1(Zi) β2(ESGi) 1 4.41 (0.48) 2.85 (3.87) 0.16 0.51 R2 = 0.18, F = 12.1 αi β1(Zi) β2(Ei) 2 5.79 (1.32) 2.90 (4.10) 0.40 (1.12) R2 = 0.17, F = 13.5 αi β1(Zi) β2(Si) 3 15.79 (2.16) 2.51 (3.54) − 0.56 (− 1.07) R2 = 0.19, F = 14.1 αi β1(Zi) β2(Gi) 4 1.32 (0.17) 2.76 (4.03) 0.99 (1.05) R2 = 0.17, F = 13.7 The explanatory variable is the stock’s return, the number in the brackets are the t statistics Table 2 demonstrates that Altman Z has a significant positive impact on stock’s return. Moreover, the existence of the ESG factors in the regression model has risen the β1 from 2.71 to 2.85 for the ESG, 2.90 for E and 2.76 for G. Only for the model that included S, the β1 was reduced to 2.51. The delta of β1 between the original model (Eq. 1) and β1 in the second model (Eq. 2) demonstrates the sensitivity of the Z impact on stock’s return due to ESG parameters. We now construct and implement an econometric model that measures the impact of the various ESG risks factors, firm’s size, and two general economic data including U.S. 10 years bonds yields and U.S. GDP (Gross Domestic Product) growth rate, on S&P500 firms Altman’s Z-score. Our aim is to evaluate the importance of the different ESG scores along with external economic data on the individual firm’s survival chances. The model implementation is presented in Eq. 3.3 AltZ=6.34-0.14E-0.28S+0.10G-0.59Yield+0.08USg+1.02Size, Tstat:7.00∗∗-3.09∗∗-4.21∗∗0.80-0.971.172.01∗∗, R2=0.21,F=6.25, where: AltZ = Altman’s Z-score, E = environmental risks, S = social risks, G = corporate governance risks, Yield = yields of U.S. government 10 years bonds, USg = U.S. economy GDP growth rate, Size = Dummy variable for the corporate market value: 1 = more than 100 B $, 0 = less than 100 B $. = significant at 95%. The econometric model indicates that Altman’s Z-score is negatively influenced by E and S and not by G. These findings are very important since they prove that high environmental and social risks may reduce corporates’ financial stability and rise their default risks incurring default costs (see e.g., Merton 1974; Koziol 2014). These findings support the hypothesis that investing resources in order to reduce its E and S risks may be increase the firm’s overall value since by doing so it reduces its default probability associated with default costs that may reduce the company value up to 25% of its pre-default value (Glover 2016). These results also support the idea that firms economically should invest resources to mitigate environmental and social damages they incur on their surroundings especially if it faces survival risks. Results also show that G risks do not affect Altman’s Z-score as the other two ESG components. From investors perspective, they should avoid investing in high environmental and social risks firms if their survival chances are slim or demand a higher return on investment that will compensate the incremental risks involved. The model results also show no significant impact of U.S. GDP growth and 10 years government bonds on Altman’s Z-score. These findings indicate that E and S individually are more important to the corporate survival chances than data from the surrounding economy. Furthermore, the model also records an expected positive impact of the company size on its survival chances. We now repeat Model 3 splitting the data into two size categories according to the firm’s market values (above and below 100 B$) and report the results in Table 3.Table 3 The Impact of ESG Risk Factors on Altman’s Z-Score by the Firm’s Size Α β1E β2S β3G Yield USg Big firms 7.76 (5.77) − 0.14* (− 1.84) − 0.31 (− 3.43) 0.10 (0.65) − 0.50 (− 0.63) 0.08 (0.79) R2 = 0.21, F = 6.22 Small firms 6.09 (4.60) − 0.13 (− 2.45) − 0.22 (− 2.27) 0.10 (0.50) − 0.84 (0.84) 0.11 (0.97) R2 = 0.15, F = 3.41 Big firms = firms with above 100 billion market value. Small firms = firms with below 100 billion market value *Significant level of 95% Significant level at 90% The numbers in the brackets are t statistics Table 3 shows a similar negative effect of E on Altman’s Z for both firms’ size categories. However, S risks have a stronger negative impact on the financial stability of big firms compared to smaller firms. This finding is probably due to the fact that big corporation employs thousands of workers in a global economic surrounding that enhances challenges of the firms’ social policies. Moreover, big firms should influence the firm’s management to adopt high social standards regarding working conditions and gender equality, in order to improve the survival chances of their firm. The split model also shows that G risks and the external economic factors do not have a significant impact on Altman’s Z for both size categories. ESG risks are evaluated periodically by the evaluation agencies and therefore the risk scores are frequently being changed over time. Model 4 tries to capture the changes of ESG components on the changes of Altman’s z.4 ΔAltZ=0.01ΔE-0.33ΔS+0.31ΔG, tstat:0.08-2.32∗∗1.07, R2=0.16,F=3.73. where: ΔAltZ = changes in Altman’s z-score,ΔE = changes in the environmental risks, ΔS = changes in the social risks, ΔG = changes in the corporate governance risks, = significant at 95%. Model 4 indicates that changes in the S score have a negative significant effect on the changes on Altman’s Z-score. This model adds information to the results of Model 3 since it emphasizes the high sensitivity of the company’s survival chances to changes in its social risks. No such sensitivities were recorded with E and G changes. These results indicate that companies with financial stability risks should first invest in reducing their social risks before their environmental and social risks. The split model into two size categories is presented in Table 4.Table 4 The Impact of Changes of ESG Risk Factors on Altman’s Z-Score by the Firm’s Size β1E β2S β3G Big firms 0.01 (0.32) − 0.15 (− 1.32) 0.06 (1.06) R2 = 0.17, F = 3.09 Small firms 0.02 (0.44) − 0.46 (− 3.18) 0.03 (0.87) R2 = 0.06, F = 0.82 Big firms = firms with above 100 billion market value. Small firms = firms with below 100 billion market value *Significant level of 95% Significant level at 90% The numbers in the brackets are t statistics Table 4 shows that Altman’s Z-score changes are more sensitive to S changes for relatively small firms than for bigger firms. Table 3 also indicates that there is no size-based difference of Altman’s Z-score changes to E and G changes. Summary and conclusions This research is the first attempt to examine the impact of corporate sustainability risks factors on its financial stability. By using S&P500 stocks data from 2019 to 2021 and calculating Altman’s Z-score, we examined the influence of the ESG risks score on the company’s survival chances. In addition to the ESG scores, we incorporate to our model external economic data od U.S. government bonds yields and U.S. economic GDP growth rate. We documented diminishing total ESG scores of S&P500 stocks in recent years pointing out that companies pay attention to sustainability issues and invest resources to reduce them. A decline in S and G has been documented while E score raised from 2019 to 2020 and then declined towards 2021. We found that the average Altman’s Z-score for S&P500 stocks was 3.71 during the examined years improving to 4.18 in 2021. We also found that Altman’s Z-score is negatively influenced by E and S and not by G. These findings are very important since they prove for the first time that high environmental and social risks may reduce corporates’ financial stability and rise their default risks incurring default costs. Moreover, we documented a high sensitivity of Altman’s Z-score changes to S changes, especially for relatively smaller firms. No such sensitivities were detected of Altman’s z changes for E and G changes. The result of this study emphasizes the importance of sustainability risk and especially social risk to a firm’s survival chances and, therefore, mitigating those risks can dramatically improve corporates’ financial stability. The current study was based on data from the years in which the Covid-19 pandemic has changed the global business circumstances and the view of investors on issues such as climate change and social sustainability. It would be interesting to reexamine the discussed issues in the future in the absence of the pandemic. Author contributions Only one contributor Professor Gil Cohen. Funding This study was supported by the Western Galilee College. Declarations Conflict of interest There is no financial or non-financial interests that are directly or indirectly related to the work submitted for publication. 1 Altman’s Z-score is based on five financial ratios. 2 ESG data provided by Sustainalytics, Inc. ==== Refs References Altman EI Financial ratios, discriminant analysis, and the prediction of corporate bankruptcy J Finance 1968 23 4 589 609 10.1111/j.1540-6261.1968.tb00843.x Altman EI A further empirical investigation of the bankruptcy cost question J Finance 1984 39 1067 1089 10.1111/j.1540-6261.1984.tb03893.x Altman EI (2000). Predicting financial distress of companies. Revisiting the Z-score and zeta models. http://www.Default Risk.com Andrade G Kaplan S How costly is financial (not economic) distress? Evidence from highly leveraged transactions that become distressed J Finance 1998 5 1443 1493 10.1111/0022-1082.00062 Cohen G What can we learn from the financial market about sustainability Environ Syst Decis 2021 41 4 Eccles R, Ioannou I, Serafeim G (2012) Is sustainability now the key to corporate success? https://www.theguardian.com/sustainable-business/sustainability-key-corporate-success Elkington J Cannibals with forks: the triple bottom line of 21st century business 1997 Oxford Capstone Publishing Ltd Garcia AS Mendes-Da-Silva W Orsato RJ Mendes-Da-Silva W Corporate sustainability, capital markets, and ESG performance Individual behaviors and technologies for financial innovations 2019 Berlin Springer Glover B The expected cost of default J Financ Econ 2016 119 284 299 10.1016/j.jfineco.2015.09.007 Hayes SK, Hodge KA, Hughes LW (2010) A study on the efficiency of Altman’s Z to predict of specialty retail firms doing business in contemporary Times. Econ Bus J: Inq Perspect 3(1) Koziol C A simple correction of the WACC discount rate for default risk and bankruptcy costs Rev Quant Financ Acc 2014 42 653 666 10.1007/s11156-013-0356-x Lee DD Faff RW Rekker SAC Do high and low-ranked sustainability stocks perform differently? Int J Account Inf Manag 2013 21 2 116 132 10.1108/18347641311312267 Lioui A Tarelli A Chasing the ESG factor J Bank Finance 2022 139 106498 10.1016/j.jbankfin.2022.106498 Lo SF Performance evaluation for sustainable business: a profitability and marketability framework Corp Soc Responsib Environ Manag 2010 17 311 319 10.1002/csr.214 Mamo AQ Applicability of Altman (1968) model in predicting financial distress of commercial banks in Kenya 2011 University of Nairobi Unpublished MBA Research Project Merton R On the pricing of corporate debt: the risky structure of interest rates J Financ 1974 29 449 469 Nizam E Ng A Dewandaru G Nagayev R Nkoba MA The impact of social and environmental sustainability on financial performance: a global analysis of the banking sector J Multinatl Financ Manag 2019 49 35 53 10.1016/j.mulfin.2019.01.002 Opler TC Titman S Financial distress and corporate performance J Financ 1994 49 1015 1040 10.1111/j.1540-6261.1994.tb00086.x Ortiz-de-Mandojana N Bansal P The long-term benefits of organizational resilience through sustainable business practices Strateg Manag J 2016 37 8 1615 1631 10.1002/smj.2410 Schaltegger S Beckmann M Hansen EG Transdisciplinary in corporate sustainability: mapping the field Bus Strateg Environ 2013 22 4 219 229 10.1002/bse.1772	36466558	PMC9709759	NO-CC CODE	2022-12-01 23:23:09	no	Environ Syst Decis. 2022 Nov 30;:1-6	utf-8	Environ Syst Decis	2,022	10.1007/s10669-022-09886-8	oa_other
==== Front J Bionic Eng J Bionic Eng Journal of Bionic Engineering 1672-6529 2543-2141 Springer Nature Singapore Singapore 298 10.1007/s42235-022-00298-7 Research Article Crisscross Harris Hawks Optimizer for Global Tasks and Feature Selection Wang Xin [email protected] 1 Dong Xiaogang [email protected] 1 Zhang Yanan [email protected] 23 http://orcid.org/0000-0002-7714-9693 Chen Huiling [email protected] 4 1 grid.440668.8 0000 0001 0006 0255 School of Mathematics and Statistics, Changchun University of Technology, Changchun, Jilin, 130012 China 2 grid.43169.39 0000 0001 0599 1243 School of Management, Xi’an Jiaotong University, Xi’an, Shaanxi, 710049 China 3 grid.440668.8 0000 0001 0006 0255 Information Construction Office, Changchun University of Technology, Changchun, Jilin, 130012 China 4 grid.412899.f 0000 0000 9117 1462 College of Computer Science and Artificial Intelligence, Wenzhou University, Wenzhou, 325035 China 30 11 2022 122 3 2 2022 19 10 2022 20 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. Harris Hawks Optimizer (HHO) is a recent well-established optimizer based on the hunting characteristics of Harris hawks, which shows excellent efficiency in solving a variety of optimization issues. However, it undergoes weak global search capability because of the levy distribution in its optimization process. In this paper, a variant of HHO is proposed using Crisscross Optimization Algorithm (CSO) to compensate for the shortcomings of original HHO. The novel developed optimizer called Crisscross Harris Hawks Optimizer (CCHHO), which can effectively achieve high-quality solutions with accelerated convergence on a variety of optimization tasks. In the proposed algorithm, the vertical crossover strategy of CSO is used for adjusting the exploitative ability adaptively to alleviate the local optimum; the horizontal crossover strategy of CSO is considered as an operator for boosting explorative trend; and the competitive operator is adopted to accelerate the convergence rate. The effectiveness of the proposed optimizer is evaluated using 4 kinds of benchmark functions, 3 constrained engineering optimization issues and feature selection problems on 13 datasets from the UCI repository. Comparing with nine conventional intelligence algorithms and 9 state-of-the-art algorithms, the statistical results reveal that the proposed CCHHO is significantly more effective than HHO, CSO, CCNMHHO and other competitors, and its advantage is not influenced by the increase of problems’ dimensions. Additionally, experimental results also illustrate that the proposed CCHHO outperforms some existing optimizers in working out engineering design optimization; for feature selection problems, it is superior to other feature selection methods including CCNMHHO in terms of fitness, error rate and length of selected features. Supplementary Information The online version contains supplementary material available at 10.1007/s42235-022-00298-7. Keywords Harris hawks optimization Bioinspired algorithm Global optimization Engineering optimization Feature selection ==== Body pmcIntroduction Optimization is the procedure of determining the optimal solution for a specific issue with a reasonable computational expense. It has received increasing attentions in a variety of scientific fields of research and engineering and is essential to propose splendid optimization algorithms [1–4]. Traditional deterministic optimization approaches cannot mostly feasible in each real-world problem [4]. As becoming more and more sophisticated, the various optimization issues are extensively solved using the promising stochastic bioinspired algorithms [5–9]. This kind of algorithm is unsensitive to the searching magnitude and has strong optimization capability and flexibility of finding high-quality solution efficiently through a brief process, even for high-dimensional difficult problems [4, 10–18]. A great number of well-regarded bioinspired optimizers, such as Differential Evolution (DE) [19], Particle Swarm Optimization (PSO) [8], Sine Cosine Algorithm (SCA) [20], Grey Wolf Optimizer (GWO) [21], Moth-Flame Optimization (MFO) [22], Whale Optimization Algorithm (WOA) [23], Bat Algorithm (BA) [24], Gravitational Search Algorithm (GSA)[25], Harris Hawks Optimizer (HHO) [26], Slime Mould Algorithm (SMA) [27], Beluga Whale Optimization (BWO) [28], Dandelion Optimizer (DO) [29], Marine Predators Algorithm (MPA) [30], and their variants have been presented to handle a various optimization problems. HHO is a newly stochastic bioinspired optimization algorithm with promising performance of dealing with continuous problems. It can provide easy implement and outstanding exploitative capability of local search. The algorithm is developed according to the inspiration of cooperative preying behavior of Harris’ hawks and skilled in converging fast and making a suitable balance between intensification and diversification. The hawks’ dynamic chasing patterns, which are resulted from the natural surrounding and time-vary escaping ways of prey, are composed of six phases, two in exploration phase and the remainder in exploitation phase. With the help of the stochastic operations of six phases, HHO can find excellent solutions compared to other well-regarded algorithms such as Genetic Algorithm (GA) [31], Biogeography-Based Optimization (BBO) [31], DE [31], PSO [31], Cuckoo Search algorithm (CS) [32], TLBO [33], BA/BAT [24], FPA [34], FA [35], GWO [21], and MFO [22]. Additionally, it can be superior to other optimizers in constrained engineering optimization tasks. HHO has already attracted many attentions in a variety of application fields soon after its occurrence [4, 36–49]. These superiority and meaningful applications reveal that HHO is a powerful enough optimizer. Notwithstanding, as a stochastic bioinspired algorithm, HHO also may undergo the shortcomings of trapping into local optima and degrading convergence rate when tackling some complex tasks. On one hand, it undergoes weak global search capability because of the levy distribution in its optimization process, which makes it difficult to run away from the local optima. On the other hand, the strong randomness in exploration phase may result in skipping the search spots at the edge of search space. Third, the “No-Free-Lunch” (NFL) theorem indicated that no optimizer can be the best general approach for all kinds of problems. With the motivation of these considerations, this study employs the Crisscross Optimization Algorithm (CSO) strategy as an operator to improve HHO. CSO [50] is a well-established stochastic algorithm inspired by crossover behaviors based on the principle of the gold mean in Confucian doctrine. As proposed in its original literature, the simple CSO offers fast convergence rate and high-quality solution with preserving the population diversity when solving not only continuous but also complex optimization problems. A dual crossover search behavior in CSO is employed to establish an interacting chain of horizontal crossover operator and vertical crossover operator. As a global optimization operator, horizontal crossover conducts a crossing over arithmetically on two distinct search agents across all their dimensions, which are, respectively, chosen from the subpopulations without repetition. This operation ensures a larger probability of searching in each subspace and diminishes effectively the unreachable search spots. Vertical crossover is to exchange the dimensional information of a single search agent, which can facilitate escaping from the stagnancy in local optima without destroying other dimensions that may be the global optimum. Considering the advantages of CSO, some papers tried to apply this algorithm on handling complex problems [50–57]. In addition, the excellent search capability of the two crossover operators can exactly make up for the deficiencies of HHO mentioned above. Therefore, this study proposes an improved HHO using these two powerful crossover mechanisms in CSO. The proposed optimizer is called CrissCross Harris Hawks Optimizer (CCHHO). CCHHO contains three phases. After the first initialization phase, the vertical crossover as a regulation mechanism in the second phase interacts with the exploitative behavior of HHO to modify partial search agents for adjusting the exploitative ability adaptively. With the help of dimensional mutation of vertical crossover, the excessive local search capability of the algorithm can be alleviated. In the third phase, the horizontal crossover is considered as an operator to boost explorative trend by diminishing the unreachable search spot. Moreover, to accelerate the convergence speed, the competitive operator conducts greedy selection after performing each crossover. The introduction of three operators in CSO into HHO can enhance the equilibrium between exploratory and exploitative tendencies in case of an accelerated convergence speed and unchanged computational complexity. Furthermore, to access the effectiveness of CCHHO, we take a series of comprehensive experiments from three kinds of issues using 4 categories of benchmark functions, 3 well-regarded engineering optimization tasks and feature selection problems from 13 UCI datasets. On the global optimization tasks of benchmark functions, to show the superiority of the proposed optimizer more rigorously, the performance of the improved optimizer on the global optimization problems of benchmark functions is, respectively, compared with that of nine well-known bioinspired optimizers and that of nine advanced algorithms. 9 well-known bioinspired optimizers contains DE [19], PSO[8], SCA [20], GWO [21], MFO [22], WOA [23], B) [24], GSA [25] and CSO[50], which is used as a strategy in the developed algorithm. 9 advanced algorithms consist of LSHADE_cnEpSin [58], SaDE [59], LSHADE[60], CGPSO [61], CLPSO [62], ALCPSO [63], ACWOA [64], IWOA [65], and CCNMHHO [49], which is the variant of HHO based on CSO and Nelder-Mead simplex. And then the scalability evaluation and sensitivity analysis of CCHHO are conducted. Besides that, the CCHHO optimizer is compared with some other previous engineering optimizers when dealing with engineering optimization problems. On feature selection, the modified HHO is discretized to construct a feature selection method, which is evaluated by comparing with other feature selection techniques including CCNMHHO. The experimental results reveal that the proposed CCHHO is significantly more effective than HHO, CSO, CCNMHHO and other competitors on diverse categories of functions, and its advantage is not influenced by the increase of problems’ dimensions. Additionally, the results also illustrate that the proposed CCHHO outperforms some existing optimizers in working out engineering design optimization; for feature selection problems, it is superior to other feature selection methods including CCNMHHO in terms of fitness, error rate and length of selected features. Notably, the proposed CCHHO in this study and CCNMHHO both adopt the horizontal crossover and vertical crossover of CSO, while CCNMHHO is modified one more strategy, Nelder-Mead simplex, than CCHHO be. Notwithstanding, the results reveal that CCNMHHO is not only less effective than CCHHO on handling the global optimization and feature selection problems but also higher computational expense than CCHHO because of the extra strategy. The main contributions of this paper are summarized as followed:An effective variant of HHO, CCHHO, is presented by introducing the vertical and horizontal crisscross of CSO into the original HHO to handle global optimization, engineering optimization and feature selection problems. The proposed algorithm utilizes the dimensional mutation of vertical crossover in CSO to avoid the stagnancy in local optima, the horizontal crossover in CSO to reduce the unreachable search spots and strengthen the exploratory capability. It improves the balance between exploration and exploitation capability. Furthermore, the convergence rate is accelerated using competitive operator. The performance of the proposed CCHHO is investigated by addressing 4 categories of benchmark functions chosen from 23 classical function and CEC2014 and 4 constrained engineering optimization tasks. The developed CCHHO is adopted to develop a binary feature selection technique for dealing with feature selection problems on 13 datasets. The proposed CCHHO outperforms CCNMHHO, which is based on the strategies of CSO and Nelder-Mead simplex, in terms of performance and computational expense on the global optimization and feature selection problems. The remainder of this paper is organized as follows: Sect. 2 explains the background of HHO and CSO and demonstrates the main structure of the proposed algorithm. And Sect. 3 provides the evaluations of the proposed approach using a set of experiments and analysis on global benchmark problems. The practicality of the proposed method in dealing with engineering optimization problems are verified in Sect. 4. Section 5 gives the application of the proposed optimizer on feature selection. Finally, conclusions and future work are drawn in Sect. 6. Literature Study The advantage of a stochastic optimization algorithm can be determined by whether the algorithm can appropriately harmonize its two significant characteristics: exploration and exploitation [66]. Nevertheless, the randomness of the stochastic optimizers leads to their drawbacks including premature convergence, poor diversity and imbalance between exploratory capability and exploitative capability [9]. Therefore, many studies were devoted to developing a novel excellent algorithm or improving an existing algorithm for solving a variety of optimization problems, which is always a challenging task [67]. As a novel bioinspired algorithm, HHO has been already paid more and more attentions on many application fields. At the very beginning, HHO was applied in the area of energy efficiency. EHHO [36] was first proposed based on chaotic strategy for closing the best solution and opposition-based theory for exploration. It is investigated on parameter estimation of photovoltaic model such as single diode module (SDM), double diode module (DDM), photovoltaic (PV) model and the manufacture’s datasheet. Experimental results demonstrated that EHHO is superior to the well-regarded competitors such as BLPSO, CLPSO, IJAYA and GOTBLBO in terms of not only root mean square error (RMSE) but also convergence speed for identifying the parameters of SDM, DDM and PV model. It also revealed that EHHO can be a beneficial approach to identify the parameters of solar cells in case of some harsh outdoor environment with high irradiance or low temperature. For the same study cases of the PV parameter extraction, Liu et al. [49] proposed an improved HHO, named CCNMHHO, using the Nelder-Mead simplex as well as the horizontal and vertical crossover of the CSO. CCNMHHO employed CSO to improve the population quality by enriching the information exchange between individuals and avoid the local optima by preventing the dimensional stagnation of individuals. Additionally, the Nelder-Mead simplex facilitated enhancing individual searching abilities in aspects of the local search phase and the acceleration of convergence. Experimental results revealed that CCNMHHO was very competitive in estimating the unknown parameters of diverse PV models when comparing to some state-of-the-art methods such as IJAYA, GOTLBO, MLBSA, CPSO, ABSO, ABC, CPSO, EHHO, and so on. And it performed well in solving the complex outdoor environments with diverse temperature and radiance. HHO has been also applied in satellite image. [39] presented a dynamic HHO (DHHO/M) as a new satellite image segmentation approach, which was incorporated with a dynamic control factor and mutation strategy. The proposed DHHO/M was more efficient to improve the search capability and escape from local optimum than HHO. And experimental results indicated that, in the aspects of fitness function evaluation, image segmentation effect and statistical tests, the DHHO/M-based thresholding approach outperformed original HHO, the advanced multilevel thresholding methods including TLBO, WOA-TH, IDSA and BDE as well as the thresholding approaches based on different criteria such as Tsallis entropy based MGOA, MABC, Otsu-based MFPA and GWO. Furthermore, the practicality and feasibility on real engineering problems of DHHO/M was evaluated using four oil pollution images. In addition, there are also some classification model optimizations using HHO. In [41], HHO provided a strengthened performance to optimize the artificial neural network (ANN) for improving the accuracy of ANN in predicting slope stability. Results illustrated that HHO raised the prediction accuracy of ANN in terms of RMSE and mean absolute error, and the correlation between actual values of the safety factor and the outputs of HHO-ANN was more significant than ANN. A HHO-CNN hybrid model for classifying hand gesture images was proposed in [44]. After tuning the hyper-parameter of CNN using HHO, the proposed model attained an accuracy of 100% and was superior to the existing models such as WOA-CNN, GSA-HHO, CSA-HHO, PSO-HHO, GA-HHO, ABC-HHO and GWO-HHO. HHO also can be used in discrete optimization problems. An upgraded binary HHO, HHOSRL, [40] was proposed based on specular reflection learning (SRL) to accurately identify the decisive factors in the early recognition and discrimination of COVID-19 severity. Experimental results showed that the indicators HHOSRL selected, such as age, PaO2, SaO2%, Na+ and LAC, were essential for early accurate estimation of COVID-19 severity. Moreover, HHOSRL performed satisfactorily when fusing with various classifiers and achieved an accuracy of almost 100.0%. The kernel extreme learning machine enabled HHOSRL performed best on blood sample dataset. HHOSRL is the best feature selection method in terms of specificity, sensitivity, accuracy, MCC and time consumption when comparing with other feature selection techniques including bMFO, BGSA, bALO, BSSA, bHHO, bGWO, BPSO, BBA and bWOA. IHHO was proposed in [68], in which salp swarm algorithm was embedded in the update stage of HHO for solving global optimization and feature selection problems. The evaluation results of the proposed IHHO demonstrated that it provided a faster convergence rate and better performance on benchmark function optimization problems than 11 conventional swarm-based algorithms including HHO, DE, GWO, WOA, SSA, MFO, SCA, PSO, MVO, ALO, and GOA as well as 11 state-of-the-art optimizer such as JADE, SaDE, jDE, CoDE, SHADE, ALCPSO, CLPSO, BLPSO, HCLPSO, EPSO and HPSO_TVAC. The IHHO-based feature selection method outperformed BBA, BSSA and BHHO in respect of fitness, feature-length and error rate on ten benchmark feature selection issues. However, from the viewpoint of computational complexity, IHHO may spend much execution time. As set forth, the own characteristics and wide usage of HHO reflect the outstanding optimization capability. As a consequence, this study proposes a variant of HHO named CCHHO using CSO strategy. However, the CSO mechanism was also introduced in CCNMHHO. It should be noted that there are some differences between the proposed HHO and CCNMHHO. First, in addition to the CSO strategy, CCNMHHO used Nelder-Mead simplex as another mechanism. Second, CCHHO and CCNMHHO are not used in the same application field. The former is employed to handle global optimization, engineering optimization and feature selection problems, while the latter is used as a parameter extraction method of photovoltaic models. We will compare their performances in experiments on global optimization and feature selection problems in later sections. Materials and Methods Overview of HHO HHO is a stochastic swarm intelligence optimizer. This optimizer imitates the collaborative behavior of Harris hawks’ swarm in hunting an escaping prey. The mathematical model is mainly based on two explorative phases and four exploitative processes, which are performed randomly. In HHO, the Harris’ hawks represent the search agents; the intended prey indicates the best or approximately optimal solutions obtained so far. In the explorative phases, Harris hawks randomly perch for locating prey. Two explorative phases are modeled based on the positions that hawks roosting. This pattern of exploration boosts the randomness of HHO. Accordingly, it is tending to enable the positions of search agents extend all over possible region of search space. While in the exploitative processes, once detecting the intended prey, the hawks attempt to employ four chasing strategies to cope with various escaping motions of the prey. Four chasing strategies are involved in four exploitative phases. In Appendix C of Supplementary material, Fig. C.1 draws each searching stage in HHO. Meanwhile, the detailed steps including the mathematical model of each stage are described as followed. Step 1: Initialization. The initial parameters such as population size N, maximum number of iterations T are determined. And the hawk population (search agents) X is defined randomly. Step 2: Evaluation. First, make sure that the current agent explores in the search space. In case of this, the best agent Xtarget is determined after evaluating the fitness of each search agent. Step 3: Updating the escaping energy. Due to the reducing escaping energy of the prey, the evolutionary process switches between exploration and exploitation. Obviously, the mutative energy greatly affects the optimization ability of HHO. This time-vary energy, marked as Escaping_energy, is defined according to the following equations [26]:1 Escaping_energy=2E01-tT 2 E0=2rand-1, where E0 represents random initial state, which is changed at each generation within the range -1,1. When Escaping_energy≥1, the search agent is updated by the explorative operation in Step 4, otherwise, when Escaping_energy<1, the search agent is modified by the exploitative operation in Step 5. Step 4: Exploration. The search agent is distributed in various areas of the search space so as to explore the best agent. According to the equal probability q, the position of the search agent is determined to be a random location when q≥0.5, while somewhere near other search agents enough when q<0.5. Therefore, there are two explorative phases expressed as Eq. 3 [26].3 Xt+1=Xrt-r1Xrt-2r2Xt,q≥0.5Xtargett-Xaveraget-r3×r4UB-LB+LB,q<0.5 4 Xaveraget=1N∑i=1NXit, where Xaveraget is the average position of the current search agents, Xit represents the location of the i-th search agent at generation t. Xt+1 and Xt, respectively, indicate the position of search agents at generation t+1 and t. Xrt denotes the position of a random agent. Xtargett is the position of the best agent. r1, r2, r3, r4 and q are produced randomly between 0 and 1 at each generation. Step 5: Exploitation. According to different escaping status of the prey, four strategies for hawks are used in hunting the prey. Therefore, four exploitative processes are constructed to acquire the search agents around the global optimum produced so far. Besides the escaping energy, a random factor r in 0,1 is employed to determine which exploitative process is performed. In addition, Jump_strength represents the prey’s jump strength ought to be updated, it is calculated using the following equation [26]:5 Jumpstrength=2×1-rand The description and execution condition of each process is given below. Based on two factors: Escaping_energy and r, one sub-step in Step 5.1–5.4 is chosen to complete the exploitation. It is remarkable that the soft besiege demonstrated in Step 5.1–5.2 occurs when Escaping_energy≥0.5 and the hard besiege described in Step 5.3–5.4 occurs when Escaping_energy<0.5. Moreover, when r<0.5 such as seeing Step 5.2 and Step 5.4, in both soft and hard besiege, the exploitative process can be the movement with progressive rapid dives, which is more intelligent because of the utilization of Levy flight. If Escaping_energy≥0.5 and r≥0.5, perform Step 5.1; If Escaping_energy≥0.5 and r<0.5, perform Step 5.2; If Escaping_energy<0.5 and r≥0.5, perform Step 5.3; If Escaping_energy<0.5 and r<0.5, perform Step 5.4; Step 5.1: The search agents are estimated as followed [26]:6 Xt+1=ΔXt-Escaping_energy×Jump_strength×Xtargett-Xt 7 ΔXt=Xtargett-Xt, where ΔXt is the difference between the position of the best agent and the current search agents. Step 5.2: The exploitative process can be modeled mathematically such that [26]:8 Y=Xtargett-Escaping_energy×Jump_strength×Xtargett-Xt 9 Z=Y+S×LevyD, where S is a random vector in the search space with size of 1×D and boundary of [0,1]. Levy flight LevyD is used to evaluate the rapid dives [26].10 Levyx=0.01×u×σv1/β,σ=Γ1+βsinπβ/2Γ1+β/2×β×2β-1/21/β, where u, v are random factors and β=1.5. Finally, the search agents can choose their next movements using the following rule [26]:11 Xt+1=Y,ifFY<FXtZ,ifFZ<FXt, where F∙ is the fitness evaluation function. Step 5.3: The search agents are approximate to the best according to the following equation [26]:12 Xt+1=Xtargett-Escapingenergy×ΔXt. Step 5.4: In this process, all of the search agents are tended to be closer to the best agent. The search agent can be evaluated according to Eq. 13 [26].13 Xt+1=Y′,ifFY′<FXtZ′,ifFZ′<FXt, where14 Y′=Xtargett-Escaping_energy×Jump_strength×Xtargett-Xaveraget 15 Z′=Y′+S×LevyD. Step 6: Repeat Steps 2–5 to perform optimization iteratively until the number of iterations has reached the iterative maximum. Step 7: Obtain the optimal solution (Xtarget). Overview of CSO The population-based CSO is an efficient stochastic search optimizer with a crisscross search strategy based on the Confucian doctrine of golden mean. Incorporating a competitive (CP) operator with a dual crossover search behavior, which is different with the crossover in genetic algorithm, CSO algorithm can address the optimization problems, especially complex ones, with high-quality solution and accelerated convergence. In the optimizing process of CSO, the dual crossover search behavior constructs an interacting chain of two searching operators, horizontal crossover (HC) operator and vertical crossover (VC) operator, respectively. HC is performed based on the capability of social recognition among individuals. While VC operation lies in the capability of each individual’s self-recognition. They are alternatively performed with respective probabilities to generate the moderation solutions in each iteration. After each crossover operating, CP operator is performed to estimate the moderation solutions (MS). The MS with better fitness values will be maintained in the new generation. The solutions reproduced in CP operator are referred to as dominant solutions (DS). DS are served as the parent population of the next HC or VC as well as the competitors of the MS stemming from one corresponding crossover operator. It is obvious that the population is modified twice in each iteration by HC and VC, respectively. And CP operator occurs twice to adjust the modified population. The interacting process based on these operators in CSO can be illustrated in Fig. C.2. Each operator in CSO mentioned above is presented specifically as the following subsections. Suppose that the population X involving N search agents is trained in a D-dimensional search space. Then, the i-th search agent can be represented as Xi=Xi,1,Xi,2,⋯,Xi,DT, i=1,2,⋯,N. HC operator HC is a cross-border search mechanism. HC operator as a global optimization operator, randomly divides the population into two parts without repetition at first and then conducts a crossing over arithmetically on two distinct search agents across all their dimensions, which are, respectively, chosen from each subpopulation. After HC operation, each search agent is updated to be MS for HC. The following arithmetical model [50] can be used to express HC operation for a pair of different search agents at the d-th dimension.16 MSHi,d=r1×Xi,d+1-r1×Xj,d+c1×Xi,d-Xj,dMSHj,d=r2×Xj,d+1-r2×Xi,d+c2×Xj,d-Xi,d, where MSHi,d and MSHj,d represent the moderation solutions produced by HC, while Xi,d and Xj,d are their corresponding distinct parent search agents at the d-th dimension, which are employed to perform HC. r1, r2, c1 and c2 are random parameters that distributed uniformly, where the values of r1 and r2 are in 0,1; and c1, c2 are considered as expansion coefficients in the range -1,1. The expansive coefficients are very influential in determining the search scope of a search agent. According to Eq. 16, in HC phase, the multi-dimensional search space is split into two parts. Each part is taken as a hypercube space with the paired parent search agents as their diagonal vertices. As we can see in Eq. 16, each MS contains two components: the first component is originated from the arithmetical crossover in genetic algorithm, which ensures HC searching in each part of the separate hypercubes space with a larger probability; while the second component applies the expansive coefficients, which ensures HC reducing effectively the unreachable blind spots by the first component and searching in the peripheral space of each part with a decreasing probability. Additionally, to adequately reach the global search ability of HC operator, the HC probability Ph that HC operation occurs is set as 1. Within each generation, a CP operator performs a greedy selection after HC operation. The MS achieved by HC compete with the corresponding parents to produce DS as the new modified population. Accordingly, the survived DS are taken as the parent population of the following VC operation. VC operator In VC operation, the information is exchanged between the paired dimensions of a search agent with VC probability Pv. Performed in the opposite direction of HC’s horizontal direction, the VC operator randomly splits the dimensions of each parent search agent into two parts without repetition and then operates a crossing over arithmetically on all search agents between two diverse dimensions. The two distinct dimensions are, respectively, selected from each part of dimensions. The parent population is the population of DS derived from HC operation. After VC operation, each search agent is updated to be MS for VC. Notably, considering that each dimension of search agent exists different units or different boundaries, the normalization and reverse normalization should be performed before and after the VC operation, respectively. Before VC, the normalization operation on the parent search agents makes sure that the dimension offspring can be produced by VC within the boundary of each dimension. The normalized search agent can be calculated as below [50]:17 NXi,j=Xi,j-LBjUBj-LBj, where NXi,j represents the i-th normalized search agent at the j-th dimension. LB and UB are the minimum and maximum of the i-th search agent Xi at the j-th dimension, respectively. VC operation on the paired dimensions of the i-th search agent can be shown as follows [50]:18 NVi,d1=r×NXi,d1+1-r×NXi,d2,d1,d2∈1,D, where NVi,d1 indicates the moderation solution with normalized form generated by VC, while NXi,d1 and NXi,d2 are, respectively, corresponding parent search agents at the d1-th and d2-th dimension, which are adopted to execute VC. r is uniformly distributed parameter whose value is randomly generated in the range 0,1. After VC, the reverse normalization operation can be expressed by the following equation [50]:19 MSVi,j=NVi,j×UBj-LBj-LBj, where MSVi,j represents the moderated solution produced by VC operator. From Eq. 18, VC search happens between two diverse dimensions of a single search agent and generates a single offspring. This exchange of dimensional information dramatically provides an opportunity to avoid dimensional stagnancy in population, hence it makes the stagnant dimension escape from local optimum without destroying other dimensions that may be the global optimum. Since certain dimensions of the population may simultaneously encounter the premature convergence in optimization process, the VC probability Pv, which determines the number of the paired dimensions taking part in VC operation, is set in 0.2,0.8 [50]. Similar to HC, after VC search operation, CP operation is also performed to generate DS as the new offspring population of the current population. It means that the outperformed DS are regarded as the parent population of HC search operation in the next generation. CP Operator As aforementioned, CP operator is applied to choose the better search agent after the population is updated by each crossover search. It plays a vital role in enhancing the search quality. CP operator employs a greedy selection mechanism to update the search agents, which can be expressed as below [50]:20 DS=X,iffXisbetterthanf(MS)MS,iffMSisbetterthanf(X), where DS represents the new modified population for the next crossover operation, which are the better solutions with better fitness between MS generated by HC or VC and its parent X. From the arithmetical model of CP operator, only that MS who are superior to their parents can be survival, otherwise, they would be dropped out of the competition. The CP mechanism can simply make the search agents maintain the better solutions and converge to the global solution rapidly. The Proposed CCHHO In this subsection, with the aid of the separate inherent excellent abilities of the three operators in CSO, we construct an enhanced optimizer, referred as CCHHO, by incorporating the operators into HHO. The structure of the proposed CCHHO is illustrated in Fig. 1, and its corresponding pseudocode is described in Appendix A of Supplementary material. The proposed CCHHO optimizer contains three phases: exploration, exploitation with dynamic adjustment, enhanced exploration. Similar to the evolutionary process of each individual in the conventional HHO, in each generation, the escaping energy determines to perform the first exploration phase or the second exploitation phase. While the third phase optimizes all of the individuals produced from the first two phases in the current generation once more.Fig. 1 Structure of the proposed CCHHO Exploration is the original exploration of HHO, which is divided into two stochastic exploring means based on the equal probability q. The search agents are randomly explored or searched around other search agents. This randomness is contributed to search the global optimal solution in an extensive search space. In this phase, a new exploring population is generated. Exploitation with dynamic adjustment is to adjust the exploitation of the conventional HHO adaptively using a VC regulatory mechanism. The current search agent is modified through one of the four exploitation stages in the conventional HHO and recorded for the subsequent competition. Considering the possibility of premature convergence caused by the Levy flight in the preceding exploitation stages of HHO, VC search mechanism is employed to provide a dimensional mutation dynamically. The dimensional mutation is tended to make the stagnant dimension of a search agent jump out of the local optimum so as to boost the other stagnant dimensions to jump out of the local optimum as soon as possible. VC as an adjustment mechanism is performed in this phase with a certain vertical crossover probability after the preceding exploitation stages. As the architecture of VC search mechanism described in Fig. 2, the dimensions of each current search agent are split into two equal parts. And then a crossover is operated on paired dimensions. There is only part of dimension of the agent updated in resulting search agent. Similar to the VC in CSO, the CP operation is used to compete the resulting search agent with the recorded search agent. An adjusted population consists of the survival search agents. It is obvious that this VC mechanism integrated with CP operator can not only eliminate the local optimality caused by the dimensional stagnancy in previous exploitation stages but also achieve an improved stability between diversification and intensification.Fig. 2 The architecture of VC search mechanism Enhanced exploration is to perform HC search mechanism on all search agents as an enhanced global search operator. This phase is performed after updating the whole population in the first two phases. Therefore, the exploring or adjusted population originated from the above two phases is regarded as the parent population of HC operator. As demonstrated in Fig. 3, the population is split into two equal subpopulations. And a search agent is, respectively, chosen from each of the subpopulations. Each paired search agents are crossed based on HC mechanism. The same as HC in CSO, the CP operator is also introduced to perform a greedy selection between the resulting population of HC and its parent population. And then, the competitive individuals are addressed as the parent population of the next generation. In this phase, HC mechanism effectively facilitates to not only achieve the better solutions in different search spaces but also decrease the inaccessible scotoma, which cannot be obtained in the above two phases, in the solution space so as to boost the ability to search the global optimum. In this case, the excellent explorative ability of HC can fill the gap in the exploration of the conventional HHO. The enhanced exploration with HC mechanism improves the quality of the solution.Fig. 3 The architecture of HC search mechanism Moreover, we can see that the VC and HC with the CP operator introduced in the suitable phases of the evolutionary process of the proposed CCHHO can explore the high-quality global best solution for a faster convergence and a diversified population. The complexity of CCHHO can be calculated based on the following five processes: initialization, exploration phase of HHO, the vertical crossover, horizontal crossover and competitive operator. First, for N search individuals, the computational complexity of initialization isO(N); Secondly, the computational complexity of exploration phase isO(T×N×D); third, the computational complexity of vertical crossover is12O(T×N×D); Fourth, the computational complexity of horizontal crossover is12O(T×N×D); finally, the competitive operator conducted after each crossover mechanism is2O(T×N). Therefore, the total computational complexity of CCHHO is2O(N×(T×D+T+1)), in which T is the maximum number of iterations and D is the dimension of the problem. It can be seen that the introduction of crisscross strategy into HHO makes the computational time increase one time, but does not rise the computational complexity. Evaluations of the Proposed CCHHO In this section, a series of experiments were conducted to investigate the efficiency and strength of incorporated mechanisms in the proposed CCHHO algorithm in a variety of cases. For all approaches in each experimental case, we recorded the fair results due to utilizing an identical experimental setup such as operating environment and parameter settings. All experiments are executed on Windows 10 (64 bit) operating system with Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40 GHz 2.39 GHz (two processors) and 8GBs RAM. The development tool MATLAB R2017a is used to code and run all algorithms. Considering evaluating the effectiveness of CCHHO optimizer from a more comprehensive perspective, the following experiments are conducted. Experiment 1: Evaluating the performance of CCHHO on benchmark function optimization using four categories of benchmark functions. The tested functions contains three unimodal, three multimodal and three fixed-dimension multimodal test functions, which are chosen from 23 classical functions [69], as well as three composition functions available from IEEE2014 [70] test problems. The CCHHO is compared with 9 well-regarded bioinspired algorithms including HHO and CSO as well as 9 state-of-the-art optimizers including CCNMHHO on above 12 test functions. In addition, we perform the analysis of CCHHO’ performance with different dimension, different population size and different maximum iteration number. Experiment 2: Evaluating the performance of CCHHO on engineering optimization problems such as tension/compression spring design, welded beam design and pressure vessel design. The optimization performance of CCHHO on these problems are compared with some existing engineering optimizer as well as CCNMHHO. Experiment 3: Evaluating the CCHHO’s capability of tackling the feature selection problems on 13 UCI datasets. CCHHO as a feature selection approach is compared with 6 feature selection methods also including CCNMHHO. Experiments on Global Optimization Problems CCHHO is compared with 9 well-regarded optimizers and 9 state-of-the-art algorithms on 12 test functions, which are of different categories, including unimodal (F1–F3), multimodal (F4–F6), fixed-dimension multimodal (F7–F9) and composition (F10–F12) test functions. In Appendix B of Supplementary material, Table B.1 and Table B.2 show the mathematical descriptions of these benchmark functions. According to the own characteristics of each type of the problems, these functions were used to validate various properties of the proposed CCHHO algorithm. The unimodal and multimodal sets are suitable for examining the performance of the tested approach in terms of convergency, exploitative capability and explorative capability. Whereas, the rest test sets of CEC 2014 test functions were applied to evaluate the overall abilities of the tested algorithm to equilibrize exploration and exploitation. In each experimental case, every chosen approach runs 30 times independently aiming at weakening the impact of randomness on experimental results. Also, the population size is set to 30. In computational intelligence, a matter of fact is the fairness of computational testing [71]. According this rule, we can make sure the obtained results are justifiably recorded without bias toward one or other algorithm [72]. The setup of initial parameters of all algorithms was identical with their original references. The setup of the parameters in all the comparing algorithms are recorded in Table1.Table 1 Parameter settings for all the comparing algorithms Algorithm Parameters CCHHO c1,c2∈-1,1;ε1,ε2,ε∈0,1;p1=1;p2=0.6 DE Fmin=0.2;Fmax=0.8;CR=0.1 PSO c1=2;c2=2;vMax=6 SCA a=2 GWO a=2,0 MFO b=1;t=-1,1;a∈-1,-2 WOA a1=2,0;a2=-2,-1;b=1 BA a=0.5;r=0.5 GSA G0=100;α=20 CSO c1,c2∈-1,1;ε1,ε2,ε∈0,1;p1=1;p2=0.6 HHO RabbitEnergy=2,0 LSHADE_cnEpSin μF=μCR=μfreq=0.5;H=5;freq=0.5;ps=0.5;pc=0.4 SaDE F1=0.9,Cr1=0.1;F2=0.9,Cr2=0.9;F3=0.5,Cr3=0.3; F4=0.5,Cr4=0.3;F5=0.5,Cr5=0.3 LSHADE rNinit=18;rarc=2.6;p=0.11;H=6 CGPSO c1=c2=1.49618;w=0.7289 CLPSO w=0.2,0.9;c=1.496 ALCPSO c1=c2=2.0;w=0.4;Θ0=60;T=2;pro=1D; Vmax=0.5∗searchrange ACWOA a1=2,0;a2=-2,-1;b=1 IWOA b=1;crossover=0.1 CCNMHHO c1,c2∈-1,1;ε1,ε2,ε∈0,1;p1=1;p2=1 In this study, to estimate the optimization capability of the comparative algorithms, the statistical performance measures such as standard deviation and mean, which can be used to show the robustness of the examined algorithm, are employed to record the experimental results. In the table recording the experimental results, the “mean” and “std” label represent, respectively, the average value and standard deviation of 30 independent runs for each function. Moreover, we employed several extensive statistical significance results to evaluate the success of CCHHO. The Wilcoxon signed rank test at the significant level of 0.05 was applied to analyze statistically the significant difference of statistical measures among the comparative methods. In each Wilcoxon signed rank test result, the results of the significant analysis are shown in the description identified by “ + / = / − “, in which “ + ” represents the number of functions that CCHHO is superior to other tested methods significantly, whereas, “ − ” represents the number of functions that CCHHO is worse than others significantly, “ = ” represents the number of functions that the significant difference does not exist between CCHHO and other comparing methods. For further statistical analysis, we employed the Friedman test to present the average ranking performance of each tested method more evidently. The “ARV” label represents the average performance ranking of the algorithm based on Friedman test for overall benchmark functions. The “rank” label represents the overall ranking of each algorithm. The best results are bolded in the experimental results. Comparison with Well-regarded Bioinspired Algorithms In this subsection, nine well-regarded bioinspired algorithms were used to assess the efficiency of the proposed CCHHO. This comprehensive group of algorithms contains DE [19], PSO [8], SCA [20], GWO [21], MFO [22], WOA [23], BA [24], GSA [25] and CSO [50]. In this experimental case, the maximum number of function evaluations is set to 30,000. Table B.3 expresses the statistical outcomes of comparing the developed CCHHO with the selected methods on tackling 12 benchmark functions. The p-values of Wilcoxon signed rank test analyzing the significant difference between paired methods are presented in Table B.4. Table 2 provides the significant statistical results from Wilcoxon signed-rank test and Friedman test.Table 2 Comparing results between CCHHO and bioinspired algorithms over 12 test functions Metric CCHHO DE PSO SCA GWO MFO + / = / − − / − / − 11/0/1 12/0/0 12/0/0 11/1/0 12/0/0 ARV 1.5292 4.1931 7.8278 7.7000 4.5056 6.7347 Rank 1 3 10 9 4 6 Metric ~ WOA BA GSA CSO + / = /- ~ 11/0/1 12/0/0 12/0/0 7/4/1 ARV ~ 6.0042 6.9139 7.2583 2.3333 Rank ~ 5 7 8 2 From the outcomes of each methods listed in Table B.3 and Table B.4, it is obviously that CCHHO can achieve the best solutions on all of the functions. For F1–F3, the optimal solutions of F1, F2 are able to be found by CCHHO. While other competitors including CSO cannot obtain the optima. It indicates that CCHHO inherits the excellent exploitative ability of original HHO, CSO and other compared optimizers are prone to premature convergence. For F4–F5, CCHHO outperforms other competitors except CSO. There are no significant difference between the performances CCHHO and CSO on F4–F6. This reveals that CCHHO gets the good explorative ability of CSO to escape from local optima. For F7–F9, CCHHO performs worse than CSO and DE on F7, equal with CSO on F9. However, CCHHO also can get the best solution on each function. For F10–F12, CCHHO is significantly superior to other comparing algorithms including CSO, which shows that CCHHO can achieve a better equilibrium between exploration and exploitation than others. Observing from Table 2, the overall statistical outcomes of significance in Wilcoxon signed rank test demonstrate that CCHHO produces 7 significantly better/4 equal/1 significantly worse when compared with CSO, which is the worst case. And CCHHO performs significantly worse than DE, WOA and CSO on only 1 function, equal with CSO and GWO on, respectively, 4 functions and 1 function. This indicates that CCHHO has overwhelmingly advantages over other approaches. It makes sense that CCHHO achieved the best ARV of 1.5292 in the Friedman test, which outperforms CSO in second place with ranking value of 2.3333. Therefore, we can draw the conclusion that the developed CCHHO was the best method with considerable success over 9 well-established bioinspired algorithms. The curves in Fig.C.3 intuitively shows the convergence rate of CCHHO, DE, PSO, SCA, GWO, MFO, WOA, BA, GSA and CSO on addressing eight functions. As we can detected from Fig.C.3, CCHHO converges faster with the best solution than other competitors on all functions except F9. CCHHO achieves a competitive convergence rate on F9. It approaches the best solutions later than CSO, but it attains the best solution, which is not significantly different with the solution obtained by CSO. For other competitors, their accelerated trends are tend to stagnate into local optimum during early evolutionary stage, whereas CCHHO can converge fastest with the high-quality solutions on these optimization tasks. Therefore, based on the enhanced exploratory trends of HC operator, the convergence speed of CCHHO is also boosted. It is obvious that these tendencies can confirm the enhancement of CCHHO in most cases of the unimodal, multimodal and composition problems. In summary, these overall results on each type of problems verify that the proposed HHO has a steadily efficient evolutionary capability at an accelerated convergence rate. This indicates that, in CCHHO, the HC mechanism facilitates strengthening the global search capability; the reasonable adjustments of HC and VC mechanisms are able to successfully make an excellent balance between exploitation and exploration trends; and the CP operator is helpful to boost the convergence rate. Comparison with State-of-the-Art Algorithms In this subsection, the advantages of the proposed CCHHO were investigated by comparing against several state-of-the-art optimization algorithms such as LSHADE_cnEpSin [58], SaDE [59], LSHADE[60], CGPSO [61], CLPSO [62], ALCPSO [63], ACWOA [64], IWOA [65], CCNMHHO [49]. In this experimental case, the maximum number of function evaluations is set to 45,000. Table B.5 records the relevant mathematical results in terms of the statistical metrics, Wilcoxon statistical test and Friedman test. The p values in Table B.6 demonstrate the statistical significance at 5% degree of the Wilcoxon test between the proposed CCHHO and each competitor. Table 3 provides the significant statistical results from Wilcoxon signed-rank test and Friedman test. Figure C.4 visibly illustrates the aforementioned results in terms of the convergence rate.Table 3 Comparing results between CCHHO and state-of-the-art algorithms over 12 test functions Metric CCHHO LSHADE_cnEpSin SaDE LSHADE CGPSO CLPSO + / = /- − / − / − 11/1/0 9/2/1 10/2/0 12/0/0 9/3/0 ARV 2.2861 5.8014 4.9319 5.7625 6.3736 5.8236 Rank 1 5 3 4 8 6 Metric ~ ALCPSO ACWOA IWOA CCNMHHO + / = /- ~ 10/2/0 8/4/0 12/0/0 8/4/0 ARV ~ 6.6514 6.1167 7.3583 3.8944 Rank ~ 9 7 10 2 As observed in Table B.5, CCHHO can exhibit the best performance on all twelve functions except F8. On unimodal cases, the CCHHO has an excellent ability of exploitative search due to finding the optimal solutions on F1, F2 and the best solution on F3, while the ACWOA, as the second top exploitative algorithm on unimodal problems, achieves the optimal solution on F1 and F2. But ACWOA cannot attain a satisfactory solution on F3. Moreover, CCNMHHO is less effective than CCHHO, which indicates that the exploitative ability of CCNMHHO is decreased resulting from the introduction of Nelder-mead simplex. On multimodal cases, the CCHHO obtains the best solutions on all multimodal functions, while the CLPSO and CCNMHHO can achieve the best solutions only on F5 and F6. This observation reveals that the enhanced explorative ability of CCHHO outperforms other methods. In other words, the integrating the Nelder-mead simplex may not boost the explorative capability of CCNMHHO. On the fixed-dimension cases, the performance of CCHHO is competitive to other comparing algorithms. CCHHO obtains the best solution on F7, which is not significantly different with LSHADE_cnEpSin, SaDE, LSHADE, CLPSO, ALCPSO and CCNMHHO. Moreover, according to the standard deviation, CCHHO is more stable than CCNMHHO and ALCPSO. For F8, CCHHO performs worse than SaDE, but it also can attain the best solution. For F9, CCHHO is superior to all other algorithms. For the composition functions, CCHHO is significantly superior to other splendid algorithms including CCNMHHO. As we can see that the coordination between exploration and exploitation of CCHHO is the best of that of all comparing algorithms. According to the significant statistical results in Table 3, CCHHO performs 11/9/10/12/9/10/8/12/8 significantly better, 1/2/2/0/3/2/4/0/4 equal, 0/1/0/0/0/0/0/0/0 significantly worse when comparing with LSHADE_cnEpSin, SaDE, LSHADE, CGPSO, CLPSO, ALCPSO, ACWOA, IWOA, CCNMHHO. Accordingly, from each perspective, the performance of CCHHO outperforms the selected outstanding competitors. Although the ACWOA also has a nice exploitive capability and SaDE and CLPSO can perform a relatively good explorative search, these are only single advantage for them. Furthermore, the results of Friedman test reveal that CCHHO ranks the first among ten algorithms, followed by CCNMHHO, SaDE, LSHADE, LSHADE_cnEpSin, CLPSO, ACWOA, CGPSO, ALCPSO, IWOA. Therefore, although CCNMHHO can achieve an excellent performance, the Nelder-mead simplex in this method weakens its exploitative capability, and then makes the coordinate between exploration and exploitation worse than CCHHO. Figure C.4 exposes the above merits of the proposed CCHHO by means of convergence curves. According to the curves, we detect that the convergence rate of CCHHO is accelerated on several functions including F1, F3, F4, F5, F11. It is also obvious that most of other advanced competitors occurs premature convergence on F1, F3, F4, F5, F11, F12. For F7, F9, F12, CGPSO traps into the local optimum with faster convergence than CCHHO, which can converge to the best solutions. These trends indicate that the convergence rate of CCHHO is accelerated in searching the global solution. Notably, the CCHHO has a faster convergence speed than CCNMHHO on each function. It confirms that CCHHO can effectively harmonizes the high search ability and accelerated convergence rate. The numerical results and convergence curves indicate that CCHHO outperforms the other comparing state-of-the-art optimizers with higher convergence rate. Therefore, it makes sense that the HC and VC mechanisms of CSO are contributed to enhancing the CCHHO’s performance comprehensively and effectively. Accordingly, the comprehensive effectiveness of the proposed CCHHO is the best among all comparing optimizers. First, CCHHO can provide high explorative capability due to combing the horizontal crossover into the later evolutionary stage of HHO. Second, CCHHO has strong exploitation ability resulting from the reasonable integration of the excellent exploitative ability of HHO’s levy flight and the adjustment of vertical crossover. Third, CCHHO attains an accelerated convergence speed. Finally, the experimental results verify that the proposed CCHHO has a suitable coordination between exploration and exploitation with the help of the CSO strategy. Hence, CCHHO has a better effectiveness than CCNMHHO, which is also improved using CSO strategy, on coping with the global optimization problems. In other words, employing more mechanisms may not enhance the optimization performance of the algorithm. In the next subsections, the effectiveness of CCHHO will be verified in some more challenging real problems such as engineering optimization and feature selection. Scalability Evaluation Above subsections presented the merits of the proposed CCHHO by comparing with other well-established and advanced methods. This section provides a scalability evaluation to analyze the impact of dimensions of the optimization task on the efficiency of CCHHO. In this evaluation, the CCHHO algorithm was compared against the original HHO on six benchmark functions with increasing dimensions of 100, 200, 500, 1000 and 2000, respectively. Table B.7 reports the statistical results for each dimension of six optimization cases tackled by CCHHO and HHO. As Table B.7 demonstrated, the comparison results on each dimension are very stable to show the superiority of the CCHHO to HHO. On each dimension, the proposed CCHHO can attain the same optimal or best solutions as HHO when addressing all functions. However, the CCHHO significantly outperforms HHO on the same magnitude when tackling F3, F5, F6. Hence, observing from the standard deviations, CCHHO is significantly more stable than HHO on any dimension. For F1, F2, F4 and F6, both of CCHHO and HHO achieve the optimal or best solution whatever the dimension is. It means that the optimization ability of the proposed CCHHO is enhanced on the problems in different dimensions due to the combination of two crisscross operations. And the success of the proposed method is not affected by the increasing dimensions of the optimization problems. In addition, it can be known that CCHHO produces the best performance when dimension is 2000, and the worst on dimension of 500. Sensitivity Analysis Since majority of metaheuristic algorithms perform on a general idea, rather than provide particular domain knowledge for each problem, diverse values of parameters set for an algorithm may result in diverse performance [73]. It is necessary to identify the parameter settings for algorithm and then can provide robustness and applicability in handling a variety of problems in different scientific areas. The performance of the proposed CCHHO is sensitive to the initial parameter such as the population size (N) and maximum number of iterations (T). A sensitivity analysis is done to examine the impact of these two parameters on CCHHO. For the sensitivity analysis, N is set to 30,50,80,100 and T is set to 100,500,800,1000. The analysis is conducted using these parameters on functions selected from four categories of unimodal, multimodal, fixed-dimension multimodal and composition test functions in Table B.1 and Table B.2. They are, respectively, F3, F5, F8 and F10. For each function, the maximum number of function evaluation is 30000. The sensitivity analyzing results are shown in aspects of mean and convergence rate for the four functions. First, from the perspective of population size (N), CCHHO is simulated for different population size and fixed number of iterations. The iteration number is fixed to 1000. Table B.8 gives the average fitness values of CCHHO on F3, F5, F8, F10 with four different population size. It reveals that CCHHO obtains the best performance with a small population size. Additionally, from the convergence curves of CCHHO on F5 related with different population size in Fig. C. 5(a), the sensitivity of CCHHO decreases with population size. Second, from the perspective of maximum number of iterations (T), CCHHO is executed for different number of iterations. Table B.9 recorded the average fitness values of CCHHO on F3, F5, F8, F10 with four different number of iterations. Figure C. 5(b) presents the convergence curves of CCHHO on F5 with different number of iterations. It can be detected from Table B.9 and Fig.C. 5(b) that CCHHO converges to optimal solution with the increasing of the number of iterations. It shows that the iterations number is essential to the convergence and robustness of CCHHO. Applications to Engineering Optimization Problem The reliability of the proposed CCHHO can be assessed on solving the real-world engineering optimization problems. In this subsection, CCHHO is employed to tackle three well-known engineering optimization tasks, which are tension/compression spring design [74], welded beam design [74] and pressure vessel design [74, 75] problem. The outcomes of optimization for solving each engineering problem were compared to a variety of standard and advanced methods proposed in previous literature. Tension/Compression Spring Design Problem The objective of the tension/compression spring design problem is to minimize the weight of a spring by determine the optima of three structural variables such as wire diameter (d), the number of active coils (N) and mean coil diameter (D). The mathematical definition of this case is as follows [74]:21 Considerx→=x1x2x3=dDN,Min.fx→=x3+2x2x12, Subjecttog1x→=1-x23x371785x14≤0,g2x→=4x22-x1x212566x2x13-x14+15108x12≤0,g3x→=1-140.45x1x22x3≤0,g4x→=x1+x21.5-1≤0 Variables range 0.05≤x1≤2,0.25≤x2≤1.30,2.00≤x3≤15 The tension/compression spring design problem is one of the most ordinary engineering optimization problems. It has already been employed to verify the effectiveness of the new optimizers in many studies. The proposed CCHHO is compared with several recent well-established methods including OBLGOA, BWOA, IFFOA, IGWO, PSOCSCALF, CMSSA, RW-GWO, RGWO, FCMHMD, HHO and CCNMHHO. CCNMHHO is coded for this problem. The results shown in Table 4 point out that the proposed CCHHO and FCMHMD can find the optimal variables for this problem with the weight of 0.1266523. The CCNMHHO, BWOA, HHO, IFFOA, CMSSA, which achieve competitive results, are respectively, ranked in the 2nd, 3rd, 4th, 5th and 6th places.Table 4 Overall results of tension/compression spring design problem Optimizer Optimal variables Optimal weight d N D OBLGOA [3] 0.0530178 9.6001616 0.38953229 0.01270136 BWOA [76] 0.051602 11.2441198 0.357488 0.0126654 IFFOA [77] 0.051657 11.334999 0.355939 0.0126655 IGWO [78] 0.051701 11.275600 0.356983 0.0126670 PSOCSCALF [79] 0.0518836 11.018738 0.36141614 0.012665923 CMSSA [79] 0.051789 11.14985 0.359116 0.0126655 RW-GWO [80] 0.05167 11.33056 0.35613 0.012674 RGWO [9] 0.0514017 11.7065 0.349818 0.012704 FCMHMD [74] 0.0516917 11.285123 0.3567832 0.01266523 HHO[26] 0.051796393 11.138859 0.359305355 0.012665443 CCNMHHO 0.051633912 11.36708866 0.355392446 0.012665288 CCHHO 0.05168488 11.29486632 0.35661714 0.01266523 Welded Beam Design Problem The main aim of this well-known case is to design a welded beam to determine four decision variables that minimize the fabrication cost. The decision variables include the weld thickness (h), length of an attached part of the bar (l), height of the bar (t) and thickness of the bar (b). The mathematical definition of this case can be expressed by the following [74]:22 Considerx→=x1x2x3x4=hltb,Min.fx→=1.10471x12x2+0.04811x3x414.0+x2, Subjecttog1x→=τx→-τmax≤0,g2x→=σx→-σmax≤0,g3x→=δx→-δmax≤0,g4x→=x1-x4≤0,g5x→→=P-Pcx→≤0,g6x→=0.125-x1≤0,g7x→=1.10471x12x2+0.04811x3x414.0+x2-5.0≤0 Variablesrange0.1≤x1,x4≤2,0.1≤x2,x3≤10 Whereτx→=τ′2+2τ′τ″x22R+τ″2,τ′=P2x1x2,τ″=MRJ,M=PL+x22,R=x224+x1+x322,J==22x1x2x224+x1+x322,σx→=6PLx4x32,δx→=6PL3Ex4x32,Pcx→=x32x46364.013EL21-x32LE4G,P=6000lb,L=14in.,δmax=0.25in.,τmax=13,600psi,σmax=30,000psi,E==30×106psi,G=12×106psi The welded beam design problem has been extensively handling by various optimizers such as BSAISA, BWOA, OBLGOA, CMSSA, GDFA, CGWO, VCS, C-ITGO, IAHLO, SFO, SFS, HHO, FCMHMD, SSC and RGWO. To evaluate the ability of the proposed CCHHO in addressing this engineering design case, the estimated variables and optimal cost obtained by this optimizer are compared against those of these optimizers and CCNMHHO. The detailed comparison results listed in Table 5 reveal that the proposed CCHHO algorithm is able to identify the best design for this engineering case by attaining the optimal cost of 1.69526617. This cost is the lower than that of other algorithms. CCNMHHO finds the optimal cost of 1.69527319, which is higher than CCHHO.Table 5 Overall results of welded beam design problem Optimizer Optimal variables Optimal cost h l t b BSAISA [81] 0.205730 3.470489 9.036624 0.205730 1.724852 BWOA [76] 0.205829 3.251922 9.034556 0.205829 1.695620 OBLGOA [3] 0.205769 3.471135 9.032728 0.2059072 1.7257 CMSSA [82] 0.205113 3.264134 9.036661 0.205729 1.695838 GDFA [83] 0.205670 3.472000 9.036680 0.205730 1.724700 CGWO [84] 0.343891 1.883570 9.03133 0.212121 1.725450 VCS [85] 0.20572964 3.47048867 9.03662391 0.20572964 1.72485231 C-ITGO [86] 0.2057296 3.4704886 9.0366239 0.2057296 1.7248523 IAHLO [87] 0.205730 3.470489 9.033624 0.205730 1.724852 SFO [66] 0.2038 3.6630 9.0506 0.2064 1.73231 SFS [88] 0.20572964 3.47048867 9.03662391 0.20572964 1.72485231 HHO[26] 0.204039 3.531061 9.027463 0.206147 1.73199057 FCMHMD [74] 0.2056756 3.4700936 9.043817 0.205693 1.725684 SSC [89] 0.1992 3.4307 9.1045 0.2051 1.7222 RGWO [9] 0.20569 3.4718 9.0365 0.20574 1.7253 CCNMHHO 0.20571349 3.253420735 9.036575827 0.205731928 1.69527319 CCHHO 0.20572195 3.253272321 9.03656748 0.20573221 1.69526617 Pressure Vessel Design Problem The main purpose of pressure vessel design problem is to design the cylindrical pressure vessel with a minimum total cost, which contains the cost of materials and welding. To tackle this case, four parameters (Ts, Th, R, L) must be optimized. Ts and Th denote the thickness of the shell and head, respectively; R is the inner radius; L represents the length of the cylindrical section of the vessel without head. The mathematical model of this case can be defined as follows [74]:23 Considerx→=x1x2x3x4=TsThRL,Min.f(x→)=0.6224x1x3x4+1.7881x2x32+3.1661x12x4+19.84x12x3, Subjecttog1x→=-x1+0.0193x3≤0,g2x→=-x12+0.00954x3≤0,g3x→=-πx32x4-43πx33+1296000≤0,g4x→=x4-240≤0 Variablesrange0≤x1,x2≤99,10≤x3,x4≤200 Some of optimization algorithms previously utilized to tackle the pressure vessel design problem include OBLGOA, RW-GWO, OBSCA, MHDA, WOA, HHO and FCGHMD. The performance of the CCHHO algorithm on optimizing this design problem is compared with that of these algorithms and CCNMHHO. The comparison outcomes in Table 6 report the optimal cost of 5826.47696325. It reveals meaningfully that the CCHHO is the most effective optimizer among above well-established ones in this engineering design case.Table 6 Overall results of pressure vessel design problem Optimizer Optimal variables Optimal cost Th Ts R L OBLGOA [3] 0.81622 0.40350 42.291138 174.811191 5966.67160 RW-GWO[80] 0.81250 0.43750 42.09840 176.63784 6059.736 OBSCA[90] 1.2500 0.0625 59.1593 70.8437 5833.9892 MHDA[91] 0.778169 0.384649 40.3196 200 5885.3353 WOA [23] 0 0.812500 0 0.437500 42 0.0982699 176 0.638998 6059 0.7410 HHO [26] 0.81758383 0.4072927 42.09174576 176.7196352 6000.46259 FCGHMD[74] 0.8129 0.401571 42.087015 176.85321 5952.89502 CCNMHHO 0.942465189 0.46586103 48.8323932 107.8873575 5833.620039 CCHHO 0.94877596 0.46898044 49.15937475 105.15766782 5826.47696325 According to the analysis of the above three constrained engineering optimization tasks, the proposed CCHHO has confirmed its efficiency in addressing real-world engineering optimization problems over other well-regarded optimizers. It indicates that the superiority of CCHHO results from the HC and VC operators that interact reasonably with the HHO mechanism. Also, due to the constrained property and uncertain search space in the engineering optimization problems, we can see that CCHHO can perform optimization in the search domain with infeasible spaces. Application to Feature Selection Feature selection (FS), as an optimization process of dimensionality reduction, is an essential stage of addressing the high-dimensional search space in classification problem. In this section, the proposed CCHHO is used in dealing with this challenging process for further assessing the availability of CCHHO in real-word application. FS problem aims at determining a subset with the fewest representative features from the overall set of properties in original dataset to achieve an optimal classification accuracy. Therefore, as the architecture of FS based on a binary optimizer illustrated in Fig. 4, it is observed that the classification accuracy of the classifier can be considered as the measurement of validating the ability of dimensionality reduction.Fig. 4 Architecture of FS based on a binary optimizer FS Based on Binary CCHHO The FS problem can be taken as a discrete optimization problem. Therefore, the proposed CCHHO is transformed into a binary version, called BCCHHO, since the solved FS task is characterized as a binary optimization problem. Hence, in the implementing process of CCHHO for solving FS problem, the continuous solutions are mapped into discrete forms with binary value. The solution ought to be represented as a binary vector with the same dimensions as the number of features in the training dataset. In the vector with binary values, the selected attribute is labelled with value 1 at its corresponding location, whereas the non-selected attribute is marked with value 0. First, we produced binary values in accordance with random thresholds to initialize the population as followed:24 xi,j=0rand≤0.51rand>0.5, where xi,j denotes the binary value in the location vector of search agent at i-th row and j-th column. Second, without influencing the architecture of CCHHO, we can employ the transfer function to map the continuous vector of search agent into binary form in the formal transformation. In this study, we chose the S-shaped transfer function [92], shown as Eq. 25, to perform the mapping to squash the continuous vectors in each dimension.25 STrans=11+e-x3, where x represents the continuous vector of search agent. S_Trans denotes a continuous intermediate form as the output of the transfer function. To achieve a binary vector, S_Trans is changed and compared with the initial binary vector generating in the initialization based on the following equation [68]:26 xi,j=outputPos=∼initialPosrand≤S_TransoutputPos=initialPosrand>S_Trans, where initialPos is the initial binary value, outputPos indicates the new binary value attained. Finally, we applied the classifier to evaluate the selected subset of features obtained from BCCHHO. FS is evidently regarded as a multi-objectives problem since it involves acquiring the highest classification accuracy with fewest feature subsets. Therefore, considering the objectives comprehensively, we can evaluate the search agent in accordance with the classification accuracy and the number of chosen features using the fitness function as follows [68]:27 fitness=α×Ns/N+β×1-accuracy, where Ns denotes the number of features filtered by BCCHHO. N is the total number of features in original dataset. accuracy represents the classification accuracy calculated from the classifier. 1-accuracy indicates the error rate. α and β are considered as the weights of the selected features’ number and classification accuracy, respectively, α∈0,1, β=1-α. They represent the importance of the selected features’ number and error rate. Experimental Setup We present a comparative study to examine the optimization behavior of the binary CCHHO compared to several state-of-the-art bioinspired algorithms including BHHO (binary HHO [26]), bGWO [93], BBA [94], bWOA (binary WOA [23]), BSSA (binary salp swarm algorithm [95]) and BCCNMHHO, which is the binary versions of CCNMHHO [49] improved using CSO and Nelder-mead simplex. Thirteen practical benchmark datasets on distinct subject areas were utilized as case studies. They are available from the UCI repository (https://archive.ics.uci.edu/ml//datasets.php, available at October 2022) and the Wielaw dataset is from the literature [96]. As presented in Table B.10, in these datasets, the sizes of instances and features are distinguished with each other. That is beneficial to assess the proposed method from distinct perspectives. To alleviate the bias of feature selection, in the training of classification, k-fold cross-validation (CV) is employed to assess the optimality of each selected feature subset. The dataset in k-fold CV was segmented into k equivalent parts. One part in the dataset is used as the testing set to estimate the classifier’s classification accuracy, while the k–1 parts is the training set for training the classifier. The evaluation metrics of FS are the average values obtained from k validations for each problem. For fairness in the investigation, each evaluation associated with the BCCHHO is performed in the same computational environment. The initial parameters of all comparison methods were identical with that in their original references. Furthermore, for each involved method, the search agents and maximum iteration are, respectively, set as 20 and 50. It runs ten independent times. And the common KNN was used as classifier in training. The fold k in k-fold CV was set as 10, and α is 0.05 [68, 97]. We adopt four evaluation metrics to express the experimental results. They are the average fitness, average error rate, average number of the selected features and average computational time. Moreover, the corresponding standard deviation (std) is accompanied in the results to estimate the performances of the investigated FS approaches on thirteen datasets on ten independent runs. According to the selected features subsets in the evaluated datasets, the average fitness and average error rate are attained. Additionally, the average ranking value (ARV) in aspects of the evaluation metrics based on the Friedman test are employed to detect the optimal of solutions among the examined optimizers. The best values of four evaluation metrics are bold in the results. Results and Discussion The simulation results attained by BCCHHO against other competitors are recorded in Table B.11, 12, 13, 14 in aspects of average fitness, error rate, numbers of selected features as well as computational time, respectively. Tables 7, 8, 9, 10 demonstrate the average ranking results of all investigated optimizers based on Friedman Test. Observing the results of average fitness in Table B.11, BCCHHO shows competitive outcomes of average fitness. The proposed optimizer can attain the best average fitness on 7 of 13 datasets, while BCCNMHHO and bGWO get the best solution on 5 and 4 datasets, respectively. However, the average fitness obtained by BCCHHO are very close to the best value. Considering Table B.11, 12, 13, BCCHHO can achieve the lowest error rate with relatively fewer features on Exactly, M-of-n, WineEW, Zoo, vehicle, wdbc and Wielaw. On other datasets, bGWO and BCCNMHHO can obtain the best fitness. Nevertheless, the lowest error rates are mostly attained by bGWO at the cost of selecting more features. And BCCNMHHO cannot realize the lowest error rates, which indicates that BCCNMHHO have missed the important information that affect the accuracy because of selecting the fewest features. This also can be concluded from Tables 7, 8, 9. According the average ranking value based on Friedman test, BCCHHO is ranked first in terms of average fitness and error rate. BCCNMHHO ranks in the first position in terms of feature-length, but it cannot outperform BCCHHO in terms of the two important metrics such as fitness and error rate.Table 7 Comparison results of the proposed BCCHHO vs. other optimizers in aspect of fitness Dataset BCCHHO BHHO bGWO BBA bWOA BSSA BCCNMHHO ARV 1.9385 6.4846 2.4308 6.5077 4.0308 4.6115 1.9962 Rank 1 6 3 7 4 5 2 Table 8 Comparison results of the proposed BCCHHO vs. other optimizers in aspect of error rate Dataset BCCHHO BHHO bGWO BBA bWOA BSSA BCCNMHHO ARV 2.2538 5.7308 2.3462 7 4.2346 4.0500 2.3846 Rank 1 6 2 7 5 4 3 Table 9 Comparison results of the proposed BCCHHO vs. other optimizers in aspect of feature-length Dataset BCCHHO BHHO bGWO BBA bWOA BSSA BCCNMHHO ARV 2.3923 5.7077 3.5654 5.2423 2.8731 5.8615 2.3577 Rank 2 6 4 5 3 7 1 Table 10 Comparison results of the proposed BCCHHO vs. other optimizers in aspect of computational time Dataset BCCHHO BHHO bGWO BBA bWOA BSSA BCCNMHHO ARV 5.8462 5.1538 2.3000 3.1308 1.6077 2.9615 7 Rank 6 5 2 4 1 3 7 Inspecting the computational time shown in Table B.14 and Table 10, BCCNMHHO expends the most running time among the tested algorithms, followed by BCCHHO, BHHO. However, as the variants of the original HHO, BCCNMHHO takes more than three times as long as that of HHO. It can be known that the time expend of BCCNMHHO is influenced by the introduced Nelder-mead simplex and CSO strategies. The computational expend of BCCHHO is the second most expensive, nevertheless, it is close to that of HHO and less half of BCCNMHHO’s time cost. Moreover, the algorithms associated with HHO spends more time than others. It indicates that the time cost of the modified algorithm depends on not only the introduced mechanism but also its original algorithm. As summary, the BCCHHO can achieve the best fitness and the lowest error rate with relatively fewer features on the datasets with low-dimensional as well as high-dimensional datasets. This indicates that the proposed BCCHHO provides a remarkable success among the tested FS techniques. It can be obviously concluded that the proposed BCCHHO considerably overcomes the limitations of the original version of BHHO by integrating the HC and VC operators. This optimizer is suite for the discrete feature selection problem. Additionally, the above analysis of the computational time reveals that the computational time of an algorithm is affected by its mechanism of finding optimum. Conclusions and Future Directions In this study, an enhanced HHO (CCHHO) optimizer was proposed, in which the horizontal crossover strategy in CSO is fused with the update stage of HHO, the vertical crossover strategy in CSO is introduced into the exploitation phase of HHO and the competitive operator in CSO is also performed after each crossover. The combination of CSO in HHO ensures not only a good balance between local search capability and global search ability but also an accelerated convergence speed. The CCHHO’s performance is comprehensively assessed on four categories of functions. The results show that CCHHO is more efficient than other well-known bioinspired algorithms and advanced optimizers in terms of convergence and the balance between exploitation and exploration. Simultaneously, on solving the engineering optimization problems, the CCHHO outperforms other previous optimizers to lower manufacturing costs. Hence, for the FS problems, binary CCHHO achieves higher accuracy by selecting fewer features than other competitors. In a word, CCHHO can handle continuous global optimization problems and discrete feature selection with an outstanding effectiveness. Additionally, we detect that CCHHO is superior to CCNMHHO on solving the above problems. It indicates that the Nelder-mead simplex introduced in CCNMHHO may not be contributed to achieving the best efficiency when addressing these problems, while it increases the algorithm’s computational time. Consequently, it is not always the case that more strategies added into an algorithm can make it perform better. Moreover, an algorithm cannot be a universal method for solving every problem, which is proven by “NFL” theorem. Accordingly, we can attempt to evaluate what other problems the CCHHO can work out. In this study, we adopt the CCHHO in single objective problems and low-dimensional feature selection. As a future study, the CCHHO can be used for multi-objective optimization problems and high-dimensional feature selection such as gene selection. Additionally, due to its verified excellent capability of optimization and applicability in feature selection, we can also employ it to enhance the classifier’s accuracy by optimizing its key parameters and simultaneously selecting the optimal feature subsets. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 2639 kb) 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 they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References 1. Gupta S Deep K Improved sine cosine algorithm with crossover scheme for global optimization Knowledge Based Syst 2019 165 374 406 10.1016/j.knosys.2018.12.008 2. Gharehchopogh FS Gholizadeh H A comprehensive survey: whale optimization algorithm and its applications Swarm Evolution Comput 2019 48 1 24 10.1016/j.swevo.2019.03.004 3. Ewees AA Abd Elaziz M Houssein EH Improved grasshopper optimization algorithm using opposition-based learning Expert Systems with Applications 2018 112 156 172 10.1016/j.eswa.2018.06.023 4. Long W Jiao J Xu M Tang M Wu T Cai S Lens-imaging learning Harris hawks optimizer for global optimization and its application to feature selection Expert Systems with Applications 2022 202 117255 10.1016/j.eswa.2022.117255 5. Beheshti Z BMPA-TVSinV: a binary marine predators algorithm using time-varying sine and V-shaped transfer functions for wrapper-based feature selection Knowledge Based Syst 2022 252 109446 10.1016/j.knosys.2022.109446 6. Zhu S Wu Q Jiang Y Xing W A novel multi-objective group teaching optimization algorithm and its application to engineering design Comput Indust Eng 2021 155 107198 10.1016/j.cie.2021.107198 7. Preeti, & Deep, K. A random walk Grey wolf optimizer based on dispersion factor for feature selection on chronic disease prediction Expert Systems with Applications 2022 206 117864 10.1016/j.eswa.2022.117864 8. Pramanik R Sarkar S Sarkar R An adaptive and altruistic PSO-based deep feature selection method for Pneumonia detection from chest X-rays Applied Soft Computing 2022 128 109464 10.1016/j.asoc.2022.109464 35966452 9. Banaie-Dezfouli M Nadimi-Shahraki MH Beheshti Z R-GWO: representative-based grey wolf optimizer for solving engineering problems Applied Soft Computing 2021 106 107328 10.1016/j.asoc.2021.107328 10. Xu Y Chen H Luo J Zhang Q Jiao S Zhang X Enhanced moth-flame optimizer with mutation strategy for global optimization Information Sciences 2019 492 181 203 10.1016/j.ins.2019.04.022 11. Tian X Li J A novel improved fruit fly optimization algorithm for aerodynamic shape design optimization Knowledge Based Syst 2019 179 77 91 10.1016/j.knosys.2019.05.005 12. Arora S Anand P Binary butterfly optimization approaches for feature selection Expert Systems with Applications 2019 116 147 160 10.1016/j.eswa.2018.08.051 13. Jadhav AN Gomathi N WGC: Hybridization of exponential grey wolf optimizer with whale optimization for data clustering Alexand Eng J 2018 57 1569 1584 10.1016/j.aej.2017.04.013 14. Abbassi R Abbassi A Heidari AA Mirjalili S An efficient salp swarm-inspired algorithm for parameters identification of photovoltaic cell models Energy Convers Manag 2019 179 362 372 10.1016/j.enconman.2018.10.069 15. Ateya AA Muthanna A Vybornova A Algarni AD Abuarqoub A Koucheryavy Y Koucheryavy A Chaotic salp swarm algorithm for SDN multi-controller networks Eng Sci Tech Intern J 2019 22 1001 1012 10.1016/j.jestch.2018.12.015 16. Faris H Mafarja MM Heidari AA Aljarah I Al-Zoubi AM Mirjalili S Fujita H An efficient binary Salp Swarm Algorithm with crossover scheme for feature selection problems Knowledge Based Syst 2018 154 43 67 10.1016/j.knosys.2018.05.009 17. Syed MA Syed R Weighted Salp Swarm Algorithm and its applications towards optimal sensor deployment J King Saud Univ Comput Inform Sci 2022 34 1285 1295 10.1016/j.jksuci.2019.07.005 18. Yang B Zhong L Zhang X Shu H Yu T Li H Sun L Novel bio-inspired memetic salp swarm algorithm and application to MPPT for PV systems considering partial shading condition J Clean Product 2019 215 1203 1222 10.1016/j.jclepro.2019.01.150 19. Kitayama S Arakawa M Yamazaki K Differential evolution as the global optimization technique and its application to structural optimization Applied Soft Computing 2011 11 3792 3803 10.1016/j.asoc.2011.02.012 20. Mirjalili S SCA: a sine cosine algorithm for solving optimization problems Knowledge Based Syst 2016 96 120 133 10.1016/j.knosys.2015.12.022 21. Mirjalili S Mirjalili SM Lewis A Grey wolf optimizer Advanc Eng Software 2014 69 46 61 10.1016/j.advengsoft.2013.12.007 22. Mirjalili S Moth-flame optimization algorithm: a novel nature-inspired heuristic paradigm Knowledge Based Syst 2015 89 228 249 10.1016/j.knosys.2015.07.006 23. Mirjalili S Lewis A The whale optimization algorithm Advanc Eng Software 2016 95 51 67 10.1016/j.advengsoft.2016.01.008 24. Yang XS Hossein Gandomi A Bat algorithm: A novel approach for global engineering optimization Engineering Computations 2012 29 464 483 10.1108/02644401211235834 25. Wu P Wang H Li B Fu W Ren J He Q Disassembly sequence planning and application using simplified discrete gravitational search algorithm for equipment maintenance in hydropower station Expert Systems with Applications 2022 208 118046 10.1016/j.eswa.2022.118046 26. Heidari AA Mirjalili S Faris H Aljarah I Mafarja M Chen H Harris hawks optimization: algorithm and applications Future Generation Computer Syst 2019 97 849 872 10.1016/j.future.2019.02.028 27. Li S Chen H Wang M Heidari AA Mirjalili S Slime mould algorithm: a new method for stochastic optimization Future Generation Computer Syst 2020 111 300 323 10.1016/j.future.2020.03.055 28. Zhong C Li G Meng Z Beluga whale optimization: a novel nature-inspired metaheuristic algorithm Knowledge Based Syst 2022 251 109215 10.1016/j.knosys.2022.109215 29. Zhao S Zhang T Ma S Chen M Dandelion optimizer: a nature-inspired metaheuristic algorithm for engineering applications Eng Appl Artific Intell 2022 114 105075 10.1016/j.engappai.2022.105075 30. Faramarzi A Heidarinejad M Mirjalili S Gandomi AH Marine predators algorithm: a nature-inspired metaheuristic Expert Systems with Applications 2020 152 113377 10.1016/j.eswa.2020.113377 31. Simon D Biogeography-based optimization IEEE Trans Evolution Comput 2008 12 702 713 10.1109/TEVC.2008.919004 32. Gandomi AH Yang X-S Alavi AH Cuckoo search algorithm: a metaheuristic approach to solve structural optimization problems Engineering Computations 2011 29 17 35 10.1007/s00366-011-0241-y 33. Rao RV Savsani VJ Vakharia DP Teaching–learning-based optimization: a novel method for constrained mechanical design optimization problems Computer Aided Design 2011 43 303 315 10.1016/j.cad.2010.12.015 34. Yang X-S Karamanoglu M He X Flower pollination algorithm: a novel approach for multiobjective optimization Eng Optimizat 2013 46 1222 1237 10.1080/0305215x.2013.832237 35. Gandomi AH Yang X-S Alavi AH Mixed variable structural optimization using Firefly Algorithm Computers & Structures 2011 89 2325 2336 10.1016/j.compstruc.2011.08.002 36. Chen H Jiao S Wang M Heidari AA Zhao X Parameters identification of photovoltaic cells and modules using diversification-enriched Harris hawks optimization with chaotic drifts J Clean Produc 2020 244 118778 10.1016/j.jclepro.2019.118778 37. Song S Wang P Heidari AA Zhao X Chen H Adaptive Harris hawks optimization with persistent trigonometric differences for photovoltaic model parameter extraction Eng Appl Artific Intell 2022 109 104608 10.1016/j.engappai.2021.104608 38. Abdel Aleem SHE Zobaa AF Balci ME Ismael SM Harmonic overloading minimization of frequency-dependent components in harmonics polluted distribution systems using harris hawks optimization algorithm IEEE Access 2019 7 100824 100837 10.1109/access.2019.2930831 39. Jia H Lang C Oliva D Song W Peng X Dynamic Harris Hawks optimization with mutation mechanism for satellite image segmentation Remote Sensing 2019 11 1421 10.3390/rs11121421 40. Hu J Han Z Heidari AA Shou Y Ye H Wang L Wu P Detection of COVID-19 severity using blood gas analysis parameters and Harris hawks optimized extreme learning machine Comput Biology Med 2022 142 105166 10.1016/j.compbiomed.2021.105166 41. Moayedi H Osouli A Nguyen H Rashid ASA A novel Harris hawks’ optimization and k-fold cross-validation predicting slope stability Engineering Computations 2021 37 369 379 10.1007/s00366-019-00828-8 42. Ramachandran M Mirjalili S Nazari-Heris M Parvathysankar DS Sundaram A Charles Gnanakkan CAR A hybrid grasshopper optimization algorithm and harris hawks optimizer for combined heat and power economic dispatch problem Eng Appl Artific Intell 2022 111 104753 10.1016/j.engappai.2022.104753 43. Issa M Samn A Passive vehicle suspension system optimization using Harris Hawk optimization algorithm Mathematic Comput Simulat 2022 191 328 345 10.1016/j.matcom.2021.08.016 44. Gadekallu TR Srivastava G Liyanage MMI Chowdhary CL Koppu S Maddikunta PKR Hand gesture recognition based on a Harris Hawks optimized convolution neural network Comput Electric Eng 2022 100 107836 10.1016/j.compeleceng.2022.107836 45. Suresh T Brijet Z Blesslin Sheeba T CMVHHO-DKMLC: a chaotic multi verse Harris Hawks optimization (CMV-HHO) algorithm based deep kernel optimized machine learning classifier for medical diagnosis Biomed Sig Proc Control 2021 70 103034 10.1016/j.bspc.2021.103034 46. Jangir P Heidari AA Chen H Elitist non-dominated sorting Harris hawks optimization: framework and developments for multi-objective problems Expert Systems with Applications 2021 186 115747 10.1016/j.eswa.2021.115747 47. Balaha HM El-Gendy EM Saafan MM CovH2SD: a COVID-19 detection approach based on Harris Hawks Optimization and stacked deep learning Expert Systems with Applications 2021 186 115805 10.1016/j.eswa.2021.115805 34511738 48. Kamboj VK Nandi A Bhadoria A Sehgal S An intensify Harris Hawks optimizer for numerical and engineering optimization problems Applied Soft Computing 2020 89 106018 10.1016/j.asoc.2019.106018 49. Liu Y Chong G Heidari AA Chen H Liang G Ye X Wang M Horizontal and vertical crossover of Harris hawk optimizer with Nelder-Mead simplex for parameter estimation of photovoltaic models Energy Convers Manag 2020 223 113211 10.1016/j.enconman.2020.113211 50. Meng A-B Chen Y-C Yin H Chen S-Z Crisscross optimization algorithm and its application Knowledge Based Syst 2014 67 218 229 10.1016/j.knosys.2014.05.004 51. Meng A Chen S Ou Z Ding W Zhou H Fan J Yin H A hybrid deep learning architecture for wind power prediction based on bi-attention mechanism and crisscross optimization Energy 2022 238 121795 10.1016/j.energy.2021.121795 52. Weng S Tan W Ou B Pan J-S Reversible data hiding method for multi-histogram point selection based on improved crisscross optimization algorithm Information Sciences 2021 549 13 33 10.1016/j.ins.2020.10.063 53. Meng A Zeng C Wang P Chen D Zhou T Zheng X Yin H A high-performance crisscross search based grey wolf optimizer for solving optimal power flow problem Energy 2021 225 120211 10.1016/j.energy.2021.120211 54. Kumar M Dhillon JS A conglomerated ion-motion and crisscross search optimizer for electric power load dispatch Applied Soft Computing 2019 83 105641 10.1016/j.asoc.2019.105641 55. Yin H Wu F Meng X Lin Y Fan J Meng A Crisscross optimization based short-term hydrothermal generation scheduling with cascaded reservoirs Energy 2020 203 117822 10.1016/j.energy.2020.117822 56. Kaur M Dhillon JS Kothari DP Crisscross differential evolution algorithm for constrained hydrothermal scheduling Applied Soft Computing 2020 93 106393 10.1016/j.asoc.2020.106393 57. Patwal RS Narang N Crisscross PSO algorithm for multi-objective generation scheduling of pumped storage hydrothermal system incorporating solar units Energy Conversion and Management 2018 169 238 254 10.1016/j.enconman.2018.05.067 58. Awad NH Ali MZ Suganthan PN Ensemble sinusoidal differential covariance matrix adaptation with euclidean neighborhood for solving CEC2017 benchmark problems. IEEE congress on evolutionary computation (CEC) Donostia Spain 2017 10.1109/CEC.2017.7969336 59. Qin AK Huang VL Suganthan PN Differential evolution algorithm with strategy adaptation for global numerical optimization IEEE Trans Evolution Comput 2009 13 398 417 10.1109/tevc.2008.927706 60. Tanabe R Fukunaga AS Improving the search performance of SHADE using linear population size reduction. 2014 IEEE congress on evolutionary computation (CEC) Beijing China 2014 2014 1658 1665 10.1109/CEC.2014.6900380 61. Jia D Zheng G Qu B Khan MK A hybrid particle swarm optimization algorithm for high-dimensional problems Comput Indust Eng 2011 61 1117 1122 10.1016/j.cie.2011.06.024 62. Liang JJ Qin AK Suganthan PN Baskar S Comprehensive learning particle swarm optimizer for global optimization of multimodal functions IEEE Trans Evolution Comput 2006 10 281 295 10.1109/tevc.2005.857610 63. Chen W-N Zhang J Lin Y Chen N Zhan Z-H Chung HS-H Shi Y-H Particle swarm optimization with an aging leader and challengers IEEE Trans Evolution Comput 2013 17 241 258 10.1109/tevc.2011.2173577 64. Elhosseini MA Haikal AY Badawy M Khashan N Biped robot stability based on an A-C parametric whale optimization algorithm J Computation Sci 2019 31 17 32 10.1016/j.jocs.2018.12.005 65. Tubishat M Abushariah MAM Idris N Aljarah I Improved whale optimization algorithm for feature selection in Arabic sentiment analysis Appl Intel 2019 49 1688 1707 10.1007/s10489-018-1334-8 66. Shadravan S Naji HR Bardsiri VK The sailfish optimizer: a novel nature-inspired metaheuristic algorithm for solving constrained engineering optimization problems Eng Appl Art Intel 2019 80 20 34 10.1016/j.engappai.2019.01.001 67. Chen H Jiao S Heidari AA Wang M Chen X Zhao X An opposition-based sine cosine approach with local search for parameter estimation of photovoltaic models Energy Conversion and Management 2019 195 927 942 10.1016/j.enconman.2019.05.057 68. Zhang Y Liu R Wang X Chen H Li C Boosted binary Harris hawks optimizer and feature selection Engineering Computations 2021 37 3741 3770 10.1007/s00366-020-01028-5 69. Yao X Liu Y Lin G Evolutionary programming made faster IEEE Trans Evolution Comput 1999 3 82 102 10.1109/4235.771163 70. Liang JJ, Qu BY, Suganthan PN (2013) Problem Definitions and Evaluation Criteria for the CEC 2014 Special Session and Competition on Single Objective Real-Parameter Numerical Optimization. Technical Report 201311, Computational Intelligence Laboratory, Zhengzhou University, Zhengzhou China and Technical Report, Nanyang Technological University, Singapore. 71. Shi K Wang J Zhong S Tang Y Cheng J Non-fragile memory filtering of T-S fuzzy delayed neural networks based on switched fuzzy sampled-data control Fuzzy Sets and Systems 2020 394 40 64 10.1016/j.fss.2019.09.001 72. Zhang H Wang Z Chen W Heidari AA Wang M Zhao X Zhang X Ensemble mutation-driven salp swarm algorithm with restart mechanism: framework and fundamental analysis Expert Systems with Applications 2021 165 113897 10.1016/j.eswa.2020.113897 73. Braik MS Chameleon swarm algorithm: a bio-inspired optimizer for solving engineering design problems Expert Systems with Applications 2021 174 114685 10.1016/j.eswa.2021.114685 74. Abd Elaziz M Yousri D Mirjalili S A hybrid Harris hawks-moth-flame optimization algorithm including fractional-order chaos maps and evolutionary population dynamics Advanc Eng Software 2021 154 102973 10.1016/j.advengsoft.2021.102973 75. Kannan BK Kramer SN An augmented lagrange multiplier based method for mixed integer discrete continuous optimization and its applications to mechanical design J Mech Design 1994 116 405 411 10.1115/1.2919393 76. Chen H Xu Y Wang M Zhao X A balanced whale optimization algorithm for constrained engineering design problems Appl Mathematic Model 2019 71 45 59 10.1016/j.apm.2019.02.004 77. Wu L Liu Q Tian X Zhang J Xiao W A new improved fruit fly optimization algorithm IAFOA and its application to solve engineering optimization problems Knowledge Based Syst 2018 144 153 173 10.1016/j.knosys.2017.12.031 78. Long W Jiao J Liang X Tang M Inspired grey wolf optimizer for solving large-scale function optimization problems Appl Mathematic Model 2018 60 112 126 10.1016/j.apm.2018.03.005 79. Chegini SN Bagheri A Najafi F PSOSCALF: A new hybrid PSO based on sine cosine Algorithm and Levy flight for solving optimization problems Applied Soft Computing 2018 73 697 726 10.1016/j.asoc.2018.09.019 80. Gupta S Deep K A novel random walk grey wolf optimizer Swarm Evolution Comput 2019 44 101 112 10.1016/j.swevo.2018.01.001 81. Wang H Hu Z Sun Y Qinghua Su Xia X Modified backtracking search optimization algorithm inspired by simulated annealing for constrained engineering optimization problems Computation Intellig Neurosci 2018 2018 1 27 10.1155/2018/9167414 82. Zhang Q Chen H Heidari AA Zhao X Xu Y Wang P Li C Chaos-induced and mutation-driven schemes boosting salp chains-inspired optimizers IEEE Access 2019 7 31243 31261 10.1109/access.2019.2902306 83. Wang C-F Song W-X A novel firefly algorithm based on gender difference and its convergence Applied Soft Computing 2019 80 107 124 10.1016/j.asoc.2019.03.010 84. Kohli M Arora S Chaotic grey wolf optimization algorithm for constrained optimization problems J Comput Design Eng 2018 5 458 472 10.1016/j.jcde.2017.02.005 85. Li MD Zhao H WeiWeng X Han T A novel nature-inspired algorithm for optimization: virus colony search Advan Eng Software 2016 92 65 88 10.1016/j.advengsoft.2015.11.004 86. Ferreira MP Rocha ML Silva Neto AJ Sacco WF A constrained ITGO heuristic applied to engineering optimization Exp Syst Appl 2018 110 106 124 10.1016/j.eswa.2018.05.027 87. Wang L Pei J Wen Y Pi J Fei M Pardalos PM An improved adaptive human learning algorithm for engineering optimization Applied Soft Computing 2018 71 894 904 10.1016/j.asoc.2018.07.051 88. Salimi H Stochastic fractal search: a powerful metaheuristic algorithm Knowledge Based Syst 2015 75 1 18 10.1016/j.knosys.2014.07.025 89. Dhiman G SSC: A hybrid nature-inspired meta-heuristic optimization algorithm for engineering applications Knowledge Based Syst 2021 222 106926 10.1016/j.knosys.2021.106926 90. Abd Elaziz M Oliva D Xiong S An improved opposition-based sine cosine algorithm for global optimization Exp Syst Appl 2017 90 484 500 10.1016/j.eswa.2017.07.043 91. Sree Ranjith KS Murugan S Memory based Hybrid Dragonfly Algorithm for numerical optimization problems Exp Syst Appl 2017 83 63 78 10.1016/j.eswa.2017.04.033 92. Mirjalili S Lewis A S-shaped versus V-shaped transfer functions for binary particle swarm optimization Swarm Evolution Comput 2013 9 1 14 10.1016/j.swevo.2012.09.002 93. Emary E Zawbaa HM Hassanien AE Binary grey wolf optimization approaches for feature selection Neurocomputing 2016 172 371 381 10.1016/j.neucom.2015.06.083 94. Mirjalili S Mirjalili SM Yang X-S Binary bat algorithm Neural Computing and Applications 2013 25 663 681 10.1007/s00521-013-1525-5 95. Mirjalili S Gandomi AH Mirjalili SZ Saremi S Faris H Mirjalili SM Salp swarm algorithm: a bio-inspired optimizer for engineering design problems Advan Eng Soft 2017 114 163 191 10.1016/j.advengsoft.2017.07.002 96. W. Pietruszkiewicz, Dynamical systems and nonlinear Kalman filtering applied in classification, 2008 7th IEEE International Conference on Cybernetic Intelligent Systems, London, UK, 2008, 1–6. 10.1109/UKRICIS.2008.4798948 97. Zhang X Xu Y Yu C Heidari AA Li S Chen H Li C Gaussian mutational chaotic fruit fly-built optimization and feature selection Exp Syst Appl 2020 141 112976 10.1016/j.eswa.2019.112976	36466727	PMC9709762	NO-CC CODE	2022-12-01 23:23:09	no	J Bionic Eng. 2022 Nov 30;:1-22	utf-8	J Bionic Eng	2,022	10.1007/s42235-022-00298-7	oa_other
==== Front Eur J Health Econ Eur J Health Econ The European Journal of Health Economics 1618-7598 1618-7601 Springer Berlin Heidelberg Berlin/Heidelberg 36449132 1545 10.1007/s10198-022-01545-8 Original Paper Social costs of obesity in the Czech Republic http://orcid.org/0000-0001-7534-2011 Landovská Petra [email protected] 1 Karbanová Martina 2 1 grid.4491.8 0000 0004 1937 116X Faculty of Social Sciences, Charles University, Opletalova 26, 110 00 Prague, Czech Republic 2 grid.10267.32 0000 0001 2194 0956 Department of Public Health, Faculty of Medicine, Masaryk University, Kamenice 5, 625 00 Brno, Czech Republic 30 11 2022 121 23 12 2021 31 10 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. Increasing prevalence of obesity (BMI > 30) is a pressing public health issue in the Czech Republic as well as world-wide, affecting up to 2.1 billion people. Increasing trend in the prevalence of obesity in adults and children generates large social costs. The main aim of this study is to estimate both direct and indirect costs of obesity in the Czech Republic. Social costs of obesity are estimated using the cost-of-illness approach. Direct costs (healthcare utilization costs and costs of pharmacotherapy of 20 comorbidities) are estimated using the top-down approach, while indirect costs (absenteeism, presenteeism and premature mortality) are estimated using the human capital approach. In aggregate, the annual costs attributable to obesity in the Czech Republic in 2018 were 40.8 bn CZK (1.6 bn EUR, 0.8% GDP). Direct costs were 14.5 bn CZK (0.6 bn EUR) and accounted for 3.4% of Czech healthcare expenditures. The highest healthcare utilization costs were attributable to type II diabetes (20.6%), ischemic heart disease (18.8%) and osteoarthritis (16.7%). The largest indirect costs were attributable to premature mortality (10 bn CZK/0.39 bn EUR), absenteeism (9.2 bn CZK/0.36 bn EUR) and presenteeism (7.1 bn CZK/0.27 bn EUR). This article demonstrates that obesity is a serious problem with considerable costs. Several preventive interventions should be applied in order to decrease the prevalence of obesity and achieve cost savings. Keywords Obesity Social costs Cost-of-illness study Czech Republic JEL Classification I12 I18 ==== Body pmcIntroduction Obesity is a serious global health problem which brings substantial economic burden to society. Worldwide rates of obesity have nearly tripled since 1975 and more than 2.1 billion people (30% of global population) suffer from overweight or obesity today. Importantly, obesity is a risk factor for cardiovascular diseases, type II diabetes and some types of cancers, which brings substantial healthcare costs, but also large indirect costs through lost productivity. The global economic impact of obesity is estimated to be $2 trillion (2.8% of global GDP), which is comparable to the impact of armed conflict, smoking or terrorism [1]. In OECD countries, 8.4% of healthcare budget is dedicated to treatment of overweight-related diseases [2]. Obesity has become a pressing concern also with regard to the COVID-19 pandemic. Numerous studies have shown that overweight and obese individuals are more at risk for being COVID-19 positive and have more severe symptoms, leading to significant increase in morbidity and mortality [3]. Moreover, due to many restrictions implemented in order to prevent the spread of COVID-19 (e.g. movement restrictions, social distancing), people lack physical activity which may aggravate current trends in the prevalence of obesity and put even larger strain on the healthcare system [3]. In the Czech Republic, the rates of obesity have been increasing since the 90’s both in adults [4] and children [5]. The goal of this study is to estimate the social costs of obesity, defined as BMI > 30 kg/m2, in the Czech Republic. The cost-of-illness (COI) approach, which views the burden of specific illness as the sum of direct (medical) and indirect costs, is implemented. The resulting social costs of obesity show us how much could be saved if the disease did not exist at all [6]. This study is being novel in estimating the social costs of obesity in the Czech Republic using the COI method, demonstrating what an extreme burden this disease brings to the society. Literature review Direct costs Numerous studies have found that obesity is associated with increased risk of cardiovascular diseases, type II diabetes and cancers [7, 8]. This leads to increased medical costs due to higher use of prescription drugs and outpatient care, or longer hospital stays as a result of post-treatment complications [9–12]. Current literature indicates that the direct costs of obesity are substantial. For instance, research from the USA indicates that obesity is associated with 36% increase in inpatient and outpatient spending and 77% increase in medication costs [13]. Several studies also show that medical costs increase proportionally with BMI [14–16]. Table 4 summarizes literature focusing on direct costs of obesity. The studies vary in the amount of comorbidities included, which ranges from 4 to 18. The share of direct costs of obesity on total healthcare costs ranges from 2.3% in Sweden [16] to 6.7% in Canada [17]. In the Czech Republic, the direct costs of obesity were estimated as 7.6 billion CZK (excluding the costs of pharmacotherapy) in 2013, which accounted for 3.45% of total healthcare costs [18]. Earlier estimate from 2007 was 9.5 billion CZK (5.2% of healthcare costs), out of which 2.6 billion CZK were costs of pharmacotherapy [19]. In general, the share of costs of overweight and obesity on total healthcare costs ranges between 2 and 8% [2]. Indirect costs Absenteeism Absenteeism refers to absence from work due to illness. The rate of absenteeism due to illness varies across countries, but Czech Republic has one of the highest rates in Europe, reaching on average 16.3 days missed in 2018 [20]. Table 5 summarizes literature which includes the costs of absenteeism associated with obesity. Almost all studies find that the costs of absenteeism are significantly larger for obese workers compared with normal-weight workers. However, the magnitude of the difference varies across studies, which can be caused by country, data or methodology differences. Usually, absenteeism is compared between obese and normal weight individuals (e.g. [9, 21, 22]), but some studies compare BMI > 25 (thus including overweight) with normal weight (e.g. [11, 23–25]), which makes some results uncomparable. Studies which examine absenteeism across three obesity classes show that the rates of absenteeism increase with BMI [21, 22, 26]. Presenteeism Presenteeism refers to reduced productivity at work due to presence of mental or physical health complications [27]. There is growing evidence that the costs of presenteeism associated with chronic conditions largely exceed the costs of absenteeism [28–30]. To the best of our knowledge, presenteeism related to obesity has not been measured in the Czech Republic thus far, therefore, our estimates are based on current literature, which is summarized in Table 6. All the studies find that obesity is positively associated with presenteeism, but the extent differs across studies. Compared with normal weight individuals, the estimates range from 1.1 to 3.8 more days lost [31–34] and reach 22.7–33 days lost for obesity class III [28, 35]. Premature mortality Excess weight is associated with substantial increases in early mortality [36]. Obesity and its related diseases are estimated to reduce life expectancy by 0.9–4.2 years [2], which leads to large productivity losses. Current literature on the costs of premature mortality related to obesity is summarized in Table 7. The estimates of number of years lost due to obesity-related premature mortality vary across studies and substantially increase with BMI [37]. In the Czech Republic, overweight-attributable reduction in life expectancy is estimated to be 3.5 years [38] and in 2013, the obesity-related costs due to premature mortality were evaluated as 1.2 billion CZK [18]. In general, the studies show that the costs are larger for men than women, typically due to higher wages, higher retirement age or higher prevalence of obesity. Estimates from the Czech Republic Only a few studies estimated the social costs of obesity in the Czech Republic. In 2007, the direct costs of obesity were estimated as 9.5 billion CZK [19]. Another estimate which is largely based on other (foreign) studies estimated direct and indirect costs of obesity to be 20.3–42.5 billion CZK and 17.2–37.8 billion CZK, respectively [39]. In 2013, direct and indirect costs of obesity were quantified as 12.1 billion CZK and corresponded to 0.3% of GDP in 2013 [18]. Data and methodology Data Table 1 summarizes the data used in the baseline model. The healthcare utilization costs and costs of pharmacotherapy in 2018 are obtained from the General Health Insurance Fund (GHIF) jointly for men and women [40], which are extrapolated to the whole population.1 The data for computation of costs of absenteeism and premature mortality in 2018 are obtained from the Institute for Health Information and Statistics (IHIS), specifically from the Information System Incapacity for work [41] and the Information System Deaths [42]. For each comorbidity, they are available for 5-year age groups for men and women separately. The estimation of costs of presenteeism is based on literature review. Prevalence of obesity in the Czech Republic is taken from the NCD Risk Factor Collaboration (NCD-RisC) study with the most recent data being from 2016. This study provides the prevalence of obesity in 200 countries based on data measured by physicians. The age-standardised prevalence of obesity in population 20 years and older is 27.3% for men and 26.5% for women [43]. Relative risks are derived from Guh et al. [7] and Dobbins et al. [8]. These studies provide a review of existing studies that identify statistically significant comorbidities of obesity, i.e. the diseases that are more likely to occur in obese population vs. normal weight population. Guh et al. (2009) conduct a review of studies coming predominantly from the USA (55%) and Europe (40%), and Dobbins et al. (2013) conduct a review of studies conducted mainly in the USA, Norway, Sweden or Japan. Paid work is valued by the average gross salary for each gender and age group in 2018 reported by the Czech Statistical Office (CSO) [44]. The average daily amount of hours spent doing housework was estimated to be 3 h for women and 2 h for men [45]. The value of unpaid work is approximated by the average wage of cleaning services workers in 2018 (the average hourly wage is 87 CZK/h [46]). Life expectancy in 2018 is derived from the CSO [47] and the number of employed people aged 25-64 years in 2018 is obtained from Eurostat [48].Table 1 Data sources Measure Data source Year Direct costs Healthcare utilization General Health Insurance Fund 2018 Pharmaceuticals General Health Insurance Fund 2018 Indirect costs Absenteeism Information System Incapacity for Work 2018 Presenteeism Literature review NA Premature mortality Information System Deaths 2018 Other sources Prevalence of obesity NCD-RisC Database 2016 Relative risks Guh et al. 1994–2006 Dobbins et al. 1985–2011 Average salary Czech Statistical Office 2018 Unpaid work Czech Academy of Sciences 2015 Value of unpaid work Average Earnings Information System 2018 Life expectancy Czech Statistical Office 2018 Employed population Eurostat 2018 Methodology Social costs of obesity are estimated using the cost-of-illness (COI) approach, which views the economic burden of disease as the sum of several categories of direct and indirect costs [49]. There are two types of approaches within the COI methodology: prevalence and incidence approach. The prevalence approach is used in our analysis as it assesses the current economic burden of illness [50]. For more details on methodology, please refer to Appendix B. Direct costs Direct costs refer to medical and non-medical expenditures related to obesity-related diseases and are estimated using the top-down approach. This approach uses population attributable fraction (PAF) which attributes part of healthcare costs to obesity. Due to data constraints specific to the Czech Republic, the computation of direct costs is divided into healthcare utilization costs and costs of pharmaceuticals. Based on two studies [7, 8], 20 comorbidities of obesity and their relative risks RR (how much more likely will these diseases occur in obese individuals compared to normal weight) are identified. Using RR and prevalence (p) of obesity in the Czech Republic from the NCD RisC study, PAFc (i.e. what portion of comorbidity’s costs are due to obesity) is computed:PAFc=p·(RRc-1)p·(RRc-1)+1 Healthcare costs attributable to obesity (HC) are computed as:HC=∑cPAFc·Cc, where PAFc is population attributable fraction for comorbidity c and Cc are the healthcare utilization costs associated with comorbidity c. Similar approach is taken in estimating the costs of pharmacotherapy attributable to obesity. Based on two studies [19, 23], five groups of pharmaceuticals are identified (e.g. pharmaceuticals used in the cure of diabetes mellitus, cardiovascular diseases etc.) along with the specific Anatomical Therapeutic Chemical (ATC) classification codes (see Appendix B for more details). Pharmaceutical costs attributable to obesity (PC) are computed as:PC=∑cPAFc·PCc, where PCc are the pharmaceutical costs of ATC group related to a comorbidity c. Indirect costs Indirect costs refer to the value of lost production due to morbidity and mortality, which we estimate using the Human capital approach (HCA). We include the costs of absenteeism, presenteeism and premature mortality. The value of both paid work (PWag) and unpaid work (UWg) is included in the costs, monetized by average gross salary (GSag) and average wage of houseworker (WHW), respectively:Pag=PWag·GSag+UWg·WHW, where Pag refers to age- (a) and gender- (g) specific evaluation of productivity lost. PAFacg specific for age (a), comorbidity (c) and gender (g) is used to determine the obesity-attributable productivity lost, because the data are stratified by 5-year age groups and genders. Absenteeism refers to missed days at work due to illness. The number of days absent (DAacg) due to obesity related comorbidities (c) for age (a) and gender (g) is multiplied by PAFacg to find the number of days absent from work due to obesity (DAacgobesity):DAacgobesity=DAacg·PAFacg, The indirect costs due to obesity-related absenteeism (ICabs) are computed as:ICabs=∑a∑c∑gDAacgobesity·Pag. Presenteeism describes lower productivity while present at work. Due to unavailability of data on obesity-related rates of presenteeism in the Czech Republic, we estimate the costs based on literature review, assuming that on average, obese individuals miss 2 days of work due to presenteeism. In case of presenteeism, we only distinguish the costs for each gender g, disregarding their age as we do not have data for it. The indirect costs due to obesity-related presenteeism (ICpres) are computed as:ICpres=∑gpg·Eg·PL·Pg, where Eg is the number of employed people in working-age population and pg·Eg is the number of obese people in labour force (pg is gender-specific prevalence of obesity in working-age population, i.e. 25–64 years old), PL stands for productive days lost due to presenteeism and Pg is the gender-specific valuation of paid and unpaid work. To estimate the value lost due to premature mortality, we use the data on number of deaths due to each comorbidity of obesity stratified by gender and age. The present value of future lost earnings (NPV) is computed using a discount rate (i) which is 3% in the baseline scenario:NPV=∑t=0nFV(1+i)t, where FV stands for future value and t is the amount of years lost. The indirect costs due to obesity-related premature mortality (ICmort) are computed as:ICmort=∑a∑c∑gPAFacg·Macg·0.5·Pag+∑t=1ret-1Pag(1+i)t+∑t=retexpUWg·WHW(1+i)t, where Macg stands for age-, comorbidity- and gender-specific number of deaths. Only half of the productivity is accounted for in the first year (t=0) to correct for different occurrences of death during the year. The productive years (i.e. before retirement age ret) are monetized by the value of paid and unpaid work, while the years after retirement until life expectancy age exp are monetized by the value of unpaid work only. Only part of these costs is attributable to obesity, which is computed using the PAFacg. Sensitivity analysis In sensitivity analysis, we test the robustness of results by varying several parameters of the model (more details are available in Sect. B.3):PAF are recomputed using the 95% confidence interval of relative risks. PAF are recomputed using the relative risks from the Dynamo project. Prevalence data from the EHES/EHIS study from 2014 [51, 52] are used. Discount rate of 1% and 5% is used for computing the costs of premature mortality. Unpaid work is completely excluded from total costs. Presenteeism is computed for missing 1, 3 and 4 days of work (baseline value is 2 days). Results Direct costs Healthcare utilization costs Table 9 lists the relevant comorbidities of obesity along with the ICD-10 codes and PAF computed based on the prevalence of obesity in the Czech Republic2 and relative risks [7, 8]. Total costs of healthcare utilization due to obesity are reported in Table 10 and amount to 11.4 billion CZK (see Fig. 1). The largest portion of these costs is due to type II diabetes mellitus (2.3 billion CZK), ischemic heart disease (2.1 billion CZK) and osteoarthritis (1.9 billion CZK).Fig. 1 Healthcare utilization costs Source: author’s computations Costs of pharmacotherapy Figure 2 summarizes the costs of pharmacotherapy. Table 11 shows the ATC groups included in the study and the costs attributable to obesity [19, 23]. Drugs used in diabetes make up the largest part of pharmacotherapy costs (820 million CZK), followed by antithrombotic agents (595 million CZK) used for the cure of cardiovascular diseases and agents acting on the renin-angiotensin system (573 million CZK) used for the cure of cancer. Total pharmacotherapy costs attributable to obesity are 3.1 billion CZK.Fig. 2 Costs of pharmacotherapy Source: author’s computations Indirect costs The indirect costs are visually summarized in Fig. 3. Absenteeism 2.8 million days were lost due to obesity in men and 2.6 million days in women in 2018. Total costs of absenteeism are 9.2 billion CZK (4.1 billion CZK for women and 5.1 billion CZK for men) and 8.1 billion CZK (3.5 billion CZK for women and 4.6 billion CZK for men) after excluding the value of unpaid work. The results are summarized in Table 12. Presenteeism The baseline value for days lost in our model is 2 days of work lost, which is associated with costs of 7.1 billion and 6.2 billion after excluding the value of unpaid work. The costs of presenteeism for 1, 3 and 4 days lost amount to 3.5, 10.6 and 14.1 billion CZK respectively (3.1, 9.3 and 12.3 billion CZK, respectively, after excluding the value of unpaid work). The costs of presenteeism are summarized in Table 13. Premature mortality In 2018 women lost 72, 670 years due to obesity, from which 2947 years were productive years. Men lost in total 89, 850 years due to obesity, from which 8737 years were productive years. The reason why the productive years make such a small part of total years lost due to obesity is that most people die due to obesity-related diseases after retirement. Using the discount rate of 3%, the costs of premature mortality due to obesity are 10 billion CZK, including unpaid work. The costs are higher for women even though the amount of productive years lost is lower compared to men because women lose more unproductive years than men. After excluding the unpaid work, the costs are 3.7 billion CZK. Table 14 shows the results for different discount rates.Fig. 3 Summary of indirect costs Source: author’s computations; UW unpaid work Summary of results Total costs of obesity in the Czech Republic for the year 2018 are summarized in Table 2. In total, they amount to 40.8 billion CZK, which corresponds to 0.8% of GDP in 2018 [53]. The indirect costs account for majority of the costs: 26.3 billion CZK (65%), whereas the direct costs are 14.5 billion CZK (35%), which accounts for 3.4% of total healthcare costs in 2018.3Table 2 Summary of results CZK % of total costs Direct costs 14,456.2 35.5 Healthcare utilization 11,373.0 27.9 Pharmacotherapy 3083.2 7.6 Indirect costs 26,295.5 64.5 Absenteeism 9237.0 22.7 Presenteeism 7050.1 17.3 Premature mortality 10,008.4 24.6 Total 40,751.7 100.0 Values are in millions CZK Sensitivity analysis Table 3 shows the change in costs attributable to obesity as the key parameters are varied. Total costs range between 32.3 billion CZK (− 20.8% from baseline values) and 51.1 billion CZK (+25.5% from baseline values). The largest changes result from using the low and high relative risks values (95% CI). The overall costs decrease by 8.4% when the 2014 data on prevalence of obesity are used.Table 3 Sensitivity analysis Direct costs % change Indirect costs % change Total costs % change Baseline 14,456.2 – 26,295.5 – 40,751.7 – Relative risks low values 10,650.0 − 26.3% 21,610.8 − 17.8% 32,260.8 − 20.8% Relative risks high values 18,902.1 30.8% 32,231.2 22.6% 51,133.2 25.5% Relative risks Dynamo project 12,975.7 − 0.2% 24,103.2 − 8.3% 37,078.9 − 9.0% EHIS/EHES prevalence (2014) 13,088.1 − 9.5% 24,260.7 − 7.7% 37,348.8 − 8.4% Discount rate 1% 14,456.2 0% 27,526.2 4.7% 41,982.4 3.0% Discount rate 5% 14,456.2 0% 25,327.9 − 3.7% 39,784.1 − 2.4% Excluding unpaid work 14,456.2 0% 17,907.6 − 31.9% 32,363.8 − 20.6% Presenteeism 1 day 14,456.2 0% 22,770.5 − 13.4% 37,226.7 − 8.7% Presenteeism 3 days 14,456.2 0% 29,820.6 13.4% 44,276.8 8.7% Presenteeism 4 days 14,456.2 0% 33,345.6 26.8% 47,801.8 17.3% Values are in millions CZK Discussion The goal of this study was to estimate the social costs of obesity in the Czech Republic in 2018. The resulting costs are equal to 40.8 billion CZK, which corresponds to 0.8% of GDP. This result should be taken as a lower-bound estimate of the costs of obesity as the prevalence data come from 2016, we use a very conservative estimate for the costs of presenteeism, we exclude intangible costs and use the top-down approach. The comparison of results across studies is complicated due to differences in methodological approach. A study from Germany, which is socio-economically similar to the Czech Republic, estimated the costs of overweight (BMI > 25) in 2008 using a similar approach as 0.5% of GDP [24]. This estimate is lower than ours mainly because it uses older data: both the prevalence of obesity and healthcare costs have increased largely since 2008. A new OECD study estimates the burden of overweight and obesity (BMI > 25) in 52 different countries to be 1.6–5.3% of GDP [2]. The specific estimate for the Czech Republic is 4% of GDP, which is much higher than our estimate. This may have several reasons. Our study focuses purely on obesity (BMI > 30), whereas the OECD study also includes overweight (i.e. BMI > 25). The prevalence of overweight is much higher than of obesity in the Czech Republic: 70% for men and 55% for women. Furthermore, the OECD study uses different methodological approach (a microsimulation model vs. a country-level COI study) and data sources (often derived from other countries or studies), so the results are not directly comparable (see Appendix C for more details). The direct costs of obesity are 14.5 billion CZK, corresponding to 3.4% of healthcare expenditures. International studies estimate the impact of overweight and obesity on health expenditures in the range of 2–7.9% [2]. The estimate from Germany from 2008, which has the same healthcare financing scheme, is 3.27% of healthcare expenditures [24]. The indirect costs are 26.3 billion CZK, which exceeds previous estimates from the Czech Republic due to inclusion of presenteeism, unpaid work, use of gross salaries and rising prevalence of obesity. Cost-of-illness methodology is the most common measurement approach to estimate the burden of disease, but it has certain drawbacks. A variety of approaches within the COI methodology can be taken, which limits the comparability of results across studies. Additionally, it measures the value of individual’s life only in terms of the production evaluated by average wage, ignoring other dimensions of illness and death, such as pain and lower quality of life [50]. However, when performed with a clear explanation, COI studies represent an important analytic tool in public health policy [55]. In this study, HCA is used to estimate the indirect costs of obesity. This method has been mainly criticised for assuming full employment in the economy, which relates mainly to the costs of absenteeism where every day the worker misses is regarded as lost production. However, the approach disregards the fact that the work can be made up by the worker after his/her return, or it can be done by his/her colleagues [50]. The friction cost approach (FCA) solves this drawback and counts the productivity losses only for the time it takes to replace the absent worker. The HCA is further criticised for evaluating the costs based on age- and gender-specific wages, implying that people earning lower wages are less valuable for the society. Willingness-to-pay approach mitigates this problem, however, it is not often employed as it requires extensive surveys of preferences and the results highly depend on the individuals’ subjective responses to hypothetical questions [55]. There are several limitations in our study, mainly related to availability of relevant data. Firstly, we use the data on prevalence of obesity from 2016, even though we estimate the costs of obesity in 2018 as no more recent data stratified by gender and age groups are available. The results of the EHES 20194 survey suggest increasing trends of obesity, which would imply even larger social costs [56]. Secondly, the relative risks used in computations of population attributable fractions (PAF) and the rate of presenteeism are based on foreign literature. This is the reason why we also perform a thorough sensitivity analysis and vary some of the key parameters of the model. It is evident that foreign data have limited relevance in the Czech Republic. For further improvement of the analysis, it will be necessary to conduct a survey in the Czech Republic. Our study demonstrates that the costs of obesity are considerable in the Czech Republic and comparable to the costs of smoking and alcohol consumption, which are estimated as 14.5 billion CZK (0.8% of GDP) in 1999 [57] and 59.5 billion CZK (1.2% of GDP) in 2016 [58], respectively. However, smoking and alcohol consumption have received more consistent attention in clinical practice and public health policy [13]. Similarly as alcohol consumption and smoking, early onset of obesity or overweight significantly increases the probability of being obese in adulthood [59]. This implies that obesity is a serious disease which should no longer be regarded as a lifestyle issue but needs to be recognised as a serious medical condition [60]. Conclusion The rising prevalence of obesity has been putting an increasing pressure on the health care system and society, which will be further aggravated due to the COVID-19 pandemic. The goal of this study was to quantify the extent of this burden in the Czech Republic using data from 2018. The social costs of obesity are estimated using the cost-of-illness approach. Total costs of obesity are estimated to be 40.8 billion CZK, which corresponds to 0.8% of GDP in 2018. Out of this, 14.5 billion CZK (35%) are attributable to direct costs and 26.3 billion (65%) are attributable to indirect costs. The direct costs account for 3.4% of total healthcare costs in 2018. Within indirect costs, the largest part is attributable to premature mortality (10 billion CZK), absenteeism (9.2 billion CZK) and presenteeism (7.1 billion CZK). This is a unique country-level COI study which focuses on the costs of obesity in the Czech Republic and accounts for several groups of direct and indirect costs. These costs are substantial which is supported by the fact that they are comparable to the costs of smoking or alcohol consumption in the Czech Republic. Moreover, with rising prevalence of overweight and obesity in children and adults, these costs are likely to increase. A comprehensive, systemic program of multiple interventions should be implemented to increase awareness, reverse the trend of growing rates of obesity and save money in the long-term horizon. Appendix A: Literature review See Tables 4, 5, 6 and 7.Table 4 Literature review—direct costs n Country Method Number of comorbidities Population Normal weight Overweight Obesity % of healthcare costs Arterburn et al. [12] 16,262 USA Econometric approach N/A 18+ $2424 $2664 $2984–4399a N/A Finkelstein et al. [22] 20,329 USA Econometric approach N/A 18–64 women 0 (base) $495 $1071–1549a N/A 18–64 men $169 $392–1591a Cawley & Meyerhoefer [10] 23,689 USA Econometric approach NA 11–64 $1763 $4458 20.6% Kleinman et al. [61]b 72,778 USA Econometric approach N/A 18+ women $4 142 $4583 $6328 N/A 18+ men $2861 $3378 $4309 An [62] 125,955 USA Econometric approach N/A 18+ women 0 (base) N/A $1525 N/A 18+ men $1160 Borg et al. [16] 23,365 Sweden Econometric approach N/A 30–60 women 0 (base) $101 millionc 2.3% (total costs) 30–60 men $169 million Sander & Bergemann [63] N/A Germany Top-down approach 4 25+ N/A N/A €2 billion N/A Konnopka et al. [11] N/A Germany Top-down approach 16 (W), 17 (M) 18+ N/A €4 854 million 2.1% Lehnert et al. [24] N/A Germany Top-down approach 16 (W), 17 (M) 18+ N/A €8 647 million 3.27% Schmid et al. [60] N/A Switzerland Top-down approach 18 15+ N/A CHF 1 077–1 615 million 2.3–3.5% Kang et al. [25]d 1.9 mil Korea Top-down approach 7 20+ N/A $270.5 million $810.5 million 3.7% Dee et al. [23] N/A Northern Ireland Top-down approach 16 N/A N/A €127 million 2.8% N/A Republic of Ireland €437 million 2.7% Krueger et al. [17] N/A Canada Top-down aproach 16 12+ women N/A $4.3 billion $7.6 billion 6.7% 13 12+ men $4.8 billion $6.6 billion Hodycová [19] N/A Czech Republic Top-down approach 7e 18+ N/A N/A 9.5 billion CZKf 5.2% Tuzarová [18] N/A Czech Republic Top-down approach 18 (W), 16 (M) 18+ N/A N/A 7.6 billion CZK 3.45% Effertz et al. [9] 146,000 Germany Bottom-up approach NA 15+ N/A N/A €29.39 billion 7.9%g aRange for obesity class I–III. bNormal weight is classified as BMI < 27, overweight: 27 ≤ BMI < 30, obesity: BMI ≥ 30. cValues are converted from SEK by a rate US$1 = SEK8. dIn Korea, classification of obesity according to BMI is different than in Europe (Overweight: 23–24.9 kg/m2; Obesity I: 25–29.9 kg/m2; Obesity II: ≥ 30 kg/m2) . eCardiovascular diseases are taken as one comorbidity. f6.7 billion healthcare utilization, 2.6 billion pharmacotherapy. gPercentage of healthcare costs stated in [2] Table 5 Literature review—absenteeism n Country Measurement unit Population Normal weight Overweight Obesity Konnopka et al. [11] N/A Germany Number of days/year 15+ 0 (base) 5,875,022 days Total yearly costs €646 milliona Lehnert et al. [64] 7990 Germany Yearly costs per person (days absent/year) 18–65 women 0 (base) €284 (3.64 days) €405 (5.19 days) Yearly costs per person (days absent/year) 18–65 men N/A (N/A) €367 (3.48 days) Total yearly costs 18–65 €2.18 billionb Lehnert et al. [24] N/A Germany Number of days/year 15+ 0 (base) 11,478,208 days Total yearly costs €1.28 billionc Effertz et al. [9] 146 000 Germany Total yearly costs 15+ 0 (base) N/A €3.87 billion Finkelstein et al. [22] 25,427 USA yearly costs per person (days absent/year) 18–64 women $0 (base)/(3.4 days) $93 (3.9 days) $302–805 (5.2–8.2 days)d yearly costs per person (days absent/year) 18–64 men $0 (base)/(3 days) $6 (3 days) $70–436 (3.5–5 days) Finkelstein et al. [28] 24,140 USA Yearly costs per person (days absent/year) 18+ women 0 (base) $147 (1.1 days) $407–1262 (3.1–9.4 days)d Yearly costs per person (days absent/year) 18+ men $85 (0.5 days) $277–1026 (1.6–5.9 days) Dall et al. [21] 225 mil USA Yearly costs per person 18+ 0 (base) $47 $104–264d Total yearly costs $3.5 billion $3.9–6.8 billion Andreyeva et al. [26] 14 975 USA yearly cost per person (days absent/year) 18+ 0 (base)/(4.25 days) N/A (4.48 days) $216–348 (5.42–6.13 days)d Total yearly costs 0 (base) N/A $8.65 billion Kleinman et al. [61]5 72,778 USA Yearly costs per person (days absent/year) 18+ women $890 (4.09 days) $1046 (5.02 days) $1175 (5.81 days) Yearly costs per person (days absent/year) 18+ men $615 (2.66 days) $640 (2.81 days) $792 (3.7 days) Kang et al. [25] 1.9 mil Korea Total yearly costs 20+ women 0 (base) $29.5 million 20+ men $44.4 million Dee et al. [23] N/A Northern Ireland Total yearly costs N/A 0 (base) €215 million Republic of Ireland Total yearly costs €136 million Neovius et al. [65] 45,920 Sweden Lifetime productivity losses 19–65 men €12,500 €15,000 €16,100 Tuzarová [18] N/A Czech Republic Number of days/year 24–60 women & 0 (base) N/A 3.7 million days Total yearly costs 25–64 men 3.2 billion CZK a€481 million without unpaid work. b€1.37 billion women and €0.81 billion men. c€858 million without unpaid work. dRange for obesity class I–III . eNormal weight: BMI <27; overweight: BMI >27 and <30 Table 6 Literature review—presenteeism n Country Measurement unit Gender Normal weight Overweight Obesity Class I Class II Class III Boles et al. [32]a 2264 USA Productivity loss (%) Both 5.6 7.1 Productivity loss (days) 14.0 17.6 Pelletier et al. [66]a 500 USA Productivity loss (%) Both 4.7 7.9 Productivity loss (days) 11.8 19.8 Burton et al. [33] 28,375 USA Productivity loss (%) Both 0 (base) N/A 1.5 Productivity loss (days) N/A 3.8 Ricci and Chee [34] 7000 USA Weekly hours lost both 4.2 4.2 4.8 Productivity loss (days) 26.3 26.3 30 Gates et al. [67]b 341 USA (KY) Productivity loss (%) Both 3.3 3.1 2.5 4.2 Productivity loss (days) 8.1 7.8 6.1 10.4 Finkelstein et al. [28] 10,262 USA Productivity loss (days) Women 0 (base) 0.9 6.3 11.0 22.7 13 878 Men − 3.3 2.3 5.8 21.9 Goetzel et al. [31]c 10,026 USA Productivity loss ($) Both 1200 1402 1416 Productivity loss (days) 5.8 6.8 6.9 Kirkham et al. [68] 17,089 USA Productivity loss (days) Both 4.2 N/A N/A 4.7 Gupta et al. [35] 31,653 FRA, DE, IT Productivity loss (%) Both 16 15.6 17.6 20.4 29.2 ESP, UK Productivity loss (days) 40 39.1 44.0 50.9 73.0 Productivity loss in days is annual; conversion from productivity loss in percent to productivity loss in days is done assuming 250 working days per year. aCompares normal weight against BMI<18.5 or >24.9 bCompares BMI categories with BMI<24.9 (normal weight + underweight) cThe study uses average wage rate $25.67/h Table 7 Literature review—premature mortality n Country Measurement unit Discount rate Population Normal weight Overweight Obesity Konnopka et al. [11] N/A Germany Annual number of deaths 5% 15+ 0 (base) 36,653 Total productivity lost €3381 million Lehnert et al. [64] N/A Germany Annual number of deaths 5% 15+ 0 (base) 47,964 Total productivity lost €5669 million Effertz et al. [9] 146,000 Germany Annual number of deaths 2% 15+ 0 (base) N/A 101,886 Amount of lost years 2072 million Total productivity lost €23.12 billion Fontaine et al. [37] 23 659 USA Years of life lost per person N/A 15–75 women N/A <1 year 3–8 yearsa 15–75 men <1 year 3–13 yearsa Dall et al. [21] 225 million USA Total productivity lost 3% 18+ 0 (base) $1.9 billion $6.9–25.9 billionb Per capita productivity lost $25 $182–1006b Borg et al. [16] 23,365 Sweden Total productivity lost 3% 30–60 women 0 (base) $1.15 million $64.0 million 30–60 men 3.3 million 298.5 million Neovius et al. [65] 45,920 Sweden Per capita productivity losses 3% 18+ men €25,100 €31,800 €52,100 Dee et al. [23] N/A Northern Ireland Total productivity lost 4% 18–75 0 (base) €147 million Republic of Ireland 18–75 €593 million Kang et al. [25] 1.9 million Korea Total productivity lost 6% 20+ women 0 (base) $70 million 20+ men $374 million Tuzarová [18] N/A Czech Republic Amount of lost years 1.5% 25–60 women 0 (base) N/A 5440 years Total productivity lost 395 million CZK Amount of lost years 25–64 men 2290 years Total productivity lost 800 million CZK aRange for different BMI groups, BMI = 24 is used as reference category bRange for obesity class I–III Appendix B: Methods appendix This study uses the cost-of-illness (COI) approach to estimate the social costs of obesity. Two approaches exist within the COI approach: (i) the prevalence approach, which estimates the costs of all new and pre-existing cases in one year, including years lost due to premature death discounted to present value, and (ii) the incidence approach, which estimates lifetime costs of all new cases/deaths in given year. Prevalence approach is used in the analysis as the aim is to assess the current economic burden of obesity. B.1 Direct costs Direct costs of illness include all resources related to its prevention, treatment and rehabilitation [50]. The top-down approach, which measures the proportion of a disease that is due to exposure to risk factor, is used. This approach uses aggregated data, along with PAF (population attributable fraction), which is used to determine the attributable costs. For example, it attributes part of costs of diabetes to obesity [55]. The computation of direct costs of obesity is divided into two parts due to data constraints specific to the Czech Republic. The reason is that the healthcare utilization costs are documented with International Classification of Diseases (ICD) codes, whereas pharmaceutical costs are documented with ATC (Anatomical Therapeutic Chemical) classification codes. B.1.1 Healthcare utilization costs Healthcare utilization costs are computed in the following way: Identify the comorbidities of obesity, i.e. the diseases that are more likely to occur if a person suffers from obesity. Find the relative risk (RRc) for each comorbidity c. That is, how much more likely is a disease to occur in population with obesity as opposed to population with normal weight. RRc=r1r2, where r1 is is the probability of disease at obese population and r2 is the probability of disease at normal weight population. Find prevalence (p) of obesity in the Czech Republic, i.e. the proportion of population whose BMI exceeds 30 kg/m2. Compute PAFc (population attributable fraction for comorbidity c), which tells us what fraction of disease’s costs is attributable to obesity: PAFc=p·(RRc-1)p·(RRc-1)+1 The PAFc does not take into account ages nor genders as the data on healthcare utilization costs are provided for each diagnosis without the distinction of age groups or gender. For instance, PAFc for stroke is equal to 11%, meaning that 11% of healthcare costs of stroke are caused by obesity. Compute the healthcare utilization costs attributable to obesity (HC): HC=∑cPAFc·Cc, where Cc are the healthcare utilization costs associated with comorbidity c. The healthcare utilization costs are obtained from the General Health Insurance Fund and contain all the costs of patients that had a comorbidity of obesity (e.g. Diabetes II mellitus with ICD codes E11, E13 or E14) as the main diagnosis, i.e. it was the main cause of receiving a treatment. B.1.2 Costs of pharmacotherapy Since healthcare utilization costs do not contain the costs of pharmacotherapy,5 we include these separately by identifying the ATC groups which are related to comorbidities of obesity. We include five groups of pharmaceuticals based on Hodycova (2009) and Dee (2015) [19, 23]:for the cure of obesity for the cure of diabetes mellitus for the cure of cardiovascular diseases for the cure of cancer for the cure of arthrosis Specific ATC groups are available in Table 11. The costs of pharmaceuticals provided by the General Health Insurance Fund contain only the costs used in the treatment of the comorbidity (e.g. if we take the example of diabetes mellitus, these costs are not the costs of pharmaceuticals incurred by all diabetic patients, but they are the costs of pharmaceuticals used in the treatment of diabetes). Pharmaceutical costs PC are computed as:PC=∑cPAFc·PCc, where PCc are the pharmaceutical costs of ATC group which is related to a comorbidity of obesity c. PAF are used to compute the part of pharmaceutical costs that are directly attributable to obesity. B.2 Indirect costs Indirect costs of illness are the value of lost production due to morbidity or mortality. This lost time is multiplied by age- and gender-specific average gross wage rates to calculate the indirect costs [50]. The Human capital approach (HCA) is used to estimate the indirect costs. It takes the patient’s perspective and counts any hour not worked as an hour lost. The value of housework is also incorporated into the lost production and it is valued as the opportunity cost of hiring a replacement from the labor market [9]. We include three common components of indirect costs: absenteeism, presenteeism and premature mortality. For premature mortality, the present value of future lost earnings is computed using a discount rate. The total lost productivity is the sum of paid work (measured in days per month) and unpaid work (measured in hours per month). Paid work is valued by gender- and age-specific monthly gross salary, whereas unpaid work is valued by the average hourly salary of a household worker:6Pag=PWag·GSag+UWg·WHW, where Pag refers to age- (a) and gender- (g) specific evaluation of productivity lost, PWag stands for the age- and gender-specific amount of paid work, GSag refers to age and gender-specific gross salary, UWg stands for the gender-specific amount of unpaid work and WHW stands for the wage of household worker.7 The productivity lost from paid work is considered until the average retirement age (63.2 years for men and 62.7 years for women [69]), whereas the productivity lost from unpaid work is considered until the average life expectancy age (76 for men and 82 for women [47]). The data on indirect costs (specifically absenteeism and premature mortality) allow for more detailistic computation of costs attributable to obesity, so PAFacg which is specific for age (a), comorbidity (c) and gender (g) is used:PAFacg=pag·(RRc-1)pag·(RRc-1)+1 B.2.1 Absenteeism Absenteeism refers to absence from work due to illness. To calculate the number of days absent from work attributable to obesity, we use the number of terminated cases of incapacity for work for obesity-related comorbidities in days from 2018 provided by the IHIS [41]. The number of days spent absent from work due to obesity for age (a), comorbidity (c) and gender (g) are computed as:DAacgobesity=DAacg·PAFacg, where DAacg stands for age- (a), comorbidity- (c) and gender- (g) specific days absent. Total costs due to obesity-related absenteeism (ICabs) are monetarily valued as:ICabs=∑a∑c∑gDAacgobesity·Pag, where Pag is the evaluation of paid and unpaid work as specified above. B.2.2 Presenteeism To the best of our knowledge, no survey measuring obesity-related presenteeism has been conducted in the Czech Republic so far. Our assumption is based on literature review summarised in Table 6. The review contains studies predominantly from the USA. Only one study focuses on five European countries (France, Germany, Italy, Spain and UK) and states that obesity-related presenteeism means 4–33 more days lost compared to normal weight, depending on obesity class. Based on the literature review, we assume that the average annual rate of presenteeism for obese individuals in the Czech Republic is 2 days lost for both men and women. This assumption is rather conservative, but we prefer a conservative approach rather than overestimating the costs. In sensitivity analysis, we also estimate the costs for 1, 3 and 4 days lost due to presenteeism. Yearly lost productivity due to presenteeism is valued as:ICpres=∑gpg·Eg·PL·Pg, where Eg is the number of employed people in working-age population (distinguished by gender g), and pg·Eg is the number of obese people in labour force (p is prevalence of obesity in working-age population, i.e. 25–64 years old), PL stands for productive days lost due to presenteeism and Pg is the gender-specific valuation of paid and unpaid work.8 B.2.3 Premature mortality To calculate the costs of premature mortality due to obesity-related diseases, we use the data from the IHIS [42]. These data contain the number of deaths at each age cathegory due to comorbidities of obesity. To evaluate the lost productivity, we take into account not only the productive years lost (i.e. years before retirement age), but also the years after retirement age as not including this would imply that life of retired people has no value [50]. The productive years are valued by average monthly salary specific for 5-year age group and gender, and by the value of unpaid work (approximated by hourly salary of cleaning services employee), assuming that both men and women perform daily unpaid work according to a survey by the Czech Academy of Sciences [45]. The years after retirement age and before life expectancy are valued by the amount of unpaid work according to a survey by the Czech Academy of Sciences [11, 45]. In COI studies, costs are computed for one given year, but in case of premature mortality, the net present value of future lost earnings is included [55]. The value of productivity losses is discounted to present value using a discount rate:NPV=∑t=0nFV(1+i)t, where i is the discount rate, FV stands for future value and t is the amount of years lost. The discount rate usually ranges between 0 and 10% [55]. We use the discount rate of 3% as suggested by Segel et al. (2006) [55], but because the discount rate affects the results largely, we also perform sensitivity analysis with discount rates 1% and 5% [70]. The obesity-attributable costs of premature mortality are computed as:ICmort=∑a∑c∑gPAFacg·Macg·0.5·Pag+∑t=1ret-1Pag(1+i)t+∑t=retexpUWg·WHW(1+i)t, where Macg stands for age-, comorbidity- and gender-specific number of deaths. Only half of the productivity is accounted for in the first year (t=0) to correct for different occurences of death during the year. Until retirement (ret), the lost productivity is valued by both paid and unpaid work. From retirement age until the average life expectancy age (exp), the lost productivity is valued by the value of unpaid work. We do not take into account background death rates in future years (i.e. deaths that would have occurred separate from obesity-related causes) as it is not standard in this methodological approach. B.3 Sensitivity analysis The robustness of our results is tested using sensitivity analysis where we vary several essential parameters used in the evaluation of costs of obesity (Table 8):PAF are recomputed using the 95% confidence interval of relative risks from Dobbins et al. [8] and Guh et al. [7]. PAF are recomputed using the relative risks from the Dynamo project. This project provides relative risks for selected diseases as indicated in Table 8 and is based on studies from Europe. Additionally, it provides adjustments of relative risks based on age. Prevalence data from the European Health Examination Survey [52] and European Health Interview Survey [52] (EHES and EHIS) are used. These surveys were conducted in 2014 in the Czech Republic.9 EHES is a survey focusing on working population (aged 25–64) and the data are collected by physicians. EHIS focuses on all population aged 15+ years and the data are self-reported. We use mainly the EHES data and complete them with the EHIS data in age groups 15–24 and 65+. The prevalence of obesity in population 15+ years is 25.3% for men and 22.9% for women [51], while the prevalence in working population (25–64 years) is 29.1% for men and 24.7% for women [52]. Discount rate of 1% and 5% is used for computing the costs of premature mortality. Unpaid work is completely excluded from total costs. Presenteeism is computed for missing 1, 3 and 4 days of work (baseline value is 2 days). Table 8 Relative risks from the Dynamo project Women Men Age adjustment Ischemic heart disease 2 2 × 0.70 age over 65 Stroke 1.55 1.5 × 0.75 from age 65 Diabetes 7 5.5 × 0.92 from age 60 × 0.90 from age 75 Breast cancer 1 1 × 1.25 over age 50 women Colorectal cancer 1.1 1.4 × 0.90 from age 45 Kidney cancer 1.8 1.55 – Gallbladder cancer 1.85 1.25 × 1.17 from age 45 men ×0.80 from age 45 women Endometrial cancer 2.5 – – Source: The Dynamo Project [71] Appendix C: Comparison with the OECD study In 2019, OECD published a study estimating the burden of obesity in 52 countries, including the Czech Republic [2]. The estimate of economic burden of overweight is estimated as 4% of GDP, which is much higher than ours estimate of 0.8% of GDP. The OECD study and our research differ in many aspects. Here are the most important ones:The OECD study does not distinguish obesity from overweight. Our paper focuses purely on obesity (BMI > 30), whereas the OECD study evaluates the costs of overweight (BMI > 25). Even though pre-obesity, or overweight (BMI 25–30), is dangerous in a way that it often leads to obesity, it is associated with significantly lower risk of developing serious complications from comorbidities, as opposed to obesity. The methodology of the OECD study and our study is completely different. While the OECD study uses a microsimulation model, we perform a country-level COI study using the top-down approach. As stated in the OECD study, studies using the top-down approach usually provide a lower-bound estimate. The microsimulation model creates a synthetic population based on national demographic characteristics and risk factors from which life expectancy, disease prevalence and disability-adjusted life years are calculated. Different data sources are used in the OECD study, for example:The data used for the estimation of healthcare utilization costs in the Czech Republic are estimated from the Netherlands data. In our study, we use the data from the Czech Republic, specifically from the General Health Insurance Fund. Additionally, each study computes the costs for different set of diseases. The data used in the estimation of costs related to unemployment and absenteeism come from the SHARE survey, which focuses on population 50-63 years old and the data are self-reported. In our study, we have data from the Institute of Health Information and Statistics (IHIS) of the Czech Republic for all the age groups. The estimation of presenteeism related costs derives the days missed from absenteeism in a specific country using a study by Goetzel et al. (2004). From this study, the ratio of presenteeism:absenteeism missed days is derived for five health conditions: asthma, COPD, CVD, cancer and diabetes. The ratio is always larger than 1 (in three cases larger than 2), leading to very high estimated costs of presenteeism. In our study, we take a conservative approach based on a detailed literature review. The OECD study does not provide the specific amount of components of total costs for each country. Instead, it provides an aggregate result. We provide a very thorough sensitivity analysis where we vary the level of several key parameters of our analysis. Appendix D: Results D.1 Healthcare utilization costs See Tables 9 and 10.Table 9 Comorbidities of obesity Diagnosis ICD-10 code PAF (%) and 95% CI Women Men Asthma J45 18.0 (9.2, 27) 10.6 (3.7, 17.9) Dorsalgia M54 33.7 (26.3, 41) 33.4 (26.0, 40.7) Type 2 diabetes mellitus E11, E13, E14 76.2 (69.3, 81.8) 61.4 (55.7, 66.5) Ischemic heart disease I20–I25 37.1 (33.7, 40.5) 16.6 (12.4, 33.4) Leukemia C91–C95 8.2 (2.2, 14.4) – Malignant melanoma C43, D03 – 6.7 (1.9, 11.7) Stroke I69.4, I64 12.1 (7.0, 17.2) 12.4 (8.4, 16.6) Obesity E66.0, E66.2, E66.8, E66.9, E65 100 100 Cholelithiasis and cholecystitis K81, K80 27.0 (4.6, 50.0) 10.6 (1.1, 21.0) Osteoarthritis M15–M19 21.2 (19.8, 22.6) 46.9 (32.7, 59.9) Pulmonary embolism I26 41.3 (31.1, 51.1) 41.0 (30.8, 50.8) Endometrial cancer C54.1, C55, D07.0, D39.0 38.4 (34.9, 41.8) – Kidney cancer C64, C65, C66, D30.0 - D30.2 31.5 (28.1, 34.8) 18.5 (14.4, 22.5) Breast cancer D05, D24, D48.6, C50 3.5 (1.4, 5.8) – Pancreatic cancer C25, D01.7, D13.6, D13.7 14.4 (4.6, 25.2) 26.3 (15.2, 37.7) Colon cancer C18, D12.0–D12.6 15.6 (12.7, 18.5) 20.8 (14.0, 27.8) Ovarian cancer C56, D27, D39.1 7.3 (5.3, 9.2) – Gallbladder cancer C23, C24, D13.5 18.7 (8.2, 29.6) 11.5 (4.5, 19.0) Congestive heart failure I50 18.0 (1.9, 35.4) 17.9 (6.2, 30.5) Hypertension I10–I15 28.5 (14.2, 42.8) 18.9 (12.4, 25.5) Source: Guh et al. (2009) [7], Dobbins et al. (2013) [8]. ICD codes are taken from De Oliveira et al. (2015) [72] Table 10 Healthcare utilization costs Diagnosis Costs (95% CI) % of total costs Type 2 diabetes mellitus 2342.4 (2128.3, 2526.7) 20.60% Ischemic heart disease 2137.8 (1833.5, 2942.9) 18.80% Osteoarthritis 1898.0 (1462.8, 2297.7) 16.69% Dorsalgia 1378.8 (1074.9, 1680.6) 12.12% Hypertension 718.6 (403.2, 1037.7) 6.32% Congestive heart failure 548.9 (124.7, 1008.0) 4.83% Kidney cancer 426.2 (362.2, 488.3) 3.75% Colon cancer 408.7 (300.2, 519.3) 3.59% Cholelithiasis and cholecystitis 268.3 (40.2, 506.1) 2.36% Obesity 246.9 (246.9, 246.9) 2.17% Pulmonary embolism 220.1 (165.6, 272.6) 1.94% Asthma 191.4 (86.4, 301.0) 1.68% Breast cancer 138.9 (54.6, 229.4) 1.22% Stroke 130 (81.9, 179.7) 1.14% Pancreatic cancer 89.0 (43.3, 137.6) 0.78% Endometrial cancer 83.1 (75.6, 90.5) 0.73% Leukemia 64.9 (17.3, 113.5) 0.57% Ovarian cancer 43.5 (31.7, 54.7) 0.38% Gallbladder cancer 20.4 (8.6, 32.9) 0.18% Malignant melanoma 17.1 (4.9, 29.9) 0.15% Total 11,373.0 (8546.9, 14,696.2) 100% Values are in millions CZK D.2 Costs of pharmacotherapy See Table 11. Table 11 Costs of pharmacotherapy ATC classification ATC code Costs (95% CI) For the cure of obesity Antiobesity preparations, excluding diet products A08 0 (0, 0) For the cure of diabetes mellitus Drugs used in diabetes A10 819.8 (744.9, 884.3) For the cure of cardiovascular diseases Antithrombotic agents B01 595.2 (363.3, 906.2) Cardiac therapy C01 83.7 (51.1, 127.4) Antihypertensives C02 95.9 (53.8, 138.4) Diuretics C03 101.4 (56.9, 146.4) Beta blocking agents C07 128.4 (72.1, 185.4) Calcium channel blockers C08 86.5 (48.5, 124.9) Agents acting on the renin-angiotensin system C09 573.3 (321.7, 827.8) Lipid modifying agents C10 430.4 (262.7, 655.3) For the cure of cancer Antineoplastic agents L01 45.4 (32, 59.2) For the cure of arthrosis Anti-inflammatory and antirheumatic products M01 123.6 (96.3, 150.6) Total 3083.2 (2103.1, 4206.9) Values are in millions CZK. ATC groups are chosen based on Hodycova (2009) and Dee (2015) ([19] and [23]) aPharmaceuticals for the treatment of obesity are not covered by insurance D.3 Costs of absenteeism See Table 12. Table 12 Results—absenteeism Days lost Including unpaid work Excluding unpaid work Women (25–64) 2,616,984 4131 3450 Men (25–64) 2,818,935 5105 4616 Total 5,435,919 9237 8067 Values are in millions CZK D.4 Costs of presenteeism See Table 13. Table 13 Results—presenteeism Including unpaid work Excluding unpaid work Days lost 1 day 2 days 3 days 4 days 1 day 2 days 3 days 4 days Women 1434 2868 4302 5736 1195 2390 3585 4780 Men 2091 4182 6273 8364 1889 3779 5668 7558 Total 3525 7050 10,575 14,100 3084 6169 9253 12,338 Values are in millions CZK D.5 Costs of premature mortality See Table 14. Table 14 Results—premature mortality Including unpaid work Excluding unpaid work Discount rate 1% 3% 5% 1% 3% 5% Women 6288 5652 5149 1215 1066 949 Men 4951 4356 3892 2948 2606 2335 Total 11,239 10008 9041 4163 3672 3285 Values are in millions CZK Declarations Conflict of interest Partial financial support was received from not-for-profit organization Czech Priorities. The authors have no relevant financial or non-financial interests to disclose. 1 The extrapolation coefficient is equal to 1.79 as there were 5.95 million people insured at GHIF in 2018 and Czech population was 10.65 million. 2 PAF are used in 5-year age groups when the data allows it. 3 Total healthcare costs in the Czech Republic were 430.9 billion CZK in 2018 [54]. 4 These data are stratified by gender and 10-year age groups. 5 ICD-10 codes are not available for prescribed medicaments. 6 We approximate this by average wage of cleaning services worker [46]. 7 The wage of household worker is not available for men and women separately. 8 We do not distinguish salaries based on age in this part: Pg=PWg·GSg+UWg·WHW. 9 A new survey started in 2019, but due to the lack of respondents, the data are not available in 5-year age groups. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References 1. Dobbs R Sawers C Thompson F Manyika J Woetzel J Child P McKenna S Spatharou A Overcoming obesity: An initial economic analysis 2014 USA McKinsey Global Institute 2. OECD. The Heavy Burden of Obesity. OECD Publishing, 2019a. 10.1787/67450d67-en. https://www.oecd-ilibrary.org/content/publication/67450d67-en 3. Popkin, B.M., Du, S., Green, W.D., Beck, M.A., Algaith, T., Herbst, C.H., Alsukait, R.F., Alluhidan, M., Alazemi, N., Shekar, M.: Individuals with obesity and COVID-19: A global perspective on the epidemiology and biological relationships. Obes. Rev. 21(11), e13128 (2020). 10.1111/obr.13128 4. Bruthans, J.: Studie Czech post-MONICA a studie Czech EUROASPIRE: kardiovaskulární rizikové faktory a jejich kontrola v obecné populaci a u osob se stabilní ischemickou chorobou srdeční/Study Czech post-MONICA and study Czech EUROASPIRE: cardiovascular risc factors and their control in general population and in population with ischemic heart disease. http://www.szu.cz/uploads/documents/szu/akce/materialy/14.10.2019/BRUTHANS.pdf (2019). Accessed 17 May 2020 5. NIPH. Výskyt nadváhy a obezity (Prevalence of overweight and obesity). http://www.szu.cz/uploads/documents/chzp/info_listy/Vyskyt_nadvahy_a_obezity_2018.pdf (2018). Accessed 17 Oct 2020 6. Byford S Torgerson DJ Raftery J Economic note: Cost of illness studies BMJ 2000 320 7245 1335 10.1136/bmj.320.7245.1335 10807635 7. Guh, D.P., Zhang, W., Bansback, N., Amarsi, Z., Birmingham, C.L., Anis, A.H.: The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis. BMC Public Health 9(1), 88 (2009). 10.1186/1471-2458-9-88 8. Dobbins, M., Decorby, K., Choi, B.C.K.: The association between obesity and cancer risk: a meta-analysis of observational studies from 1985 to 2011. ISRN Prevent. Med. 2013, 680536 (1985). 10.5402/2013/680536 9. Effertz T Engel S Verheyen F Linder R The costs and consequences of obesity in germany: a new approach from a prevalence and life-cycle perspective Eur. J. Health Econ. 2016 17 9 1141 1158 10.1007/s10198-015-0751-4 26701837 10. Cawley J Meyerhoefer C The medical care costs of obesity: an instrumental variables approach J. Health Econ. 2012 31 1 219 230 10.1016/j.jhealeco.2011.10.003 22094013 11. Konnopka, A., Bödemann, M., König, H.H.: Health burden and costs of obesity and overweight in Germany. Eur. J. Health Econ. 12(4), 345–352 (2011). 10.1007/s10198-010-0242-6 12. Arterburn, D.E., Maciejewski, M.L., Tsevat, J.: Impact of morbid obesity on medical expenditures in adults. Int. J. Obes. 29(3), 334–339 (2005). 10.1038/sj.ijo.0802896 13. Sturm R The effects of obesity, smoking, and drinking on medical problems and costs Health Aff. 2002 21 2 245 253 10.1377/hlthaff.21.2.245 14. Wang, F., McDonald, T., Bender, J., Reffitt, B., Miller, A., Edington, D.W.: Association of healthcare costs with per unit body mass index increase. J. Occup. Environ. Med. 48(7), 668–674 (2006). 10.1371/journal.pone.0247307 15. Raebel, M.A, Malone, D.C, Conner, D.A., Xu, S., Porter, J.A., Lanty, F.A.: Health services use and health care costs of obese and nonobese individuals. Arch. Int. Med. 164(19), 2135–2140 (2004). 10.1001/archinte.164.19.2135 16. Borg, S., Persson, U., Ödegaard, K., Berglund, G., Nilsson, J., Nilsson, P.M.: Obesity, survival, and hospital costs-findings from a screening project in Sweden. Value Health 8(5), 562–571 (2005). 10.1111/j.1524-4733.2005.00048.x 17. Krueger, H., Krueger, J., Koot, J.: Variation across Canada in the economic burden attributable to excess weight, tobacco smoking and physical inactivity. Can. J. Public Health 106(4), e171–e177 (2015). 10.17269/cjph.106.4994 18. Tuzarová, Kateřina: Společenské náklady obezity v České republice. Master’s thesis, Vysoká škola ekonomická v Praze, (2016) 19. Hodycová, T.: Ekonomické dopady rostoucí incidence obezity na zdravotnictví v ČR (diplomová práce)/Economic consequences of growing incidence of obesity on healthcare in the Czech Republic. Master’s thesis, Vysoká škola ekonomická, Institut managementu zdravotnických služeb (2009) 20. WHO. Absenteeism from work due to illness, days per employee per year. https://gateway.euro.who.int/en/indicators/hfa_411-2700-absenteeism-from-work-due-to-illness-days-per-employee-per-year/visualizations/#id=19398 &tab=table (2019). Accessed 27 Jan 2021 21. Dall TM FulgoniIII VL Zhang Y Reimers KJ Packard PT Astwood JD Predicted national productivity implications of calorie and sodium reductions in the american diet Am. J. Health Promot. 2009 23 6 423 430 10.4278/ajhp.081010-QUAN-227 19601482 22. Finkelstein E Fiebelkorn IC Wang G The costs of obesity among full-time employees Am. J. Health Promot. 2005 20 1 45 51 10.4278/0890-1171-20.1.45 16171161 23. Dee, A., Callinan, A., Doherty, E., O’Neill, C., McVeigh, T., Sweeney, M.R., Staines, A., Kearns, K., Fitzgerald, S., Sharp, L. et al.: Overweight and obesity on the island of Ireland: an estimation of costs. BMJ Open 5(3) (2015). 10.1136/bmjopen-2014-006189corr1 24. Lehnert, T., Streltchenia, P., Konnopka, A., Riedel-Heller, S.G., König, H.H.: Health burden and costs of obesity and overweight in Germany: an update. Eur. J. Health Econ. 16(9), 957–967 (2015). 10.1007/s10198-014-0645-x 25. Kang JH Jeong BG Cho YG Song HR Kim KA Socioeconomic costs of overweight and obesity in korean adults J. Korean Med. Sci. 2011 26 12 1533 1540 10.3346/jkms.2011.26.12.1533 22147988 26. Andreyeva, T., Luedicke, J., Wang, Y.C.: State-level estimates of obesity-attributable costs of absenteeism. J. Occup. Environ. Med. Am. Coll. Occup. Environ. Med. 56(11), 1120 (2014). 10.1097/JOM.0000000000000298 27. Johns G Presenteeism in the workplace: A review and research agenda J. Organ. Behav. 2010 31 4 519 542 10.1002/job.630 28. Finkelstein, E.A., DiBonaventura, M.dC., Burgess, S.M., Hale, B.C., et al.: The costs of obesity in the workplace. J. Occup. Environ. Med. 52(10), 971–976 (2010). 10.1097/JOM.0b013e3181f274d2 29. Collins JJ Baase CM Sharda CE Ozminkowski RJ Nicholson S Billotti GM Turpin RS Olson M Berger ML The assessment of chronic health conditions on work performance, absence, and total economic impact for employers J. Occup. Environ. Med. 2005 47 6 547 557 10.1097/01.jom.0000166864.58664.29 15951714 30. Goetzel RZ Long SR Ozminkowski RJ Hawkins K Wang S Lynch W Health, absence, disability, and presenteeism cost estimates of certain physical and mental health conditions affecting us employers J. Occup. Environ. Med. 2004 46 4 398 412 10.1097/01.jom.0000121151.40413.bd 15076658 31. Goetzel RZ Gibson TB Short ME Chu B-C Waddell J Bowen J Lemon SC Fernandez ID Ozminkowski RJ Wilson MG A multi-worksite analysis of the relationships among body mass index, medical utilization, and worker productivity J. Occup. Environ. Med. 2010 52 1S S52 S58 10.1097/JOM.0b013e3181c95b84 20061888 32. Boles M Pelletier B Lynch W The relationship between health risks and work productivity J. Occup. Environ. Med. 2004 46 7 737 745 10.1097/01.jom.0000131830.45744.97 15247814 33. Burton WN Chen C-Y Conti DJ Schultz AB Pransky G Edington DW The association of health risks with on-the-job productivity J. Occup. Environ. Med. 2005 47 8 769 777 10.1097/01.jom.0000169088.03301.e4 16093926 34. Ricci JA Chee E Lost productive time associated with excess weight in the us workforce J. Occup. Environ. Med. 2005 47 12 1227 1234 10.1097/01.jom.0000184871.20901.c3 16340703 35. Gupta S Richard L Forsythe A The humanistic and economic burden associated with increasing body mass index in the eu5 Diabetes Metab. Syndr Obes. Targets Ther. 2015 8 327 338 10.2147/DMSO.S83696 36. Peeters A Barendregt JJ Willekens F Mackenbach JP Mamun AA Bonneux L Obesity in adulthood and its consequences for life expectancy: a life-table analysis Ann. Intern. Med. 2003 138 1 24 32 10.7326/0003-4819-138-1-200301070-00008 12513041 37. Fontaine KR Redden DT Wang C Westfall AO Allison DB Years of life lost due to obesity JAMA 2003 289 2 187 193 10.1001/jama.289.2.187 12517229 38. OECD. The Heavy Burden of Obesity, Technical country notes. OECD Publishing, 2019b. URL http://www.oecd.org/health/the-heavy-burden-of-obesity-67450d67-en.htm 39. Nejedlá M Zdravotní a hospodářské důsledky epidemie obezity a možnosti její prevence ve školách/health and economic consequences of obesity epidemic and possibilities of its prevention in schools Česká Antropologie 2014 64 20 24 40. GHIF. Healthcare costs of selected diseases, 2018. Data obtained upon author’s request 41. IHIS. Information system incapacity for work, 2018a. Data obtained upon author’s request 42. IHIS. Information system deaths, 2018b. Data obtained upon author’s request 43. Abarca-Gómez, L., Abdeen, Z.A., Abdul Hamid, Z., Abu-Rmeileh, N.M., Acosta-Cazares, B., Acuin, C., Adams, R.J., Aekplakorn, W., Afsana, K., Aguilar-Salinas, C.A. et al. Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies in 128.9 million children, adolescents, and adults. Lancet, 390(10113), 2627–2642 (2017). 10.1016/S0140-6736(17)32129-3 44. CSO. Structure of Earnings Survey - 2018. https://www.czso.cz/csu/czso/structure-of-earnings-survey-2018 (2019). Accessed 20 Nov 2020 45. AVČR. Jak Češi tráví čas? Výsledky 1. ročníku výzkumu proměny české společnosti 2015 (How Czechs spend time? Results of first research of changes in Czech society) (2016). https://www.promenyceskespolecnosti.cz/aktuality/aktualita22/Jak_Cesi_travi_cas_TK_20-06-2016.pdf. Accessed 1 Mar 2021 46. ISPV. Informační systém o průměrném výdělku - Mzdová sféra ČR 2018 (Information system on average earnings). https://www.ispv.cz/cz/Vysledky-setreni/Archiv/2018.aspx (2020). Accessed 18 Dec 2020 47. CSO. Life tables for the Czech Republic, cohesion regions, and regions - 2017 - 2018. https://www.czso.cz/csu/czso/life-tables-for-the-czech-republic-cohesion-regions-and-regions-2017-2018 (2020a). Accessed 18 Dec 2020 48. Eurostat. Employment and activity by sex and age - annual data. https://ec.europa.eu/eurostat/data/database (2018). Accessed 19 Jan 2021 49. Bloom, D.E., Cafiero, E., Jané-Llopis, E., Abrahams-Gessel, S., Bloom, L.R., Fathima, S., Feigl, A.B., Gaziano, T., Hamandi, A., Mowafi, M., et al.: The global economic burden of noncommunicable diseases. Technical report, Program on the Global Demography of Aging, (2012) 50. WHO. WHO guide to identifying the economic consequences of disease and injury. Technical report, 2009 51. IHIS. Health interview surveys in the Czech Republic: European Health Interview Survey 2014. https://ehis.uzis.cz/index-en.php?pg=ehis-2014 (2014). Accessed 19 Sep 2020 52. NIPH. Evropský průzkum zdravotního stavu (European health examination survey)—EHES. http://www.szu.cz/ehes (2014). Accessed 19 Sep 2020 53. CSO. Key macroeconomic indicators. https://www.czso.cz/csu/czso/hmu_ts (2021). Accessed 01 Mar 2021 54. CSO. Výsledky zdravotnických účtů ČR 2010–2018 (Health accounts results). https://www.czso.cz/csu/czso/vysledky-zdravotnickych-uctu-cr-2010-2018 (2020). Accessed 1 Mar 2021 55. Segel JE Cost-of-illness studies - a primer RTI-UNC Center Excell. Health Promot. Econ. 2006 1 39 56. NIPH. Vybrané výsledky studie EHES 2019 (Selected results of EHES 2019). http://www.szu.cz/ehes/vybrane-vysledky-studie-ehes-2019 (2019). Accessed 5 Nov 2020 57. Ross H Critique of the philip morris study of the cost of smoking in the Czech Republic Nicot. Tob. Res. 2004 6 1 181 189 10.1080/14622200310001657000 58. Chadimova K Mlcoch T Dolejsi D Hajickova B Mazalova M Lamblova K Dolezal T The economic burden of alcohol consumption in the Czech Republic Value Health 2019 22 S686 10.1016/j.jval.2019.09.1506 59. Whitaker RC Wright JA Pepe MS Seidel KD Dietz WH Predicting obesity in young adulthood from childhood and parental obesity N. Engl. J. Med. 1997 337 13 869 873 10.1056/NEJM199709253371301 9302300 60. Schmid A Schneider H Golay A Keller U Economic burden of obesity and its comorbidities in Switzerland Soc. Prevent. Med. 2005 50 2 87 94 10.1007/s00038-004-4067-x 61. Kleinman N Abouzaid S Andersen L Wang Z Powers A Cohort analysis assessing medical and nonmedical cost associated with obesity in the workplace J. Occup. Environ. Med. 2014 56 2 161 170 10.1097/JOM.0000000000000099 24451611 62. An R Health care expenses in relation to obesity and smoking among us adults by gender, race/ethnicity, and age group: 1998–2011 Public Health 2015 129 1 29 36 10.1016/j.puhe.2014.11.003 25542741 63. Sander B Bergemann R Economic burden of obesity and its complications in Germany Eur. J. Health Econ. 2003 4 4 248 253 10.1007/s10198-003-0178-1 15609192 64. Lehnert T Stuhldreher N Streltchenia P Riedel-Heller SG König H-H Sick leave days and costs associated with overweight and obesity in Germany J. Occup. Environ. Med. 2014 56 1 20 27 10.1097/JOM.0000000000000065 24351899 65. Neovius K Rehnberg C Rasmussen F Neovius M Lifetime productivity losses associated with obesity status in early adulthood Appl. Health Econ. Health Policy 2012 10 5 309 317 10.1007/BF03261865 22827692 66. Pelletier B Boles M Lynch W Change in health risks and work productivity over time J. Occup. Environ. Med. 2004 46 7 746 754 10.1097/01.jom.0000131920.74668.e1 15247815 67. Gates DM Succop P Brehm BJ Gillespie GL Sommers BD Obesity and presenteeism: The impact of body mass index on workplace productivity J. Occup. Environ. Med. 2008 50 1 39 45 10.1097/JOM.0b013e31815d8db2 18188080 68. Kirkham HS Clark BL Bolas CA Lewis GH Jackson AS Fisher D Duncan I Which modifiable health risks are associated with changes in productivity costs? Popul. Health Manag. 2015 18 1 30 38 10.1089/pop.2014.0033 25375893 69. OECD. Ageing and employment policies—statistics on average effective age of retirement. https://www.oecd.org/els/emp/average-effective-age-of-retirement.htm (2018). Accessed 18 Dec 2020 70. Hodgson, T. A., Meiners, M.R.: Cost-of-illness methodology: a guide to current practices and procedures. Milbank Mem. Fund Q. Health Soc. 429–462 (1982). 10.2307/3349801 71. Lobstein, T., Leach, R.J.: DYNAMO-HIA: Report on data collection for overweight and obesity prevalence and related relative risks. https://webgate.ec.europa.eu/chafea_pdb/assets/files/pdb/2006116/2006116_d4_dynamo_hia.pdf, 2010. Accessed: 10 (May 2022) 72. de Oliveira, M.L., Pacheco Santos, L.M., da Silva, E.N.: Direct healthcare cost of obesity in Brazil: an application of the cost-of-illness method from the perspective of the public health system in 2011. PLOS One 10(4), e0121160 (2015). 10.1371/journal.pone.0121160	36449132	PMC9709763	NO-CC CODE	2022-12-01 23:23:39	no	Eur J Health Econ. 2022 Nov 30;:1-21	utf-8	Eur J Health Econ	2,022	10.1007/s10198-022-01545-8	oa_other
==== Front Expert Rev Vaccines Expert Rev Vaccines Expert Review of Vaccines 1476-0584 1744-8395 Taylor & Francis 17931145 11218347 10.1586/14760584.6.5.653 Version of Record Research Article News in Brief New smallpox vaccine approved by US FDA 9 1 2014 2007 6 5 653655 IntegraConverted from TF JATS 1.0 to JATS 1.2 by T&F tfjats-to-jats1.2-converter21 11 2022 © Future Drugs Ltd 2007 Future Drugs Ltd ==== Body pmcACAM2000™, a second-generation, single-dose smallpox vaccine by Acambis (UK and MA, USA), has been approved by the US FDA for vaccination of high-risk subjects and for use in an emergency, such as a bioterrorist attack. Smallpox is a contagious infection caused by variola virus. An infected individual presents with fever and body aches followed by raised, pus-filled blisters on the skin. Death is often associated with complications, such as bacterial infections, and the mortality rate is approximately 30% as estimated by the US CDC. Treatment is only supportive and vaccination is the only means of prevention. Although smallpox has been eradicated worldwide since the 1980s and vaccines are no longer produced, recent threats of bioterrorism have renewed interest and necessity of stockpiling smallpox vaccines, especially by the US Government. “Smallpox could be a particularly dangerous biological threat to us that would kill or debilitate a high percentage of the population,” said Craig Vanderwagen, assistant secretary for preparedness and response, US Department of Health and Human Services. ACAM2000 contains cell-cultured live vaccinia, a related virus to variola that results in protective immunity to smallpox without causing the disease. The cell culture technology allows fast and consistent large-scale vaccine production. Acambis has supplied 192.5 million doses to the CDC Strategic National Stockpile so far. “The licensure of ACAM2000 is a significant milestone not only for Acambis but also for the US Government in its plans to ensure a state of readiness against the threat of smallpox. This has been a highly successful collaboration between Acambis and the CDC, and we look forward to finalizing the warm-base manufacturing contract to secure this vaccine production capability for the US Government for the long term,” said Ian Garland, Acambis’ chief executive officer. This new licensure means “we are more prepared to protect the population should the virus ever be used as a weapon,” added Jesse Goodman, director of FDA’s Center for Biologics Evaluation and Research. Source: The US FDA: www.fda.gov; Acambis PLC, UK: www.acambis.com Toward a hepatitis C vaccine Results from a study using human antibodies to HCV to prevent hepatitis C were presented by Alexander Tarr of the University of Nottingham at the UK Society for General Microbiology’s 161st meeting in Edinburgh last September. The findings may be an important step toward a vaccine against HCV. HCV infection is the most common cause of liver transplantation. The virus infects 180 million people worldwide and up to 500,000 people in the UK alone. As the majority of cases are undiagnosed, long-term HCV infection often leads to liver cirrhosis and cancer. Treatment is costly and not always successful, and there is no vaccine available. The antibodies have been shown to prevent infection with many diverse strains of HCV in laboratory models. “Historically, successful vaccines against viruses have required the production of antibodies and this is likely to be the case for HCV. Identifying regions of the virus that are able to induce broadly reactive neutralizing antibodies is a significant milestone in the development of a HCV vaccine, which will have distinct healthcare benefits for hepatitis sufferers, and could also help us design vaccines for other chronic viral diseases, such as HIV,” said Tarr. “We are also currently exploring the possibility of improving liver transplantation success rates by passively infusing people with these antibodies.” Source: University of Nottingham, UK: www.notting-ham.ac.uk Promising HIV vaccine funded for further clinical development Funding has been given to develop a novel HIV vaccine that was created at the Wistar Institute (PA, USA). The 5-year, US$13.3-million grant will support further clinical development of the vaccine toward clinical trials in humans. The new vaccine is based on a simian adenoviral vector that carries HIV antigens. This novel approach eliminates the main drawback of other HIV vaccines based on human adenoviral vectors, which is the pre-existing immunity to the vectors. Approximately 45% of adults in the USA have pre-existing immunity to human adenoviruses, which interferes with the development of immune responses to HIV antigens when these adenoviral vectors are used. “We believe our vaccine, which is built on a novel chimpanzee virus backbone, has unique immunological advantages over other HIV vaccines currently in testing,” says Hildegund Ertl of the Wistar Institute, also the principal investigator for the newly funded project. “In preclinical studies, the vaccine induced a vigorous immune response in monkeys, and we are hopeful it will do the same in humans.” The project will be carried out by the Wistar Institute researchers in collaboration with other scientists at Emory University, the University of Pennsylvania, Harvard School of Public Health, MRC/UVRI Uganda Research Unit and the National Institute for Communicable Diseases in South Africa. “Based on what we know about HIV and the immune system’s response to the virus, it may not be possible to create a vaccine that generates antibodies able to neutralize HIV,” says Ertl. “For this reason, we and others are now focusing our attention on developing a vaccine that stimulates the production of anti-HIV CD8+ T cells, which have been shown to reduce viral load, although they do not prevent infection. Our vaccine has induced unprecedented levels of activated CD8+ T cells in experimental animals, and we are eager to see if it can perform as well in humans.” Source: The Wistar Institute, PA, USA: www.wistar.org First order of cattle Escherichia coli vaccine dispatched Bioniche Life Sciences Inc. (Canada) has released the first shipment of its Escherichia coli O157:H7 vaccine for cattle. The vaccine received permits from the Canadian Food Inspection Agency (CFIA) in December 2006 to release its vaccine to Canadian veterinarians, through whom cattle owners should request the vaccine if they wish their cattle to be vaccinated. The vaccine aims to reduce E. coli O157:H7 shedding in cattle. Bacterial shedding from cattle to the environment results in approximately 100,000 cases of human illness every year in North America. The vaccine is still under review by the CFIA and the US Department of Agriculture. Additional data have been sent to the CFIA, which confirmed the effectiveness of the vaccine in reducing E. coli O157:H7 shedding in vaccinated animals. For the time being, Bioniche is manufacturing the vaccine at its Belleville (ON, Canada) facility, which will be scaled up when full domestic and international license approvals are achieved. Source: Bioniche Life Sciences Inc., ON, Canada: www.bioniche.com Novavax revealed preclinical data of virus-like particle HIV vaccine Findings from a preclinical study using a virus-like particle (VLP) HIV vaccine were presented by Weimin Liu of the University of Alabama School of Medicine (AL, USA) at the AIDS Vaccine Conference, Seattle (WA, USA) in August. The VLP-based HIV-1 vaccine was developed by Novavax Inc. (MD, USA) in collaboration with researchers from the University of Alabama, Emory University and Duke University with funding from the NIH. VLP consists of viral structural proteins and sometimes a lipid bilayer envelope, hence resembling the real virus, only without any viral nucleic acid. Therefore, VLPs are not infectious but can trigger effective antibody, as well as cell-mediated, immune responses to the virus. Human papillomavirus and hepatitis B vaccines are the first VLP-based vaccines approved by the US FDA, and the VLP technology is being applied to many other viral vaccines under research. “VLPs represent an exciting new vaccine delivery platform for HIV, the full potential of which has not yet been realized,” said Beatrice Hahn of the University of Alabama. “The data presented demonstrate that immunization with VLPs containing a consensus envelope protein is a promising approach to achieving the desired broadly neutralizing antibody response against the HIV-1 virus. These data also substantiate the promise of our recombinant VLP technology to design vaccines to ultimately generate an immune response that can prevent complex diseases, such as HIV/AIDS,” commented Rahul Singhvi, Novavax’s President and Chief Executive Officer. DNA priming followed by HIV VLP boost resulted in HIV-1-neutralizing antibodies to several virus strains including some that are difficult to neutralize, such as HIV-1 subtypes B and C, which are associated with more than 50% of AIDS worldwide. In addition, vaccination with VLPs without adjuvants could trigger comparable levels of antibodies to those induced by adjuvanted protein subunit vaccine. “This vaccine approach differs from most others now in clinical trials in that VLPs are designed to produce neutralizing antibodies. Such antibodies have the potential to block the HIV infection process at its earliest stage,” said Richard Compans of Emory University School of Medicine, also the senior investigator of the research team. In this study, the HIV-1 VLP also carries a modified envelope protein that can induce cross-reactive immune responses against several HIV strains. As genetic variation remains an obstacle in HIV vaccine development, artificially designed sequences that represent most common and essential regions of different HIV envelope proteins may provide a solution. Novavax-supported preclinical studies of the consensus envelope HIV-1 VLP candidate vaccine will be completed in the next 12–18 months and human clinical trials may begin in 2009. In addition, vaccines against influenza and other viral diseases are being developed by Novavax using its proprietary VLP technology. Source: Novavax Inc., MD, USA: www.novavax.com Monoclonal antibody may reduce hospitalizations due to respiratory syncytial virus Results of a randomized, double-blind Phase III clinical trial have revealed that motavizumab, a monoclonal antibody developed by MedImmune Inc. (MD, USA), could reduce hospitalizations due to respiratory syncytial virus (RSV) by 83% compared with placebo. In addition, a 71% reduction in RSV-related lower respiratory infections (LRIs) in outpatients was also observed. RSV is the leading cause of LRIs in infants in the USA, with more than 100,000 cases of hospitalization occuring every year. RSV infections may also occur in the elderly, immunocompromised and those with underlying respiratory or cardiac disease. Motavizumab is a humanized monoclonal antibody that was designed to prevent LRIs due to RSV in pediatric patients. Phase I and II clinical trial results have showed that motavizumab has a similar safety and pharmacokinetic profile to Synagis® (palivizumab) in infants. Also developed by MedImmune, Synagis is a monoclonal antibody approved by the FDA for prevention LRIs due to RSV in high-risk pediatric patients. The product is administered by intramuscular injection and is currently available in 62 countries. This Phase III study included 1410 full-term infants of less than 6 months of age in two Native American populations that are known to have high risk of RSV infections. Motavizumab was well tolerated in these infants and reduced the risk of hospitalization due to RSV compared with placebo. “We are pleased with the results of this study, which support the positive results seen in our Phase III pivotal trial comparing motavizumab and Synagis that were previously reported at the Pediatric Academic Societies meeting in May 2007,” said Genevieve Losonsky, Vice-President of clinical development and infectious disease at MedImmune. Source: MedImmune Inc., MD, USA: www.medimmune.com	17931145	PMC9709927	NO-CC CODE	2022-12-01 23:23:09	no	Expert Rev Vaccines.; 6(5):6532-655	utf-8	Expert Rev Vaccines	2,014	10.1586/14760584.6.5.653	oa_other
==== Front Expert Rev Anti Infect Ther Expert Rev Anti Infect Ther Expert Review of Anti-Infective Therapy 1478-7210 1744-8336 Taylor & Francis 16597208 11221844 10.1586/14787210.4.2.277 Version of Record Research Article Review Smallpox antiviral drug development: satisfying the animal efficacy rule Regulatory requirements for antiorthopoxvirus drugs, Jordan & Hruby Jordan Robert Hruby Dennis Director of Virology, SIGA Technologies Inc., 4575 SW Research Way, Suite 230, Corvallis, OR 97333, USA. [email protected] Chief Scientific Officer, SIGA Technologies Inc., Corvallis, OR 97333, USA † Author for correspondence 10 1 2014 2006 4 2 277289 IntegraConverted from TF JATS 1.0 to JATS 1.2 by T&F tfjats-to-jats1.2-converter21 11 2022 © Future Drugs Ltd 2006 Future Drugs Ltd Concerns over the potential use of variola virus as a biological weapon have prompted new interest in the development of small molecule therapeutics to prevent and treat smallpox infection. Since smallpox is no longer endemic, human clinical trials designed to link antiviral efficacy to clinical outcome have been supplanted by antiviral efficacy evaluations in animal models of orthopoxvirus disease. This poses a unique challenge for drug development; how can animal efficacy data with a surrogate virus be used to establish clinical correlates predictive of human disease outcome? This review will examine the properties of selected animal models that are being used to evaluate poxvirus antivrial drug candidates, and discuss how data from these models can be used to link drug efficacy to clinical correlates of human disease. Keywords: animal model antiviral biodefense inflammatory disease orthopoxvirus pathogenesis regulatory smallpox systemic spread ==== Body pmc10.1586/14787210.4.2.277-F0001 Figure 1. Gene synteny between variola virus and selected species of orthopoxviruses. A pairwise nucleotide comparison of VARV (BSH) with CPV (BR); ECTV (MOS); MPV (ZRE); VACV (WR). Paralogs with similarity >90% fall on the diagonal axis while paralogs with lower homology are plotted away from the diagonal axis. Genes that are not found in variola virus are plotted on the y-axis and genes that found in variola virus but not in the test genome are plotted on the x-axis. The E-value cutoff was set at 0.01 and the percentage identify cutoff was set at 0% to show the maximum degree of similarity between genomes. The majority of genes that are plotted away from the diagonal shift to the x- and y-axes when the E-value cutoff is set to 1×10-25 and percent identify cutoff is set to 40%. The direction of the ORF is shown. In general, the similarity between genes on the diagonal falls from ˜99% in the middle of the genome to ˜90–95% at the termini. The sequence data and tools for this analysis can be found at [102]. CPV (BR); Cowpox virus strain Red; ECTV (MOS); Ectromelia virus strain Moscow; MPV (ZRE); Monkeypox virus strain Zaire 96-\|-16; ORF; Open reading frame; VACV (WR); Vaccinia virus strain Western Reserve; VARV (BSH); Variola virus strain Bangladesh; +: Forward reading frame; −: Backward reading frame; +/−: Forward gene in query aligns with a backward gne in target. 10.1586/14787210.4.2.277-F0002 Figure 2. A comparison of the pathogenesis of orthopovirus infection. (A) Lesion formation caused by footpad inoculation of a naturally hairless mouse with 10 PFU of ectromella virus (ECTV) at day 6 post infection (courtesy of Zentralinstitut for Versuchstiere. Federal Republic of Germany) [47], intranasal inoculation of a rabbit with 150 PFU of rabbitpox virus (RPV) at day 6 post infection (reprinted with permission from Chaed Roy, US Army Research Institute for Infectious Disease [USAMRIID]), intravenous inoculation of a rhesus macaque with 106 PFU of monkeypox virus (MPV) at day 10 post infection (reprinted with permission from Jay Hooper, USAMRIID) [53], and variola virus (VARV) infection of a human at day 22 post infection [101]. (B) Diagram showing time of death (rectangles) and duration of lesion formation (lines) for orthopovirus infections of a mouse, rabbit, monkey or human. (C) Diagram showing the time of appearance of viremia detected in spleen (rabbit, mouse) or in throat swabs (monkeys, humans) and quantified by plaque assay (RPV, ECTV, MPV) in cell culture of chorioallantoic membranes (VARV). The height of the peaks represents relative levels of viral titers and was modified from previously published data [18,47,52,53]. The ‘animal efficacy’ rule (US FDA regulation 21 Code of Federal Regulations [CFR] 314 Subpart I) is designed to evaluate new drug products when human efficacy studies are not ethical or feasible. This rule was developed to facilitate the development of drugs for the prevention and treatment of diseases caused by pathogens considered to be potential bioweapon threats. Many of these pathogens are no longer endemic or exist in isolated areas where outbreaks affect small numbers of people, thereby reducing the feasibility of conducting large-scale human trials to measure drug efficacy. Thus, regulatory approval of new drug products must rely not only on safety data in humans, but also on animal efficacy evaluations that provide data predictive of human disease outcome. Satisfying the animal efficacy rule will be challenging, especially in the area of smallpox antiviral drug development, since variola virus is no longer endemic in the human population, access to variola virus is highly restricted, and animal models of orthopoxvirus disease require the use of surrogate viruses to cause infection and disease. Moreover, current molecular markers and techniques used to describe human disease had not been developed prior to smallpox eradication. Thus, linking animal model data to human disease poses many challenges. This review will provide an overview of poxvirus pathogenesis in humans and other animal hosts. In addition, a description of several animal models currently in use to evaluate antipoxvirus drugs will be discussed. The goal of this review is to define disease parameters in these model systems that can be used to link antiviral efficacy to clinical outcome in humans. Poxvirus replication cycle Variola virus is the etiological agent of smallpox and belongs to the family Orthopoxviridae that includes vaccinia (smallpox vaccine strain), monkeypox, cowpox and camelpox viruses [1]. Orthopoxviruses are large double-stranded DNA viruses that replicate exclusively in the cytoplasm of infected cells. There are four types of infectious virus particles produced during productive infection; intracellular mature virus (IMV), intracellular enveloped virus (IEV), cell-associated enveloped virus (CEV) and extracellular enveloped virus (EEV). The intracellular and extracellular forms of virus are thought to play unique roles in orthopoxvirus pathogenesis [1,2]. IMV particles assemble from crescent-shaped membranes in virus factory areas of the cytoplasm. Particle formation requires a series of temporally regulated proteolytic cleavage events of viral core proteins that result in condensation of the viral core [3]. The core particles are enveloped by intracellular membranes to form IMV particles. Once formed, these particles remain inside the cell and are released upon cell lysis. IMV particles are stable to the environment and are thought to play a role in host transmission [2]. Approximately 10% of the total infectious particles produced during infection are wrapped in virus-modified membranes derived from post-trans Golgi or endosomal membrane systems to form IEV particles [4]. Once formed, IEV particles travel to the cell surface in a microtubule-dependent fashion where the outer membrane of the IEV containing vesicle fuses with the plasma membrane to release CEV that remain associated with the cell surface. Approximately 1% of CEV particles are released into the extracellular space as EEV particles by a process involving motile actin tail formation [5,6]. The extracellular virus particles, EEV and CEV, are responsible for efficient cell-to-cell spread and long-range dissemination of the virus in the host [4,7,8]. Virus variants containing defects in genes required for the production of extracellular virus particles produce small plaques in vitro and are attenuated for virus spread in vivo[9–11]. Virus strains that produce higher proportions of extracellular virus particles as part of the infection cycle are more virulent in mice [7]. Passive immunization with antibodies directed against IMV particles are less protective than antibodies directed against whole virus, suggesting that neutralizing antibodies to EEV antigens play a significant role in disease prevention [12]. Finally, inhibitors of extracellular virus formation or more specifically, production of EEV particles, protect animals from lethal orthopoxvirus infection [13–15]. These observations suggest that extracellular virus particles play an important role in virus spread and orthopoxvirus disease progression. At least seven virus-specific gene products are required for synthesis of extracellular virus particles. These gene products participate in the wrapping of IMV particles (B5R and F13L), transport of particles to the cell surface (F12L), actin tail formation (A33R, A34R and A36R) and release of particles from the cell (A33R, A34R and B5R). EEV particles have a higher specific infectivity compared with IMV particles and are more resistant to complement-mediated neutralization [16,17]. Complement resistance is the result of incorporation of host complement control proteins, CD46, CD55 and CD59, into the outer membrane of the virus [17]. These properties of extracellular virus particles facilitate long-range spread of the virus in the host. Natural history of human orthopoxvirus infections The natural history of orthopoxvirus infections can be divided into two stages, localized replication at the site of infection and systemic spread. Many factors influence the course of disease progression, including the amount of virus entering the host, route of entry and host response to infection. Although variola virus is no longer endemic, the natural history of variola infection and clinical manifestations of smallpox has been well documented [18]. Orthopoxvirus infection of humans produces a spectrum of diseases that range from severe disseminated lesional disease characteristic of the most common type of variola virus infection to localized lesional infection caused by vaccinia virus. Four major types of clinical disease are associated with variola virus infection and are defined by the morphology of the virus-specific lesion and severity of disease symptoms. Ordinary smallpox is characterized by raised pustular skin lesions that can be confluent or discrete. Variola sine eruptione is characterized by fever without rash and requires serological analysis to confirm diagnosis. Flat-type smallpox is characterized by confluent flat pustules, and hemorrhagic-type smallpox is characterized by widespread hemorrhages in the skin and mucous membranes. Both flat- and hemorrhagic-type smallpox are usually fatal with 97% mortality in diagnosed cases [17,19]. Three other species of orthopoxviruses have been found to infect humans and cause disease. Monkeypox virus, which is likely a disease of rodents, can cause a generalized infection in humans resembling a milder version of smallpox. Monkeypox virus is poorly transmissible from person to person, and outbreaks result in small numbers of people contracting disease. Vaccinia virus is a laboratory vaccine strain with no known natural reservoir that causes localized infection and generates protective immunity against variola virus. Cowpox virus, which is thought to be carried by rodents, can be transmitted to humans by contact with infected animals. Cowpox virus infection produces a self-limiting, localized infection similar to vaccinia virus infection [18]. Disease severity in all cases is influenced by host immune status. Individuals suffering from certain skin disorders or who are immunocompromised suffer more severe infection [20]. Other species of orthopoxviruses that are genetically related to variola virus such as camelpox and ectromelia viruses have not been found to infect humans. The genetic basis for susceptibility of poxviruses is not well understood, but is thought to be related to acquisition and adaptive evolution of host-response modifier genes [21–23]. These genes are often found to be virulence factors that downregulate the host immune response and thereby facilitate systemic virus spread. Phylogenetic analysis of poxvirus genomes has identified a number of gene families undergoing positive selection, many of which are candidate host-response modifier genes. These genes tend to localize towards the ends of the viral genome [21,23,24]. Gene families which are evolutionarily more stable encode proteins required for basic virus replication functions and are located in the central region of the genome [23]. Thus, inhibitors that target conserved replication functions developed against one species of orthopoxvirus often work equally well to inhibit replication of multiple orthopoxvirus species in cell culture [14,25]. A pair-wise genomic comparison of variola virus with selected orthopoxviruses demonstrated a number of genes that differ from variola virus at the nucleotide level. The ends of the viral genome appear to have the greatest level of divergence (Figure 1). While the function of most of these genes is unknown, several genes can be classified as potential host-response modifiers. An example of genes found in ectromelia virus that are not found in variola virus is shown in Table 1. Thus, even within Orthopoxviridae, there are species-specific differences in host-response modifier genes. While orthopoxviruses exhibit a broad host range in cell culture, the susceptibility in animals (including humans) is likely to be related to the acquisition and adaptive evolution of host-response modifier genes that provide a selective advantage for virus replication in a particular animal host. Host response to systemic infection mediates disease progression The host response to variola infection and the ability of the virus to counteract the host response, determines the severity of smallpox by limiting virus replication and spread during the early stages of infection and contributing to the immunopathology associated with late-stage infection [18,26–28]. From studies with ectromelia virus, immune cells encounter invading virus in the oral mucosa and express growth factors, chemokines, cytokines and interferon which trigger host immune-effector cells to migrate to the site of infection [29]. The virus counteracts this by producing a variety of host-response modifier proteins that downregulate host inflammatory signals [21,30]. Studies conducted with ectromelia and vaccinia virus have shown that viral mutants lacking functional interleukin-18 binding protein are attenuated for replication in mice [31,32]. Vaccinia virus mutants that lack functional soluble tumor necrosis factor receptor encoded by the cytokine response modifier E (CrmE) gene or chemokine-binding protein are less virulent in mice [33,34]. Moreover, patients who are immunocompromised fail to mount an efficient host response to infection and exhibit higher mortality rates and more severe disease [18]. Likewise, people with certain types of immune-related skin ailments suffer higher rates of complications resulting from vaccination with the vaccinia virus vaccine [20]. Thus, the ability of the virus to spread in the host and cause systemic disease is regulated by the robustness of the initial host response to viral infection and the effectiveness of viral immune evasion strategies. While the host response to infection is important for restricting virus spread, it may also contribute to late-stage disease [35]. In animals, orthopoxvirus infection causes the release of cytokines, chemokines and other mediators of inflammation into the bloodstream that result in vascular dysfunction, coagulopathy and multiorgan failure [29,34,36,37]. Moreover, in humans, the spectrum of disease from fatal hemorrhagic type to self-limiting lesional disease can be associated with the degree of inflammatory response to systemic viral infection. Autopsy results from smallpox patients show dilation of the capillaries and endothelial swelling in the capillary walls, leukocyte mobilization and disseminated intravascular coagulation, consistent with pathology resulting from an inflammatory response to infection [18,26–28]. In light of these observations, smallpox can be thought of as an acute inflammatory disease where the host response to systemic infection stimulates immune cells to release proinflammatory cytokines causing severe disease and death. It is tempting to speculate that virus-mediated inhibition of the host response may exacerbate this condition and cause a build up of activated immune cells which must elicit a stronger cytokine response to overcome virus-specific immune evasion strategies. Variola infection of humans The life-cycle of the variola virus resembles that of other orthopoxviruses in which infection of the natural human host results in virus replication at the periphery followed by systemic spread. The spread of variola in humans has been inferred from animal studies, especially those conducted in mice infected with ectromelia virus [18]. The variola virus is thought to enter the respiratory tract via aerosolized droplets, seeding mucous membranes and passing rapidly into local lymph nodes. Based on animal studies, the virus replicates in the local lymph tissue to produce a primary viremia. The virus then travels to the spleen, liver and reticulo–endothelial system where replication in these organs produces a secondary viremia. If this mechanism of intense viral multiplication occurs in humans, it is remarkable in that it is accompanied by a clinical latency in which physical symptoms of infection are absent. In animal models of orthopoxvirus disease, virus replication stimulates the production of proinflammatory cytokines that can be detected in the blood [19,35]. While techniques for cytokine profiling were not available during the time when smallpox was endemic, hematological analysis of patients suffering from normal forms of smallpox found that thrombocytopenia was relatively common with more severe hematological abnormalities found in hemorrhagic forms of disease. Thrombocytopenia associated with variola virus infection was attributed to disseminated intravascular coagulation, a clinical condition that could be related to excessive inflammatory cytokine secretion [27,28]. While induction of cytokines in response to variola virus infection likely occurs in humans, physical symptoms of an inflammatory response are not evident early in the infection process [21,30]. Clinical latency ends with the rapid onset of severe headache, backache and fever, termed the prodromal phase. The prodromal phase correlates with a secondary viremia in which infectious virus can be detected in the mucous membranes of the mouth and pharynx. The virus invades the capillary epithelium of the dermal layer in skin, perivascular cells and epidermis where replication results in necrosis and the formation of a rash. The spleen, lymph nodes, liver, bone marrow, kidneys and other viscera may contain large quantities of virus based upon data from animal studies. Replication in the epidermis may be enhanced by secretion of virus-specific growth factors that bind to cellular receptors on keratinocytes and stimulate growth [38]. Histological studies of skin lesions from smallpox patients have documented the changes in skin leading to lesion formation [18]. The earliest changes detected were a dilation of the capillaries in the papillary layer of the dermis followed by swelling of the endothelial cells in the wall of blood vessels [18,26]. Infection results in 30% mortality with the cause of death attributed to toxemia, associated with immune complexes and hypotension. Toxemia is a poorly defined clinical condition thought to be caused by an excessive inflammatory immune response similar to septicemia associated with systemic bacterial infections. Examination of dermal layers of blood vessels from autopsy patients infected with variola virus show extensive leakage of the endothelial layer consistent with the presence of high levels of proinflammatory cytokines [18,26]. Treatment of smallpox The most effective therapy for the treatment and prevention of smallpox is vaccination with vaccinia virus [18]. While vaccinia virus is closely related to variola virus, it is much less virulent causing localized infection in normal healthy adults rather than systemic disease. Localized infection generates a robust immune response that results in protective immunity. The immunogenicity of vaccinia virus forms the basis for the current smallpox vaccine which has been used to eradicate endemic smallpox [19]. The route of inoculation also contributes to the reduced virulence of vaccinia virus. Virus entering through the skin as opposed to the oral mucosa produces a localized infection that is cleared by the immune system. While intentional dermal infection or intranasal inoculation with variola virus (variolization) also produces a localized infection of the skin, systemic spread occurs more frequently with virus-induced lesions forming away from the site of inoculation, and fatal smallpox occurring in approximately 1% of patients. Variolization had been used for many years prior to the development of the vaccinia virus-based vaccine [18]. Replication of vaccinia virus is similar to variola virus except that infection remains localized and virus spread is uncommon in normal immunocompetent people. Vaccinia virus establishes productive infection in basilar epithelium following inoculation with multiple stabs from a bifurcated needle to establish a local cellular reaction. By 6–8 days post inoculation, a grayish–white pustule develops that is 1–2 cm in diameter, usually having central umbilication. Central crusting begins and spreads peripherally over a period of 3–5 days. Local edema and a hard crust remain until the third week post inoculation [18]. The host response to infection leads to the development of protective immunity against variola virus. Smallpox antiviral drug candidates Currently, there are no FDA-approved drugs to treat smallpox. A number of compounds that are in early preclinical development have been tested for antiorthopoxvirus activity [25]. In general, two classes of compounds have been found to be active in treating orthopoxvirus infection; those that directly inhibit virus replication and those that target systemic spread. Cidofovir (Vistide®), a broad-spectrum DNA polymerase inhibitor, is an example of the first compound class. Cidofovir which is currently licensed for the treatment of cytomegalovirus retinitis in AIDS patients is also active against orthopoxvirus replication in vitro and in vivo[25,39]. The compound targets the viral DNA polymerase to directly inhibit virus replication [40]. Orally bioavailable prodrugs have been developed to overcome the need for intravenous administration [41]. The second class of compounds is represented by those that target systemic spread. Compounds such as Gleevec, CI-1033 and ST-246 inhibit systemic spread without directly inhibiting virus replication [13–15]. Gleevec and CI-1033 inhibit host Abl-family tyrosine kinases and Erb-1 family of kinases, respectively to prevent formation of EEV particles [13,15]. ST-246 targets the poxvirus F13L protein, which is required for the formation of CEV and EEV particles [14]. While these compounds fail to inhibit IMV particle production in vitro, they have been shown to protect mice from lethal orthopoxvirus infection and/or reduce disease. Thus, inhibitors of poxvirus replication that block replication and/or systemic spread in animal models of orthopoxvirus disease are likely to be effective in reducing disease severity in humans. Animal models of orthopoxvirus disease The specter of bioterrorism has led to renewed interest in developing therapies to treat and prevent smallpox. Animal models of orthopoxvirus disease that are predictive for human disease outcome are an important component of the drug and vaccine development process. A number of animal models of orthopoxvirus disease have been developed to evaluate antipoxvirus compounds [42]. While these models are useful for evaluating antiviral activity of compounds in animals, each model by itself fails to capture all aspects of human disease and, therefore can not be predictive of clinical outcome. Thus, multiple models of orthopoxvirus infection will be required to evaluate antiviral efficacy of poxvirus inhibitor compounds. The most relevant animal models of orthopoxvirus infection involve the use of host-adapted virus and require replication at the periphery followed by systemic virus spread in a manner similar to variola virus infection of humans. These models are appealing because virus replication in target tissue and systemic spread is determined by the host response to infection. Some animal models of orthopoxvirus infection have been developed using nonhost-adapted orthopoxviruses to establish disease. These models require inoculation of animals with large quantities of virus sometimes delivered by unnatural routes. Delivering large amounts of virus by unnatural routes alters pathogenesis by allowing virus access to different tissues and stimulating a non-natural host response. The term ‘host adapted’ is used in a broad sense and is defined as genetic alterations that result in increased virulence for a specific host. These changes may include evolutionary adaptations such as acquisition of host-response modifier functions such as those found in ectromelia virus for example, or unknown changes that result in increased virulence, such as those found in rabbitpox virus, which is genetically very similar to vaccinia virus but more pathogenic in rabbits [43]. To better understand the pathogenesis of orthopoxvirus infection in animals and humans, the natural history of poxvirus infection in several host species is described. Intranasal vaccinia virus/cowpox virus infection of mice Vaccinia virus mouse models have been developed to measure the efficacy of antipoxvirus compounds [39,44]. Similar models have also been developed using cowpoxvirus [36,39]. The pathogenesis of infection is dictated by the route of entry and models using intracranial, intravenous or intranasal inoculation have been described [44–46]. Intranasal inoculation of mice with vaccinia or cowpox virus produces local replication and systemic disease providing a system capable of assessing the antiviral activity of compounds that inhibit multiple steps in the virus life-cycle. To establish infection, mice are inoculated with 104–106 plaque-forming units (PFU) of vaccinia virus or cowpox virus in a small volume (20 µl) to each nare. Although establishment of lethal infection requires a large inoculating dose of virus, the infection starts locally in the nasal tissue and lungs before spreading systemically [39,44]. The virus replicates in the reticulo–endothelial system and can be found in the liver, spleen, lung and kidney [39]. By day 4 post inoculation, mice begin losing weight and become lethargic. Mice continue to lose weight and their general appearance declines with most animals exhibiting signs of severe disease such as ruffled fur, hunched posture and unsteady gate. By day 8 post inoculation, mice are moribund and have lost as much as 30% of their initial body weight. Mortality is the primary end-point in this model with 100% of mice succumbing to infection by day 10 post inoculation. Disease progression can be monitored by measuring the change in weight during the course of infection. The change in weight correlates with systemic replication of virus in mice and is a noninvasive method of determining disease severity. The percentage weight change is useful for determining disease severity when treatment protects mice from lethal infection. Thus, the efficacy of compounds that prevent mortality but do not completely inhibit virus replication and all aspects of disease progression can be assessed in this model. To quantify the level of virus spread, animals must be sacrificed at selected time points post infection and virus titers measured in the liver, spleen, lung, kidney and other organs. Antiviral efficacy is measured by the decreased mortality, inhibition of virus-induced weight loss and reduction in viral titers in the liver, spleen, lung, kidney and other tissues. Ectromelia virus infection of mice Ectromelia virus is a laboratory pathogen of mice that has been used as a model system to characterize disease pathogenesis of orthopoxvirus infections [18,29,47,48]. The pathogenesis of ectromelia virus disease closely resembles human smallpox, with distinct stages of localized replication, systemic virus spread and lesion formation, however, the time course of infection and disease progression is much shorter (Figure 2)[47]. Ectromelia virus enters through abrasions in the skin and replicates in local lymphoid cells [29]. Virus can be detected in these tissues by immunofluorescence within a few hours after inoculation [49–51]. The virus multiplies in the lymphatic endothelial cells, macrophages and lymphocytes within the node over a period of 2–4 days [29]. Following this latency period, the virus spreads through the lymph and enters the bloodstream to cause a primary viremia. The virus is rapidly removed by macrophages lining the sinusoids of the liver, spleen and bone marrow [49]. Infection of the parenchymal cells of liver and lymphoid cells of the spleen produces high virus titers that are released into the bloodstream to cause a secondary viremia [50,51]. In highly susceptible animals, replication in the liver and spleen produces focal necrotic lesions, acute hepatitis and multiorgan failure. In mice that are less susceptible to infection, a rash develops following the secondary viremia [47]. The rash is caused by virus replication in the perivascular cells, dermal endothelial cells and epidermis (Figure 2). A lethal ectromelia virus mouse model has been established to evaluate the efficacy of antiviral drugs [41,48]. In susceptible mice as little as 1 PFU of ectromelia virus causes lethal infection [41]. Mice are inoculated with ectromelia virus either by footpad scarification, which is similar to the natural route of infection, or intranasal delivery [47]. The virus multiplies in the lymphatic endothelial cells, macrophages and lymphocytes within the regional node over a period of 2–4 days. By day 4 post inoculation, animals appear ill with hunched posture, ruffled coat and increased respiration. Viral replication in the liver and spleen and other internal organs cause death in the infected animal between days 6 and 10 post inoculation [47]. The rate of disease progression is similar to the intranasal vaccinia virus model with mortality occurring in 100% of mice by day 10 post inoculation [41]. Disease progression can be monitored by measuring the change in weight during the course of infection. To quantify the level of virus spread, animals are sacrificed at selected time points post infection, and virus titers measured in the liver, spleen, lung, kidney and other organs [14,41]. Detecting virus in these tissues is an excellent measure of virus spread since viral inoculums contain 50–100 PFU, and the virus is likely to be detectable only after significant levels of replication and spread. Like the intranasal vaccinia virus model, antiviral efficacy is measured by decreased mortality, inhibition of virus-induced weight loss and reduction in viral titers in the liver, spleen and other tissues. Rabbitpox virus infection of rabbits Rabbitpox virus is highly adapted to replicate in rabbits and as little as 15 PFU can establish productive infection in most rabbit strains [52]. Rabbits are inoculated with rabbitpox virus in the footpad by intradermal injection or the nasal cavity by aerosol spray. Virus replicates in the mucosa or local lymph tissue to produce a primary viremia which lasts 2–4 days [52]. The virus spreads through the lymph to the blood, ultimately seeding the liver, spleen and other internal organs. Virus replication at these sites often results in multiorgan failure and fatal disease [18]. Rabbits that survive infection of internal organs develop a secondary viremia where high levels of circulating virus seed the endothelial cells that line the dermal blood vessels to produce a rash on the skin that appears by day 6 post inoculation (Figure 2). Vertical transmission of the virus to susceptible rabbits has been observed allowing for the potential assessment of drug effects on virus spread within a population [52]. To establish infection, rabbits are inoculated with 100 PFU of rabbitpox virus by intranasal administration through aerosol delivery to the respiratory tract. By day 6 post inoculation animals develop fever, listlessness and purulent discharges from the eyes and nose [52]. A rash develops between day 6 and 8 post inoculation; however, skin lesions range from a few scattered lesions to confluency. Most animals die without developing a rash and death is accompanied by a fall in body temperature to below normal levels [18,52]. Thus, quantifying lesion number or severity of the rash is a subjective measure of systemic virus spread and may not be possible in all cases. Since rabbits can tolerate more frequent and larger volume blood draws, blood chemistries can be measured to correlate changes in hematological status with disease progression. To quantify the level of virus spread, animals are sacrificed at selected time points post infection and virus titers measured in the liver, spleen, lung, kidney and other organs. Antiviral efficacy is assessed by measuring decreased mortality, reduced virus titers in organs and improvement of hematological status. Histological examination of tissue from sacrificed animals further defines the effects of antiviral compounds on viral pathogenesis. Monkeypox virus infection of nonhuman primates Primate models of orthopoxvirus disease have been developed to evaluate efficacy of new smallpox vaccines [53]. These models are also being considered for the evaluation of antiorthopoxvirus drugs [54]. Administration of 107 PFU of monkeypox virus to nonhuman primates delivered by intravenous or intratracheal injection produces a lethal infection that reproduces the lesional disease characteristic of smallpox and monkeypox in humans (Figure 2). Typically, animals die between days 10 and 14 post infection with over 750 poxvirus lesions [53]. Monkeypox virus infection of nonhuman primates causes fever, decreased oxygen saturation, elevated heart rate and other changes in vital signs that can be measured by telemetry (J Huggins, pers. comm.). Moreover, changes in blood chemistries can be monitored over time since repeated blood draws do not compromise animal health. At selected time points post infection, skin lesions can be quantified in anesthetized animals and disease severity scores calculated based upon lesion number. In addition, virus yields can be measured in serum and whole blood samples and from throat swabs taken from anesthetized animals to quantify the level of viremia [53]. Quantifying lesion numbers and virus yields provide noninvasive methods for evaluating systemic virus spread. Viral titers and histopathology can be measured in selected organs and tissues from sacrificed animals at different time points post infection to determine the degree of virus spread and extent of disease. Thus, there are multiple secondary end-points in this model system that can be used to assess antiviral efficacy. One of the major drawbacks of delivering large amounts of infectious virus directly to the bloodstream by intravenous administration is that this route bypasses the mucosa and produces a lesional disease that is characteristic of post secondary viremia. Variola virus infection of nonhuman primates Variola virus can infect nonhuman primates to produce a disease resembling human smallpox; however, lethal infection requires administration of large quantities of virus delivered by intravenous injection [55]. A recent study has shown that intravenous delivery of 109 PFU of the Harper strain of variola virus produced a significant lesional disease, with more than 250 lesions in one animal, and 100% mortality in all three animals between day 4 and 10 post inoculation. Intravenous infection with 108 PFU produced significantly lower mortality rates (33%) [55]. Thus, high-dose virus delivered by intravenous injection is required to establish disease. Intravenous injection establishes instant viremia bypassing the normal requirement for replication at the periphery and systemic spread. Since the natural history of disease progression is dramatically altered, it is unclear how antiviral efficacy would correlate with human disease outcome in this model. In summary, animal models of orthopoxvirus infection that use host-adapted viruses to establish disease are more likely to be predictive of human disease outcome than infection models that require administration of large quantities of virus delivered by unnatural routes. While small animal models (mice and rabbits) are useful for assessing antiviral efficacy, mortality in these systems is due to multiorgan failure caused by virus replication in vital organs and does not appear to be related to an excessive inflammatory immune response to systemic infection as in human smallpox. In addition, animals die prior to the formation of lesions that are characteristic of smallpox disease eliminating the possibility of using this noninvasive measure of clinical disease severity. A comparison of the pathogenesis of orthopoxvirus infection of mice, rabbits, nonhuman primates and humans is presented in Figure 2. Linking animal efficacy to human disease The molecular pathogenesis of human smallpox is poorly defined since modern techniques such as cytokine profiling and quantitative virology were unavailable prior to eradication. Thus, surrogate markers predictive of human disease outcome have not been well established. From historical accounts, the clinical course of smallpox resembles that of ectromelia virus infection of mice, or rabbitpox virus infection of rabbits, where acute systemic infection results from inoculation with a small amount of virus at the periphery [18]. While it is recognized that no single model is likely to be predictive of human disease outcome, especially now that smallpox is no longer endemic and disease markers cannot be qualified, certain disease models that focus on infection with host-adapted viruses will be more informative in assessing antiviral activity. These models allow evaluation of antiviral activity in a coevolved system (like human smallpox) and measurement of many aspects of viral pathogenesis. Disease models that require high-dose virus administered by unnatural routes may artificially alter pathogenesis by inducing different arms of the immune system or by introducing virus to tissues and cell types not normally associated with viral infection. Animal models where infection is established with a low dose of virus and where pathogenesis is defined based upon the appearance of disease markers such as skin lesions or inflammatory cytokines, and not strictly on the levels of viral replication in the animal would be most desirable for linking antiviral efficacy to human disease. Disease markers that define orthopoxvirus infections Skin pocks The appearance of pox lesions are the hallmark of smallpox. These lesions are easily quantifiable and require local replication and systemic viral spread for their formation. The appearance of pox lesions in humans is well characterized [18]. Smallpox lesions are predictive of clinical outcome and disease grading scales based upon lesion number, have provided clinicians with a noninvasive method for assessing disease severity. Unfortunately, lesion formation is not the primary disease symptom in the most common models of orthopoxvirus pathogenesis. In the ectromelia virus mouse model, susceptible mice die without developing lesional disease. Moreover, in less susceptible mouse strains, lesions are not well-defined pox, but appear as a generalized rash, which is difficult to quantify [18,47]. This situation also occurs in the rabbitpox model of orthopoxvirus disease where most rabbits succumb to infection before developing quantifiable skin lesions [18,52]. In animals that develop rash, the onset of skin lesions occur after the peak of viremia, suggesting that animals must survive replication of the virus in the internal organs in order for the virus to seed endothelial cells in the capillaries of the skin. Mouse models have been developed that emphasize lesion formation [45]. Tail vein injection of vaccinia virus into normal mice results in the appearance of quantifiable, pox-like lesions that resolve over the course of several weeks [14,45]. While this model is useful for measuring virus replication and pox-like lesion formation in an animal, the route of viral entry, which bypasses normal routes of systemic spread, and ensuing nonlethal disease, make lesion formation in this model less predictive of human disease outcome. Lesion formation in the monkeypox model suffers from similar drawbacks. Lesion formation requires administration of large quantities of virus delivered by intravenous injection [53]. In addition, both the mouse tail vein model and the monkeypox nonhuman primate model use nonhost-adapted virus to establish disease. Proinflammatory cytokines Smallpox is an inflammatory disease and the cytokine response to infection would likely be predictive of disease outcome. However, methods for assessing cytokine levels were not available during the time when smallpox was endemic, thus no direct correlation between cytokine levels during smallpox outbreaks and clinical severity of disease can be made. Moreover, in animal models of orthopoxvirus disease, mortality is not caused by an excessive inflammatory immune response to systemic infection, but correlates more closely with virus replication in vital organs [18,36]. While the cytokine response is an important marker of the host response to infection it may not be possible to use this disease marker to predict clinical outcome in humans. Systemic virus spread Smallpox is a systemic disease with the appearance of generalized pox lesions and lethal consequences of virus infection being directly related to the degree of virus spread [18,35]. In animal models of orthopoxvirus disease, the degree of systemic spread can be quantified by measuring viral titers in the liver, spleen and other internal organs. While there is little information on the levels of virus replication in humans during natural infection, the virus has been cultivated from oropharangeal secretions of smallpox patients during the prodromal phase of infection and at beginning stages of lesion formation [18]. In addition, autopsy results from patients who died within several days after the first sign of rash showed evidence of necrotic lesion formation in the reticulo–endothelial system consistent with virus replication in internal organs [26]. Thus, measuring viral titers in the internal organs of animals as a measure of systemic spread will likely correlate most closely with human clinical outcome. A summary of disease markers that define pathogenesis of orthopoxvirus infection in different animal models is presented in Table 2. Conclusion Development of antiviral drugs for treatment of diseases where human clinical trials that measure compound efficacy are either unethical or are not feasible poses a challenge for drug development. The animal efficacy rule has been developed to provide data to support regulatory approval of drugs for these indications. The design of animal models that can be used to link antiviral efficacy data to human disease outcome will be challenging since the pathogenesis in these model systems is often caused by different mechanisms. Animal models that measure viral pathogenesis using systems that retain coevolved virus host interactions are likely to be most informative. Expert commentary Development of therapeutics in which human efficacy evaluations are not feasible will require the use of animal models to assess therapeutic efficacy. This requirement will likely be necessary to facilitate regulatory approval of therapeutics for biodefense targets where disease is no longer endemic or the numbers of patients contracting disease is not large enough to conduct statistically robust efficacy evaluations. To satisfy this requirement a link must be established between therapeutic efficacy in animal models of disease to clinical outcome in humans. For development of antiviral compounds to treat smallpox infection, no single model recapitulates all aspects of human disease. To satisfy the animal efficacy rule, a combination of animal models that mimic specific aspects of orthopoxvirus disease will be required. Models that require small amounts of virus delivered through mucosal routes to establish infection are most desirable for modeling smallpox infection. Infection by this route requires local replication and systemic virus spread to cause disease, thereby reproducing the critical aspects of human smallpox. Moreover, host-adapted virus species such as ectromelia virus in mice and rabbitpox virus in rabbits which have acquired unique strategies through positive adaptive evolution of host-response modifier genes, allow for virus spread in the host in a manner similar to variola virus in humans. Lesional disease models that involve tail vein inoculation of mice with vaccinia virus, or intravenous delivery of large quantities of monkeypox virus to nonhuman primates are limited in that viremia and disease is induced by intravenous injection bypassing normal routes of systemic spread. In addition, delivery of virus by this route may elicit a different host response relative to virus that spreads systemically following localized replication at the site of inoculation. Lesional disease models are important in that they mimic pock formation characteristic of smallpox disease, but would be less useful for assessing efficacy of therapeutics that prevent systemic infection. To link antiviral efficacy to human disease, the pharmacokinetic (PK) and pharmacodynamic (PD) variables that track with drug efficacy must be identified in multiple animal species. To establish a PK–PD link, exposure–response relationships are constructed from sequential measurements of the microbiological end-point (reduction in lesion formation, cytokine response, virus yield from organs or other disease markers) as a function of plasma drug concentration. From these analyses, the PD variable(s) that track with drug efficacy can be determined. This information can be used to establish dosing regimens based upon human drug exposure in the absence of viral infection to determine the dose of a compound that will likely provide therapeutic benefit. Establishing the PK–PD link in several animal models using multiple disease markers to define antiviral efficacy will provide data that has the potential of being predictive of human clinical outcome. Five-year view Future research will focus on standardizing existing animal models of orthopoxvirus disease with the ultimate goal towards compliance with Good Laboratory Practice (GLP) guidelines. GLP compliance will provide regulatory agencies consistent and validated data to assess the quality of therapeutics for biodefense indications. Regulatory agencies will need to come to agreement as to which features of the current disease models will act as surrogate markers of human disease and most likely to be predictive of outcome of human smallpox. This information will provide the framework for development of biodefense-specific therapeutics. Development of new animal models using host-adapted virus strains will be important for assessing efficacy of biodefense therapeutics since infection and disease are established under conditions that more closely resemble human infection. An emphasis on models that can be conducted safely under Biosafety Level 2 conditions will provide broader access to resources for companies and facilitate biodefense drug development. Acknowledgements We would like to thank Chelsea Byrd for critical review of the manuscript. In addition, we would like to thank Mark Buller, Chad Roy and Jay Hooper for supplying images of poxvirus lesions shown in Figure 2. 10.1586/14787210.4.2.277-T0001 Table 1. Ectromelia virus genes not found in variola virus. Open reading frame Start* Finish* Function Ref. EVM004 10,494 9673 Unknown EVM006 12,824 12,597 Unknown EVM007 13,077 12,766 Unknown EVM009 14,351 14,016 CD30, cytokine inhibition [56] EVM010 16,733 14,442 Unknown EVM014 22,782 22,603 Unknown EVM024 37,765 36,932 Unknown EVM047 60,995 61,492 Unknown EVM117.5 133,000 132,839 Phospholipase D-like protein [57] EVM137 152,897 153,427 Virulence factor [58] EVM153 173,249 174,760 Unknown EVM157 178,451 178,996 Tumor necrosis factor binding protein C-terminal [21] EVM159 180,883 181,146 Unknown EVM163 183,858 184,844 Interleukin-1β binding protein [21] Start and finish refer to the starting nucleotide position and ending nucleotide position in the open reading frame, respectively. EVM: Ectromelia virus strain Moscow (AF012825). 10.1586/14787210.4.2.277-T0002 Table 2. Factors that influence pathogenesis of orthopoxvirus infections. Animal Virus Route Inoculum (PFU) Lesions Systemic disease Death (days) Ref. Mouse VV IN 105 - + 6–8 [14,39] ID 105 1 - - [59] ¶ IV 104 20–50 + - [14,60] ECTV IN 1–10 - + 6–8§ [14,41] ID 1 - + 6–8§ [47] Rabbit RPV IN 1–15 Rash + 10–14 [52] Monkey MPV IT 107 >100 + 15–19 [61] IV 107 >100 + 6–16 [53] VARV IN 108.5 Mild - - [62] IV 109 >250 + 4–10 [55] Human VV ID 105 1 - - [63] VARV ID Scab material 1–50 + - [18] MPV IN Unknown >100 + 20–30 [18] VARV IN 1 >100 + 20–30‡ [18] *Mortality observed in approximately 10–15% of patients. ‡Mortality observed in approximately 30% of patients. §In highly susceptible BALB/c mouse strains. ¶Study performed in a hairless immunocompetent mouse. ECTVL Ectromelia virus; ID: Intradermal; IN: Intranasal; IT: Intratracheal; IV: Intravenous; MPV: Monkeypox virus; PFU: Plaque-forming units; RVP: Rabbitpox virus; VARV: Variola virus; VV: Vaccinia virus. • Regulatory approval of therapeutics, where human clinical trials are not possible or are unethical, requires the use of animal models of human disease for assessment of efficacy. • Therapeutic efficacy in animal models must be predictive of human disease outcome. • For smallpox antiviral drug development, no animal models exist that have been proven to be predictive of clinical outcome. • Models that require low-dose virus in the inoculum to establish disease will be most informative since they require localized virus replication and systemic spread which mirrors critical features of human smallpox. • Multiple animal models in different animal species will be required to establish clear dose–response relationships and establish pharmacokinetic variables that track with antiviral efficacy. • Human dosing regimens most likely to provide clinical benefit can be designed based on the pharmacokinetic–pharmacodynamic link established in animals and human pharmacokinetic data. ==== Refs References Papers of special note have been highlighted as: • of interest •• of considerable interest 1 Moss B. Poxviridae and their replication. In: Fields Virology. Knipe DM, Howley PM (Eds), Raven Press, Ltd., NY, USA, 2849–2884 (2001). • An excellent review of poxvirus molecular virology. 2 Smith GL, Vanderplasschen A, Law M. The formation and function of extracellular enveloped vaccinia virus. J. Gen. Virol. 83 (Pt 12), 2915–2931 (2002).12466468 • An excellent review of the role of extracellular virus in orthopoxvirus replication. 3 Byrd CM, Bolken TC, Hruby DE. The vaccinia virus I7L gene product is the core protein proteinase. J. Virol. 76 (17), 8973–8976 (2002).12163618 4 Blasco R, Moss B. Extracellular vaccinia virus formation and cell-to-cell virus transmission are prevented by deletion of the gene encoding the 37,000-Dalton outer envelope protein. J. Virol. 65 (11), 5910–5920 (1991).1920620 5 Cudmore S, Cossart P, Griffiths G, Way M. Actin-based motility of vaccinia virus. Nature 378 (6557), 636–638 (1995).8524400 6 Roper RL, Wolffe EJ, Weisberg A, Moss B. The envelope protein encoded by the A33R gene is required for formation of actin-containing microvilli and efficient cell-to-cell spread of vaccinia virus. J. Virol. 72 (5), 4192–4204 (1998).9557708 7 Payne LG. Significance of extracellular enveloped virus in the in vitro and in vivo dissemination of vaccinia. J. Gen. Virol. 50 (1), 89–100 (1980).7441216 8 Payne LG, Kristensson K. Extracellular release of enveloped vaccinia virus from mouse nasal epithelial cells in vivo. J. Gen. Virol. 66 (Pt 3), 643–646 (1985).3973566 9 McIntosh AA, Smith GL. Vaccinia virus glycoprotein A34R is required for infectivity of extracellular enveloped virus. J. Virol. 70 (1), 272–281 (1996).8523536 10 Stern RJ, Thompson JP, Moyer RW. Attenuation of B5R mutants of rabbitpox virus in vivo is related to impaired growth and not an enhanced host inflammatory response. Virology 233 (1), 118–129 (1997).9201221 11 Zhang WH, Wilcock D, Smith GL. Vaccinia virus F12L protein is required for actin tail formation, normal plaque size, and virulence. J. Virol. 74 (24), 11654–11662 (2000).11090164 12 Boulter EA, Zwartouw HT, Titmuss DH, Maber HB. The nature of the immune state produced by inactivated vaccinia virus in rabbits. Am. J. Epidemiol. 94 (6), 612–620 (1971).4332354 13 Reeves PM, Bommarius B, Lebeis S et al. Disabling poxvirus pathogenesis by inhibition of Abl-family tyrosine kinases. Nature Med. 11 (7), 731–739 (2005).15980865 14 Yang G, Pevear DC, Davies MH et al. An orally bioavailable antipoxvirus compound (ST-246) inhibits extracellular virus formation and protects mice from lethal orthopoxvirus challenge. J. Virol. 79 (20), 13139–13149 (2005).16189015 15 Yang H, Kim SK, Kim M et al. Antiviral chemotherapy facilitates control of poxvirus infections through inhibition of cellular signal transduction. J. Clin. Invest. 115 (2), 379–387 (2005).15690085 16 Vanderplasschen A, Hollinshead M, Smith GL. Antibodies against vaccinia virus do not neutralize extracellular enveloped virus but prevent virus release from infected cells and comet formation. J. Gen. Virol. 78 (Pt 8), 2041–2048 (1997).9267005 17 Vanderplasschen A, Mathew E, Hollinshead M, Sim RB, Smith GL. Extracellular enveloped vaccinia virus is resistant to complement because of incorporation of host complement control proteins into its envelope. Proc. Natl Acad. Sci. USA 95 (13), 7544–7549 (1998).9636186 18 Fenner F, Henderson DA, Arita I, Jazek Z, Ladnyi ID. Smallpox and its eradication. WHO, Geneva, Switzerland, 1–1421 (1988). •• Classic treatise of the smallpox eradication effort that includes a comprehensive review of the pathogenesis, natural history and epidemiology of variola virus infection. 19 Breman JG, Henderson DA. Diagnosis and management of smallpox. N. Engl. J. Med. 346 (17), 1300–1308 (2002).11923491 • An outstanding overview of the natural history and clinical manifestations of variola virus infection. 20 Bray M. Pathogenesis and potential antiviral therapy of complications of smallpox vaccination. Antiviral Res. 58 (2), 101–114 (2003).12742570 21 Chen N, Buller RM, Wall EM, Upton C. Analysis of host response modifier ORFs of ectromelia virus, the causative agent of mousepox. Virus Res. 66 (2), 155–173 (2000).10725549 • Identification and functional charaterization of host-response modifier genes in ectromelia virus. 22 Gubser C, Hue S, Kellam P, Smith GL. Poxvirus genomes: a phylogenetic analysis. J. Gen. Virol. 85 (Pt 1), 105–117 (2004).14718625 23 McLysaght A, Baldi PF, Gaut BS. Extensive gene gain associated with adaptive evolution of poxviruses. Proc. Natl Acad. Sci. USA 100 (26), 15655–15660 (2003).14660798 • A bioinformatic approach was used to evaluate poxvirus genome evolution to identify gene families undergoing positive selection. 24 Massung RF, Esposito JJ, Liu LI et al. Potential virulence determinants in terminal regions of variola smallpox virus genome. Nature 366 (6457), 748–751 (1993).8264798 25 Baker RO, Bray M, Huggins JW. Potential antiviral therapeutics for smallpox, monkeypox and other orthopoxvirus infections. Antiviral Res. 57 (1–2), 13–23 (2003).12615299 • A review of compounds that inhibit orthopoxvirus replication in vitro. 26 Bras G. The morbid anatomy of smallpox. Doc. Med. Geogr. Trop. 4 (4), 303–351 (1952).13033693 • A comprehensive analysis of smallpox pathogenesis using data from autopsy patients. 27 McKenzie PJ, Githens JH, Harwood ME, Roberts JF, Rao AR, Kempe CH. Haemorrhagic smallpox. 2. Specific bleeding and coagulation studies. Bull. World Health Organ. 33 (6), 773–782 (1965).5295401 28 Roberts JF, Coffee G, Creel SM et al. Haemorrhagic smallpox. I. Preliminary haematological studies. Bull. World Health Organ. 33 (5), 607–613 (1965).5295141 29 Esteban DJ, Buller RM. Ectromelia virus: the causative agent of mousepox. J. Gen. Virol. 86 (Pt 10), 2645–2659 (2005).16186218 • Outstanding review of the pathogenesis of ectromelia virus infection. 30 Johnston JB, McFadden G. Poxvirus immunomodulatory strategies: current perspectives. J. Virol. 77 (11), 6093 (2003).12743266 31 Born TL, Morrison LA, Esteban DJ et al. A poxvirus protein that binds to and inactivates IL-18, and inhibits NK cell response. J. Immunol. 164 (6), 3246–3254 (2000).10706717 32 Reading PC, Smith GL. Vaccinia virus interleukin-18-binding protein promotes virulence by reducing gamma interferon production and natural killer and T-cell activity. J. Virol. 77 (18), 9960–9968 (2003).12941906 • An example of the role of host-response modifier genes in pathogenesis of orthopoxvirus infection. 33 Reading PC, Khanna A, Smith GL. Vaccinia virus CrmE encodes a soluble and cell surface tumor necrosis factor receptor that contributes to virus virulence. Virology 292 (2), 285–298 (2002).11878931 34 Reading PC, Symons JA, Smith GL. A soluble chemokine-binding protein from vaccinia virus reduces virus virulence and the inflammatory response to infection. J. Immunol. 170 (3), 1435–1442 (2003).12538705 35 Bray M, Buller M. Looking back at smallpox. Clin. Infect. Dis. 38 (6), 882–889 (2004).14999635 • An excellent review of smallpox relating historical data to our present day understanding of viral pathogenesis. 36 Martinez MJ, Bray MP, Huggins JW. A mouse model of aerosol-transmitted orthopoxviral disease: morphology of experimental aerosol-transmitted orthopoxviral disease in a cowpox virus-BALB/c mouse system. Arch. Pathol. Lab. Med. 124 (3), 362–377 (2000).10705388 • The pathogenesis of cowpox virus infection in mice. 37 Zaucha GM, Jahrling PB, Geisbert TW, Swearengen JR, Hensley L. The pathology of experimental aerosolized monkeypox virus infection in cynomolgus monkeys (Macaca fascicularis). Lab. Invest. 81 (12), 1581–1600 (2001).11742030 • Histopathology of monkeypox infection of nonhuman primates. 38 Kim M, Yang H, Kim SK et al. Biochemical and functional analysis of smallpox growth factor (SPGF) and anti-SPGF monoclonal antibodies. J. Biol. Chem. 279 (24), 25838–25848 (2004).15070899 39 Quenelle DC, Collins DJ, Kern ER. Efficacy of multiple- or single-dose cidofovir against vaccinia and cowpox virus infections in mice. Antimicrob. Agents Chemother. 47 (10), 3275–3280 (2003).14506041 • Antiviral efficacy in lethal vaccinia and cowpox virus mouse models. 40 Magee WC, Hostetler KY, Evans DH. Mechanism of inhibition of vaccinia virus DNA polymerase by cidofovir diphosphate. Antimicrob. Agents Chemother. 49 (8), 3153–3162 (2005).16048917 41 Buller RM, Owens G, Schriewer J, Melman L, Beadle JR, Hostetler KY. Efficacy of oral active ether lipid analogs of cidofovir in a lethal mousepox model. Virology 318 (2), 474–481 (2004).14972516 • Antiviral efficacy in a lethal ectromelia virus mouse model. 42 Smee DF, Sidwell RW. A review of compounds exhibiting anti-orthopoxvirus activity in animal models. Antiviral Res. 57 (1–2), 41–52 (2003).12615302 • A comprehensive review of animal models used to evaluate antiorthopoxvirus compounds. 43 Li G, Chen N, Roper RL et al. Complete coding sequences of the rabbitpox virus genome. J. Gen. Virol. 86 (Pt 11), 2969–2977 (2005).16227218 44 Nelson JB. The behavior of pox viruses in the respiratory tract. J. Exp. Med. 68 , 401–412 (1938).19870795 45 De Clercq E, De Somer P. Effect of interferon, polyacrylin acid, and polymethacrylic acid on tail lesions on mice infected with vaccinia virus. Appl. Microbiol. 16 (9), 1314–1319 (1968).5676405 46 Thompson RL, Price M, Minton SA Jr, Falco EA, Hitchings GH. Protection of mice against the vaccinia virus by the administration of phenoxythiouracils. J. Immunol. 67 (6), 483–491 (1951).14908071 47 Fenner F, Buller RML. Mousepox. In: Viral Pathogenesis. Nathanson N (Ed.), Lippincott-Raven Publishers, PA, USA, 535–553 (1997). 48 Schriewer J, Buller RM, Owens G. Mouse models for studying orthopoxvirus respiratory infections. Methods Mol. Biol. 269 , 289–308 (2004).15114022 49 Mims CA. The response of mice to large intravenous injections of ectromelia virus. I. The fate of injected virus. Br. J. Exp. Pathol. 40 , 533–542 (1959).14422711 50 Mims CA. The response of mice to large intravenous injections of ectromelia virus. II. The growth of virus in the liver. Br. J. Exp. Pathol. 40 , 543–550 (1959).14422712 51 Mims CA. Aspects of the pathogenesis of virus diseases. Bacteriol. Rev. 28 , 30–71 (1964).14127970 52 Westwood JC, Boulter EA, Bowen ET, Maber HB. Experimental respiratory infection with poxviruses. I. Clinical virological and epidemiological studies. Br. J. Exp. Pathol. 47 (5), 453–465 (1966).4288602 53 Hooper JW, Thompson E, Wilhelmsen C et al. Smallpox DNA vaccine protects nonhuman primates against lethal monkeypox. J. Virol. 78 (9), 4433–4443 (2004).15078924 54 Huggins JW, Smee DF, Martinez M, Bray M. Cidofovir (HPMPC) treatment of monkeypox. Antiviral Res. 37 , A37 (1998). 55 Jahrling PB, Hensley LE, Martinez MJ et al. From the cover: exploring the potential of variola virus infection of cynomolgus macaques as a model for human smallpox. Proc. Natl Acad. Sci. USA 101 (42), 15196–15200 (2004).15477589 • Variola virus infection of nonhuman primates. 56 Saraiva M, Smith P, Fallon PG, Alcami A. Inhibition of type 1 cytokine-mediated inflammation by a soluble CD30 homologue encoded by ectromelia (mousepox) virus. J. Exp. Med. 196 (6), 829–839 (2002).12235215 57 Afonso CL, Tulman ER, Lu Z et al. The genome of camelpox virus. Virology 295 (1), 1–9 (2002).12033760 58 Roper RL. Characterization of the vaccinia virus A35R protein and its role in virulence. J. Virol. 80 (1), 306–313 (2006).16352555 59 Neyts J, Leyssen P, Verbeken E, De CE. Efficacy of cidofovir in a murine model of disseminated progressive vaccinia. Antimicrob. Agents Chemother. 48 (6), 2267–2273 (2004).15155231 60 Neyts J, Verbeken E, De CE. Effect of 5-iodo-2´-deoxyuridine on vaccinia virus (orthopoxvirus) infections in mice. Antimicrob. Agents Chemother. 46 (9), 2842–2847 (2002).12183236 61 Stittelaar KJ, van AG, Kondova I et al. Modified vaccinia virus Ankara protects macaques against respiratory challenge with monkeypox virus. J. Virol. 79 (12), 7845–7851 (2005).15919938 62 LeDuc JW, Jahrling PB. Strengthening national preparedness for smallpox: an update. Emerg. Infect. Dis. 7 (1), 155–157 (2001).11266310 63 Frey SE, Couch RB, Tacket CO et al. Clinical responses to undiluted and diluted smallpox vaccine. N. Engl. J. Med. 346 (17), 1265–1274 (2002).11923490 Websites 101 WHO. Smallpox slide set, 2005 www.who.int/emc/diseases/smallpox/ slideset/ 102 PBR shared orthologs syntenic plot www.poxvirus.org/synteny.asp	16597208	PMC9709928	NO-CC CODE	2022-12-01 23:23:09	no	Expert Rev Anti Infect Ther.; 4(2):27713-289	utf-8	Expert Rev Anti Infect Ther	2,014	10.1586/14787210.4.2.277	oa_other
==== Front Expert Rev Vaccines Expert Rev Vaccines Expert Review of Vaccines 1476-0584 1744-8395 Taylor & Francis 19348560 11218618 10.1586/erv.09.4 Version of Record Research Article Special Focus Issue: Influenza Vaccines - Review Candidate influenza vaccines based on recombinant modified vaccinia virus Ankara Candidate influenza vaccines based on recombinant MVA, Rimmelzwaan & Sutter Rimmelzwaan Guus F Sutter Gerd Erasmus Medical Center, Department of Virology, Rotterdam, The Netherlands. [email protected] Paul-Ehrlich-Institut, Department of Virology, Langen, Germany. [email protected] † Author for correspondence 9 1 2014 2009 8 4 447454 IntegraConverted from TF JATS 1.0 to JATS 1.2 by T&F tfjats-to-jats1.2-converter21 11 2022 © Expert Reviews Ltd 2009 Expert Reviews Ltd Recombinant modified vaccinia virus Ankara (MVA) is attractive and promising as a novel viral vector for the expression of foreign genes of interest because it possesses unique properties. In particular, its excellent safety profile and the availability of versatile vector technologies have frequently made MVA the vaccinia virus of choice for preclinical and clinical studies. Owing to its avirulence and deficiency to productively replicate after in vivo inoculation, MVA can be used under biosafety level 1 conditions. In addition to a better safety profile than replication competent vaccinia viruses, the use of MVA leads to similar levels of gene expression and has better immunostimulatory properties and improved efficacy as a recombinant vaccine. In animal models, recombinant MVA vaccines were immunogenic and induced protective immunity against various infectious agents, including viruses, bacteria and parasites. Here we review the progress that has been made in the development of recombinant MVA as a viral vector and candidate pandemic influenza H5N1 vaccine. Specifically, we will focus on the preclinical evaluation of recombinant MVA vector as pandemic influenza A/H5N1 vaccine candidates and discuss the possible future approaches for the use of these novel MVA-based vaccines. Keywords: influenza virus modified vaccinia virus pandemic vaccine ==== Body pmcInfluenza viruses are a major cause of upper respiratory tract infections and are responsible for epidemic outbreaks during the winter months. These seasonal influenza epidemics are caused by currently circulating human influenza A and B viruses. For certain patients who are at a high risk of complications due to influenza, annual vaccination against influenza is recommended. Currently used inactivated vaccines are mostly efficacious and reduce morbidity and mortality provided that the strains used for vaccine production match the epidemic strains. Occasionally, novel subtypes of influenza A viruses are introduced into the human population. These new subtypes originate from the avian reservoir of all subtypes of influenza A viruses [1]. Since the human population has never been exposed to these novel subtypes, these viruses can replicate in their new host without the interference of pre-existing virus-neutralizing antibodies. Under these circumstances, new viruses can cause worldwide influenza epidemics, also known as pandemics, with considerable morbidity and mortality in the human population. In the last century, three pandemics occurred, caused by influenza A viruses of the H1N1, H2N2 and H3N2 subtypes [2]. Currently, influenza A viruses of the H5N1 subtype pose a pandemic threat because transmissions from infected birds to humans have been reported during outbreaks of highly pathogenic avian influenza (HPAI) in poultry since 1997 [3–5]. Since 2003, 387 human cases have been reported, of which more than 60% were fatal [101]. So far, these viruses do not spread efficiently from human to human, although sporadic clusters of human-to-human transmissions have been described [6–8]. When the HPAI viruses acquire the necessary adaptations to cause sustained human-to-human transmission, they may cause a future pandemic outbreak [9]. To limit the impact of such an outbreak, the availability of safe and effective vaccines is desirable and considered a high priority by the WHO [102]. Major efforts have been made to prepare such H5N1 vaccines. However, there were a number of issues that complicated the development of such vaccines, including poor vaccine immunogenicity, long response time, limited production capacity and antigenic variation of circulating strains [103]. Some important developments have been made during the last decade. A review by Kreijtz et al. addresses some of these issues, such as the development of rapid procedures to produce vaccine strains (reverse genetics), cell culture technology to produce vaccine independent of embryonated chicken eggs and the development of adjuvants that increase the immunogenicty of conventional vaccine preparations and that would facilitate dose sparing [10]. Here, we discuss modified vaccinia virus Ankara (MVA) as a viral vector for the delivery of influenza virus antigens as a promising technology that addresses most of the issues outlined above. MVA: a replication-deficient poxvirus vector Modified vaccinia virus Ankara is an attenuated strain of vaccinia virus that was originally developed for use as safer vaccine during the last decades of the smallpox eradication campaign [11,12]. Indeed, MVA was chosen by the Bavarian State Vaccine Institute in Munich (Germany) as a basis for the evaluation of new vaccine preparations and vaccination strategies against smallpox [13,14]. From 1968 to 1988, MVA immunizations were administered to more than 120,000 individuals in Germany without significant adverse events. The excellent safety profile is also observed in the more recent clinical trials aiming at the approval of MVA as a next-generation smallpox vaccine [15–17]. The extraordinary safety profile was exemplified in studies with immunocompromized macaques. Macaques from which T lymphocytes were depleted by treatment with anti-thymocyte globulin or that received a total-body irradiation were subsequently vaccinated with a high dose of MVA. In these severely immunocompromised animals, the virus did not replicate and did not lead to generalized infection normally seen with wild-type, replication-competent vaccinia viruses [18]. After genetic modification of MVA (i.e., insertion of foreign genes under the control of a vaccinia virus promoter), its potential as viral vector was recognized after demonstrating that MVA can efficiently express foreign genes of interest in nonpermissive human cells [19]. At present, recombinant MVA serves as a vaccine-development platform due to its clinical safety and its potency to induce robust immune responses against heterologous antigens [20–22]. Various recombinant MVA vaccines are currently being evaluated in clinical trials, mostly aiming for prophylaxis or therapy of infectious diseases and cancers against which no vaccine is available [23–27]. The wealth of information that has been obtained from a vast and ever-increasing body of basic and clinical research with MVA vector vaccines has provided answers to important questions related to the development and use of viral vector vaccines. One of these questions relates to the influence of pre-existing antivector immunity and the possibility to repeatedly administer the same vector expressing the same or other antigens. Of course, the potential interference with vaccination by pre-existing antivector antibodies is a concern. However, it is of special interest to note that MVA vector vaccines differ in this respect from many other viral vectors that are not effective in the presence of pre-existing antibodies and that induce antibodies against the vaccine antigen of interest inefficiently upon a second administration of the same vector. Preclinical evaluation of MVA vectors already demonstrated its capacity to repeatedly boost immune responses directed to the recombinant antigens. This suggests that immunization with nonreplicating MVA resembles immunization with inactivated vaccines more than with replicating live vaccines, which are more sensitive to antivector immunity [28]; for a review of this see [21,22]. Recently, the first encouraging data have been obtained from the therapeutic immunization of humans [29]. This Phase II clinical trial in colorectal cancer patients tested the immunogenicity of six consecutive applications of a MVA vector encoding the tumor antigen 5T4. Despite efficient induction of MVA-specific antibodies already peaking to high levels after the second vaccination, the antibody responses to 5T4 were boosted after each vaccination, with the highest levels found after the fifth and sixth immunization. Thus, vector-specific antibodies do not seem to have a major impact on the induction of antibody responses specific for the target antigen by repeated administration of recombinant MVA vaccines. Yet, pre-existing antivector immunity may have a greater influence on target antigen-specific CD8+ T-cell responses. Data from MVA vector immunizations with simian immunodeficiency virus antigens in the macaque model firstly suggested that three prior applications of MVA vaccine limited the levels of Gag epitope-specific CD8+ T cells induced by a fourth immunization [30]. MVA delivery of the Ebola virus glycoprotein (GP) in the mouse model demonstrated that prevaccination with replication-competent vaccinia virus inhibited cellular (cytotoxic T cell) but not humoral immune responses to GP [31]. Interestingly, such hindrance of MVA immunogenicity by pre-existing vaccinia-specific immunity could be largely overcome by priming with a GP-specific DNA vaccine [31] or by the use of a new oral vaccination with recombinant MVA attached to TMPEG-modified cationic liposomes [32]. With regard to the induction of T-cell immunity, evidence is accumulating that recombinant MVA vaccines can induce more balanced antigen-specific CD4+ and CD8+ T-cell responses in animal models than other poxviral or adenoviral vectors that elicit either dominant CD4+ or CD8+ antigen-specific T-cell responses [33,34]. Typically, MVA-induced CD4+ T-cell responses are being characterized predominantly as Th1 like [35], which fits well with the recent finding that MVA vector vaccination can be used to protect against allergic sensitization [36]. Strong CD8+ T-cell responses directed against the target antigens were consistently found with various heterologous vaccine prime–MVA boost protocols (first shown with DNA vaccines) [23,24,37,38]. To most efficiently elicit CD8+ T-cell responses with MVA vectors, the delivery of full-length antigen was found to be superior to the expression of peptide antigens or rapidly degradable proteins [39]. These data suggest that the particular importance of cross-priming in MVA-mediated antigen presentation and appears to correlate with recent clinical findings from HIV-1-specific DNA/MVA prime–boost vaccinations in humans [23,40]. Recombinant MVA vaccines expressing HIV proteins as antigens were highly immunogenic, in contrast to more disappointing responses that were elicited by MVA expressing a HIV-1 fusion protein consisting of a string of CD8+ T-cell peptide epitopes [40]. Finally, a long-standing observation is that nonreplicating MVA vectors seem to be paradoxically immunogenic in comparison with fully replication-competent vaccinia viruses, which are able to deliver overall drastically higher amounts of antigen upon administration in vivo[41,42]. Further evidence supports the notion that MVA has particular immunostimulatory properties [43–45]. Recent experiments in mouse models revealed the in vivo synthesis of substantial amounts of type I interferon shortly after MVA vaccine administration and an activation of dendritic cells by both Toll-like receptor (TLR)-9-dependent and TLR-independent pathways [46,47]. Moreover, MVA infection of human monocyte-derived cells can induce or upregulate the expression of genes for host molecules involved in antigen uptake, cytokines, cytokine receptors, chemokines and chemokine receptors [48,49] New developments in MVA vector generation & vaccine production The generation of MVA vectors is straightforward, requiring genomic insertion of heterologous gene-expression cassettes. In most cases, this is achieved by homologous recombination between the MVA genome and DNA from a plasmid that carries recombinant gene sequences being placed under the control of poxvirus-specific promoters. Subsequently, the MVA vector viruses are to be clonally isolated, a procedure that is helped by well-established selection techniques [50]. Concerning influenza candidate vaccines, a very rapid generation of recombinant MVA might be desirable. Recent advances in methodology are likely to significantly shorten the time window required to obtain MVA vectors. One particularly well-performing protocol takes advantage of selective propagation on rabbit kidney RK-13 cells [51]. Another elegant method is based on the achievement to clone and engineer the entire vaccinia virus genome within a bacterial artificial chromosome (BAC) [52] and application of this BAC technology might well be an additional viable route to a more rapid and efficient generation of MVA recombinant vaccines [53]. The first MVA vaccines to be used in humans were propagated in embryonated chicken eggs or in cultures of primary chicken embryo fibroblasts (CEFs) [14]. As of today, CEF cultures are still the preferred and sole substrate for the production of MVA (vector) vaccines and, among vaccine producers, there is considerable experience in the manufacture and use of other live virus vaccines against human infections (e.g., measles and mumps). Due to the highly active development of MVA as new third-generation smallpox vaccine, MVA propagation in CEF has been adapted to a large-scale process by Bavarian Nordic [104]. In its manufacturing facility in Kvistgard, Denmark, this company reportedly uses CEF cells grown in suspension in Wave Bioreactor® bags aiming for a production capacity of up to 60 million doses of MVA vaccine per year. In addition to primary CEF culture, the potential use of cell lines could be of great interest to establish a robust and commercially viable manufacturing process with a controlled seed-lot system, full characterization of the production cells and a lower risk of introducing adventitious agents. For example, the CEF cell line DF-1 can be used to efficiently generate and amplify MVA vector viruses at the laboratory scale [54]. In addition, although MVA cannot productively replicate in most cell lines of mammalian origin, the virus has been found to efficiently multiply in the hamster kidney cell line BHK-21 [55,56] and, more recently, also in rat IEC-6 cells [57]. However, it still remains to be determined if these or other continuous cell lines fulfill the technical and regulatory requirements for development of a commercial manufacturing process for MVA vector vaccines. MVA induces protective immunity against influenza virus infection The potential of recombinant MVA to induce protective immunity was already demonstrated 15 years ago. A recombinant MVA was constructed that expressed the hemagglutinin (HA) and NP genes of influenza virus A/PR/8/34 (H1N1) under the control of the synthetic early/late vaccinia virus promotor sP[41]. A single immunization with 108 plaque-forming units (PFU) of MVA-HA-NP induced strong antibody responses that could be boosted by subsequent immunizations. Furthermore, a single immunization with as low as 105 PFU of MVA HA-NP afforded protection against a lethal challenge infection with influenza virus A/PR/8/34 4 weeks later. The protective immunity not only correlated with the induction of virus-specific antibodies but also with anamnestic cytotoxic T lymphocyte (CTL) responses detected in MVA-HA-NP-primed mice. Interestingly, the same MVA vector vaccine also provided protection against influenza challenge upon oral delivery [58]. The enteric administration of two doses (108 PFU) MVA-HA-NP elicited serum anti-H1 IgG and mucosal anti-H1 IgA antibodies and protected the upper and lower respiratory tract upon influenza virus challenge. Furthermore, vaccination with this MVA-HA-NP vaccine afforded enhanced recovery from infection with a heterosubtypic (H3N2) influenza A virus strain, which correlated with the induction of cross-reactive CTL responses in the vaccinated mice. More recently, recombinant MVA were constructed that express the HA genes of influenza A/H5N1 viruses A/Hong Kong/156/97 (A/HK/97) and A/Vietnam/1194/04 (A/VN/04), which originate from clades 0 and 1 of A/H5N1 viruses, respectively [59]. Expression of HA was under control of the vaccinia virus promoter PsynII. Initially, these MVA recombinants were evaluated in mice and, upon a single immunization with 108 PFU, both constructs proved to be immunogenic. However, higher antibody titers were achieved with the recombinant MVA expressing the HA of A/HK/97 (clade 0). After a second immunization, the homologous antibody titers against A/VN/04, in particular, increased, which crossreacted with influenza virus A/HK/97 but not with influenza virus A/Indonesia/5/05 (A/IND/05), a virus belonging to yet another clade (clade 2.1). However, MVA-HA-VN/04-vaccinated mice were not only fully protected against infection with the homologous strain but also against infection with influenza viruses A/HK/97 and A/IND/05. Protective immunity was assessed by scoring clinical signs of infections (e.g., weight loss), virus titers in the lungs and immunohistochemistry. By contrast, vaccination with MVA-HA-HK/97 only afforded protection against homologous challenge infection. Since promising results were obtained with MVA-HA-VN/04, this vaccine candidate was further evaluated in a nonhuman primate model [60]. To this end, cynomolgus macaques were immunized twice with MVA-HA-VN/04 and then challenged with influenza virus A/Vietnam/1194/04 (clade 1) or A/Indonesia/5/05 (clade 2.1) to assess the level of protective immunity. Immunization with MVA-HA-VN/04 induced antibodies and prevented replication of both viruses used for challenge infection in the upper and lower respiratory tract and the development of fever and severe necrotizing broncho–interstitial pneumonia. Furthermore, vaccination was well tolerated and did not provoke a rise of body temperature in vaccinated animals. Therefore, MVA-HA-VN/04 is a promising vaccine candidate for the induction of protective immunity in humans against highly pathogenic H5N1 avian influenza viruses that originate from clades of antigenically distinct viruses. Based on these promising results obtained in mice and macaques, further development of recombinant MVA as pandemic influenza vaccine candidates seems warranted. Collectively, recombinant MVA expressing selected influenza virus proteins are promising and attractive vaccine candidates for the induction of protective immunity against pandemic influenza. The production of vaccine seed strains is fairly easy and straightforward and production of vaccines might even be performed in a flexible way independent of embryonated chicken eggs in CEF. This may seem contradictive; however, CEF cells can be produced in advance and cryopreserved until use. If this is achievable at a large scale, production can start without delay as soon as the seed virus becomes available and independent of a source of embryonated chicken eggs, which may be in short supply when HPAI viruses are circulating. Furthermore, the use of recombinant MVA expressing the HA gene of A/H5N1 influenza viruses was highly immunogenic. A major advantage of MVA as a vaccine candidate over other vaccine preparations currently under evaluation is that adjuvant systems are not required for high immunogenicity, which will most probably increase their acceptance. Although not studied in great detail, a single dose of recombinant MVA vaccine may be sufficient to protect against infection with a (homologous) virus. However, more research is required to assess the minimal dose and number of vaccinations required to afford protection against homologous and heterologous virus strains. Nevertheless, two immunizations with a high dose of a candidate MVA-based H5 vaccine induced protective immunity against multiple clades of influenza virus A/H5N1 strains in mice and macaques. This indicates that the MVA-H5 vaccine can induce broad protective immunity. Since the immunogenicity of the MVA-based vaccines is not affected substantially by pre-existing antibodies against the vector and repeated vaccinations can induce antibody responses to the target antigens, this approach could possibly be used for repeated vaccination against seasonal influenza virus strains. Based on their favorable properties (Table 1), we conclude that recombinant MVA is an attractive platform for the development of next-generation influenza vaccines. Expert commentary During the last decade, the transmission of HPAI A viruses of the H5N1 subtype from infected poultry to humans has raised our awareness of our inadequate preparedness for the next influenza pandemic. The timely availability of safe and effective vaccines would be a cornerstone in controlling the impact of an influenza pandemic. However, the timely delivery of sufficient doses of effective and safe vaccines is a point of concern and has spurred the development of adjuvants that improve the immunogenicity of vaccines and that allow dose sparing and the development of novel production technologies. Recombinant MVA was recently evaluated preclinically as a viral vector for the delivery of the influenza A/H5N1 virus HA. The use of MVA-H5 induced protective immunity in mice and macaques against challenge infection with homologous and heterologous A/H5N1 influenza virus strains. Based on these results and the unique properties attributed to MVA, recombinant MVA expressing the influenza A/H5N1 virus hemaglutinin is a promising influenza vaccine candidate that could address most issues raised in association with pandemic influenza vaccine development, including timely delivery, production capacity, efficacy and safety. The promising results obtained in animal models warrant further evaluation and development of recombinant MVA-H5 as a pandemic influenza vaccine candidate. Five-year view Further evaluation of MVA-H5 in animal models and clinical Phase I/II/III trials is needed to confirm its suitability for use in humans. Based on the increasing clinical experience with MVA as a candidate third-generation smallpox vaccine and as an experimental vector vaccine against various other human diseases, it can be anticipated that a recombinant MVA-H5 vaccine will perform as is expected in humans and will be well tolerated. In addition, dose-finding experiments will assess the minimal dose required for the induction of protective immunity. Furthermore, the possibility to induce protective immunity by a single immunization with the MVA-H5 vaccines needs to be explored, which of course would be an ideal vaccination regimen in the face of a pandemic outbreak. In the long run, the MVA technology may be at the basis of a new generation of safe (pandemic) influenza vaccines that are immunogenic without the use of an adjuvant. To this end, the feasibility to produce MVA-based vaccines in a flexible way in CEFs at a large scale needs to be demonstrated. 10.1586/erv.09.4-T0001 Table 1. Advantages and disadvantages of recombinant modified vaccinia virus Ankara vaccines and implication for pandemic influenza vaccine development. Advantage Implication for influenza vaccine production Disadvantage Implication for influenza vaccine production Production Independent of embryonated chicken eggs, in (stockpiled) CEF cells Flexible vaccine production Use of primary/secondary CEF cells Higher risk for adventitious agent contamination Safe, BSL-1 conditions Ease of manufacturing Option to upscale Increase production capacity Efficacy Induction of strong antibody responses Use of adjuvant is not required High dose required Increased costs Induction of cross-reactive antibodies Protection against antigenically distinct variants Induction of T-cell responses Possibility for broadly protective immunity No interference by pre-existing vector immunity Allows for repeated vaccination and induction of antibodies to multiple influenza virus antigens Safety Replication deficiency and avirulence Acceptable safety profile Administration to immunocompromized individuals Vaccination of these high-risk patients possible Multivalent vaccines possible Induction of virus-specific antibodies and T cell Expression of multiple HA genes to induce broad protective antibody responses Stability Record for stability as lyophilized vaccine >4 weeks at 37°C* Stockpiling of vaccines possible *Taken from [61]. CEF: Chicken embryo fibroblast; HA: Hemagglutinin. Key issues • An influenza pandemic caused by influenza A viruses of the H5N1 subtype is imminent. • To limit the impact of a future influenza pandemic, the timely availability of sufficiently efficacious and safe vaccines is highly desirable. • Current influenza vaccine production capacity is limited, although the availability of adjuvants facilitates dose sparing and increase vaccine efficacy. • Modified vaccinia virus Ankara (MVA)-based vaccines have been proven to be safe and efficacious and constitute a promising technology for the development of recombinant vaccines. • A recombinant MVA expressing the hemagglutinin gene of influenza A/H5N1 viruses was immunogenic in mice and macaques and afforded protection against challenge infection with homologous and heterologous highly pathogenic avian influenza A/H5N1 viruses. • The development and use of recombinant MVA-H5 could address most of the issues related to the production of a pandemic influenza vaccine and further clinical evaluation seems warranted. Financial & competing interests disclosure The authors are inventors on a patent application concerning the use of MVA-H5 as an influenza virus vaccine. The authors wish to acknowledge support from The Netherlands Influenza Vaccine Research Center (NIVAREC), ZonMW (grant 91402008), the European Commission (QLK2-CT2002-01034 NOVAFLU, LSHB-CT-2006-037536 MVACTOR) and the Forschungs-Sofortprogramm Influenza (FSI) of the Federal Government of Germany. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript. ==== Refs References Papers of special note have been highlighted as: • of interest 1 Webster RG, Bean WJ. Evolution and ecology of influenza viruses: interspecies transmission. In: Textbook of Influenza. Nicholson KG, Webster RG, Hay AJ (Eds). Blackwell Science, Oxford, UK 109–119 (1998). 2 Potter CW. Chronicle of influenza pandemics. In: Textbook of Influenza. Nicholson KG, Webster RG, Hay AJ (Eds). Blackwell Science, Oxford, UK 3–18 (1998). 3 Claas EC, Osterhaus AD, van Beek R et al. Human influenza A H5N1 virus related to a highly pathogenic avian influenza virus. Lancet 351 (9101), 472–477 (1998).9482438 4 de Jong JC, Claas EC, Osterhaus AD, Webster RG, Lim WL. A pandemic warning? Nature 389 (6651), 554 (1997).9335492 5 Subbarao K, Klimov A, Katz J et al. Characterization of an avian influenza A (H5N1) virus isolated from a child with a fatal respiratory illness. Science 279 (5349), 393–396 (1998).9430591 6 Olsen SJ, Ungchusak K, Sovann L et al. Family clustering of avian influenza A (H5N1). Emerg. Infect. Dis. 11 (11), 1799–1801 (2005).16422010 7 Ungchusak K, Auewarakul P, Dowell SF et al. Probable person-to-person transmission of avian influenza A (H5N1). N. Engl. J. Med. 352 (4), 333–340 (2005).15668219 8 Yang Y, Halloran ME, Sugimoto JD, Longini IM Jr. Detecting human-to-human transmission of avian influenza A (H5N1). Emerg. Infect. Dis. 13 (9), 1348–1353 (2007).18252106 9 Kuiken T, Holmes EC, McCauley J et al. Host species barriers to influenza virus infections. Science 312 (5772), 394–397 (2006).16627737 10 Kreijtz JH, Osterhaus AD, Rimmelzwaan GF. Vaccination strategies and vaccine formulations for epidemic and pandemic influenza control. Hum. Vaccin. (Epub ahead of print) (2009). 11 Mayr A, Munz E. [Changes in the vaccinia virus through continuing passages in chick embryo fibroblast cultures]. Zentralbl Bakteriol [Orig.], 195 (1), 24–35 (1964). 12 Mayr A, Stickl H, Muller HK, Danner K, Singer H. [The smallpox vaccination strain MVA: marker, genetic structure, experience gained with the parenteral vaccination and behavior in organisms with a debilitated defence mechanism]. Zentralbl. Bakteriol. [B] 167 (5–6), 375–390 (1978). 13 Stickl H, Hochstein-Mintzel V. Ìntracuteneous smallpox vaccination with a weak pathogenic vaccinia virus (“MVA virus”). Munch. Med. Wochenschr. 113 , 1149–1153 (1971).5109577 14 Stickl H, Hochstein-Mintzel V, Mayr A et al. [MVA vaccination against smallpox: clinical tests with an attenuated live vaccinia virus strain (MVA)]. Dtsch Med. Wochenschr. 99 (47), 2386–2392 (1974).4426258 15 Frey SE, Newman FK, Kennedy JS et al. Clinical and immunologic responses to multiple doses of IMVAMUNE (modified vaccinia Ankara) followed by Dryvax challenge. Vaccine 25 (51), 8562–8573 (2007).18036708 16 Parrino J, McCurdy LH, Larkin BD et al. Safety, immunogenicity and efficacy of modified vaccinia Ankara (MVA) against Dryvax challenge in vaccinia-naive and vaccinia-immune individuals. Vaccine 25 (8), 1513–1525 (2007).17126963 17 Jones T. IMVAMUNE, an attenuated modified vaccinia Ankara virus vaccine for smallpox infection. Curr. Opin. Mol. Ther. 10 (4), 407–417 (2008).18683106 18 Stittelaar KJ, Kuiken T, de Swart RL et al. Safety of modified vaccinia virus Ankara (MVA) in immune-suppressed macaques. Vaccine 19 (27), 3700–3709 (2001).11395204 • Demonstrates that modified vaccinia virus Ankara (MVA) can be used safely in immunocompromized nonhuman primates. 19 Sutter G, Moss B. Nonreplicating vaccinia vector efficiently expresses recombinant genes. Proc. Natl Acad. Sci. USA 89 (22), 10847–10851 (1992).1438287 • First paper to describe MVA as a vector for the expression of foreign genes of interest. 20 Sutter G, Staib C. Vaccinia vectors as candidate vaccines: the development of modified vaccinia virus Ankara for antigen delivery. Curr. Drug Targets Infect. Disord. 3 (3), 263–271 (2003).14529359 21 Drexler I, Staib C, Sutter G. Modified vaccinia virus Ankara as antigen delivery system: how can we best use its potential? Curr. Opin. Biotechnol. 15 (6), 506–512 (2004).15560976 22 Gomez CE, Najera JL, Krupa M, Esteban M. The poxvirus vectors MVA and NYVAC as gene delivery systems for vaccination against infectious diseases and cancer. Curr. Gene Ther. 8 (2), 97–120 (2008).18393831 23 Sandstrom E, Nilsson C, Hejdeman B et al. Broad immunogenicity of a multigene, multiclade HIV-1 DNA vaccine boosted with heterologous HIV-1 recombinant modified vaccinia virus Ankara. J. Infect. Dis. 198 (10), 1482–1490 (2008).18808335 24 Hawkridge T, Scriba TJ, Gelderbloem S et al. Safety and immunogenicity of a new tuberculosis vaccine, MVA85A, in healthy adults in South Africa. J. Infect. Dis. 198 (4), 544–552 (2008).18582195 25 Jaoko W, Nakwagala FN, Anzala O et al. Safety and immunogenicity of recombinant low-dosage HIV-1 A vaccine candidates vectored by plasmid pTHr DNA or modified vaccinia virus Ankara (MVA) in humans in East Africa. Vaccine 26 (22), 2788–2795 (2008).18440674 26 Harrop R, Drury N, Shingler W et al. Vaccination of colorectal cancer patients with modified vaccinia Ankara encoding the tumor antigen 5T4 (TroVax) given alongside chemotherapy induces potent immune responses. Clin. Cancer Res. 13 (15 Pt 1), 4487–4494 (2007).17671134 27 Acres B, Bonnefoy JY. Clinical development of MVA-based therapeutic cancer vaccines. Expert Rev. Vaccines 7 (7), 889–893 (2008).18767940 28 Ramirez JC, Gherardi MM, Rodriguez D, Esteban M. Attenuated modified vaccinia virus Ankara can be used as an immunizing agent under conditions of preexisting immunity to the vector. J. Virol. 74 (16), 7651–7655 (2000).10906221 29 Harrop R, Drury N, Shingler W et al. Vaccination of colorectal cancer patients with TroVax given alongside chemotherapy (5-fluorouracil, leukovorin and irinotecan) is safe and induces potent immune responses. Cancer Immunol. Immunother. 57 (7), 977–986 (2008).18060404 30 Sharpe S, Polyanskaya N, Dennis M et al. Induction of simian immunodeficiency virus (SIV)-specific CTL in rhesus macaques by vaccination with modified vaccinia virus Ankara expressing SIV transgenes: influence of pre-existing anti-vector immunity. J. Gen. Virol. 82 (Pt 9), 2215–2223 (2001).11514732 31 Yang ZY, Wyatt LS, Kong WP et al. Overcoming immunity to a viral vaccine by DNA priming before vector boosting. J. Virol. 77 (1), 799–803 (2003).12477888 32 Naito T, Kaneko Y, Kozbor D. Oral vaccination with modified vaccinia virus Ankara attached covalently to TMPEG-modified cationic liposomes overcomes pre-existing poxvirus immunity from recombinant vaccinia immunization. J. Gen. Virol. 88 (Pt 1), 61–70 (2007).17170437 33 Mooij P, Balla-Jhagjhoorsingh SS, Koopman G et al. Differential CD4+ versus CD8+ T-cell responses elicited by different poxvirus-based human immunodeficiency virus type 1 vaccine candidates provide comparable efficacies in primates. J. Virol. 82 (6), 2975–2988 (2008).18184713 34 Tatsis N, Lin SW, Harris-McCoy K et al. Multiple immunizations with adenovirus and MVA vectors improve CD8+ T cell functionality and mucosal homing. Virology 367 (1), 156–167 (2007).17590405 35 Ramirez JC, Gherardi MM, Esteban M. Biology of attenuated modified vaccinia virus Ankara recombinant vector in mice: virus fate and activation of B- and T-cell immune responses in comparison with the Western reserve strain and advantages as a vaccine. J. Virol. 74 (2), 923–933 (2000).10623755 36 Albrecht M, Suezer Y, Staib C et al. Vaccination with a modified vaccinia virus Ankara-based vaccine protects mice from allergic sensitization. J. Gene Med. 10 (12), 1324–1333 (2008).18816482 37 Schneider J, Gilbert SC, Blanchard TJ et al. Enhanced immunogenicity for CD8+ T cell induction and complete protective efficacy of malaria DNA vaccination by boosting with modified vaccinia virus Ankara. Nat. Med. 4 (4), 397–402 (1998).9546783 38 El-Gogo S, Staib C, Lasarte JJ, Sutter G, Adler H. Protective vaccination with hepatitis C virus NS3 but not core antigen in a novel mouse challenge model. J. Gene Med. 10 (2), 177–186 (2008).18076128 39 Gasteiger G, Kastenmuller W, Ljapoci R, Sutter G, Drexler I. Cross-priming of cytotoxic T cells dictates antigen requisites for modified vaccinia virus Ankara vector vaccines. J. Virol. 81 (21), 11925–11936 (2007).17699574 40 Peters BS, Jaoko W, Vardas E et al. Studies of a prophylactic HIV-1 vaccine candidate based on modified vaccinia virus Ankara (MVA) with and without DNA priming: effects of dosage and route on safety and immunogenicity. Vaccine 25 (11), 2120–2127 (2007).17250931 41 Sutter G, Wyatt LS, Foley PL, Bennink JR, Moss B. A recombinant vector derived from the host range-restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus. Vaccine 12 (11), 1032–1040 (1994).7975844 • Describes for the first time that with MVA, influenza virus-specific cellular and humoral immune responses can be induced that protect against infection. 42 Hirsch VM, Fuerst TR, Sutter G et al. Patterns of viral replication correlate with outcome in simian immunodeficiency virus (SIV)-infected macaques: effect of prior immunization with a trivalent SIV vaccine in modified vaccinia virus Ankara. J. Virol. 70 (6), 3741–3752 (1996).8648709 43 Paran N, Suezer Y, Lustig S et al. Post-exposure immunization with vaccinia virus MVA or conventional Lister vaccine provides solid protection in murine model for human smallpox. J. Infect. Dis. 199 (1), 39–48 (2009).19012492 44 Earl PL, Americo JL, Wyatt LS et al. Rapid protection in a monkeypox model by a single injection of a replication-deficient vaccinia virus. Proc. Natl Acad. Sci. USA 105 (31), 10889–10894 (2008).18678911 45 Ferrier-Rembert A, Drillien R, Tournier JN, Garin D, Crance JM. Short- and long-term immunogenicity and protection induced by non-replicating smallpox vaccine candidates in mice and comparison with the traditional 1st generation vaccine. Vaccine 26 (14), 1794–1804 (2008).18336966 46 Waibler Z, Anzaghe M, Ludwig H et al. Modified vaccinia virus Ankara induces Toll-like receptor-independent type I interferon responses. J. Virol. 81 (22), 12102–12110 (2007).17855554 47 Samuelsson C, Hausmann J, Lauterbach H et al. Survival of lethal poxvirus infection in mice depends on TLR9, and therapeutic vaccination provides protection. J. Clin. Invest. 118 (5), 1776–1784 (2008).18398511 48 Guerra S, Najera JL, Gonzalez JM et al. Distinct gene expression profiling after infection of immature human monocyte-derived dendritic cells by the attenuated poxvirus vectors MVA and NYVAC. J. Virol. 81 (16), 8707–8721 (2007).17537851 49 Lehmann MH, Kastenmuller W, Kandemir JD et al. Modified vaccinia virus Ankara (MVA) triggers chemotaxisof monocytes and early respiratory immigration of leukocytes by induction of CCL2 expression. J. Virol. 83 (6), 2540–2552 (2009).19129447 50 Schnierle BS, Suezer Y, Sutter G. Recombinant poxvirus vaccines in biomedical research. In: Poxviruses. Mercer A, Schmidt A, Weber O (Eds). Birkhauser-Verl, Basel, Switzerland (2007). 51 Staib C, Lowel M, Erfle V, Sutter G. Improved host range selection for recombinant modified vaccinia virus Ankara. Biotechniques 34 (4), 694–696, 698, 700 (2003).12703290 52 Domi A, Moss B. Engineering of a vaccinia virus bacterial artificial chromosome in Escherichia coli by bacteriophage λ-based recombination. Nat. Methods 2 (2), 95–97 (2005).15782205 53 Cottingham MG, Andersen RF, Spencer AJ et al. Recombination-mediated genetic engineering of a bacterial artificial chromosome clone of modified vaccinia virus Ankara (MVA). PLoS ONE 3 (2), e1638 (2008).18286194 54 Chavan R, Marfatia KA, An IC, Garber DA, Feinberg MB. Expression of CCL20 and granulocyte-macrophage colony-stimulating factor, but not Flt3-L, from modified vaccinia virus ankara enhances antiviral cellular and humoral immune responses. J. Virol. 80 (15), 7676–7687 (2006).16840346 55 Carroll MW, Moss B. Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia virus: propagation and generation of recombinant viruses in a nonhuman mammalian cell line. Virology 238 (2), 198–211 (1997).9400593 56 Drexler I, Heller K, Wahren B, Erfle V, Sutter G. Highly attenuated modified vaccinia virus Ankara replicates in baby hamster kidney cells, a potential host for virus propagation, but not in various human transformed and primary cells. J. Gen. Virol. 79 (Pt 2), 347–352 (1998).9472619 57 Okeke MI, Nilssen O, Traavik T. Modified vaccinia virus Ankara multiplies in rat IEC-6 cells and limited production of mature virions occurs in other mammalian cell lines. J. Gen. Virol. 87 (Pt 1), 21–27 (2006).16361414 58 Bender BS, Rowe CA, Taylor SF et al. Oral immunization with a replication-deficient recombinant vaccinia virus protects mice against influenza. J. Virol. 70 (9), 6418–6424 (1996).8709274 59 Kreijtz JH, Suezer Y, van Amerongen G et al. Recombinant modified vaccinia virus Ankara-based vaccine induces protective immunity in mice against infection with influenza virus H5N1. J. Infect. Dis. 195 (11), 1598–1606 (2007).17471429 • Demonstrates that with MVA protective immunity can be induced against a lethal challenge with homologous and heterologous influenza A/H5N1 viruses. 60 Kreijtz JHCM, Süzer Y, de Mutsert G et al. Recombinant modified vaccinia virus Ankara expressing HA confers protection against homologous and heterologous H5N1 influenza virus infections in macaques. J. Infect. Dis. 199 (3), 405–413 (2009).19061423 • Shows that the results obtained in the mouse model can be reproduced in nonhuman primates confirming that recombinant MVA-H5 is a promising influenza virus H5N1 vaccine candidate. 61 Just I, Finke H. [A contribution to the stabilisation of the MVA-vaccine]. Zentralbl. Bakteriol. [Orig. A], 245 (3), 276–282 (1979). Websites 101 WHO cumulative number of confirmed human cases of avian influenza A/(H5N1) reported to WHO www.who.int/csr/disease/avian_influenza/country/cases_table_2008_09_10/en/index.html 102 WHO options for the use of human H5N1 influenza vaccines and the WHO H5N1 vaccine stockpile www.who.int/csr/resources/publications/who_hse_epr_gip_2008_1d.pdf 103 WHO Antigenic and genetic characteristics of H5N1 viruses and candidate H5N1 vaccine viruses developed for potential use as human vaccines www.who.int/csr/disease/avian_influenza/guidelines/200809_H5VaccineVirusUpdate.pdf 104 Bavarian Nordic www.bavarian-nordic.com	19348560	PMC9709929	NO-CC CODE	2022-12-01 23:23:09	no	Expert Rev Vaccines.; 8(4):4479-454	utf-8	Expert Rev Vaccines	2,014	10.1586/erv.09.4	oa_other
