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
Heliyon
Heliyon
Heliyon
2405-8440
The Author(s). Published by Elsevier Ltd.
S2405-8440(22)03615-5
10.1016/j.heliyon.2022.e12327
e12327
Research Article
Effectivity of repurposed drugs against SARS-CoV-2 infections, A hope for COVID 19: inhibitor modelling studies by docking and molecular dynamics
Yadav Pooja
Chowdhury Papia ∗
Department of Physics and Materials Science & Engineering, Jaypee Institute of Information Technology, Noida 201309, Uttar Pradesh, India
∗ Corresponding author.
10 12 2022
12 2022
10 12 2022
8 12 e12327e12327
16 7 2022
1 11 2022
6 12 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 present study, we have done a comparative study on the efficacy of some currently used repurposed drugs: Oseltamivir (O), Favipiravir (F) and Hydroxychloroquine (H) in individual and in their combinational mode against CoV-2 infections. The ADME analysis has helped us to identify the inhibitory possibility of the tested drugs towards receptor 3CLpro protein of SARS-CoV-2. Various thermodynamical parameters obtained from Molecular Docking, Molecular dynamics (MD) and MMPBSA simulations like binding affinity, potential energy (Epot), RMSD, RMSF, SASA energy, interaction energies, Gibbs free energy (ΔGbind) etc. also helped us to verify the effectivity of mentioned drugs against CoV-2 protease.
SARS-CoV-2; COVID-19; 3CLpro; Favipiravir; Hydroxychloroquine; Oseltamivir; RMSD; RMSF.
Graphical abstract
Image 1
Keywords
SARS-CoV-2
COVID-19
3CLpro
Favipiravir
Hydroxychloroquine
Oseltamivir
RMSD
RMSF
==== Body
pmc1 Introduction
In the 21st century over the last three years, world is in the pandemic situation created by novel coronavirus. Sometime in December 2019, the virus first detected in China [1]. SARS-CoV-2, a subcategory of the novel coronavirus has played a key role in this catastrophic public health emergency throughout the world [2]. As per WHO, on 5th December 2022, cumulative cases of COVID-19 disease are 640, 395, 351 and cumulative deaths are 6,618,579 [3]. CoV-2 infection can be easily spread by respiratory droplets from symptomatic, asymptomatic, and pre-symptomatic patients through surface, contact and aerosol transmission [4, 5, 6]. Community transmission and cluster formations are observed to be important factors for the rapid expansion of this disease. A lack of immunity in the majority of the population has already created a high incidence of life-threatening conditions in each and every corner of the world. Owing to these conditions there is an urgent need for a therapeutic solution and vaccine development with maximum efficacy against the disease as well as extensive research on its biology.
Only COVID-19 vaccines can provide strong resistance against death, hospitalization and serious illness. As no one is safe till everyone is safe, a mass vaccination programme has started early from December 2020. As on December 2022, around 10 vaccines are available in market for human use. There are more 31 vaccines are under different stages of human trial phases [7, 8, 9]. However even with massive efforts, till date none of the tested vaccine could significantly impact the course of COVID-19 pandemic. Also, it is a fact that some people are still getting ill from COVID-19 after vaccination. With the scientific idea obtained from recent outbreaks of Ebola virus disease, SARS-CoV and MERS-CoV, the researchers and medical practitioners have created a synchronized and fast-tracked answer to COVID-19, through the research into potential therapeutic pharmaceutical treatments. De novo development of therapeutics to bring a new drug into clinical practice has to pass through a laborious five step process. The steps include discovery & development, preclinical research, clinical research, FDA review and FDA post market drug safety monitoring. Usually, the whole process used to take a decade to act as effective option against any disease as drug invention and distribution may be refined endlessly even after the drug's final approval. So the whole procedure becomes time-consuming and expensive with a high possibility of failure. So as an alternative, another novel approach to identify new therapeutic option is drug repurposing [10, 11]. Drug repositioning is the main activity of drug repurposing which involves the investigation of new therapeutic indications of existing drugs. To combat against CoV-2 infections, many antiviral FDA approved drugs already used for known viral diseases like influenza, Ebola, HIV, hepatitis, Zika, Alzheimer's etc are being repositioning for the treatment of COVID-19 disease [12, 13]. Remdesivir, Sofosbuvir, Ribavirin etc are approved antiviral drugs for hepatitis C virus [14, 15, 16], are being repositioning for the treatment of COVID positive patients. Favipiravir and Oseltamivir, as originally approved drugs Ebola, Lassa and influenza viruses (A and B), now are being tested for the treatment of COVID-19. Antiviral drugs generally target viral proteins or cellular proteins where they do not destroy their target pathogens rather inhibit their development by any of the below mentioned actions. Drug can resist the viral protease from entering the host cell, secondly it can oppose the replication of viral protein or it can act as minimizing agent against the damages that can be created by infecting virus [14]. Not only antiviral drugs, many antibiotic drugs like Ivermectin and Doxycycline, antimalarial drugs Chloroquine and Hydroxychloroquine, also are being repositioning for COVID-19 treatment [17, 18]. Therapeutic switching is now considered as new effective approach to enhance the stereoisomers and metabolites of existing drugs. All of the above-mentioned individual antiviral and antibiotic drugs have been tested for the prevention and effective management of COVID-19 infection. Till now the available clinical data validated that most of the repurposed drugs tested for COVID-19 treatment, have shown lower or moderate effectiveness with many adverse side effects for the patient conditions like diabetes, hypertension, cardiac problems [19, 20]. Many repurposed combination drugs have also repositioned for the treatment of HIV, other coronaviruses and MERS-CoV infection [21]. Lopinavir + Ritonavir and Mycophenolic acid (MPA) + IFN-® combination are popularly used combination medicine for MERS-CoV infection [22, 23]. However, it is a big question that which repositioning showing better clinical benefits for hospitalized COVID-19 patients: individual or combination therapy [24]? It is a topic of argument that whether a repurposed drugs have shown any benefit in hospitalized COVID-19 patients or not, still they have been proved to be useful in slowing the development of illness, preventing severe disease and less hospitalization. Dose adjustments, acute/chronic toxicity, choice of appropriate delivery system and route of administration to deliver the drugs are some major challenges for the application of repurposed drugs which need a rigorous study. Also, there is an urgent need for assessment of repurposed drugs in terms of their binding efficiency with target virus, repurposing dynamics, limitations and delivery approaches to identify some effective antiviral drugs which will positively impact the COVID 19 therapy. Approved repurposed drugs have saved millions of human lives over the last decades, still it is a challenge to identify effective, low toxic, well tolerated drugs that can enhance patient recovery limit. This challenge has created the motivation for performing the present study. For the current study, the effectivity of some popularly used repurposed direct targeted and host targeted antiviral drugs in their individual and combinational modes have been checked with the help of some in silico methods like DFT, ADME, Molecular Docking, MD simulation and MMPBSA simulations against CoV-2. Our target individual drugs are Oseltamivir, Favipiravir, Hydroxychloroquine.
For SARS-CoV-2 drug development, the intriguing drug target is broad -spectrum class of viral RdRP inhibitors. The most promising RdRp inhibitors are nucleoside analogues (NAs). Favipiravir is one of popularly used RdRP inhibitor or NA [25]. NAs are incorporated into the nascent viral RNA by the RdRp. RdRP inhibitor is mistakenly considered as purine nucleotide by receptor viral protease. As a result of that target protein makes the RdRP: receptor complex structure in the form of exonuclease (ExoN) [26]. The main problem of NA incorporation by RdRp into nascent RNA is that they many causes disturbance in RNA synthesis. NAs can non-obligate the RNA synthesis process by perturbing the product RNA structure enough to stop further synthesis by polymerase which may definitely reduce their antiviral effects. NAs also can obligate the RNA chain extension as they are unable to be recognized as regular Watson-Crick nucleobases. As a result of that production of non-viable genomes occurs by a process ‘lethal mutagenesis. Favipiravir belongs to this class of antivirals which can act actively against variety of viruses (Ebola, Lassa and influenza viruses). Though for CoVs, certain NAs insertion into nascent RNA have shown their reduced antiviral effects, Favipiravir is widely and successfully used to cure COVID-19 patients with a good efficacy [27, 28, 29]. In India and Chaina, Favipiravir is being used as an approved drug for COVID 19 treatment [30]. Another popular antiparasitic drug showing anti SARS-CoV 2 activity is Hydroxychloroquine. It is an antimalarial drug which also used for the treatment many other inflammatory diseases like rheumatoid arthritis [31, 32], systemic lupus erythematosus [33]. The utility of Hydroxychloroquine for COVID-19 treatment has been described in several research papers [32, 34], but majority of studies reported no benefit with Hydroxychloroquine treatment for hospitalized patients. Moreover, the concern has been raised about the cardiac toxicity effects of Hydroxychloroquine. As a result, use of Hydroxychloroquine as proposed drug against COVID treatment has been terminated in many countries but later on emergency use authorization have been again revived by FDA [35]. Another commonly used antiviruses in the regimen is Oseltamivir. Oseltamivir is a commonly used drug for prophylaxis of infection due to influenza viruses A and B [36]. Oseltamivir is a most accepted drug for MERS-CoV and SARSCoV-2 therapy in Korea and China [37]. It is an analog to sialic acid which acts to inhibit the neuraminidase enzyme for viral replication. A large number of reported data suggested that for the hospitalized patients with several respiratory tract infections administration of Oseltamivir has reduced their of length of stay in hospitals [38].
2 Materials and methods
2.1 Identification of target protein structure of SARS-CoV-2 and its preparation
The structures of newly invented CoV-2 virus and already known CoV virus is very similar [39, 40]. Till now there are more than 160 reported compounds (co-crystallized with SARS-CoV-2 Pro/CLpro) proteins have been identified. Inhibitors of the SARS-CoV (2003) Mpro can act through a irreversible inactivation mechanism where the inhibitor associates with the Mpro to form enzyme-inhibitor complex through stable covalent bonding [41]. Among many available SARS-CoV-2 Mpro/CLpro reported [42, 43] structures, 6LU7 is used as main protease in CoV enzyme. 6LU7 plays a crucial role in mediating viral replication and transcription, making it an better drug target for this virus. We have used 6LU7, a 3CLpro proteases as main target protein. The 3D structure of the 6LU7 was retrieved from the Protein Data Bank website (https://www.rcsb.org) [44] shown in Table 1 and used as a receptor. To prepare the protein structure for simulation ready, the existing water molecules were removed and then polar hydrogens were added in protein structure by using AutoDock and MGL tools. Also, the inbuilt ligand was removed from the protein structure by Discovery studio 2020 [45, 46, 47]. The output protein structure was saved in PDB format.Table 1 Structures of receptor protein (6LU7), and targeted drugs: Oseltamivir (O), Favipiravir (F) and Hydroxychloroquine (H).
Table 1Compound Name Structure
Protease:6LU7 Image 1
Oseltamivir (O)
(C16H28N2O4) Image 2
Favipiravir (F) (C5H4FN3O2) Image 3
Hydroxychloroquine(H)
(C18H26ClN3O) Image 4
Red color-oxygen atom, blue-nitrogen atom, green – chlorine atom, light-blue-fluorine atom
2.2 Procedure of ligand drug molecules preparations
For the screening of O, F and H, the SWISS ADME (https://www.swissadme.ch) and ADMET (https://vnnadmet.bhsai.org/) software's were used to verify Drug-likeness rules like Lipinski's rule of five (Ro5) or ‘‘a rule of thumb, Veber's rule, MDDR-like rule, Egan rule, Ghose filter, Muegge rule etc., [48, 49].
For Drug preparation, the ligands in ‘SDF’ format were obtained directly from the PubChem (National Library of Medicine) (https://pubchem.ncbi.nlm.nih.gov/) and converted to ‘PDB’ format with the help of Autodock tools [46]. All ligands have been used at their optimized structures for further simulation studies. All molecular structures of the ligands and appo protein were optimized by using density functional theory (DFT) with the basis set 6.311G (d,p) [50] using the Gaussian 09 program [51]. The optimized structures were visualized with the help of Gauss View 5 molecular visualization program [52].
2.3 Molecular docking
By using Autodock Vina and Discovery studio visualizer [45, 46], the docking-based studies and analysis of the recommended inhibitors against protease of SARS-CoV-2 have been performed. The parameters for configuration file were as follows: total no. of binding modes was taken as 9, exhaustiveness as 8 and maximum energy difference was considered as 3 kcal/mol. Among all of the observed 9 poses, the best pose was chosen on the basis of maximum non bonded interaction, minimum binding affinity (kcal/mol), dreiding energy, dipole moment and also minimum inhibition constant. For inhibition constant (K i), following equation was used(1) Ki=eΔGRT
where, binding affinity is G, universal constant is R and T temperature (300K). Lowest value of K i validates the strong interaction between ligand and receptor protein 6LU7. For combinational mode of ligands sequential docking were performed.
2.4 Molecular dynamic (MD) and MMPBSA
For calculating thermodynamics parameters of the ligand: protein complex structure through MD simulation, the LINUX based platform ‘‘GROMACS 5.1 Packageˮ with GROMOS43A2 force field was used [53, 54]. For topology creation of protein: ligand complex structure or appo protein structures, TIP3P water model was used with adding 4Na+ ions to maintain the neutrality of the structure. Simulation has been performed starting with steepest descent algorithm for achieving the optimized structure with minimum potential energy (Epot) under equilibrium condition. The algorithm has a cut off up to 239.006 kcal/mol for minimizing the steric clashes which includes two phases each having 500,000 steps. The first phase includes constant particles no., volume and temperature (NVT) and the second phase includes, constant pressure and temperature (NPT). On the optimized lower energy structure MD simulation has been run ranges from 1 ns-100 ns and various thermodynamical parameter have been derived within the mentioned time scale. Some important thermodynamic parameters are root mean square deviation (RMSD), root mean square fluctuation (RMSF), Epot, radius of gyration (Rg), inter-molecular H-bonds, solvent accessible surface area (SASA) and various non-bonded interaction energies for protein:ligand complex structures. Short range Coulomb interaction (Coul-SR) and short range Lennard–Jones interaction (SR-LJ) were used to compute nonbonded interaction between protein and ligand. Different terms of energies can be expressed in additive forms as:(2) Etotal=Ebonded+Enonbonded
(3) Ebonded=Ebond+Eangle+Edihedral
(4) Enonbonded=Ehydrogenbond+Eelectrostatic+Evanderwaals
(5) Eelectrostatic=Ecoulombic+ElenardJones
To find the binding affinities of inhibitors towards receptor protein we have applied the MM/PBSA method [55]. To calculate the interaction free energies (ΔGbind) for the protein:ligand complex structure, the MMPBSA (Molecular Mechanics Poisson-Boltzmann Surface Area) method [56] sourced from the APBS and GROMACS packages have been used within 0 ns–100 ns. In MM/PBSA, method the van der Waals energy components (Evdw) and electrostatic energy components (Eelectrostatic) between inhibitor and receptor are used to determine the stability/binding affinity for all complex structures. The same process is applied for single and multiple ligand system. The ΔEMM was calculated by using molecular mechanics (MM) force field with polar part of the solvation free energy calculation by linearized Poisson-Boltzmann model and apolar part by using surface area approach. In aqueous environment, the binding free energy of the ligand: receptor complex can be computed by following equations.(6) ΔGbind,aqu=ΔH−TΔS≈ΔEMM+ΔGbind,solv−TΔS
(7) ΔEMM=ΔEcovalent+ΔEelectrostatic+ΔEvanderwaals
(8) ΔEcovalent=ΔEbond+ΔEangle+ΔEtorsion
(9) ΔGbind,solv=ΔGpolar+ΔGnonpolar
where, −TΔS, Δ and Δ, are the molecular conformational energy due to binding and solvation free energy and mechanical energy changes in gas phase, respectively.
2.5 Computational details
MD simulations and corresponding energy calculations have been computed in a single system using HP Intel Core i5 - 1035G1 CPU and 8 GB of RAM with Intel UHD Graphics and a 512 GB SSD.
3 Results and discussion
3.1 Screening and analysis of drug likeness properties of targeted drugs
All the targeted repurposed drugs: F, H, O we have considered for current study follow the Lipinski's rule of five (Ro5) and veber rule (Table 2 ). All of them also satisfy all bioavailability conditions [49]. All the targeted drugs satisfy all required conditions for Physiochemical Properties, Fraction CSp3, Water Solubility and Medicinal Chemical properties. Results from ADMET analysis (https://vnnadmet.bhsai.org/) shows that none of the tested drug has any cyto-toxicity effect and maximum suggested dose for O, F and H were obtained as 165mg/day, 170mg/day and 478mg/day.Table 2 Physiochemical, Drug-likeness, water solubility properties, Medicinal chemistry and toxicity of Oseltamivir, Favipiravir and Hydroxychloroquine.
Table 2Name of Ligand Oseltamivir Favipiravir Hydroxychloroquine
Physiochemical Properties
Molecular Formula C16H28N2O4 C5H4FN3O2 C18H26ClN3O
Molecular Weight 312.40 g/mol 157.10 g/mol 335.87 g/mol
Hydrogen Bond Donor Count 2 2 2
Hydrogen Bond Acceptor Count 5 4 3
Topological Polar Surface Area 90.65 Å2 88.84 Å2 48.39 Å2
Fraction CSP3 0.70 0 0.50
Water Solubility
Log S (SILICOS-IT) -2.47 -1.42 -6.35
Class soluble Soluble Poorly soluble
Solubility 1.05e+00 mg/ml; 3.37e-03 mol/l 6.04e+00 mg/ml; 3.85e-02 mol/l 1.50e-04 mg/ml; 4.46e-07 mol/l
Drug Likeness
Lipinski Rule Yes; 0 violation Yes; 0 violation Yes; 0 violation
Ghose Filter Yes No; 4 violations:MW < 160, WLOGP < -0.4, MR < 40, #atoms<20 Yes
Veber (GSK) Rule Yes Yes Yes
Egan (phatmacial) Filter Yes Yes Yes
Muegge (Bayer) Filter Yes No; 1 violation:MW < 200 Yes
Bioavailability (Abbott) Score 0.55 0.55 0.55
Medicinal Chemistry
PAINS (Pan Assey Interference Structures) 0 alert 0 alert 0 alert
Brenk 1 alert: phosphor 0 alert 0 alert
Leadlikeness No; 1 violation:Rotors>7 No; 1 violation:MW < 250 No; 2 violations:Rotors>7, XLOGP3>3.5
Synthetic accessibility 4.449 2.08 2.82
Toxicity
Cyto-toxicity No No No
MRTD(mg/day) 165 170 478
3.2 Analysis of molecular docking results
As per molecular docking results for O, among 9 pose structures pose 1 can be considered as the stable complex (O:6LU7) structure having a lowest value of binding energy (−6.2 kcal/mol), inhibition constant (2.8 × 10−5 M) as obtained from Eq. (1) and maximum number of intermolecular hydrogen bonded interactions, dreiding energy (91.62 kcal/mol), dipole moment (4.039 Debye) at room temperature (300 K) (Figure 1 a, Table 3 ) [57]. Similarly in case of F and H also for pose 1 we obtained the stable complex structures (F: 6LU7, H:6LU7) at 300 K (Figures 1b, 1c and Table 3). Based on their crystal structure it was identified that CoV enzyme 6LU7 have some binding pockets within its protease which act as active sites to bind with various ligands [58, 59]. The binding sites are known as S1, S2, S3 and S4 which comprise of some side chain residues and backbone of the base protein. From molecular docking study it is observed that during individual modes of interaction O gets attached with GLY143, GLN189, GLU166 chain residues of active site S4, F gets attached with GLY143, SER144, CYS145 chain residues of S2 binding sites whereas H gets attached with GLU166, MET165, ARG188, THR190 chain residues of active sites of S4 (Table 3).Figure 1 Binding energies and active binding sites for drugs a) O, b) F and c) H towards protein 6LU7 by individual docking.
Figure 1
Table 3 Various parameters like binding affinity, Hydrogen bonded interaction, dipole moment, dreiding energy, Inhibition constant for the docked structure of O, F and H with receptor protein 6LU7.
Table 3Ligand Binding affinity (kcal/mol) Hydrogen bonded interaction (donor - acceptor, distance in Å)
(A-Amino acid, drug:ligand) Dreiding energy (ligand)
(kcal/mol) Dipole moment of ligand (Debye) Inhibition
Constant (M)
Ki = eΔG/KT from equ (1)
O -6.2 (A:GLY143:HN -::O:O, 1.92)
(A:GLU166:HN -::O:O, 1.99)
(:O:H -: A:GLN189:OE1, 2.2) 91.62 4.039 2.8 × 10−5
F -4.9 (A:GLY143:HN -:F:O, 2.57)
(A:SER144:HN -:F:O, 2.15)
(A:CYS145:HN-:F:O, 2.20) 55.0 2.460 2.5 × 10−4
H -4.8 (:H:N -:A:GLU166:O, 3.23)
(:H:O -:A:MET165:SD, 3.61)
(:H:O -:A:ARG188:O, 3.29)
(:H:C -:A:THR190:O, 3.59) 126.57 2.461 3.0 × 10−4
To check the effectivity of combination drugs against CoV 2 target protein, sequential molecular docking mechanism has been used for inspection of the binding affinity of combination drugs (O + F, O + H, F + H, F + H + O) as inhibitors towards receptor 6LU7. The computed binding affinity for O + F, O + H, F + H, F + H + O combinational inhibitors and individual inhibitors towards receptor protein of 6LU7 have shown that O in individual mode and O + F + H in combination mode exhibited better affinity (−6.2, -5.3 kcal/mol) towards 6LU7 (Figure 2 a,b, Table 3, SD1). For every best pose, the donor–acceptor surface are shown in Figure 3 (a-f) in 3D and 2D view with their possible hydrogen bonding (Table 3 and supporting document (SD)1). Further through MD simulation results, we have tried to check the effectivity of ligand: receptor complex formation for our proposed individual and combinations drugs.Figure 2 Binding energies and binding sites for drug a) O, b) O + F + H towards protein 6LU7 by sequential docking.
Figure 2
Figure 3 Donor: acceptor surface for best pose in terms of H-bond interaction a), b), c) O: 6LU7, F: 6LU7, H: 6LU7 d), e), f) Possible types of interaction in pose obtained from O:6LU7, F:6LU7, H:6LU7.
Figure 3
3.3 Analysis of MD simulation results
MD simulation of appo 6LU7 and for ligand:6LU7 complexes have been carried out for the time scale of 1 ns–100 ns (Eqs. (1), (2), (3), (4), and (5)). Before MD simulation to work every structure have to be optimized by potential energy (Epot) minimization process. For energetically stable structure the Epot should be minimum with a maximum force value. For appo 6LU7 and ligand:6LU7 complexes we have obtained steady convergence of potential energy (Figure 4 ). Average energy of appo 6LU7 was obtained as -0.3 × 106 kcal/mol. The optimized potential energies were computed at the order of ∼ -0.1 × 106 kcal/mol for all complex structures (in individual and combination forms) (Figure 4).Figure 4 Potential energy of receptor 6LU7 and complex structures of O, F, H and O + F + H with 6LU7.
Figure 4
To run the MD simulation in aqueous solution a well-known water model: TIP3P has been used. In aqueous environment before MD simulation, the stability of each complex structures has been verified for a short interval of time (100 ps) under the equilibrium condition (NVT and NPT). It was observed that at 300 K during stabilization process, each structure individually has achieved its stable equilibrated phase after first few ps (∼10 ps) of computation and maintained that stable configuration by the obtained simulated data like density (D), pressure (P) and temperature (T) throughout the whole process (SD2 a-c). After achieving the fully equilibrated phase for both appo protein and ligand: protein complex forms, MD simulations were operated for complete 100 ns time trajectory. Output for all thermodynamic parameters obtained from MD simulation are shown in Table 4 , SD3.Table 4 Data obtained by MD simulations for protein 6LU7 in its bare state and for the O:6LU7, F:6LU7 and H:6LU7 complex structure.
Table 4S. No. Parameter Bare 3CLpro protease (6LU7) O:6LU7 F:6LU7 H:6LU7
Mean Range Mean Range Mean Range Mean Range
MD Simulation Result
1. RMSD (nm) 0.22 0.13–0.32 0.23 0.09–0.48 0.22 0.10–0.40 0.24 0.23–0.24
2. RMSF (nm) 0.23 0.05–0.4 0.33 0.07–0.58 0.29 0.06–0.52 0.16 0.05–0.28
3. Inter H-Bonds NA NA 2 0–3 5 0–9 3 0–5
4. Radius of gyration 2.18 2.13–2.24 2.18 2.16–2.4 2.21 2.20–2.21 2.21 2.20–2.22
5. SASA (nm2) 33 30–35 8 6.5–9 8 7–9 8.2 7.3–9
MM/PBSA Results
6. Potential Energy (kcal/mol) -0.30×106± 8.8 -7.0×105 - -1.3×106 -0.1×106± 4.94 -3.4×104 - -2.8×105 -0.1×106± 4.94 -3.4×104 - -2.8×105 -0.1×106± 4.94 -3.4×104 - -2.8×105
7. Binding energy(ΔG) (kcal/mol) NA NA -6.2 NA -4.1 NA -49.4 NA
Stability of any complex structure is generally checked by its RMSD value with respect to the reference carbon backbone structure. For the present case we have used the reference backbone structure of receptor protein 6LU7. Figure 5 a has shown the 3D view of RMSD values for carbon backbone (6LU7) and of O:6LU7. Figure 5b has shown the 2D view of RMSD values for all ligand: protein complex structures (O:6LU7, F:6LU7, H:6LU7, O + F + H:6LU7) with respect to 6LU7. RMSD variations were observed between 0.09-0.48 nm, 0.10–0.40 nm 0.23–0.24 nm for O:6LU7, F:6LU7 and H:6LU7 complex structures compared to appo 6LU7 variation of 0.13–0.32 nm. Similarly, for O + F + H:6LU7 combination mode, the RMSD variation was observed between 0.10-0.40 nm. The RMSD data indicates the existence of less fluctuation and better stability of the ligand: 6LU7 complex formation (Table 4, SD3). The flatness in 2D contour for H:6LU7 complex structure reveal that in presence of H, during complexation host 6LU7 does not show any significant change (Figure 5b).Figure: 5 a) Root mean square deviation (RMSD) graphs for bare state of receptor protein 6LU7 and in complex (O:6LU7) with 6LU7 3D view up to 10000ps and b) 2D for complex (O:6LU7, F:6LU7 and H:6LU7and O + F + H: 6LU7) view up to 10000ps.
Figure: 5
RMSF has shown average values of 0.33, 0.29, 0.16, 0.18 nm for O:6LU7, F:6LU7, H; 6LU7 and O + F + H:6LU7 complex structures whereas appo 6LU7 has shown the average value as 0.23 nm, which indicates the protein backbone (Figure 6 a) was never get effected due to complexation with any of the selected inhibitor drugs. For further verification of the stability and compressed nature of each complex structures, we have tested the Rg data also for all proposed inhibitors [60]. Rg data has shown the fluctuations between 2.16-2.24 nm, 2.20–2.21 nm, 2.20–2.22 nm and 2.12–2.25 nm for O:6LU7, F:6LU7, H:6LU7 and O + F + H:6LU7 complex structures whereas 2.13–2.24 nm for bare 6LU7 structure. The average Rg were observed in between 2.18 to 2.21nm (Figure 6b, Table 4). Small variation of Rg values showed that all our tested drugs in complex structure with target protein 6LU7 are quite stable and compressed in nature (Figure 6b, Table 4). Among all of them O:6LU7 shows better stability after complex formation due to its compactness with respect to bare protein 6LU7 (2.18 nm).Figure: 6 a): RMSF of 6LU7 in its bare state and of complex (O:6LU7, F:6LU7, H:6LU7 and O + F + H: 6LU7) structure, b) Rg of 6LU7 in its bare state and of complex (O:6LU7, F:6LU7, H:6LU7 and O + F + H: 6LU7) structure.
Figure: 6
Number of nonbonded interactions are directly related to the stability of ligand: receptor protein complex structure. For present MD simulation work we have used a 3.5 Å cut-off condition for identification of proper nonbonded hydrogen bonded interaction. We have observed the presence of 0–5 number of hydrogen bonded interactions for different complex combinations (Table 4, Figure 7 a, SD3). For the present work solvent Accessible Surface areas (SASA) also have been computed to get the information about the area of receptor contact to the solvents (Figure 7b, Table 4). Smaller value of SASA means ligand is mostly surrounded and covered by the receptor protein means better inhibitor: receptor complexation. For different complex structures O:6LU7, F:6LU7, H:6LU7 and O + F + H:6LU7, the SASA values were observed between 6.5-9, 7–9, 7.3–9 and 6–11 nm2 with average value 8, 8, 8.2 and 8.3 nm2. For appo protein structure it was computed between 30-35 nm2 with a mean value of 33 nm2. Oseltamivir and Favipiravir in complex forms have shown the minimum SASA values. Favipiravir already had proved its good clinical efficacy against COVID-19 [61]. Similarly clinical trial data shows Oseltamivir is one of the effective drugs used for treating pregnant COVID Patients [62].Figure 7 a): Intermolecular Hydrogen bond numbers for complex (O:6LU7) structure for the total time trajectory 0–10000 ps b) SASA area for protein 6LU7 in its bare state and for complex (O:6LU7, F:6LU7, H:6LU7 and O + F + H:6LU7) structure.
Figure 7
MM/PBSA method has been applied for the present study for checking the inhibitor's nonbonding binding affinity towards target protein (Table 4). By MM/PBSA method, change in free binding energy (ΔGbind) indicate the nonbonding interaction energies of the binding region for the complex formation which is dependent on energies of the complex in vacuum, in nonpolar solvation state and in polar solvation states (Eqs. (6), (7), (8), and (9)). The average ΔGbind values of -6.2, -4.1, -49.4 and -51.3 kcal/mol were computed for O:6LU7, F:6LU7, H:6LU7 and O + F + H:6LU7. Simulation result on binding energy verified that in individual and combination mode, presence of H has showed a stronger affinity towards 3CLpro protease. Several reported clinical trial data also validated the positive effectivity of H for the treatment of hospitalized COVID-19 patients. Submitted report by Mokhtari et al. [63] suggested that those COVID positive patients who were treated with H have shown reduced period of hospital stay.
4 Conclusion
For the present study we have tried to recognize the mechanism of action and effectivity of some repurposed common antiviral and antibiotic individual and combination drugs Favipiravir, Hydroxychloroquine and Oseltamivir against COVID-19 disease. The properties like physiochemical, medical chemistry and pharmacokinetics from ADME analysis have found a strong inhibitory opportunity of O, F, H towards SARS-CoV-2 protein 3CLpro. The results of molecular docking validated the good binding affinity of all tested individual drugs: O, F and H and combination drugs: F + O, F + H and O + F + H towards CoV-2 virus. Existence of strongest binding affinity (−6.2 kcal/mol) and lowest inhibition constant (2.8 × 10−5 M) established the possibility of better complexation of O with 6LU7 protease. Molecular dynamics simulation data for the whole-time scale (0–100 ns) validated the stability of all tested repurposed inhibitors in terms of their different thermodynamical parameters (Epot, T, V, D, interaction energies, Rg, ΔGbind, SASA energy). Trajectories analysis showed that all studied complexes have displayed structural stability during the molecular docking and MD runs. Among all, the perfect closeness of average RMSD of O:6LU7, F:6LU7 (0.23 nm, 0.22 nm) and host 6LU7 (0.22 nm) confirmed the total inheritance of Favipiravir and Oseltamivir inside host protein whereas RMSF variation data clearly showed that H:6LU7 and F:6LU7 complex structures does not affected the protein backbone. Lowest SASA energy for all tested inhibitors in their individual modes (∼8 nm2) validated the best stability of these individual drugs F, O, H towards receptor 6LU7 compared to their combination modes. From the simulated data of binding affinity by MM/PBSA method it is observed that H showed higher affinity than F and O against COVID 19 main protease. Supported by previous literatures and our in silico results we may conclude that all our tested repurposed drugs may act as an inhibitor for the main protease of SARS-CoV-2 since they all have considerable effect but none of them has shown to be fully effective against the CoV-2 infection. We hope our current comparative study of repurposed drugs: Favipiravir, Oseltamivir and Hydroxychloroquine in their individual and combinational modes against CoV-2 infections will help the current clinical research to develop an effective drug for the cure for COVID -19 treatment.
Declaration
Author contribution statement
Pooja Yadav: Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.
Papia Chowdhury: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper.
Funding statement
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data availability statement
Data included in article/supplementary material/referenced in article.
Declaration of interests statement
The authors declare no competing interests.
Additional information
No additional information is available for this paper.
Appendix A Supplementary data
The following is the supplementary data related to this article:Revised Supporting document.docx
Revised Supporting document.docx
==== Refs
References
1 Wang D. Hu B. Hu C. Zhu F. Liu X. Zhang J. Wang B. Xiang H. Cheng Z. Xiong Y. Zhao Y. Li Y. Wang X. Peng Z. Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus–infected pneumonia in Wuhan, China JAMA 323 11 2020 1061 32031570
2 Huang C. Wang Y. Li X. Ren L. Zhao J. Hu Y. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China Lancet 395 2020 497 506 31986264
3 https://www.who.int.
4 Rothe C. Schunk M. Sothmann P. Bretzel G. Froeschl G. Wallrauch C. Zimmer T. Thiel V. Janke C. Guggemos W. Seilmaier M. Drosten C. Vollmar P. Zwirglmaie K. Zange S. WeOlfel Hoelscher M. Transmission of 2019-nCoV infection from an asymptomatic contact in Germany, New England J. Med. 382 10 2020 970 971
5 Furukawa N.W. Brooks J.T. Sobel J. Evidence supporting transmission of severe acute respiratory syndrome coronavirus 2 while presymptomatic or asymptomatic Emerg. Infect. Dis. 26 2020 7
6 Chunyang Li. Asymptomatic and human-to-human transmission of SARS-CoV-2 in a 2-family cluster. Xuzhou, China Emerg. Infect. Dis. 26 7 2020 1626 1628 32228809
7 Chang C. Lo S.C. Wang Y.S. Hou M.H. Recent insights into the development of therapeutics against coronavirus diseases by targeting N protein Drug Discov. Today 21 4 2016 562 572 26691874
8 Grein J. Ohmagari N. Shin D. Diaz G. Asperges E. Castagna A. Feldt T. Compassionate use of remdesivir for patients with severe Covid-19 N. Engl. J. Med. 2020 1 10
9 McIntyre P. Joo Y.J. Chiu C. Flanagan K. Macartney K. COVID-19 vaccines–are we there yet? Aust. Prescr. 44 2021 19 33664546
10 Celik I. Erol M. Duzgun Z. In silico evaluation of potential inhibitory activity of remdesivir, favipiravir, ribavirin and galidesivir active forms on SARS-CoV-2 RNA polymerase Mol. Divers. 26 1 2021 279 292 33765239
11 Kabir E.R. Mustafa N. Nausheen N. Siam M.K.S. Syed E.U. Exploring existing drugs: proposing potential compounds in the treatment of COVID-19 Heliyon 7 2 2021 e06284
12 Chowdhury P. Pathak P. Neuroprotective immunity by essential nutrient “choline” for the prevention of SARS CoV2 infections: an in silico study by molecular dynamics approach Chem. Phys. Lett. 761 2020 138057
13 Cardoso W.B. Mendanha S.A. Molecular dynamics simulation of docking structures of SARS-CoV-2 main protease and HIV protease inhibitors J. Mol. Struct. 1225 2021 129143
14 Baron S. Fons M. Albrecht T. Viral pathogenesis Chapter 45 Baron S. Medical Microbiology fourth ed. 1996 University of Texas Medical Branch at Galveston Galveston (TX) Available from: https://www.ncbi.nlm.nih.gov/books/NBK8149/
15 Fried M.W. Shiffman M.L. Reddy K.R. Smith C. Marinos G. Gonçales F.L. Jr. Craxi A. Peginterferon alfa-2a plus ribavirin for chronic hepatitis C virus infection N. Engl. J. Med. 347 13 2002 975 982 12324553
16 Pagadala N.S. AZT acts as an anti-influenza nucleotide triphosphate targeting the catalytic site of A/PR/8/34/H1N1 RNA dependent RNA polymerase J. Comput. Aided Mol. Des. 33 2019 387 404 30739239
17 Kaur H. Shekhar N. Sharma S. Ivermectin as a potential drug for treatment of COVID-19: an in-sync review with clinical and computational attributes Pharmacol. Rep. 73 2021 736 749 33389725
18 Rainsford K.D. Parke A.L. Clifford-Rashotte M. Therapy and pharmacological properties of hydroxychloroquine and chloroquine in treatment of systemic lupus erythematosus, rheumatoid arthritis and related diseases Inflammopharmacology 23 2015 231 269 26246395
19 Costanzo M. Anna Rachele De Giglio M. NicolaRoviello G., SARS-CoV-2: recent reports on antiviral therapies based on lopinavir/ritonavir, darunavir/umifenovir, hydroxychloroquine, remdesivir, favipiravir and other drugs for the treatment of the new coronavirus Curr. Med. Chem. 27 2020 4536 4541 32297571
20 Shadrack D.M. Deogratias G. Kiruri L.W. Onoka I. Vianney J.M. Swai H. Nyandoro S.S. Luteolin: a blocker of SARS-CoV-2 cell entry based on relaxed complex scheme, molecular dynamics simulation, and metadynamics J. Mol. Model. 27 8 2021 1 15
21 Falzarano D. Wit De W. Rasmussen A.,L. Feldmann F. Okumura A. Scott D.P. Benecke A.G. Treatment with interferon-α2b and ribavirin improves outcome in MERS-CoV–infected rhesus macaques Nat. Med. 19 10 2013 1313 1317 24013700
22 Chan W.J.F. Yao Y. Yeung M.L. Deng W. Bao L. Jia L. Cai J.P. Treatment with lopinavir/ritonavir or interferon-β1b improves outcome of MERS-CoV infection in a nonhuman primate model of common marmoset J. Infect. Dis. 212 12 2015 1904 1913 26198719
23 Hart B.J. Dyall J. Postnikova E. Zhou H. Kindrachuk J. Johnson R.F. Olinger G.G. Frieman M.B. Holbrook M.R. Jahrling P.B. Hensley L. Interferon-β and mycophenolic acid are potent inhibitors of Middle East respiratory syndrome coronavirus in cell-based assays J. Gen. Virol. 95 2014 571 577 24323636
24 Mageed A.E. Doaa H.R. Abdelrheem A.S. Ahmed A.A. Rahman R.K.M.M.E. Ahmed S.A. El-Bassuony A.A. Mohamed H.S. Combination and tricombination therapy to destabilize the structural integrity of COVID-19 by some bioactive compounds with antiviral drugs: insights from molecular docking study Struct. Chem. 2021 1 16
25 Yadav P. Rana M. Chowdhury P. DFT and MD simulation investigation of favipiravir as an emerging antiviral option against viral protease (3CLpro) of SARS-CoV-2 J. Mol. Struct. 2021 131253
26 Shannon A. Selisko B. Le N.T.T. Huchting J. Touret F. Piorkowski G. Canard B. Favipiravir strikes the SARS-CoV-2 at its Achilles heel. the RNA polymerase bioRxiv 2020
27 Kaur R.J. Charan J. Dutta S. Sharma P. Bhardwaj P. Sharma P. Misra S. Favipiravir use in COVID-19: analysis of suspected adverse drug events reported in the WHO database Infect. Drug Resist. 13 2020 4427 33364790
28 Dabbous H.M. Abd Elsalam S. El-Sayed M.H. Sherief A.F. Ebeid F.F. Abd El Ghafar M.S. Soliman S. Elbahnasawy M. Badawi R. Tageldin M.A. Efficacy of favipiravir in COVID-19 treatment: a multi-center randomized study Arch. Virol. 166 2021 949 954 33492523
29 Pertinez H. Rajoli R. Khoo S.H. Owen A. Pharmacokinetic modelling to estimate intracellular favipiravir ribofuranosyl-5'-triphosphate exposure to support posology for SARS-CoV-2 J. Antimicrob. Chemother. 76 8 2021 2121 2128 34075418
30 Joshi S. Jalil P. Abdul A. Agam V. Deepak T. Mangesh T. Saiprasad P. Hanmant B. Role of favipiravir in the treatment of COVID-19 Int. J. Infect. Dis. 102 2021 501 508 33130203
31 Fuzimoto A.D. An overview of the anti-SARS-CoV-2 properties of Artemisia annua, its antiviral action, protein-associated mechanisms, and repurposing for COVID-19 treatment Journal of integrative medicine 19 5 2021 375 388 34479848
32 Bansal P. Goyal A. Cusick IV A. Lahan S. Dhaliwal H.S. Bhyan P. Bhattad P.B. Aslam F. Ranka S. Dalia T. Chhabra L. Hydroxychloroquine: a comprehensive review and its controversial role in coronavirus disease 2019 Ann. Med. 53 1 2021 117 134 33095083
33 D’Acquarica I. Agranat I. Chiral switches of chloroquine and hydroxychloroquine: potential drugs to treat COVID-19 Drug Discov. Today 25 2020
34 Lee Z. Rayner C.R. Forrest J.I. Nachega J.B. Senchaudhuri E. Mills E.J. The rise and fall of hydroxychloroquine for the treatment and prevention of COVID-19 Am. J. Trop. Med. Hyg. 104 2021 35 33236703
35 Thomson K. Nachlis H. Emergency use authorizations during the COVID-19 pandemic: lessons from hydroxychloroquine for vaccine authorization and approval JAMA 324 13 2020 1282 1283 32870235
36 Hurt A.C. Holien J.K. Parker M.W. Oseltamivir resistance and the H274Y neuraminidase mutation in seasonal. Pandemic and highly pathogenic influenza viruses Drugs 69 2009 2523 2531 19943705
37 Chiba S. Effect of early oseltamivir on outpatients without hypoxia with suspected COVID-19 Wien Klin. Wochenschr. 133 2021 292 297 33296027
38 Diego V. Paño-Pardo J.R. Pachón J. Riera M. López-Medrano F. Payeras A. Carmen Fariñas M. Timing of oseltamivir administration and outcomes in hospitalized adults with pandemic 2009 influenza A (H1N1) virus infection Chest 140 2011 1025 1032 21415133
39 Wu F. Zhao S. Yu B. Chen Y.-M. Wang W. Song Z.G. Hu Y. Tao Z.W. Tian J.H. Pei Y.Y. Yuan M.L. Zhang Y.L. Dai F.H. Liu Y. Wang Q.M. Zheng J.J. Xu L. Holmes E.C. Zhang Y.Z. A new coronavirus associated with human respiratory disease in China Nature 579 7798 2020 265 269 32015508
40 Choudhary M.I. Shaikh M. tul-Wahab A. ur-Rahman A. In silico identification of potential inhibitors of key SARS-CoV-2 3CL hydrolase (Mpro) via molecular docking, MMGBSA predictive binding energy calculations, and molecular dynamics simulation PLoS One 15 7 2020 e0235030
41 Di Micco S. Musella S. Scala M.C. Sala M. Campiglia P. Bifulco G. Fasano A. In silico analysis revealed potential anti-SARS-CoV-2 main protease activity by the zonulin inhibitor larazotide acetate Front. Chem. 8 2021 1271
42 Gupta S. Singh A.K. P Kushwaha P. Prajapati K.S. Shuaib M. Senapati S. Kumar S. Identification of potential natural inhibitors of SARS-CoV2 main protease by molecular docking and simulation studies J. Biomol. Struct. Dyn. 2020 1 12
43 Nayim Sepay Anti-COVID-19 terpenoid from marine sources: a docking, admet and molecular dynamics study J. Mol. Struct. 1228 2021 129433
44 Burley S.K. Berman H.M. Bhikadiya C. Bi C. Chen L. Di Costanzo L. Christie C. Dalenberg K. Duarte J.M. Dutta S. Feng Z. Ghosh S. Goodsell D.S. Green R.K. Guranovi_c V. Guzenko D. Hudson B.P. Kalro T. Liang Y. Zardecki C. RCSB Protein Data Bank: biological macromolecular structures enabling research and education in fundamental biology, biomedicine, biotechnology and energy Nucleic Acids Res. 47 D1 2019 D464 D474 30357411
45 Dassault Syst`emes Biovia Discovery Studio Modeling Environment 2017 Release San Diego Dassault Syst`emes
46 Morris G.M. Huey R. Lindstrom W. Sanner M.F. Belew R.K. Goodsell D.S. Olson A.J. AutoDock 4 and auto dock Tools 4: automated docking with selective receptor flexibility J. Comput. Chem. 30 16 2009 2785 2791 19399780
47 Tools A. 1.5. 6 (ADT)/MGL Tools 1.5. 6 2012
48 Daina A. Michielin O. Zoete V. Swiss ADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules Sci. Rep. 7 2017 42717
49 Lipinski C.A. Lead- and drug-like compounds: the rule-of-five revolution Drug Discov. Today Technol. 1 4 2004 337 341 24981612
50 Becke A. A new inhomogeneity parameter in density-functional theory J. Chem. Phys. 10 20 2002 8554 8560
51 Frisch M.J. Gaussian 09, Revision D.01 2004 Gaussian Inc
52 Dennington R. Keith T. Millam J. Gauss View 2007 Semichem, Inc. Shawnee Mission, KS Version 4.1.2
53 Benson N.C. Daggett V. A comparison of multiscale methods for the analysis of molecular dynamics simulations J. Phys. Chem. B 116 2012 29 8722–8731
54 Gunsteren W.F.V. Billeter S.R. Eising A.A. Hünenberger P.H. Krüger P. Mark A.E. Scott W.R.P. Tironi I.G. Biomolecular Simulation: the GROMOS96 Manual and User Guide 1996 Hochschulverlag AG an der ETH Zürich Zürich
55 Pooja M. Rana Chowdhury P. Influence of size and shape on optical and electronic properties of CdTe quantum dots in aqueous environment AIP Conference Proceedings 2136 2019 AIP Publishing LLC 040006
56 Rashmi K. Rajendra K. Andrew L. g_mmpbsa-a GROMACS tool for high-throughput MM-PBSA calculations J. Chem. Inf. Model. 54 7 2014 1951 1962 24850022
57 Stephen Lm L.M. DREIDING: a generic force field for molecular simulations J. Phys. Chem. 94 1990 8897 8909
58 Selvaraj C. Panwar U. Dinesh D.C. Boura E. Singh P. Dubey V.K. Singh S.K. Microsecond MD simulation and multiple-conformation virtual screening to identify potential anti-COVID-19 inhibitors against SARS-CoV-2 main protease Front. Chem. 8 2021 595273
59 Martorana Annamaria Gentile Carla Lauria Antonino In silico insights into the sars cov-2 main protease suggest nadh endogenous defences in the control of the pandemic coronavirus infection Viruses 12 8 2020 805 32722574
60 Dharmendra K.Y. Molecular insights into the interaction of RONS and Thieno [3,2-c]pyran analogs with SIRT6/COX-2: a molecular dynamics study Sci. Rep. 8 2018 4777 29556059
61 Xinhuanet Favipiravir shows good clinical efficacy in treating COVID-19 Official [Online]. Available at: http://www.xinhuanet.com/english/2020-03/17/c_138887971.htm 2020 Accessed 03 03, 2021
62 Wu R. Wang L. Kuo H.D. Shannar A. Peter R. Chou P.J. An update on current therapeutic drugs treating COVID-19 Curr Pharmacol Rep 2020 1 15 [PMC free article] [PubMed] [Google Scholar
63 Mokhtari M. Mohraz M. Gouya M.M. Namdari Tabar H. Tabrizi J.S. Tayeri K. Clinical outcomes of patients with mild COVID-19 following treatment with hydroxychloroquine in an outpatient setting Int. Immunopharm. 96 2021 107636
| 0 | PMC9737521 | NO-CC CODE | 2022-12-16 23:18:17 | no | Heliyon. 2022 Dec 10; 8(12):e12327 | utf-8 | Heliyon | 2,022 | 10.1016/j.heliyon.2022.e12327 | oa_other |
==== Front
Heliyon
Heliyon
Heliyon
2405-8440
The Author(s). Published by Elsevier Ltd.
S2405-8440(22)03485-5
10.1016/j.heliyon.2022.e12197
e12197
Research Article
Covid-19 waste facemask conundrum: A facile way of utilization through fabricating composite material with unsaturated polyester resin and evaluation of its mechanical properties
Bin Mobarak Mashrafi a
Hossain Md. Sahadat a
Chowdhury Fariha b
Ahmed Samina ac∗
a Institute of Glass & Ceramic Research and Testing (IGCRT), Bangladesh Council of Scientific and Industrial Research (BCSIR), Dhaka 1205, Bangladesh
b Biomedical and Toxicological Research Institute (BTRI), Bangladesh Council of Scientific and Industrial Research (BCSIR), Dhaka 1205, Bangladesh
c BCSIR Laboratories Dhaka, Bangladesh Council of Scientific and Industrial Research (BCSIR), Dhaka 1205, Bangladesh
∗ Corresponding author.
10 12 2022
12 2022
10 12 2022
8 12 e12197e12197
4 7 2022
11 9 2022
30 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.
Since the outbreak of novel coronavirus (COVID-19), the use of personal protective equipment (PPE) has increased profusely. Among all the PPEs, face masks are the most picked ones by the mass people for protective purpose. This spawned extensive daily use of face masks and production of masks had to augment to keep up this booming demand. Such extensive use of face masks has resulted in a huge waste generation. Lack of proper disposal, waste management and waste recycling have already led this waste to pervade in the environment. In quest of finding a solution, here in this research, a composite material was fabricated utilizing waste face mask (WFM) with unsaturated polyester resin (UPR) and the mechanical properties were evaluated. The composites were fabricated by incorporating 1%, 2%, 3%, 4% and 5% WFM (by weight) within the UPR matrix in the shredded form following hand lay-up technique. Tensile properties, i.e., tensile strength (TS), tensile modulus (TM) and percentage elongation at break (% EB) as well as flexural properties, i.e., bending strength (BS) and bending modulus (BM) were evaluated for the fabricated composites. According to the results obtained, the 2% WFM loaded composites showed highest values of TS, TM, BS and BM which are 31.61 N/mm2, 1551.41 N/mm2, 66.53 N/mm2 and 4632.71 N/mm2 respectively. These values of 2% WFM loaded composite are 69.58%, 107.78%, 129.49% and 152% higher than the values of the control sample (UPR). Such results depict the successfulness of WFM's incorporation as a reinforcing material in the composite materials. Attenuated Total Reflectance-Fourier transform infrared spectroscopy (ATR-FTIR), scanning electron microscopy (SEM), water uptake and thickness swelling tests were also carried out for the fabricated composites. FTIR of the collected WFM revealed the fiber to be of polypropylene and the existing functional groups were also identified. The SEM images confirmed the proper adhesion of WFM and UPR in terms of mechanical bonding rather than chemical bonding. Water absorption and dimension change was investigated by water uptake and thickness swelling test. To sum up, the way we have utilized WFM as a reinforcing agent in a composite material, this could be a possible solution for the face mask's waste conundrum.
Graphical abstract
Image 1
COVID-19 pandemic; Face mask waste; Composite material; Hand lay-up technique; Mechanical properties.
Keywords
COVID-19 pandemic
Face mask waste
Composite material
Hand lay-up technique
Mechanical properties
==== Body
pmc1 Introduction
The world has trembled with the outbreak of the severe acute respiratory syndrome or novel coronavirus (COVID-19) emerged from Wuhan, China back in December 2019 which was followed by SARS-CoV in 2003 and MERS-CoV in 2012 [1]. Till writing this manuscript, COVID-19 pandemic has caused demise to nearly 6.14 millions of people with 486 million confirmed cases and excruciating sufferings to countless people [2]. Right after the declaration as a pandemic form World Health Organization (WHO) [3, 4] many countries across the world have undertaken various measures to suppress the spread of this deadly virus including strict lockdown (staying at home) [5]. Strict lockdowns were withdrawn with the lessening of spread of this virus and other preventive measures like quarantine, isolation, social distancing, frequent washing of hands, avoiding mass gathering, use of personal protective equipment (PPE) (facial masks, gloves, face shields, wet wipes, gowns, shoes) etc. have been in effect according to the guidelines of WHO [6, 7, 8, 9, 10]. Among all the PPE's, face masks are the most prevalent since it is favored by both the health-care workers and mass people following the guidelines of WHO [11, 12]. The sudden use of face mask by the mass people with the hope of slowing down the transmission of coronavirus has not only led to global shortage [13, 14] but also raising concern for the colossal waste that is being produced [15, 16]. Many countries have boosted the production of disposable face mask and eased its way of distribution throughout the world [17]. This only adds more concern for the environmental pollution since improper disposal of facemask has been reported worldwide [18, 19, 20, 21, 22].
According to an international online survey, 19% of individuals throw away their disposable facemasks recklessly and even only 1% of disposable face masks by the world population would mean ∼10 million face masks (30,000–40,000 kg) discarded in the environment [23]. Since proper disposal of WFM aren't being ensured, scientists across the world have been trying to come up with alternative ideas for the utilization of such wastes. Asim et al. [24] reviewed the potential valorization techniques for discarded WFMs while Aragaw and Mekonnen et al. [25] have demonstrated a possible mitigation technique for waste-to-energy conversion using pyrolysis technique. In addition to that, Rehman et al. [26] utilized WFM with fat clay, making it an amalgamated binary admixture for better mechanical characteristics and WFM served as an additive in this regard. This whole idea of utilizing WFM into anything with potentiality might serve to mitigate the disposal problem.
Keeping this in mind, we have come up with a way to utilize WFM by making a composite material that will serve many purposes. According to Karbhari et al. [27], a composite material is a macroscopic combination of two or more distinct materials having a finite interface between them. Polymeric composite materials are usually fabricated by a combination of two phases; one is matrix which is normally organic polymer-based compound and the other one is the reinforcing material. The polymeric matrix can be either thermoplastic or thermosetting in nature. Most widely used thermoplastic matrixes are polyethylene, polypropylene (PP), poly vinyl chloride (PVC) while the most widely used thermosetting matrixes are polyester, epoxy and phenolic resins [28, 29]. The reinforcing materials that have been widely used are glass, Kevlar, aramid, carbon, natural fiber etc. [30, 31]. Such fabrication of composite using reinforcing material results in better mechanical properties than the matrix alone [32, 33]. The penchant of researchers has diverged from monolithic materials to fiber reinforced composite (FRCM) materials which is due to the nonpareil advantages of FRCM such as high strength to weight ratio, non-corrosive property and high fracture toughness [34, 35]. These advantages could serve the automotive and construction industries since they are always in quest of lighter and stronger materials of lower cost [36].
Here in this work, we have fabricated a composite material utilizing WFM and UPR. UPR is an extensively used polymeric matrix, mainly used in composite making [37, 38, 39, 40, 41], in flame retardation [42, 43, 44], in road construction [45], as corrosion resistant coating of steel [46], in furniture making [47, 48, 49] etc. owing to its noteworthy mechanical properties, dielectric properties, heat and chemical resistance, low cost, high gloss, ease of processing and good performance [50, 51]. The present study strives at preparation, characterization and evaluation of mechanical properties of the fabricated composite and utilization of WFM into a potential material was a perquisite.
2 Materials and methods
2.1 Materials
The polymeric matrix which is general purpose unsaturated polyester resin (UPR) and curing agent methyl ethyl ketone peroxide (MEKP) (Table 1 ) were purchased from Lucky Acrylic & Fiber, Dhaka, Bangladesh. MEKP functions as a hardener by cross linking the UPR matrix. These chemicals were used as received. The WFMs were collected from waste bins of different laboratories and households of close acquaintances. The chemical structure of UPR and MEKP is shown in Table 1 according to references.Table 1 Structure of unsaturated polyester resin and methyl ethyl ketone peroxide.
Table 1Chemical Name Unsaturated Polyester Resin Methyl Ethyl Ketone Peroxide
Acronym UPR MEKP
Structure Image 1 Image 2
Reference Recreated following: [52] Recreated following: [53]
2.2 WFM processing
Among all the collected WFM, only the three layer surgical masks were sort out and kept secluded under sunlight for about 7 days [54] to eliminate any chances of initial contamination. After 7 days at seclusion, the masks were then subjected to washing using detergent-water followed by sun drying for 2 days. The WFMs were then cut in 10 cm by 10 cm dimension to eliminate elastic ear bands, nose clip and stitches and weighed. Again, the masks were sorted out based on the grams per square meter (gsm) and only 80 gsm three layered WFMs were selected for composite fabrication. The next step involves the manual shredding of WFM into fragments of roughly 1 mm by 10 mm and sorting of oversized fragments were done with manual visualization.
2.3 Composite fabrication
The shredded WFM fragments were mixed with UPR in a beaker and vigorous stirring was continued for 10–15 min to make sure of proper immixture of the components. This was followed by the addition of 1% (w/w) of curing agent (MEKP) into the mixture and vigorous stirring was continued for 2–3 min. The shredded WFM loading was varied to 1%, 2%, 3%, 4% and 5% by weight and UPR was added accordingly.
The composite mixture was then poured into a pre-made mold frame of known thickness and the mold frame was placed over a glass sheet. Non-stick Teflon sheet was used beneath and above the mold frame to eliminate adhesion of the sample with the glass sheet. Another glass sheet was used on top of the mold frame making it a sandwich and composite was fabricated following the hand lay-up technique (Figure 1 ) [55]. The sample mixture was cured for 24 h inside a fume hood at temperature ∼25 °C. Finally, the composite was taken out of the mold frame, cut into appropriate dimension using a band saw and stored in air tight polyethylene bag.Figure 1 WFM and UPR based composite fabrication using hand lay-up technique.
Figure 1
2.4 Test methods
2.4.1 Tensile and flexural test
The specimens for tensile and flexural tests were prepared in accordance to the ASTM D638-01 [56] and ASTM D790-10 [57] test methods. The tensile properties, i.e., tensile strength (TS), tensile modulus (TM) and percentage elongation at break (%EB) were evaluated by using a universal testing machine (Testometric M-500-30 KNCT, UK). The initial clamp separation of the machine was 20 mm and the cross-head speed was fixed at 20 mm/min. A 3000 kgf load cell was used in this regard and the tests were halted whenever failure occurred. The flexural tests, i.e. bending strength (BS) and bending modulus (BM) were also carried out with the aid of the same instrument. The cross-head speed was fixed at 10 mm/min and a 100 kgf load cell was used. The span distance of the three-point bending rig was 40 mm and the tests were halted whenever failure occurred. All the tests were carried out at the same conditions and the samples were kept in air tight polyethylene bag prior to any test. For better accuracy of the mechanical tests, five or more samples were evaluated for each test and the average value was considered. The resultant was a Time (sec) vs Force (N) curve showing values of Force @break (N) and Elongation @break (mm). The T.S, T.M and EB were calculated using the following Eqs. (1), (2), and (3).(1) T.S=Force@breakArea
(2) %E.B=Elongation@breakSamplelength×100
(3) T.M=T.SE.B
2.4.2 Functional group analysis
The existing functional groups of the fabricated composites were detected by following fourier transform infrared spectroscopic (FTIR) analysis using IR Prestige-21 (Shimadzu Corporation, Japan) mounted with MIRacle-10 Single Reflection ATR (Attenuated Total Reflectance) accessory. Zinc selenide (ZnSe) prism plate was used for this single reflection. The analysis was done in the wavenumber range of 400–4000 cm−1 with a total number 30 scans per sample in the transmittance mode. Happ-Genzel apodization technique was implemented by the instrument where the resolution was 4 cm−1.
2.4.3 Surface morphology analysis
The scanning electron microscopic (SEM) analysis was carried out to have a look into the surface morphology of the samples using Phenom Pro Desktop (Phenom 1481) at an accelerating voltage of 15 kv. Composite samples that were fractured during the tensile tests were selected in this regard and the images of the fractured surface were recorded.
2.4.4 Water uptake test
The water uptake test was carried out following ASTM D570-98(2018) test method [58]. Initially, the composites were cut into small dimensions, cleaned to remove any dust or loose particle and weighed using an electrical analytical balance. Then they were submerged in 500 ml de-ionized (DI) water that was contained in a beaker and static motion and ambient temperature of fluid was ensured. The submergence was maintained for 30 days and the composite samples were withdrawn from the DI water periodically. After withdrawing for a brief time, the adherent water was removed using tissue paper, weighed and submerged again. This was continued up-to 30 days and the water uptake was calculated using the following Eq. (4) [59],(4) WaterUptake(%)=Wwet−WdryWdry×100
Here, W wet = weight of the wet sample withdrawn at a certain time and W dry = weight of the initial dry sample.
2.4.5 Thickness swelling test
The thickness swelling test was also carried out to investigate the dimensional changes that might occur when the composite samples are in contact with water. The procedure for this test was similar to that of the water uptake test but instead of weighing, the thicknesses of the samples were recorded. This was continued periodically for 30 days and the thickness was measured using a digital Vernier Caliper. The thickness swelling was calculated using the following Eq. (5) [60],(5) Thicknessswelling(%)=Twet−TdryTdry×100
Here, T wet = thickness of the wet sample withdrawn at a certain time and T dry = thickness of the initial dry sample.
3 Results and discussion
3.1 Tensile tests
The mechanical property is one of the decisive factors for the applicability of a composite in the desired field. Hence, the tensile and flexural tests were carried out in order to evaluate the mechanical properties of the composite. Figure 2 shows the TS of the UPR (control) and composites fabricated with UPR and WFM with respect to WFM loading percentage. The percentage of WFM loading was 1%, 2%, 3%, 4% and 5% by weight. The tensile tests were carried out at the optimal direction and only the samples that broke at the specific area were considered for further analysis. According to the data obtained, the control sample that is only made of UPR with the aid of curing agent, showed TS of 18.64 N/mm2. The addition of shredded WFM into the UPR evidently increased the TS but in different magnitude. At 2% WFM loading, TS was 31.61 N/mm2 which was highest among all the WFM loading composites and 69.58% higher than the control sample. Further addition of WFM into the composite didn't surpass the TS of 2% WFM reinforced composite but was definitely higher than the control sample. Such decrease in the TS might be due to the fact that upon the addition of WFM, the stress transfer between the matrix (UPR) and the reinforced fiber (WFM) got lessened which ultimately affected the TS of the composite. At 5% WFM loading, the TS was 25.99 N/mm2 which was 39.43% higher than the control sample but 21.62% less than the 2% WFM reinforced composite.Figure 2 Tensile strength of UPR (control) and composites (UPR + WFM) with respect to WFM loading percentage.
Figure 2
Since the TS of composite materials are dependent on the reinforcing material and amalgamation of WFM onto UPR resin has increased such, hence WFM evidently improved the mechanical property of the composite by acting as a reinforcing material.
The tensile modulus (TM) is one of the features that describes the stiffness of the composite material. Likewise to that of the TS tests, the increase of WFM loading also increased the TM of the composite. Figure 3 shows the TM values of UPR (control) and the WFM reinforced composites at different WFM loading percentages. The TM values of the reinforced composites were higher than that of the control sample. TM achieved apex at 2% WFM loading and further addition decreased the TM up to 5% WFM loading. This was anticipated since the TS values also exerted the similar phenomena. The 2% WFM reinforced composite showed highest value (1551.41 N/mm2) which was 107.78% higher than the control sample (746.64 N/mm2).Figure 3 Tensile modulus of UPR (control) and composites (UPR + WFM) with respect to WFM loading percentage.
Figure 3
The control samples with the lowest values of TS and TM have shown the highest value of percentage elongation @break (5.73%); meaning that, it stretches to the highest compared to the reinforced samples as a percentage of its initial dimension until it breaks. The percentage elongation @break (%EB) was calculated using Eq. (3) and shown in Figure 4 .Figure 4 Percentage elongation at break of UPR (control) and composites (UPR + WFM) with respect to WFM loading percentage.
Figure 4
According to the results obtained, 2% WFM loaded composite showed the highest %EB (3.65%) among the reinforced composites which was 36.3% less than that of the control sample. Further increase in the WFM loading lessened the %EB and very close values were found for 3%, 4% and 5% loading (2.07, 2.03 and 2.01% respectively). When compared with the % EB values of control and WFM loaded composites, the addition of WFM into the UPR matrix decreased the %EB. This might be due to the fact that, upon the addition of WFM, the interaction between WFM-WFM increases while UPR-WFM interaction decreases. Higher the WFM-WFM interaction, lower the %EB.
3.2 Flexural tests
The flexural properties in terms of bending strength and bending modulus were also evaluated. The bending strength is an important factor for composite materials for demanding applications. It is defined as the ability of a composite material to resist bending deflection when energy is applied to the structure [61]. Figure 5 represents the bending strength of UPR (control) and the WFM reinforced composites at different WFM loading percentages.Figure 5 Bending strength of UPR (control) and composites (UPR + WFM) with respect to WFM loading percentage.
Figure 5
Likewise to that of the tensile tests, flexural tests also transpired similar phenomena where the addition of WFM into the UPR matrix increased the BS. The control sample showed BS of 28.99 N/mm2 whereas 1% & 2% WFM reinforced composites showed 57.57 N/mm2 and 66.53 N/mm2 which were 98.58% and 129.49% greater than the control samples respectively. The BS of 2% WFM reinforced composite was the highest and further reinforcement lessened the BS. At higher loading percentages, the amount of matrix (UPR) is less and WFM is higher which invigorate the WFM-WFM interaction. On the contrary, the UPR-WFM interaction decreases which influences the flexural properties of the composite and begets lower values of BS. The phenomenon is similar for the bending modulus of the control and the WFM reinforced composites which has been shown in Figure 6 .Figure 6 Bending modulus of UPR (control) and composites (UPR + WFM) with respect to WFM loading percentage.
Figure 6
The BM illustrates the stress-strain ratio of the control and composite materials within their elastic region [62]. The 2% WFM reinforced composite with highest BS also had the highest BM (4632.71 N/mm2), 152% higher than that of the control sample (1838.36 N/mm2). Such significant modification of the flexural properties does indicate the succession of WFM reinforcement within the UPR matrix.
3.3 Functional group analysis
To get hold of the existing functional groups of the WFM, UPR and the WFM reinforced composites, ATR-FTIR spectrum was recorded which has been shown in Figure 7 . Generally, face masks are made of polypropylene, polyethylene, polystyrene, polyester, polycarbonate, polyurethane, polyacrylonitrile [63]. The classical 3-ply face masks are comprised of polypropylene or polyester that goes through melt blowing in order to make the non-woven fabrics of the face mask [64]. The ATR-FTIR spectrum that has been recorded for the outer layer of the WFM indicates its formation with polypropylene fibers [65, 66]. The symmetry deformation vibration peak of the methylene group on the aliphatic hydrocarbons appears at 1456 cm−1 (moderate intensity) and the methyl group vibration peak appears at 1375 cm−1 which are in compliance with the data reported in previous literature [25, 67].Figure 7 ATR-FTIR spectra of WFM, UPR, 1%, 2%, 3%, 4% and 5% WFM loaded composites.
Figure 7
For the spectrum of WFM, between the wavenumber 2800 cm−1 and 3000 cm−1, four adjacent peaks were observed. For the symmetric stretching vibrations of CH2 and CH3, two peaks were observed at 2918 cm−1 and 2949 cm−1. On the other hand, peaks at 2866 cm−1 and 2837 cm−1 corresponds to the asymmetric stretching vibrations of CH2 and CH3 [68,69]. According to the findings of Morent et al., the low intense peak at 1166 cm−1 can be assigned to CH3 asymmetric rocking, C–C asymmetric stretching and C–H wagging vibrations. The asymmetric rocking vibration exerted peak at the wavenumber of 997 cm−1. At the wavenumbers of 972 cm−1 and 898 cm−1 two peaks appeared for the vibration of C–C asymmetric stretching and CH3 asymmetric rocking and symmetric stretching. The rocking vibration of CH2 was confirmed from the peaks of 840 and 808 cm−1 and similar peaks were also reported [70].
The peak from 3200 cm−1 to 3600 cm−1 was identified which carried good evidence of the stretching vibration of –OH groups in the spectrum of polymer matrix (UPR). Less intense peaks at 2856 cm−1 to 2929 cm−1 were originated form the symmetric stretching vibration of C–H. One of the most intense peaks was appeared for the presence of carbonyl group (C=O) at the wavenumber of 1720 cm−1 and similar peak was reported elsewhere [37, 71, 72]. The stretching vibrations of aromatic C=C exerted peaks at 1598, 1577 and 1492 cm−1. Symmetric and asymmetric bending vibrations of the methyl (CH3) group were noticed at 1452 cm−1 and 1375 cm−1. The peaks around the fingerprint region (1273 cm−1, 1120 cm−1, 1068, cm−1 and 1041 cm−1) were identified for the existence of stretching vibrations of C–O group for ester molecules. Strong peaks at 698 cm−1, 742 cm−1 and 989 cm−1 were referred to the out-of-plane bending vibration of =C–H group in aromatic ring, vibrations of substituted aromatic ring and C–H out-of-plane bending vibration of trans –CH <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> CH– and similar peaks were documented elsewhere [73]. To summarize, all the vibrational assignments of the WFM and UPR has been shown in Table 2 .Table 2 Peak position and assignments of ATR-FTIR absorption bands of WFM and UPR.
Table 2Sample Name Peak Position (Wavenumber, cm−1) Assignments
WFM 2949 CH3 symmetric stretching
2918 CH2 symmetric stretching
2866 CH3 asymmetric stretching
2837 CH2 asymmetric stretching
1456 CH2 symmetry deformation
1375 CH3 bending
1166 C–C asymmetric stretching
CH3 asymmetric rocking
C–H wagging
997 CH3 asymmetric rocking
972 CH3 asymmetric rocking
C–C asymmetric stretching
898 CH3 asymmetric rocking
C–C asymmetric and symmetric stretching
840 CH2 rocking
808 CH2 rocking
UPR 3200–3600 –OH stretching
2856–2929 C–H symmetric stretching
1720 carbonyl group (C=O) (ester linkage)
1598, 1577 and 1492 C=C stretching within the aromatic ring
1452 CH3 asymmetric bending
1375 CH3 symmetric bending
1273, 1120, 1068 and 1041 C–O stretching (ester)
989 C–H out-of-plane bending of trans –CH=CH–
742 =C–H out-of-plane bending of aromatic ring
698 =C–H out-of-plane bending in singly substituted aromatic ring
Abbreviations: WFM = Waste Face Mask; UPR = Unsaturated Polyester Resin.
The four significant peaks of WFM 2949 cm−1, 2918 cm−1, 2866 cm−1 and 2837 cm−1 can also be observed in the FTIR spectra of 1%–5% WFM loaded composites which indicate the incorporation of WFM within the UPR matrix. Also, peaks at 997 cm−1 and 972 cm−1 can be observed within the WFM reinforced composites but at different intensities depending on the WFM loading.
3.4 Surface morphology analysis
The scanning electron micrographs of the WFM as well as the fractured surfaces of WFM loaded composites have been captured in order to investigate the interaction of the fibers with the polymer matrix. Figure 8 (a) shows the SEM image of a typical WFM which is comprised of three layers. The outer layer (1st layer) fibers are much dense than the fibers of outer layer (3rd layer). Fibers of the middle layer (2nd layer) are smaller in diameter and much dense than both the outer layers (1st and 3rd layer). Such fibril arrangement serves as a filtration system against dust or contaminant particles. The diameter of the fibers has been measured using imageJ software based on the SEM image. The average fiber diameter of the outer layer (1st layer), middle layer (2nd layer) and the outer layer (3rd layer) has been found to be 29.60 μm, 3.20 μm and 28.54 μm respectively. Figure 8(b) represents the SEM image of control which is the UPR matrix. The undisrupted surface of the control sample is smooth with no perforations or voids which is also confirmed by looking into the fractured surface. Figure 8(c)–(g) represents the fractured surfaces of the 1%–5% WFM loaded composites. The presence of fibers at the fractured surface is more apparent with increasing fiber loading. Very few holes can be seen at the fractured surfaces of the composites at lower WFM loading, indicating that the WFM fibers were strongly agglutinated with the UPR matrix which resulted in low pull out of the fibers. But slightly different scenario was observed for the composites with higher WFM loading.Figure 8 SEM images of (a) typical WFM; fractured surfaces of (b) control (UPR), (c) 1%, (d) 2%, (e) 3%, (f) 4% and (g) 5% WFM loaded composites.
Figure 8
This is due to the fact that, at higher WFM loading, the amount of UPR is decreased and this results in less agglutination between the fiber and UPR. It seems like an over-crowdedness or over-saturation of WFM fibers in the matrix and this consequently affects the mechanical property of the composites. Both the outer layer and middle layer WFM fibers are present at the fractured surfaces of the fabricated composites.
3.5 Water uptake test
The percentage of water uptake of the composite samples for a period of 30 days has been shown in Figure 9 . The test was carried out with four samples for each WFM loaded composites and the control sample from which an average value is obtained. According to the data obtained, the percentage of water uptake was found to be increased with increasing immersion time. Highest percentage of water uptake (1.0182 %) was found for 5% WFM loaded composites and the lowest percentage (0.9161 %) was found for 2% WFM loaded composites after 30 days of immersion.Figure 9 Water uptake of control (UPR) and WFM loaded composites.
Figure 9
It has been noticed that within the 30 days of test period, the percentage of water uptake didn't reach to any constant value and hence depicts the slowness of water uptake by the WFM loaded composites. Within the 15 days period, the rate increased swiftly which slowed down afterwards. An increase in WFM loading didn't necessarily increased the rate of water uptake. This may be due to a lot of reasons since the rate of water uptake greatly depends on temperature, time of immersion, loading of the WFM, orientation of WFM fiber, area of exposed surface, reaction between water and matrix, interfacial bonding, surface protection and voids, diffusivity etc. [74].
3.6 Thickness swelling test
The results of thickness swelling test have been shown in Figure 10 . This test was also carried out for a 30 days’ time period and thickness was measured periodically. The significance of this test is related to the change of dimension of the composite samples upon water uptake at certain period of time. According to the data obtained, the effect of thickness swelling is less compared to the water uptake percentage. The control sample (UPR) showed no sign of thickness swelling although it uptakes 0.988% water after 30 days. Highest percentage of swelling (1.673%) was observed for 5% WFM loaded composites where the thickness increased until the 7th day and remained constant up to 30th day. The 1% WFM loaded composite showed thickness swelling only for the first day and the rest was constant. The 2%, 3% and 4% WFM loaded composites showed initial thickness swelling up to 4th, 7th and 4th day respectively and rest days were at constant values.Figure 10 Thickness swelling of control (UPR) and WFM loaded composites.
Figure 10
The swelling effect was not that much significant for the WFM loaded composites. One possible reason might be due to the softening of composite under the influence of water that reduces the rigidity. In addition to that, the WFM fibers were shredded to a certain size rather than a continuous thread. With such case, the absorbed water by protruding WFM fibers at the boundary wall were unable to move inside the material significantly and thus low thickness swelling resulted. Since no WFM fibers are present in the control sample, thickness swelling wasn't observed as a repercussion. The 5% WFM loaded composite has the highest amount of WFM fiber and these fibers are close enough to make an inter-connection; this aided in greater water uptake as well as swelling of the composite material.
4 Conclusion
Since its emergence, the cataclysmic corona virus (covid-19) has affected our daily life as well as the environment with wastes that are most likely to persist for at least couple hundred years. The waste generation associated with PPE, especially the waste face mask is already a perturbation for the concerned society as well as for the researchers trying to find an immediate solution. Here in this research, waste face masks have been utilized by making composite material using unsaturated polyester resin. The key findings of this research include.i. Inclusion of WFM into UPR matrix has evidently increased the mechanical properties of the composites which indicate its potential applicability as a reinforcing material.
ii. The 2% WFM loaded composite exhibits highest tensile and flexural properties compared to control, 1%, 3%, 4% and 5% WFM loaded composites. This depicts the amount of WFM that is in better interaction with the UPR matrix which ultimately shows better physical properties.
iii. The functional group analysis confirmed the fibers of the collected WFM to be of polypropylene. The existing functional groups of the UPR matrix have also been confirmed. The inclusion of WFM within the UPR matrix was also confirmed by the FTIR analysis.
iv. SEM analysis confirmed strong adherence of the WFM fibers with the UPR matrix as well as very few fiber pull outs. Three different sizes of WFM fibers were confirmed within the fractured surface of the composites.
v. Water uptake and thickness swelling tests revealed the water absorption capacity and consequential dimensional change within the composite materials. The 5% WFM loaded composite showed highest thickness swelling and water uptake due to highest amount of shredded WFM in close proximity and making an inter-connection for water to enter.
According to the findings of this research, it can easily be concluded that, waste face mask can be utilized in making composite material with better mechanical properties and hence, have potential applicability in making household articles, containers, roofing, door panels, automotive structure, insulation boards, door/window frames etc. [30, 75]. The idea of utilization of a waste material that is associated with the covid-19 outbreak will allow researchers to find solutions of this waste conundrum. There is scope for further investigation with different types of face masks as well as other PPEs which will also be crucial in covid-19 breed waste utilization.
Declarations
Author contribution statement
Mashrafi Bin Mobarak: Performed the experiments; Analyzed and interpreted the data; Wrote the paper.
Md. Sahadat Hossain: Conceived and designed the experiments; Analyzed and interpreted the data.
Fariha Chowdhury: Performed the experiments; Analyzed and interpreted the data.
Samina Ahmed: Conceived and designed the experiments; Contributed reagents, materials, analysis tools or data.
Funding statement
This work was supported by IGCRT, BCSIR (R&D approval ref. 39.02.0000.011.14.134.2021/900; Date: 30/12/2021).
Data availability statement
Data included in article/supplementary material/referenced in article.
Declaration of interests statement
The authors declare no competing interests.
Additional information
No additional information is available for this paper.
==== Refs
References
1 Shereen M.A. Khan S. Kazmi A. Bashir N. Siddique R. COVID-19 infection: origin, transmission, and characteristics of human coronaviruses J. Adv. Res. 24 2020 91 32257431
2 WHO Coronavirus (COVID-19) Dashboard, (n.d.). https://covid19.who.int (accessed April 2, 2022).
3 WHO Director-General’s opening remarks at the media briefing on Covid-19-11 March 2020, (n.d.). https://www.who.int/director-general/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19–11-march-2020 (accessed April 2, 2022).
4 The WHO Just declared coronavirus COVID-19 a pandemic, time. (n.d.). https://time.com/5791661/who-coronavirus-pandemic-declaration/(accessed April 2, 2022).
5 Coronavirus: The world in lockdown in maps and charts, BBC News https://www.bbc.com/news/world-52103747 2020 (accessed April 2, 2022)
6 Wilder-Smith A. Freedman D.O. Isolation, quarantine, social distancing and community containment: pivotal role for old-style public health measures in the novel coronavirus (2019-nCoV) outbreak. J. Trav. Med. 27 2 2020 taaa020
7 Chintalapudi N. Battineni G. Amenta F. COVID-19 virus outbreak forecasting of registered and recovered cases after sixty day lockdown in Italy: a data driven model approach, J. Microbiol. Immunol. Infect. 53 2020 396 403 32305271
8 Lin Y.-H. Liu C.-H. Chiu Y.-C. Google searches for the keywords of “wash hands” predict the speed of national spread of COVID-19 outbreak among 21 countries Brain Behav. Immun. 87 2020 30 32 32283286
9 Advice for the public on Covid-19 – World Health Organization, (n.d.). https://www.who.int/emergencies/diseases/novel-coronavirus-2019/advice-for-public (accessed April 2, 2022).
10 De-la-Torre G.E. Aragaw T.A. What we need to know about PPE associated with the COVID-19 pandemic in the marine environment Mar. Pollut. Bull. 163 2021 111879 33385799
11 Rubio-Romero J.C. del Carmen Pardo-Ferreira M. Torrecilla-García J.A. Calero-Castro S. Disposable masks: disinfection and sterilization for reuse, and non-certified manufacturing, in the face of shortages during the COVID-19 pandemic Saf. Sci. 129 2020 104830 32406406
12 Leung C.C. Lam T.H. Cheng K.K. Mass masking in the COVID-19 epidemic: people need guidance Lancet 395 2020 945
13 Wu H. Huang J. Zhang C.J. He Z. Ming W.-K. Facemask shortage and the novel coronavirus disease (COVID-19) outbreak: reflections on public health measures EClinicalMedicine 21 2020 100329 32292898
14 Shortage of personal protective equipment endangering health workers worldwide, (n.d.). https://www.who.int/news/item/03-03-2020-shortage-of-personal-protective-equipment-endangering-health-workers-worldwide (accessed April 2, 2022).
15 Prata J.C. Silva A.L. Walker T.R. Duarte A.C. Rocha-Santos T. COVID-19 pandemic repercussions on the use and management of plastics Environ. Sci. Technol. 54 2020 7760 7765 32531154
16 Akhbarizadeh R. Dobaradaran S. Nabipour I. Tangestani M. Abedi D. Javanfekr F. Jeddi F. Zendehboodi A. Abandoned Covid-19 personal protective equipment along the Bushehr shores, the Persian Gulf: an emerging source of secondary microplastics in coastlines Mar. Pollut. Bull. 168 2021 112386 33901902
17 Fadare O.O. Okoffo E.D. Covid-19 face masks: a potential source of microplastic fibers in the environment, Sci. Total Environ. 737 2020 140279 32563114
18 Aragaw T.A. Surgical face masks as a potential source for microplastic pollution in the COVID-19 scenario Mar. Pollut. Bull. 159 2020 111517 32763564
19 Chowdhury H. Chowdhury T. Sait S.M. Estimating marine plastic pollution from COVID-19 face masks in coastal regions Mar. Pollut. Bull. 168 2021 112419 33930644
20 Abbasi S.A. Khalil A.B. Arslan M. Extensive use of face masks during COVID-19 pandemic:(micro-) plastic pollution and potential health concerns in the Arabian Peninsula Saudi J. Biol. Sci. 27 2020 3181 3186 33052188
21 Haddad M.B. De-la-Torre G.E. Abelouah M.R. Hajji S. Alla A.A. Personal protective equipment (PPE) pollution associated with the COVID-19 pandemic along the coastline of Agadir, Morocco, Sci Total Environ. 798 2021 149282
22 Rakib M.R.J. De-la-Torre G.E. Pizarro-Ortega C.I. Dioses-Salinas D.C. Al-Nahian S. Personal protective equipment (PPE) pollution driven by the COVID-19 pandemic in Cox’s Bazar, the longest natural beach in the world Mar. Pollut. Bull. 169 2021 112497 34022562
23 Silva A.L.P. Prata J.C. Mouneyrac C. Barcelò D. Duarte A.C. Rocha-Santos T. Risks of Covid-19 face masks to wildlife: present and future research needs Sci. Total Environ. 2021 148505 34465061
24 Asim N. Badiei M. Sopian K. Review of the valorization options for the proper disposal of face masks during the COVID-19 pandemic Environ. Technol. Innovat. 23 2021 101797
25 Aragaw T.A. Mekonnen B.A. Current plastics pollution threats due to COVID-19 and its possible mitigation techniques: a waste-to-energy conversion via Pyrolysis Environ. Syst. Res. 10 2021 1 11
26 Khalid U. Optimization of COVID-19 face mask waste fibers and silica fume as a balanced mechanical ameliorator of fat clay using response surface methodology Environ. Sci. Pollut. Res. 2021 1 16
27 Karbhari V.M. Fabrication, quality and service-life issues for composites in civil engineering, in: Durab Compos. Civil Struct. Appl. 2007 13 30
28 Malkapuram R. Kumar V. Negi Y.S. Recent development in natural fiber reinforced polypropylene composites J. Reinforc. Plast. Compos. 28 2009 1169 1189
29 Ku H. Wang H. Pattarachaiyakoop N. Trada M. A review on the tensile properties of natural fiber reinforced polymer composites, Compos. Part B Eng. 42 2011 856 873
30 Rajak D.K. Pagar D.D. Menezes P.L. Linul E. Fiber-reinforced polymer composites: manufacturing, properties, and applications Polymers 11 2019 1667 31614875
31 Rahman R. Putra S.Z.F.S. Tensile properties of natural and synthetic fiber-reinforced polymer composites Mech. Phys. Test. Biocompos. Fibre-Reinf. Compos. Hybrid Compos. 2019 81 102
32 Gull N. Khan S.M. Munawar M.A. Shafiq M. Anjum F. Butt M.T.Z. Jamil T. Synthesis and characterization of zinc oxide (ZnO) filled glass fiber reinforced polyester composites Mater. Des. 67 2015 313 317
33 Ahmad J. Zhou Z. Mechanical properties of natural as well as synthetic fiber reinforced concrete: a review Constr. Build. Mater. 333 2022 127353
34 Cheung H. Ho M. Lau K. Cardona F. Hui D. Natural fibre-reinforced composites for bioengineering and environmental engineering applications Compos. B Eng. 40 2009 655 663
35 Banea M.D. da Silva L.F. Adhesively bonded joints in composite materials: an overview Proc. Inst. Mech. Eng. Part J. Mater. Des. Appl. 223 2009 1 18
36 Cavalcanti D.K.K. Banea M.D. Neto J.S.S. Lima R.A.A. Da Silva L.F.M. Carbas R.J.C. Mechanical characterization of intralaminar natural fibre-reinforced hybrid composites Compos. B Eng. 175 2019 107149
37 Ali M.F. Hossain M.S. Moin T.S. Ahmed S. Chowdhury A.S. Utilization of waste chicken feather for the preparation of eco-friendly and sustainable composite Clean. Eng. Technol. 4 2021 100190
38 Hossain M.T. Hossain M.S. Uddin M.B. Khan R.A. Chowdhury A.S. Preparation and characterization of sodium silicate–treated jute-cotton blended polymer–reinforced UPR-based composite: effect of γ-radiation Adv. Compos. Hybrid Mater. 4 2021 257 264
39 Sahadat Hossain M. Chowdhury A.S. Khan R.A. Effect of disaccharide, gamma radiation and temperature on the physico-mechanical properties of jute fabrics reinforced unsaturated polyester resin-based composite Radiat. Eff. Defect Solid 172 2017 517 530
40 Hossain M. Mobarak M.B. Rony F.K. Sultana S. Mahmud M. Ahmed S. Fabrication and characterization of banana fiber reinforced unsaturated polyester resin based composites Nano Hybrids Compos. 2020 Trans Tech Publ 84 92
41 Oleksy M. Galina H. Unsaturated polyester resin composites containing bentonites modified with silsesquioxanes Ind. Eng. Chem. Res. 52 2013 6713 6721
42 Zhang C. Huang J.Y. Liu S.M. Zhao J.Q. The synthesis and properties of a reactive flame-retardant unsaturated polyester resin from a phosphorus-containing diacid Polym. Adv. Technol. 22 2011 1768 1777
43 Pan L.L. Li G.Y. Su Y.C. Lian J.S. Fire retardant mechanism analysis between ammonium polyphosphate and triphenyl phosphate in unsaturated polyester resin Polym. Degrad. Stabil. 97 2012 1801 1806
44 Zhang M. Ye W. Liao Z. Preparation, characterization and properties of flame retardant unsaturated polyester resin based on r-PET J. Polym. Environ. 2021 1 11
45 Shi X. Zhang H. Bu X. Zhang G. Zhang H. Kang H. Performance evaluation of BDM/unsaturated polyester resin-modified asphalt mixture for application in bridge deck pavement Road Mater. Pavement Des. 2020 1 17
46 Atta A.M. Nassar I.F. Bedawy H.M. Unsaturated polyester resins based on rosin maleic anhydride adduct as corrosion protections of steel React. Funct. Polym. 67 2007 617 626
47 Niu M. Wu Z. Lin X. Liu Z. Xie Y. Bhuiyan I.U. Wang X. Manufacturing and properties of ultra-low density fiberboards with an unsaturated polyester resin by a dry process, Eur. J. Wood Wood Prod. 76 2018 853 859
48 Mehta L.B. Wadgaonkar K.K. Jagtap R.N. Synthesis and characterization of high bio-based content unsaturated polyester resin for wood coating from itaconic acid: effect of various reactive diluents as an alternative to styrene J. Dispersion Sci. Technol. 40 2019 756 765
49 Wu Z. Aladejana J.T. Liu S. Gong X. Wang X.A. Xie Y. Unsaturated polyester resin as a nonformaldehyde adhesive used in bamboo particle boards ACS omega 7 4 2022 3483 3490 35128257
50 Zhang N. Hou X. Cui X. Chai L. Li H. Zhang H. Wang Y. Deng T. Amphiphilic catalyst for decomposition of unsaturated polyester resins to valuable chemicals with 100% atom utilization efficiency J. Clean. Prod. 296 2021 126492
51 Rajak D.K. Wagh P.H. Linul E. A review on synthetic fibers for polymer matrix composites: performance, failure modes and applications, Materials 15 2022 4790 35888257
52 Devaraju S. Alagar M. Unsaturated polyester—macrocomposites Unsaturated Polyest 2019 Elsevier Resins 43 66
53 Methyl Ethyl Ketone Peroxide (MEKP, MED, and peroxide) Gooch J.W. Encycl. Dict. Polym. 2007 Springer New York, NY 614 615
54 Chin A.W. Chu J.T. Perera M.R. Hui K.P. Yen H.-L. Chan M.C. Peiris M. Poon L.L. Stability of SARS-CoV-2 in different environmental conditions Lancet Microbe 1 2020 e10 32835322
55 Khare J.M. Dahiya S. Gangil B. Ranakoti L. Sharma S. Huzaifah M.R.M. Ilyas R.A. Dwivedi S.P. Chattopadhyaya S. Kilinc H.C. Comparative analysis of erosive wear behaviour of epoxy, polyester and vinyl esters based thermosetting polymer composites for human prosthetic applications using taguchi design Polymers 13 2021 3607 34685366
56 Standard Test Method for Tensile Properties of Plastics, (n.d.). https://www.astm.org/d0638-01.html (accessed February 14, 2022).
57 Standard Test Methods for Flexural Properties of Unreinforced and Reinforced Plastics and Electrical Insulating Materials, (n.d.). https://www.astm.org/d0790-10.html (accessed February 14, 2022).
58 Standard Test Method for Water Absorption of Plastics, (n.d.). https://www.astm.org/d0570-98r18.html (accessed March 28, 2022).
59 Mobarak M.B. Hossain M.S. Mahmud M. Ahmed S. Redispersible polymer powder modified cementitious tile adhesive as an alternative to ordinary cement-sand grout Heliyon 7 2021 e08411 34841113
60 Hamdan M.H.M. Siregar J.P. Cionita T. Jaafar J. Efriyohadi A. Junid R. Kholil A. Water absorption behaviour on the mechanical properties of woven hybrid reinforced polyester composites Int. J. Adv. Manuf. Technol. 104 2019 1075 1086
61 Azammi A.N. Ilyas R.A. Sapuan S.M. Ibrahim R. Atikah M.S.N. Asrofi M. Atiqah A. Characterization studies of biopolymeric matrix and cellulose fibres based composites related to functionalized fibre-matrix interface Interfaces Part. Fibre Reinf. Compos. 2020 Elsevier 29 93
62 M. Biron, Material selection for thermoplastic parts, (n.d.).
63 Potluri P. Needham P. 6-Technical Textiles for protection Scott R.A. Text. Prot. 2005 Woodhead Publishing 151 175
64 Leonas K.K. Jones C.R. Hall D. The relationship of fabric properties and bacterial filtration efficiency for selected surgical face masks, JTATM 3 2003 1 8
65 Krylova V. Dukštienė N. Synthesis and characterization of Ag2S layers formed on polypropylene J. Chem. 2013 2013
66 Jung M.R. Horgen F.D. Orski S.V. Rodriguez V. Beers K.L. Balazs G.H. Jones T.T. Work T.M. Brignac K.C. Royer S.-J. Validation of ATR FT-IR to identify polymers of plastic marine debris, including those ingested by marine organisms Mar. Pollut. Bull. 127 2018 704 716 29475714
67 Stuart B.H. Infrared Spectroscopy: Fundamentals and Applications 2004 John Wiley & Sons
68 Sciarratta V. Vohrer U. Hegemann D. Müller M. Oehr C. Plasma functionalization of polypropylene with acrylic acid Surf. Coat. Technol. 174 2003 805 810
69 Tammer, M. G. Sokrates: Infrared and Raman characteristic group frequencies: tables and charts. Colloid Polym. Sci. 2004, 283, 235.
70 Morent R. De Geyter N. Leys C. Gengembre L. Payen E. Comparison between XPS-and FTIR-analysis of plasma-treated polypropylene film surfaces Surf. Interface Anal. Int. J. Devoted Dev. Appl. Tech. Anal. Surf. Interfaces Thin Films 40 2008 597 600
71 Reddy K.O. Shukla M. Maheswari C.U. Rajulu A.V. Evaluation of mechanical behavior of chemically modified Borassus fruit short fiber/unsaturated polyester composites J. Compos. Mater. 46 2012 2987 2998
72 Koto N. Soegijono B. Effect of rice husk ash filler of resistance against of high-speed projectile impact on polyester-fiberglass double panel composites J. Phys. Conf. Ser. 2019 IOP Publishing 12058
73 Pavia, D. L., Lampman, G. M., Kriz, G. S. & Vyvyan, J. A. Introduction to Spectroscopy Ch. 2 (Brooks Cole, 2008).
74 Deng H. Reynolds C.T. Cabrera N.O. Barkoula N.-M. Alcock B. Peijs T. The water absorption behaviour of all-polypropylene composites and its effect on mechanical properties Compos. Part B Eng. 41 2010 268 275
75 Samanta A.K. Fibre reinforced composites: multiplicity of application Latest Trends Text Fash. 1 2018 1 3
| 0 | PMC9737522 | NO-CC CODE | 2022-12-16 23:20:00 | no | Heliyon. 2022 Dec 10; 8(12):e12197 | utf-8 | Heliyon | 2,022 | 10.1016/j.heliyon.2022.e12197 | oa_other |
==== Front
Eur J Cancer
Eur J Cancer
European Journal of Cancer
0959-8049
1879-0852
Published by Elsevier Ltd.
S0959-8049(22)01780-4
10.1016/j.ejca.2022.11.030
Original Research
AGIHO guideline on evidence-based management of COVID-19 in cancer patients: 2022 update on vaccination, pharmacological prophylaxis and therapy in light of the omicron variants
Giesen Nicola a∗1
Busch Elena b1
Schalk Enrico c
Beutel Gernot de
Rüthrich Maria M. f
Hentrich Marcus g
Hertenstein Bernd h
Hirsch Hans H. ijk
Karthaus Meinolf l
Khodamoradi Yascha m
Koehler Philipp no
Krüger William p
Koldehoff Michael qr
Krause Robert s
Mellinghoff Sibylle C. not
Penack Olaf u
Sandherr Michael v
Seggewiss-Bernhardt Ruth wx
Spiekermann Karsten y
Sprute Rosanne not
Stemler Jannik not
Weissinger Florian z
Wörmann Bernhard aa
Wolf Hans-Heinrich ab
Cornely Oliver A. notac
Rieger Christina T. ad
von Lilienfeld-Toal Marie aeaf
a Department of Hematology, Oncology and Palliative Care, Robert Bosch Hospital, Stuttgart, Germany
b Department of Medical Oncology, National Center for Tumor Diseases, University Hospital Heidelberg, Heidelberg, Germany
c Department of Hematology and Oncology, Medical Centre, Otto-von-Guericke University Magdeburg, Magdeburg, Germany
d Department for Haematology, Haemostasis, Oncology and Stem Cell Transplantation, Hannover Medical School, Hannover, Germany
e Working Party Intensive Care in Haematologic and Oncologic Patients (iCHOP) of the German Society of Haematology and Medical Oncology (DGHO)
f Department of Interdisciplinary Intensive Care Medicine, Vivantes Humboldt-Klinikum, Berlin, Germany
g Department of Hematology and Oncology, Red Cross Hospital Munich, Munich, Germany
h Klinikum Bremen-Mitte, Bremen, Germany
i Transplantation & Clinical Virology, Department Biomedicine (Haus Petersplatz), University of Basel, Basel, Switzerland
j Clinical Virology, Laboratory Medicine, University Hospital Basel, Basel, Switzerland
k Infectious Diseases & Hospital Epidemiology, University Hospital Basel, Basel, Switzerland
l Department of Hematology, Oncology and Palliative Care, Klinikum Neuperlach/Klinikum Harlaching, Munich, Germany
m Department of Internal Medicine, Infectious Diseases, Goethe University Frankfurt, Frankfurt Am Main, Germany
n University of Cologne, Faculty of Medicine and University Hospital Cologne, Translational Research, Cologne Excellence Cluster on Cellular Stress Responses in Aging-Associated Diseases (CECAD), Cologne, Germany
o University of Cologne, Faculty of Medicine and University Hospital Cologne, Department I of Internal Medicine, Center for Integrated Oncology Aachen Bonn Cologne Duesseldorf (CIO ABCD) and Excellence Center for Medical Mycology (ECMM), Cologne, Germany
p Department of Hematology and Oncology, Stem Cell Transplantation, Palliative Care, University Hospital Greifswald, Greifswald, Germany
q Department of Bone Marrow Transplantation, West German Cancer Center, University Hospital Essen, University Duisburg-Essen, Essen, Germany
r Department of Hygiene and Environmental Medicine, University Hospital Essen, University Duisburg-Essen, Essen, Germany
s Division of Infectious Diseases, Department of Internal Medicine, Medical University of Graz, Graz, Austria
t German Center for Infection Research (DZIF), Partner Site Bonn-Cologne, Cologne, Germany
u Charité – Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Department of Hematology, Oncology and Tumorimmunology, Berlin, Germany
v MVZ Penzberg, Department of Hematology and Oncology, Weilheim, Germany
w Medizinische Klinik V, Sozialstiftung Bamberg, Bamberg, Germany
x Comprehensive Cancer Center Erlangen-EMN, Erlangen, Germany
y Department of Medicine III, University Hospital, LMU Munich, Munich, Germany
z Department of Internal Medicine, Hematology, Oncology, Stem Cell Transplantation and Palliative Care, Evangelisches Klinikum Bethel, Bielefeld, Germany
aa Division of Haematology, Oncology and Tumor Immunology, Department of Medicine, Charité-Universitätsmedizin Berlin, Berlin, Germany
ab Department of Hematology, Oncology and Hemostaseology, Südharzklinikum Nordhausen, Nordhausen, Germany
ac University of Cologne, Faculty of Medicine and University Hospital Cologne, Clinical Trials Centre Cologne (ZKS Köln), Cologne, Germany
ad Hemato-Oncology Germering & Interdisciplinary Tumorcenter, Ludwig-Maximilians-University Munich, Munich, Germany
ae Department of Haematology and Medical Oncology, Clinic for Internal Medicine II, University Hospital Jena, Jena, Germany
af Leibniz Institute for Natural Product Research and Infection Biology, Hans Knöll Institute, Jena, Germany
∗ Corresponding author. Department of Hematology, Oncology and Palliative Care, Robert Bosch Hospital, Auerbachstr. 110, 70376 Stuttgart, Germany. Tel.: +49 711-8101-5506.
1 equal contribution.
10 12 2022
10 12 2022
16 11 2022
21 11 2022
22 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.
The novel coronavirus SARS-CoV-2 and the associated infectious disease COVID-19 pose a significant challenge to health care systems worldwide. Cancer patients have been identified as a high-risk population for severe infections, rendering prophylaxis and treatment strategies for these patients particularly important. Rapidly evolving clinical research resulting in the recent advent of various vaccines and therapeutic agents against COVID-19 offers new options to improve care and protection of cancer patients. However, ongoing epidemiological changes and rise of new virus variants require repeated revisions and adaptations of prophylaxis and treatment strategies in order to meet these new challenges. Therefore, this guideline provides an update on evidence-based recommendations with regard to vaccination, pharmacological prophylaxis and treatment of COVID-19 in cancer patients in light of the currently dominant omicron variants. It is developed by an expert panel of the Infectious Diseases Working Party (AGIHO) of the German Society for Hematology and Medical Oncology (DGHO) based on a critical review of the most recent available data.
Keywords
COVID-19
SARS-CoV-2
cancer
solid tumor
hematological malignancy
guideline
vaccination
prophylaxis
treatment
==== Body
pmcIntroduction
Since their first description in late 2019 the novel severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) and the associated infectious coronavirus disease COVID-19 have swept throughout the world posing a serious strain on health care systems but also triggering an unprecedented scientific outreach to meet the challenge.1 , 2 Patients with cancer were early on identified as a particular at-risk population and mortality rates of cancer patients with COVID-19 of around 30% have been reported.3, 4, 5 With a better understanding of the disease and the advent of vaccines and novel therapeutic options, outcome of COVID-19 has also improved in cancer patients.6 , 7
Since the beginning of 2022 the SARS-CoV-2 variant of concern (VOC) omicron is dominant in many regions including Europe and Northern America.8 While transmissibility seems to be increased, it has been associated with a lower hospitalization and mortality rate, in particular compared to the preceding delta variant.9, 10, 11 COVID-19, however, still remains a potentially fatal disease, in particular in immunocompromised cancer patients.12 Furthermore, protracted courses of COVID-19 are often observed in these patients leading to treatment interruptions and thus potentially endangering the success of oncologic therapies.13 Adequate prophylaxis and treatment strategies of COVID-19 are therefore of the utmost importance in this patient population.
Major advances in the development of COVID-19 vaccines,14, 15, 16, 17, 18 anti-SARS-CoV-2 monoclonal antibodies19, 20, 21, 22, 23 and specific antiviral agents24, 25, 26 have shifted the focus towards prevention of severe COVID-19 both by active or passive immunization and by early treatment of non-hospitalized patients. As immunocompromised cancer patients are at particular risk for severe COVID-19 and often do not mount an adequate immune response to SARS-CoV-2 infection or COVID-19 vaccine,27, 28, 29, 30, 31 specific strategies with regard to vaccination and treatment have to be applied to protect this high-risk patient population.
Given the rapid advancements of new prophylactic and therapeutic options in the field of SARS-CoV-2/COVID-19 and the continuous change in dominant variants of concern we feel that a new update of our previously published guidelines32 , 33 is warranted. This guideline update focusses in particular on vaccination and treatment strategies of COVID-19 in adult patients with solid tumors or hematologic malignancies in light of the new omicron variant and its dominant sublineages. It offers evidence-based recommendations to help treating physicians make informed decisions on their patients’ care.
Methods
This guideline was developed in a formalized process by an expert panel from the Infectious Diseases Working Party (AGIHO) of the German Society for Hematology and Medical Oncology (DGHO). This panel consisted of 28 specialists certified in medical oncology, hematology, infectious diseases, critical care, emergency medicine, and virology.
First, a systematic search of relevant literature on pre-defined topics was performed on MEDLINE for publications using one of the following search terms: “SARS-CoV-2”, “COVID-19”, “vaccine/vaccination”, “prophylaxis/prevention”, or “therapy/treatment”. Due to the fast developments in COVID-19 research, publications available only on the pre-print server www.medRxiv.org were also included, however, the lack of formal peer-review was considered with regard to the grading of the quality of evidence. Publications were evaluated that appeared online until October 15th, 2022.
After collection of relevant literature, a thorough review was performed, and the data was extracted and rated. Based on the results of the data analysis, preliminary recommendations were discussed and revised in a formalized step-by-step process by the expert panel. Strength of recommendation and quality of evidence were graded according to the scale proposed by the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) (Table 1 ).34 The final recommendations as presented in this guideline were discussed and agreed upon by the AGIHO general assembly.Table 1 Grading system for strength of recommendation (SoR) and quality of evidence (QoE) as proposed by ESCMID.34
Table 1Strength of recommendation
A AGIHO strongly supports a recommendation for use
B AGIHO moderately supports a recommendation for use
C AGIHO marginally support a recommendation for use
D AGIHO supports a recommendation against use
Quality of evidence
I Evidence from at least one properly designed randomized, controlled trial
II* Evidence from at least one well-designed clinical trial, without randomization; from cohort- or case-control analytic studies (preferably from more than one centre); from multiple time series; or from dramatic results from uncontrolled experiments
III Evidence from opinion of respected authorities, based on clinical experience, descriptive case studies, or report of expert committees
*added index for level II
r Meta-analysis or systematic review of randomized controlled trials
t Transferred evidence, i.e. results from different patients’ cohorts or similar immune status situation
h Comparator group is a historical control
u Uncontrolled trial
a Abstract published at an international meeting or manuscript available on preprint server only
Vaccination
Although patients with cancer have not been part of initial clinical trials, the clinical efficacy of vaccination against COVID-19 has now been proven for this population as well.35 While the rate of seroconversion as well as absolute antibody levels seem to be lower in cancer patients compared to the healthy individuals,36 , 37 overall, a clinical efficacy with an 80-90% prevention rate of symptomatic COVID-19 can be expected in patients with cancer.38 However, cancer patients remain at an increased risk of severe disease despite vaccination compared to the general population39 , 40 and thus several aspects need to be taken into account:1) Patients with cancer are often of advanced age and harbor co-morbidities, that may lead to reduced vaccine efficacy38
2) Patients with cancer in general seem to exhibit a more rapidly waning immunity than the general population which requires more frequent and earlier booster doses than the general population35 , 41 , 42
3) Patients with cancer are often undergoing therapy with immunosuppressive drugs which reduce the immune response to vaccines.43 Amongst these, therapy with B cell depleting agents such as CD20-antibodies, BTK inhibitors or BCMA-directed therapy is associated with the lowest rate of seroconversion: 0% in some studies, around 30% in most studies.38 , 43
We therefore recommend vaccines effective against COVID-19 unrestrictedly for patients with cancer as per recommendation by the respective local authorities and subject to availability (AIIt, see Table 2 ). This also applies to bivalent vaccines if available, as they may provide increased protection against current VOC.44 Of note, in clinical studies, the mRNA vaccines and particularly the mRNA-1273 vaccine appears to result in the best clinical protection. As a guidance we recommend that if given the choice, mRNA vaccines with the highest possible dose should be chosen (BII). We recommend prioritization of patients with cancer for booster vaccination (AIIt). This is especially important given the rapid decline of humoral vaccine response observed in cancer patients compared to healthy individuals.42 Booster vaccination has been shown to be effective in increasing antibody levels and lower the risk of severe disease in cancer patients, and may even lead to seroconversion in initial non-responders.39 , 45 A significant increase in antibody levels after booster vaccination was observed in cancer patients undergoing a variety of antineoplastic treatments including chemotherapy, targeted therapy, immunotherapy or prior stem cell transplantation (SCT).41 With regard to the management of cancer therapy during vaccination we do not recommend pausing ongoing cancer therapy (DIII), but if it is possible to apply the vaccine prior to the beginning of the cancer therapy, this should be attempted (AIIt). Also, we recommend simultaneous vaccination against influenza if the vaccination against COVID-19 coincides with the scheduled seasonal influenza vaccination in autumn (AIIt).Table 2 Recommendations on vaccination against COVID-19 in patients with cancer.
Table 2Population Intention Intervention SoR QoE Reference
Cancer patients Reduce rate of symptomatic COVID Full vaccination against COVID A IIt 141,142
Cancer patients Reduce rate of hospitalisation Full vaccination against COVID A IIt 141
Cancer patients Reduce mortality Full vaccination against COVID A IIt 141,142
Cancer patients Increase immunogenicity Booster as recommended by national authorities A IIt 39,41,45,143, 144, 145
Cancer patients before start of therapy Increase immunogenicity Vaccinate before start of chemotherapy* A IIt
Cancer patients due to receive other vaccination Co-administration of needed vaccinations such as influenza A III
Cancer patients Maintain immunogenicity Pause cancer therapy D IIu 146,147
Cancer patients Induce best possible vaccination response Choose mRNA-vaccines over all other vaccines B IIu 147148
Cancer patients Induce best possible vaccination response Choose highest approved vaccine dose (no increased toxicity in pts. with cancer) B IIu 141147149
Regarding toxicity, there is no evidence, that patients with cancer have a higher rate of adverse reactions than the general population.37 , 38 This is also true for patients undergoing immune-checkpoint inhibition.46 Also, there is no evidence whatsoever that vaccination against COVID-19 increases the risk of de novo or relapsed cancer. However, there are specific clinical situations which may be associated with symptoms, that may appear worrisome. One situation may be diagnostic confusion in the event of enlarged reactive lymph nodes after vaccination, which may be confused with progressive malignant disease.47 Another situation that should be kept in mind concern patients after radiation, since a radiation recall phenomenon has been described after COVID-19 vaccination.48, 49, 50 The third situation has been observed in patients after allogeneic SCT, which have been described to exhibit a 10% rate of flare of graft-versus-host disease as well as temporary cytopenias.51, 52, 53 However, none of the above mentioned situations justify refraining from vaccination.
The level of protection in this extremely vulnerable patient group is of high interest for treating physicians as for the patients.54 It is tempting to measure a serological response with the expectation of a clinically meaningful result. In the general population, there is good evidence that the level of (neutralizing) antibody response correlates with clinical efficacy55 although a clear cut-off value for protection has not been set. However, in patients with cancer, there are a number of pitfalls with routine testing of the serological response: first, in patients with cancer the humoral and cellular response may be discordant – most prominently with a higher rate of cellular response despite lack of antibody response in patients undergoing B cell depletion and a lack of cellular response despite the presence of antibodies after allogeneic SCT.43 , 56 Measuring antibody levels therefore does not capture the full immune response and in those without antibodies there may still be relevant protection by specific T cells. This is still the case after booster vaccinations.56 Second, even if neutralizing antibodies are found, a breakthrough infection may yet occur with a novel VOC which is not recognized by the antibodies present.57 Third, the significance of B cell and T cell response with regard to protection from infection and from severe course of disease is yet to be determined58 and last but not least, the presence of monoclonal antibodies after passive immunization may be confused with a vaccine response. Therefore, the result of routine testing of antibodies may not be helpful in the clinical context as it is difficult to draw conclusions from it.59
Pharmacological Prophylaxis
While vaccination is most effective in the prevention of severe COVID-19 in the general population, in some cancer patients with severe immunosuppression vaccination is either not expected to result in adequate seroprotection, e.g. in those undergoing hematopoietic stem cell transplantation, or they fail to mount an adequate immune response despite repeated boostering.43 For these patient populations, passive immunization strategies via pre- or post-exposure prophylaxis with monoclonal antibodies are valuable options (see Table 3 ).Table 3 Recommendations on prophylaxis against COVID-19 in cancer patients.
Table 3Population Intention Intervention SoR QoE Reference
Cancer patients without adequate vaccine protectiona, pre-exposure prophylaxis to prevent infection Anti-S monoclonal antibodiesb B IIt 19
Cancer patients without adequate vaccine protectiona, post-exposure prophylaxis to prevent infection Anti-S monoclonal antibodiesb A IIt 20
Cancer patients, pre- or post-exposure prophylaxis to prevent infection high-titer convalescent plasma D IIt 63
Cancer patients, pre- or post-exposure prophylaxis to prevent infection any antiviral (nirmatrelvir/ritonavir, remdesivir, molnupiravir) D III
a i.e. cancer patients in whom vaccination is not feasible, at high risk of poor vaccine response or with proven inadequate vaccine response despite full vaccination.
b if available against the locally predominant SARS-CoV-2 variant.
The long-acting anti-S monoclonal antibody combination tixagevimab/cilgavimab has been evaluated as pre-exposure prophylaxis in patients with high risk of poor vaccine response or high risk of exposure in a large placebo-controlled RCT.19 While the overall number of symptomatic COVID-19 cases was low during the observation period, the primary efficacy endpoint, a significant reduction in symptomatic COVID-19, was met (0.2% vs 1.0%).19 Given the trial design, it is important to note that these results are mainly transferable to cancer patients in whom active vaccination is unlikely to result in sufficient vaccine response or those with proven inadequate vaccine response despite full vaccination. This also includes patients undergoing SCT who are expected to loose any prior vaccine-induced seroprotection. For these patient populations, we moderately recommend monoclonal anti-S antibodies with efficacy against the locally predominant SARS-CoV-2 variants, in particular long-acting antibodies, as pre-exposure prophylaxis (BIIt). The complexity of assessing vaccine response, given the diverse impacts of anti-cancer treatments on B-cell and T-cell responses as delineated above, has to be kept in mind. Furthermore, it has to be stated clearly, that pre-exposure prophylaxis should not be used as an alternative strategy to vaccination in patients in whom successful vaccination is possible (AIII).
Monoclonal antibodies as post-exposure prophylaxis have been successfully evaluated in a large RCT in unvaccinated household contacts of a person with SARS-CoV-2 infection.20 Casirivimab/imdevimab administered within 96 hours after diagnosis were associated with a lower risk of symptomatic COVID-19 compared to placebo (1.5% vs 7.8%).20 In vitro studies suggest that neutralizing activity of casirivimab/imdevimab is severely reduced against the currently predominant omicron BA.4/5 variants compared to the ancestral strain.60 , 61 Therefore, this particular antibody combination might currently not be a suitable option. However, the strategy of applying monoclonal anti-S antibodies as post-exposure prophylaxis has been shown to be safe and efficacious and we therefore strongly recommend post-exposure prophylaxis in unvaccinated cancer patients or those with poor vaccine response with monoclonal anti-S antibodies as early as possible if antibody preparations active against the currently predominant SARS-CoV-2 variants are available (AIIt).
It is important to note that published data on RCTs evaluating monoclonal anti-S antibodies as prophylaxis or treatment strategies have all been conducted prior to the predominance of the omicron variant, in particular the currently predominant BA.4/5 subvariants. So far, data on efficacy of the various monoclonal antibodies against BA.4/5 is based almost exclusively on in vitro neutralization assays making predictions on clinical efficacy inherently fraught.60 , 61 Summing up the currently available in vitro evidence, it seems that neutralizing efficacy against omicron BA.4/5 of casirivimab/imdevimab, bamlanivimab/etesevimab, and sotrovimab is too strongly reduced to retain clinical efficacy, and of tixagevimab/cilgavimab is moderately reduced.60 , 61 Some preliminary data suggest doubling the dosage of tixagevimab/cilgavimab in the setting of less susceptible omicron subvariants may increase efficacy.62 Neutralizing activity of bebtelovimab seems to be retained against BA.4/5,60 although this monoclonal antibody is so far only approved by the FDA and not by EMA.
High-titer convalescent plasma (CP) as post-exposure prophylaxis was evaluated in a smaller RCT and failed to show prevention of SARS-CoV-2 infection.63 CP is therefore not recommended as a prophylaxis strategy (DIIt).
There is currently no data on the antiviral agents remdesivir, nirmatrelvir/ritonavir or molnupiravir as pre- or post-exposure prophylaxis in cancer patients which can therefore not be recommended (DIII).
Treatment
Treatment options in COVID-19 are a rapidly evolving field. In order to best devise treatment strategies in cancer patients with COVID-19 several important aspects have to be considered: the patient’s immune status, in particular with regard to prior COVID vaccinations and COVID vaccine response; the severity of COVID-19 and time lapse since onset of symptoms; the local epidemiology and presence of variants of concern; the national approval and local availability of anti-COVID-19 drugs. In the following, evidence-based recommendations on treatment of COVID-19 in cancer patients are given with differentiation of patients’ clinical status according to the WHO clinical progression scale, i.e. out-patient mild disease (score 1-3), hospitalized moderate disease (score 4-5), and hospitalized severe disease (score 6-9).64
Especially in the early months of the COVID-19 pandemic, a plethora of potential medications to treat COVID-19 were evaluated that had to be discarded later on. Based on current data the following drugs are not indicated for COVID-19 in cancer patients of any severity due to lack of proven efficacy: hydroxy-chloroquine +/- azithromycin (DIIt),65, 66, 67 lopinavir/ritonavir (DIIt),68 umifenovir (DIItr),69 favipiravir (DIItr),70 and ivermectin (DIItr).71
Out-patient mild disease (WHO score 1-3)
For immunocompromised cancer patients with COVID-19 and out-patient mild disease both monoclonal antibodies and antiviral agents offer effective options of early therapy. Evidence-based recommendations in this setting are summarized in Table 4 . Initiation of therapy should be initiated as early after symptom onset and/or positive test for SARS-CoV-2 as possible, ideally within 3-7 days (AIII). However, a later time point should not be considered a contraindication to therapy initiation, especially in high-risk patients with uncontrolled infection. Confirmation of a positive rapid antigen test via nucleic acid amplification technique (NAT) should be attempted. However, given the advantage of early treatment initiation, therapy should not be delayed awaiting NAT test results in such cases.Table 4 Recommendations on treatment of cancer out-patients with mild COVID-19 (WHO score 1-3).
Table 4Population Intention Intervention SoR QoE Reference
Cancer patients with Covid-19 – out-patient, mild [1-3] to prevent hospitalization and/or death anti-S monoclonal antibodiesa A IIt 21, 22, 23,73
Cancer patients with Covid-19 – out-patient, mild [1-3] to prevent hospitalization and/or death high-titer convalescent plasma C IIt 75, 76, 77
Cancer patients with Covid-19 – out-patient, mild [1-3] to prevent hospitalization and/or death nirmatrelvir/ritonavir A IIt 25
Cancer patients with Covid-19 – out-patient, mild [1-3] to prevent hospitalization and/or death remdesivir B IIt 26
Cancer patients with Covid-19 – out-patient, mild [1-3] to prevent hospitalization and/or death molnupiravir C IIt 24
a if available against the locally predominant SARS-CoV-2 variant, particularly indicated in unvaccinated patients or those at risk of poor vaccine response.
Various monoclonal antibodies against the SARS-CoV-2 spike have been evaluated in placebo-controlled RCTs in ambulatory patients with COVID-19 early after symptom onset, defined as within 3-7 days after symptom onset or diagnosis of COVID-19 across the respective trials.21, 22, 23 , 72 , 73 While these trials addressed particularly patients at high-risk for severe COVID-19 including cancer patients, cancer patients constituted only a small minority of trial participants. Furthermore, as vaccinated patients were excluded from trial participation, the evidence is best transferable to cancer patients without vaccination or with presumed or proven inadequate vaccine response. The primary endpoint of a significant reduction in the rate of hospitalization or death compared to placebo was met for the following monoclonal antibodies or antibody combinations: casirivimab/imdevimab,21 bamlanivimab/etesevimab,23 regdanvimab,72 sotrovimab22, tixagevimab/cilgavimab.73 A relative risk reduction of 50.5% to 85% was observed.21, 22, 23 , 72 , 73 It is important to acknowledge that these clinical trials were all conducted prior to the dominance of the omicron variants. With regard to efficacy of currently available monoclonal antibodies against the omicron variants, in particular the newer BA.4/BA.5 subvariants, mainly data from in vitro neutralization assays as well as from small observational studies are available. For a summary of these data please refer to the previous paragraph.
Taking into account the caveats mentioned above, monoclonal antibodies seem to be a drug class particularly suitable for immunocompromised cancer patients, as these patients are prone to mount only an inadequate vaccine response and often take multiple medications making potential drug-drug interactions (DDI), such as with some of the antiviral agents, a concern. We therefore strongly recommend the drug class of monoclonal anti-S antibodies for early treatment of out-patient cancer patients (AIIt), if preparations effective against the locally dominant SARS-CoV-2 variant are available. Given their mechanism of action and the available trial data, monoclonal anti-S antibodies are particularly recommended for either unvaccinated patients or patients with inadequate vaccine response (AIII). Laboratory assessment of vaccine response is not necessary prior to therapy initiation, rather known risk factors of inadequate vaccine response in cancer patients should be taken into consideration to guide treatment decisions.74
High-titer CP has been evaluated in several RCTs in high-risk out-patients with COVID-19 with overall mixed results.75, 76, 77 We marginally recommend CP early after symptom onset in cancer out-patients with COVID-19 (CIIt). Given the superior efficacy data of some monoclonal antibodies in this setting, CP should only be administered if monoclonal antibody preparations efficacious against the locally predominant variants or antivirals are not available.
As of today, three different antiviral agents have been approved by both FDA and EMA for early treatment of COVID-19 in high-risk patients: nirmatrelvir/ritonavir, remdesivir, and molnupiravir. While the major clinical trials on these agents have all been conducted prior to the rise of the omicron variant, recent in vitro as well as retrospective clinical studies suggest that activity of all three agents against the currently predominant BA.4/5 omicron subvariants is preserved.60 , 78 , 79. Nirmatrelvir is an oral inhibitor of the viral 3CL-protease and is used in fixed combination with ritonavir as an inhibitor of CYP3A4 in order to increase bioavailability. In the placebo-controlled EPIC-HR RCT nirmatrelvir/ritonavir met its primary endpoint of a significant reduction compared to placebo in COVID-19 related hospitalization or death in unvaccinated, high-risk patients treated no later than 5 days after onset of symptoms (0.8% vs 6.3%, relative risk reduction 88%).25 Again, despite the fact that active cancer was featured among the pre-specified risk factors, only a minority of trial participants were cancer patients. While data in the setting of dominance of the omicron VOC is still limited, a recently published large real-world data analysis reported a relative risk reduction with regard to hospitalization due to COVID-19 of 73% and with regard to death of 79% in older adults who received nirmatrelvir/ritonavir compared to those who did not.79 We strongly recommend nirmatrelvir/ritonavir as early therapy in out-patient cancer patients with COVID-19 (AIIt). As ritonavir is a potent CYP3A4 inhibitor, possible DDI have to be considered which can be a particular issue in cancer patients who often take multiple medications. A comprehensive and regularly updated list of potential DDI between nirmatrelvir/ritonavir and concomitant medications is readily accessible e.g. via the National Institutes of Health webpage and many others.80
Remdesivir applied intravenously for three days was assessed as early therapy against placebo in the similarly designed PINETREE RCT resulting in a 87% relative risk reduction of COVID-related hospitalization or death in unvaccinated high-risk outpatients with COVID-19 within seven days of symptom onset.26 While the intravenous mode of application has logistic challenges in the outpatient setting, the lesser potential of DDI compared to nirmatrelvir/ritonavir may make this an attractive treatment option for certain patients. We therefore moderately recommend remdesivir as early therapy in ambulatory cancer patients with COVID-19 (BIIt).
Molnupiravir, an oral prodrug of a nucleoside analog interfering with the viral RNA polymerase, has been evaluated in the similarly designed phase-III MOVe-OUT trial in unvaccinated high-risk outpatients with COVID-19. In contrast to favorable reports from the interim analysis, final trial results showed only a relative risk reduction of 30%. Up to now molnupiravir has not been approved by EMA but can be administered within a compassionate use program. We marginally recommend molnupiravir as early therapy in ambulatory cancer patients with COVID-19 (CIIt), in particular if more potent therapeutic options are contraindicated or not available. Due to its mutations inducing mechanism of action, molnupiravir is absolutely contraindicated in pregnant or breast-feeding women.
As antibody-based therapies and antiviral agents possess different mechanisms of action it is conceivable that combination therapies of these two drug classes might be beneficial, in particular in severely immunocompromised cancer patients. However, no published trial data is available so far, therefore no definite recommendations can be made on such a strategy. Similarly, it remains currently unclear whether prolonged treatment or repeated dosing might be an advisable strategy in immunocompromised cancer patients with prolonged viral shedding of SARS-CoV-2.
Inhaled corticosteroids such as budesonide or ciclesonide were assessed in several smaller studies in non-high-risk adults and in a large open-label RCT in high-risk adult out-patients with COVID-19. Although associated with improved patient-reported outcomes, there was no impact on hospitalization rates or mortality.81, 82, 83 We therefore don’t see sufficient evidence of benefit and recommend against administration of inhaled corticosteroids in cancer out-patients with COVID-19 who are not already on these medications as part of their standard of care (DIIt).
The serotonin re-uptake inhibitor fluvoxamine showed a lower likelihood of clinical deterioration compared to placebo in a small placebo-controlled RCT in non-high-risk adult out-patients with COVID-19. However, the study was underpowered to evaluate a potential impact on stronger endpoints.84 While fluvoxamine might be considered as an off-label treatment option, if antibody-based therapies or antivirals are unavailable, there is currently not enough evidence to make any definite recommendations in this regard. Similarly, data on the anti-inflammatory drug colchicine is inconclusive and as of yet lacking in high-risk patients thus precluding any definite recommendations.85
Immunosuppressant drugs, in particular dexamethasone, anti-IL6 or anti-IL1 monoclonal antibodies, or Janus kinase (JAK) inhibitors, are not recommended as treatment options in out-patient cancer patients with COVID-19 (DIII).
Hospitalized moderate disease (WHO score 4-5) and severe disease (WHO score 6-9)
For management of hospitalized cancer patients with COVID-19 it is helpful to differentiate between moderate disease, i.e. no or low-flow oxygen only (WHO score 4-5), and severe disease, i.e. high-flow oxygen, non-invasive (NIV) or mechanical ventilation (MV, WHO score 6-9). While in earlier phases of disease active viral replication seems to be a major pathogenetic factor, in later, more severe disease, hyperinflammation predominates resulting in different treatment strategies at different time points (see Table 5 ).86 Table 5 Recommendations on treatment of hospitalized cancer patients with moderate to severe COVID-19 (WHO score 4-9).
Table 5Population Intention Intervention SoR QoE Reference
Cancer patients with Covid-19 - hospitalized, moderate to severe [4-6] to shorten time to recovery Remdesivir B IIt 65,87,89,90,92
Cancer patients with Covid-19 - hospitalized, severe [7-9] to reduce mortality Remdesivir D IIt 65,87,89,90,92
Cancer patients with Covid-19 - hospitalized, moderate to severe [4-6], seronegative to reduce mortality Anti-S monoclonal antibodiesa B IIt 94,150
Cancer patients with Covid-19 - hospitalized, moderate to severe [4-6], seropositive to reduce mortality Anti-S monoclonal antibodiesa D IIt 94,150
Cancer patients with Covid-19 - hospitalized, moderate [4-5], seronegative to reduce mortality high-titer convalescent plasma C III 95, 96, 97
Cancer patients with Covid-19 - hospitalized, moderate [4-5], seropositive to reduce mortality high-titer convalescent plasma D IIt 95, 96, 97
Cancer patients with Covid-19 - hospitalized, severe [6-9] to reduce mortality high-titer convalescent plasma D IIt 96,98
Cancer patients with Covid-19 - hospitalized, moderate [4] to reduce mortality dexamethasone, anti-IL6 or anti-IL1 monoclonal antibodies, JAK inhibitors D IIt 99, 100, 101
Cancer patients with Covid-19 - hospitalized, moderate to severe [5-9] to reduce mortality dexamethasone A IIt 99
Cancer patients with Covid-19 - hospitalized, moderate to severe [5-6] and systemic inflammation to reduce mortality anti-IL6 monoclonal antibodies B IIt 102,103
Cancer patients with Covid-19 - hospitalized, severe [7-9] and systemic inflammation to reduce mortality anti-IL6 monoclonal antibodies C IIt 102, 103, 104
Cancer patients with Covid-19 - hospitalized, moderate to severe [5-6] and systemic inflammation to reduce mortality Anti-IL1 monoclonal antibodies C IIt 105,106
Cancer patients with Covid-19 - hospitalized, severe [7-9] and systemic inflammation to reduce mortality anti-IL1 monoclonal antibodies D IIta 104
Cancer patients with Covid-19 - hospitalized, moderate to severe [5-6] and systemic inflammation to reduce mortality JAK inhibitors C IIt 100,101,108
a if available against the locally predominant SARS-CoV-2 variant.
In hospitalized patients with moderate COVID-19 and hypoxic pneumonia and/or low-flow oxygen support, administration of remdesivir early after symptom onset was significantly associated with a shortened time to recovery in the ACTT-1 double-blind placebo-controlled RCT, with a trend towards reduced mortality in the low-flow oxygen group.87 The open-label DisCoVeRy trial and the interim analysis of the large, equally open-label WHO Solidarity trial failed to detect any clinical benefit of remdesivir treatment in hospitalized patients.88 , 89 The final analysis of the WHO Solidarity trial, however, showed a small, but significant reduction in mortality in hospitalized patients with oxygen support but without mechanical ventilation (14.6% vs 16.3%).90 No benefit was seen in patients already on mechanical ventilation.90 In a large observational multicenter study, remdesivir treated patients had a lower 14-day mortality compared to patients without, regardless of the mode of oxygen support.91 In cancer patients who might be particularly prone to impaired viral clearance, a large retrospective study also reported a potential benefit of remdesivir on mortality.92 In summary, we moderately recommend remdesivir for up to a maximum of 10 days in hospitalized cancer patients with moderate COVID-19 and with severe COVID-19 not yet on mechanical ventilation(WHO scale 4-6, BIIt), while the current evidence does not support the use of remdesivir in patients with mechanical ventilation or on ECMO (WHO scale 7-9, DIIt).
However, it is important to note that COVID-19 patients, and cancer patients in particular, often experience rapid clinical deterioration with escalation from low-flow oxygen to mechanical ventilation within 24 hours in some cases. We therefore recommend to account for the patient’s course of disease and consider remdesivir or other adjuncts such as IL-6 or JAK inhibitors (see below) within the first 24 hours after ICU admission, even if mechanical ventilation was already initiated.
With regard to the oral antivirals nirmatrelvir/ritonavir and molnupiravir, no published trial data is available in patients hospitalized because of COVID-19 with need for supplemental oxygen, so currently no recommendation can be made in this indication. One retrospective trial in hospitalized patients with mild COVID-19 without supplemental oxygen showed a reduced disease progression (including death) in nirmatrelvir/ritonavir or molnupiravir treated patients > 65 years of age.93 Obviously, to patients who are hospitalized for other reasons without specific symptoms sugggesting COVID-19 pneumonia, all considerations for the out-patient setting as described above apply.
In hospitalized patients with moderate to severe disease, the randomized, open-label RECOVERY trial evaluated the impact of the monoclonal antibody combination casirivimab/imdevimab on mortality.94 While no effect was seen in seropositive patients, mortality at day 28 in seronegative patients was significantly reduced by monoclonal anti-S antibody therapy (24% vs 30%).94 While casirivimab/imdevimab are no longer considered sufficiently active against the currently predominant omicron BA.4/5 variants,60 the strategy of administering monoclonal anti-S antibodies to hospitalized seronegative cancer patients seems sensible, if preparations with activity against the locally predominant variants are available, especially taking into account the often impaired humoral immune response in cancer patients. We therefore moderately recommend monoclonal anti-S antibodies to hospitalized seronegative cancer patients with WHO scale 4-6 disease (BIIt). As only 2% of patients in the RECOVERY trial were on mechanical ventilation, no recommendation can be made with regard to a potential benefit of monoclonal antibodies in patients with WHO scale 7-9 disease. In seropositive hospitalized patients, monoclonal anti-S antibodies are not recommended (DIIt).
Clinical data on high-titer CP in hospitalized patients with moderate disease is so far inconclusive. Some retrospective analyses have noted a benefit especially in immunocompromised patients and patients with hematological malignancies.95, 96, 97 Hence, high-titer CP administration can be discussed in selected seronegative cancer patients with moderate disease if monoclonal antibodies against the locally predominant variants are not accessible (CIII). In seropositive patients as well as in those with severe disease, the current evidence does not support the use of CP (DIIt).95 , 96 , 98
Immunosuppressive agents, in particular dexamethasone, are an important part of therapy in severely ill COVID-19 patients. In hospitalized patients without oxygen support (WHO scale 4), however, dexamethasone was associated with a trend towards increased mortality in the large RECOVERY RCT and is therefore contraindicated in these patients (DIIt).99 The same holds true for other immunosuppressive agents, such as anti-IL6 monoclonal antibodies or JAK inhibitors in this patients population (DIIt).100 , 101
In patients requiring oxygen support, the addition of dexamethasone at 6 mg per day for a 10-day course significantly improved clinical outcomes and reduced mortality in the RECOVERY RCT by about a fifth in patients with low- or high-flow oxygen (WHO scale 5-6) and by about a third in mechanically ventilated patients (WHO scale 7-9) and is therefore strongly recommended in these patient population (AIIt).99
If systemic inflammation is present, e.g. highly elevated CRP levels in the absence of bacterial infection, the addition of anti-IL-6 monoclonal antibodies, such as tocilizumab or sarilumab, to dexamethasone treatment can be considered in patients with oxygen support but without mechanical (WHO scale 5-6) and is recommended here with moderate strength (BIIt).102 , 103 Ideally, anti-IL-6 antibodies are initiated early after symptom onset and prior to MV (BIIt). As discussed above, exceptions can be made in rapidly-progressing patients, but anti-IL-6 monoclonal antibodies are only marginally recommended in patients with recently initiated mechanical (WHO scale 7-9, CIIt)102, 103, 104 and are not indicated in patients with already prolonged mechanical.
As an alternative to anti-IL6 antibodies, anti-IL1 monoclonal antibodies, such as anakinra, can be considered in hospitalized patients with systemic inflammation and low- or high-flow oxygen support (WHO scale 5-6), however, the available data is less conclusive (CIIt)105 , 106. In patients on NIV or mechanical ventilation, anti-IL1 monoclonal antibodies failed to demonstrate a survival benefit and are not indicated (DIIta).104
JAK inhibitors, such as baricitinib and tofacitinib, demonstrated a survival benefit in addition to dexamethasone in hospitalized patients, especially in patients on high-flow oxygen or NIV (WHO scale 5-6), and represent another alternative in patients with systemic inflammation (CIIt).101 , 107 , 108 Patients on mechanical ventilation or ECMO were not included in these trials, hence no recommendations can be made in this regard.
Monoclonal anti-IL6 and anti-IL1 antibodies and JAK inhibitors must not be given concomitantly (DIII) and the second immunomodulator of choice should always be given in addition to standard dexamethasone treatment (AIIt).
Supportive therapy
Correction of vitamin D deficiency in cancer patients may positively influence the clinical outcome in case of COVID-19, though the evidence is based on mainly indirect observational data, since patients with vitamin D deficiency experience inferior COVID-19 outcomes (BIIt, see Table 6 ).109 However, vitamin D supplementation is not recommended as a prophylactic measure in cancer patients without vitamin D deficiency (DIIt).110 , 111 Table 6 Recommendations on supportive care in cancer patients with COVID-19.
Table 6Population Intention Intervention SoR QoE Reference
Cancer patients, uninfected [0] or with Covid-19 [1-9], with vitamin D deficiency to improve clinical outcome in case of COVID-19 vitamin D supplementation B IIt 109
Cancer patients, uninfected [0] or with Covid-19 [1-9], without vitamin D deficiency to improve clinical outcome in case of COVID-19 vitamin D supplementation D IItr 110,111
Cancer patients with Covid-19 - ambulatory, mild [1-3] to prevent thromboembolic complications low-dose LMWH C III 112
Cancer patients with Covid-19 - hospitalized, moderate to severe [4-9] to prevent thromboembolic complications low-dose LMWH A IIt 113,114
Cancer patients with Covid-19 - hospitalized, moderate [4-5] plus additional risk factors to prevent thromboembolic complications and reduce mortality therapeutic anticoagulation B IIt 114, 115, 116, 117
Cancer patients with Covid-19 - hospitalized, severe [6-9] to prevent thromboembolic complications and reduce mortality routine intermediate-dose LMWH D IIt 115,118
Cancer patients with Covid-19 - hospitalized, severe [6-9] to prevent thromboembolic complications and reduce mortality routine therapeutic anticoagulation D IIt 114,119,151
Cancer patients are per se considered to be at increased risk of thromboembolic complications. Prevention of these complications in the setting of COVID-19 therefore deserves particular attention. In cancer patients with mild COVID-19 who are still in outpatient management, thromboembolic prophylaxis with low-dose low molecular weight heparin (LMWH) can be considered especially in patients with additional risk factors, e.g. immobilization (CIII).112 This rationale builds on the observation that COVID-19 patients have been reported to often present with thromboembolic complications already within the first 24 hours after hospital admission.112 Contraindications must be considered. In hospitalized cancer patients with moderate to severe COVID-19 thromboembolic prophylaxis with low-dose LMWH is strongly recommended (AIIt).113 , 114 In patients with moderate COVID-19 plus additional risk factors, e.g. significantly elevated D-dimer levels or prior thromboembolic complications, therapeutic anticoagulation can be considered to prevent further thromboembolic complications and to reduce mortality in patients with a low risk of bleeding (BIIt)114, 115, 116, 117. In severely-ill patients on high-flow oxygen, MV or ECMO, routine application of intermediate-dose LMWH is discouraged (DIIt).115 , 118 Likewise, routine therapeutic anticoagulation is not recommended in this patient cohort (DIIt)114 , 118 , 119 outside of specific indications. In cancer patients at high-risk for thromboembolic complications and low bleeding risk, prophylactic anticoagulation may be continued after discharge from hospitalization for COVID-19 (CIIt).120 Aspirin should not be initiated as treatment of COVID-19 in hospitalized cancer patients not already on chronic aspirin therapy due to other indications (DIIt).121
Concerning neutropenic cancer patients with COVID-19, granulocyte colony stimulating factor (G-CSF) should not be routinely administered outside of current guidelines (DIII) as G-CSF application in neutropenic cancer patients was reported to be associated with a risk of worsening respiratory situation.122
Generally, current guidelines for ICU management in COVID-19 patients should also be applied to cancer patients with COVID-19, including the timepoint of intubation (which should not be delayed to prolong NIV respiratory support) and the definition of therapy goals (AIII).
Long-COVID
A subset of patients suffers from a variety of symptoms after a SARS-CoV-2 infection.123 A clinical case definition of a post-COVID-19 condition was provided by the WHO: Long-COVID occurs in patients with a history of probable or confirmed SARS-CoV-2 infection, usually three months from onset, with symptoms that last for at least two months and cannot be explained by an alternative diagnosis.123 Common symptoms include, but are not limited to, fatigue, shortness of breath, and cognitive dysfunction, and generally have an impact on everyday functioning.123
In cancer patients, long-COVID is common with a prevalence up to 15% and in particular older cancer patients with more comorbidities seem to be at increased risk of developing COVID-19 sequelae.124 Cancer patients with COVID-19 have a higher one-year all-cause mortality than non-cancer COVID-19 patients, yet cancer patients have no more symptoms after one year post COVID-19 than other patients.125 While the currently available data is insufficient to make specific evidence-based recommendations for the prevention and treatment of long-COVID other than those presented above for the prevention and treatment of (severe) COVID-19, we think it useful to briefly summarize possible strategies.
Patients with hematological malignancies experience prolonged viral shedding and have a higher maximal viral load than patients without malignancies.126 Anti-CD20 treatment <1 year (odds ratio (OR) 3.04), SCT/cellular therapy <1 year (OR 3.64), and chronic lymphopenia (<500/μl) (OR 3.78) are predictors for SARS-CoV-2 persistence.127 It remains, however, unclear whether viral clearance contributes to preventing long-COVID.128
In a prospective non-interventional study, treatment with remdesivir was independently associated with a 35% risk reduction of long-COVID.129 Impaired pulmonary function or dyspnea are frequent symptoms in cancer patients with long-COVID.124 In a small RCT in patients with clinical-radiological suspicion of COVID-19 and requirement of oxygen support, a short course of methylprednisolone was shown to improve pulmonary function at day 120.130 Furthermore, in COVID-19 patients with interstitial lung disease, the application of prednisolone was reported to be beneficial with regard to prevention of pulmonary fibrosis with permanent functional deficit.131 In a small observational study, systemic corticosteroids administered during the acute phase of COVID-19 were associated with reduced symptoms and better quality of life one year after initial admission for COVID-19.132 Of note, in an observational study in health care workers, vaccination against COVID-19 was associated with a 84% relative risk reduction of long-COVID after three doses of vaccine, again highlighting the importance of vaccination.133
To improve a variety of long-COVID symptoms, e.g. fatigue or olfactory impairment, brain function, functional capacity or emotional well-being, a plethora of alimentary supplementation of vitamins, minerals, amino acids, and plant extracts have been suggested, as well as systemic prednisone, nasal irrigation or hyperbaric oxygen therapy, so far with little to no conclusive evidence of clinical benefit.134, 135, 136, 137, 138 Furthermore, multidisciplinary outpatient neuropsychological rehabilitation measures have been reported as helpful.139 , 140
Conclusion and outlook
Cancer patients still constitute a population at high risk of severe and prolonged COVID-19. Major advances with regard to development of vaccines and therapeutic agents against COVID-19 have significantly broadened the options for prevention and treatment of this infectious disease. Improving response to vaccination in immunocompromised cancer patients, devising optimal treatment strategies for these patients and addressing the symptoms of long-COVID still remain significant challenges.
Author contributions
All authors actively participated in the guideline panel. NG coordinated the guideline panel. NG, EB, ES, and MvLT wrote the final version of the manuscript. All authors agreed upon guideline topics, performed a systematic literature search, extracted and rated the data, discussed and agreed upon the final recommendations, helped in writing and critically revised the first draft of the manuscript, and approved the final version of the manuscript.
Funding
No external funding was provided for any part of the writing process of this guideline.
Declaration of interests
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
NG reports honoraria from AbbVie, AstraZeneca, GSK, Hexal, MSD, travel grants from Janssen, and participation on an advisory board to AstraZeneca. MH received lecture fees by Amgen, AstraZeneca, EusaPharma, Celgene, Janssen, Jazz Pharma, and Takeda, and served on advisory boards of Amgen, EusaPharma, Janssen, and Sanofi. PK reports grants or contracts from the Ministry of Education and Research (BMBF) B-FAST and NAPKON of the Network University Medicine (NUM) and the State of North Rhine-Westphalia; consulting fees Ambu GmbH, Gilead Sciences, Mundipharma Research Limited, Noxxon N.V., and Pfizer Pharma; honoraria for lectures from Akademie für Infektionsmedizin e.V., Ambu GmbH, Astellas Pharma, BioRad Laboratories Inc., European Confederation of Medical Mycology, Gilead Sciences, GPR Academy Ruesselsheim, HELIOS Kliniken GmbH, Lahn-Dill-Kliniken GmbH, medupdate GmbH, MedMedia, MSD Sharp & Dohme GmbH, Pfizer Pharma GmbH, Scilink Comunicación Científica SC and University Hospital and LMU Munich; participation on an advisory board from Ambu GmbH, Gilead Sciences, Mundipharma Research Limited and Pfizer Pharma; a pending patent currently reviewed at the German Patent and Trade Mark Office; other non-financial interests from Elsevier, Wiley and Taylor & Francis online outside the submitted work. RK received research grants from Merck and Pfizer and received speaker honoraria by Pfizer, Gilead, Astellas, Basilea, Merck, Angelini, and Shionogi. OP has received honoraria or travel support from Gilead, Jazz, Novartis, Pfizer, and Therakos; he has received research support from Incyte and Priothera; he is a member of advisory boards to Equilium Bio, Jazz, Gilead, Novartis, MSD, Omeros, Priothera, Shionogi, and SOBI. MS has received speaker honoraria from Roche, Novartis, BMS, Lilly, AstraZeneca, and Celgene. RSB received speaker honoraria and/or travel grants and/or has been consultant to AbbVie, Amgen, AstraZeneca, BMS, Celgene, Eusa Pharma, Gilead, Ipsen, Janssen, Merck, MSD, Novartis, Pfizer, Roche, and Sanofi. KS received speaker honoraria and/or travel grants by and/or has been consultant to: Abbvie, Alexa, Bristol Myers Squibb, Jazz Pharmaceuticals, Leo, Pfizer, Sanofi, Servier, Sobi, Stemline, Takeda. JS has received research grants by the Ministry of Education and Research (BMBF) and Basilea Pharmaceuticals Inc.; has received speaker honoraria by Pfizer Inc., Gilead, and AbbVie; has been a consultant to Gilead, Produkt&Markt GmbH, Alvea Vax, and Micron Researc and has received travel grants by the German Society of Infectious Diseases (DGI e.V.) and Meta-Alexander Foundation. OAC reports grants or contracts from Amplyx, Basilea, BMBF, Cidara, DZIF, EU-DG RTD (101037867), F2G, Gilead, Matinas, MedPace, MSD, Mundipharma, Octapharma, Pfizer, Scynexis; Consulting fees from Abbvie, Amplyx, Biocon, Biosys, Cidara, Da Volterra, Gilead, IQVIA, Janssen, Matinas, MedPace, Menarini, Molecular Partners, MSG-ERC, Noxxon, Octapharma, Pardes, Pfizer, PSI, Scynexis, Seres; Honoraria for lectures from Abbott, Abbvie, Al-Jazeera Pharmaceuticals, Astellas, Gilead, Grupo Biotoscana/United Medical/Knight, Hikma, MedScape, MedUpdate, Merck/MSD, Mylan, Noscendo, Pfizer, Shionogi; Payment for expert testimony from Cidara; Participation on a Data Safety Monitoring Board or Advisory Board from Actelion, Allecra, Cidara, Entasis, IQVIA, Janssen, MedPace, Paratek, PSI, Pulmocide, Shionogi, The Prime Meridian Group; A patent at the German Patent and Trade Mark Office (DE 10 2021 113 007.7); Stocks from CoRe Consulting; Other interests from DGHO, DGI, ECMM, ISHAM, MSG-ERC, Wiley. CTR reports honoraria from and has served as a consultant to abbvie, AstraZeneca, BMS, Gilead Sciences, GSK, Incyte, Ipsen, Janssen, Lilly, Pfizer Pharma, MSD, Novartis. MvLT reports honoraria from Celgene, Gilead, Chugai, Janssen, Novartis, Amgen, Takeda, BMS, Medac, Oncopeptides, Merck, CDDF, Abbvie, AstraZeneca, Pfizer, Thermofisher, GSK and Shionogi and research funding from BMBF, Deutsche Jose Carreras Leukämie-Stiftung, IZKF Jena, DFG, Novartis, Gilead, Deutsche Krebshilfe, Celgene, Oncopeptides and receives support from the German Research Foundation within the Collaborative Research Centre/Transregio 124 FungiNet, DFG project no. 210879364 (project A1) as well as from the Deutsche Krebshilfe OncoReVir Registry (No. 70113851). EB, ES, GB, MMR, BH, HHH, MKa, YK, WK, MKo, SCM, RS, FW, BW, and HHW have no conflicts of interest to declare.
Uncited References
148; 149.
==== Refs
References
1 Guan W.J. Ni Z.Y. Hu Y. Liang W.H. Ou C.Q. He J.X. Clinical Characteristics of Coronavirus Disease 2019 in China N Engl J Med 382 18 2020 1708 1720 32109013
2 Zhu N. Zhang D. Wang W. Li X. Yang B. Song J. A Novel Coronavirus from Patients with Pneumonia in China, 2019 N Engl J Med 382 8 2020 727 733 31978945
3 Pagano L. Salmanton-Garcia J. Marchesi F. Busca A. Corradini P. Hoenigl M. COVID-19 infection in adult patients with hematological malignancies: a European Hematology Association Survey (EPICOVIDEHA) J Hematol Oncol 14 1 2021 168 34649563
4 Ruthrich M.M. Giessen-Jung C. Borgmann S. Classen A.Y. Dolff S. Gruner B. COVID-19 in cancer patients: clinical characteristics and outcome-an analysis of the LEOSS registry Ann Hematol 100 2 2021 383 393 33159569
5 Lee L.Y.W. Cazier J.B. Starkey T. Briggs S.E.W. Arnold R. Bisht V. COVID-19 prevalence and mortality in patients with cancer and the effect of primary tumour subtype and patient demographics: a prospective cohort study Lancet Oncol 21 10 2020 1309 1316 32853557
6 OnCovid Study G. Pinato D.J. Patel M. Scotti L. Colomba E. Dolly S. Time-Dependent COVID-19 Mortality in Patients With Cancer: An Updated Analysis of the OnCovid Registry JAMA Oncol 8 1 2022 114 122 34817562
7 Pagano L. Salmanton-Garcia J. Marchesi F. Lopez-Garcia A. Lamure S. Itri F. COVID-19 in vaccinated adult patients with hematological malignancies: preliminary results from EPICOVIDEHA Blood 139 10 2022 1588 1592 34748627
8 WHO. World Health Organization Coronavirus disease (COVID-19) Situation reports [Available from: https://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports.
9 Madhi S.A. Kwatra G. Myers J.E. Jassat W. Dhar N. Mukendi C.K. Population Immunity and Covid-19 Severity with Omicron Variant in South Africa N Engl J Med 386 14 2022 1314 1326 35196424
10 Mahase E. Covid-19: Hospital admission 50-70% less likely with omicron than delta, but transmission a major concern BMJ 375 2021 n3151
11 Viana R. Moyo S. Amoako D.G. Tegally H. Scheepers C. Althaus C.L. Rapid epidemic expansion of the SARS-CoV-2 Omicron variant in southern Africa Nature 603 7902 2022 679 686 35042229
12 Pinato D.J. Aguilar-Company J. Ferrante D. Hanbury G. Bower M. Salazar R. Outcomes of the SARS-CoV-2 omicron (B.1.1.529) variant outbreak among vaccinated and unvaccinated patients with cancer in Europe: results from the retrospective, multicentre, OnCovid registry study Lancet Oncol 23 7 2022 865 875 35660139
13 Bestvina C.M. Whisenant J.G. Torri V. Cortellini A. Wakelee H. Peters S. Coronavirus Disease 2019 Outcomes, Patient Vaccination Status, and Cancer-Related Delays During the Omicron Wave: A Brief Report From the TERAVOLT Analysis JTO Clin Res Rep 3 8 2022 100335 35619644
14 Polack F.P. Thomas S.J. Kitchin N. Absalon J. Gurtman A. Lockhart S. Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine N Engl J Med 383 27 2020 2603 2615 33301246
15 Heath P.T. Galiza E.P. Baxter D.N. Boffito M. Browne D. Burns F. Safety and Efficacy of NVX-CoV2373 Covid-19 Vaccine N Engl J Med 385 13 2021 1172 1183 34192426
16 Baden L.R. El Sahly H.M. Essink B. Kotloff K. Frey S. Novak R. Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine N Engl J Med 2020
17 Sadoff J. Gray G. Vandebosch A. Cardenas V. Shukarev G. Grinsztejn B. Safety and Efficacy of Single-Dose Ad26.COV2.S Vaccine against Covid-19 N Engl J Med 384 23 2021 2187 2201 33882225
18 Voysey M. Clemens S.A.C. Madhi S.A. Weckx L.Y. Folegatti P.M. Aley P.K. Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil South Africa, and the UK. Lancet 2020
19 Levin M.J. Ustianowski A. De Wit S. Launay O. Avila M. Templeton A. Intramuscular AZD7442 (Tixagevimab-Cilgavimab) for Prevention of Covid-19 N Engl J Med 386 23 2022 2188 2200 35443106
20 O'Brien M.P. Forleo-Neto E. Musser B.J. Isa F. Chan K.C. Sarkar N. Subcutaneous REGEN-COV Antibody Combination to Prevent Covid-19 N Engl J Med 385 13 2021 1184 1195 34347950
21 Weinreich D.M. Sivapalasingam S. Norton T. Ali S. Gao H. Bhore R. REGEN-COV Antibody Combination and Outcomes in Outpatients with Covid-19 N Engl J Med 385 23 2021 e81 34587383
22 Gupta A. Gonzalez-Rojas Y. Juarez E. Crespo Casal M. Moya J. Falci D.R. Early Treatment for Covid-19 with SARS-CoV-2 Neutralizing Antibody Sotrovimab N Engl J Med 385 21 2021 1941 1950 34706189
23 Dougan M. Nirula A. Azizad M. Mocherla B. Gottlieb R.L. Chen P. Bamlanivimab plus Etesevimab in Mild or Moderate Covid-19 N Engl J Med 385 15 2021 1382 1392 34260849
24 Jayk Bernal A. Gomes da Silva M.M. Musungaie D.B. Kovalchuk E. Gonzalez A. Delos Reyes V. Molnupiravir for Oral Treatment of Covid-19 in Nonhospitalized Patients N Engl J Med 386 6 2022 509 520 34914868
25 Hammond J. Leister-Tebbe H. Gardner A. Abreu P. Bao W. Wisemandle W. Oral Nirmatrelvir for High-Risk, Nonhospitalized Adults with Covid-19 N Engl J Med 386 15 2022 1397 1408 35172054
26 Gottlieb R.L. Vaca C.E. Paredes R. Mera J. Webb B.J. Perez G. Early Remdesivir to Prevent Progression to Severe Covid-19 in Outpatients N Engl J Med 386 4 2022 305 315 34937145
27 Becerril-Gaitan A. Vaca-Cartagena B.F. Ferrigno A.S. Mesa-Chavez F. Barrientos-Gutierrez T. Tagliamento M. Immunogenicity and risk of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection after Coronavirus Disease 2019 (COVID-19) vaccination in patients with cancer: a systematic review and meta-analysis Eur J Cancer 160 2022 243 260 34794855
28 Barriere J. Chamorey E. Adjtoutah Z. Castelnau O. Mahamat A. Marco S. Impaired immunogenicity of BNT162b2 anti-SARS-CoV-2 vaccine in patients treated for solid tumors Ann Oncol 32 8 2021 1053 1055 33932508
29 Mair M.J. Berger J.M. Berghoff A.S. Starzer A.M. Ortmayr G. Puhr H.C. Humoral Immune Response in Hematooncological Patients and Health Care Workers Who Received SARS-CoV-2 Vaccinations JAMA Oncol 8 1 2022 106 113 34591965
30 Goubet A.G. Dubuisson A. Geraud A. Danlos F.X. Terrisse S. Silva C.A.C. Prolonged SARS-CoV-2 RNA virus shedding and lymphopenia are hallmarks of COVID-19 in cancer patients with poor prognosis Cell Death Differ 28 12 2021 3297 3315 34230615
31 Passamonti F. Romano A. Salvini M. Merli F. Porta M.G.D. Bruna R. COVID-19 elicits an impaired antibody response against SARS-CoV-2 in patients with haematological malignancies Br J Haematol 195 3 2021 371 377 34272724
32 Giesen N. Sprute R. Ruthrich M. Khodamoradi Y. Mellinghoff S.C. Beutel G. Evidence-based management of COVID-19 in cancer patients: Guideline by the Infectious Diseases Working Party (AGIHO) of the German Society for Haematology and Medical Oncology (DGHO) Eur J Cancer 140 2020 86 104 33068941
33 Giesen N. Sprute R. Ruthrich M. Khodamoradi Y. Mellinghoff S.C. Beutel G. 2021 update of the AGIHO guideline on evidence-based management of COVID-19 in patients with cancer regarding diagnostics, viral shedding, vaccination and therapy Eur J Cancer 147 2021 154 160 33676266
34 Ullmann A.J. Akova M. Herbrecht R. Viscoli C. Arendrup M.C. Arikan-Akdagli S. ESCMID* guideline for the diagnosis and management of Candida diseases 2012: adults with haematological malignancies and after haematopoietic stem cell transplantation (HCT) Clin Microbiol Infect 18 7 2012 53 67 23137137
35 Lee L.Y.W. Starkey T. Ionescu M.C. Little M. Tilby M. Tripathy A.R. Vaccine effectiveness against COVID-19 breakthrough infections in patients with cancer (UKCCEP): a population-based test-negative case-control study Lancet Oncol 23 6 2022 748 757 35617989
36 Margalit O. Shacham-Shmueli E. Itay A. Berger R. Halperin S. Jurkowicz M. Seropositivity and neutralising antibodies at six months after BNT162b2 vaccination in patients with solid tumours Eur J Cancer 168 2022 51 55 35439660
37 Mair M.J. Berger J.M. Mitterer M. Gansterer M. Bathke A.C. Trutschnig W. Third dose of SARS-CoV-2 vaccination in hemato-oncological patients and health care workers: immune responses and adverse events - a retrospective cohort study Eur J Cancer 165 2022 184 194 35248840
38 Fendler A. de Vries E.G.E. GeurtsvanKessel C.H. Haanen J.B. Wörmann B. Turajlic S. COVID-19 vaccines in patients with cancer: immunogenicity, efficacy and safety Nat Rev Clin Oncol 2022 1 17
39 Lee L.Y.W. Ionescu M.C. Starkey T. Little M. Tilby M. Tripathy A.R. COVID-19: Third dose booster vaccine effectiveness against breakthrough coronavirus infection, hospitalisations and death in patients with cancer: A population-based study Eur J Cancer 175 2022 1 10 36084618
40 Heudel P. Favier B. Solodky M.L. Assaad S. Chaumard N. Tredan O. Survival and risk of COVID-19 after SARS-COV-2 vaccination in a series of 2391 cancer patients Eur J Cancer 165 2022 174 183 35245864
41 Debie Y. Vandamme T. Goossens M.E. van Dam P.A. Peeters M. Antibody titres before and after a third dose of the SARS-CoV-2 BNT162b2 vaccine in patients with cancer Eur J Cancer 163 2022 177 179 35077960
42 Di Noia V. Pimpinelli F. Renna D. Maccallini M.T. Gariazzo L. Cosimati A. Rapid decline of humoral response to two doses of BNT162b2 vaccine in patients with solid cancer after six months: The urgent need of the additional dose! Eur J Cancer 165 2022 169 173 35245863
43 Piechotta V. Mellinghoff S.C. Hirsch C. Brinkmann A. Iannizzi C. Kreuzberger N. Effectiveness, immunogenicity, and safety of COVID-19 vaccines for individuals with hematological malignancies: a systematic review Blood Cancer J 12 5 2022 86 35641489
44 Chalkias S. Harper C. Vrbicky K. Walsh S.R. Essink B. Brosz A. A Bivalent Omicron-Containing Booster Vaccine against Covid-19 N Engl J Med 387 14 2022 1279 1291 36112399
45 Mai A.S. Lee A. Tay R.Y.K. Shapiro L. Thakkar A. Halmos B. Booster doses of COVID-19 vaccines for patients with haematological and solid cancer: a systematic review and individual patient data meta-analysis Eur J Cancer 172 2022 65 75 35753213
46 Oosting S.F. van der Veldt A.A.M. GeurtsvanKessel C.H. Fehrmann R.S.N. van Binnendijk R.S. Dingemans A.C. mRNA-1273 COVID-19 vaccination in patients receiving chemotherapy, immunotherapy, or chemoimmunotherapy for solid tumours: a prospective, multicentre, non-inferiority trial Lancet Oncol 2021
47 Garreffa E. Hamad A. O'Sullivan C.C. Hazim A.Z. York J. Puri S. Regional lymphadenopathy following COVID-19 vaccination: Literature review and considerations for patient management in breast cancer care Eur J Cancer 159 2021 38 51 34731748
48 Soyfer V. Gutfeld O. Shamai S. Schlocker A. Merimsky O. COVID-19 Vaccine-Induced Radiation Recall Phenomenon Int J Radiat Oncol Biol Phys. 110 4 2021 957 961 33677050
49 Steber C.R. Ponnatapura J. Hughes R.T. Farris M.K. Rapid Development of Clinically Symptomatic Radiation Recall Pneumonitis Immediately Following COVID-19 Vaccination Cureus 13 4 2021 e14303 33968515
50 Stewart R. McDowell L. Radiation Recall Phenomenon Following COVID-19 Vaccination Int J Radiat Oncol Biol Phys. 111 3 2021 835 836 34153381
51 Ram R. Hagin D. Kikozashvilli N. Freund T. Amit O. Bar-On Y. Safety and Immunogenicity of the BNT162b2 mRNA COVID-19 Vaccine in Patients after Allogeneic HCT or CD19-based CART therapy-A Single-Center Prospective Cohort Study Transplant Cell Ther 27 9 2021 788 794 34214738
52 Piñana J.L. López-Corral L. Martino R. Montoro J. Vazquez L. Pérez A. SARS-CoV-2-reactive antibody detection after SARS-CoV-2 vaccination in hematopoietic stem cell transplant recipients: Prospective survey from the Spanish Hematopoietic Stem Cell Transplantation and Cell Therapy Group (GETH-TC) Am J Hematol 2021
53 Ali H. Ngo D. Aribi A. Arslan S. Dadwal S. Marcucci G. Safety and Tolerability of SARS-CoV2 Emergency-Use Authorized Vaccines for Allogeneic Hematopoietic Stem Cell Transplant Recipients Transplant Cell Ther 27 11 2021 938 e931-938.e936
54 Barriere J. Carles M. Audigier-Valette C. Re D. Adjtoutah Z. Seitz-Polski B. Third dose of anti-SARS-CoV-2 vaccine for patients with cancer: Should humoral responses be monitored? A position article Eur J Cancer 162 2022 182 193 35016032
55 Gilbert P.B. Montefiori D.C. McDermott A.B. Fong Y. Benkeser D. Deng W. Immune correlates analysis of the mRNA-1273 COVID-19 vaccine efficacy clinical trial Science 2021 eab3435
56 Mellinghoff S.C. Mayer L. Robrecht S. Weskamm L.M. Dahlke C. Gruell H. SARS-CoV-2-specific cellular response following third COVID-19 vaccination in patients with chronic lymphocytic leukemia Haematologica 107 10 2022 2480 2484 35734927
57 Fendler A. Shepherd S.T.C. Au L. Wilkinson K.A. Wu M. Schmitt A.M. Immune responses following third COVID-19 vaccination are reduced in patients with hematological malignancies compared to patients with solid cancer Cancer Cell. 40 4 2022 438 35413273
58 Ruthrich M.M. Giesen N. Mellinghoff S.C. Rieger C.T. von Lilienfeld-Toal M. Cellular Immune Response after Vaccination in Patients with Cancer-Review on Past and Present Experiences Vaccines (Basel) 10 2 2022
59 Rüthrich M.M. Giesen N. Mellinghoff S.C. Rieger C.T. von Lilienfeld-Toal M. Cellular Immune Response after Vaccination in Patients with Cancer-Review on Past and Present Experiences Vaccines (Basel) 2 2022 10
60 Takashita E. Yamayoshi S. Simon V. van Bakel H. Sordillo E.M. Pekosz A. Efficacy of Antibodies and Antiviral Drugs against Omicron BA.2.12.1, BA.4, and BA.5 Subvariants N Engl J Med 387 5 2022 468 470 35857646
61 Cao Y. Yisimayi A. Jian F. Song W. Xiao T. Wang L. BA.2.12.1, BA.4 and BA.5 escape antibodies elicited by Omicron infection Nature 608 7923 2022 593 602 35714668
62 Stuver R. Shah G.L. Korde N.S. Roeker L.E. Mato A.R. Batlevi C.L. Activity of AZD7442 (tixagevimab-cilgavimab) against Omicron SARS-CoV-2 in patients with hematologic malignancies Cancer Cell. 40 6 2022 590 591 35598602
63 Shoham S. Bloch E.M. Casadevall A. Hanley D. Lau B. Gebo K. Transfusing convalescent plasma as post-exposure prophylaxis against SARS-CoV-2 infection: a double-blinded, phase 2 randomized, controlled trial Clin Infect Dis 2022
64 WHOWGotC Characterisation Management of C-i. A minimal common outcome measure set for COVID-19 clinical research Lancet Infect Dis 20 8 2020 e192 e197 32539990
65 Consortium WHOSTPan H. Peto R. Henao-Restrepo A.M. Preziosi M.P. Sathiyamoorthy V. Repurposed Antiviral Drugs for Covid-19-Interim WHO Solidarity Trial Results N Engl J Med 2020 10.1056/NEJMoa2023184
66 Axfors C. Schmitt A.M. Janiaud P. Van't Hooft J. Abd-Elsalam S. Abdo E.F. Mortality outcomes with hydroxychloroquine and chloroquine in COVID-19 from an international collaborative meta-analysis of randomized trials Nat Commun 12 1 2021 2349 33859192
67 Cavalcanti A.B. Zampieri F.G. Rosa R.G. Azevedo L.C.P. Veiga V.C. Avezum A. Hydroxychloroquine with or without Azithromycin in Mild-to-Moderate Covid-19 N Engl J Med 383 21 2020 2041 2052 32706953
68 Group RC Lopinavir-ritonavir in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial Lancet 396 10259 2020 1345 1352 33031764
69 Amani B. Amani B. Zareei S. Zareei M. Efficacy and safety of arbidol (umifenovir) in patients with COVID-19: A systematic review and meta-analysis Immun Inflamm Dis 9 4 2021 1197 1208 34347937
70 Ozlusen B. Kozan S. Akcan R.E. Kalender M. Yaprak D. Peltek I.B. Effectiveness of favipiravir in COVID-19: a live systematic review Eur J Clin Microbiol Infect Dis 40 12 2021 2575 2583 34347191
71 Cruciani M. Pati I. Masiello F. Malena M. Pupella S. De Angelis V. Ivermectin for Prophylaxis and Treatment of COVID-19: A Systematic Review and Meta-Analysis Diagnostics (Basel) 9 2021 11 35054180
72 EMA. Regkirona product information [Available from: https://www.ema.europa.eu/en/documents/product-information/regkirona-epar-procut-information_en.pdf.
73 Montgomery H. Hobbs F.D.R. Padilla F. Arbetter D. Templeton A. Seegobin S. 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
74 Fendler A. de Vries E.G.E. GeurtsvanKessel C.H. Haanen J.B. Wormann B. Turajlic S. COVID-19 vaccines in patients with cancer: immunogenicity, efficacy and safety Nat Rev Clin Oncol 19 6 2022 385 401 35277694
75 Libster R. Perez Marc G. Wappner D. Coviello S. Bianchi A. Braem V. Early High-Titer Plasma Therapy to Prevent Severe Covid-19 in Older Adults N Engl J Med 384 7 2021 610 618 33406353
76 Korley F.K. Durkalski-Mauldin V. Yeatts S.D. Schulman K. Davenport R.D. Dumont L.J. Early Convalescent Plasma for High-Risk Outpatients with Covid-19 N Engl J Med 385 21 2021 1951 1960 34407339
77 Sullivan D.J. Gebo K.A. Shoham S. Bloch E.M. Lau B. Shenoy A.G. Early Outpatient Treatment for Covid-19 with Convalescent Plasma N Engl J Med 386 18 2022 1700 1711 35353960
78 Wong C.K.H. Au I.C.H. Lau K.T.K. Lau E.H.Y. Cowling B.J. Leung G.M. Real-world effectiveness of molnupiravir and nirmatrelvir plus ritonavir against mortality, hospitalisation, and in-hospital outcomes among community-dwelling, ambulatory patients with confirmed SARS-CoV-2 infection during the omicron wave in Hong Kong: an observational study Lancet 400 10359 2022 1213 1222 36216007
79 Arbel R. Wolff Sagy Y. Hoshen M. Battat E. Lavie G. Sergienko R. Nirmatrelvir Use and Severe Covid-19 Outcomes during the Omicron Surge N Engl J Med 387 9 2022 790 798 36001529
80 NIH. [Available from: https://www.covid19treatmentguidelines.nih.gov/therapies/antiviral-therapy/ritonavir-boosted-nirmatrelvir–paxlovid-/paxlovid-drug-drug-interactions/.
81 Ramakrishnan S. Nicolau D.V. Jr. Langford B. Mahdi M. Jeffers H. Mwasuku C. Inhaled budesonide in the treatment of early COVID-19 (STOIC): a phase 2, open-label, randomised controlled trial Lancet Respir Med 9 7 2021 763 772 33844996
82 Clemency B.M. Varughese R. Gonzalez-Rojas Y. Morse C.G. Phipatanakul W. Koster D.J. Efficacy of Inhaled Ciclesonide for Outpatient Treatment of Adolescents and Adults With Symptomatic COVID-19: A Randomized Clinical Trial JAMA Intern Med 182 1 2022 42 49 34807241
83 Yu L.M. Bafadhel M. Dorward J. Hayward G. Saville B.R. Gbinigie O. Inhaled budesonide for COVID-19 in people at high risk of complications in the community in the UK (PRINCIPLE): a randomised, controlled, open-label, adaptive platform trial Lancet 398 10303 2021 843 855 34388395
84 Lenze E.J. Mattar C. Zorumski C.F. Stevens A. Schweiger J. Nicol G.E. Fluvoxamine vs Placebo and Clinical Deterioration in Outpatients With Symptomatic COVID-19: A Randomized Clinical Trial JAMA 324 22 2020 2292 2300 33180097
85 Tardif J.C. Bouabdallaoui N. L'Allier P.L. Gaudet D. Shah B. Pillinger M.H. Colchicine for community-treated patients with COVID-19 (COLCORONA): a phase 3, randomised, double-blinded, adaptive, placebo-controlled, multicentre trial Lancet Respir Med 9 8 2021 924 932 34051877
86 Siddiqi H.K. Mehra M.R. COVID-19 illness in native and immunosuppressed states: A clinical-therapeutic staging proposal J Heart Lung Transplant 39 5 2020 405 407 32362390
87 Beigel J.H. Tomashek K.M. Dodd L.E. Mehta A.K. Zingman B.S. Kalil A.C. Remdesivir for the Treatment of Covid-19-Final Report N Engl J Med 383 19 2020 1813 1826 32445440
88 Consortium WHOSTPan H. Peto R. Henao-Restrepo A.M. Preziosi M.P. Sathiyamoorthy V. Repurposed Antiviral Drugs for Covid-19-Interim WHO Solidarity Trial Results N Engl J Med 384 6 2021 497 511 33264556
89 Ader F. Bouscambert-Duchamp M. Hites M. Peiffer-Smadja N. Poissy J. Belhadi D. Remdesivir plus standard of care versus standard of care alone for the treatment of patients admitted to hospital with COVID-19 (DisCoVeRy): a phase 3, randomised, controlled, open-label trial Lancet Infect Dis 22 2 2022 209 221 34534511
90 Consortium WHOST Remdesivir and three other drugs for hospitalised patients with COVID-19: final results of the WHO Solidarity randomised trial and updated meta-analyses Lancet 399 10339 2022 1941 1953 35512728
91 Mozaffari E. Chandak A. Zhang Z. Liang S. Thrun M. Gottlieb R.L. Remdesivir Treatment in Hospitalized Patients With Coronavirus Disease 2019 (COVID-19): A Comparative Analysis of In-hospital All-cause Mortality in a Large Multicenter Observational Cohort Clin Infect Dis 75 1 2022 e450 e458 34596223
92 Rivera D.R. Peters S. Panagiotou O.A. Shah D.P. Kuderer N.M. Hsu C.Y. Utilization of COVID-19 Treatments and Clinical Outcomes among Patients with Cancer: A COVID-19 and Cancer Consortium (CCC19) Cohort Study Cancer Discov 10 10 2020 1514 1527 32699031
93 Wong C.K.H. Au I.C.H. Lau K.T.K. Lau E.H.Y. Cowling B.J. Leung G.M. Real-world effectiveness of early molnupiravir or nirmatrelvir-ritonavir in hospitalised patients with COVID-19 without supplemental oxygen requirement on admission during Hong Kong's omicron BA.2 wave: a retrospective cohort study Lancet Infect Dis 2022
94 Group RC Casirivimab and imdevimab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial Lancet 399 10325 2022 665 676 35151397
95 Group RC Convalescent plasma in patients admitted to hospital with COVID-19 (RECOVERY): a randomised controlled, open-label, platform trial Lancet 397 10289 2021 2049 2059 34000257
96 Simonovich V.A. Burgos Pratx L.D. Scibona P. Beruto M.V. Vallone M.G. Vazquez C. A Randomized Trial of Convalescent Plasma in Covid-19 Severe Pneumonia N Engl J Med 384 7 2021 619 629 33232588
97 Thompson M.A. Henderson J.P. Shah P.K. Rubinstein S.M. Joyner M.J. Choueiri T.K. Association of Convalescent Plasma Therapy With Survival in Patients With Hematologic Cancers and COVID-19 JAMA Oncol 2021
98 Writing Committee for the R-CAPIEstcourt L.J. Turgeon A.F. McQuilten Z.K. McVerry B.J. Al-Beidh F. Effect of Convalescent Plasma on Organ Support-Free Days in Critically Ill Patients With COVID-19: A Randomized Clinical Trial JAMA 326 17 2021 1690 1702 34606578
99 Group R.C. Horby P. Lim W.S. Emberson J.R. Mafham M. Bell J.L. Dexamethasone in Hospitalized Patients with Covid-19 N Engl J Med 384 8 2021 693 704 32678530
100 Kalil A.C. Patterson T.F. Mehta A.K. Tomashek K.M. Wolfe C.R. Ghazaryan V. Baricitinib plus Remdesivir for Hospitalized Adults with Covid-19 N Engl J Med 2020 10.1056/NEJMoa2031994
101 Guimaraes P.O. Quirk D. Furtado R.H. Maia L.N. Saraiva J.F. Antunes M.O. Tofacitinib in Patients Hospitalized with Covid-19 Pneumonia N Engl J Med 385 5 2021 406 415 34133856
102 Group RC Tocilizumab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial Lancet 397 10285 2021 1637 1645 33933206
103 Investigators R.-C. Gordon A.C. Mouncey P.R. Al-Beidh F. Rowan K.M. Nichol A.D. Interleukin-6 Receptor Antagonists in Critically Ill Patients with Covid-19 N Engl J Med 384 16 2021 1491 1502 33631065
104 Derde LPG. Effectiveness of Tocilizumab, Sarilumab, and Anakinra for critically ill patients with COVID-19 The REMAP-CAP COVID-19 Immune Modulation Therapy Domain Randomized Clinical Trial. medRxiv. 2021:2021.2006.2018.21259133.
105 Kyriazopoulou E. Poulakou G. Milionis H. Metallidis S. Adamis G. Tsiakos K. Early treatment of COVID-19 with anakinra guided by soluble urokinase plasminogen receptor plasma levels: a double-blind, randomized controlled phase 3 trial Nat Med 27 10 2021 1752 1760 34480127
106 group C-C Effect of anakinra versus usual care in adults in hospital with COVID-19 and mild-to-moderate pneumonia (CORIMUNO-ANA-1): a randomised controlled trial Lancet Respir Med 9 3 2021 295 304 33493450
107 Kalil A.C. Patterson T.F. Mehta A.K. Tomashek K.M. Wolfe C.R. Ghazaryan V. Baricitinib plus Remdesivir for Hospitalized Adults with Covid-19 N Engl J Med 384 9 2021 795 807 33306283
108 Marconi V.C. Ramanan A.V. de Bono S. Kartman C.E. Krishnan V. Liao R. 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 9 12 2021 1407 1418 34480861
109 Radujkovic A. Hippchen T. Tiwari-Heckler S. Dreher S. Boxberger M. Merle U. Vitamin D Deficiency and Outcome of COVID-19 Patients Nutrients 9 2020 12 33374515
110 Murai I.H. Fernandes A.L. Sales L.P. Pinto A.J. Goessler K.F. Duran C.S.C. Effect of a Single High Dose of Vitamin D3 on Hospital Length of Stay in Patients With Moderate to Severe COVID-19: A Randomized Clinical Trial JAMA 325 11 2021 1053 1060 33595634
111 Stroehlein J.K. Wallqvist J. Iannizzi C. Mikolajewska A. Metzendorf M.I. Benstoem C. Vitamin D supplementation for the treatment of COVID-19: a living systematic review Cochrane Database Syst Rev. 5 2021 CD015043
112 Lodigiani C. Iapichino G. Carenzo L. Cecconi M. Ferrazzi P. Sebastian T. Venous and arterial thromboembolic complications in COVID-19 patients admitted to an academic hospital in Milan Italy. Thromb Res. 191 2020 9 14 32353746
113 Nadkarni G.N. Lala A. Bagiella E. Chang H.L. Moreno P.R. Pujadas E. Anticoagulation, Bleeding, Mortality, and Pathology in Hospitalized Patients With COVID-19 J Am Coll Cardiol 76 16 2020 1815 1826 32860872
114 Lopes R.D. de Barros E.S.P.G.M. Furtado R.H.M. Macedo A.V.S. Bronhara B. Damiani L.P. Therapeutic versus prophylactic anticoagulation for patients admitted to hospital with COVID-19 and elevated D-dimer concentration (ACTION): an open-label, multicentre, randomised, controlled trial Lancet 397 10291 2021 2253 2263 34097856
115 Investigators I. Sadeghipour P. Talasaz A.H. Rashidi F. Sharif-Kashani B. Beigmohammadi M.T. Effect of Intermediate-Dose vs Standard-Dose Prophylactic Anticoagulation on Thrombotic Events, Extracorporeal Membrane Oxygenation Treatment, or Mortality Among Patients With COVID-19 Admitted to the Intensive Care Unit: The INSPIRATION Randomized Clinical Trial JAMA 325 16 2021 1620 1630 33734299
116 Sholzberg M. Tang G.H. Rahhal H. AlHamzah M. Kreuziger L.B. Ainle F.N. Effectiveness of therapeutic heparin versus prophylactic heparin on death, mechanical ventilation, or intensive care unit admission in moderately ill patients with covid-19 admitted to hospital: RAPID randomised clinical trial BMJ 375 2021 n2400
117 Investigators A. AC-a Investigators Investigators R.-C. Lawler P.R. Goligher E.C. Berger J.S. Therapeutic Anticoagulation with Heparin in Noncritically Ill Patients with Covid-19 N Engl J Med 385 9 2021 790 802 34351721
118 Perepu U.S. Chambers I. Wahab A. Ten Eyck P. Wu C. Dayal S. Standard prophylactic versus intermediate dose enoxaparin in adults with severe COVID-19: A multi-center, open-label, randomized controlled trial J Thromb Haemost 19 9 2021 2225 2234 34236768
119 Investigators R.-C. AC-a Investigators Investigators A. Goligher E.C. Bradbury C.A. McVerry B.J. Therapeutic Anticoagulation with Heparin in Critically Ill Patients with Covid-19 N Engl J Med 385 9 2021 777 789 34351722
120 Ramacciotti E. Barile Agati L. Calderaro D. Aguiar V.C.R. Spyropoulos A.C. de Oliveira C.C.C. Rivaroxaban versus no anticoagulation for post-discharge thromboprophylaxis after hospitalisation for COVID-19 (MICHELLE): an open-label, multicentre, randomised, controlled trial Lancet 399 10319 2022 50 59 34921756
121 Group RC Aspirin in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial Lancet 399 10320 2022 143 151 34800427
122 Morjaria S. Zhang A.W. Kaltsas A. Parameswaran R. Patel D. Zhou W. The Effect of Neutropenia and Filgrastim (G-CSF) in Cancer Patients With COVID-19 Infection medRxiv 2020 2020 2008.2013.20174565
123 Soriano J.B. Murthy S. Marshall J.C. Relan P. Diaz J.V. Condition WHOCCDWGoP-C-. A clinical case definition of post-COVID-19 condition by a Delphi consensus Lancet Infect Dis 22 4 2022 e102 e107 34951953
124 Pinato D.J. Tabernero J. Bower M. Scotti L. Patel M. Colomba E. Prevalence and impact of COVID-19 sequelae on treatment and survival of patients with cancer who recovered from SARS-CoV-2 infection: evidence from the OnCovid retrospective, multicentre registry study Lancet Oncol 22 12 2021 1669 1680 34741822
125 Chai C. Feng X. Lu M. Li S. Chen K. Wang H. One-year mortality and consequences of COVID-19 in cancer patients: A cohort study IUBMB Life 73 10 2021 1244 1256 34318585
126 Arcani R. Colle J. Cauchois R. Koubi M. Jarrot P.A. Jean R. Clinical characteristics and outcomes of patients with haematologic malignancies and COVID-19 suggest that prolonged SARS-CoV-2 carriage is an important issue Ann Hematol 100 11 2021 2799 2803 34518918
127 Lee C.Y. Shah M.K. Hoyos D. Solovyov A. Douglas M. Taur Y. Prolonged SARS-CoV-2 Infection in Patients with Lymphoid Malignancies Cancer Discov 12 1 2022 62 73 34753749
128 Jacobs J.J.L. Persistent SARS-2 infections contribute to long COVID-19 Med Hypotheses 149 2021 110538 33621843
129 Boglione L. Meli G. Poletti F. Rostagno R. Moglia R. Cantone M. Risk factors and incidence of long-COVID syndrome in hospitalized patients: does remdesivir have a protective effect? QJM 114 12 2022 865 871 34850210
130 Barros C. Freire R.S. Frota E. Rezende Santos A.G. Farias M.E.L. Rodrigues M.G.A. Short-Course of Methylprednisolone Improves Respiratory Functional Parameters After 120 Days in Hospitalized COVID-19 Patients (Metcovid Trial): A Randomized Clinical Trial Front Med (Lausanne) 8 2021 758405 34917633
131 Myall K.J. Mukherjee B. Castanheira A.M. Lam J.L. Benedetti G. Mak S.M. Persistent Post-COVID-19 Interstitial Lung Disease. An Observational Study of Corticosteroid Treatment Ann Am Thorac Soc. 18 5 2021 799 806 33433263
132 Catalan I.P. Marti C.R. Sota D.P. Alvarez A.C. Gimeno M.J.E. Juana S.F. Corticosteroids for COVID-19 symptoms and quality of life at 1 year from admission J Med Virol 94 1 2022 205 210 34436783
133 Azzolini E. Levi R. Sarti R. Pozzi C. Mollura M. Mantovani A. Association Between BNT162b2 Vaccination and Long COVID After Infections Not Requiring Hospitalization in Health Care Workers JAMA 328 7 2022 676 678 35796131
134 Vollbracht C. Kraft K. Feasibility of Vitamin C in the Treatment of Post Viral Fatigue with Focus on Long COVID, Based on a Systematic Review of IV Vitamin C on Fatigue Nutrients 4 2021 13 35010888
135 Rossato M.S. Brilli E. Ferri N. Giordano G. Tarantino G. Observational study on the benefit of a nutritional supplement, supporting immune function and energy metabolism, on chronic fatigue associated with the SARS-CoV-2 post-infection progress Clin Nutr ESPEN 46 2021 510 518 34857243
136 Vaziri-Harami R. Delkash P. Can l-carnitine reduce post-COVID-19 fatigue? Ann Med Surg (Lond) 73 2022 103145 34925826
137 Vaira L.A. Hopkins C. Petrocelli M. Lechien J.R. Cutrupi S. Salzano G. Efficacy of corticosteroid therapy in the treatment of long- lasting olfactory disorders in COVID-19 patients Rhinology 59 1 2021 21 25 33290446
138 Robbins T. Gonevski M. Clark C. Baitule S. Sharma K. Magar A. Hyperbaric oxygen therapy for the treatment of long COVID: early evaluation of a highly promising intervention Clin Med (Lond) 21 6 2021 e629 e632 34862223
139 Albu S. Rivas Zozaya N. Murillo N. Garcia-Molina A. Figueroa Chacon C.A. Kumru H. Multidisciplinary outpatient rehabilitation of physical and neurological sequelae and persistent symptoms of covid-19: a prospective, observational cohort study Disabil Rehabil 2021 1 8
140 Garcia-Molina A. Espina-Bou M. Rodriguez-Rajo P. Sanchez-Carrion R. Ensenat-Cantallops A. Neuropsychological rehabilitation program for patients with post-COVID-19 syndrome: a clinical experience Neurologia 36 7 2021 565 566
141 Embi P.J. Levy M.E. Naleway A.L. Patel P. Gaglani M. Natarajan K. Effectiveness of 2-Dose Vaccination with mRNA COVID-19 Vaccines Against COVID-19-Associated Hospitalizations Among Immunocompromised Adults - Nine States MMWR Morb Mortal Wkly Rep 70 44 2021 1553 1559 January-September 2021 34735426
142 Heudel P. Favier B. Assaad S. Zrounba P. Blay J.Y. Reduced SARS-CoV-2 infection and death after two doses of COVID-19 vaccines in a series of 1503 cancer patients Ann Oncol 32 11 2021 1443 1444 34333128
143 Shapiro L.C. Thakkar A. Campbell S.T. Forest S.K. Pradhan K. Gonzalez-Lugo J.D. Efficacy of booster doses in augmenting waning immune responses to COVID-19 vaccine in patients with cancer Cancer Cell 2021
144 Rottenberg Y. Grinshpun A. Ben-Dov I.Z. Oiknine Djian E, Wolf DG, Kadouri L. Assessment of Response to a Third Dose of the SARS-CoV-2 BNT162b2 mRNA Vaccine in Patients With Solid Tumors Undergoing Active Treatment JAMA Oncol 2021
145 Shroff R.T. Chalasani P. Wei R. Pennington D. Quirk G. Schoenle M.V. Immune responses to two and three doses of the BNT162b2 mRNA vaccine in adults with solid tumors Nat Med 27 11 2021 2002 2011 34594036
146 Peeters M. Verbruggen L. Teuwen L. Vanhoutte G. Vande Kerckhove S. Peeters B. Reduced humoral immune response after BNT162b2 coronavirus disease 2019 messenger RNA vaccination in cancer patients under antineoplastic treatment ESMO Open 6 5 2021 100274 34597941
147 Thakkar A. Gonzalez-Lugo J.D. Goradia N. Gali R. Shapiro L.C. Pradhan K. Seroconversion rates following COVID-19 vaccination among patients with cancer Cancer Cell. 39 8 2021 1081 1090 e1082 34133951
148 Fendler A. Au L. Shepherd S. Byrne F. Cerrone M. Boos L. Functional antibody and T-cell immunity following SARS-CoV-2 infection, including by variants of concern, in patients with cancer: the CAPTURE study. Res Sq 2021
149 Stampfer S.D. Goldwater M.S. Jew S. Bujarski S. Regidor B. Daniely D. Response to mRNA vaccination for COVID-19 among patients with multiple myeloma Leukemia 35 12 2021 3534 3541 34326466
150 Group AC-TfIwC-S Efficacy and safety of two neutralising monoclonal antibody therapies, sotrovimab and BRII-196 plus BRII-198, for adults hospitalised with COVID-19 (TICO): a randomised controlled trial Lancet Infect Dis 22 5 2022 622 635 34953520
151 Spyropoulos A.C. Goldin M. Giannis D. Diab W. Wang J. Khanijo S. Efficacy and Safety of Therapeutic-Dose Heparin vs Standard Prophylactic or Intermediate-Dose Heparins for Thromboprophylaxis in High-risk Hospitalized Patients With COVID-19: The HEP-COVID Randomized Clinical Trial JAMA Intern Med 181 12 2021 1612 1620 34617959
| 0 | PMC9737523 | NO-CC CODE | 2022-12-14 23:29:54 | no | Eur J Cancer. 2022 Dec 10; doi: 10.1016/j.ejca.2022.11.030 | utf-8 | Eur J Cancer | 2,022 | 10.1016/j.ejca.2022.11.030 | oa_other |
==== Front
Journal of Building Engineering
2352-7102
2352-7102
Elsevier Ltd.
S2352-7102(22)01712-0
10.1016/j.jobe.2022.105706
105706
Article
Influences of obstacle factors on the transmission trends of respiratory infectious diseases in indoor public places
Cui Ziwei a
Cai Ming a
Xiao Yao a∗
Zhu Zheng b
Chen Gongbo c
a School of Intelligent System Engineering, Shenzhen Campus of Sun Yat-Sen University, Shenzhen, Guangdong, China
b College of Civil Engineering and Architecture, Zhejiang University, Hangzhou, Zhejiang, China
c Guangdong Provincial Engineering Technology Research Center of Environmental and Health Risk Assessment, Department of Occupational and Environmental Health, School of Public Health, Sun Yat-sen University, Guangzhou, Guangdong, China
∗ Corresponding author.
10 12 2022
1 4 2023
10 12 2022
64 105706105706
20 9 2022
9 11 2022
6 12 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Public facilities are important transmission places for respiratory infectious diseases (e.g., COVID-19), due to the frequent crowd interactions inside. Usually, changes of obstacle factors can affect the movements of human crowds and result in different epidemic transmissions among individuals. However, most related studies only focus on the specific scenarios, but the common rules are usually ignored for the impacts of obstacles' spatial elements on epidemic transmission. To tackle these problems, this study aims to evaluate the impacts of three spatial factors of obstacles (i.e., size, quantity, and placement) on infection spreading trends in two-dimension, which can provide scientific and concise spatial design guidelines for indoor public places. Firstly, we used the obstacle area proportion as the indicator of the size factor, gave the mathematical expression of the quantity factor, and proposed the walkable-space distribution indicator to represent the placement factor by introducing the Space Syntax. Secondly, two spreading epidemic indicators (i.e., daily new cases and people's average exposure risk) were estimated based on the fundamental model named exposure risk with the virion-laden particles, which accurately forecasted the disease spreading between individuals. Thirdly, 120 indoor scenarios were built and simulated, based on which the value of independent and dependent variables can be measured. Besides, structural equation modeling was employed to examine the effects of obstacle factors on epidemic transmissions. Finally, three obstacle-related guidelines were provided for policymakers to mitigate the disease spreading: minimizing the size of obstacles, dividing the obstacle into more sub-ones, and placing obstacles evenly distributed in space.
Keywords
Indoor environment
Spatial design
Structural equation modeling
Pedestrian-based epidemic spreading model
Space syntax
==== Body
pmc1 Introduction
The unexpected Corona Virus Disease 2019 (COVID-19) epidemic has been a global concern in the past three years and has significantly impacted all aspects of the world [[1], [2], [3]]. The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes this disease, and COVID-19 is a respiratory infectious disease (RID) with a high transmission rate [4,5]. RIDs including COVID-19, SARS, and MERS are harmful to public health [6], which always bring sequelae such as fatigue and dyspnea, and even cause the loss of life expectancy [7,8]. To reduce the potential risk of ongoing or re-emerging RIDs, the safety of individual staying spaces is of paramount concern [9,10]. Considering the fact that most individual contacts take place in indoor public places, recognizing the effective spatial setups in indoor public spaces with human crowds is essential for the prevention and control of RIDs [[11], [12], [13], [14]].
There are some studies exploring the epidemic spreading in different scenarios with a moving crowd [[15], [16], [17], [18], [19], [20], [21], [22], [23], [24]]. For example, Garcia et al. collected detailed field data in various scenarios, and built a model to rank scenarios based on infection risks. Besides, they found that changing the geometry of queues would affect the epidemic transmission risks [15]. Mokhtari and Jahangir used a university building as the case, and constructed a multi-objective optimization problem to find the optimum occupant distribution patterns [19]. Moritz et al. conducted an experimental indoor mass gathering event under three different hygiene practices (i.e., no restrictions, moderate restrictions, and strong restrictions), and measured the contacts of each person during the event with contact tracing devices. Then, based on the data, they simulated the exposure of individuals and evaluated the contributions of several preventive and control measures [20]. The study by Ku et al. focused on the resulting burden of COVID-19 on public transport. As a lot of public transportation passengers have regular and repeated movements, stranger people with similar mobility patterns form a familiar stranger group. The researchers used smart card data and real data on infected individuals to estimate the familiar stranger group, and then simulated the epidemic spreading in the group. Furthermore, they predicted the transmission probability under mandatory mask-wearing and social distancing [16]. Vuorinen et al. constructed two Monte-Carlo models which targeted a generic public place without obstacles and a supermarket with fixed shelves, respectively. Besides, they investigated the physical processes related to aerosol and droplet dispersion in the context of COVID-19, and coupled the movements of virus-laden particles and the motion of individuals to conduct Monte-Carlo modeling [24]. However, the spatial settings of obstacles are specified and fixed in these studies.
Recently, the influences of obstacles on epidemic transmission have been explored from different perspectives. Under the constraints of non-pharmaceutical interventions such as maintaining safe social distance, the layout design of obstacles has been optimized to better prevent and control infection. Contardo and Costa aimed to seek a layout that maximized the number of people that the restaurant can accommodate under satisfying social distancing measures. Moreover, they analyzed whether adding space separators and considering the sitting sense of customers can increase the room capacity separately [25]. Similarly, an equilateral triangle-based seat pattern was provided in symmetrical spaces to achieve maximum seating capacity under the physical distancing safety guidelines [26]. Moreover, Mekawy and Gabr presented a multi-objective optimization approach to mitigate the risks of infectious diseases’ transmission in open-plan offices. In detail, they maximized the number of workers, window proximity score, “buzz score”, and minimized the “adjacency score” to respond to relative prevention and control measures [27]. Shorfuzzaman, Hossain, and Alhamid proposed a system with deep learning-based object detection models to track individuals in real-time, and further monitored and warned people who had broken the social distancing constraints [28]. Differently, the impacts of the number of equal-size obstacles on disease spreading were studied by Azmi, Hidayat, and Pramata [29]. To evaluate the best setup for the isolation room and the sanitizing machine for spreading disinfectant aerosol, they simulated the airflow in three rooms with different amounts of equal-size beds, respectively. Although these studies have provided useful findings for the spatial design in response to RIDs, they ignore details brought from the individual movements. Therefore, some researchers have coupled individual movement and spatial configuration management to study. Based on the space occupation and density of the moving crowd, Braidotti et al. used pedestrian simulations to clarify the spatial problems for infection prevention in the ship environment. Then, an alternative layout was proposed to solve the above critical issues for the ship, but it did not apply to other scenarios [13]. Xiao et al. coupled the pedestrian dynamics and the modified susceptible–exposed–infectious model to decipher the spreading process of COVID-19, and they analyzed the resulting epidemic transmission in different scenarios (i.e., three closed rooms with different exits, three corridors with different settings of impassable railings, and three winding-queue configurations). However, their fundamental model was not tested with real-world data [30].
In sum, most previous studies focus on the placement of obstacles and the resulting epidemic transmission, but the physical mechanisms behind them are usually not clarified. Moreover, the impact of other obstacle factors (e.g., size and quantity) needs further evaluation. On the other hand, existing studies only pay attention to scenarios with fixed functions, such as the office, hospital isolation room, and restaurant. However, various scene functions are more suitable for analysis in combination with personal behaviors (such as dwell time in the scene) rather than obstacle settings. For example, if there is a large classroom and a small dining room with the same obstacles, students will stay in a position for a long time to have classes after entering the classroom; and people buy food first and then stay somewhere for a short time to eat in the restaurant. These two scenarios have different functions, and they affect the epidemic spreading to varying degrees through people's activities rather than the obstacles setups. Therefore, in this paper, to get a general conclusion about obstacles, we simplified the study to temporarily ignore the functions of various scenes. Based on this, scientific and concise obstacle adjustment advice can be concluded and further applied in different types of scenarios, which may contribute to the prevention and control of RIDs in the public health system.
This study aims to estimate the influences of the size, quantity, and placement of obstacles on the transmission trends in indoor public places, and clarify the mechanisms behind them. In reality, the spread of infection in indoor public places is rarely measured, and it is difficult to set obstacles flexibly in different scenarios to make them comparable. As a result, there is a lack of real-world data to achieve our goals. To tackle this issue, we attempt to obtain the basic data by simulating the epidemic transmission between individuals in indoor rooms with various obstacles. Therefore, we need a validated simulation fundamental model which can forecast the spreading trends of RIDs between moving individuals. In addition, as the indicators of obstacle factors are the independent variables in the statistical analysis, it is necessary to find quantifiable indicators as dependent variables to represent the transmission trends.
The framework of our study is presented in Fig. 1 , and it should be pointed out that our study is conducted in two-dimension (2D). The rest of the paper is organized as follows. The methodology is presented in section 2. The simulation setups are illustrated in section 3, and the results are then reported in section 4. The in-depth discussions and future perspectives are reported in section 5. Finally, conclusions are provided in section 6.Fig. 1 There is the framework of our study.
Fig. 1
2 Methodology
In this section, three indicators of obstacle factors are first defined, and then the adopted fundamental model is introduced, based on which two transmission trend indicators are determined.
2.1 Obstacle factors
In a scenario, an obstacle is defined as a non-walkable object that does not intersect or tangent with others. The NObs refers to the quantity of obstacles in an indoor public place, and obstacles can be numbered as {1,…,n,…,NObs}. The SnObs is the size (2D area) of the n-th obstacle, and we have the total size of all obstacles as(1) SAllObs=∑n=1NObsSnObs.
Here, the dimensionless proportion RAllObs is calculated as(2) RAllobs=SAllobsSTotal×100%,
where STotal is the size of the simulation room. Hence, the first and second obstacle indicators are determined mathematically, which are respectively the proportion of the obstacle size to the simulation space size RAllObs and the quantity of obstacles NObs. It should be noted that, if there are multiple obstacles in one scenario, their different sizes and shapes may affect the epidemic transmission, but these are not the keys of this study. Thus, a simple case is considered in this paper: all obstacles in a scenario have the same size and shape.
For the placement of obstacles, previous studies have studied the locations of obstacles [26,27,29]. In fact, individuals are moving in walkable spaces, which vary with the layout of obstacles. It is more direct to analyze the variation of walkable spaces to find the effects and mechanism of the obstacle placement factor. To quantify the impact of obstacle placement and the resulting walkable area, we should describe the spaces quantitatively. Therefore, Space Syntax, which has successfully analyzed spatial relationships, is adopted here to express the space [[31], [32], [33]]. In Space Syntax, spaces can be defined in three ways: Convex Partitions, Axial Lines, and Isovists. As the indoor public place is investigated in this study, the first way is used here as it is mainly used for small buildings and their interior space. Hence, continuous walkable space is divided into a series of convex spaces, and a convex space is defined as “one space where all points within this space are visible to one another” by Bill Hillier and his colleagues [33], see Fig. 2 .Fig. 2 There are examples of the (a) convex and (b) non-convex spaces.
Fig. 2
Moreover, Bill Hillier and his colleagues have suggested an algorithm for manually constructing a convex map: “Simply find the largest convex space and draw it in, then the next largest, and so on until all the space is accounted for.“. However, there may be multiple convex spaces of the same size. Hence, for a given scenario, the algorithm cannot be used to represent a unique walkable convex map. For example, a scenario in Fig. 3 (a) is a 22 m × 22 m room with a 6 m× 6 m obstacle at the geometric center. Each blue space in Fig. 3(b) can serve as the first walkable convex space, and their corresponding convex maps are different.Fig. 3 In (a) a scenario, (b) there are four spaces that can be used as the first workable convex space in Hillier's method, and (c) a unique one can be determined as the first in our method. Then, a unique walkable convex map is obtained in (d) based on our method.
Fig. 3
To tackle this issue, we first construct a rectangular coordinate system and then use greedy strategies to get the unique convex map for a scenario. The coordinate system has “x” and “y” axes, and its origin is the geometrical center of the simulation room. Moreover, the orientation of the “x” axis is the same as that of the pedestrian outflow. Then, the coordinate of the geometrical center in each convex walkable space is measured. Therefore, for a given scenario, we get started by finding the first convex space until all walkable areas have been added to the convex map. For each time, a unique convex space can be found and added to the map based on the following three steps.➢ Step 1: Find the largest convex walkable space. If only one convex space with the largest size can be found, we draw it in the convex map; Otherwise, we transfer the spaces with the largest size to step 2.
➢ Step 2: Find the convex space with the maximum x-axis value among spaces from Step 1. If only one convex space with the largest size and the maximum x-axis value can be found, we draw it in the convex map; Otherwise, we transfer the spaces with the largest size and the maximum x-axis value to step 3.
➢ Step 3: Find the convex space with the maximum y-axis value among spaces from Step 2. As the geometric center coordinates of two convex spaces cannot be the same, the unique convex element can be identified and added to the convex map, as shown in Fig. 3(c).
Consequently, a unique convex map can be obtained for a constructed scenario (see Fig. 3(d)), and each component in the map is fixed. When there are M walkable convex spaces in a convex map, their determined sizes are respectively {S1Con,…,SmCon,…,SMCon}. Here, we propose the walkable-space distribution indicator DCon to measure the variation of walkable spaces modified by the placement of obstacles. As the standard deviation is an index that represents the dispersion degree of the dataset, it can help measure the distribution of walkable space sizes in a convex map. Hence, the indicator DCon is defined as the standard deviation of walkable convex spaces in a unique map and given as(3) DCon=∑m=1M(SmCon−SCon‾)2M,whereSCon‾=∑m=1MSmConM
For example, after moving the obstacle in Fig. 3(a) down 5 m, the resulting scenario is shown in Fig. 4 (a), and the corresponding unique walkable convex map is obtained in Fig. 4(b) based on our method. Apparently, the obstacle placements in these two scenarios are different, which leads to various walkable convex maps. Although there are 4 walkable spaces in each convex map, their sizes are different. Specifically, the size of the 1st walkable convex space in Fig. 4(b) is larger than that in Fig. 3(d), while the size of the 4th walkable convex space in Fig. 4(b) is smaller than that in Fig. 3(d). Hence, the distribution of walkable space sizes in Fig. 4(b) is larger than that in Fig. 3(d), which results in the larger value of the walkable-space distribution indicator DCon in Fig. 4(b).Fig. 4 In (a) a scenario, the unique walkable convex map is obtained in (b) based on our method.
Fig. 4
2.2 Fundamental model and transmission trends indicators
In forecasting the disease spreading between individuals, previous studies have integrated pedestrian dynamics into the epidemic spreading models [34]. These pedestrian-based epidemic models describe the disease transmission process with time-varying personal physical distances during individual movements. There are several alternative pedestrian-based epidemics fundamental models, such as the exposure risk with virion-laden particles (ERP) model [10], the EXPOSED model [35], the fixed exposure-risk unit (FERU) model [36], and the exposure risk with quality (ERQ) model [37]. As the ERP model has been verified with real-world data and exhibits superior prediction performance than FERU and ERQ models, it is adopted here as the fundamental model.
The ERP model [10] focuses on the typical symptom of most RIDs, i.e., coughing. The model has four components, and they are introduced below. Note that in the ERP model, when the susceptible individual and cough-generated particles meet in the same place, the instantaneous exposure risk is defined as the possible maximal mass of particles suffered, and the infectors’ exposure risks are 0.
2.2.1 Individual movement simulation
Based on the model inputs (i.e., number of individuals Ctotal, number of infectors among individuals Cinf, and mean dwell time of individuals Tdwell) and the pedestrian dynamic model (i.e., the social force model), individuals’ movements are reproduced in a general simulation place. Herein, each individual is represented as a circle, and the dynamic modelings of infectors and susceptible people are the same. The widely used social force model is conducted to simulate pedestrian dynamics, where the individual is driven by the resultant force from the goal, neighbor, and obstacles, as shown in Fig. 5 . Besides, the social force model considers the realistic obstacle avoidance behaviors, and it can reasonably reproduce the people movements in the room with different obstacles.Fig. 5 Diagram of the social force model.
Fig. 5
The velocity vi of individual i (with a mass of mi) at time t is estimated as(4) dvidt=Fidrv+∑inearFi,inearped+∑wFi,wobsmi
where Fidrv represents the force attracted by the goal, which indicates the desire of individual i to keep towards the goal with a certain walking velocity; Fi,inearped means the interaction force between individual i and the neighboring individual inear, which is generally a kind of repulsive force and consists of two parts: social force at the psychological level and contact force at the physical level; Fi,wobs reflects the instant interaction force between the objective individual i and the obstacle (including the wall) w. There are more details about the social force model in previous literatures [36,38,39]. During the dwell time Tdwell, people would randomly adjust the goal, and the new goals are required to be away from obstacles. Once the staying time reaches the specified time Tdwell, the individual goal changes to the room exit. Therefore, the time series positions of each individual can be obtained.
2.2.2 Typical cough simulation
The transmission process of virion-laden particles, including droplets (the diameter > 100 μm [[40], [41], [42], [43]]) and aerosols (the diameter ≤ 100 μm [[40], [41], [42], [43]]), from a typical cough is simulated in a closed environment without other air movements such as ventilation. Besides, we consider the breathing height area of susceptible people to set the cough simulation domain, as shown in Appendix A. According to previous studies [10,12,44], as the particle volume fraction in the cough flow is low, the Eulerian-Lagrangian method is selected: the continuous phase (e.g., the flow field) is modeled by the Eulerian method; the discrete phase (e.g., cough particles) is modeled by the Lagrangian method; and the discrete phase exchanges mass, energy, and momentum with the continuous phase. Here, the renormalization group (RNG) k-ε model is used as the turbulence model, and the discrete phase model (DPM) is adopted for the particle diffusion [45]. In the discrete phase modeling, the velocity uc of particle c at time t is estimated as(5) ducdt=FD(u−uc)+Fg
where u is the fluid phase velocity, FD(u−uc) is the Stokes drag force, and Fg is the gravitational force [46].
Thus, the time series positions and masses of each cough particle can be estimated. The simulation is conducted with the commercial Computational Fluid Dynamics (CFD) solver ANSYS Fluent release 2020 R2, and more details are described in Appendix A. In the cough numerical simulation, the positions of particles at two different times are shown in Fig. 6 . Results indicate that droplets fall down quickly and aerosols stay for a long time in the air, which is consistent with existing researches [47,48].Fig. 6 (a) 0.40 s and (b) 2.00 s after the cough, particles are integrated on the y-z plane in the computational domain.
Fig. 6
2.2.3 Individual exposure risk estimation
Outputs from the first two modules are coupled to estimate the Ei,j,g, i.e., the instantaneous exposure riskof individual i exposed to infector j’s g-th cough. Specifically, we use di,j,g(t) to represent the distance between individual i and the location where starting infector j’ s g-th cough at time t, tj,gstart to reflect the time when starting infector j’s g-th cough, and Tinf to show the infectious time of a typical cough. Besides, to help describe the change of cough particles with the time-varying distance, we construct several representative planes with the length of s(=0.1,0.3,…) meters along the “y” axis and more details can be found in Ref. [10]. Hence, when di,j,g(t)∈[s−0.1,s+0.1) and t≤tj,gstart+Tinf, we have(6) Ei,j,g(t)=as*exp(−(tj,ginterval−bscs)2),
where tj,ginterval is the time interval between time t and tj,gstart; as, bs and cs are parameters of the fitted Gaussian distribution function for the s-th representative plane.
Based on these, the exposure risk of individual i during the dwell time is measured as(7) Ei=∑t=tientertienter+Tidwell∑j=1J(t)∑g=1JG(j,t)Ei,j,g(t)
where tienter is the place enter time, Tidwell is the dwell time of individual i, J(t) denotes the number of infectors in the simulation scenario at time t, and JG(j,t) indicates the number of the infector j’ s infectious coughs at time t.
2.2.4 Prediction of transmission trends
Since the number of susceptible individuals Csus can be determined according to model inputs (= Ctotal−Cinf), the number of high-risk exposed people Crisk during the simulation can be estimated as(8) Crisk(α)=∑i=1Csusψ(Ei,α),whereψ(Ei,α)={1,ifEi>α0,otherwise
Here, the number of new cases CNew is assumed to be a linear equation of Crisk, and an extreme case is considered (i.e., CNew=0 if Crisk=0) [10]. Therefore, CNew is represented by(9) CNew=β*Crisk(α).
It should be noted that the cut-line of high exposure risk α and the coefficient β in Equation (9) should be estimated based on the real-world historical data first, and then be applied to predict the daily new cases CNew in the future.
Based on the ERP model, for a scenario, we can predict the exposure risk of each individual Ei and the number of new cases CNew. As CNew is a direct and useful indicator to describe the disease spreading trends, it is adopted as a transmission indicator. Another indicator is people's average exposure risk EAve, which represents the general level of all visits' exposure risks in a scenario and is defined as(10) EAve=∑iCsusEiCsus,
where Csus and Ei are the same as mentioned before.
3 Simulation setups
3.1 Scenario setups
A scenario consists of people and spaces. In this section, we first introduce the human simulation setups, and then determine the space simulation setups.
Both infectors and susceptible individuals are represented by a circle with a radius of 0.2 m for simplification (see Fig. 7 ) [10]. In the beginning, no individual is in the simulation space. Then, individuals enter the indoor room through the entrance in sequence with an average interval of 5 s, and leave through the room exit. In the sequence of people entering the room, the number of susceptible individuals between any two adjacent infectors is set the same, which brings an approximate same number of infectors in the room at any time during the simulation period. In the dwell time, individuals follow the random walking pattern with the desired velocity of 1.34 m/s [36]. Besides, after entering the room, the infected individual averagely coughs every 15 s [36], and the infectious time follows a uniform distribution from 0 to 15 s [10,36]. A cough's infectious distance is 1.70 m, and the parameters of the Gaussian distribution function in Equation (6) are shown in Table 1 , which are consistent with Ref. [10].Fig. 7 There is a sketch map of the simulation people and room in the case.
Fig. 7
Table 1 Parameters of the Gaussian distribution function in Equation (6).
Table 1s Range of di,j,g(t) as bs cs
0.1 [0.0, 0.2) 3.793 ×10−6 0.100 0.030
0.3 [0.2, 0.4) 3.337 ×10−6 0.140 0.034
0.5 [0.4, 0.6) 2.075 ×10−6 0.180 0.096
0.7 [0.6, 0.8) 8.738 ×10−7 0.220 0.284
0.9 [0.8, 1.0) 6.720 ×10−7 0.220 0.360
1.1 [1.0, 1.2) 4.455 ×10−7 0.300 0.410
1.3 [1.2, 1.4) 1.632 ×10−7 0.340 0.763
1.5 [1.4, 1.6) 5.821 ×10−8 0.340 0.882
1.7 [1.6, 1.7] 5.475 ×10−10 0.380 1.971
We change the obstacles in the same simulation space to explore the factors influencing the transmission of RIDs. Herein, a fixed 22 m × 22 m indoor room is constructed as the simulation space, i.e., STotal=484m2, and there is an entrance on one side of the room and an exit on the opposite side (see Fig. 7). There are two steps to determine the size, quantity, and placement of obstacles in each simulation scenario:➢ Step 1: Determine the size and quantity of obstacles. The obstacle size indicator SAllObs is formulated first. To clarify the impacts of the quantity of obstacles, an effective tool to separate the obstacle is needed. Here, a horizontal-vertical division rule is proposed for dividing square or rectangle obstacles, and it has four parameters: Hcut, Vcut, Ecut, and Dcut. Specifically, Hcut and Vcut are the number of divisions in the horizontal and vertical directions, respectively. Ecut is a binary variable, where Ecut=1 indicates the obstacle is divided equally and Ecut=0 represents unequal division. Dcut is the shortest distance between adjacent obstacles, whose value can be fixed or changeable in the division process.
➢ Step 2: Determine the placement of obstacles. After determining the size, quantity, and relative positions of obstacles, a minimum rectangle that can cover all obstacles can be found. Then, the geometrical center GAllObs of the rectangle is uniquely determined when the placement of obstacles in the room is fixed. Thus, when we build a coordinate system with “x” and “y” axes, whose origin is the geometrical center of the simulation space GRoom, the obstacle positions can be confirmed based on the accurate coordinate of GAllObs.
According to the above steps, we set 5 groups with different total obstacle sizes (i.e., SAllObs= 36 m2, 64 m2, 100 m2, 144 m2, 196 m2, respectively), and the obstacles are all square before segmentation. Besides, we set Hcut∈[0,1,2,3] and Vcut∈[0,1,2] in each group with the same SAllObs, and explore the simple situation when Ecut=1 and Dcut=2.0 m. For example, there are 12 patterns in the group with SAllObs=36m2 (see Fig. 8 ), and six same obstacles are obtained in the pattern with Hcut=3 and Vcut=2. Hence, independent variables of NObs and RAllObs in each scenario can be measured. Moreover, we set two coordinates of GAllObs for each pattern to explore the influence of obstacle placement, and the first coordinate is (0,0). To ensure that all obstacles are in the simulation room, for each group with SAllObs= 36 m2, 64 m2, 100 m2, 144 m2, 196 m2, the second coordinate is set as (0,−5), (0,−4), (0,−3), (0,−2), and (0,−1), respectively. For example, for the pattern Hcut=3 and Vcut=2 when SAllObs= 36 m2, two obstacle placements in the simulation space are shown in Fig. 9 . Based on these, the independent variable DCon in each scenario can be estimated.Fig. 8 The division results are based on the horizontal-vertical division rule when Snon=36m2.
Fig. 8
Fig. 9 When Snon=36m2, Hcut=3, and Vcut=2, there are two layouts with different positions of obstacle.
Fig. 9
In sum, there are 5 groups with different obstacle sizes, 12 patterns in each group, and 2 coordinates in each pattern to estimate the placement. Thus, there are 5 ×12×2=120 samples collected in this case. Each scenario is simulated at least three times to ensure that the standard deviation of their resulting new cases is smaller than CNewStd, and the average experimental results are taken as the final values. In this case, CNewStd is set as 600 through numerous tests. Each scenario is denoted by a sequence of intervention codes in the “SAllObs-[Hcut, Vcut]-GAllObs” format. For instance, Scenario #36- [1,1]-(0,0) is the scenario whose total size of obstacles SAllObs=36 m2, the pattern in the horizontal-vertical division rule is Hcut=Vcut=1, and the position of obstacle geometrical center GAllObs=(0,0); Scenario #36- [2,3]-(0,-5) represents the scenario with SAllObs=36 m2, Hcut=3, Vcut=2, and GAllObs=(0,−5). Fig. 10 shows the configuration of all 24 scenarios when SAllObs is 36 m2, and scenarios with other sizes are illustrated in Appendix B.Fig. 10 There are 24 scenarios when SAllObs is 36 m2.
Fig. 10
3.2 Model setups
As introduced in section 2.2, there are three main individual inputs in the ERP model: number of individuals Ctotal, number of infectors among individuals Cinf, and mean dwell time Tdwell. Since the impacts of obstacle factors are explored in this study, the fixed values of model inputs are enough. Thus, we use data from the United States (U.S.) during the spreading of COVID-19 on June 5th, 2020, and there are Ctotal= 257,177,921 [49,50], Cinf = 1,759,672 [51], and Tdwell= 25 min [10,36]. Meanwhile, according to the research of [10], to reduce the computational cost, Ctotal and Cinf are scaled down with a proportion ρ=4.07×10−5 for simulation, and the model results CNew are expanded with the same proportion after the simulation; the appropriate value of parameters α and β are set as 7.00 μg and 6.20×10−4, respectively. The simulation time step size of pedestrian dynamics is 0.04 s. All simulations are implemented in Microsoft Visual C++ and conducted on a Windows server with Intel Xeon CPU E5-2630 v3 2.40 GHz and 128 GB RAM. On our PC, the simulation of a scene takes about 24 h, and at most 30 scenes can be simulated at the same time.
4 Results and analysis
4.1 Simulation results
Obstacle factors are calculated based on the simulation setups, and transmission trend indicators are collected from the simulation outputs. Based on the simulation results, the descriptive statistics of the independent and dependent variables are reported in Table 2 . Moreover, as there are 120 samples in our dataset, we explore the distributions of variables with the Shapiro-Wilk test, and the results are shown in Table 2.Table 2 Descriptive statistics and the normality test results of independent and dependent variables.
Table 2Variables Number of Samples Max. Values Min. Values Mean Standard Deviation Skewness Kurtosis Shapiro-Wilk Test
Independent Variables
RAllObs 120 40.50% 7.44% 22.31% 11.82% 0.295 −1.249 0.875***
NObs 120 12 1 5 3 0.803 −0.294 0.892***
DCon 120 102.849 15.011 46.368 20.037 0.594 −0.151 0.961***
Dependent Variables
CNew 120 42,781 22,560 30,238 4697 0.535 −0.407 0.961***
EAve 120 4.781 2.998 3.644 0.415 0.668 −0.246 0.949***
Note: * p <0.1, ** p <0.05, *** p <0.01.
According to Table 2, all p values of Shapiro-Wilk test statistics are significant, and it can be preliminarily judged that all variables don't obey normal distribution characteristics initially. However, if the absolute value of kurtosis is less than 10 and the absolute value of skewness is less than 3, the data can be accepted as approximately normally distributed [52,53]. Then, the Pearson coefficients are applied for correlation analysis, and the results are reported in Table 3 .Table 3 Correlations between all variables.
Table 3Variables RAllObs NObs DCon CNew EAve
RAllObs 1
NObs −0.003 1
DCon −0.837*** −0.358*** 1
CNew 0.917*** −0.245*** −0.732*** 1
EAve 0.906*** −0.271*** −0.710*** 0.998*** 1
Note: * p <0.1, ** p <0.05, *** p <0.01.
As shown in Table 3, there is no correlation between the size indicator RAllObs and quantity indicator NObs, and the relationship between any other two variables is significant. As it is difficult to show the change of dependent variables with all three independent variables in one figure, and considering that RAllObs is not related to NObs, the transmission indicators change with the size indicator RAllObs and placement indicator DCon in Fig. 11 (a and b) and they vary with the quantity indicator NObs and placement indicator DCon in Fig. 11(c and d).Fig. 11 (a) CNew and (b) EAve vary with RAllObs and DCon; (c) CNew and (d) EAve change with NObs and DCon.
Fig. 11
4.2 Structural equation modeling
The structural equation model (SEM) is established and tested with path analysis to determine the effect of obstacle factors on transmission trends [[54], [55], [56]]. It is a powerful technique to estimate the hypothesized patterns of direct, indirect, and mediating relationships among variables. Moreover, compared with single-equation linear regression modeling, structural equation modeling allows us to simultaneously estimate the effects of multi-variables (RAllObs, NObs, and DCon) on the dependent variable (CNew or EAve).
According to the results of section 4.1 and domain knowledges, when the dependent variable is the daily new cases CNew, we present the theoretical framework of Model A in Fig. 12 (a), and formulate the study hypotheses as follows:➢ Hypothesis 1A (H 1A): the size indicator RAllObs has a positive (+) effect on CNew.
➢ Hypothesis 1B (H 1B): the quantity indicator NObs has a negative (−) effect on CNew.
➢ Hypothesis 1C (H 1C): the placement indicator DCon mediates the association between RAllObs and CNew.
➢ Hypothesis 1D (H 1D): the placement indicator DCon mediates the association between NObs and CNew.
Fig. 12 The hypothetical relations in (a) Model A and (b) Model B.
Fig. 12
Besides, when the dependent variable is the people's average exposure risk EAve, the framework of the Model B is illustrated in Fig. 12(b), and the following hypotheses are proposed:➢ Hypothesis 2A (H 2A): the size indicator RAllObs has a positive (+) effect on EAve.
➢ Hypothesis 2B (H 2B): the quantity indicator NObs has a negative (−) effect on EAve.
➢ Hypothesis 2C (H 2C): the placement indicator DCon mediates the association between RAllObs and EAve.
➢ Hypothesis 2D (H 2D): the placement indicator DCon mediates the association between NObs and EAve.
Therefore, RAllObs and NObs are independent variables; DCon is the mediating variable; CNew and EAve are respectively the dependent variables in Model A and Model B.
After using standardized data with the min-max scaler method, the maximum likelihood method is adopted to estimate the fitness of the hypothesized models [54], and several fit indices are used to evaluate: the Chi-square (χ2), the p-value of χ2 (p), the relative Chi-square (χ2/df), goodness-of-fit index (GFI), comparative fit index (CFI), root mean square error of approximation (RMSEA), and root mean square residual (RMR). The results and recommended thresholds of fit indices are summarized in Table 4 , which illustrates a satisfactory fit in each model [55,57]. In addition, the path analysis results of hypothesized models are illustrated in Table 5 and Fig. 13 .Table 4 The fit indices of model A and model B.
Table 4Model Fit Indices (Recommended Threshold [58,59])
χ2 (N/A) p (>0.05) χ2/df (<3) GFI (>0.9) CFI (>0.9) RMSEA (<0.05) RMR (<0.05)
A 0.001 0.971 0.001 1.000 1.000 0.000 0.000
B 0.001 0.971 0.001 1.000 1.000 0.000 0.000
Table 5 Path analysis results of hypothesized model A and model B.
Table 5Model Paths Estimate S.E. C.R. p Standardized Estimate
A RAllObs → CNew 0.430 0.040 10.884 *** 0.662
RAllObs → DCon −0.535 0.024 −22.254 *** −0.838
NObs → CNew −0.283 0.029 −9.877 *** −0.352
NObs → DCon −0.285 0.030 −9.567 *** −0.360
DCon → CNew −0.309 0.066 −4.657 *** −0.304
B RAllObs → EAve 0.435 0.042 10.481 *** 0.668
RAllObs → DCon −0.535 0.024 −22.254 *** −0.838
NObs → EAve −0.299 0.030 −9.920 *** −0.370
NObs → DCon −0.285 0.030 −9.567 *** −0.360
DCon → EAve −0.289 0.070 −4.152 *** −0.284
Note: * p <0.1, ** p <0.05, *** p <0.01; S.E. is short for standard error; C.R. is short for critical ratio.
Fig. 13 The analysis results of (a) Model A and (b) Model B. Note: *** p <0.01.
Fig. 13
As shown in Table 5, in Model A, the size indicator RAllObs has a significant and positive effect on the daily new cases CNew (p < 0.001), hence Hypothesis 1A is proved. Similarly, Hypotheses 1B, 2A, and 2B are accepted. Moreover, to verify Hypotheses 1C, 1D, 2C, and 2D, we should test the mediating effects of the placement indicator DCon in each model. Here, a bias-corrected bootstrapping method with 1000 resamples and a 95% confidence interval is used [60], and the results are summarized in Table 6 . Consequently, in both paths RAllObs →DCon →CNew and RAllObs →DCon →EAve, as the 95% CI of DCon does not contain zero and the indirect effect has the same sign as the total effect, DCon is a mediator in each path (proving Hypotheses 1C and 2C) [60,61]. Differently, even the 95% CI of DCon does not contain zero in paths NObs →DCon →CNew and NObs →DCon →EAve, the indirect effect has the opposite sign of the total effect, so the placement indicator DCon is a suppressor in each path, which weakens the observed relationship by its omission [61,62]. Specifically, once the placement indicator DCon is fixed, the effects of the quantity indicator NObs on transmission indicators increase. As the suppression effects are known as inconsistent mediation effects [63], Hypotheses 1D and 2D are rejected.Table 6 The effects of RAllObs and NObs on dependent variables in model A and model B.
Table 6Model Paths Total Effects Direct Effects Indirect Effects Description
Estimate S.E. 95% CI
A RAllObs → DCon → CNew 0.595*** 0.430*** 0.165*** 0.049 0.071–0.266 Mediation Effects
NObs → DCon → CNew −0.195*** −0.283*** 0.088*** 0.030 0.034–0.155 Suppressing Effects
B RAllObs → DCon → EAve 0.590*** 0.435*** 0.155*** 0.049 0.062–0.260 Mediation Effects
NObs → DCon → EAve −0.217*** −0.299*** 0.083*** 0.030 0.029–0.148 Suppressing Effects
Note: * p <0.1, ** p <0.05, *** p <0.01; S.E. is short for standard error; CI is short for confidence interval.
4.3 Analysis under controlled size and quantity
In this section, we explore the influences of obstacles’ placement on epidemic transmission when the size indicator RAllObs and the quantity indicator NNon are controlled simultaneously. Scenario #100- [1,2]-(0,0) is selected as the base, and the shortest distance between adjacent obstacles Dcut changes from 0.2 m to 6.0 m with a step of 0.2 m to create different obstacle layouts. Thus, the diverse values of Dcut bring various values of the placement indicator DCon and further affect the transmission. In sum, there are total 30 scenarios (including the base one), and these layouts are shown in Fig. 14 . Based on the simulation results, the daily new cases CNew and the average exposure risk EAve change with the DCon in Fig. 15 . Further, the influences of the placement indicator DCon are studied according to the linear regression analysis with the ordinary least square (OLS) method, i.e., the disease spreading indicators (CNew and EAve) are modeled as the linear function of the placement indicator DCon. Consequently, regression results are demonstrated in Table 7 .Fig. 14 Different layouts with Dcut changes from 0.2 m to 6.0 m with a step of 0.2 m based on scenario #100- [1,2]-(0,0).
Fig. 14
Fig. 15 (a) CNew and (b) EAve change with the placement indicator DCon in 30 scenarios with different Dcut.
Fig. 15
Table 7 Summary of linear regression results.
Table 7Variables Dependent Variable CNew Dependent Variable EAve (μg)
Coefficient S.E. t p Coefficient S.E. t p
DCon 199.612 57.550 3.469 0.002*** 0.018 0.005 3.424 0.002***
Constant 18949.711 2220.862 8.533 0.000*** 2.640 0.198 13.316 0.000***
R-Squared 0.301 – 0.295 –
Adjusted R-Squared 0.276 – 0.270 –
F-statistic 12.031 – 11.725 –
Prob (F-statistic) 0.002*** – 0.002** –
Durbin-Watson stat 0.509 – 0.535 –
Note: * p <0.1, ** p <0.05, *** p <0.01; S.E. is short for standard error.
Based on Table 7, the R-squared values of 0.301 and 0.295 indicate that the transmission variables CNew and EAve can be explained by the placement indicator DCon (30.1% and 29.5%), respectively. Besides, there is a significant and positive influence of the placement indicator DCon on epidemic transmission variables (CNew and EAve).
5 Discussions and future perspectives
5.1 Discussions
5.1.1 Influences of the size
In a scenario, the transmission indicators (the number of new cases CNew and the average exposure risk of all individuals EAve) change with the total size of obstacles SAllObs in Fig. 16 . With the growth of the obstacle size, the walkable space reduces, and the people density (= the size of walkable space/number of individuals in the simulation room) increases accordingly. Hence, infectious spaces constructed by infectors are denser, and there is a higher possibility that susceptible individuals arrive at the infectious areas and facilitates epidemic spreading. In addition, the higher density brings a lower velocity [64,65], which results in a long time to pass through the infectious space and then increases the transmission. As the desired velocity is a fixed setup in our cases, we can further explore the impacts of the various crowd walking speeds on the infection spread as in previous studies [30].Fig. 16 (a) CNew and (b) EAve vary with the total size of obstacles SObs.
Fig. 16
5.1.2 Influences of the quantity
There is a finding from the previous analysis: increasing the obstacle quantity indicator NObs reduces the epidemic infections, whose possible reason is that the growing quantity of obstacles decreases the walkable infector influence areas in the space. Specifically, with the growth of the quantity indicator NObs, the total side length of the obstacles increases, which enhances the interactions between individuals and obstacles. Then, more infection areas and obstacles overlap in two-dimension (2D), resulting in these overlapping infection areas being non-walkable, as shown in Fig. 17 . Thus, the walkable infector influence areas decrease, thus reducing the epidemic transmission.Fig. 17 When (a) an obstacle is divided into (b) two sub-ones, there are more overlapping areas between infection areas and obstacles.
Fig. 17
5.1.3 Influences of the placement
According to the path analysis results in Table 5, the direct effect of the placement indicator DCon on the new cases CNew is −0.309 (p <0.01), and the direct effect of the placement indicator DCon on the average exposure risk EAve is −0.289 (p <0.01). Hence, it can be preliminarily considered that the placement indicator DCon has a negative effect on transmission trends. However, in Fig. 11(a and b), “Simpson's paradox” is found: although there are negative relationships between the epidemic transmission variables and the placement indicator DCon based on the overall data, the opposite relations are observed in each group with the same size indicator RAllObs [66]. Specifically, when the size indicator RAllObs is controlled and large (e.g., 40.50%, 29.75%, 20.66%), the positive relationships between the epidemic spreading variables and the placement indicator DCon are obtained. It should be noted that, there isn't “Simpson's paradox” in Fig. 11(c and d): epidemic spreading indicators decrease with the increase of the placement indicator DCon whether in the overall data or each subclass. In sum, there are:1) once the size indicator RAllObs is controlled and large enough, the rise of the placement indicator DCon increases the transmission; 2) once the quantity indicator NNon is controlled, the growth of the placement indicator DCon brings a reduction in epidemic infections.
Based on the results from section 4.3, once the obstacles’ size and quantity are controlled simultaneously, with the rise of the placement indicator DCon, more and more individuals will be infected and the average exposure risk of the global will increase. The possible reasons for this are illustrated as follows. A larger value of the placement indicator DCon brings a large upper bound and a small lower bound of the walkable convex space, which leads to different widths of the “door” linking adjacent spaces, as shown in Fig. 18 . The individual gathering and congestion always happen in the narrow “door” and lead to a longer time of accompanying motion between susceptible individuals and infectors, which increases the transmission [67,68]. Therefore, once the obstacle is large enough, the more uniform distribution of walkable convex space (i.e., decreasing the placement indicator DCon) brings equal width “doors” and further lowers the epidemic spreading. It should be noted that when the total obstacle size is too small, the transmission trends affected by the quantity or placement factors are minor, which is unlike to cause significant variations.Fig. 18 The sketch map of the “door” linking adjacent spaces in the case of (a) Fig. 3 and (b) Fig. 4.
Fig. 18
5.1.4 Transmission routes
For respiratory viruses, there are three transmission routes in existing research [40]: contact with virus-containing secretions (including directly and through fomites); droplet transmission (usually falling to contaminating surfaces in a short range); and aerosol transmission (including short and long ranges).
Existing studies have found that airborne (droplet and aerosol) transmission is the dominant route of transmission in RIDs such as COVID-19 [48], so the contact transmission route is not considered in our study. Besides, long-range aerosol transmission is not contained in our study, and the reasons are explained as follows. Researchers have found that aerosols can easily be located up to 20 m from their origin under different environmental factors [69]. Our simulation room is 22 m × 22 m, which is a small room relative to the distance of long-range propagation. Hence, wherever the coughs are generated, the long-range aerosols are evenly mixed in the room to some extent, and this happens in all scenarios with different obstacles. On the other hand, viruses can remain viable to cause infection in the air for many hours and even days posing the risk of long-range aerosol transmission [48]. All people (including infectors and susceptible individuals) in our simulation have the same dwell time (i.e., 25 min) in the room, which is short relative to the alive time of viruses in long-range aerosol transmission. Therefore, whenever individuals enter the room, they are suffered from long-range aerosols during their whole dwell time, which occurs in all scenes with various obstacles. In sum, long-range aerosol transmission has a long transmission distance and a long infection time, and without considering it would bring a lower risk than the real. However, since long-range aerosol transmission happens in all scenarios with different obstacles, our obstacle conclusions are obtained based on the relative exposure risks in different obstacle setup scenarios. Hence, without considering long-range aerosol transmission has negligible influence on the conclusions in this paper.
Overall, only droplet transmission and short-range aerosol transmission are considered in our study. The threshold of the short range in this study is set the same as the cough infectious distance (i.e., 1.70 m) in the ERP model [10], and it is also consistent with the existing research (1 m–2 m) [70].
5.2 Future perspectives
There are several limitations that need to be further studied.
Firstly, due to a lack of real-world data to achieve the research goal, we use the simulated data based on the ERP fundamental model. As the ERP model has been quantitatively calibrated and validated through the macroscopic real-life data, the prediction transmission trends are reliable to some extent, and the results can be applied for scenario investigation and comparison. Once we get real-world data in the future, the findings in this paper can be further validated.
Secondly, in our cases, individuals walk freely inside the simulated space, but people's diverse behavior and points of interest in the scenario are ignored. For example, in a health club facility, the individual stays on a fixed treadmill for a long time instead of walking randomly [2]. Thus, individuals moving patterns in various scenarios should be considered in the future, which helps determine the customized prevention and control measures for a scenario.
Thirdly, as mentioned in section 2.1, all obstacles in a scenario are requested to have the same size and shape in our study. Compared with other shapes, rectangular (including square) obstacles are more conducive to dividing multiple identical sub-obstacles, and thus the obstacle's shape is fixed as rectangular here. Besides, the pedestrian's shape is set as a circle for simplicity. However, both the pedestrian and the obstacle in real scenarios have various shapes and thus bring more dynamic properties. Therefore, further studies can be carried out in applying realistic object features.
Lastly, previous studies have shown that environmental factors (e.g., temperature, humidity, and ventilation) can affect the transmission of virion-laden particles [24,43,71,72]. However, these factors are set to common values of the cough in our study based on literatures [44,73,74]. In the future, these factors can be explored as independent variables in combination with the ERP model (see Fig. 19). For example, under different ventilation rates, the cough can be modeled through Computational Fluid Dynamics (CFD) simulations. Then, the movements of cough-generated particles are obtained, which affects the second step of the ERP model and further varies the epidemic transmission trends. Meanwhile, environmental variables can be added to structural equation models A and B in section 4.2, as shown in Fig. 20. It should be mentioned that variant environmental factors could lead to a shorter/longer transmission distance of cough particles, and the threshold of cough infection distance needs to be changed accordingly.Fig. 19 Environmental factors can be explored as potential independent variables in combination with the ERP model.
Fig. 19
Fig. 20 The hypothetical relations in (a) Model A and (b) Model B after adding environmental factors.
Fig. 20
6 Conclusions
In this paper, the association between obstacle factors (i.e., size, quantity, and placement) and RIDs transmission trends is examined. There are four findings in our study. 1) Decreasing the obstacle size reduces the epidemic spreading by lowering the probability of susceptible individuals reaching the walkable infectious spaces and the time of passing through the space. 2) Increasing the obstacle quantity can decrease infections by reducing the overlapping areas between infection areas and obstacles. 3) Once the walkable-space distribution indicator presented in this paper is controlled, the effect of obstacle size on transmission decreases, and the influence of obstacle quantity on transmission increases. 4) When the total obstacle size is large enough, placing the obstacles uniformly helps decrease the epidemic transmission by creating equal-width “doors” linking adjacent walkable convex spaces. Moreover, these findings can be applied in the daily presentation and control of RIDs with other measures such as increasing ventilation rates [1].
Our findings suggest that minimizing the obstacle size, dividing the obstacle into more sub-blocks, and putting them evenly in space can help decrease the transmission of RIDs in indoor public places.
Funding sources
This work was supported by 10.13039/501100001809 National Natural Science Foundation of China (Grant No. 72101276), Shenzhen Science and Technology Program (Grant No. GXWD 20200830165201001), Fundamental Research Funds for the Central Universities, 10.13039/501100002402 Sun Yat-sen University (Grant No. 22qntd1710), Basic and Applied Basic Research Project of Guangzhou Municipal Science and Technology Bureau (Grant No. 202102020275).
CRediT authorship contribution statement
Ziwei Cui: Conceptualization, Methodology, Writing – original draft. Ming Cai: Conceptualization, Supervision, Writing – review & editing. Yao Xiao: Conceptualization, Methodology, Writing – review & editing, Funding acquisition. Zheng Zhu: Data curation, Writing – review & editing. Gongbo Chen: Data curation, Writing – review & editing.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Yao Xiao reports financial support was provided by 10.13039/501100001809 National Natural Science Foundation of China . Yao Xiao reports financial support was provided by Shenzhen Science and Technology Program. Yao Xiao reports financial support was provided by Fundamental Research Funds for the Central Universities, 10.13039/501100002402 Sun Yat-sen University . Gongbo Chen reports financial support was provided by Basic and Applied Basic Research Project of Guangzhou Municipal Science and Technology Bureau.
Appendix A For the numerical simulation of a typical cough, the computational domain is introduced first, and then the methods and parameter settings selected from existing studies are illustrated.