==== Front Expert Rev Vaccines Expert Rev Vaccines Expert Review of Vaccines 1476-0584 1744-8395 Taylor & Francis 18844596 11218494 10.1586/14760584.7.8.1225 Version of Record Research Article Review Smallpox vaccines for biodefense: need and feasibility Smallpox vaccines for biodefense: need & feasibility, Artenstein & Grabenstein Artenstein Andrew W Grabenstein John D Department of Medicine, Director, Center for Biodefense and Emerging Pathogens and Associate Professor of Medicine and Community Health, Alpert Medical School of Brown University and Department of Medicine, Memorial Hospital of RI, 111 Brewster Street, Pawtucket, RI 02860, USA. [email protected] Adult Vaccine Medical Affairs, Merck Vaccines and Infectious Diseases, 770 Sumneytown Pike, WP97-A279, West Point, PA 19486, USA. [email protected] † Author for correspondence 9 1 2014 2008 7 8 12251237 IntegraConverted from TF JATS 1.0 to JATS 1.2 by T&F tfjats-to-jats1.2-converter21 11 2022 © Expert Reviews Ltd 2008 Expert Reviews Ltd Smallpox, eradicated as a cause of natural disease through an intensive global effort in the later part of the 20th Century, has resurfaced as a possible agent of bioterrorism. For this reason, there is renewed interest in smallpox vaccines. Live vaccinia virus, an orthopoxvirus related to smallpox, has a long and successful clinical track record as an effective smallpox vaccine; however, its use is associated with uncommon yet serious adverse events. This has led to a surge of recent research into newer-generation smallpox vaccines with improved safety profiles and retained efficacy. This article will review the history of smallpox vaccines, assess the status of newer-generation vaccines and examine the overall risk-versus-benefit profile of smallpox vaccination. Keywords: bioterrorism myopericarditis smallpox smallpox vaccine vaccine safety vaccinia ==== Body pmcVaccines against infectious diseases, ranked first among the ten greatest public-health achievements of the 20th Century [1], have arguably resulted in greater benefits to the health of mankind than any other cultural, social or scientific advances. Their implementation has eradicated scourges of nature and controlled a host of lethal, communicable diseases, allowing generations of children to survive, unscathed, into adulthood. Perhaps more than any other, the smallpox vaccine provides the most compelling illustration of vaccination’s success. The impact of smallpox on human history is well documented and has been the subject of numerous textbooks, works of literature, objects of art and theses regarding the rise and fall of civilizations [2,3]. The eradication of smallpox and its theoretical resurgence as an agent of bioterrorism illuminate a number of controversial issues engendered by vaccines: safety, public acceptance and risk versus benefit are chief among them. This article will review the genealogy of smallpox vaccines and discuss their potential use in the arena of biodefense. Brief history of smallpox vaccination The history of vaccination, from a scientific standpoint, is traditionally dated from the publication, in 1798, of Edward Jenner’s landmark experiments with cowpox, in which he inoculated a neighbor’s boy with purulent material from a milkmaid’s hand lesion in Berkeley, UK [4]. The boy, 8-year-old James Phipps, was subsequently shown to be protected against a smallpox challenge. In many ways, smallpox represented a natural choice for the earliest explorations into systematic vaccination because of its historical position as the greatest disease scourge of mankind. It was commonly observed, as early as ancient times, that survivors of smallpox were protected against further episodes of the disease. Toward that end, various forms of inoculating healthy individuals with powdered scabs or lesions from infected individuals were used in Africa, China, India and the Ottoman Empire before being introduced into Europe in the early 18th Century [5]. Such procedures were termed ‘variolation’, derived from variola , the Latin word meaning ‘mark on the skin’ and the scientific name for smallpox [5]. Lady Mary Wortley Montagu, the wife of the British Ambassador to Turkey, is credited with the introduction of variolation to England in 1721 [2]. The practice also spread to the New World, where it was adopted to abort smallpox epidemics and used by General George Washington in 1777 to inoculate all susceptible members of the Continental Army, in what became the first large-scale inoculation of a military force [6]. Despite an observed mortality rate of 2–3%, variolation still offered better odds than the 15–30% mortality from naturally acquired smallpox, but because of the risks, alternative practices arose among rural agricultural societies. It was believed, although not necessarily widely known, that milkmaids who developed cowpox, generally a benign disease in humans manifested by pustular lesions on the hands or forearms following contact with infected cow udders, were protected against smallpox and failed to demonstrate cutaneous responses to variolation [2,7]. Jenner became the first to systematically study the hypothesis that cowpox infection protected against subsequent smallpox infection [8]. In An Inquiry into the Causes and Effects of the Variolae Vaccinae, a Disease Discovered in Some of the Western Counties of England, Particularly Gloucestershire and Known by the Name of Cow Pox , Jenner described in detail the vaccination of ten individuals and an additional 17 who resisted variolation after acquiring natural cowpox infection [9]. While Jenner’s work was met with initial skepticism, vaccination against smallpox using his cowpox product in lieu of variolation became widespread in the Western World by the early part of the 19th Century [2,10]. Smallpox was eradicated as a cause of natural human disease after an intensive global campaign in the 1960s and 1970s by the WHO and sponsoring countries using live vaccinia virus, a distinct species of orthopoxvirus that is of an unknown derivation, but is genetically related to Jenner’s vaccine [3]. The last naturally acquired case of smallpox occurred in Somalia in 1977 [3]; the last known human case occurred in 1978 as a result of inadvertent laboratory exposure [3,11]. Despite this, smallpox vaccination continued to be administered selectively into the late 20th Century at which time it had become clear that the risks associated with smallpox vaccine outweighed any perceived benefits [12]. However, in December 2002 after more than a 12-year hiatus, the US Department of Defense (DoD) reinstituted large-scale military vaccination using live vaccinia to mitigate against the perceived threat of bioterrorism involving smallpox. Concurrently, a program of voluntary civilian healthcare worker vaccination was initiated by the US Department of Health and Human Services (DHHS). Since that time, in excess of 1.5 million individuals have been vaccinated in the military program. Approximately 39,000 individuals received the vaccine in the civilian program before it came to an end in late 2003 due to lack of participation. Efficacy of smallpox vaccines The historical premise underlying traditional smallpox vaccines, that of a localized infection with either variola or a cross-reactive orthopoxvirus leading to immune protection, was established long before Jenner published his treatise on vaccination in 1798 [13], and continues to guide the development of newer-generation smallpox vaccines (Table 1). First-generation smallpox vaccines comprising a variety of live vaccinia viruses were used for protection against smallpox yet were hampered by uncommon, potentially life-threatening adverse events that limited their use in the absence of substantial disease risk [14]. This, in concert with concerns regarding the threat of smallpox as a potential agent of bioterrorism has prompted recent efforts toward developing new vaccines with a focus on enhancing safety while maintaining efficacy. First-generation smallpox vaccines possess a proven track record of clinical effectiveness, highlighted by their success in the global eradication campaign of the 1970s [3]. While immune determinants of protection against smallpox remain incompletely understood, the historical record provides ample data concerning a clinical correlate of protection in humans; observations from the use of variola, cowpox and vaccinia viruses document the direct relationship between a vaccine-associated major cutaneous reaction, or ‘take’, and protection against smallpox [3,14,15]. The protection appears to be of long duration and to correlate with the presence of neutralizing antibodies [3]. The cellular arm of the immune response is also known to have a significant role in containing vaccinia [12] and, by extrapolation, variola. Smallpox vaccination induces robust vaccinia-specific cytotoxic T lymphocytes (CTLs) and IFN-γ production by T cells in naive recipients, and these may correlate with neutralizing antibody responses [16]. The original production method of first-generation vaccines involved scarification of calf, sheep or water buffalo skin and viral isolation from skin scrapings containing pus, serum and extruded lymph [3,17]. The resultant liquid suspension of vaccine or ‘wet’ lymph contained viable bacteria, primarily skin commensals, which were minimized by the use of glycerol and, later, phenol in processing [3]. By the 1950s, liquid vaccine lymph preparations had largely been replaced by lyophilized preparations that enhanced preservation of vaccinia virus viability [3]. Vaccine production by animal scarification was abandoned more than 25 years ago and, because smallpox had been eradicated, essentially no first-generation vaccine has been manufactured since then. This led to the view in 2001 that the stockpiled supply was insufficient to cope with a potential large-scale bioterrorist threat. The stockpile consisted of lymph-derived vaccinia, mainly the last production lots of Dryvax®-brand smallpox vaccine, manufactured by Wyeth Laboratories using the New York City Board of Health (NYCBH) strain of vaccinia. Multiple studies have since demonstrated that existing stockpiles can be expanded by diluting the vaccine; lymph-derived, live vaccinia products retain surrogate clinical efficacy at tenfold dilutions in both vaccinia-naive and vaccinia-experienced subjects [17,18]. Second-generation smallpox vaccines (Table 1), in which full-strength vaccinia virus is grown in tissue culture rather than in the skin of large mammals, possess theoretical advantages conferred by this modern manufacturing technique: lowered risk of contamination by adventitious agents [19], viral genetic homogeneity and relative ease of large-scale, consistent production. ACAM1000, a clonal isolate derived from Dryvax and grown in human diploid lung cells (Medical Research Council [MCR]-5), demonstrates similar immunogenicity and cutaneous efficacy at comparable doses to the Dryvax gold standard in animal models, and demonstrates an improved safety profile in preclinical neurovirulence studies in suckling mice and rhesus macaques [20,21]. ACAM2000™, derived from the ACAM1000 master virus by three additional passages in Vero cells [22], has nearly identical biological characteristics to those of its progenitor in animals [23]. Randomized Phase II and III clinical trials, in which nearly 1100 vaccinia-naive subjects were vaccinated with ACAM2000, demonstrated its noninferiority compared with Dryvax at similar vaccinia virus inocula, using cutaneous responses (i.e., takes) as an efficacy end point; ACAM2000 did not meet the noninferiority measure using geometric mean neutralizing antibody titers (GMT) on day 30 after vaccination as another efficacy end point [22,201]. In vaccinia-experienced subjects, ACAM2000 only met the noninferiority threshold for the GMT end point but not for cutaneous responses [201]. Nonetheless, in August 2007, ACAM2000 became the initial second-generation smallpox vaccine to be licensed for human use by the US FDA, leading to the delivery of 192.5 million doses to the US government for stockpiling purposes [202]. The vaccine received the following clinical indication: ‘active immunization against smallpox disease for persons deemed to be at high risk for smallpox infection’ [201]. ACAM2000 is not expected to be commercially distributed in the USA in order to minimize its use and, therefore, its risk [203]. CCSV, another second-generation vaccine grown in MRC-5 cells, compared favorably with Dryvax in a single-center study of 150 vaccinia-naive and 100 vaccinia-experienced subjects [24]. However, this agent was apparently ‘deselected’ by the manufacturer for further advancement. Despite the theoretical advantages conferred by secondgeneration vaccines, they comprise replication-competent, virulent vaccinia viruses and, therefore, possess the potential for a number of uncommon but well-described serious adverse events associated with first-generation smallpox vaccines [14]. Alternative candidates based on attenuated vaccinia strains, third-generation vaccines, may offer more favorable therapeutic ratios. LC16m8, a replication-competent, highly attenuated vaccinia strain, derives from 53 serial passages of a Lister strain isolate in rabbit kidney cells [25]. LC16m8 appears to be less neurovirulent in animals than unattenuated Lister strain vaccinia [26,27]; its use in more than 100,000 Japanese children in the 1970s demonstrated take rates and neutralizing antibody responses similar to those of lymph-derived smallpox vaccines [27,28]. However, the vaccine was never formally field tested, as smallpox was no longer an epidemic threat in Japan at the time. Recently, LC16m8 was shown to engender complete protection in both a rabbit model using intradermal rabbitpox challenge and a mouse model using aerosolized ectromelia (i.e., mousepox) virus [29]. In the mouse model, LC16m8-vaccinated animals developed higher vaccinia-specific neutralizing antibody titers, enhanced neutralization of intracellular mature virus (IMV) and comparable capacity to neutralize extracellular enveloped virus (EEV), compared with Dryvax-vaccinated animals [29]. The latter finding is reassuring in that the B5R gene, required for EEV formation, but deleted during the attenuation process in LC16m8, is a neutralizing antibody target. Additional data suggest that LC16m8 may be a safer alternative to unattenuated vaccine strains in immunocompromised hosts. While comparable protection is noted between LC16m8 and Dryvax in a BALB/c mouse vaccinia challenge model, LC16m8 is nonlethal to severe combined immunodeficiency mice [30,31]. Combined data from trials involving nearly 1700 vaccinia-naive subjects demonstrate 95% take rates and neutralizing antibody seroconversions with LC16m8 [32,33], similar to rates reported with first- and second-generation vaccines in naive individuals [22]. Modified vaccinia Ankara (MVA) strain, a replication-defective, highly attenuated vaccinia virus was initially used as a priming vaccine followed by first-generation smallpox vaccination in more than 120,000 primary vaccinees in Germany in the 1970s [34]. It is attenuated via 570 serial passages in chicken embryo fibroblasts leading to DNA deletions in approximately 15% of its genome, including genes related to host range and immune evasion; thus MVA is generally replication incompetent in mammalian cells [35]. It has been advanced as a third-generation alternative vaccine of potential utility in immunocompromised hosts in whom live vaccinia vaccines are generally contraindicated [36]. Theoretically though, MVA may regain the potential for growth in certain mammalian cell lines owing to reversions at the nucleotide level [35]. Unlike replication-competent vaccinia, MVA does not result in stereotypical neurovirulence upon intracerebral inoculation of suckling mice and may protect against subsequent intracerebral live vaccinia challenge [35]. Additionally, MVA is not associated with detectable viral replication in irradiated mice and rabbits and protects irradiated mice against live vaccinia challenge [35]. Immunosuppressed cynomolgus macaques demonstrate no significant clinical, hematological or pathological abnormalities following inoculation with high-dose MVA by multiple routes, although vaccinial genomes are detectable by PCR from tissues in the majority of macaques [37]. Modified vaccinia Ankara strain is immunogenic and protective in both normal and variably immunosuppressed mice [38,39]. However, animals clearly require multiple and higher doses of MVA to achieve comparable antibody titers to those induced by replication-competent vaccinia [38], and immunosuppressed macaques may fail to develop MVA-specific IgG responses despite high vaccine doses [37]. In comparisons of first-generation vaccinia virus, LC16m8 and MVA, the latter appears to be the least immunogenic, requiring 100-fold more virus to produce similar response levels [30]. Modified vaccinia Ankara strain protects cynomolgus macaques from lethal intravenous [40] or respiratory [41] monkeypox challenges. Such studies confirm data in mice that high-dose MVA or priming with MVA followed by vaccination with first-generation vaccinia virus is necessary to generate immune responses and protection analogous to those observed with replication-competent vaccinia virus alone [40–43]. In some cases MVA-immunized animals, while protected against lethal disease, develop pox lesions following viral challenge; thus, this product may not abrogate the transmission potential of orthopoxviruses. In humans, MVA induces neutralizing antibodies in only 50% of naive subjects receiving a single dose; whereas 80% seroconvert after two doses [44]. The magnitude and duration of humoral immune responses are dose dependent; the proportion of subjects with neutralizing antibodies diminishes by at least half within 3 months following the second dose [44]. Vaccinia-experienced subjects demonstrate more rapid seroconversion or a boosting response and more durable antibody levels after a single dose of MVA [44]. When employed as a priming vaccine in vaccinia-naive subjects, MVA induces a ‘modified-take skin reaction’ with or without a vesicle upon Dryvax challenge 3 months later, similar to cutaneous responses observed in vaccinia-experienced subjects primed with MVA or administered Dryvax alone [45]. Priming with multiple doses of MVA decreases cutaneous viral shedding after Dryvax challenge in naive subjects. Neutralizing antibody titers are comparable among the vaccinated groups; higher vaccinia-specific CD8+ CTLs are noted in those receiving multiple doses of MVA than in those administered one dose of MVA or Dryvax alone [45]. In summary, MVA modifies the cutaneous reactogenicity of live vaccinia without altering its immunogenicity, and multiple MVA priming doses may enhance immune responses to live vaccinia products. Other attenuated, replication-defective vaccine candidates may show promise as priming agents in immunocompromised hosts. NYVAC, derived from the Copenhagen vaccine strain of vaccinia and attenuated by the deletion of 18 nonessential open reading frames [46,47], modulates the effects of Dryvax when used as a priming agent in immunodeficient rhesus macaques [48], yet fails to protect macaques with AIDS against a lethal, intravenous monkeypox challenge [49]. A replication-defective derivative of the Lister strain of vaccinia, bioengineered by deleting the gene encoding for an essential replication cycle enzyme, uracil-DNA-glycosylase [50], has similar preclinical characteristics to MVA, but is theoretically unable to revert to virulence because it only grows in permanent cell lines capable of complementing the enzyme deletion [50,51]. Subunit products are also under investigation as alternative smallpox vaccines. Limited preclinical data support the immunogenicity and protective effect of a vaccinia envelope protein, H3L, in BALB/c mice; passive transfer of H3L-neutralizing antibodies also appears protective [52]. Multiple immunizations with combinations of three outer membrane proteins of IMV (e.g., L1 and A27) and EEV (e.g., A33 and B5) or with combinations of the genes encoding these proteins, are protective in mice and macaque models [53,54]. The latter approach prevents viremia in immunized, challenged monkeys [54]. Animals primed with plasmid DNA encoding the four proteins, then boosted with the analogous proteins, survive lethal monkeypox challenge with significantly milder disease than those immunized with the proteins alone [55]. Safety of smallpox vaccines Substantial volumes of safety data have accumulated on first-generation vaccines through the period of widespread smallpox vaccination, the intensified eradication program and posteradication vaccination exemplified by the recent US military and civilian healthcare worker programs. Surveillance data from the late 1960s in the USA showed serious complications of smallpox vaccination in approximately four per 100,000 individuals with an overall risk of death of one per million primary vaccinations [56–58]. The rate of serious adverse events may be strain related; a retrospective meta-analysis describes a sixfold increased risk of death with the Lister compared with the NYCBH strains [59]. Serious, albeit rare, complications of vaccination are well documented and occur with higher frequency in primary vaccinees or those with immunologic abnormalities (Table 2)[56,57,60]. Postvaccinial encephalitis, a rare disorder of the CNS that generally occurs in children younger than 5 years of age during the second week following vaccination, is associated with a high mortality rate and severe neurological impairment [14,61]. Other serious adverse events are associated with specific predispositions: progressive vaccinia, a frequently fatal complication of smallpox vaccination in immunocompromised hosts, involves regional and metastatic spread of vaccinia virus as a consequence of the inability to contain the localized infection; and eczema vaccinatum, characterized by extension of the local vaccinia infection to other cutaneous areas actively or remotely affected by atopic dermatitis [14,58]. A number of other complications of smallpox vaccination, including generalized vaccinia, congenital vaccinia, inadvertent inoculation and bacterial superinfection [3,58,63,64], are all potential causes of severe morbidity (or mortality in the case of congenital vaccinia) in vaccinees or their close contacts [14,58]. The incidence of serious adverse events expected in modern mass vaccinations using first-generation vaccinia viruses could potentially be significantly higher than historical levels due to a larger population of individuals with vaccine contraindications and a larger proportion of vaccinia-naive individuals in the population [65,66]. That this higher risk did not materialize in contemporary, posteradication programs was probably due to rigorous, risk-based contraindication screening and extensive education. In the setting of a smallpox outbreak, however, fewer exemptions might be granted. Thus, a major focus of newer vaccine approaches is to improve upon safety while maintaining efficacy. Live vaccinia virus vaccines are also associated with a high incidence of local and systemic symptoms. The majority of vaccinia-naive subjects experience local symptoms related to the vaccination site and as many as 40% experience mild-to-moderate constitutional symptoms, such as headache, myalgias, malaise or fever [14]. Data from both the Lister/Elstree [59,67] and the NYCBH strains [14,56,68] of vaccinia virus confirm the higher incidence of local and systemic adverse events in primary vaccinees, compared with revaccinees [36]. While immunogenicity and efficacy in primary vaccinees are apparently not affected by diluting unattenuated vaccinia viruses up to tenfold, fever, systemic symptom score and missed activities are significantly mitigated [69]. The rates of adverse events in the ongoing DoD vaccination program (Table 2) are below historically anticipated levels [70–72] for a number of reasons, including careful screening to exclude those at predictably higher risk, enhanced vaccine education, and a generally healthy population pool. Ten military subjects with undiagnosed HIV infection, all with CD4+ counts above 280 cells/mm3, were inadvertently vaccinated and tolerated the local vaccinia infection without untoward clinical sequelae [73]. In the concurrent DHHS program, seven cases involving the well-described, serious complications of smallpox vaccination were reported: one subject experienced suspected postvaccinial encephalitis; three had confirmed or suspected generalized vaccinia; and three subjects experienced ocular autoinoculation (Table 2)[62,74,75]. The relative dearth of ‘expected’ vaccine complications in these programs is probably multifactorial with more rigorous screening for contraindications than during the era of routine vaccine use, a lower overall denominator of vaccinees than during past routine vaccination, limiting vaccines to adults and possible reporting differences being the main reasons [76]. Cardiac complications of first-generation smallpox vaccines were reported, albeit infrequently, during the era of routine use decades ago. Five cases of myopericarditis were described in association with the NYCBH strain in the USA [75]; data from Finland and Australia involving non-NYCBH vaccinia strains support rates as high as one case per 10,000 vaccinees [77] and 1.6 per million [78], respectively. Up to 3% of Swedish military recruits were found to have nonspecific, asymptomatic T-wave changes on electrocardiogram following smallpox vaccination in the 1960s [79,80]. Nonetheless, a retrospective review of death certificates in New York (NY, USA) during a 4-month period in 1947 in which 6 million people were vaccinated against smallpox using the NYCBH strain failed to show a significant increase in cardiac deaths attributable to vaccination [81]. In the recent, posteradication vaccination programs, two forms of cardiac complications associated with smallpox vaccination were recognized: ischemic events and myopericarditis. The US military identified 24 subjects with ischemic events within 4 weeks of vaccination; the civilian program identified ten [62,74,75,82]. Of these, 19 experienced myocardial infarction, three of whom died. Both the military and civilian rates of ischemic events were within the range expected for an age-matched population, and all occurred in vaccinia-experienced individuals [82,83]. In addition, four cases of dilated cardiomyopathy in the military cohort and three cases in the civilian cohort, all but one in re-vaccinees, were recognized between 1 and 7 months after vaccination [75]. Despite the lack of a clear causal relationship between ischemic cardiac events and smallpox vaccination, the US CDC promulgated new recommendations regarding cardiac prescreening, surveillance and vaccine contraindications for pre-outbreak smallpox vaccination based on the temporal associations [63]. Vaccine deferral on the basis of known heart disease or multiple cardiac risk factors was not associated with a clear reduction in ischemic cardiac events [83]. The DoD identified 140 cases of myopericarditis during its first 2 years of the program, largely in male, Caucasian, primary vaccinees [75,84], representing a rate of approximately 1.2 per 10,000 – similar to the historical rates in Finnish conscripts [77]. The rate in the civilian DHHS vaccination program in which 21 cases were identified was similar if only probable cases were considered, but was approximately 5.5 per 10,000 [74] if both suspected and probable cases were included. Both rates were higher than age-matched, unvaccinated individuals and since cases cluster in the second week after vaccination, the appropriate conclusion is that primary smallpox vaccination of adults using first-generation vaccinia is associated with a hitherto unrecognized, increased risk of myopericarditis. Second-generation vaccines, ACAM2000 [22] and CCSV [24], show no significant differences in local or systemic adverse events compared with Dryvax. While none of the rare but well described, serious adverse events related to smallpox vaccines have been noted with these newer vaccines to date, small sample sizes preclude a relative risk determination. Seven out of 2983 (0.2%) vaccinia-naive subjects who received ACAM2000 and three out of 868 (0.3%) who received Dryvax during recent Phase II and III trials were identified as cases of suspected vaccine-induced myopericarditis [17,22,201]. These rates extrapolate to approximately fivefold higher than those noted in the DoD and DHHS efforts, possibly as a result of rigorous, active surveillance for cardiac complications informed by the findings of these posteradication vaccination programs, although the distinction between suspected and confirmed cases needs to be taken into account [22,63]. Although no statistically significant differences were observed in the rates of myopericarditis between those who received ACAM2000 versus Dryvax, the Phase III trials of ACAM2000 were prematurely terminated on this basis. Since myopericarditis cases have occurred in subjects who had received first- or second-generation vaccines, this complication appears to be directly or indirectly related to vaccinia virus and unlikely to be related to an adventitious agent introduced in the processing of lymph. The higher incidence of myopericarditis observed in both treatment groups in the ACAM2000 studies, compared with the government-sponsored vaccination programs, probably results from active surveillance using routine assessments of cardiac symptoms, cardiac enzymes and electrocardiograms designed to identify asymptomatic individuals or cases involving only mild or transient symptoms. The prototypical third-generation vaccines, LC16m8 and MVA, lack large-scale human safety evaluations. LC16m8 was noted to be well tolerated in both an open-label study involving 476 primary vaccinees and 552 revaccinees [32] and in comparison with Dryvax in 153 vaccinia-naive volunteers; neither vaccinia-associated serious adverse events nor cardiovascular complications were noted, although planned cardiac evaluations were not performed [33]. In an open-label study, one primary vaccinee developed acute sensorineural deafness and one reported chest pain ascribed by the authors to musculoskeletal causes, with no further information provided [32]. MVA appears to be associated with dose-related, local reactions in the majority of recipients; these self-limited events have not led to discontinuation of subjects from Phase I studies [44]. In a small study of vaccinia-naive individuals with either a history of atopic dermatitis or with active atopic dermatitis, groups in which first-generation vaccinia vaccines are traditionally contraindicated, all subjects receiving MVA reported mild-to-moderate local reactogenicity but no serious adverse reactions [85]. MVA-primed subjects exhibit decreased reactogenicity and minimal systemic symptoms following Dryvax challenge compared with placebo-primed subjects, supporting a modulating effect of MVA in the context of safety, similar to that seen in efficacy studies. No vaccinia-associated serious adverse events or cardiac complications have been observed with MVA to date, although cardiac evaluations are uniformly lacking [45]. Feasibility & acceptability of smallpox vaccination As smallpox is no longer a cause of naturally occurring disease in humans and there is no known animal reservoir for this pathogen in nature, any human case of smallpox occurring outside of a known laboratory exposure would be tantamount to bioterrorism [86]. Thus, any discussion of smallpox mitigation strategies, specifically pre-event or postevent vaccination, hinges on the concept of ‘risk’. Risk refers to the likelihood that exposure to a hazard will lead to a negative consequence [87]. In the context of a smallpox threat, it is essential to consider both the probabilities of exposure and the potential range of consequences associated with the disease and with its vaccines in order to attempt to objectively gauge risk of this type. The probability of exposure to smallpox outside of a laboratory setting is believed to be low but not zero. It has been suggested that unreported smallpox stocks may have existed in the former Soviet Union; if this is accurate, the whereabouts of such viruses would not be known [88]. Since the exposure variable is dependent on the unpredictable tactics of terrorists, accurate, quantifiable risk appraisal is not possible [89]. However, the potential consequences of a bioterrorist attack using smallpox would be devastating. Multiple characteristics of smallpox ensure that its deliberate reintroduction into the human population would be a global health catastrophe of profound dimensions. Smallpox is stable in aerosol form, raising the possibility of a large-scale attack; it has a low infective dose, requiring minimal viral inocula to cause productive infection in humans [3]; and case–fatality rates historically approached 30%. Morbidity from smallpox was substantial and included prolonged duration of illness, scarring of survivors, secondary soft-tissue infections and blindness [3]. Secondary attack rates among unvaccinated close contacts ranged from 37–88%, although these data derive from historical studies in developing countries and may not be analogous to current circumstances [3,86,90]. Additionally, much of the world’s population is susceptible to smallpox due to the cessation of routine vaccination in the early 1970s and the absence of low-level, boosting exposures that would be expected if variola circulated naturally in the environment. Finally, other orthopoxviruses, such as monkeypox, may be pathogenic for humans in either naturally occurring outbreaks or bioterrorism scenarios and may be partially ameliorated by vaccinia immunity [88,91,92]. While improvements in medical care and infection-control procedures, and advances in health technology may mitigate some of the expected morbidity and mortality from smallpox in the 21st Century, they represent a double-edged sword. These same advancements have increased the prevalence of immunocompromised hosts, a population at higher risk of serious morbidity and mortality from smallpox. Similarly, the prevalence of atopic dermatitis in the population has increased markedly since the discontinuation of routine smallpox vaccination; up to 10% of adults and 30% of children in industrialized countries are now diagnosed with this disorder [93]. Live-virus smallpox vaccines are traditionally contraindicated in this population as well. Furthermore, mass casualties due to a smallpox outbreak could rapidly overwhelm healthcare resources. It has been variably estimated that the number of deaths in the USA after implementation of mass vaccination, using first-generation smallpox vaccines presumably in response to a realized threat, would conservatively range between 125 and 500, accompanied by thousands of serious adverse events [65,66]. In a postevent setting, where the actual smallpox ‘event’ was realized anywhere throughout the world, the risks associated with the disease would probably far outweigh the potential risks associated with vaccination; thus the benefits of vaccination would favor its deployment, although the relative merits of various strategies, ranging from mass vaccination to a more targeted, ring vaccination approach, are debatable [66,94]. In an outbreak scenario, some combination of ring and mass vaccination would probably be implemented. Dilution studies that have expanded the existing supply of first-generation vaccines in concert with the licensure of a second-generation vaccine, ACAM2000 [202], have resulted in stockpiles of clinically effective vaccines sufficient to vaccinate the entire US population, serving as a fail-safe posture in the event of a biologic attack using smallpox. In addition to the US stockpile, smallpox vaccine stockpiles are also being developed by other nations [95]. Recent experiences with posteradication vaccination and previous mass smallpox vaccination efforts [17,81] have demonstrated the feasibility of this approach. Japan has limited stockpiles of the attenuated LC16m8 vaccine, although more data would be needed to assure its efficacy and safety in individuals with vaccinia contraindications [95]. Many other nations have developed stockpiles of first- and, in some cases, second-generation vaccines; it is estimated that current capabilities would be sufficient to vaccinate approximately 10% of the world’s population [95]. To assure vaccine availability to poorer nations and provide for a nimble response by the international public-health community, the WHO has recently implemented a plan to develop a strategic smallpox vaccine stockpile of at least 200 million doses, largely derived from pledged donations from member countries and reminiscent of the WHO’s efforts during the global smallpox eradication program of the 1970s [204]. By contrast, the concept of pre-event vaccination presents a more problematic analysis. Despite the relative dearth of serious adverse events in both the recent DoD and DHHS smallpox vaccination programs [62,75], there was a low but meaningful incidence of complications related to first-generation vaccines. While the risks can be mitigated via careful screening and exclusion of those in selected higher risk categories, they cannot be completely abrogated. For instance, it has been demonstrated that more than a third of subjects with atopic dermatitis or other vaccine contraindications were unrecognized using various screening strategies [96,97]. In a setting of a very low perceived risk of smallpox, are any levels of significant vaccine-related risks acceptable? Data from the civilian healthcare worker vaccination program of 2003 address this issue [62]. Multiple, detailed evaluations of the DHHS program have been reported elsewhere [62,74,82,98,99]. While there appears to be general agreement that many aspects of the program were instructive from an operational public health standpoint, it remains unclear whether the program achieved the stated goal of enhancing national biodefense preparedness [100]. Certainly the number of civilians actually vaccinated fell far short, approximately 8%, of the 500,000 target set at the program’s inception [62]. However, this in and of itself does not necessarily constitute failure, to the extent that the program served as a pilot study to explore the feasibility, acquire experience and reveal hitherto unrecognized issues. Perhaps the most instructive aspects of the posteradication DHHS smallpox vaccination program, however, relate to the acceptability of vaccination among the targeted civilian groups, largely healthcare workers and others potentially involved in the initial response to a bioterrorist event. Revelations from this experience should inform future vaccination programs in the arena of biodefense. One contemporary study that modeled various smallpox attack scenarios demonstrates that the risk associated with pre-event vaccination of healthcare workers generally outweighs the potential health benefits when the probability of a smallpox attack is less than 22%; in order for mass pre-event vaccination of the public to be beneficial, the probability of an attack would have to be significantly higher, above 47% [101]. In large part, healthcare workers and traditional first responders who declined voluntary smallpox vaccination determined that their personal risk associated with vaccination using first-generation smallpox vaccine outweighed the perceived risk of smallpox [102,103]. A number of additional factors contributed to the risk equation that ultimately limited the acceptability of pre-event smallpox vaccination in the 2003 setting: uncertainties regarding liability for vaccine-induced injury; sources of compensation and mechanisms of remedy for illness or injury related to vaccination; inadequate education concerning the risks and potential benefits of the program; the recognition of novel cardiac adverse events; and the lack of biological weapon caches in Iraq [104,105]. Individuals will generally act according to their personal perceptions of risk, but since it is inherently impossible to quantify the probability that a terrorist will release a biological weapon, the perceived risks associated with smallpox vaccination apparently dominated the equation in 2003. Expert commentary First-generation smallpox vaccines have a long, distinguished track record of effectiveness in the control and subsequent eradication of naturally occurring smallpox. However, their utility in the posteradication setting is limited by uncommon but serious adverse effects (Table 2). The incidence of some of the more notorious of these complications can be minimized by rigorous screening for known contraindications and site hygiene; others, such as myopericarditis, have not yet had clear precipitating factors identified. A significant proportion of the population would be excluded from receiving these vaccines in nonemergent scenarios. New-generation smallpox vaccines, specifically second- (tissue culture-derived vaccinia) and third-generation (highly attenuated vaccinia) vaccines potentially have a similar efficacy to first-generation smallpox vaccines. Second-generation vaccines, as with first-generation ones, are associated with a significant risk of myopericarditis that substantially limits their utility in a pre-event setting. With the licensure of ACAM2000 and its substitution as the principal vaccine in the ongoing DoD program, the FDA has imposed a risk-minimization action plan that includes a myopericarditis case registry and Phase IV cohort study of military vaccinees to further characterize cardiac adverse events [203]. Third-generation products may possess improved safety profiles, but this has yet to be proven in adequately powered studies or experience with large numbers of vaccinees. Highly attenuated, replication-defective vaccinia MVA sacrifices degrees of immunogenicity and efficacy for its theoretically improved safety profile. For some third-generation products, multidose regimens limit their utility in outbreak settings. The risk versus benefit profile of smallpox vaccination is complex (Table 3). The risks associated with currently licensed vaccines probably do not justify their pre-event use in groups with a very low perceived risk of smallpox exposure. However, the latter type of risk is dependent on the unpredictable nature of terrorists and may be stratified among different groups; for example, deployed military forces may be at higher levels of exposure risk. Additionally, the general level of perceived risk may increase abruptly should a terrorist event occur. Such an unpredictable situation argues for continued research on safer smallpox vaccines. New-generation vaccines that are demonstrated to have significantly improved safety profiles after adequate human studies may alter the risk-versus-benefit assessment. Five-year view Current stockpiles of first- and second-generation smallpox vaccines serve as an important contingency position for emergent circumstances. Newer-generation smallpox vaccines that employ highly attenuated and/or nonreplicative forms of vaccinia or subunit vaccine approaches, some with promising preclinical data, may provide significantly safer, effective alternatives over the next 5 years that will enhance biodefense strategies. Viral subunit strategies, in particular, may provide a flexible platform in the future upon which to build capabilities for protection against genetically altered forms of smallpox. 10.1586/14760584.7.8.1225-T0001 Table 1. Smallpox vaccines and vaccine candidates (2008). Platform Product Parent strain Rationale for its use First-generation Lymph-derived vaccinia virus Dryvax® (Wyeth) NYCBH Historical experience in the USA through the era of routine use Sanofi Pasteur smallpox vaccine (SPSV) NYCBH Produced in 1956–1957 and used in the USA program of that era; in frozen storage since Elstree-RIVM (master seed stock held at the National Institute of Public Health in The Netherlands [RIVM]) Lister Historical experience in the Intensified Smallpox Eradication Programme Second-generation Replication-competent tissue-cultured vaccinia virus ACAM2000™ (Acambis): cloned virus grown in Vero cells NYCBH Defined manufacturing process; reduced theoretical risk of adventitious agents compared with lymph-derived vaccine; less neurovirulent in animal models Elstree-BN (Bavarian-Nordic) Lister Defined manufacturing process; reduced theoretical risk of adventitious agents compared with lymph-derived vaccine Third-generation Replication-competent, highly attenuated vaccinia virus LC16m8 vaccine: derived from 53 serial passages in rabbit kidney cells; temperature sensitive, small-plaque phenotype due to mutation in the B5R gene Lister Experience in more than 100,000 Japanese children between 1973 and 1975; better safety profile than traditional live vaccinia, less neurovirulent in animals but unproven clinical efficacy Replication-deficient, highly attenuated vaccinia virus MVA: derived from more than 570 serial passages in chicken embryo fibroblasts: IMVAMUNE (Bavarian-Nordic); TBC-MVA (Therion) Ankara Theoretically improved safety profile, especially for those in whom live vaccinia is contraindicated. Used in 120,000 primary vaccinees in Germany in 1970s but unproven clinical efficacy NYVAC (Sanofi-Pasteur): attenuated by the deletion of 18 open-reading frames from a plaque-cloned vaccinia isolate Copenhagen Theoretically improved safety profile, especially for those in whom live vaccinia is contraindicated dVV-L: derived from deletion of UDG gene needed for viral replication Lister Theoretically improved safety profile and can be manufactured in cell line that complements UDG deficiency, thus increased capacity for rapid production Subunit vaccines Recombinant proteins; plasmid DNA Vaccinia viruses, different sources Theoretically improved safety profile MVA: Modified vaccinia Ankara; NYCBH: New York City Board of Health; UDG: Uracil DNA glycosylase. 10.1586/14760584.7.8.1225-T0002 Table 2. Noteworthy adverse events after smallpox vaccination, USA, December 2002–June 2004. Event type Events and rates among 628,414 DoD vaccinees* Events and rates among 39,566 DHHS vaccinees Historical rate per million vaccinees Events (n) Rate per million DoD vaccinees Events (n) Rate per million DHHS vaccinees Moderate or serious Postvaccinial encephalitis 1 1.6 1 26 2.6–8.7‡ Acute myopericarditis 83§ 132 21§ 531 100 Eczema vaccinatum 0 0 0 0 2–35‡ Progressive vaccinia 0 0 0 0 1–7‡ Mild or temporary Generalized vaccinia, mild 40 64 3 77 45–212‡ Erythema multiforme major 1 1.6 0 0 NA Inadvertent inoculation, self 73¶ 116 24¶ 607 606‡ Vaccinia transfer to contact 47 75 0 0 8–27‡ *Primarily composed of uniformed military personnel plus some DoD civilian employees; a minority of this total was healthcare workers. ‡Based on adolescent and adult smallpox vaccination from 1968 studies (both primary and revaccination). §DoD events include four biopsy-confirmed, 73 probable and six suspected cases; DHHS events include none confirmed, five probable and 16 suspected cases. ¶DoD events include 59 inadvertent inoculations of the skin and 14 of the eye; DHHS events include 21 inadvertent inoculations of the skin and three of the eye. DoD: Department of Defense; DHHS: Department of Health and Human Services; NA: Not available. Data from [56,57,60–62,75,106]. 10.1586/14760584.7.8.1225-T0003 Table 3. Risk-versus-benefit considerations for pre-event smallpox vaccination. Risks/concerns Benefit/mitigating factors Unresolved issues Ref. • Probability of exposure to smallpox is difficult to reliably quantify, but is greater than zero – Weaponized viruses may be available – Growth and expansion of terrorist activities • Currently available first- and second-generation vaccines are efficacious • Variola virus is difficult to manipulate in the laboratory • No known animal or natural reservoir of infection • Second-generation vaccines demonstrate surrogate efficacy in humans; in the absence of a smallpox outbreak, their clinical effectiveness cannot be ascertained • Exposure to smallpox depends on the unpredictable acts of terrorists [14,17,89] • Reintroduction of smallpox into the human population would be potentially catastrophic, especially in settings with inadequate public health infrastructure – Variola virus can be aerosolized – Air-borne transmission has been documented – Mortality can exceed 30% – High secondary attack rates in some settings (e.g., healthcare environments) – Most people are either immunologically naive or were vaccinated decades ago – High transmission risk in healthcare settings – Genetic or molecular technology could be used to enhance virulence • Effective vaccines are available • Postexposure vaccination, within 4 days, is protective • Aerosol transmission is not the predominant route • Improvements in supportive medical care may mitigate excess mortality • Remote vaccination may afford at least partial protection • Possible role for antivirals that were not available when smallpox occurred naturally • Advances in medical care and health technology have also served to increase the prevalence of immunocompromised hosts and other subgroups in whom currently licensed smallpox vaccines are contraindicated • Medical resources may be overwhelmed by a large number of cases • Durability of vaccine-induced immunity is ill-defined in clinical settings [3,89,107] • Smallpox vaccines are associated with significant safety issues (Table 2) • Public acceptance will depend on scenario • Post-eradication programs show that careful screening can minimize some serious adverse effects but may not avert myopericarditis • Newer-generation vaccines may be associated with improved safety profiles, but currently there are insufficient data to demonstrate this Key issues • First-generation smallpox vaccines, comprising live vaccinia virus grown largely in the skin of calves, have a well-documented track record of effectiveness in preventing smallpox but are associated with uncommon, serious adverse events that may limit their use. • Newer-generation smallpox vaccines that employ either vaccinia grown in tissue culture (second generation) or highly attenuated vaccinia viruses (third generation) may retain efficacy; although both first- and second-generation vaccines are associated with a significant risk of myopericarditis that limits their acceptability in pre-event settings. Third-generation vaccines may improve upon the safety profile of other smallpox vaccines, although there are insufficient data to determine this conclusively. • The feasibility of deploying smallpox vaccines is dependent on a risk-versus-benefit assessment, in which the probability of exposure to smallpox through bioterrorism must be weighed against the risks and potential benefits of smallpox vaccines. The unpredictable nature of terrorism may compel this evaluation to be more qualitative than quantitative. • Because there remains a potential risk of smallpox exposure that depends on specific scenarios, newer-generation smallpox vaccines that retain clinical efficacy and improve upon safety are needed. Acknowledgements The authors wish to acknowledge Katherine Bollesen and Margo Katz for administrative assistance with the manuscript. Financial & competing interests disclosure AW Artenstein has served as a consultant to Acambis, Inc., but has no ongoing financial relationship with the company. JD Grabenstein is an employee of Merck & Co., Inc., which does not manufacture vaccinia or any related vaccine. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript. ==== Refs References Papers of special note have been highlighted as: • of interest •• of considerable interest 1 Center for Disease Control and Prevention. Achievements in public health, 1900–1999 impact of vaccines universally recommended for children – United States, 1990–1998. MMWR Morb. Mortal. Wkly Rep. 48 (12), 243–248 (1999).10220251 2 Barquet N, Domingo P. Smallpox: the triumph over the most terrible of the ministers of death. Ann. Intern. Med. 127 (8), 635–642 (1997).9341063 3 Fenner F, Henderson DA, Arita I, Zdenêk J, Ladnyi ID. Smallpox and its Eradication. World Health Organization, Geneva, Switzerland (1988). •• Definitive textbook of smallpox written by a group of authors who were among the leaders in the global eradication campaign of the 1960s and 1970s. 4 Stern A, Markel H. The history of vaccines and immunization: familiar patterns, new challenges. Health Aff. 24 (3), 611–621 (2005). 5 Parish HJ. A History of Immunization. E & S Livingstone Ltd, Edinburgh, UK (1965). 6 Artenstein AW, Opal JM, Opal SM et al. History of US military contributions to the study of vaccines against infectious diseases. Mil. Med. 170 (4), 3–11 (2005).15916278 7 Hammarsten JF, Tattersall W, Hammarsten JE. Who discovered smallpox vaccination? Edward Jenner or Benjamin Jesty? Trans. Am. Clin. Climatol. Assoc. 90 , 44–55 (1979).390826 8 Riedel S. Edward Jenner and the history of smallpox and vaccination. Proc. (Bayl. Univ. Med. Cen.) 18 , 21–25 (2005). 9 Baxby D. Edward Jenner’s inquiry; a bicentenary analysis. Vaccine 17 , 301–307 (1999).9987167 10 Grabenstein JD, Pittman PR, Greenwood JT, Engler RJM. Immunization to protect the US armed forces: heritage, current practice, and prospects. Epidemiol. Rev. 28 , 3–26 (2006).16763072 11 Wade N. New smallpox case seems lab-caused. Science 201 (4359), 893 (1978).684414 12 Redfield RR, Wright DC, James WD et al. Disseminated vaccinia in a military recruit with human immunodeficiency virus (HIV) disease. N. Engl. J. Med. 316 , 673–676 (1987).3821799 13 Moloo H, Artenstein AW. History of vaccines. In: Encyclopedia of Public Health. Murray CJ (Ed.). Elsevier (2008). 14 Henderson DA, Inglesby TV, Bartlett JG et al. Smallpox as a biological weapon. JAMA 281 , 2127–2137 (1999).10367824 • Excellent review of the clinical issues related to smallpox as a threat agent of bioterrorism. 15 Breman JG, Henderson DA. Diagnosis and management of smallpox. N. Engl. J. Med. 346 , 1300–1308 (2002).11923491 16 Ennis FA, Cruz J, Demkowicz Jr WE et al. Primary induction of human CD8 cytotoxic T lymphocytes and interferon-γ- producing T cells after smallpox vaccination. J. Infect. Dis. 185 , 1657–1659 (2002).12023773 17 Artenstein AW. New generation smallpox vaccines: a review of preclinical and clinical data. Rev. Med. Virol. 18 (4), 217–231 (2008).18283712 • Represents a thorough review of recent animal and human data regarding newer smallpox vaccines. 18 Frey SE, Newman FK, Yan L et al. Response to smallpox vaccine in persons immunized in the distant past. JAMA 289 , 3295–3299 (2003).12824212 19 Murphy FA, Osburn BI. Adventitious agents and smallpox vaccine in strategic national stockpile. Emerg. Infect. Dis. 11 (7), 1086–1089 (2005).16022785 20 Weltzin R, Liu J, Pugachev KV et al. Clonal vaccinia virus grown in cell culture as a new smallpox vaccine. Nat. Med. 9 , 1125–1130 (2003).12925845 21 Monath TP, Caldwell JR, Mundt W et al. ACAM2000 clonal Vero cell culture vaccinia virus (New York City Board of Health strain) – a second-generation smallpox vaccine for biological defense. Int. J. Infect. Dis. 8 (S2), S31–S44 (2004).15491873 22 Artenstein AW, Johnson C, Marbury TC et al. A novel, cell culture-derived smallpox vaccine in vaccinia-naive adults. Vaccine 23 , 3301–3309 (2005).15837236 23 Acambis, Inc. Investigator’s brochure for ACAM2000 smallpox vaccine. Acambis, Inc, MA, USA (2002). 24 Greenberg RN, Kennedy JS, Clanton DJ et al. Safety and immunogenicity of new cell-cultured smallpox vaccine compared with calf-lymph derived vaccine: a blind, single-centre, randomized controlled trial. Lancet 365 , 389–409 (2005). 25 Hashizume S; Chiba Serum Institute. Special edition future of vaccination: everything about attenuated vaccines. Basics of new attenuated vaccine strain LC16m8. Clin. Virus 3 , 229–235 (1975). 26 Hashizume S, Yoshizawa N, Morita M Suzuki K. Properties of attenuated mutant of vaccinia virus, LC16m8, derived from Lister strain. In: Vaccinia Viruses as Vectors for Vaccine Antigens. Quinnan GV (Ed.). Elsevier, NY, USA (1985). 27 Kenner J, Cameron F, Empig C et al. LC16m8: an attenuated smallpox vaccine. Vaccine 24 , 7009–7022 (2006).17052815 28 Wiser I, Balicer RD, Cohen D. An update on smallpox vaccine candidates and their role in bioterrorism related vaccination strategies. Vaccine 25 , 976–984 (2007).17074424 29 Empig C, Kenner JR, Perret-Gentil M et al. Highly attenuated smallpox vaccine protects rabbits and mice against pathogenic orthopoxvirus challenge. Vaccine 24 , 3686–3694 (2006).16430997 30 Kidokoro M, Tashiro M, Shida H. Genetically stable and fully effective smallpox vaccine strain constructed from highly attenuated vaccinia LC16m8. Proc. Natl Acad. Sci. USA 102 , 1452–1457 (2005). 31 Yokote H, Shinmura Y, Satou A et al. Safety and efficacy study of attenuated smallpox vaccine LC16m8 in animals. Presented at: 2006 ASM Biodefense Research Meeting for the American Society for Microbiology. Washington, DC, USA, 15–18 February 2006. 32 Fujii T, Kannatani Y, Matsumura T et al. Clinical evaluation of attenuated smallpox vaccine LC16m8 in Japan. Presented at: 2006 ASM Biodefense Research Meeting for the American Society for Microbiology. Washington, DC, USA, 15–18 February 2006. 33 Kenner JR, Gurwith M, Luck A et al. Safety and immunogenicity of attenuated smallpox vaccine, LC16m8: results from a Phase 2 study. Presented at: 2006 ASM Biodefense Research Meeting for the American Society for Microbiology. Washington, DC, USA, 15–18 February 2006. 34 Mayr A, Stickl H, Müller HK et al. The smallpox vaccination strain MVA: marker, genetic structure, experience gained with the parenteral vaccination and behavior in organisms with a debilitated defense mechanism (author’s transl). Zentralbl. Bakteriol. [B]. 167 , 375–390 (1978). 35 McCurdy LH, Larkin BD, Martin JE et al. Modified vaccinia Ankara: potential as an alternative smallpox vaccine. Clin. Infect. Dis. 38 , 1749–1753 (2004).15227622 36 Parrino J, Graham BS. Smallpox vaccines: past, present, and future. J. Allergy Clin. Immunol. 118 , 1320–1326 (2006).17157663 37 Stittelaar KJ, Kuiken T, de Swart RL et al. Safety of modified vaccinia virus Ankara (MVA) in immune-suppressed macaques. Vaccine 19 , 3700–3709 (2001).11395204 38 Wyatt LS, Earl PL, Eller LA et al. Highly attenuated smallpox vaccine protects mice with and without immune deficiencies against pathogenic vaccinia virus challenge. Proc. Natl Acad. Sci. USA 101 , 4590–4595 (2004).15070762 39 Meseda CA, Garcia AD, Jumar A et al. Enhanced immunogenicity and protective effect conferred by vaccination with combinations of modified vaccinia virus Ankara and licensed smallpox vaccine Dryvax in a mouse model. Virology 339 , 164–175 (2005).15993917 40 Earl PL, Americo JL, Wyatt LS et al. Immunogenicity of a highly attenuated MVA smallpox vaccine and protection against monkeypox. Nature 428 , 182–185 (2004).15014500 41 Stettelaar KJ, van Amerongen G, Kondova I et al. Modified vaccinia virus Ankara protects macaques against respiratory challenge with monkeypox virus. J. Virol. 79 , 7845–7851 (2005).15919938 42 McCurdy LH, Rutigliano JA, Johnson TR et al. Modified vaccinia virus Ankara immunization protects against lethal challenge with recombinant vaccinia virus expressing murine interleukin-4. J. Virol. 78 , 12471–12479 (2004).15507634 43 Abdalrhman I, Gurt I, Katz E. Protection induced in mice against a lethal orthopox virus by the Lister strain of vaccinia virus and modified vaccinia virus Ankara (MVA). Vaccine 24 , 4152–4160 (2006).16603280 44 Vollmar J, Arndtz N, Eckl KM et al. Safety and immunogenicity of IMVAMUNE, a promising candidate as a third generation smallpox vaccine. Vaccine 24 , 2065–2070 (2006).16337719 45 Parrino J, McCurdy LH, Larkin BD et al. Safety, immunogenicity and efficacy of modified vaccinia Ankara (MVA) against Dryvax® challenge in vaccinia-naive and vaccinia-immune individuals. Vaccine 25 , 1513–1525 (2007).17126963 46 Tartaglia J, Perkus ME, Taylor J et al. NYVAC: a highly attenuated strain of vaccinia virus. Virology 188 (1), 217–232, (1992).1566575 47 Nájera JL, Gómez CE, Domingo-Gil E et al. Cellular and biochemical differences between two attenuated poxvirus vaccine candidates (MVA and NYVAC) and role of the C7L gene. J. Virol. 80 , 6033–6047 (2006).16731942 48 Edghill-Smith Y, Venzon D, Karpova T et al. Modeling a safer smallpox vaccination regimen, for human immunodeficiency virus type 1-infected patients, in immunocompromised macaques. J. Infect. Dis. 188 , 1181–1191 (2003).14551889 49 Edghill-Smith Y, Bray M, Whitehouse CA et al. Smallpox vaccine does not protect macaques with AIDS from a lethal monkeypox virus challenge. J. Infect. Dis. 191 , 372–381 (2005).15633096 50 Ober BT, Brűhl P, Schmidt M et al. Immunogenicity and safety of defective vaccinia virus Lister: comparison with modified vaccinia virus Ankara. J. Virol. 76 , 7713–7723 (2002).12097585 51 Coulibaly S, Brűhl P, Mayrhofer J et al. The nonreplicating smallpox candidate vaccines defective vaccinia Lister (dVV-L) and modified vaccinia Ankara (MVA) elicit robust long-term protection. Virology 341 , 91–101 (2005).16061267 52 Davies DH, McCausland MM, Valdez C et al. Vaccinia virus H3L envelope protein is a major target of neutralizing antibodies in humans and elicits protection against lethal challenge in mice. J. Virol. 79 , 11724–11733 (2005).16140750 53 Fogg C, Lustig S, Whitbeck C et al. Protective immunity to vaccinia virus induced by vaccination with multiple recombinant outer membrane proteins of intracellular and extracellular virions. J. Virol. 78 , 10230–10237 (2004).15367588 54 Hooper JW, Thompson E, Wilhelmsen C et al. Smallpox DNA vaccine protects nonhuman primates against lethal monkeypox. J. Virol. 78 , 1133–1143 (2004). 55 Pulford DJ, Gates A, Bridge SH et al. Differential efficacy of vaccinia virus envelope proteins administered by DNA immunization in protection of BALB/c mice from a lethal intranasal poxvirus challenge. Vaccine 22 , 3358–3366 (2004).15308360 56 Lane JM, Ruben RL, Neff JM et al. Complications of smallpox vaccination, 1968: national surveillance in the United States. N. Engl. J. Med. 281 , 1201–1208 (1969).4186802 57 Lane JM, Ruben FL, Neff JM et al. Complications of smallpox vaccination, 1968: results of ten statewide surveys. J. Infect. Dis. 122 , 303–309 (1970).4396189 58 Fulginiti VA, Papier A, Lane JM et al. Smallpox vaccination: a review, Part II. Adverse events. Clin. Infect. Dis. 37 , 251–271 (2003).12856218 59 Kretzschmar M, Wallinga J, Teunis P et al. Frequency of adverse events after vaccination with different vaccinia strains. PLoS Med. 3 , 1341–1351 (2006). 60 Lane JM, Millar JD. Risks of smallpox vaccination complications in the United States. Am. J. Epidemiol. 93 , 238–240 (1971).4396307 61 Miravalle A, Roos KL. Encephalitis complicating smallpox vaccination. Arch. Neurol. 60 , 925–928 (2003).12873847 62 Strikas RA, Neff LJ, Rotz L et al. US civilian smallpox preparedness and response program, 2003. Clin. Infect. Dis. 46 (Suppl. 3), 157–167 (2008). • Reviews the safety data from the civilian, healthcare-worker, posteradication smallpox vaccine program. 63 Center for Disease Control and Prevention. Surveillance guidelines for smallpox vaccine (vaccinia) adverse reactions. MMWR Morb. Mortal. Wkly Rep. 55 (RR01), 1–16 (2006).16410759 64 Neff JM, Lane JM, Fulginiti VA et al. Contact vaccinia-transmission of vaccinia from smallpox vaccination. JAMA 288 , 1901–1905 (2002).12377090 65 Kemper AR, Davis MM, Freed GL. Expected adverse events in a mass smallpox vaccination campaign. Eff. Clin. Pract. 5 , 84–90 (2002).11990216 66 Lane MJ, Goldstein J. Evaluation of 21st Century risks of smallpox vaccination and policy options. Ann. Intern. Med. 138 , 488–493 (2003).12639083 67 Auckland C, Cowlishaw A, Morgan D et al. Reactions to smallpox vaccine in naive and previously-vaccinated individuals. Vaccine 23 , 4185–4187 (2003). 68 Baggs J, Chen RT, Damon IK et al. Safety profile of smallpox vaccine: insights from the laboratory worker smallpox vaccination program. Clin. Infect. Dis. 40 , 1133–1140 (2005).15791513 69 Couch RB, Winokur P, Edwards KM et al. Reducing the dose of smallpox vaccine reduces vaccine-associated morbidity without reducing vaccination success rates or immune responses. J. Infect. Dis. 195 , 826–832 (2007).17299712 70 Grabenstein JD, Winkenwerder W. US military smallpox vaccination program experience. JAMA 289 , 3278–3282 (2003).12824209 71 Sejvar JJ, Labutta RJ, Chapman LE et al. Neurologic adverse events associated with smallpox vaccination in the United States, 2002–2004. JAMA 294 , 2744–2750 (2005).16333010 72 Lewis FS, Norton SA, Bradshaw D et al. Analysis of cases reported as generalized vaccinia during the US military smallpox vaccination program, December 2002 to December 2004. J. Am. Acad. Dermatol. 55 , 23–31 (2006).16781288 73 Tasker SA, Schnepf GA, Lim M et al. Unintended smallpox vaccination of HIV-1-infected individuals in the United States military. Clin. Infect. Dis. 38 , 1320–1322 (2004).15127348 74 Casey CG, Iskander JK, Roper MH et al. Adverse events associated with smallpox vaccination in the United States, January–October 2003. JAMA 294 , 2734–2743 (2005).16333009 75 Poland GA, Grabenstein JD, Neff JM. The US smallpox vaccination program: a review of a large modern erasmallpox vaccination implementationprogram. Vaccine 23 , 2078–2081 (2005).15755574 • Reviews the large-scale, US military, posteradication smallpox vaccine program. 76 McMahon AW, Zinderman C, Ball R et al. Comparison of military and civilian reporting rates for smallpox vaccine adverse events. Pharmacoepidemiol. Drug Saf. 16 (6), 597–604 (2007).17154344 77 Karjalainen J, Heikkila J, Nieminen MD et al. Etiology of mild acute infectious myocarditis. Relation to clinical features. Acta Med. Scand. 213 , 65–73 (1983).6829323 78 MacAdam D, Whitaker W. Cardiac complications after vaccination for smallpox. Br. Heart J. 2 , 1099–1100 (1962). 79 Ahlborg B, Linroth K, Nordgren B. ECG-changes without subjective symptoms after smallpox vaccination of military personnel. Acta Med. Scand. 464 (Suppl.), 127–134 (1966). 80 Helle EP, Koskenvuo K, Heikkila J et al. Myocardial complications of immunizations. Ann. Clin. Res. 10 , 280–287 (1978).736507 81 Thorpe LE, Mostashari F, Karpati AM et al. Mass smallpox vaccination and cardiac deaths, New York City, 1947. Emerg. Infect. Dis. 10 , 917–920 (2004).15200831 82 Swerdlow DL, Roger MH, Morgan J et al. Ischemic cardiac events during the Department of Health and Human Services Smallpox Vaccination Program, 2003. Clin. Infect. Dis. 46 (Suppl. 3), 234–241 (2008). 83 Eckart RE, Shry EA, Atwood JE et al. Smallpox vaccination and ischemic coronary events in healthy adults. Vaccine 25 , 8359–8364 (2007).17981378 84 Halsell JS, Riddle JR, Atwood JE et al. Myopericarditis following smallpox vaccination among vaccinia – naive US military personnel. JAMA 289 , 3283–3289 (2003).12824210 85 von Sonnenburg F, Perona P, Schunk M et al. First clinical experience with a third generation smallpox vaccine in subjects with atopic dermatitis. Presented at: 2006 ASM Biodefense Research Meeting for the American Society for Microbiology. Washington, DC, USA, 15–18 February 2006. 86 Artenstein AW. Bioterrorism and biodefense. In: Infectious Diseases (2nd Edition). Cohen J, Powderly WG (Eds). Mosby, London, UK 99–107 (2003). 87 Ropeik D, Gray G. Risk. Houghton Mifflin Company, MA, USA (2002). 88 Breman JD, Henderson DA. Poxvirus dilemmas – monkeypox, smallpox, and biologic terrorism. N. Engl. J. Med. 339 (8), 556–559 (1998).9709051 89 Martin TM, Artenstein AW. Bioterrorism. In: The Social Ecology of Infectious Diseases. Mayer KH, Pizer HF (Eds). Academic Press, MA, USA 316–350 (2007). 90 Mack T. A different view of smallpox and vaccination. N. Engl. J. Med. 348 (5), 460–463 (2003).12496354 91 Reed KD, Melski JW, Graham MB et al. The detection of monkeypox in humans in the western hemisphere. N. Engl. J. Med. 350 (4), 342–350 (2004).14736926 92 Hammarlund EM, Lewis MW, Carter SV et al. Multiple diagnostic techniques identify previously vaccinated individuals with protective immunity against monkeypox. Nat. Med. 11 (9), 1005–1011 (2005).16086024 93 Bieber T. Atopic dermatitis. N. Engl. J. Med. 358 (14), 1483–1494 (2008).18385500 94 Ferguson NM, Keeling MJ, Edmunds WJ et al. Planning for smallpox outbreaks. Nature 425 , 681–685 (2003).14562094 95 Arita I. Smallpox vaccine and its stockpile in 2005. Lancet Infect. Dis. 5 , 647–651 (2005).16183519 96 Naleway AL, Belogia ES, Greenlee RT, Kieke BA, Chen RT, Shay DK. Eczematous skin disease and recall of past diagnoses: implications for smallpox vaccination. Ann. Intern. Med. 139 (1), 1–7 (2003).12834312 97 Belongia EA, Naleway A, Kieke B, et al. Validation of a screening instrument to identify persons for exclusion from smallpox vaccination. Clin. Infect. Dis. 40 , 620–623 (2005).15712089 98 Chapman LE, Iskander JK, Chen RT et al. A process for sentinel case review to assess causal relationships between smallpox vaccination and adverse outcomes, 2003–2004. Clin. Infect. Dis. 46 (Suppl. 3), 271–293 (2008). 99 Morgan J, Roper MH, Sperling L et al. Myocarditis, pericarditis, and dilated cardiomyopathy after smallpox vaccination among civilians in the United States, January–October 2003. Clin. Infect. Dis. 46 (Suppl. 3), 242–250 (2008). 100 Institute of Medicine. The Smallpox Vaccination Program: Public Health in an Age of Terrorism. National Academies Press, Washington, DC, USA (2005). 101 Bozzette SA, Boer R, Bhatnagar et al. A model for smallpox-vaccination policy. N. Engl. J. Med. 348 (5), 416–425 (2003).12496353 102 Wortley PM, Schwartz B, Levy PS, Quick LM, Evans B, Burke B. Healthcare workers who elected not to receive smallpox vaccination. Am. J. Prev. Med. 30 (3), 258–265 (2006).16476643 103 Kwon N, Raven MC, Chiang WK et al. Emergency physicians’ perspectives on smallpox vaccination. Acad. Emerg. Med. 10 (6), 599–605 (2003).12782519 104 Wortley PM, Levy PS, Quick L et al. Predictors of smallpox vaccination among healthcare workers and other first responders. Am. J. Prev. Med. 32 (6), 538–541 (2007).17533071 105 Millock PJ. Legal implications of the smallpox vaccination program. J. Public Health Manag. Pract. 9 (5), 411–417 (2003).15503606 106 Neff J, Modlin J, Birkhead GS et al. Monitoring the safety of a smallpox vaccination program in the United States: Report of the Joint Smallpox Vaccine Safety Working Group of the Advisory Committee on Immunization Practices and the Armed Forces Epidemiological Board. Clin. Infect. Dis. 46 , S258–S270 (2008).18284367 107 Wehrle PF, Posch J, Richter KH et al. An outbreak of smallpox in a German hospital and its significance with respect to other recent outbreaks in Europe. Bull. World Health Organ. 43 , 669–679 (1970).5313258 Websites 201 ACAM2000 Smallpox (vaccinia) vaccine, live package insert www.fda.gov/Cber/label/acam2000LB.pdf (Accessed June, 2008). 202 FDA approves second-generation smallpox vaccine www.fda.gov/bbs/topics/NEWS/2007/NEW01693.html (Accessed April, 2008). 203 ACAM2000 (smallpox vaccine) www.fda.gov/ohrms/dockets/AC/07/slides/2007–4292S2_5.ppt (Accessed June, 2008). 204 World Health Organization Fifty-eighth World Assembly A58/9 Provisional agenda item 13.6, April 7, 2005. Smallpox: global smallpox vaccine reserve. Report by the Secretariat www.who.int/gb/ebwha/pdf_files/WHA58/A58_9-en.pdf (Accessed June, 2008).	18844596	PMC9709930	NO-CC CODE	2022-12-01 23:23:09	no	Expert Rev Vaccines.; 7(8):122512-1237	utf-8	Expert Rev Vaccines	2,014	10.1586/14760584.7.8.1225	oa_other
==== Front Expert Rev Vaccines Expert Rev Vaccines Expert Review of Vaccines 1476-0584 1744-8395 Taylor & Francis 19093767 11218529 10.1586/14760584.8.1.13 Version of Record Research Article Vaccine Profile IMVAMUNE®: modified vaccinia Ankara strain as an attenuated smallpox vaccine IMVAMUNE®, Kennedy & Greenberg Kennedy Jeffrey S Greenberg Richard N Division of Infectious Diseases, Wadsworth Center, NYS Department of Health, Biggs Laboratory, C606, PO Box 509, Albany, NY 12201-0509, USA. [email protected] VA Staff Physician, Lexington VA Medical Center and Professor of Medicine, The Belinda Mason Carden and Paul Mason Professor of HIV/AIDS Research and Education, University of Kentucky School of Medicine, Department of Medicine, Room MN-672, 800 Rose Street, Lexington, KY 40536-0084, USA. [email protected] † Author for correspondence 9 1 2014 2009 8 1 1324 IntegraConverted from TF JATS 1.0 to JATS 1.2 by T&F tfjats-to-jats1.2-converter21 11 2022 © Expert Reviews Ltd 2009 Expert Reviews Ltd Smallpox vaccines based on replicating vaccinia virus are known to elicit rare yet serious adverse events, particularly in human populations with immune deficiency, atopic dermatitis and at the extremes of age. A vaccine that induces protective immune responses equivalent to first-generation smallpox vaccines while reducing the risk for severe adverse events is critical for a national stockpile of smallpox vaccines. Modified vaccinia Ankara (MVA) has been proposed as an immediate solution for vaccination of high-risk individuals. Bavarian Nordic’s vaccine MVA-BN® (IMVAMUNE®) is a MVA strain that is replication incompetent in mammalian cell lines. IMVAMUNE has been administered to more than 1900 human subjects to date, including high-risk populations (e.g., people diagnosed with atopic dermatitis or infected with HIV) in which standard replicating vaccines are contraindicated. We review the Phase I clinical trial safety profile and immune responses and compare them with other smallpox vaccines, including ACAM2000™ and Dryvax®. Keywords: adverse event clinical trial immunity modified vaccinia Ankara smallpox vaccine vaccinia ==== Body pmcIn the context of history, current smallpox vaccine development brings to mind the phrase, ‘déjà vu all over again’. The textbook by Fenner et al. reveals that scientists of 40 years ago, while successful in the eradication of smallpox, left unanswered the same questions we struggle with today: ‘what is protective immunity?’; ‘how long does immunity last?’; and ‘can we develop a safer smallpox vaccine?’ [1]. Despite our ability to understand human immune responses to viruses at an increasingly advanced cellular level, the elements of effective protective immune memory to viruses are often not well defined. For infectious diseases such as variola, where endemic disease is nonexistent, or others where field study is impractical, translation of immunological biomarkers to vaccine development remains a less than certain art. In this article, we review the current clinical development program by Bavarian Nordic A/S to develop a new strain of modified vaccinia Ankara (MVA) as a safer alternative to first-generation replicating smallpox vaccines such as Dryvax® and the second-generation vaccine ACAM2000™. Several vaccinia strains were widely used prior to the end of the compulsory smallpox vaccination program. In 1967, the WHO smallpox eradication unit surveyed 59 laboratories producing vaccinia for inoculation of humans: 51 harvested from the skin of calves, six from other bovines, three from the chorioallantoic membrane of chick embryo and three from tissue culture using bovine embryonic fibroblasts [1]. The strains included the Lister-Elstree strain (Lister Institute, UK; 39% of laboratories), New York City Board of Health (NYCBOH) (12%) and the Paris strain (12%), with the remaining laboratories using a variety of other strains (37%), including poorly characterized strains or a mixture of vaccinia and cowpox viruses [1]. Biological properties of a given vaccinia strain were known to be influenced by passage methodology but most of these vaccines had clinical efficacy against variola major. In 1963, the WHO established criteria and standards for the manufacture of vaccinia-derived smallpox vaccines, essentially reducing the derivatives of vaccinia used to three strains: Lister-Elstree, EM63 (Moscow Research Institute of Viral Preparation, Russia) and NYCBOH [2]. The WHO standards required a vaccine to produce major skin reactions in 95% of primary vaccinees and in 90% of those vaccinated 10 or more years previously, using an inoculum virus titer of 108 plaque-forming units per milliliter [1]. The three aforementioned strains of vaccinia virus were considered equivalent in protecting against smallpox during outbreaks, and provided some protection when administered within 3–4 days of exposure [3]. Although highly effective immunizing agents, each of these vaccines produced significant adverse events [4–8]. The smallpox vaccines in current global stockpiles are derived from these historical vaccinia strains. The strains differ in their pathogenicity in animal models [9,10] and, although never shown to be statistically correlated with adverse events, differences were noted in the incidence of adverse events in humans across strains [11,12]. These differences in safety profiles may also reflect changes in vaccination policy. Earlier vaccination of newborns was replaced in the late 1950s with vaccination after 1 year of age. This change significantly reduced serious adverse complications. Therefore, the exact comparator safety profile expected for a modern vaccine remains uncertain based on a number of population concerns, including prevalence of diseases that predispose risk (atopic dermatitis [AD] and immunosuppression), vaccinia strain, method used to inoculate, environmental risk factors and underlying health for those who are currently immunologically naive to vaccinia. These concerns may alter predicted frequencies of certain adverse events based on historical data [12–15]. Recent experience during the smallpox vaccination campaign demonstrated that the frequency of most adverse events that were anticipated based on historical data were lower than expected [14,16]. By contrast, myopericarditis appeared more frequently than anticipated based on historical age-based data [14,17–20]. Whether differences in the age of those typically receiving their first inoculation with vaccinia, children (<5 years historically) versus the current vaccination programs (aged > 18 years), have a role in the risk for myopericarditis is not known [15,19,21–23]. Clinical trials to date with ACAM2000 vaccine have not demonstrated definitive conclusions as to whether this serious adverse event (SAE) is more problematic with this second-generation smallpox vaccine. Vaccinia strains may differ in the frequency of adverse events elicited following inoculation [12]. NYCBOH-derived strains (as compared with other replicating strains) have historically been associated with less pathogenicity and a safety profile marked by fewer adverse events, such as postvaccinial encephalitis (PVE) deaths or resulting permanent neurological disability [24]. PVE is a complication that is not associated with excessive viral replication and, therefore, could occur as a result of immunization with IMVAMUNE®. A review of experience in humans comparing several different strains revealed that moderate-to-high pathogenic strains (Lister and Ikeda) produced more neurological events (including non-PVE) than low pathogenic strains (LC16m8, CVI and EM63), despite eliciting equivalent skin-take rates and seroconversion [24,25]. Although this secondary level of evidence suggested that less-pathogenic strains could provide improved safety without compromising efficacy, the incidence of SAEs, particularly PVE, were too rare to prove statistically significant differences between vaccinia strains. Current assessment of PVE would require clinical trials of the vaccine in millions of recipients in order to assess true differences between vaccine strains. The attenuated vaccines date back to at least 1931, when an investigator at the Rockefeller Institute (NY, USA) described serial passage of the NYCBOH strain in chick embryo cells [26]. Two strains, CVI-78 and CVII, the latter carried for a total of 235 passages over chick embryo explants, were shown to be effective with reduced adverse events by intradermal inoculation in humans [27]. The CVII vaccine was tested in 60,000 Netherland’s Army recruits and had less local and systemic reactogenicity than other replicating vaccines. However, CVII vaccination resulted in lower neutralizing antibody titers compared with a Lister-inoculated comparator group, and led to the conclusion that, without subsequent challenge with lymph-derived vaccine and induction of a skin-pock lesion, CVII would be ineffective [3]. CVI-78 was resurrected in the early 1970s by the National Institute of Allergy and Infectious Diseases, US NIH, in a study that compared four vaccines: calf lymph- and egg-derived NYCBOH, CVI-78 and Lister. The results again demonstrated that, regardless of the route of administration, induction of neutralizing antibody titers for attenuated CVI-78 were less compared with Lister or NYCBOH (75 vs 93–96% of subjects) and, upon standard challenge vaccination, the CVI-78 recipients were more likely to have primary skin responses, an indicator of nonprotective immunity [3,28,29]. Several other attenuated strains, including Darian-derived, temperature-sensitive strains (LC16m8) and G-9, an attenuated variant of the highly pathogenic vaccinia stain Temple of Heaven, were tested in human studies. For the most part, there were improved immune responses compared with prior results observed with CVI strains, although detailed data on G-9 have never been published. Modified vaccinia Ankara Modified vaccinia Ankara originates from the dermal vaccinia strain Ankara (chorioallantois vaccinia virus Ankara [CVA]) derived from a pox lesion from a horse in Ankara, Turkey, and maintained in the Vaccination Institution Ankara [30–32]. Attenuated by over 570 continuous passages in primary chick embryo fibroblast (CEF) cells, MVA was used to protect livestock against orthopoxviruses [33,34] and, later, tested as a pre-immunization to primary vaccination in human populations at risk of complications when using first-generation smallpox vaccines. Early indications that MVA could provide a safer immunogenic smallpox vaccine came from preclinical studies performed in irradiated, immunocompromised rabbits that demonstrated generation of neutralizing antibodies and protection against vaccinia virus challenge [35]. The reduction in CVA virulence through passage on CEF was reported at passage 371 [36] and the strain was renamed MVA after passage 516 in order to distinguish it from other attenuated vaccinia virus strains [33]. Early clinical development of MVA as a pre-smallpox vaccine in Europe used MVA-517 (corresponding to the 517th passage) as a priming vaccine, followed by vaccination with Lister-Elstree. These studies included subjects at risk of adverse reactions from vaccinia. In 1976, MVA derived from the MVA-571 seed stock (corresponding to the 571st passage) was tested in approximately 120,000 individuals without any reports of SAEs. Several high-risk groups were vaccinated, including young children with skin conditions [37–39]. Rigorous follow-up of those vaccinated was not performed in that setting, as is currently the standard for investigational vaccine protocols. IMVAMUNE is derived from a seed stock that corresponds to the 597th passage in CEF cells and has undergone several more rounds of limiting dilution compared with the historical MVA strain used in Germany. This MVA strain has multiple cloning sites and has been tested in humans as an experimental recombinant HIV and cancer vaccine at doses up to five-times higher than those used with the IMVAMUNE smallpox vaccination [40,41]. Bavarian Nordic has conducted an extensive preclinical development program that has demonstrated the safety, efficacy and bioequivalence of IMVAMUNE to first-generation smallpox vaccines (e.g., Dryvax and Lister-Elstree). The preclinical program has included a unique challenge model of intratracheal monkeypox and two murine orthopoxvirus challenge models. Human testing is currently ongoing, but data available to date, from Phase I and II studies, suggest that the vaccine can elicit both T- and B-cell immune responses, and the implications of how this compares to replicating vaccinia vaccines are discussed later [42,43]. IMVAMUNE properties IMVAMUNE possesses interesting properties for a new generation of attenuated smallpox vaccines. IMVAMUNE differs from other MVA strains in that it has undergone extensive plaque purification in CEF cells, is propagated in serum-free conditions and, after passage 570, is a genetically stable virus [44–46]. The attenuation characteristics of this MVA separate it from other MVA strains [101]. Most MVA strains used in research are polyclonal and contain virus subtypes that are genetically similar to a replicating vaccinia strain. These subtype viruses, if virus host-range genes (e.g., K1L) are present and thus enable mammalian cell replication, could result in the emergence of phenotypes identical to replicating vaccinia. Culture replicating subtypes can theoretically emerge in vivo as replicating vaccinia virus, particularly when injected into immunocompromised animals [46,47]. However, rescue studies on host-range defects [48] and injection of MVA at 1000-times the lethal dose of vaccinia failed to induce death in severe combined immunodeficiency (SCID) mice [49]. Although it can be speculated that human use of such MVA variants can lead to the isolation of these virulent strains in vivo, to date this has not been observed in preclinical animal testing or human clinical trials involving multiple MVA-based vaccines. IMVAMUNE is replication deficient in mammalian cells and, in comparison to direct ancestral MVA strains, produces an acceptable safety profile in severely immune compromised animals [101]. Comparison of the genomes of IMVAMUNE and the ancestral vaccinia Ankara strain CVA demonstrates a loss of 15% of the genetic information (∼31,000 bp of DNA), resulting from six major deletions in the MVA genome. Deletion of the host-range gene K1L and deletions of other virulence and host-range genes, including the gene for type A inclusion bodies, may partially explain the lack of replication in mammalian cell lines of IMVAMUNE [44,46,50]. Interestingly, even though IMVAMUNE exhibits attenuated replication in mammalian cells, efficient transcription of the viral proteins occurs within host cells and the functional result of the deletions appears to block virus assembly and egress [45,46,51,52]. This latter finding may have effects on comparative immune responses to first- or second-generation vaccines in which vaccinia virus assembly and egress occurs. However, notwithstanding the significant effect of genome deletion on attenuation and reduced virulence, preclinical studies using IMVAMUNE have demonstrated the ability to induce humoral and cellular immune responses to both vaccinia genes as well as cloned target proteins. Human cells infected with replicating vaccinia respond by induction of type I interferons (IFNs) and the expression of a soluble IL-1 receptor [47], a potential virulence factor for certain poxviruses [53,54]. Viral induction of type I IFNs and the expression of the receptor may lead to interference with human immune responses through molecular mimicry of cytokines and immune receptors. MVA does not express soluble receptors for IFN-γ, IFN-α/β, TNF or CC chemokines [50], which could influence comparative effects on innate immune responses [55] or the induction of stable long-term memory responses [47,56]. Manufacturing, potency & administration of vaccine The formulation of IMVAMUNE used in clinical trials is a purified live vaccine produced under serum-free conditions in CEF cells, supplied as a sterile liquid-frozen preparation and does not contain adjuvants or preservatives. In preparing the vaccine, virus cell suspensions are homogenized by a process that includes a freeze and thaw, Benzonase® and ultrasound treatment. The bulk vaccine is adjusted to a titer of 2 × 108 tissue culture infectious dose (TCID50)/ml, filled into 2-ml injection vials and stored below -15°C as a frozen liquid product. Prior to administering the vaccine, the vial is thawed at room temperature and gently swirled for at least 30 s. The injection volume is 0.5 ml (1.0 × 108 TCID50) per dose and is administered as a subcutaneous injection. IMVAMUNE can also be effectively administered via intramuscular injection. Stability tests are continuing but current preliminary data support stability for at least 2 years at -15°C. In reference to historical smallpox vaccine guidelines, the WHO’s efforts to develop freeze-dried vaccines (Dryvax, Lister and EM-63) were partially in response to the instability of vaccinia strains in liquid form during extended exposure to climate level temperatures. Long-term stability of IMVAMUNE is under evaluation. In addition, the vaccine, once thawed, must be administered immediately within 12 h; this further distinguishes it from Dryvax and ACAM2000, since both have extended post-thawing half-lives of 1–2 weeks. IMVAMUNE is not stable to refreeze once thawed and would require cold-chain logistics during an epidemic. Summaries of nonclinical studies The studies conducted for evaluating the safety prior to initiation of human trials included repeat administrations (subcutaneous and intramuscular) in animal models that demonstrated reversible nondose-limiting injection-site reactions and lymphoid changes. Injection-site expression levels of vaccinia by RT-PCR revealed only one weakly positive reaction at injection sites beyond day 3 (at day 7), suggesting no persistence of virus or genetic incorporation. Teratology studies in both rats and rabbits did not demonstrate teratogenic or intrauterine toxicity, and peri- and postnatal studies did not reveal any toxicity to embryos or developing offspring at doses up to 1 × 108 TCID50. Three preclinical evaluations provide the primary support for safety and efficacy. To assess the attenuation profile of different strains of MVA, their growth was compared on CEF cells and across a number of mammalian cell lines. The strains included IMVAMUNE (MVA-597), MVA-572 (a 1970s smallpox vaccine used in Germany [57]), MVA-Vero (a MVA-575 strain adapted to Vero cell line [58]) and MVA-I721 (a MVA strain deposited at the Institute Pasteur [accession number I-721]. All four strains replicated in primary CEF cells but demonstrated significant differences in the ability to replicate in mammalian cells, with IMVAMUNE showing the near absence of replication across monkey (CV-1), rabbit (RK-13), murine (AG-101) and human (143-B, HeLa) mammalian cell lines [101]. These results suggest that MVA variants may exist in the MVA-572 and MVA-1721 parental lots. This conclusion is supported by studies inoculating immune-deficient mice (AGR129 strain), which led to in vivoreplication and the emergence of altered genetic variants from these MVA strains but not from IMVAMUNE [59]. These results challenge the perception that all MVA strains are similar and suggest that lack of human cell replication, as observed with IMVAMUNE, represents a more stable clonal virus that could be uniquely suited to a human attenuated smallpox vaccine. The second line of evidence for clinical attenuation was demonstrated using an immune-compromised mouse model inoculated with IMVAMUNE. AGR129 mice have gene deletions that inactivate the function of the IFN system (IFN receptor types I [α/β] and type II [IFN-γ]) and inactivate the production of IFN-γ on a Rag-/- gene-deletion background), which prevents the formation of mature B and T cells. To evaluate IMVAMUNE, seven strains of vaccinia differing in virulence (IMVAMUNE, MVA-572, MVA-Vero, MVA-I721, smallpox vaccines Dryvax and Lister-Elstree, and a highly lethal vaccinia strain Western Reserve [WR]) were administered intraperitoneally at 1 × 107 TCID50. AGR129 mice that were inoculated with IMVAMUNE survived for more than 120 days with no replicating virus isolated from the ovaries (a marker of virus dissemination) at any time point. Animals inoculated with other MVA strains died or demonstrated virus-mediated disease at an average time to disease of 20 days (MVA-I721), 80 days (MVA-Vero) and 81 days (MVA-572) [101]. Recent reports of MVA providing protection against otherwise lethal poxvirus infection in Toll-like receptor-9-deficient mice suggests that MVA recognition and induction of immunity can occur through both Toll-like receptor-9-dependent and -independent pathways [60,61]. Replicating vaccinia strains Dryvax, Lister-Elstree and WR killed all mice within 8 days of inoculation. The combined immunodeficiency mouse model provides evidence that MVA strains differ in their preclinical safety profiles and there is lack of mammalian cell replication with IMVAMUNE. Preclinical efficacy was tested using two unique challenge models, employing both the Ectromelia virus (ECTV) and monkeypox virus (MPXV) [43]. Earlier animal studies demonstrated that IMVAMUNE induces an equivalent humoral and T-cell immune responses in mice compared with first-generation smallpox vaccines (Dryvax and Elstree) [42,62,63]. In the ECTV challenge model, a single dose of IMVAMUNE protected mice from intranasal challenge with a lethal dose of VV-WR or ECTV that was given 6 weeks after vaccination. Mice challenged 3 days after vaccination with WR virus were also protected, consistent with previous studies using first-generation vaccines and variola or vaccinia challenge [64]. In the nonhuman primate model of intratracheal delivery of MPXV, challenged animals vaccinated with IMVAMUNE were completely protected, with the exception that one animal developed skinpox lesions. Viral loads were even more diminished if IMVAMUNE was followed by conventional inoculation with the Elstree vaccine prior to the challenge. The overall clinical response to vaccination with IMVAMUNE was equivalent to first-generation smallpox vaccines [43]. These challenge studies provide important efficacy data in that the MPXV model mimics natural variola infection, producing a smallpox-like disease (i.e., fever, skin lesions and fibrinonecrotic bronchopneumonia); although 100% of placebo-treated animals die [43], which is in contrast with the lower mortality in human smallpox [65]. Clinical experience with IMVAMUNE IMVAMUNE, as a third-generation smallpox vaccine for special populations at risk of SAEs from replicating vaccinia, must demonstrate a safety profile that succeeds at eliminating common complications associated with replicating vaccinia-based vaccines. For instance, accidental spread of virus to close contacts is not probable with IMVAMUNE due to the lack of replication in human cells and intramuscular inoculation eliminating the development of skinpock lesion. The occurrence of eczema vaccinatum and progressive vaccinia also seem unlikely but these assessments will require continued monitoring in clinical trials and beyond. Assessment of the occurrence of PVE will also require long-term monitoring. The rarity of both PVE and myopericarditis requires large numbers of volunteers to receive IMVAMUNE in order to achieve statistically superior safety assessment over first- or second-generation vaccines. Overall, any comparison trial would require a treatment arm receiving a replicating vaccine. Use of a replicating vaccinia vaccine as part of a trial design to compare adverse events is not acceptable for these special populations. Historical data using the MVA vaccine in over 120,000 individuals during the 1970s provide encouraging evidence that MVA as a smallpox vaccine can be safely administered to humans. In 1976, MVA was authorized in Germany as a preimmunization vaccine in combination with Lister vaccine. This two-step inoculation program was thought to diminish potential adverse events, particularly PVE. More than 120,000 subjects received intradermal and subcutaneous injections with a low dose of MVA (1 × 106 TCID50) before the mass-vaccination program was discontinued [37]. Prior clinical trials described the safety and tolerability of the vaccine and served as the basis for licensure in Germany [66–68]. These studies included data on 7098 subjects (5691 aged <3 years and 1407 aged >3 years). Reactions at the injection site revealed no blistering, pustules or ulcerations, and no cases of PVE or other SAEs were observed. Reactogenicity was limited to approximately 4% with mild fever (>38°C; 2.28%) and approximately 4% with nonspecific systemic symptoms. However, detailed immunogenicity was never sufficiently tested. Administering MVA prior to first-generation smallpox vaccine reduced the frequency of adverse events from the replication-competent vaccinia virus. By contrast, with IMVAMUNE, single-dose vaccination with this earlier CEF-derived MVA elicited only a weak hemagglutinin inhibition and virus-neutralizing antibody titers. MVA prevaccination resulted in a strong booster effect after vaccination with a replicating vaccinia strain, indicating that priming with MVA induced specific humoral and celllar immune responses. Human safety of IMVAMUNE To date, more than 2000 subjects have been vaccinated with either IMVAMUNE or recombinant MVA-BN®-based vaccines, including at-risk populations with AD or HIV infection. Tables 1 & 2 provide an overview of ongoing and completed IMVAMUNE clinical studies. The use of MVA-BN-based vectored vaccines in clinical trials is not reviewed here, since these vaccines represent altered viruses and, therefore, comparison and interpretation of the safety and tolerability data can be complex and problematic. In summary, however, administration of MVA-BN-vectored vaccines in HIV-infected subjects has not been associated with dissemination or any severe systemic or neurological events attributable to vaccinia. Adverse vaccine reactions Clinical trials of IMVAMUNE in subjects over 18 years of age have included HIV-infected subjects and individuals with a history of AD, or active AD. The frequency of reactions occurring in the 1025 subjects enrolled in the completed IMVAMUNE clinical trials is shown in Table 3. Summaries to date suggest that the majority of the events observed in these trials have been classified as being mild to moderate and resolving without sequelae. Reported local and systemic reactions are consistent with prior observations with MVA [69]. Of particular interest are adverse events occurring in special populations with AD (POX-MVA-007 and -008) and HIV-1 infection (POX-MVA-010 and -011), which are two populations contraindicated for first- and second-generation smallpox vaccines. POX-MVA-007 was a Phase I clinical trial that used two injections of IMVAMUNE in adults aged 18–40 years. A total of 60 subjects were enrolled: 15 patients with mildly active AD and 16 patients with a history of AD were compared with healthy subjects (n = 15) and subjects with allergic rhinitis (n = 14). No SAEs or premature study withdrawal events have been reported. Mild-to-moderate transient pain and redness at the inoculation site occurred in nearly all subjects and no differences in the incidence of grade 3 and above adverse reactions during the immediate postvaccination period were observed between AD and healthy controls. A larger open-label comparative (AD vs no AD) trial, POX-MVA-008, is currently ongoing and will recruit up to 530 individuals aged between 18 and 40 years in the USA and Mexico. In trials POX-MVA-010 and -011, the objectives were to compare safety and immunogenicity data in HIV-infected subjects with uninfected subjects. Study POX-MVA-010 included 91 subjects with HIV (30 naive and 61 experienced to vaccinia). To date, there have been no reports of SAEs in the published Phase I studies [70,71]. The company will soon publish complete safety data from the completed larger cohort studies (Chaplin P, Pers. Comm.), which limits their summary in this report. A recombinant BN-MVA containing the HIV nef protein construct has been administered to 14 HIV-infected patients (CD4 > 400/µl), with only mild systemic reactions noted [40]. Special interest adverse events (cardiac events) Due to recent clinical trials with ACAM2000 and Dryvax and First Responder Vaccination Programs employing Dryvax, there is additional concern regarding the association of vaccinia inoculation and the development of myopericarditis after vaccination [17,72–74]. Myocarditis and pericarditis as a sequelae of viral infection has, for some viruses, been shown to be autoimmune related and generally occurs in later teen years or young adulthood [22,75,76]. Whether the incidence of vaccinia-related myopericarditis is indeed more frequent than recognized in the past as a consequence of changing vaccination patterns using older subjects is subject to debate. Recently reported incidence rates for developing myopericarditis following vaccination with the first-generation smallpox vaccine Dryvax (10.38 events per 1000 vaccinations) and a second-generation replicating vaccinia vaccine ACAM2000 (5.73 events per 1000 vaccinations) have been a surprise and are alarming [77,78]. Historical data related to myopericarditis must also be interpreted carefully as those studies were prior to the clinical use of echocardiograms and cardiac-specific enzymes. Adverse events from published clinical trials using IMVAMUNE and MVA-BN-based recombinant vaccines have not reported a case of myopericarditis. Further evidence supporting the cardiac safety of IMVAMUNE was obtained from a placebo-controlled Phase II study in healthy individuals (POX-MVA-005), which intensively monitored the risk of myopericarditis developing after vaccination with IMVAMUNE. There have been no cases of myopericarditis reported for this study (Chaplin P, Pers. Comm.). Immune response to IMVAMUNE in healthy individuals In order to delineate what is known to date regarding IMVAMUNE immune responses, current and completed clinical trials can be broken down into vaccinia-naive or -experienced subjects with or without either AD or HIV. POX-MVA-001, -002, -004 & -005 These four trials represent the initial IMVAMUNE trials in adults. Tables 1 & 2 depicts the design of each of these trials in terms of groups, dose tested, prior vaccinia status and numbers tested. POX-MVA-002 was independently conducted by the National Institute of Allergy and Infectious Diseases; Bavarian Nordic sponsored the others. Tables 1 & 2 presents an incomplete summary of the humoral and cellular responses reported to date. We are only able to report complete data for two trials: POX-MVA-001 and -002. Historically, a plaque-reduction neutralization titer (PRNT) greater than 1:40 is considered a positive humoral response. In Table 4, comparative IFN-γ-ELISPOT (>15 spots/million peripheral blood mononuclear cells [PBMCs] considered positive) using a standardized assay across all vaccines depicted offers an assessment of the T-cell responses observed with IMVAMUNE compared with that observed with first-generation Dryvax, experimental cell-cultured smallpox vaccine and ACAM2000 [78–82]. This assay was employed across clinical trials of smallpox vaccines, utilizing the same reagents, assay controls and methods, thus enabling comparison of IFN-γ T-cell responses across vaccine candidates [70,79–82]. In the POX-MVA-001 trial, 86 male subjects aged between 18 and 55 years were vaccinated and stratified based on the presence or absence of a prior history of smallpox vaccination. The first four dosing cohorts were naive to smallpox vaccination and received the vaccine on days 0 and 28. A fifth cohort of subjects, who had been previously vaccinated against smallpox, received a single subcutaneous dose of IMVAMUNE at 1 × 108 TCID50. Maximum seroconversion rates by ELISA, defined as titers of at least 100, were reached at day 42, with 100% seroconversion after the 1 × 108 TCID50 dose administered by subcutaneous or intramuscular injection. The analysis of variance (ANOVA) indicated a significant superiority of the 1 × 108 and 1 × 107 TCID50 doses over the lower doses. In the group of previously vaccinated subjects, all showed a rapid booster response and 100% seroconversion by total IgG ELISA [71]. In the POX-MVA-002 trial, high seroconversion rates were observed for all doses after the second vaccination (87–100%). In the POX-MVA-002 trial, measurement of neutralizing antibodies (PRNT assay) did not demonstrate a 100% seroconversion rate with the highest dose of IMVAMUNE, although a dose-dependent increase in neutralizing (PRNT) antibody responses was observed. Since complete data from the larger POX-MVA-004 trial has been published, detailed analysis of the antibody PRNT and ELISA values can be made particularly in comparison to those observed with Dryvax and ACAM2000 or in the POX-MVA-002 trial, which directly compared IMVAMUNE with Dryvax. These early results, from POX-MVA-001 and POX-MVA-002, raise the interesting question as to whether some degree of replication may be an important determinant for neutralizing antibodies [70,80–84]. This premise is supported by the work of others demonstrating that various forms (extracellular) of vaccinia virus are typically generated after infection of human cells. These forms are important in the generation of hemagglutinin and neutralizing antibodies to the virus, which may be critical early in protective immunity [85–87]. Further work is needed to clarify this issue. A peak CMI response was reached approximatley 2 weeks after the second vaccination (day 42) for groups 1–4 in the POX-MVA-001 trial. While a cell-mediated response was detected in five subjects with pre-immunity on day 0 (group 5), those subjects showed a booster response on day 28 following a single vaccination with IMVAMUNE, with a mean T-cell count of 109 cells per million PBMCs. This would indicate that IMVAMUNE was able to stimulate the memory T-cell response induced by a previous replicating smallpox vaccination. In vaccinia-naive subjects, dose responses were observed, with better responses being recorded using the higher doses of IMVAMUNE. At the highest dose (groups 3 and 4), the cell-mediated responses measured on day 42 were in the same range (102 and 152) as subjects with pre-existing immunity (group 5) on day 28. A correlation analysis showed that the T-cell response measured in the ELISPOT was highly correlated to the results of the PRNT and ELISA antibody tests (data not shown) [70,71]. These Phase I results confirm dose-finding data from the study POX-MVA-004 (see later), which demonstrated that a dose of 1 × 108 TCID50 IMVAMUNE was well tolerated and the most immunogenic dose evaluated in healthy subjects [70]. Two larger multicenter Phase II clinical trials in subjects with AD (POX-MVA-008) or HIV-1 infection (POX-MVA-011) are underway to evaluate the immunogenicity and safety of two doses of 1 × 108 TCID50 of IMVAMUNE subcutaneously, given 1 month apart. The primary study objective of the POX-MVA-008 trial is to assess the humoral immune response (measured by ELISA) induced by IMVAMUNE in subjects with AD and, in the POX-MVA-011 trial, safety of the vaccine in HIV-infected subjects. Data on the POX-MVA-011 trial is expected later this year. Expert commentary Derived from MVA, IMVAMUNE is a highly attenuated strain of vaccinia that is unable to replicate in human cells and, therefore, cannot be transmitted or cause dispersed vaccinia infection. The extensive nonclinical development has shown the vaccine to induce protective immunity in two important animal challenge studies: immune-compromised mice and primates challenged with MPXV. The issue of durable immunogenicity, both humoral and cellular, requiring the need for repeat inoculations to boost responses, raises some concerns and challenges to develop a best-use strategy during an outbreak scenario. In addition, patients with AD inoculated with vaccinia represent a unique population with altered skin host-defense mechanisms [88], and how these alterations might affect immunogenicity of IMVAMUNE remains to be determined. Finally, MVA as a vaccine strain does not produce the typical skinpock lesion, which eliminates a useful marker of successful vaccination for field assessment of protection during outbreaks. Recently, antibody profiling across species (including humans) for MVA, Dryvax and WR strains demonstrated that binding antibodies to select structural vaccinia proteins were similar across species [53]. Although this study has some limitations with respect to the panel of vaccinia proteins selected, the data demonstrate remarkable similarity of binding profiles, suggesting that the 3% genetic variation observed between MVA and Dryvax may not significantly alter the repertoire of humoral immune response. However, the deletion mutations of specific proteins in IMVAMUNE could alter the magnitude of protective levels of humoral or cellular immunity, the ability to neutralize variola major and the durability of response. These issues will require further study focusing on the site-directed nature of the neutralizing antibody response. Of concern remains the lower neutralizing geometric mean titer in MVA-vaccinated individuals and whether this reflects altered response to early vaccinia antigens, which others have suggested may be directed to specific vaccinial proteins: B5R, D8, A27, D13 and A14 [54,55]. Development of third-generation smallpox vaccines that target the key antigens required for neutralization of vaccinia forms intracellular mature virus and extracellular mature virus are new candidates entering human testing and may provide a unique boost combination with IMVAMUNE. Such boosting may ultimately contribute to more durable neutralizing antibody titers and T-cell memory. Of course, all of the previous commentary on immunogenicity and durability really depends on IMVAMUNE protection against variola; only studies or experience with variola will reveal this answer. So far, there is the expectation that IMVAMUNE will be effective. Five-year view To date, more than 1900 subjects have been vaccinated with IMVAMUNE and recombinant MVA-BN-based vaccines; no cases of myopericarditis have been observed. Therefore, these early clinical studies suggest that IMVAMUNE may offer a safer cardiac profile, that is, have a much lower rate of cardiac events compared with Dryvax and ACAM2000. Whether the unique absence of replication in human cells translates to an improved safety profile remains to be determined by expanded clinical testing and usage. In addition, the immunogenicity data generated to date suggest that both humoral and cellular immune responses are lower, particularly following one dose. Future considerations with regards to IMVAMUNE include recent development and testing of smallpox vaccines based on DNA plasmids and recombinant protein vaccines, enabling higher antigenic content and use of adjuvants to boost neutralizing antibodies to targeted vaccinia proteins. The role of IMVAMUNE may emerge as a prevaccination, which could be subsequently boosted with newer vaccine designs. Although the limited data thus far suggest two doses of IMVAMUNE induces sufficient antibody and cellular immune responses in the immediate vaccination period, the decay of the response compared with replicating smallpox vaccines needs further study to support IMVAMUNE use as a vaccine that would provide 3–5 years of protection. This concern is supported by recent data comparing three nonreplicating smallpox vaccines in mice that failed to demonstrate long-term (150 days postvaccination) protection against intranasal challenge with cowpox virus [89]. Perhaps a schedule that includes periodic IMVAMUNE booster vaccinations will solve this issue. On the other hand, in a recent nonhuman primate study, MVA did prevent infection with vaccination 4 days prior to a monkeypox challenge, while Dryvax required vaccination 6 days prior to challenge to be effective [60,90]. These primate-challenge models emphasize differences between MVA and replicating vaccinia smallpox vaccines that need to be considered in the strategies for human use in the pre- and postexposure settings. Finally, we believe that the IMVAMUNE program is well guided, providing the necessary information needed to evaluate the vaccine for US FDA approval for special populations in an emergency. 10.1586/14760584.8.1.13-T0001 Table 1. Summary of completed human trials of IMVAMUNE® attenuated smallpox vaccine. Study Population Dose (route) Dose of IMVAMUNE n GMT* (% positive) nAb‡ (% positive) POX-MVA-001§ Vaccinia naive 106 (sc.) 2 18 39 33 107 (sc.) 2 16 81 50 108 (sc.) 2 16 100 80 108 (im.) 2 18 100 87 Nonvaccinia naive 108 (sc.) 1 18 100 89 POX-MVA-002§ Vaccinia naive 2 × 107 (sc.) 2 + Dryvax® 15 100 100 5 × 107 (sc.) 2 + Dryvax 15 100 93 1 × 108 (sc.) 2 + Dryvax 15 100 87 Placebo × 2 + Dryvax 15 84 85 1 × 108 (sc.) 2 + placebo 15 100 92 1 × 108 (im.) 2 + Dryvax 15 100 100 Measuring total anti-MVA IgG by ELISA, percentage responding 14 days after last dose of IMVAMUNE or 14 days after Dryvax alone. ‡Conversion percentages are based on plaque-reduction neutralization titer GMT titer approximately 30 days after last dose. §See clinical trial [71,72]. GMT: Geometric mean titer; im.: Intramuscular; MVA: Modified vaccinia Ankara; NA: Data not available in peer-review published or reported form; nAb: Neutralizing antibody; sc.: Subcutaneous. 10.1586/14760584.8.1.13-T0002 Table 2. Study design of on-going studies for which immunogenicity data is not available. Study Population Dose (route) Doses n POX-MVA-004 Vaccinia-naive 2 × 107 (sc.) 2 55 5 × 107 (sc.) 2 55 108 (sc.) 2 55 POX-MVA-005 Vaccinia-naive 108 (sc.) 2 183 108 (sc.) 1 + placebo 181 108 (sc.) Placebo 181 Nonvaccinia-naive 108 (sc.) 1 200 POX-MVA-007 Vaccinia-naive 108 (sc.) 2 15 History of AD 108 (sc.) 2 16 Mild active AD 108 (sc.) 2 15 Mild allergic rhinitis 108 (sc.) 2 14 POX-MVA-008 Vaccinia-naive 108 (sc.) 2 230 AD, vaccinia-naive 108 (sc.) 2 300 POX-MVA-010 HIV-infected naive 108 (sc.) 2 30 HIV-infected non-naive 108 (sc.) 1 61 Naive 108 (sc.) 2 30 Non-naive 108 (sc.) 1 30 POX-MVA-011 Naive 108 (sc.) 2 90 HIV+ CD4 200–750 108 (sc.) 2 360 AD: Atopic dermatitis; MVA: Modified vaccinia Ankara; sc.: Subcutaneous. 10.1586/14760584.8.1.13-T0003 Table 3. Comparison of IMVAMUNE®/ACAM2000™/Dryvax® adverse events (possibly related) occurring in at least 5% of trial subjects. Adverse event characterization Vaccinia-naive subjects Preferred term IMVAMUNE (n = 1025) (n; %) ACAM2000 (n = 873) (n; %)‡ Dryvax (n = 289) (n; %)‡ Blood and the lymphatic system disorders Lymphadenopathy 13 (1) 72 (8) 35 (12) Lymph node pain 1 (0.1) 494 (57) 199 (69) Nervous system disorders Headache 280 (27) 433 (50) 150 (52) Respiratory, thoracic and mediastinal disorders Dyspnea 0 (0) 39 (4) 16 (6) Gastrointestinal disorders Nausea 105 (10) 170 (19) 65 (22) Diarrhea 8 (1) 144 (16) 34 (12) Constipation 0 (0) 49 (6) 9 (3) Vomiting 1 (0.1) 42 (5) 10 (3) Skin and subcutaneous tissue disorders Erythema 1 (0.1) 190 (22) 69 (24) Rash 3 (0.3) 94 (11) 30 (10) Musculoskeletal, connective tissue and bone disorders Myalgia 103 (10) 404 (46) 147 (51) General disorders and administration-site conditions Injection-site erythema 827 (81) 649 (74) 229 (79) Injection-site pain 887 (87) 582 (67) 208 (72) Injection-site pruritus 211 (21) 804 (92) 277 (96) Injection-site swelling 692 (68) 422 (48) 165 (57) Fatigue 316 (31) 423 (48) 161 (56) Malaise 5 (0.5) 327 (37) 122 (42) Rigors 31 (3) 185 (21) 66 (23) Exercise tolerance decreased 0 (0) 98 (11) 35 (12) Feeling hot 1 (0.1) 276 (32) 97 (34) Summary of published and unpublished data from incompleted and reported IMVAMUNE clinical trials. ‡Prescribing information for ACAM2000, August 2007. Modified from [91]. 10.1586/14760584.8.1.13-T0004 Table 4. Comparison of cell-mediated immunity induction for smallpox vaccines using a standardized IFN-γ-ELISPOT assay in human trials in naive subjects. Vaccine name and study n Vaccine titer Route (+) ELISPOT (%) Mean SFCs per 106 PBMC 30 days post first dose Mean SFCs per 106PBMC 30 days post second dose Dryvax® ‡ 90 1 × 108 PFU/ml id. 98 402 NA IMVAMUNE® 21 1 × 108 TCID50 im. 95 100 286 ACAM2000™ 30 1 × 108 PFU/ml id. 99 442 NA ACAM1000 30 1 × 108 PFU/ml id. 100 331 NA CCSV 40 1 × 108 PFU/ml id. 99 251 NA Controls§ 10 Naive controls 0 5.8 4.1 Variation of response was 0–15 SFC for a given well run in triplicate. Value is the mean across five studies in the table, except for 30 days post-second dose, where data are mean for IMVAMUNE study alone. *Studies cited are IMVAMUNE [71], ACAM1000 [81], CCSV [80] and ACAM2000. ‡Dryvax® (n) data: all subjects comparatively injected with Dryvax in the studies of experimental vaccines in the table. §Nonimmunized controls were used in the assay. One donor was used for every three plates. Summary data is for all studies and represents repeated measures for ten donors. CCSV: Cell-cultured smallpox vaccine; id.: Intradermal by bifurcated needle; im.: Intramuscular injection; NA: Not assessed (only one dose administered); PBMC: Peripheral blood mononuclear cell; PFU: Plaque-forming unit; SFC: Spot-forming cell; TCID: Tissue culture infectious dose. Key issues • IMVAMUNE®, an attenuated vaccinia strain, has been shown not to replicate in human cell lines, reducing the risk of transmissibility to close contacts – a key new safety advantage. • The vaccine is a modern cell culture-derived vaccine, developed for use in special populations at risk of complications from replicating vaccinia-based smallpox vaccines. • In 2007, a report of Phase I studies comparing immunogenicity of IMVAMUNE to Dryvax® revealed that at dose of 1 × 108 TCID50 IMVAMUNE could elicit robust humoral and cellular immune responses, albeit lower than Dryvax. The duration of protective antibody and T-cell responses are unclear and requires further evaluation in both low- and high-risk populations. • Numerous registration trials in special at-risk populations (atopic dermatitis and HIV-1 infected) are ongoing or planned, and key data from these trials are expected in 2009. • To date, in limited small trials, IMVAMUNE has been well tolerated and no myocardial events have been reported. • Cold-chain and two-dose requirements with respect to IMVAMUNE may result in logistic problems during outbreak scenarios. • The important unfilled niche in the supply of smallpox vaccine in the event of a bioterrorism act is to protect populations for which replicating vaccinia-based vaccines are contraindicated. IMVAMUNE based on replication-deficient modified vaccina Ankara provides an attenuated smallpox vaccine for these special populations. Such a vaccine should strive for single-dose immunogenicity comparable to first-generation vaccines, a reasonable duration of protective immunity, as well as a superior safety profile. As additional safety and immunogenicity data from IMVAMUNE clinical trials become available, the capabilities and role for IMVAMUNE will be resolved. Acknowledgements The authors would like to thank P Chaplin, A von Krempelhuber, N Arndtz, G Virgin and the scientists at Bavarian Nordic for providing the IMVAMUNE ® data included in Table 3 and the cited comments that shed light on trial data that will be published later this year. Financial & competing interests disclosure R Greenberg has performed clinical trials for Bavarian Nordic. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript. ==== Refs References 1 Fenner F, Henderson DA, Arita I, Jezek Z, Ladnyi ID. Smallpox and its Eradication. History of International Public Health. World Health Organization, Geneva, Switzerland (1988). 2 Krag P, Bentzon MW. The international reference preparation of smallpox vaccine. An international collaborative assay. Bull. World Health Organ. 29 , 299–309 (1963).14058224 3 Meltzer MI. Risks and benefits of preexposure and postexposure smallpox vaccination. Emerging Infect. Dis. 9 (11), 1363–1370 (2003). 4 Galasso GJ, Mattheis MJ, Cherry JD et al. Clinical and serologic study of four smallpox vaccines comparing variations of dose and route of administration. J. Infect. Dis. 135 (1), 183–186 (1977).833448 5 Cherry JD, Connor JD, McIntosh K et al. Clinical and serologic study of four smallpox vaccines comparing variations of dose and route of administration. Standard percutaneous revaccination of children who receive primary subcutaneous vaccination. J. Infect. Dis. 135 (1), 176–182 (1977).188953 6 Connor JD, McIntosh K, Cherry JD et al. Clinical and serologic study of four smallpox vaccines comparing variations of dose and route of administration. Primary subcutaneous vaccination. J. Infect. Dis. 135 (1), 167–175 (1977).833447 7 Vaccinia (smallpox) vaccine. Recommendations of the Immunization Practices Advisory Committee (ACIP). MMWR Recomm. Rep. 40 (RR-14), 1–10 (1991). 8 Cherry JD, McIntosh K, Connor JD et al. Clinical and serologic study of four smallpox vaccines comparing variations of dose and route of administration. Primary percutaneous vaccination. J. Infect. Dis. 135 (1), 145–154 (1977).188951 9 Morita M, Aoyama Y, Arita M et al. Comparative studies of several vaccinia virus strains by intrathalamic inoculation into cynomolgus monkeys. Arch. Virol. 53 (3), 197–208 (1977).404993 10 Morita M, Arita M, Komatsu T, Amano H, Hashizume S. A comparison of neurovirulence of vaccinia virus by intrathalamic and/or intracisternal inoculations into cynomolgus monkeys. Microbiol. Immunol. 21 (7), 417–418 (1977).409906 11 Marennikova SS, Maltseva NN. [Comparative study of some strains of vaccinia virus. II. Pathogenicity for laboratory animals]. Vopr. Virusol. 126 , 287–291 (1964).14237276 12 Kretzschmar M, Wallinga J, Teunis P, Xing S, Mikolajczyk R. Frequency of adverse events after vaccination with different vaccinia strains. PLoS Med. 3 (8), E272 (2006).16933957 13 Fulginiti VA, Papier A, Lane JM, Neff JM, Henderson DA. Smallpox vaccination: a review, part II. Adverse events. Clin. Infect. Dis. 37 (2), 251–271 (2003).12856218 14 Kemper AR, Davis MM, Freed GL. Expected adverse events in a mass smallpox vaccination campaign. Eff. Clin. Pract. 5 (2), 84–90 (2002).11990216 15 Neff JM, Lane JM, Pert JH, Moore R, Millar JD, Henderson DA. Complications of smallpox vaccination. I. National survey in the United States, 1963. N. Engl. J. Med. 276 (3), 125–132 (1967).4381041 16 Poland GA, Grabenstein JD, Neff JM. The US smallpox vaccination program: a review of a large modern era smallpox vaccination implementation program. Vaccine 23 (17–18), 2078–2081 (2005).15755574 17 Halsell JS, Riddle JR, Atwood JE et al. Myopericarditis following smallpox vaccination among vaccinia-naive US military personnel. JAMA 289 (24), 3283–3289 (2003).12824210 18 Morgan J, Roper MH, Sperling L et al. Myocarditis, pericarditis, and dilated cardiomyopathy after smallpox vaccination among civilians in the United States, January–October 2003. Clin. Infect. Dis. 46 (Suppl. 3), S242–S250 (2008).18284365 19 Lane JM, Ruben FL, Neff JM, Millar JD. Complications of smallpox vaccination, 1968: results of ten statewide surveys. J. Infect. Dis. 122 (4), 303–309 (1970).4396189 20 Matthews AW, Griffiths ID. Post-vaccinial pericarditis and myocarditis. Br. Heart J. 36 (10), 1043–1045 (1974).4279685 21 Bengtsson E, Holmgren A, Nystrom B. Smallpox outbreak and vaccination problems in Stockholm, Sweden 1963. Circulatory studies in patients with abnormal ECG in the course of postvaccinal complications. Acta Med. Scand. 464 , 113–126 (1966). 22 Bengtsson E, Hansson S, Nystrom B. Smallpox outbreak and vaccination problems in Stockholm, Sweden 1963. V. Postvaccinal reactions and complications. Acta Med. Scand. 464 , 87–104 (1966). 23 Karjalainen J, Heikkila J, Nieminen MS et al. Etiology of mild acute infectious myocarditis. Relation to clinical features. Acta Med. Scand. 213 (1), 65–73 (1983).6829323 24 Roos KL, Eckerman NL. The smallpox vaccine and postvaccinal encephalitis. Semin. Neurol. 22 (1), 95–98 (2002).12170398 25 Mo J. Report of committee on smallpox vaccination; investigation of treatment of complications caused by smallpox vaccination. J. Clin. Virol. (Japan) 3 , 269–278 (1975). 26 Rivers TM. Cultivation of vaccine virus for Jennerian prophylaxis in man. J. Exp. Med. 58 , 635–648 (1933).19870221 27 Rivers TM, Ward SM, Baird RD. Amount and duration of immunity induced by intradermal inoculation of cultured vaccine virus. J. Exp. Med. 69 , 857–866 (1939).19870882 28 Wesley RB, Speers WC, Neff JM, Ruben FL, Lourie B. Evaluation of two kinds of smallpox vaccine: CVI-78 and calf lymph vaccine. I. Clinical and serologic response to primary vaccination. Pediatr. Res. 9 (8), 624–628 (1975).1098000 29 Speers WC, Wesley RB, Neff JM, Goldstein J, Lourie B. Evaluation of two kinds of smallpox vaccine: CVI-78 and calf lymph vaccine. II. Clinical and serologic observations of response to revaccination with calf lymph vaccine. Pediatr. Res. 9 (8), 628–632 (1975).1171424 30 Hochstein-Mintzel V, Huber HC, Stickl H. [Virulence and immunogenicity of a modified vaccinia virus (strain MVA)]. Z. Immunitatsforsch. Exp. Klin. Immunol. 144 (2), 104–156 (1972).4282933 31 Hochstein-Mintzel V, Hanichen T, Huber HC, Stickl H. [An attenuated strain of vaccinia virus (MVA). Successful intramuscular immunization against vaccinia and variola]. Zentralbl. Bakteriol. (Orig. A) 230 (3), 283–297 (1975).1146441 32 Hochstein-Mintzel V. [Smallpox vaccine, then and now. From the “cow lymphe” to the cell-culture vaccine]. Fortschr. Med. 95 (2), 79–84 (1977).13033 33 Mayr A. [TC marker of the attenuated vaccinia vaccide strain “MVA” in human cell cultures and protective immunization against orthopox diseases in animals]. Zentralblatt Veterinarmedizin Reihe B 23 (5–6), 417–430 (1976). 34 Munz E, Linckh S, Renner-Muller IC, Reimann M. [The effectiveness of immunization with vaccinia virus type “MVA” against an infection with cowpox virus type “OPV 85” in rabbits]. Zentralblatt Veterinarmedizin Reihe B 40 (2), 131–140 (1993). 35 Werner GT, Jentzsch U, Metzger E, Simon J. Studies on poxvirus infections in irradiated animals. Arch. Virol. 64 (3), 247–256 (1980).6250514 36 Mayr A, Munz E. [Changes in the vaccinia virus through continuing passages in chick embryo fibroblast cultures]. Zentralbl. Bakteriol. (Orig.) 195 (1), 24–35 (1964).5890664 37 Mayr A, Danner K. Vaccination against pox diseases under immunosuppressive conditions. Dev. Biol. Stand. 41 , 225–234 (1978).223909 38 Mayr A, Stickl H, Muller HK, Danner K, Singer H. [The smallpox vaccination strain MVA: marker, genetic structure, experience gained with the parenteral vaccination and behavior in organisms with a debilitated defence mechanism (author’s transl.)]. Zentralbl Bakteriol (Orig. B) 167 (5–6), 375–390 (1978). 39 Mayr A. [Historical review of smallpox, the eradication of smallpox and the attenuated smallpox MVA vaccine]. Berl. Munch. Tierarztl. Wochenschr. 112 (9), 322–328 (1999).10507180 40 Harrer E, Bauerle M, Ferstl B et al. Therapeutic vaccination of HIV-1-infected patients on HAART with a recombinant HIV-1 nef-expressing MVA: safety, immunogenicity and influence on viral load during treatment interruption. Antivir. Ther. (Lond.) 10 (2), 285–300 (2005). 41 Meyer RG, Britten CM, Siepmann U et al. A Phase I vaccination study with tyrosinase in patients with stage II melanoma using recombinant modified vaccinia virus Ankara (MVA-hTyr). Cancer Immunol. Immunother. 54 (5), 453–467 (2005).15627214 42 Earl PL, Americo JL, Wyatt LS et al. Immunogenicity of a highly attenuated MVA smallpox vaccine and protection against monkeypox. Nature 428 (6979), 182–185 (2004).15014500 43 Stittelaar KJ, van Amerongen G, Kondova I et al. Modified vaccinia virus Ankara protects macaques against respiratory challenge with monkeypox virus. J. Virol. 79 (12), 7845–7851 (2005).15919938 44 Meyer H, Sutter G, Mayr A. Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence. J. Gen. Virol. 72 (Pt 5), 1031–1038 (1991).2033387 45 Sutter G, Wyatt LS, Foley PL, Bennink JR, Moss B. A recombinant vector derived from the host range-restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus. Vaccine 12 (11), 1032–1040 (1994).7975844 46 Carroll MW, Moss B. Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia virus: propagation and generation of recombinant viruses in a nonhuman mammalian cell line. Virology 238 (2), 198–211 (1997).9400593 47 Blanchard TJ, Alcami A, Andrea P, Smith GL. Modified vaccinia virus Ankara undergoes limited replication in human cells and lacks several immunomodulatory proteins: implications for use as a human vaccine. J. Gen. Virol. 79 (Pt 5), 1159–1167 (1998).9603331 48 Wyatt LS, Carroll MW, Czerny CP, Merchlinsky M, Sisler JR, Moss B. Marker rescue of the host range restriction defects of modified vaccinia virus Ankara. Virology 251 (2), 334–342 (1998).9837798 49 Wyatt LS, Earl PL, Eller LA, Moss B. Highly attenuated smallpox vaccine protects mice with and without immune deficiencies against pathogenic vaccinia virus challenge. Proc. Natl Acad. Sci. USA 101 (13), 4590–4595 (2004).15070762 50 Antoine G, Scheiflinger F, Dorner F, Falkner FG. The complete genomic sequence of the modified vaccinia Ankara strain: comparison with other orthopoxviruses. Virology 244 (2), 365–396 (1998).9601507 51 Sutter G, Moss B. Nonreplicating vaccinia vector efficiently expresses recombinant genes. Proc. Natl Acad. Sci. USA 89 (22), 10847–10851 (1992).1438287 52 Sutter G, Ramsey-Ewing A, Rosales R, Moss B. Stable expression of the vaccinia virus K1L gene in rabbit cells complements the host range defect of a vaccinia virus mutant. J. Virol. 68 (7), 4109–4116 (1994).8207789 53 Alcami A, Smith GL. A soluble receptor for interleukin-1β encoded by vaccinia virus: a novel mechanism of virus modulation of the host response to infection. Cell 71 (1), 153–167 (1992).1394428 54 Smith GL, Symons JA, Khanna A, Vanderplasschen A, Alcami A. Vaccinia virus immune evasion. Immunol. Rev. 159 , 137–154 (1997).9416508 55 Waibler Z, Anzaghe M, Ludwig H et al. Modified vaccinia virus Ankara induces Toll-like receptor-independent type I interferon responses. J. Virol. 81 (22), 12102–12110 (2007).17855554 56 Alcami A, Symons JA, Smith GL. The vaccinia virus soluble α/β interferon (IFN) receptor binds to the cell surface and protects cells from the antiviral effects of IFN. J. Virol. 74 (23), 11230–11239 (2000).11070021 57 Mahnel H, Mayr A. [Experiences with immunization against orthopox viruses of humans and animals using vaccine strain MVA]. Berl. Munch. Tierarztl. Wochenschr. 107 (8), 253–256 (1994).7945180 58 Mayr A. [Development of of non-immunising, paraspecific vaccine from attenuated pox viruses: a new type of vaccine.]. Berl. Munch. Tierarztl. Wochenschr. 114 (5–6), 184–187 (2001).11413711 59 Timm A, Enzinger C, Felder E, Chaplin P. Genetic stability of recombinant MVA-BN. Vaccine 24 (21), 4618–4621 (2006).16157428 60 Staib C, Suezer Y, Kisling S, Kalinke U, Sutter G. Short-term, but not post-exposure, protection against lethal orthopoxvirus challenge after immunization with modified vaccinia virus Ankara. J. Gen. Virol. 87 (Pt 10), 2917–2921 (2006).16963750 61 Samuelsson C, Hausmann J, Lauterbach H et al. Survival of lethal poxvirus infection in mice depends on TLR9, and therapeutic vaccination provides protection. J. Clin. Invest. 118 (5), 1776–1784 (2008).18398511 62 Ferrier-Rembert A, Drillien R, Meignier B, Garin D, Crance JM. Safety, immunogenicity and protective efficacy in mice of a new cell-cultured Lister smallpox vaccine candidate. Vaccine 25 (49), 8290–8297 (2007).17964011 63 Phelps AL, Gates AJ, Hillier M, Eastaugh L, Ulaeto DO. Comparative efficacy of modified vaccinia Ankara (MVA) as a potential replacement smallpox vaccine. Vaccine 25 (1), 34–42 (2007).16950548 64 Belyakov IM, Earl P, Dzutsev A et al. Shared modes of protection against poxvirus infection by attenuated and conventional smallpox vaccine viruses. Proc. Natl Acad. Sci. USA 100 (16), 9458–9463 (2003).12869693 65 Henderson DA. Smallpox: clinical and epidemiologic features. Emerging Infect. Dis. 5 (4), 537–539 (1999). 66 Stickl H, Hochstein-Mintzel V. [Intracutaneous smallpox vaccination with a weak pathogenic vaccinia virus (“MVA virus”)]. Munch. Med. Wochenschr. 113 (35), 1149–1153 (1971).5109577 67 Stickl H, Hochstein-Mintzel V, Huber HC. [Primary vaccination against smallpox after preliminary vaccination with the attenuated vaccinia virus strain MVA and the use of a new “vaccination stamp”]. Munch. Med. Wochenschr. 115 (35), 1471–1473 (1973).4740817 68 Stickl H, Hochstein-Mintzel V, Mayr A, Huber HC, Schafer H, Holzner A. [MVA vaccination against smallpox: clinical tests with an attenuated live vaccinia virus strain (MVA) (author’s transl)]. Dtsch. Med. Wochenschr. 99 (47), 2386–2392 (1974).4426258 69 Earl PL, Hugin AW, Moss B. Removal of cryptic poxvirus transcription termination signals from the human immunodeficiency virus type 1 envelope gene enhances expression and immunogenicity of a recombinant vaccinia virus. J. Virol. 64 (5), 2448–2451 (1990).2182912 70 Frey SE, Newman FK, Kennedy JS et al. Clinical and immunologic responses to multiple doses of IMVAMUNE (modified vaccinia Ankara) followed by Dryvax challenge. Vaccine 25 (51), 8562–8573 (2007).18036708 71 Vollmar J, Arndtz N, Eckl KM et al. Safety and immunogenicity of IMVAMUNE, a promising candidate as a third generation smallpox vaccine. Vaccine 24 (12), 2065–2070 (2006).16337719 72 Cassimatis DC, Atwood JE, Engler RM, Linz PE, Grabenstein JD, Vernalis MN. Smallpox vaccination and myopericarditis: a clinical review. J. Am. Coll. Cardiol. 43 (9), 1503–1510 (2004).15120802 73 Arness MK, Eckart RE, Love SS et al. Myopericarditis following smallpox vaccination. Am. J. Epidemiol. 160 (7), 642–651 (2004).15383408 74 Eckart RE, Love SS, Atwood JE et al. Incidence and follow-up of inflammatory cardiac complications after smallpox vaccination. J. Am. Coll. Cardiol. 44 (1), 201–205 (2004).15234435 75 Chen RT, Lane JM. Myocarditis: the unexpected return of smallpox vaccine adverse events. Lancet 362 (9393), 1345–1346 (2003).14585633 76 Finlay-Jones LR. Fatal myocarditis after vaccination against smallpox. Report of a case. N. Engl. J. Med. 270 , 41–42 (1964).14062126 77 Monath TP, Caldwell JR, Mundt W et al. ACAM2000 clonal Vero cell culture vaccinia virus (New York City Board of Health strain) – a second-generation smallpox vaccine for biological defense. Int. J. Infect. Dis. 8 (Suppl. 2), S31–S44 (2004).15491873 78 Greenberg RN, Kennedy JS. ACAM2000: a newly licensed cell culture-based live vaccinia smallpox vaccine. Expert Opin. Invest. Drugs 17 (4), 555–564 (2008). 79 Greenberg RN, Kennedy JS, Clanton DJ et al. Safety and immunogenicity of new cell-cultured smallpox vaccine compared with calf-lymph derived vaccine: a blind, single-centre, randomised controlled trial. Lancet 365 (9457), 398–409 (2005).15680454 80 Weltzin R, Liu J, Pugachev KV et al. Clonal vaccinia virus grown in cell culture as a new smallpox vaccine. Nat. Med. 9 (9), 1125–1130 (2003).12925845 81 Kennedy JS, Frey SE, Yan L et al. Induction of human T cell-mediated immune responses after primary and secondary smallpox vaccination. J. Infect. Dis. 190 (7), 1286–1294 (2004).15346340 82 Frey SE, Newman FK, Cruz J et al. Dose-related effects of smallpox vaccine. N. Engl. J. Med. 346 (17), 1275–1280 (2002).11923489 83 Artenstein AW, Johnson C, Marbury TC et al. A novel, cell culture-derived smallpox vaccine in vaccinia-naive adults. Vaccine 23 (25), 3301–3309 (2005).15837236 84 Ober BT, Bruhl P, Schmidt M et al. Immunogenicity and safety of defective vaccinia virus lister: comparison with modified vaccinia virus Ankara. J. Virol. 76 (15), 7713–7723 (2002).12097585 85 Mack TM, Thomas DB, Ali A, Muzaffar Khan M. Epidemiology of smallpox in West Pakistan. I. Acquired immunity and the distribution of disease. Am. J. Epidemiol. 95 (2), 157–168 (1972).5060373 86 Mack TM, Noble J Jr, Thomas DB. A prospective study of serum antibody and protection against smallpox. Am. J. Trop. Med. Hyg. 21 (2), 214–218 (1972).5061278 87 Marennikova SS, Climiskjan KL, Senkman LS, Macevic GR. Some factors determining differences in the antigenicity of vaccinia virus strains. Bull. World Health Organ. 46 (2), 159–163 (1972).4537479 88 Hata TR, Gallo RL. Antimicrobial peptides, skin infections, and atopic dermatitis. Semin. Cutan. Med. Surg. 27 (2), 144–150 (2008).18620136 89 Ferrier-Rembert A, Drillien R, Tournier JN, Garin D, Crance JM. Short- and long-term immunogenicity and protection induced by non-replicating smallpox vaccine candidates in mice and comparison with the traditional 1st generation vaccine. Vaccine 26 (14), 1794–1804 (2008).18336966 90 Earl PL, Americo JL, Wyatt LS et al. Rapid protection in a monkeypox model by a single injection of a replication-deficient vaccinia virus. Proc. Natl Acad. Sci. USA 105 (31), 10889–10894 (2008).18678911 91 Investigator’s Brochure v.10, Imvamune, Bavarian Nordic A/S, Hejreskovvej 10A, DK-3490 Kvistgård, Denmark, 15 January 2008. Patent 101 Chaplin P. Phenotypic and genotypic differences of MVA strains. Modified vaccinia Ankara virus variant. WO/2008/028665 (2006).	19093767	PMC9709931	NO-CC CODE	2022-12-01 23:23:09	no	Expert Rev Vaccines.; 8(1):134-24	utf-8	Expert Rev Vaccines	2,014	10.1586/14760584.8.1.13	oa_other
==== Front Expert Rev Vaccines Expert Rev Vaccines Expert Review of Vaccines 1476-0584 1744-8395 Taylor & Francis 19093767 11218529 10.1586/14760584.8.1.13 Version of Record Research Article Vaccine Profile IMVAMUNE®: modified vaccinia Ankara strain as an attenuated smallpox vaccine IMVAMUNE®, Kennedy & Greenberg Kennedy Jeffrey S Greenberg Richard N Division of Infectious Diseases, Wadsworth Center, NYS Department of Health, Biggs Laboratory, C606, PO Box 509, Albany, NY 12201-0509, USA. [email protected] VA Staff Physician, Lexington VA Medical Center and Professor of Medicine, The Belinda Mason Carden and Paul Mason Professor of HIV/AIDS Research and Education, University of Kentucky School of Medicine, Department of Medicine, Room MN-672, 800 Rose Street, Lexington, KY 40536-0084, USA. [email protected] † Author for correspondence 9 1 2014 2009 8 1 1324 IntegraConverted from TF JATS 1.0 to JATS 1.2 by T&F tfjats-to-jats1.2-converter21 11 2022 © Expert Reviews Ltd 2009 Expert Reviews Ltd Smallpox vaccines based on replicating vaccinia virus are known to elicit rare yet serious adverse events, particularly in human populations with immune deficiency, atopic dermatitis and at the extremes of age. A vaccine that induces protective immune responses equivalent to first-generation smallpox vaccines while reducing the risk for severe adverse events is critical for a national stockpile of smallpox vaccines. Modified vaccinia Ankara (MVA) has been proposed as an immediate solution for vaccination of high-risk individuals. Bavarian Nordic’s vaccine MVA-BN® (IMVAMUNE®) is a MVA strain that is replication incompetent in mammalian cell lines. IMVAMUNE has been administered to more than 1900 human subjects to date, including high-risk populations (e.g., people diagnosed with atopic dermatitis or infected with HIV) in which standard replicating vaccines are contraindicated. We review the Phase I clinical trial safety profile and immune responses and compare them with other smallpox vaccines, including ACAM2000™ and Dryvax®. Keywords: adverse event clinical trial immunity modified vaccinia Ankara smallpox vaccine vaccinia ==== Body pmcIn the context of history, current smallpox vaccine development brings to mind the phrase, ‘déjà vu all over again’. The textbook by Fenner et al. reveals that scientists of 40 years ago, while successful in the eradication of smallpox, left unanswered the same questions we struggle with today: ‘what is protective immunity?’; ‘how long does immunity last?’; and ‘can we develop a safer smallpox vaccine?’ [1]. Despite our ability to understand human immune responses to viruses at an increasingly advanced cellular level, the elements of effective protective immune memory to viruses are often not well defined. For infectious diseases such as variola, where endemic disease is nonexistent, or others where field study is impractical, translation of immunological biomarkers to vaccine development remains a less than certain art. In this article, we review the current clinical development program by Bavarian Nordic A/S to develop a new strain of modified vaccinia Ankara (MVA) as a safer alternative to first-generation replicating smallpox vaccines such as Dryvax® and the second-generation vaccine ACAM2000™. Several vaccinia strains were widely used prior to the end of the compulsory smallpox vaccination program. In 1967, the WHO smallpox eradication unit surveyed 59 laboratories producing vaccinia for inoculation of humans: 51 harvested from the skin of calves, six from other bovines, three from the chorioallantoic membrane of chick embryo and three from tissue culture using bovine embryonic fibroblasts [1]. The strains included the Lister-Elstree strain (Lister Institute, UK; 39% of laboratories), New York City Board of Health (NYCBOH) (12%) and the Paris strain (12%), with the remaining laboratories using a variety of other strains (37%), including poorly characterized strains or a mixture of vaccinia and cowpox viruses [1]. Biological properties of a given vaccinia strain were known to be influenced by passage methodology but most of these vaccines had clinical efficacy against variola major. In 1963, the WHO established criteria and standards for the manufacture of vaccinia-derived smallpox vaccines, essentially reducing the derivatives of vaccinia used to three strains: Lister-Elstree, EM63 (Moscow Research Institute of Viral Preparation, Russia) and NYCBOH [2]. The WHO standards required a vaccine to produce major skin reactions in 95% of primary vaccinees and in 90% of those vaccinated 10 or more years previously, using an inoculum virus titer of 108 plaque-forming units per milliliter [1]. The three aforementioned strains of vaccinia virus were considered equivalent in protecting against smallpox during outbreaks, and provided some protection when administered within 3–4 days of exposure [3]. Although highly effective immunizing agents, each of these vaccines produced significant adverse events [4–8]. The smallpox vaccines in current global stockpiles are derived from these historical vaccinia strains. The strains differ in their pathogenicity in animal models [9,10] and, although never shown to be statistically correlated with adverse events, differences were noted in the incidence of adverse events in humans across strains [11,12]. These differences in safety profiles may also reflect changes in vaccination policy. Earlier vaccination of newborns was replaced in the late 1950s with vaccination after 1 year of age. This change significantly reduced serious adverse complications. Therefore, the exact comparator safety profile expected for a modern vaccine remains uncertain based on a number of population concerns, including prevalence of diseases that predispose risk (atopic dermatitis [AD] and immunosuppression), vaccinia strain, method used to inoculate, environmental risk factors and underlying health for those who are currently immunologically naive to vaccinia. These concerns may alter predicted frequencies of certain adverse events based on historical data [12–15]. Recent experience during the smallpox vaccination campaign demonstrated that the frequency of most adverse events that were anticipated based on historical data were lower than expected [14,16]. By contrast, myopericarditis appeared more frequently than anticipated based on historical age-based data [14,17–20]. Whether differences in the age of those typically receiving their first inoculation with vaccinia, children (<5 years historically) versus the current vaccination programs (aged > 18 years), have a role in the risk for myopericarditis is not known [15,19,21–23]. Clinical trials to date with ACAM2000 vaccine have not demonstrated definitive conclusions as to whether this serious adverse event (SAE) is more problematic with this second-generation smallpox vaccine. Vaccinia strains may differ in the frequency of adverse events elicited following inoculation [12]. NYCBOH-derived strains (as compared with other replicating strains) have historically been associated with less pathogenicity and a safety profile marked by fewer adverse events, such as postvaccinial encephalitis (PVE) deaths or resulting permanent neurological disability [24]. PVE is a complication that is not associated with excessive viral replication and, therefore, could occur as a result of immunization with IMVAMUNE®. A review of experience in humans comparing several different strains revealed that moderate-to-high pathogenic strains (Lister and Ikeda) produced more neurological events (including non-PVE) than low pathogenic strains (LC16m8, CVI and EM63), despite eliciting equivalent skin-take rates and seroconversion [24,25]. Although this secondary level of evidence suggested that less-pathogenic strains could provide improved safety without compromising efficacy, the incidence of SAEs, particularly PVE, were too rare to prove statistically significant differences between vaccinia strains. Current assessment of PVE would require clinical trials of the vaccine in millions of recipients in order to assess true differences between vaccine strains. The attenuated vaccines date back to at least 1931, when an investigator at the Rockefeller Institute (NY, USA) described serial passage of the NYCBOH strain in chick embryo cells [26]. Two strains, CVI-78 and CVII, the latter carried for a total of 235 passages over chick embryo explants, were shown to be effective with reduced adverse events by intradermal inoculation in humans [27]. The CVII vaccine was tested in 60,000 Netherland’s Army recruits and had less local and systemic reactogenicity than other replicating vaccines. However, CVII vaccination resulted in lower neutralizing antibody titers compared with a Lister-inoculated comparator group, and led to the conclusion that, without subsequent challenge with lymph-derived vaccine and induction of a skin-pock lesion, CVII would be ineffective [3]. CVI-78 was resurrected in the early 1970s by the National Institute of Allergy and Infectious Diseases, US NIH, in a study that compared four vaccines: calf lymph- and egg-derived NYCBOH, CVI-78 and Lister. The results again demonstrated that, regardless of the route of administration, induction of neutralizing antibody titers for attenuated CVI-78 were less compared with Lister or NYCBOH (75 vs 93–96% of subjects) and, upon standard challenge vaccination, the CVI-78 recipients were more likely to have primary skin responses, an indicator of nonprotective immunity [3,28,29]. Several other attenuated strains, including Darian-derived, temperature-sensitive strains (LC16m8) and G-9, an attenuated variant of the highly pathogenic vaccinia stain Temple of Heaven, were tested in human studies. For the most part, there were improved immune responses compared with prior results observed with CVI strains, although detailed data on G-9 have never been published. Modified vaccinia Ankara Modified vaccinia Ankara originates from the dermal vaccinia strain Ankara (chorioallantois vaccinia virus Ankara [CVA]) derived from a pox lesion from a horse in Ankara, Turkey, and maintained in the Vaccination Institution Ankara [30–32]. Attenuated by over 570 continuous passages in primary chick embryo fibroblast (CEF) cells, MVA was used to protect livestock against orthopoxviruses [33,34] and, later, tested as a pre-immunization to primary vaccination in human populations at risk of complications when using first-generation smallpox vaccines. Early indications that MVA could provide a safer immunogenic smallpox vaccine came from preclinical studies performed in irradiated, immunocompromised rabbits that demonstrated generation of neutralizing antibodies and protection against vaccinia virus challenge [35]. The reduction in CVA virulence through passage on CEF was reported at passage 371 [36] and the strain was renamed MVA after passage 516 in order to distinguish it from other attenuated vaccinia virus strains [33]. Early clinical development of MVA as a pre-smallpox vaccine in Europe used MVA-517 (corresponding to the 517th passage) as a priming vaccine, followed by vaccination with Lister-Elstree. These studies included subjects at risk of adverse reactions from vaccinia. In 1976, MVA derived from the MVA-571 seed stock (corresponding to the 571st passage) was tested in approximately 120,000 individuals without any reports of SAEs. Several high-risk groups were vaccinated, including young children with skin conditions [37–39]. Rigorous follow-up of those vaccinated was not performed in that setting, as is currently the standard for investigational vaccine protocols. IMVAMUNE is derived from a seed stock that corresponds to the 597th passage in CEF cells and has undergone several more rounds of limiting dilution compared with the historical MVA strain used in Germany. This MVA strain has multiple cloning sites and has been tested in humans as an experimental recombinant HIV and cancer vaccine at doses up to five-times higher than those used with the IMVAMUNE smallpox vaccination [40,41]. Bavarian Nordic has conducted an extensive preclinical development program that has demonstrated the safety, efficacy and bioequivalence of IMVAMUNE to first-generation smallpox vaccines (e.g., Dryvax and Lister-Elstree). The preclinical program has included a unique challenge model of intratracheal monkeypox and two murine orthopoxvirus challenge models. Human testing is currently ongoing, but data available to date, from Phase I and II studies, suggest that the vaccine can elicit both T- and B-cell immune responses, and the implications of how this compares to replicating vaccinia vaccines are discussed later [42,43]. IMVAMUNE properties IMVAMUNE possesses interesting properties for a new generation of attenuated smallpox vaccines. IMVAMUNE differs from other MVA strains in that it has undergone extensive plaque purification in CEF cells, is propagated in serum-free conditions and, after passage 570, is a genetically stable virus [44–46]. The attenuation characteristics of this MVA separate it from other MVA strains [101]. Most MVA strains used in research are polyclonal and contain virus subtypes that are genetically similar to a replicating vaccinia strain. These subtype viruses, if virus host-range genes (e.g., K1L) are present and thus enable mammalian cell replication, could result in the emergence of phenotypes identical to replicating vaccinia. Culture replicating subtypes can theoretically emerge in vivo as replicating vaccinia virus, particularly when injected into immunocompromised animals [46,47]. However, rescue studies on host-range defects [48] and injection of MVA at 1000-times the lethal dose of vaccinia failed to induce death in severe combined immunodeficiency (SCID) mice [49]. Although it can be speculated that human use of such MVA variants can lead to the isolation of these virulent strains in vivo, to date this has not been observed in preclinical animal testing or human clinical trials involving multiple MVA-based vaccines. IMVAMUNE is replication deficient in mammalian cells and, in comparison to direct ancestral MVA strains, produces an acceptable safety profile in severely immune compromised animals [101]. Comparison of the genomes of IMVAMUNE and the ancestral vaccinia Ankara strain CVA demonstrates a loss of 15% of the genetic information (∼31,000 bp of DNA), resulting from six major deletions in the MVA genome. Deletion of the host-range gene K1L and deletions of other virulence and host-range genes, including the gene for type A inclusion bodies, may partially explain the lack of replication in mammalian cell lines of IMVAMUNE [44,46,50]. Interestingly, even though IMVAMUNE exhibits attenuated replication in mammalian cells, efficient transcription of the viral proteins occurs within host cells and the functional result of the deletions appears to block virus assembly and egress [45,46,51,52]. This latter finding may have effects on comparative immune responses to first- or second-generation vaccines in which vaccinia virus assembly and egress occurs. However, notwithstanding the significant effect of genome deletion on attenuation and reduced virulence, preclinical studies using IMVAMUNE have demonstrated the ability to induce humoral and cellular immune responses to both vaccinia genes as well as cloned target proteins. Human cells infected with replicating vaccinia respond by induction of type I interferons (IFNs) and the expression of a soluble IL-1 receptor [47], a potential virulence factor for certain poxviruses [53,54]. Viral induction of type I IFNs and the expression of the receptor may lead to interference with human immune responses through molecular mimicry of cytokines and immune receptors. MVA does not express soluble receptors for IFN-γ, IFN-α/β, TNF or CC chemokines [50], which could influence comparative effects on innate immune responses [55] or the induction of stable long-term memory responses [47,56]. Manufacturing, potency & administration of vaccine The formulation of IMVAMUNE used in clinical trials is a purified live vaccine produced under serum-free conditions in CEF cells, supplied as a sterile liquid-frozen preparation and does not contain adjuvants or preservatives. In preparing the vaccine, virus cell suspensions are homogenized by a process that includes a freeze and thaw, Benzonase® and ultrasound treatment. The bulk vaccine is adjusted to a titer of 2 × 108 tissue culture infectious dose (TCID50)/ml, filled into 2-ml injection vials and stored below -15°C as a frozen liquid product. Prior to administering the vaccine, the vial is thawed at room temperature and gently swirled for at least 30 s. The injection volume is 0.5 ml (1.0 × 108 TCID50) per dose and is administered as a subcutaneous injection. IMVAMUNE can also be effectively administered via intramuscular injection. Stability tests are continuing but current preliminary data support stability for at least 2 years at -15°C. In reference to historical smallpox vaccine guidelines, the WHO’s efforts to develop freeze-dried vaccines (Dryvax, Lister and EM-63) were partially in response to the instability of vaccinia strains in liquid form during extended exposure to climate level temperatures. Long-term stability of IMVAMUNE is under evaluation. In addition, the vaccine, once thawed, must be administered immediately within 12 h; this further distinguishes it from Dryvax and ACAM2000, since both have extended post-thawing half-lives of 1–2 weeks. IMVAMUNE is not stable to refreeze once thawed and would require cold-chain logistics during an epidemic. Summaries of nonclinical studies The studies conducted for evaluating the safety prior to initiation of human trials included repeat administrations (subcutaneous and intramuscular) in animal models that demonstrated reversible nondose-limiting injection-site reactions and lymphoid changes. Injection-site expression levels of vaccinia by RT-PCR revealed only one weakly positive reaction at injection sites beyond day 3 (at day 7), suggesting no persistence of virus or genetic incorporation. Teratology studies in both rats and rabbits did not demonstrate teratogenic or intrauterine toxicity, and peri- and postnatal studies did not reveal any toxicity to embryos or developing offspring at doses up to 1 × 108 TCID50. Three preclinical evaluations provide the primary support for safety and efficacy. To assess the attenuation profile of different strains of MVA, their growth was compared on CEF cells and across a number of mammalian cell lines. The strains included IMVAMUNE (MVA-597), MVA-572 (a 1970s smallpox vaccine used in Germany [57]), MVA-Vero (a MVA-575 strain adapted to Vero cell line [58]) and MVA-I721 (a MVA strain deposited at the Institute Pasteur [accession number I-721]. All four strains replicated in primary CEF cells but demonstrated significant differences in the ability to replicate in mammalian cells, with IMVAMUNE showing the near absence of replication across monkey (CV-1), rabbit (RK-13), murine (AG-101) and human (143-B, HeLa) mammalian cell lines [101]. These results suggest that MVA variants may exist in the MVA-572 and MVA-1721 parental lots. This conclusion is supported by studies inoculating immune-deficient mice (AGR129 strain), which led to in vivoreplication and the emergence of altered genetic variants from these MVA strains but not from IMVAMUNE [59]. These results challenge the perception that all MVA strains are similar and suggest that lack of human cell replication, as observed with IMVAMUNE, represents a more stable clonal virus that could be uniquely suited to a human attenuated smallpox vaccine. The second line of evidence for clinical attenuation was demonstrated using an immune-compromised mouse model inoculated with IMVAMUNE. AGR129 mice have gene deletions that inactivate the function of the IFN system (IFN receptor types I [α/β] and type II [IFN-γ]) and inactivate the production of IFN-γ on a Rag-/- gene-deletion background), which prevents the formation of mature B and T cells. To evaluate IMVAMUNE, seven strains of vaccinia differing in virulence (IMVAMUNE, MVA-572, MVA-Vero, MVA-I721, smallpox vaccines Dryvax and Lister-Elstree, and a highly lethal vaccinia strain Western Reserve [WR]) were administered intraperitoneally at 1 × 107 TCID50. AGR129 mice that were inoculated with IMVAMUNE survived for more than 120 days with no replicating virus isolated from the ovaries (a marker of virus dissemination) at any time point. Animals inoculated with other MVA strains died or demonstrated virus-mediated disease at an average time to disease of 20 days (MVA-I721), 80 days (MVA-Vero) and 81 days (MVA-572) [101]. Recent reports of MVA providing protection against otherwise lethal poxvirus infection in Toll-like receptor-9-deficient mice suggests that MVA recognition and induction of immunity can occur through both Toll-like receptor-9-dependent and -independent pathways [60,61]. Replicating vaccinia strains Dryvax, Lister-Elstree and WR killed all mice within 8 days of inoculation. The combined immunodeficiency mouse model provides evidence that MVA strains differ in their preclinical safety profiles and there is lack of mammalian cell replication with IMVAMUNE. Preclinical efficacy was tested using two unique challenge models, employing both the Ectromelia virus (ECTV) and monkeypox virus (MPXV) [43]. Earlier animal studies demonstrated that IMVAMUNE induces an equivalent humoral and T-cell immune responses in mice compared with first-generation smallpox vaccines (Dryvax and Elstree) [42,62,63]. In the ECTV challenge model, a single dose of IMVAMUNE protected mice from intranasal challenge with a lethal dose of VV-WR or ECTV that was given 6 weeks after vaccination. Mice challenged 3 days after vaccination with WR virus were also protected, consistent with previous studies using first-generation vaccines and variola or vaccinia challenge [64]. In the nonhuman primate model of intratracheal delivery of MPXV, challenged animals vaccinated with IMVAMUNE were completely protected, with the exception that one animal developed skinpox lesions. Viral loads were even more diminished if IMVAMUNE was followed by conventional inoculation with the Elstree vaccine prior to the challenge. The overall clinical response to vaccination with IMVAMUNE was equivalent to first-generation smallpox vaccines [43]. These challenge studies provide important efficacy data in that the MPXV model mimics natural variola infection, producing a smallpox-like disease (i.e., fever, skin lesions and fibrinonecrotic bronchopneumonia); although 100% of placebo-treated animals die [43], which is in contrast with the lower mortality in human smallpox [65]. Clinical experience with IMVAMUNE IMVAMUNE, as a third-generation smallpox vaccine for special populations at risk of SAEs from replicating vaccinia, must demonstrate a safety profile that succeeds at eliminating common complications associated with replicating vaccinia-based vaccines. For instance, accidental spread of virus to close contacts is not probable with IMVAMUNE due to the lack of replication in human cells and intramuscular inoculation eliminating the development of skinpock lesion. The occurrence of eczema vaccinatum and progressive vaccinia also seem unlikely but these assessments will require continued monitoring in clinical trials and beyond. Assessment of the occurrence of PVE will also require long-term monitoring. The rarity of both PVE and myopericarditis requires large numbers of volunteers to receive IMVAMUNE in order to achieve statistically superior safety assessment over first- or second-generation vaccines. Overall, any comparison trial would require a treatment arm receiving a replicating vaccine. Use of a replicating vaccinia vaccine as part of a trial design to compare adverse events is not acceptable for these special populations. Historical data using the MVA vaccine in over 120,000 individuals during the 1970s provide encouraging evidence that MVA as a smallpox vaccine can be safely administered to humans. In 1976, MVA was authorized in Germany as a preimmunization vaccine in combination with Lister vaccine. This two-step inoculation program was thought to diminish potential adverse events, particularly PVE. More than 120,000 subjects received intradermal and subcutaneous injections with a low dose of MVA (1 × 106 TCID50) before the mass-vaccination program was discontinued [37]. Prior clinical trials described the safety and tolerability of the vaccine and served as the basis for licensure in Germany [66–68]. These studies included data on 7098 subjects (5691 aged <3 years and 1407 aged >3 years). Reactions at the injection site revealed no blistering, pustules or ulcerations, and no cases of PVE or other SAEs were observed. Reactogenicity was limited to approximately 4% with mild fever (>38°C; 2.28%) and approximately 4% with nonspecific systemic symptoms. However, detailed immunogenicity was never sufficiently tested. Administering MVA prior to first-generation smallpox vaccine reduced the frequency of adverse events from the replication-competent vaccinia virus. By contrast, with IMVAMUNE, single-dose vaccination with this earlier CEF-derived MVA elicited only a weak hemagglutinin inhibition and virus-neutralizing antibody titers. MVA prevaccination resulted in a strong booster effect after vaccination with a replicating vaccinia strain, indicating that priming with MVA induced specific humoral and celllar immune responses. Human safety of IMVAMUNE To date, more than 2000 subjects have been vaccinated with either IMVAMUNE or recombinant MVA-BN®-based vaccines, including at-risk populations with AD or HIV infection. Tables 1 & 2 provide an overview of ongoing and completed IMVAMUNE clinical studies. The use of MVA-BN-based vectored vaccines in clinical trials is not reviewed here, since these vaccines represent altered viruses and, therefore, comparison and interpretation of the safety and tolerability data can be complex and problematic. In summary, however, administration of MVA-BN-vectored vaccines in HIV-infected subjects has not been associated with dissemination or any severe systemic or neurological events attributable to vaccinia. Adverse vaccine reactions Clinical trials of IMVAMUNE in subjects over 18 years of age have included HIV-infected subjects and individuals with a history of AD, or active AD. The frequency of reactions occurring in the 1025 subjects enrolled in the completed IMVAMUNE clinical trials is shown in Table 3. Summaries to date suggest that the majority of the events observed in these trials have been classified as being mild to moderate and resolving without sequelae. Reported local and systemic reactions are consistent with prior observations with MVA [69]. Of particular interest are adverse events occurring in special populations with AD (POX-MVA-007 and -008) and HIV-1 infection (POX-MVA-010 and -011), which are two populations contraindicated for first- and second-generation smallpox vaccines. POX-MVA-007 was a Phase I clinical trial that used two injections of IMVAMUNE in adults aged 18–40 years. A total of 60 subjects were enrolled: 15 patients with mildly active AD and 16 patients with a history of AD were compared with healthy subjects (n = 15) and subjects with allergic rhinitis (n = 14). No SAEs or premature study withdrawal events have been reported. Mild-to-moderate transient pain and redness at the inoculation site occurred in nearly all subjects and no differences in the incidence of grade 3 and above adverse reactions during the immediate postvaccination period were observed between AD and healthy controls. A larger open-label comparative (AD vs no AD) trial, POX-MVA-008, is currently ongoing and will recruit up to 530 individuals aged between 18 and 40 years in the USA and Mexico. In trials POX-MVA-010 and -011, the objectives were to compare safety and immunogenicity data in HIV-infected subjects with uninfected subjects. Study POX-MVA-010 included 91 subjects with HIV (30 naive and 61 experienced to vaccinia). To date, there have been no reports of SAEs in the published Phase I studies [70,71]. The company will soon publish complete safety data from the completed larger cohort studies (Chaplin P, Pers. Comm.), which limits their summary in this report. A recombinant BN-MVA containing the HIV nef protein construct has been administered to 14 HIV-infected patients (CD4 > 400/µl), with only mild systemic reactions noted [40]. Special interest adverse events (cardiac events) Due to recent clinical trials with ACAM2000 and Dryvax and First Responder Vaccination Programs employing Dryvax, there is additional concern regarding the association of vaccinia inoculation and the development of myopericarditis after vaccination [17,72–74]. Myocarditis and pericarditis as a sequelae of viral infection has, for some viruses, been shown to be autoimmune related and generally occurs in later teen years or young adulthood [22,75,76]. Whether the incidence of vaccinia-related myopericarditis is indeed more frequent than recognized in the past as a consequence of changing vaccination patterns using older subjects is subject to debate. Recently reported incidence rates for developing myopericarditis following vaccination with the first-generation smallpox vaccine Dryvax (10.38 events per 1000 vaccinations) and a second-generation replicating vaccinia vaccine ACAM2000 (5.73 events per 1000 vaccinations) have been a surprise and are alarming [77,78]. Historical data related to myopericarditis must also be interpreted carefully as those studies were prior to the clinical use of echocardiograms and cardiac-specific enzymes. Adverse events from published clinical trials using IMVAMUNE and MVA-BN-based recombinant vaccines have not reported a case of myopericarditis. Further evidence supporting the cardiac safety of IMVAMUNE was obtained from a placebo-controlled Phase II study in healthy individuals (POX-MVA-005), which intensively monitored the risk of myopericarditis developing after vaccination with IMVAMUNE. There have been no cases of myopericarditis reported for this study (Chaplin P, Pers. Comm.). Immune response to IMVAMUNE in healthy individuals In order to delineate what is known to date regarding IMVAMUNE immune responses, current and completed clinical trials can be broken down into vaccinia-naive or -experienced subjects with or without either AD or HIV. POX-MVA-001, -002, -004 & -005 These four trials represent the initial IMVAMUNE trials in adults. Tables 1 & 2 depicts the design of each of these trials in terms of groups, dose tested, prior vaccinia status and numbers tested. POX-MVA-002 was independently conducted by the National Institute of Allergy and Infectious Diseases; Bavarian Nordic sponsored the others. Tables 1 & 2 presents an incomplete summary of the humoral and cellular responses reported to date. We are only able to report complete data for two trials: POX-MVA-001 and -002. Historically, a plaque-reduction neutralization titer (PRNT) greater than 1:40 is considered a positive humoral response. In Table 4, comparative IFN-γ-ELISPOT (>15 spots/million peripheral blood mononuclear cells [PBMCs] considered positive) using a standardized assay across all vaccines depicted offers an assessment of the T-cell responses observed with IMVAMUNE compared with that observed with first-generation Dryvax, experimental cell-cultured smallpox vaccine and ACAM2000 [78–82]. This assay was employed across clinical trials of smallpox vaccines, utilizing the same reagents, assay controls and methods, thus enabling comparison of IFN-γ T-cell responses across vaccine candidates [70,79–82]. In the POX-MVA-001 trial, 86 male subjects aged between 18 and 55 years were vaccinated and stratified based on the presence or absence of a prior history of smallpox vaccination. The first four dosing cohorts were naive to smallpox vaccination and received the vaccine on days 0 and 28. A fifth cohort of subjects, who had been previously vaccinated against smallpox, received a single subcutaneous dose of IMVAMUNE at 1 × 108 TCID50. Maximum seroconversion rates by ELISA, defined as titers of at least 100, were reached at day 42, with 100% seroconversion after the 1 × 108 TCID50 dose administered by subcutaneous or intramuscular injection. The analysis of variance (ANOVA) indicated a significant superiority of the 1 × 108 and 1 × 107 TCID50 doses over the lower doses. In the group of previously vaccinated subjects, all showed a rapid booster response and 100% seroconversion by total IgG ELISA [71]. In the POX-MVA-002 trial, high seroconversion rates were observed for all doses after the second vaccination (87–100%). In the POX-MVA-002 trial, measurement of neutralizing antibodies (PRNT assay) did not demonstrate a 100% seroconversion rate with the highest dose of IMVAMUNE, although a dose-dependent increase in neutralizing (PRNT) antibody responses was observed. Since complete data from the larger POX-MVA-004 trial has been published, detailed analysis of the antibody PRNT and ELISA values can be made particularly in comparison to those observed with Dryvax and ACAM2000 or in the POX-MVA-002 trial, which directly compared IMVAMUNE with Dryvax. These early results, from POX-MVA-001 and POX-MVA-002, raise the interesting question as to whether some degree of replication may be an important determinant for neutralizing antibodies [70,80–84]. This premise is supported by the work of others demonstrating that various forms (extracellular) of vaccinia virus are typically generated after infection of human cells. These forms are important in the generation of hemagglutinin and neutralizing antibodies to the virus, which may be critical early in protective immunity [85–87]. Further work is needed to clarify this issue. A peak CMI response was reached approximatley 2 weeks after the second vaccination (day 42) for groups 1–4 in the POX-MVA-001 trial. While a cell-mediated response was detected in five subjects with pre-immunity on day 0 (group 5), those subjects showed a booster response on day 28 following a single vaccination with IMVAMUNE, with a mean T-cell count of 109 cells per million PBMCs. This would indicate that IMVAMUNE was able to stimulate the memory T-cell response induced by a previous replicating smallpox vaccination. In vaccinia-naive subjects, dose responses were observed, with better responses being recorded using the higher doses of IMVAMUNE. At the highest dose (groups 3 and 4), the cell-mediated responses measured on day 42 were in the same range (102 and 152) as subjects with pre-existing immunity (group 5) on day 28. A correlation analysis showed that the T-cell response measured in the ELISPOT was highly correlated to the results of the PRNT and ELISA antibody tests (data not shown) [70,71]. These Phase I results confirm dose-finding data from the study POX-MVA-004 (see later), which demonstrated that a dose of 1 × 108 TCID50 IMVAMUNE was well tolerated and the most immunogenic dose evaluated in healthy subjects [70]. Two larger multicenter Phase II clinical trials in subjects with AD (POX-MVA-008) or HIV-1 infection (POX-MVA-011) are underway to evaluate the immunogenicity and safety of two doses of 1 × 108 TCID50 of IMVAMUNE subcutaneously, given 1 month apart. The primary study objective of the POX-MVA-008 trial is to assess the humoral immune response (measured by ELISA) induced by IMVAMUNE in subjects with AD and, in the POX-MVA-011 trial, safety of the vaccine in HIV-infected subjects. Data on the POX-MVA-011 trial is expected later this year. Expert commentary Derived from MVA, IMVAMUNE is a highly attenuated strain of vaccinia that is unable to replicate in human cells and, therefore, cannot be transmitted or cause dispersed vaccinia infection. The extensive nonclinical development has shown the vaccine to induce protective immunity in two important animal challenge studies: immune-compromised mice and primates challenged with MPXV. The issue of durable immunogenicity, both humoral and cellular, requiring the need for repeat inoculations to boost responses, raises some concerns and challenges to develop a best-use strategy during an outbreak scenario. In addition, patients with AD inoculated with vaccinia represent a unique population with altered skin host-defense mechanisms [88], and how these alterations might affect immunogenicity of IMVAMUNE remains to be determined. Finally, MVA as a vaccine strain does not produce the typical skinpock lesion, which eliminates a useful marker of successful vaccination for field assessment of protection during outbreaks. Recently, antibody profiling across species (including humans) for MVA, Dryvax and WR strains demonstrated that binding antibodies to select structural vaccinia proteins were similar across species [53]. Although this study has some limitations with respect to the panel of vaccinia proteins selected, the data demonstrate remarkable similarity of binding profiles, suggesting that the 3% genetic variation observed between MVA and Dryvax may not significantly alter the repertoire of humoral immune response. However, the deletion mutations of specific proteins in IMVAMUNE could alter the magnitude of protective levels of humoral or cellular immunity, the ability to neutralize variola major and the durability of response. These issues will require further study focusing on the site-directed nature of the neutralizing antibody response. Of concern remains the lower neutralizing geometric mean titer in MVA-vaccinated individuals and whether this reflects altered response to early vaccinia antigens, which others have suggested may be directed to specific vaccinial proteins: B5R, D8, A27, D13 and A14 [54,55]. Development of third-generation smallpox vaccines that target the key antigens required for neutralization of vaccinia forms intracellular mature virus and extracellular mature virus are new candidates entering human testing and may provide a unique boost combination with IMVAMUNE. Such boosting may ultimately contribute to more durable neutralizing antibody titers and T-cell memory. Of course, all of the previous commentary on immunogenicity and durability really depends on IMVAMUNE protection against variola; only studies or experience with variola will reveal this answer. So far, there is the expectation that IMVAMUNE will be effective. Five-year view To date, more than 1900 subjects have been vaccinated with IMVAMUNE and recombinant MVA-BN-based vaccines; no cases of myopericarditis have been observed. Therefore, these early clinical studies suggest that IMVAMUNE may offer a safer cardiac profile, that is, have a much lower rate of cardiac events compared with Dryvax and ACAM2000. Whether the unique absence of replication in human cells translates to an improved safety profile remains to be determined by expanded clinical testing and usage. In addition, the immunogenicity data generated to date suggest that both humoral and cellular immune responses are lower, particularly following one dose. Future considerations with regards to IMVAMUNE include recent development and testing of smallpox vaccines based on DNA plasmids and recombinant protein vaccines, enabling higher antigenic content and use of adjuvants to boost neutralizing antibodies to targeted vaccinia proteins. The role of IMVAMUNE may emerge as a prevaccination, which could be subsequently boosted with newer vaccine designs. Although the limited data thus far suggest two doses of IMVAMUNE induces sufficient antibody and cellular immune responses in the immediate vaccination period, the decay of the response compared with replicating smallpox vaccines needs further study to support IMVAMUNE use as a vaccine that would provide 3–5 years of protection. This concern is supported by recent data comparing three nonreplicating smallpox vaccines in mice that failed to demonstrate long-term (150 days postvaccination) protection against intranasal challenge with cowpox virus [89]. Perhaps a schedule that includes periodic IMVAMUNE booster vaccinations will solve this issue. On the other hand, in a recent nonhuman primate study, MVA did prevent infection with vaccination 4 days prior to a monkeypox challenge, while Dryvax required vaccination 6 days prior to challenge to be effective [60,90]. These primate-challenge models emphasize differences between MVA and replicating vaccinia smallpox vaccines that need to be considered in the strategies for human use in the pre- and postexposure settings. Finally, we believe that the IMVAMUNE program is well guided, providing the necessary information needed to evaluate the vaccine for US FDA approval for special populations in an emergency. 10.1586/14760584.8.1.13-T0001 Table 1. Summary of completed human trials of IMVAMUNE® attenuated smallpox vaccine. Study Population Dose (route) Dose of IMVAMUNE n GMT* (% positive) nAb‡ (% positive) POX-MVA-001§ Vaccinia naive 106 (sc.) 2 18 39 33 107 (sc.) 2 16 81 50 108 (sc.) 2 16 100 80 108 (im.) 2 18 100 87 Nonvaccinia naive 108 (sc.) 1 18 100 89 POX-MVA-002§ Vaccinia naive 2 × 107 (sc.) 2 + Dryvax® 15 100 100 5 × 107 (sc.) 2 + Dryvax 15 100 93 1 × 108 (sc.) 2 + Dryvax 15 100 87 Placebo × 2 + Dryvax 15 84 85 1 × 108 (sc.) 2 + placebo 15 100 92 1 × 108 (im.) 2 + Dryvax 15 100 100 Measuring total anti-MVA IgG by ELISA, percentage responding 14 days after last dose of IMVAMUNE or 14 days after Dryvax alone. ‡Conversion percentages are based on plaque-reduction neutralization titer GMT titer approximately 30 days after last dose. §See clinical trial [71,72]. GMT: Geometric mean titer; im.: Intramuscular; MVA: Modified vaccinia Ankara; NA: Data not available in peer-review published or reported form; nAb: Neutralizing antibody; sc.: Subcutaneous. 10.1586/14760584.8.1.13-T0002 Table 2. Study design of on-going studies for which immunogenicity data is not available. Study Population Dose (route) Doses n POX-MVA-004 Vaccinia-naive 2 × 107 (sc.) 2 55 5 × 107 (sc.) 2 55 108 (sc.) 2 55 POX-MVA-005 Vaccinia-naive 108 (sc.) 2 183 108 (sc.) 1 + placebo 181 108 (sc.) Placebo 181 Nonvaccinia-naive 108 (sc.) 1 200 POX-MVA-007 Vaccinia-naive 108 (sc.) 2 15 History of AD 108 (sc.) 2 16 Mild active AD 108 (sc.) 2 15 Mild allergic rhinitis 108 (sc.) 2 14 POX-MVA-008 Vaccinia-naive 108 (sc.) 2 230 AD, vaccinia-naive 108 (sc.) 2 300 POX-MVA-010 HIV-infected naive 108 (sc.) 2 30 HIV-infected non-naive 108 (sc.) 1 61 Naive 108 (sc.) 2 30 Non-naive 108 (sc.) 1 30 POX-MVA-011 Naive 108 (sc.) 2 90 HIV+ CD4 200–750 108 (sc.) 2 360 AD: Atopic dermatitis; MVA: Modified vaccinia Ankara; sc.: Subcutaneous. 10.1586/14760584.8.1.13-T0003 Table 3. Comparison of IMVAMUNE®/ACAM2000™/Dryvax® adverse events (possibly related) occurring in at least 5% of trial subjects. Adverse event characterization Vaccinia-naive subjects Preferred term IMVAMUNE (n = 1025) (n; %) ACAM2000 (n = 873) (n; %)‡ Dryvax (n = 289) (n; %)‡ Blood and the lymphatic system disorders Lymphadenopathy 13 (1) 72 (8) 35 (12) Lymph node pain 1 (0.1) 494 (57) 199 (69) Nervous system disorders Headache 280 (27) 433 (50) 150 (52) Respiratory, thoracic and mediastinal disorders Dyspnea 0 (0) 39 (4) 16 (6) Gastrointestinal disorders Nausea 105 (10) 170 (19) 65 (22) Diarrhea 8 (1) 144 (16) 34 (12) Constipation 0 (0) 49 (6) 9 (3) Vomiting 1 (0.1) 42 (5) 10 (3) Skin and subcutaneous tissue disorders Erythema 1 (0.1) 190 (22) 69 (24) Rash 3 (0.3) 94 (11) 30 (10) Musculoskeletal, connective tissue and bone disorders Myalgia 103 (10) 404 (46) 147 (51) General disorders and administration-site conditions Injection-site erythema 827 (81) 649 (74) 229 (79) Injection-site pain 887 (87) 582 (67) 208 (72) Injection-site pruritus 211 (21) 804 (92) 277 (96) Injection-site swelling 692 (68) 422 (48) 165 (57) Fatigue 316 (31) 423 (48) 161 (56) Malaise 5 (0.5) 327 (37) 122 (42) Rigors 31 (3) 185 (21) 66 (23) Exercise tolerance decreased 0 (0) 98 (11) 35 (12) Feeling hot 1 (0.1) 276 (32) 97 (34) Summary of published and unpublished data from incompleted and reported IMVAMUNE clinical trials. ‡Prescribing information for ACAM2000, August 2007. Modified from [91]. 10.1586/14760584.8.1.13-T0004 Table 4. Comparison of cell-mediated immunity induction for smallpox vaccines using a standardized IFN-γ-ELISPOT assay in human trials in naive subjects. Vaccine name and study n Vaccine titer Route (+) ELISPOT (%) Mean SFCs per 106 PBMC 30 days post first dose Mean SFCs per 106PBMC 30 days post second dose Dryvax® ‡ 90 1 × 108 PFU/ml id. 98 402 NA IMVAMUNE® 21 1 × 108 TCID50 im. 95 100 286 ACAM2000™ 30 1 × 108 PFU/ml id. 99 442 NA ACAM1000 30 1 × 108 PFU/ml id. 100 331 NA CCSV 40 1 × 108 PFU/ml id. 99 251 NA Controls§ 10 Naive controls 0 5.8 4.1 Variation of response was 0–15 SFC for a given well run in triplicate. Value is the mean across five studies in the table, except for 30 days post-second dose, where data are mean for IMVAMUNE study alone. *Studies cited are IMVAMUNE [71], ACAM1000 [81], CCSV [80] and ACAM2000. ‡Dryvax® (n) data: all subjects comparatively injected with Dryvax in the studies of experimental vaccines in the table. §Nonimmunized controls were used in the assay. One donor was used for every three plates. Summary data is for all studies and represents repeated measures for ten donors. CCSV: Cell-cultured smallpox vaccine; id.: Intradermal by bifurcated needle; im.: Intramuscular injection; NA: Not assessed (only one dose administered); PBMC: Peripheral blood mononuclear cell; PFU: Plaque-forming unit; SFC: Spot-forming cell; TCID: Tissue culture infectious dose. Key issues • IMVAMUNE®, an attenuated vaccinia strain, has been shown not to replicate in human cell lines, reducing the risk of transmissibility to close contacts – a key new safety advantage. • The vaccine is a modern cell culture-derived vaccine, developed for use in special populations at risk of complications from replicating vaccinia-based smallpox vaccines. • In 2007, a report of Phase I studies comparing immunogenicity of IMVAMUNE to Dryvax® revealed that at dose of 1 × 108 TCID50 IMVAMUNE could elicit robust humoral and cellular immune responses, albeit lower than Dryvax. The duration of protective antibody and T-cell responses are unclear and requires further evaluation in both low- and high-risk populations. • Numerous registration trials in special at-risk populations (atopic dermatitis and HIV-1 infected) are ongoing or planned, and key data from these trials are expected in 2009. • To date, in limited small trials, IMVAMUNE has been well tolerated and no myocardial events have been reported. • Cold-chain and two-dose requirements with respect to IMVAMUNE may result in logistic problems during outbreak scenarios. • The important unfilled niche in the supply of smallpox vaccine in the event of a bioterrorism act is to protect populations for which replicating vaccinia-based vaccines are contraindicated. IMVAMUNE based on replication-deficient modified vaccina Ankara provides an attenuated smallpox vaccine for these special populations. Such a vaccine should strive for single-dose immunogenicity comparable to first-generation vaccines, a reasonable duration of protective immunity, as well as a superior safety profile. As additional safety and immunogenicity data from IMVAMUNE clinical trials become available, the capabilities and role for IMVAMUNE will be resolved. Acknowledgements The authors would like to thank P Chaplin, A von Krempelhuber, N Arndtz, G Virgin and the scientists at Bavarian Nordic for providing the IMVAMUNE ® data included in Table 3 and the cited comments that shed light on trial data that will be published later this year. Financial & competing interests disclosure R Greenberg has performed clinical trials for Bavarian Nordic. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript. ==== Refs References 1 Fenner F, Henderson DA, Arita I, Jezek Z, Ladnyi ID. Smallpox and its Eradication. History of International Public Health. World Health Organization, Geneva, Switzerland (1988). 2 Krag P, Bentzon MW. The international reference preparation of smallpox vaccine. An international collaborative assay. Bull. World Health Organ. 29 , 299–309 (1963).14058224 3 Meltzer MI. Risks and benefits of preexposure and postexposure smallpox vaccination. Emerging Infect. Dis. 9 (11), 1363–1370 (2003). 4 Galasso GJ, Mattheis MJ, Cherry JD et al. Clinical and serologic study of four smallpox vaccines comparing variations of dose and route of administration. J. Infect. Dis. 135 (1), 183–186 (1977).833448 5 Cherry JD, Connor JD, McIntosh K et al. Clinical and serologic study of four smallpox vaccines comparing variations of dose and route of administration. Standard percutaneous revaccination of children who receive primary subcutaneous vaccination. J. Infect. Dis. 135 (1), 176–182 (1977).188953 6 Connor JD, McIntosh K, Cherry JD et al. Clinical and serologic study of four smallpox vaccines comparing variations of dose and route of administration. Primary subcutaneous vaccination. J. Infect. Dis. 135 (1), 167–175 (1977).833447 7 Vaccinia (smallpox) vaccine. Recommendations of the Immunization Practices Advisory Committee (ACIP). MMWR Recomm. Rep. 40 (RR-14), 1–10 (1991). 8 Cherry JD, McIntosh K, Connor JD et al. Clinical and serologic study of four smallpox vaccines comparing variations of dose and route of administration. Primary percutaneous vaccination. J. Infect. Dis. 135 (1), 145–154 (1977).188951 9 Morita M, Aoyama Y, Arita M et al. Comparative studies of several vaccinia virus strains by intrathalamic inoculation into cynomolgus monkeys. Arch. Virol. 53 (3), 197–208 (1977).404993 10 Morita M, Arita M, Komatsu T, Amano H, Hashizume S. A comparison of neurovirulence of vaccinia virus by intrathalamic and/or intracisternal inoculations into cynomolgus monkeys. Microbiol. Immunol. 21 (7), 417–418 (1977).409906 11 Marennikova SS, Maltseva NN. [Comparative study of some strains of vaccinia virus. II. Pathogenicity for laboratory animals]. Vopr. Virusol. 126 , 287–291 (1964).14237276 12 Kretzschmar M, Wallinga J, Teunis P, Xing S, Mikolajczyk R. Frequency of adverse events after vaccination with different vaccinia strains. PLoS Med. 3 (8), E272 (2006).16933957 13 Fulginiti VA, Papier A, Lane JM, Neff JM, Henderson DA. Smallpox vaccination: a review, part II. Adverse events. Clin. Infect. Dis. 37 (2), 251–271 (2003).12856218 14 Kemper AR, Davis MM, Freed GL. Expected adverse events in a mass smallpox vaccination campaign. Eff. Clin. Pract. 5 (2), 84–90 (2002).11990216 15 Neff JM, Lane JM, Pert JH, Moore R, Millar JD, Henderson DA. Complications of smallpox vaccination. I. National survey in the United States, 1963. N. Engl. J. Med. 276 (3), 125–132 (1967).4381041 16 Poland GA, Grabenstein JD, Neff JM. The US smallpox vaccination program: a review of a large modern era smallpox vaccination implementation program. Vaccine 23 (17–18), 2078–2081 (2005).15755574 17 Halsell JS, Riddle JR, Atwood JE et al. Myopericarditis following smallpox vaccination among vaccinia-naive US military personnel. JAMA 289 (24), 3283–3289 (2003).12824210 18 Morgan J, Roper MH, Sperling L et al. Myocarditis, pericarditis, and dilated cardiomyopathy after smallpox vaccination among civilians in the United States, January–October 2003. Clin. Infect. Dis. 46 (Suppl. 3), S242–S250 (2008).18284365 19 Lane JM, Ruben FL, Neff JM, Millar JD. Complications of smallpox vaccination, 1968: results of ten statewide surveys. J. Infect. Dis. 122 (4), 303–309 (1970).4396189 20 Matthews AW, Griffiths ID. Post-vaccinial pericarditis and myocarditis. Br. Heart J. 36 (10), 1043–1045 (1974).4279685 21 Bengtsson E, Holmgren A, Nystrom B. Smallpox outbreak and vaccination problems in Stockholm, Sweden 1963. Circulatory studies in patients with abnormal ECG in the course of postvaccinal complications. Acta Med. Scand. 464 , 113–126 (1966). 22 Bengtsson E, Hansson S, Nystrom B. Smallpox outbreak and vaccination problems in Stockholm, Sweden 1963. V. Postvaccinal reactions and complications. Acta Med. Scand. 464 , 87–104 (1966). 23 Karjalainen J, Heikkila J, Nieminen MS et al. Etiology of mild acute infectious myocarditis. Relation to clinical features. Acta Med. Scand. 213 (1), 65–73 (1983).6829323 24 Roos KL, Eckerman NL. The smallpox vaccine and postvaccinal encephalitis. Semin. Neurol. 22 (1), 95–98 (2002).12170398 25 Mo J. Report of committee on smallpox vaccination; investigation of treatment of complications caused by smallpox vaccination. J. Clin. Virol. (Japan) 3 , 269–278 (1975). 26 Rivers TM. Cultivation of vaccine virus for Jennerian prophylaxis in man. J. Exp. Med. 58 , 635–648 (1933).19870221 27 Rivers TM, Ward SM, Baird RD. Amount and duration of immunity induced by intradermal inoculation of cultured vaccine virus. J. Exp. Med. 69 , 857–866 (1939).19870882 28 Wesley RB, Speers WC, Neff JM, Ruben FL, Lourie B. Evaluation of two kinds of smallpox vaccine: CVI-78 and calf lymph vaccine. I. Clinical and serologic response to primary vaccination. Pediatr. Res. 9 (8), 624–628 (1975).1098000 29 Speers WC, Wesley RB, Neff JM, Goldstein J, Lourie B. Evaluation of two kinds of smallpox vaccine: CVI-78 and calf lymph vaccine. II. Clinical and serologic observations of response to revaccination with calf lymph vaccine. Pediatr. Res. 9 (8), 628–632 (1975).1171424 30 Hochstein-Mintzel V, Huber HC, Stickl H. [Virulence and immunogenicity of a modified vaccinia virus (strain MVA)]. Z. Immunitatsforsch. Exp. Klin. Immunol. 144 (2), 104–156 (1972).4282933 31 Hochstein-Mintzel V, Hanichen T, Huber HC, Stickl H. [An attenuated strain of vaccinia virus (MVA). Successful intramuscular immunization against vaccinia and variola]. Zentralbl. Bakteriol. (Orig. A) 230 (3), 283–297 (1975).1146441 32 Hochstein-Mintzel V. [Smallpox vaccine, then and now. From the “cow lymphe” to the cell-culture vaccine]. Fortschr. Med. 95 (2), 79–84 (1977).13033 33 Mayr A. [TC marker of the attenuated vaccinia vaccide strain “MVA” in human cell cultures and protective immunization against orthopox diseases in animals]. Zentralblatt Veterinarmedizin Reihe B 23 (5–6), 417–430 (1976). 34 Munz E, Linckh S, Renner-Muller IC, Reimann M. [The effectiveness of immunization with vaccinia virus type “MVA” against an infection with cowpox virus type “OPV 85” in rabbits]. Zentralblatt Veterinarmedizin Reihe B 40 (2), 131–140 (1993). 35 Werner GT, Jentzsch U, Metzger E, Simon J. Studies on poxvirus infections in irradiated animals. Arch. Virol. 64 (3), 247–256 (1980).6250514 36 Mayr A, Munz E. [Changes in the vaccinia virus through continuing passages in chick embryo fibroblast cultures]. Zentralbl. Bakteriol. (Orig.) 195 (1), 24–35 (1964).5890664 37 Mayr A, Danner K. Vaccination against pox diseases under immunosuppressive conditions. Dev. Biol. Stand. 41 , 225–234 (1978).223909 38 Mayr A, Stickl H, Muller HK, Danner K, Singer H. [The smallpox vaccination strain MVA: marker, genetic structure, experience gained with the parenteral vaccination and behavior in organisms with a debilitated defence mechanism (author’s transl.)]. Zentralbl Bakteriol (Orig. B) 167 (5–6), 375–390 (1978). 39 Mayr A. [Historical review of smallpox, the eradication of smallpox and the attenuated smallpox MVA vaccine]. Berl. Munch. Tierarztl. Wochenschr. 112 (9), 322–328 (1999).10507180 40 Harrer E, Bauerle M, Ferstl B et al. Therapeutic vaccination of HIV-1-infected patients on HAART with a recombinant HIV-1 nef-expressing MVA: safety, immunogenicity and influence on viral load during treatment interruption. Antivir. Ther. (Lond.) 10 (2), 285–300 (2005). 41 Meyer RG, Britten CM, Siepmann U et al. A Phase I vaccination study with tyrosinase in patients with stage II melanoma using recombinant modified vaccinia virus Ankara (MVA-hTyr). Cancer Immunol. Immunother. 54 (5), 453–467 (2005).15627214 42 Earl PL, Americo JL, Wyatt LS et al. Immunogenicity of a highly attenuated MVA smallpox vaccine and protection against monkeypox. Nature 428 (6979), 182–185 (2004).15014500 43 Stittelaar KJ, van Amerongen G, Kondova I et al. Modified vaccinia virus Ankara protects macaques against respiratory challenge with monkeypox virus. J. Virol. 79 (12), 7845–7851 (2005).15919938 44 Meyer H, Sutter G, Mayr A. Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence. J. Gen. Virol. 72 (Pt 5), 1031–1038 (1991).2033387 45 Sutter G, Wyatt LS, Foley PL, Bennink JR, Moss B. A recombinant vector derived from the host range-restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus. Vaccine 12 (11), 1032–1040 (1994).7975844 46 Carroll MW, Moss B. Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia virus: propagation and generation of recombinant viruses in a nonhuman mammalian cell line. Virology 238 (2), 198–211 (1997).9400593 47 Blanchard TJ, Alcami A, Andrea P, Smith GL. Modified vaccinia virus Ankara undergoes limited replication in human cells and lacks several immunomodulatory proteins: implications for use as a human vaccine. J. Gen. Virol. 79 (Pt 5), 1159–1167 (1998).9603331 48 Wyatt LS, Carroll MW, Czerny CP, Merchlinsky M, Sisler JR, Moss B. Marker rescue of the host range restriction defects of modified vaccinia virus Ankara. Virology 251 (2), 334–342 (1998).9837798 49 Wyatt LS, Earl PL, Eller LA, Moss B. Highly attenuated smallpox vaccine protects mice with and without immune deficiencies against pathogenic vaccinia virus challenge. Proc. Natl Acad. Sci. USA 101 (13), 4590–4595 (2004).15070762 50 Antoine G, Scheiflinger F, Dorner F, Falkner FG. The complete genomic sequence of the modified vaccinia Ankara strain: comparison with other orthopoxviruses. Virology 244 (2), 365–396 (1998).9601507 51 Sutter G, Moss B. Nonreplicating vaccinia vector efficiently expresses recombinant genes. Proc. Natl Acad. Sci. USA 89 (22), 10847–10851 (1992).1438287 52 Sutter G, Ramsey-Ewing A, Rosales R, Moss B. Stable expression of the vaccinia virus K1L gene in rabbit cells complements the host range defect of a vaccinia virus mutant. J. Virol. 68 (7), 4109–4116 (1994).8207789 53 Alcami A, Smith GL. A soluble receptor for interleukin-1β encoded by vaccinia virus: a novel mechanism of virus modulation of the host response to infection. Cell 71 (1), 153–167 (1992).1394428 54 Smith GL, Symons JA, Khanna A, Vanderplasschen A, Alcami A. Vaccinia virus immune evasion. Immunol. Rev. 159 , 137–154 (1997).9416508 55 Waibler Z, Anzaghe M, Ludwig H et al. Modified vaccinia virus Ankara induces Toll-like receptor-independent type I interferon responses. J. Virol. 81 (22), 12102–12110 (2007).17855554 56 Alcami A, Symons JA, Smith GL. The vaccinia virus soluble α/β interferon (IFN) receptor binds to the cell surface and protects cells from the antiviral effects of IFN. J. Virol. 74 (23), 11230–11239 (2000).11070021 57 Mahnel H, Mayr A. [Experiences with immunization against orthopox viruses of humans and animals using vaccine strain MVA]. Berl. Munch. Tierarztl. Wochenschr. 107 (8), 253–256 (1994).7945180 58 Mayr A. [Development of of non-immunising, paraspecific vaccine from attenuated pox viruses: a new type of vaccine.]. Berl. Munch. Tierarztl. Wochenschr. 114 (5–6), 184–187 (2001).11413711 59 Timm A, Enzinger C, Felder E, Chaplin P. Genetic stability of recombinant MVA-BN. Vaccine 24 (21), 4618–4621 (2006).16157428 60 Staib C, Suezer Y, Kisling S, Kalinke U, Sutter G. Short-term, but not post-exposure, protection against lethal orthopoxvirus challenge after immunization with modified vaccinia virus Ankara. J. Gen. Virol. 87 (Pt 10), 2917–2921 (2006).16963750 61 Samuelsson C, Hausmann J, Lauterbach H et al. Survival of lethal poxvirus infection in mice depends on TLR9, and therapeutic vaccination provides protection. J. Clin. Invest. 118 (5), 1776–1784 (2008).18398511 62 Ferrier-Rembert A, Drillien R, Meignier B, Garin D, Crance JM. Safety, immunogenicity and protective efficacy in mice of a new cell-cultured Lister smallpox vaccine candidate. Vaccine 25 (49), 8290–8297 (2007).17964011 63 Phelps AL, Gates AJ, Hillier M, Eastaugh L, Ulaeto DO. Comparative efficacy of modified vaccinia Ankara (MVA) as a potential replacement smallpox vaccine. Vaccine 25 (1), 34–42 (2007).16950548 64 Belyakov IM, Earl P, Dzutsev A et al. Shared modes of protection against poxvirus infection by attenuated and conventional smallpox vaccine viruses. Proc. Natl Acad. Sci. USA 100 (16), 9458–9463 (2003).12869693 65 Henderson DA. Smallpox: clinical and epidemiologic features. Emerging Infect. Dis. 5 (4), 537–539 (1999). 66 Stickl H, Hochstein-Mintzel V. [Intracutaneous smallpox vaccination with a weak pathogenic vaccinia virus (“MVA virus”)]. Munch. Med. Wochenschr. 113 (35), 1149–1153 (1971).5109577 67 Stickl H, Hochstein-Mintzel V, Huber HC. [Primary vaccination against smallpox after preliminary vaccination with the attenuated vaccinia virus strain MVA and the use of a new “vaccination stamp”]. Munch. Med. Wochenschr. 115 (35), 1471–1473 (1973).4740817 68 Stickl H, Hochstein-Mintzel V, Mayr A, Huber HC, Schafer H, Holzner A. [MVA vaccination against smallpox: clinical tests with an attenuated live vaccinia virus strain (MVA) (author’s transl)]. Dtsch. Med. Wochenschr. 99 (47), 2386–2392 (1974).4426258 69 Earl PL, Hugin AW, Moss B. Removal of cryptic poxvirus transcription termination signals from the human immunodeficiency virus type 1 envelope gene enhances expression and immunogenicity of a recombinant vaccinia virus. J. Virol. 64 (5), 2448–2451 (1990).2182912 70 Frey SE, Newman FK, Kennedy JS et al. Clinical and immunologic responses to multiple doses of IMVAMUNE (modified vaccinia Ankara) followed by Dryvax challenge. Vaccine 25 (51), 8562–8573 (2007).18036708 71 Vollmar J, Arndtz N, Eckl KM et al. Safety and immunogenicity of IMVAMUNE, a promising candidate as a third generation smallpox vaccine. Vaccine 24 (12), 2065–2070 (2006).16337719 72 Cassimatis DC, Atwood JE, Engler RM, Linz PE, Grabenstein JD, Vernalis MN. Smallpox vaccination and myopericarditis: a clinical review. J. Am. Coll. Cardiol. 43 (9), 1503–1510 (2004).15120802 73 Arness MK, Eckart RE, Love SS et al. Myopericarditis following smallpox vaccination. Am. J. Epidemiol. 160 (7), 642–651 (2004).15383408 74 Eckart RE, Love SS, Atwood JE et al. Incidence and follow-up of inflammatory cardiac complications after smallpox vaccination. J. Am. Coll. Cardiol. 44 (1), 201–205 (2004).15234435 75 Chen RT, Lane JM. Myocarditis: the unexpected return of smallpox vaccine adverse events. Lancet 362 (9393), 1345–1346 (2003).14585633 76 Finlay-Jones LR. Fatal myocarditis after vaccination against smallpox. Report of a case. N. Engl. J. Med. 270 , 41–42 (1964).14062126 77 Monath TP, Caldwell JR, Mundt W et al. ACAM2000 clonal Vero cell culture vaccinia virus (New York City Board of Health strain) – a second-generation smallpox vaccine for biological defense. Int. J. Infect. Dis. 8 (Suppl. 2), S31–S44 (2004).15491873 78 Greenberg RN, Kennedy JS. ACAM2000: a newly licensed cell culture-based live vaccinia smallpox vaccine. Expert Opin. Invest. Drugs 17 (4), 555–564 (2008). 79 Greenberg RN, Kennedy JS, Clanton DJ et al. Safety and immunogenicity of new cell-cultured smallpox vaccine compared with calf-lymph derived vaccine: a blind, single-centre, randomised controlled trial. Lancet 365 (9457), 398–409 (2005).15680454 80 Weltzin R, Liu J, Pugachev KV et al. Clonal vaccinia virus grown in cell culture as a new smallpox vaccine. Nat. Med. 9 (9), 1125–1130 (2003).12925845 81 Kennedy JS, Frey SE, Yan L et al. Induction of human T cell-mediated immune responses after primary and secondary smallpox vaccination. J. Infect. Dis. 190 (7), 1286–1294 (2004).15346340 82 Frey SE, Newman FK, Cruz J et al. Dose-related effects of smallpox vaccine. N. Engl. J. Med. 346 (17), 1275–1280 (2002).11923489 83 Artenstein AW, Johnson C, Marbury TC et al. A novel, cell culture-derived smallpox vaccine in vaccinia-naive adults. Vaccine 23 (25), 3301–3309 (2005).15837236 84 Ober BT, Bruhl P, Schmidt M et al. Immunogenicity and safety of defective vaccinia virus lister: comparison with modified vaccinia virus Ankara. J. Virol. 76 (15), 7713–7723 (2002).12097585 85 Mack TM, Thomas DB, Ali A, Muzaffar Khan M. Epidemiology of smallpox in West Pakistan. I. Acquired immunity and the distribution of disease. Am. J. Epidemiol. 95 (2), 157–168 (1972).5060373 86 Mack TM, Noble J Jr, Thomas DB. A prospective study of serum antibody and protection against smallpox. Am. J. Trop. Med. Hyg. 21 (2), 214–218 (1972).5061278 87 Marennikova SS, Climiskjan KL, Senkman LS, Macevic GR. Some factors determining differences in the antigenicity of vaccinia virus strains. Bull. World Health Organ. 46 (2), 159–163 (1972).4537479 88 Hata TR, Gallo RL. Antimicrobial peptides, skin infections, and atopic dermatitis. Semin. Cutan. Med. Surg. 27 (2), 144–150 (2008).18620136 89 Ferrier-Rembert A, Drillien R, Tournier JN, Garin D, Crance JM. Short- and long-term immunogenicity and protection induced by non-replicating smallpox vaccine candidates in mice and comparison with the traditional 1st generation vaccine. Vaccine 26 (14), 1794–1804 (2008).18336966 90 Earl PL, Americo JL, Wyatt LS et al. Rapid protection in a monkeypox model by a single injection of a replication-deficient vaccinia virus. Proc. Natl Acad. Sci. USA 105 (31), 10889–10894 (2008).18678911 91 Investigator’s Brochure v.10, Imvamune, Bavarian Nordic A/S, Hejreskovvej 10A, DK-3490 Kvistgård, Denmark, 15 January 2008. Patent 101 Chaplin P. Phenotypic and genotypic differences of MVA strains. Modified vaccinia Ankara virus variant. WO/2008/028665 (2006).	36465857	PMC9710096	NO-CC CODE	2022-12-16 23:20:04	no	iScience. 2023 Jan 20; 26(1):105696	latin-1	iScience	2,022	10.1016/j.isci.2022.105696	oa_other
==== Front IJID Reg IJID Reg IJID Regions 2772-7076 Published by Elsevier Ltd on behalf of International Society for Infectious Diseases. S2772-7076(22)00146-1 10.1016/j.ijregi.2022.11.013 Article The impact of earlier reopening to travel in the Western Pacific on SARS-CoV-2 transmission Jin Shihui ab Lim Jue Tao ac Dickens Borame Lee a Cook Alex R ab⁎ a Saw Swee Hock School of Public Health, National University of Singapore and National University Health System, 117549, Singapore b Department of Statistics and Data Science, National University of Singapore, 117546, Singapore c Lee Kong Chian School of Medicine, Nanyang Technological University, 308232, Singapore ⁎ Corresponding author: Ms. Alex R Cook, National University of Singapore, Department of Statistics and Data Science, #10-01 Tahir Foundation Building, 12 Science Drive 2, 117549, Singapore 30 11 2022 30 11 2022 © 2022 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. Background The COVID-19 pandemic has led to a fall of over 70% in international travel, resulting in substantial economic damages. The impact is especially pronounced in the Asia-Pacific region, where governments have been slow to relax border restrictions. Methods For eight Asia-Pacific countries or regions, we utilized a retrospective approach to construct notional epidemic trajectories from June to November 2021 under hypothetical scenarios of earlier resumption of international travel and selective border-reopening. We calculated number of local infections and deaths over the prediction window accordingly. Results Had quarantine-free entry been permitted for all travellers from all the regions investigated and travel volumes recovered to the 2019 levels, Australia, New Zealand and Singapore would have been the three most severely affected regions, with at least doubled number of deaths, while infections would have increased marginally (<5%) for Japan, Malaysia and Thailand. Conclusions Earlier resumption of travel in Asia-Pacific, while maintaining a controlled degree of importation risk, could have been implemented through selective border-reopening strategies and on-arrival testing. Once countries had experienced large, localized COVID-19 outbreaks, earlier relaxation of border containment measures would not have resulted in a great increase in morbidity and mortality. Keywords Border measures Quarantine SARS-CoV-2 Testing Travel restrictions ==== Body pmc1 Introduction Globally, COVID-19 has had a substantial influence on cross-border travel, reducing the volume of international tourist arrivals to 30% of the pre-pandemic level in 2020 and 2021 (UNWTO, 2022a). The plunge can largely be explained by travel restrictions implemented from early 2020 to reduce disease importation. Such border-control measures have substantially lowered importation risks in settings where they have been carefully implemented (Chinazzi et al., 2020), most notably in the upper-middle and high income countries of Asia-Pacific, where border control was coupled with measures to control local spread until vaccines became available, thereby keeping mortality rates low (Miyawaki and Tsugawa, 2022). The economic and social consequences of travel bans have, nevertheless, been extensive. The drastic reduction in air traffic led to an estimated loss of 370 billion USD for airlines in 2020, of which 32% was borne by Asia-Pacific countries, and losses of over 100 billion USD for airports and air navigation service providers (Asian Development Bank, 2021), affecting millions of jobs in aviation around the globe (Iacus et al., 2020). Tourism was also adversely affected, with the top ten source markets, making up nearly half of all international travelers, advising against non-essential travel abroad (UNWTO, 2022b). Countries in Asia-Pacific, the initial pandemic epicenter, have had different policies for border control, but frequently these have restricted who may enter the country, with many foreigners being prohibited from entry, and travelers being subject to forms of quarantine, pre-departure and post arrival tests (Australian Government Department of Home Affairs, 2022; Ministry of Health NZ, 2022; Ullah et al., 2021). By the end of November 2021, before the global prevalence of the variant B.1.1.529 (henceforth, the Omicron variant), two thirds of destinations in this region had still completely closed their borders to international tourists (UNWTO, 2021). Such blanket travel bans were, however, not supported by the WHO, even following identification of more transmissible variants of concern (WHO, 2021a). As vaccination programs were rolled out, policies to resume intra-region travel were introduced. Singapore, for example, launched several Vaccinated Travel Lanes (Civil Aviation Authority of Singapore, 2022) in late 2021 to allow vaccinated visitors from countries or territories with low transmission risks to enter without quarantine if they provided a negative on-arrival test. Similar quarantine-free policies were also adopted by Australia and New Zealand for travelers between these two countries (Australian Government Department of Health and Aged Care, 2021). The expansion of international travel in the Asia-Pacific has nevertheless been much more cautious than in other parts of the world, even as autochthonous transmission of more transmissible variants became widespread. The question therefore arises whether the travel restrictions that were effected in the Asia-Pacific early in the course of the pandemic, and concomitant economic and social costs, were retained for an excessive length of time, and if so the degree of relaxation that would have maintained low disease prevalence. In this study, utilizing the estimated infection potential of SARS-CoV-2 based on prevailing within-country control policies, we quantified the impact of easing border control on local transmission in a basket of countries in the Western Pacific, exploring hypothetical scenarios and assessing the feasibility of earlier reopening. 2 Materials and methods 2.1 Sources of Data We focused on eight countries or territories (henceforth, regions) in the Western Pacific: Australia (AU), Japan (JP), the Republic of Korea (KR), Malaysia (MY), New Zealand (NZ), Singapore (SG), Thailand (TH) and Taiwan, China (TW). These eight were chosen because they are all upper or upper middle income regions that imposed border restrictions, including to each other, despite in many cases having little autochthonous transmission of COVID-19. We considered the period June to November 2021, when the delta variant dominated but before the omicron variant emerged (Freunde von GISAID e.V., 2022; Pulliam et al., 2022). Daily counts of COVID-19 cases and deaths were obtained from either the John Hopkins University Covid-19 database (Dong et al., 2020) or authoritative websites of the eight regions (Figure S1). We collected monthly incoming visitor data and population data from official statistical databases of the eight regions (Figure S2), whence we calculated the connectivity from region A to B at time t, ct,AB, as the (average) proportion of residents of A that arrived at B in day t (see supplementary information). Sources of the case counts, international arrival and population data are also listed in the supplementary information (Table S1–S3). 2.2 Statistical Analysis 2.2.1 Estimation of real-time infection potential We characterized the real-time infection potential using the time-varying effective reproduction number Rt, defined as the ratio of new cases generated by the total infectiousness of infected cases at time t (Cori et al., 2013), using the EpiFilter algorithm (details in the supplement) (Parag, 2021). This was assumed not to change under other scenarios. 2.2.2 Simulating new incidence curves To model infection sizes under scenarios in which travel restrictions were eased, we constructed notional epidemic curves using different connectivity matrices derived from the travel data, assuming the transmission potential of imported cases were the same as that of local cases. Let Yt,i and Y˜t,i be the observed and notional case counts for region i at time t respectively. To account for the possible effects of herd immunity, if Rt,i was the estimated effective reproduction number for region i at time t, the corresponding effective reproduction number R˜t,i for the notional scenario wasR˜t,i=Ni2021−r×∑s=1t−1Y˜s,iNi2021−r×∑s=1t−1Ys,i×Rt,i, where Ni2021 is the population size of region i in 2021, and r is a scaling factor accounting for under-reporting of COVID-19 infections. We set r to be 1.6, the ratio between the increase in seroprevalence and reported percentage infected from May to December 2021 in the US (Dong et al., 2020; Jones et al., 2022). We additionally conducted a sensitivity analysis with different r values involved in the model, but found that different rs had a limited impact on the final infection size (Figure S5, Table S5). Following Cook's method of constructing the effects of alternative intervention strategies (Cook et al., 2008), we assumed the same stochasticity for both actual and notional epidemic trajectories, by fixing the cumulative distribution functions for each datum. If qt,i is the probability that case counts in region i on day t is no larger than Yt,i, i.e.,qt,i=Ft,i(Yt,i)=Pt,i(Xt,i≤Yt,i), where Ft,i(·) is the cumulative density function for the random variable Xti that follows a Poisson distribution with parameter (mean and variance) λt,i=Rt,i∑d=1pwdYt−d,i and w1:p is the corresponding serial interval distribution, the number of new cases reported on day t for region i in the notional outbreak isY˜t,i=F˜t,i−1(qt,i). In this, F˜t,i(·) is the cumulative density function for the Poisson distribution with parameter λ˜t,i=R˜t,i∑d=1pwd(Y˜t−d,i+∑t1=t−dt−1Z˜t−d,i,t1), while R˜t,i is the time-varying effective reproduction number at time t for region i in the hypothetical scenario, and Z˜t,i,t1=(Ct1,·i)T·Y˜t=∑j=1nct1,ji×Y˜t,j the expected number in the notional outbreak of foreign cases infected at time t and imported to region i at time t1≥t. We let Y˜t=(Y˜t,1,Y˜t,2,⋯,Y˜t,n)T be the vector of differenced notional case counts for the selected regions at the t, and Ct=(ct,ij) be the time-varying connectivity matrix with zero principal diagonal elements and (i,j)th element (i≠j) representing the proportion of residents in location i who visited location j (j≠i) at time t, quantifying the travel flow from region i to region j at that time. Owing to the discreteness of Poisson distributions, we let the solution x=F−1(q) be the smallest integer such that P(X≤x)≥q. Assuming an average of 21 days from notification to death (Shim et al., 2022), we obtained the region-specific, time-varying case fatality rate, CFRt,i, by smoothing the ratio of reported deaths on day (t+21) to reported case counts on day t for region i, using the mgcv R package (Wood, 2022). The expected additional deaths for region i that would have resulted under the alternative scenarios was calculated as∑tCFRt,i×(Y˜t,i−Yt,i) which we convert to additional deaths per million population for comparison between regions. 2.2.3 Scenarios Using the travel data in 2021, we considered the baseline scenario that visitors travelling among these eight regions were able to move freely in local communities upon their arrival, i.e., not be subject to quarantine, and estimated the impact on the number of local cases from June 1 to November 30, 2021. Visitor and population data in 2019 were then used to model how COVID-19 transmission would be if intra-regional travel flows recovered to pre-pandemic levels, assumed to match the corresponding period of 2019. For this, we considered different travel-resumption dates (the first day of each month from June to November). We then considered the case when each region selectively opened its border only to some of the other seven regions. If region i were open to region j, all visitors from j would be allowed to enter i without any movement restrictions upon arrival, while travel restrictions would remain unchanged for visitors from the regions i were not yet open to (i.e., they would still be subject to quarantine and have negligible effects on local transmission). For each region, we started with the lowest risk region they could open to (see supplement) and simulated the infections from June to November 2021, then successively expanded the set of regions to which their borders were open. We assumed that increasing infections under these scenarios did not cause secondary effects on other regions, since the effects of increased travel transpired to be relatively small. Finally, we considered three strategies with varying degrees of resumption of quarantine-free entry for visitors from the eight regions:S1. without an entry test; S2. following a negative rapid antigen test (RAT); and S3. following a negative polymerase chain reaction (PCR) test. We assumed the false negative rates for RAT and PCR test, averaged over the infectious period, were 40% and 10% respectively (Chadwick et al., 2022; Kanji et al., 2021). Namely, the expected number of imported cases that could affect future local incidence curve was 100%, 40% and 10% of the total imported cases. The simulation interval for all the three strategies was from June to November 2021. 3 Results Estimated Rts from June to November 2021 for the eight regions are presented in Figure S3. 3.1 Impact of cancelling movement restrictions for foreign visitors on local cases Without travel restrictions, the epidemic trajectory would have been fundamentally altered by imported cases for New Zealand, where an epidemic wave with at least 1,000 daily infections would have been expected in November 2021 and the total case count over the 6 months’ time would be 11.5 (95% credible interval [CrI]: 8.0–16.7) times the observed value. Australia and Singapore would also have experienced larger outbreaks from August and September to November 2021, with over 500 additional local infections per day, and 1.4 (95% CrI: 1.3–1.5) times the overall number of cases by November. The other regions, by comparison, would not have been so adversely affected by importation, which would have brought fewer than 6,000 additional cases in total for each region (Figure 1 , Figure S6).Figure 1 Local cases averted by the border containment measures in 2021. Comparison between daily case counts with (‘observed’, dots in blue for reported case counts) and without (‘notional’, in red with line for mean and shade for 95% CrI) implementation of travel restrictions (i.e., quarantine requirement for foreign visitors after their arrival) from June to November 2021, assuming the number of foreign visitors were the same as that in the real situation. For presentation purposes, observed incidence is presented for every third day. Figure 1 3.2 Cases increased due to ease of travel restrictions and recovery of tourism 3.2.1 Diverse tourism recovery dates Changes in infection sizes that would have been caused by increase in international arrivals varied greatly from region to region for any fixed tourism recovery time. Had all foreign visitors not been quarantined upon arrival, Taiwan and Singapore would have been the most adversely affected by resumption of tourism to the 2019 level, where the total number of infections over the 6-month period would have been 2.18 (95% CrI: 1.8–2.6) and 1.34 (95% CrI: 1.25–1.44) times higher than in the scenario when the volume of tourists remained at 2021 levels. By comparison, influence of the rise in incoming tourists would not have been so significant in New Zealand, because lifting border restrictions would have caused surges in infections even when there were not so many international visitors (Figure 2 , Figure S7). For Japan, Malaysia and Thailand, an earlier resumption of tourism in June 2022 would have caused less than a 5% increase in total cases. The relationship between increases in infection sizes and recovery times, however, was non-linear for all eight regions, and a later recovery of tourism did not necessarily mean fewer additional infections per month (Figure S8).Figure 2 Infection size comparison for scenarios with diverse tourism recovery dates. Comparison (mean & 95% CrI) of infection sizes (from the start of tourism resumption to the end of November) between scenarios with different tourism resumption dates ranging from the first day of June to November. The reference scenario is when numbers of foreign visitors remained the same as those in 2021 throughout the six months’ simulation window. In each scenario, no movement restriction was imposed on foreign visitors upon their arrival. Figure 2 3.2.2 Selective border reopening We then considered selective reopening to low incidence regions, leading to some but not all intra-regional travel flows recovering to pre-pandemic levels. Over the period considered, as the number of regions reopened to rose from one to seven, the number of infections would have surged. Outbreak sizes in New Zealand were the most sensitive to border reopening, where visitors from Taiwan, the lowest risk region it could open to, meant the total infections over the half year would have expanded by 13 (95% CrI: 9–18) times. Other regions, by comparison, would not have been so greatly influenced by importation of cases, and withdrawing border restrictions to New Zealand and Taiwan would have increased case counts by less than 50% elsewhere. Furthermore, for Japan, Malaysia and Thailand, reopening to all the regions would have increased total infection sizes by no more than 5%. The impact of additionally reopening to regions with large infection and mortality rates, such as Thailand and Malaysia, nevertheless, would have been relatively significant. Particularly, reported cases increased by at least a quarter over the 6-month period if residents in Malaysia were eligible for quarantine-free entry elsewhere (Figure 3 , Figure S9).Figure 3 Infection size comparison for border reopening to different regions. Times of total infections (mean & 95% CrI) for each region by the end of the 6-month simulation period (from June to November 2021), compared to the real situation, under selective opening of borders to some of the other regions (from one to seven, ordered by risk), assuming communication among the regions were recovered to pre-pandemic (2019) levels throughout this time. Figure 3 3.2.3 Impact on mortality of easing travel restrictions Singapore was estimated to experience the biggest rise in mortality due to the relaxation of quarantine requirements, with an average increase of 225 deaths per million residents if visitors from the other seven regions were allowed to enter without quarantine (Table 1 ). It would be followed by Australia and New Zealand, where 60 and 54 additional deaths per million residents would be expected, respectively. The introduction of mandatory tests for international arrivals, however, would have approximately halved the increase in deaths in Singapore and Australia; the reduction was not so great for New Zealand. For the other regions, none of the three proposed strategies led to more than 35 more deaths per million residents, and testing upon arrival brought down the fatalities considerably, with fewer than half the deaths of a less cautious exit strategy (Table 1, Table S6, Figure S10).Table 1 Deaths per million residents for different travel-resumption strategies. Actual, expected increase of and total deaths per million residents for each region over the 6-month period (June–November, 2021) when quarantine-free entry were allowed for all (S1), RAT-negative (S2) and PCR-negative (S3) travellers from the eight Western Pacific regions, assuming communication among the regions were recovered to pre-pandemic levels (the corresponding period in 2019). Table 1Region Actual S1: quarantine free S2: RAT test S3: PCR test Increase Total Increase Total Increase Total Australia 42 60 (140%) 102 33 (78%) 75 21 (50%) 63 Japan 42 1.5 (3.6%) 44 0.5 (1.2%) 43 0.1 (0.2%) 42 Republic of Korea 33 4.5 (14%) 37 1.8 (5.5%) 35 0.7 (2.1%) 33 Malaysia 845 32 (3.8%) 877 9.2 (1.1%) 854 1.6 (0.2%) 847 New Zealand 4 54 (1500%) 58 42 (1200%) 46 37 (1100%) 41 Singapore 125 225 (180%) 349 148 (120%) 273 78 (62%) 202 Thailand 298 14 (4.5%) 312 5.2 (1.7%) 303 1.3 (0.4%) 300 Taiwan 31 20 (64%) 51 8.7 (28%) 39 3.7 (12%) 35 4 Discussion Since early 2020, COVID-19 has caused the severest social and economic setbacks in the past few decades (UN.ESCAP, 2021). While border control measures have proven to be effective in containing spread (Dickens et al., 2020), the resulting reduction in international travel went against the interdependency of modern economies. Forward-looking reopening policies are therefore required, to mitigate the harmful effects of the current pandemic, be prepared for future epidemic waves, and to avoid exhausting the goodwill of policymakers and the public. The number of cases arising from importation has been found to be closely associated with local incidence and local virus transmissibility (Russell et al., 2021). Our analysis suggests that in the absence of rigorous travel restrictions, new epidemic waves seeded by imported cases would have been substantially more likely to have taken place in regions with few local cases, compared to those that already had large, localized outbreaks. The vulnerability could partially be explained by the use of an estimated Rt corresponding to transmissibility as events transpired, but which might be reasonably expected to be lower had infections been markedly greater, given these regions’ otherwise successful policies. The diversified correlations between monthly increases in infection sizes and tourism resumption times proposed in this study suggested the possibly of earlier reopening in some regions. The similarity in average additional infections per month in some cases supports the notion that travel restrictions that reduce international arrivals only postpone an outbreak rather than preventing it, particularly when a region was in the midst of an epidemic (Russell et al., 2021). Meanwhile, border reopening to regions with higher risks would generate a disproportionally large number of additional infections. This would align well with a selective reopening strategy, which is in accordance with WHO's recommendation of risk-based lifting of travel restrictions (WHO, 2021b). In addition, our estimation of infection sizes in the hypothetical scenarios with varying border requirements demonstrated the significance of mandatory on-arrival tests. Though the more-accurate PCR tests could better bring down importation impact by half for most of the regions investigated compared to RAT tests, the former's high false-positive rates after recovery (Routsias et al., 2021) would cause unnecessary quarantine for recovered infections who are no longer infectious and thus extra financial burdens for individual travellers as well as the authorities. By comparison, RAT tests, despite their relatively low sensitivity, are more affordable and give more timely results (Cherian et al., 2021). Limitations of this study include the utilization of uniform, constant reporting rates under the assumption of sufficient access to COVID-19 testing in the regions considered, but the rates could be region-specific or time-varying, affected by policies adopted, such as the home-recovery programme in Singapore introduced in October 2021 (Ministry of Health Singapore, 2022). Rt was estimated under the optimal assumption of minimal leaking risks in quarantine, realized through stringent border restrictions and potent test-trace-isolate strategies (Grantz et al., 2021), while changes would possibly arise with the ease of border restrictions, since the transmission potential might be subjected to exogenous factors such as mobility and containment measures implemented as well (Jin et al., 2022). Rt is also likely to be affected by imported cases, particularly when the composition of circulating variants differs from that of local infections and disparities exist in transmissibility of different variants. Another simplified assumption was the uniform probability of getting infected for all individuals in the same place, be they foreign visitors, local residents or people planning to travel abroad. We neither considered the possibility of re-infection, nor did we account for the heterogeneity in individuals’ transmission capacity, which may be influenced by symptoms displayed during the infectious period (Cevik et al., 2020; Dickens et al., 2021). Meanwhile, simulated infection sizes could be affected by the uncertainty in Rt posterior estimates, resulted from the scarcity of case information in regions like New Zealand and Taiwan. The impact could hardly be eliminated, though greatly mitigated by EpiFilter, which provides more precise Rt estimations in low incidence scenarios compared with other prevailing methods. Furthermore, due to limited data accessibility, we only modelled the impact of foreign arrivals without accounting for residents returning from abroad, which was likely to double the expected additional infections if taken into consideration. Finally, as the estimation window for this study was mostly the time when delta variant was the dominant strain, parameters involved for prediction were relatively consistent over the period, but emergence of new variants may modify transmission patterns of the virus, as well as efficacy of vaccination and fatality rates (Eyre et al., 2022; Maslo et al., 2022). Therefore, risks of lifting border restrictions assessed in this study might not reflect the situation after the emergence of the Omicron variant, which is more transmissible yet. However, the results nevertheless suggest that the resumption of international travel could have been effected sooner within the Asia-Pacific once widespread local transmission had begun. 5 Ethical approval Not applicable. 6 Data availability The data are available upon request. 7 Authors' contributions All authors have read and agreed to the published version of the manuscript. 8 Reference Asian Development Bank. How COVID-19 Is Changing the World: A Statistical Perspective Volume 3. Asian Development Bank; 2021. Australian Government Department of Health and Aged Care. Recommencing quarantine-free travel from New Zealand to Australia 2021. https://www.health.gov.au/news/recommencing-quarantine-free-travel-from-new-zealand-to-australia (accessed July 22, 2022). Australian Government Department of Home Affairs. COVID-19 and the border 2022. https://www.homeaffairs.gov.au/covid19 (accessed July 23, 2022). Cevik M, Kuppalli K, Kindrachuk J, Peiris M. Virology, transmission, and pathogenesis of SARS-CoV-2. BMJ 2020;371:m3862. https://doi.org/10.1136/bmj.m3862. Chadwick FJ, Clark J, Chowdhury S, Chowdhury T, Pascall DJ, Haddou Y, et al. Combining rapid antigen testing and syndromic surveillance improves community-based COVID-19 detection in a low-income country. Nat Commun 2022;13:2877. https://doi.org/10.1038/s41467-022-30640-w. Cherian P, Krishna S, Menon GI. Optimizing testing for COVID-19 in India. PLOS Comput Biol 2021;17:e1009126. https://doi.org/10.1371/journal.pcbi.1009126. Chinazzi M, Davis JT, Ajelli M, Gioannini C, Litvinova M, Merler S, et al. The effect of travel restrictions on the spread of the 2019 novel coronavirus (COVID-19) outbreak. Science 2020;368:395–400. https://doi.org/10.1126/science.aba9757. Civil Aviation Authority of Singapore. CAAS \| Newsroom 2022. https://www.caas.gov.sg/who-we-are/newsroom (accessed July 22, 2022). Cook A r, Gibson G j, Gottwald T r, Gilligan C a. Constructing the effect of alternative intervention strategies on historic epidemics. J R Soc Interface 2008;5:1203–13. https://doi.org/10.1098/rsif.2008.0030. Cori A, Ferguson NM, Fraser C, Cauchemez S. A New Framework and Software to Estimate Time-Varying Reproduction Numbers During Epidemics. Am J Epidemiol 2013;178:1505–12. https://doi.org/10.1093/aje/kwt133. Dickens BL, Koo JR, Lim JT, Park M, Sun H, Sun Y, et al. Determining quarantine length and testing frequency for international border opening during the COVID-19 pandemic. J Travel Med 2021;28:taab088. https://doi.org/10.1093/jtm/taab088. Dickens BL, Koo JR, Lim JT, Sun H, Clapham HE, Wilder-Smith A, et al. Strategies at points of entry to reduce importation risk of COVID-19 cases and reopen travel. J Travel Med 2020;27:taaa141. https://doi.org/10.1093/jtm/taaa141. Dong E, Du H, Gardner L. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect Dis 2020;20:533–4. https://doi.org/10.1016/S1473-3099(20)30120-1. Eyre DW, Taylor D, Purver M, Chapman D, Fowler T, Pouwels KB, et al. Effect of Covid-19 Vaccination on Transmission of Alpha and Delta Variants. N Engl J Med 2022;386:744–56. https://doi.org/10.1056/NEJMoa2116597. Freunde von GISAID e.V. GISAID Initiative 2022. https://www.epicov.org/epi3/frontend#61d856 (accessed August 6, 2022). Grantz KH, Lee EC, McGowan LD, Lee KH, Metcalf CJE, Gurley ES, et al. Maximizing and evaluating the impact of test-trace-isolate programs: A modeling study. PLOS Med 2021;18:e1003585. https://doi.org/10.1371/journal.pmed.1003585. Iacus SM, Natale F, Santamaria C, Spyratos S, Vespe M. Estimating and projecting air passenger traffic during the COVID-19 coronavirus outbreak and its socio-economic impact. Saf Sci 2020;129:104791. https://doi.org/10.1016/j.ssci.2020.104791. Jin S, Dickens BL, Lim JT, Cook AR. EpiRegress: A Method to Estimate and Predict the Time-Varying Effective Reproduction Number. Viruses 2022;14:1576. https://doi.org/10.3390/v14071576. Jones JM, Opsomer JD, Stone M, Benoit T, Ferg RA, Stramer SL, et al. Updated US Infection- and Vaccine-Induced SARS-CoV-2 Seroprevalence Estimates Based on Blood Donations, July 2020-December 2021. JAMA 2022;328:298–301. https://doi.org/10.1001/jama.2022.9745. Kanji JN, Zelyas N, MacDonald C, Pabbaraju K, Khan MN, Prasad A, et al. False negative rate of COVID-19 PCR testing: a discordant testing analysis. Virol J 2021;18:13. https://doi.org/10.1186/s12985-021-01489-0. Maslo C, Friedland R, Toubkin M, Laubscher A, Akaloo T, Kama B. Characteristics and Outcomes of Hospitalized Patients in South Africa During the COVID-19 Omicron Wave Compared With Previous Waves. JAMA 2022;327:583–4. https://doi.org/10.1001/jama.2021.24868. Ministry of Health NZ. COVID-19: Response planning 2022. https://www.health.govt.nz/covid-19-novel-coronavirus/covid-19-response-planning (accessed July 23, 2022). Ministry of Health Singapore. Eligible for Home Recovery Programme 2022. https://www.covid.gov.sg/unwell/hrp (accessed September 26, 2022). Miyawaki A, Tsugawa Y. Health and Public Health Implications of COVID-19 in Asian Countries. Asian Econ Policy Rev 2022;17:18–36. https://doi.org/10.1111/aepr.12358. Parag KV. Improved estimation of time-varying reproduction numbers at low case incidence and between epidemic waves. PLOS Comput Biol 2021;17:e1009347. https://doi.org/10.1371/journal.pcbi.1009347. Pulliam JRC, van Schalkwyk C, Govender N, von Gottberg A, Cohen C, Groome MJ, et al. Increased risk of SARS-CoV-2 reinfection associated with emergence of Omicron in South Africa. Science 2022;376:eabn4947. https://doi.org/10.1126/science.abn4947. Routsias JG, Mavrouli M, Tsoplou P, Dioikitopoulou K, Tsakris A. Diagnostic performance of rapid antigen tests (RATs) for SARS-CoV-2 and their efficacy in monitoring the infectiousness of COVID-19 patients. Sci Rep 2021;11:22863. https://doi.org/10.1038/s41598-021-02197-z. Russell TW, Wu JT, Clifford S, Edmunds WJ, Kucharski AJ, Jit M. Effect of internationally imported cases on internal spread of COVID-19: a mathematical modelling study. Lancet Public Health 2021;6:e12–20. https://doi.org/10.1016/S2468-2667(20)30263-2. Shim E, Choi W, Song Y. Clinical Time Delay Distributions of COVID-19 in 2020–2022 in the Republic of Korea: Inferences from a Nationwide Database Analysis. J Clin Med 2022;11:3269. https://doi.org/10.3390/jcm11123269. Ullah AA, Azam N, Daud K. COVID-19 and shifting border policies in Southeast Asia. Southeast Asia Multidiscip J 2021;21:1–14. UN.ESCAP. Economic and social survey of Asia and the Pacific 2021 : towards post-COVID-19 resilient economies. United Nations; 2021. UNWTO. Country profile – inbound tourism \| Tourism Dashboard 2022a. https://www.unwto.org/tourism-data/country-profile-inbound-tourism (accessed July 23, 2022). UNWTO. Tightened Travel Restrictions Underline Current Challenges for Tourism 2022b. https://www.unwto.org/news/tightened-travel-restrictions-underline-current-challenges-for-tourism (accessed July 22, 2022). UNWTO. New COVID-19 Surges Keep Travel Restrictions in Place 2021. https://www.unwto.org/news/new-covid-19-surges-keep-travel-restrictions-in-place (accessed July 22, 2022). WHO. WHO advice for international traffic in relation to the SARS-CoV-2 Omicron variant (B.1.1.529) 2021a. https://www.who.int/news-room/articles-detail/who-advice-for-international-traffic-in-relation-to-the-sars-cov-2-omicron-variant (accessed July 23, 2022). WHO. Policy considerations for implementing a risk-based approach to international travel in the context of COVID-19, 2 July 2021 2021b. https://www.who.int/publications-detail-redirect/WHO-2019-nCoV-Policy-Brief-Risk-based-international-travel-2021.1 (accessed July 26, 2022). Wood S. mgcv: Mixed GAM Computation Vehicle with Automatic Smoothness Estimation 2022. CRediT authorship contribution statement Shihui Jin: Methodology, Formal analysis, Data curation, Writing – original draft, Visualization. Jue Tao Lim: Conceptualization, Writing – review & editing, Funding acquisition. Borame Lee Dickens: Writing – review & editing. Alex R Cook: Conceptualization, Methodology, Writing – review & editing, Supervision, Funding acquisition. Appendix Supplementary materials Image, application 1 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. Funding This work was supported by Singapore's Ministry of Education (through a Tier 1 grant) and the National University of Singapore (through a Reimagine Research grant). Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ijregi.2022.11.013.	36466213	PMC9710097	NO-CC CODE	2022-12-01 23:23:12	no	IJID Reg. 2022 Nov 30; doi: 10.1016/j.ijregi.2022.11.013	utf-8	IJID Reg	2,022	10.1016/j.ijregi.2022.11.013	oa_other
==== Front Eur J Pharm Sci Eur J Pharm Sci European Journal of Pharmaceutical Sciences 0928-0987 1879-0720 The Authors. Published by Elsevier B.V. S0928-0987(22)00230-5 10.1016/j.ejps.2022.106345 106345 Article The PBPK LeiCNS-PK3.0 framework predicts Nirmatrelvir (but not Remdesivir or Molnupiravir) to achieve effective concentrations against SARS-CoV-2 in human brain cells Saleh Mohammed A.A. 1 Hirasawa Makoto 1 Sun Ming 1 Gülave Berfin 1 Elassaiss-Schaap Jeroen 2 de Lange Elizabeth C.M. 1⁎ 1 Division of Systems Pharmacology and Pharmacy, Leiden Academic Center for Drug Research, Leiden University, Leiden, The Netherlands 2 PD-value B.V., Houten, The Netherlands ⁎ Corresponding author. 30 11 2022 1 2 2023 30 11 2022 181 106345106345 3 8 2022 17 11 2022 29 11 2022 © 2022 The Authors. Published by Elsevier B.V. 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. SARS-CoV-2 was shown to infect and persist in the human brain cells for up to 230 days, highlighting the need to treat the brain viral load. The CNS disposition of the antiCOVID-19 drugs: Remdesivir, Molnupiravir, and Nirmatrelvir, remains, however, unexplored. Here, we assessed the human brain pharmacokinetic profile (PK) against the EC90 values of the antiCOVID-19 drugs to predict drugs with favorable brain PK against the delta and the omicron variants. We also evaluated the intracellular PK of GS443902 and EIDD2061, the active metabolites of Remdesivir and Molnupiravir, respectively. Towards this, we applied LeiCNS-PK3.0, the physiologically based pharmacokinetic framework with demonstrated adequate predictions of human CNS PK. Under the recommended dosing regimens, the predicted brain extracellular fluid PK of only Nirmatrelvir was above the variants’ EC90. The intracellular levels of GS443902 and EIDD2061 were below the intracellular EC90. Summarizing, our model recommends Nirmatrelvir as the promising candidate for (pre)clinical studies investigating the CNS efficacy of antiCOVID-19 drugs. Graphical abstract Image, graphical abstract Keywords LeiCNS-PK3.0 COVID-19 Brain Pharmacokinetics ==== Body pmc1 Introduction Increasing evidence supports that COVID-19 is not only a respiratory disease but may also have serious impact on, among others, the central nervous system (CNS) (Philippens et al., 2021). The neurological manifestations associated with SARS-CoV-2 include headaches, encephalopathy (Chou et al., 2021; Guadarrama-Ortiz et al., 2020), Alzheimer's disease-like manifestations (Shen et al., 2022), and brain atrophy (Douaud et al., 2022). SARS-CoV-2 has also been demonstrated to infect (Matschke et al., 2020; Veleri, 2022) and persist in neurons for up to 230 days (Stein et al., 2021). A causal relationship of neurotropism and neurological manifestations is still, however, unestablished (Pacheco-Herrero et al., 2021; Shen et al., 2022; Yang et al., 2021). Addressing the viral infection in the brain is therefore relevant to avoid a long-term latent state of virus in the CNS, which could result in recurrent CNS pathologies. Three small molecule drugs have so far been approved for the treatment of COVID-19 in humans, which include the main protease inhibitor, Nirmatrelvir, in addition to Remdesivir and Molnupiravir that are activated intracellularly to the nucleoside analogues GS443902 and EIDD2061, respectively. The PK profiles of these drugs and their active metabolites in the human brain have not been assessed. We here apply the physiologically based LeiCNS-PK3.0 framework to predict the PK profiles of these drugs in the brain, and relate these to their in vitro EC90 (Gonçalves et al., 2020) values against the delta and omicron variants of SARS-CoV-2. By this approach we select drug(s) that seems to be promising for treating these viruses in the brain. 2 Data and Methods 2.1 Data collection We first compiled in vitro and preclinical in vivo data on CNS disposition and blood-brain barrier (BBB) transport of Nirmatrelvir, Remdesivir (and its metabolites: GS704277 and GS441524 and active form GS443902), and Molnupiravir (and its metabolite: EIDD1931 and active form EIDD2061). Molnupiravir is unstable in plasma and is efficiently and rapidly converted to EIDD1931. Therefore, EIDD1931 was used as a surrogate to describe Molnupiravir's plasma and CNS disposition. Also, Molnupiravir dosing was performed in molarity to account for the difference in molecular weight between the parent drug and its metabolite. In addition, the extent of CNS distribution of these drugs given by Kpuu,BBB (brainECF to plasma unbound drug ratio) was evaluated using the in-silico brain exposure efficiency (BEE) score (Gupta et al., 2020). Population plasma PK models were extracted from literature. Drug physicochemical properties were available from DRUGBANK (Wishart et al., 2017). The model related input is reported in Table 1 . Literature data, where required, were digitized with WebPlotDigitizer version 4.2 (https://apps.automeris.io/wpd/).Table 1 LeiCNS-PK3.0 input parameters Table 1Drug Nirmatrelvir Remdesivir Molnupiravir Parent Parent Metabolites Metabolite GS704277 GS441524 EIDD1931 Physicochemical properties (Wishart et al., 2017) MW (g/mol) 499.535 602.585 442.32 291.267 259.218 LogP (unitless) 2.12 2.01 -0.88 -0.58 -2 pKa (unitless) 7.1 10.23 2.38 12.13 12.55 pKb (unitless) -1.6 0.65 0.64 0.65 2.39 Plasma PK (European Medicines Agency, 2021a, 2021b; U.S. Food and Drug Administration, 2020) CLcen (mL min−1) 17 803.33 3500 2933.33 1281.667 Qcen-per1 (mL min−1) 7.4 84 341.667 6316.667 55.833 Qcen-per2 (mL min−1) 0 0 0 526.667 0 Vcen (mL) 8200 6340 242000 104000 72000 Vper1 (mL) 5650 6000 46000 236000 70000 Vper2 (mL) 0 0 0 233000 0 Ka (min−1) 0.3783 not applicable not applicable not applicable 0.01383 D1 (min) NA not applicable not applicable not applicable 48.12 IIV on CLcen 0.264 0.387 0.565 0.4898 0.411 IIV on Qcen-per1 0 0 0 0 0 IIV on Qcen-per2 0 0.5656 0 0 0 IIV on Vcen 0.307 0.387 0.6244 0.67 0.4 IIV on Vper1 0.699 0.5656 0.5099 0.6557 0 IIV on Vper2 0 0 0 0.5 0 IIV on Ka 0.576 NA NA NA 0 IIV on D1 0 NA NA NA 0.428 Proportional residual error 0.0336 0.45 0.44 0.31 0.442 Additive residual error (ng/ml) 399 0.884 0.604 0.511 0 Drug biological parameters fu,p 0.31 0.121 0.98 0.99 1 Kpuu,BBB 0.35a 0.14a 0.07a 0.22a 0.4b AFef,BBB 3.02 7.97 23864402 696.67 1719.71 Kpuu,LV 0.35c 0.14c 0.07c 0.22c 0.4c AFef,LV 4.67 10.46 94737456 2739.28 6773.89 Kpuu,SAS 0.35c 0.14c 0.07c 0.22c 0.4c AFef,SAS 4.71 10.48 95620650 2744.83 6893.5 BBB transport (European Medicines Agency, 2021b, 2021a, 2020) P-gp substrate substrate NA substrate not substrate BCRP not substrate not substrate NA substrate not substrate ENT1 NA NA NA substrate NA ENT2 NA NA NA NA substrate CNT1 NA NA NA not substrate substrate CNT2 NA NA NA not substrate substrate CNT3 NA NA NA substrate substrate Dosing (European Medicines Agency, 2021a, 2021b; U.S. Food and Drug Administration, 2020) Dose (mg) 300 200 + 100 NA NA 800d Dosing frequency (day−1) twice once NA NA twice Treatment duration (days) 5 5 NA NA 5 Administration route oral IV NA NA oral Efficacy EC50 (delta) (uM) 0.076 (European Medicines Agency, 2021b; Rosales et al., 2022) 0.071 (European Medicines Agency, 2020; Rosales et al., 2022) NA 0.86 (Vangeel et al., 2022) 1.43 (European Medicines Agency, 2021a; Rosales et al., 2022; Vangeel et al., 2022) EC90 (delta) (uM) 0.149 (European Medicines Agency, 2021b; Rosales et al., 2022) 0.1455 (European Medicines Agency, 2020; Rosales et al., 2022) NA NA 4.65 (Rosales et al., 2022) EC50 (omicron) (uM) 0.02 (Rosales et al., 2022) 0.02 (Rosales et al., 2022) NA 0.5 (Vangeel et al., 2022) 0.25 (Rosales et al., 2022) EC90 (omicron) (uM) 0.05 (Rosales et al., 2022) 0.09 (Rosales et al., 2022) NA NA >10 (Rosales et al., 2022) a Predicted values using the brain exposure efficiency score (Gupta et al., 2020) b Kp value was calculated based on mouse brain homogenate (Painter et al., 2019) and was corrected to Kpuu,BBB accounting for the plasma and brain binding and brain pH differences c assumed the same as Kpuu,BBB d Molnupiravir dose in the model simulations was performed in units of molarity to account for the difference of molecular weight between Molnupiravir and its metabolite EIDD1931. MW: molecular weight, LogP: octanol-water partitioning, pKa: acid dissociation constant, pKb: base dissociation constant, CLcen: drug clearance from central plasma compartment, Qcen-per: Drug clearance between central and peripheral plasma compartments, Vcen: volume of central plasma compartment, Vper: volume of peripheral plasma compartments, Ka: absorption rate constant, D1: estimated duration, IIV: interindividual variability, fu,p: plasma unbound fraction, Kpuu: unbound drug concentration ratio, AFef/in: asymmetry factor efflux/influx, P-gp: P-glycoprotein, BCRP: breast cancer receptor protein, ENT: equilibrative nucleoside transporters, CNT: concentrative nucleoside transporter, EC50/90: drug concentration for 50%/90% efficacy, NA: not available 2.2 LeiCNS-PK3.0 framework LeiCNS-PK3.0 is a physiologically based pharmacokinetic (PBPK) model of the CNS, which can predict the unbound PK profile in different CNS compartments, including the target sites in the brain extracellular (brainECF) and intracellular (brainICF) compartments and also the lumbar cerebrospinal fluid compartment. The model was previously validated and was shown to predict, independently of clinical brain PK data, the unbound PK profiles of morphine in the human brainECF and of indomethacin, oxycodone, and acetaminophen at the lumbar region of the subarachnoid space cerebrospinal fluid (CSF) compartments, both with less than two-fold error. Additional details on model structure and validation have been reported previously (Saleh et al., 2021). Here, we will use the validated LeiCNS-PK3.0 to predict the human brainECF and brainICF PK profiles of the three antiCOVID-19 drugs. It will not be possible, however, to validate these predictions since relevant brain PK measurements are unavailable. 2.3 LeiCNS-PK3.0 simulations Model simulations were performed using the physiological parameters of a healthy human adult as reported previously (Saleh et al., 2021) and the plasma PK parameters and drug physicochemical properties presented in Table 1. Fifty simulations were performed to account for interindividual variability of the population plasma PK models and the median and 95 percentiles were reported. Simulations were performed in R (version 4.1.2) (R Core Team, 2019) using the package RxODE (version 1.1.4) (Fidler et al., 2019) and the LSODA (Livermore Solver for Ordinary Differential Equations) Fortran package. 2.4 Brain intracellular PK assessment Remdesivir is a prodrug and is metabolized intracellularly to GS443902, the active nucleoside analogue. GS443902 is hydrophilic, with long elimination half-life (≈ 43 hours) as measured in human peripheral blood mononuclear cells (Humeniuk et al., 2021), which imply that GS443902 may accumulate intracellularly, producing a sustained effect. We therefore investigated the intracellular brain PK profile of GS443902. The intracellular PK profiles of Remdesivir metabolites were reported in lung epithelium cells (Calu-3 cells) (Gilead Sciences, 2020) and were used to model the intracellular brain PK of GS443902. Briefly, we assumed that the triphosphate active metabolite GS443902 is formed from GS704277 metabolite directly (half-life = 30.4 hours), given the low concentrations of the intermediate monophosphate and diphosphate metabolites. The formation rate of GS443902 was multiplied by a factor of 24.9 to correct for the slow metabolic rate of Calu-3 cell line compared to other human cell lines, for example the hepatocellular carcinoma (Huh-7), primary airway epithelium (HAE), and kidney epithelium (293T) (Pruijssers et al., 2020; Tao et al., 2021). No formation of GS443902 from GS441524 was considered, supported by the inefficiency of this process, as demonstrated by the in vitro experiments using the Huh-7, HAE, Calu-3, Caco-2, and 293T cell lines (Gilead Sciences, 2020; Tao et al., 2021). GS443902 is metabolized to GS441524 with a half-life of 43 hours (Humeniuk et al., 2021). Likewise, Molnupiravir is the prodrug of the parent nucleoside EIDD1931, which undergoes intracellular conversion to EIDD2061, the triphosphate metabolite of EIDD1931. Intracellular PK of EIDD2061 was modeled based on the mouse brain homogenate data of EIDD1931 and EIDD2061 (Painter et al., 2019). The formation and elimination half-lives of EIDD2061 were 3.5 (Painter et al., 2019) and 4.5 (European Medicines Agency, 2021a) hours, respectively. 2.5 Efficacy calculation Comparison of predicted brainECF PK profile against the EC90 was used to assess if a drug would achieve effective brain PK. Efficacy against the omicron and delta variants was considered as these are the current variants of concern (World Health Organization WHO, 2022. In addition, efficacy at a given time point (ε t) was calculated using the predicted brainECF concentrations at given time point (CECF,t) and EC50. Average efficacy (ε) of the drug across the PK profile was calculated by integrating ε t over the treatment duration (D) (Gonçalves et al., 2020). In vitro measured EC50 and EC90 were available from literature and are reported in Table 1.εt=CECF,tEC50+CECF,t ε=1D*∫0Dεtdt 2.6 Sensitivity analysis A sensitivity analysis was performed to evaluate the impact of the CNS pathophysiological changes associated with COVID-19 on brain PK profiles of the three antiCOVID-19 drugs (Saleh and de Lange, 2021). Changes of all model physiological parameters were assessed including pH values of brainECF, brainICF, lysosomes, plasma, and CSF; effective surface area of paracellular transport across the BBB and blood-CSF (BCSFB); bulk fluid flow as cerebral blood flow, brainECF bulk flow, and CSF flow; surfaces areas of BBB, BCSFB, brain cells, and lysosomes; and the volumes of brain microvasculature, brainECF, brainICF, lysosomes, brain phospholipids, lateral ventricles, third and fourth ventricles, cisterna magna, and subarachnoid space. Model parameters were changed by 10% and 200%, while pH values were altered by 0.1 and 2 pH units. The Cmax, Tmax, AUC, and half-life of the PK profiles from the healthy and altered CNS parameters were then compared. 3 Results The predicted PK profiles of plasma and brainECF of Remdesivir, GS441524, Nirmatrelvir, and EIDD1931 are presented in Fig. 1 . The predicted brainECF PK profiles are depicted against the in vitro EC90 values against the delta and omicron variants, except for GS441524 which is depicted against the EC50 value, as the EC90 values of the delta and omicron variants were not available. Extracellular PK profiles were compared against the EC90 values, since the in vitro EC90 values reflect extracellular and not intracellular drug concentrations.Fig. 1 CNS PK predictions of LeiCNS-PK3.0 of drugs approved for COVID-19 treatment. Median (solid green line) and 95 percentiles (green shaded area) of brainECF and subarachnoid space PK profiles of Nirmatrelvir, EIDD1931 (Molnupiravir plasma metabolite), and Remdesivir, in addition to Remdesivir metabolite, GS441524. The PK profiles are depicted against the EC90 (dashed lines) against the omicron (blue) and the delta (red) variants. The concentrations of Nirmatrelvir only were above the respective EC90 of both variants, with an average efficacy of 87% and 96% against the delta and omicron variants, respectively. Fig 1 The predicted brainECF PK profile of Nirmatrelvir was consistently above the EC90 value of both variants, with an average efficacy of 87% and 96% against the delta and omicron variants, respectively. Nirmatrelvir still achieved effective brainECF PK profiles following a 50% reduction of plasma Cmax (supplementary figure 1, online resource 1). The reduction of plasma Cmax was achieved with an 85% lower absorption rate constant to account for the formulation differences between tablets and oral suspensions (European Medicines Agency, 2021b). The predicted brainECF PK profiles of Remdesivir and of GS441524 were below the EC90. The predicted brainICF PK profile of GS443902 against the intracellular EC90 value are depicted in Fig. 2 . The intracellular EC90 value was calculated based on the extracellular EC90 value of Remdesivir and the average intracellular levels of GS443902 (Pruijssers et al., 2020). BrainICF concentrations profile of GS443902 increased over time with each dose, but remained, however, below the intracellular EC90.Fig. 2 Median (solid green line) and 95 percentiles (green shaded area) of the predicted intracellular levels of GS443902 (top) and EIDD2061 (bottom), the active triphosphate metabolites of Molnupiravir and Remdesivir, respectively. GS443902 is hydrophilic (logP = -5.3 (NCBI Resource Coordinators, 2018)), with an elimination half-life of 43.4 hours (Humeniuk et al., 2021), which indicate its potential for intracellular accumulation. At the recommended dosing, however, GS443902 predicted levels are below the intracellular EC90 value (1.78 pmol/million cell (Pruijssers et al., 2020), dashed blue line) against USA-WA1/2020. EIDD2061 is also hydrophilic, but with a relatively short half-life of 4.5 hr (European Medicines Agency, 2021a) and therefore does not accumulate extensively intracellularly (Painter et al., 2019). Data required for calculating intracellular EC90 of EIDD2061 were not available. EIDD2061 intracellular predicted Cmax is however ten fold lower than the EC90 reported for EIDD1931, while the average concentration ratio of EIDD2061 to EIDD1931 ranges from one-third to two. Thus, intracellular EC90 can be as low as 1.55 nmol/ml, which is still three times the Cmax of EIDD2061 at the recommended dosing regimen. Fig 2 The predicted brainECF PK of EIDD1931 was below the EC90 of the two variants. EIDD2061 brainICF PK profile, reported in Fig. 2, does not notably accumulate continuously in brain (Painter et al., 2019), mainly because of its short half-life. Data required to calculate the intracellular EC50/90 of EIDD2061 were not available. The average concentration ratio of EIDD2061 to EIDD1931 is between one-third and two, as measured in mice spleen and brain, respectively (Painter et al., 2019). This means that the intracellular EC90 of EIDD2061 can be assumed to be (at best) threefold lower than that measured extracellularly for EIDD1931 or 1.55 nmol/ml, which is still three times higher than the predicted intracellular Cmax of EIDD2061 (0.4 nmol/ml). In this study, LeiCNS-PK3.0 simulations were performed using the parameters of the healthy human CNS. Therefore, a sensitivity analysis was performed to assess the PK changes caused by the potential COVID-19 alterations of CNS physiology. Changes of pHECF and pHICF resulted in the largest change of brainECF and brainICF PK of Nirmatrelvir (pKa = 7.1, Table 1). Remdesivir and EIDD1931 are neutral molecules and thus not impacted by pH changes. Also, changes of brain cell volume and surface area impacted the PK of EIDD1931. 4 Discussion The neurotrophic characteristics and the associated neurological manifestations of SARS-CoV-2 strongly imply the need to eradicate the virus from the brain. CNS penetration of small molecule drugs approved for COVID-19 treatment have not been studied in humans. Using the LeiCNS-PK3.0 PBPK framework and the recommended dosing regimens, we predict that Nirmatrelvir alone achieves adequate brain PK profiles as based on the in vitro EC90 values against SARS-CoV-2 variants of interest, i.e. the delta and omicron variants. These results can guide clinical trials on the assessment of efficacy of antiCOVID-19 drugs in the human CNS. Based on our model simulations, the dose of Remdesivir or Molnupiravir required to achieve effective concentrations in the brain cells will exceed by several folds the highest dose that was tested during the clinical development of both drugs. A minimum dose of 300 mg twice daily of Remdesivir was needed for the brainICF Cmin of GS443902 to be higher than the calculated intracellular EC90 value (1.78 pmol/million cell (Pruijssers et al., 2020)). With regards to Molnupiravir, a dose of 4000 mg twice daily was required for the intracellular Cmin of EIDD2061 to exceed the lowest predicted intracellular EC90 value of 1.55 nmol/ml. Both doses were not explored in the dose escalation studies in humans (Humeniuk et al., 2020; Painter et al., 2021) and thus the associated potential toxicities have not been investigated. COVID-19 is associated with distinct CNS pathophysiological alterations. SARS-CoV-2 impaired the integrity of the BBB, either because of the impairment of the basement membrane without affecting tight junctions (Krasemann et al., 2022; Zhang et al., 2021) or the loss of tight junction proteins (Buzhdygan et al., 2020; Erickson et al., 2021; Reynolds and Mahajan, 2021; Wang et al., 2021). Also, the increased protein content in CSF (Jarius et al., 2022; Tandon et al., 2021) suggests a breakdown of the BCSFB (Pellegrini et al., 2020), but could also be a result of decreased CSF flow (Reiber, 1994). SARS-CoV-2 infection might result in brain atrophy, wherein the volume of gray matter significantly reduced than white matter (Douaud et al., 2022; Qin et al., 2021). No direct evidence suggests the changes in the volume and the surface area of brain cells. Many COVID-19 patients, however, present with hypoxemia (Dhont et al., 2020; Solomon et al., 2020), which in turn, results in the increase of anaerobic metabolism in the mitochondria of brain cells (Abdennour et al., 2012). The accumulation of lactic acid produced by mitochondria can cause swelling of brain cells (Duan et al., 2021; Juzekaeva et al., 2018). In addition, patients recovered from severe COVID-19 have significantly reduced cortical cerebral blood flow (Qin et al., 2021). While no reports on the impact of SARS-CoV-2 on brainECF pH, the accumulation of lactic acid due to anerobic respiration might results in a lower brain pH (Fan et al., 2020). In addition, influenza virus results in a decreased brainECF pH, by H+ export from cells (Liu et al., 2016). Hence, we performed a sensitivity analysis to study the impact of these pathophysiological changes on brain PK with a focus on Cmax and exposure given by the AUC (Supplementary figure 2, online resource 1). An increase of brain cell volume as a result of brain cell swelling will reduce the Cmax and AUC of EIDD1931. A decrease of pHECF slightly decreased the Cmax of brainECF and increased the Cmax and exposure of brainICF. Therefore, based on the sensitivity analysis results and the literature summary of CNS pathophysiology in COVID-19, small changes (10%) of CNS physiology as expected in COVID-19 will not notably impact the brain PK profiles. We therefore postulate that our simulation results using healthy CNS parameters still apply for COVID-19 patients, independent of the disease state of the CNS. In this simulation study, asymmetry factors (AF), which represent active transport activity at BBB, were calculated based on Kpuu,BBB values provided by “the brain exposure efficiency” (BEE) in silico calculator (Gupta et al., 2020) for Nirmatrelvir and Remdesivir, both drugs being P-glycoprotein substrates. The predicted Kpuu,BBB of Remdesivir was in line with total brain-to-plasma Remdesivir ratios measured in radiographic imaging studies in rats (Gilead Sciences, 2020) and in rhesus monkeys (Warren et al., 2016). No in vivo or in vitro data on P-glycoprotein activity were available for Nirmatrelvir. To assess the impact of the uncertainty associated with the predicted Kpuu,BBB (and consequently AF) on brain PK, we explored the scenario assuming a five-fold increase of BBB p-glycoprotein activity (i.e. a five-fold decrease of Kpuu,BBB), Nirmatrelvir still maintained activity against the omicron, but not the delta, variant (results not shown). Future in vitro or preclinical in vivo studies addressing the brain penetration of Nirmatrelvir are required to further substantiate these outcomes. Remdesivir and Molnupiravir are prodrugs of the parent nucleosides and undergo intracellular metabolism to the active nucleoside analogues, GS443902 and EIDD2061, respectively. EC50/90 were derived from in vitro systems that were based on animal and human cell lines and vary by the metabolic capacity of these cell lines depending on the enzyme expression. Protein expression of human brain kinase and HINT1/3 (phosphoramidase enzymes) were comparable or lower to those of human lungs and liver (Sjöstedt et al., 2020), maintaining our earlier conclusion. 5 Conclusion With the LeiCNS-PK3.0 PBPK framework, we predict that with the current dosing regimens only Nirmatrelvir and not Remdesivir or Molnupiravir will reach effective PK against the delta and the omicron variants in the human brain. Our study provides evidence-based guidance for the design of future (pre)clinical studies addressing the antiCOVID-19 drug efficacy in the human CNS. 6 Author Contributions Mohammed AA Saleh, Jeroen Elassaiss-Schaap, Elizabeth CM de Lange contributed to project conceptualization, Mohammed AA Saleh, Ming Sun, and Berfin Gülave performed the data collection, Mohammed AA Saleh and Makoto Hirasawa performed the data analysis and model simulations, Mohammed AA Saleh, Jeroen Elassaiss-Schaap and Elizabeth CM de Lange drafted and reviewed the manuscript. Supplementary materials Supplementary figures 1 and 2 are included in electronic supplementary materials ESM_1.pdf Declaration of Competing Interest Makoto Hirasawa is an employee of Daiichi-Sankyo Co., Ltd. Appendix Supplementary materials Image, application 1 Data availability No data was used for the research described in the article. Funding This research was funded by Leiden Academic Center for Drug Research (LACDR), Leiden University, Leiden, The Netherlands. Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ejps.2022.106345. ==== Refs References Abdennour L. Zeghal C. Dème M. Puybasset L. Interaction cerveau-poumon Ann. Fr. Anesth. Reanim. 31 2012 101 107 10.1016/j.annfar.2012.04.013 Buzhdygan T.P. DeOre B.J. Baldwin-Leclair A. Bullock T.A. McGary H.M. Khan J.A. Razmpour R. Hale J.F. Galie P.A. Potula R. Andrews A.M. Ramirez S.H. The SARS-CoV-2 spike protein alters barrier function in 2D static and 3D microfluidic in-vitro models of the human blood–brain barrier Neurobiol. Dis. 146 2020 105131 10.1016/j.nbd.2020.105131 Chou S.H.Y. Beghi E. Helbok R. Moro E. Sampson J. Altamirano V. Mainali S. Bassetti C. Suarez J.I. McNett M. Global Incidence of Neurological Manifestations among Patients Hospitalized with COVID-19 - A Report for the GCS-NeuroCOVID Consortium and the ENERGY Consortium JAMA Netw. Open 4 2021 1 14 10.1001/jamanetworkopen.2021.12131 Dhont S. Derom E. Van Braeckel E. Depuydt P. Lambrecht B.N. The pathophysiology of ‘happy’ hypoxemia in COVID-19 Respir. Res. 21 2020 1 9 10.1186/s12931-021-01614-1 31898493 Douaud G. Lee S. Alfaro-Almagro F. Arthofer C. Wang C. McCarthy P. Lange F. Andersson J.L.R. Griffanti L. Duff E. Jbabdi S. Taschler B. Keating P. Winkler A.M. Collins R. Matthews P.M. Allen N. Miller K.L. Nichols T.E. Smith S.M. SARS-CoV-2 is associated with changes in brain structure in UK Biobank Nature 604 2022 697 707 10.1038/s41586-022-04569-5 35255491 Duan K. Premi E. Pilotto A. Cristillo V. Benussi A. Libri I. Giunta M. Bockholt H.J. Liu J. Campora R. Pezzini A. Gasparotti R. Magoni M. Padovani A. Calhoun V.D. Alterations of frontal-temporal gray matter volume associate with clinical measures of older adults with COVID-19 Neurobiol. Stress 14 2021 10.1016/j.ynstr.2021.100326 Erickson M.A. Rhea E.M. Knopp R.C. Interactions of SARS-CoV-2 with the Blood – Brain Barrier Int. J. Mol. Sci. 22 2021 1 28 10.3390/ijms22052681 European Medicines Agency Assessment report on the Use of molnupiravir for the treatment of COVID-19 [WWW Document] Eur. Med. Agency 2021 URL https://www.ema.europa.eu/en/documents/referral/lagevrio-also-known-molnupiravir-mk-4482-covid-19-article-53-procedure-assessment-report_en.pdf accessed 2.1.22 European Medicines Agency Assessment report on paxlovid use in COVID-19 [WWW Document] Eur. Med. Agency 2021 URL https://www.ema.europa.eu/en/documents/referral/paxlovid-pf-07321332-ritonavir-covid-19-article-53-procedure-assessment-report_en.pdf accessed 2.1.22 European Medicines Agency Assessment report on the use of Veklury in the treatment of COVID-19 [WWW Document] Eur. Med. Agency 2020 URL https://www.ema.europa.eu/en/documents/assessment-report/veklury-epar-public-assessment-report_en.pdf accessed 2.1.22 Fan H. Tang X. Song Y. Liu P. Chen Y. Influence of covid-19 on cerebrovascular disease and its possible mechanism Neuropsychiatr. Dis. Treat. 16 2020 1359 1367 10.2147/NDT.S251173 32547039 Fidler, M., Hallow, M., Wilkins, J., Wang, W., 2019. RxODE: Facilities for Simulating from ODE-Based Models. Gilead Sciences Pharmacokinetics written summary (Remdesivir) [WWW Document] Prod. Doc 2020 URL https://www.pmda.go.jp/drugs/2020/P20200518003/230867000_30200AMX00455_I100_1.pdf accessed 2.1.22 Gonçalves A. Bertrand J. Ke R. Comets E. de Lamballerie X. Malvy D. Pizzorno A. Terrier O. Rosa Calatrava M. Mentré F. Smith P. Perelson A.S. Guedj J. Timing of Antiviral Treatment Initiation is Critical to Reduce SARS-CoV-2 Viral Load CPT Pharmacometrics Syst. Pharmacol. 9 2020 509 514 10.1002/psp4.12543 32558354 Guadarrama-Ortiz P. Choreño-Parra J.A. Sánchez-Martínez C.M. Pacheco-Sánchez F.J. Rodríguez-Nava A.I. García-Quintero G. Neurological Aspects of SARS-CoV-2 Infection: Mechanisms and Manifestations Front. Neurol. 11 2020 1 14 10.3389/fneur.2020.01039 32116995 Gupta M. Bogdanowicz T. Reed M.A. Barden C.J. Weaver D.F. The Brain Exposure Efficiency (BEE) Score ACS Chem. Neurosci. 11 2020 205 224 10.1021/acschemneuro.9b00650 31815431 Humeniuk R. Mathias A. Cao H. Osinusi A. Shen G. Chng E. Ling J. Vu A. German P. Safety, Tolerability, and Pharmacokinetics of Remdesivir, An Antiviral for Treatment of COVID-19, in Healthy Subjects Clin. Transl. Sci. 13 2020 896 906 10.1111/cts.12840 32589775 Humeniuk R. Mathias A. Kirby B.J. Lutz J.D. Cao H. Osinusi A. Babusis D. Porter D. Wei X. Ling J. Reddy Y.S. German P. Pharmacokinetic, Pharmacodynamic, and Drug-Interaction Profile of Remdesivir, a SARS-CoV-2 Replication Inhibitor Clin. Pharmacokinet. 60 2021 569 583 10.1007/s40262-021-00984-5 33782830 Jarius S. Pache F. Körtvelyessy P. Jelčić I. Stettner M. Franciotta D. Keller E. Neumann B. Ringelstein M. Senel M. Regeniter A. Kalantzis R. Willms J.F. Berthele A. Busch M. Capobianco M. Eisele A. Reichen I. Dersch R. Rauer S. Sandner K. Ayzenberg I. Gross C.C. Hegen H. Khalil M. Kleiter I. Lenhard T. Haas J. Aktas O. Angstwurm K. Kleinschnitz C. Lewerenz J. Tumani H. Paul F. Stangel M. Ruprecht K. Wildemann B. Cerebrospinal fluid findings in COVID-19: a multicenter study of 150 lumbar punctures in 127 patients J. Neuroinflammation 19 2022 1 33 10.1186/s12974-021-02339-0 34980176 Juzekaeva E. Gainutdinov A. Mukhtarov M. Khazipov R. Dynamics of the hypoxia—induced tissue edema in the rat barrel cortex in vitro Front. Cell. Neurosci. 12 2018 1 11 10.3389/fncel.2018.00502 29386999 Krasemann S. Haferkamp U. Pfefferle S. Woo M.S. Heinrich F. Schweizer M. Appelt-Menzel A. Cubukova A. Barenberg J. Leu J. Hartmann K. Thies E. Littau J.L. Sepulveda-Falla D. Zhang L. Ton K. Liang Y. Matschke J. Ricklefs F. Sauvigny T. Sperhake J. Fitzek A. Gerhartl A. Brachner A. Geiger N. König E.M. Bodem J. Franzenburg S. Franke A. Moese S. Müller F.J. Geisslinger G. Claussen C. Kannt A. Zaliani A. Gribbon P. Ondruschka B. Neuhaus W. Friese M.A. Glatzel M. Pless O. The blood-brain barrier is dysregulated in COVID-19 and serves as a CNS entry route for SARS-CoV-2 Stem Cell Reports 17 2022 307 320 10.1016/j.stemcr.2021.12.011 35063125 Liu H. Maruyama H. Masuda T. Honda A. Arai F. The influence of virus infection on the extracellular pH of the host cell detected on cell membrane Front. Microbiol. 7 2016 1 8 10.3389/fmicb.2016.01127 26834723 Matschke J. Lütgehetmann M. Hagel C. Sperhake J.P. Schröder A.S. Edler C. Mushumba H. Fitzek A. Allweiss L. Dandri M. Dottermusch M. Heinemann A. Pfefferle S. Schwabenland M. Sumner Magruder D. Bonn S. Prinz M. Gerloff C. Püschel K. Krasemann S. Aepfelbacher M. Glatzel M. Neuropathology of patients with COVID-19 in Germany: a post-mortem case series Lancet Neurol 19 2020 919 929 10.1016/S1474-4422(20)30308-2 33031735 NCBI Resource Coordinators Database resources of the National Center for Biotechnology Information Nucleic Acids Res 46 2018 D8 D13 10.1093/nar/gkx1095 29140470 Pacheco-Herrero M. Soto-Rojas L.O. Harrington C.R. Flores-Martinez Y.M. Villegas-Rojas M.M. León-Aguilar A.M. Martínez-Gómez P.A. Campa-Córdoba B.B. Apátiga-Pérez R. Corniel-Taveras C.N. Dominguez-García J. de J. Blanco-Alvarez V.M. Luna-Muñoz J. Elucidating the Neuropathologic Mechanisms of SARS-CoV-2 Infection Front. Neurol. 12 2021 1 19 10.3389/fneur.2021.660087 Painter G.R. Bowen R.A. Bluemling G.R. DeBergh J. Edpuganti V. Gruddanti P.R. Guthrie D.B. Hager M. Kuiper D.L. Lockwood M.A. Mitchell D.G. Natchus M.G. Sticher Z.M. Kolykhalov A.A. The prophylactic and therapeutic activity of a broadly active ribonucleoside analog in a murine model of intranasal venezuelan equine encephalitis virus infection Antiviral Res 171 2019 1 10 10.1016/j.antiviral.2019.104597 Painter W.P. Holman W. Bush J.A. Almazedi F. Malik H. Eraut N.C.J.E. Morin M.J. Szewczyk L.J. Painter G.R. Human safety, tolerability, and pharmacokinetics of molnupiravir, a novel broad-spectrum oral antiviral agent with activity against SARS-CoV-2 Antimicrob. Agents Chemother. 65 2021 10.1128/AAC.02428-20 Pellegrini L. Albecka A. Mallery D.L. Kellner M.J. Paul D. Carter A.P. James L.C. Lancaster M.A. SARS-CoV-2 Infects the Brain Choroid Plexus and Disrupts the Blood-CSF Barrier in Human Brain Organoids Cell Stem Cell 27 2020 951 961 10.1016/j.stem.2020.10.001 e5 33113348 Philippens, I.H.C.H.M., Böszörményi, K.P., Wubben, J.A., Fagrouch, Z.C., Driel, N. van, Mayenburg, A.Q., Lozovagia, D., Roos, E., Schurink, B., Bugiani, M., Bontrop, R.E., Middeldorp, J., Bogers, W.M., Geus-Oei, L.-F. de, Langermans, J.A.M., Stammes, M.A., Verstrepen, B.E., Verschoor, E.J., 2021. SARS-CoV-2 causes brain inflammation and induces Lewy body formation in macaques. BIORXIV. https://doi.org/10.1101/2021.02.23.432474. Pruijssers A.J. George A.S. Schäfer A. Leist S.R. Gralinksi L.E. Dinnon K.H. Yount B.L. Agostini M.L. Stevens L.J. Chappell J.D. Lu X. Hughes T.M. Gully K. Martinez D.R. Brown A.J. Graham R.L. Perry J.K. Du Pont V. Pitts J. Ma B. Babusis D. Murakami E. Feng J.Y. Bilello J.P. Porter D.P. Cihlar T. Baric R.S. Denison M.R. Sheahan T.P. Remdesivir Inhibits SARS-CoV-2 in Human Lung Cells and Chimeric SARS-CoV Expressing the SARS-CoV-2 RNA Polymerase in Mice Cell Rep 32 2020 10.1016/j.celrep.2020.107940 Qin Y. Wu J. Chen T. Li J. Zhang G. Wu D. Zhou Y. Zheng N. Cai A. Ning Q. Manyande A. Xu F. Wang J. Zhu W. Long-term microstructure and cerebral blood flow changes in patients recovered from COVID-19 without neurological manifestations J. Clin. Invest. 131 2021 1 12 10.1172/JCI147329 R Core Team R: A language and environment for statistical computing 2019 R Foundation for Statistical Computing Vienna, Austria Reiber H. Flow rate of cerebrospinal fluid (CSF) - A concept common to normal blood-CSF barrier function and to dysfunction in neurological diseases J. Neurol. Sci. 122 1994 189 203 10.1016/0022-510X(94)90298-4 8021703 Reynolds J.L. Mahajan S.D. SARS-COV2 Alters Blood Brain Barrier Integrity Contributing to Neuro-Inflammation J. Neuroimmune Pharmacol. 16 2021 4 6 10.1007/s11481-020-09975-y 33405097 Rosales, R., Rodriguez, M.L., Rai, D.K., Cardin, R.D., Anderson, A.S., Sordillo, E.M., Bakel, H. van, Simon, V., García-Sastre, A., White, K.M., 2022. Nirmatrelvir, Molnupiravir, and Remdesivir maintain potent in vitro activity against the SARS-CoV-2 Omicron variant. bioRxiv 2019. Saleh M.A.A. de Lange E.C.M. Impact of CNS Diseases on Drug Delivery to Brain Extracellular and Intracellular Target Sites in Human: A “WHAT-IF” Simulation Study Pharmaceutics 13 2021 1 17 10.3390/pharmaceutics13010095 Saleh M.A.A. Loo C.F. Elassaiss-Schaap J. De Lange E.C.M. Lumbar cerebrospinal fluid-to-brain extracellular fluid surrogacy is context-specific: insights from LeiCNS-PK3.0 simulations J. Pharmacokinet. Pharmacodyn. 48 2021 725 741 10.1007/s10928-021-09768-7 34142308 Shen W. Logue J. Yang Penghua Baracco L. Elahi M. Reece A. Wang B. Li L. Blanchard T.G. Han Z. Frieman M.B. Rissman R.A. Yang Peixin SARS-CoV-2 invades cognitive centers of the brain and induces Alzheimer's-like neuropathology Bior 2022 10.1101/2022.01.31.478476 Sjöstedt E. Zhong W. Fagerberg L. Karlsson M. Mitsios N. Adori C. Oksvold P. Edfors F. Limiszewska A. Hikmet F. Huang J. Du Y. Lin L. Dong Z. Yang L. Liu X. Jiang H. Xu X. Wang J. Yang H. Bolund L. Mardinoglu A. Zhang C. Feilitzen K.von Lindskog C. Pontén F. Luo Y. Hökfelt T. Uhlén M. Mulder J. An atlas of the protein-coding genes in the human, pig, and mouse brain Science 367 80 2020 eaay5947 10.1126/science.aay5947 - 32139519 Solomon I.H. Normandin E. Bhattacharyya S. Mukerji S.S. Keller Kiana Ali A.S. Adams G. Hornick J.L. Padera R.F. Sabeti P. Neuropathological Features of Covid-19 N. Engl. J. Med. 383 2020 986 989 10.1056/nejmc2001362 32877590 Stein S. Ramelli S. Grazioli A. Winkler C. Dickey J. Platt A. Pittaluga S. Herr D. Mccurdy M. Consortium N.C.-19 A. Peterson K.E. Cohen J.I. Wit E.de Vannella K.M. Hewitt S.M. Kleiner D.E. Chertow D.S. SARS-CoV-2 infection and persistence throughout the human body and brain Res. Sq. 2021 10.21203/rs.3.rs-1139035/v1 Tandon M. Kataria S. Patel J. Mehta T.R. Daimee M. Patel V. Prasad A. Chowdhary A.A. Jaiswal S. Sriwastava S. A Comprehensive Systematic Review of CSF analysis that defines Neurological Manifestations of COVID-19 Int. J. Infect. Dis. 104 2021 390 397 10.1016/j.ijid.2021.01.002 33434662 Tao S. Zandi K. Bassit L. Ong Y.T. Verma K. Liu P. Downs-Bowen J.A. McBrayer T. LeCher J.C. Kohler J.J. Tedbury P.R. Kim B. Amblard F. Sarafianos S.G. Schinazi R.F. Comparison of anti-SARS-CoV-2 activity and intracellular metabolism of remdesivir and its parent nucleoside Curr. Res. Pharmacol. Drug Discov. 2 2021 100045 10.1016/j.crphar.2021.100045 U.S. Food and Drug Administration, 2020. Clinical Pharmacology and Biopharmaceutics Review (Remdesivir) [WWW Document]. U.S. Food Drug Adm. URL https://www.accessdata.fda.gov/drugsatfda_docs/nda/2020/214787Orig1s000ClinpharmR.pdf. 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 Antiviral Res 198 2022 105252 10.1016/j.antiviral.2022.105252 Veleri S. Neurotropism of SARS-CoV-2 and neurological diseases of the central nervous system in COVID-19 patients Exp. Brain Res. 240 2022 9 25 10.1007/s00221-021-06244-z 34694467 Wang, P., Jin, L., Zhang, M., Wu, Y., Duan, Z., Chen, W., Wang, C., Liao, Z., Han, J., Guo, Yingqi, Guo, Yaqiong, Wang, Y., Lai, R., Qin, J., 2021. SARS-CoV-2 causes human BBB injury and neuroinflammation indirectly in a linked organ chip platform. bioRxiv. 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 2016 381 385 10.1038/nature17180 26934220 Wishart D.S. Feunang Y.D. Guo A.C. Lo E.J. Marcu A. Grant J.R. Sajed T. Johnson D. Li C. Sayeeda Z. Iynkkaran N.A.I. Liu Y. Maciejewski A. Gale N. AlexWilson Chin Cummings L. Le R. Pon D. Knox A. Wilson C. DrugBank 5.0: a major update to the DrugBank database for 2018 Nucleic Acids Res 46 2017 D1074 D1082 10.1093/nar/gkx1037 Yang A.C. Kern F. Losada P.M. Agam M.R. Maat C.A. Schmartz G.P. Fehlmann T. Stein J.A. Schaum N. Lee D.P. Calcuttawala K. Vest R.T. Berdnik D. Lu N. Hahn O. Gate D. McNerney M.W. Channappa D. Cobos I. Ludwig N. Schulz-Schaeffer W.J. Keller A. Wyss-Coray T. Dysregulation of brain and choroid plexus cell types in severe COVID-19 Nature 595 2021 565 571 10.1038/s41586-021-03710-0 34153974 Zhang L. Zhou L. Bao L. Liu J. Zhu H. Lv Q. Liu R. Chen W. Tong W. Wei Q. Xu Y. Deng W. Gao H. Xue J. Song Z. Yu P. Han Y. Zhang Y. Sun X. Yu X. Qin C. SARS-CoV-2 crosses the blood–brain barrier accompanied with basement membrane disruption without tight junctions alteration Signal Transduct. Target. Ther. 6 2021 10.1038/s41392-021-00719-9 World Health Organization (WHO), 2022. Tracking SARS-CoV-2 variants. URL https://www.who.int/en/activities/tracking-SARS-CoV-2-variants/(accessed 4.6.22).	36462547	PMC9710098	NO-CC CODE	2022-12-15 23:17:45	no	Eur J Pharm Sci. 2023 Feb 1; 181:106345	utf-8	Eur J Pharm Sci	2,022	10.1016/j.ejps.2022.106345	oa_other
==== Front Am J Obstet Gynecol MFM Am J Obstet Gynecol MFM American Journal of Obstetrics & Gynecology Mfm 2589-9333 Elsevier Inc. S2589-9333(22)00260-9 10.1016/j.ajogmf.2022.100830 100830 Research Letter Research Letter SARS-CoV-2 Vaccine Booster Elicits Robust Prolonged Maternal Antibody Responses and Passive Transfer to the Offspring via the Placenta and Breastmilk MARSHALL Nicole E. MD 1⁎⁎ BLANTON Madison B. BS 23 DORATT Brianna M. MS 2 MALHERBE Delphine C. PhD 2 RINCON Monica MD 1 TRUE Heather PharmD 23 MCDONALD Taylor 2 BEAUREGARD Caroline 2 ADATORWOVOR Reuben PhD 4 MESSAOUDI Ilhem PhD 2⁎ 1 Department of Obstetrics and Gynecology, Oregon Health & Science University, Portland, OR 2 Department of Microbiology, Immunology, and Molecular Genetics, College of Medicine, University of Kentucky, Lexington, KY 3 Department of Pharmaceutical Sciences, University of Kentucky, Lexington, KY 4 Department of Biostatistics, University of Kentucky, Lexington, KY ⁎ Corresponding author: Dr. Ilhem Messaoudi, University of Kentucky, Department of Microbiology, Immunology, and Molecular Genetics, College of Medicine, 760 Press Ave, Lexington, KY 40536, United States ⁎⁎ Corresponding author: Dr. Nicole Marshall, Department of Obstetrics and Gynecology, Oregon Health & Science University, 3181 SW Sam Jackson Park Rd, Portland, OR 97239, United States 30 11 2022 30 11 2022 10083023 11 2022 28 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. ==== Body pmcINTRODUCTION Infection during pregnancy can lead to adverse outcomes for both pregnant persons and offspring1 as observed during the SARS-CoV-2 global pandemic. Adverse outcomes can be mitigated by maternal vaccination, protecting the pregnant person and the neonate/infant via passive transfer of maternal antibodies either in utero via the placenta or after birth via breastmilk.2 Immunoglobulins G (IgG) transfer from maternal to fetal circulation via neonatal plasma Fc receptors (FcRN) in the placenta and fetal intestines.2 The Centers of Disease Control and Prevention (CDC) recommend vaccination against SARS-CoV-2 for persons who are pregnant or plan to become pregnant.3 Despite this recommendation, there remains a high level of vaccine hesitancy among the pregnant population.4 Vaccination decisions during pregnancy are often influenced by a primary goal of protecting neonatal health. Thus, the decision to vaccinate during pregnancy or to delay vaccination is shaped by knowledge about impact of vaccine timing and duration of protection. Previous studies investigating maternal SARS-CoV-2 vaccination included minimal longitudinal sampling and focused on one compartment (e.g., maternal blood, or breastmilk). In this study, we assessed the antibody response throughout gestation, at birth, and up to 12 months post-partum in a cohort of 121 women in maternal circulation, cord blood (UCB), newborn blood (NB), and breastmilk. Study Design The study was approved by the IRB of Oregon Health & Science University and the University of Kentucky. From March 2021 until June 2022, maternal blood and breastmilk samples were obtained longitudinally from 121 SARS-CoV-2 vaccinated participants. The overwhelming majority (90.9%) of participants received the Pfizer BN162b2 vaccine ( Table 1 ). Umbilical cord and maternal blood were collected at the time of delivery while newborn blood and colostrum were collected within 48 hours of delivery (Figure 1 A). An indirect ELISA was used to determine the IgG (total and subclasses) end-point titer (EPT) of antibodies against SARS-CoV-2 receptor-binding domain (RBD) of the spike protein in plasma, whereas breastmilk antibody levels were reported as optical density (OD) values.Table 1 Cohort Metadata Subjects are stratified by the trimester of initial maternal SARS-CoV-2 vaccination. Maternal age and gestational age at delivery are mean ± standard deviation. There is no significant difference among maternal age nor gestational age of delivery within the cohort when stratified by vaccination timepoint Table 1:All (121) Pre-pregnancy T1 T2 T3 Post-Partum N (%) 15 (12.4%) 15 (12.4%) 36 (29.8%) 27 (22.3%) 28 (23.1%) Maternal age (years) 34.9 ± 3.7 34.5 ± 3.1 34.2 ± 4.1 34.8 ± 4.7 33.6 ± 4.6 Gestational age at delivery (years) 38.9 ± 0.9 39.2 ± 1.2 38.9 ± 1.3 39.0 ± 2.0 39.0 ± 1.2 Fetal sex (% female) 33% 40% 47% 52% 36% Initial Vaccine Series Pfizer 13 13 34 26 24 Moderna 2 2 2 1 4 Received booster 14 12 27 23 15 Days post second dose that the booster was received 255 ± 29 275 ± 29 219 ± 33 234 ± 33 240 ± 31 Figure 1 (A) Experimental design to investigate the impact of maternal SARS-CoV-2 vaccination on passive transmission of RBD-specific IgG antibodies by assessing antibody titers in maternal plasma, UCB, newborn plasma, and breastmilk. (B) RBD-specific IgG antibody titers in maternal plasma relative to days post first vaccination (n=370 samples). (C) RBD-specific IgG antibody titers 50.59 ± 4.46 days before and 55.74 ± 4.14 days after booster dose (n=77 pairs). (D) RBD-specific IgG antibody titers in maternal plasma relative to days post booster dose (n=112). (E) IgG isotype levels 84.07 ± 12.34 days before and 58.47 ± 8.98 days after the booster dose (n=15 pairs). (F) RBD-specific IgG levels in breastmilk after the first and second vaccine doses (n=179). (G) Breastmilk IgG levels 37.84 ± 3.80 days prior to and 55.32 ± 5.30 days after booster (n=45 pairs). (H) RBD-specific IgG antibodies in breastmilk after maternal booster vaccination (n=123). (I) Levels of RBD specific IgG isotypes in breastmilk 57.50 ± 8.17 days before (n=28) and 117.23 ± 11.32 days after the booster dose (n=44). (J) RBD-specific IgG titers in maternal circulation and umbilical cord plasma at delivery (n=45 pairs). (K) Correlation between UCB and maternal RBD-specific IgG titers at delivery (n=45). (L) RBD-specific IgG titers in UCB relative to days since maternal first vaccine dose (n=48). (M) Overall comparison between maternal RBD-specific IgG antibodies at delivery and newborn RBD-specific IgG titers, independent of trimester of initial vaccination (n=35 pairs). (N) Correlation (n=35) of RBD-specific IgG titers in newborn and maternal plasma at delivery. (O) RBD-specific IgG titers in newborn plasma relative to days post maternal vaccination. Bar graphs show median values with the standard error of the mean (SEM). ∗ p < 0.03, ∗∗ p < 0.002, ∗∗∗ p < 0.0002, ∗∗∗∗ p<0.0001. Figure 1: Results Maternal plasma RBD-specific IgG titers strongly inversely correlated with time elapsed since first vaccination (r=0.07043 p<0.0001, half-life 56.45 days). (Figure 1 B). After the booster dose, RBD-specific IgG titers increased significantly (p<0.0001) (Figure 1 C) and exhibited a longer half-life of 128.12 days (Figure 1 B, 1 D). The booster produced a significant increase in all four IgG isotypes measured in maternal plasma (Figure 1 E). The initial 2-dose vaccination regimen resulted in detectable IgG antibody response in breastmilk (albeit reduced levels compared to maternal plasma) with a half-life of 61.34 days (Figure 1 F). Comparable to maternal circulation, the booster produced a significant increase in antibody levels (p<0.0001) (Figure 1 G) and half-life (124.67 days) (Figure 1 H). After the booster, IgG1 and IgG4 increased significantly, with IgG4 becoming dominant (Figure 1 I). RBD-specific IgG antibodies were detected in UCB plasma albeit at significantly lower levels than in maternal circulation at delivery (p=0.0012) (Figure 1 J). Interestingly, there was no correlation between UCB RBD-specific IgG titers and maternal titers at delivery ( Figure 1 K) or time since maternal first vaccination (Figure 1 L). Although UCB is often used as surrogate for newborn blood (NB), there may be differences in antibody transfer into UCB and fetal circulation. As described for UCB, titers in NB were lower than those in maternal circulation at delivery (p=0.0200) (Figure 1 M). In contrast to UCB, a significant positive correlation was observed between paired NB and maternal plasma titers at delivery (r=0.3782 p=<0.0001) (Figure 1 N). Moreover, NB titers were inversely correlated with the time since initial maternal vaccination (r=0.3130 p=0.0002) (Figure 1 O) with lower newborn IgG antibody titers in infants born to mothers vaccinated during early pregnancy. Conclusion Our results confirm the initial two dose vaccination series during gestation resulted in appreciable RBD-specific IgG response in maternal circulation, UCB, NB, and breastmilk. Longitudinal analysis of post-partum samples indicates the booster dose is essential for eliciting higher and more durable antibody levels in both maternal circulation and breastmilk. SARS-CoV-2-specific maternal antibodies generated via vaccination are passively transferred in utero and after birth via breastfeeding but wane within 6 months after first vaccination dose. Our longitudinal data indicate that breastmilk antibody levels are dramatically increased by the booster. Therefore, the best neonatal protection against SARS-CoV-2 is for pregnant persons to receive the 3-dose vaccination series at any point during pregnancy to allow for placental antibody transfer, and to subsequently breastfeed their children for at least 6 months, at which point infants are eligible for SARS-CoV-2 vaccination. 5 Continued breastfeeding throughout the first year of life is encouraged as SARS-CoV-2-specific maternal antibody levels persist in breastmilk following booster for at least 12 months. REFERENCES 1 Megli CJ, Coyne CB. Infections at the maternal–fetal interface: an overview of pathogenesis and defence. Nature Reviews Microbiology 2022;20:67-82. 2 Roopenian DC, Akilesh S. FcRn: the neonatal Fc receptor comes of age. Nature Reviews Immunology 2007;7:715-25. 3 Prevention CfDCa. COVID-19 Vaccines While Pregnant or Breastfeeding, 2022. 4 Kharbanda EO, Vazquez-Benitez G. COVID-19 mRNA Vaccines During Pregnancy. JAMA 2022;327:1451. 5 Administration USFaD. Coronavirus (COVID-19) Update: FDA Authroizes Moderna and Pfizer-BioNTech COVID-19 Vaccines for Children Down to 6 Months of Age, 2022. Declaration of Competing Interest: The authors report no conflict of interest. Funding: This research was supported by grants from the National Institutes of Health R01AI145910 (IM), R01AI142841 (IM). The funding source had no involvement in study design; in the collection, analysis and interpretation of data; in the writing of the report, and in the decision to submit the article for publication. The findings in the article will be presented at the 50th Annual Autumn Immunology Conference presented by Autumn Immunology, Inc. on November 18th-21st, 2022 in Chicago, Illinois.	36462615	PMC9710099	NO-CC CODE	2022-12-01 23:23:40	no	Am J Obstet Gynecol MFM. 2022 Nov 30;:100830	utf-8	Am J Obstet Gynecol MFM	2,022	10.1016/j.ajogmf.2022.100830	oa_other