An average-sized manikin with 1.70 m tall is considered as the infector, whose mouth (i.e., cough inlet) with an area of 3.70 × 10−4 m2 is 1.50 m from the floor [44,75]. Meanwhile, people with a height from 1.40 m to 2.00 m are studied to determine the susceptible individuals’ breathing height area: 1.20 m–1.80 m. Based on these, the computational domain is constructed as a cuboid [10], as shown in the blue area of Fig. A1. Here, the center of the cough inlet is denoted as the origin of the coordinate system with (0, 0, 0); “x”, “y”, and “z” axes represent the lateral, axial (streamwise), and vertical directions, respectively; the gravitational acceleration is −9.81 m/ s2 along the “z” axis. We generate the computational domain and grids with the Gambit 2.4.6, and the grid resolution is set as 0.01 m which results in approximately 1.80 million computational cells.Fig. A1 Schematic diagram of (a) the computational domain with infector and susceptible manikins, and (b) the numerical simulation computational domain. (reproduced from Ref. [10]).
Fig. A1
The coughing transmission medium is modeled as an incompressible ideal gas with constant properties calculated at ambient conditions. Since the flow field evolution of the cough is time-dependent, simulations are conducted under a transient condition. Moreover, the Eulerian-Lagrangian method is utilized and the energy equation is required. The renormalization group (RNG) k-ε model with scalable wall functions is used as the turbulence model. Meanwhile, the component transport model is implemented, and the discrete phase model (DPM) is adopted for the diffusion of particles [46,76]. In the DPM, we use unsteady particle tracking and set the particle time step size to 1 ×10−3 s. Then, the tracking parameters are given: the max number of steps is 50,000 and the step length factor is 5. Besides, physical methods including two-way turbulence coupling, stochastic collision, break-up, and coalescence are utilized.
The cough is simulated in a closed space with a temperature of 26 °C and relative humidity of 50% [44], and all surfaces (except the inlet) are set to be the “trap” boundary condition [44,77]. Besides, the boundary condition of the inlet is set as a velocity-inlet and the velocity direction is programmed as normal to the inlet boundary. The fluid injected from the inlet consists of air (the mole fraction is 93.80%) and water vapor (the mole fraction is 6.20%) at 35 °C [74]. Moreover, the cough particles consist of multi-components: 94% volume fraction of the water (can be evaporated) and the rest is the non-evaporative water (represents the virus-carried mucus and physiological electrolytes) [73,74]. For the size distribution of the particles produced by coughing, like some existing literatures [44,74], we use the distribution from Ref. [78] (see Table A1) and adopt the Rosin-Rammler distribution to fit the data. Next, as the same with Ref. [44,79], the inlet cough velocity is set to be a function of time with a peak velocity (around 22.06 m/s) at 0.066 s, and particles are injected in the time range of 0.042s–0.136 s. The spherical drag model and the effect of the secondary break-up with a Taylor Analogy Breakup model are considered. Besides, we use the Discrete Random Walk model to capture the effects of the turbulent flow on the particles.Table A1 The size distribution of the cough-generated particles. (reproduced from Ref. [78])
Table A1Class Diameter Class (μm) Number of Particles
1 3 76
2 6 1041
3 12 386
4 20 127
5 28 47
6 36 45
7 45 38
8 62.5 38
9 87.5 27
10 112.5 32
11 137.5 30
12 175 83
13 225 47
14 375 40
15 750 27
In the simulation solution, the pressure-based solver (suitable for incompressible flows), the Semi-Implicit Method for Pressure-Linked Equations algorithm, the first-order upwind for turbulence-related variables, and the second-order for other variables (e.g., pressure) are employed. Besides, since the bounded second-order implicit formulation could bring high accuracy and good stability, it is selected as the transient formulation. A simulation time step size of 0.01 s is used with 10 sub-iterations, and the total flow time of the simulation is 15.00 s to investigate the dynamic characteristics of the cough-generated particles.
Appendix B There are five different sizes of obstacles in our study, i.e. 36 m2, 64 m2, 100 m2, 144 m2, and 196 m2. As the scenarios with the size of 36 m2 have been shown in Fig. 10, the rest scenarios are illustrated in Fig. B1.Fig. B1 There are scenarios with the obstacle size of 64 m2, 100 m2, 144 m2, and 196 m2.
Fig. B1
Data availability
Data will be made available on request.
==== Refs
References
1 Ahmadzadeh M. Shams M. Multi-objective performance assessment of HVAC systems and physical barriers on COVID-19 infection transmission in a high-speed train J. Build. Eng. 53 2022 104544 10.1016/j.jobe.2022.104544
2 Ibrahim A.M. Hassanain M.A. Assessment of COVID-19 precautionary measures in sports facilities: a case study on a health club in Saudi Arabia J. Build. Eng. 46 2022 103662 10.1016/j.jobe.2021.103662
3 Asif Z. Chen Z. Stranges S. Zhao X. Sadiq R. Olea-Popelka F. Peng C. Haghighat F. Yu T. Dynamics of SARS-CoV-2 spreading under the influence of environmental factors and strategies to tackle the pandemic: a systematic review Sustain. Cities Soc. 81 2022 103840 10.1016/j.scs.2022.103840
4 Flaxman S. Mishra S. Gandy A. Unwin H.J.T. Mellan T.A. Coupland H. Whittaker C. Zhu H. Berah T. Eaton J.W. Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe Nature 584 7820 2020 257 261 32512579
5 Navas-Martín M.Á. Oteiza I. Cuerdo-Vilches T. Dwelling in times of COVID-19: an analysis on habitability and environmental factors of Spanish housing J. Build. Eng. 60 2022 105012 10.1016/j.jobe.2022.105012
6 Melikov A.K. COVID-19: reduction of airborne transmission needs paradigm shift in ventilation Build. Environ. 186 2020 107336
7 Huang G. Guo F. Loss of life expectancy due to respiratory infectious diseases: findings from the global burden of disease study in 195 countries and territories 1990–2017 J. Popul. Res. 2022 1 43
8 Xie Y. Bowe B. Al-Aly Z. Burdens of post-acute sequelae of COVID-19 by severity of acute infection, demographics and health status Nat. Commun. 12 1 2021 1 12 33397941
9 Chen P. Zhang D. Liu J. Jian I.Y. Assessing personal exposure to COVID-19 transmission in public indoor spaces based on fine-grained trajectory data: a simulation study Build. Environ. 218 2022 109153 10.1016/j.buildenv.2022.109153
10 Cui Z. Cai M. Xiao Y. Zhu Z. Yang M. Chen G. Forecasting the transmission trends of respiratory infectious diseases with an exposure-risk-based model at the microscopic level Environ. Res. 212 2022 113428 10.1016/j.envres.2022.113428
11 Katal A. Wang L.L. Albettar M. A real-time web tool for monitoring and mitigating indoor airborne COVID-19 transmission risks at city scale Sustain. Cities Soc. 80 2022 103810
12 Ai Z.T. Melikov A.K. Airborne spread of expiratory droplet nuclei between the occupants of indoor environments: a review Indoor Air 28 4 2018 500 524 29683213
13 Braidotti L. Bertagna S. Dodero M. Piu M. Marinò A. Bucci V. Identification of measures to contain the outbreaks on passenger ships using pedestrian simulations Procedia Comput. Sci. 200 2022 1565 1574 10.1016/j.procs.2022.01.357 35284027
14 Qian H. Miao T. Liu L. Zheng X. Luo D. Li Y. Indoor transmission of SARS‐CoV‐2 Indoor Air 31 3 2021 639 645 33131151
15 Garcia W. Mendez S. Fray B. Nicolas A. Model-based assessment of the risks of viral transmission in non-confined crowds Saf. Sci. 144 2021 105453
16 Ku D. Yeon C. Lee S. Lee K. Hwang K. Li Y.C. Wong S.C. Safe traveling in public transport amid COVID-19 Sci. Adv. 7 43 2021 g3691
17 Löhner R. Antil H. High fidelity modeling of aerosol pathogen propagation in built environments with moving pedestrians Int. J. Numer. Meth. Biomed. 37 3 2021
18 Lu T. Zhao Y. Wu P. Zhu P. Dynamic analysis of single-file pedestrian movement with maintaining social distancing in times of pandemic J. Statistical Mechanic. 2021 9 2021 93402
19 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
20 Moritz S. Gottschick C. Horn J. Popp M. Langer S. Klee B. Purschke O. Gekle M. Ihling A. Zimmermann F.D.L. Mikolajczyk R. The risk of indoor sports and culture events for the transmission of COVID-19 Nat. Commun. 12 1 2021
21 Qi B. Tan J. Zhang Q. Cao M. Wang X. Zou Y. Unfixed movement route model, non-overcrowding and social distancing reduce the spread of COVID-19 in sporting facilities Int. J. Environ. Res. Publ. Health 18 15 2021 8212
22 Salmenjoki H. Korhonen M. Puisto A. Vuorinen V. Alava M.J. Modelling aerosol-based exposure to SARS-CoV-2 by an agent based Monte Carlo method: risk estimates in a shop and bar PLoS One 16 11 2021 e260237
23 Xu Q. Chraibi M. On the effectiveness of the measures in supermarkets for reducing contact among customers during COVID-19 period Sustainability 12 22 2020 9385
24 Vuorinen V. Aarnio M. Alava M. Alopaeus V. Atanasova N. Auvinen M. Balasubramanian N. Bordbar H. Erästö P. Grande R. Modelling aerosol transport and virus exposure with numerical simulations in relation to SARS-CoV-2 transmission by inhalation indoors Saf. Sci. 130 2020 104866
25 Contardo C. Costa L. On the Optimal Layout of a Dining Room in the Era of COVID-19 Using Mathematical Optimization, arXiv 2021
26 Bañón L. Bañón C. Improving room carrying capacity within built environments in the context of COVID-19 Symmetry 12 10 2020 1683
27 Mekawy M. Gabr M.A. Against a workplace contagion: a digital approach to support hygiene-conscious office space planning Open House Int. 46 3 2021 391 400
28 Shorfuzzaman M. Hossain M.S. Alhamid M.F. Towards the sustainable development of smart cities through mass video surveillance: a response to the COVID-19 pandemic Sustain. Cities Soc. 64 2021 102582 10.1016/j.scs.2020.102582
29 Azmi N.Z.E. Hidayat M.I.P. Pramata A.D. Numerical simulation of effects of number of beds and presence of aerosol flow from sanitation machine to air circulation in hospital isolation room of COVID-19 patients Mater. Res. Commun. 2 1 2021
30 Xiao T. Mu T. Shen S. Song Y. Yang S. He J. A dynamic physical-distancing model to evaluate spatial measures for prevention of Covid-19 spread Phys. Stat. Mech. Appl. 592 2022 126734 10.1016/j.physa.2021.126734
31 Bapir S.Y. Kareem S.M. Covid-19 and functionality: by providing social distancing of indoor common spaces in residential building J. Studies Sci. Eng. 1 1 2021 36 45
32 Bapir S.Y. Kareem S.M. How to limit the spread of covid-19 in residential buildings: erbil city as a case study Des. Eng. 2021 2635 2647
33 Hillier B. Hanson J. The Social Logic of Space 1984 Cambridge university press Cambridge, UK
34 Haghani M. Crowd dynamics research in the era of Covid-19 pandemic: challenges and opportunities Saf. Sci. 153 2022 105818 10.1016/j.ssci.2022.105818
35 Ronchi E. Lovreglio R. EXPOSED: an occupant exposure model for confined spaces to retrofit crowd models during a pandemic Saf. Sci. 130 2020 104834
36 Xiao Y. Yang M. Zhu Z. Yang H. Zhang L. Ghader S. Modeling indoor-level non-pharmaceutical interventions during the COVID-19 pandemic: a pedestrian dynamics-based microscopic simulation approach Transport Pol. 109 2021 12 23 10.1016/j.tranpol.2021.05.004
37 Hernández-Orallo E. Armero-Martínez A. How human mobility models can help to deal with covid-19 Electronics 10 1 2020 33
38 Helbing D. Molnar P. Social force model for pedestrian dynamics Phys. Rev. E. 51 5 1995 4282
39 Sticco I.M. Frank G.A. Dorso C.O. Social force model parameter testing and optimization using a high stress real-life situation Phys. Stat. Mech. Appl. 561 2021 125299
40 Tellier R. COVID-19: the case for aerosol transmission Interface Focus 12 2 2022 20210072
41 Milton D.K. A Rosetta stone for understanding infectious drops and aerosols J. Pediatr. Infect. Dis. Soc. 9 4 2020 413 415 10.1093/jpids/piaa079
42 Prather K.A. Marr L.C. Schooley R.T. McDiarmid M.A. Wilson M.E. Milton D.K. Airborne transmission of SARS-CoV-2 Science 370 6514 2020 303 304
43 Wang C.C. Prather K.A. Sznitman J. Jimenez J.L. Lakdawala S.S. Tufekci Z. Marr L.C. Airborne transmission of respiratory viruses Science 373 6558 2021 d9149
44 Borro L. Mazzei L. Raponi M. Piscitelli P. Miani A. Secinaro A. The role of air conditioning in the diffusion of Sars-CoV-2 in indoor environments: a first computational fluid dynamic model, based on investigations performed at the Vatican State Children's hospital Environ. Res. 193 2021 110343 10.1016/j.envres.2020.110343
45 Quiñones J.J. Doosttalab A. Sokolowski S. Voyles R.M. Castaño V. Zhang L.T. Castillo L. Prediction of respiratory droplets evolution for safer academic facilities planning amid COVID-19 and future pandemics: a numerical approach J. Build. Eng. 54 2022 104593 10.1016/j.jobe.2022.104593
46 Zhang L. Li Y. Dispersion of coughed droplets in a fully-occupied high-speed rail cabin Build. Environ. 47 2012 58 66 10.1016/j.buildenv.2011.03.015
47 Zhao X. Liu S. Yin Y. Zhang T. Chen Q. Airborne transmission of COVID‐19 virus in enclosed spaces: an overview of research methods Indoor Air 32 6 2022 e13056
48 Rabaan A.A. Al-Ahmed S.H. Al-Malkey M. Alsubki R. Ezzikouri S. Hassan Al-Hababi F. Sah R. Mutair A.A. Alhumaid S. Al-Tawfiq J.A. Airborne transmission of SARS-CoV-2 is the dominant route of transmission: droplets and aerosols Inf. Med. 29 1 2021 10 19
49 Maryland transportation institute and center for advanced transportation Technology laboratory at the university of Maryland, trips by distance – national. 2021 https://data.bts.gov/Research-and-Statistics/Trips-by-Distance-National/6ced-86in
50 Zhang L. Darzi A. Ghader S. Pack M.L. Xiong C. Yang M. Sun Q. Kabiri A. Hu S. Interactive COVID-19 mobility impact and social distancing analysis platform Transport. Res. Rec. 2021 862848922 10.1177/03611981211043813
51 Centers for Disease Control and Preventions Covid-19 case surveillance public data access, summary, and limitations https://data.cdc.gov/Case-Surveillance/United-States-COVID-19-Cases-and-Deaths-by-State-o/9mfq-cb36 2021
52 Liu H. Official social media and its impact on public behavior during the first wave of COVID-19 in China BMC Publ. Health 22 1 2022 428 10.1186/s12889-022-12803-y
53 Kulip J. Jeffree M.S. Pang N.T.P. Nasirruddin N. Wider W. Relationships between coping styles, emotional distress, and fear of COVID-19 among workers in the oil and gas industry in Malaysia during the COVID-19 pandemic Sustainability 14 9 2022 5398 10.3390/su14095398
54 Abadi B. Mahdavian S. Fattahi F. The waste management of fruit and vegetable in wholesale markets: intention and behavior analysis using path analysis J. Clean. Prod. 279 2021 123802 10.1016/j.jclepro.2020.123802
55 Tran L.N. Xuan J. Nakagami H. Kuroki S. Ge J. Gao W. Influence of household factors on energy use in Vietnam based on path analysis J. Build. Eng. 57 2022 104834 10.1016/j.jobe.2022.104834
56 Hauashdh A. Jailani J. Rahman I.A. Structural equation model for assessing factors affecting building maintenance success J. Build. Eng. 44 2021 102680
57 Chen X. Gou Z. Bridging the knowledge gap between energy-saving intentions and behaviours of young people in residential buildings J. Build. Eng. 57 2022 104932 10.1016/j.jobe.2022.104932
58 Garg S. Misra S. Causal model for rework in building construction for developing countries J. Build. Eng. 43 2021 103180 10.1016/j.jobe.2021.103180
59 Asman G.E. Kissi E. Agyekum K. Baiden B.K. Badu E. Critical components of environmentally sustainable buildings design practices of office buildings in Ghana J. Build. Eng. 26 2019 100925 10.1016/j.jobe.2019.100925
60 Yue C. Liu C. Wang J. Zhang M. Wu H. Li C. Yang X. Association between social support and anxiety among pregnant women in the third trimester during the coronavirus disease 2019 (COVID-19) epidemic in Qingdao, China: the mediating effect of risk perception Int. J. Soc. Psychiatr. 67 2 2021 120 127 10.1177/0020764020941567
61 Rucker D.D. Preacher K.J. Tormala Z.L. Petty R.E. Mediation analysis in social psychology: current practices and new recommendations Soc. Personal. Psychol. Compass. 5 6 2011 359 371
62 Preacher K.J. Hayes A.F. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models Behav. Res. Methods 40 3 2008 879 891 18697684
63 MacKinnon D.P. Krull J.L. Lockwood C.M. Equivalence of the mediation, confounding and suppression effect Prev. Sci. 1 4 2000 173 181 11523746
64 Cao S. Seyfried A. Zhang J. Holl S. Song W. Fundamental diagrams for multidirectional pedestrian flows J. Statist. Mechanic.: Theory Experim. 2017 3 2017 33404
65 Vanumu L.D. Ramachandra Rao K. Tiwari G. Fundamental diagrams of pedestrian flow characteristics: a review Eur. Transp. Res. Rev. 9 4 2017 1 13
66 Wagner C.H. Simpson's paradox in real life Am. Statistician 36 1 1982 46 48
67 Gaisie E. Oppong-Yeboah N.Y. Cobbinah P.B. Geographies of infections: built environment and COVID-19 pandemic in metropolitan Melbourne, Sust Cities Soc 81 2022 103838 10.1016/j.scs.2022.103838
68 Hamidi S. Hamidi I. Subway ridership, crowding, or population density: determinants of COVID-19 infection rates in New York city Am. J. Prev. Med. 60 5 2021 614 620 10.1016/j.amepre.2020.11.016 33888260
69 Reinhardt É.L. Transmission of COVID-19: a brief review of droplet and aerosol transmission routes Revista Brasileira de Saúde Ocupacional 47 2022
70 Peng Z. Rojas A.P. Kropff E. Bahnfleth W. Buonanno G. Dancer S.J. Kurnitski J. Li Y. Loomans M.G. Marr L.C. Practical indicators for risk of airborne transmission in shared indoor environments and their application to COVID-19 outbreaks Environ. Sci. Technol. 56 2 2022 1125 1137 34985868
71 Smither S.J. Eastaugh L.S. Findlay J.S. Lever M.S. Experimental aerosol survival of SARS-CoV-2 in artificial saliva and tissue culture media at medium and high humidity, Emerg Microb. Infect. 9 1 2020 1415 1417
72 Dabisch P. Schuit M. Herzog A. Beck K. Wood S. Krause M. Miller D. Weaver W. Freeburger D. Hooper I. The influence of temperature, humidity, and simulated sunlight on the infectivity of SARS-CoV-2 in aerosols Aerosol Sci. Technol. 55 2 2021 142 153
73 Aliabadi A.A. Rogak S.N. Green S.I. Bartlett K.H. CFD simulation of human coughs and sneezes: a study in droplet dispersion, heat, and mass transfer Proceedings of the ASME international mechanical engineering congress and exposition 7 2010 1051 1060
74 Bi R. A Numerical Investigation of Human Cough Jet Development and Droplet Dispersion 2018 University of Western Ontario London, Canada
75 Liu L. Li Y. Nielsen P.V. Wei J. Jensen R.L. Short‐range airborne transmission of expiratory droplets between two people Indoor Air 27 2 2017 452 462 27287598
76 Zhang Y. Feng G. Kang Z. Bi Y. Cai Y. Numerical simulation of coughed droplets in conference room Procedia Eng. 205 2017 302 308 10.1016/j.proeng.2017.09.981
77 Wang J. Wu Z. Wang H. Zhong M. Mao Y. Li Y. Wang M. Yao S. Ventilation reconstruction in bathrooms for restraining hazardous plume: mitigate COVID-19 and beyond J. Hazard Mater. 439 2022 129697 10.1016/j.jhazmat.2022.129697
78 Duguid J.P. The size and the duration of air-carriage of respiratory droplets and droplet-nuclei Epidemiol. Infect. 44 6 1946 471 479 10.1017/S0022172400019288
79 Gupta J.K. Lin C.H. Chen Q. Flow dynamics and characterization of a cough Indoor Air 19 6 2009 517 525 19840145
| 0 | PMC9737524 | NO-CC CODE | 2022-12-15 23:17:45 | no | 2023 Apr 1; 64:105706 | utf-8 | null | null | null | oa_other |
==== Front
J Interprof Educ Pract
J Interprof Educ Pract
Journal of Interprofessional Education & Practice
2405-4526
Elsevier Inc.
S2405-4526(22)00099-4
10.1016/j.xjep.2022.100592
100592
Article
Common herbs for stress: The science and strategy of a botanical medicine approach to self-care
Burns Joshua
Georgian Court University, USA
10 12 2022
3 2023
10 12 2022
30 100592100592
28 9 2022
21 11 2022
5 12 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.
Frontline healthcare workers have reported elevated levels of stress and increase prevalence of burnout symptoms since the onset of the COVID-19 pandemic. With these heightened levels of stress and burnout comes a need for more evidence-based interventions to address these symptoms earlier, in both a safe and effective way. Some common botanical medicines have a measurable effect on perceived stress, neurotransmitter levels, and circulating cortisol levels indicating their ability to modify the stress response. Botanical medicines are often relatively low cost, increasingly available in retail stores and online marketplaces, and show relatively low reports of adverse effects, making these medicinal herbs an important option for addressing work-related stress for healthcare workers.
==== Body
pmc1 Stress and burnout in the health-care industry
Stress among healthcare workers has increased significantly since the onset of the COVID-19 pandemic, leaving many workers with an increased need for strategies to cope with heightened stress, feelings of burnout and other mental health conditions. Burnout is characterized by the World Health Organization (WHO) as stemming from work-related stress that has not been resolved, having symptoms of fatigue and exhaustion, feelings of negativity about the workplace, a lack of personal connection and not feeling personal accomplishment from the job (2011).1 A 2022 study examined burnout prevalence among primary health-care professionals in lower- and middle-income countries, concluding that participants reported a high prevalence of emotional exhaustion (55.7%), depersonalization (39.1%), and diminished personal accomplishment (60.0%).2 This meta-analysis, among other studies, also categorizes burnout rates as highest among nurses and primary care physicians.2 , 3 In wealthier countries burnout rates were as high as 65.1%.4 Thus, early intervention strategies to address workplace stress are essential to mitigation of burnout in essential front-line healthcare fields. Self-care strategies and stress-management interventions have been identified as effective at reducing physician burnout.5 , 6 More safe and efficacious self-care strategies are needed to alleviate chronic work-place stress and to reduce numbers of healthcare workers that are reporting feelings of burnout. Botanical medicines provide healthcare workers with a safe and efficacious way to abate symptoms of chronic stress.
Botanical medicines have a long historical use to address stress and are popular to use among the general population with an estimated 35% of Americans reporting use of at least one herbal medicine.7 The ease of access, relative perceived safety of the medicines and various forms of delivery (such as teas, tinctures, and capsules) make them an easy self-care intervention when paired with a better understanding of the science and strategy of herbal medicines. The purpose of this article is to examine the safety and science behind four commonly used botanicals for stress reduction and to provide health-care workers with evidence-based strategies for the use of these common botanicals as self-care.
2 The science and safety of common botanicals for stress
2.1 Ashwagandha
Withania somnifera (Ashwagandha) has a long history of use for stress and sleep disturbances. An ayurvedic herb, it is commonly used in formulas designed to attenuate the stress hormone cortisol. Ashwagandha lowers circulating glucocorticoids cortisol and corticosterone through alteration of the Hypothalamic-pituitary-adrenal (HPA) axis, ameliorating stress symptoms.8,9 Additionally, it has been proposed that Ashwagandha has GABAergic and serotonergic effects by altering the release of both glutamate and GABA neurotransmitters, making it a potential therapeutic option for sleeping disorders, anxiety and depression.8 , 10 Glutamate, an excitatory neurotransmitter, is elevated during the acute stress response, while GABA is an inhibitory neurotransmitter with calming effects on the central nervous system. Ashwagandha's effectiveness as a stress reduction therapy should be measured for its ability to reduce perceived stress, and not just a decrease plasma cortisol. A 2019 clinical trial saw participants have a reduction on the Perceived Stress Scale (PSS) of 33.77%–38.34% in the Ashwagandha group, depending on the dose taken, compared to a 26.74% point drop on the PSS in the placebo group, showing a modest improvement from the herb.11 Another clinical trial of Ashwagandha resulted in a 44.0% reduction of PSS in the Ashwagandha group with only a 5.5% reduction noted in the placebo group.12 A 2021 review article further showed Ashwagandha ability in clinical trials to reduce markers of stress such as serum cortisol levels and reduced PSS scores compared to baseline levels from participants in the study.8 Based upon this evidence, studies lend credibility to the use of Ashwagandha for the reduction of stress as a regular self-care intervention.
Ashwagandha is generally perceived as a safe herb with its long history of traditional use, and consistent presence in herbal formulas. Moreover, the safety of Ashwagandha has been studied and established based on its relative few adverse reactions reported,13 and a lack of statistical change in biomarkers,14 , 15 liver enzymes, thyroid hormone and immunological markers tested.16 These studies demonstrate the relative safety of the herb for up to 8 weeks of use at doses up to 600 mg per day of herbal extract, and higher doses of 1250 mg per day for up to 10 days. The safety of the herb in relative high doses, combined with its efficacy in multiple studies makes Ashwagandha an excellent choice for a self-care intervention.
2.2 Rhodiola
Rhodiola rosea (Rhodiola) is an adaptogen used to manage physical symptoms of daily stress with a long history of traditional use. As a self-care therapy, this herb can be valuable in reducing mental fatigue, depression, and improving recovery from stress-induced physical fatigue. The European Medicines Agency (EMA) herbal monograph describes its traditional use for short-term relief of mild and moderate stress-related symptoms.17 A 2022 review article examines the effectiveness of rhodiola in stress-induced depression and fatigue as well as its ability to positively affect work productivity and provide anti-inflammatory benefits.18 A 2017 trial looked at rhodiola's efficacy in treating symptoms of burnout for a period of 12 weeks and found that Rhodiola was effective in alleviating most of the tracked variables in as little as 1 week with continued reduction in these variables over time.19 Another 2022 placebo-controlled trial, which combined rhodiola with B-vitamins, green tea, and magnesium, found that there was a decrease in perceived stress in both the variable and placebo groups on the DASS-42 stress questionnaire, but that the variable group showed a statistically significant decrease in scores when compared to placebo.20 These studies lend significant evidence to the ability of rhodiola to mitigate the stress response and reduce symptoms of burnout providing benefit as a self-care strategy to healthcare professionals and the public alike.
Several mechanisms for the stress-resiliency activity of rhodiola have been proposed. One mechanism looks at the activation of Heat Shock Protein 70, a stress-sensor and a reduction of Nitric Oxide (NO) via decreased NO synthase II.21 A reduction of Nitric Oxide may reduce physical fatigue and improve musculoskeletal endurance. Additionally, whole plant extracts of Rhodiola are thought to interact with the HPA Axis and limit the release of glucocorticoids, accounting for its use as an adaptogen to mitigate the stress response.22 These mechanisms help explain Rhodiola's effect on the stress response, and lend further credibility to its use as a self-care herb.
Regarding safety, rhodiola has been studied and shown to be well tolerated over a wide range of doses in several clinical trials based on few reported adverse outcomes. A 2017 systematic review saw rhodiola used from 50 mg to 660 mg per day in capsular form and up to 1500 mg with no reported adverse reactions as a single ingredient versus placebo.23 Further clinical trials on the efficacy of rhodiola had relatively low dropout rates and only mild reactions reported, such as nervousness and dizziness in the rhodiola group, lending evidence to the safety of the herbal product.24 , 25 With significant research available in the study of its efficacy and few reports of adverse reactions, rhodiola can easily be adapted as a regular self-care intervention to reduce perceived stress, fatigue, and plasma cortisol levels.
2.3 Passionflower
Passiflora incarnata is a flowering herb native to the Southeastern part of North America. This herb has been used prolifically by traditional herbalists for its nervine and mild sedative effects. Passionflower has been studied to determine the effects of the plant on mental stress and related disorders such as anxiety. A 2022 systematic review looked into the efficacy of Passionflower for stress reduction. While included studies were limited, the authors concluded that passionflower was an effective method of treating stress reactivity, anxiety and insomnia.26 The mechanism of action of the herb is not fully identified, and proposed mechanisms may differ depending on the condition treated. Most current research looks predominately on the GABAergic mechanisms with relation to its anxiolytic effects,26, 27, 28 with one article showing an affinity for acting on the hippocampus,29 which has been shown as a potential target area for stress reduction.30
There is limited data evaluating the safety of Passionflower in the last 10 years. The European Medicines Agency categorizes Passionflower as safe based on its years of traditional use with no reported side effects.31 A 2013 study examined the safety of Passionflower as part of a multiherb formula, revealing only mild adverse events from the combination of Passionflower with Valerian and Hops--with drowsiness the most commonly reported.32 A 2019 animal study concluded that Passionflower revealed no adverse effects in mice through chronic use of the herb.33 Despite the lack of adverse effects reported in these limited studies there are still a few cautions to consider. Passionflower does contain the flavonoid 5,7-dihydroxyflavone,34 more commonly called chrysin which has been shown to inhibit the enzyme aromatase, affecting the conversion of testosterone to estradiol,35 and prolonged use should be contraindicated in individuals with hormone-balance conditions such as Polycystic Ovary Syndrome (PCOS).
2.4 Lavender
A well-known aromatherapy herb with uses for insomnia, stress, and anxiety, Lavandula angustifolia has some published evidence supporting its use. Due to its availability and its ease of use, Lavender is a great resource for those wishing to practice herbal self-care. A 2018 clinical trial looked at the efficacy of lavender aromatherapy on 38 members of a nursing staff. The study reports a statistical reduction of heart rate and blood pressure, but no statistically significant reduction on psychological stress.36 Other studies examining the efficacy of lavender aromatherapy on stress levels found a measurable reduction in plasma and salivary cortisol levels in the aromatherapy groups.37, 38, 39 A 2014 meta-analysis noted some supportive evidence of aromatherapy on perceived stress levels, but no statistical reduction on salivary cortisol levels. While this meta-analysis included studies with combined aromatherapy, lavender was the most common herb included.40 Lavender has also been studied for its use in anxiety and insomnia. Evidence from clinical trials shows its efficacy in treating General Anxiety Disorder when compared to placebo,41 and a 2020 trial reported a reduction of anxiety on the Visual Anxiety Scale (VAS) in patients undergoing bone marrow biopsy.42 For insomnia, Lavender was effective in improving sleep quality, but no statistical difference was seen in Resting Eye Movement (REM) sleep time or total sleep time between the lavender and placebo groups.43 Other studies have combined lavender aromatherapy with sleep hygiene practices for improved benefit in sleep quality, vibrancy, and energy.44 , 45 Overall, the literature does indeed support its use as a self-care therapy, but the evidence is limited in quality with low sample sizes, and lack of quality placebo-controlled groups.
The mechanism for Lavender's effect is more difficult to ascertain than some of the previously mentioned herbs as there are a limited amount of studies on its physiological effect. A 2016 study examining the mechanisms for Lavender aromatherapy found no effect on the HPA axis, but measurable changes in Chromogranin A (CgA), an indicator for catecholamine levels, demonstrating a potential effect on the acute stress response via the sympathoadrenal medullary (SAM) pathway.46 Another study examining lavender aromatherapy during sleep, showed an increase in activity in the temporal lobe and increased delta-waves, which are associated with deep sleep, in the lavender group.47 Mechanism is particularly hard to study for aromatherapy due to the identifiable nature of the strong lavender smell, which was correctly identified by 50% of participants of one study,46 illustrating the difficulty in creating quality placebo-controlled studies.
The only study which examined the safety of oral administration of lavender oil for up to 10 weeks found lavender oil to be well tolerated in this capacity.41 As lavender is primarily used externally, it is not recommended for internal use. Skin irritation is the primary adverse effect from external use of Lavender essential oil.
While more quality placebo-controlled studies are necessary to verify the efficacy and safety in long term use, for these four botanicals, the body of current scientific evidence indicates that they are both safe and effective for use as a self-care strategy to address stress and symptoms of stress. More studies are also necessary to address the effectiveness specifically for healthcare workers, but most studies agree that these herbs can address general stress and related disorders such as anxiety, insomnia, and physical manifestations of stress. When combined with the limited reports of adverse events, these interventions could be used daily for short-term relief of stress and symptoms which may help to prevent or reduce feelings of burnout in frontline healthcare workers.
3 Instructions and recommendations for use
Recommendations for choosing the best preparation of an herb are to pick the method that is easy to find, meets the needs of the self-care practitioner, and is supported in the peer-reviewed evidence. Ashwagandha has studies showing its safety and efficacy at doses as high as 600 mg–1250 mg each day. Herbal products listed in milligrams (mg), would be best consumed as a dried herb or capsule as the quantity is easily measured. Some liquid and alcohol extracts will give the total amount in milligrams, but many do not provide this information. Most self-care practitioners will likely find it easier to obtain the herb in capsular form with a maximum dose of 600 mg per day for up to 8 weeks.16 It is best to follow the instructions for use from the manufacturer when possible and adhere to any specific warnings on the label. Avoid extended use of the herb beyond 8 weeks consecutively, as more safety information for long-term use is necessary.
Rhodiola is almost exclusively found on the market in capsular form, with dosages ranging from 200 mg–1000 mg easily available. It is noteworthy that most studies used dosages in the 200–300 mg per day range, and headaches were reported at doses as low as 200 mg.23 Those individuals using Rhodiola for self-care should start with lower dosages and discontinue use prior to 12 weeks, which was the maximum time of the longest study conducted. Prolonged use of Rhodiola is not recommended as long-term studies on adverse effects have not been conducted to assess safety.
Passionflower is readily available in liquid extract, tablet, and capsular forms. Studies on its efficacy saw most clinical research using lower dose 90 mg extracts of passaflamin a singular constituent of passionflower, 260 mg tablets of the whole herb, or 5 mL aqueous extracts--containing up to 700 mg of passionflower.26 Other monograph sources indicate a posology of 60 drops of fluid extract up to three times daily.48 None of the studies indicate usage longer than 6 weeks and use should be discontinued beyond this.
Lavender is usually found as an essential oil for use in aromatherapy. Specific dosage recommendations are less necessary with this type of preparation, as essential oils are recommended for external use. One concern with essential oils is avoiding direct contact with the skin, without the use of a carrier oil. Common carrier oils can include olive oil, coconut oil, and grapeseed oil. These oils help to dilute the essential oil and reduce the risk of skin irritation from direct contact. Other recommended uses include adding small amounts of lavender oil to bed linens, and clothing to bring about the desired effect. Long-term use of Lavender is safe if used topically with a carrier oil.
4 Strategies for safe use
As botanicals are common sources of pharmaceuticals, it is important to consider interactions with other medications in the use of these products. Individuals taking pharmaceutical prescriptions should consult their prescribing physician for any potential herb-drug interaction prior to use. Likewise, there is very limited information on the safety of botanicals in individuals who are pregnant and nursing. Due to limited information on safety, individuals who are pregnant or nursing should avoid almost all botanicals as a self-care option. Finally, individuals with allergies, especially severe reactions to other plants should consult a physician prior to using any new botanical medications.
While studies have shown the relative safety of the herbs reviewed, it is important to point out that botanicals do have potential side effects, and an inventory of these side effects is a useful tool for safe use of these herbs. Ashwagandha can sometimes cause gastrointestinal discomfort, diarrhea, and headaches. Due to its ability to effect circulating levels of cortisol, Ashwagandha can also cause sleepiness. While recent studies on safety show no effect on elevation levels of thyroid hormone,16 the potential to modulate thyroid hormone level is cause for ashwagandha to be contraindicated for individuals taking thyroid medications. Furthermore, ashwagandha has been seen in recent studies to increase immunoglobulin levels and cytokine activity49 and thus should be avoided for individuals with autoimmune disorders. Due to its effects on the immune response, individuals who may be undergoing surgical procedures should notify their surgeon of the use of this herb and follow physician recommended discontinuation guidelines prior to any surgical procedure.
While most studies support the safety of Rhodiola a few individuals reported adverse events in the experimentation groups including increased salivary production, headaches and insomnia.23 Likewise, Passionflower has a sedative effect and thus may cause drowsiness or mental fatigue with use. In higher doses of dried alcoholic extract, uncoordinated muscle movements have been reported.50 Therefore, recommendations for use of Passionflower are to avoid high doses and extended use to minimize the risk of adverse effects.
Additionally, herbal remedies may need more time than pharmaceuticals to build up to therapeutic levels in the blood stream and use of botanicals for self-care may not have immediate noticeable effects. Individuals experiencing a need for more immediate symptomatic relief should discuss available options with their primary care physician (PCP) prior to botanical use. Likewise, it is a good practice to consider discussion with a PCP prior to beginning any new botanical regiment.
5 Resources for continued use
There are many free and easily accessible online resources with evidence-based information for botanical medicines. The following resources provide more information on these and many other herbs:1. European Union Monographs (https://www.ema.europa.eu/en/medicines/field_ema_web_categories%253Aname_field/Herbal/field_ema_herb_outcome/european-union-herbal-monograph-254) – Published by the EMA, this website provides options to look up herbs by scientific binomial and English common name. The full monographs are limited to a smaller number of herbs, but users can utilize the search function and find basic information on herbs when a full monograph is not available.
2. Medline Plus Herbal Database (https://medlineplus.gov/druginfo/herb_All.html) This database, provides users with a short review of evidence, usage information and information on drug-herb interactions. Medline plus is an excellent resource for self-care practitioners to find efficacy and safety information on many herbs. Disadvantages are the limited number of herbs available in the database and categorization of botanicals by common name only.
3. Memorial Sloan-Kittering Cancer Center Integrative Medicine Database (https://www.mskcc.org/cancer-care/diagnosis-treatment/symptom-management/integrative-medicine/herbs/search?letter=A) Providing information about many herbs, supplements and integrative health interventions, this tool provides users with information on safety and efficacy. The website design includes fast links that show the mechanism of action of the botanical viewed when available.
4. NCCIH Health Topics Database (https://www.nccih.nih.gov/health/atoz#linkA) The National Center for Complimentary and Integrative Health (NCCIH) provides information links on a limited number of botanical medicines. This website provides users with information on many integrative therapies and is an excellent resource for information about safety of some herbs.
5.1 Key take-aways
• Stress and burnout are serious problems in the healthcare industry and new strategies are necessary to address the symptoms of stress, mitigate the body's response to stress and reduce burnout.
• Herbal medicines such as Ashwagandha, Rhodiola, Passionflower and Lavender provide a safe, effective and easy-to-use method for mitigating the stress response with extensive research for their use.
• Herbal medicines are readily available, and easy to use when using them within evidence-based recommended ranges
• Herbal medicines may interact with other pharmaceuticals and have little data on their safety for pregnant and nursing individuals. Caution is recommended in these groups.
• Herbal medicine online resources can provide quick access to evidence and safety information for the use of these herbs and others.
Acknowledgements
The author would like to thank Eric Rosenberg and Vincent Chen for their input on the rough draft of this article, and the reviewers for their suggested revisions.
==== Refs
References
1 International Classification of Diseases, Eleventh Revision (ICD-11), World Health Organization (WHO) 2019/2021 https://icd.who.int/browse11..
2 Wright T. Mughal F. Babatunde O. Dikomitis L. Mallen C. Helliwell T. Burnout among primary health-care professionals in low- and middle-income countries: systematic review and meta-analysis Bull World Health Organ 100 2022 385 401A 06 35694622
3 Shah M.K. Gandrakota N. Cimiotti J.P. Ghose N. Moore M. Ali M.K. Prevalence of and factors associated with nurse burnout in the US JAMA Netw Open 4 2 2021
4 McGuinness S.L. Johnson J. Eades O. Mental health outcomes in Australian healthcare and aged-care workers during the second year of the COVID-19 pandemic Int J Environ Res Publ Health 19 9 2022 4951
5 Zhang X. Song Y. Jiang T. Ding N. Shi T. Interventions to reduce burnout of physicians and nurses Medicine 99 26 2020 e20992
6 Søvold L.E. Naslund J.A. Kousoulis A.A. Prioritizing the mental health and well-being of healthcare workers: an urgent global public health priority Front Public Health 9 1 2021
7 Rashrash M. Schommer J.C. Brown L.M. Prevalence and predictors of herbal medicine use among adults in the United States J Patient Experien 4 3 2017 108 113
8 Speers A.B. Cabey K.A. Soumyanath A. Wright K.M. Effects of withania somnifera (ashwagandha) on stress and the stress-related neuropsychiatric disorders anxiety, depression, and insomnia Curr Neuropharmacol 19 2021
9 Lopresti A.L. Smith S.J. Malvi H. Kodgule R. An investigation into the stress-relieving and pharmacological actions of an ashwagandha (Withania somnifera) extract: a randomized, double-blind, placebo-controlled study Medicine 98 37 2019 e17186
10 Murthy S.V. Fathima S.N. Mote R. Hydroalcoholic extract of ashwagandha improves sleep by modulating GABA/histamine receptors and EEG slow-wave pattern in in vitro - in vivo experimental models Preventive Nutrition Food Sci 27 1 2022 108 120
11 Salve J. Pate S. Debnath K. Langade D. Adaptogenic and anxiolytic effects of ashwagandha root extract in healthy adults: a double-blind, randomized, placebo-controlled clinical study Cureus 11 12 2019
12 Chandrasekhar K. Kapoor J. Anishetty S. A prospective, randomized double-blind, placebo-controlled study of safety and efficacy of a high-concentration full-spectrum extract of ashwagandha root in reducing stress and anxiety in adults Indian J Psychol Med 34 3 2012 255 262 23439798
13 Langade D. Kanchi S. Salve J. Debnath K. Ambegaokar D. Efficacy and safety of ashwagandha (withania somnifera) root extract in insomnia and anxiety: a double-blind, randomized, placebo-controlled study Cureus 11 9 2019 e5797 Published online September 28 31728244
14 Gopukumar K. Thanawala S. Somepalli V. Rao T.S.S. Thamatam V.B. Chauhan S. Efficacy and safety of ashwagandha root extract on cognitive functions in healthy, stressed adults: a randomized, double-blind, placebo-controlled study Hughes C. Evidence-Based Complementary and Alternative Medicine 2021 2021 1 10
15 Raut A.A. Rege N.N. Tadvi F.M. Exploratory study to evaluate tolerability, safety, and activity of Ashwagandha (Withania somnifera) in healthy volunteers J Ayurveda Integr Med 3 3 2012 111 114 23125505
16 Verma N. Gupta S.K. Tiwari S. Mishra A.K. Safety of ashwagandha root extract: a randomized, placebo-controlled, study in healthy volunteers Compl Ther Med 57 2021 102642
17 Anonymous. Rhodiolae roseae rhizoma et radix - European Medicines Agency 2018 European Medicines Agency Published September 17 https://www.ema.europa.eu/en/medicines/herbal/rhodiolae-roseae-rhizoma-et-radix
18 Ivanova Stojcheva E. Quintela J.C. The effectiveness of rhodiola rosea L. Preparations in alleviating various aspects of life-stress symptoms and stress-induced conditions—encouraging clinical evidence Molecules 27 12 2022 3902 35745023
19 Kasper S. Dienel A. Multicenter, open-label, exploratory clinical trial with Rhodiola rosea extract in patients suffering from burnout symptoms Neuropsychiatric Dis Treat 13 2017 889 898
20 Noah L. Morel V. Bertin C. Effect of a combination of magnesium, B vitamins, rhodiola, and green tea (L-Theanine) on chronically stressed healthy individuals—a randomized, placebo-controlled study Nutrients 14 9 2022 1863 35565828
21 Grech-Baran M. Sykłowska-Baranek K. Pietrosiuk A. Approaches of Rhodiola kirilowii and Rhodiola rosea field cultivation in Poland and their potential health benefits Ann Agric Environ Med 22 2 2015 281 285 26094524
22 Panossian A. Wikman G. Sarris J. Rosenroot (Rhodiola rosea): traditional use, chemical composition, pharmacology and clinical efficacy Phytomedicine 17 7 2010 481 493 20378318
23 Ishaque S. Shamseer L. Bukutu C. Vohra S. Rhodiola rosea for physical and mental fatigue: a systematic review BMC Compl Alternative Med 12 1 2012
24 Lekomtseva Y. Zhukova I. Wacker A. Rhodiola rosea in subjects with prolonged or chronic fatigue symptoms: results of an open-label clinical trial Complem Med Res 24 1 2017 46 52
25 Mao J.J. Xie S.X. Zee J. Rhodiola rosea versus sertraline for major depressive disorder: a randomized placebo-controlled trial. Phytomedicine Inter J Phytother Phytopharmacol 22 3 2015 394 399
26 Janda K. Wojtkowska K. Jakubczyk K. Antoniewicz J. Skonieczna-Żydecka K. Passiflora incarnata in neuropsychiatric disorders—a systematic review Nutrients 12 12 2020 3894 33352740
27 Fonseca LR da Rodrigues R. de A. Ramos A. de S. Herbal medicinal products from passiflora for anxiety: an unexploited potential Sci World J 2020 2020 1 18
28 Aman U. Subhan F. Shahid M. Passiflora incarnata attenuation of neuropathic allodynia and vulvodynia apropos GABA-ergic and opioidergic antinociceptive and behavioural mechanisms BMC Compl Alternative Med 16 2016
29 Elsas S.M. Rossi D.J. Raber J. Passiflora incarnata L. (Passionflower) extracts elicit GABA currents in hippocampal neurons in vitro, and show anxiogenic and anticonvulsant effects in vivo, varying with extraction method Phytomedicine 17 12 2010 940 949 20382514
30 Snyder J.S. Soumier A. Brewer M. Pickel J. Cameron H.A. Adult hippocampal neurogenesis buffers stress responses and depressive behaviour Nature 476 7361 2011 458 461 21814201
31 Anonymous. Passiflorae herba - European medicines agency 2018 European Medicines Agency Published September 17 https://www.ema.europa.eu/en/medicines/herbal/passiflorae-herba
32 Maroo N. Hazra A. Das T. Efficacy and safety of a polyherbal sedative-hypnotic formulation NSF-3 in primary insomnia in comparison to zolpidem: a randomized controlled trial Indian J Pharmacol 45 1 2013 34 39 23543804
33 Kim G.H. Yi S.S. Chronic oral administration of Passiflora incarnata extract has no abnormal effects on metabolic and behavioral parameters in mice, except to induce sleep Laboratory Animal Research 35 1 2019
34 Rodríguez-Landa J.F. German-Ponciano L.J. Puga-Olguín A. Olmos-Vázquez O.J. Pharmacological, neurochemical, and behavioral mechanisms underlying the anxiolytic- and antidepressant-like effects of flavonoid chrysin Molecules 27 11 2022 3551 35684488
35 Balam F.H. Ahmadi Z.S. Ghorbani A. Inhibitory effect of chrysin on estrogen biosynthesis by suppression of enzyme aromatase (CYP19): a systematic review Heliyon 6 3 2020 e03557
36 Montibeler J. Domingos T. da S. Braga E.M. Gnatta J.R. Kurebayashi L.F.S. Kurebayashi A.K. Efetividade da massagem com aromaterapia no estresse da equipe de enfermagem do centro cirúrgico: estudo-piloto Rev Esc Enferm USP 52 2018 0
37 Lee A. Cho H. The effects caused by lavender and rosemary for salivary cortisol, stress levels and mood alteration J Fashion Business 17 6 2013
38 Atsumi T. Tonosaki K. Smelling lavender and rosemary increases free radical scavenging activity and decreases cortisol level in saliva Psychiatr Res 150 1 2007 89 96
39 Heydari A. Hosseini S. Vakili M. Moghadam S. Tazyky S. Effect of lavender essence inhalation on the level of anxiety and blood cortisol in candidates for open-heart surgery Iran J Nurs Midwifery Res 21 4 2016 397 27563324
40 Hur M.H. Song J.A. Lee J. Lee M.S. Aromatherapy for stress reduction in healthy adults: a systematic review and meta-analysis of randomized clinical trials Maturitas 79 4 2014 362 369 10.1016/j.maturitas.2014.08.006 25234160
41 Kasper S. Gastpar M. Müller W.E. Lavender oil preparation Silexan is effective in generalized anxiety disorder – a randomized, double-blind comparison to placebo and paroxetine Int J Neuropsychopharmacol 17 2014 859 869 06 24456909
42 Abbaszadeh R. Tabari F. Asadpour A. The effect of lavender aroma on anxiety of patients having bone marrow biopsy Asian Pac J Cancer Prev APJCP: APJCP. 21 3 2020 771 775 32212806
43 dos Reis Lucena L. dos Santos-Junior J.G. Tufik S. Hachul H. Lavender essential oil on postmenopausal women with insomnia: double-blind randomized trial Compl Ther Med 59 2021 102726
44 Lillehei A.S. Halcón L.L. Savik K. Reis R. Effect of inhaled lavender and sleep hygiene on self-reported sleep issues: a randomized controlled trial J Alternative Compl Med 21 7 2015 430 438
45 Lillehei A.S. Halcón L. Gross C.R. Savik K. Reis R. Well-being and self-assessment of change: secondary analysis of an RCT that demonstrated benefit of inhaled lavender and sleep hygiene in college students with sleep problems EXPLORE 12 6 2016 427 435 27659004
46 Chamine I. Oken B.S. Aroma effects on physiologic and cognitive function following acute stress: a mechanism investigation J Alternative Compl Med 22 9 2016 713 721
47 Ko L.W. Su C.H. Yang M.H. Liu S.Y. Su T.P. A pilot study on essential oil aroma stimulation for enhancing slow-wave EEG in sleeping brain Sci Rep 11 1 2021
48 Godfrey A. Saunders Paul Richard Principles & Practices of Naturopathic Botanical Medicine 2010 Canadian College of Naturopathic Medicine Press
49 Tharakan A. Shukla H. Benny I.R. Tharakan M. George L. Koshy S. Immunomodulatory effect of Withania somnifera (ashwagandha) extract-A randomized, double-blind, placebo controlled trial with an open label extension on healthy participants J Clin Med 10 16 2021 3644 Published 2021 Aug 18 34441940
50 Passionflower National Center for Complimentary and Integrative Health August 2020 https://www.nccih.nih.gov/health/passionflower
| 0 | PMC9737923 | NO-CC CODE | 2022-12-14 23:45:34 | no | J Interprof Educ Pract. 2023 Mar 10; 30:100592 | utf-8 | J Interprof Educ Pract | 2,022 | 10.1016/j.xjep.2022.100592 | oa_other |
==== Front
Intern Emerg Med
Intern Emerg Med
Internal and Emergency Medicine
1828-0447
1970-9366
Springer International Publishing Cham
3167
10.1007/s11739-022-03167-7
Im - Case Record
A rare case of abdominal lymphadenopathy and fever
http://orcid.org/0000-0001-5745-7202
Gallo Antonella [email protected]
1
Macerola Noemi 1
Ibba Francesca 1
Contegiacomo Andrea 1
Montalto Massimo 12
1 grid.414603.4 Fondazione Policlinico Universitario A. Gemelli, IRCCS, Largo A. Gemelli 1, 00168 Rome, Italy
2 grid.8142.f 0000 0001 0941 3192 Università Cattolica del Sacro Cuore, Rome, Italy
10 12 2022
14
27 8 2022
29 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.
==== Body
pmcCase presentation
Dr. Macerola: A European 52-year-old man referred to our Emergency Department (ED) for high fever with chills, arthralgia of the small joints and sweating; these symptoms developed 1 month before the adm1ission. He did not take any chronic medication. In the last month he took paracetamol and, sometimes, non-steroidal anti-inflammatory drugs (NSAIDs), with partial and non-lasting benefit. His medical record was only notable for a previous episode characterized by the same symptoms, occurred about twenty years before. In that occasion, a whole body computed tomography (CT) scan only revealed a mild splenomegaly (diameter 15.5 cm) and enlarged lymph node (diameter 1.5 cm) along hepatic artery tract. A bone marrow biopsy was negative for significant alteration. Microbiological and immunological exams were all negative. Symptoms lasted a few months and slowly resolved without a conclusive diagnosis and any specific therapy.
On arrival at the ED, the patient complained of fever (the axillary temperature was 39.5°) and arthralgia; the other vital signs were normal. Physical examination revealed neither palpable masses nor hepatosplenomegaly.
A complete blood count showed mild anemia (9.5 g/dL) and thrombocytosis (417 × 10^9/L), a white blood cell (WBC) absolute count of 11.62 × 10 ^9/L, with neutrophil leukocytosis (81%). The lactate dehydrogenase level was within the normal range. C-reactive protein was significantly increased (197.5 mg/L, reference interval < 5 mg/l), serum ferritin levels were within normal range (261 µg/ml, reference interval 21–275 µg/ml). Nasopharyngeal swab for SARS-Cov-2 tested negative. A chest X-ray was negative.
The patient was admitted to our Internal Medicine Division for further diagnostic evaluations. During hospitalization, he remained hemodynamically stable with a good compensation of oxyhemoglobin in ambient air. He continued to have fever and arthralgia without clear signs of arthritis. The laboratory test confirmed the presence of anemia (Hb 10 g/dl) associated with thrombocytosis (620 × 10^9/L), high levels of C-reactive protein (243 mg/L); the procalcitonin value was within normal range.
Serum electrolytes and blood levels of transaminases, alkaline phosphatase, bilirubin, were normal (Table 1).Table 1 Laboratory data upon arrival and at discharge from hospital
Variable Reference range, adults On arrival At discharge
Hemoglobin (g/dl) 13–17 9.9 8.9
Hematocrit (%) 40–50 29.3 28.1
Mean corpuscular volume (fl) 83–101 85.8 84.4
White-cell count (per μl) 4–10 10.33 6.2
Neutrophils (%) 40–80 76.6 81.6
Lymphocytes (%) 20–40 15.3 10.3
Monocytes (%) 0.2–1 0.56 0.71
Eosinophils(%) 0–5.4 2.3 0.4
Platelet count (per μl) 150–450 555 447
Sodium (mmol/l) 135–145 139 135
Potassium (mmol/l) 3.5–5 4.3 4.2
Chloride (mmol/l) 98–108 103
Calcium (mg/dl) 8.6–10.2 8.9 8.3
Magnesium (mg/dl) 1.8–2.4 2.2
Urea nitrogen (mg/dl) 10–23 11 13
Creatinine (mg/dl) 0.67–1.17 0.75 0.89
Glucose (mg/dl) 65–100 100 91
Alanine aminotransferase (IU/l) 7–45 39
γ-glutamiltransferase (IU/l) 8–61 110 82
Alkaline phosphatase (IU/l) 40–129 94
Bilirubin (total) (mg/dl) 0.3–1.2 0.4
Albumin (g/dl) 34–48 24
Creatine kinase (IU/l) 30–170 24
Lactate dehydrogenase (IU/l) < 250 159
C-reactive protein (mg/dl) < 5 243.4 78
Procalcitonine < 0.05 0.24 0.09
International normalized ratio 0.8–1.2 1.07
APTT ratio 20–38 34
Serum iron (μg/dl) 60–160 17
Ferritin (μg/l) 21–275 261
Reference ranges used at the IRCCS Fondazione Policlinico Universitario Agostino Gemelli (Roma) are for adults who are not pregnant and do not have medical conditions that could affect the result
Further investigation and differential diagnosis
Prof. Montalto: The differential diagnostic process was immediately articulated through the evaluation of infectious (bacterial, viral, mycotic, parasitic), malignant and immune mediated disorders.
About a possible infectious etiology, multiple blood cultural tests were performed, both in the ED and during the hospitalization, all of them resulting sterile. The urine test did not show any significant alteration; urine culture showed the presence of E.faecalis but, in the absence of symptoms and non-significant load (10,000 CFU/ml), antibiotic therapy was not considered appropriate.
Serologies for hepatitis (HBV, HCV), human immunodeficiency virus, syphilis, Citomegalovirus, Parvovirus B19, Leishmania were negative, those for Epstein-Barr virus (EBV) and Toxoplasmosis were consistent with a previous infection and serum EBV genome was not significant. Serologies for pneumotropic agents documented a previous infection of Chlamydia and Mycoplasma pneumoniae; the cryptococcal antigen was negative.
Quantiferon test was undetermined. Widal-Wright and Weil-Felix were negative, as well the serum Beta-D-glucan.
The echocardiogram did not show any evidence of endocarditic vegetation.
Dr. Gallo: Once the most likely infective causes were ruled out, the patient was examined for immune mediated disorders. Electrophoretic control of serum proteins was consistent with active inflammation (mild hypoalbuminemia, high levels of alpha-1 and alpha-2 globulins). Serum immunofixation was normal, as well the individual classes of immunoglobulin.
An autoimmunity panel was performed, including search for antinuclear antibodies (ANA), extractable nuclear antigen antibodies (ENA; SSA/Ro, SSB/La, Sm, RNP), Anti SCL-70 antibodies, anti -smooth-muscle antibodies (ASMA), anti-neutrophil cytoplasmic antibodies (p-ANCA, c-ANCA), anti-mitochondrial antibodies (AMA), anti-citrulline antibodies. All these antibodies resulted negative. Also the C3 and C4 complement fractions resulted in normal ranges and the rheumatoid factor (RF)was negative. Lupus anticoagulant, anti-cardiolipin antibodies IgM and IgG were absent. The research of circulating immunocomplexes was also negative. We investigated also IgG subclasses (with particular reference to IgG4) that were normal.
Based on clinical picture, an adult onset Still’s disease (AOSD) diagnosis also resulted worthy of consideration. Yamaguchi’s criteria were used [1]; at first, in absence of radiological data, our patient already showed the presence of three major criteria (fever > 39 °C, arthralgia and neutrophil leukocytosis) and one minor criteria (negative ANA and RF). A whole body scan CT was therefore planned to investigate about lymphadenopathy/splenomegaly, which could have represented the second minor criteria, but also to rule out about an eventual malignant disease.
Prof. Montalto: The whole body CT scan showed slight enlarged lymph-nodes along the initial tract of the lower mesenteric artery and the celiac tripod (maximum diameter of about 2–3 cm) (Fig. 1a–c). No other relevant data were reported.Fig. 1 Axial (a, b) and sagittal (c) contrast-enhanced CT scans show the presence of a hypodense infiltrating tissue (arrow) that surrounds the celiac trunk, without clear signs of vascular invasion. The PET-CT study (d) shows an intense uptake (SUV max = 8.7) of 18-FDG (arrowhead). H&E staining of macrophages (e) showing emperipolesis of lymphocytes (arrow), plasma cells and granulocytes in a case of Rosai-Dorfman disease (200x)
The value of the main onco-markers was negative; beta-2-microglobulin was only slightly increased (3.4 microgr/L, reference interval < 3.2).
Finally, no significant alterations were documented by esophagogastroduodenoscopy and colonoscopy.
Although the presence of lymphadenopathies supported the hypothesis of an AOSD diagnosis, such a radiological picture seemed more likely to be expression of a lymphoproliferative disorder. Moreover, although it does not represent diagnostic criteria for AOSD, our serum ferritin levels were relatively low, that is quite unusual for a Still’s disease [2].
This diagnostic uncertainty imposed, before starting steroid therapy, a further radiological investigation by a positron emission tomography with 2-deoxy-2-[fluorine-18] fluoro-D-glucose integrated with CT (18F-FDG PET/CT) was performed.
Dr. Contegiacomo: The 18F-FDG PET/CT revealed a significant enhanced uptake of multiple lymph-nodes at the sub-diaphragmatic level, without large vessels involvement (Fig. 1d). Also this picture resulted almost unusual for the AOSD, conversely characterized by a diffusely increased FDG uptake in the spleen and bone marrow with multiple reactive hyperplasia lymph nodes symmetrically mainly distributed in the neck and axilla [3].
A lymphoprofilerative disorder (Hodgkin’s lymphoma, Castleman’s disease) became therefore the most likely diagnosis, however, deserving a histological characterization. At first, a minimally invasive approach was attempted by an echoendoscopy, which, however, was unsuccessfull because of the low visibility of the structures described at the CT.
A laparotomy was therefore performed for diagnostic purpose, collecting lymphadenomegalies at the root of the mesentery. The lymph node specimen comprised an intact, encapsulated, of 2.5, 1.2, 0.5 cm. The cytofluorometry performed on an operating sample was negative for lymphocytic B clonality. IgG4 immunohistochemical stain did not show a IgG4-positive plasma cells infiltration. The sample was subjected to a bacterioscopic examination, negative for resistant alcohol-acid forms; the cultural examination for Mycobacterium Spp. was negative.
On microscopic examination, it showed sinusoidal dilation with individual and aggregated, moderate to large histiocyte, emperipolesis of lymphocytes, plasma cells and granulocytes. This picture resulted suggestive for a Rosai–Dorfman syndrome (RDS) (Fig. 1e).
Prof: Montalto: After surgery, all symptoms quickly resolved without need for any other specific drug- Few days later, the patient recovered completely and he was discharged.
The 12 months follow-up showed the persistent resolution of clinical picture.
Discussion
Dr. Ibba: Rosai-Dorfman disease (RDD) is a rare histiocytic disorder; about 1000 cases have been reported in medical literature since 1969, when it was firstly described by Rosai and Dorfman. RDD is a heterogeneous entity that can occur as an isolated disorder or in association with autoimmune, hereditary, and malignant diseases.
Classic RDD occurs in young patients, usually in the first or second decades of life, with massive bilateral painless cervical lymphadenopathy with associated fever, loss of weight and night sweats. However, the disease can affect any organ system, and inguinal, and mediastinal lymph nodes may be involved. Retroperitoneal lymphadenopathy is not common [4, 5].
Dr. Gallo: Since the variegate and sometimes unspecific clinical and radiological picture, as in our case, differential diagnosis includes different disorders, from Hodgkin’s lymphoma and Whipple disease, to Langerhans cell histiocytosis, IgG4-related disease to AOSD [4].
RDD is usually self-limited and has a good outcome, with spontaneous remission within few months reported in up to 50% of cases. However, patients may experience different courses of disease, ranging from reappearance of symptoms after many years of full wellness, to stability of disease over the years, mainly in the extra-nodal involvement, to no full clinical remission at all [4].
Up to 10% of patients may experience direct complications of the disease, such as infection or amyloidosis, potentially leading to death [4].
Dr. Macerola: RDD represents a rare and heterogeneous disorder, still deserving many diagnostic and therapeutic challenges. Treatment strategies range from observation alone to immunomodulatory and antineoplastic agents to surgery.
Steroids are usually indicated to reduce nodal size and improve symptoms, although responses are quite variable and an optimal duration of therapy has not been recognized yet.
Surgery in RDD is usually considered for diagnostic purposes, even if resection (and debulking) is necessary in case of airway obstruction, spinal cord compression, or large lesions causing end-organ compromise. Moreover, surgery has also been reported as curative in unifocal disease, mainly in isolated intracranial disease [6].
Prof. Montalto: RDD, as other rare histiocytic disorders, may present with nonspecific but very disabling symptoms. Getting to the final diagnosis may be challenging, and it is crucial to carry out a complete differential diagnosis between infectious, malignancy and immune mediated disorders. It is also fundamental to collect a careful history, which in our case was significant for a previous episode with the same symptoms about twenty years before. After a comprehensive evaluation involving physical examination, blood tests and imaging, the conclusive diagnosis is usually histological [7]. In our patient, we decided to perform a slight surgical excision, and not only a biopsy for diagnostic purpose; it is likely that this strategy may have significantly contributed to achieve an immediate clinical success and long-term remission.
It is hopeful that increased knowledge of these rare diseases may help all clinicians facing with such intricate cases, in performing an accurate diagnostic algorithm process, also comprehensive of evaluation about the best procedures to urndergo.
Declarations
Conflict of interest
The authors have no conflict of interest.
Human and animal rights
The patient was treated following the ethical standards and cannot be identified in text or images.
Informed consent
The patient signed an informed consent and cannot be identified in the text or images.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Yamaguchi M Ohta A Tsunematsu T Preliminary criteria for classification of adult Still’s disease J Rheumatol 1992 19 424 430 1578458
2. Bagnari V Colina M Ciancio G Adult-onset Still’s disease Rheumatol Int 2010 30 855 862 10.1007/s00296-009-1291-y 20020138
3. Zhou X Li Y Wang Q FDG PET/CT used in identifying adult-onset Still’s disease in connective tissue diseases Clin Rheumatol 2020 39 2735 2742 10.1007/s10067-020-05041-3 32180040
4. Cohen Aubart F Haroche J Emile JF La maladie de Destombes-Rosai-Dorfman : évolution du concept, classification et prise en charge [Rosai-Dorfman disease: diagnosis and therapeutic challenges] Rev Med Interne 2018 39 635 640 10.1016/j.revmed.2018.02.011 29501513
5. Karami R Ghieh F Baroud J Rosai-Dorfman disease: cutaneous and parotid involvement Ann Plast Surg 2019 82 639 641 10.1097/SAP.0000000000001794 30882409
6. Bruce-Brand C Schneider JW Schubert P Rosai-Dorfman disease: an overview J Clin Pathol 2020 73 697 705 10.1136/jclinpath-2020-206733 32591351
7. Abla O Jacobsen E Picarsic J Consensus recommendations for the diagnosis and clinical management of Rosai-Dorfman-Destombes disease Blood 2018 131 2877 2890 10.1182/blood-2018-03-839753 29720485
| 36495420 | PMC9739337 | NO-CC CODE | 2022-12-14 23:29:54 | no | Intern Emerg Med. 2022 Dec 10;:1-4 | utf-8 | Intern Emerg Med | 2,022 | 10.1007/s11739-022-03167-7 | oa_other |
==== Front
Chromatographia
Chromatographia
Chromatographia
0009-5893
1612-1112
Springer Berlin Heidelberg Berlin/Heidelberg
4226
10.1007/s10337-022-04226-z
Original
Identification, Isolation, and Structural Characterization of Novel Forced Degradation Products of Darunavir Using Advanced Analytical Techniques Like UPLC–MS, Prep-HPLC, HRMS, NMR, and FT-IR Spectroscopy
Modini Arun Kumar 12
Ranga Mahesh 12
Puppala Umamaheshwar 1
Kaliyapermal Muralidharan 1
Geereddy Mahesh Kumar Reddy 1
Samineni Ramu 3
Grover Parul 4
Konidala Sathish Kumar [email protected]
3
1 Analytical Discovery Chemistry, Aragen Life Sciences Pvt. Ltd., IDA Nacharam, Hyderabad, 522017 India
2 grid.449932.1 0000 0004 1775 1708 Department of Chemistry, School of Applied Sciences and Humanities, Vignan’s Foundation for Science Technology and Research (Deemed to Be University), Vadlamudi, Guntur, 522213 Andhra Pradesh India
3 grid.449932.1 0000 0004 1775 1708 Department of Pharmaceutical Sciences, School of Biotechnology and Pharmaceutical Sciences, Vignan’s Foundation for Science Technology and Research (Deemed to Be University), Vadlamudi, Guntur, 522213 Andhra Pradesh India
4 KITE School of Pharmacy, KITE Group of Institutions, Delhi-NCR, Ghaziabad, 201206 India
10 12 2022
116
10 10 2022
18 11 2022
30 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.
Since the stability of the pharmaceuticals plays a crucial role in efficacy and safety while using them in the treatment of disorders, the evaluation of purity and impurity profiling of pharmaceuticals is of utmost importance using efficient analytical techniques. The present study explains the identification, isolation, and characterization of stress degradation products of the anti-human immunovirus drug Darunavir. The degradation study was performed to evaluate the stability profile of Darunavir in different stress conditions like hydrolytic, oxidative, thermal, and photolytic conditions as per the ICH guidelines. Degradation products were identified using ultra-performance liquid chromatography coupled with mass spectrometry, isolated using semi-preparative high-performance liquid chromatography, and structural characterization by HRMS and 1H, 13C NMR (1D, 2D). Darunavir is relatively stable in oxidative, thermal, and photolytic conditions; however, considerable degradation was observed in acid and base hydrolysis. A total of five degradation products were identified and isolated in acid and base degradation. DP-1, DP -2, & DP-3 were observed in acid conditions, whereas in base conditions, along with DP-2, two more DPs, i.e., DP-4 & DP-5, were identified. Among the five DPs, two degradation products, namely DP-1: N-(4-(N-(3-amino-2-hydroxy-4-phenylbutyl)-N-isobutylsulfamoyl) phenyl) acetimidamide. & DP-3: hexahydrofuro[2,3-b]furan-3-yl(4-((4-acetimidamido-N-isobutylphenyl)sulfonamido)-3-hydroxy-1-phenylbutan-2-yl)carbamate, are novel, remaining degradation products DP-2: 4-amino-N-(3-amino-2-hydroxy-4-phenylbutyl)-N-isobutylbenzenesulfonamide, DP-4: 4-amino-N-(((5S)-4-benzyl-2-oxooxazolidin-5-yl) methyl) -N-isobutyl benzenesulfonamide and DP-5: methyl ((3S)-4-((4-amino-N-isobutylphenyl) sulfonamido)-3-hydroxy-1-phenylbutan-2-yl) carbamate are already reported tentatively using a single analytical technique coupled with mass analysis without any evidence from NMR and IR data. Hence, the present study focused on using High-Resolution Mass, 1D, and 2D 1H, 13C NMR data for concrete confirmation of structures for degradation products.
Supplementary Information
The online version contains supplementary material available at 10.1007/s10337-022-04226-z.
Keywords
Darunavir
Force degradation studies
UPLC–MS
HRMS
NMR
==== Body
pmcIntroduction
The purity of pharmaceuticals is critical when using them to treat various ailments since the existence of unrelated products or impurities that were may arise during the manufacturing, storage, or transportation of these pharmaceuticals and can cause major harmful or adverse effects. For instance, the acid hydrolysis degradation products of tetracycline produce reversible Fanconi syndrome (a disorder of kidney tubes), thermal degradation products of Isoniazid produce genotoxicity and carcinogenicity, and oxidative and photolytic degradation product of Ibuprofen is cytotoxic to fibroblast, etc. [1]. The monitoring of impurities for their acceptable limits is a challenging task since the availability of the methods may or may not be sensitive or selective for the separation and identification of possible impurities present along with the drug substance. The utilization of antiviral drugs is drastically increased against different viral infections that were emerging currently days including COVID-19. Darunavir (DAR) is an antiretroviral drug from the protease inhibitor class, which is used to treat infection of human immunodeficiency virus and also recently used in the treatment of COVID-19 along with cobicistat drug. It selectively inhibits the cleavage of HIV-encoded Gag-polyproteins in virus-infected cells, thereby preventing the formation of mature infectious virus particles [2].
DAR is marketed in ethanolate form under the brand name Prezista, in different dose strengths, such as 75, 150, 300, 400, and 600 mg. In addition, Prezista is available as an oral suspension with a 100 mg/mL dose strength [3]. Its molecular formula is C27H27N3O7S (Mol. Wt.: 547.67 mol/gm) and the chemical name is (3R,3aS,6aR)–hexahydrofuro [2,3-b] furan-3-yl (4-((4-amino-N-isobutylphenyl) sulfonamido)-3-hydroxy-1-phenylbutan-2-yl) carbamate. DAR appears as a white, non-hygroscopic fine powder with a melting range of 111–115 °C.
The literature scrutiny revealed that few analytical methods were reported for the separation, identification, and estimation of related substances of DAR [4–7], DPs [2, 8], and for the estimation of the DAR in bulk and dosage forms along with other drugs [9–13]. LC–MS/MS studies on metabolites rat serum metabolism, and pharmacokinetics after administration of the drug were reported [14, 15].
The reported methods for separation, identification, and estimation of related substances and DPs of DAR were proposed structures of DPs of DAR only based on mass fragmentations, and none of them reported on the isolation, characterization, and complete structural conformation of DPs of DAR using any high-end analytical technique like nuclear magnetic resonance (NMR) spectroscopy (1D, and 2D), high-resolution mass spectrometry, and IR spectroscopy. Hence, the present research endeavored to provide concrete evidence for the structures of the DPs of DAR, which includes the UPLC–MS method development for efficient separation and estimation of DAR and its five DPs, along with the characterization of the separated and isolated DPs by HRMS, IR, and NMR (1D, 2D) experiments.
Materials and Methods
Chemicals and Reagents
Darunavir is a gifted sample from one of the pharmaceutical organizations in Hyderabad, water purified by Millipore (Millipore, Netherlands) was used. Analytical grade reagents and solvents were used. Acetonitrile, methanol, and formic acid, which are HPLC grade purchased from Merck India Ltd. Dimethyl sulfoxide-d6 was purchased from Cambridge isotope laboratories, hydrochloric acid, sodium hydroxide, hydrogen peroxide, which are AR grade purchased from Merck India Ltd.
Instrumentation and Software
The analytical instruments and supported software used in the present research work are LC–MS instruments manufactured by Waters, single quadrupole mass detector coupled with Acquity UHPLC front end, and Maslynx 4.2 software. Purification instrument was from Waters binary module 2545, auto-sampler 2707 with detector-2489 and Chromscope-2.1 software. HRMS instrument from Thermo Q-exactive Orbitrap MS with ESI ion source and Dionex ultimate 3000 LC front end was supported by Xcalibur software. NMR instrument was from Bruker Avance Neo 400 MHz and Topspin 4.11 software. Infrared spectroscopy instrument was from Shimadzu IR-Afinity-1S supported by lab solution software and analytical balance from Sartorius-SQP-F.
Ultra-Performance Liquid Chromatography–Mass Spectrometry (UPLC–MS)
Liquid chromatography separation was performed on Waters make Single Quadrupole Detector (SQD2) coupled with Acquity UPLC & Diode array detector (DAD) front end. Waters SQ Detector-2 single quadrupole mass spectrometer operating in dual polarity-positive and negative with electrospray ionization source (ESI) is used for mass spectral analysis. Full-scan mode 100–1500 Daltons (Da) was used for the MS optimization. The capillary voltage and source temperature were set to 3.5 kV and 140 °C, respectively. The desolvation temperature and gas flow were set as 350 °C and 650 L/h respectively, while the cone gas flow was set at 50 L/h. The liquid chromatography–mass spectrometer instrument was controlled by the Mass Lynx V4.2 application manager. The injection volume was 0.5 µL with sample manager temperature maintained at 18 °C and the chromatographic runtime of 7.0 min.
UPLC–MS Method Development (Optimization of Chromatographic Conditions)
In pursuit of a suitable, reproducible, and reliable method, different UPLC columns using different buffers were screened to optimize the resolution. During copious trials, 0.05% formic acid buffer and 0.05% formic acid acetonitrile with Acquity CSH C18 showed positive and encouraging results. Though API and 4 DPs (DP-1, DP-2, DP-3, & DP-5) were resolved in Acquity CSH C18 50 × 2.1 mm, 1.7 µ, DP-4 was not separated. Further, extensive trials were made on Acquity; BEH C18 column with various flow rate and gradient conditions. The optimum separation of all degradation impurities and API was achieved on Acquity BEH C18 100 × 2.1 mm, 1.7 µ, with 0.05% Formic acid in aqueous and 0.05% Formic acid acetonitrile, flow rate- 0.3 mL/min; column temperature-35 °C; with Binary gradient Time/ B conc. (%): 0/20, 0.5/20, 3/40, 4.5/40, 6.0/90, 7/90, 7.1/10. All 6 peaks were separated with good resolution and peak shape in this condition.
Optimized Chromatographic Conditions
The chromatographic condition used was Acquity UPLC BEH C18 column (100 × 2.1 mm, 1.7 µm), which was procured from Waters India Ltd. The mobile phase components were mobile phase-A: 0.05% formic acid in aqueous, mobile Phase-B: acetonitrile, separation was accomplished in a gradient elution program {Time (min)/B conc (%): 0/20, 0.5/20, 3/40, 4.5/40, 6/90, 7/90 at a flow rate of 0.3 mL/ min, Column temperature 35 °C. The chromatographic eluents were monitored using a photodiode array (PDA) detector, and the injection volume of 0.5 µL was used for sample injections. Water and acetonitrile in the ratio of 50:50 (v/v) were used as a diluent. The typical chromatogram is shown in in Fig. 1.Fig. 1 Chromatogram of Darunavir and its Degradation Products
High-Resolution Mass Spectrometry (HRMS)
Samples were analyzed on Thermo Q Exactive orbitrap MS with ESI ion source. The front end is UHPLC Dionex Ultimate 3000 instrument comprising the binary pump, column manager, and PDA detector. Instrument parameters (source) were Spray Voltage: 3200 V; Capillary Temperature: 300 °C; Aux gas flow rate: 14; Aux gas heater Temperature: 440 °C. Sheath gas flow rate: 50; Sweep gas flow rate: 3. Reserpine (monoisotopic mass: 608.2734 Da) was used to check the mass accuracy. Mass data were acquired using Xcalibur software.
Preparative HPLC
Waters preparative HPLC equipped with ChromScope-2.1 software, 2545 pump module, 2489 dual UV detector, and 2707 sample manager with auto-fraction collector-III. In-house packed Luna C18 10µ, 150 × 25 mm (Phenomenex) was used to isolate the degradation products, with a flow rate of 18 mL/min and mobile phase 0.1% v/v formic acid in aqueous and acetonitrile. All isolated fractions were lyophilized using a Lyofreeze lyophilizer.
Nuclear Magnetic Resonance Spectroscopy (NMR)
1H, 13C, and 2D NMR spectra of Darunavir and degradation products were recorded in DMSO-d6 solvent on Bruker Avance Neo 400 MHz NMR instrument equipped with 5 mm i- probe (Broad Band Obserbence) with Z-gradient shim system, which has the sensitivities of 550:1 and 220:1. The 1H NMR spectra referenced tetra-methyl-silane (TMS) singlet at zero (0) ppm and referenced DMSO-d6 septet at 39.5 ppm in the 13C NMR spectrum.
FT-IR Spectroscopy
Shimadzu IR-Affinity-1S model with Lab solutions software was used to know the functional groups present in the compounds. KBr was used as a medium to make the sample pellets.
Degradation Behavior of DAR
DAR was subjected to various forced degradation stress conditions, and the compound showed highly stable in oxidative, thermal, and photolytic conditions and did not form any degradation products, which confirms the stability of DAR toward the above conditions. The drug was found to be labile to acid and base hydrolysis in the presence of Acetonitrile and Methanol. As a result, ~ 15–20% of degradation was observed in both acid (1 N HCl with stirring at room temperature, up to 48 h) and base (1 N NaOH with stirring at room temperature, up to 36 h) conditions. Detailed degradation conditions [16–21] and results are displayed in Table 1, and degradation chromatograms of acid and base hydrolysis are shown in Fig. 2. In acid condition, 3 degradation products were formed (DP-1, DP-2, & DP-3), whereas in base conditions, along with DP2, 2 more additional degradation products were developed (DP-4 & DP-5). The confirmed structures of all the DPs are shown in Fig. 3. The rate of degradation was influenced by temperature. Increasing the temperature increased the rate of degradation, and the percentage of DPs formed also increased.Table 1 Darunavir forced degradation studies
Conditions DP-1 DP-2 DP-3 DP-4 DP-5 API
Darunavir API (ACN + MeOH) – – – – – 99
Acid (1 N HCl stirring at rt up to 48 h) (ACN + MeOH) 8 5 6 – – 81
Base (1 N NaOH stirring at rt up to 36 h) (ACN + MeOH) – 6 – 8 6 80
Oxidation (30% H2O2 stirring at rt up to 48 h) (ACN + MeOH) – – – – – 98
Thermal (explore to 100 °C up to 48 h) (ACN + MeOH) – – – – – 98
Photolytic (explore at 254 nm for 48 h) (ACN + MeOH) – – – – – 98
Fig. 2 Degradation behavior of Darunavir in Acid & Base
Fig. 3 Structure of Darunavir and its degradation products
Preparation of Degradation Samples for Purification
Degradation was observed in acidic and base conditions. The acid-degraded sample was neutralized with 5 N sodium bicarbonate solution, and the resultant solution was lyophilized to get a crude solid sample. The same was dissolved in 4–5 mL of Acetonitrile: Water (1:1 v/v). For base degradation, the degraded sample was neutralized with 5 N HCl solution, and the resultant solution was lyophilized to get a free solid. The same sample was dissolved in 4–5 mL of Methanol: water (1:1 v/v) for preparative HPLC purification.
Results and Discussion
Using liquid chromatography–mass spectrometry, individual samples were analyzed to determine the results of all the stress studies. All degradation products formed over a period, and from each constraint, study conditions are outlined in the experimental section. Five significant degradation products developed in the study were isolated, identified, and characterized by UPLC–MS, HRMS, NMR (1D and 2D), and FT-IR techniques.
Isolation of Degradation Products
Degradation was observed under acid and base stress conditions with a considerable percentage of impurity formation > 5%. Purification was carried out using 0.1% formic acid in aqueous and acetonitrile as a mobile phase with In-house packed Luna C18 10µ, 150 × 25 mm (Phenomenex). Crude sample solutions were injected in consecutive injections, and the fractions were collected based on UV response and later mass confirmed by UPLC–MS. After completion of the degradation procedure, the degradation products were purified in Preparative HPLC to collect all the fractions of various degradation products separately and lyophilized to get a free solid.
Structural Confirmation of Degradation Products
To get the structural information of DAR API, all analytical data were recorded for reference purposes. Under positive ESI–MS conditions, [M + H] + at m/z 548.24, which confirms the molecular formula for DAR is C27H37N3O7S. The confirmed data of DAR are mentioned at HRMS (Fig. 4A), HRMS (Table 2), IR (Table 3), and NMR 1D & 2D (Table 4). The structural conformation of DAR by 2D-NMR is shown in Fig. 5. All the other degradation products are characterized based on the comparison of this structural elucidation data.Fig. 4 HRMS data of Darunavir and its degradation products
Table 2 HRMS and melting range data for all degradation products
Products name Formula m/z Calculated m/z Obtained Mass error (ppm) Melting range (°C)
Darunavir C27H38O7N3S 548.2430 548.2416 1.56 111–115
DP-1 C22H32N4O3S 433.2273 433.2249 1.21 97–101
DP-2 C20H29N3O3S 392.2008 392.1995 1.77 95–99
DP-3 C29H40N4O7S 589.2696 589.2686 0.71 131–135
DP-4 C25H27N5O5 418.1800 418.1790 1.32 105–109
DP-5 C22H31N3O5S 450.2062 450.2054 0.64 64–68
Table 3 Comparative FTIR data of Darunavir and its degradation products
Assignment Region (cm−1) API DP-1 DP-2 DP-3 DP-4 DP-5
O–H/N–H stretching 3700–3300 ~ 3368 ~ 3410 ~ 3370 ~ 3349 ~ 3470,3368 ~ 3367
SP3 C–H stretching 2850–3000 ~ 2965 ⁓2957 ⁓2963 ⁓2962 ⁓2960 ⁓2959
C = O stretching 1760–1630 ~ 1699 ~ 1694 – ~ 1707 ~ 1754 ~ 1707
S = O asymmetric stretching 1300–1420 ~ 1338 ~ 1338 ~ 1348 ~ 1331 ~ 1317 ~ 1315
S = O symmetric stretching 1000–1200 ~ 1148 ~ 1154 ~ 1145 ~ 1153 ~ 1148 ~ 1146
C–O stretching in 2° alcohol 1087–1024 ⁓1088 ⁓1089 ⁓1088 ⁓1090 ⁓1087 ⁓1088
Table 4 Relative NMR assignments for Darunavir API
DAR-API (Recorded in DSMO-D6 solvent)
Atom serial number Name of the atom Proton resonance frequency(ppm) and coupling constant J (Hz) Carbon resonance frequency (ppm) 1H–1H-gDQCOSY 1H–13C-HMBC
1,3 CH 7.20(m,2H) 127.85 7.13 139.51
2 CH 7.13(m,1H) 125.66 7.2 129.2
4,6 CH 7.20(m,1H) 129.2 – 125.66, 35.15
5 C – 139.51 – –
7 CH2 2.43(dd,13.2 Hz,11.2 Hz,1H),3.01(dd,13.6 Hz,2.8 Hz,1H) 35.15 3.54 129.2, 139.51, 55.86, 72.8
8 CH 3.54(m,1H) 55.86 2.43,3.01,7.25 139.51
9 CH 3.64(m,1H) 72.8 4.98,2.65,3.27 –
10 CH2 2.65(m,1H),3.27(dd,14.8 Hz,3.2 Hz,1H) 52.76 3.64 57.33, 55.86
12 CH2 2.60(m,1H),2.91(dd,13.2 Hz,8.4 Hz,1H) 57.33 1.94 52.76, 26.34, 20.05
13 CH 1.94(m,1H) 26.34 2.60,2.91&0.78,0.85 57.33, 20.05
14.15 CH 0.78,0.85(d,6.4 Hz,6H) 20.05 1.94 20.05, 57.33
Fig. 5 Darunavir-1H-13C-HMBC & HSQC spectrum
Degradation Product (DP-1) Characterization
The DAR API compound on treatment with acid, a degradation product-1 (DP-1), has been formed. This degradation product has been isolated and its mass was confirmed as (M + H) + 433.2249 with 1.21 ppm error in HRMS. For the calculated molecular formula, C22H32O3N4S is shown in in Fig. 4B. The complete structure of the impurity was characterized by NMR experiments. The sample was prepared in a DMSO-D6-TFA-D mixture of solvents (0.5 ml of DMSO + 0.1 ml of TFA) and conducted NMR analysis. In proton NMR, a few proton signals are broadened, hence adding 0.1 ml of TFA to the DMSO solution to enhance the relaxation. In proton NMR, a total of 27 protons were observed, of which 9 are aromatic protons at 7.0 ppm to 8.0 ppm, and 18 protons are present in the aliphatic region at 0.6 ppm to 4.1 ppm. Two ortho-doublet signals at 7.52 ppm, 7.89 ppm with two protons integration, belong to a 1,4-disubstituted benzene ring in the structure, benzyl protons observed at 7.28 ppm and 7.35 ppm with one proton and four protons integration. A singlet signal with three proton integration at 2.39 ppm confirms the presence of a methyl group (acetamidine). Two doublets with 6 protons integration at 0.76, 0.82 ppm correspond to methyl groups forming the isobutyl group. In 13C NMR, a total of 17 carbon signals were observed, out of which 9 signals were observed around 120 to 165 ppm, and 8 carbons were around 15 to 70 ppm, due to symmetry, the four aromatic signals’ (Benzyl and 1,4-disubstituted benzene ring) intensity was high and considered as eight aromatic protonated carbons, most downfield carbon at 164.62 ppm corresponding to acetamidine carbon. In edited g-HSQC experimental data, confirmed the presence of 9 aromatic protonated methine groups, 3 methyls, 3 methylene groups, and 3 methine groups in the aliphatic region. In gDQ-COSY experiment, H13 methine proton shows a correlation with H12 methylene and H14,15 methyl groups, and H8 methine proton shows a correlation with H7 methylene and H9 methine and H9 methine showing a correlation with H10 methylene protons, H8 proton. In g-HMBC experiment, a key 3 J correlation is observed from H30 protons to C28 carbon. In 1H-15 N-HMBC experiment, H30 methyl protons showed 3j correlation with N29(118.47 ppm) and N25(132.43 ppm) Nitrogen, as well, H21,23 protons showed connectivity to N25 Nitrogen. One excess methyl group and missing hexahydrofuro[2,3-b]furan ring protons in proton NMR, presence of extra carbon at 164.62 ppm, key correlation in g-HMBC and 1H-15 N-HMBC experiments, all these NMR studies concluded that hexahydrofuro[2,3-b]furan cleaved at carbamate link and formed N-(4-(N-(3-amino-2-hydroxy-4-phenyl butyl)-N-isobutylsulfamoyl) phenyl) acetimidamide. The proton, carbon chemical shifts, and COSY, HMBC experiment connectivity were captured in Table 5, and the structural elucidation is represented in Figs. 6, 7, 8, 9, 10. In FTIR spectra, stretching frequency observed at 3410 cm−1 indicates the presence of NH group, the stretching frequency at 1089 cm−1 indicates the presence of secondary alcohol, and stretching frequency at 1338 cm−1, 1154 cm−1 indicates the presence of the sulfoxide group, as shown in Table 3. The melting range of DP-1 is 97–101 °C. The possible mechanism of DP-1 formation from DAR in acidic conditions is explained in Fig. 11.Table 5 Relative NMR assignments for DP-1
DAR-DP-1 (Recorded in DMSO-D6 + 1drop of TFA-D solvent)
Atom serial number Name of the atom Proton resonance frequency (ppm) and coupling constant J (Hz) Carbon resonance frequency (ppm) 1H–1H-gDQCOSY 1H–13C-HMBC
1,3 CH 7.35 (m,2H) 128.68 7.28 136.69
2 CH 7.28 (m,2H) 126.91 7.35 129.44
4,6 CH 7.36 (m,2H) 129.44 – 126.91, 32.60
5 C – 136.69 – –
7 CH2 2.84 and 3.04 (m, 2H) 32.6 3.49 136.69, 129.44, 55.16 and 68.49
8 CH 3.49 (broad hump, 1H) 55.16 2.84, 3.04 and 4.02 −
9 CH 4.02 (broad triplet, 1H) 68.49 3.49, 2.92 and 3.37 –
10 CH2 2.92 (dd,14.4 Hz,7.6 Hz, 1H and 3.37 (dd,14.4 Hz,4.0 Hz,1H) 50.96 4.02 –
12 CH2 2.79 (dd,14.0 Hz,6.8 Hz,1H and 3.01(m, 1H) 56.95 1.92 50.96, 26.34 and 19.9
13 CH 1.92 (M,1H) 26.34 2.79, 3.01, 0.76 and 0.82 56.95, 19.9
14,15 CH3 0.76 and 0.82 (d,6.4 Hz,6H) 19.9 1.92 56.95, 26.34
19 C – 137.87 – –
20,24 CH 7.89 (d, 8.4 Hz, 2H) 129.03 7.52 138.22
21,23 CH 7.52 (d, 8.4 Hz, 2H) 125.69 7.89 137.87
22 C – 138.22 – –
28 C – 164.62 – –
30 CH3 2.39 (s, 3H) 19.15 – 164.62
Fig. 6 1H-15N-HMBC & 1H-13C-HMBC data of DP-1
Fig. 7 13C & 1H-15N-HSQC data of DP-2
Fig. 8 1H-15N-HMBC & 1H-13C-HMBC data of DP-3
Fig. 9 1H-15N-HSQC & 1H-13C-HMBC data of DP-4
Fig. 10 1H-13C HSQC & 1H-13C-HMBC of DP-5
Fig. 11 Proposed mechanism for Acid degradation (DP-1 & DP-3)
Degradation Product (DP-2) Characterization
Degradation product-2 (DP-2) has been formed on treating DAR with acid as well as base, and this DP-2 is isolated and its mass was confirmed as (M + H) + 392.1995 with 1.77 ppm error in HRMS for the calculated molecular formula C20H30O3N3S shown in Fig. 4C. The complete study of DP-2 has been done by 1D and 2D NMR studies. We have recorded needed NMR experiments in 400 MHz instrument, dissolved ~ 10 mg of DP-2 compound in DMSO-D6 and recorded 1H, 13C, g-COSY, g-HSQC and g-HMBC experiments. In proton NMR, a total of 26 protons were observed, in which, 9 protons from the aromatic region, 15 protons from the aliphatic region, and D2O-Exchangeable protons primary amine showed at 5.99 ppm. There are two doublet signals at 6.60 ppm and 7.37 ppm with integration 2 protons, benzyl protons are present at 7.37 ppm with five proton integration, methyl protons from the isobutyl group show at 0.76 ppm with integration value 6, methine proton of the isobutyl is showing at 1.9 ppm, alcohol group attached proton is at 3.78 ppm, aliphatic amine attached methine is at 3.16 ppm, aryl group attached methylene group is at 2.55 and 2.91 ppm, N-attached methylene groups are at 2.75, 3.30, 2.62 and 2.84 ppm. In 13C NMR, 8 aromatic carbons and 7 carbons from the aliphatic region were present, most downfield carbamate carbon is missing. g-HSQC experiment confirmed the presence of three methine groups, three methylenes, two methyls from the aliphatic side, and 9 methines from the aromatic region,1H-15N- HSQC experiment confirm the presence of primary amine, which is present on the aryl ring. The g-HMBC experiment has been conducted to check overall connectivity in structure, and data compiling with the given structure, all 1D and 2D NMR experiment results confirmed the accurate structure of DP-2 as 4-amino-N-(3-amino-2-hydroxy-4-phenyl butyl)-N-isobutyl benzenesulfonamide and represented in Fig. 7. Chemical shifts of 1H and 13C are given in Table 6. In FTIR spectra, stretching frequency observed at 3370 cm−1 indicates the presence of the NH group, the stretching frequency at 1088 cm−1 indicates the presence of 2° alcohol, and the stretching frequency at 1348 cm−1, 1154 cm−1 indicates the presence of the sulfoxide group, as shown in Table 3. The melting range of DP-2 is 95–99 °C. The possible degradation mechanism of DP-2 formation from DAR in basic conditions is explained in Fig. 12.Table 6 Relative NMR assignments for DP-2
DAR-DP-2 (Recorded in DMSO-D6 solvent)
Atom serial number Name of the atom Proton resonance frequency (ppm) and coupling constant J (Hz) Carbon resonance frequency (ppm) 1H–1H-gDQCOSY 1H–13C-HMBC
1,3 CH 7.31 (m, 2H) 128.42 7.23 138.59
2 CH 7.23 (m, 1H) 126.33 7.31 129.43
4,6 CH 7.33 (m, 2H) 129.43 – 126.33, 35.43
5 C – 138.59 – –
7 CH2 2.55 (dd, 13.6 Hz, 8.8 Hz, 1H) and 2.91 (dd, 1.6 Hz, 4.0 Hz, 1H) 35.43 3.16 138.59, 129.43, 55.59 and 70.98
8 CH 3.16 (broad peak, 1H) 55.59 2.55,2.91&3.78 –
9 CH 3.78 (broad peak, 1H) 70.98 3.16,2.75&3.30 –
10 CH2 2.75 (dd, 14.4 Hz,7.2 Hz,1H) and 3.30 (14.8 Hz, 4.4 Hz, 1H) 52 3.78 55.59, 70.98 and 57.4
12 CH2 2.62 (dd, 13.6 Hz, 6.8 Hz, 1H) and 2.84(dd, 13.2 Hz, 8.0 Hz, 1H) 57.4 1.9 52, 26.59 and 20.1
13 CH 1.90 (m,1H) 26.59 2.62, 2.84, 0.76 and 0.81 57.4 and 20.1
14,15 CH3 0.76 and 0.81 (d, 6.4 Hz, 6H) 20.1 1.9 57.4, 26.59
19 C – 123.5 – –
20,24 CH 7.37 (d, 8.8 Hz, 2H) 129.09 6.6 152.84, 123.5
21,23 CH 6.60 (d, 8.8 Hz, 2H) 112.8 7.37 123.5
22 C – 152.84 – –
25 NH2 5.99 (broad singlet, 2H) – – –
Fig. 12 Proposed mechanism for Base degradation (DP-2, DP-4 & DP-5)
Degradation Product (DP-3) Characterization
DP-3 impurity has been formed from DAR on treatment with acid. Then the DP-3 is isolated and its mass was confirmed as (M + H) + 589.2686 with 0.71 ppm error in HRMS for the calculated molecular formula of C29H41O7N4S shown in Fig. 4D. It has excess 41 units than API, to find the accurate structure of DP-3, it was examined with the NMR experiments, in 1H NMR, one of the key methyl groups shows broaden, hence added one drop of TFA to the DMSO to improve the relaxation and recorded 1D and 2D experiment for the same sample, in 1H NMR, 9 aromatic protons are shown in that 5 protons belong to benzyl group with integration 5 at 7.19 ppm and 1,4-disubstituted benzene ring protons observed at 7.51 ppm, 7.91 ppm with integration of each 2, one of the key important methyl group protons observed at 2.36 ppm, and isobutyl group methyls are at 0.81 ppm 0.86 ppm, remaining 18 methine and methylene protons were observed between 1.0 and 6.0 ppm, in 13C NMR, a total 25 carbon signals were observed, in that 10 carbons are from aromatic and 15 from aliphatic region, most downfield carbons are amidine carbon showing at 164.73 ppm, carbamate carbon is at 155.45 ppm, remaining aryl carbons between 100 and 140 ppm, N-attached, O-attached methine and methylene carbons are at 45 pm to 75 ppm; methyls, isobutyl methine and methylenes carbons are between 15 and 45 ppm, most up-field carbon is acetamidine at 19.30 ppm; in g-HSQC experiment, 6 methines, 6 methylenes, and 3 methyls were confirmed in aliphatic region, and benzyl and 1,4-disubstituted benzene protonated carbons were confirmed in aromatic region. 2 J and 3 J correlations of 1H-13C are identified by 1H-13C-HMBC, and 2 J and 3 J correlations of Proton to Nitrogen were confirmed in 1H-15 N-HMBC experiment, the most important observation in g-HMBC experiment is H41 protons showing connectivity to C39 carbon, and 1H-15 N-HMBC experiment H41 protons showing connectivity to N25 (130.44 ppm) and N40 (116.65 ppm) nitrogen; as well, H21 and H23 protons showed connectivity to N25 Nitrogen. This key connectivity in 15 N-HMBC strongly confirms the structure of DP-3, here primary amine group was converted as acetamidine and formed as hexahydrofuro[2,3-b]furan-3-yl(4-((4-acetimidamido-N-isobutylphenyl)sulfonamido)-3-hydroxy-1-phenylbutan-2-yl)carbamate, 1H and 13C assignment details are given in Table 7 and confirmed structure shown in Fig. 8. In FTIR spectra, stretching frequency observed at 3349 cm−1 indicates the presence of the NH group, the stretching frequency at 1090 cm−1 indicates the presence of 2° alcohol, stretching frequency at 1331 cm−1, and 1153 cm−1 indicates the presence of the sulfoxide group, as shown in Table 3. The melting range of DP-3 is 131–135 °C. The possible degradation mechanism of DP-3 formation from API in acidic conditions is explained in Fig. 11.Table 7 Relative NMR assignments for DP-3
DAR-DP-3 (Recorded in DMSO-D6 + 1drop of TFA-D solvent)
Atom serial number Name of the atom Proton resonance frequency (ppm) and coupling constant J (Hz) Carbon resonance frequency (ppm) 1H–1H-gDQCOSY 1H–13C-HMBC
1,3 CH 7..22 (m, 2H) 128.17 7.15 139.5
2 CH 7.15 (m, 1H) 126.04 7.22 129.42
4,6 CH 7.23 (m, 2H) 129.42 – 126.04, 35.41
5 C – 139.5 – –
7 CH2 2.49 (dd, 13.6 Hz, 11.2 Hz, 1H) and 3.02 (dd, 14.0 Hz, 2.8 Hz, 1H) 35.41 3.57 139.5, 129.42, 55.92, 72.7
8 CH 3.57 (m, 1H) 55.92 2.49, 3.02 −
9 CH 3.64 (m, 1H) 72.7 2.89 –
10 CH2 2.89 (dd, 14.8 Hz, 8.0 Hz, 1H) and 3.39 (dd, 14.8 Hz, 2.4 Hz, 1H) 52.25 3.64 55.92, 72.7, 56.68
12 CH2 2.83 (m, 1H) and 3.09 (dd, 13.6 Hz, 8.8 Hz, 1H) 56.68 1.97 52.25, 26.36, 20.02, 20.10
13 CH 1.97 (m, 1H) 26.36 2.83, 3.09, 0.81, 0.86 56.68, 20.2, 20.10
14,15 CH3 0.81 and 0.86 (d, 6.4 Hz, 6H) 20.02 and 20.10 1.97 26.36, 56.68
19 C – 138.7 – –
20,24 CH 7.91 (d, 8.4 Hz, 2H) 129.19 7.51 137.9
21,23 CH 7.51 (d, 8.4 Hz, 2H) 125.79 7.91 138.7
22 C – 137.9 – –
28 CO – 155.45 – –
31 CH 4.85 (m, 1H) 72.36 3.58, 3.83, 2.76 155.45, 109.04, 25.75
32 CH2 3.58 (m, 1H) and 3.83 (dd, 9.6 Hz, 6.0 Hz, 1H) 70.68 4.85 109.04, 45.28, 72.36
34 CH 5.51 (d, 5.2 Hz, 1H) 109.04 2.76 69.09, 45.28, 70.68, 72.36, 25.75
35 CH 2.76 (m, 1H) 45.28 4.85, 5.51, 1.24, 1.40 69.09, 70.68
36 CH2 1.24 and 1.40 (m, 2H) 25.75 2.76, 3.63, 3.73 109.04, 69.09, 72.36, 45.28
37 CH2 3.63 (m, 1H) and 3.73 (td, 8.0 Hz, 1.6 Hz, 1H) 69.09 1.24, 1.40 109.04, 45.28
39 C – 164.73 – –
41 CH3 2.36 (s, 3H) 19.3 – 164.73
Degradation Product (DP-4) Characterization
API compound on treated with base DP-4 impurity has been formed. This DP-4 is isolated and its mass confirmed as (M + H)+ 418.1790 with 1.32 ppm error in HRMS for the calculated molecular formula of C21H28O4N3S shown in Fig. 4E. Mass is less than parent API compound 124 units and to find the precise structure of DP-4 1D and 2D NMR, experiments have been conducted, used DMSO-D6 solvent to solubilize the compound and acquired in 400 MHz instrument, under the 1D experiments conducted 1H and 13C experiments; in proton NMR, a total 27 protons are observed, in that 9 protons are from aromatic region, 15 protons from aliphatic, and 3 labile protons, in aromatic region at benzyl protons are at 7.25 ppm and 7.33 ppm with respective 3 and 2 protons integration, 1,4 di-substituted ring protons present at 6.56 ppm and 7.19 ppm with the integration of two protons each and primary amine protons are at 6.00 ppm and oxazolidone ring proton is at 7.65 ppm as a singlet, isobutyl group methyls and methine protons are at 0.78, 0.84 ppm and 1.82 ppm, remaining methine and methylenes are between 2.5 ppm and 5.0 ppm. In 13C NMR, a total 17 carbons are present in that 9 carbons are aromatic region between 110 and 160 ppm, most downfield carbon is 2-Oxazolidone carbon at 157.64 ppm, and primary amine attached carbon is at 152.90 ppm, symmetric carbons in phenyl and 1,4-disubstituted ring carbons are showing high-intensity signals, in aliphatic region, carbon signals are between 15 and 80 ppm and most unfilled carbons are isobutyl group methyl carbons and downfield carbon is O-attached methine showing at 77.58 ppm, Under the 2D experiments conducted g-HSQC, 1H-15 N-HSQC and g-HMBC experiments, to identity proton attached carbons performed g-HSQC experiment, here got the clear information about every methylene carbons and methine carbons, further conducted 1H-15 N-HSQC experiment, it gave clarity on the presence of amine and amide protons in the structure, and in g-HMBC experiment observed, one of the key correlations is H8,H9 and H27 protons showing connectivity to C28 carbon. Here H9 to C28 correlation should possible in cyclized structure only, all the experiments data confirmed that carbamate carbon got cyclized with secondary alcohol group and formed 2-Oxazolidone ring contained structure as 4-amino-N-(((5S)-4-benzyl-2-oxooxazolidin-5-yl)methyl)-N-isobutylbenzenesulfonamide and structure is given at Fig. 9. Its detailed proton and carbon values are captured in Table 8. In FTIR spectra, stretching frequency observed at 3470 cm−1, 3368 cm−1 indicates presence of NH group, stretching frequency at 1087 cm−1 indicates the presence of 2° alcohol, stretching frequency at 1317 cm−1,1148 cm−1 indicates the presence of sulfoxide group, as shown in Table 3. The melting range of DP-4 is 105–109 °C. The possible degradation mechanism of DP-4 formation from DAR in the basic condition is explained in Fig. 12.Table 8 Relative NMR assignments for DP-4
DAR-DP-4 (Recorded in DMSO-D6 solvent)
Atom Serial number Name of the atom Proton resonance frequency (ppm) and coupling constant J (Hz) Carbon resonance frequency (ppm) 1H–1H-gDQCOSY 1H–13C-HMBC
1,3 CH 7.33 (t, 7.2 Hz, 2H) 128.51 7.27 137.4
2 CH 7.27 (t, 7.2 Hz, 2H) 126.4 7.33 128.81
4,6 CH 7.23 (d, 8.0 Hz, 2H) 128.81 7.33 126.4, 35.03
5 C – 137.4 – –
7 CH2 2.69 (14.4 Hz, 7.6 Hz, 1H) and 2.85 (14.4 Hz, 6.8 Hz, 1H) 35.03 4.26 137.4, 128.81, 54.25 and 77.58
8 CH 4.23 (q, 7.2 Hz, 1H) 54.25 2.69, 2.85 and 4.69 137.4, 35.03, 77.58 and 48.88, 157.64
9 CH 4.69 (td, 10 Hz, 2.4 Hz, 1H) 77.58 4.23, 3.14 and 3.23 157.64, 48.88, 54.25 and 35.03
10 CH2 3.14 (dd,15.2 Hz,9.6 Hz,1H) and 3.23 (dd, 15.2 Hz, 2.4 Hz, 1H) 48.88 4.69 54.25, 77.58, 57.02
12 CH2 2.57 (dd, 13.6 Hz, 6.0 Hz, 1H) and 2.79 (dd, 13.6 Hz, 8.80 Hz, 1H) 57.02 1.84 48.88, 26.27 and 19.84, 19.93
13 CH 1.84 (m, 1H) 26.27 2.57, 2.79 and 0.78, 0.84 57.02, 19.84, 19.93
14,15 CH3 0.78 and 0.84 (d, 6.8 Hz, 6H) 19.84 and 19.93 1.84 57.03, 26.27
19 C – 122.74 – –
20,24 CH 7.19 (d, 8.8 Hz, 2H) 129.01 6.56 122.74, 152.9
21,23 CH 6.56 (d, 8.8 Hz, 2H) 112.76 7.19 122.74
22 C – 152.9 – –
25 NH2 6.00 (s, 2H) – – 112.76
27 NH 7.65 (s, 1H) – – 54.25, 77.58 and 157.64
28 CO – 157.64 – –
Degradation Product (DP-5) Characterization
API compound on treatment with base DP-5 impurity has been formed, this degradation product is isolated and its mass confirmed as (M + H) + 450.2054 with 0.64 ppm error in HRMS for the calculated molecular formula of C22H32O5N3S shown in Fig. 4F. Compound mass shows 98 units, it is lower than API compound, to confirm the precise structure of DP-5 performed NMR experiments. In 1H NMR, the most significant observation is no protons observed from hexahydrofuro[2,3-b]furan ring, and singlet signal is shown at 3.39 ppm with integration value 3. It clearly indicates the presence of O-methoxy group in the structure, two doublet methyls observed at 0.78 ppm & 0.84 ppm corresponding to isobutyl group, 3 methylene groups and 3 methines groups are present between 1.5 and 4.0 ppm, secondary alcohol proton is showing at 4.95 ppm and amidic –NH proton is at 7.02 ppm, benzyl ring protons are showing at 7.22 ppm with integration 5 protons, 1,4-di-substituted benzene ring protons are shown at 6.59 ppm and 7.37 ppm with integration each signal two protons. The total number of protons in the structure is 31, in 13C NMR, most downfield carbon corresponding to methyl carbamate is shown at 156.29 ppm, and benzyl and 1,4-di-substituted benzene ring carbons are shown between 110 and 152 ppm, and a total number of aromatic carbons are 9, in aliphatic region, 8 carbons are present, O-methoxy carbon is at 51.09 ppm and alcohol group attached carbon is at 72.44 ppm, and isobutyl group methyl carbons are shown at 20.05 ppm, N-attached methylene carbons are at 52.67 ppm & 57.24 ppm, benzyl attached methylene group carbon is at 34.86 ppm, methine carbons are at 55.69 & 26.32 ppm, conducted HSQC experiment for confirmation of methine, methylene and methyl carbons. Here H31 protons attached carbon is shown at 51.09 ppm. This chemical shift confirms the presence of O-methoxy group in the compound, in g-HMBC experiment, O-methoxy protons are shown 3 J connectivity to carbamate carbon, and this key correlation confirms the position of O-methoxy position. Based on above key correlations, structure of DP5 is methyl ((3S)-4-((4-amino-N-isobutylphenyl)sulfonamido)-3-hydroxy-1-phenylbutan-2-yl)carbamate. Detailed assignment of proton and carbon values is captured in Table 9, and structure is shown at Fig. 10. In FTIR spectra, stretching frequency observed at 3367 cm−1, indicates the presence of the NH group, the stretching frequency at 1088 cm−1 indicates the presence of 2° alcohol, and the stretching frequency at 1315 cm−1, and 1146 cm−1 indicates the presence of sulfoxide group, as shown in Table 3. The melting range of DP-5 is 64–68 °C. The possible degradation mechanism of DP-5 formation from DAR in the basic condition is explained in Fig. 12.Table 9 Relative NMR assignments for DP-5
DAR-DP-5 (Recorded in DMSO-D6 solvent)
Atom serial number Name of the atom Proton resonance frequency(ppm) and coupling constant J (Hz) Carbon resonance frequency (ppm) 1H–1H-gDQCOSY 1H–13C-HMBC
1,3 CH 7.22 (m, 2H) 127.89 7.14 139.63
2 CH 7.14 (m, 1H) 125.7 7.22 129.12
4,6 CH 7.23 (m, 2H) 129.12 – 125.7, 34.86
5 C – 139.63 – –
7 CH2 2.47 (m, 1H) and 2.97 (14.0 Hz, 2.8 Hz, 1H) 34.86 3.54 139.63, 129.12, 55.69
8 CH 3.54 (m, 1H) 55.69 2.47, 2.97, 7.02, 3.64 −
9 CH 3.64 (m, 1H) 72.44 2.66, 3.25 and 3.54, 4.95 –
10 CH2 2.66 and 3.25 (m, 2H) 52.67 3.64 72.44, 57.24
12 CH2 2.59 (dd, 13.2 Hz, 6.4 Hz, 1H) and 2.90 (13.2 Hz, 8.4 Hz, 1H) 57.24 1.93 52.67, 26.32 and 20.05
13 CH 1.93 (m, 1H) 26.32 2.59, 2.90 and 0.78, 0.84 57.24, 20.05
14,15 CH3 0.78 and 0.84 (d, 6.4 Hz, 6H) 20.05 1.93 57.24, 26.32
19 C – 123.45 – –
20,24 CH 7.37 (d, 8.8 Hz, 2H) 129.02 6.59 152.7
21,23 CH 6.59 (d, 8.8 Hz, 2H) 112.64 7.37 123.45
22 C – 152.7 – –
25 NH2 5.97 (s, 2H) – – 112.64
26 OH 4.95 (d, 6.4 Hz, 1H) – 3.64 –
27 NH 7.02 (d, 9.2 Hz, 1H) – 3.54 –
28 CO – 156.29 – –
31 OCH3 3.39 (s, 3H) 51.09 – 156.29
Conclusion
The stress degradation study of Darunavir was examined as per ICH prescribed guidelines. The drug was subjected to acidic, alkaline, oxidative, thermolytic, and photolytic degradation conditions. The drug was stable in oxidative, thermal, and photolytic conditions as it did not show any impurity formation. However, five degradation products were formed in acid and base stress hydrolysis conditions. These degradation products were isolated and fully characterized by various analytical techniques like NMR (1D & 2D experiments) and HRMS experiments. FT-IR data gave an additional add-on to confirm the structures. Among five DPs, DP-1, DP-3 are novel products, and DP-2, DP-4, and DP-5 are reported degradation products with limited data as the structures were characterized only by mass spectral data, and no structural elucidation was done by 2D-NMR and HRMS. The current study provides the complete structural interpretation of Darunavir and all 5 degradation impurities using HRMS, FTIR, and 2D-NMR studies. It also reports well-developed UPLC–MS method to separate all the impurities with good resolution. The method used is simple, sensitive, and stability indicating with very short runtime i.e., 7.0 min. Using this we can perform qualitative and quantitative analyses of degradation products. The complete structural characterization of degradation products by HRMS, NMR, and FTIR will help the complete studies and the nature of the degradation products, which will help in the process, shelf-life time, and safety of the API products. These stress studies will provide knowledge about possible degradation products and the pathways of API and help to elucidate the structure of the degradation products. This stress degradation can be a useful tool to predict the stability of a drug product with effects on purity, potency, and safety. It is important to know the impurity profile and behavior of a drug product under various stress conditions. The isolation method of the degradation products will help to know the individual characterization of degradation products. The complete spectral analysis of 1HNMR,13C NMR, 2D-NMR, FT-IR, and HR-MS was reported for the first time and confirmed the proposed chemical structures of degradation impurities.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 1847 KB)
Acknowledgements
Sincere thanks to Aragen Life Sciences Pvt. Ltd. management for support and providing the laboratory facility to perform this research.
Author contributions
Arun Kumar Modinia carried out experimentation work and prepared the manuscript Sathish Kumar Konidala planed and guided the work and prepared the manuscript Mahesh Ranga, Umamaheshwar Puppala, Muralidharan Kaliyapermala, and Mahesh Kumar Reddy Geereddy did data collection, analysis. Ramu Samineni and parul Grover did the manuscript review and suggestions
Data availability statement
The data can be available as the supplementary data file attached to this article.
Declarations
Conflict of Interest
The authors declare that they have no conflict of interests.
Consent for Publication
We authorize to publish the article without any conflict.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. de Oliveira Melo SR Homem-de-Mello M Silveira D Simeoni LA Advice on degradation products in pharmaceuticals: a toxicological evaluation PDA J Pharm Sci Technol 2014 68 3 221 238 10.5731/pdajpst.2014.00974 25188345
2. Bana AA Patel P Mehta PJ Forced degradation study of Darunavir ethanolate and ritonavir combination in acidic, basic and oxidative conditions establishing degradation products Int J Pharm Sci Res 2020 11 11 5875 5883 10.13040/IJPSR.0975-8232.11(11)
3. Corrêa JC, Serra CH, Salgado HR. Stability study of darunavir ethanolate tablets applying a new stability-indicating HPLC method. Chromatography Research International. 2013;2013: Article ID 834173.
4. Rathod S Pounikar AR Umekar MJ Gupta KR Development of HPLC method for estimation of Darunavir related substance in formulation Biomed J Sci Tech Res 2020 28 2 21444 21460 10.26717/BJSTR.2020.28.004624
5. Rao RN Ramachandra B Santhakumar K RP-HPLC separation and characterization of unknown impurities of a novel HIV-protease inhibitor Darunavir by ESI-MS and 2D NMR spectroscopy J Pharm Biomed Anal 2013 5 75 186 191
6. Patel KP Jasani M Patel CJ Patel MM Development and validation of stability indicating RP HPLC Method for estimation of Darunavir and its related substance in tablet dosage form World J Pharm Res 2018 7 8 709 722 10.20959/wjpr20188-11587
7. Yusop Z Jaafar J Aris AB Majid ZA Umar K Talib J Development and validation of a selective, sensitive and stability indicating UPLC–MS/MS method for rapid, simultaneous determination of six process related impurities in darunavir drug substance J Pharm Biomed Anal 2016 128 141 148 10.1016/j.jpba.2016.05.026 27262107
8. Rao RN Ramachandra B Sravan B Khalid S LC–MS/MS structural characterization of stress degradation products including the development of a stability indicating assay of Darunavir: an anti-HIV drug J Pharm Biomed Anal 2014 15 89 28 33 10.1016/j.jpba.2013.10.007
9. Deshpande P Butle S Development and validation of stability-indicating HPTLC method for determination of darunavir ethanolate and ritonavir Int J Pharm Pharm Sci 2015 7 6 66 71
10. Patel BN Suhagia BN Patel CN RP-HPLC method development and validation for estimation of darunavir ethanolate in tablet dosage form Int J Pharm Pharm Sci 2012 4 3 270 273
11. Reddy BR Jyothi G Reddy BS Raman NV Reddy KS Rambabu C Stability-Indicating HPLC method for the determination of darunavir ethanolate J Chromatogr Sci 2013 51 5 471 476 10.1093/chromsci/bms165 23097581
12. Satyanarayana L Naidu SV Rao MN Kumar A Suresh K The estimation of darunavir in tablet dosage form by RP-HPLC Asian J Res Pharm Sci 2011 1 3 74 76
13. Nalini MV Veni PR Haribabu B Determination of darunavir and cobicistat simultaneously using stability indicating RP-HPLC method Marmara Pharmaceutical Journal 2016 20 3 293 302 10.12991/mpj.20162036176
14. Rao RN Prasad KG LC–Q-TOF-MS/MS determination of darunavir and its metabolites in rat serum and urine: application to pharmacokinetics J Pharm Biomed Anal 2014 1 94 92 98
15. Vermeir M Lachau-Durand S Mannens G Cuyckens F van Hoof B Raoof A Absorption, metabolism, and excretion of darunavir, a new protease inhibitor, administered alone and with low-dose ritonavir in healthy subjects Drug Metab Dispos 2009 37 4 809 820 10.1124/dmd.108.024109 19131522
16. ICH Guideline Q2 (R1), Validation of Analytical Procedures, Text and Methodology, 2005.
17. ICH Guideline Q1A(R2) Stability Testing of New Drug Substances and Products, 2003.
18. ICH Guideline Q3B(R2) Impurities in New Drug Products, 2006.
19. WHO, Draft Stability Testing of Active Pharmaceutical Ingredients and Pharmaceutical Products, World Health Organization: Geneva, 2007.