==== Front Arab J Chem Arab J Chem Arabian Journal of Chemistry 1878-5352 1878-5379 The Author(s). Published by Elsevier B.V. on behalf of King Saud University. S1878-5352(22)00784-5 10.1016/j.arabjc.2022.104468 104468 Review Article Chloroquine chaos and COVID-19: Smart delivery perspectives through pH sensitive polymers/micelles and ZnO nanoparticles Manuja Anju ⁎ Chhabra Dharvi Kumar Balvinder ⁎ ICAR-National Research Centre on Equines, Hisar, Haryana 125001, India ⁎ Corresponding authors at: ICAR-National Research Centre on Equines, Hisar-125001, Haryana, India. 30 11 2022 2 2023 30 11 2022 16 2 104468104468 10 5 2022 24 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. Graphical abstract The global pandemic of COVID-19 had a consequential impact on our lives. (Hydroxy)chloroquine, a well-known drug for treatment or prevention against malaria and chronic inflammatory conditions, was also used for COVID patients with reported potential efficacy. Although it was well tolerated, however in some cases, it produced severe side effects, including grave cardiac issues. The variable reports on the administration of (hydroxy)chloroquine in COVID19 patients led to chaos. This drug is a well-known zinc ionophore, besides possessing antiviral effects. Zinc ionophores augment the intracellular Zn2+ concentration by facilitating the zinc ions into the cells and subsequently impair virus replication. Zinc oxide nanoparticles (ZnO NPs) have been reported to possess antiviral activity. However, the adverse effects of both components are also reported. We discussed in depth their possible mechanism as antiviral and smart delivery perspectives through pH-sensitive polymers/ micelles and ZnO NPs. Keywords ZnO nanoparticles COVID 19 Chloroquine pH responsive Micelles Polymers ==== Body pmc1 Introduction Coronaviruses are enveloped, single-stranded positive-sense RNA viruses that cause diseases in mammals and birds. In humans, coronaviruses cause respiratory disorders ranging from mild to lethal infections. They became a matter of concern for the whole world due to their lethality and transmissibility from person to person. Transmissible gastroenteritis virus in swines, bovine coronavirus in cattle, and infectious bronchitis virus in fowls are of veterinary significance (Weiss and Navas-Martin, 2005). The world observed the infection in humans as “Severe-acute-respiratory syndrome (SARS-CoV)”, “Middle-East-respiratory syndrome coronavirus (MERS-CoV)”, “SARS CoV2” in 2002, 2012, 2019 respectively (Rathore and Ghosh, 2020, Wang et al., 2013). SARS and MERS were both responsible for high mortality rates. Involvement of multi-organs and variable symptoms has been reported in SARS CoV2/COVID-19 affected patients. The most common manifestations reported in COVID-19 patients are fever, dry cough, and tiredness (Wu et al., 2020) and the people showing only these signs generally recovered without hospitalization. However, diabetes is associated with the severity of the disease in COVID-19 (Huang et al., 2020). Severe symptoms include shortness of breath, loss of speech, loss of movement, hypoxemia, shocks, heart injury, renal injury, etc (Mehta et al., 2021). In addition to this, bacterial and fungal infections may also attack COVID-19 patients. Currently, there is no treatment for coronavirus, and it is based on symptoms only. However, according to a recent study, three new oral antivirals (molnupiravir/fluvoxamine/paxlovid) are effective in lowering the mortality and hospitalization rates of COVID-19 patients (Wen et al., 2022). Chloroquine phosphate, an aminoquinoline, is an old drug known mainly for the treatment of protozoan disease, malaria, and lingering inflammatory ailments (rheumatoid arthritis (RA), systemic lupus erythematosus, etc.). Another related drug, hydroxychloroquine which is also derived from quinoline molecules, is meant for similar clinical conditions. Hydroxychloroquine differs from chloroquine owing to the hydroxyl group (OH) and is preferred over chloroquine in its use against malaria & viral diseases due to its low ocular toxicity. Both these drugs demonstrated inhibition of SARS COV2 in vitro (Wang et al., 2020, Liu et al, 2020, Yao et al., 2020, Vincent et al., 2005). The drug has shown noticeable efficacy against COVID-19 associated pneumonia in China's clinical trials (Gao et al., 2020, Jie et al., 2020, Gautret et al., 2020). Hydroxychloroquine and chloroquine phosphate were consequently advocated for prophylactic and therapeutic usage for COVID-19 patients by the Indian Council of Medical Research, India and the National Health Commission, China (Gautret et al., 2020, Jie et al., 2020). (Hydroxy) chloroquine emerged as a potential drug for the treatment and prevention of the SARS COV2 infection in many countries (China, France, USA, and India) (Gao et al., 2020, Jie et al., 2020, Gautret et al., 2020). Initial medical information confirmed that (hydroxy)chloroquine avoided the aggravation of pneumonic cases and, reduced the viral load in SARS COV2 infected patients. Hydroxychloroquine in combination with azithromycin was granted restricted emergency-use approval by US-FDA to deal with COVID-19 patients (Gautret et al., 2020). A total of 290 clinical trials have been done on hydroxychloroquine, of which 96 have been completed (http.//Clinical trials.gov, 2022). Forty studies on hydroxychloroquine were terminated; 8 studies were suspended; 42 studies had been withdrawn, and 57 were considered as unknown status. Out of 96, only two studies were found to be associated with zinc. Clinical trials conducted on hydroxychloroquine in a few countries yielded variable outcomes. Some studies have indicated the benefits of chloroquine/ hydroxychloroquine as therapeutics for COVID-19, whereas other trials have reported adverse effects that have been discussed in detail elsewhere (Bansal, 2021). The World Health Organization suspended global trials of chloroquine, and its derivative hydroxychloroquine in treating the global pandemic of COVID-19 given its potentially toxic effects, but later resumed the solidarity trials for the drug. The emergency authorization of chloroquine/hydroxychloroquine use was then suspended due to a lack of efficacy in clinical trials by the FDA and other regulatory agencies. Lancet editors provided an expression of concern for the retracted study “Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis” (Lancet, n.d.) which failed to verify a positive impact of hydroxychloroquine or chloroquine, either alone or in combination with a macrolide, on COVID-19 in-hospital outcomes. When used to treat COVID-19, each of these medication regimens was linked to a lower rate of in-hospital survival and a higher incidence of ventricular arrhythmias. Four repurposed antiviral medications, including hydroxychloroquine and three others (remdesivir, lopinavir, and interferon β-1a), were suggested for mortality studies by World Health Organization expert committees in patients who were hospitalized with COVID-19 (WHO Solidarity Trial Consortium, 2021). The inefficacy of hydroxychloroquine in those trials was not due to the lack of antiviral activity of the molecule, which was repeatedly and reproducibly shown by many groups throughout the World (Liu et al., 2020). The reasons for the lack of efficacy of hydroxychloroquine in the clinical trials are two fold: (i) Cells from the lower respiratory airways become insensitive to chloroquine/hydroxychloroquine due to activation of transmembrane serine protease 2 (TMPRSS2) (Ou et al., 2021). Therefore, it is expected to become inactive in the later stages of the disease when pneumonia is developed. ii) The virus-inhibitory concentrations only in part match the tissue and, especially, plasma concentrations of the drug. Therefore, mixed results were expected in the early stages of COVID-19, when the virus remains in the nasopharynx (Tarek and Savarino, 2020). The further description about the efficacy of the drug against the virus is given in the next section under possible mechanism of (hydroxy)chloroquine action against SARS COV2. 2 Possible mechanism of (hydroxy)chloroquine action against SARS COV2 Numerous phases in the coronaviral lifecycle have been proposed to be restricted by (hydroxy)chloroquine, but it's inhibitory effect on viral entrance as a lysosomotropic drug is well described. (Hydroxy)chloroquine comprises an ‘amino group’ connected to a ‘quinoline ring’, which are feeble, susceptible ‘diprotic bases’ that possibly accrue within intra-cellular acidic cubicles including lysosomes (Kaufmann and Krise, 2007). It augments the lysosomes’ pH and causes enlargement, vacuole formation, and impaired lysosomes’ function (Yoon et al., 2010). Both receptor activation and fusion activation by proteolytic processing of the S glycoproteins are necessary for coronavirus entrance. SARS CoV2 requires the angiotensin-converting enzyme 2 receptor (ACE 2) for entrance and the serine protease TMPRSS2 for S protein priming. The drug increases endosomal and lysosomal pH, which in turn prevents cathepsin L from functioning as one of coronaviruses' entrance factors. It effectively interferes with the glycosylation of its host receptor, ACE 2, suggesting similar effects of the drug against this virus at its point of entry into the cells (Hoffmann et al., 2020). These drugs, after accumulating within the intracellular compartment and raising the endosomal pH, rapidly transport into acidic vesicles, and impede terminal glycosylation. Consequently, this hinders the fusion procedure and puts off the release of viral RNA particles, therefore inhibiting the multiplication of viruses intracellularly, thus restricting the virus's survival (Al-Bari, 2017, Akpovwa, 2016). The first pathway is dependent on the endosomal protease cathepsin and susceptible to the drug, while the second process relies on TMPRSS2, which is suggested to be unaffected by hydroxychloroquine (Ou et al., 2021). The literature reports that chloroquine inhibits pH-dependent multiplication of human immunodeficiency, influenza, dengue, Japanese encephalitis, West Nile and Zika viruses and thus restricts their infection (Tsai et al., 1990, Ooi et al., 2006, Farias et al., 2013, Boonyasuppayakorn et al., 2014, Zhu et al., 2012, Delvecchio et al., 2016). The drug also diminishes the expression of phosphatidyl inositol-binding clathrin assembly (PICALM), an abundant protein in clathrin-coated pits (Hu et al., 2020). PICALM regulates the size of the clathrin cage which senses and drives membrane curvature, thus controlling the pace of endocytosis. Literature also suggests that drug acts by hampering quinine reductase 2, which is required for sialic acid biosynthesis (Kwiek et al., 2004). Sialic acid was involved in the binding of MERS-CoV and hCoV-OC43 and its access into host cells, and thus chloroquine hinders the entry of viruses into the host cell via this mechanism. However, the effects on viral glycosylation have been hypothesized to be not a direct consequence of it's effects on quinone reductase 2 but, rather, on the possible effects of the drug on UDP-N-acetylglucosamine 2-epimerases, which are structurally related but distinct enzymes. If the drug is provided early enough, (hydroxy)chloroquine may have an impact on the magnitude of the viral load peak/viral clearance (i.e., when the virus is still confined within the pharyngeal cavity). This mechanism has however been derived from bioinformatic results and is still hypothetical at present. Chloroquine/ hydroxychloroquine's effect can only be completely understood when taking into account its ability to raise the virucidal effect on SARS CoV2-infected cells, in this case through boosting the cell-mediated immune response. Another merit of chloroquine includes its activity to alter iron metabolism, affecting its homeostasis at several stages (Roldan et al., 2020). It suppresses the iron release from transferrin incorporated in endosomes. Iron is essential for the production of deoxyribonucleotides, through the iron-dependent enzyme deoxyribonucleotide reductase. Chloroquine/hydroxychloroquine inhibits lymphocyte proliferation by engaging in this action. It also induces cellular iron starvation which is considered as a possible inhibition of SARS COV2 and useful modulation of immune responses. Hormone hepcidin is increased during infection and inflammation, chloroquine reduces this hormone release and inflammatory cytokines, thus help in recovering anemia and thrombosis (Roldan et al., 2020). Chloroquine and hydroxychloroquine are powerful ‘anti inflammatory’ drugs and ‘immunomodulators’. These were broadly used for decades as a remedial measure for ‘malaria’ and ‘autoimmune diseases’ inclusive of ‘rheumatoid arthritis’, ‘systemic lupus erythematosus’, etc. They can lower the production of proinflammatory cytokines, such as ‘tumor necrosis factor α (TNFα)’, interleukins ‘IL-1, IL-6 and, interferon-γ’ (Jang et al., 2006, Picot et al., 1991, Van den Borne et al., 1997, Sperber et al., 1993). Moreover, they inhibit the innate immune response by (i) blocking the interaction of cytosolic DNA with the nucleic acid sensor cGAMP synthase (Zhang et al., 2020), (ii) interaction of ‘toll-like receptors’ with ‘nucleic acid ligands’ (Hong et al., 2004, Zhu et al., 2012, Yasuda et al., 2008). In severe COVID-19 infection, the hyperactive immune response is responsible for pneumonia. Hydroxychloroquine is also a lysosomotropic inhibitor, an interferon blocker, which decreases the TNFα release, and suppresses TNF receptors of monocytes, thus reducing the inflammatory immune response to viral infection. 3 Adverse side effects of (hydroxy)chloroquine The most widely recognized unsafe impacts of (hydroxy) chloroquine are gastrointestinal side effects, like nausea, vomiting, and stomach distress, unusually hepatotoxicity, blindness/visual complications, toxic epidermal necrolysis, cardiotoxicity; and, very rare urticaria, ototoxicity, and neurological symptoms (Munster et al., 2002). The incidence of visual complications is typically uncommon. It’s binding to melanin can lead to ocular pigmentation, leading to visualization-related complications (Yam and Kwok, 2006, Michaelides et al., 2011, Wolfe and Marmor, 2010). Proximal myopathy related to respiratory failure in older patients receiving any of these drugs due to ceaseless rheumatoid arthritis or immune system disorders has also been reported (Siddiqui et al., 2007, Kwon et al., 2010, Abdel-Hamid et al., 2008, Becerra-Cunat et al., 2003). (Hydroxy) chloroquine-associated cardiac disease is an exceptional but may cause severe adverse effect resulting in death (Chatre et al., 2018). Human-ether-a-go-go-related gene (hERG) dysfunction causes chronic QT syndrome and sudden death, which occurs in patients with cardiac ischemia. Although small doses of (hydroxy)chloroquine are generally safe, both of which can block the hERG channel (Giudicessi et al., 2020, Naksuk et al., 2020). Adversity is dose-dependent and may vary from person to person. It has been reported that “an amount of 600 mg increased the mean ‘QTc’ by 6.1 ms, whereas a 1200 mg dose led to the increase of the mean QTc by 28 ms” (Pukrittayakamee et al., 2014, Mzayek et al., 2007, Zhang et al., 2020). Long-term use of (Hydroxy)chloroquine has been reported to cause prolonged QT intervals and severe arrhythmia (Chen et al., 2006, Stas et al., 2008). These drugs are reported to block the Ik current, slow down the rate of deactivation and increase the transport of hERG protein. In addition, if (hydroxy)chloroquine is combined with CYP3A4 inhibitors like anti-flu medicines like lopinavir/ritonavir/azithromycin, the risk of QT prolongation may increase and could cause serious arrhythmias by blocking or inhibiting cardiac potassium channels (Zhang et al., 2020, Zequn et al., 2021, Wu et al., 2020). The possible mechanism of SARS COV2 infection causing pathologies in multiple organs is shown in Fig. 1 .Fig. 1 Possible mechanism of SARS COV2 infection causing pathologies in multiple organs (Violet arrows). Black dotted arrows show the (hydroxy) chloroquine therapeutic action to compete with receptors on different organs of SARS COV2 affected individuals and potential toxic effects. The left side of the figure shows its effect on the hERG channel inducing prolonged QT interval and cardiac arrhythmia. 4 Zinc oxide nanoparticles Due to the toxic side effects of (hydroxy)chloroquine or their combinations with anti-flu medicines, there is a pressing need for new, and safe chemotherapeutic agents. The development of potential therapeutics requires understanding the entry of the virus into the cells, a key element in viral infection. Various researchers have elucidated insights into the endocytic path and the autophagy route in viral entrance and viral multiplication. As an outcome, the endosomes and lysosomes are considered significant targets for developing therapeutic strategies to combat diseases caused by CoVs. Cellular enzymes and transcription factors use zinc ions to play a crucial role in various activities. Intracellular zinc concentrations can constrain RNA-dependent RNA polymerases and other proteins critical for completing various viral life cycle stages. Zinc has an intrinsic antiviral property, as mentioned above in the article. Zinc deficiency is responsible for 16 % of all deep respiration infections globally (Wessels et al., 2020, World Health Organization, 2002) suggesting the association of zinc deficiency with the threat of infection and excessive development of COVID-19. Zinc supplementation improves mucociliary clearance (Darma et al., 2020) strengthens epithelial integrity (Roscioli et al., 2017) reduces viral replication (Hamdi et al., 2021), and maintains antiviral immunity (Razzaque, 2021, Maares and Haase, 2016). As a result, it reduces lung damage and secondary infections (Wessels et al., 2020, Razzaque, 2021). Given the well-established therapeutic use of ZnO against herpes simplex and influenza viruses (Tavakoli et al., 2018, Ghaffari et al., 2019), exploiting ZnO NPs against the virus would be a potential approach, due to the unique and distinctive characteristics of nanoparticles as compared to conventional materials. ZnO NPs possess attractive optical, piezoelectric, magnetic, and sensing characteristics. Due to nano size, their outer surfaces can be tailored to present cationic, anionic, polar, nonpolar, or neutral faces to the nearby milieu (Kim, 2007). ZnO NPs interact electro-statically with viral-like proteins using a core–shell model approach (Phan and Hoang, 2019). ZnO NPs bind with DNA/RNA, preferably with the ring nitrogen atom or top position of the nucleobases. A recent report regarding the inactivation of the H1N1 influenza virus by ZnO NPs (Ghaffari et al., 2019) is influential and suggests exploiting its antiviral activity against coronaviruses. In silico molecular docking suggested the possible interactions between ZnO NPs and COVID-19 targets, including ACE 2 receptors, RNA-dependent RNA polymerase, and COVID-19 protease (Hamdi et al., 2021). The most likely mechanism includes a potential attachment with viral virions (spike proteins) and blocking the host receptors ACE 2 for interaction, internalization of ZnO NPs inhibit early viral replication cycle followed by the release of zinc ions which disturbs the plasmid/viral integrity; lastly, reactive oxygen species are produced photo catalytically to potentially deprive lipid/ protein, and nucleic structure of SARS CoV2. At very low concentrations (10 µM), zinc ions can inhibit ACE 2′s capacity to metabolize substrates to provide antiviral effects (Sportelli et al., 2022). Zinc ionophores augment the intracellular zinc concentration by facilitating the zinc ions into the cells and subsequently impairing virus replication. Chloroquine phosphate acts as a zinc ionophore and it directs zinc to lysosomes (Xue et al., 2014). Both zinc and chloroquine are FDA-approved and readily available. In another molecular docking investigation, the binding site interaction showed that the communication between ‘Zn (Chloroquine) Cl2 (H2O)’ and the ‘protease’ of SARS CoV2 showed 3 hydrogen bonds, while the ‘Zn (hydroxychloroquine)Cl2 (H2O)’ depicted the solid binding to ‘protease’ receptors due to 8 hydrogen bond formation (Hussein and Elkhair, 2021), which is suggestive of the strong binding of hydroxychloroquine as compared to chloroquine. Chloroquine is water soluble and has two basic groups that correspond to the “quinoline-ring nitrogen” and the “diethylamino side-chain nitrogen,” with an ionization constant’ of 8.1 and 10.2′, respectively. The ability of chloroquine to traverse biological membranes and accumulate in acidic organelles is connected to both its acid-base characteristics and the protonation state of nitrogen atoms, which are closely tied to the chloroquine coordination (Paulikat et al., 2022). At 7.2–7.4 (physiological pH), chloroquine binds to Zn2+ via the ‘quinoline-ring nitrogen’ in a tetrahedral complex forming the coordination sphere to produce either a ‘zwitterionic’ complex which is stable at pH 7 (neutral) or ‘cationic’ complex. The metal coordination is lost at somewhat low pH below 6, suggesting that Zn2+ ions are released into the lysosomal lumen. Fig. 2 depicts the schematic illustration of binding of zinc ions with quinoline nitrogen of chloroquine to form chloroquine-zinc complex. Eighteen percent of chloroquine is monoprotonated at physiologic pH, but it is still lipid soluble and may penetrate cell membranes. Chloroquine that has been biprotonated and is present in lysosomes at pH 4–5 is sequestered and prohibited from returning to the cytoplasm. Although the quantity of free chloroquine in the blood is negligible at physiologic pH, this form of the drug determines its distribution between body tissues and blood.Fig. 2 Schematic illustration of binding of zinc ions with quinoline nitrogen of chloroquine to form chloroquine-zinc complex. Zinc also exhibits anti-inflammatory properties barring NF-kB signaling and regulatory T-cell roles that may limit the cytokine storm. Zinc ions interact directly with the hERG channel, and the interaction leads to an adjustment of the channel deactivation mechanism (Anumonwo et al., 1999, Piscopo and Brown, 2018). It has been reported that it significantly reduces the rate of hERG current inactivity during the stabilization, accelerates the closure of the channel during renewable tails, and shows significant fluctuations in current performance or power dependencies. It is tough to obtain dispersible ZnO or ZnO NPs in aqueous solutions. In our work, we developed a novel method for preparing dispersible suspensions of ZnO NPs possessing flowerlike morphology with good dispersion and high yield in a short period (Manuja et al., 2020). The dispersed ZnO NPs may penetrate easily through the host cell membrane to combat the virus particles. Although nanoparticles hold novel properties that can enhance their efficacy, but they can be toxic when they get in touch with biological systems. Numerous unique characteristics of NPs, including their size, shape, charge, crystal structure, surface area, sensitivity to certain cell types, and other characteristics, impact both the toxicity of NPs as well as their mode of action in biological applications (Manuja et al., 2021). For instance, DNA has a diameter of 2 nm, but the typical cell membrane thickness is approximately10 nm; as a result, the particle size will aid with the possible internalization of NP within a cell. Similarly, the surface charge drives a predictable aspect of NP uptake in cells. Cationic NPs stimulate endocytosis by expressing affinity for anionic phospholipid membranes. When exposed to a biological environment, ZnO NPs have a tendency to scatter and release ions, which causes the generation of reactive oxygen species (ROS) and oxidative stress (Raguvaran et al., 2017). ZnO NPs' toxicity has been thoroughly investigated and has been proven to have an impact on a variety of cell types and animal systems. We have already reported the toxicity of ZnO NPs (Raguvaran et al., 2017, Raguvaran et al., 2015) and the resolving through polymeric delivery in our previous work (Raguvaran et al., 2017, Chopra Meenu et al., n.d., Raguvaran et al., 2017, Manuja et al., 2020). To address the toxicity issues, the proper delivery of ZnO NPs along with (hydroxy)chloroquine (as an ionophore and antiviral property) may pave the way for more efficient therapy. 5 Delivery options Carriers/nanocarriers play a vital role in drug delivery to overcome the toxic effects. The primary benefit of the delivery vehicle is reducing the adverse effects and improving therapeutic efficacy at low concentrations. It should have proven characteristics of enhancing the drug bioavailability and prolonging the duration in blood circulation with sustained release and targeting ability. Liver, spleen, bone marrow, and lung tissues serve as the main route of elimination. Large numbers of phagocytic cells (such as macrophages), which identify nanoparticles as foreign objects and effectively remove them from the bloodstream, are present in these tissues. The opsonization of the nanoparticle by serum proteins, such as immunoglobulins and complement proteins, which results in more effective phagocyte recognition, can increase the pathway's effectiveness. Contrarily, by making nanoparticles more “stealthy” through methods like PEG conjugation, this clearance mechanism can be delayed (Klibanov et al., 1990). The PEGylation can alter the hydrophilicity, diameter, and shape of nanoparticles, among other physicochemical characteristics. Nanomaterials used for delivery have been shown to lower the toxicity of antiviral agents (Lembo and Cavalli, 2010, Cojocaru et al., 2020). Since (hydroxy) chloroquine is a zinc ionophore besides an antiviral agent. It is pertinent to focus on Zn/chloroquine's delivery applications like (i) copolymer micelles, (ii) pH-sensitive delivery, and (iii) encapsulation of the drugs or molecules or their combinations. We have demonstrated that when metal nanoparticles are added to a polymer hydrogel matrix, their toxicity is reduced and their efficacy is improved due to sustained and controlled release (Raguvaran et al., 2017, Chopra Meenu et al., n.d., Raguvaran et al., 2017, Manuja et al., 2020).(i) Copolymer micelles Copolymer micelles are quickly becoming dominant platforms for drug delivery applications because of their small size, ability to solubilize water-insoluble drugs, and extended blood circulation. The application or potential application of ZnO NPs in therapy and vaccine to fight COVID-19 has been discussed (Faizan, 2021, Croy and Kwon, 2006). ZnO NPs are mentioned as antiviral, easy to prepare economically but toxic, whereas polymer micelles preparation is comparatively complex, costly. Polymeric micelles are more biocompatible, soluble, elicit high immune response as compared to ZnO NPs. Table 1 summarized the comparison between ZnO NPs and polymer/micelles used in therapy or used in the vaccine to fight COVID-19 impact (Supplementary file).Table 1 Comparison between ZnO NPs and polymer/micelles used in therapy or used in the vaccine to fight COVID-19 impact. ZnO NPs POLYMER Micelles Antiviral yes – Preparation Easy Complex Cost Economical Costly Solubility Less More Biocompatibility Less High Immune response Comparatively less immune response. strong cellular Immune response. Cytokines Minimum secretion of cytokines Increased secretion of cytokines Antigen Antibody Interaction. Decreased levels of antibodies and antigen specific antibodies. Increased levels of antibodies and antigen-specific antibodies (i.e., IgA, IgG, etc.) Adjuvant properties Adjuvant properties are not much advanced Having advanced adjuvant properties Dosage formulations More than single dose formulations single dose formulations Toxicity Toxic Comparatively safe The drawbacks of polymeric micelles include early and insufficient release into the diseased tissue (Miller et al., 2013). However, stabilization of the suitable micelle or strong drug interaction through hydrogen/covalent bonding can solve this problem. PEGylation is a pertinent process in which polyethylene glycol (PEG) is combined with another molecule with promising therapeutic properties. PEG with good bio-compatibility and hydrophilic formation of core in micelles can reduce the nonspecific adhesion to different components of the bloodstream and prolong the duration of its blood circulation. The chloroquine incorporated Zn particles and PEG/PLA may be linked by various interactions such as Hydrophobic-hydrophobic, Ion-dipole, H-bond, and Dipole-dipole as shown in Fig. 3 A. Among all interactions, solute-hydrophobic and hydrophobic-hydrophobic interactions are dominant over solute-hydrophilic and hydrophilic-hydrophobic interactions (Masood et al., 2020). Hydrophobic (e.g., PLA, etc.) and hydrophilic (PEG) polymers/entities create the hydrophobic core and the hydrophilic covering shell parts, respectively, of the copolymer micelles (Fig. 3B). The drug is sheltered from enzymatic degradation by the micelle's shell; it remains in the blood for a prolonged period. The micelles' surface properties, size, and stability primarily determine their biodistribution. They traverse the fenestrated blood vessels/capillaries that arise in most tissues (Hill et al., 2012). The inflamed tissues usually have permeable blood vessels/capillaries with big fenestra, so the permeation of micellar drug complexes into such tissues is quicker than into healthy tissues because of passive targeting and selective distribution to the diseased site. One can achieve specific targeting by attaching it to the human monoclonal or polyclonal antibodies directed against spike proteins of the SARS COV2, thus preventing the virus from latching onto the other cells.(ii) pH responsive micelles Fig. 3 A. Possible interactions between chloroquine incorporated Zn particles and PEG/PLA. Hydrophobic-hydrophobic, Ion-dipole, H-bond, and Dipole-dipole are shown by red, blue, dotted brown and green colors respectively. B. Schematic illustration of intracellular delivery of (hydroxy)chloroquine and zinc through copolymer micelles. The micelles contain hydrophilic and hydrophobic polymer entities. The drug may suppress the expression of phosphatidyl inositol binding clathrin assembly protein (PICALM) and reduce the pace of endocytosis in the hydrophobic core. Smart block copolymers, which are responsive to pH, temperature, ultrasound, or light, can allow controlled dissociation of the micelle and well-organized drug release. The pH-responsive micelles can be designed and synthesized for target delivery, exploiting the pH-dependent interaction of the virus and cell membrane. For various biological applications, pH-responsive systems ought to be reactive and steady to somewhat lower (5.0–6.5 pH), and 7.4 pH (physiological pH). The pH-sensitive formulation can be fabricated to deliver the zinc/(hydroxy)chloroquine with or without another antiviral agent allowing the release in an acidic milieu. This design can help the Zn/chloroquine liberation within the cell and adjacent tissues. It is due to the endo lysosomal sections formed upon the internalization of cargo. Micelles will be stable at neutral pH but allow the fast drug release in endocytic pH. Positively charged micelles can be safeguarded by the negatively charged entities at pH 7.4 and they can be broken or unprotected at the diseased sites’ pH due to pH-responsive entity. The pH-sensitive linkages responsive to low pH (endosomal pH) can be utilized in the design. Following endocytosis, the cationic entity will be protonated in the endosomal area resulting in the disintegration of the micelle and destabilizing the endosomal membrane and thus supporting the delivery of the zinc/(hydroxy)chloroquine to the cytosol. Although some pH-sensitive nanoparticles have good in vitro reactivity or activity, they may be hampered in vivo by a complicated physiological or pathological milieu. The low levels of the drug/molecule may inhibit the response activity of a given pH-sensitive component. The breakage of pH-sensitive chemical bonds may take longer, affecting the release of drug/molecule from the delivery system (Mu et al., 2021). A more sensitive and particular pH should be required with a subtle polymer design (Zhuo et al., 2020). A well-built library of materials with varied pH conversions to get suitable polymers can be created (Ma et al., 2014).(iii) Encapsulation The exploitation of natural polymers is valuable based on established biocompatibility. Chitosan can be used as a functional carrier for zinc and chloroquine, reducing both components' adverse effects. The delivery materials could also function as camouflage to deter immune responses or as promoters that could propel or respond to specific molecules or chemical processes. Chitosan is a customized biopolymer accomplished by de-acetylation of natural chitin, comprising N-acetyl glucosamine and glucosamine units (Rashki et al., 2020). It is a cationic polysaccharide polymer that can bind easily with the drug and is biocompatible and biodegradable. Additionally, it enhances the transport of drugs across the cell membrane. It can suppress multiple efflux pumps, thereby avoiding the problem of drug resistance (Ngo et al., 2015). The quick efflux lowers the intracellular drug concentration (Boroumand et al., 2021). The metabolism of chloroquine in the liver and small intestine, which produces quick clearance, is another explanation put forth for its decreased bioavailability. Higher dosages must be used because of the medicine's low bioavailability, which may also promote the development of drug resistance. The use of degradable polymers like chitosan as a carrier may protect the drug and avoid its renal clearance. The chitosan itself will be metabolized and degraded by enzymes in the body, eventually being removed by renal clearance. Several researchers employ chitosan as a delivery vehicle for antiviral drugs such as saquinavir, a protein inhibitor affecting viral proliferation of HIV (Ramana et al., 2014), Acyclovir against Herpes simplex virus-1 (Donalisio et al., 2018), and influenza vaccine (Dehghan et al., 2014) with greater efficacy as compared to their counterparts. Chitosan has a remarkable water absorption capacity and it swells as a hydrogel. At physiologic/basic pH, it becomes insoluble. However, it may dissolve in acidic conditions. This property can be exploited for the release of chloroquine/hydroxychloroquine to the affected areas with an acidic environment. It was noted that the hydrogel showed higher drug release at pH 5.7 than at pH 7.4. The chemical and graphical illustration of encapsulation of chloroquine-zinc complex and decapsulation at different pH of the body is shown in Fig. 4 A and B respectively. At physiologic/basic pH, it becomes insoluble. It dissolves in acidic conditions. Given the acidic milieu of endosomes (pH 4.5–6.0) and lysosomes (pH 4.5–4.8) compared to a body pH of 7.4, chitosan-based stimulant-sensitive hydrogels can be synthesized and developed for enhanced drug release in the endosomal or lysosomal compartments using internal pH stimuli (Chen et al., 2009, Du et al., 2005).Fig. 4 Chemical (A) and Graphic illustration (B) of encapsulation of chloroquine-zinc complex and decapsulation at different pH of the body. At physiologic/basic pH, it becomes insoluble. It dissolves in acidic conditions. The chitosan hydrogel showed higher drug release at pH 5.7. As referred to earlier reports, chitosan carriers are becoming increasingly popular because of their numerous benefits. However, this polysaccharide still faces several limitations, such as the low solubility at blood pH, premature release, and its structural instability after cellular uptake. Therefore, it is obligatory to find some improvement in chitosan-related nanocarriers. In a study, chitosan integrated albumin was observed as a perfect carrier for muco-inhalable delivery of silymarin/curcumin against SARS COV2 (Hanafy and El-Kemary, 2022). Chitosan can alter the physiochemical characteristics of nanoparticles, increasing their dispersibility and bioavailability in the lungs. Researchers used the chemical alteration of chitosan using PEG (PEGylation) to improve the melting of chitosan, even though extreme PEGylation may reduce its solubility, charge, and binding capacity to DNA/RNA (Boroumand et al., 2021). Despite all of the progress in understanding ZnO nanostructures' mechanisms of action and biological impacts, there is still a paucity of knowledge regarding the long-term implications. As a result, further information is needed to determine if the numerous proven benefits of ZnO outweigh the potential hazards. In summary, the toxicity issues or adverse effects relating to the decades-old known drug (hydroxy)chloroquine are associated with its improper use concerning dose and its interaction with other medicines provided to COVID-19 patients. Chloroquine is a well-known zinc ionophore besides possessing an antiviral effect. Smart and targeted delivery, including nanotechnology approaches to deliver ZnO/hydroxychloroquine, can be exploited. It will enable us to deliver inaccessible drugs with diminished toxicity, improved solubility, controlled/ sustained release and site-specific delivery through pH-sensitive copolymer micelles/drug encapsulation. One of the most difficult obstacles has been the low pH range, which requires the micellar form to hold the medicine for a longer duration before releasing it. Because of the moderate acidic environment and dynamic distribution of nanoparticles in vivo, a quick response time is required for precise drug release. We discussed the delivery options like copolymer micelle/pH-sensitive/encapsulation of ZnO NPs along with (hydroxy) chloroquine (as an ionophore) which would be the better option for therapeutic management of COVID-19. The use of FDA-approved materials for the delivery or various therapies has favorable risk: benefit profiles and may shorten the development and approval procedures. 6 Authors’ contributions Anju Manuja and Balvinder Kumar conceived the idea; Anju Manuja wrote the article, Dharvi Chhabra collected the matter and assisted in figures. Balvinder Kumar edited 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. Acknowledgment Authors thank Director, ICAR-National Research Centre on Equines, Hisar, Haryana, India for providing administrative support and Ms Swati Rani for drawing the chemical structures. Peer review under responsibility of King Saud University. ==== Refs References Abdel-Hamid H. Oddis C.V. Lacomis D. Severe hydroxychloroquine myopathy Muscle Nerve: Off. J. Am. Assoc. Electrodiag. Med. 38 3 2008 Sep 1206 1210 Akpovwa H. Chloroquine could be used for the treatment of filoviral infections and other viral infections that emerge or emerged from viruses requiring an acidic pH for infectivity Cell Biochem. Funct. 34 4 2016 Jun 191 196 27001679 Al-Bari M.A. Targeting endosomal acidification by chloroquine analogs as a promising strategy for the treatment of emerging viral diseases Pharmacol. Res. Perspect. 5 1 2017 Feb e00293 28596841 Anumonwo J.M.B. Horta J. Delmar M. Taffet S.M. Jalife J. Proton and zinc effects on HERG currents Biophys. J. 77 1999 282 298 10.1016/S0006-3495(99)76889-X 10388757 Bansal et al., 2021. Hydroxychloroquine: a comprehensive review and its controversial role in coronavirus disease, Annals of Medicine, 53:1, 117-134 Becerra-Cunat J.L. Coll-Cantí J. Gelpí-Mantius E. Ferrer-Avellí X. Lozano-Sánchez M. Millán-Torné M. Ojanguren I. Ariza A. Olivé A. Chloroquine-induced myopathy and neuropathy: progressive tetraparesis with areflexia that simulates a polyradiculoneuropathy Two Case Rep. Rev. Neurol. 36 6 2003 Mar 1 523 526 Boonyasuppayakorn S. Reichert E.D. Manzano M. Nagarajan K. Padmanabhan R. Amodiaquine, an antimalarial drug, inhibits dengue virus type 2 replication and infectivity Antivir. Res. 106 2014 125 134 24680954 Boroumand H. Badie F. Mazaheri S. Seyedi Z.S. Nahand J.S. Nejati M. Baghi H.B. Abbasi-Kolli M. Badehnoosh B. Ghandali M. Hamblin M.R. Chitosan-based nanoparticles against viral infections Front. Cell. Infect. Microbiol. 17 11 2021 Mar 175 Chatre C. Roubille F. Vernhet H. Jorgensen C. Pers Y.M. Cardiac complications attributed to chloroquine and hydroxychloroquine: a systematic review of the literature Drug Saf. 41 10 2018 Oct 919 931 29858838 Chen W. Meng F. Li F. Ji S.J. Zhong Z. pH-responsive biodegradable micelles based on acid-labile polycarbonate hydrophobe: synthesis and triggered drug release Biomacromolecules 10 7 2009 Jul 13 1727 1735 19469499 Chen C.Y. Wang F.L. Lin C.C. Chronic hydroxychloroquine use associated with QT prolongation and refractory ventricular arrhythmia Clin. Toxicol. 44 2 2006 Jan 1 173 175 Chopra Meenu, Bernela Manju, Kaur Pawan, Manuja Anju, Kumar Balvinder, Thakur Rajesh Alginate/gum acacia bipolymeric nanohydrogels—Promising carrier for Zinc oxide nanoparticles.International Journal of biological macromolecules.72c:827-833. Cojocaru F.D. Botezat D. Gardikiotis I. Uritu C.M. Dodi G. Trandafir L. Rezus C. Rezus E. Tamba B.I. Mihai C.T. Nanomaterials designed for antiviral drug delivery transport across biological barriers Pharmaceutics 12 2 2020 Feb 171 32085535 Croy S.R. Kwon G.S. Polymeric micelles for drug delivery Curr. Pharm. Des. 12 36 2006 4669 4684 17168771 Darma A. Ranuh I.G. Merbawani W. Setyoningrum R.A. Hidajat B. Hidayati S.N. Endaryanto A. Sudarmo S.M. Zinc supplementation effect on the bronchial cilia length, the number of cilia, and the number of intact bronchial cell in zinc deficiency rats Indonesian Biomed. J. 12 1 2020 78 84 Dehghan S. Tafaghodi M. Bolourieh T. Mazaheri V. Torabi A. Abnous K. Kheiri M.T. Rabbit nasal immunization against influenza by dry-powder form of chitosan nanospheres encapsulated with influenza whole virus and adjuvants Int. J. Pharm. 475 1–2 2014 Nov 20 1 8 25148732 Delvecchio R. Higa L.M. Pezzuto P. Valadão A.L. Garcez P.P. Monteiro F.L. Loiola E.C. Dias A.A. Silva F.J. Aliota M.T. Caine E.A. Chloroquine, an endocytosis blocking agent, inhibits Zika virus infection in different cell models Viruses 8 12 2016 322 27916837 Donalisio M. Leone F. Civra A. Spagnolo R. Ozer O. Lembo D. Cavalli R. Acyclovir-loaded chitosan nanospheres from nano-emulsion templating for the topical treatment of herpesviruses infections Pharmaceutics 10 2 2018 Jun 46 29642603 Du J. Tang Y. Lewis A.L. Armes S.P. pH-sensitive vesicles based on a biocompatible zwitterionic di block copolymer J. Am. Chem. Soc. 127 51 2005 Dec 28 17982 17983 16366531 Faizan M. Zinc oxide nanoparticles to fight the COVID-19 Acta Scient. Agric. 5 7 2021 14 16 Farias K.J.S. Machado P.R.L. da Fonseca B.A.L. Chloroquine inhibits dengue virus type 2 replication in Vero cells but not in C6/36 cells Sci. World J. 2013 282734 Gao J. Tian Z. Yang X. Breakthrough: chloroquine phosphate has shown apparent efficacy in treatment of COVID19 associated pneumonia in clinical studies Biosci. Trends 2020 Gautret P. Lagier J.C. Parola P. Meddeb L. Mailhe M. Doudier B. Courjon J. Giordanengo V. Vieira V.E. Dupont H.T. Honoré S. Hydroxychloroquine and azithromycin as a treatment of COVID19: results of an open-label non-randomized clinical trial Int. J. Antimicrob. Agents 56 1 2020 Jul 1 105949 Ghaffari H. Tavakoli A. Moradi A. Tabarraei A. Bokharaei-Salim F. Zahmatkeshan M. Farahmand M. Javanmard D. Kiani S.J. Esghaei M. Pirhajati-Mahabadi V. Inhibition of H1N1 influenza virus infection by zinc oxide nanoparticles: another emerging application of nanomedicine J. Biomed. Sci. 26 1 2019 1 30602371 Giudicessi JR, Noseworthy PA, Friedman PA, Ackerman MJ. Urgent guidance for navigating and circumventing the QTc-prolonging and torsadogenic potential of possible pharmacotherapies for coronavirus disease 19 (COVID19). In Mayo Clinic Proceedings 2020 Jun 1 (Vol. 95, No. 6, pp. 1213-1221). Elsevier. Hamdi M. Abdel-Bar H.M. Elmowafy E. El-Khouly A. Mansour M. Awad G.A.S. Investigating the internalization and COVID19 antiviral computational analysis of optimized nanoscale zinc oxide ACS Omega 6 10 2021 6848 6860 33748599 Hanafy N.A. El-Kemary M.A. Silymarin/curcumin loaded albumin nanoparticles coated by chitosan as muco-inhalable delivery system observing anti-inflammatory and anti COVID-19 characterizations in oleic acid triggered lung injury and in vitro COVID-19 experiment Int. J. Biol. Macromol. 198 2022 101 110 34968533 Hill RG, Richards D, editors. Drug Discovery and Development E-Book: Technology in Transition. Elsevier Health Sciences; 2012 Hoffmann M. Kleine-Weber H. Schroeder S. Krüger N. Herrler T. Erichsen S. Schiergens T.S. Herrler G. Wu N.H. Nitsche A. Müller M.A. SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor Cell 2020 Mar 5 Hong Z, Jiang Z, Liangxi W, Guofu D, Ping L, Yongling L, Wendong P, Minghai W. Chloroquine protects mice from challenge with CpG ODN and LPS by decreasing proinflammatory cytokine release. International immunopharmacology. 2004 Feb1;4(2):223 http.//Clinical trials.gov.in as assessed on date 11/11/2022. Hu T.Y. Frieman M. Wolfram J. Insights from nanomedicine into chloroquine efficacy against COVID19 Nat. Nanotechnol. 2020 1 3 Huang I. Lim M.A. Pranata R. Diabetes mellitus is associated with increased mortality and severity of disease in COVID19 pneumonia–a systematic review, meta-analysis, and meta-regression Diabetes Metab. Syndr. 14 4 2020 Jul 1 395 403 32334395 Hussein R.K. Elkhair H.M. Molecular docking identification for the efficacy of some zinc complexes with chloroquine and hydroxychloroquine against main protease of COVID19 J. Mol. Struct. 1231 2021 129979 Jang C.H. Choi J.H. Byun M.S. Jue D.M. Chloroquine inhibits production of TNF-α, IL-1β and IL-6 from lipopolysaccharide-stimulated human monocytes/macrophages by different modes Rheumatology 45 6 2006 Jun 1 703 710 16418198 Jie Z. He H. Xi H. Zhi Z. Expert consensus on chloroquine phosphate for the treatment of novel coronavirus pneumonia Zhonghua Jie He He Hu Xi Za Zhi 43 3 2020 185 188 32164085 Kaufmann A.M. Krise J.P. Lysosomal sequestration of amine-containing drugs: analysis and therapeutic implications J. Pharm. Sci. 96 4 2007 Apr 1 729 746 17117426 Kim KY. Nanotechnology platforms and physiological challenges for cancer therapeutics. Nanomedicine: Nanotechnology, Biology and Medicine. 2007 Jun 1;3(2):103-10. Klibanov A.L. Maruyama K. Torchilin V.P. Huang L. Amphipathic polyethyleneglycols effectively prolong the circulation time of liposomes FEBS Lett. 268 1 1990 235 237 2384160 Kwiek J.J. Haystead T.A. Rudolph J. Kinetic mechanism of quinone oxidoreductase 2 and its inhibition by the antimalarial quinolines Biochemistry 43 2004 4538 4547 15078100 Kwon J.B. Kleiner A. Ishida K. Godown J. Ciafaloni E. Looney R.J. Jr Hydroxychloroquine-induced myopathy JCR J. Clin. Rheumatol. 16 1 2010 Jan 1 28 31 20051753 Lancet, T., Expression of concern: Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis. Lancet (London, England). Lembo D. Cavalli R. Nanoparticulate delivery systems for antiviral drugs Antivir. Chem. Chemother. 21 2 2010 Dec 53 70 21107015 Liu J, Cao R, Xu M. Hydroxychloroquine, a less toxic derivative of chloroquine, is effective in inhibiting SARS-CoV-2 infection in vitro [published online March 18, 2020]. Cell Discovery. Ma X. Wang Y. Zhao T. Li Y. Su L.C. Wang Z. Huang G. Sumer B.D. Gao J. Ultra-pH-sensitive nanoprobe library with broad pH tunability and fluorescence emissions J. Am. Chem. Soc. 136 31 2014 Aug 6 11085 11092 25020134 Maares M. Haase H. Zinc and immunity: an essential interrelation Arch. Biochem. Biophys. 1 611 2016 Dec 58 65 Manuja A. Kumar B. Riyesh T. Talluri T.R. Tripathi B.N. Microwave assisted fast fabrication of zinc/iron oxides based polymeric nanocomposites and evaluation on equine fibroblasts Int. J. Biol. Macromol. 165 2020 71 81 32987081 Manuja A. Raguvaran R. Kumar B. Kalia A. Tripathi B.N. Accelerated healing of full thickness excised skin wound in rabbits using single application of alginate/acacia based nanocomposites of ZnO nanoparticles Int. J. Biol. Macromol. 2020 Mar 28 Manuja A. Kumar B. Kumar R. Chhabra D. Ghosh M. Manuja M. Brar B. Pal Y. Tripathi B.N. Prasad M. Metal/metal oxide nanoparticles: toxicity concerns associated with their physical state and remediation for biomedical applications Toxicol. Rep. 8 2021 1970 1978 34934635 Masood S. Rehman W. Begum S. Khan Z. Gulnar L. Drug-drug and drug-solvent interaction studies of Chloroquine phosphate, Acefylline piperazine and Gentamicin sulfate in polymeric systems Arab. J. Chem. 13 7 2020 6221 6235 Mehta O.P. Bhandari P. Raut A. Kacimi S.E. Huy N.T. Coronavirus disease (COVID-19): comprehensive review of clinical presentation Front. Public Health 15 8 2021 Jan 1034 Michaelides M. Stover N.B. Francis P.J. Weleber R.G. Retinal toxicity associated with hydroxychloroquine and chloroquine: risk factors, screening, and progression despite cessation of therapy Arch. Ophthalmol. 129 1 2011 Jan 10 30 39 21220626 Miller T. Breyer S. Van Colen G. Mier W. Haberkorn U. Geissler S. Voss S. Weigandt M. Goepferich A. Premature drug release of polymeric micelles and its effects on tumor targeting Int. J. Pharm. 445 1–2 2013 Mar 10 117 124 23384729 Mu Y. Gong L. Peng T. Yao J. Lin Z. Advances in pH-responsive drug delivery systems OpenNano 10 2021 Dec 100031 Munster T. Gibbs J.P. Shen D. Baethge B.A. Botstein G.R. Caldwell J. Dietz F. Ettlinger R. Golden H.E. Lindsley H. McLaughlin G.E. Hydroxychloroquine concentration–response relationships in patients with rheumatoid arthritis Arthrit. Rheumat.: Off. J. Am. College Rheumatol. 46 6 2002 Jun 1460 1469 Mzayek F. Deng H. Mather F.J. Wasilevich E.C. Liu H. Hadi C.M. Chansolme D.H. Murphy H.A. Melek B.H. Tenaglia A.N. Mushatt D.M. Randomized dose-ranging controlled trial of AQ-13, a candidate antimalarial, and chloroquine in healthy volunteers PLoS Clin. Trials 2 1 2007 Jan 5 e6 17213921 Naksuk N. Lazar S. Peeraphatdit T. Cardiac safety of off-label COVID19 drug therapy: a review and proposed monitoring protocol Eur. Heart J. Acute Cardiovasc. Care 9 3 2020 Apr 215 221 32372695 Ngo D.H. Vo T.S. Ngo D.N. Kang K.H. Je J.Y. Pham H.N. Byun H.G. Kim S.K. Biological effects of chitosan and its derivatives Food Hydrocoll. 51 2015 Oct 1 200 216 Ooi E.E. Chew J.S.W. Loh J.P. Chua R.C.S. In vitro inhibition of human influenza A virus replication by chloroquine Virol. J. 3 2006 39 16729896 Ou T. Mou H. Zhang L. Ojha A. Choe H. Farzan M. Hydroxychloroquine-mediated inhibition of SARS-CoV-2 entry is attenuated by TMPRSS2 PLoS Pathog. 17 1 2021 e1009212 33465165 Paulikat, M., Vitone, D., Schackert, F.K., Schuth, N., Barbanente, A., Piccini, G., Ippoliti, E., Rossetti, G., Clark, A.H., Nachtegaal, M. and Haumann, M., 2022. Molecular dynamics and structural studies of zinc chloroquine complexes. Phan AD, Hoang TX. The pH-dependent electrostatic interaction of a metal nanoparticle with the MS2 virus-like particles. Chemical Physics Letters. 2019 Sep1;730:84-8 Picot S, Peyron F, Vuillez JP, Polack B, Ambroise-Thomas P. Chloroquine inhibits tumor necrosis factor production by human macrophages in vitro. J Infect Dis. 1991 Oct1;164(4):830 Piscopo S, Brown ER. Zinc Oxide Nanoparticles and Voltage‐Gated Human Kv11. 1 Potassium Channels Interact through a Novel Mechanism. Small. 2018 Apr;14(15):1703403 Pukrittayakamee S. Tarning J. Jittamala P. Charunwatthana P. Lawpoolsri S. Lee S.J. Hanpithakpong W. Hanboonkunupakarn B. Day N.P. Ashley E.A. White N.J. Pharmacokinetic interactions between primaquine and chloroquine Antimicrob. Agents Chemother. 8 6 2014 Jun 5 3354 3359 Raguvaran R. Manuja A. Singh S. Chopra M. Manuja B.K. Dimri U. Zinc oxide nanoparticles induced haemolytic cytotoxicity in horse red blood cells Int. J. Pharm. Sci. Res. 6 3 2015 Mar 1 1166 Raguvaran R. Manuja A. Manuja B.K. Riyesh T. Singh S. Kesavan M. Dimri U. Sodium alginate and gum acacia hydrogels of zinc oxide nanoparticles reduce hemolytic and oxidative stress inflicted by zinc oxide nanoparticles on mammalian cells Int. J. Biol. Macromol. 1 101 2017 Aug 967 972 Raguvaran R. Manuja B.K. Chopra M. Thakur R. Anand T. Kalia A. Manuja A. Sodium alginate and gum acacia hydrogels of ZnO nanoparticles show wound healing effect on fibroblast cells Int. J. Biol. Macromol. 1 96 2017 Mar 185 191 Ramana LN, Sharma S, Sethuraman S, Ranga U, Krishnan UM. Evaluation of chitosan nanoformulations as potent anti-HIV therapeutic systems. Biochimica et Biophysica Acta (BBA)-General Subjects. 2014 Jan 1;1840(1):476-84. Rashki S, Asgarpour K, Tarrahimofrad H, Hashemipour M, Ebrahimi MS, Fathizadeh H, Khorshidi A, Khan H, Salavati-Niasari M, Mirzaei H. Chitosan-based nanoparticles against bacterial infections. Carbohydrate Polymers. 2020 Sep 20;117(108) Rathore JS, Ghosh C. Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), a newly emerged pathogen: an overview. Pathogens and disease. 2020 Aug;78(6):ftaa042 Razzaque M.S. COVID-19 pandemic: Can zinc supplementation provide an additional shield against the infection? Comput. Struct. Biotechnol. J. 1 19 2021 Jan 1371 1378 Roldan E.Q. Biasiotto G. Magro P. Zanella I. The possible mechanisms of action of 4-aminoquinolines (chloroquine/hydroxychloroquine) against Sars-Cov-2 infection (COVID-19): a role for iron homeostasis? Pharmacol. Res. 1 158 2020 Aug 104904 Roscioli E. Jersmann H.P. Lester S. Badiei A. Fon A. Zalewski P. Hodge S. Zinc deficiency as a codeterminant for airway epithelial barrier dysfunction in an ex vivo model of COPD Int. J. Chron. Obstruct. Pulmon. Dis. 12 2017 3503 29255357 Siddiqui A.K. Huberfeld S.I. Weidenheim K.M. Einberg K.R. Efferen L.S. Hydroxychloroquine-induced toxic myopathy causing respiratory failure Chest 131 2 2007 Feb 1 588 590 17296665 Sperber K, Quraishi HU, Kalb TH, Panja AS, Stecher V, Mayer L. Selective regulation of cytokine secretion by hydroxychloroquine inhibition of interleukin 1 alpha (IL-1-alpha) and IL-6 in human monocytes and T cells. The Journal of rheumatology.1993 May 1;20(5):803-8. Sportelli M.C. Izzi M. Loconsole D. Sallustio A. Picca R.A. Felici R. Chironna M. Cioffi N. On the efficacy of ZnO nanostructures against SARS-CoV-2 Int. J. Mol. Sci. 23 6 2022 3040 35328455 Stas P, Faes D, Noyens P. Conduction disorder and QT prolongation secondary to long-term treatment with chloroquine. International journal of cardiology. 2008 Jul 4;127(2):e80-2.Stas 2008. Tarek M. Savarino A. Pharmacokinetic basis of the hydroxychloroquine response in COVID-19: implications for therapy and prevention Eur. J. Drug Metab. Pharmacokinet. 45 6 2020 715 723 32780273 Tavakoli A. Ataei-Pirkooh A. Mm Sadeghi G. Bokharaei-Salim F. Sahrapour P. Kiani S.J. Polyethylene glycol-coated zinc oxide nanoparticle: an efficient nanoweapon to fight against herpes simplex virus type 1 Nanomedicine 13 21 2018 2675 2690 30346253 Tsai W.P. Nara P.L. Kung H.F. Oroszlan S. Inhibition of human immunodeficiency virus infectivity by chloroquine AIDS Res. Hum. Retrovir. 6 1990 481 489 1692728 Van den Borne BE, Dijkmans BA, De Rooij HH, Le Cessie S, Verweij CL. Chloroquine and hydroxychloroquine equally affect tumor necrosis factor-alpha, interleukin 6, and interferon-gamma production by peripheral blood mononuclear cells. rheumatology.1997 Jan 1;24(1):55-60. Vincent M.J. Bergeron E. Benjannet S. Erickson B.R. Rollin P.E. Ksiazek T.G. Seidah N.G. Nichol S.T. Chloroquine is a potent inhibitor of SARS coronavirus infection and spread Virol. J. 2 1 2005 Dec 1 15631631 Wang M. Cao R. Zhang L. Yang X. Liu J. Xu M. Shi Z. Hu Z. Zhong W. Xiao G. Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro Cell Res. 30 3 2020 Mar 269 271 32020029 Wang N. Shi X. Jiang L. Zhang S. Wang D. Tong P. Guo D. Fu L. Cui Y. Liu X. Arledge K.C. Structure of MERS-CoV spike receptor-binding domain complexed with human receptor DPP4 Cell Res. 23 8 2013 Aug 986 993 23835475 Weiss S.R. Navas-Martin S. Coronavirus pathogenesis and the emerging pathogen severe acute respiratory syndrome coronavirus Microbiol. Mol. Biol. Rev. 69 4 2005 Dec 635 664 16339739 Wen W. Chen C. Tang J. Wang C. Zhou M. Cheng Y. Zhou X. Wu Q. Zhang X. Feng Z. Wang M. Efficacy and safety of three new oral antiviral treatment (molnupiravir, fluvoxamine and Paxlovid) for COVID-19: a meta-analysis Ann. Med. 54 1 2022 Dec 31 516 523 35118917 Wessels I. Rolles B. Rink L. The potential impact of zinc supplementation on COVID-19 pathogenesis Front. Immunol. 2020 1712 32754164 WHO Solidarity Trial Consortium. Repurposed antiviral drugs for Covid-19—interim WHO solidarity trial results. New England journal of medicine. 2021 Feb 11;384(6):497-511 Wolfe F. Marmor M.F. Rates and predictors of hydroxychloroquine retinal toxicity in patients with rheumatoid arthritis and systemic lupus erythematosus Arthritis Care Res. 62 6 2010 Jun 775 784 World Health Organization. The world health report 2002: reducing risks, promoting healthy life. World Health Organization; 2002 Wu Y.C. Chen C.S. Chan Y.J. The outbreak of COVID-19: an overview J. Chin. Med. Assoc. 83 3 2020 Mar 217 32134861 Wu C.I. Postema P.G. Arbelo E. Behr E.R. Bezzina C.R. Napolitano C. Robyns T. Probst V. Schulze-Bahr E. Remme C.A. Wilde A.A. SARS-CoV-2, COVID19, and inherited arrhythmia syndromes Heart Rhythm. 17 9 2020 Sep 1 1456 1462 32244059 Xue J. Moyer A. Peng B. Wu J. Hannafon B.N. Ding W.Q. Chloroquine is a zinc ionophore PLoS One 9 10 2014 Yam J.C. Kwok A.K. Ocular toxicity of hydroxychloroquine Hong Kong Med. J. 12 4 2006 Aug 12 294 16912357 Yao X. Ye F. Zhang M. Cui C. Huang B. Niu P. Liu X. Zhao L. Dong E. Song C. Zhan S. In vitro antiviral activity and projection of optimized dosing design of hydroxychloroquine for the treatment of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Clin. Infect. Dis. 71 15 2020 Jul 28 732 739 32150618 Yasuda H, Leelahavanichkul A, Tsunoda S, Dear JW, Takahashi Y, Ito S, Hu X, Zhou H, Doi K, Childs R, Klinman DM. Chloroquine and inhibition of Toll-like receptor 9 protect from sepsis-induced acute kidney injury. American Journal of Physiology-Renal Physiology. 2008 May 29; 4(5):1050-8 Yoon Y.H. Cho K.S. Hwang J.J. Lee S.J. Choi J.A. Koh J.Y. Induction of lysosomal dilatation, arrested autophagy, and cell death by chloroquine in cultured ARPE-19 cells Invest. Ophthalmol. Vis. Sci. 51 11 2010 Nov 1 6030 6037 20574031 Zequn Z. Yujia W. Dingding Q. Jiangfang L. Off-label use of chloroquine, hydroxychloroquine, azithromycin and lopinavir/ritonavir in COVID19 risks prolonging the QT interval by targeting the hERG channel Eur. J. Pharmacol. 15 893 2021 Feb 173813 Zhang X.L. Li Z.M. Ye J.T. Lu J. Ye L.L. Zhang C.X. Liu P.Q. Duan D.D. Pharmacological and cardiovascular perspectives on the treatment of COVID19 with chloroquine derivatives Acta Pharmacol. Sin. 23 2020 Sep 1 Zhu X. Pan Y. Li Y. Jiang Y. Shang H. Gowda D.C. Cui L. Cao Y. Targeting Toll-like receptors by chloroquine protects mice from experimental cerebral malaria Int. Immunopharmacol. 13 4 2012 Aug 1 392 397 22659438 Zhu Y. Xu Q. Wu D. Ren H. Zhao P. Lao W. Wang Y. Tao Q. Qian X. Wei Y.-H. Japanese encephalitis virus enters rat neuroblastoma cells via a pH-dependent, dynamin and caveola-mediated endocytosis pathway J. Virol. 86 2012 13407 13422 23015720 Zhuo S. Zhang F. Yu J. Zhang X. Yang G. Liu X. pH-sensitive biomaterials for drug delivery Molecules 25 23 2020 Jan 5649 33266162	36466721	PMC9710101	NO-CC CODE	2022-12-14 23:45:35	no	Arab J Chem. 2023 Feb 30; 16(2):104468	utf-8	Arab J Chem	2,022	10.1016/j.arabjc.2022.104468	oa_other
==== Front Psychol Sport Exerc Psychol Sport Exerc Psychology of Sport and Exercise 1469-0292 1878-5476 Elsevier Ltd. S1469-0292(22)00216-3 10.1016/j.psychsport.2022.102348 102348 Article The association of families’ socioeconomic and demographic characteristics with parents’ perceived barriers to returning to youth sport following the COVID-19 pandemic☆ Fleming Daniel J.M. a∗ Dorsch Travis E. a Serang Sarfaraz b Hardiman Amand L. a Blazo Jordan A. c Farrey Tom d Lerner Jennifer Brown d Solomon Jon d a Utah State University, United States b University of South Carolina, United States c Louisana Tech University, United States d Aspen Institute Project Play Initiative, United States ∗ Corresponding author. Department of Human Development and Family Studies, Utah State University, Logan, Utah, 84322, United States. 30 11 2022 3 2023 30 11 2022 65 102348102348 21 6 2022 28 11 2022 30 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. Developmentally appropriate sport contexts have the potential to positively influence young people’s physiological, psychological, and social outcomes. However, little is known about how families returned to sport in the wake of COVID-19-related restrictions or how socioeconomic and demographic factors influenced parents’ perceptions of barriers to returning. A nationally representative sample (N = 6183) of American youth sport parents completed a questionnaire in which they provided demographic information and answered questions related to the barriers they perceived in returning to sport, such as the risk of their child getting sick. Structural equation modeling was used to examine the relationships among a range of socioeconomic and demographic factors and these barriers to returning. Results suggest that parents from racially minoritized and urban neighborhoods held higher levels of concern over health-related and practical barriers to returning to sport. Findings highlight the importance of designing available, equitable, and appropriate youth sport contexts. Keywords Youth sport Sport parenting Socioeconomic status COVID-19 Barriers ==== Body pmcAn estimated 36 million youth take part in organized sport every year in the United States (Sport & Fitness Industry Association, 2020). Childhood and adolescence represent periods of life where sport participation rates are typically at their highest (McKay et al., 2019) and well-designed youth sport contexts have the potential to foster positive physical and psychosocial health and well-being (Côté & Vierimaa, 2014; Eime et al., 2013; Fraser-Thomas et al., 2005). Specifically, optimizing participation contexts can improve short- and long-term physical outcomes such as cardiovascular fitness, weight control, muscular strength, and endurance (Fraser-Thomas et al., 2005). Importantly, increased physical activity has also been associated with a reduced likelihood of taking up smoking and developing diseases such as heart disease, stroke, diabetes, osteoporosis, and cancer. Concerning psychological outcomes, those who participate in regular physical activity have a greater likelihood of experiencing more enjoyment, increased self-esteem, greater levels of happiness and well-being, reduced stress, and lower depression (Fraser-Thomas et al., 2005). Socially, sport experiences provide opportunities to develop citizenship, positive peer relationships, leadership skills, cooperation, empathy, responsibility, and self-control (Eime et al., 2013; Fraser-Thomas, et al., 2005). When evaluating the potential of youth sport to positively influence the development of children and adolescents, it is important to understand the factors related to its design and delivery. It is these factors that foster more or less accessible, developmentally appropriate, and positive experiences for participants. Considering these factors has become particularly important in the wake of the COVID-19 pandemic. COVID-19 is an infectious disease caused by the pathogen known as “severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)” (CDC, 2020). The recent global pandemic led to drastic changes to everyday life around the world. Rapid transmission of the virus led to an unprecedented closure of schools and businesses, along with other domains of modern life. Leisure time pursuits, such as travel, arts, and sport, were some of the more heavily impacted activities. Deemed as “non-essential”, many sporting events (training, competition, and even free play) were either postponed or cancelled during the initial stages of the pandemic (Grix et al., 2020). Indeed, McGuine et al. (2021) found that in the fall of 2020, 69% of high school athletes did not have access to interscholastic sport. Further, findings from a national survey of youth sport parents in the United States noted a 60.3% reduction in the number hours per week youth spent training and a 66.8% reduction in hours of competition during the pandemic (Dorsch & Blazo, 2021). The decrease in hours of participation has the potential to impact young people’s athletic identity as researchers assert those with strong athletic identities may experience maladaptive emotions with abrupt decrease in sport participation (Edison et al., 2021). Athletic identity is the self-perception of an individual who identifies with the athlete role (Lamont-Mills & Christensen, 2006). Brewer et al. (1993) posited that athletic identity is influenced by interactions with coaches, family members, peers, and other key stakeholders within the sport system. Youth sport participants with strong athletic identities are more likely to have better health and physical development, higher self-esteem and perceived confidence, better social relationships, and increased sport participation rates (Brewer et al., 1993; Tasiemski et al., 2004). Researchers note that individuals with strong athletic identities who undergo abrupt career-ending events (e.g., injury) experience significant loss of identity (Sanders & Stevinson, 2017) and that this loss of identity can be associated with increased levels of anxiety and depression (Sanders & Stevinson, 2017). The abrupt pause in sport due to COVID-19 forced many older adolescent athletes to “retire” unexpectedly, potentially having similar effects on well-being. Similarly, COVID-19 decreased the social connectedness of teammates and competitors. Consequently, individuals’ athletic identities were infringed upon due to the lack of training and competition, and therefore less frequent interactions between athletes, peers, and coaches (Malina, 2009). The cancellation of youth sport is thought to have impacted family functioning as well as the normative development of children and adolescence. Sanderson and Brown (2020) highlighted ways families were forced to adjust their daily routines, as sport once occupied a salient space in children’s and families’ public lives. At an intrapersonal level, the authors suggested that the cancellation of youth sport was a strong source of stress for many athletes, with some even potentially experiencing grief given the sudden loss of such an instrumental part of their identity. Graupensperger et al. (2020) found that prioritizing social connections to peers fostered the maintenance of identity during the pandemic. However, this proved to be difficult as physical distancing protocols reduced much opportunity for regular social contacts among peers and teammates. Although there is evidence that a lack of sport led to deleterious consequences for youth, emerging research also suggests that continued engagement in sport during the pandemic had the potential to lead to positive outcomes (McGuine et al., 2021). Specifically, those who continued to play sport experienced lower levels of anxiety and depression, as well as higher perceptions of quality of life (McGuine et al., 2021). The COVID-19 pandemic also led to a large amount of economic stress for families (Sanderson & Brown, 2020). A large number of jobs were lost during the initial months of the pandemic, causing many parents to seek additional and often non-traditional sources of income. This resulted in fewer opportunities for parents to support children’s structured and unstructured play, as well as independent or virtual training. In April 2020, the unemployment rate in the United States peaked at 14.8%, thought to be the highest since the Great Depression and the highest on record since data collection began in 1948 (Congressional Research Service, 2021). This hidden consequence of the pandemic struck those individuals and families of lower socioeconomic status particularly hard. It is clear the pandemic amplified the systemic disadvantages that individuals and families faced prior to the pandemic, with minoritized families experiencing higher levels of peak unemployment in 2020 and a slower recovery since (Congressional Research Service, 2021). In addition to income as a lead indicator, there were also differences in pandemic-related unemployment as a function of education. Namely, adults in the United States with less than a high school diploma peaked at 21% unemployment, whereas those with a bachelor’s degree or higher peaked at 8.4%. Socioeconomic status (SES) is defined as the “social standing or class of an individual or group. It is often measured as a combination of education, income, and occupation.” (APA, 2022, Socioeconomic Status). In the United States and other industrialized countries, SES is strongly associated with life and health-related outcomes (Braveman et al., 2011). Notably, individuals of lower SES are more likely to be frontline/essential workers and were therefore more likely to experience higher levels of exposure to the virus (Blau, Meyerhofer, & Koebe, 2020). In contrast, individuals of higher SES were more likely to work from home during the pandemic, with modern comforts such as stable internet, plentiful access to food, and comfortable living arrangements (Patel et al., 2020; Wanberg et al., 2020). In sport, children from less educated and less affluent backgrounds have traditionally been less likely to have access to participation opportunities and the potential positive outcomes associated with them (Baxter-Jones & Maffulli, 2003). These access and outcome gaps have been exacerbated in recent years, as the financial support required from parents to allow their children to participate in sport has risen exponentially, particularly within the travel, club, and “elite” participation domain. Current literature suggests that parents spend between 3 and 12% of their gross annual income for one child to participate in sport (Dunn et al., 2016). Recent literature also suggests that children from higher-earning households engage in more weekly hours of sport participation when compared to those from lower-income households (Aspen Institute, 2019). This rise in families’ sport investment highlights the important role parents play as gatekeepers to their children’s sport opportunities, and how affluence is directly related to the opportunity to engage in, and benefit from, youth sport experiences (Fredricks & Eccles, 2004). In this light, Merkel (2013) suggests that living in lower SES neighborhoods contributes to inactivity among youth due to the limited access to adequate sport programming as well as the facilities and infrastructure to support it. Evidence also suggests that family structure may influence a youth’s likelihood of participating in sport (McMillan et al., 2016), specifically that children from ‘traditional’ families (i.e., dual parent homes) were more likely to participate in youth sport and that this relationship was mediated by perceived familial wealth. In the realm of sport, there is little empirical research that highlights whether one’s race is associated with SES. While there is an over-representation of Black athletes in intercollegiate and professional sport leagues when compared to the broader United States population (Harper et al., 2013), there remain systemic barriers to youth sport participation for many minoritized groups (Kanters et al., 2012). This may be due, in part, to the relatively high financial commitment required for youth sport participation (see Dunn et al., 2016). Existing literature relating to barriers to sport participation highlights a number of practical and person-centered barriers (Somerset & Hoare, 2018). Specifically, three main practical barriers have been identified by sport stakeholders: (1) Time, this may be either the parent or child’s schedule or transport related, (2) Cost, in the form of accessibility of quality equipment or facilities in the local area, and (3) Location, which includes issues with transport in the area as well as space, suitability, and access (Somerset & Hoare, 2018). Furthermore, person-centered barriers have the potential to include sex- or race-based stereotypes, bad experiences in other sport contexts, and negative appraisal (Somerset & Hoare, 2018), as well as insurance status and race (Pandya, 2021). These patterns, and the potential for them to have an exclusionary role in families’ sport participation decision making, appear to have long been an issue in the United States (Seefeldt & Ewing, 1997; Kingsley & Spencer-Cavaliere, 2015). Recent work from the Aspen Institute, 2019, Aspen Institute highlights discrepancies in youth sport participation rates by race from pre-to post-pandemic. Importantly, race is closely tied to the neighborhood in which individuals reside, and data suggest that minoritized individuals are more likely to live in structurally vulnerable neighborhoods (Berkowitz et al., 2020). This is concerning in light of the fact that lower SES neighborhoods were disproportionately impacted by the COVID-19 pandemic (Tai et al., 2020), specifically with regard to health care access and outcomes. Given present understanding, it appears that dimensions of family SES may be linked to the COVID-19-related opportunities and outcomes American youth experienced during the pandemic. Therefore, close examination of SES and race has the potential to be a fruitful research pathway. In pursuing this work, the integrated model of the youth sport system (Dorsch et al., 2022) provides an appropriate theoretical frame to highlight the importance of the numerous personal characteristics, persons, and contexts that influence athletes’ experiences and outcomes in youth sport. Specifically, it can be inferred that parents behave differently depending on the communities and societies in which they exist. Different communities and societies have different values, interests, and opportunities, altering the way in which parents interact with their children and sport. Further, parents are considered co-participants in their children’s youth sport experience and therefore have the potential to influence their athletes’ participation through their provisions of support (e.g., providing opportunities for participation), while also being influenced by their children and the contexts in which they interact. This makes it extremely important to consider parents’ perceptions of barriers related to their children’s return to sport in a post-COVID-19 landscape. Systems thinking places the individual and their personal characteristics at the center (Dorsch et al., 2022) and accounts, importantly, for development across time. The present study was designed to address this gap by examining families’ socioeconomic and demographic factors in relation to parents’ perceived barriers to returning to sport following the COVID-19 pandemic. It was hypothesized that multiple aspects of families’ socioeconomic and demographic characteristics (i.e., parent and child age and sex; family income and neighborhood; parent race, relationship status, and employment status) would be significantly associated with parents’ perceptions of health (e.g. getting sick if a child starts playing sports again) and practical barriers (e.g. difficulty fitting sports into the family’s schedule again), as well as their levels of sport participation as their children to returned following the COVID-19 pandemic. 1 Method 1.1 Participants Participants (N = 6183) were parents of children competing in youth sport in the United States during the COVID-19 pandemic. Parents were defined as the biological, adoptive, or otherwise regular caregivers of a child. This included mothers and fathers as well as any other individual who served as the primary caregiver of a child or adolescent athlete. Data were collected between May 2020 and September 2021. The sample included 2738 self-identified females, 3430 self-identified males, and 15 who did not identify. Parents ranged in age from 18 to 89 years (M age = 39.31, SD = 8.9). The nationally representative sample was 58.98% White, 18.11% Hispanic, Latino, or Spanish origin (Latinx), 13.38% Black or African American, 5.76% Asian, 1.99% Multiracial, 1.00% American Indian or Alaskan Native, 0.27% “other”, and 0.23% Native Hawaiian or Pacific Islander. Seventeen parents (0.27%) did not identify a specific race. The median gross household income was $69,000, in line with the national household average of $68,703 (United States Census Bureau, 2019). Children were reported to have a mean age of 12.21 years (SD = 3.22, range = 6 to 18) with 3386 males and 2789 females identified by parents. Eight parents did not report on their children’s gender. 1.2 Procedure An institutional review board approved study procedures prior to data collection. Parent respondents were recruited via a paid Qualtrics panel and through youth sport industry partners between June 2020 and August 2021. Participants were selected to participate based on socioeconomic and demographic characteristics (i.e., sex, race, income, state, size of community, etc.) to achieve a relatively representative sample of American youth sport parents. Respondents provided informed consent online, and then answered 13 socioeconomic and demographic questions about themselves, their oldest sport-participating child, and their family. Parents also responded to questions related to potential barriers to their children resuming participation in youth sport once their community of state lifted COVID-19-related restrictions. 1.3 Measures Socioeconomic and demographic characteristics. Parents were asked about their age, the racial category they identify with, their household annual income across all earners, before taxes, the type of neighborhood they reside in (urban, suburban, or rural), as well as their relationship and employment status. Parents also indicated their children’s sex and age. Barriers to sport. Single study-designed items assessed parents’ perceptions of a range of potential barriers to returning to sport post-pandemic. These included the fear of child illness “I am afraid of my child getting sick if he/she starts playing sports again”, the fear of parent illness “I am afraid of myself getting sick if my child starts playing sports again”, schedule conflicts “It will be difficult to fit sports into our schedule again”, transport difficulty “It will be too difficult to transport my child to play sports”, and their child losing interest in sport “My child is not interested in playing sports again”. All items were measured on a 5-point scale from 1 (strongly disagree) to 5 (strongly agree). In the present study, barrier variables were grouped into two factors (‘practical barriers’ and ‘health barriers’). A rationale for this decision is provided in the Data Analysis sub-section below. Sport Participation. Parents reported how many hours per week their children were currently taking part in youth sport. This item asked parents to specifically report on their children’s hours spent in virtual training (e.g., watching film, online coaching), practice and training, games and competition, and pickup/free play. Responses across these categories were summed to represent the total number of hours per week in which youth were actively engaged in sport during the pandemic. 1.4 Data Analysis Statistical analyses were conducted using R (R Core Team, 2021). Descriptive statistics were examined based on the recommendations of Tabachnick and Fidell (2018), using the psych package (Revelle, 2021). The structural equation models were fit using the lavaan package in R (Rosseel, 2012). We constructed a three-factor measurement model with latent variables “health barriers”, “practical barriers”, and “sport participation”. For comparison, we also fit a two-factor measurement model, comprised of “sport participation” and “barriers to sport” (see table 2 for final factor structure). We compared the two models using a likelihood ratio test, χ2 (1) = 3817.7, p < .001). Prior to creating the structural model, we created dummy codes for each categorical variable using the fastDummies package in R (Kaplan, 2020). The category with the largest membership was selected to serve as the reference category (parent sex = male; parent race = white; relationship status = married; employment = full-time, neighborhood type = urban; child sex = male). These referent groups were omitted from the regressions within the model, which allows them to be used as the reference. As a result, all parameter estimates presented are relative to the average for these referent groups. Subsequent to deciding on an appropriate measurement model and establishing referent groups, we fit a MIMIC model (Joreskog & Goldberger, 1975) to examine the association of socioeconomic and demographic variables with the latent outcomes of interest. The model can be seen in Fig. 1 . The structural regression portion of the model utilized dummy coded variables to represent each category of parent sex, race, neighborhood type, relationship status, employment status (Kaplan, 2020). The model also included the continuous predictors of parent age and family income. To aid in model convergence, family income was scaled by a factor of .001 (income in thousands) to reduce the difference in scales. Child age and sex were statistically accounted for by including them as covariates in the model, this allowed us to examine parent perceptions when child age and sex are held constant. We tested study hypotheses by examining regression coefficients within the model and related significance tests. Model fit was assessed using RMSEA, SRMR, CFI, and TLI. (Bentler, 2007; Hu & Bentler, 1995; McDonald & Ho, 2002). RMSEA and SRMR values of less than 0.06 are indicative of good fit, while CFI and TLI values of more than 0.95 indicate good model fit.Fig. 1 Proposed Structural Regression Model – simplified version, estimates will be made for each category within multicategory variables. Fig. 1 2 Results 2.1 Descriptive statistics Descriptive statistics are provided in Table 1 . As shown in Table 1, children participated in, on average, 1.8 virtual hours, 2.8 pickup hours, 2.5 practice hours, and 1.9 competition hours per week. Parents reported mean scores on health-related barriers of 3.5 for child health and 3.3 for parent health. Practical barrier mean scores were 2.8 for scheduling conflicts, 2.3 for children’s lack of interest in sport, and 2.5 for transportation-related concerns.Table 1 Descriptive statistics of study variables. Table 1Variable Mean SD Range Virtual Hours 1.8 3.75 0–40 Pickup Hours 2.8 4.13 0–40 Practice Hours 2.5 4.21 0–40 Competition Hours 1.9 3.50 0–40 Child Health - Barrier 3.5 1.35 1–5 Parent Health - Barrier 3.3 1.32 1–5 Schedule - Barrier 2.8 1.34 1–5 Interest - Barrier 2.3 1.37 1–5 Transport - Barrier 2.5 1.36 1–5 Table 2 Measurement model confirmatory factor analysis loadings and standard errors. Table 2Items Estimate SE Health Barriers Parent Health 1.00 Child Health 0.97 0.01 Sport Participation Virtual Hours 1.00 Pick-up Hours 0.96 0.03 Practice Hours 1.32 0.04 Competition Hours 1.37 0.04 Practical Barriers Interest 1.00 Schedule 0.99 0.01 Transport 1.08 0.01 CFI 0.98 TLI 0.97 RMSEA 0.06 SRMR 0.03 Note: The first factor loading for each factor was fixed to 1.00 for identification purposes. 2.2 Structural equation models Comparing measurement models using the likelihood ratio test showed strong support for the proposed three-factor structure, ͼ2(2) = 3817.7, p < .001. Factor structure and item loadings can be found in Table 2. Fit indices offer evidence of good fit for the measurement model (RMSEA = 0.03, SRMR = 0.01 CFI = 0.98, and TLI = 0.97). Loadings and standard errors for the three-factor measurement model can be found in Table 2. The MIMIC model yielded a number of significant findings, which are detailed in Table 3 as unstandardized betas. Fit indices offer evidence of good fit for the structural model (RMSEA = 0.03, CFI = 0.99, and TLI = 0.99). Concerning Health Barriers, the most salient findings involved race, employment status, and neighborhood type. Specifically, Asian (β = 0.27, p = .01), Black (β = 0.22, p = .01), and Latinx (β = 0.11, p = .03) parents were more likely to report higher levels of health-related barriers than white parents. While those with divorced parents reported lower levels of concern compared to married couples (β = −0.16, p = .03), as were parents in part-time (β = −0.14, p = .04) or self-employment (β = −0.21, p = .01) when compared to full-time working parents. On the contrary, those unable to work reported higher levels of concern regarding health-related barriers (β = 0.26, p = .03). Finally, those living in suburban or rural neighborhoods reported lower levels of concern when compared to those living in urban neighborhoods (β = −0.28, p < .001; β = −0.33, p = .01). Parent age was also a significant predictor, but its small effect size is unlikely to be of practical value (β = 0.01, p = .01).Table 3 Structural Regression Estimates, Standard Errors, and p-values. Table 3 Estimate SE P Health Barriers Sex Female 0.06 0.04 0.13 Race American Indian or Alaskan −0.05 0.18 0.77 Asian 0.27 0.08 0.01 Black or African American 0.22 0.06 0.01 Latinx 0.11 0.05 0.03 Multiracial −0.02 0.13 0.88 Relationship Status Single, Never Married 0.05 0.06 0.43 Living with Partner −0.05 0.07 0.44 Widowed −0.21 0.15 0.17 Divorced −0.16 0.08 0.03 Separated −0.12 0.13 0.38 Employment Status Part-Time −0.14 0.07 0.04 Self-Employed −0.21 0.08 0.01 Out of Work - COVID 0.02 0.08 0.84 Out of Work - Non-COVID −0.09 0.16 0.57 Homemaker 0.01 0.07 0.84 Student 0.20 0.17 0.24 Retired −0.05 0.14 0.71 Unable to Work 0.26 0.11 0.03 Neighborhood Type Suburban −0.28 0.04 0.01 Rural −0.33 0.06 0.01 Child Sex Female −0.04 0.04 0.27 Parent Age 0.01 0.00 0.01 Family Income 0.00 0.00 0.53 Child Age −0.02 0.01 0.01 Participation Sex Female −1.39 0.20 0.01 Race American Indian or Alaskan −0.60 0.88 0.50 Asian −1.75 0.38 0.01 Black or African American 0.04 0.27 0.88 Latinx −1.18 0.24 0.01 Multiracial −1.22 0.62 0.05 Relationship Status Single, Never Married −1.22 0.28 0.01 Living with Partner −0.88 0.31 0.01 Widowed −0.53 0.73 0.47 Divorced −1.28 0.36 0.01 Separated −1.80 0.64 0.01 Employment Status Part-Time −0.62 0.33 0.06 Self-Employed −2.26 0.39 0.01 Out of Work - COVID −3.06 0.40 0.01 Out of Work - Non-COVID −2.57 0.74 0.01 Homemaker −2.38 0.32 0.01 Student −1.79 0.83 0.03 Retired −1.26 0.67 0.06 Unable to Work −1.42 0.54 0.01 Neighborhood Type Suburban −1.62 0.19 0.01 Rural −1.98 0.27 0.01 Child Sex Female −0.16 0.18 0.37 Parent Age −0.10 0.01 0.00 Family Income 0.00 0.00 0.02 Child Age 0.09 0.03 0.01 Practical Barriers Sex Female −0.16 0.04 0.00 Race American Indian or Alaskan −0.01 0.18 0.98 Asian 0.06 0.08 0.44 Black or African American −0.10 0.06 0.07 Latinx −0.14 0.05 0.00 Multiracial −0.19 0.13 0.13 Relationship Status Single, Never Married −0.02 0.06 0.79 Living with Partner −0.19 0.06 0.00 Widowed 0.22 0.15 0.15 Divorced −0.13 0.07 0.07 Separated −0.09 0.13 0.52 Employment Status Part-Time −0.02 0.07 0.72 Self-Employed −0.26 0.08 0.00 Out of Work - COVID −0.36 0.08 0.00 Out of Work - Non-COVID −0.22 0.15 0.16 Homemaker −0.13 0.06 0.04 Student 0.04 0.17 0.81 Retired −0.28 0.14 0.04 Unable to Work 0.12 0.11 0.30 Neighborhood Type Suburban −0.33 0.04 0.00 Rural −0.35 0.06 0.00 Child Sex Female 0.09 0.04 0.02 Parent Age 0.00 0.00 0.07 Family Income 0.00 0.00 0.10 Child Age −0.02 0.01 0.00 Covariances Health Barriers - Participation 0.19 0.12 0.11 Health Barriers - Practical Barriers 1.19 0.03 0.00 Participation - Practical Barriers 1.00 0.06 0.00 Variances Health Barriers 1.75 0.04 0.00 Participation 31.21 0.24 0.00 Practical Barriers 1.65 0.03 0.00 Parent Health 0.28 0.02 0.00 Child Health 0.41 0.02 0.00 Virtual Hours 60.57 0.20 0.00 Pick-up Hours 80.48 0.26 0.00 Practice Hours 72.03 0.23 0.00 Competition Hours 51.03 0.16 0.00 Interest 0.85 0.02 0.00 Schedule 0.61 0.02 0.00 Transport 0.44 0.02 0.00 CFI 0.98 TLI 0.97 RMSEA 0.03 SRMR 0.01 Note: Significant relationships indicated in bold. Significant relationships were also identified between the socioeconomic and demographic variables of interest and Practical Barriers. Female parents reported lower levels of perceived practical barriers (β = −0.16, p < .001), as did parents that identified as Latinx when compared to White parents (β = −0.14, p < .001), and those living with a partner compared to married couples (β = −0.19, p < .001). Self-employed parents (β = −0.26, p < .001), those out of work because of COVID-19 (β = −0.36, p < .001), homemakers (β = −0.13, p = .04), and retired parents (β = −0.28, p = .04) all reported lower levels of perceived practical barriers compared to those in full-time employment. Finally, those living in suburban (β = −0.33, p < .001) and rural (β = −0.35, p < .001) neighborhoods reported lower levels of perceived practical barriers compared to those living in urban neighborhoods. There were several significant relationships between socioeconomic and demographic variables and sport participation. Firstly, female parents reported less participation (less time spent in the sport context) than male parents (β = −1.39, p = .01). On a similar note, Asian (β = −1.75, p = .01), Latinx (β = −1.18, p = .01), and Multiracial parents (β = −1.22, p = .05) reported lower levels of participation than White parents. Single never married parents (β = −1.22, p = .01), those living with partners (β = −0.88, p = .01), divorced (β = −1.28, p = .01), and separated parents (β = −1.80, p = .01) all reported less participation in youth sport when compared to married couples. Further, parents who are self-employed (β = −2.26, p = .01), out of work because of COVID-19 (β = −3.06, p = .01), out of work not related to COVID-19 (β = −2.57, p = .01), homemakers (β = −2.38, p = .01), students (β = −1.79, p = .01), and those unable to work (β = −1.42, p = .03) reported significantly lower levels of participation compared to full-time employed parents. Living in a suburban (β = −1.62, p = .01) or rural (β = −1.98, p = .01) neighborhood was also related to lower levels of sport participation. Finally, a higher parent age was associated with lower levels of participation (β = −0.10, p < .001). 3 Discussion The present study was designed to examine families’ socioeconomic and demographic factors in relation to parents’ perceived barriers to returning to sport following the COVID-19 pandemic. We hypothesized that multiple aspects of families’ socioeconomic and demographic characteristics would significantly predict parents’ perceived barriers and sport participation as their children returned to sport following the COVID-19 pandemic. We tested these hypotheses using a MIMIC model with the results presented in Table 2, Table 3. Regarding perceived Health Barriers, findings suggest that individuals who identified as a member of a minoritized group were more likely to report higher levels of concern when considering their children’s return to youth sport following the pandemic. Furthermore, individuals living in an urban neighborhood also reported higher levels of concern when compared to those residing in a suburban or rural area, even after for controlling for family income. A plausible explanation for this is that Black and Latinx families are more likely to reside in structurally vulnerable neighborhoods (Berkowitz et al., 2020). Previous literature notes that families who reside in these neighborhoods often have less access to safe parks and recreational facilities (Basch, 2011; Spengler, 2012). This is important because research highlights strong associations between the presence of neighborhood parks and recreational facilities and youth’s physical activity (Roemmich et al., 2006). However, systematic barriers such as recreational budgets often plague these neighborhoods from having safe and well-run recreational facilities for children and adolescents. Emerging epidemiological data suggest that there was a higher concentration of COVID-19 cases and related deaths in low-SES than high-SES neighborhoods (Tai et al., 2020). Research highlights disparity in access to affordable and proximal health care infrastructure and suggests that individuals from low-SES neighborhoods see higher mortality rates when infected ((Townsend, Kyle, & Stanford, 2020)). Indeed, the CDC (2020) noted that as of December 2020, individuals who identified as Black or Latinx had a 2.8 times higher mortality rate when compared to White individuals. Overall, findings from the present study suggest that membership in a minoritized racial group was associated with higher levels of health-related concerns upon children’s return to youth sport in the wake of the COVID-19 pandemic, net of income status. When addressing participation in youth sport during the COVID-19 pandemic, we also found significant differences in the amount of time spent in the youth sport setting between different races, relationship and employment statuses, and neighborhood type. First, male parents reported higher levels of sport participation, as did White parents when compared to Asian, Latinx, and Multiracial parents, but there was no significant difference between White and Black parents. Married parents also reported higher levels of sport participation, as did those parents who reported being employed full-time, likely due to the increased financial resources from this intersection. Finally, those living in urban neighborhoods reported higher levels of sport participation. These findings can be interpreted by considering them through an integrated lens of the youth sport system (Dorsch et al., 2022). Dorsch and colleagues suggest that the environmental subsystem influences the opportunities, infrastructure, access, and support for individuals to participate in sport. In the present study, results suggest that those living in urban areas are likely to have greater access to sport facilities, and thus opportunities, as a result of living in this type of environment where opportunities for participation may be more likely. This is directly supported by both theory and empirical research within the “community” level of the environmental subsystem of Dorsch and colleagues’ (2022) integrated model of the youth sport system. This work highlights the importance of initiatives, access, and infrastructure (O’Reilly et al., 2015). It may also be the case that at the “Society” level, there are relationships between local policy, resources, and values, that influence parent’s perceptions of barriers when returning to youth sport (Strittmatter & Skille, 2016; Tomik et al., 2012). Regarding race and participation, the function of youth sport and identity may play a salient role in explaining the difference between racial groups. This is a more nuanced finding to explicate when we consider the innumerable bidirectional relationships between the athlete and parents in the family subsystem with the community and society levels of the environmental subsystem. The present study examines both practical and person-cantered barriers (Somerset & Hoare, 2018). Data indicate that different person-cantered attributes such as parent sex, race, and marital status, influenced individual’s perceptions of health related and practical barriers. This aligns with existing literature regarding barriers to youth sport participation. Particularly the findings of Pandya (2021), who highlighted how certain individuals not only face a more difficult entry into sport, but also face disparities in the care received when injured. Given the health-related concerns parents had during the COVID-19 pandemic, the present findings appear to support the notion that race and SES have a significant influence on individuals’ experiences in youth sport (Pandya, 2021). However, the present study also demonstrates that these disparities continued throughout an international pandemic. The presence of a crisis on the scale of COVID-19 did not appear to place all individuals on a level playing field when considering their experiences of barriers to return to sport. Rather, the disparities that had been observed in society prior to Spring 2020 were likely exacerbated during the pandemic. This highlights the urgent need to reassess structural policy and youth sport opportunities (see Pandya, 2021; Somerset & Hoare, 2018), which have long been identified as needs of the American youth sport system (Seefeldt & Ewing, 1997). Given the abrupt loss of sport opportunities in the wake of the pandemic, it is important to consider the influence of COVID-19-related restrictions on individual’s identities. Extant literature notes that Black youth are encouraged more than other minoritized groups to participate in sport (Harris, 1994). Therefore, the sport may be more salient to Black youth -- in particular Black males' -- identity (Messner & Play, 1992). Meanwhile, parents of White athletes from higher SES groups are more likely to invest in sport than parents from other racial groups. Potential explanations for this include the prevalence of sport specialization as a means to increase the chances of obtaining a college athletic scholarship (Brooks et al., 2018). The final factor, Practical Barriers, also saw some significant differences between socioeconomic and demographic characteristics. Most differences were found between those in full-time employment and those in other employment categories. Parents employed on a full-time basis reported higher levels of practical barriers when compared to self-employed parents, those out of work due to COVID-19, homemakers, and retired parents. This makes practical sense in that those working full-time jobs may have more difficulties in scheduling and transport than those in more flexible positions, leading them to perceive more barriers. As illustrated by these findings, there remains significant support for the notion that socioeconomic and demographic factors had a significant impact on parents’ perceptions surrounding youth sport and the COVID-19 pandemic. Beyond the present results, there are several practical reasons to examine how and why the socioeconomic and demographic subcomponents of SES may shape sport participation experiences and outcomes among youth in minoritized groups. First, research indicates that Black and Latinx families tend to have more negative experiences concerning health, financial, and food security than White families (Berkowitz et al., 2020). Additionally, Black, Latinx, and Native American families are more likely to live in structurally vulnerable neighborhoods (Berkowitz et al., 2020). Importantly, these vulnerable neighborhoods may lack the necessary funding, support, and infrastructure to offer well-designed and safe sport experiences. This may result in fewer opportunities to play unorganized and free play sport in developmentally appropriate environments (Fraser-Thomas et al., 2005; Holt et al., 2009). Of note, structurally vulnerable neighborhoods had a higher concentration of COVID-19 cases and deaths during the peak of the pandemic (Verma et al., 2021). Existing literature suggests that some urban areas have an overrepresentation of sport facilities despite having a lower per capita youth population (O’Reilly et al., 2015), while American suburbs boast higher quality facilities in fewer number. Even if the necessary funding, support, and infrastructure are in place to provide equitable opportunities, it does not guarantee that youth or their parents would have different perceptions related to returning to organized sport due to heightened concerns of health-related barriers. This speaks to the intersectionality of health barriers and practical barriers, and the “double jeopardy” that may have limited rates of return-to-play in structurally vulnerable neighborhoods following the COVID-19 pandemic. In short, findings from the present research appear to be consistent with the broader COVID-19 literature in that racial minorities in structurally vulnerable neighborhoods have been disproportionately impacted by the pandemic (Berkowitz et al., 2020; CDC, 2020; Tai et al., 2020). 3.1 Practical implications Findings from the present study highlight the importance of actively working with minoritized groups in urban neighborhoods, where health related barriers and concerns seemed to be highest, to more effectively define a “new normal” in youth sport as society emerges from the COVID-19 pandemic. This process would be enhanced by re-evaluating federal, state, and local policy to create safe and developmentally appropriate spaces for children to participate, irrespective of their socioeconomic and demographic backgrounds. The benefits from participating in such environments are vast (Fraser-Thomas et al., 2005; Hills et al., 2015; Eime et al., 2013). It is therefore incumbent upon research-practitioners to design and deliver on the promise of “Sport for All” (see Wicker et al., 2009) by affording all children the opportunity to experience the benefits that sport and physical activity have to offer. 3.2 Limitations and future directions Despite the many strengths of this research, there are inherent limitations to consider. First, the present study was cross-sectional and is therefore useful for identifying statistical associations among variables and potential pathways of influence. Future literature should be designed to examine how stakeholder perceptions change over time, as such an approach would allow scholars to establish a developmental sequence of events while minimizing recall bias in participant responses (Caruana et al., 2015). In survey-based research, it is also important to consider the sources of information. In the present study, we utilized parents’ self-report measures of perceived barriers to returning to sport. This limitation is captured in a quote from the social anthropologist Margaret Mead, who said, “What people say, what people do, and what people say they do, are entirely different things” (Ewing, 2011, p. 80). Future work could therefore measure parent and child perceptions in an effort to capture a more holistic understanding of the family subsystem in youth sport. Alternatively, observational or behavior-based research could be utilized to better understand the objective manifestations of athletes’, parents’, and families’ health-related concerns. A second limitation of this study lies in the fact that results can only be considered within the context in which they were collected (i.e., youth sport parents in the United States). Although the sample of youth sport parents was drawn from all 50 states and the District of Columbia and is considered “representative” based on the socioeconomic and demographic composition of the sample, there are unique aspects of the larger population that could be considered. As highlighted by Dorsch et al. (2022) conceptualization of the youth sport system, factors related to the individual, family, team, organization, community, and society influence and are influenced by every athlete’s youth sport experience. Therefore, although the present results offer a sharper understanding of the socioeconomic and demographic factors that shape sport experiences and outcomes in the United States, they are not generalizable to an international audience. To be clear, every society around the world was impacted by the COVID-19 pandemic; however, it would behove future scholars to document the similarities and differences in the ways these societies, their constituent individuals, and youth sport contexts reacted to, and were impacted by, the onset of COVID-19. Furthermore, barriers may be perceived in different ways, or new barriers may be present in other settings. Future research could be designed to examine these barriers in other geographical regions, while potentially including new barriers such as financial stress. There are a variety of potential barriers that could have been explored in the present study. The barrier items that were included were specifically designed for this study in partnership and consultation with the funding agency, per their internal knowledge priorities. Moving forward qualitative work may be a useful tool for researchers to adopt in order to better understand why some families walked away from sport during the pandemic. This would also aid in the understanding of how the absence of sport had an influence on youth’s identities (Sanders & Stevinson, 2017). A final (de)limitation relates to the choice not to apply an intersectional lens to this research. Intersectional theoretical approaches have not been widely used in sport (Dagkas, 2016) and it has been relatively uncommon for such an approach to be adopted in quantitative research (Dill & Zambrana, 2009). In large part, this is due to the relative complexity of the models required to capture the constellation of variables that comprise one’s social identity. While the present study measured a number of key components of SES, further research should be designed to explore potential systemic differences and inequalities among groups. Theoretically informed work in this area holds the potential to influence policy and legislation that might positively affect those most vulnerable to the permanent (e.g., structures and systems) and temporary (e.g., the COVID-19 pandemic) factors that have the potential to affect youth sport participation in the United States. 4 Conclusion Youth sport is a relatively ubiquitous activity among youth in the United States. Sport has the potential to offer a range of positive developmental outcomes when it is designed and delivered toward the end users: children and adolescents. However, the recent COVID-19 pandemic caused unforeseen hardships for individuals, families, and communities, leaving many parents to make decisions about when, whether, and in what capacity their children would return to youth sport once COVID-19-related restrictions were lifted. The present study examined a range of socioeconomic and demographic factors with regard to this decision. Salient findings suggest that identifying as a minority and living in an urban neighborhood were related to the perception of health-related barriers in sport. Further, those who lived in urban neighborhoods and had higher family incomes reported higher levels of practical barriers to returning to sport. Theoretically informed intersectional research is needed to expand these findings and to identify areas of implementation for new policy and practice. Declaration of competing interest The authors have no existing conflicts of interest that would have the potential to influence the present study. Data availability The authors are unable or have chosen not to specify which data has been used. ☆ Research for this project was conducted with the support of the Aspen Institute Sports & Society Program. We thank the many sport parents from around the United States who took part in the research as well as the Aspen Institute’s Project Play 2020 partners for making this research possible. ==== Refs References American Psychological Association Socioeconomic status 2022 American Psychological Association https://www.apa.org/topics/socioeconomic-status Aspen Institute State of play report https://www.aspenprojectplay.org/national-youth-sport-survey 2019 Aspen Institute State of Play report https://www.aspenprojectplay.org/state-of-play-2021/introduction 2021 Basch C.E. Physical activity and the achievement gap among urban minority youth Journal of School Health 81 10 2011 626 634 21923875 Baxter-Jones A.D.G. Maffulli N. Parental influence on sport participation in elite young athletes The Journal of Sports Medicine and Physical Fitness 43 2003 250 255 12853909 Bentler P.M. On tests and indices for evaluating structural models Personality and Individual Differences 42 5 2007 825 829 10.1016/j.paid.2006.09.024 Berkowitz R.L. Gao X. Michaels E.K. Mujahid M.S. Structurally vulnerable neighborhood environments and racial/ethnic COVID-19 inequities Cities & Health 2020 1 4 10.1080/23748834.2020.1792069 Blau F.D. Meyerhofer P.A. Koebe J. Essential and Frontline Workers in the COVID-19 Crisis \| Econofact 2020 https://econofact.org/essential-and-frontline-workers-in-the-covid-19-crisis Braveman P.A. Kumanyika S. Fielding J. LaVeist T. Borrell L.N. Manderscheid R. Troutman A. Health disparities and health equity: The issue is justice American Journal of Public Health 101 S1 2011 S149 S155 10.2105/ajph.2010.300062 21551385 Brewer B.W. Van Raalte J.L. Linder D.E. Athletic identity: Hercules' muscles or Achilles' heel International Journal of Sport Psychology 24 1 1993 237 254 Brooks M.A. Post E.G. Trigsted S.M. Schaefer D.A. Wichman D.M. Watson A.M. McGuine T.A. Bell D.R. Knowledge, attitudes, and beliefs of youth club athletes toward sport specialization and sport participation Orthopedic Journal of Sports Medicine 6 5 2018 1 8 Caruana E. Roman M. Hernandez-Sanchez J. Solli P. Longitudinal studies Journal of Thoracic Disease 7 11 2015 537 540 Centers for Disease Control Risk for Covid-19 infection, hospitalization, and death by race/ethnicity https://www.cdc.gov/coronavirus/2019-ncov/covid-data/investigations-discovery/hospitalization-death-by-race-ethnicity.html 2020 Congressional Research Service Unemployment rates during the COVID-19 pandemic 2021 https://sgp.fas.org/crs/misc/R46554.pdf Côté J. Vierimaa M. The developmental model of sport participation: 15 years after its first conceptualization Science & Sports 29 1 2014 S63 S69 10.1016/j.scispo.2014.08.133 Dagkas S. Problematizing social justice in health pedagogy and youth sport: Intersectionality of race, ethnicity, and class Research Quarterly for Exercise & Sport 87 3 2016 221 229 10.1080/02701367.2016.1198672 27463227 Dill B.T. Zambrana R.E. Critical thinking about inequality: An emerging lens Emerging Intersections 1 2009 1 21 10.36019/9780813546513-003 Dorsch T.E. Blazo J.A. Findings from phase I of the COVID-19 national parenting survey: June, 2020 2021 Aspen Institute Sports & Society Program 1 59 Dorsch T.E. Smith A.L. Blazo J.A. Coakley J. Côté J. Wagstaff C.R.D. Warner S. King M.Q. Toward an integrated understanding of the youth sport system Research Quarterly for Exercise & Sport 93 1 2022 105 119 10.1080/02701367.2020.1810847 32960153 Dunn R.C. Dorsch T.E. King M.Q. Rothlisberger K.J. The impact of family financial investment on perceived parent pressure and child enjoyment and commitment in organized youth sport Family Relations 65 2 2016 287 299 10.1111/fare.12193 Edison B.R. Christino M.A. Rizzone K.H. Athletic identity in youth athletes: A systematic review of the literature International Journal of Environmental Research and Public Health 18 14 2021 1 18 10.3390/ijerph18147331 Eime R.M. Young J.A. Harvey J.T. Charity M.J. Payne W.R. A systematic review of the psychological and social benefits of participation in sport for children and adolescents: Informing development of a conceptual model of health through sport International Journal of Behavioral Nutrition and Physical Activity 10 1 2013 98 10.1186/1479-5868-10-98 23945179 Ewing S. Architecture and field/work 2011 Routledge Fraser-Thomas J.L. Côté J. Deakin J. Youth sport programs: An avenue to foster positive youth development Physical Education and Sport Pedagogy 10 1 2005 19 40 10.1080/1740898042000334890 Fredricks J.A. Eccles J.S. Parental influences on youth involvement in sports Weiss M.R. Developmental sport and exercise psychology: A lifespan perspective 2004 145 164 Fit Information Technology Graupensperger S. Benson A.J. Kilmer J.R. Evans M.B. Social (un)distancing: Teammate interactions, athletic identity, and mental health of student-athletes during the COVID-19 pandemic Journal of Adolescent Health 67 5 2020 10.1016/j.jadohealth.2020.08.001 Grix J. Brannagan P.M. Grimes H. Neville R. The impact of Covid-19 on sport International Journal of Sport Policy and Politics 13 1 2020 1 12 10.1080/19406940.2020.1851285 Harper S.R. Williams C.D. Blackman H.W. Black male student-athletes and racial inequities in NCAA Division I college sports 2013 University of Pennsylvania, Center for the Study of Race and Equity in Education Philadelphia Harris O. Race, sport, and social support Sociology of Sport Journal 11 1 1994 40 50 Hills A.P. Dengel D.R. Lubans D.R. Supporting public health priorities: Recommendations for physical education and physical activity promotion in schools Progress in Cardiovascular Diseases 57 4 2015 368 374 10.1016/j.pcad.2014.09.010 25269062 Holt N.L. Cunningham C.-T. Sehn Z.L. Spence J.C. Newton A.S. Ball G.D.C. Neighborhood physical activity opportunities for inner-city children and youth Health & Place 15 4 2009 1022 1028 10.1016/j.healthplace.2009.04.002 19457701 Hu L.T. Bentler P.M. Evaluating model fit Hoyle R.H. Structural equation modeling: Concepts, issues, and applications 1995 Sage Publications Inc 76 99 Joreskog K.G. Goldberger A.S. Estimation of a model with multiple indicators and multiple causes of a single latent variable Journal of the American Statistical Association 70 351 1975 631 10.2307/2285946 Kanters M.A. Bocarro J.N. Edwards M.B. Casper J.M. Floyd M.F. School sport participation under two school sport policies: Comparisons by race/ethnicity, gender, and socioeconomic status Annals of Behavioral Medicine 45 S1 2012 113 121 10.1007/s12160-012-9413-2 Kaplan J. Fastdummies: Fast creation of dummy (binary) columns and rows from categorical variables 2020 R package version 1.6.3 https://CRAN.R-project.org/package=fastDummies Kingsley B.C. Spencer-Cavaliere N. The exclusionary practices of youth sport Social Inclusion 3 3 2015 24 10.17645/si.v3i3.136 Lamont-Mills A. Christensen S.A. Athletic identity and its relationship to sport participation levels Journal of Science and Medicine in Sport 9 6 2006 472 478 16765643 Malina R.M. Children and adolescents in the sport culture: the overwhelming majority to the select few Journal of Exercise Science & Fitness 7 2 2009 1 10 McDonald R.P. Ho M.-H.R. Principles and practice in reporting structural equation analyses Psychological Methods 7 1 2002 64 82 10.1037/1082-989x.7.1.64 11928891 McGuine T.A. Biese K. Hetzel S.J. Schwarz A. Kliethermes S. Reardon C.L. Bell D.R. Brooks M.A. Watson M.D. A M. High school sports during the COVID-19 pandemic: The impact of sport participation on the health of adolescents Journal of Athletic Training 57 1 2021 10.4085/1062-6050-0121.21 McKay C. Cumming S. Blake T. Youth sport: Friend or foe? Best Practice & Research Clinical Rheumatology 33 2019 141 157 31431268 McMillan R. McIsaac M. Janssen I. Family structure as a correlate of organized sport participation among youth PLoS One 11 2 2016 e0147403 10.1371/journal.pone.0147403 Merkel D. Youth sport: Positive and negative impact on young athletes Open Access Journal of Sports Medicine 4 4 2013 151 10.2147/oajsm.s33556 24379720 Messner M.A. Play P.A. Sports and the problem of masculinity 1992 Power at Play O'Reilly N. Berger I.E. Hernandez T. Parent M.M. Séguin B. Urban sportscapes: An environmental deterministic perspective on the management of youth sport participation Sport Management Review 18 2 2015 291 307 10.1016/j.smr.2014.07.003 Pandya N.K. Disparities in youth sports and barriers to participation Current Reviews in Musculoskeletal Medicine 14 1 2021 441 446 10.1007/s12178-021-09716-5 34622353 Patel J.A. Nielsen F.B.H. Badiani A.A. Assi S. Unadkat V.A. Patel B. Ravindrane R. Wardle H. Poverty, inequality and COVID-19: The forgotten vulnerable Public Health 183 2020 110 111 10.1016/j.puhe.2020.05.006 32502699 R Core Team R: A language and environment for statistical computing 2021 R Foundation for Statistical Computing Vienna, Austria URL Https://www.R-project.org/ Revelle W. Psych: Procedures for personality and psychological research 2021 Northwestern University Evanston, Illinois, USA https://CRAN.R-project.org/package=psych Version = 2.1.9 Roemmich J.N. Epstein L.H. Raja S. Yin L. Robinson J. Winiewicz D. Association of access to parks and recreational facilities with the physical activity of young children Preventive Medicine 43 6 2006 437 441 16928396 Rosseel Y. Lavaan: An R package for structural equation modeling Journal of Statistical Software 48 2 2012 1 36 Sanderson J. Brown K. COVID-19 and youth sports: Psychological, developmental, and economic Impacts International Journal of Sport Communication 13 3 2020 313 323 10.1123/ijsc.2020-0236 Sanders G. Stevinson C. Associations between retirement reasons, chronic pain, athletic identity, and depressive symptoms among former professional footballers European Journal of Sport Science 17 10 2017 1311 1318 28911275 Seefeldt V.D. Ewing M.E Youth Sports in America: An Overview President’s Council on Physical Fitnes and Sports 2 1 1997 Somerset S. Hoare D.J. Barriers to voluntary participation in sport for children: A systematic review BMC Pediatrics 18 1 2018 10.1186/s12887-018-1014-1 Spengler J.O. Promoting physical activity through the shared use of school and community recreational resources. A research brief 2012 Active Living Research, a National Program of the Robert Wood Johnson Foundation Princeton, NJ Sports & Fitness Industry Association 2020 sports, fitness, and leisure activities topline participation report https://www.sfia.org/reports/802_2020-Sports%2C-Fitness%2C-and-Leisure-Activities-Topline-Participation-Report 2020 Strittmatter A.-M. Skille E.Å. Boosting youth sport? Implementation of Norwegian youth sport policy through the 2016 lillehammer winter youth olympic games Sport in Society 20 1 2016 144 160 10.1080/17430437.2015.1124568 Tabachnick B.G. Fidell L.S. Using multivariate statistics 2018 Pearson Tai D.B.G. Shah A. Doubeni C.A. Sia I.G. Wieland M.L. The disproportionate impact of COVID-19 on racial and ethnic minorities in the United States Clinical Infectious Diseases 72 4 2020 10.1093/cid/ciaa815 Tasiemski T. Kennedy P. Gardner B.P. Blaikley R.A. Athletic identity and sports participation in people with spinal cord injury Adapted Physical Activity Quarterly 21 4 2004 364 378 Tomik R. Olex-Zarychta D. Mynarski W. Social values of sport participation and their significance for youth attitudes towards physical education and sport Studies in Physical Culture and Tourism 19 2 2012 99 104 Townsend M.J. Kyle T.K. Stanford F.C. Outcomes of COVID-19: disparities in obesity and by ethnicity/race International Journal of Obesity 44 9 2020 1807 1809 10.1038/s41366-020-0635-2 32647359 United States Census Bureau Survey redesigns make Comparisons to years before 2017 difficult. Census.gov 2019 https://www.census.gov/library/stories/2019/09/us-median-household-income-not-significantly-different-from-2017.html Verma R. Yabe T. Ukkusuri S.V. Spatiotemporal contact density explains the disparity of COVID-19 spread in urban neighborhoods Scientific Reports 11 1 2021 10.1038/s41598-021-90483-1 Wanberg C.R. Csillag B. Douglass R.P. Zhou L. Pollard M.S. Socioeconomic status and well-being during COVID-19: A resource-based examination Journal of Applied Psychology 105 12 2020 1382 1396 10.1037/apl0000831 33090858 Wicker P. Breuer C. Pawlowski T. Promoting sport for all to age-specific target groups: The impact of sport infrastructure European Sport Management Quarterly 9 2 2009 103 118 10.1080/16184740802571377	36465329	PMC9710102	NO-CC CODE	2022-12-09 23:14:53	no	Psychol Sport Exerc. 2023 Mar 30; 65:102348	utf-8	Psychol Sport Exerc	2,022	10.1016/j.psychsport.2022.102348	oa_other
==== Front J Infect Public Health J Infect Public Health Journal of Infection and Public Health 1876-0341 1876-035X The Author(s). Published by Elsevier Ltd on behalf of King Saud Bin Abdulaziz University for Health Sciences. S1876-0341(22)00333-1 10.1016/j.jiph.2022.11.030 Editorial Comparison of the Causes of Death Associated With Delta and Omicron SARS-CoV-2 Variants Infection Kim A Reum Lee Jiyoung Park Somi Kang Sung Woon Lee Yun Woo Lim So Yun Chang Euijin Bae Seongman Jung Jiwon Kim Min Jae Chong Yong Pil Lee Sang-Oh Choi Sang-Ho Kim Yang Soo Kim Sung-Han ⁎ Department of Infectious Diseases, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea ⁎ Correspondence to: Department of Infectious Diseases, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro-43-gil, Songpa-gu, Seoul, 05505, South Korea. Fax: +82-2-3010-6970. 30 11 2022 30 11 2022 7 9 2022 24 11 2022 27 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. Keywords coronavirus COVID-19 SARS-CoV-2 Delta omicron Cause of Death ==== Body pmcDear Editor The recent study reported that substantial excess mortality occurred during the Omicron-dominant era, although Omicron variant may cause milder COVID-19 [1]. However, these data were based on the mortality statistics record, so the exact causes of deaths were not known and uncertainty largely remains regarding the relative contribution of Omicron-variant infection to deaths. Therefore, we investigated the causes of death among COVID-19 patients with Delta- and Omicron-variant infections. We retrospectively reviewed the medical records of adult patients with COVID-19 who were admitted to Asan Medical Center, Seoul, South Korea, between July 2021 and March 2022. The causes of death were classified into COVID-19 pneumonia, other causes, and indeterminate cause. The study was approved by the institutional review board of Asan Medical Center (IRB No 2022-0154), and informed consent was waived because of the retrospective nature of this study. A total of 1,020 COVID-19 patients were hospitalized at our center between July 2021 and March 2022, among whom 366 were admitted during the Delta-dominant period (Jul 2021- Dec 2021), and 654 were admitted during the Omicron-dominant period (Feb 2022- Mar 2022). The demographic and clinical characteristics, along with the causes of death of the COVID-19 patients, are shown in Table 1. During the Delta-dominant period, 42 (11%) of 366 patient with COVID-19 were admitted died. During the Omicron-dominant period, 42 (6%) of 654 patients with COVID-19 were admitted died (Supplemental Figure 1). The primary cause of death was COVID-19–associated pneumonia in both the Omicron (64%, 27/42) and Delta (88%, 37/42) eras (P = 0.01). Univariable analysis revealed that old age, COVID-19 severity, variant types, and solid cancer were associated with COVID-19-pneumonia-associated deaths. Multivariable analysis exhibited that old age and underlying solid cancer were independently associated with COVID-19-pneumonia-associated deaths (Supplemental Table 1). Unadjusted odds ratio (OR) for COVID-19-pneumonia-associated deaths in Delta group compared with those in Omicron group was 4.11 (95% CI 1.33-12.69, p value=0.01). However, after adjustment of potential confounders, there was a trend toward being higher COVID-19-pneumonia-associated deaths in Delta variant infection than in Omicron variant infection (OR=3.84, 95% CI 0.95-18.65, p value=0.07) (Supplemental Table 1).Table 1 Baseline clinical characteristics and causes of deaths between patients during delta variant dominant period and omicron variant dominant period. Table 1Characteristics Total (n=84) Delta (n=42) Omicron(n=42) Pvalue Age, years, median (IQR) 72.5 (65.0-81.0) 74.0 (66.8-82.0) 71.5 (60.8-81.0) 0.282 Male gender 57 (67.9) 27 (64.3) 30 (71.4) 0.483 Comorbidities Diabetes mellitus 22 (26.2) 9 (21.4) 13 (31.0) 0.321 Hypertension 40 (47.6) 22 (52.4) 18 (42.9) 0.382 Cardiovascular disease 23 (27.4) 10 (23.8) 13 (31.0) 0.463 Chronic lung disease 13 (15.5) 5 (11.9) 8 (19.0) 0.365 Liver disease 10 (11.9) 3 (7.1) 7 (16.7) 0.178 Renal disease 15 (17.9) 9 (21.4) 6 (14.3) 0.393 Solid cancer 31 (36.9) 13 (31.0) 18 (42.9) 0.258 Hematologic malignancy 4 (4.8) 0 4 (9.5) 0.116 Rheumatic disease 4 (4.8) 4 (9.5) 0 0.116 Obesity (BMI>25) 21 (25.0) 12 (28.6) 9 (21.4) 0.450 Smoking 2 (2.4) 0 2 (4.8) 0.494 Symptoms at diagnosis Fever 18 (21.4) 11 (26.2) 7 (16.7) 0.287 Chill 4 (4.8) 3 (7.1) 1 (2.4) 0.616 Cough 20 (23.8) 15 (35.7) 5 (11.9) 0.010 Sputum 15 (17.9) 8 (19.0) 7 (16.7) 0.776 Sore throat 5 (6.0) 3 (7.1) 2 (4.8) >0.999 Dyspnea 50 (59.5) 22 (52.4) 28 (66.7) 0.182 Rhinorrhea 0 0 0 NA Hemoptysis 1 (1.2) 0 1 (2.4) >0.999 Chest pain 4 (4.8) 0 4 (9.5) 0.116 Diarrhea 0 0 0 NA Headache 4 (4.8) 3 (7.1) 1 (2.4) 0.616 Myalgia 3 (3.6) 3 (7.1) 0 0.241 Hypogeusia 1 (1.2) 1 (2.4) 0 >0.999 Pneumonia at admission 74 (88.1) 40 (95.2) 34 (81.0) 0.043 Severity 0.175 Mild to moderate 19 7 12 Severe 34 21 13 Critical 31 14 17 Deaths 84/1,020 (8.2) 42/366 (11.5) 42/654 (6.4) COVID-19 pneumonia 64/84 (76.2) 37/42 (88.1) 27/42 (64.3) 0.010 Other causes 19/84 (22.6) 5/42 (11.9) 14/42 (33.3) 0.019 Underlying disease 5 1 4 Cardiovascular disease 6 1 5 Bleeding 3 1 2 Sepsis 5 2 3 Indeterminate cause 1/84 (1.2) 0 1/42 (2.4) >0.999 Data are presented as number (%) unless otherwise indicated. Abbreviations: BMI, body mass index; COVID-19, coronavirus disease 2019, IQR, interquartile range One study revealed that COVID-19 was documented as the direct cause of death in more than 90% of hospitalized patients with COVID-19 who eventually died [2]. On the other hand, the government’s COVID-19 death figures are based on total deaths from any cause in patients recently diagnosed with COVID-19 [3], so the overestimations are inevitable. Our data showed that about two-thirds of the deaths among hospitalized patients with COVID-19 during the Omicron era were attributed directly to COVID-19, as were the majority of deaths among patients with COVID-19 during the Delta era. Overwhelming number of patients with Omicron infections can cause collateral damage of healthcare system. For example, inaccessibility of intensive care service or emergency service due to overwhelming COVID-19 patients may result in excess deaths during the large community outbreaks. Further studies are urgently needed on the collateral damage of COVID-19 to healthcare system in SARS-CoV-2-uninfected patients. Although there are several retrospective studies [4], [5], [6], [7], [8], [9], [10] on the severity and risk factor of COVID-19 infection according to variant types, there are few prospective studies on this area. And data directly comparing the causes of death between Omicron variant and Delta variant infection are also limited. The small number of deaths and the analysis solely of patients admitted to a tertiary care hospital may limit the generalizability of our findings. Assuming that in-hospital mortality of COVID-19 patients was 15.1% during Delta period and 4.9% during later Omicron period [10], 135 participants were needed for each group for 80% power of the study and 5% margin of error. Therefore, our findings of Delta variant infection having marginal statistical significance on COVID-19-pneumonia-associated mortality after adjustment of potential confounders might be due to low study power. Taken together, our data clearly demonstrated that the excess mortality during the Omicron-dominant period in highly vaccinated area like South Korea could be explained by about one-third indirect contribution of COVID-19 in SARS-CoV-2-infected dead patients as well as by about two-third direct contribution of COVID-19 in SARS-CoV-2-infected dead patients. Although we could not access the collateral damage of COVID-19 in SARS-CoV-2-uninfected dead patients, our data provide important insight for us to figure out the mortality of SARS-CoV-2-infected patients with the relative contribution to excess mortality during the Omicron outbreaks. Despite some limitations, our data suggest that the Omicron variant has a relatively lower contribution to deaths than the Delta variant. However, during Omicron-dominant waves, large numbers of hospitalized patients still overwhelm health systems, and absolute mortality figures remain high despite relatively lower mortality rates compared with pandemic waves in which more virulent variants predominate. Therefore, our findings provide valuable and timely insight to facilitate preparedness for the emergence of less-virulent and more-transmissible variants. Funding Disclosure This work was supported by a grant from 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]. Appendix A Supplementary material Supplementary material Acknowledgements Not applicable. Conflicts of Interest Disclosure All authors have no potential conflicts of interest. Appendix A Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jiph.2022.11.030. ==== Refs References 1 Faust J.S. Du C. Liang C. Excess mortality in Massachusetts during the delta and omicron waves of COVID-19 JAMA. 328 1 2022 74 76 35594035 2 Slater T.A. Straw S. Drozd M. Kamalathasan S. Cowley A. Witte K.K. Dying ‘due to’ or ‘with’ COVID-19: a cause of death analysis in hospitalized patients Clin Med (Lond) 20 5 2020 e189 e190 32753516 3 Iacobucci G. Covid-19: unravelling the conundrum of omicron and deaths BMJ. 376 2022 o254 35091392 4 Ulloa A.C. Buchan S.A. Daneman N. Brown Ka. Estimates of SARS-CoV-2 Omicron Variant severity in Ontario, Canada JAMA. 327 13 2022 1286 1288 35175280 5 Bouzid D. Visseaux B. Kassasseya C. Comparison of Patients Infected with Delta Versus Omicron COVID-19 Variants Presenting to Paris Emergency Departments: A Retrospective Cohort Study Ann Intern Med 2022 M22 0308 6 Wald I.L. Bermingham C. Risk of covid-19 related deaths for SARS-CoV-2 omicron (B.1.1.529) compared with delta (B.1.617.2): retrospective cohort study BMJ. 378 2022 e070695 7 Nyberg Tommy Comparative analysis of the risks of hospitalization and death associated with SARS-CoV-2 omicron (B.1.1.529) and delta (B.1.617.2) variants in England: a cohort study The Lancet 399 10332 2022 1303 1312 8 Mary F.B. Talisa V.B. Castro A.D. Shaikh O.S. Omer S.B. Butt A.A. COVID-19 disease severity in US Veterans infected during Omicron and Delta variant predominant periods Nat commun 13 1 2022 3647 35752687 9 Butt A.A. Darhgam S.R. Tang P. COVID-19 disease severity in persons infected with the Omicron variant compared with Delta variant in Qatar J Glob Health 12 2022 05032 35788085 10 Adjei S. Hong K. Molinari N.M. Mortality Risk Among Patients Hospitalized Primarily for COVID-19 During the Omicron and Delta Variant Pandemic Periods – United states. April 2020- June 2022 MMWR. Morb Mortal Wkly Rep 71 37 2022 1182 1189 36107788	36516648	PMC9710103	NO-CC CODE	2022-12-13 23:16:25	no	J Infect Public Health. 2023 Jan 30; 16(1):133-135	utf-8	J Infect Public Health	2,022	10.1016/j.jiph.2022.11.030	oa_other
==== Front Actas Dermosifiliogr Actas Dermosifiliogr Actas Dermo-Sifiliograficas 0001-7310 1578-2190 AEDV. Published by Elsevier España, S.L.U. S0001-7310(22)01007-9 10.1016/j.ad.2022.07.027 Original Comportamiento de las principales infecciones de transmisión sexual bacterianas durante la pandemia por SARS-CoV-2 [[Translated article]]Epidemiologic Profile of the Main Bacterial Sexually Transmitted Infections During the SARS-CoV-2 PandemicCasanova-Esquembre A. 1⁎ Fuster Escrivá B. 2 Lorca Spröhnle J. 1 Labrandero-Hoyos C. 1 Peñuelas-Leal R. 1 Gimeno Cardona C. 2 Pérez-Ferriols A. 1 Hernández-Bel P. 1 1 Servicio de Dermatología, Hospital General Universitario de Valencia, España 2 Servicio de Microbiología, Hospital General Universitario de Valencia, España ⁎ Corresponding author: Servicio de Dermatología, Hospital General Universitario de Valencia, Av. Tres Creus, 2, 46014 Valencia, Spain 30 11 2022 30 11 2022 10 4 2022 8 7 2022 © 2022 AEDV. Published by Elsevier España, S.L.U. 2022 AEDV 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. Introducción y objetivo: Entre 2020-2021 hemos sufrido una pandemia por el virus SARS-CoV2. Debido a los cambios sociales por la pandemia, nos hemos propuesto estudiar el comportamiento epidemiológico de las principales infecciones de transmisión sexual (ITS) bacterianas (clamidia, gonorrea y sífilis) durante este periodo y compararlo con periodos previos. Material y método: Se recogieron las muestras mensuales de clamidia, gonorrea y sífilis entre los años 2018-2021 y las muestras positivas mensuales de SARS-CoV2 entre los años 2020-2021 del Hospital General Universitario de Valencia, analizadas por técnica PCR Multiplex. Se recogieron datos clínicos y demográficos de los pacientes con ITS. Resultados: Durante los años 2020-2021 (pandemia) se diagnosticaron más casos de ITS total (664) que los años 2018-2019 (pre-pandemia) (570), con tasas de incidencia superiores y tasas de positividad similares. Se ha observado una correlación cronológica negativa entre las muestras positivas de SARS-CoV2 y las muestras positivas de ITS. La edad media de los pacientes con diagnóstico de clamidia, gonorrea y sífilis fue de 29,64 (19,33-41,14 IC 95%), 30,86 (20,24-42,45 IC 95%) y 37,04 (26,01-51,00 IC 95%), años respectivamente. El número de casos de clamidia en varones ha aumentado un 13.85% (6,39-21,08 IC 95%; p_valor=0,0003) en los años de pandemia. Conclusión: Durante los años de pandemia, ha existido una correlación negativa entre los casos de SARS-CoV2 y los casos de ITS, con más casos de clamidia en varones. Las ITS han aumentado en los dos últimos años, por lo que son un importante problema de salud en la población joven y adulta que merece especial atención. [[en]]Abstract Background and objective: The COVID-19 pandemic brought about social changes in 2020 and 2021. The aim of this study was to evaluate the epidemiologic profiles of the main sexually transmitted infections (STIs) of bacterial origin (chlamydia, gonorrhea, and syphilis) diagnosed during this period and compare them to findings from previous years. Material and methods: Drawing on data from Hospital General Universitario in Valencia, Spain, we recorded the number of chlamydia, gonorrhea, and syphilis cases diagnosed monthly by multiplex polymerase chain reaction (PCR) in 2018-2021 and the number of PCR-confirmed SARS-CoV-2 cases diagnosed monthly in 2020-2021. We also collected clinical and demographic information on all patients diagnosed with STIs during the years studied. Results: The total number of STIs diagnosed increased from 570 in 2018-2019 to 664 in 2020-2021. PCR positivity rates were similar in the 2 periods, but the incidence rates were higher during the pandemic. The chronologic correlation between SARS-CoV-2 and STI positivity was negative. Mean age at diagnosis was 29.64 years (95% CI, 19.33-41.14 years) for chlamydia, 30.86 years (95% CI, 20.24-42.45 years) for gonorrhea, and 37.04 years (95% CI, 26.01-51.00 years) for syphilis. The number of men diagnosed with chlamydia increased by 13.85% (95% CI, 6.39-21.08; P = .0003) during the pandemic. Conclusions: We observed a negative correlation between SARS-CoV2 infections and STIs during the pandemic and an increase in chlamydia cases among men. STI cases rose during 2020-2021, indicating that they remain a significant problem that needs to be addressed in young and adult populations. Palabras Clave Infecciones de transmisión sexual Chlamydia Gonorrhea Sífilis SARS-CoV2 Keywords Sexually transmitted infections Chlamydia Gonorrhea Syphilis SARS-CoV-2 ==== Body pmc	36462671	PMC9710105	NO-CC CODE	2022-12-01 23:23:12	no	Actas Dermosifiliogr. 2022 Nov 30; doi: 10.1016/j.ad.2022.07.027	utf-8	Actas Dermosifiliogr	2,022	10.1016/j.ad.2022.07.027	oa_other
==== Front Neurologia (Engl Ed) Neurologia (Engl Ed) Neurologia (Barcelona, Spain) 2173-5808 Sociedad Española de Neurología. Published by Elsevier España, S.L.U. S2173-5808(22)00184-5 10.1016/j.nrleng.2022.01.007 Letter to the Editor Parsonage-Turner syndrome associated with COVID-19: About 2 family cases Síndrome de Parsonage Turner asociado a Covid-19: A propósito de 2 casos familiaresPivaral Cabrera C.E. a Rincón Sánchez A.R. b Dávalos Rodríguez N.O. c Ramirez Garcia S.A. d⁎ a Departamento de Salud Pública, Centro Universitario de Ciencias de la Salud, Universidad de Guadalajara, Guadalajara, Jalisco, Mexico b Instituto de Biología Molecular y Terapia génica, Centro Universitario de Ciencias de la Salud, Universidad de Guadalajara, Guadalajara, Jalisco, Mexico c Instituto de Genética Humana, Centro Universitario de Ciencias de la Salud, Universidad de Guadalajara, Guadalajara, Jalisco, Mexico d Universidad de la Sierra Sur, Miahuatlán de Porfirio Díaz, Oaxaca, Mexico ⁎ Corresponding author. 30 11 2022 30 11 2022 © 2022 Sociedad Española de Neurología. Published by Elsevier España, S.L.U. 2022 Sociedad Española de Neurología 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 pmcWe would like to underscore the relevance of the case of Parsonage-Turner syndrome (PTS) secondary to SARS-CoV-2 infection reported by Alvarado et al.,1 as this syndrome has an incidence of 1.6 cases per 100 000 population. PTS has been associated with viral infections triggering an immune-mediated response against the brachial plexus; however, this case associated with SARS-CoV-2 infection is particularly relevant as the syndrome may constitute a clinical manifestation or neurological complication of COVID-19, even after vaccination.2, 3 Familial aggregation has been observed in PTS, which suggests the involvement of genetic factors.4 Cases have been reported of hereditary neuralgic amyotrophy, with a clinical phenotype similar to that of PTS, linked to mutations at a locus on chromosome 17q25 (SEPT9 gene, between the markers D17S1301 [centromeric] and D17S784 [telomeric]). However, and unlike in PTS, hereditary neuralgic amyotrophy presents from the second decade of life, and a founder effect has been described in families from the United States.4, 5 Few candidate gene studies have been conducted for PTS. Previous studies helped us to establish the clinical diagnosis of familial neuralgic amyotrophy in a rural setting in 2 siblings with SARS-CoV-2 infection (beta variant, PANGO lineage B1.1.35, clade GH/501Y.V2 [clade-specific TaqMan RT-qPCR]), a man and a woman aged 44 and 45 years, respectively. Both patients presented severe pain in the left shoulder associated with a tingling sensation; the physical examination detected atrophy of the deltoid, supraspinatus, and scapular muscles. They also presented odynophagia, fever of 40°, and palatopharyngeal vesicular enanthem (an early finding of COVID-19)6; anteroposterior chest radiography revealed ground-glass opacification. Electromyoneurography revealed abnormal postganglionic sensory nerve action potentials in the left brachial plexus at the level of the trunk and cervical vertebrae, from the C2 to the C5 nerve roots, suggestive of acute brachial plexopathy. Both patients were intolerant to corticosteroids. They gave informed consent to treatment. They were treated on an outpatient basis, receiving ciclosporin A dosed at 5 mg/kg/day for 3 days (this drug may inhibit viral replication, blocking 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine, which prevents mitochondrial dysfunction associated with COVID-19 and blocks the cytokine storm, inhibiting the calcineurin inflammatory pathway and NF-kB7), ivermectin 6 mg/12 h for 7 days, azithromycin 500 mg/day for 5 days, lopinavir/ritonavir 2 tablets/12 h for 14 days, subcutaneous interferon-beta 0.25 mg every 24 h for 14 days, hydroxychloroquine 200 mg/12 h for 14 days, and OM85 one capsule/day for 30 days (Broncho-Vaxom, an immunostimulant and immunoregulator promoting opsonisation and mediating immune response via β-defensins and interferons), and subcutaneous enoxaparin 60 mg/day for 30 days. Brachial amyotrophic neuralgia was initially managed with non-steroidal anti-inflammatory drugs, opioids, and neuromodulators such as gabapentin, due to the patients’ intolerance to corticosteroids. During the first 22 days of infection, COVID-19 symptoms and brachial neuralgia did not respond to treatment in either patient; RT-qPCR yielded positive results for the SARS-CoV-2 E-gene, and non-steroidal anti-inflammatory drugs, opioids, and neuromodulators were withdrawn. We added prolonged-release pirfenidone (600 mg/12 h) given the drug’s strong anti-inflammatory, antioxidant, antiviral, and antifibrogenic properties and its ability to stabilise oxygen saturation; this drug was used as an adjuvant to control viraemia, neuralgia, and perineurial oedema associated with SARS-CoV-2 infection.8 At day 30 after COVID-19 symptom onset, both patients were asymptomatic and RT-qPCR yielded negative results, and they presented oxygen saturation levels of 96% and 98%, respectively. However, left brachial plexus neuralgia persisted and the patients continued to test positive for SARS-CoV-2 IgG antibodies; therefore, pirfenidone was not discontinued. Symptoms of brachial plexus neuralgia resolved on day 21 post-infection in the woman and on day 26 in the man. Both patients tested negative for SARS-CoV-2 IgG antibodies; pirfenidone was withdrawn. In the rural setting of Sierra Sur de Oaxaca, in Mexico, the management of neuropathy is extremely challenging due to the scarcity of resources and tools for neurophysiological and imaging diagnosis; in this respect, the study by Alvarado et al.1 was very helpful for clinical diagnosis. Neither of our patients received corticosteroids to inhibit the cytokine storm and manage neuralgia, as they both were intolerant to this drug group; instead, they received the first-line drugs ciclosporin A and pirfenidone, which, in addition to their anti-inflammatory properties, also have an antiviral effect against SARS-CoV-2. In conclusion, we present 2 familial cases of PTS associated with SARS-CoV-2 (beta variant, PANGO lineage B1.1.35, clade GH/501Y.V2) in whom the viral infection was successfully managed. Pirfenidone should be regarded as a new adjuvant drug for the treatment of PTS, as resolution of neuralgia may be linked to the reduction in viral load. Funding The authors have received no funding for this study. Conflicts of interest None. ==== Refs References 1 Alvarado Y. Lin Y.-M. Carrillo A. Parsonage-Turner syndrome post-infection by SARS-CoV-2: a case report Neurología 36 2021 548 576 10.1016/j.nrl.2021.04.008 34172405 2 Crespo J.B. Loriente M.C. García A.C. Mora P.F. Neuralgia amiotrófica secundaria a vacuna contra COVID-19 Vaxzevria (AstraZeneca) Neurología 36 2021 571 572 10.1016/j.nrl.2021.05.007 34330677 3 Jauregui-Larrañaga C. Ostolaza-Ibáñez C.A. Martín-Bujanda M. Mielitis transversa aguda asociada a infección por SARS-CoV-2 Neurologia 36 2021 572 574 10.1016/j.nrl.2021.05.008 34315679 4 Meuleman J. Kuhlenbaumer G. Schirmacher A. Wehnert M. De Jonghe P. De Vriendt E. Genetic refinement of the hereditary neuralgic amyotrophy (HNA) locus at chromosome 17q25 Eur J Hum Genet 7 1999 920 927 10.1038/sj.ejhg.5200384 10602368 5 Watts G.D.J. O’Briant K.C. Chance P.F. Evidence of a founder effect and refinement of the hereditary neuralgic amyotrophy (HNA) locus on 17q25 in American families Hum Genet 110 2002 166 172 10.1007/s00439-001-0647-5 11935323 6 Domínguez J.R. Ramírez S.A. Dávalos N.O. Cabrera C.E. Enantema vesicular palatofaríngeo, hallazgo temprano de COVID-19 Cir Cir 89 2021 1 2 10.24875/CIRU.21000279 7 Sanchez-Pernautea O. Romero-Buenoa F.I. Selva-O’Callaghanb A. Why choose cyclosporin A as first-line therapy in COVID-19 pneumonia Reumatol Clin 17 2021 554 557 34756320 8 Seifirad S. Pirfenidone: a novel hypothetical treatment for COVID-19 Med Hypotheses 144 2020 110005 10.1016/j.mehy.2020.110005	36462620	PMC9710106	NO-CC CODE	2022-12-05 23:15:28	no	Neurologia (Engl Ed). 2022 Nov 30; doi: 10.1016/j.nrleng.2022.01.007	utf-8	Neurologia (Engl Ed)	2,022	10.1016/j.nrleng.2022.01.007	oa_other
==== Front Int J Nurs Stud Adv Int J Nurs Stud Adv International Journal of Nursing Studies Advances 2666-142X Published by Elsevier Ltd. S2666-142X(22)00050-9 10.1016/j.ijnsa.2022.100111 100111 Article COVID-19 Related Negative Emotions and Emotional Suppression are Associated with Greater Risk Perceptions among Emergency Nurses: A Cross-Sectional Study Huff Nathan R. 1 Liu Guanyu 1 Chimowitz Hannah 1 Gleason Kelly 2 Isbell Linda M. 1⁎ 1 Psychological and Brain Sciences, University of Massachusetts Amherst, 135 Hicks Way, Amherst, Massachusetts, 01003 United States 2 School of Nursing, Johns Hopkins University, 525 N. Wolfe Street, Baltimore, Maryland, 21205 United States ⁎ Corresponding author: Linda M. Isbell. University of Massachusetts-Amherst, 135 Hicks Way, Amherst, Massachusetts, 01003, Phone: 413-335-7236. 30 11 2022 30 11 2022 10011123 5 2022 28 11 2022 29 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. Background As the COVID-19 pandemic began, frontline nurses experienced many emotions as they faced risks relevant to both patients (e.g., making errors resulting in patient harm) and themselves (e.g., becoming infected with COVID-19). Although emotions are often neglected in the patient safety literature, research in affective science suggests that emotions may significantly impact nurses’ perceptions of risk, which can have downstream consequences. Further, the use of chronic emotion regulation strategies that are known to differ in adaptability and effectiveness (i.e., emotional suppression, reappraisal) can impact risk perceptions. Objective To investigate the relationship between nurses’ emotional experiences in response to the pandemic and their estimates of how likely they would be to experience adverse outcomes related to both patients and themselves within the next six months. Additionally, we investigated the extent to which the use of suppression and reappraisal processes to manage emotions are associated with these risk perceptions. Design Cross-sectional survey. Setting Online survey distributed via email to emergency nurses at eight hospitals in the northeastern United States during fall 2020. Participants 132 emergency nurses (Mage = 37.05; 81.1% Female; 89.4% White). Methods Nurses reported the extent to which they experienced a variety of positive (e.g., hope, optimism) and negative (e.g., fear, sadness) emotions in response to the COVID-19 pandemic, and reported their perceptions of risk to both patients and themselves. Nurses also completed the Emotion Regulation Questionnaire, a measure of chronic tendencies to engage in emotional suppression and reappraisal. Immediately prior to providing data for this study, nurses completed an unrelated decision-making study. Results Nurses’ negative emotions in response to COVID-19 were associated with greater perceptions of both patient safety risks (b = 0.31, p < .001) and personal risks (b = 0.34, p < .001). The relationships between positive emotions and risk perceptions were not statistically significant (all p values > .66). Greater chronic tendencies to suppress emotions uniquely predicted greater perceptions of patient safety risks (b = 2.91, p = .036) and personal risks (b = 2.87, p = .040) among nurses; however, no statistically significant relationships with reappraisal emerged (all p values > .16). Conclusions Understanding factors that influence perceptions of risk are important, given that these perceptions can motivate behaviours that may adversely impact patient safety. Such an understanding is essential to inform the development of interventions to mitigate threats to patient safety that emerge from nurses’ negative emotional experiences and their use of different emotion regulation strategies. Tweetable abstract Covid-related negative emotions and emotional suppression are associated with greater patient and personal risk perceptions among emergency nurses @lindamisbell @Nathan_Huff_1 Keywords Attitudes COVID-19 Emergency Department Emotion Nurses Patient Safety Risk Judgements ==== Body pmc What is already known • Researchers in psychology have demonstrated that negative emotions and emotional suppression are related to greater perceptions of risk in a variety of domains. • Inflated risk perceptions have been linked to suboptimal judgements and behaviors in both non-clinical and clinical contexts. • Despite the significant emotional toll that the COVID-19 pandemic has had on nurses, there is a considerable gap in knowledge concerning the relationship between nurses’ negative emotional experiences and perceptions of risks to both patients and themselves, as well as how suppressing these emotions may adversely impact risk perceptions. What this paper adds • In this cross-sectional study, we found that emergency nurses’ negative emotions in response to COVID-19 were associated with increased perceptions of both patient safety and personal risks related to COVID-19. • Emotional suppression was also related to greater perceptions of both types of risk; however, no statistically significant relationships emerged for positive emotions and emotion reappraisal. • Given that risk perceptions are known to relate to objective risk markers, these findings suggest that addressing nurses’ negative emotions and maladaptive coping strategies may be key intervention targets to reduce safety risks to both patients and nurses, while also increasing nurse well-being. 1 Introduction 1.1 Background In the beginning of the COVID-19 pandemic, healthcare workers in the emergency department experienced many emotions while facing both known and unknown risks. Emotions were largely negative, like fear and sadness, but also positive, like hope and optimism. Risks that providers faced were relevant to their patients (e.g., making errors resulting in patient harm) as well as themselves (e.g., becoming infected with the virus) (Welsh et al., 2021). A well-established body of research in psychology has demonstrated that negative emotions increase risk perceptions in many domains, such as disease susceptibility, smoking cessation, and threats due to natural disaster and terrorism (Johnson and Tversky, 1983, Lerner and Keltner, 2001, Lerner et al., 2003). Furthermore, it has been shown that risk perceptions directly influence judgements, decisions, and behaviours (Ferrer and Klein, 2015), and a meta-analysis demonstrated that manipulating risk perceptions resulted in consistent behaviour changes across a variety of domains (Sheeran et al., 2014). While largely unexamined among nurses, this possibility has implications for clinical care and patient safety, as nurses’ risk perceptions could directly inform their decision to engage with, assess, and advocate for patients. In addition to negative emotions themselves, maladaptive emotion regulation strategies may also increase risk perceptions (Gross and John, 2003). One such strategy is suppression, which involves inhibiting emotional expression (Gross and John, 2003, Gross, 1998, Gross, 2015). A large body of research demonstrates that chronically suppressing one's emotions is associated with adverse influences on cognitive functioning and memory (Gross and John, 2003, Richards and Gross, 2006, Johns et al., 2008, Richards and Gross, 2000), as well as reduced well-being and relationship quality (Chervonsky and Hunt, 2017). Cognitive reappraisal, which involves re-evaluating one's original interpretation of a situation to influence what one feels, is considered a more adaptive strategy and is generally associated with positive outcomes (Hu et al., 2014). 1.2 Objectives Despite clearly established links among emotions, regulation strategies, and consequential risk perceptions, as well as recent work highlighting the urgent need to investigate affective factors (Wyer et al., 1999, Heyhoe et al., 2016, Liu et al., 2022, Isbell et al., 2020, Isbell et al., 2020) that can influence clinical information processing and decision making in the emergency department (Kozlowski et al., 2017, Djulbegovic and Elqayam, 2017, Croskerry, 2015), the role of emotions is often neglected in patient safety research (Croskerry et al., 2008, Croskerry et al., 2010, Committee on Diagnostic Error in Health Care 2015). The current study focused on a key aspect of clinical information processing within an important emergency provider population: risk perceptions among nurses. Nurses play a critical role in emergency care, as they are often responsible for triaging patients, advocating for their care, and conveying critical information to other providers through patient handoffs. Indeed, a recent report found that emergency nurses substantially shape patient safety due to their close and sustained contact with patients, role as information communicators, and tendency to be the first medical provider to assess patients and estimate the severity of illness (Manojlovich et al., August 2022). Thus, nurses’ assessments of risk are particularly consequential. We investigated the relationship between emergency nurses’ emotions in response to COVID-19 and their perceptions of both patient safety and personal risks, as well as the effect of chronic emotion regulation strategies on these risk perceptions. 2 Methods 2.1 Study Participants and Recruitment Emergency nurses in the northeastern United States (US) were recruited via email during fall 2020 to complete a survey online, following an unrelated study investigating medical decision making. Study invitations were sent over email by hospital leaders at eight institutions between October and December 2020. At six hospitals, invitations were distributed by hospital representatives. At two hospitals, invitations were shared by other nurses. Eligible nurses who indicated interest were sent a personalized, one-time-use Qualtrics study link and were given two weeks to participate. Those who did not access the study were sent reminder emails, provided a new link, and offered an additional week. Nurses who did not complete the study after three weeks were again re-invited, sent another link if interested, and given one week to participate. Nurses were compensated 100 US Dollars for the hour required to complete both studies. We aimed to recruit a minimum of 132 nurses to adequately power regression analyses. 2.2 Measurement Tools 2.2.1 Emotion Regulation Questionnaire Nurses completed the Emotion Regulation Questionnaire (Gross and John, 2003), which assesses the use of emotional suppression and cognitive reappraisal strategies. The suppression subscale contains four items (e.g., “I keep my emotions to myself”); the reappraisal subscale contains six items (e.g., “When I want to feel less negative emotion, I change what I'm thinking about”). Responses were reported along a scale from strongly disagree (1) to strongly agree (7). The Emotion Regulation Questionnaire has good internal consistency, reliability, temporal stability, test-retest reliability, and good convergent and discriminant validity (Gross and John, 2003, Ioannidis and Siegling, 2015) and has previously been used with healthcare providers (Măirean, 2016, Kafetsios et al., 2016). 2.2.2 Emotions in Response to COVID-19 To assess emotional experience, nurses reported the extent to which they felt 12 emotions in response to COVID-19 along unnumbered sliding scales from not at all (0) to very much (100): afraid, angry, anxious, calm, exhausted, frustrated, hopeful, hopeless, optimistic, overwhelmed, sad, and stressed. Emotion items were generated from the Positive and Negative Affect Schedule – Expanded Form, a widely used emotion measurement tool in psychological research (Clark and Watson, 1994). Specific emotions were selected due to their relevance to the COVID-19 pandemic. Emotion items were presented in a random order. 2.2.3 Personal and Patient Safety Risk Perceptions To assess risk perceptions, we developed an 8-item risk perception scale related to COVID-19 (Table 1 ). Four items informed by prior work (Ho et al., 2005) were designed to capture personal risk perceptions (e.g., nurse will become infected), and four items developed for this research were designed to assess patient safety risk perceptions (e.g., nurse will fail to adequately assess a patient). Nurses reported how likely it is these outcomes would occur in the next six months using unnumbered sliding scales from not at all (0) to very likely (100). Risk judgement items were presented in a random order. The emotion and risk judgement measurement tools were counterbalanced: half of the participants completed the emotion items before the risk judgement items, while the other half completed the risk judgement items before the emotion items.Table 1 Patient Safety and Personal Risk Perception Items Table 1In the next 6 months, how likely is it that you will … Patient Safety Risks: …make a medical error that results in harm to a patient. …lack the skills you need to provide care to one or more of your patients. …fail to adequately assess a patient. …contribute to the misdiagnosis of a patient. Personal Risks: …become infected with COVID-19. …bring COVID-19 home and infect a family member or loved one. …die from COVID-19. …experience the death of a family member or loved one due to COVID-19. Note. Items were presented in a random order, on the same page. Responses were provided on unnumbered sliding scales ranging from 0 (Not at all likely) to 100 (Very likely). 2.2.4 Participant Demographic Questions Finally, participants responded to demographic questions. Free response items assessed participants’ age, years of nursing experience, years of emergency medicine experience, and highest degree achieved. Multiple choice questions with a free response option were used to capture participant race, ethnicity, and gender. Multiple choice questions with a free response option were also used to capture the type of work institution (Private, Government, Other), the academic affiliation of the work institution (Academic, Nonacademic, Other), and the emergency department role (Registered Nurse, Other) in which participants worked. 2.3 Research Ethics The study was approved by our institutional review board (IRB#196: 2016-3291). Participants reviewed an informed consent document and provided their consent prior to initiating the study. At the end of the study, participants were debriefed and made aware of study purposes. 2.4 Data Analysis First, exploratory factor analyses were used to aid in the construction of positive and negative emotion subscales, as well as personal and patient safety risk perception subscales. Second, we examined bivariate correlations to assess relationships between study variables. Third, we ran a series of linear regressions to test our hypothesized relationships between COVID-19-related emotions, chronic emotion regulation strategies, and risk perceptions. In Model 1, we entered demographic control variables and questionnaire order, then added the positive and negative COVID-19 emotion subscales (Model 2), and finally added chronic emotion regulation strategy subscales (Model 3). Data were analysed using SPSS v.25. 3 Results 3.1 Participant Demographics Our sample included 132 emergency department nurses (Meanage = 37.05, Standard Deviation [SD]age = 9.58), who were 89.4% white (n = 118), 81.1% female (n = 107), and reported an average of 8.54 years of emergency medicine experience (SD = 7.64). Over ninety-eight percent of participants (n = 130) indicated their emergency department role was “Registered Nurse”, and 1.5% (n = 2) selected “Other”. These two participants indicated their emergency department role was “Emergency Staff Nurse” and “Primary Patient Care and Resource”. Additional participant demographic information is presented in Table 2 .Table 2 Participant Demographics Table 2Years M SD Age 37.05 9.58 Nursing Experience 11.55 9.16 Emergency Medicine Experience 8.54 7.64 Gender n % Female 107 81.1 Male 24 18.2 Non-Binary 1 0.8 Race White 118 89.4 Black/African American 6 4.5 Asian 4 3.0 White and Black/African American 2 1.5 White and American Indian/Alaskan Native 1 0.8 Race Not Provided 1 0.8 Ethnicity Not Hispanic and/or Latino 125 94.7 Hispanic and/or Latino 7 5.3 Highest Degree Earned Bachelor's Degree 111 84.1 Associate's Degree 14 10.6 Graduate Level Degree 7 5.3 Academic Affiliation of Work Institution Academic 99 75.0 Nonacademic 29 22.0 Other 4 3.0 Type of Work Institution Private 92 69.7 Government 27 20.5 Other 13 9.8 Note. Descriptive statistics presented include mean (M), standard deviation (SD), number (n) and percentage (%). Due to rounding, percentages do not always add up to 100. 3.2 Scale Creation Because the Emotion Regulation Questionnaire is an established and widely-used measure (Gross and John, 2003) with well-documented psychometric properties (Ioannidis and Siegling, 2015), composites were created by averaging subscale items for the suppression (Cronbach's α = .82) and the reappraisal (Cronbach's α = .88) subscales. We subjected the 12 emotion items to an exploratory factor analysis to facilitate the creation of subscales emergent from the data. An exploratory factor analysis using principal axis factor extraction and promax rotation supported a two-factor structure. Specifically, following interpretation of Eigenvalues, scree plots, parallel analysis, and minimal average partial values, exactly two factors had Eigenvalues greater than one. These two factors accounted for 55% of the total variance and are supported by theory. As such, subscales were created by averaging negative emotions (afraid, anxious, hopeless, sad, angry, frustrated, stressed, overwhelmed, exhausted; Cronbach's α = .89), and positive emotions (hopeful, optimistic, calm; Cronbach's α = .81). A second exploratory factor analysis was conducted on the eight risk judgement items to verify that the hypothesized subscales of personal risk and patient safety risk were emergent from the data. Exploratory factor analysis using principal axis factor extraction and promax rotation supported a two-factor structure. Specifically, following interpretation of Eigenvalues, scree plots, parallel analysis, and minimal average partial values, exactly two factors had Eigenvalues greater than one. The two factors accounted for 50% of the total variance and were consistent with our hypothesized personal and patient safety risk perception subscales. As such, subscales were created by averaging personal risk (Cronbach's α = .75) and patient safety risk items (Cronbach's α = .80). 3.3 Bivariate Correlations Bivariate correlations among the positive and negative emotion subscales, the personal and patient safety risk judgement subscales, and the reappraisal and suppression subscales of the Emotion Regulation Questionnaire are reported in Table 3 .Table 3 Bivariate Correlations Among Study Variables Table 3Variable 1 2 3 4 5 6 1. Positive Emotions Toward COVID-19 — -.58⁎⁎ -.19* -.21* .21* -.03 2. Negative Emotions Toward COVID-19 — .34⁎⁎ .34⁎⁎ -.11 -.00 3. Personal Risk Judgements — .37⁎⁎ -.09 .13 4. Patient Safety Risk Judgements — -.14 .19* 5. Use of Emotional Reappraisal Strategies — .01 6. Use of Emotional Suppression Strategies — Note. ⁎ p < .05. ⁎⁎ p < .01. Positive Emotions Toward COVID-19 (3 items; Cronbach's α = .81). Negative Emotions Toward COVID-19 (9 items; Cronbach's α = .89). Personal Risk Judgements (4 items; Cronbach's α = .75). Patient Safety Risk Judgements (4 items; Cronbach's α = .80). Use of Emotional Reappraisal Strategies (6 items; Cronbach's α = .88). Use of Emotional Suppression Strategies (4 items; Cronbach's α = .82). 3.4 Regression Models Predicting Patient Safety and Personal Risk Perceptions Regression weights from all models predicting both patient safety risk perceptions and personal safety risk perceptions appear in Table 4 . Negative emotions predicted increased perceptions of both patient safety risk (b = 0.31, SE = 0.11, p < .001) and personal risk (b = 0.34, SE = 0.11, p < .001) when controlling for positive emotions, suppression, reappraisal, the order in which emotions and risk were assessed, and participants’ gender, race, and age (Models 3; Table 4). Also, nurses higher in chronic tendency to suppress emotions reported greater perceptions of patient safety risk (b = 2.91, SE = 1.37, p = .036) and personal risk (b = 2.87, SE = 1.39, p = .040) controlling for all predictors. No other effects were statistically significant, all p values > .16.Table 4 Negative Emotions and Emotional Suppression Predict Personal Risk and Patient Safety Risk Accounting for Control Variables Table 4 Patient Safety Risk Personal Safety Risk Variable Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Nurse Gender -0.70 (5.03) -5.22 (4.89) -1.93 (5.03) 4.23 (4.99) 0.39 (4.91) 3.60 (5.10) Nurse Race -4.97 (6.41) -5.90 (6.19) -5.98 (6.31) 2.50 (6.36) 2.68 (6.22) 3.81 (6.39) Nurse Age -0.05 (0.20) 0.07 (0.20) 0.08 (0.20) -0.16 (0.20) -0.08 (0.20) -0.09 (0.20) Order of Emotion and Risk Judgement Questionnaires 0.17 (3.88) 2.61 (3.72) 1.50 (3.68) -4.40 (3.85) -2.34 (3.74) -3.39 (3.73) Positive Emotions Toward COVID-19 -0.06 (0.10) -0.03 (0.10) 0.02 (0.10) 0.04 (0.10) Negative Emotions Toward COVID-19 0.33 (0.11)⁎⁎ 0.31 (0.11)⁎⁎ 0.35 (0.11)⁎⁎ 0.34 (0.11)⁎⁎ Emotion Regulation Strategy - Suppression 2.91 (1.37)* 2.87 (1.39)* Emotion Regulation Strategy - Reappraisal -2.48 (1.74) -0.97 (1.77) R2 .005 .123 .169 .024 .119 .152 F 0.16 2.88* 3.07⁎⁎ 0.75 2.77* 2.70⁎⁎ ∆ R2 .005 .118 .046 .024 .095 .033 ∆ F 0.16 8.26⁎⁎ 3.34* 0.75 6.66⁎⁎ 2.33 Note. Sample Size (N) = 132. ⁎ p < .05, ⁎⁎ p < .01. Unstandardized regression weights are presented with standard error in parenthesis. Nurse Gender: 0 = male, 1 = female. Nurse Race: 0 = non-white, 1 = white. Order of Emotion and Risk Judgement Questionnaires: 0 = Emotion items first, 1 = Risk judgement items first. Positive Emotions Toward COVID-19 (3 items; Cronbach's α = .81). Negative Emotions Toward COVID-19 (9 items; Cronbach's α = .89). Emotional Suppression (4 items; Cronbach's α = .82). Emotional Reappraisal (6 items; Cronbach's α = .88). Patient Safety Risk (4 items; Cronbach's α = .80). Personal Safety Risk (4 items; Cronbach's α = .75). Nursing experience was substituted for nurse age in Model 1, and results were identical in interpretation; as such, age was used in subsequent models. We also examined whether emotion regulation strategies moderated the relationship between negative emotions and risk judgements. We found no statistically significant evidence of moderation, all p values > .19 (see supplemental material Table S1). Finally, although factor analysis revealed a negative and positive emotion factor structure best fit our data, researchers have specifically linked fear to pessimistic risk judgements (Lerner and Keltner, 2001, Ferrer et al., 2015). Thus, we created a fear composite (afraid, anxious; Cronbach's α = .70) and re-ran all analyses with fear in place of negative emotions. Results were analogous to those reported. 4 Discussion Emergency nurses’ negative emotions related to COVID-19 and nurses’ use of emotional suppression, a maladaptive emotion regulation strategy, both predicted increased perceptions of patient safety and personal risks. These associations held when controlling for the effect of the other, as well as when controlling for positive emotions, tendencies to engage in emotion reappraisal, and several demographic variables. This finding suggests that not just negative emotions themselves, but also tendencies to suppress emotions are unique predictors of greater perceptions of patient safety and personal risks. Risk perceptions are consequential, as they are known to relate to behaviour in many domains (Ferrer and Klein, 2015) and likely influence clinical decisions, such as the extent to which a nurse engages with, assesses, and advocates for patients. Interestingly, positive emotions (e.g., hope, optimism) as well as emotion reappraisal (i.e., an emotion regulation strategy associated with positive outcomes (Hu et al., 2014)) were not statistically significant predictors of risk perceptions in the regression models. However, when examining bivariate correlations (Table 2), a statistically significant relationship emerged between positive emotions and both types of risk judgements, revealing that positive emotions were associated with reduced perceptions of risk. Emotion reappraisal was similarly associated with reduced risk perceptions, but this relationship was not statistically significant. This suggests that while positive emotions may be tenuously related to these consequential risk perceptions, negative emotions and tendencies to suppress emotions are more salient predictors of COVID-19 risk judgements. Considering the robust links among negative emotions, emotion suppression, and risk perceptions, efforts to reduce both the experience of negative emotions and the use of emotional suppression will likely reduce risk perceptions that can interfere with clinical decisions. This may be accomplished through interventions designed to enhance emotional intelligence or one's ability to regulate one's emotions and to be aware of and understand emotional states in oneself and others (Gross and John, 2003, Salovey and Mayer, 1990, Bulmer Smith et al., 2009). Because emotions are a natural part of emergency care (Isbell et al., 2020), emotional intelligence trainings encourage participants to recognize their and others’ emotions and communicate with their peers about emotional experiences. In medical contexts, emotional intelligence is associated with improved patient care (Nightingale et al., 2018), enhanced communication, and interpersonal sensitivity (Libbrecht et al., 2014, Bourgeon et al., 2016) and is indirectly related to reduced malpractice risk among physicians (Shouhed et al., 2019). In addition to improving patient care quality and safety, such interventions are likely to improve healthcare provider well-being, as prolonged experiences of negative emotion place nurses at risk for burnout and mental health conditions (Vahey et al., 2004, Rushton and Boston-Leary, 2022). Given that emergency physicians and nurses experience negative emotions in response to patient, hospital, and system factors (Isbell et al., 2020), among others (the pandemic) (Welsh et al., 2021), multipronged interventions that target negative emotions emerging from different sources will yield the best outcomes. Because we found that negative emotions were the more salient and robust predictor of these risk perceptions relative to positive emotions, interventions should specifically pursue the reduction of negative emotions, rather than the promotion of positive emotions. Importantly however, interventions that target the promotion of positive emotions may still be beneficial; we suggest that their benefit may largely be reaped via the reduction of negative emotion engendered by a promotion of positive emotion. 5 Limitations Our study has several limitations. First, we captured risk perceptions, not objective risk markers. Objective risk markers, such as exposure to COVID-19 (i.e., personal risk) and the making of medical errors (i.e., patient safety risk) would be difficult if not impossible to determine amongst frontline healthcare providers during the early stages of the pandemic. Nonetheless, research demonstrates that risk perceptions influence behaviour in a wide range of domains (Ferrer and Klein, 2015, Sheeran et al., 2014) and thus are likely to similarly influence clinical decision making and behaviour. For example, emergency physicians who estimated greater risk of adverse outcomes for chest pain patients were more likely to admit these patients, even when actual risk was low (Schriger et al., 2018). Even so, more research is needed to link specific risk assessments to clinical behaviors. Second, we found similar results regardless of whether we used negative affect or fear alone to predict risk perceptions, even though research suggests that fear uniquely predicts risk perceptions (Johnson and Tversky, 1983, Lerner and Keltner, 2001, Lerner et al., 2003, Ferrer et al., 2015). While our emotion items were selected from the Positive and Negative Affect Schedule – Expanded Form, which is a well-established tool (Clark and Watson, 1994), this finding may reflect the inability of our COVID-19 emotion measure to detect meaningfully distinct emotion categories. However, it may also reflect the possibility that nurses themselves did not distinguish between fear and anger, for example, with respect to COVID-19 during the start of the pandemic, and instead experienced many overlapping negative emotions. Third, the extent to which our results would generalize to emergency nurses outside the northeastern US is unknown, as our sample is limited to a specific geographic region. Finally, our sample was largely female, reflecting the demographics of this profession in the US and was also disproportionately white. The extent to which our findings would generalize to more diverse participant samples is unknown. 6 Conclusions Actual, objective risk indicators are shaped by how nurses perceive the likelihood of these risks unfolding, and we demonstrated that such perceptions were influenced by nurses’ negative emotions. In this sense, emotions serve as a nexus from which threats to patient safety and medical errors likely emerge. Addressing nurses’ emotions and coping is crucial to explore, as emergency nurses serve a vital role in preserving continuity of care, and negative emotions may be particularly susceptible to targeted interventions that improve clinical care and reasoning. Funding sources National Institute of Health (NIH); Agency for Healthcare Research and Quality (AHRQ); Project # (MASKED) Role of the Funder/Sponsor The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation and processing of the manuscript; and decision to submit the manuscript for publication. Declaration of competing interest None. Appendix Supplementary materials Image, application 1 Image, application 2 Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ijnsa.2022.100111. ==== Refs References Welsh M Chimowitz H Nanavati JD A qualitative investigation of the impact of coronavirus disease 2019 (COVID-19) on emergency physicians’ emotional experiences and coping strategies J Am Coll Emerg Physicians Open 2 2021 10.1002/emp2.12578 Johnson EJ Tversky A. Affect, generalization, and the perception of risk J Pers Soc Psychol 45 1983 20 31 10.1037/0022-3514.45.1.20 Lerner JS Keltner D. Fear, anger, and risk J Pers Soc Psychol 81 2001 146 159 10.1037/0022-3514.81.1.146 11474720 Lerner JS Gonzalez RM Small DA Effects of Fear and Anger on Perceived Risks of Terrorism: A National Field Experiment Psychol Sci 14 2003 144 150 10.1111/1467-9280.01433 12661676 Ferrer RA Klein WM. Risk perceptions and health behavior Curr Opin Psychol 5 2015 85 89 10.1016/j.copsyc.2015.03.012 26258160 Sheeran P Harris PR Epton T. Does heightening risk appraisals change people's intentions and behavior? A meta-analysis of experimental studies Psychol Bull 140 2014 511 543 10.1037/a0033065 23731175 Gross JJ John OP. Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being J Pers Soc Psychol 85 2003 348 362 10.1037/0022-3514.85.2.348 12916575 Gross JJ. The Emerging Field of Emotion Regulation: An Integrative Review Rev Gen Psychol 2 1998 271 299 10.1037/1089-2680.2.3.271 Gross JJ. Emotion Regulation: Current Status and Future Prospects Psychol Inq 26 2015 1 26 10.1080/1047840X.2014.940781 Richards JM Gross JJ. Personality and emotional memory: How regulating emotion impairs memory for emotional events J Res Pers 40 2006 631 651 10.1016/j.jrp.2005.07.002 Johns M Inzlicht M Schmader T. Stereotype threat and executive resource depletion: Examining the influence of emotion regulation J Exp Psychol Gen 137 2008 691 705 10.1037/a0013834 18999361 Richards JM Gross JJ. Emotion regulation and memory: The cognitive costs of keeping one's cool J Pers Soc Psychol 79 2000 410 424 10.1037/0022-3514.79.3.410 10981843 Chervonsky E Hunt C. Suppression and expression of emotion in social and interpersonal outcomes: A meta-analysis Emotion 17 2017 669 683 10.1037/emo0000270 28080085 Hu T Zhang D Wang J Relation between Emotion Regulation and Mental Health: A Meta-Analysis Review Psychol Rep 114 2014 341 362 10.2466/03.20.PR0.114k22w4 24897894 Wyer RS Clore GL Isbell LM. Affect and Information Processing Advances in Experimental Social Psychology 1999 Elsevier 1 77 10.1016/S0065-2601(08)60271-3 Heyhoe J Birks Y Harrison R The role of emotion in patient safety: Are we brave enough to scratch beneath the surface? J R Soc Med 109 2016 52 58 10.1177/0141076815620614 26682568 Liu G Chimowitz H Isbell LM. Affective influences on clinical reasoning and diagnosis: insights from social psychology and new research opportunities Diagnosis 2022 0 10.1515/dx-2021-0115 Isbell LM Tager J Beals K Emotionally evocative patients in the emergency department: a mixed methods investigation of providers’ reported emotions and implications for patient safety BMJ Qual Saf 29 1 2020 10.1136/bmjqs-2019-010110 3-2 Isbell LM Boudreaux ED Chimowitz H What do emergency department physicians and nurses feel? A qualitative study of emotions, triggers, regulation strategies, and effects on patient care BMJ Qual Saf 29 1 2020 10.1136/bmjqs-2019-010179 5-2 Kozlowski D Hutchinson M Hurley J The role of emotion in clinical decision making: an integrative literature review BMC Med Educ 17 2017 255 10.1186/s12909-017-1089-7 29246213 Djulbegovic B Elqayam S. Many faces of rationality: Implications of the great rationality debate for clinical decision-making J Eval Clin Pract 23 2017 915 922 10.1111/jep.12788 28730671 Croskerry P. Clinical decision making Barach PR Jacobs JP Lipshultz SE Pediatric and Congenital Cardiac Care: Volume 2: Quality Improvement and Patient Safety 2015 Springer London London 397 409 Croskerry P Abbass AA Wu AW. How doctors feel: affective issues in patients’ safety Lancet 372 2008 1205 1206 10.1016/S0140-6736(08)61500-7 19094942 Croskerry P Abbass A Wu AW. Emotional Influences in Patient Safety J Patient Saf 6 2010 199 205 10.1097/PTS.0b013e3181f6c01a 21500605 Committee on Diagnostic Error in Health Care Board on Health Care Services, Institute of Medicine, et al. Improving Diagnosis in Health Care 2015 National Academies Press Washington, D.C. 10.17226/21794 Manojlovich M Krein SL Kronick SL Distributed cognition and the role of nurses in diagnostic safety in the emergency department August 2022 Agency for Healthcare Research and Quality Rockville, MD AHRQ Publication No. 22-0026-2-EF Ioannidis CA Siegling AB. Criterion and incremental validity of the emotion regulation questionnaire Front Psychol 6 2015 10.3389/fpsyg.2015.00247 Măirean C. Emotion Regulation Strategies, Secondary Traumatic Stress, and Compassion Satisfaction in Healthcare Providers J Psychol 150 2016 961 975 10.1080/00223980.2016.1225659 27629057 Kafetsios K Hantzara K Anagnostopoulos F Doctors’ Attachment Orientations, Emotion Regulation Strategies, and Patient Satisfaction: A Multilevel Analysis Health Commun 31 2016 772 777 10.1080/10410236.2014.993497 26529518 Clark LA, Watson D. The PANAS-X: Manual for the Positive and Negative Affect Schedule - Expanded Form. 1994. doi:10.17077/48vt-m4t2 Ho SMY Kwong-Lo RSY Mak CWY Fear of Severe Acute Respiratory Syndrome (SARS) Among Health Care Workers J Consult Clin Psychol 73 2005 344 349 10.1037/0022-006X.73.2.344 15796643 Ferrer R Klein W Lerner J Chapter 4: Emotions and health decision making: extending the appraisal tendency framework to improve health and healthcare Roberto CA Kawachi I Behavioral Economics and Public Health 2015 Oxford University Press Oxford 101 132 Salovey P Mayer JD. Emotional Intelligence Imagin Cogn Pers 9 1990 185 211 10.2190/DUGG-P24E-52WK-6CDG Bulmer Smith K Profetto-McGrath J Cummings GG. Emotional intelligence and nursing: An integrative literature review Int J Nurs Stud 46 2009 1624 1636 10.1016/j.ijnurstu.2009.05.024 19596323 Nightingale S Spiby H Sheen K The impact of emotional intelligence in health care professionals on caring behaviour towards patients in clinical and long-term care settings: Findings from an integrative review Int J Nurs Stud 80 2018 106 117 10.1016/j.ijnurstu.2018.01.006 29407344 Libbrecht N Lievens F Carette B Emotional intelligence predicts success in medical school Emotion 14 2014 64 73 10.1037/a0034392 24219393 Bourgeon L Bensalah M Vacher A Role of emotional competence in residents’ simulated emergency care performance: a mixed-methods study BMJ Qual Saf 25 2016 364 371 10.1136/bmjqs-2015-004032 Shouhed D Beni C Manguso N Association of Emotional Intelligence With Malpractice Claims: A Review JAMA Surg 154 2019 250 10.1001/jamasurg.2018.5065 30698614 Vahey DC Aiken LH Sloane DM Nurse Burnout and Patient Satisfaction Med Care 42 2004 10.1097/01.mlr.0000109126.50398.5a II–57 Rushton CH Boston-Leary K. Nurses suffering in silence: Addressing the stigma of mental health in nursing and healthcare Nurs Manag (Harrow) 53 2022 7 11 10.1097/01.NUMA.0000853148.17873.77 Schriger DL Menchine M Wiechmann W Emergency Physician Risk Estimates and Admission Decisions for Chest Pain: A Web-Based Scenario Study Ann Emerg Med 72 2018 511 522 10.1016/j.annemergmed.2018.03.003 29685372	36467310	PMC9710107	NO-CC CODE	2022-12-11 23:16:20	no	Int J Nurs Stud Adv. 2023 Dec 30; 5:100111	utf-8	Int J Nurs Stud Adv	2,022	10.1016/j.ijnsa.2022.100111	oa_other
==== Front Travel Med Infect Dis Travel Med Infect Dis Travel Medicine and Infectious Disease 1477-8939 1873-0442 Published by Elsevier Ltd. S1477-8939(22)00260-5 10.1016/j.tmaid.2022.102514 102514 Article Post-vaccination seropositivity against SARS-CoV-2 in peruvian health workers vaccinated with BBIBP-CorV (Sinopharm) Cvetkovic-Vega Aleksandar a Urrunaga-Pastor Diego ab Soto-Becerra Percy a Figueroa Morales Luis Edgardo c Fernández-Bolivar Lizzete d Alvizuri-Pastor Sergio e Oyanguren-Miranda Martin f Vera Ibeth Melania Neyra d Carrillo Ramos Elizabeth Emilia g Sagástegui Arturo Ampelio g Contreras Macazana Roxana Milagros h Rengifo Diana Elizabeth Lecca i Castro Nikolai Grande j Apolaya-Segura Moises a Maguina Jorge L. a∗ a Instituto de Evaluación de Tecnologías Sanitarias e Investigación - IETSI, EsSalud, Lima, Peru b Facultad de Ciencias de la Salud, Universidad Científica del Sur, Lima, Peru c Servicio de Patología Clínica, Hospital III Suárez Angamos, EsSalud, Lima, Peru d Departamento de Patología Clínica, Servicio de Inmunología y Bioquímica, Hospital Nacional Edgardo Rebagliati Martins, EsSalud, Lima, Peru e Servicio de Inmunología, Hospital Nacional Guillermo Almenara Irigoyen, EsSalud, Lima, Peru f Unidad de Cuidados Intensivos, Hospital Nacional Edgardo Rebagliati Martins, EsSalud, Lima, Peru g Departamento de Patología Clínica, Hospital Nacional Edgardo Rebagliati Martins, EsSalud, Lima, Peru h Departamento de Patología Clínica, Hospital Nacional Alberto Sabogal Sologuren, EsSalud, Lima, Peru i Subgerencia de Proyectos Especiales, Gerencia de Oferta Flexible, EsSalud, Lima, Peru j Departamento de Patología Clínica, Unidad de Inmuno-diagnóstico, Hospital Nacional Guillermo Almenara Irigoyen, EsSalud, Lima, Peru ∗ Corresponding author. Jirón Domingo Cueto 109, Jesús María, Zip code: 15072, Peru. 30 11 2022 30 11 2022 10251410 5 2022 23 9 2022 29 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. Objective To estimate the prevalence of post-vaccination seropositivity against SARS-CoV-2 and identify its predictors in Peruvian Social Health Insurance (EsSalud) personnel in 2021. Methods We conducted a cross-sectional study in a representative simple stratified sample of EsSalud workers. We evaluated IgG anti-SARS-CoV-2 antibodies response (seropositivity) by passive (previous infection) and active immunization (vaccination), and epidemiological and occupational variables obtained by direct interview and a data collection form. Descriptive and inferential statistics were used with correction of sample weights adjusted for non-response rate, and crude and adjusted odds ratio (OR) and geometric mean ratio (GMR) with their respective 95% confidence intervals (95%CI) were estimated. Results We enrolled 1077 subjects. Seropositivity was 67.4% (95%CI: 63.4–71.1). Predictors of seropositivity were age (negative relation; p < 0.001), previous infection (aOR = 11.7; 95%CI: 7.81–17.5), working in COVID-19 area (aOR = 1.47; 95%CI: 1.02–2.11) and time since the second dose. In relation to antibody levels measured by geometric means, there was an association between male sex (aGMR = 0.77; 95%CI: 0.74–0.80), age (negative relation; p < 0.001), previous infection (aGMR = 13.1; 95%CI:4.99–34.40), non-face-to-face/licensed work modality (aGMR = 0.78; 95%CI: 0.73–0.84), being a nursing technician (aGMR = 1.30; 95%CI: 1.20–1.41), working in administrative areas (aGMR = 1.17; 95%CI: 1.10–1.25), diagnostic support (aGMR = 1.07; 95%CI: 1.01–1.15), critical care (aGMR = 0.85; 95%CI: 0.79–0.93), and in a COVID-19 area (aGMR = 1.30; 95%CI: 1.24–1.36) and time since receiving the second dose (negative relation; p < 0.001). Conclusions Seropositivity and antibody levels decrease as the time since receiving the second dose increases. Older age and no history of previous infection were associated with lower seropositivity and antibody values. These findings may be useful for sentinel antibody surveillance and the design of booster dose strategies. Keywords COVID-19 Vaccine Serology Antibodies Doses (source: MeSH) ==== Body pmc1 Introduction The pandemic caused by the SARS-CoV-2 virus has reported figures of up to 458 million cases and 6 million deaths worldwide up to March 13, 2022 [1]. To mitigate the contagion and spread of the disease, many countries have responded with the development of vaccines. One of the vaccines designed was BBIBP-CorV, developed by the Beijing Bio-Institute of Biological Products of the Chinese national BIOTEC group and known internationally as Sinopharm [2]. This vaccine was produced from the HB02 strain and consists of viral particles cultured in the laboratory and inactivated to lose their ability to produce disease while stimulating the host immune response [3]. Despite being one of the first countries to initiate mandatory social immobilization to reduce the spread of COVID-19, Peru has registered more than 2 million cases and has one of the highest mortality rates in the world at 9.3% [4]. These figures can be explained considering labor informality, agglomeration, precariousness of the health system and intradomiciliary overcrowding which prevail in Peruvian society [5,6]. Nonetheless, in order to address the further spread of COVID-19, one of the fundamental pillars implemented by the Peruvian government was the acquisition and administration of vaccines to immunize the population, starting with high-risk target groups such as health personnel [7]. This first group was inoculated with the BBIBP-CorV vaccine, requiring the application of 2 doses with a 21-day interval between doses [8]. Although efforts have been focused on maximizing vaccine uptake and coverage, the question of passive immunity conferred arose taking into consideration studies showing that active immunization did not necessarily lead to the generation of antibodies [9] and/or in which a drop in these antibodies was described months after completing the vaccination schedule [10]. In addition, we can introduce terms related to active immunity, such as infection-induced immunity (defined as immune protection in an unvaccinated individual after an episode of SARS-CoV-2), vaccine-induced immunity (immune protection in someone who has not previously been infected with SARS -CoV-2 but have received at least one dose of vaccine) and hydrid immunity (occurs in people who suffered at least one episode of COVID-19 and have received at least one dose of vaccine) [11]. This aspect is currently the subject of studies worldwide, focusing their attention on neutralizing antibodies as a strategy for monitoring the individual's immune response to infection and vaccination [12]. Studies with the BNT162b2 vaccine reported an exponential increase in neutralizing antibodies on days 11 and 21 after vaccination [13]. However, studies on vaccines with inactivated virus technologies, such as BBIBP-CorV and CoronaVac, are still scarce. A study in Chile, in health care workers who completed the 2 doses in 0–14 day schedules, reported activation of interferon gamma secreting T cells and favorable antibody levels at 14, 28 and 42 days after immunization [14]. In Peru, it was possible to vaccinate health personnel at the beginning of the second wave of COVID-19 (February to April 2021), despite various controversies regarding the efficacy and effectiveness of the BBIBP-CorV vaccine, the social and scientific scandal of “vacunagate” [15] and the deep-rooted infodemics and misinformation surrounding COVID-19 [16]. A study in health personnel vaccinated with 2 doses of BBIBP-CorV reported an effectiveness of 50.4% in preventing infection and 94% in preventing mortality due to SARS-CoV-2 [17], while a study on the CoronaVac vaccine reported an effectiveness in preventing infection of 65.9%, hospitalization of 87.5% and mortality of 86.3% [18]. These effectiveness were lower compared to efficacy reported by the World Health Organization (WHO) Phase III report for the same vaccine in relation to preventing symptomatic infection [19]. It has been shown that the BBIBP-CorV vaccine provides protection against severe forms of COVID-19 that can lead to hospitalization and death. However, to date evidence regarding the prevention of symptomatic infection is questionable. Even within the current context of the circulation of different SARS-CoV-2 variants (Omicron, Delta, and Lambda) and heterologous vaccination schemes with booster doses (BBIBP-CorV + BNT162b2, BBIBP-CorV + ChAdOx1, BNT162b2 + ChAdOx1), there is no evidence of the generation of immune response by the BBIBP-CorV vaccine and even less by the heterologous vaccination schemes. This situation highlights the importance of immunological monitoring of antibody seropositivity in vaccinees to identify specific groups of low seropositivity, as well as temporal trends of the antibodies generated. Therefore, the objective of this study was to estimate post-vaccination seropositivity against COVID-19 in Peruvian Social Health Insurance (EsSalud) personnel vaccinated with two doses of BBIBP-CorV in Lima, Peru, 2021. 2 Methods 2.1 Study design and population We conducted a cross-sectional study in a representative sample of health workers from five secondary and tertiary level hospitals of the Peruvian Social Health Insurance (EsSalud). The hospitals were Hospital Nacional Edgardo Rebagliati Martins, Hospital Nacional Guillermo Almenara Irigoyen, Hospital Nacional Alberto Sabogal Sologuren and Villas Panamericana and Mongrut. EsSalud health workers in whom there was an interval of at least 14 days since the first vaccination with the BBIBP-CorV vaccine, and who provided consent to participate in the study were enrolled. Those with contraindications for venous blood collection, active symptoms suggestive of COVID-19 and any condition related to hospitalization or quarantine hospitalization were excluded. 2.2 Sample Based on the sampling frame defined by the list of vaccinated workers of the Health Care Centers (N = 2539), a probability, uni-stage, stratified sampling was performed with independent and representative strata corresponding to the domains represented by the occupational groups (physicians, nurses, nursing technicians, others and administrative personnel). A sample size per domain was calculated considering a nonresponse rate of 20% and a precision of 9% for each of the 24 area-occupancy strata (6 areas for each occupational group), and an estimated prevalence of seropositivity based on a previous study of post-vaccination IgG antiprotein S antibody production of 79.5% [20]. The sample size calculated was a total of 1436 participants. The sample weights were adjusted for nonresponse by the propensity score-matched class method [21]. The present analysis was restricted to health personnel who received two doses of the BBIBP-CorV vaccine in Peru as part of the vaccination campaigns promoted by the Peruvian government. 2.3 Participant recruitment, data and blood sample collection All the participants were invited to participate by telephone and agreed on a specific date to come to the enrollment site to sign the informed consent form and to provide a blood sample. At recruitment, the participants were given a data collection form prepared by the research team to collect information on the variables of interest. All doubts or questions the participants had, were answered by a team of professionals assigned for this purpose. In addition, 5 cc of venous blood were drawn from each participant, to which EDTA was added and the sample was transported and stored in the laboratory for processing. A cold chain was maintained at all times to ensure sample stability. In personnel who confirmed participation but were unable to do so for reasons of distance and workload, an additional period of sample collection from May to July 2021 was developed in 3 hospital sites to facilitate sample collection in these participants. 2.4 Laboratory methods Blood samples were processed at the Clinical Pathology Service Laboratory of the Hospital Nivel II Suárez Angamos following standardized protocols and the manufacturer's recommendations. IgG anti-SARS-CoV-2 antibodies were measured using the LIAISON® SARS-CoV-2 TrimericS IgG test (DiaSorin Inc., Stillwater, USA), Stillwater, USA). This chemiluminescence immunoassay has a positive and negative concordance greater than 96% with the microneutralization plate test and has proven to be an excellent substitute for the Plate Reduction Neutralization Test - PRNT (gold standard) [22]. Likewise, the equipment complied with the verification method recommended by the National Institute of Quality - INACAL [23] and the Clinical & Laboratory Standards Institute - CLSI, under the EP06-A, EP12-A2 and EP15A3 evaluation protocols [[24], [25], [26]]. 3 Variables 3.1 Outcomes and covariates A participant was defined as seropositive with antibody levels greater than or equal to 33.8 BAU/ml, which is the cut-off value recommended by the WHO harmonization process [27]. Antibody levels were also analyzed as a quantitative variable after transformation as a logarithm. Variables related to sex, age, occupation, work area, work modality (referring to the participants main work and classified as non-attendance, face-to-face or mixed), work in COVID-19 area, comorbidities, full dose of BBIBP-CorV vaccine, and time from the first/second dose to sampling were measured. In addition, we collected history of previous SARS-CoV-2 infection through self-report. 3.2 Statistical analysis All analyses were performed with the R statistical program [28]. The data were entered into the REDCap® capture platform [29] and were subjected to a quality control process that checked for missing, extreme and/or inconsistent values. Missing data were completed by simple multivariate imputation processing by random Forest [30]. Numerical variables were described as means (standard deviation [SD]) or medians (25th and 75th percentiles), as appropriate. Categorical variables were described as absolute and relative frequencies. Bivariate analyses were performed using the Wald or Mann-Whitney U test (both adjusted for the sample design) to compare numerical variables between groups; and the Chi-2 test with Rao-Scott second-order correction for association of categorical variables. Prevalences of seropositivity were reported together with 95% confidence intervals (95%CI) obtained by the logit method. The association between the probability of seropositivity with predictors of interest was assessed using a logistic generalized additive model (GAM). Odds ratios (OR) were estimated with their respective 95%CI. On the other hand, the relationship between antibody level and predictors of interest was evaluated by tobit GAM, with identity link function and censoring on both sides corresponding to the lower (3.81 BAU/mL) and upper (2080 BAU/mL) limit of the test. Considering evidence that the response to vaccination would vary differentially according to age, time since vaccination, and the existence of previous infection [31], we constructed models that evaluated the interaction between these three variables. The interactions were evaluated by specifying a tensor product of B-splines that allowed modeling the nonlinearity among these variables. To reduce the risk of overfitting we performed a smoothing penalty by restricted maximum likelihood. We assessed collinearity using generalized variance inflation factor and concurvity, a generalization of collinearity that can make estimates unstable, as previously described [32]. We selected variables a priori based on an epidemiological approach, considering previous studies [13,20]. 3.3 Ethical issues The study was approved by the Institutional Research Ethics Committee of the National Heart Institute (INCOR) (12/2021-CEI). Participants provided informed consent prior to enrollment in the study. The information was anonymized and coded to avoid any subsequent identification of the participant. 4 Results 4.1 General characteristics according to seropositivity A total of 1077 subjects were enrolled. Seventeen participants were excluded for only having received one dose and 18 because they did not have the variables of interest. The prevalence of seropositivity was 67.4% (95%CI: 63.4–71.1) with a coefficient of variation of 2.9%. Among the main characteristics of the sample, we found that 66.7% (n = 786) were female, the median age was 44.9 years (IQR: 35.0–55.0), 69.6% (n = 720) had no comorbidities, 58.5% (n = 651) had no previous SARS-CoV-2 infection, 82.1% (n = 866) worked in an office and 50.9% (n = 527) worked in an area with COVID-19 patients. In addition, the median time since receipt of the second dose was 130 days (interquartile range: 124 to 134). There were no statistically significant differences between groups according to seropositivity and sex, reported comorbidities, area of work, work setting, number of BBIBP-CorV doses received, additional vaccination abroad, and time in days since receiving the first and second doses (Table 1 ).Table 1 Characteristics of the study sample including ineligible individuals. Table 1Characteristics Total Seropositivity Missing data n = 1077 Negative Positive P valuea (n = 378) (n = 687) n (%) n (%) Sexrowhead 0 0.200 Female 786 (66.7%) 268 (30.2%) 514 (69.8%) Male 291 (33.3%) 110 (35.5%) 173 (64.5%) Age (years) 0 <0.001 Mean (SD) 45.1 (11.8) 47.6 (11.7) 43.9 (11.7) Median (p25-p75) 0 44.9 (35.0–55.0) 49.0 (36.0–57.7) 42.0 (35.0–52.0) Range (minimum-maximum) 24.0–70.0 25.0–69.0 24.0–70.0 Age 0 0.004 18 to 44 500 (49.8%) 150 (25.7%) 345 (74.3%) 45 to 59 395 (34.1%) 152 (37.0%) 237 (63.0%) 60 or more 182 (16.1%) 76 (40.5%) 105 (59.5%) Nationality 0 0.018 Peruvian 1069 (99.6%) 377 (32.0%) 680 (68.0%) Foreign 8 (0.4%) 1 (4.7%) 7 (95.3%) Comorbidities 10 0.300 None 720 (69.6%) 250 (30.5%) 465 (69.5%) One 266 (24.2%) 92 (35.0%) 169 (65.0%) Two or more 81 (6.2%) 34 (40.8%) 45 (59.2%) Previous SARS-CoV-2 infection 4 <0.001 No 651 (58.5%) 344 (49.7%) 297 (50.3%) Yes 422 (41.5%) 34 (7.5%) 386 (92.5%) Profession 0 <0.001 Physician 289 (16.0%) 130 (45.4%) 151 (54.6%) Administrative or other 283 (45.3%) 103 (32.3%) 177 (67.7%) Nurse 284 (25.9%) 100 (31.7%) 183 (68.3%) Nursing technician 221 (12.9%) 45 (14.7%) 176 (85.3%) Work modality 1 0.002 Non-attendance 137 (11.9%) 65 (46.1%) 71 (53.9%) Face-to-face 866 (82.1%) 278 (29.0%) 577 (71.0%) Mixed 38 (3.2%) 22 (55.8%) 16 (44.2%) Licensed 35 (2.8%) 12 (29.0%) 23 (71.0%) Work area 0 0.076 Hospitalization/Surgery 226 (46.9%) 67 (29.2%) 155 (70.8%) Administrative or other related 151 (11.3%) 56 (32.8%) 95 (67.2%) Diagnostic support and other related 145 (12.3%) 54 (37.9%) 88 (62.1%) Outpatient, extramural and other related 191 (8.3%) 81 (45.2%) 107 (54.8%) Critical care 185 (6.2%) 69 (35.7%) 116 (64.3%) Emergency or urgent care 179 (15.0%) 51 (26.1%) 126 (73.9%) Main work area 66 0.034 ICU 167 (8.8%) 54 (26.4%) 112 (73.6%) Emergency 174 (16.9%) 43 (22.5%) 127 (77.5%) Hospitalization 212 (28.7%) 59 (27.6%) 153 (72.4%) Non-COVID-19 Clinic 89 (5.1%) 32 (41.5%) 55 (58.5%) Home care 10 (2.6%) 5 (36.3%) 5 (63.7%) Administrative care 61 (7.6%) 26 (37.3%) 34 (62.7%) Research 1 (0.0%) 1 (100.0%) 0 (0.0%) Remote work 86 (7.7%) 45 (52.3%) 40 (47.7%) Other 211 (22.6%) 85 (34.6%) 124 (65.4%) Works in COVID-19 area 40 0.002 No 527 (50.9%) 205 (36.8%) 317 (63.2%) Yes 510 (49.1%) 154 (24.5%) 349 (75.5%) EsSalud hospital of work 0 0.600 I Octavio Mongrut Muñoz Hospital 52 (4.5%) 18 (40.7%) 34 (59.3%) Alberto Sabogal Sologuren National Hospital 341 (23.3%) 128 (33.4%) 212 (66.6%) Edgardo Rebagliati Martins National Hospital 465 (42.3%) 156 (31.3%) 308 (68.7%) Guillermo Almenara Irigoyen National Hospital 198 (24.3%) 70 (32.5%) 118 (67.5%) Villa Panamericana 21 (5.5%) 6 (20.5%) 15 (79.5%) BBIBP-CorV doses 0 0.070 One dose 17 (2.4%) 1 (7.9%) 16 (92.1%) Two doses 1060 (97.6%) 377 (32.5%) 671 (67.5%) Additional vaccination abroad 0 0.150 No 1071 (99.6%) 378 (32.0%) 681 (68.0%) Yes 6 (0.4%) 0 (0.0%) 6 (100.0%) Time since first dose (days) 3 0.130 Median (p25-p75) 152.0 (145.0–155.0) 153.0 (145.0–155.0) 152.0 (145.0–155.0) Range (minimum-maximum) 13.0–163.0 95.0–163.0 13.0–163.0 Time since second dose (days) 19 0.033 Median (p25-p75) 130.0 (124.0–134.0) 131.0 (124.0–134.0) 130.0 (123.0–134.0) Range (minimum-maximum) 14.0–142.0 14.0–142.0 14.0–142.0 n: unweighted absolute frequency; %: weighted percentage; ICU: intensive care unit; SD: standard deviation. Ineligible individuals were those who had only one dose of vaccine, were vaccinated with a vaccine other than BBIBP-CorV and/or were vaccinated abroad. Likewise. we also excluded those who did not have complete data on the response variable. a Squared chi-square test with Rao and Scott second-order correction; Wilcoxon rank sum test for complex samples. 4.2 Seropositivity predictors In the adjusted analysis, we found a negative association between seropositivity and age (p < 0.001). In addition, belonging to the occupational group of nursing technicians (aOR = 2.24; 95%CI: 1.14–4.37), belonging to the Villa Mongrut or Panamericana group (aOR = 0.27; 95%CI: 0.14–0.53) and Sabogal Hospital (aOR = 0.60; 95%CI: 0.40–0.94), time since second vaccination (p < 0.001), previous SARS-CoV-2 infection (aOR = 11.7; 95%CI: 7.81–17.5) and working in a COVID-19 area (aOR = 1.47; 95%CI: 1.02–2.11) (Table 2 ) were associated with presenting seropositivity (see Table 3).Table 2 Predictors of seropositivity 14 to 142 days after receiving the second dose in health personnel vaccinated with BBIBP-CorV. Table 2Characteristics 14–142 days post second dose Crude analysis Adjusted analysis Negative Positive P valuea cORb 95%CIc P value aORd 95%CIc P value (n = 377) (n = 665) n (%) n (%) Sex 0.200 Female 267 (31.0%) 495 (69.0%) – – – – Male 110 (36.0%) 170 (64.0%) 0.80 0.61–1.04 0.091 0.82 0.58–1.17 0.300 Age (years) 0.001 <0.001 <0.001 Mean (SD) 47.5 (11.6) 44.0 (11.6) Median (p25-p75) 49.0 (36.0–57.3) 43.0 (35.0–52.0) Range (minimum-maximum) 25.0–69.0 24.0–70.0 Comorbidities 0.200 None 252 (31.1%) 460 (68.9%) – – – – One or more 125 (36.4%) 205 (63.6%) 0.79 0.60–1.03 0.086 0.96 0.68–1.35 0.800 Previous SARS-CoV-2 infection <0.001 No 343 (49.9%) 292 (50.1%) – – – – Yes 34 (7.8%) 373 (92.2%) 11.7 8.00–17.2 <0.001 11.7 7.81–17.5 <0.001 Profession <0.001 Physician 130 (46.3%) 147 (53.7%) – – – – Administrative or other 102 (32.9%) 172 (67.1%) 1.76 1.25–2.49 0.001 1.24 0.77–2.02 0.400 Nurse 100 (32.2%) 178 (67.8%) 1.82 1.24–2.66 0.002 1.53 0.91–2.58 0.110 Nursing technician 45 (15.6%) 168 (84.4%) 4.67 2.73–8.00 <0.001 2.24 1.14–4.37 0.019 Work modality 0.017 Face-to-face/Mixed 300 (30.7%) 576 (69.3%) – – – – Non-attendance/Licensed 77 (43.9%) 89 (56.1%) 0.57 0.40–0.80 0.001 0.94 0.58–1.51 0.800 Work area 0.089 Hospitalization/Surgery 66 (30.3%) 146 (69.7%) – – – – Administrative or other related 56 (32.8%) 94 (67.2%) 0.89 0.58–1.35 0.600 0.61 0.36–1.03 0.065 Diagnostic support or other related 54 (38.4%) 86 (61.6%) 0.70 0.47–1.04 0.078 0.77 0.46–1.28 0.300 Outpatient. extramural and other related 81 (45.9%) 102 (54.1%) 0.51 0.32–0.81 0.005 0.83 0.48–1.42 0.500 Critical care 69 (36.1%) 113 (63.9%) 0.77 0.45–1.31 0.300 0.53 0.28–1.02 0.058 Emergency or urgent care 51 (26.2%) 124 (73.8%) 1.22 0.82–1.82 0.300 0.77 0.47–1.27 0.300 Works in COVID-19 area <0.001 No 220 (39.0%) 318 (61.0%) – – – – Yes 157 (25.6%) 347 (74.4%) 1.86 1.44–2.40 <0.001 1.47 1.02–2.11 0.040 EsSalud hospital of work >0.900 Edgardo Rebagliati Martins National Hospital 155 (31.6%) 298 (68.4%) – – – – Guillermo Almenara Irigoyen National Hospital 70 (33.2%) 114 (66.8%) 0.93 0.67–1.29 0.700 0.87 0.57–1.32 0.500 I Octavio Mongrut Muñoz Hospital/Villa Panamericana 24 (33.9%) 45 (66.1%) 0.90 0.57–1.42 0.600 0.27 0.14–0.53 <0.001 Alberto Sabogal Sologuren National Hospital 128 (33.5%) 208 (66.5%) 0.92 0.66–1.26 0.600 0.61 0.40–0.94 0.024 Time since second dose (days) 0.030 <0.001 <0.001 Median (p25-p75) 131.0 (124.0–134.0) 130.0 (123.0–134.0) Range (minimum-maximum) 14.0–142.0 14.0–142.0 n: unweighted absolute frequency; %: weighted percentage. The associations of seropositivity with age and time since second vaccination are shown in Fig. 1, Fig. 2. a Squared chi-square test with Rao and Scott second-order correction; Wilcoxon rank sum test for complex samples. b cOR: crude odds ratio. c CI: confidence interval. d aOR: adjusted odds ratio; SD: standard deviation. Table 3 Predictors of anti-SARS-CoV-2 antibody levels 14 to 142 days after receiving the second dose in health personnel vaccinated with BBIBP-CorV. Table 3Characteristics Crude analysis Adjusted analysis cGMRa 95%CI2 P value aGMR3 95%CI2 P value Sex Female – – – – Male 0.88 0.83–0.93 <0.001 0.77 0.74–0.80 <0.001 Age (years) <0.001 <0.001 Comorbidities None – – – – More than one 0.87 0.83–0.92 <0.001 1.00 0.96–1.05 0.900 Previous SARS-CoV-2 infection No – – – – Yes 7.7 7.40–8.02 <0.001 13.1 4.99–34.40 <0.001 Profession Physician – – – – Administrative or other 1.57 1.46–1.69 <0.001 0.98 0.92–1.05 0.600 Nurse 1.41 1.30–1.52 <0.001 1.01 0.94–1.08 0.800 Nursing technician 3.18 2.90–3.50 <0.001 1.30 1.20–1.41 <0.001 Work modality Face-to-face/Mixed – – – – Non-attendance/Licensed 0.48 0.45–0.51 <0.001 0.78 0.73–0.84 <0.001 Work area Hospitalization/Surgery – – – – Administrative or other related 1.39 1.28–1.51 <0.001 1.17 1.10–1.25 <0.001 Diagnostic support or other related 0.92 0.85–1.00 0.053 1.07 1.01–1.15 0.045 Outpatient. extramural and other related 0.65 0.59–0.72 <0.001 1.04 0.97–1.12 0.300 Critical care 0.98 0.88–1.08 0.600 0.85 0.79–0.93 <0.001 Emergency or urgent care 1.52 1.41–1.63 <0.001 1.06 1.00–1.13 0.057 Works in COVID-19 area No – – – – Yes 1.71 1.62–1.80 <0.001 1.30 1.24–1.36 <0.001 EsSalud hospital of work Edgardo Rebagliati Martins National Hospital – – – – Guillermo Almenara Irigoyen National Hospital 1.01 0.95–1.08 0.700 1.02 0.97–1.07 0.500 I Octavio Mongrut Muñoz Hospital/Villa Panamericana 1.09 0.99–1.20 0.067 0.62 0.58–0.68 <0.001 Alberto Sabogal Sologuren National Hospital 0.99 0.93–1.05 0.700 0.84 0.80–0.89 <0.001 Time since second dose (days) <0.001 <0.001 Interaction (Time since second dose (days) Previous SARS-CoV-2 infection) <0.001 a cGMR: crude geometric mean ratio; 2CI: confidence interval; 3aGMR: adjusted geometric mean ratio. The associations of seropositivity and time since the second vaccination are shown in the graphs in Fig. 1 . Fig. 1A shows the trend to a decrease in the predicted probability of seropositivity to Ac IgG SARS-CoV-2 with a longer time since the second vaccination dose, showing a notable reduction after day 110. In addition, in Fig. 1B we describe a slightly negative association between SARS-CoV-2 IgG antibody positivity and age. This is also shown in Fig. 1C in individuals aged 25, 45 and 60 years, with a sustained reduction in individuals aged 60 years after day 110. The predicted probability of seropositivity remained high over time in individuals who reported having had a previous infection compared to those who did not. Fig. 1E, F and 1G show the predicted probability of seropositivity according to occupational groups, working in a COVID-19 area and the hospital work site.Fig. 1 Association between SARS-CoV-2 IgG antibody positivity and time since receiving the second vaccination dose (days) (A), age (B), according to age group (C), history of previous SARS-CoV-2 infection (D), profession (E), work in COVID-19 area (F) and EsSalud hospital (G). Fig. 1 4.3 Predictors of anti-SARS-CoV-2 antibody levels In the adjusted analysis, a statistically significant association was found for male sex (aGMR = 0.77; 95%CI: 0.74–0.80), non-face-to-face/licensed work modality (aGMR = 0.78; 95%CI: 0.73–0.84), type of administrative or other related service (aGMR = 1.17; 95%CI: 1.10–1.25), work in critical care (aGMR = 0.85; 95%CI: 0.79–0.93), the nursing technician occupational group (aGMR = 1.30; 95%CI: 1.20–1.41), having worked at Villa Mongrut/Panamericana (aGMR = 0.62; 95%CI: 0.58–0.68) and Sabogal Hospital (aGMR = 0.84; 95%CI: 0.80–0.89), having reported a previous SARS-CoV-2 infection (aGMR = 13.1; 95%CI: 4.99–34.3), having worked in a COVID-19 area (aGMR = 1.30; 95%CI: 1.24–1.36), age (p < 0.001) and time since second vaccination (p < 0.001) (Table 2). The association between the geometric mean of SARS-CoV-2 IgG antibody levels and time since second vaccination is shown in Fig. 2 . Fig. 2A shows the trend to a reduction in the geometric mean of antibody levels in individuals 62 years of age or older. In addition, Fig. 2B presents a negative relation between antibody levels and time since second dose. Likewise, Fig. 2D and E presents the geometric mean of antibody levels and time since second vaccination comparing previous SARS-CoV-2 infection and disease presentation in individuals 25, 45 and 60 years old. In all these cases there was a reduction in the time of the geometric mean which was especially notable in those without previous infection and 60 years of age. Graphs 2F and 2G show the behavior of the curve according to previous infection and by sex. In Fig. 3 , it is important to highlight that the occupational group of nursing technicians and those who worked in the on-site modality had a higher predicted geometric mean of IgG SARS-CoV-2 antibodies. In addition, we showed in supplementary material the behavior of the curve according to groups by EsSalud hospital of work and working areas (by previous SARS-CoV-2 infection).Fig. 2 Association between the SARS-CoV-2 IgG antibodies (BAU/mL) levels and age (A), time (days) since receipt of the second vaccination dose (B) according to: age groups with no history of previous SARS-CoV-2 infection (C), age groups with this history (D), previous SARS-CoV-2 infection (E), groups according to sex and no history of previous SARS-CoV-2 infection (F), groups by sex with this history (G). Fig. 2 Fig. 3 Association between the SARS-CoV-2 IgG antibodies (BAU/mL) levels and time since receipt of the second vaccination dose (days) according to: groups by work modality according to previous SARS-CoV-2 infection, groups by profession according to previous SARS-CoV-2 infection, groups according to work in COVID-19 area and -CoV-2 infection and groups by working area according to previous SARS-CoV-2 infection. Fig. 3 5 Discussion Our study estimated the prevalence of post-vaccination seropositivity against SARS-CoV-2 and identified the predictors in Peruvian health personnel during 2021. Approximately 70% of the participants were seropositive, with younger age, having a history of COVID-19, and working in a COVID-19 area being associated with higher seropositivity. In addition, being male, younger age, having previous COVID-19 infection, working in a non-face-to-face modality, as well as in a COVID-19 area were predictors of higher antibody levels. We also found that antibody levels progressively decreased from day 110 after receiving the second vaccine dose. About seven out of ten participants presented seropositivity for anti-SARS-CoV-2 IgG anti-SARS-CoV-2 protein S antibodies generated from vaccination. There was a negative significant association with age and in subcategories of age group, highlighting a marked difference in seropositivity between the groups of 18–44, 45 to 59, and greater than or equal to 60 years of age. Some studies have reported similar findings on the relationship of age and antibody quantification. One study, by Ferenci et al. evaluated the levels of neutralizing antibodies after receiving a second dose of the BBIBP-CorV vaccine, and reported that 90% of participants younger than 50 had detectable antibodies, while 50% of those older than 80 had no detectable antibodies [37]. Likewise, a cohort study evaluating antibody response in participants after receiving the ChAdOx1 and BNT162b2 vaccines identified a group of non-responders, mainly composed of those over 75 years of age, males and individuals with chronic health problems [35]. The decrease in antibody levels with increasing age could be explained by immunosenescence [38], which would produce a reduced adaptive immune response and a decline in humoral and cellular immune response [39,40], indicating a greater need for booster doses in this age group. We found that being female was predictive of higher antibody levels compared to males. However, we observed a similar sustained reduction in antibody levels from day 110 onwards regardless of having had previous infection or not. Our finding is consistent with that described by Wei et al. who reported that being male was a predictor of lower antibody positivity [35]. Likewise, sex differences have been described after natural infection with COVID-19 [41]. This finding could be explained by the fact that biological sex affects the response of the innate and adaptive immune system, inducing different responses to a pathogen or vaccines [42,43]. In addition, males are at higher risk for diseases caused by X-linked alleles [44,45] and epigenetic expression in this group would determine exposure to sex steroids that would have a direct effect on immune function [[46], [47], [48]]. Our results showed that working in an area with patients with COVID-19 infection was a predictor of seropositivity and higher antibody levels and working in the non-face-to-face modality was a predictor of lower antibody levels. These findings could be explained in that the participants who worked in a face-to-face setting as well as in COVID-19 areas had an additional risk of becoming infected. These two characteristics are important because they could be associated with having a history of COVID-19. Thus, passive re-exposure would cause the immune system to produce a greater proportion of antibodies [49] compared to individuals without these occupational characteristics. Self-reported previous COVID-19 infection was a predictor of seropositivity and higher antibody levels. In addition, this group of participants recorded a lower reduction in antibody levels during the time after receiving the second dose compared to those who had no history of COVID-19. Previous studies have reported similar findings, describing higher antibody levels in vaccinated persons with a history of COVID-19 [50,51] independently of the age group [35]. Thus, it was proposed that, in a scenario of vaccine shortage, a dose of mRNA-type BNT162b2 vaccine could generate a robust immune response in persons who have had a COVID-19 infection three to six months prior to vaccination [35,52]. Immune response following inoculation with a COVID-19 vaccine involves two processes; the first is related to cellular response involving the production of different T-cell lineages, interleukins and interferons, while the second process is triggered by the first and involves the production of IgG immunoglobulins against viral antigens, such as protein S [53]. This process must be understood as a whole for the correct study of the immune response of a particular individual. These antibodies usually persist for up to six months [[54], [55], [56]] and then decline by 5 to 10-fold [57,58]. However, B and T cells can be detected even longer and are essential for protection against possible reinfections [[56], [57], [58], [59]], highlighting the role of cellular immunity [56,60]. Thus, our finding on the reduction of antibody levels 110 days after receiving the second vaccination dose indicate the need to administer a booster dose in the event of the circulation of new variants. Previous studies have highlighted the reduction in antibody levels generated after passive immunization over time. One study showed a decrease in antibody levels three months after receiving the second dose of the BNT162b2 vaccine (mRNA type vaccine produced by the Pfizer laboratory) [61]. On the other hand, another report indicated that between days 21–70 after receiving the second dose of ChAdOx1 (viral vector vaccine produced by the University of Oxford and the AstraZeneca laboratory) and BNT162b2 there was a 5-fold and 2-fold reduction in antibody levels, respectively [62]. Similarly, a study evaluating antibody levels in two groups after receiving a type of inactivated virus vaccine produced by the Sinopharm laboratory described a significant reduction in antibody levels after the third- or fourth week following receipt of the second dose. However, after receiving a third dose, the humoral immune response was very high [63]. The evidence is consistent in showing a decrease in the level of antibodies three months after receiving the second dose, a finding that is consistent with our results. This evidence supports the need for a booster dose after this period to increase antibody levels. Although low antibody values would not imply greater vulnerability and decreased protection against the virus, the need for a booster dose should be considered due to the adaptive mutability of the virus. This characteristic has generated different variants that can compromise the protection of vaccines against severe forms of the disease, hospitalization, admission to the intensive care unit and death [64,65]. By compromising vaccine protection, new waves, collapse of health systems and greater impact on the population could be generated [6,66], all of which are preventable scenarios. Our results elucidate the need for follow-up and immunologic surveillance (at humoral and cellular levels) in the risk group of health personnel. This surveillance would generate information for evidence-based decision making to identify the times at which a booster dose is necessary to maintain elevated antibody levels and provide better protection against SARS-CoV-2 and its variants [67]. Our study has limitations. Since antibody levels tend to decline over time, some of the seronegative measurements could previously have been seropositive. Despite having considered non-response and loss rates, we were unable to reach the estimated sample size for the study due to the high workload of the participants, limiting their availability for sample collection. To deal with this drawback, we went to the hospitals in which the health personnel worked to try to extract samples and increase the recruitment of participants. Another limitation is that memory bias could affect the validity of the information obtained in some questions such as the date of vaccination, self-report of previous SARS-CoV-2 infection, and a history of a positive diagnostic test for COVID-19. However, to reduce this bias, we provided support for filling out the survey and obtaining the vaccination date from the official portal of the Peruvian Ministry of Health. In addition, although previous studies have evaluated seropositivity and antibody production after receiving different vaccines worldwide, information related to the BBIBP-CorV vaccine is still scarce, thereby limiting comparison with other vaccines. Finally, this study only evaluated the humoral response of the immune system but not the cellular response. Studies on cellular immunity are needed for better evidence-based decision making. To our knowledge, this study is one of the first reports of active surveillance of SARS-CoV-2 anti-SARS-CoV-2 IgG antiprotein S antibody levels and post-vaccination seropositivity in a high-risk group, such as health care workers. 6 Conclusions Two out of three participants achieved seropositivity after receiving both doses of the inactivated BBIBP-CorV vaccine produced by the Sinopharm laboratory. We found predictors of seropositivity to be male sex, younger age, self-reported COVID-19 and working in an area with COVID-19 patients. In addition, a longer time after receiving the second dose was a predictor of lower antibody levels. We described a sustained drop in antibody levels after day 110, elucidating the need for a booster dose three months after the second dose. Funding This research was carried out with funding from the Peruvian Social Health Insurance - EsSalud. Uncited References [33]; [34]; [36]. CRediT authorship contribution statement Aleksandar Cvetkovic-Vega: Conceptualization, Investigation, Methodology, Writing – original draft, Writing – review & editing. Diego Urrunaga-Pastor: Conceptualization, Investigation, Methodology, Writing – original draft, Writing – review & editing. Percy Soto-Becerra: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Luis Edgardo Figueroa Morales: Conceptualization, Investigation, Writing – review & editing. Lizzete Fernández-Bolivar: Conceptualization, Investigation, Writing – review & editing. Sergio Alvizuri-Pastor: Conceptualization, Investigation, Writing – review & editing. Martin Oyanguren-Miranda: Conceptualization, Investigation, Writing – review & editing. Ibeth Melania Neyra Vera: Conceptualization, Investigation, Writing – review & editing. Elizabeth Emilia Carrillo Ramos: Conceptualization, Investigation, Writing – review & editing. Arturo Ampelio Sagástegui: Conceptualization, Investigation, Writing – review & editing. Roxana Milagros Contreras Macazana: Conceptualization, Investigation, Writing – review & editing. Diana Elizabeth Lecca Rengifo: Conceptualization, Investigation, Writing – review & editing. Nikolai Grande Castro: Conceptualization, Investigation, Writing – review & editing. Moises Apolaya-Segura: Conceptualization, Investigation, Methodology, Writing – review & editing. Jorge L. Maguiña: Conceptualization, Investigation, Methodology, Writing – review & editing. Declaration of competing interest The authors declare no conflicts of interest. Appendix A Supplementary data The following is the Supplementary data to this article:Multimedia component 1 Multimedia component 1 Acknowledgments The research team would like to thank all the health and non-health personnel who participated in the study. We would also like to thank the EsSalud 107 call team, as well as the teams of the IETSI units and the Flexible Supply Management that provided on-site support in the execution of the research and to the health personnel of the clinical pathology service of the EsSalud Hospital Nivel III Suárez Angamos. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.tmaid.2022.102514. ==== Refs References 1 COVID-19 Map [Internet]. Johns Hopkins Coronavirus Resource Center [Available from: https://coronavirus.jhu.edu/map.html 2 Farmacovigilancia de vacunas para COVID-19 - Sinopharm/BIBP [Internet]. Farmacovigilancia de vacunas para COVID-19. [Available from: https://covid-19pharmacovigilance.paho.org/ 3 Wang H. Zhang Y. Huang B. Deng W. Quan Y. Wang W. Development of an inactivated vaccine candidate, BBIBP-CorV, with potent protection against SARS-CoV-2 Cell 182 3 2020 713 721 e9 32778225 4 de Salud Ministerio Sala situacional COVID-19 perú [Available from: https://covid19.minsa.gob.pe/sala_situacional.asp 2020 5 Alcalde-Rabanal J.E. Lazo-González O. Nigenda G. Sistema de salud de Perú Salud Publica Mex 53 2011 s243 s254 21877089 6 Gianella C. Gideon J. Romero M.J. What does COVID-19 tell us about the Peruvian health system? Canadian Journal of Development Studies/Revue canadienne d'études du développement. 42 1–2 2021 55 67 7 de Salud Ministerio Resolución ministerial N° 345-2021-MINSA.pdf 2021 [Internet] 8 Xia S. Zhang Y. Wang Y. Wang H. Yang Y. Gao G.F. 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 9 He Z.R.L. Yang J. Guo L. Feng L. Ma C. Seroprevalence and humoral immune durability of anti-SARS-CoV-2 antibodies in Wuhan, China: a longitudinal, population-level, cross-sectional study 2021 The Lancet 10 Meng H. Mao J. Ye Q. Booster vaccination strategy: necessity, immunization objectives, immunization strategy and safety J Med Virol 2022 11 World Health Organization Interim statement on hybrid immunity and increasing population seroprevalence rates 2022 12 Food and Drug Administration La FDA autoriza la primera prueba que detecta anticuerpos neutralizantes de una infección reciente o anterior de SARS-CoV-2 [Internet] https://www.fda.gov/news-events/press-announcements/actualizacion-sobre-el-coronavirus-la-fda-autoriza-la-primera-prueba-que-detecta-anticuerpos 2020 [Available from: 13 Garcia-Beltran W.F. Lam E.C. Astudillo M.G. Yang D. Miller T.E. Feldman J. COVID-19-neutralizing antibodies predict disease severity and survival Cell 184 2 2021 476 488 e11 33412089 14 Khoury D.S. Cromer D. Reynaldi A. Schlub T.E. Wheatley A.K. Juno J.A. What level of neutralising antibody protects from COVID-19? medRxiv 2021 15 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 7 2021 1205 1211 34002089 16 Vickers M.A. Sariol A. Leon J. Ehlers A. Locher A.V. Dubay K.A. Exponential increase in neutralizing and spike specific antibodies following vaccination of COVID-19 convalescent plasma donors Transfusion 61 7 2021 2099 2106 33829513 17 Bueno S.M. Abarca K. González P.A. Gálvez N.M. Soto J.A. Duarte L.F. Interim report: safety and immunogenicity of an inactivated vaccine against SARS-CoV-2 in healthy chilean adults in a phase 3 clinical trial medRxiv 2021 18 Kenyon G. Vacuna-gate escalates in Peru Lancet Infect Dis 21 4 2021 463 33773129 19 Alvarez-Risco A. Mejia C.R. Delgado-Zegarra J. Del-Aguila-Arcentales S. Arce-Esquivel A.A. Valladares-Garrido M.J. The Peru approach against the COVID-19 infodemic: insights and strategies Am J Trop Med Hyg 103 2 2020 583 32500853 20 Silvia Valencia J. Soto Becerra P. Escobar Agreda S. Fernández Navarro M. Moscoso Porras M. Solari L. Efectividad de la vacuna BBIBP-CorV para prevenir infección y muerte en personal de salud 2021 2021 Perú 21 Jara A. Undurraga E.A. González C. Paredes F. Fontecilla T. Jara G. Effectiveness of an inactivated SARS-CoV-2 vaccine in Chile N Engl J Med 385 10 2021 875 884 34233097 22 WHO Evidence assessment: Sinopharm/BBIBP COVID-19 vaccine 2021 World Health Organization 23 Stefanelli P. Bella A. Fedele G. Fiore S. Pancheri S. Benedetti E. Longevity of seropositivity and neutralizing titers among SARS-CoV-2 infected individuals after 4 months from baseline: a population-based study in the province of Trento 2020 medRxiv 24 Valliant R. Dever J.A. Survey weights: a step-by-step guide to calculation 2018 Stata Press College Station, TX 25 Favresse J. Gillot C. Di Chiaro L. Eucher C. Elsen M. Van Eeckhoudt S. Neutralizing antibodies in COVID-19 patients and vaccine recipients after two doses of BNT162b2 Viruses 13 7 2021 1364 34372570 26 Perkmann T. Perkmann-Nagele N. Koller T. Mucher P. Radakovics A. Marculescu R. Anti-spike protein assays to determine SARS-CoV-2 antibody levels: a head-to-head comparison of five quantitative assays Microbiol Spectr 9 1 2021 e00247-21 27 Instituto Nacional de la Calidad 2022 INACAL Lima Available at: https://www.gob.pe/inacal 28 Clinical and Laboratory Standards Institute Wayne P. User verification of precision and estimation of bias; approved guideline 3rd 2014 CLSI USA CLSI document EP15-A3 29 Wayne P. For CaLSIUp, guideline-2nd eoqtpa 2008 CLSI USA CLSI document EP12-A2 30 Institute CaLS Evaluation of the linearity of quantitative measurement procedures: a statistical approach; approved guideline. CLSI Document EP06-a 2003 CLSIe Wayne (PA) 31 Infantino M. Pieri M. Nuccetelli M. Grossi V. Lari B. Tomassetti F. The WHO International Standard for COVID-19 serological tests: towards harmonization of anti-spike assays Int Immunopharm 100 2021 108095 32 Team R.C.R. A language and environment for statistical computing 2013 R Core Team Vienna, Austria 2021 33 Harris P.A. Taylor R. Minor B.L. Elliott V. Fernandez M. O'Neal L. The REDCap consortium: building an international community of software platform partners J Biomed Inf 95 2019 103208 34 Breiman L. Random forests Mach Learn 45 1 2001 5 32 35 Wei J. Stoesser N. Matthews P.C. Ayoubkhani D. Studley R. Bell I. Antibody responses to SARS-CoV-2 vaccines in 45,965 adults from the general population of the United Kingdom Nature Microbiology 6 9 2021 1140 1149 36 Wood S.N. Generalized additive models: an introduction with R 2006 chapman and hall/CRC 37 Ferenci T. Sarkadi B. Virus neutralizing antibody responses after two doses of BBIBP-CorV (Sinopharm, Beijing CNBG) vaccine 2021 medRxiv 38 Blomberg B.B. Frasca D. Quantity, not quality, of antibody response decreased in the elderly J Clin Invest 121 8 2011 39 Gustafson C.E. Kim C. Weyand C.M. Goronzy J.J. Influence of immune aging on vaccine responses J Allergy Clin Immunol 145 5 2020 1309 1321 32386655 40 Frasca D. Blomberg B.B. Aging induces B cell defects and decreased antibody responses to influenza infection and vaccination Immun Ageing 17 1 2020 1 10 31911808 41 Grzelak L. Velay A. Madec Y. Gallais F. Staropoli I. Schmidt-Mutter C. Sex differences in the evolution of neutralizing antibodies to severe acute respiratory syndrome coronavirus 2 J Infect Dis 224 6 2021 983 988 33693749 42 Klein S.L. Flanagan K.L. Sex differences in immune responses Nat Rev Immunol 16 10 2016 626 638 27546235 43 Markle J. Fish E.N. SeXX matters in immunity Trends Immunol 35 3 2014 97 104 24239225 44 Carrel L. Brown C.J. When the Lyon (ized chromosome) roars: ongoing expression from an inactive X chromosome Phil Trans Biol Sci 372 1733 2017 20160355 45 Tukiainen T. Villani A.-C. Yen A. Rivas M.A. Marshall J.L. Satija R. Landscape of X chromosome inactivation across human tissues Nature 550 7675 2017 244 248 29022598 46 Furman D. Hejblum B.P. Simon N. Jojic V. Dekker C.L. Thiébaut R. Systems analysis of sex differences reveals an immunosuppressive role for testosterone in the response to influenza vaccination Proc Natl Acad Sci USA 111 2 2014 869 874 24367114 47 Sex and sex steroids impact influenza pathogenesis across the life course Vom Steeg L.G. Klein S.L. Seminars in immunopathology 2019 Springer 48 Vom Steeg L.G. Klein S.L. Sex steroids mediate bidirectional interactions between hosts and microbes Horm Behav 88 2017 45 51 27816626 49 Gobbi F. Buonfrate D. Moro L. Rodari P. Piubelli C. Caldrer S. Antibody response to the BNT162b2 mRNA COVID-19 vaccine in subjects with prior SARS-CoV-2 infection Viruses 13 3 2021 422 33807957 50 Krammer F. Srivastava K. Alshammary H. Amoako A.A. Awawda M.H. Beach K.F. Antibody responses in seropositive persons after a single dose of SARS-CoV-2 mRNA vaccine N Engl J Med 384 14 2021 1372 1374 33691060 51 Demonbreun A.R. Sancilio A. Velez M.P. Ryan D.T. Saber R. Vaught L.A. Comparison of IgG and neutralizing antibody responses after one or two doses of COVID-19 mRNA vaccine in previously infected and uninfected individuals EClinicalMedicine 38 2021 101018 52 Buonfrate D. Piubelli C. Gobbi F. Martini D. Bertoli G. Ursini T. Antibody response induced by the BNT162b2 mRNA COVID-19 vaccine in a cohort of health-care workers, with or without prior SARS-CoV-2 infection: a prospective study Clin Microbiol Infect 27 12 2021 1845 1850 34329793 53 Burckhardt R.M. Dennehy J.J. Poon L.L. Saif L.J. Enquist L.W. Are COVID-19 vaccine boosters needed? The science behind boosters J Virol 96 3 2021 e01973-21 54 Müller L. Andrée M. Moskorz W. Drexler I. Walotka L. Grothmann R. Age-dependent immune response to the Biontech/Pfizer BNT162b2 coronavirus disease 2019 vaccination Clin Infect Dis 73 11 2021 2065 2072 33906236 55 Doria-Rose N. Suthar M.S. Makowski M. O'Connell S. McDermott A.B. Flach B. Antibody persistence through 6 months after the second dose of mRNA-1273 vaccine for Covid-19 N Engl J Med 384 23 2021 2259 2261 33822494 56 Pegu A. O'Connell S.E. Schmidt S.D. O'Dell S. Talana C.A. Lai L. Durability of mRNA-1273 vaccine–induced antibodies against SARS-CoV-2 variants Science 373 6561 2021 1372 1377 34385356 57 Cho A. Muecksch F. Schaefer-Babajew D. Wang Z. Finkin S. Gaebler C. Anti-SARS-CoV-2 receptor-binding domain antibody evolution after mRNA vaccination Nature 600 7889 2021 517 522 34619745 58 Sette A. Crotty S. Adaptive immunity to SARS-CoV-2 and COVID-19 Cell 184 4 2021 861 880 33497610 59 Dan J.M. Mateus J. Kato Y. Hastie K.M. Yu E.D. Faliti C.E. Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection Science 371 6529 2021 eabf4063 60 Goel R.R. Painter M.M. Apostolidis S.A. Mathew D. Meng W. Rosenfeld A.M. mRNA vaccines induce durable immune memory to SARS-CoV-2 and variants of concern Science 374 6572 2021 abm0829 34648302 61 Favresse J. Bayart J.-L. Mullier F. Elsen M. Eucher C. Van Eeckhoudt S. Antibody titres decline 3-month post-vaccination with BNT162b2 Emerg Microb Infect 10 1 2021 1495 1498 62 Shrotri M. Navaratnam A.M. Nguyen V. Byrne T. Geismar C. Fragaszy E. Spike-antibody waning after second dose of BNT162b2 or ChAdOx1 Lancet 398 10298 2021 385 387 34274038 63 Saeed U. Uppal S.R. Piracha Z.Z. Uppal R. SARS-CoV-2 spike antibody levels trend among Sinopharm vaccinated people Iran J Public Health 50 7 2021 1486 34568189 64 Bian L. Gao Q. Gao F. Wang Q. He Q. Wu X. Impact of the Delta variant on vaccine efficacy and response strategies Expet Rev Vaccine 20 10 2021 1201 1209 65 Vasireddy D. Vanaparthy R. Mohan G. Malayala S.V. Atluri P. Review of COVID-19 variants and COVID-19 vaccine efficacy: what the clinician should know? J Clin Med Res 13 6 2021 317 34267839 66 da Silva S.J.R. Pena L. Collapse of the public health system and the emergence of new variants during the second wave of the COVID-19 pandemic in Brazil One Health 13 2021 100287 67 Rubin R. COVID-19 vaccines vs variants—determining how much immunity is enough JAMA 325 13 2021 1241 1243 33729423	36462747	PMC9710108	NO-CC CODE	2022-12-12 23:21:01	no	Travel Med Infect Dis. 2023 Nov 30 March-April; 52:102514	utf-8	Travel Med Infect Dis	2,022	10.1016/j.tmaid.2022.102514	oa_other
==== Front J Affect Disord J Affect Disord Journal of Affective Disorders 0165-0327 1573-2517 Elsevier B.V. S0165-0327(22)01304-0 10.1016/j.jad.2022.11.044 Research Paper Predicting depression and anxiety of Chinese population during COVID-19 in psychological evaluation data by XGBoost Tian Zhanxiao a Qu Wei a Zhao Yanli a Zhu Xiaolin a Wang Zhiren a Tan Yunlong a Jiang Ronghuan b Tan Shuping a⁎ a Peking University HuiLongGuan Clinical Medical School, Beijing HuiLongGuan Hospital, Beijing 100096, China b The First Medical Center of Chinese People's Liberation Army General Hospital, No.100 West Fourth Ring Road, Fengtai District, Beijing 100853, China ⁎ Corresponding author. 30 11 2022 30 11 2022 16 3 2022 27 10 2022 18 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 Due to the onset of sudden stress, COVID-19 has greatly impacted the incidence of depression and anxiety. However, challenges still exist in identifying high-risk groups for depression and anxiety during COVID-19. Studies have identified how resilience and social support can be employed as effective predictors of depression and anxiety. This study aims to select the best combination of variables from measures of resilience, social support, and alexithymia for predicting depression and anxiety. Methods The eXtreme Gradient Boosting (XGBoost1) model was applied to a dataset including data on 29,841 participants that was collected during the COVID-19 pandemic. Discriminant analyses on groups of participants with depression (DE2), anxiety (AN3), comorbid depression and anxiety (DA4), and healthy controls (HC5), were performed. All variables were selected according to their importance for classification. Further, analyses were performed with selected features to determine the best variable combination. Results The mean accuracies achieved by three classification tasks, DE vs HC, AN vs HC, and DA vs HC, were 0.78, 0.77, and 0.89. Further, the combination of 19 selected features almost exhibited the same performance as all 56 variables (accuracies = 0.75, 0.75, and 0.86). Conclusions Resilience, social support, and some demographic data can accurately distinguish DE, AN, and DA from HC. The results can be used to inform screening practices for depression and anxiety. Additionally, the model performance of a limited scale including only 19 features indicates that using a simplified scale is feasible. Keywords Machine learning Depression Anxiety Resilience Social support COVID-19 pandemic ==== Body pmc1 Introduction Since its outbreak, COVID-19 rapidly became a pandemic (Wang et al., 2020a). Several factors including demographic characteristics (e.g., gender, occupation, education level, health status) and those related to COVID-19 (e.g., physical symptoms, contact history, worry level, and preventive measures) significantly impacted people's mental health, which, in some cases, further developed into psychiatric disorders (Banerjee and Rai, 2020; Minihan et al., 2020; Wang et al., 2020b; de Figueiredo et al., 2021), such as depression, anxiety, insomnia, and post-traumatic stress symptoms (Bao et al., 2020; Huang and Zhao, 2020; Luo et al., 2020, Luo et al., 2020; Shader, 2020; Li et al., 2022). Typically, diagnoses for depression and anxiety depend on the clinical evaluation of symptoms, as well as scales, such as the Patient Health Questionnaire-9 (PHQ-9) and Generalized Anxiety Disorder-7 (GAD-7). However, medical resource shortages during the COVID-19 pandemic made it increasingly challenging to identify these psychiatric disorders and intervene (Emanuel et al., 2020). This necessitated the development of psychiatric screening tools with minimal demand on the already limited resources of clinical staff. Although the aforementioned measures are readily accessible, they only offer short-term evaluations based on patients' subjective experiences, which may only detect the recent abnormal (last two weeks) psychological fluctuations of such patients (Garabiles et al., 2020). Therefore, it is difficult for PHQ-9 and GAD-7 to effectively describe the risk of depression or anxiety. We hope to use some indicators that can describe the risk of depression or anxiety to predict depression and anxiety, so as to quantify the probability of depression and anxiety. In addition, because there are many risk factors related to depression and anxiety, it is difficult for participants to complete if all the factors are included. It may ultimately affect the prediction results. Thus, we hope to find some stable key variables to simplify the whole process without affecting the prediction effect. The incidence of depression and anxiety, especially during COVID-19, were affected by many factors, such as knowledge and concerns related to COVID-19 (Tee et al., 2020; Wang et al., 2021a), more physical symptoms (Wang et al., 2021b), facemask use (Wang et al., 2020, Wang et al., 2020, Wang et al., 2020), loss of confidence in doctors (Wang et al., 2021, Wang et al., 2021, Wang et al., 2021), and number of children in the family (Le et al., 2020). In addition to the degree of enthusiasm about the government's response. A meta analysis showed that reduction in the prevalence of depression was significantly related to a rapid and strict response from the government (Lee et al., 2021). We plan to build upon the collective and uncertain factors like government response by examining more individual and stable indicators. Psychological resilience is an individual's ability to recover from negative experiences and flexibly adapt to the changing external environment (Werner, 1995) or withstand a high level of destructive changes without being significantly influenced (Lazarus, 1993). This is regarded as a dynamic mechanism for mitigating the impact of adverse events (Tusaie and Dyer, 2004). Many studies have demonstrated that resilience is an essential factor affecting depression and anxiety (Kanako et al., 2018; Morete et al., 2018). For example, resilience can mitigate the adverse effects of stress (Garmezy and Masten, 1986; Sheerin et al., 2018), regulate depressive symptoms caused by personality characteristics and family dysfunctions (Chang et al., 2019; Gong et al., 2019), and help reduce the risk of depression for individuals with negative childhood experiences (A et al., 2017). Moreover, adolescents with low resilience levels are at high risk of lifelong use of antidepressants and anxiolytics (Ayako et al., 2015). Therefore, evaluating an individual's resilience could help predict mental health outcomes. Another significant factor affecting the incidence of depression and anxiety is social support, which generally includes objective support, subjective experience, and the utilization of social support. Presently, many reports have confirmed the relationship between social support and psychiatric disorders (Rothon et al., 2012; Koelmel et al., 2016; Tomás et al., 2016; Cao et al., 2018). For example, patients with depression generally have an abnormal social support system (Nr et al., 2020). The lack of social support exerts an adverse effect on depression by serving as a stressor (Li et al., 2017). Improvements in depressive symptoms positively correlate with improved utilization of social support (Gariepy et al., 2016). Therefore, these studies revealed that social support could be an important factor for depression. Alexithymic patients are unable to properly describe their emotional experience, and lack fantasy and practical thinking (Hogeveen and Grafman, 2021). Alexithymia is positively related to the severity of mental symptoms (Mcgillivray et al., 2017). Specifically, one study found that there was an indirect relation between alexithymia and affective disorder symptoms with emotion regulation as the intermediate variable (Preece et al., 2022). This means that alexithymia, as a relatively stable risk factor, may make an individual more prone to affective disorder by influencing emotional regulation. Thus, alexithymia can also be employed as a good predictor of depression and anxiety. The current study aimed to predict depression and anxiety using psychological resilience, social support, and alexithymia as predictors, and also select some key predictors for simplifying the whole process. 2 Methods 2.1 Participants Participants were from different occupations in different provinces of China and were recruited online during the early stages of the COVID-19 pandemic (between February 2020 and May 2020). Participants were required to provide their personal information and complete the Connor–Davidson resilience scale (CD-RISC), Social Support Questionnaire (SSQ), Diagnostic Criteria for Psychosomatic Research (DCPR), followed by the PHQ-9 and GAD-7 evaluations. Participants were informed of the purpose and significance of the study and signed informed consent before undergoing any assessment. Data quality was controlled by employing the following rules: 1) each participant must have a unique IP address; 2) all the items were accomplished. A total of 31,017 participants were registered in our online evaluation system. Of these, 1176 failed to pass the quality control, so 29,841 were included in the final analysis. 2.2 Measures This was a cross-sectional study. The online evaluation consisted of two parts: general information and psychological evaluation. The general information part consisted of demographic data (e.g., age, gender, education level, marital status, occupation) and information relating to COVID-19 included variables, such as contact with COVID-19 patients, worry about COVID-19, and general health status. Psychological assessments included CD-RISC, SSQ, DCPR, PHQ-9, and GAD-7. 2.3 Assessment of psychological resilience The Chinese version of the CD-RISC (Kathryn et al., 2003) is a self-report measure employed to measure personal psychological resilience within the past 30 days. The CD-RISC consisted of 25 items with the options for each item rated from 0 to 4 (not at all (0), rarely (1), sometimes (2), often (3), and almost always (4)). The scale contains items measuring three factors: 1) tenacity (11–23), 2) strength (1, 5, 7, 8, 9, 10, 24, and 25), and 3) optimism (2, 3, 4, and 6). A high score represents enhanced psychological resilience. The CD-RISC has demonstrated significant reliability and validity within different populations (Windle et al., 2011; Ye et al., 2017). 2.4 Assessment of social support The SSQ (Sarason et al., 1983) is a self-report measure for evaluating the level of individual social support with a high score corresponding with a high social support level. The consistency of the total score, when retested using Chinese college students, was 0.92 (p < 0.01); each item was between 0.89 and 0.94, which corresponded to good reliability and validity. 2.5 Assessment of alexithymia The DCPR (Porcelli and Sonino, 2007) is a simple, effective, and reliable regular interview tool, which was developed by an international psychosomatic research group and can be employed to screen and diagnose psychosomatic and psychophysiological disorders. In the revised DCPR, a minimum of three items were considered alexithymia from the following six items: 1) inability to utilize appropriate emotions, 2) tendency to describe details rather than feelings, 3) lack of an interesting life, 4) exhibiting thought patterns that are more related to external events than fantasies or emotions, (5) being unaware of the relationship between common physical reactions and the various emotional experiences, and (6) displaying occasionally violent and often inappropriate emotional behaviors. 2.6 Assessment of anxiety and depression The PHQ-9 (Wang et al., 2014) is a self-report measure for identifying whether individuals are suffering from depression. Scores correspond with normal (0–4), mild (5–9), moderate (10–14), moderate to severe (15–19), and severe (20–27) depression. In the current study, a score of 4 was used as the boundary between healthy control and depression. The PHQ-9 exhibited strong reliability and validity in Chinese individuals (the internal consistency was 0.86). A recent online evaluation via smartphones, in addition to a paper evaluation, obtained similar results (Zhen et al., 2020). This GAD-7 (He et al., 2010) is a self-report measure for identifying whether subjects suffer from anxiety. Scores correspond with normal (0–4), mild (5–9), moderate (10–14), and severe (≥15) levels. The current study designated 4 as the boundary between healthy control and anxiety. The retest reliability for the Chinese version of the GAD-7 was 0.85. 2.7 Descriptive and data analysis Based on results of the PHQ-9 and GAD-7, participants were labeled DE (depression, PHQ-9 ≥ 5 & GAD-7 ≤ 4), AN (anxiety, GAD-7 ≥ 5 & PHQ-9 ≤ 4), DA (depression and anxiety comorbidity, PHQ-9 ≥ 5 & GAD-7 ≥ 5), and HC (health control, PHQ-9 ≤ 4 & GAD-7 ≤ 4). The HC group was combined separately with DE, AN and DA group to form three new datasets. We use univariate analysis to analyse the relations between features and labels. To be specific, the Mann–Whitney U test was employed to compare the continuous data of the non-normal distribution. Pearson chi-square (χ2) test was applied for categorical and dichotomous variables. Two-tailed test of significance used: p < 0.01. 2.8 Machine learning model After comparing several models based on performance, including the support vector machine, random forest, logistic regression, and Xgboost; Xgboost was selected as our classifier. Xgboost was first proposed by Tianqi Chen (Chen and Guestrin, 2016). It is a widely recognized and efficient machine learning technique, which assembles weak prediction models through continuous feature splitting, as well as the addition of new trees, to generate a more accurate model. It is also an open-source package. 2.9 Data preprocessing Each dataset was split into the training and testing sets in an 8:2 ratio. Further, a 10-fold cross-validation was conducted within the training set to optimize the algorithm. The holdout testing set was only employed to measure the performance of the model. 2.10 Predictions and evaluation The area under the curve (AUC) was employed as a primary indicator to evaluate the model. An AUC of 0.8–0.9 is generally considered to be good, while an AUC of >0.9 is considered excellent (Hosmer and Lemeshow, 2000). Other performance indicators, including the overall accuracy, precision, recall, and F1 score, were also employed. 2.11 Feature selection “Gain” is a built-in method of the Xgboost model, employed to determine the significance of selected features during prediction. The F-score represents the degree of feature significance (the higher the F-score, the more significant the feature). Afterward, feature selection was performed based on the significance of the feature, as well as its predicted performance. Feature combination was also performed to fit three groups of people according to the feature selection result. 3 Results 3.1 Descriptive and data analysis of demographic characteristics, COVID-19 related factors, current health status, and psychological factors The results of the descriptive analysis of the 29,841 participants (male: 10,592, female: 19,249) is presented in Table 1 .Table 1 Sample characteristics based on DE, AN, DA, and HC (n = 29,841). Table 1 DE (n = 4632) AN (n = 1219) DA (n = 8055) HC (n = 15,935) Sex: N (%) a,⁎ b,⁎ c,⁎ Female 2714 (58.6) 765 (62.8) 5259 (65.3) 6268 (39.3) Age: Mean (SD) 30.9 ± 8.3a,⁎ 32.9 ± 8.9b,⁎ 31.0 ± 8.2c,⁎ 30.0 ± 7.6 Education level: Mean (SD) 15.7 ± 2.5a,⁎ 16.4 ± 2.5b,⁎ 15.9 ± 2.5c,⁎ 15.5 ± 2.5 Marital status: N (%) a,⁎ b,⁎ c,⁎ Married 2212 (47.8) 756 (62.0) 4201 (52.2) 7693 (48.3) Divorce 107 (2.3) 27 (2.2) 238 (3.0) 204 (1.3) Cohabitating 38 (0.8) 12 (1.0) 85 (1.0) 49 (0.3) Single 2231 (48.2) 409 (33.6) 3408 (42.3) 7855 (49.3) Others 44 (0.9) 15 (1.2) 123 (1.5) 134 (0.8) COVID-19 exposure: N (%) a,⁎ b,⁎ c,⁎ FFI 871 (18.8) 234 (19.2) 1526 (18.9) 3656 (22.9) CNI 1785 (38.5) 426 (34.9) 3259 (40.5) 6522 (40.9) CH 33 (0.7) 11 (0.9) 36 (0.4) 53 (0.3) SC 112 (2.4) 31 (2.5) 220 (2.7) 145 (0.9) NCH 1814 (39.2) 513 (42.1) 2993 (37.2) 5541 (34.8) Patient 17 (0.4) 4 (0.3) 21 (0.2) 18 (0.1) Worried status: N (%) a,⁎ b,⁎ c,⁎ Not at all 649 (14.0) 114 (9.4) 729 (9.1) 3214 (20.2) A little 2165 (46.7) 489 (40.1) 3281 (40.7) 7209 (45.2) Some 456 (9.8) 148 (12.1) 971 (12.1) 1224 (7.7) Worry 843 (18.2) 268 (22.0) 1602 (19.9) 2688 (16.9) Very much 519 (11.2) 200 (16.4) 1472 (18.3) 1600 (10.0) Current health status: N (%) a,⁎ b,⁎ c,⁎ Good 3886 (83.4) 1036 (85.0)b,⁎ 5532 (68.7)c,⁎ 15,327 (96.2) Ok 693 (15.0)a,⁎ 172 (14.1)b,⁎ 2200 (27.3)c,⁎ 579 (3.6) Not very well 45 (1.0)a,⁎ 11 (0.9)b,⁎ 277 (3.4)c,⁎ 23 (0.1) Bad 8 (0.2)a,⁎ 0 (0.0)b,⁎ 46 (0.6)c,⁎ 6 (0.0) Social support: Mean (SD) Objective support 8.2 (2.9)a,⁎ 9.1 (3.0)b,⁎ 7.4 (2.8)c,⁎ 10.4 (3.3) Subjective support 21.8 (5.0)a,⁎ 23.5 (4.8)b,⁎ 20.1 (5.0)c,⁎ 26.1 (4.7) Used of support 7.8 (1.9)a,⁎ 8.2 (1.9)b,⁎ 7.1 (1.9)c,⁎ 9.4 (2.0) Total support 37.8 (7.8)a,⁎ 40.8 (7.7)b,⁎ 34.7 (7.9)c,⁎ 45.9 (8.0) Psychological resilience: Mean (SD) Toughness 30.4 (8.1)a,⁎ 32.1 (8.1)b,⁎ 26.4 (8.2)c,⁎ 37.7 (8.1) Strength 22.1 (4.8)a,⁎ 22.9 (4.5)b,⁎ 19.2 (5.0)c,⁎ 26.2 (4.5) Optimism 9.5 (2.5)a,⁎ 9.8 (2.5)b,⁎ 8.3 (2.6)c,⁎ 10.7 (2.5) Total resilience 62.0 (13.9)a,⁎ 64.8 (13.6)b,⁎ 53.9 (14.4)c,⁎ 74.6 (13.6) Alexithymia: Mean (SD) 2.2 (1.1)a,⁎ 2.2 (1.1)b,⁎ 2.2 (1.1)c,⁎ 2.3 (0.9) DE: depression; AN: anxiety; DA: depression and anxiety comorbidity; HC: health control. FFI: family or friends infected; CNI: colleagues or neighbors infected; CH: has contact history; SC: suspected case; NCH: does not have contact history. a Significant difference between DE and HC. b Significant difference between AN and HC. c Significant difference between DA and HC. The demographic data reveals that older, higher education level and divorced women are more likely to suffer from depression and anxiety (p<0.01). As for the factors related to the COVID-19 epidemic, having patients with infection (including family members, friends or colleagues) around them, having COVID-19 contact history or being infected are tend to depression and anxiety (p<0.01). In terms of general health status, the worse the general health status, the more prone to depression and anxiety (p<0.01). In psychological assessment, social support and resilience are protective factors of depression and anxiety (p<0.01). The higher the score of SSQ or CDRISC, the less likely to suffer from depression and anxiety. The results of the descriptive analysis showed all factors are related to depression and anxiety. However, it is hard to explain the impact of these variables on depression and anxiety, so it is necessary to further quantify the predictive effect of these variables on depression and anxiety with machine learning models. 3.2 Predictive performance Among the four groups in the discriminant analysis, the prediction of DA was the most accurate, followed by DE and AN. The AUC of the three prediction tasks were ≥0.85, indicating that the model exhibited high stability and reliability for the three tasks. The predictive performance of the model for the three tasks is summarized in Table 2 , and their receiver operating characteristics (ROC) are shown in Fig. 1 .Table 2 Performance of the three groups based on the Xgboost model. Table 2Groups Accuracy Precision Recall F1-score AUC DE vs HC* 0.78 0.77 0.80 0.78 0.86 AN* vs HC 0.77 0.78 0.80 0.79 0.85 DA* vs HC 0.89 0.89 0.89 0.89 0.95 DE: depression group (n = 4632); AN: anxiety group (n = 1219); DA*: depression and anxiety comorbidity group (n = 8055). Fig. 1 ROC curves and AUC values for the three prediction tasks. Fig. 1 3.3 Feature selection and feature importance 3.3.1 Feature selection The feature screening results for DE, AN and DA is summarized in Fig. 2 . The red label indicates that the accuracy of the model was relatively high when enough features were selected. (DE: feature numbers = 22, accuracy = 0.76; AN: feature numbers = 19, accuracy = 0.77; DA: feature numbers = 28, accuracy = 0.88.)Fig. 2 Feature selections of DE (a), AN (b) and DA (c) prediction (the X-axis represents the feature numbers, as screened by the “Gain” method, and the Y-axis represents accuracy, which is the relative number of features that were employed for the prediction). The red circle and square represent the ideal accuracy and related feature numbers, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 2 3.3.2 Feature importance The top features of DE, AN and DA are plotted in Fig. 3 .Fig. 3 The significance of the top features in DE (a), AN (b), DA (c) prediction. The X-axis indicates the feature name, and the Y-axis indicates the F-score. Fig. 3 Fig. 3(a) shows the top 22 features of DE. The inclusion of five demographic characteristics accounted for 22.7 % of all the 22 features. Therefore, “AGE,” “WorkingPlace,” “EducationLevel,” and “WorryofnewCvirus” accounted for the top four features among the 22; eleven CD-RISC items (50.0 %) were included, and Item 3 (“Sometimes fate or god can help”) scored the highest; six SSQ items (27.3 %) were included, and Item 6 (“Whether there is someone to talk when encountering troubles in the past year”) was the most significant; and no DCPR items were included in the top 22 features. Fig. 3(b) shows the top 19 features of AN. The inclusion of six demographic characteristics accounted for 31.6 % of all the 19 features. “AGE,” “WorkingPlace,” and “EducationLevel” accounted for the top three features among the 19; eight CD-RISC items (42.1 %) were included, and “Sometimes fate or god can help” still scored the highest; five SSQ items (26.3 %) were included, and “Whether there is someone to talk, when encountering troubles in the past year” was the most significant. DCPR did not account for any item in the 19 features. Fig. 3(c) shows the top 28 features of DA. The inclusion of five demographic characteristics accounted for 17.9 % of all the 28 features: “AGE” was still the most significant feature among the 28; 14 CD-RISC items (50.0 %) were included, and Item 2 (“I have close and safe relationship”) scored the highest; seven SSQ items (25.0 %) were included, and Item 7 (“Whether there is someone for help when encountering trouble in the past year”) ranked 2nd among the 28 features. Two DCPR items: Item 6 (“Have you ever had an occasional but violent outburst of anger, crying or joy, which is incompatible with the relationship between the events at that time or your normal behavior”) and Item3 (“Do you often fantasize”) (7.1 %) were included, and they ranked as the last two among the 28 features. 3.4 Feature combination After the first feature-selection step, twenty overlapping features were observed among the DE, AN, and DA groups. We built a combination of 19 selected features, including only the overlapping variables, but excluding the features with a correlation coefficient of >0.8 (SSQ7). The Xgboost model based on 19 selected features achieved a good predictive power (accuracy: DE = 0.75; AN = 0.75; DA = 0.86). A description of the definition for the 19 selected features can be found in Online supplementary Table S1. 4 Discussion Based on the Xgboost model, this study explored the use of the CD-RISC, SSQ, DCPR, and demographic characteristics for predicting DE, AN and DA and achieved good model performances. Among the three groups of classification tasks, the DA group performed significantly better than the other two. On the one hand, there was a certain gap between the data volumes of the three groups. Thus, we down-sampled the data of the normal control group according to the sample size of the patient group to avoid the bias caused by unbalanced data. However, since downsampling is actually just reducing the overall sample size, the small sample size might result in underfitting, followed by reduced accuracy. On the other hand, DA combines the depression and anxiety characteristics, so this makes their recognition easier compared to normal controls. Following feature selection, it was easy for the participants to accept 19 features, and the performance of such a combination was only slightly lower than those of 28 and 36 features, indicating that these features were key features for predicting DE, AN, and DA. As a result, our study provides a simplified instrument to screen depression and anxiety disorders, which may improve the efficiency of clinical evaluation. “AGE” was the most significant feature during feature selection. The demographic data show that the age distribution of the patient groups were higher than those of the HC, indicating that older people might have higher risks of depression, anxiety, or comorbidity during the pandemic than younger people, possibly due to the cognitive control deficits (Dotson et al., 2020). A recent study demonstrated that young age was related to reduced depressive symptoms (Alonso Debreczeni and Bailey, 2021), which aligns with the findings reported here. Additionally, “WorkingPlace,” “EducationLevel,” “WorryofnewCvirus,” and “Contactwithvirus” accounted for the overlapping features, and they all ranked high among the screened features of the three groups. The demographic data revealed that there were significant differences between these four variables among the groups of DE, AN, DA and HC (p < 0.01). Studies have demonstrated that the incidence of depression and anxiety increased following the COVID-19 pandemic (Choi et al., 2020; Luo et al., 2020, Luo et al., 2020). In our study, “WorkingPlace”, “WorryofnewCvirus” and “Contactwithvirus” were associated with the COVID-19 pandemic; which corresponds with findings from previous studies. Conversely, one study showed that COVID-19 patients did not show significant depression and anxiety tendencies compared with psychiatric patients and healthy controls (Hao et al., 2020). A meta-analysis on the association between patients with mood disorders and COVID-19 outcomes, found that mood disorders were a high-risk factor for COVID-19 infection and death (Ceban et al., 2021, Ceban et al., 2021). Although our study shows that there is a significant difference in the proportion of COVID-19 patients among the four groups (p < 0.01), after feature selection, the proportion of COVID-19 patients does not show a good predictive effect (i.e., it is not among the top variables), which is also consistent with previous studies to some extent. Since the proportion of COVID-19 patients in this study is <1 %, it may not be enough to conclude that the proportion of COVID-19 patients will affect the prediction results. In addition, when examining the association between patients with COVID-19 symptoms that subsided within 12 weeks and the high incidence of depression, a systematic review found that the severity of acute COVID-19 does not lead to an increase in the incidence of depression (Renaud-Charest et al., 2021). However, fatigue and cognitive impairment are more likely to occur after 12 weeks of COVID-19 infection (Ceban et al., 2021, Ceban et al., 2021). While patients with depression have extensive cognitive impairment (Wang et al., 2022), perhaps cognitive impairment is the mediator of depression 12 weeks or more after a COVID-19 infection. Finally, the studies involving the China family panel that explored the relationship between depression and educational achievement confirmed that high educational achievement reduced the risk of depression (Shen, 2020), similar to what was observed in our research. In our study, the patient groups generally had high education levels. Additional data would be required to further explore the reason. Regarding the CD-RISC scale, Items 2 (“I have close and safe relationship”), 3 (“Sometimes fate or god can help”) and 20 (“I had to act on my hunch”) ranked as the top three, respectively. According to the three-dimensional score method of CD-RISC (Yu et al., 2007), Items 2 and 3 represented optimism, indicating that these patient groups might be less-optimistic. A strong negative correlation has been noted between optimism and the COVID-19 pandemic (Ran et al., 2020). Item 20 belongs in the tenacity category. A recent study indicated that tenacity correlated with depression and anxiety (Ran et al., 2020). The more tenacious the participants were, the less likely they were to be affected by the pandemic. For SSQ, Items 3 (“How is your relationship with your neighbors”), 6 (“Whether there is someone to talk when encountering troubles in the past year”), and 7 (“Whether there is someone for help when encountering troubles in the past year”) ranked as the top three. Item 3 represented the participant's relationship with their neighbors. A previous study indicated that the neighborhood was key to reducing depressive symptoms by building a kind of social cohesion among the members of a community (Miao et al., 2019). Items 6 and 7 represent how the participants confided and asked for help when they encountered troubles, respectively. They exhibited very strong correlation (r = 0.8). The patient group tended to talk to and ask for help from fewer people compared with the other groups (p < 0.01). A systematic review revealed that the presence of confidants was identified as a factor of social relations that was significantly associated with depression (Schwarzbach et al., 2014). The results of this study demonstrated that alexithymia was one of the etiological factors that determined generalized anxiety disorder and depression (Lenzo et al., 2020). It is generally believed that alexithymia abnormally regulates emotional processes, and this is a critical risk factor regarding the occurrence and development of mood disorders and psychosomatic diseases (Panayiotou et al., 2021). Item3 (“Do you often fantasize”) and Item 6 (“Have you ever had an occasional but violent outburst of anger, crying or joy, which is incompatible with the relationship between the events at that time or your normal behavior”) are included in the DA group, similar to previous studies (Lisha et al., 2018; Palser et al., 2018). Considering the role of emotion regulation in alexithymia and affective disorder, cognitive behavior therapy (CBT), as a kind of intervention aimed at emotion regulation can be used as a part of comprehensive treatment of alexithymia (Ho et al., 2020). Especially during the COVID-19 pandemic, online CBT has great potential (Zhang and Ho, 2017). In conclusion, based on the CD-RISC, SSQ, and DCPR measures, as well as demographic data, an excellent performance regarding the predictions of DE, AN, and DA was obtained. The performance supported the possibility for screening potential patients of depression and anxiety online. After the feature selection, the final 19 variables almost achieved the same performance as the total 56 variables, indicating that the simplified scale was theoretically feasible and is expected to improve the efficiency of screening patients online. Similar to our study, Ren et al., 2021, Ren et al., 2021 collected data on the demographic characteristics of Chinese college students, such as gender, major, and grade, in addition to variables on personal views associated with COVID-19, and predicted the impact of COVID-19 on college students' mental health using a machine learning model (logistic regression). The results showed that the accuracy of 12 variables in predicting anxiety and depression were 81 % and 74 %, respectively. The AUC value for each model was >0.8, indicating that those models are as stable and reliable as ours. There were also some differences between two studies. First, the sample size of Ren's study was quite small compared to ours (478 vs 29,841), as mentioned in the study's limitations. Second, Ren's study included more variables only relevant to college students, such as internship status, examination scores, and school-related situations. In contrast, social support, resilience, and alexithymia can be applicable to a wider range of people, not just school students. The current study was cross-sectional, which is insufficient for providing evidence on the causal relationship between the selected features and the incidences of depression and anxiety. In a four-week longitudinal study, the authors found that the scores of the DASS-21 subscale in patients with depression and anxiety had no statistically significant change. Factors associated with higher DASS-21 subscale scores included physical symptoms, general health status, chronic medical history (Cw et al., 2020), but did not include social support, psychological resilience, or alexithymia. Therefore, additional longitudinal studies are needed to verify the role of these features in predicting the risks of depression and anxiety. Further, extensively retaining participant's data and synthesizing it for the machine learning model may introduce an unbalanced data distribution, which might be a potential factor affecting performance. Finally, the variables in the current study were limited. Studies also have shown that suicidal ideation and suicide attempts increased during the COVID-19 pandemic (Ec et al., 2020; Berardelli et al., 2021). Although in some countries, national prevention and control led to a decline in the suicide rate (McIntyre et al., 2021), it began to rise significantly after the lock-down (Montalbani et al., 2020). It is still unclear how suicide interacts with depression and COVID-19. Another study found that the willingness to vaccinate is related to the severity of depression and anxiety (Hao et al., 2021). In addition, some groups, such as adolescents (Ren et al., 2021, Ren et al., 2021) and pregnant women (Nguyen et al., 2022), have different effects for depression and anxiety. These factors, in addition to other factors, such as post-traumatic stress symptoms, social state, and the government's response to the mental health crisis during COVID-19 should also be measured and included in future studies examining the impact on mental health. The following is the supplementary data related to this article.Table S1 The description of definition of 19 selected features Table S1 Funding This work was supported by the 10.13039/501100004826 Beijing Natural Science Foundation Grant [grant number 7202072]; the 10.13039/501100009592 Beijing Municipal Science and Technology Commission Grant [grant number Z191100006619104]; and Beijing Hospitals Authority Ascent Plan [grant number DFL20192001]. CRediT authorship contribution statement Zhanxiao Tian: Investigation, Validation, Formal analysis, Data curation, Writing - original draft. Wei Qu: Software, Validation, Formal analysis, Data curation. Yanli Zhao: Formal analysis, Data curation. Xiaolin Zhu: Formal analysis, Data curation. Zhiren Wang:: Conceptualization, Methodology, Formal analysis. Yunlong Tan: Conceptualization, Resources, Project administration, Writing – review. Ronghuan Jiang: Resources, Project administration. Shuping Tan: Conceptualization, Methodology, Resources, Writing - review & editing, Supervision, Project administration, Funding acquisition. Uncited reference Schulz and Becker, 2014 Conflict 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. Acknowledgments We are grateful to all of the project participants and all of the individuals involved in the data collection. 1 The eXtreme Gradient Boosting. 2 Depression. 3 Anxiety. 4 Depression and anxiety comorbidity. 5 Health control. ==== Refs References Schulz Andrea Becker The impact of childhood trauma on depression: does resilience matter? Population-based results from the Study of Health in Pomerania J. Psychosom. Res. 77 2 2014 97 103 10.1016/j.jpsychores.2014.06.008 25077849 Jcp A. Ksd A. Dp B. Childhood adversity and adult depression: the protective role of psychological resilience Child Abuse Negl. 64 2017 89 100 10.1016/j.chiabu.2016.12.012 28056359 Alonso Debreczeni F. Bailey P.E. A systematic review and meta-analysis of subjective age and the association with cognition, subjective well-being, and depression J. Gerontol. B Psychol. Sci. Soc. Sci. 76 2021 471 482 10.1093/geronb/gbaa069 32453828 Ayako Hiyoshi Ruzan Stress resilience in adolescence and subsequent antidepressant and anxiolytic medication in middle aged men: Swedish cohort study Soc. Sci. Med. 134 2015 43 49 10.1016/j.socscimed.2015.03.057 Jun 25884415 Banerjee D. Rai M. Social isolation in Covid-19: the impact of loneliness Int. J. Soc. Psychiatry 66 6 2020 525 527 10.1177/0020764020922269 32349580 Bao Y. Sun Y. Meng S. Shi J. Lu L. 2019-nCoV epidemic: address mental health care to empower society Lancet 395 10224 2020 e37 e38 10.1016/S0140-6736(20)30309-3 32043982 Berardelli I. Sarubbi S. Rogante E. Cifrodelli M. Pompili M. The impact of the COVID-19 pandemic on suicide ideation and suicide attempts in psychiatric patients in Italy Psychiatry Res. 303 2021 114072 10.1016/j.psychres.2021.114072 2021 Sep Cao X. Yang C. Wang D. The impact on mental health of losing an only child and the influence of social support and resilience Omega(Westport) 80 4 2018 666 684 10.1177/0030222818755284 2020 Mar 29380659 Ceban F. Ling S. Lui L. Lee Y. Mcintyre R.S. Fatigue and cognitive impairment in post-COVID-19 syndrome: a systematic review and meta-analysis Brain Behav. Immun. 101 2021 93 135 10.1016/j.bbi.2021.12.020 2022 Mar 34973396 Ceban F. Nogo D. Carvalho I.P. Lee Y. Nasri F. Xiong J. Lui L. 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 Psychiatry 78 2021 1079 1091 10.1001/jamapsychiatry.2021.1818 34319365 Chang L.Y. Wu C.C. Yen L.L. Chang H.Y. The effects of family dysfunction trajectories during childhood and early adolescence on sleep quality during late adolescence: resilience as a mediator Soc. Sci. Med. 222 2019 162 170 10.1016/j.socscimed.2019.01.010 30641286 Chen T. Guestrin C. XGBoost: A Scalable Tree Boosting System. Knowledge Discovery and Data Mining 2016 ACM Choi E. Hui B. Wan E. Depression and anxiety in Hong Kong during COVID-19 Int. J. Environ. Res. Public Health 17 10 2020 3740 10.3390/ijerph17103740 32466251 Cw A. Rp A. Xw A. Yt A. Lx A. Mi B. Fnc E. Btc D. Rhe F. Vksh I. A longitudinal study on the mental health of general population during the COVID-19 epidemic in China Brain Behav. Immun. 87 2020 40 48 10.1016/j.bbi.2020.04.028 32298802 de Figueiredo C.S. Sandre P.C. Portugal L. COVID-19 pandemic impact on children and adolescents' mental health: biological, environmental, and social factors Prog. Neuro-Psychopharmacol. Biol. Psychiatry 106 2021 110171 10.1016/j.pnpbp.2020.110171 Dotson V.M. McClintock S.M. Verhaeghen P. Kim J.U. Draheim A.A. Syzmkowicz S.M. Gradone A.M. Bogoian H.R. Wit L. Depression and cognitive control across the lifespan: a systematic review and meta-analysis Neuropsychol. Rev. 30 2020 461 476 10.1007/s11065-020-09436-6 32385756 Ec A. Cdb B. Ac A. Fc A. Rn A. Mbc D. Mca B. Psychiatric emergency care during Coronavirus 2019 (COVID 19) pandemic lockdown: results from a Department of Mental Health and Addiction of northern Italy Psychiatry Res. 293 2020 113463 10.1016/j.psychres.2020.113463 2020 Nov 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 10.1056/NEJMsb2005114 32202722 Garabiles M.R. Lao C.K. Yip P. Chan E. Mordeno I. Hall B.J. Psychometric validation of PHQ-9 and GAD-7 in Filipino migrant domestic workers in Macao (SAR),China J. Pers. Assess. 102 6 2020 833 844 10.1080/00223891.2019.1644343 31361153 Gariepy G. Honkaniemi H. Quesnel-Vallee A. Social support and protection from depression: systematic review of current findings in Western countries Br. J. Psychiatry 209 4 2016 284 293 10.1192/bjp.bp.115.169094 2016 Oct 27445355 Garmezy N. Masten A.S. Stress, competence, and resilience: common frontiers for therapist and psychopathologist Behav. Ther. 17 5 1986 500 521 10.1016/S0005-7894(86)80091-0 Gong Y. Shi J. Ding H. Zhang M. Han J. Personality traits and depressive symptoms: the moderating and mediating effects of resilience in Chinese adolescents J. Affect. Disord. 265 2019 611 617 10.1016/j.jad.2019.11.102 31787420 Hao F. Tam W. Hu X. Tan W. Jiang L. Jiang X. Zhang L. Zhao X. Zou Y. Hu Y. Luo X. McIntyre R.S. Quek T. Tran B.X. Zhang Z. Pham H.Q. Ho C. Ho R. A quantitative and qualitative study on the neuropsychiatric sequelae of acutely ill COVID-19 inpatients in isolation facilities Transl. Psychiatry 10 1 2020 355 10.1038/s41398-020-01039-2 33077738 Hao F. Wang B. Tan W. Husain S.F. Mcintyre R.S. Tang X. Zhang L. Han X. Jiang L. Chew N.W.S. Attitudes toward COVID-19 vaccination and willingness to pay: comparison of people with and without mental disorders in China BJPsych Open 7 5 2021 e146 10.1192/bjo.2021.979 34422295 He X.Y. Li C.B. Qian J. Cui H.S. Wu W.Y. Reliability and validity of a generalized anxiety disorder scale in general hospital outpatients Shanghai Arch. Psychiatry 22 4 2010 200 203 10.3969/j.issn.1002-0829.2010.04.002 Ho C.S. Chee C.Y. Ho R.C. Mental health strategies to combat the psychological impact of COVID-19 beyond paranoia and panic Ann. Acad. Med. Singap. 49 3 2020 155 160 PMID:32200399 32200399 Hogeveen J. Grafman J. Alexithymia Handbook of Clinical Neurology 183 2021 47 62 10.1016/B978-0-12-822290-4.00004-9 34389125 Hosmer D. Lemeshow S. Model-building strategies and methods for logistic regression Applied Logistic Regression Second edition 2000 0-471-72214-6 Huang Y. Zhao N. Generalized anxiety disorder, depressive symptoms and sleep quality during COVID-19 outbreak in China: a web-based cross-sectional survey Psychiatry Res. 288 2020 112954 10.1016/j.psychres.2020.112954 Kanako N. Takamasa N. Kanako I. Tomoko O. Yuji T. Takashi Y. Kazuyuki N. Resilience and depression/anxiety symptoms in multiple sclerosis and neuromyelitis optica spectrum disorder Mult.Scler.Relat.Disord. 25 2018 309 315 10.1016/j.msard.2018.08.023 30176401 Kathryn M. Connor Jonathan R.T. Davidson Li-Ching Lee Spirituality, resilience, and anger in survivors of violent trauma: a community survey J. Trauma. Stress. 16 3 2003 487 494 10.1023/A:1025762512279 14584633 Koelmel E. Hughes A.J. Alschuler K.N. Ehde D.M. Resilience mediates the longitudinal relationships between social support and mental health outcomes in multiple sclerosis Arch. Phys. Med. Rehabil. 98 2016 1139 1148 10.1016/j.apmr.2016.09.127 27789238 Lazarus R.S. From psychological stress to the emotions: a history of changing outlooks Annu. Rev. Psychol. 44 1 1993 1 21 8434890 Le X. Dang A.K. Toweh J. Nguyen Q.N. Ho R. Evaluating the psychological impacts related to COVID-19 of Vietnamese people under the first nationwide partial lockdown in Vietnam Front.Psychiatry 11 2020 824 10.1146/annurev.ps.44.020193.000245 32982807 Lee Y. Lui L. Chen-Li D. Liao Y. Mansur R.B. Brietzke E. Rosenblat J.D. Ho R. Rodrigues N.B. Lipsitz O. Nasri F. Cao B. Subramaniapillai M. Gill H. Lu C. McIntyre R.S. Government response moderates the mental health impact of COVID-19: a systematic review and meta-analysis of depression outcomes across countries J. Affect. Disord. 290 2021 364 377 10.1016/j.jad.2021.04.050 34052584 Lenzo V. Barberis N. Cannavò M. Filastro A. Verrastro V. Quattropani M.C. The relationship between alexithymia, defense mechanisms, eating disorders, anxiety and depression Riv.Psichiatr. 55 2020 24 30 10.1708/3301.32715 32051622 Li W. Zhao Y.J. Zhang S.F. Yang B. Cheung T. Jackson T. Sha S. Xiang Y.T. Mapping post-traumatic stress disorder symptoms and quality of life among residents of Wuhan, China after the COVID-19 outbreak: a network perspective J. Affect. Disord. 318 2022 80 87 10.1016/j.jad.2022.08.074 36030998 Li Y. Long Z. Cao D. Cao F. Social support and depression across the perinatal period: a longitudinal study J. Clin. Nurs. 26 17-18 2017 2776 2783 10.1111/jocn.13817 28334472 Lisha Dai Yi Zhou Jing Hu. Yunlong Deng Effect of alexithymia on health anxiety: mediating role of cognition and meta-cognition J.Cent.South Univ.Med.Sci. 28 43(9) 2018 1026 1031 10.11817/j.issn.1672-7347.2018.09.015 Luo M. Guo L. Yu M. Jiang W. Wang H. The psychological and mental impact of coronavirus disease 2019 (COVID-19) on medical staff and general public - a systematic review and meta-analysis Psychiatry Res. 291 2020 113190 10.1016/j.psychres.2020.113190 Luo M. Guo L. Yu M. Wang H. The psychological and mental impact of coronavirus disease 2019 (COVID-19) on medical staff and general public: a systematic review and meta-analysis Psychiatry Res. 291 2020 113190 10.1016/j.psychres.2020.113190 Mcgillivray L. Becerra R. Harms C. Prevalence and demographic correlates of Alexithymia: a comparison between Australian psychiatric and community samples J. Clin. Psychol. 73 1 2017 76 83 10.1002/jclp.22314 27129142 McIntyre R.S. Lui L.M. Rosenblat J.D. Ho R. Gill H. Mansur R.B. Teopiz K. Liao Y. Lu C. Subramaniapillai M. Nasri F. Lee Y. Suicide reduction in Canada during the COVID-19 pandemic: lessons informing national prevention strategies for suicide reduction J. R. Soc. Med. 114 2021 473 479 10.1177/01410768211043186 34551280 Miao J. Wu X. Sun X. Neighborhood, social cohesion, and the Elderly's depression in Shanghai Soc. Sci. Med. 229 2019 134 143 10.1016/j.socscimed.2018.08.022 30194018 Minihan E. Gavin B. Kelly B.D. McNicholas F. COVID-19, mental health and psychological first aid Ir. J. Psychol. Med. 37 4 2020 259 263 10.1017/ipm.2020.41 32404221 Montalbani B. Bargagna P. Mastrangelo M. Sarubbi S. Comparelli A. The COVID-19 outbreak and subjects with mental disorders who presented to an Italian psychiatric emergency department J. Nerv. Ment. Dis. 209 4 2020 246 250 10.1097/NMD.0000000000001289 Morete M.C. Solano J. Boff M. Jacob-Filho W. Ashmawi H. Resilience, depression, and quality of life in elderly individuals with chronic pain followed up in an outpatient clinic in the city of São Paulo, Brazil J. Pain Res. 11 2018 2561 2566 10.2147/JPR.S166625 30464576 Nguyen L.H. Nguyen L.D. Ninh L.T. Nguyen H. Nguyen A.D. Dam V. Nguyen T.T. Do H.P. Vu T. Tran B.X. Latkin C.A. Ho C. Ho R. COVID-19 and delayed antenatal care impaired pregnant women's quality of life and psychological well-being: what supports should be provided? Evidence from Vietnam J. Affect. Disord. 298 2022 119 125 10.1016/j.jad.2021.10.102 34715160 Nr A. Kz A. MD B. St C. Sm D. Perinatal depression: the role of maternal adverse childhood experiences and social support J. Affect. Disord. 263 2020 576 581 10.1016/j.jad.2019.11.030 31759669 Palser E.R. Palmer C.E. Alejandro G.P. Ricci H. Aikaterini F. Kilner J.M. Elizabeth A.J. Alexithymia mediates the relationship between interoceptive sensibility and anxiety PloS one 13 2018 e0203212 10.1371/journal.pone.0203212 Panayiotou G. Panteli M. Vlemincx E. Adaptive and maladaptive emotion processing and regulation, and the case of alexithymia Cognit. Emot. 35 2021 488 499 0.1080/02699931.2019.1671322 31556808 Porcelli P. Sonino N. Psychological factors affecting medical conditions : a new classification for DSM-V Psychological Factors Affecting Medical Conditions: A New Classification for DSM-V.2007 186 2007 978-3-8055-8331-2 Preece D.A. Mehta A. Becerra R. Chen W. Allan A. Robinson K. Boyes M. Hasking P. Gross J.J. Why is alexithymia a risk factor for affective disorder symptoms? The role of emotion regulation J. Affect. Disord. 296 2022 337 341 10.1016/j.jad.2021.09.085 34606815 Ran L. Wang W. Ai M. Kong Y. Chen J. Kuang L. Psychological resilience, depression, anxiety, and somatization symptoms in response to COVID-19: a study of the general population in China at the peak of its epidemic Soc. Sci. Med. 262 2020 113261 10.1016/j.socscimed.2020.113261 Ren Z. Xin Y. Ge J. Zhao Z. Liu D. Ho R. Ho C. Psychological impact of COVID-19 on college students after school reopening: a cross-sectional study based on Mach. Learn. 12 2021 641806 10.3389/fpsyg.2021.641806 Ren Z. Xin Y. Wang Z. Liu D. Ho R. Ho C. What factors are most closely associated with mood disorders in adolescents during the COVID-19 pandemic? A cross-sectional study based on 1,771 adolescents in Shandong Province,China Front. Psychiatry 12 2021 728278 10.3389/fpsyt.2021.728278 Renaud-Charest O. Lui L. Eskander S. Ceban F. Ho R. Di Vincenzo J.D. Rosenblat J.D. Lee Y. Subramaniapillai M. McIntyre R.S. Onset and frequency of depression in post-COVID-19 syndrome: a systematic review J. Psychiatr. Res. 144 2021 129 137 10.1016/j.jpsychires.2021.09.054 34619491 Rothon C. Goodwin L. Stansfeld S. Family social support, community "social capital" and adolescents' mental health and educational outcomes: a longitudinal study in England Soc.Psychiatry Psychiatr.Epidemiol. 47 2012 697 709 21557090 Sarason I.G. Levine H.M. Basham R.B. Sarason B.R. Assessing social support: the social support questionnaire J. Pers. Soc. Psychol. 44 1983 127 139 10.1037/0022-3514.44.1.127 Schwarzbach M. Luppa M. Forstmeier S. König H.H. Riedel-Heller S.G. Social relations and depression in late life-a systematic review Int.J.Geriatr.Psychiatry 29 1 2014 1 21 10.1002/gps.3971 23720299 Shader R.I. COVID-19 and depression Clin. Ther. 42 6 2020 962 963 10.1016/j.clinthera.2020.04.010 32362345 Sheerin C.M. Lind M.J. Brown E.A. Gardner C.O. Kendler K.S. Amstadter A.B. The impact of resilience and subsequent stressful life events on MDD and GAD Depress. Anxiety 35 2 2018 140 147 10.1002/da.22700 29172241 Shen W. A tangled web: the reciprocal relationship between depression and educational outcomes in China Soc. Sci. Res. 85 2020 102353 10.1016/j.ssresearch.2019.102353 Tee M.L. Tee C.A. Anlacan J.P. Aligam K. Ho R.C. Psychological impact of COVID-19 pandemic in the Philippines J. Affect. Disord. 277 2020 379 391 10.1016/j.jad.2020.08.043 32861839 Tomás C.C. Oliveira E. Sousa D. Ubachupel M. Furtado G. Rocha C. Teixeira A. Ferreira P. Alves C. Gisin S. Proceedings of the 3rd IPLeiria's international health congress BMC Health Serv. Res. 16 Suppl 3 Suppl. 3 2016 200 10.1186/s12913-016-1423-5 27409075 Tusaie K. Dyer J. Resilience: a historical review of the construct Holist. Nurs. Pract. 18 1 2004 3 8 10.1097/00004650-200401000-00002 14765686 Wang C. Horby P.W. Hayden F.G. Gao G.F. A novel coronavirus outbreak of global health concern Lancet 395 10223 2020 470 473 10.1016/S0140-6736(20)30185-9 31986257 Wang C. Chudzicka-Czupaa A.E. Grabowski D. Pan R. Ho C. The association between physical and mental health and face mask use during the COVID-19 pandemic: a comparison of two countries with different views and practices Front.Psychiatry 11 2020 569981 10.3389/fpsyt.2020.569981 Wang C. Fardin M.A. Shirazi M. Pan R. Ho R. Mental health of the general population during the 2019 coronavirus disease (COVID-19) pandemic: a tale of two developing countries PsychiatryInt. 2 1 2021 71 84 10.3390/psychiatryint2010006 Wang C. López-Núez M.I. Pan R. Wan X. García M. The impact of 2019 coronavirus disease (COVID-19) pandemic on physical and mental health: a comparison between China and Spain: cross-sectional study JMIR Formative Res. 2021 5 10.2196/27818 Wang C. Pan R. Wan X. Tan Y. Xu L. Ho C.S. Ho R.C. Immediate psychological responses and associated factors during the initial stage of the 2019 coronavirus disease (COVID-19) epidemic among the general population in China Int. J. Environ. Res. Public Health 17 5 2020 1729 10.3390/ijerph17051729 32155789 Wang C. Tripp C. Sears S.F. Xu L. Ho R. The impact of the COVID-19 pandemic on physical and mental health in the two largest economies in the world: a comparison between the United States and China J. Behav. Med. 2021 Wang M. Yin D. Liu L. Zhou S. Liu Q. Tian H. Wei J. Zhang K. Wang G. Chen Q. Zhu G. Wang X. Si T. Yu X. Lv X. Zhang N. Features of cognitive impairment and related risk factors in patients with major depressive disorder: a case-control study J. Affect. Disord. 307 2022 29 36 10.1016/j.jad.2022.03.063 35358550 Wang W. Bian Q. Zhao Y. Li X. Wang W. Du J. Zhang G. Zhou Q. Zhao M. Reliability and validity of the Chinese version of the patient health questionnaire (PHQ-9) in the general population Gen. Hosp. Psychiatry 36 2014 539 544 10.1016/j.genhosppsych.2014.05.021 25023953 Werner E.E. Resilience in development Curr. Dir. Psychol. Sci. 4 3 1995 81 10.1111/1467-8721.EP10772327 Windle G. Bennett K.M. Noyes J. A methodological review of resilience measurement scales Health Qual. Life Outcomes 9 2011 1 18 10.1186/1477-7525-9-8 21223594 Ye Z.J. Qiu H.Z. Li P.F. Validation and application of the Chinese version of the 10-item Connor-Davidson resilience scale (CD-RISC-10) among parents of children with cancer diagnosis Eur. J. Oncol. Nurs. 27 2017 36 44 10.1016/j.ejon.2017.01.004 28279394 Yu X. Zhang J. Yu X.N. Zhang J.X. Factor analysis and psychometric evaluation of the Connor-Davidson Resilience Scale (CD-RISC) with Chinese people Soc.Behav.Personal.Int.J. 35 2007 19 30 10.2224/sbp.2007.35.1.19 Zhang M. Ho R. Moodle: the cost effective solution for internet cognitive behavioral therapy (I-CBT) interventions Technol.Health Care 25 2017 163 165 10.3233/THC-161261 27689560 Zhen L. Wang G. Xu G. Xiao L. Feng L. Chen X. Liu M. Zhu X. Evaluation of the paper and smartphone versions of the quick inventory of depressive symptomatology-self-report (QIDS-SR16) and the patient health questionnaire-9 (PHQ-9) in depressed patients in China Neuropsychiatr. Dis. Treat. 16 2020 993 1001 10.2147/NDT.S241766 32368061	36462608	PMC9710109	NO-CC CODE	2022-12-08 23:16:23	no	J Affect Disord. 2023 Feb 15; 323:417-425	utf-8	J Affect Disord	2,022	10.1016/j.jad.2022.11.044	oa_other