20. FDA, Guidance for Industry: Stability Testing of Drug Substances and Drug Products (Draft Guidance). Food and Drug Administration, Rockville, MD, 1998.
21. Salakolusu S Sharma GVR Katari NK Puppala U Kaliyapermal M Vijay R Identification, isolation, and structural characterization of novel forced degradation products of apixaban using advanced analytical techniques J Sep Sci 2022 45 3942 3954 10.1002/jssc.202200466 36048725
| 0 | PMC9739338 | NO-CC CODE | 2022-12-14 23:29:54 | no | Chromatographia. 2022 Dec 10;:1-16 | utf-8 | Chromatographia | 2,022 | 10.1007/s10337-022-04226-z | oa_other |
==== Front
Mycopathologia
Mycopathologia
Mycopathologia
0301-486X
1573-0832
Springer Netherlands Dordrecht
694
10.1007/s11046-022-00694-x
Original Article
Assessment of Risk Factors and Clinical Outcomes in Hospitalized COVID-19 Patients with Candida spp. Co-infections: Species Distribution and Antifungal Susceptibility Patterns of Isolates
Yazdanpanah Somayeh 12
Ahmadi Mohammad 3
Zare Zahra 3
Nikoupour Hamed 3
Arabsheybani Sara 3
Jabrodini Ahmad 12
Eghtedarnejad Esmaeel 12
Chamanpara Parisa 4
Geramizadeh Bita 3
Anbardar Mohammad Hossein 3
Malekizadeh Zahra 3
Gashtasebi Maryam 3
Mohsenzadeh Mehdi 5
Shafiekhani Mojtaba [email protected]
367
Zomorodian Kamiar [email protected]
28
1 grid.412571.4 0000 0000 8819 4698 Student Research Committee, School of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran
2 grid.412571.4 0000 0000 8819 4698 Department of Medical Parasitology and Mycology, School of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran
3 grid.412571.4 0000 0000 8819 4698 Shiraz Transplant Center, Abu-Ali Sina Hospital, Shiraz University of Medical Sciences, Shiraz, Iran
4 grid.412571.4 0000 0000 8819 4698 Department of Biostatistics, School of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran
5 grid.512375.7 0000 0004 4907 1301 Cellular and Molecular Research Center, Gerash University of Medical Sciences, Gerash, Iran
6 grid.412571.4 0000 0000 8819 4698 Department of Clinical Pharmacy, Faculty of Pharmacy, Shiraz University of Medical Sciences, Shiraz, Iran
7 grid.412571.4 0000 0000 8819 4698 Transplant Research Center, Shiraz University of Medical Sciences, Shiraz, Iran
8 grid.412571.4 0000 0000 8819 4698 School of Medicine, Basic Sciences in Infectious Diseases Research Center, Shiraz University of Medical Sciences, Shiraz, Iran
Handling Editor: Abdullah Mohammed Said Al-Hatmi.
10 12 2022
112
2 9 2022
12 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.
Introduction
Fungal co-infections are considered an important complication in hospitalized patients with SARS-CoV-2 that can be attributed to disease aggravation, increased mortality, and poor outcomes. This study was conducted to determine the species distribution and antifungal susceptibility patterns of Candida isolates from hospitalized COVID-19 patients in Shiraz, Iran, in addition to associated risk factors and outcomes of co-infections with Candida species.
Materials and Methods
In this single-center study, a total of 106 hospitalized COVID-19 patients were evaluated for clinical characteristics and outcomes. Species identification was performed by ITS1-5.8S-ITS2 gene sequencing. Antifungal susceptibility testing to fluconazole, itraconazole, voriconazole, posaconazole, caspofungin, amphotericin B, and nystatin was determined according to the M27-A3/S4 CLSI protocol.
Results
Candida species were recovered from 48% (51/106) of hospitalized COVID-19 patients. Statistical analysis showed that patients who had heart failure, bacterial co-infection, and were receiving empirical antifungal therapy had a higher risk of developing Candida co-infection. In total, 71 Candida isolates were recovered, of which C. albicans (69%) was the most prevalent isolate. The majority of the Candida isolates were susceptible to all classes of tested antifungal drugs.
Discussion
Our results elucidate a high rate of Candida co-infections among hospitalized COVID-19 patients. Comorbidities such as heart failure, HTN, COPD, bacterial infections as well as therapeutic interventions including catheterization, mechanical ventilation, and ICU admission increased the risk of Candida spp. isolation from the bloodstream, respiratory tract and urine samples, which led to a higher in-hospital mortality rate. Additionally, obtained data clarified that empirical antifungal therapy was not as successful as anticipated.
Keywords
Co-infection
COVID-19
Candida
Antifungal susceptibility
Candidiasis
http://dx.doi.org/10.13039/501100013041 Vice-Chancellor for Research, Shiraz University of Medical Sciences 26040 Zomorodian Kamiar
==== Body
pmcIntroduction
Microbial co-infections in hospitalized patients with COVID-19 have been documented in many investigations from the onset of the global pandemic in Wuhan. Co-infections in SARS-CoV-2 infected individuals were caused by a variety of pathogens, and it was determined that these concomitant infections resulted in disease aggravation and poor outcomes [1–3].
In patients with SARS-CoV-2 infection, predisposing conditions such as mechanical ventilation and ICU admission in critically ill patients, the use of broad-spectrum antibiotics, immunosuppressive/anti-inflammatory treatments, catheterization, and underlying diseases all increase the risk of secondary infections [4, 5]. Among fungal pathogens, opportunistic fungi such as Aspergillus and Candida species cause fungal co-infections in COVID-19 patients [6, 7]. Since species of the genus Candida are normal inhabitants of such various sites and internal organs as skin, mucous membranes, respiratory tract, digestive system, and urinary tract, they can give rise to infections due to the presence of favorable conditions in COVID-19 patients [6, 8].
Iran has been one of the countries most afflicted by the COVID-19 pandemic in the Middle East, with 7,553,169 confirmed cases and 144,502 deaths as of October 16, 2022 [9]. Early diagnosis and management of concomitant fungal infections in COVID-19 patients with effective antifungal agents leads to improved clinical outcomes. As a result, clinicians can benefit from a better understanding of the epidemiological evidence of co-infections in COVID-19 patients, such as risk factors, species distribution, and susceptibility profiles of isolates, to manage and control super-infections. In view of these considerations, this study was conducted to assess the risk factors as well as clinical outcomes of Candida co-infections in patients with COVID-19 admitted to Abu-Ali Sina hospital in Shiraz, during the COVID-19 pandemic. Furthermore, the species distribution and antifungal susceptibility profiles of isolates recovered from COVID-19 patients were determined in this study.
Materials and Methods
Study Design and Participants
For this observational and single-center study, all patients with confirmed COVID-19 admitted to Abu-Ali Sina hospital in Shiraz, one of the largest hospitals in the south of Iran, were evaluated for Candida co-infections from September to November 2021. This center has 600 individual hospital beds, and the number of beds was increased during the COVID-19 pandemic. SARS-CoV-2 was diagnosed based on positive Real-time Polymerase Chain Reaction (PCR) tests for SARS-CoV-2 or according to the clinical guidelines definition for COVID-19 [10]. All yeast isolates were recovered from clinical samples of COVID-19 patients (blood, urine, tracheal aspirate) during their hospital stay. Moreover, no clinical intervention in antifungal treatment was included in this study.
The patient data was extracted from the electronic medical record system of the hospital considering several variables, such as demographic characteristics, comorbidities, COVID-19 severity, and management such as antivirals, corticosteroids, antibacterial and immunomodulators, bacterial co-infections, empirical/definitive antifungal therapy, length of hospitalization, and length of ICU stay. In addition, relative laboratory results for the patients who participated in the study were collected. Moreover, risk factors associated with yeast co-infections and also clinical outcomes were investigated. Informed consent was obtained from all individual participants included in the study, which was approved by the ethics committee of the Shiraz University of Medical Sciences (IR.SUMS.REC.1401.210).
Fungal Culture and Isolation
To detect fungal strains, all clinical samples obtained from confirmed COVID-19 patients (blood, urine, tracheal aspirate) were cultured on Sabouraud Dextrose Agar (SDA, HiMedia, India), and HiCrome Candida Differential Agar (HiMedia,India) in the Department of Microbiology of Abu-Ali Sina Medical Center.
Species Identification by Sequencing the rDNA Region
DNA extraction of pure colonies was performed according to the method as previously described [11]. Definitive species identification of isolates were performed by PCR-sequencing method following amplification of ITS region [12].The ITS region sequence for each isolate was assembled and used for BLAST explores (http://blast.ncbi.nlm.nih.gov/Blast.cgi?). Sequence data of ITS region were deposited in NCBI and GenBank, and accession numbers of sequences are available for all isolates as follows: ON 312540-69, ON312571-99, ON479767-77, ON514607.
Antifungal Susceptibility Testing (AFST)
Minimum Inhibitory concentration (MIC) values were determined using broth microdilution method according to the M27-A3/S4 protocol documented by Clinical and Laboratory Standards Institute (CLSI) [13]. All clinical isolates were evaluated for susceptibility to antifungals including Fluconazole (FLZ, Sigma,USA), Posaconazole (PSZ,Sigma,Germany), Voriconazole (VRZ, Pfizer, New York, USA), Itraconazole (ITR,Sigma,USA), Caspofungin (CSP,sigma,USA), Nystatin (NYS, Sigma, Germany), and also amphotericin B AMB,Sigma,Germany). RPMI 1640 (Sigma, St. Louis, Missouri,USA) was prepared as directed by the manufacturer, and buffered to pH 7.0 using 0.165 N-morpholino propanesulfonic acid (MOPS) (Sigma, USA). Pure isolates were grown on SDA by incubation for 24 h at 35 °C. Following the growth of isolates, the inoculum suspensions were made by suspending the colonies in NaCl, and the turbidity adjusted to 0.5 McFarland at 530 nm.Then, prepared suspension was diluted to 0.5–2.5 103 cells/ml in RPMI 1640 media. Two-fold serial dilutions of antifungals were made in 96-well plates, and 100 µlit of yeast inoculum was added to each well with in equal volume. Finally, the plates were incubated for 24 h at 35 °C.
Interpretation of data was performed using clinical breakpoints that described in former CLSI documents [14, 15], and new editions of defined breakpoints [16, 17].When the clinical breakpoint was not available, the epidemiological cutoff value (ECV) was used. MIC was defined as the lowest concentration of antifungal agents that inhibits the growth of yeast isolates by 50% for CSP, FLZ, VRZ, ITR, and PSZ in comparison to the controls (drug-free wells). Moreover, the lowest concentration that resulted in any visible growth of isolates (100% inhibition) was considered as the MIC for NYS and AMB. The final concentrations of the antifungal agents were 0.032–16 μg/ml for AMB, ITR, PSZ, VRZ, and NYS, 0.125–64 μg/ml for FLZ, and 0.015–8 μg/ml for CSP. C. krusei (ATCC 6258) was included as a reference strain for quality control.
Statistical Analysis
Data were analyzed using SPSS statistical software version 18.00 (SPSS Inc.IBM.USA). Continuous variables are shown as median and interquartile range (IQR). Also, continuous data with normal distribution were expressed either as means ± standard deviation. Categorical variables were reported as numbers and percentages. Statistical differences between groups were analyzed using the exact Chi-square and/or Fisher’s exact test, and the χ2 test or Mann–Whitney U test. p-value < 0.05 was considered statistically significant.
To highlight the risk factors associated with Candida co-infections, univariable and multivariable logistic regression models were used. Factors with a p-value < 0.2 were retained for multivariate analysis, and those demonstrating statistical significance (p-value < 0.05) on multivariate analysis were considered verifiable risk factors.
Results
A total number of 106 patients were included in this study (51 patients with Candida spp. positive culture and 55 patients without Candida spp. positive culture). All the statistical data analyzed in this study is shown in Table 1.Table 1 Demographic characteristics, clinical data, laboratory findings, treatments and outcomes of COVID-19 patients (N = 106)
Total (106) Patients with Candida positive culture (51) Paients without Candida positive culture (55) p-value
Demographics characteristics
Age, years (mean ± SD) 61 ± 16 66 ± 15 57 ± 11 0.003
Gender
Male 64 (60.4%) 28 (54.9%) 36 (65.4%) 0.26
Female 42 (39.6%) 23 (45.1%) 19 (34.5%)
Clinical interventions
Nasal cannula oxygenation 90 (85.7%) 44 (86.3%) 46 (83.6%) 0.73
ICU admission 39 (36.8%) 24 (47.1%) 15 (27.3%) 0.03
MV 38 (35.8%) 24 (47.1%) 14 (25.5%) 0.02
CVC 36 (34%) 16 (31.4%) 20 (36.4%) 0.58
Urinary catheterization 45 (42.5%) 28 (54.9%) 17 (30.9%) 0.01
Comorbid and clinical conditions
Hepatic failure 16 (15.2%) 5 (10%) 11 (20%) 0.15
Renal failure 32 (30.2%) 18 (35.3%) 14 (25.5%) 0.27
Heart failure 35 (33.0%) 26 (50.9%) 9 (16.4%) < 0.001
DM 46 (43.4%) 25 (49.0%) 21 (38.2%) 0.26
HTN 69 (65.1%) 38 (74.5%) 31 (56.4%) 0.05
ACS 9 (8.5%) 4 (7.8%) 5 (9.1%) 0.81
COPD 15 (14.2%) 12 (23.5%) 3 (5.5%) 0.008
Malignancy 8 (7.5%) 5 (9.8%) 3 (5.5%) 0.39
Neutropenia 2 (1.9%) 0 2 (3.6%) 0.16
Hemodialysis 23 (21.9%) 10 (19.6%) 13 (24.1%) 0.58
Solid organ Transplantation 23 (21.7%) 5 (9.8%) 18 (32.7%) 0.004
Septic shock 28 (26.4%) 17 (33.3%) 11 (20.0%) 0.12
Bacterial co-infection 49 (46.2%) 31 (60.8%) 18 (32.7%) 0.004
Disease severity
Mild/moderate 69 (65.1%) 28 (54.9%) 41 (74.5%) 0.03
Severe/critical 37 (34.9%) 23 (45.1%) 14 (25.5%)
Treatments
Antibacterial treatment 99 (93.4%) 49 (96.1%) 50 (90.9%) 0.44
Empirical antifungal treatments 50 (47.2%) 30 (58.8%) 20 (36.4%) 0.02
Anti-viral treatments 87 (82.1%) 40 (78.4%) 47 (85.5%) 0.34
IL-6 antagonist 21 (19.8%) 11 (21.6%) 10 (18.2%) 0.66
High dose corticosteroids 102 (96.2%) 47 (92.2%) 55 (100%) 0.05
Outcomes
In-hospital mortality 29 (27.4%) 18 (35.3%) 11 (20%) 0.07
Duration of hospitalization, days 12 (8–17) 12 (8–16) 11 (8–18) 0.9
Duration of ICU stay, days 0 (0–6) 8 (5–14) 9 (5–23) 0.61
Duration of MV,days 0 (0–5) 8 (4–13) 7 (5–23) 0.98
Duration of CVC, days 0 (0–4) 1 (0–8) 4 (0–22) 0.19
Duration of urinary catheterization 0 (0–9) 5 (0–10) 0 (0–6) 0.036
Laboratory findings
White blood cell count, ×103 per microliter 12 (7.25–20) 15 (11–23) 12 (4–16) 0.005
Neutrophil (%) 84 (79–90) 86.5 (81–92) 82 (76.75–89.25) 0.1
Serum Ferritin (ng/mL) 1218.50 ( 490.75–2476.50) 1219 (359–3001) 1218 (630–2373) 0.84
LDH (U/L) 735 (499–1236) 960 (564–1560) 643 (458–953) 0.005
D-dimer (µg/mL) 3.6 (1.49–13.48) 3.34 (1.43–18.80) 3.75 (1.54–11.02) 0.80
C-Reactive Protein (mg/mL) 128 (64–256) 128 (64–256) 128 (48–256) 0.1
Data are median (IQR) or n (%), *p- values were calculated by Mann–Whitney U test, χ2 test, or Fisher's exact test, as appropriate. ICU:Intensive Care Unit, MV:Mechanical Ventilation, CVC: Central Venous Catheter, DM:Diabetese mellitus, HTN:Hypertension, ACS: Acute coronary syndrome, COPD: Chronic obstructive pulmonary disease, LDH: lactate dehydrogenase.
The mean age of total patients was 61 ± 16 years, ranging from 18 to 99 years. Of the total patients, 64 (60.4%) were male and 42 (39.6%) were female. Our results showed about 48% of hospitalized COVID-19 patients during this study period had urinary tract, respiratory system, and bloodstream co-infections with Candida species. According to statistical analysis of the data, there was a significant difference between patients with and without Candida spp. in terms of age (66 vs. 57 years; p-value: 0.003),whereas no significant difference was found in sex distribution between the two groups of patients. In patients with Candida positive culture, the median time from hospital admission to positive fungal culture was 6 days, ranging from 2 to 36 days (IQR: 4–11). As far as the time of staying in hospital in patients is concerned, there was not found a statistical difference in length of hospitalization between two groups (p-value: 0.9).
Compared to patients without Candida positive culture, a higher percentage of patients with Candida co-infections were admitted in the ICU ( 47.1% vs 27.3%; p-value:0.03), used mechanical ventilation (47.1% vs 25.5; p-value:0.02) and used urinary catheter (54.9% vs 30.9%; p-value:0.013). Also, the median time of urinary catheterization in patients with Candida spp. positive culture was 5 days, with a significant difference in comparison to patients without Candida spp. positive culture (p-value: 0.036).Although hypertension (HTN) and diabetes mellitus (DM) were the most common comorbidities among the study population, there was not a statistically significant association between the presence of Candida co-infection and HTN/DM in COVID-patients. In regards to other underlying conditions evaluated in this study, significant differences were observed in the presence of chronic obstructive pulmonary disease (COPD), heart failure, and solid organ transplantation. Among patients with COVID-19, those with Candida positive cultures showed a higher proportion of bacterial co-infection (60.8% vs. 32.7; p-value: 0.004). Moreover, in 24 of 45 patients with positive culture of Candida spp. from respiratory samples, isolation of bacterial species was also reported. Enterococcus and Acinetobacter spp. were the most commonly identified bacteria.
Regarding treatments, 93.4% of patients received antibiotics, 96.2% received high doses of corticosteroids, 82.1% received antiviral treatments, and 47.2% received antifungals as empirical therapy. Toclizumab, a recombinant anti-interleukin-6 receptor (IL-6R) monoclonal antibody, was prescribed for 19.8% of patients. Moreover, fluconazole was the most common antifungal agent used as empirical therapy in hospitalized patients with COVID-19 (31/50).
Antifungal treatment was administered to 34/51 patients with a Candida positive culture. The most commonly prescribed antifungal agents were as follows: fluconazole(12/34), combination therapy with caspofungin and fluconazole (6/34), liposomal amphotericin B (4/34), and nystatin (3/34). Although a higher proportion of in-hospital mortality was observed in patients with Candida positive cultures than in patients without Candida positive cultures, no significant difference in mortality rate was observed between the two groups (35.3% vs 20%; p-value = 0.07).
Regression analysis demonstrates that odds for Candida co-infection were higher in patients who admitted in ICU, had mechanical ventilatory supports, urinary catheterization, bacterial infection and patients with severe /critical diseases (Table 2). With regards to underlying conditions, HTN, COPD, and heart failure are associated with an increased risk of the development of Candida co-infection among patients with COVID-19. Moreover, patients who had severe /critical SARS-CoV-2 infection were at a greater risk for Candida co-infections.Table 2 Risk factors associated with isolation of Candida spp. in hospitalized COVID-19 patients (N = 106)
Univariable OR (95% CI) p-value Multivariable OR (95% CI) p-value
Demographics characteristics
Age,yearsa 1.03 (1.01–1.06) 0.005 – –
Clinical interventions
ICU admission 2.37 (1.05–5.32) 0.003 – –
MV 2.6 (1.14–5.9) 0.02 – –
Urinary catheterization 2.72 (1.22–6.02) 0.01 – –
Comorbid and clinical conditions
Heart failure 5.31 (2.15–13.8) < 0.001 7.37 (2.4–22.5) < 0.001
HTN 2.26 (0.99–5.16) 0.05 – –
COPD 5.33 (1.4–20.19) 0.014 – –
Solid organ transplantation 0.22 (0.07–0.65) 0.007 – –
Bacterial co-infection 3.18 (1.43–7.06) 0.004 4.97 (1.8–13.6) 0.002
Disease severityb 2.4 (1.06–5.46) 0.036
Treatments
Empirical antifungal therapy 2.5 (1.14–5.46) 0.02 3.17 (1.1–8.4) 0.02
Laboratory findings
Neutrophil (%) 1.04 (0.99–1.09) 0.07 – –
C-Reactive protein 1 (0.99–1) 0.09 – –
aPer 1 unit increase
bSevere/Critical versus Mild/Moderte disease
OR Odds Ratio, CI Confidence Interval, ICU Intensive Care Unit, MV Mechanical Ventilation, HTN Hypertension, COPD Chronic Pulmonary Obstructive Diseas
In univariable analysis, the odds of in-hospital mortality was higher in patients with positive Candida culture (RR: 2.18, 95% confidence interval: 0.9, 5.2). Moreover, a multiple logistical regression showed that heart failure, bacterial co-infection, and prescribed antifungals as empirical treatment were significant risk factors for Candida co-infection among hospitalized COVID-19 patients.
Overall, 67 episodes of Candida positive cultures were documented from different clinical samples (45 teracheal aspirate samples, 20 urine samples, and 2 blood samples) collected from 106 COVID-19 patients during the study period. With regards to the categorization of fungal infections according to European Organisation for Research and Treatment of Cancer and Mycoses Study Group (EORTC-MSG), two patients had proven invasive candidiasis [18]. Moreover, 45 patients were identified with Candida airway colonization and 20 patients with candiduria. Also, 4 samples (2 urine and 2 teracheal aspirate) presented mixed infection. Of the patients with Candida positive culture, 11 patients showed multiple episodes of Candida infections during their hospitalization. Totally, 71 Candida spp. were isolated from 51 COVID-19 patients with Candida positive culture as follows: C. albicans (n:49, 69%), C.glabrata (n:14, 19.7%), C. tropicalis (n: 7, 9.9%), and C. dubliniensis (n:1, 1.4%) (Table 3).Table 3 Species distribution of Candida isolates from different clinical specimens
Candida isolates Tracheal aspirate Urine Blood Total
C.albicans 31 16 2 49
C.glabrata 10 4 – 14
C.tropicalis 5 2 – 7
C.dubliniensis 1 – – 1
Total 47 22 2 71
The categorization of all Candida isolates in this study was based on obtained MICs according to recommended breakpoints by CLSI as shown in Table 4. Moreover, in-vitro antifungal susceptibility patterns of all isolates are shown in Table 5 in detail.Table 4 MIC interpretation of tested antifungal drugs for Candida spp. recovered from COVID-19 patients
Antifungal drugs
Candidia species Susceptibility profile FLZ ITZ VRZ CSP PSZ AMB NYS
C. albicans (n = 49) ≤ ECV 47 39 27 49 32 49
> ECV 2 10 22 0 17 0
S 39 29 34 49 NA 49
SDD/I 6 10 8 0 NA 0
R 4 10 7 0 NA 0
C. glabrata (n = 14) ≤ ECV 12 14 12 14 13 13
> ECV 2 0 2 0 1 1
S 0 7 14 NA 14
SDD/I 14 5 0 NA 0
R 0 2 0 NA 0
C. tropicalis (n = 7) ≤ ECV 6 6 5 7 6 7
> ECV 1 1 2 0 1 0
S 6 4 5 7 NA 7
SDD/I 0 2 2 0 NA 0
R 1 1 0 0 NA 0
C.dubliniensis (n = 1) ≤ ECV 1 0 1 1 1 1
> ECV 0 1 0 0 0 0
S 1 1 NA 1
SDD/I 0 0 NA 0
R 0 0 NA 0
ECV Epidemiological cut-off value, S Susceptible, SDD/I Susceptible-dose-dependant/Intermediate, R Resistant, FLZ fluconazole, ITZ itraconazole, VRZ voriconazole, CSP caspofungin, PSZ posaconazole, AMB amphotericin B, NYS nystatin, NA not applicable, MIC ≤ ECV wild-type, MIC > ECV non-wild-type
Table 5 In-vitro antifungal susceptibility patterns of Candida spp. recovered from COVID-19 patients using micro broth dilution method
Candida spp. (71) Antifungal drug Range GMa MIC50 MIC90 MICb (µg/ml)
0.03 0.06 0.12 0.25 0.5 1 2 4 8 16 32
C.albicans (49) Fluconazole 0.12–16 0.91 0.5 4 1 10 15 7 6 6 2 2
Itraconazole 0.03–4 0.17 0.12 1 4 14 11 9 1 7 3
Voriconazole 0.03–8 0.11 0.03 2.2 18 9 7 4 4 1 4 2
Posaconazole 0.03–1 0.07 0.06 0.3 16 16 6 6 3 2
Caspofungin 0.03–0.06 0.03 0.03 0.06 40 9
Nystatin 0.06–8 1.16 2 4 1 13 6 2 8 18 1
Amphotericin B 0.06–2 0.37 0.5 1 1 4 16 20 6 2
C.glabrata (14) Fluconazole 0.12–32 1.55 1 6.8 1 4 3 4 1 1
Itraconazole 0.03–1 0.16 0.12 0.8 2 3 2 3 2 2
Voriconazole 0.03–1 0.09 0.06 0.4 6 2 2 2 1 1
Posaconazole 0.03–2 0.08 0.06 0.3 4 4 4 1 1
Caspofungin 0.03–0.12 0.03 0.03 0.12 11 1 2
Nystatin 0.25–8 0.8 0.3 6.8 7 1 1 1 2 2
Amphotericin B 0.06–2 0.4 0.5 1 1 4 3 5 1
C.tropicalis (7) Fluconazole 0.25–8 0.74 NA NA 2 3 1 1
Itraconazole 0.06–2 0.18 NA NA 2 2 2 1
Voriconazole 0.03–0.5 0.15 NA NA 2 1 2 1 1
Posaconazole 0.03–0.25 0.08 NA NA 1 3 2 1
Caspofungin 0.03–0.06 0.03 NA NA 6 1
Nystatin 0.5–4 1.64 NA NA 2 1 1 3
Amphotericin B 0.25–1 0.55 NA NA 2 2 3
C.dubliniensis (1) Fluconazole – – – – 1
Itraconazole – – – – 1
Voriconazole – – – – 1
Posaconazole – – – – 1
Caspofungin – – – – 1
Nystatin – – – – 1
Amphotericin B – – – – 1
aGM Geometric Mean, bMIC Minimum Inhibitory Concentration, NA not applicable
All Candida isolates presented low MIC values for CSP (MIC range: 0.03–0.12) that classified as susceptible according to M-60 CLSI breakpoints. Resistant to itraconazole were observed with a high proportion among Candida species regarding breakpoints documented by M27-S3. According to the M59 CLSI document, all C. albicans isolates were wild-type (WT) for AMB. Among non-albicans Candida species, one C.glabrata isolate had a non-wild-type (NWT) phenotype for AMB. In regards to another polyene drug, nystatin, all Candida isolates were categorized as susceptible. Although no resistant isolate was found among all Candida species against posaconazole, few isolates were resistant to other azoles tested in this study. One C. albicans isolate was cross-resistant to VRZ (≥ 1 μg/ml) and ITZ (≥ 1 μg/ml), and also two C.albicans were cross-resistant toFLZ (≥ 8 μg/ml), VRZ (≥ 1 μg/ml) and ITZ (≥ 1 μg/ml). Cross-resistant species against FLZ (≥ 8 g/ml) and ITZ (≥ 1 g/ml) were also found in one C.albicans and one C.tropicalis isolate.
Discussion
The presence of fungal co-infections is considered an important complication in COVID-19 hospitalized patients by association with increased mortality [19–21]. COVID-19-Associated Candidiasis has been studied in different parts of the world with a range of 0.7–23.5% [4]. According to obtained data, almost half of hospitalized patients with COVID-19 have shown at least one positive culture of urinary, respiratory, or blood specimens for Candida species during their stay in hospital. To the best of our knowledge, this is the first comprehensive study on COVID-19-associated candidiasis (CAC) that included all clinical samples collected from patients with COVID-19 in Iran. It should be noted that the inclusion of both invasive and non-invasive candidiasis cases in this study led to a considerable difference in our reported rates compared to earlier studies.
As we know, Candida spp. are a part of the human microbiome that resides in different sites. So, dysregulation of the immune system following infection with SARS-CoV-2 leaves patients vulnerable to Candida superinfections. In general, previous evidence supports our findings that ICU-admission, mechanical ventilation, and urinary catheterization represent important risks for the development of candidiasis [4, 22–28].
Moreover, underlying conditions in COVID-19 patients, including DM, hypertension, and organ failure, in addition to a number of various antimicrobial and immunosuppressive medications, may be related to fungal co-infections [21, 29–32]. However, in the present work we identified HTN, COPD, hepatic and heart failures as risk factors for CAC that are in consistent with other previous studies in Iran and other parts of the world [25, 27, 28, 33].
Similar to other investigations, our results revealed that more severe conditions of pneumonia associated with SARS-Cov-2 infection increased the probability of fungal co-infections [25, 34]. Certainly, such therapeutic approaches in severe cases of COVID-19 patients as steroids, mechanical ventilation, broad-spectrum antimicrobial therapy, and ICU admission make them more susceptible to candidiasis, which has been previously reported in the Iranian population [27, 28, 35]. However, further studies are needed to investigate the possibility that Candida infection may increase the severity of the disease in patients with COVID-19.
Despite the fact that isolation of Candida spp. from respiratory samples occurs particularly in patients on mechanical ventilation, the link between this colonization and pneumonia is still controversial. Because the definitive diagnosis of Candida pneumonia is confirmed by examining the invasive form of the fungus in biopsy tissue, that is a limited diagnostic procedure. Since the lung is the main affected organ in COVID-19 patients, Candida colonization following immunosuppressive treatments and antibacterial medications can lead to overgrowth and morphogenesis of Candida spp. due to induced defects in cell-mediated immunity of the respiratory epithelium [36–39].
According to our research findings, patients with bacterial co-infection were at higher risk for Candida infections. Also, a high proportion of COVID-19 patients received antibacterial treatment, which is in agreement with the previous literature [5, 40]. Aligned to these results, it could be explained that overprescribing of broadspectrum antibiotics in COVID‐19 patients causes an imbalance of normal flora and overgrowth of Candida, which could result in Candida co-infections [41]. Furthermore, the coexistence of bacterial and fungal infections could contribute to increased pathogenicity, host damage, and inflammation [42]. Our findings showed that mortality among COVID-19 patients with Candida co-infection was higher in comparison to COVID-19 alone, which is in agreement with many previous studies [25, 34, 43–45].
Efficacy of empirical or preemptive antifungal therapy in high-risk patients has been studied in a number of researches [46, 47]. In nearly half of our research population, antifungal agents were provided as empirical therapy, despite the fact that there is no treatment guideline for fungal co-infections in COVID-19-positive individuals. Excessive use of empiric antifungal regimens not only has not been proven to be safe or beneficial but also could lead to resistance and therapeutic failures in patients who have previously been exposed, in addition to imposing an economic burden. Obviously, drug-drug interactions, toxicity, and adverse side effects of antifungals used in COVID-19 patients should all be considered [38].
Based on recent studies, C.albicans is known as the most prevalent species responsible for candidiasis in COVID-19 patients [48–51]. However, the identification of C. auris, a multi-drug resistant species, has been reported in previous studies with a high mortality rate [25]. In accordance with the findings presented in previous reports, C. albicans was the major species for candidiasis in COVID-19 patients in this study. Moreover, the species distribution of isolates from CAC with the highest proportion of C.albicans and C.glabrata is in accordance with previous reports from different countries[31, 34, 49, 51–54]. Although it is known that C.albicans accounts for half the isolates recovered from different types of Candida infections, it should be noted that non-albicans Candida species have emerged as prominent species in recent years [55]. Importantly, outbreaks of C.auris infections have been reported among COVID-19 patients in some regions of the world [25, 56, 57].
Although many reports have been published, there are a few studies about the susceptibility of fungal isolates from COVID-19 patients. Also, there have been few studies about the susceptibility patterns of Candida species isolated from non-invasive candidiasis. In total, resistance to azole drugs was observed among Candida species recovered from different samples in our study in spite of their high level of susceptibility to all tested antifungals. As reported from previous studies in Shiraz, the current study found azole-resistant Candida species among clinical isolates [58, 59]. Moreover, assessment of the azole susceptibility profiles of Candida isolates showed that C. albicans had cross-resistant profile to azoles. [38]Also, the resistant rate among C.glabrata was low in agreement with previous reports from Iran [60, 61]. Of note, itraconazole resistance among Candida species has been reported in previous studies[62–65]. Collectively, various degrees of resistance to azole antifungals have been declared previously in many reports from Iran [66, 67]. As a consequence, the presence of azole-resistant species could be a result of high background usage of azoles as empirical or first-line choice treatments and repeated exposure in our reviewed medical center.
Although posaconazole has no defined breakpoint in CLSI documents, it totally showed low MIC values for Candida species in the current study, which is consistent with the other studies [67, 68]. Considering the potent activity of posaconazole against Candida species, administration of this agent is proposed for the treatment of Candida infections. Additionally, the efficacy of posaconazole for treatment of mucormycosis, as a post-COVID-19 fungal infection, makes this azole antifungal an efficacious choice for treatment of other mycoses than yeast infections. In this survey, the most potent antifungal agent was caspofungin, with low MIC values for all species in accordance with prior research [32, 65, 69]. From the perspective of availability and cost-effectiveness, echinocandins appear to be an appropriate drug of first choice to define the most effective approach to the management of candidiasis in our country.
Some limitations must be acknowledged. In this study,we analyzed the available information for a small research population in one center that follow-up of patients after discharge was not included. The unavailability of new antifungals such as Isavuconazole in our country precludes the possibility of evaluating their efficacy in therapeutic procedures.
In conclusion, the results of our study declared some fundamental issues for patients infected with SARS-CoV-2 that had Candida co-infections. The main finding of this study reveals a high rate of Candida co-infections among hospitalized COVID-19 patients. In addition, comorbidities such as heart failure, HTN, COPD, bacterial infections as well as therapeutic interventions including catheterization, mechanical ventilation, and ICU admission increased the risk of infection by Candida spp. in COVID-19 patients. The isolation of Candida species from urinary/respiratory tracts and the bloodstream led to poor clinical outcomes, which presented as a higher in-hospital mortality rate. Also, obtained data clarified that empiric antifungal therapy did not achieve expected effectiveness. Finally, our results could assist infectious disease specialists with a better understanding of COVID-19-associated candidiasis. Definitely, selection of the appropriate antifungal therapies considering the species distribution and susceptibility patterns of causative agents can result in more effective therapeutic interventions and desired outcomes.
Author Contributions
SY: Conceptualization,Investigation,Writing—original draft, Writing—review and editing. PC: Formal analysis. MA, ZZ, AJ, EE, ZM, MG: Investigation. HN, SA, BG, MHA, MM: Validation, Data curation. MS: Supervision, Writing—review and editing, Project administration. KZ: Project administration,Supervision, Resources, Funding acquisition, review and editing. All authors read and approved the final manuscript.
Funding
This work was supported by the Research Council of Shiraz University of Medical Sciences [Grant No. 26040].
Declarations
Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Zhou F Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study Lancet 2020 395 10229 1054 1062 10.1016/S0140-6736(20)30566-3 32171076
2. Silva DL Fungal and bacterial coinfections increase mortality of severely ill COVID-19 patients J Hosp Infect 2021 113 145 154 10.1016/j.jhin.2021.04.001 33852950
3. Shafiekhani M Bacterial and fungal co-infections with SARS-CoV-2 in solid organ recipients: a retrospective study Virol J. 2022 19 1 1 7 10.1186/s12985-022-01763-9 34980196
4. Arastehfar A COVID-19-associated candidiasis (CAC): an underestimated complication in the absence of immunological predispositions? J Fungi. 2020 6 4 211 10.3390/jof6040211
5. Chen N Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study Lancet 2020 395 10223 507 513 10.1016/S0140-6736(20)30211-7 32007143
6. Chen X The microbial coinfection in COVID-19 Appl Microbiol Biotechnol 2020 104 18 7777 7785 10.1007/s00253-020-10814-6 32780290
7. Song G Liang G Liu W Fungal co-infections associated with global COVID-19 pandemic: a clinical and diagnostic perspective from China Mycopathologia 2020 185 4 599 606 10.1007/s11046-020-00462-9 32737747
8. Rolling T Hohl TM Zhai B Minority report: The intestinal mycobiota in systemic infections Curr Opin Microbiol 2020 56 1 6 10.1016/j.mib.2020.05.004 32599521
9. World health organization (WHO). WHO coronavirus (COVID-19) Dashboard [Internet]. Available from: https://covid19.who.int/.
10. Pascarella G COVID-19 diagnosis and management: a comprehensive review J Intern Med 2020 288 2 192 206 10.1111/joim.13091 32348588
11. Jafari M Exoenzyme activity and possibility identification of Candida dubliniensis among Candida albicans species isolated from vaginal candidiasis Microb Pathog 2017 110 73 77 10.1016/j.micpath.2017.06.024 28642006
12. Mirhendi H Differentiation of Candida albicans and Candida dubliniensis using a single-enzyme PCR-RFLP method Jpn J Infect Dis 2005 58 4 235 16116258
13. Wayne, P., Clinical and Laboratory Standards Institute: Reference method for broth dilution antifungal susceptibility testing of yeasts; approved standard. CLSI document M27-A3 and Supplement S, 2008;3:6–12.
14. Wayne, P., Reference method for broth dilution antifungal susceptibility testing of yeasts, approved standard. CLSI document M27-A2, 2002.
15. Wayne P Reference method for broth dilution antifungal susceptibility testing of yeasts Clin Lab Standards Inst. 2008 3 M27 M33
16. Clinical and L.S. Institute, Performance standards for antifungal susceptibility testing of yeasts. CLSI supplement M60, 2017.
17. Clinical and L.S. Institute, Epidemiological cutoff values for antifungal susceptibility testing. 2016, Clinical and Laboratory Standards Institute Wayne, PA.
18. De Pauw B Revised definitions of invasive fungal disease from the European organization for research and treatment of cancer/invasive fungal infections cooperative group and the national institute of allergy and infectious diseases mycoses study group (EORTC/MSG) consensus group Clin Infect Dis 2008 46 12 1813 1821 10.1086/588660 18462102
19. Falces-Romero I Isolation of Aspergillus spp. in respiratory samples of patients with COVID-19 in a Spanish tertiary care hospital Mycoses 2020 63 11 1144 1148 10.1111/myc.13155 32749040
20. Zhou P Bacterial and fungal infections in COVID-19 patients: a matter of concern Infect Control Hosp Epidemiol 2020 41 9 1124 1125 10.1017/ice.2020.156 32317036
21. Norberg C Candida infections associated with COVID-19: an underestimated risk WJPPS 2021 10 9 48 64
22. Clancy CJ Nguyen MH Coronavirus disease 2019, superinfections, and antimicrobial development: what can we expect? Clin Infect Dis 2020 71 10 2736 2743 10.1093/cid/ciaa524 32361747
23. Chen G Clinical and immunological features of severe and moderate coronavirus disease 2019 J Clin Investig 2020 130 5 2620 2629 10.1172/JCI137244 32217835
24. Huang C Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China Lancet 2020 395 10223 497 506 10.1016/S0140-6736(20)30183-5 31986264
25. Chowdhary A Multidrug-resistant Candida auris infections in critically Ill coronavirus disease patients, India, April–July 2020 Emerg Infect Dis 2020 26 11 2694 10.3201/eid2611.203504 32852265
26. Bishburg E Fungemia in covid-19 ICU patients, a single medical center experience J Med Virol 2021 93 5 2810 2814 10.1002/jmv.26633 33107617
27. Salehi M Opportunistic fungal infections in the epidemic area of COVID-19: a clinical and diagnostic perspective from Iran Mycopathologia 2020 185 4 607 611 10.1007/s11046-020-00472-7 32737746
28. Nazari T COVID-19-associated fungal infections in Iran: A systematic review PLoS ONE 2022 17 7 e0271333 10.1371/journal.pone.0271333 35816494
29. Seagle EE The landscape of candidemia during the COVID-19 pandemic Clin Infect Dis 2021 74 5 802 811 10.1093/cid/ciab562
30. Ezeokoli OT Gcilitshana O Pohl CH Risk factors for fungal co-infections in critically ill COVID-19 patients, with a focus on immunosuppressants J Fungi. 2021 7 7 545 10.3390/jof7070545
31. Antinori S Tocilizumab for cytokine storm syndrome in COVID-19 pneumonia: an increased risk for candidemia? Autoimmun Rev 2020 19 7 102564 10.1016/j.autrev.2020.102564 32376396
32. Kordalewska M Antifungal drug susceptibility and genetic characterization of fungi recovered from COVID-19 patients J Fungi. 2021 7 7 552 10.3390/jof7070552
33. Herc E et al. Characteristics and outcomes of COVID-19 patients with fungal infections. In: Open Forum Infectious Disease. Oxford University Press 2020.
34. Segrelles-Calvo G Candida spp. co-infection in COVID-19 patients with severe pneumonia: Prevalence study and associated risk factors Respir Med 2021 188 106619 10.1016/j.rmed.2021.106619 34555702
35. Vinayagamoorthy K Pentapati KC Prakash H Prevalence, risk factors, treatment and outcome of multidrug resistance Candida auris infections in Coronavirus disease (COVID-19) patients: a systematic review Mycoses 2022 65 6 613 624 10.1111/myc.13447 35441748
36. Pittet D Candida colonization and subsequent infections in critically ill surgical patients Ann Surg 1994 220 6 751 10.1097/00000658-199412000-00008 7986142
37. Shoham S Levitz SM The immune response to fungal infections Br J Haematol 2005 129 5 569 582 10.1111/j.1365-2141.2005.05397.x 15916679
38. Silva LN Fungal infections in COVID-19-positive patients: a lack of optimal treatment options Curr Top Med Chem 2020 20 22 1951 1957 10.2174/156802662022200917110102 33040728
39. Benelli JL Fungal bloodstream co-infection by trichosporon asahii in a COVID-19 critical patient: case report and literature review Mycopathologia 2022 187 4 397 404 10.1007/s11046-022-00637-6 35661958
40. Goncalves Mendes Neto A Bacterial infections and patterns of antibiotic use in patients with COVID-19 J Med Virol 2021 93 3 1489 1495 10.1002/jmv.26441 32808695
41. Riad A et al. Oral candidiasis in non‐severe COVID‐19 patients: call for antibiotic stewardship. Oral Surg. 2020;15(3):465–466.
42. Morales DK Hogan DA Candida albicans interactions with bacteria in the context of human health and disease PLoS Pathog 2010 6 4 e1000886 10.1371/journal.ppat.1000886 20442787
43. Cultrera R Co-infections in critically ill patients with or without COVID-19: a comparison of clinical microbial culture findings Int J Environ Res Public Health 2021 18 8 4358 10.3390/ijerph18084358 33923992
44. Cruz AB LeRose J Chopra T Comparison of Outcomes in Candidemia Between COVID-19 and Non–COVID-19 Patients Antimicrob Steward Healthc Epidemiol. 2021 1 S1 s58 s58 10.1017/ash.2021.112
45. Kayaaslan B Characteristics of candidemia in COVID-19 patients; increased incidence, earlier occurrence and higher mortality rates compared to non-COVID-19 patients Mycoses 2021 64 9 1083 1091 10.1111/myc.13332 34085319
46. Whitney L Effectiveness of an antifungal stewardship programme at a London teaching hospital 2010–16 J Antimicrob Chemother 2019 74 1 234 241 10.1093/jac/dky389 30376118
47. Morrissey C Consensus guidelines for the use of empiric and diagnostic-driven antifungal treatment strategies in haematological malignancy, 2014 Intern Med J 2014 44 12b 1298 1314 10.1111/imj.12596 25482742
48. Ventoulis I Bloodstream infection by Saccharomyces cerevisiae in two COVID-19 patients after receiving supplementation of Saccharomyces in the ICU J Fungi. 2020 6 3 98 10.3390/jof6030098
49. White PL A national strategy to diagnose coronavirus disease 2019–associated invasive fungal disease in the intensive care unit Clin Infect Dis 2021 73 7 e1634 e1644 10.1093/cid/ciaa1298 32860682
50. Posteraro B Pan-echinocandin-resistant Candida glabrata bloodstream infection complicating COVID-19: a fatal case report J Fungi. 2020 6 3 163 10.3390/jof6030163
51. Al-Hatmi AM COVID-19 associated invasive candidiasis J Infect 2021 82 2 e45 e46 10.1016/j.jinf.2020.08.005
52. White PL et al. Candidemia in Coronavirus Disease 2019: A Link to Disease Pathology or Increased Clinical Pressures? Clin Infect Dis. 2020;73(9):e2839–2841
53. Arastehfar A Candidemia among Iranian patients with severe COVID-19 admitted to ICUs J Fungi 2021 7 4 280 10.3390/jof7040280
54. Salehi M Oropharyngeal candidiasis in hospitalised COVID-19 patients from Iran: species identification and antifungal susceptibility pattern Mycoses 2020 63 8 771 778 10.1111/myc.13137 32609906
55. Kullberg BJ Arendrup MC Invasive candidiasis N Engl J Med 2015 373 15 1445 1456 10.1056/NEJMra1315399 26444731
56. Villanueva-Lozano H Outbreak of Candida auris infection in a COVID-19 hospital in Mexico Clin Microbiol Infect 2021 27 5 813 816 10.1016/j.cmi.2020.12.030 33429028
57. Almeida JND Emergence of Candida auris in Brazil in a COVID-19 intensive care unit J Fungi. 2021 7 3 220 10.3390/jof7030220
58. Arastehfar A Clinical and microbiological features of candiduria in critically ill adult patients in Shiraz, Iran (2016–2018): Deviations from international guidelines and fluconazole therapeutic failure Med Mycol 2021 59 6 600 607 10.1093/mmy/myaa092
59. Arastehfar A Epidemiology of candidemia in Shiraz, southern Iran: A prospective multicenter study (2016–2018) Med Mycol 2021 59 5 422 430 10.1093/mmy/myaa059 32692816
60. Arastehfar A Low level of antifungal resistance in Iranian isolates of Candida glabrata recovered from blood samples in a multicenter study from 2015 to 2018 and potential prognostic values of genotyping and sequencing of PDR1 Antimicrob Agents Chemother 2019 63 7 e02503 e2518 10.1128/AAC.02503-18 30936110
61. Amanloo S Drug susceptibility profile of Candida glabrata clinical isolates from Iran and genetic resistant mechanisms to caspofungin Rev Iberoam Micol 2018 35 2 88 91 10.1016/j.riam.2018.01.002 29685375
62. Badiee P Susceptibility of Candida species isolated from immunocompromised patients to antifungal agents East Mediterr Health J 2011 17 5 425 430 10.26719/2011.17.5.425 21796956
63. Haddadi P et al. Yeast colonization and drug susceptibility pattern in the pediatric patients with neutropenia. Jundishapur J Microbiol. 2014;7(9): e11858
64. Badiei P et al. Molecular identification and in-vitro susceptibility of Candida albicans and C. dubliniensis isolated from immunocompromised patients. 2009.
65. Erami M Clinical impact of Candida respiratory tract colonization and acute lung infections in critically ill patients with COVID-19 pneumonia Microb Pathog 2022 166 105520 10.1016/j.micpath.2022.105520 35405278
66. Badiee P Antifungal susceptibility testing of Candida species isolated from the immunocompromised patients admitted to ten university hospitals in Iran: comparison of colonizing and infecting isolates BMC Infect Dis 2017 17 1 1 8 10.1186/s12879-017-2825-7 28049444
67. Mahmoudabadi AZ, Rezaei-Matehkolaei A, Ghanavati F. The susceptibility patterns of Candida species isolated from urine samples to posaconazole and caspofungin. Jundishapur J Microbiol. 2015;8(3).
68. Castanheira M Isavuconazole and nine comparator antifungal susceptibility profiles for common and uncommon Candida species collected in 2012: application of new CLSI clinical breakpoints and epidemiological cutoff values Mycopathologia 2014 178 1 1 9 10.1007/s11046-014-9772-2 24952015
69. Cataldo MA Incidence of bacterial and fungal bloodstream infections in COVID-19 patients in intensive care: An alarming “collateral effect” J Glob Antimicrob Resist. 2020 23 290 10.1016/j.jgar.2020.10.004 33130024
| 36495418 | PMC9739341 | NO-CC CODE | 2022-12-14 23:29:54 | no | Mycopathologia. 2022 Dec 10;:1-12 | utf-8 | Mycopathologia | 2,022 | 10.1007/s11046-022-00694-x | oa_other |
==== Front
Sex Cult
Sex Cult
Sexuality & Culture
1095-5143
1936-4822
Springer US New York
10046
10.1007/s12119-022-10046-y
Original Article
Minority Stressors and Attitudes Toward Intimate Partner Violence Among Lesbian and Gay Individuals
http://orcid.org/0000-0002-5280-1315
Eric S. Reyes Marc [email protected]
1
Camille M. Alday Angeli 1
Jay J. Aurellano Alexa 1
Raven R. Escala Sahara 1
Ermelo V. Hernandez Piolo 1
Esrom P. Matienzo John 1
Marian R. Panaguiton Khim 1
Charmaine C. Tan Angeli 1
Zsila Ágnes 2
1 grid.412775.2 0000 0004 1937 1119 Department of Psychology, College of Science, University of Santo Tomas, Manila, Philippines
2 grid.425397.e 0000 0001 0807 2090 Institute of Psychology, Pázmány Péter Catholic University, Budapest, Hungary
10 12 2022
121
2 8 2022
12 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.
Sexual minority individuals experience more intimate partner violence (IPV) than those in heterosexual relationships. Issues of mistrust, stigma, and anticipation of abuse contribute to these rates. Lesbian and gay IPV victims have distinct experiences from their abuses with exposure to homophobia, heterosexism, discrimination, and threats of sexual disclosure, among others. These unique and additive minority stressors can lead to adverse health concerns, increase vulnerability to victimization, and elevate abuse perpetration. This study aimed to investigate whether experiences of minority stressors are associated with attitudes toward intimate partner violence among a sample of 240 lesbian and gay Filipinos (155 lesbian and 85 gay participants) aged 20 to 40. Through convenience sampling, lesbian and gay Filipinos completed the Sexual Minority Stress Scale (SMSS) and Intimate Partner Violence Attitude Scale-Revised (IPVAS-Revised). Comparing the minority stressors levels among the participants, lesbians expressed higher expectations of rejection, while gay men experienced more sexual minority adverse events. Lesbians also reported higher satisfaction with outness. Regarding IPV, gay men expressed slightly more favorable attitudes toward abuse, which could make them at risk of becoming victims or perpetrators. Internalized homophobia was associated with more favorable attitudes toward abuse and control, indicating its contribution to more favorable IPV attitudes, although the explanatory power was modest.
Keywords
gay
lesbian
minority stressors
intimate partner violence
attitudes
Ágnes Zsila was supported by the ÚNKP-22-4 New National Excellence Program of the Ministry for Culture and Innovation from the source of the National Research, Development, and Innovation Fund.Ágnes Zsila was supported by the ÚNKP-22-4 New National Excellence Program of the Ministry for Culture and Innovation from the source of the National Research, Development, and Innovation Fund.
==== Body
pmcIntroduction
Intimate Partner Violence (IPV) cases continue to increase yearly (Centers for Disease Control and Prevention, 2020). On average, one in nine men and one in four women experience IPV, including physical, sexual, psychological, and economic violence, as well as stalking (Huecker et al., 2021). The same study revealed that IPV accounted for 15% of all violent crimes recorded. In the Philippines, one in four Filipino women aged 15 to 49 has experienced abuse from their partner or husband (Philippine Statistics Authority, 2018), while 12 to 15 Filipino men out of every 100 couples have experienced it (Khidhir, 2020). As COVID-19 continues, the reports on IPV among Filipinos tripled, becoming the pandemic’s silent consequence (Galang, 2021). Lockdown implementation, strict stay-at-home orders, curfew, lack of public transportation, and other measures to curb the virus significantly restricted a person’s opportunity to seek help.
The Forms of Intimate Partner Violence
According to the Centers for Disease Control and Prevention [CDC] (2020), Intimate Partner Violence (IPV) behaviors include (a) physical violence, the intended use of physical force toward a partner that might result in injury, harm, or even death (Ali et al., 2016), (b) sexual violence, any sexual act a partner attempted or committed without the victim’s consent or someone unable to respond to the advances (Breiding et al., 2015) that also includes physically coercing a partner to have sexual intercourse, humiliating a partner through sexual acts, and harming them during sex (Ali et al., 2016), (c) stalking, behavior that depicts a pattern of repeated unwanted attention and contact from another person, resulting in concern for the victim’s safety (Breiding et al., 2015) and (d) psychological aggression, usage non-verbal and verbal communication against a victim that can hurt them emotionally and mentally (Postmus et al., 2018). Other IPV dimensions include financial, economic, social, and spiritual abuse (Hegarty et al., 1999; Dehan& Levi, 2009; Postmus et al., 2018), but these are not as established as those mentioned by the CDC.
IPV is a worldwide concern initially thought of as an issue that only exists in heteronormative relationships, but it happens in all kinds of relationships, including queerness (Harden et al., 2020). Sexual minorities experience IPV at higher rates than those in heterosexual relationships, and mistrust, stigma, and anticipation of abuse contribute to these rates (Russell & Sturgeon, 2018). Sexual minorities refer to individuals with gender identities, sexual orientations, and gender expressions that differ from most cultural norms and are usually composed of but not limited to lesbian, gay, bisexual, and transgender (LGBT) individuals (Cochat Costa Rodrigues et al., 2017).
LGBT IPV victims have distinct experiences from their abusers compared with their heterosexual counterparts, as they also encounter homophobia, heterosexism, transphobia, and threats to disclose sexuality, among others (Russell & Sturgeon, 2018). In his Minority Stress Model, Ilan Meyer (2015) proposed that the unique stressors experienced by sexual minorities are multiple and additive. Their stress is excessive because of prejudice, stigma, and discrimination, leading to various physical and mental health issues and risk-taking behaviors. Therefore, minority stress plays a vital role in the increased rates of IPV within the LGBT community (Finneran & Stephenson, 2014; Longobardi & Badenes-Ribera, 2017; Swan et al., 2019).
Furthermore, compared to heterosexual persons, lesbian and gay individuals were at higher risk of committing IPV or becoming victims themselves (AyhanBalik & Bilgin, 2019; Boston, 2019). The overall lifetime rates of IPV victimization are higher at 43.8% among lesbians and 26.0% among gay men in comparison to heterosexual women (35.0%) and heterosexual men (29.0%) (Walters et al., 2013). Walters and colleagues also reported that more than 75% of lesbians and 50% of gay men were victims of psychological abuse. Similarly, a study by Goldberg et al. (2013) estimated that 26.9% of gay men experienced IPV in their lifetimes, while another study by Messinger (2011) estimated a much lower prevalence of 3.1%. Messinger’s (2011) study only sampled 32 gay men, while Goldberg and colleagues had a sample of 415 gay men. Hence, substantial differences in the prevalence rates can be due to the differences in the sample size of these studies. Regarding lesbians, a report from Swift (2019, as cited in the Centers for Disease Control and Prevention [CDC], n.d.) found that their overall IPV lifetime rate is 44%, while Messinger (2011) found 3.6%. Despite these numbers, same-sex IPV cases are more likely to be ignored or unreported, especially in cultures that view same-sex relationships as immoral and unacceptable (Chong et al., 2013).
Attitudes towards intimate partner violence also influenced these same-sex IPV cases. They were closely linked to factors such as gender, gender roles, behaviors, cultural norms, and familial and societal views (Copp et al., 2019). Minority stressors are relevant in understanding IPV in same-sex relationships (Longobardi & Badenes-Ribera, 2017). As Jacobson et al. (2015) highlighted, there is a high attitudinal acceptance of intimate partner violence among the LGBT community. Specifically, male participants were most likely to incite verbal and physical victimization and justify violence perpetration.
Stigma Towards Intimate Partner Violence
However, public discussion of IPV among the LGBT community was silenced since it recognized such acts were stigmatizing and added to the existing oppression and social marginalization (Ard & Makadon, 2011; Calton et al., 2015; Rolle et al., 2018). Nonetheless, IPV remained eminent irrespective of marital status, age, and sexual orientation (Ali et al., 2016). Messinger (2011) found that gay men committed more IPV than heterosexual men. The high percentage of IPV among same-sex relationships resulted from various factors linked to minority stress. In a similar vein, Finneran and Stephenson (2014) found strong correlations between IPV and minority stress stemming from internalized homophobia, racism, and homophobic discrimination among people involved in same-sex relationships.
Stigmatization and Discrimination of the LGBTQI + Community in the Philippines
According to the Psychological Association of the Philippines’ (PAP) 2011 non-discrimination statement, Filipinos who identify with the LGBT community still experience stigma, prejudice, and discrimination. Even though the Philippines has a lower homonegativity attitude than neighboring countries such as Indonesia and Malaysia, 31% of the Filipino respondents in the most recent data of the World Values Survey (WVS) considered lesbian and gay sexual orientations as not justifiable or morally unacceptable (Manalastas et al., 2017). Religion is a significant factor in the public’s perception of the LGBT community. The vast majority of Filipinos are Roman Catholics. Catholicism and its teachings suppress the acceptance of sexual minorities as they deem same-sex marriage and homosexuality immoral practices (UNDP, 2014). Discrimination and stigma of the LGBT community are still present in the Philippine society in the following forms: bullying gay children, banning transgender individuals in business establishments, labeling gay and lesbian adults as “sinful” or “abnormal,“ comedic and sexually predatory media portrayals of gay men, sexual abuse of lesbians to “correct” their sexuality, and increasingly documented violence targeting people perceived to be part of the LGBT community (Manalastas & Torre, 2013). Because of these negative experiences, PAP (2020) reiterated its support for the fundamental human rights of all people, including sexual minorities.
Intimate Partner Violence in Filipino Sexual Minorities
Research by OutRight Action International (2018) indicated that Filipino lesbian and gay IPV victims tended to avoid seeking help from other people, police, and social or legal services because of the fear of revealing their relationships and families, humiliation, and receiving inappropriate reactions. These were the main reasons why cases of IPV within the Filipino LGBT community were unreported. Additionally, the Philippine government has no existing means for documenting and recording LGBT Violence (Commission on Human Rights of the Philippines, 2019). Thus, statistics on the number of lesbian and gay victims and perpetrators were unavailable. Moreover, only a few studies addressed IPV in the LGBT community, but none in the Philippines (Fehringer & Hindin, 2013). These studies focused on the relationship between minority stress and IPV and found a positive association between internalized homophobia and IPV (Badenes-Ribera et al., 2017; Finneran & Stephenson, 2014; Lewis et al., 2017; Stephenson & Finneran, 2016).
There is a lack of psychological research on the experiences of Filipino sexual minority individuals, and our study aimed to examine the relationship between sexual minority stressors and intimate partner violence attitudes among LG Filipinos to fill this gap. We hypothesized that incidents of sexual minority stressors by lesbian and gay Filipinos were associated with their attitudes towards IPV irrespective of their socio-demographic characteristics (i.e., sexual orientation, age, outness, and relationship status).
Method
Participants
The participants included a total of 240 self-identified Filipino lesbians (n = 155; 64.58%) and gay men (n = 85; 35.42%) aged 20 to 40 years old (M = 26 years old; SD = 5.41) with either undisclosed or disclosed sexual orientation. The minimum age was set explicitly at 20 since this is when the energy of emerging young adults is primarily devoted to developing romantic relationships (Kelley et al., 2015), and the risk for IPV victimization is most significant for people aged 20 to 40 (Rivara et al., 2009).
A non-probability convenience sampling method was used in gathering the participants who have experienced being in a relationship. They were not necessarily required to be in a relationship or have a partner at the time of data collection. Likewise, participants did not necessarily need to be perpetrators or victims of IPV or have any IPV history since we only assessed IPV attitudes. Those with no relationship history or with unanswered or missing items in the tests were excluded from the data analysis. Thus, from the 269 participants who voluntarily participated without remuneration, 29 were excluded resulting in a total of 240 LG Filipinos. Table 1 presents the socio-demographic characteristics of our study sample.
Table 1 Socio-demographic Characteristics of Study Sample
Characteristics n %
Assigned Sex at Birth
Female 160 66.67
Male 80 33.33
Gender
Cis Female 123 51.2
Cis Male 59 24.6
Third Gender/Nonbinary/Transgender 47 4.6
Prefer not to say 11 19.6
Sexual Orientation
Gay
Lesbian
Outness
85
155
35.42
64.58
Sexual orientation is out to someone 230 95.8
Sexual orientation is not out to someone 10 4.2
Outness Reception
Ambivalent/Uncertain 79 32.9
Negative 13 5.4
Positive 135 56.3
Prefer not to say 13 5.4
Current Relationship Status
In a relationship 151 62.92
Married 4 1.67
Single 85 35.42
Note. N = 240
Measures
Sexual Minority Stress Scale (SMSS). It is a 58-item self-report questionnaire developed by Goldblum et al. (unpublished manuscript), which was adapted and validated by Iniewicz et al. (2017. It assesses the minority stress levels of LGB individuals, which includes five subscales that measure proximal stressors: Internalized Homophobia (IH), Expectations of Rejection (ExR), Concealment (Clm), Satisfaction with Outness (SO), and Sexual Minority Negative Events (SMNE). The Satisfaction with Outness is further divided into (1) levels of disclosure of the person’s sexual orientation to others (SOa) and (2) degree of satisfaction with the disclosure (SOb). The SMNE has three categories: events related to the examined person, events that the person had witnessed or heard about, and items about infectious diseases. Meyer’s Sexual Minority Stress Model was the basis of all other subscales except for the SO subscale. The answers are given on a checklist and in 4 to 6-point Likert-type formats depending on the subscale. Sample items are, for IH, “Have you tried to stop being attracted to persons of the same sex?“ (ranging from1 Often to 4 Never); for ExR, “Most employees will not hire a person like you” (1 Strongly Agree, 2 Somewhat Agree, 3 Somewhat Disagree, 4 Strongly Disagree), for Clm, “I have concealed my sexual orientation by telling someone that I was straight or denying that I was LGB” (1 Not at all, 2 A little bit, 3 Somewhat, 4 Very much, 5 All the time), for SOa, “Are you out to your family about your sexual and gender identity?“ (Yes or No), for SOb, “How satisfied are you with your level of outness to your family?“ (ranging from 1 Very Dissatisfied to 6 Extremely Satisfied), and for SMNE (one for each category; checklist format), “I was treated unfairly by peers and siblings,“ “I heard negative statements about LGB or gender nonconforming people,“ and “I have been diagnosed with HIV or other chronic sexually transmitted diseases.“ In the SMSS, there is no total score, and each subscale is scored separately. The range of each subscale’s overall score differs: IH total score ranges from 10 to 40, ExR total score ranges from 6 to 24, Clm total score ranges from 6 to 30, SO total score ranges from 5 to 30, and SMNE total score ranges from 0 to 69. These total subscale scores are computed by adding the items of each subscale with question 10 of Internalized Homophobia reversely scored (1 = 4, 2 = 3, 3 = 2, 4 = 1). Scoring high on a subscale means the stress level is high. The minimal values that indicate sexual minority stress on each subscale are IH ≥ 3, ExR ≥ 3, Clm ≥ 3, SO ≥ 4, SMNE, and any item endorsed. In the present study, the SMSS had Cronbach’s alpha reliabilities ranging from 0.73 to 0.90: IH (α = 0.84), ExR (α = 0.85), Clm (α = 0.83), SO (α = 0.73), SMNE (α = 0.90). The scale has not yet been validated in the Philippines.
Intimate Partner Violence Attitude Scale-Revised (IPVAS-Revised). It is a 17-item self-report instrument that measures one’s attitudes toward intimate partner violence (IPV) (Smith et al., 2005). It has three subscales: abuse (eight items), violence (four items), and control (five items) (Fincham et al., 2008). Items are rated on a 5-point Likert-type scale. Sample items included: for abuse, “As long as my partner doesn’t hurt me, ‘threats’ are excused,“ for violence, “I think it is wrong to ever damage anything that belongs to my partner.“ For control, “I would not like my partner to ask me what I did every minute of the day.“ Its total score ranges from 17 to 85 and is calculated by adding the three subscale scores. Higher scores indicate favorable attitudes toward IPV behaviors, while lower scores indicate unfavorable attitudes. The IPVAS-Revised scale in the present study had a Cronbach’s alpha from 0.63 to 0.76:0.71 (abuse subscale), 0.63 (control subscale), 0.67 (violence subscale), and 0.76 (total attitude towards IPV). The IPVAS-Revised scale has not yet been validated in the Philippines.
Procedure
Ethical Approval
from the College of Science Ethics Review Committee (protocol number: ERC# 21-0702-0035) was secured before data collection. Subsequently, the scales were converted into web-based questionnaires using Google Forms with the authors’ permission. A call for Filipino participants currently residing in the Philippines was posted on social media sites (e.g., Facebook, Twitter, Instagram) and sent to LGBT organizations to recruit potential participants. The online questionnaire was divided into six sections: (1) informed consent, (2) participant’s agreement with regards to voluntary involvement and withdrawal option, (3) demographic profile (age, sexual orientation, assigned sex at birth, current relationship status, and sexual orientation disclosure), (4) test battery composed of the SMSS and IPVAS-Revised, (5) validity check, and (6) debriefing. The online questionnaire took approximately 10–15 min to complete. The order of the two scales (SMSS & IPVAS-Revised) was programmed to be randomized for every participant to control for possible systematic order effects using allocate.monster. The duration of data collection lasted four months following the data gathering period given by the University of Santo Tomas - College of Science. The 240 valid responses were analyzed using the IBM Statistical Package for the Social Sciences (SPSS) Statistics 28.0.1 software. Mplus 7.4 (Muthén & Muthén, 1998–2015) was used for the multivariate multiple regression analysis, which allowed the exploration of the associations between the five dimensions of minority stressors and the three subscales representing the attitudes towards intimate partner violence simultaneously. Socio-demographic characteristics (i.e., age, sexual orientation, outness, and relationship status) were added to the model as control variables. A fully saturated model was estimated with manifest variables; therefore, fit indices were set at 2 = 0; df = 0, Comparative Fit Index (CFI) = 1.00; Tucker-Lewis Index (TLI) = 1.00; Root-Mean-Square Error of Approximation (RMSEA) = 0.00 by default. The robust maximum-likelihood (MLR) estimator was applied, which is robust to non-normal data distribution.
Results
First, group comparisons between gay and lesbian individuals regarding minority stressors and intimate partner violence attitudes were carried out. For this purpose, we performed independent samples t-tests and Mann-Whitney U tests based on the data distribution of the respective variable. Lesbian individuals reported higher expectations of rejection, while gay individuals reported more sexual minority adverse events (see Table 2). The effect sizes were small-to-moderate. Lesbians also expressed higher satisfaction with outness, while gay men showed more favorable attitudes towards abuse. However, the effect sizes were small.
Table 2 Group Comparisons among Gay and Lesbian Individuals Concerning Minority Stressors and Intimate Partner Violence Attitudes
Variables
(M, SD) Total sample (N = 240) Gay
(n = 85) Lesbian
(n = 155) t/U Cohen’s d
Minority Stressors
Internalized Homophobia 17.72 (6.01) 17.99 (5.22) 17.57 (6.41) 0.54 –
Expectations of Rejection 12.45 (4.52) 11.21 (3.82) 13.13 (4.75) -3.41** 0.45
Satisfaction with Outness 12.12 (4.98) 11.18 (4.85) 12.64 (4.99) -2.19* 0.30
Concealment 12.03 (5.09) 12.60 (5.03) 11.72 (5.11) 1.29 –
Sexual Minority Negative Events 16.77 (10.95) 20.40 (10.95) 14.78 (10.47) 3.91*** 0.52
IPV Subscales
Abuse 12.10 (4.22) 12.85 (4.33) 11.68 (4.12) 2.06* 0.28
Control 9.92 (3.66) 9.81 (3.65) 9.98 (3.68) -0.34 –
Violence 4.93 (2.51) 4.73 (2.49) 5.04 (2.53) 6119.50 –
***p < .001; ** p < .01; *p < .05
Mann-Whitney U test was conducted for violence due to the non-normal distribution of the data, while independent samples t-tests were performed for all other variables
Second, the associations between minority stressors and intimate partner violence attitudes were explored. For this purpose, Pearson and Spearman rank-order correlations were performed based on the data distribution of the respective variable. Results in Table 3 showed that among the five minority stressors, internalized homophobia had a significant association with abuse (r = .231, p < .001) and control (r = .145, p = .03). This result implies higher internalized homophobia is associated with more favorable attitudes toward abuse and control dimensions of intimate partner violence. There is also a significant positive relationship between the concealment of minority stressors and abuse (r = .15, p = .02), indicating that higher concealment is associated with more favorable attitudes towards abuse. However, these associations were generally weak.
Table 3 Zero-order Correlations for Sexual Minority Stressors and Intimate Partner Violence Attitude Subscales
Variables 1 2 3 4 5 6 7 8
Minority Stressors
1. Internalized Homophobia 1
2. Expectations of Rejection 0.10 1
3. Satisfaction with Outness 0.23*** 0.24*** 1
4. Concealment 0.41*** 0.11 0.38*** 1
5. Sexual Minority Negative Events 0.001 0.18** 0.18** 0.24*** 1
IPV Subscales
6. Abuse 0.23*** 0.04 0.03 0.15* 0.04 1
7. Control 0.15* − 0.04 0.05 − 0.03 − 0.10 0.36*** 1
8. Violence 0.01 − 0.02 − 0.03 0.001 − 0.10 0.26*** 0.26*** 1
M 17.72 12.45 12.12 12.03 16.77 12.10 9.92 4.93
SD 6.01 4.52 4.98 5.09 10.95 4.22 3.66 2.51
Skewness 0.93 0.20 0.47 0.89 1.01 1.05 0.56 3.54
Kurtosis 0.68 -0.97 -0.50 0.43 0.94 1.31 0.01 14.47
***p < .001; ** p < .01; *p < .05
Spearman correlations were conducted for IPV Violence and its associations due to the high skewness and kurtosis, while Pearson correlations were performed for all other variables
In the final step, a multivariate multiple regression model was performed to determine whether minority stressors can predict intimate partner violence attitudes while controlling for socio-demographic characteristics (i.e., age, sexual orientation, outness, and relationship status). Table 4 showed that only age had a weak association with control and violence among socio-demographic characteristics, indicating that favorable attitudes towards control and violence slightly increased with age. Moreover, higher internalized homophobia was again associated with more favorable attitudes towards abuse and control. However, these variables explained only a small proportion of the total variance of abuse (8%) and control (7%).
Table 4 Multivariate Multiple Regression Model Representing the Associations between Socio-demographic Characteristics, Minority Stressors, and Intimate Partner Violence Attitudes (N = 240)
Predictor variables Outcome variables (IPV subscales) β (SE)
Abuse Control Violence
Socio-demographic Characteristics
Age (years) 0.03 (0.08) 0.13 (0.07)* 0.18 (0.09)*
Sexual Orientation -0.13 (0.07) 0.01 (0.07) 0.10 (0.07)
Outness -0.04 (0.07) 0.08 (0.06) -0.01 (0.11)
Relationship Status 0.02 (0.07) 0.02 (0.07) -0.04 (0.06)
Minority Stressors
Internalized Homophobia 0.21 (0.07)** 0.19 (0.07)** 0.08 (0.08)
Expectations of Rejection 0.06 (0.07) -0.05 (0.07) -0.02 (0.09)
Satisfaction with Outness -0.04 (0.07) 0.12 (0.07) -0.03 (0.09)
Concealment 0.06 (0.07) -0.08 (0.07) 0.12 (0.12)
Sexual Minority Negative Events 0.001 (0.07) -0.07 (0.08) < 0.001 (0.07)
R 2 0.08* 0.07* 0.05
**p < .01 level; *p < .05 level
Sexual orientation (1 = gay, 2 = lesbian), outness (0 = sexual orientation is not out to someone, 1 = sexual orientation is out to someone), and relationship status (1 = single, 2 = married or in a relationship) were dichotomized for the sake of clarity
Discussion
Sexual Minority Stressors and Attitudes Towards Intimate Partner Violence
The current study aimed to investigate whether sexual minority stressors, mainly internalized homophobia, expectations of rejection, satisfaction with outness, concealment, and sexual minority adverse events, has a relationship with attitudes towards intimate partner violence (IPV). We found that internalized homophobia as a sexual minority stressor is significantly associated with intimate partner violence attitudes (IPV), such as abuse and control. Age was weakly associated with control and violence dimensions, indicating that they exhibit slightly more favorable attitudes toward control and violence as people get older. It is contrary to the work of Ali et al. (2016), wherein, regardless of marital status, age, or sexual orientation, IPV cases remained eminent. The research by Volpe et al. (2013) discussed how women with older male partners are more likely to have psychosexual problems. They stated that the low relationship control felt by the younger women evoked IPV issues, such as their partners telling them how to dress or demanding more time together. It could be associated with individuals with more masculine characteristics expressing lower health literacy. In contrast, those with feminine expressions engage in increased transactional sex (e.g., sugar babies), wherein they receive gifts, money, or services from their partners (Ramos et al., 2021).
Internalized Homophobia and Attitudes Towards Intimate Partner Violence
The relationship between internalized homophobia and intimate partner violence attitudes is consistent with Jacobson et al. (2015). They found that LGBTQ + members reported high internalized homophobia with verbal and physical victimization. The findings in this study are also supported by previous research showing associations between IPV attitudes and internalized homophobia (Badenes-Ribera et al., 2017; Finneran & Stephenson, 2014; Lewis et al., 2017; Stephenson & Finneran, 2016). However, the associations were generally weak, similar to Badenes-Ribera et al. (2017), indicating that LGB individuals with negative feelings about their sexual orientation might project violence toward their same-sex partners. The latter may see themselves deserving of such treatment due to their sexual orientation. Those who viewed their sexual orientation negatively were likely to perceive their victimization as deserved act or a consequence of being identified as LGB (Stiles-Shields & Carroll, 2014). Lesbians with a negative view of homosexuality were most likely to stay in abusive relationships (AyhanBalik & Bilgin, 2019). Various research also documented associations between internalized homophobia and negative relationship quality (Cao et al., 2017; Totenhagen et al., 2018). It involves romantic relationship problems in responses (Okutan et al., 2016), lack of commitment, decreased ability to communicate appropriately and decision-making (Stachowski & Stephenson, 2015), and vulnerability to greater severity of relationship conflict (Totenhagen et al., 2018). These are all pathways to elevating the risk of experiencing victimization and perpetration of IPV (Finneran & Stephenson, 2014).
Moreover, the present study shows that IH predicts IPV attitudes, leading to more favorable attitudes. According to Pepper and Sand (2015), women in same-sex relationships reported high condemnation of lesbians, and lesbian relationships predict the perpetration of sexual abuse. The dissonance between their desire to engage in a sexual relationship and their denunciation of lesbians and being lesbian would result in intense feelings of shame and self-loathing. Thus, higher levels of internalized homophobia predict a greater chance of perpetrating physical aggression (Kelley et al., 2014). On the other hand, it is also notable that various factors could mediate attitudes on IPV and IH, such as fusion (Milletich et al., 2014), rumination (Lewis et al., 2014), and relationship quality (Balsam & Szymanski, 2005). These factors could affect the predictability of IH to favorable and unfavorable attitudes toward IPV.
Abuse and Internalized Homophobia
We likewise found an association between the abuse dimension of IPV attitudes and internalized homophobia. Bartholomew and colleagues (2008) found a relationship between internalized homophobia and physical and psychological abuse perpetration. Members of sexual minorities who view their identity negatively may think they deserve to be treated abusively in an intimate relationship (Stiles-Shields & Carroll, 2014). Chong et al. (2013) also reported an association between internalized homophobia and physical and psychological abuse in gay relationships, while IH is associated with sexual and physical abuse in lesbian relationships.
Furthermore, Bartholomew et al. (2008) reported internalized homophobia as a consistent predictor of physical and psychological abuse perpetration.
Control and Internal Homophobia
The control dimension of IPV attitudes was found to be associated with IH. A study by Donovan & Hester (2014) found that coercive control that is eminent among heterosexual relationships is also experienced across same-sex relationships. Stiles-Shields and Carroll (2014) found that control and power in relationships were the strongest predictors of IPV in same-sex relationships; however, only a few studies examined the association of IH with control in same-sex relationships. McKenry et al. (2006) proposed that sexual minority women with negative perceptions and beliefs about homosexuality may have low self-worth and incite physical aggression as an avenue for them to regain control in the relationship. A loss of control and power in intimate relationships could perpetuate physical Violence (Milletich et al., 2014). Furthermore, control can occur in the form of “fusion” or the lack of boundaries between partners (Kimmes et al., 2017), often leading to a loss of sense of self, which is highly associated as a mediator between IPV and IH. Higher levels of IH are associated with higher fusion levels, whereas higher levels are also related to intimate partner violence perpetration (Milletich et al., 2014).
Expectations of Rejection and Attitudes Towards Intimate Partner Violence
Inman and London (2021) stated that a person’s sensitivity to rejection could predict the perpetration of IPV. However, the results showed no relationship between expectations of rejection (ExR) and intimate partner violence attitudes. It contradicts other studies that showed an association between ExR and IPV. According to Carvalho et al. (2011), individuals with an IPV history may have been predisposed to anticipate rejection because of their sexual preference and orientation. Armenti & Babcock (2018) supported the results obtained since their research also reported no association between rejection sensitivity and variables of IPV. It can make them hesitant to interact or socialize with others and isolate themselves, decreasing their chances of becoming victims or perpetrators of abuse. Likewise, Rostosky and Riggle (2017a) concluded that most are careful with whom they share their relationship with the same sex and are probably reluctant to enter into one. Thus, the lack of experience and involvement could result in little to no knowledge about IPV attitudes.
Additionally, gay men’s experiences with prejudice, stigma, and rejection are augmented by the heteronormative and sexist culture that contributes to the belief that men cannot be victims of violence in any way (Finneran & Stephenson, 2013). Due to their fear of stigma and discrimination, sexual minorities are likely socially isolated and reluctant to seek support from the queer community (McConnell et al., 2018). Their experience with rejection may have clouded their judgment when they answered the scales. Therefore, the participant’s answers in the current study may have been affected — underreported or understated.
Satisfaction with Outness and Attitudes Towards Intimate Partner Violence
The present study also shows no relationship between satisfaction with outness (SO) and IPV attitudes, contrary to the results of Kelley et al. (2014) and Hines (2015). Similarly, SO is not associated with the three IPV dimensions. In parallel, some studies demonstrated no link between abuse and outness to family and religion (Longares et al., 2018). The lack of association between SO and IPV attitudes could be explained by the relationship between their level of outness and their partner and relationship satisfaction (Knoble & Linville, 2010). A study on lesbians revealed that those who feel more “out” regarding their sexual orientation showed greater satisfaction with their relationships (Lavner, 2016). Correspondingly, lesbian and gay couples who are open with their sexual preferences reported greater relational satisfaction (Rostosky & Riggle, 2017b). A direct relationship was also found between outness and relational quality in lesbian relationships (LaSala, 2013). Contrarily, only a limited number of studies explored the link between relationship satisfaction and IPV attitudes. One study concluded that relationship satisfaction was indirectly associated with IPV, wherein people, specifically lesbians, with increased relationship satisfaction, are less likely to be perpetrators or victims of IPV (Hines, 2015). Another possible reason for the lack of relationship between satisfaction with outness and IPV is the limited variability of participants’ outness, wherein most have disclosed their sexual orientation to someone.
Concealment and Attitudes Towards Intimate Partner Violence
The present study shows no significant relationship between concealment (Clm) minority stressors with overall attitude towards IPV and only a negligible association with the abuse dimension. Metheny (2019) stated that the concealment of minority identity is an effort to reduce anticipated stigma. It is in line with Edwards and Sylaska’s (2012) research, wherein concealment was not strongly related to IPV perpetration. The authors explained that not all cases of concealment are rooted in shame, and concealment can also be used adaptively to avoid discrimination. The same research found that when compared to IH, negative feelings about one’s sexual orientation were not always evident by concealment, which can be why it was less consistent in IPV. Sexual minority members are always mindful of their actions and could be aggressive or engage in risk-taking when their sexual orientation is discovered (Balsam & Szymanski, 2005; Freire, 2022; Schrimshaw et al., 2013).
Sexual Minority Negative Events and Attitudes Towards Intimate Partner Violence
Contrary to previous research suggesting that sexual minority adverse events (SMNE) are associated with intimate partner violence attitudes (Longobardi & Badenes-Ribera, 2017), this study argues otherwise. Neither actual experiences nor perceiving events such as discrimination, isolation, bullying, aggression, and diagnosis of Human Immunodeficiency Virus (HIV) and other chronic sexually transmitted diseases happening to someone because they are part of the LGBTQ + community are unrelated to attitudes towards same-sex relationship violence. According to Edwards and Sylaska (2012), the lack of association between SMNE and IPVAS is because, despite the high rates of externalized minority stressors faced by LGBTQ + members, their display of resilience takes over, and they do not internalize these negative experiences. These externalized minority stressors (e.g., sexual orientation-related victimization) are not the driving factors in IPV perpetration. The extent to which these individuals internalize these experiences is most influential in the perpetration of IPV. Our findings are further supported by Steele et al., (2017), who concluded that IPV is not a gendered phenomenon. Instead, it involves power and control and is influenced by racism and classism. As gender is irrelevant to the risk of IPV, it could be that breaking heteronormative scripts in same-sex relationships renders traditional, gendered models of IPV less applicable. More recent research suggests differences between behavior and perceptions worthy of future exploration regarding the discrepancy between existing studies that show a relationship between sexual minority discrimination and IPV. It is recommended to study further the connection between external sexual minority stigma and perceptions of psychological IPV (Islam, 2021). Besides, Balsam and Szymanski (2005) argued that same-sex couples are better equipped to cope with experiences of minority stressors outside their dyadic relationship. Their relationship can serve as their haven for such experiences since lesbian and gay couples have better communication and negotiating skills regarding their differences and tend to be more egalitarian than their heteronormative counterparts (Lev, 2015). Thus, they are more satisfied in their relationships and resolve problems more effectively.
Minority Stressors and Violence Dimension
None of the sexual Minority Stressors (internalized homophobia, satisfaction with outness, expectations of rejection, concealment, and sexual minority adverse events) have shown a significant relationship with the violence dimension of IPV attitudes in the current study. This finding is contrary to our hypothesis but consistent with some research indicating no associations. Balsam and Szymanski (2005) concluded no association between IH and physical and sexual violence. IH is a hidden belief that is a personal and sensitive subject to be discussed and dealt with carefully. It is rare for LGB research participants with low levels of IH to partake in an LGB study (Milletich et al., 2014). On the other hand, satisfaction with outness is not associated with attitudes (favorable/unfavorable) toward violence. Aside from contentment with one’s outness degree, it is more important to consider personal situations, the outness degree of one’s partner, and the same-sex support couples can receive from others and provide for each other (Knoble & Linville, 2010).
Depending on the social context, outness can have its advantages and disadvantages. Some lesbians prefer not to disclose their sexual orientation to their family, friends, and colleagues to protect themselves against the perceived dangers of abuse and discrimination (Hines, 2015). Specifically, those raised conservatively and who find their loved one’s valuable support are more likely to remain closeted. Likewise, expectations of rejections have the same findings. Anticipated or actual feelings of unwanted judgment from others due to being part of a sexual minority can hinder an individual from socializing with others or even establishing intimate relationships (Armenti & Babcock, 2018). Therefore, minimizing the chances of becoming a perpetrator or victim of violence. Rather than being friendly, gays who expect rejection tend to use politeness strategies when talking to someone. They are more attentive to others, noting every subtle hint that could mean rejection by the person they are interested in (Ferlotti, 2020). However, the present study only tackles how lesbians and gays perceive rejection but not how they respond psychologically and emotionally. Determining the internal experiences and processing these expectations could provide a more detailed perspective and address which areas are deemed unsafe. In examining these minority stressors, it is crucial to note that the insufficient discourse on violence due to the topic being sensitive may contribute to the lack of significant results. The silence contributes to the lack of public discussion of the phenomenon since recognizing such acts only adds to the stigmatization and discrimination of the community (Rolle et al., 2018). It is also challenging to disclose abuse history as it can be traumatizing. The participants may hesitate to share their experiences openly and honestly. Some participants could have practiced “faking bad” or “faking good,“ wherein they exaggerated or downplayed their answers on scales and tests.
Intimate Partner Violence in the Philippine Context
McDonagh et al.‘s (2021) assessment of IPV perpetrators who completed self-reported tests on personality pathology revealed that some exaggerated their answers to gain a high pathology score while others minimized their answers to achieve a low pathology score. Furthermore, the participants included in the study are all residents of the Philippines, a conservative and religious country wherein the majority has not entirely embraced the LGBTQ + community. Despite claims that the Philippines is “gay friendly,“ the years of debate regarding gender equality and SOGIE in congress state otherwise due to religious and conservative politicians’ rejection of their moral concerns on homosexuality (Manalastas & Torre, 2016). The LGBTQ + community’s efforts and voice advocating equality has been silenced despite using their religious freedom as a reason because conservative Christian organizations have retaliated by using the same argument. (Cornelio &Dagle, 2019). Some traditional Filipino beliefs do not align with the LGBT community’s lifestyle. Likewise, religiosity has been reported to predict attitudes about gays and lesbians in the Philippines (Reyes et al., 2019). Understanding the situation and views on the LGBT + community in the Philippines is essential. Participants can be affected by these societal circumstances and public attitudes toward them, which may be reflected in the study results. Indeed, cultural norms and familial and societal views are closely associated with attitudes toward intimate partner violence (Copp et al., 2019).
Implications for Clinical Practice and Public Policy
We found that internalized homophobia as a minority stressor is positively associated with intimate partner attitudes such as abuse and control. The present findings can contribute to the growing research exploring sexual minority stressors concerning attitudes toward IPV. These findings were demonstrated using a sample of Filipino LGBTQ + individuals, which can highlight the importance of research on IPV in this cultural context. Indeed, research on IPV in the Filipino LGBTQ + community is scarce.
Fear of stigma and discrimination, which could lead to the concealment of sexual identity (Edwards & Sylaska, 2012), social isolation, and reluctance to seek support or help (McConnell et al., 2018), are some of the difficulties faced by the Filipino LGBT + community. Thus, the government should prioritize this problem by passing the Sexual Orientation and Gender Identity Expression Equality (SOGIE) Bill or Anti-Discrimination Bill, which has not proceeded for more than twenty years (Press Release - Hontiveros Renews Call to Pass SOGIE Equality Bill, 2022). Moreover, educating and sensitizing individuals about the forms of IPV and supporting vulnerable individuals can facilitate favorable public attitudes towards this community, fostering their mental health. More attention should be paid to this vulnerable group in psychological healthcare to cope with IPV (e.g., facilitating adaptive coping strategies).
Limitations and Future Directions
We must interpret our findings meticulously, as some limitations should be considered. Our results may not be generalizable to the LGBTQ + community since the age range of participants is restricted to ages 20 to 40 and only covers lesbian and gay individuals. In addition, most of them have disclosed their sexual orientation to someone making concealment a minority stressor challenging to investigate: the variability in their sexual outness is significantly minimized. It is also crucial to note the wide gap in the local review of related literature on IPV in terms of sexual minorities as participants. Most data focus on women and children.
In contrast, most research regarding sexual minority stressors was conducted outside the country. Since the topic is sensitive, the participants may have had response sets geared to more favorable ones in completing the online questionnaire. Moreover, other forms of violence may have been underestimated as the study is centered on the participants’ overall IPV attitudes.
The convenience sampling method is another limitation, as the present results may not be generalizable to Filipino LGBTQ + individuals. Thus, to better understand sexual minority stressors and attitudes toward intimate partner violence and ensure its generalizability among Filipino LGBTQ + members, we suggest that future studies should have a more significant sample size representative of the population of LGBTQ + Filipinos. Another limitation to be considered is the quantitative data analysis which lacks a more detailed inquiry about the participants’ views on IPV. Survey questionnaires have a structured pattern with close-ended questions, which cannot allow participants to explain their choices. Thus, the answers provided are limited categories. Future studies may explore the phenomenon using a qualitative approach to investigate the attitudes toward intimate partner violence of lesbians and gays in more depth. A more comprehensive age range (e.g., youth and above 40 years old) and focusing on other sexual orientations and identities must also be considered. The limited sample size of subgroups concerning gender identity did not allow for an investigation of possible group differences in IPV. Future studies should focus on the perceptions and attitudes towards IPV across different gender identities to gain a more nuanced knowledge of whether individuals with non-binary identities show more permissive attitudes towards IPV. Other mediating factors such as relationship satisfaction, resilience, fusion, and sex should be examined to gain a different perspective on predicting IH and IPV attitudes. While the present study only focuses on IPV attitudes, exploring the different types of violence against sexual minorities, such as physical, sexual, emotional, and psychological, would be beneficial. The supposed presence of internalized homophobia can heighten the risk of Filipino sexual minorities becoming perpetrators or victims of IPV due to their seemingly favorable or accepting attitudes towards IPV. In this case, a qualitative investigation may help shed more understanding on our result. Moreover, other research should focus on therapeutic strategies to prevent and mitigate the influence of internalized homophobia on IPV and provide evidence-based data to strengthen existing IPV-related laws by seeking the inclusion of queer relationships.
General Conclusions
The present findings suggested that gay men had more favorable attitudes towards abuse than lesbians, which can elevate the risk of involvement in IPV. Internalized homophobia was also associated with more permissive attitudes towards abuse and control. These findings highlight the importance of alleviating stress in sexual minority individuals to prevent them from cultivating favorable attitudes towards IPV, which can elevate the risk of experiencing IPV as a victim and/or perpetrator. The present study also points out the importance of investigating IPV in the Filipino LGBTQ + community, which can contribute to developing targeted mental healthcare programs to support sexual minority individuals with IPV experiences and prevent them from severe mental health concerns.
Funding:
Ágnes Zsila was supported by the ÚNKP-22-4 New National Excellence Program of the Ministry for Culture and Innovation from the source of the National Research, Development, and Innovation Fund.
Data Availability
not applicable.
Declarations
Conflicts of interest/Competing Interests:
The authors have no conflicts of interest to declare relevant to the content of this article.
Ethics Approval:
All procedures performed in the present study that involved human participants were per the ethical standards of the Ethics Review Committee (ERC) of the College of Science, the University of Santo Tomas, with ERC# 21-0702-0035.
Consent to Participate:
The current study gave informed consent before voluntary participation. In addition, participants were briefed on the nature of the study and were assured that all data collected would be kept confidential and that participation was purely voluntary without remuneration.
Consent for publication
not applicable.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Ali PA Dhingra K McGarry J A literature review of intimate partner violence and its classifications Aggression and Violent Behavior 2016 31 1–2 16 25 10.1016/j.avb.2016.06.008
Ard KL Makadon HJ Addressing intimate partner violence in lesbian, gay, bisexual, and transgender patients Journal of General Internal Medicine 2011 26 8 930 933 10.1007/s11606-011-1697-6 21448753
Armenti, N. A., & Babcock, J. C. (2018). Borderline personality features, anger, and intimate partner violence: An experimental manipulation of rejection. Journal of Interpersonal Violence, 36(5–6). 10.1177/0886260518771686
AyhanBalik, C. H., & Bilgin, H. (2019). Experiences of minority stress and intimate partner violence among homosexual women in Turkey. Journal of Interpersonal Violence, 1–24. 10.1177/0886260519864371.
Badenes-Ribera L Sánchez-Meca J Longobardi C The relationship between internalized homophobia and intimate partner violence in same-sex relationships: Ameta-analysis Trauma Violence & Abuse 2017 20 3 331 343 10.1177/1524838017708781
Balsam KF Szymanski DM Relationship quality and domestic violence in women’s same-sex relationships: the role of minority stress Psychology of Women Quarterly 2005 29 258 269 10.1111/j.1471-6402.2005.00220.x
Bartholomew K Regan KV Oram D White MA Correlates of Partner abuse in male same-sex Relationships Violence and Victims 2008 23 3 344 360 10.1891/0886-6708.23.3.344 18624099
Boston, N. (2019). Intimate partner violence in the LGBTQ Community.Georgia Southern University. https://digitalcommons.georgiasouthern.edu/cgi/viewcontent.cgi?article=1504&context=honors-theses
Breiding, M., Basile, K., Smith, S., Black, M., & Mahendra, R. (2015). Intimate partner violence surveillance: uniform definitions and recommended data elements (2nd ed.). Centers for Disease Control and Prevention; National Center. https://stacks.cdc.gov/view/cdc/31292 for Injury Prevention.
Calton JM Cattaneo LB Gebhard KT Barriers to help seeking for Lesbian, Gay, Bisexual, Transgender, and Queer Survivors of intimate Partner violence Trauma Violence & Abuse 2015 17 5 585 600 10.1177/1524838015585318
Cao H Zhou N Fine M Liang Y Li J Mills-Koonce WR Sexual minority stress and same-sex Relationship Well-being: a Meta-analysis of Research Prior to the U.S. Nationwide legalization of same-sex marriage Journal of Marriage and Family 2017 79 5 1258 1277 10.1111/jomf.12415 28989184
Carvalho AF Lewis RJ Derlega VJ Winstead BA Viggiano C Internalized sexual minority stressors and same-sex intimate partner violence Journal of Family Violence 2011 26 7 501 509 10.1007/s10896-011-9384-2
Centers for Disease Control and Prevention (2020). Prevent domestic violence. CDC. https://www.cdc.gov/injury/features/intimate-partner-violence/index. html
Chong ES Mak WW Kwong MM Risk and protective factors of same sex intimate partner violence in Hong Kong Journal of Interpersonal Violence 2013 28 7 1476 1497 10.1177/0886260512468229 23295381
Cochat Costa Rodrigues, M. C., Leite, F., & Queirós, M. (2017). Sexual minorities: The terminology. European Psychiatry, 41. https://doi.org/10.1016/j. eurpsy.2017.01.1680
Commission on Human Rights of the Philippines (2019). Inputs on protection against violence and discrimination based on sexual orientation and gender identity with focus on data collection. CHR. https://www.ohchr.org/Documents/Issues/SexualOrientation/Data/PhilippinesCommissionHumanRight.docx
Copp JE Giordano PC Manning WD Longmore MA Neighborhood norms, disadvantage, and intimate partner violence perpetration Sociological Forum 2019 34 3 594 615 10.1111/socf.12516 34658505
Cornelio J Daggle RC Weaponising Religious Freedom: same-sex marriage and gender equality in the Philippines Religion & Human Rights 2019 14 2 65 94 10.1163/18710328-13021146
Dehan N Levi Z Spiritual abuse: an additional dimension of abuse experienced by abused Haredi (ultraorthodox) jewish wives Violence Against Women 2009 15 11 1294 1310 10.1177/1077801209347619 19809096
Donovan, C., & Hester, M. (2014). Domestic violence and sexuality: What’s love got to do with it? (1st ed.) [E-book]. Bristol University Press, Policy Press. 10.2307/j.ctt9qgzhk
Edwards KM Sylaska KM The perpetration of intimate partner violence among LGBTQ college youth: the role of minority stress Journal of Youth and Adolescence 2012 42 11 1721 1731 10.1007/s10964-012-9880-6 23233160
Fehringer JA Hindin MJ "I get angry if he’s always drinking and we have no money”: exploring motivations for male and female perpetrated intimate partner violence in the Philippines Health Care for Women International 2013 35 4 476 491 10.1080/07399332.2013.77000 23659284
Ferlotti, R. (2020). Romantic rejection in gay potential relationships: A gay blog analysis John Carroll University. https://collected.jcu.edu/cgi/viewcontent.cgi?article=1045&context=masterstheses
Fincham FD Cui M Braithwaite S Pasley K Attitudes toward intimate partner violence in dating relationships Psychological Assessment 2008 20 3 260 269 10.1037/1040-3590.20.3.260 18778162
Finneran C Stephenson R Gay and bisexual men’s perception of police helpfulness in response to male-male intimate partner violence Western Journal of Emergency Medicine 2013 14 4 354 362 10.5811/westjem.2013.3.15639 23930149
Finneran C Stephenson R Intimate partner violence, minority stress, and sexual risk-taking among U.S. men who have sex with men Journal of Homosexuality 2014 61 2 288 306 10.1080/00918369.2013.839911 24383859
Freire, F. R. (2022). Outness, internalized homophobia, and intimate partner violence in same-gender couples (Master’s thesis, Instituto Universitário Egas Moniz). http://hdl.handle.net/10400.26/39920
Galang, P. (2021, March 11). Violence against women, a major concern during pandemic -Manila News. Manila News: The Manila News - Intelligencer https://manilanews.ph/violence-against-women-a-major-concern-during-pandemic/
Goldberg NG Meyer IH Sexual orientation disparities in history of intimate partner violence Journal of Interpersonal Violence 2013 28 5 1109 1118 10.1177/0886260512459384 23008053
Goldblum, P., Waelde, L., Skinta, M., & Dilley, J. (2011). The Sexual Minority Stress Scale. Unpublished material, personal information
Harden, J., McAllister, P., Spencer, C. M., & Stith, S. M. (2020). The dark side of the rainbow: Queer women’s experiences of intimate partner violence. Trauma Violence & Abuse, 1–13. 10.1177/1524838.
Hegarty K Sheehan M Schonfeld C A Multidimensional Definition of Partner abuse: development and preliminary validation of the composite abuse scale Journal of Family Violence 1999 14 4 399 415 10.1023/a:1022834215681
Hines, J. (2015). Internalized Heterosexism, Outness, Relationship Satisfaction, and Violence in Lesbian Relationships University of Illinois at Chicago. https://indigo.uic.edu/articles/thesis/Internalized_Heterosexism_Outness_Relationship_Satisfaction_and_Violence_in_Lesbian_Relationships/10861568/1
Huecker, M., King, K., & Jordan, G. (2021). Domestic violence. Stat Pearls Publishing LLC. https://www.ncbi.nlm.nih.gov/books/NBK499891/.
Iniewicz G Sałapa K Wrona M Marek N Minority stress among homosexual and bisexual individuals – from theoretical concepts to research tools: the sexual minority stress scale Archives of Psychiatry and Psychotherapy 2017 19 3 69 80 10.12740/app/75483
Inman EM London B Self-silencing mediates the relationship between rejection sensitivity and intimate partner violence Journal of Interpersonal Violence 2021 10.1177/0886260521997948
Islam S Perceptions of psychological intimate Partner violence: the influence of sexual minority stigma and childhood exposure to domestic violence among bisexual and lesbian women International Journal of Environmental Research and Public Health 2021 18 10 5356 10.3390/ijerph18105356 34069849
Jacobson LE Daire AP Abel EM Lambie G Gender expression differences in same-sex intimate partner violence victimization, perpetration, and attitudes among LGBTQ college students Journal of LGBT Issues in Counseling 2015 9 3 199 216 10.1080/15538605.2015.1068144
Kelley ML Milletich RJ Lewis RJ Winstead BA Barraco CL Padilla MA Lynn C Predictors of perpetration of men’s same-sex partner violence Violence and Victims 2014 29 5 784 796 10.1891/0886-6708.vv-d-13-00096 25905128
Kelley ML Lewis RJ Mason TB Discrepant alcohol use, intimate partner violence, and relationship adjustment among lesbian women and their same-sex intimate partners Journal of Family Violence 2015 30 8 977 986 10.1007/s10896-015-9743-5 26478657
Khidhir, S. (2020, June 6). Philippines’ abused husbands. The ASEAN post. https://theaseanpost.com/article/philippines-abused-husbands
Kimmes, J. G., Mallory, A. B., Spencer, C., Beck, A. R., Cafferky, B., & Stith, S. M. (2017). A Meta-analysis of risk markers for intimate Partner violence in same-sex Relationships. Trauma Violence & Abuse, 152483801770878. 10.1177/1524838017708784.
Knoble NB Linville D Outness and relationship satisfaction in same-gender couples Journal of Marital and Family Therapy 2010 38 2 330 339 10.1111/j.1752-0606.2010.00206.x 22512295
LaSala MC Out of the darkness: three waves of Family Research and the emergence of Family Therapy for Lesbian and Gay People Clinical Social Work Journal 2013 41 3 267 276 10.1007/s10615-012-0434-x
Lavner JA Relationship satisfaction in lesbian couples: review, methodological critique, and research agenda Journal of Lesbian Studies 2016 21 1 7 29 10.1080/10894160.2016.1142348 27557338
Lev, A. I. (2015). Resilience in Lesbian and Gay couples. Couple Resilience, 45–61. 10.1007/978-94-017-9909-6_3.
Lewis RJ Milletich RJ Derlega VJ Padilla MA Sexual minority stressors and psychological aggression in lesbian women’s intimate relationships: the mediating roles of rumination and relationship satisfaction Psychology of Women Quarterly 2014 38 535 550 10.1177/0361684313517866
Lewis RJ Mason TB Winstead BA Kelley ML Empirical investigation of a model of sexual minority specific and general risk factors for intimate partner violence among lesbian women Psychology of Violence 2017 7 1 110 119 10.1037/vio0000036 28239508
Longares L Escartín J Barrientos J Rodríguez-Carballeira L Insecure attachment and perpetration of psychological abuse in same-sex couples: a relationship moderated by Outness Sexuality Research and Social Policy 2018 17 1 1 12 10.1007/s13178-018-0363-8
Longobardi C Badenes-Ribera L Intimate partner violence in same-sex relationships and the role of sexual minority stressors: a systematic review of the past 10 years Journal of Child and Family Studies 2017 26 8 2039 2049 10.1007/s10826-017-0734-4
Manalastas, E. J. D., & Torre, B. A. (2013). Social Psychological Aspects of Advocating LGBT Human Rights in the Philippines. Gender and Justice Action Research Program, Institute of Human Rights, University of the Philippines Law Center. https://books.google.com.ph/books/about/Social_Psychological_Aspects_of_Advocati.html?id=GsUjMvyH54cC&redir_esc=y
Manalastas E Torre B LGBT psychology in the Philippines Psychology of Sexualities Review 2016 7 1 60 72
Manalastas EJ Ojanen TT Torre BA Ratanashevorn R Hong BCC Kumaresan V Veeramuthu V Homonegativity in Southeast Asia: attitudes toward lesbians and gay men in Indonesia, Malaysia, the Philippines, Singapore, Thailand, and Vietnam Asia-Pacific Social Science Review 2017 17 1 25 33
McConnell EA Janulis P Phillips G Truong R Birkett M Multiple minority stress and LGBT community resilience among sexual minority men Psychology of Sexual Orientation and Gender Diversity 2018 5 1 1 12 10.1037/sgd0000265 29546228
McDonagh, T., Travers, Á., Murphy, S., & Elklit, A. (2021). Assessing personality pathology response sets in perpetrators of intimate partner violence. Journal of Interpersonal Violence, 088626052110014. 10.1177/08862605211001482.
McKenry PC Serovich JM Mason TL Mosack K Perpetration of gay and lesbian partner violence: a disempowerment perspective Journal of Family Violence 2006 21 233 243 10.1007/s10896-006-9020-8
Messinger AM Invisible victims: same-sex IPV in the national violence against women survey Journal of Interpersonal Violence 2011 26 11 2228 2243 10.1177/0886260510383023 20829231
Metheny, N. (2019). A Fundamental Cause Approach to Intimate Partner Violence in Marginalized Populations. University of Michigan. https://deepblue.lib.umich.edu/bitstream/handle/2027.42/149983/metheny_1.pdf?sequence=1&isAllowed=y
Meyer IH Resilience in the study of minority stress and health of sexual and gender minorities Psychology of Sexual Orientation and Gender Diversity 2015 2 3 209 213 10.1037/sgd0000132
Milletich RJ Gumienny LA Kelley ML D’Lima GM Predictors of women’s same-sex partner violence perpetration Journal of Family Violence 2014 29 653 664 10.1007/s10896-014-9620-7
Muthén, L., & Muthén, B. (1998–2015). Mplus User’s Guide (8th ed.). Muthén&Muthén. Okutan, N., BuyuksahinSunal, A., & SakalliUgurlu, N. (2016). Comparing Heterosexuals’ and Gay Men/Lesbians’ Responses to Relationship Problems and the Effects of Internalized Homophobia on Gay Men/Lesbians’ Responses to Relationship Problems in Turkey. Journal of Homosexuality, 64(2), 218–238. 10.1080/00918369.2016.1174028
OutRight Action International (2018, June 28). Enhancing domestic violence protections for LGBT in the Philippines and Sri Lanka. OutRight Action International. https://outrightinternational.org/asia/enhancing-Domestic-violence-protections-lgbt-philippines-and-sri-lanka
Pepper BI Sand S Internalized homophobia and intimate Partner Violence in Young Adult Women’s same-sex Relationships Journal of Aggression Maltreatment & Trauma 2015 24 6 656 673 10.1080/10926771.2015.1049764
Psychological Association of the Philippines (2011). Statement of the Psychological Association of the Philippines on Non-Discrimination Based on Sexual Orientation, Gender Identity, and Expression.Philippine Journal of Psychology, 44(2). http://ejournals.ph/form/cite.php?id=3955
Psychological Association of the Philippines (2020). PAP Statement Upholding the Basic Human Rights and Well-being of LGBT Persons Based on the Psychological Science of SOGIE https://pap.ph/position-paper/12
Philippine Statistics Authority (2018). One in four women have ever experienced spousal violence (Preliminary results from the 2017 national demographic and health survey) Philippine Statistics Authority. https://psa.gov.ph/content/one-four-women-have-ever-experienced-spousal-violence-preliminary-results-2017-nationalIndianJPsychiatry_74_17
Postmus JL Hoge GL Breckenridge J Sharp-Jeffs N Chung D Economic abuse as an invisible form of domestic violence: a multi-country review Trauma Violence & Abuse 2018 21 2 261 283 10.1177/1524838018764160
Ramos SR Lardier DT Opara I Turpin RE Boyd DT Gutierrez JI Williams CN Nelson LE Kershaw T Intersectional Effects of sexual orientation concealment, internalized Homophobia, and gender expression on sexual identity and HIV risk among sexual minority men of Color: a path analysis Journal of the Association of Nurses in AIDS Care 2021 32 4 495 511 10.1097/jnc.0000000000000274
Reyes M Ballesteros K Bandol P Jimenez K Malangen S Religiosity, gender role beliefs, and Attitudes toward Lesbians and Gays in the Philippines North American Journal of Psychology 2019 21 3 559 571
Rivara FP Anderson ML Fishman P Reid RJ Bonomi AE Carrell D Thompson RS Age, period, and cohort effects on intimate partner violence Violence and Victims 2009 24 5 627 638 10.1891/0886-6708.24.5.627 19852403
Rolle L Giardina G Caldarera AM Gerino E Brustia P When intimate partner violence meets same-sex couples: a review of same-sex intimate partner violence Frontiers in Psychology 2018 9 1506 10.3389/fpsyg.2018.01506 30186202
Rostosky SS Riggle EDB Same-sex couple relationship strengths: a review and synthesis of the empirical literature (2000–2016) Psychology of Sexual Orientation and Gender Diversity 2017 4 1 1 13 10.1037/sgd0000216
Rostosky SS Riggle EDB Same-sex relationships and minority stress Current Opinion in Psychology 2017 13 29 38 10.1016/j.copsyc.2016.04.011 28813290
Russell B Sturgeon JA Police evaluations of intimate partner violence in heterosexual and same-sex relationships: do experience and training play a role? Journal of Police and Criminal Psychology 2018 34 1 34 44 10.1007/s11896-018-9279-8
Schrimshaw EW Siegel K Downing MJ Parsons JT Disclosure and concealment of sexual orientation and the mental health of non-gay-identified, behaviorally bisexual men Journal of Consulting and Clinical Psychology 2013 81 1 141 153 10.1037/a0031272 23276123
Smith BA Thompson S Tomaka J Buchanan AC Development of the intimate partner violence attitude scales (IPVAS) with a predominantly mexican American College sample Hispanic Journal of Behavioral Sciences 2005 27 4 442 454 10.1177/0739986305281233
Stachowski C Stephenson R Homophobia and Communal coping for HIV risk management among Gay Men in Relationships Archives of Sexual Behavior 2015 44 2 467 474 10.1007/s10508-014-0417-9 25614049
Steele, S. M., Everett, B. G., & Hughes, T. L. (2017). Influence of Perceived femininity, masculinity, Race/Ethnicity, and socioeconomic status on intimate Partner Violence among sexual-minority women. Journal of Interpersonal Violence, 088626051668317. 10.1177/0886260516683176.
Stephenson R Finneran C Minority stress and intimate partner violence among gay and bisexual men in Atlanta American Journal of Men’s Health 2016 11 4 952 961 10.1177/1557988316677506
Stiles-Shields C Carroll RA Same-sex domestic violence: prevalence, unique aspects, and clinical implications Journal of Sex & Marital Therapy 2014 41 6 636 648 10.1080/0092623x.2014.958792 25189218
Swan, L. E., Henry, R. S., Smith, E. R., Arelis, A., Barajas, A. R., B. V., & Perrin, P. B. (2019). Discrimination and intimate partner violence victimization and perpetration among a convenience sample of LGBT individuals in Latin America. Journal of Interpersonal Violence, 1–18. 10.1177/0886260519844774.
Swift, J. (2019, November 11). Bisexual Women and Intimate Partner Violence. Gender Policy Report. https://genderpolicyreport.umn.edu/bisexual-women-and-intimate-partner-violence/
Totenhagen CJ Randall AK Lloyd K Stress and relationship functioning in same-sex couples: the vulnerabilities of internalized Homophobia and Outness Family Relations 2018 67 3 399 413 10.1111/fare.12311
UNDP, U. S. A. I. D. (2014). A participatory review and analysis of the legal and social environment for Lesbian, Gay, Bisexual, and Transgender (LGBT) individuals and civil society Being LGBT in Asia: The Philippines Country Report. https://pdf.usaid.gov/pdf_docs/PBAAA888.pdf
Volpe EM Hardie TL Cerulli C Sommers MS Morrison-Beedy D What’s Age got to Do with it? Partner Age Difference, Power, intimate Partner Violence, and sexual risk in Urban Adolescents Journal of Interpersonal Violence 2013 28 10 2068 2087 10.1177/0886260512471082 23345572
Walters, M., Chen, J., & Breiding, M. (2013). The national intimate partner and sexual violence survey: 2010 findings on victimization by sexual orientation. National Center for Injury Prevention; Control Centers for Disease Control and Prevention. https://www.cdc.gov/violenceprevention/pdf/nisvs_sofindings.pdf.
| 0 | PMC9739342 | NO-CC CODE | 2022-12-14 23:29:54 | no | Sex Cult. 2022 Dec 10;:1-21 | utf-8 | Sex Cult | 2,022 | 10.1007/s12119-022-10046-y | oa_other |
==== Front
J Clin Psychol Med Settings
J Clin Psychol Med Settings
Journal of Clinical Psychology in Medical Settings
1068-9583
1573-3572
Springer US New York
9914
10.1007/s10880-022-09914-4
Article
Moving Beyond Words: Leveraging Financial Resources to Improve Diversity, Equity, and Inclusion in Academic Medical Centers
http://orcid.org/0000-0002-1047-7633
Clark Shawnese Gilpin [email protected]
12
Cohen Alyssa 13
Heard-Garris Nia 123
1 grid.16753.36 0000 0001 2299 3507 Department of Pediatrics, Feinberg School of Medicine, Northwestern University, Chicago, IL USA
2 grid.413808.6 0000 0004 0388 2248 Mary Ann & J. Milburn Smith Child Health Outcomes, Research and Evaluation Center, Stanley Manne Children’s Research Institute, Ann & Robert H. Lurie Children’s Hospital of Chicago, 225 E Chicago Avenue, Box 162, Chicago, IL 60611 USA
3 grid.413808.6 0000 0004 0388 2248 Division of Advanced General Pediatrics and Primary Care, Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL USA
10 12 2022
17
25 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.
Diversity, equity, and inclusion (DEI) efforts at academic medical centers (AMCs) began prior to 2020, but have been accelerated after the death of George Floyd, leading many AMCs to recommit their support for DEI. Institutions crafted statements to decry racism, but we assert that institutions must make a transparent, continuous, and robust financial investment to truly show their commitment to DEI. This financial investment should focus on (1) advocacy efforts for programs that will contribute to DEI in health, (2) pipeline programs to support and guide minoritized students to enter health professions, and (3) the recruitment and retention of minoritized faculty. While financial investments will not eliminate all DEI concerns within AMCs, investing significant financial resources consistently and intentionally will better position AMCs to truly advance diversity, equity, and inclusion within healthcare, the community, and beyond.
Keywords
Academic medical center
Diversity
Equity and inclusion
Pipeline
Financial commitment
Healthcare workforce
==== Body
pmcIntroduction
Diversity, equity, and inclusion (DEI) efforts at academic medical centers (AMCs) started prior to 2020, but significantly increased after the death of George Floyd, which resulted in protests and a call for racial justice in the United States. The Centers for Medicare and Medicaid Services defines diversity as characteristics and experiences that make everyone unique, while equity ensures that everyone has fair access to opportunity and experiences, and inclusion aims to make all feel welcomed, valued, and supported (Centers for Medicare & Medicaid Services, 2022). Many AMCs committed and recommitted their support for DEI with displays of statements on their organizations’ websites and social media accounts. The statements were followed by an increasing number of DEI task forces, committees, and postings for executive and senior-level DEI positions—all charged with the goal of improving DEI within institutions (Sotto-Santiago et al., 2021). As discourse amplified, so did the realization that diversity statements may be laden with hollow messages and influenced by isomorphism, or the pressures placed on an organization to mimic the actions of peers (Carnes et al., 2019; Sotto-Santiago et al., 2021). Despite institutions’ public commitment to DEI and anti-racism, there was evidence that AMCs were not truly invested in this work, such as stories that surfaced about mistreatment of minoritized faculty (Blackstock, 2020; Thompson & Lozano, 2021). For this reason, institutions must not only craft statements that decry racism, but also must move beyond their words and follow through with actions to create more diverse, equitable, inclusive, and anti-racist environments. Here, we assert that, for institutions to truly show their commitment, a transparent, continuous, and robust financial investment must be made. This financial investment should focus on (1) advocacy efforts for programs that will contribute to DEI in health, (2) pipeline programs to support and guide minoritized students to enter health professions, and (3) the recruitment and retention of minoritized faculty. While the landscape of DEI work in AMCs can take many forms beyond what is explored here, financial investment represents an important component of a comprehensive approach.
AMCs as Advocates for Increased Funding
Due to budgetary restrictions and complex financing within AMCs, it is often necessary to obtain dedicated funding for individual programs. AMCs are mostly funded by three entities—faculty practice plans and affiliate hospitals, medical schools, and tax appropriations for public institutions (Reuter, 1997). Medicare and Medicaid finance up to fifty percent of patient care revenue for AMCs, and additional Medicare funds are provided for graduate medical education (GME) training of residents and fellows (Association of American Medical Colleges & Task Force on Graduate Medical Education, 2019; Grover et al., 2014). The financing of AMCs is unique in two ways. First, although AMCs account for only five percent of hospitals in the United States, they provide almost 40% of charity care or free medical treatment for people otherwise unable to pay, and 26% of Medicaid hospitalizations (Grover et al., 2014). Second, although Medicare finances a sizable portion of GME, funding allotments were based on the number of residents and fellows from 1996 and were not increased until 2020 (Association of American Medical Colleges, 2020). AMCs have to supplement the additional funding for GME, charity care, and other activities like research within their budgets, which may make it difficult to allocate additional funds to activities focused on DEI (Grover et al., 2014). Thus, it is crucial for AMCs to leverage their resources to advocate for additional monies from government, private, and philanthropic sources, as well as money from their own budget. AMCs can achieve this by ensuring there are roles dedicated to advocacy within the institution and empowering their healthcare workforce to serve as advocates.
Healthcare advocacy can be defined as an “action by a physician [or other healthcare professional] to promote those social, economic, educational, and political changes that ameliorate the suffering and threats to human health and well-being” (Earnest et al., 2010). Healthcare workers (HCWs) are well positioned to promote changes that would improve DEI in healthcare through advocacy. In particular, given their direct interface with the healthcare system and the community, physicians, clinical psychologists, nurses, and other HCWs hold a vast amount of knowledge and experiences that could impact policies and expand access to funding for DEI activities in AMCs. AMCs can encourage HCW advocacy by recognizing it as scholarly work for faculty promotion and other incentives, which has already been proposed and outlined in the literature (Jindal et al., 2020; Nerlinger et al., 2018). Institutions can also provide protected time for faculty to meet with policy makers, to serve on advisory boards or medical societies, and to provide advocacy education to students or trainees (Dharamsi et al., 2011; Landers & Sehgal, 2000; Louisias et al., 2022). HCWs can make a significant impact via avenues such as becoming involved with professional associations, becoming a school board advisor, or writing opinion pieces for newspapers and online media sites (Earnest et al., 2010; Fernandez Lynch et al., 2020). Given time and work constraints, HCWs may be limited in their ability to consistently advocate; however, there have been major healthcare policy successes in the past from advocacy led by HCWs.
A popular example of advocacy to impact is the work done by Dr. Mona Hanna-Attisha, who in 2015 published an article analyzing blood lead levels in children living in Flint, Michigan before and after the source of drinking water changed in the city (Hanna-Attisha et al., 2016). Her findings suggested that the water source change caused a marked rise in elevated blood lead levels among local children. This prompted her strong advocacy efforts, which included holding a press release, reaching out to news stations, writing opinion pieces, contacting government officials, and testifying before Congress. These efforts were crucial in getting the attention of the United States Environmental Protection Agency, which provided a grant to help improve Flint’s drinking water and helped the state pass legislation enforcing stricter interrogation and mitigation of the drinking water (Whitmer, 2021). Similarly, HCWs at AMCs should be encouraged to advocate for funding that could be used for DEI work.
In addition to encouraging HCWs to be advocates, AMCs as entities, including their administrators, need to be strong advocates for changes that would maximize funding for this work. Although money will not fix every issue, major and rapid changes can occur when tied to policy enforced by funding. This was seen in the 1960s after the passage of the Civil Rights Act of 1964. Title VI of the Act prohibited federal government funds to institutions that discriminated against people on the basis of race, creed, and national origin, which included institutions that were still segregated. Prior to the implementation of the Act, only 83% of hospitals in the North offered integrated patient care, while 6% of hospitals in the South did. After two years of enforcing the Act, 85% of hospitals in the country were compliant with Title VI, meaning they were providing integrated care and purported to not discriminate based on race, creed, and nationality (Reynolds, 1997). These changes occurred so quickly because these institutions could otherwise not receive funds from the federal government, which includes Medicare. Rapid change and advancement of work promoting DEI would manifest if AMCs, who are heavily funded by government sources and foundations, supported and advocated for policy changes that tied the elimination of discriminatory behaviors and DEI metrics and outcomes more closely to funding.
AMC’s Financial Commitment to Increasing the Pipeline
DEI efforts in the workplace started in the 1960s and were prompted by the influx of anti-discrimination legislation that passed during that time. Since then, research has shown that a more diverse workforce enhances a company’s productivity (Saxena, 2014). More recently, reports have shown a positive correlation between increasing gender and racial diversity among executives and a company’s operating profitability (Hunt et al., 2018). Likewise, research in clinical medicine reveals that increased diversity, which more easily allows for patient-physician racial and gender concordance, can increase preventive screening and health care utilization among minorities and women and improve Black infant mortality rates (Alsan et al., 2019; Greenwood et al., 2020; LaVeist et al., 2003). Despite evidence of the benefits of a diverse workforce, diversity among practicing healthcare providers continues to lag, making the need for DEI efforts imperative.
From 1997 to 2017, the percent of medical school matriculants from underrepresented groups (American Indians or Alaska Natives, Blacks, and Hispanics or Latinos) decreased 16 percent compared to the number of first-year medical school slots (Talamantes et al., 2019). This trend remains as trainees move up the academic ranks, with 64.4% of faculty positions held by White males, compared to 2.7% of positions held by Black males and 3.3% held by Hispanic males in 2021 (Association of American Medical Colleges, 2021). Given that many AMCs are in underserved and minoritized areas, they have an opportunity to leverage the communities’ strengths, along with their infrastructure and financial resources to develop and strengthen their programs to improve the diversity in the healthcare workforce (Lale et al., 2010).
Pipeline programs, including pre-professional and enrichment programs, aim to target students prior to medical (or other health professional) school to increase interest in the health care field and provide mentorship and guidance (Smith et al., 2009a). Most pipeline programs are funded by the federal government and foundations, with lower levels of financial contributions originating from institutions (Health Resources & Services Administration, 2018; US Department of Health & Human Services, 2009). Institutions that receive funding from external sources are more likely to have established pipeline programs even after the funding period is over; however, without institutional contribution, the programs are only sustainable for a short time (Association of American Medical Colleges Public-School/Health-Professional School Partnerships National Program Office, 2004; Terrell, 2006). A financial investment into these programs from AMCs could be used to hire dedicated staff to run programs that expose minoritized students to various healthcare fields, and provide stipend support for students to carry out research projects or participate in mentorship programs led by HCWs in different fields (Smith et al., 2009a, 2009b; Taylor, 2018). An ongoing financial investment could support data collection and analysis to assess the impact of these programs. Programs that have a positive impact, such as showing students from disadvantaged backgrounds successfully go on to pursue a clinical career, and even return back to the AMC to work, could increase the likelihood of sustainable funding from internal and external sources. AMC’s financial commitment into pipeline programs would help demonstrate the institution's commitment to DEI, which could be discussed as a community benefit and also may be attractive to current and prospective minoritized faculty and staff.
AMC’s Financial Commitment to Recruiting and Retaining Minoritized Faculty
Supporting minoritized healthcare professionals after they have completed their training should be of the utmost importance to DEI efforts at AMCs, in part, because they are typically led by minoritized faculty. These efforts include leading DEI committees, pipeline programs, and research on health inequities. Institutional support for minoritized faculty has been lacking, as evidenced by decreasing rates of racially minoritized physicians ascending the academic ranks (Association of American Medical Colleges, 2021). There are many elements contributing to a lack of faculty diversity, including lower academic salaries for minoritized HCWs and the minority tax—or increased academic responsibilities placed on minoritized faculty in the name of diversity. This is coupled with inadequate compensation for work that often advances DEI, such as providing mentorship and leading DEI committees (Rodríguez et al., 2015). AMCs need to financially commit to improve faculty salaries, provide compensation for DEI work and research, address medical school debt, and expand paid parental leave.
Unpaid labor and a lack of protected time to carry out many diversity-related tasks are often cited as a huge burden for minoritized faculty. If AMCs want to improve DEI at their institutions, it is important to ensure that their minoritized faculty are not burdened by the minority tax. These responsibilities are often uncompensated and unrecognized as scholarly activities. Although AMCs must do many things to improve the minority tax felt by their faculty, their willingness to increase protected time and compensation is one major component that could alleviate the burden, specifically on racially minoritized faculty, and further support scholarly pursuits (Campbell & Rodríguez, 2019).
Salary surveys have shown that there is a discrepancy between the income of Black and other racially minoritized physicians compared to White physicians, even after controlling for specialty, faculty rank, and number of hours worked (Dander & Lautenberger, 2021). In addition to racial disparities, salary disparities exist among gender, religion, and sexual orientation, leading the American College of Physicians to call for increased compensation equity and salary transparency in academic medicine (American College of Physicians, 2017). Although the Association of American Medical Colleges (AAMC) offers a yearly faculty salary report by gender, race, and geographic location, these data are not institution-specific and other existing resources are costly to access. By providing salary transparency and accessibility, especially for senior-level executives, institutions could demonstrate their commitment to equity. This would ensure their minoritized faculty have competitive salaries, as this is one major reason for faculty attrition and a barrier to recruitment in AMCs (Budhu, 2022; Cropsey et al., 2008). In addition to competitive salaries, loan repayment, adequate paid parental leave, and financial support for research are additional tools to recruit and retain minoritized HCWs.
Loan repayment is a form of compensation that can be offered to support recruiting and retaining minoritized faculty, given their disproportionate loan burden (Dugger et al., 2013). Black, Indigenous, and Latinx people in the United States have significantly less net worth in comparison to their white counterparts, which is reflected in the debt burden of medical school matriculants and graduates (AAMC, 2016; Association of American Medical Colleges, 2019; Bhutta et al., 2020). Analyses of federal and state loan repayment programs demonstrate that they are effective in recruiting physicians and other healthcare professionals (Podolsky & Kini, 2016). Junior faculty have cited financial challenges, including debt management, as one of the obstacles to entering academic medicine (Kubiak et al., 2012). Federal loan repayment programs exist for faculty; however, these funds are limited and an institution providing additional loan repayment could help with debt relief more immediately (Duke, 2002).
Paid parental leave is also an effective recruitment and retention strategy for minoritized early career HCWs, given that disparities exist in paid parental leave among minorities. Data obtained from the San Francisco Bay area from 2016 to 2017 revealed that, compared to white women, Asian, Hispanic, and non-Hispanic Black women received 0.9, 2, and 3.6 less weeks of paid maternity leave (Goodman et al., 2021). Paid parental leave within healthcare, especially among physicians, continues to also be extremely challenging (Freeman et al., 2021). This is especially important because paid parental leave can be protective against peripartum health complications, which are higher in minorities (Montoya-Williams et al., 2020; Van Niel et al., 2020). Therefore, AMCs investing in paid parental leave for minoritized faculty stands to not only close the gap in receipt of this benefit, but also is an important investment in the health and well-being of HCWs that demonstrates AMC’s commitment to supporting diverse staff.
Institutional financial support for research activities is also a major challenge. Minoritized faculty are underrepresented among National Institutes of Health-funded researchers, due to many barriers including a lack of institutional support (Shavers et al., 2005). While this must be addressed, bridge or pilot funding for minoritized faculty is paramount to advancing their faculty careers, as it helps to increase their scholarly productivity and develop research skills. In addition, this will make them more competitive as they apply for external funding (Kubiak et al., 2012; Shavers et al., 2005). Along with providing startup funds, AMCs can also increase protected time for minoritized faculty to allow them time for scholarship, research, DEI work, and other activities that contribute to their productivity and overall mission (Daley et al., 2008). Additionally, AMCs can consider early career endowed chair positions to support minoritized faculty and their research, especially those engaged in DEI work.
Potential Costs of These Recommendations
The recommendations for AMCs and other institutions to financially invest in their DEI work is important, however come at two major costs—the financial cost to the AMC and the potential decrease in availability of HCWs to provide direct patient care.
It is important to address the budgetary constraints many AMCs face, that would limit their ability to contribute financially to DEI programs. The COVID-19 pandemic further strained many institutions’ budgets. At a time when the push for anti-racism in healthcare peaked, so too did AMCs’ financial losses, caused in part by a pause in elective procedures and increased cost of supplies that happened during the COVID-19 pandemic (Carroll & Smith, 2020). AMCs and other hospital systems, therefore have been proceeding with caution when adding line items to their budget, while interrogating ways to ensure financial sustainability and longevity for the company.
While this caution is justified, it is important to consider two things. The financial investment into DEI has the potential to have a positive impact on the profitability of the AMC, based on research from companies in the private sector (Dixon-Fyle et al., 2020; Turner, 2018). Also, reallocating monies from executive salaries, which have grown by up to 93% from 2005 to 2015, compared to HCWs’ wages which increased by 8% in the same time frame, could be a feasible option to consider (Du et al., 2018). This is especially pertinent to AMCs given that CEO compensation is highest at major AMCs compared to other hospital types (Saini et al., 2022; Urbaniec, 2020). Although it will not solve all the financial concerns, these are important aspects to consider regarding the financial consequence of allocating money to DEI.
Greater time away from direct patient care that would result from increasing protected time among HCWs to carry out DEI work could also be seen as a negative consequence. This is especially important because minoritized faculty engage in this work the most and there is already a dearth of minoritized physicians, clinical psychologists, nurses, clinical social workers, and other HCWs available to provide patient care (Association of American Medical Colleges, 2021). To alleviate this issue, it is important to encourage non-minoritized faculty to engage in DEI work as well, so the burden does not fall only on minoritized faculty. This would ensure that minoritized HCWs maintain adequate time to provide direct patient care if they choose. It is also important to acknowledge that increased protected time allows for faculty to do work that previously may have been uncompensated, which could help work satisfaction and potentially decrease attrition. Physicians’ intention to leave patient care has been tied to work dissatisfaction and burnout, all of which has been worsened by COVID-19 (Degen et al., 2015; Garcia et al., 2020). In addition, extant literature has shown high attrition rates among minoritized physicians who have been treated unfairly in academia (Avakame et al., 2021; Blackstock, 2020). Therefore, allowing for more protected time could help career longevity that would provide more direct patient care in the long run.
Conclusion
DEI work in healthcare is a complex and continuous process that AMCs can significantly contribute to given their resources and position at the interface of research, education, and provision of clinical care to the underserved. AMCs can show true commitment to this work by allocating consistent funds to improve some of the critical issues within academia, including the diversity of the healthcare workforce and equitable pay for minoritized faculty and staff. In addition to allocating the institution’s funds to these activities, AMCs must promote continued advocacy of this work at a higher level by encouraging advocacy among HCWs and serving as strong advocates themselves. AMCs can use their role in healthcare to truly advance health equity with tangible investments into DEI. While financial investments will not eliminate all DEI problems within AMCs, they could significantly move this work forward and ensure that DEI work at AMCs is meaningful, sustainable, and actionable.
Author Contribution
Conceptualization: SGC; Writing-original draft preparation: SGC; Writing-review and editing: AC and NHG.
Funding
Not applicable.
Data Availability
Not applicable.
Code Availability
Not applicable.
Declarations
Conflict of interest
Not applicable.
Ethical Approval
Not applicable.
Informed Consent
Not applicable.
Consent for Publication
Not applicable.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
AAMC. (2016). AAMC facts & figures 2016, table 36. https://aamcdiversityfactsandfigures2016.org/report-section/medical-schools/#tablepress-36
Alsan M Garrick O Graziani G Does diversity matter for health? Experimental evidence from Oakland American Economic Review 2019 109 12 4071 4111 10.1257/aer.20181446
American College of Physicians. (2017). Position statement on compensation equity and transparency in the field of medicine. Retrieved from https://www.acponline.org/acp_policy/policies/compensation_equity_and_transparency_position_statement_2017.pdf
Association of American Medical Colleges. (2019). Figure 10. Amount of premedical education debt for U.S. medical school matriculants by race/ethnicity, academic year 2018–2019 (Diversity in Medicine: Facts and Figures 2019, Issue. Retrieved from https://www.aamc.org/data-reports/workforce/interactive-data/figure-10-amount-premedical-education-debt-us-medical-school-matriculants-race/ethnicity-academic
Association of American Medical Colleges. (2020). AAMC statement on inclusion of physician training support in year-end legislation. Retrieved from https://www.aamc.org/news-insights/press-releases/aamc-statement-inclusion-physician-training-support-year-end-legislation
Association of American Medical Colleges. (2021). U.S. medical school faculty trends. Faculty Roster, U.S. Medical School Faculty, Issue. Retrieved from https://www.aamc.org/data-reports/faculty-institutions/interactive-data/us-medical-school-faculty-trends-percentages
Association of American Medical Colleges Public-School/Health-Professional School Partnerships National Program Office. (2004). Learning from others: A literature review and how to guide from the health professions partnership initiative. Association of American Medical Colleges.
Association of American Medical Colleges. Task Force on Graduate Medical Education. (2019). Medicare payments for graduate medical education: what every medical student, resident, and advisor needs to know. Retrieved from https://aamc-black.global.ssl.fastly.net/production/media/filer_public/64/77/6477adae-c4c6-4e0e-8c6f-adcdabf2bdec/dgme_-_medicare_gme_payments_what_you_need_to_know_-_20190430.pdf
Avakame EF October TW Dixon GL Antiracism in academic medicine: Fixing the leak in the pipeline of black physicians ATS Scholar 2021 2 2 193 201 10.34197/ats-scholar.2020-0133PS 34409414
Bhutta, N., Chang, A., Dettling, L., & Hsu, J. (2020). Disparities in wealth by race and ethnicity in the 2019 Survey of Consumer Finances. FEDS Notes, 2020. 10.17016/2380-7172.2797
Blackstock, U. (2020). Why black doctors like me are leaving faculty positions in Academic Medical Centers. STAT. https://www.statnews.com/2020/01/16/black-doctors-leaving-faculty-positions-academic-medical-centers/comment-page-3/
Budhu, J. (2022). If academic medicine is serious about diversity, its institutions need to pay up. STATNews. https://www.statnews.com/2022/02/28/academic-medicine-serious-about-diversity-institutions-pay-up/
Campbell, K. M., & Rodríguez, J. E. (2019). Addressing the minority tax: Perspectives from two diversity leaders on building minority faculty success in academic medicine. Academic Medicine. https://journals.lww.com/academicmedicine/Fulltext/2019/12000/Addressing_the_Minority_Tax__Perspectives_From_Two.13.aspx
Carnes M Fine E Sheridan J Promises and pitfalls of diversity statements: Proceed with caution Academic Medicine 2019 94 1 20 24 10.1097/ACM.0000000000002388 30067539
Carroll, N. W., & Smith, D. G. (2020). Financial implications of the CoviD-19 epidemic for hospitals: A case study. Journal of Health Care Finance
Centers for Medicare and Medicaid Services. (2022). Diversity, equity, and inclusion. Retrieved October 26th from https://www.cms.gov/about-cms/careers-cms/diversity-equity-and-inclusion
Cropsey KL Masho SW Shiang R Sikka V Kornstein SG Hampton CL Why do faculty leave? Reasons for attrition of women and minority faculty from a medical school: Four-year results Journal of Women's Health (2002) 2008 17 7 1111 1118 10.1089/jwh.2007.0582
Daley SP Palermo A-G Nivet M Soto-Greene ML Taylor VS Butts GC Johnson J Strelnick AH Lee-Rey E Williams B Dorscher J Lipscomb WD McDowell S Cornbill R Mindt MR Herbert-Carter J Fry-Johnson YW Smith QT Rust G Kondwani K Diversity in academic medicine no 6 successful programs in minority faculty development: Ingredients of success Mount Sinai Journal of Medicine 2008 75 6 533 551 10.1002/msj.20084 19021192
Dander, V., & Lautenberger, D. (2021). Exploring faculty salary equity at U.S. Medical Schools by gender and race/ethnicity.
Degen C Li J Angerer P Physicians' intention to leave direct patient care: An integrative review Human Resources for Health 2015 13 74 10.1186/s12960-015-0068-5 26350545
Dharamsi S Ho A Spadafora SM Woollard R The physician as health advocate: Translating the quest for social responsibility into medical education and practice Academic Medicine 2011 86 9 1108 1113 10.1097/ACM.0b013e318226b43b 21785306
Dixon-Fyle S Dolan K Hunt V Prince S Diversity wins: How inclusion matters 2020 McKinsey & Company. Inc
Du JY Rascoe AS Marcus RE The growing executive-physician wage gap in major US nonprofit hospitals and burden of nonclinical workers on the US healthcare system Clinical Orthopaedics and Related Research 2018 476 10 1910 1919 10.1097/corr.0000000000000394 30001293
Dugger RA El-Sayed AM Dogra A Messina C Bronson R Galea S The color of debt: Racial disparities in anticipated medical student debt in the United States PLoS ONE 2013 8 9 e74693 e74693 10.1371/journal.pone.0074693 24019975
Duke, B. (2002). Faculty loan repayment program: Making more effective use of program funds. Retrieved from https://oig.hhs.gov/oei/reports/oei-12-01-00510.pdf
Earnest MA Wong SL Federico SG Perspective: Physician advocacy: What is it and how do we do it? Academic Medicine 2010 85 1 63 67 10.1097/ACM.0b013e3181c40d40 20042825
Fernandez Lynch H Bateman-House A Rivera SM Academic advocacy: Opportunities to influence health and science policy under U.S. Lobbying Law Academic Medicine 2020 95 1 44 51 10.1097/acm.0000000000003037 31599758
Freeman G Bharwani A Brown A Ruzycki SM Challenges to navigating pregnancy and parenthood for physician parents: A framework analysis of qualitative data Journal of General Internal Medicine 2021 36 12 3697 3703 10.1007/s11606-021-06835-0 33959880
Garcia LC Shanafelt TD West CP Sinsky CA Trockel MT Nedelec L Maldonado YA Tutty M Dyrbye LN Fassiotto M Burnout, depression, career satisfaction, and work-life integration by physician race/ethnicity JAMA Network Open 2020 3 8 e2012762 10.1001/jamanetworkopen.2020.12762 32766802
Goodman JM Williams C Dow WH Racial/ethnic inequities in paid parental leave access Health Equity 2021 5 1 738 749 10.1089/heq.2021.0001 34909544
Greenwood BN Hardeman RR Huang L Sojourner A Physician-patient racial concordance and disparities in birthing mortality for newborns Proceedings of the National Academy of Sciences 2020 117 35 21194 21200 10.1073/pnas.1913405117
Grover A Slavin PL Willson P The economics of academic medical centers New England Journal of Medicine 2014 370 25 2360 2362 10.1056/NEJMp1403609 24826948
Hanna-Attisha M LaChance J Sadler RC Schnepp AC Elevated blood lead levels in children associated with the flint drinking water crisis: A spatial analysis of risk and public health response American Journal of Public Health 2016 106 2 283 290 10.2105/ajph.2015.303003 26691115
Health Resources and Services Administration. (2018). Health careers pipeline and diversity programs: Academic year 2017–2018. https://bhw.hrsa.gov/sites/default/files/bureau-health-workforce/data-research/health-careers-pipeline-and-diversity-2018.pdf
Hunt V Prince S Dixon-Fyle S Yee L Delivering through diversity 2018 McKinsey & Company
Jindal M Heard-Garris N Empey A Perrin EC Zuckerman KE Johnson TJ Getting "our house" in order: Re-building academic pediatrics by dismantling the anti-black racist foundation Academic Pediatric 2020 20 8 1044 1050 10.1016/j.acap.2020.08.019
Kubiak NT Roman J Guidot DM Franklin Trimm R Kamen DL Recruitment and retention in academic medicine—What junior faculty and trainees want department chairs to know The American Journal of the Medical Sciences 2012 344 1 24 27 10.1016/S0002-9629(15)30914-9 22744375
Lale A Moloney R Alexander GC Academic medical centers and underserved communities: Modern complexities of an enduring relationship Journal of the National Medical Association 2010 102 7 605 613 10.1016/s0027-9684(15)30638-6 20690324
Landers SH Sehgal AR How do physicians lobby their members of Congress? Archives of Internal Medicine 2000 160 21 3248 3251 10.1001/archinte.160.21.3248 11088085
LaVeist TA Nuru-Jeter A Jones KE The association of doctor-patient race concordance with health services utilization Journal of Public Health Policy 2003 24 3–4 312 323 10.2307/3343378 15015865
Louisias M Hicks R Jacobs S Foggs MB The role of physician advocacy in achieving health equity: Where is the allergist-immunologist? The Journal of Allergy and Clinical Immunology 2022 10 4 910 917 10.1016/j.jaip.2022.01.033 35131512
Montoya-Williams D Passarella M Lorch SA The impact of paid family leave in the United States on birth outcomes and mortality in the first year of life Health Services Research 2020 55 Suppl 2 807 814 10.1111/1475-6773.13288 32249413
Nerlinger AL Shah AN Beck AF Beers LS Wong SL Chamberlain LJ Keller D The advocacy portfolio: A standardized tool for documenting physician advocacy Academic Medicine 2018 93 6 860 868 10.1097/acm.0000000000002122 29298182
Podolsky, A., & Kini, T. (2016). How effective are loan forgiveness and service scholarships for recruiting teachers? Policy brief. Learning Policy Institute.
Reuter JA The financing of academic health centers: A chart book 1997 Commonwealth Fund
Reynolds P The Federal Government’s use of title VI and medicare to racially integrate hospitals in the United States, 1963 through 1967 American Journal of Public Health 1997 87 11 1850 1858 10.2105/AJPH.87.11.1850 9366643
Rodríguez JE Campbell KM Pololi LH Addressing disparities in academic medicine: What of the minority tax? BMC Medical Education 2015 15 1 6 10.1186/s12909-015-0290-9 25638211
Saini V Garber J Brownlee S Nonprofit hospital CEO compensation: How much is enough? Health Affairs Forefront 2022 10.1377/forefront.20220208.925255
Saxena A Workforce diversity: A key to improve productivity Procedia Economics and Finance 2014 11 76 85 10.1016/S2212-5671(14)00178-6
Shavers VL Fagan P Lawrence D McCaskill-Stevens W McDonald P Browne D McLinden D Christian M Trimble E Barriers to racial/ethnic minority application and competition for NIH research funding Journal of the National Medical Association 2005 97 8 1063 1077 16173321
Smith SG Nsiah-Kumi PA Jones PR Pamies RJ Pipeline programs in the health professions, part 1: preserving diversity and reducing health disparities Journal of the National Medical Association 2009 101 9 836 851 10.1016/s0027-9684(15)31030-0 19806840
Smith SG Nsiah-Kumi PA Jones PR Pamies RJ Pipeline programs in the health professions, part 2: The impact of recent legal challenges to affirmative action Journal of the National Medical Association 2009 101 9 852 863 10.1016/s0027-9684(15)31031-2 19806841
Sotto-Santiago S Sharp S Mac J Messmore N Haywood A Tyson M Yi V Reclaiming the mission of academic medicine: An examination of institutional responses to (anti)racism AEM Education and Training 2021 5 S1 S33 S43 10.1002/aet2.10668 34616971
Talamantes E Henderson MC Fancher TL Mullan F Closing the gap: Making medical school admissions more equitable New England Journal of Medicine 2019 380 9 803 805 10.1056/NEJMp1808582 30811906
Taylor, S. (2018). The Health Care Career Pipeline: A program director’s reflection on extending the resources of the university to the minority student community. NAM Perspectives. 10.31478/201803b
Terrell C Foreword: The health professions partnership initiative and working toward diversity in the health care workforce Academic Medicine 2006 81 6 S2 S4 10.1097/01.Acm.0000225248.61509.47 16723829
Thompson, P., & Lozano, A. (2021). ‘No glass ceiling’: Tulane doctor files discrimination lawsuit against medical school. NBC News.
Turner, A. (2018). The business case for racial equity. W. K. K. Foundation. https://wkkf.issuelab.org/resource/business-case-for-racial-equity.html
Urbaniec, J. (2020, October 26th). Top 10 highest paid CEOs at nonprofits 2021. C-Suite and Executives. https://www.erieri.com/blog/post/top-10-highest-paid-ceos-at-nonprofits-2021
US Department of Health and Human Services Pipeline programs to improve racial and ethnic diversity in the health professions: An inventory of federal programs, assessment of evaluation approaches, and critical review of the research literature 2009 US Department of Health and Human Services
Van Niel MS Bhatia R Riano NS de Faria L Catapano-Friedman L Ravven S Weissman B Nzodom C Alexander A Budde K Mangurian C The impact of paid maternity leave on the mental and physical health of mothers and children: A review of the literature and policy implications Harvard Review of Psychiatry 2020 28 2 113 126 10.1097/hrp.0000000000000246 32134836
Whitmer, G. (2021). Executive directive 2021 - 9. Retrieved from https://www.michigan.gov/whitmer/news/state-orders-and-directives/2021/11/04/executive-directive-2021-9
| 36495346 | PMC9739343 | NO-CC CODE | 2022-12-14 23:29:54 | no | J Clin Psychol Med Settings. 2022 Dec 10;:1-7 | utf-8 | J Clin Psychol Med Settings | 2,022 | 10.1007/s10880-022-09914-4 | oa_other |
==== Front
Multimed Tools Appl
Multimed Tools Appl
Multimedia Tools and Applications
1380-7501
1573-7721
Springer US New York
14276
10.1007/s11042-022-14276-y
Article
Texture classification for visual data using transfer learning
http://orcid.org/0000-0002-3859-9034
Goyal Vinat [email protected]
Sharma Sanjeev [email protected]
grid.503420.0 0000 0004 7471 7960 Indian Institute of Information Technology, Pune, India
10 12 2022
124
10 3 2022
1 6 2022
19 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 texture is the most fundamental aspect of a picture that contributes to its recognition. Computer vision challenges such as picture identification and segmentation are built on the foundation of texture analysis. Various images of satellite, forestry, medical etc. have been identifiable because of textures. This work aims to offer texture classification models that will outperform previously presented methods. In this work, transfer learning was applied to attain this goal. MobileNetV3 and InceptionV3 are the two pre-trained models employed. Brodatz, Kylberg, and Outex texture datasets were used to evaluate the models. The models achieved excellent results and achieved the objective in most cases. Classification accuracy obtained for the Kylberg dataset were 100% and 99.89%. For the Brodatz dataset, the classification accuracy obtained was 99.83% and 99.94%. For the Outex datasets, the classification accuracy obtained was 99.48% and 99.48%. The model outputs the corresponding label of the texture of the image.
Keywords
Texture classification
Computer vision
Transfer learning
MobileNetV3
InceptionV3
Deep learning
==== Body
pmcIntroduction
The texture is the fundamental quantity of an image that aids in its identification. Texture analysis forms the foundation for computer vision problems like image recognition, image retrieval [37] and segmentation. Various images of satellite [30], forestry [27], medical [10], etc have been identifiable because of textures in them. The texture of an object provides important insights into the properties and behaviour of these objects. These insights later help in the computer vision tasks related to such objects when their shape doesn’t help. Texture today is one of the key components in the analysis of images. This makes the task of texture classification important. For the past years, there has been a lot of effort to develop models that can identify and classify these textures efficiently.
Classic machine learning approaches used for this task include using hand-engined features to extract information and using a statistical algorithm like SVM in the final layer for classification [43]. These approaches were previously preferred, but in recent times these approaches have been outperformed by deep learning methods, particularly convolutional neural networks. After the win of AlexNet [18] in the 2012 ImageNet large-scale visual recognition challenge, there has been an exponential growth in the usage of convolutional neural networks for image classification tasks. Today, significant models in computer vision for tasks like image classification, segmentation, recognition, etc., use convolutional neural networks.
CNN’s learn feature vectors with weight sharing and local connectivity, which detects patterns at all locations in the image. Initial layers of a CNN learn simple features like the edges, and the deeper ones learn more complex features. CNN’s can learn texture patterns of various complexity and scales. Novel convolution neural network models have better performance than the classic machine learning algorithms. This paper aims to propose models that would perform better than the previously proposed models and improvise the texture classification approach.
This paper proposes a transfer learning approach for the texture classification problem. Transfer learning is an approach wherein the intuition uses the knowledge gained while learning to classify classes of one dataset to a different data set of related problems. Transfer learning aims to focus on leveraging labelled data from one feature space to enhance the classification of other entirely different learning spaces. This approach works well when the source dataset (on which the model is trained) and the target dataset (the one in the study) are of a similar domain, making their feature spaces similar. In transfer learning, the top layer of the pre-trained model is replaced by a new layer with the number of neurons equal to the number of classes of the target dataset.
There are two types of transfer learning approaches. The first is feature extraction, wherein only the top layer is trained on the target dataset, freezing the rest of the dataset. The frozen layers are used as feature extractors on the target dataset, training only the top layer. The idea is that a feature vector trained on one kind of data set can extract valuable features on another data set. The second type is the fine-tuning of the model wherein only a few or none of the layers are frozen, and the rest of the layers along with the top layer are trained on the target dataset.
Transfer learning helps leverage the knowledge learnt by a model on one data set to extract information on another data set. Transfer learning also reduces the time of learning all the weights of the convolution layer. Using knowledge of a pre-trained model might also help in complete learning of the problem task compared to building a model from scratch. The pre-trained models used in this paper are MobileNetV3 and InceptionV3. The presented work focusses on: Study about transfer learning on texture datasets.
Achieving better results on the provided benchmark datasets than previous work on the same datasets.
The rest of the paper is organised as follows. Section 2 discuss the literature survey of the related work. Section 4 cover the study of material and methods. Section 4 presents the experiments and results. At last, we are concluding work in Section 5.
Literature review
There has been a lot of research dedicated to texture analysis owing to the importance it holds in the field of computer vision. In 1993, [29] used two powerful algorithms, Principal Component Analysis and Multiscale Autoregressive models, on the Brodatz dataset. The variety of homogenous and non-homogenous images studied in this paper was more significant than those in the previous work. This approach got better results than the models proposed before it. In 1994 an energy-based approach was proposed in [38]. This model got an accuracy of over 90 for the classification of images.
Statistical methods are considered one of the earliest methods for texture analysis of the image, which have given good results on standard texture datasets. Ramola et al. [31] discusses the different statistical approaches like grey level concurrence matrix (GLCM), Local binary pattern(LBP), auto-correction function(ACF) and histogram pattern. Their research and discussion concluded that GCLM is the best approach for texture analysis. The major drawback of the GLCM model is the high matrix dimensionality and high correlation between harlick features. Feng et al. [9] and [5] have also implemented such statistical models on standard data sets and got good results.
Xu et al. [42] proposed a novel robust texture descriptor on variance in rotation, scale and illumination, which combines the dominant orientation analysis and multifractal analysis based on the Gabor filter. This approach was then implemented on the Brodatz and Outex datasets.
Sana and Islam [32] proposed power-law transform (PLT) to extract new spectral texture features. This technique outperformed the widely used Gabor features. As seen, machine learning approaches have had excellent results on standard datasets for texture analysis. However, these algorithms require handmade features for feature extraction. Also, such models cannot be used for feature extraction of images of another dataset, as seen in deep learning architectures with the help of transfer learning.
Zheng et al. [44] proposed an eight feature learning model alongside a deep learning perceptron based architecture. This paper showed the deep learning model’s advantage over the other model. In recent years, convolutional neural networks have surpassed the standard artificial neural networks in the field of computer vision. CNNs have also revolutionised other fields like natural language processing, image and video recognition, information retrieval, grayscale colourisation, and multi-dimensional data processing and have surpassed many machine learning algorithms. Y LeCun proposed CNNs, Boser [21] (1989), three decades ago but did not get popular then because of lack of data and computational power. Today there is abundant data available, the computational power of computers has drastically increased, and there has been a lot of development in developing better optimisation algorithms. Algorithms like stochastic gradient descent with momentum (SGDM) and RMSprop have emerged as the favourites for optimisation. All these factors have contributed to the success of CNNs today[23]. Many CNN architectures such as AlexNet [17], VGG [36], ResNet [11], MobileNet [33], etc have emerged and are being used widely.
Simon and Vijayasundaram [35] Proposed a standard convolution neural network for the task of classification of images of flower and KTH data sets. This paper achieved excellent results as compared to its predecessors. A modified version of CNN is proposed in [2] called T-CNN, which is built on the intuition that the overall shape information extracted by the fully connected layers of a classic CNN is of minor importance in texture analysis. Therefore, an energy measure from the last convolution layer is pooled, connected to a fully connected layer. This idea was inspired by the classic neural networks and filter bank approach. Jain et al. [13] proposed an Optimal Probability-Based Deep Neural Network (OP-DNN) for multi-type skin disease prediction and achieved an accuracy of 95%.
Dixit et al. [6] proposes another approach to classification where whale optimisation algorithm (WOA) is used along with the CNN. Results of this model on the Kylberg, Brodatz and Outex datasets are compared to the results obtained by other models on the same data set. This model gained excellent results and beat other models in comparison. Another such work was [14] where the authors used a new optimisation module Knowledge-Based-Search (KBS), along with Moth–Flame Optimization (MFO). Their work performed well in a dynamic environment.
As discussed, Deep Neural Networks require a large amount of data. When trained on a small dataset, their generalisation performance is limited. Liu et al. [22] proposes the use of relative position network (RPN) and relative mapping network (RMN) for skin lesion image classification with a small dataset. They were able to achieve an accuracy of 85%. Deep learning architectures have the advantage that one model trained on a vast dataset can extract features of images from another dataset. This approach is called transfer learning.
Kazi and Panda [16] uses the transfer learning technique to determine three different types of fruits and their relative freshness and got great results. Kundo et al. [19] proposes a bagging ensemble of three transfer learning models, InceptionV3, ResNet34 and DenseNet201, that outperformed the state of the art methods by 1.56%. Nadeem et al. [26] uses transfer learning for Pakistani traffic-sign recognition. They use a model trained on the German traffic-sign recognition, and with additional pre-processing and regularisation, they achieved competitive results on a small available dataset. This paper uses the approach of transfer learning. Transfer learning has also been widely employed in the medical domain. Arora et al. [3] used a transfer learning-based approach for detecting COVID-19 ailment in lung CT scan. They achieved a precision of 100% using the MobileNet architecture on the SARS-COV-2 CT-Scan dataset.
In recent years transformer-based architectures have revolutionised every domain of deep learning. A transformer-based architecture was originally proposed in [41] where authors proposed an attention mechanism based architecture, dispensing with recurrence and convolutions entirely. Their model was experimented on two translation tasks and outperformed the other models in terms of results and training time. Dosovitskiy et al. [7] proposed Vision Transformers(ViT) inspired from the transformer architectures for Natural Language Processing (NLP) tasks. Their study showed that ViT outperformed the conventional convolutional networks in terms of results and training time on standard datasets like the ImageNet.
The following sections of this paper discuss the materials and methods used and the experiments and results obtained. The last section summarises the paper and talks about the future scope.
Materials and methods
Figure 1 depicts the flowchart followed. The first step was to find the problem statement. The following step was to collect the related dataset to the problem statement. After the data was collected, it was preprocessed to make it of desirable format and size. The pre-processing stage also included data augmentation, which was done to avoid over-fitting the model. After pre-processing, models were designed for the problem statement, then tested on the pre-processed dataset. Transfer learning models are used to classify the different datasets collected. We use the MobileNetV3 and the InceptionV3 models for the classification task. Fig. 1 Flow graph
Dataset
We have used three standard benchmark datasets of the texture classification problem. These are the Brodatz dataset, Kylberg dataset and the Outex dataset. Below is the summary of these datasets.
Brodatz dataset
Brodatz dataset [4] is a very popular dataset for texture classification problems. The dataset has been referred from the University of Southern California[]. The original dataset did not contain the rotated images. In this paper, we have proposed these rotations using 40 different rotation angles on these images. This dataset has 112 classes. The samples of this dataset are displayed in Fig. 2. The summary of this dataset is given in Table 1. Fig. 2 Samples of the Brodatz dataset
Table 1 Summary of the Brodatz dataset
Features Value
Number of Classes 112
Number of samples/ class 40
Total number of samples 4480
Texture patch size 640*640pixels
Format of image 8 bit grey scale PNG
Total size of dataset 1.02 GB
Kylberg dataset
The Kylberg dataset is another widely used dataset for texture classification problems. This dataset has 2 versions (1) with rotation patches and (2) without rotation patches [20]. We have used v1.0, which is the version without rotation patches. The classes of this dataset are blanket1, blanket2, canvas1, ceiling1, ceiling2, cushion1, floor1, floor2, grass1, lentils1, linseds1, oatmeal1, pearlsugar1, rice1, rice2, rug1, sand1, scarf1, scarf2, screen1,seat1, seat2, sesameseeds1, stone1, stone2,stone3, stoneslab1 and wall1. The samples of this dataset are displayed in Fig. 3. The summary of this dataset is given in Table 2. Fig. 3 Samples of the Kylberg dataset
Table 2 Summary of the Kylberg dataset
Features Value
Number of Classes 28
Number of samples/ class 160
Total number of samples 4480
Texture patch size 576*576
Format of image 8 bit grey scale PNG
Total size of dataset 1.76 GB
Outex dataset
The Outex [28] database has a lot of datasets. We are using the Outex_TC_00012 dataset of this database. We have referred to this dataset from the University of OULU. The classes of this dataset are canvas001, canvas002, canvas003, canvas005, canvas006, canvas009, canvas011, canvas021, canvas022, canvas023,canvas025,canvas026 ,canvas031 ,canvas032, canvas033, canvas035, canvas038,canvas039,tile005 ,tile006 ,carpet002 ,carpet004 ,carpet005 and carpet009. The samples of this dataset are displayed in Fig. 4. The summary of this dataset is given in Table 3. Fig. 4 Samples of the Outex dataset
Table 3 Summary of the Outex dataset
Features Value
Number of Classes 24
Number of samples/ class 40
Total number of samples 960
Texture patch size 128*128
Format of image 8 bit grey scale GIF
Total size of dataset 0.16 GB
Data preprocessing and splitting
Data preprocessing is one of the most critical steps. This step makes the raw data compatible with the deep learning model. Images in the Outex and Brodatz datasets are in GIF format and converted to compatible models.
After the data is converted to a compatible format, the images are then resized to a size of 224*224*3, making it compatible with the pre-trained model. After preprocessing, the data is split. The Kylberg, Outex_TC_00012 and the brodatz dataset are split in a ratio of 80:20 into training and testing data.
Data augmentation
As discussed earlier, data augmentation refers to the act of creating more data out of the already existing data. The intuition is that the image of the surface of texture rotated by an angle or flipped along an axis remains the image of that surface. Since there is only 1 image available for each class in the Brodatz dataset, more images are produced by rotating the original images by different angles. Data augmentation is also done for all the data of all three datasets. Figure 5 shows a sample of data received after subjecting the data of the Kylberg dataset to data re-scaling and data augmentation. Figure 6 shows a sample of data received after subjecting the data of the Brodatz dataset to data re-scaling and data augmentation. Figure 7 shows a sample of data received after subjecting the data of the Outex dataset to data re-scaling and data augmentation. Fig. 5 Sample images of Kylberg data set after pre-processing and data augmenation
Fig. 6 Sample images of Brodatz data set after pre-processing and data augmenation
Fig. 7 Sample images of Outex_TC_00012 data set after pre-processing and data augmenation
Proposed model
This paper uses the methodology of using pre-trained models called transfer learning. Intuition uses the knowledge gained by a model on one problem to solve another similar problem. This methodology reduces the time spent on training a model from scratch. Also, using a pre-trained model might be able to learn the problem entirely compared to a model trained from scratch. This paper uses the MobileNetV3 and InceptionV3 models. For each approach, the last dense layer (classification layer) of the pre-trained model is replaced with a softmax layer suitable for classifying the texture of classes of that dataset. In this work, the following transfer learning techniques were implemented (Fig. 8): Feature extraction: Here, we froze all the model layers and trained only the added dense layer. Here the pre-trained model is only used as a feature extractor for the classifier.
Full fine-tuning: Here, the whole pre-trained model was fine-tuned using the data in use.
Fig. 8 Proposed model
Transfer learning
It becomes difficult to collect enough data to build a model from scratch in many world applications. In such scenarios, the idea of transfer learning comes in. As discussed earlier, transfer learning is an approach wherein a model trained on a vast data set is used to solve a related problem. In the medical domain, the number of samples is limited because the procedure of collecting the data is both expensive and complicated. In such situations, using a pre-trained model is more effective than training a model from scratch. One such example is breast cancer classification [34] where the goal is to classify whether a cancer is malignant or benign. The paper compared the results from a pre-trained model and a model trained from scratch. Results obtained by transfer learning surpassed those obtained by a model trained from scratch. In this paper, we have used TensorFlow Hub to import such pre-trained models without their top layers. A softmax layer is then added to these layers. For the Kylberg dataset, only the last layer is trained. For the Outex and brodatz datasets, the models were fine-tuned.
MobileNetV3
MobileNet was proposed by Sandler, Howard [33]. This model has achieved a great balance between performance and computation cost. MobileNet offers an extremely efficient network architecture that can easily match the requirements for mobile and embedded applications. This paper makes use of the MobileNetV3 small model, which was proposed in [12]. TensorFlow Hub is used to use the MobileNetV3 model, which has been trained on ImageNet (ILSVRC-2012-CLS) data. The model is used as feature extraction for the Kylberg dataset without tuning. The model is fully fine-tuned for Outex and the brodatz datasetsuned. Figure 9 summarises the MobileNetV3 architecture. Fig. 9 Summary of the MobileNetV3 model
InceptionV3
InceptionV3 [40] is the third edition of Google’s Inception Convolutional Neural Network. The Inception modules are well-designed convolution modules that can generate discriminatory features and reduce the number of parameters. The InceptionV1 model was introduced at the 2014 ILSVRC classification challenge, where VGGNet [36] was also presented for the first time. Both gained similar results. However, Inception architecture had the advantage of performing well even under strict constraints on memory and computational budget.
The Inceptionv1 [39] model overcame the problem of variation of information by having different sizes of filters and a wider network. It is 22 layers deep (27, including the pooling layers). It uses global average pooling at the end of the last inception module. It is a deep network and is subject to the vanishing gradient problem. To prevent the middle part of the network from “dying out,” it uses two auxiliary classifiers.
Neural networks perform better when convolutions don’t alter the dimensions of the input drastically. Reducing the dimensions too much may cause loss of information, known as a “representational bottleneck.” InceptionV2 [40] model overcame this problem by expanding the filterbanks. InceptionV2 also used clever factorization methods to make the convolution more efficient in terms of computation complexity.
The InceptionV3 had all the upgrades that InceptionV2 had. In addition, it used RMSProp Optimizer, BatchNorm in the Auxillary Classifiers, and Label Smoothing to prevent overfitting. Figure 10 summarises the MobileNetV3 architecture. Fig. 10 Summary of the InceptionV3 model
Experiments and results
Hardware and software setup
Tesla K80 GPU and 13 GB RAM used for training along with TensorFlow, Keras, and Scikit-learn libraries in Google Colab, coded in Python 3.7.10.
Training and testing data
The Kylberg, Brodatz and the Outex datasets are split into training data (80%) and testing data (20%). Adam optimisation and categorical cross-entropy loss functions are used in all cases. A learning rate of 0.01 has been used. The batch size for the training was set to 32. The proposed model 1 for the Kylberg dataset is only fully trained on the training data. Rest in all other cases, the pre-trained model is used as a feature vector, and only the top added layer is trained on the training data.
Evaluation criteria
In the prediction phase, seven quantitative performance measures were computed to access the reliability of trained models using the validation data, including precision, recall, f1-score, accuracy, macro-avg, weighted-avg and Cohen kappa score. These metrics are computed based on True Positive (TP), True Negative (TN), False Positive (FP), False Negative (FN).
1 Precision=TPTP+FP
2 Recall=TPTP+FN
3 F1Score=2∗Precision∗RecallPrecision+Recall
4 Accuracy=TP+TNTP+FN+TN+FP
5 Weightedavg=F1class1∗W1+F1class2∗W2+F1class3∗W3+⋯+F1classn∗Wn
F1classm : F1 score of class m 6 Macroavg=F1class1+F1class2+F1class3+⋯+F1classn
F1classm : F1 score of class m Cohen kappa score: 7 K=p0−pe1−pe
p0 = relative observed agreement among raters, pe = the hypothetical probability of chance agreement.
Training single convolution mode
All the images in the .gif or the .ras format were converted to a compatible format. After that, All the images of the three datasets in the study were rescaled to a size of 224*224. The images were then normalised to make the values of their pixels range from 0-1. The Kyllberg and Brodatz datasets were then subjected to data augmentation before passing them to the proposed model.
Kylberg dataset
The first dataset to be studied was the Kylberg dataset. The first model is developed using the MobileNetV3 small model, trained on the ImageNet dataset. The top layer of the pre-trained model is removed and replaced by a softmax layer with 28 classes. The model was fully fined tuned, i.e. all the model layers were trained on the training dataset. The proposed model was trained for 10 epochs on the training dataset. The model achieved an accuracy of 100% on the testing dataset. The classification report and confusion matrix of model 1 on testing it on testing data are shown in Table 4 and Fig. 11 respectively. The accuracy vs epochs graph and the loss vs epochs graph of model1 for the Kylberg dataset while training is shown in Fig. 12. Table 4 Classification report for model 1 Kylberg dataset
precision recall f1-score support
Accuracy − − 1.00 896
Macro Avg 1.00 1.00 1.00 896
Weighted Avg 1.00 1.00 1.00 896
Fig. 11 Model 1 confusion matrix for the Kylberg datset
Fig. 12 Model 1 accuracy and losses graph for the Kylberg datset
The second model is developed using the InceptionV3 model trained on the ImageNet dataset. The top layer of the pre-trained model is removed and replaced by a softmax layer with 28 classes. The pre-trained model was used as a feature extractor, i.e. all the layers of the pre-trained model were frozen, and only the top layer was trained on the training dataset. The proposed model was trained for 10 epochs on the training dataset. The model achieved an accuracy of 99.8883% on the testing dataset. The classification report and confusion matrix of model 2 on testing it on testing data is shown in Table 5 and Fig. 13 respectively. The accuracy vs epochs graph and the loss vs epochs graph of model1 for the Kylberg dataset while training is shown in Fig. 14. Table 5 Classification report for model 2 Kylberg dataset
precision recall f1-score support
Accuracy − − 1.00 896
Macro Avg 1.00 1.00 1.00 896
Weighted Avg 1.00 1.00 1.00 896
Fig. 13 Model 2 confusion matrix for the Kylberg dataset
Fig. 14 Model 2 accuracy and losses graph for the Kylberg datset
Brodatz dataset
The second dataset to be studied was the Brodatz dataset. The first model is developed using the MobileNetV3 small model, trained on the ImageNet dataset. The top layer of the pre-trained model is removed and replaced by a softmax layer with 112 classes. The pre-trained model was used as a feature extractor, i.e. all the layers of the pre-trained model were frozen, and only the top layer was trained on the training dataset. The proposed model was trained for 7 epochs on the training dataset. The model achieved an accuracy of 99.6651% on the testing dataset. The classification report of model 1 on testing it on the testing data is shown in Table 6. The accuracy vs epochs graph and the loss vs epochs graph of model1 for the Brodtz dataset while training is shown in Fig. 15. Table 6 Classification report for model 1 Brodatz dataset
precision recall f1-score support
Accuracy − − 0.9967 896
Macro Avg 1.00 1.00 1.00 896
Weighted Avg 1.00 1.00 1.00 896
Fig. 15 Model 1 accuracy and losses graph for the Brodatz datset
The second model is developed using the InceptionV3 model trained on the ImageNet dataset. The top layer of the pre-trained model is removed and replaced by a softmax layer with 112 classes. The pre-trained model was used as a feature extractor, i.e. all the layers of the pre-trained model were frozen, and only the top layer was trained on the training dataset. The proposed model was trained for 7 epochs on the training dataset. The model achieved an accuracy of 99.8884% on the testing dataset. The classification report of model 2 on testing it on the testing data is shown in Table 7. The accuracy vs epochs graph and the loss vs epochs graph of model1 for the Brodatz dataset while training is shown in Fig. 16. Table 7 Classification report for model 2 Brodatz dataset
precision recall f1-score support
Accuracy − − 0.9933 896
Macro Avg 0.99 0.99 0.99 896
Weighted Avg 0.99 0.99 0.99 896
Fig. 16 Model 2 accuracy and losses graph for the Brodatz datset
Outex dataset
The third dataset to be studied was the Outex dataset. The first model is developed using the MobileNetV3 small model, trained on the ImageNet dataset. The top layer of the pre-trained model is removed and replaced by a softmax layer with 112 classes. The pre-trained model was used as a feature extractor, i.e. all the layers of the pre-trained model were frozen, and only the top layer was trained on the training dataset. The proposed model was trained for 5 epochs on the training dataset. The model achieved an accuracy of 99.479% on the testing dataset. The classification report and confusion matrix of model 1 on testing it on testing data is shown in Table 8 and Fig. 17 respectively. The accuracy vs epochs graph and the loss vs epochs graph of model1 for the Outex dataset while training is shown in Fig. 18. Table 8 Classification report for model 1 Outex dataset
precision recall f1-score support
Accuracy − − 1.00 192
Macro Avg 1.00 1.00 1.00 192
Weighted Avg 1.00 1.00 1.00 192
Fig. 17 Model 1 confusion matrix for the Outex datset
Fig. 18 Model 1 accuracy and losses graph for the Outex datset
The second model is developed using the InceptionV3 model trained on the ImageNet dataset. The top layer of the pre-trained model is removed and replaced by a softmax layer with 112 classes. The pre-trained model was used as a feature extractor, i.e. all the layers of the pre-trained model were frozen, and only the top layer was trained on the training dataset. The proposed model was trained for 5 epochs on the training dataset. The model achieved an accuracy of 99.479% on the testing dataset. The classification report and confusion matrix of model 2 on testing it on testing data are shown in Table 9 and Fig. 19 respectively. The accuracy vs epochs graph and the loss vs epochs graph of model1 for the Outex dataset while training is shown in Fig. 20. Table 9 Classification report for model 2 Outex dataset
precision recall f1-score support
Accuracy − − 1.00 192
Macro Avg 1.00 1.00 1.00 192
Weighted Avg 1.00 1.00 1.00 192
Fig. 19 Model 2 confusion matrix for the Outex datset
Fig. 20 Model 2 accuracy and losses graph for the Outex datset
Comparative study
The results of the 2 proposed models are compared with other recently proposed models. Table 10 shows the comparison between the two proposed models and other recently applied models on the Kylberg dataset. Table 11 shows the comparison between the two proposed models and other recently applied models on the Brodatz dataset. Table 12 shows the comparison between the two proposed models and other recently applied models on the Outex dataset. Table 10 Performance comparison of our models with the existing techniques for the Kylberg dataset
Paper (reference Model/technique Classifcation accuracy (%)
Andrearczyk and Whelan [2] T-CNN-3 99.4 ± 0.2
Kaya et al. [15] KNN+nLBP(d = 1) 99.64
El Khadiri et al. [8] RALBGC 99.23
Kaya et al. [15] LBP 97.97
Dixit et al. [6] Modifed CNN+WOA 99.71
Proposed model 1 MobileNetV3 (Fully fined tuned) 100
Proposed model 2 InceptionV3 (Feature extraction) 100
Table 11 Performance comparison of our models with the existing techniques for the Brodatz dataset
Paper (reference) Model/technique Classifcation accuracy (%)
Kaya et al. [15] LBPû2̂ and nLBP_d 99.26
El Khadiri et al. [8] RALBGC, RLBGC 100
de Mesquita Sá Junior and Backes [25] ELM based Signature (Ψ19,39) 99.42
Ahmadvand and Daliri [1] Hybrid feature vector 89.28
Dixit et al. [6] Modifed CNN+WOA 97.43
Proposed model 1 MobileNetV3 (Feature extraction) 99.67
Proposed model 2 InceptionV3 (Feature extraction) 99.33
Table 12 Performance comparison of our models with the existing techniques for the Outex TC-00012 dataset
Paper (reference Model/technique Classifcation accuracy (%)
Ahmadvand and Daliri [1] Hybrid feature vector 90.78
Mehta and Egiazarian [24] FbLBP 96.00
Dixit et al. [6] Modifed CNN+WOA 97.70
Proposed model 1 MobileNetV3 (Feature extraction) 100
Proposed model 2 InceptionV3 (Feature extraction) 100
Discussion
In this experiment, the pre-trained models used are trained on the ImageNet dataset and openly available for use. The models were trained and tested using two cases. In the first case, the pre-trained model used a feature extractor, and only the last layer was trained on the dataset. The whole model was trained on the training dataset in the second case. The feature extraction case yielded better results and lesser training time in most cases. As mentioned in Section 3.2, all the images were rescaled to a size of 224*224*3 to make them compatible with the pre-trained models. The datasets were then split in a ratio of 80:20 for training and testing data. Tables 10, 11, and 12 in Section 4.5 showcase the comparison of the results of our method with the previously proposed methods. From the tables, it is evident that our methods have outperformed the previously proposed methods.
Conclusion and future Scope
Texture classification is an essential area of research that has attracted many researchers to propose different models. From the comparative study, it can be concluded that our models give better results than most of the existing models for the Kylberg and Outex datasets. Both models got a testing accuracy of 100 on the Kylberg and Outex datasets. Our models gave competitive results for the Brodatz dataset too. Despite using the models as only feature extractors (except for MobileNetV3 on the Kylberg dataset), the models have attained outstanding results. It means that the datasets in the study and the ImageNet dataset have very similar feature space. Hence, it can be concluded that transfer learning can be used to quickly solve tasks where the feature space of the target dataset is similar to the feature space of the dataset on which the pre-trained model is trained.
In future, we would like to test our models on more texture datasets and even use them for other domains like medical and aerial imagery. It is evident that the similarity of feature space of the source and target dataset has a massive impact on the model performance. This study used models which were trained on the ImageNet datasets. The authors also aim to extend this work to transformer based architectures. We would also like to expand the study by using the same model architectures trained on a different dataset. Using different source models for standard architectures and different target models can help understand transfer learning deeper.
Funding
No funding was received to assist with the preparation of this manuscript.
Data Availability
The Brodatz [4] and the Kylberg [20] datasets are publicly available and can be accessed using the link mentioned in the citation. The Outex [28] dataset is available on request from the authors of the cited paper.
Declarations
Conflict of Interests
The authors declare that they have no confict of interest.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Ahmadvand A Daliri MR Invariant texture classification using a spatial filter bank in multi-resolution analysis Image Vis Comput 2016 45 1 10 10.1016/j.imavis.2015.10.002
2. Andrearczyk V Whelan P Using filter banks in convolutional neural networks for texture classification Pattern Recogn Lett 2016 84 63 69 10.1016/j.patrec.2016.08.016
3. Arora V Ng EYK Leekha RS Darshan M Singh A Transfer learning-based approach for detecting covid-19 ailment in lung ct scan Comput Biol Med 2021 135 104575 10.1016/j.compbiomed.2021.104575 34153789
4. Brodatz P (1966) Textures: A photographic album. Accessed June 2021. http://sipi.usc.edu/database/database.php?volume=textures
5. Di Ruberto C Histogram of radon transform and texton matrix for texture analysis and classification IET Image Process 2017 11 9 760 766 10.1049/iet-ipr.2016.1077
6. Dixit U Mishra A Shukla A Tiwari R Texture classification using convolutional neural network optimized with whale optimization algorithm SN Appli Sci 2019 1 6 655 10.1007/s42452-019-0678-y
7. Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X, Unterthiner T, Dehghani M, Minderer M, Heigold G, Gelly S, et al. (2020) An image is worth 16x16 words: transformers for image recognition at scale. arXiv:2010:11929
8. El Khadiri I Kas M El Merabet Y Ruichek Y Touahni R Repulsive-and-attractive local binary gradient contours: new and efficient feature descriptors for texture classification Inf Sci 2018 467 634 653 10.1016/j.ins.2018.02.009
9. Feng J Liu X Dong Y Liang L Pu J Structural difference histogram representation for texture image classification IET Image Process 2017 11 118 125 10.1049/iet-ipr.2016.0495
10. Haralick RM, Shanmugam K, Dinstein I (1973) Textural features for image classification. IEEE Transactions on Systems Man, and Cybernetics SMC-3(6),610–621. 10.1109/TSMC.1973.4309314
11. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: 2016 IEEE conference on computer vision and pattern recognition (CVPR), pp 770–778. 10.1109/CVPR.2016.90
12. Howard A, Sandler M, Chu G, Chen LC, Chen B, Tan M, Wang W, Zhu Y, Pang R, Vasudevan V et al (2019) Searching for mobilenetv3. In: Proceedings of the IEEE/CVF international conference on computer vision, pp 1314–1324
13. Jain A Rao ACS Jain PK Abraham A Multi-type skin diseases classification using op-dnn based feature extraction approach Multimedia Tools and Applications 2022 81 5 6451 6476 10.1007/s11042-021-11823-x 35035267
14. Kalita DJ Singh VP Kumar V A dynamic framework for tuning svm hyper parameters based on moth-flame optimization and knowledge-based-search Expert Syst Appl 2021 168 114139 10.1016/j.eswa.2020.114139
15. Kaya Y Ertuğrul OF Tekin R Two novel local binary pattern descriptors for texture analysis Appl Soft Comput 2015 34 C 728 735 10.1016/j.asoc.2015.06.009
16. Kazi A Panda SP Determining the freshness of fruits in the food industry by image classification using transfer learning Multimedia Tools and Applications 2022 81 6 7611 7624 10.1007/s11042-022-12150-5
17. Krizhevsky A, Sutskever I, Hinton GE (2012) Imagenet classification with deep convolutional neural networks. In: Proceedings of the 25th International conference on neural information processing systems-volume 1, NIPS’12. Curran Associates Inc., Red Hook, NY, USA, pp 1097–1105
18. Krizhevsky A Sutskever I Hinton GE Imagenet classification with deep convolutional neural networks Commun ACM 2017 60 6 84 90 10.1145/3065386
19. Kundu R Singh PK Ferrara M Ahmadian A Sarkar R Et-net: an ensemble of transfer learning models for prediction of covid-19 infection through chest ct-scan images Multimedia Tools and Applications 2022 81 1 31 50 10.1007/s11042-021-11319-8 34483709
20. Kylberg G (2011) The kylberg texture dataset v. 1.0. External report (Blue series) 35, Centre for Image Analysis, Swedish University of Agricultural Sciences and Uppsala University, Uppsala, Sweden. Accessed June 2021. http://www.cb.uu.se/gustaf/texture/
21. LeCun Y Boser B Denker JS Henderson D Howard RE Hubbard W Jackel LD Backpropagation applied to handwritten zip code recognition Neural Comput 1989 1 4 541 551 10.1162/neco.1989.1.4.541
22. Liu XJ Li KL Luan HY Wang WH Chen ZY Few-shot learning for skin lesion image classification Multimedia Tools and Applications 2022 81 4 4979 4990 10.1007/s11042-021-11472-0
23. Lu SY Wang SH Zhang YD A classification method for brain mri via mobilenet and feedforward network with random weights Pattern Recogn Lett 2020 140 252 260 10.1016/j.patrec.2020.10.017
24. Mehta R Egiazarian K Dominant rotated local binary patterns (drlbp) for texture classification Pattern Recogn Lett 2016 71 C 16 22 10.1016/j.patrec.2015.11.019
25. de Mesquita Sá Junior JJ Backes AR Elm based signature for texture classification Pattern Recogn 2016 51 395 401 10.1016/j.patcog.2015.09.014
26. Nadeem Z Khan Z Mir U Mir UI Khan S Nadeem H Sultan J Pakistani traffic-sign recognition using transfer learning Multimed Tools Appl 2022 81 6 8429 8449 10.1007/s11042-022-12177-8
27. Nasirzadeh M, Khazael AA, Khalid MB (2010) Woods recognition system based on local binary pattern. In: Proceedings of the 2010 2nd international conference on computational intelligence, communication systems and networks, CICSYN ’10. IEEE Computer Society, USA, pp 308–313, DOI 10.1109/CICSyN.2010.27
28. Ojala T Pietikainen M Maenpaa T Multiresolution gray-scale and rotation invariant texture classification with local binary patterns IEEE Trans Pattern Anal Mach Intell 2002 24 7 971 987 10.1109/TPAMI.2002.1017623
29. Picard R, Kabir T, Liu F (1993) Real-time recognition with the entire brodatz texture database. In: Proceedings of IEEE conference on computer vision and pattern recognition, pp 638–639. 10.1109/CVPR.1993.341050
30. Pritt M, Chern G (2017) Satellite image classification with deep learning. In: 2017 IEEE applied imagery pattern recognition workshop (AIPR), pp 1–7. 10.1109/AIPR.2017.8457969
31. Ramola A Shakya AK Van Pham D Study of statistical methods for texture analysis and their modern evolutions Eng Reports 2020 2 4 e12149 10.1002/eng2.12149
32. Sana JK Islam MM Plt-based spectral features for texture image retrieval IET Image Process 2018 12 11 2065 2074 10.1049/iet-ipr.2018.5604
33. Sandler M, Howard A, Zhu M, Zhmoginov A, Chen LC (2018) Mobilenetv2: inverted residuals and linear bottlenecks. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 4510–4520
34. Shallu MR Breast cancer histology images classification: Training from scratch or transfer learning? ICT Express 2018 4 4 247 254 10.1016/j.icte.2018.10.007
35. Simon P Vijayasundaram U Deep learning based feature extraction for texture classification Procedia Comput Sci 2020 171 1680 1687 10.1016/j.procs.2020.04.180
36. Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. arXiv:1409.1556
37. Singh VP Srivastava R Improved image retrieval using fast colour-texture features with varying weighted similarity measure and random forests Multimedia Tools and Applications 2018 77 11 14435 14460 10.1007/s11042-017-5036-8
38. Smith J, Chang SF (1994) Transform features for texture classification and discrimination in large image databases. In: Proceedings of 1st international conference on image processing, vol 3, pp 407–411. 10.1109/ICIP.1994.413817
39. Szegedy C, Liu W, Jia Y, Sermanet P, Reed S, Anguelov D, Erhan D, Vanhoucke V, Rabinovich A (2015) Going deeper with convolutions. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1–9
40. Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z (2016) Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 2818–2826
41. Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser Ł, Polosukhin I (2017) Attention is all you need. Advances in neural information processing systems 30
42. Xu P, Yao H, Ji R, Sun X, Liu X (2011) A robust texture descriptor using multifractal analysis with gabor filter. In: Proceedings of the second international conference on internet multimedia computing and service, ICIMCS ’10, pp 147–150. Association for Computing Machinery, New York, NY, USA. 10.1145/1937728.1937763
43. Yuan X, Yang Z, Zouridakis G, Mullani N (2006) Svm-based texture classification and application to early melanoma detection. In: 2006 international conference of the IEEE engineering in medicine and biology society, pp. 4775–4778. 10.1109/IEMBS.2006.260056
44. Zheng Y, Zhong G, Liu J, Cai X, Dong J (2014) Visual texture perception with feature learning models and deep architectures. 10.1007/978-3-662-45646-0_41
| 0 | PMC9739347 | NO-CC CODE | 2022-12-14 23:29:54 | no | Multimed Tools Appl. 2022 Dec 10;:1-24 | utf-8 | Multimed Tools Appl | 2,022 | 10.1007/s11042-022-14276-y | oa_other |
==== Front
J Fam Econ Issues
J Fam Econ Issues
Journal of Family and Economic Issues
1058-0476
1573-3475
Springer US New York
9877
10.1007/s10834-022-09877-6
Original Paper
How Relational Conflict Harms Family Firm Performance: The Mediating Role of Family Social Capital and the Moderating Role of Family Ownership
http://orcid.org/0000-0003-4777-5498
Rosecká Nikola [email protected]
Machek Ondřej [email protected]
grid.266283.b 0000 0001 1956 7785 Faculty of Business Administration, Prague University of Economics and Business, Nam. W. Churchilla 4, 130 67 Prague, Czech Republic
10 12 2022
116
23 4 2022
10 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.
While many researchers suggest that relational conflict has adverse performance effects in family firms, the exact mechanisms through which conflict harms performance are rarely empirically investigated. This paper explores the role of family social capital in the relationship between relational conflict and family firm performance. We hypothesize that the negative relationship between relational conflict and family firm performance is partially mediated by family social capital, while family ownership moderates the relationship between relational conflict and family social capital. In a sample of 175 U.S.-based small and medium-sized family firms recruited through Prolific Academic, we find that relational conflict harms firm performance indirectly through the erosion of family social capital. However, no evidence of a direct negative effect of relational conflict on performance is found. Our results also indicate that the negative relationship between relational conflict and family social capital is intensified by family ownership. We discuss the implications and contributions and present relevant directions for future research.
Keywords
Relational conflict Family social capital Family ownership Family firm performance
Internal Grant Agency, Prague University of Economics and BusinessF3/8/2021
==== Body
pmcIntroduction
Because of the family and business overlap in family firms, family businesses are said to be “plagued by conflicts” (Levinson, 1971). Interpersonal conflicts in family firms have attracted the attention of professionals and scholars since the emergence of family business research. Most often, they are perceived as a dysfunctional phenomenon that harms the family and the firm (Kubíček & Machek, 2020). The psychology literature distinguishes multiple types of conflict. Disagreements about tasks or problems at hand, referred to as substantive or cognitive conflicts, can benefit firm performance as they allow for increased understanding of the tasks and initiate the exchange of ideas and opinions (Jehn, 1995). However, personal and emotional conflicts, known as affective or relational conflicts, do not seem to have any benefits related to performance (De Dreu & Weingart, 2003). The family business literature corroborates these results (Eddleston & Kellermanns, 2007; Ensley et al., 2007; Nosé et al., 2017), finding that relational conflict is detrimental to family firm performance. Nevertheless, the exact mechanisms of how relational conflict harms family firm performance are rarely empirically evaluated.
To understand the adverse effects of relational conflict in family firms, it is critical to discuss the factors that make family firms different from their nonfamily counterparts. One of these factors is the presence of family ties in the business. Compared to other organizational forms of business, these ties are considered stronger, more intense, and more long-lasting (Hoffman et al., 2006). Consequently, a substantial body of family business literature concentrated on the uniqueness of family ties and their potential to create competitive advantage (Arregle et al., 2007; Hoffman et al., 2006; Pearson et al., 2008). Broadly, the wealth embedded in and engendered by social ties is referred to as social capital (Adler & Kwon, 2002). The academic literature presents several categorizations of social capital. First, social capital can refer to internal (bonding) social ties, which connect people inside a social group or organization, and external (bridging) social ties related to outside networks of contacts (Putnam, 2000). Social capital is a multidimensional construct, having a structural dimension (represented by closure, density, and connectivity of social ties), relational dimension (trust, norms, and obligations), and cognitive dimension (shared representations and meanings;Nahapiet & Ghoshal, 1998). In the unique context of family firms, social capital takes a particular form because of the temporal stability, interrelatedness, and closure of family ties (Pearson et al., 2008). The internal social capital in family firms is referred to as family social capital (Arregle et al., 2007). Like organizational social capital, family social capital is assumed to be composed of its structural, relational, and cognitive dimensions (Carr et al., 2011; Herrero & Hughes, 2019). However, it has several unique attributes: it is hardly imitable by nonfamily firms (Herrero, 2018), it is transferred and inherited during the succession process (Aragón‑Amonarriz et al., 2019), and it takes longer to develop (Arregle et al., 2007).
Apart from family social capital, another distinguishing feature of family firms is their pursuit of family-centred goals, which is theoretically explained by their propensity to protect and build their socioemotional wealth (Gómez-Mejía et al., 2007). Socioemotional wealth (SEW) can be defined as the non-economic utility derived from ownership and involvement in the family firm (Martin & Gómez-Mejía, 2016). The extent to which family firms emphasise family-centred goals varies. While family firms need to maintain family involvement to be considered “family firms” (Chrisman et al., 2012), the degree to which they do so can vary from complete family control to “enough but not too much” (Stewart & Hitt, 2012). In other words, some family firms are “pure family firms”, while others are more similar to nonfamily businesses. Thus, family firms display vast within-group differences, and recent family business literature calls for further investigation of their heterogeneity (Daspit et al., 2021; Haynes et al., 2021). One factor that creates this heterogeneity is the concentration of firm ownership in family hands. Family ownership has important consequences related to family firms’ behaviors as the firm’s actions are driven by utilities derived by dominant family owners (Gómez-Mejía et al., 2011). For instance, high family ownership “leads to an emphasis on particular types of stakeholders and a particular type of socially oriented behavior” (Gómez-Mejía et al., 2011, p. 681), making family firms responsive to environmental concerns (Berrone et al., 2010). Family ownership gives family members greater power to influence a firm’s strategy and decisions, providing the firm with the ability to behave in a way that is distinctively supportive of the family (Evert et al., 2018). When family ownership is high, the financial and socioemotional wealth of family members is at stake, making family firms take actions that are risk-averse and focused on SEW preservation (Evert et al., 2018). Although family ownership is not directly linked to family firm performance (Gómez-Mejía et al., 2011), in the context of relational conflicts in family firms, it becomes an important factor to study. Under normal circumstances, family ownership produces emotional attachment and leads to family-firm identification and seeing the firm as “our business” (Kotlar et al., 2020). However, relational conflicts are associated with socioemotional costs (Rousseau et al., 2018), and under high emotional attachment and family-firm identification of family members, they can be even more detrimental.
To shed more light on how relational conflict harms family performance, the current study investigates how relational conflict affects family firms’ performance and their family social capital, as well as the role of family ownership in these relationships. The rest of this paper proceeds as follows. First, we present the relevant theoretical background and develop the hypotheses. Second, we discuss the sampling and data analysis methods. Subsequently, we present and discuss the results. Finally, we provide concluding remarks.
Theoretical Background and Hypotheses
Relational Conflict and Family firm Performance
There is a consensus in the intragroup conflict literature that relational conflict has the potential to have direct adverse effects on the performance of social groups. Relational conflict produces negative feelings such as anxiety and fear, severely affecting individuals. Jehn (1995) found that relational conflict reduces group members’ satisfaction and willingness to remain in the social group. The negative feelings associated with relational conflict also reduce group members’ cognitive functioning and ability to process information (Jehn, 1995) and inhibit their creative thinking (Carnevale & Probst, 1998). Instead of focusing on tasks that must be accomplished, group members spend their time and energy resolving (or trying to ignore) the conflicts (Jehn & Mannix, 2001). De Dreu and Weingart’s (2003) meta-analysis revealed that relational conflict is indeed negatively associated with individuals’ satisfaction and team performance. The findings are supported in the context of family firms, where relational conflict produces animosity, stress, hostile behaviours, and insufficient attention devoted to family business needs (Eddleston & Kellermanns, 2007). Thus, like previous authors, we expect that:
Hypothesis 1
There is a negative relationship between relational conflict and family firm performance.
Family Social Capital and Family firm Performance
In their seminal paper, Pearson et al. (2008) discuss family firms’ unique resources and capabilities caused by family members’ involvement and interactions and posit that “familiness”, or family social capital (Arregle et al., 2007), is composed of structural, relational, and cognitive dimensions, just as organizational social capital. It should be noted that family social capital is most often considered to be an “internal”, bonding type of social capital (“strong ties”), and our discussion will therefore not focus on bridging social capital (“weak ties”). The entrepreneurship literature does not present a consensus on whether internal social capital is always beneficial for firms. On the one hand, due to close relationships, trust, and shared vision, social capital creates the potential for solidarity and collaboration, making people in a firm more attentive to the firm’s goals (Adler & Kwon, 2002). When internal social capital is strong, employees are willing to subordinate their individual goals and actions to collective goals and actions (Leana & Van Buren, 1999). Sharing of information, which results in richer information for decision-making, and a shared understanding of roles also increase the quality of decision-making (Mustakallio et al., 2002). Thus, internal social capital has the potential to contribute to firm performance. On the other hand, some authors argue that social capital also has its “dark side” (Gargiulo & Benassi, 1999). Excessive internal social capital can lead to group-thinking, rigidity, or inability to access novel ideas and resources (Stam et al., 2014). For instance, excessive trust can lead to an overreliance on close contacts and their networks, thus restricting the ability to find, evaluate and exploit opportunities (Shi et al., 2015). Thus, excessive internal social capital can limit growth opportunities.
Arregle et al. (2007) and Pearson et al. (2008) assume that family social capital presents a competitive advantage vis-à-vis nonfamily firms. In the family business setting, one of the rare studies suggesting evidence of “dark side” of family social capital was presented by Herrero and Hughes (2019). In their study, social capital’s relational and cognitive dimensions contribute to performance, but the structural dimension displays an inverse U-shaped relationship with performance. Apart from Herrero and Hughes’s (2019) study, the family business literature is relatively unanimous regarding the positive effects of family social capital on performance (e.g., Duarte Alonso et al., 2020). Family social capital has been found to improve firm performance by enhancing knowledge integration (Kansikas & Murphy, 2011), knowledge internalization and product development (Chirico & Salvato, 2016), innovation (Sanchez-Famoso et al., 2019), fostering social cohesion and work climate (Ruiz Jiménez et al., 2013), improving family firm resilience during difficult times (Wiatt et al., 2020), and creating strong bonds with the community (Duarte Alonso et al., 2020). As all the above factors can be drivers of firm performance, then, we also expect that:
Hypothesis 2
There is a positive relationship between family social capital and family firm performance.
Relational Conflict and Family Social Capital
Family studies suggest that family members who experience conflict with other family members tend to avoid mutual interactions. For instance, adolescents avoid interaction with parents if they observe hostile, impulsive and inconsistent family conflict (Cooper, 1988) or have a bad conflict experience (Noller et al., 2006). When marital conflict is present in the family, parents tend to devote less engagement to the parent-child relationship and become less attentive, thus reducing their interactions with children (Buehler & Gerard, 2002). Conflicts between siblings are characterized by a lack of active conflict resolution efforts; often, siblings simply ignore each other, which can have long-term consequences (Vandell & Bailey, 1992). In extremely conflicted families, family ties can be broken when, for instance, an adolescent is told to leave or a family member proposes divorce (Aldridge, 1984). Hence, relational conflict has the potential to harm social interactions among family members, weaken family ties and their connectivity, thus harming the essential components of the structural dimension of social capital. Moreover, relational conflict can also affect “appropriable organization” (Coleman, 1988). Family members involved in conflict tend to perpetuate the toxic behaviours in their peer- or nonfamily social groups. For example, adolescents adopt their parents’ hostility and adverse conflict behaviours (Taylor & Segrin, 2010). Children’s direct exposure to parental conflicts increases the likelihood of imitating parents’ conflict behaviours in relationships with nonfamily members (Buehler et al. 2009). Thus, family conflict can permeate the whole organization (Barach & Ganitsky, 1995), reducing the ability to use and reuse family ties (Pearson et al., 2008) in the family-firm context. In sum, relational conflict can be expected to harm the structural dimension of family social capital.
The emotionality of relational conflict also seems to affect the relational dimension of social capital, i.e., trust, norms, obligations, and identification. Broadly, trust is developed through perceived trustworthiness and affect (Williams, 2001). Perceived trustworthiness is a cognitive determinant of trust related to perceived abilities, benevolence, and integrity (Mayer & Davis, 1999). On the other hand, positive affect, i.e. positive emotions and moods, is one of the predictors of affective (“deep”) trust. Negative affect produced by relational conflict signals the presence of a threatening environment, thus creating negative feelings and perceptions of other people involved in the conflict (Lu et al., 2017). These feelings and perceptions, in turn, determine how an individual’s trust develops over time (Williams, 2001). Through emotional contagion, individuals’ negative emotions can spread throughout the social group, leading to a low level of trust in the whole group (Fulmer & Gelfand, 2012). The psychology literature presents similar findings for identification, suggesting that an individual’s identification with the group decreases with negative emotions such as anger (Kessler & Hollbach, 2005). Thus, by virtue of negative affect, relational conflict can be assumed to hinder the relational dimension of family social capital.
Relational conflict is also likely associated with the cognitive dimension of social capital, i.e. shared vision and language. First, one of the most commonly discussed cognitive manifestations of interpersonal conflict is disagreement. Barki and Hartwick (2004) define conflict as a “process that occurs between independent parties as they experience negative emotional reactions to perceived disagreements and interference with the attainment of their goals”. Thus, it can be assumed that relational conflict reduces shared cognitions about collective goals (Knight et al., 1999). Second, relational conflict weakens social ties among family members and their social interactions. However, developing shared vision, shared language, and collective narratives among family members requires close social relations and frequent social interactions (Nahapiet & Ghoshal, 1998; Mustakallio et al., 2002). Social interactions can only be built with a positive affective tone (Pelled & Xin, 1999); negative affect, typical for relational conflict, will hinder their development. Overall, then, it can be assumed that relational conflict harms the cognitive dimension of social capital by reducing mutual agreement, quality of relationships, and frequency of social interactions among family members.
As we have shown, psychology literature and family business literature present compelling arguments why relational conflict can harm all three dimensions of family social capital: structural, relational, and cognitive. In addition, according to the process perspective of social capital, the development of family social capital requires stability, closure, interdependence, and interactions (Pearson et al., 2008), all of which can be assumed to be negatively affected by an escalated interpersonal conflict between family members. To sum up, it can be expected that:
Hypothesis 3
There is a negative relationship between relational conflict and family social capital.
Hypotheses 1, 2, and 3 constitute a partial mediation model in which relational conflict has both direct and indirect negative effects (through family social capital) on family firm performance. Consequently, taking the arguments as a whole, we expect that:
Hypothesis 4
The relationship between relational conflict and family firm performance is partially mediated by family social capital.
Moderating Effects of Family Ownership
Hypothesis 3
expects a negative relationship between relational conflict and family social capital. Presumably, in family firms, relational conflict escalates more easily and rapidly than in nonfamily firms (Frank et al., 2011). It can be expected that the differences in family-firm identification and emotional attachment among family firms will also cause differences in conflict and emotional dynamics. First, the negative affect produced by relational conflict represents a socioemotional cost (Rousseau et al., 2018). While relational conflict produces negative affect itself, ceteris paribus, in families that strongly perceive the firm as “their business”, the level of stress and negative affect can be considered to be higher because family members feel emotionally invested, and the wellbeing of the family is put at danger (Kotlar et al., 2020). Second, a possible economic and socioemotional loss can create the need to leave the situation (Rousseau et al., 2018); yet, family members face great exit barriers and cannot simply leave. These exit barriers can be assumed to be greater under high family ownership since for most owners, the loss of the firm represents a highly emotional event (Berrone et al., 2012), and thus, family ownership has high emotional value (Zellweger & Astrachan, 2008). When family ownership is high and relational conflict occurs, family members face the need to leave while being unable to leave, which results in additional strain and negative affect. Third, under normal circumstances, when a family member feels threatened, trhey are likely to approach other family members to address the threat. When relational conflict occurs in the family, the family system cannot coordinate family members to collectively protect the sense of belonging and emotional attachment created by family ownership (Kotlar et al., 2020). Family members are left “alone” in a socioemotional loss perspective, intensifying the already-existing negative affect. The intensified negative affect acts as an “exacerbator” of the conflict-outcome relationship (Adams & Laursen, 2007; Jehn & Bendersky, 2003). Finally, family firms with concentrated family ownership can be reluctant to seek help outside of the family circle due to their emotional resistance (McCann et al., 2004) and hubristic overestimation of own abilities to resolve the conflicts (Haynes et al., 2015). The negative effects of relational conflict on family social capital cannot be effectively mediated by (or be transferred to) nonfamily members who can otherwise diffuse family tensions and reduce family conflicts (Rosecká & Machek, 2021). In sum, it can be expected that high family ownership magnifies the adverse effects of relational conflict on family social capital by increasing the extent of negative affect produced by relational conflict and by virtue of lower ability to use conflict-resolution mechanisms. In other words, we posit that:
Hypothesis 5
The relationship between relational conflict and family social capital is moderated by family ownership.
The Moderated Mediation Model
The combination of the expectations of Hypotheses 4 and 5 leads to a moderated mediation model between relational conflict and family firm performance, in which family social capital acts as a mediating variable and family ownership as a moderating variable. In other words, besides harming family firm performance directly, relational conflict is expected to harm family social capital (with family ownership acting as a moderator), which, in turn, negatively affects performance. Formally, we expect that:
Hypothesis 6
The relationship between relational conflict and family firm performance is partially mediated by family social capital, with family ownership acting as a moderator of the relational conflict-family social capital relationship.
Methods
Data Collection
An online questionnaire was prepared in Qualtrics and distributed through Prolific Academic (e.g., Derfler-Rozin et al., 2021) to managers of small and medium-sized family firms in the United States. The choice of the data collection method builds on the fact that online panel data has been increasingly used in recent years (Porter et al., 2019) while providing results that converge to conventional datasets in terms of data quality (Walter et al., 2019). The survey was carried out in the first quarter of 2021 and consisted of two waves. The objective of the first wave (“prescreening”) was to recruit managers who (1) would describe their firm as a family business and (2) were members of the controlling family. Thus, in the sample, we identify family firms based on their self-perception as a family firm (Zahra, 2003). In the prescreening stage, we addressed a total of 1,700 managers, out of which 297 met our selection criteria. In the second wave, we distributed the full survey to these 297 respondents and obtained 229 complete responses. Besides firm demographic variables and constructs employed in this study (the Appendix provides a summary list of measures), we asked the respondents to briefly describe their family business. This description served as an initial check of respondents’ efforts to provide meaningful responses. We eliminated 41 observations in case we were dissatisfied with the description (e.g. the person did not seem to be a family member or provided only superficial information). The survey also contained several attention checks (e.g., “Please write below the word ‘red’). Eight respondents failed to pass the attention checks, and their responses were removed from the sample (Jackson et al., 2016). Finally, we did not admit five respondents who seemed to use virtual private servers to hide their location (Winter et al., 2019).
The final sample consists of 175 U.S. family business managers. The companies in the sample can be classified as microenterprises with 5–10 employees (27.6%), small companies with 10–49 employees (65.4%), and medium-sized companies with 50–499 employees (14.1%). The family firms are controlled by the first generation (24%), second generation (47.5%), third generation (25.1%), or fourth generation (3.4%). The majority of companies were affiliated with the services sector (50.8%), followed by manufacturing (17.1%), wholesale and retail (16%), construction (6.8%), and other industries (9.1%). Several factors complicate the assessment of sample representativeness. The descriptive data of the respondents should be compared with those of the population of U.S. family firms. Because there is no official database of such businesses, and the U.S Census does not track them (Chang et al., 2008), it is hard to know exactly what characteristics apply to this population. In addition, we are unable to draw valid comparisons between our sample and the samples of previous family business studies carried out in the US, as these studies frequently use convenience samples (Gedajlovic et al., 2012). The existing surveys of American family businesses are very diverse, too. Table 1 illustrates this diversity by showing family firms’ industry affiliation in selected surveys conducted in the US The examples include two representative family business surveys (e.g., Haynes et al., 2021), one large-sample survey conducted among SMEs in Florida, and two recent family business reports of advisory organizations.
Table 1 Industry affiliation in the sample and in previous representative family business surveys
Survey Sample size Manufacturing [%] Construction [%] Wholesale and
retail trade [%] Services [%] Notes
This study 175 17.1 6.8 16 50.8
National Family Business Survey (2000) 708 7.2 n/a 23.4 48 Construction not presented as a separate category.
American Family Business Survey (2003) 1,143 24.5 12.2 27.7 n/a Wholesale and distribution represents one broader category. Services not presented as a separate category.
Florida Small Business Development Center (2020) 3,633 12.3 7.6 13.8 66.3 The figures refer to nine regions in Florida.
KPMG UAE Family Business Report: COVID-19 Edition (2021) 498 20 8 n/a 66 The figures refer to “Americas”, including South America and the Caribbean. Wholesale and retail not presented as a separate category.
Family Enterprise USA Family Business Survey (2021) 172 23.5 6 n/a n/a Wholesale not presented as a separate category. Instead, retail & consumer durables is considered.
Notes: The frequencies for the National Family Business Survey were taken from Puryear et al. (2008)
The National Family Business Survey (NFBS) was conducted in 1997/2000. Most NFBS survey participants were microenterprises with fewer than five employees (Puryear et al., 2008). According to the U.S. Census, most American companies are indeed microenterprises—with different industrial characteristics from small and medium-sized firms—and thus NFBS could not serve as a reference survey for our study. Further, the American Family Business Survey (AFBS), conducted in 2003, presents wholesale and distribution as a broader industry category and does not present the percentage of firms operating in services. On the other hand, the generational involvement of firms surveyed in AFBS is remarkably similar to our sample (first generation = 27%, second generation = 43%, third generation = 23%). A survey among 3,633 small businesses was conducted in 2019 by Florida Small Business Development Center (SBDC). The industry affiliation of the firms in the sample is fairly similar to ours (services = 66.3%, manufacturing = 12.3%, construction = 7.6%, wholesale and retail = 13.8%). The fourth example, KPMG UAE Family Business Report conducted in 2021, surveyed family firms operating in the broader region of “Americas” and provides similar figures (services = 66%, manufacturing = 20%, construction = 8%). However, the survey does not present wholesale and retail as a separate sector. Finally, we consider the annual survey conducted by Family Enterprise USA in 2021. While the proportion of construction (6%) is comparable to our study, the survey does not publish information on wholesale trade and services, thus preventing us from assessing the similarities. While the descriptive elements of our data set are, to some extent, comparable to previous representative surveys, our sample must be interpreted as a convenience sample, which is likely not representative of the entire population of family firms in the U.S. Thus, our sample offers particular insights into relational conflict in small and medium-sized U.S. family firms, rather than family microenterprises (see Table 1).
Measures
Dependent variable. To assess family firm performance, we asked respondents to indicate on a five-point Likert scale the extent to which they were satisfied with the firm’s current performance relative to the rivals. Performance was measured using three individual items: net profit growth, market share, and sales (Cooper & Artz, 1995).
Independent variable. To measure relational conflict, we employ the intragroup conflict scale developed by Jehn (1995) with item wordings adapted to the family business context (Paskewitz & Beck, 2017). Respondents were asked to indicate their agreement on a five-point Likert scale with statements such as “There is much relationship conflict among family members in our firm” or “Family members often get angry with each other while working in our firm”.
Mediator. The mediator, family social capital, was measured using the ISC-FB scale developed by Carr et al. (2011). The scale has twelve items and consists of the structural, relational, and cognitive dimensions (Nahapiet & Ghoshal, 1998). Due to the high interdependence between the three dimensions, we included them under a single construct.
Moderator. To measure family ownership, we employ a single item in which respondents were asked to indicate the percentage of ownership in the hands of the controlling family (e.g., Adithipyangkul et al., 2021; Chrisman et al., 2012; Memili et al., 2013).
Control variables. In the analysis, we control for additional variables that can be assumed to affect firm performance (e.g., Lee, 2006). First, to capture maturation effects, we control for the age of the business (i.e., the number of years since the date of incorporation). Second, we control for firm size by asking the respondents to indicate how many people, including management, were employed in the firm on average in the past 12 months. Third, we control for generational involvement using a single item in which the respondents indicated the number of generations (one, two, three, or more) currently involved in the firm’s operations (Arzubiaga et al., 2018). Finally, we introduce four binary dummy variables, each representing a broad industry affiliation (manufacturing, construction, wholesale, services).
Reliability and Validity
To assess the quality of the measurement model, we conducted a confirmatory factor analysis (CFA) in Amos. In the first step, we evaluated the overall model fit, and subsequently, to assess the reliability and validity of the latent variables, we observed their composite reliabilities (CR), average variances extracted (AVE), maximum shared variances (MSV), and average shared variances (ASV). The measurement model consists of relational conflict (α = 0.944, CR = 0.945, AVE = 0.863; MSV = 0.396, ASV = 0.246), family social capital (α = 0.924, CR = 0.985, AVE = 0.956; MSV = 0.533, ASV = 0.369), and family firm performance (α = 0.862, CR = 0.862, AVE = 0.677; MSV = 0.218, ASV = 0.146). Overall, the model displays a good fit to the data (χ2 = 333.99, df = 221, CMIN/df = 1.511, CFI = 0.963, RMSEA = 0.054, PCLOSE = 0.266). All constructs display good reliability as the individual CRs are greater than 0.7 (Nunnally, 1978). Moreover, Cronbach’s alpha values are all greater than 0.8, thus indicating a good internal consistency of the measures. Convergent validity is acceptable since all CRs are greater than the AVEs, and AVEs are greater than 0.5 (Hair et al., 2010). The constructs also display a good discriminant validity as both the MSVs and ASVs are less than AVEs (Hair et al., 2010). Further, to support the convergent validity of the outcome variable (firm performance), we use a single item to measure respondent’s overall satisfaction with the firm. Specifically, we employ a single item (“How overall satisfied are you with the business?”) with answers ranging from 1 = Completely dissatisfied to 5 = Extremely satisfied. The correlation between these two variables is positive and significant (r = .595, p < .0001), showing a good convergent validity with respect to different measures of a single construct (Barringer & Bluedorn, 1999). Thus, there are no construct reliability or validity concerns in our study. Because the data were collected in the same period using the same methods, we evaluate the presence of common method bias using Harman’s single factor test (Podsakoff et al., 2003). Exploratory factor analysis does not provide a single-factor solution, suggesting that common method bias is not an issue in our data.
Results
Descriptive statistics along with Pearson correlations between the individual variables are presented in Table 2. Relational conflict is negatively correlated with family social capital (r = –.553, p < .01), firm performance (r = –.171, p < .05), and family ownership (r = –.306, p < .01). Family social capital is also positively correlated with firm performance (r = .375, p < .01). Firm performance is further negatively correlated with family ownership (r = –.174, p < .05), and positively correlated with generational involvement (r = .165, p < .05). Family ownership is negatively correlated with firm age (r = –.187, p < .05), firm size (r = –.251, p < .01), and generational involvement (r = –.156, p < .05). As expected, there is a positive correlation between firm age and firm size (r = .338, p < .01) and firm age and generational involvement (r = .302, p < .01). Table 2.
Table 2 Descriptive statistics and Pearson correlation matrix
M SD Min Max 1 2 3 4 5 6
1. Relational conflict 2.174 1.069 1.000 5.000 1.000
2. Family social capital 4.145 0.655 2.000 5.000 –0.553** 1.000
3. Firm performance 3.798 0.713 1.670 5.000 –0.171* 0.375** 1.000
4. Family ownership 89.926 18.321 25.00 100.00 –0.306** 0.049 –0.174* 1.000
5. Firm age 23.060 19.468 2.00 121.00 0.069 –0.127 0.025 –0.187* 1.000
6. Firm size 35.058 60.660 5.00 499.00 0.048 − 0.086 0.061 –0.251** 0.338** 1.000
7. Generational involvement 2.080 0.791 1.00 4.00 0.094 –0.054 0.165* –0.156* 0.302** 0.141
Note: M = mean, SD = standard deviation
* p < .05, ** p < .01
To test our hypotheses, we use the conditional process analysis tool PROCESS for SPSS (Hayes, 2018). The significance of direct and indirect effects is estimated using 95% bias-corrected bootstrap confidence intervals based on 5000 bootstrap samples. A statistically significant effect at the 0.05 level exists if the confidence interval does not contain the zero value (Hayes, 2018). To reduce structural multicollinearity, all variables defining product terms are centred before the analysis. The regressions use heteroscedasticity-consistent standard errors.
To test hypotheses 1–4, we evaluate a simple mediation model and report the total, direct, and indirect effects (Table 3). Considering firm performance as the outcome variable, we did not observe any significant association between relational conflict and performance (b = –0.013, p = .837). Thus, hypothesis 1 is not supported. However, there is a positive and significant relationship between family social capital and family firm performance (b = 0.391, p < .01), which supports hypothesis 2. Relational conflict is significantly and negatively associated with family social capital (b = –0.379, p < .01), supporting hypothesis 3. In the bottom part of Table 3, we display the estimated total, direct and indirect effects and their 95% confidence intervals. While the total effect is negative and significant, we did not find evidence of a significant direct effect, suggesting that family social capital fully mediates the relationship between relational conflict and family firm performance. Thus, we only find partial support for hypothesis 4 as no evidence of a direct effect of relational conflict on firm performance has been found (see Table 3).
Table 3 Testing Hypothesis 4 - Mediating role of family social capital
Variables / Outcome variable Family social capital Firm performance
Intercept 4.732** (0.249) 1.663** (0.499)
Control variables
Firm age –0.002 (0.002) –0.001 (0.003)
Firm size –0.001 (0.001) 0.001 (0.001)
Generational involvement –0.006 (0.047) 0.163* (0.070)
Manufacturing 0.383* (0.177) 0.201 (0.234)
Construction –0.023 (0.252) 0.060 (0.205)
Wholesale 0.231 (0.189) 0.205 (0.220)
Services 0.394* (0.175) 0.169 (0.187)
Independent variable
Relational conflict –0.379** (0.062) –0.013 (0.064)
Mediator
Family social capital 0.391** (0.099)
R2 0.409 0.202
F-statistics 9.529** 6.394**
Direct and indirect effects of relational conflict on firm performance
Outcome variable Firm performance
b SE LLCI ULCI
Total effect –0.162* 0.066 –0.271 –0.052
Direct effect –0.013 0.064 –0.119 0.093
Indirect effect –0.148* 0.042 –0.212 –0.077
Note: LLCI = lower limit of the 95% confidence interval, ULCI = upper limit of the 95% confidence interval. Standard errors are presented in parentheses (.)
* p < .05, ** p < .01
To test hypothesis 5, we assess a simple moderation model while observing the significance of the interaction term and the significance of the conditional effect of the focal predictor (relational conflict) at three conditioning values of the moderator (family ownership): low (minus one standard deviation from the mean), moderate (mean), and high. Because one standard deviation above the mean is above the maximum observed in the data for family ownership, in all subsequent analyses, the maximum measurement for family ownership (100%) is used for conditioning instead. The results are displayed in Table 4. They suggest the existence of a significant and negative relationship between relational conflict and family social capital (b = –0.398, p < .01). Furthermore, the interaction term is significant and negative (b = –0.006, p < .01), which indicates that family ownership strengthens the negative relationship between relational conflict and family social capital. The moderating effect is illustrated in Fig. 1, which presents the interaction plot. The conditional effects of relational conflict on family social capital are statistically significant for all the three conditioning values of family ownership (low, moderate, and high). Thus, hypothesis 5 is supported.
Table 4 Testing Hypothesis 5 - Moderating role of family ownership
Variables Estimates
Intercept 3.940** (0.193)
Control variables
Firm age –0.002 (0.002)
Firm size –0.000 (0.001)
Generational involvement –0.025 (0.046)
Manufacturing 0.400* (0.178)
Construction 0.025 (0.245)
Wholesale 0.280 (0.185)
Services 0.421* (0.173)
Independent variable
Relational conflict –0.398** (0.065)
Moderator
Family ownership –0.001 (0.002)
Interaction term
Relational conflict × Family ownership –0.006* (0.002)
R2 0.434
F-statistics 9.307**
Conditional effects of relational conflict on family social capital
Conditioning values of family ownership b SE LLCI ULCI
Low –0.317* 0.051 –0.401 –0.232
Medium –0.398* 0.065 –0.505 –0.291
High –0.442* 0.076 –0.568 –0.315
Note: Outcome variable: family social capital. LLCI = lower limit of the 95% confidence interval, ULCI = upper limit of the 95% confidence interval. Standard errors are presented in parentheses (.)
* p < .05, ** p < .01
Fig. 1 Moderating effect of family ownership on the relationship between relational conflict and family social capital: The interaction plot
Hypothesis 6
is evaluated using a moderated mediation model. We observe the significance of the direct effect and conditional indirect effect of relational conflict on performance and the significance of the index of moderated mediation (Hayes, 2015). Table 5 presents the results. Consistent with our previous findings, there is a significant and negative relationship between relational conflict and family social capital (b = –0.398, p < .01). Since the interaction term is significant and negative (b = –0.006, p < .01), this relationship is negatively moderated by family ownership. Moreover, a significant and positive relationship is found between family social capital and family firm performance (b = 0.391, p < .01). Again, we have not found any direct association between relational conflict and family firm performance. On the other hand, there are significant and negative indirect effects when family ownership is low, moderate, and high. Finally, the index of moderated mediation is statistically significant at the 0.05 level. In conclusion, we found evidence of a moderated mediation, as predicted by hypothesis 6. However, relational conflict does not harm family firm performance directly but indirectly by harming family social capital. The results are graphically displayed in Fig. 2.
Table 5 Testing Hypothesis 6 - Moderated mediation
Variables / Outcome variable Family social capital Firm performance
Intercept 3.940** (0.193) 1.636** (0.411)
Control variables
Firm age –0.002 (0.002) –0.001 (0.003)
Firm size –0.000 (0.001) 0.001 (0.001)
Generational involvement –0.025 (0.046) 0.163* (0.070)
Manufacturing 0.400* (0.178) 0.201 (0.234)
Construction 0.025 (0.245) 0.060 (0.205)
Wholesale 0.280 (0.185) 0.205 (0.220)
Services 0.421* (0.173) 0.169 (0.187)
Independent variable
Relational conflict –0.398** (0.065) –0.013 (0.064)
Mediator
Family social capital 0.391** (0.099)
Moderator
Family ownership –0.001 (0.002)
Interaction term
Relational conflict × Family ownership –0.006* (0.002)
R2 0.434 0.202
F-statistics 9.307** 6.394**
Direct and indirect effects
Outcome variable Firm performance
b SE LLCI ULCI
Direct effect –0.013 0.064 –0.119 0.093
Conditional indirect effect when family ownership is:
Low –0.124* 0.035 –0.179 –0.064
Medium –0.156* 0.043 –0.221 –0.081
High –0.173* 0.049 –0.248 –0.087
Index of moderated mediation –0.002* 0.001 –0.004 –0.001
Note: LLCI = lower limit of the 95% confidence interval, ULCI = upper limit of the 95% confidence interval. Standard errors are presented in parentheses (.)
* p < .05, ** p < .01
Fig. 2 Results of the moderated mediation model
Note: The indirect effect is significant at the 0.05 level for all levels of family ownership. * p < .05, ** p < .01
Robustness Check
Since the data is cross-sectional, we tested an alternative causal model, in which the order of the first two variables predicting family firm performance was switched (e.g. Van Gils et al., 2019). The alternative model specification thus considered that the relationship between family social capital and family firm performance relationship is mediated by relational conflict, with family ownership acting as a moderator of the relational conflict-family firm performance relationship. Consistent with our expectations, there is a positive and significant direct effect of family social capital on family firm performance (b = 0.371, p < .01), but the indirect effect is not significant for any level of family ownership (low, medium, and high), thus providing evidence against the alternative causal path and in favour of our hypotheses (Van Gils et al., 2019). The complete results are displayed in Table 6.
Table 6 Robustness check – Alternative causal path, relational conflict as a mediator
Variables / Outcome variable Relational conflict Firm performance
Intercept 3.699** (0.508) 1.727** (0.433)
Control variables
Firm age –0.002 (0.002) –0.001 (0.003)
Firm size 0.001 (0.001) 0.001 (0.001)
Generational involvement –0.024 (0.072) 0.154* (0.070)
Manufacturing 0.039 (0.275) 0.203 (0.224)
Construction 0.009 (0.311) 0.112 (0.199)
Wholesale –0.188 (0.232) 0.231 (0.217)
Services 0.118 (0.230) 0.202 (0.181)
Independent variable
Family social capital –0.893** (0.102) 0.371** (0.106)
Mediator
Relational conflict –0.039 (0.073)
Moderator
Family ownership –0.006 (0.005)
Interaction term
Relational conflict × Family ownership –0.001 (0.003)
R2 0.365 0.216
F-statistics 14.281** 6.292**
Direct and indirect effects
Outcome variable Firm performance
b SE LLCI ULCI
Direct effect 0.371 0.106 0.162 0.581
Conditional indirect effect when family ownership is:
Low 0.024 0.077 –0.130 0.180
Medium 0.035 0.071 –0.077 0.201
High 0.041 0.081 –0.083 0.300
Index of moderated mediation 0.001 0.004 –0.005 0.010
Note: LLCI = lower limit of the 95% confidence interval, ULCI = upper limit of the 95% confidence interval. Standard errors are presented in parentheses (.)
* p < .05, ** p < .01
Discussion
Despite the resilience of family firms, relational conflicts can make them especially fragile (Kubíček & Machek, 2020). Previous family business authors found that relational conflict only has detrimental consequences in family firms (Eddleston & Kellermanns, 2007; Ensley et al., 2007; Nosé et al., 2017). While it is known that relational conflict is harmful, to our knowledge, no study investigated empirically how relational conflict harms firm performance. Previous authors explain their findings using the arguments of the intragroup conflict literature, which assumes that relational conflict is associated with the dissatisfaction of individual team members (Jehn, 1995). Thus, in family firms, relational conflict is believed to reduce goodwill and mutual understanding, induce negative emotions and hostile behaviours, and distract attention from business needs (Eddleston & Kellermanns, 2007). This led us to the expectation that relational conflict has a direct negative effect on firm performance. However, contrary to our expectations, we do not find evidence of any direct effect. This finding is, to some extent, similar to the findings of Hoelscher (2014) who does not find any impact of relational conflict on performance once “family capital” is controlled for1. While there does not seem to be a direct link between relational conflict and firm performance, we show that family social capital, as a unique family-firm-specific variable, enters into the relational conflict-performance relationship as a mediator. Relational conflict harms family social capital, whose lower levels, in turn, deteriorate family firm performance.
We also find that family ownership strengthens the negative relationship between relational conflict and family social capital. This finding can be attributed to an intensified level of negative affect in a situation when family members feel emotionally attached to and identified with the family firm, but relational conflict escalates among family members. In Jehn and Bendersky’s (2003) terminology, family ownership becomes an “exarcebator” of the conflict-outcome relationship, increasing the adverse effects of relational conflict. Intuitively, this finding means that the greater the “family character” of the firm, the worse the negative impact of relational conflict on family social capital becomes. Thus, family firms with high family ownership can be particularly vulnerable to relational conflicts (Eddleston & Kellermanns, 2007). It should be noted that we do not find that relational conflict emerges more easily in family firms than in other types of organizations. However, we do argue that the exact extent of relational conflict, ceteris paribus, does more harm in family firms that display high family ownership.
Our study has theoretical implications for the family business but also the intragroup conflict literature. First, we demonstrate that relational conflict does not harm firm performance directly. Instead, the adverse effects of relational conflict on family firm performance can all be attributed to the erosion of family social capital. This does not, however, contradict previous findings (e.g. Jehn, 1995). For instance, while relational conflict is believed to reduce mutual understanding, it, in fact, deteriorates the relational dimension of social capital. Likewise, when relational conflict is said to shift attention from collective goals, it actually harms the cognitive dimension of social capital. From a broader perspective, the intragroup conflict literature could be enriched by considering that the adverse effects of relational conflict can be attributed to the damage to social capital shared by social group members. Likewise, while the conflict literature finds that trust mediates the conflict-performance relationship (Lau & Cobb, 2010), we provide a broader picture, showing that not only trust, but generally, social capital, can act as this mediator. Overall, we believe that social capital can be used as an overarching concept for studying the adverse effects of dysfunctional conflicts in family firms and other types of organizations or social groups, presenting a contribution beyond family business research.
From the viewpoint of family business literature, there are two major theoretical implications. First, we show that the damage to family social capital, a resource that is unique to family firms, is the primary reason why dysfunctional conflicts are harmful in family firms. Thus, we are among the few authors to investigate the family outcomes of relational conflicts in family firms (Kubíček & Machek, 2020). Second, our findings suggest that family ownership, or more broadly, emotional attachment to the firm and family-firm identification, determines the damage caused by relational conflict. From this point of view, emotional attachment and identification can represent not only endowment but also a burden (Kellermanns et al., 2012); on the one hand, they can help prevent the development of relational conflict, but they can also intensify its negative consequences when the conflict escalates.
For sure, this study is not free of limitations. First, we used an online panel to collect data. We followed recent management studies’ findings that online panels provide results comparable to conventional data collection methods (Porter et al., 2019; Walter et al., 2019). Moreover, we did our best to reduce potential bias and were very selective in admitting observations to the research sample. Second, while we observed enough variance to test our hypotheses, like some other previous authors (e.g., Carr et al., 2011; Eddleston & Kellermanns, 2007; Memili et al., 2013; Merchant et al., 2018; Rousseau et al., 2018; Ruiz Jiménez et al., 2013), we employed a convenience sample of family firms. Although some descriptive statistics of our sample display a reasonable match with previous family business surveys (Table 1), we cannot guarantee that the sample is fully representative of the U.S. population of family firms, whose characteristics are not precisely known. In general, data collection in family business research is constrained by the lack of national databases of family firms (Chang et al., 2008; Lussier & Sonfield, 2009). Additionally, the existing family business studies diverge in their sample structures, making it difficult to compare with prior research. Microenterprises, which represent the vast majority of U.S. family firms, can display different natures of interpersonal conflict and different forms of family social capital than larger firms. While we believe that the theoretical justification of our hypotheses remains valid for all types of businesses, applying our findings to microenterprises should be done with caution. Hence, our study offers implications especially for small and medium-sized firms with more than five employees. Third, while the respondents are family managers, we lack information on their individual ownership stake. Nevertheless, we assume that family managers are well-informed about the company ownership structure and naturally tend to hold an ownership stake (Rousseau et al., 2018), making them qualified enough to reliably answer the survey questions. Fourth, our study is cross-sectional. Thus, whenever possible, we avoid referring to causality and emphasize the existence of directed associations/relationships between variables. Further, we only consider family ownership as the moderator, making it difficult to comment on the extent to which a family firm follows family-centred goals. A fine-grained measure of socioemotional wealth (e.g., Berrone et al., 2012) could provide additional insights into how relational conflict harms family firm performance. Finally, our analysis does not control for the effects of the Covid-19 pandemics. Social distancing and virtual interactions have caused disruptions in social relationships, possibly affecting family firms’ social capital. Likewise, the strain on family members’ physical and psychological health could exacerbate the existing tensions, contributing to existing conflicts among family members and creating new ones (De Massis & Rondi, 2020). Since not all industries have been equally affected by the pandemics, its adverse performance effects are, at least partially, reflected in the effects of industry control variables. Nevertheless, the effects of Covid-19 on social relationships among family members are not considered in the analysis.
Conclusion
This paper hypothesizes and finds supportive evidence for a moderated mediation model in which relational conflict harms firm performance indirectly through the deterioration of family social capital. We also show that family ownership strengthens the adverse effects of relational conflict. Contrary to our expectations, we do not find that relational conflict harms family firm performance directly.
The findings have several practical implications. Family firm owners, especially those holding a significant portion of ownership shares, should be aware of the potential harm that relational conflict can do to interpersonal relationships. In turn, poor relationships can severely harm the family firm. From a practical standpoint, solving this problem in family businesses is difficult. For fear of opening the door to more damaging issues that lie under the surface (Kaye, 1991), some family members tend to postpone conflict resolution, often indefinitely. In family firms, sacrifices are made to preserve even conflicted relationships in the hope they will repair over time, sometimes leading to situations when it becomes unhealthy for the family to remain in the business together (Kaye, 1996). Timely prevention and, if necessary, conflict resolution can help reduce the adverse effects of relational conflict (Alvarado-Alvarez et al., 2020; Caputo et al., 2018) on family social capital in family firms. The emphasis on cooperation, which helps family members understand deeply held beliefs that can be clarified and shared only through intense interaction over time, offers one method to counteract these impacts. Collaboration helps formulate shared norms, which serve as a compass when a family business is threatened and contribute to the development of family social capital (Sorenson et al., 2009). In general, to mitigate the adverse effects of relational conflict, family firms should not hesitate to address nonfamily members as they are often able to diffuse family tensions (Rosecká & Machek, 2021).
Our analysis also shows that there is a need for further research. First, while family social capital is traditionally considered an internal type of social capital (Arregle et al., 2007), it would be worthy to investigate how relational conflict harms the bridging social capital (or “weak ties”) of family firms. Second, psychology literature acknowledges that conflicts can also have positive effects, and future studies could investigate how cognitive conflicts (i.e. task and process conflicts) affect firm performance. While the family business literature already presents findings related to the performance effects of cognitive conflicts in family firms (Eddleston & Kellermanns, 2007), the role of family social capital in this relationship remains unexplored. Another major avenue for future research is the moderating role of family involvement and family essence. Our study only considered family ownership, representing a source of power and family firms’ ability to pursue family-centred goals. However, of equal importance is the willingness to follow these goals (Evert et al., 2018). While family ownership is considered to be associated with family-firm identification (Kotlar et al., 2020), this only represents one of many dimensions of socioemotional wealth (Berrone et al., 2012). Thus, future research could investigate the role of the individual facets of socioemotional wealth in conflict-outcome relationships in family firms, including whether socioemotional wealth represents an endowment or a burden (Kellermanns et al., 2012) when it comes to conflicts in family firms. Finally, family ownership forms part of the organizational context. Showing that family ownership matters, we believe that future research should explore more in depth the broader role of contextual factors that attenuate or exacerbate conflicts within family firms. Those factors can include, for instance, the succession process stage, the presence of nonfamily CEOs, family climate, or cultural norms and rules (Kubíček & Machek, 2020).
Appendix Questionnaire items
Variables Range Source
Relational conflict I strongly disagree (1) - I strongly agree (5) Paskewitz & Beck (2017)
There is much relationship conflict among family members in our firm.
Family members often get angry with each other while working in our firm.
There is much emotional conflict among family members in our firm.
There is much personal animosity among family members in our firm.
Family social capital I strongly disagree (1) - I strongly agree (5) Carr et al. (2011)
Family members who work in this firm engage in honest communication with one another.
Family members who work in this firm have no hidden agendas.
Family members who work in this firm willingly share information with one another.
Family members who work in this firm take advantage of their family relationships to share information.
Family members who work in this firm have confidence in one another.
Family members who work in this firm show a great deal of integrity with each other.
Overall, family members who work in this firm trust each other.
Family members who work in this firm are usually considerate of each other’s feelings.
Family members who work in this firm are committed to the goals of this firm.
There is a common purpose shared among family members who work in this firm.
Family members who work in this firm view themselves as partners in charting the firm’s direction.
Family members who work in this firm share the same vision for the future of this firm’s direction.
Family firm performance Completely dissatisfied (1) - Extremely satisfied (5) Cooper & Artz (1995)
Relative to your rivals, how satisfied are you with your current performance in terms of net profit growth?
Relative to your rivals, how satisfied are you with your current performance in terms of market share?
Relative to your rivals, how satisfied are you with your current performance in terms of sales?
Acknowledgements
We appreciate the funding support received from the Czech Science Foundation for the project entitled “Intrafamily Conflicts in Family Firms: Antecedents, Effects and Moderators” (registration no.: GA20-04262S).
Declarations
Confirmation
With the submission of this manuscript we would like to undertake that all authors agreed to the submission and that the manuscript is not published, in press, or submitted elsewhere. We confirm that all the research meets the ethical guidelines. We have seen, read, and understood your guidelines on copyright.
1 In Hoelscher’s (2014) study, “family capital” is understood as information channels, moral infrastructure, identity, and family expectations.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Adams RE Laursen B The correlates of conflict: disagreement is not necessarily detrimental Journal of Family Psychology 2007 21 3 445 458 10.1037/0893-3200.21.3.445 17874930
Adler PS Kwon SW Social capital: prospects for a new concept Academy of Management Review 2002 27 1 17 40 10.5465/amr.2002.5922314
Adithipyangkul P Hung HY Leung TY An auditor’s perspective of executive incentive pay and dividend payouts in Family Firms Journal of Family and Economic Issues 2021 42 4 697 714 10.1007/s10834-020-09729-1
Aldridge D Family interaction and suicidal behaviour: a brief review Journal of Family Therapy 1984 6 3 309 322 10.1046/j.1467-6427.1984.00652.x
Alvarado-Alvarez C Armadans I Parada MJ Tracing the roots of constructive conflict management in family firms Negotiation and Conflict Management Research 2020 13 2 105 126 10.1111/ncmr.12164
Aragón-Amonarriz C Arredondo AM Iturrioz-Landart C How can responsible family ownership be sustained across generations? A Family Social Capital Approach Journal of Business Ethics 2019 159 2 161 185 10.1007/s10551-017-3728-7
Arregle JL Hitt MA Sirmon DG Very P The development of organizational social capital: attributes of family firms Journal of Management Studies 2007 44 1 73 95 10.1111/j.1467-6486.2007.00665.x
Arzubiaga U Iturralde T Maseda A Kotlar J Entrepreneurial orientation and firm performance in family SMEs: the moderating effects of family, women, and strategic involvement in the board of directors International Entrepreneurship and Management Journal 2018 14 1 217 244 10.1007/s11365-017-0473-4
Barach JA Ganitsky JB Successful succession in family business Family Business Review 1995 8 2 131 155 10.1111/j.1741-6248.1995.00131.x
Barki H Hartwick J Conceptualizing the construct of interpersonal conflict International Journal of Conflict Management 2004 15 3 216 244 10.1108/eb022913
Barringer BR Bluedorn AC The relationship between corporate entrepreneurship and strategic management Strategic Management Journal 1999 20 5 421 444 10.1002/(SICI)1097-0266(199905)20:5<421::AID-SMJ30>3.0.CO;2-O
Berrone P Cruz C Gomez-Mejia LR Larraza-Kintana M Socioemotional wealth and corporate responses to institutional pressures: do family-controlled firms pollute less? Administrative Science Quarterly 2010 55 1 82 113 10.2189/asqu.2010.55.1.82
Berrone P Cruz C Gomez-Mejia LR Socioemotional wealth in family firms: theoretical dimensions, assessment approaches, and agenda for future research Family Business Review 2012 25 3 258 279 10.1177/0894486511435355
Buehler C Gerard JM Marital conflict, ineffective parenting, and children’s and adolescents’ maladjustment Journal of Marriage and Family 2002 64 1 78 92 10.1111/j.1741-3737.2002.00078.x
Buehler C Franck KL Cook EC Adolescents’ triangulation in marital conflict and peer relations Journal of Research on Adolescence 2009 19 4 669 689 10.1111/j.1532-7795.2009.00616.x
Caputo A Marzi G Pellegrini MM Rialti R Conflict management in family businesses: a bibliometric analysis and systematic literature review International Journal of Conflict Management 2018 29 4 519 542 10.1108/IJCMA-02-2018-0027
Carnevale PJ Probst TM Social values and social conflict in creative problem solving and categorization Journal of Personality and Social Psychology 1998 74 1300 1309 10.1037/0022-3514.74.5.1300
Carr JC Cole MS Ring JK Blettner DP A measure of variations in internal social capital among family firms Entrepreneurship Theory and Practice 2011 35 6 1207 1227 10.1111/j.1540-6520.2011.00499.x
Chang EP Chrisman JJ Chua JH Kellermanns FW Regional economy as a determinant of the prevalence of family firms in the United States: a preliminary report Entrepreneurship Theory and Practice 2008 32 3 559 573 10.1111/j.1540-6520.2008.00241.x
Chirico F Salvato C Knowledge internalization and product development in Family Firms: when Relational and affective factors Matter Entrepreneurship Theory and Practice 2016 40 1 201 229 10.1111/etap.12114
Chrisman JJ Chua JH Pearson AW Barnett T Family involvement, family influence, and family–centered non–economic goals in small firms Entrepreneurship Theory and Practice 2012 36 2 267 293 10.1111/j.1540-6520.2010.00407.x
Coleman JS Social capital in the creation of human capital American Journal of Sociology 1988 94 S95 S120 10.1086/228943
Cooper, C. (1988). Commentary: The role of conflict in adolescent parent relationships. In M. Gunnar (Ed.), 21st Minnesota Symposium on Child Psychology (pp. 181–197). Lawrence Erlbaum Associates.