==== Front Int J Pharm Int J Pharm International Journal of Pharmaceutics 0378-5173 1873-3476 The Author(s). Published by Elsevier B.V. S0378-5173(22)01018-3 10.1016/j.ijpharm.2022.122463 122463 Article Novel trehalose-based excipients for stabilizing nebulized anti-SARS-CoV-2 antibody Noverraz François a1 Robin Baptiste c1 Passemard Solène d Fauvel Bénédicte e Presumey Jessy e Rigal Emilie e Cookson Alan f Chopineau Joël b Martineau Pierre g Villalba Martin h Jorgensen Christian h Aubert-Pouëssel Anne b Morille Marie b⁎ Gerber-Lemaire Sandrine a⁎ a Group for Functionalized Biomaterials, Institute of Chemical Sciences and Engineering Ecole Polytechnique Fédérale de Lausanne, EPFL SB ISIC SCI-SB-SG, Station 6, CH-1015 Lausanne, Switzerland b ICGM, Univ Montpellier, CNRS, ENSCM, Montpellier, France c MedXCell Science, Bâtiment Cyborg 1 (IRMB), Hôpital Saint-Eloi, 80 avenue Augustin Fliche, 34295 Montpellier, France d Montpellier Life Science Bâtiment Cyborg 1 (IRMB), Hôpital Saint-Eloi, 80 avenue Augustin Fliche, 34295 Montpellier, France e CYTEA BIO, Bâtiment Cyborg 1 (IRMB), Hôpital Saint-Eloi, 80 avenue Augustin Fliche, 34295 Montpellier, France f MedXCell SA, Av. des Planches 20C, 1820 Montreux, Suisse g IRCM, Univ Montpellier, INSERM, ICM, Montpellier, France h IRMB, Univ Montpellier, INSERM, CNRS, CHU Montpellier, Montpellier, France ⁎ Corresponding authors. 1 contributed equally. 30 11 2022 5 1 2023 30 11 2022 630 122463122463 19 10 2022 23 11 2022 27 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. Graphical abstract COVID-19 is caused by the infection of the lungs by SARS-CoV-2. Monoclonal antibodies, such as sotrovimab, showed great efficiency in neutralizing the virus before its internalization by lung epithelial cells. However, parenteral routes are still the preferred route of administration, even for local infections, which requires injection of high doses of antibody to reach efficacious concentrations in the lungs. Lung administration of antibodies would be more relevant requiring lower doses, thus reducing the costs and the side effects. But aerosolization of therapeutic proteins is very challenging, as the different processes available are harsh and trigger protein aggregation and conformational changes. This decreases the efficiency of the treatment, and can increase its immunogenicity. To address those issues, we developed a series of new excipients composed of a trehalose core, a succinyl side chain and a hydrophobic carbon chain (from 8 to 16 carbons). Succinylation increased the solubility of the excipients, allowing their use at relevant concentrations for protein stabilization. In particular, the excipient with 16 carbons (C16TreSuc) used at 5.6 mM was able to preserve colloidal stability and antigen-binding ability of sotrovimab during the nebulization process. It could also be used as a cryoprotectant, allowing storage of sotrovimab in a lyophilized form during weeks. Finally, we demonstrated that C16TreSuc could be used as an excipient to stabilize antibodies for the treatment against COVID-19, by in vitro and in vivo assays. The presence of C16TreSuc during nebulization preserved the neutralization capacity of sotrovimab against SARS-CoV-2 in vitro; an increase of its efficacy was even observed, compared to the non-nebulized control. The in vivo study also showed the wide distribution of sotrovimab in mice lungs, after nebulization with 5.6 mM of excipient. This work brings a solution to stabilize therapeutic proteins during storage and nebulization, making pulmonary immunotherapy possible in the treatment of COVID-19 and other lung diseases. Keywords Protein delivery Lung disease Nebulization Antibody formulation Stabilizing excipients Succinylated trehalose ==== Body pmc1 Introduction Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causative virus infects epithelial cells in the respiratory tract, which makes it a worldwide major burden causing coronavirus disease-19 (COVID-19) with severe inflammatory conditions to the lungs (Carsana et al., 2020). The viral membrane Spike protein is responsible for cell entry by interacting with the angiotensin-converting enzyme ACE2 present on the pneumocyte membrane (Song et al., 2018). To fight against the infection, anti-SARS-CoV-2 treatments based on monoclonal antibodies (mAbs) have gained considerable interest, as proteins often show high efficacy, and few side effects, thanks to their strong specificity towards their target (Leader et al., 2008). Some of them are able to neutralize SARS-CoV-2 virus before their entry into pneumocytes. In particular, S309 - an antibody isolated from a patient infected with SARS-CoV-1 - showed promising results against SARS-CoV-2 in prophylactic and therapeutic settings in animal models (Corti et al., 2021, Pinto et al., 2020). This antibody was optimized to obtain sotrovimab, which was marketed and administered by parenteral route for the treatment of patients being at high risk of developing a severe form of COVID-19 (EMA, 2021). As far as therapeutic proteins on the market are concerned, the preferred routes of administration are parenteral because it maximizes bioavailability, as well as rapidity of action (Respaud et al., 2015). However, in the context of a pulmonary infection such as COVID-19, the passage of mAbs from the blood to the lung is very limited, requiring injection of very high doses of antibodies to obtain an effective concentration in the lungs (Roifman et al., 1987). The delivery of antibody drugs directly to the site of action via the pulmonary route is therefore an attractive alternative (Respaud et al., 2015). Antibody formulations developed for parenteral administration may not easily be repurposed for inhalation delivery. Among key parameters that must specifically be addressed, the formulation must preserve the integrity of inhaled proteins when they undergo mechanical and/or thermal stresses during aerosolization (Respaud et al., 2014). These different stresses can indeed trigger aggregation, conformation changes, or even cleavage of the proteins, resulting in a loss of efficacy and an increase in immunogenicity (Cui et al., 2017). The amphiphilic properties of proteins can trigger changes in conformation, after their adsorption to an interface (Lee et al., 2011). As the bulk liquid is nebulized, the surface is dramatically increased, hence enhancing the risk of protein unfolding and aggregation following their adsorption at the air–liquid interface. A commonly used method to prevent interfacial damage on proteins is the use of nonionic surfactants. Among them, polysorbates (PS 20 and PS 80), widely used for parenteral administration of antibodies, have been successfully included in formulations for inhalation delivery of mAbs and proteins (Hertel et al., 2015, Respaud et al., 2014). However, several limitations of these excipients have been reported. First, polysorbates impair long-term storage due to the presence of peroxide residuals causing protein oxidation and enhancing protein degradation (Ha et al., 2002). Finally, it contains poly(ethylene glycol) (PEG), which is now known to provoke side effects due to the polymer itself after administration (specific and/or non-specific immunological response) (Guerrini et al., 2022, Hamad et al., 2008, Stone et al., 2019, Szebeni et al., 2011, Turecek et al., 2016, Yang et al., 2016). Finally, even if a number of protein therapeutics have been stably nebulized as part of chronic treatment regimens, to date, there has been no report on stable nebulization of a fully human mAb that has been advanced through late-stage clinical trials (Fröhlich and Salar-Behzadi, 2021). The challenges rely not only on the selection of appropriated excipients to be added in the formulation to stabilize mAb during nebulization, but also on the paucity of toxicity data on inhaled excipients and their potential impact on formulation properties, and thereby device performances (Liang et al., 2020, Mayor et al., 2021). Consequently, the nature and the concentration of these stabilisers should be kept to the minimum possible limit to reduce potential side effects (Montharu et al., 2010, Respaud et al., 2015). In this framework, sugar-based surfactants are a good alternative to polysorbate surfactants. They display low toxicity, good biodegradability, good biocompatibility and are produced from cheaper and renewable sources (Naughton et al., 2019). While pharmaceutical, food, cosmetic, textile, oil, and agricultural applications manifest great interest in using biosurfactants for their benefits, their uses in pharmaceutical industries are limited due to the risk of endotoxin contamination when they are synthesized by microorganisms (Kügler et al., 2014, Kuyukina et al., 2015). Moreover, this mean of production results in glycolipids mixtures with different properties leading to inconsistent and unclear results (Ismail et al., 2021). Nevertheless, sugar-based surfactants have been recently synthetically produced circumventing those limitations. For instance, trehalose lipids that are constituted of trehalose attached through an ester bond to a lipid at various positions have been successfully synthesized (Bird et al., 2018, Jana et al., 2017, Jana et al., 2020, Kale and Akamanchi, 2016, Schiefelbein, 2010). Thus, this work focused on the synthesis and formulation of surfactants based on natural and safe materials, trehalose, fatty acids, and succinate functionalities, for the stabilization of antibodies during nebulization. Trehalose is a two glucose units nonreducing disaccharide found in many plants, microorganisms, and animals. It is involved in many metabolism pathways and has been recognized to be very efficient for the protection and stabilization of cellular membranes and proteins through different mechanisms (Elbein et al., 2003, Jain and Roy, 2009, Lerbret et al., 2007, Lins et al., 2004, Moiset et al., 2014, Paul and Paul, 2015, Sudrik et al., 2017). This disaccharide is widely approved in biomolecules formulations for pharmaceutical development (Ohtake and Wang, 2011). Yet, to our knowledge, a trehalose fatty acid derivative with or without succinyl functionalities has never been used for the delivery of antibodies to the lung through nebulization. During this study, trehalose-based excipients of different hydrophobic chain lengths were synthesized, and a succinyl moiety was added on the trehalose polar head. The ability of these excipients to preserve protein colloidal stability and binding ability was assessed after freeze-drying and/or nebulization. Biological in vitro efficacy and preliminary in vivo biodistribution studies were also performed. The proof of concept was carried out with sotrovimab, nebulized with a commercially available mesh nebulizer. 2 Material and methods 2.1 Material D-(+)-Trehalose anhydrous, Succinic anhydride (TCI Europe), O-(Benzotriazol-1-yl)-N,N,N’,N’-tetramethyluronium-tetrafluorborat (TBTU) (Carl Roth GmbH), Pyridine extra dry over molecular sieve (Thermo Scientific), Benzyl alcohol (Fluka), Myristic acid, Decanoic acid (Acros Organics), Octanoic acid, Palmitic acid (Fluorochem) were used as received without further purification. For antibody formulation, sotrovimab antibody (PM = 146 kDa, pI = 8.42, monomer percentage = 97.1 %, sequence given in Supplementary Information S3) was synthesized by Evitria (Switzerland). Trehalose dihydrate and trehalose surfactants (C8Tre, C10Tre, C14Tre and C16Tre) were provided by Sigma-Aldrich (Merck, USA). Eumulgin SMO 20 (better known as polysorbate 80, or PS80) was purchased to BASF (Germany). PBS was provided by ThermoFisher Scientific (USA). Water was purified (resistivity of 18.2 MΩ.cm at 25 °C) using a MilliQ Direct Type 1 Ultrapure water system (Merck, USA). For cytotoxicity, NIH-3 T3 cells were provided by American Type Culture Collection (ATCC) (USA). Cell viability assay (CellTiter 96® Aqueous Non-Radioactive Cell Proliferation Assay) was purchased from Promega (USA). For antigen binding assays, SARS-CoV-2 Spike expressing HEK293 cells, blasticidin and normocin were provided by InvivoGen (USA), while Cohesion Bioscience (UK) provided mouse anti-human IgG antibody labelled with Alexa Fluor 488. For in vitro neutralization assays, Vero E6 cells were provided by ECACC (UK), while SARS-CoV-2 viral strain n°2020A00935-34 (described by CPP Ile-de-France III) was from the CRB collection of Centre Hospitalier Universitaire de Montpellier (France). Control neutralizing antibodies (rabbit antibody, reference 40592-R001) were bought from Sino Biological (China), and Viral ToxGlo™ Assay was purchased from Promega (USA). Common items such as RPMI and DMEM culture medium, FBS, penicillin, streptomycin and glutamine were purchased from ThermoFisher Scientific (USA). For in vivo lung distribution studies, C57BL/6 wild type male mice were provided by Janvier labs (France). Isoflurane, ketamine and xilazine were purchased from Virbac (France), while PBS solution with PFA 4 % was purchased from Thermo Fisher Scientifics (USA), as well as Normal Fish Serum. Sucrose, Triton X-100 and Rabbit anti-human IgG were purchased from Sigma-Aldrich (USA). Secondary goat anti-rabbit coupled with Alexa Fluor 488 were purchased from Abcam (UK), Optimal Cutting Temperature compound NEG-50 MM was provided by MM Microm Microtech (France), Dako fluorescent mounting medium was provided by Agilent (USA), and ELISA kit for human IgG quantification was provided by BioRad (USA). 2.2 Characterization of trehalose derivatives 2.2.1 Nuclear magnetic resonance (NMR) NMR spectra were recorded on Bruker Avance III-400 or Bruker Avance NEO spectrometers at room temperature (rt) (400 MHz) (Bruker, Billerica, MA, USA). 1H frequency is at 400.13 MHz, 13C frequency is at 100.62 MHz. Chemical shifts are expressed in parts per million (ppm) and coupling constants (J) in hertz (Hz). Solvent used for NMR spectroscopy was deuterated methanol (CD3OD, Eurisotop). 2.2.2 Mass spectrometry Mass spectrometry analyses were performed either on:LTQ Orbitrap FTMS instrument (LTQ Orbitrap Elite FTMS, Thermo Scientific, Bremen, Germany) operated in the positive ionization mode coupled with a robotic chip-based nano-ESI source (TriVersa Nanomate, Advion Biosciences, Ithaca, NY, U.S.A.). A standard data acquisition and instrument control system was utilized (Thermo Scientific) whereas the ion source was controlled by Chipsoft 8.3.1 software (Advion BioScience). Samples were loaded onto a 96-well plate (Eppendorf, Hamburg, Germany) within an injection volume of 5 µL. The experimental conditions for the ionization voltage was + 1.4 kV and the gas pressure was set at 0.30 psi. The temperature of ion transfer capillary was 275 °C. FTMS spectra were obtained in the 100–1200 m/z range in the reduce profile mode with a resolution set to 120,000. In all spectra one microscan was acquired with a maximum injection time value of 1000 ms. Xevo G2-S QTOF mass spectrometer coupled to the Acquity UPLC Class Binary Solvent manager and BTN sample manager (Waters, Corporation, Milford, MA). The sample manager system temperature was maintained at 10 °C and the injection volume was 5 μL. Mass spectrometer detection was operated in positive ionization using the Zspray™ dual-orthogonal multimode ESI/APCI/ESCi® source. The TOF mass spectra were acquired in the resolution mode over the range of m/z 50–1200 at an acquisition rate of 0.036 sec/spectra. The instrument was calibrated using a solution of sodium formate (0.01 mg/L in isopropanol/H2O 90:10). A mass accuracy better than 5 ppm was achieved using a Leucine Enkephalin solution as lock-mass (200 pg/mL in ACN/H2O (50:50)) infused continuously using the LockSpray source. Source settings were as follows: cone, 25 V; capillary, 3 kV, source temperature, 150 °C; desolvation temperature, 500 °C, cone gas, 10 L/h, desolvation gas, 500 L/h. Data were processed using MassLynx™ 4.1 software. Waters Acquity-I-UPLC Classsystem (Waters Corporation, Milford, MA, USA) coupled with a Waters Vion IMs-QTof Mass Spectrometer equipped with LockSpray (Leucine-enkephalin (200 pg/µL). Nitrogen was used as collision gas. The instrument was controlled by Waters UNIFI 1.9.4 (3.1.0, Waters Corporation, Milford, MA, USA). Injection volume was 5 µL in bypass mode with a flow rate was 0.1 mL/min. The instrument was operated in positive polarity, sensitivity mode (33,000 FWHM at 556.2766 m/z). Data was acquired in HDMSe mode with a scan time of 1 s. The recorded mass range was from 50 to 1200 m/z for both low and high energy spectra. The collision energy was ramped from 20 to 40 V. The cone voltage was set to 30 V, capillary voltage was set to 2 kV and source offset was set to 50 V. Source temperature was set to 120 ◦C and desolvation temperature set to 500 °C. Cone gas flow rate was set to 50 L/h and desolvation gas flow rate was set to 1000 L/h. 2.2.3 Fourier-transform infrared spectroscopy (FTIR) IR spectra were recorded on a PerkinElmer Frontier FT/IR spectrometer outfitted with a QUEST ATR accessory as neat films compressed onto a Zinc Selenide window. The spectra are reported in cm−1; s, strong; m, medium, w, weak. 2.2.4 Critical micelle concentration (CMC) Surfactants solutions in distilled water were made via serial dilution in the concentration range between 1 µM and 100 mM. The solutions were analyzed by dynamic light scattering (DLS) on a Zetasizer Nano ZS (Malvern Panalytical, UK). Each dilution was scanned 3 times at 25 °C. The micellar formation was measured at ∼ 10 nm. The recorded scattering intensity (Derived Count Rate expressed in kilo counts per second [kpcs]) was used to describe the quantity of micelles in solution. Critical micelle concentrations were determined by plotting the log of kpcs versus the log of surfactant concentration and fitting to two linear regressions at high and low concentration. The intersection of the two linear regressions correspond to the CMC. 2.2.5 In vitro cytoxicity NIH-3T3 cell line was cultured at 37 °C in a 5 % CO2 atmosphere, in DMEM growth medium, supplemented with 100 U/mL penicillin and 100 μg/mL streptomycin, and 10 % of fetal bovine serum. In vitro cytotoxicity of trehalose-based surfactants was then assessed, following a standardized protocol described in normative documents (International Organization of Standardization, 2009) and publications (Arechabala et al., 1999). In a 96-well plate, 5.103 cells were deposited in each well, and left for 1 day at 37 °C in a 5 % CO2 atmosphere. Surfactant solutions ranging from 0.0112 to 11.2 mM were prepared by dilution of the stock solution into the cell culture medium. Those diluted solutions were then deposited to the wells, and after 48 h of incubation a CellTiter 96® Aqueous Non-Radioactive Cell Proliferation Assay (Promega, Wisconsin, USA) was used to evaluate the cell viability following manufacturer’s instructions. The absorbance was then recorded at 490 nm using Multiskan™ GO Microplate Spectrophotometer (Thermo Scientific™, Waltham, Massachusetts, USA). Mean triplicate values of cell viability were plotted against the surfactant concentration, and the half-maximal effective concentration (EC50) was determined by non-linear regression, using a “sigmoidal dose–response with variable hill slope” (the bottom was constraint to a constant value of 0 and the top to 100). This analysis was performed with the software GraphPad Prism 9.2.0 (GraphPad Software, USA). 2.3 Synthesis protocols All reactions were performed under argon atmosphere (1 atm). Glassware was dried for 12 hr in an oven (T greater than 100 °C) or flamed dried under vacuum. Reactions were monitored by TLC (Merck silica gel 60F254 plates, Merck, Darmstadt, Germany). Detection was performed by KMnO4 stain. Purifications were performed by flash chromatography on silica gel (SiliCycle SiliaFlash P60, 230–400 mesh) or C18-reversed phase silica gel (Merck 60757, 230–400 mesh). Solvent used for flash column chromatography elution: Dichloromethane and Methanol (Thommen-Furler AG). In this section are described the characterizations of C16TreSuc and its intermediates. Others derivatives characterizations (CxxTre, CxxTreSucBn and CxxTreSuc for xx = 8, 12, 14, 16) are reported in detail in Supplementary Information S5-S23. 2.4 General procedure for the preparation of CxxTre (xx = 8, 12, 14, 16) In a flame-dried round-bottomed flask equipped with a magnetic stir bar, the corresponding fatty acid (1.1 eq.) and TBTU (1.1 eq.) were dissolved in anhydrous pyridine (2 mL / 0.1 g of Trehalose). Trehalose (1 eq.) was then poured into the reaction mixture, and stirring was continued at r.t. for 96 hr under argon atmosphere. Pyridine was then removed under vacuum, and the resulting residue was purified by flash column chromatography using a solvent gradient of 5–20 % methanol in EtOAc:DCM (1:1), yielding the intermediate CxxTre mono esters as white solids (yield 30–65 %). Characterization data for Trehalose 6-hexadecanoate (C16Tre): 1H NMR (400 MHz, MeOD)δ 5.10 (dd, J = 8.7, 3.7 Hz, 2H, CH e, e’), 4.38 (dd, J = 11.9, 2.2 Hz, 1H, OOC-CH2′), 4.21 (dd, J = 11.9, 5.1 Hz, 1H, OOC-CH2′’), 4.03 (ddd, J = 10.2, 5.1, 2.1 Hz, 1H, CH a), 3.88 – 3.75 (m, 4H, CH c, c’, a’, a’–CH2′), 3.69 (dd, J = 12.0, 5.6 Hz, 1H, a’–CH2′’), 3.49 (dt, J = 9.8, 3.8 Hz, 2H, CH d, d’), 3.39 – 3.34 (m, 2H, CH b, b’), 2.36 (t, J = 7.4 Hz, 2H, [CH2]–CH2COO), 1.62 (m, 2H, [CH2]–CH2-CH2COO-), 1.31 (s, 24H, -[CH2]-), 0.96 – 0.88 (m, 3H, CH3). HRMS (ESI/QTOF) m/z: [M + Na]+ Calcd for C28H52NaO12 + 603.3351; Found 603.3356. 2.5 General procedure for preparation of CxxTreSucBn (xx = 8, 12, 14, 16) In a flame-dried round-bottomed flask equipped with a magnetic stir bar, succinic acid monobenzyl ester (SAMBE, for synthesis procedure, see Supplementary Information S4) (1.1 eq.) and TBTU (1.1 eq.) were dissolved in anhydrous pyridine (2 mL / 0.1 g of CxxTre). The corresponding intermediate CxxTre (1 eq.) was poured into the reaction mixture, and stirring was continued at r.t. for 96 hr under argon atmosphere. Pyridine was then removed under vacuum, and the resulting residue was purified by flash column chromatography using a solvent gradient of 5–20 % methanol in EtOAc:DCM (1:1), yielding the intermediate CxxTreSucBn as white solids (yield 30–40 %). Characterization data for Trehalose 6-hexadecanoate 6′-benzylsuccinate (C16TreSucBn): 1H NMR (400 MHz, MeOD)δ 7.40 – 7.28 (m, 5H, ar.), 5.15 (d, J = 1.6 Hz, 2H, CH2-ar.), 5.07 (dd, J = 8.8, 3.7 Hz, 2H, CH e, e’), 4.39 (ddd, J = 11.9, 6.6, 2.2 Hz, 2H, a-CH2′, a’–CH2′), 4.26 – 4.17 (m, 2H, a-CH2′’, a’–CH2′’), 4.04 (tt, J = 5.5, 2.7 Hz, 2H, CH a, a’), 3.80 (td, J = 9.3, 1.7 Hz, 2H, CH c, c’), 3.50 (dt, J = 9.8, 3.4 Hz, 2H, CH d, d’), 3.40 – 3.36 (m, 2H, CH b, b’), 2.70 (d, J = 1.9 Hz, 4H, OOC-CH2-CH2-COO), 2.34 (t, J = 7.4 Hz, 2H, [CH2]–CH2COO), 1.66 –1.58 (m, 2H, [CH2]–CH2-CH2COO), 1.30 (s, 24H, -[CH2]-), 0.92 (t, J = 6.8 Hz, 3H, CH3). 13C NMR (101 MHz, MeOD) δ 172.5, 128.12, 127.76, 93.60, 73.14, 72.38, 71.84, 70.54, 66.11,33.41, 31.67–28.96, 28.78, 24.63, 22.33, 13.03. HRMS (ESI/QTOF) m/z: [M + Na]+ Calcd for C39H62NaO15 + 793.3981; Found 793.3979. 2.6 General procedure for the preparation of CxxTreSuc (xx = 8, 12, 14, 16) The intermediate CxxTreSucBn (1 eq.) was dissolved in MeOH (1 mL / 0.1 g of intermediate) in a 3 necks flask and the solution was placed under argon atmosphere before adding 10 % Pd/C (0.1 eq.). Argon was removed from the reactor and the reaction mixture was back flushed with hydrogen (1 atm, 3 times). The reaction was vigorously stirred at r.t for 4 hr (monitored by TLC) under hydrogen atmosphere. After completion of the reaction, the solution was filtered through a Celite pad and the filter cake was washed with MeOH. The filtrate was concentrated under reduced pressure and the residue was purified by reversed phase flash chromatography (H2O:MeOH 20:80 with 0.1 % TFA). The fractions corresponding to the product were collected, concentrated in vacuo and lyophilized to yield the desired ionic surfactants as white solids (yield 58 – 64 %). Characterization data for Trehalose 6-hexadecanoate 6′-succinate (C16TreSuc): 1H NMR (400 MHz, MeOD)δ 5.06 (dd, J = 5.1, 3.8 Hz, 2H, CH e, e’), 4.38 (ddd, J = 16.3, 11.8, 2.1 Hz, 2H, a-CH2′, a’–CH2′), 4.21 (dt, J = 11.8, 4.8 Hz, 2H, a-CH2′’, a’–CH2′’), 4.04 (ddd, J = 10.1, 4.3, 2.2 Hz, 2H, CH a, a’), 3.79 (td, J = 9.3, 2.6 Hz, 2H, CH c, c’), 3.50 (dt, J = 9.8, 3.8 Hz, 2H, CH d, d’), 3.39 – 3.34 (m, 2H, CH b, b’), 2.72 – 2.56 (m, 4H, 2 × OOC-CH2), 2.35 (t, J = 7.4 Hz, 2H, [CH2]–CH2COO), 1.61 (q, J = 7.2 Hz, 2H, [CH2]–CH2-CH2COO-), 1.29 (s, 24H, -[CH2]-), 0.90 (t, J = 6.8 Hz, 3H, CH3). 13C NMR (101 MHz, MeOD) δ 174.16, 173.18, 93.98, 73.09, 72.98, 71.72, 71.61, 70.49, 70.19,70.07, 63.44, 63.01, 33.65, 31.67–29.03, 31.67–29.00, 28.81, 24.66, 22.33. HRMS (ESI/QTOF) m/z: [M + Na]+ Calcd for C32H56NaO15 + 703.3511; Found 703.3497. 2.7 Preparation of antibody solution The surfactant was weighed and dissolved in PBS, whose pH was adjusted to 5.8 by use of the suitable amount of HCl 0.67 M prior the dissolution. The solution of surfactant was then mixed with the solution of antibody 6.3 mg/mL, for a final concentration of antibody set at 1 mg/mL. 2.8 Nebulization and retrieving of antibody solution A volume of 500 µL of antibody solution was placed in the atomizing cup of a vibrating mesh nebulizer NEB-001, provided by Briutcare (China), and a 15-mL tube was fixed at the exit of the nozzle. The nebulizer was started, and the nebulization rate was set at 0.4 mL/min (“medium rate”). After the whole volume was nebulized, the aerosol was left to coalesce and fall to the bottom of the tube during 2 min. Then the solution could be retrieved and antibody stability could be assessed. 2.9 Freeze-drying of antibody solution A storage stability study was performed on freeze-dried antibody. Shortly, antibody solutions were frozen by dipping them in liquid nitrogen, and then placed in a freeze-drier (Heto Powerdry LL3000, Thermo Fisher Scientific, US) overnight. The tubes were then hermetically sealed and kept at 4 °C until further stability assays. 2.10 Stability assays 2.10.1 Aggregation monitoring The solutions were analyzed by dynamic light scattering (DLS) on a Malvern Zetasizer Nano ZS (Malvern Panalytical, UK) operating at 25 °C with a 633-nm He-Ne laser and 173° scattering angle. A sample volume of 75 µL was transferred undiluted to a polymethyl methacrylate sub-microcuvette. ZetaSizer software was fed with the characteristics of particle material (“protein”, refraction index = 1.45 and absorption = 0.01) and dispersant (water, viscosity = 0.8872 cP and refraction index = 1.330). The data processing setting was set as “General Purposes”. Each sample was analyzed three times, to determine the total area under curve (AUC) of the peaks corresponding to antibody aggregates on the non-negative least-square distribution. All the peaks whose distribution mean was higher than the one of the monomer peak (around 10 nm) were considered to represent aggregates. 2.10.2 Binding efficiency by antigen recognition assays The stability of the antibody was evaluated by studying its ability to bind antigen expressed at the surface of a stably transfected cell line. Shortly, SARS-CoV-2 spike (D614)-expressing Human Embryonic Kidney (HEK) cells were cultured in RPMI medium, with 10 % of fetal bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin. Blasticidin (10 µg/mL) was added to select the transfected cells only. After distributing 250 000 cells per well in a 96-well plate, cells were incubated with 50 µL of formulations at 500 ng/ml of antibody, for 15 min at 4 °C. Cells were washed in PBS, and Alexa Fluor 488-conjugated mouse monoclonal anti-human Fc IgG (Cohesion Bioscience) diluted at 1/500 in PBS was added. After 15 min of incubation in the dark, at 4 °C, cells were washed again and analyzed by flow cytometry using FACS CANTO II (BD Bioscience, USA). Each condition was done in triplicate. Data treatment was then performed using FlowJo software. Live cells were gated, and the percentage of fluorescent-labelled cells (i.e. cells interacting with the antibody of interest) was estimated among those latter. A negative control was performed without the antibody of interest and used to set the baseline at 0. 2.10.3 In vitro neutralization assay Vero E6 cells were seeded in culture in DMEM medium, with 10 % of fetal bovine serum, 1 % penicillin, and 1 % streptomycin. In a 96-well plate, 30,000 cells were deposited in each well, and left for 1 day at 37 °C in a 5 % CO2 atmosphere. Antibody solutions were prepared, and diluted in DMEM 0 % FBS at concentrations ranging from 6.8 ng/mL to 5 µg/mL. A SARS-CoV-2 viral suspension was also prepared at 100-fold its TCID50 (i.e. 5.106 PFU/mL). It was then mixed to the antibody solutions, and incubated for 1 h at 37 °C. These mixtures were then added to the Vero cells, in the corresponding well, before incubating for 96 h at 37 °C. Each plate contains negative controls (cells infected with no antibody) and positive controls (cells not infected, cells in contact with virus and control neutralizing antibody from 5 μg/mL to 40 ng/mL). Cell viability was quantified using Viral ToxGlo™ Assay kit (Promega) according to the manufacturer's recommendations. The signal was read with an EnVision microplate reader (Perkin Elmer). For each condition, an average of the 3 replicates was calculated, and the data were processed with GraphPad Prism software (GraphPad Software, USA). The neutralizing titer NT50 was calculated using [Inhibitor] vs response model (variable slope, four parameters). To confirm the trends observed with luminescence values, the cells were also observed in brightfield with Evos FL Microscope (ThermoFisher Scientific, USA). An image of each well was recorded for qualitative assessment of the cytopathic effect of the virus on the cells in the presence or absence of the antibody. Those results were shown in Supplementary Information S27 only, with only 1 image for each triplicate (as they were always similar). 2.11 Droplet size measurement A volume of 3.6 mL of antibody solution was deposited in the nebulizing cup of a vibrating mesh nebulizer NEB-001 (Briutcare, China). The nebulizer was then mounted on a Spraytec laser diffraction system (Malvern Panalytical, UK): the distance between the laser and the detector was of 8 cm, while nebulizer outlet was placed at 2 cm from the laser. The nebulization flow rate was set at 0.4 mL/min, and the experiment was performed at room temperature and ambient relative humidity. Six measurements were performed on the same sample: four measurements with a duration of 1 min, and two measurements with a duration of 4 min, in order to analyze the totality of the solution. Note the last analysis was stopped before the end of the measurement, due to the lack of solution. Data were then analyzed using Spraytec software (Malvern Panalytical, UK). An average value of d10, d50, d90 was given for the six measurements. 2.12 In vivo lung distribution studies Four C57BL/6 wild type mice were treated with a solution of sotrovimab at 1 mg/mL, formulated with 5.6 mM of C8TreSuc in PBS and one untreated mouse as a control was used in this study. The laboratory procedures comply with French legislation, which implements the European Directives (Reference Number: D3417223, APAFIS#23920-2020020320279696v3). Mice from the treated group were anaesthetized individually using 5 % of isoflurane for 2 min. A nebulization mask was homemade by setting the finger of a latex glove at the tip of a vibrating mesh nebulizer NEB-001 (Briutcare, China), and piercing it gently. The 4 mice were then individually treated by placing the muzzle in the nebulization mask and aerosol was administered during 1 min at a flow rate of 0.4 mL/min. Blood and lung were collected from 2 mice at 30 min (n = 2) and from the 2 other mice at 24 h (n = 2) after treatment for IgG quantification in blood and qualitative biodistribution assessment by histopathology. 2.12.1 Blood sampling and ELISA human IgG analysis At each time point, the animals were anaesthetized using ketamine/xilazine mixture (100 and 10 mg/kg respectively) administered by intraperitoneal injection at 10 mL/kg. In anesthetized animals, 1 mL of blood was sampled by cardiac puncture and collected in a tube containing heparine-Na as anticoagulant. Samples were centrifuged for 15 min at 1000 × g (or 3000 rpm) at 2–8 °C within 30 min of collection. 200 µL of supernatant (plasma) was stored at −20 °C for ELISA analysis. Human IgG quantification of undiluted samples (100 µL/well) was performed in duplicated by ELISA method. 2.12.2 Lung sampling and immunohistology analysis After blood sampling, mice were sacrificed by cervical dislocation. Then, each animal was transcardially perfused using PBS solution with PFA 4 % during 2 min. After perfusion, the chest was opened and both lungs were sampled. Lungs were collected in a fresh tube and fixed overnight with PFA 4 % at 4 °C. Following fixation, samples were incubated 24–48 h in two successive baths of 6 % and 30 % sucrose, embedded in Optimal Cutting Temperature and stored at −80 °C. Coronal sections (10 μm of thickness) were cut using cryostat apparatus (LEICA CM3050). Proximal, central and distal lung sections (three segments) were cut. For immunolabeling, three cryosections were blocked with 5 % Normal Fish Serum (NFS) and 0.1 % triton X-100 in PBS, incubated overnight at 4 °C with rabbit anti-human IgG diluted 1/500 in NFS/Triton/PBS (5 % NFS and 0.05 % Triton in 100 mL PBS 1X pH 7.4), washed with PBS and then incubated 1 h at RT with green secondary goat anti-rabbit Alexa Fluor 488 diluted at 1/10,000 in NFS/Triton/PBS (5 % NFS and 0.05 % Triton in 100 mL PBS 1X pH 7.4). After several PBS washes, cryosections were mounted in Dako fluorescent mounting medium (S3023) with DAPI (1/10,000). Confocal laser scanning microscope (CLSM) (SP8-UV confocal microscope, Leica, Germany) was used to obtain images. The fluorescence intensity, corresponding to the IgG concentration was quantified for each lung section using ImageJ software. 3 Results and discussion 3.1 Preparation and analysis of the excipients The ionic surfactants were synthesized by desymmetrization of D-(+)-Trehalose via mono esterification of its primary alcohols with a fatty acid (n = 6, 8, 12, 14) in the presence of TBTU as coupling agent (Fig. 1 ). The resulting CxxTre mono esters were obtained in 30 to 65 % yield (Table 1 ). Purification of these intermediates by flash chromatography, using a slow gradient of eluent (MeOH 5 to 20 % in EtOAc:DCM 1:1), allowed to efficiently remove the side products resulting from the esterification of secondary alcohols. The second esterification was carried out in the presence of succinic acid mono benzyl ester (SAMBE) and TBTU, leading to the functionalization of the second primary alcohol in 30 to 49 % yield. The benzyl group was removed by Pd catalyzed hydrogenolysis to afford the ionic surfactants CxxTreSuc, presenting a carboxylate moiety and aliphatic chains of various lengths. An alternative pathway involving temporary protection of the secondary alcohols as trimethylsilyl ethers, followed by sequential esterification of the remaining primary alcohols, was investigated. However, the higher number of synthetic steps led to drastically lower yields and this route was consequently discarded.Fig. 1 Synthesis of Trehalose derived ionic surfactants. Reagents and conditions: i- Fatty acid (n = 6, 8, 12, 14), TBTU, pyridine 96 hr. ii- SAMBE, TBTU pyridine 96 hr. iii- H2 Pd/C, MeOH 3 hr. Table 1 Yields of intermediates and final surfactants with respect to the length of the fatty acid used (n = 6, 8, 12, 14). Fatty acid CxxTre CxxTreSucBn CxxTreSuc Octanoate (n = 6) 40 % 49 % 60 % Decanoate (n = 8) 30 % 40 % 63 % Tetradecanoate (n = 12) 35 % 34 % 58 % Hexadecanoate (n = 14) 65 % 30 % 64 % As observed in Table 1, the yield varied between 39 and 65 % for the first step, 30 to 49 % for the second step and 58 to 64 % for the final step. These yields were reproducible throughout the numerous batches produced during this study. Also, there is no correlation between the yield of the reactions and the length of the fatty acid used but it is dependent on the type of reaction and purification strategy. The critical micelle concentration (CMC) of the excipient can bring important insight on the behaviour of trehalose-based surfactants in aqueous medium, and provide valuable clues for the formulation. As a population of micelles must appear from CMC and at higher concentrations, CMC was assessed by DLS, measuring the derived count rate (correlated to the number of particles in suspension) as a function of the concentration for the different excipient solutions (graphical determinations are shown in the Supplementary Information, Figures S1, S2, S3). The results are shown in Table 2 . First, it must be noticed that the derived count rate increase was mainly caused by the apparition of a peak at 5–10 nm (data not shown), which corresponds to spherical micelles of monoalkyl trehalose as described in the literature (Schiefelbein et al., 2010). However, C8TreSuc did not seem to form this type of micelles, and the increase of scattering intensity was caused by larger structures (from few tens to few hundred nanometers). No further analysis of those structures was performed, but they might correspond to merged vesicles as suggested by another team working with non-succinylated C8Tre (Kanemaru et al., 2012). Therefore, the DLS data were not exploitable for the CMC determination of C8TreSuc. For the other surfactants, CMC results showed that the longer the alkyl chain, the lower the CMC, increasing from 0.07 mM for C16TreSuc to 0.96 mM for C10TreSuc. This is in good agreement with results obtained with other surfactants (Hanson et al., 2020). More interestingly, other groups have studied the CMC of non-succinylated trehalose monoalkyl derivatives (Chen et al., 2007). They showed the same increase of the CMC while reducing the length of the alkyl chain (30 mM for C8Tre, 2.2 mM for C10Tre, 0.21 mM for C14Tre and 0.045 mM for C16Tre). The influence of the succinylation could also be studied through CMC values. However, the direct comparison of CMC values between succinylated and non-succinylated trehalose surfactants is not relevant, as several publications showed that CMC values could be significantly different depending on the measurement method used (Hanson et al., 2020). For example, the CMC of non-succinylated C8Tre was measured at 30 mM by the drop weight method (Chen et al., 2007), when another group who confronted tensiometry and pyrene fluorescence method obtained a CMC of 2.14 and 1.92 mM respectively (Schiefelbein et al., 2010).Table 2 Critical Micellar Concentration of succinylated trehalose-based excipients, determined by DLS (n.d.: not determined). Surfactant C16TreSuc C14TreSuc C10TreSuc C8TreSuc CMC (mM) 0.07 0.67 0.96 n.d. The cytotoxicity of all trehalose-based surfactants (succinylated or not) was assessed on NIH-3T3 cells after 48 h with a MTS cell viability test. C16Tre could not be studied, due to its very low solubility (less than 5 mM for C16Tre whereas C16TreSuc was still soluble at a concentration of 800 mM in milli-Q water). EC50 were calculated by fitting dose–response models with experimental data, as shown in Supplementary Information S25) (Table 3 ). First, C8Tre showed an EC50 value higher than 11.2 mM, while C10Tre and C14Tre had EC50 of 1.91 mM and 0.21 mM respectively. This trend suggests that the longer the alkyl chain, the higher the cytotoxicity. This was confirmed by the EC50 of succinylated compounds, progressively decreasing from 3.61 mM (for C8TreSuc) to 0.34 mM (for C14TreSuc). Also, succinylation of the trehalose appeared to slightly increase the toxicity, as EC50 decreased from more than 11.2 mM to 3.61 mM after succinylation of C8Tre. The same trend was observed for C10Tre, although less pronounced (from 3.91 to 1.75 mM after succinylation). Indeed, succinyl group was probably negatively charged in the conditions of the experiment, as cells were maintained at pH = 7.4 (which is higher than the pKa of most carboxylic acid). Yet, it is well known that anionic surfactants are more toxic than non-ionic surfactants, as they can generate electrostatic interactions with proteins (Lémery et al., 2015). However, if succinylation and long carbon chains seemed to increase cell toxicity, it was noticed that the chains of 14 or 16 carbons had very similar EC50 (0.21–0.34 mM), whether the trehalose was succinylated or not. In other words, there might be a toxicity threshold, above which increased number of carbons and succinylation have no more influence on the cytotoxicity of the compound. Interestingly, four of the trehalose-based surfactants studied were less toxic than PS80 (EC50 = 0.96 mM, which is close to other values found in the literature for similar experiments (Arechabala et al., 1999)), known itself for its low acute toxicity (Lansdown and Grasso, 1972). The three other surfactants were more toxic than PS80, but their respective EC50 remained in the same order of magnitude.Table 3 EC50 of the different excipients depending on the length of their carbon chain and the presence of a succinyl moiety, determined on NIH-3 T3 cells after 48 h of contact. Carbon chain C8 C10 C14 C16 Trehalose greater than 11.2 mM 1.91 mM 0.21 mM n.e. Succinyl trehalose 3.61 mM 1.75 mM 0.34 mM 0.24 mM n.e.: not evaluated. 3.2 Trehalose-based excipients as antibody stabilizers during nebulization process A first screening of the stabilizing conditions was made by monitoring the formation of antibody aggregates during nebulization of freshly prepared formulations. The nebulized mixture always contained 1 mg/mL of antibody, solubilized in PBS to have an isotonic medium, which avoids post-inhalation coughing reflex (Bonvini et al., 2016, Sahakijpijarn et al., 2020). The pH was adjusted to 5.8, thus remaining in the range of pH of commercial products (Wang and Ohtake, 2019, Warne, 2011). Depending on the samples, a trehalose-based excipient was also present, in various amounts. In the range of surfactants described in this study, we focused on the ones with the shortest and the longest carbon chain, respectively 8 and 16 carbons. This choice was made to show significant differences of behavior, depending on the chain length. However, as C16Tre was not sufficiently soluble to be used, we worked mainly on the new succinylated trehalose-based surfactants, C8TreSuc and C16TreSuc. The stability of the antibody during nebulization was assessed by DLS, following the area under the curve of aggregate peaks on the volume-weighed size distribution. As shown in Fig. 2 , the presence of trehalose-based excipients can prevent the formation of aggregates, and their effect is proportional to their concentration. Indeed, when no excipient was used, antibody aggregates represented 0.6 % of the AUC in the fresh solution, while reaching 9.7 % after nebulization. When C8TreSuc was added to the mixture (Fig. 2A), aggregation was dramatically restrained; aggregates increased from 2.1 to 4.6 % after nebulization with only 1 mM of C8TreSuc. C16TreSuc showed similar protective abilities, with aggregates proportion of 0.6 % before nebulization versus 4.9 % after nebulization with 1 mM of surfactant. If the excipient concentration was increased to 5.6 mM, a full protection of antibody against aggregation was achieved, whatever the carbon chain length (for C8TreSuc and C16TreSuc, shown in Fig. 2, but also for C14TreSuc shown in Supplementary Information S26). The screening of the excipient concentration was performed once to select the best concentration of excipient in the formulation. Complete assessment of the formulations submitted to successive stress conditions (freeze-drying, storage and nebulization) can be found in triplicates in the next part of this study (see section 3.3).Fig. 2 Effect of nebulization on mAb aggregation in the presence of different concentrations of (A) C8TreSuc or (B) C16TreSuc (n = 1). Aggregation was monitored following the AUC of the aggregate peaks, relative to the total AUC of the volume-weighed size distribution obtained by DLS. Noticeably, the effective concentration was not related to the excipient CMC. Indeed, other excipients studied stabilized the antibody during nebulization when their concentration was set at 5.6 mM (see Supplementary Information S26), even if their CMC values were very different, ranging from 0.07 to 0.96 mM (Table 2). This absence of correlation between effective concentration and CMC is in good agreement with other results found in the literature (Cui et al., 2017, Mahler et al., 2009). This phenomenon arises from the fact that the optimal concentration for protein stabilization depends greatly on the interactions between the protein, the surfactant, and the interfaces. Indeed, the optimal concentration is reached when the surfactant molecules are numerous enough to cover all the interfaces and/or all the protein surface. This concentration can be considered as the CMC of the surfactant in solution containing proteins (Lee et al., 2011). It differs from the classical CMC, measured in pure water, hence the absence of correlation between the CMC measured in this work and the optimal stabilizing concentration for sotrovimab nebulization. Then, the ability of the excipients to preserve the antibody functionality during nebulization was assessed in vitro. The antibody was solubilized with an effective concentration of excipient (5.6 mM of C8TreSuc or C16TreSuc), and nebulized. Nebulized and non-nebulized samples were then incubated with SARS-CoV-2 spike-expressing HEK293 cells. The proportion of antibodies that bind to cells was then assessed by FACS thanks to an AlexaFluor488-labelled secondary antibody. The binding of excipient-stabilized antibody, nebulized or not, was compared to the binding of non-stabilized antibody, in PBS only (Fig. 3 ). The results first show that 5.6 mM of trehalose-based excipient does not prevent the non-nebulized antibody from binding to Spike protein, as antibodies bound to 78 ± 1.6 % and 84 ± 1.2 % of cells when protected by C8TreSuc or C16TreSuc respectively, versus 80 ± 2.8 % of labelled-cells when no excipient was used. After nebulization, the proportion of positive cells dramatically dropped to 32 ± 3.2 % when no excipient was used whereas in presence of trehalose-based excipient (5.6 mM) the proportion of positive cells remained unchanged (73 ± 3.2 % with C8TreSuc, 82 ± 1.4 % with C16TreSuc). Those results confirmed the stabilizing role of both C8TreSuc and C16TreSuc when used at 5.6 mM to protect antibodies against the different stresses occurring during nebulization.Fig. 3 Effect of nebulization on mAb binding to Spike-expressing HEK cells, depending on the presence or not of C8TreSuc or C16TreSuc at 5.6 mM. mAb binding was assessed by measuring the percentage of cells labelled by a fluorescent secondary antibody, by FACS and is expressed as the mean ± SD (3 independent experiments). 3.3 Trehalose-based excipients as a cryoprotectant While trehalose-based excipients were proven to be good stabilizers for nebulized antibodies in solution, we also investigated the interest of our synthesized excipients for long-term storage of the proteins in a dry form. Indeed, lyophilization is a common method to guarantee the stability of antibody formula stored for a long period of time. However, this process requires a freezing step and a drying step, that can trigger protein instability leading to a change of its conformation and thus its efficacy (Cui et al., 2017, Wang and Ohtake, 2019). Addition of sugar-based compounds (e.g. sucrose, trehalose) or surfactants (e.g. PS 80) have shown promise for preservation of protein structure during freeze-drying, the former acting as remarkable lyoprotectants and the latter as protein stabilizers during freezing (Wang, 2000). To demonstrate the cryopreservative property of our trehalose surfactants, antibody solutions with 5.6 mM of C8TreSuc or C16TreSuc were lyophilized and stored for 28 days at 4 °C, before assessing both the aggregation and the antigen-binding ability of the antibody. As shown in Fig. 4 A, the formulation containing C8TreSuc presented some aggregation (0.4 ± 0.1 % of aggregates on the volume weighed distribution before lyophilization, 3.6 ± 2.7 % after lyophilization and 28 days of storage) similar to the results obtained with formulation in PBS only (0.1 ± 0.1 % of aggregates before lyophilization, and 4.4 ± 2.7 % after lyophilization and 28 days of storage). Notably, the standard errors were high for those two conditions. Due to the non-quantitative value of DLS results, the proportion of detected aggregates was highly variable, when present in large amounts. This variability led us to consider the aggregation results with caution, despite the apparently high level of aggregation. Thus, antigen-binding assays were performed on the same samples, and the results were confirmed (Fig. 4B). A dramatic decrease of the cell binding was observed when antibody was solubilized in PBS alone (from 84 ± 0.6 to 9 ± 4 % after the process) or in the presence of 5.6 mM of C8TreSuc (from 76 ± 0.6 % to 11 ± 9 %). On the contrary, C16TreSuc prevented antibody from aggregating (Fig. 4A, 0.2 ± 0.1 % in the initial solution, and 0.1 ± 0.1 % after lyophilization and storage) and preserved its antigen-binding ability (Fig. 4B, 75 ± 1.6 % in the initial solution, versus 63 ± 7 % after lyophilization and storage). Thus, fatty acid chain length positively correlates with the antibody protection during lyophilization and storage. This can be explained by the stabilization mechanism itself. During lyophilization, protein conformation can change, and hydrophobic regions can be exposed to the outer surface, inducing destabilization of the proteins and increasing aggregation. The surfactant hydrophobic chain interacts with newly exposed hydrophobic regions of the protein, which prevents protein hydrophobic regions from interacting with each other, limiting irreversible conformation changes, and aggregation (Lee et al., 2006). Thus, C8TreSuc might be less cryoprotective than C16TreSuc because its shorter carbon chain confers lower hydrophobicity, hence decreased interaction with proteins during lyophilization.Fig. 4 Effect of freeze-drying, storage and nebulization (A) on mAb aggregation (monitored following the AUC of the aggregate peaks, relative to the total AUC of the volume-weighed size distribution obtained by DLS) and (B) on mAb binding to Spike-expressing HEK cells (assessed by measuring the percentage of cells labelled by a fluorescent secondary antibody, by FACS), depending on the presence or not of PS80, C8TreSuc or C16TreSuc at 5.6 mM. (n.p.: non performed). All values are expressed as the mean ± SD (3 independent experiments). Then, we challenged the cryoprotective performance of C16TreSuc against other excipients commonly used in protein formulation. The cryoprotective performances of this innovative excipient were comparable to a formula containing 5 % of trehalose, commonly used for protein lyophilization (Kaushik and Bhat, 2003). However, trehalose alone had no protective effect during nebulization (data not shown), which made it less versatile than C16TreSuc. More interestingly, C16TreSuc was compared to PS80, recognized as a gold standard for protein stabilization, in liquid-form storage (Warne, 2011), but also during nebulization (Respaud et al., 2014, Sécher et al., 2022) or lyophilization (Ji et al., 2014). We studied the influence of C16TreSuc and the reference PS80 surfactants on antibody aggregation and antigen-binding over the whole lifecycle of the drug. Thus, the antibody solution was prepared, lyophilized, stored for 28 days before being resuspended in purified water and nebulized. It appeared that PS80 did protect the antibody during its whole lifecycle. Antibody aggregates did not exceed 0.1 % of the total volume-weighed particle size distribution (Fig. 4A), and antigen-binding ability of antibody remained stable, between 68 ± 2 % before lyophilization, and 71 ± 4 % at the end of the lyophilization/storage/nebulization cycle (Fig. 4B). On the other hand, the novel excipient C16TreSuc which previously showed great cryopreservation capacity also demonstrated its ability to preserve antibody colloidal stability (0.3 ± 0.2 % of aggregates) and binding ability (70 ± 2 %) after the whole lyophilization/storage/nebulization cycle. Thus, C16TreSuc equals PS80 as a versatile excipient for antibody stabilization. But it also showed some advantages compared to polysorbates. Indeed, C16TreSuc was conceived to be biodegradable, with hydrolysable ester bounds. After cleavage, the three released compounds are trehalose, succinic acid and palmitic acid, which are naturally present in the environment. It does not contain any PEG, unlike PS80, which is clearly an advantage regarding the issues recently highlighted by several groups about massive PEG use (Guerrini et al., 2022, Hamad et al., 2008, Stone et al., 2019, Szebeni et al., 2011, Turecek et al., 2016, Yang et al., 2016). 3.4 Therapeutic potential of the antibody after nebulization in the presence of trehalose-based excipients 3.4.1 In vitro neutralization assays After demonstrating that trehalose-based excipients could help preserving antibody colloidal stability and antigen-binding ability, we aimed to show the interest of stabilizing Sotrovimab with these excipients for the treatment of COVID-19 by pulmonary way. To do so, we first demonstrated the potential of the formula to neutralize SARS-COV-2 virus in vitro after being nebulized. Wuhan variant virus was first mixed with nebulized antibody formulations, and then incubated with Vero E6 cells. The cytopathic effect caused by viral infection was measured and NT50 values were calculated (Table 4 ) (obtained from cell viability curves showed in Supplementary Information S25). The results indicated that, in the conditions of the experiment, at a concentration of 2.90 µg/mL, the non-nebulized antibody in PBS prevented half of the cells lysis. In the presence of 5.6 mM of nebulized C8TreSuc formulation, the NT50 slightly decrease (NT50 = 2.02 µg/mL) and was divided by 3 with C16TreSuc (NT50 = 0.95 µg/mL). The reliability of these surprising results was supported by the NT50 of the control neutralizing antibody, that was found to be 0.38 µg/mL in the conditions of this experiment, which is in the same order of magnitude as the provider’s claims (0.29 µg/mL with WT SARS-CoV-2 Spike Pseudovirus). To explain the remarkable increase of neutralizing capacity of the antibody when nebulized in presence of C16TreSuc, we first considered that the sotrovimab solution in PBS could have undergone instabilities during the incubation time with the virus (1 h at 37 °C). However, for all samples, no aggregation was detected by DLS after the incubation period (see Supplementary Information S28). A second hypothesis would be a favorable conformation change of sotrovimab, that could have been induced by the presence of C16TreSuc, thus improving neutralization effect. Indeed, it was shown that PS80 and 20 can change the antibody conformation, without triggering aggregation (Singh et al., 2017). Further studies are needed to confirm this result.Table 4 NT50 of the antibody after nebulization in the presence of C8TreSuc or C16TreSuc 5.6 mM, measured in the conditions of the in vitro infection of Vero E6 cells by SARS-CoV-2. The control was a solution of antibody in PBS, with no excipient and no nebulization stress. Sample Non nebulized mAb in PBS Nebulized mAb with C8TreSuc Nebulized mAb with C16TreSuc NT50 (µg/mL) 2.90 2.02 0.95 3.4.2 In vivo lung biodistribution studies Finally, a preliminary in vivo study was conducted to assess the antibody distribution in the lungs. C57BL/6 wild type mice were treated with a formula containing 1 mg/mL of antibody in PBS pH 5.8, stabilized with 5.6 mM of trehalose-based surfactant. Lung distribution was determined by labelling the human antibody of interest by immunohistochemistry in lung superior, middle and inferior lobes collected after 30 min or 24 h. As shown by CLSM images (Fig. 5 A) and Mean Fluorescence Intensity measurements (MFI) (Fig. 5B), the nebulized antibody reached the superior and middle lobes and remained there in significant amounts after 30 min (93 AU in the superior lobe, 58 AU in the middle lobe). Comparatively, the inferior lobe contained very few antibodies, even though the MFI was higher than for the control mouse. This phenomenon was already observed for mice, and was explained by a preferential ventilation of the superior lobes in small rodents (Gu and Darquenne, 2021).Fig. 5 Antibody biodistribution after treatment of C57BL/6 wild type mice, with 0.4 mL of a formula containing 1 mg/mL of antibody in PBS pH 5.8, stabilized with 5.6 mM of C8TreSuc. (A, B) Lung distribution was determined by labelling human IgG of interest by immunohistochemistry on lung sections collected after 30 min or 24 h. Images shown in A are representative of what was observed on mice receiving the same treatment, and scale bar represents 100 µm. They were used to evaluate the mean fluorescence intensity of Alexa488, shown in B (1 value for the negative control, 2 independent values for the other conditions; mean ± SD). (C) Blood levels of human IgG determined by ELISA on blood samples, collected after 30 min or 24 h (1 value for the negative control, 2 independent values for the other conditions; mean ± SD). In superior and middle lobes, the antibody was homogenously distributed all over the cross section: this might mean the aerosolized formulation did not only deposit in the larger airways, but also in the alveoli (Bivas-Benita et al., 2005). This assumption was supported by the droplet size measurements, performed with Briutcare nebulizer by laser diffraction (Table 5 ). Nebulization of a solution of sotrovimab, at 1 mg/mL concentration, with or without the presence of excipient generated particles of median diameter d50 around 4.80 µm. This droplet size distribution was not influenced by the presence of trehalose-based excipient, the d10 and d90 being similar with or without C16TreSuc at 5.6 mM (remaining respectively around 2.3 µm and 9.0 µm). According to recent reports, particles with a diameter below 5 µM deposit in the deep lungs, by diffusion, which means more than 50 % of the droplets would be able to reach the alveoli, while the rest would deposit in the bronchi and bronchioles (Carvalho et al., 2011). A more cautious threshold is sometimes proposed, considering that only particles smaller than 3 µm can deposit in the alveoli (Carvalho et al., 2011). This would still allow more than 10 % of the produced aerosol to reach the deeper lung (d10 = 2.41 µm), which supports the assumption that the aerosol was distributed in the whole airways. Considering that SARS-CoV-2 infection is characterized, in part, by alveoli injuries (Upadhya et al., 2022), such aerosol depth deposition might be useful in the perspective of a treatment for COVID-19 disease.Table 5 Droplet size distribution of aerosols containing the antibody 1 mg/mL in PBS pH 5.8, alone or stabilized with C16TreSuc at 5.6 mM. Nebulization was performed with a vibrating mesh nebulizer NEB-001 (Briutcare) at a 0.4-mL/min flow rate. Six measurements were done on each sample, using a Spraytec laser diffraction system (Malvern Panalytical). Values are expressed as the mean ± SD (6 replicate measurements). Sample d10 (µm) d50 (µm) d90 (µm) Sotrovimab in PBS 2.28 ± 0.02 4.82 ± 0.05 9.10 ± 0.12 Sotrovimab in PBS + C16TreSuc 5.6 mM 2.41 ± 0.13 4.78 ± 0.04 8.92 ± 0.18 Beside the antibody distribution 30 min after the treatment administration, it is noteworthy that the antibody remained in the lung after 24 h, with MFI values of 62 AU in the superior lobe and 42 AU in the middle lobe (Fig. 5B). It is generally acknowledged that free antibody is eliminated from airways in less than 24 h by mucociliary clearance and by macrophages (Freches et al., 2017, Todoroff and Vanbever, 2011). However, other publications showed the presence of antibody several days after the nebulization treatment by immunochemistry assays (Guilleminault et al., 2014). We assumed that the antibody detected on CLSM images during immunohistochemistry assays was actually internalized by the lung epithelium. Finally, the presence of the antibody in the plasma was quantified by ELISA, 30 min and 24 h after nebulization (Fig. 5C). The results showed a slow absorption of the antibody: antibody level was too low to be detected 30 min after nebulization, and it reached 3.8 ng/mL after 24 h. However, this level was still very low, compared to the concentration that would be obtained if the totality of the aerosolized antibody reached the lung and was absorbed through the alveoli walls (around 0.2 mg/mL for a mouse with 1.8 mL of blood). This suggests that a significant portion of the deposited antibody remained in the lung for at least 24 h, which is the targeted organ for COVID-19 treatment. 4 Conclusion We developed new excipients that preserve antibody stability during nebulization process, allowing the administration of sotrovimab in lungs, for the treatment of COVID-19. Indeed, we synthesized a series of novel excipients composed of a succinylated trehalose polar head and a hydrophobic carbon chain of various length (from 8 to 16 carbons). Those excipients showed a higher solubility than their non-succinylated counterparts, allowing their use at relevant concentrations for protein stabilization. In particular, C16TreSuc was proven to be an excellent candidate to preserve colloidal stability and antigen-binding ability of an antibody during the nebulization process. It also showed great advantages as a cryoprotectant, allowing to store antibodies in a lyophilized form for one month, at least. Finally, we demonstrated that C16TreSuc could have potential uses in immunotherapy against COVID-19. The presence of C16TreSuc during nebulization preserved the neutralization capacity of sotrovimab against SARS-CoV-2 in vitro; an increase of its efficacy was even observed, compared to the control. A preliminary in vivo study also showed the wide distribution of sotrovimab in mice lung, after nebulization with 5.6 mM of trehalose-based excipient. This study demonstrates the potential of trehalose-based excipients to stably nebulize human mAbs with no loss in binding and no aggregation. This opens the road for alternative excipients to replace usual stabilizers, such as polysorbates, for protein lung delivery. 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 The raw data associated to this study are available on Zenodo at: https://doi.org/10.5281/zenodo.7243350 Acknowledgements The authors thank L. Menin and D. Ortiz (EPFL ISIC-MSEAP) for their support with MS characterizations and LC-MS quantifications, A. Bornet (EPFL ISIC-NMRP) for his assistance with NMR measurements. The present work has also benefited from the work of IDD-Xpert (France) for the droplet size analysis and InVivex (France) for the in vivo lung distribution study. We also acknowledge Centre d’Etudes des Maladies Infectieuses et Pharmacologie Anti-Infectieuse (UAR3725 CNRS – Université de Montpellier, France). MV and PM thank Labex MabImprove (ANR-10-LABX-53) for its support. This work was funded by research industrial grants at EPFL and ICGM. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijpharm.2022.122463. ==== Refs References Arechabala B. Coiffard C. Rivalland P. Coiffard L.J.M. Roeck-Holtzhauer Y.D. Comparison of cytotoxicity of various surfactants tested on normal human fibroblast cultures using the neutral red test, MTT assay and LDH release J. Appl. Toxicol. 19 1999 163 165 10.1002/(SICI)1099-1263(199905/06)19:3<163::AID-JAT561>3.0.CO;2-H 10362266 Bird J.H. Khan A.A. Nishimura N. Yamasaki S. Timmer M.S.M. Stocker B.L. Synthesis of branched trehalose glycolipids and their mincle agonist activity J. Org. Chem. 83 2018 7593 7605 10.1021/acs.joc.7b03269 29781274 Bivas-Benita M. Zwier R. Junginger H.E. Borchard G. Non-invasive pulmonary aerosol delivery in mice by the endotracheal route Eur. J. Pharm. Biopharm. 61 2005 214 218 10.1016/j.ejpb.2005.04.009 16039104 Bonvini S.J. Birrell M.A. Grace M.S. Maher S.A. Adcock J.J. Wortley M.A. Dubuis E. Ching Y.-M. Ford A.P. Shala F. Miralpeix M. Tarrason G. Smith J.A. Belvisi M.G. Transient receptor potential cation channel, subfamily V, member 4 and airway sensory afferent activation: Role of adenosine triphosphate J. Allergy Clin. Immunol. 138 2016 249 261.e12 10.1016/j.jaci.2015.10.044 26792207 Carsana L. Sonzogni A. Nasr A. Rossi R.S. Pellegrinelli A. Zerbi P. Rech R. Colombo R. Antinori S. Corbellino M. Galli M. Catena E. Tosoni A. Gianatti A. Nebuloni M. Pulmonary post-mortem findings in a series of COVID-19 cases from northern Italy: a two-centre descriptive study Lancet Infect. Dis. 20 2020 1135 1140 10.1016/S1473-3099(20)30434-5 32526193 Carvalho T.C. Peters J.I. Williams R.O. Influence of particle size on regional lung deposition – What evidence is there? Int. J. Pharm. 406 2011 1 10 10.1016/j.ijpharm.2010.12.040 21232585 Chen J. Kimura Y. Adachi S. Surface activities of monoacyl trehaloses in aqueous solution LWT Food Sci. Technol. 40 2007 412 417 10.1016/j.lwt.2005.11.006 Corti D. Purcell L.A. Snell G. Veesler D. Tackling COVID-19 with neutralizing monoclonal antibodies Cell 184 2021 3086 3108 10.1016/j.cell.2021.05.005 34087172 Cui Y. Cui P. Chen B. Li S. Guan H. Monoclonal antibodies: formulations of marketed products and recent advances in novel delivery system Drug Dev. Ind. Pharm. 43 2017 519 530 10.1080/03639045.2017.1278768 28049357 Elbein A.D. Pan Y.T. Pastuszak I. Carroll D. New insights on trehalose: a multifunctional molecule Glycobiology 13 2003 17R 27R 10.1093/glycob/cwg047 Ema COVID-19: EMA recommends authorisation of antibody medicine Xevudy [WWW Document] accessed 8.3.22 Eur. Med. Agency. 2021 https://www.ema.europa.eu/en/news/covid-19-ema-recommends-authorisation-antibody-medicine-xevudy Freches D. Patil H.P. Machado Franco M. Uyttenhove C. Heywood S. Vanbever R. PEGylation prolongs the pulmonary retention of an anti-IL-17A Fab’ antibody fragment after pulmonary delivery in three different species Int. J. Pharm. 521 2017 120 129 10.1016/j.ijpharm.2017.02.021 28192159 Fröhlich E. Salar-Behzadi S. Oral inhalation for delivery of proteins and peptides to the lungs Eur. J. Pharm. Biopharm. 163 2021 198 211 10.1016/j.ejpb.2021.04.003 33852968 Gu W. Darquenne C. Heterogeneity in lobar and near-acini deposition of inhaled aerosol in the mouse lung J. Aerosol Sci 151 2021 105642 10.1016/j.jaerosci.2020.105642 Guerrini G. Gioria S. Sauer A.V. Lucchesi S. Montagnani F. Pastore G. Ciabattini A. Medaglini D. Calzolai L. Monitoring Anti-PEG Antibodies Level upon Repeated Lipid Nanoparticle-Based COVID-19 Vaccine Administration Int. J. Mol. Sci. 23 2022 8838 10.3390/ijms23168838 36012103 L, Guilleminault, N, Azzopardi, C, Arnoult, J, Sobilo, V, Hervé, J, Montharu, A, Guillon, C, Andres, O, Herault, A, Le Pape, P, Diot, E, Lemarié, G, Paintaud, V, Gouilleux-Gruart, N, Heuzé-Vourc’h, Fate of inhaled monoclonal antibodies after the deposition of aerosolized particles in the respiratory system. Journal of Controlled Release 196 2014, 344 354. 10.1016/j.jconrel.2014.10.003. Ha E. Wang W. Wang Y.J. Peroxide formation in polysorbate 80 and protein stability J. Pharm. Sci. 91 2002 2252 2264 10.1002/jps.10216 12226852 Hamad I. Hunter A.C. Szebeni J. Moghimi S.M. Poly(ethylene glycol)s generate complement activation products in human serum through increased alternative pathway turnover and a MASP-2-dependent process Mol. Immunol. 46 2008 225 232 10.1016/j.molimm.2008.08.276 18849076 Hanson M.G. Katz J.S. Ma H. Putterman M. Yezer B.A. Petermann O. Reineke T.M. Effects of hydrophobic tail length variation on surfactant-mediated protein stabilization Mol. Pharmaceutics 17 2020 4302 4311 10.1021/acs.molpharmaceut.0c00737 Hertel S.P. Winter G. Friess W. Protein stability in pulmonary drug delivery via nebulization Adv. Drug Delivery Rev. Protein stability in drug delivery applications 93 2015 79 94 10.1016/j.addr.2014.10.003 International Organization of Standardization, 2009. ISO 10993-5:2009 : Biological evaluation of medical devices — Part 5: Tests for in vitro cytotoxicity. Ismail R. Baaity Z. Csóka I. Regulatory status quo and prospects for biosurfactants in pharmaceutical applications Drug Discov. Today 26 2021 1929 1935 10.1016/j.drudis.2021.03.029 33831583 Jain N.K. Roy I. Effect of trehalose on protein structure Protein Sci. 18 2009 24 36 10.1002/pro.3 19177348 Jana S. Kulkarni S.S. Synthesis of trehalose glycolipids Org. Biomol. Chem. 18 2020 2013 2037 10.1039/D0OB00041H 32115587 Jana S. Mondal S. Kulkarni S.S. Chemical Synthesis of Biosurfactant Succinoyl Trehalose Lipids Org. Lett. 19 2017 1784 1787 10.1021/acs.orglett.7b00550 28328221 Ji C. Sun M. Yu J. Wang Y. Zheng Y. Wang H. Niu R. Trehalose and Tween 80 Improve the Stability of Marine Lysozyme During Freeze-Drying Biotechnol. Biotechnol. Equip. 2014 Kale S.S. Akamanchi K.G. Trehalose monooleate: a potential antiaggregation agent for stabilization of proteins Mol. Pharmaceutics 13 2016 4082 4093 10.1021/acs.molpharmaceut.6b00686 Kanemaru M. Yamamoto K. Kadokawa J. Self-assembling properties of 6-O-alkyltrehaloses under aqueous conditions Carbohydr. Res. 357 2012 32 40 10.1016/j.carres.2012.05.014 22704195 J.K, Kaushik, R, Bhat, Why Is Trehalose an Exceptional Protein Stabilizer: An analysis of the thermal stability of proteins in the presence of the compatible osmolyte trehalose. Journal of Biological Chemistry 278 2003 26458–26465. 10.1074/jbc.M300815200. Kügler J.H. Muhle-Goll C. Kühl B. Kraft A. Heinzler R. Kirschhöfer F. Henkel M. Wray V. Luy B. Brenner-Weiss G. Lang S. Syldatk C. Hausmann R. Trehalose lipid biosurfactants produced by the actinomycetes Tsukamurella spumae and T. pseudospumae Appl. Microbiol. Biotechnol. 98 2014 8905 8915 10.1007/s00253-014-5972-4 25091045 Kuyukina M.S. Ivshina I.B. Baeva T.A. Kochina O.A. Gein S.V. Chereshnev V.A. Trehalolipid biosurfactants from nonpathogenic Rhodococcus actinobacteria with diverse immunomodulatory activities New Biotechnol. Eur. Congress of Biotechnol. - ECB 16 32 2015 559 568 10.1016/j.nbt.2015.03.006 Lansdown A.B.G. Grasso P. Physico-chemical factors influencing epidermal damage by surface active agents Br. J. Dermatol. 86 1972 361 378 10.1111/j.1365-2133.1972.tb05049.x 5023897 Leader B. Baca Q.J. Golan D.E. Protein therapeutics: a summary and pharmacological classification Nat Rev Drug Discov 7 2008 21 39 10.1038/nrd2399 18097458 Lee R.C. Despa F. Guo L. Betala P. Kuo A. Thiyagarajan P. Surfactant copolymers prevent aggregation of heat denatured lysozyme Ann. Biomed. Eng. 34 2006 1190 1200 10.1007/s10439-006-9139-z 16786393 Lee H.J. McAuley A. Schilke K.F. McGuire J. Molecular origins of surfactant-mediated stabilization of protein drugs Adv. Drug Delivery Rev. Formulating biomolecules: mechanistics insights in molecular interactions 63 2011 1160 1171 10.1016/j.addr.2011.06.015 Lémery E. Briançon S. Chevalier Y. Bordes C. Oddos T. Gohier A. Bolzinger M.-A. Skin toxicity of surfactants: structure/toxicity relationships Colloids Surf A Physicochem. Eng. Asp. 469 2015 166 179 10.1016/j.colsurfa.2015.01.019 Lerbret A. Bordat P. Affouard F. Hédoux A. Guinet Y. Descamps M. How do trehalose, maltose, and sucrose influence some structural and dynamical properties of lysozyme? insight from molecular dynamics simulations J. Phys. Chem. B 111 2007 9410 9420 10.1021/jp071946z 17629322 Liang W. Pan H.W. Vllasaliu D. Lam J.K.W. Pulmonary Delivery of Biological Drugs. Pharma. 12 2020 1025 10.3390/pharmaceutics12111025 Lins R.D. Pereira C.S. Hünenberger P.H. Trehalose–protein interaction in aqueous solution. proteins: structure Function, and Bioinformatics 55 2004 177 186 10.1002/prot.10632 Mahler H.-C. Senner F. Maeder K. Mueller R. Surface activity of a monoclonal antibody J. Pharm. Sci. 98 2009 4525 4533 10.1002/jps.21776 19655376 A, Mayor, B, Thibert, S, Huille, R, Respaud, H, Audat, N, Heuzé-Vourc’h, Inhaled antibodies: formulations require specific development to overcome instability due to nebulization. Drug Deliv. and Transl. Res. 11 2021 1625–1633. 10.1007/s13346-021-00967-w. Moiset G. López C.A. Bartelds R. Syga L. Rijpkema E. Cukkemane A. Baldus M. Poolman B. Marrink S.J. Disaccharides impact the lateral organization of lipid membranes J. Am. Chem. Soc. 136 2014 16167 16175 10.1021/ja505476c 25316578 Montharu J. Le Guellec S. Kittel B. Rabemampianina Y. Guillemain J. Gauthier F. Diot P. de Monte M. Evaluation of lung tolerance of ethanol, propylene glycol, and sorbitan monooleate as solvents in medical aerosols J. Aerosol Med. Pulm. Drug Deliv. 23 2010 41 46 10.1089/jamp.2008.0740 20131984 Naughton P.J. Marchant R. Naughton V. Banat I.M. Microbial biosurfactants: current trends and applications in agricultural and biomedical industries J. Appl. Microbiol. 127 2019 12 28 10.1111/jam.14243 30828919 Ohtake S. Wang Y.J. Trehalose: current use and future applications J. Pharm. Sci. 100 2011 2020 2053 10.1002/jps.22458 21337544 Paul S. Paul S. Molecular Insights into the Role of Aqueous Trehalose Solution on Temperature-Induced Protein Denaturation J. Phys. Chem. B 119 2015 1598 1610 10.1021/jp510423n 25558880 Pinto D. Park Y.-J. Beltramello M. Walls A.C. Tortorici M.A. Bianchi S. Jaconi S. Culap K. Zatta F. De Marco A. Peter A. Guarino B. Spreafico R. Cameroni E. Case J.B. Chen R.E. Havenar-Daughton C. Snell G. Telenti A. Virgin H.W. Lanzavecchia A. Diamond M.S. Fink K. Veesler D. Corti D. 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 R, Respaud, D, Marchand, C, Parent, T, Pelat, P, Thullier, J.-F, Tournamille, M.-C, Viaud-Massuard, P, Diot, M, Si-Tahar, L, Vecellio, N, Heuzé-Vourc’h, Effect of formulation on the stability and aerosol performance of a nebulized antibody. mAbs 6 2014 1347–1355. 10.4161/mabs.29938. R, Respaud, L, Vecellio, P, Diot, N, Heuzé-Vourc’h, Nebulization as a delivery method for mAbs in respiratory diseases. Expert Opinion on Drug Delivery 12 2015 1027–1039. 10.1517/17425247.2015.999039. Roifman, ChaimM., Levison, H., Gelfand, ErwinW., 1987. High-dose versus low-dose intravenous immunoglobulin in hypogammaglobulinaemia and chronic lung disease. The Lancet, Originally published as Volume 1, Issue 8541 329, 1075–1077. 10.1016/S0140-6736(87)90494-6. Sahakijpijarn S. Smyth H.D.C. Miller D.P. Weers J.G. Post-inhalation cough with therapeutic aerosols: Formulation considerations Adv. Drug Deliv. Rev. 165–166 2020 127 141 10.1016/j.addr.2020.05.003 Schiefelbein L. Synthesis, characterization and assessment of suitability of trehalose fatty acid esters as alternatives for polysorbates in protein formulation Eur. J. Pharm. Biopharm. 9 2010 Schiefelbein L. Keller M. Weissmann F. Luber M. Bracher F. Frieß W. Synthesis, characterization and assessment of suitability of trehalose fatty acid esters as alternatives for polysorbates in protein formulation Eur. J. Pharm. Biopharm. 76 2010 342 350 10.1016/j.ejpb.2010.08.012 20816956 T, Sécher, E, Bodier-Montagutelli, C, Parent, L, Bouvart, M, Cortes, M, Ferreira, R, MacLoughlin, G, Ilango, O, Schmid, R, Respaud, N, Heuzé-Vourc’h, Aggregates Associated with Instability of Antibodies during Aerosolization Induce Adverse Immunological Effects. Pharmaceutics 14 2022 671. 10.3390/pharmaceutics14030671. Singh S.M. Bandi S. Jones D.N.M. Mallela K.M.G. Effect of Polysorbate 20 and Polysorbate 80 on the Higher-Order Structure of a Monoclonal Antibody and Its Fab and Fc Fragments Probed Using 2D nuclear magnetic resonance spectroscopy J. Pharm. Sci. 106 2017 3486 3498 10.1016/j.xphs.2017.08.011 28843351 Song W. Gui M. Wang X. Xiang Y. Cryo-EM structure of the SARS coronavirus spike glycoprotein in complex with its host cell receptor ACE2 PLoS Pathog. 14 2018 e1007236 30102747 Stone C.A. Liu Y. Relling M.V. Krantz M.S. Pratt A.L. Abreo A. Hemler J.A. Phillips E.J. Immediate hypersensitivity to polyethylene glycols and polysorbates: more common than we have recognized. the journal of allergy and clinical immunology In Pract. 7 2019 1533 1540.e8 10.1016/j.jaip.2018.12.003 Sudrik C. Cloutier T. Pham P. Samra H.S. Trout B.L. Preferential interactions of trehalose, L-arginine.HCl and sodium chloride with therapeutically relevant IgG1 monoclonal antibodies MAbs 9 2017 1155 1168 10.1080/19420862.2017.1358328 28758834 Szebeni J. Muggia F. Gabizon A. Barenholz Y. Activation of complement by therapeutic liposomes and other lipid excipient-based therapeutic products: prediction and prevention Adv. Drug Delivery Rev., Complement Monotoring of Nanomed. Implants 63 2011 1020 1030 10.1016/j.addr.2011.06.017 Todoroff J. Vanbever R. Fate of nanomedicines in the lungs Curr. Opin. Colloid Interface Sci. 16 2011 246 254 10.1016/j.cocis.2011.03.001 Turecek P.L. Bossard M.J. Schoetens F. Ivens I.A. PEGylation of Biopharmaceuticals: a review of chemistry and nonclinical safety information of approved drugs J. Pharm. Sci. 105 2016 460 475 10.1016/j.xphs.2015.11.015 26869412 Upadhya S. Rehman J. Malik A.B. Chen S. Mechanisms of Lung Injury Induced by SARS-CoV-2 Infection Physiology 37 2022 88 100 10.1152/physiol.00033.2021 34698589 Wang W. Lyophilization and development of solid protein pharmaceuticals Int J Pharm 203 2000 1 60 10.1016/s0378-5173(00)00423-3 10967427 Wang W. Ohtake S. Science and art of protein formulation development Int. J. Pharm. 568 2019 118505 10.1016/j.ijpharm.2019.118505 Warne N.W. Development of high concentration protein biopharmaceuticals: the use of platform approaches in formulation development Eur. J. Pharma. Biopharmaceutics, Unmet Needs in Protein Formulation Sci. 78 2011 208 212 10.1016/j.ejpb.2011.03.004 Yang Q. Jacobs T.M. McCallen J.D. Moore D.T. Huckaby J.T. Edelstein J.N. Lai S.K. Analysis of Pre-existing IgG and IgM Antibodies against Polyethylene Glycol (PEG) in the General Population Anal. Chem. 88 2016 11804 11812 10.1021/acs.analchem.6b03437 27804292	36462738	PMC9710110	NO-CC CODE	2022-12-07 23:16:15	no	Int J Pharm. 2023 Jan 5; 630:122463	utf-8	Int J Pharm	2,022	10.1016/j.ijpharm.2022.122463	oa_other
==== Front Neurologia (Engl Ed) Neurologia (Engl Ed) Neurologia (Barcelona, Spain) 2173-5808 Sociedad Española de Neurología. Published by Elsevier España, S.L.U. S2173-5808(22)00183-3 10.1016/j.nrleng.2022.01.006 Letter to the Editor Telematic adaptation to home mechanical ventilation in patients with amyotrophic lateral sclerosis Adaptación telemática a la ventilación mecánica domiciliaria en pacientes con esclerosis lateral amiotróficaBalañá A. ab Rubio M.Á. ab Bertran B. ab Martínez Llorens J. abcd⁎ a Servicio de Neumología, Hospital del Mar, Cataluña, Spain b Unidad Multidisciplinar de ELA, Cataluña, Spain c CEXS, Universitat Pompeu Fabra, Cataluña, Spain d CIBER de Enfermedades Respiratorias (CIBERES), ISC III. Madrid, Spain ⁎ Corresponding author. 30 11 2022 30 11 2022 © 2022 Sociedad Española de Neurología. Published by Elsevier España, S.L.U. 2022 Sociedad Española de Neurología 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	36462623	PMC9710147	NO-CC CODE	2022-12-01 23:23:40	no	Neurologia (Engl Ed). 2022 Nov 30; doi: 10.1016/j.nrleng.2022.01.006	utf-8	Neurologia (Engl Ed)	2,022	10.1016/j.nrleng.2022.01.006	oa_other
==== Front Am J Otolaryngol Am J Otolaryngol American Journal of Otolaryngology 0196-0709 1532-818X Saunders S0196-0709(22)00345-3 10.1016/j.amjoto.2022.103718 103718 Article COVID-19 and transtympanic injections for sudden sensorineural hearing loss Adams Jason K. Marinelli John P. Travis Newberry R. Spear Samuel A. Erbele Isaac D. ⁎ Brooke Army Medical Center, San Antonio, TX, United States of America Uniformed Services University of the Health Sciences, Bethesda, MD, United States of America ⁎ Corresponding author at: 3551 Roger Brooke Dr., Department of Otolaryngology-Head and Neck Surgery, San Antonio Uniformed Services Health Education Consortium, JBSA-Ft Sam Houston, TX 78234, United States of America. 30 11 2022 March-April 2023 30 11 2022 44 2 103718103718 15 9 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 Multiple reports have linked COVID-19 infection with sudden sensorineural hearing loss (SSNHL), although other studies have failed to demonstrate this association. The current study was conceived to examine the rates of SSNHL across a large, principally national, population by characterizing the rate of transtympanic injections for SSNHL during the pandemic. Methods Retrospective review of all patients that underwent transtympanic injection from 2019 to 2020. Results Covering a unique beneficiary population of 9.6 million individuals of all ages in the United States, a statistically significant decrease in transtympanic injections for SSNHL was performed from 2019 to 2020 (p = 0.04, IRR = 0.91, 95 % CI = 0.84–0.99). No patient receiving a transtympanic injection also had a COVID-19 diagnosis. Conclusions These findings support the idea that COVID-19 infections do not clinically significantly increase patients' risk of developing SSNHL. In fact, the decreased exposure through social isolation to other common viruses implicated in causing SSNHL may have actually led to a lower rate of SSNHL during the pandemic. Keywords COVID Sudden sensorineural hearing loss Population-based United States Abbreviations SSNHL, Sudden Sensorineural Hearing Loss CPT, Current Procedural Terminology ICD10, International Classification of Disease, Tenth Revision IRR, Incidence Rate Ratio ==== Body pmc1 Introduction Sudden sensorineural hearing loss (SSNHL) is an otologic emergency requiring aggressive, timely intervention to prevent permanent hearing loss. Early in the COVID-19 pandemic, reports surfaced linking SSNHL to COVID-19 infections [1], [2]. Multiple other limited reports have since suggested COVID-19 infection may place patients at increased risk for SSNHL [3], although several other studies have failed to demonstrate an association [4], [5], [6]. Since many of the conflicting reports surrounding COVID-19 and SSNHL extend from smaller studies from single institutions, the current study was conceived to examine the rates of SSNHL across a large, principally national, population by characterizing the rate of transtympanic injections for SSNHL during the pandemic. 2 Methods This study was approved by the local Institutional Review Board (IRB #C.2022.002n). Within an insurance beneficiary population of 9.6 million (Tricare), a comprehensive beneficiary database was queried for patients undergoing transtympanic injections (Current Procedural Terminology (CPT) 69801). This database included patients of all ages treated at national and international military treatment facilities, as well as those procedures for which Tricare paid to have performed at civilian centers outside the military health system. Patients with an International Classification of Disease-10 (ICD10) code for SSNHL (H91.2, H91.21, H91.22) were included. Within this population, the ICD10 codes given at the time of injection was also queried for COVID-19 infection (U0.71, B34.2, B97.29). ICD10 codes for a personal history of COVID-19 infection (Z86.16) and a suspected exposure to COVID-19 (Z20.822) were not included, since these were added January 2021. The procedures performed were compared by year using an incidence rate ratio (IRR). Analysis was performed with R version 3.5.3 (R Foundation for Statistical Computing; Vienna, Austria) and RStudio version 1.1.463 (RStudio, Inc.; Boston, USA), with the additional packages “tidyverse” version 1.2.1 [7] and “fmsb” version 0.7.3 [8]. 3 Results A total of 2129 transtympanic injections were performed in 946 patients with a diagnosis of SSNHL. There were 496 patients who received 1110 injections in 2019, and 473 patients received 1019 injections in 2020. Twenty-three patients had injections in both 2019 and 2020, likely representing a series of injections started in 2019. There was a statistically significant decrease in injections performed from 2019 to 2020 (p = 0.04, IRR = 0.91, 95 % CI = 0.84–0.99) (Fig. 1 ), however, there was no statistically significant difference in the number of patients who received injections (p = 0.41, IRR = 0.95, 95 % CI = 0.84–1.08). No patient receiving a transtympanic injection also had a COVID-19 diagnosis at the time of injection.Fig. 1 Decrease in transtympanic injections performed in patients with sudden sensorineural loss between 2019 and 2020, *p = 0.04. Fig. 1 4 Discussion Across a large national patient population in the United States, the current study demonstrates a decline in transtympanic injections performed for SSNHL in 2020. Despite covering a population of 9.6 million subjects, no patients who received a transtympanic injection also had a diagnosis of COVID-19 infection. These findings support the idea that COVID-19 infections do not clinically significantly increase patients' risk of developing SSNHL. This work reinforces findings in studies from Israel and Massachusetts on ICD10 codes [5], [6], survey data of otolaryngologists on the perception of COVID-19's effect on SSNHL [9], and a study testing for COVID-19 in patients presenting for SSNHL [4]. Interestingly, similar to the current work, other reports have also observed a decrease in rates of SSNHL during the pandemic, perhaps secondary to decreased exposure to common viruses that actually cause SSNHL [10]. The absence of patients receiving a transtympanic injection for SSNHL with a co-diagnosis of COVID-19 is notable, particularly in the setting of an increased concern for the relationship between COVID-19 and SSNHL. It may be influenced by a potential lack of surgeon awareness of the ICD10 code added in April 2020, or that it was not added because symptomatic patients were not seeing otolaryngologists. Notable limitations of this analysis include that the pandemic prevented some patients from pursuing care. As alluded to previously, miscoding is possible, especially with regards to a COVID-19 diagnosis. Additionally, this study does not explore a possibly salient relationship between COVID-19 and Meniere's. 5 Conclusion Based on transtympanic injections from this large patient population, there was not a clinically significant increase in SSNHL. These data support the mounting evidence that COVID-19 infection does not increase patients' risk of developing SSNHL. Funding source None. Prior presentation This material has not been previously presented or published. Department of defense disclosure The views expressed herein are those of the authors and do not reflect the official policy of position of Brooke Army Medical Center, the U.S. Army Medical Department, the U.S. Army Office of the Surgeon General, the Department of the Army, the Department of the Air Force, the Department of Defense, or the U.S. Government. CRediT authorship contribution statement Jason K. Adams: ethics approval, data analysis and interpretation, manuscript drafting and review, final approval, accountability for all aspects of the work; John P. Marinelli: data analysis and interpretation, manuscript drafting and review, final approval, accountability for all aspects of the work; Travis R. Newberry: data analysis and interpretation, manuscript drafting and review, final approval, accountability for all aspects of the work; Samuel A. Spear: data analysis and interpretation, manuscript drafting and review, final approval, accountability for all aspects of the work; Isaac D. Erbele: study design, ethics approval, data collection, data analysis and interpretation, manuscript drafting and review, final approval, accountability for all aspects of the work. Declaration of competing interest None. Acknowledgements The authors wish to thank Ms. Stacy Leonard, Ms. Sandra Walker, Ms. Rincy Varughese, and Ms. Elsa Granato of the Department of Defense Hearing Center of Excellence for their support in data extraction. ==== Refs References 1 Kilic O. Kalcioglu M.T. Cag Y. Could sudden sensorineural hearing loss be the sole manifestation of COVID-19? An investigation into SARS-COV-2 in the etiology of sudden sensorineural hearing loss Int J Infect Dis 97 2020 208 211 32535294 2 Kaliyappan K. Chen Y.C. Krishnan Muthaiah V.P. Vestibular Cochlear manifestations in COVID-19 cases Front Neurol 13 2022 850337 3 Meng X. Wang J. Sun J. Zhu K. COVID-19 and sudden sensorineural hearing loss: a systematic review Front Neurol 13 2022 883749 4 van Rijssen L.B. Derks W. Hoffmans R. No COVID-19 in patients with sudden sensorineural hearing loss (SSNHL) Otol Neurotol 43 2 2022 170 173 34889826 5 Doweck I. Yanir Y. Najjar-Debbiny R. Shibli R. Saliba W. Sudden sensorineural hearing loss during the COVID-19 pandemic JAMA Otolaryngol Head Neck Surg 148 4 2022 373 375 35084455 6 Chari D.A. Parikh A. Kozin E.D. Reed M. Jung D.H. Impact of COVID-19 on presentation of sudden sensorineural hearing loss at a single institution Otolaryngol Head Neck Surg 165 1 2021 163 165 33228476 7 Wickham H. Averick M. Bryan J. Welcome to the tidyverse J Open Source Softw 4 2019 1686 1692 8 Nakazawa M. fmsb: functions for medical statistics book with some demographic data 2022 9 Pool C. King T.S. Pradhan S. Isildak H. Sudden sensorineural hearing loss and COVID-19 J Laryngol Otol 1–15 2022 10 Parrino D. Frosolini A. Toninato D. Matarazzo A. Marioni G. de Filippis C. Sudden hearing loss and vestibular disorders during and before COVID-19 pandemic: an audiology tertiary referral Centre experience Am J Otolaryngol 43 1 2022 103241	36470008	PMC9710149	NO-CC CODE	2022-12-02 23:16:58	no	Am J Otolaryngol. 2023 Nov 30 March-April; 44(2):103718	utf-8	Am J Otolaryngol	2,022	10.1016/j.amjoto.2022.103718	oa_other
==== Front Biosens Bioelectron Biosens Bioelectron Biosensors & Bioelectronics 0956-5663 1873-4235 Elsevier B.V. S0956-5663(22)01019-3 10.1016/j.bios.2022.114979 114979 Article High-intensity vector signals for detecting SARS-CoV-2 RNA using CRISPR/Cas13a couple with stabilized graphene field-effect transistor Sun Yang a Yang Cheng b Jiang Xiaolin d Zhang Pengbo a Chen Shuo b Su Fengxia a Wang Hui a Liu Weiliang a He Xiaofei a Chen Lei c∗∗∗ Man Baoyuan b∗∗ Li Zhengping a∗ a Department of Chemistry and Biological Engineering, University of Science and Technology Beijing 30 Xueyuan Road, Haidian District, Beijing, 100083, PR China b Department of Physics and Electronics, Shandong Normal University 1 Daxue Road, Changqing District, Jinan, Shandong Province, 250014, PR China c Department of of Life Sciences, Shandong Normal University 1 Daxue Road, Changqing District, Jinan, Shandong Province, 250014, PR China d Shandong Provincial Key Laboratory of Communicable Disease Control and Prevention, Shandong Center for Disease Control and Prevention, 16992 Jingshi Road, Lixia District, Jinan, Shandong Province, 250014, PR China ∗ Corresponding author. Corresponding author. * Corresponding author. 30 11 2022 15 2 2023 30 11 2022 222 114979114979 7 10 2022 25 11 2022 29 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. False detection of SARS-CoV-2 is detrimental to epidemic prevention and control. The scalar nature of the detected signal and the imperfect target recognition property of developed methods are the root causes of generating false signals. Here, we reported a collaborative system of CRISPR-Cas13a coupling with the stabilized graphene field-effect transistor, providing high-intensity vector signals for detecting SARS-CoV-2. In this collaborative system, SARS-CoV-2 RNA generates a “big subtraction” signal with a right-shifted feature, whereas any untargets cause the left-shifted characteristic signal. Thus, the false detection of SARS-CoV-2 is eliminated. High sensitivity with 0.15 copies/μL was obtained. In addition, the wide concerned instability of the graphene field-effect transistor for biosensing in solution environment was solved by the hydrophobic treatment to its substrate, which should be a milestone in advancing it's engineering application. This collaborative system characterized by the high-intensity vector signal and amazing stability significantly advances the accurate SARS-CoV-2 detection from the aspect of signal nature. Keywords Vector signal SARS-CoV-2 detection Stabilized graphene field-effect transistor CRISPR/Cas13a Amplification-free ==== Body pmc1 Introduction The characteristics of human-to-human transmission and a high proportion of asymptomatic infections caused the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spread rapidly and large-scale outbreaks worldwide, posing a massive threat to public health (Arons et al., 2020; Boyton and Altmann, 2021). The severe epidemic has prompted scientists to explore various strategies for SARS-CoV-2 detection (Kevadiya et al., 2021; Orooji et al., 2021), such as surface-enhanced Raman spectroscopy (Yang et al., 2021; Sitjar et al., 2021), microfluidic integrated biochip (Swank et al., 2021; Welch et al., 2022), colorimetric assay (Talebian et al., 2020), loop-mediated isothermal amplification (Bokelmann et al., 2021), and real-time reverse transcription-polymerase chain reaction (RT-PCR) (Liu et al., 2020; Smyrlaki et al., 2020). These works have actively promoted the detection of SARS-CoV-2 from different aspects. Determining the intensity of the detected signal is the primary basis for these reported strategies to distinguish positive samples from negative samples. However, the low abundance of biomarkers in mildly or asymptomatic positive patients and the unperfect detection mechanism focusing on signal intensity alone, leading to the detected signal may be false-negative or false-positive (Cohen et al., 2020; Long et al., 2020; Nath et al., 2021; Song et al., 2020; Woloshin et al., 2020). Inaccurate test results have brought a lot of inconvenience to people's health, finance, education, and psychological convenience, affecting the government's decision-making and causing unnecessary loss of social resources (Surkova et al., 2020). The root cause of false-negative signals is that trace biomarkers cannot effectively generate high-intensity detectable signals. The root cause of false-positive signals is that the reported detection signals are scalar signals rather than vector signals, where untargeted biomarkers can generate similar-intensity signals to target biomarkers. Suppose the signal generation ability of trace target biomarkers can be amplified and the detected signals could have a directional characteristic in addition to the intensity characteristics alone, it will undoubtedly strongly prevent the false detection results. Hailed by nature as one of the seven technologies that will majorly impact science in 2022 (Eisenstein, 2022), the clustering regularly interspaced short palindromic repeats (CRISPR) based diagnostic technology, especially the CRISPR-Cas13a system, has shown significant advantages in the field of virus diagnosis and attracted widespread attention (Kaminski et al., 2021; Tang et al., 2021). SARS-CoV-2 is a positive-strand RNA virus (Chan et al., 2020). For the CRISPR-Cas13a system, the CRISPR is a programmable spacer sequence that can recognize target RNA with single-base mismatch specificity and activates the Cas13a. The activated Cas13a performs a “collateral cleavage” effect and could indiscriminately cleave any surrounding single-strand RNA molecules (Gootenberg et al., 2017). The impressive cleavage to RNA, high base resolution, and programmability upon nucleic acid recognition make the CRISPR-Cas13a system highly promising to specifically amplify the detection signal of SARS-CoV-2 RNA. Biosensors based on graphene field effect transistors (GFET) have attracted widespread attention in the rapid and high-precision detection of various diseases by analyzing nucleic acid information due to their abundant analyzable signals, short response time, high sensitivity, and low cost. In recent years, the rapid and sensitive detection of SARS-CoV-2 (Wang et al., 2022a), Alzheimer's disease, Down's trisomy, and various cancers has been effectively realized by GFET biosensors (Bonanni et al., 2012; Liu et al., 2019; Mao et al., 2017). For the GFET, graphene shows a strong ambipolar electric field effect (Novoselov et al., 2004), where the exogenous electron doping makes graphene's Dirac point move to the left and the exogenous hole doping makes graphene's Dirac point move to the right (Chen et al., 2009). Type change of exogenous doping modulates the Dirac point to move in different directions (Ang et al., 2008). In addition, graphene's high carrier mobility and good electrical conductivity make the Dirac point change significantly responding to minor charge disturbances on the graphene's surface (Schedin et al., 2007). These make GFET could sensitively detect charged molecules (such as RNA or DNA) through a directional signal. Here, we reported a collaborative system of CRISPR-Cas13a coupling with graphene field-effect transistor for detecting SARS-CoV-2 RNA by generating a high-intensity vectorized signal. CRISPR-Cas13a system makes the detected signal with high intensity, and the GFET makes the detected signal directional. In this collaborative system, the reporter probes (RP) are saturable bound to the graphene surface and dope the graphene with the electron, making Dirac point shift to the left. CRISPR-Cas13a activated by SARS-CoV-2 RNA can massively cleavage the reporter probes and sharply attenuate the reporter probe's electron doping to graphene, resulting in an amplified signal with a significant right shift of the Dirac point and thus generating a “big subtraction” signal. Inability to activate CRISPR-Cas13a and unspecific adsorption with graphene, make untargeted molecules play the same role as reporter probes, doping graphene with electrons and making the Dirac point furtherly shift to the left. Thus, the false results for detecting SARS-CoV-2 can be effectively eliminated by the opposite signal direction. As a result, high sensitivity as low as 0.15 copies/ μL was obtained. More importantly, as well known, the instability of graphene field-effect transistor resulting from the time-dependent left shift of graphene Dirac point in analytical liquid is one of the most important factors that plague it's practicability for many years. In this work, we firstly found that hydrophobic treatment of the field-effect transistor's substrate can successfully eliminate this instability, which should be a milestone in advancing the reliable application of the field-effect transistor in biosensing. 2 Material and methods 2.1 Reagents LawCas13a protein was purchased from HuicH (Shanghai, China). 10XNEB buffer r2.1 was purchased from NEW ENGLAND BioLabs Inc. (Ipswich, UK). Recombinant RNase Inhibitor (RRI), DEPC-treated water, DNase I, Taq DNA Polymerase, dNTP mixture, NTP mixture, forward primer and reverse primer were obtained from TaKaRa Biotechnology Co., Ltd. (Dalian, China). T7 RNA Polymerase was purchased from New England Biolabs (Beijing, China). RNA clean Kit was obtained from TIANGEN Biotech (Beijing) Co., Ltd. EZ-10 Spin Column Viral Total RNA Extraction Kit acquired from Sangon Biotech (Shanghai, China). The indium tin oxide (ITO)/SiO2 substrate was purchased from GULUO GLASS (Luo yang) Co., Ltd. 1-Pyrenebutanoic acid succinimidyl ester (PBASE), ethanolamine, n-octadecyl trichlorosilane (OTS), and Dimethylsulfoxide (DMSO) were purchased from Aladdin Co., Ltd. The detailed information on the used nucleic acids in this work is shown in Table S1. 2.2 Immobilization of RP PBASE functionalized the fabricated GFET to immobilize the RP. The pyrene group of PBASE can bind with graphene by π-π stacking and thus form a self-assembled uniformly monolayer film on graphene surface (Hinnemo et al., 2017; Zhang et al., 2007), as show in Fig. S1. Then the 5′ terminus amine group-modified RP consisting of 5 uracil ribonucleotides (U) at 5′ terminus and 16 thymine deoxyribotides at 3′ terminus, were immobilized on the GFET surface through the conjugation reaction between the amine group and the amine-reactive succinimide group of PBASE (Fig. S2). The functionalization and immobilization results were characterized by transfer characteristic curve, X-ray photoelectron spectroscopy (XPS), and Raman spectrum. Firstly, after PBASE functionalization, the Dirac point, value of VGS corresponding to minimum IDS, exhibits an obvious right shift from 73 mV to 121 mV (Fig. S3, orange line), which is due to its pyrene group causing hole doping to graphene. Negatively charged RP contribute electron doping to graphene, making the Dirac point left-shifted from 121 mV to 76 mV when finishing the immobilization process (Fig. S3, purple line). In addition, the gradual decrease in the sharpness of the resistivity near the Dirac point also indicates the successful functionalization and immobilization (Tan et al., 2007). Secondly, the Raman spectrum, which is sensitive to doping (Das et al., 2008), was selected to characterize the functionalization and immobilization processes further. The fingerprint 2D peak of graphene shifts by 13 cm−1 to a positive direction after PBASE modification (Figs. S4A–4B), indicating graphene was effectively functionalized and has been hole-doped (Das et al., 2008; Wu et al., 2017). Then, the position of the 2D peak shifts by 5 cm−1 to a negative direction after reacting with the RP (Figs. S4B–4C), indicating RP was successfully immobilized and graphene has been electron-doped (Das et al., 2008; Wu et al., 2017). XPS of the functionalized graphene shows the clear N 1s peak, C-N, and C <svg xmlns="http://www.w3.org/2000/svg" version="1.0" width="20.666667pt" height="16.000000pt" viewBox="0 0 20.666667 16.000000" preserveAspectRatio="xMidYMid meet"><metadata> Created by potrace 1.16, written by Peter Selinger 2001-2019 </metadata><g transform="translate(1.000000,15.000000) scale(0.019444,-0.019444)" fill="currentColor" stroke="none"><path d="M0 440 l0 -40 480 0 480 0 0 40 0 40 -480 0 -480 0 0 -40z M0 280 l0 -40 480 0 480 0 0 40 0 40 -480 0 -480 0 0 -40z"/></g></svg> N, which is caused by the N element in the Pyrene group of PBASE (Fig. S5A). The significant P 2p peak appears upon RP immobilization since RP consists of the P element (Fig. S5B). The above characterizations reveal the successful functionalization of graphene and the immobilization of RP. Here,the process of PBASE functionalizing graphene was finished in two steps instead of one step in the conventional method. Detailly, the pure graphene channel first reacts with PBASE solution in 30 min and then reacts with new PBASE solution in another 30 min after DMSO flushing. The two-step method shows a larger change of Dirac point than the one-step method with the same reaction time (Figs. S6A–6B). Similarly, the RP was also immobilized on the graphene surface by a two-step method. In the first step, 50 μL of the RP solution at a concentration of 10 mM was added to the liquid storage tank to react with PBASE. After 3 h, the liquid storage tank was rinsed several times with RNase-free water. In the second step, another 50 μL of the RP solution with a concentration of 10 mM was added again to the liquid storage tank to react with graphene for 3 h at room temperature. 2.3 Fabrication of proposed GFET Details about preparing graphene was described in text S1 and Fig. S7. The fabrication process of liquid-gated large-scale graphene field-effect transistor based on hydrophobic substrate consists of 10 steps (Fig. S8). The scale of obtained graphene/copper foil is 5 cm × 10 cm. Step 1: the integral graphene/copper was cut as several small graphene/Copper samples with a scale of 0.8 cm × 1.2 cm (Fig. S8A). Step 2: The acetone solution of polymethyl methacrylate (PMMA) was spin-coated on the surface of the graphene/copper sample. After drying at 100oC for 1 h, PMMA film was formed to serve as the support layer for transferring graphene (Fig. S8B). Step 3: Put the PMMA/graphene/copper foil sample on the surface of the ferric chloride solution to etch the copper foil away (Fig. S8C). Step 4: After completing the etching step, the PMMA/graphene samples are transferred to the surface of deionized water by the coverslip for cleaning FeCl3 and CuCl2, and the cleaning process was repeated three times (Fig. S8D). To make the graphene field-effect transistor stable in solution, the substrate of the graphene field-effect transistor (glass substrate coated with ITO on both sides of the surface) was modified with hydrophobic function. Step 5: The substrates and 10 mL tubes were pre-dried at 70 °C for 1 h (Fig. S8E). Step 6: The substrate and 1% OTS solution were loaded into the pre-dried 10 mL tube and reacted for 20 min (Fig. S8F). Note that the 10 mL tube should be protected from light. After the modification process, the hydrophobic substrate was rinsed with toluene, acetone, alcohol, and deionized water. Step 7: The PMMA/graphene samples floating on the surface of deionized water were fished up with the hydrophobic substrate. During the fishing process, the PMMA/graphene sample should be accurately connected to the ITO electrodes on both sides of the glass substrate surface (Fig. S8G). Step 8: The PMMA/graphene/substrate sample was heated at 130 °C for 1.5 h to bond the graphene to the substrate surface (Fig. S8H). Step 9: To fully remove PMMA, the PMMA/graphene/substrate samples were soaked in acetone overnight (Fig. S8I). The graphene/substrate samples were washed sequentially with acetone, alcohol, and deionized water to obtain pure graphene. Step 10: The homemade polymethyl methacrylate liquid storage tank was pasted precisely above the graphene with UV-curable glue. Up to here, the liquid-gate large-scale graphene field-effect transistor based on the hydrophobic substrate is completed (Fig. S8J). 2.4 The collaborative system for detecting SARS-COV-2 RNA Before using this collaborative system to detect SARS-CoV-2 RNA, it is necessary to mix SARS-CoV-2 RNA and CRISPR-Cas13a system in advance and pre-react in a test tube at 37 °C for 1 h to ensure that the RNP is fully activated. In addition, the pre-reaction could make the large-size targets were fully degraded before adding to GFET, avoiding the Cas13 enzymes be trapped on the sensor surface by the unspecific interaction between the large-size targets and the graphene. As shown in Fig. S9, the activated CRISPR-Cas13a system after 1 h pre-reaction just work to cleavage the reporter probes and would not contribute any unspecific absorption. After 1 h the pre-reaction, the mixed solution was mixed again and added to the liquid storage tank of GFET to react with the RP at 37 °C for 2 h, where the liquid storage tank should be sealed to avoid interference from the external environment. The detailed component of the mixture is shown in Table S2. 3 Results and discussion The shift of Dirac points (ΔDP) has been widely chosen as the characteristic signal for graphene field-effect transistor sensors to detect biomarkers (Gao et al., 2018; Hwang et al., 2020; Wang et al. 2019, 2021, 2022b). However, the Dirac point will spontaneously shift over time in solution (Gao et al., 2022; Hu et al., 2019), which causes extreme noise to the detected signal. The Dirac point of traditional GFET undergoes a clear left shift in 4 h without adding any target biomarkers to the test solution, and the maximum difference is 63 mV (Fig. 1 A). Water in the test solution can penetrate the graphene and form a sub-nanometer-thin icelike film between the graphene and the substrate (Hong et al., 2019; Lee et al. 2012, 2014). Icelike water film modulates the charge transfer between graphene and the substrate, and it's thickness would increase with a positive charge and decrease with a negative charge (Chiu et al., 2019; Dollekamp et al., 2017; Hong et al., 2019). For GFET, the sweep gate voltage (charge varies from −1 v to 1 v) is a necessary condition for reading the Dirac point. Thus, the presence of the icelike water film makes it impossible to accurately determine that the detected signal is entirely caused by the target biomarkers, resulting in unconvincing detection results. We found that the hydrophobic treatment to GFET substrate can effectively solve this instability. Here, the substrate of GFET was treated by OTS, whose alkyl chain could form a hydrophobic film on the substrate, preventing the formation of the icelike water film between graphene and substrate (Fig. S10). The hydrophobic modification can be simply and quickly completed by immersing the graphene substrate in the toluene solution of OTS for 20 min. The liquid contact angle of the graphene substrate changed from 26 °C to 105 °C after hydrophobic modification (Fig. S11). Benefiting from the hydrophobic treatment, the Dirac point almost keeps a constant with a low RSD (1.1%) even if the GFET was immersed in a liquid environment for 31 h (Fig. 1B). Improving the stability of graphene in solution environment is a necessary technological breakthrough to promote reliable biosensing for GFET.Fig. 1 Stability performance. The Dirac point shift over time of (A) traditional FET and (B) developed GFET with the hydrophobic-treated substrate. Fig. 1 The working principle of this collaborative system for detecting SARS-COV-2 RNA is illustrated in Fig. 2 . CRISPR contains a single stem-loop domain (Tracr RNA) that could specifically bind to Cas13a protein and a protospacer domain (Guide RNA, gRNA) which can specifically recognize the target sites of SARS.Fig. 2 Working principle of the collaborative system in detecting SARS-CoV-2 RNA. (A) Structure of CRISPR-Cas13a system including Trac RNA, Guide RNA, and Cas13a protein. (B-C) Mechanism that Cas13a protein is activated by Guide RNA recognizing the target sites of SARS-CoV-2 RNA. (D) Functionalization of graphene. (E) Immobilization of RP. (F) RP is cut away from graphene by activated RNP. (G) Mechanism of activated RNP (ARNP)-cleavage-mediated signal change. Ⅰ, Ⅱ and Ⅲ are respectively the distribution of Fermi level from the immobilization to the cleavage of RP; Two inserts from left to right are the characteristic signal changes corresponding to Fermi level changes. Fig. 2 -COV-2 RNA (Fig. 2A). In the absence of the target RNAs, CRISPR combines with Cas13a to form the nuclease-inactive ribonucleo protein (RNP). When the gRNA recognizes the target sites of SARS-COV-2 RNA, the nonspecific RNA cleavage activity of CRISPR-Cas13a is activated (Fig. 2B and C). Single activated RNP can indiscriminately degrade tens of thousands of RP. The GFET was employed as the sensor to detect the SARS-COV-2 RNA since graphene's ambipolar electric field effect could make the detected signal directional. The characterizations of GFET are detailed described in text S2 and Figs. S12–S13. The GFET was functionalized by PBASE to immobilize the RP(Fig. 2D–E). Electron doping to P-type graphene by the immobilized RP leads to an upward shift of the graphene Fermi level, resulting in an E signal (Figs. 2G, 1–2). In the presence of SARS-COV-2 RNA, the immobilized RP was massively cleaved by the activated RNP and left the graphene surface (Fig. 2F). RP's electron doping to graphene is sharply attenuated and the graphene Fermi level shifts downward, resulting in a “Cut-E″ signal (Fig. 2G, 2–3). Since untargeted molecules do not have the targeted site for the pre-designed CRISPR, there is no activated RNP to cleavage the RP to generate a “Cut-E″ signal. Instead, untargeted molecules can unspecifically adsorb on graphene surface and act like the RP to electronically dope the graphene, resulting in an enlarged E signal. Experimentally, the E signal, enlarged E signal and “Cut-E″ signal can be determined by reading the Dirac point from the transfer characteristic curve of the GFET (text S3 and Fig. S14). The left shift of the Dirac point corresponds to the E signal or the enlarged E signal, and the right shift of the Dirac point corresponds to the “Cut-E″ signal. In conclusion, the directional “Cut-E″ signal generated by cutting the RP-induced E signal via SARS-COV-2 RNA-activated RNP is the indicator of whether the collaborative system has specifically and successfully detected the SARS-COV-2 RNA. It is worth mentioning that, unlike the RT-PCR method where the RP is uniformly dispersed in a solution with large size, the RP in this collaborative system is immobilized on the graphene surface with limited location, which facilitates its can be cleavaged more intensively by the activated RNP. More importantly, the RP consisted of 6 ribonucleotides close to graphene and 16 deoxyribonucleotides away from graphene (Fig. 2E). If one ribonucleotide in the RP is cleavaged, at least 16 deoxyribonucleotides can be removed from the graphene surface (Fig. 2F). This dramatically increases signal generation efficiency compared to the RP composed entirely by ribonucleotides. The signal detection mechanism caused by the specific super cleavage properties of activated RNP to RP and the optimized RP structure bring a directional “big subtraction” signal for detecting SARS-COV-2 RNA. The in vitro transcriptions N gene RNA of SARS-CoV-2 was first employed to evaluate this collaborative system's detection performance (text S4). Activating more RNP by one SARS-CoV-2 N gene RNA facilities the higher efficiency to cleavage RP and thus improves sensitivity (Fozouni et al., 2021). Here, three designed CRISPRs whose gRNA targets different sites of the same N gene RNA were used to complex with Cas13a protein to form three different RNP populations, making a single N gene RNA active 3 RNP (Fig. 3 A). In the other hand, large E signal is facilized to detect and distinguish the “Cut-E″ signal caused by SARS-COV-2 RNA. Here, the RP was immobilized on the graphene surface by a two-step method rather than the conventional one-step method as described in method section, generating a larger E signal (Fig. 3B). The RP pre-reaction in the first step can extrude water molecules out of the graphene surface, which greatly reduces the repulsion of the water molecules on the graphene surface to the RP, thus improving the diffusion speed and the reaction efficiency of the RP in the second step (Elder and Jayaraman, 2013). To measure the sensitivity of this collaborative system, the immobilized RP were respectively cleavaged by the RNP activated by N gene RNA with concentrations changing from 0.1 aM to 100 aM. Each experiment was repeated ten times. After the RP & GFET reacted with the RNP activated by 0.1 aM N gene RNA (about 3 copies in 50 μL solution) for 2 h, the response signal (Cut-E/E %) was hardly generated (Fig. 3C and Fig. S15). The reason may be that such a low density of the activated RNPs make it difficult to fully diffuse to the graphene surface to generate an obviously “Cut-E″ signal within 2 h. Moreover, it is difficult to accurately pick up the N gene RNA molecules with 3 copies by dividing a certain volume of the solution. After the RP & GFET reacted with the RNP activated by 0.25 aM N gene RNA for 2 h, an unmistakable “Cut-E″ signal with 18 mV was generated (Fig. S16A), and the average cleavage efficiency of activated RNP to RP is 39.7% (Figs. 3C and 0.25 aM). The monitored results show a significant difference between 0.25 aM N gene and RNP with ***p < 0.000001, proving the detection limit of this collaborative system for detecting N gene RNA can easily reach 0.15 copies/ μL. 0.5 aM N gene RNA activats more RNP and thus cleavage more RP, resulting in a larger “Cut-E″ signal 27 mV (Fig. S16B), and the average cleavage efficiency of the activated RNP to RP is 47.1% (Figs. 3C and 0.5 aM). Logically, more E signals will be cut by growingly activated RNP and larger “Cut-E″ signals should generate, as N gene RNA concentration increases. However, when the RP & GFET reacted with the RNP activated by 1 aM N gene RNA and 10 aM N gene RNA for 2 h, the “Cut-E″ signal presented by 23 mV and 26 mV, respectively (Figs. S17A–17B), and the average cleavage efficiencies of activated RNP to RP were 47.3% and 49% (Fig. 1, Fig. 3 aM-10 aM). There is almost no significant increase in the “Cut-E″ signal compared to the signal caused by the RNP activated by 0.5 aM N gene RNA. This can be explained by that the space-occupying effect make the density of activated RNP reach an upper limit on the surface of RP with the increasing activated RNP. The upper density makes the additional activated RNP be blocked in the outer layer and cannot directly interact with the RP.Fig. 3 SARS-CoV-2 N gene RNA testing. (A) Genome map of SARS-CoV-2 RNA and the target sites (TS) of N gene RNA to hybridize with gRNA. (B) Comparison of E signal induced by RP respectively in a one-step method and a two-step method to immobilize the RP. (C) Response of this collaborative system to detect SARS-CoV-2 N gene RNA with gradient concentration; Response signal (Cut-E/E %) were compared to the RNP control through an analysis of covariance (ANOVA): p > 0.5, ****p < 0.000001, P < 0.05 indicates that there is a significant difference between N gene RNA and RNP. (D) Verifying the space-occupying effect: activated RNP sequentially cleavages the RP on the same device. Fig. 3 The same device's E signal was orderly cut by the RNPs respectively activated by the N gene RNA with the concentration changing from 0.25 aM to 10 aM to verify the space-occupancy effect. Between every two cutting processes, there is a washing step. RNPs activated by a previous concentration of N gene RNA are washed off the RP surface while making room for RNPs activated by the next concentration of N gene RNA, thereby further cleavage the RP. In the experiment, an E signal of about 53 mV is formed by the immobilization of RP (Fig. 3D; PBASE-RP). After the RP & GFET reacted with the RNP activated by 0.25 aM N gene RNA for 2 h, an unmistakable “Cut-E″ signal with 16 mV occurred, and a 32% E signal was cut off (Fig. 3D; 0.25 aM). Following the wash step, the RNP activated by 0.5 aM N gene RNA was used to furtherly cut the E signal of the same device. The results showed that a 71% E signal was cut off (Fig. 3D; 0.5 aM). By the same steps, the E signal is 100% cut off after the RP & GFET reacted with RNP activated by 1 aM N gene RNA (Fig. 3D; 1 aM). The space-occupancy effect was effectively proved. By the way, there is no further change of Dirac point after the RP & GFET reacted with the RNP activated by 10 aM N gene RNA (Fig. 3D; 10 aM), demonstrating the activated RNP system does not contribute any unspecific effect to graphene, just work to cleavage RP. Given its outstanding performance in detecting N gene RNA, the collaborative system was further applied to detect SARS-CoV-2 full-length RNA (Fig. 4 A). Unlike N gene RNA, full-length RNA contains approximately 30,000 nucleotides (Chan et al., 2020; Liu et al., 2020). To fully exploit the signal-generating ability of these 30,000 nucleotides, another CRISPR whose gRNA targets the E gene RNA was added. The RNAs extracted from 6 RT-PCR confirmed positive samples were employed to characterize this collaborative system's sensitivity and reliability (Fig. S18 and text S5). The results show that this collaborative system exhibits the negative response signal in detecting 0.1 aM full-length RNA, which means enlarged E signals occurred and no generation of “Cut-E″ signals (Fig. 4B; 0.1 aM). We speculate that the enlarged E signal is caused by the strong unspecific adsorption between graphene and full-length RNA. Without sufficient activated RNP to fully cleavage these full-length RNAs, it can play the same role as the RP to contribute electron doping to graphene. Recently, some literature has verified this strong unspecific adsorption by studying the behaviour of oligonucleotides at the graphene−water interface using theoretical simulation (Cortés-Arriagada, 2021; Manna and Pati, 2013; Ranganathan et al., 2016). Here, we experimentally explored this unspecific adsorption. In the experiment, the full-length RNA was directly added to the graphene surface where there was no immobilization of specific probes. After reacting for 2 h, the graphene surface was flushed three times with RNase water. The results show that GFET generated a significant E signal about 25 mV (Fig. 4C), which proves full-length RNA can firmly bond to the graphene surface by unspecific interaction. This unspecific adsorption makes the traditional FET based on the hybridization-driven “small addition” signal generation mechanism face the challenge of signal purity in detecting long-chain nucleic acids, which simultaneously proves the superiority of the directional “big subtraction” signal generation mechanism of this collaborative system. This collaborative system exhibited approximately 42% E signal was cut off by the RNP activated by 0.25 aM full-length RNA (Fig. 4B; 0.25 aM). In the range of 0.25 aM–10 aM, the response signal grows with the increasing full-length RNA concentration (Fig. 4B; 0.25 aM–10 aM). Like N gene RNA assay results, the growing response signal stopped when RNP was activated by a higher concentration of 100 aM due to the space-occupying effect (Fig. 4B; 100 aM). 0.25 aM full-length RNAs extracted from the other five positive samples were independently tested. The results showed that these response signal's intensities were all about 44%, proving that this collaborative system has high sensitivity and excellent reproducibility for SARS-CoV-2 RNA detection (Fig. 4D; PS1-PS5). Notably, at any concentration, full-length RNA-activated RNP produced a more significant response signal than N gene-activated RNP in this collaborative system, which is caused by the additive effect of the CRISPR that targets E gene RNA. The RNAs extracted from six negative samples were also tested. Because these RNAs do not have the targeting site of the pre-designed CRISPR in this collaborative system, the RNP had not been activated and thus there is no generation of the “Cut-E″ signal instead of generating enlarged E signals by the unspecific adsorption (Fig. 4C; NS1-NS6). Thus the false-positive signal is eliminated by the opposite signal direction between the “Cut-E″ signal of targets and the enlarged E signals of untargets. The possibilities that RP has flushed away in the wash step or that the reading solution degraded RP were studied to prove that the “Cut-E″ signal is specifically generated by the full-length RNA. Since the collaborative system works in a 1 × NEB buffer where the salt concentration is 170 mM, the Debye length of the sensing graphene surface is only 0.74 nm, much shorter than the length of the RP (1.2 nm) (Stern et al., 2007). If we read the signal in the 1 × NEB buffer directly, some RP will exceed the Debye length and thus part RP cannot be effectively detected. Therefore, the 1× NEB buffer was flushed away and replaced by a 0.1× PBS buffer (reading solution) with a Debye length of 2.5 nm in the reading signal step. This would lead us to suspect that the RP was flushed away during the washing step instead of being cleavaged by the activated RNP during the step of reading the “Cut-E″ signal. To prove that, firstly, GFET & RP was flushed three times by reading solution. The result shows the Dirac points of GFET before and after flushing are the same, and there is almost no generation of the “Cut-E″ signal (Fig. 4E; RP-flushing), which proves that RP can be firmly bound to the graphene surface by PBASE and cannot be washed off. Secondly, the RNP activated by 0.5 aM full-length RNA was added and reacted with GFET & RP for 2 h. A significant “Cut-E″ signal with 21 mV was generated, and about 46% E signal was cut off (Fig. 4E; flushing-RNP). Finally, GFET & RP was rewashed three times by reading solution, the position of Dirac points had not changed, and there was no generation of the “Cut-E″ signal (Fig. 4E; RNP-flushing). The above progressive experiments demonstrated that the RP could not be affected by flushing steps. The Dirac point of GFET & RP were compared before and after reacting with the reading solution for 2 h h. The absence of a significant Dirac point shift suggests the RP cannot be degraded by the reading solution (Fig. 4F). The collaborative system also detected the RNAs extracted from three other SARS-like coronaviruses. Like negative sample assay results, the response signal is negative, suggesting the enlarged E signals were generated without generating any “Cut-E″ signal (Fig. 4G). In addition, the GFET sensor can still perform good ambipolar electric field effect and shows the similar response to the targets as the fresh GFET after 3 weeks storage, as shown in Fig. S19.Fig. 4 Detection of SARS-CoV-2 full-length RNA. (A)Workflow of obtaining SARS-CoV-2 RNA from positive samples. (B) Response signal of this collaborative system to detect SARS-COV-2 RNA with increasing concentrations from 0.1 aM to 100 aM. (C) Untargeted molecules can absorb on graphene, causing enlarged E signal. (D) Response signals to SARS-CoV-2 RNA extracted from positive samples PS1−PS5 and negative samples NS1−NS6. (E-F) Specificity verification: RP cannot be degraded by test buffer or RNP mixture. (G) Specificity verification: Response signals to the RNAs of SARS-like coronavirus. Fig. 4 4 Discussion Up to now, changing the type or configuration of the probe molecules (Cai et al., 2014; Hajian et al., 2019; Hwang et al., 2018; Kong et al., 2021; Mei et al., 2018) or changing the sensing material (Hwang et al., 2020; Liang et al., 2020; Liu et al., 2019) of the FET sensor is still the mainstream research direction to enhance FET's performance in detecting nucleic acid. In these studies, the probe molecules are immobilized on the surface of the sensing material to recognize and capture target nucleic acid molecules through complementary base pairing (Fu et al., 2017). The hybridization-driven mechanism for grabbing target molecules results in that one target molecule generate one unit signal and multiple target molecules linearly generates multiple unit signals. Here, we call the detection signal generated by capturing the target molecule in the hybridization mechanism as a “small addition” signal. Although the sensitivity of FET sensors has been improved a lot by these research works, the “small addition” signal generation mechanism prevents these sensors from showing a breakthrough sensitivity. Therefore, developing a new signal generation mechanism with higher efficiency is desirable. Here, thanks to the CRISPR-Cas13a system's strong cleavage properties to RP, the coupling of multiple CRISPRs, the optimized structure of RP and the highly sensitive electrical properties of GFETs, this collaborative system presented a “big subtraction” signal generating mechanism and obtained a higher intensity directional signal (more than 30%) compared with the detection signal (less than 7%) in similar graphene field-effect transistor (Wang et al., 2022a), and exhibits a lower detection limit of 0.15 copies/ μL, compared with the pure CRISPR/Cas13a technology (100 copies/ μL) (Fozouni et al., 2021), in detecting SARS-CoV-2 RNA. For the GFET to detect the real samples of SARS-CoV-2, it must be considered that the RNA of SARS-CoV-2 is a single-stranded structure containing nearly 30,000 nucleic acid molecules, and the nucleosides of these 30,000 nucleic acids are very easy to form a non-cohesive valence binding with graphene. That leads to the fact that even without probe molecules, the RNA of SARS-CoV-2 can still be captured by graphene forcefully. The probe molecules of the traditional FET for detecting SARS-CoV-2 are usually composed of only dozens of nucleic acid molecules, but the corresponding dozens of target sites are hidden in a single-stranded structure containing 30,000 nucleic acid molecules. This makes it hard to believe that it is the probe molecules rather than graphene are responsible for capturing the target molecules to generate the detection signal. Therefore, it is necessary to develop a new signalling mechanism with higher specificity. Here, benefiting from the programmability of the CRISPR and the right-shifted characteristic detection signal (“Cut-E″ signal), untargeted RNAs adsorbed on the graphene produces the left-shifted characteristic enlarged E signal, but not the right-shifted characteristic “Cut-E signal”. This mechanism makes the collaborative system achieve highly specificity to detecting SARS-CoV-2 RNA. The volume of target RNAs extracted from clinical COVID-19 positive samples by commercial kits is usually from several microliters to several milliliters. At present, the size of the sensing material of the reported FET nucleic acid sensors is in the order of micrometers, and the scale of the liquid pool above the FET is designed to be in the order of millimeters (Dai et al., 2021; Kong et al., 2021; Seo et al., 2020; Wang et al., 2022a). By simple calculation, we can easily find that the area of the sensing material only accounts for a few thousandths of the liquid pool's surface, which means that the probability of a single SARS-CoV-2 RNA diffusing from the space of the liquid pool to the surface of the sensing material may be only one in a few million, or even lower. This means that even after a long diffusion time, only partial or even no target RNAs can diffuse to the surface of the sensing material and be caught by the probe molecules. Here, we believe that field-effect transistors based on micron-scale sensing materials have played an irreplaceable and positive role in scientific exploration at the laboratory level. However, the existing deficiencies make it necessary to develop the field-effect transistors based on larger-scale sensing materials to meet the needing for practical clinical testing. Here, benefiting from the large-area graphene with a centimeter diameter and a liquid storage tank of the same size (Fig. S7), the SARS-CoV-2 RNA & CRISPR-Cas13a analysis solution can fully and directly contact graphene & RP, which dramatically improves the probability of SARS-CoV-2 RNA & CRISPR-Cas13a diffusing from the liquid pool space to graphene & RP surface. This enables the collaborative system to achieve high-efficiency detection of SARS-CoV-2 RNA. 5 Conclusion In the proposed collaborative system, SARS-CoV-2 RNAs can generate high-intensity signals with right shift features, and untargeted molecules generate left shift signals, eliminating the false-negative and false-positive signals effectively. The high sensitivity (0.25 aM) and high specificity caused by the opposite direction of the signals between targeted molecules and untargeted molecules, are not only of great significance for GFET to realize the effective detection of SARS-CoV-2 RNA but also very enlightening for GFET to realize the analysis and detection of other nucleic acid-related diseases. All the nucleic acid-related diseases are highly promising to be detected by our collaborative system by replacing the CRISPR-Cas system that specifically recognizes the related DNA/RNA targets. In addition, eliminating the instability of the field-effect transistor by hydrophobic treatment of it's substrate should greatly promote the engineering application of the field-effect transistor. CRediT authorship contribution statement Yang Sun: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Visualization, Supervision, Project administration, Funding acquisition, Writing – original draft, Writing – review & editing. Cheng Yang: Conceptualization, Data curation, Writing – original draft, Writing – review & editing, Supervision. Xiaolin Jiang: Validation, Resources. Pengbo Zhang: Methodology, Software. Shuo Chen: Methodology, Software, Formal analysis. Fengxia Su: Resources. Hui Wang: Resources. Weiliang Liu: Validation, Writing – original draft. Xiaofei He: Investigation. Lei Chen: Supervision, Project administration. Baoyuan Man: Conceptualization, Project administration. Zhengping Li: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration, Funding acquisition. 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 will be made available on request. Acknowledgements This study was funded by 10.13039/501100001809 National Natural Science Foundation of China (21775012). Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bios.2022.114979. ==== Refs References Ang P.K. Chen W. Wee A.T.S. Loh K.P. Solution-gated epitaxial graphene as pH sensor J. Am. Chem. Soc. 130 44 2008 14392 14393 18850701 Arons M.M. Hatfield K.M. Reddy S.C. Kimball A. James A. Jacobs J.R. Taylor J. Spicer K. Bardossy A.C. Oakley L.P. Presymptomatic SARS-CoV-2 infections and transmission in a skilled nursing facility N. Engl. J. Med. 382 22 2020 2081 2090 32329971 Bokelmann L. Nickel O. Maricic T. Pääbo S. Meyer M. Borte S. Riesenberg S. Point-of-care bulk testing for SARS-CoV-2 by combining hybridization capture with improved colorimetric LAMP Nat. Commun. 12 1 2021 1 8 33397941 Bonanni A. Chua C.K. Zhao G. Sofer Z.k. Pumera M. Inherently electroactive graphene oxide nanoplatelets as labels for single nucleotide polymorphism detection ACS Nano 6 10 2012 8546 8551 22992186 Boyton R.J. Altmann D.M. The immunology of asymptomatic SARS-CoV-2 infection: what are the key questions? Nat. Rev. Immunol. 21 12 2021 762 768 34667307 Cai B. Wang S. Huang L. Ning Y. Zhang Z. Zhang G.-J. Ultrasensitive label-free detection of PNA–DNA hybridization by reduced graphene oxide field-effect transistor biosensor ACS Nano 8 3 2014 2632 2638 24528470 Chan J.F.-W. Kok K.-H. Zhu Z. Chu H. To K.K.-W. Yuan S. Yuen K.-Y. Genomic characterization of the 2019 novel human-pathogenic coronavirus isolated from a patient with atypical pneumonia after visiting Wuhan Emerg. Microb. Infect. 9 1 2020 221 236 Chen F. Qing Q. Xia J. Li J. Tao N. Electrochemical gate-controlled charge transport in graphene in ionic liquid and aqueous solution J. Am. Chem. Soc. 131 29 2009 9908 9909 19572712 Chiu U.-T. Lee B.-F. Hu S.-K. Yu T.-F. Lee W.-Y. Chao L. Graphene memory based on a tunable nanometer-thin water layer J. Phys. Chem. C 123 17 2019 10842 10848 Cohen A.N. Kessel B. Milgroom M.G. Diagnosing SARS-CoV-2 infection: the danger of over-reliance on positive test results medRxiv 2020 10.1101/2020.04.26.20080911 Cortés-Arriagada D. Intermolecular driving forces on the adsorption of DNA/RNA nucleobases to graphene and phosphorene: an atomistic perspective from DFT calculations J. Mol. Liq. 325 2021 115229 Dai C. Guo M. Wu Y. Cao B.-P. Wang X. Wu Y. Kang H. Kong D. Zhu Z. Ying T. Ultraprecise antigen 10-in-1 pool testing by multiantibodies transistor assay J. Am. Chem. Soc. 143 47 2021 19794 19801 34792340 Das A. Pisana S. Chakraborty B. Piscanec S. Saha S.K. Waghmare U.V. Novoselov K.S. Krishnamurthy H.R. Geim A.K. Ferrari A.C. Monitoring dopants by Raman scattering in an electrochemically top-gated graphene transistor Nat. Nanotechnol. 3 4 2008 210 215 18654505 Dollekamp E. Bampoulis P. Faasen D.l.P. Zandvliet H.J. Kooij E.S. Charge induced dynamics of water in a graphene–mica slit pore Langmuir 33 43 2017 11977 11985 28985466 Eisenstein M. Seven technologies to watch in 2022 Nature 601 7894 2022 658 661 35079149 Elder R.M. Jayaraman A. Structure and thermodynamics of ssDNA oligomers near hydrophobic and hydrophilic surfaces Soft Matter 9 48 2013 11521 11533 Fozouni P. Son S. de León Derby M.D. Knott G.J. Gray C.N. D'Ambrosio M.V. Zhao C. Switz N.A. Kumar G.R. Stephens S.I. Amplification-free detection of SARS-CoV-2 with CRISPR-Cas13a and mobile phone microscopy Cell 184 2 2021 323 333 e329 33306959 Fu W. Jiang L. van Geest E.P. Lima L.M. Schneider G.F. Sensing at the surface of graphene field‐effect transistors Adv. Mater. 29 6 2017 1603610 Gao Z. Xia H. Zauberman J. Tomaiuolo M. Ping J. Zhang Q. Ducos P. Ye H. Wang S. Yang X. Detection of sub-fM DNA with target recycling and self-assembly amplification on graphene field-effect biosensors Nano Lett. 18 6 2018 3509 3515 29768011 Gao J. Wang C. Chu Y. Han Y. Gao Y. Wang Y. Wang C. Liu H. Han L. Zhang Y. Graphene oxide-graphene Van der Waals heterostructure transistor biosensor for SARS-CoV-2 protein detection Talanta 240 2022 123197 Gootenberg J.S. Abudayyeh O.O. Lee J.W. Essletzbichler P. Dy A.J. Joung J. Verdine V. Donghia N. Daringer N.M. Freije C.A. Nucleic acid detection with CRISPR-Cas13a/C2c2 Science 356 6336 2017 438 442 28408723 Hajian R. Balderston S. Tran T. DeBoer T. Etienne J. Sandhu M. Wauford N.A. Chung J.-Y. Nokes J. Athaiya M. Detection of unamplified target genes via CRISPR–Cas9 immobilized on a graphene field-effect transistor Nat. Biomed. Eng 3 6 2019 427 437 31097816 Hinnemo M. Zhao J. Ahlberg P. Hagglund C. Djurberg V. Scheicher R.H. Zhang S.-L. Zhang Z.-B. On monolayer formation of pyrenebutyric acid on graphene Langmuir 33 15 2017 3588 3593 28350965 Hong Y. Wang S. Li Q. Song X. Wang Z. Zhang X. Besenbacher F. Dong M. Interfacial icelike water local doping of graphene Nanoscale 11 41 2019 19334 19340 31423505 Hu S.-K. Lo F.-Y. Hsieh C.-C. Chao L. Sensing ability and formation criterion of fluid supported lipid bilayer coated graphene field-effect transistors ACS Sens. 4 4 2019 892 899 30817891 Hwang M.T. Wang Z. Ping J. Ban D.K. Shiah Z.C. Antonschmidt L. Lee J. Liu Y. Karkisaval A.G. Johnson A.T.C. DNA nanotweezers and graphene transistor enable label‐free genotyping Adv. Mater. 30 34 2018 1802440 Hwang M.T. Heiranian M. Kim Y. You S. Leem J. Taqieddin A. Faramarzi V. Jing Y. Park I. van der Zande A.M. Ultrasensitive detection of nucleic acids using deformed graphene channel field effect biosensors Nat. Commun. 11 1 2020 1 11 31911652 Kaminski M.M. Abudayyeh O.O. Gootenberg J.S. Zhang F. Collins J.J. CRISPR-based diagnostics Nat. Biomed. Eng 5 7 2021 643 656 34272525 Kevadiya B.D. Machhi J. Herskovitz J. Oleynikov M.D. Blomberg W.R. Bajwa N. Soni D. Das S. Hasan M. Patel M. Diagnostics for SARS-CoV-2 infections Nat. Mater. 20 5 2021 593 605 33589798 Kong D. Wang X. Gu C. Guo M. Wang Y. Ai Z. Zhang S. Chen Y. Liu W. Wu Y. Direct SARS-CoV-2 nucleic acid detection by Y-shaped DNA dual-probe transistor assay J. Am. Chem. Soc. 143 41 2021 17004 17014 34623792 Lee M.J. Choi J.S. Kim J.-S. Byun I.-S. Lee D.H. Ryu S. Lee C. Park B.H. Characteristics and effects of diffused water between graphene and a SiO2 substrate Nano Res. 5 10 2012 710 717 Lee D. Ahn G. Ryu S. Two-dimensional water diffusion at a graphene–silica interface J. Am. Chem. Soc. 136 18 2014 6634 6642 24730705 Liang Y. Xiao M. Wu D. Lin Y. Liu L. He J. Zhang G. Peng L.-M. Zhang Z. Wafer-scale uniform carbon nanotube transistors for ultrasensitive and label-free detection of disease biomarkers ACS Nano 14 7 2020 8866 8874 32574035 Liu J. Chen X. Wang Q. Xiao M. Zhong D. Sun W. Zhang G. Zhang Z. Ultrasensitive monolayer MoS2 field-effect transistor based DNA sensors for screening of down syndrome Nano Lett. 19 3 2019 1437 1444 30757905 Liu R. Han H. Liu F. Lv Z. Wu K. Liu Y. Feng Y. Zhu C. Positive rate of RT-PCR detection of SARS-CoV-2 infection in 4880 cases from one hospital in Wuhan, China, from Jan to Feb 2020 Clin. Chim. Acta 505 2020 172 175 32156607 Long Q.-X. Tang X.-J. Shi Q.-L. Li Q. Deng H.-J. Yuan J. Hu J.-L. Xu W. Zhang Y. Lv F.-J. Clinical and immunological assessment of asymptomatic SARS-CoV-2 infections Nat. Med. 26 8 2020 1200 1204 32555424 Manna A.K. Pati S.K. Theoretical understanding of single-stranded DNA assisted dispersion of graphene J. Mater. Chem. B 2013 91 100 32260616 Mao S. Chang J. Pu H. Lu G. He Q. Zhang H. Chen J. Two-dimensional nanomaterial-based field-effect transistors for chemical and biological sensing Chem. Soc. Rev. 46 22 2017 6872 6904 28933459 Mei J. Li Y.-T. Zhang H. Xiao M.-M. Ning Y. Zhang Z.-Y. Zhang G.-J. Molybdenum disulfide field-effect transistor biosensor for ultrasensitive detection of DNA by employing morpholino as probe Biosens. Bioelectron. 110 2018 71 77 29602033 Nath H. Mallick A. Roy S. Sukla S. Basu K. De A. Biswas S. Archived dengue serum samples produced false-positive results in SARS-CoV-2 lateral flow-based rapid antibody tests J. Med. Microbiol. 70 6 2021 001369 Novoselov K.S. Geim A.K. Morozov S.V. Jiang D.-e. Zhang Y. Dubonos S.V. Grigorieva I.V. Firsov A.A. Electric field effect in atomically thin carbon films Science 306 5696 2004 666 669 15499015 Orooji Y. Sohrabi H. Hemmat N. Oroojalian F. Baradaran B. Mokhtarzadeh A. Mohaghegh M. Karimi-Maleh H. An overview on SARS-CoV-2 (COVID-19) and other human coronaviruses and their detection capability via amplification assay, chemical sensing, biosensing, immunosensing, and clinical assays Nano-Micro Lett. 13 1 2021 1 30 Ranganathan S.V. Halvorsen K. Myers C.A. Robertson N.M. Yigit M.V. Chen A.A. Complex thermodynamic behavior of single-stranded nucleic acid adsorption to graphene surfaces Langmuir 32 24 2016 6028 6034 27219463 Schedin F. Geim A.K. Morozov S.V. Hill E. Blake P. Katsnelson M. Novoselov K.S. Detection of individual gas molecules adsorbed on graphene Nat. Mater. 6 9 2007 652 655 17660825 Seo G. Lee G. Kim M.J. Baek S.-H. Choi M. Ku K.B. Lee C.-S. Jun S. Park D. Kim H.G. Rapid detection of COVID-19 causative virus (SARS-CoV-2) in human nasopharyngeal swab specimens using field-effect transistor-based biosensor ACS Nano 14 4 2020 5135 5142 32293168 Sitjar J. Liao J.-D. Lee H. Tsai H.-P. Wang J.-R. Liu P.-Y. Challenges of SERS technology as a non-nucleic acid or-antigen detection method for SARS-CoV-2 virus and its variants Biosens. Bioelectron. 181 2021 113153 Smyrlaki I. Ekman M. Lentini A. Rufino de Sousa N. Papanicolaou N. Vondracek M. Aarum J. Safari H. Muradrasoli S. Rothfuchs A.G. Massive and rapid COVID-19 testing is feasible by extraction-free SARS-CoV-2 RT-PCR Nat. Commun. 11 1 2020 1 12 31911652 Song L. Xiao G. Zhang X. Gao Z. Sun S. Zhang L. Feng Y. Luan G. Lin S. He M. A case of SARS-CoV-2 carrier for 32 days with several times false negative nucleic acid tests medRxiv 2020 10.1101/2020.03.31.20045401 Stern E. Wagner R. Sigworth F.J. Breaker R. Fahmy T.M. Reed M.A. Importance of the Debye screening length on nanowire field effect transistor sensors Nano Lett. 7 11 2007 3405 3409 17914853 Surkova E. Nikolayevskyy V. Drobniewski F. False-positive COVID-19 results: hidden problems and costs Lancet Respir. Med. 8 12 2020 1167 1168 33007240 Swank Z. Michielin G. Yip H.M. Cohen P. Andrey D.O. Vuilleumier N. Kaiser L. Eckerle I. Meyer B. Maerkl S.J. A high-throughput microfluidic nanoimmunoassay for detecting anti–SARS-CoV-2 antibodies in serum or ultralow-volume blood samples P Natl. A. Sci 118 18 2021 e2025289118 Talebian S. Wallace G.G. Schroeder A. Stellacci F. Conde J. Nanotechnology-based disinfectants and sensors for SARS-CoV-2 Nat. Nanotechnol. 15 8 2020 618 621 32728083 Tan Y.-W. Zhang Y. Bolotin K. Zhao Y. Adam S. Hwang E. Sarma S.D. Stormer H. Kim P. Measurement of scattering rate and minimum conductivity in graphene Phys. Rev. Lett. 99 24 2007 246803 Tang Y. Gao L. Feng W. Guo C. Yang Q. Li F. Le X.C. The CRISPR–Cas toolbox for analytical and diagnostic assay development Chem. Soc. Rev. 50 2021 11844 11869 34611682 Wang Z. Yi K. Lin Q. Yang L. Chen X. Chen H. Liu Y. Wei D. Free radical sensors based on inner-cutting graphene field-effect transistors Nat. Commun. 10 1 2019 1 10 30602773 Wang Z. Hao Z. Wang X. Huang C. Lin Q. Zhao X. Pan Y. A flexible and regenerative aptameric graphene–Nafion biosensor for cytokine storm biomarker monitoring in undiluted biofluids toward wearable applications Adv. Funct. Mater. 31 4 2021 2005958 Wang L. Wang X. Wu Y. Guo M. Gu C. Dai C. Kong D. Wang Y. Zhang C. Qu D. Rapid and ultrasensitive electromechanical detection of ions, biomolecules and SARS-CoV-2 RNA in unamplified samples Nat. Biomed. Eng 6 3 2022 276 285 35132229 Wang Z. Hao Z. Yang C. Wang H. Huang C. Zhao X. Pan Y. Ultra-sensitive and rapid screening of acute myocardial infarction using 3D-affinity graphene biosensor Cell Rep. Phys. Sci 5 3 2022 100855 Welch N.L. Zhu M. Hua C. Weller J. Mirhashemi M.E. Nguyen T.G. Mantena S. Bauer M.R. Shaw B.M. Ackerman C.M. Multiplexed CRISPR-based microfluidic platform for clinical testing of respiratory viruses and identification of SARS-CoV-2 variants Nat. Med. 2022 1-1 Woloshin S. Patel N. Kesselheim A.S. False negative tests for SARS-CoV-2 infection—challenges and implications N. Engl. J. Med. 383 6 2020 e38 32502334 Wu G. Tang X. Meyyappan M. Lai K.W.C. Doping effects of surface functionalization on graphene with aromatic molecule and organic solvents Appl. Surf. Sci. 425 2017 713 721 Yang Y. Peng Y. Lin C. Long L. Hu J. He J. Zeng H. Huang Z. Li Z.-Y. Tanemura M. Human ACE2-functionalized gold “virus-trap” nanostructures for accurate capture of SARS-CoV-2 and single-virus SERS detection Nano-Micro Lett. 13 1 2021 1 13 Zhang Y. Yuan S. Zhou W. Xu J. Li Y. Spectroscopic evidence and molecular simulation investigation of the π–π interaction between pyrene molecules and carbon nanotubes J. Nanosci. Nanotechnol. 7 7 2007 2366 2375 17663254	36463654	PMC9710152	NO-CC CODE	2022-12-13 23:17:20	no	Biosens Bioelectron. 2023 Feb 15; 222:114979	utf-8	Biosens Bioelectron	2,022	10.1016/j.bios.2022.114979	oa_other
==== Front BMJ Case Rep BMJ Case Rep bmjcr bmjcasereports BMJ Case Reports 1757-790X BMJ Publishing Group BMA House, Tavistock Square, London, WC1H 9JR 36442910 bcr-2022-253383 10.1136/bcr-2022-253383 Case Reports: Reminder of important clinical lesson 2474 1524 1307 1560 1372 1330 196 Post-COVID mRNA vaccine myocarditis in children: report of two cases Shamekh Ahmed 1 Powell Colin 2 Ashabani Abdusamea 1 http://orcid.org/0000-0002-8676-0953 Abdelgadir Ibtihal Siddiq 3 1 Paediatric Emergency, Sidra Medicine, Doha, Ad Dawhah, Qatar 2 Paediatrics, Cardiff and Vale University Health Board, Cardiff, UK 3 Paediatric Emergency, Sidra Medical and Research Center, Doha, Qatar Correspondence to Dr Ahmed Shamekh; [email protected] 2022 28 11 2022 28 11 2022 15 11 e25338317 11 2022 © BMJ Publishing Group Limited 2022. No commercial re-use. See rights and permissions. Published by BMJ. 2022 The SARS-COV-2 pandemic led to the development of several vaccinations to contain the disease. The Pfizer-BioNTech COVID-19 (BNT162b2) vaccine was recommended on May 2021 for use in children above 12 years and older. The vaccine is safe, well tolerated and highly effective. Initial reports showed no serious adverse events; however, cases of myocarditis in young healthy male adolescents have been reported. We report two cases of myocarditis/perimyocarditis who presented with short history of chest pain following administration of the second dose of the MRN COVID-19 vaccine. Cardiovascular system Paediatrics (drugs and medicines) Vaccination/immunisation COVID-19 ==== Body pmcBackground Although post-COVID-19 vaccination myocarditis is rare, it is a known side effect of the vaccine.1 Studies from the USA show that the risk of myocarditis after receiving mRNA-based COVID-19 vaccines is increased across all age and sex groups. Epidemiological studies reported an incidence of 20–30 per million patients, and the risk is highest after the second vaccination dose in adolescent males.2 3 The Centers for Disease Control and Prevention (CDC) recommends that COVID-19 primary series vaccines should be given to everyone aged 6 months and older with COVID-19 boosters for everyone 5 years and older.4 The vaccine shows excellent efficacy and safety outcomes, but the long-term side effects are still under investigations.5 We report two adolescent males who developed acute myocarditis, post Pfizer-BioNTech vaccine for COVID-19. The healthcare provider is to suspect myocarditis in healthy children who recently received COVID-19 vaccinations and presented with chest pain with or without cardiac symptoms. Case presentation Case 1 A boy in early adolescence with no significant medical history, apart from Perthe’s disease. He presented with a 1-day history of chest pain radiating to the left arm. There were no other symptoms. He received the second Pfizer-BioNTech vaccine dose 4 days before presentation to the paediatric emergency department (PED). On arrival, all his vital signs vitals were stable. His aural temperature was 37°C, heart rate was: 67 beats per minute (bpm), respiratory rate (RR) was 16 bpm, blood pressure was 126/59 mm Hg and oxygen saturation - SpO2 was 100% in air. His weight was 71 kg (90th centile), height was 165 cm (25th centile) and body mass index (BMI) 26.1 kg/m2. Serum troponin I level, on arrival to PED, was raised at 586 ng/L (normal values <20 ng/L) (see figure 1). ECG showed diffuse ST segment changes and cardiacecho was normal with left ventricular ejection fraction of 67%. The cardiac MRI (cMRI) showed myocardial oedema, wall motion abnormality and transmural delayed enhancement at the apical lateral region of left ventricle supporting the diagnosis of myocarditis (figure 2). His COVID-19 nasal swab antigen test was negative; however, his COVID-19 IgG antibodies were positive with a titre of (binding antibody units -BAU)1930 BAU/mL. He was admitted to paediatric intensive care unit (PICU) for close monitoring of his clinical condition and serum troponin levels (figure 1). He received intravenous immunoglobulin, but this was stopped soon after been started as he developed severe chest pain. Figure 1 Case 1: troponin I titre reading recorded over 3 weeks. Figure 2 Case 1: Cardiac MRI (A) long axis view showing an abnormal high signal intensity of myocardium highlighted by the arrows. (B) four-chamber views post gadolinium showing late gadolinium enhancement. Case 2 A boy in early adolescence with a medical history of allergic rhinitis, presented with 1-day history of non-radiating central chest pain. He had no other symptoms. He had received the second Pfizer-BioNTech vaccine 5 days before attending the PED. On arrival, all his vital signs were stable: his aural temperature was 37°C heart rate was 78 bpm, respiratory, rate was 22, blood pressure was 115/46 mm Hg and SpO2 in air was 99%. He weighed 125 kg (>97th centile), his height was 183 cm (90th centile) and BMI 37.3 kg/m2. Serum troponin I on arrival to PED was raised at 23 390 ng/L (see figure 3) and creatine kinase was 1338 IU/L (normal value: 0–178 IU/L)(figure 4). ECG showed diffuse ST segment elevation, cardiac echo was normal and left ventricular ejection fraction was 64%. cMRI showed evidence of perimyocarditis with borderline impairment of right ventricular systolic function figure 5). His COVID-19 nasal swab antigen test was negative however his COVID-19 IgG antibodies were not tested. He did not require PICU admission. However, he was admitted to the cardiology ward for observation but did not receive any further active treatment such as intravenous immunoglobulin. Figure 3 Case 2: Troponin I titre recorded over 2 weeks. Figure 4 Case 2: CK level. CK, creatine kinase. Figure 5 Case 2: Cardiac MR (A) Short-axis view non-contrast showing abnormal high signal intensity involve the perimyocardium (B) - highlighted by the arrows. Long-axis view, late gadolinium enhancement showing perimyocardium enhancement of left ventricle. Outcome and follow-up Both patients recovered completely without sequalae, and discharged home within 5 days. The chest pain resolved within 48 hours for both patients. Blood result and cardiac enzymes improved before discharge. Follow-up with cardiology in 4 weeks showed complete recovery and good cardiac function for both patients. Discussion In December 2020, the Food and Drug Administration (FDA) issued an emergency use authorisation (EUA) for the Pfizer-BioNTech COVID-19 (BNT162b2) vaccine for young adults aged ≥16 years. In May 2021, FDA expanded the EUA for the Pfizer-BioNTech COVID-19 vaccine to include adolescents aged 12–15 years.6 7 This was revised recently by the CDC, and the current recommendation is to use the COVID-19 primary series vaccines for everyone ages 6 months and older, and COVID-19 boosters for everyone ages 5 years and older. Most side effects of the vaccine were reported as mild that included fatigue, injection site pain, headaches, chills, fever, and muscle aches and pain. Systemic adverse reactions were more commonly reported after the second dose, with the usual onset of 1–4 days after vaccine receipt.6 This case report reports cases of a temporal but not proven causal association between the vaccine and myocarditis. Marshall et al described seven adolescents who developed myocarditis within 4 days following the second dose of the Pfizer-BioNTech COVID-19 vaccine.8 Larson et al reported further cases of myocarditis post-COVID-19 vaccination in Italy.9 A recent study from Israel reported that mRNA COVID-19 vaccination was associated with an increased risk for myocarditis (RR 3.24; 95% CI 1.55 to 12.44); with stronger risk for myocarditis from the COVID-19 (RR 18.28, 95% CI 3.95 to 25.12)9 10 Patients with COVID-19 had nearly 16 times the risk for myocarditis compared with patients who did not have COVID-19.11 According to the US CDC, in children and young adults, myocarditis/pericarditis rates are up to 12.6 cases per million doses of second-dose mRNA vaccine. Usual presentation is 2–3 days after the second dose of the vaccine. Patients with myocarditis invariably presented with chest pain, elevated cardiac troponin levels, ST elevations in the ECG, with no evidence of acute COVID-19 or other viral infections.12 Male predominance in myocarditis/pericarditis cases has been described, however, the reason for that was not clear.12 The CDC Advisory Committee on Immunisation Practices COVID-19 advises that the benefits of vaccinating all recommended age groups with mRNA COVID-19 vaccine clearly outweigh the risks of COVID-19 infection.7 There have been no deaths reported in Qatar from a vaccine associated myocarditis. Learning points The incidence of myocarditis and pericarditis among young people following mRNA COVID-19 vaccine is low compared with the actual viral infection. Most cases are benign, responded well to non-steroid anti-inflammatory drugs and recovered completely without sequala. The vaccine is recommended for all eligible populations because the benefit is outweighing the risks. Children and young people who presented with chest pain, tachycardia and/or difficulty in breathing following COVID-19 vaccine should be assessed for the possibility of myocarditis. Ethics statements Patient consent for publication Consent obtained from parent(s)/guardian(s). Twitter: @Colin Powell Contributors: This is to state that all authors contributed to this submission. AS: involved in the medical care of the patients, data collection and writing the manuscript. CP: revised the manuscript critically and approved it for publication. AA: participated in the writing of the manuscript. ISA: participated in the data collection and writing the manuscript. All Authors approved the version of the manuscript to be published. Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Case reports provide a valuable learning resource for the scientific community and can indicate areas of interest for future research. They should not be used in isolation to guide treatment choices or public health policy. Competing interests: None declared. Provenance and peer review: Not commissioned; externally peer reviewed. ==== Refs References 1 Gellad WF. Myocarditis after vaccination against covid-19. BMJ 2021;16 :n3090. 10.1136/bmj.n3090 2 Oster ME, Shay DK, Su JR, et al . Myocarditis cases reported after mRNA-based COVID-19 vaccination in the US from December 2020 to August 2021. JAMA 2022;327 :331–40. 10.1001/jama.2021.24110 35076665 3 Power JR, Keyt LK, Adler ED. Myocarditis following COVID-19 vaccination: incidence, mechanisms, and clinical considerations. Expert Rev Cardiovasc Ther 2022;20 :241–51. 10.1080/14779072.2022.2066522 35414326 4 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. FDA news release, 2021. Available: https://www.fda.gov/news-events/press- [Accessed 22 Jun 2022]. 5 Polack FP, Thomas SJ, Kitchin N, et al . Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. N Engl J Med 2020;383 :2603–15. 10.1056/NEJMoa2034577 33301246 6 Wallace M, Woodworth KR, Gargano JW, et al . The Advisory Committee on Immunization Practices' Interim Recommendation for Use of Pfizer-BioNTech COVID-19 Vaccine in Adolescents Aged 12-15 Years - United States, May 2021. MMWR Morb Mortal Wkly Rep 2021;70 :749–52. 10.15585/mmwr.mm7020e1 34014913 7 Gargano JW, Wallace M, Hadler SC, et al . Use of mRNA COVID-19 Vaccine After Reports of Myocarditis Among Vaccine Recipients: Update from the Advisory Committee on Immunization Practices - United States, June 2021. MMWR Morb Mortal Wkly Rep 2021;70 :977–82. 10.15585/mmwr.mm7027e2 34237049 8 Marshall M, Ferguson ID, Lewis P, et al . Symptomatic acute myocarditis in 7 adolescents after Pfizer-BioNTech COVID-19 vaccination. Pediatrics 2021;148 :e2021052478. 10.1542/peds.2021-052478 34088762 9 Larson KF, Ammirati E, Adler ED, et al . Myocarditis after BNT162b2 and mRNA-1273 vaccination. Circulation 2021;144 :506–8. 10.1161/CIRCULATIONAHA.121.055913 34133884 10 Barda N, Dagan N, Ben-Shlomo Y, et al . Safety of the BNT162b2 mRNA Covid-19 vaccine in a nationwide setting. N Engl J Med Overseas Ed 2021;385 :1078–90. 10.1056/NEJMoa2110475 11 Boehmer TK, Kompaniyets L, Lavery AM, et al . Association Between COVID-19 and Myocarditis Using Hospital-Based Administrative Data - United States, March 2020-January 2021. MMWR Morb Mortal Wkly Rep 2021;70 :(2021):1228–32. 10.15585/mmwr.mm7035e5 34473684 12 Bozkurt B, Kamat I, Hotez PJ. Myocarditis with COVID-19 mRNA vaccines. Circulation 2021;144 :471–84. 10.1161/CIRCULATIONAHA.121.056135 34281357	36442910	PMC9710324	NO-CC CODE	2022-12-01 23:23:14	no	BMJ Case Rep. 2022 Nov 28; 15(11):e253383	utf-8	BMJ Case Rep	2,022	10.1136/bcr-2022-253383	oa_other
==== Front Environ Health Perspect Environ Health Perspect EHP Environmental Health Perspectives 0091-6765 1552-9924 Environmental Health Perspectives EHP12241 10.1289/EHP12241 Invited Perspective Invited Perspective: Uncovering the Hidden Burden of Tropical Cyclones on Public Health Locally and Worldwide https://orcid.org/0000-0002-7916-1717 Parks Robbie M. 1 Guinto Renzo R. 2 3 1 Department of Environmental Health Sciences, Mailman School of Public Health, Columbia University, New York, New York, USA 2 Planetary and Global Health Program, St. Luke’s Medical Centre College of Medicine–William H. Quasha Memorial, Quezon City, Philippines 3 Sunway Centre for Planetary Health, Sunway University, Selangor, Malaysia Address correspondence to Robbie M. Parks, 722 W. 168th St., New York, NY 10032 USA. Email: [email protected] 30 11 2022 11 2022 130 11 11130605 10 2022 21 10 2022 24 10 2022 https://ehp.niehs.nih.gov/about-ehp/license EHP is an open-access journal published with support from the National Institute of Environmental Health Sciences, National Institutes of Health. All content is public domain unless otherwise noted. The authors declare they have no conflicts of interest. Note to readers with disabilities: EHP strives to ensure that all journal content is accessible to all readers. However, some figures and Supplemental Material published in EHP articles may not conform to 508 standards due to the complexity of the information being presented. If you need assistance accessing journal content, please contact [email protected]. Our staff will work with you to assess and meet your accessibility needs within 3 working days. Refers to https://doi.org/10.1289/EHP11252 ==== Body pmcAs climate change and health researchers of Filipino heritage, we both are all too directly familiar with the cycles of devastation that typhoons (such as Typhoon Haiyan1 in 2013 and Typhoon Noru2 in 2022) and other tropical cyclones can leave in their path. For our families, the destruction wrought by tropical cyclones has manifested as downed trees, damaged or destroyed homes, and uprooted lives. Sadly, such traumatic experiences are commonplace when tropical cyclones make landfall, be it in the Philippines, China, Mexico, the United States, or elsewhere. Tropical cyclones, intense rotating storms that form over warm tropical waters, are characterized by a panoply of mortal hazards, including high wind speeds (≥39 mph), storm surges, rip currents, and heavy precipitation.3 The strongest tropical cyclones, with wind speeds of ≥74 mph, are known by different names throughout the world (hurricanes, typhoons, cyclones) but are universally devastating to societies.4 The health impacts can be wide-ranging; physical trauma can result directly from the force of tropical cyclones, whereas outbreaks of diarrheal and mosquito-borne diseases can result from unsafe or unhealthy conditions in their aftermath.5 Recent tropical cyclone seasons—which have yielded stronger,6 more active,7 and longer-lasting8 storms than previously recorded—demonstrate that these events will remain a critical public health concern; Hurricane Ian’s rapid intensification in North America during 2022 is a reminder of how climate change is modifying tropical cyclones.9 In a scoping review in this issue of Environmental Health Perspectives, Ghosh et al. assess the literature published through the end of 2021 evaluating the association of tropical cyclones with cardiovascular health.10 They found emerging evidence of short-term increases in adverse cardiovascular health outcomes following tropical cyclones, particularly in people with existing health conditions. Specifically, observed increases in cardiovascular-related illness11 and death12 have been linked to heart attacks and cardiac arrests from physical overexertion,13 increases in stress,14 and disrupted treatment of chronic cardiovascular conditions.15 Ghosh et al. also noted that the majority of the research they reviewed focused on the United States—a focus that should be expanded. Low- and middle-income countries (LMICs), such as the Philippines, a country of more than 100 million people with a year-round tropical cyclone season, are some of the most affected by climate change.16 Yet LMICs remain some of the least studied with respect to climate-related exposures and public health.17 More high-quality public health research on tropical cyclones focused on LMICs is essential10; from our experience, doing so requires greater cooperation between researchers from high-income countries and LMICs. Addressing worldwide knowledge gaps on the health impacts of tropical cyclones and other climate-related hazards also requires a multidisciplinary approach, involving diverse contributions from climate science, public health, and the social sciences to capture the lived experiences of people affected by tropical cyclones. The public health impact of tropical cyclones certainly reaches beyond injuries and cardiovascular diseases. Associations of tropical cyclones with neuropsychiatric conditions, respiratory diseases, and infectious and parasitic diseases are all evident in recent research.11,12 The long-term mental health consequences of repeated tropical cyclones need to be more completely understood,18 as well as the influence on childhood neurodevelopment, educational attainment, and DNA methylation. Even counting the number of dead after a tropical cyclone is a challenge; mortality estimates of the same hurricane can vary greatly, such as Hurricane Maria in 2017, for which official death counts were up to 70 times lower than the total number of estimated excess deaths.19,20 Research identifying those “hidden burdens” of tropical cyclones on life expectancy, illness, and mortality is critical to mitigating their overall impact worldwide. Although tropical cyclones will inevitably arrive each year, the worst consequences on public health and society are often avoidable with an equitable, long-term approach5; resilience to tropical cyclones is built over a long time via robust societal infrastructure, including adequate social services, housing stock, and power distribution. In the United States, the same tropical cyclone can affect communities differently, with differences likely driven by demographic, economic, and social factors21; in nonaffluent communities, impacts are often exacerbated by systemic inequity due to institutional neglect.22 The recovery after a tropical cyclone is also often inequitable, with federal aid and private insurance particularly difficult to obtain by Black and low-income individuals.23 Evacuation is a useful way to provide short-term relief from a tropical cyclone. However, this luxury is not available to all, due to a lack of early warning systems, financial resources, or adequate transport.24 Others are constrained by health conditions that they or a family member experience.25 Some simply have nowhere to go. In short, many cannot leave. It is almost too painful to read the news stories every time a powerful tropical cyclone makes landfall; we are all regularly reminded that typhoons and hurricanes are some of the most deadly and frequent climate-related hazards there are. For those of us not directly affected by extreme weather, it is much too easy to move on with our lives once the headlines fade. However, the long shadow of a tropical cyclone can destroy lives and result in hospitalization and death. Recovery and rehabilitation can seem to move in slow motion over the subsequent months and years. Giving appropriate attention and funding to understanding the full impacts of tropical cyclones on health locally and worldwide is critical to the fight for social, environmental, and climate justice. ==== Refs References 1. Lagmay AMF, Agaton RP, Bahala MAC, Briones JBLT, Cabacaba KMC, Caro CVC, et al. 2015. Devastating storm surges of Typhoon Haiyan. Int J Disaster Risk Reduct 11 :1–12, 10.1016/j.ijdrr.2014.10.006. 2. Broomby KJC. Noru became a super typhoon in 6 hours. Scientists say powerful storms are becoming harder to forecast. CNN. 1 October 2022. https://www.cnn.com/2022/10/01/asia/philippines-super-typhoon-noru-karding-climate-intl-dst-hnk/index.html [accessed 4 October 2022]. 3. Wallace JM, Hobbs PV. 2006. Atmospheric Science: An Introductory Survey, vol 92. 2nd ed. New York, New York: Academic Press. 4. Peduzzi P, Chatenoux B, Dao H, De Bono A, Herold C, Kossin J, et al. 2012. Global trends in tropical cyclone risk. Nat Clim Chang 2 (4 ):289–294, 10.1038/nclimate1410. 5. Shultz JM, Russell J, Espinel Z. 2005. Epidemiology of tropical cyclones: the dynamics of disaster, disease, and development. Epidemiol Rev 27 (1 ):21–35, PMID: , 10.1093/epirev/mxi011.15958424 6. Wang S, Toumi R. 2021. Recent migration of tropical cyclones toward coasts. Science 371 (6528 ):514–517, PMID: , 10.1126/science.abb9038.33510027 7. National Oceanic and Atmospheric Administration. 2020 Atlantic Hurricane Season Takes Infamous Top Spot for Busiest on Record. 10 November 2020. https://www.noaa.gov/news/2020-atlantic-hurricane-season-takes-infamous-top-spot-for-busiest-on-record [accessed 24 November 2020]. 8. Li L, Chakraborty P. 2020. Slower decay of landfalling hurricanes in a warming world. Nature 587 (7833 ):230–234, PMID: , 10.1038/s41586-020-2867-7.33177666 9. Bhatia KT, Vecchi GA, Knutson TR, Murakami H, Kossin J, Dixon KW, et al. 2019. Recent increases in tropical cyclone intensification rates. Nat Commun 10 (1 ):1–9, PMID: , 10.1038/s41467-019-11922-2.30602773 10. Ghosh AK, et al. 2022. Impact of hurricanes and associated extreme weather events on cardiovascular health: a scoping review. Environ Health Perspect 130 (11 ):116003, 10.1289/EHP11252. 11. Parks RM, Anderson GB, Nethery RC, Navas-Acien A, Dominici F, Kioumourtzoglou MA. 2021. Tropical cyclone exposure is associated with increased hospitalization rates in older adults. Nat Commun 12 (1 ):1–12, PMID: , 10.1038/s41467-021-21777-1.33397941 12. Parks RM, Benavides J, Anderson GB, Nethery RC, Navas-Acien A, Dominici F, et al. 2022. Association of tropical cyclones with county-level mortality in the US. JAMA 327 (10 ):946–955, PMID: , 10.1001/jama.2022.1682.35258534 13. Rappaport EN, Blanchard BW. 2016. Fatalities in the United States indirectly associated with atlantic tropical cyclones. Bull Amer Meteor Soc 97 (7 ):1139–1148, 10.1175/BAMS-D-15-00042.1. 14. Cruz-Cano R, Mead EL. 2019. Causes of excess deaths in Puerto Rico after Hurricane Maria: a time-series estimation. Am J Public Health 109 (7 ):1050–1052, PMID: , 10.2105/AJPH.2019.305015.30998411 15. Swerdel JN, Janevic TM, Cosgrove NM, Kostis JB, Myocardial Infarction Data Acquisition System (MIDAS 24) Study Group. 2014. The effect of Hurricane Sandy on cardiovascular events in New Jersey. J Am Heart Assoc 3 (6 ):e001354, PMID: , 10.1161/JAHA.114.001354.25488295 16. IPCC, 2022: Climate Change 2022: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Pörtner HO, Roberts DC, Tignor M, Poloczanska ES, Mintenbeck K, Alegría A, et al. , (eds.). Cambridge University Press. Cambridge University Press, Cambridge, UK and New York, NY, USA, 3056 pp., 10.1017/9781009325844. 17. Harvey C. Climate Studies Have Focused on Rich Countries. Scientific American , E&E News Climate Change section. 13 October 2021. https://www.scientificamerican.com/article/climate-studies-have-focused-on-rich-countries/ [accessed 4 October 2022]. 18. Guinto RR, Alejandre JCP, Bongcac MK, Guilaran J, Salcedo SS, Sunglao JA. 2021. An agenda for climate change and mental health in the Philippines. Lancet Planet Health 5 (11 ):e755–e757, PMID: , 10.1016/S2542-5196(21)00284-9.34774114 19. Kishore N, Marqués D, Mahmud A, Kiang MV, Rodriguez I, Fuller A, et al. 2018. Mortality in Puerto Rico after Hurricane Maria. N Engl J Med 379 (2 ):162–170, PMID: , 10.1056/NEJMsa1803972.29809109 20. Santos-Burgoa C, Sandberg J, Suárez E, Goldman-Hawes A, Zeger S, Garcia-Meza A, et al. 2018. Differential and persistent risk of excess mortality from Hurricane Maria in Puerto Rico: a time-series analysis. Lancet Planet Health 2 (11 ):e478–e488, PMID: , 10.1016/S2542-5196(18)30209-2.30318387 21. Keim ME. 2008. Building human resilience: the role of public health preparedness and response as an adaptation to climate change. Am J Prev Med 35 (5 ):508–516, PMID: , 10.1016/j.amepre.2008.08.022.18929977 22. Houston D. 2013. Crisis is where we live: environmental justice for the anthropocene. Globalizations 10 (3 ):439–450, 10.1080/14747731.2013.787771. 23. Mahoney A. Black Louisianans Still Haven’t Recovered From 2020’s Storms. Capital B. https://capitalbnews.org/louisiana-fema-private-insurance-housing-crisis-hurricane-season/ [accessed 4 October 2022]. 24. Paul BK, Rashid H, Islam MS, Hunt LM. 2007. Cyclone evacuation in Bangladesh: tropical cyclones Gorky (1991) vs. Sidr (2007). Environ Hazards 9 (1 ):89–101, 10.3763/ehaz.2010.SI04. 25. Rosenkoetter MM, Covan EK, Cobb BK, Bunting S, Weinrich M. 2007. Perceptions of older adults regarding evacuation in the event of a natural disaster. Public Health Nurs 24 (2 ):160–168, PMID: , 10.1111/j.1525-1446.2007.00620.x.17319888	36448793	PMC9710374	NO-CC CODE	2022-12-01 23:23:14	no	Environ Health Perspect. 2022 Nov 30; 130(11):111306	utf-8	Environ Health Perspect	2,022	10.1289/EHP12241	oa_other
==== Front Environ Health Perspect Environ Health Perspect EHP Environmental Health Perspectives 0091-6765 1552-9924 Environmental Health Perspectives EHP11252 10.1289/EHP11252 Review Impact of Hurricanes and Associated Extreme Weather Events on Cardiovascular Health: A Scoping Review https://orcid.org/0000-0002-5887-3301 Ghosh Arnab K. 1 Demetres Michelle R. 2 Geisler Benjamin P. 3 4 Ssebyala Shakirah N. 1 Yang Tianyi 5 Shapiro Martin F. 1 Setoguchi Soko 6 Abramson David 7 1 Department of Medicine, Weill Cornell Medical College, Cornell University, New York, New York, USA 2 Samuel J. Wood Library and C.V. Starr Biomedical Information Center, Weill Cornell Medicine, New York, New York, USA 3 Department of Medicine, Massachusetts General Hospital/Harvard Medical School, Boston, Massachusetts, USA 4 Institute for Medical Information Processing, Biometry, and Epidemiology, Ludwig Maximilian University of Munich, Munich, Germany 5 Rollins School of Public Health, Emory University, Atlanta, Georgia, USA 6 Rutgers Robert Wood Johnson Medical School, Rutgers University, New Brunswick, New Jersey, USA 7 Center of Public Health Disaster Science, School of Global Public Health, New York University, New York, New York, USA Address correspondence to Arnab K. Ghosh. Telephone: (212) 746-4071. Email: [email protected] 30 11 2022 11 2022 130 11 11600314 3 2022 21 10 2022 24 10 2022 https://ehp.niehs.nih.gov/about-ehp/license EHP is an open-access journal published with support from the National Institute of Environmental Health Sciences, National Institutes of Health. All content is public domain unless otherwise noted. Background: The frequency and destructiveness of hurricanes and related extreme weather events (e.g., cyclones, severe storms) have been increasing due to climate change. A growing body of evidence suggests that victims of hurricanes have increased incidence of cardiovascular disease (CVD), likely due to increased stressors around time of the hurricane and in their aftermath. Objectives: The objective was to systematically examine the evidence of the association between hurricanes (and related extreme weather events) and adverse CVD outcomes with the goal of understanding the gaps in the literature. Methods: A comprehensive literature search of population-level and cohort studies focused on CVD outcomes (i.e., myocardial infarction, stroke, and heart failure) related to hurricanes, cyclones, and severe storms was performed in the following databases from inception to December 2021: Ovid MEDLINE, Ovid EMBASE, Web of Science, and The Cochrane Library. Studies retrieved were then screened for eligibility against predefined inclusion/exclusion criteria. Studies were then qualitatively synthesized based on the time frame of the CVD outcomes studied and special populations that were studied. Gaps in the literature were identified based on this synthesis. Results: Of the 1,103 citations identified, 48 met our overall inclusion criteria. We identified articles describing the relationship between CVD and extreme weather, primarily hurricanes, based on data from the United States (42), Taiwan (3), Japan (2), and France (1). Outcomes included CVD and myocardial infarction–related hospitalizations (30 studies) and CVVD-related mortality (7 studies). Most studies used a retrospective study design, including one case–control study, 39 cohort studies, and 4 time-series studies. Discussion: Although we identified a number of papers that reported evaluations of extreme weather events and short-term adverse CVD outcomes, there were important gaps in the literature. These gaps included a) a lack of rigorous long-term evaluation of hurricane exposure, b) lack of investigation of hurricane exposure on vulnerable populations regarding issues related to environmental justice, c) absence of research on the exposure of multiple hurricanes on populations, and d) absence of an exploration of mechanisms leading to worsened CVD outcomes. Future research should attempt to fill these gaps, thus providing an important evidence base for future disaster-related policy. https://doi.org/10.1289/EHP11252 Supplemental Material is available online (https://doi.org/10.1289/EHP11252). The authors declare they have nothing to disclose. Note to readers with disabilities: EHP strives to ensure that all journal content is accessible to all readers. However, some figures and Supplemental Material published in EHP articles may not conform to 508 standards due to the complexity of the information being presented. If you need assistance accessing journal content, please contact [email protected]. Our staff will work with you to assess and meet your accessibility needs within 3 working days. ==== Body pmcIntroduction Hurricanes and related extreme weather events (e.g., cyclones and severe storms) are destructive events and are likely to increase in both number and destructiveness driven by climate change. In the United States alone, severe storms such as hurricanes have increased by 37% between 1990 and 2020,1 costing an estimated USD $1.8 trillion.1 Substantial populations across the globe, including the United States, reside in coastal regions, often in large, diverse, urban environments, making them vulnerable to the adverse effects of hurricanes,2 including loss of life and other morbidity, property damage, and homelessness. There is growing evidence that stress induced by hurricanes increases the risk of adverse cardiovascular disease (CVD) events, including increased risk of CVD-related mortality and hospitalization from coronary heart disease (CHD)3 and stroke.3 Additionally, studies have suggested an increased incidence and worsened control of chronic CVD risk factors such as diabetes4,5 and hypertension.6 These effects have been documented immediately after such events, but also in the months to years following, suggesting that the consequences of hurricanes extend beyond their immediate effects.5,7 Concerns exists that certain racial/ethnic groups,8 older individuals,9,10 and residents of socioeconomically disadvantaged neighborhoods11 are at increased risk after hurricanes of CVD-related morbidity12 and death3,13,14 in comparison with other segments of the population. Why this occurs is unclear but may be driven by predisaster factors such as insurance status15 and postdisaster factors such as access to health care and financial resources that would otherwise insulate vulnerable individuals from short- and long-term harm.15–17 Furthermore, evidence suggests these vulnerable subgroups may perceive threats from hurricanes differently18 and lack the means to leave,19–22 forcing them to endure the hurricanes’ effects while sheltering in place.23 Hurricanes and related extreme weather events are likely to worsen into the future.24 Currently no systematic reviews of the literature exist that focus on the relationship between hurricanes (and associated extreme weather events) and CVD outcomes. Therefore, to better understand the association between hurricanes and CVD outcomes, our objective was to systematically examine the peer-reviewed evidence that describes the association between hurricanes and adverse CVD outcomes. A specific goal of this study was to understand the gaps in the scientific literature to develop a research agenda that ultimately will provide the scientific community with the means to mitigate the adverse effects of hurricanes. Methods We performed a scoping review to qualitatively describe the evidence from studies that examined the exposure of hurricanes and related extreme weather events on CVD. This study was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews (PRISMA-ScR).25 In accordance with these standards, a protocol was submitted and preregistered by the International Prospective Register of Systematic Reviews (PROSPERO; CRD42022299034). The PRISMA flow diagram is described in Figure S1. Search Strategy A medical librarian (MRD) performed comprehensive searches to identify studies that addressed the effects of hurricanes or extreme weather events on CVD outcomes, including a range of search terms related to myocardial ischemia and/or myocardial infarction (MI), heart failure, and stroke. Full description of the search terms is described in the Supplemental Material, “Search Terms by Database.” Searches were run on 27 December 2021 in the following databases: Ovid MEDLINE (ALL - 1946 to present), Ovid EMBASE (1974 to present), The Cochrane Library (Wiley, 1996 to present), and Web of Science (Clarivate, 1900 to present). The search strategy included all appropriate controlled vocabulary and keywords for the concepts of extreme weather events (e.g., cyclone, hurricane, tropical storm, monsoon, tropical storm, tropical depression) and the cardiovascular outcomes of interest (e.g., MI, heart attack, heart failure, stroke, brain infarct). To limit publication bias, there were no language, publication date, or article type restrictions on the search strategy. However, due to the limitations of language competency of the study team, we subsequently excluded non-English publications. Study Selection Retrieved studies were screened (AKG, BPG, MRD, SNS, TY) for inclusion using Covidence systematic review software. Titles and abstracts were reviewed against predefined inclusion/exclusion criteria by two independent reviewers. Discrepancies were resolved by consensus. For final inclusion, full texts were retrieved, and then they were screened by two independent reviewers. Inclusion and Exclusion Criteria Our inclusion criteria were: a) Population: adult participants ≥18y old; b) Exposure: extreme weather events involving wet precipitation, including heavy rainfall and flooding (i.e., not snow), such as hurricanes, cyclones, tropical storms, typhoons, or storm surge; and c) Outcomes studied: reported CVD-related outcomes, such as myocardial ischemia and/or MI, heart failure, stroke, and death relate to CVD, collectively defined as major adverse cardiovascular or cerebrovascular events (MACE). Studies were excluded according to the following criteria: a) They were non-English; b) They were review articles, commentaries, editorials; c) Articles were case reports (i.e., n=1); and d) Articles did not provide enough data to determine whether CVD outcomes were related to MI or ischemia, heart failure, or stroke. The fourth exclusion criterion was included to ensure consistency in the CVD outcomes under study and because these CVD-related diagnoses are prevalent. Data Extraction Data extraction was performed independently in duplicate with predefined, standardized templates. Each extraction was reviewed independently by a secondary reviewer after extraction by the primary reviewer. Data points defined for extraction were study location, study setting, number of participants, exposure type, name and year, study objective, study design, data sources, CVD outcome, and key findings. Data Synthesis Following data extraction, CVD outcomes were synthesized qualitatively, assessed for the quality of evidence based on study design and number of participants and then categorized by the time frame of associated CVD outcomes, where short-term was defined as <1y and long-term as ≥1y. To reflect the importance of stress and broader social determinants of health as possible mechanisms moderating or mediating CVD outcomes with hurricane exposure, articles were grouped (AKG, SNS) by vulnerable populations, such as racial/ethnic groups; those with existing CVD-related comorbidities, including end-stage renal disease; diabetic patients; and patients with psychiatric comorbidities. These groups were not mutually exclusive and reflect our choices in highlighting important findings from our scoping review. Gaps in the literature were then identified based on this qualitative synthesis. No quantitative assessment of the literature (i.e., meta-analysis) was performed because of the heterogeneity in CVD outcomes. Results Summary of Articles Figure 1 summarizes the full PRISMA flow diagram outlining the study selection process. Of the 1,103 citations identified after deduplicating, 48 met our inclusion criteria (Table 1). Table 1 Table of studies. First author, year City, country Study setting Study design Study number Exposure type and name, year CVD outcome Key findings Becquart 201826 New Orleans, Jefferson, United States of American (USA) Countywide analysis Retrospective cohort study n=383,552 Hurricane Katrina, 2005 CVD hospitalization (including CHD, stroke) Of individuals with CHD, hospitalizations for CVD were two times higher for Black patients than for White patients (9.8% vs. 4.7%, respectively). Chang 202154 Taiwan Nationwide, longitudinal using National Registry Retrospective cohort study n=715,244 Typhoon Morakot, 2009 AMI and stroke-related incidence For patients on dialysis, higher incidence of AMI events and stroke events in areas severely affected by hurricane than in those moderately affected. Cifelli 201549 New York, USA Emergency department Retrospective cohort study n=23,776 (before hurricane); n=24,815 (after hurricane) Hurricane Sandy, 2012 AMI-related hospitalization There was a statistically significant increase of 39% in daily AMI visits after the storm, and non-statistically significant increase of 14% in daily tachyarrhythmia visits after the storm. Cruz-Cano 201965 Puerto Rico, USA Statewide analysis Retrospective time-series analysis n=3,493,593 (September); n=3,489,119 (October) Hurricane Maria, 2017 CVD-related mortality Estimated 253 excess deaths (1 in 5) from heart disease associated with Hurricane Maria. Deere 201827 New Orleans, USA Single center hospital Retrospective cohort study n=150 (before hurricane); n=2,724 (after hurricane) Hurricane Katrina, 2005 MI-related hospitalization 3.0% of patients presented with AMI 11 years after vs. 0.7% before the Hurricane, a 4-fold increase; prevalent psychosocial, behavioral, and traditional CHD risk factors were significantly higher among the post-Katrina group. Edmondson 201328 New Orleans, USA Community-based Prospective cohort study n=391 Hurricane Katrina, 2005 CVD-related hospitalization and CVD-related mortality Positive depression screening was significantly associated with increased risk of CVD-related hospitalization and mortality HR 1.33) but not PTSD Gautam 20097 New Orleans, USA Single center hospital Retrospective cohort study n=150 (before hurricane); n=246 (after hurricane) Hurricane Katrina, 2005 AMI-related hospitalization 2.18% of the 2-y post-Katrina cohort were admitted for AMI vs. 0.7% before the hurricane, an almost 3-fold increase. Gonzales 201531 New Orleans, USA Single center hospital Retrospective cohort study n=299 (before hurricane); n=1,479 (after hurricane) Hurricane Katrina, 2005 AMI-related hospitalization Compared to pre-Katrina data, 8 y post-Katrina AMI incidence decreased on Mondays, weekdays, and mornings. Post-Katrina AMI incidence increased on nights and weekends. Gonzales 201629 New Orleans, USA Single center hospital Retrospective cohort study n=150 (before hurricane); n=1,982 (after hurricane) Hurricane Katrina, 2005 AMI-related hospitalization 2.5% of the combined 9-year post-Katrina cohort were admitted for AMI vs. 0.7% before the hurricane - a 4-fold increase. Gonzales 201630 New Orleans, USA Single center hospital Retrospective cohort study n=299 (before hurricane); n=1,606 (after hurricane) Hurricane Katrina, 2005 AMI-related hospitalization Compared to pre-Katrina data, 9 y post-Katrina incidence of AMI decreased on Mondays, weekdays, and mornings. Further, the post-Katrina cohort had increased AMI incidence at night and on weekends. Gonzalez 202066 Puerto Rico, USA Single center hospital Retrospective cohort study n=235 admitted 2017; n=373 admitted 2016 Hurricane Maria, 2017 CHD-related hospitalization (all ACSs) AMI incidence of 9.79% (23 of 235) post hurricane, in comparison prehurricane MI incidence of 2.95% (11 of 373) [p=0.004]; Posthurricane CHD-related admissions was 27.17% (2017) vs. 33.04% prehurricane. Hameed 201232 New Orleans, USA Single center hospital Retrospective cohort study n=21,079 (before hurricane); n=28,597 (after hurricane) Hurricane Katrina, 2005 AMI-related hospitalization 2.2% of patients presented with AMI after vs. 0.7% before the Hurricane - a 3-fold increase; post-Katrina group had a higher prevalence of unemployment, lack of medical insurance, smokers, medical noncompliance, substance abuse, psychiatric comorbidities, history of coronary artery disease, and percutaneous coronary interventions. Harrison 202133 New Orleans, USA Single center hospital Retrospective cohort study n=3,619 Hurricane Katrina, 2005 AMI-related hospitalization 3.0% of patients presented with AMI after vs. 0.7% before the Hurricane - 3-fold increase; the posthurricane group had higher prescription rates but also larger nonadherence and was more likely to be unemployed and unmarried. Hendrickson 199757 Hawaii, USA Outpatient and emergency department chart review Retrospective cohort study n=1,584 Hurricane Iniki, 1992 CHD, hypertension, and stroke presentation Physician visits for CVD complaints were significantly increased in the post-Iniki period (RR 2.73, 95% CI: 1.51, 4.94); individuals aged 65 years and over experienced significant increases in CVD complaints (RR 2.67). Howe 200834 New Orleans, USA Outpatient clinic chart review Retrospective cohort study n=465 Hurricane Katrina, 2005 Presentation with hypertension, hypertensive urgency/emergency, MI, HF, arrhythmia, stable/unstable angina Among 410 illness complaints, 11% were CVD-related diagnoses. Hua 202158 Florida, USA Emergency department Retrospective time-series analysis n=30,358 Hurricane Irma, 2017 CHD, stroke, or HF emergency presentation No statistically significant differences pre- vs. posthurricane in CVD-related emergency department outcomes: HF (25.5.% vs. 25.4%); CHD (47.5% vs. 47.0%); stroke (7.0% vs. 7.1%). Huang 201752 Taiwan Major medical centers and regional hospitals in Taiwan Retrospective cohort study Not reported Typhoon Morakot, 2009 CHD-related incidence During the week of Typhoon Morakot, there was a significantly higher incidence (9.14 cases per day) of ACS than the study average (6.48 cases per day). Jiao 201235 New Orleans, USA Single center hospital Retrospective cohort study n=21,093 (before hurricane); n=150 (after hurricane) Hurricane Katrina, 2005 AMI-related hospitalization 2.0% of the combined 3-year post-Katrina cohort were admitted for AMI vs. 0.7% before the hurricane - an almost 3-fold increase. Kanaoka 202064 Japan Nationwide analysis using registry Retrospective cohort study n=10,782 Six typhoon landfalls between April 2013–March 2015 HF diagnosis, specifically TTS Admissions for TTS were significantly higher on the day and immediately after the typhoon landfalls with IRR 2.84 (95% CI: 1.53, 5.30) on the day of the landfall. Kim 201760 New Jersey, USA Statewide analysis Retrospective time-series analysis Not reported Hurricane Sandy, 2012 CVD-related mortality CVD mortality during the Sandy quarter showed significant 6% increased risk in mortality compared to same time period in previous 3 years, but during month of Sandy, CVD-related mortality for elderly significantly higher by 10% and persisted over the quarter. Kim 202014 Florida, USA Hospital Retrospective time-series analysis n=3,372,993 Tropical storms before 2007 HF-related hospitalization No significant correlation was found between frequency of storms in each quarter with HF discharges. Kostis 201450 New Jersey, USA County-wide analysis Retrospective cohort study Not reported Hurricane Sandy, 2012 AMI-related hospitalization and CVD-related death 23% increase in the number of AMI hospitalizations; the total number of AMI-related deaths (in-hospital plus out-of-hospital) increased by 28% all compared to same period in 5 prior years). Lawrence 20195 New York, USA Outpatient, hospitalization, and emergency department visits Retrospective cohort study n=217,873 Hurricane Sandy, 2012 CVD-related hospitalizations In the 4-month period following the storm, CVD risk (RR: 2.13) was significantly higher than corresponding period in other years [2 control groups of 5 y prior (2007–2011) and 1 y after (2013–2014)]. Comparing affected areas versus nonaffected areas, CVD had the greatest risk immediately (RR: 2.65, 4 months (RR: 2.62), and 12 months (RR: 2.64) after the hurricane period (all significantly different). Lee 201651 New York, NY, USA Emergency Department Retrospective cohort study Not reported Hurricane Sandy, 2012 AMI-related ED presentation Statistically significant increase MI presentations as a primary diagnosis in the week after Sandy’s landfall compared to the weeks preceding landfall. Lenane 201936 New Orleans, USA Community-based Prospective cohort study n=2,194 Hurricane Katrina, 2005 CVD-related hospitalization and CVD-related mortality (i.e., stroke, AMI, HF) The proportion of the sample experiencing a CVD event during follow-up was 18.0% and 11.0% among those with and without PTSD symptoms, respectively. Matthew 201337 New Orleans, USA Hospital Retrospective cohort study n=707 Hurricane Katrina, 2005 MI-related hospitalization Post-Katrina AMI presentations increased on nights and weekends and decreased on Mondays, weekdays, and mornings in comparison with pre-Katrina AMI presentations. McKinney 201167 Florida, USA County-wide analysis Retrospective cohort study n=624 Hurricanes Charley, Frances, Jeanne, 2004 All-cause mortality and also CVD mortality as subgroup CVD-related deaths significantly higher up to 1 month after Hurricane Ivan, 2 months after Hurricane Frances and Charley’s landfall in comparison with average mortality in the 3 y preceding. Moscona 201239 New Orleans, LA, USA Single center hospital Retrospective cohort study n=999 Hurricane Katrina, 2005 AMI-related hospitalization 2.2% of the 5-y post-Katrina cohort were admitted for AMI vs. 0.7% before the hurricane, an almost 3-fold increase. Moscona 201338 New Orleans, USA Single center hospital Retrospective cohort study n=1,177 Hurricane Katrina, 2005 AMI-related hospitalization 2.8% of patients presented with AMI 6 y after vs. 0.7% before the Hurricane, a 4-fold increase. Moscona 201940 New Orleans, USA Single center hospital Retrospective cohort study n=2,491 Hurricane Katrina, 2005 AMI-related hospitalization 2.8% of the combined 3-y post-Katrina cohort were admitted for MI vs. 0.7% before the hurricane, a 4-fold increase. Nakhle 202041 New Orleans, USA Single center hospital Retrospective cohort study n=3,067 Hurricane Katrina, 2005 AMI-related hospitalization 3.4% of patients presented with AMI after vs. 0.7% before the Hurricane - an almost 5-fold increase; Patients suffered higher rates of cardiovascular risk factors and psychiatric risk factors of CHD. Nakhle 202042 New Orleans, USA Single center hospital Retrospective cohort study n=3,278 (before hurricane); n=150 (after hurricane) Hurricane Katrina, 2005 AMI-related hospitalization 3.3% of the 13-y post-Katrina cohort were admitted for AMI vs. 0.7% before the hurricane, an almost 5-fold increase. Parks 202163 USA Nationwide analysis, multiple hospitals Retrospective cohort quasi-experimental study n=69,682,674 Multiple tropical cyclones from 1999 to 2014 Hospitalization (general); all CVD-related hospitalizations CVD was leading cause of hospitalizations (30%); CVD-related hospitalizations (from MI, pulmonary heart disease, general vascular disease in particularly) decreased on day of tropical cyclone, increased then peaked from day 1 to 3, and then dropped to baseline rate of cause-specific hospitalization (defined as average hospitalization rate between 1999–2014) by day 7. Peters 201443 New Orleans, USA Single center hospital Retrospective cohort study n=1,528 Hurricane Katrina, 2005 AMI-related hospitalization 2.24% of patients presented with AMI after vs. 0.7% before the Hurricane, a 3-fold increase. Peters 201444 New Orleans, USA Single center hospital Retrospective cohort study n=1,595 (before hurricane); n=1,296 (after hurricane) Hurricane Katrina, 2005 AMI-related hospitalization Post-Katrina AMI presentations increased on nights and weekends, and decreased on Mondays, weekdays, and mornings compared to pre-Katrina AMI presentations. Peters 201446 New Orleans, USA Single center hospital Retrospective cohort study n=1,476 Hurricane Katrina, 2005 AMI-related hospitalization 2.4% of the 6-y post-Katrina cohort were admitted for AMI vs. 0.7% before the hurricane, a 3-fold increase. Peters, 201345 New Orleans, USA Single center hospital Retrospective cohort study n=707 Hurricane Katrina, 2005 AMI-related hospitalization Post-Katrina AMI presentations increased on nights and weekends, and decreased on Mondays, weekdays, and mornings compared to pre-Katrina AMI presentations. Quast 201962 4 states (Louisiana, Mississippi, Texas, or Alabama) County-wide analysis Retrospective cohort study with propensity score matching n=170,328 Hurricane Katrina and Hurricane Rita, 2005 All-cause and CVD-related mortality Affected cohort had 34% higher risk of CVD-related mortality at 1 month, 17.4% at 6 months, and 14.6% at 1 year in comparison with unaffected matched controls. Sen 201855 Houston, USA Single center hospital Retrospective cohort study n=88 Hurricane Harvey, 2017 CVD-related ICU admission There were 40 flood-related admissions, of which 6 were CHD- or HF-related (15%). Sharma 200847 New Orleans, USA Hospital and community-based Retrospective cohort study Not reported Hurricane Katrina, 2005 Presenting with cardiovascular disease (CVD) CVD diagnoses presented the most common condition (32.8%); the proportion of men presenting with CVD diagnoses increased with age from 14.0% among those ages 0–19 y to 54.9% among those ages 80 y or older. Among all women, the most common diagnosis was CVD (29.2%). Shih 202053 Taiwan Nationwide analysis using registry Case–control study with propensity score matching n=715,244 Typhoon Morakot, 2009 CVD-related (i.e., stroke and CHD) hospitalizations Compared to matched individuals in moderately affected areas, individuals in severely affected area had 9.6% higher odds of MI, 12.9% higher risk of stroke, and 15.2% higher risk of HF. Singh 201548 New Orleans, USA Single center hospital Retrospective cohort study n=1,851 Hurricane Katrina, 2005 AMI-related hospitalization 2.57% of the 8-y post-Katrina cohort were admitted for AMI vs. 0.7% before the hurricane, an almost 4-fold increase. Sunohara 202169 Nagano, Japan Multiple hospitals Retrospective cohort study n=2,426 Reiwa First Year East Japan Typhoon, 2019 CVD-related hospitalization (including ACS, stroke) CVD hospitalizations increased 2 weeks after flooding, compared to same periods 2 y prior; unstable angina cases increased up to 2 months after flooding, and stroke increased 2 wk after flooding; also reported an increased incidence of missed medications and dyslipidemia. Swerdel 20143 New Jersey, USA Statewide analysis Retrospective cohort study Not reported Hurricane Sandy, 2012 AMI and stroke-related hospitalization, and CVD-related mortality In high-impact areas, AMI incidence increased by 22% in comparison with previous years, and 30-d AMI-related mortality increased by 31%. Stroke incidence increased by 7% and no significant change was observed in 30-d stroke-related mortality. Tarnoki 201761 Florida, USA Multiple hospitals Retrospective cohort study n=1,410 Hurricanes Karl (2010), Irene (2011), Cina (2011), Isaac (2012), Sandy (2012), Barbara (2013) and Andrea (2013) Ischemic stroke- and subarachnoid hemorrhages-related hospitalizations Non-significant increased incidence rate of ischemic stroke was consistent with the daily lowest and highest air pressure, highest air temperature compared to periods of normal air pressure and temperature, and also not significantly associated with presence of hurricanes or storms in comparison with time periods where there were no hurricanes or storms. Trebouet 201259 Vendée district, France Single center hospital Retrospective cohort study n=6 Cyclone Xynthia, 2010 TTS In 50% of patients, TTS was diagnosed. Yan 202168 USA Nationwide analysis, multiple hospitals Retrospective cohort study Not reported Multiple tropical cyclones between 1990–2010 CVD hospitalizations, defined as HF, stroke, arrythmia, CHD There was a significant 3% higher risk of CVD-related hospitalization during the storm period, associated with heart failure (8% increased risk) and AMI (5% increased risk) and CHD (3% increased risk) in comparison with counties not exposed to storms. Zane 201156 Texas, USA Statewide analysis Retrospective cohort study n=74 Hurricane Ike, 2008 MACE defined as cardiovascular failure MACE (n=12, 16%) was the leading cause of illness-related deaths. All of these were classified as indirectly (n=8) or possibly (n=4) caused by Hurricane Ike. Note: ACS, acute coronary syndrome; AMI, acute myocardial infarction; CHD, coronary heart disease; CI, confidence interval; CVD, cardiovascular disease; HF, heart failure; HR, hazard ratio; IRR, incidence rate ratio; MACE, major adverse cardiovascular event; MI, myocardial infarction; PTSD, posttraumatic stress disorder; RR, risk ratio; TTS, Takotsubo syndrome. Only 6 out of 48 (12.5%) of the articles described the relationship between CVD and hurricanes outside the United States and were based on data from Taiwan, Japan, and France. Of those articles from the United States, the majority (25/42; 54.3%) described exposures in New Orleans after 2005 Hurricane Katrina.7,26–48 Of particular note, 19 of these 25 articles were authored by the same research group that replicated the same pre- and posthurricane analysis to compare a) MI-related admission rates and b) MI admission chronobiology from the period before Hurricane Katrina to the period after Hurricane Katrina over ever-increasing periods of time (i.e., 2, 3 y after Hurricane Katrina, up to 14 y after Hurricane Katrina). Other hurricanes and related weather events that were described were 2012 Hurricane Sandy (six studies),3,14,5,49–51 2009 Typhoon Morakot that struck Taiwan (three studies),52–54 2017 Hurricane Harvey (one study),55 2008 Hurricane Ike (one study),56 1992 Hurricane Iniki (one study),57 2017 Hurricane Irma (one study),58 and 2010 Cyclone Xynthia (one study).59 Six of the 48 studies described population-level effects of multiple hurricanes/cyclones/typhoons over an extended period of time. Thirty-two of the 48 articles (66.7%) focused on the individual patient-level CVD outcomes in the hospitalized setting by analyzing hospital data from single centers. CVD- and MI-related hospitalizations were the most common CVD outcome studied, representing 30 of the 48 studies (62.5%). The second most common outcome (7 studies; 14.5%) studied was CVD-related mortality. Three studies reported no association between hurricanes and CVD outcomes60; the first focused on heart failure exacerbations in Florida after tropical storms, the second focused on strokes in southern Florida,61 and the third focused on emergency room presentations for residents in assisted living facilities.58 Sources and Directions of Bias The majority of the studies (46/48; 95%) employed a retrospective study design using secondary data, reflecting the observational nature of hurricane-related studies in general. Of these, one used a case–control design, 39 employed a retrospective cohort study design, and 4 employed a retrospective time-series analysis approach. As a result of the retrospective nature of these studies, unobserved confounders including posthurricane migration and displacement between hurricane exposure and CVD outcomes are likely a source of bias, particularly in the Hurricane Katrina studies that examined this relationship at a single institution authored by the same research group. Several studies may have limited these sources of bias, owing to their large study population sizes over broad geographies (five used nationwide data, four used statewide data, four used countywide data) and through the use of causal methods, including propensity score matching53,62 and quasi-experimental design63 (Table 1). The largest of these studies63 used Medicare fee-for-service beneficiary-level data between 1999 and 2014, yielding >69 million individual observations in a quasi-experimental design (pre- and posthurricane observation over a short time frame), thus improving the quality of the study findings through large numbers and limiting unobserved exogenous confounding as a natural experiment. Hereafter, we provide a synthesis of findings from the scoping review that describe the CVD outcomes by short- and long-term outcomes, followed by studies that focused on special populations. A summary of final list of studies is described in the Table 1. Short-term (<1y) CVD-related outcomes related to hurricanes. Three studies26,63,64 examined CVD-related outcomes exclusively in the immediate aftermath (within 1 wk) of hurricane/cyclone exposure. Based on findings from single institution study after Hurricane Katrina26 and a larger study63 that examined acute posthurricane hospitalizations from multiple hurricane exposures during the period 1999–2014, CVD hospitalizations peaked between days 2 and 6, peaking at an approximate 3%–4% increase in comparison with prehurricane levels. Six studies14,5,49–51,3 met criteria for inclusion that focused on the short-term CVD outcomes related to Hurricane Sandy. Three of these studies used population-level measures to examine the effect of the hurricane on CVD events and CVD mortality in the weeks to months after landfall. Using New Jersey vital statistics, the first study found a 6% increase [95% confidence interval (CI): 2%, 11%) in CVD-related mortality overall the month after landfall compared to before,14 whereas two others, demonstrated a 22%–23% increased risk (95% CI: 1.16, 1.28) of MI-related hospitalization50,3 [including a 31% increased risk (95% CI: 1.22, 1.41) of 30-d mortality], and a 7% increased risk (95% CI: 1.03, 1.11) of stroke-related hospitalization compared to the same periods in the years preceding.3 Two studies65,66 examined short-term CVD outcomes in aftermath of Hurricane Maria in 2017 in Puerto Rico. The first, which examined CVD-related mortality in the month following landfall,65 calculated 253 excess deaths from CVD using a time-series analytical approach, which represented overall 1 in 5 excess deaths. Separately, results from a single site observation study66 suggested a significant increase in MI hospitalizations after landfall in comparison with MI hospitalizations before landfall [post: 23/235 (9.75%) vs. pre: 11/373 (2.95%)]. Four studies focused on the exposures of multiple hurricanes either across a hurricane season,67 or multiple hurricanes over many years.61,63,68 Similar to previous studies examining CVD-related mortality, authors67 performed a countywide analysis of the 2004 hurricane season in Florida and showed increased mortality 1–2 months after landfall in comparison with weeks before landfall. In another,63 CVD was the leading cause of posthurricane hospitalization (30% of total hospitalizations from the period were CVD-related) across multiple hurricane seasons. Two studies described higher rates of short-term (>1 month) CVD outcomes after hurricanes in comparison with outcomes before hurricanes. CVD reports increased in primary care and emergency departments at various locations in Hawaii after Hurricane Iniki in 1992 in comparison with prehurricane primary care visits [relative risk (RR) 2.73; 95% CI: 1.51, 4.94] prior to landfall.57 After the Reiwa East Japan Typhoon,69 MI-related hospitalizations significantly increased up to 2 months after landfall, and stroke hospitalizations increased 2 wk after landfall in comparison with stroke hospitalizations in prior years. After Hurricane Harvey, intensive care admissions related to worsened heart failure and CHD represented 15% of intensive care unit (ICU) cases related to hurricane flooding, although it was unclear whether this represented a deviation from prehurricane norms.55 Two studies59,64 reported the increased incidence of Takotsubo’s cardiomyopathy (a rare form of heart failure often related to severe acute stresses) in the aftermath of hurricanes in comparison with incidences in prehurricane timeframes. The larger of these two studies64 showed, using a Japanese CVD registry, that Takotsubo’s cardiomyopathy admissions were significantly higher on the day of and immediately after typhoon landfalls [incidence rate ratio (IRR) 2.84; 95% CI: 1.53, 5.30] but not subsequent days. Long-term (≥1y) CVD-related outcomes related to hurricanes. Of the 25 Hurricane Katrina studies, 1927,7,29–33,37–46,48,66,36 describe statistically significant 3- to 5-fold increases in MI-related hospitalization rates at a single institutional site from 2 y before Hurricane Katrina’s landfall to up to 14 y after landfall. In patient populations under investigation, these same 19 studies also reported significant differences in rates of baseline CHD and other CVD risk factors, such as smoking, hypertension, diabetes, substance abuse, and psychiatric disease, and lower socioeconomic factors (including higher unemployment rates), with higher rates found in the post-Katrina patient population. Furthermore, a subset of these studies30,31,37,40,44,45 further report a consistent increase in MI-related hospital admissions over nights and weekends but decreased MI-related hospitalizations on weekdays and mornings both relative to prehurricane times. Vulnerable populations. Racial disparities in CVD-related outcomes. In a post-Katrina study26 examining racial differences in CVD-hospitalization based on Medicare claims, the authors reported that within the first month, Black patients were almost twice as likely White patients to be admitted within the first month after landfall (Black patients: 26.3±23.7 cases per day vs. White patients: 16.6±11.7 cases per day). Patients with end-stage renal disease. In a study64 that used Taiwan’s National Health Insurance Registry to understand the relationship between 2009 Typhoon Morakot and associated CVD outcomes for end-stage renal disease (ESRD) patients, authors found that ESRD patients had lower rates of MI- and stroke-related hospitalizations after landfall compared to such hospitalizations before landfall (111 vs. 127 cases, respectively, per 1,000 population). Patients with psychiatric comorbidities. Two studies, both of prospective cohort design created for other purposes, have highlighted the link between depression and posttraumatic stress disorder (PTSD) and their association with CVD outcomes in hurricane-affected communities. In the first study,28 which focused on patients with ESRD, positive depression screening (but not positive PTSD screening) was significantly associated with an increased risk of CVD-related hospitalization and mortality [hazard ratio (HR) 1.33; 95% CI: 1.06, 1.76]. In the second study36 that evaluated CVD event incidence (for stroke, MI, and heart failure) in community-based older patients post-Katrina with hypertension, CVD event incidence was overall high (11.6%) and higher still among those patients with evidence of PTSD (18%). Diabetic patients. Two studies specifically focused on diabetic patients. The first70 used Medicare claims to examine the association between CVD-related mortality in diabetic patients and the exposure of Hurricanes Katrina and Maria across multiple affected counties, finding that those affected had a 35% increased odds [odds ratio (OR) 1.35; 95% CI: 1.21, 1.51] of CVD-related mortality at 1 month, 17.4% (OR 1.17; 95% CI: 1.12, 1.23) at 6 months, and 14.6% (OR 1.15; 95% CI: 1.11, 1.19) at 1 y in comparison with unaffected matched beneficiaries. Examining the effects of Hurricane Sandy in New York City, one study51 focused on diabetic patients reported a significant increase in MI-related hospitalizations in the week after landfall, whereas the other examined older patients, using Medicare and Medicaid claims data to highlight that CVD-related hospitalizations continued to be higher than prehurricane time periods even 1 y after landfall (RR=2.64; 95% CI: 2.64, 2.65). Discussion In this scoping review, we identified 48 English-language, peer-reviewed, original research articles that described the association between CVD and extreme weather events, including hurricanes. Of these, the majority focused on events in the United States and in particular the effects of Hurricane Katrina in 2005. This review provides an overview of publications that focus on a range of CVD outcomes and their association with hurricanes and other extreme weather events. Particularly, our scoping review summarizes the current literature on the short-term associations between hurricanes and CVD-related hospitalization and CVD-related mortality—evidence that has been described in the aftermath of Hurricanes Katrina,26–28,7,29–48,62 Sandy,14,3,5,49–51 Maria,65 Irma,58 Iniki,57 and Harvey55 and Typhoon Morakot,57,58 as well as broadly across multiple hurricanes.52,60,61,63,64,67,68 Furthermore, the scoping review sheds light on certain vulnerable populations at increased risk of adverse CVD outcomes. Several studies described the impact on older patients (including those in the Medicare58,63 and dually eligible Medicare–Medicaid population),5 patients with underlying CVD risk factors such as diabetes51,70 and ESRD,54 and people suffering from psychiatric conditions such as depression and PTSD.28,36 Most important, this scoping review also highlights several important gaps in knowledge. As hurricanes increase in frequency and destructiveness, developing a research agenda to fill these gaps will become increasingly important in the future to mitigate the adverse CVD-related effects of hurricanes. This significant need to increase knowledge is particularly the case as the importance of environmental justice becomes more prominent in the era of climate change,71 with the increasing acknowledgment of the role that structural factors play in worsening health outcomes for vulnerable populations affected by hurricanes. Rigorous Long-Term Evaluation of Hurricane Exposure on CVD Outcomes, Including outside the United States It remains unclear but important to understand whether the CVD-related associations of hurricanes persist over longer durations. In the United States72 and broadly,73 disaster-related frameworks focused on health outcomes often concentrate resources in the immediate postdisaster period (often the first few months). If evidence suggests that the effects of hurricane exposure linger and potentially accelerate CVD progression, then such frameworks should be reconceptualized over a longer time frame. Our scoping review did find a number of studies that examined longitudinal outcomes for hurricane victims. However, in the studies that did so, findings were limited by the potential confounding inherent in a single institution focused pre- and posthurricane design, the lack of adequate controls as a comparison group to define differences in outcomes. Furthermore, these findings lack generalizability outside the United States, given the high prevalence of studies performed in the United States. Therefore, to fill this important gap in the literature, more rigorous, well-defined cohorts to follow over time are required to quantify the long-term CVD risk resulting from hurricane exposures, particularly in non-U.S. settings. Vulnerable Populations with Focus on Environmental Justice Our scoping review reported one study that examined racial differences in MI-related hospitalization after Hurricane Katrina.26 Although the relationship between hurricanes such as Hurricane Katrina and vulnerable populations has been clearly described in the general literature,10,74 our scoping review highlighted the dearth of peer-reviewed literature focused on the differential role hurricanes play in affecting CVD outcomes by race/ethnicity, gender, socioeconomic status, older age, and children’s and adolescents’ health and development, as well as other important structural factors such as home ownership and social connectedness. This finding reflected, in our views, the unclear nature of how vulnerable populations were defined (other than Black vs. White individuals, as described in Becquart et al.).61 Given the prevalence of CVD and the clear socioeconomic gradient linking CVD outcomes to lower socioeconomic status,75 we believe this to be a surprising finding stemming from the scoping review. Further research is required to cast an important light on the health-related disparities created by climate change-driven weather phenomenon. Association between Multiple and Repeated Hurricanes Strikes and CVD Outcomes Hurricanes and related weather phenomena have historically occurred in specific geographies during certain times of the year. Therefore, at-risk populations will continue to be exposed to hurricanes. In our scoping review, the majority of studies were limited to the exposure to one hurricane, not repeated or cumulative exposures over time. Nonetheless, the cumulative impact of multiple hurricanes or other extreme weather events over time is likely to have deleterious effects on CVD outcomes for affected individuals. The need for peer-reviewed evidence to describe and quantify the repeated and cumulative impacts of hurricanes on these CVD outcomes will be even more important as hurricanes become more frequent and more destructive. Exploration of the Mechanisms for Increased CVD Outcomes Two studies found in the scoping review describe the role depression and PTSD play on CVD outcomes in relation to hurricanes.28,36 Although stress has been commonly attributed to the causal pathway from hurricane impacts to CVD outcomes, this has not been rigorously studied, based on our findings. One challenge in the exploration of mechanisms of CVD outcomes related to hurricanes is the lack of prospective epidemiological cohorts to follow over time. This challenge is likely the reason why most studies used a retrospective study design involving secondary data sources. Although this is a limitation in future work focused on understanding the mechanistic pathway linking CVD and hurricanes, novel, validated causal inference methods that allow identification are now available, making this work possible.76 Future research should seek to use these tools and large data sources to develop an understanding of potential mechanisms. A major limitation of this study is that it excluded non-English texts. Hurricanes and related events happen all across the world. It is highly likely that published non-English peer-reviewed literature exists that focuses on CVD outcomes and that we could not access this research because of the language barrier. In an attempt to offset this risk, our initial search did not exclude non-English language texts. However, subsequently these were excluded from further review because we lacked the resources to confidently review and appraise non-English texts. In conclusion, our scoping review demonstrated a clear link between hurricanes and related extreme weather events exposure and CVD outcomes in the prevailing peer-reviewed, published, original research studies in English. Most studies linked hurricanes to CVD outcomes in the short-term, and also highlighted special populations, including those with other medical comorbidities. An important result is that our review revealed a number of gaps in the knowledge, which includes a lack of rigorous long-term evaluation of hurricane impacts, their effects on vulnerable populations including structural drivers (e.g., access to care, systematic inequalities in health care delivery, neighborhood-level social determinants of health) that place such individuals at risk predisaster and exacerbate adverse health outcomes postdisaster, the role that multiple hurricanes play on population health, and exploration of the mechanisms that lead to worsened CVD outcomes. Future research is currently underway to generalize these findings across multiple hurricanes and extreme weather events, and such research should attempt to fill these gaps. This enhancement will provide an important evidence base for future disaster-related policy that seeks to address environmental injustices as well as the overall adverse effects of these extreme weather events. Supplementary Material Click here for additional data file. Click here for additional data file. ==== Refs References 1. National Centers for Environmental Information. 2020. Billion-Dollar Weather and Climate Disasters. https://www.ncdc.noaa.gov/billions/ [accessed 1 July 2020]. 2. Crossett KM, Culliton TJ, Wiley PC, Goodspeed TR. 2004. Population Trends Along the Coastal United States: 1980–2008. Washington, DC: NOAA National Ocean Service Special Projects. 3. Swerdel JN, Janevic TM, Cosgrove NM, Kostis JB. 2014. The effect of Hurricane Sandy on cardiovascular events in New Jersey. J Am Heart Assoc 3 (6 ):e001354, PMID: , 10.1161/JAHA.114.001354.25488295 4. Fonseca VA, Smith H, Kuhadiya N, Leger SM, Yau CL, Reynolds K, et al. 2009. Impact of a natural disaster on diabetes: exacerbation of disparities and long-term consequences. Diabetes Care 32 (9 ):1632–1638, PMID: , 10.2337/dc09-0670.19542210 5. Lawrence WR, Lin Z, Lipton EA, Birkhead G, Primeau M, Dong G-H, et al. 2019. After the storm: short-term and long-term health effects following Superstorm Sandy among the elderly. Disaster Med Public Health Prep 13 (1 ):28–32, PMID: , 10.1017/dmp.2018.152.30841951 6. Baum A, Barnett ML, Wisnivesky J, Schwartz MD. 2019. Association between a temporary reduction in access to health care and long-term changes in hypertension control among veterans after a natural disaster. JAMA Netw Open 2 (11 ):e1915111, PMID: , 10.1001/jamanetworkopen.2019.15111.31722027 7. Gautam S, Menachem J, Srivastav SK, Delafontaine P, Irimpen A. 2009. Effect of Hurricane Katrina on the incidence of acute coronary syndrome at a primary angioplasty center in New Orleans. Disaster Med Public Health Prep 3 (3 ):144–150, PMID: , 10.1097/DMP.0b013e3181b9db91.19713855 8. Burger J, Gochfeld M, Lacy C. 2019. Ethnic differences in risk: experiences, medical needs, and access to care after Hurricane Sandy in New Jersey. J Toxicol Environ Health A 82 (2 ):128–141, PMID: , 10.1080/15287394.2019.1568329.30722754 9. Malik S, Lee DC, Doran KM, Grudzen CR, Worthing J, Portelli I, et al. 2018. Vulnerability of older adults in disasters: emergency department utilization by geriatric patients after Hurricane Sandy. Disaster Med Public Health Prep 12 (2 ):184–193, PMID: , 10.1017/dmp.2017.44.28766475 10. Adams V, Kaufman SR, van Hattum T, Moody S. 2011. Aging disaster: mortality, vulnerability, and long-term recovery among Katrina survivors. Med Anthropol 30 (3 ):247–270, PMID: , 10.1080/01459740.2011.560777.21590581 11. Burger J, Gochfeld M, Lacy C. 2019. Concerns and future preparedness plans of a vulnerable population in New Jersey following Hurricane Sandy. Disasters 43 (3 ):658–685, PMID: , 10.1111/disa.12350.30990925 12. Flores AB, Collins TW, Grineski SE, Chakraborty J. 2020. Disparities in health effects and access to health care among Houston area residents after Hurricane Harvey. Public Health Rep 135 (4 ):511–523, PMID: , 10.1177/0033354920930133.32539542 13. Stephens KU, Grew D, Chin K, Kadetz P, Greenough PG, Burkle FM, et al. 2007. Excess mortality in the aftermath of Hurricane Katrina: a preliminary report. Disaster Med Public Health Prep 1 (1 ):15–20, PMID: , 10.1097/DMP.0b013e3180691856.18388597 14. Kim S, Kulkarni PA, Rajan M, Thomas P, Tsai S, Tan C, et al. 2017. Hurricane Sandy (New Jersey): mortality rates in the following month and quarter. Am J Public Health 107 (8 ):1304–1307, PMID: , 10.2105/AJPH.2017.303826.28640678 15. Sharp MJ, Sun M, Ledneva T, Lauper U, Pantea C, Lin S. 2016. Effect of Hurricane Sandy on health care services utilization under Medicaid. Disaster Med Public Health Prep 10 (3 ):472–484, PMID: , 10.1017/dmp.2016.75.27181259 16. Davidow AL, Thomas P, Kim S, Passannante M, Tsai S, Tan C. 2016. Access to care in the wake of Hurricane Sandy, New Jersey, 2012. Disaster Med Public Health Prep 10 (3 ):485–491, PMID: , 10.1017/dmp.2016.79.27292171 17. Burger J, Gochfeld M, Pittfield T, Jeitner C. 2017. Responses of a vulnerable Hispanic population in New Jersey to Hurricane Sandy: access to care, medical needs, concerns, and ecological ratings. J Toxicol Environ Health A 80 (6 ):315–325, PMID: , 10.1080/15287394.2017.1297275.28644717 18. Elder K, Xirasagar S, Miller N, Bowen SA, Glover S, Piper C. 2007. African Americans’ decisions not to evacuate New Orleans before Hurricane Katrina: a qualitative study. Am J Public Health 971 (suppl 1 ):S124–S129, PMID: , 10.2105/AJPH.2006.100867.17413086 19. Fussell E. 2015. The long-term recovery of New Orleans’ population after Hurricane Katrina. Am Behav Sci 59 (10 ):1231–1245, PMID: , 10.1177/0002764215591181.26880853 20. Bethel JW, Foreman AN, Burke SC. 2011. Disaster preparedness among medically vulnerable populations. Am J Prev Med 40 (2 ):139–143, PMID: , 10.1016/j.amepre.2010.10.020.21238861 21. Eisenman DP, Cordasco KM, Asch S, Golden JF, Glik D. 2007. Disaster planning and risk communication with vulnerable communities: lessons from Hurricane Katrina. Am J Public Health 97 (suppl 1 ):S109–S115, PMID: , 10.2105/AJPH.2005.084335.17413069 22. Meyer MA. 2017. Elderly perceptions of social capital and age-related disaster vulnerability. Disaster Med Public Health Prep 11 (1 ):48–55, PMID: , 10.1017/dmp.2016.139.27839520 23. Elliott JR, Pais J. 2006. Race, class, and Hurricane Katrina: social differences in human responses to disaster. Soc Sci Res 35 (2 ):295–321, 10.1016/j.ssresearch.2006.02.003. 24. Kossin JP, Knapp KR, Olander TL, Velden CS. 2020. Global increase in major tropical cyclone exceedance probability over the past four decades. Proc Natl Acad Sci U S A 117 (22 ):11975–11980, PMID: , 10.1073/pnas.1920849117.32424081 25. Tricco AC, Lillie E, Zarin W, O’Brien KK, Colquhoun H, Levac D, et al. 2018. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med 169 (7 ):467–473, PMID: , 10.7326/M18-0850.30178033 26. Becquart NA, Naumova EN, Singh G, Chui KKH. 2018. Cardiovascular disease hospitalizations in Louisiana parishes’ elderly before, during and after Hurricane Katrina. Int J Environ Res Public Health 16 (1 ):74, 10.3390/ijerph16010074.30597886 27. Deere B, Moscona J, Srivastav S, Zimmerman M, Mallya S, Razavi M, et al. 2018. Incidence of acute myocardial infarction and Hurricane Katrina: four-fold increase eleven years after the storm. J Am Coll Cardiol 71 (11 ):A1903, 10.1016/S0735-1097(18)32444-6. 28. Edmondson D, Gamboa C, Cohen A, Anderson AH, Kutner N, Kronish I, et al. 2013. Association of posttraumatic stress disorder and depression with all-cause and cardiovascular disease mortality and hospitalization among Hurricane Katrina survivors with end-stage renal disease. Am J Public Health 103 (4 ):e130–137–e137, PMID: , 10.2105/AJPH.2012.301146.23409901 29. Gonzales H, Albashaireh D, Raja M, et al. 2016. Continued impact of Hurricane Katrina on incidence of acute myocardial infarction: nine years after the storm. J Investig Med 64 (2 ):647–647. 30. Gonzales H, Lawrence J, Peters M, Yadav K, Albashaireh D, Baydoun H, et al. 2016. Persistent alterations in myocardial infarction circadian rhythm after Hurricane Katrina: nine years after the storm. J Am Coll Cardiol 67 (13 ):470, 10.1016/S0735-1097(16)30471-5. 31. Gonzales H, Singh K, Deandrade K, et al. 2015. Variation in the timing of myocardial infarctions after Hurricane Katrina: eight years after the storm. Circulation 132 (suppl 3) . 32. Hameed I, Moscona J, Kakoulides S, Srivastav S, Delafontaine P, Irimpen A. 2012. Acute myocardial infarction before and after the storm: Hurricane Katrina. J Investig Med 60 (1 ):401–402. 33. Harrison D, Rawal H, Quan M, et al. 2021. Incidence of acute myocardial infarction and Hurricane Katrina: fourteen years after the storm. J Investig Med 69 (2 ):494. 34. Howe E, Victor D, Price EG. 2008. Chief complaints, diagnoses, and medications prescribed seven weeks post-Katrina in New Orleans. Prehosp Disaster Med 23 (1 ):41–47, PMID: , 10.1017/s1049023x00005549.18491660 35. Jiao Z, Kakoulides SV, Moscona J, Whittier J, Srivastav S, Delafontaine P, et al. 2012. Effect of Hurricane Katrina on incidence of acute myocardial infarction in New Orleans three years after the storm. Am J Cardiol 109 (4 ):502–505, PMID: , 10.1016/j.amjcard.2011.09.045.22154089 36. Lenane Z, Peacock E, Joyce C, Frohlich ED, Re RN, Muntner P, et al. 2019. Association of post-traumatic stress disorder symptoms following Hurricane Katrina with incident cardiovascular disease events among older adults with hypertension. Am J Geriatr Psychiatry 27 (3 ):310–321, PMID: , 10.1016/j.jagp.2018.11.006.30581139 37. Matthew NP, Morgan JK, John CM, et al. 2013. Alteration in the chronobiology of onset of acute myocardial infarction in New Orleans residents following Hurricane Katrina. J Am Coll Cardiol 61 (10 ):E48, 10.1016/S0735-1097(13)60049-2. 38. Moscona J, Tiwari S, DeAndrade K, Quevedo H, Peters M, Munshi K, et al. 2013. Increased incidence of acute coronary syndrome following Hurricane Katrina in New Orleans: the impact continues. J Am Coll Cardiol 61 (10 ):E12, 10.1016/S0735-1097(13)60013-3. 39. Moscona J, Tiwari S, Munshi K, Srivastav S, Delafontaine P, Irimpen A. 2012. The effects of Hurricane Katrina on acute myocardial infarction five years after the storm. J Am Coll Cardiol 59 (13 ):E354, 10.1016/S0735-1097(12)60355-6. 40. Moscona JC, Peters MN, Maini R, Katigbak P, Deere B, Gonzales H, et al. 2019. The incidence, risk factors, and chronobiology of acute myocardial infarction ten years after Hurricane Katrina. Disaster Med Public Health Prep 13 (2 ):217–222, PMID: , 10.1017/dmp.2018.22.29644946 41. Nakhle A, Ayoub A, Subedi R, Panhwar M, Sangani D, Razavi M, et al. 2020. Incidence of acute myocardial infarction and Hurricane Katrina: thirteen years after the storm. J Am Coll Cardiol 75 (11 ):138, 10.1016/S0735-1097(20)30765-8. 42. Nakhle A, Deere BP, Razavi M, et al. 2020. Incidence of acute myocardial infarction and Hurricane Katrina: twelve years after the storm. J Investig Med 68 (2 ):475. 43. Peters MN, Deandrade K, Diaz HQ, Andrew BR, Tiwari S, Singh K, et al. 2014. Hurricane Katrina and myocardial infarction incidence: seven years after the storm. J Am Coll Cardiol 63 (12 ):A127, 10.1016/S0735-1097(14)60127-3. 44. Peters MN, Deandrade K, Diaz HQ, Burchett AR, Tiwari S, Singh K, et al. 2014. Hurricane Katrina and the timing of myocardial infarction: seven years after the storm. J Am Coll Cardiol 63 (12 ):A50, 10.1016/S0735-1097(14)60050-4. 45. Peters MN, Katz MJ, Moscona JC, Alkadri ME, Khazi Syed RH, Turnage TA, et al. 2013. Effect of Hurricane Katrina on chronobiology at onset of acute myocardial infarction during the subsequent three years. Am J Cardiol 111 (6 ):800–803, PMID: , 10.1016/j.amjcard.2012.10.050.23291089 46. Peters MN, Moscona JC, Katz MJ, Deandrade KB, Quevedo HC, Tiwari S, et al. 2014. Natural disasters and myocardial infarction: the six years after Hurricane Katrina. Mayo Clin Proc 89 (4 ):472–477, PMID: , 10.1016/j.mayocp.2013.12.013.24656058 47. Sharma AJ, Weiss EC, Young SL, Stephens K, Ratard R, Straif-Bourgeois S, et al. 2008. Chronic disease and related conditions at emergency treatment facilities in the New Orleans area after Hurricane Katrina. Disaster Med Public Health Prep 2 (1 ):27–32, PMID: , 10.1097/DMP.0b013e31816452f0.18388655 48. Singh KY, Gonzales H, Diaz HQ, Raja M, Deandrade K, Baydoun H, et al. 2015. Hurricane Katrina and myocardial infarction incidence: eight years after the storm. J Am Coll Cardiology 65 (10 ):A40, 10.1016/S0735-1097(15)60040-7. 49. Cifelli MA, Allegra JR, Allegra CM. 2015. Cardiac events in New York metropolitan emergency departments after Hurricane Sandy. Acad Emergency Med 22 (5 suppl 1 ):S209. 50. Kostis JB, Cosgrove NM, Swerdel JN, Janevic TM. 2014. The effect of Hurricane Sandy on cardiovascular events. European Heart Journal 35 (suppl 1 ):555. 51. Lee DC, Gupta VK, Carr BG, Malik S, Ferguson B, Wall SP, et al. 2016. Acute post-disaster medical needs of patients with diabetes: emergency department use in New York city by diabetic adults after Hurricane Sandy. BMJ Open Diabetes Res Care 4 (1 ):e000248, PMID: , 10.1136/bmjdrc-2016-000248.27547418 52. Huang C-H, Lin H-C, Tsai C-D, Huang H-K, Lian I-B, Chang C-C. 2017. The interaction effects of meteorological factors and air pollution on the development of acute coronary syndrome. Sci Rep 7 (101563288 ):44004, PMID: , 10.1038/srep44004.28276507 53. Shih H-I, Chao T-Y, Huang Y-T, Tu Y-F, Sung T-C, Wang J-D, et al. 2020. Increased medical visits and mortality among adults with cardiovascular diseases in severely affected areas after Typhoon Morakot. Int J Environ Res Public Health 17 (18 ):6531, PMID: , 10.3390/ijerph17186531.32911725 54. Chang C-M, Chao T-YS, Huang Y-T, Tu Y-F, Sung T-C, Wang J-D, et al. 2021. Maintaining quality of care among dialysis patients in affected areas after Typhoon Morakot. Int J Environ Res Public Health 18 (14 ):7400. PMID: 34299851, 10.3390/ijerph18147400.34299851 55. Sen A, Ayad M, Karanth S, Patil S, Luther K, Patel B. 2018. Hurricane Harvey: impact on ICU admission. Am J Respir Crit Care Med 197 :A6310. 10.1164/ajrccm-conference.2018.197.1_MeetingAbstracts.A6310. 56. Zane DF, Bayleyegn TM, Hellsten J, Beal R, Beasley C, Haywood T, et al. 2011. Tracking deaths related to Hurricane Ike, Texas, 2008. Disaster Med Public Health Prep 5 (1 ):23–28, PMID: , 10.1001/dmp.2011.8.21402823 57. Hendrickson LA, Vogt RL, Goebert D, Pon E. 1997. Morbidity on Kauai before and after Hurricane Iniki. Prev Med 26 (5 Pt 1 ):711–716, PMID: , 10.1006/pmed.1997.0196.9327481 58. Hua CL, Thomas KS, Peterson LJ, Hyer K, Dosa DM. 2021. Emergency department use among assisted living residents after Hurricane Irma. J Am Med Dir Assoc 22 (4 ):918–922.E1, PMID: , 10.1016/j.jamda.2020.10.010.33234448 59. Trebouet E, Prieur S, Dimet J, Lipp D, Orion L. 2012. Cardiovascular emergencies related to the Xynthia storm. Am J Emerg Med 30 (2 ):377–379, PMID: , 10.1016/j.ajem.2011.09.022.22100467 60. Kim I, Locascio J, Sarin RR, Hart A, Ciottone G. 2020. Time series analysis of congestive heart failure discharges in Florida post tropical storms. Acad Emergency Med 27 (suppl 1 ):S298. 61. Tarnoki AD, Turker A, Tarnoki DL, Iyisoy MS, Szilagyi BK, Duong H, et al. 2017. Relationship between weather conditions and admissions for ischemic stroke and subarachnoid hemorrhage. Croat Med J 58 (1 ):56–62, PMID: , 10.3325/cmj.2017.58.56.28252876 62. Quast T, Andel R, Sadhu AR. 2019. Long-term effects of disasters on seniors with diabetes: evidence from Hurricanes Katrina and Rita. Diabetes Care 42 (11 ):2090–2097, PMID: , 10.2337/dc19-0567.31548250 63. Parks RM, Navas-Acien A, Kioumourtzoglou M-A, Anderson GB, Nethery RC, Dominici F. 2021. Tropical cyclone exposure is associated with increased hospitalization rates in older adults. Nat Commun 12 (1 ):1545, PMID: , 10.1038/s41467-021-21777-1.33750775 64. Kanaoka K, Okayama S, Terasaki S, Nakano T, Ishii M, Nakai M, et al. 2020. Role of climatic factors in the incidence of Takotsubo syndrome: a nationwide study from 2012 to 2016. ESC Heart Fail 7 (5 ):2629–2636, PMID: , 10.1002/ehf2.12843.32715646 65. Cruz-Cano R, Mead EL. 2019. Causes of excess deaths in Puerto Rico after Hurricane Maria: a time-series estimation. Am J Public Health 109 (7 ):1050–1052, PMID: , 10.2105/AJPH.2019.305015.30998411 66. Gonzalez JB, Quilichini-Oliver A, Cordero-Jimenez J. 2020. Impact of Hurricane Maria in the incidence of acute coronary syndromes: a single center observational study. J Am Coll Cardiol 75 (11 ):3650, 10.1016/S0735-1097(20)34277-7. 67. McKinney N, Houser C, Meyer-Arendt K. 2011. Direct and indirect mortality in Florida during the 2004 hurricane season. Int J Biometeorol 55 (4 ):533–546, PMID: , 10.1007/s00484-010-0370-9.20924612 68. Yan M, Wilson A, Dominici F, Wang Y, Al-Hamdan M, Crosson W, et al. 2021. Tropical cyclone exposures and risks of emergency Medicare hospital admission for cardiorespiratory diseases in 175 urban United States counties, 1999–2010. Epidemiology 32 (3 ):315–326, PMID: , 10.1097/EDE.0000000000001337.33591048 69. Sunohara D, Miura T, Komatsu T, Hashizume N, Momose T, Kono T, et al. 2021. Relationship between the flood disaster caused by the Reiwa first year east Japan typhoon and cardiovascular and cerebrovascular events in Nagano City: the SAVE trial. J Cardiol 78 (5 ):447–455, PMID: , 10.1016/j.jjcc.2021.06.003.34183228 70. Quast T, Andel R, Sadhu AR. 2019. Long-term effects of disasters on seniors with diabetes: evidence from Hurricanes Katrina and Rita. Diabetes Care 42 (11 ):2090–2097, PMID: , 10.2337/dc19-0567.31548250 71. Brulle RJ, Pellow DN. 2006. Environmental justice: human health and environmental inequalities. Annu Rev Public Health 27 :103–124, PMID: , 10.1146/annurev.publhealth.27.021405.102124.16533111 72. U.S. Department of Homeland Security. 2016. National Disaster Recovery Framework. Washington, DC: Department of Homeland Security. 73. World Health Organization. 2019. Health Emergency and Disaster Risk Management Framework. Geneva, Switzerland: World Health Organization. 74. Davidson TM, Price M, McCauley JL, Ruggiero KJ. 2013. Disaster impact across cultural groups: comparison of Whites, African Americans, and Latinos. Am J Community Psychol 52 (1–2 ):97–105, PMID: , 10.1007/s10464-013-9579-1.23709270 75. Schultz WM, Kelli HM, Lisko JC, Varghese T, Shen J, Sandesara P, et al. 2018. Socioeconomic status and cardiovascular outcomes. Circulation 137 (20 ):2166–2178, PMID: , 10.1161/CIRCULATIONAHA.117.029652.29760227 76. Nethery RC, Katz-Christy N, Kioumourtzoglou MA, Parks RM, Schumacher A, Anderson GB. 2021. Integrated causal-predictive machine learning models for tropical cyclone epidemiology. Biostatistics, 10.1093/biostatistics/kxab047.	36448792	PMC9710380	NO-CC CODE	2022-12-01 23:23:15	no	Environ Health Perspect. 2022 Nov 30; 130(11):116003	utf-8	Environ Health Perspect	2,022	10.1289/EHP11252	oa_other
==== Front Energy Res Soc Sci Energy Res Soc Sci Energy Research & Social Science 2214-6296 2214-6296 Elsevier Ltd. S2214-6296(22)00407-8 10.1016/j.erss.2022.102904 102904 Corrigendum Corrigendum to “Towards a theory of just transition: A neo-Gramscian understanding of how to shift development pathways to zero poverty and zero carbon” [Energy Res. Soc. Sci. 70 (2020) 101789] Winkler Harald University of Cape Town, South Africa 30 11 2022 1 2023 30 11 2022 95 102904102904 © 2020 Elsevier Ltd. All rights reserved. 2020 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 pmcThe authors regret a missing entry [8] in the bibliography. Instead of “!!! INVALID CITATION !!! [8]”, the entry should read:[8] Köhler J, Geels FW, Kern F, Markard J, Onsongo E, Wieczorek A, Alkemade F, Avelino F, Bergek A, Boons F, Fünfschilling L, Hess D, Holtz G, Hyysalo S, Jenkins K, Kivimaa P, Martiskainen M, McMeekin A, Mühlemeier MS, Nykvist B, Pel B, Raven R, Rohracher H, Sandén B, Schot J, Sovacool B, Turnheim B, Welch D, Wells P. 2019, An agenda for sustainability transitions research: State of the art and future directions. Environmental Innovation and Societal Transitions (31) 1–32, https://doi.org/10.1016/j.eist.2019.01.004. The author would like to apologise for any inconvenience caused.	36471756	PMC9710474	NO-CC CODE	2022-12-02 23:21:28	no	Energy Res Soc Sci. 2023 Jan 30; 95:102904	utf-8	Energy Res Soc Sci	2,022	10.1016/j.erss.2022.102904	oa_other
==== Front Ethics Med Public Health Ethics Med Public Health Ethics, Medicine, and Public Health 2352-5525 Elsevier Masson SAS. S2352-5525(22)00107-4 10.1016/j.jemep.2022.100858 100858 Letter to the Editor The resurgence of polio: The effect of the Covid-19 pandemic on polio eradication Niaz F. a Tariq S. a Rana M.S. b Nashwan A.J. c* Fatima I. a Afzal Y. a Tariq R. a a Dow Medical College, Dow University of Health Sciences, Karachi, Pakistan b National Institute of Health, Islamabad, Pakistan c Hamad Medical Corporation, P.O. Box 3050, Doha, Qatar ⁎ Corresponding author. 24 11 2022 2 2023 24 11 2022 26 100858100858 19 11 2022 20 11 2022 © 2022 Elsevier Masson SAS. All rights reserved. 2022 Elsevier Masson SAS 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 Covid-19 Poliomyelitis Re-emergence Vaccination Virus Abbreviations IPV, Injected Poliovirus Vaccine OPV, Oral Poliovirus Vaccine cVDPV, circulating Vaccine Derived Poliovirus WPV, Wild Poliovirus WHO, World Health Organization GPEI, Global Polio Eradication Initiative ==== Body pmcDear Editor, The international focus on the ongoing COVID-19 pandemic has seen a reduction in childhood vaccinations. Owing to disruptions in immunization services, reduced outreach, funding shortfalls, and vaccine misinformation, a staggering 23 million children failed to receive basic childhood vaccines [1]. This number is 3.7 million more than in 2019, and the highest since 2009 [1]. As polio vaccination is among the standard immunization procedures administered to children, this statistic also reflects a global reduction in immunization rates against polio. In 2019, before the COVID-19 pandemic had begun, the global polio immunization coverage amongst 1-year-olds had reached 86%, the highest ever [2]. In 2020, however, when COVID-19 reached a pandemic status, these numbers saw an abrupt fall to 82% [2]. The trend continued in 2021, in which the coverage fell to 80% [2]. This is the lowest global polio immunization coverage since 2008. Poliovirus immunization is carried out via oral or injected vaccines. While the Injected Poliovirus Vaccine (IPV) grants immunity via a killed specimen, the Oral Poliovirus Vaccine (OPV), which is the most commonly used, instead uses an inactivated virus with little to no virulence. The OPV's inactivated virus, however, can undergo mutations as it replicates in the gut. These mutations can rarely restore the virus's transmissibility as well as virulence [3]. These vaccine-derived variants, termed circulating Vaccine Derived Poliovirus (cVDPV) can then spread from person to person and cause infections no different from the naturally found Wild Poliovirus (WPV) in unvaccinated individuals. cVDPV can be further categorized into three different types based on the inactivated poliovirus strain that has undergone mutations. These are numbered as cVDPV1, cVDPV2, and cVDPV3. Among these, cVDPV2 is the most frequently transmitted. Fig. 1 highlights countries with new cVDPV and WPV cases in 2022.Figure 1 Countries with new Poliovirus cases in 2022. The reduced polio immunization coverage has been accompanied by an increment in the emergence of both WPV and cVDPV worldwide. In 2019, a total of 378 cases of cVDPV were reported across only 19 countries [4]. Following this, between January 2020 and April 2022, a report by the CDC found that 33 countries reported 1,856 cases of cVDPV [3]. The majority of these cases involved cVDPV2, with 1081 and 682 cases of cVDPV2 in 2020 and 2021 respectively [4]. Meanwhile, cVDPV1 was responsible for 35 and 16 cases in 2020 and 2021 [4]. As of this report, in September 2022, 14 countries have had 247 new cVDPV cases, including the United States of America [5]. 10 of these are cVDPV1 cases, while 236 are cVDPV2 [4]. The only case of cVDPV3 was reported in Israel [4]. To combat the spread of polio, the World Health Organization (WHO) launched the Global Polio Eradication Initiative (GPEI) in 1988. Since then, two of the three strains of Wild Poliovirus (WPV), WPV2 and WPV3, have been successfully eradicated, reducing WPV cases by approximately 99.99%, with the number of cases going down from over 350,000 in 1988 to just six in 2021 [6]. The GPEI has successfully eliminated WPV in all except two countries - Pakistan and Afghanistan. Together, the two countries have reported 16 WPV1 cases so far in 2022 [7]. 15 of these were in Pakistan and 1 in Afghanistan [7]. It is thus evident that the two nations have faced difficulties in conducting immunization drives in polio-endemic areas. These can be attributed to two factors: Political unrest and civilian refusal to comply with vaccination attempts. The political unrest in Afghanistan rendered vaccination inaccessible to children, a result of the ban on house-to-house Supplementary Immunization Activities since May 2018 [3]. In addition, non-compliant areas have a history of attacking polio vaccination teams due to the myths that envelop immunization. These myths also increase the number of vaccine-hesitant parents, further reducing vaccination rates. Another important reason for this aversion to immunization is security. Many polio-endemic areas regularly find themselves subjected to death threats from nearby militant groups. These groups perceive vaccination efforts as intelligence-gathering attempts, dissuading citizens from complying with immunization teams. The majority of cVDPV cases were reported in regions of Africa and Asia, in countries such as Nigeria, Chad, Yemen, and the Democratic Republic of the Congo. However, In July 2022, a young adult in the US was diagnosed with paralysis caused by cVDPV2 [5]. This is the country's first reported case since 2013 and is not the only reoccurrence of cVDPV2 this year. In June, the UK, where poliovirus was eradicated in 1984, reported traces of genetically related cVDPV from sewage samples in London. Although no cases have been reported, the similar genetic makeup indicates the strong possibility of person-to-person spread. India found itself in a similar situation earlier this year, with cVDPV being detected in sewage from Kolkata despite its eradication in 2014. With this re-emergence of cVDPV in countries free from its grasp for nearly a decade, the question arises: Why? The answer, as has always been for the topic of polio eradication, is immunization. Covid-19-induced lapses in surveillance, reduced available staff and hospital capacities, as well as funding, have had an effect even in the UK, where only 35% of teenagers received the routine polio booster dose in 2021 [8]. Now, as the pandemic slows down, and resurgences of vaccine-preventable diseases occur around the world, the reality of the situation is quickly becoming clear as cVDPV is not the only variant of polio undergoing a resurgence. Malawi and Mozambique, two countries that eradicated poliovirus in 1992 and 1993 respectively, have reported cases this year [9], [10]. The fact that these outbreaks, which can be called side effects of the Covid-19 pandemic, are already palpable, hints at a dire future. Any country with similar vaccination rates of < 80% is thus vulnerable, its population at risk of debilitating paralysis [9], [10]. Even with the presence of ever-improving vaccines, the solution lies not with the equipment but its use. As ever, the key to reaching global polio eradication is immunization coverage. To get there, countries must recognize the potentially debilitating effects, both physical and economic, that the virus can have on a population. Myths surrounding the vaccines must be debunked, and stricter implementation of vaccination drives encouraged. Missed immunizations from the recent past must be covered, and Covid-19-induced laxity is made up for. These initial findings can thus serve as warnings of a future outbreak. One which, with enough preparation, can be weathered better than in recent times. Disclosure of interest The authors declare that they have no competing interest. Human and animal rights The authors declare that the work described has not involved experimentation on humans or animals. Informed consent and patient details The authors declare that the work described does not involve patients or volunteers. Funding This work did not receive any grant from funding agencies in the public, commercial, or not-for-profit sectors. Author contributions All authors attest that they meet the current International Committee of Medical Journal Editors (ICMJE) criteria for Authorship. ==== Refs References 1 Covid-19 pandemic leads to major backsliding on childhood vaccinations, new WHO, UNICEF data shows. https://www.unicef.org/press-releases/covid-19-pandemic-leads-majorbacksliding-childhood-vaccinations-new-who-unicef-data. 2 GHO \| By category \| Polio (Pol3) - Immunization coverage estimates by WHO region. WHO https://apps.who.int/gho/data/view.main.81605?lang=en. 3 Rachlin A. Patel J.C. Burns C.C. Jorba J. Tallis G. O’Leary A. Progress toward polio eradication – Worldwide January 2020-April 2022 Morb Mortal Wkly Rep 71 2022 650 655 4 GPEI-Circulating vaccine-derived poliovirus. https://polioeradication.org/polio-today/polionow/this-week/circulating-vaccine-derived-poliovirus/. 5 Jazeera A. What to know about polio and the first US case in nearly 10 years. https://www.aljazeera.com/news/2022/7/22/what-to-know-about-polio-and-the-first-us-casein-nearly-a-decade-explainer. 6 Poliomyelitis (polio). https://www.who.int/health-topics/poliomyelitis. 7 GPEI-This Week. https://polioeradication.org/polio-today/polio-now/this-week/. 8 Devlin H. Stewart H. Low polio vaccination rates among teenagers risks ‘virulent infection’in UK The Guardian 2022 9 Wild poliovirus type 1 (WPV1) - Malawi. https://www.who.int/emergencies/diseaseoutbreak-news/item/wild-poliovirus-type-1-(WPV1)-malawi. 10 Wild poliovirus type 1 (WPV1) – Mozambique. https://www.who.int/emergencies/diseaseoutbreak-news/item/2022-DON395.	36471883	PMC9710480	NO-CC CODE	2022-12-02 23:21:27	no	Ethics Med Public Health. 2023 Feb 24; 26:100858	utf-8	Ethics Med Public Health	2,022	10.1016/j.jemep.2022.100858	oa_other
==== Front Global Finance Journal 1044-0283 1044-0283 Elsevier Inc. S1044-0283(21)00043-0 10.1016/j.gfj.2021.100645 100645 Article High-speed railway opening and urban green productivity in the post-COVID-19: Evidence from green finance Kong Qunxi a Shen Chenrong a Li Rongrong a Wong Zoey b⁎ a School of Industrial Development, Nanjing University of Finance & Economics, Nanjing, 210003, China b School of Business, Nanjing University, Nanjing, 210093, China ⁎ Corresponding author. 14 5 2021 8 2021 14 5 2021 49 100645100645 10 2 2021 11 5 2021 12 5 2021 © 2021 Elsevier Inc. All rights reserved. 2021 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. In an era during which the COVID-19 pandemic continues to spread, high-speed railway (HSR), as one of the key influencers of urban green development, has a significant impact on urban green finance and green productivity. This paper uses HSR as a quasi-natural experiment to study the effect of HSR openings on green productivity in Chinese cities. The empirical results show that, first, the opening of HSR is conducive to the sustained improvement of green productivity in Chinese cities. Second, the opening of HSR makes a significant contribution to the improvement of green productivity in large-scale cities as well as cities in the east and central regions. Third, the opening of HSR can positively impact urban green productivity through the mechanism of green finance development. However, this positive impact tends to first increase and then decrease over time. As the relationship between “finance” and “environment,” green finance has an important impact on the green development of cities. These findings will provide positive and useful references for cities to formulate reasonable green development plans in the post-COVID-19 era. Keywords HSR Green total factor productivity Green finance Urban development Quasi-natural experiments ==== Body pmc1 Introduction Over the past 40 years of reform and increased global trade, China's rapid economic growth has been accompanied by the process of urbanization. In order to meet growing travel demands and promote urban economic growth, China's State Council approved the Medium- and Long-Term Railway Network Plan in 2004 to connect the country's major cities with bullet trains. The plan's goal was to reach a total of 100,000 km of railroads by 2020, with 12,000 km of high-speed railroads operating at 200 km/h and above. Since 2008, China's high-speed railway (HSR) construction has accelerated. In February of 2008, the Ministry of Railways and the Ministry of Science and Technology signed a cooperation agreement to develop a new generation of high-speed trains; in August, the Beijing-Tianjin Intercity Railway was opened for operation. This was the first high-speed railroad in China with fully independent intellectual property rights on the global stage. In October of the same year, the State Council approved the Medium- and Long-Term Road Network Planning (2008 Adjustment) (after this referred to as Planning), which set the construction target at 120,000 km by 2020. The Planning also focuses on the “four vertical and four horizontal”1 trains and other small passenger lines, along with intercity passenger transport networks in economically developed and densely populated areas. By the end of 2019, the national railroad network reached 139,000 km, including 35,000 km of HSR, accounting for more than 70% of the world's HSR.2 China has already achieved the world's longest HSR network, more than the rest of the world's HSR tracks combined, with HSR construction projects still ongoing. Even more ambitious plans to extend China's HSR network to Europe and Singapore are in the works, along with plans to have an “eight vertical and eight horizontal” HSR network by 2030. HSR is changing and reshaping the urban and economic geography of China. However, the January 2020 outbreak of COVID-19 in Wuhan, a major railroad transportation hub in China, led to partial or even total traffic blockades in most cities across the country, seriously affected the normal operations of high-speed railroad transportation and disrupted the orderly flow of people and materials. Especially considering that the outbreak occurred during the Spring Festival and lasted for a long time, it impacted the nationwide flow of resources and even economic productivity. Urban economic growth was severely affected by the delayed return of workers to jobs in the city under the impact of the pandemic—a large number of enterprises shut down or reduced production, there was a significant reduction in the transportation activities of industrial manufacturing goods and production-oriented raw materials, and a basic stagnation of short-term development in certain manufacturing and infrastructure investment activities set in. Simultaneously, foreign economic policies toward China became highly uncertain due to the pandemic, which in turn has continued to affect China's economic performance (Kong, Peng, Ni, Jiang, & Wang, 2020; Lei & Song, 2020). Foreign policy shocks may affect China's factor markets, financial markets, and the investment behavior of its firms (Zhang, Guo, Wang, & Chen, 2020; Kong, Guo, Wang, Sui, & Zhou, 2020; Kong, Tong, Peng, Wong, & Chen, 2021), while free trade can improve the overall information efficiency of financial markets (Baig, Blau, & Sabah, 2021). Therefore, this paper argues that, in the context of the COVID-19 pandemic, it is crucial to examine the impact of HSR openings on urban productivity from a green finance perspective. The construction of HSR has not only changed the way people travel and transport goods but has also had a profound impact on the regional economic distribution patterns, resource allocation, and industrial structure in China, bringing opportunities for urban green development (Ran, Zhang, & Yang, 2020; Sun, Pofoura, Mensah, Li, & Mohsin, 2020). The opening of HSR can have a direct impact on the improvement of urban green productivity. Firstly, total factor productivity positively impacts expansion in the number of foreign investments by enterprises (Kong, Peng, Zhang, & Wong, 2021), and HSR construction directly affects urban green total factor productivity by increasing regional capital investment (Peng & Wang, 2019). The opening of HSR facilitates foreign and industrial capital from developed to developing and rural cities. It improves the factor utilization efficiency of less advanced cities through the demonstration and technology spillover effects, thus promoting urban green innovation (Sun, Edziah, Kporsu, Sarkodie, & Taghizadeh-Hesary, 2021). Secondly, compared with traditional railroad transportation, HSR also plays a positive role in promoting energy conservation, emission reduction, and energy use efficiency. The relationship between pollution emissions and economic development is critical to the goal of sustainable growth. The construction of HSR can stimulate companies to develop green energy technologies, which can effectively improve energy use efficiency and promote a sustainable green economy (Zhang, Mohsin, Rasheed, Chang, & Taghizadeh-Hesary, 2021; Zhang & Vigne, 2021). The question therefore becomes, does the opening of HSR affect urban green productivity indirectly through other paths, in addition to its direct effect on urban green productivity? Existing studies show that the opening of HSR can affect the allocation of resources, particularly green financial resources. Meanwhile, innovation plays a crucial role in enterprise development, and a good financing environment to support enterprise R&D investment is conducive to promoting enterprise innovation activities (Zhang & Guo, 2019; Zhang, Zhuge, & Freeman, 2020). Green finance is an important way to influence green production efficiency (Naeem et al., 2021). Financial institutions can encourage green and energy-saving enterprises to innovate, upgrade, and improve their own production efficiency (Yoshino et al., 2019). Financial institutions can also attract private participation in green project investment by improving green credit guarantees and increasing green financial investment opportunities, thus promoting green economic development (Taghizadeh-Hesary and Yoshino, 2019). Therefore, this paper argues that green finance is an important means to influence the relationship between HSR openings and green total factor productivity. Based on the above analysis, this paper's motivation is to examine the impact and effects of HSR construction on green total factor productivity in Chinese cities, from the perspective of green finance and in the context of the current reality of the COVID-19 pandemic. First, this paper uses the panel data of 285 prefecture-level and above cities in China from 2003 to 2018 to measure urban green total factor productivity using the Super Slack-based Measure (SBM) model with “non-expected output.” Second, this paper adopts the difference-in-difference method (DID) to study the effect of HSR on the green total factor productivity of the urban economy. This paper further examines the heterogeneous effects of HSR construction on green total factor productivity in Chinese cities by dividing different city subsamples. Finally, based on the perspective of green finance, the mechanism of the effect of HSR openings on green total factor productivity in Chinese cities is examined. The main contributions of this paper are primarily in the following three aspects: (1) With the increasing threat of environmental problems due to industry, green total factor productivity has received more and more attention as the key to achieving green economic development. However, existing studies have mainly focused on the impact of HSR construction on total factor productivity. Only a small number of scholars to date have explored the relationship between infrastructure and specifically green total factor productivity. This paper enriches existing research by studying the impact of HSR openings on urban green total factor productivity from the macro-institutional context of HSR construction in China. (2) Existing studies generally agree that HSR construction has a significant impact on productivity but lack a rigorous causal identification framework for the relationship between the two. For this paper, a quasi-natural experiment has been conducted with the opening of HSR, and a difference-in-difference (DID) method was used to test the empirical evidence. The potential endogeneity problem is solved by replacing variable measures and instrumental variables to ensure the robustness of this paper's findings while providing reliable empirical support for related studies. (3) Green finance, as the relationship between “finance” and “environment,” will have an important impact on the green development of cities. This paper incorporates green finance into the analysis framework and provides an in-depth analysis of the path of green total factor productivity in cities due to the opening of HSR, which provides a useful reference for cities to formulate reasonable green development plans in the post-COVID-19 era. The structure of the remainder of this paper is organized as follows: Section 2 reviews the relevant literature; Section 3 develops identification strategies and conducts validity tests; Section 4 consists of empirical tests and analysis of results; Section 5 discusses the mechanism of the effect of HSR openings on urban green productivity based on a green finance perspective; Section 6 presents the conclusions and policy recommendations of this paper. 2 Literature review The relationship between transportation infrastructure development and various aspects of economic development has been richly studied in academia (Diao, Leonard, & Sing, 2017; Dong & Zhu, 2016; Pan & wen Luo, 2020). Scholars in China and abroad have discussed the impact of transportation infrastructure construction on economic growth in terms of long-term economic growth effects, trade gains, land value, technology spillover from local markets, residential employment, population growth, and urbanization (Zhu & Diao, 2016). These studies largely focus on two main aspects of the problem. The first aspect is concerned with the causal relationship between transportation infrastructure construction and economic growth. The second aspect is concerned with the relationship between transportation infrastructure construction and regional economic agglomeration—does the reduction in transportation costs lead to the diffusion of economic activities to the surrounding areas, or does it reinforce the concentration of production in existing urban centers? Empirical research efforts have focused on assessing the firm performance, economic benefits, and social welfare of the construction of infrastructure such as rail networks, road networks, interstate highway systems, and intra-city rail transit systems in cities across countries over time (Banerjee, Duflo, & Qian, 2012; Wang, Lin, & Gao, 2020). Regarding HSR construction and green economic growth, some scholarly found that the opening of HSR will promote urban air quality improvement and enhance green development in cities along the constructed route (Sun & Zhang, 2021). It will also promote green total factor productivity in cities by alleviating the distorted situation of labor factor allocation, especially the significant effect on green total factor productivity in large and medium-sized cities (Peng & Wang, 2019). HSR investments can enhance the externalities of firm agglomeration and increase productivity by increasing urban density, generating “broader economic benefits” (Huang & Wu, 2020; Venables, 2007; Zhang, Sun, & Yao, 2019). Thus, HSR construction and urban development can mutually reinforce each other and jointly promote economic development. As long as the three conditions of “positive economic externalities” (e.g., agglomeration, labor market economics, and labor quality), “investment factors” (including the availability, size, location, and timing of investment), and “political factors” (including any accompanying policy measures) are present, urban economies will achieve sustainable growth (Banister & Berechman, 2001; Zhang & Li, 2019； Cheng, Wang, Peng, & Kong, 2020). The construction of HSRs can promote the concentration of local financial resources. The impact of railroads in neighboring regions on the concentration of local financial resources is manifested as the spatial agglomeration effect and the spatial spillover effect (Wang et al., 2020). The spatial agglomeration of financial services also positively impacts the quality of urban economic growth (Wong, Li, Zhang, Kong, & Cai, 2021). Zhang et al. (2021) found that the concentration of green financial resources can attract capital and technology-oriented production activities by studying the development of green economies along the Belt and Road. Taghizadeh-Hesary et al., 2020, Taghizadeh-Hesary et al., 2021 also found that green finance plays a crucial role in promoting the development of green industries by studying the new energy economy in Japan. Green financial markets take environmental costs into account to internalize the negative externalities of environmental pollution (Zhang, Du, Zhuge, et al., 2019; Zhang, Tong, & Li, 2020), prompting economic agents with high environmental costs and low factor productivity to withdraw from market competition (Zhang, Du, & Chen, 2019) and thereby improving the green total factor productivity. Cities are one of the most important vectors of socio-economic development in China. The development of cities and surrounding metropolitan areas promotes industrial infrastructure development, urban-rural development, information sharing, technological development, and environmental protection. Through these integrations, common economic benefits and social welfare are generated across urban areas (Fang & Yu, 2017; Kong, Shen, Sun, & Shao, 2020; Ni, 2008). In the case of China, combined with the current context of green development, transportation, as one of the three major areas of energy consumption and carbon emissions in the economy and society, must practice ecological civilization, strengthen energy conservation and emission reduction, and meet additional requirements to achieve green urban development through green finance. 3 Identification strategy and validity test 3.1 Data sample The opening of the Qin–Shen passenger line in 2003 launched China's railroad system into the era of high speed. This paper takes 2003 as the starting point of the study to examine the impact of the opening of HSRs on green total factor productivity in Chinese cities. Since the State Council adjusted the establishment of individual prefecture-level cities during this study's sample examination, four cities, Lhasa, Sansha, Haidong, and Chaohu, were excluded from the screening sample for the sake of uniformity. This study selected 285 prefecture-level and above cities in China from 2003 to 2018 as the research sample and obtained 4560 city-annual observations. Information regarding cities opening HSR has been taken mainly from the China Railway Yearbook, the China Railway Corporation website, and the National Railway Administration of the People's Republic of China. The economic data at the prefecture city level, such as GDP, population, industry, and government expenditure, have mainly been obtained from the China City Statistical Yearbook and the China Regional Statistical Yearbook. In response to the missing data of some cities, this paper supplements certain data by adopting the mean value and smoothing methods. 3.2 Variable definition Urban Green Total Factor Productivity (GTFP) is the main explained variable in this study. The inclusion of environmental factors in the analysis framework of traditional total factor productivity improves traditional total factor productivity. This improvement is also more in line with China's current demand for the high-quality economic development of “innovation + green.” In this paper, each city is considered as a production decision unit to construct the best practice boundary for urban development in each period. Using the Super Slack-based Measure (SBM) model with “undesired outputs,” this paper uses MaxDEA7 Ultra software to measure urban green total factor productivity under environmental constraints by integrating desired and undesired outputs into a unified analytical framework. The input, desired output, and non-desired output indicators used in measuring green total factor productivity are: (1) Input variables, including capital input and labor input. Capital inputs are calculated using the perpetual inventory method and are calculated as follows: k t = k t−1(1 − δ t) + I t/p t, k tand k t−1 are the capital stocks in periods t and t + 1. Depreciation rate (δ t) is set to 10.96% regarding Shan (2008) article, I t is fixed asset input, and p t is the investment price index of the city's province in the current period. The labor input is measured by the number of employees in each city in the calendar year. (2) Desired output is measured by using the real GDP of each city in constant 1990 prices. (3) The non-desired output uses industrial wastewater, industrial sulfur dioxide, and industrial smoke (dust) emissions. It should be noted that because of the high correlation between these three pollutants, this paper refers to Li, Liu, and Wang (2019) to represent the non-desired output with the composite pollution index calculated using the entropy method.3 The interaction term HSR × post is the key explanatory variable in this paper. HSR represents a dummy variable for whether the city opened the HSR. If city i opened HSR in year t, then the value is 1; otherwise, the value is 0. post represents the time variable of HSR opening. If a year is the year of HSR opening in city i or after, it takes the value of 1; otherwise, it takes 0. Seven control variables are selected in this paper, namely foreign capital dependence (fdi), industrial structure (ind), infrastructure level (road), administrative level (aeg), government financial support (gov), population size (poe), and economic development level (pgdp). Foreign capital dependence (fdi) is expressed by the proportion of the main business income of the three enterprises to the total income of the high-tech industry in the region. Industrial structure (ind) is expressed using the ratio of tertiary industry output value to secondary industry output value in the city, referring to He, Luo, and Chen (2020). Infrastructure level (road) refers to Wang and Miao (2015) and uses the road area per capita as a proxy variable for infrastructure. Administrative level (aeg) is a dummy variable. It takes the value of 1 if the city is the province's capital city; otherwise, it takes the value of 0. Government financial support (gov) is expressed using the ratio of government financial expenditure to regional GDP. Population size (poe) is measured using the number of permanent residents in the city. Economic development level (pgdp) is measured using the growth rate of urban real GDP per capita. The information on the variable measures is summarized in Table 1 .Table 1 Variable names and measurement methods. Table 1Variable name Measurement method Urban Green Total Factor Productivity (GTFP) Inputs, desired outputs, and non-desired outputs. Foreign capital dependence (FDI) The main business income of the three enterprises/the total income of the high-tech industry. Industrial structure (ind) The ratio of tertiary industry output value to secondary industry output value. Infrastructure level (road) The road area per capita. Administrative level (AEG) A dummy variable: if the city is the province's capital city, it takes 1; otherwise, it takes 0. Government financial support (Gov) The ratio of government financial expenditure to regional GDP. Population size (poe) The number of permanent residents in the city. Economic development level (PGDP) The growth rate of urban real GDP per capita. Organized by the author. 3.3 Identification strategy To test the effect of the opening of HSR on urban green total factor productivity, this paper uses a difference-in-difference model (DID) for empirical analysis. The DID method is the result of a foreign economics community which emerged in the late 1980s, an econometric method based on experimental design or quasi-experimental design (Zhao, 2017) which uses the idea of randomized experiments as the basis of causal effects and can effectively solve the endogeneity problem in regression (Angrist & Pischke, 2009). The empirical model is set as follows:(1) GTFPit=∂0+∂1HSRit×postit+∂2controlsit+ui+λt+εit where GTFP it is the explanatory variable of this paper's study and denotes the urban green total factor productivity of city i in year t. HSR it denotes whether city i opened high in year t, and post it denotes the time variable for the opening of HSR in city i. The interaction term coefficient ∂ 1 is the coefficient of focus in this model, capturing the effect of the initiative of HSR opening on the green total factor productivity of the city. controls it represents the set of all control variables. u i is the city fixed effect, which captures other city characteristics that do not vary over time; λ t is the year fixed effect, which controls for factors that vary over time but cannot be observed; and ε it is the random disturbance term. 3.4 Validity test This paper uses a univariate DID method to investigate the effect of the opening of HSR on urban green total factor productivity. Table 2 presents the estimated results for the control and experimental groups before and after the opening of HSR. Table 2 shows that the regression results are divided into the control and experimental groups. The coefficient estimates before and after the opening of HSR are positive, and the coefficient of the experimental group is larger than that of the control group. Horizontally, Diff is positive for both, and Diff of the experimental group is larger than Diff of the control group; vertically, Diff is significantly negative before the opening of HSR, but it is significantly positive after the opening of HSR. The above results indicate that the experimental group's green total factor productivity has improved after the opening of HSR. In terms of economics, the opening of HSR increases the green total factor productivity of cities in the control group by 0.022 units and increases the green total factor productivity of cities in the experimental group by 0.039 units. Therefore, the empirical results show that the opening of HSR can promote the green total factor productivity of cities.Table 2 Univariate DID results. Table 2GTFP Control group Experimental group Diff Before 0.0177 0.0034 −0.0143⁎⁎⁎ (−6.1011) Post 0.0221 0.0394 0.0173⁎⁎⁎ (5.5563) Diff1 0.0044 (0.4659) 0.0360 (0.5741) 0.0316⁎⁎⁎ (7.5676) , , and ** denote significance levels of 10%, 5%, and 1%; t-values are in parentheses. Diff denotes the mean GTFP value of cities with high-speed rail minus the mean GTFP value of cities without high-speed rail; Diff1 denotes the mean GTFP value of cities in the Post period minus the mean GTFP value of cities in the Before period. 4 Empirical results and analysis 4.1 Baseline regression results Referring to Banister and Berechman (2001), this paper further applies the DID method to examine the impact of HSR openings on urban green total factor productivity. The results are shown in Table 3 . The coefficient estimates of the interaction terms HSR × before1 and HSR × before2 are negative but not significant, indicating that the model passed the parallel trend test. Meanwhile, the coefficient estimates of the interaction terms HSR × post, HSR × post1, HSR × post2, and HSR × post3 are all positive and pass the significance test at the 1% level. This indicates that the opening of HSR has a longer-term promotion effect on the improvement of green total factor productivity in cities. Therefore, combining the regression results in Table 3, Table 4 , this paper concludes that the opening of HSR can promote urban green total factor productivity. In addition, the regression results of control variables show that the level of industrial structure, infrastructure level, government financial support, real GDP per capita growth rate, and whether the city in question is a provincial capital city have positive effects on the enhancement of green total factor productivity in cities. In contrast, the foreign investment dependence and population sizes of cities have a negative impact on the enhancement of green total factor productivity in cities.Table 3 Impact of HSR openings on urban green total factor productivity: DID test. Table 3Variables Urban Green Total Factor Productivity(GTFP) (1) (2) (3) (4) (5) (6) HSR×before1 −0.0032 (−0.2244) −0.0008 (−0.6531) HSR×before2 −0.0012 (−0.8008) −0.0080 (−0.5508) HSR×post 0.0125⁎⁎⁎ (4.6685) 0.0711⁎⁎⁎ (6.0494) HSR×post1 0.0044⁎⁎⁎ (4.3743) 0.0777⁎⁎⁎ (5.8083) HSR×post2 0.0040⁎⁎⁎ (6.2429) 0.0664⁎⁎⁎ (5.5514) HSR×post3 0.0206⁎⁎⁎ (5.0474) 0.0671⁎⁎⁎ (4.6600) FDI −0.0094⁎⁎ (−2.1349) −0.0186⁎⁎ (−2.0636) −0.7346⁎⁎ (−2.2196) ind 0.1099⁎⁎⁎ (6.1607) 0.1515⁎⁎⁎ (5.2103) 0.5240⁎⁎⁎ (4.1641) road 0.0218⁎⁎⁎ (5.8258) 0.0382⁎⁎⁎ (4.3365) 0.0280⁎⁎⁎ (6.3732) AEG 0.0013⁎⁎⁎ (6.3106) 0.0035⁎⁎⁎ (4.5530) 0.0228⁎⁎⁎ (5.3567) Gov 0.0633⁎⁎⁎ (3.4495) 0.0314⁎⁎⁎ (4.5163) 0.0861⁎⁎⁎ (4.5830) poe −0.0467 (−1.0754) −0.1305 (−0.9744) −0.0774 (−0.6291) PGDP 0.0419⁎⁎⁎ (4.2078) 0.0308⁎⁎⁎ (5.7704) 0.0104⁎⁎⁎ (4.2080) Cons 0.0301 (1.0156) 0.1776 (0.5478) 0.0029 (0.2548) 0.3863 (0.8846) 0.1637 (0.9523) 0.0569 (0.7260) Year Yes Yes Yes Yes Yes Yes City Yes Yes Yes Yes Yes Yes N 4560 4560 4560 4560 4560 4560 R2(Within) 0.5003 0.6023 0.6586 0.5061 0.6527 0.5892 , , and ** denote significance levels of 10%, 5%, and 1%, respectively; t-values are in parentheses. Year denotes time fixed effects and city denotes city fixed effects. Table 4 Regression results of samples at different time periods. Table 4Variables Urban Green Total Factor Productivity(GTFP) 2011–2012 2013–2014 2015–2016 2017–2018 (1) (2) (3) (4) HSR×post 0.0079⁎⁎⁎ (5.4145) 0.0601⁎⁎⁎ (3.8386) 0.0188⁎⁎⁎ (2.8365) 0.0035⁎⁎⁎ (3.1884) FDI −0.0587⁎⁎⁎ (−2.7410) −0.1130⁎⁎ (−2.1836) −0.0502⁎⁎ (−2.2948) −0.0035⁎⁎ (−2.4884) ind 0.1129⁎⁎⁎ (2.8802) 0.1389⁎⁎⁎ (4.3125) 0.0327⁎⁎⁎ (3.2349) 0.0058⁎⁎⁎ (3.4723) road 0.1402⁎⁎ (2.3238) 0.6787⁎⁎⁎ (3.2963) 0.7401⁎⁎⁎ (3.6085) 0.0010⁎⁎⁎ (3.0968) AEG 0.6937⁎⁎⁎ (3.3176) 0.0250⁎⁎⁎ (3.0747) 0.0305⁎⁎⁎ (3.7331) 0.0046⁎⁎ (2.4398) Gov 0.0247⁎⁎⁎ (3.0591) 0.0038⁎⁎⁎ (3.6122) 0.0035⁎⁎⁎ (4.4069) 0.0012⁎⁎ (2.0150) poe −0.0038 (−0.6158) −0.0544 (−0.7736) −0.1248 (−1.0804) −0.0023 (−0.3170) PGDP 0.0537⁎⁎⁎ (3.7632) 0.0454⁎⁎⁎ (4.6037) 0.0206⁎⁎⁎ (3.1695) 0.0238⁎⁎⁎ (4.5041) Cons 0.0023 (0.3143) 1.1591 (1.0116) 0.0023 (0.3170) 1.1144 (0.6671) Year Yes Yes Yes Yes City Yes Yes Yes Yes N 570 570 570 570 R2(Within) 0.5701 0.4063 0.5098 0.6062 , , and ** denote significance levels of 10%, 5%, and 1%, respectively; t-values are in parentheses. 4.2 Robustness tests 4.2.1 Sample of different periods To examine the impact of HSR openings on urban green total factor productivity as well as to exclude the impact of other events on the level of urban economic development during the sample period, the sample interval of 2011–2018 has been retained for robustness testing. First, as Table 4 shows, the sample interval in column (1) is 2011–2012, column (2) is 2013–2014, column (3) is 2015–2016, and column (4) is 2017–2018. Each column has a time interval of two years, and the time interval is narrowed for more precise results and more robust conclusions. Second, the explanatory variable HSR × post is significantly positive at the 1% level in each column, indicating that the opening of HSR is beneficial to the improvement of urban green total factor productivity. After adding the control variables, the explanatory variable HSR × post is still significantly positive, verifying the basic conclusion of this paper. Thus, the results of the benchmark regression are consistent with the conclusion that the opening of HSR can promote the improvement of urban green total factor productivity. 4.2.2 Re-measure urban GFTP In order to ensure the robustness and reliability of the estimation results, this paper remeasures GTFP after adjusting the expected output indicator by drawing on Zhang and Luo (2019). Urban green development is a comprehensive concept, and the desired output should include economic aspects, social welfare, and ecological environment aspects. Therefore, this paper selects multiple output indicators from economic growth output, social welfare, and ecological environment; applies the principal component analysis to downscale them; and combines them into desired output indicators. Specifically, the real GDP of each city in constant 1990 prices is selected to measure the economic output of the city; the total retail sales of social consumer goods per capita, public library collections per 100 people, and hospital and health center beds per 10,000 people are selected to reflect the social welfare level of the city; and the green area per capita is selected to measure the ecological environment output. The regression results obtained using the remeasured GTFP are shown in Table 5 . From Table 5, it can be seen that the coefficient estimates of the key explanatory variables HSR × post, HSR × post1, HSR × post2, and HSR × post3 are significantly positive. This indicates that urban green total factor productivity continues to increase immediately after the opening of HSR, as well as one, two, and three years after the opening of HSR. This means that the impact effect of HSR opening has some dynamic persistence. In addition, the results of the control variables are also fundamentally consistent with the results in Table 4. With the exceptions of foreign capital dependence (FDI) and population size (Poe), all items are significantly positive. It can therefore be concluded that the baseline regression results have not changed significantly, and the results of this paper are robust.Table 5 Changing the measurement of urban GTFP. Table 5Variables Remeasured Urban Green Total Factor Productivity(GTFP) (1) (2) (3) (4) HSR×post 0.0136⁎⁎⁎ (3.0241) 0.0162⁎⁎⁎ (4.1859) HSR×post1 0.0075⁎⁎⁎ (3.5448) 0.0056⁎⁎⁎ (3.3121) HSR×post2 0.0503⁎⁎ (2.0153) 0.0082⁎⁎⁎ (4.7379) HSR×post3 0.0195⁎⁎⁎ (2.9410) 0.0021⁎⁎⁎ (3.1358) FDI −0.0086⁎⁎ (−2.1309) −0.0711⁎⁎ (−2.2473) ind 0.0086 (0.8309) 0.0855 (0.7134) road 0.0094⁎⁎⁎ (3.8243) 0.0276⁎⁎⁎ (2.7451) AEG 0.0518⁎⁎⁎ (0.4538) 0.0305⁎⁎⁎ (0.0745) Gov 0.0970 (1.3458) 0.0449 (0.7755) poe −0.0240 (−0.5821) −0.0999 (−1.5369) PGDP 0.0221⁎⁎⁎ (4.8774) 0.0044⁎⁎⁎ (3.1194) Cons 0.0038 (0.4814) 0.0263 (0.5857) Year Yes Yes Yes Yes City Yes Yes Yes Yes N 4560 4560 4560 4560 R2(Within) 0.4706 0.5025 0.6805 0.5942 , , and ** denote significance levels of 10%, 5%, and 1%, respectively; t-values are in parentheses. 4.2.3 Instrumental variable method Considering the potential endogeneity issues that may affect the impact of HSR openings on green total factor productivity, this study used the 2SLS method to re-run the regression to further test the robustness of the empirical results. Before conducting the regression, this study used spatial geographic information to calculate the geographic development cost as an instrumental variable for the opening of HSR. The regression results are shown in Table 6 , where control variables are added in the even columns (not in the odd columns). Observing Table 5, it is evident that columns (1)–(2) are the regression results of the first stage, the IV (minimum spanning tree) results are significantly negative, and the F-value of the first stage is 63.0829, which excludes the problem of weak instrumental variables. Columns (3)–(4) show the regression results of the second stage, and the coefficient estimates of the key explanatory variable HSR × post are significantly positive at the 1% level. As can be seen, solving the endogeneity problem also did not change the baseline regression results, indicating that the conclusions of this paper are reliable.Table 6 Estimation results of instrumental variable method. Table 6Variables HSR GTFP (1) (2) (3) (4) HSR×post 0.0377⁎⁎⁎ (5.0328) 0.0193⁎⁎⁎ (6.1697) IV (Minimum Spanning Tree) −0.0356⁎⁎⁎ (−5.0412) −0.0131⁎⁎⁎ (−4.4973) FDI −0.0098⁎⁎ (−2.1396) −0.1450⁎⁎ (−2.2563) ind 0.1130 (1.1907) 0.0378 (1.3130) road 0.0220⁎⁎⁎ (3.8384) 0.0034⁎⁎⁎ (3.5427) AEG 0.0013⁎⁎⁎ (4.3092) 0.0309⁎⁎⁎ (5.5075) Gov 0.0635 (1.4535) 0.1293 (1.0551) poe −0.0469 (−1.0789) −0.0301 (−0.7727) PGDP 0.0319⁎⁎⁎ (3.2076) 0.0608⁎⁎⁎ (4.0532) Cons 0.1718 (0.5293) 0.4003 (0.9163) 0.1776 (0.5477) 0.3857 (0.8832) First stage F value 85.0987 63.0829 Year Yes Yes Yes Yes City Yes Yes Yes Yes N 4560 4560 4560 4560 R2(Within) 0.6124 0.5565 0.4823 0.5964 , , and ** denote significance levels of 10%, 5%, and 1%, respectively; t-values are in parentheses. 4.3 Heterogeneity analysis In this paper, the total sample is divided into large-scale cities and small- and medium-scale cities by city size; the sample is also divided and into eastern, central, and western cities by city region. This study conducted a heterogeneity analysis of the impact effect of the opening of HSR. The empirical results are shown in Table 7 . The results of the grouped regression based on city size show that the coefficient estimates of HSR × post, HSR × post1, HSR × post2, and HSR × post3 are positive in each column and pass the 1% significance level test. This indicates that the impact effect of HSR opening is not affected by the heterogeneity of city size. The results based on different regional groupings show that the coefficient estimates of the core explanatory variables pass the significance level test of at least 5% in the eastern and central regions, although they are not significant in the western region. This indicates that the opening of HSR makes a significant contribution to the improvement of green total factor productivity in cities in the eastern and central regions, but that this contribution is not significant for cities in the western region. Meanwhile, a longitudinal comparison of the magnitude of the coefficient of HSR × post reveals that the opening of HSR gives a longer-term boost to urban green total factor productivity. This effect gradually strengthens over time in large-scale and central cities; however, the opposite is true in small- and medium-scale cities and eastern cities. Therefore, this paper concludes that there are differences in the impact of HSR openings on green total factor productivity in cities of different sizes and regions. The differences are manifested in whether the boosting effect on green total factor productivity is significant and whether this boosting effect gradually strengthens over a certain period.Table 7 Regression results for subgroups. Table 7Variables Panel A: Grouped by City Size Panel B: Grouped by Urban Area Large-scale cities Small and medium-sized cities East area Central area West area (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) HSR×post 0.1594⁎⁎⁎ (3.4304) 0.0172⁎⁎⁎ (4.3552) 0.1154⁎⁎⁎ (3.8547) 0.0317⁎⁎ (2.0912) 0.0156 (1.1267) HSR×post1 0.1485⁎⁎⁎ (3.4253) 0.0488⁎⁎⁎ (4.0526) 0.1640⁎⁎⁎ (3.4098) 0.0178⁎⁎⁎ (3.5638) 0.0081 (1.0854) HSR×post2 0.2540⁎⁎⁎ (4.2889) 0.0295⁎⁎⁎ (3.4174) 0.1183⁎⁎⁎ (3.7726) 0.0753⁎⁎⁎ (3.3574) 0.0163 (0.9652) HSR×post3 0.2820⁎⁎⁎ (5.6667) 0.0060⁎⁎⁎ (3.8212) 0.1047⁎⁎⁎ (4.6868) 0.0591⁎⁎ (2.2176) 0.0478 (1.2053) Cons 0.0044 (0.0486) 0.0022 (0.2077) 0.0524 (0.4672) 0.0070 (0.8707) 0.0611 (0.5908) 0.0063 (0.7520) 0.0461 (0.4747) 0.0073 (0.6702) 0.0219 (0.5071) 0.0052 (0.7125) controls Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Year Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes City Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes N 2016 2016 2544 2544 1616 1616 1744 1744 1200 1200 R2(Within) 0.4043 0.5222 0.6047 0.5392 0.5226 0.6158 0.6073 0.5198 0.5370 0.5936 (1) Cities with populations above 1 million are classified as large-scale cities and those with populations below 1 million are classified as small- and medium-scale cities by city size; (2) , , and ** denote significance levels of 10%, 5%, and 1%, respectively; t-values are in parentheses. Controls denote a set of control variables controlling for the previously described. 5 Mechanism testing: a green finance perspective The empirical results in the previous section show that the opening of HSR significantly increases the green total factor productivity of cities. The next question to be considered is, how does the opening of HSR affect urban green total factor productivity? That is, what is the mechanism of action by which the opening of HSR affects the green total factor productivity of cities? The opening of HSR facilitates the flow and development of financial capital between cities, which in turn promotes the movement of green financial resources between cities. Therefore, from the perspective of green finance, this paper examines how the opening of HSRs can “lend” green finance to green development in order to investigate the mechanism of influence of the opening of HSRs on the green total factor productivity of cities. 5.1 Impact of HSR on GFTP from a green finance perspective To illustrate the mechanism of action by which the opening of HSR promotes the development of green finance and thus urban green total factor productivity, the model in this section is set up as follows:(2) GTFPit=∂0+∂1HSRit×postit×GFit+∂2HSRit×postit+∂3HSRit×GFit+∂4postit×GFit+∂5controlsit+ui+λt+εit where GF it denotes the green financial resources of city i in year t. At present, scholars measure green financial resources in the following ways: (1) Yin, Sun, & Xing (2021) constructed an evaluation index system of green finance development levels, including five secondary index systems of green credit, green securities, green insurance, green investment, and carbon finance. They used a combination of subjective and objective weighting methods to measure the regional green finance development level. The indicator weights refer to the weights of Li and Xia (2014) in the China Green Finance Report (2014). (2) Zhu, Zhu, Huang, and Huang (2021) measured with the scale of green bonds owned by cities. Wen, Lin, and Liu (2021) argued that the indicator of the allocation of financial resources in the green environmental protection field can be used as a proxy indicator of green finance, representing the greening of the financial industry's operation and investment decisions, i.e., all financial activities that support the green energy-saving industry belong to the scope of green finance. The measurement is as follows: green finance indicator = financial resources of green industries/financial resources of all industries. The third approach is consistent with the frontier trend of modern green finance research. This paper also draws on that approach by dividing the loans obtained by energy-saving and environment-friendly enterprises in a city by the loans obtained by all listed companies in that city as an indicator of green finance activities. The calculation formula is: GF it = the sum of borrowings of environmental protection enterprises in city i in year t/the sum of borrowings of enterprises in all industries in city i in year t. If the coefficient of ∂ 1 is significantly positive, it indicates that the opening of HSR promotes the factor flow of green financial resources between cities; if the coefficient of ∂ 1 is significantly negative, it indicates that the opening of HSR hinders the factor flow of green financial resources between cities. The meanings of the remaining explanatory variables are consistent with the baseline regression model, and the regression results are shown in Table 8 .Table 8 Mechanisms testing the impact of HSR openings on urban GTFP. Table 8Variables Urban Green Total Factor Productivity(GTFP) (1) (2) (3) (4) HSR×post×GFit 0.0194⁎⁎⁎ (4.5995) 0.0024⁎⁎⁎ (3.5841) HSR×post1 × GFit 0.0605⁎⁎⁎ (3.2286) 0.0562⁎⁎⁎ (4.1749) HSR×post2 × GFit 0.1625⁎⁎⁎ (4.4299) 0.1152⁎⁎⁎ (3.5666) HSR×post3 × GFit 0.1043⁎⁎ (2.1110) 0.0734⁎⁎⁎ (4.3062) Cons 0.0124 (0.3525) 0.0092 (0.2742) 0.0032 (0.6672) 0.0072 (1.3566) controls No Yes No Yes Year Yes Yes Yes Yes City Yes Yes Yes Yes N 4560 4560 4560 4560 R2(Within) 0.5205 0.5023 0.6028 0.6346 (1) , , and ** denote significance levels of 10%, 5%, and 1%, respectively; t-values are in parentheses. (2) controls denote a set of control variables controlling for the previously described. In Table 8, the coefficient estimates of the interaction term HSR × post1 × GF it are positive and pass the significance level test of at least 5%. This indicates that green financial resources enhance the positive impact of HSR opening on green total factor productivity. Meanwhile, comparing the coefficient magnitudes of HSR × post1 × GF it, HSR × post2 × GF it, and HSR × post3 × GF it reveals that the opening of HSR can have a positive effect on green total factor productivity in a longer period. However, the effect of this contribution tends to first increase and then decrease. The above results indicate that the opening of HSR can positively affect urban green total factor productivity through the mechanism of promoting the development of green finance. 5.2 Heterogeneity test from a green finance perspective Table 9 presents the results of grouped regressions on the impact of HSR opening on urban green total factor productivity from green finance. When the regression samples include large-scale cities, small- and medium-sized cities, and east and central cities, the coefficient estimates of the interaction term HSR × post × GF it are positive and pass the 1% significance level test. When the regression sample includes western cities, the coefficient estimates of the interaction term HSR × post × GF it are positive but insignificant. The above results indicate that with the development of green finance, the opening of HSR has a significant contribution to the green total factor productivity development of large-scale cities, small- and medium-sized cities, and cities in the east and central regions, but that the effect on the green total factor productivity development of cities in the west is not significant. Meanwhile, the regression coefficients of HSR × post1 × GF it, HSR × post2 × GF it, and HSR × post3 × GF it show that the opening of HSR will have a boosting effect on urban green total factor productivity through the development of green finance over a certain period of time. This boosting effect tends to strengthen gradually in large-scale cities. However, this result is reversed in small- and medium-sized, eastern, and western cities. Therefore, the impact of HSR opening on green total factor productivity development is different in cities of different sizes and regions from the perspective of green finance. The above results further support the theoretical mechanism that the opening of HSR triggers the development of green finance and thus affects urban green total factor productivity.Table 9 Mechanisms testing the impact of HSR openings on urban GTFP: Regression in difference groups. Table 9Variables Panel A: Grouped by City Size Panel B: Grouped by Urban Area Large-scale cities Small and medium-sized cities East area Central area West area (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) HSR×post×GFit 0.1271⁎⁎⁎ (3.4954) 0.0195⁎⁎⁎ (4.3727) 0.2453⁎⁎⁎ (4.2722) 0.0599⁎⁎⁎ (3.7942) 0.0619 (1.2874) HSR×post1 × GFit 0.1043⁎⁎⁎ (4.0906) 0.0464⁎⁎⁎ (5.8027) 0.1451⁎⁎⁎ (5.4566) 0.0845⁎⁎⁎ (4.9723) 0.0643 (0.9054) HSR×post2 × GFit 0.1487⁎⁎ (2.1124) 0.0123⁎⁎⁎ (4.2872) 0.1172⁎⁎⁎ (4.3825) 0.0631⁎⁎ (2.0189) 0.0452 (1.1768) HSR×post3 × GFit 0.1686⁎⁎⁎ (4.5213) 0.0062⁎⁎⁎ (4.5532) 0.0906⁎⁎ (2.3146) 0.0598⁎⁎ (2.2054) 0.0245 (0.6592) Cons 0.0456 (0.4865) 0.0756 (0.8677) 0.0345 (0.0082) 0.03492 (0.2956) 0.0141 (0.4092) 0.0238 (0.1863) 0.0770 (0.7427) 0.0634 (0.5803) 0.0234 (0.7427) 0.0318 (0.6157) controls Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Year Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes City Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes N 2016 2016 2544 2544 1616 1616 1744 1744 1200 1200 R2(Within) 0.5173 0.6222 0.7047 0.6392 0.5026 0.6558 0.4673 0.6198 0.5364 0.6815 (1) Cities with populations above 1 million are classified as large-scale cities and those with populations below 1 million are classified as small- and medium-scale cities by city size; (2) , , and ** denote significance levels of 10%, 5%, and 1%, respectively; t-values are in parentheses. Controls denote a set of control variables controlling for the previously described. 6 Conclusions and policy implications Investments in transportation infrastructure are thought to promote local economic growth by improving accessibility. The construction of HSR in China provides an empirical setting to study the impact of transportation infrastructure development on urban economic growth. This paper uses data from a sample of 285 prefecture-level and above cities in China from 2003 to 2018, using the opening of HSRs as a quasi-natural experiment to conduct an empirical regression on the opening of HSRs and green total factor productivity in Chinese cities based on the DID method. The benchmark regression results show that the opening of HSR is conducive to the sustained improvement of green total factor productivity in Chinese cities. After a series of robustness and endogeneity tests, the regression coefficients of the core explanatory variables remain significantly positive and the conclusions of the benchmark regression still hold. The heterogeneity regression results show that, first, after dividing the total sample into large-scale cities and small- and medium-scale cities by city size, the coefficient estimates of the core explanatory variables are all significantly positive, and the coefficient estimates are larger for the large-scale city sample. This indicates that the opening of HSR is more conducive to promoting green total factor productivity in large-scale cities. Second, after dividing the total sample by urban areas, the coefficient estimates of the core explanatory variables are significantly positive only for the eastern and central urban samples. This indicates that the opening of HSR can play a positive contribution to green total factor productivity in cities in the east and central regions. The results of the mechanism test show that, first, the opening of HSR can positively impact urban green total factor productivity through the promotion of green finance development. This positive impact shows a trend of first increasing and then decreasing. Second, there are differences in the impact of HSR openings on green financial development at different scales and in different regions. Specifically, the opening of HSR makes a significant contribution to the development of green finance in large-scale cities, small- and medium-scale cities, and eastern and central cities, but the impact on the development of green finance in western cities is not significant. The above findings provide useful insights for understanding the changing urban economy in China, especially as the rapidly expanding HSR network plays an important role in this process. The findings of this paper can also help policymakers and micro enterprises in developing and emerging economies understand the spatial changes in the urban economy from the perspective of green financial development and maximize the benefits of HSR-induced economic growth in the post-COVID-19 pandemic period. Firstly, the government should establish a green credit guarantee system in response to the non-optimistic investment expectations caused by the pandemic. Using fiscal funds as guarantee leverage, government departments should moderately enlarge the scale of bank investment in green credit, encourage investment in cross-regional and transportation-dependent sectors, and mitigate the impact of the pandemic on cross-regional investment. Secondly, government departments should also reasonably diversify the credit risks of commercial banks in supporting the financing of external sustainable infrastructure projects through fiscal subsidies and tax incentives, thus leveraging and leading a large amount of social capital into external green projects and green industries in the post-COVID-19 pandemic period. They must also give full play to the positive effects of HSR networks on urban economic development. Developing and emerging economies should increase investment in rail networking in city clusters and metropolitan areas to increase the density of urban HSR networks. In this way, local economies can take advantage of HSR connections and benefit from involvement in a virtuous cycle by ensuring frequent and efficient HSR services. Finally, China should strengthen green financial cooperation with countries along the “Belt and Road” and improve the ability and level of financial institutions to “go global”. Financial institutions should actively integrate into sustainable investment-oriented investment and financing networks and practice responsible investment and value investment, so as to better face the uncertainty of the economic environment. The government should vigorously develop new infrastructure construction such as intercity high-speed railroads and urban railways in China, and increase support for green infrastructure and green industries in developing countries. The government should also guide firms to carry out green production capacity, technology and investment and financing cooperation with host countries and relevant market players, and give focused financial support to green technology innovation and advanced management models. Acknowledgements This work was supported by the 10.13039/501100001809 National Natural Science Foundation of China (NO. 71303105), the National Social Science Foundation (NO. 19FJYB039), the Postgraduate Research & Practice Innovation Program of Jiangsu Province (NO. KYCX20_1286), and the Nanjing University Innovation Program for PhD Candidates (NO. CXYJ21-03). 1 The four verticals include: Beijing–Shanghai high-speed railroad, connecting Beijing and Tianjin to the economically developed eastern coastal region of the Yangtze River Delta; Beijing–Hong Kong passenger special line, connecting northern and southern China; Beijing–Harbin passenger special line, connecting northeast China and the Guanzhou region; and the Hangzhou–Fuzhou–Shenzhen passenger special line (southeast coastal passenger special line), connecting the Yangtze River, the Pearl River Delta, and the southeast coastal region. The four crosses include: Shanghai–Hanrong high-speed railroad, connecting northwest and eastern China; Xulan passenger line, connecting southwest, central and eastern China; Shanghai–Kunshan high-speed railroad, connecting southwest and eastern China; and the Qingtai passenger line, connecting north and eastern China. 2 Data from the 2019 Transportation Industry Development Statistics Bulletin. 3 The specific content of the indicators and the measurement process are not shown here due to space constraints, but are kept for reference. ==== Refs References Angrist J.D. Pischke J.-S. Mostly harmless econometrics: An empiricist’s companion number 8769. Economics Books 2009 Princeton University Press Baig A.S. Blau B.M. Sabah N. Free trade and the efficiency of financial markets Global Finance Journal 48 2021 100545 Banerjee A. Duflo E. Qian N. On the road: Access to transportation infrastructure and economic growth in China 2012 National Bureau of Economic Research Banister D. Berechman Y. Transport investment and the promotion of economic growth Journal of Transport Geography 9 3 2001 209 218 Cheng H. Wang Z.Q. Peng D. Kong Q. Firm’s outward foreign direct investment and efficiency loss of factor price distortion: Evidence from Chinese firms International Review of Economics and Finance 67 2020 176 188 Diao M. Leonard D. Sing T.F. Spatial-difference-in-differences models for impact of new mass rapid transit line on private housing values Regional Science and Urban Economics 67 2017 64 77 Dong Y.M. Zhu Y.M. Can high-speed rail construction reshape the layout of China’s economic space——Based on the perspective of regional heterogeneity of employment, wage and economic growth China Industrial Economics 10 2016 92 108 Fang C.L. Yu D.L. Urban agglomeration: An evolving concept of an emerging phenomenon Landscape and Urban Planning 162 2017 126 136 He X.G. Luo Q. Chen J.L. High-quality human capital and upgrading of urban industrial structure in China: Evidence from enrollment expansion Economic Review 4 2020 3 19 Huang Z.Y. Wu L.C. Impact of Beijing-Shanghai high-speed railway on the economy of cities along the route: An empirical analysis based on the theory of space economics Macroeconomics 2 2020 165 175 Kong Q.X. Guo R. Wang Y. Sui X.P. Zhou S.M. Home-country environment and firms’ outward foreign direct investment decision: Evidence from Chinese firms Economic Modelling 85 2020 390 399 Kong Q.X. Peng D. Ni Y.H. Jiang X.Y. Wang Z.Q. Trade openness and economic growth quality of China: Empirical analysis using ARDL model Finance Research Letters 101488 2020 Kong Q.X. Peng D. Zhang R.J. Wong Z. Resource misallocation, production efficiency and outward foreign direct investment decisions of Chinese enterprises Research in International Business and Finance 55 2021 101343 Kong Q.X. Shen C.R. Sun W. Shao W. KIBS import technological complexity and manufacturing value chain upgrading from a financial constraint perspective Finance Research Letters 2020 101843 Kong Q.X. Tong X. Peng D. Wong Z. Chen H. How do factor market distortions affect OFDI: An explanation based on investment propensity and productivity effects International Review of Economics and Finance 73 2021 459 472 Lei A.C.H. Song C. Economic policy uncertainty and stock market activity: Evidence from China Global Finance Journal 2020 100581 10.1016/j.gfj.2020.100581 Li W.B. Liu F.W. Wang B. Can environmental regulation promote GTFP?——Evidence from the two control zones policy Journal of Huazhong University of Science and Technology(Social Science Edition) 33 1 2019 72 82 Li X.X. Xia G. China green financial report 2014 China Financial Naeem M.A. Nguyen T.T.H. Nepal R. Ngo Q.-T. Taghizadeh-Hesary F. Asymmetric relationship between green bonds and commodities: Evidence from extreme quantile approach Finance Research Letters 2021 101983 10.1016/j.frl.2021.101983 Ni P.F. Report of Chinese cities’ competitiveness 2008 Social Science Literature Press Pan Y.R. wen Luo L. The impact of infrastructure investment on high-quality economic development: Mechanism and heterogeneity research Reform 6 2020 100 113 Peng X.H. Wang J.Y. High-speed rail construction and green total factor productivity: Based on factor allocation distortion China Population, Resources and Environment 29 11 2019 11 19 Ran Q.Y. Zhang J.N. Yang X.D. Does high-speed railway improve the efficiency of urban green development——An empirical test based on difference in difference model Journal of Guizhou University of Finance and Economics 05 2020 100 110 Shan H.J. Reestimating the capital stock of China:1952~2006 The Journal of Quantitative & Technical Economics 25 10 2008 17 31 Sun H. Edziah B.K. Kporsu A.K. Sarkodie S.A. Taghizadeh-Hesary F. Energy efficiency: The role of technological innovation and knowledge spillover Technological Forecasting and Social Change 167 2021 120659 Sun H. Pofoura A.K. Mensah I.A. Li L. Mohsin M. The role of environmental entrepreneurship for sustainable development: Evidence from 35 countries in sub-Saharan Africa Science of the Total Environment 741 2020 140132 Sun W.H. Zhang J. High-speed rail service and urban total factor productivity Shanghai Journal of Economics 01 2021 90 104 Taghizadeh-Hesary F. Mortha A. Farabi-Asl H. Sarker T. Chapman A. Shigetomi Y. Fraser T. Role of energy finance in geothermal power development in Japan International Review of Economics and Finance 70 2020 398 412 Taghizadeh-Hesary F. Yoshino N. The way to induce private participation in green finance and investment Finance Research Letters 31 2019 98 103 Taghizadeh-Hesary F. Yoshino N. Inagaki Y. Morgan P.J. Analyzing the factors influencing the demand and supply of solar modules in Japan–does financing matter International Review of Economics and Finance 74 2021 1 12 Venables A.J. Evaluating urban transport improvements: Cost–benefit analysis in the presence of agglomeration and income taxation Journal of Transport Economics and Policy (JTEP) 41 2 2007 173 188 Wang F.R. Miao M. Tax competition, regional environment and inter-regional capital flow : An empirical research based on the perspective of inter-province M&A Economic Research Journal 50 2 2015 16 30 Wang Z.L. Lin X.Y. Gao H.W. A study on the impact of transportation infrastructure on financial agglomeration: An evidence based on railways and highways Macroeconomics 12 2020 47-61+120 Wen S.Y. Lin Z.F. Liu X.L. Green finance and economic growth quality: Construction of general equilibrium model with resource constraints and empirical test Chinese Journal of Management Science 2021 1 11 Wong Z. Li R.R. Zhang Y.D. Kong Q.X. Cai M. Financial services, spatial agglomeration, and the quality of urban economic growth–based on an empirical analysis of 268 cities in China Finance Research Letters 2021 101993 10.1016/j.frl.2021.101993 Yoshino N. Taghizadeh-Hesary F. Nakahigashi M. Modelling the social funding and spill-over tax for addressing the green energy financing gap Economic Modelling 77 2019 34 41 Zhang D. Mohsin M. Rasheed A.K. Chang Y. Taghizadeh-Hesary F. Public spending and green economic growth in BRI region: Mediating role of green finance Energy Policy 153 2021 112256 Zhang D. Vigne S.A. The causal effect on firm performance of China’s financing–pollution emission reduction policy: Firm-level evidence Journal of Environmental Management 279 2021 111609 33218832 Zhang D.Y. Du P.C. Chen Y.W. Can designed financial systems drive out highly polluting firms? An evaluation of an experimental economic policy Finance Research Letters 2019 31 Zhang D.Y. Du W.C. Zhuge L.Q. Tong Z.M. Freeman R.B. Do financial constraints curb firms’ efforts to control pollution? Evidence from Chinese manufacturing firms Journal of Cleaner Production 215 2019 1052 1058 Zhang D.Y. Guo Y.M. Financing R&D in Chinese private firms: Business associations or political connection? Economic Modelling 79 2019 247 261 Zhang D.Y. Guo Y.M. Wang Z.R. Chen Y.B. The impact of US monetary policy on Chinese enterprises’ R&D investment Finance Research Letters 35 2020 101301 10.1016/j.frl.2019.09.016 Zhang D.Y. Tong Z.M. Li Y. The role of cash holding towards cleaner production in China’s manufacturing sectors: A financial constraint perspective Journal of Cleaner Production 245 2020 118875 Zhang D.Y. Zhuge L.Q. Freeman R.B. Firm dynamics of hi-tech start-ups: Does innovation matter? China Economic Review 59 2020 101370 Zhang M.Z. Sun T. Yao P. Influence of high-speed rail opening on agglomeration efficiency of urban service industry Soft Science 33 8 2019 44 48 Zhang Z.D. Li F.Y. Infrastructure, spatial spillover and the upgrading of industrial structure: Based on the empirical analysis of prefecture-level cities in the Yangtze Economic Belt Journal of Yunnan University of Finance and Economics 35 5 2019 55 63 Zhang Z.Y. Luo X.H. Measurement of urban green development efficiency Urban Problem 2 2019 12 20 Zhao X.L. Basically useful econometrics 2017 Peking University Press Zhu X.D. Zhu S.J. Huang Y.Y. Huang H.F. How does green finance impact Chinese urban environmental pollution?——A case study of haze pollution Tropical Geography 41 01 2021 55 66 Zhu Y. Diao M. The impacts of urban mass rapid transit lines on the density and mobility of high-income households: A case study of Singapore Transport Policy 51 2016 70 80 Yin Z.B. Sun X.Q. Xing M.Y. Research on the impact of green finance development on green total factor productivity Statistics & Decision 37 03 2021 139 144	0	PMC9710515	NO-CC CODE	2022-12-02 23:21:28	no	2021 Aug 14; 49:100645	utf-8	null	null	null	oa_other
==== Front JACC Adv JACC Adv Jacc Advances 2772-963X The Authors. Published by Elsevier on behalf of the American College of Cardiology Foundation. S2772-963X(22)00220-4 10.1016/j.jacadv.2022.100143 100143 Original Research Changes in Practice/Outcomes of Pediatric/Congenital Catheterization in Response to the First Wave of COVID Quinn Brian MD a Barry Oliver M. MD b Batlivala Sarosh P. MD, MSCI c Boe Brian A. MD d Glatz Andrew C. MD, MSCE e Gauvreau Kimberlee PhD a Goldstein Bryan H. MD f Gudausky Todd M. MD g Hainstock Michael R. MD h Holzer Ralf J. MD, MSc i Nicholson George T. MD j Trucco Sara M. MD f Whiteside Wendy MD k Yeh Mary a Bergersen Lisa MD, MPH a O'Byrne Michael L. MD, MSCE lm∗ a Department of Cardiology, Boston Children’s Hospital and Department of Pediatrics Harvard Medical School, Boston, Massachusetts, USA b Division of Cardiology, Morgan Stanley Children’s Hospital of New York and Columbia University Vagelos College of Physicians and Surgeons, New York, New York, USA c The Heart Institute Cincinnati Children’s Hospital Medical Center and Department of Pediatrics, University of Cincinnati College of Medicine, Cincinnati, Ohio, USA d Division of Cardiology, Nationwide Children’s Hospital and Department of Pediatrics Ohio State University Medical School, Columbus, Ohio, USA e Division of Cardiology, Department of Pediatrics, St. Louis Children’s Hospital and Washington University School of Medicine, St. Louis, Missouri, USA f Heart Institute, UPMC Children’s Hospital of Pittsburgh and Department of Pediatrics, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA g Division of Cardiology, Children’s Wisconsin and Department of Pediatrics, Medical College of Wisconsin, Milwaukee, Wisconsin, USA h Division of Cardiology, UVA Health, and Department of Pediatrics University of Virginia School of Medicine, Charlottesville, Virginia, USA i Division of Cardiology New York-Presbyterian Hospital and Department of Pediatrics, Weill Cornell Medical School, New York, New York, USA j Division of Cardiology Monroe Carrell Jr. Children’s Hospital and Department of Pediatrics, Vanderbilt University School of Medicine, Nashville, Tennessee, USA k Division of Pediatric Cardiology, University of Michigan C.S. Mott Children’s Hospital, Ann Arbor, Michigan, USA l Division of Cardiology and Clinical Futures, The Children’s Hospital of Philadelphia and Department of Pediatrics, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA m Leonard Davis Institute and Cardiovascular Outcomes, Quality, and Evaluative Research Center, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA ∗ Address for correspondence: Dr Michael L. O'Byrne, Division of Cardiology, The Children's Hospital of Philadelphia, 3401 Civic Center Blvd, Philadelphia, Pennsylvania 19104, USA. 30 11 2022 30 11 2022 10014316 9 2022 20 9 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 The Coronavirus disease 2019 (COVID-19) pandemic has posed tremendous stress on the health care system. Its effects on pediatric/congenital catheterization program practice and performance have not been described. Objectives To evaluate how case volumes, risk-profile, and outcomes of pediatric/congenital catheterization procedures changed in response to the first wave of COVID-19 and after that wave. Methods A multicenter retrospective observational study was performed using Congenital Cardiac Catheterization Project on Outcomes Registry (C3PO) data to study changes in volume, case mix, and outcomes (high-severity adverse events [HSAEs]) during the first wave of COVID (March 1, 2020, to May 31, 2020) in comparison to the period prior to (January 1, 2019, to February 28, 2020) and after (June 1, 2020, to December 31, 2020) the first wave. Multivariable analyses adjusting for case type, hemodynamic vulnerability, and age group were performed. Hospital responses to the first wave were captured with an electronic study instrument. Results During the study period, 12,557 cases were performed at 14 C3PO hospitals (with 8% performed during the first wave of COVID and 32% in the postperiod). Center case volumes decreased from a median 32.1 cases/mo (interquartile range: 20.7-49.0) before COVID to 22 cases/mo (interquartile range: 13-31) during the first wave (P = 0.001). The proportion of cases with risk factors for HSAE increased during the first wave, specifically proportions of infants and neonates (P < 0.001) and subjects with renal insufficiency (P = 0.02), recent cardiac surgery (P < 0.001), and a higher hemodynamic vulnerability score (P = 0.02). The observed HSAE risk did not change significantly (P = 0.13). In multivariable analyses, odds of HSAE during the first wave of COVID (odds ratio: 0.75) appeared to be lower than that before COVID, but the difference was not significant (P = 0.09). Conclusion Despite increased case-mix complexity, C3PO programs maintained, if not improved, their performance in terms of HSAE. Exploratory analyses of practice changes may inform future harm-reduction efforts. Central Illustration Key words catheterization health services research outcomes research pediatric cardiology Abbreviations and Acronyms AE, adverse events C3PO, Cardiac Catheterization Project on Outcomes Registry CI, confidence interval FTR, failure to rescue HSAE, high-severity adverse events IQR, interquartile range PCCL, pediatric/congenital catheterization laboratory PREDIC3T, Procedure Risk in Congenital Cardiac Catheterization US, United States ==== Body pmcThe Coronavirus Disease 2019 (COVID-19) pandemic has posed an unprecedented stress on all aspects of health care. Although the clinical burden of the first wave of COVID-19 infections was disproportionately felt in older adults and largely spared children,1 , 2 the pandemic still affected the delivery of medical care to children with chronic medical conditions in ways that had the potential to lead to harm. In the spring of 2020, rapidly rising infection rates and early outbreaks raised the concern that scarcity of medical resources (specifically ventilators and intensive care beds) would result in excess mortality. Preemptive delay or cancelation of elective medical procedures, which would both reduce unnecessary exposure to nosocomial infection and preserve finite medical resources, was proposed as a temporizing measure. This led to both voluntary and government-mandated delays and cancelations of elective medical and surgical procedures across the United States (US). Pediatric and congenital cardiac programs faced unique challenges in this context. The incidence of congenital heart disease is a stable proportion of live births (of which a similarly stable proportion had critical congenital heart disease obligating neonatal intervention), as a result of which the demand for cardiac procedures is less elastic than in other areas of medicine.3, 4, 5 As a part of congenital heart programs, pediatric/congenital cardiac catheterization laboratories (PCCLs) confronted the complicated decision of choosing which procedures should be delayed and which should be performed, balancing the risk of delay for individual patients against the risk of iatrogenic exposure for patients, families, and staff along with the societal risk of occupying potentially necessary hospital resources in a time of scarcity. These tensions were reflected in both multicenter surveys of the immediate response of queried centers to the first wave of COVID-19,6 and a report from Morgan Stanley Children’s Hospital of New York (located in one of the hardest hit metropolitan areas in the US) described their PCCL’s response to the pandemic.7 To our knowledge, no multicenter study has, to date, described how PCCL programs changed their practice in response to the pandemic and whether these changes were accompanied by measurable changes in outcome. In response to these pressures, we anticipated 1) that case volumes would have decreased in this period, and 2) that the case-mix would be characterized by increased preprocedural risk of adverse events (AEs)—measureable increases in the proportion of patients with indicators of vulnerability (based on age, procedure-associated risk, other comorbidities, or hemodynamic condition). At the same time, we were curious if changes in practice in response could have allowed centers to maintain the quality of care provided in the face of these adverse conditions. If so, identifying practices that facilitated delivery of high-quality care during these trying times could provide meaningful benefit beyond this time period. There are 2 major obstacles to stringently studying these questions. First, a large sample from diverse centers is necessary to determine if changes in case mix and outcomes are legitimate. Second, willingness of centers to share how their practices changed is necessary to explore the changes in practice made in response to the first wave of COVID. To accomplish this, we leveraged data from the Congenital Cardiac Catheterization Project on Outcomes Registry (C3PO) to perform a retrospective cohort study evaluating both changes in practice and outcomes in the face of the first wave of COVID in the US. Methods Data source C3PO is a collaborative composed of 14 centers at the time of this study, focused on improving outcomes of pediatric/congenital cardiac catheterization through both quality improvement and facilitating clinical research. Centers contribute data from all cases to a multicenter registry that began collecting data in 2007. Each member center collects data and sends a deidentified data set using a standard electronic data-collection tool. Data management and analysis are managed by the C3PO data-coordinating center at the Boston Children’s Hospital. Data quality and reliability are assured through regular auditing of submitted data. Data sharing is governed by a series of data use agreements between C3PO and member institutions. These prohibit sharing of subject-level data. Statistical methods will be shared upon request. Analysis of deidentified data does not constitute human subjects research in accordance with the Common Rule (45 CFR 46.102(f)). Study design and population The overall goal of this study was to evaluate how practice and outcomes changed at contributing PCCL programs. This was accomplished in 3 parts. First, we sought to describe changes in number of cases and case mix in response to COVID at C3PO centers. Second, we sought to describe how outcomes changed at these same centers. Finally, as an exploratory aim, we distributed an online instrument to evaluate the specific changes that each C3PO center made in response to COVID. For the first 2 parts, we studied all cases performed at contributing centers from January 1, 2019, to December 31, 2020. These cases were divided into pre-first-wave (January 1, 2019, to February 28, 2020), first wave (March 1, 2020, to May 31, 2020), and post-first-wave (June 1, 2020, to December 31, 2020) periods. These periods were chosen to establish a stable pre-COVID baseline, against which the immediate changes in response to COVID were compared, and also to see if changes persisted beyond the first wave. There were no exclusion criteria. In the third part of the study, we evaluated the specific changes each C3PO hospital made in response to the first wave of COVID, through an electronically distributed study instrument. All active C3PO sites were eligible and invited to participate in this section. Study measures Data were directly extracted from the C3PO database. For each subject, demographics, cardiac diagnosis, and preprocedural risk factors were extracted. The primary exposure was on the day of the PCCL procedure, which were divided into 3 time periods as described above. Outcomes collected were AEs as well as unplanned admission and death at ≤72 hours of catheterization procedure. AEs are stratified in the current version of C3PO using Strata (Stratacorp) similar to those described for previous iterations,8, 9, 10, 11 specifically stratified into 5 levels of increasing severity. A change in the most recent iteration is division of level 3 events between 3a events and more severe 3bc events. For the purpose of this analysis, high-severity AEs (HSAEs) were defined as level 3bc, 4, and 5 events. Catastrophic AEs were defined as level 4 or 5 AE. Failure to rescue (FTR) was defined as a level 5 AE and/or death within 72 hours in cases in which another less-severe AE (levels 1-4) also occurred.12 Potential covariates were identified from previous studies in large registries and databases after adjusting for case-mix8 , 9 , 12, 13, 14, 15, 16 and extracted covariates were cardiothoracic surgery in the preceding 90 days, noncardiac medical conditions (coagulation disorder, renal insufficiency, and other), single ventricular vs biventricular circulation, indicators of hemodynamic vulnerability (elevated systemic ventricular end-diastolic pressure, low mixed venous saturation, low systemic arterial saturation, elevated pulmonary pressure, and elevated indexed pulmonary vascular resistance),8 preprocedural cardiac status (a novel ordinal marker of preprocedural risk developed at the Boston Children’s Hospital), and Procedure Risk in Congenital Cardiac Catheterization (PREDIC3T) case types.9 This panel of covariates includes measures (eg, PREDIC3T case type and cardiac status) that have been identified since the previous C3PO risk adjustment models and are included in the hopes of providing the most accurate depiction of risk in this cohort. Age of patients was divided into neonates (≤30 days), infants (>30 days and <1 year), children (≥1 year and <18 years), and adults (≥18 years) as described previously.8 , 12, 13, 14, 15 COVID response survey A novel study instrument was developed to formally evaluate how hospitals adjusted their practices in response to COVID. No formal focus group or field testing was performed during survey development. The instrument requested information about 5 domains: patient selection, scheduling, staffing, recovery/observation, and changes outside the PCCL. The instrument combined discrete questions (eg, “Did you restrict cases based on their perceived urgency?”) with opportunities for narrative comments, which were used to clarify responses to discrete questions. The instrument was electronically distributed to each C3PO institution using Research Electronic Data Capture tools hosted by the Boston Children’s Hospital. Electronic mail reminders were sent to encourage participation. No other incentives (financial or otherwise) were applied. Responses to survey questions were automatically anonymized, but whether a specific individual responded was known. These data enabled calculation of the response rate and permitted a limited description of the respondents (eg, number of centers and geographic spread). Statistical analysis Descriptive statistics were calculated. PCCL cases per month for each month of the study period were described across the entire collaborative and per center to provide complementary measures of procedural volume. Differences in monthly case volumes were compared between 1) pre-COVID and first wave, and 2) between pre-COVID and post-first-wave periods using the Wilcoxon signed rank test. Next, we sought to study changes in practice and outcomes in response to COVID. The primary exposure was time period. First, we evaluated whether case mix had changed over the study period. We did this by comparing the distribution of potential risk factors between the 3 preidentified time periods, using Kruskal-Wallis for continuous variables and Fisher’s exact tests for categorical variables. Second, we compared the risk of HSAE over the same time periods, expressed as likelihood with 95% confidence intervals (CIs) calculated using the exact binomial method. Finally, we evaluated whether the risk of HSAE changed after adjustment for measurable confounders, calculating multivariable logistic regression models for HSAE, with time period as the primary exposure and adjusted for prespecified covariates (PREDIC3T category, hemodynamic vulnerability score, and age category). The pre-first-wave period was the referent. As secondary analyses, we also evaluated for changes in the risk of catastrophic AE and FTR. For the novel study instrument, response rate was calculated and reported. The results were tabulated and reported using standard descriptive statistics. Missing data were limited with 2 exceptions. For PREDIC3T (>5%) and hemodynamic vulnerability scores (<1%), data were missing. Multiple imputation was not used to address missingness because there was no obvious way to predict the missing values based on other data. To avoid bias that might result from case restriction, a category named “missing” was created as described previously.12 , 17 , 18 Results Procedural volume During the study period, a total of 12,557 cases were performed at 14 hospitals. Of these, 60% (n = 7,536) were performed in the pre-first-wave period, 8% (n = 1,053) during the first wave of COVID, and 32% (n = 3,968) after the first wave (Table 1 ). Prior to the first wave of COVID, the median monthly total case volume at C3PO sites was 32.1 cases/mo (interquartile range [IQR]: 20.7-49.0). During the first wave, case volumes decreased significantly to 22.2 cases/mo (IQR: 13-31, P = 0.001) (Central Illustration A ), with the monthly procedural volume decreasing at 100% of centers. After June 2020, there was no significant difference in case volumes compared to the pre-COVID period (median: 34.8 cases/mo; IQR: 22.3-52.3; P = 0.06).Table 1 Study Population 1/2019-2/2020 (n = 7,536) 3/2020-5/2020 (n = 1,053) 6/2020-12/2020 (n = 3,968) P Value Male sex 54% (4051) 55% (577) 54% (2146) 0.80 Age at procedure (y) 3.0 (0.4, 11) 1.0 (0.2, 6) 3.0 (0.4, 12) <0.001 Age group <0.001 ≤30 d 10% (753) 17% (181) 9% (347) 31 d-1 y 25% (1,881) 32% (336) 24% (952) 1-17 y 53% (3,968) 44% (459) 54% (2,133) ≥18 y 934 (12%) 77 (7%) 536 (14%) Height (cm) (n = 7,533; 1,053; 3,968) 92 (61, 142) 73 (53, 112) 96 (62, 148) <0.001 Weight (kg) 14.0 (6.0, 39) 9.0 (4.0, 19) 14.5 (6.2, 42) <0.001 Any noncardiac problem 23% (1,734) 21% (218) 18% (719) <0.001 Chronic lung disease 6% (424) 7% (72) 5% (212) 0.18 Renal insufficiency 1% (58) 2% (17) 1% (44) 0.02 Coagulation disorder 1% (35) 0.2% (3) 1% (21) 0.65 Cardiothoracic surgery within 90 d 12% (876) 16% (173) 13% (496) <0.001 Cardiac catheterization within 90 d 12% (909) 14% (152) 12% (479) 0.09 Genetic syndrome 16% (1,207) 15% (161) 165 (619) 0.76 Single ventricle (n = 7,504; 1,050; 3,954) 29% (2,173) 30% (312) 28% (1,107) 0.42 Hemodynamic vulnerability score (n = 7,504; 1,050; 3,954) 0.02 0 50% (3,756) 49% (513) 53% (2,082) 1 20% (1,481) 18% (184) 19% (748) 2 15% (1,145) 17% (176) 15% (577) ≥3 15% (1,122) 17% (177) 14% (547) Indicators of hemodynamic vulnerability (n = 7,504; 1,050; 3,954) Elevated systemic ventricular end-diastolic pressure 5% (345) 4% (43) 5% (181) 0.79 Low mixed venous saturation 13% (974) 15% (154) 11% (422) <0.001 Low systemic arterial saturation 31% (2,313) 31% (325) 28% (1,104) 0.004 Elevated pulmonary artery pressure 18% (1,362) 21% (225) 19% (736) 0.04 Elevated ratio of pulmonary to systemic blood flow 11% (790) 10% (104) 10% (396) 0.64 Elevated indexed pulmonary vascular resistance 14% (1,072) 14% (146) 14% (539) 0.63 PREDIC3T category (performed procedure) <0.001 1 26% (1,955) 21% (225) 26% (1,032) 2 26% (1,966) 25% (264) 24% (959) 3 18% (1,324) 21% (225) 17% (682) 4 14% (1,085) 18% (188) 15% (593) 5 9% (694) 7% (72) 9% (344) Not categorized 7% (512) 8% (79) 9% (358) PREDIC3T category for anticipated procedure <0.001 0 0.1% (5) 0% (0) 0.02% (1) 1 26% (1,973) 21% (225) 25% (1,000) 2 26% (1,983) 25% (266) 25% (979) 3 17% (1,277) 21% (224) 17% (661) 4 14% (1,078) 18% (190) 15% (601) 5 10% (728) 7% (72) 9% (368) Not categorized 7% (492) 7% (76) 9% (358) Performed procedure differs from anticipated procedure 5% (414) 1% (15) 3% (117) <0.001 Preprocedural cardiac status <0.001 1 39% (2,912) 35% (373) 43% (1,701) 2 30% (2,264) 31% (329) 29% (1,159) 3 16% (1,243) 16% (172) 13% (513) Not categorized 15% (1,117) 17% (179) 15% (595) Preprocedural cardiac status (n = 6,419; 874; 3,373) <0.001 1 45% (2,912) 43% (373) 50% (1,701) 2 35% (2,264) 38% (329) 34% (1,159) 3 19% (1,243) 20% (172) 15% (513) Data are presented as % (N) or median (interquartile range). PREDIC3T = Procedure Risk in Congenital Cardiac Catheterization. Central Illustration Changes in Center Case Volume and Adjusted Risk of High Severity Adverse Events During the First Wave of COVID-19 (A) Comparison of catheterization procedure volumes between 2019 and 2020. Pediatric/congenital cardiac catheterization case volumes at C3PO centers are depicted in 2019 (blue) and 2020 (red) for the entire C3PO collaborative. Center case volumes decreased from the pre-COVID period (32.1 cases/mo; IQR: 20.7-49.0). It decreased significantly during the first wave of COVID (22.2 cases/mo; IQR: 13-31; P = 0.001). There was no significant difference between pre- and post-COVID case volumes (median: 34.8 cases/mo; IQR: 22.3-52.3; P = 0.06). (B) Multivariable model for high-severity adverse events. This forest plot depicts the main effects of multivariable model for the association between time period and odds of high severity adverse events adjusted for PREDIC3T procedure risk category, hemodynamic vulnerability score, and age category. The point estimate for odds ratio is depicted (blue diamond) along with 95% confidence intervals (brackets). Odds ratios to the left of unity (red hashed line) reflect reduced odds of high severity adverse events, while those to the right reflect higher odds. Changes in case-mix During the first wave of COVID, increases in the proportion of procedures performed in neonates and older infants (P < 0.001), cardiothoracic surgery within 90 days (P < 0.001), renal insufficiency (P = 0.02), and higher hemodynamic vulnerability scores (P = 0.02) were observed. The proportion of cases with PREDIC3T scores for the anticipated procedure in the higher-risk categories (classes 3 and 4) increased from 17% and 14% to 21% and 18%, respectively (P < 0.001). The proportion of cases with the more-severe preprocedural cardiac status (P < 0.001) also increased. Unexpectedly, the proportion of procedures in which the performed procedure differed from the anticipated one decreased from 5% to 1% (P < 0.001). The proportion of cases in which the patient had a chronic lung disease, a genetic syndrome, single ventricle physiology, or abnormal coagulation did not change significantly (all P > 0.05). Interestingly, the reported proportion of cases with noncardiac conditions decreased in both the first wave of COVID and post-first-wave periods relative to the pre-COVID period (P < 0.001). Overall, these findings are consistent with a case-mix that is associated with a higher risk of HSAE and predicts greater numbers of observed HSAE. Changes in outcome The observed proportion of cases with HSAE was 4.8% (95% CI: 4.3-5.3%) in the pre-COVID period. In the first-wave period, it was 3.9% (95% CI: 2.8-5.3%), and in the post-first-wave period, it was 4.0% (95% CI: 3.4-4.7%). Although suggestive, this difference in the observed HSAE risk was not significant (P = 0.13). The risk of secondary outcomes (all AE, catastrophic AE, and FTR) was not significantly different during the first-wave period (Table 2 ). The risk of unplanned admission decreased from 3.7% in the pre-first-wave period to 1.6% during the first wave, and it remained low in the post-first-wave period (1.8%, P < 0.001).Table 2 Outcomes 1/2019-2/2020 (n = 7,536) 3/2020-5/2020 (n = 1,053) 6/2020-12/2020 (n = 3,968) P Value Major adverse events 4.8% (358) 95% CI: 4.3%-5.3% 3.9% (n = 41) 95% CI: 2.8%-5.3% 4.0% (159) 95% CI: 3.4%-4.7% 0.13 Secondary outcomes Any adverse event 12.2% (917) 95% CI: 11.4%-12.9% 10.8% (114) 95% CI: 9.0%-12.9% 11.3% (447) 95% CI: 10.3%-12.3% 0.23 Catastrophic adverse events 1.8% (132) 95% CI: 1.5%-2.1% 1.5% (16) 95% CI: 0.9%-2.5% 1.5% (58) 95% CI: 1.1%-1.9% 0.50 Unplanned admission 3.7% (280) 95% CI: 3.3%-4.2% 1.6% (17) 95% CI: 0.9%-2.6% 1.8% (70) 95% CI: 1.4%-2.2% <0.001 Failure to rescue N = 917 2.3% (21) 95% CI: 1.4%-3.5% N = 114 0% (0) 95% CI: 0.0%-3.2% N = 447 2.2% (10) 95% CI: 1.1%-4.1% 0.30 In multivariable adjusted models, the point estimate of the odds of HSAE decreased in both first-wave COVID (OR: 0.75; 95% CI: 0.54-1.04; P = 0.09) and post-first-wave periods (OR: 0.85; 95% CI: 0.70-1.03; P = 0.10), but the difference was not statistically significant (Central Illustration B and Supplemental Table 1). A similar pattern was seen for catastrophic AE, with the point estimates suggesting an association between the first-wave and post-first-wave periods and lower odds of catastrophic AE, although it again was not statistically significant (P = 0.29, Supplemental Table 2). Because of the low event rate of FTR during the first-wave period, multivariable models were not calculated. Study instrument Responses were collected from 14 of 14 (100%) C3PO member institutions (Table 3 ). All centers (14/14) reported restricting cases based on urgency in response to COVID, with 7% (n = 1) closing their catheterization laboratory and referring all patients to an associated program. Of all C3PO programs, 57% (n = 8) restricted cases to emergent cases, and 36% (n = 5) limited cases to those that were deemed urgent. This was similar to changes in scheduling surgical cases (1 program closing completely and the remaining 13 performing only urgent cases). Changes were mandated outside of the cardiac center in 93% (n = 13) of centers; 79% (11/14) of programs received mandates from the hospital/health system leadership, 21% (3/14) from their city government, and 36% (5/14) from their state government.Table 3 Changes in Hospital Practices N = 14 Were catheterization laboratory cases restricted/halted?a Complete closure 7% (1) Only emergency cases 57% (8) Only urgent cases 36% (5) If so, who mandated these changes Hospital 79% (11) Local government 21% (3) State government 36% (5) Not mandated 7% (1) Were changes made in the process of reviewing cases prior to scheduling them? 57% (8) Were changes made to the precatheterization testing process? 93% (13) Addition of COVID testing 93% (13) Conversion to telemedicine visit 21% (3) As additional clinical review added prior to scheduling cases? 71% (10) Were there changes in the number of case slots available? More 7% (1) Same 0% (0) Less 93% (13) Were there changes in staffing of cases? Attending interventional cardiologist More 0% (0) Same 64% (9) Less 36% (5) Nurses and/or technologists More 0% (0) Same 36% (2) Less 64% (9) Anesthesiologists More 0% (0) Same 64% (9) Less 36% (5) Trainees More 0% (0) Same 43% (6) Less 57% (8) Were multidisciplinary teams cohorted to avoid exposure? 42% (6) Were any team members reassigned to other duties because of COVID? 42% (6) Were changes made to recovery location/practice? 14% (2)b Did staffing of recovery unit change? More 0% (0) Same 64% (9) Less 36% (5) Changes in surgical casesc Complete closure 7% (1) Only urgent cases 93% (13) Was cardiac bed space converted to care of COVID patients? 50% (7) If yes, did these changes affect scheduling? 71% (5/7) a Of these, all reported that cases were restricted to those that were urgent or would otherwise suffer from a delay in scheduling. b General postanesthesia care unit cohort created. c 100% mandated. Procedure scheduling changes were common. The number of case slots was reduced at 93% (13/14) of programs (1 program opened a new lab during this period, and their available slots increased). Ninety-three percent (13/14) of programs added preprocedural COVID testing for cases. No program increased the use of preprocedural testing in other (ie, noncardiac) areas, and 21% (3/14) offered telemedicine visits for preprocedural assessments. Finally, 57% (8/14) of programs added an additional review of cases prior to scheduling them. In terms of staffing, 42% (6/14) of programs experienced staff reassignments to cover COVID patients within their institution. Staffing was impacted across the board, with 64% (9/14) of programs reporting lower staffing for nurses and/or technologists, 36% (5/14) reporting reduced staffing for interventional cardiologists and anesthesiologists, and 57% (8/14) reporting reductions in trainee availability. A sizeable minority (42%, 6/14) of centers created team cohorts to reduce the risk of occupational transmission of COVID. In terms of recovery from procedures, 36% (5/14) of programs had reductions in their recovery unit staffing, and 36% (5/14) reported conversion of recovery unit beds to service COVID patients. Discussion In this multicenter cohort study from the C3PO registry, we evaluated the response of 14 US PCCL programs to the first wave of COVID. All programs reported canceling or delaying elective cases, and this has resulted in a reduction in the total number of cases attended, both across the collaborative and at individual centers. As expected, prioritizing urgent and emergent cases resulted in cases with a higher-severity case mix. At the same time, there were reductions in staffing at many centers as well as unmeasured external stressors. However, contrary to expectations, the risk of HSAE with catheterization was not significantly different in unadjusted analyses, implying that programs maintained their safety performance in spite of a higher-risk case mix. Moreover, in analyses adjusted for measurable confounders, although the association is still insignificant, the point estimates for odds of HSAE were lower in both the first-wave and post-first-wave periods. In a study with a limited sample size relative to the incidence of HSAE, these findings raise the possibility that performance was better than it was in the pre-first-wave period. This study is among the first in pediatric/congenital cardiology to utilize the unique circumstances in health care delivery introduced by COVID and an existing, audited, multicenter registry to 1) examine the impact of health care disruption on clinical outcomes, and 2) translate lessons from this unusual period to improve future care delivery. Understanding the factors underlying these changes could have potential public health impact beyond the COVID pandemic. PCCL procedures represent episodes with increased risk of harm in vulnerable patients.13 , 19 , 20 Our study period serves as a potential natural experiment during which individual programs were able to maintain if not improve their performance despite challenging circumstances inside and outside of the hospital. Although quality-improvement efforts were in place prior to the COVID pandemic, these were disrupted during this period. Each center made a series of choices that governed how they would schedule cases and conduct procedures. It is possible that the observed trends in outcome are the result of dedicated care teams exerting themselves in the face of inclement conditions, which alone would be admirable. However, the observed improvements may also be the result of adoption of practices at individual centers in response to stresses resulting from COVID and associated restrictions. Identifying what changes were beneficial is challenging but would be advantageous to ongoing (non-COVID-related) harm-reduction and quality-improvement efforts in PCCLs. We tried to determine what specific responses were employed using an electronically distributed novel instrument. Although the number of centers and variety of responses precluded quantitative evaluation of the relative contributions of each change, they do point to some practice changes that might warrant further investigation. First, as noted above, case volumes decreased. Over longer periods of time, studies have demonstrated that higher average case volumes at a center are associated with reduced likelihood of harm (specifically catastrophic AEs15 , 16 , 18 and FTR12). However, for a program with an established capacity (number of cases per year, historically), reducing the number of cases requested or performed might allow for better allocation of otherwise scarce resources (eg, staffing, equipment, and/or attention) and improved outcomes. The benefit from this acute drop in volume may be transient, as prolonged reductions in volume might result in erosion of the team experience and resources underlying associations between volume and performance. In a fee-for-service system, reducing catheterization case volumes below historical norms is unlikely to be popular. Moreover, the potential benefit may have arisen from other changes that can be implemented without affecting the PCCL capacity, and therefore, this is not sustainable. During the study period, the majority of programs reported instituting a formalized case review process prior to scheduling cases. Although this procedure was likely intended to ensure that cases were performed according to their medical priority, case review processes also provide an opportunity for rational allocation of equipment, personnel, and time. It also provides an opportunity to ensure that all levels of the care team are aware of high-risk patients and an opportunity to discuss strategies to prevent AE and/or be better prepared to rescue them. The observation that the match between expected procedure and the procedure performed was better during the first wave and that the risk of unanticipated admission was reduced during the first wave both support the idea that increased preprocedural review was successful. We acknowledge that a possible alternative is that higher-priority cases may be more goal-directed than the usual case mix. Despite this limitation, we would contend that these findings support preprocedural review, risk assessment, and planning that are part of ongoing harm-reduction work within the C3PO collaborative and that the COVID period may have accelerated progress in this area. Finally, almost all centers reported cohorting staff (specific combinations of interventionalist, anesthesiologist, nurses, and technologists) to maintain team function in the face of a communicable disease, but these efforts may have had unanticipated knock-on benefits. Formal team training programs to improve performance have been an area of active research in medicine and surgery following successes in other industries. Although not uniformly successful,21, 22, 23, 24, 25 these programs aim to improve team-wide communication, create more accurate shared mental models, and reduce authority gradients in ways that will translate into improved outcomes. Although most research has focused on formalized training programs, there is evidence that familiarity either from long-standing working relationships26 or from cohorting staff27 is both associated with improved performance. The time course of team cohorting efforts in the first wave of COVID was short, but anecdotally, cohorting teams facilitated clear communication and improved efficiencies in cases. The connection between improved communication and outcomes may appear self-evident on face, but efforts are necessary to evaluate both the implementation and durability of improvement outside of these extenuating circumstances. Subsequent waves of infections due to the Delta and Omicron COVID variants invite a postscript to this analysis. In contrast to the first wave of infections described here, subsequent waves have been characterized by much more widespread cases including a higher proportion of children, which inflict a different set of stresses on PCCL programs. High community prevalence with large volumes of asymptomatic or mild infections extending into the pediatric population has resulted in an increase in late or last-minute cancelations of scheduled cases due to close contacts or documented COVID infections. These late substitutions complicate preprocedural review and communication. Moreover, widespread community infection also has had a greater direct impact on staffing with losses due to either personal infection of team members or their care responsibilities, losses that are superimposed on staffing shortages, and turnover that has been reported across medical fields. These losses can undermine the team dynamics that may have helped maintain high-quality care in the first wave, as well as adversely affecting morale. Finally, in contrast to the first wave, there have not been widespread restrictions. If anything, previous waves have left backlogs of untreated patients and pressure to increase or maximize capacity (including asking teams to work more shifts or longer days). Cumulatively, these conditions undermine all the potential factors that we posit might have resulted in preserving the high performance of centers during the first wave. It is beyond the scope of this study, but future analyses of the subsequent waves may (regrettably) provide further evidence to support the benefit of practices we identified during this study period. Limitations In addition to those mentioned previously, there are a number of limitations that should be recognized. First, we acknowledge that we cannot determine the cause of the observed trends seen. Although a number of centers contribute data, the combination of centers and practice changes observed does not result in a sufficient variety of combinations to differentiate which changes were more or less strongly associated with the observed changes in outcome. Moreover, as noted, the short period of the first wave also results in a small sample size and a high risk of type II error. A second important issue is that the impact of COVID’s first wave on centers was extremely variable depending on their location, and it is impossible to account for this in our analysis (with significant variation in infection rates in different regions). Although every attempt was made to adjust for measurable confounders, there remains the possibility of unmeasured confounding. We acknowledge that the potential improvements in outcome are not attributable solely to catheterization laboratory teams and that they are potentially the product of the efforts of the entire care teams at each center. Information about COVID infection is not available across the cohort, so could not be reported. Finally, centers contributing to C3PO represent a fraction of the pediatric/congenital catheterization procedures in the US. In addition, C3PO centers are all actively engaged in quality-improvement efforts. For both these reasons, the observations from this sample may not be universally applicable. Conclusions While acknowledging these limitations, we conclude that during the first wave of COVID, PCCL volume decreased while these cases were characterized by a higher-risk case mix. Paradoxically, in both observed and adjusted analyses, the likelihood of a major AE was at worst similar to that in the earlier period (with some indications that the odds were lower during this period). Substantial changes in practice made in response to COVID (eg, case scheduling, reducing volume below historical capacity, preprocedural team review of cases, and cohorting staff) were observed and may guide future efforts to develop targeted interventions to improve PCCL safety.PERSPECTIVES COMPETENCY IN SYSTEMS-BASED PRACTICE: Delaying or canceling elective procedures during the first wave of COVID decreased total PCCL case volumes and increased the risk profile of cases being performed. However, programs maintained their performance in terms of major adverse events (and may have improved it). A multiinstitutional survey demonstrated that in addition to reducing case volumes, centers also implemented centralized preprocedural reviews and cohorted personnel (interventionalist cardiologists, anesthesiologists, nurses, and technologists). TRANSLATIONAL OUTLOOK: For future quality-improvement and/or research efforts, the following changes 1) reductions in case volume relative to the established capacity, 2) formal centralized preprocedural case review, and 3) cohorting staff into teams should be evaluated to determining if these changes are associated with improved outcomes outside of the pandemic setting. Funding support and author disclosures The proposed research and manuscript were reviewed by the Congenital Cardiac Catheterization Project on Outcomes (C3PO) Research and Publications Committee, but the resulting manuscript represents the views of the authors and not that of C3PO. Dr Goldstein has received consulting fees from Medtronic, W.L. Gore & Associates, Mezzion Pharmaceuticals, and PECA Labs. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Supplementary data Supplemental Tables 1 and 2 The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center. ==== Refs References 1 Shekerdemian L.S. Mahmood N.R. Wolfe K.K. Characteristics and outcomes of children with coronavirus disease 2019 (COVID-19) infection admitted to US and Canadian pediatric intensive care units JAMA Pediatr 174 2020 868 873 32392288 2 Bialek S. Gierke R. Hughes M. McNamara L.A. Pilishvili T. Skoff T. Coronavirus disease 2019 in children — United States, February 12–April 2, 2020 MMWR Morb Mortal Wkly Rep 69 2020 422 426 32271728 3 Gillum R.F. Epidemiology of congenital heart disease in the United States Am Heart J 127 1994 919 927 8154432 4 Hoffman J.I.E. Kaplan S. The incidence of congenital heart disease J Am Coll Cardiol 39 2002 1890 1900 12084585 5 van der Bom T. Zomer A.C. Zwinderman A.H. Meijboom F.J. Bouma B.J. Mulder B.J.M. The changing epidemiology of congenital heart disease Nat Rev Cardiol 8 2011 50 60 21045784 6 Morray B.H. Gordon B.M. Crystal M.A. Resource allocation and decision making for pediatric and congenital cardiac catheterization during the novel coronavirus SARS-CoV-2 (COVID-19) pandemic: a U.S. multi-institutional perspective J Invasive Cardiol 32 2020 E103 E109 32269177 7 Oshiro K.T. Turner M.E. Torres A.J. Crystal M.A. Vincent J.A. Barry O.M. Non-elective pediatric cardiac catheterization during COVID-19 pandemic: a New York center experience J Invasive Cardiol 32 2020 E178 E181 32610270 8 Bergersen L. Gauvreau K. Foerster S.R. Catheterization for congenital heart disease adjustment for risk method (CHARM) JACC Cardiovasc Interv 4 2011 1037 1046 21939947 9 Quinn B.P. Yeh M. Gauvreau K. Procedural risk in congenital cardiac catheterization (PREDIC3T) J Am Heart Assoc 11 2021 e022832 10 Bergersen L. Gauvreau K. Marshall A. Procedure-type risk categories for pediatric and congenital cardiac catheterization Circ Cardiovasc Interv 4 2011 188 194 21386090 11 Bergersen L. Marshall A. Gauvreau K. Adverse event rates in congenital cardiac catheterization - a multi-center experience Catheter Cardiovasc Interv 75 2010 389 400 19885913 12 O'Byrne M.L. Kennedy K.F. Jayaram N. Failure to rescue as an outcome metric for pediatric and congenital cardiac catheterization laboratory programs: analysis of data from the IMPACT registry J Am Heart Assoc 8 2019 e013151 13 Jayaram N. Beekman R.H. Benson L. Adjusting for risk associated with pediatric and congenital cardiac catheterization: a report from the NCDR Impact Registry Circulation 132 2015 1863 1870 26481778 14 Jayaram N. Spertus J.A. Kennedy K.F. Modeling major adverse outcomes of pediatric and adult patients with congenital heart disease undergoing cardiac catheterization: observations from the NCDR IMPACT registry (National Cardiovascular Data Registry improving pediatric and adult congenital treatment) Circulation 136 2017 2009 2019 28882885 15 Jayaram N. Spertus J.A. O'Byrne M.L. Relationship between hospital procedure volume and complications following congenital cardiac catheterization: a report from the IMproving Pediatric and Adult Congenital Treatment (IMPACT) registry Am Heart J 183 2017 118 128 27979036 16 O'Byrne M.L. Glatz A.C. Shinohara R.T. Effect of center catheterization volume on risk of catastrophic adverse event after cardiac catheterization in children Am Heart J 169 2015 823 832.e5 26027620 17 O'Byrne M.L. Glatz A.C. Song L. Association between variation in preoperative care before arterial switch operation and outcomes in patients with transposition of the great arteries Circulation 138 2018 2119 2129 30474422 18 O'Byrne M.L. Kennedy K.F. Kanter J.P. Berger J.T. Glatz A.C. Risk factors for major early adverse events related to cardiac catheterization in children and young adults with pulmonary hypertension: an analysis of data from the IMPACT (improving adult and congenital treatment) registry J Am Heart Assoc 7 2018 e008142 19 Vincent R.N. Moore J. Beekman R.H. Procedural characteristics and adverse events in diagnostic and interventional catheterisations in paediatric and adult CHD Cardiol Young 26 2016 70 78 25705856 20 Moore J.W. Vincent R.N. Beekman R.H. Procedural results and safety of common interventional procedures in congenital heart disease: initial report from the National Cardiovascular Data Registry J Am Coll Cardiol 64 2014 2439 2451 25500227 21 Weaver S.J. Dy S.M. Rosen M.A. Team-training in healthcare: a narrative synthesis of the literature BMJ Qual Saf 23 2014 359 372 22 Neily J. Mills P.D. Young-Xu Y. Association between implementation of a medical team training program and surgical mortality JAMA 304 2010 1693 1700 20959579 23 Duclos A. Peix J.L. Piriou V. Cluster randomized trial to evaluate the impact of team training on surgical outcomes Br J Surg 103 2016 1804 1814 27642053 24 Truong H. Sullivan A.M. Abu-Nuwar M.R. Operating room team training using simulation: hope or hype? Am J Surg 222 2021 1146 1153 33933207 25 Armour Forse R. Bramble J.D. McQuillan R. Team training can improve operating room performance Surgery 150 2011 771 778 22000190 26 Finnesgard E.J. Pandian T.K. Kendrick M.L. Farley D.R. Do not break up the surgical team! Familiarity and expertise affect operative time in complex surgery Am J Surg 215 2018 447 449 29174774 27 Grant M.C. Hanna A. Benson A. Dedicated operating room teams and clinical outcomes in an enhanced recovery after surgery pathway for colorectal surgery J Am Coll Surg 226 2018 267 276 29274837	36471862	PMC9710529	NO-CC CODE	2022-12-02 23:21:28	no	JACC Adv. 2022 Nov 30;:100143	utf-8	JACC Adv	2,022	10.1016/j.jacadv.2022.100143	oa_other
==== Front Arch Pediatr Arch Pediatr Archives De Pediatrie 0929-693X 1769-664X French Society of Pediatrics. Published by Elsevier Masson SAS. S0929-693X(22)00244-5 10.1016/j.arcped.2022.11.017 Short Communication c-ANCA-associated vasculitis with predominant CNS demyelination after COVID-19 Thabet F. ab⁎ Yahyaoui A. a Besbes H. ab Salem R.Hadj ab Zayani S. ab Chouchane C. ab Chouchane S. ab a Pediatric Department, Fattouma Bourguiba University Hospital, Monastir, Tunisia b Faculty of Medicine of Monastir, Tunisia ⁎ Corresponding author at: Pediatric Department, Fattouma Bourguiba University Hospital, 5000 Monastir, Tunisia. 30 11 2022 30 11 2022 1 1 2022 3 7 2022 11 11 2022 © 2022 French Society of Pediatrics. Published by Elsevier Masson SAS. All rights reserved. 2022 French Society of Pediatrics 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 infection may trigger the presentation or exacerbation of autoimmune diseases. c-Antineutrophil cytoplasmic antibody (c-ANCA)-associated vasculitis after COVID-19 mainly involves the kidneys and lungs, and is rarely reported. We describe the case of a 13-year-old girl with a history of chronic immunologic thrombocytopenic purpura who presented with transverse myelitis and central nervous system demyelination, and was subsequently diagnosed with c-ANCA-associated vasculitis following COVID-19. The patient's condition improved after pulse therapy with methylprednisolone and rituximab. To our knowledge, this is the first reported pediatric case of ANCA-associated vasculitis with predominant central nervous system involvement after COVID-19 infection. Keywords Vasculitis ANCA Myelitis COVID-19 Editor: B. Chabrol ==== Body pmc1 Introduction ANCA-associated vasculitis (AAV) is a pauci-immune small-vessel vasculitis, characterized by neutrophil-mediated vasculitis and granulomatosis [1]. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been shown to trigger the presentation or exacerbation of autoimmune diseases in genetically susceptible patients [2]. AAV has rarely been reported in patients with coronavirus disease 2019 (COVID-19). Renal and pulmonary system involvement are mostly reported in cases of AAV after COVID-19 [3,4]. The central nervous system (CNS) is involved in 15% of AAV cases, and no AAV cases have been reported after COVID-19 infection [5]. We discuss a challenging diagnosis of pediatric AAV with predominant CNS involvement. CNS symptoms in AAV may hinder early diagnosis, causing treatment delays and disease progression, which lead to relapse or even death. 2 Case report A 13-year-old girl with a history of chronic immune thrombocytopenic purpura was admitted to the pediatric department in June 2021. She presented with walking difficulties and numbness of the lower limbs that had begun 3 weeks prior. Four weeks before her presentation, she was exposed to SARS-CoV-2 through her parents, who had mild COVID-like symptoms. She experienced only mild symptoms of fatigue, myalgia, and low-grade fever. Physical examination revealed a normal Glasgow coma scale score and intact cranial nerves; however, motor strength showed an MRC grade of 4/5 throughout the right lower extremity. Upper motor neuron signs were present in the bilateral lower extremities, with Grade 3+ reflex, positive bilateral Babinski reflex, and decreased sensation in the right lower extremity. Complete blood count as well as renal and hepatic function test results were normal. Magnetic resonance imaging (MRI) of the spine showed thickening of the spinal cord, with hyperintense and diffuse edematous lesions extending from C4 to the cone, with heterogeneous contrast (Fig. 1 ). A lesion showing nodular contrast at the C5 level was also observed. The MRI of the brain revealed demyelinating lesions in the right capsulo-thalamic, right temporal (Fig. 2 ), and left cortico-subcortical paramedian hemi-cerebellar white matter, which was enhanced after gadolinium injection. MRI of the lung showed a pulmonary infiltrate (Fig. 3 ). A nasal swab polymerase chain reaction (PCR) assay for COVID-19 was negative. However, the serology test, utilizing electrochemiluminescence immunoassay, demonstrated a recent COVID-19 infection (total immunoglobulin index = 2.11 [normal < 1.0]). Cerebrospinal fluid analysis showed mild pleocytosis (10 cells/mm3 lymphocytes), high protein level (0.59 g/L), and normal glucose level (2.4 mmol/L). Bacterial cultures and a PCR test for COVID-19 were both negative. Serology test results for cytomegalovirus, human parvovirus B19, herpes simplex virus, and HIV were also negative. The initial diagnosis was acute CNS demyelination following COVID-19. Pulse steroid therapy was administered for 5 days followed by oral prednisolone. The patient presented with generalized tonic–clonic status epilepticus 2 weeks later, requiring ICU admission. MRI revealed dura mater thickening with focal and diffuse contrast enhancement, cervical and thoracic cord enlargement, swelling with diffuse hyperintensities, and cerebral demyelinating lesions. A whole-body computed tomography (CT) scan revealed bilateral pulmonary infiltrates and nodules with a halo sign, mild hepatomegaly, mild splenomegaly, and multiple retroperitoneal adenomegaly. Blood analysis showed a white blood cell count of 3.0 × 109/L and a slightly elevated C-reactive protein level. Cerebrospinal fluid analysis showed slightly elevated leukocyte and protein levels. The renal profile was normal. c-ANCA test results revealed elevated levels with a titer of >1:1000 and a proteinase-3 antibody (PR3) level of 250 (normal < 1.0). Antinuclear antibody, anti-dsDNA, aquaporin-4 antibodies (IFT HEK 293 cells technique), and MOG antibodies (cell-based assay technique), were negative, and serum complement was normal. A few days later, the patient developed diffuse dermatologic manifestations, including palpable purpura and maculopapular exanthema (Fig. 4 ). The biopsy specimen revealed features of chronic inflammatory leukocytoclastic vasculitis with nonspecific patterns. The patient was treated with pulse intravenous methylprednisolone (1 g/day) for 5 days, followed by oral prednisolone and intravenous rituximab. She underwent physical therapy and rehabilitation, resulting in gradual improvement of neurological symptoms. Repeated MRI showed regression of the demyelinating lesions. A clear reduction in pulmonary nodule size, micro- nodules, and infiltrates on chest CT (Fig. 5 ) was also observed. All laboratory markers were within normal limits. Two months after initiation of therapy, the repeated ANCA test results were negative, as were the serum MOG and aquaporin antibody results.Fig. 1 Magnetic resonance imaging of the spine: T2-weighted sequence shows diffuse thickening of the dura mater and swelling of the spinal cord. Fig 1 Fig. 2 Magnetic resonance imaging of the brain: FLAIR sequence shows demyelinating lesion. Fig 2 Fig. 3 Computed tomography of the chest revealed bilateral pulmonary infiltrates and nodules with a halo sign. Fig 3 Fig. 4 Extensive palpable purpura and maculopapular exanthema. Fig 4 Fig. 5 Repeated computed tomography of the chest shows reduction in the size of pulmonary nodules and infiltrates. Fig 5 3 Discussion Our patient presented with transverse myelitis, brain demyelination, and pachymeningitis after COVID-19. She later developed skin lesions and pulmonary involvement, and was diagnosed with AAV. CNS involvement has been observed in both AAV and COVID-19 infection [6], [7], [8]. Typically, COVID-19 neurological manifestations include stroke, encephalitis, and CNS inflammatory disorders, such as acute disseminated encephalomyelitis, vasculitis, transverse myelitis, acute hemorrhagic necrotizing encephalopathy, multiple sclerosis, Guillain–Barré syndrome, and myasthenia gravis [7]. Only a few case reports of COVID-19-related spinal cord disorders have been described in the literature, including transverse myelitis and acute disseminated encephalomyelitis [8]. In our patient, criteria suggestive of a c-ANCA-associated vasculitis included pulmonary involvement, skin involvement, very high ANCA titer, negativity of other antibodies and infectious work-up, and the improvement of symptoms on rituximab. We can also add the history of chronic autoimmune thrombocytopenia, which increase the risk of autoimmune-related diseases. The diagnosis of c-ANCA vasculitis relies on the combination of clinical findings, e.g., fever; joint pain; disease of the upper and lower respiratory tract, kidney, and other organs; and basic and nonspecific laboratory tests (inflammatory markers such as C-reactive protein level, complete blood count, renal parameters, and urine sediment analysis) as well as more specific ones, including c-ANCA testing of course and, when feasible, a biopsy of an affected organ [9]. Our patient fulfilled these criteria. Hepatosplenomegaly is a rare finding in the systemic manifestation of c-ANCA vasculitis. Isolated case reports, however, have been published mentioning hepatic and splenic involvement in c-ANCA patients [10,11]. CNS presentations in AAV patients vary and may include headache, ischemic infarction, intracranial hemorrhage, encephalopathy, and, rarely, spinal cord symptoms. These symptoms are caused by the involvement of corresponding CNS structures, such as the dura mater, brain parenchyma, spinal cord, and leptomeninges [6]. To date, only six cases of AAV after COVID-19 have been reported in the literature. All these cases involved adults, and their presentation included a combination of pulmonary and kidney injury without CNS manifestations [5]. To our knowledge, this is the first reported pediatric case of AAV following COVID-19 infection with predominant CNS involvement. Notably, the history of chronic immune thrombocytopenic purpura in this patient may increase the risk of other autoimmune diseases and suggests a possible genetic predisposition. 4 Conclusion The diagnosis of new-onset AAV with predominant CNS involvement can be challenging in COVID-19 patients because of the similarity in symptoms and clinical manifestations of both diseases. Timely diagnosis and treatment are crucial for this life-threatening disease. Funding None. Declaration of Competing Interest The authors declare that they have no competing interest ==== Refs References 1 Calatroni M. Oliva E. Gianfreda D. ANCA-associated vasculitis in childhood: recent advances Ital J Pediatr 43 2017 46 28476172 2 Liu Y. Sawalha A.H. Lu Q COVID-19 and autoimmune diseases Curr Opin Rheumatol 33 2021 155 162 33332890 3 Morris D. Patel K. Rahimi O. ANCA vasculitis: a manifestation of post-COVID-19 syndrome Respir Med Case Rep 34 2021 101549 4 Reiff D.D. Meyer C.G. Marlin B. New onset ANCA-associated vasculitis in an adolescent during an acute COVID-19 infection: a case report BMC Pediatr 21 2021 333 34353302 5 Izci Duran T. Turkmen E. Dilek M. ANCA-associated vasculitis after COVID-19 Rheumatol Int 41 2021 1523 1529 34100115 6 Zheng Y. Zhang Y. Cai M. Central nervous system involvement in ANCA-associated vasculitis: what neurologists need to know Front Neurol 9 2019 1166 30687221 7 Zamani R. Pouremamali R. Rezaei N. Central neuroinflammation in COVID-19: a systematic review of 182 cases with encephalitis, acute disseminated encephalomyelitis, and necrotizing encephalopathies Rev Neurosci 33 2021 397 412 34536341 8 Siracusa L. Cascio A. Giordano S. Neurological complications in pediatric patients with SARS-CoV-2 infection: a systematic review of the literature Ital J Pediatr 47 2021 123 34078441 9 Jennette J.C. Falk R.J. Bacon P.A. 2012 revised international Chapel Hill consensus conference nomenclature of vasculitides Arthritis Rheum 65 2013 1 11 23045170 10 Boissy C. Bernard E. Chazal M. Wegener's granulomatosis disclosed by hepato-splenic involvement Gastroenterol Clin Biol 21 1997 633 635 9587507 11 Pelechas E. Zouzos G. Voulgari P.V. An uncommon presentation of granulomatosis with polyangiitis Mediterr J Rheumatol 29 2018 49 51 32185298	36462988	PMC9710565	NO-CC CODE	2022-12-02 23:21:29	no	Arch Pediatr. 2022 Nov 30; doi: 10.1016/j.arcped.2022.11.017	utf-8	Arch Pediatr	2,022	10.1016/j.arcped.2022.11.017	oa_other
==== Front Heart Lung Circ Heart Lung Circ Heart, Lung & Circulation 1443-9506 1444-2892 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier B.V. S1443-9506(22)01144-1 10.1016/j.hlc.2022.10.014 Article Aerosol Generation During High Intensity Exercise—Implications for COVID-19 Transmission Cowie Brian MBBS, FANZCA ab∗ Wadlow Imogen BSc cd Yule Andrew MSc e Janssens Kristel MSc ai Ward Jason BSc d Foulkes Steve PhD ai Humphries Ruhi PhD d McGain Forbes PhD fg Dhillon Rana MBBS, FRACS h La Gerche André PhD aij a Sports Cardiology Laboratory, Baker Heart and Diabetes Institute, Melbourne, Vic, Australia b Department of Anaesthesia, St Vincent’s Hospital, Melbourne, Vic, Australia c Department of Atmospheric Science, University of Melbourne, Melbourne, Vic, Australia d Climate Science Centre, CSIRO Oceans and Atmosphere, Melbourne, Vic, Australia e Australian Radiation Protection and Nuclear Safety Agency, Melbourne, Vic, Australia f Department of Anaesthesia and Intensive Care, Western Health, Victoria, Australia g Department of Critical Care, University of Melbourne, Melbourne, Vic, Australia h Department of Neurosurgery, St. Vincent’s Hospital, Melbourne, Vic, Australia i Department of Cardiometabolic Health, University of Melbourne, Melbourne, Vic, Australia j National Centre for Sports Cardiology, St Vincent’s Hospital Melbourne, Melbourne, Vic, Australia ∗ Corresponding author at: Dr Brian Cowie, Department of Anaesthesia, St. Vincent’s Hospital, Melbourne, Vic, Australia 30 11 2022 30 11 2022 24 2 2022 10 10 2022 24 10 2022 © 2022 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier B.V. All rights reserved. 2022 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ) 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 and Aim COVID-19 can be transmitted through aerosolised respiratory particles. The degree to which exercise enhances aerosol production has not been previously assessed. We aimed to quantify the size and concentration of aerosol particles and evaluate the impact of physical distance and surgical mask wearing during high intensity exercise (HIE). Methods Using a prospective observational crossover study, three healthy volunteers performed high intensity cardiopulmonary exercise testing at 80% of peak capacity in repeated 5-minute bouts on a cycle ergometer. Aerosol size and concentration was measured at 35, 150 and 300 cm from the participants in an anterior and lateral direction, with and without a surgical face mask, using an Aerodynamic Particle Sizer (APS) and a Mini Wide Range Aerosol Spectrometer (MiniWRAS), with over 10,000 sample points. Results High intensity exercise generates aerosol in the 0.2–1 micrometre range. Increasing distance from the rider reduces aerosol concentrations measured by both MiniWRAS (p=0.003 for interaction) and APS (p=0.041). However, aerosol concentrations remained significantly increased above baseline measures at 300 cm from the rider. A surgical face mask reduced submicron aerosol concentrations measured anteriorly to the rider (p=0.031 for interaction) but not when measured laterally (p=0.64 for interaction). Conclusions High intensity exercise is an aerosol generating activity. Significant concentrations of aerosol particles are measurable well beyond the commonly recommended 150 cm of physical distancing. A surgical face mask reduces aerosol concentration anteriorly but not laterally to an exercising individual. Measures for safer exercise should emphasise distance and airflow and not rely solely on mask wearing. Keywords Coronavirus Covid 19 Aerosol Exercise Sport Indoor air quality ==== Body pmcIntroduction Aerosol production during high intensity exercise (HIE) has not been definitively demonstrated or quantified. This has significant implications in the current COVID-19 pandemic, in which the transmission of aerosolised particles is increasingly considered the dominant cause of viral transmission. Many nations have attempted to minimise the impact of the pandemic by entering various stages of lockdown, with closure of facilities deemed nonessential [1]. This has included major professional sporting organisations, postponement of the Olympic Games and closure of local amateur sport and gyms [[2], [3], [4]]. COVID-19 outbreaks have been reported in indoor fitness facilities, with over 20% of exposed participants of a fitness dance class becoming infected [[5], [6], [7]]. Additional outbreaks have occurred during basketball [8], indoor recreational squash and ice hockey [9]. These risks have to be balanced against the mental, physical and economic benefits of sport which are well documented [3,10]. SARS-CoV-2, the virus responsible for COVID-19, is spread by close contact, and almost certainly by airborne transmission [4,[11], [12], [13], [14], [15], [16]]. SARS-CoV-2 has a diameter of around 0.1 micrometres and viable RNA virus has been cultured from air samples exhibiting aerosol sizes ranging from submicron to several microns [12]. In a health care setting, SARS-CoV-2 has been found in aerosol in two peaks of 0.25–1.0 micrometres and >2.5 micrometres [17]. Based on concerns of viral transmission, there has been considerable research into the generation and dispersion of droplet and aerosol during normal breathing, talking, coughing, and sneezing [[13], [14], [15],[18], [19], [20]]. These droplets/aerosols range in size from 0.05 to 500 micrometres, depending on their origin [[18], [19], [20]]. Aerosols are arbitrarily defined by the World Health Organization (WHO) and the US Centers for Disease Control and Prevention (CDC) as small respirable particles <5 micrometres that remain airborne for potentially prolonged periods (hours or days). These can penetrate into alveolar spaces [21]. Droplets are arbitrarily defined as particles >5 micrometres that tend to fall towards the ground, with 10 micrometre particles falling within 20 minutes and 100 micrometre particles falling to the floor in seconds [21]. Aerosol generation during HIE research, in the setting of a global pandemic is difficult to perform. It requires sophisticated, expensive, high fidelity aerosol measurement equipment with operator expertise; a low background particle count to detect small increases in aerosol concentration; comparison between interventions such as face masks and circumferential distance from the participant; and accurate measures of minute ventilation and work rate. To date, the available data for aerosol generation during HIE has limitations [[22], [23], [24]]. Public health recommendations on the topic have extrapolated findings from other aerosol generating activities. Many nations have introduced and enforced the concept of ‘social distancing’ or physical distancing of individuals 1.5–2 metres apart, including during indoor sport [2,3]. The evidence base supporting this recommendation is limited [25]. Given coughing and sneezing can propel droplets and aerosols further than 7–8 metres, there are concerns that 1.5 metres may not be adequate during HIE if significant aerosol is generated [2,26]. In the setting of high intensity exercise, we aimed to:1. Determine if droplets and aerosol are generated. 2. Characterise the size, concentration and distribution of droplet and aerosol particles. 3. Determine the impact of a surgical mask and increasing physical distance from the participant on droplet and aerosol concentration. Materials and Methods The project was approved by the Alfred Health Human Research Ethics Committee (Project 709/20). The project was registered with the Australian and New Zealand Clinical Trials Registry (ANZCTR) (Registration number ACTRN12621000130864). Study Cohort The inclusion criteria were healthy, active adult volunteer participants, free of major medical co-morbidities, >18 years of age, who met the WHO recommendations of 150 minutes per week of moderate to vigorous exercise [10] and were available to exercise during the weekend of data collection. Three [3] participants were recruited. Study Design This was a prospective observational crossover study. Exercise and Measurements The participants underwent a cardiopulmonary exercise test (CPET) 48 hours before the experiment according to published guidelines [27]. A continuous ramp protocol was conducted on an electronically braked cycle ergometer (Lode Excalibur Sport, Lode BV Medical Technology, Groningen, NL) to obtain measures of heart rate, respiratory rate, minute ventilation, peak flow, and oxygen uptake at given power outputs and maximum workload. The same CPET protocol was suitable for all participants to achieve an optimal exercise duration between 8 and 12 minutes. Two (2) minutes passive rest preceded a minute warm-up at an initial resistance of 50 Watts. Thereafter, workload increased progressively at a rate of 30 Watts per minute until volitional fatigue. Gas exchange data was collected continuously throughout the test using a calibrated metabolic cart (Vyntus CPX, Carefusion, San Diego, CA, USA). VO2 peak was calculated as the maximum value from a 30-second rolling average of six consecutive 5 second averaged data points with the peak workload recorded. A second CPET at 80% of peak workload for a duration of 5 minutes, was performed on a different day, to replicate conditions during the experiment and to derive measurements of ventilation and respiratory rate. We opted for 80% of maximum workload for the 5-minute exercise bursts as the acceptable intensity that would be sustainable for the multiple repeat measures and allow sufficient exertional breathing consistent with working out in a gym setting. Two (2) methods to detect droplet/aerosol were utilised. An Aerodynamic Particle Sizer (APS model 3320, TSI Incorporated, Shoreview, MN, USA) spectrometer measures particles in the size 0.5–20 micrometres at 1 second intervals using the acceleration of particles between two laser beams. To extend the measurement range to smaller sizes, a Mini Wide Range Aerosol Spectrometer (MiniWRAS model 1371, GRIMM Aerosol Technik, Ainring, Germany) was employed which measures 0.01–35 micrometre particles using electrical mobility and optical light scattering at 1-minute intervals. The sampling inlet utilised during the study limited the measurable size range to an upper limit of 10 micrometres for both instruments. The sampling device was carefully positioned at the same level as the head, with subjects in the cycling position. Air was sampled at 6 litres per minute, through an inlet common to both instruments constructed of a stainless-steel funnel (100 mm opening, tapering to 12 mm over a 100 mm distance) inserted into a 2,500 mm long, 12 mm diameter conductive silicone tubing. Setting It has been well demonstrated that it is difficult to detect aerosol production in a standard indoor environment setting in which background particulate counts are high [24]. We found that measures of ambient particulate matter (‘noise’) within our exercise laboratory to be prohibitively high for detection of aerosol generation. Thus, we used a 7 x 6 x 3 m operating room (OR) at 3.3 Pascals positive pressure, at 21.5°C with a relative humidity of 40%, with 25 volume air exchanges per hour and high efficiency particulate air (HEPA) filtration to provide a ‘clean’ environment with negligible background particulate count [28,29]. This HEPA filtered air entered the OR from a large central outlet in the roof and exited via four smaller air-return vents at each corner (floor level), with the cyclist and sampling devices positioned adjacent to this flow path. Baseline particle measurements were obtained overnight within this operating room to accurately quantify background particulate counts, and this was used as the zero steady-state condition. Participants performed HIE at 80% of their peak work rate for 5 minutes. The aim was to perform sampling at 35, 150 and 300 cm directly anterior and lateral to the head position of the cyclist, with participants wearing no mask, a standard three-ply surgical face mask (Purist Australian, Sydney, NSW, Australia) and a fit checked N95 surgical face mask (BYD Care, Los Angeles, CA, USA). A single 150 cm posterior measurement was also performed. Subjects reached 80% of their peak work rate and then maintained this level for a 5-minute period, with the MiniWRAS and APS sampling continuously. The 5-minute exercise bout was used as a standardised representation of a typical HIE session that could be sustained for a reasonable period of time. It approximates settings such as in indoor cycling ‘spin class’ sessions or high intensity interval classes. Each participant commenced exercise only after background aerosol levels had returned to near zero baseline. All investigators and participants wore an N95 face mask when entering and leaving the OR between exercise bursts to minimise background contamination. The experimental set up is demonstrated in Figure 1 .Figure 1a Photographic representation of exercise participant on cycle ergometer in the operating room with the APS and MiniWRAS sampling inlet at 150 cm posterior to the rider in this example above. Abbreviations: APS, aerodynamic particle sizer; MiniWRAS, Mini Wide Range Aerosol Spectrometer. Figure 1b: Schematic showing the sampling points (marked as ‘x’) at 30 cm, 150 cm and 300 cm anteriorly, laterally and posteriorly from a top down view (not to scale). Participants wore an N95 mask between exercise bursts as background aerosol concentration returned to near zero baseline. Analysis Aerosol size distributions are predominantly lognormal and are presented using normalised concentration in aerosol literature as dN/dlogDp, where dN is the particle concentration and dlogDp is the log of the midpoint particle diameter [30]. Total concentrations from instruments are reported as particles per cm3 of air. For each subject in each setting (varied distance with and without a mask) sampling was acquired continuously, and data was selected after an equilibrium was reached, resulting in an average of 245 and 5 sample points per subject per setting, for the APS and MiniWRAS, respectively. The average size distribution and total concentration of the background period was subtracted from each measurement resulting in just the perturbation above background. The average of the resulting samples was used to provide a single value for each rider in each experimental setting. Repeated measured factorial analysis of variance (ANOVA) was used with distance and mask wearing considered as within subject variables. Thus, the significance of an interaction for distance (i.e. was there a significant impact of distance from the subject on aerosol concentration?) and mask wearing (i.e. did the wearing of a mask significantly impact on aerosol concentration?). Measures were obtained anterior and lateral to the subject. Thus, in addition, an interaction between direction of sampling and mask wearing was considered (i.e. was the effect of mask wearing on aerosol concentration different in an anterior vs lateral position?). Results Demographics and CPET Results Demographics and the summary of the baseline and 5-minute high intensity burst CPET results of the three participants are displayed in Table 1 .Table 1 Participant demographics, cardiopulmonary exercise test and 5-minute steady state exercise measures. Participant 1 2 3 Age (years) 41 49 28 Sex Female Male Male Height (cm) 171 191 173 Weight (kg) 61 87 55 Body Mass Index (kg/m2) 21 24 18 CPET (cardiopulmonary exercise test) Peak workload (W) 280 330 360 Peak heart rate (beats/min) 160 167 194 Peak respiratory rate (breaths/min) 54 63 69 Peak oxygen consumption - VO2 (mL/kg/min) 48 40 72 Peak minute ventilation (L/min) 113 163 164 Peak flow rate (L/min) 282 378 420 5 MIN EFFORT Average workload (W) 210 260 273 Peak heart rate (beats/min) 156 155 188 Peak respiratory rate (breaths/min) 43 60 54 Peak oxygen consumption - VO2 (mL/kg/min) 50 41 69 Minute ventilation (L/min) 104 147 139 Peak flow rate (L/sec) 4.3 5.9 6.7 Average heart rate (beats/min) 150 148 179 Average respiratory rate (breaths/min) 37 51 46 Average oxygen consumption - VO2 (mL/kg/min) 45 36 62 Average minute ventilation (L/min) 84 124 114 A baseline was established in the OR during two overnight sampling periods, during which the OR was closed with the particle count very close to zero (<0.02 cm-3 identified using the APS). All three participants were able to perform the 5-minute burst of high intensity exercise with and without a surgical face mask at the planned work rate. None of the three riders could achieve the target workload using an N95 mask. Given that none of the three fit and healthy cyclists could exercise beyond mild to moderate intensity exercise wearing an N95 mask, this line of inquiry in the experiments was considered futile and subsequent planned bouts of exercise using an N95 mask were abandoned. In total, each participant performed 14 separate HIE bursts over 2 days, seven without a face mask, six with standard surgical mask and one with an N95 mask. As depicted in Figure 2a, the three participants produced a statistically similar quantity and distribution of aerosols. During 5 minutes of high intensity with sampling at 35 cm anterior to the face, the highest concentration of aerosol generation occurred within the 0.2–1.0 micrometre range. There were almost no particles > 5 micrometres (Figure 2a and Table 2 ). The pattern of aerosol generation between the three riders was consistent in all conditions and positions and are subsequently presented as a single mean value.Figure 2a Aerosol generation by each rider. Distribution of aerosol measured by the MiniWRAS with sampling anterior to the rider exercising without a face mask. The shaded area indicates the standard deviation of the average of the three riders. Abbreviations: MiniWRAS, Mini Wide Range Aerosol Spectrometer. Figure 2b: Impact of distance on aerosol concentration. Average of three riders measured in front of the rider with no mask at increasing distances of 35 cm, 150 cm and 300 cm. The mean aerosol size distribution during exercise is shown on the left, with the mean and standard deviation of the total concentration of each case shown in panels on the right. Statistically significant reductions are present with increasing distance. Abbreviations: APS, aerodynamic particle sizer. Table 2 Average particle concentration per cm3 generated with exercise sampling at different positions, with and without a face mask. Surgical Mask 35 cm 150 cm 300 cm Distance Interaction P-value Mask Interaction P-value Anterior MiniWRAS No 2.29±0.34 1.49±0.56 0.76±0.29∗# 0.003 0.031 Yes 1.41±0.43 0.66±0.18 0.34±0.11∗ APS No 0.084±0.01 0.070±0.008 0.068±0.005 0.041 0.32 Yes 0.078±0.001 0.069±0.005 0.061±0.002∗ Lateral MiniWRAS No 1.37±0.59 1.04±0.38 0.36±0.30# 0.15 0.64 Yes 0.72±0.38 0.88±0.45 0.68±0.17 APS No 0.074±0.008 0.071±0.006 0.064±0.005 0.15 0.58 Yes 0.071±0.002 0.068±0.001 0.063±0.003# Values in bold are Statistically significant. ∗ P<0.05 for comparison with 35 cm. # P<0.05 for comparison with 150 cm. Abbreviations: APS, aerodynamic particle sizer; MiniWRAS, Mini Wide Range Aerosol Spectrometer. Exercise resulted in the generation of aerosol particles that were measured almost exclusively in the range 0.25 – 1 micrometres (Figure 2). Impact of Distance on Aerosol Concentration The impact of increasing distance on droplet/aerosol count from the exercise participant was evaluated at 35, 150 and 300 cm from the rider, with results shown in Table 2 and Figure 2b. With sampling anterior to the rider, in direct line with expiration, there was a 35% and 65% reduction, respectively, in MiniWRAS determined aerosol counts at 150 cm and 300 cm relative to 35 cm (p=0.003 for distance interaction). Using APS aerosol quantification, there was an 18% and 20% reduction, respectively (p=0.041 for the distance interaction). With lateral sampling, there was a 24% and 74% reduction in aerosols at 150 and 300 cm, respectively (p=0.037). The 4% and 15% reduction in APS values was not significant (p=0.15, Table 2 and Figure 2b) reflecting the lower concentrations of larger particle aerosols at all distances. Impact of Mask Wearing on Aerosol Concentration With a standard surgical mask there was a 38%, 56% and 58% reduction in submicron aerosol counts measured with the MiniWRAS at 35 cm, 150 cm and 300 cm, respectively, when measured in front of the rider (p=0.031 see Table 2 and Figure 3 ). There were no significant reductions associated with mask wearing in the larger aerosols measured with the APS (−8% at 35 cm, −1% at 150 cm and −9% at 300 cm, p=0.32).Figure 3 Average aerosol concentrations with anterior sampling at a distance of 35 cm with different masks. The mean aerosol size distribution during exercise is shown on the left, with the mean and standard deviation of the total concentration of each case shown in panels on the right. There were statistically significant reductions in aerosol concentration with both masks for submicron aerosols (See Supplementary Table 1). Participants were unable to reach target work rate wearing an N95 mask. Abbreviations: APS, aerodynamic particle sizer; MiniWRAS, Mini Wide Range Aerosol Spectrometer. The potential for further improvements in aerosol counts was assessed comparing a standard surgical mask as compared with a tightly fitting N95 grade mask at 35 cm from the riders with an anterior sampling position. The N95 did reduce MiniWRAS aerosol counts more than a standard surgical mask as compared with no mask (58% vs 38%, p=0.031, Figure 3 and Supplementary Table 1). Impact of Mask Wearing and Distance on Aerosol Concentration The impact of mask wearing combined with physical distancing revealed statistically significant reductions in aerosol concentration (Figure 4 , Table 2). For the submicron aerosols measured by the MiniWRAS, both increasing distance and mask wearing reduced aerosol concentrations but the reductions with distance were greater when a mask was not worn (interaction for mask and distance p=0.035, see Supplementary Figure 1). This was due to the concentrations being much higher at closer distances without a mask and convergence of values at 300 cm regardless of whether the subject was wearing a mask. When measured using the APS, the small reductions in large aerosol particles with increasing distance occurred to a similar extent regardless of whether the subject was wearing a mask (p=0.85 for distance by mask interaction; Supplementary Figure 1b).Figure 4 Average aerosol concentrations when wearing a surgical mask measured in front of the rider at increasing distances. The mean aerosol size distribution during exercise is shown on the left, with the mean and standard deviation of the total concentration of each case shown in panels on the right. Abbreviations: APS, aerodynamic particle sizer; MiniWRAS, Mini Wide Range Aerosol Spectrometer. Impact of Angle on Aerosol Concentration The influence of angle on aerosol concentration was evaluated by comparing sampling in the anterior position, and in a 90° lateral position (Figure 5 ). There was a significant reduction in aerosol concentration when measured in the lateral position relative to anterior using MiniWRAS (p=0.029). (Supplementary Figure 2a). The effect of position on aerosols of larger particle size measured with APS was similar but did not reach statistical significance (p=0.10, Supplementary Figure 2b).Figure 5 Average aerosol concentrations obtained 150 cm away from the rider at varying angles with no mask. The mean aerosol size distribution during exercise is shown on the left, with the mean and standard deviation of the total concentration of each case shown in panels on the right. Abbreviations: APS, aerodynamic particle sizer; MiniWRAS, Mini Wide Range Aerosol Spectrometer. Impact of Angle and Mask on Aerosol Concentration In contrast to anterior position, mask wearing did not alter small or larger aerosol counts as compared to no mask when sampling was lateral to the rider (p=0.64 and p=0.58 for miniWRAS and APS respectively, Table 2 and Figure 6 ).Figure 6 Average aerosol concentrations in anterior and lateral positions at 300 cm away with riders wearing a surgical mask. Mean aerosol size distribution during exercise is shown on the left, with the mean and standard deviation of the total concentration of each case shown in panels on the right. Abbreviations: APS, aerodynamic particle sizer; MiniWRAS, Mini Wide Range Aerosol Spectrometer. Discussion Using state-of-the-art measures of aerosol concentration in a unique experimental setting with trivial background particle confounding, we demonstrated that HIE generates a significant number of expired aerosols, predominantly in the 0.2–1.0 micrometre range. Aerosol concentration was attenuated with the use of a surgical mask when sampled in front of the rider. However, the use of a surgical mask did not reduce aerosol concentration when measured lateral to the rider. Aerosol concentrations were reduced further with the use of an N95 surgical mask but high intensity exercise was not able to be sustained by any of the riders and was deemed incompatible with exercise beyond mild exertion. Our methodology was enhanced by measuring particle concentration by two techniques enabling an assessment over a large range of particle size. Although there was some overlap in the range of detection, the greatest number concentrations of particulate at all distances were measured in the 0.25–0.75 micrometre range that was exclusively measured with the MiniWRAS. Thus, the MiniWRAS results provided greater sensitivity than APS for detecting changes due to distance, mask wearing and position of sampling relative to the rider. Despite this, the two techniques were generally complementary with the direction and nature of effect proving similar in most settings. These findings should inform the design of future experiments aiming to quantify aerosol generation during increased respiratory ventilation. It is noted that both these devices measure number, rather than mass, concentration. Aerosol microphysics dictates that number concentrations are dominated by smaller aerosols, whereas most of the mass is present in the larger aerosols. This has important implications for pathogen spread via aerosols where the ability of an aerosol to transmit a disease depends on where in the respiratory tract aerosols can penetrate (smaller aerosols making it ‘deeper’), and the total viral load reaching the patient (the larger aerosols with more mass are likely to carry more virus). High background particle concentrations in our daily environment make detection of relatively small concentrations of aerosol challenging [24]. We achieved a clean background signal through the use of an ultralow baseline particle operating room. A standard gym has in the order of 800 particles per litre background with a size spectrum from 0.3–10 micrometres [24] and typically 6–10 air changes per hour [2]. The high baseline particle count of such facilities explains the difficulty detecting aerosol in this environment [24]. Previous data has demonstrated that glottic closure manoeuvres such as coughing and sneezing generate somewhere between 1,000−10,000 particles per episode [19]. Talking generates between 60–3,000 particles per litre [2]. At first glance, it would appear that high intensity exercise produces fewer particles. However, based on the typical 45-minute duration of an indoor cycling class, a single participant would be expected to generate a total aerosol count comparable to that of a glottic-closure manoeuvre. Furthermore, with participants in such an indoor cycling class having a minute volume in the order of 100–150 litres per minute, it is likely that an individual will inhale large numbers of small particles generated by neighbouring participants in their immediate vicinity. Because the aerosol generated during HIE is in the submicron range, it is likely to remain suspended in the air for extended periods and thus be at an even higher risk of being inhaled. Lastly, exercise often stimulates upper airway irritation resulting in an increased rate of coughing and upper airway clearance that would compound aerosol concentration. Respiratory droplets and aerosol can be generated from distal respiratory bronchioles and alveoli, larger airways, periglottic/laryngeal structures, and the oropharynx [14,19]. In general, smaller submicron aerosol is generated in terminal airways via a mechanism that relates to closure then reopening of terminal respiratory bronchioles. Terminal bronchioles collapse at smaller lung volumes and then reopen with inspiration, forming a plug that breaks up and generates small particles. These particles contain phospholipids and proteins commonly found in surfactant. This is particularly prominent with deeper respiration [19,20]. Forced exhalation with high velocity flow leaving large central airways generates larger particles in the 0.5–5 micrometre range. These have a different size and contain little to no surfactant material, supporting the different origin of the particles [19]. They are not prominent during quiet respiration and are usually deposited in the larger airways and larynx at rest. They may be more likely to appear with the high velocity short expiratory time typical of HIE [19]. As might be expected, we demonstrated a reduction in aerosol concentrations with greater distance from the source. However, we observed significant concentrations of aerosols at the 150 cm of physical distancing that is currently recommended in many Western countries [2,3]. There were further reductions at 300 cm, but even at this distance there were measurable concentrations of aerosol. Our findings suggest that we should not consider 150 cm to be a safe distance from an exercising individual in terms of aerosol exposure and should maintain as great a distance as possible, especially when indoors. The benefits of wearing a surgical mask have been widely debated during the COVID-19 pandemic [31,32]. Our data suggests that aerosol concentration reduces both with distance and with the use of a surgical mask. However, aerosol concentration with mask wearing is complex in that it clearly reduces aerosol concentration in the direction of exhalation (anteriorly) but not the concentration of aerosol particles when measured lateral to the rider at both 150 and 300 cm using either sampling technique. It is possible that wearing of a surgical mask may direct the plume of aerosol laterally from the face. The increased resistance of air flow imposed by the mask could lead to gas and aerosol favouring directions of least resistance, such as gaps between the face and the side of the mask. The greater the minute volume and flow rate of expired air, the more aerosol could be directed in this manner. This could be an issue in indoor fitness facilities where exercise apparatus is often organised in a line with participants less than three metres lateral to one another. These findings have the potential to inform public policy in that the use of masks appear to reduce overall aerosol concentration but the efficacy of this intervention is far from absolute. The effect on aerosol concentration of wearing a mask appears similar to that resulting from distancing by an additional 1.5 metres. Furthermore, the effect of wearing a mask has lesser impact on exposure to aerosols lateral to the athlete. Therefore, the effect of mask wearing must be considered as only part of the strategy to reduce aerosol exposure in exercise environments. All participants were able to perform HIE with the surgical mask with subjective discomfort but with no objective deterioration in the work rate achieved. Other recent studies have also suggested that standard surgical face masks have minimal impact on performance during HIE [[33], [34], [35]]. In part, this is probably consistent with the idea that inhaled and exhaled air finds pathways around the mask. On the other hand, our participants were unable to perform HIE wearing an N95 mask. An N95 mask is associated with a significant increase in the work of breathing [36] and an 8 mmHg increase in end tidal carbon dioxide compared with no mask or a standard surgical mask, even in young adults [34]. This likely explains the inability to exercise at the same intensity wearing N95 masks in our cohort. The strengths of this study include the collaborative approach between clinicians, exercise physiologists and experts in atmospheric aerosols, using high fidelity aerosol detection equipment. The participants are typical of those found in an indoor fitness facility that exercise regularly. Formal cardiopulmonary testing adds additional value to quantify the volume of gas expiration occurring during high intensity exercise. Using an extremely low baseline aerosol concentration in an operating room enabled detection of a small signal. Our study examined aerosol generation in individual participants. It is likely that multiple participants performing simultaneous HIE would generate a higher total aerosol concentration. We used cycling as our model for exercise as cycling is widely used in gyms and in clinical exercise testing settings. It is possible that other forms of exercise and other equipment apparatus generate a different aerosol profile. The APS was unable to detect aerosol particles less than 0.78 micrometres in size. This likely explained the better sensitivity of the MiniWRAS to changes in aerosol concentrations with increasing distances and angle of the sampling devices relative to the rider. The techniques are complementary, detecting aerosol over a range of particle sizes. Patient susceptibility to SARS-CoV-2 depends on a range of factors including viral load, duration of exposure and underlying comorbidities [25]. The minimum infectious ‘dose’ is currently unknown, and this study did not assess the infectivity of particles generated. The size distribution of aerosol generated during exercise make viral transmission plausible. The safe reopening of indoor fitness facilities will likely require a combination of improvements in airflow, ventilation, in conjunction with physical distancing, screening of high-risk individuals and widespread vaccination. Conclusion High intensity exercise is an aerosol generating procedure with particles predominantly detected in the 0.2–1.0 micrometre range. Aerosol concentration is reduced with increasing distance in all directions but even at 300 cm from the source, aerosol concentrations are measurable. The commonly recommended ‘1.5 m rule’ would not safely protect other users from aerosol exposure in an indoor setting. The use of standard surgical masks results in modest reductions in aerosol concentration anteriorly but no reductions in concentration laterally. These findings have direct implications for public health advice as summarised in Table 3 . If we were to extrapolate from aerosol concentration to risk of infection with a virus such as COVID-19, we may conclude that distances between exercising individuals need to be maximised and that masks provide an additional means of reducing aerosol exposure. However, the relatively modest protection from masks was only observed only in the direction of expiration of the exercising subject. Indoor exercise activities should not rely on masks alone for exposure minimisation but also emphasise distancing and ventilation.Table 3 Summary of findings and implications for public health policy and practice New findings. • High intensity exercise generates aerosol predominantly in the 0.2–1 micrometre range. • Increasing physical distance from the exercise participant reduces, but does not completely eliminate, exposure to aerosol. • A surgical face mask worn during exercise reduces aerosol concentration measured anteriorly but not when measured laterally to an exercising subject. This may be due to high velocity jets between the face and mask during the higher minute volume and peak flow generated with exercise. Impact on practice• High intensity exercise in an indoor setting is likely to result in aerosol exposure to nearby participants. • The commonly recommended 1.5 m distance rule reduces, but does not eliminate, aerosol exposure. • Mask wearing has a relatively modest impact on aerosol spread that is only appreciable when measured in the direction of expiration. It should not be relied upon as the sole means of protection from airborne infections during exercise. Contributors BC – study concept, design, data acquisition, data analysis and manuscript preparation AY, IW, JW and RH — study design, data acquisition, data analysis and manuscript preparation. KJ and SF — study design, data acquisition and manuscript preparation. FM and RD — study design and manuscript preparation. ALG — study design, data analysis and manuscript preparation. All authors take responsibility for the integrity and accuracy of the data analyses. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted. Funding This study was funded by the 10.13039/501100014643 Baker Heart and Diabetes Institute Sports Cardiology Lab using internal funds. The manuscript has been written by the experienced clinician researchers involved with no professional medical writers. Competing Interests None of the authors have any competing interests or financial relationships that influenced this work. Transparency Statement The lead author affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted. Dissemination to Participants Participants will receive a full copy of the peer reviewed manuscript. Data Sharing The authors commit to making the relevant anonymised patient data available on reasonable request. Appendices Supplementary Data Supplementary Figure 1 Supplementary Figure 2 Supplementary Table 1 Appendices Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.hlc.2022.10.014 Trial registration - The project was registered with the Australian and New Zealand Clinical Trials Registry (ANZCTR) (Registration number ACTRN12621000130864). ==== Refs References 1 Cyranoski D. How to stop restaurants from driving COVID infections Nature 587 7834 2020 344 33173217 2 Blocken B. van Druenen T. van Hooff T. Verstappen P.A. Marchal T. Marr L.C. Can indoor sports centers be allowed to re-open during the COVID-19 pandemic based on a certificate of equivalence? Build Environ 180 2020 107022 3 Hughes D. Saw R. Perera N.K.P. Mooney M. Wallett A. Cooke J. The Australian Institute of Sport framework for rebooting sport in a COVID-19 environment J Sci Med Sport 23 7 2020 639 663 32451268 4 Sparrow A.K. Brosseau L.M. Harrison R.J. Osterholm M.T. Protecting Olympic participants from Covid-19 - the urgent need for a risk-management approach N Engl J Med 385 1 2021 e2 34033274 5 Jang S. Han S.H. Rhee J.Y. Cluster of coronavirus disease associated with fitness dance classes, South Korea Emerg Infect Dis 26 8 2020 1917 1920 32412896 6 Blocken B. van Druenen T. Ricci A. Kang L. van Hooff T. Qin P. Ventilation and air cleaning to limit aerosol particle concentrations in a gym during the COVID-19 pandemic Build Environ 193 2021 107659 7 Bae S. Kim H. Jung T.Y. Lim J.A. Jo D.H. Kang G.S. Epidemiological characteristics of COVID-19 outbreak at fitness centers in Cheonan, Korea J Korean Med Sci 35 31 2020 e288 32776726 8 Costello B.T. Climie R.E. Wright L. Janssens K. Mitchell A. Wallace I. Athletes with mild COVID-19 illness demonstrate subtle imaging abnormalities without exercise impairment or arrhythmias Eur J Prev Cardiol 29 6 2022 e220 e223 34669943 9 Atrubin D. Wiese M. Bohinc B. An outbreak of COVID-19 associated with a recreational hockey game - Florida, June 2020 MMWR Morb Mortal Wkly Rep 69 41 2020 1492 1493 33056952 10 Polero P. Rebollo-Seco C. Adsuar J.C. Pérez-Gómez J. Rojo-Ramos J. Manzano-Redondo F. Physical activity recommendations during COVID-19: narrative review Int J Environ Res Public Health 18 1 2020 11 Noorimotlagh Z. Jaafarzadeh N. Martínez S.S. Mirzaee S.A. A systematic review of possible airborne transmission of the COVID-19 virus (SARS-CoV-2) in the indoor air environment Environ Res 193 2020 110612 12 Comber L. O Murchu E. Drummond L. Carty P.G. Walsh K.A. De Gascun C.F. Airborne transmission of SARS-CoV-2 via aerosols Rev Med Virol 2020 e2184 13 Asadi S. Bouvier N. Wexler A.S. Ristenpart W.D. The coronavirus pandemic and aerosols: does COVID-19 transmit via expiratory particles? Aerosol Sci Technol 0 0 2020 1 4 32308568 14 Borak J. Airborne transmission of COVID-19 Occup Med (Lond) 70 5 2020 297 299 32476011 15 Jarvis M.C. Aerosol transmission of SARS-CoV-2: physical principles and implications Front Public Health 8 2020 590041 16 Morawska L. Cao J. Airborne transmission of SARS-CoV-2: the world should face the reality Environ Int 139 2020 105730 17 Liu Y. Ning Z. Chen Y. Guo M. Gali N.K. Sun L. Aerodynamic analysis of SARS-CoV-2 in two Wuhan hospitals Nature 582 7813 2020 557 560 32340022 18 Anand S. Mayya Y.S. Size distribution of virus laden droplets from expiratory ejecta of infected subjects Sci Rep 10 1 2020 21174 19 Bake B. Larsson P. Ljungkvist G. Ljungström E. Olin A.C. Exhaled particles and small airways Respir Res 20 1 2019 8 30634967 20 Holmgren H. Ljungstrom E. Almstrand A.-C. Bake B. Olin A.-C. Size distribution of exhaled particles in the range from 0.01 to 2.0 mm J Aerosol Sci 41 2010 439 446 21 Anderson E.L. Turnham P. Griffin J.R. Clarke C.C. Consideration of the aerosol transmission for COVID-19 and public health Risk Anal 40 5 2020 902 907 32356927 22 Varga C.M. Kwiatkowski K.J. Pedro M.J. Groepenhoff H. Rose E.A. Gray C. Observation of aerosol generation by human subjects during cardiopulmonary exercise testing using a high-powered laser technique: a pilot project J Med Biol Eng 42 1 2022 1 10 35095378 23 Sajgalik P. Garzona-Navas A. Csécs I. Askew J.W. Lopez-Jimenez F. Niven A.S. Characterization of aerosol generation during various intensities of exercise Chest 160 4 2021 1377 1387 33957100 24 Helgeson S.A. Lee A.S. Patel N.M. Taylor B.J. Lim K.G. Niven A.S. Cardiopulmonary exercise and the risk of aerosol generation while wearing a surgical mask Chest 159 4 2021 1567 1569 32956718 25 Jones N.R. Qureshi Z.U. Temple R.J. Larwood J.P.J. Greenhalgh T. Bourouiba L. Two metres or one: what is the evidence for physical distancing in covid-19? BMJ 370 2020 m3223 32843355 26 Fennelly K.P. Particle sizes of infectious aerosols: implications for infection control Lancet Respir Med 8 9 2020 914 924 32717211 27 Balady G.J. Arena R. Sietsema K. Myers J. Coke L. Fletcher G.F. Clinician's guide to cardiopulmonary exercise testing in adults: a scientific statement from the American Heart Association Circulation 122 2 2010 191 225 20585013 28 Dhillon R.S. Rowin W.A. Humphries R.S. Kevin K. Ward J.D. Phan T.D. Aerosolisation during tracheal intubation and extubation in an operating theatre setting Anaesthesia 76 2 2021 182 188 33047327 29 Scott D.A. Humphries R.S. Dhillon R.S. Confirming estimates of aerosol clearance time Anaesthesia 76 Suppl 3 2021 22 23 33368167 30 Finlay W.H. Darquenne C. Particle size distributions J Aerosol Med Pulm Drug Deliv 33 4 2020 178 180 32598205 31 Cowling B.J. Leung G.M. Face masks and COVID-19: don't let perfect be the enemy of good Euro Surveill 25 49 2020 2001998 32 Matuschek C. Moll F. Fangerau H. Fischer J.C. Zänker K. van Griensven M. Face masks: benefits and risks during the COVID-19 crisis Eur J Med Res 25 1 2020 32 32787926 33 Shaw K. Butcher S. Ko J. Zello G.A. Chilibeck P.D. Wearing of cloth or disposable surgical face masks has no effect on vigorous exercise performance in healthy individuals Int J Environ Res Public Health 17 21 2020 8110 33153145 34 Epstein D. Korytny A. Isenberg Y. Marcusohn E. Zukermann R. Bishop B. Return to training in the COVID-19 era: the physiological effects of face masks during exercise Scand J Med Sci Sports 31 1 2021 70 75 32969531 35 Mapelli M. Salvioni E. De Martino F. Mattavelli I. Gugliandolo P. Vignati C. “You can leave your mask on”: effects on cardiopulmonary parameters of different airway protective masks at rest and during maximal exercise Eur Respir J 58 3 2021 2004473 36 Lee H.P. Wang d.Y. Objective assessment of increase in breathing resistance of N95 respirators on human subjects Ann Occup Hyg 55 8 2011 917 921 21893677	36463077	PMC9710566	NO-CC CODE	2022-12-12 23:20:59	no	Heart Lung Circ. 2022 Nov 30; doi: 10.1016/j.hlc.2022.10.014	utf-8	Heart Lung Circ	2,022	10.1016/j.hlc.2022.10.014	oa_other
==== Front J Clean Prod J Clean Prod Journal of Cleaner Production 0959-6526 1879-1786 Elsevier Ltd. S0959-6526(21)01923-5 10.1016/j.jclepro.2021.127705 127705 Article Impact of COVID-19 pandemic on socio-economic, energy-environment and transport sector globally and sustainable development goal (SDG) Nundy Srijita a Ghosh Aritra b∗ Mesloub Abdelhakim c Albaqawy Ghazy Abdullah c Alnaim Mohammed Mashary c a School of Advanced Materials Science and Engineering, Sungkyunkwan University, Suwon, 16419, Republic of Korea b College of Engineering, Mathematics and Physical Sciences, Renewable Energy, University of Exeter, Cornwall, TR10 9FE, UK c Department of Architectural Engineering, Ha'il University, Ha'il, 2440, Saudi Arabia ∗ Corresponding author. 31 5 2021 20 8 2021 31 5 2021 312 127705127705 1 9 2020 22 5 2021 25 5 2021 © 2021 Elsevier Ltd. All rights reserved. 2021 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 United Nation's Sustainable Development Goals (SDGs) want to have a peaceful world where human life will be in a safe, healthy, sustainable environment without any inequalities. However, the year 2020 experienced a global pandemic due to COVID-19. This COVID-19 created an adverse impact on human life, economic, environment, and energy and transport sector compared to the pre-COVID-19 scenario. These above-mentioned sectors are interrelated and thus lockdown strategy and stay at home rules to reduce the COVID-19 transmission had a drastic effect on them. With lockdown, all industry and transport sectors were closed, energy demand reduced greatly but the time shift of energy demand had a critical impact on grid and energy generation. Decreased energy demand caused a silver lining with an improved environment. However, drowned economy creating a negative impact on the human mind and financial condition, which at times led to life-ending decisions. Transport sector which faced a financial dip last year trying to coming out from the losses which are not feasible without government aid and a new customer-friendly policy. Sustainable transport and the electric vehicle should take high gear. While people are staying at home or using work from home scheme, building indoor environment must specially be taken care of as a compromised indoor environment affects and increases the risk of many diseases. Also, the energy-efficient building will play a key role to abate the enhanced building energy demand and more generation from renewable sources should be in priority. It is still too early to predict any forecast about the regain period of all those sectors but with vaccination now being introduced and implemented but still, it can be considered as an ongoing process as its final results are yet to be seen. As of now, COVID-19 still continue to grow in certain areas causing anxiety and destruction. With all these causes, effects, and restoration plans, still SDGs will be suffered in great order to attain their target by 2030 and collaborative support from all countries can only help in this time. Graphical abstract Image 1 Keywords COVID-19 Pandemic Social Economic Environment Energy Transport Lockdown Quarantine Sustainable development goal (SDG) United nation (UN) Handling editor: Prof. Jiri Jaromir Klemeš ==== Body pmc1 Introduction RNA-enveloped coronaviruses ranging from 60 nm–140 nm in diameter along with a crown-like appearance, can be witnessed in humans, other mammals, and birds which reasons respiratory, enteric, hepatic, and neurologic diseases (Lu et al., 2020). They are well-known to mutate and recombine (Zhou et al., 2020) to create human diseases including 229E, OC43, NL63, HKU1, severe acute respiratory syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV) (Baghizadeh Fini, 2020). Among them, the first four are common and elicits simple cold symptoms while the severe SARS-CoV and MERS-CoV have a zoonotic origin that cause fatal illness. Previously, in 2002–03 because of SARS, more than 8000 people suffered serious sickness while 774 people died. Further, in 2012, MERS-CoV instigated 2494 infections, with over 858 deaths worldwide (Chakraborty and Maity, 2020). SARS-CoV-2 has 88–89% resemblance to bat-SL-CoVZC45 and bat-SL-CoVZXC21 (two bat-derived severe acute respiratory syndrome-like coronaviruses) (Lai et al., 2020) and, 79% and ~50% similarity to SARS-CoV and MERS-CoV respectively (Lu et al., 2020). In December 2019, patients suffering from pneumonia, known to evolve from an unknown cause, which has its epidemiological link with the wet animal wholesale market, was reported in Wuhan, Hubei Province, China. Wuhan, the capital city of Hubei province and one of the largest cities in central China, located in the middle of the Yangtze River delta, experiences subtropical humid, warm summer, cold winter, and monsoon climate with a population of 10 million as of 2017. On Jan 9, 2020, World Health Organisation (WHO) identified the mysterious cause of the disease to be the coronavirus and on Jan 12, 2020, WHO officially avowed this fast-spreading virus as “2019-novel coronavirus (2019-nCoV), which was further reformed to be SARS-CoV-2 on Feb 11, 2020, and also officially professed this disease to be COVID-19 (CO-Corona; VI-Virus; D-Disease; 19: year) (Tay et al., 2020). This was declared as the sixth public health emergency of international concern following H1N1 in 2009, polio in 2014, Ebola in 2014 in West Africa, Zika in 2016 and Ebola in 2019 in the Democratic Republic of Congo (Chakraborty and Maity, 2020). Eventually, the market of Wuhan was shut down from Jan 1, 2020, as part of efforts to contain the outbreak, however, a large number of patients still confirmed to have COVID-19 even without exposure to the market but either had a travel history from Wuhan or any close physical connexion with a COVID-19 patient (including health-care workers). This proposed a strong human-to-human transmission of the virus, thereby leading to 162,506 infections by Mar 12, 2020, worldwide. Thus, on Mar 11, 2020, WHO announced this to be a pandemic, when in that very same day a total number of confirmed cases reached 118,319 including 4292 deaths, worldwide. Recessions, down-turns, wars, revolutions, earthquakes, and volcanos seemed like minor blips when it was compared to the global lockdowns, an expensive state interventionism on a scale previously witnessed not for a millennium. Quarantine, lockdown, and social distancing are few of the popular terms, which soon gained recognition during this pandemic. Researchers started focusing on writing articles about different sectors of society, which were severely affected due to COVID-19. More than 66,000 articles have been published between Dec 2019 and Jan 2021 (web of science), concerning the rise of this pandemic. A major section of the researches includes the medical approaches or pathological findings to study the clinical characteristics of the corona virus, its clinical course, risk factors, association or similarity with previously occurred diseases, case study of various patients with COVID-19 and other diseases, treatment or clinical trials of hydroxychloroquine and azithromycin for the cure of COVID-19 and vaccine development. Also, researches were made on various strategies to diagnose patients with COVID-19 at early stages either clinically, statistically or computationally by studying and finding correlation between environmental changes and the number of patients in each area, or scrutinizing community and localities such as employing wastewater management to detect coronavirus. Apart from the medical study, researchers also studied widely on various sectors of society, which are greatly damaged due to COVID-19 pandemic. Researchers also greatly indulged themselves in scrutinizing how COVID-19 and its associated system of lockdown or social distancing affected economy of each nation (Ghosh et al., 2020c), tourism, science (Zhu et al., 2020), community, work, family life (Lehmiller et al., 2020), nationalism (Woods et al., 2020), politics, relationships (Brown et al., 2020), physical activities, pollution (Zheng et al., 2020), mental health (Killgore et al., 2020), food behaviour (Smith and Wesselbaum, 2020), education sector (Ahimi et al., 2020), etc. This work tried to highlight how COVID-19 propagated around the world creating a drastic impact on human social life, economy, environment, and secondary fields like transport and energy, which were reviewed elaborately. Adverse impacts of COVID-19 could compromise to achieve UN's 2030 SDGs Agenda. Recent progress in terms of vaccine and treatment was also briefly discussed. For this work, we searched relevant databases, including, Google Scholar, Science Direct, Web of Science, to investigate published literature in the past few decades. The search keywords were COVID-19, SARS-CoV-2, Pandemic, COVID and environment, Economy, Energy and Transport. However, because of the diverse terminology, we also employed other terms to obtain more work to review which were included and investigated for this work. The rest of the paper is structured as follows. Section 2 explores the Chronological history for COVID-19. Section 3 discusses the affected area due to the COVID-19 including human social life, environment, economy, energy, transport. Finally, section 4 summarises the discussion and section 5 draw the main conclusions. 2 Methodology The systematic literature review method is always better than a traditional review because it helps to identify the gaps in studies and provides information on areas where the majority have been undertaken. However, these COVID-19 cases and associated incidents are very new and at the time of the first submission, it was only 6 months happened after COVID-19. Thus, most of the information was covered including either government documents, quick submission or the different published research article. It was not either very clear at that time how and for how long the global impact will remain active. However, after the first revision and during the time of the second submission, it was clear that the impact of COVID-19 remained drastic and aligned with the previous work. For the discussion section, a wider review has been included which has an immense influence on building energy has been included. In our study, we did not investigate more on details of COVID-19 virus. Fig. 1 illustrates the mechanism of performing this review work. At the first step keywords such as COVID-19- energy, COVID-19 –social, COVID-19-environment, COVID-19-transport, COVID-19 –economy, COVID-19 sustainable were employed to obtain published work. Details of the COVID-19 virus and its genetic structure and comparative relation between the genetic structures of other SARS viruses were excluded. This information was added only in the introduction section to start the topic. For the 2nd steps publication period between 2020 and 2021 was included. In this process, we excluded the work, which was based on perspective and added only which included real time data analysis. However, to make a clear and positive discussion more relevant work in the building sides were included and for them, publication ranged varied between 2015 and 2021 however priority was given to the most recent updates for each specific topic.Fig. 1 Employed literature search approach in the selection of reviewed studies. (Red arrow indicates exclusion and sky-blue arrow indicates inclusion of study). Fig. 1 3 A chronological history of COVID-19 3.1 Spread of COVID-19 Established on rudimentary observations on 1099 COVID-19 confirmed patients, it was recognized that COVID-19 undertakes 14 days (median time) to transfer from symptom onset to death. SARS-CoV-2 has a briefer median incubation period than SARS (4 days) (Lessler et al., 2009) and MERS (7 days) (Cho et al., 2016) but high latency period of maximum 24 days (Wang et al., 2020e), encouraging high COVID-19 transmission risk (Wang et al., 2020c). Also, coronaviruses are stable and can be detectable on aerosol (3h), copper (4h), cardboard (24h), and on stainless steel and plastic (2–3 days) (Doremalen et al., 2020). These results demonstrate crucial information about the stability of these pathogens which implies that transmission/infection is conceivable to people from touching of contaminated objects along with human-to-human transmission via droplets or direct contact, and this type of infection has revealed a basic reproduction number of about 2.24–3.58 (Remuzzi and Remuzzi, 2020). Starting from China (Lau et al., 2020a), COVID-19 spread out more rapidly around the world primarily through the air travel (Nakamura and Managi, 2020) and cruise travel (Ito et al., 2020). Fig. 2 shows the present COVID-19 cases in the world (230 countries infected due to COVID-19) and Fig. 3, Fig. 4 show the chronological event of COVID-19 cases globally. Now globe is suffering from the hit of second wave (Hafeez et al., 2021).Fig. 2 Cumulative number of COVID-19 cases in the world for a time period of 1 year: Total number of cases, number of active cases, number of death cases and total number of cured cases daily. Fig. 2 Fig. 3 Chronological major event for COVID-19 globally (from Dec 2019–July 2020). Fig. 3 Fig. 4 Chronological major event for COVID-19 globally (July 2020–March 2021)). Fig. 4 3.2 Techniques to combat with COVID-19 transmission: social distance, lockdown, and sewage management To inhibit the spread of the infection, standard recommendations including regular washing of hands, mouth, and nose during coughing and sneezing and usage of masks outdoors were implemented. Another commendation is to maintain 1–2 m of social distance to avoid close contact with anyone showing symptoms of respiratory illness. Thus, social distancing became a highly recommended and appreciated practice to eliminate unnecessary transmission spread (Wilder-Smith and Freedman, 2020) (Sen-Crowe et al., 2020). Over a hundred countries in the world started lockdown to combat with COVID-19 at the end of March (BBC, 2020a). Wastewater management in different localities became one of the means to detect early COVID-19 and prevent its spread in the community. Even though the social distance measure designed by the WHO for influenzas was 1m, different distance measures are employed worldwide: the UK and New Zealand held 2 m, while 1.8 m, 1.5 m, and 1 m were exercised in the USA, Australia, and Singapore, respectively. COVID-19, not being a flu, all the distance measures the different country adopted were not considered scientific. It is still not explicit how far an infectious droplet can travel. Real-time experimental data is not accessible as analysis includes the presence of several variables: the number of infectious particles and their airborne survival, humidity and the speed of expulsion, thereby making the process complicated. Still, for the secure side, all countries enhanced the measurements. Australia executed the distance based on the length of the available shopping trolley in the market. In the UK, national health services and other health bodies adopted higher distance criteria for more safety precautions. The USA previously estimated that a safe distance of 1.8 m, declines the flu transmission. However, during activities such as running and exercising, a safe distance of 5 m, while 20 m for cyclists or usage of personal protective equipment (PPEs) amongst Chinese health care workers led to the merest spread of the ongoing coronavirus within hospitals (Brook, 2020). Furthermore, >25% of health care workers in Sweden suffered severe infection due to the lack of PPEs. Additionally, it is also critical that healthy individuals remain disease-free to curb the pandemic. Therefore, the application of facemasks is empowered as proactive protective measures to safeguard health care personnel, patients, and healthy individuals during not only this pandemic but applicable for future viral outbreaks. Prevention of spreading the virus from infected people is possible by using facemasks. The first countries to initiate mobility prohibitions to colleges, universities, and apply telework due to the COVID-19 situation were: Mongolia and China. South Korea methodically controlled this outbreak and stood as an epitome for the world to learn from their techniques that they have experienced and developed during the MERS outbreak back in 2015, which had 19% fatality and 40–90% infection rate (Her, 2020). Brazil declared a public health emergency on Feb 3, along with social restriction set up in two most populous states, São Paulo and Rio de Janeiro. In Rio de Janeiro, a series of events took place. From 21st March, partial lockdown was started when schools and universities were closed, bars, restaurants, beaches, shopping centres, and commerce in general (except for food and medicines) were closed, public events were cancelled, public transport within the city was limited and work at home was recommended. The first lockdown on Feb 22, order issued in a cluster of cities in Lombardy and Veneto regions in the north, which further expanded till March 8 to all of Lombardy and 14 other northern provinces (Ren, 2020). By March 9, lockdowns began to be applied in other countries, where, not all countries carried out a lockdown, there were some exceptions such as Kazakhstan (Astana), Romania and Indonesia, where a state of emergency was declared by March 15, March 24, and on April 2, respectively. However, some countries eased the lockdown with restrictions. Slovakia (Bratislava) allowed its people to walk or exercise outdoors with mask protection. Mexico City declared a voluntary quarantine, whereas Bangkok (Thailand) and Belgrade (Serbia) declared a curfew since April 4. Saudi authorities reported its first COVID-19 case on March 2. This case was imported by a Saudi national returning from Iran via Bahrain. The Saudi Ministry of Sports announced, all sports competitions to be indoors from March 7, in addition to the suspension of the 2020 Saudi Olympics Games, planned to start on March 23, 2020. March 8, the Saudi Ministry of Education declared the closure of schools and on March 20, they suspended all domestic public transportations: flights, trains, buses, and taxis in a heightened effort to stop the spread of the virus (Yezli and Khan, 2020). Lockdown measure was announced in the UK, on March 23, except essential businesses. People were only allowed to go outside for shopping, necessities, health reasons, and one form of exercise a day, or work if it was considered ‘essential’ such as firefighters, police, or electricity provision. Belgian government took strong measures on March 12, and ordered the closure of schools and cafes, along with the cancellation of all public gatherings. Strict strategies issued on March 17, ordered the closure of non-essential shops, prohibited non-essential travel, and banned all gatherings. Fig. 5 shows the implemented lockdown dates for different countries.Fig. 5 Lockdown time globally for different countries for COVID-19 pandemic. Fig. 5 Even though major lockdown restrictions were uplifted from most of the countries after august, as the number of cases decreased, but still some countries continued to stay under lockdown or restricted movements with varying degrees. Also, reportedly, a new variant of the novel coronavirus has been detected in the UK (Wise, 2020), South Africa (BBC, 2020a), Denmark and the Netherlands resulting a sharp increase in number of infections in these countries (News, 2020). Thus, these countries toughened their restrictions in order to control the spread of the new variant. The UK implemented local lockdown measures under tier system, with restrictions such as banning inter-household mixing and curfew in various sectors. These lockdown restrictions are expected to be implemented until end of February. Other countries like Australia, Austria, and Denmark also employed similar lockdown restrictions, keeping all shopping centres and arcades closed with only essential shops kept open. Germany, Greece, Ireland, Israel, Italy and Mexico continued with their partial-lockdown restrictions, after facing their second wave. The new rules mandate only gatherings of up to five people inside and 10 people outside. Further, the Netherlands, Poland, Portugal and Switzerland extended their lockdown for two weeks more till end of January or even till February. Schools, shops, restaurants and all entertainment zones continued to remain closed. The lockdown in the USA remained controversial situation since the start. More than 50 US states have reopened while high alert zones are still at a pause. India have been the among the most infected country due to its huge population, but unfortunately it started uplifting its lockdown restrictions to preserve its largely falling economy. Public transport and shops all reopened with mandatory wearing mask rule. Economy dropped greatly, with 19 million people losing jobs, however, India faced the lowest death toll in comparison to US and Brazil. Some countries still spoke out against the localised lockdowns like Spain. South Korea was one of the few countries, who managed to control the spread at very early stage. Schools reopened and people returned to their normal life with lifting of restrictions. However, wearing masks and taking precautions still persist. Iran however faced a third wave of outbreak, nevertheless lockdown was lifted, and schools reopened. Countries like Hungary and Lithuania, reopened schools, open-air restaurants but kept their borders closed. Singapore, Dubai, Thailand, began phased reopening with reopening schools and a combination of in-person and virtual learning and work, encouraging people to work from home, reopening stores, bars, restaurants with limited capacity. Russia on the other hand eased restrictions, opened borders, and reportedly became the first country to approve the coronavirus vaccine. Among the other countries, Saudi Arabia reopened mosque and eased restrictions, Colombia surprisingly reopened tourism even with still-rising cases. The situation in different parts of the world looks different and is expected to change regularly until there is a permanent solution to this COVID-19 pandemic. Previous studies with various viral diseases focusing on screening communities and scrutinizing sewage for traces of a pathogen have provided an indication of whether or not the pathogen is existent in the population along with its corresponding transmission pattern (Larsen and Wigginton, 2020). Apart from respiratory diseases, Diarrhoea is also reported as a significant symptom in COVID-19 cases, which was also quite prominent during the outbreak of SARS (Xiao et al., 2020). Thus, various researchers across the world employed different approaches aiming towards surveillance and detection of wastewater data to track down its relationship with coronavirus and the number of COVID-19 cases (Chen et al., 2020b). “Wastewater”, or “sewage,” includes water from household/building use (i.e., toilets, showers, sinks) that can contain human faecal waste, as well as water from non-household sources (e.g., rainwater and industrial use.) which can be tested for RNA from SARS-CoV-2, the virus that causes COVID-19. Medema et al. used four qRT-PCR assays to test sewage samples of cities and airport area during the outbreak of COVID-19 in Netherlands. A distinct direct and highly sensitive correlation was observed with increase in RNA virus in sewage with increasing COVID-19 cases (Medema et al., 2020). Even in Spain, faecal shedding of SARS-CoV-2 RNA from COVID-19 patients were reported, where Randazzo et al. investigated the occurrence of the virus in wastewater treatments plants in major municipalities using aluminium hydroxide adsorption-precipitation concentration method and real-time RT-PCR, revealing detection of virus in wastewater in early stages of the spread of COVID-19 (Randazzo et al., 2020). Peccia.et al. demonstrated the concentrations of SARS-CoV-2 RNA in primary sewage sludge obtained from COVID-19 cases in hospital during the primary outbreak in the New Haven, Connecticut. They reported a high-resolution dataset generated from sewage sludge along with statistical analysis to infer the lead-time their data may provide over epidemiological indicators. These studies strengthen the evidence that wastewater monitoring could be a powerful tool in tracking the spread of COVID-19 (Peccia et al., 2020). The Centres for Disease Control and Prevention (CDC), US Department of Health and Human Services (HHS), and agencies throughout the federal government, are developing a National Wastewater Surveillance System (NWSS) in the state, tribal and local areas in response to the COVID-19 pandemic. The data generated by NWSS will help public health action and a better understanding of the extent of COVID-19 infections in communities (NWSS, Times, 2020). Therefore, quantitative SARS-CoV-2 measurements in untreated sewage can provide information on changes in total COVID-19 infection in the community, depending on the frequency of testing, sewage surveillance can be a leading indicator of changes in COVID-19 in a community and its detection in sewage serves as a COVID-19 indicator that is independent of healthcare-seeking behaviours and access to clinical testing. 3.3 Vaccine and treatment Most common symptoms of COVD-19 cases include cough, dyspnoea fatigue, fever, sputum production, muscle ache, gastrointestinal issues, sore throat, headache, rhinorrhoea, sneezing, nasal congestion (Xu et al., 2020b). According to research, the immunity of COVID-19 patients declines at a higher rate within a month after recovering from it (Chen et al., 2020a). A first longitudinal study using 90 patients and healthcare workers at Guy's and St Thomas' NHS foundation trust found levels of antibodies that can destroy the virus peaked about three weeks after the onset of symptoms then swiftly declined (Seow et al., 2020). Table 1 shows explicitly different levels or stages of vaccine development including preclinical test, safety tests, animal trials and human trials. As of July 2020, 155 vaccines were being developing, and 23 vaccines were under human trial (Table 2 ). A rapid success in the field of development of vaccine was seen and as of January 2021, there are around 64 vaccines in clinical trials on humans. Among these 3 vaccines have been approved for full use, 7 vaccines are in their early or limited use stage, 20 vaccines are in large scale efficacy tests (Phase 3), 20 of them are in expanded safety trials (Phase 2) and 43 of them are testing their safety and dosage criteria (Phase 1) and 85 preclinical vaccines are under active investigation in animals. Table 3 shows a list of vaccines that have reached approval and trials in humans, along with a selection of promising vaccines being tested for their final approvals. By the end of February 2021, 256 COVID-19 vaccines have been developed, 182 under pre-clinical trials and 74 in clinical trials (Li et al., 2021). Fig. 6 shows currently active different vaccines in different countries.Table 1 Stages of vaccine development and testing (T. N. Times, 2020). Table 1Step 1 Preclinical Test A vaccine to animals such as mice or monkeys to see if it produces an immune response. Step 2 Phase I Safety Trials vaccine is given to a small number of people to test safety and dosage as well as to confirm that it stimulates the immune system Step 3 PHASE II Expanded Trials Vaccine is given to hundreds of people including children and adults to see if it acts differently in them. Further, this trial checks the vaccine's safety and ability to stimulate the immune system. Step 4 PHASE III Efficacy Trials Vaccine is given to thousands of people and wait to see how many become infected, compared with volunteers who received a placebo. These trials can determine if the vaccine protects against the coronavirus Step 5 Approval Regulators in each country review the trial results and decide whether to approve the vaccine or not. During a pandemic, a vaccine may receive emergency use authorization before getting formal approval. Table 2 Vaccine under various phases: I/II, II and III (as of July 2020). Table 2Phase Name Developer Type Design, Product Description Location Start date Phase III Moderna mRNA-1273 Moderna/NIAID RNA Double-blind, mRNA-1273, encodes for a form of the spike (S) protein on the virus USA 27/07/2020 Phase III Sinovac vaccine Ege University Inactivated Double-blind, Inactivated (inactivated + alum); CoronaVac (formerly PiCoVacc) Brazil 01/07/2020 Phase II/III Oxford AZD1222/ChAdOx1-S Immunomic Therapeutics/EpiVax/PharmaJet Non-replicating viral vector Single-blind, Non-replicating viral vector; AZD 1222 (formerly ChAdOx1) UK 28/05/2020 Phase II AZLB protein subunit vaccine Protein subunit Double-blind China 12/07/2020 Phase II Cansino Ad5-nCoV Non-replicating viral vector Double-blind China 12/04/2020 Phase II Moderna mRNA-1273 Symvivo RNA Observer-blind, doseconfirmation, RNA; LNPencapsulated mRNA (mRNA 1273) USA 29/05/2020 Phase I/II Aivita AV-COVID-19 Other Double-blind, dose-finding USA 01/07/2020 Phase I/II Altimmune T-COVID Non-replicating viral vector TBC TBC 01/06/2020 Phase I/II Bharat Covaxin Inactivated Double-blind India 13/07/2020 Phase I/II BIBP/Sinopharm BBIBP-CorV BioNet Asia Inactivated Double-blind, dose-finding, Inactivated China 28/04/2020 Phase I/II BioNTech BNT162 Takis/Applied DNA Sciences/Evvivax RNA 3 LNP-mRNAs; BNT162 USA 29/04/2020 Phase I/II BioNTech BNT162 RNA Open-label, dose-finding Germany 23/04/2020 Phase I/II CAMS vaccine Chula Vaccine Research Center Inactivated Double-blind, dose-finding, Inactivated China 15/05/2020 Phase I/II Cansino Ad5-nCoV Mediphage Bioceuticals/University of Waterloo Non-replicating viral vector Double-blind, dose-finding, Non-replicating viral vector; Adenovirus Type 5 vector (Ad5-nCoV) Canada 01/08/2020 Phase I/II Genexine GX-19 Genexine Consortium (GenNBio, International Vaccine Institute, (KAIST), (POSTECH)/Binex DNA Double-blind South Korea 17/06/2020 Phase I/II Inovio INO-4800 DNA Open-label (A), double-blind (B), dose-finding South Korea 22/06/2020 Phase I/II Sinovac vaccine Inactivated Double-blind, dose-finding China 16/04/2020 Phase I/II Sinovac vaccine Inactivated Double-blind, dose-finding China 20/05/2020 Phase I/II WIBP vaccine Entos Pharmaceuticals/Cytiva Inactivated Double-blind, dose-finding, Inactivated China 11/04/2020 Phase I/II Zydus Cadila DNA vaccine DNA Double-blind India 13/07/2020 Table 3 Vaccine under various phases: II/III, and III (as of January 2021). Table 3Phase Name Developer Type Design, Product Description Efficacy Country Status Phase 2/3 Comirnaty Pfizer/BioNTech mRNA Muscle Injection/Freezer storage (−70C), 2 doses, 3 weeks apart 95% USA Approved in several countries, emergenecy in USA, elsewhere Phase 3 mRNA-1273 Moderna mRNA Muscle Injection/30 days with refrigeration, 6 months at −20C, 2 doses, 4 weeks apart 94.50% USA Approved in canada, emergency use in U.S., E.U., Israel Phase 3 Sputnik V Gamaleya Ad5 and AD26 Muscle injection, Freezer storage. Developing an alternative formulation that can be refrigerated, 2 doses, 3 weeks apart 91.40% Russia Early use in Russia, elsewhere Phase 2/3 AZD1222 Oxford/AstraZeneca ChAdOx1 Muscle injection, Stable in refrigerator for at least 6 months, 2 doses, 4 weeks apart 62%–90% (depending on the dosage) UK/Sweden Emergency use in Britain, India, other countries. Phase 3 Convidecia CanSinoBIO Ad5 Muscle injection, refrigerated, single dose Unknown China Limited use in China Phase 3 EpiVacCorona BEKTOP Protein Muscle injection, Stable in refrigerator for upto 2 years, 2 doses, 3 weeks apart Unknown Russia Early use in Russia Phase 3 BBIBP-CorV SINOPHARM Inactivated Muscle injection, 2 doses, 3 weeks apart 79.34% China Approved in China, elsewhere Phase 3 CoronaVac Sinovac Inactivated Muscle injection, 2 doses, Refrigerated, 2 weeks apart 78% China Limited use in China Phase 3 Covaxin Bharat BIOTECH Inactivated Atleast a week at room temperature, 2 doses, 4 weeks apart unknown india Emergency use in India Phase 3 CVnCoV CUREVAC Inactivated Muscle injection, Stable at 3 months at 2–8C, 2 doses, 4 weeks apart unknown USA Under Trial Phase 2/3 AG0302-COVID1 AnGes Inactivated Skin injection, over a year at room temperature, 2 doses, 2 weeks apart unknown japan Under Trial Phase 3 ZyCoV-D Zydus Cadila Inactivated Skin injection, stable at room temperature for three months, 3 doses, 4 weeks apart unknown India Under Trial Phase 3 Ad26.COV2.S Johnson-Johnson Ad26 Muscle injection, Upto 2 years at -4C, and upto 3 months refrigerated at 2–8C, 1 dose unknown USA/Israel Under Trial Phase 3 NVX-CoV2373 NOVAVAX inactivated Muscle injection, Stable in refrigerator, 2 doses, 3 weeks apart unknown USA under Trial Phase 3 ZF2001 Anhui Zhifei Longcom inactivated Muscle injection, 3 doses, 4 weeks apart unknown China under Trial Phase 3 CoVLP Medicago inactivated Muscle injection, Stable in refrigerator, 2 doses, 3 weeks apart unknown Canada under Trial Fig. 6 Presently acting COVID-19 vaccines in different countries. Fig. 6 3.4 Climate dependency on COVID-19 transmission The coronavirus has some impending features such as mutation and recombination while spreading, thereby causes severe health issues to patients of old age and those with an existing health condition. Influenza prototype diseases wreck humankind under low daily temperature and with humidity up to 70% (Park et al., 2020). Initially, investigations explicated COVID-19 transmission decreased with an increase in temperature (Tobías and Molina, 2020). Another study that involved 429 cities suggested that temperature may have a strong relation to COVID-19 infection and transmission, which collected data for only 16 days (Jan 20 ~ Feb 4) (Wang et al., 2020d). Using these climatic correlations with COVID-19 cases data (Jan 20 to Feb 29: 2299 COVID-19 death counts) in Wuhan, temperature and humidity proved to have an impact on mortality, and increased temperature showed a slight decline in the rate of death. A positive association with COVID-19 daily death counts during the diurnal temperature range (r = 0.44) and, a negative association for relative humidity (r = −0.32) was observed (Ma et al., 2020). Additionally, investigation of daily COVID-19 cases association with daily average temperature and relative humidity in 30 Chinese provinces (Hubei: Dec 1 ~ Feb 11 for; other districts: Jan 20 ~ Feb 11) showed both temperature and humidity had a negative association with COVID-19 cases, though inconsistent results throughout mainland China were observed (Qi et al., 2020). Results from capital cities of 30 provinces in China (Jan 20 ~ Mar 2) showed that after controlling a population, migration meteorological factors played an independent role in the COVID-19 transmission, which is plausible if local weather has a low temperature, humidity, and mild diurnal temperature range (Liu et al., 2020b). Data from five Brazilian (Brasilia, Manaus and Fortaleza, Rio de Janeiro, São Paulo) cities showed that higher mean temperatures and average relative humidity favoured the COVID-19 transmission (Auler et al., 2020). Since Feb 29, (national onset) to Mar 31, Mexican capital, and, other 31 states showed temperature associated negatively with the local confirmed COVID-19 positive cases (Méndez-Arriaga, 2020). Statistical analysis involved investigation from Jakarta (Tosepu et al., 2020), Indonesia, and California (Bashir et al., 2020), the USA reported that temperature has a moderate impact on COVID-19 transmission. Data were analyzed from 12 cities of Turkey and observed that the crowd has a positive relationship with several cases, wind speed has an inflectional impact, and the temperature has a negative relation with COVID-19 cases (Şahin, 2020). Additionally, it is reported that places with similar COVID-19 transmission had the same temperature and humidity (Sajadi et al., 2020). Data from 185 countries/region regarding COVID-19 cases (Center for Systems Science and Engineering at Johns Hopkins University; more than 3,750,000 confirmed COVID-19 cases) between Jan 21 ~ May 6, showed 60.0% of confirmed COVID-19 cases happened in places where the ambient temperature ranged from 5 °C to 15 °C (Huang et al., 2020). Oslo climate, maximum and minimum temperature were positively, and precipitation was negatively associated with COVID-19 (Menebo, 2020). In German federal states, COVID-19 mortality negatively correlates with local air humidity (Biktasheva, 2020). Impact of temperature and humidity on COVID-19 cases, were investigated in 166 countries (excluding China) till March 27, which showed negative relation to new daily cases and deaths (Wu et al., 2020) which also supported by another work (Sobral et al., 2020). For the USA, vulnerable absolute humidity range for COVID-19 spread came into existence (Gupta et al., 2020). The air quality index has a strong effect on COVID-19, while the temperature range varies from 10 to 20 °C (Xu et al., 2020a). Investigation in Iran showed that low solar radiation, humidity, and wind speed promote the spread of COVID-19. Nevertheless, high population density escalates the rate in higher-order, since populated cities such as Alborz, Gilian, Mazandaran, Qom, and Tehran had higher number of infection (Ahmadi et al., 2020). In Japan, a positive correlation of mean temperature and cumulative COVID-19 cases and no evidence on infectivity and temperature was noticeable (Ujiie et al., 2020). (Xie and Zhu, 2020) investigated 122 Cities in China, where the temperature has a positive linear relationship with the number of COVID-19 cases with no evidence to support that COVID-19 cases counts could decline due to warm weather. However, Yao et al., 2020 reported that temperature and UV radiation had no strong association with COVID-19 cases in Chinese Cities (Yao et al., 2020). Hubei, Hunan, and Anhui province had a positive relationship between temperature and COVID-19 while, Shandong and Zhejiang provinces had a negative relation (Shahzad et al., 2020). Collected data of daily confirmed cases from the capital and 27 states in Brazil also confirmed that COVID-19 cases do not decline due to temperature above 25.8 °C (Prata et al., 2020). Additionally, inconsistent results between temperature and COVID-19 in the provinces of Spain were observed (Briz-Redón and Serrano-Aroca, 2020). It is not entirely prominent that daily temperature and humidity possess an impact on this COVID-19 transmission or reduction. According to WHO, temperature and humidity may have some relation to COVID-19, especially their survival capacity outside the human body, but population density and human contact play a critical role in the spread of COVID-19 (WHO, 2020). Hence, non-meteorological factors such as population density should get attention to obtain more reliable results. In another study based in Iran showed that temperature has a high and population density has a low sensitivity to COVID-19 cases (Jahangiri et al., 2020). The efficiency of the data collection is also a crucial task, which could be the ground for not having a suitable correlation. Authors and data providers in the (Gov.UK, NHS) betokened that there is a lag of data entry. In addition, an infected virus whose majority of nature is unknown will be a critical task to predict. Some authors suspected that lowering air pollution can be an influential factor to abate COVID-19 cases. As COVID-19 is primarily respiratory in nature, hence the contribution from air pollution for this disease susceptibility or outcome is a great interest among the researchers (Contini and Costabile, 2020). suggested that air pollution might have been a contributing factor to the high number of COVID-19 fatalities in Italy. In another work (Dutheil et al., 2020), showed that the COVID-19 death rate decreased due to low air pollution in China (3158 in China and 4607 worldwide reported deaths). However, this data also showed that the COVID-19 transmission rate at an early stage in the first week of March 2020 and during the lockdown in China from Jan and Feb, improved the air quality (Dutheil et al., 2020). Italy showed a clear correlation between air pollution and COVID-19 spread (Fattorini and Regoli, 2020). In northern Italy, Lombardy and Emilia Romagna, the most polluted regions in Europe, experienced the highest COVID-19 cases. The author claimed that a potential positive correlation between COVID-19 cases and pollution was present in that region (Conticini et al., 2020). COVID-19 cases and air pollution indicators from Jan 1 ~ Apr 30 showed that surface levels of air pollution and dry air intensified the fast diffusion effects in Milan (Zoran et al., 2020). Further, London revealed a strong correlation between air pollutants (PM 2.5 and NO2) indicator and COVID-19 cases (Sasidharan et al., 2020). The high presence of NO2 was associated with COVID-19 mortality in Italy and Spain (Ogen, 2020). Hazardous waste sites are also sources of air pollution. Recently a “double-hit hypothesis” was proposed. Due to chronic exposure to PM 2.5, alveolar ACE-2 receptor overexpression, which might increase viral load in patients exposed to pollutants, in turn, depleting ACE-2 receptors and impairing host defences. High atmospheric NO2 may provide a second hit causing a severe form of SARS-CoV-2 in ACE-2 exhausting lungs resulting in a worse outcome (Frontera et al., 2020). In general, prolonged exposure to air pollution creates acute respiratory disease, inflammation, and asthma attack, which may lead to death, distress syndrome, and eventually causes death. However, for COVID-19 cases, no direct link was found with the air pollution that can work as a cofactor. 4 Findings In this section finding from the different publication has been reviewed. To restrict the transmission of COVID-19, lockdown measure, and social distancing was adopted. Starting from March and April, extended up to May, many countries underwent lockdown; shut their industry, transport, and other non-essential sectors (Haleem et al., 2020). Back in 2015, United Nation defined 17 sustainable development goal, (SDG) which developed, and developing both countries will try to attain by making new strategies and working together to have a peaceful and sustainable world. These seventeen SDGs include no poverty (SDG1), zero hunger (SDG2), good health and well-being (SDG3), quality education (SDG4), gender equality (SDG5), clean water and sanitation (SDG), affordable and clean energy (SDG7), decent work and economic growth (SDG8), industry innovation and infrastructure (SDG9), reduced inequalities (SDG10), sustainable cities (SDG11), responsible consumption and production (SDG12), climate action (SDG13), life below water (SDG14), life on land (SDG15), peace justice and strong institution (SDG16), partnerships for the goals (SDG17). It is widely discussed that progress towards these 17 SDGs was mixed before the pandemic (Huan et al., 2021). In the world, still, 673 million have no toilet, 736 million people suffer from poverty, 785 million people have no basic drinking water, 821 million people are undernourished, 840 million people have electricity. At least 28 poor countries will not attain the SDGs 1–4, 6 and 7 by 2030 (Moyer and Hedden, 2020). After the pandemic achieving this target is projected to be slower (Lancet and Health, 2020). Keeping in mind these goals, this work tried to focus on the key five sector, which got immense negative affect during this COVID-19 and specially during lockdown and broadly represents these 17 SDGs. These fives sectors are social/human life, environment, economy, energy, transport. For social impact, we tried to focus on factor, which can shades lights on SDG1-6 while for environment impact our goal was to bring alignment with SDG 6, 7, 13, 15. Impact on economic has relation with SDG 8–10; while for Energy SDG 7, 11–13 are crucial. For transport SDG 6,7,13 and 15 all, of these together had an impact on SDG 16 and 17. It is evident that most of the sectors are connected to each other (as shown in Fig. 7 ) and thus the impact of this pandemic is so drastic. The following section will discuss areas, which got adversely impacted due to COVID-19 or precisely lockdown and now suffering from social distancing.Fig. 7 Five key sector, Social, Environment, Economy, Energy and Transport in terms of 17 sustainable development goal. Fig. 7 4.1 Impact on social/human life Nonetheless, throughout history, pandemics have wrought a considerable number of social changes such as from the black plague in the middle ages to the Spanish Flu in the early 20th century, and now similar effects are happening due to COVID-19 pandemic. COVID-19 created a variety of changes in daily human life. To subdue the transmission, most of the countries started lockdown strategy and social distancing (Mitjà et al., 2020). During the lockdown, people were encouraged to stay at home and only go out if it is inevitably essential, e.g., buying food, and thus advised to use masks and maintain social distance, to reduce the droplet transmission (Lau et al., 2020b). These rules are similarly applicable to any respiratory disease. Because of this social distancing, Olympic and para Olympic games were suspended (BBC, 2020b). However, coronavirus made an immensely positive impact on society, to get in touch frequently with distant friends and family to share experiences and stories and to know their health and wellbeing. Staying at home gave them ample amount of time and opportunity to nurture their old hobbies, interest, and exploring more creative stuff. People kept themselves busy with making or watching different online videos (Sheth, 2020). Though people spent more time at home with a partner, because of stress arising from pandemic and risk of job loss, sexual function did not improve. Collected data from 89 women in Rome showed that high stress and pandemic death in Italy increased their intercourse interval (Schiavi et al., 2020). Unfortunately, COVID-19 has affected all levels of the education system, pre-school to tertiary education. Over 100 countries initiated the cessation of schools. UNESCO estimates 990 million learners to have been affected by the closure of educational institutions, as shown in Fig. 8 (UNESCO, 2020). Children and adolescents faced immense negative psychological impact because of the fear of infection, boredom, lack of personal space at home, stress, lack of in-person contact, family financial issue. For adolescents, life quality improvement, independence achieved through socialization, was absent during the lockdown period. Mean post-traumatic stress scores were four times lower in non-lockdown children than in those under lockdown (Wang et al., 2020b). Home-schooling and widespread use of remote teaching via online learning modules and television were brought into use, to tackle this issue. However, these modes of education are accessible to an economically built country with the availability of internet and computers. Countries with not-so-upgraded systems suffered a lot. These young pupils, being deprived of the right to education during the lockdown, suffered mental trauma leading to emotional breakdown and suicides (Lathabhavan and Griffiths, 2020). It might appear to be a rare issue but not at all an ignorable area. The online education system started, still can't replace the real-world classroom education (Schwarz et al., 2020).Fig. 8 Number of affected students from education due to global lockdown (UNESCO, 2020). Fig. 8 Changes in medical treatment facilities, like oral and telephonic Medicare, were also adopted to avoid COVID-19 transmission (Machado et al., 2020). In addition, exclusive emergency dental procedures conducted to protect the medical personnel and the patients and to reduce as much as possible the consumption of personal protective equipment. It is expected that because of the unavailability of reproductive and sexual health service for women, 2.7 million extra unsafe abortions was being carried out (Wenham et al., 2020). Fear of no food and household material haunted people and reflected in their purchasing style. Additionally, the scarcity effect and stockpiling as unnecessary buying habits in every country, during the lockdown periods, were visible. Toilet paper and cleaning pieces of stuff were on the prime list. People stood in a long queue before entering the shop, which was later restricted to limit the number of customers and handle them simultaneously. In the United Kingdom, new purchase limits, online, and home delivery services, and priority delivery slots for vulnerable or elderly customers started (e.g., Waitrose and Ocado shops) (Pantano et al., 2020). However, ignorance created huge chaos to tackle this transmission. Negligence in the use of masks and maintaining social distancing occurred in every country. Understandably, maintaining social distance in a populated country like India is hard, but, in the USA, people were not eager enough at the beginning to keep them at a safe level. Thus, a mixed reaction among people from different countries, with excessive food storage but not abiding rules, was observed. In contrast, the Kingdom of Saudi Arabia (KSA) implemented strict measures for social distancing where it was hard for them due to their social and religious norms, level of urbanization, and religious mass gatherings annually (Yezli and Khan, 2020). Further, the negative impact of COVID-19 on the economy, daily life, and social activity, created psychological difficulties (Cao et al., 2020). Previously for the SARS outbreak, due to quarantine, high rates of depression and anxiety among people were visible. This pandemic also created psychological issues including depression, frustration and stress while survey was conducted with 1182 individual in New Delhi, India which included different age groups and genders (Chaturvedi et al., 2021). In London 70,000 adults data between 23rd March and 9th August, showed in the early stage of lockdown depression and anxiety was present, which reduced later, may be because of the adaption with circumstances (Fancourt et al., 2021). Similar outcome was also found in Germany (Bendau et al., 2021). Using 500 adult samples from nationwide, the USA community showed the positive impact of COVID-19 on daily life associated with health anxiety, financial worry, and social support, and a negative association with loneliness, due to self-isolation and no social life was prominent (Tull et al., 2020). In the Greek population, insomnia was prevalent for women and people living in an urban area. Financial pressure, changes in social life, and the daily routine increased health issues during a virus outbreak (Voitsidis et al., 2020). Depression, anxiety, and PTSD symptoms were prominent among the USA young adults, age between 18 and 30 years, with high levels of COVID-19-specific worry and loneliness (898 participants from April 13, 2020, to May 19, 2020) (Liu et al., 2020a). Based on different published reports, due to economic hardship, isolation, quarantine suicide rate increased during this time. In some cases, unavailability of food and alcohol was also a reason for suicide. Six different couples committed suicide in Bangladesh, India, and the USA for various reasons such as public harassment, fear from COVID-19, and financial constraint (Griffiths and Mamun, 2020). Suicide for financial distress was higher in economically hard countries (Rajkumar, 2020). Increasing levels of domestic violence, which includes physical, emotional, and sexual abuse increased (Roesch et al., 2020) in Brazil (40%–50%), and in Spain, Cyprus, UK, and Singapore helpline received 20%, 30%, 25%, 33% higher call respectively because of the domestic violence (Bradbury-Jones and Isham, 2020). Domestic violence tripled during February 2020 compared to February 2019 in Hubei, an increase of 30% in France, and 25% in Argentina were observed since they initiated a lockdown in March 17 and March 20 respectively (Boserup et al., 2020). Lockdown adversely affected the life of refugee in Uganda because of the insecurity in income while gender-based and sexual violence and anxiety increased (Bukuluki et al., 2020). During this COVID-19 pandemic, technology played a key role (Oztemel and Gursev, 2020). Worldometer’ renders a real-time update on the genuine number of people known to have COVID-19 worldwide, including diurnal new cases, distribution by countries, and austerity of the disease in each country (recovered, critical condition or death). What's app was employed in Singapore to inform people about the updated COVID-19 details (Wang and Tang, 2020). COVID-19 Intelligent Diagnosis and Treatment Assistant Program (nCapp)” based on the Internet of Things, contributed to the long-term follow-up of patients diagnosed with COVID-19. The ultimate goal is to facilitate different levels of COVID-19 investigation and medication among different doctors from various hospitals to upgrade to the national and international level through the nCapp system (Bai et al., 2020). In India, Arogya Setu was launched to develop a connection between the potential healthcare assistance and the people of India (Singh et al., 2020). A mobile application named Close Contact was launched for chinses civilians to track the corona-positive person (Wang et al., 2020a). The use of video conferencing technologies such as Zoom, Microsoft Teams during the pandemic skyrocketed, as they have morphed from an obscure brand name to a household verb. However, fatigue from excessive use of web conferences also became another cause of illness (Kirk and Rifkin, 2020). Excessive use of digital media and video games during the lockdown and home confinement decreased sleep quality (Cellini et al., 2020). In the event of natural disasters, pandemics, riots, terrorist attacks, criminal activity, home theft rate decreased, thereby leading to a decline in the crime rate, worldwide (Hodgkinson and Andresen, 2020). However, some commercial burglary increased as most of the markets and shops were closed for a prolonged period without any workers or owners around the shop (Mohler et al., 2020). Some habits after this pandemic such as using masks, remote working, less traveling, increased security checking of health in the airport to test the presence of virus, will be remarkably altered (Sheth, 2020). Out of the 17 SDGs, preventing the deaths of new-borns and under-fives, and sending all under five children into primary schools were the two SDGs (SDG 3–4), which were close to being achieved before the COVID-19 pandemic (Fisher, 2020). However, it is evident that COVID-19 has changed the scenario again. Increase of domestic violence during this pandemic proved again that how essentials are the gender equality and women's empowerment (SDG 5 and SDG 10). Existing inequalities in socio-economic and health sector has now enhanced the issues in higher order. COVID-19 impact on the most vulnerable and poor people is now most staggering as action on SDGs were not taken seriously since 2015. 4.2 Impact on the environment Global warming and preventing the rise of global temperature is one of the global challenges. Presently, 90% of the CO2 emission occurs due to human activity such as burning fossil fuels, while 10% comes from deforestation (Jackson et al., 2018). Air pollution, which is a complex mixture of particulate matter (PM) (2.5,10), NO2, SO2, ozone (O3), has an adverse impact not only on the environment but also on human health (Yang, 2020). Combustion of fossil fuels and road transportation (motor exhaust; brake, wear and road erosion; resuspension due to wheel-generated turbulence) emits nitrogen dioxide (NO2), which is appalling for human health and, long-term exposure can even increase the mortality rate. Presence of particulate matter (PM, 2.5 to 10-μm in diameter), in the ambient and engendering from biomass and fireworks burning has an adverse effect on human health, causing asthma and COPD (Liu et al., 2016). Probably the environment is the only sector that got an immensely positive impact form this COVID-19 scenario. International energy agency reported that global coal use was 8% lower in the first quarter in 2020. Due to the Locked down, transport, industry, and all non-essential sectors were closed, which reduced emission significantly. NASA (National Aeronautics and Space Administration) and ESA (European Space Agency) published recent data (Fig. 9 ) declaring that compared to last year, NO2 emission reduced by 30% (Dutheil et al., 2020). The decline in PM2.5 was significant in the US, UAE, Italy, and Spain, in the month of March, due to cumulative lockdown (Chauhan and Singh, 2020). Noticeably, in China, the overall air quality improved as NO2 reduced by 22.8 μg/m3, PM 2.5 decreased by 1.4 μg/m3 particularly in Wuhan (Zambrano-Monserrate et al., 2020) and by 18.9 μg/m3 in 367 other cities (Lal et al., 2020). However, some cities also witnessed the air quality index over 100. These reductions accounted for lowering the particle loadings (Wang et al., 2020f). Air quality showed improvement near the Yangtze River Delta (YRD) region, which is one of the economic city-clusters in Eastern China. However, the percentage of PM 2.5 attributed to residences and long-range transport (Li et al., 2020). Additionally, 44 cities of northern China marked 69.5% reduction in human mobility improving the air quality as SO2, PM2.5, PM10, NO2, and CO decreased by 6.76%, 5.93%, 13.66%, 24.67%, and 4.58%, respectively (Bao and Zhang, 2020). In 2017, the energy sector in Italy (industry and transport) contributed 80% of the total country GHG emissions. COVID-19 related lockdown caused an overall 20% reduction of GHG emission, lower than emissions of March and April in 2015–2019 (Rugani and Caro, 2020). In Milan, Italy, partial lockdown restricted the people movement, and total lockdown terminated industry and transport activities. Reduction of PM10, PM2.5, BC, benzene, CO, and NOx level was observed because of a decrease in road transport (Collivignarelli et al., 2020). In Barcelona, PM10 reduced by 31% (Tobías et al., 2020) and NO2 by 50% (Baldasano, 2020). Initially, Madrid and Barcelona contributed 55% and 56% of NO2 emission from traffic. However, due to the COVID-19 scenario-based lockdown, since March, Barcelona and Madrid (Spain), emitted 50% and 62% less NO2 respectively (Baldasano, 2020). In the continental USA, PM2.5 reduced during the lockdown, especially in urban counties and wherever non-essential businesses were closed (Berman and Ebisu, 2020). During the lockdown period (March 19th to April 14th, 2020), reduction in PM2.5, NO2, and CO concentration by 21%, by 35%, CO by 49%, was noticed in Almaty, Kazakhstan (Kerimray et al., 2020). Sao Paolo Brazil also encountered a reduction in CO and NO2 emission by 64.8%, and 77.3% (Nakada and Urban, 2020). Further, PM10, NO2, and SO2 emissions decreased by more than half during the COVID-19 lockdown period in Salé City, Morocco (Otmani et al., 2020). India, every year battles more than 350,000 new cases of childhood asthma and 16000 premature death attributed to air pollution, mostly NO2 and PM (2.5–10 μm) generated from fossil fuels and transportation sector (CREA, 2020). Additionally, lockdown resulted in the suspension of transportation, and industries, the primary sources of air pollution. The first phase-locked down showed betterment of air quality with a reduction of NO2. Delhi, the capital of India, reported air quality index to change from 900 to below 20 because of the absence of 11 million registered cars from the road, with an alarming reduction in PM2.5 in Delhi (Sharma et al., 2020). In Malaysia, open, burning motor vehicles and industrial emissions are primary sources of PM2.5, which reduced up to 24% due to Movement Control Order from the Malaysian Government (Abdullah et al., 2020a). Studies supported that quarantine and lockdown reduced the PM2.5 for Dhaka (14%), Kampala (35%), Delhi (40%), Bogotá (57%), and Kuwait City (42%). Maximum reduction in the capitals of America, Asia, and Africa was recognized (Rodríguez-Urrego and Rodríguez-Urrego, 2020). PM2.5 concentration reduction was only possible from automobiles or industry, and not from any residential sectors. A report from Ontario, Canada, showed that residential sector emits 56% of PM2.5 emissions. Hence, during the lockdown, 28% of PM2.5 resulted from outdoor cooking using barbeques. However, NO2 and NOX both were lowered because of the reduction of automobiles (Adams, 2020). In amidst of COVID-19 lockdown situation, even though the major air pollutants like PM2.5, PM10, NO2, SO2 reduced largely, Ozone (O3) appeared to increase in various parts of the world: Milan (Collivignarelli et al., 2020), China (Wang et al., 2020f), Rio de Janeiro, Brazil (Siciliano et al., 2020), Barcelona (Tobías et al., 2020), mostly produced from household VOCs in lockdown. An investigation reflected that four European cities (Nice, Rome, Turin, Valencia) and Wuhan in China showed a drastic increase in ozone ~17% and 36%, respectively. Further, the reduction of PM2.5 and PM10 led to less scattering of solar radiation which, eventually increased the solar radiation, favouring O3 formation (Sicard et al., 2020).Fig. 9 NO2 emissions in (a) Spain, (b) France, (c) USA, (d) China before and after lockdown (ESA, 2020) Fig. 9 Another significant factor to be influenced is noise pollution, which reduced due to a decline in road transport. Barcelona indicated a 50% decrease in sound pressure (Baldasano, 2020), whereas, Dwarka river basin of Eastern India marked a drop in noise level from 85 dB to <65 dB (Mandal and Pal, 2020). Another global concern being water pollution also revealed remarkable improvement in water quality during lockdown days. The lagoon of Venice, being affected by the regional geomorphological evolution, anthropogenic stressors, and global change stress from human activities, marked a decline in water traffic due to mobility restrictions during the lockdown (Braga et al., 2020). India witnessed prosperity in water quality: Vembanad lake, longest freshwater lake in Kerala, experienced a 15.9% reduction in suspended PM concentration (Yunus et al., 2020). Groundwater in the proximity of the Tuticorin industrial city noticed a drop in the amount of NO3, As, Fe, Se, Pb, total coliforms, and faecal coliforms (Selvam et al., 2020). National River of India, Ganga (declared in 2018), with over 29 cities, 97 towns and thousands of villages along the banks marked a sudden decrease in the quantity of dissolved oxygen (DO), biological oxygen demand (BOD) and nitrate (NO3-), securing the quality of water nearly at drinkable level (Dutta et al., 2020). Another harmful measure taken to prevent COVID-19 transmission is disinfecting urban public areas by spraying corrosive chlorine-releasing agents, quaternary ammonium cation with the help of trucks, drones, and mini-tankers, destroying the wildlife and human settlement in these areas. Both the physical and mental health of humans is immensely hampered due to the death of wild animals, causing a massive biodiversity massacre (Nabi et al., 2020). Previously, heavy traffic and daylight human activities caused several species to adopt a nocturnal lifestyle. However, current mobility restrictions limited human intervention in wildlife, resulting in free animal wandering even during the daytime. Nevertheless, the negative impact of COVID-19 on wildlife remains elusive, as the feeding of animals depends on human activities, tourism. A prolonged pandemic may endanger such animals due to the scarcity of sufficient, nutritious, and safe food. Compared to 2019, global CO2 emissions for 2020 were estimated to be between 4 and 7%, which was achieved by limiting mobility, and sacrificing societal cost (Fisher, 2020). Thus, this development in the environment is temporary; the scenario will alter after the COVID-19 scenario. Developed and developing both will employ fossil fuel sources to secure productivity (Geography, 2020). To keep global temperature rise below 2 °C and 1.5 °C, need reduction of 3% and 8% a year respectively (Gillingham et al., 2020). To achieve the required target, imposing tax on pollution and emitting other environment pollutant gasses will be good practice (Yoshino et al., 2021). 4.3 Impact on economy Due to COVID-19, every analysis showed that 2020 experienced a negative or reduced growth of the economy. Moody Investor Services estimated 0.5% contraction, an organisation for Economic Co-operation and Development predicted a 1.5% reduction (assessment as on March 3, 2020) and the Institute of International Finance expected a 1.6% reduction. First quarter of 2020 in China faced 6.8% national economic output contraction which was its worst performance in last two decades (Liu, 2021). The United Nations Conference on Trade and Development estimated US$2 trillion shortfalls in global income (Srivastava, 2020), while the USA predicted at least three years of recovery time to cope up with the COVID-19 dip. Areas like commercial aerospace, travel, and insurance might see a more delayed restoration. EU GDP anticipated declination by 7.5% as the IMF states that the global economy will shrivel by 3% by the end of this year (BBC, 2020c). In the second quarter of 2020 global GDP declined by over 4.9% though, it was better than 2007–08 global financial crisis. These factors negatively affect the job market. More than 300 million people lost jobs due to COVID-19 in the second quarter of 2020, higher than the recession faced in 2008–2009 (Kenny, 2020). To cope up with this scenario, some companies already adopted sacking steps. Uber disclosed its plan to lay off 3700 drivers (Heater, 2020), while in the UK, over 600,000 people lost their job between March and July. British Airways, BP, Rolls Royce, restaurants and builders have cut jobs (BBC, 2020d). Even though India did not release official job loss data but Centre for Monitoring the Indian Economy data widely accepted that unemployment increased by 14.2% since March to April (BBC, 2020e). America also marked more than 2.9 million employment as of May 14, 2020, bearing a two-month total of 36 million, and with 20 million jobs lost in April, and the rate of unemployment rose to 14.7% The sectors, which got immensely affected by COVID-19, includes travel & tourism, aviation, automobile. Tourism sector experienced drastic declination as more than 50 million jobs were at risk, as declared by the World Travel and Tourism Council. On 7 May, the UN World Tourism Organisation predicted a decline of 80% from the earnings of international tourism as compared to last year ($1.7 ton) along with 120 million job layoffs. Tourism contributes ~15% of Spain's and some 13% of Italy's GDP (The Guardian, 2020). The next sector to be affected by COVID-19 is the automobile industry. June accounted for more than 6000 automotive job layoffs jobs in the UK. These layoffs affect thousands of jobs affecting an industry, employing around 800,000 people in the UK (Guardian, 2020). Various automobile companies experienced various drawbacks during this pandemic: Toyota Motor minimized its global production capacity by 2% for the month of August along with a temporary halt of production at Bidadi, Karnataka, India, Nissan Motor delayed its production and aimed at 30% more production, by the end of Dec 2020, Volvo Cars experienced a revenue drop of 14.1% along with an operating loss of 989 SEK since January to June, Mercedes-Benz stopped making C-Class in Tuscaloosa, Alabama plant. However, among all this chaos, South Korea reported positive news were light vehicles sold increased by 41%. Maruti Suzuki and Hyundai India both announced enhancement of sales after the retrieval of lockdown in India (Roberts, 2020). The aviation industry is another concerned sector during COVID-19 (Iacus et al., 2020). On March 23, the International Air Transport Association (IATA) budgeted revenue loss from globally passenger airlines ~ $252 billion and contributed ~ $200 billion in government assistance (Forbes, 2020). Previously this sector faced bankruptcy due to oil refusal, airline deregulation, terrorist attacks (9/11 attack) and, SARS (Took three years to overcome the losses) (Sobieralski, 2020). International, low-cost, and regional airlines suffered job layoffs (Sobieralski, 2020). UK airports squandered £10,000 per minute between March and June (Independent, 2020). Since June, 40% of the aircraft has returned to the line and, total seat capacity increased by 32% compared to the previous month but remained 35% below the level (IATA, 2020). Direct impact on agriculture due to COVID-19 was less affected as compared to other sectors. The agricultural sector saw a price drop of 20% attributed to demand crash from restaurants and hotels during the COVID-19 outbreak (T. E. Times, 2020). Labour – intensive agricultural production systems got affected due to social distance and lockdown measures (OECD, 2020). Real time data collection from ship tracking before and after lock down showed disruption was not that much bad as it was expected and countries, which had strong trade link with China suffered lot (Verschuur et al., 2021). During COVID-19 pandemic (March ~ April), facemasks and N95 respirators became a worldwide healthcare necessity causing a shortage of supply in various countries along with rising the prices exorbitantly. Consecutively, industries have changed their production process based on high demand. Non-renewable and biodegradable petroleum and polymer-based materials like polypropylene, polystyrene, polycarbonate, polyethylene, and polyesters were used for the production of environment-friendly masks to fight COVID-19 and pollution (Das et al., 2020). Production line of various companies altered due to change of demand and for corporate social responsibility: Ford automotive industry (vehicles to modified respiratory and ventilator), Tesla (Electrical vehicles to ventilators), Airbus (aerospace to ventilators), Dyson Tech (Vacuum cleaners to hand dryers), Ventilators Ineos (Oil, gas, plastics Chemicals and other products to Hand sanitizer and other healthcare products), Gucci (Luxury clothing to Masks), and Zara (Apparel to Surgical masks). In the economy, consumers play a significant role. Panic buying of household items (e.g., toilet paper, groceries) at the starting of this pandemic increased and, suppliers were not ready to meet the demands. Hence, the supply-demand chain disrupted, and soon restriction on the maximum purchase was enforced. Globalization might suffer due to current trends (He and Harris, 2020). Governments prepared emergency plans, and compensation packages to support their economies. The UK offered £330 billion as an emergency loan to help those in financial difficulty, the People Bank of China and the Bank of Japan granted $240 billion and $43 billion for maintaining bank movement, respectively (Nicola et al., 2020) Germany offered unlimited loans to protect companies from collapsing. Small and large businesses were provided with loans to protect their employees, which affected the tax system. In Denmark, 75% of wage bills were covered by the government, which helped companies to struggle against the drop in the economy and, employees were entitled to take five days' leave from work (Forum, 2020). Social distances created problems, particularly in the industries, where contactless working is difficult. Remote working facilities like digital seminar conferences, work from home concept gained importance due to lockdown measures, which remained a challenge in populated cities: India (Population: 1.38 billion), China (Population: 1.43 billion), Singapore (Population density: 8358/km2) and Hong Kong (Population density: 6754/km2). By October 2020, US$12.7 trillion was committed by G20 countries to recover the economic down from COVID-19. However, only US$3.7 trillion was directed to environment and carbon emissions sector (Griffiths et al., 2021). It is evident that trade losses during lockdown was higher and it is recommended that shorter but strict rules could minimize the overall losses. Lifting up the restriction with go slow approach is only valid if further lockdowns are avoided (Guan et al., 2020). However the economic down will force to slow down to attaint he SDG goal. Previously because of the recession, global investors were less interested to invest in SDG, which implies that the achievements of the SDGs in the post Covid-19 is fully dependent on the government support (Shulla et al., 2021). To promote investment towards SDGs, investment institution should look for the optimal portfolio allocation. Ellen MacArthur Foundation also batted the circular economy as the key to creating resilient supply chains after the COVID-19 pandemic. Government must fund long-term green recovery policies as well as short-term emergency packages. 4.4 Impact on energy Energy is essential commodity for poverty reduction and economic growth. Energy security for a nation is important to be interdependent in international scenario (Le and Nguyen, 2019). Nation must supply adequate energy in affordable and reliable price to the people. Energy security is the most popular term in 20th century because of the need to decarbonise the energy sector, associated gas supplies issues in Europe and increase in demand in Asia. From 1960s this term is well known and because of the oil crisis in 1970 it became most popular term (Cherp and Jewell, 2014). To ensure the energy security global energy trade is most common fashion nowadays however it fully depends on any countries own strategic decision (Sutrisno et al., 2021). Understanding the necessity and importance of it, UN has made clear that affordable and clean energy should be one of SDG. Even though the energy sector has a strong correlation with the environment and economy, it remained the most ignored area by the researchers, who only investigated the economic, social, and environmental impact during the pandemic (Henry et al., 2020). In the current world of modernization and urbanization, uninterrupted energy and electrical power supply was a great boom to society, contributing to continuous work from home facilities and balancing the consumption demand, which was otherwise reduced by the industrial sectors (Mastropietro et al., 2020). Energy consumption and demand pattern were different in different countries based on lockdown strictness, adopted measures, and industry closures (Bahmanyar et al., 2020). The reduction of global primary energy demand from 5% to 52% between mid –March and the end of April 2020, shrunk the global economy 4.4% in the same year (Griffiths et al., 2021). Previously, electricity demand from the residential sector was dominant only during Sundays, which now became a daily scenario (Abu-rayash and Dincer, 2020). In New York City, overall industrial and commercial energy consumption decreased ~7% while domestic household consumption increased ~23% in March and ~10% in April 2020. In the UK, 30% and in the USA, 20% increment of electricity consumption was experienced during the middle of the day (9 a.m.–5 p.m.) in lock down period (Rouleau and Gosselin, 2021). Building energy consumption reduction now in highest priority globally by employing advanced building envelop (Ghosh et al., 2015) but this pandemic clearly showed the essence to take this matter seriously (Nundy et al., 2021). Energy insecurity and inability to pay the extra utility bills previously created tremendous issues among low to middle income family (Graff and Carley, 2020). Thus, it was expected that similar condition will rise during pandemic (Memmott et al., 2021). However, in reality it was not the case. Various measures were taken to reduce excessive energy bill (space heating and cooling load increased) in residential and abate the burden from the customers during lockdown (previously consumed 40% of the total energy), as listed in Table 4 (Qarnain et al., 2020). The measures were inclined more towards discounts on electricity bills rather than advising consumers to lessen consumption.Table 4 Details of electricity supply during COVID-19 global lockdown period. Table 4Argentina Starting from 24th March 2020, no disconnection of electricity services for non-payment of bills up to three consecutive bills or alternate bills (Buenos. Aires.Times, 2020) Australia At crisis time, no one would be deprived of electricity at a residential and commercial building. Financial support for the energy consumer (Energy.gov.au, 2020) Canada Uninterrupted power supply until pandemic exists. (The.Canadian.Press, 2020) China Price of electricity was reduced France from midnight of 16 March 2020 for a period of 30 days, all utility bill was Suspended for all business establishments SKWAWKBOX (2020) Germany Consumer having loss of income no need to pay electricity payments till 30 June 2020 (Germany.VISA, 2020) India The Indian government announced a three months moratorium for state-owned electricity distribution companies to make payments for their power purchased by them; it also reduced the payment security to 50% for future power purchases. Mint (2020) Indonesia Free electricity for poor people starting from 24 April 2020 (The.Jakarta.Post, 2020) Italy Until 30th April all electricity bills were suspended Williams (2020) Japan Japanese government requested all the electricity companies to present a bill on providing moratorium for bill electricity bill payment for 3 months. Malaysia The people of Sabah Province will get a 30% discount on Electricity bill for 3 months starting from 1 April 2020. Mail (2020) UK No power interruption for energy users, the energy supply is ensured with support and initiatives from the Government of the UK. (GOV.UK, 2020) USA No interruption during the prevalent pandemic times KSLA (2020) Full lockdowns decreased daily electricity demand by at least 15% (France, India, Italy, Spain, the United Kingdom, and the US northwest), which was later recovered after the ease of confinement (April ~ May), and later by June, electricity demand, decreased >10%, except in India, where the recovery is more pronounced (Fig. 10 ). Thus, during the lockdown, renewable energy got attention to overcome the electricity demand. In May, renewables have strengthened their second position after natural gas. In India, coal energy and renewables managed to acquire a significantly equal position after the first lockdown, and thus, coal energy's share in the electricity mix stayed under 70%. In Germany, renewable energy penetration in the net electricity generation was above 55% in 2020 compared 47% in 2019 (Halbrügge et al., 2021). In late May, levels of electricity demand started recovering while the rising share of renewables in the mix reflected their seasonal availability. In late June, electricity demand grew with rising temperatures, where share of coal energy increased in the electricity mix while the share of wind energy decreased (IEA, 2020a). For energy industry, the change of energy demand and consumption pattern was damaging. In the USA, at least 19 energy companies bankrupted due to this change. Although the overall energy demand dropped, not only residential but medical industry consumed energy to produce medical products and personal protective equipment (Klemeš et al., 2020a). Change of the spatial and temporal distributions of energy consumption have shifted the peaks of electricity consumption. The reduced electricity demand created negative impact on the power generation from fossil fuels like coal. To maintain the grid dynamic energy generation by coal, oil and nuclear was reduced in favour of intermittent renewable sources (Werth et al., 2021). On the other hand, continuous development of vaccines can enhance energy consumption. At the beginning of the pandemic limited time was there for energy managers to deal with this pandemic and balance system for energy demand and supply (Jiang et al., 2021). More solutions should be discussed on the use of reusable masks, appealed to minimize the plastic waste, energy, and environmental footprints during and after the COVID-19 pandemic (Klemeš et al., 2020b). It was also reported that new construction for energy facilities faced challenges. Production and global delivery of solar panels wind turbines and batteries were in halt from China while India's 3000 MW RE installation faced serious slowdown from the lockdown (Zhang et al., 2021a).Fig. 10 Decrease in daily electricity demand during lockdown (118 days): France- Mar 14; Germany- Mar 15; India- Mar 18; Italy- Mar 4; Spain- Mar 9; UK- Mar 19 (IEA, 2020a).). Fig. 10 In summary, it is evident that during the lockdown closure of industry sectors reduced the demand for fossil fuel energy, which in turn improved the environment. Even though the energy demand was higher in the residential areas, fossil fuel engendering did not increase. Post COVID-19, to maintain renewable energy following the energy generation mix, all governments should create a strict energy policy, where enhancing subsidies in renewable energy can be one of the solutions (Akrofi and Antwi, 2020). The IEA expected that the net renewable energy expansion capacity would be 13% in 2020 compared to 2019. However, COVID-19 slowed down this pace. Subsidy on renewable energy could improve the situation. According to IEA, spend of US$1 trillion per year between 2020 and 2024, can improve the sustainable energy goal. The International Renewable Energy Agency (IRENA) has suggested between 2021 and 2023, spending US$2 trillion per year on clean energy and related infrastructure to address the global climate agenda. These steps can improve the scenario to attain UN's SDG 7, 11–13 goal. On the other hand, accurate prediction of daily energy demand must be in line to protect the national grid from future disturbances either from a pandemic or similar issues (Lu et al., 2021). 4.5 Impact on transport Human mobility occurs due to various reason, which can include travel for shopping, work, personnel essential services, military services. The spread of infectious disease has a direct relation to human mobility (Peak et al., 2018). Thus, travel restriction is indispensable during pandemic (Yan et al., 2018). Travel restrictions are generally stricter for travel from a medium to a high-risk area. The spread of COVID-19 accelerated due to different modes of public transport (Zhao et al., 2020). Spread from one country to another one occurred through the commercial air flight (Shen et al., 2020). It was found that near the airport (25 miles) had 1.392 times higher COVID-19 cases and 1.545 times higher deaths due to COVID-19 in comparison to places that are over 50 miles away from an airport (Gaskin et al., 2021). Domestic land travel was also full of risk for the pandemic. Thus to limit the spread, transport sector was under lockdown. Different countries adopted different degrees of restrictions to tackle and abate COVID-19 spread, which affected largely on peoples' lifestyles as explained in section 3.1. On the other hand, the pandemic, lockdown and travel restriction due to COVID-19 created long lasting damaging in transport sector. Impact on transport sector for the 2003 SARS epidemic and 2008 swine flu outbreak were less compared to 2019 COVID cases (Vickerman, 2021). Transportation, a non-ignorable part of daily life, suffered mobility restrictions due to the COVID-19 lockdown measure, which in turn reduced 57% of global oil demand. Additionally, automobiles also have a strong relation to the environment, as they produce environmental pollutants, which in turn decreased during the lockdown, as discussed in section 3.2. In the lockdown regions, road transport dropped between 50% and 75% (as shown in Fig. 11 ) while, with global average road transport activity fell to 50% as of the 2019 level by the end of March (IEA, 2020b). Countries that didn't impose strict lockdown, also faced revenue losses, as people avoided public transport (Kanda and Kivimaa, 2020). Public transport usages in Stockholm decreased to 60%, and 75% in Nashville and Chattanooga, TN, US (Jenelius and Cebecauer, 2020). Due to official stay at home in the USA, 7.87% of human mobility was reduced. The rise of the infection rate from 0% to 0.0003% reduced the 2.31% mobility rate. By the second week of March in 2020, Switzerland experienced a travel reduction of up to 60% within the country and female travellers were less compared to male travellers (Abdullah et al., 2020b). In Germany, the use of single user car increased from 53% before lockdown to 66% during lockdown (Eisenmann et al., 2021). The use of bicycle was also increased during the lock down (Przybylowski et al., 2021). In Tokyo Japan, it was observed that travel for leisure and eating out was reduced in greater order while outing for grocery shopping and other type of shopping was increased (Parady et al., 2020). However just seeing the present COVID-19 spread, it is evident that just travel restriction is not enough for limiting a pandemic. During the consideration of overall epidemic size, high- and low-risk communities should be identified with mobility restrictions to have effective results (Espinoza et al., 2020). Restrictions in the mobility of airlines also created economic downfall (Kraemer et al., 2020), as discussed in section 3.3. Probably one of the positive aspects that occurred due to the lockdown in the transport sector is less incident of accident. According to WHO, around the world, 1.35 million people are killed in a road accident and 50 million injuries, which is a huge loss in terms of material damage and economic point of view of a nation. It is also reported that traffic accident was reduced significantly during the lockdown and due to pandemic in urban (Qureshi et al., 2020), suburban (Saladié et al., 2020) and rural (Zhang et al., 2021b) areas. Human error, bad weather and visibility, road characteristics, vehicle design, are the among the main factors behind traffic accidents (Retallack and Ostendorf, 2019). However, due to less traffic (Inada et al., 2021) congestion irresponsible driving was also experienced (Meyer, 2020). To regain the economy, overall growth in the transport sector is necessary. Mobility improves the access to good employment. UN 2030 has sustainable transport is an agenda. For urban areas, sustainable public transport is crucial. Thus, positive stand, translated to a frequent yet safe use of public transport while maintaining safe distance between travellers are required as much as possible.Fig. 11 Automobile usage during and post COVID-19 lockdown (image courtesy and source(IEA, 2020b):). Fig. 11 Thus, post-COVID-19 and lockdown, responsible transport choices will play a crucial role in daily life. Manageable work from home or necessary physical travel will be a new question in life, along with the selection of transport with the lowest environmental and social impact (Budd and Ison, 2020). Studies already showed that travellers who own car are now less likely to use public transport (Li et al., 2021). It will be critical if, after the pandemic, people just prefer the private transport over public transport. This will increase the no of car on the road, which in turn increase the traffic congestion and also if those cars are fossil fuel driven then the environment will be in danger again. In addition, both these scenarios are against the UN SDG plan. Thus, choice of transport and behaviour of public will play a crucial role. However, it is seen that travel pattern after COVID-19 is a complicated topic for travel policymakers. A mixed outcome was observed from different countries study regarding the travel pattern of people. Studies in the city of Gdańsk, in Poland, showed that almost 75% of respondent eager to use public transport, once the epidemic is stabilized and the rest are completely lost hope regarding the safe use of public transport ever (Przybylowski et al., 2021). In the Netherlands, data from 2500 respondents showed that 80% of people limited or reduced their outdoor activities while 44% of workers are now working from home to avoid transport (Haas et al., 2020). Cycling gained importance in different countries, as the most suitable road transport (De Vos, 2020). It is interesting that this mode of transport is not part of UN's SDG. Bogota, the capital of Colombia, changed its 100 Km bus lane to cycle lane, Berlin expanded yellow tapes marks to allow cycle movements. Mexico City aspires to quadruple its cycle lane capacity, Canadian city in Vancouver restricted vehicles inside Stanley Park (Source: World economic forum), Budapest introduced cycle lanes and lowered up to 300% of tariff for a cycle (Bucsky, 2020). In addition, in the UK, the government is supporting to use of more cycles. However, it will be a real problem for a crowded and overpopulated city like London, where every 15 min, more than 325,000 people use underground (BBC, 2020f). Changing the transport policy with the inclusion of electric vehicles (EV) could be the future boom (Bhattacharjee et al., 2020) and part of UN sustainable transport goal. In 2010, 17000 EVs were on the global road while it reached to 7.2 million by 2019 (Ghosh, 2020a). Most of the countries around the globe have the plan to reach EV by 2050, which can be modified to achieve targets. Grid disturbances, which was observed in section 3.4 due to reduction of energy demand, could have create same issues if electrification is increased in transport sector (Peng et al., 2021). After pandemic transport sector will be in the high discussion as different factors are associate with it. Staying at home and less use of transport obviously an environmentally benign solution but these could drastically damage the transport industry. On the other hand, the enhancement of the private vehicle will definitely increase the potential risk of traffic and accident. In addition, electrification of transport sector can increase the demand for energy, which still rely on the fossil fuel and negatively associated with the UN SDG target. 5 Discussion It is clearly visible that potential danger from COVID-19 occurred because of the underestimation of previous public health crises including the 1918 Spanish Flu, SARS in 2003, Zika fever in 2005 and 2016, H1N1 influenza virus in 2009, Ebola in 2014. Human society has not learned much from the past pandemics. Undoubtedly, social/human, environment, economy, energy and transport are the major segments of society, which got immensely affected due to this COVID-19 pandemic. Responses to the impact of coronavirus was very different for different countries, for e.g. while South Korea implemented testing facility, Italy, the U.K., and the USA suffered huge losses. It is also analyzed that reduction of mortality could save 40.76 trillion USD globally (Yoo and Managi, 2020). Impact of this virus was so dreadful, that even vaccine has been developed within a year in comparison to other diseases (Mahase, 2020), where, generally it takes a long time. For example, it took about 40 years for polio vaccine, 5 years for Ebola, and an average of 15 years for most vaccines development (Wibawa, 2020). COVID-19 lockdowns decreased carbon emissions from top three greenhouse gas emitters in the world: China, the EU, and the USA. A broader range of environmental benefits were obtained from cleaner air, reduced air travel and vehicle traffic, shipping manufacturing, and other activities. It is noteworthy that no such world event in the 20th century could decreased the global environment pollutant emission significantly at such a level in comparison to COVID-19 (Perkins et al., 2020). Thus, COVID-19 pandemic inadvertently minimized emissions more than any individual action, policy, or intervention to date, also aligning with the Climate Action Sustainable Development Goal targets of holding warming below 1.5 °C above preindustrial levels (Perkins et al., 2020). However, containing people at home is against the sustainable living. Sustainable society and cities which were growing rapidly, now stopped due to COVID-19. It is evident that social distances are key to trim down the spread of the virus, thus, densely populated cities became unsuitable for sustainable living (Ghosh et al., 2020b), which led to focus us more about uncontrolled global urbanization (Liu, 2020). Human health is strongly related to the economy and if economy is not protected due to the COVID-19 scenario world will face tragic health issue (McKee and Stuckler, 2020). One of the most challenging areas which needs to be improved is the building sector (Pinheiro and Luís, 2020). People spend 90% of their time indoors, which includes bedroom, office room, gym, movie hall, shopping mall (Ghosh and Norton, 2019). Now though they will stay inside more than ever only in their own home. Thus, building indoor environment quality assessment is in high demand. Indoor Environmental Quality (IEQ) is crucial parameters, which quantify the quality of a building's environment in terms of health and well-being of buildings occupants. IEQ is a combined factor of thermal comfort (Ghosh et al., 2018a), visual comfort (Ghosh et al., 2021), interior light (Ghosh and Norton, 2017a) and air temperature (Hemaida et al., 2020), psychosocial impact (Pollard et al., 2021). However, there is no standard is available which can assesses the occupants health. IEQ directly affects the comfort and well-being of occupants (Aggarwal et al., 2020). Compromised IEQ possess increased risk for diseases which are exacerbated by both socio-economic factors (Awada et al., 2021). During this pandemic while people are working from maintaining the IEQ becoming more critical (Wang et al., 2021). High occupant density in buildings increase the risk of virus transmission and also enhance the energy consumption. It was evaluated that an optimal distribution of occupant in a building can decrease 56% of infection rates and 32% of energy consumption. Though further study is required to consider and validate this claim (Mokhtari and Jahangir, 2021). Thus, it is still unclear how the antivirus-built environment would look (Megahed and Ghoneim, 2020). According to WHO, COVID-19 can be transmitted by air and dangerous for closed environment (Greenhalghu et al., 2021). Therefore, dilution ventilation, correct direction of airflow, pressure differential, etc. offered by the well-maintained HVAC system could effectively mitigate the risk of COVID-19 transmission. ASHRAE even stated two statements officially opposing the advice not to run residential or commercial HVAC systems. HVAC-related institutions including the Architectural Society of China, the Chinese Association of Refrigeration, the American Society of Heating, Refrigeration, and Air Conditioning Engineers, the Federation of European of Heating, Ventilation, Air-conditioning Associations, and the Society of Heating, Air-Conditioning and Sanitary Engineering in Japan have all issued documents in response to COVID-19 (Guo et al., 2021). ASHRAE pointed out that staying away from crowded and poorly ventilated areas may help reduce infection risk. Previously, to minimize the risk of infectious diseases, interior design, architecture, cities, and infrastructure were redesigned. The pandemic has highlighted the lack of how we manage our built environment and presented certain lessons from this forced experiment, alerting the architects, planners, and policymakers to react wisely (Ghosh, 2020b) and think about how post-pandemic housing and office spaces or any indoor built environment should look like (Megahed and Ghoneim, 2020). For city, urban heat island is a critical issue (Ghosh et al., 2016a), which is caused by anthropogenic heat emission, changes in urban land use and heat emission from buildings (Ghosh et al., 2016b). It is estimated that COVID-19 restrictions resulted a 0.13 °C temperature reduction in urban areas which improved the urban heat island issue (Nakajima et al., 2021). Strong sustainability by preserving natural resources and not exploiting them for financial benefits could be the great lessons from this pandemic. Hence, energy sector should now lean more towards renewable energy resources. Major energy consumption sectors are building, industry and transport. As building energy demand increased during lockdown and even after lockdown the work from home culture increased the energy demand, integration renewable energy technology in the building sector now should be in high demand (Krarti and Aldubyan, 2021). Integration of photovoltaics in to the building as a form of building integrated Photovoltaic (BIPV) (Reddy et al., 2020) can be considered which can generate onsite electricity and replace the traditional roofs, walls, windows (Alrashidi et al., 2020b), as well as shading devices (Mesloub and Ghosh, 2020). PV shading devices already showed potential suitability for hot and desert climate (Mesloub et al., 2020). Exploitation of solar energy can further be improved by employing solar water heating, solar thermal cooling (Khalid et al., 2021). Building wall can be improved by employing phase change material (PCM) (Roy et al., 2020a) or PCM-aerogel (Buratti et al., 2021) or PCM terracotta brick wall (Chelliah et al., 2021). PCM has latent heat storage capacity which is beneficial for colder climate and also to shift the peak of higher temperature from mid-day to afternoon time (Karthick et al., 2020). Building windows play crucial role to manage the overall energy performance of a building (Ghosh et al., 2019a). Thus, replacing the traditional window with advanced or smart window technology as per their climatic requirement can be a good option (Ghosh et al., 2019b). Energy efficient building windows are either highly insulated static transparent or solar hat control switchable smart window (Ghosh et al., 2016c). Currently smart windows are operated by either electrical, optical or thermal actuation (Ghosh and Mallick, 2018b). Electrically activated smart windows (Ghosh and Mallick, 2018a) are particularly interesting because by changing its transmission state it can tune the incoming solar radiation depends on the occupants comfort and also can trim down the overall building energy demand (Ghosh and Norton, 2018). At present electrically activated smart windows includes suspended particle device (SPD) (Ghosh and Norton, 2017b), electrochromic (EC), and polymer dispersed liquid crystal (PDLC) (Ghosh and Mallick, 2018b). For colder climate, replacing single or double glazing window with vacuum glazing will be beneficial as they have lower heat loss (Ghosh et al., 2017a). As vacuum glazing limit or reduce the conductive, convective and radiative heat flow, it has 53% higher heating load demand reduction potential over double glazing (Ghosh et al., 2016d). As vacuum glazing has similar transmission (Ghosh et al., 2017b), it is possible to make it switchable which will have controllable transmission and low heat loss capacity (Nundy and Ghosh, 2020). Employing more smart window technology to reduce building energy consumption (Ghosh et al., 2018b) and generation of energy from benign sources should get more priority (Ghosh et al., 2018c). Solar energy which is one of the most promising renewable energy sources can get more priorities and should experience the benefit from cleaner environment (Ghosh et al., 2020a). To utilize solar energy solar cells are the key technologies (Alrashidi et al., 2020a). At present solar cell technologies includes first generation crystalline silicon based, second generation thin film CdTe (Alrashidi et al., 2019), CIGS and a-Si, and third generation DSSC (Selvaraj et al., 2019), Perovskite (Bhandari et al., 2019) and carbon (Roy et al., 2020b) types. At present, first and second generations are commercially available. However because of the potential, research should be more focus on third generation (Bhandari et al., 2020). Globally because of air pollution and suspend particle in the air and deposition of them on photovoltaic (PV) technology reduces the power generation from the PV which is lower than its rated values (Ghosh, 2020c). Globally depends on the location, this deposition varies from 5% 70% (Chanchangi et al., 2020a). Cleaning those PV system is essential to keep the power generation high either by using manual (Chanchangi et al., 2020b) or anti-soiling coating (Nundy et al., 2020). Cleaner environment will allow more sunlight to incident on PV panel to generate more power (Naderipour et al., 2020). In the future, the transport sector stands as a real question. Currently work from home (WFH) concept is sufficiently popular and also it was found that productivity has been increased significantly (Hensher, 2020). Technological firms are happy to maintain this WFH and most of the Universities have been developed online teaching. Also it is expected that people will choose such job which will allow them to work from home (Junyi Zhang et al., 2021c). Now, environmentally this is a good option as IC engine based vehicles will be less on the road and there will be possible improvement of the transport sector (Chen et al., 2021). However, this causes a serious revenue loss from transport sector and resulting in associated job losses. Airline industry faced around US$250 billion revenue losses in 2020 (Amankwah-Amoah, 2020). It is very hard for some firms to sustain financially by maintaining environmental sustainability orientation. For connectivity point of view, several unprofitable routes will be closed and remain unknown when they will open to meet the standard as in 2019. Thus, government support is now the pragmatic approach (Abate et al., 2020). On the other hand, to improve the public transport and increase the crowd, fare-free public transport policies can be introduced. This policy is not uncommon and visible in Estonia, some remote location in China and the USA (Hess, 2017). Recently after pandemic, three cities in China, Hangzhou, Ningbo, and Xiamen, implemented fare-free policies to attract passengers back to public transport (Dai et al., 2021). Sustainable transport should get more priority to maintain a cleaner environment which was achieved during pandemic lockdown period (Shokouhyar et al., 2021). Also as people will use less transport hence penetration of electric vehicles will be easier now to change the human habit and also the charging facilities will be easier now because of the less number of Car on the street (Basu and Ferreira, 2021). In July 2020, UN recognized that digitization and virtual delivery is now essential because of the COVID-19. Though, school students are disappointed from online learning experiences due to lack of informal student engagement opportunities (Piyatamrong et al., 2021), still, undoubtedly this digital technology helped to maintain the education sector. During social isolation lockdown condition, digital space became the necessity for economic, educational, and leisure activities as well as for social interactions. Digital inequalities were prominent between developed and developing country and also between urban and rural areas (Beaunoyer et al., 2020). The pandemic highlights the importance of distributing smaller units such as health facilities, schools, and services across more of the urban tissue and strengthen local centres. After this COVID -19, application of 6G will probably take high gear (Allam and Jones, 2021). Also, artificial intelligence and use of machine learning will be increased in research area (Chandran et al., 2021). An exit strategy from this quarantine and lockdown is essential (Peto et al., 2020). Sustainability, which is multidimensional and complex, has the potential to bring the different sector together with help of policymakers, government, and practice and habit change of common people. This pandemic jeopardized basic health, well-being, life quality, education and social needs which should now on high demand to attain the sustainable goal. Based on the UN report, 690 million people, which accounts of 8.9% of the population of the world were hungry before pandemic and this COVID-19 created the issues more critical to solve because of the overall damage to the economy. One analysis pointed out that because of the pandemic, 4 countries from Asia, 6 countries from Oceania, 10 countries from Latin America and 15 countries from Africa can face food issues because of the import dependency and can have an impact on SDG2 food security (Udmale et al., 2020). Imposing a green tax and innovation on industry could enhance the economy. However, it should be remembered that earning money from low subsidy on fossil fuel would only be fruitful when clean affordable alternative energy will be available for everybody. Otherwise, societal inequality will still be in the society and effort to achieve the SDGs will be hampered. For developing countries, it will be difficult to create policy aligned to SDGs after COVID-19 pandemic and grow towards all 17 SDGs. Without sacrificing the other SDGs, cost effective and policy and raising and saving revenue should be implemented may require additional investment fund. Subsidy swap from fossil fuel to clean energy to invest in rural areas, subsidy on irrigation for better water supply (Barbier et al., 2020), waste water and sanitation, and employment of carbon tax could be effective (Barbier and Burgess, 2020). Because of so many issue, there is a sought that whether UN should revise their practice (Degai and Petrov, 2021). However, a difference in opinion is also there which says that may be time to consider this pandemic a positive way, which can now trigger to achieve those SDGs promptly. UN, which was created to bring the world together with peace and save from war threaten, must now also think about the threat from pandemic (Naidoo and Fisher, 2020). We must understand now that no one is safe until everyone is safe (Tikkinen et al., 2020). 6 Conclusion In December 2019, one of the deadliest viruses in the last 100 years is reported. Because of its destructive nature, human life changed completely and made people confine themselves at home. Vaccination has just developed within a year to tackle this virus. However, in the beginning, lack of tackling methods of this COVID-19 transmission, the old methods of self-isolation, lockdown, and home confinement were employed in the play and still valid until vaccines are widely available. Because of the lockdown and total closure of the major industrial sectors, which created a dip in the economy, caused job losses, financial uncertainty, and probably a recession for the near future. Lockdown created tremendous change in the human mind and their social life. Domestic violence, self-harming was prominent which indicated that 24 h staying at home had a negative psychological impact on the human mind. Working from home and using immense technology is now more prominent. The industrial and transport sector was fully closed, which eventually improved the environment. The emission of air pollutants decreased due to the reduction of fossil fuel power generation as industry and transport sector consumption was low. It is expected that to reach the pre COVID-19 situation considerable amount of time will be taken. Particular attention should be given to building an indoor environment, as working from home is very popular. With the presence of so many different negative aspects due to COVID-19 pandemic, SDGs are expected to get a much longer time to achieve. 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 Abate M. Christidis P. Purwanto A.J. Government support to airlines in the aftermath of the COVID-19 pandemic J. Air Transport. Manag. 89 2020 101931 10.1016/j.jairtraman.2020.101931 Abdullah M. Dias C. Muley D. Transportation Research Interdisciplinary Perspectives Exploring the impacts of COVID-19 on travel behavior and mode preferences Transp. Res. Interdiscip. Perspect. 8 2020 100255 10.1016/j.trip.2020.100255 Abdullah S. Mansor A.A. Napi N.N.L.M. Mansor W.N.W. Ahmed A.N. Ismail M. Ramly Z.T.A. Air quality status during 2020 Malaysia Movement Control Order (MCO) due to 2019 novel coronavirus (2019-nCoV) pandemic Sci. Total Environ. 729 2020 139022 10.1016/j.scitotenv.2020.139022 Abu-rayash A. Dincer I. Analysis of the electricity demand trends amidst the COVID-19 coronavirus analysis of the electricity demand trends amidst the COVID-19 coronavirus Energy Res. Soc. Sci. 2020 101682 10.1016/j.erss.2020.101682 Adams M.D. Air pollution in Ontario, Canada during the COVID-19 state of emergency Sci. Total Environ. 742 2020 140516 10.1016/j.scitotenv.2020.140516 Aggarwal V. Meena C.S. Kumar Ashok Alam T. Kumar Anuj Ghosh Arijit Ghosh Aritra Potential and future prospects of geothermal energy in space conditioning of buildings: India and worldwide review Sustain. Times 12 2020 1 19 10.3390/su12208428 Ahimi A.E. Ahimi E. Soheil D. Ani Esfa A. Ni H.A. Ebrahimi S. COVID-19 could change medical education curriculum J. Adv. Med. Educ. Prof. Adv. Med. Educ. Prof. 8 2020 144 145 10.30476/jamp.2020.86090.1217.Received Ahmadi M. Sharifi A. Dorosti S. Jafarzadeh Ghoushchi S. Ghanbari N. Investigation of effective climatology parameters on COVID-19 outbreak in Iran Sci. Total Environ. 729 2020 10.1016/j.scitotenv.2020.138705 Akrofi M.M.C. Antwi S.H. COVID-19 energy sector responses in Africa: a review of preliminary government interventions Energy Res. Soc. Sci. 68 2020 10.1016/j.erss.2020.101681 Allam Z. Jones D.S. Future (post-COVID) digital, smart and sustainable cities in the wake of 6G: digital twins, immersive realities and new urban economies Land Use Pol. 101 2021 105201 10.1016/j.landusepol.2020.105201 Alrashidi H. Ghosh A. Issa W. Sellami N. Mallick T.K. Sundaram S. Evaluation of solar factor using spectral analysis for CdTe photovoltaic glazing Mater. Lett. 237 2019 332 335 10.1016/j.matlet.2018.11.128 Alrashidi H. Ghosh A. Issa W. Sellami N. Mallick T.K. Sundaram S. Thermal performance of semitransparent CdTe BIPV window at temperate climate Sol. Energy 195 2020 536 543 10.1016/j.solener.2019.11.084 Alrashidi H. Issa W. Sellami N. Ghosh A. Mallick T.K. Sundaram S. Performance assessment of cadmium telluride-based semi-transparent glazing for power saving in façade buildings Energy Build. 215 2020 109585 10.1016/j.enbuild.2019.109585 Amankwah-Amoah J. Stepping up and stepping out of COVID-19: new challenges for environmental sustainability policies in the global airline industry J. Clean. Prod. 271 2020 123000 10.1016/j.jclepro.2020.123000 Auler A.C. Cássaro F.A.M. da Silva V.O. Pires L.F. Evidence that high temperatures and intermediate relative humidity might favor the spread of COVID-19 in tropical climate: a case study for the most affected Brazilian cities Sci. Total Environ. 729 2020 10.1016/j.scitotenv.2020.139090 Awada M. Becerik-Gerber B. Hoque S. O'Neill Z. Pedrielli G. Wen J. Wu T. Ten questions concerning occupant health in buildings during normal operations and extreme events including the COVID-19 pandemic Build. Environ. 188 2021 107480 10.1016/j.buildenv.2020.107480 Baghizadeh Fini M. What dentists need to know about COVID-19 Oral Oncol. 105 2020 104741 10.1016/j.oraloncology.2020.104741 Bahmanyar A. Estebsari A. Ernst D. The impact of different COVID-19 containment measures on electricity consumption in Europe Energy Res. Soc. Sci. 68 2020 101683 10.1016/j.erss.2020.101683 Bai L. Yang D. Wang Xun Tong L. Zhu X. Zhong N. Bai C. Powell C.A. Chen R. Zhou J. Song Y. Zhou X. Zhu H. Han B. Li Q. Shi G. Li S. Wang C. Qiu Z. Zhang Y. Xu Y. Liu J. Zhang D. Wu C. Li J. Yu J. Wang J. Dong C. Wang Yaoli Wang Q. Zhang L. Zhang M. Ma X. Zhao L. Yu W. Xu T. Jin Y. Wang Xiongbiao Wang Yuehong Jiang Y. Chen H. Xiao K. Zhang X. Song Z. Zhang Z. Wu X. Sun J. Shen Y. Ye M. Tu C. Jiang J. Yu H. Tan F. Chinese experts' consensus on the Internet of Things-aided diagnosis and treatment of coronavirus disease 2019 (COVID-19) Clin. eHealth 3 2020 7 15 10.1016/j.ceh.2020.03.001 Baldasano J.M. COVID-19 lockdown effects on air quality by NO2 in the cities of Barcelona and Madrid (Spain) Sci. Total Environ. 741 2020 140353 10.1016/j.scitotenv.2020.140353 Bao R. Zhang A. Does lockdown reduce air pollution? Evidence from 44 cities in northern China Sci. Total Environ. 731 2020 139052 10.1016/j.scitotenv.2020.139052 Barbier E.B. Burgess J.C. Sustainability and development after COVID-19 World Dev. 135 2020 105082 10.1016/j.worlddev.2020.105082 Barbier E.B. Lozano R. Rodríguez C.M. Troëng S. Tropical forests Nature 578 2020 213 216 32051595 Bashir M.F. Ma B. Bilal Komal B. Bashir M.A. Tan D. Bashir M. Correlation between climate indicators and COVID-19 pandemic in New York, USA Sci. Total Environ. 728 2020 138835 10.1016/j.scitotenv.2020.138835 Basu R. Ferreira J. Sustainable mobility in auto-dominated Metro Boston: challenges and opportunities post-COVID-19 Transport Pol. 103 2021 197 210 10.1016/j.tranpol.2021.01.006 BBC South Africa Coronavirus Variant: what Is the Risk? 2020 BBC Tokyo 2020: Olympic and Paralympic Games Postponed Because of Coronavirus 2020 https://www.bbc.co.uk/sport/olympics/52020134 BBC Coronavirus: A Visual Guide to the Economic Impact 2020 https://www.bbc.co.uk/news/business-51706225 BBC Coronavirus: Job Cuts Warning as 600,000 Roles Go in Lockdown 2020 https://www.bbc.co.uk/news/business-53060529 BBC Coronavirus Lockdown: India Jobless Numbers Cross 120 Million in April 2020 https://www.bbc.co.uk/news/world-asia-india-52559324 BBC Coronavirus: How Will Transport Need to Change? 2020 https://www.bbc.co.uk/news/explainers-52534135 Beaunoyer E. Dupéré S. Guitton M.J. COVID-19 and digital inequalities: reciprocal impacts and mitigation strategies Comput. Hum. Behav. 111 2020 10.1016/j.chb.2020.106424 Bendau A. Lydia S. Wyka S. Bruno M. Plag J. Asselmann E. Str A. Longitudinal changes of anxiety and depressive symptoms during the COVID-19 pandemic in Germany : the role of pre-existing anxiety , depressive , and other mental disorders J. Anxiety Disord. 79 2021 10.1016/j.janxdis.2021.102377 Berman J.D. Ebisu K. Changes in U.S. air pollution during the COVID-19 pandemic Sci. Total Environ. 739 2020 139864 10.1016/j.scitotenv.2020.139864 Bhandari S. Roy A. Ghosh A. Mallick T.K. Sundaram S. Performance of WO 3 -incorporated carbon electrodes for ambient mesoscopic perovskite solar cells ACS Omega 5 2019 422 429 10.1021/acsomega.9b02934 31956789 Bhandari S. Roy A. Ghosh A. Mallick T.K. Sundaram S. Perceiving the temperature coefficients of carbon-based perovskite solar cells Sustain. Energy Fuels 4 2020 6283 6298 10.1039/d0se00782j Bhattacharjee A. Mohanty R.K. Ghosh A. Design of an optimized thermal management system for Li-ion batteries under different discharging conditions Energies 13 2020 5695 10.3390/en13215695 Biktasheva I.V. Role of a habitat's air humidity in Covid-19 mortality Sci. Total Environ. 736 2020 138763 10.1016/j.scitotenv.2020.138763 Boserup B. McKenney M. Elkbuli A. Alarming trends in US domestic violence during the COVID-19 pandemic Am. J. Emerg. Med. 91 2020 3 5 10.1016/j.ajem.2020.04.077 Bradbury-Jones C. Isham L. The pandemic paradox: the consequences of COVID-19 on domestic violence J. Clin. Nurs. 29 2020 2047 2049 10.1111/jocn.15296 32281158 Braga F. Scarpa G.M. Brando V.E. Manfè G. Zaggia L. COVID-19 lockdown measures reveal human impact on water transparency in the Venice Lagoon Sci. Total Environ. 736 2020 10.1016/j.scitotenv.2020.139612 Briz-Redón Á. Serrano-Aroca Á. A spatio-temporal analysis for exploring the effect of temperature on COVID-19 early evolution in Spain Sci. Total Environ. 728 2020 10.1016/j.scitotenv.2020.138811 Brook B. Coronavirus Social Distancing: Why Distance Differs Around the World 2020 new.com.au Brown S.M. Doom J.R. Lechuga-Peña S. Watamura S.E. Koppels T. Stress and parenting during the global COVID-19 pandemic Child Abus. Negl. 110 2020 10.1016/j.chiabu.2020.104699 Bucsky P. Modal share changes due to COVID-19: the case of Budapest Transp. Res. Interdiscip. Perspect. 2020 100141 10.1016/j.trip.2020.100141 Budd L. Ison S. Responsible Transport: a post-COVID agenda for transport policy and practice Transp. Res. Interdiscip. Perspect. 6 2020 100151 10.1016/j.trip.2020.100151 Buenos. Aires.Times Government decree blocks providers from cutting utilities for three months [WWW Document] https://www.batimes.com.ar/news/economy/government-decree-blocks-providers-from-cutting-utilities-for-three-months.phtml 2020 Bukuluki P. Mwenyango H. Peter S. Sidhva D. Palattiyil G. Social Sciences & Humanities Open the socio-economic and psychosocial impact of Covid-19 pandemic on urban refugees in Uganda Soc. Sci. Humanit. Open 2 2020 100045 10.1016/j.ssaho.2020.100045 Buratti C. Belloni E. Merli F. Zinzi M. Aerogel glazing systems for building applications: a review Energy Build. 231 2021 110587 10.1016/j.enbuild.2020.110587 Cao W. Fang Z. Hou G. Han M. Xu X. Dong J. The psychological impact of the COVID-19 epidemic on college students in China Psychiatr. Res. 287 2020 112934 10.1016/j.psychres.2020.112934 Cellini N. Canale N. Mioni G. Costa S. Changes in sleep pattern, sense of time and digital media use during COVID-19 lockdown in Italy J. Sleep Res. 1–5 2020 10.1111/jsr.13074 Chakraborty I. Maity P. COVID-19 outbreak: migration, effects on society, global environment and prevention Sci. Total Environ. 728 2020 138882 10.1016/j.scitotenv.2020.138882 Chanchangi Yusuf N. Ghosh A. Sundaram S. Mallick T.K. Dust and PV performance in Nigeria: a review Renew. Sustain. Energy Rev. 121 2020 109704 10.1016/j.rser.2020.109704 Chanchangi Yusuf N. Ghosh A. Sundaram S. Mallick T.K. An analytical indoor experimental study on the e ff ect of soiling on PV , focusing on dust properties and PV surface material Sol. Energy 203 2020 46 68 10.1016/j.solener.2020.03.089 Chandran V. Patil C.K. Karthick A. Ganeshaperumal D. Rahim R. Ghosh A. State of charge estimation of lithium ‐ ion battery for electric vehicles using machine learning algorithms World Electr. Veh. J. 12 2021 38 Chaturvedi K. Vishwakarma D.K. Singh N. Children and Youth Services Review COVID-19 and its impact on education , social life and mental health of students : a survey. Child Youth Serv. Rev. 121 2021 105866 10.1016/j.childyouth.2020.105866 Chauhan A. Singh R.P. Decline in PM 2 . 5 concentrations over major cities around the world associated with COVID-19 Environ. Res. 187 2020 109634 10.1016/j.envres.2020.109634 Chelliah A. Saboor S. Ghosh A. Kontoleon K.J. Thermal behaviour analysis and cost-saving opportunities of PCM-integrated terracotta brick buildings Adv. Civ. Eng. 2021 2021 1 15 10.1155/2021/6670930 Chen N. Zhou M. Dong X. Qu J. Gong F. Han Y. Qiu Y. Wang J. Liu Y. Wei Y. Xia J. Yu T. Zhang X. Zhang L. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study Lancet 395 2020 507 513 10.1016/S0140-6736(20)30211-7 32007143 Chen N. Zhou M. Xuan D. Qu J. Gong F. Han Y. Qiu Y. Wang J. Yiu L. Wei Y. Xia J. Ting Y. Zhan X. Li Z. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in wuhan, China: a DescriptNanshan chen 1 , min Zhou 2 , Xuan dong 1 , Jieming Qu 2 , Fengyun Gong 3 , Yang han 4 , Yang Qiu 5 , Jingli wang 3 , ying liu 6 , Yuan Lancet 395 2020 507 513 32007143 Chen Z. Hao X. Zhang X. Chen F. Have traffic restrictions improved air quality? A shock from COVID-19 J. Clean. Prod. 279 2021 123622 10.1016/j.jclepro.2020.123622 Cherp A. Jewell J. The concept of energy security : beyond the four as Energy Pol. 75 2014 415 421 10.1016/j.enpol.2014.09.005 Cho S.Y. Kang J.M. Ha Y.E. Park G.E. Lee Ji Yeon Ko J.H. Lee Ji Yong Kim J.M. Kang C.I. Jo I.J. Ryu J.G. Choi J.R. Kim S. Huh H.J. Ki C.S. Kang E.S. Peck K.R. Dhong H.J. Song J.H. Chung D.R. Kim Y.J. MERS-CoV outbreak following a single patient exposure in an emergency room in South Korea: an epidemiological outbreak study Lancet 388 2016 994 1001 10.1016/S0140-6736(16)30623-7 27402381 Collivignarelli M.C. Abbà A. Bertanza G. Pedrazzani R. Ricciardi P. Carnevale Miino M. Lockdown for CoViD-2019 in Milan: what are the effects on air quality? Sci. Total Environ. 732 2020 1 9 10.1016/j.scitotenv.2020.139280 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 CREA Air quality improvements due to COVID 19 lock-down in India https://energyandcleanair.org/air-quality-improvements-due-to-covid-19-lock-down-in-india/ 2020 Dai J. Liu Z. Li R. Improving the subway attraction for the post-COVID-19 era: the role of fare-free public transport policy Transport Pol. 103 2021 21 30 10.1016/j.tranpol.2021.01.007 Das O. Neisiany R.E. Capezza A.J. Hedenqvist M.S. Försth M. Xu Q. Jiang L. Ji D. Ramakrishna S. The need for fully bio-based facemasks to counter coronavirus outbreaks: a perspective Sci. Total Environ. 736 2020 10.1016/j.scitotenv.2020.139611 De Vos J. The effect of COVID-19 and subsequent social distancing on travel behavior Transp. Res. Interdiscip. Perspect. 5 2020 100121 10.1016/j.trip.2020.100121 Degai T.S. Petrov A.N. Rethinking Arctic sustainable development agenda through indigenizing UN sustainable development goals sustainable development goals Int. J. Sustain. Dev. World Ecol. 2021 1 6 10.1080/13504509.2020.1868608 Doremalen N. Morris D. Holbrook M.G. Gamble A. Williamson B. Tamin A. Llyod-Smith J. Wit E.D. Aerosol and surface stability of SARS-CoV-2 as compared with SARS-CoV-1 N. Engl. J. Med. 2020 Dutheil F. Baker J.S. Navel V. COVID-19 as a factor influencing air pollution? Environ. Pollut. 263 2020 2019 2021 10.1016/j.envpol.2020.114466 Dutta V. Dubey D. Kumar S. Cleaning the River Ganga: impact of lockdown on water quality and future implications on river rejuvenation strategies Sci. Total Environ. 743 2020 140756 10.1016/j.scitotenv.2020.140756 Eisenmann C. Nobis C. Kolarova V. Lenz B. Winkler C. Transport mode use during the COVID-19 lockdown period in Germany : the car became more important , public transport lost ground Transport Pol. 103 2021 60 67 10.1016/j.tranpol.2021.01.012 Energy.gov.au Support for Australian households [WWW Document] https://www.energy.gov.au/energy-sector-response-novel-coronavirus-covid-19/information-australian-households 2020 Espinoza B. Castillo-chavez C. Perrings C. Mobility restrictions for the control of epidemics : when do they work ? PloS One 15 2020 1 14 10.1371/journal.pone.0235731 Fancourt D. Steptoe A. Bu F. Trajectories of anxiety and depressive symptoms during enforced isolation due to COVID-19 in England : a longitudinal observational study Lancet Psychiatr. 8 2021 141 149 10.1016/S2215-0366(20)30482-X Fattorini D. Regoli F. Role of the chronic air pollution levels in the Covid-19 outbreak risk in Italy Environ. Pollut. 264 2020 114732 10.1016/j.envpol.2020.114732 Fisher B. Time to revise the sustainable development goals Nature 583 2020 Forbes How COVID-19 is Transforming global aviation's outlook [WWW Document] https://www.forbes.com/sites/oliverwyman/2020/04/06/how-covid-19-is-transforming-global-aviations-outlook/ 2020 Forum W.E. This is how Europe is helping companies and workers as the coronavirus crisis deepens [WWW Document] https://www.weforum.org/agenda/2020/03/covid-19-quarantine-sick-pay/ 2020 Frontera A. Cianfanelli L. Vlachos K. Landoni G. Cremona G. Severe air pollution links to higher mortality in COVID-19 patients: the “double-hit” hypothesis J. Infect. 2020 10.1016/j.jinf.2020.05.031 Gaskin D.J. Zare H. Delarmente B.A. Geographic disparities in COVID-19 infections and deaths : the role of transportation Transport Pol. 102 2021 35 46 10.1016/j.tranpol.2020.12.001 Geography N. Pollution Made COVID-19 Worse. Now, Lockdowns Are Clearing the Air 2020 https://www.nationalgeographic.com/science/2020/04/pollution-made-the-pandemic-worse-but-lockdowns-clean-the-sky/ Germany.VISA Nine new rules affecting Foreigners in Germany as of April 2020 [WWW Document] https://www.germany-visa.org/news/nine-new-rules-affecting-foreigners-in-germany-as-of-april-2020/ 2020 Ghosh A. Possibilities and challenges for the inclusion of the electric vehicle ( EV ) to reduce the carbon footprint in the transport Sector : a review Energies 13 2020 2602 Ghosh A. Potential of building integrated and attached/applied photovoltaic (BIPV/BAPV) for adaptive less energy-hungry building's skin: a comprehensive Review J. Clean. Prod. 2020 123343 10.1016/j.jclepro.2020.123343 Ghosh A. Soiling Losses : a barrier for India ’ s energy security dependency from photovoltaic power Challenges 11 2020 1 9 10.3390/challe11010009 Ghosh A. Mallick T.K. Evaluation of optical properties and protection factors of a PDLC switchable glazing for low energy building integration Sol. Energy Mater. Sol. Cells 176 2018 391 396 10.1016/j.solmat.2017.10.026 Ghosh A. Mallick T.K. Evaluation of colour properties due to switching behaviour of a PDLC glazing for adaptive building integration Renew. Energy 120 2018 126 133 10.1016/j.renene.2017.12.094 Ghosh A. Norton B. Interior colour rendering of daylight transmitted through a suspended particle device switchable glazing Sol. Energy Mater. Sol. Cells 163 2017 218 223 10.1016/j.solmat.2017.01.041 Ghosh A. Norton B. Durability of switching behaviour after outdoor exposure for a suspended particle device switchable glazing Sol. Energy Mater. Sol. Cells 163 2017 178 184 10.1016/j.solmat.2017.01.036 Ghosh A. Norton B. Advances in switchable and highly insulating autonomous (self-powered) glazing systems for adaptive low energy buildings Renew. Energy 126 2018 1003 1031 10.1016/j.renene.2018.04.038 Ghosh A. Norton B. Optimization of PV powered SPD switchable glazing to minimise probability of loss of power supply Renew. Energy 131 2019 993 1001 10.1016/j.renene.2018.07.115 Ghosh A. Norton B. Duffy A. Measured overall heat transfer coefficient of a suspended particle device switchable glazing Appl. Energy 159 2015 362 369 10.1016/j.apenergy.2015.09.019 Ghosh A. Norton B. Duffy A. Daylighting performance and glare calculation of a suspended particle device switchable glazing Sol. Energy 132 2016 114 128 10.1016/j.solener.2016.02.051 Ghosh A. Norton B. Duffy A. First outdoor characterisation of a PV powered suspended particle device switchable glazing Sol. Energy Mater. Sol. Cells 157 2016 1 9 10.1016/j.solmat.2016.05.013 Ghosh A. Norton B. Duffy A. Behaviour of a SPD switchable glazing in an outdoor test cell with heat removal under varying weather conditions Appl. Energy 180 2016 695 706 10.1016/j.apenergy.2016.08.029 Ghosh A. Norton B. Duffy A. Measured thermal & daylight performance of an evacuated glazing using an outdoor test cell Appl. Energy 177 2016 196 203 10.1016/j.apenergy.2016.05.118 Ghosh A. Norton B. Duffy A. Effect of sky clearness index on transmission of evacuated (vacuum) glazing Renew. Energy 105 2017 160 166 10.1016/j.renene.2016.12.056 Ghosh A. Norton B. Duffy A. Effect of atmospheric transmittance on performance of adaptive SPD-vacuum switchable glazing Sol. Energy Mater. Sol. Cells 161 2017 424 431 10.1016/j.solmat.2016.12.022 Ghosh A. Norton B. Mallick T.K. Influence of atmospheric clearness on PDLC switchable glazing transmission Energy Build. 172 2018 257 264 10.1016/j.enbuild.2018.05.008 Ghosh A. Norton B. Mallick T.K. Daylight characteristics of a polymer dispersed liquid crystal switchable glazing Sol. Energy Mater. Sol. Cells 174 2018 572 576 10.1016/j.solmat.2017.09.047 Ghosh A. Sundaram S. Mallick T.K. Investigation of thermal and electrical performances of a combined semi- transparent PV-vacuum glazing Appl. Energy 228 2018 1591 1600 10.1016/j.apenergy.2018.07.040 Ghosh A. Sarmah N. Sundaram S. Mallick T.K. Numerical studies of thermal comfort for semi-transparent building integrated photovoltaic ( BIPV ) -vacuum glazing system Sol. Energy 190 2019 608 616 10.1016/j.solener.2019.08.049 Ghosh A. Sundaram S. Mallick T.K. Colour properties and glazing factors evaluation of multicrystalline based semi-transparent Photovoltaic-vacuum glazing for BIPV application Renew. Energy 131 2019 730 736 10.1016/j.renene.2018.07.088 Ghosh A. Bhandari S. Sundaram S. Mallick T.K. Carbon counter electrode mesoscopic ambient processed & characterised perovskite for adaptive BIPV fenestration Renew. Energy 145 2020 2151 2158 10.1016/j.renene.2019.07.119 Ghosh A. Nundy S. Ghosh S. Mallick T.K. Study of COVID-19 pandemic in London (UK) from urban context Cities 106 2020 102928 10.1016/j.cities.2020.102928 Ghosh A. Nundy S. Mallick T.K. How India is dealing with COVID-19 pandemic Sens. Int. 1 2020 100021 10.1016/j.sintl.2020.100021 Ghosh A. Mesloub A. Touahmia M. Ajmi M. Visual comfort analysis of semi-transparent perovskite based building integrated photovoltaic window for hot desert Energies 14 2021 1043 Gillingham K.T. Knittel C.R. Li J. Ovaere M. Reguant M. The short-run and long-run effects of covid-19 on energy and the environment Joule 4 2020 1337 1341 10.1016/j.joule.2020.06.010 32835174 GOV.UK Government agrees measures with energy industry to support vulnerable people through COVID-19 [WWW Document] https://www.gov.uk/government/news/government-agrees-measures-with-energy-industry-to-support-vulnerable-people-through-covid-19 2020 Graff M. Carley S. COVID-19 assistance needs to target energy insecurity Nat. Energy 5 2020 352 354 10.1038/s41560-020-0620-y Greenhalghu T. Jimenez J.L. Prather K.A. Tufekci Z. Fisman D. Schooley R. Ten scientific reasons in support of airborne transmission of Lancet 397 2021 1603 1605 10.1016/S0140-6736(21)00869-2 33865497 Griffiths M.D. Mamun M.A. COVID-19 suicidal behavior among couples and suicide pacts: case study evidence from press reports Psychiatr. Res. 289 2020 10.1016/j.psychres.2020.113105 Griffiths S. Del D.F. Sovacool B. Policy mixes to achieve sustainable mobility after the COVID-19 crisis Renew. Sustain. Energy Rev. 143 2021 110919 10.1016/j.rser.2021.110919 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. Hum. Behav. 4 2020 577 587 10.1038/s41562-020-0896-8 32493967 Guardian T. UK car industry ’could lose one in six jobs due to Covid-19 crisis [WWW Document] https://www.theguardian.com/business/2020/jun/23/uk-car-industry-jobs-due-covid-19-smmt-redundancies 2020 Guo M. Xu P. Xiao T. He R. Dai M. Miller S.L. Review and comparison of HVAC operation guidelines in different countries during the COVID-19 pandemic Build. Environ. 187 2021 107368 10.1016/j.buildenv.2020.107368 Gupta S. Raghuwanshi G.S. Chanda A. Effect of weather on COVID-19 spread in the US: a prediction model for India in 2020 Sci. Total Environ. 728 2020 138860 10.1016/j.scitotenv.2020.138860 Haas M. De Faber R. Hamersma M. Transportation Research Interdisciplinary Perspectives How COVID-19 and the Dutch ‘ intelligent lockdown ’ change activities , work and travel behaviour : evidence from longitudinal data in The Netherlands Transp. Res. Interdiscip. Perspect. 6 2020 100150 10.1016/j.trip.2020.100150 Hafeez S. Din M. Zia F. Ali M. Shinwari Z.K. Emerging concerns regarding COVID ‐ 19 ; second wave and new variant J. Med. Virol. 2021 10.1002/jmv.26979 Halbrügge S. Schott P. Weibelzahl M. Buhl H.U. Fridgen G. Schöpf M. How did the German and other European electricity systems react to the COVID-19 pandemic? Appl. Energy 285 2021 116370 10.1016/j.apenergy.2020.116370 Haleem A. Javaid M. Vaishya R. Effects of COVID-19 pandemic in daily life Curr. Med. Res. Pract. 10 2020 78 79 10.1016/j.cmrp.2020.03.011 32292804 He H. Harris L. The impact of Covid-19 pandemic on corporate social responsibility and marketing philosophy J. Bus. Res. 116 2020 176 182 10.1016/j.jbusres.2020.05.030 32457556 Heater B. Uber Is Laying off 3,700 as Rides Plummet Due to COVID-19 2020 https://techcrunch.com/2020/05/06/uber-is-laying-off-3700-as-rides-plummet-due-to-covid-19/ Hemaida A. Ghosh A. Sundaram S. Mallick T.K. Evaluation of thermal performance for a smart switchable adaptive polymer dispersed liquid crystal ( PDLC ) glazing Sol. Energy 195 2020 185 193 10.1016/j.solener.2019.11.024 Henry M.S. Bazilian M.D. Markuson C. Just transitions: histories and futures in a post-COVID world Energy Res. Soc. Sci. 68 2020 101668 10.1016/j.erss.2020.101668 Hensher D.A. What might Covid-19 mean for mobility as a service (MaaS)? Transp. Rev. 40 2020 551 556 10.1080/01441647.2020.1770487 Hess D.B. Decrypting fare-free public transport in Tallinn, Estonia Case Stud. Transp. Pol. 5 2017 690 698 10.1016/j.cstp.2017.10.002 Hodgkinson T. Andresen M.A. Show me a man or a woman alone and I’ll show you a saint: changes in the frequency of criminal incidents during the COVID-19 pandemic J. Crim. Justice 69 2020 101706 10.1016/j.jcrimjus.2020.101706 Huan Y. Liang T. Li H. Zhang C. Science of the Total Environment A systematic method for assessing progress of achieving sustainable development goals : a case study of 15 countries Sci. Total Environ. 752 2021 141875 10.1016/j.scitotenv.2020.141875 Huang Z. Huang J. Gu Q. Du P. Liang H. Dong Q. Optimal temperature zone for the dispersal of COVID-19 Sci. Total Environ. 736 2020 139487 10.1016/j.scitotenv.2020.139487 Iacus S.M. Natale F. Santamaria C. Spyratos S. Vespe M. Estimating and projecting air passenger traffic during the COVID-19 coronavirus outbreak and its socio-economic impact Saf. Sci. 129 2020 104791 10.1016/j.ssci.2020.104791 IATA Airline financial monitor https://www.iata.org/en/iata-repository/publications/economic-reports/airlines-financial-monitor---june-2020/ 2020 IEA Covid-19 Impact on Electricity 2020 Paris https://www.iea.org/reports/covid-19-impact-on-electricity IEA Evolution of Road Passenger Transport Activity in Selected Countries in Early 2020 2020 Paris https://www.iea.org/data-and-statistics/charts/evolution-of-road-passenger-transport-activity-in-selected-countries-in-early-2020 Inada H. Ashraf L. Campbell S. COVID-19 lockdown and fatal motor vehicle collisions due to speed- related traffic violations in Japan: a time- series study Inj. Prev. 27 2021 98 100 10.1136/injuryprev-2020-043947 33067222 Independent UK airports lost £10,000 per minute between MArch and June https://www.independent.co.uk/travel/news-and-advice/airports-money-loss-coronavirus-flights-uk-passengers-a9631241.html 2020 Ito H. Hanaoka S. Kawasaki T. The cruise industry and the COVID-19 outbreak Transp. Res. Interdiscip. Perspect. 5 2020 100136 10.1016/j.trip.2020.100136 Jackson R.B. Le Quéré C. Andrew R.M. Canadell J.G. Korsbakken J.I. Liu Z. Peters G.P. Zheng B. Global energy growth is outpacing decarbonization Environ. Res. Lett. 13 2018 10.1088/1748-9326/aaf303 Jahangiri Mehdi Jahangiri Milad Najafgholipour M. The sensitivity and specificity analyses of ambient temperature and population size on the transmission rate of the novel coronavirus (COVID-19) in different provinces of Iran Sci. Total Environ. 728 2020 138872 10.1016/j.scitotenv.2020.138872 Jenelius E. Cebecauer M. Impacts of COVID-19 on public transport ridership in Sweden: analysis of ticket validations, sales and passenger counts Transp. Res. Interdiscip. Perspect. 8 2020 100242 10.1016/j.trip.2020.100242 Jiang P. Fan Y. Van Klemeš J.J. Impacts of COVID-19 on energy demand and consumption: challenges, lessons and emerging opportunities Appl. Energy 285 2021 10.1016/j.apenergy.2021.116441 Kanda W. Kivimaa P. What opportunities could the COVID-19 outbreak offer for sustainability transitions research on electricity and mobility? Energy Res. Soc. Sci. 68 2020 101666 10.1016/j.erss.2020.101666 Karthick A. Ramanan P. Ghosh A. Stalin B. Kumar R.V. Baranilingesan I. Performance enhancement of copper indium diselenide photovoltaic module using inorganic phase change material Asia Pac. J. Chem. Eng. 2020 1 11 10.1002/apj.2480 Kenny P. UN Body Warns of up to 25M Job Losses Due to COVID-19 2020 https://www.aa.com.tr/en/economy/un-body-warns-of-up-to-25m-job-losses-due-to-covid-19/1771040 Kerimray A. Baimatova N. Ibragimova O.P. Bukenov B. Kenessov B. Plotitsyn P. Karaca F. Assessing air quality changes in large cities during COVID-19 lockdowns: the impacts of traffic-free urban conditions in Almaty Kazakhstan. Sci. Total Environ. 730 2020 139179 10.1016/j.scitotenv.2020.139179 Khalid M. Shanks K. Ghosh A. Tahir A. Sundaram S. Mallick T.K. Temperature regulation of concentrating photovoltaic window using argon gas and polymer dispersed liquid crystal fi lms Renew. Energy 164 2021 96 108 10.1016/j.renene.2020.09.069 Killgore W.D.S. Cloonan S.A. Taylor E.C. Dailey N.S. Loneliness: a signature mental health concern in the era of COVID-19 Psychiatr. Res. 290 2020 113117 10.1016/j.psychres.2020.113117 Kirk C.P. Rifkin L.S. I’ll trade you diamonds for toilet paper: consumer reacting, coping and adapting behaviors in the COVID-19 pandemic J. Bus. Res. 117 2020 124 131 10.1016/j.jbusres.2020.05.028 32834208 Klemeš J.J. Fan Y. Van Jiang P. The energy and environmental footprints of COVID-19 fighting measures – PPE, disinfection, supply chains Energy 211 2020 10.1016/j.energy.2020.118701 Klemeš J.J. Fan Y. Van Tan R.R. Jiang P. Minimising the present and future plastic waste, energy and environmental footprints related to COVID-19 Renew. Sustain. Energy Rev. 127 2020 10.1016/j.rser.2020.109883 Kraemer M.U.G. Yang C.H. Gutierrez B. Wu C.H. Klein B. Pigott D.M. The effect of human mobility and control measures on the COVID-19 epidemic in China Science 4218 80 2020 1 9 10.1126/science.abb4218 Krarti M. Aldubyan M. Review analysis of COVID-19 impact on electricity demand for residential buildings Renew. Sustain. Energy Rev. 143 2021 110888 10.1016/j.rser.2021.110888 KSLA Utility customers won't lose service during COVID-19 coronavirus emergency [WWW Document] https://www.ksla.com/2020/03/16/utility-customers-wont-lose-service-during-covid-coronavirus-emergency/ 2020 Lai C.C. Shih T.P. Ko W.C. Tang H.J. Hsueh P.R. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 (COVID-19): the epidemic and the challenges Int. J. Antimicrob. Agents 55 2020 105924 10.1016/j.ijantimicag.2020.105924 Lal P. Kumar A. Kumar S. Kumari S. Saikia P. Dayanandan A. Adhikari D. Khan M.L. The dark cloud with a silver lining: assessing the impact of the SARS COVID-19 pandemic on the global environment Sci. Total Environ. 732 2020 139297 10.1016/j.scitotenv.2020.139297 Lancet T. Health P. Will the COVID-19 pandemic threaten the SDGs ? Lancet Public Heal 5 2020 e460 10.1016/S2468-2667(20)30189-4 Larsen D.A. Wigginton K.R. Tracking COVID-19 with wastewater Nat. Biotechnol. 38 2020 1151 1153 10.1038/s41587-020-0690-1 32958959 Lathabhavan R. Griffiths M. First case of student suicide in India due to the COVID-19 education crisis: a brief report and preventive measures Asian J. Psychiatr. 53 2020 102202 10.1016/j.ajp.2020.102202 Lau H. Khosrawipour V. Kocbach P. Mikolajczyk A. Ichii H. Zacharski M. Bania J. Khosrawipour T. The association between international and domestic air traffic and the coronavirus (COVID-19) outbreak J. Microbiol. Immunol. Infect. 53 2020 467 472 10.1016/j.jmii.2020.03.026 32299783 Lau H. Khosrawipour V. Kocbach P. Mikolajczyk A. Schubert J. Bania J. Khosrawipour T. The positive impact of lockdown in Wuhan on containing the COVID-19 outbreak in China J. Trav. Med. 2020 1 7 10.1093/jtm/taaa037 Le T.H. Nguyen C.P. Is energy security a driver for economic growth? Evidence from a global sample Energy Pol. 129 2019 436 451 10.1016/j.enpol.2019.02.038 Lehmiller J.J. Garcia J.R. Gesselman A.N. Mark K.P. Less sex, but more sexual diversity: changes in sexual behavior during the COVID-19 coronavirus pandemic Leisure Sci. 2020 1 10 10.1080/01490400.2020.1774016 Lessler J. Reich N.G. Brookmeyer R. Perl T.M. Nelson K.E. Cummings D.A. Incubation periods of acute respiratory viral infections: a systematic review Lancet Infect. Dis. 9 2009 291 300 10.1016/S1473-3099(09)70069-6 19393959 Li L. Li Q. Huang L. Wang Q. Zhu A. Xu J. Liu Ziyi Li H. Shi L. Li R. Azari M. Wang Y. Zhang X. Liu Zhiqiang Zhu Y. Zhang K. Xue S. Ooi M.C.G. Zhang D. Chan A. Air quality changes during the COVID-19 lockdown over the Yangtze River Delta Region: an insight into the impact of human activity pattern changes on air pollution variation Sci. Total Environ. 732 2020 10.1016/j.scitotenv.2020.139282 Li Y. Tenchov R. Liu C. Watkins S. A comprehensive review of the global E ff orts on COVID-19 vaccine development ACS Cent. Sci. 2 2021 10.1021/acscentsci.1c00120 Liu L. Emerging study on the transmission of the Novel Coronavirus (COVID-19) from urban perspective: evidence from China Cities 103 2020 102759 10.1016/j.cities.2020.102759 Liu K. COVID-19 and the Chinese economy: impacts, policy responses and implications Int. Rev. Appl. Econ. 35 2021 308 330 10.1080/02692171.2021.1876641 Liu Y. Yan S. Poh K. Liu S. Iyioriobhe E. Sterling D.A. Impact of air quality guidelines on COPD sufferers Int. J. COPD 11 2016 839 872 10.2147/COPD.S49378 Liu C.H. Zhang E. Wong G.T.F. Hyun S. “Chris” Hahm H. Factors associated with depression, anxiety, and PTSD symptomatology during the COVID-19 pandemic: clinical implications for U.S. young adult mental health Psychiatr. Res. 290 2020 10.1016/j.psychres.2020.113172 Liu J. Zhou J. Yao J. Zhang X. Li L. Xu X. He X. Wang B. Fu S. Niu T. Yan J. Shi Y. Ren X. Niu J. Zhu W. Li S. Luo B. Zhang K. Impact of meteorological factors on the COVID-19 transmission: a multi-city study in China Sci. Total Environ. 726 2020 138513 10.1016/j.scitotenv.2020.138513 Lu R. Zhao X. Li J. Niu P. Yang B. Wu H. Wang W. Song H. Huang B. Zhu N. Bi Y. Ma X. Zhan F. Wang L. Hu T. Zhou H. Hu Z. Zhou W. Zhao L. Chen J. Meng Y. Wang J. Lin Y. Yuan J. Xie Z. Ma J. Liu W.J. Wang D. Xu W. Holmes E.C. Gao G.F. Wu G. Chen W. Shi W. Tan W. Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding Lancet 395 2020 565 574 10.1016/S0140-6736(20)30251-8 32007145 Lu H. Ma X. Ma M. A hybrid multi-objective optimizer-based model for daily electricity demand prediction considering COVID-19 Energy 219 2021 119568 10.1016/j.energy.2020.119568 Ma Y. Zhao Y. Liu J. He X. Wang B. Fu S. Yan J. Niu J. Zhou J. Luo B. Effects of temperature variation and humidity on the death of COVID-19 in Wuhan, China Sci. Total Environ. 724 2020 138226 10.1016/j.scitotenv.2020.138226 Machado R.A. de Souza N.L. Oliveira R.M. Martelli Júnior H. Bonan P.R.F. Social media and telemedicine for oral diagnosis and counselling in the COVID-19 era Oral Oncol. 105 2020 104685 10.1016/j.oraloncology.2020.104685 Mahase E. Covid-19: UK approves Oxford vaccine as cases of new variant surge BMJ 371 2020 4968 10.1136/bmj.m4968 Mail M. Covid-19: water bill exemption, electricity bill discount for Sabahans starting April [WWW Document] https://www.malaymail.com/news/malaysia/2020/03/25/covid-19-water-bill-exemption-electricity-bill-discount-for-sabahans-starti/1850187 2020 Mandal I. Pal S. COVID-19 pandemic persuaded lockdown effects on environment over stone quarrying and crushing areas Sci. Total Environ. 732 2020 139281 10.1016/j.scitotenv.2020.139281 Mastropietro P. Rodilla P. Batlle C. Emergency measures to protect energy consumers during the covid-19 pandemic: global review and critical analysis Robert Schuman Cent. - Pol. Br. 68 2020 101678 10.1016/j.erss.2020.101678 McKee M. Stuckler D. If the world fails to protect the economy, COVID-19 will damage health not just now but also in the future Nat. Med. 26 2020 640 642 10.1038/s41591-020-0863-y 32273610 Medema G. Heijnen L. Elsinga G. Italiaander R. Brouwer A. Presence of SARS-coronavirus-2 RNA in sewage and correlation with reported COVID-19 prevalence in the early stage of the epidemic in The Netherlands Environ. Sci. Technol. Lett. 7 2020 511 516 10.1021/acs.estlett.0c00357 Megahed N.A. Ghoneim E.M. Antivirus-built environment: lessons learned from Covid-19 pandemic Sustain. Cities Soc. 61 2020 102350 10.1016/j.scs.2020.102350 Memmott T. Carley S. Graff M. Konisky D.M. Sociodemographic disparities in energy insecurity among low-income households before and during the COVID-19 pandemic Nat. Energy 6 2021 186 193 10.1038/s41560-020-00763-9 Méndez-Arriaga F. The temperature and regional climate effects on communitarian COVID-19 contagion in Mexico throughout phase 1 Sci. Total Environ. 735 2020 10.1016/j.scitotenv.2020.139560 Menebo M.M. Temperature and precipitation associate with Covid-19 new daily cases: a correlation study between weather and Covid-19 pandemic in Oslo, Norway Sci. Total Environ. 737 2020 139659 10.1016/j.scitotenv.2020.139659 Mesloub A. Ghosh A. Daylighting performance of light shelf photovoltaics (LSPV) for office buildings in hot desert-like regions Appl. Sci. 10 2020 1 24 10.3390/app10227959 Mesloub A. Ghosh A. Albaqawy G.A. Noaime E. Alsolami B.M. Energy and daylighting evaluation of integrated semitransparent photovoltaic windows with internal light shelves in open-office buildings Adv. Civ. Eng. 2020 2020 Meyer M.W. COVID lockdowns , social distancing , and fatal car Crashes : more deaths on hobbesian Highways ? Cambridge J. Evidence-Based Polic. 4 2020 238 259 Mint L. India announces discom’ relief measures to ensure round-the-clock power supply https://www.livemint.com/industry/energy/india-announces-discom-relief-measures-to-ensure-round-the-clock-power-supply-11585375723511.html 2020 Mitjà O. Arenas À. Rodó X. Tobias A. Brew J. Benlloch J.M. Experts' request to the Spanish Government: move Spain towards complete lockdown Lancet 395 2020 1193 1194 10.1016/S0140-6736(20)30753-4 32224297 Mohler G. Bertozzi A.L. Carter J. Short M.B. Sledge D. Tita G.E. Uchida C.D. Brantingham P.J. Impact of social distancing during COVID-19 pandemic on crime in Los Angeles and Indianapolis J. Crim. Justice 68 2020 101692 10.1016/j.jcrimjus.2020.101692 Mokhtari R. Jahangir M.H. The effect of occupant distribution on energy consumption and COVID-19 infection in buildings: a case study of university building Build. Environ. 190 2021 107561 10.1016/j.buildenv.2020.107561 Moyer J.D. Hedden S. Are we on the right path to achieve the sustainable development goals ? World Dev. 127 2020 104749 10.1016/j.worlddev.2019.104749 Nabi G. Wang Y. Hao Y. Khan S. Wu Y. Li D. Massive use of disinfectants against COVID-19 poses potential risks to urban wildlife Environ. Res. 188 2020 9 11 10.1016/j.envres.2020.109916 Naderipour A. Abdul-Malek Z. Ahmad N.A. Kamyab H. Ashokkumar V. Ngamcharussrivichai C. Chelliapan S. Effect of COVID-19 virus on reducing GHG emission and increasing energy generated by renewable energy sources: a brief study in Malaysian context Environ. Technol. Innov. 20 2020 101151 10.1016/j.eti.2020.101151 Naidoo R. Fisher B. Goals : pandemic reset Nature 583 2020 198 201 32632244 Nakada L.Y.K. Urban R.C. COVID-19 pandemic: impacts on the air quality during the partial lockdown in São Paulo state Brazil. Sci. Total Environ. 730 2020 139087 10.1016/j.scitotenv.2020.139087 Nakajima K. Takane Y. Kikegawa Y. Furuta Y. Takamatsu H. Human behaviour change and its impact on urban climate: restrictions with the G20 Osaka Summit and COVID-19 outbreak Urban Clim. 35 2021 100728 10.1016/j.uclim.2020.100728 Nakamura H. Managi S. Airport risk of importation and exportation of the COVID-19 pandemic Transport Pol. 96 2020 40 47 10.1016/j.tranpol.2020.06.018 News S. COVID-19: New Strain Found in Italy, Denmark, Netherlands, Australia and Gibraltar 2020 Nicola M. Alsafi Z. Sohrabi C. Kerwan A. Al-jabir A. Since January 2020 Elsevier Has Created a COVID-19 Resource Centre with Free Information in English and Mandarin on the Novel Coronavirus COVID- 19 2020 The COVID-19 resource centre is hosted on Elsevier Connect , the company ’ s public news and information Nundy S. Ghosh A. Thermal and visual comfort analysis of adaptive vacuum integrated switchable suspended particle device window for temperate climate Renew. Energy 156 2020 1361 1372 10.1016/j.renene.2019.12.004 Nundy S. Ghosh A. Mallick T.K. Hydrophilic and superhydrophilic self-cleaning coatings by morphologically varying ZnO microstructures for photovoltaic and glazing applications ACS Omega 5 2020 1033 1039 10.1021/acsomega.9b02758 31984259 Nundy S. Mesloub A. Alsolami B.M. Ghosh A. Electrically actuated visible and near-infrared regulating switchable smart window for energy positive building : a review J. Clean. Prod. 301 2021 126854 10.1016/j.jclepro.2021.126854 ( NWSS), N.W.S.S. A new public health tool to understand COVID-19 spread in a community [WWW Document]. Centers Dis. Control Prev. URL https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/wastewater-surveillance.html 2020 OECD COVID-19 and the food and agriculture sector issues and policy response https://read.oecd-ilibrary.org/view/?ref=130_130816-9uut45lj4q&title=Covid-19-and-the-food-and-agriculture-sector-Issues-and-policy-responses 2020 Ogen Y. Assessing nitrogen dioxide (NO2) levels as a contributing factor to coronavirus (COVID-19) fatality Sci. Total Environ. 726 2020 138605 10.1016/j.scitotenv.2020.138605 Otmani A. Benchrif A. Tahri M. Bounakhla M. Chakir E.M. El Bouch M. Krombi M. Impact of covid-19 lockdown on PM10, SO2 and NO2 concentrations in Salé city (Morocco) Sci. Total Environ. 735 2020 139541 10.1016/j.scitotenv.2020.139541 Oztemel E. Gursev S. Literature review of Industry 4.0 and related technologies J. Intell. Manuf. 31 2020 127 182 10.1007/s10845-018-1433-8 Pantano E. Pizzi G. Scarpi D. Dennis C. Competing during a pandemic? Retailers' ups and downs during the COVID-19 outbreak J. Bus. Res. 116 2020 209 213 10.1016/j.jbusres.2020.05.036 32501307 Parady G. Taniguchi A. Takami K. Transportation Research Interdisciplinary Perspectives Travel behavior changes during the COVID-19 pandemic in Japan : analyzing the effects of risk perception and social in fl uence on going-out self-restriction Transp. Res. Interdiscip. Perspect. 7 2020 100181 10.1016/j.trip.2020.100181 Park J.E. Son W.S. Ryu Y. Choi S.B. Kwon O. Ahn I. Effects of temperature, humidity, and diurnal temperature range on influenza incidence in a temperate region Influenza Respi. Viruses 14 2020 11 18 10.1111/irv.12682 Peak C.M. Wesolowski A. Erbach-schoenberg E. Tatem A.J. Wetter E. Lu X. Power D. Weidman-grunewald E. Ramos S. Moritz S. Population mobility reductions associated with travel restrictions during the Ebola epidemic in Sierra Leone : use of mobile phone data Int. J. Epidemiol. 47 2018 1562 1570 10.1093/ije/dyy095 29947788 Peccia J. Zulli A. Brackney D.E. Grubaugh N.D. Kaplan E.H. Casanovas-Massana A. Ko A.I. Malik A.A. Wang D. Wang M. Warren J.L. Weinberger D.M. Arnold W. Omer S.B. Measurement of SARS-CoV-2 RNA in wastewater tracks community infection dynamics Nat. Biotechnol. 38 2020 1164 1167 10.1038/s41587-020-0684-z 32948856 Peng K. Wei Z. Chen J. Li H. International Journal of Electrical Power and Energy Systems Hierarchical virtual inertia control of DC distribution system for plug-and-play electric vehicle integration Int. J. Electr. Power Energy Syst. 128 2021 106769 10.1016/j.ijepes.2021.106769 Perkins K.M. Munguia N. Ellenbecker M. Moure-Eraso R. Velazquez L. COVID-19 pandemic lessons to facilitate future engagement in the global climate crisis J. Clean. Prod. 2020 125178 10.1016/j.jclepro.2020.125178 Peto J. Alwan N. Godfrey K. Universal weekly testing as the UK COVID-19 lockdown exit strategy Lancet (London, England) 2020 1420 1421 10.1016/s0140-6736(20)30936-3 32325027 Pinheiro M.D. Luís N.C. COVID-19 could leverage a sustainable built environment Sustain. Times 12 2020 10.3390/su12145863 Piyatamrong T. Derrick J. Nyamapfene A. Technology-mediated higher education provision during the COVID-19 Pandemic Qual. Assess. Eng. Student Exp. Sentiments 34 2021 290 297 Pollard B. Held F. Engelen L. Powell L. Dear R. De Science of the Total Environment Data fusion in buildings : synthesis of high-resolution IEQ and occupant tracking data Sci. Total Environ. 776 2021 146047 10.1016/j.scitotenv.2021.146047 Prata D.N. Rodrigues W. Bermejo P.H. Temperature significantly changes COVID-19 transmission in (sub)tropical cities of Brazil Sci. Total Environ. 729 2020 138862 10.1016/j.scitotenv.2020.138862 Przybylowski A. Stelmak S. Suchanek M. Mobility behaviour in view of the impact of the COVID-19 pandemic — public transport users in Gdansk case study Sustain. Times 13 2021 1 12 Qarnain S.S. Muthuvel S. Bathrinath S. Review on government action plans to reduce energy consumption in buildings amid COVID-19 pandemic outbreak Mater. Today Proc. 2020 10.1016/j.matpr.2020.04.723 Qi H. Xiao S. Shi R. Ward M.P. Chen Y. Tu W. Su Q. Wang W. Wang X. Zhang Z. COVID-19 transmission in Mainland China is associated with temperature and humidity: a time-series analysis Sci. Total Environ. 728 2020 138778 10.1016/j.scitotenv.2020.138778 Qureshi A.I. Huang W. Khan S. Lobanova I. Siddiq F. Mandated societal lockdown and road traffic accidents Accid. Anal. Prev. 146 2020 105747 10.1016/j.aap.2020.105747 Rajkumar R.P. Suicides related to the COVID-19 outbreak in India: a pilot study of media reports Asian J. Psychiatr. 53 2020 102196 10.1016/j.ajp.2020.102196 Randazzo W. Truchado P. Cuevas-Ferrando E. Simon P. Allende A. Sanchez G. SARS-CoV-2 RNA in wastewater anticipated COVID-19 occurrence in a low prevalence area Water Res. 181 2020 115942 Reddy P. Gupta M.V.N.S. Nundy S. Karthick A. Status of BIPV and BAPV system for less energy-hungry building in India — a review Appl. Sci. 10 2020 2337 Remuzzi A. Remuzzi G. COVID-19 and Italy: what next? Lancet 395 2020 1225 1228 10.1016/S0140-6736(20)30627-9 32178769 Ren X. Pandemic and lockdown: a territorial approach to COVID-19 in China, Italy and the United States Eurasian Geogr. Econ. 2020 1 12 10.1080/15387216.2020.1762103 Retallack A.E. Ostendorf B. Current understanding of the E ff ects of congestion on Tra ffi c accidents Int. J. Environ. Res. Publ. Health 16 2019 3400 Roberts G. Updated- Daily automotive coronoa virus briefing free to read https://www.just-auto.com/news/updated-daily-automotive-coronavirus-briefing-free-to-read_id194210.aspx 2020 Rodríguez-Urrego D. Rodríguez-Urrego L. Air quality during the COVID-19: PM2.5 analysis in the 50 most polluted capital cities in the world Environ. Pollut. 266 2020 115042 10.1016/j.envpol.2020.115042 Roesch E. Amin A. Gupta J. García-Moreno C. Violence against women during covid-19 pandemic restrictions BMJ 369 2020 2 3 10.1136/bmj.m1712 Rouleau J. Gosselin L. Impacts of the COVID-19 lockdown on energy consumption in a Canadian social housing building Appl. Energy 287 2021 116565 10.1016/j.apenergy.2021.116565 Roy A. Ghosh A. Benson D. Mallick T.K. Sundaram S. Emplacement of screen - printed graphene oxide coating for building thermal comfort discernment Sci. Rep. 10 2020 1 13 10.1038/s41598-020-72670-8 31913322 Roy A. Ghosh A. Bhandari S. Sundaram S. Mallick T.K. Realization of poly ( methyl methacrylate ) -encapsulated solution- processed carbon-based solar Cells : an emerging candidate for buildings ’ comfort Ind. Eng. Chem. Res. 2020 10.1021/acs.iecr.9b06902 Rugani B. Caro D. Impact of COVID-19 outbreak measures of lockdown on the Italian Carbon Footprint Sci. Total Environ. 737 2020 139806 10.1016/j.scitotenv.2020.139806 Şahin M. Impact of weather on COVID-19 pandemic in Turkey Sci. Total Environ. 728 2020 10.1016/j.scitotenv.2020.138810 Sajadi M.M. Habibzadeh P. Vintzileos A. Miralles-wilhelm F. Amoroso A. Temperature, Humidity, and Latitude Analysis to Predict Potential Spread and Seasonality for COVID-19 6–7 2020 Saladié Ò. Bustamante E. Gutiérrez A. Transportation Research Interdisciplinary Perspectives COVID-19 lockdown and reduction of traf fi c accidents in Tarragona province , Spain Transp. Res. Interdiscip. Perspect. 8 2020 100218 10.1016/j.trip.2020.100218 Sasidharan M. Singh A. Torbaghan M.E. Parlikad A.K. A vulnerability-based approach to human-mobility reduction for countering COVID-19 transmission in London while considering local air quality Sci. Total Environ. 741 2020 140515 10.1016/j.scitotenv.2020.140515 Schiavi M.C. Spina V. Zullo M.A. Colagiovanni V. Luffarelli P. Rago R. Palazzetti P. Love in the time of COVID-19: sexual function and quality of life analysis during the social distancing measures in a group of Italian reproductive-age women J. Sex. Med. 17 2020 1407 1413 10.1016/j.jsxm.2020.06.006 32653391 Schwarz M. Scherrer A. Hohmann C. Heiberg J. Brugger A. Nuñez-Jimenez A. COVID-19 and the academy: it is time for going digital Energy Res. Soc. Sci. 68 2020 1 3 10.1016/j.erss.2020.101684 Selvam S. Jesuraja K. Venkatramanan S. Chung S.Y. Roy P.D. Muthukumar P. Kumar M. Imprints of pandemic lockdown on subsurface water quality in the coastal industrial city of Tuticorin, South India: a revival perspective Sci. Total Environ. 738 2020 139848 10.1016/j.scitotenv.2020.139848 Selvaraj P. Ghosh A. Mallick T.K. Sundaram S. Investigation of semi-transparent dye-sensitized solar cells for fenestration integration Renew. Energy 141 2019 516 525 10.1016/j.renene.2019.03.146 Seow J. Graham C. Merrick B. Acors S. Longitudinal Evaluation and Decline of Antibody Responses in SARS-CoV-2 Infection 2 21 2020 1–9 Shahzad F. Shahzad U. Fareed Z. Iqbal N. Hashmi S.H. Ahmad F. Asymmetric nexus between temperature and COVID-19 in the top ten affected provinces of China: a current application of quantile-on-quantile approach Sci. Total Environ. 736 2020 139115 10.1016/j.scitotenv.2020.139115 Sharma S. Zhang M. Anshika Gao J. Zhang H. Kota S.H. Effect of restricted emissions during COVID-19 on air quality in India Sci. Total Environ. 728 2020 138878 10.1016/j.scitotenv.2020.138878 Shen J. Duan H. Zhang B. Wang Jiaqi Ji J.S. Wang Jiao Liang C. Sun H. Lv Y. Li Y. Li T. Li L. Liu H. Zhang L. Wang L. Shi X. Prevention and control of COVID-19 in public transportation : experience from China Environ. Pollut. 266 2020 115291 10.1016/j.envpol.2020.115291 Sheth J. Impact of Covid-19 on consumer behavior: will the old habits return or die? J. Bus. Res. 117 2020 280 283 10.1016/j.jbusres.2020.05.059 32536735 Shokouhyar S. Shokoohyar S. Sobhani A. Gorizi A.J. Shared mobility in post-COVID era: new challenges and opportunities Sustain. Cities Soc. 67 2021 102714 10.1016/j.scs.2021.102714 Shulla K. Friedrich B. Stefan V. Giuseppe C. Edna S. Filip M. Effects of COVID - 19 on the sustainable development goals ( SDGs ) Discov. Sustain. 2 2021 1 15 10.1007/s43621-021-00026-x Sicard P. De Marco A. Agathokleous E. Feng Z. Xu X. Paoletti E. Rodriguez J.J.D. Calatayud V. Amplified ozone pollution in cities during the COVID-19 lockdown Sci. Total Environ. 735 2020 10.1016/j.scitotenv.2020.139542 Siciliano B. Dantas G. da Silva C.M. Arbilla G. Increased ozone levels during the COVID-19 lockdown: analysis for the city of Rio de Janeiro, Brazil Sci. Total Environ. 737 2020 139765 10.1016/j.scitotenv.2020.139765 Singh R.P. Javaid M. Haleem A. Suman R. Internet of things (IoT) applications to fight against COVID-19 pandemic Diabetes Metab. Syndr. Clin. Res. Rev. 14 2020 521 524 10.1016/j.dsx.2020.04.041 SKWAWKBOX From midnight, France is cancelling all utility bills to help citizens cope [WWW Document] https://skwawkbox.org/2020/03/16/from-midnight-france-is-cancelling-all-utility-bills-to-help-citizens-cope/ 2020 Smith M.D. Wesselbaum D. COVID-19, food insecurity, and migration J. Nutr. 150 2020 2855 2858 10.1093/jn/nxaa270 32840610 Sobieralski J.B. COVID-19 and airline employment: insights from historical uncertainty shocks to the industry Transp. Res. Interdiscip. Perspect. 5 2020 100123 10.1016/j.trip.2020.100123 Sobral M.F.F. Duarte G.B. da Penha Sobral A.I.G. Marinho M.L.M. de Souza Melo A. Association between climate variables and global transmission oF SARS-CoV-2 Sci. Total Environ. 729 2020 138997 10.1016/j.scitotenv.2020.138997 Srivastava D.K. COVID-19: How India Can Revive Economic Growth 2020 Sutrisno A. Nomaler Ӧnder Alkemade F. Has the global expansion of energy markets truly improved energy security? Energy Pol. 148 2021 10.1016/j.enpol.2020.111931 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. 2020 1 12 10.1038/s41577-020-0311-8 31792373 The Guardian The end of tourism? https://www.theguardian.com/travel/2020/jun/18/end-of-tourism-coronavirus-pandemic-travel-industry 2020 The.Canadian.Press Coronavirus: power rates in Canada not being cut despite orders to work from home [WWW Document] https://globalnews.ca/news/6723366/coronavirus-electricity-rates-canada/ 2020 The.Jakarta.Post Jokowi announces free electricity, discounts for households hardest hit by COVID-19 impacts [WWW Document] https://www.thejakartapost.com/news/2020/03/31/jokowi-announces-free-electricity-discounts-for-households-hardest-hit-by-covid-19-impacts.html 2020 Tikkinen K.A.O. Malekzadeh R. Schlegel M. Rutanen J. Glasziou P. COVID-19 puts the Sustainable Development Goals center stage Nature 26 2020 10.1038/s41591-020-1077-z Times T.E. Prices of agricultural commodities drop 20% post COVID-19 outbreak https://economictimes.indiatimes.com/news/economy/agriculture/prices-of-agricultural-commodities-drop-20-post-covid-19-outbreak/articleshow/74705537.cms 2020 Times T.N. Coronavirus vaccine tracker https://www.nytimes.com/interactive/2020/science/coronavirus-vaccine-tracker.html 2020 Tobías A. Molina T. Is temperature reducing the transmission of COVID-19 ? Environ. Res. 186 2020 109553 10.1016/j.envres.2020.109553 Tobías A. Carnerero C. Reche C. Massagué J. Via M. Minguillón M.C. Alastuey A. Querol X. Changes in air quality during the lockdown in Barcelona (Spain) one month into the SARS-CoV-2 epidemic Sci. Total Environ. 726 2020 138540 10.1016/j.scitotenv.2020.138540 Tosepu R. Gunawan J. Effendy D.S. Ahmad L.O.A.I. Lestari H. Bahar H. Asfian P. Correlation between weather and covid-19 pandemic in Jakarta, Indonesia Sci. Total Environ. 725 2020 10.1016/j.scitotenv.2020.138436 Tull M.T. Edmonds K.A. Scamaldo K.M. Richmond J.R. Rose J.P. Gratz K.L. Psychological outcomes associated with stay-at-home orders and the perceived impact of COVID-19 on daily life Psychiatr. Res. 289 2020 113098 10.1016/j.psychres.2020.113098 Udmale P. Pal I. Szabo S. Pramanik M. Large A. Progress in Disaster Science Global food security in the context of COVID-19 : a scenario-based exploratory analysis Prog. Disaster Sci. 7 2020 100120 10.1016/j.pdisas.2020.100120 Ujiie M. Tsuzuki S. Ohmagari N. Effect of temperature on the infectivity of COVID-19 Int. J. Infect. Dis. 95 2020 301 303 10.1016/j.ijid.2020.04.068 32360939 UNESCO Education: from disruption to recovery [WWW Document] https://en.unesco.org/covid19/educationresponse 2020 Verschuur J. Koks E.E. Hall J.W. Observed impacts of the COVID-19 pandemic on global trade Nat. Hum. Behav. 5 2021 305 307 10.1038/s41562-021-01060-5 33633376 Vickerman R. Will Covid-19 put the public back in public transport? A UK perspective Transp. policy 103 2021 95 102 10.1016/j.tranpol.2021.01.005 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 Psychiatr. Res. 289 2020 113076 10.1016/j.psychres.2020.113076 Wang Z. Tang K. Combating COVID-19: health equity matters Nat. Med. 26 2020 458 10.1038/s41591-020-0823-6 32284617 Wang G. Zhang Y. Zhao J. Zhang J. Jiang F. Mitigate the effects of home confinement on children during the COVID-19 outbreak Lancet 395 2020 945 947 10.1016/S0140-6736(20)30547-X Wang L. Wang Y. Ye D. Liu Q. Review of the 2019 novel coronavirus (SARS-CoV-2) based on current evidence Int. J. Antimicrob. Agents 55 2020 105948 10.1016/j.ijantimicag.2020.105948 Wang M. Jiang A. Gong L. Luo L. Guo W. Li C. Zheng J. Li C. Yang B. Zheng J. Chen Y. Zheng K. Li H. Temperature significant change COVID-19 Transmission in 429 cities medRxiv 2020 1689 1699 10.1101/2020.02.22.20025791 Wang W. Tang J. Wei F. Updated understanding of the outbreak of 2019 novel coronavirus (2019-nCoV) in Wuhan, China J. Med. Virol. 92 2020 441 447 10.1002/jmv.25689 31994742 Wang Y. Yuan Y. Wang Q. Liu C.G. Zhi Q. Cao J. Changes in air quality related to the control of coronavirus in China: implications for traffic and industrial emissions Sci. Total Environ. 731 2020 139133 10.1016/j.scitotenv.2020.139133 Wang C.J. Ng C.Y. Brook R.H. Response to COVID-19 in Taiwan: big data analytics, new technology, and proactive testing JAMA, J. Am. Med. Assoc. 323 2020 1341 1342 10.1001/jama.2020.3151 Wang C. Zhang F. Wang J. Doyle J.K. Hancock P.A. Mak C.M. Liu S. How indoor environmental quality affects occupants' cognitive functions: a systematic review Build. Environ. 193 2021 107647 10.1016/j.buildenv.2021.107647 Wenham C. Smith J. Davies S.E. Feng H. Grépin K.A. Harman S. Herten-crabb A. Morgan R. Women are most affected by pandemics- lessons from past outbreaks Nature 583 2020 194 202 32641809 Werth A. Gravino P. Prevedello G. Impact analysis of COVID-19 responses on energy grid dynamics in Europe Appl. Energy 281 2021 116045 10.1016/j.apenergy.2020.116045 WHO Do weather and climate determine where COVID-19 occurs? [WWW Document] https://www.who.int/news-room/q-a-detail/q-a-on-climate-change-and-covid-19#:~:text=There%20is%20no%20evidence%20of,transmission%20and%20treating%20patients 2020 Wibawa T. COVID-19 vaccine research and development: ethical issues Trop. Med. Int. Health 26 2020 14 19 10.1111/tmi.13503 33012020 Williams W.F. The Italian maritime and energy industries and COVID-19 2020 https://www.wfw.com/articles/the-italian-maritime-and-energy-industries-in-the-time-of-covid-19/ [WWW Document]. Wise J. Covid-19: new coronavirus variant is identified in UK BMJ 371 2020 m4857 10.1136/bmj.m4857 33328153 Woods E.T. Schertzer R. Greenfeld L. Hughes C. Miller-Idriss C. COVID-19, nationalism, and the politics of crisis: a scholarly exchange Nations Natl. 2020 1 19 10.1111/nana.12644 Wu Y. Jing W. Liu J. Ma Q. Yuan J. Wang Y. Du M. Liu M. Effects of temperature and humidity on the daily new cases and new deaths of COVID-19 in 166 countries Sci. Total Environ. 729 2020 1 7 10.1016/j.scitotenv.2020.139051 Xiao F. Tang M. Zheng X. Liu Y. Li X. Shan H. Evidence for gastrointestinal infection of SARS-CoV-2 Gastroenterology 2507 2020 1 9 Xie J. Zhu Y. Association between ambient temperature and COVID-19 infection in 122 cities from China Sci. Total Environ. 724 2020 138201 10.1016/j.scitotenv.2020.138201 Xu H. Yan C. Fu Q. Xiao K. Yu Y. Han D. Wang W. Cheng J. Possible environmental effects on the spread of COVID-19 in China Sci. Total Environ. 731 2020 139211 10.1016/j.scitotenv.2020.139211 Xu X.W. Wu X.X. Jiang X.G. Xu K.J. Ying L.J. Ma C.L. Li S.B. Wang H.Y. Zhang S. Gao H.N. Sheng J.F. Cai H.L. Qiu Y.Q. Li L.J. Clinical findings in a group of patients infected with the 2019 novel coronavirus (SARS-Cov-2) outside of Wuhan, China: retrospective case series BMJ 368 2020 1 7 10.1136/bmj.m606 Yan Q.L. Tang S.Y. Xiao Y.N. Impact of individual behaviour change on the spread of emerging infectious diseases Stat. Med. 37 2018 948 969 10.1002/sim.7548 29193194 Yang K. Research and countermeasures of the influence of air pollution on human body IOP Conf. Ser. Earth Environ. Sci. 450 2020 10.1088/1755-1315/450/1/012047 Yao Y. Pan J. Liu Z. Meng X. Wang Weidong Kan H. Wang Weibing No association of COVID-19 transmission with temperature or UV radiation in Chinese cities Eur. Respir. J. 55 2020 7 9 10.1183/13993003.00517-2020 Yezli S. Khan A. COVID-19 social distancing in the Kingdom of Saudi Arabia: bold measures in the face of political, economic, social and religious challenges Trav. Med. Infect. Dis. 2020 101692 10.1016/j.tmaid.2020.101692 Yoo S. Managi S. Global mortality benefits of COVID-19 action Technol. Forecast. Soc. Change 160 2020 10.1016/j.techfore.2020.120231 Yoshino N. Taghizadeh-Hesary F. Otsuka M. Covid-19 and optimal portfolio selection for investment in sustainable development goals Finance Res. Lett. 38 2021 101695 10.1016/j.frl.2020.101695 Yunus A.P. Masago Y. Hijioka Y. COVID-19 and surface water quality: improved lake water quality during the lockdown Sci. Total Environ. 731 2020 139012 10.1016/j.scitotenv.2020.139012 Zambrano-Monserrate M.A. Ruano M.A. Sanchez-Alcalde L. Indirect effects of COVID-19 on the environment Sci. Total Environ. 728 2020 10.1016/j.scitotenv.2020.138813 Zhang H. Yan J. Yu Q. Obersteiner M. Li W. Chen J. Zhang Q. Jiang M. Wallin F. Song X. Wu J. Wang X. Shibasaki R. 1.6 Million transactions replicate distributed PV market slowdown by COVID-19 lockdown Appl. Energy 283 2021 116341 10.1016/j.apenergy.2020.116341 Zhang Jie Feng B. Wu Y. Id P.X. Ke R. Id N.D. The effect of human mobility and control measures on traffic safety during COVID-19 pandemic PloS One 16 2021 1 9 10.1371/journal.pone.0243263 Zhang Junyi Hayashi Y. Frank L.D. COVID-19 and transport: findings from a world-wide expert survey Transport Pol. 103 2021 68 85 10.1016/j.tranpol.2021.01.011 Zhao S. Zhuang Z. Ran J. Lin Y. He D. The association between domestic train transportation and novel coronavirus ( 2019-nCoV ) outbreak in China from 2019 to 2020 : a data-driven correlational report Trav. Med. Infect. Dis. 33 2020 2019 2021 10.1016/j.tmaid.2020.101568 Zheng H. Kong S. Chen N. Yan Y. Liu D. Zhu B. Xu K. Cao W. Ding Q. Lan B. Zhang Z. Zheng M. Fan Z. Cheng Y. Zheng S. Yao L. Bai Y. Zhao T. Qi S. Significant changes in the chemical compositions and sources of PM2.5 in Wuhan since the city lockdown as COVID-19 Sci. Total Environ. 739 2020 10.1016/j.scitotenv.2020.140000 Zhou P. Yang X. Lou Wang X.G. Hu B. Zhang L. Zhang W. Si H.R. Zhu Y. Li B. Huang C.L. Chen H.D. Chen J. Luo Y. Guo H. Jiang R. Di Liu M.Q. Chen Y. Shen X.R. Wang X. Zheng X.S. Zhao K. Chen Q.J. Deng F. Liu L.L. Yan B. Zhan F.X. Wang Y.Y. Xiao G.F. Shi Z.L. A pneumonia outbreak associated with a new coronavirus of probable bat origin Nature 579 2020 270 273 10.1038/s41586-020-2012-7 32015507 Zhu N. Zhang D. Wang W. Li X. Yang B. Song J. Zhao X. Huang B. Shi W. Lu R. Niu P. Zhan F. Ma X. Wang D. Xu W. Wu G. Gao G.F. Tan W. A novel coronavirus from patients with pneumonia in China, 2019 N. Engl. J. Med. 382 2020 727 733 10.1056/NEJMoa2001017 31978945 Zoran M.A. Savastru R.S. Savastru D.M. Tautan M.N. Assessing the relationship between surface levels of PM2.5 and PM10 particulate matter impact on COVID-19 in Milan Italy. Sci. Total Environ. 738 2020 139825 10.1016/j.scitotenv.2020.139825	36471816	PMC9710714	NO-CC CODE	2022-12-02 23:21:29	no	J Clean Prod. 2021 Aug 20; 312:127705	utf-8	J Clean Prod	2,021	10.1016/j.jclepro.2021.127705	oa_other
==== Front J Obstet Gynecol Neonatal Nurs J Obstet Gynecol Neonatal Nurs Journal of Obstetric, Gynecologic, and Neonatal Nursing 0884-2175 1552-6909 AWHONN, the Association of Women's Health, Obstetric and Neonatal Nurses. Published by Elsevier Inc. S0884-2175(22)00356-2 10.1016/j.jogn.2022.11.003 Review Scoping Review of Racial and Ethnic Representation of Participants in Mental Health Research Conducted in the Perinatal Period During the COVID-19 Pandemic Goyal Deepika ∗ Dol Justine Leckey Madeline Naraine Sarah Dennis Cindy-Lee Chan Emily K. Basu Geetali ∗ Correspondence Deepika Goyal, PhD, MS, FNP-C, San José State University, The Valley Foundation School of Nursing, One Washington Square, San Jose, CA 95192-0057. 30 11 2022 30 11 2022 14 11 2022 © 2022 AWHONN, the Association of Women's Health, Obstetric and Neonatal Nurses. Published by Elsevier Inc. 2022 AWHONN, the Association of Women’s Health, Obstetric and Neonatal Nurses 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 the racial and ethnic representation of participants in mental health research conducted in the perinatal period during the COVID-19 pandemic. Data Sources MEDLINE, CINAHL, Cochrane Library, PsycINFO, Scopus, Web of Science. Study Selection We included peer-reviewed research articles in which researchers reported mental health outcomes of women during the perinatal period who were living in the United States or Canada during the COVID-19 pandemic. We included 25 articles in the final review. Data Extraction We extracted the citation, publication date, design, aim, country of origin, participant characteristics, sampling method, method of measurement of race and ethnicity, and mental health outcome(s). Data Synthesis The combined racial and ethnic representation of the 16,841 participants in the included studies was White (76.5%), Black (9.8%), other/multiracial (6.2%), Asian (3.9%), Hispanic/Latina (2.6%), Indigenous or Ethnic Minority Canadian (0.9%), and Native American or Alaska Native (0.1%). Most studies were conducted in the United States, used a cross-sectional design, and incorporated social media platforms to recruit participants. Depression, anxiety, and stress were the most frequently assessed mental health outcomes. Conclusion Relatively few women of color who were pregnant or in the postpartum period during the pandemic participated in mental health research studies. Future studies should develop intentional recruitment strategies to increase participation of women of color. Researchers should use updated guidance on reporting race and ethnicity to accurately represent every participant, minimize misclassification of women of color, and report meaningful results. The authors review the racial and ethnic diversity of women who participated in perinatal mental health research studies conducted during the pandemic. Keywords COVID-19 mental health mothers postpartum period pregnancy ==== Body pmcCommon mental health disorders that occur during the perinatal period between pregnancy and the first year postpartum include anxiety, depression, stress, and posttraumatic stress. In the United States, up to 14% of women experience symptoms of depression across the perinatal period (Gavin et al., 2005; Ko et al., 2017; Wisner et al., 2013), and approximately 15% to 21% experience depression and anxiety symptoms in pregnancy (Wisner et al., 2013). Depression and anxiety rates are similar among Canadian women (Gheorghe et al., 2021; Statistics Canada, 2019). Evidence suggests that between 3.3% and 4% of women in the United States experience posttraumatic stress symptoms after childbirth (Yildiz et al., 2017). The COVID-19 public health crisis intensified the risk of antenatal mental health disorders because of related stressors that included (a) a shift to virtual health care and the inability to bring a partner or support person to in-person visits (C. H. Liu, Goyal, et al., 2021); (b) fear and the unknown effects of the virus on the fetus or newborn (Basu et al., 2021; Burgess et al. 2022; Kotlar et al., 2021); (c) the increased demand of juggling work from home, loss of child care, and school closures (Ajayi et al., 2021; Basu et al., 2021; Lee & Parolin, 2021; Mollard et al., 2021; Perzow et al., 2021); and (d) missed milestone celebrations, such as baby showers and gender reveals (Goyal, De La Rosa, et al., 2021; Goyal, Han, et al., 2021). Fear of exposure to COVID-19, partner/support restrictions during the birth, and being separated from the infant after the birth increased the risk of depression and anxiety in women in the postnatal period (Burgess et al., 2022; C. H. Liu et al., 2022; Shuman et al., 2022).The racial and ethnic representation of participants in maternal mental health research conducted in North America during the COVID-19 pandemic is unknown. The COVID-19 pandemic also highlighted racial inequities among women during the childbearing period. Even before the pandemic, women of color were at an increased risk of developing symptoms of depression in the perinatal period (Bauman et al., 2020; Daoud et al., 2019; Gheorghe et al., 2021; Groulx et al., 2021; Hetherington et al., 2020; Kendig et al., 2017; Miller et al., 2022; Mukherjee et al., 2016; Soffer et al., 2019; Wisner et al., 2013). Although data from 240,147 birth certificates in California indicated an increase in COVID-19 diagnosis across all racial and ethnic groups (Karasek et al., 2021), COVID-19 disproportionately affected the birth experiences of women of color. For example, compared with White women, those of color were more likely to experience discrimination in health care (e.g., Black women, 40.0% and Latina women, 35.3%; Janevic et al., 2021), be less satisfied with the birth (Breman et al., 2021; Janevic et al., 2021), and have less access to health care (Masters et al., 2021). Additionally, in a retrospective design, Pope et al. (2021) compared public health surveillance data of 162 pregnant women infected with COVID-19 to identify risk factors among Black (n = 81, 50%) and non-Black (n = 81, 50%) women. The results indicated that Black women infected with COVID-19 were significantly more likely to have preterm birth (p = .026) and be exposed to COVID-19 at work (p = .020) than non-Black women infected with COVID-19 (Pope et al., 2021). Race and Ethnicity of Participants in Perinatal Mental Health Research Before COVID-19 Before the pandemic, in several large-scale studies on perinatal mental health issues conducted in the United States and Canada, researchers reported that most participants were White. For example, 80% of participants in a sequential case series study to identify symptoms of depression among 10,000 women who recently gave birth in the United States were White (Wisner et al., 2013). McCall-Hosenfeld et al. (2016) identified symptoms of depression among 3,006 women in the United States who participated in the First Baby Cohort Study, and 83.2% self-identified as White. In a secondary analysis of 8,784 pregnant women in the United States conducted by Miller et al. (2022) to identify the trajectory of symptoms of depression, 62.3% of the participants self-identified as White. In Canada, in a survey by Hetherington et al. (2020) survey of 3,387 women to assess symptoms of depression in the perinatal period, 78.6% self-identified as White. The percentage of White participants in these large studies is not surprising because it mirrors the demographic characteristics of the North American population, in which 75% of Americans and Canadians self-identify as White (Statistics Canada, 2022; U.S. Census Bureau, 2021a). However, the population in North America is becoming increasingly racially and ethnically diverse. In the United States, the non-Hispanic White population decreased from 63.7% in 2010 to 57.8% in 2020 (U.S. Census, 2021b). Conversely, the Hispanic or Latino population grew from 16% of the total population in 2010 to 18.7% in 2020; similarly, the Asian population grew from 5% in 2010 to 6.2% in 2020 (U.S. Census Bureau, 2021b). In Canada, census data reveal that approximately one fifth (22.3%) of the population is classified as a visible minority, defined as “persons, other than Aboriginal peoples, who are non-Caucasian in race or non-White in colour” (Statistics Canada, 2021, para. 1), and this is projected to increase to 33% by 2036 (Morency et al., 2017). The increasing racial and ethnic diversity in North America, together with the recent national focus on the increased maternal mortality and morbidity among women of color (Howell, 2018; Kozhimannil et al., 2020), has led to renewed social justice movements that call out structural racism and implicit bias (Harris et al., 2021; Huggins et al., 2020; Matthews et al., 2021; Taylor, 2020). Although review articles have mapped the influence of the COVID-19 pandemic on maternal mental health issues across the globe, the race and ethnicity of research participants have not been discussed (Iyengar et al., 2021; Kotlar et al., 2021; Suwalska et al., 2021; Yan et al., 2020) and remain unknown. Therefore, we conducted a scoping review to identify the racial and ethnic representation of participants in mental health research conducted in the perinatal period during the COVID-19 pandemic. Methods Design We used the scoping review methodology of Arksey and O’Malley (2005), which includes the following steps: identifying the research question; identifying relevant studies; study selection; charting the data; and collating, summarizing, and reporting the results. We developed a review protocol to identify research studies that examined mental health outcomes among women who were pregnant or in the postpartum period during the COVID-19 pandemic. Next, we consulted with two research librarians on the topic and scope of the review. We conducted literature searches between December 9, 2021, and December 12, 2021, using MEDLINE (via PubMed), CINAHL, Cochrane Library, PsycINFO, Scopus, and Web of Science. Searches are inclusive of the results indexed at that time. Study Selection We included articles if they were published between January 1, 2020, and December 12, 2021, in English in peer-reviewed journals and if authors examined the influence of the pandemic on mental health outcomes (e.g., depression, anxiety, stress, posttraumatic stress) among women who were pregnant or within the first year postpartum at the time of data collection and who were living in the United States or Canada. We excluded studies on combined maternal and paternal outcomes, paternal outcomes only, and ones that were published in countries other than the United States or Canada. We also excluded studies with randomized controlled designs, short communications, psychometric evaluation of screening instruments, or reports of preliminary studies. See Supplementary Table S1 for the final search strings. Two research librarians conducted the searches (G.B. and E.K.C.). G.B. exported 2,640 citations into the review software manager Covidence and removed 1,028 duplicate citations, which left us to review 1,612 citations. Figure 1 depicts the search strategy.Figure 1 Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) diagram of scoping review methods. Data Extraction and Synthesis We screened articles in two phases. First, we conducted a title and abstract review and then a full-text review. The first and third authors (D.G. and M.L.) independently completed the title and abstract review for 1,612 articles, of which 1,484 were excluded and 128 were identified for full-text screening. At this stage, we excluded 103 articles for the following reasons: research was conducted outside of the United States or Canada; main outcomes were not depression symptoms, such as anxiety, stress, or posttraumatic stress disorder; inclusion of paternal or parental results; research was conducted outside the perinatal period; or the articles reported on preliminary studies or short communications. We discussed any discrepancies with the second author (J.D.). Twenty-five articles met the review criteria (see Figure 1). Data abstraction included the following information: publication date, design, aim, country of origin, participant sociodemographic characteristics, sampling method, method of measurement of race and ethnicity, and mental health outcome(s) measured. We extracted racial and ethnic data from the results sections, tables, and supplementary materials of the published articles. We compiled a spreadsheet to count the frequency of the racial and ethnic categories reported in the included studies and collapsed them into the following categories for this review: Asian, Hawaiian Pacific Islander; Black, African American, non-Hispanic Black; White, Non-Hispanic White, Caucasian; Hispanic, Latina; multiracial; Native American, Indigenous, First Nations; and other. The third (M.L.) and fourth (S.N.) authors independently extracted data, and the first author (D.G.) verified all data extraction and discussed discrepancies with the rest of the research team. We used the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) flowchart to guide screening along with the PRISMA extension for scoping reviews checklist (Tricco et al., 2016). Results We identified 25 articles that met the inclusion criteria: 15 were reports of quantitative studies, four were reports of qualitative studies, and six were reports of studies that used mixed methods. Most of the studies took place between March 2020 and January 2021, were conducted in the United States (n = 20, 80%), used cross-sectional designs (n =15, 60%), and incorporated social media platforms to recruit participants (n = 19, 76%). The total combined number of participants in the included studies was 16,841; individual sample sizes ranged from 31 (Farewell et al., 2020) to 4,604 (Groulx et al., 2021). See Table 1 for a summary of the articles and Supplementary Table S2 for detailed information on the data extracted from each article.Table 1 Summary Characteristics of Studies Included in Review (N = 25) Characteristics n (%) Country United States 20 (80) Canada 5 (20) Study design Quantitative 15 (60) Qualitative 4 (16) Mixed methods 6 (24) Recruitment strategy Social media, online 19 (76) Ongoing study 3 (12) Electronic medical records 2 (8) No method provided 1 (4) Language criteria English only 22 (88) French 2 (8) Spanish 1 (4) Method of race/ethnicity data collection Self-report 23 (92) Electronic medical records 2 (8) Participant race/ethnicity Asiana 665 (3.9) Blackb 1,655 (9.8) Canadianc 157 (0.9) Hispanic or Latina 445 (2.6) Native American, Alaskan Native 10 (0.1) Other, non-White, multiracial 1,041 (6.2) Whited 12,882 (76.5) Perinatal period Pregnancy 10 (40) Postpartum period 6 (24) Pregnancy and postpartum 9 (36) Mental health outcomes Anxiety alone 1 (4) Depression alone 2 (8) Stress alone 4 (16) Anxiety and depression 8 (32) Anxiety, depression, and stress 7 (28) Anxiety, depression, and posttraumatic stress 3 (12) a Includes Asian, Asian Indian, Asian/Hawaiian Pacific Islander, Chinese, Filipino, and Korean. b Includes Black, African American, and non-Hispanic Black. c Includes Indigenous, Ethnic Minority, First Nations, Inuit, and Metis. d Includes Caucasian and non-Hispanic White. Race and Ethnicities of Participants Most researchers in the included studies, including two who used electronic medical record data, used conventional race categories outlined by the Office of Management and Budget (1997): American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, White, and Hispanic ethnicity (see Table 2 and Supplementary Table S2). The racial and ethnic representation of the 16,841 participants in the included studies was White (76.5%), Black (9.8%), other/multiracial (6.2%), Asian (3.9%), Hispanic/Latina (2.6%), Indigenous or Ethnic Minority Canadian (0.9%), and Native American or Alaska Native (0.1%; see Table 2). Three researchers (12%) provided details for White participants without any further detail on the rest of the sample; McMillian et al. (2021) described participants as “primarily White, Non-Hispanic, Non-Latino” (83.7%); Moyer et al. (2020) described participants as “mostly White” (87.7%); and Ollivier et al. (2021) described participants as “primarily Caucasian” (96.6%).Table 2 Races and Ethnicities of Participants, % Authors and Year Asian, Hawaiian, and Pacific Islander Black African American and Non-Hispanic Black White, Non-Hispanic White, and Caucasian Hispanic and Latina Multiracial Native American, Indigenous, and First Nations Other Ahlers-Schmidt et al. (2020) — 19.3 43 26.3 5.3 — 6.1 Anderson et al. (2022) 1.7 3.3 71.7 15 8.3 — — Barbosa-Leiker et al. (2021) 4 5 79 7 5 — — Claridge et al. (2021) 2.6 2.4 88.2 16.7 6.9 — Farewell et al. (2020) — 7.1 85.7 14.8 — — 3.6 Goyal, Beck, et al. (2021) 8.4 0.8 82 6.1 — — 2.7 Goyal, De La Rosa, et al. (2021) 86.8 — — — — — 13.2 Groulx et al. (2021) 7.5 1.7 81.6 2.1 4.9 2.2 — Joy et al. (2020)a — — — — — — — Khoury et al. (2021) 6.9 — 84.8 — 3.0 0.7 4.6 Kinser et al. (2021) 3 8 83 2 3 1 2 Kornfield et al. (2021) — 18.1 5 71.9 — — — Lebel et al. (2020) 6 0.7 87.1 1.1 3.3 2 — C. H. Liu, Erdei, & Mittal (2021) 3.5 0.9 89.9 3.6 — — 2.1 C. H. Liu, Hyun, et al. (2021) 3 1 92.9 3.1 — — J. Liu et al. (2021) — 44.1 38.9 9.4 — — 7.7 McMillan et al. (2021) — — 83.7 — — — Mollard et al. (2021) 2.9 1.4 84.8 9.2 — — 1.8 Moyer et al. (2020) — — 88 — — — — Ollivier et al. (2021) — — 96.6 — — — — Omowale et al. (2021) — 38 62 — — — — Perzow et al. (2021) 5.1 11.1 54.8 25.9 — 6.7 — Silverman, Burgos, et al. (2020) and Silverman, Medeiros, & Burgos (2020)b — — — — — — — Wheeler et al. (2021) — 100 — — — — — a Majority Caucasian, predominantly White or non-Hispanic White. b Hispanic or African American (90%) and Asian (10%). In 17 studies (68%), researchers categorized participants as ethnic minorities, other, mixed race, one or more races, multiracial, or non-White, accounting for 6.2% of the total sample (Ahlers-Schmidt et al., 2020; Anderson et al., 2022; Barbosa-Leiker et al., 2021; Claridge et al., 2021; Farewell et al., 2020; Goyal, Beck, et al., 2021; Goyal, De La Rosa, et al. 2021; Groulx et al., 2021; Khoury et al., 2021; Kinser et al., 2021; Kornfield et al., 2021; Lebel et al., 2020; C. H. Liu, Erdei, & Mittal, 2021; J. Liu et al., 2021; McMillan et al., 2021; Mollard et al., 2021; Perzow et al., 2021). Two of the 25 studies (Groulx et al., 2021; Lebel et al., 2020) provided in-depth racial and ethnic descriptions of non-White participants: Black Chinese, Filipino, First Nations, Hispanic, Korean, Metis, mixed, South Asian, South-east Asian, and West Asian. Both studies were conducted in Canada, and the researchers reported on participants from the same prospective study; Lebel et al. (2020) evaluated data collected in April 2020, and Groulx et al. (2021) reported data collected from April to June 2020.The results of our review indicate that few women of color participated in maternal mental health research conducted in North America during the COVID-19 pandemic. In two studies, researchers evaluated the experiences of women of color during the pandemic. Goyal, De La Rosa, et al. (2021) assessed symptoms of depression and experiences in women of Asian American descent. Researchers asked participants to self-report their Asian ethnicity resulting in the following 10 ethnicities: Asian Indian, Chinese, Filipino, Hmong, Japanese, Laotian, Korean, Thai, and Vietnamese (Goyal, De La Rosa, et al., 2021). In a prospective, longitudinal cohort design, Wheeler et al. (2021) evaluated stress and coping among 33 pregnant Black women before and during the COVID-19 pandemic using data from an ongoing study. Mental Health Outcomes Of the 25 included studies that reported mental health outcomes, the most frequently assessed mental health outcomes were symptoms of depression, anxiety, and stress (n = 15, 60%; see Table 1). Mental health outcomes were assessed in 10 studies during pregnancy (Claridge et al., 2021; Groulx et al., 2021; Khoury et al., 2021; Lebel et al., 2020; C. H. Liu, Hyun, et al., 2021; J. Liu et al., 2021; McMillan et al., 2021; Moyer et al., 2020; Silverman, Medeiros, & Burgos, 2020; Wheeler et al., 2021), in six studies during the postnatal period (Goyal, Beck, et al., 2021; Goyal, Han, et al., 2021; Joy et al., 2020; Mollard et al., 2021; Ollivier et al., 2021; Silverman, Burgos, et al., 2020), and in nine studies across the perinatal period (Ahlers-Schmidt et al., 2020; Anderson et al., 2022; Barbosa-Leiker et al., 2021; Farewell et al., 2020; Kinser et al., 2021; Kornfield et al., 2021; C. H. Liu, Erdei, & Mittal, 2021; Omowale et al., 2021; Perzow et al., 2021). Mental Health Outcomes by Race and Ethnicity In four studies, researchers evaluated mental health outcomes by race and ethnicity with mixed results (Khoury et al., 2021; Kornfield et al., 2021; J. Liu et al., 2021; Mollard et al., 2021). In the study by J. Liu et al. (2021), a greater percentage of non-Hispanic White women (60.4%) reported symptoms of depression compared with Hispanic (44.8%), non-Hispanic Black (14.0%), or non-Hispanic other (32.7%) participants. Conversely, Black, Indigenous, or other women of color were more likely to report stress symptoms than White women (Mollard et al., 2021). Kornfield et al. (2021) assessed postpartum depression risk and found no significant difference between Black and White participants. Finally, the results of Khoury et al. (2021) results indicated an association between race and symptoms of anxiety and depression, without any further detail. Discussion In our scoping review, we mapped the racial and ethnic representation of participants in perinatal mental health studies conducted during the COVID-19 pandemic. Our findings indicate that very few Black, Indigenous, or other people of color (23.5%) participated in these pandemic studies. The stigma that is associated with mental health issues among African American (Ward et al., 2013), American Indian/Alaska Native (Grandbois, 2005), Asian American (Goyal et al., 2015; Han et al., 2020; Ta Park et al., 2017, 2019), and Hispanic individuals (Benuto et al., 2019) may account for the lower numbers of participants of color. Moreover, the fear, distrust (George et al., 2014), and historical misconduct associated with the Tuskegee study (Shavers et al., 2000) may be related to the low percentage of Black or African American participants (8.9%). On a more practical level, during the pandemic, women of color were more likely to be classified as essential workers (Pope et al., 2021; Rogers et al., 2020) and faced unemployment if they did not continue to work (Pew Research Center, 2020a), which may have left women without the time or desire to participate in research. Researchers’ implicit bias during study design and data reporting phases may also contribute to the underrepresentation of women of color. Most of the studies in our review used cross-sectional designs (n =15, 60%) and incorporated social media platforms to recruit participants (n = 19, 76%), which may be a problem, given that Facebook and Instagram are used more frequently by individuals with higher income levels (Pew Research Center, 2021). Researchers should consider using other social media platforms that attract younger and lower-income individuals, such as TikTok, to recruit a more diverse sample (Pew Research Center, 2021).To increase the participation of women of color in research, culturally sensitive recruitment strategies should be used and effort should be made to develop trusting relationships within diverse communities. The articles in our review reported racial and ethnic categories outlined by the Office of Management and Budget (1997): White, Black or African American, Latino or Hispanic, Asian American, Native Hawaiian and Pacific Islander, and American Indian and Alaska Native. The use of these broad racial and ethnic categories fails to accurately describe every participant and perpetuates the misrepresentation of many. Although more than 75% of the population in the United States and Canada self-identifies as White (Statistics Canada, 2022; U.S. Census Bureau, 2021a), it is important to note that the category “White” may include persons from Western/Eastern Europe, North Africa, or the Middle East. This may promote the misrepresentation of large groups of people with different experiences, cultures, and histories (Kauh et al., 2021). Misrepresentation also occurs when other broad racial and ethnic categories, such as Black, Asian, and Hispanic/Latino, are used to classify study participants (Kauh et al., 2021). To ensure accurate representation in prospective research studies, participants should self-identify their racial and ethnic identity (Ross et al., 2020). Updated guidance on the reporting of race and ethnicity in medical and science journals suggests that researchers should collect specific racial and ethnic categories versus broad terms and describe how racial and ethnic data were collected (e.g., self-report, electronic health record; Flanagin et al., 2021). American Indian and Alaska Native individuals accounted for 0.1% of the participants in our review, which may be due to misrepresentation. Researchers should consider best practices for American Indian and Alaska Native data collection recommendations, including allowing participants to report multiple races to accurately represent all American Indian and Alaska Native individuals (Urban Indian Health Institute, 2022). Limitations The purpose of our review methodology (Arksey & O’Malley, 2005) was to rapidly map key concepts in a research area, and we did not include a quality appraisal of the included articles, so the rigor and potential biases of the included studies are unknown. Our review was limited to English language articles with samples from the United States or Canada. Sampling bias and limitations inherent in the primarily descriptive, quantitative, and cross-sectional designs of the studies included in the review may also be reasons for inadequate racial and ethnic representation. Conclusion Based on our review, we conclude that the majority of women who participated in studies about mental health while pregnant or in the postpartum period during the COVID-19 pandemic were White. Our findings call for urgent action from researchers to make a targeted effort to recruit participants that reflect population demographics. Furthermore, researchers must use standard methods of reporting race and ethnicity (Flanagin et al., 2021) to shed light on how public health crises such as COVID-19 affect the mental health of pregnant and postpartum women of color. Additional recommendations to increase the participation of women of color in future research studies include building a diverse research team, developing culturally sensitive recruitment materials, and identifying key community contacts (Shavers et al., 2002; Webber-Ritchey et al., 2021). Uncited References Pew Research Center, 2019, Pew Research Center, 2020b Deepika Goyal, PhD, MS, FNP-C, is a professor in the Valley Foundation School of Nursing, San José State University, San Jose, CA. Justine Dol, PhD, is a postdoctoral fellow, St Michael’s Hospital, Toronto, Canada. Madeline Leckey, is an MSc student, Lawrence S. Bloomberg Faculty of Nursing, University of Toronto, Toronto, Canada. Sarah Naraine, is an MSc student, Lawrence S. Bloomberg Faculty of Nursing, University of Toronto, Toronto, Canada. Cindy-Lee Dennis, PhD, is a professor in the Lawrence S. Bloomberg Faculty of Nursing, University of Toronto, Toronto, Canada; Li Ka Shing Knowledge Institute, St. Michael’s Hospital, Toronto, Canada; and Department of Psychiatry, University of Toronto, Toronto, Canada. Emily K. Chan, is the Associate Dean for Research & Scholarship, Dr. Martin Luther King, Jr. Library, San José State University, San Jose, CA. Geetali Basu, is an academic liaison librarian, San José State University, Dr. Martin Luther King, Jr. Library, San Jose, CA. Supplementary Material Supplementary Tables S1 and S2 Conflict of Interest The authors report no conflicts of interest or relevant financial relationships. Funding None. Note: To access the supplementary material that accompanies this article, visit the online version of the Journal of Obstetric, Gynecologic, & Neonatal Nursing at http://jognn.org and at https://doi.org/10.1016/j.jogn.2022.11.003. ==== Refs References Ahlers-Schmidt C.R. Hervey A.M. Neil T. Kuhlmann S. Kuhlmann Z. Concerns of women regarding pregnancy and childbirth during the COVID-19 pandemic Patient Education and Counseling 103 12 2020 2578 2582 10.1016/j.pec.2020.09.031 33010997 Ajayi K.V. Harvey I.S. Panjwani S. Uwak I. Garney W. Page R.L. Narrative analysis of childbearing experiences during the COVID-19 pandemic MCN. The American Journal of Maternal/Child Nursing 46 5 2021 284 292 10.1097/nmc.0000000000000742 34162794 Anderson M.R. Salisbury A.L. Uebelacker L.A. Abrantes A.M. Battle C.L. Stress, coping and silver linings: How depressed perinatal women experienced the COVID-19 pandemic Journal of Affective Disorders 298 2022 2022 329 336 10.1016/j.jad.2021.10.116 34715180 Arksey H. O’Malley L. Scoping studies: Towards a methodological framework International Journal of Social Research Methodology 8 1 2005 19 32 10.1080/1364557032000119616 Barbosa-Leiker C. Smith C.L. Crespi E.J. Brooks O. Burduli E. Ranjo S. Gartstein M.A. Stressors, coping, and resources needed during the COVID-19 pandemic in a sample of perinatal women BMC Pregnancy Childbirth 21 1 2021 10.1186/s12884-021-03665-0 Article 171 Basu A. Kim H.H. Basaldua R. Choi K.W. Charron L. Kelsall N. Koenen K.C. A cross-national study of factors associated with women’s perinatal mental health and wellbeing during the COVID-19 pandemic PLOS ONE 16 4 2021 10.1371/journal.pone.0249780 Article e0249780 Bauman B.L. Ko J.Y. Cox S. D’Angelo D.V. Warner L. Folger S. Barfield W.D. Vital signs: Postpartum depressive symptoms and provider discussions about perinatal depression—United States, 2018 MMWR. Morbidity and Mortality Weekly Report 69 19 2020 575 581 10.15585/mmwr.mm6919a2 32407302 Benuto L.T. Gonzalez F. Reinosa-Segovia F. Duckworth M. Mental health literacy, stigma, and behavioral health service use: The case of Latinx and non-Latinx Whites Journal of Racial and Ethnic Health Disparities 6 6 2019 1122 1130 10.1007/s40615-019-00614-8 31327136 Breman R.B. Neerland C. Bradley D. Burgess A. Barr E. Burcher P. Giving birth during the COVID-19 pandemic, perspectives from a sample of the United States birthing persons during the first wave: March–June 2020 Birth 48 4 2021 524 533 10.1111/birt.12559 34114262 Burgess A. Breman R.B. Roane L.A. Dada S. Bradley D. Burcher P. Impact of COVID-19 on pregnancy worry in the United States Birth 49 3 2022 420 429 10.1111/birt.12608 34997646 Claridge A.M. Beeson T. Wojtyna A. Hoxmeier J. Pregnant women’s experiences during the COVID-19 pandemic: A mixed method exploration of prenatal depression Couple and Family Psychology 10 3 2021 168 178 10.1037/cfp0000178 Daoud N. O’Brien K. O’Campo P. Harney S. Harney E. Bebee K. Smylie J. Postpartum depression prevalence and risk factors among Indigenous, non-Indigenous and immigrant women in Canada Canadian Journal of Public Health 110 4 2019 440 452 10.17269/s41997-019-00182-8 30767191 Farewell C.V. Jewell J. Walls J. Leiferman J.A. A mixed-methods pilot study of perinatal risk and resilience during COVID-19 Journal of Primary Care & Community Health 11 2020 10.1177/2150132720944074 Article 2150132720944074 Flanagin A. Frey T. Christiansen S.L. Updated guidance on the reporting of race and ethnicity in medical and science journals JAMA 326 7 2021 621 627 10.1001/jama.2021.13304 34402850 Gavin N.I. Gaynes B.N. Lohr K.N. Meltzer-Brody S. Gartlehner G. Swinson T. Perinatal depression: A systematic review of prevalence and incidence Obstetrics & Gynecology 106 5 Pt. 1 2005 1071 1083 10.1097/01.AOG.0000183597.31630.db 16260528 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 American Journal of Public Health 104 2 2014 e16 e31 10.2105/AJPH.2013.301706 Gheorghe M. Varin M. Wong S.L. Baker M. Grywacheski V. Orpana H. Symptoms of postpartum anxiety and depression among women in Canada: Findings from a national cross-sectional survey Canadian Journal of Public Health 112 2 2021 244 252 10.17269/s41997-020-00420-4 33079328 Goyal D. Beck C.T. Webb R. Ayers S. Postpartum depressive symptoms and experiences during COVID-19 MCN. The American Journal of Maternal/Child Nursing 47 2 2021 77 84 10.1097/nmc.0000000000000802 Goyal D. De La Rosa L. Mittal L. Erdei C. Liu C.H. Unmet prenatal expectations during the COVID-19 pandemic MCN. The American Journal of Maternal/Child Nursing 47 2 2021 66 70 10.1097/nmc.0000000000000801 Goyal D. Han M. Feldman-Schwartz T. Le H.N. Perinatal experiences of Asian American women during COVID-19 MCN. The American Journal of Maternal/Child Nursing 47 2 2021 71 76 10.1097/nmc.0000000000000796 Goyal D. Park V.T. McNiesh S. Postpartum depression among Asian Indian mothers MCN. The American Journal of Maternal/Child Nursing 40 4 2015 256 261 10.1097/nmc.0000000000000146 26121757 Grandbois D. Stigma of mental illness among American Indian and Alaska Native nations: Historical and contemporary perspectives Issues in Mental Health Nursing 26 10 2005 1001 1024 10.1080/01612840500280661 16283996 Groulx T. Bagshawe M. Giesbrecht G. Tomfohr-Madsen L. Hetherington E. Lebel C.A. Prenatal care disruptions and associations with maternal mental health during the COVID-19 pandemic Frontiers in Global Women’s Health 2 2021 10.3389/fgwh.2021.648428 Article 648428 Han M. Goyal D. Lee J. Cho H. Kim A. Korean immigrant women’s postpartum experiences in the United States MCN. The American Journal of Maternal/Child Nursing 45 1 2020 42 48 10.1097/NMC.0000000000000585 31687983 Harris L.M. Forson-Dare Z. Gallagher P.G. Critical disparities in perinatal health: Understanding risks and changing the outcomes Journal of Perinatology 41 2 2021 181 182 10.1038/s41372-020-00913-7 33462341 Hetherington E. McDonald S. Williamson T. Tough S. Trajectories of social support in pregnancy and early postpartum: Findings from the all our families cohort Social Psychiatry and Psychiatric Epidemiology 55 2 2020 259 267 10.1007/s00127-019-01740-8 31256206 Howell E.A. Reducing disparities in severe maternal morbidity and mortality Clinical Obstetrics and Gynecology 61 2 2018 387 399 10.1097/grf.0000000000000349 29346121 Huggins B. Jones C. Adeyinka O. Ofomata A. Drake C. Kondas C. Racial disparities in perinatal mental health Psychiatric Annals 50 11 2020 489 493 10.3928/00485713-20201007-02 Iyengar U. Jaiprakash B. Haitsuka H. Kim S. One year into the pandemic: A systematic review of perinatal mental health outcomes during COVID-19 Frontiers in Psychiatry 12 2021 10.3389/fpsyt.2021.674194 Article 674194 Janevic T. Maru S. Nowlin S. McCarthy K. Bergink V. Stone J. Howell E.A. Pandemic birthing: Childbirth satisfaction, perceived health care bias, and postpartum health during the COVID-19 pandemic Maternal Child Health 25 6 2021 860 869 10.1007/s10995-021-03158-8 Joy P. Aston M. Price S. Sim M. Ollivier R. Benoit B. Iduye D. Blessings and curses: Exploring the experiences of new mothers during the COVID-19 pandemic Nursing Reports 10 2 2020 207 219 10.3390/nursrep10020023 34968364 Karasek D. Baer R.J. McLemore M.R. Bell A.J. Blebu B.E. Casey J.A. Jelliffe-Pawlowski L.L. The association of COVID-19 infection in pregnancy with preterm birth: A retrospective cohort study in California Lancet Regional Health. Americas 2 2021 10.1016/j.lana.2021.100027 Article 100027 Kauh T.J. Read J.G. Scheitler A.J. The critical role of racial/ethnic data disaggregation for health equity Population Research and Policy Review 40 1 2021 1 7 10.1007/s11113-020-09631-6 33437108 Kendig S. Keats J.P. Hoffman M.C. Kay L.B. Miller E.S. Simas T.A.M. Lemieux L.A. Consensus bundle on maternal mental health: Perinatal depression and anxiety Journal of Midwifery Women’s Health 62 2 2017 232 239 10.1111/jmwh.12603 Khoury J.E. Atkinson L. Bennett T. Jack S.M. Gonzalez A. COVID-19 and mental health during pregnancy: The importance of cognitive appraisal and social support Journal of Affective Disorders 282 2021 1161 1169 10.1016/j.jad.2021.01.027 33601691 Kinser P.A. Jallo N. Amstadter A.B. Thacker L.R. Jones E. Moyer S. Salisbury A.L. Depression, anxiety, resilience, and coping: The experience of pregnant and new mothers during the first few months of the COVID-19 pandemic Journal of Women’s Health 30 5 2021 654 664 10.1089/jwh.2020.8866 Ko J.Y. Rockhill K.M. Tong V.T. Morrow B. Farr S.L. Trends in postpartum depressive symptoms—27 states, 2004, 2008, and 2012 MMWR. Morbidity and Mortality Weekly Report 66 6 2017 153 158 10.15585/mmwr.mm6606a1 28207685 Kornfield S.L. White L.K. Waller R. Njoroge W. Barzilay R. Chaiyachati B.H. Gur R.E. Risk and resilience factors influencing postpartum depression and mother-infant bonding during COVID-19 Health Affairs 40 10 2021 1566 1574 10.1377/hlthaff.2021.00803 34606353 Kotlar B. Gerson E. Petrillo S. Langer A. Tiemeier H. The impact of the COVID-19 pandemic on maternal and perinatal health: A scoping review Reproductive Health 18 1 2021 10.1186/s12978-021-01070-6 Article 10 Kozhimannil K.B. Interrante J.D. Tofte A.N. Admon L.K. Severe maternal morbidity and mortality among Indigenous women in the United States Obstetrics & Gynecology 135 2 2020 294 300 10.1097/AOG.0000000000003647 31923072 Lebel C. MacKinnon A. Bagshawe M. Tomfohr-Madsen L. Giesbrecht G. Elevated depression and anxiety symptoms among pregnant individuals during the COVID-19 pandemic Journal of Affective Disorders 277 2020 5 13 10.1016/j.jad.2020.07.126 32777604 Lee E.K. Parolin Z. The care burden during COVID-19: A national database of child care closures in the United States Socius 7 2021 Article 23780231211032028 Liu C.H. Erdei C. Mittal L. Risk factors for depression, anxiety, and PTSD symptoms in perinatal women during the COVID-19 Pandemic Psychiatry Research 295 2021 10.1016/j.psychres.2020.113552 Article 113552 Liu C.H. Goyal D. Mittal L. Erdei C. Patient satisfaction with virtual-based prenatal care: Implications after the COVID-19 pandemic Maternal and Child Health Journal 25 2021 1735 1743 10.1007/s10995-021-03211-6 34410565 Liu C.H. Hyun S. Erdei C. Mittal L. Prenatal distress during the COVID-19 pandemic: Clinical and research implications Archives of Gynecology and Obstetrics 306 2 2021 397 405 10.1007/s00404-021-06286-2 34716818 Liu C.H. Koire A. Erdei C. Mittal L. Unexpected changes in birth experiences during the COVID-19 pandemic: Implications for maternal mental health Archives of Gynecology and Obstetrics 306 3 2022 687 697 10.1007/s00404-021-06310-5 34724569 Liu J. Hung P. Alberg A.J. Hair N.L. Whitaker K.M. Simon J. Taylor S.K. Mental health among pregnant women with COVID-19–related stressors and worries in the United States Birth 48 4 2021 470 479 10.1111/birt.12554 34008216 Masters G.A. Asipenko E. Bergman A.L. Person S.D. Brenckle L. Moore Simas T.A. Byatt N. Impact of the COVID-19 pandemic on mental health, access to care, and health disparities in the perinatal period Journal of Psychiatric Research 137 2021 126 130 10.1016/j.jpsychires.2021.02.056 33677216 Matthews K. Morgan I. Davis K. Estriplet T. Perez S. Crear-Perry J.A. Pathways to equitable and antiracist maternal mental health care: Insights from Black women stakeholders Health Affairs 40 10 2021 1597 1604 10.1377/hlthaff.2021.00808 34606342 McCall-Hosenfeld J.S. Phiri K. Schaefer E. Zhu J. Kjerulff K. Trajectories of depressive symptoms throughout the peri- and postpartum period: Results from the first baby study Journal of Women’s Health 25 11 2016 1112 1121 10.1089/jwh.2015.5310 McMillan I.F. Armstrong L.M. Langhinrichsen-Rohling J. Transitioning to parenthood during the pandemic: COVID-19 related stressors and first-time expectant mothers’ mental health Couple and Family Psychology 10 3 2021 179 189 10.1037/cfp0000174 Miller E.S. Saade G.R. Simhan H.N. Monk C. Haas D.M. Silver R.M. Grobman W.A. Trajectories of antenatal depression and adverse pregnancy outcomes American Journal of Obstetrics and Gynecology 226 1 2022 108.e1 108.e9 10.1016/j.ajog.2021.07.007 Mollard E. Kupzyk K. Moore T. Postpartum stress and protective factors in women who gave birth in the United States during the COVID-19 pandemic Women’s Health 17 2021 10.1177/17455065211042190 Article 17455065211042190 Morency J.-D. Malenfant É.C. MacIsaac S. Immigration and diversity: Population projections for Canada and its regions, 2011 to 2036. Statistics Canada https://www150.statcan.gc.ca/n1/pub/91-551-x/91-551-x2017001-eng.htm 2017, January 25 Moyer C.A. Compton S.D. Kaselitz E. Muzik M. Pregnancy-related anxiety during COVID-19: A nationwide survey of 2740 pregnant women Archives of Women’s Mental Health 23 6 2020 757 765 10.1007/s00737-020-01073-5 Mukherjee S. Trepka M.J. Pierre-Victor D. Bahelah R. Avent T. Racial/ethnic disparities in antenatal depression in the United States: A systematic review Maternal and Child Health Journal 20 9 2016 1780 1797 10.1007/s10995-016-1989-x 27016352 Office of Management and Budget Revisions to the standards for the classification of federal data on race and ethnicity. Federal Register https://www.govinfo.gov/content/pkg/FR-1997-10-30/pdf/97-28653.pdf 1997 Ollivier R. Aston D.M. Price D.S. Sim D.M. Benoit D.B. Joy D.P. Nassaji N.A. Mental health & parental concerns during COVID-19: The experiences of new mothers amidst social isolation Midwifery 94 2021 10.1016/j.midw.2020.102902 Article 102902 Omowale S.S. Casas A. Lai Y.-H. Sanders S.A. Hill A.V. Wallace M.L. Mendez D.D. Trends in stress throughout pregnancy and postpartum period during the COVID-19 pandemic: Longitudinal study using ecological momentary assessment and data from the postpartum mothers mobile study JMIR Mental Health 8 9 2021 10.2196/30422 Article e30422 Perzow S.E.D. Hennessey E.-M.P. Hoffman M.C. Grote N.K. Davis E.P. Hankin B.L. Mental health of pregnant and postpartum women in response to the COVID-19 pandemic Journal of Affective Disorders Reports 4 2021 10.1016/j.jadr.2021.100123 Article 100123 Pew Research Center Around the world, more say immigrants are a strength than a burden https://www.pewresearch.org/global/2019/03/14/around-the-world-more-say-immigrants-are-a-strength-than-a-burden/ 2019, March 14 Pew Research Center Unemployment rate is higher than officially recorded, more so for women and certain other groups June 30 https://www.pewresearch.org/fact-tank/2020/06/30/unemployment-rate-is-higher-than-officially-recorded-more-so-for-women-and-certain-other-groups/ 2020 Pew Research Center Key findings about U.S. immigrants August 20 https://www.pewresearch.org/fact-tank/2020/08/20/key-findings-about-u-s-immigrants/ 2020 Pew Research Center Social media fact sheet https://www.pewresearch.org/internet/fact-sheet/social-media/ 2021, April 7 Pope R. Ganesh P. Miracle J. Brazile R. Wolfe H. Rose J. Gullett H. Structural racism and risk of SARS-CoV-2 in pregnancy EClinicalMedicine 37 2021 10.1016/j.eclinm.2021.100950 Article 100950 Rogers T.N. Rogers C.R. VanSant-Webb E. Gu L.Y. Yan B. Qeadan F. Racial disparities in COVID-19 mortality among essential workers in the United States World Medical & Health Policy 12 3 2020 311 327 10.1002/wmh3.358 32837779 Ross P.T. Hart-Johnson T. Santen S.A. Zaidi N.L.B. Considerations for using race and ethnicity as quantitative variables in medical education research Perspectives on Medical Education 9 5 2020 318 323 10.1007/s40037-020-00602-3 32789666 Shavers V.L. Lynch C.F. Burmeister L.F. Knowledge of the Tuskegee study and its impact on the willingness to participate in medical research studies Journal of the National Medical Association 92 12 2000 563 572 11202759 Shavers V.L. Lynch C.F. Burmeister L.F. Racial differences in factors that influence the willingness to participate in medical research studies Annals of Epidemiology 12 4 2002 248 256 10.1016/s1047-2797(01)00265-4 11988413 Shuman C.J. Morgan M.E. Pareddy N. Chiangong J. Veliz P. Peahl A. Dalton V. Associations among postpartum posttraumatic stress disorder symptoms and COVID-19 pandemic-related stressors Journal of Midwifery & Women’s Health 67 5 2022 626 634 10.1111/jmwh.13399 Silverman M.E. Burgos L. Rodriguez Z.I. Afzal O. Kalishman A. Callipari F. Loudon H. Postpartum mood among universally screened high and low socioeconomic status patients during COVID-19 social restrictions in New York City Scientific Reports 10 1 2020 10.1038/s41598-020-79564-9 Article 22380 Silverman M.E. Medeiros C. Burgos L. Early pregnancy mood before and during COVID-19 community restrictions among women of low socioeconomic status in New York City: A preliminary study Archives of Women’s Mental Health 23 6 2020 779 782 10.1007/s00737-020-01061-9 Soffer M.D. Adams Z.M. Chen Y.S. Fox N.S. Risk factors for positive postpartum depression screen in women with private health insurance and access to care Journal of Maternal-Fetal & Neonatal Medicine 32 24 2019 4154 4158 10.1080/14767058.2018.1484096 29852802 Statistics Canada Maternal mental health in Canada, 2018/2019 https://www150.statcan.gc.ca/n1/daily-quotidien/190624/dq190624b-eng.htm 2019, June 24 Statistics Canada Visible minority of a person https://www23.statcan.gc.ca/imdb/p3Var.pl?Function=DEC&Id=45152 2021, January 11 Statistics Canada Census profile, 2021 census of population https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/index.cfm?Lang=E 2022, February 9 Suwalska J. Napierała M. Bogdański P. Łojko D. Wszołek K. Suchowiak S. Suwalska A. Perinatal mental health during COVID-19 pandemic: An integrative review and implications for clinical practice Journal of Clinical Medicine 10 11 2021 10.3390/jcm10112406 Article 2406 Ta Park V.M. Goyal D. Nguyen T. Lien H. Rosidi D. Postpartum traditions, mental health, and help-seeking considerations among Vietnamese American women: A mixed-methods pilot study Journal of Behavioral Health Services Research 44 3 2017 428 441 10.1007/s11414-015-9476-5 26276422 Ta Park V.M. Goyal D. Suen J. Win N. Tsoh J. Chinese American women’s experiences with postpartum depressive symptoms and mental health help-seeking behaviors MCN. The American Journal of Maternal/Child Nursing 44 3 2019 150 156 10.1097/NMC.0000000000000518 31034454 Taylor J.K. Structural racism and maternal health among Black women Journal of Law, Medicine & Ethics 48 3 2020 506 517 10.1177/1073110520958875 Tricco A.C. Lillie E. Zarin W. O’Brien K. Colquhoun H. Kastner M. Straus S.E. A scoping review on the conduct and reporting of scoping reviews BMC Medical Research Methodology 16 1 2016 10.1186/s12874-016-0116-4 Article 15 Urban Indian Health Institute Best practices for American Indian and Alaska Native data collection https://aipi.asu.edu/sites/default/files/best-practices-for-american-indian-and-alaska-native-data-collection.pdf 2022, August 6 U.S. Census Bureau Quick facts: United States https://www.census.gov/quickfacts/fact/table/US/PST045219 2021 U.S. Census Bureau 2020 U.S. population more racially and ethnically diverse than measured in 2010 https://www.census.gov/library/stories/2021/08/2020-united-states-population-more-racially-ethnically-diverse-than-2010.html 2021 Ward E.C. Wiltshire J.C. Detry M.A. Brown R.L. African American men and women’s attitude toward mental illness, perceptions of stigma, and preferred coping behaviors Nursing Research 62 3 2013 185 194 10.1097/NNR.0b013e31827bf533 23328705 Webber-Ritchey K.J. Aquino E. Ponder T.N. Lattner C. Soco C. Spurlark R. Simonovich S.D. Recruitment strategies to optimize participation by diverse populations Nursing Science Quarterly 34 3 2021 235 243 10.1177/08943184211010471 34212805 Wheeler J.M. Misra D.P. Giurgescu C. Stress and coping among pregnant Black women during the COVID-19 pandemic Public Health Nursing 38 4 2021 596 602 10.1111/phn.12909 33844868 Wisner K.L. Sit D.K. McShea M.C. Rizzo D.M. Zoretich R.A. Hughes C.L. Hanusa B.H. Onset timing, thoughts of self-harm, and diagnoses in postpartum women with screen-positive depression findings JAMA Psychiatry 70 5 2013 490 498 10.1001/jamapsychiatry.2013.87 23487258 Yan H. Ding Y. Guo W. Mental health of pregnant and postpartum women during the coronavirus disease 2019 pandemic: A systematic review and meta-analysis Frontiers in Psychology 11 2020 10.3389/fpsyg.2020.617001 Article 617001 Yildiz P.D. Ayers S. Phillips L. The prevalence of posttraumatic stress disorder in pregnancy and after birth: A systematic review and meta-analysis Journal of Affective Disorders 208 2017 634 645 10.1016/j.jad.2016.10.009 27865585	36462529	PMC9710717	NO-CC CODE	2022-12-02 23:21:29	no	J Obstet Gynecol Neonatal Nurs. 2022 Nov 30; doi: 10.1016/j.jogn.2022.11.003	utf-8	J Obstet Gynecol Neonatal Nurs	2,022	10.1016/j.jogn.2022.11.003	oa_other