Cooper AC Artz KW Determinants of satisfaction for entrepreneurs Journal of Business Venturing 1995 10 6 439 457 10.1016/0883-9026(95)00083-K
Daspit JJ Chrisman JJ Ashton T Evangelopoulos N Family firm heterogeneity: a definition, common themes, Scholarly Progress, and directions Forward Family Business Review 2021 34 3 296 322 10.1177/08944865211008350
De Dreu CK Weingart LR Task versus relationship conflict, team performance, and team member satisfaction: a meta-analysis Journal of Applied Psychology 2003 88 4 741 749 10.1037/0021-9010.88.4.741 12940412
De Massis AV Rondi E COVID-19 and the future of family business research Journal of Management Studies 2020 57 8 1727 1731 10.1111/joms.12632
Derfler-Rozin R Sherf EN Chen G To be or not to be consistent? The role of friendship and group-targeted perspective in managers’ allocation decisions Journal of Organizational Behavior 2021 42 6 814 833 10.1002/job.2490
Duarte Alonso A Kok S O’Brien S Understanding the impact of family firms through social capital theory: a south american perspective Journal of Family and Economic Issues 2020 41 4 749 761 10.1007/s10834-020-09669-w
Eddleston KA Kellermanns FW Destructive and productive family relationships: a stewardship theory perspective Journal of Business Venturing 2007 22 4 545 565 10.1016/j.jbusvent.2006.06.004
Ensley MD Pearson AW Sardeshmukh SR The negative consequences of pay dispersion in family and non-family top management teams: an exploratory analysis of new venture, high-growth firms Journal of Business Research 2007 60 10 1039 1047 10.1016/j.jbusres.2006.12.012
Evert RE Sears JB Martin JA Payne GT Family ownership and family involvement as antecedents of strategic action: a longitudinal study of initial international entry Journal of Business Research 2018 84 301 311 10.1016/j.jbusres.2017.07.019
Frank H Kessler A Nosé L Suchy D Conflicts in family firms: state of the art and perspectives for future research Journal of Family Business Management 2011 1 2 130 153 10.1108/20436231111167219
Fulmer CA Gelfand MJ At what level (and in whom) we trust: trust across multiple organizational levels Journal of Management 2012 38 4 1167 1230 10.1177/0149206312439327
Gargiulo, M., & Benassi, M. (1999). The dark side of social capital. In S. Gabbay, & R. Leenders (Eds.), Social capital and liability (pp. 298–322). Kluwer.
Gedajlovic E Carney M Chrisman JJ Kellermanns FW The adolescence of family firm research: taking stock and planning for the future Journal of Management 2012 38 4 1010 1037 10.1177/0149206311429990
Gómez-Mejía LR Cruz C Berrone P De Castro J The bind that ties: socioemotional wealth preservation in family firms Academy of Management Annals 2011 5 1 653 707 10.5465/19416520.2011.593320
Gómez-Mejía LR Haynes KT Núñez-Nickel M Jacobson KJ Moyano-Fuentes J Socioemotional wealth and business risks in family-controlled firms: evidence from spanish olive oil mills Administrative Science Quarterly 2007 52 1 106 137 10.2189/asqu.52.1.106
Hair, J., Black, W., Babin, B., & Anderson, R. (2010). Multivariate Data Analysis (7th Ed.) Prentice Hall Inc.
Hayes AF An index and test of linear moderated mediation Multivariate Behavioral Research 2015 50 1 1 22 10.1080/00273171.2014.962683 26609740
Hayes, A. F. (2018). Introduction to Mediation, Moderation, and conditional process analysis: a regression-based Approach. Second Edition. Guilford Press.
Haynes KT Hitt MA Campbell JT The dark side of leadership: towards a mid-range theory of hubris and greed in entrepreneurial contexts Journal of Management Studies 2015 52 4 479 505 10.1111/joms.12127
Haynes G Marshall M Lee Y Zuiker V Jasper CR Sydnor S Valdivia C Masuo D Niehm L Wiatt R Family business research: reviewing the past, contemplating the future Journal of Family and Economic Issues 2021 42 1 70 83 10.1007/s10834-020-09732-6
Herrero I How familial is family Social Capital? Analyzing Bonding Social Capital in Family and Nonfamily Firms Family Business Review 2018 31 4 441 459 10.1177/0894486518784475
Herrero I Hughes M When family social capital is too much of a good thing Journal of Family Business Strategy 2019 10 3 100271 10.1016/j.jfbs.2019.01.001
Hoelscher ML Does family capital outweigh the negative effects of conflict on firm performance? Journal of Family Business Management 2014 4 1 46 61 10.1108/JFBM-03-2013-0009
Hoffman J Hoelscher M Sorenson R Achieving sustained competitive advantage: a family capital theory Family Business Review 2006 19 2 135 145 10.1111/j.1741-6248.2006.00065.x
Jackson SA Gopalakrishna-Remani V Mishra R Napier R Examining the impact of design for environment and the mediating effect of quality management innovation on firm performance International Journal of Production Economics 2016 173 142 152 10.1016/j.ijpe.2015.12.009
Jehn KA A multimethod examination of the benefits and detriments of intragroup conflict Administrative Science Quarterly 1995 40 2 256 282 10.2307/2393638
Jehn KA Bendersky C Intragroup conflict in organizations: a contingency perspective on the conflict-outcome relationship Research in Organizational Behavior 2003 25 187 242 10.1016/S0191-3085(03)25005-X
Jehn KA Mannix EA The dynamic nature of conflict: a longitudinal study of intragroup conflict and group performance Academy of Management Journal 2001 44 2 238 251 10.5465/3069453
Kansikas J Murphy L Bonding family social capital and firm performance International Journal of Entrepreneurship and Small Business 2011 14 4 533 550 10.1504/IJESB.2011.043474
Kaye K Penetrating the cycle of sustained conflict Family Business Review 1991 4 1 21 44 10.1111/j.1741-6248.1991.00021.x
Kaye K When the family business is a sickness Family Business Review 1996 9 4 347 368 10.1111/j.1741-6248.1996.00347.x
Kellermanns FW Eddleston KA Zellweger TM Article commentary: extending the socioemotional wealth perspective: a look at the dark side Entrepreneurship Theory and Practice 2012 36 6 1175 1182 10.1111/j.1540-6520.2012.00544.x
Kessler T Hollbach S Group-based emotions as determinants of ingroup identification Journal of Experimental Social Psychology 2005 41 6 677 685 10.1016/j.jesp.2005.01.001
Knight D Pearce CL Smith KG Olian JD Sims HP Smith KA Flood P Top management team diversity, group process, and strategic consensus Strategic Management Journal 1999 20 5 445 465 10.1002/(SICI)1097-0266(199905)20:5<445::AID-SMJ27>3.0.CO;2-V
Kotlar J De Massis A Frattini F Kammerlander N Motivation gaps and implementation traps: the paradoxical and time-varying effects of family ownership on firm absorptive capacity Journal of Product Innovation Management 2020 37 1 2 25 10.1111/jpim.12503
Kubíček A Machek O Intrafamily conflicts in family businesses: a systematic review of the literature and agenda for future research Family Business Review 2020 33 2 194 227 10.1177/0894486519899573
Lau RS Cobb AT Understanding the connections between relationship conflict and performance: the intervening roles of trust and exchange Journal of Organizational Behavior 2010 31 6 898 917 10.1002/job.674
Leana CR Van Buren HJ Organizational social capital and employment practices Academy of Management Review 1999 24 3 538 555 10.5465/amr.1999.2202136
Lee J Family firm performance: further evidence Family Business Review 2006 19 2 103 114 10.1111/j.1741-6248.2006.00060.x
Levinson H Conflicts that plague family businesses Harvard Business Review 1971 49 90 98
Lussier RN Sonfield MC Founder influence in family business: analyzing combined data from six countries Journal of Small Business Strategy 2009 20 1 103 118
Lu SC Kong DT Ferrin DL Dirks KT What are the determinants of interpersonal trust in dyadic negotiations? Meta-analytic evidence and implications for future research Journal of Trust Research 2017 7 1 22 50 10.1080/21515581.2017.1285241
Martin G Gómez-Mejía L The relationship between socioemotional and financial wealth: re-visiting family firm decision making Management Research 2016 14 3 215 233 10.1108/MRJIAM-02-2016-0638
Mayer RC Davis JH The effect of the performance appraisal system on trust for management: a field quasi-experiment Journal of Applied Psychology 1999 84 1 123 136 10.1037/0021-9010.84.1.123
McCann G Hammond C Keyt A Schrank H Fujiuchi K A view from afar: rethinking the director’s role in university-based family business programs Family Business Review 2004 17 3 203 219 10.1111/j.1741-6248.2004.00014.x
Memili E Zellweger TM Fang HC The determinants of family owner-managers’ affective organizational commitment Family Relations 2013 62 3 443 456 10.1111/fare.12015
Merchant P Kumar A Mallik D Factors influencing family business continuity in indian small and medium enterprises (SMEs) Journal of Family and Economic Issues 2018 39 2 177 190 10.1007/s10834-017-9562-3
Mustakallio M Autio E Zahra SA Relational and contractual governance in Family Firms: Effects on Strategic decision making Family Business Review 2002 15 3 205 222 10.1111/j.1741-6248.2002.00205.x
Nahapiet J Ghoshal S Social capital, intellectual capital, and the organizational advantage Academy of Management Review 1998 23 2 242 266 10.2307/259373
Noller, P., Atkin, S., Feeney, J. A., & Peterson, C. (2006). Family conflict and adolescents. In L. H. Turner, & R. West (Eds.), The family communication sourcebook (pp. 165–182). Sage Publications.
Nosé L Korunka C Frank H Danes SM Decreasing the effects of relationship conflict on family businesses: the moderating role of family climate Journal of Family Issues 2017 38 1 25 51 10.1177/0192513X15573869
Nunnally, J. C. (1978). Psychometric theory. McGraw Hill.
Paskewitz EA Beck SJ When work and family merge: understanding intragroup conflict experiences in family farm businesses Journal of Family Communication 2017 17 4 386 400 10.1080/15267431.2017.1363757
Pearson AW Carr JC Shaw JC Toward a theory of familiness: a social capital perspective Entrepreneurship Theory and Practice 2008 32 6 949 969 10.1111/j.1540-6520.2008.00265.x
Pelled LH Xin KR Down and out: an investigation of the relationship between mood and employee withdrawal behavior Journal of Management 1999 25 6 875 895 10.1177/014920639902500605
Podsakoff PM MacKenzie SB Lee JY Podsakoff NP Common method biases in behavioral research: a critical review of the literature and recommended remedies Journal of Applied Psychology 2003 88 5 879 903 10.1037/0021-9010.88.5.879 14516251
Porter CO Outlaw R Gale JP Cho TS The use of online panel data in management research: a review and recommendations Journal of Management 2019 45 1 319 344 10.1177/0149206318811569
Puryear AN Rogoff EG Lee MS Heck RK Grossman EB Haynes GW Onochie J Sampling minority business owners and their families: the understudied entrepreneurial experience Journal of Small Business Management 2008 46 3 422 455 10.1111/j.1540-627X.2008.00251.x
Putnam, R. (2000). Bowling alone: the collapse and revival of american community. Simon & Schuster.
Rosecká N Machek O Non-family members and conflict processes in family firms: a systematic review of literature Journal of Business Economics 2021 92 235 281 10.1007/s11573-021-01054-9
Rousseau MB Kellermanns F Zellweger T Beck TE Relationship conflict, family name congruence, and socioemotional wealth in family firms Family Business Review 2018 31 4 397 416 10.1177/0894486518790425
Ruiz Jiménez MC Martos V Martínez Jiménez R Organisational Harmony as a value in Family businesses and its influence on performance Journal of Business Ethics 2013 126 2 259 272 10.1007/s10551-013-1941-6
Sanchez-Famoso V Pittino D Chirico F Maseda A Iturralde T Social capital and innovation in family firms: the moderating roles of family control and generational involvement Scandinavian Journal of Management 2019 35 3 1 13 10.1016/j.scaman.2019.02.002
Shi HX Shepherd DM Schmidts T Social capital in entrepreneurial family businesses: the role of trust International Journal of Entrepreneurial Behavior & Research 2015 21 6 814 841 10.1108/IJEBR-04-2015-0090
Sorenson RL Goodpaster KE Hedberg PR Yu A The family point of view, family social capital, and firm performance: an exploratory test Family Business Review 2009 22 3 239 253 10.1177/0894486509332456
Stam W Arzlanian S Elfring T Social capital of entrepreneurs and small firm performance: a meta-analysis of contextual and methodological moderators Journal of Business Venturing 2014 29 1 152 173 10.1016/j.jbusvent.2013.01.002
Stewart A Hitt MA Why can’t a family business be more like a nonfamily business? Modes of professionalization in family firms Family Business Review 2012 25 1 58 86 10.1177/0894486511421665
Taylor M Segrin C Perceptions of parental gender roles and conflict styles and their association with young adults’ relational and psychological well-being Communication Research Reports 2010 27 3 230 242 10.1080/08824096.2010.496326
Van Gils A Huybrechts J Minola T Cassia L Unraveling the impact of family antecedents on family firm image: a serial multiple-mediation model Journal of Family Business Strategy 2019 10 1 17 27 10.1016/j.jfbs.2019.02.001
Vandell, D. L., & Bailey, M. D. (1992). Conflicts between siblings. In C. U. Shantz, & W. W. Hartup (Eds.), Conflict in child and adolescent development (pp. 242–269). Cambridge University Press.
Walter SL Seibert SE Goering D O’Boyle EH A tale of two sample sources: do results from online panel data and conventional data converge? Journal of Business and Psychology 2019 34 4 425 452 10.1007/s10869-018-9552-y
Wiatt RD Lee YG Marshall MI Zuiker VS The Effect of Cash Flow problems and resource intermingling on Small Business Recovery and Resilience after a natural disaster Journal of Family and Economic Issues 2020 42 203 214 10.1007/s10834-020-09710-y 32958989
Williams M In whom we trust: Group membership as an affective context for trust development Academy of Management Review 2001 26 3 377 396 10.5465/amr.2001.4845794
Winter, N., Burleigh, T., Kennedy, R., & Clifford, S. (2019). A simplified protocol to screen out VPS and international respondents using Qualtrics. Available at SSRN 3327274.
Zahra SA International expansion of US manufacturing family businesses: the effect of ownership and involvement Journal of Business Venturing 2003 18 4 495 512 10.1016/S0883-9026(03)00057-0
Zellweger TM Astrachan JH On the emotional value of owning a firm Family Business Review 2008 21 4 347 363 10.1177/08944865080210040106
| 0 | PMC9739348 | NO-CC CODE | 2022-12-14 23:29:54 | no | J Fam Econ Issues. 2022 Dec 10;:1-16 | utf-8 | J Fam Econ Issues | 2,022 | 10.1007/s10834-022-09877-6 | oa_other |
==== Front
J Relig Health
J Relig Health
Journal of Religion and Health
0022-4197
1573-6571
Springer US New York
1714
10.1007/s10943-022-01714-2
Original Paper
The Effect of Praying on Endogenous Pain Modulation and Pain Intensity in Healthy Religious Individuals in Lebanon: A Randomized Controlled Trial
http://orcid.org/0000-0003-2639-6560
Najem Charbel [email protected]
http://www.paininmotion.be/
123
Meeus Mira http://www.paininmotion.be/
134
Cagnie Barbara 1
Ayoubi Farah 25
Al Achek Mikel 2
Van Wilgen Paul http://www.paininmotion.be/
378
Van Oosterwijck Jessica http://www.paininmotion.be/
1346
De Meulemeester Kayleigh http://www.paininmotion.be/
13
1 grid.5342.0 0000 0001 2069 7798 Spine, Head and Pain Research Unit Ghent, Department of Rehabilitation Sciences, Faculty of Medicine and Health Sciences, Ghent University, Ghent, Belgium
2 grid.444431.2 0000 0001 2218 8962 Department of Physiotherapy, Faculty of Public Health, Antonine University, Hadat, Lebanon
3 grid.512583.8 Pain in Motion International Research Group, Ghent, Belgium
4 grid.5284.b 0000 0001 0790 3681 MOVANT Research Group, Department of Rehabilitation Sciences and Physiotherapy, Faculty of Medicine and Health Sciences, University of Antwerp, Antwerp, Belgium
5 grid.411324.1 0000 0001 2324 3572 Department of Physiotherapy, Faculty of Public Health, Lebanese University, Beirut, Lebanon
6 grid.434261.6 0000 0000 8597 7208 Research Foundation-Flanders (FWO), Brussels, Belgium
7 grid.491510.8 Transcare Transdisciplinary Pain Management Center, Groningen, The Netherlands
8 grid.8767.e 0000 0001 2290 8069 PAIN–VUB Pain in Motion Research Group, Department of Physiotherapy, Human Physiology and Anatomy, Faculty of Physical Education & Physiotherapy, Vrije Universiteit Brussel, Brussels, Belgium
10 12 2022
124
29 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.
Prayer is considered to be the most common therapy used in alternative medicine. This study aimed to explore the effect of prayers on endogenous pain modulation, pain intensity, and sensitivity in healthy religious participants. A total of 208 healthy religious participants were enrolled in this study and randomly distributed into two groups, a prayer group (n = 156) and a poem reading or control group (n = 52). Participants from the prayer group were then selectively allocated using the prayer function scale to either an active prayer group (n = 94) receiving an active type of praying or to a passive prayer group (n = 62) receiving a passive type of praying. Pain assessments were performed before and following the interventions and included pressure pain threshold assessment (PPT), conditioned pain modulation (CPM), and a numerical pain rating scale. A significant group-by-time interaction for PPT (p = 0.014) indicated post-intervention increases in PPT in the prayer group but not in the poem reading control group. Participants experienced a decrease in CPM efficacy (p = 0.030) and a reduction in their NPRS (p < 0.001) following the interventions, independent of their group allocation. The results showed that prayer, irrespective of the type, can positively affect pain sensitivity and intensity, but does not influence endogenous pain inhibition during hot water immersion. Future research should focus on understanding the mechanism behind “prayer-induced analgesia.”
Keywords
Pressure pain threshold
Conditioned pain modulation
Prayer
Religion
Pain
==== Body
pmcIntroduction
Religion can be defined as a “Sentiment of learned behaviors and social expressions that reflect cultural values”(White et al., 2011). Prayers, religious activities, and seeking spiritual guidance all refer to religion (Tzeng & Yin, 2008; Wachholtz et al., 2007). Religion is also defined as a belief system, a connection with the divine being, a relationship with the supernatural, and a philosophy (Narayanasamy, 2004). Religiosity can be divided into three major dimensions (Levin et al., 1995). The first dimension comprises organizational religious activity (ORA), which reflects the social dimension of religiousness and includes attending church, synagogue, and taking part in prayer or Bible study groups. The second dimension of religiosity is the non-organizational religious activity (NORA), and it comprises more private and personal religious behaviors such as prayer, meditation, reading the Bible, or other religious literature. The final and third dimension is subjective or intrinsic religiosity (IR), and it reflects the extent to which religion is the primary motivating factor in people’s lives and how it influences decision-making and behavior (Koenig et al., 2004).
Prayer is considered the most common alternative medicine therapy (South & McDowell, 2018; Tippens et al., 2009). In pain management, traditional strategies do not always ease pain or improve quality of life, leading to alternative pain relief approaches (Breivik et al., 2006). Previous studies showed that prayer for self (43%) and prayer for others (24.4%) as being two of the most used alternative medicine practices in the USA (Barnes et al., 2004) and that the inclusion of prayer in the definition of alternative and complementary medicine resulted in a significant increase in its usage (Robles et al., 2017). Recently, researchers showed interest in understanding the role of spirituality on pain experience (Ferreira-Valente et al., 2019; Illueca & Doolittle, 2020; O’Beirne et al., 2020) based on the need for a model that incorporates spirituality in the biopsychosocial frame of pain (Wachholtz A.B. et al., 2007). However, many of these studies identify prayer as a coping mechanism and do not focus on the therapeutic effect of prayer in pain management. In addition, one can also distinguish different forms of prayer (Laird et al., 2004): adoration, confession, thanksgiving, reception, and supplication defined also as petitionary prayer (Poloma & Pendleton, 1991).
The current study focuses on supplication or petitionary prayer, which is a specific request for (a) oneself or (b) others (Jors et al., 2015). The praying ritual is structured as follows: a motive to pray (a problem), an action to perform (ask something), and an effect to be sought (the solution to the problem). Depending on the individual’s relationship with God, we can distinguish 3 methods of problem-solving or 3 methods of praying to address a problem (K. Pargament & Mahoney, 2005; Pargament et al., 1988). In the first type, known as “self-directing,” the individual is very active, and God is passive, giving people the freedom and resources to direct their own life. The second type describes a style in which the individual takes no active steps and passively waits for God to solve the problem known as “deferring.” The third type describes a pattern of coping in which the individual and God both take active roles, in partnership with each other, to solve a problem known as “collaborative” (Pargament et al., 1988). While the deferring type represents a passive type of prayer and coping, the collaborative and self-directing types represent a more active type of prayer and coping.
Biologically, there are multiple potential pathways through which prayer may affect pain modulation (Seybold, 2007). Spiritual/religious activities are associated with an increase in serotonin levels (Mohandas, 2008). This raises the possibility that the serotonin system, which plays an important role in endogenous pain modulation through the facilitatory and inhibitory pathways, serves as a biological basis for spiritual experiences (Borg et al., 2003). Besides, prayer and other religious practices such as meditation activate various brain regions, including the medial prefrontal cortex (mPFC) and posterior cingulate (Neubauer, 2014). The mPFC is important for pain processing and its involvement in the modulation of pain catastrophizing (Seminowicz & Davis, 2006), reduction of pain-induced sympathetic activity (Perlaki et al., 2015), and decrease in facial expressions of pain (Karmann et al., 2016). Previous studies (Jegindø et al., 2013; Wiech et al., 2008) have demonstrated that religious participants perceived painful stimulation as less intense after prayer or after meditating over religious images. In addition, an active style of prayer and in contrast to passive prayer is associated with greater pain tolerance for participants with religious beliefs undergoing an experimental painful procedure (Meints et al., 2018). However, none of these studies investigated the effect of prayer on endogenous analgesia.
More research should reveal if prayers affect endogenous pain modulation. Conditioned pain modulation (CPM) has recently been coined for the psychophysical protocols that assess the functioning of descending pain inhibitory pathways in humans and could thus assess the effect of prayer on endogenous pain modulation. Besides, pressure pain threshold (PPT) assessments are a way of quantifying the sensitivity of deep structures to mechanical pain (Balaguier et al., 2016b). PPT provides a quantitative value related to deep structures sensitivity, allowing researchers to make comparisons over time (Balaguier et al., 2016a), and could be used to evaluate the effect of prayer on pain sensitivity.
The primary purpose of this study was to explore the effect of petitionary praying on endogenous pain modulation. It was hypothesized that prayer would increase PPTs, CPM efficacy, and reduce pain intensity during painful hot water immersion compared to a no-prayer control group in a healthy religious population. The secondary purpose of this study was to investigate the effect of different types of praying on pain outcomes since the style of praying has been shown to affect health outcomes in different ways. For instance, active prayers are associated with better mental health outcomes than passive prayers (Bade & Cook, 2008; Tait et al., 2016). Therefore, it was hypothesized that participants engaging in active prayer would show greater improvements in pain outcomes compared to those engaging in passive prayer.
Methods
Design Overview and Setting
The experiment trial took place from October 2020 to February 2021 in Rehabzone clinic, a rehabilitation clinic affiliated with the physical therapy department of Antonine University in Lebanon. The local ethics committees from Antonine University approved the trial. All participants signed informed consent. The full study protocol is registered at ClinicalTrials.gov (NCT04614272.). In the present paper, the effects of two types of prayer (active and passive) versus a control condition (poem reading), on CPM, PPT, and pain intensity rated on a numeric pain rating scale (NPRS) in healthy religious university students are reported. Outcome measures were assessed at baseline and directly after the intervention. The trial is reported following the CONSORT guidelines (http://www.consortstatement.org). Since the study was performed during the Covid-19 outbreak, a hygiene policy was adopted to ensure the safety of the participants and the assessors.
Study Design
The present study is a double-blind randomized controlled experiment. The study participants were blinded to the study hypothesis, and the therapist collecting the data was also blinded to the randomization sequence.
Study Population and Sample Size
Healthy Christian and Muslim male and female participants were recruited through different sources: Flyers distributed at the Antonine University and Rehabzone clinic, emails sent to the Antonine University students, and adverts on social media. People interested to take part in the study were asked to fill out an online questionnaire that screened for inclusion and exclusion criteria. The inclusion criteria were Lebanese English-speaking students aged between 18 and 25 and with a minimum score of at least two over six on the second question from the Duke University Religion Index (DUREL): “How often do you spend time in private religious activities, such as prayer, meditation, or Bible study?” the scores were from one (rarely or never) to six with (more than once a day) (Koenig & Büssing, 2010). This question was chosen from the DURELL since it reflects the NORA and it helped to define the religious activities performed by the participants in private, such as prayer. Subjects who scored low in religiosity (< two/six) were excluded from the study. Subjects were also excluded in case of regular use of medication, pregnancy, severe allergic reactions, systemic, neurological, metabolic, cardiovascular pathologies, chronic pain, psychiatric disease (being under pharmacological or psychiatric treatment), or suffering from hypertension (> 140/90 mm Hg) (Chalaye et al., 2013). People meeting the criteria were called to set up an appointment.
To minimize the risk of bias, confounding variables affecting both the autonomic and the central nervous systems were controlled. While scheduling appointments, participants were asked to consume a light meal must no later than two hours (heavy meals no later than four hours) before the initiation of the experiments (Anjana & Reetu, 2014; Zmarzty et al., 1997) and requested to refrain from physical exertion 24 h before the experiments (Flood et al., 2017; Lemley et al., 2015; Lima et al., 2017; Stolzman & Bement, 2016), to abstain from analgesic medications 48 h before the experiments (Niesters et al., 2013), and to refrain from smoking (Ditre et al., 2016; Perkins et al., 1994), alcohol (Horn-Hofmann et al., 2019), and caffeine (Sawynok, 2011) in the two hours before the experiments. On the day of the experiments, participants were questioned regarding their adherence to these requests.
Sample Size and randomization
The sample size needed for this study was calculated using the software program G*Power 3.1. To detect an average effect size (f = 0.25) based on Cohen’s conventional standards for the interpretation of effect sizes (Cohen, 2013) with a power of P = 0.8 and a significance threshold of α = 0.05 using a one-way between-subject ANOVA, a total sample size of 159 individuals was warranted, with 53 individuals per group (active prayer group, passive prayer group, poem control group) (Faul et al., 2007).
Randomization
Randomization of the 208 participants was performed using a permuted block allocation (block size of four) with 52 blocks and a ratio of 3:1, with three being the prayer group and one representing the control (poem) group. Unequal randomization was used to allocate the participants to an intervention (prayer) or control group (poem). Selective allocation was later used to allocate the participants into an active or passive prayer group based on the style of praying. However, unequal randomization has consequences for statistical power and a 3:1 randomization scheme requires 33% more patients (Hey & Kimmelman, 2014). Therefore, a total sample size of 208 participants was required to provide adequate power for the analyses.
Style of Praying
All participants filled out a self-reported questionnaire called “the prayer function scale” (PFS) which describes ways that people use prayer to deal with personal difficulties (Bade & Cook, 2008). The PFS helped to identify the style of praying (Bade & Cook, 2008). It is a self-report instrument that assesses the motivation or purpose behind an individual’s prayer, while she or he is coping with difficult circumstances. This scale comprises 58 items that are scored on a five-point Likert-type scale ranging from one (almost never) to five (a great deal), and it is divided into four scales: provides acceptance (17 items), provides calm, and focus (11 items), deferring/avoiding (16 items), and provides assistance (14 items). While the deferring/avoiding scale represents a passive type of prayer and coping, the assistance scale represents an active type of prayer and coping. The PFS deferring/avoiding scale and the assistance scale were used to allocate participants from the prayer group into, respectively, a passive or an active prayer group.
Intervention
While the “deferring/avoiding” group was given a script for a passive type of praying, the “ask for assistance group” was given a script for an active type of praying. The two types of prayers were inspired by the PFS (Bade & Cook, 2008). The passive prayer script (i.e., “Please God, take the pain away”) was inspired by the questions in the PFS related to the deferring/avoiding style, and the script for the active prayer (i.e., “Please God, help me endure this pain”) was inspired from the PFS questions related to the ask for assistance style of praying. The control group received the script of a poem and was asked to read this (i.e., “The earth is our home, so blue and so green, let’s do our part to keep the earth clean’’). The poem was chosen to be emotionless, to avoid the psychophysiological responses related to a poetry reading (Wassiliwizky et al., 2017). All three groups received the instructions and scripts on a piece of paper and were asked to repeat the prayer or the poem for a duration of three minutes. The instructions read: “In the next three minutes you are asked to repeat the following sentence, during which you can choose your preferred posture (sitting, standing, or kneeling).”
Outcome Measures
Outcome measures were PPT, CPM, and NPRS assessed before and directly following the intervention. Sociodemographic data such as religious affiliation, age, gender, body mass index (BMI), hand dominance, smoking, alcohol intake, caffeine intake, and physical activity level were also collected at baseline using a self-reported questionnaire. In addition, the DUREL, which is a five-item self-report measure of religious involvement, was used to assess the religiosity level. It assesses the three major dimensions of religiosity: ORA, NORA, and IR (Koenig & Büssing, 2010).
PPT
PPT assessment is considered a reliable method for measuring mechanical pain thresholds (Cathcart & Pritchard, 2006). PPTs were assessed in a sitting position using a digital algometer (FPX 50, Wagner Instruments, Greenwich, USA) unilaterally (at the side of the dominant hand) at two different body sites. The investigator applied the pressure in a perpendicular direction relative to the muscle while increasing the force at a rate of one kg//s until the participant said to stop when the sensation became intolerable. The pressure marked at that moment was determined as the PPT, measured in kg/cm2. The first location was the trapezius belly, with PPTs being assessed at mid-distance between the acromion and the spinous process of the seven cervical vertebrae (Salavati et al., 2017). The trapezius muscle is a reliable test location for measuring the PPT (Persson et al., 2004). The second location was on the calf belly, with PPTs being measured at the proximal one-third of the calf (Giesbrecht & Battié, 2005; Meeus et al., 2010). The PPT was taken at each of the two anatomical sites with an interval of 30 s until the circuit was repeated a total of two times, starting with the trapezius as the first measurement and then proceeding with the measurement of the calf (Bisset et al., 2015). The time between two PPT measures of the same body location was enough to prevent the pain wind-up effect that might be induced by temporal summation (Cathcart et al., 2009). Four PPT measurements were taken, from which a mean PPT was calculated using the following formula: (PPT calf 1 + PPT calf 2 + PPT trapezius 1 + PPT trapezius 2) / 4.
CPM
Conditioning stimulus (CS). The CS consisted of thermal, hot water stimulation of the non-dominant hand. Participants were comfortably seated next to a water bath and instructed to immerse the non-dominant in hot water for one minute. The temperature water of 45,5 °C, was achieved using an immersion circulator (Immersion Circulator LX, Polyscience, Illinois, USA). The temperature of 45,5 °C has been shown to elicit a robust CPM effect, without potential ceiling or floor effects (Nir et al., 2011, 2012). A line was drawn 10 cm proximally of the wrist crease marking until where the hand needs to be immersed, to ensure whole-hand immersion. Participants were instructed to keep their hands still and unclenched and motivated to complete one full minute of hand immersion. Participants could see a countdown timer of the immersed time. If the participant could not complete the entire one minute, the duration of immersion was recorded. Previous research has shown fair to excellent reliability for the use of a hot noxious water bath as a CS (Kennedy et al., 2016).
Test stimulus (TS). The TS existed of mechanical pain stimulation applied using algometry and assessed by determining the PPT. Therefore, PPTs were taken prior to and following the application of the CS as described in the section PPT. The use of PPT has been validated as a proper TS for measuring CPM (Klyne et al., 2015). It has been shown that CPM is still active five minutes after the removal of the CS in studies using experimental pain (France & Suchowiecki, 1999; Motohashi & Umino, 2001) and argued that the purest CPM effect is obtained by measuring immediately following the CS (i.e., sequential) and not during (i.e., parallel) (Yarnitsky et al., 2015). In line with this recommendation, a sequential CPM paradigm protocol was used.
The CPM outcome score was calculated using the following formula: average of the two consecutive PPTs per location following CS—the average of the two consecutive PPTs per location before the CS (i.e., (T1 (PPT1 trapezius + PPT2 trapezius + PPT1 calf + PPT2 calf)/4)-(T0 (PPT1 trapezius + PPT2 trapezius + PPT1 calf + PPT2 calf)/4)). Hence, higher CPM values reflect better functioning of endogenous pain inhibition. The CPM protocol was repeated before and following the intervention (prayer or poem reading). The post-intervention CPM protocol took place at least 10 min after the pre-intervention or baseline CPM protocol to ensure wash-out of the CPM effect (France & Suchowiecki, 1999; Motohashi & Umino, 2001). The intervention was delivered during the 10-min break, each participant moved to another room to practice either three minutes of active praying, passive praying, or three minutes of poetry reading.
NPRS
Pain intensity for thermal hot water stimulation was evaluated using a NPRS from 0 to 100, with 0 referring to “no pain” and 100 to “maximal pain” felt. It was assessed after the first 30 s of immersion and once more immediately after removing the CS.
Statistical Analysis
All analyses were conducted using SPSS version 2.6 (IBM, New York, USA). The normality of the data was assessed using the Shapiro–Wilk test. Descriptive analyses were used to present the sociodemographic and clinical group characteristics, which were described for the control and prayer group as a whole, and separate for the active prayer group and the passive prayer group. Mean, median, standard deviation, interquartile, and confidence intervals were calculated for the description of continuous variables, whereas frequencies and percentages were calculated for the categorical variables. To evaluate differences in sociodemographic features between all groups, the Chi-square was used for categorical variables, the Kruskal–Wallis was used for the continuous variables with non-normal data distribution, and the Mann–Whitney was used for the continuous variables with normal data distribution. The presence of CPM effects before the intervention was examined using the Wilcoxon signed-rank test to compare the PPT post-conditioning vs. pre-conditioning.
To answer the first research question which evaluated the effect of prayer on PPT, CPM, and NPRS compared to the poetry reading in religious individuals, linear mixed models (LMM) were constructed to test for mean differences between groups (prayer vs control) with the factors “time” (pre, post) and “group” (prayer, control). To answer the second research question which evaluated the effect of two types of prayer (active, passive) on PPT, CPM, and NPRS compared to a poetry reading in religious individuals, LMMs were constructed to test for mean differences between three groups (active, passive, control), with the factors “time” (pre, post) and “group” (active, passive, control). The residuals of the LMMs were checked for normal distribution. When required, post hoc pairwise comparisons were taken using a Bonferroni correction. An intervention-by-time interaction for fixed effects and a main effect for the factor time were analyzed, and random intercept for subject was included to account for within-subject variability. Statistical significance was accepted at a p level of 0.050. Imbalances in demographic data were considered covariates and were included in the analysis.
Results
Participant Characteristics
The participant flow diagram reflected in Fig. 1 shows the number of total responders and participants who were assessed for eligibility, and those who were randomized and underwent allocated intervention and measurements for each group. A total of 208 religious individuals (age range, 17–25 years) took part in the study. While 156 participants were allocated to the prayer group, 52 were allocated to the control group. Within the prayer group, 94 subjects (62%) were allocated to the active prayer group, whereas 62 individuals (38%) were allocated to the passive prayer group in line with the results of the PFS. All participants reported having performed the prayer or read the poem in a sitting position.Fig. 1 CONSORT flow diagram that shows the number of total responders and participants who were assessed for eligibility, and those who were randomized and underwent allocated intervention and measurements for each group
Group Differences over Sociodemographic Variables
There was a significant difference in alcohol consumption (p = 0.010) between the two groups (prayer and control), whereas significant imbalances between the three groups (active prayer, passive prayer, and control) were observed for age (p = 0.027), religion (p = 0.025), and alcohol consumption (p = 0.022). These imbalances were included as covariates in the LMM analyses. The three groups showed no differences in the NORA (p = 0.434). Demographic features of the prayer and control group are summarized in Table 1, whereas features of the active prayer, passive prayer, and control are represented in Table 2.Table 1 Socio-demographic Factors of the Prayer Group and the Control Group at Baseline
Variables Values Prayer Control Chi-square value p value
Gender n (%)
Male 74 (47.4%) 23 (44.2%) 0.161 0.680*
Female 82 (52.6%) 29 (55.8%)
Religion n (%)
Christian 137 (87.8%) 44 (84.6%) 0.360 0.550*
Muslim 19 (12.2%) 8 (15.4%)
Hand dominance n (%)
Right 141 (90.4%) 47 (90.4%) 0.000 1.000*
Left 15 (9.6%) 5 (9.6%)
Smoking n (%)
No 115 (73.7%) 39 (75%) 1.067 0.780*
1 pack /day 1 (0.6%) 1 (1.9%)
½ pack /day 39 (25%) 12 (23.1%)
1 pack/ week 1 (0.6%) 0 (0,0%)
Alcohol n (%)
No 134 (85.9%) 46 (88.5%) 11.283 0.010*
2/ week 22 (14.1%) 3 (5.8%)
1/day 0 (0,0%) 2 (3.8%)
3/week 0 (0,0%) 1 (1.9%)
Caffeine n (%)
No 73 (46.8%) 22 (42.3%) 1.550 0.670*
1/ day 78 (50%) 28 (53.8%)
2/day 3 (1.9%) 2 (3.8%)
3/day 2 (1.3%) 0 (0,0%)
Menstrual phase n (%)
Follicular 42 (56.8%) 11 (47.8%) 2.018 0.360*
During menses 10 (13.5%) 6 (26.1%)
Ovulation 22 (29.7%) 6 (26.1%)
Physical activity n (%)
Yes 86 (55.1%) 30 (57.7%) 0.162 0.690*
No 70 (44.9%) 22 (42.3%)
Durel ORA n (%)
Never 2 (1.3%) 2 (3.8%) 4.292 0.510*
Once/year or less 7 (4.5%) 0 (0,0%)
Few times a year 57 (36.5%) 18 (34.6%)
Few times/month 43 (27.6%) 15 (28.8%)
Once/week 37 (23.7%) 12 (23.1%)
More than once/week 10 (6.4%) 5 (9.6%)
Durel NORA n (%)
Few times/month 75 (48.1%) 18 (34.6%) 3.570 0.470*
Once/week 13(8.3%) 5 (9.6%)
Two or more /week 22 (14.1%) 11 (21.2%)
Daily 34 (21.8%) 12 (23.1%)
More than once /day 12 (7.7%) 6 (11.5%)
Durel IR Q1 n (%)
Definitely not true 1 (0.6%) 1 (1.9%) 4.060 0.400*
Tends not to be true 1 (0.6%) 2 (3.8%)
Unsure 14 (9%) 6 (11.5%)
Tends to be true 32 (20.5%) 11 (21.2%)
Definitely true of me 108 (69.2%) 32 (61.5%)
Durel IR Q2 n (%)
Definitely not true 5 (3.2%) 2 (3.8%) 1.239 0.870*
Tends not to be true 11 (7.1%) 3 (5.8%)
Unsure 25 (16%) 7 (13.5%)
Tends to be true 70 (44.9%) 21 (40.4%)
Definitely true of me 45 (28.8%) 19 (36.5%)
Durel IR Q3 n (%)
Definitely not true 10 (6.4%) 3 (5.8%) 4.300 0.370*
Tends not to be true 15 (9.6%) 6 (11.5%)
Unsure 37 (23.7%) 6 (11.5%)
Tends to be true 55 (35.3%) 19 (36.5%)
Definitely true of me 39 (25%) 18 (34.6%)
Age Mean (SD) 20.39 (1.98) 20.27 (2.05) 0.630◻
Median (IQR) 20 (3) 20 (4)
BMI Mean (SD) 23.67 (3.98) 23.13 (4.21) 0.680◻
Median (IQR) 23.16 (4.54) 22.8 (4.8)
% percentage; *: P-values were calculated using Chi-square tests; ◻: P-values were calculated using Mann–Whitney test; BMI: body mass index; Durel: Duke University Religion index; IR intrinsic religiosity; n frequency; NORA non-organizational religious activity; ORA organizational religious activity
Table 2 Socio-demographic Factors of the Active and Passive Prayer Group and the Control Group at Baseline
Variables Values Active Passive Control Chi-square value p
value
Gender n (%)
Male 47 (50%) 35 (56.5%) 23 (44.2%) 0.786 0.675*
Female 47 (50%) 27 (43.5%) 29 (55.8%)
Religion n (%)
Christian 88 (93.6%) 49 (79%) 44 (84.6%) 7.390 0.025*
Muslim 6 (6.4%) 13 (21%) 8 (15.4%)
Hand dominance n (%)
right 83 (83.3%) 58 (93.5%) 47 (90.4%) 1.185 0.550*
left 11 (11.7%) 4 (6.5%) 5 (9.6%)
Smoking n (%)
No 68 (72.3%) 47 (75.8%) 39 (75%) 2.475 0.871*
1 pack/day 1 (1.1%) 0 (0,0%) 1 (1.9%)
½ pack/day 24 (25.5%) 15 (24.2%) 12 (23.1%)
1 pack/week 1 (1.1%) 0 (0%)
Alcohol n (%)
No 77 (81.9%) 57 (91.9%) 46 (88.5%) 14.838 0.022*
2 drinks/week 17 (18, 1%) 5 (8.1%) 3 (5.8%)
1drink/day 0 (0,0%) 0 (0%) 2 (3.8%)
3 drinks/week 0 (0,0%) 0 (0%) 1 (1.9%)
Caffeine n (%)
No 42 (44.7%) 31 (50%) 22 (42.3%) 2.152 0.905*
1/ day 49 (52.1%) 29 (46.8%) 28 (53.8%)
2/day 2 (2.1%) 1 (1.6%) 2 (3.8%)
3/day 1 (1.1%) 1 (1.6%) 0 (0,0%)
Menstrual phase n (%)
Follicular 27 (57.4%) 15 (55.6%) 11 (47.8%) 2.370 0.668*
During menses 7 (14.9%) 3 (11.1%) 6 (26.1%)
Ovulation 13 (27.7%) 9 (33.3%) 6 (26.1%)
Physical activity n (%)
Yes 49 (52.1%) 36 (58.1%) 30 (57.7%) 0.695 0.707*
No 45 (47.9) 26 (41.9%) 22 (42.3%)
Durel ORA n (%)
Never 1 (1.1%) 1 (1.6%) 2 (3.8%) 5.710 0.839*
Once /year or less 3 (3.2%) 4 (6.5%) 0 (0,0%)
Few times a year 34 (36.2%) 23 (37.1%) 18 (34.6%)
Few times/month 27 (28.7%) 16 (25.8%) 15 (28.8%)
Once/week 23 (24.5%) 14 (22.6%) 12 (23.1%)
More than once/week 6 (6.4%) 4 (6.5%) 5 (9.6%)
Durel NORA n (%)
Few times/month 43 (45.7%) 32 (51.6%) 18 (34.6%) 7.991 0.434*
Once/week 9 (9.6%) 4 (6.5%) 5 (9.6%)
Two or more /week 16 (17%) 6 (9.7%) 11 (21.2%)
Daily 17 (18.1%) 17 (27.4%) 12 (23.1%)
More than once /day 9 (9.6%) 3 (4.8%) 6 (11.5%)
Durel IR Q1 n (%)
Definitely not true 0 (0,0%) 1 (1.6%) 1 (1.9%) 5.847 0.664*
Tends not to be true 1 (1.1%) 0 (0%) 2 (3.8%)
Unsure 8 (8.5%) 6 (9.7%) 6 (11.5%)
Tends to be true 21 (22.3%) 11 (17.7%) 11 (21.2%)
Definitely true of me 64 (68.1%) 44 (71%) 32 (61.5%)
Durel IR Q2 n (%)
Definitely not true 5 (5.3%) 0 (0,0%) 2 (3.8%) 4.737 0.785*
Tends not to be true 6 (6.4%) 5 (8.1%) 3 (5.8%)
Unsure 14 (14.9%) 11 (17.7%) 7 (13.5%)
Tends to be true 42 (44.7%) 28 (45.2%) 21 (40.4%)
Definitely true of me 27 (28.7%) 18 (29%) 19 (36.5%)
Durel IR Q3 n (%)
Definitely not true 8 (8.5%) 2 (3.2%) 3 (5.8%) 7.395 0.495*
Tends not to be true 10 (10.6%) 5 (8.1%) 6 (11.5%)
Unsure 21 (22.3%) 16 25.8%) 6 (11.5%)
Tends to be true 30 (31.9%) 25 (40.3%) 19 (36.5%)
Definitely true of me 25 (26.6%) 14 (22.6%) 18 (34.6%)
Age Mean (SD) 20.36 (1.99) 19.9 (1.91) 20.27 (2.05) 7.198 0.027◻
Median (IQR) 21 (3.00) 19 (3.00) 20 (4.00)
BMI Mean (SD) 23.53 (4.04) 23.25 (3.88) 23.13 (4.21) 1.476 0.478◻
Median (IQR) 23.37 (4.51) 23.1 (4.30) 22.9 (4.80)
% percentage; * P-values were calculated using Chi-square tests; ◻ P-values were calculated using Mann–Whitney test; BMI body mass index; Durel Duke University Religion index; IR intrinsic religiosity; N frequency; NORA non-organizational religious activity; ORA organizational religious activity
CPM effect
The results of the Wilcoxon signed-rank assessing the occurrence of the CPM effect before the intervention showed that the CS elicits a significant change in the average PPT before the intervention (Z = − 4.29, p< 0.001) which indicates the overall presence of a CPM effect. Looking at individual responses, 128 participants out of 208 participants (61.5%) showed an increase in the PPT following CS indicating that they were CPM responders, while 80 participants (38.5%) were considered to be non-responders.
Results to Answer the First Research Question.
Effects of prayer versus control on PPT, CPP, and NPRS
Descriptive statistics of PPT, CPM, and NPRS for the data related to the first research question are shown in Table 3.Table 3 CPM, NPRS, and PPT of the Prayer Group and the Control Group
Outcomes Prayer Control
Mean(SD) Median (IQR) 95% CI Mean(SD) Median (IQR) 95% CI
LB UB LB UB
CPM pre 0.54(1.89) 0.39 (1.83) 0.25 0.84 0.72(2.27) 0.45 (2.08) 0.08 1.35
CPM post 0.15(1.80) 0.81 (1.54) -0.13 0.43 0.38(1.54) 0.30(1.84) −0.50 0.81
NPRS pre 54.80(25.6) 60.00(40.00) 50.78 58.88 52.54(24.35) 57.50(34.13) 45.76 59.32
NPRS post 45.06(24.59) 50.00(39.63) 41.17 48.90 45.47 (24.07) 49.25(34.38) 38.76 52.17
PPT pre 10.23(4.25) 9.10(5.55) 9.57 10.90 11.09 (4.51) 10.17(6.90) 9.82 12.33
PPT post 12.03 (5.90) 10.64(7.40) 11.10 12.97 11.76(5.25) 10.51(7.06) 10.30 13.22
CI confidence interval; CPM conditioned pain modulation; IQR interquartile; LB lower bond; NPRS numeric pain ration scale; PPT pressure pain threshold; pre pre-intervention; post post-intervention; SD standard deviation, UB upper bond
PPT The LMM analysis showed a significant group-by-time interaction for PPT (p = 0.014). Post hoc pairwise comparisons showed a significant increase in the PPT after the intervention in the prayer group (p < 0.001) (mean difference (MD): 1.806; 95% CI, 1.357 to 2.25) which was not the case for the control or poem group (p = 0.085) (MD: 0.682; 95% CI, − 0.95 to 1.460).
CPM No significant group-by-time interaction effects were found for CPM (p > 0.050). However, a significant main effect for time was observed (p = 0.030). Participants presented a drop in their CPM scores following the intervention ((EM mean post-intervention 1.058; 95% CI, 0.198 to 1.919; EM mean pre-intervention 1.440; 95% CI, 0.579 to 2.301) regardless of being in the prayer or the control group.
NPRS No significant group-by-time interaction effects were found for NPRS (p > 0.050). However, a significant main effect for time was shown (p < 0.001). Participants experienced a drop in their NPRS scores following the intervention (EM mean post-intervention 28.96; 95% CI, 14.416 to 43.49; EM mean pre-intervention 38.049; 95% CI, 52.59 to 23.51) independent of the group which they were in.
Results to Answer the Second Research Question
Effects of active prayer versus passive prayer versus control on PPT, CPP, and NPRS
Descriptive statistics of PPT, CPM, and NPRS for the data related to the second research question are shown in Table 4.Table 4 CPM, NPRS, and PPT of the Active and Passive Prayer Group and the Control Group
Outcomes Active Passive Control
Mean(SD) Median(IQR) 95% confidence of interval Mean(SD) Median(IQR) 95%CI Mean (SD)(SD) Median(IQR) 95% CI
LB UB LB UB LB UB
PPT pre 10.64(4.80) 9.08(6.24) 9.65 11.62 9.6(3.18) 9.06(5.23) 8.80 10.42 11.09(4.51) 10.17(6.90) 9.82 12.33
PPT1 post 12.84(6.57) 10.81(8.34) 11.50 14.20 10.8(4.45) 10.20(6.04) 9.67 11.93 11.76(5.25) 10.51(7.06) 10.30 13.22
CPM pre 0.68(2.18) 0.38(2.18) 0.23 1.13 0.34 (1.32) 0.39(1.60) 0.11 0.68 0.72(2.27) 0.46(2.08) 0.08 1.35
CPM post 0.25(2.02) 0.12(1.93) -0.16 0.66 0.065(1.42) 0.25(1.20) -0.35 0.36 0.38(1.54) 0.30(1.84) -0.50 0.81
NPRS pre 52.75 (26.86) 57.50(39.13) 47.24 58.25 57.99 (23.42) 65.00(37.50) 52.04 63.94 52.54(24.35) 57.50(34.13) 45.76 59.32
NPRS post 42.59(24.66) 45(37.75) 37.53 57.64 48.81(24.19) 55.00(37.50) 42.67 54.95 45.47(24.07) 49.25(34.38) 38.76 52.17
CI confidence interval; CPM conditioned pain modulation; IQR interquartile; LB lower bond; NPRS0 pain numeric ration scale at baseline; NPRS numeric pain rating; PPT pressure pain threshold; pre pre-intervention; post post-intervention; SD standard deviation; UB upper bond
PPT The LMM analysis showed a significant group-by-time interaction for PPT (p = 0.005). Bonferroni post hoc analyses for group-by-time interaction effects revealed a significant increase in the PPT following the active prayer intervention (p < 0.001) (MD 2.21; 95% CI, 1.63 to 2.78) and the passive prayer intervention (p = 0.001) (MD: 1.2; 95% CI, 0.49 to 1.9), compared to the control intervention (p = 0.082) (MD: 0.682; 95% CI, -0.09 to 1.45). There was no significant difference (p = 0.165) between the active (EM mean 16.45) and the passive prayer group (EM mean 14.84). The differences between the active prayer group and the control group (EM mean 15.13) did also not reach statistical significance (p = 0.400). All results can be found in Table 5.Table 5 Group-by-time Interaction for PPT, Comparing Active, Passive, and Control
Within-group differences
Group Time EM mean (LB; UB) Mean Difference (LB;UB) SE p
Active Pre 14.25 (11.07; 17.43) 2.21 (1.63; 2.78) 1.61 < 0.001
Post 16.45 (13.28; 19.63) 1.61
Passive Pre 13.63 (10.41; 16.86) 1.20 (0.49; 1.90) 1.64 0.010
Post 14.84 (11.61; 18.10) 1.64
Control Pre 14.44 (11.41; 17.48) 0.68 (-0.09; 1.45) 1.54 0.082
Post 15.13 (12.1; 18.16) 1.54
Between-group differences
Active vs. Passive 1.62 (− 0.41; 3.64) 0.84 0.165
Active vs. Control 1.33 (− 0.80; 3.46) 0.89 0.400
Passive vs. Control − 0.29 (− 2.60; 2.02) 0.95 1.000
EM mean estimated marginal mean; LB lower bond; PPT pressure pain threshold, pre pre-intervention; post post-intervention; UB upper bond
CPM No significant group-by-time interaction effects were found for CPM (p > 0.050). However, a significant main effect for time was shown (p = 0.030). Participants experienced a reduction in CPM scores following the intervention (EM mean post-intervention 0.968; 95% CI, 0.088 to 1.848; EM mean pre-intervention 1.350; 95% CI, 0.470 to 2.23) independent of their group allocation.
NPRS No significant group-by-time interaction effects were found for NPRS (p > 0.050). However, a significant main effect for time was established (p < 0.001). Participants experienced a reduction in NPRS scores following the intervention (EM mean post-intervention 30.82; 95% CI, 15.99 to 45.65; EM mean pre-intervention 39.92; 95% CI, 25.08 to 54.74), regardless of the group they were in.
Discussion
This study aimed at investigating the pain modulating effect of prayer in a sample of healthy religious individuals. The primary aim of this study was to determine the effect of prayer on mechanical pain sensitivity, endogenous pain modulation, and pain intensity compared to poem reading. It was hypothesized that engaging in prayer would lead to increases in PPT and CPM efficacy, and a decrease in NPRS, while no prayer would not induce any changes, and that these increases would be greater following active prayer than when engaging in passive prayer. The findings provide some support for these hypotheses.
Concerning mechanical pain sensitivity, results showed a significant increase in PPT over time in the prayer groups, regardless of the type of prayer, and this effect was not present in the poem reading control group. However, when the types of prayer were compared to each other or with the poem reading control group, statistics did not reach significance.
Regarding endogenous pain modulation, and in contrast to our hypotheses, both prayer groups and the poem reading control group showed a decrease in CPM efficacy following the intervention. Their effects were similar between groups. To explain the reduced CPM following the intervention, several hypotheses can be proposed: (1) The decrease in CPM efficacy could be explained by the use of a fixed and not an adapted conditioning paradigm. Previous studies (Nir et al., 2011; Oono et al., 2011) showed that CPM could be intensity-dependent and thus an increase in the intensity of the CS would induce better CPM results. Prior studies also showed a decreased CPM efficacy during a second CPM testing (Coppieters et al., 2016; Meeus et al., 2015). It may be that each successive conditioned noxious stimulus decreases CPM efficacy. Coppieters et al. (2016) investigated the effect of relaxation on CPM in chronic whiplash and fibromyalgia patients compared to healthy controls and found a decreased CPM efficacy in the three groups after the intervention, regardless of the type of intervention. (2) It may be that the 10-min break maintained between the two CS could not have been enough to avoid a carry-over effect; therefore, a longer recovery period may be necessary after a previous CPM activation. It is possible that adequate CPM activation after the intervention was affected by all of these factors. (3) Additionally, it could be that prayer and CPM do not rely on the same mechanisms. Pain modulation through religious prayer like mindfulness meditation (Zeidan et al., 2016) seems to rely on non-opioidergic systems (Elmholdt et al., 2017) which suggests the involvement of a non-opioidergic cognitive pain modulation system and the notion of multiple pathways in pain control independent of descending inhibitory mechanisms. Therefore, it was hypothesized that prayers and CPM might rely on different mechanisms and do not reinforce each other.
Regarding pain intensity, NPRS findings for both prayer groups and the poetry reading control group resulted in a significant decrease in scores over time, with no significant differences between groups. The decrease in the poem group could be explained by the distraction from hot water, causing pain by focusing on reading the poem. Distraction is an effective approach to reducing pain (Bukola & Paula, 2017).
As expected, and in line with earlier studies (Elmholdt et al., 2017; Meints et al., 2018), results showed that prayer decreases pain sensation for religious individuals regardless of the type of prayer. Active prayers are related to better health when compared to passive prayers (Bade & Cook, 2008; Tait et al., 2016), and active praying is considered an active or self-management approach to pain, while passive praying is considered a passive style of coping. However, in our study, there were no significant differences between the two styles of praying on pain sensitivity.
Although the exact underlying mechanisms are unclear, several hypotheses may explain how prayer reduces pain. Previous studies showed that the cognitive activity of positive re-appraisal mediated the relationship between prayer and pain. Positive reappraisal involves cognitively reframing an event as more positive or valuable allowing individuals to adapt successfully to stressful life events (Garland et al., 2009). Also, other theories have been elaborated, such as conscious re-appraisal, which can alter the meaning of pain without targeting the sensory aspects of the percept (Woo et al., 2015). Other studies (Elmholdt et al., 2017; Jegindø et al., 2013) highlighted the power of strong expectations driven by beliefs and previous religious coping experiences to explain ‘‘religion-induced analgesia.’’
Strength, Limitations, and Future Research
The present study has several strengths. This is the first study, to our knowledge, to investigate the effect of prayer on CPM. Participants were blinded to the study objectives, and the assessor of the outcome measures was blinded to the intervention allocation. However, when interpreting the results, some limitations must be considered. First, all participants were young and pain-free; thus, the findings cannot be generalized to all ages or individuals suffering from pain conditions. In addition, the prayer was not personalized, which could have reduced its meaning and effects. Future research is needed on the analgesic effect of praying in which it would be necessary to personalize the experiment by allowing the participant to pray in their way to reduce the pain and then allocate them to an active or a passive group, according to their style of praying. Moreover, research should focus on extending the follow-up period to observe the long-term effects of “religious induced analgesia.’’ Furthermore, it would be interesting to inventory expectations and previous religious coping experiences, to examine how these potentially influence the results. Also, subjects were selectively allocated using the prayer function scale to either an active prayer group or to a passive prayer group, rather than being randomly allocated; self-selection bias may have affected the results.
Conclusion
The results suggest that prayer, regardless of the used style, reduces mechanical pain sensitivity and self-reported pain intensity in a healthy religious population. Endogenous pain modulation, assessed using a CPM paradigm, decreased in response to both prayer and poem reading, indicating that CPM and praying probably rely on different mechanisms which do not interact.
Acknowledgements
Jessica Van Oosterwijck is a post-doctoral research fellow funded by the Research Foundation–Flanders (FWO) [grant number 12L5616N].
Author Contributions
The authors (Charbel Najem, DPT, Mira Meeus, Ph.D., Barbara Cagnie, Ph.D., Farah Ayoubi, Ph.D., Paul Van Wilgen, Ph.D., Mikel Al Achek, B.S., Jessica Van Oosterwijck, Ph.D., and Kayleigh De Meulemeester, Ph.D.) confirm contribution to the paper as follows: Study conception: All authors conceived the idea of the study. Study design: C.N., M.M., B.C., F.A., and K.D.M. contributed to the design and protocol of the study. Data collection: C.N., M.A.A. carried out the experiment and collected data supervised by F.A. Analysis and interpretation of results: C.N. made the analysis and interpretation of the results with support from K.D.M. and M.M. Draft manuscript preparation: C.N. drafted the manuscript. M.M., B.C., P.V.W., J.V.O., and K.D.M. critically revised the manuscript. Final version: Approved by all authors.
Funding
The authors received no specific funding for this work.
Data Availability
The data that support the findings of this study are available from the corresponding author (C.N).
Code Availability
Not applicable.
Declarations
Conflict of interests
The authors declare that they have no conflicts of interest.
Ethical approval
The local ethics committees from Antonine University approved the trial. The authors certify that the study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki.
Consent to Participate
All participants signed consent forms.
Consent for Publication
Patients signed informed consent regarding publishing their data.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Anjana, Y., & Reetu, K. (2014). Effect of food intake on pain perception in healthy human subjects. Journal of Evolution of Medical and Dental Sciences, 3(29), 7984–7989. https://link.gale.com/apps/doc/A467680812/AONE?u=anon~6e7e2edf&sid=googleScholar&xid=3a146b7b
Bade MK Cook SW Functions of christian prayer in the coping process Journal for the Scientific Study of Religion 2008 47 1 123 133 10.1111/j.1468-5906.2008.00396.x
Balaguier R Madeleine P Vuillerme N Is one trial sufficient to obtain excellent pressure pain threshold reliability in the low back of asymptomatic individuals? A test-retest study PLoS ONE 2016 10.1371/journal.pone.0160866
Balaguier R Madeleine P Vuillerme N Intra-session absolute and relative reliability of pressure pain thresholds in the low back region of vine-workers: Effect of the number of trials BMC Musculoskeletal Disorders 2016 17 1 350 10.1186/s12891-016-1212-7 27538914
Barnes PM Powell-Griner E McFann K Nahin RL Complementary and alternative medicine use among adults: United States, 2002 Advance Data 2004 343 1 19 10.1016/j.sigm.2004.07.003
Bisset LM Evans K Tuttle N Reliability of 2 protocols for assessing pressure pain threshold in healthy young adults Journal of Manipulative and Physiological Therapeutics 2015 38 4 282 287 10.1016/j.jmpt.2015.03.001 25925018
Borg J Andrée B Soderstrom H Farde L The serotonin system and spiritual experiences American Journal of Psychiatry 2003 160 11 1965 1969 10.1176/appi.ajp.160.11.1965 14594742
Breivik H Collett B Ventafridda V Cohen R Gallacher D Survey of chronic pain in Europe: Prevalence, impact on daily life, and treatment European Journal of Pain 2006 10 4 287 333 10.1016/j.ejpain.2005.06.009 16095934
Bukola IM Paula D The effectiveness of distraction as procedural pain management technique in pediatric oncology patients: A meta-analysis and systematic review Journal of Pain and Symptom Management 2017 54 4 589 600.e1 10.1016/j.jpainsymman.2017.07.006 28712986
Cathcart S Pritchard D Reliability of pain threshold measurement in young adults The Journal of Headache and Pain 2006 7 1 21 26 10.1155/2009/523098 16440140
Cathcart S Winefield AH Rolan P Lushington K Reliability of temporal summation and diffuse noxious inhibitory control Pain Research & Management 2009 14 6 433 438 10.1155/2009/523098 20011713
Chalaye P Devoize L Lafrenaye S Dallel R Marchand S Cardiovascular influences on conditioned pain modulation Pain 2013 154 8 1377 1382 10.1016/j.pain.2013.04.027 23692807
Cohen J Statistical power analysis for the behavioral sciences 2013 Routledge
Coppieters I Cagnie B Nijs J van Oosterwijck J Danneels L De Pauw R Meeus M Effects of stress and relaxation on central pain modulation in chronic whiplash and fibromyalgia patients compared to healthy controls Pain Physician 2016 19 3 119 130 10.36076/ppj/2019.19.119 27008285
Ditre JW Heckman BW Zale EL Kosiba JD Maisto SA Acute analgesic effects of nicotine and tobacco in humans: A meta-analysis Pain 2016 157 7 1373 1381 10.1097/j.pain.0000000000000572 27023418
Elmholdt E-M Skewes J Dietz M Møller A Jensen MS Roepstorff A Wiech K Jensen TS Reduced pain sensation and reduced BOLD signal in parietofrontal networks during religious prayer Frontiers in Human Neuroscience 2017 10.3389/fnhum.2017.00337
Faul F Erdfelder E Lang A-G Buchner A G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences Behavior Research Methods 2007 39 2 175 191 10.3758/BF03193146 17695343
Ferreira-Valente A Sharma S Torres S Smothers Z Pais-Ribeiro J Abbott JH Jensen MP Correction to: Does religiosity/spirituality play a role in function, pain-related beliefs, and coping in patients with chronic pain? A systematic review Journal of Religion and Health 2019 10.1007/s10943-019-00928-1
Flood A Waddington G Thompson K Cathcart S Increased conditioned pain modulation in athletes Journal of Sports Sciences 2017 35 11 1066 1072 10.1080/02640414.2016.1210196 27454706
France CR Suchowiecki S A comparison of diffuse noxious inhibitory controls in men and women Pain 1999 81 1–2 77 84 10.1016/s0304-3959(98)00272-3 10353495
Garland E Gaylord S Park J The role of mindfulness in positive reappraisal Explore 2009 5 1 37 44 10.1016/j.explore.2008.10.001 19114262
Giesbrecht RJ Battié MC A comparison of pressure pain detection thresholds in people with chronic low back pain and volunteers without pain Physical Therapy 2005 85 10 1085 1092 10.1093/ptj/85.10.1085 16180957
Hey SP Kimmelman J The questionable use of unequal allocation in confirmatory trials Neurology 2014 82 1 77 79 10.1212/01.wnl.0000438226.10353.1c 24306005
Horn-Hofmann C Capito ES Wolstein J Lautenbacher S Acute alcohol effects on conditioned pain modulation, but not temporal summation of pain Pain 2019 160 9 2063 2071 10.1097/j.pain.0000000000001597 31276454
Illueca M Doolittle BR The use of prayer in the management of pain: A systematic review Journal of Religion and Health 2020 59 2 681 699 10.1007/s10943-019-00967-8 31912350
Jegindø E-ME Vase L Skewes JC Terkelsen AJ Hansen J Geertz AW Roepstorff A Jensen TS Expectations contribute to reduced pain levels during prayer in highly religious participants Journal of Behavioral Medicine 2013 36 4 413 426 10.1007/s10865-012-9438-9 22772583
Jors K Büssing A Hvidt NC Baumann K Personal prayer in patients dealing with chronic illness: A review of the research literature Evidence-Based Complementary and Alternative Medicine 2015 2015 e927973 10.1155/2015/927973
Karmann AJ Maihöfner C Lautenbacher S Sperling W Kornhuber J Kunz M The role of prefrontal inhibition in regulating facial expressions of pain: A repetitive transcranial magnetic stimulation study The Journal of Pain 2016 17 3 383 391 10.1016/j.jpain.2015.12.002 26705973
Kennedy DL Kemp HI Ridout D Yarnitsky D Rice ASC Reliability of conditioned pain modulation: A systematic review Pain 2016 157 11 2410 2419 10.1097/j.pain.0000000000000689 27559835
Klyne DM Schmid AB Moseley GL Sterling M Hodges PW Effect of types and anatomic arrangement of painful stimuli on conditioned pain modulation The Journal of Pain: Official Journal of the American Pain Society 2015 16 2 176 185 10.1016/j.jpain.2014.11.005
Koenig HG Büssing A The Duke University religion index (DUREL): A five-item measure for use in epidemological studies Religions 2010 1 1 78 85 10.3390/rel1010078
Koenig HG George LK Titus P Religion, spirituality, and health in medically Ill hospitalized older patients Journal of the American Geriatrics Society 2004 52 4 554 562 10.1111/j.1532-5415.2004.52161.x 15066070
Laird SP Snyder CR Rapoff MA Green S Measuring private prayer: Development, validation, and clinical application of the multidimensional prayer Inventory International Journal for the Psychology of Religion 2004 14 4 251 272 10.1207/s15327582ijpr1404_2
Lemley KJ Hunter SK Bement MKH Conditioned pain modulation predicts exercise-induced hypoalgesia in healthy adults Medicine and Science in Sports and Exercise 2015 47 1 176 184 10.1249/MSS.0000000000000381 24870571
Levin JS Chatters LM Taylor RJ Religious effects on health status and life satisfaction among black Americans The Journals of Gerontology: Series B 1995 50B 3 S154 S163 10.1093/geronb/50B.3.S154
Lima LV Abner TSS Sluka KA Does exercise increase or decrease pain? Central mechanisms underlying these two phenomena The Journal of Physiology 2017 595 13 4141 4150 10.1113/JP273355 28369946
Meeus M Hermans L Ickmans K Struyf F Van Cauwenbergh D Bronckaerts L De Clerck LS Moorken G Hans G Grosemans S Nijs J Endogenous pain modulation in response to exercise in patients with rheumatoid arthritis, patients with chronic fatigue syndrome and comorbid fibromyalgia, and healthy controls: A double-blind randomized controlled trial Pain Practice: THe Official Journal of World Institute of Pain 2015 15 2 98 106 10.1111/papr.12181 24528544
Meeus M Roussel NA Truijen S Nijs J Reduced pressure pain thresholds in response to exercise in chronic fatigue syndrome but not in chronic low back pain: An experimental study Journal of Rehabilitation Medicine 2010 42 9 884 890 10.2340/16501977-0595 20878051
Meints SM Mosher C Rand KL Ashburn-Nardo L Hirsh AT An experimental investigation of the relationships among race, prayer, and pain Scandinavian Journal of Pain 2018 18 3 545 553 10.1515/sjpain-2018-0040 29794272
Mohandas E Neurobiology of spirituality Mens Sana Monographs 2008 6 1 63 80 10.4103/0973-1229.33001 22013351
Motohashi K Umino M Heterotopic painful stimulation decreases the late component of somatosensory evoked potentials induced by electrical tooth stimulation Brain Research. Cognitive Brain Research 2001 11 1 39 46 10.1016/s0926-6410(00)00062-8 11240110
Narayanasamy A The puzzle of spirituality for nursing: A guide to practical assessment British Journal of Nursing 2004 13 19 1140 1144 10.12968/bjon.2004.13.19.16322 15573007
Neubauer RL Prayer as an interpersonal relationship: A neuroimaging study Religion, Brain & Behavior 2014 4 2 92 103 10.1080/2153599X.2013.768288
Niesters M Aarts L Sarton E Dahan A Influence of ketamine and morphine on descending pain modulation in chronic pain patients: A randomized placebo-controlled cross-over proof-of-concept study British Journal of Anaesthesia 2013 110 6 1010 1016 10.1093/bja/aes578 23384733
Nir R-R Granovsky Y Yarnitsky D Sprecher E Granot M A psychophysical study of endogenous analgesia: The role of the conditioning pain in the induction and magnitude of conditioned pain modulation European Journal of Pain (london, England) 2011 15 5 491 497 10.1016/j.ejpain.2010.10.001 21035364
Nir R-R Yarnitsky D Honigman L Granot M Cognitive manipulation targeted at decreasing the conditioning pain perception reduces the efficacy of conditioned pain modulation Pain 2012 153 1 170 176 10.1016/j.pain.2011.10.010 22119318
O'Beirne S Katsimigos AM Harmon D Forgiveness and chronic pain: A systematic review Irish Journal of Medical Science 2020 189 4 1359 1364 10.1007/s11845-020-02200-y 32125628
Oono, Y., Wang, K., Svensson, P., & Arendt-Nielsen, L. (2011). Conditioned pain modulation evoked by different intensities of mechanical stimuli applied to the craniofacial region in healthy men and women. Journal of Orofacial Pain, 25(4), 364–375. http://www.quintpub.com/journals/ofph/index.php
Pargament K Kennell J Hathaway W Grevengoed N Newman J Jones W Religion and the problem-solving process: Three styles of coping Journal for the Scientific Study of Religion 1988 27 90 10.2307/1387404
Pargament KI Mahoney A Theory: “Sacred matters: Sanctification as a vital topic for the psychology of religion” The International Journal for the Psychology of Religion 2005 15 3 179 198 10.1207/s15327582ijpr1503_1
Perkins KA Grobe JE Stiller RL Scierka A Goettler J Reynolds W Jennings JR Effects of nicotine on thermal pain detection in humans Experimental and Clinical Psychopharmacology 1994 2 1 95 106 10.1037/1064-1297.2.1.95
Perlaki G Orsi G Schwarcz A Bodi P Plozer E Biczo K Aradi M Doczi T Komoly S Hejjel L Kovacs N Janszky J Pain-related autonomic response is modulated by the medial prefrontal cortex: An ECG–fMRI study in men Journal of the Neurological Sciences 2015 349 1 202 208 10.1016/j.jns.2015.01.019 25623806
Persson AL Brogårdh C Sjölund BH Tender or not tender: Test-retest repeatability of pressure pain thresholds in the trapezius and deltoid muscles of healthy women Journal of Rehabilitation Medicine 2004 36 1 17 27 10.1080/16501970310015218 15074434
Poloma MM Pendleton BF The effects of prayer and prayer experiences on measures of general well-being Journal of Psychology and Theology 1991 19 1 71 83 10.1177/009164719101900107
Robles B Upchurch DM Kuo T Comparing complementary and alternative medicine use with or without including prayer as a modality in a local and diverse United States jurisdiction Frontiers in Public Health 2017 5 56 10.3389/fpubh.2017.00056 28377918
Salavati M Akhbari B Ebrahimi Takamjani I Ezzati K Haghighatkhah H Reliability of the upper trapezius muscle and fascia thickness and strain ratio measures by ultrasonography and sonoelastography in participants with myofascial pain syndrome Journal of Chiropractic Medicine 2017 16 4 316 323 10.1016/j.jcm.2017.06.003 29276464
Sawynok J Caffeine and pain Pain 2011 152 4 726 729 10.1016/j.pain.2010.10.011 21041029
Seminowicz DA Davis KD Cortical responses to pain in healthy individuals depends on pain catastrophizing Pain 2006 120 3 297 306 10.1016/j.pain.2005.11.008 16427738
Seybold KS Physiological mechanisms involved in religiosity/spirituality and health Journal of Behavioral Medicine 2007 30 4 303 309 10.1007/s10865-007-9115-6 17549618
South RM McDowell L Use of prayer as complementary therapy by christian adults in the bible belt of the United States Religions 2018 9 11 350 10.3390/rel9110350
Stolzman S Bement MH does exercise decrease pain via conditioned pain modulation in adolescents? Pediatric Physical Therapy : THe Official Publication of the Section on Pediatrics of the American Physical Therapy Association 2016 28 4 470 473 10.1097/PEP.0000000000000312
Tait R Currier JM Harris JI Prayer coping, disclosure of trauma, and mental health symptoms among recently deployed United States Veterans of the Iraq and Afghanistan conflicts The International Journal for the Psychology of Religion 2016 26 1 31 45 10.1080/10508619.2014.953896
Tippens K Marsman K Zwickey H Is prayer CAM? Journal of Alternative and Complementary Medicine 2009 15 4 435 438 10.1089/acm.2008.0480 19388867
Tzeng H-M Yin C-Y Religious activities of inpatients and their family visitors in Taiwan Journal of Holistic Nursing. 2008 10.1177/0898010107310616
Wachholtz AB Pearce MJ Koenig H Exploring the relationship between spirituality, coping, and pain Journal of Behavioral Medicine 2007 30 4 311 318 10.1007/s10865-007-9114-7 17541817
Wassiliwizky E Koelsch S Wagner V Jacobsen T Menninghaus W The emotional power of poetry: Neural circuitry, psychophysiology and compositional principles Social Cognitive and Affective Neuroscience 2017 12 8 1229 1240 10.1093/scan/nsx069 28460078
White ML Peters R Schim SM spirituality and spiritual self-care: Expanding self-care deficit nursing theory Nursing Science Quarterly 2011 10.1177/0894318410389059
Wiech K Farias M Kahane G Shackel N Tiede W Tracey I An fMRI study measuring analgesia enhanced by religion as a belief system Pain 2008 139 2 467 476 10.1016/j.pain.2008.07.030 18774224
Woo C-W Roy M Buhle JT Wager TD Distinct brain systems mediate the effects of nociceptive input and self-regulation on pain PLOS Biology 2015 13 1 e1002036 10.1371/journal.pbio.1002036 25562688
Yarnitsky D Bouhassira D Drewes AM Fillingim RB Granot M Hansson P Landau R Marchand S Matre D Nilsen KB Stubhaug A Treede RD Wilder-Smith OHG Recommendations on practice of conditioned pain modulation (CPM) testing European Journal of Pain (london, England) 2015 19 6 805 806 10.1002/ejp.605 25330039
Zeidan F Adler-Neal AL Wells RE Stagnaro E May LM Eisenach JC McHaffie JG Coghill RC Mindfulness-meditation-based pain relief is not mediated by endogenous opioids The Journal of Neuroscience 2016 36 11 3391 3397 10.1523/JNEUROSCI.4328-15.2016 26985045
Zmarzty SA Wells AS Read NW The influence of food on pain perception in healthy human volunteers Physiology & Behavior 1997 62 1 185 191 10.1016/s0031-9384(97)00038-3 9226361
| 36495356 | PMC9739350 | NO-CC CODE | 2022-12-14 23:29:54 | no | J Relig Health. 2022 Dec 10;:1-24 | utf-8 | J Relig Health | 2,022 | 10.1007/s10943-022-01714-2 | oa_other |
==== Front
Heart Lung Circ
Heart Lung Circ
Heart, Lung & Circulation
1443-9506
1444-2892
The Author(s). Published by Elsevier B.V. on behalf of Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ).
S1443-9506(22)01143-X
10.1016/j.hlc.2022.10.013
Original Article
Impact of Nationwide COVID-19 Lockdowns on the Implantation Rate of Cardiac Implantable Electronic Devices
Wood-Kurland Hannah K. MS abc∗
Phelps Matthew MS, PhD d
Thune Jens Jakob MD, PhD a
Philbert Berit MD, PhD e
Larroudé Charlotte Ellen MD, PhD b
Schou Morten MD, PhD b
Hansen Morten Lock MD, PhD b
Gislason Gunnar H. MD, PhD bd
Bang Casper N. MD, PhD ac
a Department of Cardiology, Bispebjerg & Frederiksberg Hospitals, Copenhagen, Denmark
b Department of Cardiology, Herlev-Gentofte Hospital, Hellerup, Denmark
c Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark
d Danish Heart Foundation, Copenhagen, Denmark
e Department of Cardiology, Rigshospitalet, Copenhagen, Denmark
∗ Corresponding author at: Department of Cardiology Bispebjerg and Frederiksberg Hospitals, Bispebjerg Bakke 23, 2400 Copenhagen, Denmark.
10 12 2022
10 12 2022
12 6 2022
5 9 2022
20 10 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.
Aim
The COVID-19 pandemic resulted in a significant decrease in the number of hospital admissions for severe emergent cardiovascular diseases during lockdowns worldwide. This study aimed to determine the impact of both the first and the second Danish nationwide lockdown on the implantation rate of cardiac implantable electronic devices (CIEDs).
Methods
We retrospectively analysed the number of CIED implantations performed in Denmark and stratified them into 3-week intervals.
Results
The total number of de novo CIED implantations decreased during the first lockdown by 15.5% and during the second by 5.1%. Comparing each 3-week interval using rate ratios, a significant decrease in the daily rates of the total number of de novo and replacement CIEDs (0.82, 95% CI [0.70, 0.96]), de novo CIEDs only (0.82, 95% CI [0.69, 0.98]), and non-acute pacemaker implantations (0.80, 95% CI [0.63, 0.99]) was observed during the first interval of the first lockdown. During the second lockdown (third interval), a significant decrease was seen in the daily rates of de novo CIEDs (0.73, 95% CI [0.55, 0.97]), and of pacemakers in total during both the second (0.78, 95% CI [0.62, 0.97]) and the third (0.60, 95% CI [0.42, 0.85]) intervals. Additionally, the daily rates of acute pacemaker implantation decreased during the second interval (0.47, 95% CI [0.27, 0.79]) and of non-acute implantation during the third interval (0.57, 95% CI [0.38, 0.84]). A significant increase was observed in the number of replacement procedures during the first interval of the second lockdown (1.70, 95% CI [1.04, 2.85]).
Conclusions
Our study found only modest changes in CIED implantations in Denmark during two national lockdowns.
Keywords
COVID-19
Cardiovascular implantable electronic devices
Pacemaker
Implantable cardioverter defibrillator
Cardiac resynchronisation therapy
==== Body
pmc What’s New?
• The total number of de novo cardiac implantable electronic devices (CIEDs) were reduced during both the first and the second COVID-19 lockdowns in Denmark.
• The daily rate of non-acute pacemaker implantations decreased during the first interval of the first COVID-19 lockdown.
• The daily rates of both acute and non-acute pacemaker implantations decreased during different intervals of the second COVID-19 lockdown.
• Overall, the daily rates of CIED implantation changed only modestly during two national COVID-19 lockdowns in Denmark.
Introduction
The COVID-19 pandemic has put most health care systems worldwide under enormous pressure. A number of societal lockdowns have been put in place since the start of 2020. On 13 March 2020, Denmark was one of the first European countries to impose a nationwide lockdown, which meant, for example, the reprioritisation and relocation of health care workers, the postponement of scheduled (elective) treatments, the closing of cultural and educational institutions, and non-essential government employees being required to work from home. This reprioritisation and postponement of treatment was controversial, and may have resulted in a lower risk of COVID-19-related death, but a higher chance of cardiovascular death during the pandemic [1].
Studies have shown that, in Italy, which was severely affected by COVID-19, a significant decrease in the number of hospital admissions for severe emergent cardiovascular diseases during the pandemic was observed [2]. Other studies have shown that this was also the case in Denmark, even though it has been less severely affected by COVID-19 than Italy [[3], [4], [5]]. In addition, countries including Greece, Peru, England, and Germany experienced a reduced number of implantations of cardiac implantable electronic devices (CIEDs) during the pandemic [[6], [7], [8], [9]].
The purpose of this study was to investigate if there was a significant reduction in the number of CIED implantations during both COVID-19 lockdowns in Denmark, and whether non-acute surgery was cancelled or postponed, as recommended by health authorities.
Methods
Study Population
The data presented in this study were collected from the Danish National Patient Registry (DNPR), which is a population-based administrative registry. The registry collects data on diagnoses, treatments, and examinations related to all in-hospital and outpatient clinic visits, as well as administrative data. DNPR data are updated continuously [10]. The registry is linked to the Danish Civil Registration System (CRS) on an individual level. The CRS contains data on, among other things, the name, sex, and date of birth of all people living in Denmark. Furthermore, indications for device implantation were collected from the Danish Pacemaker and Implantable Cardioverter Defibrillator (ICD) register. The register is an official clinical quality database and records details of all cardiac device implants and explants in Denmark [11]. The Danish health care system is funded by taxes and is available to the Danish population at no additional cost. The present study included all men and women aged ≥18 years with either an implantable cardioverter defibrillator (ICD; NOMESCO-code: BFCB00-02, BFCB04), pacemaker (NOMESCO codes: BFCA01, BFCA03, and BFCA07), cardiac resynchronisation therapy (CRT; NOMESCO codes: BFCB03 and BFCA04-06), or implantable loop recorder (ILR; NOMESCO code: BFCA50) implanted, in any Danish hospital. Pacemakers and ICDs were not further differentiated on subtype. In a secondary analysis, all replacement procedures across all three device types were included (NOMESCO codes: BFCA1, BFCA10-12, and BFCB5, BFCB50-53).
Our study complied with the principles of the Declaration of Helsinki.
Stratifications
The data on de novo implantations were stratified in two ways: by implantation type and indication; and by implementation type only. The indications were categorised into two groups: non-acute and acute. For pacemakers, acute indications were defined as third-degree atrioventricular (AV) block, or syncope 2 weeks before implantation, and non-acute indications as all other indications in relation to pacemaker implantation. For ICDs, acute indications were defined as cardiac arrest or ventricular arrhythmia with no diagnosis of ischaemic heart disease, myocardial infarction, or bypass 6 months prior to implantation. Non-acute ICDs were defined as all other implantation-related indications. Both hospitalised and ambulatory patients were included. In the case of one person having several indications, the most recent was used. If a person had more than one indication on the same date, the indications were ranked and used as follows: (1) ventricular arrhythmia/cardiac arrest; (2) third-degree AV block; and (3) all other indications. The numbers of CRTs were too few to stratify in this way. The replacement procedures and ILRs were not stratified further.
Approval
Register-based studies do not need ethical approval in Denmark. Furthermore, this study was approved by the Capital Region of Denmark (approval P-2019-191) in accordance with General Data Protection Regulation.
Study Period
The Danish government imposed the first nationwide lockdown on 13 March 2020 and began lifting restrictions on 15 April 2020. The first study periods were then defined as weeks 11–16 of 2019 and 2020. The second lockdown was imposed on 17 December 2020 and lasted until 28 February 2021. Our data covered up to 29 January 2021. Therefore, the second study periods were defined as weeks 51 of 2019 and 2020 up to and including week 5 of 2020 and week 4 of 2021.
As our second study period spanned 1 January, we analysed the data using a custom year, which started on 29 January. In addition, each lockdown was further divided into 3-week intervals. Consequently, our first study periods were defined as the third and fourth 3-week interval, and our second periods as the 16th, 17th, and 18th 3-week interval.
Analysis
We retrospectively analysed the number of implantations during weeks 11–16 of 2020 to data from the same weeks in 2019 and weeks 51–04 of 2020/21, to data from weeks 51–05 of 2019/20. We compared the total number of de novo implantations (ICDs, pacemakers, CRT, and ILRs), shown as a graphical illustration of the calculated daily rates. We also compared the number of implantations of ICDs, pacemakers, CRTs, and ILRs. This was done by calculating the rate ratios between the years for each device by using an exact Poisson test, using 2019 as the reference for the first study period, and 2019 spanning into 2020 for the second. The exact Poisson test performed an exact test of our null hypothesis that the ratio between two rate parameters was 1.0. The rate ratios were shown graphically as forest plots. Furthermore, the number of pacemaker and ICD implantations were stratified by indication. The indications were divided into two groups: non-acute and acute. The proportions between these numbers were shown graphically. Stratification by indication was not possible for CRTs and ILRs, as the numbers would be too few to maintain the patients’ anonymity in this study, as was the case for rate ratios of the number of ICDs implanted with an acute indication during the second and third 3-week interval. For replacement procedures both daily rates and rate ratios were calculated and were shown in the same way as de novo implantations.
Results
The total number of de novo CIED implantations decreased during the first lockdown by 15.5%, from 524 in 2019 to 443 in 2020. Divided by device type there was a change of –16.8% in pacemaker implantation (316 in 2020 and 380 in 2019), –18.1% in ICDs (68 in 2020 and 83 in 2019), and +4.4% in CRTs (47 in 2020 and 45 in 2019), while ILRs were too few to calculate (Table 1 ). Regarding pacemakers, 310 were implanted with a non-acute indication and 70 with an acute indication in 2019. In 2020, 256 were implanted with a non-acute indication (–17.4% vs 2019), while 60 were implanted with an acute indication (–14.3%; Table 1). Regarding ICDs, 66 were implanted with a non-acute indication and 17 with an acute indication in 2019. In 2020, 49 were implanted with a non-acute indication (–25.8% vs 2019), while 19 were implanted with an acute indication (+11.8%; Table 1).Table 1 Numbers and daily implantation rates of cardiac implantable electronic devices in the first COVID-19 lockdown.
Total (de novo + replacements) De novo (total) Replacements (total) CRT (total) ICD (total) ICD (non-acute) ICD (acute) PM (total) PM (non-acute) PM (acute) ILR (total)
First interval 2019 (n) 359 298 61 26 45 35 10 215 177 38 12
First interval 2020 (n) 294 245 49 24 39 26 13 177 141 36 5
Second interval 2019 (n) 266 226 40 19 38 31 7 165 133 32 NAa
Second interval 2020 (n) 242 198 44 23 29 23 6 139 115 24 7
Total 2019 (n) 625 524 101 45 83 66 17 380 310 70 NAa
Total 2020 (n) 536 443 93 47 68 49 19 316 256 60 12
Daily rate, first interval 2019 25.6 21.3 4.4 1.9 3.2 2.5 0.7 15.4 12.6 2.7 0.9
Daily rate, first interval 2020 21 17.5 3.5 1.7 2.8 1.9 0.9 12.6 10.1 2.6 0.4
Daily rate, second interval 2019 19 16.1 2.9 1.4 2.7 2.2 0.5 11.8 9.5 2.3 NAa
Daily rate, second interval 2020 17.3 14.1 3.1 1.6 2.1 1.6 0.4 9.9 8.2 1.7 0.5
The first four rows show data for the first and second 3-week intervals for both 2019 and 2020, stratified by de novo or replacement, de novo by device type, and pacemakers and implantable cardioverter defibrillators (ICDs) by non-acute or acute indication. The next two rows show the corresponding numbers in total for the reference period and during lockdown, respectively. The final four rows show the corresponding daily rates again divided between the first and second 3-week intervals.
a The number of implantable loop recorder (ILR) implantations during these periods was too few to be included.
Abbreviations: CRT, cardiac resynchronisation therapy; PM, pacemaker; NA, not applicable; ICD, implantable cardioverter defibrillator; PM, pacemaker; ILR, implantable loop recorder.
During the second lockdown, the total number of de novo CIED implantations decreased by 5.1% from 623 in 2019/20 to 591 in 2020/21. Divided by device type, there was a change of –8.9% in pacemaker implantations (410 in 2020/2021 and 450 in 2019/20), +10.5% in ICDs (95 in 2020/21 and 86 in 2019/20), +11.0% in CRTs (61 in 2020/21 and 55 in 2019/20), and –21.9% in ILRs (25 in 2020/21 and 32 in 2019/20; Table 2 ). In 2019/20, regarding pacemakers, 343 were implanted with a non-acute indication and 107 with an acute indication. In 2020/21, 325 were implanted with a non-acute indication (–5.2%), while 85 were implanted with an acute indication (–20.6%; Table 2). The ICDs implanted in 2019/20 were too few to be further stratified. In 2020/21, 71 were implanted with a non-acute indication, while 24 were implanted with an acute indication (Table 2). Daily rates of the total number of de novo implantations of all three types of CIED appeared to be lower during both lockdowns compared to the year before (Figure 1 ).Table 2 Numbers and daily implantation rates of cardiac implantable electronic devices in the second COVID-19 lockdown.
Total (de novo + replacements) De novo (total) Replacements (total) CRT (total) ICD (total) ICD (non-acute) ICD (acute) PM (total) PM (non-acute) PM (acute) ILR (total)
First interval 2019 (n) 282 255 27 24 37 29 8 187 142 45 7
First interval 2020 (n) 322 276 46 20 34 26 8 214 164 50 8
Second interval 2019 (n) 287 248 39 21 32 NAa NAa 179 132 47 16
Second interval 2020 (n) 265 215 50 22 46 35 11 139 117 22 8
Third interval 2019 (n) 135 120 15 10 17 NAa NAa 84 69 15 9
Third interval 2020 (n) 124 100 24 19 15 10 5 57 44 13 9
Total 2019 (n) 704 623 81 55 86 NAa NAa 450 343 107 32
Total 2020 (n) 711 591 120 61 95 71 24 410 325 85 25
Daily rate, first interval 2019 20.1 18.2 1.9 1.7 2.6 2.1 0.6 13.4 10.1 3.2 0.5
Daily rate, first interval 2020 23 19.7 3.3 1.4 2.4 1.9 0.6 15.3 11.7 3.6 0.6
Daily rate, second interval 2019 20.5 17.7 2.8 1.5 2.3 NAa NAa 12.8 9.4 3.4 1.1
Daily rate, second interval 2020 18.9 15.4 3.6 1.6 3.3 2.5 0.8 9.9 8.4 1.6 0.6
Daily rate, third interval 2019 16.9 15 1.9 1.3 2.1 NAa NAa 10.5 8.6 1.9 1.1
Daily rate, third interval 2020 13.8 11.1 2.7 2.1 1.7 1.1 0.6 6.3 4.9 1.4 1
The first six rows show the number of implantations done in the first, second, and third 3-week intervals for both 2019 and 2020/21 in total, stratified by de novo or replacement, de novo by device type, and pacemakers and implantable cardioverter defibrillators (ICDs) by non-acute or acute indications. The next two rows show the corresponding numbers in total for the reference period and during lockdown, respectively. The final six rows show the corresponding daily rates again divided between the first, second, and third 3-week intervals. 2020 was the year of the second lockdown, although it spanned into 2021.
a The number of acute ICD implantations could have been calculated from the total and non-acute ICDs and would have been too few, and have therefore been left out for 2019.
Abbreviations: CRT, cardiac resynchronisation therapy; PM, pacemaker; ILR, implantable loop recorder; ICD, implantable cardioverter defibrillator.
Figure 1 Daily rates of implantations of cardiac devices, including cardiac resynchronisation therapy, pacemakers, implantable cardioverter defibrillators, and implantable loop recorders. Each week group contains data on the number of implantations grouped into 3-week intervals, illustrated by the blue line for the reference period of 2019 spanning into 2020, and the orange line for the time of the lockdowns in 2020 spanning into 2021. The black vertical lines indicate in pairs the period of the first and second Danish nationwide lockdowns, respectively.
Abbreviation: CIED, cardiac implantable electronic device.
In order to show whether the differences were statistically significant, we used rate ratios of the daily rates during the first and second lockdown, respectively, using the year before as a reference.
During the first lockdown both the total number of de novo and replacement procedures, as well as the number of de novo procedures, decreased significantly during the first 3-week interval (Figure 2 ). In relation, the number of pacemakers implanted with a non-acute indication decreased significantly during the first 3-week interval (Figure 3 ).Figure 2 Rate ratios of the daily rates of cardiac implantable electronic device (CIED) implantations defined by both device type and 3-week interval during the first and second COVID-19 lockdowns in Denmark with 2019 as the reference. The first interval corresponds to the first 3 weeks of lockdown, the second interval to the next 3 weeks of lockdown, and the third interval to the last 3 weeks, where applicable. Implantable loop recorders (ILRs) during the second interval of the first lockdown were too few to be calculated as rate ratios, and are therefore not included in the graph.
Abbreviations: CI, confidence interval; CRT, cardiac resynchronisation therapy; ICD, implantable cardioverter defibrillator; PM, pacemaker; ILR, implantable loop recorder.
Figure 3 Rate ratios of the daily rates of pacemaker and implantable cardioverter defibrillator (ICD) implantations defined by an non-acute or acute indication and 3-week interval during the first and second COVID-19 lockdown in Denmark, with the previous year as the reference. The first interval corresponds to the first 3 weeks of lockdown, the second interval to the next 3 weeks of lockdown, and the third interval to the last 3 weeks, where applicable. The ICDs implanted according to an acute indication were too few to be calculated as rate ratios for the second and third 3-week interval, and are therefore not included in the graph.
Abbreviations: CI, confidence interval; PM, pacemaker.
During the second lockdown, the daily rates of de novo implantations across all device types decreased significantly during the third 3-week interval, while the daily rates of replacement procedures increased significantly during the first 3-week interval (Figure 2). Additionally, the total number of pacemaker implantations decreased during both the second and third 3-week intervals (Figure 2). In relation, the daily rates of acute pacemaker implantations decreased significantly during the second 3-week interval, whereas the daily rates of non-acute pacemaker implantations decreased during the third 3-week interval (Figure 3).
In the time between the two lockdowns all other daily rates were comparable to the year before, except that there was a higher rate of ILRs during 3-week interval 14 and a lower rate of non-acute ICDs during interval 15 (Supplementary Table 1).
Discussion
The present study is the first study to investigate the consequences of the two Danish nationwide lockdowns during the COVID-19 pandemic, imposed on 13 March and 17 December 2020, respectively, on the number of CIED implantations. The total number of de novo CIEDs was reduced in both the first and the second lockdown. Overall, we observed mostly lower daily rates of the total number of CIEDs implanted. When divided into acute and non-acute indications for CIED, a significant decrease was seen in the daily rates of non-acute pacemaker implantations during the first interval of the first lockdown. In the second lockdown, both acute and non-acute pacemaker implantations decreased significantly during the second and third 3-week intervals, respectively.
Several national studies have been published providing national rates of CIED procedures during COVID-19 lockdowns. In northern Greece, a decrease of 48% was seen in CIED implantations during their first lockdown, while the number of implantations during their second lockdown was comparable to the year before [6]. Catalonia has reported an overall decrease of 56.5% in pacemaker and ICD implantations during the first wave of COVID-19, affecting both elective and urgent implantations, while a study from England also showed decreases in pacemaker, CRT, and ICD implantations for both elective and urgent indications during the first lockdown [8,12]. Peru reported a decrease of 73% in de novo pacemaker implantations for all aetiologies after the beginning of their first social isolation period, and in Germany a decrease of 18% in pacemaker implantations was recorded during their first lockdown [7,9]. Our study also revealed lower daily rates; however, when comparing the changes using rate ratios, we found more significant changes during the second lockdown compared with the first. Keeping in mind that most other studies have been done on the first lockdown/COVID-19 wave only, it appears that several other countries have experienced a more extensive and significant impact of COVID-19 on CIED implantation compared with Denmark. In line with our study, another Danish study that investigated cardiovascular disease during the COVID-19 lockdown found a significant drop in admission rates for new-onset ischaemic stroke and ischaemic heart disease during week 4 of the second lockdown [13], which corresponds to the drop in acute pacemaker implantations during the second 3-week interval found in our study. The Danish study found even larger reductions during the first lockdown, which was not reflected in our study [4].
In order to evaluate if non-acute electronic device procedures were postponed, as recommended by the Danish government, we stratified our data accordingly. During the first interval of the first lockdown, compliance with recommendations from the Danish government were demonstrated, to some extent, as only non-acute pacemaker implantation rates decreased significantly. This was only the case during the first 3-week interval, which could suggest a different prioritisation of patients further into the first lockdown. During the second lockdown, however, we observed an increase in replacement procedures during the first interval, a decrease in acute pacemaker implantations during the second interval, and a decrease of the number of non-acute pacemaker implantations during the third 3-week interval. The increase in replacements could be due to an excess left over from the first lockdown, and the following decrease in acute and non-acute pacemakers could be down to patients staying away from hospitals for fear of contracting COVID-19. Other possible explanations could be the inadequate redistribution of health care resources and prioritisation of patients, or a larger reduction of available staff due to COVID-19-related absenteeism during the second lockdown vs the first, even though instructions issued by the Danish health authorities on how to redistribute health care resources during the lockdowns tried to account for the latter [14]. However, these reasons are speculative, and the only information that can be concluded is that our results showed a greater discrepancy between recommendations from the Danish government and the surgeries performed during the second lockdown. It should be noted that guidelines from the Danish health authorities mainly provided recommendations on how to manage the large number of hospitalised patients with COVID-19 [14]. Furthermore, it was highlighted in the guidelines that surgical procedures should only be postponed on the basis of the clinical assessment of each patient. No definite list of indications that should result in surgery being postponed or not was announced, only a few examples were given: in cardiology, elective pacemaker implantations were given as an example of a procedure to be postponed. However, to our knowledge, most Danish invasive cardiologists agreed on the prioritisation of indications, which was also the basis for how they were stratified in this study, and the general impression is that they were followed, although our study cannot rule out the opposite.
On the contrary, patients in need of a non-acute ICD seem to have not had their procedure postponed during either lockdown. Only some data on acute ICDs were available, and these indicate sufficient treatment, at least during the first interval of both lockdowns. The number of CRTs and ILRs were too few to differentiate by indication, and therefore too few to determine the types of patients affected by any changes.
Overall, a decrease in acute life-saving procedures is problematic and should be avoided. Knowledge obtained from the pandemic and lockdowns is important, and attention to the effects of subsequent lockdowns during this and future pandemics is essential to maintaining the balance of health care resources and the right level of care for cardiovascular patients.
The prioritisation of acute and life-threatening conditions, and patients in need of a CIED during future and subsequent lockdowns or pandemics could prevent a later increase in the need for more extensive care and possibly even in mortality.
Strengths and Limitations
This was a nationwide study and the inclusion and selection biases were therefore minimal. In comparison to most other studies published on CIED implantation during lockdown, this study compared implantation rates with the year before using rate ratios. This is a strength of this study; however, these stricter analytics result in higher requirements, making it more difficult to show significant changes. A limitation is that Denmark is a relatively small country with a population of 5.8 million people, and the number of implantations performed is therefore equally small. This means, as mentioned earlier, that we could not stratify our data on CRTs and ILRs based on indication, as the numbers would have been too small. In addition, this was an observational study, and therefore analyses on changes in implantation rates and the COVID-19 lockdowns are only observational and do not necessarily prove a causal relationship.
Conclusion
Our study found a significant decrease in both the total daily rates of de novo and replacement CIEDs, de novo CIEDs only, and in the daily rates of non-acute pacemaker implantations during the first Danish nationwide lockdown. During the second lockdown, a significant decrease in the total daily rates of pacemaker implantations was observed, divided between both non-acute and acute indications, while an increase in the number of replacements was seen. The reasons behind these findings cannot be concluded, but our results reveal that CIED implantation rates were only modestly affected in Denmark during two national lockdowns.
Funding Sources
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Appendices Supplementary Data
Supplementary Material
Conflicts of Interest
There are no conflicts of interest to disclose.
Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.hlc.2022.10.013.
==== Refs
References
1 Butt J.H. Fosbøl E.L. Gerds T.A. Andersson C. Kragholm K. Biering-Sørensen T. All-cause mortality and location of death in patients with established cardiovascular disease before, during, and after the COVID-19 lockdown: a Danish Nationwide Cohort Study Eur Heart J 42 2021 1516 1523 33624011
2 Toniolo M. Negri F. Antonutti M. Masè M. Facchin D. Unpredictable fall of severe emergent cardiovascular diseases hospital admissions during the COVID-19 pandemic: experience of a single large center in Northern Italy J Am Heart Assoc 9 2020 e017122
3 Andersson C. Gerds T. Fosbøl E. Phelps M. Andersen J. Lamberts M. Incidence of new-onset and worsening heart failure before and after the COVID-19 epidemic lockdown in Denmark: a nationwide cohort study Circ Heart Fail 13 2020 e007274
4 Butt J.H. Fosbøl E.L. Østergaard L. Yafasova A. Andersson C. Schou M. Effect of COVID-19 on first-time acute stroke and transient ischemic attack admission rates and prognosis in Denmark: a nationwide cohort study Circulation 142 2020 1227 1229 32755320
5 Holt A. Gislason G.H. Schou M. Zareini B. Biering-Sørensen T. Phelps M. New-onset atrial fibrillation: incidence, characteristics, and related events following a national COVID-19 lockdown of 5.6 million people Eur Heart J 41 2020 3072 3079 32578859
6 Bechlioulis A. Sfairopoulos D. Korantzopoulos P. Impact of COVID-19 pandemic on cardiac electronic device implantations in Northwestern Greece Am J Cardiovasc Dis 11 2021 489 493 34548948
7 Gonzales-Luna A.C. Torres-Valencia J.O. Alarcón-Santos J.E. Segura-Saldaña P.A. Impact of COVID-19 on pacemaker implant J Arrhythm 36 2020 845 848 32837668
8 Leyva F. Zegard A. Okafor O. Stegemann B. Ludman P. Qiu T. Cardiac operations and interventions during the COVID-19 pandemic: a nationwide perspective Europace 23 2021 928 936 33778881
9 Mathew S. Fraebel C. Johnson V. Abdelgwad S. Schneider N. Müller P. Cardiac arrhythmias in patients with SARS-CoV-2 infection and effects of the lockdown on invasive rhythmological therapy Herzschrittmacherther Elektrophysiol 32 2021 108 113 33355696
10 Schmidt M. Schmidt S.A. Sandegaard J.L. Ehrenstein V. Pedersen L. Sørensen H.T. The Danish National Patient Registry: a review of content, data quality, and research potential Clin Epidemiol 7 2015 449 490 26604824
11 Danish Pacemaker and ICD Register. Danish Pacemaker and ICD Register Annual Report 2019 2020. Available at: https://www.sundhed.dk/content/cms/21/109821_dpir_annual_report_2019_final_211220.pdf 2020
12 Arbelo E. Angera I. Trucco E. Rivas-Gándara N. Guerra J.M. Bisbal F. Reduction in new cardiac electronic device implantations in Catalonia during COVID-19 Europace 23 2021 456 463 33595062
13 Christensen D.M. Butt J.H. Fosbøl E. Køber L. Torp-Pedersen C. Gislason G. Phelps M. Nationwide cardiovascular disease admission rates during a second COVID-19 lockdown Am Heart J 241 2021 35 37 34274314
14 Sundhedsstyrelsen. Notat om reduktion af hospitalsaktivitet ifm COVID-19. In: Health TMo, (ed): The Ministry of Health. Translated title: Information note on the reduction of hospital activities in relation to COVID-19 2020. Available at: https://www.sst.dk/-/media/Udgivelser/2020/Corona/Hospitalskapacitet/Notat-om-reduktion-af-hospitalsaktivitet-ifm-med-COVID-19 2020
| 36513581 | PMC9741195 | NO-CC CODE | 2022-12-14 23:29:54 | no | Heart Lung Circ. 2022 Dec 10; doi: 10.1016/j.hlc.2022.10.013 | utf-8 | Heart Lung Circ | 2,022 | 10.1016/j.hlc.2022.10.013 | oa_other |
==== Front
Neural Comput Appl
Neural Comput Appl
Neural Computing & Applications
0941-0643
1433-3058
Springer London London
8090
10.1007/s00521-022-08090-8
Original Article
Real-time automated detection of older adults' hand gestures in home and clinical settings
Huang Guan [email protected]
1
http://orcid.org/0000-0002-5912-293X
Tran Son N. [email protected]
1
Bai Quan [email protected]
1
Alty Jane [email protected]
23
1 grid.1009.8 0000 0004 1936 826X College of Sciences and Engineering, University of Tasmania, Sandy Bay, TAS 7005 Australia
2 grid.1009.8 0000 0004 1936 826X Wicking Dementia Research and Education Centre, University of Tasmania, Hobart, TAS 7000 Australia
3 grid.1009.8 0000 0004 1936 826X School of Medicine, University of Tasmania, Hobart, TAS 7000 Australia
10 12 2022
114
21 3 2022
22 11 2022
© The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
There is an urgent need, accelerated by the COVID-19 pandemic, for methods that allow clinicians and neuroscientists to remotely evaluate hand movements. This would help detect and monitor degenerative brain disorders that are particularly prevalent in older adults. With the wide accessibility of computer cameras, a vision-based real-time hand gesture detection method would facilitate online assessments in home and clinical settings. However, motion blur is one of the most challenging problems in the fast-moving hands data collection. The objective of this study was to develop a computer vision-based method that accurately detects older adults’ hand gestures using video data collected in real-life settings. We invited adults over 50 years old to complete validated hand movement tests (fast finger tapping and hand opening–closing) at home or in clinic. Data were collected without researcher supervision via a website programme using standard laptop and desktop cameras. We processed and labelled images, split the data into training, validation and testing, respectively, and then analysed how well different network structures detected hand gestures. We recruited 1,900 adults (age range 50–90 years) as part of the TAS Test project and developed UTAS7k—a new dataset of 7071 hand gesture images, split 4:1 into clear: motion-blurred images. Our new network, RGRNet, achieved 0.782 mean average precision (mAP) on clear images, outperforming the state-of-the-art network structure (YOLOV5-P6, mAP 0.776), and mAP 0.771 on blurred images. A new robust real-time automated network that detects static gestures from a single camera, RGRNet, and a new database comprising the largest range of individual hands, UTAS7k, both show strong potential for medical and research applications.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00521-022-08090-8.
Keywords
TAS Test
Hand gesture classification
Similar object detection
Motion blur
Dementia
Bradykinesia
==== Body
pmcIntroduction
In recent years, hand gesture detection has been increasingly explored in human computer interaction research, including sign language detection, video games and virtual reality. The rapid development of deep learning [1, 2] has significantly improved the accuracy of hand gesture detection; for example, researchers have used 3DCNN models to accurately classify hand gestures used in Arabic sign language with 90% accuracy [3].
There is now growing interest in how these technologies can be applied to medical and neuroscience research applications as there is an urgent need, accelerated by the COVID-19 pandemic, for methods that allow remote evaluation of hand movements. Hand movement assessments play a key part in the detection and monitoring of brain disorders such as Parkinson’s and stroke. These disorders are particularly prevalent in older adults and usually require participants to attend clinics or research institutes for face-to-face tests. This is problematic for people who live in rural or remote locations, those with limited mobility and for the majority of patients and research participants during the COVID-19 pandemic. With the wide accessibility of computer cameras, a vision-based method that can detect hand gestures in real time would facilitate online assessments in the home and clinical settings, and this would transform the accessibility and efficiency of medical assessments and research studies.
Apraxia is a neurological disorder characterised by difficulties carrying out precise movements despite the physical ability to perform them. It is usually due to impaired brain connections that integrate the planning, sequencing and motor–sensory integration of movement. Causes in adults include stroke and neurodegenerative disorders, such as Alzheimer’s disease and Parkinson’s disease [4]. We are developing a new online test, the TAS Test, that analyses hand movement features [5] associated with ageing and neurodegenerative disorders, in a large established cohort of older adults. Participants access the online test using their own laptop or desktop computer via a website and then follow a series of instructions to record their hand movements with a webcam. The test is designed to be completed remotely from the research centre without any researcher supervision [6]. A robust real-time automated method is important for clinician and researchers to monitor whether the participants are following the instructions and therefore to further evaluate and analyse the level of apraxia in the hand movement.
However, real-time gesture recognition of fast-moving hands is challenging for several reasons. First, the validated hand movements tests, such as finger tapping and whole hand opening–closing require participants to repeat these movements ‘as fast as possible’ and this creates motion blur, especially for home computer cameras that tend to have a relatively low rate of frames per second (fps). Second, accurately recognising similar gestures such as finger and thumb together (closed position) at the start of the finger tapping cycle or a few centimetres apart (open position) partway through the finger tapping cycle results in inherent errors and confusion, which decreases the detection accuracy. Third, in the home and clinic settings, the backgrounds are typically cluttered and there are variations in ambient lighting and distance of the hands to the camera.
The hand movement tests include repetitive finger tapping (tapping the index finger against the thumb, in the phase and anti-phase) and repetitive hand opening–closing (of all the digits). Both are well-validated tests for evaluating human movement function. Figure 1 illustrates how the anti-phase (or alternate) finger tapping test is performed; the participants are instructed to switch between ‘Gesture 1’ and ‘Gesture 2’ quickly and repeatedly [7].Fig. 1 Two images were taken from a video recording of the anti-phase (alternate) finger tapping test. a Gesture 1 with the right-hand finger and thumb opposed and the left-hand finger and thumb separated. b Gesture 2 with the left-hand finger and thumb opposed and the right-hand finger and thumb separated
The overall aim of this study was to develop a robust method to detect fast-moving hand gestures in real time. Our first objective was to develop a large dataset of hand gestures collected in home environments and clinical settings (with real-life cluttered backgrounds), and split into clear and motion-blurred images. Our second objective was to develop an accurate method for discriminating similar hand gestures in clear and blurred images while remaining real time and compare the accuracy of this to other established methods.
In this study, we establish a new dataset, UTAS7k, with 20 per cent of blurred images included. We compare the detection accuracy of different network structures on hand gesture classification, where we find multi-scale detection technique was effective. To develop an optimal network structure for hand gesture detection, we embedded different types of neural networks into the detector network. We implemented an attention-based hand gesture network to detect the hand gestures performed in the hand movement tests. We developed a new model, RGRNet, for fast-moving hand gesture detection, inspired by CSPDarknet-53 [8]. We have implemented the multi-scale detection technique and embedded one more detection head for the detection of hands in different sizes. We also adopted attention layers in the feature extraction blocks to increase the prediction capability. To further increase the performance of detection, data augmentation was employed during training, including mosaic and left or right flip.
Our experiments revealed that multi-scale detection techniques help to increase the overall performance of similar gesture classification. Also, the experimental results show that our new model, RGRNet improved the accuracy of similar gesture classification on clear images and performs more robustly on both clear and motion-blurred hand gesture images.
The key contributions of this research study are as follows: We develop a novel model, RGRNet, with attention and transformer layers to improve the classification performance on similar gestures and blurred images extracted from hand movement videos. Our model achieved near real-time hand gesture detection and classification with a total processing speed of 18.8ms.
We establish UTAS7k, a real-world dataset comprising more than 7,000 images of similar hand gestures with cluttered backgrounds and split 4:1 into clear and motion-blurred images. Then, we provide a comprehensive comparison of the classification accuracy of different network structures on the hand gesture videos with motion blur and similar gestures and show that our model, RGRNet, has a better performance than the state-of-the-art network structure. All of these contributions have strong potential for medical, neuroscience and computer science research applications.
The organisation of the paper is as follows: in Sect. 2, we summarise the literature related to our research, including hand gesture detection methods, object detection algorithms and network structures for object detection. In Sect. 3, we describe our network structure in detail, including the structure of each block. In Sect. 4, we describe the collection and processing steps to develop UTAS7k, our dataset of hand gestures. In Sect. 5, we present the experiment methods and results of classification accuracy and detection speed of our new network, RGRNet, compared to a range of other popular network structures. Finally, in Sect. 6, we summarise the findings from our work and discuss future directions.
Related work
Hand gesture detection methods
Methods to detect hand gestures are generally categorised into two types: wearable device-based gesture recognition or computer and vision-based gesture recognition. The first method typically measures the angle between fingers and proximal interphalangeal joints to estimate the gestures [9] although the activity of the muscles in the digits and upper limb has also been used [10]. A range of wearable devices have been employed, including data gloves (with embedded sensors), tactile switches, optical goniometers and resistance sensors to measure the bending of finger joints [11]. This approach mainly focuses on increasing the accuracy to pinpoint the position of the hand in the 3D model. However, the main limitations of the wearable sensor approach are accessibility, cost and infection. The clinician or researcher needs to find a robust method of delivering the sensors to patients or participants with clear instructions, or bring the participants into the research laboratory setting. Also, sensors can be very expensive; for example, commercial wearable data gloves typically cost in the range of $1000 to $20,000 each [12]. Furthermore, any multi-user wearable devices will have infection control issues as there is a need for thorough cleaning between participants. All of these barriers limit the usefulness of wearable devices for hand gesture detection in medical and large scale studies.
The second method, the vision-based method, holds much more potential for remote or large scale studies as participants’ hand gestures are captured as image data by video cameras and then processed and analysed through computer vision algorithms [13]. Before the popularity of CNNs in hand gesture recognition, the traditional vision-based approach focuses on extracting image features and then using a classifier to differentiate features into different gestures. Statistical methods were the most widely used. For example, Lee and Kim [14] first introduced Hidden Markove Model (HMM) to calculate likelihood of the detected gestures. Many subsequent efforts have been made to improve the classification performance, for example, IOHMM [15] and the combination of HMM and recurrent neural networks (RNN) [16]. Some work focus on how to extract features effectively, such as stochastic context-free grammar (SCFG) [17].
With the development of deep learning and convolutional neural network (CNN), researchers have been employing CNNs for hand gesture recognition thanks to their ability to learn visual features from images; hence, feature extraction is not required [18]. Real-time hand gesture recognition has benefited greatly from this as many popular object detection and image classification algorithms have been developed recently. They include inception V2 for MITI hand dataset [19], SSD for American Manual Alphabet detection [20], YOLOV3 on a custom hand gesture dataset [21] and Temporal Segment Networks (TSN) [22] for IPN Hand dataset [23]. Many of those approaches have achieved high accuracy, confirming that vision-based hand gesture detection methods can be a reliable method for hand gesture detection problems. Unlike static hand gesture detection, dynamic hand gesture recognition is more challenging because blurriness boundaries of the hand gestures [24]. Deep learning-based methods also show promising results on dynamic gesture recognition, Kopuklu et al. implemented ResNet-10 and achieved 77.39% accuracy on nvGesture Dataset [25]. Do et al. also achieved 96.07% accuracy on a custom dynamic hand gesture dataset by using a ConvLSTM model [26]. However, previous literature shows limited information about similar gesture detection and how real-world hand gesture classification problem were analysed.
Object detection algorithms used in recent years have tended to be either two-stage detectors or one-stage detectors and each has its own advantage: two-stage detectors are good at improving the accuracy of detection and one-stage detectors generally have faster detection speeds. YOLO, which was first introduced by Redmon in 2016 [27], is regarded as one of the most successful one-stage object detectors and has been widely adopted for real-time hand gesture detection. For example, Ni et al. implemented a hand gesture detection system based on YOLOV2 for hand gesture recognition in scenes with a cluttered background [28] and Mujahid et al. proposed a YOLOV3 system for real-time gesture recognition in detecting hand gestures and then denoted numbers from 1 to 5 [21].
The YOLO object detection model has been updated and improved constantly. From the first version through to the latest version, YOLOV5, many techniques such as batch normalisation, anchor boxes, multi-scale training, feature pyramid networks for object detection, mosaic data augmentation and model size reduction have been implemented to improve the performance [8, 29–31]. Nevertheless, despite the successes of YOLO, some researchers have found it lacks capabilities in detecting small objects such as fingers of the hand or small images of pedestrians on the road [32].
In summary, the development of vision-based object detection techniques has dramatically improved both the accuracy and speed of hand gesture detection, but there remains a lack of research into real-world challenges. Two key challenges include how to detect hand gestures in real time when there is motion blur and how to discriminate very similar gestures. These challenges are commonplace in medical and neuroscience applications, especially during the COVID-19 pandemic, when patients and participants are increasingly using their own laptop cameras, or clinic webcams to collect hand data remotely.