==== Front Semin Vasc Surg Semin Vasc Surg Seminars in Vascular Surgery 0895-7967 1558-4518 Published by Elsevier Inc. S0895-7967(21)00028-4 10.1053/j.semvascsurg.2021.04.003 Review Article The impact of the COVID-19 pandemic on wellness among vascular surgeons Drudi Laura M. a Nishath Thamanna b Ma Xiya c Mouawad Nicolas J. d O'Banion Leigh Ann e Shalhub Sherene f⁎ a Division of Vascular Surgery, Department of Surgery, Centre Hospitalier de l'Université de Montréal, Montreal, Quebec, Canada b Division of Vascular Surgery, Department of Surgery, University of Washington School of Medicine, Seattle, WA c Faculty of Medicine, Université de Montréal, Montreal, Quebec, Canada d Vascular and Endovascular Surgery, McLaren Health System, Bay City, MI e Division of Vascular and Endovascular Surgery, University of California San Francisco–Fresno, Fresno, CA f Division of Vascular Surgery, Department of Surgery, University of Washington School of Medicine, 1959 NE Pacific Street, Box 356410, Seattle, WA, 98195 ⁎ Corresponding author. 21 5 2021 6 2021 21 5 2021 34 2 4350 © 2021 Published by Elsevier Inc. 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 coronavirus disease 2019 (COVID-19) pandemic has placed significant strain on the health and welfare of all health care professionals, including vascular surgeons. This review summarizes the implications of the pandemic on the health and wellness of surgeons and trainees, with a particular focus on those in vascular surgery (VS). A literature review was completed using common resource databases. We provide a brief history of burnout in VS and explore burnout and wellness in VS during this unprecedented pandemic. We then offer recommendations to address mental health needs by the VS workforce and highlight opportunities to address the gaps in the literature. The impact of COVID-19 on the professional and personal lives of surgeons and trainees in VS is notable. More than half of vascular surgeons reported some degree of anxiety. Factors associated with anxiety and burnout include COVID-19 exposure, moral injury, practice changes, and financial impacts. Trainees appeared to have more active coping strategies with dampened rates of anxiety compared to those in practice. Women appear to be disproportionately affected by the pandemic, with higher rates of anxiety and burnout. Groups underrepresented in medicine seemed to have more resilience when it came to burnout, but struggled with other inequities in the health care environment, such as structural racism and isolation. Strategies for addressing burnout include mindfulness practices, exercise, and peer and institutional support. The COVID-19 pandemic has had a substantial mental health impact on the VS workforce globally, as shifts were made in patient care, surgical practice, and work–home life concerns. ==== Body pmc1 Introduction Wellness and occupational burnout are increasingly recognized as formidable forces in the retention (or lack thereof) of the health care workforce [1]. Wellness is defined as “a quality or state of being in good health” [2], while burnout is a mental state of emotional exhaustion, depersonalization, and a diminished sense of personal accomplishment [3,4]. Several factors have been identified as causes of occupational burnout in medicine: work environment dissatisfaction, work–life integration challenges, ergonomic stresses, and the everchanging health care environment affecting health care resource use [5]. As of February 2021, there have been more than 2.4 million deaths secondary to coronavirus disease 2019 (COVID-19) [6]. Acknowledging the lives lost and the lives impacted by COVID-19, it is clear that individuals, communities, and global societies have suffered because of this pandemic. In addition, the pandemic resulted in an unprecedented strain on health care systems and caused distress to health care professionals [7], [8], [9]. In this review, we aim to explore the implications of the COVID-19 pandemic on wellness and burnout of vascular surgeons in practice and in training. We review the impact of the COVID-19 pandemic on patients cared for by vascular surgeons, on the practices and careers of vascular surgeons, and on the institutions that support vascular surgery (VS) practices. We then discuss how wellness and burnout are shaped by the impact of the pandemic on vascular surgeons, with specific attention paid to the differential impact on women and groups underrepresented in medicine (URiM). Finally, we review lessons learned to mitigate burnout and increase a sense of wellness during these challenging times to address the ongoing wellness needs of vascular surgeons. 2 Methods This literature review sought to provide a brief history of burnout and explore the current landscape of burnout and wellness in VS during the unprecedented COVID-19 pandemic. A literature review was completed using the PubMed/MEDLINE, Scopus, EMBASE, and Google Scholar databases for articles from October 2019 to February 2021. The literature was reviewed for articles relevant to “COVID-19,” “vascular surgery,” “wellness,” and “burnout” in English and French. The review was further broadened to include other surgical and medical subspecialties when there was a knowledge and evidence gap specific to VS. Key concepts were identified, and sources of evidence were summarized as they pertain to patient care, practice changes, and education. The impact on women and those groups URiM was also summarized where applicable. The American Association of Medical Colleges defines URiM as “racial and ethnic populations that are underrepresented in the medical profession relative to their numbers in the general population” [10]. Finally, strategies for coping and resilience were summarized. 3 COVID-19 impact on patients cared for by vascular surgeons The peak of the COVID-19 pandemic led to the redistribution of health care resources and cancellation of elective surgical operations. From March to May 2020, during a 12-week period of peak disruption to hospital services due to COVID-19, the COVIDSurg Collaborative estimated that more than 28 million operations would be cancelled or postponed globally [11]. Similar trends were being evaluated by the Vascular and Endovascular Research Network's (VERN) COVER (Covid-19 Vascular Service) study [12], which is a global survey aimed at evaluating the disruption in vascular services. This disruption was a considerable strain on all health care systems and has been associated with distress in health care personnel [13]. One recent survey of vascular surgeons in practice and training during the peak of the pandemic showed that the overwhelming majority of the survey respondents (91.7%) had cancellations of elective VS operations, with the Northeast and Southeast regions of the United States having the most case cancellations proportional to the pandemic impact in these areas [14]. Other surveys also demonstrated an overwhelming majority of vascular surgeons were performing urgent and emergent cases only [15], with significant VS practice changes [16]. Similar events were simultaneously happening internationally during the peak of the pandemic [11,17]. During that time, triaging of patients with VS needs became critical and the Vascular Surgery Activity Condition (VASCCON) [18] system was borne. VASCCON was modeled after DEFCON (the Defense Readiness Condition) used by the US military to describe the various stages of readiness in response to an external threat, such as a global pandemic. VASCCON offers a stepwise de-escalation of surgical activities from 5 (normal vascular surgery scheduling of cases) to 1 (no surgical activity) [18]. Given the global cancellation of elective cases and shifts in resource use, the COVID-19 pandemic had a considerable impact on the way patients with vascular disease were cared for. There was a considerable reduction in screening for abdominal aortic aneurysms and higher size thresholds for repair [19]. There was also a shift to treating aortic and peripheral arterial pathologies with endovascular interventions, even when open repair options were preferred in order to minimize intensive care unit and overall hospital stays [19,20]. This shift toward treating patients with a primary endovascular strategy during the pandemic was in addition to the already present endovascular shift being made in VS practice before the pandemic [14,21]. Furthermore, delays in providing timely vascular care were prevalent during the COVID-19 pandemic. One study [22] found that 4.5% of patients cared for by vascular surgeons experienced adverse events “attributed to a delay in surgery.” The adverse events were defined as disease progression, hospitalization, urgent operations, or death related to the index procedure or diagnosis [22]. For example, there was a trend of increased limb loss among patients with chronic limb-threatening ischemia. Musajee et al [23] demonstrated that patients with chronic limb-threatening ischemia being treated during the pandemic had significantly worse amputation-free survival and limb salvage compared to patients treated prepandemic. Furthermore, undergoing treatment during the pandemic was an independent predictor of worse primary patency and freedom from major adverse limb events [23]. This was attributed to treatment delays associated with late presentations because of social distancing and lockdown measures [23,24]. The long-term sequelae of COVID-19 pandemic on clinical and patient-centered outcomes in VS and other surgical specialties are unknown, and will become a critical piece as we retrospectively evaluate the larger impact of the pandemic on patient care [25,26]. In the decade before the pandemic, telemedicine was used modestly and predominantly in primary and mental health care [27]. The adoption of telemedicine grew exponentially in response to the social distancing and lockdown measures necessitated by the pandemic. To make telemedicine a feasible alternative to in-person clinic visits, many regulatory health agencies allowed temporary waivers for reimbursement and HIPAA (Health Insurance Portability and Accountability Act) restrictions [27], [28], [29], [30]. Vascular surgeons quickly adapted and embraced telemedicine as a tool for triaging and postoperative monitoring for patients with vascular disease [31], [32], [33]. In the outpatient setting, patients cared for by vascular surgeons had significantly reduced travel distances, commuting time, and costs associated with telemedicine utilization. Also, studies have found that there was significant patient satisfaction concerning their health care experience using telemedicine [34,35]. 4 COVID-19 impact on vascular surgeons and burnout Burnout is associated with worsening performance in surgeons and trainees and can lead to increased stress and risk of depression [36], [37], [38]. Before the pandemic, vascular surgeons appeared to be at a high risk for burnout, suicide, low career satisfaction, and quality of life [39], [40], [41]. Recognizing the association between surgeon burnout, patient outcomes, and workforce retention [42], [43], [44], the Society for Vascular Surgery (SVS) Wellness Taskforce was created in 2017 and was charged with assessing and addressing vascular surgeon burnout, with the aim of improving member wellness. In 2018, the Wellness Task Force disseminated a survey to the SVS society members to identify the prevalence of burnout, depression, and suicidal ideation among practicing vascular surgeons to guide future policy, advocacy, and member programs to support this workforce crisis. This demonstrated that one-third of survey respondents reported burnout, 37% endorsed symptoms of depression in the past month, and 8% indicated that they had considered suicide in the past year [45,46]. Addressing burnout among vascular surgeons with the compounding stress of the COVID-19 pandemic became a central area of focus for the Wellness Task Force, who dedicated their efforts to mitigate distress in the VS workforce. The impact of occupational and personal stressors on the risk of physician burnout during the COVID-19 pandemic is growing (Table 1 ). In April of 2020, the SVS Wellness Task Force sponsored an anonymous Pandemic Practice, Anxiety, Coping, and Support Survey designed to capture the cross-sectional experiences of vascular surgeons and trainees during a time of uncertainty. Of the 1,609 respondents, 23.3% reported moderate and severe anxiety, and more than half reported some degree of anxiety. Factors that were significantly associated with this reporting were directly related to the day-to-day COVID-19 impact, such as staying in a separate room at home or at an alternate lodging, donning and doffing personal protective equipment, worrying about potential treatment delays with adverse patient outcomes, and financial concerns [47]. Furthermore, other factors associated with self-reported anxiety were female sex and avoidant coping strategies, such as substance abuse, disengagement, and self-blame [47].Table 1 The impact of COVID-19 in health care professionals and vascular surgeon burnout. Table 1Evidence for burnout during COVID-19 pandemic among health care workers [48] Evidence for burnout during COVID-19 pandemic among vascular specialists [42,47] Burnout was prevalent in health care workers during the COVID-19 pandemic. There were higher rates of burnout in nonfrontline workers compared to frontline workers. Burnout was associated with female gender, long work hours, fears of infections, exposed to COVID-19 infection, and lack of perceived support by friends. Prevent burnout and occupational stress through support, self-awareness and mindfulness activities. Half of vascular surgeons reported some degree of anxiety with >20% reporting moderate or severe anxiety. Burnout in vascular surgery trainees is associated with depression, perceived stress, and lower levels of social support and self-efficacy. Anxiety associated with surgeons having a separate room at home or staying at the hospital or a hotel after work, donning and doffing personal protective equipment, worry about potential adverse patient outcomes due to care delay, and financial concerns. Active coping strategies was associated with less anxiety. Abbreviation: COVID-19, coronavirus disease 2019. Current evidence suggests that the type of physician practice may influence how susceptible a physician is to developing burnout during a crisis. A systematic review highlighted how physicians’ prepandemic work setting may influence the risk of burnout during the COVID-19 pandemic, with frontline physicians paradoxically less affected than others [48]. This phenomenon may be explained by the nature of frontline physicians’ work, where uncertainty is part of their professional daily routine, potentially offering them baseline resilience and coping strategies in an environment of ongoing practice changes during the pandemic [49]. However, this perceived advantage may be outweighed by a greater risk of exposure to COVID-19, with resulting higher rates of distress and occupational burnout [50]. Although the practices of vascular surgeons vary across the United States and globally, only 23% of vascular surgeons had moderate to severe anxiety and many were coping well with active coping strategies, such as emotional support, positive reframing, and strategic planning [47]. 4.1 Moral injury The inability to provide optimal high-quality care amidst this pandemic has led to a resounding moral injury among health care providers [51]. Moral injury occurs when individuals perpetuate acts that go against their value system or moral beliefs. The concept of moral injury was first described in the context of service members returning from the Vietnam War with symptoms resembling post-traumatic stress disorder related to conflicts with their value systems [52]. Similarly, surgeons and physicians may have numerous conflicts weighing “individual clinical ethics versus public health ethics, best medical practices versus resource scarcity, and expert practice versus practicing at the edge or beyond one's competencies” [53]. During the pandemic, cancellations of operations, delays in care, lack of resources, and shifts in resource management may have led to multiple moral insults resulting in a compounding effect of unresolved conflicts leading to moral distress and occupational burnout. 4.2 Vascular surgery practice changes The day-to-day activities of surgeons, including in-person clinics, multidisciplinary meetings, education, and professional activities, were quickly replaced with virtual clinics and meetings. Although the virtual interaction facilitated ongoing access for physician support, complex decision-making, and access to care [19], it may have also led to psychological distress, a sense of isolation, mental fatigue, and increasing sedentary time [54]. This is speculative as the impact of a virtual world on wellness has not yet been extensively explored among physicians and surgeons during the pandemic. Previous work has focused on the effect of screen time among adolescents and children reporting more depressive symptoms, burnout, and sleep disturbances, with a higher rate impact among females compared to males [55], [56], [57]. It has also been demonstrated that there appears to be a dose-dependent response with longer duration of screen time having a more profound impact on levels of stress and anxiety in this young population [58], [59], [60]. It is unclear whether these findings translate to adults or physicians during the pandemic. Furthermore, the prolonged use of smart devices might also have an impact on physical health due to increased sedentary time; however, this has not been shown consistently [61]. The hunched over postures while looking at smart devices and screens for extended periods of time may lead to long-term ergonomic issues impacting physical and emotional health [62]. 4.3 The financial impact of COVID-19 on vascular surgery practices The redistribution of resources and suspension of elective surgical operations has also placed financial stress in VS and across all interventional disciplines. In the United States, the financial impact in a tertiary care hospital VS division resulted in losses ranging from 39% to 65% compared to the prepandemic period the previous year [13]. In a recent survey, solo, private, and community-based vascular surgeons reported significantly higher anxiety levels related to economic changes compared with those working in academic centers, given the extreme reduction of revenues and a lack of significant financial reserves to absorb the economic impact [47]. 5 Career-related gender differences in the impact of COVID-19 The struggles women face with professional activities and work–life integration were more pronounced during the COVID-19 pandemic. The impact of the pandemic on women in medicine is widely heterogeneous, depending on career stage, scope of practice (teaching, research, private), and personal life commitments [63]. Collectively, however, the pandemic adds to the existing gender-related systemic inequities that plague trainees and physicians, threatening to undo the significant progress made in the past decades, and leaving a legacy of difficulties for the generation to come. Among surgeons, women are more likely to experience anxiety, depression, and burnout compared to male surgeons [64,65]. This has been attributed to societal sexism [66,67], gender bias [68], workplace discrimination [69,70], and familial obligations that disproportionately affect women [48,63,71]. In the recent VS pandemic impact survey, vascular surgeons who are women reported higher rates of anxiety compared to vascular surgeons who are men [47]. Similarly, physician trainees who are women reported significantly more stress at work than their peers who are men [50]. A Brazilian study of head and neck surgeons also reported how women experience significantly more symptoms of anxiety and burnout, and decreased productivity at work and at home [72]. There is a pervasive notion that women in medicine are more likely to compromise their career advancement in order to meet home- or family-related needs compared to men in medicine (maternal bias) [63]. In addition, the pandemic-related increase in demands of childcare, meal preparation, and grocery shopping has disproportionately affected women, including physicians who are women. This has translated to fewer hours for clinical activities [63]. Telemedicine and remote teaching can be significantly more challenging for women who need to simultaneously care for children at home [73]. Before the pandemic, more than half of female vascular surgeons did not believe they had enough balance between their professional and personal lives [46]. Furthermore, women compared to men were more likely to have had a recent conflict between work and home responsibilities (68.5% v 57.4%; P < .01) and to have resolved this conflict in favor of work (57.5 v 42.8%; P < .005) [46]. Beyond the impact on clinical work, the COVID-19 pandemic has taken a toll on women's involvement in research. JAMA Surgery noticed a 4%, 6%, and 5% decrease in submitted manuscripts with female first author, last author, and corresponding author, respectively, when comparing the two peak months of the first COVID-19 wave relative to the same time period 1 year prior [73]. Similarly, while COVID-19-related research soared since March 2020, the opposite was true for the proportion of female authors of COVID-19 submissions [74]. These changes can further amplify the difficulties that women in surgery face when seeking academic promotion and tenure, which is based largely on academic advancement with research productivity and publications. Furthermore, there is an underrepresentation of women in leadership positions in medicine and surgery. This leads to a lack of understanding of the perspectives of a diverse group of women and rather allows for making decisions based on inaccurate stereotypical assumptions based on an outdated notion of the “female” architype [75]. Although initiatives to promote diversity, equity, and inclusion in medicine and surgery have increased amidst the COVID-19 pandemic, it has drawn attention to gender inequities. Our profession can and should support and respect all physicians, irrespective of gender, and create work environments that achieve workplace equity and equality, without stigmatizing women or derailing their career progression [63]. 6 The impact of COVID-19 on URiM physicians Several studies on URiM physicians have demonstrated inconsistent findings on whether physicians who are URiM have greater burnouts compared to their White counterparts [76]. However, current wellness and burnout assessment tools do not take into account the experiences faced by URiM physicians, such as structural racism, sexism, discrimination, isolation, lack of inclusion, and social support [77,78]. In addition, the current instruments that categorize URiM physicians rely on the US Census Bureau classification and fail to capture the experiences of some physicians, such as those who identify as being of Middle Eastern and North African descent, but still face discrimination due to their origin or religion. Discrimination of URiM may manifest in the devaluing of research and skill sets, resulting in race-based assumptions of competency and workplace micro-aggressions [79]. Furthermore, URiM physicians face the “minority tax,” which is the expectation to participate in additional, largely unpaid, responsibilities related to diversity, such as mentoring and advising URiM students [80]. Some have suggested that URiM physicians may face lower rates of burnout because of the resilience, coping, and social support they have already developed to face the perpetual battle of adversity and inequities facing URiM physicians [81]. 7 The impact of COVID-19 on surgical trainees Surgical education and training were also impacted due to suspensions or delays of conferences, displacement of residents and attendings to emergency and critical care services instead of their specialty training, and research projects being suspended or re-aligned to tackle the COVID-19 pandemic [16,82]. A subset analysis of the Pandemic Practice, Anxiety, Coping, and Support Survey was distributed to VS trainees, integrated residents, and fellows in the United States. Vascular trainees were noted to have significant changes in clinical responsibilities: 91% reporting cancellation of elective procedures, 82% having call schedule changes, 24% having duties other than those related to VS, and 24% participating in outpatient care delivery. Major stressors included concerns about educational and professional development, infection risk to family/friends, and impact of care delay on patients. Despite these considerable changes to clinical responsibilities and personal and professional stressors, there were low rates of anxiety in VS trainees and VS trainees employed mostly active coping strategies and used online support systems [82]. 8 Lessons learned from the COVID-19 pandemic Despite the devastating impact of the COVID-19 pandemic, there are some benefits worth acknowledging as a “silver lining.” First, this pandemic provided a time of global introspection and opportunity for self-reflection [83]. The year 2020 was aptly named the year of vision and clarity. This pandemic opened our eyes with 20/20 vision to the fragile health care system; the care we deliver to our patients, colleagues, families, and ourselves; and the importance of wellness in a time of extreme uncertainty. This pandemic has also ushered in a wave of international collaborations and partnerships that would have been unlikely in the past, or would have occurred over a longer expanse of time. Many were necessary due to the urgency of the pandemic, spanning from policymakers and the pharmaceutical industry to physicians and scientists, but therein rose opportunity [31], [32], [33], [34]. Others increased collaborative work with the widespread and rapid adoption of teleconferencing and group communications platforms, such as WhatsApp. For example, COVIDSurg Collaborative is an international collaborating group of surgeons and anesthetists who work on a spectrum from offering guidance on how to deliver surgical services appropriately to assessing the outcomes of surgery in patients diagnosed with COVID-19 [35,36]. There are countless more formations, like the COVID-19 Clinical Research Coalition targeting resources available in low- and middle-income countries, that have been critical in battling this pandemic, in addition to furthering preparedness for future outbreaks and contributing to research for other fields of medicine and public health [37]. 9 Mitigating burnout: strategies and solutions The World Health Organization defines self-care as “what people do for themselves to establish and maintain health, and to prevent and deal with illness,” including hygiene, nutrition, lifestyle, environmental, and socioeconomic factors. Coping with the adverse effects of burnout compounded by the COVID-19 pandemic was a necessity. Critical to mitigating the detrimental effects of burnout is first formulating strategies and implementing skills that promote increased wellbeing and improve resilience strategies, not only during the burnout crisis in this pandemic period but at large. The following is a nonexhaustive list of proposed strategies for mitigating burnout.1 Mindfulness: Mindfulness is a practice of “compassionate and intentional awareness” [84]. Mindfulness transitioned into health care pervasively as a practice of “nonjudgmental, curious, and self-compassionate awareness of one's moment-to-moment experiences” [85]. Over the past decade, physicians have begun to delve into the concept of mindfulness in efforts to reduce burnout and improve quality of life. Furthermore, providing physicians with tools to implement mindfulness has been shown to decrease burnout and improve wellbeing in health care providers [86]. 2 Exercise: Developing a consistent exercise routine has been shown to decrease anxiety, depression, and burnout [87]. With social distancing measures and lockdown measures in place, group exercises transitioned to an online platform fostering a safe and protective community. Innovative solutions to foster wellness, physical health, and collegiality were created during this pandemic. For example, physicians and surgeons created Peloton groups or bubbles to exercise in a safe space to foster and enhance their own wellness [88]. 3 Peer support: The necessary public health principles of social distancing and density reduction have been extremely challenging due to the lack of in-person interactions. This new era has forced a virtual format for regular day-to-day activities and relationships. A variety of both formal and informal peer-to-peer connections have blossomed from the devastation of the pandemic. Amid the COVID-19 strain on physicians, vascular surgeons were assertive in building strong peer support structures. The use of social media allowed vascular surgeons to share best practices in real time. Multiple social media platforms have been used by vascular surgeons, including Twitter, Facebook, LinkedIn, WhatsApp, as well as SVS Connect. These various platforms have allowed surgeons to meet others globally, share concerns and worries, provide peer support, and assist in clinical questions and research dissemination. Importantly, whichever medium was used, it provided the opportunity for closeness despite being physically apart [89]. 4 Institutional support: The SVS organizational leadership recognized the importance of peer and community support for its members during the pandemic. The SVS developed formal programs through the SVS Wellness Taskforce, as well as dedicated town halls [90]. These town halls focused on clinical concerns, as well as wellness and coping strategies. Peer support awareness and discussions were at the forefront of these endeavors and were very well attended, with more than 4,500 viewers across the livestreaming event and additional views on YouTube, Twitter, and Facebook. 10 Conclusions The COVID-19 pandemic has affected physician wellness through the multidimensional aspects of professional and personal lives, and the interplay between the two. This literature review was designed to explore the current landscape of the impact of COVID-19 on vascular surgeon wellbeing, and we may have failed to include evidence that might have been captured in a rigorous systematic process. As we continue to evolve in these times of uncertainly, thoughtful and mindful leadership addressing the psychological impact and open nonjudgmental dialogue to address these mental health concerns is paramount to maintaining a healthy workforce. Identifying vulnerable populations at risk for burnout and implementing institutional policies and strategies to address these psychological concerns should be in place for those needing mental health resources and services. Health systems must prioritize the health and wellbeing of their providers before, during, and after crises like the COVID-19 pandemic. 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 Shanafelt TD Balch CM Bechamps GJ Burnout and career satisfaction among American surgeons Ann Surg 250 2009 463 471 19730177 2 Merriam-Webster. Wellness. Available at: https://www.merriam-webster.com/dictionary/wellness. Accessed February 23, 2021. 3 Shanafelt TD Bradley KA Wipf JE Burnout and self-reported patient care in an internal medicine residency program Ann Intern Med 136 2002 358 367 11874308 4 Dimou FM Eckelbarger D Riall TS. Surgeon burnout: a systematic review J Am Coll Surg 222 2016 1230 1239 27106639 5 Campbell DA Jr Sonnad SS Eckhauser FE Burnout among American surgeons Surgery 130 2001 696 702 discussion 5 11602901 6 Worldometer COVID-19 Coronavirus pandemic February 15, 2021 Available at: https://www.worldometers.info/coronavirus/?utm_campaign=homeAdvegas1? UpdatedAccessed February 15, 2021 7 Di Tella M Romeo A Benfante A Mental health of healthcare workers during the COVID-19 pandemic in Italy J Eval Clin Pract 26 2020 1583 1587 32710481 8 Greenberg N Docherty M Gnanapragasam S Managing mental health challenges faced by healthcare workers during covid-19 pandemic BMJ 368 2020 m1211 32217624 9 Choudhury T Debski M Wiper A COVID-19 pandemic: looking after the mental health of our healthcare workers J Occup Environ Med 62 2020 e373 e376 32730043 10 Association of American Medical Colleges. Underrepresented in medicine definition. Available at: https://www.aamc.org/what-we-do/diversity-inclusion/underrepresented-in-medicine. Accessed March 2, 2021. 11 Collaborative COVIDSurg Elective surgery cancellations due to the COVID-19 pandemic: global predictive modelling to inform surgical recovery plans BJS 107 2020 1440 1449 12 The Vascular and Endovascular Research Network Executive Committee The COvid-19 Vascular sERvice (COVER) study: an International Vascular and Endovascular Research Network (VERN) Collaborative Study Assessing the Provision, Practice, and Outcomes of Vascular Surgery During the COVID-19 Pandemic Eur J Vasc Endovasc Surg 60 2020 156 157 32410815 13 Fang ZB Simons JP Judelson DR Financial implications of COVID-19 on a tertiary academic vascular surgery practice J Vasc Surg 2021 Feb 3 10.1016/j.jvs.2021.01.024 [Epub ahead of print] 14 Mouawad NJ Woo K Malgor RD The impact of the COVID-19 pandemic on vascular surgery practice in the United States J Vasc Surg 73 2021 772 779 e4 32889073 15 Latz CA Boitano LT Png CYM Early vascular surgery response to the COVID-19 pandemic: results of a nationwide survey J Vasc Surg 73 2021 372 380 32454233 16 Aziz F Bath J Smeds MR. Implications of the severe acute respiratory syndrome associated with the novel coronavirus-2 on vascular surgery practices J Vasc Surg 73 2021 4 11 e2 32891807 17 Metelmann IB Busemann A. Elective surgery in times of COVID-19: A two-centre analysis of postponed operations and disease-related morbidity and mortality Z Evid Fortbild Qual Gesundhwes 158–159 2020 62 65 18 Forbes TL. Vascular surgery activity condition is a common language for uncommon times J Vasc Surg 72 2020 391 392 32360373 19 Vascular and Endovascular Research Network (VERN) COVER Study Collaborative Global impact of the first coronavirus disease 2019 (COVID-19) pandemic wave on vascular services Br J Surg 107 2020 1396 1400 33405234 20 Al-Jabir A Kerwan A Nicola M Impact of the coronavirus (COVID-19) pandemic on surgical practice–part 2 (surgical prioritisation) Int J Surg 79 2020 233 248 32413502 21 Bozzani A Arici V Ticozzelli G Endovascular surgery during COVID-19 virus pandemic as a valid alternative to open surgery Ann Vasc Surg 71 2021 101 102 33157245 22 Sarfati MR Griffin CL Kraiss LW Vascular surgery triage during the COVID-19 pandemic J Vasc Surg 2020 Nov 27 10.1016/j.jvs.2020.11.026 [Epub ahead of print] 23 Musajee M Zayed H Thulasidasan N Impact of COVID-19 pandemic on the outcomes in patients with critical limb threatening ischaemia and diabetic foot infection Ann Surg 2020 Dec 22 10.1097/SLA.0000000000004677 [Epub ahead of print] 24 Bashar AHM Hakim ME Rahman MM Vascular surgery practice guidelines during COVID-19 pandemic in a setting of high work volume against limited resources: perspective of a developing country Ann Vasc Surg 70 2021 306 313 32889161 25 Brown TS Bedard NA Rojas EO The effect of the COVID-19 pandemic on electively scheduled hip and knee arthroplasty patients in the United States J Arthroplasty 35 2020 S49 S55 32376163 26 Poeran J Zhong H Wilson L Cancellation of elective surgery and intensive care unit capacity in New York state: a retrospective cohort analysis Anesth Analg 131 2020 1337 1341 33079852 27 Contreras CM Metzger GA Beane JD Telemedicine: patient-provider clinical engagement during the COVID-19 pandemic and beyond J Gastrointest Surg 24 2020 1692 1697 32385614 28 Lanham NS Bockelman KJ McCriskin BJ. Telemedicine and orthopaedic surgery: the COVID-19 pandemic and our new normal JBJS Rev 8 7 2020 e20 29 Shokri T Lighthall JG. Telemedicine in the era of the COVID-19 pandemic: implications in facial plastic surgery Facial Plast Surg Aesthet Med 22 2020 155 156 32302225 30 Eilenberg W Busch A Wagenhäuser M Vascular surgery in unreal times Eur J Vasc Endovasc Surg 60 2020 167 168 32605851 31 Hemingway JF Singh N Starnes BW. Emerging practice patterns in vascular surgery during the COVID-19 pandemic J Vasc Surg 72 2020 396 402 32361072 32 Ajibade A Younas H Pullan M Telemedicine in cardiovascular surgery during COVID-19 pandemic: a systematic review and our experience J Cardiac Sug 35 2020 2773 2784 33 Al-Jabir A Kerwan A Nicola M Impact of the coronavirus (COVID-19) pandemic on surgical practice–part 1 Int J Surg 79 2020 168 179 32407799 34 Paquette S Lin JC. Outpatient telemedicine program in vascular surgery reduces patient travel time, cost, and environmental pollutant emissions Ann Vasc Surg 59 2019 167 172 31077768 35 Erben Y Franco-Mesa C Hamid O Telemedicine in vascular surgery during the coronavirus disease 2019 (COVID-19) pandemic—a multi-site healthcare system experience J Vasc Surg 2020 Dec 16 10.1016/j.jvs.2020.12.012 [Epub ahead of print] 36 Lebares CC Guvva EV Ascher NL Burnout and stress among US surgery residents: psychological distress and resilience J Am Coll Surg 226 2018 80 90 29107117 37 Eckleberry-Hunt J Van Dyke A Lick D Changing the conversation from burnout to wellness: physician well-being in residency training programs J Grad Med Educ 1 2009 225 230 21975983 38 West CP Dyrbye LN Erwin PJ Interventions to prevent and reduce physician burnout: a systematic review and meta-analysis Lancet 388 10057 2016 2272 2281 27692469 39 Ji YD Robertson FC Patel NA Assessment of risk factors for suicide among US health care professionals JAMA Surg 155 2020 713 721 32520355 40 Balch CM Shanafelt TD Sloan JA Distress and career satisfaction among 14 surgical specialties, comparing academic and private practice settings Ann Surg 254 2011 558 568 21946217 41 Pulcrano M Evans SR Sosin M. Quality of life and burnout rates across surgical specialties: a systematic review JAMA Surg 151 2016 970 978 27410167 42 Cimbak N Stolarski A Moseley J Burnout leads to premature surgeon retirement: a nationwide survey J Surg Res 2 2019 159 169 43 Panagioti M Geraghty K Johnson J Association between physician burnout and patient safety, professionalism, and patient satisfaction: a systematic review and meta-analysis JAMA Intern Med 178 2018 1317 1331 30193239 44 Shanafelt TD Balch CM Bechamps G Burnout and medical errors among American surgeons Ann Surg 251 2010 995 1000 19934755 45 Coleman DM Money SR Meltzer AJ Vascular surgeon wellness and burnout: a report from the SVS wellness task force J Vasc Surg 2020 Nov 25 10.1016/j.jvs.2020.10.065 [Epub ahead of print] 46 Drudi LM, Mitchell EL, Chandra V, et al. A gender-based analysis of the predictors and sequela of burnout amongst practicing American vascular surgeons. J Vasc Surg. 47 Shalhub S Mouawad NJ Malgor RD Global vascular surgeons' experience, stressors, and coping during the coronavirus disease 2019 pandemic J Vasc Surg 73 2021 762 771 e4 32882345 48 Amanullah S Ramesh Shankar R The impact of COVID-19 on physician burnout globally: a review Healthcare (Basel) 8 2020 421 33105757 49 Dimitriu MCT Pantea-Stoian A Smaranda AC Burnout syndrome in Romanian medical residents in time of the COVID-19 pandemic Med Hypotheses 144 2020 109972 50 Kannampallil TG Goss CW Evanoff BA Exposure to COVID-19 patients increases physician trainee stress and burnout PLoS One 15 8 2020 e0237301 51 Coleman DM. Wellness: of dreaming in blood and the ‘moral injury’ contained in a healthcare setting Vascular Specialist March 25, 2020 Available at: https://vascularspecialistonline.com/wellness-of-dreaming-in-blood-and-the-moral-injury-contained-in-a-healthcare-setting/ PublishedAccessed March 4, 2021 52 Currier JM Holland JM Malott J. Moral injury, meaning making, and mental health in returning veterans J Clin Psychol 71 2015 229 240 25331653 53 Dunham AM Rieder TN Humbyrd CJ. A Bioethical perspective for navigating moral dilemmas amidst the COVID-19 pandemic J Am Acad Orthop Surg 28 2020 471 476 32282442 54 Wolf CR. Virtual platforms are helpful tools but can add to our stress May 14, 2020 Psychology Today Available at: https://www.psychologytoday.com/us/blog/the-desk-the-mental-health-lawyer/202005/virtual-platforms-are-helpful-tools-can-add-our-stress PublishedAccessed February 7, 2021 55 Twenge JM Campbell WK. Associations between screen time and lower psychological well-being among children and adolescents: evidence from a population-based study Prev Med Rep 12 2018 271 283 30406005 56 Mheidly N Fares MY Fares J. Coping with stress and burnout associated with telecommunication and online learning Front Public Health 8 2020 574969 57 Sansone RA Sansone LA. Cell phones: the psychosocial risks Innov Clin Neurosci 10 2013 33 37 58 Višnjić A Veličković V Sokolović D Relationship between the manner of mobile phone use and depression, anxiety, and stress in university students Int J Environ Res Public Health 15 2018 697 29642471 59 Khouja JN Munafò MR Tilling K Is screen time associated with anxiety or depression in young people? Results from a UK birth cohort BMC Public Health 19 2019 82 30654771 60 Madhav KC Sherchand SP Sherchan S. Association between screen time and depression among US adults Prev Med Rep 8 2017 67 71 28879072 61 Schmidt SCE Anedda B Burchartz A Physical activity and screen time of children and adolescents before and during the COVID-19 lockdown in Germany: a natural experiment Sci Rep 10 1 2020 21780 33311526 62 Fares J Fares MY Fares Y. Musculoskeletal neck pain in children and adolescents: risk factors and complications Surg Neurol Int 8 2017 72 28584675 63 Brubaker L. Women physicians and the COVID-19 pandemic JAMA 324 2020 835 836 32735329 64 Hu YY Ellis RJ Hewitt DB Discrimination, abuse, harassment, and burnout in surgical residency training N Engl J Med 381 2019 1741 1752 31657887 65 Templeton K, Bernstein C., Sukhera J, et al. Gender-based differences in burnout: issues faced by women physicians. National Academy of Medicine. Available at: https://nam.edu/gender-based-differences-in-burnout-issues-faced-by-women-physicians/. Accessed XXXX. 66 Heath I. Women in medicine BMJ 329 7463 2004 412 413 15321881 67 Wang K Dovidio JF. Perceiving and confronting sexism: the causal role of gender identity salience Psychol Women Q 41 2017 65 76 29051685 68 Barnes KL McGuire L Dunivan G Gender bias experiences of female surgical trainees J Surg Educ 76 2019 e1 14 31601487 69 Dossa F, Simpson AN, Sutradhar R, et al. Sex-based disparities in the hourly earnings of surgeons in the fee-for-service system in Ontario, Canada. JAMA Surg 019;154:1134–42. 70 Bruce AN Battista A Plankey MW Perceptions of gender-based discrimination during surgical training and practice Med Educ Online 20 2015 25923 25652117 71 Ceppa DP Dolejs SC Boden N Gender bias and its negative impact on cardiothoracic surgery Ann Thorac Surg 109 2020 14 17 31445047 72 Civantos AM Bertelli A Gonçalves A Mental health among head and neck surgeons in Brazil during the COVID-19 pandemic: a national study Am J Otolaryngol 41 6 2020 102694 73 Kibbe MR. Consequences of the COVID-19 pandemic on manuscript submissions by women JAMA Surg 155 2020 803 804 32749449 74 Oleschuk M. Gender equity considerations for tenure and promotion during COVID-19 Can Rev Sociol 57 2020 502 515 32779307 75 Jones Y Durand V Morton K Collateral damage: how COVID-19 is adversely impacting women physicians J Hosp Med 15 2020 507 509 32804615 76 Lawrence JA Davis BA Corbette T Racial/ethnic differences in burnout: a systematic review J Racial Ethn Health Disparities 2021 Jan 11 10.1007/s40615-020-00950-0 [Epub ahead of print] 77 Pololi L Cooper LA Race Carr P. disadvantage and faculty experiences in academic medicine J Gen Intern Med 25 2010 1363 1369 20697960 78 Hassouneh D Lutz KF Beckett AK The experiences of underrepresented minority faculty in schools of medicine Med Educ Online 19 2014 24768 25472784 79 Zambrana RE Harvey Wingfield A Blatant, subtle, and insidious: URM faculty perceptions of discriminatory practices in predominantly White institutions Sociol Inquiry 87 2017 207 232 80 Campbell KM Rodríguez JE. Addressing the minority tax: perspectives from two diversity leaders on building minority faculty success in academic medicine Acad Med 94 2019 1854 1857 31192803 81 Dyrbye LN Power DV Massie FS Factors associated with resilience to and recovery from burnout: a prospective, multi-institutional study of US medical students Med Educ 44 2010 1016 1026 20880371 82 Johnson AP Wohlauer MV Mouawad NJ The impact of the COVID-19 pandemic on vascular surgery trainees in the United States Ann Vasc Surg 72 2021 182 190 33157252 83 Romanelli J Gee D Mellinger JD The COVID-19 reset: lessons from the pandemic on burnout and the practice of surgery Surg Endosc 34 2020 5201 5207 33051763 84 Beach SR. Is mindfulness a religion? Huffpost November 13, 2014 Available at: https://www.huffpost.com/entry/is-mindfulness-a-religion_b_6136612 PublishedUpdated December 6, 2017. Accessed April 24, 2021 85 Raski MP. Mindfulness: what it is and how it is impacting healthcare UBCMJ 7 1 2015 Available at https://ubcmj.med.ubc.ca/ubcmj-volume-7-issue-1/mindfulness-what-it-is-and-how-it-is-impacting-healthcare/mindfulne-ss-what-it-is-and-how-it-is-impacting-healthcare/ Accessed March 4, 2021 86 Goodman MJ Schorling JB. A mindfulness course decreases burnout and improves well-being among healthcare providers Int J Psychiatry Med 43 2012 119 128 22849035 87 Smith JM. Exercise to prevent surgeon burnout Surgeon Masters November 29, 2016 Available at https://surgeonmasters.com/blog/exercise-to-prevent-surgeon-burnout PublishedAccessed March 4, 2021 88 Berkowitz LR Dzara K Simpkin AL. Building your “educational Peloton: ” cycling together for success during uncertain times J Contin Educ Health Prof 41 2021 8 9 33433126 89 Shalhub S Mouawad NJ. COVID-19: socially distant yet never closer April 6, 2020 Vascular Specialist Available at: https://vascularspecialistonline.com/covid-19-socially-distant-yet-never-closer/ PublishedAccessed February 16, 2021 90 Society for Vascular Surgery. COVID-19. Available at: https://vascular.org/news-advocacy/covid-19-resources. Accessed March 4, 2021.	34144747	PMC9710729	NO-CC CODE	2022-12-02 23:21:30	no	Semin Vasc Surg. 2021 Jun 21; 34(2):43-50	utf-8	Semin Vasc Surg	2,021	10.1053/j.semvascsurg.2021.04.003	oa_other
==== Front Clin Nutr Clin Nutr Clinical Nutrition (Edinburgh, Scotland) 0261-5614 1532-1983 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. S0261-5614(21)00315-0 10.1016/j.clnu.2021.06.019 Covid-19 Anti-COVID-19 measures threaten our healthy body weight: Changes in sleep and external synchronizers of circadian clocks during confinement Baquerizo-Sedano Luis a∗∗ Chaquila José A. a Aguilar Luis b Ordovás José M. cd González-Muniesa Pedro efgh Garaulet Marta ijk∗ a Faculty of Health Sciences, San Ignacio de Loyola University, Av. La Molina 430, 15012, Lima, Peru b Institute of Food Sciences and Nutrition, San Ignacio de Loyola University, Av. La Molina 430, 15012, Lima, Peru c JM-USDA-HNRCA at Tufts University, 419 Boston Ave, Medford, MA 02155, USA d IMDEA Food, Crta. de Canto Blanco Institute, 8, E-28049 Madrid, Spain e University of Navarra; Department of Nutrition, Food Science and Physiology; School of Pharmacy and Nutrition. C/ Irunlarrea, 1, 31008 Pamplona, Spain f University of Navarra, Center for Nutrition Research, School of Pharmacy and Nutrition, Pamplona, C/ Irunlarrea, 1, 31008 Pamplona, Spain g IdISNA- Navarra Institute for Health Research, C/ Irunlarrea, 3, 31008 Pamplona, Spain h CIBER Fisiopatología de la Obesidad y Nutrición (CIBERobn), Instituto de Salud Carlos III (ISCIII), Monforte de Lemos, 5. Pabellón 12. 28029. Madrid, Spain i Department of Physiology, University of Murcia, Campus de Espinardo, s/n. 30100, Murcia, Spain j Medical Chronobiology Program, Division of Sleep and Circadian Disorders, Departments of Medicine and Neurology, Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115, USA k Research Biomedical Institute of Murcia (IMIB-Arrixaca) 30120 El Palmar, Murcia, Spain ∗ Corresponding author. Department of Physiology, University of Murcia, Campus de Espinardo, s/n. 30100, Murcia, Spain. Fax: +34 868 88 39 63. ∗∗ Corresponding author. 25 6 2021 12 2022 25 6 2021 41 12 29882995 26 2 2021 18 6 2021 © 2021 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. 2021 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism 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 & aims Emergency measures in the face of the recent COVID-19 pandemic have been different among countries, although most have opted for confinement and restrictions on social contact. These measures have generated lifestyle changes with potential effects on individuals' health. The disturbances in daily routines due to confinement and remote work have impacted circadian rhythms and energy balance; however, the consequences of these disruptions have not been studied in depth. The objective was to evaluate the impact of 12-week confinement on body weight, considering changes in several external synchronizers of the biological clock. Methods The participants, 521 university students (16–35 years), responded to 52 questions oriented to determine light exposure, sleep patterns, sedentary lifestyle, and eating times. Results We found a reduction in sunlight exposure and sleep duration, an increment in sedentarism and screen exposure, and a delay in the timing of the main meals and sleep in the whole cohort. These behavioral changes were associated with a twofold increase in obesity. Subjects who increased their sedentary hours and shortened their sleep to a higher degree were those who gained more bodyweight. The most influential factors in body weight variation during confinement were sleep duration, physical activity (sedentarism), and light (timing of screen exposure). The mediation model explained 6% of the total body weight variation. Conclusions Results support a significant impact of confinement on several external synchronizers of the biological clock and on body weight. Health-related recommendations during the pandemic must include behavioral recommendations to mitigate the adverse effects on the biological clock. Keywords Covid-19 Confinement Biological clocks External synchronizer Obesity ==== Body pmc1 Introduction June 2021, eigthteen months after the first reported case of severe acute respiratory syndrome coronavirus 2 (SARS Cov-2) infection, the COVID-19 pandemic has affected more than 171 million people, with over 3,5 million deaths globally [1]. The measures taken by governments have dramatically changed our lifestyle, which is now characterized by confinement and social isolation. While needed to prevent contagion, these measures appear to have also negative health consequences [[2], [3], [4]]. Thus, the decades-long obesity pandemic may have worsened due to the measures imposed by the COVID-19 pandemic. Although the etiology of obesity identifies excessive energy intake as the main factor (positive energy balance), there are other critical factors involved, such as sedentarism, genetics, altered microbiota and sleep deprivation [[5], [6], [7], [8]]. Furthermore, obesity has been shown as the main modifiable risk factor for severity and mortality in COVID-19 [9], which is of particular concern at the public health level [10]. The internal clock of mammals has rhythmicity of approximately 24 h. It controls physical activity patterns, endocrine functions and multiple physiological processes, and it is synchronized by internal and external regulators [11]. Examples of external synchronizers are the changes in sleep/wake cycles, eating/fasting, or light/darkness. Alterations in these external synchronizers, as happens during shift work, social jet lag, or light pollution, can desynchronize our biological clock [12]. In connection with this, mandatory confinement has been shown to alter the sleep/wake patterns [13] and to decrease the light time exposure [14], desynchronizing the internal clocks, mainly due to sleep deprivation and increased sedentary lifestyle, which has been shown to increase the risk of metabolic diseases, such as obesity [15], triggering a vicious cycle. The current study aimed to evaluate the impact of 12-week confinement on body weight, considering changes in several external synchronizers of the biological clock in university students. 2 Methods details 2.1 Population and study design A descriptive exploratory study was carried out in 2020, between June 13 and 20, in the province of Lima, Perú, after 12-weeks of confinement. During the confinement period, the right to free transit was restricted. Only those who worked in essential services such as electricity, water, food, and health services could go out from their houses. Furthermore, only one person per family was allowed to go out to buy necessities like food or medicine. These restrictions were continuously prolonged and slightly modified by the emergency decree of the Peruvian state [16]. The information was collected using a questionnaire developed in Google Forms that was sent to the students' institutional mail of San Ignacio de Loyola University. Initially, 820 subjects responded to the questionnaire; the sample included adolescents and young adults (up to 35 years old). Postgraduate students were excluded from the survey because they were not full-time students and their schedules were significantly different from undergraduates'. Indeed, they usually had an external job and this would have differentially influenced the time of exposure to screens and sleep habits. In total, those students who were older than 35 years, postgraduate and those who had eating disorders or morbid obesity were also excluded (n = 40). After reviewing for quality control, 259 were excluded due to wrong or incomplete answers. The final sample of the study was of 521 subjects (Fig. 1 ) (female = 63.72%, male = 36.28%, age = 20.93 ± 3.20 years, range = 16–35 years). Participants declared their willingness to participate in the study through informed consent. The study was conducted following the standards of the Declaration of Helsinki and was approved by the institutional review board of San Ignacio de Loyola University.Fig. 1 Sample collection chart. Fig. 1 2.2 Questionnaire The questionnaire (supplemental material 1), which the research group developed, reflected the status of the participants before and after the 12-weeks of confinement. It consisted of 52 questions (4 questions about light exposure time, 6 questions about sleep patterns, 2 questions about sedentary lifestyle, and 6 about food timing, the remaining questions were not taken into account for this manuscript). Two investigators reviewed the initial data to identify illogical values in the responses. For example: a. durations longer than 24 h per day (such as a sleep duration of 12 h and a screen exposure time of 16 h), b. incongruences in the timing of activities (such as reporting falling asleep at 21 h and having dinner at 23 h, or waking up at 10 h and having breakfast at 8 h). Illogical or incomplete responses on any of the variables ruled out participation. 2.2.1 Body weight variables To estimate the body weight change during confinement, 3 questions were asked: The first question was about the self-reported body weight; the second was related to whether they had changed their body weight during confinement (with two options, yes or not). If the answer was yes, the last question was to what extent they changed their body weight (with the option of positive and negative values). Finally, only completed answers were considered. If the person was not sure about the variation of its weight and left that answer blank, the participation was suspended. The body mass index was calculated by dividing body weight by squared height and classified according to the World Health Organization (WHO) [17]. 2.2.2 Chronotype and sleep habits Individuals' chronotypes were assessed by the Munich Chronotype Questionnaire [18]; the midpoint of sleep was calculated as the midpoint between time falling asleep and rise time. Chronotypes were classified into morning, intermediate and evening types by dividing the sample of 521 students according to their midpoint of sleep, into tertiles, with the first tertile defining the morning type (midpoint of sleep < 3:45 h), the second tertile as intermediate (3:45 h - 4:52 h) and the third tertile as evening type (>4:52 h). 2.2.3 Screen exposure time Screen exposure time (duration) was obtained through open questions such as: “During the first week of confinement, on average, how many hours per day did you spend in front of a screen?” 2.2.4 Sedentarism To determine the sedentary time, the following question was formulated “During the first week of confinement, on average, how many hours did you sit per day?” 2.2.5 Food timing and midpoint of intake The timing of the three main meals of the day, such as breakfast, lunch and dinner, were obtained through questions such as “Currently, on average, at what time do you have lunch?” or “During the first week of confinement, on average, at what time did you eat dinner?” The midpoint of intake was estimated by subtracting the last intake and the first intake, the result was divided by 2, and the first intake was added to that result, as previously reported by Lopez-Miguez et al. [19]. 2.2.6 Fasting at night determination Night fasting duration was calculated by subtracting to the timing of breakfast, the timing of the day before dinner. 2.3 Statistical analysis The normality of the distribution of the variables was assessed using the Shapiro–Wilk Test. For data without a normal distribution, the median was compared using the Mann–Whitney U test. Median comparisons were used to determine the change in body mass index, sedentarism, sleep duration, external synchronizers and meal timing between the thirteenth week and the first week of confinement. In contrast, a comparison of proportions was used to evaluate the change in the body mass index classification and the chronotype classification. The One-Way ANOVA Test was used to evaluate differences between groups, and Tukey's Post Hoc Test to compare groups. The statistical package used was STATA v16. The mediation analysis was performed with the IBM SPSS Amos program. A significance of 95% was used in all statistical analyses. 3 Results We have evaluated the impact of 12-week confinement on body mass index (BMI) and several circadian synchronizers. Daily habits, sleep habits and feeding times in college students are presented in Table 1 .Table 1 General characteristics of college student at first and thirteenth week of confinement. Table 1N First week of confinement Thirteenth week of confinement p value 521 521 Age, years 21 (19–23) Women, % (n) 63.72 (332) BMI, kg/m2 23.61 (23.33–23.90) 24.17 (23.85–24.49) <0.001 Daily habits Sunlight exposure duration, h 2:06 (1:55–2:17) 1:27 (1:13–1:36) <0.001 Sedentarism duration, h 7:06 (6:49–7:24) 9:34 (9:16–9:52) <0.001 Screen exposure time, h 7:15 (6:59–7:32) 9:49 (9:31–10:06) <0.001 Sleep habits Ready for sleeping, h 23:42 (23:34–23:50) 00:34 (0:25–0:42) <0.001 Duration of sleep latency, h 0:44 (0:40–0:48) 0:43 (0:40–0:47) 0.387 Sleep time, h 0:25 (0:16–0:35) 1:17 (1:08–1:27) <0.001 Time to wake up, h 8:34 (8:25–8:42) 8:19 (8:11–8:28) 0.029 Sleep duration, h 8:06 (7:58–8:14) 7:01 (6:53–7:09) <0.001 MPS, h 4:30 (4:22–4:38) 4:49 (4:42–4:57) <0.001 Feeding time Breakfast, h 9:16 (9:09–9:24) 9:29 (9:21–9:36) 0.006 Lunch, h 13:55 (13:49–14:01) 14:13 (14:07–14:19) <0.001 Dinner, h 20:07 (20:00–20:13) 20:28 (20:21–20:34) <0.001 MPI, h 14:41 (14:35–14:46) 14:58 (14:52–15:04) <0.001 Night fasting duration, h 13:10 (13:03–13:18) 13:00 (12:52–13:08) 0.073 BMI: Body mass index; MPS: Midpoint of sleep; MPI: Midpoint of intake. Values were mean (CI 95%), p value represent the delta of both periods derived from U-The Man Whitney test. Overall, we observed statistically significant increases in BMI and a shift towards an evening chronotype (Fig. 2 ). The percentage of subjects below 18.5 (kg/m2) and between 25 and 29.9 (kg/m2) increased significantly (p < 0.001) (Fig. 2B) while the proportion of morning chronotypes decreased (p = 0.02), resulting in a significantly increased number of subjects in the evening chronotype (p < 0.001) (Fig. 2C).Fig. 2 Variation of main outcomes. A) Variation of BMI, B) Variation of BMI classification and C) Variation of chronotype classification. A shows the mean, and the dispersion is expressed as the Standard Error of the Mean and the U-The Man Whitney test was used. For B and C comparison of proportions was used. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05. BMI: Body mass index. Fig. 2 These changes in BMI and chronotype paralleled changes in factors known to synchronize the biological clock (i.e., external synchronizers) (Fig. 3 ). We found significant decreases in the duration of sunlight exposure (Δ −0:39h [−0:53h, −0:24h]) and of sleep (Δ −1:04h [−1:16h, −0:53h]) and significant increases in the duration of screen exposure (Δ 2:33h [2:09h, 2:57h]), and sedentarism (Δ 2:27h [IC 2:03h, 2:52h]) and a trend was found for decreases in night fasting duration (Δ −0:11h [−0:27h, −0:03h]) (p = 0.07) (Fig. 3A). Likewise, we found significant delays in food timing (a synchronizer of peripheral clocks) such as breakfast timing (Δ 0:12h [0:01h, 0:22h]), lunch timing (Δ 0:19h [0:11h, 0:27h]) and dinner timing (Δ 0:22h [0:12h, 0:31h]), and consequently in the midpoint of intake (Δ 0:16h [0:09h, 0:25h]). We also found a delay in the midpoint of sleep (Δ 0:18h [0:07h, 0:29h]) (p < 0.05) (Fig. 3B). All these external synchronizers changed significantly with confinement (p < 0.001), except for night fasting duration that only showed a trend (p = 0.07).Fig. 3 Changes in external synchronizers with confinement. Y axis represents duration in Figure A and timing in Figure B. Mean (SEM). The U-The Man Whitney test was used. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, ns: not significant. Fig. 3 Changes in body weight were classified into five categories (Fig. 4 ), a) Body weight gain: greater than 10% (BWG >10%), between 10 and 5% (BWG 10-5%), less than 5% (BWG <5%); b) no variation (NV) and c) body weight loss (BWL). The percentage of people who gained body weight was large (64% of the total population, n = 332 from the total n = 521). From this population, a 3.3% gained more than the 10% of their initial body weight and a 26%, gained more than the 5% of their initial body weight.Fig. 4 External synchronizers and bodyweight variation during confinement. A) Sleep duration, B) Screen exposure time and C) Sedentarism. The capital letter represents the difference in the groups. Mean (SEM). The U-The Man Whitney test was used. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05. Fig. 4 Those subjects who gained more body weight (BWG >10% group) decreased their sleep duration to a greater extent (Δ −2:19h [−3:32h, −1:07h]) than those who gained less body weight or who even lost body weight during confinement (Fig. 4A). Furthermore, BWG 10-5% group showed a greater increase in screen exposure time (Δ 3:10h [2:21h, 4:00h]) compared to BWL (Δ 1:34h [0:43h, 2:25h]) (Fig. 4B). Also, the BWL group had a lower increase in sedentary hours (Δ 0:46h [−0:04h, 1:39h]) compared to the other groups (Fig. 4C). Mediation models for the external synchronizers' changes with confinement (Fig. 5 ) showed a direct effect of sedentarism, sleep duration, and screen exposure duration on body weight variation (BWV). The model was able to predict the variation in body weight in 6% (R2 = 0.064), showing a positive correlation between sedentarism and BWV (β = 0.302) and a negative correlation between sleep duration and BWV (β = −0.211). For every 1 h of increase in sedentarism, there was a 0.3% increase in body weight, and for every 1 h of increase in sleep duration, there was a 0.2% decrease in body weight (Fig. 5).Fig. 5 Mediation models of external synchronizerson body weight variation. Body weight variation (BWV) as the dependent variable, Δ sleep duration, Δ sedentarism, Δ screen exposure time as independent variables. Δ was calculated by subtracting the variables from the thirteenth week to the first week of confinement. The straight arrows represent the correlation between variables of interest; the curved arrows denote the covariance among the three represented external synchronizers. The beta coefficients are shown in italic letter while beta coefficients of the adjusted model are shown in bold letter. Cov. = Covariance. BMI = Body Mass Index. Δ = Delta. ∗∗∗p < 0.001, ∗p < 0.05. Fig. 5 Regarding meal timing, the model explained 0.6% (R2 = 0.006) of the variation in body weight, which suggests that the influence of meal timing on the BWV was low (data not shown). The coefficients of determination showed that sleep duration was the most influential variable on BWV when the model was not adjusted by the other external synchronizers (R2 = 0.022, p = 0.001); In contrast, after adjusting for external synchronizers, the most influential variable on BWV was sedentarism (R2 = 0.058, p < 0.001), followed by screen exposure time (R2 = −0.031, p < 0.001), and finally by sleep duration (R2 = 0.022, p < 0.001). Overall, the adjusted model predicted a 6% of the BWV (R2 = 0.064, p < 0.001). Figure 6 represents a summary of the impact of 12-week confinement on body weight, considering changes in several external synchronizers of the biological clock in these university students.Fig. 6 Changes after twelve weeks of confinement on biological clock synchronizers. Sunlight exposure time and sleep duration decreased. Sedentarism, screen exposure time and body weight increased while meals timing and midpoint of sleep were delayed. All changes were statistically significant p < 0.05. Fig. 6 4 Discussion In this study, including 521 young university students, we found a significant increase in BMI and body weight together with variations in the duration and timing of various external synchronizers of the biological clock after 12-weeks of confinement. This confinement was part of the emergency measures implemented by the Peruvian Government against COVID-19 [16]. These changes included a twofold increase in the prevalence of obesity together with decreases in sunlight exposure and sleep duration. Moreover, confinement was associated with an increase in the screen exposure time and in sedentarism and a delay in the timing of the main meals and in the midpoint of sleep. Mediation analyses suggested that from these various external synchronizers, the most influential factors in the body weight variation with confinement were sleep (duration), physical activity (sedentarism), and light (timing of screen exposure). Our data showed an overall increase in BMI both in men and women, with no significant sex interaction. In addition, according to BMI classification, the percentage of obesity doubled and the percentage of normal BMI decreased (~10%). When the population was classified according to the variation in body weight with confinement, our data showed that changes in body weight were mediated by three factors: sleep duration (R2 = 0.02), sedentarism (R2 = 0.06), and screen exposure (R2 = 0.03). The regression model explained a 6% of the total body weight variation; while this number does not seem to be relevant, it is higher than other classical known factors influencing obesity such as genetics, which has been shown to account for a ~2.7% of the variation in BMI [20]. Our results are in agreement with the body weight gain shown in other studies examining shorter lockdowns [21]. Furthermore, in the current study, subjects who gained more body weight (>10%) had a more significant increase in sedentarism (Δ 3:28h) vs. those who lost body weight (Δ 0:46h) during the 12 weeks of confinement. This increase would be related to a more positive energy balance due to a more sedentary lifestyle. Nevertheless, other circadian system alterations could also be involved, such as the more prolonged exposure to screens, the shorter exposure to sunlight, and the delay in the timing of sleep and food intake, which may be influencing the changes in chronotype towards a more evening-type. Unlike genetics, these factors are modifiable and should be considered a fundamental component in the recommendations to prevent the adverse effects associated with obesity. The recent confinement has affected sunlight exposure, known as the strongest synchronizer of the circadian system [22]. In our study, the time of exposure to sunlight decreased by more than half an hour with confinement (Δ −0:39h). Conversely, we observed an increase of more than two hours in screen exposure (Δ 2:33h) in agreement with other studies examining the effects of confinement [23]. These changes may be related to variations in the study and work habits due to confinement, which demands greater use of electronics and increased exposure to digital media [24]. Decreased sunlight, together with continuous exposure to screens and delayed meals and sleep timings, affects the circadian system and may alter the synchronization between the central and peripheral clocks and between the biological clocks and the environment, causing chronodisruption and subsequently obesity [25]. It has been described that, in addition to light, which synchronizes our central clock, there are other external synchronizers directed towards specific peripheral clocks, such as those in adipose tissue [26]. Among these peripheral synchronizers, examples are related to changes in activity and resting and in eating and fasting. In addition, sleep can be understood as a result of- or a synchronizer of-the circadian system [27]. Circadian system alterations have been related to changes in adipose tissue metabolism [28], obesity [7,8] and cardiovascular risk [[29], [30], [31]]. Therefore, and considering that confinement has affected several known external synchronizers of the biological clock, we would expect a significant change in body weight as the one observed in our study. There is concern about sleep habit disturbances that may have arisen during the pandemic [[32], [33], [34]]. We found an overall reduction in sleep duration of approximately one hour (Δ −1:04) in both genders, in agreement with other studies [13]. Our results show that those who shortened their sleep with confinement to a higher degree, more than two hours, (Δ −2:19h) were those who gained more body weight (11.9%). Furthermore, the significant delays observed for sleep timing in combination with the delay in the timing of the main meals: breakfast (Δ 0:12h), lunch (Δ 0:19h), and dinner (Δ 0:22h), explained the increase in the proportion of evening-types: 11% of the students became evening-types during the 12 weeks of confinement. Our results are consistent with other studies that show that confinement represents a parenthesis in sleep habits according to the natural changes in society [35]. In this context, it is now a challenge to maintain adequate metabolic health [36] with current and upcoming measures regarding confinement and social isolation. Furthermore, confinement within homes, mobility restrictions, and social isolation rules have been traduced to less accessibility to fitness centers and public spaces for physical activity. In accordance, our results show an increase in sedentarism, similar to previous reports [21,23,37], although one study showed a slight increase in physical activity, especially for body weight training [38]. These results show a lower energy expenditure due to physical activity, one of the factors of body weight gain, with a probable loss of muscle mass and increased fat accumulation. Furthermore, sedentarism may lead to a flattening of the daily activity rhythms, resulting in a weakening of the signals to the internal clock, and therefore, chronodisruption. Although the etiology of obesity recognizes dietary habits as the main factor with a positive energy balance, there are other intervening factors such as physical activity, genetics, microbiota, and sleep habits [[5], [6], [7], [8]]. Therefore, the obesity pandemic that has been going on for decades has found itself in the worst scenario with this new pandemic. Considering that, in the short term, there is still uncertainty about the duration of the state of emergency, we contemplate that it is important to monitor changes in behavior and a good alternative is the use of smartphones and other wearable devices [39]. The current study has strengths and limitations. This is the first study that considers global changes in several external synchronizers of the circadian clock during confinement by using a comprehensive questionnaire consisted of 52 questions that addressed several aspects related to the circadian system such as exposure to sunlight, exposure to screens, sleeping habits, sedentary lifestyle, and timing of food intake, together with body weight changes. A relevant aspect is that due to the low variation of daylight length in Lima, sun hours were practically the same at the beginning and at the end of confinement during the 12 weeks of the study. Therefore, changes in sun exposure during the study were not affected by the shortening of days [40]. There are limitations associated with our study; first, the data are based on self-completed questionnaires. Moreover, we have no information about the students' dietary habits, apart from food intake timing. The results and conclusions apply to the specific group under investigation (young students). Nevertheless, data support a significant impact of confinement on several external synchronizers and body weight change in this population. 5 Conclusion This study analyses how acute changes in our biological clock's external synchronizers may affect body weight in an unusual real-life situation. During confinement, there was an increase in BMI and in the proportion of obese students. Sleep patterns and daily habits were altered, finding a decrease in sleep duration and sunlight exposure time and an increase in the duration of exposure to screens. In addition, there was a delay in feeding schedules and in the midpoint of sleep. Finally, sedentarism seemed to be more relevant in body weight variation than sleep duration and screen exposure time. As long as the vaccine for SARS-CoV-2 is not widely available, there is great concern about the possible impact of confinements. Healthier lifestyle changes appear to be more urgent now than ever. This study has shown the association between changes in the external circadian rhythm synchronizers and body weight gain. Funding statement This work has been supported in part by The Spanish Government of Investigation, Development and Innovation (SAF2017-84135-R) including 10.13039/501100002924 FEDER co-funding; The Autonomous Community of the Region of Murcia through the Seneca Foundation (20795/PI/18) and 10.13039/100000062 NIDDK R01DK105072 granted to M. Garaulet. In addition, P. González-Muniesa has received support from the CIBER Physiopathology of Obesity and Nutrition (10.13039/501100012514 CIBERobn ), 10.13039/501100004587 Carlos III Health Research Institute (CB12/03/30002). Author contributions Baquerizo-Sedano L, designed the study, analyzed and interpreted the data and wrote the first draft of the manuscript. Chaquila JA, analyzed and interpreted the data and wrote the first draft of the manuscript. Aguilar L, designed the study and contributed to data collection. Ordovás JM, contributed to data interpretation and review of the manuscript. González-Muniesa P, analyzed and interpreted the data and reviewed the manuscript. Garaulet M, analyzed and interpreted the data and reviewed the manuscript. All authors edited the manuscript. Conflict of interest The authors declare no conflict of interest. Appendix A Supplementary data The following is the Supplementary data to this article:Multimedia component 1 Multimedia component 1 Acknowledgements We are grateful to all the students of the San Ignacio de Loyola University who responded to the survey. We acknowledge the contribution of our colleagues from the San Ignacio de Loyola University: Hans Donayre, Melanie Agustin and Ariana Espino (Faculty of Health Sciences) in carrying out this study. And a special thanks to the director of the Nutrition and Dietetics career, Dayana Barriga, for her unconditional support. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.clnu.2021.06.019. ==== Refs References 1 COVID-19 Map - Johns Hopkins Coronavirus Resource Center n.d https://coronavirus.jhu.edu/map.html 2 Jia P. Liu L. Xie X. Yuan C. Chen H. Guo B. Impacts of COVID-19 lockdown on diet patterns among youths in China: the COVID-19 Impact on Lifestyle Change Survey (COINLICS) Appetite 158 2020 105015 10.1038/s41366-020-00710-4 33121998 3 Galandra C. Cerami C. Santi G.C. Dodich A. Cappa S.F. Vecchi T. Job loss and health threatening events modulate risk - taking behaviours in the Covid - 19 emergency Sci Rep 2020 1 10 10.1038/s41598-020-78992-x 31913322 4 Violant-Holz V. Gallego-Jiménez M.G. González-González C.S. Muñoz-Violant S. Rodríguez M.J. Sansano-Nadal O. Psychological health and physical activity levels during the covid-19 pandemic: a systematic review Int J Environ Res Publ Health 17 2020 1 19 10.3390/ijerph17249419 5 González-Muniesa P. Mártinez-González M.-A. Hu F.B. Després J.-P. Matsuzawa Y. Loos R.J.F. Obesity Nat Rev Dis Prim 3 2017 17034 10.1038/nrdp.2017.34 28617414 6 San-Cristobal R. Navas-Carretero S. Martínez-González M.Á. Ordovas J.M. Martínez J.A. Contribution of macronutrients to obesity: implications for precision nutrition Nat Rev Endocrinol 16 2020 305 320 10.1038/s41574-020-0346-8 32235875 7 Garaulet M. Ordovás J.M. Madrid J.A. The chronobiology, etiology and pathophysiology of obesity Int J Obes (Lond) 34 2010 1667 1683 10.1038/ijo.2010.118 20567242 8 Garaulet M. Gómez-Abellán P. Timing of food intake and obesity: a novel association Physiol Behav 134 2014 44 50 10.1016/j.physbeh.2014.01.001 24467926 9 Huang Y. Lu Y. Huang Y.M. Wang M. Ling W. Sui Y. Obesity in patients with COVID-19: a systematic review and meta-analysis Metabolism 113 2020 154378 10.1016/j.metabol.2020.154378 33002478 10 Cuschieri S. Grech S. Obesity population at risk of COVID-19 complications Glob Heal Epidemiol Genom 2020 4 9 10.1017/gheg.2020.6 11 Morris C.J. Aeschbach D. Scheer F.A.J.L. Circadian system, sleep and endocrinology Mol Cell Endocrinol 349 2012 91 104 10.1016/j.mce.2011.09.003 21939733 12 Lopez-Minguez J. Gómez-Abellán P. Garaulet M. Circadian rhythms, food timing and obesity Proc Nutr Soc 75 2016 501 511 10.1017/S0029665116000628 27339810 13 Pinto J. van Zeller M. Amorim P. Pimentel A. Dantas P. Eusébio E. Sleep quality in times of Covid-19 pandemic Sleep Med 74 2020 81 85 10.1016/j.sleep.2020.07.012 32841849 14 Górnicka M. Drywień M.E. Zielinska M.A. Hamułka J. Dietary and lifestyle changes during COVID-19 and the subsequent lockdowns among polish adults: a cross-sectional online survey PLifeCOVID-19 study Nutrients 12 2020 10.3390/nu12082324 15 Garaulet M. Gómez-Abellán P. Chronobiology and obesity Nutr Hosp 28 Suppl 5 2013 114 120 10.3305/nh.2013.28.sup5.6926 24010751 16 Gobierno del Perú Normativa sobre estado de emergencia por coronavirus Compendio 2021 https://www.gob.pe/institucion/pcm/colecciones/787-normativa-sobre-estado-de-emergencia-por-coronavirus 17 Weir C.B. Jan A. BMI classification percentile and cut off points 2020 Treasure Island (FL) 18 Roenneberg T. Wirz-Justice A. Merrow M. Life between clocks: daily temporal patterns of human chronotypes J Biol Rhythm 18 2003 80 90 10.1177/0748730402239679 19 Lopez-Minguez J. Dashti H.S. Madrid-Valero J.J. Madrid J.A. Saxena R. Scheer F.A.J.L. Heritability of the timing of food intake Clin Nutr 38 2019 767 773 10.1016/j.clnu.2018.03.002 29571565 20 Locke A.E. Kahali B. Berndt S.I. Justice A.E. Pers T.H. Day F.R. Genetic studies of body mass index yield new insights for obesity biology Nature 518 2015 197 206 10.1038/nature14177 25673413 21 Barrea L. Pugliese G. Framondi L. Di Matteo R. Laudisio D. Savastano S. Does Sars-Cov-2 threaten our dreams? Effect of quarantine on sleep quality and body mass index J Transl Med 18 2020 1 11 10.1186/s12967-020-02465-y 31900168 22 Legates T.A. Fernandez D.C. Hattar S. Light as a central modulator of circadian rhythms, sleep and affect Nat Rev Neurosci 15 2014 443 454 10.1038/nrn3743 24917305 23 Górnicka M. Drywień M.E. Zielinska M.A. Hamułka J. Dietary and lifestyle changes during COVID-19 and the subsequent lockdowns among polish Adults : PLifeCOVID-19 study Nutrients 12 2020 2324 32756458 24 Cellini N. Canale N. Mioni G. Costa S. Changes in sleep pattern, sense of time, and digital media use during COVID-19 lockdown in Italy 2020 10.31234/osf.io/284mr 25 Gómez-Abellán P. Madrid J.A. Ordovás J.M. Garaulet M. [Chronobiological aspects of obesity and metabolic syndrome] Endocrinol y Nutr organo la Soc Esp Endocrinol y Nutr 59 2012 50 61 10.1016/j.endonu.2011.08.002 26 Johnston J.D. Ordovás J.M. Scheer F.A. Turek F.W. Circadian rhythms, metabolism, and chrononutrition in rodents and humans Adv Nutr 7 2016 399 406 10.3945/an.115.010777 26980824 27 Lewis P. Oster H. Korf H.W. Foster R.G. Erren T.C. Food as a circadian time cue — evidence from human studies Nat Rev Endocrinol 16 2020 213 223 10.1038/s41574-020-0318-z 32055029 28 Froy O. Garaulet M. The circadian clock in white and Brown adipose tissue: mechanistic, endocrine, and clinical aspects Endocr Rev 39 2018 261 273 10.1210/er.2017-00193 29490014 29 Lopez-Minguez J. Gómez-Abellán P. Garaulet M. Timing of breakfast, lunch, and dinner. Effects on obesity and metabolic risk Nutrients 11 2019 2624 10.3390/nu11112624 31684003 30 Scheer F.A.J.L. Hilton M.F. Mantzoros C.S. Shea S.A. Adverse metabolic and cardiovascular consequences of circadian misalignment Proc Natl Acad Sci U S A 106 2009 4453 4458 10.1073/pnas.0808180106 19255424 31 Potter G.D.M. Skene D.J. Arendt J. Cade J.E. Grant P.J. Hardie L.J. Circadian rhythm and sleep disruption: causes, metabolic consequences, and countermeasures Endocr Rev 37 2016 584 608 10.1210/er.2016-1083 27763782 32 Casagrande M. Favieri F. Tambelli R. Forte G. The enemy who sealed the world: effects quarantine due to the COVID-19 on sleep quality, anxiety, and psychological distress in the Italian population Sleep Med 75 2020 12 20 10.1016/j.sleep.2020.05.011 32853913 33 Ramar K. The COVID-19 pandemic: reflections for the fi eld of sleep medicine 2020 34 Zhou S.-J. Wang L.-L. Yang R. Yang X.-J. Zhang L.-G. Guo Z.-C. Sleep problems among Chinese adolescents and young adults during the coronavirus-2019 pandemic Sleep Med 74 2020 39 47 10.1016/j.sleep.2020.06.001 32836185 35 Wright K.P. Linton S.K. Withrow D. Casiraghi L. Lanza S.M. de la Iglesia H. Sleep in university students prior to and during COVID-19 Stay-at-Home orders Curr Biol 30 2020 R797 R798 10.1016/j.cub.2020.06.022 32693068 36 King A.J. Burke L.M. Halson S.L. Hawley J.A. The challenge of maintaining metabolic health during a global pandemic Sports Med 50 2020 1233 1241 10.1007/s40279-020-01295-8 32449141 37 Zhang C. Yang L. Liu S. Ma S. Wang Y. Cai Z. Survey of insomnia and related social psychological factors among medical staff involved in the 2019 novel coronavirus disease outbreak Front Psychiatr 11 2020 10.3389/fpsyt.2020.00306 38 Di Renzo L. Gualtieri P. Pivari F. Soldati L. Attinà A. Cinelli G. Eating habits and lifestyle changes during COVID-19 lockdown: an Italian survey J Transl Med 18 2020 229 10.1186/s12967-020-02399-5 32513197 39 Sun S. Folarin A.A. Ranjan Y. Rashid Z. Conde P. Stewart C. Using smartphones and wearable devices to monitor behavioral changes during COVID-19 J Med Internet Res 22 2020 e19992 10.2196/19992 40 Calendario Solar Año 2020 Perú 2021 https://www.vercalendario.info/es/sol/peru-ano-calendario-2020.html%3E	34246488	PMC9711511	NO-CC CODE	2022-12-02 23:21:30	no	Clin Nutr. 2022 Dec 25; 41(12):2988-2995	utf-8	Clin Nutr	2,021	10.1016/j.clnu.2021.06.019	oa_other
==== Front J Emerg Med J Emerg Med The Journal of Emergency Medicine 0736-4679 0736-4679 Published by Elsevier Inc. S0736-4679(22)00652-7 10.1016/j.jemermed.2022.10.026 Article Therapeutic versus prophylactic anticoagulation for patients admitted to the hospital with COVID-19 and elevated d-dimer concentration (ACTION): an open-label, multicenter, randomized, controlled trial Lopes RD, Melo de Barros e Silva PG, Furtado RM, et. al Lancet. 2021; 297: 2253-2263 30 11 2022 10 2022 30 11 2022 63 4 615616 Copyright © 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. ==== Body pmcIncreased arterial and venous thrombotic events have been reported in patients with COVID-19compared to those with other respiratory viruses. These thrombotic events are thought to be related to a thromboinflammatory state brought about by the virus. Observational data has suggested that starting patients on therapeutic or prophylactic anticoagulation on admission to the hospital may lower in-hospital mortality for patients with COVID-19. However, there is currently not enough data to know the optimal strategy in terms of type, dose, and duration of anticoagulation treatment. This study aimed to determine whether therapeutic anticoagulation is effective in preventing complications in patients hospitalized with COVID-19 and elevated d-dimer concentrations. It was an open-label multicenter, randomized controlled trial in patients hospitalized in Brazil with COVID-19 diagnosis, symptoms for up to 14 days prior to randomization and elevated d-dimer concentrations. Patients were randomized via a 1:1 ratio in permuted blocks of variable size, stratified based on clinical condition (stable vs unstable) to either receive therapeutic anticoagulation for 30 days (with rivaroxaban if clinically stable, or with enoxaparin if clinically unstable) or prophylactic anticoagulation (with either enoxaparin or unfractionated heparin). Neither patients nor investigators were blinded to group allocation. The clinically unstable group received subcutaneous enoxaparin 1mg/kg twice per day or unfractionated heparin dosed to achieve a target anti-Xa concentration (0.3-0.7IU/mL). Once stabilized, these patients were transitioned to oral rivaroxaban dosed at 20mg daily. All patients in the therapeutic group continued treatment with rivaroxaban to day 30. The prophylactic group received standard venous thromboembolism prophylaxis dosing. If patients developed an indication for therapeutic anticoagulation, they were allowed to receive it. The prophylactic group of patients was only kept on anticoagulation while inpatient. Follow up was performed at 30 and 60 days. The primary outcome studied was a hierarchical composite of time to death, duration of hospitalization or duration of supplemental oxygen use through 30 days. The primary safety outcome was major or clinically relevant non-major bleeding. Intention-to-treat analysis was used, and the primary outcome was reported using a win ratio method. Each patient in the treatment group was incrementally compared to each control patient to determine the “winners” or “ties” in relation to time to death, length of stay, and days of oxygen-free support in escalating challenges if “ties” occurred. The win ratio was then reported. There were 310 patients in the therapeutic group's and 304 patients in the prophylactic group's primary analysis. Baseline characteristics were similar between the therapeutic and prophylactic groups. The total number of wins between the groups was not statistically different (win ratio 0.86, p=0.40, 95% CI 0.59-1.22) with 28,899 (34.8%) wins in the therapeutic group versus 34,288 (41.3%) wins in the prophylactic group. There were 19,837 (23.9%) ties. There was no statistical difference between the 8-point ordinal scale at day 30, disease progression measured on day 7, 15, and 30, or duration of invasive mechanical ventilation at the end of 30 days. There was also no statistical difference between individual thrombotic events or composite VTE, myocardial infarction, stroke, systemic embolism, or major adverse limb events. Differences remained insignificant when comparing clinically stable and unstable patients in each group. As to safety outcomes, there were 26 (8%) major or clinically relevant bleeding events in the therapeutic group compared to 7 (2%) in the prophylactic group (RR 3.64, 95% CI 1.61-8.27, p=0.001). The authors concluded that therapeutic anticoagulation did not result in clinically better outcomes than prophylactic anticoagulation in the specific population of patients studied, and that therapeutic was in fact associated with a higher risk of major or clinically relevant bleeding than was prophylactic anticoagulation. They therefore recommended against using therapeutic anticoagulation in this population unless specifically indicated for other clinical reasons. The authors do note limitations such as performing follow-up interviews and pill counts over a phone call may have influenced their results or that the open label type study may have introduced bias although possibly mitigated by blind adjudication process. Elizabeth M. Hanson, MD Zachary B. Lewis, MD University of Arkansas for Medical Sciences Little Rock, AR Comment: Thus far, there is a limited collection of data regarding the best management of hospitalized COVID-19 patients especially when considering thrombotic events. This study provides moderate quality evidence that therapeutic anticoagulation likely provides more harm than benefit and should not be used for patients with COVID-19 unless they are being treated for another indication. There is still much work to be done before we can definitively create high quality evidence practice guidelines on the role of anticoagulation in COVID-19 patients.	0	PMC9711512	NO-CC CODE	2022-12-02 23:21:30	no	J Emerg Med. 2022 Oct 30; 63(4):615-616	utf-8	J Emerg Med	2,022	10.1016/j.jemermed.2022.10.026	oa_other
==== Front Pers Individ Dif Pers Individ Dif Personality and Individual Differences 0191-8869 0191-8869 Elsevier Ltd. S0191-8869(21)00355-X 10.1016/j.paid.2021.110980 110980 Article Early COVID-19 quarantine: A machine learning approach to model what differentiated the top 25% well-being scorers Kyriazos Theodoros ⁎ Galanakis Michalis Karakasidou Eirini Stalikas Anastassios Department of Psychology, Panteion University, Athens, Greece ⁎ Corresponding author at: Panteion University, 136 Syngrou Av., 17671 Athens, Greece. 12 5 2021 10 2021 12 5 2021 181 110980110980 3 11 2020 29 4 2021 5 5 2021 © 2021 Elsevier Ltd. All rights reserved. 2021 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. This study focused on the interaction of demographics and well-being. Diener's subjective well-being (SWB) was successfully validated with Exploratory Graph Analysis and Confirmatory Factor Analysis to track well-being differences of the COVID-19 quarantined individuals. Six tree-based Machine Learning models were trained to classify top 25% SWB scorers during COVID-19 quarantine, after data-splitting (train 70%, test 30%). The model input variables were demographics, to avoid overlapping of inputs-outputs. A 10-fold cross-validation method (70%–30%) was then implemented in the training session to select the optimal Machine Learning model among the six tested. A CART classification was the optimal algorithm (Train-Accuracy = 0.77, Test-Accuracy = 0.75). A clean, three-node tree suggested that if someone spends time on perceived creative activities during the COVID-19 quarantine, under clearly described conditions, he/she had high probabilities to be a top subjective well-being scorer. The key importance of creative activities was subsequently cross-validated with three different model configurations: (1) a different tree-based model (Test-Accuracy =0.75); (2) a different operationalization of subjective well-being (Test-Accuracy =0.75) and (3) a different construct (depression; Test-Accuracy =0.73). This is an integrative approach to study individual differences in subjective well-being, bridging Exploratory Graph Analysis and Machine Learning in a single research cycle with multiples cross-validations. Keywords Subjective well-being Machine learning Tree-based models Network psychometrics Exploratory Graph Analysis Confirmatory Factor Analysis COVID-19 ==== Body pmc1 Introduction Loneliness and controlled socializing during early COVID-19 quarantine posed a significant threat for the well-being of quarantined individuals (Brooks et al., 2020). Physical distancing, financial uncertainty, bereavement, and unemployment can put quarantined individuals at risk of post-traumatic stress, depression, anxiety (Mazza et al., 2020), suicide, self-harm, lack of life meaning, relationship breakdown (Holmes et al., 2020), confusion, and social withdrawal (Ingram & Luxton, 2005). Individual differences in demographics were reported during the early COVID-19 containment measures across studies. Females were more distressed than males (Mazza et al., 2020). Unmarried females were also more distressed than married. The same was true for males at a lower magnitude (Srilakshmidevi & Suseela, 2020). Similarly, people with a psychiatric diagnosis, vulnerable health, rich contact history, or lack of daily routines were more distressed (García-Dantas, Justo-Alonso, Rio-Casanova, González-Vázquez, & Sánchez-Martín, 2020). However, years of education and age showed controversial results (Mazza et al., 2020). 1.1 Well-being, individual differences and syndemics The current ongoing research suggests that COVID-19 containment measures can affect well-being with the dynamic approach of syndemics (Holmes et al., 2020). Syndemics are global demographic tendencies (like population aging) interacting with health conditions (like diabetes) to generate individual differences through comorbidities. Well-being is an umbrella term for a number of constructs tapping positive human functioning (Boniwell, David, & Ayers, 2013). It mainly involves subjective well-being (SWB; Diener, Suh, Lucas, & Smith, 1999), flourishing (Seligman, 2011), and happiness (Seligman, 2002). Over the past decades, SWB dominated well-being literature (Boniwell et al., 2013). SWB is related to hedonic well-being tradition (Ryan & Deci, 2001) and means experiencing more positive emotions than negative, and satisfaction in most life domains (Ruini, 2017). Therefore, it involves a life appraisal both affectively and cognitively (Diener et al., 1999). The affective appraisal refers to the experience of moods or emotions in momentary events. The cognitive appraisal refers to satisfaction from how an individual perceives life and the potential discrepancy between the present life situation and the perceived ideal (Boniwell et al., 2013). The term “subjective” implies the potential contrast of objective living conditions (like material goods or health status), and SWB ratings (Diener et al., 1999). This subjective dimension was the main reason it was selected for measuring well-being differences during the COVID-19 containment measures. SWB it has been operationalized by an affect measure and combined with life satisfaction measure (Diener et al., 2010). 1.2 The present study This study focused on the interactions of demographics and subjective well-being differences through syndemics, due to physical distancing of the Greek adult population during early COVID-19 containment measures. The dynamic complexity of differences in subjective well-being through syndemics can be effectively studied with multivariate approaches. Machine Learning (ML) techniques can effectively classify the multivariate complexity of differences in subjective well-being through syndemics. The objectives of this study were to (a) Validate Diener et al.'s (1999) SWB model, facilitating COVID-19 research on well-being differences during the COVID-19 containment measures; (b) Contribute to applied research of individual differences a Machine Learning research cycle with multiple cross-validating steps by comparing Machine Learning models to study what differentiate the SWB of the quarantined individuals; (c) Examine the most important demographic differentiating variables for the optimal Machine Learning model, with an SWB syndemics approach; (d) Examine the replicability of the most important demographic differentiating variables for the SWB in multiple model configurations. 2 Material and methods 2.1 Participants The sample involved 759 adults (78% females). The 25%, of the sample, was 18–40 years, 42% was 41–60 years, 3% was 61–70 years and 1% was over 70. Almost 1 in 2 participants were single/not married (47%), married/living together (40%), divorced/widowed (13%). 59% did not have children. Most participants received tripartite education (88%), or lower (13%). The 31% were private-sector employees, public-sector employees (26%), self-employed (17%), students (10%), unemployed (7%), retired (4%), other (6%). Monthly income ranged from none (13%), ≤600€ (13%), 601–1200€ (41%), 1201–1800€ (21%), >2500€ (13%). There were 98.8% of no-COVID-19 respondents. Respondent's families included 97.5% no-COVID-19 cases. 84% did not belong to a vulnerable group and 64% had a vulnerable family member. There were three demographics rated on a 5-point Likert scale (Table 1 ).Table 1 Frequencies of COVID-19 related demographics answered on a Likert scale. Table 1 Likert scale points Question 1 2 3 4 5 1. Do you engage in creative activities during the quarantine?a 3% 13% 26% 42% 16% 2.The financial impact from quarantine was for you…b 27% 19% 33% 15% 5% 3. Has your daily routine changed during the quarantine?a 2% 12% 21% 40% 25% a 1 = Not at all, 3 = Neither slightly nor strongly, 5 = Very strongly. b 1 = Very low, 3 = Moderate, 5 = Very high. 2.2 Measures 2.2.1 Scale of Positive and Negative Experience 8 (SPANE-8) SPANE-8 (Kyriazos, Stalikas, Prassa, & Yotsidi, 2018a) is a shorter version of SPANE-12 (Diener et al., 2010) with 4 items for SPANE P (Pleasant, Happy, Joyful, Contented) and 4 for SPANE N (Bad, Sad, Afraid, Angry). Items are rated on a 5-point Likert scale from 1 (Very Rarely or Never) to 5 (Very Often or Always), midpoint = Sometimes. The positive and negative scores (SPANE Positive and SPANE Negative) can range from 4 to 20. Their difference (Affect Balance or SPANE B) can range from −16 (unhappiest possible) to 16 (happiest possible). 2.2.2 Satisfaction with Life Scale (SWLS) The SWLS (Diener, Emmons, Larsen, & Griffin, 1985) measures perceived global satisfaction with life (e.g. ‘So far I have gotten the important things I want in life’) on a 7-point scale, from 1 (Strongly Disagree) to 7 (Strongly Agree), midpoint = Neither Agree nor Disagree. The score ranges from 1 to 9 = Extremely dissatisfied, 20 = Neutral, 31–35 = Extremely satisfied. 2.2.3 Mental Health Continuum–Short Form (MHC–SF) MHC–SF (Keyes et al., 2008) is a 14-item questionnaire, measuring emotional (i.e. subjective), social, and psychological well-being in 3 factors (EWB, SoWB, PWB respectively). Items are rated on a 6-point frequency scale (never –every day). Higher scores suggest higher frequency. 2.2.4 Depression Anxiety Stress Scale, short-form (DASS-9) DASS-9 (Yusoff, 2013 and in Greek by Kyriazos, Stalikas, Prassa, & Yotsidi, 2018b) is a briefer DASS-21 (Lovibond & Lovibond, 1995). It measures depression, anxiety, and stress in three 3-item factors, rated on a 4-point scale from 0 (did not apply to me at all) to 3 (applied to me very much, or most of the time). 2.3 Procedure This is a cross-sectional design, using the network sampling method. Data were collected via a web-link posted on webpages and Facebook accounts of the team members. The fields of the digital form were set as “required”. The link was available online from April, 5th 2020 until May 4th, 2020, 6:30 A.M. Note that Greece took early containment measures on February 282,020 locally, and on March 23 nationally. On May 4th containment measures were loosened and eventually removed by June 2020. 2.4 Analytic strategy Generally, Machine Learning (ML) is searching for generalizable predictive patterns to make predictions of optimal precision from a dataset. In contrast, traditional statistics focus on inferring relationships between variables from a sample. An important advantage of ML models is that researchers do not need to assume the distribution of the dependent/independent variables while traditional statistical models work based on a number of distributional and other assumptions (Kassambara, 2018). Additionally, in comparison to traditional statistical analyses ML offers the major advantage of statistical model-training in train data to improve predictions in test data. To identify the most important factors related to SWB, classification and regression three-based ML models were tested. Table S1 (Supplementary material) presents each step of the adopted analytic strategy. Data were analysed with R version 4.0.2. 3 Results Table 2 presents descriptive statistics.Table 2 Descriptive statistics of the study dimensions (N = 759). Table 2Dimension (latent variable) M Mdn SD Range SPANE 8 Positive Experiences (SPANE-8 P) 12.95 13 3.36 4–20 SPANE 8 Negative Experiences (SPANE-8 N) 10.08 10 3.49 4–20 Satisfaction with Life Scale (SWLS) 24.07 25 5.69 7–35 DASS 9 Depression (DASS-9 D) 2.6 2 2.11 0–9 Mental Health Emotional Well-being (EWB) 10.51 11 2.94 0–15 Note. M = Mean, Mdn = Median, SD=Standard deviation. 3.1 Building and validating Diener et al.'s (1999) SWB model Univariate normality was examined with Kolmogorov-Smirnov, Shapiro-Wilk, Shapiro-Francia, and Anderson-Darling tests, p < .001. Multivariate normality was examined with Mardia's multivariate kurtosis and skewness, Henze–Zirkler's consistent test, Doornik–Hansen test, and Energy-test, p < .001. There were no missing values and 16 multivariate outliers, D 2 Critical Value > χ2(13) = 34.53, p < .001. Outliers did not impair findings and they were kept in the dataset, N = 759. 3.1.1 Exploratory Graph Analysis (EGA): building the SWB model EGA is a network psychometrics technique (see Epskamp, Maris, Waldorp, & Borsboom, 2018). It evaluates the number of dimensions without a priori assumptions. Dimensions are equivalent to latent variables (Golino & Epskamp, 2017). A 3-cluster network was identified (Fig. 1 ) with the Glasso estimator (γ = 0.5), as expected (Diener et al., 1999). The first cluster (dimension) contained the 4 SPANE Positive items (2, 3, 6, 8). The second cluster grouped the 4 SPANE Negative items (1,4,5,7). The third contained the 5 SWLS items. The edges connected to the three clusters showed the expected pattern of positive and negative correlations (Fig. 1).Fig. 1 The 3-cluster EGA network to establish Diener et al.'s (1999) SWB model, using SPANE −8 Positive (red cluster with 4 nodes or 1), SPANE-8 Negative (blue cluster with 4 nodes or 2), SWLS (green cluster with 5 nodes or 3). Green edges (positive partial correlations) connected the SWLS nodes with the nodes of the SPANE Positive and red edges (negative partial correlations) connected the SWLS cluster with the nodes of the SPANE Negative. Red edges connected the two SPANE-8 clusters. Edge width indicated the magnitude of the partial correlations. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 1 3.1.2 Confirmatory Factor Analysis (CFA): confirming the SWB model To evaluate the generated EGA network, an equivalent CFA model was estimated (WLSMV estimator). Goodness fit criteria were RMSEA ≤0.06, RMSEA 90% CI ≤ 0.06, CFI ≥ 0.95, TLI ≥ 0.95, SRMR ≤0.08. The results showed a remarkably good fit, χ2(62) = 59.99 (p = .549), RMSEA = 0.000 [90% CI = 0.000, 0.021], CFI = 1.000, TLI = 0.994, SRMR = 0.033. Factor loadings for SPANE Positive ranged from 0.721–0.871, for SPANE Negative 0.547–0.866, and for SWLS 0.633–898. Inter-factor correlations were Ft1 → Ft2 = −0.744, Ft2 → Ft3 = −0.451, and Ft3 → Ft1 = 0.562 (Fig. 2 ). Statistical power based on population RMSEA (MacCallum, Browne, & Sugawara, 1996) suggested N ≥ 183, α = 0.80 (df = 62, N = 759).Fig. 2 Path diagram of Diener et al.'s (1999) SWB model specified to calculate the EGA model fit. SPANE-8 Positive = (Ft1), SPANE-8 Negative = (Ft2), SWLS = (Ft3). Fig. 2 3.1.3 Internal consistency reliability, model-based reliability and validity The reliability of SPANE Positive, SPANE Negative and SWLS was α = 0.88 [95% CI = 0.87, 0.89], 0.79 [95% CI = 0.77, 0.82], and 0.87 [95% CI = 0.85, 0.88] respectively. All calculated ω coefficients (see Table S2 in Supplementary material), and Average Variance Extracted (see Kyriazos, 2018) suggested adequate model-based reliability and convergent validity respectively (Table S2). The greatest lower bound (Jackson & Agunwamba, 1977) was ≥α (see Table S2). 3.2 Using SWB to compare and select machine learning models 3.2.1 Calculation of the SWB class The Affect Balance (SPANE-B) and SWLS scores were rescaled (0–1) with min-max scaling to calculate the SWB scoring rule (Table 3 ).Table 3 The binary SWB scoring rule. Table 3If SWB > 3rd quartile (Q3) then =1 else = 0 3.2.2 Validation dataset The binary SWB classifying rule (Table 3) was used to train ML models. Therefore, the dataset (N = 759) was split into training dataset (70%) and test dataset (30%), see Table 4 .Table 4 The frequencies of the SWB binary classes were held constant across datasets. Table 4 Dataset SWB classes Total (N = 759) Train (n = 532) Test (n = 227) N % N % N % 1 188 30 132 30 56 30 0 571 70 400 70 171 70 3.2.3 Cross-validation dataset In the training dataset, the k-fold cross-validation method (k = 10) was used. Specifically, during training, the model was trained on k-1 (=9) subsets. This process was repeated 10 times to build an overall accuracy metric for the entire train dataset (n = 532). 3.2.4 ML model building and optimal model selection Tree-based ML models were used. Tree-based models generate relatively simple, easy-to-follow if-then conditions, unlike other ML algorithms (Burger, 2018). Six different three-based ML algorithms were evaluated to find the most robust for evaluating differences of the top 25% SWB individuals: Classification and regression trees or CART, C5.0, Random Forest, Conditional Inference Tree (CTREE), Stochastic Gradient Boosting, and Bagged CART. Model evaluation metrics were Accuracy (correctly predicted instances rate), Cohen's kappa (κ), Sensitivity (true positive rate), and Specificity (true negative rate), see Brownlee, 2014. All 6 models had the same 16 demographic input variables, converted into numerical variables (Fig. 3 ). Centering, scaling, and transforming inputs were omitted. Zero-variance and near-zero-variance inputs were not examined because for tree-based models, this may cause model crashes or unstable fit (Brownlee, 2014).Fig. 3 Box and whisker plots by SWB class value for each input (N = 759). Note. area = respondent lives in an absolute lock-down area due to a large number of COVID-19 cases, Creative = perceived creativity of activities during quarantine, routin = change in daily routine during quarantine, finan = financial impact of quarantine, job = respondent's job, diag = respondent was tested positive, f_diag = a family member was tested positive, quaran = be on quarantine, vul = respondent belongs to a vulnerable group, fam_vul = A family member belongs to a vulnerable group, Age = respondent's age, Sex = respondent's gender, Marital = Marital Status, kids = respondent is a parent, Ed lev = Educational level, income = monthly income. Fig. 3 3.2.5 Optimal model Table 5 summarizes the Acc of the 6 models (Train dataset, n = 532), based on the distributions from the 10-fold cross-validation. A mean Acc > 0.50 was the baseline (>chance, Carpenter, Sprechmann, Calderbank, Sapiro, & Egger, 2016), therefore the Acc across models was acceptable, suggesting a learnable problem (Brownlee, 2016). CART showed the highest mean Acc (M = 0.77), and κ (M = 0.19), see Table 5 and Fig. 4 .Table 5 Descriptive statistics for the accuracy of the 6 models evaluated in the train dataset (n = 532). The distributions were calculated from the 10-fold cross-validation method (number of resamples = 10). Table 5Model (R name) Accuracy Kappa Min Q1 Mdn M Q3 Max M CART (rpart) 0.72 0.74 0.77 0.77 0.79 0.83 0.19 C5.0 (c50) 0.70 0.74 0.75 0.75 0.77 0.79 0.18 Random Forest (ranger) 0.70 0.74 0.77 0.76 0.77 0.79 0.16 Conditional Inference Tree (CTREE) 0.74 0.74 0.75 0.75 0.75 0.77 0.05 Stochastic Gradient Boosting (gbm) 0.66 0.70 0.77 0.75 0.79 0.85 0.19 Bagged CART method (treebag) 0.67 0.70 0.732 0.73 0.74 0.81 0.15 Note. Models are compared in terms of mean accuracy and Cohen's Kappa (κ) presented in bold typeface. Fig. 4 Dotplot comparing the accuracy and Cohen's Kappa of the ML models tested in the train dataset (n = 532). Note. rpart = CART, c50 = C5.0, ranger = Random Forest, CTREE = Conditional Inference Tree, gbm = Stochastic Gradient Boosting, treebag = Bagged CART method. Fig. 4 To evaluate further the optimal model, pair-wise differences in Acc were calculated between the distributions of the 6 models tested in the train dataset (n = 532; see Table S3 in Supplementary material), with a Bonferroni correction. CART model generated a clean tree-diagram with 3 terminal nodes, no need for fine-tuning (Fig. 5 ). Specifically, the root node split into 1 terminal node for cases with perceived creativity of activities ≤4 (1–5 scale) and a brunch of 2 more terminal nodes. The first brunch-node classified cases with perceived creativity ≤4 and the second brunch-node cases with perceived creativity = 5. The score code that describes the scoring algorithm reads as follows. Respondents (79% in the train dataset) with a perceived creativity ≤4 had 21% probability to be into the top 25% SWB scorers during the quarantine. Respondents (9% in the train dataset) with a perceived creativity = 5 had a 30% probability to be into the top 25% of SWB scorers during quarantine, on condition that their perceived financial impact of the quarantine was >2 (1–5 scale). Likewise, respondents (7% in the test dataset) with perceived creativity of activities rating = 5, had a 66% probability to be classified into the top 25% of SWB scorers during quarantine, provided that they perceived a financial impact of the quarantine ≤2 (Fig. 5).Fig. 5 Tree-diagram generated by the CART model (train dataset, n = 532). Root-node (perceived creativity of activities during quarantine ≤4) is divided into a terminal node (2) for respondents with perceived creativity of activities rating ≤ 4 rated on a 1–5 scale and a brunch (3) with 2 additional terminal nodes with perceived creativity of activities rating = 5 (6 & 7). Fig. 5 3.2.6 Optimal model evaluation The CART model classification rules were assessed with a confusion matrix in test dataset (n = 227). CART predicted correctly 7/56 cases of SWB = 1, and 164/171 cases of SWB = 0, Acc = 75%, Sensitivity = 96%, Specificity = 13% (Positive Class: 0). Table 6 summarizes the metrics of the CART model in the train and test datasets.Table 6 The metrics of the CART model in the train (n = 532) and test (n = 227) datasets. Table 6 Acc 95% CI Dataset Model Acc Lower Upper p value Sensitivity Specificity Train (n = 532) CART 0.77 0.74 0.81 0.124 0.97 0.19 Test (n = 227) CART 0.75 0.69 0.81 0.535 0.96 0.13 Note. Acc = accuracy, 95% CI = 95% confidence interval. 3.2.7 Most important classification variable Eight classification variables had variable importance ≠ 0. Variable importance is the sum of all goodness of split measures for each split it was the main variable (Therneau, Atkinson, Ripley, & Ripley, 2015). The most important variable for the CART model was the perceived creativity of activities during quarantine, i.e. ‘Do you engage in creative activities during the quarantine?’ (variable importance = 8.99), followed by the financial impact of quarantine (importance = 7.52). Additionally, 8 variables had zero importance (Fig. 6 ).Fig. 6 Variable importance plot in the train dataset (n = 532) showing how each variable contributes to the model (CART). Note. Creativ = perceived creativity of activities during quarantine, finan = financial impact of quarantine, fam_vul = A family member belongs to a vulnerable group, Marital = Marital Status, Ed_lev = Educational Level, vul = respondent belongs to a vulnerable group, quaran = be on quarantine, Sex = respondent's gender, f_diag = a family member was tested positive, area = respondent lives in a universally lock-down area due to the large number of COVID-19 cases, diag = respondent was tested positive, kids = respondent is a parent, routin = change in daily routine during quarantine, income = monthly income, Age = respondent's age, job = respondent's job. Fig. 6 3.3 Cross-validation of the most important classification variable in different model configurations 3.3.1 Different tree-based model × SWB The popular unbiased recursive partitioning algorithm (Schlosser, Hothorn, & Zeileis, 2019) CTREE also showed high accuracy, Acc (M = 0.75), and κ (M = 0.05), in train data (see Table 5 in Optimal model section). The most important classification variable for the CTREE model remained the perceived creativity of activities during quarantine. This cross-validation corroborated the CART variable importance ranking. Two more model configurations were tested for the CART model: (a) On the top 25% SWB scorers during COVID-19 quarantine, training an SWB model operationalized with MHC-SF Emotional Well-Being (Keyes et al., 2008) and (b) On the lowest 50% DASS-9 Depression scorers (Kyriazos et al., 2018b; Lovibond & Lovibond, 1995; Yusoff, 2013) during COVID-19 quarantine, training depression model (train dataset, n = 532). 3.3.2 CART × different SWB operationalization Repeating the above process, an equivalent scoring rule was calculated for MHC-SF EWB (Keyes et al., 2008) to classify the 25% top MHC-SF EWB scorers. Then, the 10-fold cross-validation method was used to train the CART model on the SWB variable operationalized by a different instrument (see model metrics in Table 7 ).Table 7 The goodness of fit metrics of the 3 tree-based models, cross-validating perceived creativity of activities during quarantine was a variable with major contribution to the classification (test data, n = 232). Table 7Model configuration Test purpose MIV Train data Acc (Kappa) Test data Acc (Kappa) 1. CTREE Diener et al.’s (1999) SWB Testing a popular unbiased recursive partitioning algorithm Perceived creativity 0.75 (0.00) 0.75 (0.00) 2. CART × MHC-SF EWB (Keyes et al., 2008) Testing a different SWB operationalization Perceived creativity 0.79 (0.23) 0.75 (0.07) 3. CART × DASS D Testing a different construct Perceived creativity 0.73 (0.39) 0.73 (0.39) Note. MIV = most important variable, Acc = accuracy, Kappa = Cohen's kappa (κ). 3.3.3 CART × different construct (DASS-9 Depression) In yet another replication, when the DASS-9 Depression score was < the 2nd quartile (Q2) it was coded 1, else 0. This procedure generated a binary outcome variable, out of the 50% scorers with the lowest depression. Then, the CART model was trained with the 10-fold cross-validation method in train data n = 532 (see model metrics in Table 7). Perceived creativity of activities during quarantine was ranked the most important classification variable both for the model trained on the emotional well-being (MHC-SF EWB; Keyes et al., 2008) and the model trained on DASS-9 Depression. Table 7 presents the metrics of the three models used for cross-validating the importance of perceived creativity of activities during quarantine (Table 7). 4 Discussion This study attempted to: (a) Use EGA and CFA to validate Diener et al.'s (1999) SWB model facilitating COVID-19 research on well-being differences during the COVID-19 containment measures; (b) Support applied research of individual differences a Machine Learning research cycle of multiple cross-validating steps by comparing six Machine Learning models to select the optimal model. Initially, an EGA network was successfully evaluated with three SWB dimensions (Diener et al., 1999). Subsequently, six ML models were trained on classifying the top 25% SWB scorers during COVID-19 containment measures. All inputs were demographics because (A) The syndemics complexity necessitated multivariate research methods (Holmes et al., 2020). (B) When using demographics to train the ML models, inputs do not affect the calculation of the scoring rule, avoiding indirect overlapping of inputs and outcome variables. For example, even if we excluded all SWB scoring items, an input like ‘On the whole, I am satisfied with myself’ (Rosenberg Self-Esteem Scale) would be highly correlated to ‘I am satisfied with my life’ (SWLS; Diener et al., 1985), used for the SWB score. This could cause problems to training algorithms and generate meaningless node rules, similar to the chicken-egg dilemma. Decision trees and rule-based models were expected to perform better than regression and instance-based models for this classification (Brownlee, 2016). Tree-based models are white-box algorithms, generating easy-to-follow if-then rules (Burger, 2018). Six different tree-based ML models were trained. CART had the highest Accuracy. Additionally, CART was considered suitable for reproducing human behavior in a ‘user-friendly’ classification, unlike the rest of the models generating complicated classifications (Burger, 2018). CART model Accuracy of about 75%, both seen in unseen data was satisfactory for human behavior. The tree rules showed that respondents with lower perceived creativity of activities rating had only 21% probabilities to be in the top 25% SWB scorers. In contrast, respondents with higher perceived creativity of activities had a 30% probability to be into the top 25% of SWB scorers when the perceived financial impact of the quarantine was low (within the 25% range). Respondents with both high perceived creativity and low perceived financial impact of the quarantine had 66% probabilities to be into the top 25% of SWB scorers. The importance of creative activities was subsequently cross-validated using three different model configurations, yielding adequate Accuracy. First, perceived creativity remained the most influencing classification variable for the CTREE model (a popular unbiased recursive partitioning algorithm), corroborating CART results. Second, perceived creativity remained an important contributor to the CART model when SWB was operationalized by the MHC-SF EWB factor (Keyes et al., 2008). Lastly, the same procedure was replicated again to evaluate if perceived creativity remained important when the CART model was trained to predict low depression scorers during the quarantine. Crucially, perceived creativity of activities during quarantine had the greatest influence in all cross-validating classifications, supporting the previous findings. Generally, objective life circumstances can explain a maximum of 10% change in SWB (Diener et al., 1999). Thus, reaching SWB during COVID-19 containment is possible. Moreover, the positive affect SWB component is associated with creativity, resourcefulness (Seligman, 2002), and openness to new experiences (Fredrickson, 2001). Interestingly, environment interacts with cognitive and personality variables to fuel creative behavior. Studies of families and creativity argued that dysfunctional environments may be associated with extreme creativity levels, and happier families with more moderate creativity levels (Kerr, 2009). This association of creativity with dysfunctional contexts seems to hold for the COVID-19 quarantine too. Kapoor and Kaufman (2020) proposed that engaging in creative activities during the COVID-19 quarantine could buffer against the negative effects of the pandemic context. Similarly, a French sample showed significant increase in everyday creative activities during lockdown but not in professional ones. Likewise, creative growth during COVID-19 was associated with a higher flourishing on a sample consisting of 1420 employees from China, Germany, and the United States (Tang, Hofreiter, Reiter-Palmon, Bai, & Murugavel, 2021). 4.1 Conclusions In the nutshell, if individuals get involved in creative activities, regardless of their other individual differences they are likely to be the top 25% SWB scorers. Certainly, it is not clear whether someone performs creative activities and then reaches higher SWB levels or the opposite because temporal precedence requires an experimental setting (see Shadish, Cook, & Campbell, 2002). A limitation is the imbalanced sample across gender and some COVID-19 demographics. Furthermore, multiple assessment methods were impossible. However, ML models with only demographic inputs have minimal self-report bias. Future research could include longitudinal studies on SWB during COVID-19, and the study of the other well-being models with ML. The following are the supplementary data related to this article.Table S1 Description of the analyses performed. Table S1 Table S2 Internal consistency reliability, model-based reliability, model-based convergent validity, and the greatest lower bound estimate for the 3 latent variables in Diener et al.'s (1999) SWB model tested with CFA in the entire sample (N = 759). Table S2 Table S3 Accuracy differences between the distributions of the 6 models tested in the train dataset, n = 532. Upper diagonal = estimates of the difference, lower diagonal = p-value for H0 (Difference = 0). Table S3 Funding This research did not receive any funding. CRediT authorship contribution statement Theodoros Kyriazos: Methodology, Software, Validation, Formal analysis, Data curation, Writing – original draft, Writing – review & editing, Visualization. Michalis Galanakis: Conceptualization, Investigation, Resources, Writing – original draft, Writing – review & editing, Visualization. Eirini Karakasidou: Investigation, Resources, Writing – original draft, Writing – review & editing, Visualization. Anastassios Stalikas: Conceptualization, Writing – review & editing, Supervision, Project administration. Declaration of competing interest The authors declare they have no known conflict of interest. ==== Refs References Boniwell I. David S.A. Ayers A.C. The Oxford handbook of happiness 2013 Oxford University Press London Brooks S.K. Webster R.K. Smith L.E. Woodland L. Wessely S. Greenberg N. Rubin G.J. The psychological impact of quarantine and how to reduce it: Rapid review of the evidence The Lancet 395 2020 912 920 Brownlee J. Machine learning mastery Available at http://machinelearningmastery 2014 (Accessed July, 2020) Brownlee J. Machine learning mastery with R 2016 Author Melbourne Burger S.V. Introduction to machine learning with R 2018 O’Reilly Sebastopol Carpenter K.L.H. Sprechmann P. Calderbank R. Sapiro G. Egger H.L. Quantifying risk for anxiety disorders in preschool children: A machine learning approach Plos One 11 11 2016 e0165524 Diener E. Emmons R.A. Larsen R.J. Griffin S. The satisfaction with life scale Journal of Personality Assessment 49 1985 71 75 16367493 Diener E. Suh E.M. Lucas R.E. Smith H.L. Subjective well-being: Three decades of progress Psychological Bulletin 125 2 1999 276 Diener E. Wirtz D. Tov W. Kim-Prieto C. Choi D.W. Oishi S. Biswas-Diener R. New well-being measures: Short scales to assess flourishing and positive and negative feelings Social Indicators Research 97 2 2010 143 156 Epskamp S. Maris G. Waldorp L.J. Borsboom D. Network psychometrics Irwing P. Hughes D. Booth T. Handbook of psychometrics 2018 Wiley New York 953 986 Fredrickson B.L. The role of positive emotions in positive psychology: The broaden-and-build theory of positive emotions American Psychologist 56 3 2001 218 11315248 García-Dantas A. Justo-Alonso A. Rio-Casanova L.D. González-Vázquez A.I. Sánchez-Martín M. Immediate psychological responses during the early stage of the coronavirus pandemic (COVID-19) in the general population in Spain 2020 (Available at SSRN 3576927) Golino H.F. Epskamp S. Exploratory graph analysis: A new approach for estimating the number of dimensions in psychological research Plos One 12 6 2017 e0174035 Holmes E.A. Multidisciplinary research priorities for the COVID-19 pandemic: A call for action for mental health science Lancet Psychiatry 7 2020 547 560 32304649 Ingram R.E. Luxton D.D. Vulnerability-stress models Development of psychopathology: A vulnerability-stress perspective 2005 32 46 Jackson P. Agunwamba C. Lower bounds for the reliability of the total score on a test composed of nonhomogeneous items: I: Algebraic lower bounds Psychometrika 42 1977 567 578 Kapoor H. Kaufman J.C. Meaning-making through creativity during COVID-19 Frontiers in Psychology 11 2020 595990 33391115 Kassambara A. Machine Learning Essentials (Ed. 1) Available at http://www.sthda.com/english 2018 Kerr B. Creativity Lopez S. The encyclopedia of positive psychology 2009 Blackwell Chichester 254 256 Keyes C.L. Wissing M. Potgieter J.P. Temane M. Kruger A. Van Rooy S. Evaluation of the mental health continuum–short form (MHC–SF) in Setswana-speaking South Africans Clinical Psychology & Psychotherapy 15 3 2008 181 192 19115439 Kyriazos T.A. Applied psychometrics: Writing-up a factor analysis construct validation study with examples Psychology 9 2018 2503 2530 Kyriazos T.A. Stalikas A. Prassa K. Yotsidi V. A 3-faced construct validation and a bifactor subjective well-being model using the Scale of Positive and Negative Experience, Greek version Psychology 9 2018 1143 1175 Kyriazos T.A. Stalikas A. Prassa K. Yotsidi V. Can the depression anxiety stress scales short be shorter? Factor structure and measurement invariance of DASS-21 and DASS-9 in a Greek, non-clinical sample Psychology 9 2018 1095 1127 Lovibond P.F. Lovibond S.H. The structure of negative emotional states: Comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories Behaviour Research and Therapy 33 3 1995 335 343 7726811 MacCallum R.C. Browne M.W. Sugawara H.M. Power analysis and determination of sample size for covariance structure modeling Psychological Methods 1 2 1996 130 Mazza C. A nationwide survey of psychological distress among Italian people during the COVID-19 pandemic: Immediate psychological responses and associated factors International Journal of Environmental Research and Public Health 17 2020 3165 32370116 Ruini C. Positive psychology in the clinical domains: Research and practice 2017 Springer Switzerland Ryan R.M. Deci E.L. On happiness and human potentials: A review of research on hedonic and eudaimonic well-being Annual Review of Psychology 52 2001 141 166 Schlosser L. Hothorn T. Zeileis A. The power of unbiased recursive partitioning: A unifying view of CTree, MOB, and GUIDE E-Print Archive https://arXiv.org/abs/1906.10179 2019 Seligman M.E.P. Authentic happiness: Using the new positive psychology to realize your potential for lasting fulfillment 2002 Free Press New York Seligman M.E.P. Flourish: A new understanding of happiness and wellbeing and how to achieve them 2011 Nicholas Brealey London Shadish W.R. Cook T.D. Campbell D.T. Experimental and quasi-experimental designs for generalized causal inference 2002 Houghton Mifflin Boston Srilakshmidevi B. Suseela V. Psychological issues based on gender and marital status during Covid-19 lockdown period Tathapi 19 8 2020 755 764 Tang M. Hofreiter S. Reiter-Palmon R. Bai X. Murugavel V. Creativity as a means to well-being in times of COVID-19 pandemic Frontiers in Psychology 12 2021 601389 33767644 Therneau T. Atkinson B. Ripley B. Ripley M.B. Package ‘rpart’ 2015 Available online: cran.ma.ic.ac.uk/web/packages/rpart/rpart.pdf Yusoff M.S.B. Psychometric properties of the depression anxiety stress scale in a sample of medical degree applicants International Medical Journal 20 2013 295 300	36471777	PMC9711521	NO-CC CODE	2022-12-02 23:21:30	no	Pers Individ Dif. 2021 Oct 12; 181:110980	utf-8	Pers Individ Dif	2,021	10.1016/j.paid.2021.110980	oa_other
==== Front Lancet Child Adolesc Health Lancet Child Adolesc Health The Lancet. Child & Adolescent Health 2352-4642 2352-4650 Elsevier Ltd S2352-4642(22)00351-0 10.1016/S2352-4642(22)00351-0 Corrections Correction to Lancet Child Adolesc Health 2022; 6: 788–98 30 11 2022 1 2023 30 11 2022 7 1 e1e1 Published by Elsevier Ltd. 2023 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 pmcKracalik I, Oster ME, Broder KR, et al. Outcomes at least 90 days since onset of myocarditis after mRNA COVID-19 vaccination in adolescents and young adults in the USA: a follow-up surveillance study. Lancet Child Adolesc Health 2022; 6: 788–98—In figure 4 of this Article, some of the labels were incorrect. These corrections have been made as of November 30, 2022.	36462507	PMC9711564	NO-CC CODE	2022-12-14 23:54:36	no	Lancet Child Adolesc Health. 2023 Jan 30; 7(1):e1	utf-8	Lancet Child Adolesc Health	2,022	10.1016/S2352-4642(22)00351-0	oa_other
==== Front Lancet Gastroenterol Hepatol Lancet Gastroenterol Hepatol The Lancet. Gastroenterology & Hepatology 2468-1253 Elsevier Ltd. S2468-1253(22)00403-4 10.1016/S2468-1253(22)00403-4 Correspondence An Australian perspective on monkeypox virus and stool transplants Koh Bryant a Cheng Anthea a Chaw Khin a Gosbell Iain B ab a Australian Red Cross Lifeblood, Melbourne, VIC, Australia b School of Medicine, Western Sydney University, Penrith, NSW, Australia 30 11 2022 30 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 pmcWe read with interest Ianiro and colleagues' recommendations1 about monkeypox virus and faecal microbiota transplantation (FMT), also known as stool transplants. This European expert panel indicated donor screening should include the identification of high-risk donors and additional evaluation of prodromal non-specific symptoms and skin lesions, or contact with a case. Additional laboratory testing was deemed not clinically justified. Australian Red Cross Lifeblood screens, collects, and processes stool from voluntary non-remunerated donors at our Perth (Western Australia) facility, primarily providing FMT as a treatment for Clostridioides difficile infection. Key factors for our response for monkeypox included an assessment of local epidemiology, evidence for potential transmission, and review of screening methods. As of Nov 3, 2022, 141 cases of monkeypox were reported nationally, with seven cases in Western Australia.2 Reflecting global reports,3 most cases have been in men who have sex with men and acquired overseas or from a recently returned traveller. Particularly concerning is evidence of viable (and therefore potentially transmissible) virus isolated from rectal swabs. However, emerging laboratory evidence suggests a short duration of viable virus. In a preprint longitudinal study of 74 patients who had rectal swabs collected up to 57 days from symptom onset, monkeypox was not isolated after 15 days.4 By use of a linear mixed effects model, the study estimated viral clearance in 95% of patients by 42 days (95% CI 27–76) in rectal samples. Screening requirements for FMT donors in Australia are mandated nationally by the Therapeutic Goods Order 105.5 The requirements include a screening questionnaire to establish eligibility to donate and before each donation. Screening questions that Australian Red Cross Lifeblood ask donors include general health status within the last 2 weeks, and whether they have had male-to-male sex within the past 12 months, contracted any sexually transmitted infections in the last 12 months, or travelled outside of Australia or New Zealand within the past 3 months. In the case of smallpox vaccination, donors who have received a live vaccine or vaccination overseas are deferred for 12 months. In addition, a physical examination is done by a trained clinician before the first donation and every 90 days in repeat donors. Similar to Ianiro and colleagues' recommendations,1 our assessment concluded that the risk of monkeypox virus in FMT is adequately mitigated for with existing donor exclusions, without the need for additional questions. There are low case numbers with clear epidemiology, and deferral periods which are likely to exceed the duration of viral excretion in stool. In our context, the addition of laboratory testing would not improve safety for donors or recipients. We declare no competing interests. Australian governments fund the Australian Red Cross Lifeblood to provide blood, blood products, and services to the Australian community. ==== Refs References 1 Ianiro G Mullish BH Iqbal TH Minimising the risk of monkeypox virus transmission during faecal microbiota transplantation: recommendations from a European expert panel Lancet Gastroenterol Hepatol 7 2022 979 980 36116455 2 Australian Government Department of Health and Aged Care Monkeypox (MPX) health alert https://www.health.gov.au/health-alerts/monkeypox-mpx/about 3 WHO Multi-country outbreak of monkeypox, external situation report #8—19 October 2022 https://www.who.int/publications/m/item/multi-country-outbreak-of-monkeypox-external-situation-report-8-19-october-2022 4 Suñer C Ubals M Tarín-Vicente EJ Viral dynamics in patients with monkeypox Infection: a prospective cohort study in Spain SSRN 2022 10.2139/ssrn.4248017 (preprint). 5 Australian Government Therapeutic goods (standard for faecal microbiota transplant products) (TGO 105) order 2020 https://www.legislation.gov.au/Details/F2021C01065	36462510	PMC9711856	NO-CC CODE	2022-12-03 23:19:52	no	Lancet Gastroenterol Hepatol. 2022 Nov 30; doi: 10.1016/S2468-1253(22)00403-4	utf-8	Lancet Gastroenterol Hepatol	2,022	10.1016/S2468-1253(22)00403-4	oa_other
==== Front Lancet Microbe Lancet Microbe The Lancet. Microbe 2666-5247 The Author(s). Published by Elsevier Ltd. S2666-5247(22)00334-2 10.1016/S2666-5247(22)00334-2 Editorial Vaccines need equity to really work The Lancet Microbe 30 11 2022 12 2022 30 11 2022 3 12 e888e888 © 2022 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license 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. ==== Body pmcOn Nov 9, 2022, WHO published its Global Vaccine Market Report 2022, which analyses data on global vaccine distribution framed in the context of the COVID-19 pandemic. The document highlights some positive achievements over the past years, notably the development of vaccines against severe diseases mainly affecting low-income and middle-income countries; the strengthening of vaccine distribution networks in 2021 compared with 2019, primarily driven by COVID-19 vaccines; and the demonstration that vaccine manufacture can be scaled up very rapidly. However, the main conclusions of the report are bleak, if unsurprising. Vaccines are inequitably distributed, particularly those in high demand in high-income countries. Vaccine manufacturers’ priorities, driven by profit potential, are unaligned with WHO's. Few, regionally concentrated manufacturers control most of the global vaccine supply. These predicaments were laid bare during the COVID-19 pandemic, and one would have hoped that the public awareness of the value of immunisation, bolstered by the pandemic, would have prompted governments and manufacturers to change their attitude towards vaccine supply. But the 2022 infectious disease landscape includes multiple examples showing that this change has not happened and that few lessons have been learnt from the COVID-19 pandemic. The pandemic itself, heralded as “over” by many high-income countries, still has a large burden, with over 10 million COVID-19 cases and 40 000 deaths reported globally in the past 4 weeks (as of Nov 16, 2022), and vaccine distribution remains highly uneven—only 12% of the approximately 15 billion doses distributed globally were delivered through COVAX. The ongoing monkeypox outbreak is still classified as a Public Health Emergency of International Concern, arguably primarily due to poor management of the global supply of the vaccine approved for use against this disease (Imvanex, Bavarian Nordic, Kvistgård, Denmark). Indications of human-to-human transmission and possible sexual transmission reported in the 2017 monkeypox outbreak in Nigeria should have prompted the investment of resources in surveillance and further investigation of the epidemiology of the disease. High-income countries with a stockpile of vaccine, most notably the USA, could have contributed vaccine doses to endemic regions to contain outbreaks, which would have reduced the likelihood of the disease being exported to non-endemic regions and of the current outbreak happening. Even as the outbreak began, high-income countries scrambled to secure doses of Imvanex for domestic use from the single manufacturer worldwide, once again disregarding the needs of endemic countries. 2022 was also marked by the environmental detection of circulating vaccine-derived polio virus type 2 in the UK and the USA, previously considered polio-free, and a resurgence of wild poliovirus, with 29 cases reported thus far (up from six cases in 2021), including in non-endemic southeast Africa. Suboptimal immunisation coverage underlies the potential for the emergence and spread of vaccine-derived and wild polio, underscoring the need for enhanced vaccination campaigns to eradicate the disease globally. It is encouraging that on Oct 18, global leaders pledged US$2·6 billion towards this end. A surge in cholera outbreaks—reported in 29 countries in 2022—is another ongoing crisis affected by vaccine supply issues. The global stockpile of oral cholera vaccine doses is running very low, forcing the International Coordinating Group, which oversees the emergency use of this vaccine, to temporarily adopt a one-dose strategy to respond to the outbreaks, in lieu of the standard two-dose regimen. The steep rise in demand could not have been foreseen, but the inability of manufacturers to increase production in the short term is emblematic of the scarce flexibility of the vaccine supply system to respond rapidly to fluctuations in demand—one of the limitations identified in WHO's report. The 2022 monkeypox, polio, and cholera outbreaks exemplify the issues identified by WHO's report, not least the tendency of high-income countries to treat outbreaks as a domestic problem instead of a global one. We urge governments, vaccine manufactures, and international organisations to embrace the recommendations delineated in the report to adopt strategic and internationally coordinated approaches to generate vaccine supplies, with the goal of achieving a more equitable global distribution of vaccines and improved preparation to face future outbreaks and pandemics. For the Global Vaccine Market Report 2022 see https://www.who.int/publications/m/item/global-vaccine-market-report-2022 For the global COVID-19 burden see https://coronavirus.jhu.edu/map.html For more on global leaders’ pledge for ending polio see https://www.who.int/news/item/18-10-2022-global-leaders-commit-usd-2.6-billion-at-world-health-summit-to-end-polio For more on the International Coordinating Group change in strategy see https://www.who.int/news/item/19-10-2022-shortage-of-cholera-vaccines-leads-to-temporary-suspension-of-two-dose-strategy--as-cases-rise-worldwide Vaccine inequality and unfair global vaccination rollout as rich privileged versus poor underprivileged countries health inequity and unequal distribution of vaccines in a pandemic or world virus outbreak with 3D illustration elements.© 2022 wildpixel/iStock 2022	36462523	PMC9711857	NO-CC CODE	2022-12-03 23:19:52	no	Lancet Microbe. 2022 Dec 30; 3(12):e888	utf-8	Lancet Microbe	2,022	10.1016/S2666-5247(22)00334-2	oa_other
==== Front J Affect Disord J Affect Disord Journal of Affective Disorders 0165-0327 1573-2517 Published by Elsevier B.V. S0165-0327(21)00607-8 10.1016/j.jad.2021.06.021 Review Article Prevalence of mental health problems among children and adolescents during the COVID-19 pandemic: A systematic review and meta-analysis Ma Lu a1 Mazidi Mohsen b1 Li Ke a Li Yixuan a Chen Shiqi a Kirwan Richard d Zhou Haixia a Yan Na a Rahman Atif c Wang Weidong e⁎ Wang Youfa a⁎⁎ a Global Health Institute, School of Public Health, Xi'an Jiaotong University Health Science Center, Xi'an, Shaanxi, China b Department of Twin Research and Genetic Epidemiology, King's College London, London, United Kingdom c Department of Primary Care and Mental Health, Institute of Population Health, University of Liverpool, Liverpool, United Kingdom d School of Biological and Environmental Sciences, Faculty of Science, Liverpool John Moores University, Liverpool, United Kingdom e Department of Sociology, School of Sociology and Population Studies, Renmin University of China, Beijing, China ⁎ Corresponding author at: Department of Sociology, School of Sociology and Population Studies, Renmin University of China, Beijing 100000, China. ⁎⁎ Corresponding author at: Global Health Institute, School of Public Health, Xi'an Jiaotong University Health Science Center, Xi'an, Shaanxi 710061, China. 1 Equal contributors 18 6 2021 1 10 2021 18 6 2021 293 7889 27 9 2020 6 6 2021 13 6 2021 © 2021 Published by Elsevier B.V. 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. Background This systematic review and meta-analysis examined the prevalence of depression, anxiety, sleep disorders, and posttraumatic stress symptoms among children and adolescents during global COVID-19 pandemic in 2019 to 2020, and the potential modifying effects of age and gender. Methods A literature search was conducted in PubMed, Web of Science, PsycINFO, and two Chinese academic databases (China National Knowledge Infrastructure and Wanfang) for studies published from December 2019 to September 2020 that reported the prevalence of above mental health problems among children and adolescents. Random-effects meta-analyses were used to estimate the pooled prevalence. Results Twenty-three studies (21 cross-sectional studies and 2 longitudinal studies) from two countries (i.e., China and Turkey) with 57,927 children and adolescents were identified. Depression, anxiety, sleep disorders, and posttraumatic stress symptoms were assessed in 12, 13, 2, and 2 studies, respectively. Meta-analysis of results from these studies showed that the pooled prevalence of depression, anxiety, sleep disorders, and posttraumatic stress symptoms were 29% (95%CI: 17%, 40%), 26% (95%CI: 16%, 35%), 44% (95%CI: 21%, 68%), and 48% (95%CI: -0.25, 1.21), respectively. The subgroup meta-analysis revealed that adolescents and females exhibited higher prevalence of depression and anxiety compared to children and males, respectively. Limitations All studies in meta-analysis were from China limited the generalizability of our findings. Conclusions Early evidence highlights the high prevalence of mental health problems among children and adolescents during the COVID-19 pandemic, especially among female and adolescents. Studies investigating the mental health of children and adolescents from countries other than China are urgently needed. Keywords Mental health problems COVID-19 pandemic Children Adolescents Review ==== Body pmc1 Introduction The COVID-19 (Coronavirus Disease 2019) pandemic has affected the mental health (e.g., depression, anxiety, sleep disorders, and posttraumatic stress symptoms) of children and adolescents (Golberstein et al., 2020). As of April 8, 2020, schools have been suspended nationwide in 188 countries (Lee, 2020). Prolonged school closures, strict social isolation from peers, teachers, extended family, and community networks, economic shutdown, and the pandemic itself have contributed to the mental health problems of children and adolescents (Holmes et al., 2020; Tan et al., 2020). While some children may benefit from increased interaction with parents and siblings, many have experienced elevated levels of emotional distress (Sprang and Silman, 2013; Xie et al., 2020). Being confined to home leads to disturbances in sleep/wake cycles and physical exercise routines, and promotes excessive use of technology (Xie et al., 2020). The pandemic may increase family financial stressors and parental unemployment, which were associated with short- and long-term consequences on child mental health (Costello et al., 2003). There is also an increased risk of seeing or experiencing domestic violence and emotional, physical and/or sexual abuse (Costello et al., 2003). It is assumed that relaxing lockdown restrictions and returning to school might improve the mental health status of children as the economy and social practices begin to normalize globally (Tan et al., 2020). Understanding the psychological impact of the COVID-19 pandemic on children and adolescents would provide a theoretical basis for designing interventions, planning resources, and promulgating policies necessary to protect young people from such occurrences in future (Pappa et al., 2020). Several original studies have found high levels of mental health problems among children and adolescents during the COVID-19 pandemic (Duan et al., 2020; Pınar Senkalfa et al., 2020; Türkoğlu et al., 2020). However, to the best of our knowledge to date, no systematic review to synthesize the impact of the pandemic on their mental health has been performed. While there are some systematic reviews on the psychological impacts of COVID-19 on patients and healthcare workers (Pappa et al., 2020; Luo et al., 2020; Rogers et al., 2020), evidence in children and adolescents is lacking. The aim of this systematic review and meta-analysis was to examine the emerging evidence of the effects of the COVID-19 outbreak on the mental health of children and adolescents aged 18 years and under. In particular, we aimed to examine the prevalence of depression, anxiety, sleep disorders, and posttraumatic stress symptoms among uninfected/not known to be infected children and adolescents during the active phase of the pandemic during 2019 to 2020. The potential modifying effects of age and gender on the prevalence were also examined. 2 Methods and materials This study was developed and reported according to the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) (Moher et al., 2009) and other standards (Johnson and Hennessy, 2019). The study protocol was registered with the International Prospective Register of Systematic Reviews, PROSPERO (registration no: CRD42020205166). 3 Literature search and study selection A systematic search was performed in three English electronic bibliographic databases: PubMed, PsycINFO, and Web of Science, and two Chinese academic databases: China National Knowledge Infrastructure (CNKI) and Wanfang. The following search terms were used: (“Novel coronavirus” OR “SARS-COV-2” OR “COVID-19” OR “2019-nCov”) AND (“depression” OR “anxiety” OR “sleep” OR “posttraumatic stress symptoms”, “mental health” OR “psychological*” OR “psychiatry” OR “insomnia”). The specific search algorithm is provided in Supplemental Table 1. Studies reported the prevalence of self-reported mental health problems and symptoms were included. Two authors independently searched the same database with these search terms to ensure that none of the relevant studies was missed. Titles and abstracts of the articles identified were screened against the study selection criteria by two independent reviewers. Potentially relevant articles were retrieved for an evaluation of the full text. Inter-rater agreement was assessed using the Cohen's kappa (k=0.64). Disagreements were reviewed and resolved through discussion with third author to resolve persistent inconsistencies. This search strategy was further supplemented with hand searching of reference lists of included articles and through tracking the citations of eligible references in Google Scholar. Articles identified from the reference lists were further screened and evaluated by using the same criteria. Reference searches were repeated on all newly identified articles until no additional relevant articles were found. 3.1 Study selection criteria Studies were included if they: (a) evaluated the prevalence of depression, anxiety, sleep disorders, and posttraumatic stress symptoms using validated assessment method among children and adolescents aged 18 years and under; (b) were written in English or Chinese; (c) were carried out between December 2019 to September 2020; and (d) were cross-sectional or longitudinal studies. When there were studies involving the same participants, only the most comprehensive or recent publication was included. Studies were excluded if they: (a) were qualitative studies, case reports, editorials, protocols, meta-analysis, or reviews, (b) computer-based simulation studies with no human participants, c) included participants with COVID-19 infected, d) studies that did not provide data on the levels of the outcomes of interest, or e) studies focused on the prevalence of suicidal behaviours, suidal ideations and attempts among children and adolescents during COVID-19 pandemic. 3.2 Data extraction and preparation A standardized data extraction form was developed to extract the following data from each article: author, study design, country, survey years, average age of participants, sample size (percentage of male participants), sampling strategy, mental health problems, diagnostic or screening instrument used, specific diagnostic criteria or screening instrument cutoff, and reported prevalence estimates of mental health problems. The data were extracted independently by two independent reviewers, and disagreements were reviewed and resolved through discussion with third reviewer to resolve persistent inconsistencies. 3.3 Study quality assessment Two authors independently assessed the quality of the articles using the U.S. National Heart, Lung, and Blood Institute's Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies. (Study Quality Assessment Tools, 2021) The assessment tool rates each study based on 14 criteria. For each criterion, a score of one was assigned if “yes” was the response, whereas a score of zero was assigned otherwise (i.e., an answer of “no,” “not applicable,” “not reported,” or “cannot determine”). Overall quality was rated based on the total score of the scale: “7 ≤ total score” = good, “4<total score≤6” = fair, “total score<4” = poor. The risk of bias of each study decreased with the increase in the total score. 3.4 Statistical analysis Prevalence estimates of mental health problems were calculated by pooling the study-specific estimates using random-effects (using the DerSimonian‐Laird method) meta-analyses that accounted for between-study heterogeneity (Borenstein et al., 2010). When studies reported point prevalence estimates made at different periods within the year, the overall period prevalence was used. Study heterogeneity was assessed using the I2 index and Tau-squared (T2). The level of heterogeneity represented by I2 was interpreted as modest (I2≤25%), moderate (25%<I2≤50%), substantial (50%<I2≤75%), or considerable (I2>75%). Sensitivity analyses was performed by serially excluding each study to determine the influence of individual studies on the overall prevalence estimates. Results from studies grouped according to prespecified study-level characteristics were compared using stratified meta-analysis (gender, diagnostic criteria or screening instrument, region, and country). Publication bias was assessed by a visual inspection of contour-enhanced funnel plots and Egger's regression tests. All statistical analyses were conducted in STATA with specific commands (e.g., Metan and Metareg) (Version 14.0; Stata Corp., College Station, Texas, U.S.). All analyses used two-sided tests, and p-value < 0.05 was considered statistically significant. 4 Results 4.1 Characteristics of included studies A total of 23 studies were included in the systematic review, the characteristics of which are summarized in Table 1 . These studies were published predominantly from February to May 2020 with one longitudinal study including data from October 2019. The vast majority of studies were from China (21 studies), with the remaining studies from Turkey (2 studies). The sample size of these studies varied greatly, ranging from 46 to 9,554 participants.Table 1 Characteristics of the 23 studies included in the review Table 1Author/Survey time (Year, month) Study design Country Age, years (mean±SD or range) Sample size (Boys, %) Participant type a Assessment method & cutoff score Mental health problems (n, %/M±SD) Depression Anxiety Sleep disorders Posttraumatic stress symptoms 1.Türkoğlu S/ 2020, May (Türkoğlu et al., 2020) Cross-sectional Turkey Mean:7.89/4-17 46(82.6%) Autism Spectrum Disorder AuBC, CSHQ (>41); Diagnosed by health providers NR NR Total CSHQ scores increased from 47.82 ± 7.13 to 50.80 ± 8.15 NR 2.Senkalfa BP/2020, April (Pınar Senkalfa et al., 2020) Cross-sectional Turkey Cystic Fibrosis group 0-18 Control group 0-18 Cystic Fibrosis group 45 (51.1%) Control group 90(51.1%) Cystic Fibrosis STAI; Diagnosed by health providers NR Children aged 13–18 years in the control group:29.0 (27.8-32.3); Age-matched children with Cystic Fibrosis 41.5 (35.5-46.3) NR NR 3.Chen F/2020, April (Chen et al., 2020) Cross-Sectional China 6-15 Children:6-12 Adolescents:13-15 1036(51.0%) General DSRS-C (≥15), SCARED (≥25); Self-reported by participants 122(11.8%) 196(18.9%) NR NR 4.Chen IH/2019, October-2020, March Chen et al., 2020 Longitudinal China 10.88±0.72 543(49.0%) General DASS-21; Self-reported by participants Mean:1.22 95%CI: (1.19,1.25) NR NR NR 5. Qi M/2020, March Qi et al., 2020 Cross-sectional China Adolescents:14-18 7202(46.4%) General PHQ-9 (≥5), GAD-7 (≥5); Self-reported by participants 3207(44.5%) 2736(38.0%) NR NR 6.Zhou SJ/2020, March Zhou et al., 2020 Cross-sectional China Adolescents:12-18 8079(46.5%) General PHQ-9 (≥5), GAD-7 (≥5); Self-reported by participants 3533(43.7%) 3020(37.4%) NR NR 7.Xie X/ 2020, February-2020, March Xie et al., 2020 Cross-sectional China NR 1784(56.7%) General CDI-S (≥7), SCARED (≥23); Self-reported by participants 403(22.6%) 337(18.9%) NR NR 8.Zhu KH/2020, February-2020, March Zhu et al., 2020 Cross-sectional China NR 1264(55.9 %) General SCARED (≥23); Self-reported by participants NR 234(18.5 %) NR NR 9.Lin L/2020, February Lin et al., 2020 Cross-sectional China NR 76(NR) General ISI(≥10), PHQ-9 (≥10), GAD-7 (≥10), ASDS (≥28); Self-reported by participants NR NR 24(31.6%) NR 10.Liu Z/2020, February Vindegaard and Benros, 2020 Longitudinal China Children: 4-6 1619(48.9%) General CSHQ (≥41); Reported by caregivers of participants NR NR 900(55.6%) NR 11.Qi H/2020, February Qi et al., 2020 Cross-sectional China Adolescents:11-20 9554(NR) General GAD-7 (≥5); Self-reported by participants NR 1814(19.0%) NR NR 12.Zhou J/2020, February Zhou et al., 2020 Cross-sectional China Adolescents:11-18 4805(0.0%) General CES-D (≥16); Self-reported by participants 1899(39.5%) NR NR NR 13.Li SW/2020, February Li et al., 2020 Cross-sectional China 12.82±2.61/8-18 396(50.3%) General SCARED(≥25); Self-reported by participants NR 87(22.0%) NR NR 14.Mo DM/2020, February Mo et al., 2020 Cross-sectional China 7-16 Children:7-12 Adolescents:13-16 5392(54.5%) General SCARED(≥23); Self-reported by participants NR 1045(19.4%) NR NR 15.Tang S/2020, February Tang and Pang, 2020 Cross-sectional China 640 primary school students and 233 junior high school students: NR 873(52.3%) General SAS(standard score ≥50) CDI(>19); Self-reported by participants Children: 41(6.4%); Adolescents: 61(26.2%) Children: 19(3.0%); Adolescents:46(19.7%) NR NR 16.Wang Y/2020, February Wang et al., 2020 Cross-sectional China 12.82±2.61/8-18 396(50.3%) General DSRS(≥15); Self-reported by participants 41(10.4%) NR NR NR 17.Yu QX/2020, February Yu et al., 2020 Cross-sectional China NR 2074(52.4%) General Psychological Questionnaire for Sudden Public Health Events (each factor score≥2); Self-reported by participants 53(2.6%) 13(0.6%) NR NR 18.Zhang Y/2020, February Zhang et al., 2020 Cross-sectional China NR 4225(47.4%) General PCL-C(≥39); Self-reported by participants NR NR NR 448(10.6%) 19.Liu X/ 2020, January-2020, February Liu et al., 2020 Cross-sectional China NR 34(NR) General STAI, SDS (≥50); Self-reported by participants 13(38.2%) NR NR NR 20.Hou TY/2020, NR Hou et al., 2020 Cross-sectional China NR 859(61.4%) General PHQ-9 (≥10), GAD-7 (≥8), IES-R(≥26); Self-reported by participants 614(71.5%) 468(54.5%) NR 735(85.5%) 21.Li D/ 2020, NR Duan et al., 2020 Cross-sectional China 7-18 Children:7-12 Adolescents:13-18 3613(50.2%) General SCAS, CDI(≥19); Self-reported by participants 805(22.3%) Children: 23.87 ± 15.79 Adolescents:29.27 ± 19.79 NR NR 22.Tang L/2020, NR Tang and Ying, 2020 Cross-sectional China 14.01±1.56 3512(49.1%) General MMHI-60(each factor score ≥2); Self-reported by participants 924(26.3%) 1047(29.8%) NR NR 23.Wang NX/2020, NR Wang and Xu, 2020 Cross-sectional China NR 410(31.5%) General GAD-7(≥5); Self-reported by participants NR 197(48.0%) NR NR NR: Not reported. DSRS-C: Depression Self-Rating Scale for Children; SCARED: Screen for Child Anxiety Related Emotional Disorders; DASS-21: Depression, Anxiety, Stress Scale 21; PHQ-9: 9-item Patient Health Questionnaire; GAD-7: 7-item Generalized Anxiety Disorder Scale; IES-R: Impact of Events Scale - Revised; SCAS: Spence Child Anxiety Scale; STAI: State and Trait Anxiety Inventory; SDS: Self-rating Depression Scale; CSHQ: Children's Sleep Habit Questionnaire; SCL-90: Symptom Checklist-90; AuBC: Autism, Behavior Checklist; CDI-S: Children's Depression Inventory–Short Form; ISI: Insomnia Severity Index; CES-D: Center for Epidemiologic Studies Depression Scale; SAS: Self-Rating Anxiety Scale; DSRS: Depression Self-rating Scale for Children; PCL-C: The PTSD Cheeklist-CivilianVersion; MMHI-60: Mental Health Inventory of Middle-school students. Two of the studies used teleconference survey, the others used online survey. Two of the studies used random cluster sampling, the others used purposive sampling, snowball sampling, and convenient sampling (Table 2 ).Table 2 Survey and sampling method of the 23 studies included in the review. Table 2Author/Survey time (Year, month) Survey method Sampling method Probability sampling Nonprobability sampling 1. Türkoğlu S/ 2020, May Teleconference survey Purposive sampling 2. Senkalfa BP/2020, April Teleconference survey Control group: Purposive sampling Age-matched children with Cystic Fibrosis: Snowball sampling 3. Chen F/2020, April Online survey Purposive sampling 4. Chen I/2019, October-2020, March Online survey Purposive sampling 5. Qi M/2020, March Online survey Purposive sampling 6. Zhou S/2020, March Online survey Purposive sampling 7. Xie X/ 2020, February-2020, March Online survey Purposive sampling 8. Zhu KH/2020, February-2020, March Online survey Random cluster sampling 9. Lin L/2020, February Online survey Snowball sampling 10. Liu Z/2020, February Online survey Convenient sampling 11. Qi H/2020, February Online survey Snowball sampling 12.Zhou J/2020, February Online survey Snowball sampling 13. Li SW/2020, February Online survey Snowball sampling 14. Mo DM/2020, February Online survey Purposive sampling 15.Tang S/2020, February Online survey Purposive sampling 16. Wang Y/2020, February Online survey Snowball sampling 17. Yu QX/2020, February Online survey Purposive sampling 18. Zhang Y/2020, February Online survey Purposive sampling 19. Liu X/ 2020, January-2020, February Online survey Snowball sampling 20. Hou T/2020, NR NR Random cluster sampling 21. Li D/ 2020, NR Online survey Convenient sampling 22. Tang L/2020, NR Online survey Purposive sampling 23. Wang NX/2020, NR Online survey Purposive sampling NR: Nor reported. The study design and populations were diverse. There were 21 cross-sectional studies and 2 longitudinal studies. The majority of studies were carried out in healthy populations (21 studies), in a population with cystic fibrosis (1 study) and autism spectrum disorder (1 study). Some studies included adult participants in which case only data from child/adolescent participants was used in these analyses and participants’ ages ranged from 0 to 18 years. Of particular note is the diversity of mental health-related scales used among these studies which included. A brief description of each mental health scale follows (in order of descending frequency):• 7-item Generalized Anxiety Disorder Scale (GAD-7) (6 studies): a self-report screening tool for generalized anxiety symptoms in the primary care setting consisting of 7 questions and validated in adolescents (Mossman et al., 2017); • Screen for Child Anxiety Related Emotional Disorders (SCARED) (5 studies): a self-report instrument for children and their parents that screens for several types of anxiety disorders including generalized anxiety disorder, separation anxiety disorder, panic disorder, and social anxiety disorder (Monga et al., 2000); • 9-item Patient Health Questionnaire (PHQ-9) (4 studies): a self-questionnaire consisting of nine items that assess the presence and severity of depressive symptoms based on the DSM-IV criteria for major depressive disorder (MDD) (Richardson et al., 2010); • Depression Self-Rating Scale for Children (DSRS-C) (4 studies) is widely used to measure children's depressive symptoms and consists of 18 items (Ivarsson et al., 1994); • Children's Depression Inventory (including short form) (CDI-S) (3 studies): a self-report scale consisting of 27-items which evaluates the severity of depression in children and adolescents (Allgaier et al., 2012). • State and Trait Anxiety Inventory (STAI) (2 studies): assesses state and trait anxiety in children for the determination of anxiety disorder and contains two scales of 20 items each (Nunn, 1988); • Self-rating Depression Scale (SDS) (2 studies): is used to assess depressive syndrome and is validated in Chinese urban children (Su et al., 2003); • Children's Sleep Habit Questionnaire (CSHQ) (2 studies): a parent administered survey to assess children's sleep problems and consists of as 48 items divided into 5 scales focusing on different aspects of sleep behaviour (Tan et al., 2018); • Autism, Behavior Checklist (AuBC) (1 study): designed for the identification of children suspected of having autism and consisting of a list of atypical behaviors characteristic of the pathology (Sevin et al., 1991); • Center for Epidemiologic Studies Depression Scale (CES-D) (1 study):screens for depressive disorders in population‐based samples and is based on a multidimensional approach to measuring depression in children and adolescents aged 6 and 17 years (Li et al., 2010); • Self-Rating Anxiety Scale (SAS) (1 study): a norm-referenced screener that, in conjunction with the Self-rating Depression Scale has been shown to discriminate anxiety from mood disorders (Dunstan and Scott, 2020); • Impact of Events Scale-Revised (IES-R) (1 study):a widely used, 22 item questionnaire used to determine the degree of distress a patient feels in response to trauma and for identifying traumatic stress (Creamer et al., 2003); • Spence Child Anxiety Scale (SCAS) (1 study): a 38-item parents-report measure of anxiety symptoms for children and adolescents developed using community samples (Wang et al., 2016); • Insomnia Severity Index (ISI) (1 study): a brief self-report instrument measuring the patient's perception of both nocturnal and diurnal symptoms of insomnia and comprising seven items (Gagnon et al., 2013); • Depression, Anxiety, Stress Scale 21 (DASS-21) (1 study): a set of three self-report scales designed to measure the emotional states of depression, anxiety and stress with each scale containing 7 items (Wang et al., 2016); • The PTSD Checklist-Civilian Version (PCL-C) (1 study): a standardized self-report rating scale comprising 17 items that correspond to the key symptoms of PTSD (Blanchard et al., 1996); • Mental Health Inventory of Middle-school students (MMHI-60) (1 study): a total of 60 items in the scale are used to measure the level of mental health of middle school students (Wang et al., 1997); • Psychological Questionnaire for Sudden Public Health Events (PQSPHE) (1 study): a total of 25 items to measure depression, neurosism, fear, obsessive anxiety, and hypochondria among adolescents (Zhang, 2005). 4.2 Prevalence of mental health problems among children and adolescents 4.2.1 Depression 12 studies provided data on the prevalence of depression among children and adolescents during the COVID-19 pandemic. Meta-analysis of the results from these studies showed that the pooled prevalence of depression among children was 29% (95%CI: 17%, 40%) with a pooled heterogeneity of 99.9% (p < 0.001). The prevalence of depression reported in individual study ranges from 10% to 71% ( Fig. 2 ).Fig. 1 Flowchart of the literature search and study selection according to the PRISMA standard Fig. 1 Fig. 2 Meta-analysis of the pooled prevalence of depression among children and adolescents (n=13) Abbreviations: DSRS-C: Depression Self-Rating Scale for Children; PHQ-9: 9-item Patient Health Questionnaire; CDI-S: Children's Depression Inventory–Short Form; CES-D: Center for Epidemiologic Studies Depression Scale; CDI: Children's Depression Inventory; DSRS-C: Depression Self-Rating Scale for Children; PQSPHE: Psychological Questionnaire for Sudden Public Health Events; SDS: Self-rating Depression Scale; MMHI-60: Mental Health Inventory of Middle-school students. Prevalence was calculated based on the random-effect models. Fig. 2 Sub-group analysis by age indicated that the prevalence of depression in adolescents age 13-18 years (34.4%, 95%CI: 18.2%, 50.7%; p<0.001) was higher than that of children age ≤ 12 years (11.8%, 95%CI: 1.3%, 22.3%, p=0.028). Sub-group analysis by gender showed that the prevalence of depression in females (33.9%, 95%CI: 24.6%, 43.1%, p<0.001) was higher than that in males (28.9%, 95%CI: 14.1%, 43.7%, p<0.001) (Table 3 ).Table 3 Total and subgroup meta-analysis of pooled prevalence d (%, 95%CI) of depression, anxiety, sleep disorders, and posttraumatic stress symptoms among children during the COVID-19 pandemic based on the included studies a Table 3Type of analysis Groups N of studies Prevalence (%, 95% CI) P value Heterogeneity I2 (%) χ2 P Tau-squared Depression Total 12 28.6 (17.2, 40.1) <0.001 99.9 8025.91 <0.001 0.0405 Anxiety Total 13 25.5 (16.0, 35.1) <0.001 99.9 10690.46 <0.001 0.0307 Sleep disorders b Total 2 44.2 (20.7, 67.7) <0.001 94.8 19.22 <0.001 0.0273 Posttraumatic stress symptoms c Total 2 48.0 (-25.4, 121.4) 0.200 100 3364.24 <0.001 0.2804 Depression Children (≤12 years) 3 11.8 (1.3, 22.3) 0.028 98.9 183.35 <0.001 0.0085 Adolescents (13-18 years) 8 34.4 (18.2, 50.7) <0.001 99.9 7695.86 <0.001 0.0548 Anxiety Children 6 15.7 (9.0, 22.3) <0.001 98.7 389.71 <0.001 0.0066 Adolescents 11 29.1 (17.1, 41.1) <0.001 99.9 10269.07 <0.001 0.0407 Depression Male 4 28.9 (14.1, 43.7) <0.001 99.6 670.02 <0.001 0.0228 Female 5 33.9 (24.6, 43.1) <0.001 99.2 506.21 <0.001 0.0110 Anxiety Male 7 22.3 (14.2, 30.4) <0.001 99.1 650.31 <0.001 0.0118 Female 7 27.4 (20.3, 34.6) <0.001 98.6 431.08 <0.001 0.0091 a All the studies included in meta-analysis were from China and among general children and adolescents, so no subgroup meta-analysis was conducted based on country and pre-existing conditions of children and adolescents. b Only two articles were found on sleep disorders, one was conducted among children and both boys and girls, the age and gender of participants in the other study were not reported, so no subgroup analysis was conducted based on age and gender. c Only two articles were found on posttraumatic stress symptoms, both studies did not report the age of participants and the gender-stratified prevalence, so no subgroup meta-analysis was conducted based on age and gender. d Prevalence was calculated based on the random-effect models. 4.2.2 Anxiety A total of 13 studies provided data on the prevalence of anxiety among children and adolescents during the pandemic. Meta-analysis of the results from these studies showed that the pooled prevalence of anxiety among children and adolescents was 26% (95%CI: 16%, 35%) with a pooled heterogeneity of 99.9% (p< 0.001). The prevalence of anxiety reported in individual study ranges from 7% to 55% (Fig. 3 ).Fig. 3 Meta-analysis of the pooled prevalence of anxiety among all children and adolescents (n=12) Abbreviations: SCARED: Screen for Child Anxiety Related Emotional Disorders; GAD-7: 7-item Generalized Anxiety Disorder Scale; SAS: Self-Rating Anxiety Scale; PQSPHE: Psychological Questionnaire for Sudden Public Health Events; MMHI-60: Mental Health Inventory of Middle-school students. Prevalence was calculated based on the random-effect models. Fig. 3 Sub-group analysis by age indicated that prevalence of anxiety in adolescents age 13-18 years (29.1%, 95%CI: 17.1%, 41.1%, p<0.001) was higher than that in children age ≤ 12 years (15.7%, 95%CI: 9.0%, 22.3%, p<0.001). Sub-group analysis by gender showed that the prevalence of anxiety of females (27.4%, 95%CI: 20.3%, 34.6%, p<0.001) was higher than that of males (22.3%, 95%CI: 14.2%, 30.4%, p<0.001) (Table 3). 4.2.3 Sleep disorders Only 2 studies provided data on the prevalence of sleep disorders among children and adolescents. Meta-analysis of the results of the two studies showed that the pooled prevalence of sleep disorders was 44% (95%CI: 21%, 68%) with a pooled heterogeneity of 94.8% (p<0.001). The prevalence of sleep disorders of the two studies were 32% to 56%, respectively (Fig. 4 ). Sub-group analyses by age and gender were not performed due to lack of data.Fig. 4 Meta-analysis of the pooled prevalence of sleep disorders among all children and adolescents (n=2) Abbreviations: ISI: Insomnia Severity Index; CSHQ: Children's Sleep Habit Questionnaire. Prevalence was calculated based on the random-effect models. Fig. 4 4.2.4 Posttraumatic stress symptoms Only 2 studies provided data on the prevalence of post-traumatic stress symptoms among children and adolescents. In the pooled analysis, the prevalence of post-traumatic stress symptoms were not be statistically significant in children and adolescents (pooled prevalence 48% (95%CI: -0.25, 1.21, p=0.200) with a pooled heterogeneity of 100% (p<0.001) (Fig. 5 ). Sub-group analyses by age and gender were not performed due to lack of data.Fig. 5 Meta-analysis of the pooled prevalence of posttraumatic stress symptoms among all children and adolescents (n=2) Abbreviations: PCL-C: The PTSD Cheeklist-CivilianVersion; IES-R: Impact of Events Scale-Revised. Prevalence was calculated based on the random-effect models. Fig. 5 4.3 Results of sensitivity analysis and meta-regression analysis Sensitivity analysis consistently showed that removing individual studies from the meta-analysis did not lead to any change in the prevalence of depression or anxiety. Because only two studies were included in meta-analyses of sleep disorders and posttraumatic stress symptoms, sensitivity analysis was not conducted (Supplemental Table 3). Meta-regression analysis was performed on the prevalence of depression (12 studies) and anxiety (13 studies). Results indicated that neither age nor sample size were significant factors contributing to the heterogeneity of studies (Depression: β=-0.02, 95%: -0.07 to 0.03, p=0.378; Anxiety: β=-0.01, 95%: -0.01, 0.01, p=0.285). However, the questionnaire used for assessment of anxiety or depression did significantly contribute to heterogeneity of studies. In the analysis of depression prevalence, PHQ-9 and CES-D, and in the analysis of anxiety prevalence, SCARED, SAS and PQSPHE contributed to heterogeneity ( Table 4 ).Table 4 Results of meta-regression analyses on the prevalence of depression and anxiety based on 12 studies on depression and 13 studies on anxiety a,b Table 4Type of analysis β 95% CI P Depression (n=12) Age -0.02 -0.07, 0,03 0.378 Sample size Assessment (Reference: CDI (≥19)) 0.01 -0.01, 0.01 0.285 DSRS-C (≥15) -0.06 -0.20, 0.08 0.270 PHQ-9 (≥5) 0.27 0.13, 0.41 0.008 PHQ-9 (≥10) 0.54 0.37, 0.72 0.002 CDI-S (≥7) 0.06 -0.12, 0.23 0.382 CES-D (≥16) 0.22 0.05, 0.39 0.025 PQSPHE (each factor score≥2) -0.14 -0.31, 0.02 0.072 SDS (≥50) 0.21 -0.10, 0.52 0.122 MMHI-60 (each factor score≥2) 0.09 -0.08, 0.26 0.184 Anxiety (n=13) Age -0.01 -0.05, 0.03 0.722 Sample size Assessment (Reference: GAD-7 (≥5)) 0.01 -0.01, 0.01 0.813 SCARED (≥25) -0.15 -0.33, 0.03 0.092 SCARED (≥23) -0.16 -0.32, -0.01 0.045 SAS (standard score≥50) -0.28 -0.51, -0.05 0.026 PQSPHE (each factor score≥2) -0.35 -0.58, -0.12 0.011 GAD-7 (≥8) 0.19 -0.04, 0.43 0.094 MMHI-60 (each factor score≥2) -0.06 -0.29, 0.18 0.581 DSRS-C: Depression Self-Rating Scale for Children; SCARED: Screen for Child Anxiety Related Emotional Disorders; PHQ-9: 9-item Patient Health Questionnaire; GAD-7: 7-item Generalized Anxiety Disorder Scale; SDS: Self-rating Depression Scale; CDI-S: Children's Depression Inventory–Short Form; CES-D: Center for Epidemiologic Studies Depression Scale; SAS: Self-Rating Anxiety Scale; MMHI-60: Mental Health Inventory of Middle-school students; PQSPHE: Psychological Questionnaire for Sudden Public Health Events. a : Because only two articles were included for sleep disorders and posttraumatic stress symptoms, so no meta-regression analyses were conducted. b : Meta-regression analysis was used to evaluate the heterogeneity of different studies, adjusting age, gender, and measurement scale of depression and anxiety. Numbers in bold indicate significance. 4.4 Assessment of publication bias The funnel plot and assessment of Egger's and Begg's tests did not reveal any significant publication bias in the prevalence of depression, anxiety, sleep disorders or post-traumatic stress symptoms (Supplemental Table 2, Supplemental Figure 1). 5 Discussion A recent position paper in The Lancet Psychiatry identified the long-term consequences of COVID-19 for the younger generations are unknown and must be a priority (Holmes et al., 2020). This systematic review and meta-analyses of 23 studies and a total of 57,927 participants provides evidence that 28.6%, 25.5%, 44.2%, and 48.0% of children and adolescents experienced depression, anxiety, sleep disorders, and posttraumatic stress symptoms, respectively, during the COVID-19 pandemic. All the studies included in meta-analysis were from China and conducted among general children and adolescents. The prevalence of depression and anxiety was higher among adolescents and females compared with children and males, respectively. The prevalence of depression and sleep disorders in children and adolescents during the COVID-19 were higher than the respective rates 19.9% for depression (Rao et al., 2019) and 21.6% for sleep disorders (Xiao et al., 2019), reported for the children and adolescents prior to the pandemic in China. However, no data on the prevalence of anxiety and posttraumatic stress symptoms were found among children and adolescents prior to the pandemic in China, thus, no comparisons could be made. Social isolation, school closures, and socioeconomic effects of the policies (increasing unemployment, financial insecurity, and poverty) during the COVID-19 pandemic have been reported to contribute to the mental health problems among children and adolescents (Holmes et al., 2020; Lee, 2020). While there was some research on the psychological impact of severe acute respiratory syndrome (SARS) and middle east respiratory syndrome coronavirus (MERS) on patients and health-care workers, such evidence in children and adolescents is scarce (Lee, 2020). Therefore, no direct comparison of the prevalence of mental health problems with previous pandemics could be made. However, COVID-19 is much more widespread than SARS, MERS, and other previous epidemics. As the pandemic continues, monitoring young people's mental health status over the long term and implementation of interventions and policies to support them are urgent and important. Our study revealed that sleep disorders and posttraumatic stress symptoms were the most severe mental health problems among children and adolescents, and about half of them experienced these disorders during the COVID-19 pandemic. These findings indicate that the COVID-19 pandemic has a substantial impact on young people's sleep. Many children and adolescents may be exposed to unconstrained sleep schedules, prolonged screen exposure, and limited access to outdoor activities and peer interactions and these could have contributed to reported sleep disorders (Liu et al., 2020). Sleep disturbances are often a precursor to other more severe mental problems and it is necessary and urgent to disseminate sleep health education and sleep hygiene behavior interventions to children and adolescents (Lin et al., 2020). The COVID-19 pandemic is a traumatic event, and it is well known that surviving critical illness can induce posttraumatic stress symptoms (Vindegaard and Benros, 2020). The COVID-19 pandemic may be an independent factor that cause posttraumatic stress symptoms in children and adolescents. Children might also be exposed to greater interpersonal violence and abuse, and this too might contribute to the high prevalence. However, the evidence is limited as only two studies reported on posttraumatic stress symptoms in this age group. The subgroup meta-analysis revealed that the prevalence of depression and anxiety may be higher among adolescents and females. The higher prevalence among female reflects the already established gender gap for anxiety and depressive symptoms (Pappa et al., 2020). Again, adolescents exhibited much higher prevalence estimates both for depression and anxiety compared to younger children in our study. This may attribute to education is highly valued and regarded as the main path to success in traditional Chinese culture (Hou et al., 2020). As adolescents face the most important tests of their lives (e.g., the college or high school entrance examination), the uncertainty and potential negative effects on academic development of prolonged school closure had more adverse effects on adolescents than children (Zhou et al., 2020), thus, adolescents had more depressive and anxiety symptoms. However, no subgroup meta-analysis based on age and gender could be conducted for sleep disorders and posttraumatic stress symptoms, because there is no data available. Future such studies should take the potential modifying effects of age and gender into consideration. No subgroup meta-analysis based on preexisting conditions and country could be conducted in this study. Only two studies included were conducted among children and adolescents with preexisting conditions (i.e., Autism Spectrum Disorder and Cystic Fibrosis). However, the two studies were not used for meta-analysis because they did not report the prevalence of mental health problems. All the studies included in our meta-analysis were from China. In our review, only two studies included were conducted in other countries (i.e., Turkey). However, the two studies from Turkey did not report the prevalence of mental health problems, thus they were not included in the meta-analysis. Though the fact that China was severely affected, our findings may provide a reliable indication of the effects of COVID-19 pandemic on the mental health of children and adolescents globally. However, considering the severity of COVID-19, economic status, and healthcare systems vary greatly between countries, more such studies from other countries are warranted. We found that the most frequently used scale to measure depression and anxiety were PHQ-9 (3 of 13) and GAD-7 (5 of 12), respectively. However, each of the two studies used a different scale to measure sleep disorders and posttraumatic stress symptoms. The PHQ-9 is a simple, widely used, and highly effective self-assessment tool for depressive symptoms during the last 2 weeks (Kroenke et al., 2001). GAD-7 measured seven anxiety symptoms that bothered participants during the last 2 weeks (Zhou et al., 2020). Both PHQ-9 and GAD-7 are widely used among children and adolescents. Using the same scale and cutoff point for specific mental health problems could be better for comparison across studies. This study has several limitations. First, all the studies included in meta-analysis were conducted in China, thus, the generalizability of findings to other countries is limited. Moreover, most of the studies used online survey method and nonprobability sampling, which further limit its generalizability. Second, a variety of assessment scales were utilized to measure mental health problems and different cut-offs were used even though several studies used the same tests. Third, due to the limited number of studies, we could not explore the potential modifying effects of preexisting conditions and country on the prevalence of mental health problems. Fourth, only two studies focused on children age <6 years were included; thus, the mental health status of these children warrants further research. To advance research in this area, future studies target children and adolescents are warranted to improve the following aspects. First, longitudinal studies to examine the long-term implications of COVID-19 pandemic on mental health are needed. Second, further studies to examine the prevalence of sleep disorders and posttraumatic stress symptoms are needed. Third, besides the general children and adolescents, more studies are needed to focus on children and adolescents with preexisting conditions, such as chronic diseases and psychiatric conditions. Fourth, studies from countries other than China are needed to provide insight on the global impacts of COVID-19 pandemic on mental health. Despite its limitations, this study is the first to examine the pooled prevalence of depression, anxiety, sleep disorders, and posttraumatic stress symptoms among children and adolescents during the COVID-19 pandemic. We conducted comprehensive literature search based on both English and Chinese databases, the findings have important clinical and public health implications. Furthermore, our subgroup analysis of depression and anxiety based on age and gender provided additional valuable insights of potential particular vulnerabilities. In conclusion, our study highlighted the high prevalence of depression, anxiety, sleep disorders, and posttraumatic stress symptoms among children and adolescents during the COVID-19 pandemic, in particular, among the females and adolescents. Further research is needed to identify strategies for preventing and treating these disorders in this population. Funding The project is supported in part by research grants from the 10.13039/100001547 China Medical Board [Grant number: 16-262], the University Alliance of the Silk Road [Grant number: 2020LMZX002], and Xi'an Jiaotong University Global Health Institute. The funding sources had no role in the design of this study, its execution, analyses, interpretation of the data, or decision to submit results. Authors' contribution The authors's responsibilities were as follows: YFW and WDW designed the research; LM and MM wrote the protocol; KL, SQC and HXZ managed the literature searches and selection; RK and KL performed data extraction; NY performed verification of data extraction, YXL performed meta-analysis; SQC assessed the quality of the included articles; LM wrote the first draft of the manuscript; ML, AR, and MM revised the manuscript; and all authors read and approved the final 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 Supplementary materials Image, application 1 Acknowledgements None. Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jad.2021.06.021. ==== Refs References Allgaier AK Frühe B Pietsch K Saravo B Baethmann M Schulte-Körne G. Is the children’s depression inventory Short version a valid screening tool in pediatric care? A comparison to its full-length version J. Psychosom. Res. 73 5 2012 Nov 369 374 23062811 Blanchard EB Jones-Alexander J Buckley TC Forneris CA. Psychometric properties of the PTSD Checklist (PCL) Behav. Res. Ther. 34 8 1996 Aug 669 673 8870294 Borenstein M Hedges LV Higgins JP Rothstein HR A basic introduction to fixed-effect and random-effects models for meta-analysis Res. Synth. Methods 1 2 2010 97 111 26061376 Chen F Zheng D Liu J Gong Y Guan Z Lou D Depression and anxiety among adolescents during COVID-19: a cross-sectional study Brain Behav. Immun. 88 2020 36 38 32464156 Chen IH Chen CY Pakpour AH Griffiths MD Lin CY Internet-related behaviors and psychological distress among schoolchildren during COVID-19 school suspension J. Am. Acad. Child Adolesc. Psychiatry 2020 Costello EJ Compton SN Keeler G Angold A Relationships between poverty and psychopathology: a natural experiment JAMA 290 15 2003 2023 2029 14559956 Creamer M Bell R Failla S. Psychometric properties of the impact of event scale - revised Behav. Res. Ther. 41 12 2003 Dec 1489 1496 14705607 Duan L Shao X Wang Y Huang Y Miao J Yang X Zhu G An investigation of mental health status of children and adolescents in china during the outbreak of COVID-19 J. Affect. Disord. 275 2020 112 118 32658812 Dunstan DA Scott N. Norms for Zung’s self-rating anxiety scale BMC Psychiatry 20 1 2020 Feb 28 90 32111187 Gagnon C Bélanger L Ivers H Morin CM. Validation of the insomnia severity index in primary care J. Am. Board Fam. Med. 26 6 2013 Nov-Dec 701 710 24204066 Golberstein E Wen H Miller BF Coronavirus disease 2019 (COVID-19) and mental health for children and adolescents JAMA Pediatr. 2020 Holmes EA O'Connor RC Perry VH Tracey I Wessely S Arseneault L Ballard C Christensen H Cohen Silver R Everall I Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science Lancet Psychiatry 7 6 2020 547 560 32304649 Hou TY Mao XF Dong W Cai WP Deng GH Prevalence of and factors associated with mental health problems and suicidality among senior high school students in rural China during the COVID-19 outbreak Asian J. Psychiatry 54 2020 102305 Ivarsson T Lidberg A Gillberg C. The Birleson Depression Self-Rating Scale (DSRS). Clinical evaluation in an adolescent inpatient population J. Affect. Disord. 32 2 1994 Oct 115 125 7829763 Johnson BT Hennessy EA Systematic reviews and meta-analyses in the health sciences: best practice methods for research syntheses Soc. Sci. Med. 233 2019 237 251 31233957 Kroenke K Spitzer RL Williams JB The PHQ-9: validity of a brief depression severity measure J. Gen. Intern. Med. 16 9 2001 606 613 11556941 Lee J Mental health effects of school closures during COVID-19 Lancet Child Adolesc Health 4 6 2020 421 32302537 Li HC Chung OK Ho KY. Center for Epidemiologic Studies Depression Scale for Children: psychometric testing of the Chinese version J. Adv. Nurs. 66 11 2010 Nov 2582 2591 20825514 Li SW Wang Y Yang YY Lei XM Yang YF Analysis of influencing factors of anxiety and emotional disorder in children and adolescents isolated at home during the epidemic of new coronavirus pneumonia Chin. J. Child Health Care 28 04 2020 407 410 Lin LY Wang J Ou-Yang XY Miao Q Chen R Liang FX Zhang YP Tang Q Wang T The immediate impact of the 2019 novel coronavirus (COVID-19) outbreak on subjective sleep status Sleep Med. 2020 Liu X Luo WT Li Y Li CN Hong ZS Chen HL Xiao F Xia JY Psychological status and behavior changes of the public during the COVID-19 epidemic in China Infect. Dis. Poverty 9 1 2020 58 32471513 Liu Z Tang H Jin Q Wang G Yang Z Chen H Yan H Rao W Owens J Sleep of preschoolers during the coronavirus disease 2019 (COVID-19) outbreak J. Sleep Res. 2020 e13142 32716566 Luo M Guo L Yu M Jiang W Wang H The psychological and mental impact of coronavirus disease 2019 (COVID-19) on medical staff and general public - A systematic review and meta-analysis Psychiatry Res. 291 2020 113190 Mo DM Yan JW Li X Liu S Guo PF Hu SW Zhong H Detection rate and influencing factors of anxiety symptoms in children and adolescents under the new crown pneumonia epidemic Sichuan Ment. Health 33 03 2020 202 206 Moher D Liberati A Tetzlaff J Altman DG Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement PLoS Med. 6 7 2009 e1000097 Monga S Birmaher B Chiappetta L Brent D Kaufman J Bridge J Cully M. Screen for Child Anxiety-Related Emotional Disorders (SCARED): convergent and divergent validity Depress. Anxiety 12 2 2000 85 91 Mossman SA Luft MJ Schroeder HK Varney ST Fleck DE Barzman DH Gilman R DelBello MP Strawn JR. The generalized anxiety disorder 7-item scale in adolescents with generalized anxiety disorder: signal detection and validation Ann. Clin. Psychiatry 29 4 2017 Nov 227 234A 29069107 Nunn GD. Concurrent validity between the Nowicki-Strickland locus of control scale and the state-trait anxiety inventory for children Educ. Psychol. Meas. 48 2 1988 435 438 Pappa S Ntella V Giannakas T Giannakoulis VG Papoutsi E Katsaounou P Prevalence of depression, anxiety, and insomnia among healthcare workers during the COVID-19 pandemic: a systematic review and meta-analysis Brain Behav. Immun. 88 2020 901 907 32437915 Pınar Senkalfa B Sismanlar Eyuboglu T Aslan AT Ramaslı Gursoy T Soysal AS Yapar D İlhan MN Effect of the COVID-19 pandemic on anxiety among children with cystic fibrosis and their mothers Pediatr. Pulmonol. 55 8 2020 2128 2134 32530552 Qi H Liu R Chen X Yuan XF Li YQ Huang HH Zheng Y Wang G Prevalence of anxiety and associated factors for Chinese adolescents during the COVID-19 outbreak Psychiatry Clin. Neurosci. 2020 Qi M Zhou SJ Guo ZC Zhang LG Min HJ Li XM Chen JX The effect of social support on mental health in chinese adolescents during the outbreak of COVID-19 J. Adolesc. Health: Off. Public. Soc. Adolesc. Med. 2020 Rao WW Xu DD Cao XL Wen SY Che WI Ng CH Ungvari GS He F Xiang YT Prevalence of depressive symptoms in children and adolescents in China: a meta-analysis of observational studies Psychiatry Res. 272 2019 790 796 30832200 Richardson LP McCauley E Grossman DC McCarty CA Richards J Russo JE Rockhill C Katon W. Evaluation of the patient health questionnaire-9 item for detecting major depression among adolescents Pediatrics 126 6 2010 Dec 1117 1123 21041282 Rogers JP Chesney E Oliver D Pollak TA McGuire P Fusar-Poli P Zandi MS Lewis G David AS 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 32437679 Sevin JA Matson JL Coe DA Fee VE Sevin BM. A comparison and evaluation of three commonly used autism scales J. Autism Dev. Disord. 21 4 1991 Dec 417 432 1778958 Sprang G Silman M Posttraumatic stress disorder in parents and youth after health-related disasters Disaster Med. Public Health Prep. 7 1 2013 105 110 24618142 Study Quality Assessment Tools 2021 https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools. Su LY Wang K Zhu Y Luo XR Yang ZW. Norm of the depression self-rating scale for children in chinese urban children Chin. Ment. Health J. 08 2003 547 549 Tan TX Wang Y Cheah CSL Wang GH. Reliability and construct validity of the Children's Sleep Habits Questionnaire in Chinese kindergartners Sleep Health 4 1 2018 Feb 104 109 29332670 Tan W Hao F McIntyre RS Jiang L Jiang X Zhang L Zhao X Zou Y Hu Y Luo X Is returning to work during the COVID-19 pandemic stressful? A study on immediate mental health status and psychoneuroimmunity prevention measures of Chinese workforce Brain Behav. Immun. 2020 Tang L Ying B Investigation and analysis of middle school students’ mental health status and influencing factors during the new coronary pneumonia epidemic Ment. Health Educ. Primary Secondary Sch. 10 2020 57 61 Tang S Pang HW Anxiety and depression of children and adolescents during the new crown pneumonia epidemic Ment. Health Educ. Primary Secondary Sch. 19 2020 15 18 Türkoğlu S Uçar HN Çetin FH Güler HA Tezcan ME The relationship between chronotype, sleep, and autism symptom severity in children with ASD in COVID-19 home confinement period Chronobiol. Int. 2020 1 7 Vindegaard N Benros ME COVID-19 pandemic and mental health consequences: systematic review of the current evidence Brain Behav. Immun. 2020 Wang JS Li Y He ES. Development and standardization of mental health scale for middle school students in China Psychosoc. Sci. 4 1997 15 20 Wang K Shi HS Geng FL Zou LQ Tan SP Wang Y Neumann DL Shum DH Chan RC. Cross-cultural validation of the depression anxiety stress scale-21 in China Psychol. Assess. 28 5 2016 May e88 e100 26619091 Wang M Meng Q Liu L Liu J. Reliability and validity of the spence children’s anxiety scale for parents in Mainland Chinese children and adolescents Child Psychiatry Hum. Dev. 47 5 2016 Oct 830 839 26667807 Wang NX Xu PF Investigation and research on adolescents’ psychological stress and coping styles during the new coronary pneumonia J. Dali Univ. 5 07 2020 123 128 Wang Y Yang YY Li SW Lei XM Yang YF Investigation of depression among children and adolescents at home during the epidemic of novel coronavirus pneumonia and analysis of influencing factors Chin. J. Child Health 28 03 2020 277 280 Xiao D Guo L Zhao M Zhang S Li W Zhang WH Lu C Effect of sex on the association between nonmedical use of opioids and sleep disturbance among Chinese adolescents: a cross-sectional study Int. J. Environ. Res. Public Health 16 22 2019 Xie X Xue Q Zhou Y Zhu K Liu Q Zhang J Song R Mental health status among children in home confinement during the coronavirus disease 2019 outbreak in Hubei Province, China JAMA Pediatr. 2020 Yu QX Zeng YM Lu WJ Investigation and analysis of the mental health of middle school students during the period of the new crown pneumonia epidemic Jiangsu Educ. 32 2020 44 47 Zhang Y Zhuang LY Yang W Survey of symptoms of post-traumatic stress disorder in middle school students during the COVID-19 pandemic: taking Chengdu Shude Middle School as an example Educ. Sci. Forum 17 2020 45 48 Zhang ZJ. Handbook of Behavioral Medicine Scale[M] 2005 China Medical Electronic Audiovisual Publishing House Beijing 267 270 Zhou J Yuan X Qi H Liu R Li Y Huang H Chen X Wang G Prevalence of depression and its correlative factors among female adolescents in China during the coronavirus disease 2019 outbreak Globalization Health 16 1 2020 69 32723373 Zhou SJ Zhang LG Wang LL Guo ZC Wang JQ Chen JC Liu M Chen X Chen JX Prevalence and socio-demographic correlates of psychological health problems in Chinese adolescents during the outbreak of COVID-19 Eur. Child Adolesc. Psychiatry 29 6 2020 749 758 32363492 Zhu KH Zhou Y Xie XY Wu H Xue Q Liu Q Wan ZH Song RR Anxiety status of primary school students in Hubei Province during the new crown pneumonia epidemic and its influencing factors Chin. Public Health 36 05 2020 673 676	34174475	PMC9711885	NO-CC CODE	2022-12-02 23:21:31	no	J Affect Disord. 2021 Oct 1; 293:78-89	utf-8	J Affect Disord	2,021	10.1016/j.jad.2021.06.021	oa_other
==== Front Biochim Biophys Acta Proteins Proteom Biochim Biophys Acta Proteins Proteom Biochimica et Biophysica Acta. Proteins and Proteomics 1570-9639 1878-1454 Elsevier B.V. S1570-9639(22)00131-5 10.1016/j.bbapap.2022.140884 140884 BBA Research Letter Peptide from NSP7 is able to form amyloid-like fibrils: Artifact or challenge to drug design? Garmay Yuri a Rubel Aleksandr b Egorov Vladimir acd⁎ a Petersburg Nuclear Physics Institute named by B. P. Konstantinov of National Research Center, Kurchatov Institute, 1 mkr. Orlova roshcha, Gatchina 188300, Russia b Saint Petersburg State University, 7/9 Universitetskaya Emb., St Petersburg 199034, Russia c Smorodintsev Research Institute of Influenza, Russian Ministry of Health, 15/17 Ulitsa Prof. Popova, St. Petersburg 197376, Russia d Institute of Experimental Medicine, 12 Ulitsa Akademika Pavlova, St. Petersburg 197376, Russia ⁎ Corresponding author at: Petersburg Nuclear Physics Institute named by B. P. Konstantinov of National Research Center, Kurchatov Institute, 1 mkr. Orlova roshcha, Gatchina 188300, Russia. 1 12 2022 1 2 2023 1 12 2022 1871 2 140884140884 5 10 2022 25 11 2022 29 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. • We found potential amyloidogenic fragment in NSP7 SARS-CoV2 protein in silico • NSP7 (52–62) fragment is able to form amyloid-like fibrils • The possibility of using such a peptide as the basis for an antiviral drug is discussed Keywords Amyloidogenic peptides Antiviral peptides SARS-CoV-2 NSP7 ==== Body pmcOne of the currently frequently discussed methods of antivirals design is the specific induction of changes in the structure and functionality of proteins by a prion-like mechanism [1]. Indeed, viral proteins tend to form amyloid-like fibrils [2], and at the same time, they often have low homology with human proteins. For most proteins, the induction of a conformational transition by peptides is a process that depends on the coincidence of the primary structure of the protein and the peptide acting on it [3]. Together with the ability to amyloid chain reaction, this suggests that such peptides will be effective at low concentrations and, at the same time, will not affect host proteins. A number of peptides are known to have antiviral activity and act according to this mechanism [4,5]. The general strategy for the search for such drugs is the search for an amyloidogenic protein determinant and the design of a peptide that carries this determinant. Such a peptide can specifically affect the parental protein at low concentrations, causing its conformational transition and loss of activity [6]. At the same time, it must be kept in mind that the effect of amyloid-like viral structures on the host has not been studied enough [7] however, for example, in the case of the influenza virus, the formation of amyloid-like fibrils by the PB1-F2 protein during the life cycle of the virus does not have a significant effect on the host organism [8]. The NSP7 protein is a processivity subunit of SARS-CoV-2 RNA-dependent RNA polymerase [9]. In this work, we analyzed the primary structure of NSP7 and proposed sequences of the peptide capable of forming amyloid-like fibrils and having the potential to influence the conformation of this subunit through a prion-like mechanism. We analyzed the primary structure of the NSP7 protein using three programs – Arches, which allows predicting the formation of hairpins characteristic of amyloid-like proteins [10], FoldAmyloid [11], which analyzes the local amino acid composition, and an original program for searching for mirror symmetry motifs [12]. The choice of these programs was due to the fact that in the course of our previous studies, the programs separately made it possible to predict amyloidogenic peptides. Protein sequence of NSP7 (P0DTD1) from the SWISSPROT database were used. Fig. S1 (A) shows the results of the search for potential amyloidogenic regions in NSP7. The peptide corresponding to the NSP7 sequence from 53 to 61 amino acid residues was determined as amyloidogenic by all three programs. Peptides PINP73 (MVSLLSVLLSM, 1191.66 Da) corresponding to NSP7 potentially amyloidogenic region to 52–62 amino acid residues (comprising 2 symmetrical methionines at the N- and C-terminus) and PINP74 (predicted only by FoldAmyloid and Arches, corresponding to 27–37 NSP amino acid residues, KLWAQCVQ, 975.17 Da) was chemically synthesized at OOO NPF VERTA, Russia, purity more than 80%. Peptides were dissolved in 5 μL of DMSO an then phosphate buffered saline (PBS) buffer was added to the 1 mg/mL peptide concentration (0.5% DMSO), then solutions were incubated for 1 h with agitation at orbital shaker (Eppendorf, USA) at 55°С, 600 rpm. Obtained PINP73 or PINP74 peptide samples were diluted 70 times with water, after which 10 μL of solution were applied onto a freshly cleaved mica substrate. After 1 min of incubation, the sample was dried in compressed air. Images (topography of the sample surface) were obtained in the semi-contact mode on an atomic force microscope “NT-MDT” (NT-MDT, Russia), with NSG01 probe. On the mica surface, filaments similar in morphology to amyloid-like fibrils (about 3 nm in height) were observed for PINP73 (Fig. 1 ), not for PINP74 peptide sample (Fig. S2). PINP73 fibrils were observed using atomic force microscopy of three independent suspension preparations, and were not observed when analyzing the PINP74 suspension. Image processing was performed using “Gwyddion” software.Fig. 1 Atomic force microscopy topography image of PINP73 suspension on mica. Fig. 1 In order to determine the amyloid-like nature of the observed filaments, we performed fluorimetry in the presence of Thioflavin T. As a control sample, we used a peptide PINP74 which has a similar to PINP73 molecular weight. Aliquots (30 μL) of peptide solutions were added to 970 μL of 10 mM ThT solution in PBS buffer. Measurements were performed in a HITACHI F-4010 fluorescence spectrophotometer (Hitachi, Japan) with excitation wavelength 440 nm (bandpass 5 nm) and emission wavelength 478 nm (bandpass 5 nm). An increase in Thioflavin T fluorescence was observed in the presence of PINP73, but not PINP74 (Spectra are shown in Fig. S1 (B)) indicates amyloid nature of PINP73 filaments observed by means of atomic force microscopy. Congo Red assay was also performed. Congo Red dye (Sigma Aldrich, USA) at a concentration of 50 μM in PBS was mixed with the 10 μM peptide solution. As a control, a similar sample was used in which a buffer was added instead of the peptide. Absorption spectra were recorded on a BMG Clariostar (BMG, Germany), and difference spectrum was analyzed. Similarly, a comparison of the absorption spectrum of Congo red with PINP74 or PINP73 peptide showed the presence of a peak in the Congo Red with PINP73 spectrum in the region of 550 nm which is characteristic for amyloid-like fibrils (Fig. S1 (C), peak is shown in difference spectrum). There have been previous attempts to predict the amyloidogenic regions of SARS-CoV-2 proteins. Gour et al. [6] analyzed the known SARS-CoV2 proteins for the presence of amyloidogenic regions. In the cited work, theoretical prediction was performed using the FoldAmyloid, Waltz, and AGGRESCAN software. Interestingly, the PINP74 peptide is located in the region predicted by the FoldAmyloid as amyloidogenic, but does not exhibit amyloidogenicity. Peptide PINP73 is also defined by FoldAmyloid as amyloidogenic, but it also falls within the arches region predicted by the Arches software and is a part of a mirror-symmetric motif. It cannot be said with certainty that the use of these three programs (Arches, FoldAmyloid, and symmetry search) can reliably predict amyloidogenic peptides derived from the whole protein, however, in this work, such an approach made it possible to detect an amyloidogenic peptide. The combination of beta-turn forming proneness, symmetry, and context of amino acid residues may be the key to predicting the ability to form amyloid-like fibrils. In the light of recently appearing in databases of fibril structures obtained using cryo-electron microscopy, such a relationship between the features of the primary structure and the ability to homooligomerization is the subject of our further research. It should be noted that the specified fragment of NSP7 is located in the core part of the protein (Fig. S3). This part is not involved in the formation of a dimer, while the fragment that did not show the ability to form fibrils (27–37) is located in the region of the terminal helix involved in dimerisation [13]. It should be noted that the potentially fibrillogenic region of NSP7 is available for interaction both in the monomer and in the dimer. It is possible that the interactions leading to the formation of higher-order oligomers described in [13] are due to interactions between such protein regions. With regard to viruses, in particular SARS-CoV2, the ability of its proteins and protein fragments to amyloidogenesis is unlikely to be an artifact. Interestingly, for the S-protein, the ability of its fragment to form amyloid-like fibrils was confirmed experimentally [14]. Also, the ability of SARS-CoV2 proteins to form amyloid-like fibrils is considered by a number of researchers as one of the possible pathogenicity factors [7,15]. In this work, we demonstrate for the first time the ability of the NSP7 fragment to form amyloid-like fibrils in vitro. Further studies will be aimed at studying the ability of fibrils formed by this peptide to induce coaggregation of the full-length recombinant NSP7 protein and to study the ability of the peptide to act as an antiviral agent in a cell culture model. Author statement VVE – writing the manuscript, spectroscopy methods, sequence analysis; YPG – Atomic force microscopy; AAR – sequence analysis, writing the manuscript. Declaration of Competing Interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Vladimir Egorov reports financial support was provided by 10.13039/501100002261 Russian Foundation for Basic Research . Appendix A Supplementary data Supplementary material Image 1 Data availability Data will be made available on request. Acknowledgements The reported study was funded by 10.13039/501100002261 RFBR , project number 20-04-60491. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bbapap.2022.140884. ==== Refs References 1 Kurpe S.R. Grishin S.Y. Surin A.K. Panfilov A.V. Slizen M.V. Chowdhury S.D. Galzitskaya O.V. Antimicrobial and amyloidogenic activity of peptides. Can antimicrobial peptides be used against sars-cov-2? Int J Mol Sci 21 2020 1 37 10.3390/ijms21249552 2 Tetz G. Tetz V. Prion-like domains in eukaryotic viruses Sci Rep 8 2018 10.1038/s41598-018-27256-w 3 Egorov V. Grudinina N. Vasin A. Lebedev D. Peptide-induced amyloid-like conformational transitions in proteins Int J Pept 2015 2015 10.1155/2015/723186 4 Zabrodskaya Y.A. Lebedev D.V. Egorova M.A. Shaldzhyan A.A. Shvetsov A.V. Kuklin A.I. Vinogradova D.S. Klopov N.V. Matusevich O.V. Cheremnykh T.A. Dattani R. Egorov V.V. The amyloidogenicity of the influenza virus PB1-derived peptide sheds light on its antiviral activity Biophys Chem 234 2018 10.1016/j.bpc.2018.01.001 5 Michiels E. Rousseau F. Schymkowitz J. Mechanisms and therapeutic potential of interactions between human amyloids and viruses, cellular and molecular Life Sci 78 2021 2485 2501 10.1007/s00018-020-03711-8 6 Gour S. Yadav J.K. Aggregation hot spots in the SARS-CoV-2 proteome may constitute potential therapeutic targets for the suppression of the viral replication and multiplication J Proteins Proteom 12 2021 1 13 10.1007/s42485-021-00057-y 33613009 7 Reiken S. Sittenfeld L. Dridi H. Liu Y. Liu X. Marks A.R. Alzheimer’s-like signaling in brains of COVID-19 patients Alzheimer’s Dementia 2022 10.1002/alz.12558 8 Vidic J. Richard C.A. Péchoux C. da Costa B. Bertho N. Mazerat S. Delmas B. Chevalier C. Amyloid assemblies of influenza a virus PB1-F2 protein damage membrane and induce cytotoxicity J Biol Chem 291 2016 739 751 10.1074/jbc.M115.652917 26601953 9 Hillen H.S. Kokic G. Farnung L. Dienemann C. Tegunov D. Cramer P. Structure of replicating SARS-CoV-2 polymerase Nature. 584 2020 154 156 10.1038/s41586-020-2368-8 32438371 10 Kajava A.V. Baxa U. Steven A.C. β arcades: recurring motifs in naturally occurring and disease-related amyloid fibrils FASEB J 24 2010 1311 1319 10.1096/fj.09-145979 20032312 11 Garbuzynskiy S.O. Lobanov M.Y. Galzitskaya O.V. FoldAmyloid: A method of prediction of amyloidogenic regions from protein sequence Bioinformatics. 26 2009 326 332 10.1093/bioinformatics/btp691 20019059 12 Egorov V.V. Garmaj Y.P. Solovyov K.V. Grudinina N.A. Aleinikova T.D. Sirotkin A.K. Kiselev O.I. Shawlovsky M.M. Amyloidogenic peptide homologous to beta-domain region of alpha-lactalbumin Dokl Biochem Biophys 414 2007 152 154 http://www.ncbi.nlm.nih.gov/pubmed/17695325 17695325 13 Wilamowski M. Hammel M. Leite W. Zhang Q. Kim Y. Weiss K.L. Jedrzejczak R. Rosenberg D.J. Fan Y. Wower J. Bierma J.C. Sarker A.H. Tsutakawa S.E. Pingali S.V. O’Neill H.M. Joachimiak A. Hura G.L. Transient and stabilized complexes of Nsp7, Nsp8, and Nsp12 in SARS-CoV-2 replication Biophys J 120 2021 3152 3165 10.1016/j.bpj.2021.06.006 34197805 14 Nyström S. Hammarström P. Amyloidogenesis of SARS-CoV-2 Spike Protein J Am Chem Soc. 25 144 20 2022 May 8945 8950 10.1021/jacs.2c03925 Epub 2022 May 17. PMID: 35579205; PMCID: PMC9136918. 15 Rhodes C.H. Priemer D.S. Karlovich E. Perl D.P. Goldman J.E. Title: β-Amyloid Deposits in Young COVID Patients https://ssrn.com/abstract=4003213 2022	36462605	PMC9711895	NO-CC CODE	2022-12-05 23:15:31	no	Biochim Biophys Acta Proteins Proteom. 2023 Feb 1; 1871(2):140884	utf-8	Biochim Biophys Acta Proteins Proteom	2,022	10.1016/j.bbapap.2022.140884	oa_other
==== Front Comput Methods Programs Biomed Comput Methods Programs Biomed Computer Methods and Programs in Biomedicine 0169-2607 1872-7565 Elsevier B.V. S0169-2607(22)00676-9 10.1016/j.cmpb.2022.107295 107295 Article Potential of vibrational spectroscopy coupled with machine learning as a non-invasive diagnostic method for COVID-19 Zhao Bingqiang Zhai Honglin ⁎ Shao Haiping Bi Kexin Zhu Ling College of Chemistry & Chemical Engineering, Lanzhou University; South Tianshui Road 222, Lanzhou, Gansu 730000, P. R. China ⁎ Corresponding author: Dr. Hong Lin Zhai, Lanzhou University, China, Tel: 86-931-8912596; Fax: 86-931-8912582 1 12 2022 1 12 2022 10729513 9 2022 10 11 2022 29 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. Background and Objective Efforts to alleviate the ongoing coronavirus disease 2019 (COVID-19) crisis showed that rapid, sensitive, and large-scale screening is critical for controlling the current infection and that of ongoing pandemics. Methods Here, we explored the potential of vibrational spectroscopy coupled with machine learning to screen COVID-19 patients in its initial stage. Herein presented is a hybrid classification model called grey wolf optimized support vector machine (GWO-SVM). The proposed model was tested and comprehensively compared with other machine learning models via vibrational spectroscopic fingerprinting including saliva FTIR spectra dataset and serum Raman scattering spectra dataset. Results For the unknown vibrational spectra, the presented GWO-SVM model provided an accuracy, specificity and F1_score value of 0.9825, 0.9714 and 0.9778 for saliva FTIR spectra dataset, respectively, while an overall accuracy, specificity and F1_score value of 0.9085, 0.9552 and 0.9036 for serum Raman scattering spectra dataset, respectively, which showed superiority than those of state-of-the-art models, thereby suggesting the suitability of the GWO-SVM model to be adopted in a clinical setting for initial screening of COVID-19 patients. Conclusions Prospectively, the presented vibrational spectroscopy based GWO-SVM model can facilitate in screening of COVID-19 patients and alleviate the medical service burden. Therefore, herein proof-of-concept results showed the chance of vibrational spectroscopy coupled with GWO-SVM model to help COVID-19 diagnosis and have the potential be further used for early screening of other infectious diseases. Graphical Abstract Image, graphical abstract Keywords Vibrational spectroscopy Fourier transform infrared Raman scattering Tchebichef curve moments Grey wolf optimized support vector machine Abbreviations COVID-19, coronavirus disease 2019 SI, swarm intelligence GWO-SVM, grey wolf optimized support vector machine RT-PCR, real-time reverse transcription polymerase chain reaction CT, computed tomography POC, point-of-care FTIR, Fourier transform infrared AI, artificial intelligence ML, machine learning LDA, linear discriminative analysis k-NN, k-nearest neighbors RF, random forest NB, Naïve Bayes SVM, support vector machines RBF, radical basis function GWO, grey wolf optimization SG, Savitzky-Golay MSC, multiplicative scatter correction airPLS, adaptive iteratively reweighted penalized least squares TCM, Tchebichef curve moments FOM, figures of merit AUROC, area under the receiver operating characteristics curve AUPRC, area under the precision-recall curve TP, true positives TN, true negatives FN, false negatives FP, false positives TPR, true positive rate FPR, false positive rate PRC, precision-recall curve FNR, false negative rate ACE2, angiotensin-converting enzyme 2 ==== Body pmc1 Introduction The ongoing coronavirus disease 2019 (COVID-19) pandemic has led to over 632 million confirmed infections and almost 6.6 million deaths up to now (https://coronavirus.jhu.edu/map.html) with multiple emerging severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants have recently been identified[1]. Significant obstacles for the diagnosis of SARS-CoV-2 infection continue to hamper large-scale population-based screening to control the COVID-19 pandemic in the absence of widely available antiviral therapeutics[2]. Nasopharyngeal swabs nucleic acid amplification test, namely real-time reverse transcription polymerase chain reaction (RT-PCR)[3], is the gold standard for SARS-CoV-2 detection, plays a critical role in detecting infection, determining infection rate, characterizing the disease progression, and guiding clinical decision-making. However, some technical shortcomings such as long processing time, laborious, involving specialized instruments and skilled personal encourage the further efforts to develop more reliable diagnosis strategy. Hence, some other testing methods including chest computed tomography (CT)[4] , [5], chest X-ray images[6], [7], [8], enzyme-linked immunosorbent assays[9], chemiluminescence immunoassays[10], lateral flow immunoassays[11], electrochemical biosensor[12], and mass spectrometry[13], [14], [15], were presented as an alternative method for SARS-CoV-2 detection. Though these assays have been effective in reducing transmission rate, it is imperative to develop individual-friendly biofluids such as saliva or serum for COVID-19 patients screening and diagnosis. Vibrational spectroscopy[16], based on Fourier transform infrared (FTIR) spectroscopy[17] , [18] or Raman scattering spectroscopy[19], [20], [21] provides a detailed fingerprinting of a biological sample from its chemical composition. Diagnostic tools based on these technologies reveal the potential to revolutionize clinical systems leading to evolutionary patient outcome, more efficient public services and significant economic savings. Moreover, vibrational spectroscopy possesses label-free disease detection and diagnosis in a single step[22] , [23]. However, obtaining clinical diagnosis speeds and accuracies remains challenging due to weak infrared absorption or Raman scattering signals from samples. Thus, developing rapid, sensitive, and large-scale screening methods could be advantageous to prevent its spread and mitigate the pandemic. Fortunately, artificial intelligence (AI) has been deployed at various levels of the healthcare systems, including diagnosis, public health, clinical decision-making, and therapeutics. Fong et al. discussed how AI can help fight this deadly virus, from early warnings, prompt emergency responses, and critical decision-making to surveillance drones[24]. Meanwhile, many machine learning (ML) methods have been used for classification or disease diagnosis to obtain meaningful information from vibrational spectroscopy[25] , [26]. Linear discriminative analysis (LDA)[27], k-nearest neighbors (k-NN)[28] , [29], random forest (RF)[30], Naïve Bayes (NB)[31], and support vector machines (SVM)[32] , [33] are the widely used ML algorithms for classification in disease diagnosis purpose. The presence of ML could be said to accelerate the speed for the development of fast, accurate and sensitive methods which allowing for the detection and diagnosis of infectious diseases[34], [35], [36]. Interestingly, SVM can deal with high dimensional data comprising multiple features from vibrational spectroscopy, which has been verified to have outstanding performance in many fields and is often considered the best classifiers[37] , [38]. The flexible characteristic of employing different kernel functions (e.g., linear, polynomial, RBF and sigmoid) to demarcate the hyperplane boundary enables to discriminate high dimensional data. Particularly, SVM based on the radical basis function (RBF) kernel function is widely used for pattern recognition[39]. The penalty coefficient C and the kernel coefficient γ are two significant parameters of the SVM with RBF kernel function. Recently, many swarm intelligence (SI) optimization algorithms derived from meta-heuristics have been proposed in recent years[40] , [41]. Grey wolf optimization (GWO) is one such algorithm proposed by Seyedali Mirjalili et al.[42] , [43], which is a population-based optimization algorithm inspired by the leadership hierarchy and hunting mechanism of grey wolf. A detailed mathematical description of GWO optimizer was included in the Supplementary Material section 4. Here, we reported a vibrational spectroscopy fingerprinting-ML model for the discrimination of saliva and serum specimen, which could remarkably discriminate the signal of positive patients infected by SARS-CoV-2 from healthy individuals or suspected with symptoms like COVID-19. Specifically, we developed a flexible discriminatory tool by combining GWO optimizer and SVM classifier (GWO-SVM) for the diagnostic of COVID-19 patients based on their vibrational spectroscopy. To obtain a model with better generalization property, we selected the best hyperparameters (C and γ) combination by GWO to avoid overfitting or local minima problems. The fitness function used was classification accuracy, which is one of the most popular metrics in classification models and it was directly computed from the confusion matrix. The proposed architecture consists of two steps. Firstly, GWO optimizer was employed to adjust the hyperparameter of SVM classifier to predict the vibrational spectroscopy for COVID-19 diagnosis. Secondly, GWO-SVM model was carried out to discriminate unknown vibrational spectroscopy of saliva and serum specimens and calculate the classification accuracy based on the optimal hyperparameter combinations obtained from GWO optimizer. Simultaneously, five other ML methods including LDA, kNN, NB, RF and SVM, were tested for their classification performances as the references. Within this context, present study showed that vibrational spectroscopy in coupled with GWO-SVM model for analysis of saliva or serum biofluids from suspected patients can become a novel rapid, cost-effective diagnostic tool for COVID-19 and have the potential be further used for early screening of other infectious diseases. The remainder of the paper is organized as follows. In section 2, we reviewed the current state-of-the-art literature related to the COVID-19 detection and vibrational spectroscopy applications in biomedical engineering. In section 3, we described the proposed method and experimental analysis. In sections 4, we presented the performance metrics and experimental results. Section 5 contains a discussion and outlook on the provided issue. Finally, a brief summary was described in last section. 2 Related works In this section, we briefly reviewed the current state-of-the-art literature about the application of various techniques to COVID-19 detection and vibrational spectroscopy for biomedical scenarios. 2.1 Diagnosis methods of COVID-19 The ongoing COVID-19 pandemic caused by SARS-CoV-2 infection has led to severe economic burdens worldwide. Multiple emerging SARS-CoV-2 variants have been identified and are now spreading internationally. Efficient outbreak control will then need cost-effective and easy-to-operate detection tools that can be easily deployed in low-resource situations[44]. Currently, three approaches, including RT-PCR, serological/immunological antigen-based test and chest CT are generally used for COVID-19 diagnosis. For instance, Ketan et al. reported a rapid RNA extraction-free lateral flow assay for molecular point-of-care (POC) detection of SARS-CoV-2 augmented by chemical probes. The assay uses highly specific 6-carboxyfluorescein and biotin labeled antisense oligonucleotides as probes designed to target N-gene sequence of SARS-CoV-2. Besides, they utilized cysteamine capped gold-nanoparticles to augment the signal[45]. Zhang et al. established an integrated system, which incorporates a ML-based FTIR for rapid COVID-19 screening and air-plasma-based disinfection modules to prevent potential secondary infectious. A partial least squares discrimination analysis and a convolutional neural network model were built using the collected infrared spectral dataset of serum samples. The sensitivity, specificity and accuracy all reach over 94% from the blind test samples[46]. Zhang et al. reported a rapid and sensitive magneto fluidic immune-PCR platform that can address the current gap in POC serological testing for COVID-19. They evaluated this magneto fluidic immune-PCR platform with 108 clinical serum samples and achieved 93.8% sensitivity and 98.3 specificity, demonstrating its potential as a rapid and sensitive POC serological test for COVID-19[47]. However, the utility of most of these techniques are limited. While serological tests suffer from the cross-reactivity with other pathogens, such as other human coronaviruses, immunological ones is limited by a detectable antibody response at the early stages of infection. Of the above, chest CT is a key screening tool for patients with COVID-19 symptoms. Mei et al. used artificial intelligence algorithms to integrate chest CT findings with clinical symptoms, exposure history and laboratory testing to rapidly diagnose patients who are positive for COVID-19. In a test set of 279 patients, the AI systems achieved an area under the curve of 0.92 and had equal sensitivity as compared to a senior thoracic radiologist[48]. Despite be widely available in cities, CT facilities typically do not reliably detect COVID-19 infection in its early stage, making them unsuitable for intensive patient surveillance. 2.2 Vibrational spectroscopy for biomedical applications The potential of vibrational spectroscopy for biomedical applications has been well established through many proofs of concept studies over the past decades[49] , [50]. Due to its unique fingerprinting capability, vibrational spectroscopy can play significant role in disease diagnosis and drug discovery. However, these vibrational spectroscopies are often possessed multivariate signatures that allow one to differentiate between patients with disease and healthy controls. Marcia et al. developed a noninvasive diagnostic for COVID-19 from saliva biofluid via FTIR spectroscopy and multivariate analysis. They evaluated a mid-infrared dataset of saliva samples obtained from symptomatic patients using unsupervised and supervised strategy. This method presents an important tool for a fast, noninvasive diagnostic technique, reducing costs and allowing for infection risk reduction[51]. Du et al. proposed a method based on Raman spectroscopy combined with generative adversarial network and multiclass SVM to classify foodborne pathogenic bacteria. Better classification results are obtained by optimizing the parameters of the multiclass SVM[52]. Shu et al. developed a deep learning guided fiberoptic Raman diagnostic platform to assess its ability of real-time in vivo nasopharyngeal carcinoma diagnosis and post-treatment follow-up patients. The robust Raman diagnostic platform was established using multi-layer Raman-specified convolutional neural networks together with simultaneous fingerprint and high-wavenumber spectra acquired within sub-seconds. The optimized model provides an overall diagnostic accuracy of 82.09% for identifying nasopharyngeal carcinoma from control and post-treatment patients[22]. Based on the evidence provided herein, vibrational spectroscopy in combination with ML provides the first glimmer of hope for the development of an accurate, inexpensive, rapid, and non-invasive method for universal biomedical applications. 3 Methods 3.1 Data preparation The first dataset consists of total 171 saliva FTIR spectra that 87 spectra for positive with SARS-CoV-2 and 84 spectra for negative with SARS-CoV-2, derived from 29 patients testing positive and 28 individuals testing negative with COVID-19 like symptoms receiving treatment in the Royal Melbourne Hospital by RT-PCR[53]. According to previously described, saliva was collected by asking the patient to cough out saliva from their throat into a sterile container, and viral transport medium was added. 10 mL saliva was deposited onto an infrared reflective substrate with three deposits per slide from a patient cohort and FTIR spectra (4000-800 cm−1) were recorded on a Perkin-Elmer Spectrum 2 spectrometer with a dedicated reflection accessory. Totally, 183 spectra were acquired from 61patients and suspected for anomalies. However, spectra belonging to four patients (2 negative and 2 positive) were identified as outliers, due to a high content of viral transport medium, or the spectra had no contribution on the 2058 cm−1 band, typical of a saliva spectrum. The second dataset comprised of total 465 serum Raman scattering spectra (1800-600 cm−1) from the serum of 53 patients who were confirmed with COVID-19 by RT-PCR, 50 healthy individuals and 54 individuals with flu symptoms like COVID-19 patients. According to the Gang Yin et al.34, the Raman spectra for each serum sample was implemented by three experimenters and repeated five times, respectively. Finally, the five spectra collected from each serum sample by three experimenters were respectively averaged. Among the total serum samples in the suspected group, only two Raman scattering spectra were obtained in subjects 16 to 21. The details of the sample preparation, instrumentation, and vibrational spectra analysis can be referred to previously work by Gang Yin et al[54]. The original saliva FTIR spectra and serum Raman scattering spectra were illustrated in Fig. 1 . The total datasets were randomly divided into a training phase (2/3 of the dataset) and a test phase (1/3 of the dataset) using the Kennard-Stone sample selection algorithm[55] to develop our model.Fig. 1 (A) Raw saliva FTIR spectra dataset. (B) Raw serum Raman scattering spectra dataset. Fig 1 3.2 Spectral preprocessing Data preprocessing is essential in vibrational spectra, as the FTIR and Raman scattering spectra are often affected by multiple noise sources, such as instrument, environmental condition, nature of the sample and other factors[56]. For the saliva FTIR spectra dataset, the spectra were firstly truncated in the bio-fingerprint region (1300-800 cm−1), this is where many important RNA and glycoprotein marker bands are located and there is less interference from viral transport media, thus increase the robustness of the modeling, because the more variables fed into the model, such as the variables account for viral transport media in the sample, the more chance of finding spurious correlations. Secondly, following preprocessing procedure were employed: Savitzky-Golay (SG) smoothing[57] to remove unnecessary noise from the original spectra (window = 9 points, 2nd order polynomial function), multiplicative scatter correction (MSC)[58] to adjust the light scattering and adaptive iteratively reweighted penalized least squares (airPLS)[59] to remove baseline absorptions. The total and mean spectra of two groups after preprocessing were referred to Fig. S2-4. For the serum Raman spectra dataset, an outlier detection[60] was implemented to exclude those samples away from others (Fig. S5). In addition, the original spectra were truncated in the bio-fingerprint region (1800-600 cm−1), SG smoothing (window = 15 points, 2nd order polynomial function), MSC and airPLS were also implemented for preprocessing. The total and mean spectra of three groups after preprocessing were referred to Fig. S6-10. It can be noted that the difference between the two vibrational spectra datasets is very tiny. Particularly, closely related spectra show almost identical mean FTIR or Raman scattering spectra. Therefore, vibrational spectroscopy needed to be analyzed using advanced statistical methods. 3.3 Feature extraction In most cases, variables from spectroscopy, e.g., Raman scattering spectroscopy and FTIR spectroscopy, are proportional to noise signals, which usually lead to a collinearity problem. Supervised models that often minimize bias (i.e., error in the estimates) on training data tend to be overfitting when the number of object or spectra sample is less than the number of variables, so the model generalizability would be poor on test data. Feature extraction is an alternative strategy to deal with the highly correlated variables before modeling. Here the Tchebichef curve moments (TCM)[61] was employed for feature extraction. As one of the discrete orthogonal moments, TCM exhibit the characteristic of powerful curve description capability, multi-resolution property, invariance property and thus can be used to capture important features of an spectra curve[62] , [63]. A brief note on TCM instructions and their mathematical description can be referred in Supplementary Material 3.4 Discriminant analysis using ML algorithms The proposed GWO-SVM model consists of four main procedures (Scheme 1 ). Firstly, a herd of grey wolves were randomly created. There were two hyperparameters (C and γ) to be optimized in the SVM classifier, thus, each grey wolf population was termed as a two-dimensional array. Secondly, the fitness function of the GWO-SVM was determined based on the cross-validation accuracy. Thirdly, the fitness of each population would be obtained after the initial population was established. These fitness values were ranked to find the three individuals with the highest fitness values and marked as the three grey wolves with the highest hunting ability, which were called as α, β and δ wolves, and then they were used to guide the position updating for other wolves. Hereby, a new population with the updated grey wolf positions were formed and the individual fitness was obtained and evaluated. The above process was repeated consistently until the maximum number of iterations accomplished. Finally, when the iteration process completed, the ideal solution would be fed into SVM classifier to make it the finest classifier. And the classification performance of the GWO-SVM model was measured by the test phase divided from the original data set.Scheme 1 Illustration of the whole procedure of grey wolves optimized support vector machine (GWO-SVM). Scheme 1 To fully gauge the discrimination capabilities of GWO-SVM model for saliva FTIR spectra and serum Raman scattering spectra, several other supervised ML algorithms were considered in this work, including LDA, kNN, NB, RF and SVM. These were introduced succinctly in Supplementary Material with their main features. 3.5 Figures of merit Cross-validation was employed to determine the validity of a model on test phase by evaluating if the model is overfitted to noise. Considering the small number of samples, hyperparameters of models were finetuned by a leave-one-spectra-out cross-validation (LOSOCV) and the determined one was chosen according to the global optimum accuracy. In this regard, the original calibration dataset was randomly divided into two phase, C and C\i, where C was the training phase and C\i was the cross-validation phase. Firstly, the training phase was used to train the model, then the cross-validation phase was used to verify the accuracy of the classification model. When the final models were achieved using different ML algorithms, the performance of each one would be back-evaluated according to the prediction accuracy for the training phase. Finally, the optimal model was validated by independent samples in the test phase with prediction accuracy. The model validation of the ML models was performed based on seven figures of merit (FOM)[64] , [65], namely accuracy, precision, recall, F1-score, specificity, area under the receiver operating characteristics curve (AUROC) and area under the precision-recall curve (AUPRC). Accuracy indicates the proportion of correctly predicted events using the optimized classification model. Precision and recall are the two basic FOM for classification model, while F1-score is their complementary parameter. Precision is a FOM expressing the proportion of correctly classified positive samples among all samples classified as positive. Thus, this is a more robust FOM to evaluate the detection of positive samples. Recall is the proportion of true positives (TP) among all positives, while specificity is the rate of true negative (TN) predictions. F1-score is the weighted average value between precision and recall, which considers both false negatives (FN) and false positives (FP) into account, thus, which is particularly useful for unbalanced dataset. For the multi-class classification, the same previous parameters were calculated, but in macro-averaging level as mentioned in Marina Sokolova et al.[66]. The equations used to calculate these performance metrics were as follows:Accuracy=TP+TNTP+TN+FP+FNprecision=TPTP+FPRecall=TpTP+FNF1_score=2×precision×Recallprecision+RecallSpecificity=TNTN+FP The receiver operating characteristics (ROC) analysis[67] is based on statistical decision theory and has been applied extensively to the evaluation of classification methods. The ROC curve can manifest the relationship between the true positive rate (TPR) and false positive rate (FPR) with the variations of decision threshold[68]. Specifically, AUROC is one of the most widely used metrics for overall discrimination ability of a classification model. It ranges between 0.5 and 1.0. A model with an AUROC of 1.0 suggests perfect separation ability, while an AUROC of 0.5 suggests there is no class separation. Similar to the ROC curve, the precision-recall curve (PRC) is also a useful metrics for unbalanced dataset, which shows the tradeoff between precision and recall during different threshold. A higher AUPRC indicates both higher precision and higher recall, where higher precision relates to a slow false positive rate (FPR), and higher recall relates to a low false negative rate (FNR). 3.6 Implementation TCM for feature extraction routines were a custom-written program in m-file. LIBSVM toolbox was downloaded from https://www.csie.ntu.edu.tw/∼cjlin/libsvm/. All algorithms of spectral preprocessing, feature extraction, optimal wavelength selection, model calibration, and visualization were performed in MATLAB 7.0 (The MathWorks, Inc., Natick, USA) via lab-made routines on a desktop with Intel(R) Core (TM) i7-4770K CPU @ 3.50GHz and 64 GB RAM, with Windows 10 operating system (professional version). 4 Results According to the results, the classification performance of the SVM that its hyperparameter combination was optimized by grid search (GS-SVM) was not desired during the test phase, as the accuracy was lower than 0.9. For saliva FTIR spectra dataset, in the case of GS-SVM, the test accuracy was 0.8947 and the AUROC was 0.9662. Moreover, in the Wood's work, the model recall and specificity were 0.93 and 0.82, respectively[53]. However, the accuracy obtained by GWO-SVM during the test phase can reach over 0.98, which was almost close to 1.0, while other ML algorithms cannot achieve such values even after many tests, suggesting the GWO-SVM model exhibited superiority to classify saliva FTIR spectra dataset (Fig. 2 ). In addition, it can be noted that the GWO-SVM got better results than other ML models. In particularly, the recall and AUROC were equals to 1.0, suggesting that a randomly positive sample spectra (i.e., saliva FTIR spectra from patient with COVID-19) will be predicted more likely to be COVID-19 than a randomly negative sample (i.e., saliva FTIR spectra from patient with non-COVID-19 with probability 1.0 (Fig. 3 ).Fig. 2 Summary of the accuracy, precision, recall, F1-score, specificity, AUROC and AUPRC values obtained from six machine learning methods for unknown saliva FTIR spectra. Fig 2 Fig. 3 (A) ROC curves of six machine learning methods for unknown saliva FTIR spectra. (B) PRC curves of six machine learning methods for unknown saliva FTIR spectra. (LDA: Linear discriminative analysis. k-NN: k-nearest neighbors. RF: Random Forest. NB: Naïve Bayes. GS-SVM: Grid search optimized support vector machines. GWO-SVM: Grey wolf optimized support vector machine). Fig 3 In addition, the overall accuracy of GS-SVM classifier for serum Raman scattering spectra dataset was only 0.7843. Moreover, the results presented significant variation when the vibrational spectra in the two datasets were classified several times. However, GWO-SVM classifier significantly outperformed other ML models to classify the serum Raman spectra dataset with an overall accuracy of 0.9085 and AUROC of 0.9995 in the test phase (Fig. 4 ). Moreover, Fig. 3 A and Fig. 5 A compared the ROC curves of the GWO-SVM and other ML models with the optimal hyperparameters. In the two vibrational spectra datasets, the highest AUROC both belonged to the GWO-SVM. As mentioned above, the higher the AUROC (the perfect value is 1.0) is, the better classification performance of the model will be. It was suggested that GWO-SVM possess a well-behaved classification capability for positive and negative samples based on saliva FTIR spectra dataset. Furthermore, the PRC of the classification model was considered. As presented in the Fig. 3 B and Fig. 5 B, the AUPRC for both vibrational spectra datasets were superior to GS-SVM as well as other machine learning models. In addition to the modeling procedure, the GWO-SVM converges faster than the GS-SVM. Considering that the GWO-SVM had better classification performance than the GS-SVM in test phase for both vibrational spectra datasets, it deduced that a faster convergence rate can be achieved by GWO-SVM model based on these two vibrational spectroscopy datasets while ensuring the classification performance.Fig. 4 Summary of the macroAccuracy, macroPrecision, macroRecall, macroF1-score, macroSpecificity, AUROC and AUPRC values obtained from six machine learning methods for unknown serum Raman scattering spectra. Fig 4 Fig. 5 (A) ROC curves of six machine learning methods for unknown serum Raman scattering spectra. (B) PRC curves of six machine learning methods for unknown serum Raman scattering spectra. (LDA: Linear discriminative analysis. k-NN: k-nearest neighbors. RF: Random Forest. NB: Naïve Bayes. GS-SVM: Grid search optimized support vector machines. GWO-SVM: Grey wolf optimized support vector machine). Fig 5 5 Discussion Clinical decisions are often complex and include continuous against trade-offs between numerous and frequently clashing targets. The rapid ongoing spread of COVID-19 over the world results to force the healthcare service systems. Generally, infrared spectroscopy is a routine analytical technique for molecular functional groups identification in organic chemistry and material chemistry. When infrared light meets vibrational modes of molecules, a unique fingerprinting of the sample will be generated. Saliva is emerging as an attractive medium for POC diagnosis of COVID-19 in the current pandemic. The SARS-CoV-2 virus has a preferential tropism to human airway epithelial cells that express the cellular receptor angiotensin-converting enzyme 2 (ACE2)[69]. Besides, ACE2 was found to be higher in salivary glands compared to the lungs, suggesting that salivary glands could be a potential target for SARS-CoV-2 virus[70]. In addition, serum testing as a routine testing item in clinical situations, can provide low-cost and rapid screening of patients in hospitals. In clinical applications, Raman scattering testing can be performed in routine serum testing items for COVID-19 screening. Once high-risk patients are found, they are immediately quarantined and further confirmed with RT-PCR technique, thus reducing the risk of infection in medical institutions. In this paper, a hybrid classification model called GWO-SVM was proposed to explore the potential of vibrational spectroscopy coupled with ML to screen COVID-19 patients in its initial stage. The GWO-SVM model was tested and comprehensively compared with other ML models via vibrational spectroscopic fingerprinting including saliva FTIR spectra dataset and serum Raman scattering spectra dataset. 5.1 Feature extraction The maximum order of TCM not only affect the extracted information from the original vibrational spectroscopy but also contribute to the total number of variables fed into ML models. For saliva FTIR spectra dataset, the reconstruction error decreases as the maximum order of TCM increases, and reconstruction error variation was gradually tended to be stable when mM = 61 (Fig. S11). The reconstructed FTIR spectra under different order mM were referred to Fig. S12. Hence, the maximum order of TCM for saliva FTIR spectra dataset was determined and as a result a total of 62 TCMs were obtained. The calculation of the maximum order of TCM for serum Raman scattering spectra dataset was like that of saliva FTIR spectra dataset (Fig. S13). The reconstructed Raman scattering spectra under different order mM were referred to Fig. S14. Further, stepwise regression was used to select valid independent moments variables and develop classification models for the discrimination of vibrational spectroscopy dataset. 5.2 GWO parameter GWO as a novel SI optimization algorithm, which has good performance in global search and convergence. The core idea of GWO algorithm is to simulate various behavior of grey wolves, including the hierarchy and hunting process within the wolf population, to find the optimal solution of the target problem. The grey wolf population in nature is divided into four grades, namely α, β, δ and ω, in order of social status from high to low. Define the current optimum solution in the wolf population as α wolf, the second-best solution as β wolf, the third-best solution as δ wolf, and other solutions as ω wolf to construct the hierarchy model of the grey wolf. The hunting behavior of the whole population consists of three steps: (1) Tracking the prey; (2) Encircling the prey; (3) Attacking the prey. In the GWO algorithm, the hunting task is performed by α, β and δ wolf. ω wolf follows the three wolves to carry out the prey tracking, encirclement, and suppression. It is worth mentioning that the GWO algorithm has very few parameters to be finetuned compared to other meta-heuristics methods according to its mechanism detailed in the Supplementary Materials section 4. The adaptive values of GWO parameters allow a smooth transition between exploration and exploitation. This provides GWO a greater ability to avoid stagnation in local optimal solution and converge quickly. Therefore, the motivation of this study is to explore the feasibility of the hybrid classification model of GWO-SVM in predicting vibrational spectroscopy. Moreover, the original GWO algorithm is applicable for continuous single objective optimization problems. However, in this study, two SVM hyperparameter (penalty factor C and RBF kernel parameter γ) combination selection is inherently multi-objective. Herein, we proposed an improved strategy namely multi-objective binary GWO algorithm. Consequently, it has been modified and represented in a way that is suitable for parameter selection task. The proposed strategy can avoid stagnation in local optima by maintaining a balance between exploration and exploitation. Based on the above, GWO algorithm will seek for the hyperparameter combinations in a more reasonable space under the restriction of the upper and lower bounds. Meanwhile, the optimal solution of GWO was affected by the maximum iteration numbers. Here, the lower bound was set as 1.0 × 10−4 for C and 1.0 × 10−2 for γ, while the upper bound was set as 1.0 × 104 for C and 1.0 × 102 for γ. The maximum number of iterations was 30. 5.3 Model performance In this section, the GWO-SVM model performance was compared with other ML algorithms. Simultaneously, the original model performance results from relevant literatures for these two vibrational spectra datasets were used as a reference to evaluate the performance for current study. It should be noticed that a one versus rest strategy[71] was implemented for serum Raman scattering spectra dataset to split the three-class classification task into three binary classification problem. Thus, for COVID-19 diagnosis, decision-making can be obtained by COVID-19 individuals versus healthy and suspected individuals. When evaluating the classification performance, confusion matrix can be introduced to judge whether the predicted labels of the model are consistent with the real one. The higher value in the upper left (TP) and lower right (TN) in the confusion matrix, the higher the consistency between the predicted results and the real labels. According to the optimal hyperparameters of several ML models, the confusion matrix of the six classification models for independent test phase from two vibrational spectra datasets can be obtained (Fig. S15-16). It can be noted that the TP and TN values of GWO-SVM were both higher than other ML models, especially for GS-SVM, suggesting that the predicted labels of the GWO-SVM for test phase were almost consistent with the real labels. Specifically, for the unknown vibrational spectra, the presented GWO-SVM model provided an accuracy, specificity and F1_score value of 0.9825, 0.9714 and 0.9778 for saliva FTIR spectra dataset, respectively, while an overall accuracy, specificity and F1_score value of 0.9085, 0.9552 and 0.9036 for serum Raman scattering spectra dataset, respectively, which showed superiority than those of state-of-the-art models, thereby suggesting the suitability of the GWO-SVM model to be adopted in a clinical setting for initial screening of COVID-19 patients. In summation, the focus of this work is to explore the feasibility and reliability of the newly proposed GWO-SVM model in classifying vibrational spectroscopy of COVID-19 patients with other ML models as a reference. According to the obtained results, the discussed vibrational spectroscopy in combination with GWO-SVM model strategy can be applied for COVID-19 detection to improve the accuracy of provisional and clinical diagnosis. The proposed technique demonstrates better or comparable results with respect to those other ML techniques. Further studies may include larger cohort or large-scale multicenter trials to prove its applicability in clinical settings and to demonstrate joint applicability of vibrational spectroscopy and ML models for biomedical scenarios. 6 Conclusion The identification of COVID-19 patients in its early stages, where treatment could provide maximum therapeutic benefits, is not only likely to slow down disease progression but also to potentially provide a cure. In the current study, we have successfully developed a vibrational spectroscopy-based approach for COVID-19 diagnosis and introduced a new ML algorithm, namely GWO-SVM where GWO was used to finetune the hyperparameters of SVM. Simultaneously, two vibrational spectra datasets were selected to train and evaluate the performance of GWO-SVM and other ML models. Through the analysis of figures of merits, it can be noted that the GWO-SVM exhibited superiority to classify vibrational spectra dataset than others ML models. In addition, the proposed GWO-SVM model was applicable to both binary and multi-classification problems. As a result, the reported saliva FTIR spectra and serum Raman scattering spectra examination have the potential to complement clinical nucleic acid testing, make early COVID-19 detection quickly, accurate, and inexpensive. The results indicate that vibrational spectroscopy coupled with GWO-SVM can be employed as adjuvant or alternative approach in the clinical diagnosis of COVID-19 patients. While this study showed promise using a small sample set, further method validation on a large scale is required to indicate the true strength of the proposed strategy. 7 Data and Code availability All study data were included in the article or Supplementary Material. Custom-built codes may be available by contacting the corresponding author on reasonable request. 8 Ethical approval statement Not required for this study because no human or animals directly participated in this study. 9 CRediT authorship contribution statement Bingqiang Zhao and Hongling Zhai conceived the study, devised the algorithms, developed the model, and performed preliminary evaluations. Bingqiang Zhao implemented and extended the algorithms and evaluation metrics, performed large scale experiments, analyzed the results and wrote the manuscript. Hongling Zhai was involved in review, editing and supervision. All authors participated in drafting the article and gave approval to the final version of the manuscript. Declaration of Competing Interest The authors declare that they have no known competing financial interest or personal relationships that could have appeared to influence the work reported in this paper. Appendix Supplementary materials Image, application 1 Acknowledgements This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cmpb.2022.107295. ==== Refs References 1 Liao Z. Song Y. Ren S. VOC-DL: Deep learning prediction model for COVID-19 based on VOC virus variants Comput Methods Programs Biomed 224 2022 106981 10.1016/j.cmpb.2022.106981 2 Shinde G.R. Kalamkar A.B. Mahalle P.N. Data Analytics for Pandemics: A COVID-19 Case Study 2020 CRC Press 10.1201/9781003095415 3 Shen M. Zhou Y. Ye J. Recent advances and perspectives of nucleic acid detection for coronavirus Journal of Pharmaceutical Analysis 10 2020 97 101 10.1016/j.jpha.2020.02.010 32292623 4 Hassan H. Ren Z. Zhou C. Supervised and weakly supervised deep learning models for COVID-19 CT diagnosis: A systematic review Comput Methods Programs Biomed 218 2022 106731 10.1016/j.cmpb.2022.106731 5 Chen Y. Lin Y. Xu X. Classification of lungs infected COVID-19 images based on inception-ResNet Comput Methods Programs Biomed 225 2022 107053 10.1016/j.cmpb.2022.107053 6 Khan A.I. Shah J.L. Bhat M.M. CoroNet: A deep neural network for detection and diagnosis of COVID-19 from chest x-ray images Comput Methods Programs Biomed 196 2020 105581 10.1016/j.cmpb.2020.105581 7 Wang G. Liu X. Shen J. A deep-learning pipeline for the diagnosis and discrimination of viral, non-viral and COVID-19 pneumonia from chest X-ray images Nature Biomedical Engineering 5 2021 509 521 10.1038/s41551-021-00704-1 8 Brunese L. Mercaldo F. Reginelli A. Explainable Deep Learning for Pulmonary Disease and Coronavirus COVID-19 Detection from X-rays Comput Methods Programs Biomed 196 2020 105608 10.1016/j.cmpb.2020.105608 9 MacMullan M.A. Ibrayeva A. Trettner K. ELISA detection of SARS-CoV-2 antibodies in saliva Scientific Reports 10 2020 20818 10.1038/s41598-020-77555-4 33257702 10 Yin Q. Zhang Y. Lian L. Chemiluminescence Immunoassay Based Serological Immunoassays for Detection of SARS-CoV-2 Neutralizing Antibodies in COVID-19 Convalescent Patients and Vaccinated Population Viruses 2021 13 10.3390/v13081508 35062217 11 Owen S.I. Williams C.T. Garrod G. Twelve lateral flow immunoassays (LFAs) to detect SARS-CoV-2 antibodies Journal of Infection 2021 10.1016/j.jinf.2021.12.007 12 Peng Y. Pan Y. Sun Z. An electrochemical biosensor for sensitive analysis of the SARS-CoV-2 RNA Biosensors and Bioelectronics 186 2021 113309 10.1016/j.bios.2021.113309 13 Nachtigall F.M. Pereira A. Trofymchuk O.S. Detection of SARS-CoV-2 in nasal swabs using MALDI-MS Nature Biotechnology 38 2020 1168 1173 10.1038/s41587-020-0644-7 14 Delafiori J. Navarro L.C. Siciliano R.F. Covid-19 Automated Diagnosis and Risk Assessment through Metabolomics and Machine Learning Analytical Chemistry 93 2021 2471 2479 10.1021/acs.analchem.0c04497 33471512 15 Yan L. Yi J. Huang C. Rapid Detection of COVID-19 Using MALDI-TOF-Based Serum Peptidome Profiling Analytical Chemistry 93 2021 4782 4787 10.1021/acs.analchem.0c04590 33656857 16 Ralbovsky N.M. Lednev I.K. Towards development of a novel universal medical diagnostic method: Raman spectroscopy and machine learning Chem Soc Rev 49 2020 7428 7453 10.1039/d0cs01019g 32996518 17 AlMasoud N. Muhamadali H. Chisanga M. Discrimination of bacteria using whole organism fingerprinting: the utility of modern physicochemical techniques for bacterial typing Analyst 146 2021 770 788 10.1039/d0an01482f 33295358 18 Duarte J.M. Sales N.G.S. Braga J.W.B. Discrimination of white automotive paint samples using ATR-FTIR and PLS-DA for forensic purposes Talanta 240 2021 123154 10.1016/j.talanta.2021.123154 19 Yan S. Wang S. Qiu J. Raman spectroscopy combined with machine learning for rapid detection of food-borne pathogens at the single-cell level Talanta 226 2021 122195 10.1016/j.talanta.2021.122195 20 Rebrošová K. Bernatová S. Šiler M. Raman spectroscopy—a tool for rapid differentiation among microbes causing urinary tract infections Analytica Chimica Acta 2021 10.1016/j.aca.2021.339292 21 Bratchenko I.A. Bratchenko L.A. Khristoforova Y.A. Classification of skin cancer using convolutional neural networks analysis of Raman spectra Comput Methods Programs Biomed 219 2022 106755 10.1016/j.cmpb.2022.106755 22 Shu C. Yan H. Zheng W. Deep Learning-Guided Fiberoptic Raman Spectroscopy Enables Real-Time In Vivo Diagnosis and Assessment of Nasopharyngeal Carcinoma and Post-treatment Efficacy during Endoscopy Analytical Chemistry 93 2021 10898 10906 10.1021/acs.analchem.1c01559 34319713 23 Cialla-May D. Krafft C. Rosch P. Raman Spectroscopy and Imaging in Bioanalytics Analytical Chemistry 94 2022 86 119 10.1021/acs.analchem.1c03235 34920669 24 Fong S.J. Dey N. Chaki J. Artificial intelligence for coronavirus outbreak 2021 Springer 10.1007/978-981-15-5936-5 25 Mehta K. Atak A. Sahu A. An early investigative serum Raman spectroscopy study of meningioma Analyst 143 2018 1916 1923 10.1039/c8an00224j 29620771 26 Ami D. Duse A. Mereghetti P. Tear-Based Vibrational Spectroscopy Applied to Amyotrophic Lateral Sclerosis Analytical Chemistry 93 2021 16995 17002 10.1021/acs.analchem.1c02546 34905686 27 Paraskevaidi M. Morais C.L.M. Freitas D.L.D. Blood-based near-infrared spectroscopy for the rapid low-cost detection of Alzheimer's disease Analyst 143 2018 5959 5964 10.1039/c8an01205a 30183030 28 Li Q. Li W. Zhang J. An improved k-nearest neighbour method to diagnose breast cancer Analyst 143 2018 2807 2811 10.1039/c8an00189h 29863729 29 Hartatik H. Tamam M.B. Setyanto A. 2020 2nd International Conference on Cybernetics and Intelligent System (ICORIS) 2020 1 5 30 Abdoh S.F. Abo Rizka M. Maghraby F.A. Cervical Cancer Diagnosis Using Random Forest Classifier With SMOTE and Feature Reduction Techniques IEEE Access 6 2018 59475 59485 10.1109/access.2018.2874063 31 Do B.H. Langlotz C. Beaulieu C.F. Bone Tumor Diagnosis Using a Naive Bayesian Model of Demographic and Radiographic Features Journal of Digital Imaging 30 2017 640 647 10.1007/s10278-017-0001-7 28752323 32 Yang Y. Yang Y. Liu Z. Microcalcification-Based Tumor Malignancy Evaluation in Fresh Breast Biopsies with Hyperspectral Stimulated Raman Scattering Analytical Chemistry 93 2021 6223 6231 10.1021/acs.analchem.1c00522 33826297 33 Lin Y.T. Chu C.Y. Hung K.S. Can machine learning predict pharmacotherapy outcomes? An application study in osteoporosis Comput Methods Programs Biomed 225 2022 107028 10.1016/j.cmpb.2022.107028 34 Yang H. Li X. Cao H. Using machine learning methods to predict hepatic encephalopathy in cirrhotic patients with unbalanced data Comput Methods Programs Biomed 211 2021 106420 10.1016/j.cmpb.2021.106420 35 Guleken Z. Jakubczyk P. Wieslaw P. Characterization of Covid-19 infected pregnant women sera using laboratory indexes, vibrational spectroscopy, and machine learning classifications Talanta 237 2022 122916 10.1016/j.talanta.2021.122916 36 Guleken Z. Tuyji Tok Y. Jakubczyk P. Development of novel spectroscopic and machine learning methods for the measurement of periodic changes in COVID-19 antibody level Measurement 196 2022 111258 10.1016/j.measurement.2022.111258 37 Mazo C. Alegre E. Trujillo M. Classification of cardiovascular tissues using LBP based descriptors and a cascade SVM Comput Methods Programs Biomed 147 2017 1 10 10.1016/j.cmpb.2017.06.003 28734525 38 Wang Y. Xu J. Cui D. Classification and Identification of Archaea Using Single-Cell Raman Ejection and Artificial Intelligence: Implications for Investigating Uncultivated Microorganisms Analytical Chemistry 93 2021 17012 17019 10.1021/acs.analchem.1c03495 34910467 39 Ryzhikova E. Ralbovsky N.M. Sikirzhytski V. Raman spectroscopy and machine learning for biomedical applications: Alzheimer's disease diagnosis based on the analysis of cerebrospinal fluid Spectrochim Acta A Mol Biomol Spectrosc 248 2021 119188 10.1016/j.saa.2020.119188 40 Dey N. Advancements in applied metaheuristic computing 2017 IGI global 41 Tang, R.; Fong, S.; Dey, N., Metaheuristics and chaos theory. 2018, 182-196. doi:10.5772/intechopen.72103. 42 Mirjalili S. Mirjalili S.M. Lewis A. Grey Wolf Optimizer Advances in Engineering Software 69 2014 46 61 10.1016/j.advengsoft.2013.12.007 43 Mirjalili S. How effective is the Grey Wolf optimizer in training multi-layer perceptrons Applied Intelligence 43 2015 150 161 10.1007/s10489-014-0645-7 44 Vandenberg O. Martiny D. Rochas O. Considerations for diagnostic COVID-19 tests Nat Rev Microbiol 19 2021 171 183 10.1038/s41579-020-00461-z 33057203 45 Dighe K. Moitra P. Alafeef M. A rapid RNA extraction-free lateral flow assay for molecular point-of-care detection of SARS-CoV-2 augmented by chemical probes Biosens Bioelectron 200 2022 113900 10.1016/j.bios.2021.113900 46 Zhang D. Guo Y. Zhang L. Integrated System for On-Site Rapid and Safe Screening of COVID-19 Anal Chem 94 2022 13810 13819 10.1021/acs.analchem.2c02337 36184789 47 Zhang P. Chen L. Hu J. Magnetofluidic immuno-PCR for point-of-care COVID-19 serological testing Biosens Bioelectron 195 2022 113656 10.1016/j.bios.2021.113656 48 Mei X. Lee H.C. Diao K.Y. Artificial intelligence-enabled rapid diagnosis of patients with COVID-19 Nature Medicine 26 2020 1224 1228 10.1038/s41591-020-0931-3 49 Baker M.J. Byrne H.J. Chalmers J. Clinical applications of infrared and Raman spectroscopy: state of play and future challenges Analyst 143 2018 1735 1757 10.1039/c7an01871a 29504623 50 Morais C.L.M. Lima K.M.G. Singh M. Tutorial: multivariate classification for vibrational spectroscopy in biological samples Nat Protoc 15 2020 2143 2162 10.1038/s41596-020-0322-8 32555465 51 Nascimento M.H.C. Marcarini W.D. Folli G.S. Noninvasive Diagnostic for COVID-19 from Saliva Biofluid via FTIR Spectroscopy and Multivariate Analysis Anal Chem 94 2022 2425 2433 10.1021/acs.analchem.1c04162 35076208 52 Du Y. Han D. Liu S. Raman spectroscopy-based adversarial network combined with SVM for detection of foodborne pathogenic bacteria Talanta 237 2022 122901 10.1016/j.talanta.2021.122901 53 Wood B.R. Kochan K. Bedolla D.E. Infrared Based Saliva Screening Test for COVID-19 Angew Chem Int Ed Engl 60 2021 17102 17107 10.1002/anie.202104453 34043272 54 Yin G. Li L. Lu S. An efficient primary screening of COVID-19 by serum Raman spectroscopy Journal of Raman Spectroscopy 2021 10.1002/jrs.6080 55 Stone R.W.K.A.J.T. Computer Aided Design of Experiments 11 1969 137 148 56 Baker M.J. Hussain S.R. Lovergne L. Developing and understanding biofluid vibrational spectroscopy: a critical review Chem Soc Rev 45 2016 1803 1818 10.1039/c5cs00585j 26612430 57 Savitzky A. Golay M.J.J.A.c. Smoothing and differentiation of data by simplified least squares procedures Analytical Chemistry 36 1964 1627 1639 10.1021/ac60214a047 58 Geladi P. MacDougall D. Martens H. Linearization and Scatter-Correction for Near-Infrared Reflectance Spectra of Meat Applied Spectroscopy 39 1985 491 500 10.1366/0003702854248656 59 Gan F. Ruan G. Mo J. Baseline correction by improved iterative polynomial fitting with automatic threshold Chemometrics and Intelligent Laboratory Systems 82 2006 59 65 10.1016/j.chemolab.2005.08.009 60 Morais C.L.M. Paraskevaidi M. Cui L. Standardization of complex biologically derived spectrochemical datasets Nature Protocols 14 2019 1546 1577 10.1038/s41596-019-0150-x 30953040 61 Li S.S. Yin B. Zhai H.L. An effective approach to the quantitative analysis of skin-whitening agents in cosmetics with different substrates based on conventional UV-Vis determination Analytical Methods 11 2019 1500 1507 10.1039/c9ay00007k 62 Yin B. Zhai H.L. Zhao B.Q. Chemometrics-assisted simultaneous voltammetric determination of multiple neurotransmitters in human serum Bioelectrochemistry 139 2021 107739 10.1016/j.bioelechem.2021.107739 63 Lu S.H. Zhai H.L. Zhao B.Q. Novel Approach to the Analysis of Chemical Third-Order Data Journal of Chemical Information and Modeling 60 2020 4750 4756 10.1021/acs.jcim.0c00554 32955255 64 El-Kenawy E.M. Ibrahim A. Mirjalili S. Novel Feature Selection and Voting Classifier Algorithms for COVID-19 Classification in CT Images IEEE Access 8 2020 179317 179335 10.1109/ACCESS.2020.3028012 34976558 65 Sokolova M. Lapalme G. A systematic analysis of performance measures for classification tasks Information Processing & Management 45 2009 427 437 10.1016/j.ipm.2009.03.002 66 Huang T.Y. Yu J.C.C. Development of Crime Scene Intelligence Using a Hand-Held Raman Spectrometer and Transfer Learning Analytical Chemistry 93 2021 8889 8896 10.1021/acs.analchem.1c01099 34134486 67 Fawcett T. An introduction to ROC analysis Pattern Recognition Letters 27 2006 861 874 10.1016/j.patrec.2005.10.010 68 Li J. Fine J.P. ROC analysis with multiple classes and multiple tests: methodology and its application in microarray studies Biostatistics 9 2008 566 576 10.1093/biostatistics/kxm050 18304996 69 Chandrasekaran S.S. Agrawal S. Fanton A. Rapid detection of SARS-CoV-2 RNA in saliva via Cas13 Nature Biomedical Engineering 2022 10.1038/s41551-022-00917-y 70 Xu J. Li Y. Gan F. Salivary Glands: Potential Reservoirs for COVID-19 Asymptomatic Infection Journal of Dental Research 99 2020 989 10.1177/0022034520918518 32271653 71 Kumar M.A. Gopal M.J.E.S.w.A. Reduced one-against-all method for multiclass SVM classification Expert Systems with Applications 38 2011 14238 14248 10.1016/j.eswa.2011.04.237	0	PMC9711896	NO-CC CODE	2022-12-02 23:21:31	no	Comput Methods Programs Biomed. 2022 Dec 1;:107295	utf-8	Comput Methods Programs Biomed	2,022	10.1016/j.cmpb.2022.107295	oa_other
==== Front J Infect J Infect The Journal of Infection 0163-4453 1532-2742 The British Infection Association. Published by Elsevier Ltd. S0163-4453(22)00687-9 10.1016/j.jinf.2022.11.026 Letter to the Editor Viral dynamics during SARS-CoV-2 omicron infection highlight presymptomatic and asymptomatic infectiousness Jiang Li a1 Tang Lu b1 Zhu Linyu c1 Zhu Yufang d1 Yang Song e Chen Wenjie f Fan Yi g Yang Xuejiao h Yang Shuai h Zheng Yulan h Xu Yunsheng bc Hong Peng bij⁎ a Health Management Center, West China Guang'an Hospital, Sichuan University, Guang'an, Sichuan, China b Department of Research, The Seventh Affiliated Hospital, Sun Yat-sen University School of Medicine, Shenzhen, Guangdong, China c Department of Dermatology, The Seventh Affiliated Hospital, Sun Yat-sen University School of Medicine, Shenzhen, Guangdong, China d Laboratory of Clinical Immunology, Division of Laboratory Medicine, Xiangyang Central Hospital, Affiliated Hospital of Hubei University of Arts and Science, Xiangyang, Hubei, China e Department of Radiology and Medical Imaging, West China Guang'an Hospital, Sichuan University, Guang'an, Sichuan, China f Department of Nephrology, Yuechi County People's Hospital, Guang'an, Sichuan, China g Program of Clinical Medicine, Sun Yat-sen University School of Medicine, Shenzhen, Guangdong, China h Department of Respiratory and Critical Care Medicine, Xiangyang Central Hospital, Affiliated Hospital of Hubei University of Arts and Science, Xiangyang, Hubei, China i Division of Research and Development, VA NY Harbor Healthcare System Brooklyn Campus, 800 Poly Place, Main Bldg 3-201, Brooklyn, NY 11209, USA j Department of Cell Biology, State University of New York Downstate Health Sciences University, Brooklyn, NY 11203, USA ⁎ Corresponding author at: Department of Research, The Seventh Affiliated Hospital, Sun Yat-sen University School of Medicine, Shenzhen, Guangdong, China. 1 These authors contributed equally to this work. 1 12 2022 1 12 2022 28 11 2022 © 2022 The British Infection Association. Published by Elsevier Ltd. All rights reserved. 2022 The British Infection 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. ==== Body pmcTo the editor, Peaking of SARS-CoV-2 viral shedding usually correlates with high infectiousness.1 Despite of early awareness of prevalent presymptomatic viral shedding that may evade control measures and drive stealth transmission,2 SARS-CoV-2 viral kinetics during presymptomatic stage or asymptomatic infections was still poorly understood due to requiring repeated tests in asymptomatic population. SARS-CoV-2 omicron infections were more likely to be asymptomatic or milder comparing to delta infections,3, 4, 5 which were associated with shifting virological characteristics of omicron in animal models,6, 7, 8 and might offer key evolutionary advantage over prior variants.9 Here, we provided evidence of peak viral shedding at presymptomatic stage and during asymptomatic omicron infections by analyzing viral RNA kinetics during the entire infection course. 1,085 SARS-CoV-2 PCR-positive cases in Guang'an city of western China, who were admitted to West China Guang'an Hospital between 2022/5/9 and 2022/6/3 were enrolled in this study. Characteristics of the cohort were summarized in Table 1 . Methods of population screening for omicron infections and in-hospital procedures including PCR testing and symptom monitoring were detailed in the Supplementary Materials. The human study protocols have been approved by West China Guang'an Hospital and all patient identifications were replaced by anonymous codes during abstraction as stipulated by the Declaration of Helsinki.Table 1 Demographic and clinical characteristics of the cohort. Table 1 All cases (n = 1085) Asymptomatic (n = 766) Symptomatic (n = 319) P* Age in years, median (IQR) 43 (17–56) 42 (16–55) 47 (18–57) .039 <18, n (%) 291 (26.8) 215 (28.1) 76 (23.8) .079 18–59, n (%) 606 (55.9) 430 (56.1) 176 (55.2) ≥60, n (%) 188 (17.3) 121 (15.8) 67 (21.0) Female sex at birth, n (%) 652 (60.1) 443 (57.8) 209 (65.5) .021 Days of RNA-positive, median (IQR) 13 (11–17) 13 (10–17) 14 (11–17) .049 Days from first RNA-positive to symptom onset, median (IQR) [n] 0 (0–4) [319] / 0 (0–4) [319] / Days from first RNA-positive to peak RNA load, median (IQR) [n] 3 (2–5) [298] 3 (2–5) [222] 3 (2–5) [76] .583 Vaccination status, n (%) Unvaccinated 29 (2.7) 18 (2.4) 11 (3.5) .063 Partial vaccination 9 (0.8) 3 (0.4) 6 (1.9) Full vaccination without booster 346 (31.9) 248 (32.4) 98 (30.8) Full vaccination with booster 699 (64.4) 496 (64.8) 203 (63.8) Unknown 2 (0.2) 1 (0.1) 1 (0.3) Days from last vaccination to first RNA-positive, median (IQR) [n] 152 (133–265) [782] 152 (137–258) [524] 151 (128–267) [258] .916 COVID-19 severity, n (%) Asymptomatic 766 (70.6) 766 (100) 0 (0) / Mild 319 (29.4) 0 (0) 319 (100) COVID-19 symptoms, n (%) Respiratory 285 (26.3) 0 (0) 285 (90.5) / Neuromuscular 75 (6.9) 0 (0) 75 (23.8) / Gastrointestinal 17 (1.6) 0 (0) 17 (5.4) / Other 4 (0.4) 0 (0) 4 (1.3) / Comorbidities, n (%) Respiratory disorders 13 (1.2) 8 (1.0) 5 (1.6) .541 Cardiovascular disorders 96 (8.8) 67 (8.8) 29 (9.1) .907 Metabolic disorders 31 (2.9) 22 (2.9) 9 (2.8) 1.00 Immune disorders 7 (0.6) 3 (0.4) 4 (1.3) .204 Other infectious diseases 11 (1.0) 8 (1.0) 3 (0.9) 1.00 Cancer 12 (1.1) 6 (0.8) 6 (1.9) .122 Other 3 (0.3) 3 (0.4) 0 (0) .560 ⁎ Categorical variables were assessed by Fisher's exact test or Chi-square test and continuous variables were assessed by Mann-Whitney U test. Tests were performed between asymptomatic and symptomatic groups. Abbreviations: IQR, interquartile range. Guang'an reported no prior SARS-CoV-2 outbreaks and >95% vaccine coverage. Patient zero with asymptomatic infection of omicron BA.2.2 sub-variant arrived in Guang'an on 2022/5/4 and was the index of and outbreak that lead to more than 1,000 cases despite early interventions (Fig. S1). Viral RNA kinetics of 962 cases showed prevalent asymptomatic and presymptomatic peaks (Fig. S2). 69% (668/962) of them, who were asymptomatic till the end of follow-up, showed comparable peak viral RNA levels (geometric mean 4.4 × 105 vs 2.7 × 105 copies/ml, p=.173) and peak timing (median 5 vs 5 days, p=.378) with symptomatic cases (Fig. 1A and B). Interestingly, 21% (61/294) symptomatic cases showed pre-symptomatic peaking of viral RNA, which were rarely seen in pre-alpha and alpha variant infections,1 with higher peak viral loads than other symptomatic (geometric mean 1.1 × 106 vs 1.9 × 105 copies/ml, p<.001) or asymptomatic cases (geometric mean 1.1 × 106 vs 4.4 × 105 copies/ml, p=.014) (Fig. 1 C). Simultaneous or post-symptomatic peaks were dominant in pre-alpha and alpha variant infections,1 whereas they were less common in early-stage (≤5 days post-lockdown, 41/83) than late-stage omicron cases in our cohort (192/211, p<.0001).Fig. 1 Peak viral RNA loads and timing according to symptomatic status or stage of outbreak. (A,B) Comparison of peak viral RNA levels (A) and time from viral RNA-positive to peak viral RNA (B) between asymptomatic (n = 668) and symptomatic cases (n = 294). Lines indicate geometric means ± 95% confidence intervals (A) or medians (B). Viral RNA concentration was calculated based on the Ct value of Orf1ab gene as described in the Supplementary Materials. Lower limit of viral RNA concentration was set at 19 copies/mL, which corresponded to cut-off Ct value of 38 for Orf1ab gene. (C) Comparison of peak viral RNA levels between asymptomatic (n = 668), pre-symptomatic peaking (n = 61) and post-symptomatic peaking cases (n = 233). The pre-symptomatic and post-symptomatic peaking cases together made the symptomatic group in (A). Lines indicate geometric means ± 95% confidence intervals. (D) Comparison of peak viral RNA levels between early- (≤5 days after lockdown, n = 436) and late-stage cases (n = 526). Lines indicate geometric means ± 95% confidence intervals. Statistical significance was assessed by Mann-Whitney U tests. Fig 1 Intriguingly, late-stage infections had lower peak viral loads (geometric mean 1.3 × 106 vs 1.5 × 105 copies/ml, p<.0001) at the time of increased preventive measures (Fig. 1D), whereas among those infected, vaccination was not associated with peak viral load, duration of viral clearance, or symptomatic disease (Fig. S3). Of note, endocrine disorders such as diabetes were associated with higher peak viral loads, whereas chronic infections such as HBV infections were also associated with higher peak viral loads despite of not reaching statistical significance due to small number of positive cases (Fig. S3). These findings together provided evidence of an early and asymptomatic window of infectiousness during some omicron infections. Such front-loaded infectiousness of omicron poses significant challenges to pandemic control policies, and measure undertaken earlier were less likely to be effective in this circumstance. Symptom-based testing and extended quarantine used to be an effective strategy to block SARS-CoV-2 transmission in pre-omicron era due to the long infectious window post-symptom onset.10 However, ongoing epidemic in countries with strict COVID-19 policies suggested that omicron carriers may already spread infections before being quarantined,10 which had exactly happened in Guang'an and could only be contained by early detection and interventions. Hopefully, upcoming omicron-adapted vaccines could control the pandemic; otherwise, more efficient and cost-effective screening techniques for asymptomatic population might be needed for future variants with high virulence. Role of the funding source This study was supported by the National Natural Science Foundation of China (grants 82072862 and 82272949 to YX). Author contributions LJ abstracted case information and did clinical investigations; LT did epidemiological investigations and statistical analysis; LZ, SY and WC collected PCR Ct values; YZhu analyzed viral kinetics and interpreted findings; YF did survey of Guang'an residents; XY, SY, YZheng and YX interpreted findings and did investigations; PH conceived the study, analyzed data, and wrote the manuscript. No authors were precluded from accessing data in the study, and they accept responsibility to submit for publication. All authors participate in the revision of the manuscript and approve the final manuscript. Declaration of Competing Interest The authors declare no competing interests. Appendix Supplementary materials Image, application 1 Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jinf.2022.11.026. ==== Refs References 1 Hakki S. Zhou J. Jonnerby J. Singanayagam A. Barnett J.L. Madon K.J. Onset and window of SARS-CoV-2 infectiousness and temporal correlation with symptom onset: a prospective, longitudinal, community cohort study Lancet Respir Med 10 11 2022 1061 1073 NovPubMed PMID: 35988572. Pubmed Central PMCID: PMC9388060. Epub 20220818 35988572 2 He X. Lau E.H.Y. Wu P. Deng X. Wang J. Hao X. Temporal dynamics in viral shedding and transmissibility of COVID-19 Nat Med 26 5 2020 672 675 MayPubMed PMID: 32296168. Epub 20200415 32296168 3 Olaiz-Fernandez G. Vicuna de Anda F.J. Diaz-Ramirez J.B. Fajardo Dolci G.E. Bautista-Carbajal P. Angel-Ambrocio A.H. Effect of Omicron on the prevalence of COVID-19 in international travelers at the Mexico city international airport. December 16th, 2021 to January 31st, 2022 Travel Med Infect Dis 49 2022 102361 May 29PubMed PMID: 35640809. Pubmed Central PMCID: PMC9148423. Epub 20220529 4 Nyberg T. Ferguson N.M. Nash S.G. Webster H.H. Flaxman S. Andrews N. 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 Apr 2PubMed PMID: 35305296. Pubmed Central PMCID: PMC8926413. Epub 20220316 35305296 5 Iuliano A.D. Brunkard J.M. Boehmer T.K. Peterson E. Adjei S. Binder A.M. 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 MMWR Morb Mortal Wkly Rep 71 4 2022 146 152 Jan 28PubMed PMID: 35085225. Epub 20220128 35085225 6 Yuan S. Ye Z.W. Liang R. Tang K. Zhang A.J. Lu G. Pathogenicity, transmissibility, and fitness of SARS-CoV-2 Omicron in Syrian hamsters Science 377 6604 2022 428 433 Jul 22PubMed PMID: 35737809. Epub 20220623 35737809 7 Suzuki R. Yamasoba D. Kimura I. Wang L. Kishimoto M. Ito J. Attenuated fusogenicity and pathogenicity of SARS-CoV-2 Omicron variant Nature 603 7902 2022 700 705 MarPubMed PMID: 35104835. Pubmed Central PMCID: PMC8942852. Epub 20220201 35104835 8 Shuai H. Chan J.F. Hu B. Chai Y. Yuen T.T. Yin F. Attenuated replication and pathogenicity of SARS-CoV-2 B.1.1.529 Omicron Nature 603 7902 2022 693 699 MarPubMed PMID: 35062016. Epub 20220121 35062016 9 Berkhout B. Herrera-Carrillo E. SARS-CoV-2 evolution: on the sudden appearance of the Omicron variant J Virol 96 7 2022 e0009022 Apr 13PubMed PMID: 35293771. Pubmed Central PMCID: PMC9006888. Epub 20220316 10 Mefsin Y.M. Chen D. Bond H.S. Lin Y. Cheung J.K. Wong J.Y. Epidemiology of infections with SARS-CoV-2 Omicron BA.2 variant, Hong Kong, January-March 2022 Emerg Infect Dis 28 9 2022 1856 1858 SepPubMed PMID: 35914518. Pubmed Central PMCID: PMC9423929. Epub 20220801 35914518	36462585	PMC9711897	NO-CC CODE	2022-12-14 23:52:33	no	J Infect. 2022 Dec 1; doi: 10.1016/j.jinf.2022.11.026	utf-8	J Infect	2,022	10.1016/j.jinf.2022.11.026	oa_other
==== Front Clin Microbiol Infect Clin Microbiol Infect Clinical Microbiology and Infection 1198-743X 1469-0691 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. S1198-743X(22)00595-X 10.1016/j.cmi.2022.11.022 Commentary Platform trials as the way forward in infectious disease’ clinical research: the case of COVID-19 Pericàs Juan M. 123 Derde Lennie PG. 45 Berry Scott M. 67 EU-PEARL and REMAP-CAP investigators 1 Liver Unit, Vall d’Hebron University Hospital, Barcelona, Spain 2 Vall d’Hebron Institut de Recerca (VHIR), Vall d’Hebron Barcelona Campus Hospitalari, Barcelona, Spain 3 Centro de Investigación Biomédica en Red de enfermedades digestivas y hepáticas (CIBERehd), Madrid, Spain 4 Department of Intensive Care Medicine, University Medical Center Utrecht, Utrecht, The Netherlands 5 Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands 6 Berry Consultants, LLC, Austin, TX, USA; 7 Department of Biostatistics, University of Kansas Medical Center, Kansas City, KS, USA 1 12 2022 1 12 2022 6 9 2022 22 11 2022 24 11 2022 © 2022 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved. 2022 European Society of Clinical Microbiology and Infectious Diseases 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 Adaptive designs Bayesian statistics COVID-19 Master protocols Platform trials Editor: Professor L Leibovici ==== Body pmcClinical research in an era of pandemics In addition to its overwhelming impacts on global health and health systems’ resilience, the COVID-19 pandemic has elicited an unprecedented global scientific effort. Since early in the onset of the pandemic, numerous randomized clinical trials (RCTs) were set up to give response to the clinical unmet needs [1]. Remarkably, platform trials (PT) emerged as a largely employed appraisal in clinical research in the wake of COVID-19, e.g. REMAP-CAP, WHO Solidarity, RECOVERY, ATTACC, PRINCIPLE, AGILE, COPPS, OPTIMISE C-19, among others. The results from some of these trials have led to the current standard of care recommendations issued in guidelines worldwide for the treatment of COVID-19. However, for most clinicians and investigators, the basics of PT, as well as their constraints and advantages, remain largely unknown. In this Commentary, we focus on the basic concepts and methodological keys of PT using some of the recent developments in COVID-19 to illustrate the potential of these trial designs. Master protocols and platform trials: the basics Alongside umbrella and basket trials, PT are one of the possible designs of master protocols (MP) (Figure 1 and Glossary in Table 1 ). MP allow testing for multiple hypotheses within a single protocol, and stratification by (biomarker) subgroup, leading to a more personalized evaluation of therapies compared to traditional RCTs [2, 3, 4]. Whereas traditional trials are based on a “one population, one drug, one disease” strategy, PT, umbrella and basket trials share key design components and operational aspects that aim to increase the efficiency (both in time and resources) at achieving high-quality evidence for medical interventions by evaluating multiple populations, and/or multiple interventions, and/or multiple diseases at the same time [2].Figure 1 Scheme of Master Protocols’ design. a. Umbrella trial; b. Basket trial; and c. Platform trial. Figure 1 Box 1 Glossary of basic terms related to platform trials Box 1Main definitions Adaptive designs for clinical trials: an adaptive trial uses data that accumulates during the study to modify study elements in a flexible though prespecified manner. Elements that may be modified include the sample size, end points, eligible populations, randomization ratio, and interventions. Master protocol: overarching trial characteristics and rules guiding several parallel subtrials (either concomitant or sequential) under the same protocol and governance structure. Multi-arm multi-stage (MAMS) trials: adaptive design and set up that aims to answer multiple questions simultaneously under the same regulatory framework. Several different treatment options can be compared simultaneously, usually against a single control arm. Tested treatments can be different drugs at single dosages, combination of drugs, or various doses of the same drug. Adaptation rules and interim analyses allow for changes in the subtrials/arms that make part of the trial. The master protocol of MAMS can be adopt different forms. Basket and umbrella trials: both are multiarm, often adaptive designs of randomized clinical trials under a master protocol, but while in the former multiple diseases or phenotypes are treated with the same treatment, in the latter multiple treatments are tested against the same disease. Platform trial: derivation of MAMS that consist on prospective, disease-focused, adaptive, randomized clinical trial that compares multiple interventions against a single, constant control group. Bayesian statistics: statistical set of methods characterised by an approach where probability expresses the degree of certainty of an event to happen within a timeframe. In platform trials, this allows to decide whether to continue or discontinue an arm containing active treatment based on interim analysis and the prediction that the drug might or not achieve a certain threshold of efficacy (i.e., futility/efficacy stop rules). Specific Appendices (Domain and Intervention specific): these appendices contain the specific details that concern to each subtrial or arm that is incorporated to the platform trial, whereas the master protocol provides the overarching information. Integrated Research Platform: the combination of the governing bodies, the network of clinical sites, the methodological tools including the master protocol, and the regulatory and legal aspects that allow setting up a platform trial. Before the outbreak of COVID-19, MP were mostly used in oncology and hematology (e.g., I-SPY in breast cancer and GBM AGILE in brain tumors), and less frequently in mental illnesses (e.g. DIAN-TU in Alzheimer and CATIE in schizophrenia) and infectious disease (e.g. ADAPT for the treatment of multi-resistant pathogens, PREVAIL II in Ebola, and REMAP-CAP in pneumonia) [3]. MP might be built leveraging the infrastructure and similarities of previous individual trials or be set up de novo, and they can both compare directly competing interventions or evaluate these in parallel with their respective controls. Due to PT’ adaptive design, new arms can be added along the way, and those reaching a pre-specified probability of failure or success can be dropped (and become the new standard of care in case of success) [4]. As non-efficacious compounds are dropped early, patients are at lower risk of being exposed to these compounds, and pharmaceutical companies can quickly move on to investigate novel compounds at lower costs. The use of a shared control arm for multiple active compounds is efficient, and the multifactorial design (allowing participants to receive a randomized allocation for more than one aspect of their treatment) adds even more efficiency. From the participant’s perspective, this reduces the probability of receiving no active treatment at all. The main advantages of PT are the likely smaller sample size and shorter time to answer any given question compared to a fixed design. However, PT are not without shortcomings, which might be more or less challenging depending on the context, e.g. during COVID-19 many of the steps involved were shortened. Amongst the potential disadvantages, the two most relevant refer on the one hand to the increased design and operational complexity, which is largely accompanied by an increased difficulty explaining the concepts to funders, the general public and the research community; and on the other hand, the longer trial set up time including protocol writing, ethics and institutional approval. REMAP-CAP: an example of global multi-collaborative success The Randomized Embedded Multifactorial Adaptive Platform for Community Acquired Pneumonia (REMAP-CAP) was conceived after the Influenza A H1N1(pdm2009) pandemic, and first funded in 2014 as a global trial investigating simultaneously multiple interventions for severe CAP with the specific aim to be prepared for the next pandemic [5]. Even before the COVID-19 pandemic, a “sleeping” Pandemic Appendix to the Core Protocol was approved, which described potential adaptations applicable during a pandemic, e.g.,primary end-point, platform eligibility, statistical model. The latter serves to enable separate analysis of pandemic patients, enrolment of patients in different disease severity states, and to evaluate the differential efficacy of interventions depending on illness severity [5]. REMAP-CAP findings, alongside RECOVERY [6], WHO Solidarity [7], and other relevant trials, critically contributed to informing COVID-19 treatment guidelines. The design approach of REMAP-CAP was specifically chosen for its value infuture pandemics that involved hospitals and specifically Intensive Care Units (ICUs). The trial was set up in a modular way, with a Core Protocol describing all general aspects of the trial, and Domain Specific Appendices (DSAs) for each of the therapeutic areas of interest (Domains). Within each Domain, two or more mutually exclusive interventions are compared. Patients may be eligible for one or more domains. The modular structure of the protocol allows new interventions and domains to be submitted as substantial amendments without changing the Core Protocol. There are additional Appendices for the statistical analysis, pandemics, and for each region the trial is active in, to allow for regional differences in legislation (Region Specific Appendices). This modular structure is also reflected in the governance structure [5]. The statistical framework underpinning the REMAP design is Bayesian. There is one overarching Bayesian cumulative logistic model driving all adaptations, rules for statistical parametrization, and result summaries. The decision to use a Bayesian analysis was driven in part by the uncertainty of the extent of the pandemic. The sample size could be small or large, and there may be important external events, such as other trial results, that alter the design of REMAP-CAP. Operationally, the challenges of running a global trial in the midst of a pandemic were substantial. Some challenges, like limited availability of research personnel and other resources, are unavoidable. The trial design, embedding research in clinical care and aiming to optimize the trade-off between learning and doing, alleviated some of the operational challenges [8]. The biggest challenges however were not related to the design. The most important hurdles to delivering the trial during the pandemic were in facilitating collaboration, data collection and sharing, leveraging funding, and optimizing prioritization at a global level. These challenges cannot be solved by any individual trial, but are extremely important for the global research community, funders, and regulators to address together [9,10]. Future perspectives Boosting adaptive designs through reusable trial infrastructures as PT might help overcoming some of the major challenges and limitations of standalone RCTs conducted in the field of infectious diseases, as overly optimistic effect sizes, unnuanced conclusions based on dichotomization of results, limited focus on patient-centred outcomes other than mortality, lack of flexibility and ability to adapt, increasing the risk of inconclusive results and limiting knowledge gains before trial completion; and inefficiency due to lack of re-use of trial infrastructure [11]. However, clinical research and drug development through a broader use of adaptive platform designs in COVID-19 specifically and in infectious diseases in general would require some large scientific, cultural and infrastructural changes. Amongst some salient scientific challenges, there are still some gaps to be filled in COVID-19 (e.g., treatment of long/persistent COVID-19), and biomarkers need to be developed that allow for stratification in other relevant infectious diseases, including neglected tropical diseases. From the cultural or educational perspective, further steps should be taken before achieving a scenario in which investigators, sponsors and health authorities are familiar with the benefits of PT and aligned in joint undertakings that expand PT use. Moreover, infrastructures and collaborative pathways are to be built from which further PT in infectious diseases might emerge. For instance, electronic health records can be embedded in trials design, with automatic integration leading to substantial logistic improvements regarding data collection, integration of randomisation modules, and alerts about potentially eligible patients. The EU Patient-cENtric clinical tRIal pLatforms (EU-PEARL) Consortium is creating such structure and framework for four diseases including tuberculosis, relying on the concept of integrated research platforms (IRPs)[12]. The European Clinical Research Alliance on Infectious Diseases (Ecraid) offers a coordinated approach that aims to build a permanent, not-for-profit, pan-European clinical research network capable of rapidly initiating and completing high-quality clinical studies with greater speed and efficiency [13]. In conclusion, MP and particularly adaptive PT designs constitute an innovative, flexible, multi-collaborative and efficient way forward to boost clinical research in infectious disease that has already proved valuable in COVID-19. Authors’ contributions Conceptualization: JMP, LD, SB; Writing- Original Draft: JMP, LD, SB; Writing- Review & Editing: JMP, LD, SB; Approval of the final version: JMP, LD, SB; Visualization, Supervision & Administration: JMP. Disclosures The REMAP-CAP trial is funded by the Platform for European Preparedness Against (Re-) emerging Epidemics (PREPARE) consortium by the European Union, FP7-HEALTH-2013-INNOVATION-1 (grant 602525), the Australian National Health and Medical Research Council (grant APP1101719), the New Zealand Health Research Council (grant 16/631), the Canadian Institute of Health Research Strategy for Patient-Oriented Research Innovative Clinical Trials Program (grant 158584), the UK National Institute for Health Research (NIHR) and the NIHR Imperial Biomedical Research Centre, the Health Research Board of Ireland (grant CTN 2014-012), the UPMC Learning While Doing Program, the Breast Cancer Research Foundation, the French Ministry of Health (grant PHRC-20-0147), and the Minderoo Foundation; EU-PEARL has received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 853966-2. This Joint Undertaking receives support from the European Union’s Horizon 2020 research and innovation programme and EFPIA, Children’s Tumor Foundation, Global Alliance for TB Drug Development non-profit organization, and Springworks Therapeutics Inc. This publication reflects only the author's views. The JU is not responsible for any use that may be made of the information it contains. ==== Refs References 1 Fragkou P.C. Belhadi D. Peiffer-Smadja N. Moschopoulos C.D. Lescure F.X. Janocha H. Review of trials currently testing treatment and prevention of COVID-19 Clin Microbiol Infect 26 2020 988 998 10.1016/j.cmi.2020.05.019 32454187 2 Berry S.M. Connor J.T. Lewis R.J. The platform trial: an efficient strategy for evaluating multiple treatments JAMA 313 2015 1619 1620 10.1001/jama.2015.2316 25799162 3 Woodcock J. LaVange L.M. Master Protocols to Study Multiple Therapies, Multiple Diseases, or Both N Engl J Med 377 2017 62 70 10.1056/NEJMra1510062 28679092 4 Adaptive Platform Trials Coalition Adaptive platform trials: definition, design, conduct and reporting considerations Nat Rev Drug Discov 18 2019 797 807 10.1038/s41573-019-0034-3 31462747 5 REMAP-CAP. A Randomised, Embedded, Multi-factorial, Adaptive Platform Trial for Community-Acquired Pneumonia. https://www.remapcap.org/Last accessed August 22, 2022. 6 RECOVERY. Randomized Evaluation of COVID-19 Therapy. https://www.recoverytrial.net/Last accessed August 22, 2022. 7 WHO COVID-19 Solidarity Therapeutics Trial. https://www.who.int/emergencies/diseases/novel-coronavirus-2019/global-research-on-novel-coronavirus-2019-ncov/solidarity-clinical-trial-for-covid-19-treatments Last accessed August 22, 2022. 8 Angus D.C. Optimizing the Trade-off Between Learning and Doing in a Pandemic JAMA 323 19 2020 1895 1896 10.1001/jama.2020.4984 32227198 9 Goossens H. Derde L. Horby P. Bonten M. The European clinical research response to optimise treatment of patients with COVID-19: lessons learned, future perspective, and recommendations Lancet Infect Dis 22 5 2022 e153 e158 10.1016/S1473-3099(21)00705-2 34951954 10 Depoortere E. Sowinski S. van Hengel A. Kerstiëns B. Norstedt I. COVID-19 kick-starts a new era for clinical trials and pandemic preparedness in Europe Lancet Infect Dis 22 3 2022 315 10.1016/S1473-3099(22)00059-7 11 Granholm A. Alhazzani W. Derde L.P.G. Angus D.C. Zampieri F.G. Hammond N.E. Sweeney R.M. Myatra S.N. Azoulay E. Rowan K. Young P.J. Perner A. Møller M.H. Randomised clinical trials in critical care: past, present and future Intensive Care Med 48 2 2022 Feb 164 178 10.1007/s00134-021-06587-9 34853905 12 EU-PEARL. European Union Patient-Centric Clinical Trial Platforms. https://eu-pearl.eu/ 13 European Clinical Research Alliance on Infectious Diseases. www.ecraid.eu Last accessed August 23, 2022.	36462745	PMC9711898	NO-CC CODE	2022-12-02 23:21:31	no	Clin Microbiol Infect. 2022 Dec 1; doi: 10.1016/j.cmi.2022.11.022	utf-8	Clin Microbiol Infect	2,022	10.1016/j.cmi.2022.11.022	oa_other
==== Front J Infect J Infect The Journal of Infection 0163-4453 1532-2742 The British Infection Association. Published by Elsevier Ltd. S0163-4453(22)00683-1 10.1016/j.jinf.2022.11.022 Letter to the Editor No evidence of clinical efficacy of famotidine for the treatment of COVID-19: a systematic review and meta-analysis Cheema Huzaifa Ahmad a⁎ Shafiee Arman bc Athar Mohammad Mobin Teymouri d Shahid Abia a Awan Rehmat Ullah e Afifi Ahmed M f Shah Jaffer g Jalal Prasun K h a Division of Infectious Diseases, Department of Medicine, King Edward Medical University, Nila Gumbad Chowk, Neela Gumbad, Lahore, Punjab 54000, Pakistan b Clinical Research Development Unit, Alborz University of Medical Sciences, Karaj, Iran c Student Research Committee, School of Medicine, Alborz University of Medical Sciences, Karaj, Iran. d School of Medicine, Shahid Beheshti University of Medical Sciences, Tehran, Iran e Department of Medicine, Ochsner Rush Medical Center, Meridian, MS, United States f Department of Gastroenterology, Hepatology & Nutrition, University of Texas MD Anderson Cancer Center, Houston, TX, United States g New York State Department of Health, Albany, NY, United States h Section of Gastroenterology and Hepatology, Department of Medicine, Baylor College of Medicine, Houston, TX, United States ⁎ Corresponding author. 1 12 2022 1 12 2022 24 11 2022 © 2022 The British Infection Association. Published by Elsevier Ltd. All rights reserved. 2022 The British Infection 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. Keywords Famotidine COVID-19 SARS-CoV-2 Histamine H2-receptor antagonists ==== Body pmcDear Editor, We read with great interest the recent article by Qian et al. that reported paxlovid as an efficacious treatment for COVID-19 patients.1 However, the relatively high cost of paxlovid and the other currently available oral antiviral, molnupiravir, will likely curtail their access in lower-income countries.2 , 3 Therefore, as the quest to find efficacious and cost-effective therapies for COVID-19 patients is still ongoing, repurposing drugs, that are approved for other indications, to treat COVID-19 is an attractive option. A recent trial by Brennen Christina et al.4 showed that the use of famotidine for treating mild to moderate COVID-19 patients was associated with improved recovery without any significant adverse events. However, another study by Jiandong et al.5 reported that famotidine use is associated with a higher risk of severe COVID-19 disease. Considering the newly published studies evaluating the effectiveness of famotidine and their conflicting results, we conducted a meta-analysis to determine its potential role in treating COVID-19 patients. A systematic search was conducted on MEDLINE (via PubMed), Embase, and the Cochrane Library for relevant studies published before 15 September 2022. The inclusion criteria were comparative studies investigating famotidine for treating COVID-19 patients. Studies that used famotidine as a prophylactic agent were excluded from our analysis. The quality of the included studies was assessed using the National Institutes of Health (NIH) Quality Assessment Tool for observational cohort studies (Supplementary Table 1) and the revised Cochrane Risk of Bias Tool (RoB 2.0) for randomized controlled trials (RCTs) (Supplementary Fig. 1). The primary outcomes were all-cause mortality and the rate of no recovery. The secondary outcomes included intensive care unit (ICU) admission, time to symptom resolution, and length of hospital stay. We used RevMan 5.4 to conduct random-effects meta-analyses with odds ratios (ORs) and mean differences (MDs) as effect measures. A total of 10 studies were included in our meta-analysis (Supplementary Table 2), of which 3 were RCTs,4 , 6 , 7 and 7 were retrospective cohort studies.8, 9, 10, 11, 12, 13, 14 Our analysis showed that famotidine has no effect on mortality (OR 0.96; 95% CI: 0.34, 2.70; I2 = 100%; Fig. 1 ); the results remained non-significant in a sensitivity analysis conducted by excluding low-quality studies (OR 0.90; 95% CI: 0.74, 1.10; I2 = 66%).9, 10, 11 In terms of time to symptom resolution, a non-significant reduction was observed in the famotidine group (MD -1.86; 95% CI: -4.08, 0.36; I2 = 81%; Fig. 2 ). Famotidine use reduced the length of hospital stay (MD -1.74; 95% CI: -2.30, -1.19; I2 = 0%; Supplementary Fig. 2) but had no effect on the number of patients with no recovery (OR 0.55; 95% CI: 0.20, 1.52; I2 = 0%; Supplementary Fig. 3) and the number of patients requiring admission to ICU (OR 1.04; 95% CI: 0.86, 1.25; I2 = 0%; Supplementary Fig. 4).Fig. 1 Effect of famotidine use on all-cause mortality in COVID-19 patients. Fig 1 Fig. 2 Effect of famotidine use on time to symptom resolution. Fig 2 The idea of suppressing the immunological dysregulation caused by SARS-CoV-2 as a suggested treatment is evident in the literature.15 By producing histamine, prostaglandin D2 (PGD2), and leukotriene C4 (LTC4), it is believed that mast cells, stimulated by coronaviruses, were able to cause lung inflammation.16 Hence, famotidine, by virtue of its histamine H2-receptor (H2R) antagonism, may be involved in modifying the pulmonary pathogenic process. There is a paucity of data regarding the impact of famotidine on COVID-19 patients. Only studies examining famotidine as a treatment modality were included in our review. However, it is worth noting that several studies have also reported a beneficial prophylactic effect of this drug.17 , 18 Freedberg et al. found that the initial use of famotidine in hospitalized COVID-19 patients was significantly associated with a decreased rate of intubation and death.17 These results were further corroborated by a retrospective study by Mather Jeffrey et al.18 Among the included studies in our meta-analysis, only six reported a beneficial effect of famotidine as a potential treatment.4 , 6 , 7 , 9 , 11 , 13 It is worth noting that two studies have reported increased COVID-19 severity in the group receiving famotidine5 , 10; it is likely that the highest benefit of famotidine might be seen with use early in the course of the disease. The results of our meta-analysis for the outcome of all-cause mortality are in line with those reported by Chiu et al. and Sun et al. in their meta-analyses.19 , 20 However, these reviews only included 5 and 4 studies, respectively. Our study pooled the results from 10 studies, thereby extending the results of the previous meta-analyses with a significantly increased statistical power. Furthermore, we found that using famotidine reduced the length of hospital stay; there was also a trend towards benefit in the famotidine group for the outcomes of time to symptom resolution and the rate of no recovery. However, these estimates are based on the results of two studies each, precluding any strong conclusions, and requiring further confirmation by well-designed, large-scale RCTs which have not been conducted so far for famotidine. Another limitation of our meta-analysis is that the pooled estimates are based mostly on observational studies due to the scarcity of available RCTs; hence, they are susceptible to confounding bias. In conclusion, famotidine is not effective in reducing mortality or increasing the rate of recovery in COVID-19 patients; hence it should not be used for this indication until large-scale RCTs have established its efficacy. Financial support No financial support was received for this study. Human and animal participants Research involving human participants and/or animals: No animals or human subjects were used in the current study. Informed consents No informed consents were required for the purpose of the current study. Availability of data The data that support the findings of this study are available from the corresponding author, HAC, upon reasonable request. Declaration of Competing Interest The authors report no relationships that could be construed as a conflict of interest. Appendix Supplementary materials Image, application 1 Acknowledgements Not applicable. Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jinf.2022.11.022. ==== Refs References 1 Qian Z. Pengfei M. Mingwei W. Yongran C. Mengyun Z. Lan Y. Efficacy and safety of Paxlovid for COVID-19:a meta-analysis J Infect 2022 10.1016/j.jinf.2022.09.027 2 Junzheng W. Jacob L. Leah E. Andrew H. Minimum manufacturing costs, national prices, and estimated global availability of new repurposed therapies for Coronavirus Disease 2019 Open Forum Infect Dis 9 1 2022 10.1093/ofid/ofab581 3 Maurish F. Saleha A. Junaid S. Abia S. Ahmad C.H. Efficacy and safety of molnupiravir for COVID-19 patients Eur J Intern Med 102 2022 118 121 10.1016/j.ejim.2022.05.024 35649740 4 Brennan Christina M. Nadella S. Xiang Z. Dima R.J. Jordan-Martin N. Demestichas B.R. Oral famotidine versus placebo in non-hospitalised patients with COVID-19: a randomised, double-blind, data-intense, phase 2 clinical trial Gut 71 5 2022 879 888 10.1136/gutjnl-2022-326952 35144974 5 Jiandong Z. Xiansong W. Sharen L. Wu W.K.K. Yung C.B.M. Qingpeng Z. Proton pump inhibitor or famotidine use and severe COVID-19 disease: a propensity score-matched territory-wide study Gut 70 10 2021 2012 2013 10.1136/gutjnl-2020-323668 33277346 6 Suraksha P. Mahesh J. Aperna D. Mehak G. Darshan L. Rakesh F.N.U. Efficacy of oral famotidine in patients hospitalized with severe acute respiratory syndrome Coronavirus 2 Cureus 14 2 2022 e22404 10.7759/cureus.22404 35345695 7 Reza S.H. Hassani A.M. Maryam H. Mohsen A. Dariush H. Mitra K. The efficacy of famotidine in improvement of outcomes in hospitalized COVID-19 patients: a phase III randomised clinical trial Res Sq 2021 10.21203/RS.3.RS-462937/V1 8 Azza S. Patrick F.S. Rachel W. Berlin J.A. Patrick R. Comparative effectiveness of famotidine in hospitalized COVID-19 patients Am J Gastroenterol 116 4 2021 692 699 10.14309/ajg.0000000000001153 33982938 9 Farhana S. Nazia M. Syed Mudasir Q. Suhail M. Afshan S. Tajamul H. Efficacy of various treatment therapies on patient related outcome in hospitalized COVID-19 patients – a retrospective study Ski J Med Sci 25 2022 10.33883/jms.v25i2.1177 2 SE-Original Articles 10 Eugene S. Bara E.K. Allison H. Sylvia K. Saatchi K. Aziz T. S1301 a retrospective review: famotidine use is not associated with improved outcomes in hospitalized patients with COVID-19 Am J Gastroenterol 116 1 2021 S598 10.14309/01.ajg.0000778736.01714.cd -S598 11 Justin W. Aaron D. Nikolas S.C. Joshua F. John T. S1458 famotidine versus pantoprazole use and clinical outcomes in patients hospitalized with COVID-19: a retrospective study Am J Gastroenterol 116 1 2021 S668 S669 10.14309/01.ajg.0000779364.64867.ec 12 Samrat Y. Pratik D. Kenneth S. Mandelin C. Dax K. Gregg F. Famotidine use is not associated with 30-day mortality: a coarsened exact match study in 7158 hospitalized patients with Coronavirus Disease 2019 from a large healthcare system Gastroenterology 160 3 2021 919 921 10.1053/j.gastro.2020.10.011 e3 33058865 13 Cameron M. Saskia P. Susanne N. Max H. Bourne P.E. Robert P. Real-world evidence for improved outcomes with histamine antagonists and aspirin in 22,560 COVID-19 patients Signal Transduct Target Ther 6 1 2021 267 10.1038/s41392-021-00689-y 34262013 14 Toshiki K. Matsuo S. Mai T. Egorova N.N. The association between famotidine and in-hospital mortality of patients with COVID-19 J Med Virol 94 3 2022 1186 1189 10.1002/jmv.27375 34609001 15 Fatemeh A. Arman S. Sayeh R. Zakiye M. Mahshid S. Somayeh Y. Comparative study of CNR1 and CNR2 cannabinoid receptors expression levels in COVID-19 patients with and without diabetes mellitus: recommendations for future research targets Diabetes Metab Syndr Clin Res Rev 16 5 2022 102499 10.1016/j.dsx.2022.102499 16 Kritas S.K. Ronconi G. Caraffa A. Gallenga C.E. Ross R. Conti P. Mast cells contribute to coronavirus-induced inflammation: new anti-inflammatory strategy J Biol Regul Homeost Agents 34 1 2020 9 14 10.23812/20-Editorial-Kritas 17 Freedberg D.E. Conigliaro J. Wang T.C. Tracey K.J. Callahan M.V. Abrams J.A. Famotidine use is associated with improved clinical outcomes in hospitalized COVID-19 patients: a propensity score matched retrospective cohort study Gastroenterology 159 3 2020 1129 1131 10.1053/j.gastro.2020.05.053 e3 32446698 18 Mather Jeffrey F. Seip Richard L. McKay Raymond G. Impact of famotidine use on clinical outcomes of hospitalized patients with COVID-19 Am J Gastroenterol 115 10 2020 1617 1623 10.14309/ajg.0000000000000832 32852338 19 Leonard C. Max S. Chun-Han L. Nicholas C. Austin C. Joon S.H. Effect of famotidine on hospitalized patients with COVID-19: a systematic review and meta-analysis PLoS One 16 11 2021 e0259514 10.1371/journal.pone.0259514 20 Chenyu S. Yue C. Lei H. Yile W. Mingming L. Ahmed Mubashir A. Does famotidine reduce the risk of progression to severe disease, death, and intubation for COVID-19 patients? A systemic review and meta-analysis Dig Dis Sci 66 11 2021 3929 3937 10.1007/s10620-021-06872-z 33625613	36462586	PMC9711899	NO-CC CODE	2022-12-05 23:15:29	no	J Infect. 2022 Dec 1; doi: 10.1016/j.jinf.2022.11.022	utf-8	J Infect	2,022	10.1016/j.jinf.2022.11.022	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)01049-2 10.1016/j.cjca.2022.11.012 Clinical Research Differentiation of Multi-system Inflammatory Syndrome Associated with COVID-19 Versus Kawasaki Disease Using Cardiac Biomarkers Fridman Michael D. MD 1 Tsoukas Paul MD 23 Jeewa Aamir MD 1 Yeung Rae S.M. MD, PhD 23 Gamulka Beth D. MDCM 4 McCrindle Brian W. MD, MPH 1# 1 Department of Pediatrics, University of Toronto, Labatt Family Heart Centre, Hospital for Sick Children, Toronto, Ontario, Canada 2 Department of Pediatrics, University of Toronto, Division of Rheumatology, Hospital for Sick Children, Toronto, Ontario, Canada 3 Department of Immunology and Institute of Medical Science, University of Toronto, Toronto, ON, Canada 4 Department of Pediatrics, University of Toronto, Division of Paediatric Medicine, Hospital for Sick Children, Toronto, Ontario, Canada # Corresponding Author: Dr. Brian McCrindle, Labatt Family Heart Centre – Department of Pediatrics, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada, M5G 1X8, Phone 416-813-7654 ext207609. 1 12 2022 1 12 2022 21 10 2022 17 11 2022 27 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. BACKGROUND Multisystem inflammatory syndrome in children (MIS-C) after COVID-19 shares clinical similarities to Kawasaki disease (KD). We sought to determine whether cardiac biomarker levels differentiate MIS-C from KD and their association with cardiac involvement. METHODS Subjects included 38 MIS-C patients with confirmed prior CoVID-19 and 32 pre-pandemic and 38 contemporaneous KD patients with no evidence of COVID-19. Patient, clinical, echocardiographic, electrocardiographic and laboratory data timed within 72 hours of cardiac biomarker assessment were abstracted. Groups were compared, and regression analyses were used to determine associations between biomarker levels, diagnosis and cardiac involvement, adjusting for clinical factors. RESULTS MIS-C patients had fewer KD clinical features, with more frequent shock, ICU admission, inotrope requirement, and ventricular dysfunction, with no difference regarding coronary artery involvement. Multivariable regression analysis showed that both higher N-terminal pro-B-type natriuretic peptide (NTproBNP) and cardiac troponin I (TnI) were associated with MIS-C versus KD, after adjusting for significant covariates. Receiver operating characteristic curves for diagnosis showed that any detectable TnI greater than 10 ng/L was predictive of MIS-C versus KD with 91% sensitivity and 76% specificity. NTproBNP >2000 ng/L predicted MIS-C versus KD with 82% sensitivity and 82% specificity. Higher TnI but not NTproBNP was associated with lower LV ejection fraction; neither biomarker was associated with coronary artery involvement. CONCLUSIONS Positive TnI and higher NTproBNP may differentiate MIS-C from KD, which may become more relevant as evidence of prior COVID-19 becomes more challenging to determine. Cardiac biomarkers may have limited associations with cardiac involvement in this setting. ==== Body pmcIntroduction: Early in the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, children were thought to have asymptomatic or less severe illness when infected. However, it soon became evident that some children were manifesting a novel and delayed multisystem inflammatory syndrome (MIS-C) after infection.1 , 2 The clinical presentation of MIS-C resembles Kawasaki disease (KD). However, it became clear that MIS-C was clinically distinct, resulting in the development of a case definition from the World Health Organization (WHO).3 The WHO case definition was understandably broad and focused primarily on the major organ systems that seem to be affected by MIS-C – specifically, cardiac, hematologic (coagulopathy), and gastrointestinal. The WHO and subsequent other case definitions have had considerable overlap with the case definition for KD.4 In addition to echocardiography and cardiac magnetic resonance imaging (MRI), cardiac biomarkers, specifically cardiac troponin I (TnI) and amino-terminal pro-B-type natriuretic peptide (NTproBNP), have been used as biomarkers of cardiac involvement. For adults, both TnI and NT-proBNP have been validated for diagnosis and monitoring of ischemic cardiac injury5 and heart failure.6 Their role in the diagnosis and management of KD has been variable,7 and they are not generally recommended.8 However, abnormalities of both NTproBNP and TnI have been described in the setting of MIS-C.9 Due to the similarities between MIS-C and KD in their clinical presentation and course, and possibly pathophysiology,10, 11, 12 these cardiac biomarkers have more recently been assessed in children and adolescents in this clinical setting.13 However, there is limited understanding as to their role in diagnosis, monitoring and prognosis of MIS-C versus KD. Therefore, we sought to determine 1) how elevations of TnI and NTproBNP are associated with clinical, biochemical, electrocardiographic, and/or echocardiographic changes in patients with KD versus MIS-C, and 2) how TnI or NTproBNP may help differentiate KD from MIS-C. Methods: Study Design and Population: This was a single-centre retrospective cohort study approved by the institutional Research Ethics Board under a waiver of consent. Three groups of patients were reviewed: 1) Patients with complete and incomplete KD diagnosed before January 1, 2020 as per the American Heart Association (AHA) 2017 guideline criteria8 (PreKD) and who had a stored blood samples obtained prior to treatment with intravenous immunoglobulin (IVIG) from which NTproBNP could be assessed (insufficient sample volume to assess TnI); 2) Patients with complete and incomplete KD diagnosed between January 1, 2020 and March 31, 2021 who had no evidence of SARS-CoV-2 exposure and do not meet criteria for MIS-C as per the WHO case definition (PostKD); and patients from January 1, 2020 to March 31, 2021 who were diagnosed with MIS-C as per WHO criteria with confirmed preceding SARS-CoV-2 (any of test-confirmed household exposure, positive PCR, positive serology). For PreKD patients, biospecimens were obtained via SickKids COVID Biobank (REB# 1000070060, BEAT KD REB# 1000043263). All patients in the PostKD and MIS-C groups who had bloodwork including assessment of NTproBNP and TnI and who had an echocardiogram within 72 hours of that assessment were included. Measurements: Information regarding demographics, clinical presentation and findings, laboratory and cardiac imaging assessments, electrocardiography (ECG), infectious disease assessment, management and outcomes were abstracted from the medical record. Biosamples were obtained in treatment naïve patients prior to administration of immunomodulatory therapy in all patient groups. Standardized and uniform preanalytical operating processes were applied to all sample collections, processing and storage. Plasma was obtained from P100 blood collection tubes with spray dried anticoagulant (K2EDTA) and proprietary protease inhibitors (Becton and Dickson (BD) Bioscience) as per the manufacturer’s protocol. Once processed, samples were stored at -80°C in 200 μL aliquots at the SickKids COVID-19 biobank at The Hospital for Sick Children (Toronto, ON). The Abbott Architect analyzer was used to test for high-sensitivity Troponin I and NTproBNP. Level of detection for NTproBNP was 5 ng/L with a measuring range of 5–35 000 ng/L, and for data analysis purposes levels >35,000 were arbitrarily set to 35,001. Level of detection for TnI was 10 ng/L, and for data analysis purposes levels <10 were arbitrarily set to 0.001. Institutional normal values for TnI were <10 ng/L and for NTproBNP were <125ng/L. In the event that there were multiple assessments of TnI and NTproBNP on the same day as an echocardiogram, the most abnormal value was selected for analysis. From the ECGs obtained closest to the blood sample, basic intervals (PR, QRS duration, QT, QTc), heart rate, presence of arrhythmia, presence of T-wave abnormalities, and index of cardiac electrophysiologic balance (QT or QTc / QRSd) were collected from finalized ECG reports. Ventricular functional indices and quantitative and qualitative assessment of the coronary arteries were abstracted from echocardiogram reports. All echocardiograms were obtained using a standardized institutional protocol. Coronary artery dimensions were converted to body surface area-adjusted Z scores,14 and coronary artery involvement was classified using the criteria of the AHA guidelines.8 As detailed time course of body temperature changes were not available, for practical purposes intravenous immunoglobulin (IVIG) resistance was defined by the patient having received after an initial dose of IVIG a second dose of IVIG or second-line therapies (e.g. immunomodulators). If IVIG and corticosteroids were given concomitantly as initial therapy, IVIG resistance was then defined by the patient having received subsequent second dose of IVIG or second-line therapies. Data Analysis: Data are described as frequencies, means with standard deviations, and medians with values at the 5th and 95th percentile, as appropriate to the level of measurement and the distribution of values for a given variable. Comparisons of features between PreKD and PostKD groups and between all KD and MIS-C groups were performed. Categorical variables between groups were compared using Fisher’s Exact Test and Mantel-Hanzel chi-squared analysis. Normally distributed variables were compared using Student’s t-test. Highly skewed continuous variables were compared using Kruskal-Wallis ANOVA. Statistical significance was set at a p<0.05. Clinical status, laboratory values and cardiac assessment closest to the time of measurement of TnI and NTproBNP, though no greater than 72 hours away, were compared between diagnostic groups using generalized linear regression models. Since the distribution of values of cardiac biomarkers was highly skewed, the logarithm of the values was used to normalize the distribution and used in all regression analyses. SAS statistical software Version 9.4 (Cary, NC) was used for analyses with p<0.05 set as the threshold for statistical significance. RESULTS Comparison of Kawasaki disease patients before and after January 1, 2020 There were 32 patients with KD included from before January 1, 2020 (PreKD) and 36 patients after January 1, 2020 (PostKD). Details and comparison of their characteristics are summarized in Table 1 . PreKD patients were more likely to have complete KD criteria with a greater number of diagnostic features than PostKD patients. The likelihood of admission to the intensive care unit (ICU) was lower for preKD patients, the length of stay in both the intensive care unit (ICU) and hospital overall was shorter f, and their bloodwork at cardiac biomarker assessment showed higher white blood cell count and alanine aminotransferase (ALT) levels. From a cardiac perspective, there were no significant differences between PreKD and PostKD patients regarding NT-ProBNP levels, or findings from electrocardiography and echocardiography. PostKD patients were more likely to have received enteral corticosteroids steroids than PreKD patients, though the proportion with IVIG resistance was similar between groups.Table 1 Demographic, biochemical, and cardiologic characteristics of pre- and post-pandemic patients with acute Kawasaki disease. Pre-pandemic Post-pandemic P value (n = 32) (n = 36) Demographics Age, years 3.3 (1.2 - 11.5) 3 (0.4 - 15.8) >0.99 Sex, male/female 20/12 21/15 0.73 Body mass index, kg/m2 15.5 ± 1.7 18.2 ± 6.6 0.03 Body mass index Z-score -0.6 ± 1.6 -0.1 ± 1.6 0.19 Clinical Course Days of fever 7 (4 - 12) 6 (3-18) 0.44 Number of Kawasaki disease clinical features 5 (3 - 5) 3 (1 - 5) <0.001 Conjunctivitis 31 (97%) 27 (75%) 0.01 Cervical lymphadenopathy 20 (63%) 11 (31%) 0.008 Rash 30 (94%) 22 (61%) 0.001 Extremity changes 30 (94%) 22 (61%) 0.001 Oral mucosal changes 30 (94%) 28 (78%) 0.06 Kawasaki disease diagnosis <0.001 Complete 28 (88%) 11(31%) Incomplete 4 (12%) 25 (69%) Shock 1 (3%) 4 (11%) 0.2 Intensive care requirements Admission to ICU 1 (3%) 3 (8%) 0.36 Length of stay, days 0 (0 - 1) 3 (2 - 5) <0.001 Inotropic support 0 0 >0.99 Total hospital length of stay, days 3 (2 - 7) 4 (3 - 18) 0.002 Treatment Days from admission to first immunomodulatory treatment 0 (0, 1) 1 (1, 5) <0.001 Intravenous immunoglobulin 32 (100%) 31 (86%) 0.03 Corticosteroids, intravenous 2 (6%) 11 (31%) 0.01 Corticosteroids, oral 3 (9%) 20 (56%) <0.001 Immunomodulating agents 0 2 (6%) 0.17 Intravenous immunoglobulin resistance 9 (28%) 3 (8%) 0.06 Biochemistry Days from admission to assessment 0 (-1, 0) 1 (0, 7) <0.001 Troponin I, ng/L - 0 (0 - 88.5) - N-terminal pro-B-type natriuretic peptide, ng/L 718 (35 - 8615) 562 (41 - 14573) 0.99 C-reactive protein, mg/L 94.4 (3.7 - 288) 93.7 (7.9 - 381.7) 0.72 Ferritin, mcg/L - 182.1 (41.2 - 6936) - Erythrocyte sedimentation rate, mm/hr 73 ± 28 59 ± 33 0.06 Hemoglobin, g/L 112 ± 12 108 ± 12 0.18 Mean corpuscular volume, fL 78.6 ± 5.9 79 ± 5 0.76 White blood cells x10ˆ9/L 14.9 ± 6 11.6 ± 5.5 0.02 Platelets x10ˆ9/L 308 ± 96 311 ± 146 0.94 Creatinine, micromol/L 29 ± 8 33 ± 26 0.46 Albumin, g/L 37 ± 5 34 ± 8 0.17 Alanine aminotransferase, U/L 39 (10 - 281) 26 (12 - 252) 0.04 International Normalized Ratio - 1.3 ± 0.2 - Partial thromboplastin time, seconds - 32 ± 5 - Echocardiography Days from admission to echocardiogram 1 (0, 4) 2 (0, 4) 0.43 Days from cardiac biomarkers assessment to echocardiogram -1.5 (-4, 0) 1 (-4, 8) 0.006 Coronary Artery Parameters Left main coronary artery Z-score 0.43 (-0.6 - +2.64) 0.16 (-1.35 - +3.52) 0.3 Left anterior descending artery Z-score 0.36 (-0.97 - +2.68) 0.34 (-0.95 - +4.07) 0.85 Right coronary artery Z-score 0.88 (-0.87 - +2.07) 0.94 (-0.45 - +5.12) 0.61 Maximum coronary artery Z-score 1.33 (-0.86 - +3.19) 1.17 (-0.01 - +9.77) 0.67 Severity of coronary artery abnormality 0.94 Normal dimensions 22 (68%) 26 (74%) Dilation only (Z-score 2 - <2.5) 5 (16%) 4 (11%) Small aneurysms (Z-score ≥ 2.5 - <5) 5 (16%) 3 (9%) Medium aneurysms (Z-score ≥5 - <10 and <8mm) 0 1 (3%) Large/giant aneurysms (Z-score ≥10 or ≥8mm) 0 1 (3%) Not well visualized 0 1 (3%) Ventricular Parameters LV End-Diastolic Dimension Z-score 1 (-1.5 - +2.6) 0.8 (-2.6 - +2.6) 0.37 Fractional shortening, % 35 ± 4 36 ± 5 0.5 LV ejection fraction, % 66 ± 6 67 ± 8 0.67 RV dysfunction 0 1 (3%) 0.36 LV dysfunction 0 2 (6%) 0.19 Electrocardiography Non-sinus rhythm, No. (%) 0 1 (3%) 0.35 HeartrRate, bpm 131 ± 26 129 ± 33 0.72 PRiInterval, ms 121 ± 18 124 ± 21 0.71 QRS duration, ms 69 ± 9 71 ± 14 0.32 QT interval, ms 289 ± 39 292 ± 53 0.79 Corrected QT interval, ms 418 ± 23 418 ± 26 0.98 Index of cardioelectrophysiologic balance 4.2 ± 0.5 4.1 ± 0.6 0.5 Corrected index of cardioelectrophysiologic balance 6.2 ± 0.8 6 ± 1.1 0.57 T-wave abnormality 16 (50%) 12 (33%) 0.16 Continuous variables are displayed as either mean ± standard deviation or median (5th, 95th percentile). ICU, intensive care unit; LV, left ventricle; RV, right ventricle Comparison of Kawasaki disease and MIS-C patients Given the relative similarities between the PreKD and PostKD groups, they were combined for comparison to the MIS-C patients. In total, 68 KD and 38 MIS-C patients were included. Details and comparison of these two groups are shown in Table 2 . The KD group was younger than the MIS-C group, with lower body mass index (BMI) by both absolute value and Z-score. KD patients had a median of 4 KD clinical criteria, whereas MIS-C patients had a median of 1.5 criteria, with 53% of MIS-C patients also meeting criteria for diagnosis of complete or incomplete KD. In both populations, conjunctivitis was the most common KD feature, followed by non-vesicular rash, oral mucosal changes and extremity changes. Lymphadenopathy was an infrequent finding for MIS-C patients.Table 2 Demographic, biochemical, and cardiologic characteristics of patients with Kawasaki disease (KD) compared to multisystem inflammatory syndrome in children (MIS-C). Kawasaki Disease MIS-C P value (n = 68) (n = 38) Demographics Age, y 3.2 (0.4 - 15.2) 9.1 (0.8 - 17) <0.001 Sex, male/female 41/27 24/14 0.77 Body mass index, kg/m2 16.9 ± 5.1 20.1 ± 5.2 0.002 Body mass index Z-score -0.3 ± 1.6 0.7 ± 1.1 0.001 Clinical Course Days of fever 7 (3 - 14) 6 (3 - 14) 0.11 Number of Kawasaki disease clinical features 4 (1 - 5) 1.5 (0 - 4) <0.001 Conjunctivitis 58 (85%) 20 (53%) <0.001 Cervical lymphadenopathy 31 (46%) 1 (3%) <0.001 Rash 52 (76%) 18 (47%) 0.002 Extremity changes 52 (76%) 9 (24%) <0.001 Oral mucosal changes 58 (85%) 10 (26%) <0.001 Kawasaki disease diagnosis <0.001 Complete 39 (57%) 3 (8%) Incomplete 29 (43%) 15 (40%) Does not meet criteria 0 20 (53%) Shock 5 (7%) 15 (40%) <0.001 Intensive care unit Admission to intensive care unit 4 (6%) 18 (47%) <0.001 Length of stay, days 0 (0 - 3) 2 (1 - 9) <0.001 Inotropic support 0 13 (34%) <0.001 Total hospital length of stay, days 4 (3 - 10) 6 (3 - 13) 0.003 Treatment Days from admission to first immunomodulatory treatment 1 (0, 1) 1 (0, 3) 0.15 Intravenous immunoglobulin 63 (90%) 32 (80%) 0.18 Corticosteroids, intravenous 13 (19%) 26 (68%) <0.001 Corticosteroids, oral 23 (34%) 33 (87%) <0.001 Immunomodulatory agents 2 (3%) 2 (5%) 0.56 Intravenous immunoglobulin resistance 12 (18%) 8 (21%) 0.56 Biochemistry Days from admission to cardiac biomarker assessment 0 (-1, 4) 1 (0, 5) <0.001 Troponin I, ng/L <10 (<10 - 88.5) 61 (<10 - 11052) <0.001 N-terminal pro-B-type natriuretic peptide, ng/L 634 (41 - 10752) 4948 (107 - 27313) <0.001 C-reactive protein, mg/L 94.4 (6.7 - 342.3) 109.1 (15.9 - 517.6) 0.12 Ferritin, mcg/L 195.6 (41.7 - 3046.4) 512.3 (113.3 - 3137.5) <0.001 Erythrocyte sedimentation rate, mm/hr 66 ± 31 59 ± 37 0.34 Hemoglobin, g/L 110 ± 12 103 ± 18 0.01 Mean corpuscular volume, fL 78.8 ± 5.4 79.4 ± 3.8 0.53 White blood cells x10ˆ9/L 13.2 ± 5.9 13.1 ± 9.2 0.98 Platelets x10ˆ9/L 310 ± 124 223 ± 113 <0.001 Creatinine, micromol/L 31 ± 20 50 ± 40 0.002 Albumin, g/L 35 ± 7 29 ± 6 <0.001 Alanine aminotransferase, U/L 30 (12 - 247) 30 (13 - 134) 0.13 International Normalized Ratio 1.3 ± 0.2 1.2 ± 0.2 0.36 Partial thromboplastin time, seconds 32 ± 5 29 ± 5 0.07 Echocardiography Days from admission to echocardiogram 2 (0, 4) 1 (1, 6) 1 Days from cardiac biomarkers assessment to echocardiogram -1 (-4, 2) 0 (-5, 3) <0.001 Coronary Artery Parameters Left main coronary artery Z-score 0.28 (-1.15 - +2.97) 0.96 (-1.2 - +2.42) 0.46 Left anterior descending artery Z-score 0.34 (-0.95 - +3.19) 0.79 (-1.03 - +2.81) 0.15 Right coronary artery Z-score 0.91 (-0,74 - +3.37) 0.64 (-0.68 - +2.2) 0.56 Maximum coronary artery Z-score 1.24 (-0.01 - +3.66) 1.41 (0.05 - +2.81) 0.85 Severity of coronary artery abnormality 0.13 Normal dimensions 48 (72%) 29 (76%) Dilation only (Z-score 2 - <2.5) 9 (13%) 6 (16%) Small aneurysms (Z-score ≥ 2.5 - <5) 8 (12%) 3 (8%) Medium aneurysms (Z-score ≥5 - <10 and <8mm) 1 (1.5%) 0 Large/giant aneurysms (Z-score ≥10 or ≥8mm) 1 (1.5%) 0 Not measured/visualized 1 (1.5%) 0 Ventricular Parameters LV end-diastolic dimension Z-score 0.9 (-1.7 - +2.5) 0.5 (-2 - +1.9) 0.18 Fractional shortening, % 36 ± 5 33 ± 7 0.04 LV ejection fraction, % 66 ± 7 61 ± 10 0.004 RV dysfunction 1 (2%) 5 (13%) <0.001 LV dysfunction 2 (3%) 9 (24%) <0.001 Electrocardiography Non-sinus rhythm 1 (2%) 1 (3%) 0.68 Heart rate, bpm 130 ± 30 124 ± 25 0.33 PR interval, ms 123 ± 20 143 ± 28 <0.001 QRS duration, ms 70 ± 12 78 ± 12 0.005 QT interval, ms 290 ± 47 304 ± 48 0.15 Corrected QT interval, ms 418 ± 24 429 ± 35 0.07 Index of cardioelectrophysiologic balance 4.2 ± 0.6 4 ± 0.4 0.04 Corrected index of cardioelectrophysiologic balance 6.1 ± 1 5.6 ± 0.7 0.01 T-wave abnormality, No. (%) 28 (41%) 22 (58%) 0.07 Continuous variables are displayed as either mean ± standard deviation or median (5th, 95th percentile). Regarding severity of illness, MIS-C patients were more likely to have shock and admission to ICU with need for inotropic support. MIS-C patients had both longer median ICU and total hospital lengths of stay. Regarding treatment, there was no difference in the use of IVIG or immunomodulators. MIS-C patients were more likely to have received both enteral and intravenous corticosteroids. There was no difference regarding the prevalence of IVIG resistance between the groups. Cardiac involvement in MIS-C versus Kawasaki disease Patients with MIS-C had higher median TnI, NTproBNP and ferritin levels, with lower hemoglobin, platelet counts, and albumin levels. There were no differences in the markers of general inflammation (erythrocyte sedimentation rate [ESR], C-reactive protein [CRP]). The MIS-C group had higher creatinine than the KD group (50±40 vs 31±20 mcmol/L, p<0.01), though this may be related to the differences in age and body size between the groups. Regarding echocardiography, patients with MIS-C were more likely to have at least mild ventricular dysfunction for both the right and left ventricles, both qualitatively and quantitatively, as noted by reductions in both left ventricular (LV) fractional shortening and ejection fraction. Notably, there were no statistically significant differences regarding coronary artery Z-scores between the KD and MIS-C groups. There was a trend towards a greater number and severity of coronary artery aneurysms for the KD cohort, although this did not reach statistical significance. From electrocardiography, heart rates for the KD and MIS-C groups were similar despite age differences, suggesting perhaps a relative tachycardia for the MIS-C patients. Differences in PR interval, QRS duration, iCEB, and iCEBc were likely age-related and appropriate. There was a trend towards more non-specific T-wave abnormalities for the MIS-C group, which might be reflective of cardiac involvement in MIS-C, although this did not reach statistical significance. Associations of cardiac biomarkers with diagnosis From a multivariable linear regression model including patient characteristics, laboratory features, ECG findings, and diagnosis group (Table 3 ), higher log(NTproBNP) was independently associated with a diagnosis of MIS-C vs KD after controlling for significant factors of male sex, lower albumin, higher heart rates and longer PR intervals. Using logistic regression for diagnosis group, a receiver-operating characteristic curve was derived (Figure 1 A). From this, an NTproBNP >2000 ng/L predicted a diagnosis of MIS-C vs KD with a sensitivity of 84% and specificity of 79%, and an overall area under the curve of 0.84.Table 3 Patient, clinical, laboratory and electrographic factors associated with log(NTproBNP).* Parameter Estimate Variable (Standard error) P value Clinical, Demographic, Biochemical, and ECG Parameters Intercept 2.53 (0.66) Male sex -0.27 (0.12) 0.03 Diagnosis of MIS-C (vs KD) 0.54 (0.14) <0.001 Lower albumin (per g/L) -0.044 (0.009) <0.001 Higher heart rate (per bpm) 0.009 (0.002) <0.001 Longer PR Interval (per msec) 0.006 (0.003) 0.05 Multivariable linear regression model; model R2 0.49. Figure 1 Receiver operating characteristic curves for differentiating multisystem inflammatory syndrome in children (MIS-C) from Kawasaki disease for A. NTproBNP and B. Troponin I., AUC, area under the curve; ROC, receiver operating characteristic TnI was measured in all PostKD and MIS-C patients (n=73). For this combined group, 41 patients (56%) had no detectable TnI. Of note, log(TnI) and log(NTproBNP) were significantly correlated (r=0.60; p<0.001). A multivariable linear regression analysis for log(TnI) included patient characteristics, laboratory features, ECG findings and diagnosis group (Table 4 ). Higher log(TnI) was independently associated with a diagnosis of MIS-C versus KD after controlling for significant factors of lower platelets, requirement of ICU care, longer QRS duration, and any T-wave abnormality. Using logistic regression for diagnosis group, a receiver-operating characteristic curve was derived (Figure 1B). From this, the presence of any detectable TnI (TnI >10 ng/L) predicted a diagnosis of MIS-C vs KD with a sensitivity of 88% and specificity of 78%, with an overall area under the curve of 0.83. Of note, when both NTproBNP and TnI were included in a logistic regression model, the area under the curve only improved to 0.87.Table 4 Patient, clinical, laboratory and electrocardiographic factors associated with log(TnI). Variable Parameter estimate (Standard error) P value Clinical, Demographic, Biochemical, and ECG Parameters Intercept -7.82 (1.64) Diagnosis of MIS-C (vs KD) 2.32 (0.5) <0.001 Intensive care unit admission 1.16 (0.61) 0.06 Lower platelet count (per x10ˆ9/L) -0.0044 (0.002) 0.03 Longer QRS duration (per msec) 0.075 (0.018) <0.001 Any non-specific T-wave abnormality 1.37 (0.47) 0.009 Multivariable linear regression model; model R2 0.67. Associations of cardiac biomarkers with left ventricular function NTproBNP was not significantly associated with LV ejection fraction at the time of assessment, either with or without diagnosis in the regression model, and there was no significant interaction between NTproBNP and diagnosis. However, higher TnI was significantly associated with lower LV ejection fraction, but only as log(TnI). Similar results were obtained both with and without diagnosis in the model, and there was no significant interaction with diagnosis. Associations of cardiac biomarkers with coronary artery involvement NTproBNP was not significantly associated with maximum coronary artery Z score in any branch at the time of assessment, either with or without diagnosis in the regression model, and there was no significant interaction between NTproBNP and diagnosis. Similar results were obtained using log(NTproBNP). Likewise, there was no significant association with TnI or log(TnI). Discussion Multisystem inflammatory syndrome in children associated with SaRS-CoV-2 (MIS-C) is a novel syndrome that has rapidly been characterized through the course of the pandemic. Cardiovascular involvement is one hallmark of the syndrome with a variety of presentations including depressed ventricular function, cardiogenic shock, myocarditis, and coronary artery abnormalities in the form of dilation and aneurysms.13 , 15 While the diagnostic criteria posed by the WHO have been helpful in raising suspicion for the disease, unless there is evidence confirming an association with COVID-19, they have limited ability to differentiate MIS-C from KD or provide prognostic value. In this study, cardiac biomarkers were identified as a useful tool to help differentiate KD from MIS-C. With the presence of any positive TnI (TnI >10 ng/L) or NTproBNP >2000 ng/L, MIS-C is the more likely diagnosis. Thus, determination of these cardiac biomarkers may be useful in the clinical setting for decision making, evaluation and management. Few studies have compared contemporaneous KD and MIS-C patients, or pre-pandemic and post-pandemic KD patients. Our study further confirms important clinical similarities and differences between MIS-C and KD patients, with MIS-C patients having higher median age with a greater risk of shock and ICU admission. Further, we found few clinical differences between KD patients immediately prior to versus during the pandemic. Biochemically, compared to KD patients, MIS-C patients had greater elevations in TnI, NTproBNP, and creatinine, along with thrombocytopenia, lower albumin, and PTT. Although the frequency of coronary artery abnormalities was similar, MIS-C patients were less likely to have medium or large coronary artery aneurysms, but were more likely to have greater degrees of dilation. Lastly, LV ejection fraction and qualitative biventricular functional variables were all lower in MIS-C versus KD, with no significant differences in LV dimensions. Features of histological myocarditis have been reported for patients with acute KD.16 The majority of these patients, as in our study, are minimally symptomatic or have no evidence of functional impairment on echocardiography. In a recent study by Desjardins et al., patients with KD with elevations in NTproBNP were more likely to have transient echocardiographic features of presumed mild myocarditis.17 Greater biochemical evidence of myocarditis noted with MIS-C is associated with observed greater clinical and functional cardiac abnormalities. Matsubara et al. in a small retrospective single institution case series noted similar findings in a comparison of MIS-C and KD patients, together with more subtle abnormalities of strain and diastolic dysfunction in the MIS-C patients.18 The fact that elevations in TnI and NTproBNP may contribute to differentiation of MIS-C from KD, with TnI being the more sensitive marker in adjusted regression models and associated with high sensitivity and specificity, may help inform diagnosis and clinical decision-making. From a clinical perspective, there have been many studies characterizing the clinical differences in patients with MIS-C and KD. In New York State, the initial MIS-C series highlighted a very severe inflammatory syndrome with a high prevalence of myocarditis.1 In their cohort, 71% of the patients with confirmed MIS-C had positive troponins and 80% were admitted to the ICU. Several months later, Kaushik et al. published the experience from New York City alone which clarified the broader spectrum of disease,19 which was echoed in more recent data from the US Centers for Disease Control and Prevention.20 In the United Kingdom, severe disease appeared to be less common, but the cohort was skewed to younger age groups which was associated with less severe illness overall.21 KD has always been strictly a clinical diagnosis, with heterogeneity in terms of identifying a clear cause or etiologic mechanism.8 , 15 Some patients with MIS-C clearly have a KD phenotype, and thus there has been speculation from the very first descriptions of MIS-C as to whether the two conditions shared pathophysiology.11 Nonetheless, KD and MIS-C appear to be somewhat biochemically and clinically distinct.22 , 23 As such, these biomarkers may become useful in combination with other features to further clarify the pathophysiologic mechanism and to aide in both differentiation and prognostication. Limitations There were several potential limitations to this study. The first was the lack of samples for measurement of TnI pre-pandemic for the KD patients, lowering the overall power of the study and eliminating comparison between pre-pandemic and during-pandemic KD patients. Furthermore, there is a possible presence of selection bias in the PreKD population as there may have been certain factors that would have increased the likelihood of their participation, whereas the PostKD population was a more representative sample of the general population at the time based on any presentation to the emergency department. In addition, the PreKD patients had no evidence of MAS or KD shock, which is likely representative of both the small patient numbers and the aforementioned possible selection bias for this group. Given that this may be a less sick population than a less-selected population, the impact on known acute phase reactants (anemia, thrombocytopenia, hypoalbuminemia, and hyperferritinemia) may have been dampened. Similarly, it has been reported in the literature that NTproBNP in the setting of inflammatory disease is an acute phase reactant, thereby suggesting that it may be more of a general marker of inflammation rather than a specific marker of cardiac involvement in this scenario.24 , 25 Further, although strict case definition criteria were applied, we cannot exclude that some of the earlier postKD patients actually had MIS-C, since early in the pandemic PCR and serologic testing were not widely available to exclude prior COVID-19 infection. Another key limitation was the age differences in patients between the cohorts. As age is collinear with many ECG parameters, it is unclear if the measured ECG duration differences have any clinical utility. Finally, there were often differences in the timing of echocardiography and collection of serum for biomarker testing. The window of 72 hours allowed for the capture of the majority of patients, but several patients were excluded, usually those with milder disease, though this was similar between the KD and MIS-C cohorts. As this was a standard practice, it is unlikely to have substantially impacted the results but may have resulted in missing more acute changes. Conclusions: In conclusion, this study revealed that both elevated and greater elevations of TnI and NTproBNP were associated with a diagnosis of MIS-C rather than KD with good sensitivity and specificity. Elevations in either biomarker were associated with male sex and were variably associated with other biochemical markers. Although, neither TnI nor NTproBNP were useful in detecting risk of coronary artery abnormalities, it was clear that the KD population had a higher risk of developing coronary artery aneurysms compared to the MIS-C group, while the MIS-C patients were much more likely to present with shock, dysfunction, and require intensive care. Future directions for this work include determining if these biomarkers are capable of predicting clinical course, allowing for preemptive optimization of management prior to the development of critical illness. Long-term data will also be needed to determine how myocardial and coronary artery involvement differs between the KD and MIS-C populations. Future studies regarding the different and shared pathophysiology of KD and MIS-C might consider the mechanisms for differences in these cardiac biomarkers. Acknowledgements Special thanks to the team of the International Kawasaki Disease Registry (IKDR) for their assistance in identifying local patients with Kawasaki Disease for analysis. Sources of Funding P.T. is supported by the Clinician-Scientist Training Program at The Hospital for Sick Children. R.S.M.Y. is supported by the Hak-Ming and Deborah Chiu Chair in Paediatric Translational Research, Hospital for Sick Children, University of Toronto. Competing Interests None of the authors have any conflicts of interest to disclose. Summary: Multi-system inflammatory syndrome in children (MIS-C) associated with prior COVID-19 has clinical overlap with Kawasaki disease (KD), including cardiac involvement, and cardiac biomarkers may help to differentiate. In a comparison of contemporaneous patients, MIS-C patients had higher levels of NTproBNP and were more likely to have detectable troponin I. Cardiac biomarkers differentiated MIS-C from KD with acceptable sensitivity and specificity, with higher troponin I also being a marker of ventricular dysfunction. ==== Refs References 1 Dufort E.M. Koumans E.H. Chow E.J. Multisystem Inflammatory Syndrome in Children in New York State The New England journal of medicine 383 2020 347 358 32598830 2 Feldstein L.R. Rose E.B. Horwitz S.M. Multisystem Inflammatory Syndrome in U.S. Children and Adolescents The New England journal of medicine 383 2020 334 346 32598831 3 Multisystem inflammatory syndrome in children and adolescents temporally related to COVID-19. World Health Organization, 2020. (Accessed October, 2022, at https://www.who.int/news-room/commentaries/detail/multisystem-inflammatory-syndrome-in-children-and-adolescents-with-covid-19.) 4 Hoste L. Van Paemel R. Haerynck F. Multisystem inflammatory syndrome in children related to COVID-19: a systematic review European journal of pediatrics 180 2021 2019 2034 33599835 5 Bularga A. Lee K.K. Stewart S. High-Sensitivity Troponin and the Application of Risk Stratification Thresholds in Patients With Suspected Acute Coronary Syndrome Circulation 140 2019 1557 1568 31475856 6 Nieminen M.S. Bohm M. Cowie M.R. Executive summary of the guidelines on the diagnosis and treatment of acute heart failure: the Task Force on Acute Heart Failure of the European Society of Cardiology Eur Heart J 26 2005 384 416 15681577 7 Dionne A. Dahdah N. A Decade of NT-proBNP in Acute Kawasaki Disease, from Physiological Response to Clinical Relevance Children (Basel) 5 2018 8 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 2017 e927 e999 28356445 9 Tajbakhsh A. Gheibi Hayat S.M. Taghizadeh H. COVID-19 and cardiac injury: clinical manifestations, biomarkers, mechanisms, diagnosis, treatment, and follow up Expert Rev Anti Infect Ther 19 2021 345 357 32921216 10 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 2020 246 248 32511667 11 Consiglio C.R. Cotugno N. Sardh F. The Immunology of Multisystem Inflammatory Syndrome in Children with COVID-19 Cell 183 2020 968 981 e7 32966765 12 Ghosh P. Katkar G.D. Shimizu C. An Artificial Intelligence-guided signature reveals the shared host immune response in MIS-C and Kawasaki disease Nature communications 13 2022 2687 13 Clark B.C. Sanchez-de-Toledo J. Bautista-Rodriguez C. Cardiac Abnormalities Seen in Pediatric Patients During the SARS-CoV2 Pandemic: An International Experience Journal of the American Heart Association 9 2020 e018007 14 McCrindle B.W. Li J.S. Minich L.L. Coronary artery involvement in children with Kawasaki disease: risk factors from analysis of serial normalized measurements Circulation 116 2007 174 179 17576863 15 Kabeerdoss J. Pilania R.K. Karkhele R. Kumar T.S. Danda D. Singh S. Severe COVID-19, multisystem inflammatory syndrome in children, and Kawasaki disease: immunological mechanisms, clinical manifestations and management Rheumatology international 41 2021 19 32 33219837 16 Fujiwara H. Hamashima Y. Pathology of the heart in Kawasaki disease Pediatrics 61 1978 100 107 263836 17 Desjardins L. Dionne A. Meloche-Dumas L. Fournier A. Dahdah N. Echocardiographic Parameters During and Beyond Onset of Kawasaki Disease Correlate with Onset Serum N-Terminal pro-Brain Natriuretic Peptide (NT-proBNP) Pediatr Cardiol 41 2020 947 954 32172336 18 Matsubara D. Kauffman H.L. Wang Y. Echocardiographic Findings in Pediatric Multisystem Inflammatory Syndrome Associated With COVID-19 in the United States J Am Coll Cardiol 76 2020 1947 1961 32890666 19 Kaushik S. Aydin S.I. Derespina K.R. Multisystem Inflammatory Syndrome in Children Associated with Severe Acute Respiratory Syndrome Coronavirus 2 Infection (MIS-C): A Multi-institutional Study from New York City J Pediatr 224 2020 24 29 32553861 20 Godfred-Cato S. Bryant B. Leung J. COVID-19-Associated Multisystem Inflammatory Syndrome in Children - United States, March-July 2020 MMWR Morb Mortal Wkly Rep 69 2020 1074 1080 32790663 21 Swann O.V. Holden K.A. Turtle L. Clinical characteristics of children and young people admitted to hospital with covid-19 in United Kingdom: prospective multicentre observational cohort study BMJ 370 2020 m3249 32960186 22 Zhang Q.Y. Xu B.W. Du J.B. Similarities and differences between multiple inflammatory syndrome in children associated with COVID-19 and Kawasaki disease: clinical presentations, diagnosis, and treatment World J Pediatr 17 2021 335 340 34013488 23 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 130 2020 5942 5950 32701511 24 Di Somma S. Pittoni V. Raffa S. IL-18 stimulates B-type natriuretic peptide synthesis by cardiomyocytes in vitro and its plasma levels correlate with B-type natriuretic peptide in non-overloaded acute heart failure patients Eur Heart J Acute Cardiovasc Care 6 2017 450 461 24585936 25 Yanagisawa D. Ayusawa M. Kato M. Factors affecting N-terminal pro-brain natriuretic peptide elevation in the acute phase of Kawasaki disease Pediatr Int 58 2016 1105 1111 26991905	36462758	PMC9711900	NO-CC CODE	2022-12-02 23:21:31	no	Can J Cardiol. 2022 Dec 1; doi: 10.1016/j.cjca.2022.11.012	utf-8	Can J Cardiol	2,022	10.1016/j.cjca.2022.11.012	oa_other
==== Front Int J Infect Dis Int J Infect Dis International Journal of Infectious Diseases 1201-9712 1878-3511 Published by Elsevier Ltd on behalf of International Society for Infectious Diseases. S1201-9712(22)00624-5 10.1016/j.ijid.2022.11.035 Article Effectiveness of COVID-19 Vaccines in People Living with HIV in British Columbia and comparisons with a matched HIV-Negative Cohort: A Test Negative Design Fowokan Adeleke 1* Samji Hasina 12#* Puyat Joseph 134 Janjua Naveed 13 Wilton James 1 Wong Jason 13 Grennan Troy 13 Chambers Catherine 5 Kroch Abigail 6 Costiniuk Cecilia T. 7 Cooper Curtis L. 8 Burchell Ann N. 59 Anis Aslam 34 1 British Columbia Centre for Disease Control, Vancouver, BC V5Z 4R4, Canada 2 Simon Fraser University, Faculty of Health Sciences, Burnaby, Canada 3 School of Population and Public Health, University of British Columbia, Vancouver, BC V6T 1Z4, Canada 4 Centre for Health Evaluation and Outcome Sciences, St Paul's Hospital, Vancouver, BC, Canada 5 Dalla Lana School of Public Health, University of Toronto, Toronto, ON M5T 3M7, Canada 6 The Ontario HIV Treatment Network, Toronto, ON M4T 1X3, Canada 7 Division of Infectious Diseases, McGill University, Montreal, QC, H4A 3J1, Canada 8 Department of Medicine, University of Ottawa, Ottawa, ON K1H 8M5, Canada 9 Department of Family and Community Medicine, St. Michael's Hospital, Toronto, ON M5B 1W8, Canada # Corresponding author. Hasina Samji, PhD, Senior Scientist BC Centre for Disease Control, Provincial Health Services Authority, Assistant Professor \| Faculty of Health Sciences, Simon Fraser University, 655 West 12th Avenue, Vancouver BC, V5Z 4R4 ⁎ Joint first authors 1 12 2022 1 12 2022 18 8 2022 24 10 2022 25 11 2022 © 2022 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. Background We estimated the effectiveness of COVID-19 vaccines against laboratory-confirmed SARS-CoV-2 infection among people living with HIV (PLWH) and compared estimates with a matched HIV-negative cohort. Method We used the British Columbia COVID-19 Cohort, a population-based data platform which integrates COVID-19 data on SARS-CoV-2 tests, laboratory-confirmed cases, and immunizations with provincial health services data. Vaccine effectiveness (VE) was estimated with a test-negative design using multivariable logistic regression. Results Adjusted VE against SARS-CoV-2 infection was 79.2% (52.5, 90.9%) 7-59 days after two doses, rising to 91.6% (75.2, 97.2%) between 60-89 days. VE was preserved four to six months following receipt of 2 doses after which slight waning was observed (72.7% [39.1, 87.8%]). In the matched cohort (n=375 043), VE peaked at 91.0% (90.5, 91.5%) 7-59 days after 2 doses and was sustained for up to four months after which evidence of waning was observed, dropping to 83.8% (82.9, 84.7%) between four to six months. Conclusion Receipt of two COVID-19 vaccines doses was effective against SARS-CoV-2 infection among PLWH pre-Omicron. VE estimates appeared to peak later in PLWH compared to the matched HIV-negative cohort and the degree of waning was relatively quicker in PLWH; however, overall estimates were comparable in both populations. Keywords Coronavirus disease SARS-CoV-2 HIV Vaccine Effectiveness Canada ==== Body pmcIntroduction People living with HIV (PLWH) appear to be at higher risk for severe Coronavirus disease (COVID-19) (Boulle et al., 2021; Ge et al., 2021; Nomah et al., 2021; Ssentongo et al., 2021; Yang et al., 2021).Several studies have now shown that those with low CD4 count (<200 cells/mm3) or CD4 nadir even with virologic suppression are at higher risk for worse outcomes including severe COVID and death (Bhaskaran et al., 2021; Dandachi et al., 2021; Liu et al., 2021). Yet, the evidence regarding effectiveness of COVID-19 vaccines in this high-risk group remains sparse, as PLWH have been largely underrepresented in vaccine trials (Ong & Bryce, 2022). Given the paucity of research in this area and the ongoing pandemic despite mass vaccination roll-out, it is important to evaluate COVID-19 vaccine effectiveness (VE) to inform COVID-19 vaccine strategies for PLWH. In British Columbia (BC), Canada, there were an estimated 10 682 PLWH in 2022, over 90% of whom are currently linked to HIV care (British Columbia Centre for Excellence in HIV/AIDS, 2022) and can be identified within population-level provincial health services databases. Integration of provincial COVID-19 immunization records and COVID-19 outcome data enables assessment of VE at a population level including for previously understudied population subgroups, like PLWH. One such approach is the Test-Negative Design (TND), which has been widely used to assess VE for influenza (Chua et al., 2020.), and more recently, COVID-19 (Chung et al., 2021; Skowronski, et al., 2022a). The TND is a modified case-control design, where vaccination status is compared between test-positive cases and test-negative controls (De Serres et al., 2013). However, the TND has been shown to be less prone to the selection and misclassification biases that commonly plague case-control studies via selecting cases and controls among those who present for testing (Jackson & Nelson, 2013; Sullivan et al., 2016; Vandenbroucke et al., 2020). Findings from Canadian studies using the TND to estimate VE against infection in the general population have reported estimates ≥ 90% (Buchan et al., 2022; Chung et al., 2021; Skowronski, et al., 2022a), yet it remains unclear to what extent these estimates apply to PLWH. Our best knowledge of COVID-19 vaccines in PLWH come from immunogenicity studies which have shown that immune response in PLWH with CD4 count ≥250 cells/mm3 appear comparable to those in the larger population (Brumme et al., 2022; Nault et al., 2022). However, these studies provide us with limited insight into the real-world impact of COVID-19 vaccines in PLWH. Consequently, population-based studies provide us with the best available opportunity to evaluate the real-world impact of COVID-19 vaccines in understudied groups like PLWH. We estimated the VE of COVID-19 vaccines against laboratory confirmed infection and compared VE estimates with a matched HIV-negative cohort in the pre-Omicron era. METHODS Study Population, Data Sources and Design This was a TND study using the British Columbia (BC) COVID-19 Cohort (BCC19C) to estimate VE among PLWH in BC. The BCC19C was established as a collaboration among the BC Centre for Disease Control (BCCDC), the Data, Analytics Reporting and Evaluation (DARE), the Provincial Health Services Authority (PHSA), and the BC Ministry of Health (MoH) to support the COVID-19 pandemic response. The BCC19C include population-level province wide COVID-19 datasets including SARS-CoV-2 testing, COVID-19 case surveillance, hospitalizations, and vaccinations, which are integrated with data from other provincial administrative data holdings and registries including: 1) Medical Services Plan (MSP); 2) Chronic Disease Registry (CDR); 3) Death Records (Vital Stats); 4) Client Roster; 5) Discharge Abstract Database (DAD); and 6) National Ambulatory Care Reporting System (NACRS) and Vital Statistics. Detailed information related to the datasets is included in supplemental file 1. We included PLWH residing in BC, ≥19 years old, accessing healthcare, and alive on December 15, 2020, who received a laboratory test for SARS-Cov-2 between December 15, 2020, and November 21, 2021. In BC, testing policies during the study period required individuals to display symptoms consistent with COVID-19 before being tested (with exceptions for travellers). We excluded those who tested positive for SARS Cov-2 before the start of the study period and those who received ≥3 doses from VE analyses as we will seek to explore 3-dose VE in future studies. We specified the study period to coincide with the mainstreaming of vaccines in BC and the time before the first case of Omicron was detected in BC to account for the attenuating impact of the Omicron variant on COVID-19 VE which has been widely reported in the research literature (Andrews et al., 2022; Buchan et al., 2022). Ascertainment of PLWH status We adapted a previously validated case-finding algorithm which has been previously described elsewhere (Nosyk et al., 2013) using international classification of disease (ICD) diagnostic codes (ICD-9 and ICD-10) that have been associated with HIV (see supplemental file 2 for details on case-finding algorithm and full list of diagnostic codes) to create a retrospective cohort of PLWH. Briefly, individuals who had >3 physician visits (MSP from 2008 to 2021), >1 hospitalization (DAD) or >1 ED visit for any of the HIV-related codes in Appendix 1 were considered HIV cases. Modifications were made to the initial algorithm to include individuals with a positive HIV lab test results based on provincial HIV laboratory test interpretation guidelines and those in the HIV/AIDS surveillance system (Janjua et al., 2016). Study Variables Outcome Our primary study outcome was SARS-Cov-2 infection defined as a positive laboratory confirmed test identified using a provincial database of COVID-19 test data. Those who tested positive for SARS-CoV-2 during the study period were considered test-positive “cases”, while those who tested negative were considered test-negative “controls”. For cases with multiple positive tests, the first positive test was selected. We used the first test positive only to minimize confounding the VE estimates with natural immunity from previous SARS-CoV-2 infection. For controls with multiple test negative results, a randomly selected negative test was chosen. Exposure of Interest (COVID-19 Vaccination Status) In BC, three vaccine products – BNT162b2 (Pfizer-BioNTech), mRNA-1273 (Moderna) and ChAdOx1 (Oxford-AstraZeneca)) – have been mostly used in its vaccination program, in addition to a small number of Janssen (Johnson and Johnson) vaccines. Information on the specific vaccine type received (i.e., BNT162b2 (Pfizer-BioNTech), mRNA-1273 (Moderna), or ChAdOx1 (Oxford-AstraZeneca)) and the number of doses were obtained from the BC Provincial Immunization Registry. For this study, individuals with a record of a single dose of any vaccine type were considered for inclusion in VE analyses. Because vaccination strategies for multiple doses varied between homologous and heterologous vaccine schedules, owing to variations in vaccine supply, any combination of any two of the vaccines were considered vaccinated with two doses. We also included information on vaccine dose timing from index date, which was defined as the date the test specimen was collected. Covariates Demographic variables such as age (categorized in 10-year intervals), sex, socioeconomic status (using neighbourhood income quintiles) based on census data and health region (i.e., Fraser Health, Interior Health, Northern Health, Vancouver Coastal Health and Vancouver Island Health) to which the individual belonged to were obtained from the Client Roster. The health authorities delineate important geographic distinctions relating to healthcare service delivery that might provide insight into disparities, particularly among PLWH (see supplemental file 3 for detailed information about health regions). Clinical information such as history of comorbidities known to be associated with increased risk of adverse COVID-19 outcomes were identified from the NACRS, CDR, DAD, and MSP databases using relevant ICD-9 and ICD-10 codes. This information was subsequently used to calculate the Elixhauser Comorbidity Index, categorizing an individual's history of comorbidities as either 0 (no comorbidity), 1 (history of one comorbidity), 2 (a history of two comorbidities), or 3 or more comorbidities (Elixhauser et al., 1998). People who inject drugs (PWID) were identified in the BCC19C using previously validated algorithm (Janjua et al., 2018). To account for the time-variations in COVID-19 cases and vaccine roll-out throughout the length of the study period, we categorized dates of testing into bi-weekly calendar time periods and epidemic waves. Our study period spanned COVID-19 epidemic waves 2 - 4 (Wave 2: December 15, 2020, to February 6, 2021; Wave 3: February 7, 2021, to July 3, 2021; and Wave 4: July 4, 2021, to November 21, 2021). Matched HIV-Negative Cohort We matched each PLWH included in this study to an HIV negative individual on the following variables: age (5-year intervals), sex, community health service area, and SARS-CoV-2 outcome status. We defined HIV-negative individuals as those who did not meet the PLWH algorithm. To increase precision of the VE estimates obtained from the matched HIV-negative cohort, we applied a one-to-many coarsened exact matching (Ripollone et al., 2020) approach to obtain as many HIV-negative matches from the general population. Statistical Analyses To describe the baseline characteristics of PLWH and the matched HIV-negative cohort we used means and standard deviation for continuous variables and frequencies and percentages for categorical variables. Standardized differences (SD) were then used to compare baseline characteristics of test positive and test negative PLWH cases, and test positive and test negative HIV-negative cases. SD values of >0.10 were used to determine clinically meaningful differences (Austin, 2009). Multivariable logistic regression was used to estimate the odds ratio (OR) comparing the odds of COVID-19 vaccination between test-positive cases and test negative controls. Adjusted analyses included covariates chosen based on comparable VE studies and literature evidence documenting their correlation with SARS-Cov-2 and HIV infection (Chung et al., 2021; Skowronski et al., 2022b). The following covariates were included in the adjusted models: age (10-year age bands), sex, area-level income, health authority, number of COVID-19 tests 3 months prior to study period, Elixhauser comorbidity index, and bi-weekly testing periods. Vaccine effectiveness was computed using the formula (1-OR) x100%. We conducted two separate regression models to estimate and indirectly compare VE for the cohort of PLWH and the matched HIV-negative cohort, respectively. We estimated VE by time since receipt of vaccine dose (i.e., ≥14 days after first dose; 7-59, 60-89, 90-119, 120-179 after 2 doses). Secondary VE analyses In addition to the stratified VE analyses described above, we conducted secondary analyses combining both the PLWH and HIV-negative group in a single logistic regression model, specifying an interaction term between the composite vaccination status variable and PLWH status to estimate VE for each population. This was done to account for known baseline and clinical differences between both the PLWH and matched HIV-negative groups that might not have been accounted for in the stratified VE analyses. We adjusted for the same covariates included in the stratified VE analyses. All data analyses were conducted using SAS Version 9.4 (SAS Institute Inc., Cary, NC). All tests were two sided, with p<0.05 used as the level of statistical significance. RESULTS There were 8200 PLWH identified in the BCC19C dataset. Between December 15, 2020, when vaccines became available, through November 21, 2021, a total of 2700 PLWH tested for SARS-Cov-2 and were eligible to be included in this study. Of the eligible cohort, 351 (13.0%) tested positive while 2349 (87.0%) tested negative, constituting our test-positive “cases” and test negative “controls”, respectively (Figure 1 ). After matching, we included a total of 375, 043 (103, 049 [27.5%] test-positive cases and 271, 994 [72.5%] test-positive “controls”) in the matched HIV-negative group who tested for SARS-CoV-2 within the same study period and formed the comparator cohort for this study. The bi-weekly testing patterns across both PLWH and the HIV-negative groups were mostly comparable, except for between March to April, 2021 (epidemic wave 3) and September to October, 2021 (epidemic wave 4), where noticeable spikes in the proportion of HIV-negative individuals and PLWH, who tested negative, respectively, were observed (Figure 2 ).Figure 1 Study Flow Diagram for PLWH cohort. Figure 1 Figure 2 Proportion of PLWH and matched HIV-negative cohort who tested positive and negative for SARS-CoV-2 by bi-weekly period. HIV= human immunodeficiency virus; PLWH =People living with HIV; C19+ = Test positive (cases); C19- = test negative (controls). Figure 2 PLWH Baseline and clinical characteristics of the study population by COVID-19 testing status are described in Table 1 . The cohort of PLWH identified were predominantly male (71.4%, 1927), which is broadly representative of PLWH in BC (British Columbia Centre for Excellence in HIV/AIDS, 2022). 2377 (88.4%) of participants had received at least one vaccine dose by the end of the study period. Of those vaccinated, 1773 (65.7%) had received two vaccine doses at the study index date. Compared to test negative controls, test positive PLWH were younger (mean age: 48.7 [SD= 11.9] years vs. 50.6 [SD= 13.2] years), had higher proportion of females, people in the 40-49 age group, people who live in the Northern health authority and PWIDs. Additionally, test positive PLWH had lower proportions of people who received three vaccine doses, people with only one comorbidity, epidemic wave 3 infection (February 7, 2021, to July 3, 2021), people who have received any of the three main vaccine types, and people who live in the Vancouver Island health authority. (Table 1).Table 1 Demographic and Clinical Characteristics of Study participants by SARS-CoV-2 Testing Status Table 1Study Characteristics PLWH (n =2700) Standardized differences Matched HIV-Negative cohort (n = 375 043) Standardized differences n, %a Test-Positive Case (n=351) Test-Negative Control (n= 2349) Positive (n=103049) Negative (n=271994) Mean Age (SD) 48.7 (11.9) 50.6 (13.2) 0.15 44.8 (17.3) 51.2 (19.1) 0.35 Age group 19-29 19 (5.4%) 146 (6.2%) 0.03 23447 (22.8%) 43737 (16.1%) 0.17 30-39 68 (19.4%) 393 (16.7%) 0.07 22557 (21.6%) 44341 (16.3%) 0.14 40-49 97 (27.6%) 497 (21.2%) 0.15 19098 (18.5%) 42814 (15.7%) 0.07 50-59 100 (28.5%) 729 (31.0%) 0.06 16810 (16.3%) 43039 (15.8%) 0.01 60-69 51 (14.5%) 414 (17.6%) 0.08 11642 (11.3%) 42149 (15.5%) 0.12 70-79 14 (4.0%) 135 (5.8%) 0.08 5924 (5.8%) 36881 (13.6%) 0.27 ≥80 <5 35 (1.5%) 0.09 3871 (3.8%) 19037 (7.0%) 0.14 Sex Female 127 (36.2%) 646 (27.5%) 0.19 51354 (49.8%) 138719 (51.0%) 0.02 Neighborhood Income (quintiles) Lowest 157 (44.7%) 971 (41.3%) 0.07 23631 (22.9%) 54962 (20.2%) 0.07 2 60 (17.1%) 461 (19.6%) 0.07 21143 (20.5%) 51626 (19.0%) 0.04 3 57 (16.2%) 421 (17.9%) 0.04 20170 (19.6%) 55468 (20.4%) 0.02 4 51 (14.5%) 319 (13.6%) 0.03 20389 (19.8%) 56614 (20.8%) 0.03 Highest 26 (7.4%) 174 (7.4%) 0 17522 (17.0%) 52928 (19.5%) 0.06 Persons who Inject Drugs Yes 176 (50.1%) 922 (39.3%) 0.22 5703 (5.5%) 10557 (3.9%) 0.08 Number of Vaccine Doses 0 75 (21.4%) 248 (10.6%) 0.30 21153 (20.5%) 23694 (8.7%) 0.34 1 26 (7.4%) 155 (6.6%) 0.03 6437 (6.3%) 9804 (3.6%) 0.12 2 224 (63.8%) 1549 (65.9%) 0.04 71217 (69.1%) 208509 (76.7%) 0.17 3 26 (7.4%) 397 (16.9%) 0.29 4242 (4.1%) 29987 (11.0%) 0.26 Pandemic Wave Wave 2: Dec. 15, 2020 to Feb. 6, 2021 74 (21.1%) 467 (19.9%) 0.03 16886 (16.4%) 47159 (17.3%) 0.03 Wave 3: Feb. 7, 2021 to Jul. 3, 2021 124 (35.3%) 963 (41.0%) 0.12 45220 (43.9%) 106567 (39.2%) 0.10 Wave 4: Jul. 4, 2021 to Dec 4, 2021 153 (43.6%) 919 (39.1%) 0.09 40943 (39.7%) 118268 (43.5%) 0.08 Time since 1st dose (days) 0-13 <5 15 (0.64) 0.16 548 (0.53) 671 (0.25) 0.01 ≥14 11 (3.13) 54 (2.30) 0.14 1282 (1.24) 2906 (1.07) 0.19 Time since 2nd dose (Days) 0-6 <5 12 (0.51) 0.04 603 (0.59) 2041 (0.75) 0.23 7-59 14 (3.99) 171 (7.28) 0.58 2687 (2.61) 24652 (9.06) 0.94 60-119 18 (5.13) 255 (10.86) 0.70 3209 (3.11) 41527 (15.27) 0.86 120-179 22 (6.27) 121 (5.15) 0.23 3581 (3.48) 16981 (6.24) 0.65 ≥180 <5 11 (0.47) 0.02 3452 (3.35) 1766 (0.65) 0.14 Time since 3rd dose 0-6 <5 10 (0.43) 0.07 110 (0.11) 564 (0.21) 0.15 ≥7-29 <5 25 (1.06) 0.36 135 (0.13) 1654 (0.61) 0.32 Elixhauser Comorbidity Index 0 39 (11.1%) 268 (11.4%) 0.01 32433 (31.5%) 64317 (23.7%) 0.18 1 37 (10.5%) 342 (14.6%) 0.12 25722 (25.0%) 60743 (22.3%) 0.06 2 58 (16.5%) 339 (14.4%) 0.06 16982 (16.5%) 47039 (17.3%) 0.02 3 or more 217 (61.8%) 1400 (59.6%) 0.05 27912 (27.1%) 99895 (36.7%) 0.21 Health Authority Interior 23 (6.6%) 189 (8.1%) 0.06 19046 (18.5%) 54524 (20.1%) 0.04 Fraser 111 (31.6%) 651 (27.7%) 0.09 42175 (40.9%) 70581 (26.0%) 0.32 Vancouver Coastal 175 (49.9%) 1218 (51.9%) 0.04 22764 (22.1%) 63013 (23.2%) 0.03 Vancouver Island 20 (5.7%) 217 (9.2%) 0.14 8077 (7.8%) 57540 (21.2%) 0.39 Northern 22 (6.3%) 71 (3.0%) 0.15 10798 (10.5%) 25976 (9.6%) 0.03 Vaccine Received Pfizer 217 (61.8%) 1592 (67.8%) 0.12 57959 (56.2%) 172630 (63.5%) 0.15 Moderna 74 (21.1%) 659 (28.1%) 0.16 27878 (27.1%) 98048 (36.1%) 0.19 AstraZeneca 20 (5.7%) 222 (9.5%) 0.14 5335 (5.2%) 17078 (6.3%) 0.05 Other 0 <5 0.04 151 (0.2%) 174 (0.1%) 0.03 Number of tests 3 months before Dec. 15, 2020 0 255 (72.7%) 1707 (72.7%) 0.00 88105 (85.5%) 226567 (83.3%) 0.06 1 70 (19.9%) 455 (19.4%) 0.01 12826 (12.5%) 38294 (14.1%) 0.04 2 14 (4.0%) 124 (5.3%) 0.06 1573 (1.5%) 5395 (2.0%) 0.03 3 12 (3.4%) 63 (2.7%) 0.04 545 (0.5%) 1738 (0.6%) 0.01 a Unless otherwise specified HIV= human immunodeficiency virus; PLWH= People living with HIV; SD=standardized difference. Standardized differences of >0.10 are considered clinically relevant VE estimates from unadjusted and adjusted models are presented in Table 2 and Figure 3 , respectively. Adjusted VE against laboratory confirmed infection ≥14 days after first dose was 54.7% (95% CI = -9.3 to 81.2%). Vaccine effectiveness 7 to 59 days after second vaccine dose was 79.2% (95% CI = 52.5 to 90.9%), this increased to 91.6% (95% CI = 75.2 to 97.2%) 60-89 days after dose 2 and was preserved up to 90 to 119 days. We found evidence suggestive of vaccine waning 120 to 179 days after dose 2 (VE = 72.7% (95% CI = 39.1 to 87.8%) (Figure 3).Table 2 Unadjusted Vaccine Effectiveness Estimates of COVID-19 vaccines against laboratory-confirmed infection during the study period, by time since vaccine dose Table 2 PLWH (n =2700) Matched HIV-negative cohort (n = 375043) VE (%) Lower CI (%) Upper CI (%) VE (%) Lower CI (%) Upper CI (%) 1st dose (≥14 days) 32.6 -35.4 66.5 50.6 47.1 53.8 2nd dose (7 to 59 days) 72.9 50.5 85.2 87.8 87.2 88.3 2nd dose (60 to 89 days) 83.0 59.9 92.8 82.7 81.9 83.4 2nd dose (90 to 119 days) 71.2 45.3 84.9 80.7 79.9 81.5 2nd dose (120 to 179 days) 39.9 -1.4 64.3 77.2 76.3 78.1 Figure 3 Adjusted Vaccine Effectiveness Estimates of COVID-19 vaccines against laboratory-confirmed infection during the study period, by time since vaccine dose. Figure 3 Matched HIV-negative Cohort Among the matched cohort of people who were HIV-negative by the end of the study period, 330,196 (88.0%) had received at least a single vaccine dose; of those vaccinated, 279,726 (75.4%) had received two vaccine doses at the study index date. Compared to test negative controls, test positive participants were younger (mean age: 44.3 (SD= 17.3) years vs 51.2 (SD= 19.1) years), had higher proportions of people aged 19-29 and 30-39, with no comorbidities, pandemic wave 3 infections, those residing in the Fraser health authority, and people who received either no vaccines or one vaccine dose. Conversely, test positive participants had lower proportions of those in age groups 60-69, 70-79 and ≥80, people who received 2 and 3 vaccine doses, people with three or more comorbidities, those who reside in the Vancouver Island health authority and people who received a Pfizer or Moderna vaccine dose (Table 1). Among the matched cohort, adjusted VE against infection ≥14 days after first dose was 53.5% (95% CI = 49.9 to 56.8%). This increased 7 to 59 days after second vaccine dose, peaking at 91.0% (95% CI = 90.5 to 91.5%). VE was preserved 60-89 days after dose 2 (VE = 89.4% (95% CI = 88.8 to 89.9%) and up to 90 to 119 days after dose 2 (VE = 87.6% (95% CI = 86.9 to 88.2%). VE 120 to 179 days after dose 2 was 83.8% (95% CI = 82.9 to 84.7%) (Figure 3). Secondary analyses comparing VE by HIV status Adjusted VE estimates from secondary VE analyses are presented in Table 3 . Overall, the findings from secondary VE analyses appear comparable to those from stratified analyses presented above. For example, VE peaked earlier in the matched cohort 7-59 days after receipt of 2nd dose at 91.0% (95% CI = 90.5 to 91.5%) compared to 76.6% (95% CI = 52.2 to 88.5%). Similar to findings from stratified analyses, waning was observed 120-179 days after receipt of second dose for both cohorts, however the degree of waning was more pronounced in PLWH [VE = 65.1% (95% CI = 30.3 to 82.6%) compared to 83.8% (95% CI = 82.9 to 84.7%) in the matched cohort]. Findings from interaction term analyses show that the differences in VE for PLWH and HIV negative participants 7-59 days (p= 0.008) and 120-179 days (p= 0.03) after receipt of the second dose were statistically significant (supplemental file 4)Table 3 . Combined Test-Negative Design Estimate of Vaccine Effectiveness Against Laboratory-Confirmed Infection Table 3 PLWH (n =2700) Matched HIV-negative cohort (n = 375043) VE (%) Lower CI (%) Upper CI (%) VE (%) Lower CI (%) Upper CI (%) 1st dose (≥14 days) 45.6 -13.2 73.8 53.4 49.8 56.8 2nd dose (7 to 59 days) 76.6 52.2 88.5 91.0 90.5 91.5 2nd dose (60 to 89 days) 90.5 75.4 96.3 89.4 88.8 89.9 2nd dose (90 to 119 days) 84.2 66.3 92.5 87.6 86.9 88.2 2nd dose (120 to 179 days) 65.1 30.3 82.6 83.8 82.9 84.7 CI = Confidence Interval; HIV= human immunodeficiency virus; PLWH =People living with HIV; VE= Vaccine Effectiveness Estimates were adjusted for baseline differences between PLWH and the matched HIV-negative cohort on the following variables: age, sex, area-level income, health authority, number of COVID-19 tests 3 months prior to study period, Elixhauser comorbidity index, and bi-weekly testing periods DISCUSSION Applying the TND to a retrospective cohort of 2,700 PLWH and over 375,000 matched non-PLWH in British Columbia from December 15, 2020, to November 21, 2021, we found that two doses of COVID-19 vaccines offered considerable protection against lab-confirmed infection among both PLWH and HIV-negative individuals during the pre-Omicron period. Among PLWH, VE estimates one week to two months after receipt of two vaccine doses were 79.2% (52.5 to 90.9%), rising to 91.6% (75.2, 97.2%) up to three months after receipt of vaccine doses, and relatively sustained for up to four months. We found evidence of vaccine waning four to six months post receipt of two COVID-19 vaccine doses, with VE against infection declining to 72.7%. When compared with the matched HIV-negative cohort, we observed different patterns in VE estimates. For example, VE peaked earlier in the matched HIV-negative cohort at 91% 7 to 59 days after the second dose, whereas in PLWH the VE peak was observed later during the period 60-89 days after the second dose. Similarly, differences in the pattern of waning between the two populations were evident. After reaching their individual peaks, VE was relatively sustained in the matched HIV-negative cohort up to three months after the second dose, with waning observed after the four-month mark with a VE of 83.8%. While similar patterns of waning were observed in the PLWH cohort at the four-month mark, the degree to which waning occurred was higher in PLWH with a VE estimate of 72.7% four months after the receipt of two doses. These findings when compared to our VE estimates in PLWH affirm our findings that although COVID-19 vaccines are effective in PLWH, a longer period might be required to achieve effectiveness levels noticed earlier in the larger non-HIV population, and that waning might occur earlier in PLWH than in otherwise healthier cohorts. However, these results appear in contrast to findings from immunological studies that show that similar antibody response in PLWH with high CD4 counts compared to healthy controls (Brumme et al., 2022; Levy et al., 2021; Nault et al., 2022). It is possible that the VE patterns we observed for PLWH relative to the HIV-negative cohort might be explained by CD4 count distribution, as was reported in a study of VE of the Sputnik vaccines in PLWH (Gushchin et al., 2022), but CD4 count data were unavailable for PLWH in this cohort. VE against infection for those with CD4 count <350 cells/µl was 73% compared to 79% in those with CD4 ≥ 350 cells/µl (Gushchin et al., 2022). Consequently, understanding the role HIV clinical parameters play in impacting VE will be integral to fully informing COVID-19 strategies in PLWH. Our VE estimates among the HIV-negative cohort are in line with other studies estimating the real-world VE of COVID-19 vaccines in the general population during pre-Omicron periods. Although the study periods we report were different, in Ontario, Canada, Chung et al reported a VE of 91% against symptomatic infection 7 days after receipt of 2 doses (Chung et al., 2021). Likewise in a cohort of BC and Quebec participants, two-dose mRNA VE against symptomatic infection was sustained at 90% through the 3rd month, with slight declines noted, but sustained at ≥80% up until the period 6 to 7 months (Skowronski et al., 2022a). It should be noted that the period specified in these studies were also prior to the spread of the Omicron variant. To our knowledge, there are no known studies estimating the real-world effectiveness of the included vaccines solely among PLWH, thus limiting our ability to contextualize our VE findings within the broader PLWH context. However, mixed patterns were observed in other immunocompromised populations. In a meta-analysis estimating the pooled, short term, two-dose VE of COVID-19 vaccines against symptomatic infection in immunocompromised individuals (n =42, 821 including recipients of hematopoietic cell or solid organs transplant, patients with inflammatory disorders, PLWH, patients under immunosuppressive therapy, asplenia, and chronic renal failure: advanced kidney disease, dialysis, or nephrotic syndrome etc.), VE was 70.4% (95% CI = 18.9 – 89.2%) (Marra et al., 2022). Individual study estimates from the included studies, however, ranged from 63% to 80%. One of the studies (Tenforde et al., 2022) included in the meta-analysis enrolled a merged immunocompromised cohort that contained a population of PLWH, yet the sampling approach adopted precluded us from making direct HIV-related VE comparisons. Another review comparing VE in immunocompromised populations to the general population by vaccine type found that the Pfizer BioNTech (BNT162b2) had the highest VE in immunocompromised populations (90% in immunocompromised population to 93% in the general population). The lowest VE reported among immunocompromised cohort was for the Ad26.COV2.S (Janssen vaccine) (64% VE in immunocompromised cohort to 79% in the general population) (Di Fusco et al., 2022). Taken together, these findings suggest that while VE in immunocompromised population might be lower relative to the broader population, the vaccine type and health conditions are relevant factors to consider. Future studies, where possible should refrain from sampling approaches that treat individuals considered to be immunocompromised as a homogenous cohort. Lastly, while we provide findings highlighting the real-world effectiveness of COVID-19 vaccines against SARS-CoV-2infection, low hospitalizations and death counts impeded our ability to estimate VE against hospitalizations and deaths in PLWH. Future studies are needed to examine severe outcomes in order to provide a complete outlook on the impact of COVID-19 vaccines in PLWH. This will help provide the totality of real-world evidence to inform vaccine strategies in this priority population. Study Limitations and Strengths Low event counts prevented us from generating more precise estimates and inhibited our ability to estimate VE against hospitalizations and death. In addition, the absence of HIV clinical characteristics impeded our ability to provide information on the HIV profile of the cohort. This prevents the identification of sub-populations of PLWH who may be more likely to experience lower VE (e.g., people not on antiretroviral treatment, worse immune status etc.). Available estimates, however, suggest that of diagnosed PLWH in BC, about 92% are on antiretrovirals while about 95% have suppressed viral loads (Public Health Agency of Canada, 2022). Although the use of the algorithms enabled PLWH case identification within administrative holdings, the imprecise sensitivity of the adapted algorithm at 88% means some PLWH might have been missed, potentially impacting study findings. While the modifications (Janjua et al., 2016) made to the algorithm helped to address some of this, validation information on the modifications might be needed to better inform its use in PLWH case identification. We were also limited in our ability to stratify findings by the dominant variants during our study period. However, data from other studies showing comparable VE rate during similar time period may suggest that stratification by the dominant strains within our specified period might not be as important. Additionally, the use of specimen collection date as a proxy of SARS-Cov-2 symptom onset restricted our ability to limit the VE analysis to individuals who were tested within a specific time period since the onset of symptoms, potentially resulting in outcome misclassification. The study strengths include the novel nature of this research and the methodological approach adopted, which allowed us to estimate the real-world VE in this population and provide us with critical insight into the impact of COVID-19 in this population. CONCLUSION Our study demonstrates that the effectiveness of two doses of COVID-19 vaccines among PLWH is broadly similar to VE in the general population against confirmed SARS-CoV2 infection up until November 2021 representing the period before Omicron circulation in Canada. Future work will evaluate the impact of Omicron and other variants on VE among PLWH, test whether observed trends in earlier waning are confirmed, and evaluate VE against hospitalizations and deaths among PLWH. Authors’ Contributions: HS, NJ, AB, AA conceptualized the study, secured research funding, contributed to study design, and reviewed the manuscript. AF drafted the initial manuscript, contributed to the analysis, and revised the manuscript. JP conducted data analysis and reviewed the manuscript. AK, CC, CTC, CLC, JW and TG provided methodological and analytical input and reviewed the manuscript. All authors approved the manuscript. Disclaimer: All inferences, opinions, and conclusions drawn in this manuscript are those of the authors, and do not reflect the opinions or policies of the Data Steward(s). Acknowledgements: We acknowledge the assistance of the Provincial Health Services Authority, BC Centre for Disease Control, BC Ministry of Health and Regional Health Authority staff involved in data access, procurement, and management. We gratefully acknowledge the population of PLWH in British Columbia whose data have been used in this manuscript. The BCC19C was established and is maintained through operational support from Data Analytics, Reporting and Evaluation (DARE), and BC Centre for Disease Control (BCCDC) at the Provincial Health Services Authority. We would also like to acknowledge the contribution of our community partners – Darren Lauscher and Monte Strong – who provided valuable insight that helped inform data analysis and writing. Conflict of Interest: The authors have no conflicts of interest to disclose. Funding Source: This project is being supported by funding from the Public Health Agency of Canada, through the Vaccine Surveillance Reference group and the COVID-19 Immunity Task Force and the Canadian Institutes for Health Research Canadian HIV Trials Network. Joseph Puyat is supported by the Michael Smith Foundation for Health Research (Scholar Awards) Ethical Approval Statement: This study was reviewed and approved by the University of British Columbia Research Ethics Board (REB) (REB#: H20-02097). Declaration of 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. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: References Andrews N, Stowe J, Kirsebom F, Toffa S, Rickeard T, Gallagher E et al. Covid-19 Vaccine Effectiveness against the Omicron (B.1.1.529) Variant. N Engl J Med. 2022;386(16):1532-1546. doi: 10.1056/NEJMoa2119451. Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38(6):1228-34.. https://doi.org/10.1080/03610910902859574 Bhaskaran K, Rentsch CT, MacKenna B, Schultze A, Mehrkar A, Bates CJ et al. HIV infection and COVID-19 death: a population-based cohort analysis of UK primary care data and linked national death registrations within the OpenSAFELY platform. Lancet HIV. 2021;8(1):e24-e32. https://doi.org/10.1016/S2352-3018(20)30305-2 Boulle A, Davies MA, Hussey H, Ismail M, Morden E, Vundle Z, Zweigenthal V et al. Risk Factors for Coronavirus Disease 2019 (COVID-19) Death in a Population Cohort Study from the Western Cape Province, South Africa. Clin Infect Dis. 2021;73(7):e2005-e2015. https://doi.org/10.1093/cid/ciaa1198 British Columbia Centre for Excellence in HIV/AIDS. HIV Monitoring Semi-Annual Report For British Columbia - Second Quarter 2022. https://stophivaids.ca/qmr/2022-Q2/#/bc, 2022 (Accesssed 10 August 2022) Brumme ZL, Mwimanzi F, Lapointe HR, Cheung PK, Sang Y, Duncan MC et al. Humoral immune responses to COVID-19 vaccination in people living with HIV receiving suppressive antiretroviral therapy. NPJ Vaccines. 2022;7(1):28. https://doi.org/10.1038/s41541-022-00452-6 Buchan SA, Chung H, Brown KA, Austin PC, Fell DB, Gubbay J et al. Effectiveness of COVID-19 vaccines against Omicron or Delta infection. MedRxiv. 2022:2021-12. https://doi.org/10.1101/2021.12.30.21268565 Chua H, Feng S, Lewnard JA, Sullivan SG, Blyth CC, Lipsitch M et al. The Use of Test-negative Controls to Monitor Vaccine Effectiveness: A Systematic Review of Methodology. Epidemiology. 2020;31(1):43-64. doi: 10.1097/EDE.0000000000001116. Chung H, He S, Nasreen S, Sundaram ME, Buchan SA, Wilson SE et al. Effectiveness of BNT162b2 and mRNA-1273 covid-19 vaccines against symptomatic SARS-CoV-2 infection and severe covid-19 outcomes in Ontario, Canada: test negative design study. BMJ. 2021;374:n1943.. https://doi.org/10.1136/BMJ.N1943 Dandachi D, Geiger G, Montgomery MW, Karmen-Tuohy S, Golzy M, Antar AAR et al. Characteristics, Comorbidities, and Outcomes in a Multicenter Registry of Patients With Human Immunodeficiency Virus and Coronavirus Disease 2019. Clin Infect Dis. 2021;73(7):e1964-e1972. https://doi.org/10.1093/CID/CIAA1339 De Serres G, Skowronski DM, Wu XW, Ambrose CS. The test-negative design: validity, accuracy and precision of vaccine efficacy estimates compared to the gold standard of randomised placebo-controlled clinical trials. Euro Surveill. 2013;18(37):20585. https://doi.org/10.2807/1560-7917.ES2013.18.37.20585 Di Fusco M, Lin J, Vaghela S, Lingohr-Smith M, Nguyen JL, Scassellati Sforzolini T et al. COVID-19 vaccine effectiveness among immunocompromised populations: a targeted literature review of real-world studies. Expert Rev Vaccines. 2022;21(4):435-451. https://doi.org/10.1080/14760584.2022.2035222 Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8-27. https://doi.org/10.1097/00005650-199801000-00004 Ge E, Li Y, Wu S, Candido E, Wei X. Association of pre-existing comorbidities with mortality and disease severity among 167,500 individuals with COVID-19 in Canada: A population-based cohort study. PLoS One. 2021;16(10):e0258154. https://doi.org/10.1371/journal.pone.0258154 Gushchin VA, Tsyganova EV, Ogarkova DA, Adgamov RR, Shcheblyakov DV, Glukhoedova NV et al. Sputnik V protection from COVID-19 in people living with HIV under antiretroviral therapy. EClinicalMedicine. 2022;46:101360. https://doi.org/10.1016/J.ECLINM.2022.101360 Jackson ML, Nelson JC. The test-negative design for estimating influenza vaccine effectiveness. Vaccine. 2013;31(17):2165-8. https://doi.org/10.1016/J.VACCINE.2013.02.053 Janjua NZ, Islam N, Kuo M, Yu A, Wong S, Butt ZA et al. Identifying injection drug use and estimating population size of people who inject drugs using healthcare administrative datasets. Int J Drug Policy. 2018;55:31-39. https://doi.org/10.1016/J.DRUGPO.2018.02.001 Janjua NZ, Kuo M, Chong M, Yu A, Alvarez M, Cook D et al. Assessing Hepatitis C Burden and Treatment Effectiveness through the British Columbia Hepatitis Testers Cohort (BC-HTC): Design and Characteristics of Linked and Unlinked Participants. PLoS One. 2016;11(3):e0150176. https://doi.org/10.1371/JOURNAL.PONE.0150176 Levy I, Wieder-Finesod A, Litchevsky V, Biber A, Indenbaum V, Olmer L, Huppert A et al. Immunogenicity and safety of the BNT162b2 mRNA COVID-19 vaccine in people living with HIV-1. Clin Microbiol Infect. 2021;27(12):1851-1855. https://doi.org/10.1016/J.CMI.2021.07.031 Liu Y, Xiao Y, Wu S, Marley G, Ming F, Wang X et al. People living with HIV easily lose their immune response to SARS-CoV-2: result from a cohort of COVID-19 cases in Wuhan, China. BMC Infect Dis. 2021;21(1):1029. https://doi.org/10.1186/S12879-021-06723-2/TABLES/3 Marra AR, Kobayashi T, Suzuki H, Alsuhaibani M, Tofaneto BM, Bariani LM et al. Short-term effectiveness of COVID-19 vaccines in immunocompromised patients: A systematic literature review and meta-analysis. J Infect. 2022 Mar;84(3):297-310. https://doi.org/10.1016/J.JINF.2021.12.035 Nault L, Marchitto L, Goyette G, Tremblay-Sher D, Fortin C, Martel-Laferrière V et al. Covid-19 vaccine immunogenicity in people living with HIV-1. Vaccine. 2022;40(26):3633-3637. https://doi.org/10.1016/J.VACCINE.2022.04.090 Nomah DK, Reyes-Urueña J, Díaz Y, Moreno S, Aceiton J, Bruguera A et al. Sociodemographic, clinical, and immunological factors associated with SARS-CoV-2 diagnosis and severe COVID-19 outcomes in people living with HIV: a retrospective cohort study. Lancet HIV. 2021;8(11):e701-e710. https://doi.org/10.1016/S2352-3018(21)00240-X Nosyk B, Colley G, Yip B, Chan K, Heath K, Lima VD et al. Application and validation of case-finding algorithms for identifying individuals with human immunodeficiency virus from administrative data in British Columbia, Canada. PLoS One. 2013;8(1):e54416. https://doi.org/10.1371/JOURNAL.PONE.0054416 Ong S & Bryce E. How effective are Covid-19 vaccines for people with HIV? British Broadcasting Corporation. https://www.bbc.com/future/article/20220419-how-effective-are-covid-19-vaccines-for-people-with-hiv, 2022 (Accessed 10 August 2022) Public Health Agency of Canada. Estimates of HIV incidence, prevalence and Canada's progress on meeting the 90-90-90 HIV targets, 2020. https://www.canada.ca/en/public-health/services/publications/diseases-conditions/estimates-hiv-incidence-prevalence-canada-meeting-90-90-90-targets-2020.html#a4.3 (Accessed 19 October 2022) Ripollone JE, Huybrechts KF, Rothman KJ, Ferguson RE, Franklin JM. Evaluating the Utility of Coarsened Exact Matching for Pharmacoepidemiology Using Real and Simulated Claims Data. Am J Epidemiol. 2020;189(6):613-622. https://doi.org/10.1093/AJE/KWZ268 Skowronski DM, Febriani Y, Ouakki M, Setayeshgar S, El Adam S, Zou M et al. Two-dose SARS-CoV-2 vaccine effectiveness with mixed schedules and extended dosing intervals: test-negative design studies from British Columbia and Quebec, Canada. Clin Infect Dis. 2022:ciac290. https://doi.org/10.1093/CID/CIAC290 Skowronski DM, Setayeshgar S, Zou M, Prystajecky N, Tyson JR, Galanis E et al. Single-dose mRNA Vaccine Effectiveness Against Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), Including Alpha and Gamma Variants: A Test-negative Design in Adults 70 Years and Older in British Columbia, Canada. Clin Infect Dis. 2022;74(7):1158-1165.. https://doi.org/10.1093/CID/CIAB616 Ssentongo P, Heilbrunn ES, Ssentongo AE, Advani S, Chinchilli VM, Nunez JJ, Du P. Epidemiology and outcomes of COVID-19 in HIV-infected individuals: a systematic review and meta-analysis. Sci Rep. 2021;11(1):6283. https://doi.org/10.1038/s41598-021-85359-3 Sullivan SG, Tchetgen Tchetgen EJ, Cowling BJ. Theoretical Basis of the Test-Negative Study Design for Assessment of Influenza Vaccine Effectiveness. Am J Epidemiol. 2016;184(5):345-53. https://doi.org/10.1093/AJE/KWW064 Tenforde MW, Patel MM, Ginde AA, Douin DJ, Talbot HK, Casey JD et al. Effectiveness of Severe Acute Respiratory Syndrome Coronavirus 2 Messenger RNA Vaccines for Preventing Coronavirus Disease 2019 Hospitalizations in the United States. Clin Infect Dis. 2022;74(9):1515-1524. https://doi.org/10.1093/CID/CIAB687 Vandenbroucke JP, Brickley EB, Vandenbroucke-Grauls CMJE, Pearce N. A Test-Negative Design with Additional Population Controls Can Be Used to Rapidly Study Causes of the SARS-CoV-2 Epidemic. Epidemiology. 2020 Nov;31(6):836-843. https://doi.org/10.1097/EDE.0000000000001251 Yang X, Sun J, Patel RC, Zhang J, Guo S, Zheng Q et al. Associations between HIV infection and clinical spectrum of COVID-19: a population level analysis based on US National COVID Cohort Collaborative (N3C) data. Lancet HIV. 2021;8(11):e690-e700. https://doi.org/10.1016/S2352-3018(21)00239-3 Appendix Supplementary materials Image, application 1 Image, application 2 Image, application 3 Image, application 4 Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ijid.2022.11.035.	36462571	PMC9711901	NO-CC CODE	2022-12-08 23:16:12	no	Int J Infect Dis. 2022 Dec 1; doi: 10.1016/j.ijid.2022.11.035	utf-8	Int J Infect Dis	2,022	10.1016/j.ijid.2022.11.035	oa_other
==== Front Am J Kidney Dis Am J Kidney Dis American Journal of Kidney Diseases 0272-6386 1523-6838 Published by Elsevier Inc. on behalf of the National Kidney Foundation, Inc. S0272-6386(22)01051-4 10.1053/j.ajkd.2022.10.010 Original Investigations SARS-CoV-2 Vaccine Effectiveness and Breakthrough Infections Among Patients Receiving Maintenance Dialysis Manley Harold J. PharmD 1# Li Nien Chen PhD, MPH 1 Aweh Gideon N. MS 1 Hsu Caroline M. MD 2 Weiner Daniel E. MD 2 Miskulin Dana MD 2 Harford Antonia M. MD 1 Johnson Doug MD 1 Lacson Eduardo Jr. MD, MPH 12 1 Dialysis Clinic Inc., Nashville, TN 2 Division of Nephrology, Tufts Medical Center, Boston, Massachusetts # Correspondence: Harold J. Manley, PharmD, Dialysis Clinic, Inc., 1633 Church Street, Nashville, TN 37203. 1 12 2022 1 12 2022 23 12 2021 11 10 2022 © 2022 Published by Elsevier Inc. on behalf of the National Kidney Foundation, 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. Rationale & Objective SARS-CoV-2 vaccine effectiveness and immunogenicity threshold associated with protection against COVID-19 related hospitalization or death in the dialysis population is unknown. Study Design Retrospective, observational study. Setting & Participants Adult patients receiving maintenance dialysis through a national dialysis provider without COVID-19 history treated between February 1, 2021 and December 18, 2021 with follow up through January 17, 2022. Predictor(s) SARS-CoV-2 vaccination status. Outcome(s) All SARS-CoV-2 infections, composite of hospitalization or death following COVID-19. Analytical Approach Logistic regression was used to determine COVID-19 case rates and vaccine effectiveness. Results Of 16,213 patients receiving dialysis during the study period, 12,278 (76%) were fully vaccinated, 589 (4%) were partially vaccinated and 3,346 (21%) were unvaccinated by the end of follow-up. Of 1,225 COVID-19 cases identified, 550 (45%) occurred in unvaccinated patients, while 891 (73%) cases occurring during the Delta period. Between pre-Delta period and Delta periods vaccine effectiveness against a severe COVID-19 related event (hospitalization or death) was 84% and 70%, respectively. In the subset of 3,202 vaccinated patients with at least one anti-spike IgG assessment, lower anti-spike IgG levels were associated with higher case rates per 10,000 days and adjusted hazard ratios for both infection and COVID-related hospitalization or death. Limitations Observational design, residual biases and confounding may exist. Conclusions Among maintenance dialysis patients, SARS-CoV-2 vaccination was associated with a lower risk of COVID-19 diagnosis and associated hospitalization or death. Among vaccinated patients, low anti-spike IgG level is associated with worse COVID-19 related outcomes. Key Words Dialysis COVID-19 infection breakthrough antibody levels ==== Body pmcIntroduction Patients receiving maintenance dialysis experience significant COVID-19–associated morbidity and mortality.1 , 2 Vaccines are an effective tool for combatting COVID-19, with early studies showing that two doses of a SARS-CoV-2 mRNA vaccine elicit a seroresponse in most (>90%) maintenance dialysis patients, albeit with lower levels than in the general population.3, 4, 5 In late June 2021, the Delta variant became the dominant SARS-CoV-2 strain in the United States.6 At this time, the rate of breakthrough infection among fully vaccinated patients was higher than expected, possibly due to waning antibody concentrations. Tartof et al reported a decrease of protection offered by the BNT162b2/Pfizer vaccine from 93% at baseline to 53% after at least 4 months.7 Among patients receiving maintenance dialysis, a notable proportion (10-15%) did not respond to two doses of an mRNA vaccine; additionally, among initial responders to vaccine, more than half experience waning immunity by 4-6 months, particularly those whose initial response was lesser.5 , 8 The impact of lesser initial vaccine response and subsequent waning antibody levels on clinical outcomes among maintenance dialysis patients is not known. Additionally, the impact of the Delta variant on vaccine effectiveness in this population is unknown. Accordingly, we describe the incidence of COVID-19 diagnoses and COVID-19 related hospitalization or death among unvaccinated, partially vaccinated and fully vaccinated adult dialysis patients during both the pre-Delta and Delta dominant periods. Additionally, among the subset of vaccinated patients with available SARS-CoV-2 vaccine immunoglobulin G spike antibody (anti-spike IgG) titers, we explore the association between antibody levels and clinical outcomes. Methods Study Population Dialysis Clinic, Incorporated (DCI) is the largest non-profit dialysis provider in the United States, operating 260 outpatient dialysis clinics in 29 states. All adult (age ≥18 years), non-transient (e.g. visiting from a non-DCI clinic) maintenance dialysis patients treated between February 1 and December 18, 2021 in DCI clinics contributed time-at-risk, excluding patients with known COVID-19 diagnosis prior to February 1, 2021 or before care at DCI, enrolled in SARS-CoV-2 vaccine trial and/or patients treated for acute kidney injury who were not certified as end-stage renal disease (ESRD). Screening for COVID-19 All maintenance dialysis patients treated at DCI outpatient facilities, including home dialysis and in-center dialysis patients, are screened at each clinic encounter for COVID-19 symptoms (e.g., fever, sore throat, new or worsening cough, shortness of breath, and loss of taste or smell) and any known exposure to infected person(s). Any patient who screens positive is classified as a “Patient Under Investigation” and is tested for SARS-CoV-2 either locally or through the DCI laboratory. Most dialysis patients have multiple comorbid illnesses and have frequent contact with the healthcare system where they are tested even when asymptomatic. Similarly, many patients who are residents of congregate home settings, such as nursing homes, are frequently screened and tested as well. All diagnoses of COVID-19 are entered into DCI’s electronic health record (EHR), primarily based on a report of a positive SARS-CoV-2 test result, which if performed via the DCI laboratory, utilizes a reverse transcriptase–polymerase chain reaction (RT-PCR) Cobas SARS-CoV-2 Assay [Roche Diagnostics]. SARS-CoV-2 Antibody Testing The DCI central laboratory assessed serum anti-spike IgG antibody against the receptor binding domain of the S1 subunit of SARS-CoV-2 spike antigen using a US-FDA-EUA-approved chemiluminescent assay ADVIA Centaur® XP/XPT COV2G between January 1, 2021 and September 30, 2021 and had an Index Value between 0 and ≥20. Since October 1, 2021, the DCI central laboratory used a US-FDA-EUA-approved chemiluminescent assay ADVIA Centaur® sCOVG which reports an Index Value between 0 and ≥100. Excepting upper limit values of the COV2G assay, we are able to convert values measured using the COV2G assay to their equivalent value on the sCOVG assay. The manufacturer defined threshold for a minimum antibody detection level is ≥1 for both assays.9 Physicians may order this test either one time or as part of a clinical testing protocol, but is not measured in all maintenance dialysis patients routinely, per present CDC guidelines.10 The clinical testing protocol follows monthly anti-spike IgG titers using residual blood from routine monthly lab draws until the level is <1 for two consecutive months or for up to 12 consecutive months. Normalized anti-spike IgG index values were converted to World Health Organization standard antibody binding units (BAU) per mL.11 Outcomes All new COVID-19 diagnoses that occurred during the study period were each assigned to the appropriate vaccination status at the time of diagnosis. CDC definitions for “unvaccinated” (to include the period extending up to 13 days after the first vaccine dose), “partially vaccinated (applicable to mRNA vaccines only; defined as the period starting 14 days after the first mRNA vaccine dose up to 13 days after the 2nd mRNA vaccine dose) and “fully vaccinated” (≥14 days after 2 mRNA-1273/Moderna or BNT162b2/Pfizer vaccines or 1 Ad26.COV2.S/Janssen vaccine).3 For all breakthrough cases, defined as a COVID-19 diagnosis in a patient deemed fully vaccinated, the clinic was contacted to verify the reason for SARS-CoV-2 testing: for exposure to a person with known COVID-19 or for positive symptom screening or as required by a non-dialysis clinic/hospital protocol prior to providing a COVID-unrelated service/procedure or as part of protocolized screening (e.g. nursing home). COVID-19 related hospitalizations or deaths were defined by documented primary diagnosis for the episode of care as COVID-19 (ICD-10 code U01.7). COVID-19 cases, hospitalizations and deaths were identified throughout the study period, and further subdivided into pre-COVID-19 delta variant (February 1, 2021 - June 25, 2021) and COVID-19 delta variant dominant periods (June 26, 2021 - December 18, 2021). Those diagnosed with COVID-19 were followed for hospitalization or death through January 17, 2022. Data Extraction All baseline patient demographic and clinical variables used in this analyses were retrospectively obtained from the DCI EHR. These variables include: SARS-CoV-2 vaccine name and dates administered, age, sex, race, ethnicity, US state and county of residence, congregate living status (e.g., nursing home, long term care facility), modality, date of ESRD, body mass index, dialysis dose delivered (Kt/V), serum albumin, hepatitis B surface antibody, immunosuppression (immune-modulating medications, prior transplant, immunodeficiency disorder), substance abuse disorder (tobacco, alcohol or drug), and other comorbid conditions. Any SARS-CoV-2 infection occurring more than 14 days after completing SARS-CoV-2 vaccination was deemed a breakthrough COVID-19 case even if asymptomatic. Data integrity checks for COVID-19 documentation are performed weekly. These include comparing consistency of all available concurrent documentation, including: de novo ICD-10 COVID-19 diagnoses within the problem list, Patient Under Investigation status during symptom screening every visit, new COVID-19 diagnosis screening indicator during every visit, as well as any lab report indicating a positive test for SARS-CoV-2. Inconsistencies detected are referred to corporate nurses who directly communicate with individual clinic staff for resolution. Statistical Analyses Each eligible patient who received dialysis treatment for at least one day during the study period contributed days-at-risk. Patients could move from being unvaccinated (defined by the CDC to include the period extending up to 13 days after the first vaccine dose)3 to partially vaccinated (for mRNA vaccines only; defined by the CDC as the period starting 14 days after the first mRNA vaccine dose up to 13 days after the 2nd mRNA vaccine dose) to fully vaccinated (≥14 days after completing the manufacturer recommended final dose). Patients could contribute days-at-risk for each applicable category unless they experienced a SARS-CoV-2 infection or censoring event (e.g., end of study, non-COVID-related death, transplantation, loss-of-follow-up, received another SARS-CoV-2 vaccination). Thus, a patient might contribute 1 to 3 “sub-period at risk”. For example, consider a patient who was finally fully vaccinated during the period 2/1/2021 to 12/18/2021. Before reaching this status, the patient would have experienced two sub-periods at risk, namely “unvaccinated sub-period” and “partially vaccinated sub-period”. We noticed that the follow up time (i.e., exposure time) was quite different between patients and among vaccination categories. Thus the logarithmic transformation of the follow up time was used as the offset variable in the logistic regression models.12 Case rates and odds ratio (OR) and 95% confidence intervals (CI) from logistic-regression models for COVID infection and the composite outcome of COVID-related hospitalization or death within 30 days of COVID-19 diagnosis were compared, first by vaccine status with unvaccinated patients as the reference group, second by vaccine type among those fully vaccinated. In the logistic regression, the dependent variable is the logit transformation of the probability of event in the sub-period at risk. Both unadjusted and adjusted ORs were derived. The latter (aOR) was derived from the multivariable logistic model was adjusted for age, sex, race, diabetes, dialysis modality, congregate living status, dialysis adequacy, albumin, hepatitis-B seroimmunity, disability, comorbidity, number of and specific comorbidities (diabetes, chronic obstructive pulmonary disease, chronic heart failure, hypertension, peripheral vascular disease, cancer, alcohol or drug abuse), immunocompromise status and the 75th percentile of county COVID infection rate (to account for geotemporal variability in the intensity of the epidemic13, 14 during each study period. Vaccine effectiveness was calculated using the following: (1 – aOR) x 100.15 Vaccine effectiveness for Ad26.COV2.S/Janssen could not be determined due to small sample size and uneven geographic distribution. To assess association of anti-spike IgG titer values with COVID-19 cases or hospitalization/death among patients with breakthrough COVID-19 diagnoses, we evaluated the subset with available anti-spike IgG titer results. Patient anti-spike IgG titer values were included if obtained 7-45 days prior to COVID-19 diagnosis or if the last known post vaccination anti-spike IgG value was <1, a level considered undetectable. Results subsequently were de-identified and aggregated, and the association between the most proximate anti-spike IgG titer result to the identified case and clinical outcomes was evaluated descriptively. In addition, case rates and aOR for COVID-19 infection and composite for COVID-related hospitalization or death within 30 days of COVID-19 diagnosis were calculated for anti-spike IgG values grouped as follows: < 1, 1 – < 2, 2 - < 7, 7 - < 10 and ≥ 10. We selected the various cut-points to compare outcomes in those with undetectable levels of < 1 (reported by assay manufacturer; 45 BAU/mL) or < 2 (internal DCI laboratory validation; 78 BAU/mL), at assay threshold above which COVID-related hospitalization or death were not observed (anti-spike IgG ≥7; 212 BAU/mL) and assay level recently reported as having higher odds for breakthrough infection (anti-spike IgG < 10; 282 BAU/mL).16 This study was reviewed and approved for exemption by the WCG IRB Work Order 1-1456342-1. Statistical analyses were performed using SAS v9.4. Results Among 18,028 maintenance dialysis patients at DCI facilities during the study period, 15,942 (88%) were included (Figure S1). Among eligible patients, 12,403 (78%) were fully vaccinated by December 18, 2021, 6,853 (55%) with mRNA-1273/Moderna, 5,132 (41%) with BNT162b2/Pfizer, 368 (3%) with Ad26.COV2.S/Janssen, and 50 (0.4%) with some combination of either vaccine. An additional 480 (3%) patients were partially vaccinated (276 with mRNA-1273/Moderna and 204 with BNT162b2/Pfizer), while 3,059 (19%) were unvaccinated. Patients mean age (years) and dialysis vintage (months) were 63±15 years old and 43 ± 56 months, respectively. The majority (87%) of patients were receiving in-center hemodialysis, 57% had diabetes and 26% were considered immunocompromised per CDC criteria16 (Table 1 ).Table 1 Patient baseline characteristics at study entry Demographics All Patients (N=15,942) Fully Vaccinated (N=12,403) Partially Vaccinated (N=480) Unvaccinated (N=3,059) P-value Age years (mean ± SD) 63 ± 15 64 ± 14 61 ± 15 59 ± 16 <0.001 Age ≥ 65 years 7,785 (49) 6,379 (51) 207 (43) 1,199 (39) <0.001 Age – Decade <0.001 <55 4,293 (27) 2,952 (24) 156 (32) 1,185 (39) 55-64 3,864 (24) 3,072 (25) 117 (24) 675 (22) 65-74 4,443 (28) 3,613 (29) 119 (25) 701 (23) 75+ 3,352 (21) 2,766 (22) 88 (18) 498 (16) Female 6,708 (42) 5,114 (41) 224 (47) 1,370 (45) <0.001 Race/Ethnicity <0.001 Non-Hispanic Black 5,586 (35) 4,294 (35) 183 (38) 1,109 (36) Hispanic 968 (6) 801 (6) 20 (4) 147 (5) Other 1,083 (7) 925 (7) 21 (4) 137 (4) Unknown 1,400 (9) 1,027 (8) 51 (11) 322 (11) Non-Hispanic White 6,905 (43) 5,356 (43) 205 (43) 1,344 (44) Dialysis Vintage months (mean ± SD) 43 ± 56 44 ± 56 38 ± 51 42 ± 56 0.01 Body Mass Index (kg/m2) (mean ± SD) 29 ± 8 29 ± 8 30 ± 8 29 ± 8 0.06 Congregate Living a 638 (4) 510 (4) 19 (4) 109 (4) 0.38 Home Dialysis 1,953 (13) 1,571 (13) 40 (9) 342 (12) 0.02 Peritoneal Dialysis 1,821 (12) 1,465 (12) 38 (8) 318 (11) Home Hemodialysis 132 (1) 106 (1) 2 (1) 24 (1) Adequate Dialysis Dose b 13,729 (90) 10,962 (91) 388 (85) 2,379 (87) <0.001 Serum Albumin (g/dl) (mean ± SD) 3.8 ± 0.5 3.8 ± 0.4 3.7 ± 0.5 3.7 ± 0.5 <0.001 History Hepatitis B Seroimmunity c 8,921 (58) 7,196 (59) 242 (52) 1,483 (53) <0.001 Potential Immunosuppression 4,066 (26) 3,190 (26) 120 (26) 756 (27) 0.82 Immune-modulating Medications 2,166 (14) 1,689 (14) 64 (14) 413 (15) 0.59 Prior Transplant 1,096 (7) 879 (7) 24 (5) 193 (7) 0.21 Immunodeficiency Disorder 2,880 (19) 2,273 (19) 81 (18) 526 (19) 0.82 Disability 599 (4) 461 (4) 27 (6) 111 (4) 0.08 Tobacco Use 1,602 (10) 1,195 (10) 66 (14) 341 (12) <0.001 Alcohol Abuse Disorder 491 (3) 375 (3) 27 (6) 89 (3) 0.004 Drug Abuse Disorder 382 (3) 257 (2) 19 (4) 106 (4) <0.001 Number of Comorbidities (mean ± SD) 2.9 ± 1.8 2.9 ± 1.8 3.1 ± 1.9 2.8 ± 1.8 0.003 Diabetes Mellitus 8,860 (57) 7,101(58) 288 (62) 1,471 (52) <0.001 Hypertension 12,266 (80) 9,689 (80) 357 (77) 2,220 (79) 0.24 Congestive Heart Failure 3,138 (20) 2,394 (20) 117 (25) 627 (22) <0.001 COPD 2,138 (14) 1,668 (14) 67 (15) 403 (14) 0.68 Stroke/Cerebrovascular Disorder 1,270 (8) 992 (8) 45 (10) 233 (8) 0.48 Peripheral Vascular Disease 1,767 (11) 1,400 (12) 60 (13) 307 (11) 0.36 Thyroid Disorder 2,107 (13) 1,690 (14) 52 (11) 365 (12) 0.01 History of Cancer 1,483 (10) 1,193 (10) 48 (9) 242 (9) 0.11 a Residing in nursing home or long-term care facility b Adequate dialysis defined by hemodialysis single pool Kt/V≥1.2 or peritoneal dialysis weekly Kt/V≥1.7 c Hepatitis B seroimmunity defined as hepatitis B surface antibody ≥ 10 mIU/mL Results presented as count (N) and percent (%) unless otherwise specified. Patients are grouped based on their vaccination status at the end of follow-up. Unvaccinated includes patients who never received a vaccine or recipients of a single dose of a vaccine within 14 days of vaccine receipt; partially vaccinated includes patients who were ≥14 days after the first mRNA vaccine dose but <14 days after the 2nd mRNA vaccine dose; fully vaccinated patients includes all patients ≥14 days after the last vaccine dose. For the table, categories are mutually exclusive. IQR = interquartile range; SD = standard deviation; COPD = Chronic Obstructive Pulmonary Disease Vaccination status and vaccine effectiveness against COVID-19 There were 1,173 documented COVID-19 cases, with 826 (70%) occurring during the Delta period. COVID-19 rates per 10,000 patient days and vaccine effectiveness for each time period are shown in Table 2 . There were 535 (46%) cases that occurred among those considered fully vaccinated, with most of these breakthrough cases (n=511; 96%) occurring during the Delta dominant period; 137 (26%) of patients with breakthrough cases met CDC criteria for immunosuppressed. The median (IQR) follow-up time for all is 57 (28, 198) days. This included time to infection or censoring point for non-event cases. However, among those diagnosed with COVID-19, the median (IQR) from being considered fully vaccinated was 153 (119, 198) days.Table 2 COVID-19 infection rates per 10,000 patient days and vaccine effectiveness between February 1 and December 18, 2021. Number of individuals contributing to time at risk Total Days at Risk Median [IQR] Days at Risk COVID-19 diagnoses Rate per 10,000 days Odds Ratio (95% CI) Vaccine Effectiveness (95% CI) Unadjusted Adjusted February 1 – December 18, 2021 Unvaccinated 14,806 1,467,403 57 [31, 116] 535 3.65 Reference Reference Reference Partially Vaccinated 12,433 398,222 28 [21, 28] 103 2.59 0.67 (0.54, 0.83) 0.67 (0.54, 0.84) 33% Fully Vaccinated 12,403 2,418,643 208[157, 242] 535 2.21 0.58 (0.51, 0.65) 0.55 (0.48, 0.63) 45% BNT162b2/Pfizer 5,132 964,666 199 [148, 234] 237 2.46 0.66 (0.56, 0.77) 0.63 (0.54, 0.75) 37% mRNA-1273/Moderna 6,853 1,377,141 215 [166, 246] 267 1.94 0.52 (0.44, 0.60) 0.50 (0.42, 0.58) 50% Ad26.COV2.S/Janssen 368 71,477 240 [143, 250] 29 4.06 1.13 (0.77, 1.68) February 1- June 19, 2021 – pre-Delta variant period Unvaccinated 12,055 740,332 52 [26, 80] 191 2.58 Reference Reference Reference Partially Vaccinated 10,101 272,431 28 [21, 28] 56 2.06 0.78 (0.58, 1.66) 0.77 (0.56, 1.66) 23% Fully Vaccinated 10,031 716,297 69 [51, 94] 23 0.32 0.12 (0.08, 0.19) 0.12 (0.08, 0.19) 88% BNT162b2/Pfizer 3,987 286,575 69 [40, 101] 9 0.31 0.12 (0.07, 0.21) 0.11 (0.07, 0.23) 89% mRNA-1273/Moderna 5,750 411,262 72 [54, 88] 13 0.32 0.12 (0.06, 0.23) 0.13 (0.06, 0.23) 87% Ad26.COV2.S/Janssen 278 17,478 62 [59, 76] 1 0.57 0.22 (0.03, 1.56) June 20 - December 18, 2021 – Delta variant period Unvaccinated 3,265 422,187 182 [72, 182] 274 6.49 Reference Reference Reference Partially Vaccinated 1,440 70,021 28 [21, 54] 41 5.86 0.86 (0.61, 1.20) 0.90 (0.63, 1.28) 10% Fully Vaccinated 11,647 1,571,118 144 [101, 182] 511 3.25 0.46 (0.39, 0.54) 0.46 (0.39, 0.54) 54% BNT162b2/Pfizer 4,814 619,530 125 [99,182] 227 3.66 0.59 (0.50, 0.70) 0.54 (0.44, 0.66) 46% mRNA-1273/Moderna 6,458 897,918 152 [103, 182] 254 2.83 0.45 (0.38, 0.54) 0.40 (0.33, 0.49) 60% Ad26.COV2.S/Janssen 335 50,056 182 [123, 182] 28 5.59 0.85 (0.57, 1.28) CI = confidence interval; Unvaccinated includes patients who never received a vaccine or recipients of a single dose of a vaccine within 14 days of vaccine receipt; partially vaccinated includes patients who were ≥14 days after the first mRNA vaccine dose but <14 days after the 2nd mRNA vaccine dose; fully vaccinated patients includes all Median days at risk included time to event (i.e., infection) or time to censoring point for non-event cases (e.g., end of study, non-COVID-related death, transplantation, loss-of-follow-up, received another SARS-CoV-2 vaccination). The median time was calculated for all patients. Multivariable logistic model was used to derive odds ratio, adjusted for age, sex, race, diabetes, dialysis modality, congregate living status, dialysis adequacy, albumin, hepatitis-B seroimmunity, disability, comorbidity, number of and specific comorbidities (diabetes, chronic obstructive pulmonary disease, chronic heart failure, hypertension, peripheral vascular disease, cancer, alcohol or drug abuse), immunocompromise status and 75th percentile of county COVID infection rate to account for geotemporal variability in the intensity of the epidemic. Median (IQR) days to breakthrough infection among vaccines was: mRNA-1273/Moderna 174 (138, 222) days; BNT162b2/Pfizer 159 (126, 204) days; and Ad26.COV2.S/Janssen 156 (133, 180) days; p=0.02. Over the entire study period, the COVID-19 case rate was significantly lower among fully vaccinated than among unvaccinated patients: 2.21 vs 3.65 per 10,000 patient days [aOR 0.55 (0.48, 0.63)](Table 2). Vaccine effectiveness was 45% overall, with mRNA-1273/Moderna vaccine having highest vaccine effectiveness at 50% followed by BNT162b2/Pfizer at 37%. During the Delta period, across all vaccination status groups, COVID-19 case rates increased. Fully vaccinated patients had lower COVID-19 case rates (3.25 vs. 6.49 per 10,000 patient days [aOR 0.46 (0.39, 0.54)]), with 54% vaccine effectiveness compared to unvaccinated patients; the mRNA-1273/Moderna vaccine had the highest vaccine effectiveness at 60%. Vaccine effectiveness against COVID-19 related hospitalization or death There were 424 COVID-19 related hospitalizations/deaths during the study period, including 112 COVID-19 related deaths, with 60 deaths among unvaccinated, 5 among partially vaccinated, and 47 among fully vaccinated patients (mRNA-1273/Moderna N=22; BNT162b2/Pfizer N=22; A26.COV2.S/Janssen N=3). Of the 112 COVID-related deaths, 33 (30%) COVID-related deaths occurred in immunocompromised patients. During the overall study period, the incidence of COVID-19 related hospitalization/death was 1.45 per 10,000 patient days among unvaccinated and 0.78 per 10,000 patient-days among vaccinated patients [aOR 0.47 (0.38, 0.58)]. For fully vaccinated patients, the vaccine effectiveness was 53% overall (63% with mRNA-1273/Moderna and 39% with BNT162b2/Pfizer against hospitalization or death. (Table 3 ).Table 3 COVID-related hospitalization/death rates per 10,000 patient days and vaccine effectiveness between February 1 and December 18, 2021. Number of individuals contributing to time at risk Total Days at Risk Median [IQR] Days at Risk Events within 30 days COVID-19 diagnosis Rate per 10,000 days Odds Ratio (95% CI) Vaccine Effectiveness (95% CI) Unadjusted Adjusted February 1 – December 18, 2021 Unvaccinated 14,806 1,404,729 54 [30, 98] 200 1.45 Reference Reference Reference Partially Vaccinated 12,433 396,399 28 [21, 28] 41 1.03 0.70 (0.50, 0.98) 0.67 (0.47, 0.95) 33%† Fully Vaccinated 12,403 2,346,397 207 [153, 242] 183 0.78 0.51 (0.41, 0.63) 0.47 (0.38, 0.58) 53% BNT162b2/Pfizer 5,132 933,674 198 [142, 234] 91 0.97 0.67 (0.52, 0.86) 0.61 (0.47, 0.79) 39% mRNA-1273/Moderna 6,853 1,339,710 214 [161, 245] 79 0.59 0.40 (0.31, 0.52) 0.37 (0.28, 0.49) 63% Ad26.COV2.S/Janssen 368 67,853 240 [123, 250] 13 1.92 1.35 (0.76, 2.39) February 1- June 19, 2021 – pre-Delta variant period Unvaccinated 12,055 737,051 52 [27, 79] 62 0.84 Reference Reference Reference Partially Vaccinated 10,101 272,714 28 [21, 28] 20 0.73 0.87 (0.52, 1.44) 0.89 (0.52, 1.50) N/A Fully Vaccinated 10,031 715,583 69 [51, 94] 11 0.15 0.17 (0.09, 0.33) 0.16 (0.08, 0.34) 84% BNT162b2/Pfizer 3,987 286,045 69 [40, 101] 6 0.21 0.25 (0.11, 0.57) 0.22 (0.09, 0.55) 78% mRNA-1273/Moderna 5,750 411,078 72 [54, 88] 4 0.10 0.11 (0.04, 0.32) 0.13 (0.05, 0.35) 87% Ad26.COV2.S/Janssen 278 17,478 62 [59, 76] 1 0.57 0.68 (0.09, 4.91) June 20 - December 18, 2021 – Delta variant period Unvaccinated 3,265 402,236 182 [57, 182] 123 3.06 Reference Reference Reference Partially Vaccinated 1,440 68,762 28 [21, 50] 18 2.62 0.84 (0.51, 1.38) 0.80 (0.48, 1.34) N/A Fully Vaccinated 11,647 1,534,287 143 [99, 182] 171 1.11 0.34 (0.27, 0.43) 0.30 (0.23, 0.38) 70% BNT162b2/Pfizer 4,814 604,472 124 [96,182] 84 1.39 0.44 (0.33, 0.58) 0.39 (0.29, 0.53) 61% mRNA-1273/Moderna 6,458 878,136 151 [101, 182] 75 0.85 0.27 (0.20, 0.36) 0.23 (0.17, 0.32) 77% Ad26.COV2.S/Janssen 335 48,205 182 [114, 182] 12 2.49 0.81 (0.44, 1.48) CI = confidence interval; * = Although patients can contribute time to any vaccination status, the N in the first column refers to patients' status at the end of follow-up. Unvaccinated includes patients who never received a vaccine or recipients of a single dose of a vaccine within 14 days of vaccine receipt; partially vaccinated includes patients who were ≥14 days after the first mRNA vaccine dose but <14 days after the 2nd mRNA vaccine dose; fully vaccinated patients includes all patients ≥14 days after the last vaccine dose.; †= patient hospitalized for COVID-19 prior to receipt of second mRNA vaccine.; N/A = not applicable Median days at risk included time to event (i.e., infection) or time to censoring point for non-event cases (e.g., end of study, non-COVID-related death, transplantation, loss-of-follow-up, received another SARS-CoV-2 vaccination). The median time was calculated for all patients. Multivariable logistic model was used to derive odds ratio, adjusted for age, sex, race, diabetes, dialysis modality, congregate living status, dialysis adequacy, albumin, hepatitis-B seroimmunity, disability, comorbidity, number of and specific comorbidities (diabetes, chronic obstructive pulmonary disease, chronic heart failure, hypertension, peripheral vascular disease, cancer, alcohol or drug abuse), immunocompromise status and 75th percentile of county COVID infection rate to account for geotemporal variability in the intensity of the epidemic. Among all vaccination status groups, both COVID-19 case rates and vaccine effectiveness against COVID-19 related hospitalization/death worsened during the Delta period. In the model comparing patients by vaccination status, fully vaccinated patients had lowest case rate per 10,000 patient days and aOR for COVID-19 infection and related hospitalization/death in both pre-Delta and Delta dominant periods. In the model comparing both vaccination status and vaccine types, patients fully vaccinated with mRNA-1273/Moderna experienced lowest case rate per 10,000 patient days and highest vaccine effectiveness against COVID-19 related hospitalization/death in both pre-Delta and Delta variant dominant periods. Anti-spike IgG level and breakthrough COVID-19 Anti-spike IgG levels were available in 3,152 (20%) patients over the study period. Compared to anti-spike IgG levels ≥10 (282 BAU/ml), each lower anti-spike IgG threshold evaluated was associated with higher case rates and aORs for infection (Table 4 ). Anti-spike IgG levels < 1 were significantly associated with COVID-related hospitalization or death (1.43 [0.67, 3.05]; Table 5 ).Table 4 Association of peri-infection anti-spike IgG values with risk for COVID diagnosis. Anti-spike IgG threshold Number of individuals contributing to time at risk Person Days at Risk COVID-19 diagnoses Rate per 10,000 days Odds Ratio (95% CI) Unadjusted Adjusted < 1 2,515 164,908 55 3.34 2.60 (1.72, 3.94) 2.41 (1.46, 3.98) 1 – < 2 854 38,964 15 3.85 2.97 (1.62, 5.42) 2.60 (1.22, 5.56) 2 - < 7 1,844 67,836 23 3.39 2.58 (1.54, 4.35) 2.49 (1.37, 4.53) 7 - < 10 528 21,302 5 2.35 1.78 (0.70, 4.56) 1.28 (0.39, 4.24) ≥ 10 7,042 293,817 39 1.33 Reference Reference Total 12,783 586,827 137 2.33 Each anti-spike IgG group odds ratio was adjusted using a multivariable logistic model that included age, sex, race, diabetes, dialysis modality, congregate living status, dialysis adequacy, albumin, hepatitis-B seroimmunity, disability, comorbidity, number of and specific comorbidities (diabetes, chronic obstructive pulmonary disease, chronic heart failure, hypertension, peripheral vascular disease, cancer, alcohol or drug abuse), immunocompromise status and 75th percentile of county COVID infection rate to account for geotemporal variability in the intensity of the epidemic. Table 5 Association of peri-infection anti-spike IgG values with risk for COVID-related hospitalization or death. Anti-spike IgG threshold Number of individuals contributing to time at risk Person Days at Risk Events within 30 days COVID-19 diagnosis Rate per 10,000 days Odds Ratio (95% CI) Unadjusted Adjusted < 1 2,515 165,750 21 1.27 1.98 (1.06, 3.70) 1.43 (0.67, 3.05) 1 – < 2 854 39,037 6 1.54 2.39 (0.95, 6.03) 2.50 (0.90, 6.92) 2 - < 7 1,844 68,312 4 0.59 0.90 (0.31, 2.66) 0.50 (0.12, 2.19) 7 - < 10 528 21,344 2 0.94 1.45 (0.34, 6.27) 0.77 (0.10, 5.82) ≥ 10 7,042 293,652 19 0.65 Reference Reference Total 12,783 588,095 52 0.88 Each anti-spike IgG group odds ratio was adjusted using a multivariable logistic model that included age, sex, race, and 75th percentile of county COVID infection rate to account for geotemporal variability in the intensity of the epidemic. Among the population with antibody assessment, there were 137 breakthrough cases (BNT162b2/Pfizer N=53, mRNA-1273/Moderna N=52, and Ad26.COV2.S/Janssen N=32) with the level most proximate to diagnosis measured a median (IQR) of 25 (12, 55) days prior to COVID-19 diagnosis. The frequency of COVID-19 cases requiring hospitalization across anti-spike IgG and BAU/mL levels is shown in Figure 1 A and 1B, respectively. Nearly half of breakthrough cases (67 of 137; 49%) and the majority of COVID-related hospitalizations (27 of 52; 52%) occurred when the anti-spike IgG level was undetectable. The majority of COVID-19 cases (109 of 137; 80%) and COVID-related hospitalizations (45of 52; 87%) occurred at an anti-spike IgG level < 10 (282 BAU/mL). Among 20 COVID-19 related deaths, anti-spike IgG levels in patients were undetectable in 12 (60%), between 1 and 10 (45-282 BAU/mL) in 5 (25%) and ≥10 (≥ 282 BAU/mL) in 3 (15%) of patients. The majority of COVID-related hospitalizations (50 of 52; 96%) and COVID-related deaths (18 of 20; 90%) occurred at BAU/mL level < 400.Figure 1 Patient anti-spike IgG value (panel A) and BAU/mL (panel B) at time of COVID diagnosis (N=137) Discussion SARS-CoV-2 vaccines are highly effective in maintenance dialysis patients, with lower risk for COVID-19 cases and COVID-19 related hospitalization or death among those who are fully vaccinated. Breakthrough COVID-19 cases and COVID-19 related hospitalizations or death among dialysis patients increased when Delta variant became dominant, even among those considered fully vaccinated. Overall, while vaccines remained protective, vaccine effectiveness during the Delta dominant period was approximately 39% lower than that observed in the pre-Delta period for breakthrough COVID-19 infection. The lower vaccine effectiveness may reflect weaker antibody production or T-cell response among maintenance dialysis patients.18 Although anti-spike IgG antibodies decline over time in vaccinated patients,8 vaccine effectiveness in preventing COVID-19 infection and hospitalization/death remains high when compared to unvaccinated patients. For COVID-related hospitalization, vaccine effectiveness was similar between pre-Delta and Delta periods and remained high when compared to unvaccinated patients. There may be differences in outcomes by vaccine type. Overall and during each time period evaluated, vaccine effectiveness against COVID-related hospitalization or death was highest with mRNA-1273/Moderna thanBNT162b2/Pfizer when compared to those unvaccinated. In our study population Ad26.COV2.S/Janssen vaccine was not only associated with lower antibody response,5 , 8 but likely was associated with higher breakthrough and COVID-19 related hospitalization rates than the mRNA vaccines, particularly when compared to the mRNA-1273/Moderna vaccine. While the CDC did not specifically designate dialysis patients as immunocompromised persons who should receive routine administration of a third COVID-19 mRNA vaccine dose, they cited dialysis patients as a possible immunocompromised group where clinical judgment is important.17 The CDC recommends that moderately to severely immunocompromised patients receive a third dose of mRNA vaccine four weeks after the second mRNA vaccine dose or four weeks after the first dose of Ad26.COV2.S/Janssen vaccine; an additional first booster dose should be administered three months after the third dose of vaccine. Alternatively, if not considering dialysis patients immunocompromised, a booster dose of vaccine should be administered 4-5 months after their second mRNA vaccine dose or 2 months after their initial adenovirus vector vaccine.17 Recognizing that not all maintenance dialysis patients produce anti-spike IgG antibodies to the same degree and that antibodies decline over time, the administration of additional vaccine doses should be not arbitrarily be based on time. Since anti-spike IgG antibody titers correlate with COVID neutralizing titers and clinical efficacy,19 , 20 many clinicians associate detectable antibodies with clinical protection. Presently the CDC and FDA does not recommend using COVID-19 antibody testing to guide clinical decision-making. 21 , 22 However, adopting a test and treat approach with routine measurement of anti-spike IgG levels followed by additional doses of vaccine as needed to maintain adequate antibody levels may be warranted, although requires confirmation. Recently Anand et al. similarly reported that anti-spike IgG value < 10 were associated with higher odds for breakthrough infection.16 In our results (Tables 4 and 5), a vast majority of COVID-19 cases (109 of 137; 80%) and COVID-related hospitalizations (45 of 52; 87%) occurred at an anti-spike IgG level < 10 (< 282 BAU/mL). This approach, where vaccine administration is predicated on maintaining antibody levels, has been well-demonstrated with hepatitis B vaccination among dialysis patients.23 Study strengths include the national population of a mid-size dialysis provider in the US with real world clinical outcomes. However, there are limitations associated with this study. Due to the observational design, residual biases (e.g., misclassification of vaccine exposure in patients vaccinated outside the clinic, inability to identify all asymptomatic infections or those identified outside facility, reasons for unvaccinated status) and confounding may exist. The electronic health records do not contain standardized documentation of COVID-19 symptoms and therefore we could not estimate vaccine effectiveness with regard to mitigating or tempering symptom severity. Individual patient adherence to mask and social distancing recommendations is not known, and there was a relatively low number of patients with antibody titer measurements and infection or hospitalization. We did not know the specific SARS-CoV-2 variant for each infection and attributed all infections to the Delta variant after June 19, 2021. Finally, our study did not include the Omicron surge. In conclusion, SARS-CoV-2 vaccines were effective in maintenance dialysis patients, reducing the risk of both COVID-19 cases and COVID-related hospitalization or death during pre-Delta and Delta dominant periods. COVID-19 cases surged during the Delta variant dominant period and current immunosuppression criteria are limited in identifying dialysis patients at highest breakthrough risk. Further research is needed to evaluate SARS-CoV-2 vaccine effectiveness and the utility of antibody titer monitoring to determine patients at highest risk for COVID-19 and to guide the timing of additional vaccine administration. Article Information Authors’ Contributions: Research idea and study design: HJM, EKL, GA, NCL, DEW, DCM, AMH, DSJ; data acquisition: HJM, GA, NCL, EKL; data analysis/interpretation: HJM, GA, NCL, DEW, DCM, CMH, EKL, AMH. Each author contributed important intellectual content during manuscript drafting or revision and agrees to be personally accountable for the individual’s own contributions and to ensure that questions pertaining to the accuracy or integrity of any portion of the work, even one in which the author was not directly involved, are appropriately investigated and resolved, including with documentation in the literature if appropriate. Support: This report was supported by Dialysis Clinic, Inc. Dialysis Clinic, Inc. had no role in study design, data collection, analysis, reporting, or the decision to submit for publication. CMH receives support from ASN Kidney Cure’s Ben J. Lipps Research fellowship. CMH’s funder had no role in study design, data collection, reporting, or the decision to submit. Supplementary Material Plain-language summary: SARS-CoV-2 vaccine effectiveness and the association between antibody levels and severe COVID-19 clinical outcomes (i.e., hospitalization or death) among maintenance dialysis patients is poorly defined. Between February 1-June18, 2021 (pre-Delta period) and June 19-December 18, 2021 (Delta period) vaccine effectiveness against a severe COVID events was 84% and 70%, respectively. Greater time since full vaccination status was associated with higher risk for severe COVID events. Anti-spike IgG level ≥ 10 was associated with lower risk of COVID-19 diagnosis and COVID-related hospitalization or death. Barring changes associated with new SARS-CoV-2 variants, our findings demonstrate high effectiveness of SARS-CoV-2 vaccines and suggest that monitoring SARS-CoV-2 antibody levels and administering additional vaccine doses to maintain adequate immunity may be beneficial. Financial Disclosure: Dr. Manley, Mr. Aweh, Dr. Nien Chen Li, Dr. Harford, Dr. Johnson and Dr. Lacson Jr are all employees of DCI, where Dr. Johnson is Vice Chair of the Board. Dr. Weiner and Dr. Miskulin receive salary support to their institution from DCI. Dr. Hsu declares that she has no relevant financial interests. Prior Presentation: A preprint version of this article was posted December 21, 2021 at medRxiv with doi 10.1101/2021.12.20.21268124. Peer Review: Received December 23, 2021. Evaluated by 2 external peer reviewers, with direct editorial input from a Statistics/Methods Editor, an Associate Editor, and a Deputy Editor who served as Acting Editor-in-Chief. Accepted in revised form October 11, 2022. The involvement of an Acting Editor-in-Chief was to comply with AJKD’s procedures for potential conflicts of interest for editors, described in the Information for Authors & Journal Policies. ==== Refs References 1 Hsu C.M. Weiner D.E. Aweh G. Miskulin D.C. Manley H.J. Stewart C. COVID-19 Among US Dialysis Patients: Risk Factors and Outcomes From a National Dialysis Provider Am J Kidney Dis 77 5 2021 748 756 e741 33465417 2 Hsu C.M. Weiner D.E. Aweh G. Salenger P. Johnson D.S. Lacson E. Epidemiology and Outcomes of COVID-19 in Home Dialysis Patients Compared with In-Center Dialysis Patients J Am Soc Nephrol 32 7 2021 1569 1573 10.1681/asn.2020111653 34108232 3 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. Accessed September 18, 2021 4 Grupper A. Sharon N. Finn T. Cohen R. Israel M. Agbaria A. Humoral Response to the Pfizer BNT162b2 Vaccine in Patients Undergoing Maintenance Hemodialysis Clin J Am Soc Nephrol 16 7 2021 1037 1042 10.2215/cjn.03500321 33824157 5 Lacson E, Argyropoulos CP, Manley HJ, Aweh G, Chin AI, Salman LH, et al.: Immunogenicity of SARS-CoV-2 Vaccine in Dialysis. medRxiv, 2021 doi:10.1101/2021.04.08.21254779 6 Centers for Disease Control and Prevention: COVID Data Tracker. Available at: https://covid.cdc.gov/covid-data-tracker/#variant-proportions. Accessed September 7, 2021 7 Tartof S.Y. Slezak J.M. Fischer H. Hong V. Ackerson B.K. Ranasinghe O.N. Effectiveness of mRNA BNT162b2 COVID-19 vaccine up to 6 months in a large integrated health system in the USA: a retrospective cohort study Lancet 398 10309 2021 1407 1416 10.1016/s0140-6736(21)02183-8 34619098 8 Hsu C.M. Weiner D.E. Aweh G.N. Manley H.J. Ladik V. Frament J. Seroresponse to SARS-CoV-2 Vaccines Among Maintenance Dialysis Patients Am J Kidney Dis 79 2 2022 307 310 10.1053/j.ajkd.2021.10.002 34758369 9 Siemens Healthcare Diagnostics Inc.: COV2G, ADVIA Centaur XP and ADVIA Centaur CPT Systems. Available at: https://www.fda.gov/media/140704/download. Accessed September 17, 2021 10 Centers for Disease Control and Prevention: Interim Guidelines for COVID-19 Antibody Testing. Available at: https://www.cdc.gov/coronavirus/2019-ncov/lab/resources/antibody-tests-guidelines.html Accessed December 12, 2021 11 Freeman J, Conklin J. Standardization of two SARS-CoV-2 serology assays to the WHO 20/136 human standard reference material. J Virol Methods. 2022;300:114430. doi: 10.1016/j.jviromet.2021.114430. Epub 2021 Dec 13. PMID: 34915088; PMCID: PMC8667347. 12 Casualty Actuarial Society E-Forum -Winter 2009: Yan J, Guszcza J, Flynn M, Wu CSP. Applications of the Offset in Property-Casualty Predictive Modeling/ Available at: https://www.casact.org/sites/default/files/database/forum_09wforum_completew09.pdf Accessed September 23, 2022 13 USAFacts: US COVID-19 cases and deaths by state. Available at: https://usafacts.org/visualizations/coronavirus-covid-19-spread-map/. Accessed November 12, 2021 14 Centers for Disease Control and Prevention: COVID-19 Vaccinations in the United States, County. Available at: https://data.cdc.gov/Vaccinations/COVID-19-Vaccinations-in-the-United-States-County/8xkx-amqh. Accessed November 12, 2021 15 CDC: Measures of risk: Vaccine efficacy or vaccine effectiveness, 2012. Available at:https://www.cdc.gov/csels/dsepd/ss1978/lesson3/section6.html. Accessed August 25, 2022. 16 Anand S. Montez-Rath M.E. Han J. Garcia P. Cadden L. Hunsader P. SARS-CoV-2 Vaccine Antibody Response and Breakthrough Infection in Patients Receiving Dialysis Ann Intern Med 175 3 2021 371 378 10.7326/M21-4176 34904856 17 Centers for Disease Control and Prevention: COVID-19 Vaccines for Moderately to Severely Immunocompromised People. Available at: https://www.cdc.gov/coronavirus/2019-ncov/vaccines/recommendations/immuno.html. Accessed April 19, 2022 18 Karakizlis H. Nahrgang C. Strecker K. Chen J. Aly M. Slanina H. Schüttler C.G. Esso I. Wolter M. Todorova D. Jessen S. Adamik A. Ronco C. Seeger W. Weimer R. Sester M. Birk H.W. Husain-Syed F. Immunogenicity and reactogenicity of homologous mRNA-based and vector-based SARS-CoV-2 vaccine regimens in patients receiving maintenance dialysis Clin Immunol 236 2022 108961 10.1016/j.clim.2022.108961 Epub 2022 Feb 25. PMID: 35227871; PMCID: PMC8875769 19 Earle K.A. Ambrosino D.M. Fiore-Gartland A. Goldblatt D. Gilbert P.B. Siber G.R. Evidence for antibody as a protective correlate for COVID-19 vaccines Vaccine 39 32 2021 4423 4428 10.1016/j.vaccine.2021.05.063 34210573 20 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 7 2021 1205 1211 10.1038/s41591-021-01377-8 34002089 21 Centers for Disease Control and Prevention: Summary Document for Interim Clinical Considerations for Use of COVID-19 Vaccines Currently Authorized in the United States. Available at: https://www.cdc.gov/vaccines/covid-19/downloads/summary-interim-clinical-considerations.pdf. Accessed April 19, 2022 22 US Food and Drug Administration: Antibody Testing Is Not Currently Recommended to Assess Immunity After COVID-19 Vaccination: FDA Safety CommunicationAvailable at: https://www.fda.gov/medical-devices/safety-communications/antibody-testing-not-currently-recommended-assess-immunity-after-covid-19-vaccination-fda-safety. Accessed April 19, 2022 23 Freedman M.S. Ault K. Bernstein H. Advisory Committee on Immunization Practices Recommended Immunization Schedule for Adults Aged 19 Years or Older — United States, 2021 MMWR Morb Mortal Wkly Rep 70 6 2021 193 196 33571173	36462570	PMC9711902	NO-CC CODE	2022-12-02 23:21:31	no	Am J Kidney Dis. 2022 Dec 1; doi: 10.1053/j.ajkd.2022.10.010	utf-8	Am J Kidney Dis	2,022	10.1053/j.ajkd.2022.10.010	oa_other
==== Front iScience iScience iScience 2589-0042 The Author(s). S2589-0042(22)01963-0 10.1016/j.isci.2022.105690 105690 Article mRNA vaccines elicit potent neutralization against multiple SARS-CoV-2 Omicron subvariants and other variants of concern Wang Gang 1# Shi Juan 1# Verma Abhishek K. 2 Guan Xiaoqing 1 Perlman Stanley 23 Du Lanying 1∗ 1 Institute for Biomedical Sciences, Georgia State University, Atlanta, GA, USA 2 Department of Microbiology and Immunology 3 Department of Pediatrics, University of Iowa, Iowa City, IA, USA ∗ Correspondence and Lead Contact: (L.D.). # These authors contributed equally to this article. 1 12 2022 1 12 2022 1056905 9 2022 25 10 2022 25 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. SARS-CoV-2 variants of concern (VOCs) have shown resistance to vaccines targeting the original virus strain. An mRNA vaccine encoding the spike protein of Omicron BA1 (BA1-S-mRNA) was designed, and its neutralizing activity, with or without the original receptor-binding domain (RBD)-mRNA, was tested against SARS-CoV-2 VOCs. First-dose of BA1-S-mRNA followed by two-boosts of RBD-mRNA elicited potent neutralizing antibodies (nAbs) against pseudotyped and authentic original SARS-CoV-2; pseudotyped Omicron BA1, BA2, BA2.12.1 and BA5 subvariants, and Alpha, Beta, Gamma and Delta VOCs; authentic Omicron BA1 subvariant and Delta VOC. By contrast, other vaccination strategies, including RBD-mRNA first-dose plus BA1-S-mRNA two-boosts, RBD-mRNA or BA1-S-mRNA three-doses, or their combinations, failed to elicit high nAb titers against all of these viruses. Overall, this vaccination strategy was effective for inducing broadly and potent nAbs against multiple SARS-CoV-2 VOCs, particularly Omicron BA5, and may guide the rational design of next-generation mRNA vaccines with greater efficacy against future variants. Graphical abstract Keywords Coronavirus COVID-19 SARS-CoV-2 Variants of concern Omicron subvariants Neutralizing antibodies ==== Body pmc	36471872	PMC9711903	NO-CC CODE	2022-12-13 23:17:22	no	iScience. 2022 Dec 22; 25(12):105690	utf-8	iScience	2,022	10.1016/j.isci.2022.105690	oa_other
==== Front J Glob Antimicrob Resist J Glob Antimicrob Resist Journal of Global Antimicrobial Resistance 2213-7165 2213-7173 The Author(s). Published by Elsevier Ltd on behalf of International Society for Antimicrobial Chemotherapy. S2213-7165(22)00258-2 10.1016/j.jgar.2022.11.011 SARS-CoV-2 Dispatches Effect of Farnesyltransferase Inhibitors on SARS-CoV-2 Weber Lea 1‡ Mautner Lena 2‡ Hoyos Mona 2 Ehrhardt Anja 3 Baiker Armin 2 Bachmann Hagen Sjard 1⁎ 1 Institute of Pharmacology and Toxicology, Centre for Biomedical Education and Research (ZBAF), Witten/Herdecke University, Witten, Germany 2 Bayerisches Landesamt für Gesundheit und Lebensmittelsicherheit (LGL), Oberschleißheim, Germany 3 Virology and Microbiology, Centre for Biomedical Education and Research (ZBAF), Witten/Herdecke University, Witten, Germany ⁎ Corresponding author: Hagen Sjard Bachmann, Institute of Pharmacology and Toxicology, Faculty of Health, Witten/Herdecke University, Stockumer Str. 10, 58453 Witten (Germany) ‡ Authors contributed equally 1 12 2022 1 12 2022 12 4 2022 6 10 2022 18 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 SARS-CoV-2 in 2019 led to a severe pandemic situation. Treatment options are limited and the efficacy of vaccines decreases due to mutations in SARS-CoV-2 strains. Therefore, new treatment options are urgently needed, and computational compound screenings are used to predict drugs quickly. One of these screenings revealed farnesyltransferase inhibitors (FTIs) as potential candidates. We demonstrated that the FTIs lonafarnib and tipifarnib have an impact on SARS-CoV-2 Wildtype and the Delta variant. Both FTIs dose-dependently reduced morphological changes and the formation of cytopathic effects in SARS-CoV-2 infected Calu-3 cells. Editor: Prof Carlo F Perno ==== Body pmcIntroduction The SARS-CoV-2 pandemic has necessitated the fast development of vaccines and new treatment options. Computational compound screenings are a common method to identify new drugs quickly. Such screenings revealed that the FTIs lonafarnib und tipifarnib might have antiviral potency against SARS-CoV-2 [1] [2]. Both are farnesyltransferase inhibitors (FTIs), finding application in different cancer therapies and Hepatitis D virus infection [3]. Due to such predictions based on computational screenings, we investigated the putative effect of that FTIs on different variants. Here, both FTIs reduced the viral replication in Wildtype and Delta. The impact of the FTIs on Omicron needs to be further elucidated due to inefficient viral replication. Results We infected Calu-3 cells with Wildtype, Delta or Omicron variants, treated them with FTIs and evaluated the morphological changes and cytopathic effects (CPE). The infection of Calu-3 cells with Wildtype and Delta led to massive morphological changes and cytopathic effects (CPE), which were dose-dependently reduced after FTI treatment (fig. 1 A). Both FTIs show stronger reductions of the CPE for the Delta variant compared to the Wildtype variant (fig. 1 A). The effect of remdesivir used as positive control was less pronounced (fig. 1 A).Fig. 1 (A) Representative morphological images of SARS-CoV-2 infected Calu-3 cells after treatment with lonafarnib and tipifarnib for 30 h. Cells infected with Wildtype, Delta or Omicron were treated with lonafarnib and tipifarnib (0.1 µM, 20 µM and 30 µM), DMSO and Remdesivir (2 µM) as controls. Scale bar: 300 µm. (B)-(C) Cell viability test on Calu-3 cells after treatment with lonafarnib (B) and tipifarnib (C). Cells were incubated with lonafarnib and tipifarnib in different concentrations, ranging from 0.1 µM – 50 µM for 48 h. (D)-(I) Effect of FTIs on fold change viral replication of SARS-CoV-2 on Calu-3 cells. Effect of lonafarnib (left) and tipifarnib (right) on Wildtype (D, E), Delta (F, G) and Omicron (H, I). A triangle for the fold-change in viral replication for remdesivir was added in (D) – (I). Fig 1 After the optical evaluation, we quantified the effect of the FTIs on the virus replication in Calu-3 cells. We first elucidated the impact of the FTIs on the cells using MTT assays (fig. 1B). Both FTIs evoked a dose-dependent reduction in cell viability, which was more prominent in higher concentrations. The CC50 values were 31.28 for lonafarnib and 37.78 for tipifarnib. After the evaluation of the virus-independent effects of the FTIs, we infected Calu-3 with Wildtype, Delta or Omicron, treated them with FTIs and calculated the fold change viral replication. To address this question, we measured extracellular viral RNA (vRNA) by RT-qPCR. We observed a dose-dependent inhibitory effect of both FTIs on Wildtype and Delta. In Wildtype, lonafarnib and tipifarnib showed IC50-values of 3.978 and 4.362, respectively. In Delta, lonafarnib and tipifarnib showed IC50-values of 6.024 and 7.287, respectively (fig. 1D-G). The effect of the FTIs on Omicron was not assessable due to the insufficient viral replication. Remdesivir was used as a control drug to reduce viral replication similar to IC50 values of previous studies at 2 µM [4]. As displayed in fig. 1 D-I remdesivir reduces but not completely inhibits viral replication. Discussion Here, we aimed to characterize the effect of the FTIs lonafarnib and tipifarnib on SARS-CoV-2 Wildtype, Delta and Omicron infected cells. We demonstrated a dose-dependent antiviral effect of both FTIs on Wildtype and Delta, whereas the effect on Omicron was not assessable due to insufficient infection of Calu-3. The reduction of the CPE on infected Calu-3 cells corresponds to the quantification of the fold change viral replication via RT-qPCR. Drug repurposing is commonly used in order to identify drugs against rapidly spreading diseases and this approach fastens the approval of drugs and uses existing production capacities quickly [5]. In a high-throughput virtual screening approach [2], lonafarnib was identified as a potential drug candidate, which might disrupt the binding of the viral cofactors NSP7 and NSP8 to the highly active NSP12 polymerase complex [2]. In another computational screening of DrugBank compounds lonafarnib was capable of simultaneously inhibiting the three viral targets 3CLpro, PLpro and RdRp [1]. However, to the best of our knowledge, lonafarnib has never been tested in cell culture approaches on SARS-CoV-2. In this study, we demonstrated an antiviral effect on viral replication of both Wildtype and Delta variants, whereas the effect on Omicron was not measurable. The second FTI, tipifarnib, was found to be potentially effective against SARS-CoV-2 in a virtual screening [4]. Furthermore, its antiviral activity against the Wildtype strain was validated in Vero cells and Calu-3 cells, showing even synergistic effects when combined with omipalisib and remdesivir [4]. We confirmed these findings and, furthermore, demonstrated similar antiviral effects on Delta. Here, the IC50 value for tipifarnib (7.29) is in accordance with the IC50 value generated by Jang et al. in Wildtype (11.01) [4]. Acknowledgment We thank Anna Arnold for technical support during experiments. Conflict of interest 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 The authors would like to acknowledge the financial support of the Bavarian State Ministry of the Environment and Consumer Protection to LM and MH. List of Abbreviations DMSO – Dimethyl sulfoxide FTI – Farnesyltransferase inhibitor Wildtype - SARS-CoV-2 non-VOC/B.1.1 Delta - SARS-CoV-2 Delta/B.1.617.2 Omicron - SARS-CoV-2 Omicron/B.1.1.529 Literature 1. Murugan, N.A., et al., Searching for target-specific and multi-targeting organics for Covid-19 in the Drugbank database with a double scoring approach. Sci Rep, 2020. 10(1): p. 19125. 2. Ruan, Z., et al., SARS-CoV-2 and SARS-CoV: Virtual screening of potential inhibitors targeting RNA-dependent RNA polymerase activity (NSP12). J Med Virol, 2021. 93(1): p. 389-400. 3. Koh, C., et al., Oral prenylation inhibition with lonafarnib in chronic hepatitis D infection: a proof-of-concept randomised, double-blind, placebo-controlled phase 2A trial. The Lancet. Infectious diseases, 2015. 15(10): p. 1167-1174. 4. Jang, W.D., et al., Drugs repurposed for COVID-19 by virtual screening of 6,218 drugs and cell-based assay. Proc Natl Acad Sci U S A, 2021. 118(30). 5. Ellinger, B., et al., A SARS-CoV-2 cytopathicity dataset generated by high-content screening of a large drug repurposing collection. Sci Data, 2021. 8(1): p. 70. 6. Mautner, L., et al., Rapid point-of-care detection of SARS-CoV-2 using reverse transcription loop-mediated isothermal amplification (RT-LAMP). Virol J, 2020. 17(1): p. 160. Appendix Supplementary materials Image, application 1 Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jgar.2022.11.011.	36462736	PMC9711904	NO-CC CODE	2022-12-02 23:21:31	no	J Glob Antimicrob Resist. 2022 Dec 1; doi: 10.1016/j.jgar.2022.11.011	utf-8	J Glob Antimicrob Resist	2,022	10.1016/j.jgar.2022.11.011	oa_other
==== Front CJC Pediatric and Congenital Heart Disease 2772-8129 2772-8129 The Author(s). Published by Elsevier Inc. on behalf of the Canadian Cardiovascular Society. S2772-8129(22)00118-X 10.1016/j.cjcpc.2022.11.005 Review The impact of COVID-19 on the cardiovascular health of emerging adults aged 18-25: findings from a scoping review Rezler Zachary V. BHSc ab∗ Ko Emma BMSc b Jin Elaine BHSc b Ishtiaq Misha BHSc ab Papaioannou Christina b Kim Helena BHSc b Hwang Kyobin a Lin Yu-Hsin (Sophy) BSc c Colautti Jake BHSc ab Davison Karen M. MSc, PhD c Thakkar Vidhi MSc, PhD ac a Bachelor of Health Sciences (Honours) Program, Faculty of Health Sciences, McMaster University, 1280 Main Street West, Hamilton, ON L8S 4L8, Canada b Michael G. DeGroote School of Medicine, Faculty of Health Sciences, McMaster University, 1280 Main Street West, Hamilton, ON L8S 4K1, Canada c Health Science Program, Faculty of Science and Horticulture, Kwantlen Polytechnic University, 12666 72 Avenue, Surrey, BC V3W 2M8, Canada ∗ Correspondence: Zachary Rezler; 140 Main Street West, Hamilton, ON L8P 0B8, Canada; (519) 564-9048; 1 12 2022 1 12 2022 17 10 2022 15 11 2022 28 11 2022 © 2022 The Author(s). 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. There is limited knowledge regarding the cardiovascular impact of coronavirus disease 2019 (COVID-19) on emerging adults aged 18-25, a group which disproportionately contracts COVID-19. To guide future cardiovascular disease (CVD) research, policy, and practice, a scoping review was conducted to: i) examine the impact of the COVID-19 pandemic on the cardiovascular health of emerging adults; and ii) identify strategies to screen for and manage COVID-19-related cardiovascular complications in this age group. A comprehensive search strategy was applied to several academic databases and grey literature sources. An updated search yielded 6738 articles, 147 of which were extracted and synthesized. Reports identified COVID-19-associated cardiac abnormalities, vascular alterations, and multisystem inflammatory syndrome in emerging adults; based on data from student-athlete samples, prevalence estimates of myocarditis and cardiac abnormalities were 0.5-3% and 0-7%, respectively. Obesity, hypertension, CVD, congenital heart disease, and marginalization are potential risk factors for severe COVID-19, related cardiovascular complications, and mortality in this age group. As a screening modality for COVID-19-associated cardiac involvement, it is recommended that cardiac magnetic resonance imaging be indicated by a positive cardiac history and/or abnormal ‘triad’ testing (cardiac troponin, electrocardiogram, and transthoracic echocardiogram) to improve diagnostic utility. To foster long-term cardiovascular health among emerging adults, cardiorespiratory fitness, health literacy and education, and telehealth accessibility should be priorities of health policy and clinical practice. Ultimately, surveillance data from the broader emerging adult population will be crucial to assess the long-term cardiovascular impact of both COVID-19 and vaccination, guide screening and management protocols, and inform CVD prevention efforts. ==== Body pmcIntroduction Coronavirus disease 2019 (COVID-19) illness severity varies among those infected, and though the majority of individuals present with a mild cough and flu-like symptoms, others may experience pneumonia, virus-related cardiac injury, and death.1 Individuals with cardiovascular comorbidities are at a greater risk of experiencing more severe COVID-19, likely due to the virus exacerbating pre-existing complications through cardiac involvement.2 , 3 However, even in otherwise healthy individuals, infection can lead to alterations in heart structure (e.g. fibrosis) and function. Karbalai Saleh et al.4 found that nearly 30% of their COVID-19 hospitalized patients (mean age=59 years) experienced cardiac injury, which was associated with a nearly two-fold increase in risk of short-term mortality. Moreover, there are data to suggest that individuals may be delaying or avoiding medical treatment for cardiovascular emergencies in response to the pandemic, likely resulting in worse cardiovascular outcomes and mortality. For example, cardiac centres in Canada reported a 30% reduction in emergency visits for ST-elevation myocardial infarctions earlier in 2020.5 The impact of COVID-19 on the cardiovascular health of emerging adults The majority of research on COVID-19 has focused on cardiovascular complications in vulnerable populations, such as older adults and those with pre-existing chronic conditions.4 , 6 Research specific to emerging adults (i.e. those aged 18-257 , 8) is lacking; however, there is preliminary evidence demonstrating myocarditis and cardiac abnormalities in as many as 58% of college athletes following COVID-19.9, 10, 11 As of February 2022, Canadians between the ages of 20-29 (accounting for 13% of the total population12) represented nearly one-fifth (n=610151) of COVID-19 cases and accounted for 4.8% (n=6346), 3.1% (n=699), and 0.3% (n=106) of hospitalizations, ICU admissions, and deaths, respectively.13 Cardiovascular complications, including cardiogenic shock and arrhythmias, have been observed in COVID-19 patients 18 years old and younger.14 Multisystem inflammatory syndrome in children/adults (MIS-C/A) and life-threatening cardiovascular presentations (e.g. myocarditis) have also been reported among young adults with or following COVID-19, the long-term consequences of which remain unknown.15 Furthermore, in light of global vaccination efforts, new evidence has emerged concerning the cardiovascular safety of COVID-19 vaccines (e.g. vaccine-related myocarditis) in younger age groups.16 , 17 Since emerging adults disproportionately contract COVID-1918 , 19, determining the short- and long-term impacts of COVID-19 on cardiovascular function and CVD risk will be of value to clinicians and policymakers and better inform COVID-19 vaccination policy. Objectives The current knowledge gaps surrounding the impact of COVID-19 on the cardiovascular health of emerging adults relate to: i) the prevalence of COVID-19-related cardiovascular presentations; ii) the appropriate screening and management of such conditions; and iii) cardiovascular care. Therefore, to address the needs of various knowledge users (e.g. policymakers, program planners, health care providers), a scoping review was conducted to: i) describe the impact of the COVID-19 pandemic on the cardiovascular health of emerging adults; and ii) identify strategies to screen for and manage cardiovascular complications in emerging adults. Methods The Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR)20 guided this scoping review. Scoping reviews are conducted to examine the extent, range, and nature of the evidence surrounding a given topic and often precede systematic reviews when evidence in an area is new or limited. These reviews map findings from a broader evidence set and identify current gaps in the literature to support future research.20 In comparison, systematic reviews additionally appraise the literature (e.g. assess risk of bias) and are more appropriate when there are clearly defined research questions. Search strategy In consultation with a research librarian from McMaster University, the search strategy was developed and applied to the following bibliographic databases: MEDLINE (via Ovid), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Embase, Web of Science, PsycInfo (via ProQuest), and Sociological Abstracts. All identified keywords and index terms related to COVID-19, emerging adults, and cardiovascular conditions were included in the search and adapted for each database (Supplemental Table S1). Bibliographic databases were initially searched on January 22, 2021 and updated on January 16, 2022; no date limits were applied. Additional searches were conducted up until January 16, 2022 to locate grey literature from national and international sources, including the Cardiac Health Foundation of Canada, Heart and Stroke Foundation of Canada, Hypertension Canada, McMaster Health Forum, and the World Health Organization. Inclusion criteria English-language articles were required to include the following: i) emerging adults (individuals aged ∼18-25 +/-2) as the main participant group or subgroup, or post-secondary student samples with a mean age between 18-25; ii) a COVID-19 context; and iii) at least one of the cardiovascular conditions of interest (Box 1 ). The age range defining emerging adults varies across the literature, often focusing on 18-25-year-olds but occasionally spanning 18-29.7 , 8 The former definition (i.e. ages 18-25) was selected by the research team; however, when applicable, 18-29-year-old cohorts were included in the review. All study designs were considered for inclusion.Box 1 Cardiovascular conditions included in the screening criteria. ● Hypertension or high blood pressure ● Arrhythmia (includes tachycardia, bradycardia, atrial fibrillation) ● Myocardial infarction ● Heart failure ● Cardiac arrest ● Coronary heart disease or ischemic heart disease ● Stroke or cerebrovascular accident ● Transient ischemic attack ● Valvular heart disease ● Cardiomyopathy (includes myocarditis) ● Cardiovascular abnormalities Exclusion criteria Studies were excluded if they: i) focused on emerging adults who were pregnant or studying/working in health care; ii) did not reference COVID-19 or the cardiovascular conditions outlined in the inclusion criteria; or iii) were not available in full text, with the exception of case reports and case series. Articles related to COVID-19 vaccines were excluded in the updated search as they were not part of the initial scope of this project; however, these articles were collected during the screening phase to provide additional context and evidence for discussion. Selection of studies Citations were collated and imported into Covidence21, and duplicates were automatically removed. Following a pilot test of the screening protocol, each title and abstract was screened independently by two reviewers (MI, HK, KH, SL, JC) and assessed against the established inclusion/exclusion criteria. Sources that appeared to satisfy the inclusion criteria were then retrieved as full-text articles and examined by two independent reviewers (MI, HK, KH, SL, JC). Disagreements that occurred between reviewers were resolved by the primary author (ZVR). Data extraction, quality assessment, and analysis A standardized data extraction form was completed for each included study to gather the following information: publication details (publication year, full citation), participant characteristics (country, population, age), study characteristics (aim, design, methods, inclusion/exclusion criteria, sample size, measures, interventions), cardiovascular health outcomes, pertinent findings (e.g. risk and protective factors, prevalence and incidence estimates), conclusions, and implications. Data extracted from included articles were organized into Supplemental Table S2 and S3. In accordance with the PRISMA-ScR20, articles were not appraised for risk of bias or methodological quality. Thematic content analysis, based on the scoping review objectives, was conducted by the primary author (ZVR) with assistance from researcher assistants (EK, EJ, MI, CP) and guidance from two senior researchers (KMD, VT). Results Search results The academic database searches yielded 6738 articles following the automatic removal of duplicates, of which 6233 were deemed irrelevant during abstract and title screening. 505 articles then underwent full-text screening, of which 130 were excluded for reasons such as: being COVID-19 vaccine-related (n=41); including participants outside of the established age criteria (n=24); not being available in full text (n=19); not being directly related to COVID-19 (n=14); or failing to report the mean age of participants (n=11) (see Figure 1 ). 248 articles captured cardiovascular health outcomes, though 33 were excluded during the data extraction phase as they were either duplicates or did not meet the inclusion criteria. Of the 215 articles, 147 featured cardiovascular conditions among emerging adult populations with a COVID-19 context: 36 and 111 articles from the initial and updated search, respectively. The grey literature search located no additional sources. Altogether, the evidence in this review includes 117 case reports/series57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 28 observational studies10 , 11 , 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, and two reviews/editorials165 , 166. Nearly half (49%; n=72) are from the United States (US), followed by Iran (6.1%; n=9), the United Kingdom (UK) (5.4%; n=8), and France (5.4%; n=8).Figure 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart of the search results. Objective #1: Impact of the COVID-19 pandemic on the cardiovascular health of emerging adults Of the 147 articles included, 117 case reports/series identified 123 cases of emerging adults aged 18-25 (mean age=21; 66% male) with active or previous COVID-19 and cardiovascular presentations/complications, 47 of which were classified by authors as hyperinflammatory syndromes, namely MIS. Tachycardia (n=65) was the most common cardiovascular presentation (accompanied by MIS or another cardiovascular complication in 89% of cases), followed by ventricular dysfunction (n=44), hypotension (n=37), thrombosis (n=23), including pulmonary embolism and stroke, cardiomyopathy (n=28), including myocarditis and myopericarditis, heart failure (n=27), hypertension (n=8), bradycardia (n=9), myocardial infarction (n=5), and cardiac arrest (n=4); a detailed report of these cases can be found in Supplemental Table S2. The majority of these cases occurred in emerging adults with no reported history of CVD (91%) or other comorbidities (60%), suggesting that COVID-19 can lead to severe and sometimes fatal cardiovascular health outcomes (8.9%; n=11) in otherwise healthy young adults. The following describes the types of cardiovascular manifestations in emerging adults who contracted COVID-19 as well as related impacts on cardiovascular care. Reports of myocarditis and cardiac abnormalities In the initial search, two observational studies10 , 11 focusing on COVID-19-associated myocarditis and cardiac abnormalities (i.e. findings not meeting the 2018 Lake Louise Criteria167) among emerging adults were identified. Starekova et al.11 examined electronic health records of 145 university student-athletes aged 17-23 recovering from COVID-19 and found that 1.4% (n=2) had cardiac magnetic resonance imaging (cMRI) findings consistent with myocarditis. In a cross-sectional study by Brito et al.10, sequential cMRI was performed on 48 college student-athletes (mean age=19) who experienced mild or asymptomatic COVID-19; cardiac abnormalities were observed in 56% (n=27) of participants, including pericardial (n=13), myocardial (n=8), and myopericardial involvement (n=6). These results initially suggested that myocarditis and general cardiac effects may be relatively common among healthy emerging adults who contract COVID-19. Since then, 10 additional articles have been identified139, 140, 141, 142, 143, 144, 145 , 147 , 165 , 166, providing further insight into the prevalence and associations of myocarditis and cardiac abnormalities following COVID-19 in emerging adults. The majority of these data come from observational studies on student-athlete populations.139, 140, 141, 142, 143, 144, 145 , 165 , 166 In the broader emerging adult population, hospital-based administrative data from 2019 through 2021 from over 900 sites in the US identified 121 inpatients aged 16-24 diagnosed with COVID-19 and myocarditis, yielding an adjusted risk 7.4 times greater than that of their uninfected counterparts (0.098% versus 0.013%, respectively).147 A recent systematic review by van Hattum et al.166 provided prevalence estimates of COVID-19-associated myocarditis and other cardiac abnormalities (e.g. arrhythmias) among student-athletes post-infection. Among 2326 college athletes (median age=22), the weighted prevalence of myocarditis found with cMRI was 2.1% using the established Lake Louise Criteria. The majority (59%) of these individuals were mildly symptomatic while infected; 22% were asymptomatic, and the remainder experienced moderate (19%) or severe (0.2%) illness. There were no observed arrhythmias and only one resuscitated cardiac arrest unlikely attributable to COVID-19. Several studies in our evidence set specific to COVID-19-related myocarditis were included in the review166; prevalence estimates ranged from 0-15%10 , 11 , 139, 140, 141, 142, 143, 144, 145, 146 , 165 , 166, which included both asymptomatic and symptomatic cases of COVID-19. Additional evidence has accumulated regarding the prevalence of general cardiac involvement (e.g. myocardial edema, pericardial effusion) in emerging adults with COVID-19. In a large cohort of 3018 college athletes (mean age=21), 0.7% (95% confidence interval [CI]: 0.4, 1.1) were determined to have definite, probable, or possible SARS-CoV-2-related myocardial or pericardial involvement with cMRI or ‘triad’ testing: cardiac troponin levels, an electrocardiogram (ECG), and a transthoracic echocardiogram (TTE).140 An analysis by Petek et al.141 utilized data from 44 US colleges, known as the Outcomes Registry for Cardiac Conditions in Athletes (ORCCA), with a sample of 3597 student-athletes (mean age=20) diagnosed with COVID-19. Investigation of individuals with exertional cardiopulmonary symptoms (n=137) revealed 10 cases of cardiovascular sequelae (7.3%) and five cases (3.6%) of definite or probable cardiac involvement via cMRI, representing 21% of athletes with chest pain post-COVID-19. In another group of 170 athletes aged 18-25, 3.5% (n=6) had abnormal cardiac rhythms and 1.2% (n=2) were diagnosed with viral pericarditis using cMRI.139 In a sample of 137 student-athletes aged 18-27, 82% (n=112) of whom were symptomatic, algorithm-guided screening following COVID-19 identified trace pericardial effusions with echocardiography in 2.9% (n=4) of participants.143 Similarly, although Malek et al.144 identified no cases of myocarditis via cMRI in a cohort of 26 athletes recovering mainly from asymptomatic or mild infections (median age=24), 19% (n=5) demonstrated cardiac abnormalities, including signs of myocardial edema and pericardial effusion. Compared to myocarditis, most studies suggest that general cardiac involvement among emerging adults post-COVID-19 is more common (0-58%).10 , 11 , 139, 140, 141, 142, 143, 144, 145, 146 , 166 However, it is of note that data from larger cohorts (i.e. N>100) provided smaller prevalence estimates in athletes imaged with cMRI (0-7%).11 , 139, 140, 141 , 143 Risk factors Cohort analyses using cMRI or triad testing identified several risk factors for cardiac involvement, including White Hispanic race (odds ratio [OR]: 7.6; 95% CI: 2.2, 26.1), cardiopulmonary symptoms before or during infection or return to exercise (adjusted odds ratio [aOR]: 3.1; 95% CI: 1.2-7.8), or one or more abnormal triad test results potentially associated with COVID-19 (aOR: 37.4; 95% CI: 13.3-105.3).140 , 141 The association between biological sex and COVID-19-associated cardiac abnormalities (e.g. abnormal ECG findings, diagnosis of myocarditis) was unclear.139 , 147 With respect to age, the risk of COVID-19-associated myocarditis in hospitalized patients was lowest among individuals aged 16-24 when compared with older age cohorts.147 Ultimately, our review of the evidence suggests the prevalence of COVID-19-associated myocarditis among otherwise healthy emerging adult student-athletes to be nearing the lower end of 0.5-3%11 , 139, 140, 141 , 143 , 165 , 166, though general cardiac involvement (e.g. myocardial edema, pericardial effusions) is likely more prevalent (0-7%)11 , 139, 140, 141 , 143, particularly in those with lingering cardiopulmonary symptoms.141 Other cardiac events, such as arrhythmias and cardiac arrests, seem to be less common in this age cohort following infection.166 There was a relatively low risk of clinical cardiac events (i.e. significant arrhythmias, heart failure, sudden cardiac arrest or death) in short-term follow-up (∼0.03%).140 Reports of hyperinflammatory syndromes Belay et al.148 conducted the largest US cohort study to date (n=1733) describing the clinical characteristics and geographical and temporal distribution of patients under 21 years of age with MIS-C. Of the 55 emerging adults aged 18-20, 58% (n=32) were admitted to the ICU, and 11% (n=6) died. Cardiovascular presentations in this age group included hypotension in 53%, cardiac dysfunction in 42%, myocarditis in 31%, pericardial effusion in 27%, and coronary artery dilation or aneurysm in 15%. Compared to the pediatric population, patients aged 18 to 20 had the highest proportion of myocarditis (31% in 18-20-year-olds versus 9.2-28% in 0-17-year-olds; p<0.001). The 18-20-year-old cohort, however, had the lowest incidence of MIS at 0.4 per 100,000 infected (p<0.001). In this review, 47 cases (mean age=21; 70% male) of emerging adults with cardiovascular complications in the context of confirmed or suspected hyperinflammatory syndromes and active or previous COVID-19 were identified.22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56 If not concomitant COVID-19, previous infections typically occurred 3-8 weeks prior to onset of symptoms.23 , 30 , 32, 33, 34 , 40 , 43, 44, 45, 46 , 52 , 53 , 56 Alongside MIS-C/A, cardiovascular presentations/complications included ventricular dysfunction (n=33), tachycardia (n=28), hypotension (n=16), cardiogenic shock (n=14), cardiomyopathy (n=11), including myocarditis and pericarditis, valve insufficiency (n=4), atrial fibrillation (n=2), and non-ST-elevation myocardial infarction (n=2); refer to Supplemental Table S2 for additional case details. Cardiac involvement with echocardiographic changes (e.g. reduced EF, global hypokinesia) and elevated NT-proBNP and troponin were frequently reported among these cases. Of note, 79% (n=37) of cases were previously healthy with no past medical history, and 98% (n=46) had no reported history of CVD. Due to prompt identification and treatment, there were few reports (4.3%; n=2) of emerging adults with MIS who died.48 , 55 A case series described the post-mortem examinations of four deaths due to maternal or pediatric ante-mortem COVID-19. In this report, the death of a 19-year-old female was attributed to MIS causing coagulopathy. 55 A separate retrospective cohort study by Whitworth et al.149 identified the incidence of thrombosis in children and adolescents under 21 years of age hospitalized with COVID-19 or MIS-C (n=564). 11 of the 20 cases of thrombosis (55%) occurred in patients aged 16-21, 36% (n=4) of which died; all cases were either African American or Hispanic individuals. Patients aged 12 and older with MIS-C displayed the highest rate of thrombotic events (19%). Despite a lower incidence than younger age cohorts, emerging adults seem to be at an elevated risk of MIS-related cardiovascular complications, including myocarditis, and death. Evidence outlining risk factors and long-term prognosis after recovery in this age group is lacking. Reports of vascular alterations Multiple studies reported signs of vascular dysfunction among young adults following COVID-19.122 , 160, 161, 162, 163, 164 Ratchford et al.160 performed a cross-sectional analysis on healthy young adults (mean age=20) to examine the effects of COVID-19 on markers of vascular function and arterial stiffness. When compared with uninfected individuals, those with COVID-19 experienced significantly lower brachial artery flow-mediated dilation (FMD) (2.7 ± 1.2% versus 8.8 ± 3.0%; p<0.01) and femoral artery blood flow response (-3 ± 91 mL versus 118 ± 114 mL; p<0.01) 3-4 weeks after infection. This same group performed another cross-sectional study161 and found higher carotid artery stiffness among young adults with COVID-19 (6 ± 1 m/s) compared with healthy controls (5 ± 1 m/s; p=0.02). Aortic augmentation index was also greater in the COVID-19 group (12.7 ± 9.1% versus 3.3 ± 12.6%; p=0.03), suggesting aortic stiffening and potential atherosclerotic risk progression. Another study by Stute et al.163 found that resting muscle sympathetic nerve activity, a measure of arterial stiffness, was higher in individuals recovering from COVID-19 (n=16; mean age=20) compared with uninfected controls (285 ± 101 a.u./min versus 159 ± 46 a.u./min, respectively; p=0.001). Growing evidence indicates that endothelial inflammation associated with COVID-19 may contribute to cerebrovascular disease. A case series by Arandela et al.122 identified two patients between the ages of 18-25 who developed reversible cerebral vasoconstriction syndrome in the context of COVID-19, suggesting a potential risk in this age cohort among those using vasoactive agents (e.g. marijuana). Alterations in cerebral and peripheral vasculature were further investigated in a cohort study conducted by Nandadeva et al.162 Analysis found that only peripheral vascular function was impaired in young adults with lingering COVID-19 symptoms (n=8; mean age=24); this impairment was not seen in those who were no longer symptomatic (n=8; mean age=22). Taken together, these findings suggest that the effects of COVID-19 on central large arteries may be a transient phenomenon. Whereas the previously mentioned studies reported vascular alterations at rest, Stute et al.164 aimed to elucidate the effect of COVID-19 on central and peripheral hemodynamics during a rhythmic handgrip exercise. Brachial artery blood flow was significantly lower in the COVID-19 group (n=13; mean age=21) compared with controls (n=13; mean age=27): 386.3 ± 132.5 mL/min versus 507.4 ± 109.9 mL/min, respectively (p=0.002). The COVID-19 group also displayed greater increases in systolic blood pressure, systolic arterial pressure, and rate pressure product upon exertion. Reports of individuals with cardiovascular-related vulnerabilities that increase risk for severe COVID-19, related cardiovascular complications, and mortality Cardiovascular disease, hypertension, and obesity A few studies offered insight into cardiovascular-related risk factors that may confer an increased risk of severe COVID-19. Fathi et al.157 developed a model to predict two-week mortality using data from 57705 inpatients with COVID-19, which included a ‘young’ cohort (n=1049; aged 15-24). Among those who died in this subsample (n=50; 4.8%), hypertension was significantly associated with two-week mortality (OR: 54.3; 95% CI: 19.9, 168.2). A retrospective study of young adult COVID-19 patients aged 18-35 admitted to New York City public hospitals also found that cardiac comorbidities and hypertension were associated with increased mortality; however, those with these comorbidities in the 18-23-year-old cohort all recovered.155 To provide context, data from the National Health Interview Survey and an undergraduate student sample during the pandemic show that pre-existing heart conditions affect 0.5-1.9% of this cohort.153 , 156 Furthermore, evidence from after a COVID-19 lockdown found that the prevalence of hypertension among undergraduate students (n=325; mean age=22) remained at 1%.154 A hospital system-based retrospective chart review159 conducted in Texas identified risk factors for severe disease and readmission among young adults aged 18-29 diagnosed with COVID-19 (mean age=24). The study identified 1853 patients with COVID-19, 8% (n=148) of whom experienced a composite disease outcome (e.g. a severe respiratory or cardiovascular event) within 30 days of their first encounter. In this cohort of young adults, older age, obesity, previous CVD (i.e. myocardial infarction, congestive heart failure, cerebrovascular disease) and diabetes were among significant risk factors for composite disease outcomes (p≤0.03). A history of CVD and obesity were also predictors of severe disease and/or readmission within 30 days (p<0.05). In addition, the authors highlighted the relationship between race and ethnicity (e.g. Hispanic ethnicity) and poorer health outcomes. A preliminary analysis (abstract) by Sands-Lincoln et al.168 of patients aged 18-24 with COVID-19 (n=6648; mean age=22) found African American (OR: 2.4; 95% CI: 1.6, 3.5) and ‘other race’ identity (OR: 5.0; 95% CI: 2.6, 9.1), CVD (OR: 4.0; 95% CI: 2.8, 5.7), and obesity (OR: 3.0; 95% CI: 2.1, 4.3) to be associated with increased odds of hospitalization. Another study by Richardson et al.158 analyzed data from hospitalized patients aged 18-39 at acute care hospitals in New York City. Notably, among patients aged 18-24 (n=119), those who died (n=4) or required invasive mechanical ventilation (n=7) were obese, and five of these patients had additional comorbidities, including Down syndrome and congestive heart failure. In summary, the majority of cardiovascular-related medical vulnerabilities in emerging adults are rare153 , 154 , 156; however, when present, data suggest that the risk for severe COVID-19, related cardiovascular complications, and mortality in this age group is not insignificant, particularly among those with comorbidities. Congenital heart disease and genetic syndromes Predicting the COVID-19 response in emerging adults with existing congenital heart disease (CHD) is challenging given the heterogeneity of the population. Two separate retrospective reviews investigating factors associated with severe COVID-19 and mortality across the CHD population found that the presence of a structural congenital heart defect did not confer an increased morbidity or mortality risk.150 , 151 Lewis et al.150 detailed the experience of four emerging adults aged 21-25 with CHD who were described to have moderate-to-severe COVID-19. These individuals did not appear to be disproportionately impacted unless they were at an advanced physiological stage (i.e. class C or D) (OR: 19.4) and/or had genetic syndromes (OR: 35.8), such as Down syndrome and DiGeorge syndrome (p≤0.002). Similarly, Broberg et al.151 found that a worse physiological stage of CHD (e.g. Eisenmenger physiology, cyanosis) was associated with mortality (p=0.001), whereas anatomic complexity or defect group were not. These findings suggest that susceptibility to severe COVID-19 among emerging adults with CHD is based primarily on physiological factors, and, when accompanied by certain genetic conditions, such as Down syndrome and DiGeorge syndrome, CHD is associated with increased COVID-19 hospitalization.71 , 150, 151, 152 , 158 The role of genetic disorders in COVID-19 severity among emerging adults with CVD seems to vary depending on the underlying disorder. Adults with Duchenne muscular dystrophy (DMD) are at risk for cardiorespiratory compromise (i.e. cardiomyopathy) and so were thought to be vulnerable to worse COVID-19 outcomes.138 However, Quinlivan et al.138 reported on five emerging adult males aged 18-23 with DMD who contracted COVID-19 and did not develop moderate or severe COVID-19. Despite a history of moderate-to-severe cardiomyopathy, long-term immunosuppressive treatment, and respiratory insufficiency, all patients recovered fully with no complications. This evidence indicates that emerging adults with DMD may not be at an elevated risk of severe COVID-19 and related cardiovascular complications. Impacts on cardiovascular care Cardiovascular care for emerging adults was impacted by the COVID-19 pandemic, which in some cases affected cardiovascular outcomes. For example, Warraich et al.130 described a 19-year-old patient who delayed treatment by self-isolating for two weeks during the pandemic due to a persistent cough that was later understood to be a symptom of a posterior circulation ischemic (POCI) stroke. Other reports of emerging adults with cardiovascular presentations, including stroke130 , 132, intracardiac thromboses135, pulmonary hypertension131, and sinus tachycardia with atrioventricular block133, found that many avoided seeking treatment due to lockdown measures or fear of contracting COVID-19. In addition, the pandemic has resulted in misguided clinical judgement and management. In the case presented by Warraich et al.130, the young patient with POCI stroke lacked risk factors for stroke, and so an initial diagnosis of COVID-19 was made. In another case, a 19-year-old male presenting to the emergency department with constitutional symptoms was repeatedly tested for COVID-19, treated with antibiotics for COVID-19 pneumonia, discharged, and later diagnosed with Coxsackie A myocarditis upon readmission.134 Similarly, Balfe et al.137 presented the case of an 18-year-old female suffering from rheumatic mitral stenosis whose condition was worsened by IV fluids she was administered under the presumption that she had COVID-19. Reports were also identified in which life-saving cardiovascular care for emerging adults (e.g. extracorporeal cardiopulmonary resuscitation) was nearly prevented or complicated due to the patient’s COVID-19 status.60 , 63 , 136 Overall, these cases highlight evidence of crucial cardiovascular care being delayed, misguided, or complicated as a result of the COVID-19 pandemic influencing public and clinical decision making. Objective #2: Strategies to screen for and manage cardiovascular complications in emerging adults Myocarditis and cardiac abnormalities The majority of evidence surrounding screening protocols for COVID-associated myocarditis and cardiac abnormalities in emerging adults focuses on student-athletes.10 , 11 , 139, 140, 141, 142, 143, 144, 145, 146 , 165 , 166 Initial evidence explored the utility of extensive cardiorespiratory and hematological screening in athletes recovering from COVID-19 to identify post-infection cardiovascular abnormalities. Gervasi et al.146 examined a cohort of 30 professional soccer players aged 19-27, 18 of whom tested positive for SARS-CoV-2 IgG antibodies and reported previous asymptomatic or mild COVID-19. Following comprehensive screening (e.g. blood tests, spirometry, ECG), none of the participants demonstrated clinically relevant cardiovascular abnormalities (e.g. myocarditis). Furthermore, both Starekova et al.11 and Brito et al.10 conducted studies in which university athletes recovering from COVID-19 underwent screening with cMRI, most of which were recovering from mild-to-moderate or asymptomatic illness. Starekova et al.11, with a sample of athletes aged 17-23, found that only 1.4% (n=4) of those screened met the Lake Louise Criteria for myocarditis. Brito et al.10 identified myocardial, pericardial, or myopericardial abnormalities in 58% of athletes, but no signs of ongoing myocarditis. Due to the low prevalence of clinical myocarditis in these cohorts, the overall utility of cMRI as a standard screening tool for myocarditis was deemed unwarranted, especially for individuals with asymptomatic or mild COVID-19 and those with a normal ECG and cardiac troponin. Evidence from the updated search continues to support cMRI as a sensitive and specific screening modality for myocarditis and other cardiac abnormalities in this population.140, 141, 142 , 145 , 165 , 166 Still, larger cohort studies recommend cMRI only for individuals with a heightened risk based on an initial, comprehensive cardiac evaluation.139, 140, 141, 142, 143, 144 Multiple studies screened for elevated cardiac troponin139, 140, 141, 142, 143, 144 , 165 or abnormal ECG or echocardiogram findings139, 140, 141 , 143 , 144 , 165 to indicate cMRI for further diagnostic workup. A sample of 1597 athletes screened for post-COVID-19 cardiac abnormalities demonstrated the utility of four respective screening strategies165 , 169:I. A positive cardiac history (e.g. chest pain) to indicate triad testing; abnormal triad testing to indicate cMRI II. Abnormal triad testing to indicate cMRI III. A positive cardiac history (e.g. chest pain) or abnormal triad testing to indicate cMRI IV. cMRI without prior screening These approaches would have identified 5 (0.3%), 13 (0.8%), 17 (1.1%), and 37 (2.3%) cases of myocarditis in the cohort, respectively. Moulson et al.140 similarly found that 82% of athletes diagnosed with definite or probable cardiac involvement following COVID-19 would have been identified with a stepwise approach using moderate symptom severity, cardiopulmonary symptoms, or abnormal triad testing to indicate cMRI. Despite such evidence showcasing an increased specificity when screening algorithms included a cardiac history and triad testing140 , 165 , 169, the previously discussed systematic review by van Hattum et al.166 found that, among ten studies (n=4171), there was no clear association between post-recovery cardiac troponin levels and cardiac abnormalities. In contrast, a comprehensive history appears to have clinical utility, as the Moulson et al.140 and Petek et al.141 analyses of data from more than 40 US academic institutions found cardiopulmonary symptoms, specifically chest pain and dyspnea, to increase the risk of COVID-19-related cardiac involvement (aOR: 3.1; 95% CI: 1.2-7.8). Of those who had chest pain following return to exercise, 21% had evidence of myocardial and/or pericardial involvement on cMRI.141 Ultimately, the impact of these screening algorithms on the clinical course of emerging adult athletes recovering from COVID-19 remains unknown, as few, if any, adverse cardiac events have been reported among this population following return to play.140 , 165 , 166 MIS-C/A Regarding the identification of MIS-C/A in emerging adults with concomitant or previous COVID-19, Belay et al.148 found that only 71% of the 18-20-year-old MIS cases demonstrated SARS-CoV-2 polymerase chain reaction (PCR) positivity, and 58% had positive serology. In our review of MIS case reports/series involving cardiovascular complications, 47% (n=17) reported negative PCR results, whereas 70% (n=33) had positive serology. Given the significant proportion of patients with MIS and negative PCR results, authors have recommended SARS-CoV-2 antibody testing to classify MIS patients without active COVID-19 or those with atypical COVID-19 presentations.24 Compared to younger age groups, 18-20-year-olds were more likely to report COVID-19-like illness seven or more days prior to MIS onset (63% versus 18-44%; p<0.001).148 In the cohort study by Belay et al.148, mainstay treatments for MIS included intravenous immunoglobulin (IVIG) and steroids. Across the 47 cases included in this review22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 70% (n=33) were treated with steroids (e.g. methylprednisolone, prednisolone, prednisone) and 62% (n=29) with IVIG. Additional therapies included interleukin-1 receptor antagonists, interleukin-6 receptor antagonists, and supportive care measures, such as aspirin and anticoagulation (Supplemental Table S2). Discussion This scoping review examined current literature to describe the impact of the COVID-19 pandemic on the cardiovascular health and care of emerging adults. We identified multiple cases of cardiovascular presentations (e.g. tachycardia, ventricular dysfunction), cardiac abnormalities (e.g. myocardial/pericardial involvement), MIS-C/A, and vascular alterations among emerging adults who contracted COVID-19, the long-term effects of which remain unknown. The evidence is insufficient to determine the true incidence and prevalence of these complications in this age cohort; however, prevalence estimates from student-athlete samples of COVID-19-associated myocarditis and cardiac involvement on cMRI ranged from 0.5-3%11 , 139, 140, 141 , 143 , 165 , 166 and 0-7%11 , 139, 140, 141 , 143, respectively. In some groups, medical vulnerabilities such as obesity158 , 159 , 168, hypertension155 , 157, previous CVD159 , 168, and certain genetic syndromes71 , 150 , 152 , 158 posed an increased risk of severe illness, cardiovascular complications, and/or mortality from COVID-19. There were also reports of emerging adults with cardiovascular presentations where the provision of cardiovascular care was negatively impacted due to COVID-19.60 , 130, 131, 132, 133, 134, 135 , 137 Regarding the appropriate screening and management of these cardiovascular abnormalities, the majority of age-specific evidence pertains to COVID-19-associated myocarditis and MIS. Based on the findings of this scoping review, a framework which highlights the negative impacts of the COVID-19 pandemic on the cardiovascular health of emerging adults and corresponding health promotion opportunities to prevent and mitigate these effects was constructed (see Figure 2 ).Figure 2 A framework to promote cardiovascular health among emerging adults in the context of the COVID-19 pandemic (created in part with BioRender.com). Dashed lines represent pathways to prevent or manage COVID-19-related cardiovascular health concerns. *Health literacy and education initiatives may benefit from including parents/guardians as part of the target audience as they often play a role in an emerging adult’s decision to seek medical care, Abbreviations: CV, cardiovascular; CVD, cardiovascular disease; COVID-19, coronavirus disease 2019; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2 In the following, accumulated evidence from the review is highlighted with corresponding suggestions for future research, policy, and practice. Investigations surrounding COVID-19 and cardiovascular health In this review, cardiovascular presentations among emerging adults with current or previous COVID-19 ranged from milder manifestations such as sinus tachycardia170, to more serious complications, including myocarditis, stroke, cardiogenic shock, heart failure, thrombosis, and MIS. Investigators suggest that these cardiovascular complications may lead to cardiomyopathy, cardiac arrhythmias, and sudden cardiac arrest in the long term.169 , 171, 172, 173 Therefore, ongoing research involving emerging adults who contract COVID-19 and develop cardiovascular sequelae is recommended. Long-term surveillance of these patients can help assess and inform cardiovascular screening and treatment protocols. An emphasis should also be placed on ensuring study samples are representative of the broader emerging adult population, given most of the cohort evidence found in this review focused on student-athletes.10 , 11 , 139, 140, 141, 142, 143, 144, 145 , 148 , 165 , 166 Myocarditis and cardiac abnormalities Extrapolating prevalence estimates of COVID-19-associated myocarditis and cardiac abnormalities from student-athlete cohorts to the entire emerging adult population is cautioned. Experts believe that athletes who exercise with COVID-19 are at an increased risk of developing myocarditis169 , 174. At the same time, their overall heightened cardiovascular health is not representative of the broader population. Regarding the long-term implications of these findings, updated ORCCA data175 demonstrate complete or partial resolution of cMRI abnormalities in 80% (n=8) of participants. Follow-up after more than one year identified no adverse cardiac outcomes in the subsample of athletes with initial cardiac involvement. Though promising, additional surveillance is required to establish the true long-term risk of COVID-19-associated myocarditis and other cardiovascular sequelae in emerging adults. With respect to screening, data from student-athletes may be used to aid clinical decision-making given limited evidence from the broader emerging adult population. Additional analysis of the ORCCA cohort176 continues to indicate the limited diagnostic utility of cardiac troponin as a screening modality for those with COVID-19-associated myocardial involvement; it is recommended only for those with a high clinical pretest probability of disease. Moreover, the data in this review do not support routine cMRI in lower-risk individuals.10 , 11 , 139, 140, 141, 142, 143, 144 , 146; this approach would be quite costly and increase the likelihood of false positive.165 , 177 Rather, a screening approach guided by cardiac symptoms (e.g. chest pain) and/or triad testing to indicate cMRI will increase its diagnostic yield and feasibility.140 , 169 A 2022 expert consensus paper from the American College of Cardiology178 is in agreement, recommending against cardiac testing for asymptomatic cases and those with only mild-to-moderate non-cardiopulmonary symptoms or with previous COVID-19 in the absence of ongoing cardiopulmonary symptoms. Individuals with COVID-19-related cardiopulmonary symptoms are recommended for triad testing and, if results are abnormal, a cardiology consultation to consider subsequent cMRI and additional cardiac testing for diagnosis. The body of evidence in this review is limited with respect to treating COVID-19-associated myocarditis in this age group. However, case reports75 , 98 , 120 , 179 and supporting literature180 indicate that in the absence of additional cardiovascular complications (e.g. acute heart failure), rest, supportive measures (e.g. intravenous/oral hydration, beta-blockers), and immunosuppressive therapy yield promising outcomes. MIS-C/A Emerging adults with MIS demonstrated significant mortality (Supplemental Table S2) and were more often reported to have COVID-19-like illness prior to presentation and subsequent findings of myocarditis.148 Ultimately, these findings emphasize the importance of recognizing and including MIS as a differential in the context of recent COVID-19 in this cohort. As of now, it remains unclear whether MIS is a manifestation of acute COVID-19 or an entirely post-acute phenomenon. Given that not all MIS patients present with positive PCR test results, SARS-CoV-2 antibody testing is recommended as a potentially crucial diagnostic measure to classify and recognize these patients. 24 However, in the current state of the pandemic, the majority of adults demonstrate seroprevalence of SARS-CoV-2 antibodies181, limiting its clinical utility. The additional use of laboratory tests for inflammation, hypercoagulability, and organ damage (e.g. CRP, D-dimer, cardiac enzymes) may assist in early identification and subsequent management. To our knowledge, no consensus guidelines are available for MIS-A; current recommendations are extrapolated from evidence specific to MIS-C, applicable to ages less than 21 years. The CDC supports serologic testing in addition to PCR testing when feasible, as well as workup for cardiac involvement in affected individuals.182 Both the CDC and National Institutes of Health discuss IVIG in combination with glucocorticoids (e.g. methylprednisolone) as the first-line treatment strategy, citing its benefits for faster recovery of cardiac function.182 , 183 The latest clinical guideline from the American College of Rheumatology additionally recommends higher doses of steroids, anakinra, or infliximab for refractory disease.184 Future research into MIS is required to better elucidate pertinent risk factors and facilitate diagnosis, management, and the long-term cardiovascular implications in this age group. Changes in systemic vasculature Reports of young adults recovering from COVID-19 found transient alterations in arterial, cerebral, and peripheral vasculature in those with lingering symptoms and even after recovery.122 , 160, 161, 162, 163, 164 Participants in these studies were all assessed 3-8 weeks after their first positive COVID-19 test, therefore there is limited long-term data to substantiate these findings. Of note, one study reported a 6% difference in brachial artery FMD among those with previous COVID-19 compared with those who were uninfected.160 A meta-analysis by Inaba et al.185 (mean age>50) found that for every 1% decrease in brachial artery FMD, there is a corresponding 8% increase in risk of future cardiovascular events, including stroke, heart attack, and death. In addition, these findings represent dysautonomia post-COVID-19, a phenomenon observed among those experiencing ‘long COVID’, the symptoms of which also include fatigue and shortness of breath.186, 187, 188 These sequelae may lead to exercise intolerance among those recovering from COVID-19 and pose future implications for CVD risk among emerging adults. Moving forward, follow-up among emerging adults is needed to better understand the relationship between COVID-19 and long-term vascular function. Risk stratification of emerging adults for COVID-19 severity, related cardiovascular complications, and mortality There is increasing evidence that certain cardiovascular comorbidities (e.g. obesity, hypertension, CVD)153, 154, 155, 156, 157, 158, 159 , 168 are relevant risk factors among emerging adults in the context of COVID-19. Among emerging adults with CHD who contract COVID-19, these patients may still be considered high-risk due to the variability in this population’s clinical presentation and response to treatments.150 , 151 In the future, retrospective analyses should focus on emerging adults who have experienced severe COVID-19 and related cardiovascular complications to further clarify the relationship between these outcomes and presumptive risk factors, particularly CVD. Since most sample sizes in the literature were modest, it is recommended that multicentre registries specific to emerging adults, such as the ORCCA141 and COVID-19 CVD Registry189, be established to enable the aggregation of data and provide adequate statistical power to conduct such analyses. The cardiovascular safety of COVID-19 vaccines for emerging adults and public education As vaccination efforts progress around the globe, there have been signals of adverse cardiovascular events from COVID-19 vaccines, specifically cases of myocarditis and pericarditis among emerging adults, which were not observed in clinical trials.16 , 17 , 190 In a sample of approximately 23 million Nordic residents, the number of excess cases of myocarditis among males aged 16-24 associated with a second dose of the BNT162b2 (Pfizer) and mRNA-1273 (Moderna) vaccines 28 days post-administration was estimated at 4-7 and 9-28 per 100,000 doses, respectively.16 A US study utilizing the Vaccine Adverse Event Reporting System (VAERS) data17 similarly found an elevated risk among males aged 18-24, with 52 and 56 cases of myocarditis per million following second doses of the Pfizer and Moderna vaccine, respectively; comparatively, there were only 4-7 cases of myocarditis per million doses in the female cohort. No reported deaths in vaccinated individuals younger than 30 were attributed to myocarditis, apart from one potential case. Of note, VAERS reports can be submitted by any member of the public and are not verified for cause-and-effect relationships, making this passive surveillance system particularly susceptible to false reports and under- or over-reporting. Still, meta-analysis reveals the incidence of myopericarditis associated with COVID-19 vaccines to be either comparable or lower than that of other vaccines.190 Cardiovascular safety concerns of COVID-19 vaccines must be contextualized with the cardiovascular risks associated with COVID-19. A recent CDC analysis found that the risk of cardiovascular complications (i.e. myocarditis, pericarditis, or MIS) among males aged 18-29 was 7-8 times greater following COVID-19 compared with vaccination; a similar observation was found for females.191 A population-based cohort study192 of individuals in Ontario, Canada found that the highest rate was among males aged 18-24 receiving Moderna versus Pfizer as the second dose (300 versus 59 cases per million doses, respectively) and was significantly higher when the interdose interval was ≤30 days compared with ≥56 days (95-377 versus 11-132 per million doses, respectively). In addition, UK data193 found an association between COVID-19 and pericarditis and cardiac arrhythmias among 16-29-year-olds, though this same association was not established with vaccines. For additional context, data suggest that one million second mRNA COVID-19 vaccine doses could prevent 11000 cases, 560 hospitalizations, 138 ICU admissions, and six deaths among 12-29-year-olds, compared with 39-47 potential cases of vaccine-related myocarditis.194 Vaccine-related myocarditis also tends to be milder when compared with COVID-19-related cases, which have been associated with a greater risk of hospitalization and death.193 Concerns regarding cardiovascular complications following COVID-19 in this cohort are partly due to the fact that emerging adults contract COVID-19 at disproportionately higher rates than older adults.156 Though in agreement with the aforementioned recommendation for COVID-19 vaccination in this cohort, we encourage continued surveillance and analysis of data in this area195. This measure is especially important in light of waning vaccine immunity, increasing natural immunity, and the emergence of new SARS-CoV-2 variants that seem to pose a lower risk of hospitalization and severe disease.196 Therefore, a risk-benefit analysis of COVID-19 vaccines should be regularly updated for emerging adults as the pandemic progresses. To mitigate public hesitancy, as well as the cardiovascular risks associated with COVID-19 vaccines and infection in this age group, we recommend researchers, policymakers, and clinicians consider the following:● The development and distribution of strain-specific boosters that improve vaccine efficacy against transmission and symptomatic illness197 , 198; ● Longer COVID-19 vaccine interdose intervals and age-based product considerations 192 to reduce the incidence of cardiovascular side effects; ● Ongoing surveillance of COVID-19 and vaccine-associated cardiovascular sequelae (e.g. myocarditis, pericarditis, MIS)15 , 195; and ● Clinician and public education (for emerging adults and parents/guardians) surrounding the signs and symptoms and expected timelines of both infection- and vaccine-associated cardiovascular sequelae; see Table 1 .Table 1 Signs and symptoms of COVID-19-associated cardiovascular sequelae that affect emerging adults and recommendations for when to seek medical care, as per the Centers for Disease Control and Prevention15,195 Cardiovascular Sequelae Recommendation Myocarditis and Pericarditis Seek medical care if you or your child have any of the following symptoms after COVID-19 infection or vaccination∗: ● Chest pain ● Shortness of breath ● Feelings of having a fast-beating, fluttering, or pounding heart MIS-C/A Seek medical care if you or your child have any of the following signs/symptoms after COVID-19∗∗: ● Ongoing fever PLUS >1 of the following: ○ Stomach pain ○ Bloodshot eyes ○ Diarrhea ○ Dizziness or lightheadedness ○ Skin rash ○ Vomiting Seek emergency medical attention if you or your child have any of the additional signs/symptoms∗∗: ● Trouble breathing ● Persistent pain or pressure in the chest ● New confusion ● Inability to wake or stay awake ● Pale, grey, or blue-coloured skin, lips, or nail beds (depending on skin tone) Abbreviations: COVID-19, coronavirus disease 2019; multisystem inflammatory syndrome in children/adults, MIS-C/A ∗ More often after the second dose; usually within one week post-vaccination (median time=3 days)192 ∗∗ Usually within six weeks post-infection or exposure Preventing misdiagnoses, delayed care, and future COVID-19-related CVD This review suggests that when emerging adults present with COVID-19-like symptoms, the differential diagnosis should include potentially life-threatening cardiovascular manifestations, such as POCI stroke and other forms of viral myocarditis.130 , 134 , 137 In addition, emerging adults who have cardiovascular symptoms, along with their parents/guardians, need to be educated on the critical importance of seeking medical care despite the risk of contracting COVID-19.130, 131, 132, 133, 134, 135 , 137 The increasing availability of telehealth and digital health interventions199 may assist younger patients in receiving timely diagnoses and treatment for potentially life-threatening cardiovascular events. From a developmental perspective, emerging adulthood is a critical time to establish behaviours that promote long-term cardiovascular health. In a national longitudinal study by Clark et al.200, the average 30-year risk of developing CVD for young adults (mean age=29) ranged from 4.4-18%. Health literacy and education initiatives199 , 201 , 202 that emphasize CVD prevention (e.g. healthy diets, physical activity, avoiding substance use) and the recognition of cardiovascular events need to be developed and targeted at emerging adults and parents/guardians.203, 204, 205, 206 Promoting cardiorespiratory fitness is also be an integral aspect of CVD prevention. An analysis of over 1.5 million Swedish men conscripted to the military (mean age=18.3) found that higher cardiorespiratory fitness in late adolescence and early adulthood was significantly protective against COVID-19-associated hospitalization (OR: 0.76), ICU admission (OR: 0.61), and mortality later in life (OR: 0.56).207 Encouraging cardiorespiratory fitness among emerging adults now will likely improve cardiovascular outcomes if faced with future pandemics. Strengths and limitations The strengths of this review include its comprehensive search strategy, an updated search, and an in-depth analysis of the acquired evidence. As per scoping review guidelines20, included articles were not critically appraised. Therefore, any conclusions drawn from this evidence set should act as a foundation, rather than a guideline, for future research, policy, and practice. The authors also acknowledge the potential for bias and subjectivity in how data were reported given the qualitative nature of scoping reviews. In addition to the writing team including multiple members as a means of limiting individual bias, data are reported thoroughly in the supplemental files and available for review. The major limitation of this review, however, is that the majority of available cohort analyses focused on student-athlete populations; this reduces the generalizability of our findings to the broader emerging adult population. In addition, despite including articles from non-English-speaking countries when available in English, excluding non-English-language articles likely skews the evidence set and prevents extrapolation to the global population. Moreover, the COVID-19 pandemic has been dynamic with respect to the transmissibility and pathogenicity of the current variants at large, rendering older evidence less relevant. Nonetheless, this review provides a novel perspective and foundation for future research, policy, and practice. Conclusions To the best of our knowledge, this review is the first to focus on studies that describe the impact of COVID-19 on the cardiovascular health of emerging adults. Among otherwise healthy emerging adults who contracted COVID-19, rare and sometimes fatal cardiovascular presentations were reported. Compared to younger age cohorts, MIS in emerging adults has been associated with a greater risk of cardiovascular involvement, specifically myocarditis, thrombosis, and mortality. Limited data for emerging adults suggest that obesity, hypertension, CVD, or belonging to marginalized groups increase the risk of severe COVID-19, related cardiovascular complications, and mortality. Future research needs to further define the prevalence and risk factors associated with COVID-19-associated cardiovascular complications in this demographic. Screening and treatment protocols for COVID-19-related cardiovascular complications (e.g. triad testing, cMRI) still require further development and validation using data from the broader emerging adult population. Within health care practice, several measures are encouraged to foster long-term cardiovascular health among emerging adults: i) conducting differential diagnoses for cardiovascular issues in those with COVID-19-like symptoms; ii) promoting health literacy and education among emerging adults and parents/guardians, including cardiorespiratory fitness; and iii) expanding telehealth accessibility. While COVID-19 vaccines are still recommended for emerging adults, policymakers and clinicians should be aware of the potential cardiovascular risks and utilize the most recent surveillance data to guide vaccination policies for this age group. These actions should prove useful while the scientific community continues to unravel the long-term cardiovascular implications of COVID-19 among emerging adults. Funding Sources Funding for publication was provided by the senior co-authors: i) KMD, supported by the Canada Research Chair in Nutrition Informatics grant; and ii) VT, supported by the Canadian Institutes of Health Research Health System Impact Fellowship grant (#177475). The agencies responsible for these grants had no involvement in this project. Disclosures None. Uncited reference 97., 99., 100., 101., 102., 103., 104., 105., 106., 107., 108., 109., 110., 111., 112., 113., 114., 115., 116., 117., 118., 119., 121., 123., 124., 125., 126., 127., 128., 129.. Supplementary data Acknowledgments The authors express their gratitude to: Denise Smith, a librarian at the McMaster University Health Sciences Library, for helping develop and refine the final search strategy; Dr. Philip Joseph, MD, associate professor at McMaster University, for his feedback; and several research assistants, including Bismah Jameel, who assisted with article screening and data extraction. ==== Refs References 1 Grant M.C. Geoghegan L. Arbyn M. The prevalence of symptoms in 24,410 adults infected by the novel coronavirus (SARS-CoV-2; COVID-19): a systematic review and meta-analysis of 148 studies from 9 countries PLoS ONE 15 2020 10.1371/journal.pone.0234765 e0234765 2 Nishiga M. Wang D.W. Han Y. Lewis D.B. Wu J.C. COVID-19 and cardiovascular disease: from basic mechanisms to clinical perspectives Nat Rev Cardiol 17 2020 543 558 10.1038/s41569-020-0413-9 32690910 3 Matsushita K. Ding N. Kou M. The relationship of COVID-19 severity with cardiovascular disease and its traditional risk factors: a systematic review and meta-analysis Glob Heart 15 1 2020 10.5334/gh.814 4 Karbalai Saleh S. Oraii A. Soleimani A. The association between cardiac injury and outcomes in hospitalized patients with COVID-19 Intern Emerg Med 15 2020 1415 1424 10.1007/s11739-020-02466-1 32772283 5 Heart and Stroke Foundation of Canada. Don’t wait for COVID-19 to end to address health concerns. Available at: https://www.heartandstroke.ca/en/articles/don-t-wait-for-covid-19-to-end-to-address-serious-health-concerns/. Accessed on March 31, 2021. 6 Puntmann V.O. Carerj M.L. Wieters I. Outcomes of cardiovascular magnetic resonance imaging in patients recently recovered from coronavirus disease 2019 (COVID-19) JAMA Cardiol 1265 2020 5 10.1001/jamacardio.2020.3557 7 Arnett J.J. Emerging adulthood: a theory of development from the late teens through the twenties Am Psychol 55 5 2000 469 10.1037/0003-066X.55.5.469 10842426 8 Lally M, Valentine-French S. Emerging and Early Adulthood. In: Hanson A, Elder A, eds. Parenting and Family Diversity Issues. Iowa State University Digital Press, 2020. 9 American College of Cardiology. Intermediate and Long-Term Impact of COVID-19 on Cardiovascular Disease. Available at: https://www.acc.org/latest-in-cardiology/articles/2021/04/21/13/08/intermediate-and-long-term-impact-of-covid-19-on-cardiovascular-disease. Accessed on June 26, 2021. 10 Brito D. Meester S. Yanamala N. High prevalence of pericardial involvement in college student athletes recovering from COVID-19 JACC Cardiovasc Imaging 14 3 2020 541 555 10.1016/j.jcmg.2020.10.023 33223496 11 Starekova J. Bluemke D.A. Bradham W.S. Evaluation for myocarditis in competitive student athletes recovering from coronavirus disease 2019 with cardiac magnetic resonance imaging JAMA Cardiol 6 8 2021 945 950 10.1001/jamacardio.2020.7444 33443537 12 Statistics Canada. Population estimates on July 1st, by age and sex. Available at: https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1710000501. Accessed on November 11, 2022. 13 Government of Canada. COVID-19 daily epidemiology update. Available at: https://health-infobase.canada.ca/covid-19/epidemiological-summary-covid-19-cases.html. Accessed on May 13, 2021. 14 Valverde I. Singh Y. Sanchez-de-Toledo J. Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with COVID-19 infection in Europe Circulation 143 2021 21 32 10.1161/CIRCULATIONAHA.120.050065 33166189 15 Centers for Disease Control and Prevention. Multisystem Inflammatory Syndrome in Adults (MIS-A). Available at: https://www.cdc.gov/mis-c/mis-a.html. Accessed on May 31, 2021. 16 Karlstad Ø Hovi P. Husby A. SARS-CoV-2 vaccination and myocarditis in a Nordic cohort study of 23 million residents JAMA Cardiol 7 6 2022 600 612 10.1001/jamacardio.2022.0583 35442390 17 Oster M.E. Shay D.K. Su J.R. Myocarditis cases reported after mRNA-based COVID-19 vaccination in the US from December 2020 to August 2021 JAMA 327 4 2022 331 340 10.1001/jama.2021.24110 35076665 18 Oster A.M. Transmission dynamics by age group in COVID-19 hotspot counties - United States Morb Mortal Wkly Rep 2020 69 April-September 2020 10.15585/mmwr.mm6941e1 19 Salvatore P.P. Sula E. Coyle J.P. Recent increase in COVID-19 cases reported among adults aged 18-22 years - United States, May 31-September 5, 2020 Morb Mortal Wkly Rep 69 2020 1419 1424 10.15585/mmwr.mm6939e4 20 Tricco A.C. Lillie E. Zarin W. PRISMA Extension for Scoping Reviews (PRISMA-ScR): checklist and explanation Ann Intern Med 169 2018 467 473 10.7326/M18-0850 30178033 21 Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia. www.covidence.org. 22 Chau V.Q. Giustino G. Mahmood K. Cardiogenic shock and hyperinflammatory syndrome in young males with COVID-19 Circ Heart Fail 13 2020 10.1161/CIRCHEARTFAILURE.120.007485 e007485 23 Razavi A.C. Chang J.L. Sutherland A. Niyogi A. Menard G.E. A 23-year-old man with multisystem inflammatory syndrome after mild COVID-19 J Investig Med High Impact Case Rep 2020 8 10.1177/2324709620974200 2324709620974200 24 Morris S.B. Schwartz N.G. Patel P. Case series of multisystem inflammatory syndrome in adults associated with SARS-CoV-2 infection — United Kingdom and United States, March–August 2020 Morb Mortal Wkly Rep 69 2020 1450 1456 10.15585/mmwr.mm6940e1 25 Kofman A.D. Sizemore E.K. Detelich J.F. Albrecht B. Piantadosi A.L. A young adult with COVID-19 and multisystem inflammatory syndrome in children (MIS-C)-like illness: a case report BMC Infect Dis 716 2020 20 10.1186/s12879-020-05439-z 26 Othenin-Girard A. Regamey J. Lamoth F. Multisystem inflammatory syndrome with refractory cardiogenic shock due to acute myocarditis and mononeuritis multiplex after SARS-CoV-2 infection in an adult Swiss Med Wkly 150 2020 10.4414/smw.2020.20387 w20387 27 Vieira C.B. Ferreira A.T. Cardoso F.B. Paulos J.P. Germano N. Kawasaki-like syndrome as an emerging complication of SARS-CoV-2 infection in young adults Eur J Case Rep Intern Med 7 2020 10.12890/2020_001886 28 Cogan E. Foulon P. Cappeliez O. Dolle N. Vanfraechem G. De Backer D. Multisystem inflammatory syndrome with complete Kawasaki disease features associated with SARS-CoV-2 infection in a young adult. A case report Front Med 428 2020 7 10.3389/fmed.2020.00428 29 Singh V.P. Thalji M. Singh S. Fulminant myocarditis and multisystem inflammatory syndrome in children, in the light of corona virus (Abst) J Am Coll Cardiol 77 18_Suppl_1 2021 1964 https://doi.org/10.1016/S0735-1097(21) 03320–03329 30 Wojnowski K. Ladna J. Kumar S. Multisystem inflammatory syndrome of the adult: an important consequence of COVID-19 (Abst) Chest 160 4 2021 A347 10.1016/j.chest.2021.07.349 31 Nwachukwu I. Fernandes M. Severe multisystem inflammatory syndrome in an adult with myocarditis (Abst) Chest 2021 10.1016/j.chest.2021.07.476 A486 32 Bulut H. Herbers A.H.E. Hageman I.M.G. SARS-CoV-2-induced multisystem inflammatory syndrome in a young adult: case report SN Compr Clin Med 3 8 2021 1773 1779 10.1007/s42399-021-00998-x 34179694 33 MacDonald A. Hussein R. Gowtham S. Multisystem inflammatory syndrome in Pennsylvania: the intersection of adult and pediatric care (Abst) Crit Care Med 77 2021 10.1097/01.ccm.0000726620.93375.08 34 Bonnet M. Champagnac A. Lantelme P. Harbaoui B. Endomyocardial biopsy findings in Kawasaki-like disease associated with SARS-CoV-2 Eur Heart J 2020 10.1093/eurheartj/ehaa588 ehaa588 35 Carvalho V. Damasco P.H. Mello T.S. Gonçalves B. Para-aortic lymphadenopathy associated with adult COVID-19 multisystem inflammatory syndrome BMJ Case Rep 14 12 2021 e246884 10.1136/bcr-2021-246884 36 Moghadam P. Blum L. Ahouach B. Multisystem inflammatory syndrome with particular cutaneous lesions related to COVID-19 in a young adult (Lett) Am J Med 134 1 2021 e36 e37 10.1016/j.amjmed.2020.06.025 32712145 37 Faller E. Barry R. O'Flynn O. Kearney P. Sadlier C. Kawasaki-like multisystem inflammatory syndrome associated with SARS-CoV-2 infection in an adult BMJ Case Rep 14 7 2021 e240845 10.1136/bcr-2020-240845 38 Campoy N.R. Gulati K. Morris K. Delgado F. Multisystem inflammatory syndrome in children (MIS-C) associated with COVID-19/incomplete Kawasaki disease Consultant 61 7 2021 e8 11 10.25270/con.2020.08.00005 39 Aggarwal A, Cohen E, Figueira M, et al. Multisystem inflammatory syndrome in an adult with COVID-19—a trial of anakinra: a case report. Infect Dis Clin Pract (Baltim Md) 2021;e420–e423. 10.1097/IPC.0000000000001028. 40 Al-Mashdali A.F. Al Samawi M.S. A case of post COVID-19 multisystem inflammatory syndrome and Bell’s palsy in a young adult Clin Case Rep 9 9 2021 10.1002/ccr3.4801 e04801 41 Mittal N, Abohelwa M, Brogan J, Nichols J. A case report of multi-system inflammatory syndrome in adults (MIS-A) associated with heart failure. Eur Heart J Case Rep 2021;5(10):ytab381. 10.1093/ehjcr/ytab381. 42 Ejaz K. Patel N. Ramos J. Sharma A. A young adult with COVID-19 associated multisystem inflammatory syndrome (Abst) Am J Respir Crit Care Med 203 2021 10.1164/ajrccm-conference.2021.203.1_MeetingAbstracts.A2468 A2468- 43 Rajendraprasad S. Ahmad F. Nair S. Case of multisystem inflammatory syndrome in children (MIS-C) presenting with rash and shock (Abst) Chest 160 4 2021 A670 10.1016/j.chest.2021.07.637 44 Pombo F. Seabra C. Soares V. Sá A.J. Ferreira I. Mendes M. COVID-19-related multisystem inflammatory syndrome in a young adult Eur J Case Rep Intern Med 8 4 2021 10.12890/2021_002520 45 Vujaklija Brajković A. Zlopaša O. Gubarev Vrdoljak N. Goran T. Lovrić D. Radonić R. Acute liver and cardiac failure in multisystem inflammatory syndrome in adults after COVID-19 (Lett) Clin Res Hepatol Gastroenterol 45 3 2021 101678 10.1016/j.clinre.2021.101678 33716189 46 Chug L. Cabrera N.M. Mathew J. Lock J. Burke L. Sendon C. Multisystem inflammatory syndrome in an adult associated with COVID-19 Crit Care Med 2021 10.1016/j.clinre.2021.101678 92–92 47 Dabas R. Varadaraj G. Sandhu S. Bhatnagar A. Pal R. Kawasaki-like multisystem inflammatory syndrome associated with COVID-19 in an adult: a case report (Abst) Br J Dermatol 185 4 2021 859 861 10.1111/bjd.20574 34115386 48 Salzman M.B. Huang C.W. O’Brien C.M. Castillo R.D. Multisystem inflammatory syndrome after SARS-CoV-2 infection and COVID-19 vaccination Emerg Infect Dis 27 7 2021 1944 1948 10.3201/eid2707.210594 34034858 49 Ciochetto Z. Havens P.L. Aldrete S. Two cases of multi-inflammatory syndrome in children (MIS-C) in adults in 2020 BMC Infect Dis 21 1 2021 1228 10.1186/s12879-021-06911-0 34876052 50 Ronit A. Jørgensen S.E. Roed C. Host genetics and antiviral immune responses in adult patients with multisystem inflammatory syndrome Front Immunol 12 2021 10.3389/fimmu.2021.718744 51 Hékimian G. Kerneis M. Zeitouni M. Coronavirus disease 2019 acute myocarditis and multisystem inflammatory syndrome in adult intensive and cardiac care units (Lett) Chest 159 2 2021 657 662 10.1016/j.chest.2020.08.2099 32910974 52 Bastug A. Aslaner H. Aybar Bilir Y. Multiple system inflammatory syndrome associated with SARS-CoV-2 infection in an adult and an adolescent Rheumatol Int 41 5 2021 993 1008 10.1007/s00296-021-04843-1 33742229 53 Bulathsinghala M. Samson R. A case of COVID-19 associated multisystem inflammatory syndrome resulting in new onset heart failure in an adult (Abst) J Invest Med 69 2 2021 431 54 Szawarski P. Whittaker R. Riyat M. Mandal A. Adult with a paediatric diagnosis after COVID-19 infection (Abst) Intensive Care Med Exp. ESICM LIVES 2021 10.1186/s40635-021-00415-6 Part 2 2021;9(Suppl_1):001226 55 Malik F. Kasten J. VandenHeuvel K. Leino D. Bernieh A. Spectrum of autopsy findings in COVID-19 related pediatric and fetal deaths: 4 cases from a tertiary care children’s center (Abst). Pediatr Dev Pathol Abstracts of the 2021 Fall Meeting 24 6 2021 592 617 10.1177/10935266211057428 56 Nguyen V.T. Zaccarini C. Klawonn M.A. Turk M. Multisystem inflammatory syndrome associated with SARS-CoV-2 infection: a case report (Abst). PM&R AAPM&R Annual Assembly Abstracts 13 Suppl_1 2021 10.1002/pmrj.12735 57 Molina M.F. Al Saud A.A. Al Mulhim A.A. Liteplo A.S. Shokoohi H. Nitrous oxide inhalant abuse and massive pulmonary embolism in COVID-19 Am J Emerg Med 38 2020 1549 10.1016/j.ajem.2020.05.023 e1-1549.e2 58 Rodriguez J.A. Rubio-Gomez H. Roa A.A. Miller N. Eckardt P.A. Co-infection with SARS-CoV-2 and parainfluenza in a young adult patient with pneumonia: case report IDCases 20 2020 10.1016/j.idcr.2020.e00762 e00762 59 Mantovani Cardoso E. Hundal J. Feterman D. Magaldi J. Concomitant new diagnosis of systemic lupus erythematosus and COVID-19 with possible antiphospholipid syndrome. Just a coincidence? A case report and review of intertwining pathophysiology Clin Rheumatol 39 2020 2811 2815 10.1007/s10067-020-05310-1 32720260 60 Kandori K. Narumiya H. Iizuka R. Extracorporeal cardiopulmonary resuscitation should not be performed on confirmed or suspected COVID-19 patients (Lett) Resuscitation 153 2020 6 7 10.1016/j.resuscitation.2020.05.040 32492456 61 Wongkittichote P. Watson J.R. Leonard J.M. Toolan E.R. Dickson P.I. Grange D.K. Fatal COVID-19 infection in a patient with long-chain 3-hydroxyacyl-CoA dehydrogenase deficiency: a case report JIMD Rep 56 2020 40 45 10.1002/jmd2.12165 33204595 62 Jeantin L. Pichereau C. Pineton de Chambrun M. Myocarditis, paraparesia and ARDS associated to COVID-19 infection Heart Lung 50 2021 6 8 10.1016/j.hrtlng.2020.10.008 63 Soquet J. Rousse N. Moussa M. Heart retransplantation following COVID-19 illness in a heart transplant recipient J Heart Lung Transplant 39 2020 983 985 10.1016/j.healun.2020.06.026 32718694 64 Garau G. Joachim S. Duliere G.-L. Sudden cardiogenic shock mimicking fulminant myocarditis in a surviving teenager affected by severe acute respiratory syndrome coronavirus 2 infection ESC Heart Fail 8 2021 766 773 10.1002/ehf2.13049 33190387 65 Crippa S. Kagi G. Graf L. Meyer Sauteur P.M. Kohler P. Stroke in a young adult with mild COVID-19 suggesting endotheliitis New Microbes New Infect 100781 2020 38 10.1016/j.nmni.2020.100781 66 Beşler M.S. Arslan H. Acute myocarditis associated with COVID-19 infection Am J Emerg Med 38 2020 2489 10.1016/j.ajem.2020.05.100 e1-2489.e2 67 Garot J. Amour J. Pezel T. SARS-CoV-2 fulminant myocarditis JACC Case Rep 2 2020 1342 1346 10.1016/j.jaccas.2020.05.060 32835276 68 Kim I.-C. Kim J.Y. Kim H.A. Han S. COVID-19-related myocarditis in a 21-year-old female patient Eur Heart J 1859 2020 41 10.1093/eurheartj/ehaa288 69 Alizadehasl A. Salehi P. Roudbari S. Peighambari M.M. Infectious endocarditis of the prosthetic mitral valve after COVID-19 infection Eur Heart J 4604 2020 41 10.1093/eurheartj/ehaa852 70 Fatehi P. Hesam-Shariati N. Abouzaripour M. Fathi F. Hesam Shariati M.B. Acute ischemic and hemorrhagic stroke and COVID-19: case series SN Compr Clin Med 2 2020 2396 2401 10.1007/s42399-020-00559-8 33024934 71 Krishnan U.S. Krishnan S.S. Jain S. SARS-CoV-2 infection in patients with Down syndrome, congenital heart disease, and pulmonary hypertension: is Down syndrome a risk factor? J Pediatr 225 2020 246 248 10.1016/j.jpeds.2020.06.076 32610168 72 Carneiro T. Dashkoff J. Leung L.Y. Intravenous tPA for acute ischemic stroke in patients with COVID-19 J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc 105201 2020 29 10.1016/j.jstrokecerebrovasdis.2020.105201 73 Calhoun A. Fatade Y. Garcia M. Fulminant myocarditis in COVID-19 with rapid recovery of systolic function (Abst) J Am Coll Cardiol 77 18_Suppl_1 2021 2012-. https://doi.org/10.1016/S0735-1097(21) 03368–4 74 Choudhary K. Regante R. Fitzgibbons T. Ram S. Miskovsky J. Acute myocarditis and acute decompensated heart failure with reduced ejection fraction in COVID-19 (Abst) J Am Coll Cardiol 77 18_Suppl_1 2021 3015-. https://doi.org/10.1016/S0735-1097(21) 04370–04379 75 Volis I. Livneh I. Hussein K. Raz-Pasteur A. COVID-19-associated suspected myocarditis as the etiology for recurrent and protracted fever in an otherwise healthy adult Am J Med Sci 361 4 2021 522 525 10.1016/j.amjms.2020.11.001 33546881 76 Laleh Far V. Najafizadeh S.R. Eslami M. Mollazadeh R. A flare up of idiopathic hypereosinophilic syndrome due to COVID-19 Eur Heart J 42 9 2021 10.1093/eurheartj/ehaa714 954- 77 Venditti L. Rousseau A. Ancelet C. Papo T. Denier C. Susac syndrome following COVID-19 infection (Lett) Acta Neurol Belg 121 3 2021 807 809 10.1007/s13760-020-01554-5 33236280 78 Bhattarai P. Allen H. Aggarwal A. Madden D. Dalton K. Unmasking of Addison’s disease in COVID-19 SAGE Open Med Case Rep 9 2021 10.1177/2050313X211027758 2050313X211027758 79 Bozan Ö Atiş Ş.E. Çekmen B. A rare complication of COVID-19: spontaneous pneumothorax following pneumomediastinum; case report Am J Emerg Med 47 2021 342 10.1016/j.ajem.2021.02.067 e1-342.e2 80 O’Connor A. Loganathan S. Aziz R. COVID-19 related myocarditis following laparoscopic appendicectomy (Abst) Br J Surg Suppl_7 2021 108 10.1093/bjs/znab311.028 znab311-028 81 Recalcati S. Piconi S. Franzetti M. Barbagallo T. Prestinari F. Fantini F. Colchicin treatment of COVID-19 presenting with cutaneous rash and myopericarditis (Lett) Dermatol Ther 33 6 2020 e13891 10.1111/dth.13891 32584431 82 Dahou S. Leach T. Bostock K. Louden J. Raj J. Ghedia S. Challenges in treating new onset systemic lupus erythematosus with lupus nephritis and COVID-19 infection overlap (Abst) Rheumatol Adv Pract Suppl_1 2020 4 10.1093/rap/rkaa052.015 rkaa052-015 83 Sheha D. El-Shayeb M. Eid Y. Unfolding of sickle cell trait by coronavirus disease 2019 (COVID-19) infection Br J Haematol 191 2 2020 e38 40 10.1111/bjh.17089 32966591 84 Singh S. Foster A. Khan Z. Siddiqui A. Atere M. Nfonoyim J.M. COVID-19-induced diabetic ketoacidosis and acute respiratory distress syndrome in an obese 24-year-old type I diabetic Am J Case Rep 21 2020 10.12659/AJCR.925586 e925586-1-e925586-6 85 Han D. Sun Y. Xie R. Zhu X. Zhong Z. A case of severe COVID-19 in a patient with acute graft-versus-host disease after haploidentical transplantation Case Rep Hematol 21 2020 10.12659/AJCR.925586 e925586–1 86 Strause J. Atsina K.B. Bialo D. Orwitz J. Wolf R.L. Cucchiara B. Ischemic stroke associated with aneurysmal lenticulostriate vasculopathy and symmetric reversible basal ganglia lesions in COVID-19 (Lett) J Neurol Sci 117484 2021 426 10.1016/j.jns.2021.117484 87 Merdad G.A. Seadawi L.E. Mustafa A.A. Peptic ulcer associated with COVID-19 in Saudi Arabia Saudi Med J 42 9 2021 1036 1040 10.15537/smj.2021.42.9.20210224 34470844 88 Bunawan N.C. Mokoagow M.I. Djojo A.Y. Delayed RT-PCR time-to-positivity in an adult with SARS-CoV-2 infection J Infect Dev Ctries 15 7 2021 913 917 10.3855/jidc.14766 34343115 89 Gaglani B. Westphal N. Bryant C. Successful management of COVID-19-induced ARDS using VV ECMO in a patient with a BMI of 73 kg/m2 (Abst) Crit Care Med 70 1 2021 49 10.1097/01.ccm.0000726568.91826.1f 90 Pasqualetto M.C. Sorbo M.D. Vitiello M. Pulmonary hypertension in COVID-19 pneumoniae: it is not always as it seems Eur J Case Rep Intern Med 7 12 2020 002160 10.12890/2020_002160 33457379 91 Rashed E. Cagliostro M. Kamran M. Acute myocardial infarction in a young man in the COVID-19 era: a case report (Abst). Circulation Suppl_3 2020 142 10.1161/circ.142.suppl_3.17112 A17112- 92 Freeman H. Moulton B. Propofol-related infusion syndrome in a patient with ARDS secondary to the novel COVID-19 (Abst) Chest 160 4 2021 A815 10.1016/j.chest.2021.07.768 93 Ouzts P.,M.D. McKeag I.,M.D. Asif I. Elevated troponins in a cross-country athlete diagnosed with COVID-19 (Abst). Clin J Sport Med AMSSM Case Podium Presentations 31 2 2021 e31 10.1097/JSM.0000000000000917 94 Edwards K. Hussain I. Two cases of severe autoimmune thyrotoxicosis following SARS-CoV-2 infection J Investig Med High Impact Case Rep 2021 9 10.1177/23247096211056497 23247096211056497 95 Munoz D. Malik H. Eickenhorst D. Newman S. Varughese C. Ali F. Cardiac screening in a young adult male leading to discovery of post-COVID myocarditis with asymptomatic large apical left ventricular thrombus CASE (Phila) 5 5 2021 309 312 10.1016/j.case.2021.07.008 34712875 96 Eid M.M. Co-infection with COVID-19 and malaria in a young man Dubai Med J 4 2 2021 164 166 10.1159/000514254 97 Clerico M. Dogliotti I. Calcagno A. COVID-19 in a post-transplant heart recipient who developed aggressive lymphoma: a biphasic course during rituximab treatment HemaSphere 5 7 2021 e592 10.1097/HS9.0000000000000592 34131632 98 Hasnie U. Andrikopoulou E. Lloyd S. False start: delayed COVID-19 myocarditis in a young athlete (Abst) J Am Coll Cardiol 77 18_Suppl_1 2021 1996-. https://doi.org/10.1016/S0735-1097(21) 03352–0 99 Sikandar B.H. Butler S. Battula A. Shetty R. ST-elevation myocardial infarction in a 23-year-old female: the mystery of thrombus formation Cureus 13 5 2021 10.7759/cureus.15302 100 Hekmatikar A.H.A. Shamsi M.M. Ashkazari Z.S.Z. Suzuki K. Exercise in an overweight patient with COVID-19: a case study Int J Environ Res Public Health 18 11 2021 5882 10.3390/ijerph18115882 34070847 101 Guendouz C. Quenardelle V. Riou-Comte N. Pathogeny of cerebral venous thrombosis in SARS-CoV-2 infection Med (Baltimore) 100 10 2021 e24708 10.1097/MD.0000000000024708 102 Bozorgmehr R. Tajabadi Z. Hemoptysis and hematuria as the initial symptoms of COVID-19: a case report Tanaffos 20 1 2021 75 78 34394374 103 Hirschbaum J.H. Bradley C.P. Kingsford P. Mehra A. Kwan W. Recurrent massive pulmonary embolism following catheter directed thrombolysis in a 21-year-old with COVID-19: a case report Eur Heart J Case Rep 5 4 2021 10.1093/ehjcr/ytab140 ytab140 104 Hussein M.H. Alabdaljabar M.S. Alfagyh N. Badran M. Alamiri K. Splanchnic venous thrombosis in a nephrotic patient following COVID-19 infection: a case report BMC Nephrol 22 1 2021 1 5 10.1186/s12882-021-02643-0 33397327 105 Doğan A.C. Güner A. Avcı Y. Zencirkiran Agus H. Güner E.G. Pulmonary embolism in a young man infected with COVID-19 pneumonia Turk Kardiyol Dern Ars 48 7 2020 714 10.5543/tkda.2020.03688 33034574 106 Bhasin V. Carrillo M. Ghosh B. Moin D. Maglione T.J. Kassotis J. Reversible complete heart block in a patient with coronavirus disease 2019 Pacing Clin Electrophysiol 44 11 2021 1939 1943 10.1111/pace.14321 34289133 107 Abdullah A. Neurath M.F. Atreya R. Mild COVID-19 symptoms in an infliximab-treated ulcerative colitis patient: can ongoing anti-TNF therapy protect against the viral hyperinflammatory response and avoid aggravated outcomes? Visc Med 36 4 2020 338 342 10.1159/000508740 32999889 108 Ghosh R. Roy D. Sengupta S. Benito-León J. Autonomic dysfunction heralding acute motor axonal neuropathy in COVID-19 J Neurovirol 26 6 2020 964 966 10.1007/s13365-020-00908-2 32918164 109 Long A. Grimaldo F. Spontaneous hemopneumothorax in a patient with COVID-19 Am J Emerg Med 40 2021 228 10.1016/j.ajem.2020.07.065 e1-228.e2 110 Seneviratna A. Fat embolism syndrome or COVID-19 pneumonia: a diagnostic dilemma Sri Lankan J Anaesthesiol 29 2 2021 109 113 10.4038/slja.v29i2.8844 111 Flower L. Gale A. Elfar E. Manson J. Tattersal R. Quick V. Adult onset PIMS-TS with secondary haemophagocytic lymphistiocytosis: into the eye of the cytokine storm Rheumatol Adv Pract Suppl_1 2020 4 10.1093/rap/rkaa053 rkaa053 112 Simpson H.D. Johnson E. Britton J. Braksick S. Alternating hemiparesis in the context of hemolytic uremic syndrome and COVID-19 positivity Epilepsy Behav Rep 100468 2021 16 10.1016/j.ebr.2021.100468 113 Pąchalska M. Effect of individually-tailored TDCS and symbolic art therapy for chronic associative prosopagnosia after infection by SARS-COV-2, neuroCOVID-19 and ischemic stroke (Abst) Acta Neuropsychologica 2021 329 345 114 Aikawa T. Ogino J. Kudo T. Kashiwagi Y. Late-onset endocarditis after coronavirus disease 2019 infection Eur Heart J 42 32 2021 10.1093/eurheartj/ehab065 3108- 115 Marcinkiewicz K. Petryka-Mazurkiewicz J. Nowicki M.M. Acute heart failure in the course of fulminant myocarditis requiring mechanical circulatory support in a healthy young patient after coronavirus disease 2019 Kardiol Pol (Pol Heart J) 79 5 2021 583 584 10.33963/KP.15888 116 Al-Kaf F.A. Garni T.A. Nahes A.H. Sandokji H. Samargandy S. Cardiac tamponade, severe hypothyroidism and acute respiratory distress syndrome (ARDS) with COVID-19 infection J Saudi Heart Assoc 33 1 2021 71 76 10.37616/2212-5043.1235 33936940 117 Krishna V. Morjaria J. Jalandari R. Omar F. Kaul S. Autoptic identification of disseminated mucormycosis in a young male presenting with cerebrovascular event, multi-organ dysfunction and COVID-19 infection IDCases 25 2021 10.1016/j.idcr.2021.e01172 e01172 118 Alam A. Kumar D. A case of COVID-19 myocarditis presenting as inferior wall MI (Abst) Chest 160 4 2021 A244 10.1016/j.chest.2021.07.252 119 O’Sullivan J.S. Lyne A. Vaughan C.J. COVID-19-induced postural orthostatic tachycardia syndrome treated with ivabradine BMJ Case Rep 14 6 2021 e243585 10.1136/bcr-2021-243585 120 Hedayat B. Hosseini K. Chest pain and high troponin level without significant respiratory symptoms in young patients with COVID-19 Caspian J Intern Med 11 Suppl 1 2020 561 565 10.22088/cjim.11.0.561 33425276 121 Kenniff S. Mawed S.A. Al-Ghazawi S. Corcho A.R. Qureshi M. COVID-19 associated diffuse myoclonus and dystonia (CADMAD) syndrome: an immune mediated para-infectious or post-infectious condition? Abst). Neurol 15_Suppl_1 2021 96 122 Arandela K. Samudrala S. Abdalkader M. Reversible cerebral vasoconstriction syndrome in patients with coronavirus disease: a multicenter case series J Stroke Cerebrovasc Dis 30 12 2021 10.1016/j.jstrokecerebrovasdis.2021.106118 106118 123 Petracek L.S. Suskauer S.J. Vickers R.F. Adolescent and young adult ME/CFS after confirmed or probable COVID-19 Front Med 2021 8 10.3389/fmed.2021.668944 668944 124 Elikowski W. Fertała N. Zawodna-Marszałek M. Marked self-limiting sinus bradycardia in COVID-19 patients not requiring therapy in the intensive care unit - case series report Pol Merkur Lekarski 49 292 2021 295 302 34464372 125 Heidarpour M. Vakhshoori M. Haghighatpanah M.A. Ashrafi L. Khorvash F. Iraj B. Rhabdomyolysis plus hypocalcemia and diabetic ketoacidosis as concurrent rare COVID-19 manifestations Case Rep Med 2021 2021 10.1155/2021/6625086 e6625086 126 Grisanti S. Schenone C. Biassoni E. Neurological complications of COVID-19: a monocentric experience of a neurological outpatient clinic (Abst) J Neurol Sci 429 2021 10.1016/j.jns.2021.119826 127 Croci G.A. Vaira V. Trabattoni D. Emergency lung transplantation after COVID-19: immunopathological insights on two affected patients Cells 10 3 2021 611 10.3390/cells10030611 33801959 128 Mitchell W.B. Davila J. Keenan J. Children and young adults hospitalized for severe COVID-19 exhibit thrombotic coagulopathy Pediatr Blood Cancer 68 7 2021 e28975 10.1002/pbc.28975 33661561 129 Sawalha K. Habash F.J. Vallurupalli S. Paydak H. Theophylline in treatment of COVID-19 induced sinus bradycardia Clin Pract 11 2 2021 332 336 10.3390/clinpract11020047 34205865 130 Warraich M. Bolaji P. Das S. Posterior circulation stroke presenting as a new continuous cough: not always COVID-19 BMJ Case Rep 14 2021 10.1136/bcr-2020-240270 131 Cruz-Utrilla A. Segura De la Cal T, Escribano-Subias P. Giant T wave inversion and dyspnea in the time of coronavirus pandemic Circulation 142 2020 906 909 10.1161/CIRCULATIONAHA.120.049194 32795150 132 Cankay T.U. Besenek M. What do we over-look during COVID-19 pandemic? An adolescent stroke case presumed conversion disorder Psychiatr Danub 32 2020 300 302 10.24869/PSYD.2020.300 32796802 133 Toth E. Dancy L. Amin-Youssef G. Papachristidis A. Dworakowski R. Collateral implications of the COVID-19 pandemic: belated presentation of infective endocarditis in a young patient Eur Heart J 4365 2020 41 10.1093/eurheartj/ehaa633 134 Atuaka C. Foziljonova K. Duarte C.D.A. Gonuguntla V.T. Rodriguez C.A. Kuhn-Basti M. Not everything is COVID: persistent fever and malaise mimicking COVID-19 pneumonia (Abst) Chest 160 4 2021 A422 A423 10.1016/j.chest.2021.07.419 135 Greenfeld S.M. Tadmor T. ‘Catastrophic’ thrombosis in a young patient with acute myeloid leukemia presenting early in the COVID-19 pandemic – a case report Vivo 35 5 2021 2951 2955 10.21873/invivo.12588 136 Daliri M. Hosseini S. Amin A. Possible role of ECMO in multiorgan failure and prolonged CPR: aluminum phosphide poisoning Asia Pac J Med Toxicol 9 4 2020 154 158 137 Balfe C. O’Connor C. Giblin G. Presentation of severe rheumatic mitral stenosis at the peak of the COVID-19 pandemic and the presumptive treatment as severe coronavirus illness Eur J Case Rep Intern Med 7 12 2020 001957 10.12890/2020_001957 33457355 138 Quinlivan R. Desikan M. Cruces F. Pietrusz A. Savvatis K. Clinical outcome of SARS-CoV-2 infection in 7 adults with Duchenne muscular dystrophy attending a specialist neuromuscular centre Neuromuscul Disord 31 7 2021 603 606 10.1016/j.nmd.2021.04.005 34049779 139 Erickson J.L. Poterucha J.T. Gende A. Use of electrocardiographic screening to clear athletes for return to sports following COVID-19 infection Mayo Clin Proc Innov Qual Outcomes 5 2 2021 368 376 10.1016/j.mayocpiqo.2021.01.007 33585801 140 Moulson N. Petek B.J. Drezner J.A. SARS-CoV-2 cardiac involvement in young competitive athletes Circulation 144 4 2021 256 266 10.1161/CIRCULATIONAHA.121.054824 33866822 141 Petek B.J. Moulson N. Baggish A.L. Prevalence and clinical implications of persistent or exertional cardiopulmonary symptoms following SARS-CoV-2 infection in 3597 collegiate athletes: a study from the Outcomes Registry for Cardiac Conditions in Athletes (ORCCA) Br J Sports Med 56 16 2022 913 918 10.1136/bjsports-2021-104644 34725052 142 Vago H. Szabo L. Dohy Z. Merkely B. Cardiac magnetic resonance findings in patients recovered from COVID-19 JACC Cardiovasc Imaging 14 6 2021 1279 1281 10.1016/j.jcmg.2020.11.014 33341416 143 Hendrickson B.S. Stephens R.E. Chang J.V. Cardiovascular evaluation after COVID-19 in 137 collegiate athletes: results of an algorithm-guided screening Circulation 143 19 2021 1926 1928 10.1161/CIRCULATIONAHA.121.053982 33970675 144 Małek ŁA Marczak M. Miłosz-Wieczorek B. Cardiac involvement in consecutive elite athletes recovered from COVID-19: a magnetic resonance study J Magn Reson Imaging 53 6 2021 1723 1729 10.1002/jmri.27513 33474768 145 Rajpal S. Tong M.S. Borchers J. Cardiovascular magnetic resonance findings in competitive athletes recovering from COVID-19 infection (Lett) JAMA Cardiol 6 1 2021 116 118 10.1001/jamacardio.2020.4916 32915194 146 Gervasi S.F. Pengue L. Damato L. Monti R. Pradella S. Pirronti T. Is extensive cardiopulmonary screening useful in athletes with previous asymptomatic or mild SARS-CoV-2 infection? Br J Sports Med 55 2021 54 61 10.1136/bjsports-2020-102789 33020140 147 Boehmer TK, Kompaniyets L, Lavery AM, et al. Association between COVID-19 and myocarditis using hospital-based administrative data — United States, March 2020–January 2021. Morb Mortal Wkly Rep 2021;70(35):1228–1232. 10.15585/mmwr.mm7035e5. 148 Belay E.D. Abrams J. Oster M.E. Trends in geographic and temporal distribution of US children with multisystem inflammatory syndrome during the COVID-19 pandemic JAMA Pediatr 175 8 2021 837 845 10.1001/jamapediatrics.2021.0630 33821923 149 Whitworth H. Sartain S.E. Kumar R. Rate of thrombosis in children and adolescents hospitalized with COVID-19 or MIS-C Blood 138 2 2021 190 198 10.1182/blood.2020010218 33895804 150 Lewis M.J. Anderson B.R. Fremed M. Impact of coronavirus disease 2019 (COVID-19) on patients with congenital heart disease across the lifespan: the experience of an academic congenital heart disease center in New York City J Am Heart Assoc 9 2020 10.1161/JAHA.120.017580 e017580 151 Broberg C.S. Kovacs A.H. Sadeghi S. COVID-19 in adults with congenital heart disease J Am Coll Cardiol 77 13 2021 1644 1655 10.1016/j.jacc.2021.02.023 33795039 152 Fisher J.M. Badran S. Li J.T. Votava-Smith J.K. Sullivan P.M. Characteristics and outcomes of acute COVID-19 infection in paediatric and young adult patients with underlying cardiac disease Cardiol Young 32 8 2022 1261 1267 10.1017/S1047951121004029 34588090 153 Kibbey M.M. Fedorenko E.J. Farris S.G. Anxiety, depression, and health anxiety in undergraduate students living in initial US outbreak “hotspot” during COVID-19 pandemic Cogn Behav Ther 2021 1 13 10.1080/16506073.2020.1853805 32954958 154 Majumdar P. Biswas A. Sahu S. COVID-19 pandemic and lockdown: cause of sleep disruption, depression, somatic pain, and increased screen exposure of office workers and students of India Chronobiol Int 37 2020 1191 1200 10.1080/07420528.2020.1786107 32660352 155 Altonen B.L. Arreglado T.M. Leroux O. Murray-Ramcharan M. Engdahl R. Characteristics, comorbidities and survival analysis of young adults hospitalized with COVID-19 in New York City PLoS ONE 15 2020 10.1371/journal.pone.0243343 e0243343 156 Adams S.H. Park M.J. Schaub J.P. Brindis C.D. Irwin C.E. Medical vulnerability of young adults to severe COVID-19 illness—data from the National Health Interview Survey J Adolesc Health 67 2020 362 368 10.1016/j.jadohealth.2020.06.025 32674964 157 Fathi M. Markazi Moghaddam N. Kheyrati L. Development and validation of models for two-week mortality of inpatients with COVID-19 infection: a large prospective cohort study Stat Anal Data Min 15 5 2022 586 597 10.1002/sam.11572 158 Richardson S. Gitlin J. Kozel Z. In-hospital 30-day survival among young adults with coronavirus disease 2019: a cohort study Open Forum Infect Dis 8 6 2021 10.1093/ofid/ofab233 ofab233 159 Sandoval M. Nguyen D.T. Vahidy F.S. Graviss E.A. Risk factors for severity of COVID-19 in hospital patients age 18-29 years PLoS ONE 16 7 2021 10.1371/journal.pone.0255544 e0255544 160 Ratchford S.M. Stickford J.L. Province V.M. Stute N. Augenreich M.A. Koontz L.K. Vascular alterations among young adults with SARS-CoV-2 Am J Physiol Heart Circ Physiol 320 2021 H404 H410 10.1152/ajpheart.00897.2020 33306450 161 Szeghy R.E. Province V.M. Stute N.L. Carotid stiffness, intima-media thickness and aortic augmentation index among adults with SARS-CoV-2 Exp Physiol 107 7 2022 694 707 10.1113/EP089481 33904234 162 Nandadeva D. Young B.E. Stephens B.Y. Blunted peripheral but not cerebral vasodilator function in young otherwise healthy adults with persistent symptoms following COVID-19 Am J Physiol Heart Circ Physiol 321 3 2021 H479 H484 10.1152/ajpheart.00368.2021 34296966 163 Stute N.L. Stickford J.L. Province V.M. Augenreich M.A. Ratchford S.M. Stickford A.S.L. COVID-19 is getting on our nerves: sympathetic neural activity and haemodynamics in young adults recovering from SARS-CoV-2 J Physiol 599 18 2021 4269 4285 10.1113/JP281888 34174086 164 Stute N.L. Stickford A.S.L. Stickford J.L. Altered central and peripheral haemodynamics during rhythmic handgrip exercise in young adults with SARS-CoV-2 Exp Physiol 107 7 2022 708 721 10.1113/EP089820 34311498 165 Udelson J.E. Rowin E.J. Maron B.J. Return to play for athletes after COVID-19 infection: the fog begins to clear JAMA Cardiol 6 9 2021 997 999 10.1001/jamacardio.2021.2079 34042956 166 van Hattum J.C. Spies J.L. Verwijs S.M. Cardiac abnormalities in athletes after SARS-CoV-2 infection: a systematic review BMJ Open Sport Exerc Med 7 4 2021 10.1136/bmjsem-2021-001164 e001164 167 Ferreira V.M. Schulz -Menger Jeanette Holmvang G. Cardiovascular magnetic resonance in nonischemic myocardial inflammation J Am Coll Cardiol 72 24 2018 3158 3176 10.1016/j.jacc.2018.09.072 30545455 168 Sands-Lincoln M. Huang H. Jackson G.P. Wang S. Towards a better understanding of COVID-19 among young adults ages 18-24 (Abst) Value Health 24 2021 10.1016/j.jval.2021.04.1269 S121 169 Daniels C.J. Rajpal S. Greenshields J.T. Prevalence of clinical and subclinical myocarditis in competitive athletes with recent SARS-CoV-2 infection: results from the Big Ten COVID-19 Cardiac Registry JAMA Cardiol 6 9 2021 1078 1087 10.1001/jamacardio.2021.2065 34042947 170 National Library of Medicine. Sinus Tachycardia. Available at: https://www.ncbi.nlm.nih.gov/books/NBK553128/. Accessed on August 29, 2022. 171 Mitrani R.D. Dabas N. Goldberger J.J. COVID-19 cardiac injury: implications for long-term surveillance and outcomes in survivors Heart Rhythm 17 2020 1984 1990 10.1016/j.hrthm.2020.06.026 32599178 172 Yadav R. Bansal R. Budakoty S. Barwad P. COVID-19 and sudden cardiac death: a new potential risk Indian Heart J 72 5 2020 333 336 10.1016/j.ihj.2020.10.001 33189190 173 Bohm P. Scharhag J. Egger F. Sports-related sudden cardiac arrest in Germany Can J Cardiol 37 1 2021 105 112 10.1016/j.cjca.2020.03.021 32464107 174 Tilles J.G. Elson S.H. Shaka J.A. Abelmann W.H. Lerner A.M. Finland M. Effects of exercise on Goxsackie A9 myocarditis in adult mice Proc Soc Exp Biol Med 117 3 1964 777 782 10.3181/00379727-117-29696 14244953 175 Petek B.J. Moulson N. Drezner J.A. Cardiovascular outcomes in collegiate athletes after SARS-CoV-2 infection: 1-year follow-up from the outcomes registry for cardiac conditions in athletes Circulation 145 22 2022 1690 1692 10.1161/CIRCULATIONAHA.121.058272 35545946 176 Moulson N. Petek B.J. Churchill T.W. Cardiac troponin testing as a component of return to play cardiac screening in young competitive athletes following SARS-CoV-2 infection J Am Heart Assoc 11 16 2022 10.1161/JAHA.122.025369 e025369 177 Ibrahim E.S.H. Frank L. Baruah D. Value CMR: towards a comprehensive, rapid, cost-effective cardiovascular magnetic resonance imaging Int J Biomed Imaging 2021 2021 10.1155/2021/8851958 8851958 178 Committee Writing Gluckman T.J. Bhave N.M. 2022 ACC expert consensus decision pathway on cardiovascular sequelae of COVID-19 in adults: myocarditis and other myocardial involvement, post-acute 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 79 17 2022 1717 1756 10.1016/j.jacc.2022.02.003 35307156 179 Beşler M.S. Arslan H. Acute myocarditis associated with COVID-19 infection Am J Emerg Med 38 11 2020 2489 10.1016/j.ajem.2020.05.100 e1-2489.e2 180 Haussner W. DeRosa A.P. Haussner D. COVID-19 associated myocarditis: a systematic review Am J Emerg Med 51 2022 150 155 10.1016/j.ajem.2021.10.001 34739868 181 COVID-19 Immunity Task Force. Seroprevalence in Canada. Available at: https://www.covid19immunitytaskforce.ca/seroprevalence-in-canada/. Accessed on September 28, 2022. 182 Centers for Disease Control and Prevention. Information for Healthcare Providers about Multisystem Inflammatory Syndrome in Children (MIS-C). Available at: https://www.cdc.gov/mis/mis-c/hcp/index.html?CDC_AA_refVal=https%3A%2F%2Fwww.cdc.gov%2Fmis%2Fhcp%2Findex.html. Accessed on November 10, 2022. 183 National Institutes of Health. Therapeutic Management of Hospitalized Pediatric Patients With Multisystem Inflammatory Syndrome in Children (MIS-C) (With Discussion on Multisystem Inflammatory Syndrome in Adults [MIS-A]). Available at: https://www.covid19treatmentguidelines.nih.gov/management/clinical-management-of-children/hospitalized-pediatric-patients–therapeutic-management-of-mis-c/. Accessed on November 10, 2022. 184 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 74 4 2022 e1 20 10.1002/art.42062 35118829 185 Inaba Y. Chen J.A. Bergmann S.R. Prediction of future cardiovascular outcomes by flow-mediated vasodilatation of brachial artery: a meta-analysis Int J Cardiovasc Imaging 26 2010 631 640 10.1007/s10554-010-9616-1 20339920 186 Lopez-Leon S. Wegman-Ostrosky T. Ayuzo del Valle N.C. Long-COVID in children and adolescents: a systematic review and meta-analyses Sci Rep 9950 2022 12 10.1038/s41598-022-13495-5 187 Centers for Disease Control and Prevention. Long COVID - Household Pulse Survey. Available at: https://www.cdc.gov/nchs/covid19/pulse/long-covid.htm. Accessed on September 24, 2022. 188 Han Q. Zheng B. Daines L. Sheikh A. Long-term sequelae of COVID-19: a systematic review and meta-analysis of one-year follow-up studies on post-COVID symptoms Pathog 11 2 2022 269 10.3390/pathogens11020269 189 American Heart Association. COVID-19 CVD Registry. Available at: https://www.heart.org/en/professional/quality-improvement/covid-19-cvd-registry. Accessed on June 3, 2021. 190 Ling R.R. Ramanathan K. Tan F.L. Myopericarditis following COVID-19 vaccination and non-COVID-19 vaccination: a systematic review and meta-analysis Lancet Respir Med 10 7 2022 679 688 10.1016/S2213-2600(22)00059-5 35421376 191 Block J.P. Boehmer T.K. Forrest C.B. Cardiac complications after SARS-CoV-2 infection and mRNA COVID-19 vaccination—PCORnet, United States, January 2021–January 2022 Morb Mortal Wkly Rep 71 14 2022 517 10.15585/mmwr.mm7114e1 192 Buchan S.A. Seo C.Y. Johnson C. Epidemiology of myocarditis and pericarditis following mRNA vaccination by vaccine product, schedule, and interdose interval among adolescents and adults in Ontario, Canada JAMA Netw Open 5 6 2022 e2218505 10.1001/jamanetworkopen.2022.18505 35749115 193 Patone M. Mei X.W. Handunnetthi L. Risks of myocarditis, pericarditis, and cardiac arrhythmias associated with COVID-19 vaccination or SARS-CoV-2 infection Nat Med 28 2 2022 410 422 10.1038/s41591-021-01630-0 34907393 194 Moreira H.G. de Oliveira Júnior MT. Valdigem B.P. Martins C.N. Polanczyk C.A. Position statement on cardiovascular safety of vaccines against COVID-19 – 2022 Arq Bras Cardiol 118 2022 789 796 10.36660/abc.20220179 35508059 195 Centers for Disease Control and Prevention. Investigating Long-Term Effects of Myocarditis. Available at: https://www.cdc.gov/coronavirus/2019-ncov/vaccines/safety/myo-outcomes.html. Accessed on August 7, 2022. 196 Wolter N. Jassat W. Walaza S. Early assessment of the clinical severity of the SARS-CoV-2 omicron variant in South Africa: a data linkage study Lancet 399 10323 2022 437 446 10.1016/S0140-6736(22)00017-4 35065011 197 World Health Organization. Interim statement on decision-making considerations for the use of variant updated COVID-19 vaccines. Available at: https://www.who.int/news/item/17-06-2022-interim-statement-on-decision-making-considerations-for-the-use-of-variant-updated-covid-19-vaccines. Accessed on August 4, 2022. 198 Chalkias S. Harper C. Vrbicky K. A bivalent Omicron-containing booster vaccine against COVID-19 N Engl J Med 387 14 2022 1279 1291 10.1056/NEJMoa2208343 36112399 199 Widmer R.J. Collins N.M. Collins C.S. West C.P. Lerman L.O. Lerman A. Digital health interventions for the prevention of cardiovascular disease: a systematic review and meta-analysis Mayo Clin Proc 90 2015 469 480 10.1016/j.mayocp.2014.12.026 25841251 200 Clark C.J. Alonso A. Spencer R.A. Pencina M. Williams K. Everson-Rose S.A. Predicted long-term cardiovascular risk among young adults in the National Longitudinal Study of Adolescent Health Am J Public Health 104 12 2014 e108 e115 10.2105/AJPH.2014.302148 201 Sørensen K. Van den Broucke S. Fullam J. Health literacy and public health: a systematic review and integration of definitions and models BMC Public Health 80 2012 12 10.1186/1471-2458-12-80 202 Gandrakota N. Ali M.K. Shah M.K. Trends in health information technology use among the US population with and without cardiovascular risk factors, 2012–2018: evidence from the National Health Interview Survey JMIR Public Health Surveill 7 9 2021 10.2196/29990 e29990 203 Lynch E.B. Liu K. Kiefe C.I. Greenland P. Cardiovascular disease risk factor knowledge in young adults and 10-year change in risk factors: the Coronary Artery Risk Development in Young Adults (CARDIA) Study Am J of Epidemiol 164 12 2006 1171 1179 10.1093/aje/kwj334 17038418 204 Afroze R. Gulati C. Mora-Almanza J.G. Thakkar V. Davison K.M. Nutrition and other wellness indicators are associated with healthy weight status in emerging adults (Abst) J Nutr Educ Behav 52 7 2020 10.1016/j.jneb.2020.04.085 S34 205 Lawrence E. Mollborn S. Hummer R. Health lifestyles across the transition to adulthood: implications for health Soc Sci Med 193 2017 23 32 10.1016/j.socscimed.2017.09.041 28992538 206 Gooding H.C. Gidding S.S. Moran A.E. Challenges and opportunities for the prevention and treatment of cardiovascular disease among young adults: report from a National Heart, Lung, and Blood Institute working group J Am Heart Assoc 9 2020 10.1161/JAHA.120.016115 e016115 207 Af Geijerstam A. Mehlig K. Börjesson M. Fitness, strength and severity of COVID-19: a prospective register study of 1 559 187 Swedish conscripts BMJ Open 11 7 2021 10.1136/bmjopen-2021-051316 e051316	0	PMC9711905	NO-CC CODE	2022-12-02 23:21:31	no	2022 Dec 1; doi: 10.1016/j.cjcpc.2022.11.005	utf-8	null	null	null	oa_other