Feature extraction network
So far, the detection of hand gestures in real time has relied heavily on feature extraction networks as the backbone for the majority of solutions. Such networks are commonly used to extract deep features from the images. For example, ResNet [33] has been widely adopted as a feature extractor and backbone in many one-stage detectors, including RetinaNet [34] and SSD [35]. ResNet introduced a shortcut connection that guaranteed the gradient would not be vanished. EfficientNet [36] is another popular feature extraction network and this network used a neural architecture search method to explore the optimistic model depth, width and resolution of input images. This provides a way to adaptively scale the model to optimise the computational cost for different computer vision tasks.
In order to extract more useful features, recent approaches have employed transformers to pay attention to the discriminable information in the inputs. Attention was originally designed as a useful tool in natural language processing (NLP), but has shown potential in wider applications. For computer image classification, it enables the neural network to learn the relevant information of the images for the tasks and increase the performance [37].
Rapid gesture recognition net (RGRNet)
Network structure
We proposed a novel network structure called ‘Deep Robust Hand Gesture Network’ (RGRNet), which includes attention mechanisms and multi-scale detection techniques to increase the accuracy for detecting fast-moving hands and for classifying similar gestures. Our proposed network consists of three blocks as shown in Fig. 2: Block 1 is designed as the feature extractor, Block 2 is for feature fusion and Block 3 is for detection.Fig. 2 Rapid gesture recognition net (RGRNet)
Block 1 - Feature extractor
Our network structure was designed as a classic one-stage detector framework. Our feature extractor includes traditional CSPNet structure [38] and attention blocks. In Block 1, the size of the convolution kernel in front of each CSP module is 3x3 with a stride equal to 2. This architecture allows the network to have different sizes of feature maps from top to the bottom.
Block 2 - Feature fusion
In Block 2, feature pyramid networks (FPN) [39] and PAnet (PAN) [40] structure were employed. The FPN can effectively propagate semantic visual information in a bottom-up manner while PAN would enhance the localisation of discriminable features in the lower layer and link them to the top layer. The idea to combine the two different structures would encourage better parameter aggregation between different layers and a more effective fusion of visual features.
Block 3 - Detection
Block 3 inherits the output from the feature fusion block, followed by two layers of convolution. The final outputs of Block 3 will include: (1) bounding boxes and their confidence scores and (2) a SoftMax layer of N, where N is the number of gestures.
Focus layer
The focus layer [31] is the first layer in the feature extraction network. A 608x608 image after the focus layer will be sliced into 4 different parts; then, we use 32 convolution kernels to convert them into 304×304×32 feature maps. This technique was also used by Tal et al. [41] who called it ‘SpaceToDepth’, and stated it allowed the network to rearrange spatial data into depth. Such a component will thus help our network to reduce the resolution of the input images and save computational costs.
Conv layer
The Conv layer is a fundamental building block in our network. In the Conv layer, the input feature maps will go through a 3x3 convolution with strike equals 2, followed by batch normalisation and a SiLU activation function [42].
C3 layer
The C3 layer is a stackable layer in our network, which consists of CBL and multiple CSP units. In this layer, CBL refers to convolution, batch normalisation and Leaky ReLU activation function. In the CSP unit, the input feature map will go through two CBL blocks and then add the previous output to its original feature map becoming the output. This approach ensures that the amount of information under the feature maps of the image is increased while guaranteeing that the dimensions of the feature maps are not increased. This operation will increase the amount of information and is beneficial to the final classification of the image.
SPP layer
In the SPP layer [8], we use four different types of max-pooling to fuse the feature maps. They are 1×1, 5×5, 9×9 and 13×13 max-pooling. This block enables the information from different sizes of feature maps to be combined.
Upsampling layer
The upsampling layer enables the feature map size to be increased. In our network, we adopted the nearest neighbour in the upsampling calculation.
Concat layer
The concat layer refers to an operation that combines the feature maps in different sizes. This layer enable us combine features from different layers and fuse to a new feature.
Attention layers
The core element in our architecture is the attention module, as shown in the green box of Fig. 2. This module consists of several attention layers, including squeeze-and-excitation (SE) layer and transformer layer.
Squeeze-and-excitation (SE) layer
The squeeze-and-excitation (SE) layer enhances the ability of models to learn correlations between visual channels (R–G–B, H–S–V, etc.). The SE-Block was first proposed in SENet [43] and showed better performance than ResNet. Although the SE layer will slightly increase the computation costs, the performance degradation is within acceptable limits, and the loss of SENet is not significant in terms of GFLOPs, the number of parameters and run-time experiments. The architecture of the SE layer is demonstrated in Fig. 3a. The SE layer is normally embedded after a traditional convolution block. Firstly, we use global average pooling to reduce the dimensionality of the feature maps from 3D to 1D, this step can also be referred to as ‘squeeze’.
The squeeze operation is calculated as the following equation :1 Zc=FsqUc=1H×W∑i=1H∑j=1WUci,j
where Fsq refers to the feature map after ‘squeeze’, Uc is the transformation output from the previous layer, H and W are height and width, respectively. By using the average pooling method, all the information contained in this feature map is averaged. ‘Excitation’ is done by two fully connected layers. The first fully connected layer will squeeze the number of channels from C to C/r, where r refers to the reduction ratio. The second fully connected layer is adopted to ensure the feature map can be returned to its original channel size. This attention mechanism allows the network to focus more on the most informative channel features and suppress the less important ones.Fig. 3 SE layer and transformer layer
Transformer layer
The transformer layer is shown in Fig. 3b, which is inspired by the vision transformer. The vision transformer has implemented a skip-connection-based network block and demonstrated having better performance than the state of the art on image classification tasks [37].
In our model, the transformer consists of multilayer perception, two normalisation layers and one multi-head attention. It enables the neural network to extract global information from image features. By adding attention, the global relationship and distinctions between five hand gestures frequently used in the finger tapping and hand opening tasks can be learned from an input image. In the transformer layer, the global information is learnt through the similarity calculation. Vaswani et al. have used Query and Key-Value pairs to represent the similarity of the features [44], where the similarity is calculated as:2 fQ,ki,i=1,2,...,m
The similarity function f is normalised by applying a SoftMax operator, followed by calculating the weighted sum of all the values in V to obtain the attention vector. We can use the following equation to calculate attention in the transformer:3 AttentionQ,K,V=softmaxQTKdkV,
where 1dk is a scaling factor. With multi-head, we are able to perform attention multiple times in parallel (we set 4 heads in our experiment) without sharing parameters. The fusion of the attention and transformer layers enables our network to focus on specific areas for solving particular tasks (hand gesture recognition), rather than evaluating the whole image.
Dataset
UTAS7k dataset—a new dataset for hand gesture detection
Subjects and setting
As part of the TAS Test [5] project, we invited adults aged 50 years and older from an established cohort in Tasmania, Australia (The ISLAND Project) [45] to perform a series of hand movement tests. The TAS Test project aims to track movement and cognitive changes associated with ageing and degenerative brain disorders over a 10-year period. Ethical approval has been granted for the TAS Test study by the Human Research Ethics committee at the University of Tasmania (reference H0021660), and all participants provide informed online consent.
Hand movement data were collected via TAS Test, an online programme, that uses a short demonstration video to instruct participants how to perform each of the hand movement tests and then records their hand movements as they complete the test using a standard laptop camera or desktop webcam (typically with 30 fps). The test is designed to be completed without any in-person researcher assistance or supervision.
So far, 1,900 participants aged between 50 and 90 years have completed a range of hand movement tests, with the majority completing tests in their own home and some in the clinical research facility at the University of Tasmania.
Dataset processing
The hand movements video consisted of a sequence of images (video frames). By stopping at a specific frame in the sequence of a hand movement video frame, we could extract a still image. Our model works by detecting a single frame in the video and then returning the result of the detection to each frame in the video. In total, more than 20,000 image frames of hand gestures were collected. In this dataset, we processed and labelled more than 7,071 images for training, validation and testing. Most data frames were collected through high-definition video (720P) or full high-definition (1080P) web cameras. Their resolutions are 1280 x 720 pixels and 1920 x 1080 pixels, respectively. To unify the size of the input image for our network and accelerate the training speed, we scaled down the image systematically (via FFmpeg) and scaling down time is not calculated; however, the average processing time for 720P video is 7.8ms/frame and 1080P at 9.1ms/frame. The data were split into 4:1 with 5996 clear images and 1075 blurred images and this dataset was named ‘UTAS7k’. Table 1 outlines how the UTAS7k dataset compares to other established hand datasets, and highlights that it has a far larger population size of individual hands (n = 1900 participants) than previous datasets (n < 643 participants). Moreover, we are the only group to have also included data with motion blur.Table 1 Table of datasets for hand gesture detection comparison
Datasets Total images Population Hand side Left–right Age recorded Include motion blur
11k Hands [46] 11,076 190 Palm-dorsal Both Yes No
CASIA [47] 5,502 312 Palm Both No No
Bosphorus [48] 4,846 642 Palm Both No No
llTD [49] 2,601 230 Palm Both No No
GPDS150hand [50] 1,500 150 Palm Right No No
UTAS7k (ours) 7,071 1,900 Palm-dorsal Both Yes Yes
Developing the UTAS7k dataset
Hand gestures
To establish the UTAS7k dataset, 5 different hand gestures were extracted from the fast finger tapping and hand opening–closing hand movement tests and called these ‘open’, ‘close’, ‘pinch open’, ‘pinch close’ and ‘flip’ as shown in Fig. 4.Fig. 4 Five hand gestures in the UTAS7k dataset
Motion blur
Quantifying the blurriness of the images is essential for us to classify and pre-process the training data. Pech-Pacheco et al. proposed a method to calculate the blurriness of the images by calculating the standard deviation of a convolution operation after a Laplace mask [51]. The Laplace mask equation is listed below:4 LAPI=∑nM∑mNLx,y,
where L( m, n) is the convolution of the input image I(m,n) with the mask L and the mask is calculated by the following equation:5 L=160-10-14-10-10
Figure 5 shows the blurriness score after calculation; if an image has a high variance (low blurriness), it means there are many edges in the image as is commonly seen in a normal, accurately focused picture. On the other hand, if the image has a small variance (high blurriness), then there are fewer edges in the image, which is typical of a motion-blurred image.
The variance (blurriness) of 7,071 images was calculated and classified into two categories: they are clear (Fuzzy ≥ 50) and blurred images (Fuzzy < 50), respectively. Table 2 shows the number of clear images and blurred images in each dataset.Fig. 5 Image blurriness calculation. The ‘fuzzy’ score is a number that quantifies the average quality of the image with Fuzzy ≥ 50 indicating the image is clear, Fuzzy score < 50 indicating motion blur
Table 2 Number of images included in the sub-datasets of UTAS7k - split into the clear dataset, motion-blurred dataset and testing dataset
Dataset Total Clear Motion blur
Clear dataset 5,376 5,376 0
Motion-blurred dataset 6,450 5,376 859
Testing dataset 621 497 124
Experiments
Setup
Presetting In our experiment, the performance of the models was compared with state-of-the-art network structures in object detection. The following competitors were tested: DarkNet-53 [30], GhostNet [52], TinyNet [30], CSPDarknet-53 [31], MobileNetv3-small[53], EfficientNet-B1 [36] and RGRNet. (Network structure is shown in Fig. 2.) For fair comparison, all models are implemented using YOLOv5 framework. We then conduct an intensive evaluation of our model using multiple metrics including the number of parameters, as what will be discussed in the next section.
Data preparation 80 per cent of the data were split for training and 20 per cent for validation. The testing dataset was an independent dataset comprising 621 images (of 5 different gestures, 80% clear and 20% motion blurred) that were not seen by the models before. In the default settings, we set the training images as image size 640 × 640 pixels.
Training techniques To improve the quality of the training of all models, several popular data augmentation techniques were implemented, including translate, scale, flip from left to right and mosaic augmentation. This process would enrich the data which is needed for deep learning.
Stochastic gradient descent (SGD) was employed as the optimising function with a decaying learning rate of 0.0005 where the initial learning rate is set as 0.01. Before that, the training process started with very low learning rate for warm-up training to help the models gradually adapt to the data.
Evaluation metrics
mAP To evaluate whether the detection was successful, ‘IOU’ was used to describe the intersection over union (IOU) area between ground truth and predicted bounding boxes. IOU indicates how accurate the bounding boxes are in terms of localising objects. Normally, a threshold t (in our experiments, we set t = 0.65) will be assigned to determine whether the detection is successful, i.e. if the IOU of a detected bounding box is larger than t then it will be accepted as a true bounding box; otherwise, it will be classified as incorrect. [email protected]:0.95 was adopted as our primary evaluation metric for detection accuracy in our experiments. [email protected]:0.95 is the primary evaluation metric from the MS COCO challenge [54] and denotes the mAP at different thresholds (from 0.5 to 0.95 in steps of 0.05), which is calculated by the following equation:6 [email protected]:0.95=([email protected][email protected]+…[email protected])/10
GigaFLOPS (GFLOPs) We employed Giga floating point operations per second (GFLOPs) to evaluate the computational cost. Generally, the more GFLOPs a model has, the greater the cost of the computer to run the model.
Parameters The number of parameters often determines the learning capacity of a model, the more parameters a model has, the more learning capacity it poses. The unit for this evaluation metric is ‘M’, meaning a million parameters.
Storage Size The storage size evaluates the amount of space for the model to be stored in the computer. The unit for storage size is Megabyte (MB).
Inference speed per image Inference speed per image evaluate the inference time for the model to process an image with a 640 × 640 pixel image size. We used milliseconds (ms) as the unit.
Experimental results
Performance analysis on datasets comprising clear images of hand gestures
The network comparisons are displayed in Table 3, where we evaluate different network structures performed on the testing dataset and the fastest inference speed and highest accuracy are highlighted. Our network RGRNet had a mean average precision of 0.782 which was superior to all other network structures on the clear images of hand gesture dataset. The end-to-end image processing speed at 720P image size is 17.5 ms (57FPS) and 18.8 ms (57FPS) at 1080P image size, which is still within the range of near real-time detection (≥30FPS). Although the processing speed was longer at 9.7ms per image, as a one-stage detector, the inference time is already within the range of real-time detection. Adding a multi-scale detection and attention layer will increase the parameters of the network considerably. EfficientNet is one of the SOTA convolutional neural networks by setting certain parameter values to balance the depth, width and input image size of the convolutional neural network. We applied efficientNet-B1 to our data and found that it can achieve a decent result, 0.757 mAP and GFLOP (6.7) with only 9.98 M parameters. CSPNet was designed to minimise duplicate gradient information within the network and reduce the complexity of the network. In our experiment, CSPDarknet-53 also shows effectiveness on hand gesture classification by achieving 0.753 mAP and 7.5 ms image inference speed at an image size 640×640. We have taken the extra step of adding in a transformer layer and an SE layer and achieved 0.782 mAP and 9.7 ms inference speed, which is significantly higher than our baseline models, MobileNetv3-Small (0.701), EfficientNet-B1 (0.757) and CSPNet-53 (0.753). Table 3 Network structure comparison on testing dataset (image tested on NVIDIA GTX 1660 SUPER)
Network structure Parameters GFLOPs Storage size Inference speed [email protected]:0.95
GhostNet 4.17 M 9.2 7.8 MB 7.4 ms 0.735
CSPDarknet-53 7.27 M 16.9 14.00 MB 7.5 ms 0.753
MobileNetv3-Small 3.55 M 6.3 4.21 MB 3.0 ms 0.701
TinyNet 2.18 M 3.3 5.85 MB 3.1 ms 0.703
Darknet-53 6.99 M 19.0 14.20 MB 7.4 ms 0.755
EfficientNet-B1 9.98 M 6.7 19.40 MB 9.5 ms 0.757
YOLOV5-P6 12.36 M 16.7 25.1 MB 9.5 ms 0.776
RGRNet (ours) 12.54 M 17.0 25.5 MB 9.7 ms 0.782
According to the experiment results, we categorised network structures into three types: medium structures, large structures and light structures. Large network structures, such as CSPDarknet-53, EfficientNet-B1 and Darknet-53, had similar detection accuracy on our dataset and their weight sizes ranged from 14MB to 20MB. Medium network structures usually have a smaller weights size because some have adopted ghost convolution which reduces the number of parameters and kernel sizes in the feature extraction blocks. However, this type of network usually has lower detection accuracy too. Although we assumed smaller models would have a faster detection speed, the image process speed is surprisingly similar for CSPDarknet-53 on GTX 1660 SUPER GPU, in comparison with other lighter models. Finally, light network structures, including TinyNet and MobileNet, showed efficient inference computation with reasonable detection speed; however, their performance in terms of mAP is not promising.
Performance analysis on the noisy dataset comprising clear and blurred images (4:1) of hand gestures
In this experiment, we included the blurred images in the motion-blurred dataset to create a dataset with 80% clear images and 20% blurred images. As shown in Table 4, the performances of all network structures dropped, we have also highlighted the highest detection accuracy, lowest accuracy drop and fastest inference speed in bold text. The result shows our model still achieved better mAP than the other baselines.Table 4 Network performance with noisy dataset (image tested on GTX 1660 SUPER)
Network structure mAP without noise mAP with noise mAP dropped Process speed per image
CSPDarknet(YOLOV5) 0.753 0.745 0.008 7.4ms
GhostNet 0.735 0.726 0.009 7.5ms
MobileNetv3 0.701 0.694 0.007 3.0ms
Darknet-53 0.755 0.747 0.008 7.4ms
EfficientNet-B1 0.757 0.745 0.012 9.5ms
TinyNet 0.703 0.701 0.002 3.1ms
CSPDarknet-P6 0.776 0.769 0.007 9.5ms
RGRNet (ours) 0.782 0.771 0.011 9.7ms
Overall, we found that complex network structures had a higher drop in mAP; see Fig. 6. The cause of such drop is mainly due to the higher number of layers. We can see that, with more layers and parameters, these models usually have more learning capacity that encourages the negative effect of noisy data to be amplified during the learning. Although the number parameters of our model, RGRNet, are the greatest among all the models tested, the amount of accuracy dropped is still only as much as a medium-sized model.Fig. 6 Parameters vs mAP dropped on motion-blurred dataset
Why attention layers would work?
In the learning process of the neural network, the network generates different features to cover different semantic information. Xie et al. found that more information is beneficial to the training of the neural network [55]. By introducing an attention mechanism, our network will learn how to better capture finger-specific attention information, thus helping the network to effectively distinguish different gestures. On the other hand, transformer layers enable deep neural networks to obtain global information. The transformer layer and attention layer used in our network thus make our network able to learn both the local feature and the global feature, thereby providing greater effectiveness on similar hand gesture detection tasks. To analyse the effectiveness of our attention layers, we have printed out the attention map in Fig. 7 to highlight the important regions in the image for the detection of two similar gestures. We can see that after adding the attention layer, our model focuses on the recognition of the gesture as a whole in the detection of similar gestures, and takes into account the fingertip part in the detection of both pinch open and pinch close gestures. More importantly, we have added an extra step to analyse how attention layers would impact detection accuracy in Table 5, in bold text, we have highlighted that SELayer achieved the highest precision, recall, [email protected] and [email protected]:0.95. Our result shows attention layers can help improve the performance where the SE layer performs better than transformer on motion-blurred dataset, which means learning local features enables our network to perform better on blurred images.Fig. 7 Attention map for the models to predict similar gestures (pinch open and pinch close gestures)
Table 5 Comparison of how different attention mechanisms performed on the noisy dataset (80% clear and 20% blurred images) (image tested on GTX 1660 SUPER) classification accuracy of similar gestures
Network structure Precision Recall mAP 0.5 mAP 0.5:0.95
YOLOV5-P6 0.915 0.902 0.887 0.761
YOLOV5-P6+Transformer 0.921 0.9 0.891 0.759
YOLOV5-P6 +SE layer 0.927 0.914 0.903 0.765
Similar gesture classification
In our experiment, we have analysed the performance of different networks to detect the two similar gestures on the clear dataset, ‘pinch open’ and ‘pinch close’; see Fig. 8. In Table 6, we show how different network structures performed on classifying 5 different gestures in the UTAS7k dataset, where ‘all classes’ evaluates the average detection accuracy for all gestures. In general, all models perform well on classifying ‘open’ gestures and have relatively lower detection accuracies on ‘pinch open’ gestures.
We also found that in many cases, most of the neural networks had mistaken ‘pinch close’ for ‘pinch open’. As the result, there is a higher detection accuracy for ‘pinch close’ than for ‘pinch open’, but generally this accuracy would not be reliable enough for most medical and neuroscience research applications as a balance error rate of these two gestures is desirable.Fig. 8 Visualisation result from CSPDarknet-53 and our network, RGRNet. Note that CSPDarknet-53 incorrectly detects the fingers as ‘pinch open’ when the gesture is ‘pinch close’, see supplementary video 1
Table 6 Comparison of how accurately different networks detect UTAS7k gestures (image tested on GTX 1660 SUPER)
Network structure All classes Open Close Flip Pinch open Pinch close
GhostNet 0.735 0.851 0.770 0.769 0.511 0.775
CSPDarknet-53 0.753 0.876 0.788 0.796 0.541 0.761
MobileNetv3-small 0.701 0.774 0.714 0.714 0.57 0.73
TinyNet 0.703 0.800 0.825 0.711 0.542 0.739
Darknet-53 0.755 0.858 0.766 0.774 0.612 0.766
EfficientNet-B1 0.757 0.857 0.773 0.785 0.599 0.772
YOLOV5-P6 0.776 0.863 0.796 0.769 0.627 0.827
RGRNet (ours) 0.782 0.861 0.783 0.763 0.685 0.820
We can also see that adding the transformer block not only increases the overall gesture detection performance but also significantly increases the model’s detection efficiency for the ‘pinch open’ gesture, which reduces the probability of the model misclassifying similar gestures. Moreover, results show that adding attention layers increase the total mAP for similar gestures and decreased the mAP difference between similar gestures, which also indicates that the model is more balanced in detecting similar gestures and minimise the possibility of misclassification.
Conclusions and future work
In this paper, we have implemented a novel network structure, RGRNet, for accurately classifying similar hand gestures and for motion-blurred image detection. Although previous methods had achieved real-time hand gesture detection, there had not been any focus on real-world fast-moving hand gesture detection in the home and clinical settings with cluttered backgrounds and ambient lighting, nor on hand detection when motion blur is present or when similar gestures are present in those blurred images. We have also developed UTAS7k, a new dataset of 7071 images (videos) with the widest variety of individual hands from 1,900 older adults and including 4: 1 clear:motion-blurred images.
We compared the detection performance of different network structures on classifying similar gesture and motion-blurred gestures, where we found multiple scale detection is effective. More importantly, our method RGRNet achieved optimising results on both similar gesture and motion-blurred gesture classification. Our assessment of a range of networks, including our new network, on these images makes a significant research contribution with a range of real-life applications. Our new dataset UTAS7k provides an important resource for the study of motion blur on hand gesture detection.
In this paper, we have shown attention mechanism is effective in classifying motion-blurred hand gestures and similar gestures. We have only used one attention module, the squeeze-and-excitation block, in our experiments, and it remains to be seen whether other attention modules will give better performance. We have used several strategies to improve the accuracy of the model and have succeeded in improving the classification accuracy for similar gestures. Essentially, we are sacrificing a portion of the speed of detection to improve performance accuracy, but still maintain real-time efficiency. Moreover, the implementation of the transformer block requires additional computing resources in the training process.
The proposed network can be embedded into a user–computer interface for clinical and neuroscience applications. It can be used for detecting different hand gestures in the hand movement tests performed by older adults, which increase the robustness of the data collection process. For future work, we will improve our model with a de-blurring attention mechanism and analyse how high resolution images would impact the inference speed of hand gesture detection. We will also investigate how complex background such as human skin-like background would impact the performance of hand gesture detection.
Supplementary information We have also included a video file named ‘Supplementary Video 1’ as the accompanying supplementary file, this file illustrates how our model detect all five gestures in UTAS7k in real time.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file 1 (mp4 8104 KB)
Acknowledgements
We would like to thank all the participants of The ISLAND Project who provided the video data for this study. We would also like to thank Professor James Vickers and all the staff at the University of Tasmania who work on The ISLAND Project for their support. We acknowledge the funding contributions for this project from the Medical Research Future Fund, St Lukes Health, Tasmanian Masonic Medical Research Foundation and the J.O. and J.R. Wicking Trust (Equity Trustees). We acknowledge funding from the National Health and Medical Research Council for the TAS Test Project.
Data availability
The datasets generated during and/or analysed during the current study are available from the authors on reasonable request.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Alex K, Sutskever I, Hinton GE Imagenet classification with deep convolutional networks. In: NIPS’12 Proceedings of the 25th international conference on neural information processing systems, Vol. 1; pp. 1097–1105
2. LeCun Y Boser B Denker JS Henderson D Howard RE Hubbard W Jackel LD Backpropagation applied to handwritten zip code recognition Neural Comput 1989 1 4 541 551 10.1162/neco.1989.1.4.541
3. Al-Hammadi M Muhammad G Abdul W Alsulaiman M Bencherif MA Mekhtiche MA Hand gesture recognition for sign language using 3dcnn IEEE Access 2020 8 79491 79509 10.1109/ACCESS.2020.2990434
4. Zadikoff C Lang AE Apraxia in movement disorders Brain 2005 128 7 1480 1497 10.1093/brain/awh560 15930045
5. Alty J Bai Q Li R Lawler K St George RJ Hill E Bindoff A Garg S Wang X Huang G The TAS Test project: a prospective longitudinal validation of new online motor-cognitive tests to detect preclinical alzheimer’s disease and estimate 5-year risks of cognitive decline and dementia BMC Neurol 2022 22 1 1 13 10.1186/s12883-022-02772-5 34979972
6. Alty J Bai Q George RJS Bindoff A Li R Lawler K Hill E Garg S Bartlett L King AE Vickers JC Tastest: moving towards a digital screening test for pre-clinical Alzheimer’s disease Alzheimer’s Dementia 2021 17 S5 058732 10.1002/alz.058732
7. Goetz CG Fahn S Martinez-Martin P Poewe W Sampaio C Stebbins GT Stern MB Tilley BC Dodel R Dubois B Movement disorder society-sponsored revision of the unified Parkinson’s disease rating scale (mds-updrs): process, format, and clinimetric testing plan Movement Disorders 2007 22 1 41 47 10.1002/mds.21198 17115387
8. Bochkovskiy A, Wang C-Y, Liao H-YM (2020) Yolov4: optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934
9. Lee M Bae J Deep learning based real-time recognition of dynamic finger gestures using a data glove IEEE Access 2020 8 219923 219933 10.1109/ACCESS.2020.3039401
10. Jung P-G Lim G Kim S Kong K A wearable gesture recognition device for detecting muscular activities based on air-pressure sensors IEEE Trans Ind Inf 2015 11 2 485 494
11. Premaratne P Historical development of hand gesture recognition 2014 Cham Springer 5 29
12. Ahmed M, Zaidan B, Zaidan A, Alamoodi A, Albahri O, Al-Qaysi Z, Albahri A, Salih MM (2021) Real-time sign language framework based on wearable device: analysis of msl, dataglove, and gesture recognition. Soft Comput, 1–22
13. Zhu Y, Yang Z, Yuan B (2013) Vision based hand gesture recognition. In: 2013 international conference on service sciences (ICSS), pp. 260–265. IEEE
14. Lee H-K Kim J-H An hmm-based threshold model approach for gesture recognition IEEE Trans Pattern Anal Mach Intell 1999 21 10 961 973 10.1109/34.799904
15. Marcel S, Bernier O, Viallet J-E, Collobert D (2000) Hand gesture recognition using input-output hidden Markov models. In: proceedings fourth IEEE international conference on automatic face and gesture recognition (Cat. No. PR00580), pp. 456–461. IEEE
16. Ng CW Ranganath S Real-time gesture recognition system and application Image Vis Comput 2002 20 13–14 993 1007 10.1016/S0262-8856(02)00113-0
17. Chen Q Georganas ND Petriu EM Hand gesture recognition using haar-like features and a stochastic context-free grammar IEEE Trans Instrum Meas 2008 57 8 1562 1571 10.1109/TIM.2008.922070
18. Mohanty A, Rambhatla SS, Sahay RR (2017) Deep gesture: static hand gesture recognition using CNN. In: proceedings of international conference on computer vision and image processing, pp. 449–461. Springer
19. Bose SR Kumar VS Efficient inception v2 based deep convolutional neural network for real-time hand action recognition IET Image Process 2020 14 4 688 696 10.1049/iet-ipr.2019.0985
20. Yi C, Zhou L, Wang Z, Sun Z, Tan C (2018) Long-range hand gesture recognition with joint ssd network. In: 2018 IEEE international conference on robotics and biomimetics (ROBIO), pp. 1959–1963. IEEE
21. Mujahid A Awan MJ Yasin A Mohammed MA Damaševičius R Maskeliūnas R Abdulkareem KH Real-time hand gesture recognition based on deep learning yolov3 model Appl Sci 2021 11 9 4164 10.3390/app11094164
22. Benitez-Garcia G Prudente-Tixteco L Castro-Madrid LC Toscano-Medina R Olivares-Mercado J Sanchez-Perez G Villalba LJG Improving real-time hand gesture recognition with semantic segmentation Sensors 2021 21 2 356 10.3390/s21020356 33430214
23. Benitez-Garcia G, Olivares-Mercado J, Sanchez-Perez G, Yanai K (2021) IPN hand: a video dataset and benchmark for real-time continuous hand gesture recognition. In: 2020 25th international conference on pattern recognition (ICPR), pp. 4340–4347. IEEE
24. Gupta P, Kautz K, et al (2016) Online detection and classification of dynamic hand gestures with recurrent 3d convolutional neural networks. In: CVPR, vol 1, p. 3
25. Köpüklü O, Gunduz A, Kose N, Rigoll G (2019) Real-time hand gesture detection and classification using convolutional neural networks. In: 2019 14th IEEE international conference on automatic face & gesture recognition (FG 2019), pp. 1–8. IEEE
26. Do N-T Kim S-H Yang H-J Lee G-S Robust hand shape features for dynamic hand gesture recognition using multi-level feature lstm Appl Sci 2020 10 18 6293 10.3390/app10186293
27. Redmon J, Divvala S, Girshick R, Farhadi A (2016) You only look once: unified, real-time object detection. In: proceedings of the IEEE conference on computer vision and pattern recognition, pp. 779–788
28. Ni Z, Chen J, Sang N, Gao C, Liu L (2018) Light yolo for high-speed gesture recognition. In: 2018 25th IEEE international conference on image processing (ICIP), pp. 3099–3103. IEEE
29. Redmon J, Farhadi A (2017) Yolo9000: better, faster, stronger. In: proceedings of the IEEE conference on computer vision and pattern recognition, pp. 7263–7271
30. Redmon J, Farhadi A (2018) Yolov3: an incremental improvement. arXiv preprint arXiv:1804.02767
31. Jocher G, et al. (2021) ultralytics/yolov5: V5.0 - YOLOv5-P6 1280 Models, AWS, Supervise.ly and YouTube integrations. 10.5281/zenodo.4679653
32. Xianbao C, Guihua Q, Yu J, Zhaomin Z (2021) An improved small object detection method based on yolo v3. Pattern Anal Appl 1–9
33. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778
34. Ross T-Y, Dollár G (2017) Focal loss for dense object detection. In: proceedings of the IEEE conference on computer vision and pattern recognition, pp. 2980–2988
35. Liu W, Anguelov D, Erhan D, Szegedy C, Reed S, Fu C-Y, Berg AC (2016) SSD: single shot multibox detector. In: European conference on computer vision, pp. 21–37. Springer
36. Tan M, Le Q (2019) Efficientnet: rethinking model scaling for convolutional neural networks. In: international conference on machine learning, pp. 6105–6114. PMLR
37. Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X, Unterthiner T, Dehghani M, Minderer M, Heigold G, Gelly S, et al (2020) An image is worth 16x16 words: transformers for image recognition at scale. arXiv preprint arXiv:2010.11929
38. Wang C-Y, Liao H-YM, Wu Y-H, Chen P-Y, Hsieh J-W, Yeh I-H (2020) CSPNet: a new backbone that can enhance learning capability of CNN. In: proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, pp. 390–391
39. Lin T-Y, Dollár P, Girshick R, He K, Hariharan B, Belongie S (2017) Feature pyramid networks for object detection. In: proceedings of the IEEE conference on computer vision and pattern recognition, pp. 2117–2125
40. Wang K, Liew JH, Zou Y, Zhou D, Feng J (2019) Panet: few-shot image semantic segmentation with prototype alignment. In: proceedings of the IEEE/CVF international conference on computer vision, pp. 9197–9206
41. Ridnik T, Lawen H, Noy A, Ben Baruch E, Sharir G, Friedman I (2021) TRESNet: high performance GPU-dedicated architecture. In: proceedings of the IEEE/CVF winter conference on applications of computer vision, pp. 1400–1409
42. Elfwing S Uchibe E Doya K Sigmoid-weighted linear units for neural network function approximation in reinforcement learning Neural Netw 2018 107 3 11 10.1016/j.neunet.2017.12.012 29395652
43. Hu J, Shen L, Sun G (2018) Squeeze-and-excitation networks. In: proceedings of the IEEE conference on computer vision and pattern recognition, pp. 7132–7141
44. Vaswani A Shazeer N Parmar N Uszkoreit J Jones L Gomez AN Kaiser Ł Polosukhin I Attention is all you need Adv Neural Inf Process Syst 2017 30 5998 6008
45. Bartlett L Doherty K Farrow M Kim S Hill E King A Alty J Eccleston C Kitsos A Bindoff A Island study linking aging and neurodegenerative disease (island) targeting dementia risk reduction: protocol for a prospective web-based cohort study JMIR Res Protoc 2022 11 3 34688 10.2196/34688
46. Afifi M 11k hands: gender recognition and biometric identification using a large dataset of hand images Multimed Tools Appl 2019 10.1007/s11042-019-7424-8
47. Sun Z, Tan T, Wang Y, Li S (2005) Ordinal palmprint representation for personal identification. In: proceedings of the IEEE conference on computer vision and pattern recognition
48. Abdesselam A, Al-Busaidi A (2012) Person identification prototype using hand geometry. 10.13140/2.1.2181.9844
49. Kumar A (2008) Incorporating cohort information for reliable palmprint authentication. In: 2008 Sixth Indian conference on computer vision, graphics & image processing, pp. 583–590. IEEE
50. Ferrer MA, Morales A, Travieso CM, Alonso JB (2007) Low cost multimodal biometric identification system based on hand geometry, palm and finger print texture. In: 2007 41st annual IEEE international Carnahan conference on security technology, pp. 52–58. IEEE
51. Pech-Pacheco JL, Cristóbal G, Chamorro-Martinez J, Fernández-Valdivia J (2000) Diatom autofocusing in brightfield microscopy: a comparative study. In: proceedings 15th international conference on pattern recognition. ICPR-2000, vol. 3, pp. 314–317. IEEE
52. Han K, Wang Y, Tian Q, Guo J, Xu C, Xu C (2020) GhostNet: more features from cheap operations
53. Howard A, Sandler M, Chu G, Chen L-C, Chen B, Tan M, Wang W, Zhu Y, Pang R, Vasudevan V, Le QV, Adam H (2019) Searching for MobileNetV3
54. Lin T-Y, Maire M, Belongie S, Hays J, Perona P, Ramanan D, Dollár P, Zitnick CL (2014) Microsoft coco: common objects in context. In: European conference on computer vision, pp. 740–755. Springer
55. Xie T Deng J Cheng X Liu M Wang X Liu M Feature mining: a novel training strategy for convolutional neural network Appl Sci 2022 12 7 3318 10.3390/app12073318
| 0 | PMC9741488 | NO-CC CODE | 2022-12-14 23:29:55 | no | Neural Comput Appl. 2022 Dec 10;:1-14 | utf-8 | Neural Comput Appl | 2,022 | 10.1007/s00521-022-08090-8 | oa_other |
==== Front
Public Health Pract (Oxf)
Public Health Pract (Oxf)
Public Health in Practice
2666-5352
The Authors. Published by Elsevier Ltd on behalf of The Royal Society for Public Health.
S2666-5352(22)00126-4
10.1016/j.puhip.2022.100350
100350
Article
Differences in case-fatality-rate of emerging SARS-CoV-2 variants
Liu Jing a
Wei Haozhen a
He Daihai ab∗
a Department of Applied Mathematics, Hong Kong Polytechnic University, Hong Kong, China
b Research Institute for Future Food, The Hong Kong Polytechnic University, Hong Kong, China
∗ Corresponding author. Department of Applied Mathematics, Hong Kong Polytechnic University, Hong Kong, China.
10 12 2022
6 2023
10 12 2022
5 100350100350
16 9 2022
29 11 2022
5 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.
Objects
Variants of Severe-Acute-Respiratory-Syndrome Coronavirus-2 (SARS-CoV-2) has caused tremendous impact globally. It has been widely reported that the Omicron (B.1.1.529) variant is less deadly than the Delta (B.1.617.2) variant, presumably due to immunity from vaccination and previous infection. When measuring the severity of a variant, Case-Fatality-Rate (CFR) is often estimated. The purpose of this work is to calculate the change in CFR of different variants over time from a large number of countries/regions since the start of the pandemic in 2020.
Study design
A Cross-sectional study.
Methods
We extend the comparison to all previous VOCs in 58 counties/regions. We use reported death divided by reported cases in 30-day sliding window with a two-week shift between reported death and reported cases.
Results
The drop from Delta variant to Omicron variant is substantial and the difference between subvariants of Omicron is not evident.
Conclusion
We showed that the CFR dropped over time, presumably due to vaccine-induced immune and infection induced immune. Population age structure and prevalence of comorbidity influence CFR.
Keywords
COVID-19
Omicron variant
Delta variant
CFR
==== Body
pmc1 Background
The Coronavirus Disease 2019 (COVID-19) pandemic, caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), hit the humankind tremendously with 544,946,610 reported cases and 6,342,015 reported deaths by June,2022 [1,2]. The rapid evaluation of the viruses led to variants with high transmissibility and high immune evasion ability and posed challenges for prevention and control [3]. Among the new variants, the Delta variant (B.1.617.2) was first identified in December 2020 [4] and classified by the World Health Organization (WHO) as Variant of Concern (VOC) on June 2021, updated as Variant Being Monitored (VBM) on April 2022 [3]. The Omicron variant (B.1.1.529) was first detected in specimens collected on November 2021, and classified as a Variant of Concern (VOC) [5]. Several subvariants of Omicron variant has been reported so far. Previous studies on the case-fatality-rate (CFR) of different variants were conducted in individual countries. Large-scale comparison across multiple countries is missing. The CFR is an important indicator of disease severity. Many factors may affect the estimates. Infection attack rate (proportion of population being infected) with previous variants, vaccination coverage, age structure of the population and the medical system preparedness may affect the CFR in different population/locations. In this work, we combine the reported cases and deaths and variant proportion over time in 58 locations (see Appendix) to calculate the raw CFR for 8 variants (or subvariants), thus, to get a large-scale picture of the changing pattern of the severity of SARS-CoV-2.
The Omicron variant spreads faster than the Delta variant due to the combined effects of increased transmissibility and immune evasion ability. In South Africa, the proportion of patients with Omicron infection presenting to emergency departments has fallen to half its previous level, and the proportion of Omicron patients presenting with acute respiratory conditions and requiring oxygen therapy and mechanical ventilation has fallen dramatically [6].
In United States, the number of deaths in the Omicron wave (analyzed from December 15, 2021 to March 15, 2022) was very similar to that seen in the Delta wave (analyzed from July 15, 2021 to November 15, 2021). However, the number of confirmed cumulative cases during this period was twofold higher with Delta. The CFR of Omicron variant is about half that of Delta variant in the both US and South Africa [7].
2 CFR of variants across 58 locations
2.1 Study design
A Cross-sectional study of Case-Fatality-Rate during the epidemic in each variant was conducted in 58 countries/regions, with controls for age group and vaccination status. We compared and calculated the relative difference in CFR between infection with Delta and Omicron variants under different age groups and the efficacy of vaccination in reducing CFR under Delta or Omicron variant prevalence time breaks.
2.2 Data collection
The data collected were incremental confirmations of different variants in 58 countries or regions on a two-week cycle. We selected the countries with relatively well-developed data among them to be collected in order of the total number of confirmed diagnoses, from most to least.
2.3 Method
Here we extend the comparison to all previous VOCs in 58 counties/regions. We use reported death divided by reported cases in 30-day sliding window with a two-week shift between reported death and reported cases.CaseFatalityRatio(CFRin%)=Numberofdeathsfromdiseaseinthepast30daysNumberofconfirmedcasesofdiseaseinthepast15to45days×100%
Based on the biweekly variant proportion data [8,9], we determine the dominant time interval for each variant in each location when the proportion of a variant is above 60% among all samples processed. Namely the raw CFR in a time interval when the proportion of variant is above 60% is assigned as the CFR for the variant in that location. We calculated the mean CFR values for each variant in its dominant time interval in each location. Fig. 1 summarized the mean CFR for each variant across 58 locations.Fig. 1 The summary of the mean case-fatality rate (CFR) during its dominant time interval in 58 locations for ten variants or subvariants.
Fig. 1
2.4 Statistical analysis
We conduct a two-sample t-test to determine whether the mean CFR values of one variant is different to that of another variant (subvariant). The null hypothesis H0: μ1=μ2, the alternative hypothesis H1: μ1≠μ2 (μi is the mean CFR value of different variants).
The mean CFR values for Delta variant was higher than that of Omicron variant (p-value = 0.01531, rejected the null hypothesis.) (Fig. 1). The mean CFR values was similar between subvariants of Omicron (p-value = 0.581, accept the null hypothesis.) (Fig. 1).
3 Age stratified CFR
In South Africa, the CFR during Omicron variant dominant time interval (November 21, 2021, to January 22, 2022) was substantially lower than that during Delta variant dominant time interval (May 9 to September 18, 2021), for 20+ age groups (see Table 1 ) [10]. This could be due to implementation of vaccination among adults and infection induced immunity across all ages.Table 1 The summary of case-fatality rate (CFR) of Delta variant and Omicron variant among different age groups in South Africa.
Table 1CFR Delta Omicron Relative difference
All ages 2.60% 0.78% 70%
Age 20–39 years 0.45% 0.24% 46.7%
Age 40–59 years 2.54% 0.64% 74.4%
Age ≥60 years 11.71% 2.38% 79.7%
In Table 1, the relative difference is defined as (1−CFR_omicronCFR_delta)%. For all ages, the CFR of Delta variant is more than three-fold of the CFR of Omicron for overall age group and 40+ ages groups. The 5–19 age group has the lowest CFR for both variants.
Italy has a high proportion of elderly people (>65 age accounting 23% of the population) in 2019, which could have led to a higher CFR compared to other countries. Among 70–79 age group, the CFR is about 12.8% in Italy and 8.0% in China; among 80+ age group, the CFR is about 20.2% in Italy and 14.8% in China, in early 2020 [11]. 86% of patients in the Washington ICU have underlying chronic conditions such as kidney disease and heart failure [12]. Population age structure and prevalence of comorbidity influence CFR.
4 Impact of vaccination
Zhao et al. showed that, in the United Kingdom during May and June 2021, Delta variant had a smaller CFR than pre-variant and the CFR of pre-Delta variant dropped substantially while the vaccine coverage increased and the drop of the CFR of Delta variant is less evident (probably due to short study period) [13]. In the United States, Johnson et al. compared the CFR among unvaccinated and fully vaccinated individuals in 25 US jurisdictions and the vaccine efficacy substantially dropped over time (see Table 2 ) [14].Table 2 The summary of case-fatality rate (CFR) of Delta variant and Omicron variant in unvaccinated people and fully vaccinated people in United States.
Table 2Time CFR among Unvaccinated CFR among Fully vaccinated Vaccine Efficacy in reduction CFR
Pre-Delta (2021, 4–5) 1.05% 0.10% 90.5%
Delta emergence (2021, 6) 1.37% 0.24% 82.5%
Delta predominance (2021, 7–11) 1.22% 0.32% 73.8%
Omicron emergence (2021, 12) 0.11% 0.03% 72.7%
The vaccine efficacy (VE) is defined as (1−CFR_vaccinatedCFR_unvaccinated)%. The VE drop substantially over time, due to a combined effect of natural waning of immune protection and immune evasion of Delta and Omicron variant.
5 Limitation
The work utilized bi-weekly variant sequencing data and reported deaths and cases. The under reporting of COVID-19 deaths and cases will impact the estimate of CFR. The CFR is an overestimate of the true infection-fatality-rate. If the reporting is consistent, the CFR across variants should be a fair comparison.
6 Conclusion
In this work, we compared the CFR for ten variants (including ancestral strain and Omicron subvariants) across 58 locations. We showed that the CFR dropped over time, presumably due to vaccine-induced immune and infection induced immune. The drop from Delta variant to Omicron variant is substantial while the difference between subvariants of Omicron is not evident.
Ethical approval statement and consent to participate
The data used in this study were collected originally via the public domains, and thus neither ethical approval nor individual consent was applicable.
Availability of materials
Data are publicly available online.
Consent for publication
Not applicable.
Funding source
We were supported by the Collaborative Research Fund (Grant Number C7123-20G) of the Research Grants Council (RGC) of Hong Kong, China, two projects of Otto Poon Charitable Foundation (Q-CDBA and Q-CDAV) and one project of Research Institute for Future Food (1-CD52).
Disclaimer
The funding agencies had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
Author's contributions
DH and JL conceived the study, carried out the analysis and drafted the manuscript. All authors discussed the results, and revised the manuscript, and approved it for publishing.
Declaration of competing interest
All authors have no conflict of interest.
Appendix List of countries/regions used in the study
Argentina Australia Bangladesh Belgium
Brazil Bulgaria Cambodia Canada
Chile Colombia Costa Rica Croatia
Czechia Denmark Ecuador France
Georgia Germany Greece Hong Kong, China
India Indonesia Ireland Israel
Italy Japan Jordan Lithuania
Luxembourg Malaysia Mexico Netherlands
Norway Pakistan Peru Poland
Portugal Qatar Russia Singapore
Slovakia Slovenia South Africa South Korea
Spain Sweden Switzerland Turkey
Ukraine United Kingdom United States Philippines
Romania Thailand Egypt Austria
Latvia Paraguay
Acknowledgments
None.
==== Refs
References
1 Gorbalenya A.E. Baker S.C. Baric R.S. de Groot R.J. Drosten C. Gulyaeva A.A. Haagmans B.L. Lauber C. Leontovich A.M. Neuman B.W. Penzar D. Perlman S. Poon L.L.M. Samborskiy D.V. Sidorov I.A. Sola I. Ziebuhr J. amp; Coronaviridae Study Group of the International Committee on Taxonomy of, V The species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2 Nature Microbiology 5 4 2020 536 544 10.1038/s41564-020-0695-z
2 The Worldometers COVID-19 Coronavirus Pandemic 2022 https://www.worldometers.info/coronavirus/
3 Centers for Disease Control and Prevention SARS- CoV-2 Variant Classifications and Definitions 2022 https://www.cdc.gov/coronavirus/2019-ncov/variants/variant-info.html
4 Yang W. Shaman J. COVID-19 Pandemic Dynamics in India and Impact of the SARS-CoV-2 Delta (B.1.617.2) Variant 2021 medRxiv 10.1101/2021.06.21.21259268 2021.2006.2021.21259268
5 Centers for Disease Control and Prevention Omicron Variant What You Need to Know 2021 https://www.cdc.gov/coronavirus/2019-ncov/variants/omicron-variant.html
6 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 327 6 2022 583 584 10.1001/jama.2021.24868 34967859
7 Sigal A. Milo R. Jassat W. Estimating disease severity of Omicron and Delta SARS-CoV-2 infections Nat. Rev. Immunol. 22 5 2022 267 269 10.1038/s41577-022-00720-5 35414124
8 EB H. CoVariants: SARS-CoV-2 mutations and variants of interest https://covariants.org/ 2021
9 Hannah Ritchie E.M. Lucas Rodés-Guirao Cameron Appel Charlie Giattino Esteban Ortiz-Ospina Joe Hasell Bobbie Macdonald Diana Beltekian Max Roser Coronavirus Pandemic (COVID-19). Our World in Data 2020
10 Jassat W. Abdool Karim S.S. Mudara C. Welch R. Ozougwu L. Groome M.J. Govender N. von Gottberg A. Wolter N. Wolmarans M. Rousseau P. group D.a. Blumberg L. Cohen C. Clinical severity of COVID-19 patients admitted to hospitals during the Omicron wave in South Africa medRxiv 2022 10.1101/2022.02.22.21268475 2022.2002.2022.21268475
11 Onder G. Rezza G. Brusaferro S. Case-fatality rate and characteristics of patients dying in relation to COVID-19 in Italy JAMA 323 18 2020 1775 1776 10.1001/jama.2020.4683 32203977
12 Arentz M. Yim E. Klaff L. Lokhandwala S. Riedo F.X. Chong M. Lee M. Characteristics and outcomes of 21 critically ill patients with COVID-19 in Washington state JAMA 323 16 2020 1612 1614 10.1001/jama.2020.4326 32191259
13 Zhao S. Lou J. Cao L. Chong K.C. Zee B.C.Y. Chan P.K.S. Wang M.H. Differences in the case fatality risks associated with SARS-CoV-2 Delta and non-Delta variants in relation to vaccine coverage: an early ecological study in the United Kingdom Infect. Genet. Evol. 97 2022 105162 10.1016/j.meegid.2021.105162
14 Johnson A.G. A A. Ali A.R. COVID-19 incidence and death rates among unvaccinated and fully vaccinated adults with and without booster doses during periods of Delta and Omicron variant emergence — 25 U.S. Jurisdictions, April 4–december 25, 2021 MMWR Morb. Mortal. Wkly. Rep. 71 2022 132 138 35085223
| 0 | PMC9741492 | NO-CC CODE | 2022-12-14 23:53:32 | no | Public Health Pract (Oxf). 2023 Jun 10; 5:100350 | utf-8 | Public Health Pract (Oxf) | 2,022 | 10.1016/j.puhip.2022.100350 | oa_other |
==== Front
Kidney Int Rep
Kidney Int Rep
Kidney International Reports
2468-0249
Published by Elsevier Inc. on behalf of the International Society of Nephrology.
S2468-0249(22)01899-X
10.1016/j.ekir.2022.12.004
Research Letters
Serum neutralization of Omicron BA.5, BA.2 and BA.1 in triple vaccinated kidney transplant recipients
Pedersen Rune M. 1
Bang Line L. 1
Tornby Ditte S. 1
Nilsson Anna C. 2
Nielsen Christian 2
Madsen Lone W. 3
Johansen Isik S. 3
Sydenham Thomas V. 1
Jensen Thøger G. 1
Justesen Ulrik S. 1
COVAC-TX study group
Vitved Lars 4
Palarasah Yaseelan 4
Bistrup Claus 5
Andersen Thomas E. 1∗
1 Department of Clinical Microbiology, Odense University Hospital and Research Unit for Clinical Microbiology, University of Southern Denmark, Odense, Denmark
2 Department of Clinical Immunology, Odense University Hospital and Research Unit for Clinical Immunology, University of Southern Denmark, Odense, Denmark
3 Department of Infectious Diseases, Odense University Hospital and Research Unit for Infectious Diseases, University of Southern Denmark, Odense, Denmark
4 Department of Cancer and Inflammation, University of Southern Denmark, Odense, Denmark
5 Department of Nephrology, Odense University Hospital and the Nephrology Research Unit, University of Southern Denmark, Odense, Denmark
∗ Corresponding author: , J. B. Winsløws Vej 21.2, 5000 Odense, Denmark
10 12 2022
10 12 2022
24 11 2022
5 12 2022
© 2022 Published by Elsevier Inc. on behalf of the International Society of Nephrology.
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
Kidney transplant recipients (KTRs) develop a lower-than-normal antibody (Ab) response against the COVID-19 mRNA vaccines and are consequently more vulnerable to breakthrough infections [1, 2]. While booster vaccinations given to these patients increase serum-levels of spike Abs [3], SARS-CoV-2 variants have emerged with increasing vaccine-evading properties, most recently the Omicron variant of concern (VOC). The original Omicron lineage, BA.1 (also known as B.1.1.529), is neutralized less efficiently by booster vaccinated KTRs than the ancestral and the Delta variant, as demonstrated by authentic virus and pseudovirus neutralization assays conducted with blood from these patients [4, 5, 6]. During spring 2022, Omicron BA.1 was rapidly replaced by the Omicron BA.2 subvariant in Europe and the United States and a descendant from Omicron BA.2, the even more transmissible Omicron BA.5 subvariant, has now replaced other subvariants worldwide and to date (October 2022) continues to dominate the COVID-19 pandemic with more than 85% prevalence in most countries [7]. BA.5 deviates from BA.2 by four additional mutations, including two in the receptor binding domain of the spike protein, which makes this subvariant even more immune-evasive in otherwise healthy individuals [8]. While the neutralization status in KTRs post vaccination against the original Omicron BA.1 was earlier estimated [4, 5, 6], the neutralization capacity in KTRs, and solid organ transplanted recipients in general, against the currently dominating descendants of Omicron BA.1, remains to be elucidated.
Here we report the spike Ab levels and serum neutralization capacity against isolates of Omicron BA.1, BA.2, and BA.5 in three times BNT162b2 Pfizer-BioNTech mRNA vaccinated KTRs (n=44) and compare these values to the BA.5 neutralization capacity in healthy controls at the same level of vaccination (n=20). In addition, we report the SARS-CoV-2 reactive T-cell response in a subset of the KTRs (n=25).
Results
The KTRs in this cross-sectional study is part of a KTR cohort that has been described previously, together with the plaque reduction neutralization test (PRNT) [1]. Here, the PRNT90 was measured, which indicates the serum titer which reduces the plaque forming ability of the virus by >90%. A titer of ten was applied as the cut-off indicating the lowest titer that yield actual neutralization in our assay, as a PRNT90 titer of 8.8 was recently shown to indicate real-world protection against the Omicron VOC in humans [S7]. Three clinical Omicron strains, BA.1, BA.2 and BA.5, and one Delta strain, were used for the PRNT90 (Genome sequences: GenBank accession no. ON055874 for BA.1, ON055857 for BA.2, OP225643 for BA.5 and ON055856 for the Delta strain). All experimental work with SARS-CoV-2 was conducted in approved biosafety level 3 facilities. The serum samples were also analysed for spike Ab levels using the LIAISON® SARS-CoV-2 TrimericS IgG assay. Finally, we performed flow cytometry on freshly collected blood to quantify SARS-CoV-2 inducible T cells in 25 KTRs and three healthy controls (for methods, statistics, and patient flow chart, see supplementary material: Cohorts, Supplementary methods, Statistics and Supplementary Figure S1-S2).
The PRNT90 was performed on serum collected at a median of 38 days (IQR 33-49) after the third BNT162b2 vaccination, from 44 KTRs, (19 females, 25 males) with a median age of 60 years (IQR 51-67). As a reference for BA.5 neutralization, PRNT90 was also performed on serum collected at a median of 42 days (IQR 41-42) after the third BNT162b2 vaccination, from 20 healthy control subjects (12 females, 8 males) with a median age of 57 years (IQR 50-60). The characteristics of the KTR cohort are shown in Table 1 . We found that 43% (19/44), 34% (15/44), 41% (18/44) and 39% (17/44) of the KTRs displayed above threshold neutralization of the Delta, Omicron BA.1, BA.2, and BA.5 strains, respectively (Figure 1 A). The PRNT90 titer range of the KTRs towards Delta was <10-160, BA.1 was <10-80, BA.2 was <10-80 and BA.5 was <10-40, with titer levels of the latter subvariant being significantly lower for the KTRs than for the healthy controls (p<0.0001, Mann Whitney test) (Figure 1A). The median Ab levels in sera from the KTRs as measured on the Liaison platform was 193 Binding Antibody Units (BAU)/mL (IQR 8-1743), whereas the median Ab level of the controls was 4115 BAU/mL (IQR 2943-5800). This difference was statistically significant (p<0.0001, Student's t-tests) and Ab levels correlated with the PRNT90 titer of Delta (r=0.88, p <0,0001, Spearman´s correlation [SC]), BA.1 (r=0.82, p<0.0001, SC), BA.2 (r=0.86, p<0.0001, SC) and BA.5 (r=0.86, p<0.0001, SC) among the KTRs as well as with BA.5 among the controls (r=0.74, p=0.0002, SC) (Figure 1B; Supplementary Figure S3). Finally, the cellular immunity was evaluated in a subset of KTRs (n=25) and three healthy controls using T cell flow cytometry. This analysis showed that 21 of 25 (84%) had detectable levels of cytotoxic CD8+ T cells directed against the SARS-CoV-2 spike protein, and 24 of 25 KTRs (96%) had spike protein- reactive CD4+ T helper cells. The KTRs had similar levels of T helper cells compared to both of the vaccinated controls, and similar levels of cytotoxic T-cells compared to the two-times vaccinated, infection-naïve control. The infection convalescent, two-times vaccinated control had considerably higher levels of cytotoxic T-cells than all other, indicating the significant boost in these cells upon infection (Figure 1C).Table 1 | Characteristics of the kidney transplant recipient cohort according to neutralization response against BA.5 after the third dose of the BNT162b2 (Pfizer-BioNTech) vaccine
Demographic characteristics BA.5 neutralizers BA.5 Non-neutralizers P
Numbers (%) 17 (39) 27 (61) N/A
Age Y (IQR) 54 (47-63) 63 (58-73) 0.04
Female (%) 8 (47) 11 (41) 0.76
BMI (IQR) 27.5 (23.7-31.6) 27.4 (23.9-30.0) 0.51
TX characteristics
Time from TX Y (IQR) 7.6 (4.9-12.0) 6.0 (2.4-15.0) 0.95
TX number 0.83
First TX (%) 14 (82) 19 (70) -
Second TX (%) 3 (18) 7 (26) -
Third TX (%) 0 (0) 1 (4) -
Deceased donor (%) 9 (53) 16 (59) 0.76
Induction
Rituximab (%) 1 (6) 0 (0) N/A
anti-CD25 (%) 11 (65) 15 (55) 0.75
anti-CD25 + Rituxmab (%) 1 (6) 1 (4) N/A
Thymoglobuline (%) 3 (18) 6 (22) N/A
Thymoglobuline + Rituxmab (%) 1 (6) 5 (19) N/A
Maintenance
Tacrolimus (%) 16 (94) 19 (70) 0.12
Tacrolimus CO ng/mL (IQR) 5.3 (4.9-6.2) 5.3 (4.9-5.9) 0.62
Ciclosporin A (%) 1 (6) 5 (19) N/A
Ciclosporin A CO nmol/L (IQR) 355 471 (277-551) N/A
Everolimus (%) 0 (0) 0 (0) N/A
MMF/MPA (%) 14 (82) 26 (96) 0,28
MMF (%) 9 (53) 21 (77) 0.11
MPA (%) 5 (29) 5 (19) 0.47
aMMF/MPA: Fraction of full dose (IQR) 1.0 (0.67-1.00) 0.75 (0.67-1.0) 0.44
MMF/kg (IQR) 17.0 (11.3-21.0) 17.1 (14.0-21.3) 0.50
MPA/kg (IQR) 12.4 (6.2-16.2) 7.7 (6.2-11.2) 0.32
Azathioprine (%) 3 (17) 1 (3) N/A
Azathioprine mg (individual dosings) 25-50-100 75 N/A
Steroids (%) 1 (6) 4 (15) N/A
Plasma creatinine μmol/L (IQR) 106 (97-167) 158 (105-207) 0.09
eGFR mL/min (IQR) 59 (33-71) 39 (23-57) 0.06
Underlying disease 0.91
bNon-immune disease (%) 7 (41) 10 (37) -
cImmune disease (%) 7 (41) 8 (30) -
Diabetes mellitus (%) 1 (6) 3 (11) -
Infection (%) 1 (6) 2 (7) -
Unknown (%) 1 (6) 4 (15) -
BMI, body mass index; CO, concentration in plasma; eGFR, estimated glomerular filtration rate; IQR, interquartile range; MMF, mycophenolate mofetil; MPA, mycophenolic acid; N/A, not applicable; TX, transplant; Y, years.
Continuous variables are presented in medians and IQR. Binomial variables are presented in numbers and percentages. Differences were analyzed with the Student's t-test and Fisher's exact test. We have not performed comparisons for data with n<10, which have been designated N/A.
aFull dose of MMF is 1000mg bid except in patients treated with tacrolimus then full dose is 750mg bid. Full dose MPA is 720mg bid except in patients treated with tacrolimus then full dose is 540mg bid. bNon-immune disease designates diseases such as cystic kidney diseases, Alports disease, urinary outlet obstruction etc. cImmune disease designates diseases such as glomerulonefritis, systemic lupus, ANCA associated vasculitis etc.
Figure 1 Neutralization of authentic SARS-CoV-2 Delta, Omicron BA.1, BA.2, and BA.5 strains, antibody levels and SARS-CoV-2 specific T cells among BNT162b2 triple vaccinated kidney transplant recipients. A. Neutralization titers (PRNT90) of KTRs (n=44) against the Delta, Omicron BA.1, Omicron BA.2 and Omicron BA.5 and neutralization titers against Omicron BA.5 of healthy controls (n=20) of sera obtained 5-6 weeks after the third BNT162b2 vaccine dose. Overall, there were significant differences between the PRNT90 values for two or more of the tested SARS-CoV-2 subvariants (p<0.0001, Friedman test). Significant differences as calculated in the subsequent one-to-one subvariant comparison are indicated with asterisks. Medians and IQRs are indicated for the individual groups by gray bars. Horizontal red line indicates the neutralization threshold. B. Dot plot showing spike Ab levels as measured with the LIAISON® SARS-CoV-2 TrimericS IgG assay in relation to the neutralization titers of the BA.5 strain by sera from KTRs (blue dots) and healthy controls (red dots). Overall, the summed-up Ab levels of the KTRs and the healthy controls correlated with the neutralization titer (r=0.91, Spearman's rank correlation, p<0.0001). Ab levels are shown in BAU/mL. The manufacturer provided Ab threshold is 34.8 BAU/mL. Horizontal red line indicates the neutralization threshold. C. The proportion of total CD4+ and CD8+ T cells inducible by SARS-COV-2 S-protein peptides measured in a subpopulation of the KTRs (n=25, open squares) and three healthy controls (closed dots). The healthy controls are *1: a naïve person with respect to both vaccination and infection; *2: a person who had received two mRNA vaccinations, and *3: an infection convalescent person who had received two mRNA vaccinations (for more details, see Supplementary Material). Median values are indicated by horizontal lines. A negative control subject is included who is both SARS-CoV-2 and COVID-19 vaccination naïve (solid circles). *p=0.05-0.01, ****p<0.0001. Abbreviations: Ab, Antibody; BAU, binding antibody units; COVID-19, coronavirus disease 2019; IQR, interquartile range; KTR, kidney transplant recipient; PRNT90, 90% plaque reduction neutralization test; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.
Discussion
Our results show that despite the splitting of the original Omicron VOC into increasingly transmissible subvariants during 2022, KTRs boosted with the BNT162b2 vaccine remain at approximately the same neutralization levels regardless of the Omicron subvariant tested. Compared to the distribution of SARS-CoV-2 PRNT90 titers among healthy persons after three doses [S8] and PRNT90 titers in the current KTR cohort after the second BNT162b2 dose [1], the KTRs’ PRNT90 titers after the third dose now appears to divide into two subpopulations; a group of non-neutralizers that remains below PRNT90 threshold and a neutralizing group which has responded well to the booster by raising neutralization capacities to levels close to the healthy control group. Looking at the antibody levels, many of the non-neutralizers generate detectable levels after the third (booster) vaccination with measurable spike Abs in 82% (36/44) of the KTRs compared to only 49% (28/57) after the second vaccination [1]. With respect to the KTR’s T-cell response, we found that 84% of the KTRs had measurable levels of both cytotoxic and T helper cells which were inducible by the SARS-CoV-2 spike protein. T cell responses against coronaviruses are typically more long-lived than Ab responses against this virus, and coronavirus specific T cells are critical in the protection against severe disease [9].
In conclusion, our results indicate that the KTRs’ neutralization capacity against the currently dominating Omicron BA.5 approximates their ability to neutralize earlier Omicron subvariants. Moreover, a considerable proportion of the KTRs show a relatively robust neutralization response against all tested Omicron subvariants after the first BNT162b2 booster. In addition, we observed a tendency towards a division of KTRs into low- and high-responders, with the first group, despite a general increase in antibody levels, still being unable to neutralize recent SARS-CoV-2 variants. Additional boosters including the now available bivalent mRNA vaccines may increase Ab levels and affinities even in this group, thus providing an important fundamental level of protection against currently circulating SARS-CoV-2 strains.
Supplementary Material
Conflicts of interest
The authors declare no conflicts of interest.
Disclosures
Nothing to disclose
Data availability statement
The data underlying this article will be shared on reasonable request to the corresponding author.
==== Refs
References
1 Pedersen R.M. Bang L.L. Tornby D.S. The SARS-CoV-2-neutralizing capacity of kidney transplant recipients 4 weeks after receiving a second dose of the BNT162b2 vaccine Kidney Int 100 2021 1129 1131 34547366
2 Qui C.X. More L.W. Anjan S. Risk of Breakthrough SARS-CoV-2 Infections in Adult Transplant Recipients Transplantation 105 2021 e265 e266 34310531
3 Caillard S. Thaunat O. Benotmane I. Antibody Response to a Fourth Messenger RNA COVID-19 Vaccine Dose in Kidney Transplant Recipients: A Case Series Ann Intern Med 175 2022 455 456 35007148
4 Benning L. Morath C. Bartenschlager M. Neutralizing antibody response against the B.1.617.2 (delta) and the B.1.1.529 (omicron) variants after a third mRNA SARS-CoV-2 vaccine dose in kidney transplant recipients Am J Transplant 5 2022 10.1111/ajt.17054
5 Charmetant X. Espi M. Benotmane I. Infection or a third dose of mRNA vaccine elicits neutralizing antibody responses against SARS-CoV-2 in kidney transplant recipients Sci Transl Med 14 2022 eabl6141
6 Jurdi A.A. Gassen R.B. Borges T.J. Suboptimal antibody response against SARS-CoV-2 Omicron variant after third dose of mRNA vaccine in kidney transplant recipients Kidney Int 101 2022 1282 1286 35429496
7 Gangavarapu K. Latif A.A. Mullen J. Outbreak.info genomic reports: scalable and dynamic surveillance of SARS-CoV-2 variants and mutations MedRxiv 2022 10.1101/2022.01.27.22269965
8 Tuekprakhon A, Nutalai R, Dijokaite-Guraliuc A, et al. Antibody escape of SARS-CoV-2 Omicron BA.4 and BA.5 from vaccine and BA.1 serum. Cell. 2022;185:2422-33.e13.
9 Channappanavar R. Fett C. Zhao J. Virus-specific memory CD8 T cells provide substantial protection from lethal severe acute respiratory syndrome coronavirus infection J. Virol 88 2014 11034 11044 25056892
| 0 | PMC9741493 | NO-CC CODE | 2022-12-14 23:29:55 | no | Kidney Int Rep. 2022 Dec 10; doi: 10.1016/j.ekir.2022.12.004 | utf-8 | Kidney Int Rep | 2,022 | 10.1016/j.ekir.2022.12.004 | oa_other |
==== Front
Addict Behav
Addict Behav
Addictive Behaviors
0306-4603
1873-6327
Published by Elsevier Ltd.
S0306-4603(22)00343-4
10.1016/j.addbeh.2022.107577
107577
Article
Supporting People Affected by Problematic Alcohol, Substance Use and Other Behaviours Under Pandemic Conditions: A Pragmatic Evaluation of How SMART Recovery Australia Responded to COVID-19
Beck Alison K. a⁎
Larance Briony ab
Baker Amanda L. c
Deane Frank P. ab
Manning Victoria d
Hides Leanne e
Kelly Peter J. ab
a School of Psychology, Faculty of Arts, Social Sciences and Humanities, University of Wollongong, Australia
b Illawarra Health and Medical Research Institute, University of Wollongong, Australia
c School of Medicine and Public Health, University of Newcastle, Australia
d Eastern Health Clinical School, Faculty of Medicine, Nursing and Health Sciences, Monash University, Australia
e Centre for Youth Substance Abuse Research, Lives Lived Well Group, School of Psychology, University of Queensland, Australia
⁎ Corresponding author at: School of Psychology, University of Wollongong, Northfields Ave, Wollongong, NSW, Australia, 2522.
10 12 2022
10 12 2022
10757719 7 2022
28 11 2022
7 12 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
The COVID-19 pandemic prompted rapid, reflexive transition from face-to-face to online healthcare. For group-based addiction services, evidence for the impact on service delivery and participant experience is limited.
Methods
A 12-month (plus 2-month follow-up) pragmatic evaluation of the upscaling of online mutual-help groups by SMART Recovery Australia (SRAU) was conducted using The Reach Effectiveness Adoption Implementation Maintenance (RE-AIM) framework. Data captured by SRAU between 1st July 2020 and 31st August 2021 included participant questionnaires, Zoom Data Analytics and administrative logs.
Results
Reach: The number of online groups increased from just 6 pre-COVID-19 to 132. These groups were delivered on 2786 (M=232.16, SD=42.34 per month) occasions, to 41752 (M=3479.33, SD=576.34) attendees. Effectiveness: Participants (n=1052) reported finding the online group meetings highly engaging and a positive, recovery supportive experience. 91% of people with experience of face-to-face group meetings rated their online experience as equivalent or better. Adoption: Eleven services (including SRAU) and five volunteers delivered group meetings for the entire 12-months. Implementation: SRAU surpassed their goal of establishing 100 groups. Maintenance: The average number of meetings delivered [t(11.14)=-1.45, p=0.1737] and attendees [t(1.95)=-3.28, p=0.1880] per month were maintained across a two-month follow-up period.
Conclusions
SRAU scaled-up the delivery of online mutual-help groups in response to the COVID-19 pandemic. Findings support the accessibility, acceptability and sustainability of delivering SMART Recovery mutual-help groups online. Not only are these findings important in light of the global pandemic and public safety, but they demonstrate the potential for reaching and supporting difficult and under-served populations.
Keywords
SMART Recovery
Mutual-help
Digital Recovery Support Services
Substance Use Disorders, COVID-19
RE-AIM
==== Body
pmc1 Introduction
The COVID-19 pandemic has had a profound impact on mental health and wellbeing. A range of mental health related consequences have been documented, (Ornell et al., 2020) including elevated rates of stress, anxiety and depression worldwide. (Torales et al., 2020, Hossain et al., 2020) Hazardous rates of alcohol use, smoking and other substances have also increased. (Ornell et al., 2020) Other potentially problematic behaviours associated with internet use, gambling and gaming have risen to unprecedented levels. (Dubey et al., 2020) People with pre-existing experience of mental health conditions and/ or addictive behaviours are particularly vulnerable to the psychological impact of the pandemic. (Ornell et al., 2020, Hossain et al., 2020)
The isolation and physical distancing measures introduced to mitigate the impact of COVID-19 significantly disrupted face-to-face service provision for people with experience of addictive behaviours, (Du et al., 2020) necessitating increased use of online technologies by health and social care providers. Remote delivery allows the provision of accessible, flexible, tailored support even under restrictive pandemic conditions. (Rauschenberg et al., 2021) Accumulating evidence supports the feasibility and acceptability of using video-conferencing platforms (e.g. Zoom) to deliver healthcare. (Kruse et al., 2017) Preliminary evidence also supports the clinical and cost effectiveness of telehealth for people with addictive behaviours. (Kruse et al., 2020, Jiang et al., 2017, Lin et al., 2019) However, evidence for group-based telehealth services is currently limited. (Gentry et al., 2019) Given that social connectedness is central to recovery from addictive behaviours; (Bathish et al., 2017) remote access to group-based services may be particularly important for addressing the mental and behavioural challenges arising from the COVID-19 pandemic.
Mutual-help groups are one of the most common, accessible and valued methods for accessing support for addictive behaviours. (Kaskutas et al., 2014, Kelly and Yeterian, 2011) A range of mutual-help groups are available, including 12-step programs (e.g. Alcoholics Anonymous) and other secular options, including SMART Recovery, Women for Sobriety and LifeRing. (Zemore, 2017) Increasing evidence supports the benefits of mutual-help groups for improving substance use, (Beck et al., 2017, Kelly et al., 2020) and mental health outcomes. (Bassuk et al., 2016) Mutual-help groups have been found to enhance recovery supportive social connections, coping skills, self-efficacy and recovery motivation. (Tracy and Wallace, 2016, Kelly, 2017) People may also be more willing to attend mutual-help groups due to the stigma associated with accessing specialist addiction services. (Faulkner et al., 2013, du Plessis et al., 2020, Watson, 2019, Eddie et al., 2019) However, compared to current understanding of face-to-face mutual-help groups, comparatively less is known about remote access delivery. (Ashford et al., 2020, Bergman et al., 2018) Much of the evidence is derived from either evaluations of asynchronous groups (e.g. forums), (Haug et al., 2020, Schwebel and Orban, 2022, Bergman et al., 2017, Chambers et al., 2017) or virtual delivery of 12-step groups, (Penfold and Ogden, 2021, Bender et al., 2022, Hoffmann and Dudkiewicz, 2021, Galanter et al., 2022, Senreich et al., 2022, Barrett and Murphy, 2021) with scant evidence for online SMART Recovery groups. (Timko et al., 2022, Beck et al., 2022) Given that not all individuals engage well with 12-step approaches, research examining participant experience of alternative approaches is an important priority. (Bergman et al., 2021)
SMART Recovery mutual-help groups incorporate evidence-based principles and strategies (e.g. motivational interviewing and cognitive behavioural therapy) to offer support for a range of addictive behaviours. (Kelly et al., 2017) SMART Recovery group based meetings are based on a four-point program (building motivation, coping with urges, problem solving and lifestyle balance) and all meetings are led by a trained facilitator. (SMART, xxxx) SMART Recovery Australia (SRAU) partners with volunteers and a range of general health, mental health and addiction service providers across the private, public and not for profit sectors to deliver mutual-help groups nationwide.
1.1 The Current Evaluation
Prior to COVID-19, Australian SMART Recovery groups provided support to approximately 2200 people each week across 346 face-to-face groups and just six online groups. To meet the continued support needs of people with addictive behaviours during the pandemic, SRAU was awarded funding by the Commonwealth Government of Australia under the Drug and Alcohol Program to upscale online delivery of SMART Recovery groups. SRAU sought to establish at least 100 online groups during the 12-month funded project. To maximise sustainability, SRAU set a goal of 80% of groups delivered by third-party providers (including trained facilitators located within private, not-for-profit and government run health and social care organisations), and 20% by SRAU staff and volunteers. This goal was based on the staff, time and resources available in-house for delivering mutual-help groups. To address inequity in service provision, SRAU also sought to establish groups for targeted cohorts (women and culturally, linguistically and gender diverse people). The current evaluation examined the Reach Effectiveness Adoption Implementation and Maintenance (RE-AIM) of the scaling up of SRAU’s online mutual-help groups in response to the COVID-19 pandemic.
2 Methods
Ethics approval was granted by the Joint University of Wollongong and Illawarra Shoalhaven Local Health District Health and Medical Human Research Ethics Committee (2020/ETH02893). This evaluation was conducted across a 14-month period, comprising the 12-month funded SRAU project, and subsequent 2-months. It was informed by the RE-AIM (Reach Effectiveness Adoption Implementation Maintenance) framework. (Glasgow et al., 2019) This framework is a well-utilised, evidence based approach that is employed within health-care and community settings to direct the planning and evaluation of innovations to service delivery. (Glasgow et al., 2019) To maximise the uptake and sustainability of healthcare innovations, this framework guides evaluators to consider key factors across the level of the individual, service provider, setting and organisation. Aligned with published recommendations for ensuring that these considerations guide the evaluation of real-world initiatives, we adopted a pragmatic approach (Glasgow and Estabrooks, 2018) that leveraged real-world data captured by SRAU. The definition of each RE-AIM domain, (Glasgow and Estabrooks, 2018) operationalisation within the current evaluation and data sources are summarised in Table 1 . Briefly, Reach focuses on data collection at the level of attendees and meetings; Effectiveness is indexed by participant evaluation; Adoption focuses on data collection at the level of the provider and organisation; Implementation outcomes focus on the overall initiative and are therefore derived from apriori targets established by SRAU and Maintenance focuses on the two months following the 12-month funding period with regard to attendees and meetings. For context, we begin by summarising how SRAU approached the task of upscaling online service provision.Table 1 Summary of the five domains of the RE-AIM framework according to the definition, operationalisation and data source.
Domain Definition Operationalisation Data Source
Zoom Data Analytics Participant Questionnaire(n=1052) SRAU Administrative Logs
Reach
Number and representativeness of people willing to engage with a given initiative • Number of meetings delivered
✓
• Attendees (number of log-ins)a
✓
• Average number of attendees per meeting
✓
• Consistency of deliveryb
✓
• Characteristics of participants accessing the online groups
✓
Effectiveness
Impact of an initiative on service user outcomes • Engagement
✓
• Experience
✓
• Contribution to recovery
✓
Adoption
Number and representativeness of settings and providers willing to deliver an initiative • Number of services with the capacity to deliver online groups
✓
• Number of services that delivered online groups during the evaluation period
✓ ✓
• Growth or drop-off of online service provision across the 12-month evaluation periodc
✓ ✓
• Number of facilitators trained
✓
Implementation
Degree to which an initiative is delivered ‘as intended’ • The degree to which SRAU met a-priori goals regarding:
-Total number of online groups establishedd ✓ ✓
-Engagement of third-party providers ✓ ✓
-Provision of support to targeted cohorts ✓ ✓
Maintenance
The extent to which an initiative becomes routine within organisational practices and/ or policies • Ongoing delivery of online groups in the 2-months following the conclusion of the 12-month study period, including:
✓ ✓
Number of meetings
Number of attendees
Note.aTo account for the facilitator, ‘attendees’ are defined as (total number of participants -1) x the total number of groups; bAverage delivery per month per group is derived from the total number of times the each group was delivered, divided by the number of months it was delivered at least once; cIn addition to descriptive statistics, the number of volunteers and service providers who delivered online groups at least once within every month of the evaluation was calculated and used as an index of ongoing engagement; c Calculated as the number of Unique meeting IDs used to deliver groups (n=174) minus the number of IDs only used once (n=42) minus the six existing meetings.
2.1 Summary of SRAU Methods
To begin, SRAU supported existing facilitators to transition their face-to-face groups online. Contact was made with volunteers and third-party providers to explore the option of moving their group(s) online. SRAU provided interested parties with log-in details for Zoom, supported them to set-up their Zoom account and helped orient them to the Zoom platform (as needed). Facilitators then nominated the preferred date and time of their group, which was added by SRAU to the list of advertised groups available on the SRAU website. SRAU endeavoured to offer a support call following the first online group conducted by each facilitator.
New facilitators were also trained using a purpose built online training platform developed in 2019. This two phase training involves four modules of training content, and a collection of video role-plays demonstrating SMART Recovery group facilitation skills. A skill development session with a SRAU trainer and four trainees conducted via Zoom comprised the second phase. Following completion of training, facilitators were asked to notify SRAU if they wished to start an online group. Interested facilitators were then supported to establish a group as per the methods outlined above. Over the course of the project SRAU developed several resources to support facilitators. This included a PowerPoint slideshow to help guide the structure of online groups, and a Facilitator Network Facebook group for all online and in-person facilitators. SRAU also offered fortnightly facilitator support groups via Zoom to any interested facilitators.
2.2 RE-AIM Evaluation
2.2.1 Data Source
Throughout the total 14-month evaluation period (1st July 2020 to 31st August 2021) SRAU routinely collected data using three primary methods: self-report participant questionnaire, Zoom data analytics and administrative logs. The research team also conducted a concurrent qualitative evaluation to explore participant and facilitator experience with online groups. These qualitative findings will be reported separately.
2.2.1.1 Self-Report Participant Questionnaire.
A self-selected, convenience sample of participants was recruited. SRAU embedded a link to an online Survey Monkey questionnaire in the post-group Zoom exit page. This brief questionnaire was developed in-house by SRAU to routinely capture participant data and was presented at the end of every meeting held across the duration of the evaluation. Participants were asked to complete the questionnaire only once, based on their most recent online group experience. The questionnaire captured basic demographic information and reason/s for attending the online group that day. It also contained items to assess participant experience of the group. A series of five-point Likert scale items assessed participant engagement (the degree to which participants felt they were welcomed, supported, and had an opportunity to contribute); experience (skill of the facilitator, experience of technical difficulties); and self-reported contribution of the online group to recovery (acquisition of practical information and strategies, degree to which the group was experienced as helpful and intention to continue attending). Participants’ use of the ‘seven-day plan’ in the current and preceding group was also assessed. The seven-day plan is a change plan, (SMART, 2016, SMRT, 2015) that comprises one or more realistic, personally meaningful goals for the upcoming week. For example, participants from Australian SMART Recovery groups (Gray et al., 2020) have described how they use the seven-day plan to set targets for changing their behaviour of concern (e.g. reducing or abstaining from use) and/ or more general lifestyle goals (e.g. commencing, increasing or maintaining engagement in self-care, social, recreational or vocational activities). Progress towards this plan is reviewed in subsequent groups and the plan revised as needed following feedback and self-reflection. Although a range of tools are available to SMART Recovery participants, the survey focused specifically on participant use of the seven-day plan because goal setting is central to the structure of Australian SMART Recovery groups and leaving a session with a seven-day plan is an essential strategy for taking specific action toward goal achievement. For example, prior research in Australian SMART Recovery groups found that participants who left meetings with a seven-day plan were more likely to report the use of behavioural activation. (Kelly et al., 2015) Finally, participants were asked to rate their satisfaction with online delivery relative to face-to-face groups. Completion of the questionnaire was anonymous, voluntary and no incentive was provided.
A total of 1107 questionnaires were completed across the 12-month period of funding. Duplicate respondents were identified based on a combination of IP address, gender and age (n=21). Thirty-four participants declined to have their anonymous data used for research purposes, leaving a sample of 1052 survey respondents. Participant postcode was used to classify the location of respondents according to the five categories of ‘remoteness’ defined by the Australian Standard Geographical Classification (ASGC; Major City, Inner Regional, Outer Regional, Remote, Very Remote).
2.2.1.2 Zoom Data Analytics.
Information pertaining to the usage of online groups was automatically captured by Zoom throughout the 14-month evaluation period. The ‘meetings’ section of the Zoom dashboard was used to download information about each SMART Recovery group meeting held (including ‘meeting ID’, host, topic, ‘number of participants’ and ‘meeting duration’). ‘Meeting ID’ is the unique identifier assigned by Zoom to an online group (based on the specified host, time, day and whether or not the meeting is recurring). We defined an ‘established’ group as one that occurred more than once. Therefore, the number of online groups established is defined as (total number of unique meeting IDs used to deliver an online group) – (meeting IDs only used once) – (the number of online groups running prior to the current evaluation). The number of ‘groups delivered’ (i.e. ‘meetings’) is defined as the total number of online SMART Recovery groups delivered via the SMART Recovery zoom account across the evaluation period (i.e. including one-off meetings). The ‘number of participants’ metric in Zoom is derived from the number of log-ins to a given group and captured and stored as an aggregate value. Although we are unable to account for multiple log-ins by the same attendee across the study period, to account for the facilitator, we defined “attendees” as: (total ‘number of participants’ – 1) x the total number of ‘groups delivered’.
2.2.1.3 Administrative Logs.
SRAU maintained administrative logs using Arlo Training Management Software and Microsoft Excel to capture data pertaining to project milestones, third-party organisations and facilitator training (e.g. training completion; registration as a facilitator; facilitator name and organisation; facilitator email address). The number of third-party providers with the potential ‘capacity’ to deliver online groups is based on the number of third-party providers who a) were provided with log-in details and were oriented to the Zoom platform, b) had one or more facilitators complete training and/ or c) delivered one or more online groups during the 12-month funded project period.
2.2.2 Statistical Analysis
Each element of the RE-AIM framework was analysed separately. The CSV files for the participant questionnaire and all Zoom Data analytics were downloaded and analysed using SPSS or Microsoft Excel. All quantitative data was summarised using descriptive statistics (mean, standard deviation, range, sum and/ or proportion), as appropriate. To examine consistency of group delivery across time, we firstly calculated the total number of times that each group was delivered and then the number of months in which it was delivered at least once. These values were used to generate an estimate of the average number of times per month each group was delivered while accounting for the varied duration of delivery across the 12–months of the evaluation. Aligned with recommendations for reducing bias and controlling Type I error when comparing two independent groups, (Delacre et al., 2017) we adopted a conservative approach and compared the average number of groups and attendees per month during the maintenance phase to the preceding 12-month evaluation period using separate Welch’s t-tests. (Gaetano, 2019)
2.3 Results
2.3.1 Reach
Between 1st July 2020 and 30th June 2021 a total of 2786 meetings were delivered to approximately 41752 attendees. In any one month, an average of 232.16 (SD=42.34; Range=178-312) meetings were delivered to an average of 3479.33 (SD=576.34; Range=2512 to 4538) attendees. For each meeting delivered there was an average of 14.98 (SD=10.85; Range = 1-73) attendees. On average, each group was delivered on approximately 16 occasions (M=16.98, SD=17.54; Range=1-66) across approximately four months (M=4.89, SD=4.07; Range=1-12), equating to roughly three meetings per month, per group.
The characteristics of survey respondents are presented in Table 2 . Approximately half of respondents were male (n=547, 52%) and aged 35-44 (n=275, 26.1%). Approximately 2% were of Aboriginal and/ or Torres Strait Islander descent (n=26). The majority of respondents were located in ‘major city’ locations (n=753, 71%). Use of alcohol was the most frequently endorsed reason for attending an online SMART Recovery group (n=761; 72%), followed by tobacco (n=139; 13.2%) and methamphetamine (n=136; 12%) use. More than a third selected multiple behaviours (n=406, 38.6%; M=1.63, SD=1.17; Range = 0-11) as their reason for attending. A total of 485 (46.1%) participants attended an online SMART Recovery group for the first time on the day that they completed the survey.Table 2 Demographic characteristics and behaviour(s) of concern for the sample of online participants (n=1052) who completed the participant questionnaire
N (%)
Gender
Male 546 (51.9)
Female 492 (46.8)
Transgender female 0 (0)
Transgender male 1 (0.09)
Non-binary/ indeterminate 8 (0.7)
Not stated 4 (0.3)
Age
Under 18 3 (0.2)
18-24 67 (6.3)
25-34 219 (20.8)
35-44 275 (26.1)
45-54 265 (25.1)
55-64 186 (17.6)
65+ 37 (3.5)
Ethnicity and Cultural Identification
Neither Aboriginal nor Torres Strait Islander 1026 (97.5)
Aboriginal but not Torres Strait Islander 22 (2.09)
Both Aboriginal and Torres Strait Islander 4 (0.38)
Participant Locationa
New South Wales 543 (51.6)
Victoria 272 (25.8)
Queensland 120 (11.4)
South Australia 37 (3.5)
Western Australia 34 (3.2)
Australian Capital Territory 15 (1.4)
Tasmania 8 (0.76)
Northern Territory 4 (0.38)
International 5 (0.47)
Participant remoteness categorya,b
Major City 753 (71.5)
Inner Regional 210 (19.9)
Outer Regional 51 (4.8)
Remote 9 (0.8)
Very remote 9 (0.8)
Behaviour(s) of concern that prompted group attendancec
Alcohol 761 (72.33)
Tobacco 139 (13.21)
Methamphetamine 136 (12.92)
Cannabis 132 (12.54)
Other drugs 187 (17.77)
Food 96 (9.12)
Gambling 51 (4.84)
Sex 45 (4.27)
Shopping 42 (3.99)
Porn 30 (2.85)
Internet 18 (1.71)
Other behaviours 83 (7.88)
None 16 (1.52)
Note.aMissing data for 15 participants did not provide their postcode; bDoes not include the 5 international participants; cParticipants could select more than one behaviour
2.3.2 Effectiveness
Participant responses regarding engagement, experience and recovery are presented in Table 3 . Self-reported engagement with the online group meetings was strong, with the majority of survey respondents endorsing ‘agree’ or ‘strongly agree’ with feeling welcomed (n=986, 93%), supported (n=961, 91%) and having an opportunity to contribute (n=962, 91%). Participant experience was largely positive, with the majority of participants endorsing ‘agree’ or ‘strongly agree’ to the meeting being well facilitated (n=970, 92%), although one in five respondents (n=219, 20.81%) endorsed either ‘agree’ or ‘strongly agree’ to experiencing technical difficulties. Online groups made a positive contribution to the recovery of the majority of respondents with 85% (n=900) leaving the meeting with practical information, strategies and/ or resources; 90% experiencing the group as helpful (n=948) and 91% (n=965) intending to continue attending online SMART Recovery groups. Of the 567 (53.9%) participants who had attended a SMART Recovery meeting previously, 370 (35%) had attended the previous week. Of those, 262 (70.8%) left that meeting with a seven-day plan. Of the participants who completed a seven-day plan the preceding week, the majority reported that they found the plan to be very (n=118, 45%) or extremely (n=80, 31%) helpful.Table 3 Self-reported engagement, experience and impact on recovery reported by the subsample of online group participants who completed the online questionnaire (n=1052). Findings are presented as M (SD) and the proportion of participants who endorsed each Likert scale category (1=Strongly disagree to 5 = Strongly agree).
M (SD) Strongly Disagree Disagree Slightly Agree Agree Strongly Agree
Engagement
I felt welcome at today’s meeting 4.55 (0.89) 43 (4.1%) 6 (0.6%) 17 (1.6%) 241 (22.9%) 745 (70.8%)
I felt supported and understood by people attending the meeting 4.43 (0.92) 42 (4%) 10 (1%) 39 (3.7%) 320 (30.4%) 641 (60.9%)
I had an opportunity to contribute to the group discussion 4.47 (0.93) 44 (4.2%) 9 (0.9%) 37 (3.5%) 279 (26.5%) 683 (64.9%)
Experience
Today’s group was well facilitated 4.50 (0.92) 45 (4.3%) 7 (0.7%) 30 (2.9%) 256 (24.3%) 714 (67.9%)
I experienced technical difficulties during the meeting 2.14 (1.41) 522 (49.6%) 202 (19.2%) 109 (10.4%) 94 (8.9%) 125 (11.9%)
Contribution to Recovery
I took away practical strategies/ideas/ tools from today’s group to help me manage my behaviour 4.27 (0.98) 43 (4.1%) 19 (1.8%) 90 (8.6%) 357 (33.9%) 543 (51.6%)
Overall, I found todays group helpful 4.41 (0.93) 41 (3.9%) 12 (1.1%) 51 (4.8%) 310 (29.5%) 638 (60.6%)
I plan on continuing to attend SMART online 4.51 (0.87) 35 (3.3%) 6 (0.6%) 46 (4.4%) 265 (25.2%) 700 (66.5%)
A total of 555 participants responded to the question about how online groups compared to face-to-face (SRAU added this question after data collection had commenced). Just over one-third (n=210, 37%) had only attended online meetings, so could not make a comparison. Of the remaining 345 participants, approximately half indicated that online meetings were either ‘better’ (n=89, 26%) or ‘much better’ (n=89, 26%) than face-to-face meetings and just over one third felt they were ‘about the same’ (n=129, 37%).
2.3.3 Adoption
Half (n=38, 50%) of the third-party service providers with the potential capacity to deliver online groups delivered at least one online meeting during the 12-month evaluation period. The total number of SRAU staff, volunteers and third-party providers per month who delivered online meetings across the initial 12-month evaluation period is presented in Figure 1 . A total of 11 services (including SRAU) and five volunteers delivered online meetings for the entire 12-month duration of the evaluation.Figure 1 Total number of providers per month who delivered at least one online meeting
A total of 130 new facilitators completed training during the initial 12-month evaluation period, comprising 109 individuals working across 47 different services. Service information for the remaining 21 individuals was not provided. Of these 47 services, less than one third (n=13, 27.6%) went on to deliver an online meeting.
2.3.4 Implementation
The number of new online SMART Recovery mutual-help groups established during the initial 12-month evaluation period (n=126) exceeded SRAU’s a-priori target of 100. Aligned with their target involvement of third-party providers, almost 80% of all meetings (n=2118, 76%) were delivered by third-party providers. Third-party providers included a range of private, not-for-profit and government run organisations delivering addiction services, general health or social care within community, residential and/ or inpatient settings. Regarding equity, of the 2786 meetings delivered, 375 (13.4%) were targeted to provide a dedicated support space for: women (n=134, 4.8%); men (n=55, 1.9%); young people (n=56, 2%); family and friends (n=118, 4.3%); people of Aboriginal and/ or Torres Strait Islander descent (n=11, 0.3%) and Korean language speakers (n=1, 0.03%).
2.3.5 Maintenance
In the two-months following the funded 12-month project, a further 500 meetings (M=250, SD=1.41, Range= 249-251 per month) were delivered. These meetings were attended by approximately 8988 attendees (M=4494, SD=367.69, Range=4234-4754 per month), with an average of 17.97 (SD=12.73) attendees per meeting (Range = 1 to 103). The average number of meetings [t(11.14)=-1.45, p=0.1737] and attendees [t(1.95)=-3.28, p=0.1880] per month during the maintenance phase did not significantly differ from the preceding 12-month evaluation period.
2.4 Discussion
The COVID-19 pandemic triggered widespread closure of SMART Recovery mutual-help groups across the world. (Kelly et al., 2021) To meet the continued support needs of people affected by addictive behaviours, SRAU undertook a project to expand the delivery of online mutual-help groups. In light of limited evidence for the delivery of online groups for addictive behaviours, the purpose of the current evaluation was to assess the impact of this real-world healthcare innovation. Although we are unable to partial out the increase in online service provision that would have occurred naturally within the context of the COVID-19 pandemic, this pragmatic evaluation demonstrated positive change across all domains of the RE-AIM framework. (Glasgow et al., 2019, Glasgow and Estabrooks, 2018) Moreover, this study contributes new knowledge on the characteristics of attendees and how these online-groups were experienced by participants during the COVID-19 pandemic.
Regarding participant characteristics, the age, location and primary behaviour of concern reported by participants in the current study appears broadly consistent with published Australian data for face-to-face groups. (Raftery et al., 2019, Kelly et al., 2021, Beck et al., 2021) However, consistent with recent evidence for the gender distribution of online mutual-help groups, (Timko et al., 2022, Beck et al., 2022) the current sample does appear to comprise a larger proportion of females (46.8%) than that of Australian face-to-face groups (31.9%-39%). (Raftery et al., 2019, Kelly et al., 2021, Beck et al., 2021) This may be due to the increase in the number of women only groups offered during the study period. Indeed, recent evidence suggests that women seeking treatment for addictive behaviours may prefer single gender groups. (Sugarman et al., 2022) Telehealth may also help to overcome a range of barriers encountered by women when trying to access treatment and support. (Goldstein et al., 2018)
Our findings also point to the potential therapeutic benefit of online mutual-help groups. Participant evaluation of online SMART Recovery groups was extremely positive with regard to engagement, experience and impact on recovery. The majority of survey respondents felt welcomed, supported and able to contribute. This is especially promising given that group cohesion may be adversely affected in online settings. (Sugarman et al., 2021) Indeed, of those who had previously attended face-to-face groups 89% felt that online groups were just as good (37%) or better (52%). Similarly, in a survey of 12-step attendees, the majority found that online meetings were at least ‘as effective’ for promoting abstinence as face-to-face meetings. (Galanter et al., 2022) Given the widespread and rapid transition of face-to-face healthcare services to virtual delivery seen during COVID-19, (Centers for Disease Control and Prevention. Using Telehealth to Expand Access to Essential Health Services during the COVID-19 Pandemic., 2020) concerns regarding the resultant impact on participant experience and the quality of service provision have been raised. (Mark et al., 2021, Sugarman et al., 2021) The current study adds to a growing body of evidence (Galanter et al., 2022, Senreich et al., 2022, Timko et al., 2022) (but see also (Barrett and Murphy, 2021) that goes some way to allaying these concerns.
Although 20% of respondents in the current evaluation experienced technical difficulties, the majority still found the groups to be helpful. Participants gained knowledge that supported recovery and many engaged in the seven-day plan, an important behaviour change strategy. These findings are promising since active engagement and coping skills represent important predictors of treatment outcome within face-to-face mutual-help groups. (Marcovitz et al., 2020, Kelly et al., 2009) Evaluations to examine the contribution of online mutual-help groups to participant recovery are needed. However, given that online treatment and support may be less effective, (Gentry et al., 2019, Jenkins-Guarnieri et al., 2015) and less suitable for certain clinical groups, (Bergman and Kelly, 2021) clarifying the participant and contextual variables that influence the effectiveness of online mutual-help groups represents an important challenge for future research.
This study also provides preliminary evidence in support of the feasibility of delivering SMART Recovery mutual-help groups online. SRAU surpassed its project target and established a total of 126 new online groups in 12-months. Prior to this initiative, only six online groups were available. Under pandemic conditions, SRAU worked with volunteers and third-party service providers to deliver over 2700 meetings to approximately 42,000 attendees. In the subsequent 2-months, at a time when face-to-face service provision began to resume, a further 500 meetings were delivered to over 8,000 attendees, with the average number of meetings delivered and attendees per month maintained. Evidence regarding virtual support groups for addictive behaviours is limited, but promising. (Oesterle et al., 2020, Bergman and Kelly, 2021) Within the broader literature, evidence for the feasibility of online service provision for addictive behaviours, (Mark et al., 2021, Molfenter et al., 2021) and the willingness of providers (Molfenter et al., 2021, Cantor et al., 2021) and patients (Sugarman et al., 2021) to engage with this mode of service delivery is growing. However, virtual delivery of groups also comes with a range of practical and clinical challenges (e.g., connectivity, safety and size). (Sugarman et al., 2022, Sugarman et al., 2021) The impact of these challenges on the delivery, experience and effectiveness of online treatment and support is unclear. Further research is needed to understand how best to optimise online service provision for addictive behaviours.
Our findings also lend insight into two key opportunities for extending the reach of online SMART Recovery mutual-help groups. Firstly, there is a need to improve access for priority clinical groups. (Institute, 2020) This includes young people, older people, culturally and linguistically diverse populations and people identifying as gay, lesbian, bisexual, transgender or intersex. (Institute, 2020) Although we are unable to comment on the actual number of people within these priority groups who attended an online meeting across the evaluation period, consistent with published Australian data from face-to-face SMART Recovery meetings, (Raftery et al., 2019, Beck et al., 2021) the majority of participants who completed the online questionnaire identified with a binary gender, were aged between 35 and 54 and did not identify as Aboriginal and/ or Torres Strait Islander, suggesting a need to continue actively targeting the priority clinical groups noted above. Improved understanding of how best to address the support needs of these individuals is essential. (Kelly et al., 2021) World-first evidence to inform the cultural adaptation of SMART Recovery groups for Aboriginal and/ or Torres Strait Islander peoples is now available (Dale et al., 2021, Dale et al., 2021) and an evaluation of an adapted version of SMART Recovery informed by young people for young people is currently underway. Further research is needed to understand how these findings may also inform the delivery of online groups.
Secondly, to maximise access and accessibility of online SMART Recovery groups, there is a need to understand and address barriers to online delivery. Based on the number of third-party providers delivering face-to-face groups pre-COVID, and the number subsequently trained during the evaluation period, the 38 third-party providers who delivered online groups represents approximately half of those with the capacity to do so. This is driven in part by the number of organisations who completed training during the evaluation but did not go on to deliver an online group. Provider attitudes (e.g. preference and comfort) play a key role in willingness to deliver telehealth services for addictive behaviours. (Mark et al., 2021) Practical considerations, for example provider and patient access to adequate technology have also been implicated. (Bergman and Kelly, 2021) Although accessibility issues are more challenging to overcome (indeed, one-fifth of survey respondents encountered significant technical difficulties), ensuring that providers are well-equipped to handle the unique challenges of online group delivery is essential. Drawing from the face-to-face training literature, ongoing supervision, feedback and self-reflection is likely to be key. (Schwalbe et al., 2014, Frank et al., 2020, Caron and Dozier, 2021) We also observed wide variation in the duration that each online group was implemented. Although we can speculate that changes in COVID-19 restrictions and participant/ provider preference for a return to face-to-face service provision may have played a role in whether or not online groups were maintained, further evidence is needed. To help characterise the individual, provider and organisational characteristics that may influence the sustainability of online groups, qualitative evaluations informed by established implementation frameworks (e.g. the Consolidated Framework for Implementation Research) (Schwebel and Orban, 2022) may be of benefit.
2.4.1 Strengths and Limitations
A key strength of this evaluation is the use of an established, evidence-based framework within the context of a real-world innovation in service delivery. The RE-AIM (Glasgow et al., 2019, Glasgow and Estabrooks, 2018) framework ensured that this evaluation was comprehensive and conducted at the level of the individual, group, service provider and organisation. However, several limitations are also worth mentioning. Firstly, findings are largely based on Zoom data analytics. Although this provides unique insight into participant use of online groups, the number of attendees is derived from the number of log-ins and is therefore likely to be a slightly inflated estimate (e.g. due to people logging into the same group multiple times). As this data is stored in aggregate by Zoom, we are also unable to identify the number of unique attendees. Secondly, data pertaining to the characteristics and experience of participants are subject to bias. Findings are derived from a small subsample of people who self-selected to complete the online questionnaire. Administration of the questionnaire at the end of the group also means that we did not capture the characteristics of those attendees who left early (and therefore may have been less satisfied with the group). The majority of the participants also completed the questionnaire at the end of their first meeting, meaning we are unable to comment on whether and how the experience and characteristics of participants may have changed across time. Thirdly, although it is not uncommon to use self-reported Likert scale questionnaires as a pragmatic index of effectiveness (e.g.), (Miller et al., 2021) these findings would be strengthened via the use of standardised, validated instruments, for example by assessing quality of life and other participant reported outcomes. Recent developments in routine outcome monitoring (Kelly et al., 2021, Beck et al., 2021, Kelly et al., 2020) may be of use. Finally, the design of this pragmatic evaluation is such that we are unable to draw causal inferences regarding the increase in online mutual-help provision observed.
2.4.2 Conclusions
The current evaluation describes how SRAU scaled-up the delivery of online SMART Recovery mutual-help groups in response to the COVID-19 pandemic. SRAU worked alongside volunteers and a diverse range of third-party service providers to deliver online SMART Recovery groups to more than 50,000 attendees across the 14-month evaluation period. Testimony to the feasibility and acceptability of delivering mutual-help groups online, groups were well attended and evaluated favourably by participants. Efforts to maximise existing capacity within partner organisations, enhance engagement with priority client groups and identify and address the training and support needs of online facilitators are warranted. Together with the current evaluation, these findings can be used to ensure that SRAU is well positioned to continue delivering an important and accessible support option to a diverse range of people affected by addictive behaviours.
Author Disclosures
3 Role of Funding Source
This project was commissioned by SMART Recovery Australia and is supported by funding from the Commonwealth Government of Australia under the Alcohol, Tobacco and Other Drugs - COVID-19 Response Grant. Neither SMART Recovery Australia, nor the funding body had direct input into the design of the study; analysis and interpretation of data; or writing and submission of the manuscript.
Authorship Statement
Authorship follows ICMJE recommendations (Mark et al., 2021). All authors made substantial contributions to conception, design, methods and/ or the content of the current publication.
CRediT authorship contribution statement
Alison K. Beck: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Project administration. Briony Larance: Conceptualization, Methodology, Writing – review & editing, Supervision. Amanda L. Baker: Conceptualization, Methodology, Writing – review & editing. Frank P. Deane: Conceptualization, Methodology, Writing – review & editing. Victoria Manning: Conceptualization, Methodology, Writing – review & editing. Leanne Hides: Conceptualization, Methodology, Writing – review & editing. Peter J. Kelly: Conceptualization, Methodology, 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.
==== Refs
References
Ornell F. Moura H.F. Scherer J.N. Pechansky F. Kessler F.H.P. von Diemen L. The COVID-19 pandemic and its impact on substance use: Implications for prevention and treatment Psychiatry Res. 289 2020 113096 10.1016/j.psychres.2020.113096
Torales J. O'Higgins M. Castaldelli-Maia J.M. Ventriglio A. The outbreak of COVID-19 coronavirus and its impact on global mental health Int J Soc Psychiatry. 66 4 2020 317 320 10.1177/0020764020915212 32233719
Hossain M.M. Tasnim S. Sultana A. Faizah F. Mazumder H. Zou L. Epidemiology of mental health problems in COVID-19: A review F1000Res. 9 636 2020 10.12688/f1000research.24457.1
Dubey M.J. Ghosh R. Chatterjee S. Biswas P. Chatterjee S. Dubey S. COVID-19 and addiction Diabetes Metab Syndr. 14 5 2020 817 823 10.1016/j.dsx.2020.06.008 32540735
Du J. Fan N. Zhao M. Hao W. Liu T. Lu L. Expert consensus on the prevention and treatment of substance use and addictive behaviour-related disorders during the COVID-19 pandemic Gen Psychiatr. 33 4 2020 e100252 34192233
Rauschenberg C. Schick A. Hirjak D. Seidler A. Apfelbacher C. Riedel-Heller S. Digital interventions to mitigate the negative impact of the COVID-19 pandemic on public mental health: A rapid meta-review JMIR. 23 e23365 2021 10.31234/osf.io/uvc78
Kruse C.S. Krowski N. Rodriguez B. Tran L. Vela J. Brooks M. Telehealth and patient satisfaction: A systematic review and narrative analysis BMJ open. 7 8 2017 10.1136/bmjopen-2017-016242
Kruse C.S. Lee K. Watson J.B. Lobo L.G. Stoppelmoor A.G. Oyibo S.E. Measures of effectiveness, efficiency, and quality of telemedicine in the management of alcohol abuse, addiction, and rehabilitation: Systematic review J Med Internet Res. 22 1 2020 e13252 32012048
Jiang S. Wu L. Gao X. Beyond face-to-face individual counseling: A systematic review on alternative modes of motivational interviewing in substance abuse treatment and prevention Addict Behav. 73 2017 216 235 10.1016/j.addbeh.2017.05.023 28554033
Lin L.A. Casteel D. Shigekawa E. Weyrich M.S. Roby D.H. McMenamin S.B. Telemedicine-delivered treatment interventions for substance use disorders: A systematic review J Subst Abuse Treat. 101 2019 38 49 10.1016/j.jsat.2019.03.007 31006553
Gentry M.T. Lapid M.I. Clark M.M. Rummans T.A. Evidence for telehealth group-based treatment: A systematic review J Telemed Telecare. 25 6 2019 327 342 10.1177/1357633x18775855 29788807
Bathish R. Best D. Savic M. Beckwith M. Mackenzie J. Lubman D.I. “Is it me or should my friends take the credit?” The role of social networks and social identity in recovery from addiction Journal of Applied Social Psychology. 47 1 2017 35 46 10.1111/jasp.12420
Kaskutas L.A. Borkman T.J. Laudet A. Ritter L.A. Witbrodt J. Subbaraman M.S. Elements that define recovery: The experiential perspective JSAD. 75 6 2014 999 1010 10.15288/jsad.2014.75.999
Kelly J.F. Yeterian J.D. The role of mutual-help groups in extending the framework of treatment Alch Res Health. 33 4 2011 350 355
Zemore S.E. Implications for future research on drivers of change and alternatives to Alcoholics Anonymous Addict. 112 6 2017 940 942 10.1111/add.13728
Beck A.K. Forbes E. Baker A.L. Kelly P.J. Deane F.P. Shakeshaft A. Systematic review of SMART Recovery: Outcomes, process variables, and implications for research Psychol Addict Behav. 31 1 2017 1 20 10.1037/adb0000237 28165272
Kelly J.F. Humphreys K. Ferri M. Alcoholics Anonymous and other 12-step programs for alcohol use disorder Cochrane Database of Systematic Reviews. 3 2020 10.1002/14651858.CD012880.pub2
Bassuk E.L. Hanson J. Greene R.N. Richard M. Laudet A. Peer-Delivered Recovery Support Services for Addictions in the United States: A Systematic Review J Subst Abuse Treat. 63 2016 1 9 10.1016/j.jsat.2016.01.003 26882891
Tracy K. Wallace S.P. Benefits of peer support groups in the treatment of addiction Substance abuse and rehabilitation. 7 2016 143 154 10.2147/SAR.S81535 27729825
Kelly J.F. Is Alcoholics Anonymous religious, spiritual, neither? Findings from 25 years of mechanisms of behavior change research Addict. 112 6 2017 929 936 10.1111/add.13590
Faulkner A. Sadd J. Hughes A. Thompson S. Nettle M. Wallcraft J. Mental health peer support in England: Piecing together the jigsaw 2013 Mind London
du Plessis C. Whitaker L. Hurley J. Peer support workers in substance abuse treatment services: A systematic review of the literature Journal of Substance Use. 25 3 2020 225 230 10.1080/14659891.2019.1677794
Watson E. The mechanisms underpinning peer support: A literature review J Ment Health. 28 6 2019 677 688 10.1080/09638237.2017.1417559 29260930
Eddie D. Hoffman L. Vilsaint C. Abry A. Bergman B. Hoeppner B. Lived experience in new models of care for substance use disorder: A systematic review of peer recovery support services and recovery coaching Frontiers in Psychology. 10 1052 2019 10.3389/fpsyg.2019.01052
Ashford R.D. Bergman B.G. Kelly J.F. Curtis B. Systematic review: Digital recovery support services used to support substance use disorder recovery Human Behavior and Emerging Technologies. 2 1 2020 18 32 10.1002/hbe2.148
Bergman B.G. Claire Greene M. Hoeppner B.B. Kelly J.F. Expanding the reach of alcohol and other drug services: Prevalence and correlates of US adult engagement with online technology to address substance problems Addict Behav. 87 2018 74 81 10.1016/j.addbeh.2018.06.018 29960132
Haug N.A. Morimoto E.E. Lembke A. Online mutual-help intervention for reducing heavy alcohol use J Addict Dis. 38 3 2020 241 249 10.1080/10550887.2020.1747331 32314667
Schwebel F.J. Orban D.G. Online support for all: Examining participant characteristics, engagement, and perceived benefits of an online harm reduction, abstinence, and moderation focused support group for alcohol and other drugs Psychol Addict Behav. 2022 10.1037/adb0000828
Bergman B.G. Kelly N.W. Hoeppner B.B. Vilsaint C.L. Kelly J.F. Digital recovery management: Characterizing recovery-specific social network site participation and perceived benefit Psychol Addict Behav. 31 4 2017 506 512 10.1037/adb0000255 28206780
Chambers S.E. Canvin K. Baldwin D.S. Sinclair J.M.A. Identity in recovery from problematic alcohol use: A qualitative study of online mutual aid Drug Alcohol Depend. 174 2017 17 22 10.1016/j.drugalcdep.2017.01.009 28282522
Penfold K.L. Ogden J. Exploring the experience of Gamblers Anonymous meetings during COVID-19: A qualitative study Curr Psychol. 2021 10.1007/s12144-021-02089-5
Bender A.K. Pickard J.G. Webster M. Exploring the impact of COVID-19 on older adults in 12-step programs J Soc Work Pract Addict. 2022 10.1080/1533256X.2022.2047561
Hoffmann B. Dudkiewicz M. The Internet as a space for Anonymous Alcoholics during the SARS-COV-2 (Covid-19) pandemic Global Journal of Health Science. 13 8 2021 61 71 10.5539/gjhs.v13n8p61
Galanter M. White W.L. Hunter B. Virtual Twelve Step Meeting Attendance During the COVID-19 Period: A Study of Members of Narcotics Anonymous Journal of Addiction Medicine. 16 2 2022
Senreich E. Saint-Louis N. Steen J.T. Cooper C.E. The Experiences of 12-Step Program Attendees Transitioning to Online Meetings during the COVID-19 Pandemic Alcoholism Treatment Quarterly. 2022 10.1080/07347324.2022.2102456
Barrett A.K. Murphy M.M. Feeling Supported in Addiction Recovery: Comparing Face-to-Face and Videoconferencing 12-Step Meetings Western Journal of Communication. 85 1 2021 123 146 10.1080/10570314.2020.1786598
Timko C. Mericle A. Kaskutas L.A. Martinez P. Zemore S.E. Predictors and outcomes of online mutual-help group attendance in a national survey study J Subst Abuse Treat. 138 2022 108732 10.1016/j.jsat.2022.108732
Beck A.K. Larance B. Manning V. Hides L. Baker A.L. Deane F.P. Online SMART Recovery mutual support groups: Characteristics and experience of adults seeking treatment for methamphetamine compared to those seeking treatment for other addictive behaviours Drug Alcohol Rev. 2022 10.1111/dar.13544
Bergman B.G. Kelly J.F. Fava M. Eden E.A. Online recovery support meetings can help mitigate the public health consequences of COVID-19 for individuals with substance use disorder Addict Behav. 113 2021 106661 - 10.1016/j.addbeh.2020.106661 33038682
Kelly P.J. Raftery D. Deane F.P. Baker A.L. Hunt D. Shakeshaft A. From both sides: Participant and facilitator perceptions of SMART Recovery groups Drug and Alcohol Review. 36 3 2017 325 332 10.1111/dar.12416 27246699
SMART Recovery Australia. SMART Recovery Australia [Available from: https://smartrecoveryaustralia.com.au/.
Glasgow R.E. Harden S.M. Gaglio B. Rabin B. Smith M.L. Porter G.C. RE-AIMpPlanning and evaluation framework: Adapting to new science and practice with a 20-year review. Frontiers Public Health. 7 64 2019 10.3389/fpubh.2019.00064
Glasgow R.E. Estabrooks P.E. Pragmatic applications of RE-AIM for health care initiatives in community and clinical settings Prev Chronic Dis. 15 2018 E02 10.5888/pcd15.170271 29300695
SMART Recovery Australia. SMART Recovery Australia participants' manual: Tools and strategies to help you manage addictive behaviours2016.
SMART Recovery Australia. SMART Recovery facilitator training manual: Practical information and tools to help you facilitate a SMART Recovery group2015.
Gray R.M. Kelly P.J. Beck A.K. Baker A.L. Deane F.P. Neale J. A qualitative exploration of SMART Recovery meetings in Australia and the role of a digital platform to support routine outcome monitoring Addict Behav. 101 2020 106144 10.1016/j.addbeh.2019.106144
Kelly P.J. Deane F.P. Baker A.L. Group cohesion and between session homework activities predict self-reported cognitive-behavioral skill use amongst participants of SMART Recovery groups J Subst Abuse Treat. 51 2015 53 58 10.1016/j.jsat.2014.10.008 25535099
Delacre M. Lakens D. Leys C. Why psychologists should by default use Welch’s t-test instead of student’s t-test. International Review of Social Psychology. 2017 30
Gaetano J. Welch’s t-test for comparing two independent groups: An excel calculator (1.01) [Micorsoft Excel workbook]2019. Available from: https://www.researchgate.net/publication/332217175_Welch's_t-test_for_comparing_two_independent_groups_An_Excel_calculator_101/link/5ca6bd6a92851c64bd50ba0c/download.
Kelly P.J. McCreanor K. Beck A.K. Ingram I. O'Brien D. King A. SMART Recovery International and COVID-19: Expanding the reach of mutual support through online groups J Subst Abuse Treat. 108568 2021 10.1016/j.jsat.2021.108568
Raftery D. Kelly P. Deane F. Baker A. Dingle G. Hunt D. With a little help from my friends: Cognitive- behavioral skill utilization, social networks, and psychological distress in SMART Recovery group attendees Journal of Substance Use. 2019 10.1080/14659891.2019.1664654
Kelly P.J. Beck A.K. Deane F.P. Larance B. Baker A.L. Hides L. Feasibility of a Mobile Health App for Routine Outcome Monitoring and Feedback in SMART Recovery Mutual Support Groups: Stage 1 Mixed Methods Pilot Study J Med Internet Res. 23 10 2021 e25217 34612829
Beck A.K. Larance B. Deane F.P. Baker A.L. Manning V. Hides L. The use of Australian SMART Recovery groups by people who use methamphetamine: Analysis of routinely-collected nationwide data Drug Alcohol Depend. 225 2021 108814 10.1016/j.drugalcdep.2021.108814
Sugarman D.E. Meyer L.E. Reilly M.E. Greenfield S.F. Women's and men's experiences in group therapy for substance use disorders: A qualitative analysis The American Journal on Addictions. 31 1 2022 9 21 10.1111/ajad.13242 34730866
Goldstein K.M. Zullig L.L. Dedert E.A. Alishahi Tabriz A. Brearly T.W. Raitz G. Telehealth Interventions Designed for Women: An Evidence Map Journal of General Internal Medicine. 33 12 2018 2191 2200 10.1007/s11606-018-4655-8 30284173
Centers for Disease Control and Prevention. Using Telehealth to Expand Access to Essential Health Services during the COVID-19 Pandemic. 2020.