==== Front J Dev Econ J Dev Econ Journal of Development Economics 0304-3878 0304-3878 Published by Elsevier B.V. S0304-3878(22)00168-7 10.1016/j.jdeveco.2022.103026 103026 Regular Article Measuring consumption over the phone: Evidence from a survey experiment in urban Ethiopia Abate Gashaw T. a de Brauw Alan a∗ Hirvonen Kalle ab Wolle Abdulazize c a International Food Policy Research Institute (IFPRI), Washington D.C, United States b The United Nations University World Institute for Development Economics Research (UNU-WIDER), Helsinki, Finland c Economics Department, State University of New York at Albany, Albany, NY, United States ∗ Corresponding author. 1 12 2022 1 12 2022 1030267 1 2022 12 7 2022 28 11 2022 © 2022 Published by Elsevier B.V. 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 paucity of reliable, timely household consumption data in many low- and middle-income countries have made it difficult to assess how global poverty has evolved during the COVID-19 pandemic. Standard poverty measurement requires collecting household consumption data, which is rarely collected by phone. To test the feasibility of collecting consumption data over the phone, we conducted a survey experiment in urban Ethiopia, randomly assigning households to either phone or in-person interviews. In the phone survey, average per capita consumption is 23 percent lower and the estimated poverty headcount is twice as high than in the in-person survey. We observe evidence of survey fatigue occurring early in phone interviews but not in in-person interviews; the bias is correlated with household characteristics. While the phone survey mode provides comparable estimates when measuring diet-based food security, it is not amenable to measuring consumption using the ‘best practice’ approach originally devised for in-person surveys. Keywords Survey experiment Phone survey Survey fatigue Food consumption Household surveys ==== Body pmc1 Introduction When it became clear the spread of COVID-19 would become a pandemic in March 2020, many surveys that had been taking place in-person could no longer be fielded due to the concern they would contribute to virus spread. Yet in-person surveys are a key component to many research efforts and monitoring outcomes such as those measuring progress towards the Sustainable Development Goals (SDGs). Without in-person surveys such as the Demographic and Health Surveys (DHS), Household Consumption Expenditure Surveys (HCES), Living Standards Measurement Surveys (LSMS), and other similar surveys conducted by national statistical offices, it is impossible to know what kind of progress is being made towards meeting the SDGs or reducing poverty in general. The main pivot by many researchers during the early part of the pandemic was to begin conducting phone surveys.1 There was a veritable explosion of efforts to collect some type of data to monitor situations over the phone, including major coordinated efforts by Innovations for Poverty Action (RECOVR) and the World Bank (Gourlay et al., 2021). These efforts have played an important role in helping us to understand some of the socioeconomic consequences of the pandemic. In terms of living standards, these surveys generally asked about job loss and loss of income, and they tend to show substantial negative effects (Egger et al., 2021; Josephson et al., 2021; Miguel and Mobarak, 2021). Yet these findings are all based on crude measures, e.g., asking whether household income was lower, the same, or higher than it had been at the same time of the year 12 months ago. Although these surveys provided valuable information about how living standards were qualitatively changing during the early part of the pandemic, there remain obvious ways that phone surveys cannot replace in-person surveys. Some variables require physical measurement; for example, it is impossible to study how stunting prevalence is evolving among children under 5 years of age without in-person data collection. Similarly, collecting data on household consumption expenditures to estimate poverty incidence requires complex measurement.2 The standard household consumption expenditure and poverty measurement involves administering detailed food and non-food consumption modules covering more than 100 items typically consumed in the country (Deaton and Grosh, 2000; Deaton and Zaidi, 2002).3 Consequently, most phone surveys have not attempted to collect such data, in trying to minimize the time spent on the phone. As researchers have shied away from collecting complex data over the phone, we lack data on specific trends through the pandemic. In reviewing impacts on incomes, Miguel and Mobarak (2021) do not even attempt to speak directly to trends in poverty incidence. Despite the fact that modelers have predicted large increases in poverty incidence and rising food insecurity due to policies associated with the pandemic (e.g., Laborde, et al., 2021; Lakner et al., 2021; Sánchez-Páramo et al., 2021; Sumner et al., 2020), the lack of data collected in-person means it is difficult to tell whether their predictions have come true. The surveys that have tried to collect consumption data over the phone during the pandemic suggest the increases in poverty incidence are not as severe as either the crude income measures or models would suggest. Egger et al. (2021) report on phone surveys in Kenya and Sierra Leone that collected data on food consumption in both countries and non-food consumption in Kenya, and find that the value of food consumption increased in both countries, offset by a decline in non-food consumption in Kenya.4 Janssens et al. (2020) study a sample of households in Kenya collecting financial diaries, and find that households sold assets to maintain food consumption levels. Hirvonen et al. (2021) also find no material change in the value of overall food consumption in a representative sample of Addis Ababa between an in-person survey conducted in 2019 and a phone survey conducted at the same time of year in 2020, though the composition of food consumption changed. These surveys suggest it might be plausible to conduct phone surveys to measure consumption as it had been before and therefore poverty incidence, particularly if survey efforts first attempt to develop some rapport with households before the long consumption survey, as is true in all the surveys described above. But it is important to quantify differences between phone and in-person measures of consumption before making such conclusions. Therefore, here we test whether consumption data collected over the phone has a comparable distribution to data collected in-person, using a sample that has been asked about food consumption several times in the past. We randomly select half of the sample to be enumerated about consumption in-person, with the other half enumerated over the phone. We do not include other modules in the survey, so we cannot test other differences between phone and in-person surveys. However, note that we can generate other indicators that are often enumerated in phone surveys, such as the household diet diversity score (HDDS) and a food consumption score (FCS) providing alternative measures of the household's food security. We can then compute poverty incidence using both the consumption measures generated by our phone sample, versus the in-person sample. Note that it is best to at least initially be agnostic about which sample provides closer to a “true” approximation of the distribution of consumption, and therefore poverty incidence. Indeed, an important challenge in survey experiments such as ours is that we do not observe the “true value” against to which to benchmark our estimates (De Weerdt et al., 2020). However, when we test for survey fatigue by randomly changing the order in which the food groups appear in the food consumption module, we observe evidence of survey fatigue occurring very early on in the phone interviews but not in the in-person interviews. It seems then that the in-person survey mode does perform better, resulting in less measurement error than the phone survey mode. Our assessment of data quality based on Benford's law also suggest that the consumption data from the in-person survey are of higher quality than the data from the phone survey. In heterogeneity analysis, we find that bias is attenuated among more educated household heads, and is positively related to household size.5 This finding implies that the measurement error in phone survey mode is not classical and, as a result, cannot be easily corrected with standard methods used in the literature (Bound et al., 2001). This paper contributes to the understanding of how variation in survey designs can shape data quality and ensuing analyses (De Weerdt et al., 2020; McKenzie and Rosenzweig, 2012; Zezza et al., 2017). Much of the previous work has focused on improving consumption measures used to measure poverty incidence (Abate et al., 2020; Ameye et al., 2021; Backiny-Yetna et al., 2017; Beaman and Dillon, 2012; Beegle et al., 2012; Caeyers et al., 2012; De Weerdt et al., 2016; Friedman et al., 2017; Gibson et al., 2015; Gibson and Kim, 2007; Jolliffe, 2001; Kilic and Sohnesen, 2019; Troubat and Grünberger, 2017). We add to this literature by systematically comparing consumption and poverty estimates generated from a phone survey to those from an in-person survey. Finally, many researchers have hypothesized that the phone survey mode is likely to be considerably more vulnerable to response fatigue than the in-person mode, leading to the widespread recommendation to keep phone-based interviews short, and to avoid complex questions (Dabalen et al., 2016; Gourlay et al., 2021). Our results on consumption measurement provide empirical support to this hypothesis. However, in our case, both survey modes result in similar estimates when measuring diet-based food security suggesting that the phone survey mode is appropriate for measuring simpler and cognitively less demanding indicators, as long as the interview time is kept relatively short (Abay et al., 2021a). 2 The survey experiment, data and methods 2.1 The survey experiment We designed a survey experiment to understand the implications of using a phone survey mode for household consumption measurement by systematically contrasting responses from computer assisted personal interviews (CAPI, or in-person) and computer assisted telephone interviews (CATI, or phone). The survey instrument in both survey modes were identical and had four sections. The interview began with a brief section containing only three questions needed to construct household size and its dependency ratio. In the first main section, respondents were asked to report on the household's food consumption for each item from a list of 118 food items, grouped into eight food groups. We first went through the list of 118 items asking whether the household consumed the item in the past seven days or not. The survey instrument was programmed to carry forward all items that were consumed in the past seven days to the next sub-section that asked about the consumption frequency (‘on how many days was the item consumed’) and quantity (‘amount consumed’) within the 7-day period. The second main section of the questionnaire included a short module asking household's food consumption outside of home within the same 7-day recall period. The final main section of the survey included a non-food consumption module, which asked respondents to recall household expenditures during the last month (e.g., toiletries or electricity expenditures) and during the last 12 months (e.g., school fees or health expenditures). The questionnaire administered for the two groups differed, then, only by the interview mode. For all other aspects, the questionnaire designs for the two groups were identical (Table 1 ). The full questionnaire is included in the Online Appendix.Table 1 Comparison of in-person and phone data collection. Table 1 In-person Phone Method of data capture Computer-assisted personal interviewing (CAPI) Computer-assisted telephone interviewing (CATI) Recall period in the food consumption modules 7 days 7 days Recall period in the non-food consumption module () 1 month or 12 months 1 month or 12 months Designated respondent Household member who decides on food purchase and/or preparation Household member who decides on food purchase and/or preparation Consumption measurement 118 food items (frequency and quantity consumed) 118 food items (frequency and quantity consumed) Note: () 1 month for non-food expenditures such as toiletries and utilities and 12 months for expenditures such as school fees and health expenses. 2.2 Household sample The household sample for this survey experiment originates from a randomized control trial (RCT) conducted to assess the impact of video-based behavioral change communication on fruit and vegetable consumption in Addis Ababa, Ethiopia (Abate et al., 2021). The baseline and endline surveys for the RCT took place in September 2019 and February 2020, respectively.6 The sample of 930 households randomly selected from six sub-cities, 20 woredas (districts), and 40 ketenas (neighborhoods; or clusters of households) within Addis Ababa.7 Comparison of household characteristics against those reported in other surveys from Addis Ababa suggest that the sample is representative of the households residing in the city (Hirvonen et al., 2020). The endline survey was administered just before the COVID-19 pandemic was declared in 2020, a setup that was highly optimal for launching COVID-19 phone surveys. Phone numbers were collected from 887 households of the 895 households (99%) that took part in the February 2020 survey. To monitor the food security situation in Addis Ababa during the pandemic, we selected a random subsample of 600 households for monthly phone surveys (Hirvonen et al., 2021). In total, four phone survey rounds were carried out between June and August 2020. In the August 2020 phone survey round, we administered the same food consumption module described above for all households selected for the phone surveys (Hirvonen et al., 2021). Table A1 in the Appendix summarizes the various surveys with the sample of households used in this study. The survey experiment contrasting consumption data collected via in-person and phone modes was administered over a 10-day period in September 2021 (i.e., one year after the last COVID-19 phone survey).8 The sampling frame for this study was based on 895 households that were interviewed during the in-person survey conducted in February 2020, the endline survey of the video RCT. Out of the 895 households, 448 were randomly selected for an in-person interview and 447 for a phone interview.9 A total of 797 households were interviewed; 421 in the in-person group and 376 in the phone group.10 Administering the consumption modules over the phone took 41 min on average (median) and while the average (median) interview duration was 43 min for an in-person visit. The quality of the connection was generally good for the phone interviews, and based on enumerators’ assessment, rarely affected the interview quality.11 The survey team tasked with the in-person surveys followed recommended COVID-19 preventive measures when visiting the households. First, both the enumerators and respondents were provided with facemasks that they were required to wear during the interview. Second, the enumerators were required to thoroughly wash their hands with soap for 20 s or use disinfectant (containing more than 70% alcohol) before entering and when leaving the respondent's premises. Third, the survey coordinator conducted daily check-ups with enumerators regarding any COVID-19 related symptoms. Finally, the interview was conducted outdoors with at least 2-m distance between the enumerator and the respondent. Ethical approval for the survey experiment was obtained from the institutional review boards (IRB) of the International Food Policy Research Institute (IFPRI) and the College of Medicine and Health Sciences at Hawassa University in Ethiopia. Informed oral consent was obtained from all participants at the start of the interview. Enumerators provided respondents a brief overview of the study objectives and informed them that their participation in the study was entirely voluntary. 2.3 Data Food consumed at home was reported in terms of quantities consumed, which we converted into local currency units (Ethiopian birr) using retail price data collected by the Central Statistical Agency (CSA) of Ethiopia. We used the retail price data for Addis Ababa from February 2020 (the latest month available to us) and then used a food-specific consumer price index for Addis Ababa to express our food consumption data in September 2021 prices. Food consumption outside the home as well as non-food expenditure were collected in birr terms, thus requiring no price adjustments. Each household's total consumption was calculated by first converting all consumption expenditure data to weekly terms and then adding up the three consumption components: food consumption at home; food consumption expenditures outside the home; and non-food expenditures. The official poverty data in Ethiopia come from the Household Consumption Expenditure Survey (HCES) collected every five years. The HCES survey is conducted throughout the Ethiopian calendar year to address consumption seasonality and covers nearly 400 food items and more than 850 non-food items. The latest HCES was administered in 2015/16, after which food prices and prices of non-food items have both been rising annually at a double-digit rate. Considering the high inflation rate and the considerable methodological differences between our survey and the HCES, we do not attempt to update the HCES poverty line for September 2021. Instead, we calibrate our poverty line for the in-person sample to match the 16.8 percent poverty headcount based on the national poverty line and reported for Addis Ababa using the 2015/16 HCES (FDRE, 2018). We also use our food consumption data to study how the phone survey mode affects household dietary diversity, an indicator of household food security (Hoddinott and Yohannes, 2002). First, we computed the HDDS of Swindale and Bilinsky (2006) by grouping the 118 food items in our consumption module into 12 food groups: cereals; roots and tubers; vegetables; fruits; meat, poultry and offal; eggs; fish and seafood; pulses, legumes and nuts; milk and milk products; oil and fats; sugar and honey; and miscellaneous foods. The HDDS is a sum of all food groups from which the household consumed food items during the 7-day recall period, with a minimum of one and maximum of 12. Second, we constructed the food consumption score (FCS) developed by the WFP (2008). The FCS combines dietary diversity and consumption frequency by grouping the consumed food items into nine groups and allocating more weight to protein rich foods.12 The weighted FCS index ranges between zero and 112, with higher scores indicating a better food security situation. After dropping two households with implausible consumption values, the final sample of 795 households is formed, out of which 421 are from the in-person group and 374 are from the phone group. Table 2 shows that the in-person and phone groups are similar in terms of basic household characteristics. Moreover, the households in the two sub-samples are balanced in terms of the number of times they had been interviewed since September 2019. We also see no meaningful differences in the household per capita food consumption collected in September 2019, whether we examine means (Table 2) or full distributions (Figure A1 in the Appendix).Table 2 Household characteristics, by survey mode. Table 2Variable In person Phone Difference t-test Mean/[SE] Mean/[SE] p-value Female respondent 0.922 0.917 0.005 0.843 [0.017] [0.018] Household size 4.800 4.832 −0.032 0.792 [0.110] [0.092] Male-headed household () 0.568 0.572 −0.004 0.898 [0.029] [0.036] Head's education in years () 6.675 6.543 0.132 0.655 [0.297] [0.310] Household asset index () −0.035 −0.009 −0.026 0.828 [0.124] [0.161] Number of times the household has been interviewed since September 2019 5.684 5.805 −0.121 0.315 [0.086] [0.082] (log) Household per capita food consumption in September 2019 () 5.570 5.534 0.036 0.416 [0.037] [0.042] Number of households: 421 374 Clusters: 40 Note: Unit of observation is household. Standard errors (SE) are clustered at enumeration area level. Difference in means between the groups tested with a t-test (null-hypothesis: difference in means = 0). Note: N = 795 households. () Based on data collected in previous survey rounds. 2.4 Estimation methods We quantify the difference in reported household per capita consumption values across the two groups using ordinary least squares (OLS). In the most basic model, we regress both the per capita consumption value and its logarithm on a binary treatment variable valued one if the household was randomly selected into the phone group, and zero if into the in-person group. In subsequent models, we control for differences in basic household characteristics (household size, and household head's sex and level of education in years) as well as sub-city fixed effects. Finally, when we discuss percentage differences derived from the coefficients in semi-log regressions they are based on the approximate unbiased variance estimator proposed by van Garderen and Shah (2002): 100×(eβˆ−0.5Vˆ(βˆ)−1), where βˆ refers to the estimated coefficient and Vˆ to the estimated variance. Finally, the standard errors in all household level regressions are clustered at the enumeration area (ketena) level. 3 Results 3.1 Household total per capita consumption Fig. 1 contrasts the full distributions of (log) household weekly per capita consumption measured in birr between households that received an in-person visit and households that were interviewed over the phone. The estimated household consumption distribution for the phone group lies to the left of the distribution estimated for the in-person group, indicating that the whole distribution of total consumption values resulting from the phone survey resulted in lower values than that of the in-person survey.Fig. 1 Distribution of (ln) weekly consumption per capita (in birr), by survey mode. Fig. 1 The regression estimates reported in Table 3 quantify the difference in household weekly food consumption when the data were collected over the phone relative to when the in-person survey mode was used. In columns 1 and 2, the dependent variable is the natural logarithm (ln) of household per capita consumption value in birr, whereas non-logged values are used in columns 3 and 4. Unadjusted estimates are reported in odd columns, whereas estimates in even columns are adjusted for differences in basic household characteristics as described above. Because the differences between the unadjusted and adjusted regressions are negligible, we focus our reporting and discussion on the adjusted regression results.Table 3 Impact of phone survey mode on household weekly per capita consumption. Table 3 (1) (2) (3) (4) Dependent variable: (ln) Household per capita consumption (birr) Household per capita consumption (birr) Phone survey mode −0.271 −0.262* −207.69* −200.61* (0.059) (0.054) (58.16) (52.65) Household level controls? No Yes No Yes Sub-city fixed effects? No Yes No Yes Observations 795 795 795 795 R2 0.051 0.288 0.031 0.232 In-person group mean of the dependent variable n/a n/a 966.27 966.27 Note: Ordinary least squares regression. Unit of observation is household. Household level controls include household size (number of members), indicator variable for male-headed households, and household head's education in years. Standard errors are clustered at the enumeration area level and reported in parentheses. Statistical significance denoted with * p < 0.10, p < 0.05, p < 0.01. Relative to the in-person survey, on average the phone survey mode decreases the reported household per capita consumption expenditures by 23 percent (Table 3, column 2).13 The 95% confidence interval (CI) for this estimate ranges between −14.2 and −31.1. The estimates based on non-logged per capita consumption variable are similar. Considering that the mean per capita consumption in the in-person group is 966 birr, the 201 birr difference reported in Column 4 of Table 3 translates into 21 percent lower average per capita consumption in the phone survey group. 3.2 Components of consumption Food consumed at home represents 50.3 percent of the total consumption among the in-person group and 55.8 percent among the phone survey group.14 The regression estimates reported in Column 1 of Table 4 indicate that the reported per capita food consumption values are 13 percent lower on average when the phone survey mode is used (95-% CI: −5.5; −20.7). However, we do not find strong evidence to suggest that some food groups were more affected than others. We re-estimated the main regression using the value of food consumption for each of seven categories of food as the dependent variable; in Figure A2 in the Appendix, we observe that all the coefficient estimates are negative and suggest 5 to 25 percent lower consumption, with overlapping confidence intervals.Table 4 Impact of phone survey mode on components of household consumption. Table 4 (1) (2) (3) (4) Dependent variable: (ln) Household per capita food consumption at home Household consumed food outside home (0/1) Household per capita food consumption outside home (ln) Household per capita non-food consumption Phone survey mode −0.143 −0.129 −21.66 −0.35* (0.043) (0.056) (8.34) (0.09) Household level controls? Yes Yes Yes Yes Sub-city fixed effects? Yes Yes Yes Yes Observations 795 795 795 795 R2 0.221 0.079 0.062 0.226 In-person group mean of the dependent variable n/a 0.660 53.92 n/a Note: Ordinary least squares regression. Unit of observation is household. 0/1 = binary variable. Household level controls include household size (number of members), indicator variable for male-headed households, and household head's education in years. Standard errors are clustered at the enumeration area level and reported in parentheses. Statistical significance denoted with * p < 0.10, p < 0.05, p < 0.01. About 60 percent of the households in our sample report to have consumed food items outside of their home in the past 7 days. This reporting incidence varies by survey mode with households in the phone survey group being 13 percentage points less likely to report to have consumed foods outside their home (Table 4, column 2). A regression based on a non-logged outcome variable shows that the food expenditures outside of the home are 40.2 percent lower in the phone group relative to the in-person group (Table 4, column 3).15 All the households in our sample report positive (non-zero) non-food consumption values. Column 4 in Table 4 shows the impact of the phone survey mode when the dependent variable is logged weekly per capita non-food consumption. On average, the phone survey mode lowers the reported non-food consumption by 30.1 percent (95-% CI: −15.5; −42.1). 3.3 Poverty estimates Next, we estimate the impact of using phone survey mode on poverty estimates. Since poverty is defined at the individual level, we need to convert our data from household to individual level. To do so, we use a weighted least squares regression method where the weights are frequency weights based on household size. Using our calibrated poverty line, in Table 5 we estimate that poverty rate is 17 percentage points higher when phone survey mode is used compared to when consumption data are collected through in-person visits (95-% CI: 9.99; 24.1). Since the poverty rate in the in-person sample is calibrated at 16.8 percent, using the phone survey mode effectively doubles the poverty rate in this context.Table 5 Impact of phone survey mode on poverty rate. Table 5 (1) (2) Dependent Variable: Consumption Below Poverty Line (0/1) Phone survey mode 0.168 0.170* (0.036) (0.035) Household level controls? No Yes Sub-city fixed effects? No Yes Observations (weighted) 3828 3828 Households 795 795 R2 0.038 0.181 In-person group mean of the dependent variable 0.168 0.168 Note: Weighted least square regression with household size used as a frequency weight. After applying the weight, the unit of observation is individual. Dependent variable obtains value 1 if the household's per capita consumption is below the poverty line, zero otherwise. 0/1 = binary variable. Household level controls include household size (number of members), indicator variable for male-headed households, and household head's education in years. Standard errors are clustered at the enumeration area level and reported in parentheses. Statistical significance denoted with * p < 0.10, p < 0.05, p < 0.01. 3.4 Measures of food security In Table 6 , we report the impacts of using the phone survey mode on two widely used diet-based food security measures, HDDS and FCS. Both can be computed from the food consumption survey data. All four reported impact estimates are relatively small in magnitude and not statistically different from zero. The HDDS and FCS do not require respondents to estimate quantities consumed, only whether the food item was consumed in the past 7 days (HDDS) or the consumption frequency in terms of number of days in the past 7 days (FCS). In contrast, collecting data for food consumption measures is cognitively more demanding because it requires respondents to also estimate quantities consumed in the household during the recall period. Our results therefore indicate that the phone survey mode appears to lead to similar estimates when measuring diet-based food security to in-person surveys but leads to much lower estimates of the value of household food or non-food consumption.Table 6 Impact of phone survey mode on household dietary diversity indicators. Table 6 (1) (2) (3) (4) Dependent variable: Household diet diversity score (HDDS) Food consumption score (FCS) Phone survey mode 0.060 0.058 −2.120 −2.055 (0.132) (0.135) (1.629) (1.646) Household level controls? No Yes No Yes Sub-city fixed effects? No Yes No Yes Observations 795 795 795 795 R2 0.000 0.121 0.003 0.111 In-person group mean of the dependent variable 9.07 9.07 63.97 63.97 Note: Ordinary least squares regression. Unit of observation is household. Household level controls include household size (number of members), indicator variable for male-headed households, and household head's education in years. Standard errors are clustered at the enumeration area level and reported in parentheses. Statistical significance denoted with p < 0.10, p < 0.05, p < 0.01. 4 Mechanisms, extensions, and robustness 4.1 Survey fatigue Our survey experiment shows that the phone survey mode leads households to underestimate their food and non-food consumption expenditures. As a result, if we trusted the phone survey mode and tried to use it in the same manner that we had used in-person surveys to measure poverty prior to the pandemic, we would conclude that the poverty headcount is twice as high using the phone survey data than the data collected in-person. Here, we study whether survey fatigue can help explain differences between results of the two survey modes. The large difference in consumption and poverty incidence estimates between the two survey modes could result from respondent or enumerator fatigue. For example, fatigued respondents pay less attention when responding to cognitively demanding questions (e.g., amount or value of consumption), increasing the risk of measurement error. Survey experts have hypothesized that the risk of respondent fatigue is considerably higher in phone surveys than in in-person surveys (Dabalen et al., 2016; Gourlay et al., 2021). Consequently, it has been widely recommended to keep the phone survey duration short to minimize the risk of survey fatigue (Glazerman et al., 2020; Hoogeveen et al., 2014; Hughes and Velyvis, 2020; Jones and von Engelhardt, 2020; Kopper and Sautmann, 2020). While it is certainly intuitive that the risk of survey fatigue is higher in phone surveys, to the best of our knowledge, no studies have attempted to compare survey fatigue between phone and in-person modes using the same survey form. Evidence from in-person surveys suggests that survey fatigue can lead to under reporting and overall deterioration of data quality in some settings (Ambler et al., 2021; Baird et al., 2008; Schündeln, 2018), but not always (Laajaj and Macours, 2021).16 In a recent phone survey conducted in rural Ethiopia, Abay et al. (2021a) estimate that delaying the timing of a dietary diversity module by 15 min increased the likelihood that the respondents reported not to have consumed from certain food groups, resulting in an 8 percent decline in the mothers’ dietary diversity score.17 To explore the role of survey fatigue, we cross-randomized the order in which the food groups appeared in the first main section of the survey, the “food consumed at home” module.18 Specifically, we implemented two versions of this food consumption module, ordering the food groups differently (see Appendix Table A2). For example, in version 1, mango appeared as the 5th item while in version 2, it appeared as the 73rd item. Similarly, in version 1, rice was the 52nd item on the list while in version 2, it was the 11th item on the list. Exploiting this variation, we use the food item level data to construct a variable that takes on the value of 1 when each food appears later in the questionnaire relative to the other version, and 0 otherwise.19 Using the example above, this variable would be 1 when mangoes appear as the 73rd item, and when rice appears as the 52nd item. Using our food item level data, we then regressed the weekly household per capita consumption of the food item on this binary variable capturing the item's relative position in the questionnaire, and the indicator variable for the phone survey mode. To assess whether the impact of delaying when the item is asked in the module differs between phone and in-person survey modes, we interact the two variables and include the interaction term in the regression. In these regressions we control for food item fixed effects, meaning that our estimates are identified from variation in the survey mode or relative position in the questionnaire for the same food items. As additional controls, we include household size, an indicator variable for male-headed households, the head's years of education, and sub-city fixed effects. Table 7 provides the results. In column 1, we estimate the model without the interaction term. Moving the item later in the questionnaire results in a report that is, on average, 5.8 percent lower for the item than if it takes on its earlier position.20 Using the phone survey mode, the average report suggest the value of consumption is 15.5 percent lower than found with the in-person survey mode. In column 2, we estimate the model with the interaction term. The basic variable now captures the effect of placing the item later in the questionnaire in the in-person survey; this coefficient is close to zero and not statistically significant. The CI is relatively tight around zero (95-% CI: −0.0167; 0.0016) indicating that survey fatigue does not play a role in the in-person survey mode, at least in this relatively early part of the questionnaire. In contrast, the coefficient on the interacted variable is negative, relatively large in magnitude, and statistically different from zero; it suggests that delaying an item in the phone survey mode leads to a report that is 11.9 percent lower on average than an item occurring later in the in-person survey. This finding is strongly suggestive that the in-person mode leads to less survey fatigue than the phone survey mode.Table 7 Impact of item's relative position in the questionnaire and phone survey mode on reported per capita food consumption value measured in birr. Table 7 (1) (2) Dependent Variable: (ln) Household per capita consumption of the food item Item appeared later in the questionnaire −0.230* −0.014 (0.101) (0.159) Phone survey mode −0.615*** −0.368 (0.203) (0.239) Item appeared later in the questionnaire * Phone survey mode −0.458** (0.222) Household level controls? Yes Yes Sub-city fixed effects? Yes Yes Food item fixed effects? Yes Yes Observations 93,810 93,810 In-person group mean of the dependent variable 3.97 3.97 Note: Ordinary least squares regression. Unit of observation is food item consumed (or not) in each household. Number of food items is 118 and number of households is 795 resulting in 93,810 observations. Dependent variable is household per capita consumption of the food item measured in birr. Standard errors are clustered at the food item level and reported in parentheses. Statistical significance denoted with * p < 0.10, p < 0.05, p < 0.01. In Appendix Table A3 we replicate this analysis, only considering the responses to the Yes/No questions regarding whether the household consumed the item or not during the 7-day period. Interestingly, all coefficients in the interacted model appear insignificant implying that only consumption quantity reports are affected, but not responses on whether the household consumed the item or not. This finding is in line with our earlier result according to which diet-based food security measures do not seem to be affected by variation in survey mode. 4.2 Data quality We next use Benford's law as a benchmark for assessing data quality. According to Benford (1938), the distribution of first-digits in many numerical data sets approximately follow the probability (P):P(d)=log10(d+1)−log10(d) where d ∈ {1, …,9} refers to the first-digit of the observation. It is unlikely that survey data perfectly conform to the Benford's law distribution (Kaiser, 2019), but previous work (Abate et al., 2020; Garlick et al., 2020; Schündeln, 2018) has used the distance between the observed distribution and the predicted distribution under Benford's law as a measure of data quality. Here, we calculate this distance separately for the data collected by phone and for the data collected by in-person visits. Following Schündeln (2018), we compute normalized Euclidean distances between the observed first-digit distribution and the one predicted by Benford's law.21 We use the digits of the quantities consumed as reported by the households in the food consumption module. The specific question asks for the quantity consumed and the unit (e.g., kg, litre, cup, or a locally used unit such as tassa). Of note is that Benford's law is scale-invariant; the law holds irrespective of the unit in which the consumed quantities were reported. Figure A3 in the Appendix reports the observed first-digit distributions in our data and compares them to the distribution predicted by Benford's law.22 The null hypothesis that the observed distributions follow Benford's law is rejected for both groups (p < 0.001). However, relative to the in-person group, the phone group is much more likely to report the smallest possible value (i.e., value 1) as the first digit, possibly indicating limited cognitive engagement with the question. Next, we calculate the Euclidean distances separately for each of the 33 consumption units reported by the households and for both survey mode groups. We then test whether the consumption unit specific average Euclidean distances for the two groups are statistically different by regressing the mean distance on our binary treatment variable. Table A4 in the Appendix shows that the coefficient on the treatment variable is positive and statistically different from zero, indicating that the data collected via the phone survey deviate more from the Benford's law than data collected via the in-person survey. This finding suggests that the consumption data from the in-person survey are of higher quality than data from the phone survey. 4.3 Heterogeneity The results show that using the phone survey mode leads to substantial underestimation of household consumption expenditures. It is tempting to think that it could be possible to devise relatively simple adjustment factors to correct for this attenuation bias. Unfortunately, evidence from previous survey experiments suggests that because the measurement error is usually not independent of household characteristics (i.e., non-classical), such adjustment factors do not exist (De Weerdt et al., 2020). To explore the possibility that the phone survey mode varies by household type, we interacted the phone survey indicator variable with the household head's level of education and household size. Table 8 provides the results when household per capita food consumption (Columns 1–2) and non-food consumption (Columns 3–4) is used as the dependent variable. For household food consumption, we observe that the bias decreases with household head's education and increases with household size.23 The former result suggests that respondents from more educated households better overcome survey fatigue in phone surveys. In contrast, the cognitive burden increases with household size as the number of consumption events is higher within the recall period (Fiedler and Mwangi, 2016; Gibson and Kim, 2007). Larger households are bound to have more consumption events than smaller households, making them more vulnerable to survey fatigue. For non-food consumption, the coefficients are of the same sign and magnitude but not statistically different from zero, possibly because of the larger variation in the data relative to the food consumption data. Overall, these heterogenous impacts imply that adjustment factors to account for the bias caused by the phone survey mode cannot be easily developed.Table 8 Regression results from interaction models. Table 8 (1) (2) (3) (4) Dependent variable: (ln) Household food consumption per capita (ln) Household non-food consumption per capita Phone survey mode −0.223 0.073 −0.427* −0.224 (0.060) (0.117) (0.143) (0.183) Phone survey mode * Head's education in years 0.015** 0.011 (0.007) (0.014) Phone survey mode * Household size −0.041* −0.027 (0.022) (0.027) Household level controls? Yes Yes Yes Yes Sub-city fixed effects? Yes Yes Yes Yes Observations 795 795 795 795 R2 0.595 0.595 0.227 0.227 Note: Ordinary least squares regression. Unit of observation is household. Household level controls include household size (number of members), indicator variable for male-headed households, and head's education in years. Standard errors are clustered at the enumeration area level and reported in parentheses. Statistical significance denoted with * p < 0.10, p < 0.05, p < 0.01. 4.4 Enumerator effects The survey team of 21 enumerators were all trained together and supervised by the same survey coordinator. To simplify survey logistics, the enumerators were tasked with conducting either phone interviews or in-person interviews. This collinearity between enumerator assignment and survey mode raises a concern that the estimated survey mode effects could be completely driven by enumerator effects.24 To address this concern, we conduct three robustness checks. First, we show that our main findings are robust to controlling for enumerator characteristics: age, level of education, and past survey experience (see Column 2 in Table A5 in the Appendix). Second, to explore whether one poorly performing enumerator in the phone survey group could explain our results, we assess the sensitivity of our result to omitting one enumerator at a time from the sample. Results are remarkably robust to running the main regression across these 21 sub-samples (see Figure A4 in the Appendix). Third, we show that our results are robust to the controlling for enumerator random effects (Table A5, column 3 in the Appendix) as well as Mundlak (1978) correlated random effects (Table A5, column 4 in the Appendix).25 Though we cannot use enumerator fixed effects, the combination of this evidence suggests that we can conclude enumerator effects could not have had much influence on the difference between in-person and phone survey results. 4.5 Cost considerations Compared to in-person surveys, phone surveys are typically considerably less costly to administer (Gourlay et al., 2021). In this case, the cost per interview was approximately 3 times lower for phone surveys than in-person surveys. The cost differences are mainly due to survey logistical costs (which are marginal for the phone survey but represent about a third of the total cost of the in-person survey) and survey personnel costs due to differences in the number of interviews per day. While there was not much difference in the time phone and in-person surveys took, phone enumerators were able to conduct about three times as many interviews in a day than in-person enumerators because the survey mode allows them to make the next call as soon as they were ready, while the in-person survey requires enumerators to travel to the next household. However, there are a few ways that the in-person costs were minimized in this urban context. For instance, travel costs were relatively low, as enumerators could travel to the neighborhoods on their own, so vehicle rental was limited to supervisory vehicles. Had households been more spread out (e.g., in a rural survey), the cost difference would have been much larger. The cost difference suggests that with the same resources, using phone surveys would allow for a sample size roughly three times larger than in-person surveys, in the same type of urban setting. Increasing the sample size that much implies a sizable gain in statistical power and thus improvement in the precision of consumption and poverty estimates.26 However, as we have shown above, the phone survey mode comes with a systematic downward bias. Consequently, survey experts interested in measuring household consumption using the standard method face a trade-off between precision and accuracy when deciding between in-person and phone survey mode. In our view, the bias introduced by the phone survey mode in this context is too large to be ignored over potential gains in precision. If poverty incidence is to be measured with phone surveys, different methods of doing so consistent with current methods of poverty estimation are necessary. 5 Conclusions Pre-pandemic, development economists and practitioners were using phone surveys in only a few contexts. In research, they were used when projects required high-frequency data or in contexts that were difficult to reach (Dabalen et al., 2016; Dillon, 2012; Hoogeveen et al., 2014). Meanwhile, WFP (2017) was building up knowledge about how to use phone surveys to monitor food insecurity. As the pandemic began, phone surveys suddenly became the only option for many types of data collection, and research on living standards and food insecurity shifted rapidly to phone surveys, to understand the socioeconomic implications of the pandemic. The subsequent COVID-19 phone surveys have provided important information about the socioeconomic consequences of the pandemic in many low- and middle-income countries with limited infrastructure to provide real-time economic or employment data to inform policy decisions. However, the economic information collected at the household level has been largely restricted to subjective indicators measuring income or employment losses, offering limited information about the severity or depth of the crisis (De Weerdt, 2008; Hirvonen et al., 2021).27 Indeed, there have been only few attempts to measure household consumption to inform how the progress toward meeting the first Sustainable Development Goal of ‘No Poverty’ has been affected by the pandemic. Finally, there remains considerable uncertainty on the implications on the use of the phone survey mode on data quality, particularly in low- and middle-income country contexts where the pre-pandemic roll out of phone survey technology and testing had been relatively slow (Gourlay et al., 2021). Our research begins to address some of these important methodological knowledge gaps. To measure the extent of bias on household consumption measures in phone surveys, we conducted a survey experiment in Addis Ababa, Ethiopia, randomly assigning a balanced and representative sample either to a phone or an in-person interview mode. We find the phone survey mode leads to a statistically significant and large underestimation of household consumption. Relative to the in-person survey mode, the phone survey mode decreases the reported household per capita consumption expenditures by 23 percent, on average. Consequently, the estimated poverty rate is twice as high when the phone survey mode is used. We therefore should reinterpret results in Hirvonen et al. (2021), which used the same household sample to show that the total value of food consumption expenditures had not changed much between August–September 2019 and August 2020. The former survey was collected in-person, and the latter by phone; if we use the results here to re-interpret that paper, it seems that if anything the average value of food consumption rose by August 2020; moreover, that paper shows that the value of relatively nutritious foods might have declined; that concern is far lower given those results likely underestimate all categories of food consumption. The mechanism appears to be linked to survey fatigue that results in phone survey respondents to greatly under-estimate consumption quantities, but not whether they consumed the item during the recall period. Our heterogeneity analysis suggests the bias increases when more people eat within the household, possibly because of the increased cognitive burden in remembering larger number of consumption events. In contrast, the bias is attenuated by education, suggesting that more educated individuals can overcome issues of attention. Our study has some important limitations. First, our sample is not nationally representative and importantly does not cover rural households that are typically poorer and consume fewer food and non-food consumption items. Consumption surveys in rural areas could take less time to complete than in urban areas, making the phone survey mode more feasible.28 Another external validity concern relates to the fact that the household sample used in this study had responded to two or three food consumption surveys prior to this survey experiment (see Table A1 in the Appendix). Consequently, the household in our sample may have become more attuned to recalling consumption events than a new, randomly selected sample of households. Finally, while we hypothesize that the documented survey fatigue is driven by respondents, the design of our experiment does not allow us to distinguish whether the fatigue is driven by fatigue among respondents or fatigue among enumerators. These limitations aside, our findings suggest that while phone surveys can provide large cost savings, they cannot replace in-person surveys for standard household consumption and poverty measurement, as outlined in Deaton and Grosh (2000). However, the phone survey mode does appear to be useful for monitoring diet-based food security indicators that do not require information about the quantities consumed, as used by the WFP (2017) in their Vulnerability Analysis and Mapping surveys. Given the prevalence of cell phone ownership, figuring out how to use phone survey data to best contribute to accurate consumption and poverty measurement in low- and middle-income countries forms an important future research agenda. One option is to substantially shorten the consumption modules to accommodate the greater risk of survey fatigue in phone surveys. However, the available evidence from low- and middle-income country contexts suggest that shorter modules systematically underestimate consumption levels and thus overestimate poverty headcounts (Beegle et al., 2012; Jolliffe, 2001; Pradhan, 2009). Therefore, when adjusting the consumption module length, survey practitioners need to balance between accuracy and survey fatigue. Finding a balance in which accuracy is maximized and risk of survey fatigue minimized in phone surveys constitutes an important task for future survey methodology research.29 Another option is to rely on cross-survey imputation methods. In recent years, these methods have become popular among poverty economists to estimate poverty in contexts and periods lacking consumption survey data (e.g., Dang, et al., 2021; Douidich et al., 2016; Stifel and Christiaensen, 2007). These types of imputation methods typically begin by using a household consumption survey and by regressing household consumption expenditures on a set of household characteristics, such as household demographics, employment status, and asset and education levels. Then another survey that collected data on the same characteristics is used, as the estimated model parameters can be applied to these household characteristics to predict household consumption expenditures and poverty rates. Phone surveys could be used to (relatively inexpensively) collect data on these household characteristics, link these data to a previous household consumption expenditure survey, and estimate poverty using cross survey imputation methods. However, the validity of this approach rests on some important assumptions. First, the relationship between household consumption expenditures and its predictors should remain stable over time (Christiaensen et al., 2012). Considering relative price changes occurring as a consequence of the COVID pandemic and the conflict between Russia and Ukraine, it remains an open question about where and when this assumption would hold. Second, linking parameters estimated from in-person consumption survey to household characteristics obtained from a phone survey assumes that survey mode effects do not matter (Kilic and Sohnesen, 2019). Considering the evidence presented here and other emerging work testing survey mode effects (e.g., Garlick, et al., 2020), this assumption is clearly is a strong assumption requiring further validation. Third, one must always be cognizant that phone ownership is correlated with income, and lower income people with phones may be less likely to keep them turned on (and therefore answer calls), to preserve their batteries. Finally, it would be useful to experiment with split questionnaire designs in a phone survey setup. In this method, respondents are randomly assigned fractions of the full questionnaire and the missing data are then imputed using multiple imputation techniques (Raghunathan and Grizzle, 1995). Recent applications of a split questionnaire design with in-person surveys suggest that the approach can produce reliable consumption and poverty estimates with considerably shorter interview durations (Pape, 2021; Pape and Mistiaen, 2015). It remains an open question about whether split designs could be used to generate low bias estimates of poverty incidence with phone surveys. Credit author statement Gashaw Abate: Conceptualization, Formal Analysis, Methodology, Investigation, Writing and Editing; Alan de Brauw: Conceptualization, Methodology, Writing and Editing, Funding Acquisition; Kalle Hirvonen: Conceptualization, Formal Analysis, Methodology, Writing and Editing; Abdulazize Wolle: Investigation, Formal Analysis, Data Curation. Uncited references AuthorAnonymous, 2008; AuthorAnonymous, 2010. Appendix Fig. A1 Distribution of (ln) weekly food consumption per capita (in birr) in September-2019, by survey mode in August-2021. Note: N = 795 households. Fig. A1 Table A1 Surveys administered to the household sample used in this study Table A1Survey Date N Relevant questionnaire features Baseline survey (in-person) September 2019 900 Food consumption module + video screening Endline survey (in-person) February 2020 900 Food consumption module + bounded recall experiment Phone surveys May, June, and July 2020 600 Food security modules Phone survey August 2020 600 Food consumption module Phone & in-person survey August 2021 800 Food and non-food consumption modules Table A2 Order of the food groups in the two versions of the ‘food consumed at home’ module Table A2Food group Order in version 1 Order in version 2 Fruits 1 6 Vegetables 2 7 Cereals 3 1 Pulses 4 2 Meat and fish 5 3 Eggs and dairy 6 4 Oils and butter 7 5 Spices and beverages 8 8 Note: Both phone and in-person surveys included two types of food consumption modules with different order in which the food groups appeared in the questionnaire. This table shows the order of food groups in both questionnaire types. Fig. A2 Impact of phone survey mode on household consumption of different food groups. Note: Based on ordinary least squares regression. Unit of observation is household; N = 795. All regressions included household level controls (household size, indicator variable for male-headed households, and head's education in years) and sub-city fixed effects. Dots quantify the difference in household per capita consumption-expenditure (in birr) when the phone survey method is used relative to when the in-person method is used. The difference is measured as a percent of the mean household per capita consumption-expenditure value reported in the in-person group. Capped bars are 95-% confidence intervals, calculated from standard errors clustered at the enumeration area level. Fig. A2 Table A3 Replicating Table 7, but using binary consumption variable as the dependent variable Table A3 (1) (2) Dependent Variable: Household consumed the food item (0/1) Item appeared later in the questionnaire −0.007** −0.004 (0.003) (0.004) Phone survey mode −0.008* −0.004 (0.005) (0.005) Item appeared later in the questionnaire * Phone survey mode −0.008 (0.005) Household level controls? Yes Yes Sub-city fixed effects? Yes Yes Food item fixed effects? Yes Yes Observations 93,810 93,810 In-person group mean of the dependent variable 0.211 0.211 Note: Ordinary least squares regression. Unit of observation is food item consumed (or not) in each household. Number of food items is 118 and number of households is 795 resulting in 93,810 observations. Dependent variable obtains a value 1 if the household reported to have consumed the item in the past week, zero otherwise. 0/1 = binary variable. Standard errors are clustered at the food item level and reported in parentheses. Statistical significance denoted with * p < 0.10, p < 0.05, p < 0.01. Fig. A3 Predicted and observed first-digit distributions, by survey mode. Note: N = 10,526 for ‘In-person group’ and N = 9042 for the ‘Phone group’. Fig. A3 Table A4 Testing differences in Euclidean distance to the distribution predicted by Benford's law Table A4 (1) (2) Phone survey mode 0.328* 0.328** (0.156) (0.156) Consumption unit fixed effects? No Yes (N = 33) Observations: 66 66 Note: Dependent variable is Euclidean distance to the distribution predicted by Benford's law. Unit of observation is unit in which the quantity consumed was reported in (one for each group). Coefficients measure Z-scores. Standard errors clustered at food item level and they are reported in parentheses. Statistical significance denoted with * p < 0.10, p < 0.05, p < 0.01. Table A5 Robustness to controlling for enumerator characteristics Table A5 (1) (2) (3) (4) Dependent Variable: (ln) Household per capita consumption Phone survey mode −0.262 −0.269* −0.263* −0.263* (0.054) (0.054) (0.059) (0.062) Household level controls? Yes Yes Yes Yes Sub-city fixed effects? Yes Yes Yes Yes Enumerator characteristics? No Yes No No Enumerator random effects? No No Yes Yes Enumerator means of household level controls? No No No Yes Observations 795 795 795 795 R2 0.288 0.290 n/a n/a R2 within n/a n/a 0.224 0.224 R2 between n/a n/a 0.600 0.652 R2 overall n/a n/a 0.286 0.290 Note: Ordinary least squares regression. Unit of observation is household. Dependent variable is (ln) household total per capita consumption (in birr). Household level controls include household size (number of members), indicator variable for male-headed households, and household head's education in years. Enumerator characteristics include enumerator's age, level of education, and survey experience (number of surveys involved in since September 2019). Standard errors are clustered at the enumeration area level and reported in parentheses. Statistical significance denoted with * p < 0.10, p < 0.05, p < 0.01. Fig. A4 Robustness of leaving one enumerator out of the dataset at a time. Note: The blue solid dot represents the benchmark OLS estimate for the full sample reported in column 2 of Table 3. The maroon hollow dots are equivalent OLS estimates for 21 different sub-samples when one enumerator is dropped from the dataset. The capped vertical lines represent the corresponding 95% confidence intervals. Fig. A4 Appendix A Supplementary data The following is the Supplementary data to this article:Multimedia component 1 Multimedia component 1 Data availability Part of the data are publicly available; the remainder are in process of being made available. Code will be made available once review process is complete. Acknowledgements We thank Kibrom Abay, Joachim De Weerdt, two anonymous reviewers, and conference participants in the Methods and Measurement Conference organized by the IPA-Northwestern Research Methods Initiative and seminar participants at the Helsinki GSE for useful comments. We are grateful to the households that participated in this study. We acknowledge NEED NUTRITIONAL and their survey staff (Abraha Weldegerima, Alemayehu Deme, Nadiya Kemal, Fikirte Sinkineh, Meskerem Abera, Rediet Dadi, Fitsum Aregawi, Asnakech Yosef, Ekiram Shimelis, Biruktawit Abebe, Yared Tilahun, Tesfaye Eana, Yosef Regasa, Ashenafi Hailemariam, Selamawit Genene, Shambel Asefa, Teshale Hirpesa, Getachew Buko, Habtamu Ayele, Ayinalem Reta, Mohamed Meka, Kibrom Tadesse, and Huluhager Endashaw) for collecting the data. Any remaining errors are the sole responsibility of the authors. This work was undertaken as part of, and funded by, the 10.13039/501100015815 CGIAR Research Program on Agriculture for Nutrition and Health (A4NH). The opinions expressed here belong to the authors, and do not necessarily reflect those of A4NH or CGIAR. This study is registered in the AEA RCT Registry with a unique identifying number AEARCTR-0008097. 1 We ainly cover the relevant literature in low- and middle-income countries. Over the past 40 years, phone surveys have become the most frequently used data collection method in high income countries. For a review of the key methodological issues in this context, see Chapter 10 in Tourangeau et al. (2000). 2 Based on the most recent data for each country reported in the World Bank's PovcalNet database, more than 90 percent of the poverty statistics in low and lower-middle income countries originate from household consumption surveys. 3 Although these guidelines were developed more than 20 years ago, they remain relevant and are still widely used to monitor global poverty (see Mancini et al., 2021). 4 However, they do find a concurrent rise in some measures of food insecurity. 5 This finding is in line with growing literature documenting non-classical measurement error in household surveys conducted in low- and middle-income countries (e.g., Abay, et al., 2019; Abay et al., 2021b; Carletto et al., 2013; Desiere and Jolliffe, 2018; Gibson et al., 2015; Gibson and Kim, 2010; Gourlay et al., 2019). 6 The endline survey also included a survey experiment to quantify the degree of telescoping bias in recalled food consumption by experimentally varying the recall method, see Abate et al. (2020) for more details. 7 Melesse et al. (2019) provide a detailed description of the sampling strategy. 8 The exact dates were 31 August to 9 September 2021. 9 To ensure balance between the two groups, we block-randomized using the following variables: sex, age and education of the household head, household size, and an asset index. The data for these variables were collected in the previous in-person visits. 10 Out of the 70 households in the phone survey group that were not interviewed, 16 did not answer the call, 37 had their phone switched off or not working, 10 had wrong numbers, and 5 had no phone numbers. Only 2 households refused to take part in the phone survey. 11 At the end of each phone interview, we asked enumerators to rate the quality of the connection during the call. 74 percent of the phone interviews were rated as ‘very good’ (“we heard each other very well”), 19 percent as ‘good’, 5 percent as ‘OK/average’ and only 2 percent (5 interviews) as ‘bad’ or ‘very bad’. 12 The FCS food groups are: main staples (weight: 2); pulses (3); vegetables (1); fruits (1); meat, eggs, fish (4); dairy products (4); sugar (0.5); oil/butter (0.5); and condiments (0). 13 Recall that we use the equation reported at the end of Section 2 to interpret the coefficients in semi-log regressions. As a result, the numbers reported in the text will differ slightly from the commonly used interpretation of 100 βˆ, where βˆ is the coefficient estimate reported in the regression tables. 14 The difference is statistically significant (p = 0.003). 15 Considering that the mean value in the in-person group is 53.91 birr, the difference of 21.66 birr estimated with OLS translates to 40.2 percent (21.66/53.91). 16 Evidence from survey experiments conducted in high-income countries have documented respondent fatigue in phone survey mode (e.g., Eckman, et al., 2014), Roberts, et al. (2010). 17 Garlick et al. (2020) randomly assigned small firms to weekly phone and in-person surveys finding that phone survey respondents systematically under-reported labor supply, stock, and inventory relative to in-person respondents. However, the authors did not explicitly test whether these differences could be driven by survey fatigue. 18 Laajaj and Macours (2021), Ambler, et al. (2021), and Abay et al. (2021a) also randomize the order in which questions are asked in their surveys to study survey fatigue. 19 As can be seen from Appendix Table A2, we administered two different versions of the food consumption module by simply changing the ordering of the food groups. As a result, we do not have sufficient variation in our data to test this with a ‘distance variable’ that captures the number of items between the version 1 and version 2. 20 The calculations in this paragraph are as follows: 5.8 percent lower is calculated as −0.230/3.97 and 15.5 percent lower is calculated as −0.615/3.97, using the estimates reported in Table 7, column 1, and 11.9 percent lower is calculated as [-0.014+(-0.458)]/3.97, using the estimates reported in Table 7, column 2. 21 The Euclidian distance is calculated as the square root of the sum of squared differences between the observed percentage and the percentage predicted by the Benford's law. We further normalize the calculated distances by taking a Z-score: subtracting the mean distance and dividing this by the standard deviation calculated using the pooled data. 22 We calculated these distributions using a user-written Stata routine devised by Jann (2007). 23 Table 2 reports that the difference in household size between the two household groups is not statistically different from zero. 24 Previous work in this area has found that the enumerator effects play a negligible role in shaping survey responses, unless the questions are sensitive in nature (Di Maio and Fiala, 2020). 25 The random effects estimator controls for enumerator heterogeneity by decomposing the unobserved heterogeneity to variance occurring between enumerators and within enumerators (i.e., across different interviews conducted by the same enumerator). The key assumption of the random effect estimator is that the correlation between the treatment status and the random effects is zero, or in the correlated random effects model, that it takes on a specific parameter. We acknowledge that, in our application, this assumption may not hold. However, simulation studies suggest that the ‘heterogeneity bias’ stemming from the violation of this assumption is relatively small (see Bell and Jones, 2015). Considering this point and the fact that the estimated coefficient based on the random effects estimator is very close to the coefficient reported in column 2 of Table 3, we believe that unobserved enumerator effects are not driving our results. 26 There is another channel through which phone surveys can be more efficient than in-person surveys. In-person surveys typically require cluster sampling to simplify logistics and reduce potentially sizable transportation costs (particularly in rural areas). As the same logistical concerns are absent in phone surveys, they permit applying a simple random sampling through random direct dial techniques that is more efficient than cluster sampling. 27 At the same time, with imperfect and non-random mobile phone access in rural areas, the data may not be representative as the poor and people in more remote areas may have less access to phones or be outside of coverage areas when phone surveys are fielded (Ambel et al., 2021; Brubaker et al., 2021). 28 However, a limited and unequal access to phones can be a major obstacle to administering representative phone surveys in rural areas. For example, in Ethiopia, only 40 percent of rural households have access to a phone, and those that have, tend to be more educated and wealthier (Wieser et al., 2020). Furthermore, rural households tend to be larger than urban households, potentially exacerbating bias related to household size. 29 It is important to note, however, that such major adjustments to survey design compromise the comparability to earlier consumption and poverty statistics that were based on different methodologies. Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jdeveco.2022.103026. ==== Refs References Abate G.T. de Brauw A. Gibson J. Hirvonen K. Wolle A. Telescoping causes overstatement in recalled food consumption: evidence from a survey experiment in Ethiopia IFPRI Discussion Paper 01976 2020 International Food Policy Research Institute Washington D.C. (IFPRI) Abate G.T. Baye K. de Brauw A. Hirvonen K. Wolle A. Video-based behavioral change communication to change consumption patterns IFPRI Discussion Paper 02052 2021 International Food Policy Research Institute Washington D.C. (IFPRI) Abay K.A. Abate G.T. Barrett C.B. Bernard T. Correlated non-classical measurement errors,‘Second best’policy inference, and the inverse size-productivity relationship in agriculture J. Dev. Econ. 139 2019 171 184 Abay K.A. Berhane G. Hoddinott J.F. Tafere K. Washington D.C 2017 Assessing Response Fatigue in Phone Surveys: Experimental Evidence on Dietary Diversity in Ethiopia 2021 International Food Policy Research Institute (IFPRI) Abay K.A. Bevis L.E. Barrett C.B. Measurement error mechanisms matter: agricultural intensification with farmer misperceptions and misreporting Am. J. Agric. Econ. 103 2021 498 522 Ambel A. McGee K. Tsegay A. Reducing bias in phone survey samples: effectiveness of reweighting techniques using face-to-face surveys as frames in four african countries Policy Research Working Paper 9676 2021 The World Bank Washington D.C. Ambler K. Herskowitz S. Maredia M.K. Are we done yet? Response fatigue and rural livelihoods J. Dev. Econ. 153 2021 102736 Ameye H. De Weerdt J. Gibson J. Measuring macro-and micronutrient intake in multi-purpose surveys: evidence from a survey experiment in Tanzania Food Pol. 102 2021 102042 -- Food consumption analysis: calculation and use of the food consumption score in food security analysis World Food Programme (WFP) 2008 Vulnerability Analysis and Mapping Branch (ODAV) Rome -- Non‒classical measurement error in long‒term retrospective recall surveys Oxf. Bull. Econ. Stat. 72 2010 687 695 Backiny-Yetna P. Steele D. Djima I.Y. The impact of household food consumption data collection methods on poverty and inequality measures in Niger Food Pol. 72 2017 7 19 Baird S. Hamory J. Miguel E. Tracking, Attrition and Data Quality in the Kenyan Life Panel Survey Round 1 (KLPS-1) 2008 Center for International and Development Economics Research Berkeley, CA, UC Berkeley Beaman L. Dillon A. Do household definitions matter in survey design? Results from a randomized survey experiment in Mali J. Dev. Econ. 98 2012 124 135 Beegle K. De Weerdt J. Friedman J. Gibson J. Methods of household consumption measurement through surveys: experimental results from Tanzania J. Dev. Econ. 98 2012 3 18 Bell A. Jones K. Explaining fixed effects: random effects modeling of time-series cross-sectional and panel data Political Science Research and Methods 3 2015 133 153 Benford F. The law of anomalous numbers Proc. Am. Phil. Soc. 1938 551 572 Bound J. Brown C. Mathiowetz N. Measurement error in survey data Handbook of Econometrics 2001 Elsevier 3705 3843 Brubaker J. Kilic T. Wollburg P. Representativeness of individual-level data in COVID-19 phone surveys Policy Research Working Paper 9660 2021 The World Bank Washington D.C. Caeyers B. Chalmers N. De Weerdt J. Improving consumption measurement and other survey data through CAPI: evidence from a randomized experiment J. Dev. Econ. 98 2012 19 33 Carletto C. Savastano S. Zezza A. Fact or artifact: the impact of measurement errors on the farm size–productivity relationship J. Dev. Econ. 103 2013 254 261 Christiaensen L. Lanjouw P. Luoto J. Stifel D. Small area estimation-based prediction methods to track poverty: validation and applications J. Econ. Inequal. 10 2012 267 297 Dabalen A. Etang A. Hoogeveen J. Mushi E. Schipper Y. von Engelhardt J. Mobile Phone Panel Surveys in Developing Countries: a Practical Guide for Microdata Collection 2016 The World Bank Washington D.C. Dang H.-A.H. Kilic T. Carletto C. Abanokova K. Poverty imputation in contexts without consumption data Policy Research Working Paper 9838 2021 The World Bank Washington D.C. De Weerdt J. Field notes on administering shock modules J. Int. Dev. 20 2008 398 402 De Weerdt J. Beegle K. Friedman J. Gibson J. The challenge of measuring hunger through survey Econ. Dev. Cult. Change 64 2016 727 758 De Weerdt J. Gibson J. Beegle K. What can we learn from experimenting with survey methods? Annual Review of Resource Economics 12 2020 431 447 Deaton A. Grosh M. Consumption Grosh M. Glewwe P. Designing Household Survey Questionaires for Developing Countries: Lessons from 15 Years of Living Standards Measurement Study 2000 World Bank Washington D.C. 91 133 Deaton A. Zaidi S. Guidelines for Constructing Consumption Aggregates for Welfare Analysis 2002 World Bank Publications Desiere S. Jolliffe D. Land productivity and plot size: is measurement error driving the inverse relationship? J. Dev. Econ. 130 2018 84 98 Di Maio M. Fiala N. Be wary of those who ask: a randomized experiment on the size and determinants of the enumerator effect World Bank Econ. Rev. 34 2020 654 669 Dillon B. Using mobile phones to collect panel data in developing countries J. Int. Dev. 24 2012 518 527 Douidich M. Ezzrari A. Van der Weide R. Verme P. Estimating quarterly poverty rates using labor force surveys: a primer World Bank Econ. Rev. 30 2016 475 500 Eckman S. Kreuter F. Kirchner A. Jäckle A. Tourangeau R. Presser S. Assessing the mechanisms of misreporting to filter questions in surveys Publ. Opin. Q. 78 2014 721 733 Egger D. Miguel E. Warren S.S. Shenoy A. Collins E. Karlan D. Parkerson D. Mobarak A.M. Fink G. Udry C. Falling living standards during the COVID-19 crisis: quantitative evidence from nine developing countries Sci. Adv. 7 2021 eabe0997 FDRE Poverty and Economic Growth in Ethiopia 1995/96-2015/16." Addis Ababa, Planning and Development Commission of the Federal Democratic Republic of Ethiopia (FDRE) 2018 Fiedler J.L. Mwangi D.M. Improving household consumption and expenditure surveys' food consumption metrics: developing a strategic approach to the unfinished agenda IFPRI Discussion Paper 1570 2016 International Food Policy Research Institute Washington D.C. (IFPRI) Friedman J. Beegle K. De Weerdt J. Gibson J. Decomposing response error in food consumption measurement: implications for survey design from a randomized survey experiment in Tanzania Food Pol. 72 2017 94 111 Garlick R. Orkin K. Quinn S. Call me maybe: experimental evidence on frequency and medium effects in microenterprise surveys World Bank Econ. Rev. 34 2020 418 443 Gibson J. Kim B. Measurement error in recall surveys and the relationship between household size and food demand Am. J. Agric. Econ. 89 2007 473 489 Gibson J. Beegle K. De Weerdt J. Friedman J. What does variation in survey design reveal about the nature of measurement errors in household consumption? Oxf. Bull. Econ. Stat. 77 2015 466 474 Glazerman S. Rosenbaum M. Sandino R. Shaughnessy L. Remote Surveying in a Pandemic: Handbook 2020 Newark, DE Innovation for Poverty Action (IPA) Gourlay S. Kilic T. Lobell D.B. A new spin on an old debate: errors in farmer-reported production and their implications for inverse scale-Productivity relationship in Uganda J. Dev. Econ. 141 2019 102376 Gourlay S. Kilic T. Martuscelli A. Wollburg P. Zezza A. High-frequency phone surveys on COVID-19: good practices, open questions Food Pol. 2021 102153 Hirvonen K. Abate G.T. de Brauw A. Food and nutrition security in Addis Ababa, Ethiopia during COVID-19 pandemic: may 2020 report IFPRI-ESSP Working Paper 143. Washington D.C., Ethiopia Strategy Support Program 2020 ESSP) of the International Food Policy Research Institute (IFPRI) Hirvonen K. de Brauw A. Abate G.T. Food consumption and food security during the COVID-19 pandemic in Addis Ababa Am. J. Agric. Econ. 103 2021 772 789 33821007 Hoddinott J. Yohannes Y. Dietary diversity as a food security indicator IFPRI-FCND Discussion Paper 136 2002 International Food Policy Research Institute Washington, DC (IFPRI) Hoogeveen J. Croke K. Dabalen A. Demombynes G. Giugale M. Collecting high frequency panel data in Africa using mobile phone interviews Canadian Journal of Development Studies/Revue canadienne d'études du développement 35 2014 186 207 Hughes S. Velyvis K. Tips to quickly switch from face-to-face to home-based telephone interviewing Mathematica 2020 Accessed Jann B. DIGDIS: Stata module to analyze the distribution of digits Statistical Software Components S456853 2007 Boston College Department of Economics revised 18 Jan 2021 Janssens W. Pradhan M. de Groot R. Sidze E. Donfouet H.P.P. Abajobir A. The short-term economic effects of COVID-19 on low-income households in rural Kenya: an analysis using weekly financial household data World Dev. 138 2020 105280 Jolliffe D. Measuring absolute and relative poverty: the sensitivity of estimated household consumption to survey design J. Econ. Soc. Meas. 27 2001 1 23 Jones L. von Engelhardt J. Insights into the utility and management of mobile phone panel surveys: evidence from surveys of household resilience in Myanmar Available at: https://ssrn.com/abstract=3732529 2020 Accessed Josephson A. Kilic T. Michler J.D. Socioeconomic impacts of COVID-19 in low-income countries Nat. Human Behav. 5 2021 557 565 33785897 Kaiser M. Benford's law as an indicator of survey reliability—can we trust our data? J. Econ. Surv. 33 2019 1602 1618 Kilic T. Sohnesen T.P. Same question but different answer: experimental evidence on questionnaire design's impact on poverty measured by proxies Rev. Income Wealth 65 2019 144 165 Kopper S. Sautmann A. Best practices for conducting phone surveys Abdul Latif Jameel Poverty Action Lab 2020 J-PAL) Accessed: Laajaj R. Macours K. Measuring skills in developing countries J. Hum. Resour. 56 2021 1254 1295 Laborde D. Martin W. Vos R. Impacts of COVID‒19 on global poverty, food security, and diets: insights from global model scenario analysis Agric. Econ. 2021 Lakner C. Yonzan N. Mahler D.G. Aguilar R.A.C. Wu H. Updated Estimates of the Impact of COVID-19 on Global Poverty: Looking Back at 2020 and the Outlook for 2021 2021 The World Bank Accessed Mancini G. Vecchi G. On the Construction of a Consumption Aggregate for Inequality and Poverty Analysis 2021 International Bank for Reconstruction and Development and The World Bank Washington D.C. McKenzie D. Rosenzweig M. Preface for symposium on measurement and survey design J. Dev. Econ. 1 2012 1 2 Melesse M.B. van den Berg M. de Brauw A. Abate G.T. Understanding urban consumers' food choice behavior in Ethiopia: promoting demand for healthy foods IFPRI-ESSP Working Paper 131 2019 International Food Policy Research Institute Washington D.C. (IFPRI) Miguel E. Mobarak A.M. The economics of the COVID-19 pandemic in poor countries NBER Working Paper 29339 2021 National Bureau of Economic Research (NBER) Cambridge, MA Mundlak Y. On the pooling of time series and cross section data Econometrica: J. Econom. Soc. 1978 69 85 Pape U. Measuring poverty rapidly using within-survey imputations The World Bank Policy Research Working Paper 9530 2021 The World Bank Washington D.C. Pape U. Mistiaen J. Measuring household consumption and poverty in 60 minutes: the Mogadishu high frequency survey The World Bank Policy Research Working Paper 8430 2015 The World Bank Washington D.C. Pradhan M. Welfare analysis with a proxy consumption measure: evidence from a repeated experiment in Indonesia Fisc. Stud. 30 2009 391 417 Raghunathan T.E. Grizzle J.E. A split questionnaire survey design J. Am. Stat. Assoc. 90 1995 54 63 Roberts C. Eva G. Allum N. Lynn P. Data quality in telephone surveys and the effect of questionnaire length: a cross-national experiment ISER Working Paper Series No. 2010-36 2010 Institute for Social and Economic Research, University of Essex UK Sánchez-Páramo C. Hill R. Mahler D.G. Narayan A. Yonzan N. COVID-19 Leaves a Legacy of Rising Poverty and Widening Inequality 2021 The World Bank Accessed Schündeln M. Multiple visits and data quality in household surveys Oxf. Bull. Econ. Stat. 80 2018 380 405 Stifel D. Christiaensen L. Tracking poverty over time in the absence of comparable consumption data World Bank Econ. Rev. 21 2007 317 341 Sumner A. Hoy C. Ortiz-Juarez E. Estimates of the impact of COVID-19 on global poverty WIDER Working Paper 2020/43. Helsinki, UNU-WIDER 2020 Swindale A. Bilinsky P. Household dietary diversity score (HDDS) for measurement of household food access: indicator guide Food and Nutrition Technical Assistance Project, Academy for Educational Development vol. 360 2006 FANTA FHI Washington, DC Washington D.C. Tourangeau R. Rips L.J. Rasinski K. The Psychology of Survey Response 2000 Cambridge University Press New York Troubat N. Grünberger K. Impact of survey design in the estimation of habitual food consumption: a study based on urban households of Mongolia Food Pol. 72 2017 132 145 van Garderen J.K. Shah C. Exact interpretation of dummy variables in semilogarithmic equations Econom. J. 5 2002 149 159 WFP Conducting mobile surveys responsibly - a field book for WFP staff World Food Programme (WFP) 2017 Vulnerability Analysis & Mapping (VAM) Rome Wieser C. Ambel A.A. Bundervoet T. Tsegay A.H. Monitoring COVID-19 Impacts on Households in Ethiopia: Results from a High-Frequency Phone Survey of Households 2020 The World Bank Report #1 Zezza A. Carletto C. Fiedler J.L. Gennari P. Jolliffe D. Food counts. Measuring food consumption and expenditures in household consumption and expenditure surveys (HCES). Introduction to the special issue Food Pol. 72 2017 1 6	36471688	PMC9711906	NO-CC CODE	2022-12-13 23:16:26	no	J Dev Econ. 2023 Mar 1; 161:103026	utf-8	J Dev Econ	2,022	10.1016/j.jdeveco.2022.103026	oa_other
==== Front Prog Cardiovasc Dis Prog Cardiovasc Dis Progress in Cardiovascular Diseases 0033-0620 1873-1740 The Authors. Published by Elsevier Inc. S0033-0620(22)00135-9 10.1016/j.pcad.2022.11.014 Article COVID-19 and elite sport: Cardiovascular implications and return-to-play Faghy Mark A. abc⁎ Ashton Ruth E.M. ac Parizher Gary d Smith Andy c Arena Ross abc Gough Lewis A. e Emery Michael S. d a School of Human Sciences, University of Derby, Derby, UK b Department of Physical Therapy, College of Applied Sciences, University of Illinois at Chicago, Chicago, IL, USA c Healthy Living for Pandemic Event Protection (HL – PIVOT) Network, Chicago, IL, USA d Sports Cardiology Center, Department of Cardiovascular Medicine, Heart Vascular and Thoracic Institute, Cleveland Clinic, Cleveland, OH, USA e Human Performance and Health Group, Centre for Life and Sport Sciences (CLaSS), School of Health Sciences, Birmingham City University, Birmingham, UK ⁎ Corresponding author at: Biomedical Research Theme, School of Human Sciences, University of Derby, UK. 1 12 2022 1 12 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. Curtailing elite sports during the coronavirus disease 2019 (COVID-19) pandemic was necessary to prevent widespread viral transmission. Now that elite sport and international competitions have been largely restored, there is still a need to devise appropriate screening and management pathways for athletes with a history of, or current, COVID-19 infection. These approaches should support the decision-making process of coaches, sports medicine practitioners and the athlete about the suitability to return to training and competition activities. In the absence of longitudinal data sets from athlete populations, the incidence of developing prolonged and debilitating symptoms (i.e., Long COVID) that affects a return to training and competition remains a challenge to sports and exercise scientists, sports medicine practitioners and clinical groups. As the world attempts to adjust toward ‘living with COVID-19’ the very nature of elite and international sporting competition poses a risk to athlete welfare that must be screened for and managed with bespoke protocols that consider the cardiovascular implications for performance. Keywords COVID-19 Cardiology Sports competition Long COVID Athletes Abbreviations ACC, American College of Cardiology AHA, American Heart Association CMRI, Cardiac magnetic resonance imaging CPET, Cardiopulmonary exercise testing CTA, Computed tomography angiography COVID-19, Coronavirus disease 2019 ECG, Electrocardiogram FIFA, Federation Internationale de Football Association GRTP, Graduated return to play HR, Heart rate IOC, International Olympic Committee LGE, Late gadolinium enhancement NGOs, Nongovernmental organizations PET, Positron emission tomography RPE, Rating of perceived exertion RTP, return-to-play RV, Right Ventricle SARS-CoV-2, severe acute respiratory syndrome coronavirus 2 WHO, World Health Organization ==== Body pmcThroughout the coronavirus disease 2019 (COVID-19) pandemic, elite sports and prestigious competitions were either suspended, pos tponed, or cancelled to reduce the risk of viral transmission and mitigate the threat to public health. Indeed, the postponement of the 2020 Tokyo Olympics and Paralympics was the only instance of its kind since its inception in 1896, which had only been previously impacted by World War I and World War II. When the games were able to go ahead, 12 months later, they took place in the absence of spectators. A global spectacle such as the Olympics which is founded on core values of friendship, respect and excellence also draws the attention of the world's richest international corporations and is broadcasted by powerful media corporations.1 It is estimated that ticket sales from spectators only contribute 5% of total competition revenue, which explains why the games proceeded as planned in 2021, without spectators and despite strong opposition from national and international bodies. Another instance where powerful corporate and commercial entities appear to have influenced decisions that are driven by financial implications rather than public health occurred during the height of the pandemic in the United Kingdom. Football competitions and leagues were curtailed and most leagues in Europe were cancelled in March 2020, but the pressure to complete the Premier League season from the media and commercial partners was at the heart of key decisions to restore and complete unfinished competitions.2 Football and specifically the English Premier League is a global market and is arguably the highest-earning and most commercialized football league in Europe, if not the world3 In football, media and commercial investment has forged a way for external bodies to be an integral part of the football ecosystem4 and associated revenue represent a large proportion of income streams which can be as high as 59%.5 This pressure led to drastic action being taken from governing bodies and clubs to develop protocols that allowed players to return to training and complete all outstanding fixtures. Strict protocols for training and competition were developed by governing bodies including regular COVID-19 testing of players and coaching and regular cleaning and quarantining of facilities and equipment6 to allow completion of the 2019–20 season, this period and the subsequent 2020–21 season was completed without spectators. Despite the measures taken to complete international leagues, fixtures and competitions, the football world governing body, Federation Internationale de Football Association (FIFA), estimates that $14 billion of revenue and income were lost in that period. Whilst football has a monumental commercial foundation, it is also a powerful vehicle in communities globally and plays a key citizenship role in community settings,7 which was likely important at a time when morale and well-being were adversely affected by imposed restrictions and lockdowns that limited social interaction.8 During a time of international crisis where global health was in the most precarious state in modern history where the health inequality gap was becoming increasingly worse, the pressure and sporadic approaches that were taken to reinstate elite sporting competitions prematurely might have been considered a risk to athlete welfare. Even as the pandemic progressed and global steps to restore social and economic activities, including the re-introduction of elite sports schedules and international competitions, the risk of infection due to sustained transmission remained a very real threat. As is the case with all acute illnesses, prevention is the preferred solution. However, with the removal of all social distancing restrictions, free testing and mandatory wearing of personal protective equipment and a rise in variants of concern, infection rates have and will continue to increase.9 The risk associated with long-term disability and cardiovascular sequelae following infection with COVID-19 represents a real challenge to practitioners and sports medicine professionals to determine when it is appropriate for athletes to return to training and competition activities.10 Whilst the knowledge about acute and chronic implications is still developing, there is a need to devise appropriate assessment and management strategies that prioritize athlete health, well-being, and welfare. Sports cardiology and COVID-19 The significance of myocarditis in athletes Observations of cardiac injury provoked by COVID-19 in the general population prompted healthcare providers to turn their attention toward the effect of the disease on the hearts of competitive athletes.10 The chief concern was the dangerous prospect of myocarditis. A robust line of evidence links exercise-induced sudden cardiac death to myocarditis in otherwise healthy young individuals.11 Studies in United States Military recruits yielded an association between myocarditis and rare incidents of exercise-induced sudden cardiac death.12 , 13 Analysis of young competitive athletes corroborated this association, mostly in males.14 As a result, the American College of Cardiology (ACC) and American Heart Association (AHA) recommend a thorough evaluation of competitive athletes presenting with myocarditis, with restrictions on exertion in individuals showing evidence of active myocardial inflammation.15 Strategies for screening and risk stratification in the general population of adolescent and young adult athletes require an accurate assessment of the incidence of the disease, which in turn requires adherence to a workable definition of cases. The spectrum of presentation of myocarditis in the general population can vary from a mild syndrome of dyspnea and/or chest discomfort to a rare but fulminant life-threatening emergency characterized by malignant arrhythmias and cardiogenic shock.16 Most cases are mild and self-limited without lasting sequelae.17 The gold standard for diagnosis of acute myocarditis is histopathological, with endomyocardial biopsy demonstrating lymphocytic infiltrate.18 However, an endomyocardial biopsy is rarely performed in practice outside of life-threatening presentations because of sampling error in patchy disease, procedural complication risks, and availability of safer testing strategies which offer excellent specificity and sensitivity.19 Electrocardiography, echocardiography, and serum biomarkers of inflammation and cardiac injury are vital components for the diagnosis but can also be normal in some cases. Cardiac magnetic resonance imaging (CMRI) with gadolinium contrast has become central to the diagnosis of myocarditis with the use of the Lake Louise Criteria, which in 2018 underwent an update to include T1 and T2 mapping to the framework.20 , 21 In addition to its ability to show focal wall motion abnormalities, CMRI's capacity for tissue characterization enables visualization of intramyocardial oedema, a hallmark of cellular injury. A non-coronary anatomic distribution of edematous segments, early gadolinium enhancement suggesting capillary leak and hyperemia, and late mid-myocardial or sub-epicardial gadolinium enhancement (LGE) characteristic of necrosis and/or scar can all be evidence of myocarditis. In a study of 40 patients with myocarditis compared to 26 controls, the updated 2018 Lake Louise Criteria yielded a sensitivity and specificity of 87.5% and 96.2%, respectively; notably this study was conducted in a population with a high pre-test probability for myocarditis.22 However, while CMR findings can demonstrate myocardial inflammation, in isolation they are insufficient to secure a diagnosis of myocarditis. The European Society of Cardiology proposed diagnostic criteria for myocarditis in 2013.23 A typical clinical presentation, in combination with at least one of four corroborative abnormal testing features (electrocardiography, echocardiography, serum biomarkers, and CMR), in the absence of an alternative explanation for the syndrome, are necessary for a diagnosis of myocarditis. Diagnosing an asymptomatic patient with myocarditis requires that at least two of these testing features be present. COVID-19 and myocarditis: initial observations A complex narrative intertwines COVID-19 with the cardiovascular care of competitive athletes. Initial observations in patients hospitalized with COVID-19 in Wuhan, China suggested a 12.5–20% incidence of biomarker-evident cardiac injury in that population.24 , 25 Given the possibility of nonspecific troponin elevation in patients hospitalized with severe viral infections, a number of investigators further explored the possibility of COVID-19 myocarditis with CMRI. In one such study including 100 German patients recovering from COVID-19, 78 had abnormal CMRI findings.26 Two patients in this cohort with high-risk findings were referred for endomyocardial biopsy, which revealed lymphocytic infiltration with no viral genome. However, in the absence of a correlation between imaging findings and symptoms, it was not possible to determine the incidence of true myocarditis from this study. An autopsy study demonstrated active severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) replication within the myocardium in 24 of 39 consecutive cases, suggesting a high incidence of COVID-19 related myocarditis.27 However, subsequent pathologic studies have since called into question the incidence of myocarditis associated with COVID-19, showing that viral replication within cardiac myocytes is rare.28 Data from patients who died of fulminant infection also may not be generalizable. A later epidemiologic study using more stringent criteria for COVID-19-associated myocarditis in hospitalized patients suggests that it is rarer than initially suspected, on the order of 1 case per 1000 hospitalizations.29 The incidence of myocarditis among all comers infected with SARS-CoV-2, including those not requiring hospitalization, has not been determined. CMRI and COVID-19 in athletes Findings from CMRI in patients recovering from hospitalization prompted healthcare providers involved in the care of competitive athletes to direct their attention to the possibility of an increased risk of exercise-related sudden cardiac arrest driven by COVID-19 myocarditis. Early in the pandemic, data informing a risk stratification strategy for return-to-play (RTP) in athletes convalescing from COVID-19 were lacking. One key question was the incidence of post-COVID-19 myocarditis in athletes. Between June and August 2020, Rajpal et al. performed comprehensive CMRI on 26 competitive collegiate athletes who presented to the Ohio State University sports medicine clinic after testing positive for COVID-19.30 Twelve athletes had LGE; four had findings suggestive of myocarditis, two of which had mild symptoms of dyspnea; however, it is difficult to discern whether the dyspnea was attributable to myocarditis or COVID-19. No athlete had abnormal serum cardiac biomarkers and follow-up for clinical events was unavailable. The high reported prevalence of cardiac involvement in this study was concerning during the initial stages of the pandemic. However, since athletic remodeling can result in characteristic changes in CMRI, including LGE at the right ventricle (RV) insertion point into the septum, it is difficult to interpret the clinical significance of the abnormalities discovered in this small cohort lacking a control group of healthy athletes.31 The lack of follow-up outcome data, as well as the overlap between symptoms of COVID-19 and symptoms of heart disease, make risk stratification more challenging. Conflicting concurrent reports also made the available data difficult to interpret; in July 2020, when a cohort of COVID-19-positive student-athletes at the University of West Virginia underwent CMRI, none reportedly showed features of myocarditis, though almost a third showed pericardial involvement.32 Nonetheless, at the time of publication of these data, organizations at the collegiate and professional athletic levels had cancelled competition in the interest of protecting athletes from complications of infection. This brought significant public and academic attention to the important and unclear issue of viral myocarditis in the pandemic.33 With concern about potential COVID-19 myocarditis developing, investigators embarked on larger and more robust investigations. The COMPETE CMR study enrolled 59 COVID-19-positive athletes, 60 athletic controls, and 27 healthy non-athletes, all of whom underwent comprehensive evaluation including CMRI.34 Two asymptomatic athletes had CMRI findings consistent with myocardial inflammation, one of whom went on to develop dyspnea and left ventricular systolic dysfunction consistent with myocarditis. Approximately one-fifth of each group of athletes demonstrated focal LGE isolated to the inferoseptal RV insertion point, a finding that the authors emphasized should not be conflated with myocarditis. The authors also made a point to encourage the use of clinical judgment when contextualizing abnormal CMRI findings and called for longer-term follow-up studies to determine the rates of complications. Soon after the publication of the COMPLETE CMR study, the University of Wisconsin published data on 145 student-athletes recovering from COVID-19, all of whom underwent cardiac MRI.35 Only two athletes had imaging evidence of myocardial inflammation, neither of whom was symptomatic. The Big 10 COVID-19 Cardiac Registry enrolled 9255 collegiate athletes in the United States Big Ten Athletic Conference, representing the largest cohort to date at the time.30 2810 (30.4%) individuals tested positive for COVID-19, of which 1597 underwent comprehensive CMRI evaluation. Thirty-seven athletes, of whom twenty-seven were male, were diagnosed with myocarditis according to the study definition. Only nine of them reported symptoms, whereas the remainder of the cases were asymptomatic and defined as “subclinical myocarditis.” The authors estimated a prevalence of study-defined overt or subclinical myocarditis of 2.1% among athletes testing positive for COVID-19 and proposed that CMRI increases the sensitivity of screening for myocarditis in this population. However, CMRI has not been studied for screening a population with a low pre-test probability of myocarditis. If the true prevalence of myocarditis is assumed to be 1% in this population, provided a sensitivity of 87.5% and a specificity of 96.2%, the positive predictive value of a CMRI suggestive of myocardial inflammation for myocarditis is only 18%. In the absence of a compatible clinical syndrome, these abnormal test results are more likely to represent false positives than “subclinical” cases. The nine symptomatic athletes with compatible CMRI findings who were diagnosed with myocarditis represent only 0.5% of the tested cohort, a figure that is comparable to results from other studies discussed below. Indeed, the authors acknowledged the lack of clinical outcome data available to guide care in asymptomatic athletes with isolated CMRI abnormalities. Nonetheless, this study raised the question of universal CMRI screening in all athletes testing positive for COVID-19. This contrasted with guideline documents and published expert opinion available at the time, which suggested cardiovascular testing before RTP was unnecessary in asymptomatic and mildly symptomatic athletes.36 , 37 Later in 2021, data with clinical outcomes began surfacing to better inform screening recommendations. Professional North American athletic leagues implemented mandatory cardiovascular screening before RTP, and their data included crucial documentation of clinical cardiac events. 789 COVID-19-positive symptomatic and asymptomatic professional athletes were included in an analysis of RTP screening.38 All COVID-19 cases in this study, including asymptomatic individuals, underwent screening, including serum cardiac troponin levels, a resting electrocardiogram, and a resting echocardiogram. No athlete had severe symptoms, but thirty athletes (3.8%) were sent for additional testing after abnormalities were detected upon screening, and twenty-seven of them underwent CMRI. Five CMRIs showed evidence of myocarditis, and 2 showed evidence of pericarditis; these athletes were held from returning to play according to published guidelines.15 Critically, throughout competition throughout the year 2020, no cardiac events were reported in this cohort, corroborating a conservative screening strategy using CMRI as a selective downstream test appropriate for symptomatic athletes. The authors also emphasized the low prevalence of COVID-19 myocarditis in this cohort. A similarly low prevalence of myocarditis and a low clinical event rate were demonstrated in a large cohort of 19,378 collegiate athletes, 3018 of whom tested positive. 21 (0.7%) of cases demonstrated evidence of myocarditis.39 Selective use of CMRI, compared to primary screening CMRI, showed better positive predictive value in this latter study. Only one cardiac event was recorded, which was felt to be unlikely related to SARS-CoV-2 infection. Considering this low event rate, the authors proposed that asymptomatic or mildly symptomatic athletes without cardiopulmonary complaints may return to play without further cardiac testing after recovery from their initial infection. Table 1 summarizes the available case series and registries evaluating the prevalence of cardiac involvement in athletes following SARS-CoV2 infection.Table 1 Published case series and registries to date evaluating athletes recovering from infection with SARS-CoV-2. CMRI, cardiac magnetic resonance imaging. Triad testing includes an electrocardiogram (EKG), serum cardiac troponin measurement, and a transthoracic echocardiogram (TTE). Table 1Authors Study Type COVID+ Athletes Included Competition Level Findings Rajpal, et al.30 Case Series 26 Collegiate No abnormal biomarkers, EKG, or TTE. All underwent CMRI. 4 (15%) had abnormal CMRI, and two had dyspnea. Brito, et al32 Case Series 60 Collegiate 54/60 had echocardiography, serum cardiac troponin measurement, and EKG. Forty-six underwent CMRI for symptoms and/or abnormal triad testing. 27/46 (56%) had abnormal CMRI findings. 19 (40%) had pericardial late enhancement with associated pericardial effusion. No specific imaging features of myocardial inflammation were identified. Vago, et al72 Case Series 12 Professional All asymptomatic. All had normal serum cardiac troponin and CMRI compared to healthy controls. Clark, et al.34 Case Series 59 Collegiate All had normal EKG, serum cardiac troponin, and TTE. All underwent CMRI. Two COVID+ athletes (3%) had CMRI features of myocardial inflammation. One (2%) developed clinically evident myocarditis. Notable CMRI findings were documented in healthy and COVID+ athletes which were absent in non-athletes, such as RV insertion point LGE. Starekova, et al.35 Case Series 145 Collegiate All underwent CMRI. 81% had symptoms. Two had evidence of myocardial inflammation on CMRI, one was minimally symptomatic; the other had mild-moderate symptoms. Malek, et al.37,36 Case Series 26 Professional All underwent CMRI. 4 (15%) had abnormal serum troponin. CMRI revealed abnormalities in five, none met updated criteria for myocarditis. Martinez, et al38 Registry 789 Professional All underwent triad testing; thirty screened abnormal. Twenty-seven underwent CMRI; five had findings of myocardial inflammation. No deaths or cardiac events were recorded. Moulson, et al.,33 Registry 3018 Collegiate Selective triad testing followed by CMRI if clinically indicated. 21/2999 had abnormal EKG; 24/2719 had abnormal troponin; 24/2556 had abnormal TTE. 198 underwent CMRI with three showing cardiac involvement. One cardiac event was recorded, unrelated to COVID-19. Hendrickson, et al.74 Case Series 137 Collegiate All underwent triad testing. CMRI was done when clinically indicated. Five athletes had CMRI, and none were abnormal. Daniels, et al.75 Registry 2810 Collegiate All recommended CMRI, 1597 completed. 37 (2.3%) had abnormal CMRI, 9 (0.5%) of which had symptoms consistent with myocarditis. Twenty-eight were asymptomatic. Hwang, et al.76 Case Series 55 Collegiate All underwent triad testing; CMRI was used selectively. One case of pericarditis. No cases of myocarditis. No cardiac events were reported. RV, right ventricle; LGE, late gadolinium enhancement. In summary, while initial reports regarding prevalence and risk of myocarditis in athletes recovering from COVID-19 were concerning, thorough subsequent investigations of imaging and outcomes in large cohorts have been reassuring. Screening recommendations have continued to be conservative, suggesting that asymptomatic or minimally symptomatic athletes can return to play without further testing once symptoms resolve40; the latest update to RTP guidelines, published in May of 2022, suggests CMRI is reserved for athletes with symptoms highly suggestive of myocarditis as well as abnormal initial testing. Individuals with no symptoms, or mild or moderate non-cardiopulmonary symptoms, following SARS-CoV2 infection may resume training after three days of abstinence without additional testing. This can be done while continuing to self-isolate for the CDC-recommended period, which is currently five days. Those with cardiopulmonary symptoms should undergo triad testing consisting of serum cardiac troponin measurement, electrocardiogram (ECG), and echocardiography. Only those with abnormal triad testing, or persistent symptoms prompting high suspicion of myocarditis, should undergo CMRI. Synthesis of clinical, biochemical, electrocardiographic, and imaging findings is crucial for the accurate diagnosis of myocarditis in a young competitive athlete. Management of the athlete with myocarditis Once the true diagnosis of myocarditis has been established in an athlete, there is insufficient evidence to suggest the management should differ based on the underlying infectious agent. Thus, current guideline recommendations for myocarditis in athletes apply to COVID-19-associated myocarditis.15 Left ventricular systolic dysfunction and cardiac arrhythmias should be managed according to current ACC/AHA guidelines.41 Athletes with myocarditis should abstain from training while symptomatic and for at least 3–6 months following symptom resolution. They should also undergo repeat testing before RTP. Testing in this setting should include a resting echocardiogram, measurement of serum cardiac troponin and inflammatory markers, a 24-h Holter monitor, and an exercise ECG with attention to arrhythmic burden. Repeat CMRI can be considered as well. Complete resolution of all evidence of cardiac inflammation is reassuring, but some athletes may manifest persistent LGE even after other abnormalities have resolved. While the presence of LGE indicating scar may convey a heightened risk for arrhythmias, it is not clear whether the presence of isolated LGE after an episode of myocarditis should preclude an athlete from participating in competitive sport; an informed risk-benefit discussion with the athlete is important in these cases.42 Managing COVID-19 infections in athletes The management of COVID-19 infections and the return to play is a unique challenge to sports scientists and sports medicine practitioners, which resulted in the sporadic approaches across sports worldwide.43 Despite evolving knowledge and understanding to inform the development of safe and athlete-centered approaches, there remains a lack of consistency. Most recently, Rafał Majka tested positive during the 2022 Tour de France cycling event, however, continued to compete as the medical team deemed it ‘safe’ to ride due to a ‘low viral load.’ This decision was taken despite existing knowledge of acute respiratory infections and decreased exercise performance due to neural (impaired coordination and speed in the performance of motor skills, reductions in submaximal force generation), physical (cardiorespiratory capacities, and reflecting muscle protein catabolism), cognition (decreased attention and vigilance) challenges.44 , 45 Conversely and at a similar time, Matteo Berrettini was forced to withdraw from the Wimbledon tennis championships due to a positive COVID-19 test. Whilst the details of the infection and progression is not public knowledge, it appears at first glance that approaches to managing performance and wellbeing are disparate and disease specific scientific protocols and consensus is needed to manage acute infection, recovery and re-introduction to training and competition. The disparity is also in part due to nongovernmental organizations (NGOs) and event organizers who in the absence of international guidelines have been able to implement their own approaches with dealing with COVID-19 infected athletes. Guidelines have also been adapted multiple times by NGOs and governments throughout the pandemic with variable testing regimes and protocols for isolation. This is surprising given there are multiple protocols offered by experts in the field, such as the graduated return to play (GRTP) protocol,46 , 47 in respect of assisting an athlete return to play in a manner to protect athlete welfare and the support staff around them. Initial response protocols to a positive COVID-19 infection in athletes focused on myocarditis and myocardial injury (i.e., heart inflammation), due to the concern that athletes undertaking a high load of training/exercise may exacerbate the myocardial injury and precipitate malignant ventricular arrhythmias with viral myocarditis following infection.36 The authors suggest that ECG, CMRI, Computed Tomography Angiography (CTA), cardiopulmonary exercise testing (CPET), and nuclear positron emission tomography (PET) could all be part of a screening process. Whilst this has clinical merit in these approaches,48 , 49 it may be difficult to implement this across the board due to budget constraints, considering a recent article suggested the cost per athlete for similar screening procedures could be between $632 ± 651 and $1357 ± 757 per athlete.49 , 50 More importantly, Kim et al.51 suggested these procedures are not merited, due to the incidence of myocardial issues was as low as 0.6–0.7% within a large study of professional (n = 789) and collegiate (n = 3018) athletes following a COVID-19 infection.38 , 39 It is logical therefore to suggest that the handling of athletes with a COVID-19 infection should be based on a priori clinical probability based on symptoms presented and not a universal screening approach of costly measures. To determine clinical probability, it is advised that the traditional ‘above the neck’ and ‘below neck’ symptoms should be identified, and this should guide the next procedure to follow, which is also in line with the GRTP protocol, highlighted in Table 2 . Generally, athletes require around 10 days to be back to normal training based on the median duration for players to report no symptoms, although as many as 14% could still be unavailable for 28 days later and beyond.52 However, athletes with ‘above neck’ symptoms will typically recover faster than those with ‘below neck’ symptoms,53 and the GRTP reflects this, as the latter has an additional 5-day rest period to allow for longer recovery and greater monitoring (Table 2). Using the GRTP will produce consistency among athlete treatment and care, and therefore reduce the confusion currently surrounding when athletes should/should not compete with a recent COVID-19 infection. This approach also is cost-effective and can be conducted by most athletes and sport science support teams globally. A caveat to this approach, however, is if clinical symptoms of COVID-19 re-appear or become more severe. In this case, athletes should cease the GRTP protocol and seek medical advice. These ‘red flag’ symptoms, as termed by the GRTP, are typically chest pains, unusual high rating of perceived exertion (RPE) and heart rate (HR) during exercise, mental health concerns, syncope, fatigue and/or dyspnea. This is where invasive scanning and biochemical blood measures, such as those proposed by Phelen et al.,36 could be employed. These more invasive procedures may also be necessary for athletes that have unvaccinated status due to the likelihood of the illness being more severe. Finally, if athletes have any concomitant medical conditions (e.g., cardiovascular or respiratory or renal disease) then medical evaluation should be conducted before completing the GRTP.54 Table 2 The Graduated Return to Play Protocol. Redrawn from Elliot et al. (2022). Table 2 Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Stage 6 Stage 7 Activity description Minimum rest period Light activity Frequency of training increases Duration of training increases Intensity of training increases Resume normal training RETURN TO COMPETITION IN SPORT SPECIFIC TIMELINES Exercise allowed Walking, activities of daily living Walking, light jogging, stationary cycle. No resistance training Simple movement activities e.g., running drills Progression to more complex training activities Normal training activities Resume normal training progressions % Heart rate max <70% <80% <80% <80% Resume normal training progressions Duration 10 days <15 mins <30 mins <45 mins <60 mins Resume normal training progressions Objective Allow recovery time, protect cardio-respiratory system Increase heart rate Increase load gradually, manage any post viral fatigue symptoms Exercise, coordination, and skills/tactics Restore confidence and assess functional skills Resume normal training progressions Monitoring Subjective symptoms, resting HR, I-PRRS Subjective symptoms, resting HR, I-PRRS, RPE Subjective symptoms, resting HR, I-PRRS, RPE Subjective symptoms, resting HR, I-PRRS, RPE Subjective symptoms, resting HR, I-PRRS, RPE Subjective symptoms, resting HR, I-PRRS, RPE I-PRRS, Injury – Psychological Readiness to Return to Sport; RPE, Rated Perceived Exertion Scale. Note: This guidance is specific to sports with an aerobic component. The International Olympic Committee (IOC) recently published a two-part consensus statement that covers acute respiratory infections55 and non-infective acute respiratory illness. The drive behind these statements is to provide guidance to sport and exercise science/medicine practitioners working with athletes and to uphold the Medical and Scientific Commission's value of protecting athletes, whilst focusing on prevention and management and enabling the development of effective return to sport protocols following acute illness. Data from the report highlights that almost half of all athlete's medical consultations at international events such as the Olympics and Para Olympics relate to acute respiratory illness (4.2 per 1000 athlete days).55 Whilst the work of the IOC task force began before the COVID-19 pandemic, there are clear lessons that have been incorporated into the report by the authors which broadly cover acute respiratory infection, but these apply directly to COVID-19. One of these is to highlight the importance of symptom recognition and the implementation of early and precise viral pathogen identification so that athletes can be quarantined to prevent the spread of further spread of infection. This will be of particular importance when travel and ‘athlete villages’ are common practice, such as during the Olympic games. Mitigating risks COVID-19 and some variants (e.g., Omicron, B.1.1.529) are highly transmissible, and the severity of acute infection is also variable ranging from asymptomatic to a mild-severe clinical presentation. Whilst the development and administration of current vaccinations have been effective against emerging variants, it is well established that immunity is timebound and there is a need for regular boosters.57 Recent data highlights that 20 to 30% of SARS-CoV-2 infections in athletes are asymptomatic, which creates additional considerations for sports medicine practitioners as screening, testing and quarantine protocols will be needed to limit the impact within a team and sports environment. Whilst vaccines offer the greatest protection from severe outcomes with COVID-19, vaccine hesitancy in general populations has been widely reported in the general population58 but with less coverage in athletes.59 From the available information, it appears that hesitancy is caused by a lack of knowledge on the impact of COVID-19 vaccines upon sports performance and also issues relating to side effects.59 , 60 Whilst athletes maintain a right to personal choice, this may create issues regarding international travel in countries where vaccines have been declared mandatory. The most notable case here relates to Novak Djokovic, who was unable to participate in the 2022 Australian Open and was subsequently deported by the Australian Government for not adhering to laws around vaccination. Despite the development and widespread administration of efficacious COVID-19 vaccines,61 waning immunity,62 athlete hesitancy,63 and the risk of sustained transmission and emerging variants64 will undoubtedly lead to spikes in the transmission that need to be considered with the development and introduction of screening/testing procedures that prioritize athlete welfare over commercial and business needs. Counterintuitively, the relaxing of mandatory social distancing, wearing of face coverings and the removal of access to free testing in western societies means that sports science and medicine practitioners will result in governing bodies and sports organizations having to develop, implement and uphold protocols that protect athletes, coaches, and spectators against the continued risk of infection and long-term sequelae. Long-COVID There remains a paucity of data to highlight the prevalence of the longitudinal challenges faced by athletes following a COVID-19 infection and the development of a more long-term illness that prevents a return to pre-COVID-19 activities. Long COVID or post COVID syndrome, is defined by the WHO as a condition that occurs in individuals with a history of probable or confirmed SARS CoV-2 infection, usually 3 months from the onset of COVID-19 with symptoms that last for at least 2 months and cannot be explained by an alternative diagnosis.65 Presently, there is limited understanding of the prevalence, severity, or impact upon an athlete's long term health and the ability to resume to training and competition schedules.66 Data from Hull et al highlights in 147 Olympic standard athletes (25 Paralympic and 37% female) that median symptom duration was 10 days but 14% reported symptoms greater than >28 days.52 Whilst the issues reported in this study highlight lasting issues there is no detail on the full time to recovery/return to training and the timeframe is not covered by the current WHO definition of Long COVID. There is, however, a growing number of reports within the media that highlight longstanding issues that athletes are facing in their attempts to return to training and competition activities.67, 68, 69. Whilst datasets remain infrequent, it is estimated that >144 million people globally are living with multi-dimensional and episodic symptoms that broadly impact functional status, quality of life and physical and mental wellbeing.70 Individuals with a history of elevated levels of cardiorespiratory fitness will observe marginal changes in their training and performance capabilities over the long term compared with the functional capacity of those with a history of reduced health status and multiple morbidities. However, more research and surveillance from governing bodies and leading organizations are needed to quantify and understand the longitudinal issues experienced by athletes regarding training, performance and their general health and wellbeing. The acute and chronic implications of COVID-19 will inevitably play a key role in the role of practitioners at least for the near future and further research and support from leading agencies in the governance of sport should direct the development of diligent and comprehensive protocols that support athletes and practitioners to restore pre-COVID-19 status. Future of elite sport As highlighted in this article, managing the elite sport environment is confounded by a myriad of complexities that exceeds sports performance and competition. As evidenced by the COVID-19 pandemic, global issues have a direct influence of upon elite sports and consideration of current and future challenges should be met with proactive rather than reactive strategies. The need to assess and manage athletes exposed to COVID-19 and even Long COVID is likely to evolve in respect of increased mechanistic understanding and with the development of efficacious treatments. However, the risk of future variants of concern COVID-19 and future new pandemics is inevitable, posing a sustained challenge to global health.71 The very nature of the elite sports will continue to change creating new working environments for sport scientists and medics. Therefore, we must take proactive steps to review recent reactive approaches and make considered and informed strategies that manage athlete physical, mental, and emotional wellbeing when future health threats are realized. The authors accept that the magnitude and impact of COVID-19 was and remains unprecedented, but the ability to apply these skills in competition and to train to enhance them is likely to be hampered by any viral infection including COVID-19. Conclusions The move to restore sporting competition was primarily due to commercial and contractual agreements, which posed a risk to athlete and sports scientist/medicine experts health and wellbeing. Despite attempts from international and national governing bodies implementing protocols to mitigate against, the absence of knowledge to inform decision making could have resulted in sustained illness for some athletes that have developed long term complication and/or not yet achieved clinical resolution. Whilst the knowledge base to inform decision making is advancing, the risk of sustained transmission and future variants of concern pose a continued risk to athlete and practitioner welfare. Appropriate screening and management guidance and mitigation strategies must be revised regularly and be developed with interdisciplinary collaborative approaches. Declaration of Competing Interest None. ==== Refs References 1 Lee Ludvigsen J.A. Rookwood J. Parnell D. The sport mega-events of the 2020s: Governance, impacts and controversies Sport Soc 25 4 2022 705 711 10.1080/17430437.2022.2026086 2 Manoli A.E. COVID-19 and the solidification of media’s power in football Manag Sport Leisure 27 1–2 2022 73 77 10.1080/23750472.2020.1792802 3. Deloitte Annual review of football finance 2021 Deloitte UK Deloitte United Kingdom Accessed July 13, 2022 https://www2.deloitte.com/uk/en/pages/sports-business-group/articles/annual-review-of-football-finance.html 4 Chadwick S. Parnell D. Widdop P. Anagnostopoulos C. Routledge handbook of football business and management 2019 Routledge London 5 Quansah T. Frick B. Lang M. Maguire K. The importance of Club revenues for player salaries and transfer expenses—how does the coronavirus outbreak (COVID-19) impact the English premier league? Sustainability. 13 9 2021 5154 6 Guard A. Brenneman A. Bradley M. Chiampas G.T. Facilitating national football teams return to training and competition during the COVID-19 pandemic BMJ Open Sport Exerc Med 8 2 2022 e001295 10.1136/bmjsem-2021-001295 7 Smith D.C.V.L. Footballers’ citizenship during COVID-19: A case study of Premier League players’ community support Int Rev Sociol Sport 2022 10.1177/10126902211045679 Published online October 16, 2021 8 Greyling T. Rossouw S. Adhikari T. The good, the bad and the ugly of lockdowns during Covid-19 PLoS One 16 1 2021 e0245546 9 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 The Lancet 399 10332 2022 1303 1312 10.1016/S0140-6736(22)00462-7 10 Phelan D. Kim J.H. Chung E.H. A game plan for the resumption of sport and exercise after coronavirus disease 2019 (COVID-19) infection JAMA Cardiol 5 10 2020 1085 1086 10.1001/jamacardio.2020.2136 32402054 11 Baggish A. Drezner J.A. Kim J. Martinez M. Prutkin J.M. Resurgence of sport in the wake of COVID-19: cardiac considerations in competitive athletes Br J Sports Med 54 19 2020 1130 1131 32561518 12 Phillips M. Robinowitz M. Higgins J.R. Boran K.J. Reed T. Virmani R. Sudden cardiac death in air force recruits: A 20-year review Jama 256 19 1986 2696 2699 3773175 13 Karjalainen J. Heikkilä J. Incidence of three presentations of acute myocarditis in young men in military service. A 20-year experience Eur Heart J 20 15 1999 1120 1125 10413642 14 Maron B.J. Doerer J.J. Haas T.S. Tierney D.M. Mueller F.O. Sudden deaths in young competitive athletes: analysis of 1866 deaths in the United States, 1980–2006 Circulation. 119 8 2009 1085 1092 19221222 15 Maron B.J. Zipes D.P. Kovacs R.J. Eligibility and disqualification recommendations for competitive athletes with cardiovascular abnormalities: preamble, principles, and general considerations: a scientific statement from the American Heart Association and American College of Cardiology J Am Coll Cardiol 66 21 2015 2343 2349 26542655 16 Cooper L.T. Jr. Myocarditis N Engl J Med 360 15 2009 1526 1538 19357408 17 Ammirati E. Cipriani M. Moro C. Clinical presentation and outcome in a contemporary cohort of patients with acute myocarditis: Multicenter Lombardy registry Circulation. 138 11 2018 1088 1099 29764898 18 Aretz T.H. Myocarditis a histopathologic definition and classification Am J Cardiovasc Pathol 1 1986 3 14 19 Cooper L.T. Baughman K.L. Feldman A.M. 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 Endorsed by the Heart Failure Society of America and the heart failure Association of the European Society of cardiology Eur Heart J 28 24 2007 3076 3093 17959624 20 Friedrich M.G. Sechtem U. Schulz-Menger J. Cardiovascular magnetic resonance in myocarditis: a JACC white paper J Am Coll Cardiol 53 17 2009 1475 1487 19389557 21 Ferreira V.M. Schulz-Menger J. Holmvang G. Cardiovascular magnetic resonance in nonischemic myocardial inflammation: Expert recommendations J Am Coll Cardiol 72 24 2018 3158 3176 30545455 22 Luetkens J.A. Faron A. Isaak A. Comparison of original and 2018 Lake Louise criteria for diagnosis of acute myocarditis: Results of a validation cohort Radiol: Cardiothor Imaging 1 3 2019 23 Caforio A.L. Pankuweit S. Arbustini E. Current state of knowledge on aetiology, diagnosis, management, and therapy of myocarditis: A position statement of the European Society of Cardiology Working Group on myocardial and pericardial diseases Eur Heart J 34 33 2013 2636 2648 23824828 24 Shi S. Qin M. Shen B. Association of cardiac injury with mortality in hospitalized patients with COVID-19 in Wuhan, China JAMA Cardiol 5 7 2020 802 810 32211816 25 Han H. Xie L. Liu R. Analysis of heart injury laboratory parameters in 273 COVID-19 patients in one hospital in Wuhan, China J Med Virol 92 7 2020 819 823 32232979 26 Puntmann V.O. Carerj M.L. Wieters I. Outcomes of cardiovascular magnetic resonance imaging in patients recently recovered from coronavirus disease 2019 (COVID-19) JAMA Cardiol 5 11 2020 1265 1273 32730619 27 Lindner D. Fitzek A. Bräuninger H. Association of cardiac infection with SARS-CoV-2 in confirmed COVID-19 autopsy cases JAMA Cardiol 5 11 2020 1281 1285 32730555 28 Kawakami R. Sakamoto A. Kawai K. Pathological evidence for SARS-CoV-2 as a cause of myocarditis: JACC review topic of the week J Am Coll Cardiol 77 3 2021 314 325 33478655 29 Ammirati E. Lupi L. Palazzini M. Prevalence, characteristics, and outcomes of COVID-19–associated acute myocarditis Circulation. 145 15 2022 1123 1139 35404682 30 Rajpal S. Tong M.S. Borchers J. Cardiovascular magnetic resonance findings in competitive athletes recovering from COVID-19 infection JAMA Cardiol 6 1 2021 116 118 32915194 31 Domenech-Ximenos B. Sanz-de la Garza M. Prat-González S. Prevalence and pattern of cardiovascular magnetic resonance late gadolinium enhancement in highly trained endurance athletes J Cardiovasc Magn Reson 22 1 2020 1 9 31898543 32 Brito D. Meester S. Yanamala N. High prevalence of pericardial involvement in college student athletes recovering from COVID-19 Cardiovascular Imaging 14 3 2021 541 555 33223496 33 Committee B.T.C.R.S. Rink L.D. Daniels C.J. Competitive sports, the coronavirus disease 2019 pandemic, and big ten athletics Circ Cardiovasc Qual Outcomes 13 12 2020 e007608 34 Clark D.E. Parikh A. Dendy J.M. COVID-19 myocardial pathology evaluation in athletes with cardiac magnetic resonance (COMPETE CMR) Circulation. 143 6 2021 609 612 33332151 35 Starekova J. Bluemke D.A. Bradham W.S. Evaluation for myocarditis in competitive student athletes recovering from coronavirus disease 2019 with cardiac magnetic resonance imaging JAMA Cardiol 6 8 2021 945 950 33443537 36 Phelan D. Kim J.H. Elliott M.D. Screening of potential cardiac involvement in competitive athletes recovering from COVID-19: An expert consensus statement Cardiovascul Imaging 13 12 2020 2635 2652 37 Kim J.H. Levine B.D. Phelan D. Coronavirus disease 2019 and the athletic heart: Emerging perspectives on pathology, risks, and return to play JAMA Cardiol 6 2 2021 219 227 10.1001/jamacardio.2020.5890 33104154 38 Martinez M.W. Tucker A.M. Bloom O.J. Prevalence of inflammatory heart disease among professional athletes with prior COVID-19 infection who received systematic return-to-play cardiac screening JAMA Cardiol 6 7 2021 745 752 33662103 39 Moulson N. Petek B.J. Drezner J.A. SARS-CoV-2 cardiac involvement in young competitive athletes Circulation. 144 4 2021 256 266 33866822 40. Jone P.N. John A. Oster M. SARS-CoV-2 infection and Associated Cardiovascular manifestations and complications in children and young adults: A scientific statement from the American Heart Association \| Circulation Accessed August 25, 2022 https://www.ahajournals.org/doi/full/10.1161/CIR.0000000000001064 2022 41. 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines J Am Coll Cardiol 79 17 2022 e263 e421 Accessed August 25, 2022 https://www.jacc.org/doi/abs/10.1016/j.jacc.2021.12.012 35379503 42 Grun S. Schumm J. Greulich S. Long-term follow-up of biopsy-proven viral myocarditis J Am Coll Cardiol 59 18 2012 1604 1615 10.1016/j.jacc.2012.01.007 22365425 43 Lavie C.J. Bond S. Phillips S.A. Respiratory muscle performance screening for infectious disease management following COVID-19: A highly pressurized situation Am J Med 113 9 2022 1025 1032 10.1016/j.amjmed.2020.04.003 Published online 2020 44 Friman G. Wesslén L. Infections and exercise in high-performance athletes Immunol Cell Biol 78 5 2000 510 522 11050534 45 Derman E.W. Hawarden D. Schwellnus M.P. Allergic rhinoconjunctivitis in athletes-mechanisms of impaired performance and implications for management Current Aller & Clin Immunol 23 2 2010 59 62 46 Elliott N. Biswas A. Heron N. Graduated return to play after SARS-CoV-2 infection–what have we learned and why we’ve updated the guidance 2022 Published online 2022 47 Elliott N. Martin R. Heron N. Elliott J. Grimstead D. Biswas A. Infographic graduated return to play guidance following COVID-19 infection Br J Sports Med 54 19 2020 1174 1175 32571796 48 Faghy M.A. Sylvester K.P. Cooper B.G. Hull J.H. Cardiopulmonary exercise testing in the COVID-19 endemic phase Br J Anaesth 125 4 2020 447 449 10.1016/j.bja.2020.06.006 32571569 49 Arena R. Faghy M.A. Cardiopulmonary exercise testing as a vital sign in patients recovering from COVID-19 Expert Rev Cardiovasc Ther 19 10 2021 877 880 34623198 50 MacNamara J.P. McCoy C.W. Hendren N.S. The cost of return to play protocols in collegiate athletes recovering from coronavirus disease 2019 Med Sci Sports Exerc 54 7 2022 1051 1057 10.1249/mss.0000000000002896 Published online 2022 35220368 51 Kim J.H. Editorial commentary: Screening cardiac magnetic resonance imaging for athletes after COVID-19: Is it time to end the debate? Trends Cardiovasc Med 32 3 2022 151 34999022 52 Hull J.H. Wootten M. Moghal M. Clinical patterns, recovery time and prolonged impact of COVID-19 illness in international athletes: The UK experience Br J Sports Med 56 1 2022 4 11 10.1136/bjsports-2021-104392 34340972 53 Hull J.H. Loosemore M. Schwellnus M. Respiratory health in athletes: Facing the COVID-19 challenge Lancet Respir Med 8 6 2020 557 558 10.1016/S2213-2600(20)30175-2 32277869 54 Calcaterra G. Fanos V. Cataldi L. Cugusi L. Crisafulli A. Bassareo P.P. Need for resuming sports and physical activity for children and adolescents following COVID-19 infection Sport Sci Health 2022 1 7 Published online 2022 55 Schwellnus M. Adami P.E. Bougault V. International Olympic Committee (IOC) consensus statement on acute respiratory illness in athletes part 1: acute respiratory infections Br J Sports Med 56 2022 1089 1103 Published online 2022 57 Juno J.A. Wheatley A.K. Boosting immunity to COVID-19 vaccines Nat Med 27 11 2021 1874 1875 10.1038/s41591-021-01560-x 34764485 58 Troiano G. Nardi A. Vaccine hesitancy in the era of COVID-19 Public Health 194 2021 245 251 10.1016/j.puhe.2021.02.025 33965796 59 Rankin A. Hull J. Wootten M. Ranson C. Heron N. Infographic: Safety of the SARS-CoV-2 vaccination and addressing vaccine hesitancy in athletes Br J Sports Med 56 18 2022 1055 1056 Published online 2022 60 Neil H. Rankin A. McLarnon M. Hull J.H. Gomes C. A journey around the COVID-19 vaccine for athletes J Sci Cycling 10 1 2021 63 66 61 Creech C.B. Walker S.C. Samuels R.J. SARS-CoV-2 vaccines Jama. 325 13 2021 1318 1320 33635317 62 Chemaitelly H. Abu-Raddad L.J. Waning effectiveness of COVID-19 vaccines The Lancet 399 10327 2022 771 773 10.1016/S0140-6736(22)00277-X 63 Narducci D.M. Diamond A.B. Bernhardt D.T. Roberts W.O. COVID vaccination in athletes and updated interim guidance on the Preparticipation physical examination during the SARS-Cov-2 pandemic Clin J Sport Med 32 1 2022 e1 e6 10.1097/JSM.0000000000000981 34723865 64 Markov P.V. Katzourakis A. Stilianakis N.I. Antigenic evolution will lead to new SARS-CoV-2 variants with unpredictable severity Nat Rev Microbiol 2022 1 2 Published online 2022 65. World Health Organisation Coronavirus disease (COVID-19): Post COVID-19 condition Accessed April 13, 2022 https://www.who.int/news-room/questions-and-answers/item/coronavirus-disease-(covid-19)-post-covid-19-condition 2022 66 Lindsay R.K. Wilson J.J. Trott M. What are the recommendations for returning athletes who have experienced long term COVID-19 symptoms? Ann Med 53 1 2021 1935 1944 10.1080/07853890.2021.1992496 34726085 67 McDonnell D. ‘After Covid, he’s had problems with his heart, irregular heartbeats, getting dizzy’ – Stephen Bradley ‘gutted’ for rising rovers star Independent 2022 Published 2022. Accessed August 25, 2022 https://www.independent.ie/sport/soccer/league-of-ireland/after-covid-hes-had-problems-with-his-heart-irregular-heartbeats-getting-dizzy-stephen-bradley-gutted-for-rising-rovers-star-41934310.html 68 Loader Wilkinson T. Former ironman champion brought low by ‘long Covid’ describes the symptoms and what it’s like to suffer South China Morning Post 2022 Published 2021. Accessed August 25, 2022 https://www.scmp.com/lifestyle/health-wellness/article/3121237/long-covid-symptoms-what-its-suffer-and-why-fitness 69. Mackinnon K. Sacramento Showdown: TMC hosts make their picks - Frodeno, Iden, Sanders or … Svenningsson? Triathlon Magazine Canada Published October 20, 2021. Accessed August 25, 2022 https://triathlonmagazine.ca/news/sacramento-showdown-tmc-hosts-make-their-picks-frodeno-iden-sanders-or-svenningsson/ 2022 70 Hanson S.W. Abbafati C. Aerts J.G. A global systematic analysis of the occurrence, severity, and recovery pattern of long COVID in 2020 and 2021 medRxiv 2022 Published online 2022 71 Faghy M.A. Owen R. Thomas C. Is long COVID the next global health crisis? J Glob Health 2022 Published online In Press 72 Alka Vago Szabo L. Dohy Z. Merkely B. Cardiac magnetic resonance findings in patients recovered from COVID-19: Initial experiences in elite athletes JACC Cardiovasc Imaging 14 6 2022 1279 1281 Accessed August 25, 2022 https://www.jacc.org/doi/full/10.1016/j.jcmg.2020.11.014 74 Hendrickson B.S. Stephens R.E. Chang J.V. Cardiovascular evaluation after COVID-19 in 137 collegiate athletes: Results of an algorithm-guided screening Circulation. 143 19 2021 1926 1928 10.1161/CIRCULATIONAHA.121.053982 33970675 75. Prevalence of clinical and subclinical myocarditis in competitive athletes with recent SARS-CoV-2 infection: Results from the big ten COVID-19 cardiac registry \| Cardiology \| JAMA cardiology \| JAMA network Accessed August 25, 2022 https://jamanetwork.com/journals/jamacardiology/article-abstract/2780548 2022 76 Hwang C.E. Kussman A. Christle J.W. Froelicher V. Wheeler M.T. Moneghetti K.J. Findings from cardiovascular evaluation of National Collegiate Athletic Association Division I Collegiate Student-Athletes after Asymptomatic or mildly symptomatic SARS-CoV-2 infection Clin J Sport Med 32 2 2022 103 107 10.1097/JSM.0000000000000954 34173780	36462554	PMC9711907	NO-CC CODE	2022-12-06 23:15:40	no	Prog Cardiovasc Dis. 2022 Dec 1; doi: 10.1016/j.pcad.2022.11.014	utf-8	Prog Cardiovasc Dis	2,022	10.1016/j.pcad.2022.11.014	oa_other
==== Front J Clin Virol J Clin Virol Journal of Clinical Virology 1386-6532 1873-5967 Elsevier B.V. S1386-6532(22)00282-7 10.1016/j.jcv.2022.105350 105350 Article Performance evaluation of the Viasure PCR assay for the diagnosis of monkeypox: A multicentre study Tan Ngee Keong a⁎ Madona Cindy P. b Taylor Joshua F. a Fourali Lynda Hadjilah c Sehmi Jasveen K. c Stone Madeline J. a Pond Marcus J. b1 Cliff Penelope R. c1 Pope Cassie F. de1 a Department of Medical Microbiology, Infection and Immunity, South West London Pathology, St George's University Hospitals NHS Foundation Trust, London, SW17 0QT, United Kingdom b Department of Virology, Infection and Immunity, North West London Pathology, Imperial College Healthcare NHS Trust, London, W6 8RF, United Kingdom c Department of Infection Sciences, Synnovis, Guy's and St Thomas’ NHS Foundation Trust, London, SE1 7EH, United Kingdom d Infection Care Group, St George's University Hospitals NHS Foundation Trust, London, SW17 0QT, United Kingdom e Institute for Infection and Immunity, St George's, University of London, London, SW17 0RE, United Kingdom ⁎ Corresponding author. 1 These authors contributed equally to this work. 1 12 2022 1 2023 1 12 2022 158 105350105350 8 10 2022 27 11 2022 © 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. Background Monkeypox virus (MPXV) is the causative agent of the 2022 monkeypox global outbreak. Rapid detection of MPXV infection is essential to inform patient management and public health response. Currently, there is a lack of established real-time PCR assays to support a rapid diagnosis of monkeypox. Objectives To evaluate the performance characteristics of the Viasure MPXV PCR assay in three London teaching hospitals. Study design Prospectively collected paired patient swabs from matched or unmatched anatomical sites were evaluated by the reference laboratory and Viasure MPXV PCR assays. A subset of samples were also tested for HSV, VZV, and/or Treponema pallidum DNA. Results 217 paired samples were evaluated. 91.2% of the paired swabs generated concordant results whilst 8.8% generated discordant results. The accuracy, diagnostic sensitivity, diagnostic specificity, positive predictive value, negative predictive value, likelihood ratio positive, and likelihood ratio negative of the Viasure PCR assay across the hospitals were 93.2 – 96.3%, 90.0 – 100%, 88.2 – 100%, 94.9 – 100%, 87.9 – 100%, 8.50 – 14.41, and 0.00 – 0.10 respectively. MPXV co-infections with HSV were detected in two patients. Five patients were negative for monkeypox but positive for herpes or chickenpox. Conclusions The Viasure MPXV PCR assay demonstrated excellent performance characteristics, was easy to use, and is fit for routine diagnostic purpose. Where implemented, the assay would allow rapid and accurate laboratory diagnosis of MPXV infections and support a timely management of monkeypox. To reduce the risk of false negative detections, vesicular lesions from any anatomical site should be preferentially and optimally sampled. Keywords Monkeypox Sensitivity Specificity Predictive value Likelihood ratio Diagnosis ==== Body pmc1 Background Monkeypox virus (MPXV), an Orthopoxvirus of the Poxviridae family, is the causative agent of the 2022 monkeypox global outbreak [1]. In an ongoing epidemic, a rapid detection of monkeypox virus infection is essential to inform patient management, infection control and public health response. Currently, there is a lack of established PCR assays that are relatively easy to be implemented by routine diagnostic laboratories to support a rapid diagnosis of MPXV infections. To support a rapid detection, CerTest Biotec (Zaragoza, Spain) recently introduced the Viasure MPXV PCR assay which can generate up to 96 results in approximately 105 mins [2]. The lyophilised reagents enable storage between 2–40 °C for up to 2 years and only require reconstitution with a provided buffer prior to PCR setup. The ease of handling makes the assay attractive for use by routine diagnostic laboratories. The aim of the current study was to evaluate the performance characteristics of the Viasure MPXV PCR assay in three London teaching hospitals. 2 Study design 2.1 Collection and pre-analytical processing of samples Paired patient swabs from matched (e.g. throat/throat) or unmatched (e.g. chest/leg) anatomical sites collected in viral transport media (Σ-Virocult®, MWE, UK) and submitted for the investigation of monkeypox at three London hospital Trusts were evaluated (Imperial College Healthcare NHS Trust (hospital 1), St George's University Hospitals NHS Foundation Trust (hospital 2), and Guy's and St Thomas’ NHS Foundation Trust (hospital 3)). One sample was tested by the reference PCR assay at the national reference laboratory (Rare and Imported Pathogens Laboratory (RIPL), UK) and the other was tested by the Viasure MPXV PCR assay (CerTest Biotec, Spain) at the respective hospital pathology laboratory (North West London Pathology (NWLP) at hospital 1, South West London Pathology (SWLP) at hospital 2, and Synnovis at hospital 3). Discordant MPXV PCR results were resolved by a laboratory-developed test (LDT) (NWLP), or repeat testing at RIPL (SWLP and Synnovis). A subset of samples tested at SWLP were also tested for HSV-1, HSV-2, VZV, and/or Treponema pallidum DNA as part of the clinical investigation of a rash illness. The protocol for the pre-analytical processing of samples is summarised in Table 1 .Table 1 Sample inactivation and extraction protocols for the Viasure PCR assay. Table 1Protocol Volume (µl) NWLP (Hospital 1) SWLP (Hospital 2) Synnovis (Hospital 3) Inactivation Sample 125 200 200 Lysis buffer (LB) 125 (External LB, Roche) 200 (NucliSENS® LB, bioMerieux) 430 (ATL/ACL LB, Qiagen) Heating Not performed Not performed 68 °C (15 mins) Extraction Input 200 400 630 Output 60 60 60 Instrument EZ1 Advanced XL (Qiagen) EZ1 Advanced XL (Qiagen) QIAsymphony SP (Qiagen) 2.2 RIPL PCR assay The RIPL MPXV RT-PCR assay was adapted from published literature [3]. The assay targets the G2R region common to all MPXV sequences. Samples were extracted on either the EZ1 Advanced XL (Qiagen, Germany) or MagNA Pure 96 (Roche Diagnostics, UK) instruments. MS2 phage was used as extraction and amplification control. PCR thermocycling was performed on the ABI ViiA 7 system. 2.3 NWLP LDT The NWLP MPXV PCR assay was adapted from published literature [3,4]. The assay targets the G2R_WA and E9L-NVAR sequences of MPXV. Samples were extracted using the 200 µl input and 60 µl output protocol with the Virus Mini Kit v2.0 on the EZ1 Advanced XL (Qiagen, Germany). PCR thermocycling was performed with QuantiFast® Pathogen PCR +IC kit (Qiagen, Germany) according to the manufacturer's instructions on the ABI 7500 Fast system. 2.4 Viasure MPXV PCR assay The Viasure MPXV PCR assay was performed according to the manufacturer's instructions [2]. Briefly, the lyophilised and ready-to-use PCR mastermix was reconstituted with the provided buffer prior to dispense (15 µl) into 96-well PCR plates or Rotor-Gene PCR tubes. For each sample, 5 µl extract was subsequently added to the well or tube. The hands-on time required for setting up 20 samples was 5 mins. PCR thermocycling was performed on the ABI 7500 Fast system (NWLP and Synnovis) or Rotor-Gene Q (SWLP). The assay targets the G2R_G and F3L genes of MPXV and uses the haemoglobin-β gene as extraction, amplification and sample adequacy control. 2.5 Data analysis Data analysis was performed using MedCalc Software (MedCalc Software Ltd, Belgium). Accuracy, diagnostic sensitivity, diagnostic specificity, positive predictive value (PPV), negative predictive value (NPV), likelihood ratio positive (LR+), and likelihood ratio negative (LR-) of the Viasure PCR assay with 95% confidence interval (95% CI) were computed. 3 Results In total, 217 paired clinical samples were evaluated by the RIPL and Viasure MPXV PCR assays (Fig. 1 ). Most patients presented at sexual health clinics and had a recent contact with monkeypox. 91.2% (198/217) of the paired swabs generated concordant results whilst 8.8% (19/217) generated discordant results. 72.2% (143/198) and 27.8% (55/198) of the concordant results were obtained from paired swabs collected from matched and unmatched anatomical sites respectively. Of the discordant results, 31.6% (6/19) were obtained from paired swabs collected from matched anatomical sites whilst 68.4% (13/19) were obtained from unmatched sites. MPXV co-infections with HSV-1 (n = 1) and HSV-1 and HSV-2 (n = 1) were detected in two patients that had additional testing at SWLP. In addition, five patients were negative for monkeypox but positive for herpes (HSV-2, n = 2) or chickenpox (n = 3).Fig. 1 Paired swab samples evaluated by the RIPL and Viasure PCR assays. Fig 1 Overall, the performance of the Viasure MPXV PCR assay was satisfactory (Table 2 ). The assay demonstrated good accuracy (88.6 – 94.7%), diagnostic sensitivity (89.7 – 100%), PPV (87.2 – 96.3%), and NPV (87.9 – 100%). The LR+ (4.00 – 41.24) indicated that the Viasure assay may be useful to rule in (i.e. confirm) a MPXV infection when the result is detected, and the LR- (0.00 – 0.11) may be sufficiently small to rule out an infection when the result is not detected. The assay exhibited a lower diagnostic specificity of 82.9% (95% CI 66.4 – 93.4%) and 75.0% (95% CI 50.9 – 91.3%) at NWLP and Synnovis respectively due to the testing of paired swabs predominantly collected from unmatched anatomical sites (Table 3 ). The paired swabs collected from matched anatomical sites i.e., throat/throat (sample 7) at NWLP and lesion/lesion and skin/skin (sample 17 and 18) at Synnovis, were two different swabs. Discordant testing of all swabs (where available and depending on which swab was tested) confirmed the primary testing results by the RIPL or Viasure assay. If the discordant testing results (Table 3) were included in the analysis, the accuracy, diagnostic sensitivity, diagnostic specificity and PPV of the Viasure MPXV PCR assay improved to 93.2 – 96.3%, 90.0 – 100%, 88.2 – 100%, and 94.9 – 100% respectively, and as a result, improves the LR+ (8.50 – 14.41) and LR- (0.00 – 0.10) of the Viasure assay.Table 2 Performance characteristics of the Viasure MPXV PCR assay at NWLP, SWLP and Synnovis laboratories. Table 2 NWLP (Hospital 1) SWLP (Hospital 2) Synnovis (Hospital 3) RIPL (reference assay for all laboratories) Detected Not detected Detected Not detected Detected Not detected Viasure Detected 49 6⁎⁎ 26 1⁎⁎⁎⁎ 34 5⁎⁎⁎⁎⁎ Not detected 4* 29 3⁎⁎⁎ 45 0 15 Accuracy (95%CI) 88.6 (80.1 – 94.4) 94.7 (86.9 – 98.5) 90.7 (79.7 – 96.9) Diag. Sensitivity (95%CI) 92.5 (81.8 – 97.9) 89.7 (72.7 – 97.8) 100.0 (89.7 – 100.0) Diag. Specificity (95%CI) 82.9 (66.4 – 93.4) 97.8 (88.5 – 99.9) 75.0 (50.9 – 91.3) PPV (95% CI) 89.1 (79.7 – 94.4) 96.3 (78.8 – 99.5) 87.2 (76.1 – 93.6) NPV (95% CI) 87.9 (73.6 – 95.0) 93.8 (83.7 – 97.8) 100.0 LR+ (95% CI) 5.39 (2.59 – 11.22) 41.24 (5.91 – 287.7) 4.00 (1.87 – 8.55) LR- (95% CI) 0.09 (0.04 – 0.24) 0.11 (0.04 – 0.31) 0.00 NWLP, North West London Pathology; SWLP, South West London Pathology; RIPL, Rare and Imported Pathogens Laboratory; PPV, positive predictive value; NPV, negative predictive value; LR+, likelihood ratio positive; LR-, likelihood ratio negative; CI, confidence interval. ⁎ Sample 1 – 4 in Table 3. ⁎⁎ Sample 5 – 10 in Table 3. ⁎⁎⁎ Sample 11 – 13 in Table 3. ⁎⁎⁎⁎ Sample 14 in Table 3. ⁎⁎⁎⁎⁎ Sample 15 – 19 in Table 3. Table 3 Discordant PCR results. Table 3Laboratory Sample No Assay Primary testing Discordant testing RIPL Viasure LDT RIPL Result Site Result (Ct) Site Result (Ct) Site Result (Ct) Site NWLP 1 D Lesion ND Throat ND Throat – – 2 D Penile ND Throat D (39.0) Penile – – 3 D Swab ND Throat ND Throat – – 4 D Swab ND Skin ND Skin – – 5 ND Lesion D (33.9) Throat Inhibitory Lesion – – 6 ND Skin D (29.0) Throat – – – – 7 ND Throat D (23.3) Throat D (28.8) Throat – – 8 ND Skin D (27.9) Perianal D (32.9) Perianal – – 9 ND Hand D (14.1) Perianal D (18.8) Perianal – – 10 ND Swab D (16.3) Perianal D (23.0) Perianal – – SWLP 11 D Skin ND Throat – – – – 12 D Groin ND Throat – – – – 13 D Swab ND Throat – – – – 14 ND Throat D (14.3) Lesion – – D Lesion Synnovis 15 ND Chest D (19.5) Leg – – D (21.4) Leg 16 ND Groin D (29.0) Ear – – D (30.9) Ear 17 ND Lesion D (14.5) Lesion – – D (19.1) Lesion 18 ND Skin D (38.1) Skin – – – – 19 ND Perianal D (38.0) Lesion – – – – RIPL, Rare and Imported Pathogens Laboratory; LDT, laboratory developed test; Ct, cycle threshold; NWLP, North West London Pathology; SWLP, South West London Pathology; D, detected; ND, not detected; -, Not performed. 4 Discussion The response to the ongoing epidemic of monkeypox necessitates the availability of diagnostics that enable rapid detection of MPXV from clinical samples. We found the commercially available Viasure MPXV PCR assay demonstrated excellent accuracy, diagnostic sensitivity and diagnostic specificity. The assay requires minimal user input and hands-on time, and is suitable for routine diagnostic laboratory investigations of monkeypox. As MPXV shedding in bodily fluids may be low in the early and late phases of infection, and at certain anatomical sites relative to skin lesions [5], optimal swab sampling is essential to reduce the risk of false negative results and improve the accuracy of the Viasure assay. We found vesicular lesions from any anatomical site, if present, should be preferentially sampled. In addition, to improve the sensitivity of the assay, testing swabs collected from different anatomical sites may be warranted. In the context of detecting MPXV in a predominantly at-risk population that may be co-infected with other organisms, or presents with an undifferentiated rash disease, an assay with a high specificity or LR+ that can rule in a MPXV infection is necessary. In-silico analysis [6] and our analytical specificity (data not shown) and diagnostic specificity data have demonstrated that the Viasure assay is specific, and provides a clinically useful LR+ that would support clinicians in their diagnosis of MPXV infections. To reduce the risk of false positive results caused by contamination from high titre MPXV samples, the incorporation of negative controls at various testing steps is suggested. In conclusion, the Viasure MPXV PCR assay demonstrated excellent performance characteristics and is fit for routine diagnostic purpose. If implemented, the assay would allow a rapid and accurate laboratory diagnosis of MPXV infections, and facilitate a timely management of monkeypox. Ethical approval Not required. Funding The Viasure MPXV PCR assay reagents were supplied free of charge by CerTest Biotec. CerTest Biotec had no role in the study design, data collection and analysis, result interpretation, writing of the manuscript and the decision to submit the article for publication. CRediT authorship contribution statement Ngee Keong Tan: Conceptualization, Methodology, Validation, Formal analysis, Data curation, Writing – original draft, Visualization. Cindy P. Madona: Investigation, Writing – review & editing. Joshua F. Taylor: Investigation, Data curation, Writing – review & editing. Lynda Hadjilah Fourali: Investigation, Data curation, Writing – review & editing. Jasveen K. Sehmi: Investigation, Data curation, Writing – review & editing. Madeline J. Stone: Investigation, Writing – review & editing. Marcus J. Pond: Conceptualization, Methodology, Investigation, Data curation, Writing – review & editing, Supervision. Penelope R. Cliff: Conceptualization, Methodology, Writing – review & editing, Supervision. Cassie F. Pope: Conceptualization, Methodology, Writing – review & editing, Supervision. Declaration of Competing Interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: JFT received conference registration expense from Pro-Lab Diagnostics to attend the 2022 Federation of Infection Societies Conference. All other authors have no relevant financial or non-financial competing interests to declare. Acknowledgements We thank the laboratory teams at NWLP, SWLP, Synnovis and RIPL for their assistance in processing patient samples. We also thank RIPL, Charlotte Duncan at Pro-Lab Diagnostics and Henar Alonso at CerTest Biotec for their invaluable support in this study. ==== Refs References 1 Thornhill J.P. Monkeypox virus infection in humans across 16 countries - April-June 2022 N. Engl. J. Med. 387 8 2022 679 691 35866746 2 CerTest Biotec., Viasure Monkeypox virus real-time PCR detection assay (Instructions for use). 2022. 3 Li Y. Real-time PCR assays for the specific detection of monkeypox virus West African and Congo basin strain DNA J. Virol. Methods 169 1 2010 223 227 20643162 4 Li Y. Detection of monkeypox virus with real-time PCR assays J. Clin. Virol. 36 3 2006 194 203 16731033 5 Peiro-Mestres A. Frequent detection of monkeypox virus DNA in saliva, semen, and other clinical samples from 12 patients, Barcelona, Spain, May to June 2022 Euro Surveill 27 28 2022 6 CerTest Biotec., Viasure Monkeypox virus Technical Report. 2022.	36473345	PMC9711911	NO-CC CODE	2022-12-05 23:15:17	no	J Clin Virol. 2023 Jan 1; 158:105350	utf-8	J Clin Virol	2,022	10.1016/j.jcv.2022.105350	oa_other
==== Front Br J Anaesth Br J Anaesth BJA: British Journal of Anaesthesia 0007-0912 1471-6771 Elsevier S0007-0912(21)00669-3 10.1016/S0007-0912(21)00669-3 Article Contents 17 11 2021 12 2021 17 11 2021 127 6 ivvi 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. ==== Body pmc	0	PMC9711913	NO-CC CODE	2022-12-10 23:20:29	no	Br J Anaesth. 2021 Dec 17; 127(6):iv-vi	utf-8	Br J Anaesth	2,021	10.1016/S0007-0912(21)00669-3	oa_other
==== Front J Mol Diagn J Mol Diagn The Journal of Molecular Diagnostics : JMD 1525-1578 1943-7811 American Society for Investigative Pathology S1525-1578(21)00336-6 10.1016/S1525-1578(21)00336-6 Article Table of Contents 25 11 2021 12 2021 25 11 2021 23 12 A5A7 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. ==== Body pmc	0	PMC9711914	NO-CC CODE	2022-12-10 23:20:29	no	J Mol Diagn. 2021 Dec 25; 23(12):A5-A7	utf-8	J Mol Diagn	2,021	10.1016/S1525-1578(21)00336-6	oa_other
==== Front JACC Heart Fail JACC Heart Fail Jacc. Heart Failure 2213-1779 2213-1787 by the American College of Cardiology Foundation. Published by Elsevier. S2213-1779(21)00441-8 10.1016/j.jchf.2021.09.003 Covid Rapid Reports Editorial Comment Look After You Leap∗ Whellan David J. MD, MHS ∗ Division of Cardiology, Department of Medicine, Thomas Jefferson University, Philadelphia, Pennsylvania, USA ∗ Address for correspondence: Dr David J. Whellan, Jefferson Medical College, Department of Medicine, 2400 Pratt Street, Durham, North Carolina 27705, USA. ∗ Editorials published in JACC: Heart Failure reflect the views of the authors and do not necessarily represent the views of JACC: Heart Failure or the American College of Cardiology. 29 11 2021 12 2021 29 11 2021 9 12 925926 © 2021 by the American College of Cardiology Foundation. Published by Elsevier. 2021 American College of Cardiology Foundation 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. Corresponding Author Key Words COVID19 quality of care telehealth ==== Body pmcWhen the COVID-19 pandemic surged across the United States in March 2020, health care systems, providers, and patients pivoted to telehealth to keep patients safe and away from public spaces like clinics and hospitals. This change in health care delivery was an unprecedented event in modern history. We were fortunate to have the necessary tools and technologies available, but never had they been used so quickly or on such a massive scale. In many ways, we were flying blind. Consistent with quality improvement procedures, it is important to continually assess the new strategy and consider modifications that can lead to better results. In this issue of JACC: Heart Failure, Sammour et al (1) review the changes to HF care during COVID-19 and assess the impact of those changes in a large Midwestern health care system. Their retrospective analysis compares the care provided by a cardiology group (66 cardiologists and 47 advance practice providers) over the initial 3 months of the pandemic (March 15 to June 15, 2020) with prepandemic care of patients with HF during the same months in 2018 and 2019, representing approximately 5,000 visits for each time period. The cardiology practice pivoted almost 180 degrees in 2020, going from no telehealth visits to approximately 88.5% in 2020. The investigators concluded that telehealth visits provided an effective substitute for in-person office visits. In the propensity-matched analysis, patients with HF cared for by telehealth had lower emergency department visits or hospitalizations at 30 and 90 days than did similar patients seen in person, with similar intensive care unit admissions and deaths. Before using telehealth as the new HF management strategy, we may want to note some significant limitations of the current analysis. First and foremost is that it looks at only 3 months of care. HF is a chronic disease that requires ongoing interactions with patients, management of therapies, and consideration of new strategies. As noted by the authors, they have not provided any information on the management of guideline-directed therapies. Three months may not be enough time to understand the impact on clinical events or survival, particularly if there were no adjustment to guideline-directed therapies. Adjusting a diuretic dose in response to worsening symptoms in order to avoid decompensation has short-term benefits, but the long-term effects of high diuretic doses are unclear (2). For many reasons, including ethical and logistical, these results do not reflect the outcomes from the traditional criterion standard of a randomized controlled trial. The investigators attempted to account for this by using a propensity-matched cohort, but they recognize that there are limitations to this approach. The telehealth cohort represented a selected group of patients, and that choice likely inserted bias into the analysis. In addition, several components were not considered in the analysis, including the type of provider or the method of telehealth visit. The current analysis involved a single practice group committed to using the technology because of the critical situation, which indicates that they were prepared to interact with patients and provide care through the platform. These results may have differed if the analysis had been conducted across several practices during more normal circumstances. There are several examples in randomized controlled studies in which participating investigators and sites do not use the data being provided (3) or do not close the loop by providing appropriate changes in treatment (4,5), either by not responding to data or by overreacting to alerts. This may be due in part to feeling constrained by protocols that dictate adjustments. In the current article, providers used the information as they wished, integrating the information obtained with their knowledge about their patients, including the success of previous medication adjustments. Access to a technology platform is the intervention being evaluated; there were not protocols for adjusting medications. Providers and practices considering the implementation of a telehealth strategy, particularly video visits, will need to consider barriers to access for both in-person and telehealth visits. Studies have consistently found that younger patients and those who speak English feel more comfortable with telehealth visits (6,7). In addition, patients are more likely to select telehealth visits when in-person visits have higher out-of-pocket costs (eg, parking charges or clinic copayments) or greater time commitment (eg, travel time) (6). Inasmuch as HF is a disease of the elderly, the consistent finding that older age is associated with less access to the internet and delayed adoption of technology, including digital health, is an issue that needs to be addressed. Even more, patients are less likely to participate in telehealth visits if they have low household income or, as in the current study, have Medicaid coverage. Although this pandemic has inflicted a significant amount of pain and suffering on the world, there are small benefits that we can appreciate. One is that humans are an immensely adaptable group, ready to use what is available to solve a problem. Telehealth was an available technology, and by using it as broadly as we did during the pandemic, we have likely altered the way medicine will be practiced from now on. There is no going back; we have leapt. Yet, this is not the time to go blindly into this new paradigm. We need to continue to evaluate and adjust. Funding Support and Author Disclosures The author has reported that he have no relationships relevant to the contents of this paper to disclose. The author attests he is in compliance with human studies committees and animal welfare regulations of the author’s institution and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center. ==== Refs References 1 Sammour Y. Spertus J.A. Austin B.A. Outpatient management of heart failure during the COVID-19 pandemic after adoption of a telehealth model J Am Coll Cardiol HF 9 2021 916 924 2 Hasselblad V. Gattis Stough W. Shah M.R. Relation between dose of loop diuretics and outcomes in a heart failure population: results of the ESCAPE trial Eur J Heart Fail 9 2007 1064 1069 17719273 3 Fiuzat M. Ezekowitz J. Alemayehu W. Assessment of limitations to optimization of guideline-directed medical therapy in heart failure from the GUIDE-IT trial: a secondary analysis of a randomized clinical trial JAMA Cardiol 5 2020 757 764 32319999 4 Loh J.P. Barbash I.M. Waksman R. Overview of the 2011 Food and Drug Administration Circulatory System Devices Panel of the Medical Devices Advisory Committee Meeting on the CardioMEMS Champion Heart Failure Monitoring System J Am Coll Cardiol 61 2013 1571 1576 23352783 5 van Veldhuisen D.J. Braunschweig F. Conraads V. Intrathoracic impedance monitoring, audible patient alerts, and outcome in patients with heart failure Circulation 124 2011 1719 1726 21931078 6 Reed M.E. Huang J. Graetz I. Patient characteristics associated with choosing a telemedicine visit vs office visit with the same primary care clinicians JAMA Netw Open 3 2020 e205873 7 Eberly L.A. Kallan M.J. Julien H.M. Patient characteristics associated with telemedicine access for primary and specialty ambulatory care during the COVID-19 pandemic JAMA Netw Open 3 2020 e2031640	34857176	PMC9711939	NO-CC CODE	2022-12-02 23:21:31	no	JACC Heart Fail. 2021 Dec 29; 9(12):925-926	utf-8	JACC Heart Fail	2,021	10.1016/j.jchf.2021.09.003	oa_other
==== Front Int J Surg Int J Surg International Journal of Surgery (London, England) 1743-9191 1743-9159 IJS Publishing Group Ltd. Published by Elsevier Ltd. S1743-9191(20)30723-8 10.1016/j.ijsu.2020.09.048 Commentary A commentary on ‘Global prevalence and reasons for case cancellation on the intended day of surgery: A systematic review and meta analysis' (Int. J. Surg. 2020; Epub ahead of Print) Chellam Shrividya ∗ Dalal Kajal Toal Pratibha Department of Anaesthesia, BARC Hospital, Mumbai, 400094, India ∗ Corresponding author. 22 10 2020 12 2020 22 10 2020 84 8586 17 9 2020 22 9 2020 © 2020 IJS Publishing Group Ltd. Published by Elsevier Ltd. 2020 IJS Publishing Group 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. Keywords Cancellation Surgery Community hospital LMIC ==== Body pmcDear Editor, We congratulate the authors Abate SM et al. on their meta-analysis of global prevalence and reasons for case cancellation on the intended day of surgery [1]. The authors have addressed an important aspect of postponement of surgeries, as it is an indirect indicator of efficiency of system and quality of patient care [2]. The authors have documented the global prevalence rates of cancellation at 18% with maximum cancellations in the low and middle-income countries. Unavailability of operation theater (OT) was the commonest reason for cancellations globally as documented by them. It would be worthwhile though to know the distribution of reasons of delay across various country income groups to better understand the health system related delays in a low middle income setting. The authors also point to corona affected patients in the section of selection criteria but the issue remains unaddressed further in results and discussion. We documented much lower prevalence of cancellations of surgeries in an urban community hospital in Mumbai, India compared to that mentioned by Abate et al. Our facility caters to 100,000 population covered under universal health coverage scheme (UHC) provided by employees’ healthcare scheme by the Government of India. We were inspired to analyze the delays at our facility over one year period 2019–2020. Postponement on day of surgery was defined as any surgery scheduled on the final OT list which was not performed on that day. Reasons for postponement were listed as 1) inadequate pre-operative optimization 2) change in clinical status 3) lack of OT time, personnel, bed, or equipment 4) patient unwilling or not showed up 5) miscellaneous [2]. Out of 3148 scheduled surgeries, 118 (3.74%) were cancelled on the scheduled day. Various reasons for cancellation documented, are given in Table 1 . The postponement rate in our hospital was significantly lower than that documented by Abate et al. in LMICs (23–40%). Though we found similar rates of cancellations due to unavailability of OT and infrastructure (25%)to that of Abate et al. patient refusal or no show was the commonest reason for cancellation in our study (37%).Table 1 Reasons for cancellation. Table 1Reason for cancellation Number of patients (n = 118) Patient not turned up 39 (33.05%) Inadequate medical optimization 31(26.27%) Lack of OT, personnel, bed, or equipment 30 (25.42%) Change in clinical status 8 (6.77%) Patient non consent 5 (4.23%) Miscellaneous 5(4.23%) Currently, the rates of cancellations of surgeries range from 10% to 40% across hospitals in India [2]. We observed that failure of patients to show up on the day of surgery was the commonest cause in our set up. Under the UHC, employees get a lifelong healthcare coverage and no monetary transactions are involved. Last minute cancellations do not lead to financial losses for them and rescheduling is easy due to good accessibility and availability. Hence, major reason for postponement in our set up was not facility related issues but cancellations by patients. Our UHC reduces the well-established barriers of accessibility, availability and affordability to surgical care delivery [3]. A repeat preoperative screening, a day prior to scheduled date of surgery by anesthetists, helped in optimizing the medical condition of the patient, in our hospital. This allowed the teams to gear up for challenges and reduce the delay due to changing medical status or optimization as is well documented to reduce cancellations of surgeries [4]. Provenance and peer review Commentary, internally reviewed ==== Refs References 1 Abate SM, Chekole YA, Minaye SY et al. Global prevalence and reasons for case cancellation on the intended day of surgery: A systematic review and Meta-Analysis, Int. J. Surg. Open 26(2020)55-63, 10.1016/j.ijso.2020.08.006. 2 Kumar R. Gandhi R. Reasons for cancellation of operation on the day of intended surgery in a multidisciplinary 500 bedded hospital J. Anaesthesiol. Clin. Pharmacol. 28 1 2012 66 69 10.4103/0970-9185.92442 22345949 3 Meara John G. Leather Andrew J.M. Hagander Lars Global surgery 2030: evidence and solutions for achieving health, welfare, and economic development Lancet 386 9993 2015 Aug 569 624 10.1016/S0140-736(15)60160-X 25924834 4 Ferschl Marla Tung Avery Sweitzer Bobbiejean Huo Dezheng Glick David Preoperative clinic visits reduce operating room cancellations and delays Anesthesiology 103 2005 855 859 10.1097/00000542-200510000-00025 16192779	33132143	PMC9711966	NO-CC CODE	2022-12-02 23:22:03	no	Int J Surg. 2020 Dec 22; 84:85-86	utf-8	Int J Surg	2,020	10.1016/j.ijsu.2020.09.048	oa_other
==== Front Cities Cities Cities (London, England) 0264-2751 0264-2751 Elsevier Ltd. S0264-2751(21)00174-8 10.1016/j.cities.2021.103274 103274 Article Collaborative neighborhood governance and its effectiveness in community mitigation to COVID-19 pandemic: From the perspective of community workers in six Chinese cities Liu Zhilin a Lin Sainan b Shen Yue c Lu Tingting d⁎ a School of Public Policy and Management, Hang Lung Center for Real Estate, Tsinghua University, China b School of Urban Design, Wuhan University, China c Research Center for China Administrative Division, East China Normal University, China d School of International and Public Affairs & China Institute for Urban Governance, Shanghai Jiao Tong University, China ⁎ Corresponding author. 24 5 2021 9 2021 24 5 2021 116 103274103274 6 8 2020 4 3 2021 20 5 2021 © 2021 Elsevier Ltd. All rights reserved. 2021 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 COVID-19 pandemic is a governance challenge for nations and cities across the world. While early observations have primarily focused on government actions, neighborhoods are at the frontline for coordinating grassroots level joint actions to fight against the pandemic. We draw from the collaborative governance theory and develop a theoretical framework for understanding the horizontal and hierarchical dynamics of collaborative neighborhood governance during crisis responses in urban China. Using a large-scale questionnaire survey of frontline community workers operated in six Chinese cities in February 2020, we conduct statistical analyses and find that the effectiveness of neighborhood collaboration in the pandemic control is predicted by both neighborhood social capital (i.e. civic engagement and citizen participation) and hierarchical steering by the government through setting policy priorities and providing support. Our research contributes to the international literature on neighborhood governance dynamics and provides policy lessons for improving neighborhood governance capacity in crisis response situations. Keywords Neighborhood governance Collaborative governance Residents' committees Public health crisis COVID-19 China ==== Body pmc1 Introduction The COVID-19 pandemic is more than a public health crisis; it is also an urban governance challenge for both developed and developing countries (Dodds et al., 2020). Urban settlements are subject to greater risks of virus transmission due to high density and spatial mobility (Kapucu, 2012). Furthermore, the pandemic has increased existing inequalities and divisions across social groups and residential communities (Kim & Bostwick, 2020; Maroko et al., 2020). A myriad of studies has been published on how countries responded to this public health crisis while minimizing social and economic losses caused by the pandemic. Most studies have focused on government actions (e.g. Benavides & Nukpezah, 2020; Mallinson, 2020; Mei, 2020; Migone, 2020; Yan et al., 2020), while relatively few studies examined how COVID-19 responses have been organized and coordinated at the grassroots level. Some scholars have observed the important roles of community-layer organizations in adopting and enforcing public health measures, as well as responding to the needs and concerns of local communities (Brodkin, 2021; Cheng et al., 2020). In developing countries such as Brazil, Thailand and Kenya, community health workers helped bridge the gap between government and citizens, provide critical pandemic-related information, enforce public health measures, and provide assistance to vulnerable populations (Lotta et al., 2020; Sudhipongpracha & Poocharoen, 2021). During COVID-19, China was quick to tame the spread of the virus, with most cities beginning the phased re-opening in late February 2020. Whereas many studies attributed China's success in controlling the pandemic to the firm leadership provided by the central government and coordination between central and local governments (Liu et al., 2021; Mei, 2020), some scholars have stressed the effective cross-sector collaboration at the neighborhood level, coordinated by residents' committees, as another critical factor (Cheng et al., 2020). Led and supported by city governments, residents' committees mobilized and collaborated with residents, other community-level organizations, as well as outside private and non-profit organizations, to form community-level joint action groups in response to the pandemic (Cheng et al., 2020; Zhao & Wu, 2020). This phenomenon provides urban scholars a unique opportunity to revisit the long-time theoretical and policy debate in the international literature, with respect to state-society relations in governing grassroots level public affairs in urban settlements. On the one hand, the conventional wisdom vouches for more spontaneous organization of the civil society in neighborhood governance, built upon horizontal ties, social capital, and civic organizations in residential neighborhoods (Sampson et al., 1997). On the other, in recent decades, cities in the US, China, and other countries have seen more active involvement of the government in steering and sponsoring reforms in public service delivery and neighborhood governance (Bray, 2006; Chen et al., 2009; Fagotto & Fung, 2006; Li et al., 2019; Liu, 2008). Nonetheless, how do the state-society dynamics play out and shape the effectiveness of collaborative neighborhood governance during COVID-19 pandemic responses remains unclear. In this paper, we draw from the collaborative governance literature (Ansell & Gash, 2007; Emerson et al., 2011), and adapt it to the context of urban neighborhood governance during public health emergencies. Using a questionnaire survey of frontline community workers from six cities—conducted in early February 2020—we examined the determinants of the neighborhood governance collaboration during the early stage of COVID-19 responses in China. We adopted multiple regression analyses to test the extent to which the effectiveness of such collaboration was predicted by both the internal conditions (i.e., neighborhood social capital) and external factors (i.e., hierarchical steering by the government). Controlling for perceived crisis level, city features, and respondents' socio-demographic features, we find that strong civic engagement and community participation, rather than neighborhood social trust, predicted a higher perceived level of collaboration effectiveness in COVID-19 responses. In addition, the role of hierarchical steering was primarily reflected in the government setting policy priorities and providing critical support for joint pandemic-control efforts in urban neighborhoods, rather than simply asserting a top-down incentive structure or an oversight authority. Our intellectual contributions are threefold. Firstly, scholars have written extensively about the emerging forms and power structures of neighborhood governance, mostly based on observations of non-crisis situations (Hemphill et al., 2006; Li et al., 2020; Parés et al., 2017; Provan et al., 2005; Wang, 2016). Focusing on real-time experiences of frontline community workers during COVID-19, we hope to revisit the structure, agencies, and dynamics of neighborhood governance in a crisis situation. Our empirical assessments of China's experiences of collaborative neighborhood governance in COVID-19 responses also seek to inform urban policy and governance for better preparedness and responses to public health crises. Secondly, we draw from the collaborative governance literature to develop a theoretical framework for understanding the dynamics of neighborhood collaboration in public health crisis responses. Previous studies have primarily focused on the internal factors of horizontal interaction and civic engagement in neighborhood governance (McGuire & Silva, 2010; Provan & Milward, 1995; Wang, 2016). Recently scholars have begun to recognize the importance of hierarchical mechanisms that facilitate horizontal collaboration at the grassroots level (Acar et al., 2008; Hafer, 2018; Rodríguez et al., 2007). Our empirical findings of residents' committees in Chinese cities expand the knowledge of collaborative neighborhood governance by highlighting the positive role of hierarchical steering during crisis response situations. Thirdly, existing studies have considered neighborhood governance in urban China as a contentious area of urban institutional reforms in the context of marketization and privatization (Bray, 2005; Fu et al., 2015; He, 2015; Wu, 2002, Wu, 2018). Yet the majority of existing studies have adopted qualitative case study approaches. Built on a unique dataset from a large-scale survey of local community workers from multiple Chinese cities, we are able to statistically test our theoretical hypotheses drawn from the existing literature as well as further explore heterogeneity across cities and locations. The rest of the paper is divided into four sections. The next section provides a review of neighborhood governance in China and collaborative governance in the urban context, as well as an explanation of the theoretical framework. The data and method section presents the survey strategies and data briefs. Following a descriptive analysis, the results of the multiple regression analyses are discussed. A conclusive summary is provided in the final section. 2 Literature review and theoretical framework 2.1 China's urban neighborhood governance in transition Over the past four decades, a rapid spatial and institutional transition has led to transformative changes in how neighborhoods are organized and governed in Chinese cities. In the pre-reform era, urban neighborhoods were primarily organized around the state work-unit (danwei) system, which provided housing and welfare services to state employees (Bray, 2005). Meanwhile, residents' committees, as subsidiaries of the government, recruited community workers to organize neighborhoods that were not affiliated with any specific work unit (Lu & Perry, 1997). With the economic reforms came the transformation of neighborhood governance. Residents' committees had taken over the responsibility of neighborhood service provision since the dismantling of the work-unit system (Wu, 2002). Although residents' committees were legally deployed to represent and serve the interests of all the residents of a particular neighborhood, in practical terms, they had only limited autonomy, capacity, leverage, or representation to organize grassroots public affairs in a bottom-up manner (Liu, 2008). Other dimensions of the urban reforms further complicated the functioning of residents' committees in urban neighborhood governance in China. First, housing marketization has brought the private sector into neighborhood governance, leading to a market-based provision of neighborhood services (Lu et al., 2020). The existing literature has highlighted state–market–society tensions and even conflicts in neighborhood governance, e.g., among residents' committees representing the state, property management companies functioning as a market player, and homeowner associations—as well as the residents themselves—representing civil society (Fu & Lin, 2014; Read, 2003). Second, social polarization and residential segregation also challenged China's urban neighborhood governance. Ever since the housing marketization process began, residents have found themselves in heterogeneous urban neighborhoods (Fang et al., 2020; Li & Wu, 2008). Especially in relatively deprived neighborhoods, residents' committees have had to serve as a liaison between the government and residents in order to secure alternative means of neighborhood service delivery, which in turn has increased the dependence of residents' committees on government funding and resources. Third, the traditionally strong social fabric of urban neighborhoods has been dissolved by increased residential mobility toward new residential spaces in the suburbs, the massive redevelopment of inner-city old neighborhoods, and the massive influx of rural-to-urban migrants (Lin et al., 2020; Shen et al., 2015). Neighborhood governance has faced the challenge of enhancing neighborhood social capital and civic engagement through the promotion of participatory and reciprocal activities (Read, 2003). The above circumstances highlight the peculiar position of residents' committees in neighborhood government in urban China. While being positioned as a grassroots community organization, residents' committees also function as an extension of the government apparatus for social organization and control. While historically having only a marginal role and inadequate resources, residents' committees are now being called upon by the government as well as residents to ensure adequate service provision and maintain governance efficacy (Wu, 2018). In fact, since the late 1990s, the state has initiated multiple waves of community-building campaigns with the aim of re-asserting the power of the state in local/neighborhood governance and ensuring the governability of urban spaces (Bray, 2006; Fu & Lin, 2014). These campaigns have primarily revolved around strengthening the presence and capacity of residents' committees in neighborhood public affairs, ranging from facilitating the social-service delivery function of sub-district governments (jiedao); resolving conflicts between property management companies, homeowner associations, and individual residents; and coordinating bottom-up neighborhood activity organizations and the engagement of non-governmental organizations in neighborhood governance. This transitional form of neighborhood governance has been tested during the fight against the COVID-19 public health crisis since January 2020. Residents' committees have played a key role in enforcing government lockdown measures, ensuring timely contact-tracing to contain the transmission of the virus, and providing necessary lifeline supports and social services to local residents (Zhao & Wu, 2020). It was reported that nearly four million community workers from 650,000 residents' committees in both rural and urban areas were mobilized to coordinate the grassroots joint actions to fight against the community transmission of the coronavirus (Zhao & Wu, 2020). Typically deprived of sufficient amounts of personnel and resources, members of residents' committees have had to mobilize volunteers and coordinate with other stakeholders (e.g., property management companies, homeowner associations, non-profit organizations inside and outside the neighborhoods, government agencies, etc.) to ensure effective neighborhood collaboration in the pandemic control. Therefore, it is important to examine, from the experiences and perceptions of community workers, the mechanisms and determinants of effective collaboration at the neighborhood level during the COVID-19 pandemic response in urban China. 2.2 Collaborative governance theory in the urban context The theory of collaborative governance was first proposed in the 1990s to help facilitate an understanding of the formation and operation of new modes of public policy making and service delivery (Agranoff & McGuire, 2001; Ansell & Gash, 2007; Emerson et al., 2011). Given the increasing complexity of policy challenges, such as environmental management and social-service delivery, governments have found themselves increasingly dependent on other agencies from the same or other tiers of government, as well as non-state sectors, such as private businesses and civil society, to solve policy problems (Leach, 2006; Lubell et al., 2002; Provan & Milward, 1995). Crisis events such as natural disasters or public health emergencies often involve a higher level of uncertainty and complexity that cannot be resolved by any single government agency, or even by the government alone, and this makes multi-stakeholder collaboration indispensable in crisis situations (Kapucu, 2012; Kapucu & Garayev, 2012; McGuire & Silva, 2010). Urban and policy scholars have also examined the emerging forms of multi-stakeholder collaboration or partnership in the urban context (Kapucu, 2012; Wang, 2016). However, urban governance that involves non-government stakeholders can be contentious or problematic rather than collaborative, leading to ineffective outcomes in neighborhood governance (Parés et al., 2017). Relatively few studies have systematically investigated the determinants of effective collaboration in urban neighborhood governance, particularly from local community workers' perspective (Li et al., 2019). Most empirical studies on collaborative governance have been based on Western contexts, where neighborhood governance has long been built on assumptions of clear boundaries between the state, market, and civil society. However, reforms of public service delivery in the past few decades have blurred these boundaries with cross-sector collaboration ensued (Lowery, 1998; Provan & Milward, 1995). Some local governments in the US, for instance, also initiated neighborhood governance reforms by setting up and sponsoring neighborhood associations or councils, which aimed at enhancing the links between city governments and residents in areas such as neighborhood planning, community development, and social service delivery (Chen et al., 2009; Fagotto & Fung, 2006; Li et al., 2019). In recent years, collaborative governance has been adopted to understand the administrative modernization process in China (Jing, 2015). Yet the extent to which urban neighborhood governance in China can be framed as collaborative governance has been subject to scholarly debates (Tomba, 2014). Nonetheless, persistent and active involvement of the state does not preclude collaborative dynamics between the state and community-level organizations, which has historical roots in the Chinese state governing the grassroots society through extended arms of local elites (Read, 2012). Neither does it preclude cross-sector collaborations in neighborhood governance that has emerged along with housing marketization and public service delivery (Wang, 2016; Wen, 2017). Based on the above discussion, we have adapted the collaborative governance theory to the empirical context of crisis governance in urban China, and developed a theoretical framework for understanding both the horizontal and hierarchical dynamics of multi-stakeholder neighborhood collaboration in response to the COVID-19 crisis. 2.3 Theoretical framework Fig. 1 illustrates our theoretical framework. Studies of collaborative governance in Western contexts have primarily focused on the horizontal mechanisms of collaborative networks in community service provision (McGuire & Silva, 2010; Provan & Milward, 1995). Collaborative governance, by definition, refers to horizontal interactions among public, private, and non-profit sectors (Ansell & Gash, 2007). Effective collaboration in neighborhood governance, therefore, would rely on informal mechanisms of trust, reciprocity, engagement, and negotiation with the aim of “making collective decisions” (Emerson et al., 2011). Recently, however, scholars have begun to acknowledge the role of hierarchical mechanisms in the formation and maintenance of horizontal collaborative governance for complex policy problems (Acar et al., 2008) — a role that is indicative of a form of “mandated collaboration” (Hafer, 2018; Rodríguez et al., 2007). Compared to Western contexts, top-down, hierarchical mechanisms play a more important role in neighborhood governance in urban China including, in particular, steering collaborative responses to the COVID-19 pandemic (Cheng et al., 2020; Zhao & Wu, 2020). Therefore, we have considered both the internal conditions (i.e., neighborhood social capital) and external factors (i.e., hierarchical steering by the government) that could determine the effectiveness of collaborative neighborhood governance in public health crisis responses.Fig. 1 Theoretical framework. Fig. 1 On the one hand, a virtuous cycle of interaction and engagement among stakeholders can provide the internal conditions for effective collaborative responses to a crisis incident. Social capital, which refers to the “stock” of trust, reciprocity, and civic engagement in a neighborhood (Putnam, 1995), has long been argued to facilitate collective actions and neighborhood governance efficacy (Sampson et al., 1997). Strong civic engagement can help build up a sense of shared purpose and identity, thus increasing the willingness and determination to collaborate across sectors to achieve a common goal (Cooper et al., 2006). This makes civic engagement key to successful collaborative governance (Emerson et al., 2011). The lack of civic engagement in public affairs can be the main barrier to the establishment and maintenance of a collaborative relationship in neighborhood planning and governance (Frieling et al., 2012). Previous experiences of successful cooperation can also create a high level of trust and social capital with which to produce a virtuous cycle of collaboration (Ansell & Gash, 2007; Kathi & Cooper, 2005). Moreover, citizen participation is vital if there are to be effective collaborative responses during emergencies in urban settings (Kapucu, 2012). On the other hand, simply focusing on the internal condition of horizontal collaboration within the neighborhood ignores the role played by political and other institutions outside the neighborhood. For example, local governments are deeply involved in the formation and functioning of local community partnerships or collaborations (Maloney et al., 2000). Top-down intervention by the government has played an important role in facilitating community participation in urban regeneration projects (Li et al., 2020). Furthermore, local governments can cultivate and control the strategic direction of collaborative partnerships through their own commitment and leadership when facing a policy challenge (Therrien & Normandin, 2020). They can also provide incentives to local stakeholders to collaborate by creating shared motivations and developing an institutional and procedural arrangement for collaboration (Hafer, 2018). Their key role in developing capacities for joint action with shared knowledge, expertise, and resources is widely acknowledged (Emerson et al., 2011; Therrien & Normandin, 2020). 3 Data and methods 3.1 Data In February 2020, at the height of the nationwide mobilization to control the COVID-19 outbreak, which had started in Wuhan City (Hubei Province) and quickly spread across the country, we conducted a large-scale questionnaire survey of community workers (i.e., chairs and members, as well as hired socialworkers in neighborhood residents' committees), selected from 20 sub-districts in six Chinese cities. The purpose of the survey was to understand the real-time experiences and perceptions of community workers regarding grassroots level mobilization and collaboration efforts during the critical stage of the pandemic control. At the time of the survey, it was impossible to conduct a probability sampling because most community workers were working around the clock and not available to participate in the survey, and because the stay-at-home policies prevented the authors from taking field research travels and conducting face-to-face interviews. Given these unique circumstances, we adopted a multi-stage, snowball sampling method to recruit participants for the online questionnaire survey. First, we purposefully selected six Chinese cities that were most affected by the COVID-19 outbreak outside Hubei Province.1 The six cities included four megacities, namely Beijing, Shanghai, Guangzhou, and Shenzhen, and two other cities that were most heavily hit by COVID-19 in their respective provinces, namely Wenzhou in Zhejiang Province and Nanyang in Henan Province. Though not statistically representative of all Chinese cities, these sample cities represent cities of diverse location, population size, economic structure, and administrative hierarchy. Moreover, they were all main destinations for Wuhan's out-flow population during China's Spring Festival, and thus faced the greatest risk of a COVID-19 outbreak in January–February 2020 (Table 1 ).Table 1 Characteristics of case cities. Table 1 Permanent population (10,000 persons) Pct. migrant population (%) Pct. population in-flow from Wuhan in total Wuhan's out flow population1 (%) Number of COVID-19 cases (as of March 19th 2020) Number of COVID-19 cases (per 10, 000 persons) Number of surveyed sub-districts Beijing 2153.60 36.88 0.88 480 0.22 4 Shanghai 2428.14 40.13 0.67 371 0.15 5 Guangzhou 1530.59 64.90 0.50 359 0.24 3 Shenzhen 1343.88 62.80 0.49 427 0.33 2 Wenzhou 830.55 32.11 0.21 504 0.54 4 Nanyang 1003.16 0.35 0.69 156 0.16 2 Note: The migration data is from Baidu Map (http://qianxi.baidu.com/), which records traveling data from January 10th (the beginning of China's Spring Festival) to January 23rd (Wuhan's lockdown). Second, we selected two to five sub-districts in different locations (inner city vs. suburbs) in each city. While following the stay-at-home orders, we managed to obtain access to a total of 20 sub-district governments, mostly through our contacts and key informants in the selected cities. Finally, assisted by sub-district government officials, we distributed the survey by sending specifically designated links to the online survey instrument to a maximum of 50 residents' committee staffers in each of the 20 sub-districts we surveyed. We acknowledge possible selection bias associated with the non-probability sampling method. This was the only feasible, though by no means most ideal, approach to respondent recruitment as we were trying to capture real-time experiences of frontline community workers during the most difficult time of COVID-19 responses in China. Nonetheless, we adopted several strategies to minimize the potential selection bias. First, we maximized the diversity of sub-districts in geographical location and demographic structure, selected from a diverse group of cities. Second, based on our preliminary interviews, there are typically 10–20 staff members in one neighborhood residents' committee. Therefore, we decided to recruit 50 respondents in each sub-district to ensure that our respondents came from a diversity of neighborhoods and represented community workers of different age cohorts and in different positions in residents' committees. Third, as a robustness check in the empirical analysis, we adopted weighted regression models to partially address the overrepresentation of megacity residents in the sample. The survey finally yielded 820 valid samples out of a total of 910 returned questionnaires, with a valid response rate of 90.1%. The majority of our survey respondents were female (65.0%), reflecting the overall gender composition of residents' committee staff. Moreover, 17.3% of the respondents were chair or vice-chair of their residents' committees, whereas 60% were hired social workers (see Table 2 for the socio-demographic profile of the respondents).Table 2 Summary statistics of sample structures (N = 820). Table 2Variable Value N Pct. Gender Male 287 35.0% Female 533 65.0% Age Mean = 37.9 (SD = 9.1) Position Residents' committee Chair/Vice-Chair 142 17.3% Residents' committee member 186 22.7% Ordinary workers 492 60.0% Neighborhood location City center 404 49.3% Suburb 416 50.7% City Beijing 192 23.4% Shanghai 206 25.1% Shenzhen 94 11.5% Guangzhou 110 13.4% Wenzhou 142 17.3% Nanyang 76 9.3% 3.2 Dependent variable: perceived effectiveness of collaboration Evaluating governance collaboration is challenging because of the difficulty in operationalizing and measuring the effects, impacts, or outcomes (Emerson et al., 2011; Li et al., 2019). Assessing the outcomes of neighborhood collaborative governance during the COVID-19 pandemic was even more difficult since the outbreak had not been fully contained at the time of our survey, and indeed the pandemic is still far from over in the world. Therefore, we did not attempt to assess the outcomes or impacts of collaborative governance (for instance, in terms of a reduced number of confirmed COVID-19 cases in the neighborhood). Rather, we chose to assess, from the perspective of community workers, the perceived effectiveness of the ongoing multi-stakeholder collaborative relationship within the neighborhood in terms of collectively containing the coronavirus during this unprecedented public health crisis. Our dependent variable in the study, perceived effectiveness of neighborhood collaboration, was captured by a question in the survey asking each respondent to rate, on a 0–10 scale, the effectiveness of neighborhood collaborative responses in containing COVID-19, with a score of zero representing least effectiveness, and 10 representing maximum effectiveness. We acknowledge that solely relying on the survey to capture perceived collaborative effectiveness may introduce measurement bias because community workers were likely to give favorable assessments to their work. While distributing the online survey instrument, we ensured that our respondents understood the anonymity of their answers by sending specifically designated links and by purposively omitting questions about the neighborhoods where they worked. Additionally, we followed previous studies (e.g., Li et al., 2019) and performed the Harman's single factor test to check the estimated variance for the models (Harman, 1976). The variance explained was 26.8%, suggesting the common source bias was not an issue with the independent and dependent variables. 3.3 Independent variables There are two sets of key independent variables: neighborhood social capital and hierarchical steering (refer to Table 3 for descriptive statistics).Table 3 Summary statistics of key independent variables. Table 3Variable name Variable type Mean SD Min Max Neighborhood social trust Ordinal (1–5) 4.40 0.76 1 5 Neighborhood civic engagement Ordinal (1–5) 4.30 0.82 1 5 Lack of community participation Binary (0/1) 19.9% / 0 1 Perceived priority Ordinal (0–10) 9.43 1.22 0 10 Perceived pressure Binary (0/1) 41.6% / 0 1 Lack of incentives Binary (0/1) 41.1% / 0 1 Perceived government support Mean value of three ordinal questions (1–5) 3.70 1.07 1 5 COVID case Binary (0/1) 38.9% / 0 1 Perceived uncertainty Binary (0/1) 20.2% / 0 1 Perceived difficult of virus control Ordinal (0–10) 8.66 1.83 0 1 Neighborhood social capital was captured by three variables: social trust, civic engagement, and community participation in COVID-19 responses. First, neighborhood social trust was measured by a Likert-scale question asking the respondent's level of agreement with the statement that “residents in my neighborhood maintain good trust and provide mutual assistance to each other”, with a value of 1 indicating total disagreement and 5 indicating total agreement. Second, neighborhood civic engagement, measured by a similar Likert-scale question, referred to the perception that “residents in my neighborhood care and actively participate in neighborhood public affairs”. Third, we included perceived lack of community participation, a binary variable measuring whether a respondent perceived a lack of participation by citizens and organizations in the neighborhood during the COVID-19 responses. We expected to find a higher level of perceived collaboration effectiveness alongside greater social trust and civic engagement in the neighborhood, and lower collaboration effectiveness alongside a perceived lack of community participation. Hierarchical steering was captured by four variables, namely perceived priority, accountability, incentives, and the support that neighborhoods received from the city government while fighting the COVID-19 pandemic. First, perceived priority referred to community workers' perceptions regarding the upper-level government's prioritization of the fight to contain COVID-19, with the score ranging from zero (indicating no priority at all) to 10 (indicating the highest level of prioritization). Second, perceived pressure reflected whether community workers perceived excessive pressure from government oversight and accountability. Third, lack of incentive referred to whether the survey respondents felt that insufficient incentives were being provided by the upper-level government. Fourth, perceived support reflected the perceived level of support which neighborhoods were receiving from the upper-level government to fight the COVID-19 pandemic. This was measured in terms of the mean value of three Likert-scale questions asking about the extent to which a respondent agreed that the neighborhood was receiving sufficient support from the upper-level government in relation to personnel, supplies, and public health professional assistance. 3.4 Control variables We first controlled for the possible effects of the perceived crisis level. Collaboration across boundaries becomes inevitable during emergency events, especially in urban settlements (Kapucu, 2012; Parker et al., 2020; Therrien & Normandin, 2020). A greater risk of COVID-19 transmission, as well as a higher uncertainty associated with the risk, can generate a greater willingness among multiple stakeholders to collaborate on finding collective solutions (Emerson et al., 2011). Therefore, we included three variables to capture the crisis level: whether there was any positive case reported in the neighborhood (COVID case), whether the respondent perceived the virus' spread to be too rapid to control (uncertainty), and whether there was a perceived difficulty in containing the virus in the neighborhood (difficulty), ranging from zero (not difficult at all) to 10 (extremely difficult). Finally, we controlled for the respondent's socio-demographic features, including age, gender, and position (1 = leader in the residents' committee), as well as the hierarchical status of the city (1 = megacity) and the location of the sub-district (1 = inner city) to investigate whether the perceived effectiveness of neighborhood collaborative governance varied across cities and locations. 4 Empirical findings 4.1 Descriptive statistics As shown in Fig. 2 , the average perceived effectiveness of neighborhood collaboration was 8.95, indicating a rather high level of perceived effectiveness among all community workers in our sample. The vast majority of the respondents (97.2%) described it as relatively effective (6–10), while 47.2% described the effectiveness as maximum (10). Note that the numbers may not reflect an objective assessment of the actual effectiveness of collaboration governance in urban neighborhoods in China. Nevertheless, the survey allowed us to compare frontline community workers' real-time perceptions of neighborhood collaborative governance across residential locations and city sizes.Fig. 2 Perceived effectiveness of neighborhood collaboration (0 = the least effective, 10 = the most effective). Fig. 2 As shown in Fig. 3 , no significant difference was found in the mean values of perceived effectiveness between community workers of inner-city neighborhoods and those of suburban neighborhoods. But perceived effectiveness is significantly different across six case cities, with megacities having overall lower levels of average perceived effectiveness than the other two cities. Nanyang had the highest level of perceived effectiveness (with a mean value of 9.26), while reporting one of the highest levels of neighborhood social trust, civic engagement, and government support but the lowest level insufficient incentives from the government (see Table 4 ). In contrast, respondents from Guangzhou reported the lowest level of perceived effectiveness (8.55; see Fig. 3), but also the lowest level of government support and the highest levels of perceived pressure and lack of incentives from the government.Fig. 3 Perceived effectiveness of neighborhood collaboration by city and by location. Fig. 3 Table 4 Comparing key independent variables among six case cities. Table 4Variable name NY WZ SH BJ SZ GZ F-statistics/Chi-square Neighborhood social trusta 4.54 4.24 4.45 4.47 4.28 4.38 2.75⁎⁎ Neighborhood civic engagementa 4.38 4.26 4.40 4.34 4.05 4.21 2.99⁎⁎ Lack of community participationb 15.8% 21.8% 17.5% 24.5% 14.9% 20.9% 1.20 Perceived prioritya 9.31 9.48 9.46 9.49 9.62 9.09 2.49⁎⁎ Perceived pressureb 34.2% 33.8% 33.5% 38.5% 60.6% 60.9% 8.89⁎⁎⁎ Lack of incentivesb 25.0% 40.1% 38.4% 40.6% 48.9% 52.7% 3.53⁎⁎⁎ Perceived government supporta 3.91 3.77 3.87 3.51 3.74 3.42 4.52⁎⁎⁎ a Reporting mean values and standard deviations and results from ANOVA tests. b Reporting percentages and results of chi-square tests. ⁎⁎⁎ p < 0.001. ⁎⁎ p < 0.01. An interesting contrast can also be found among the four megacities: respondents from Guangzhou and Shenzhen, both in Southern China, reported lower levels of perceived effectiveness than Beijing and Shanghai (Fig. 3), although Guangzhou and Shenzhen did not necessarily face a greater risk of COVID-19 spreading or population inflow from Wuhan (shown in Table 1). Further comparisons (see Table 4) showed that respondents in Beijing and Shanghai reported relatively higher levels of neighborhood social trust and civic engagement than respondents from Shenzhen and Guangzhou. By contrast, higher percentages of respondents in Shenzhen and Guangzhou perceived excessive pressure while insufficient incentives from the government. Nonetheless, these findings may not be conclusive and we further adopt multivariate regression analysis to determine to what extent these factors may explain the variance in the perceived effectiveness of neighborhood collaboration among community workers. 4.2 Regression results from the full-sample models We conducted multiple linear regression analyses to estimate the effects of neighborhood social capital and hierarchical steering on perceived collaborative effectiveness (refer to Table 5 for model results for the full sample analyses). Model 1 included only the two sets of independent variables, i.e., neighborhood social capital and hierarchical steering factors. Model 2 further included the three sets of control variables, namely crisis levels, respondent's socio-demographic features, and neighborhood locational features. As robustness checks, we ran weighted regression with weights proportional to the size of the permanent urban population to address the selection bias from non-probability sampling (results were shown in model 3). Additionally, we ran an ordinal logistic regression analysis, treating the dependent variable as an ordinal variable (model 4). We employed the cluster-robust estimator approach—with estimations of standard errors clustered by sub-district—to account for the nested nature of the survey data, in which individual respondents were clustered in the 20 sub-districts we surveyed. All models show largely consistent results, indicating an overall robustness of the empirical findings.Table 5 Regression results on the effectiveness of neighborhood collaboration across six cities in China. Table 5 Model 1 Model 2 Model 3 (weighted) Model 4 B S.E. B S.E. B S.E. B S.E. Neighborhood social trust 0.046 0.079 0.051 0.079 0.096 0.096 0.149 0.131 Neighborhood civic engagement 0.169⁎ 0.075 0.131 0.098 0.079 0.095 0.250 0.164 Lack of community participation −0.271⁎⁎ 0.096 −0.280⁎ 0.101 −0.266⁎ 0.118 −0.353⁎ 0.159 Perceived priority 0.477⁎⁎⁎ 0.032 0.449⁎⁎⁎ 0.075 0.470⁎⁎⁎ 0.065 0.851⁎⁎⁎ 0.124 Perceived pressure −0.136 0.083 −0.128 0.089 −0.143 0.088 −0.219 0.157 Lack of incentives −0.054 0.085 −0.035 0.086 −0.090 0.087 −0.168 0.16 Perceived government support 0.147⁎⁎ 0.044 0.168⁎⁎ 0.046 0.190⁎⁎⁎ 0.053 0.307⁎⁎ 0.106 COVID case 0.159+ 0.082 0.146+ 0.081 0.332⁎ 0.166 Perceived uncertainty −0.123 0.078 −0.144 0.098 −0.101 0.200 Perceived difficulty 0.082⁎ 0.032 0.088⁎⁎ 0.026 0.172⁎⁎⁎ 0.054 Age 0.015⁎⁎ 0.004 0.016⁎⁎⁎ 0.004 0.033⁎⁎⁎ 0.009 Female 0.000 0.055 −0.020 0.077 −0.095 0.104 Leader −0.056 0.123 −0.043 0.105 −0.179 0.225 Inner-city location −0.042 0.090 −0.078 0.080 −0.084 0.177 Megacity −0.370⁎⁎⁎ 0.087 −0.359⁎⁎⁎ 0.090 −0.739⁎⁎⁎ 0.174 Constant 3.111⁎⁎⁎ 0.348 2.418⁎⁎ 0.690 2.131⁎⁎⁎ 0.594 0.149 0.131 N 820 820 820 820 R2 0.338 0.379 0.402 / Log likelihood / / / −954.946 Note: Models 1–3 report results from multiple linear regression and model 4 report results from ordinal logistic regression, with standard errors clustered by sub-district in all models. ⁎⁎⁎ p < 0.001. ⁎⁎ p < 0.01. ⁎ p < 0.05. + p < 0.1. In all four models, coefficients for neighborhood social trust are positive but insignificant (Table 5). Civic engagement was also positive, but only significant on 0.05 level in model 1 (B = 0.169), when control variables were not included. Rather, a community worker tended to perceive less effective neighborhood collaboration in coronavirus responses if he or she perceived a lack of community participation in the joint efforts, with all control variables held constant (B = −0.280, p < 0.05, model 2). These findings indicate that, rather than simply forging neighborliness and trustworthiness among residents, having strong citizen engagement in public affairs, particularly in a crisis situation, is a key internal determinant to more effective joint efforts to fight a public health crisis. The finding also echoes previous observations that widespread mobilization of the general public was a critical part of China's early success in controlling the virus spread (China Watch Institute, 2020). Residents were mobilized to follow the health care guidelines (e.g., wearing masks, temperature checking) and comply with the quarantine or lockdown measures. In addition, community-based organizations and community volunteers provided residents' committees with critical personnel support and resources for residents' committee (Cheng et al., 2020). Among the four hierarchical steering variables, neither perceived pressure or lack of incentive significantly predicted more effective neighborhood collaboration in fighting the coronavirus pandemic: the coefficients were negative but insignificant (Table 5). This is interesting given that much of the policy debate has been centered around providing more incentives (i.e. monetary compensation or career development) for community workers along with sufficient oversight and accountability. Our survey data indicated that the incentive-accountability mechanism may not necessarily enable community workers to coordinate better horizontal collaboration in the neighborhood. Instead, hierarchical steering was more effective in facilitating collaborative governance in the neighborhood by setting up a clear policy priority and providing sufficient support to local community workers. All else equal, a one-unit increase in perceived priority contributed to collaborative response effectiveness by 0.449 (p < 0.001, model 2; see Table 5). A community worker also tended to perceived a higher level of collaborative effectiveness if he or she perceived to have received more support from the government (B = 0.168, p < 0.01, model 2). As we expected, perceived problem severity regarding the coronavirus transmission tends to contribute to more active collaboration among neighborhood stakeholders (Table 5). A community worker tended to perceive a higher level of collaborative effectiveness if there have been positive COVID-19 cases reported in the neighborhood (p < 0.1 in model 2 and model 3, and p < 0.05 in model 4), and if he or she perceived a greater challenge of the pandemic control facing the neighborhood (p < 0.05 in model 2, p < 0.01 in model 3, and p < 0.001 in model 4). Although perceived uncertainty of the virus spread was insignificant, the findings in general confirmed previous findings that greater risk perception associated with a crisis situation leads to more effective collaboration in responses (McGuire & Silva, 2010). Among other control variables, it is not surprising to find respondents of an older age tended to report a higher performance of neighborhood governance in the pandemic (Table 5). Furthermore, regression results confirmed findings from the descriptive analysis that, on a 0.001 significance level, community workers in megacities perceived a lower level of collaborative effectiveness during the COVID-19 responses than in regular cities, whereas the difference was not significant between inner-city and suburban locations. The difference by city size may result from the higher population density and mobility in megacities, which tended to complicate the responses to a pandemic crisis. It may also reflect the intense challenge in megacities, where more dynamic urban transformation and greater population heterogeneity within and across neighborhoods have made collaboration among diffused interests more difficult. 4.3 Geographical differences We further explored possible geographical heterogeneity with respect to the determinants of collaborative effectiveness in the neighborhood responses to COVID-19. Table 6 presents results of separate linear regression models for subsamples in megacities (model 5), inner-city neighborhoods (model 6), and suburban neighborhoods (model 7).Table 6 Results from separate regression models for megacity samples, inner city samples, and suburban samples. Table 6 Model 5 (megacity sample) Model 6 (inner-city sample) Model 7 (suburban sample) B S.E. B S.E. B S.E. Neighborhood social trust 0.083 0.079 0.052 0.112 0.038 0.119 Neighborhood civic engagement 0.152 0.133 0.101 0.165 0.114 0.116 Lack of community participation −0.202 0.128 −0.161 0.147 −0.329⁎ 0.113 Perceived priority 0.483⁎⁎⁎ 0.092 0.436⁎⁎ 0.111 0.480⁎⁎ 0.103 Perceived pressure −0.169 0.107 −0.162 0.149 −0.043 0.113 Lack of incentives −0.091 0.095 −0.041 0.119 −0.034 0.149 Perceived government support 0.222⁎⁎⁎ 0.048 0.211⁎⁎ 0.063 0.156⁎ 0.052 COVID case 0.103 0.100 0.303⁎ 0.096 −0.004 0.085 Perceived uncertainty −0.173 0.099 −0.168 0.125 −0.123 0.085 Perceived difficulty 0.097⁎⁎⁎ 0.042 0.021 0.037 0.154⁎⁎ 0.042 Age 0.017⁎⁎ 0.005 0.013⁎⁎ 0.003 0.016+ 0.008 Female −0.041 0.068 0.096 0.075 −0.090 0.064 Leader −0.067 0.132 0.108 0.114 −0.156 0.202 Inner-city location −0.106 0.108 / / / / Megacity / / −0.498⁎⁎⁎ 0.051 −0.217 0.120 Constant 1.222+ 0.618 3.044⁎ 1.016 1.637+ 0.810 N 602 404 416 R2 0.414 0.406 0.384 Note: Standard errors were clustered by sub-district in all models. ⁎⁎⁎ p < 0.001. ⁎⁎ p < 0.01. ⁎ p < 0.05. + p < 0.1. In all three models, perceived collaborative effectiveness was predicted by the same set of hierarchical steering variables across different geographical locations as in the full sample model. In other words, whether in a megacity, an inner-city neighborhood, or a suburban neighborhood, local community workers tended to perceived a significantly higher level of collaborative effectiveness in cases of higher perceived priority and government support, but not incentives or pressure (Table 6). It was noted that the effects of hierarchical steering factors exhibited greater significance to the collaboration effectiveness in megacity neighborhoods samples (p < 0.001, model 5) than in rest samples. On the other hand, model results in Table 6 also indicated heterogeneous effects of neighborhood social capital variables on collaborative effectiveness in different locations of neighborhoods. In fact, none of the neighborhood social capital variables was significant in the model for inner-city respondents, whereas lack of community participation was significant and negatively associated with collaborative effectiveness perceived by suburban respondents. It possibly indicated a more important role of neighborhood participation in forging better neighborhood governance in suburban areas. 5 Conclusions COVID-19, as the biggest public health crisis in a century, has presented a unique opportunity to revisit the functioning and performance of collaborative neighborhood governance in a crisis response situation. In this paper, we drew from the theory of collaborative governance and the literature on China's urban governance to investigate the determinants of neighborhood-scale collaborative efforts to control the pandemic in six Chinese cities. We focused on the real-time experiences and perceptions of frontline community workers, who played a key coordinating role in the community-level pandemic responses in China. We believed that this study contributed to the scholarly knowledge of the transitional urban governance in China, and more generally, to the international literature on the horizontal and hierarchical dynamics underlying neighborhood governance, particularly in a crisis situation. Our findings confirmed the role of civic engagement in the effective collaborative neighborhood governance, as have been argued in the existing literature (Cooper et al., 2006; Frieling et al., 2012). Active citizen participation is a critical component of neighborhood governance and key to its success in crisis responses. Our statistical analysis based on a multi-city survey also revealed that, in China, the effect of citizen participation was more salient in suburban locations than central locations to foster a collaborative neighborhood governance. Our empirical analysis also highlighted the important roles of hierarchical steering by public authorities in facilitating horizontal collaboration, which was not only a unique feature in China's neighborhood governance (Li et al., 2020; Wu, 2018) but also a factor that has attracted growing interest from Western scholars (e.g., Hafer, 2018; Maloney et al., 2000; Therrien & Normandin, 2020). Our survey research confirmed that frontline community workers would report greater effectiveness in neighborhood collaboration to fight against the pandemic if city and sub-district governments set up clear policy priorities and provided critical support. The effects of hierarchical steering were overtly seen in megacities where high residential mobility had already imposed greater challenges on forming collaborative partnerships within neighborhoods. The above findings provide important policy implications as well. First, although much scholarly knowledge has been produced on neighborly interactions and trustworthiness in urban China, civic engagement – the willingness of residents to engage in neighborhood public affairs – rather holds to be the key to more effective neighborhood governance. Local governments' community building campaigns should strive to provide opportunities, venues, platforms, and even techniques to encourage more citizen participation in collective decision-making within neighborhoods, thereby cultivating the level of “civic-ness” that can transform into active community participation in joint crisis responses. Government in suburban districts in particular need to foster civic engagement and citizen participation in neighborhood public affairs. Second, the state has an important steering role in neighborhood collaborative governance, particularly during crisis responses. But hierarchical steering should focus more on setting up a clear policy priority and providing sufficient support, while avoiding adding unnecessary oversight and burden on community organizations such as residents' committees. This is specifically important for the collaborative neighborhood governance in megacities where the public health challenge is relatively severe. This research has several limitations, largely because the travel restrictions during the pandemic prevented us from conducting in-depth field observations. We had to resort to the online questionnaire survey as an instrument for collecting real-time data as the grassroots mobilization was well underway in February 2020. Nevertheless, our measurement of the effectiveness of neighborhood collaboration may be superficial and potentially biased, and we were not able to control more objective measures of neighborhood characteristics in our model. We made our best efforts to conduct informal interviews with community workers before, during, and after the survey, as well as rely on media reports and our own participant observations while living through the pandemic in various Chinese cities, in order to both inform our research design and interpret model results. Nonetheless, our cross-sectional survey data can only help us determine the correlations of key variables. We hope that, with the pandemic largely under control and travel restrictions lifted in China, future research may focus on more in-depth interviews of both residents' committees and other stakeholders to understand how collaborative efforts were organized within neighborhoods and promoted by various internal and external conditions. CRediT authorship contribution statement Zhilin Liu: Conceptualization; Formal analysis; Funding acquisition; Investigation; Project administration; Writing - original draft & review. Sainan Lin: Formal analysis; Investigation; Visualization; Writing - original draft & review. Yue Shen: Formal analysis; Investigation; Visualization; Writing - original draft & review. Tingting Lu: Formal analysis; Funding acquisition; Investigation; Methodology; Writing - original draft & review. 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. Acknowledgement We would like to acknowledge the funding support from the 10.13039/501100001809 National Natural Science Foundation of China (42071208/42001175). This project also received funding support from the Institute for Urban Governance and Sustainable Development, 10.13039/501100004147 Tsinghua University (20201160065). 1 We deliberately chose not to include Wuhan, the epicenter of the outbreak, in our study. We thought that it would be unethical and unfeasible to impose on community workers the additional burden of participating in the survey while already being under enormous pressure to stop the community transmission of the virus in the city. ==== Refs References Acar M. Guo C. Yang K. Accountability when hierarchical authority is absent: Views from public–private partnership practitioners The American Review of Public Administration 38 1 2008 3 23 Agranoff R. McGuire M. Big questions in public network management research Journal of Public Administration Research and Theory 11 3 2001 295 326 Ansell C. Gash A. Collaborative governance in theory and practice Journal of Public Administration Research and Theory 18 4 2007 543 571 Benavides A.D. Nukpezah J.A. How local governments are caring for the homeless during the COVID-19 pandemic The American Review of Public Administration 50 6–7 2020 650 657 Bray D. Social space and governance in urban China: The danwei system from origins to reform 2005 Stanford University Press Stanford, Calif Bray D. Building “community”: New strategies of governance in urban China Economy and Society 35 4 2006 530 549 Brodkin E.Z. Street-level organizations at the front lines of crises Journal of Comparative Policy Analysis: Research and Practice 23 1 2021 16 29 Chen B. Cooper T.L. Sun R. Spontaneous or constructed? Neighborhood governance reforms in Los Angeles and Shanghai Public Administration Review 69 s1 2009 S108 S115 Cheng Y.D. Yu J. Shen Y. Huang B. Coproducing responses to COVID-19 with community-based organizations: Lessons from Zhejiang Province, China Public Administration Review 80 5 2020 866 873 32836448 China Watch Institution China’s fight against COVID-19 2020 China Watch Beijing Available at: http://www.chinadaily.com.cn/pdf/2020/Chinas.Fight.Against.COVID-19-0420-final-2.pdf Cooper T.L. Bryer T.A. Meek J.W. Citizen-centered collaborative public management Public Administration Review 66 2006 76 88 Dodds K. Broto V.C. Detterbeck K. Jones M. Mamadouh V. Ramutsindela M. …Woon C.Y. The COVID-19 pandemic: Territorial, political and governance dimensions of the crisis Territory Politics Governance 8 3 2020 289 298 Emerson K. Nabatchi T. Balogh S. An integrative framework for collaborative governance Journal of Public Administration Research and Theory 22 1 2011 1 29 Fagotto E. Fung A. Empowered participation in urban governance: The Minneapolis Neighborhood Revitalization Program International Journal of Urban and Regional Research 30 3 2006 638 655 Fang Y. Liu Z. Chen Y. Housing inequality in urban China: Theoretical debates, empirical evidences, and future directions Journal of Planning Literature 35 1 2020 41 53 Frieling M.A. Lindenberg S.M. Stokman F.N. Collaborative communities through coproduction: Two case studies The American Review of Public Administration 44 1 2012 35 58 Fu Q. He S. Zhu Y. Li S.-M. He Y. Zhou H. Lin N. Toward a relational account of neighborhood governance American Behavioral Scientist 59 8 2015 992 1006 Fu Q. Lin N. The weaknesses of civic territorial organizations: Civic engagement and homeowners associations in urban China International Journal of Urban and Regional Research 38 6 2014 2309 2327 Hafer J.A. Understanding the emergence and persistence of mandated collaboration: A policy feedback perspective of the United States’s model to address homelessness The American Review of Public Administration 48 7 2018 777 788 Harman H.H. Modern factor analysis 1976 University of Chicago press Chicago, IL He S. Homeowner associations and neighborhood governance in Guangzhou, China Eurasian Geography and Economics 56 3 2015 260 284 Hemphill L. McGreal S. Berry J. Watson S. Leadership, power and multisector urban regeneration partnerships Urban Studies 43 1 2006 59 80 Jing Y. The road to collaborative governance in China 2015 Springer Kapucu N. Disaster and emergency management systems in urban areas Cities 29 2012 S41 S49 32287822 Kapucu N. Garayev V. Designing, managing, and sustaining functionally collaborative emergency management networks The American Review of Public Administration 43 3 2012 312 330 Kathi P.C. Cooper T.L. Democratizing the administrative state - Connecting neighborhood councils and city agencies Public Administration Review 65 5 2005 559 567 Kim S.J. Bostwick W. Social vulnerability and racial inequality in COVID-19 deaths in Chicago Health Education & Behavior 47 4 2020 509 513 32436405 Leach W.D. Collaborative public management and democracy: Evidence from western watershed partnerships Public Administration Review 66 SI 2006 100 110 Li H. Wen B. Cooper T. What makes neighborhood associations effective in urban governance? Evidence from neighborhood council boards in Los Angeles The American Review of Public Administration 49(8) 8 2019 931 943 Li X. Zhang F. Hui E.C.-M. Lang W. Collaborative workshop and community participation: A new approach to urban regeneration in China Cities 102 2020 10.1016/j.cities.2020.102743 Li Z. Wu F. Tenure-based residential segregation in post-reform Chinese cities: A case study of Shanghai Transactions of the Institute of British Geographers 33 3 2008 404 419 Lin S. Wu F. Li Z. Social integration of migrants across Chinese neighbourhoods Geoforum 112 2020 118 128 Liu C. Empowered autonomy - The politics of community governance innovations in Shanghai China Public Administration Review 5 1/2 2008 61 71 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 Lotta G. Coelho V.S.P. Brage E. How COVID-19 has affected frontline workers in Brazil: A comparative analysis of nurses and community health workers Journal of Comparative Policy Analysis: Research and Practice 23 1 2020 63 73 Lowery D. Consumer sovereignty and quasi-market failure Journal of Public Administration Research and Theory 8 2 1998 137 172 Lu T. Zhang F. Wu F. The variegated role of the state in different gated neighbourhoods in China Urban Studies 57 8 2020 1642 1659 Lu X. Perry E.J. Introduction: The changing Chinese workplace in historical and comparative perspective Lu X. Perry E.J. Danwei: The changing Chinese workplace in historical and comparative perspective 1997 M. E. Sharpe Armonk, NY 3 20 Lubell M. Schneider M. Scholz J. Mete M. Watershed partnerships and the emergence of collective action institutions American Journal of Political Science 46 1 2002 148 163 Mallinson D.J. Cooperation and conflict in state and local innovation during COVID-19 The American Review of Public Administration 50 6–7 2020 543 550 Maloney W. Smith G. Stoker G. Social capital and urban governance: Adding a more contextualized “top-down” perspective Political Studies 48 2000 802 820 Maroko A.R. Nash D. Pavilonis B.T. COVID-19 and inequity: A comparative spatial analysis of New York City and Chicago hot spots Journal of Urban Health: Bulletin of the New York Academy of Medicine 2020 10.1007/s11524-020-00468-0 McGuire M. Silva C. The effect of problem severity, managerial and organizational capacity, and agency structure on intergovernmental collaboration - Evidence from local emergency management Public Administration Review 70 2 2010 279 288 Mei C. Policy style, consistency and the effectiveness of the policy mix in China’s fight against COVID-19 Policy and Society 39 3 2020 1 17 Migone A.R. Trust, but customize: Federalism’s impact on the Canadian COVID-19 response Policy and Society 39 3 2020 382 402 35039727 Parés M. Boada J. Canal R. Hernando E. Martínez R. Challenging collaborative urban governance under austerity: How local governments and social organizations deal with housing policy in Catalonia (Spain) Journal of Urban Affairs 39 8 2017 1066 1084 Parker C.F. Nohrstedt D. Baird J. Hermansson H. Rubin O. Baekkeskov E. Collaborative crisis management: A plausibility probe of core assumptions Policy and Society 39 4 2020 510 529 Provan K.G. Milward H.B. A preliminary theory of interorganizational network effectiveness - A comparative study of 4 community mental-health systems Administrative Science Quarterly 40 1 1995 1 33 Provan K.G. Veazie M.A. Staten L.K. Teufel‐Shone N.I. The use of network analysis to strengthen community partnerships Public administration review 65 5 2005 603 613 Putnam R.D. Bowling alone: America’s declining social capital Journal of Democracy 6 1 1995 65 78 Read B. Democratizing the neighborhood? New private housing and home-owner self-organization in urban China The China Journal 49 1 2003 31 59 Read B. Roots of the state: Neighborhood organization and social networks in Beijing and Taipei 2012 Stanford University Press Stanford, CA Rodríguez C. Langley A. Béland F. Denis J.-L. Governance, power, and mandated collaboration in an interorganizational network Administration and Society 39 2 2007 150 193 Sampson R.J. Raudenbush S.W. Earls F. Neighborhoods and violent crime: A multilevel study of collective efficacy Science 277 5328 1997 918 924 9252316 Shen Y. Chai Y. Kwan M.P. Space–time fixity and flexibility of daily activities and the built environment: A case study of different types of communities in Beijing suburbs Journal of Transport Geography 47 2015 90 99 Sudhipongpracha T. Poocharoen O. Community health workers as street-level quasi-bureaucrats in the COVID-19 Pandemic: Cases of Kenya and Thailand Journal of Comparative Policy Analysis: Research and Practice 23 2 2021 234 249 Therrien M.-C. Normandin J.-M. From policy challenge to implementation strategy: Enabling strategies for network governance of urban resilience Risk, Hazards & Crisis in Public Policy 2020 10.1002/rhc3.12192 Tomba L. The government next door: Neighborhood politics in urban China 2014 Cornell University Press Wang W. Exploring the determinants of network effectiveness: The case of neighborhood governance networks in Beijing Journal of Public Administration Research and Theory 26 2 2016 375 388 Wen Z. Government purchase of services in China: Similar intentions, different policy designs Public Administration and Development 37 1 2017 65 78 Wu F. China's changing urban governance in the transition towards a more market-oriented economy Urban Studies 39 7 2002 1071 1093 Wu F. Housing privatization and the return of the state: Changing governance in China Urban Geography 39 8 2018 1177 1194 Yan B. Zhang X. Wu L. Zhu H. Chen B. Why do countries respond differently to COVID-19? A comparative study of Sweden, China, France, and Japan The American Review of Public Administration 50 6–7 2020 762 769 Zhao T. Wu Z. Citizen-state collaboration in combating COVID-19 in China: Experiences and lessons from the perspective of co-production The American Review of Public Administration 50 6–7 2020 777 783	36471788	PMC9712009	NO-CC CODE	2022-12-02 23:21:31	no	Cities. 2021 Sep 24; 116:103274	utf-8	Cities	2,021	10.1016/j.cities.2021.103274	oa_other
==== Front Transp Res Part C Emerg Technol Transp Res Part C Emerg Technol Transportation Research. Part C, Emerging Technologies 0968-090X 1879-2359 The Author(s). Published by Elsevier Ltd. S0968-090X(22)00376-X 10.1016/j.trc.2022.103963 103963 Article Estimating on-board passenger comfort in public transport vehicles using incomplete automatic passenger counting data Roncoli Claudio a Chandakas Ektoras b Kaparias Ioannis c⁎ a Department of Built Environment, Aalto University, Espoo, Finland b LVMT UMR-T 9403, Ecole des Ponts, Université Gustave Eiffel, Champs-sur-Marne, France c Transportation Research Group, University of Southampton, UK ⁎ Corresponding author. 1 12 2022 1 2023 1 12 2022 146 103963103963 23 9 2021 20 11 2022 21 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 prevention of crowding inside buses, trams and trains is an important component of on-board passenger comfort and is central to the provision of good public transport services. In light of the COVID-19 pandemic and the associated significant reduction in public transport patronage and, more importantly, in passenger confidence, the avoidance of crowds by passengers and operators alike becomes even more critical. This is where the provision of information on on-board comfort becomes a necessity. The present study, therefore, proposes a new Kalman filter based estimation scheme for on-board comfort levels, employing historical and current (same-day) non-exhaustive Automatic Passenger Counting data, as well as Automatic Vehicle Locating measurements. The accuracy and reliability of the estimation is, then, evaluated through application to the tramway network of the French city of Nantes. The results suggest that the proposed method is able to deliver good estimation accuracy, both in terms of absolute passenger numbers, but also, more crucially, in terms of on-board comfort Levels of Service. ==== Body pmc1 Introduction The number of passengers on board a public transport vehicle is a prominent constituent factor of on-board passenger comfort and is, therefore, critical for both operators and passengers. It plays an important role in implementing control strategies and improving schedule adherence and is also a key determinant of the quality of service. In addition, knowledge about current and anticipated on-board volumes is key to preventing crowding and ensuring effective observation of social distancing in the post-COVID-19 reality. Indeed, keeping a social distance of 1 or 2 m at stops and stations, and, more crucially, on-board public transport vehicles, is likely to remain desirable by passengers, regardless of whether it is a legal or advisory requirement. Early evidence on public transport occupancy levels from cities around the world from the onset of the COVID-19 pandemic, unfortunately, shows that travellers’ confidence towards public transport may have been significantly dented (Transport Focus, 2021). And while it can be expected that mass vaccination and the decreasing virulence of the disease over time will restore some of this confidence in the long-run, it looks likely that much of the damage may be irreparable in the short- to medium-term, and that it may be a long time until passengers feel again comfortable travelling on public transport systems operating at or near capacity (Przybylowski et al., 2021). As a result of the effects of the COVID-19 pandemic, hence, obtaining information on on-board passenger comfort is now no longer just a desirable feature for operators and passengers (Zhang et al., 2017), but actually a necessary-one that can provide additional confidence to travellers and can, consequently, make a direct positive contribution to the economic viability and sustainability of public transport services (Transport Focus, 2021, Gkiotsalitis and Cats, 2021). Up until recently, due to the absence of the relevant enabling technology, the only way of obtaining the relevant data was through the conduct of exhaustive manual passenger counts. These would be carried out infrequently and would provide data of questionable accuracy, as they would only give a snapshot of the condition at the time of the survey. As a result, the key issue of estimating on-board comfort in real-time, which has now risen to prominence due to the COVID-19 effects, has received only little attention in the literature. Nevertheless, technological advances in data acquisition, transmission and storage have accelerated the development and implementation of automated data collection systems (Koutsopoulos et al., 2017, Koutsopoulos et al., 2019), with Automatic Passenger Counting (APC) being a prominent example. APC systems make use of a standard system architecture of sensors on-board the public transport vehicle, thus providing much more accurate estimates of the vehicle on-board loads. As such, APC allows operators to gain an insight into the number of passengers boarding and alighting at each station, and therefore helps them in long-term planning, but can also, more crucially, assist them in real-time operations. APC systems are therefore increasingly being installed on public transport vehicles in various cities around the world. However, the cost of each APC installation and maintenance is high, and as a result only a small subset of the vehicles can usually be equipped. The common practice of public transport operators in France, for example, is that roughly 10–20 % of the total fleet is equipped, especially when legacy vehicles are considered. This means that only partial knowledge on the loadings can be obtained, which makes real-time loading predictions a particularly complex problem. The aim of the present study is, hence, to propose an estimation method of on-board passenger loads, based on non-exhaustive (incomplete) APC datasets, in conjunction with Automatic Vehicle Locating (AVL) data, which is an Intelligent Transport Systems (ITS) feature that is installed almost as standard on public transport systems worldwide (Koutsopoulos et al., 2017, Koutsopoulos et al., 2019). The estimation is to be achieved per vehicle run, line stop, and time interval of a typical working day, in order to be suitable for use as input by operators in planning and management processes, particularly as concerns the provision of real-time crowding information to passengers in the post-COVID-19 era. The proposed method is then tested and validated using actual APC and AVL datasets from the tramway network of the French city of Nantes, covering a period of three months. The rest of the paper is structured as follows. Section 2 reviews the background of the study, as reported in the available scientific literature in the related fields of passenger comfort measurement and quantification and public transport passenger loading forecasting and estimation methods. Section 3, then, presents the new on-board passenger load estimation method, including the formal problem definition, models for boarding and alighting passengers, and the resulting Kalman filter based approach used for the estimation. Section 4 provides a description of the study area and datasets, defines the evaluation metrics and processes, and reports and discusses the results obtained. Section 5 presents the numerical results produced with the proposed estimation method, considering various performance metrics. Section 6, finally, concludes the paper and identifies areas of further work. 2 Background In order to establish the background of the present study, the relevant scientific literature is reviewed. This includes the topics of passenger comfort measurement and quantification, and passenger loading forecasting and estimation methods. These, then, lead to the identification of the research gap that the study addresses. 2.1 Passenger comfort measurement and quantification With the increase of public transport patronage in the years prior to the COVID-19 pandemic, passenger comfort had already become a major issue in scheduling and operating public transport services, as it was seen as being related to significant welfare costs (Haywood et al., 2017). Comfort evaluation is, actually, a multi-criteria assessment problem, as defined by European Standard EN13816 (European Committee for Standardization, 2002). According to Mohammadi et al. (2020), the comfort level on-board public transport vehicles can be broken down to five critical factors: thermal, vibration, noise, lighting and air quality. Furthermore, the level of comfort can vary with respect to the volume of on-board passengers, where in-vehicle comfort and crowding have a significant impact on passenger (and, hence, customer) satisfaction (Cox et al., 2006), which is subjectively evaluated. Consequently, the measurement of in-vehicle crowding can be performed using both subjective (perception-based) and objective (actually measured) metrics (Turner et al., 2005), and both have advantages and drawbacks. Subjective metrics, for instance, give a much more accurate idea of how passengers really perceive and rate their on-board experience. However, they are typically backed by very limited empirical evidence, and are also heavily influenced by external factors, such as geographical and cultural differences, which makes them difficult to assess, use, and generalise from (Li and Hensher, 2013). Objective metrics, on the other hand, may be much more difficult to relate to real passenger perceptions, but they can be much more easily measured and used as a common standard of what would constitute good or bad on-board comfort by most passengers. For example, Tirachini et al. (2012) evaluated a number of objective metrics, such as the density of standing passengers and the proportion of seats occupied on-board, and found that they represented a good approximation of what passengers actually experienced. Looking at some examples of past studies on the topic, Kroes et al. (2013) conducted qualitative and quantitative surveys in order to quantify and measure in-vehicle comfort in the Paris metro. The study provided a typology of passengers with respect to their attitude towards travel time and comfort, which was obtained from stated preference survey models exploring the willingness to wait for a next, less crowded, train in relation to relative crowding levels. On the other hand, Haywood and Koning (2013) looked at the relationship between in and vehicle comfort and seating availability, and by carrying out surveys with public transport users in Paris, found that passenger inconvenience increased with decreasing in-vehicle comfort, and that there was a non-linear trade-off rate between “comfortable” and “uncomfortable” travel time. They also concluded that passengers are less keen on trading-off travel time for greater comfort in the morning peak hour, likely due to the constraints of morning commuting trips (e.g. punctual arrival at the workplace). Similarly, Batarce et al. (2015) evaluated in-vehicle comfort on the basis of mixed stated and revealed preference data in Santiago, Chile, using discrete choice models, and found a twofold increase of the marginal disutility between a “low” density of 1 passenger/m2 and a “higher” density of 6 passengers/m2, linearly related with the travel time. Tirachini et al. (2017) built on that research to identify relationships between the crowding level, perceived comfort and security. This body of research motivates the definition of levels of in-vehicle comfort. For example, the US Transit Capacity and Quality of Service Manual (TCQSM) defines specific levels of on-board crowding from A to F, where the latter represents “crushing” loading levels (i.e., more than 5 standing passengers/m2) (US Transportation Research Board, 2013). On-board comfort, naturally, has direct implications on passenger demand and public transport operations. De Palma et al. (2015) discussed the necessity of distinguishing seating and standing as two different states of comfort and provided an analytical expression of the discomfort that can be employed in order to derive optimal timetables and tariffs. Several demand models followed the distinction of the two passenger states and provided specific algorithms to address the difference between them. For instance, Leurent and co-authors (Leurent and Liu, 2009, Leurent et al., 2013) built on previous models to formulate an integrated framework of transit assignment that considers comfort-related factors, such as train line capacity, vehicle passenger capacity and in-vehicle comfort. Trozzi et al. (2013) also provided a dynamic user equilibrium model for bus networks, considering capacity constraints at the vehicle level. In light of COVID-19 and the associated changes in passenger habits and comfort thresholds, some of the assumptions behind several of these studies will likely need to be revisited. The principles, however, remain the same and highlight the need for a reliable way of estimating crowding levels on-board public transport vehicles. 2.2 Passenger loading forecasting and estimation methods Several decades’ worth of research have extensively explored the topic of travel demand estimation and forecasting. A considerable body of literature has focused on vehicle traffic; this is comprehensively appraised by Vlahogianni and co-authors (Vlahogianni et al., 2004, Vlahogianni et al., 2014), with methods typically being categorised into parametric and non-parametric ones. More recent research has attempted to transfer several of these methods onto the public transport domain in order to estimate or forecast passenger demand in the short-term. Examples of approaches adopted in this respect include autoregressive integrated moving average (ARIMA) and generalised autoregressive conditional heteroscedasticity (GARCH) (Ding et al., 2018), neural networks (Tsai et al., 2009, Jia et al., 2019); Kalman filtering (Guo et al., 2014, Gong et al., 2014), random forests (Cheng et al., 2019), and deep belief networks (Bai et al., 2017). A related problem that has also received considerable attention has been the estimation and prediction of Origin-Destination (OD) flows and matrices for public transport, usually on the basis of passenger counts or Automatic Fare Collection (AFC) systems. Methods adopted include optimisation (Gur and Ben-Shabat, 1997, Liu et al., 2021), elasticity (van Oort et al., 2015), trip chaining (Wang et al., 2011, Li et al., 2011), Iterative Proportional Fitting (IPF) (Ji et al., 2015), clustering (Huang et al., 2020), Bayesian inference (Sun et al., 2021), as well as data fusion and Kalman filtering (Tao and Tang, 2019). Some research has also explored the nature and patterns of prediction and forecasting errors and has created inferential statistics models aiming to address them (Jung and Casello, 2020). The problem of estimating real-time on-board passenger loading, and consequently also passenger comfort, has received much less attention in the literature, however, primarily due to the lack of reliable data sources to date. Some research used manual passenger counts, such as for example, He et al. (2018), who developed a scheme employing Monte Carlo simulation, neural networks and Markov chains in order to more efficiently control bus air-conditioning systems in Beijing on the basis of the anticipated loading. Other research attempted to estimate on-board occupancy using WiFi, but with rather limited success, mostly due to the inability of probes to exclude WiFi-enabled devices outside the vehicle and to count passengers without WiFi-enabled devices on-board (Mikkelsen et al., 2016, Oransirikul et al., 2014). More successful attempts have been carried out using a combination of AFC and AVL data. For instance, the method by Zhang et al. (2017) first estimated the on-board load of buses on the basis of trip chaining analysis and a probability model, and then predicted it using an extended Kalman filter, with promising results in terms of prediction accuracy when applied to the bus network of the city of Shenzhen. The approach of Noursalehi (2017), on the other hand, made use of random forests and gradient boosting for predicting passenger arrivals and their destinations on the London Underground network, along with an online simulation performing transit assignment. Random forests and gradient boosting, along with some other supervised learning methods (namely neural networks and k-Nearest-Neighbours), were also compared in the study by Heydenrijk-Ottens et al. (2018) for the prediction of both long- and short-term on-board loading of trams in the Hague, with, again, promising results. Nevertheless, the main disadvantage of AFC is that in rail systems passengers are usually required to scan their smartcard at station entries and exits rather than on-board the vehicle, while in bus and tram systems they are usually only required to scan it when they board and not when they alight. As a result, a considerable amount of inference is needed in order to estimate on-board loads, which can make the process overly complex and can compromise accuracy. Sun et al. (2021) used vehicle dwell measured through AVL data to relate passenger flows to passenger activity and then formulate a Bayesian inference model to predict boarding and alighting flows, as well as passenger loads. APC systems can also make a difference, and several studies have made use of them recently. For example, Khomchuk et al. (2018) used a Bayesian estimation approach to predict train loads on the basis of real-time APC and historical data, which they validated on a simulated network, while Pasini et al. (2020) and Hu et al. (2020) both used neural networks to predict train loads in suburban Paris and the San Francisco Bay Area respectively based on temporal features and recent (same-day) previous measurements. Jung and Casello (2020) used AVL and APC data to examine transit ridership errors. Pasini et al. (2019) additionally experimented with time-series modelling and machine learning methods (specifically random forests and gradient boosting trees) and found that they were able to adequately consider the temporal irregularity of train services. Wang et al. (2021), on the other hand, developed a two-stage prediction process of bus passenger on-board loads, whereby an initial short-term prediction is effectuated using an adaptive Kalman filter, and a further prediction is made using support vector regression; the performance of the method was then evaluated on the bus network of the city of Suzhou. Finally, Jenelius (2020) used a number of methods (stepwise regression, lasso regression and boosted tree ensembles) in order to predict real-time car-specific on-board crowding on the Stockholm metro network on the basis of APC (on-board passenger counts estimated through weight measurements of the train cars) and AVL data and found that when considering real-time data, the prediction accuracy improved. APC systems, however, have two main limitations. The first limitation is that, due to the dynamic nature of the phenomenon observed (passengers entering and exiting buses, trams or trains), many of the enabling technologies (such as weight/pressure, optical or radar sensors) are unable to deliver high precision, which means that APC systems are often prone to downtime as well as measurement errors. The second limitation is that, as mentioned already, due to their high cost, APC systems would typically have very low penetration rates in a public transport fleet, and as such, they would only be able to deliver partial information. Several of the studies carried out so far have sufficiently addressed the first limitation (malfunctions), but have generally not dealt with the second one (the partial availability), having usually explicitly or implicitly assumed complete data availability. A notable exception to this has been the work of Jenelius (2019), who, extending their previous work (Jenelius, 2020), used their developed lasso regression approach to predict real-time on-board crowding on buses in Stockholm, taking into account the fact that only 20 % of the vehicles were equipped. The results suggested that run-specific load prediction improved as the target run approached the departure time from the station. While this approach is capable of delivering sufficiently accurate estimates, however, it requires a fairly extensive training and calibration phase every time it is applied on a new case study. 2.3 Summary Consequently, from the review of the available literature, two research gaps can be identified. The first one is that, despite real-time and short-term passenger comfort information having been identified as an important factor of passenger mode choice, particularly post-COVID-19, and even though several passenger comfort quantification models have been developed, a link with on-board loading estimation has not been made to date. The second is that the majority of studies having attempted to estimate or predict on-board loading on the basis of APC have assumed access to complete datasets, which, however, is unrealistic in practice. Therefore, the present study addresses these gaps by developing a Kalman filter based on-board passenger comfort estimation method. An advantage of the approach is that it is largely off-the-shelf: as opposed to data-driven methods, it does not require a substantial amount of preparatory work to be carried out (such as data collection, data processing, model fitting, etc.), and is capable of producing estimates as soon as the first measurement point becomes available and of subsequently improving the accuracy of these estimates as more data become available. The approach is described in the next section. 3 On-board passenger load estimation methodology This section presents the new estimation method of real-time in-vehicle comfort, as expressed by on-board passenger numbers, proposed by the present study. The overall estimation framework is described first, followed by the mathematical notation used and an outline of the modelling assumptions and conventions adopted. Dynamic models for representing boarding and alighting passengers are, then, formulated, and the observability of the proposed systems is analysed and assessed. The section is, then, concluded with the formulation of the proposed Kalman filter based estimation method. 3.1 Overall estimation framework The aim of the proposed framework is the estimation the on-board comfort. “Estimation” here refers to the computation of the on-board comfort level at the current station and time. This is different to “prediction”, which refers to the establishment of the on-board comfort level at a later station and/or at a future time point, and which lies beyond the scope of the present study. The proposed estimation method aims at taking advantage of the available measurements in terms of vehicle location and passengers. In particular, measurements originate from:• AVL systems, installed in all vehicles, providing the position of a vehicle in real-time (including the time when a vehicle stops at each station); and • APC systems, installed on a limited number of vehicles, providing the number of passengers on-board, as well as the numbers of boarding and alighting passengers at each station. This results in a situation where full passenger information is available for some vehicles, but no passenger information is available for the remaining vehicles, which should therefore be estimated. However, this setting is not desirable for developing a vehicle-based estimation method, i.e., a method that directly processes vehicle-based measurements to calculate vehicle-based estimates, since no meaningful relation can be assumed between the passenger load of a vehicle and that of preceding (or subsequent) vehicles; this does not allow relating the available APC measurements to the quantities that are to be estimated. On the other hand, a more reasonable way is to perform vehicle-based estimation by first estimating station-based quantities and then derive the vehicle-based quantities that are of interest. In fact, such “indirect” estimation is preferable, since, by casting the problem into an estimation problem of station-based quantities, partial passenger loading information is available for every station, i.e., provided when a vehicle with APC is at the station. Moreover, this allows formulating analytical (data-driven) models for station-based passenger dynamics, which have a clear physical meaning and are based on reasonable assumptions, resulting in a more rigorous estimation problem. For instance, it is reasonable to assume that the number of passengers arriving at a station does not exhibit strong fluctuations at a given time of the day and, except in exceptional circumstances, is not affected by a variation in the public transport schedule, such as, for example, a minor train delay. On the other hand, an unexpected change of a vehicle headway may strongly affect the number of passengers boarding a vehicle and will likely affect also all successive runs. There are multiple ways of modelling the passenger arrival, boarding, and alighting processes at a station. The very nature of the problem can result in some complex models characterised by several parameters, whose calibration will likely require the availability of vast quantities of data (e.g., Gur and Ben-Shabat, 1997, Liu et al., 2021, van Oort et al., 2015, Wang et al., 2011, Li et al., 2011, Ji et al., 2015, Huang et al., 2020, Sun et al., 2021, Tao and Tang, 2019). Here, simplified models are developed and employed. These are characterised by linear dynamics, which allow to represent the boarding and alighting processes at each station according to simplified, yet reasonable assumptions. These are, then, complemented by linear measurement models, which allow the incorporation of APC measurements, reformulated as station-based quantities, as well as historical information, obtained, for example, by pre-processing AVL and APC data from preceding days. The resulting models are therefore capable of assimilating real-time data, as well as historical data, resulting in a data fusion approach, which can be tailored to the data availability in order to achieve the best possible estimation performance. Based on the developed models, the estimation is performed by employing a Kalman filter (KF) (Kalman and Bucy, 1961, Anderson and Moore, 1979), which is an effective methodology for state estimation of linear systems in the presence of limited and/or noisy measurements. The KF is an optimal state estimator applied to a dynamic system that involves random noise and includes a limited amount of noisy real-time measurements. In particular, the KF and its variants have been successfully applied in several domains, including transport (see, e.g., Szeto and Gazis, 1972, Wang and Papageorgiou, 2005, Bekiaris-Liberis et al., 2016, Roncoli et al., 2016, Bekiaris-Liberis et al., 2016, Antoniou et al., 2010, Achar et al., 2020). To summarise, the proposed approach consists of the following basic components:1. a station-based data-driven model for boarding passengers; 2. a station-based data-driven model for alighting passengers, formulated in terms of alighting rates; 3. the utilisation of vehicle position information and vehicle-based passenger measurements, where the latter are provided by a limited amount of vehicles equipped with APC systems; 4. the use of a KF for the real-time estimation of station-based boarding passengers and station-based alighting rates; and 5. a conservation-of-passengers equation for calculating the vehicle-based passenger load for each operating vehicle. The different components are described in detail in the next sub-sections. 3.2 Problem notation, conventions, and assumptions A public transport network is modelled by a set of stations I and a set of lines L, whereby an individual station is indexed by i∈I and a specific line is indexed by l∈L. A single run of a public transport vehicle (train, tram or bus) along a certain line l is denoted j∈Jl, where Jl is the set of all runs along line l over an observation period, which is assumed being one full operational day. Here, dynamic models are considered, which are defined in the discrete-time domain, introducing a step size T (e.g. of the order of 30–120 s), where time is indexed by k, such that actual time t=kT. The following variables are defined: pj(k) number of passengers on-board run j at time.k bjr(k) number of passengers boarding run j during.(k-1,k] ajr(k) number of passengers alighting from run j during.(k-1,k] γjr(k) alighting rate of run j during.(k-1,k] wi,l(k) number of passengers on the platform at station i waiting to board a vehicle of line l at time.k ei,l(k) number of passengers entering station i platform to board a vehicle of line l during.(k-1,k] bi,ls(k) number of passengers boarding a vehicle of line l at station i during.(k-1,k] ai,ls(k) number of passengers alighting from a vehicle of line l at station i during.(k-1,k] γi,ls(k) alighting rate of vehicles of line l at station i during.(k-1,k] ηi,jr(k) binary variable indicating if run j is at (or departs from) station i during.(k-1,k] ηi,ls(k) binary variable indicating if a line l vehicle is at (or departs from) station i during.(k-1,k] βjr binary variable indicating if run j is equipped with APC providing passenger load information. βi,ls(k) binary variable indicating if a run of line l departing from station i during (k-1,k] is equipped with APC, providing passenger load information. φ binary variable indicating if historical data is used for estimation. Also, for any variable ω, its measured value is denoted ω¯, its value calculated on the basis of “historical” observations is denoted ω∼, and its estimated value on the basis of the proposed method is denoted ω^. The objective of the proposed method is to estimate the number of on-board passengers pj(k) for all runs over a certain period, by employing combined AVL and APC information, which is available from both historical and real-time data. It is assumed that the estimation algorithm runs on a daily basis, considering an “operational” day, which typically starts in the morning of a calendar day (usually at 4 or 5 AM) and finishes in the early hours of the next calendar day (usually at 1 or 2 AM). Therefore, historical data comprise any data originating from previous “operational” days (which have been appropriately aggregated and pre-processed – an example of such pre-processing is documented in Section 4.3), while real-time data are received and processed whenever available, assuming no communication delays. In developing the proposed estimation method, the following measurements are assumed to be available at any time k: • Real-time AVL information for all runs and at all stations, providing ηi,jrk,∀i∈I,j∈Jl,l∈L. • Real-time APC data for a limited number of runs J¯l⊂Jl, providing b¯jr(k), a¯jr(k), and p¯j(k), for j∈J¯l, ∀l∈L. This allows to assign βjr=1 if j∈J¯l and βjr=0 otherwise. • Historical information obtained by processing AVL and APC data available for the previous days, providing e∼i,l(k) and γ∼i,lsk,∀i∈I,l∈L. Before proceeding to formulate station-based models, a correspondence between station-based quantities and vehicle-based quantities is formulated by first introducing two assumptions, which are, in general, trivially satisfied for public transport networks, considering a reasonably sized time-step T (e.g., of the order of 30 s to 2 min), depending on the resolution of the data and the public transport mode considered. For example, tram systems tend to exhibit longer headways and could accommodate a larger time step compared to urban bus systems that would require a shorter step size. In the test case provided in the following sections, the time-step is set 60 s to match the resolution of the data used. Specifically:Assumption 1 There is only one runjoperating on linelthat departs from stationiduring time interval(k-1,k], i.e. (1) ∑j∈Jlηi,jrk=ηi,lsk,∀i∈I,k. Assumption 2 A runjcan depart from only one station during a time interval(k-1,k], i.e. (2) ∑i∈Iηi,jrk=1,∀j∈Jl,l∈L,k. These assumptions allow introducing the following relations for boarding and alighting passengers:(3) bjrk=∑i∈Iηi,jrkbi,lsk,∀l∈L,j∈Jl,k (4) ajrk=∑i∈Iηi,jrkai,lsk,∀l∈L,j∈Jl,k. These imply that by estimating station-based quantities bi,ls(k) and ai,ls(k), vehicle-based quantities bjr(k) and ajrk can then be directly calculated. Hence, in the following sub-sections, models for estimating the former quantities are presented. 3.3 Dynamic model for boarding passengers In order to estimate the number of passengers boarding a vehicle of line l at station i, bi,ls(k), a dynamic model is introduced for the number of passengers on the platform at a stop waiting to board a run of a specific line. This evolves according to the following dynamics:(5) wi,lk+1=wi,lk+ei,lk-bi,lsk. The following assumption is, then, introduced:Assumption 3 At the time that any vehicle operating on lineldeparts from stationi, all passengers waiting on the platform to travel on linelat timekwill board the vehicle during(k-1,k], i.e., (6) bi,lsk=ηi,lsk·wi,lk+ξi,lbk, where ξi,lb(k) is an unknown modelling error, which can be, for example, described by zero-mean Gaussian noise. Describing random variables as zero-mean Gaussian is a typical approach in filtering design, as this allows specifying such stochastic process solely by its mean and variance, which, despite not matching exactly the process modelled, are deemed sufficient statistics for filtering purposes. In this case, a KF is employed, which has been rigorously proven optimal under the assumptions of a linear model and Gaussian noise. Still, it has been shown that one can successfully use KF even when the noise is not Gaussian (as almost always the case in real life), and that makes KF the best linear filter (Simon, 2006). It should be noted that Assumption 3 is typically satisfied in public transport networks, where there are no passengers left behind. However, situations of extreme passenger congestion can occur in practice in some public transport networks during peak times, and in such cases the assumption does not hold. This, however, is not a limitation of the proposed model, but rather an inherent limitation of APC as a measurement technology. This is because APC is capable of capturing only the passengers that board a vehicle but is unable to provide any information on the actual demand of passengers (and/or any left-behind passengers). Substituting (6) into (5) leads to:(7) wi,lk+1=1-ηi,ls(k)·wi,lk+ei,lk+ξi,lbk. Since there is no available information on the number of passengers entering the platforms to board vehicles, ei,lk is treated as constant (or, effectively, slowly varying), being characterised by random walk dynamics,1 i.e.(8) ei,lk+1=ei,lk+ξi,lek, where ξi,le(k) is, for example, zero-mean Gaussian noise. Although this may seem a crude approach, such simplified dynamic model is widely used for model-based estimation in the absence of a descriptive dynamic model (e.g., Wang and Papageorgiou, 2005, Bekiaris-Liberis et al., 2016). The overall (deterministic part of) system (7)-(8) is next written in a compact state-space form by defining the state vector of the system as(9) xi,lb=wi,lei,l, whose dynamics evolve according to(10) xi,lbk+1=Abk·xi,lbk, where(11) Abk=Abηi,lsk=1-ηi,lsk101. System (10)-(11) is a linear-parameter-varying (LPV) system, where parameter ηi,ls(k) is assumed to be known (measured), as stated in Section 3.2. Available real-time measurements for system (10)-(11) are obtained from APC measurements, which are, however, available only when a run equipped with APC is leaving the station, i.e., when βi,lsk=1, where βi,lsk is calculated from measured quantities as(12) βi,lsk=∑j∈Jlβjr·ηi,jk. Following the rationale of Assumption 3 and (12), when an APC equipped vehicle of line l is at station i, we can treat the available measurement for bi,lsk as a (noisy) measurement for wi,lk; this is formulated as(13) zi,lwk=βi,lsk·wi,lk+ψi,lwk, where ψi,lwk is a measurement error in the form of zero-mean Gaussian noise. Nevertheless, as real-time data are available only at specific times, i.e., when a vehicle with APC is at or departs from a station (βi,lsk=1), there could be long periods for which no measurements are available. This may cause issues due to, for example, daily recurrent fluctuations of passenger arrivals, demand peaks, etc., which, if not “observed” via a measurement, may cause a deterioration of the estimation performance. In order to overcome this issue, it is proposed to employ also historical data to feed the estimator when real-time information is not available. In particular, availability of historical data is assumed in terms of the number of passengers entering the platform of a station to board a specific line e∼i,l(k), which can be obtained by processing AVL and APC data from previous days. The resulting measurement model reads:(14) zi,lek=φ1-βi,lrk·e∼i,lk+ψi,lek, where ψi,lek is the measurement error associated with historical data in the form of a zero-mean Gaussian noise and φ is a binary parameter indicating whether historical data are used (φ=1) or not (φ=0). To summarise, system (10)-(11) is complemented by associating an output vector zi,lb, described by the following (deterministic part of the) measurement model:(15) zi,lbk=zi,lw(k)zi,le(k)=Cbk·xi,lbk, where Cbk is obtained by known (measured) parameters as(16) Cbk=Cbβi,lsk,φ=βi,lsk00φ1-βi,lsk The noisy (measured) version of zi,lbk, which holds the passenger measurements that are available for estimation, either from real-time APC data or from historical data, is:(17) z¯i,lbk=w¯i,lke∼i,lk, where w¯i,lk is obtained from measured quantities available at any time step k as(18) w¯i,lk=∑j∈Jlηi,jr(k)·b¯jr(k). 3.4 Dynamic model for alighting passengers A second model for estimating the number of passengers alighting at any station of a line is now formulated. In this case, instead of modelling directly the number of alighting passengers, a relationship between the number of passengers on-board a vehicle and the number of passengers alighting at a station is introduced, namely:(19) γjrk=ajrkpjk,∀l∈L,j∈Jl,k. As previously stated, real-time estimation of vehicle-based variables is challenging, since any small disturbances in the schedules or passenger patterns may create an estimation bias that would be difficult to identify and correct. For this reason, the alighting rate is re-defined as a station-based variable, denoted by γi,lsk, as (from Assumption 2):(20) γi,lsk=∑j∈Jlηi,jrkγjrk,∀i∈I,l∈L,k. Variable γi,lsk represents the percentage of passengers on-board any vehicle operating on line l that alights at station i. Under the reasonable assumption that such value does not feature strong fluctuations in time (i.e. can be considered as slowly-varying), and in absence of a descriptive dynamic model, its dynamics is modelled as constant via a random walk, i.e.(21) γi,lsk+1=γi,lsk+ξi,lγk, where ξi,lγ(k) is, for example, zero-mean Gaussian noise. It should be noted that the deterministic part of system (21) is a linear time-invariant (LTI) system. Real-time measurements for system (21) are again assumed to be available when an APC-equipped vehicle of line l is at station i; this results in the following measurement equation:(22) zi,lγ¯k=βi,lsk·γi,lsk+ψi,lγ¯k, where ψi,lγ¯k is the measurement error associated with real-time data in the form of a zero-mean Gaussian noise. Similarly as for the boarding passenger model, since real-time data are available only at specific times, i.e. when a vehicle with APC is at or departs from a station (βi,lsk=1), the measurement data may be complemented by historical information that is fed to the estimator when there is no real-time information available. In this case, availability of historical data on the alighting rate γ∼i,ls(k) is assumed, which can be extracted by processing AVL and APC data from previous days. The resulting measurement model reads:(23) zi,lγ∼k=φ1-βi,lsk·γ∼i,lsk+ψi,lγ∼k, where ψi,lγ∼k is the measurement error associated with historical data in the form of a zero-mean Gaussian noise. Therefore, to complete model (21), defined for passenger alighting rate, the output vector zi,lγ is introduced, described by the following (deterministic part of the) measurement model:(24) zi,lγk=Cγ(k)·γi,lsk, where Cγk is obtained from measured quantities available at any time step k as:(25) Cγk=Cγβi,lsk,φ=βi,ls(k)φ1-βi,lsk The noisy (measured) version of zi,lγk, which holds all the passenger measurements that are available for estimation, either from real-time APC data or from historical data, is:(26) z¯i,lγk=γ¯i,lskγ∼i,lsk, where γ¯i,ls(k) is obtained from measured quantities available at any time step k as:(27) γ¯i,ls(k)=∑j∈Jlηi,jrk·a¯jrkp¯jk. 3.5 Observability of the proposed systems Before proceeding to design estimators for boarding and alighting passengers, the observability of the systems formulated in the previous sections is investigated. In order to support readers that may not be familiar with the concept of observability, some physical implications of the formal definitions of observability are provided first (e.g., Antsaklis and Michel, 2006, Liu et al., 2013). In simple terms, the observability property of a system guarantees that the dynamic evolution of its internal states (i.e., the states that are not directly measured) may be determined (observed) by measuring only some specific states (or, more generally, some outputs of the system). In particular, while dealing with real-time state estimation, observability is a property that guarantees that the state of a system, such as the boarding passengers or alighting rates, can be reproduced, in real-time in an unbiased way from the available (partial) measurements by use of an estimator, such as a KF. The observability of a system is usually studied employing certain algebraic conditions (see, e.g., Antsaklis and Michel, 2006), related in particular to the A and C matrices characterising the system. However, for time-varying or parameter-varying systems (as in the case of the models considered in this study), it may not be trivial to formally check and guarantee these conditions, since the parameters affect the system’s matrices in real-time. For this reason, an alternative graph-theoretic approach is employed, which allows studying the observability property of a system by looking into its structure, defined by the zero and non-zero elements of the A and C matrices (see, e.g., Liu et al., 2013, Lin, 1974, Reissig et al., 2014). Moreover, the study of the structural observability properties of a system is useful in order to determine under which measurement configurations a system is actually observable. It should be noted that structural observability is a necessary condition for observability, as it provides an intuitive way to the study of observability which, in practice, typically implies, indeed, system observability. However, the loss of observability of a structurally observable system may happen for some time intervals as a consequence of a combination of parameters that cause the elements of the A and C matrices to satisfy some specific conditions (e.g., Liu et al., 2013, Lin, 1974, Reissig et al., 2014). On the other hand, if no combinations of parameters guaranteeing (structural) observability exist, no estimator would be able to reconstruct the system state from the measured outputs. Thus, in practice, as it is also suggested from the estimation results in Section 5, structural observability implies a proper operation of an estimation scheme as the one presented here, even though the system may not always be formally completely observable at any time. In order to study the structural observability for the proposed systems, the structure matrices A and C are introduced, representing the patterns of zero and non-zero elements of system matrices A and C, respectively. A useful representation of such patterns is via the construction of graphs GAT,CT, which are shown in Fig. 1 for both the boarding passenger model, considering (10), (15), and the alighting rate model, considering (21), (24).Fig. 1 The graphs GAT,CT for patterns A and C that include matrices A and C, respectively, of system (10), (15) (left), and of system (21), (24) (right). Black circles relate to the process models and red circles relate to the measurement models. Dashed lines indicate that the edge may exist, depending on the condition of parameters listed next to it (from which the time dependence is omitted). The following condition for structural observability is considered (as per, for example, Liu et al., 2013, Lin, 1974): Condition 1: A linear system (A,C) is structurally observable if and only if: i) The graph GAT,CT contains no non-accessible vertex; and ii) the graph GAT,CT contains no dilation. Considering the definition stated above, it can be established that both systems generally satisfy the conditions for structural observability, that is that there exist combinations of parameters that guarantee observability. This can be demonstrated by assuming βi,ls = 1 (or, more generally, non-zero) and observing that, for both systems, all vertices can be accessed, while no dilation exists in the graphs. In addition, Condition 1 allows determining for which combination of parameter values the system is observable or when it may temporarily lose observability; this can be investigated by looking at the resulting graphs when some of the dashed edges are removed. In particular, the following claims are established:1. If historical data are utilised for the boarding passenger model (φ=1), when βi,ls(k)=0 the system is temporally only partially observable due to the non-accessibility of vertex wi,l, while vertex ei,l remains always accessible. 2. If historical data are not utilised for the boarding passenger model (φ=0), the system is temporally not observable when βi,ls(k)=0 due to the non-accessibility of both vertices. 3. For both previous cases related to the boarding passenger model, observability is fully restored when βi,ls(k)=1 (i.e., when a vehicle operating on line l equipped with APC stops at station i). 4. If historical data are utilised for the alighting rate model (φ=1), the system is always observable, since, irrespectively of the value βi,ls(k), one of the two dashed edges is present. 5. If historical data are not utilised for the alighting rate model (φ=0), the system is temporally not observable when βi,ls(k)=0, since no dashed lines are present. Thus, in practice, as will also be shown by the estimation results in Section 5, apart from the cases described above, in which observability conditions are not met (partially, i.e., only for some states, or completely, i.e. for all states), the structural observability property holds and implies, as a general rule, the proper operation of an estimation scheme like the one proposed by the present study. In fact, since the cases in which observability is lost are only temporary occurrences, at the time when observability is restored the estimation capabilities of the proposed scheme are again guaranteed. Finally, it is noted that, if neither historical nor real-time data are available at any time, the system would be unobservable. 3.6 Estimation method The KF algorithm that is employed to estimate boarding passengers and alighting rates using the models previously described is introduced here. The estimation equations for a KF are given by:(28) x^-(k)=Ak-1x^(k-1) (29) P-k=Ak-1P+k-1A(k-1)T+Q (30) Kk=P-kCkTR+CkP-kCkT-1 (31) x^k=x^-k+Kkz¯k-Ckx^k (32) P+k=I-KkC(k)P-k, where x^- and x^ denote, respectively, the a-priori (i.e. predicted) and a-posteriori (i.e. updated) estimates of variable (vector) x; z¯ is a (noisy) measurement of x; A and C describe the state-transition and observation models of x; P- and P+ are the a-priori (i.e. predicted) and a-posteriori (i.e. updated) estimated co-variance matrices; K is the optimal Kalman gain; and variables Q=QT>0 and R=RT>0 are tuning parameters that represent the (ideally known) covariance matrices of the process and measurement noise, respectively. Eq. (28) calculates the predicted (a priori) state estimate, i.e., the estimate of the system’s state considering the previous (estimated) state and the system dynamics, whereas (29) calculates the predicted (a priori) covariance, i.e., a measure of the estimated uncertainty of the prediction of the system’s state when employing only the system’s dynamics. Eq. (30) calculates the optimal Kalman gain K, i.e., the gain that minimises the residual error in the minimum mean-square-error sense. Finally, Eq. (31) calculates the updated (a posteriori) state estimate, accounting for the correction due to the available measurements, while (32) calculates the updated (a posteriori) estimate covariance, i.e., a measure of the estimated uncertainty of the prediction of the system’s state after measurements are taken into account. The algorithm is initialised as:(33) x^k0=μ (34) Pk0=H, where μ and H=HT>0 represent, in the ideal case where x^k0 is a Gaussian random variable, the mean and auto-covariance of x^k0 and Pk0, respectively. In particular, two separate KFs are implemented: one for the estimation of boarding passengers and one for the estimation of alighting rates. For estimating boarding passengers, an estimator for xi,lb is designed considering process model (10)-(11), measurement model (15)-(16), and employing measurements (17); moreover, initial values μ=0ei,l0T and H=I2×2 are considered. The estimator delivers estimates x^i,lb, from which w^i,l is to be extracted; the estimated boarding passengers can be derived from (3), (6) as:(35) b^jrk=∑i∈Iηi,jrk·w^i,lk,∀j∈Jl. For estimating alighting passengers, an estimator for γi,ls is designed, considering process model (21), measurement model (24)-(25) and employing measurements (26); initial values are set as μ=γi,ls,0 and H=1. The estimator delivers estimates γ^i,ls, from which γ^jr is calculated as:(36) γ^jrk=∑i∈Iηi,jrk·γ^i,lsk. In both cases, it is possible that the KF delivers negative estimates at some steps, which are physically unrealistic. This is handled here in a heuristic manner, by bounding, at each step, the resulting estimates to be non-negative, and then using the bounded value at the next iteration. Even though some more complex methods exist to deal with this issue (Simon, 2010), testing them here led to virtually identical results. Finally, in order to estimate the passengers on board of vehicle j, p^jk, a conservation-of-passengers equation is employed at each time step, of the form(37) p^jk+1=p^jk+b^jrk-a^jrk, where p^j0=0, i.e. the vehicle is empty at the beginning of the service. Combining (37) with (19) results in:(38) p^jk+1=1-γ^jkp^jk+b^jrk, which is calculated at each discrete time interval after estimates for b^jr and γ^jr are computed. The overall estimation methodology is illustrated in Fig. 2 .Fig. 2 The proposed estimation scheme. 4 Data acquisition and processing The developed estimation method for real-time on-board passenger loads on the basis of AVL and APC data is applied on a real public transport network in this study, and this section sets out the core principles and methods used in that respect. The study area and dataset are introduced first and are followed by an outline of the data cleansing and processing tasks and by a description of the assimilation of the historical dataset used in the estimation. Finally, a brief description of the in-vehicle comfort measurement framework used is provided. 4.1 Study area and dataset The present study focuses on the tramway system of the French city of Nantes. Nantes is located on the Loire River in Western France, close to the Atlantic coast. It is the sixth largest city of France, with a metropolitan population of 900,000. Its tramway network is operated by Semitan, and with its opening in 1985 Nantes became the first city to introduce a modern generation tramway, built from scratch. With its subsequent extensions, the network now consists of three tramway lines (numbered 1, 2 and 3) running on 44 km of track and serving a total of 83 stations, as well as a “Busway” Bus Rapid Transit (BRT) line (numbered 4). The Nantes tramway is shown in Fig. 3 . Line 1, shown in green, has a length of 18.4 km and serves 34 stations. It consists of two branches at each end (Beaujoire and Ranzay in the East, and François Mitterand and Jamet in the West) and a central trunk between the branches with 19 stations. Its frequency reaches 15 vehicles per hour during peak times, and it is the busiest line on the network (and with 120,000 passengers per day, it is also one of the busiest of the whole of France), serving several principal locations of the city, including the main railway station and the city’s stadium. Line 2, shown in red, runs from Orvault in the North to Gare de Pont-Rousseau in the South, has a length of 11.7 km and serves 25 stations, including important educational (university) and health establishments. It has a frequency of 8 vehicles per hour during peak times, and its patronage approaches roughly 80,000 passengers per day. Lastly, Line 3, shown in blue, runs from Marcel Paul in the North to Neustrie in the South, has a length of 14.1 km and serves 34 stations. It has a similar operation with Line 2, with which it shares the track for seven stations (Commerce to Gare de Pont Rousseau) in the city centre. It serves several major commercial sites and is used by 75,000 passengers per day. The three lines run radially off the city centre but meet at Commerce. They are combined with Park and Ride (P + R) facilities on the outskirts, and also have major transfer points with the other public transport modes: the Busway (exclusive right-of-way line), the Chronobus (buses with limited segregated lines), the local buses and the regional coaches.Fig. 3 The Nantes tramway network (Source. https://www.tan.fr). The tramway system is served by three types of rolling stock, irrespectively of the line: the Alstom TFS, the Bombardier Incentro, and the CAF Urbos. The Alstom TFS is a 39 m long vehicle with a capacity of 236 passengers (including 74 seats) which began operation in 1985. Each Alstom vehicle is composed of two high floor carriages with three-step accesses (of which one mobile step) and a lower floor carriage in the middle; access is provided by six double length doors and two simple doors per vehicle side. The Bombardier Incentro is 36 m long with a capacity of 252 passengers (including 72 seats) and started operating in 2000. It has an integral low floor and six double (1.30 m) doors per side. Finally, the CAF Urbos is the newest vehicle in the network, having started operations in 2012. It is 37 m long with a capacity of 249 passengers (including 68 seats) and has an integral low floor and six double doors per vehicle side. The data used in this study have been collected from the Opthor and Ineo systems, used by the operator. Opthor is an APC system measuring the number of passengers boarding and alighting at each station, as well as the dwell time and other performance-related measures. The system detects passengers using infrared sensors installed at each door of the vehicle with a 95 % measurement accuracy. Ineo, on the other hand, is an AVL system that enables comprehensive real-time tracking of each tramway vehicle, and can therefore deliver data on the actual arrival and departure time of each vehicle at each stop. All vehicles circulating in the network are equipped with Ineo; however, the operator has only installed Opthor on a limited number of vehicles (of Alstom TFS and Bombardier Incentro class only) and uses it to run frequent counts with it for various purposes. 4.2 Dataset processing and cleansing The passenger data employed here were collected during an Opthor count conducted between 16 September and 20 December 2019 on Line 1 of the Nantes tramway network. As part of the count, the operator collected 52,350 valid entries from 42 weekdays and a total of 1946 runs. This sample corresponds to roughly 10 % of the total number of daily runs. Each data entry reports the total number of boarding and alighting passengers (as detected by the APC sensors), as well as the number of passengers on-board. The Opthor data are, then, complemented by the corresponding Ineo dataset for the same period. Fig. 4 illustrates the number of runs covered by AVL and APC data throughout the study period; days with no APC data availability or AVL malfunctions have been discarded.Fig. 4 Daily APC and AVL coverage during the study period. A two-step process is performed to clean the AVL data and merge them with the APC relevant data. First, the AVL data are processed in order to remove erroneous data and fill missing data. This cleaning process involves correcting entries from inaccurate geo-localisation, namely:• removal of double entries of a station for a single run; • addition of missing stops, including terminal stops, to specific runs, and inference of travel times through distance-based linear interpolation between known stops; and • removal of runs that do not provide service coherence. Consequently, out of 20,360 commercial runs throughout the study period monitored by the AVL systems, 18,672 valid runs are kept. The APC and the AVL systems are operated independently and use different data identifiers (station and run IDs, as well as timestamps), and as such an exact matching of APC runs with the “clean” AVL runs does not exist. Therefore, a run-matching process by approximation is employed in order to ensure as good a match as possible between the datasets, as the second step of the process. The method relies on a matching algorithm employed for each day and for each origin station throughout the day, which involves conducting a matching test for each AVL run with respect to selected “matching candidate” APC runs. According to the algorithm, a match between an AVL run and an APC run is established if the AVL and APC runs share the same stops, and:• the departure times of the AVL run and the APC run (either the theoretical/scheduled departure time or the actual departure time) are within 2 min of each other; or • the departure times are within 2–15 min and the APC run is the one nearest to the AVL run. If the difference between the departure times of the APC and AVL runs is greater than 15 min, the match is considered invalid and the relevant APC run is discarded. The parameters used in the run-matching process are justified as follows. Firstly, the 2-min threshold value determining immediate run-matching is based on the usual headway between runs at peak-time, as per the present dataset and the authors’ experience with other similar datasets. Then, the reason why both theoretical/scheduled and actual departure times from the AVL data need to be considered is the fact that the actual departure time may be inaccurate due to a vehicle dwelling at the origin station between runs without the AVL detectors on the ground distinguishing between its arrival (in the previous run) and departure (in the next run). Finally, the 15-min threshold is used in order to allow matching an APC run with the next recorded AVL run, if the latter is affected by lower geo-localisation precision accuracy – which can be a common source of error in AVL data. Out of the 1946 initial APC runs from the entire study period, 29 (1,5%) are discarded by the run-matching process, leaving 1917 APC runs matched with an AVL run. Of the matched runs, 95 % have a time difference of up to 1 min, with only 1 % greater than 5 min, as illustrated in Fig. 5 .Fig. 5 AVL-APC time difference distribution for matched runs. 4.3 Assimilation of historical data The methodology presented in Section 3 requires availability of historical data, which can be determined by processing AVL and APC data from days prior to the one for which estimation is carried out. In particular, historical data are required in order to compute e∼i,l(k) and γ∼i,ls(k), i.e. the number of passengers entering the platform of station i to board a vehicle of line l and the alighting rate of vehicles of line l at station i, as well as ei,l0 and γi,ls,0, i.e. the initial values for the estimates applied at the beginning of an operational day. The historical data extracted from the APC counts may vary with respect to the day of the week. As suggested previously, historical data observations are not homogenously distributed throughout the days of the week. In addition, a number of days (specifically Tuesdays) have been discarded from the initial dataset due to technical issues and major incidents. Furthermore, some variance is observed at the within-day profile of the number of boardings per stop, run and hour (Fig. 6 ). For example, Friday has a higher afternoon peak and the service is extended at night, whereas Wednesday has a higher midday peak. These differences suggest only broad day-to-day regularity and no within-day regularity, which is consistent with the mobility patterns of French urban areas (De Solere, 2012).Fig. 6 Observations per day of the week (left) and average volume of boardings per stop and run per hour and per day of the week (right). Consequently, two strategies for calculating historical data are employed here: a) aggregating data from the same weekday; and b) aggregating all available days together. Regardless of the aggregation strategy, historical data are averaged and aggregated into a set of bins, each of a 30-min duration, that include all APC measurements available during such time. As the operational day starting at 4:00 AM and ending the next day at 3:59 AM is considered, a total of 48 30-min bins are used. The historical value e∼i,lk∼, i.e. the historical value of the number of passengers per interval T entering the platform of station i to board a vehicle of line l during (k∼T∼,k∼+1T∼-1], is then calculated as(39) e∼i,lk∼=∑d∈D∑k=k∼T∼k∼+1T∼-1b¯i,l,dd(k)ti,l,dd(k)∑d∈D∑k=k∼T∼k∼+1T∼-1βi,l,dd(k), where k∼ is the bin index; T∼ is the bin size (in this case, 30 min); D is the set of days considered, which, depending on the chosen strategy, may include all available previous days or just a subset of them (e.g. only the days corresponding to the same weekday, as per case a defined above); bi,l,dd(k) is the number of passengers boarding a vehicle of line l at station i during k-1,k in day d; ti,l,dd(k) is the time elapsed between when the previous vehicle of line l stopped at station i and current time k; and βi,l,dd(k) is binary variable indicating if a run of line l departing from station i during (k-1,k] on day d provides passenger load information. Correspondingly, γ∼i,ls(k) is calculated as(40) γ∼i,lsk∼=∑d∈D∑k=k∼T∼k∼+1T∼-1γ¯i,l,dd(k)∑d∈D∑k=k∼T∼k∼+1T∼-1βi,l,dd(k), where γ¯i,l,dd(k) is the alighting rate of vehicles of line l at station i during (k-1,k] in day d. 4.4 Passenger comfort levels The Level of Service (LOS) framework of Chandakas (2009) is used to measure in-vehicle comfort on-board a tramway vehicle, which assigns a comfort level on the basis of the on-board density of standing passengers (D) and of the availability of occupied seats (R), as shown in Table 1 .Table 1 In-vehicle comfort LOS (Chandakas, 2009). LOS Conditions Description 6 D>4 Density greater than 4 standees per m2, exceptional situation 5 1.8<D≤4 No tolerance from standees, reliability issues 4 0<D≤1,8 Standing conditions tolerated 3 0.5<R≤1 Possibility to travel seated, limited choice available 2 0.25<R≤0.5 Possibility to choose a seat, constrained to availability 1 R≤0,25 Free choice of a seat The in-vehicle comfort level is matched to specific locations (or stops) at specific times of the day. Fig. 7 illustrates the average LOS of the eastbound (left) and westbound (right) Line 1 per time step of the historical data. LOS 4, where standing takes place in tolerable conditions, is observed in the densest section of the central trunk of the line throughout the day, but extends to the start and end of the line during the morning, midday and afternoon peak hours. LOS 5 conditions are also observed sporadically during peak times.Fig. 7 In-vehicle comfort LOS for eastbound (left) and westbound (right) Line 1 of the Nantes tramway (the direction of the run is from bottom to top). 5 Evaluation and validation This section documents the validation and evaluation process and results of the developed estimation method for real-time on-board passenger loads on the basis of AVL and APC data, based on its application on the Nantes tramway network. The evaluation metrics and methodology employed are presented first, and then the results obtained are reported and discussed. 5.1 Evaluation metrics and methodology The proposed method is particularly aimed at estimating the passenger load for the runs that do not feature an APC system. However, as information is not available for such runs, an alternative validation approach is developed, which employs only data from runs equipped with APC. Specifically, when the estimation algorithm processes an APC-equipped run, estimation is first performed assuming that no APC measurement is available, denoting such estimate ωˇ (for a generic variable ω) and storing the respective estimated values; then, the estimates are re-calculated considering the available APC measurements, and these latter values are then used for the subsequent estimation steps. In this way, the estimation method can be evaluated in a fair manner through comparison of the estimates obtained without considering APC measurements (ωˇ) with their corresponding actual measurements (ω¯). In order to evaluate the estimation performance and assess its effectiveness, a series of experiments are conducted and different metrics are calculated. First, the main components of the proposed methodology are evaluated, namely the boarding and alighting estimators described in 3.3, 3.4. The mean absolute error (MAE) metric is employed in this respect, defined as:(41) MAE=∑k=0K∑i∈I∑l∈Lωˇi,lsk-ω¯i,ls(k)∑k=0K∑i∈I∑l∈Lβi,ls(k), where K is the estimation horizon, i.e. an operational day; and ωi,ls represents the variables considered for evaluation, which, in this case, are bi,ls(k), γi,ls(k), and (indirectly) ai,ls(k). In addition, the weighted mean absolute percentage error (WMAPE) is employed as a relative metric, defined as:(42) WMAPE=∑k=0K∑i∈I∑l∈Lωˇi,lsk-ω¯i,ls(k)∑k=0K∑i∈I∑l∈Lω¯i,ls(k), This metric overcomes the infinite error issue of other relative metrics (such as the mean absolute percentage error, MAPE); however, it is still expected to suffer from the limitation of over-penalising cases when real values are low. The focus then shifts to the estimated passenger comfort, which is calculated according to the method described in Section 4.4, assuming such information is available every time a vehicle stops at a station. In this case, the sequence of estimated comfort levels (again, assuming no APC measurement was available) is compared with the sequence of comfort level calculated from actual measured data. In this comparison, besides investigating aggregated measures such as the mean absolute comfort error, more disaggregated information is further considered, such as the distribution of errors for different error classes. 5.2 Results and discussion The performance of the individual components of the estimation method is first investigated, i.e. the boarding and the alighting rate estimators. The estimation results for the estimator of boarding passengers are presented in Table 2, Table 3 . As can be observed, the estimation MAEs and WMAPEs for all days investigated are overall low; all MAEs are lower than 10 passengers and most WMAPEs remain below 1 (bearing in mind that they are also affected by the low actual values). Moreover, it can be seen that the performance of the estimator improves by incorporating historical data, with a maximum improvement of the order of 25 % on Friday 20 December 2019 in both the MAE and the WMAPE values. In addition, it can be observed that the case where the estimator is fed with historical data from all days outperforms the case that employs historical data only from the same weekdays. This suggests that the assumption that different days may have different peak hour patterns (which is the reason why the use of historical data only from the same weekdays was proposed in the first place) is uninfluential and that its effects are most probably outweighed by the fact that greater data availability allows obtaining better results. A notable exception is Thursday 12 December 2019, which is also characterised by considerably higher MAE and WMAPE values than the other days; a deeper investigation into that day, though, has revealed that a service disruption lasting about 40 min that happened during the afternoon has caused the estimator to produce some unreasonably high errors.Table 2 MAE for the estimated number of boarding passengers, employing different historical data settings. Day No historical data (φ=0) Historical data only from same weekdays Historical data from all days Wed 11 Dec 6.411 5.768 5.320 Thu 12 Dec 9.100 9.484 9.327 Fri 13 Dec 6.282 6.422 5.696 Mon 16 Dec 7.050 6.184 6.147 Wed 18 Dec 5.458 5.214 4.351 Thu 19 Dec 7.071 7.094 6.497 Fri 20 Dec 7.944 6.696 5.914 Table 3 WMAPE for the estimated number of boarding passengers, employing different historical data settings. Day No historical data (φ=0) Historical data only from same weekdays Historical data from all days Wed 11 Dec 0.881 0.803 0.733 Thu 12 Dec 1.209 1.217 1.271 Fri 13 Dec 0.726 0.689 0.621 Mon 16 Dec 0.959 0.950 0.935 Wed 18 Dec 0.773 0.760 0.658 Thu 19 Dec 0.854 0.735 0.801 Fri 20 Dec 1.033 0.835 0.756 The performance of the alighting passengers’ estimator is demonstrated in Table 4, Table 5 , where a similar pattern as the one observed for the boarding passengers’ estimator can be observed. On the other hand, the MAE and WMAPE values calculated for the alighting rates without using historical data already look reasonably low. Still, even in this case the use of historical data improves the estimation performance even further and, again, best results are obtained when all days are used to assimilate the historical data.Table 4 MAE for the estimated alighting rate and number of alighting passengers, using different historical data settings. Day Alighting rate Alighting passengers No historical data (φ=0) Historical data only from same weekdays Historical data from all days No historical data (φ=0) Historical data only from same weekdays Historical data from all days Wed 11 Dec 0.106 0.099 0.092 7.819 7.810 7.816 Thu 12 Dec 0.091 0.089 0.075 7.610 7.613 7.642 Fri 13 Dec 0.092 0.083 0.078 9.155 9.145 9.149 Mon 16 Dec 0.099 0.085 0.077 8.022 8.016 8.041 Wed 18 Dec 0.092 0.086 0.078 7.613 7.632 7.618 Thu 19 Dec 0.079 0.073 0.071 8.594 8.611 8.612 Fri 20 Dec 0.090 0.090 0.076 8.689 8.682 8.678 Table 5 WMAPE for the estimated alighting rate and number of alighting passengers, using different historical data settings. Day Alighting rate Alighting passengers No historical data (φ=0) Historical data only from same weekdays Historical data from all days No historical data (φ=0) Historical data only from same weekdays Historical data from all days Wed 11 Dec 0.633 0.573 0.555 0.940 0.945 0.946 Thu 12 Dec 0.552 0.475 0.496 0.945 0.956 0.946 Fri 13 Dec 0.550 0.493 0.470 0.939 0.934 0.934 Mon 16 Dec 0.608 0.536 0.569 0.944 0.956 0.956 Wed 18 Dec 0.608 0.555 0.525 0.944 0.962 0.951 Thu 19 Dec 0.558 0.517 0.514 0.912 0.921 0.917 Fri 20 Dec 0.570 0.536 0.504 0.937 0.937 0.933 The results in terms of passenger comfort (absolute) errors are investigated next, calculated as described in Section 4.4. The results are shown in Table 6 , where it can be observed that the resulting comfort levels obtained via the estimation scheme are accurate, with an absolute error lower than one comfort level for all examined scenarios, except one (the service disrupted impacted Thursday 12 December 2019, without using historical data). From these results, the value of employing historical data in addition to real-time measurements is highlighted again.Table 6 Absolute comfort level error employing different settings for historical data. Day No historical data (φ=0) Historical data only from same weekdays Historical data for all days Wed 11 Dec 0.859 0.744 0.734 Thu 12 Dec 1.031 0.715 0.712 Fri 13 Dec 0.694 0.630 0.588 Mon 16 Dec 0.837 0.651 0.692 Wed 18 Dec 0.864 0.710 0.646 Thu 19 Dec 0.728 0.609 0.564 Fri 20 Dec 0.821 0.553 0.504 Considering the passenger comfort estimation results in more detail, Fig. 8 shows how the comfort errors (calculated as estimated level minus measured comfort level) are distributed. Firstly, it is interesting to observe that the number of data points with a comfort error equal to zero have by far the highest count, reaching almost 50 % of all observations for the case when historical data for all days are used. Secondly, it can be seen that the occurrence of error points decreases as the (absolute) error increases, i.e., the second most frequent errors are ±1, while higher errors (e.g. 4 and 5) are either extremely rare or even completely absent. Thirdly, the use of historical data not only increases the average performance, as discussed earlier, but also makes the distribution of errors narrower, meaning fewer higher error occurrences.Fig. 8 Distribution of the comfort errors for all days, considering a total of 10,523 data points. Moreover, it is interesting to mention that, from a practical viewpoint, a negative error (i.e. when a more comfortable condition than the actual one is estimated) is more critical than a positive error (i.e. when a less comfortable condition than the actual one is estimated), since the former would tell end-users that a run is less crowded than what it is in reality, which may, in turn, lead to poor decisions (e.g. even more crowding in an already crowded run). In this respect, it is interesting to observe that by using historical data, the samples with negative error diminish more than the ones with positive error. Furthermore, Fig. 9, Fig. 10 present an example of disaggregated information on the passenger comfort estimation for a selected day. In particular, Fig. 9 shows contour plots for the measured comfort, the estimated comfort, and the comfort error, calculated for all the runs equipped with APC for Wednesday 18 December 2019, and where estimates are produced without utilising historical data. Correspondingly, Fig. 10 shows the same contour plots, but in this case estimates are produced utilising historical data for all days. Again, the good performance of the proposed estimation scheme is demonstrated, as generally very low errors are produced, while it can be clearly seen that, in the case where historical data for all days are utilised, large areas in the contour plot are white (denoting zero comfort error) and swathes of purple and pink coloured cells (denoting negative comfort errors) are eliminated. It can also be observed how, at the disaggregated level, the negative errors reduce both in total amount and in magnitude when utilising historical data, which is therefore a preferable choice.Fig. 9 Measured comfort level (left); its estimates using the proposed method (centre); and the estimation error calculated in terms of comfort error (right), for Wednesday 18 December 2019, without utilising historical data. Fig. 10 Measured comfort level (left); its estimates using the proposed method (centre); and the estimation error calculated in terms of comfort error (right), for Wednesday 18 December 2019, utilising historical data for all days. The findings obtained are further supported by the graphs in Fig. 11 , which show the distribution of the comfort errors for all days, grouped by the measured comfort levels. As can be seen, the impact of the use of historical data in terms of the estimation accuracy becomes most prominent at the higher levels of on-board crowding (namely levels 3 and 4), where most of the severe negative discrepancies (of two or three levels) are either completely eliminated or, at least, attenuated. Indeed, just like in the example day of Fig. 9, Fig. 10, the pink and purple areas are severely diminished as more historical data are employed. This suggests that the proposed estimation scheme is not only able to reliably estimate the actual on-board comfort level with only few errors, but is also capable of ensuring that any remaining errors are low and not critical (i.e. no instances of a crowded vehicle estimated as empty).Fig. 11 Distribution of the comfort errors for all days, grouped by the measured comfort levels. Furthermore, it is investigated how the comfort error metric changes with the amount of historical data available to the estimation methodology, and to this end, experiments are performed by utilising an increasing number of days from the set of available historical data, from 5 to 34 days (i.e. all available days), and evaluating the performance for the same days as tested in the previous experiments. The results are presented in Fig. 12 , where it can be observed that as more historical data become available, the estimation performance improves. It is particularly interesting to highlight that, after a certain amount of historical data are available, adding further data improves the estimation performance only marginally; in the experiments carried out in this study, for example, this happens when more than 15 days of data are available.Fig. 12 Absolute comfort level error obtained via utilising different amounts of historical data. Finally, it is investigated how the performance of the proposed estimator varies with different amounts of APC data. These experiments are conducted by randomly selecting a percentage of the available APC equipped runs and performing estimation so that the proposed estimator employs only APC data from the selected runs. In particular, it is tested how the estimator performs using 25 %, 50 % and 75 % of the available APC data, which roughly corresponds to 2.5 %, 5 % and 7.5 % of the total number of daily runs, respectively (as per Section 4.2). The same percentages are used to sample the data treated as historical, as well as the data treated as real-time. In order to reduce noise due to the stochastic nature of the selection, five different random seeds are employed, and the arithmetic averages of the estimation metrics obtained for five replications for each of the tested percentages are reported. The results are presented in Fig. 13 , where the case where all the available APC data are used (i.e. 100 %) is also shown. As can be observed, and as can be expected, in virtually all tested days the estimation performance improves by increasing the amount of APC data available, with improvements being more significant at lower levels of APC data availability (i.e. from 25 % to 50 %).Fig. 13 Absolute comfort level error obtained by utilising different amounts of APC data. 6 Conclusions The provision of accurate and reliable passenger comfort information on-board public transport vehicles has long been identified as an important constituent component of good customer service. In light of the COVID-19 pandemic and its associated impacts on public transport patronage, however, the provision of such information has become an absolute necessity for the “survival” of public transport services going forward. The present study, therefore, has proposed a new KF-based estimation scheme for on-board comfort levels, which employs historical and current non-exhaustive APC data, as well as AVL measurements. The method has been successfully validated through application to the tramway network of the French city of Nantes, with the results demonstrating that overall low estimation errors are delivered, which are also non-critical. But while the study has made an initial contribution to the topic on on-board passenger comfort estimation, research in this direction continues. For instance, a limitation of the present study is that, as identified in Section 3.5, the system is not always fully observable. This means that there are some caveats in the estimation, such as in early-morning runs (due to the non-availability of previous measurements (apart from historical data)) and conditions with high passenger flow volatility (e.g. service disruptions). Indeed, it can be observed from the experiments of this study that the vast majority of the (remaining) errors originate in such situations, and it is the objective of future work to address this aspect, by including additional data sources, such as WiFi and mobile data. The development of these device-based capturing techniques could impact the necessity and the extent to which APC systems are deployed. In addition, it would be interesting to investigate how the frequency and distribution of APC-equipped vehicles affect estimation performance, which could be properly evaluated if a larger amount of APC data were available. To this extent, further analysis could explore how the estimation performance of the proposed scheme would compare with that of other methods, such as that of Jenelius (2019), and how it could be potentially improved by combining and integrating these methods. An alternative approach could involve using data produced by a simulator, which would allow randomly determining whether any (simulated) run is APC-equipped or not (naturally with limitations in terms of the realism of the demand and arrival patterns). Moreover, a limitation of the present study is that the proposed estimation method has so far not been applied in a real-time context, but only in an “offline” one. The reason for this is not inherent to the estimation method itself, but rather the fact that the two systems used in the validation case study (APC and AVL) operate independently from each other and their data are therefore not synchronised. A considerable amount of data pre-processing has therefore been necessary here in order to assimilate both the historical and the current datasets. In order to be able to perform real-time on-board comfort estimation, synchronisation of the APC and AVL data feeds using a vehicle label and a timestamp would be required so as to improve the run-matching process. In practice this is usually enabled by an architecture with continuous or semi-continuous communication between the systems. It would be, hence, useful to explore how this could be achieved in the current case study, or to apply the method in a real-time context on a different case study. This would additionally enable the monitoring of passenger behaviour and perceptions in response to the information provided by the estimation scheme. Finally, further research will concentrate on extending the capabilities and accuracy of the proposed method. For example, it would be interesting to investigate the performance of the estimation scheme in other sites, featuring more complex network topologies and a greater variety of modes, and to also extend the remit of the method from estimation (at the current station and time) to prediction (at a later station and/or future time point). Moreover, additional scenarios covering the phenomena of overcrowding and left-behind passengers (which were not present in the dataset employed here), may be further studied; these will likely require some special treatments and potentially also some modifications to the proposed method. Such extensions could provide a much different insight into how on-board comfort information would influence a priori and real-time passenger planning decisions in multi-modal trips and would consequently allow modelling the impacts at the route/mode choice and assignment levels. CRediT authorship contribution statement Claudio Roncoli: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Visualization. Ektoras Chandakas: Conceptualization, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Visualization. Ioannis Kaparias: Conceptualization, Investigation, Resources, Visualization, Writing – original draft, Writing – review & editing, Project administration. 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 would like to thank Semitan for supplying the data used in this study, and in particular the ‘‘Direction de la Performance et de l’Innovation’’ for their assistance in analysing the Autumn 2019 Opthor and Ineo datasets. The author Claudio Roncoli also acknowledges the support of the Academy of Finland project ALCOSTO (349327). 1 A random walk is an approach for modelling a stochastic process composed of a series of random variables through time. A Gaussian random walk is used in this study, in which the time series data are assumed to be generated based on a normal distribution. ==== Refs References Achar A. Bharathi D. Kumar B.A. Vanajakshi L. Bus arrival time prediction: a spatial Kalman filter approach IEEE Trans. Intell. Transp. Syst. 21 2020 1298 1307 Anderson B.D.O. Moore J.B. Optimal Filtering 1979 Prentice-Hall Antoniou C. Ben-Akiva M. Koutsopoulos H.N. Kalman Filter applications for traffic management Kordic V. Kalman Filter 2010 IntechOpen Antsaklis, P., Michel, A.N., 2006. Linear Systems. Birkhäuser. Bai Y. Sun Z. Zeng B. Deng J. Li C. A multi-pattern deep fusion model for short-term bus passenger flow forecasting Appl. Soft Comput. 58 2017 669 680 Batarce M., Muñoz J.C., Ortúzar J. de D., Raveau S., Mojica C., Ríos A., 2015. Valuing crowding in public transport systems using mixed SP/RP data: the case of Santiago. Transport. Res. Record 2535, 73-78. Bekiaris-Liberis N. Roncoli C. Papageorgiou M. Highway traffic state estimation with mixed connected and conventional vehicles IEEE Trans. Intell. Transp. Syst. 17 2016 3484 3497 Bekiaris-Liberis N. Roncoli C. Papageorgiou M. Highway traffic state estimation per lane in the presence of connected vehicles Transp. Res. B 106 2016 1 28 Chandakas E., 2009. La capacité des transports ferroviaires d’Ile-de-France face à la hausse du trafic à long-terme. Master Thesis, Ecole Nationale des Ponts et Chaussées. Cheng L. Chen X. De Vos J. Lai X. Witlox F. Applying a random forest method approach to model travel mode choice behavior Travel Behav. Soc. 14 2019 1 10 Cox T. Houdmont J. Griffiths A. Rail passenger crowding, stress, health and safety in Britain Transp. Res. A 40 2006 244 258 De Palma A. Kilani M. Proost S. Discomfort in mass transit and its implication for scheduling and pricing Transp. Res. B 71 2015 1 18 De Solere, R. (Ed), 2012. La mobilité urbaine en France: Enseignements des années 2000-2010. Editions du CERTU, Collection Références. Ding C. Duan J. Zhang Y. Wu X. Yu G. Using an ARIMA-GARCH modeling approach to improve subway short-term ridership forecasting accounting for dynamic volatility IEEE Trans. Intell. Transp. Syst. 19 2018 1054 1064 European Committee for Standardization (CEN) (2002). EN 13816:2002 Transportation – Logistics and services – Public passenger transport – Service quality definition, targeting and measurement, CEN/TC 320 - Transport - Logistics and services. Available. Gkiotsalitis K. Cats O. Public transport planning adaption under the COVID-19 pandemic crisis: literature review of research needs and directions Transp. Rev. 41 2021 374 392 Gong M. Fei X. Wang Z.H. Qiu Y.J. Sequential framework for short-term passenger flow prediction at bus stop Transp. Res. Rec. 2417 2014 58 66 Guo J. Huang W. Williams B.M. Adaptive Kalman filter approach for stochastic short-term traffic flow rate prediction and uncertainty quantification Transp. Res. C 43 2014 50 64 Gur Y.J. Ben-Shabat E. Estimating bus boarding matrix using boarding counts in individual vehicles Transp. Res. Rec. 1607 1997 81 86 Haywood L. Koning M. Estimating crowding costs in public transport DIW Berlin Discussion Paper No. 1293 2013 Haywood L. Koning M. Monchambert G. Crowding in public transport: Who cares and why? Transp. Res. A 100 2017 215 227 He H. Yan M. Sun C. Peng J. Li M. Jia H. Predictive air-conditioner control for electric buses with passenger amount variation forecast Appl. Energy 227 2018 249 261 Heydenrijk-Ottens, L., Degeler, V., Luo, D., van Oort, N., van Lint, H., 2018. Supervised learning: Predicting passenger load in public transport. Conference on Advanced Systems in Public Transport (CASPT) 2018. Hu R. Chiu Y.-C. Hsieh C.-W. Crowding prediction on mass rapid transit systems using a weighted bidirectional recurrent neural network IET Intel. Transport Syst. 14 2020 196 203 Huang D. Yu J. Shen S. Li Z. Zhao L. Gong C. A method for bus OD matrix estimation using multisource data J. Adv. Transp. 2020 2020 5740521 Jenelius E. Data-driven metro train crowding prediction based on real-time load data IEEE Trans. Intell. Transp. Syst. 21 2020 2254 2265 Jenelius, E. (2019). “Data-driven bus crowding prediction based on real-time passenger counts and vehicle locations”. 6th International Conference on Models and Technologies for Intelligent Transport Systems (MT-ITS 2019). Ji Y. Mishalani R.G. McCord M.R. Transit passenger origin–destination flow estimation: Efficiently combining onboard survey and large automatic passenger count datasets Transp. Res. C 58 2015 178 192 Jia F. Li H. Jiang X. Xu X. Deep learning-based hybrid model for short-term subway passenger flow prediction using automatic fare collection data IET Intel. Transport Syst. 13 2019 1708 1716 Jung Y.J. Casello J.M. Assessment of the transit ridership prediction errors using AVL/APC data Transportation 47 2020 2731 2755 Kalman R.E. Bucy R.S. New results in linear filtering and prediction theory Transactions of the ASME Series D 83 1961 95 108 Khomchuk, P., Tuladhar, S.R., Sivananthan, S., 2018. Predicting passenger loading level on a train car: A Bayesian approach. arXiv eprints, p. 1808.06962. Koutsopoulos, H.N., Noursalehi, P., Zhu, Y., Wilson, N.H.M., 2017. Automated data in transit: Recent developments and applications. 5th IEEE International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS 2017), Naples, Italy. Koutsopoulos H.N. Ma Z. Noursalehi P. Zhu Y. Chapter 10 - Transit data analytics for planning, monitoring, control, and information Actoniou C. Dimitriou L. Pereira F. Mobility Patterns, Big Data and Transport Analytics: Tools and Applications for Modeling 2019 Elsevier 229 261 Kroes, E., Kouwenhoven, M., Debrincat, L., Pauget, N., 2013. “On the value of crowding in public transport for Île-de-France. International Transport Forum Discussion Papers, no. 2013/18, Organisation for Economic Co-operation and Development (OECD). Leurent, F., Liu, K., 2009. On seat congestion, passenger comfort and route choice in urban transit: a network equilibrium assignment model with application to Paris, 88th Annual Meeting of the Transportation Research Board, Washington DC, USA. Leurent F. Chandakas E. Poulhes A. A traffic assignment model for passenger transit on a capacitated network: Bi-layer framework, line sub-models and large-scale application Transp. Res. C 47 2013 3 27 Li, D., Lin, Y., Zhao, X., Song, H., Zou, N., 2011. Estimating a transit passenger trip Origin-Destination matrix using Automatic Fare Collection system. In: Xu, J., Yu, G., Zhou, S., Unland, R. (eds), Database Systems for Advanced Applications (DASFAA 2011), Lecture Notes in Computer Science, vol. 6637. Springer. Li Z. Hensher D.A. Crowding in public transport: a review of objective and subjective measures J. Public Transp. 16 2013 107 134 Lin C.-T. Structural controllability IEEE Trans. Autom. Control 19 1974 201 208 Liu, Y-Y., Slotine, J-J. , Barabasi, A-L., 2013. Observability of complex systems. Proceedings of the National Academy of Sciences, vol. 110, pp. 2460–2465. Liu X. van Hentenryck P. Zhao X. “Optimization models for estimating transit network Origin-Destination flows with big transit data J. Big Data Anal. Transport. 3 2021 247 262 Mikkelsen, L., Buchakchiev, R., Madsen, T., Schwefel, H.P., 2016. Public transport occupancy estimation using WLAN probing. 8th International Workshop on Resilient Networks Design and Modeling (RNDM), 302-308. Mohammadi A. Amador-Jimenez L. Nasiri F. A multi-criteria assessment of the passengers’ level of comfort in urban railway rolling stock Sustain. Cities Soc. 53 2020 101892 Noursalehi, P., 2017. Decision Support Platform for Urban Rail Systems: Real-time Crowding Prediction and Information Generation, PhD thesis, Northeastern University, Boston, MA, USA. Oransirikul T. Nishide R. Piumarta I. Takada H. Measuring bus passenger load by monitoring Wi-Fi transmissions from mobile devices Procedia Technol. 18 2014 120 125 Pasini, K., Khouadjia, M., Ganansia, F., Oukhellou, L., 2019. Forecasting passenger load in a transit network using data driven models”. 12th World Congress on Railway Research (WCRR 2019). Pasini, K., Khouadjia, M., Samé, A., Ganansia, F., Oukhellou, L., 2020. LSTM encoder-predictor for short-term train load forecasting. In: Brefeld, U., Fromont, E., Hotho, A., Knobbe, A., Maathuis, M., Robardet, C. (eds), Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2019, Lecture Notes in Computer Science, vol. 11908, pp 535-551. Springer. Przybylowski A. Stelmak S. Suchanek M. Mobility behaviour in view of the impact of the COVID-19 pandemic – Public transport users in Gdansk case study Sustainability 13 2021 Reissig G. Hartung C. Svaricek F. Strong structural controllability and observability of linear time-varying systems IEEE Trans. Autom. Control 59 2014 3087 3092 Roncoli C. Bekiaris-Liberis N. Papageorgiou M. Use of speed measurements for highway traffic state estimation: case studies on NGSIM data and Highway A20, Netherlands Transp. Res. Rec. 2559 2016 90 100 Simon D. Kalman filtering with state constraints: a survey of linear and nonlinear algorithms IET Control Theory Appl. 4 2010 1303 1318 Simon, D., 2006. Optimal State Estimation: Kalman, H Infinity and Nonlinear Approaches. John Wiley & Sons. Sun W. Schmöcker J.-D. Fukuda K. Estimating the route-level passenger demand profile from bus dwell times Transp. Res. C 130 2021 103273 Szeto M. Gazis D. Application of Kalman Filtering to the surveillance and control of traffic systems Transp. Sci. 6 1972 419 439 Tao Z. Tang J. Real-time estimation of urban rail transit passenger flow status based on multi-source data J. Phys. Conf. Ser. 1187 2019 052070 Tirachini, A., Heshner, D.A., Rose, J.M. (2012). “Multimodal pricing and optimal design of public transport services: The interplay between traffic congestion and bus crowding”. 12th International Conference on Advanced Systems of Public Transport, Santiago, Chile. Tirachini A. Hurtubia R. Dekker T. Daziano R. Estimation of crowding discomfort in public transport: results from Santiago in Chile Transp. Res. A 103 2017 311 326 Transport Focus, 2021. “Will there be space on board?”. Available: https://www.transportfocus.org.uk/publication/will-there-be-space-onboard. Trozzi V. Gentile G. Bell M.G.H. Kaparias I. Dynamic user equilibrium in public transport networks with passenger congestion and hyperpaths Transp. Res. B 57 2013 266 285 Tsai T.-H. Lee C.-K. Wei C.-H. Neural network based temporal feature models for short-term railway passenger demand forecasting Expert Syst. Appl. 36 2009 3728 3736 Turner S. Corbett E. O’Hara R. White J. Health and safety effects of rail crowding: hazard identification 2005 RSSB R&D Programme T307 Report, US Transportation Research Board, 2013. Transit Capacity and Quality of Service Manual, Third Edition. The National Academies Press. van Oort N. Brands T. de Romph E. Short-term prediction of ridership on public transport with smart card data Transp. Res. Rec. 2535 2015 105 111 Vlahogianni E. Golias J.C. Karlaftis M.G. Short-term traffic forecasting: Overview of objectives and methods Transp. Rev. 24 2004 533 557 Vlahogianni E. Karlaftis M.G. Golias J.C. Short-term traffic forecasting: where we are and where we’re going Transp. Res. C 43 2014 3 19 Wang W. Attanucci J. Wilson N. Bus passenger Origin-Destination estimation and related analyses using automated data collection systems J. Public Transp. 14 2011 131 150 Wang P. Chen X. Chen J. Hua M. Pu Z. A two-stage method for bus passenger load prediction using automatic passenger counting data IET Intel. Transport Syst. 15 2021 248 260 Wang Y. Papageorgiou M. Real–time freeway traffic state estimation based on extended Kalman filter: A general approach Transp. Res. B 39 2005 141 167 Zhang Y. Jenelius E. Kottenhoff K. Impact of real-time crowding information: a Stockholm metro pilot study Public Transport 9 2017 483 499 Zhang J. Shen D. Tu L. Zhang F. Xu C. Wang Y. Tian C. A real-time passenger flow estimation and prediction method for urban bus transit systems IEEE Trans. Intell. Transport. Syst. 18 2017 3168 3178	36471757	PMC9712033	NO-CC CODE	2022-12-02 23:21:31	no	Transp Res Part C Emerg Technol. 2023 Jan 1; 146:103963	utf-8	Transp Res Part C Emerg Technol	2,022	10.1016/j.trc.2022.103963	oa_other

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.