Mark T.L. Treiman K. Padwa H. Henretty K. Tzeng J. Gilbert M. Addiction treatment and telehealth: Review of efficacy and provider insights during the COVID-19 pandemic Psychiatr Serv. 2021 10.1176/appi.ps.202100088
Sugarman D.E. Busch A.B. McHugh R.K. Bogunovic O.J. Trinh C.D. Weiss R.D. Patients' perceptions of telehealth services for outpatient treatment of substance use disorders during the COVID-19 pandemic The American Journal on Addictions. 30 5 2021 445 452 10.1111/ajad.13207 34405475
Oesterle T.S. Kolla B. Risma C.J. Breitinger S.A. Rakocevic D.B. Loukianova L.L. Substance use disorders and telehealth in the COVID-19 pandemic era: A new outlook Mayo Clin Proc. 95 12 2020 2709 2718 10.1016/j.mayocp.2020.10.011 33276843
Bergman B.G. Kelly J.F. Online digital recovery support services: An overview of the science and their potential to help individuals with substance use disorder during COVID-19 and beyond J Subst Abuse Treat. 120 2021 108152 10.1016/j.jsat.2020.108152
Molfenter T. Roget N. Chaple M. Behlman S. Cody O. Hartzler B. Use of telehealth in substance use disorder services during and after COVID-19: Online survey study JMIR Ment Health. 8 2 2021 e25835 33481760
Cantor J. McBain R.K. Kofner A. Hanson R. Stein B.D. Yu H. Telehealth adoption by mental health and substance use disorder treatment facilities in the COVID-19 pandemic Psychiatr Serv. 2021 10.1176/appi.ps.202100191
Marcovitz D.E. McHugh K.R. Roos C. West J.J. Kelly J. Overlapping mechanisms of recovery between professional psychotherapies and alcoholics anonymous Journal of addiction medicine. 14 5 2020 367 375 10.1097/ADM.0000000000000601 32058337
Kelly J.F. Magill M. Stout R.L. How do people recover from alcohol dependence? A systematic review of the research on mechanisms of behavior change in Alcoholics Anonymous Addiction Research & Theory. 17 3 2009 236 259 10.1080/16066350902770458
Jenkins-Guarnieri M.A. Pruitt L.D. Luxton D.D. Johnson K. Patient perceptions of telemental health: Systematic review of direct comparisons to in-person psychotherapeutic treatments Telemed J E Health. 21 8 2015 652 660 10.1089/tmj.2014.0165 25885491
Australian Institute of Health and Welfare. National Drug Strategy 2017-2026: A national framework for building safe, healthy and resilient Australian communities through preventing and minimising alcohol, tobacco and other drug related health, social and economic harms among individuals, families and communities. Canberra: AIHW; 2020. Contract No.: Cat. no. PHE 270. .
Dale E. Lee K.S.K. Conigrave K.M. Conigrave J.H. Ivers R. Clapham K. A multi-methods yarn about SMART Recovery: First insights from Australian Aboriginal facilitators and group members Drug Alcohol Rev. 2021 10.1111/dar.13264
Dale E. Conigrave K.M. Kelly P.J. Ivers R. Clapham K. Lee K.S.K. A Delphi yarn: Applying Indigenous knowledges to enhance the cultural utility of SMART Recovery Australia Addiction Science & Clinical Practice. 16 1 2021 2 10.1186/s13722-020-00212-8 33407873
Schwalbe C.S. Oh H.Y. Zweben A. Sustaining motivational interviewing: A meta-analysis of training studies Addiction. 109 8 2014 1287 1294 10.1111/add.12558 24661345
Frank H.E. Becker-Haimes E.M. Kendall P.C. Therapist training in evidence-based interventions for mental health: A systematic review of training approaches and outcomes Clinical Psychology: Science and Practice. 27 3 2020 e12330 34092941
Caron E.B. Dozier M. Self-coding of fidelity as a potential active ingredient of consultation to improve clinicians' fidelity Adm Policy Ment Health. 49 2 2021 237 254 10.1007/s10488-021-01160-4 34499299
Miller M.J. Pak S.S. Keller D.R. Barnes D.E. Evaluation of pragmatic telehealth physical therapy implementation during the COVID-19 pandemic Phys Ther. 101 1 2021 10.1093/ptj/pzaa193
Beck A.K. Kelly P.J. Deane F.P. Baker A.L. Hides L. Manning V. Developing a mHealth Routine Outcome Monitoring and Feedback App (“Smart Track”) to Support Self-Management of Addictive Behaviours. Front Psychiatry. 2021
Kelly P.J. Beck A.K. Baker A.L. Deane F.P. Hides L. Manning V. Feasibility of a Mobile Health App for Routine Outcome Monitoring and Feedback in Mutual Support Groups Coordinated by SMART Recovery Australia: Protocol for a Pilot Study JMIR Res Protoc. 9 7 2020 e15113 32673272
| 0 | PMC9741494 | NO-CC CODE | 2022-12-16 23:21:38 | no | Addict Behav. 2023 Apr 10; 139:107577 | utf-8 | Addict Behav | 2,022 | 10.1016/j.addbeh.2022.107577 | oa_other |
==== Front
Int J Disaster Risk Reduct
Int J Disaster Risk Reduct
International Journal of Disaster Risk Reduction
2212-4209
Elsevier Ltd.
S2212-4209(22)00709-9
10.1016/j.ijdrr.2022.103490
103490
Article
Impacts of community-level grassroots organizations on household food security during the COVID-19 epidemic period in China
Liang Yajia
Zhong Taiyang ∗
School of Geography and Ocean Science, Nanjing University, Nanjing 210023, China
∗ Corresponding author. School of Geography and Ocean Science, Nanjing University, China; 163 Xianlin Avenue, Nanjing, Jiangsu Province, 210023, China. Phone number:
10 12 2022
10 12 2022
10349013 6 2022
18 11 2022
9 12 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.
Purchasing food from community-level grassroots organizations was a novel and unforgettable experience for Wuhan residents during the COVID-19 lockdown, but little attention was paid to it. The study examined the relationship between community-level grassroots organizations and household food insecurity based on an online survey of household food insecurity in Wuhan in March 2020. The study found that problems in all three domains of food insecurity including food anxiety, food quality and food quantity existed but were uneven. The COVID-19 epidemic affected household food quality the most, while it had the least impact on household food quantity. Community-level grassroots organizations played an important role in promoting food security including reducing worries about food supply and providing enough food intake, but did not ensure households had adequate food quality due to increasing food prices, fewer varieties of food and decreased food freshness. Compared to other grassroots organizations, the community committee had actually become an extension of the government to run administrative grassroots affairs before the epidemic, so its tight relationship with local government made it become the major grassroots power in ensuring household food security at the residential community level.
Keywords
Apartment complex
Community organization
Contingency food provisioning
Food quality
==== Body
pmc1 Introduction
Food-based community organizations have played a non-negligible role in reducing food insecurity caused by disaster, contingency and emergency. Food-based community organizations can be grouped into two types, those who aim to increase access to food and those who aim to provide food-relevant education [1]. Community kitchens are community-based organizations focusing on food and nutrition-relevant education [2]. Food banks or food pantries are food-based community organizations aiming to provide emergency food supply to those in need [3]. There is a difference between food banks and food pantries in the US, Canada, and Australia. In Canada, food banks usually receive food resources from the public and corporate sectors, then distribute food to food pantries and soup kitchens [4,5]. While in the US and Australia, food pantries tend to be smaller than food banks and provide food directly to those in need [6,7]. In the UK, food banks include both operations [8]. Generally, food banks and food pantries have been well operated under normal circumstances, and they have been a critical food source for people with low income in some high-income countries [9].
However, the COVID-19 epidemic has caused challenges for those food-based community organizations to continue their service [10]. Lockdown measures were essential to halting the spread of the virus, but also led to economic and social consequences, including global food insecurity. A set of compulsory measures to reduce community transmission of COVID-19 posed a growing threat to stable food availability because of disrupted food production, and caused huge challenges for people to generate daily incomes to access adequate and preferred food. According to FAO's 2021 report, during the COVID-19 epidemic close to 318 million more people worldwide faced food insecurity at moderate or severe levels in 2020 than in 2019 [11]. Traditional channel interruption due to the COVID-19 outbreak has caused a surge in demand for contingent food distribution [10]. Food banks in Canada provided more people with food assistance in March 2020 compared to the previous year [3]. At the same time, about 42% of food banks in Canada encountered a decrease in volunteers who could dedicate their time, physical labor and knowledge to the organizations in 2020 [3]. The loss of available volunteers could lead to the reduction of the service of food-based community organizations. For instance, about 14% of communal food service sites were closed in the Canadian city of Hamilton during the COVID-19 epidemic [3]. During the COVID-19 epidemic, people seeking assistance from food banks have been almost overwhelming in the UK [12]. The surge of demand for food has caused challenges for food banks because of reduced revenue, as well as a lack of human resources and volunteers [13,14]. Large food assistance organizations have struggled to effectively protect the food security of households or individuals since the outbreak of COVID-19 [15,16]. As the largest federal food assistance program in the United States, the Supplemental Nutrition Assistance Program (SNAP) encountered increasing applications due to the COVID-19 epidemic and the daily application rate in April 2020 was nearly 4 times higher than usual. In California, nearly half of the applicants were first-time participants [17], while studies demonstrated that people using the SNAP program had a higher likelihood of experiencing food insecurity [18].
Besides food-based community organizations, non-food-based community organizations have also been involved in food distribution during the COVID-19 epidemic. One kind of community-level grassroots power is local NGOs (non-government organizations). They took action to provide food to vulnerable groups during the lockdown when it was realized that the public administration was facing great difficulties making effective emergency food plans [[19], [20], [21]]. In Greenland, the volunteers of local NGOs distributed between sixty to ninety meals per day to homeless people during 5 weeks of lockdown [22]. Community-level organizations quickly and spontaneously organized food support campaigns, which played a significant role in ensuring food access, provisioning, and distribution for residents in communities [23,24]. Roughly 9% of households used a community organization to access free food in Canada [25]. In the Andes, different grassroots communities organized to respond to the epidemic effectively. They sold and delivered Agroecological Baskets to each home in middle and upper-level class neighborhoods, then the generated profit was used to distribute food to the neediest poor homes [26]. Community organizers also paid attention to marginalized people's food security; for instance, they were the first to provide food support for Black and Brown communities during the emergence of COVID-19 in the USA [27].
There remain notable research gaps in understanding the role of community-level grassroots organizations in food security. Firstly, existing studies have generally focused on the effect of food-based community organizations on food insecurity mitigation. Despite growing studies of community organizations' influence on food insecurity, the role of non-food-based community organizations in ensuring large-scale emergency food security remains largely understudied. Secondly, most existing studies have placed emphasis on the overall security of food, but little attention has been paid to community-level grassroots organizations’ different influences on three domains of food security - food anxiety, food consumption quality and food intake quantity. Furthermore, there are numerous community-level grassroots organizations such as the community residential committee (Shequ jumin weiyuanhui) and property management company (Wuye gongsi) in China, which are not food-relevant organizations. Some of those community-level grassroots organizations were involved in addressing food provisioning during the COVID-19 epidemic period. Few empirical studies have examined the role of those non-food-based community organizations in emergency food supply during COVID-19 epidemic, especially in the early stage of COVID-19 epidemic. This study aims to narrow the gaps. This study attempts to investigate household food purchasing supported by community-level grassroots organizations during the lockdown period, then uses a regression model to examine the relationship between community-level grassroots organizations and three domains of household food insecurity based on cross-sectional data. The rest of the paper is structured as follows. Section two provides a brief background on community and its relevant grassroots organizations in China. Section three briefly presents data sources and processing, and describes variables and estimation methods. The descriptive statistics and estimated results of models are shown in section four. In section five, the study is analyzed and the reasons why community-level grassroots organizations had different impacts on three domains of household food security are discussed. Section six concludes the findings and provides policy implications.
2 Theoretical framework
2.1 Community-level grassroots organizations and emergency food provision
During the lockdown period in Wuhan, China, governments released policies and recommendations such as mobility restrictions and social distancing due to the high transmission and easy infection of COVID-19. These measures were beneficial to reduce the risk of the virus spreading, but indeed increased the difficulties of ensuring household food security. Thus, community-level grassroots organizations and campaigns emerged quickly and had non-negligible effects on households’ food security. There are mainly three effects of community-level grassroots on ensuring emergency food provision (Fig. 1 ).Fig. 1 Role of community-level grassroots organizations in ensuring household food security.
Fig. 1
The first effect is termed as “recovering effect” in this study, referring to how community-level grassroots organizations contributed to recovering food availability that was broken by the COVID-19 epidemic. Due to the outbreak of COVID-19 in early 2020, the operation of food wholesale markets in Wuhan was affected and led to decreased availability at wholesale markets. Moreover, most wet markets and many supermarkets were closed, which are the primary outlets of urban household food sources. Food delivery labor force shortages and limited food transportation also contributed to decreased food availability at the community or neighborhood level. Grassroots organizations could directly cooperate with local farmers and food retailers to formulate a simple but stable food supply chain. This kind of purchase services increased the availability of food at the community level, which was beneficial for households to purchase sufficient and diverse foods.
Another is termed as “filling up effect”, which refers to that community-level grassroots organizations providing physical access to food outlets during the COVID-19 epidemic. The infection risk increased households' willingness to stay at home while in addition communities that implemented partial lockdowns strictly limited the times for households going out to buy food, such as only three times a week. Familiar food outlets could even be closed. Therefore, time and food selection for households’ food purchasing were reduced massively, which increased the difficulties in ensuring household food quality and quantity. The additional service from grassroot organizations could fill the decreased physical access to food to a certain extent, by organizing food distribution insides neighborhoods. Those food delivery services included handling, sanitization, and home delivery, which contributes to solve the last hundred yards problem (from the neighborhood gate to household door).
The term “alleviating effect” is used to refer to the role of community-level grassroots organizations in reducing household food anxiety. The epidemic was unprecedented and caused wide disruptions in the urban food provisioning system. Actions of grassroots organizations involved in food provisioning probably provided an informal promises for residents who were being faced with disrupted food supply chains. and helped to alleviate households' anxiety, or even panic, about food shortages. WeChat groups provided a platform for grassroots organizations to collect the information of households' demands for food diversity. Grassroots organizations also paid attention to the food demands of vulnerable groups such as the elderly and the disabled, who are not as likely to be familiar with using smartphones and/or have mobility issues. With the growing maturity of the practices, the food purchase and delivery services from grassroots organizations not only ensured households' food quality and quantity, but also played an important role in decreasing the worries and concerns about households’ food supply and consumption.
2.2 Community and its relevant grassroots organizations in China
There are two meanings for “community (Shequ)” in China in view of government administration and geographic space. The first one is that “community” is viewed as a geographical unit (Fig. 2 ). A community is a spatial unit composed of several neighboring residential complexes which are commonly served by a property management company. Those neighboring residential complexes could be separated by open spaces such as woodland, river or park.Fig. 2 Street, community and residential complex in China.
Fig. 2
The second meaning is that a community is also an administrative unit. There are five levels of government in China, including central, provincial, prefectural, county and township-level governments. A township-level government in a city is commonly called a street office. A region governed by a street office (township-level government) is called a “street” (Jiedao), which is usually composed of several communities. A community residential committee is a self-governing mass organization in law that is financially and personally supported by the government, which thus could be regarded as a quasi-administrative organization. Therefore, community is de jure an administrative spatial unit.
As a grassroots organization, a community residential committee provides services for residents and conducts administration required by the government. For a residential complex, members of the property owner committee are elected by the apartment or house owners. The property owner committee is subordinate to the community residential committee and is responsible for selecting and employing a property management company. A community residential committee has the power to supervise and require property management companies to conduct or implement governmental policies such as lockdown measures. Moreover, both the community residential committee and property management company can also establish volunteer teams that can be involved in conventional and emergent management. Thus, there could be three kinds of grassroots organizations acting within a residential complex when combating COVID-19, including a community residential committee, a property management company and a volunteer team. All these community-level of grassroots organizations are non-food-based. Before the outbreak of the epidemic, the responsibilities of community residential committees focused on carrying out the initiatives to improve public etiquette and ethical standards and the maintenance of residents’ public welfare [28,29]. Property management companies worked to improve the buildings and facilities in the residential complex [30], while the volunteer teams which focused on purchasing food were almost nonexistent.
3 Methods and data sources
3.1 Data sources
Wuhan was chosen as the study area. Wuhan was the first city that reported confirmed cases of COVID-19. Residents in Wuhan experienced both partial and complete enclosures of residential complexes between February 11th and April 8th, 2020 [31]. Partial enclosure policy required residents to stay at home and allowed one member per household to leave and return to their residential complex with limited frequency; while complete enclosure policy allowed no one to go outside their residential complex [31]. The policy of “community group buying” was implemented as a contingent food provision policy in response to disrupted access to food outlets caused by complete enclosure [31]. Under the lockdown measures such as partial and complete enclosure, people's mobility was restricted and public and airtight places including wet markets were asked to close, which directly increased the difficulty for residents to access food. Community-level grassroots organizations thus played an indispensable role in securing the food supply of households. As a secondary-tier city in China, Wuhan has a large number of grassroots organizations. It has a total resident population of 12.45 million and 13 districts (county-level administrative units), with 1431 residential communities (neighborhoods) [32], which means that a community committee served about 8700 persons or about 3522 households in 2020. A community residential committee usually consists of 5–9 staff members [29,33]. This suggests that community-level grassroots organizations could have been involved in food provisioning. For instance, the community committee could have been required to collect information about households' food demands.
As mobility restrictions were implemented, the study distributed an online questionnaire survey between March 24th and 31st, 2020. The electronic questionnaire was compiled on a platform called Wenjuanxing, which is a convenient tool to design questionnaires as well as retrieve and analyze data. WeChat is the most widely used social media in China, and WeChat groups are generally established on kinship, work and life where people from different backgrounds can gather [34], so it is a good channel to invite people from different regions and classes to participate in the questionnaire. By limiting the IP address, we ensured the online survey could only be seen and filled in by respondents who stayed in Wuhan at the time. This was the only feasible method to capture real-time experience of households during the COVID-19 responses in Wuhan, although it was not the most ideal.
The strategy used to collect data was similar to snowball sampling. First, our research team shared the online questionnaire with friends who lived in Wuhan based on personal relationships, and encouraged them to invite more respondents to join in various WeChat groups. Then, in order to keep the samples representative of the community's diversity, our team tried our best to obtain interviewed households of diverse locations, structures, and economic levels. Finally, 874 respondents filled in the questionnaire with those respondents spatially covering almost areas of Wuhan. There were 820 valid cases after removing cases with missing data for relevant variables.
The questionnaire aimed to investigate the status of household food security in Wuhan, and mainly included three aspects of information. The first aspect was basic information about the household, such as household size, the head of household, and household living environment; The second aspect was household food security status, which was measured based on the Household Food Insecurity Access Score (HFIAS). HFIAS is proposed by the FANTA project, and it has been assumed that it is able to effectively distinguish whether the household is food secure, and it is widely used in cross-sectional surveys due to its low cost [35]; The third aspect was the information about household food purchase approaches, food consumption and difficulties or challenges for household access to food during the COVID-19 lockdown. In this part, the questionnaire asked households whether and how often they purchased food from community-level grassroots organizations, as well as for comments about this purchase approach if they were willing to share.
3.2 Household food insecurity measurement and its dependent variables
It is not enough to know how long a household experienced food insecurity, it is also reasonable to reflect how severe the experience was and what characteristics of food insecure status the household had. The household food insecurity happens on a gradient and mainly includes three periods [36]: (1) the first period is when a household feels anxiety and uncertainty about future physical or economic access to food, this period is an initial and mild household food insecure status; (2) the second period is when a household's diet quality and diversity have to change because of difficulties of food access, but family members do not need to endure hunger, this period is a moderate household food insecure status; (3) the last period is when a household encounters limited food quality and quantity, family members have to skip meals and even starvation appears, this is a severe household food insecure status. These periods correspond to three domains of household food insecurity based on HFIAS metrics. Table 1 shows the nine questions for measuring the household food insecurity scale [37]. The nine questions can be grouped as three domains, including the domain of food anxiety, the domain of food quality (insufficient quality) and the domain of food quantity (inadequate food intake). Thus, three dependent variables were used to represent three domains of household food insecurity, symbolized as FA, FQL, and FQT, respectively. The study comprehensively examined relationships between grassroots organizations and the three domains of food security.Table 1 Nine questions used to measure household food insecurity scale.
Table 1Question (During the lockdown period) Occurrence frequency options
Food anxiety (generate variable FA)
Q1: Did you worry about household would not have sufficient food? 0 = Never, 1 = Rarely,
2 = Sometimes,
3 = Often or all the time
Food quality (generate variable FQL)
Q2: Were your household not able to eat preferred food? For each question, options include:
0 = Never, 1 = Rarely,
2 = Sometimes,
3 = Often or all the time
Q3: Did your household have to eat only a few kinds of foods?
Q4: Did your household have to eat some kinds of foods that not really want?
Food quantity (generate variable FQT)
Q5: Did your household have to eat a smaller meal than needed? For each question, options include:
0 = Never, 1 = Rarely,
2 = Sometimes,
3 = Often or all the time
Q6: Did your household have to eat fewer meals in a day?
Q7: Did your household have no food to eat?
Q8: Did your household have to go to sleep hungry?
Q9: Did your household eat nothing for a whole day and night?
Source: Household Food Insecurity Access Scale (HFIAS) for measurement of food access: indicator guide(version 3) [37].
As shown in Table 1, the nine-item scale of HFIAS measures people's perceptions about food insecurity including anxiety about food supply, limited food variety and quality, and insufficient food quantity. Generally, each question had four response options in the questionnaire including ‘Never’, ‘Rarely’, ‘Sometimes’ and ‘Often’. HFIAS set up an algorithm that guides users to respectively measure three different domains of household food insecurity [37]. The study added the option of ‘All the time’ due to the complete lockdown policy in Wuhan, which was unprecedented and lasted more than four weeks, then divided the five response options into four groups, the last group including two options - ‘Often’ and ‘All the time’. The coding rules of variables FA, FQL, and FQT were presented as follows (Fig. 3 ). For variable FA, if a surveyed household never felt anxiety about food access, the value of variables FA equals 1, if it felt anxiety about food access sometimes, the value of variables FA equals 2, and if it always felt anxiety about food access, FA equals 3. The value of variable FQL was given based on Q2-4, and the value of variable FQT was given based on Q5-9. The rule of applying 1, 2, or 3 is shown in Fig. 3. The value 1, 2 and 3 represent the levels of trivial (including none and mild), moderate and severe, respectively.Fig. 3 Value assignment for the variable FA, FQL and FQT.
Fig. 3
Table 2 shows the definition and statistics of three dependent variables. Variables FA, FQL and FQT were used in model 1, model 2 and model 3, respectively. The highest mean among them was variable FQL, while the lowest was variable FQT. Meanwhile, standard deviation of variable FQL also meant that the distribution of household food quality was more concentrated than the two other domains. These data indicated that the lockdown may mainly have worsened the severity of household food quality.Table 2 Definition and statistics of dependent variables.
Table 2Variable Definition Model Mean Standard deviation
1 2 3
FA Household food anxiety, never food anxiety = 1, moderate food anxiety = 2, severe food anxiety = 3 ✓ 2.36 0.76
FQL Household food quality and variety, adequate food quality = 1, moderate insufficient food quality = 2, severe insufficient food quality = 3 ✓ 2.60 0.61
FQT Household food quantity, enough food quantity = 1, moderate insufficient food quantity = 2, severe insufficient food quantity = 3 ✓ 1.92 0.85
3.3 Independent variables
The binary variable GRO was used to reflect whether households purchased food through grassroots organizations. A food contingency planning policy has been in place since 2004 (after the 2003 SARS outbreak) in China, however, the COVID-19 epidemic confining millions of people in residential communities was unprecedented, which led to an urgent need for additional and effective actions to address “last mile” and “last hundred yards” problems of food distribution. The “last mile” refers to food transportation from wholesale markets or supermarkets to a residential complex and the “last hundred yards” refers to food delivery from the residential complex gate to every household's doorstep. Some community-level grassroots organizations were involved in the last mile and last hundred yards problems. In terms of formality, there are two main types of community-level grassroots organizations. One is formal grassroots organizations, which includes community committees and property management companies, and the other is informal grassroots organizations such as volunteer teams. Along with rapid urbanization and marketization, the Chinese central government officially adopted “community service as an alternative way of providing the supplemental safety net in urban areas in 1994” [38]. Therefore, community committees take an important role in providing neighborhood security, sanitation, welfare, mediation services and promoting neighborhood development [39]. A property management agent is a management company that provides professional housing management and services for homeowners, which is beneficial to keeping a safe and clean living environment [40], and is a good go-between for mediating conflicts between homeowners as a third-party [41]. Volunteer teams work as informal grassroots organizations, for instance, some residents undertake work such as purchasing and distributing food. The variable GRO was computed by answers to the question - “Since January 23, did your household buy food through any of the following means?” If the surveyed household chose any of the options including ‘Residential property management committee’, ‘Local government (community committee)’ and ‘Volunteers’, the value of variable GRO equals 1, and 0 for otherwise. Using the service from community-level grassroots organizations could decrease household food security, the coefficient of variable GRO thus is assumed to be negative.
The binary variable FC was used to indicate whether an interviewed household is female-centered. The variable is based on the question - “Which of the following best describes your household structure?” If the interviewed household chose the option of ‘Female-centered (No husband/male partner in the household, may include relatives, children, friends)’, variable FC equals 1, 0 for otherwise. It has generally been observed that female-headed households were more likely to experience higher rates of poverty compared to male-headed households [42]. The reasons are mainly attributed to gender inequality, including females facing more disadvantages in the labor market, having more time spent on child care and household chores, and having fewer opportunities to gain a high level of education [43,44]. Therefore, as female-centered households faced more challenges that could increase household food insecurity, the coefficient of variable FC thus is assumed to be positive.
The variable NFM was used to represent the number of family members. It was computed based on the question - “During quarantine, how many household members are you living with?” The options included 0 to 9 (more than 9). Social transformation has caused a surge in one-person households and multigenerational households in China [45]. Variable NFM thus is measured using three dummy variables to distinguish household size and structure: one dummy variable for 2 to 4 family members, one dummy variable for 5 to 8 family members, and one dummy variable for more than 8 family members. There is a complex relationship between household size and household food insecurity, the coefficient of variable NFM thus is uncertain. It is generally assumed that moderate household size is beneficial to ensuring household food security [46], but as the number of family members increases, the incidence of severe food insecurity increases significantly [47], especially for households that have many children or adolescents.
Previous studies indicated that living in a leased or rented home is associated with household food insecurity [48]. Housing costs are a big financial burden for a low-income household and the financial constraints force the household to try to balance other aspects of expenditure, which causes difficulties for those households to have access to nutritious, preferred and adequate food [2]. Therefore, the binary variable RP was used to reflect whether the household rented a property. If interviewed household answered ‘I'm renting the property’ to the question - “Do you own your current place of residence or are you renting it?“, the value of RP equals 1, and equals 0 otherwise. Additional financial pressure caused by renting a house would increase the risk of household food insecurity, the variable RP thus is assumed to have a positive coefficient.
Economic access is one important aspect of household food security. An increase in food prices would harm household economic access to food. For example, the sharp increase in food prices caused about 1 billion undernourished people in the world from 2005 to 2008 [49]. The study used a dummy variable FPI to examine whether food price affected household food insecurity during the epidemic period. The variable FPI was generated based on the question - “Since January 23, did you or any member of your household experience any of the following challenges?” If the surveyed household chose the option of ‘Food price increase’, the value of FPI equals 1, and equals 0 otherwise. Decreased food affordability would increase household food insecurity, therefore the coefficient of FPI is assumed to be positive.
Conventional food patronage is undoubtedly an indispensable factor in influencing household food security. Although purchasing food in online stores is not a traditional method for households, with the rapid development of e-commerce, online food purchasing has become a more and more popular option for urban residents in China [50]. Purchasing food from public food markets (wet markets), supermarkets and online were conventional food purchasing methods in economically-developed cities such as Nanjing before the outbreak of COVID-19. Therefore, variable RA was used to reflect whether surveyed households experienced restricted access to physical markets and online food markets. The variable RA was generated based on the answer to the multiple-choice question - “Since January 23, did you or any member of your household experience any of the following challenges?” If a surveyed household chose any of the options including, ‘Restricted access to food markets and supermarkets’ or ‘Restricted access to online stores’, variable RA equals 1, it equals 0 otherwise. Variable LFA was used to denote whether surveyed households experienced limited variety and amount of food. If a surveyed household chose any of the options including, ‘Limited food availability and lack of food variety at wet markets or supermarkets’ or ‘Limited food availability and lack of food variety at online stores’, the value of variable LFA equals 1, otherwise it equals 0. Restricted access to food and insufficient food supply would increase the risk of facing household food insecurity, variable RA and LFA are thus assumed to have a positive coefficient.
For the model with FA as dependent variable, there were ten additional binary control variables to reflect whether and the extent to which different types of food groups affect household food anxiety. The ten variables were based on the multiple-choice question - “Has the COVID-19 outbreak affected your consumption of the following foods?” The definitions of variables were made according to the household dietary diversity scores (HDDS) by FAO [51]. Variables HCC, HRTC, HMEC, HVC, HFC, HSC, HEC, HFSC, HMIC, HSCC and denoted whether the epidemic affected household consumption of cereal, roots and tubers, meat, vegetables, fruits, foods made from soybeans, eggs, fish and seafood, milk and food made from milk, spice and condiments, respectively. If surveyed households confirmed that the COVID-19 outbreak affected the consumption of one or some food groups, the value of the related variable is 1, otherwise it is 0. As the variables HMEC and HVC included several food groups, once the household chose any of the groups, the value of the variable equals 1, 0 for otherwise. Variable HMEC consisted of organ meat and flesh meat. Variable HVC included several food groups, such as green leafy vegetables, vitamin A-rich vegetables and tubers, and other vegetables. Any kind of affected food consumption would increase the probability of household food insecurity, therefore the coefficients of these variables are assumed to be positive.
The definition and statistics of all independent variables are shown in Table 3 . Variables HCC, HRTC, HMEC, HVC, HFC, HSC, HEC, HFSC, HMIC, HSCC are only used in model 1.Table 3 Definition and statistics of independent variables.
Table 3Variable Definition Expected
Sign Mean Standard deviation
GRO Grassroots organization, GO = 1 for household obtained food from grassroots organization, otherwise, GO = 0 – 0.79 0.40
FC Female-centered household, FC = 1 for the head of household was female, otherwise, FC = 0 + 0.06 0.24
NFM Number of Family members, NFM range from 1 to 4 2.28 0.66
RP Rent the property, RP = 1 for household rented the property, otherwise, RP = 0 + 0.12 0.32
FPI Food price increase, FPI = 1 for the price of food increased, otherwise, FPI = 0 + 0.60 0.49
LA Limited access to stores, LA = 1 for household experienced limited access to offline or online stores, otherwise, LA = 0 + 0.89 0.32
LFA Limited food availability in stores, LFA = 1 for household Limited food availability in offline or online stores, otherwise, LFA = 0 + 0.45 0.50
HCC Household cereal consumption, HCC = 1 for the COVID-19 epidemic affected household cereal consumption, otherwise, HCC = 0 + 0.21 0.41
HRTC Household roots and tubers consumption, HRTC = 1 for the COVID-19 epidemic affected household roots and tubers consumption, otherwise, HRTC = 0 + 0.12 0.33
HMEC Household meat consumption, HMEC = 1 for epidemic affected household meat consumption, otherwise, HMEC = 0 + 0.60 0.49
HVC Household vegetables consumption, HVC = 1 for the COVID-19 epidemic affected household vegetables consumption, otherwise, HVC = 0 + 0.38 0.48
HFC Household fruits consumption, HFC = 1 for the COVID-19 epidemic affected household fruits consumption, otherwise, HFC = 0 + 0.31 0.46
HSC Household consumption of food made from soybeans, HSC = 1 for the COVID-19epidemic affected household consumption of food made from soybeans, otherwise, HSC = 0 + 0.34 0.47
HEC Household eggs consumption, HEC = 1 for the COVID-19 epidemic affected household eggs consumption, otherwise, HEC = 0 + 0.07 0.26
HFSC Household fish and seafood consumption, HFSC = 1 for the COVID-19 epidemic affected household fish and seafood consumption, otherwise, HFSC = 0 + 0.42 0.49
HMIC Household consumption of milk and food made from milk, HMIC = 1 for the COVID-19 epidemic affected household consumption of milk and food made from milk, otherwise, HMIC = 0 + 0.25 0.44
HSCC Household spice and condiments consumption, HSCC = 1 for the COVID-19 epidemic affected household spice and condiments consumption, otherwise, HSCC = 0 + 0.10 0.30
Note: The expected sign shows the relationship of this variable to the independent variables, with ‘+’ indicating a positive correlation and ‘−’ indicating a negative correlation.
3.4 Ordered logit model
The dependent variables in models 1–3 were variables FA, FQL, FQT, respectively, independent variables in models 2 and 3 were the same, but model 1 added additional variables. As variables FA, FQL, FQT are ordinal and discrete, ordered logit/probit models are a reasonable regression approach. However, the parallel-lines assumption needs to be satisfied when using ordered logit model, which assumes that the relationship between any pairs of outcome categories is equal. If some explanatory variables violate the parallel-lines assumptions, the results of the ordered model are biased and misleading. One major method to test the assumption is using the Brant test. The specification of the ordinary ordered logit model is as follows (Equation (1)). Fi* is a latent but continuous variable in the function, whose value determines what the observed but ordinal Fi equals, the specific functions are shown in Equation (2).(1) Fi*=a*X+εi
(2) Fi={1if−∞<Fi*<w1triviallevel2ifw1<Fi*<w2moderatelevel3ifFi*>w2severelevel
where the dependent variable Fi represents three types of food insecurity status, which include food anxiety, food quality and food intake. The higher the value, the more severe the food insecurity status. X is a vector of independent variables including explanatory variables, household characteristics and food consumption characteristics. a* are unknown coefficients of explanatory variables that need to be estimated, εi is the random error term that is assumed to obey logistic distribution, and wk is also an unknown threshold that needs to be estimated.
Unlike the linear regression model, the logistics model uses odds ratio (OR) rather than coefficient to indicate how the explanatory variable affects the dependent variable on the underlying scale [52]. OR is a commonly used tool to compare proportions, and can be calculated as follows (Equation (3)).(3) {OR=p1−p=exp(X′a)p=(F=1|X)=exp(X′a)1+exp(X′a)
4 Results
4.1 Household food security and community-based grassroots organization
A total of 820 interviewed households were included in the analysis (Table 4 ). There were 651 households out of 820 which bought food via grassroots organizations. Regarding food anxiety, about 17%, 29% and 53% of households experienced trivial, moderate and severe anxiety about food access. For those households with trivial and moderate food anxiety, about 81% used services from grassroots organizations. 78% of households with severe food anxiety bought food through the assistance of grassroots organizations. As far as limited food quality, 7%, 26% and 67% of households encountered trivial, moderate and severe insufficiency in food consumption quality. For those households with trivial food quality insufficiency, 71% purchased food from grassroots organizations, while the proportions of households with moderate and severe food quality insufficiency who purchased through grassroots organizations were 82% and 80%, respectively. With regard to limited food quantity, 40%, 28% and 32% of households experienced trivial, moderate and severe inadequate food intake during the COVID-19 epidemic outbreak in 2020. For those households with trivial food quantity insufficiency, 82% purchased food assisted by grassroots organizations. 80% and 76% of households with moderate and severe food intake insufficiency purchased food with assistance from grassroots organizations.Table 4 Descriptive statistics of three domains of food insecurity.
Table 4Value Food anxiety Food quality Food quantity
No. % % using service No. % % using service No. % % using service
1 143 17.44 81.12 55 6.71 70.91 330 40.24 81.82
2 241 29.39 80.91 215 26.22 82.33 228 27.81 79.82
3 436 53.17 77.98 550 67.07 79.09 262 31.95 75.95
Note: value refers to the value of the variable FA, FQL and FQT. % using service refers to the percentage of household buying food via community-level grassroots organization.
The results demonstrated that the COVID-19 outbreak had severe impacts on household food security. Among the three domains of food insecurity, households faced the least problem in obtaining sufficient food quantity, while food quality was the most prominent challenge to food security during this period of the COVID-19 epidemic. Most households purchased food with assistance from community-based grassroots organizations during the lockdown. At the descriptive level, therefore, we can hypothesize that community-based grassroots organizations made very important contributions to conveying and distributing food to households in Wuhan, especially when households got stranded in a severe food insecure status.
4.2 Model estimation results
Table 5, Table 6, Table 7 presents the estimation results of Model 1, Model 2 and Model 3. The P-value of the Brant test of models 1–3 shows that the assumptions in the three models were not violated (p-value>0.10), which indicates traditional ordered logit model is suitable to examine the relationship between household food security and grassroots organizations.Table 5 Estimation results of model 1 with FA as dependent variable.
Table 5variables Coef. Odds Ratio Standard error Z-score 95% Confidence Interval
GRO −0.42** 0.66 0.12 −2.28 0.46 0.94
FC 0.21 1.24 0.39 0.68 0.67 2.29
NFM(ref:live alone)
2–4 0.59** 1.81 0.48 2.24 1.08 3.05
5–8 0.69** 1.99 0.56 2.41 1.14 3.46
more than 8 0.00 1.00 0.51 −0.01 0.37 2.70
RP 0.45* 1.57 0.39 1.83 0.97 2.54
FPI 0.70*** 2.02 0.32 4.38 1.48 2.77
LA 0.45** 1.57 0.35 2.02 1.01 2.44
LFA 0.00 1.00 0.16 0.00 0.74 1.36
HCC 0.46** 1.58 0.33 2.22 1.06 2.37
HRTC 0.65** 1.91 0.56 2.22 1.08 3.39
HMEC 0.32** 1.38 0.23 1.94 1.00 1.91
HVC 0.56*** 1.76 0.29 3.36 1.26 2.44
HFC 0.35** 1.42 0.26 1.95 1.00 2.02
HEC 0.14 1.15 0.19 0.80 0.82 1.60
HFSC 0.35 1.42 0.55 0.92 0.67 3.02
HSC 0.02 1.02 0.17 0.10 0.73 1.41
HMIC 0.42** 1.53 0.30 2.12 1.03 2.26
HSCC 0.04 1.04 0.30 0.13 0.59 1.83
Log likelihood −737.87 LR Chi-square 164.77 (p-value = 0.00)
Pseudo R2 0.10 Brant test 14.02 (p-value = 0.78)
Note: *, **, *** represents significant at the level of 10%, 5% and 1%, respectively.
Table 6 Estimation results of model 2 with FQL as dependent variable.
Table 6variables Coef. Odds Ratio Standard error Z-score 95% Confidence Interval (OR)
GRO −0.13 0.88 0.17 −0.65 0.61 1.29
FC −0.40 0.67 0.21 −1.29 0.38 1.27
NFM(ref:live alone)
2–4 −0.42 0.66 0.20 −1.36 0.36 1.19
5–8 −0.29 0.75 0.25 −0.89 0.39 1.44
more than 8 −0.79 0.45 0.25 −1.44 0.15 1.33
RP 0.39 1.47 0.40 1.43 0.86 2.49
FPI 0.70*** 2.01 0.32 4.33 1.48 2.78
LA 1.05*** 2.86 0.63 4.78 1.80 4.12
LFA 0.74*** 2.10 0.35 4.50 1.51 2.89
Log likelihood −609.42 LR Chi-square 93.32 (p-value = 0.00)
Pseudo R2 0.07 Brant tes 14.02 (p-value = 0.14)
Note: *, **, *** represents significant at the level of 10%, 5% and 1%, respectively.
Table 7 Estimation results of model 3 with FQT as dependent variable.
Table 7variables Coef. Odds Ratio Standard error Z-score 95% Confidence Interval (OR)
GRO −0.32** 0.73 0.12 −1.96 0.53 1.00
FC −0.25 0.78 0.21 −0.92 0.46 1.33
NFM(ref:live alone)
2–4 −0.54** 0.58 0.14 −2.20 0.36 0.94
5–8 −0.76*** 0.47 0.12 −2.87 0.28 0.79
more than 8 −0.80* 0.45 0.22 −1.65 0.17 1.16
RP 0.36* 1.44 0.30 1.75 0.96 2.16
FPI 0.39*** 1.48 0.21 2.74 1.12 1.96
LA 0.66*** 1.94 0.43 2.97 1.25 3.00
LFA 0.28** 1.33 0.18 2.04 1.01 1.75
Log likelihood −868.11 LR Chi-square 46.03 (p-value = 0.00)
Pseudo R2 0.03 Brant test 4.49 (p-value = 0.88)
Note: *, **, *** represents significant at the level of 10%, 5% and 1%, respectively.
4.2.1 Estimation results for model 1 with FA (food anxiety) as dependent variable
The estimation results of Model 1 show that twelve different independent variables had a significant relationship with household food anxiety (Table 5). Firstly, the coefficients of the first two dummy variables of NFM are significant at a 5%-level, which indicates that bigger household size had a greater impact on household food anxiety. Compared to households of one person, households with the number of family members ranging from 2 to 8 had roughly two times greater odds of feeling anxiety or uncertainty about their food supply, given all other variables as constant. Second, the coefficient of RP is 0.45 with a significance of 10%, which indicated that renting a property had a higher probability of increasing the level of food anxiety than not renting. Households who rented property had 1.57 times greater odds of feeling anxiety about food, with all other variables constant. Third, the coefficient of FPI is 0.70, significant at a 1%-level, indicating that increasing food prices was a key factor that led to household food anxiety. Increasing food prices caused 2.02 times greater odds of living in a household experiencing food anxiety, holding constant all other variables. Additionally, the coefficient of LA is 0.45, significant at a 5%-level, which indicated that households who experienced limited access to offline or online stores were more likely to feel food anxiety. Households who claimed that they experienced limited access to offline or online stores during the epidemic, had the odds of feeling more anxiety about household food supply of 1.57 times than those who did not, with all other variables constant. Finally, different categories of impaired food consumption influenced a household's food anxiety to varying degrees. The coefficient of HCC, HRTC, HMEC, HFC, HMIC ranges from 0.35 to 0.65, with a significance of 5%, while the coefficient of HVC is significant at a 1%-level. This demonstrated that households had a higher likelihood to worry about food supply when it was affected by the impaired food groups including cereal, roots and tubers, vegetables, fruits, and milk and food made from milk. Meanwhile, households who claimed difficulties in obtaining roots and tubers had odds of being more likely to worry about food 1.91 times than those who did not have difficulties, given all other variables as constant.
4.2.2 Estimation results for model 2 with FQL (food quality) as dependent variable
In model 2, three different independent variables that had a significant relationship with food quality (Table 6). Firstly, the coefficient of FPI is 0.70, significant at a 1%-level, indicating that increasing food prices was more likely to result in household insufficient food quality. Households who judged food prices increased during the epidemic were 2.01 times more likely to experience insufficient food quality than those who did not affirm it, given all other variables as constant. Second, the coefficient of LA and LFA is 1.05 and 0.74, respectively, significant at a 1%-level. This indicates that households who had difficulties in conventional food patronage were more likely to encounter insufficient food quality. For households who stated that they experienced limited access to stores and limited food availability and variety in stores, respectively, the odds of facing severe insufficient food quality was 2.86 and 2.10 times that of those who did not experience these issues, holding constant all other variables.
4.2.3 Estimation results for model 3 with FQT (food quantity) as dependent variable
In model 3, the results show that five different independent variables had a significant relationship with food quantity (Table 7). The coefficient of three dummy variables of NFM was −0.54, −0.76, and −0.80, respectively, and significant at a 5%-level, 1%-level and 10%-level. This indicates that a bigger household size was beneficial to reduce the risk of insufficient food quantity. Households who had a small, medium and large family experienced insufficient food quantity, respectively, 0.58, 0.47 and 0.45 times that of those who lived alone, with all other variables constant. The coefficient of RP is 0.36 with 10%-level significance, which meant that households who rented property were more likely to experience inadequate food quantity. Households who rented property had 1.44 times greater odds of experiencing insufficient food quantity, all other variables being constant. The coefficient of FPI is 0.39 with 1%-level significance, indicating that increasing food prices was more likely to increase the severity of insufficient food quantity. Increasing food prices caused 1.48 times greater odds of living in a household experiencing inadequate food quantity, holding constant all other variables. Similarly, the coefficient of LA is 0.66 with 1%-level significance and LFA is 0.28 with 5%-level significance, which indicated that households who faced difficulties in conventional food patronage had a high likelihood to encounter insufficient food quantity. For households who experienced limited access to offline or online stores and limited food availability and variety in stores, respectively, the odds of facing more severe insufficient food quality was 1.94 and 1.33 times that of those who did not experience these limits, given all other variables as constant.
5 Discussions
5.1 Grassroots organizations contributed to ensuring household food security
The estimation results show that households who purchased food from grassroots organizations significantly reduced food anxiety and ensured the household would have adequate food quantity. The coefficient of variable GRO in models 1 and 3 is −0.42 and −0.32, respectively, with 5%-level significance (Table 5, Table 7). For households who did not purchase food from grassroots organizations, the odds of feeling more anxious about food and facing more severe insufficient food quality was 0.66 and 0.73 times that of those who did, given all other variables as constant. The results confirmed the observation that grassroots organizations helped residents decrease food anxiety and increase their access to food. In terms of food anxiety, previous studies confirmed that when households encountered inadequate food supply, they were more likely to feel stressed [53,54]. Owing to the existence of grassroots organizations in China, households had an additional approach to purchase food and realized that people care about them and provided kind care for them. This assistance from grassroots organizations greatly alleviated their anxiety about food supply and the panic of sudden lockdown. Because of community cohesion based on trust, cooperation and support, this kind of harmonious environment was also beneficial to decrease the risk of food anxiety [55,56].
Grassroots organizations provided an alternative method that increased households' access to food. Households could purchase enough food without going out, thus reducing the risk of infection by avoiding gatherings of people. Usually, households obtaining food from grassroots organizations consisted of three steps: firstly, residents submitted their food requests based on food supply lists posted by grassroots organizations in the WeChat group; then an organization's members purchased the food from wet markets or supermarkets, or even from wholesale markets; finally, the organizations delivered food to households or notified households to pick up the food in a designated spot. This kind of food purchasing approach was similar to online food purchase and delivery, but the organization members took on every job between food purchase and distribution.
5.2 Community committee played a leading role at residential community-level
Community committee is the major force of organizations in ensuring household food security at the community level with 52% of households reporting that they purchased food with assistance from a community committee, while the proportion of purchasing food from a property management company or a volunteer team was 39% and 36%, respectively (Fig. 4 ). The result is not surprising because community committee is an organization that has a tight relationship with local government. In China, community committees are a social institution led by the Communist Party of China, and the committees are encouraged to have a degree of their own autonomy and run their own affairs. In practice, the levels of government above community committee treated community committee as an anchor or leg to run grassroots affairs [57]. Meanwhile, the community committee members receive office expenses from the municipal government and the community committee accepts performance assessments from the township-level government (street office). Therefore, the city-level government required that the community committees took the responsibility of ensuring basic living services for residents during the epidemic. Moreover, community committees have the basic information on permanent resident populations in their respective complexes, which helped the community committees decide the sequence of assisting households according to the degree of obtaining food difficulty. Compared to other social organizations, the community committees adapted to the circumstances more effectively and flexibly because of greater government involvement and support. The most vital role of the community committee in ensuring household food security is that it solved the last hundred yards problem based on the community grid governance scheme [58]. Wuhan constructed the community grid governance scheme by equipping one member for one grid in 2019, and one grid has 300–500 households or 1000 permanent residents. These grid members played an important part in contacting residents and delivering goods and materials during the lockdown. Meanwhile, the community committees were effective at recruiting volunteers and mobilizing property management companies as well. Guided by community committees, roughly 24,000 volunteers took on the task of delivering food to households in need, such as the elderly and persons with disability in Wuhan [59].Fig. 4 Statistics of households re: purchasing food from community-level grassroots organizations.
Fig. 4
Furthermore, community committees were believed by residents to be the backbone force for contingency food provisioning. One question was asked in the questionnaire - “What do you perceive to be the best measures to purchase food since the outbreak of COVID-19?” More than 60% of participants shared their views and experience, and 87 participants mentioned that purchasing food through their community committee was a good channel to purchase food in an emergency situation. Community committees met their residents' daily food needs, which made a significant contribution to making residents willingly stay at home during the extended lockdown [60]. Besides the leading role of community committees, property management companies and volunteer teams played a supplementary role in ensuring household food security during the COVID-19 epidemic, reflecting their partnerships with community committees. As for volunteer teams, their major work was delivering food from the complex's gate to a household's doorstep, which reduced the workload for the community committees. The “Volunteer service care action” for preventing and controlling the epidemic in Wuhan recruited volunteers to provide local residents with daily necessities such as grain, oil, vegetables, medicines and other purchasing and delivery services, which greatly ensured the basic livelihood of residents during the lockdown period [61].
5.3 Difficulties of grassroots organizations to effectively ensure household food consumption quality
The results of model 3 show that grassroots organizations did not make a difference in ensuring households’ adequate food quality. The coefficient of variable FQL is −0.13 in model 2, but not significant (Table 6). Answers to one question in the online questionnaire were helpful to find out the reasons. The question was - “Your comments about any aspect of your personal food-related experience during the COVID-19 epidemic.” More than 200 interviewees shared their food-related experiences and opinions about contingent food purchasing activities in Wuhan. Based on the responses, the study concluded there were three aspects of problems causing inadequate food quality, including the rising prices of food, limited diversity of food, and decreased food freshness.
5.3.1 Soaring food prices and inadequate food consumption quality
Soaring food prices was a significant problem that affected the capacity and choices of household food purchase, the results of variable FPI in models 1–3 all provided the confirmation that grassroots organizations were hard-pressed to ensure sufficient food consumption quality for households (Table 5, Table 6, Table 7). Firstly, lockdown measures caused havoc on the distribution of food, and inadequate food supply in the short term made food prices increase sharply; grassroots organizations did not have the capacity to control those increasing food prices. About 60% of households said that food prices were higher than before the COVID-19 outbreak, and roughly 38% of households reported that they were spending more than twice as much on food. Meanwhile, food prices are closely related to their storability, thus the price of perishable food was more likely to increase during the lockdown [62]. Therefore, one of the reasons for inadequate food consumption quality could be attributed to decreased consumption of perishable food such as leafy vegetables. Additionally, the increase in food prices decreased the economic access to food, which led to difficulties for households to ensure adequate food consumption quality. People were asked to stay at home, which could lead to an income loss in some households. This caused a negative impact on people's purchasing capacity. For those residents or households with no or little savings, income loss could cause a more severe problem to maintain food affordability during COVID-19 epidemic period. Some responders also reported that food prices decreased after the first several weeks of lockdown, this is consistent with the findings of other studies [63]. Since the quarantine measures were implemented, more than half of households thought the most difficult period in food access they experienced was from February 1st to February 16th, while less than 10% of households faced huge difficulty in food access after March 11th. This indicated that the impact of lockdown policies on the deterioration of food consumption quality and an increase in food prices was mitigated when the reaction of government and markets rapidly changed and became more positive.
5.3.2 Fewer varieties of food and insufficient food consumption quality
Another problem is food provided by grassroots organizations did not have a rich diversity during the lockdown period. Dietary diversity is a key element of high-quality and healthy diets, which ensures the intake of essential nutrients and energy required to perform daily work efficiently. But during the lockdown period in Wuhan, it was hard to ensure rich dietary diversity [31,64]. Many interviewed households complained about the limited diversity of food. Fig. 5 showed that the consumption of different kinds of food groups was affected by the epidemic. The consumption of meat was most affected; roughly 60% of households complained that they did not have enough kinds of meat during the lockdown. Nearly 42% of households had difficulties in purchasing fish and other seafood. 38% of households reported their consumption of vegetables was affected by the lockdown, in particular leafy vegetables. Households also reported that the consumption of fruits and milk and food made from milk was limited, by 31% and 25%, respectively. The consumption of cereal, roots and tubers was less affected by the epidemic, with only 7% of households claiming that COVID-19 affected their consumption of eggs. This finding is consistent with results in Pakistan, that is, cereals remained the largest source to support daily calorie consumption for households, and non-perishable food items accounted for a large proportion [65,66]. Thus, it is reasonable that households who had difficulties in purchasing roots and tubers had the highest odds to encounter severe food anxiety (Table 5).Fig. 5 Proportion of households reported the consumption of food items were affected by the COVID-19 epidemic.
Fig. 5
5.3.3 Decreased food freshness and insufficient food consumption quality
The last reason community-based grassroots organizations did not make a difference in food consumption quality could arise from the difficulty in ensuring the freshness of food during the COVID-19 epidemic. The freshness of food matters in terms of the food being healthy, safe and tasty [67]. Therefore, freshness is the primary consideration in food shopping, and this characteristic is especially strong in southern China [68,69]. For instance, in order to meet the residents’ meat requirements, both the central government and municipal government released reserved frozen meat to market, but many surveyed households complained it was not fresh based on how it looked, smelled and felt. There were definitely some constraints that reduced the freshness of food during the lockdown in Wuhan. Firstly, there were issues with transportation capacity and delayed distribution. The people working with the grassroots organizations were perhaps less efficient due to restrictions on movement and the fear of infection, while the lack of food transport vehicles made it impossible for them to achieve the same level of efficiency of purchasing food by households before the COVID-19 outbreak. On the other hand, the frequency of food purchasing organized by grassroots organizations was limited in both timing and product diversity. Studies also show that people may have deliberately reduced the frequency of food shopping to diminish the risk of infection [70], so households chose storable food such as tubers rather than perishable leafy vegetables, which caused a loss of food freshness to some extent.
5.4 Research limitations
Findings presented in the study must be interpreted within the context of limitations. Given that the study adopted cross-sectional data, the reported findings might not be able to present the direct cause-effect relationships between grassroots organizations and household food insecurity status throughout the period of lockdown in Wuhan. Meanwhile, the movement restriction led to the study gaining interviews with households by using the method of snowball sampling, which resulted in a weak heterogeneity of participants. In particular marginalized households such as the elder who was not familiar with smartphones but lived in hardship may have been excluded, which may cause a lower proportion of food insecure households in the whole sample. The study was a quick online survey that aimed to investigate residents' specific food experiences during the lockdown, and due to travel restrictions the survey needed to be completed independently without professional guidance. It was hard to evaluate whether their understandings of questions were accurate, which led to a certain bias in the estimation result. Additionally, the socio-economic characteristics of communities was not considered in this research because of data availability, more efforts could be made to distinguish whether grassroots organizations’ coping capacity had spatial disparities or not and the influencing factors in any such disparities.
6 Conclusions
The COVID-19 crisis brought an enormous challenge to society and exposed a lot of problems within the food emergency system. Residents of Wuhan were placed under an unprecedented lockdown period of uncertainty and disrupted work, which greatly increased difficulties in purchasing food due to unstable food supply and decreased physical and economic access to food. The fear of infection and a series of lockdown measures led to traditional food retail modes not operating effectively and generated urgent demand for innovative food purchasing methods to adapt to COVID-19-related challenges. Thus, the COVID-19 epidemic provoked a remarkable surge of social public food assistance programs and organizations. Community-level grassroots organizations could be regarded as the first to perceive and react to the changes at the initial stage of the COVID-19 outbreak, in turn, residents’ daily food needs also increased their dependence on purchasing food from community organizations. Therefore, the new mode of purchasing food from non-food-based community grassroots organizations became an important feature in ensuring household food security during the epidemic in Wuhan.
Using data from an online survey conducted during March 2020 in Wuhan, the study found that the most serious problem households faced was insufficient food quality, as less than 10% of households had adequate food quality and variety during the lockdown. Meanwhile, the rate of purchasing food from community grassroots organizations was high among all households, regardless of food insecurity status. Thus, it is not surprising that the estimated results of the ordered logit model showed that community grassroots organizations played a significant role in ensuring household food security, especially in alleviating anxiety about food supply and providing enough food quantity by increasing food access and supplementing food availability. But community grassroots organizations faced challenges including soaring food prices, fewer varieties and decreased freshness of food, making it hard to protect adequate household food quality.
On the other hand, the study found that community committees were an effective tool to promote household food security in Wuhan. This finding is attributed to the administrative system in urban China, where the community committee is a grassroots mass autonomous organization in law, but has actually become an extension of the government's administrative functions, and serves as a strong liaison between the government and residents in practice. The local government put pressure on community committees to take responsibility to take care of residents, meanwhile, greater government support gave community committees the capacity to purchase and provide enough food to households. Additionally, community committees can cooperate with property management companies and property owner committees to construct work teams with plenty of human and information resources, whereas other community grassroots organizations such as volunteer teams were mainly dependent on their own initiatives and found it more difficult to operate effectively.
This study contributes to the literature of emergency food supply. To our knowledge, this is the first study to examine the relationship between purchasing food from non-food-based community grassroots organizations and household food insecurity mitigation. Unlike existing studies focusing on the overall effect [58], based on theoretical and empirical analysis, this study explored the specific influence of grassroots organizations on three domains of food insecurity including food anxiety, food quality and food quantity. This study obtained a new finding that grassroots organizations played only an important role in reducing worries about food supply and ensuring food intake, but not in ensuring food quality. This study also contributes to the empirical understanding of the role of community-level grassroots organization in large-scale emergency food provisioning.
Facing such a large-scale disaster like COVID-19, government and public organizations need to work together. The evidence and experiences of community-level grassroots organizations in Wuhan combating COVID-19 provide important lessons to adjust existing food emergency systems. Government should realize that non-food-based community-level grassroots engagement is also an essential element in combating COVID-19, as these groups had huge contributions to ensuring grassroots food security for both individuals and households. There are planned steps government managers can take to establish a community level of food emergency planning. Moreover, with roughly 30 years of development, community committees can become an important sector for urban residents in solving public affairs. A new relationship of multi-party participation including community committees, businesses and volunteer teams was constructed during the epidemic. The next step is to enhance the cooperative relationship between community committees and other major community-level grassroots organizations and then try to build sustainable and stable local food production and distribution networks by cooperating with farms and food retailers in the nearby area under the leadership of a community committee, which can improve the resilience of the community-level food system to respond to unexpected disasters.
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.
Acknowledgments
This research was funded by an operating grant from the Canadian Novel Coronavirus (COVID-19) Rapid Research Funding Opportunity with funding from the 10.13039/501100000155 Social Sciences and Humanities Research Council (Grant No. 440234).
==== Refs
References
1 Poulos N.S. Best practices among food-based community organizations: a qualitative analysis Ecol. Food Nutr. 59 1 2020 104 116 10.1080/03670244.2019.1671836 31549860
2 Sandel M. Unstable housing and caregiver and child health in renter families Pediatrics 2 2018 141 10.1542/peds.2017-2199
3 Higgins C.D. Changes in accessibility to emergency and community food services during COVID-19 and implications for low income populations in Hamilton, Ontario Soc. Sci. Med. 291 2021 114442 10.1016/j.socscimed.2021.114442 34655939
4 Tarasuk V. A survey of food bank operations in five Canadian cities BMC Publ. Health 14 2014 1234 10.1186/1471-2458-14-1234
5 Tarasuk V. Fafard St-Germain A.-A. Loopstra R. The relationship between food banks and food insecurity: insights from Canada Voluntas 31 5 2020 841 852 10.1007/s11266-019-00092-w
6 Hosler A.S. Cong X. Alharthy A. Food pantry use and its association with food environment and food acquisition behavior among urban adults J. Hunger Environ. Nutr. 16 3 2021 356 369 10.1080/19320248.2020.1848687
7 Pollard C.M. Charitable food systems' Capacity to address food insecurity: an Australian capital city audit Int. J. Environ. Res. Publ. Health 15 6 2018 1249 10.3390/ijerph15061249
8 Bacon C.M. Baker G.A. The rise of food banks and the challenge of matching food assistance with potential need: towards a spatially specific, rapid assessment approach Agric. Hum. Val. 34 4 2017 899 919 10.1007/s10460-017-9783-y
9 Simmet A. Tinnemann P. Stroebele-Benschop N. The German food bank system and its users-A cross-sectional study Int. J. Environ. Res. Publ. Health 15 7 2018 1485 10.3390/ijerph15071485
10 Palinkas L.A. A rapid Assessment of disaster preparedness Needs and Resources during the COVID-19 pandemic Int. J. Environ. Res. Publ. Health 18 2 2021 425 10.3390/ijerph18020425
11 FAO The state of food security and nutrition in the world 2021 Transforming food systems for food security, improved nutrition and affordable healthy diets for all. Rome, FAO. 2021
12 Beck D.J. Gwilym H. The food bank: a safety-net in place of welfare security in times of austerity and the covid-19 crisis Soc. Pol. Soc. 2022 10.1017/s1474746421000907
13 Simmet A. Stroebele-Benschop N. Food Bank operations during the COVID-19 pandemic in Germany. Journal of hunger & environmental nutrition 10.1080/19320248.2021.1943590 2021
14 Castro A.N. The impact of the COVID-19 pandemic on food distribution at emergency food assistance organizations in the southwestern United States: a qualitative investigation Nutrients 13 12 2021 4267 10.3390/nu13124267 34959818
15 Bene C. Global assessment of the impacts of COVID-19 on food security Global Food Security-Agriculture Policy Economics and Environment 31 2021 100575 10.1016/j.gfs.2021.100575
16 Bene C. Resilience of local food systems and links to food security - a review of some important concepts in the context of COVID-19 and other shocks Food Secur. 12 4 2020 805 822 10.1007/s12571-020-01076-1 32837646
17 Nagata J.M. Seligman H.K. Weiser S.D. Perspective: the convergence of coronavirus disease 2019 (COVID-19) and food insecurity in the United States Adv. Nutr. 12 2 2021 287 290 10.1093/advances/nmaa126 32970098
18 Clay L.A. Rogus S. Food access worries, food assistance use, purchasing behavior, and food insecurity among new yorkers during COVID-19 Front. Nutr. 8 2021 647365 10.3389/fnut.2021.647365 34513897
19 Chevee A. Mutual aid in north london during the covid-19 pandemic Soc. Mov. Stud. 2021 10.1080/14742837.2021.1890574
20 Pakravan-Charvadeh M.R. Determinants of household vulnerability to food insecurity during COVID-19 lockdown in a mid-term period in Iran Publ. Health Nutr. 24 7 2021 1619 1628 10.1017/s1368980021000318
21 Park S. Kim K. Food acquisition through private and public social networks and its relationship with household food security among various socioeconomic statuses in South Korea Nutrients 10 2 2018 121 10.3390/nu10020121 29370127
22 Arnfjord S. Greenland's emerging social conscience - Voluntary food delivery to people experiencing homelessness in Nuuk Qual. Soc. Work 20 1 2021 433 438 10.1177/1473325020973209
23 Pyle A. Disease and disaster: navigating food insecurity in a community affected by crises during COVID-19 Journal of Agriculture Food Systems and Community Development 10 3 2021 37 54 10.5304/jafscd.2021.103.005
24 Zollet S. Towards territorially embedded, Equitable and resilient food systems? Insights from grassroots Responses to COVID-19 in Italy and the City Region of Rome. Sustainability 13 5 2021 2425 10.3390/su13052425
25 Polsky J.Y. Gilmour H. Food insecurity and mental health during the COVID-19 pandemic Health Rep. 31 12 2020 3 11 10.25318/82-003-x202001200001-eng
26 Cordoba D. Peredo A.M. Chaves P. Shaping alternatives to development: solidarity and reciprocity in the Andes during COVID-19 World Dev. 139 2021 105323 10.1016/j.worlddev.2020.105323
27 Rodriguez M. More than one pandemic: how New York City community members fight against coronavirus and systematic oppression Soc. Work. Educ. 39 8 2020 983 992 10.1080/02615479.2020.1825665
28 Zhuang Y. Source separation of household waste: a case study in China Waste Manag. 28 10 2008 2022 2030 10.1016/j.wasman.2007.08.012 17920856
29 The National People's Congress of the People's Republic of China Organic Law of Urban Residents Committees of the People's Republic of China(((Amended December 29th 2018) 2019
30 Shiu J.Y. Fuzzy multicriteria decision-making tools for selecting a professional property management company Int. Trans. Oper. Res. 26 4 2019 1527 1557 10.1111/itor.12356
31 Zhong T.Y. Emergency food supplies and food security in Wuhan and Nanjing, China, during the COVID-19 pandemic: evidence from a field survey Dev. Pol. Rev. 2021 10.1111/dpr.12575
32 Wuhan Municipal Bureau of Statistics and NBS Survey Office in Wuhan Wuhan Statistical Yearbook 2021 2021 China Statistics Press 13 14
33 XINHUA NET White Paper: 4 Million Community Workers Fight in 650,000 Urban and Rural Communities in China 2020 [cited 2022; Available from:
34 Wang Y. Huang X. Wang C. Chinese WeChat users and their behavioural features: a case study based on the grounded theory Journal of Applied Business and Economics 23 5 2021 141 153
35 Desiere S. D'Haese M. Niragira S. Assessing the cross-sectional and inter-temporal validity of the household food insecurity access scale (HFIAS) in Burundi Publ. Health Nutr. 18 15 2015 2775 2785 10.1017/s1368980015000403
36 Castell G.S. Household food insecurity access scale (HFIAS) Nutr. Hosp. 31 2015 272 278 10.3305/nh.2015.31.sup3.8775 25719795
37 Coates J. Swindale A. Bilinsky P. Household Food Insecurity Access Scale (HFIAS) for Measurement of Food Access: Indicator Guide 2007
38 Xu Q. Chow J.C. Urban community in China: service, participation and development Int. J. Soc. Welfare 15 3 2006 199 208
39 Chen J. Coordinating community organizations in urban China Int. J. Soc. Welfare 18 4 2009 385 394
40 Yau Y. Collectivism and activism in housing management in Hong Kong Habitat Int. 35 2 2011 327 334 10.1016/j.habitatint.2010.11.006 32287696
41 Yau Y. Ho D.C.W. Destech Publicat I. A new paradigm for evaluating performance of property management agents 2nd International Conference on Social Science and Development (ICSSD) 2016 (Chengdu, PEOPLES R CHINA)
42 Felker-Kantor E. Wood C.H. Female-headed households and food insecurity in Brazil Food Secur. 4 4 2012 607 617 10.1007/s12571-012-0215-y
43 Fuwa N. The poverty and heterogeneity among female-headed households revisited: the case of Panama World Dev. 28 8 2000 1515 1542 10.1016/s0305–750x(00)00036-x
44 Lutomia C.K. Determinants of gender differences in household food security perceptions in the Western and Eastern regions of Kenya Cogent Food & Agriculture 5 1 2019 1694755 10.1080/23311932.2019.1694755
45 Shen K. Changing society, changing lives: Three decades of family change in China Int. J. Soc. Welfare 30 4 2021 453 464 10.1111/ijsw.12492
46 Ehebhamen O.G. Rural households' food security status and coping strategies in Edo State Nigeria International Food Research Journal 24 1 2017 333 340
47 Joshi N. Raghuvanshi R.S. Determinants of household food insecurity in rural areas of the hilly region of kumaun, uttarakhand, India: a pilot study Ecol. Food Nutr. 60 3 2021 351 376 10.1080/03670244.2020.1852228 33334196
48 McIntyre L. Homeowner versus non-homeowner differences in household food insecurity in Canada J. Hous. Built Environ. 31 2 2016 349 366 10.1007/s10901-015-9461-6
49 Anriquez G. Daidone S. Mane E. Rising food prices and undernourishment: a cross-country inquiry Food Pol. 38 2013 190 202 10.1016/j.foodpol.2012.02.010
50 Yuan L. China Thinks it Has Figured Out Online Grocery Shopping. 2017
51 Kennedy G. Ballard T. Dop M.C. Guidelines for measuring household and individual dietary diversity Food and Agriculture Organization of the United Nations 2011
52 Rita H. Komonen A. Odds ratio: an ecologically sound tool to compare proportions Ann. Zool. Fenn. 45 1 2008 66 72 10.5735/086.045.0106
53 Fitzpatrick K.M. Harris C. Drawve G. Living in the midst of fear: depressive symptomatology among US adults during the COVID-19 pandemic Depress. Anxiety 37 10 2020 957 964 10.1002/da.23080 32667117
54 Khan A.H. The impact of COVID-19 pandemic on mental health & wellbeing among home-quarantined Bangladeshi students: a cross-sectional pilot study J. Affect. Disord. 277 2020 121 128 10.1016/j.jad.2020.07.135 32818775
55 Matthews S.A. Yang T.-C. Exploring the role of the built and social neighborhood environment in moderating stress and health Ann. Behav. Med. 39 2 2010 170 183 10.1007/s12160-010-9175-7 20300905
56 Nebbitt V. Correlates of food security among low-resource young people: an assessment of community protective factors within public housing neighborhoods J. Health Care Poor Underserved 27 3 2016 1126 1142 10.1353/hpu.2016.0119 27524756
57 Xu F. Gated Communities and Migrant Enclaves: the conundrum for building 'harmonious community/shequ' J. Contemp. China 17 57 2008 633 651 10.1080/10670560802253303
58 Qian Y. Hanser A. How did Wuhan residents cope with a 76-day lockdown? Chinese Sociological Review 53 1 2021 55 86
59 Wuhan Civilization Office More than 20,000 "Care Action" Volunteers in Wuhan Are Fighting in the Frontline of Epidemic Prevention and Control 2020 Available from:
60 Cheng Y. Coproducing responses to COVID-19 with community-based organizations: lessons from zhejiang province, China Publ. Adm. Rev. 80 5 2020 866 873 10.1111/puar.13244
61 QSTHEORY The Sparks Became Torch : the Heroes of Volunteer Service in Fighting against COVID-19 Pandemic 2020 Available from:
62 Wang Y.P. Shimokawa S. Which food prices were more vulnerable to the COVID-19 blockade regulations in China? Comparison between perishable and storable vegetables J. Agribus. Dev. Emerg. Econ. 2022 10.1108/jadee-09-2021-0249
63 Yuan X.L. The changing patterns of consumers' behavior in China: a comparison during and after the COVID-19 pandemic Int. J. Environ. Res. Publ. Health 18 5 2021 2447 10.3390/ijerph18052447
64 Zhao A. Dietary diversity among Chinese residents during the COVID-19 outbreak and its associated factors Nutrients 12 6 2020 1699 10.3390/nu12061699 32517210
65 Shahbaz P. Gender-based implications of the COVID-19 pandemic on household diet diversity and nutritional security in Pakistan Br. Food J. 124 3 2022 951 967 10.1108/bfj-05-2021-0464
66 Bracale R. Vaccaro C.M. Changes in food choice following restrictive measures due to Covid-19 Nutr. Metabol. Cardiovasc. Dis. 30 9 2020 1423 1426 10.1016/j.numecd.2020.05.027
67 Evans D.M. The ontological politics of freshness: qualities of food and sustainability governance Environment and Planning a-Economy and Space 54 3 2022 461 476 10.1177/0308518x211059834
68 Zhong T.Y. The impact of proximity to wet markets and supermarkets on household dietary diversity in nanjing city, China Sustainability 10 5 2018 1465 10.3390/su10051465
69 Zhong S.R. Crang M. Zeng G.J. Constructing freshness: the vitality of wet markets in urban China Agric. Hum. Val. 37 1 2020 175 185 10.1007/s10460-019-09987-2
70 Tan S.T. Tan C.X. Tan S.S. Changes in dietary intake patterns and weight status during the COVID-19 lockdown: a cross-sectional study focusing on young adults in Malaysia Nutrients 14 2 2022 280 10.3390/nu14020280 35057460
| 0 | PMC9741495 | NO-CC CODE | 2022-12-14 23:45:31 | no | Int J Disaster Risk Reduct. 2022 Dec 10;:103490 | utf-8 | Int J Disaster Risk Reduct | 2,022 | 10.1016/j.ijdrr.2022.103490 | oa_other |
==== Front
Pediatr Neurol
Pediatr Neurol
Pediatric Neurology
0887-8994
1873-5150
Elsevier Inc.
S0887-8994(22)00268-5
10.1016/j.pediatrneurol.2022.12.003
Research Paper
Neurologic and neuroradiologic manifestations in neonates born to mothers with coronavirus disease 2019
Kurokawa Mariko a
Kurokawa Ryo a∗
Lin Ava Yun b
Capizzano Aristides A. a
Baba Akira a
Kim John a
Johnson Timothy c
Srinivasan Ashok a
Moritani Toshio a
a Division of Neuroradiology, Department of Radiology, University of Michigan, 1500 E. Medical Center Dr., Ann Arbor, MI, 48109, USA
b Division of Neurology, University of Michigan, 1500 E. Medical Center Dr., Ann Arbor, MI, 48109, USA
c Department of Biostatistics, University of Michigan School of Public Health, 1415 Washington Heights, Ann Arbor, MI, 48109, USA
∗ Corresponding author: Ryo Kurokawa, M.D., Ph.D. Division of Neuroradiology, Department of Radiology, University of Michigan, Ann Arbor, Michigan, 1500 E Medical Center Dr, UH B2, Ann Arbor, MI 48109. , Tel.: +1-784-219-2884, Fax: +1-734-615-9800
10 12 2022
10 12 2022
3 3 2022
7 11 2022
5 12 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 investigate the complications that occurred in neonates born to mothers with COVID-19, focusing on neurological and neuroradiological findings, and to compare differences associated with the presence of maternal symptoms.
Methods
Ninety neonates from 88 mothers diagnosed with coronavirus disease 2019 (COVID-19) during pregnancy were retrospectively reviewed. Neonates were divided into two groups: symptomatic (Sym-M-N, n=34) and asymptomatic mothers (Asym-M-N, n=56). The results of neurological physical examinations were compared between the groups. Data on electroencephalography, brain ultrasound, and magnetic resonance imaging abnormalities were collected for neonates with neurological abnormalities.
Results
Neurological abnormalities at birth were found in nine neonates (Sym-M-N, 7/34, 20.6%). Decreased tone was the most common physical abnormality (n=7). Preterm and very preterm birth (p < 0.01), very low birth weight (p < 0.01), or at least one neurological abnormality on physical examination (p = 0.049) were more frequent in Sym-M-N neonates . All infants with abnormalities on physical examination showed neuroradiological abnormalities. The most common neuroradiological abnormalities were intracranial hemorrhage (n=5; germinal matrix, n=2; parenchymal, n=2; intraventricular, n=1), and hypoxic brain injury (n=3).
Conclusions
Neonates born to mothers with symptomatic COVID-19 showed an increased incidence of neurological abnormalities. Most of the mothers (96.4%) were unvaccinated before the COVID-19 diagnosis. Our results highlight the importance of neurological and neuroradiological management in infants born to mothers with COVID-19 and the prevention of maternal COVID-19 infection.
Keywords
COVID-19
pregnancy
newborn
neuroradiologic manifestations
Abbreviations
Asym-M, Mothers with asymptomatic coronavirus disease 2019
Asym-M-N, Neonates born to mothers with asymptomatic coronavirus disease 2019
BMI, Body mass index
COVID-19, coronavirus disease 2019
NICU, Neonatal intensive care unit
ICU, intensive care unit
RT-PCR, real-time polymerase chain reaction
SARS-CoV-2, severe acute respiratory syndrome coronavirus 2
SpO2, Oxygen saturation
Sym-M, Mothers with symptomatic coronavirus disease 2019
Sym-M-N, Neonates born to mothers with symptomatic coronavirus disease 2019
==== Body
pmcIntroduction
In December 2019, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was first identified in Wuhan, China, and spread worldwide thereafter. The World Health Organization declared the coronavirus disease 2019 (COVID-19) outbreak a global pandemic on March 11, 2020 [1]. As of August 2022, over 598 million confirmed cases had been reported globally, with approximately 6.4 million deaths [2]. Various neonatal impacts, such as intensive care unit admission, severe infection, and severe neonatal morbidity indices [3], have been reported.
Pregnancy complicated by COVID-19, and the adverse effects of this disease on mothers and neonates, have recently been of interest [[3], [4], [5], [6]]. Pregnant women are a high-risk group for severe COVID-19 [7] and are at higher risk for intensive care unit (ICU) admission [8], severe infection, preeclampsia/eclampsia, and maternal death [3]. In addition, several factors, including increased maternal age, high body mass index [BMI], and pre-existing maternal comorbidities, increase the risk of maternal death [8].
Complications, such as preterm birth, low birth weight, and higher risk of neonatal intensive care unit (NICU) admission, have been reported for neonates born to mothers with COVID-19 [[3], [4], [5], 9]; however, the neurological complications and neuroradiological features of neonates born to mothers with COVID-19 remain largely unknown. Therefore, the purpose of the present study was to investigate the complications that occurred in neonates born to mothers with COVID-19, focusing on the neurological findings and neuroradiological features, and to compare differences associated with the presence of maternal symptoms.
Methods
Study design and participants
Between January 2020 and August 2021, all pregnant women were tested for COVID-19 in our hospital, and mothers diagnosed with COVID-19 using real-time polymerase chain reaction (RT-PCR) tests during pregnancy were included in this retrospective study. Mothers without clinical information on age, the date of the RT-PCR test, and the presence or absence of symptoms due to COVID-19, were excluded. We divided the mothers into two groups, based on the presence of symptoms: symptomatic (Sym-M) or asymptomatic (Asym-M). The neonates were divided into two groups, based on the presence of maternal symptoms: born to symptomatic (Sym-M-N) and asymptomatic mothers (Asym-M-N).
Standard protocol approvals, registration, and patient consent
We obtained an institutional review board exemption for including cases from our hospital. Data were acquired in compliance with all applicable Health Insurance Portability and Accountability Act regulations.
Data collection
The demographic and clinical data of mothers and neonates were extracted from the electronic medical records. Maternal data included maternal age, BMI, oxygen saturation (SpO2) at delivery, previous medical history, complications during pregnancy and delivery, trimester of COVID-19 diagnosis (first: 1−12 weeks, second: 13−28 weeks, third: 29−40 weeks), whether antenatal steroids were given, and associated symptoms. Neonatal data included sex, term of birth (very preterm < 32 weeks gestation; preterm = 32−37 weeks gestation; term: > 37 weeks gestation), birth weight, 1- and 5-min Apgar scores (score < 7 considered abnormal), respiratory distress, abnormal neurological findings on physical examination at birth, complications at birth, neurological imaging findings (ultrasound and magnetic resonance imaging [MRI]), and the results of COVID-19 PCR tests within 2 weeks of birth. The physical examinations and assessments were performed by pediatric neurologists who were not blinded to the mothers’ COVID-19 statuses during the study period. All the clinical data were retrospectively assessed by a board-certified radiologist under the supervision of a pediatric neurologist. To determine the presence of ‘neurological abnormalities,’ neurological physical examinations were performed at birth, and data on mental status, cranial nerves (pupils, hearing, sucking, and swallowing), motor function, and reflexes (Moro, asymmetric tonic neck reflex, and palmar grasp) were collected [10]. Furthermore, data on placental pathology were extracted from the pathology reports.
Radiological analysis
The decision to order an MRI was made by the examining physician based on the clinical findings in the neonates. All neuroradiological examinations were independently and blindly interpreted by two board-certified radiologists. Any discrepancies between the radiologists were resolved through discussion until consensus was reached.
Statistical analysis
Continuous and categorical variables are presented as means with standard deviations and rates (%), respectively. They were compared between the Sym-M-N and Asym-M-N groups, or the Sym-M and Asym-M groups, using the t-test or Mann–Whitney U test as appropriate. The Shapiro–Wilk test was performed to assess the normality of the distribution of the studied groups before each comparison of the continuous variables. When the OR was calculated between the two groups and either group contained 0, 0.5 was added to both the denominator and the numerator (Haldane–Anscombe correction [11]). We performed family-wise error correction using the false-discovery rate approach. Family-wise error-corrected two-sided p-values < 0.05 were considered statistically significant. All the statistical analyses were performed using R software (version 4.0.0; The R Foundation for Statistical Computing).
Data availability
Data were de-identified before the analysis. The underlying de-identified data can be accessed upon reasonable request from the corresponding author.
Results
Demographic and clinical characteristics of the neonates
A total of 90 neonates, including two pairs of twins, and their 88 mothers with confirmed COVID-19, were included in the analysis. The neonates were divided into Sym-M-N (n = 34; M:F = 20:14) and Asym-M-N groups (n = 56; M:F = 30:26). The demographic and clinical characteristics of the neonates are presented in Table 1 .Table 1 Descriptive data for neonates born to mothers with COVID-19
Total neonates Born to mother with symptomatic COVID-19 (Sym-M-N group, n = 34 from 33 mothers) Born to mother with asymptomatic COVID-19 (Asym-M-N group, n = 56 from 55 mothers P value
Sex (n, %)
Male : Female 50 : 40 20 : 14 30 : 26 0.81
Term of birth (n, %)
Very Preterm (<32 weeks) 7/90 (7.8%) 7/34 (20.6%) 0/56 (0%) <0.01
Preterm (32−37 weeks) 24/90 (26.7%) 13/34 (38.2%) 11/56 (19.6%)
Normal term 59/90 (65.6%) 14/34 (41.2%) 45/56 (80.4%)
Body weight (gram, mean ± SD) 3064 ± 875 2766 ± 1156 3243 ± 595 0.08
Very low birth weight (under 1500g) 5/90 (5.6%) 5/34 (14.7%) 0/56 (0%) <0.01
Infant Complications (n, %)
No complications 42/90 (46.7%) 8/34 (23.5%) 34/56 (60.7%) <0.01
Abnormal symptoms
Fever 5/90 (5.6%) 2/34 (5.9%) 3/56 (5.4%) >0.99
Vomiting 9/90 (10.0%) 7/34 (20.6%) 2/56 (3.6%) 0.047
Neurological abnormality 10/90 (11.1%) 8/34 (23.5%) 2/56 (3.6%) 0.016
Abnormal Apgar score (under 7) 30/90 (33.3%) 19/34 (55.9%) 11/56 (19.6%) <0.01
1 minute (mean ± SD) 6.1 ± 2.8 4.9 ± 3.3 6.9 ± 2.5 0.006
5 minute (mean ± SD) 7.6 ± 2.1 6.7 ± 2.3 8.2 ± 1.7 <0.001*
Meconium aspiration syndrome 3/90 (3.3%) 3/34 (8.8%) 0/56 (0%) 0.02
Small for gestational age 7/90 (7.8%) 3/34 (8.8%) 4/56 (7.1%) >0.99
NICU admission 27/90 (30.0%) 19/34 (55.8%) 8/56 (14.3%) <0.01
Respiratory distress 23/90 (25.6%) 17/34 (50.0%) 6/56 (10.7%) <0.01
Neuroimaging abnormality 9/17 (36.8%) 7/11 (63.6%) 2/6 (33.3%) 0.48
Stillbirth 1/90 (1.1%) 1/34 (2.9%) 0/56 (0%) 0.25
Maternal age at the time of delivery (mean ± SD) 29.6 ± 5.4 30.5 ± 3.5 29.1 ± 6.3 0.73
Term at the diagnosis of COVID-19 (n, %)
1st trimester (1 to 12 weeks) 1/90 (1.1%) 1/34 (2.9%) 0/55 (0%) 0.027
2nd trimester (13 to 26 weeks) 9/90 (10.0%) 7/34 (20.6%) 2/55 (3.6%)
3rd trimester (27 to the end of the pregnancy) 80/90 (88.9%) 26/34 (76.5%) 54/55 (98.2%)
Maternal COVID-19 symptom (n, %)
Admission 8/86 (9.3%) 8/30 (26.7%) 0/55 (0%) <0.01
ICU admission 6/86 (7.0%) 6/30 (20.0%) 0/55 (0%) <0.01
Invasive ventilation 5/86 (5.8%) 5/30 (16.7%) 0/55 (0%) <0.01
Death 0/86 (0%) 0/30 (0%) 0/55 (0%) >0.99
Maternal BMI (mean ± SD) 33.8 ± 7.5 34.7 ± 8.0 33.3 ± 7.2 0.76
Maternal past medical history (n, %)
Positive Smoking history 12/81 (14.8%) 8/30 (26.7%) 4/50 (8.0%) 0.09
Current Smoking 3/81 (3.7%) 2/30 (6.6%) 1/50 (2.0%) 0.73
Former Smoking 9/81 (11.1%) 6/30 (20.0%) 3/50 (1.5%) 0.12
Chronic hypertension 6/81 (7.4%) 4/30 (13.3%) 2/50 (4.0%) 0.29
Pre-existing diabetes 5/81 (6.2%) 4/30 (13.3%) 1/50 (2.0%) 0.11
Maternal complications during pregnancy (n, %)
Gestational hypertension 10/81 (12.3%) 3/30 (10.0%) 7/50 (14.0%) 0.85
Gestational diabetes 5/81 (6.2%) 4/30 (13.3%) 1/50 (2.0%) 0.11
Maternal SpO2 (%) at the time of delivery 98.3 ± 2.2 97.3 ± 3.0 98.9 ± 1.2 0.07
Maternal complications at the time of delivery (n, %)
Chorioamnionitis 6/81 (7.4%) 3/30 (10.0%) 3/50 (6.0%) 0.59
Placental abruption 3/81 (3.7%) 1/30 (3.3%) 2/50 (4.0%) >0.99
Preeclampsia 10/81 (12.3%) 4/30 (13.3%) 6/50 (12.0%) >0.99
COVID-19, Coronavirus disease 2019; SD, standard deviation; NICU, newborn intensive care unit; ICU, intensive care unit; BMI, body mass index; Inf, infinity
Neurological abnormalities were significantly more frequent in the Sym-M-N group (7/33, 20.6%) than in the Asym-M-N group (2/56, 3.6%, p = 0.049). The most common abnormality was decreased tone (7/9, 77.8%), followed by delayed or absent reflexes (5/9, 55.6%). The neurological characteristics of these nine neonates are summarized in Table 2 . Of the neonates with neurological abnormalities, six (66.7%) were born preterm or very preterm, and five (55.6%) were born by emergency cesarean section. Two neonates had status epilepticus needing three anti-seizure medications. Of the neonates with neurological abnormalities, those with decreased sulcation (patient 3) and holoprosencephaly (patient 5) were considered unrelated to their mothers’ COVID-19 infection since these abnormalities start early in the first trimester. Of the nine neonates with neurological abnormalities, electroencephalography was performed in six.Table 2 Patient characteristics
Gestational age Sex Body weight, g Infant comorbidities Mother's COVID-19 symptoms and other comorbidities Emergency c- section Mother's past medical history Time from COVID-19 diagnosis to delivery, days Apgar score, 1 min, 5 min Neurological physical examinations at birth EEG Purpose of MRI Neuroradiological findings
1 24w0d M 680 Very preterm, respiratory distress syndrome, extremely low birth weight Symptomatic (cough, sneezing, headache, and vomiting) - Diabetes 17 4, 8 ↓ tone and activity n/a n/a US (day 1) = grade 1 germinal matrix hemorrhage
2 28w0d F 975 Very preterm, respiratory distress syndrome, extremely low birth weight Symptomatic (septic shock and acute organ dysfunction) requiring ICU admission + Chronic hypertension 11 1, 6 ↓ tone and activity n/a n/a US (day 1) = grade 1 germinal matrix hemorrhage
3 31w1d F 1650 Very preterm, respiratory distress syndrome, low birth weight, MD twin, IUGR Symptomatic (shortness of breath, nasal congestion, and tachycardia) + None 51 2, 4 Hypotonic and floppy, intermittent extremity flexion Standard HIE cEEG protocol; + status epilepticus requiring three anti-seizure medications; background attenuated with severe encephalopathy n/a US (day 1) = asymmetric ventriculomegaly, ↓ sulcation
4 31w1d F 1985 Very preterm, respiratory distress syndrome, low birth weight, shoulder dystocia, Pfeiffer syndrome, MD twin, IUGR, meconium aspiration syndrome, Cloverleaf deformity, low birth weight, HIE Symptomatic (shortness of breath, nasal congestion, and tachycardia) + None 51 2, 4 Minimal constriction to light, poor grasp reflex 2-day cEEG for event capture due to concern for seizure-like activity; focal epileptiform activity without seizures; background mildly encephalopathic Clinically suspected HIE MRI (day 5) = restricted diffusion in the bilateral parietal and temporal regions with cortical laminar necrosis
5 35w1d M 3015 Preterm, respiratory failure with hypoxia and hypercapnia, multiple congenital anomalies including semi lobar holoprosencephaly, preductal coarctation of the aorta, VSD Symptomatic (acute bronchitis) - Diabetes, preeclampsia, BMI > 35 kg/m2 44 4, 7 Delayed reflex plantar stimulation, ↓ sucking reflex, hypotonia n/a Prenatally diagnosed hydrocephalus MRI (day 1) = semilobar holoprosencephaly, ventriculomegaly right frontal lobe hematoma
6 35w6d F 2440 Preterm infant, neonatal asphyxia, respiratory failure that required intubation, severe metabolic acidosis, HIE Asymptomatic, placental abruption - Current smoker 1 1, 1 ↓ tone and activity Prolonged HIE cEEG protocol due to ongoing seizures (9 days); + status epilepticus needing three anti-seizure medications; background profoundly suppressed, initially with prolonged interburst interval (burst suppression pattern) that became more continuous toward the end of the study Clinically suspected HIE MRI (day 4) = diffuse diffusion restriction in cerebral white matter, intraventricular hemorrhage
7 40w3d M 3800 Meconium aspiration syndrome Symptomatic (tachycardia and ache days) + Former smoker 14 1, 3 ↓ consciousness, sluggish pupils, no Moro reflex, no gag reflex Standard HIE cEEG protocol; multifocal epileptiform discharges without seizures; background with moderate encephalopathy due to excessive discontinuity and paucity of normal state cycling Neurological symptoms MRI (day 6) = ↓ NAA/Cr ratio on MRS
8 40w5d M 5440 Meconium aspiration syndrome, respiratory failure, HIE Symptomatic - Gestational diabetes 5 1, 5 ↓ tone and activity, constricted pupils, no gag reflex, no sucking reflex Standard HIE cEEG protocol; multifocal epileptiform discharges without seizures; background profoundly suppressed, initially with prolonged interburst interval (burst suppression pattern) that became more continuous toward the end of the study; bursts asymmetric, suggesting worse right neuronal function Clinically suspected HIE MRI (day 12) = tiny foci of prior ischemic change in the periventricular and deep white matter
9 40w6d M 3165 Respiratory failure Asymptomatic, chorioamnionitis + None 3 1, 3 Lower extremity tone poor, weakness of suck Standard HIE cEEG protocol; multifocal epileptiform discharges without seizures; background continuous with good state modulation and variability Neurological symptoms MRI (day 5) = ↓ NAA/Cr ratio on MRS
Abbreviations: BMI = body mass index; COVID-19 = coronavirus disease 2019; Cr = creatine; EEG = electroencephalography; F = female; HIE cEEG = hypoxic ischemic encephalopathy continuous EEG; ICU = intensive care unit, IUGR = intrauterine growth restriction; M = male; MD = monochorionic diamniotic; MRS = magnetic resonance spectroscopy; n/a = not applicable; NAA = n-acetylaspartate; VSD = ventricular septal defect, HIE = hypoxic ischemic encephalopathy
Abnormal Apgar scores (i.e., score < 7; p < 0.01), very preterm and preterm births (p < 0.01), NICU admission (p < 0.01), respiratory distress (p < 0.01), very low birth weight (p < 0.01), and meconium aspiration syndrome (p = 0.049) were significantly more common in the Sym-M-N group than in the Asym-M-N group. One neonate, born to a mother who had a cytokine storm in the Sym-M group and was given antenatal steroids, was born preterm, with low birth weight.
COVID-19 testing of the neonates
RT-PCR tests for COVID-19 infection were performed in 54 neonates, and three of the 18 neonates in the Sym-M-N group and one of the 36 neonates in the Asym-M-N group tested positive. All neonates with neurological abnormalities underwent RT-PCR testing for COVID-19 and showed negative results.
Maternal demographic and clinical characteristics
The gestational week in which the COVID-19 diagnosis was made was significantly earlier in the Sym-M group than in the Asym-M group (1 vs. 0, 7 vs. 2, and 26 vs. 54 in the first, second, and third trimesters, respectively; p < 0.01). In the Sym-M group, eight of the 30 mothers (26.7%) were admitted, and five (16.7%) required invasive ventilation. The mothers’ age and BMI at the time of delivery, the term at the date of COVID-19 diagnosis, past medical history (smoking history or pre-existing diabetes), and complications (gestational hypertension or diabetes, chorioamnionitis, placental abruption, and preeclampsia) were not statistically different between the two groups. In 25 of the 31 preterm birth mothers, including one mother who had a cytokine storm, antenatal steroids were given for anticipated preterm birth. Of the mothers with a known vaccination status (31/33 mothers in the Sym-M group and 52/55 mothers in the Asym-M group), all the mothers in the Sym-M group and all but three mothers in the Asym-M group were not vaccinated (vaccinated, 3/83 [3.6%]; unvaccinated, 80/83 [96.4%]). Data on placental pathology was available for 55 mothers (20/33 in the Sym-M group and 35/55 in the Asym-M group). Among the 55 mothers, SARS-CoV-2 RNA in situ hybridization was tested in 13 mothers’ placental tissues (four symptomatic mothers and nine asymptomatic mothers). All but one placental tissue showed negative results. One placenta in the Asym-M group showed rare positive cells as well as acute chorioamnionitis. The neonate born to this mother was normal and healthy.
Neuroradiological findings in neonates born to mothers with COVID-19
All of the nine neonates with neurological abnormalities had abnormal neuroradiological findings. The neuroradiological characteristics of the nine neonates are summarized in Table 2. Of the nine neonates with neuroradiological abnormalities, seven were in the Sym-M-N group. All the neonates with neurological abnormalities developed respiratory failure that needed NICU admission. Of these, four (44.4%) were very preterm, and two (22.2%) were preterm. The most common imaging feature was intracranial hemorrhage (germinal matrix hemorrhage, n = 2 [22.2%]; parenchymal hemorrhage, n = 2 [22.2%]; intraventricular hemorrhage, n = 1 [11.1%]), followed by hypoxic brain injury (diffuse white matter and magnetic resonance spectroscopy abnormalities, n = 2 [22.2%] each; Fig 1, Fig 2, Fig 3 ). Patient 5, with a right frontal lobe hematoma, was diagnosed with holoprosencephaly, and patient 6, with hypoxic-ischemic injury (HIE), was born to a mother with placental abruption.Fig 1 Representative cases with ultrasound abnormalities. A 28-weeks-and-0-days old female born to a COVID-19 symptomatic mother (Patient 2). The neonate had very low birth weight and respiratory distress syndrome. She showed decreased muscle tone and decreased activity. Ultrasound on the first day after birth shows a high-echoic lesion in the left caudate nucleus, indicating a grade 1 germinal matrix hemorrhage (A, B; arrows). A 31-weeks-and-1-day old female, born to a COVID-19 symptomatic mother (Patient 3). The neonate had meconium aspiration syndrome and low birth weight. She was hypotonic and floppy. Ultrasound on the first day after birth shows decreased sulcation indicating prematurity and ventriculomegaly (C, D). A 24-weeks-and-0-days old male born to a COVID-19 symptomatic mother (Patient 1). The neonate had neurobehavioral instability. Ultrasound on the first day of birth shows a high-echoic lesion, indicating hemorrhage in the right caudate nucleus (E, arrow) and germinal matrix (F, arrows).
Fig 2 Cases with hypoxic-ischemic encephalopathy, A 31-weeks-and-1-day old female born to a COVID-19 symptomatic mother. The neonate had minimal reaction to physical examination and a cloverleaf skull deformity (Patient 4). MRI on the fifth day after birth shows bilateral diffuse white matter abnormalities on a fluid-attenuated inversion recovery image (A, arrows), diffusion-weighted image (B, arrows), and T1-weighted image (C, arrows) with cortical T1 hyperintensity corresponding to cortical laminar necrosis (arrowheads). On the susceptibility-weighted image, a microbleed is observed in the right periventricular white matter (D, arrow). A 35-weeks-and-6-day old female born to a COVID-19 asymptomatic mother who had a placental abruption (Patient 6). The neonate had decreased muscle tone and activity. Diffuse signal abnormality in bilateral deep nuclei and white matter is observed on a T2-weighted image (E), T1-weighted image (F), and diffusion-weighted image (G). Intraventricular hemorrhage is observed in the dorsal horn of the right lateral ventricle (H; arrow; susceptibility-weighted image).
Fig 3 Cases with intraparenchymal hemorrhage. A 35-weeks-and-1-day old male born to a COVID-19 symptomatic mother (Patient 5). The neonate had delayed bilateral Babinski reflexes, neurobehavioral instability, and hypotonia. MRI on the first day after birth shows bilateral ventriculomegaly (arrowheads), semilobar holoprosencephaly, and intraparenchymal hemorrhage in the right frontal lobe (short arrows; A: T2-weighted image, B: T1-weighted image, C: diffusion-weighted image, D: susceptibility-weighted imaging). A 40-weeks-and-5-day old male born to a COVID-19 symptomatic mother (Patient 8). The neonate had pinpoint pupils and decreased muscle tone, with concern for hypoxemic-ischemic encephalopathy in the setting of meconium aspiration syndrome. MRI on the 12th day after birth shows a tiny focus of prior ischemic change in the left periventricular white matter (arrows; E: T1-weighted image, F: susceptibility-weighted image).
Discussion
This retrospective cohort study found that neurological abnormalities occurred significantly more frequently in neonates born to mothers with symptomatic COVID-19 compared to those born to asymptomatic mothers (7/34 [20.6%] vs. 2/56 [3.6%]; p = 0.049). The most common neuroradiological abnormality in the symptomatic newborns born to mothers with COVID-19 infection was intracranial hemorrhage, followed by hypoxic brain injury. Conversely, other clinical features, including age, BMI, past medical history, previous or pregnancy-related complications, and complications at the time of delivery were not statistically different between the two maternal groups. To our knowledge, the present study was the first to evaluate and compare the symptoms of the neonates by focusing on the presence or absence of symptoms in the mothers with COVID-19 during pregnancy. The results indicated the increased risk for prematurity and neurological findings in the neonates born to symptomatic mothers.
Given that neurological abnormalities were significantly more common in the Sym-M-N group, symptomatic COVID-19 infection during pregnancy may have resulted in neonatal neurological abnormalities at birth. We suggest that there are two reasons for this: preterm birth and low birth weight and/or the systemic inflammatory response induced by maternal COVID-19 infection.
Increased risks for complications related to COVID-19 and severe COVID-19 infection in pregnant women, compared to non-pregnant women, have been reported [[6], [7], [8],[12], [13], [14], [15], [16]]. The risk of ICU admission (1.1–4.2%) and the need for invasive mechanical ventilation (0.2–3%) were reported to be significantly higher among pregnant women [7,8,13,14]. For neonates born to mothers with COVID-19, increased risks of complications, such as stillbirth (0.5–1.6%), preterm birth (21–37.7%), and NICU admission (23.2–46.2%), have been reported [6,8,12,[14], [15], [16], [17]]. Previous systematic reviews have reported that 16.6% low birth weight [12], 8.1–8.3% small for gestational age [6, 18], 1.8–13.2% birth asphyxia [12, 18], and 6.4 % respiratory distress syndrome [18] occurred among neonates born to mothers with COVID-19, respectively. The high incidence of preterm and very preterm birth (31/90, 34.5%) and emergency cesarean delivery (12/90, 13.3%) in our population were consistent with the results of a previous meta-analysis [8]; therefore we considered that there were associations between maternal COVID-19 and increased incidences of preterm and very preterm birth and cesarean delivery. Moreover, the high frequency of very low birth weight neonates in the Sym-M-N group in the present study (14.7%) was consistent with previous studies, and might be the result of affected fetal growth by maternal infection, as reported in severe acute respiratory syndrome caused by SARS-CoV-1, which is genetically close to SARS-CoV-2 [19, 20]. A previous report of 201 neonates born to (mostly symptomatic) mothers with a confirmed diagnosis of SARS-CoV-2 showed slightly low Apgar scores at 1 min [21]. In contrast to these studies, respiratory distress syndrome (23/90 [25.6%]) and abnormal Apgar scores (30/90 [33.3%]) were much more frequent in the present study. These increased frequencies could be partly attributable to the patient population, which included more severe cases, represented by higher ICU admission rates, (6/86 [7%]) in the current study compared to previous studies. Another possible explanation for this discrepancy is that the previous studies were based on data before the emergence of the SARS‐CoV‐2 delta variant (B.1.617.2 lineage) [22], which has higher hospital admission and emergency care attendance risk [23] and lower efficacy of vaccines [24], compared to the alpha variant.
Maternal inflammation during pregnancy may affect fetal brain development and may lead to neuronal dysfunction and behavioral abnormalities [25]. COVID-19 is known to induce a cytokine storm and subsequent multiple organ inflammation, such as acute respiratory distress syndrome [[26], [27], [28], [29], [30]]. In the present study, one mother in the Sym-M group was diagnosed with a cytokine storm, and the neonate born to this mother was preterm and had a low birth weight (appropriate for gestational age). Cases with pathologically proven placental vasculopathy complicated with chronic villitis [31] and direct placental SARS-CoV-2 infection [24] have also been reported; however, no conclusive evidence of vertical transmission of SARS-CoV-2 to the fetus has been found. These findings indicate that inflammation in the maternal body caused by COVID-19 may have adverse implications for the development of the fetus due to increased cytokine activation and/or viral infection at the maternal-fetal interface.
Reports about the neurological and neuroradiological features of neonates born to mothers with COVID-19 are limited. A prospective cohort study from Sweden observed moderate to severe HIE in three of the 2323 infants (0.1%) born to mothers with COVID-19, and severe brain injury, including intraventricular hemorrhage grades 3-4 or cystic periventricular leukomalacia, was observed in one of the 34 infants with very preterm births (2.9%) [32]; however, information on the severity of the illness in their mothers was not available. Sukhikh et al. reported on the case of a 27-year-old woman, diagnosed with COVID-19 in the 21st week of pregnancy, who was admitted to the hospital with moderate symptoms (fever and bilateral pneumonia) [33]. Neurosonography in the 25th week of pregnancy revealed periventricular leukomalacia, intraventricular hemorrhage, and partial agenesis of the corpus callosum. The neonate died two days after birth, and the presence of SARS-CoV-2 in placental tissue and umbilical cord blood was indicated. Placental pathology revealed extensive infarctions of the placenta on the maternal and fetal surfaces.
The neuroradiological abnormalities found in the present study can be classified into the following categories: hypoxic/ischemic, hemorrhagic (germinal matrix, intraparenchymal), hypoplasia/aplasia, and others. The mechanisms of these abnormalities may be complicated by multiple contributing factors. Neonatal HIE and germinal matrix hemorrhage are typically caused by severe asphyxia and decreased gestational age respectively [34]. The incidence of HIE is 1.5 per 1000 live births [35]. The overall incidence of germinal matrix hemorrhage ranges from 20% to 25% among preterm infants [36]. The incidence of HIE (Sym-M-N group, 2/34; Asym-M-N group, 1/56; 3.0% in total) and germinal matrix hemorrhage among very low birth weight infants (Sym-M-N group, 2/5 [40%]) were higher in the present study when compared to what was reported in the literature on neonates born to mothers irrespective of COVID-19 diagnosis [38, 39]. However, the number of the present study is small, and we cannot conclude that there is a statistically significant difference. Regarding patient 5, with right frontal lobe hematoma and holoprosencephaly, and patient 6 with HIE and maternal placental abruption, the neurological and neuroradiological sequelae might not be associated with maternal COVID-19 infection. The incidence of HIE in the neonates in the present study was 3.0% (3/90 neonates: one term neonate in the Sym-M-N group, one preterm neonate in the Asym-M-N group, and one very preterm neonate in the Sym-M-N group). Although this incidence was higher than that in the term population (0.1–0.2% [37]), the high incidence of preterm and very preterm births in the present study (26.7% and 7.8%, respectively) might have affected the result. Nayak et al. reported a similar incidence of neonatal HIE (3.6%) in neonates born to mothers with COVID-19; however, whether the neonates with HIE were term or not was not described in their study [38]. Further study is necessary to clarify whether maternal COVID-19 infection increases the incidence of HIE in term, preterm, and very preterm neonates. Given the significantly higher occurrence of very preterm, preterm, and very low birth weight neonates in the Sym-M-N group, a higher occurrence of germinal matrix hemorrhage is expected due to decreased gestational age and birth weight. Two neonates in the Sym-M-N group showed ventriculomegaly in the present study (patient 3 and patient 5). Although ventriculomegaly, secondary to holoprosencephaly, is likely unrelated to maternal COVID-19 infection as mentioned above, ventriculomegaly in patient 3 could be associated with maternal COVID-19 infection, similar to the previous reported cases [39,40].
The results of the present study indicate that symptomatic maternal COVID-19 infection may increase the risk of neurological adverse events in neonates. To reduce the risk of unfavorable outcomes in neonates, various measures to prevent maternal COVID-19 infection should be taken, and pregnant women should seek prompt medical care if they are infected. Recent studies have shown that receiving an mRNA COVID-19 vaccine prior to conception or during pregnancy is not associated with an increased risk of adverse effects on pregnancy course and outcomes [41, 42]. Since vaccinations are highly effective in preventing symptomatic and asymptomatic COVID-19 infections [43], clear communication to improve mothers’ awareness of the safety and efficacy of vaccination is important.
Our study has several limitations. First, this was a single-center retrospective study with a small number of patients. Second, the risk of including cases with false-positive COVID-19 RT-PCR test results and excluding cases with false-negative test results exists in both groups. Third, information about the virus variants could not be obtained. Since the delta variant is a more contagious virus that may cause more severe illness [23], determining the correlation between neurological outcomes and the type of virus variants is desirable. Fourth, the possibility that the neurological abnormalities of the neonates were not related to the mothers’ COVID-19 but accidental complications cannot be ruled out. Furthermore, the correlation and independence of each abnormality has not been elucidated due to the lack of a sufficient number of cases. Fifth, the frequency of the neuroradiological findings might have been affected by selection bias since neuroradiological examinations were performed only in symptomatic neonates. The neuroradiological abnormalities were considered to be associated with neurological symptoms, so their frequencies were not independent of each other, but correlated. Sixth, since almost all the mothers were not vaccinated in the present study, the difference in maternal and neonatal outcomes between vaccinated and unvaccinated pregnant women is unknown. Further studies are necessary to address this issue. Seventh, we only collected data on SpO2 at the time of delivery. Continuous SpO2 monitoring may have provided a better understanding of the relationship between maternal oxygenation status and neonatal neurological outcomes.
In conclusion, the Sym-M-N group tended to have more neurological and associated neuroradiological abnormalities than the Asym-M-N group. Very preterm and preterm birth, very low birth weight, low Apgar score, NICU admission, and respiratory distress were frequently found in both groups. Almost all the mothers (96.4%) were not vaccinated before COVID-19 infection. These results highlight the importance of neurological and physical assessment, including neuroradiological examinations, in neonates born to mothers with COVID-19, as well as the importance of prevention of maternal COVID-19 infection.
Declarations
Conflicts of interest/Competing interests: All authors declare that they have no competing interests.
Acknowledgments:
None
Funding:
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declarations of interest:
None
==== Refs
References
1 Huang C. Wang Y. Li X. Ren L. Zhao J. Hu Y. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China Lancet 395 2020 497 506 10.1016/S0140-6736(20)30183-5 31986264
2 World Health Organization. Weekly epidemiological update on COVID-19 – 31 August 2022 n.d, https://www.who.int/publications/m/item/weekly-epidemiological-update-on-covid-19–-31-August-2022 [accessed August 31, 2022].
3 Villar J. Ariff S. Gunier R.B. Thiruvengadam R. Rauch S. Kholin A. Maternal and neonatal morbidity and mortality among pregnant women with and without COVID-19 infection: the INTERCOVID multinational cohort study JAMA Pediatr 175 2021 817 826 10.1001/jamapediatrics.2021.1050 33885740
4 Chen L. Li Q. Zheng D. Jiang H. Wei Y. Zou L. Clinical characteristics of pregnant women with Covid-19 in Wuhan, China N Engl J Med 382 2020 e100 10.1056/NEJMc2009226 32302077
5 Karasek D. Baer R.J. McLemore M.R. Bell A.J. Blebu B.E. Casey J.A. The association of COVID-19 infection in pregnancy with preterm birth: a retrospective cohort study in California Lancet Reg Health Am 100027 2021 10.1016/j.lana.2021.100027
6 Neef V. Buxmann H. Rabenau H.F. Zacharowski K. Raimann F.J. Characterization of neonates born to mothers with SARS-CoV-2 infection: review and meta-analysis Pediatr Neonatol 62 2021 11 20 10.1016/j.pedneo.2020.10.001 33334687
7 Zambrano L.D. Ellington S. Strid P. Galang R.R. Oduyebo T. Tong V.T. Update: characteristics of symptomatic women of reproductive age with laboratory-confirmed SARS-CoV-2 infection by pregnancy status - United States, January 22-October 3, 2020 MMWR Morb Mortal Wkly Rep 69 2020 1641 1647 10.15585/mmwr.mm6944e3 33151921
8 Allotey J. Stallings E. Bonet M. Yap M. Chatterjee S. Kew T. Clinical manifestations, risk factors, and maternal and perinatal outcomes of coronavirus disease 2019 in pregnancy: living systematic review and meta-analysis BMJ 370 2020 m3320 10.1136/bmj.m3320 32873575
9 Wei S.Q. Bilodeau-Bertrand M. Liu S. Auger N. The impact of COVID-19 on pregnancy outcomes: a systematic review and meta-analysis CMAJ 193 2021 E540 E548 10.1503/cmaj.202604 33741725
10 Yang M. Newborn neurologic examination Neurology 62 2004 E15 E17 10.1212/wnl.62.7.e15 15079049
11 Lawson R. Small sample confidence intervals for the odds ratio Communications in Statistics - Simulation and Computation 33 2004 1095 1113
12 Lassi Z.S. Ana A. Das J.K. Salam R.A. Padhani Z.A. Irfan O. A systematic review and meta-analysis of data on pregnant women with confirmed COVID-19: clinical presentation, and pregnancy and perinatal outcomes based on COVID-19 severity J Glob Health 05018 2021 11 10.7189/jogh.11.05018
13 Khan D.S.A. Pirzada A.N. Ali A. Salam R.A. Das J.K. Lassi Z.S. The differences in clinical presentation, management, and prognosis of laboratory-confirmed COVID-19 between pregnant and non-pregnant women: a systematic review and meta-analysis Int J Environ Res Public Health 5613 2021 18 10.3390/ijerph18115613
14 Ko J.Y. DeSisto C.L. Simeone R.M. Ellington S. Galang R.R. Oduyebo T. Adverse pregnancy outcomes, maternal complications, and severe illness among US delivery hospitalizations with and without a coronavirus disease 2019 (COVID-19) diagnosis Clin Infect Dis 73 2021 S24 31 10.1093/cid/ciab344 33977298
15 Mullins E. Hudak M.L. Banerjee J. Getzlaff T. Townson J. Barnette K. Pregnancy and neonatal outcomes of COVID-19: coreporting of common outcomes from PAN-COVID and AAP-SONPM registries Ultrasound Obstet Gynecol 57 2021 573 581 10.1002/uog.23619 33620113
16 Matar R. Alrahmani L. Monzer N. Debiane L.G. Berbari E. Fares J. Clinical presentation and outcomes of pregnant women with coronavirus disease 2019: a systematic review and meta-analysis Clin Infect Dis 72 2021 521 533 10.1093/cid/ciaa828 32575114
17 Jafari M. Pormohammad A. Sheikh Neshin S.A. Ghorbani S. Bose D. Alimohammadi S. Clinical characteristics and outcomes of pregnant women with COVID-19 and comparison with control patients: a systematic review and meta-analysis Rev Med Virol 31 2021 1 16 10.1002/rmv.2208
18 Yoon S.H. Kang J.M. Ahn J.G. Clinical outcomes of 201 neonates born to mothers with COVID-19: a systematic review Eur Rev Med Pharmacol Sci 24 2020 7804 7815 10.26355/eurrev_202007_22285 32744708
19 Lam C.M. Wong S.F. Leung T.N. Chow K.M. Yu W.C. Wong T.Y. A case-controlled study comparing clinical course and outcomes of pregnant and non-pregnant women with severe acute respiratory syndrome BJOG 111 2004 771 774 10.1111/j.1471-0528.2004.00199.x 15270922
20 Wong S.F. Chow K.M. Leung T.N. Ng W.F. Ng T.K. Shek C.C. Pregnancy and perinatal outcomes of women with severe acute respiratory syndrome Am J Obstet Gynecol 191 2004 292 297 10.1016/j.ajog.2003.11.019 15295381
21 Chamseddine R.S. Wahbeh F. Chervenak F. Salomon L.J. Ahmed B. Rafii A. Pregnancy and neonatal outcomes in SARS-CoV-2 infection: a systematic review J Pregnancy 4592450 2020 2020 10.1155/2020/4592450
22 Sheikh A. McMenamin J. Taylor B. Robertson C. Public Health Scotland and the EAVE II Collaborators. SARS-CoV-2 Delta VOC in Scotland: demographics, risk of hospital admission, and vaccine effectiveness Lancet 397 2021 2461 2462 10.1016/S0140-6736(21)01358-1 34139198
23 Twohig K.A. Nyberg T. Zaidi A. Thelwall S. Sinnathamby M.A. Aliabadi S. Hospital admission and emergency care attendance risk for SARS-CoV-2 delta (B.1.617.2) compared with alpha (B.1.1.7) variants of concern: a cohort study Lancet Infect Dis 22 2022 35 42 10.1016/S1473-3099(21)00475-8 34461056
24 Tao K. Tzou P.L. Nouhin J. Gupta R.K. de Oliveira T. Kosakovsky Pond S.L. The biological and clinical significance of emerging SARS-CoV-2 variants Nat Rev Genet 22 2021 757 773 10.1038/s41576-021-00408-x 34535792
25 Mor G. Aldo P. Alvero A.B. The unique immunological and microbial aspects of pregnancy Nat Rev Immunol 17 2017 469 482 10.1038/nri.2017.64 28627518
26 Marini J.J. Gattinoni L. Management of COVID-19 respiratory distress JAMA 323 2020 2329 2330 10.1001/jama.2020.6825 32329799
27 Conti P. Ronconi G. Caraffa A. Gallenga C.E. Ross R. Frydas I. Induction of pro-inflammatory cytokines (IL-1 and IL-6) and lung inflammation by Coronavirus-19 (COVI-19 or SARS-CoV-2): anti-inflammatory strategies J Biol Regul Homeost Agents 34 2020 327 331 10.23812/CONTI-E
28 Basu S. Agarwal P. Anupurba S. Shukla R. Kumar A. Elevated plasma and cerebrospinal fluid interleukin-1 beta and tumor necrosis factor-alpha concentration and combined outcome of death or abnormal neuroimaging in preterm neonates with early-onset clinical sepsis J Perinatol 35 2015 855 861 10.1038/jp.2015.86 26226245
29 Sherer M.L. Lei J. Creisher P.S. Jang M. Reddy R. Voegtline K. Pregnancy alters interleukin-1 beta expression and antiviral antibody responses during severe acute respiratory syndrome coronavirus 2 infection Am J Obstet Gynecol 225 2021 301 10.1016/j.ajog.2021.03.028 e1–4
30 Garcia-Flores V. Romero R. Xu Y. Theis K.R. Arenas-Hernandez M. Miller D. Maternal-fetal immune responses in pregnant women infected with SARS-CoV-2 Nat Commun 320 2022 13 10.1038/s41467-021-27745-z
31 Hsu A.L. Guan M. Johannesen E. Stephens A.J. Khaleel N. Kagan N. Placental SARS-CoV-2 in a pregnant woman with mild COVID-19 disease J Med Virol 93 2021 1038 1044 10.1002/jmv.26386 32749712
32 Norman M. Navér L. Söderling J. Ahlberg M. Hervius Askling H. Aronsson B. Association of maternal SARS-CoV-2 infection in pregnancy with neonatal outcomes JAMA 325 2021 2076 2086 10.1001/jama.2021.5775 33914014
33 Sukhikh G. Petrova U. Prikhodko A. Starodubtseva N. Chingin K. Chen H. Vertical transmission of SARS-CoV-2 in second trimester associated with severe neonatal pathology Viruses 447 2021 13 https://doi.org/10.3390/v13030447
34 Tan A.P. Svrckova P. Cowan F. Chong W.K. Mankad K. Intracranial hemorrhage in neonates: a review of etiologies, patterns and predicted clinical outcomes Eur J Paediatr Neurol 22 2018 690 717 10.1016/j.ejpn.2018.04.008 29731328
35 Kurinczuk J.J. White-Koning M. Badawi N. Epidemiology of neonatal encephalopathy and hypoxic–ischaemic encephalopathy Early Hum Dev 86 2010 329 338 10.1016/j.earlhumdev.2010.05.010 20554402
36 Parodi A. Govaert P. Horsch S. Bravo M.C. Ramenghi L.A. eurUS.brain group. Cranial ultrasound findings in preterm germinal matrix haemorrhage, sequelae and outcome Pediatr Res 87 2020 13 24 10.1038/s41390-020-0780-2
37 Gopagondanahalli K.R. Li J. Fahey M.C. Hunt R.W. Jenkin G. Miller S.L. Preterm hypoxic-ischemic encephalopathy Front Pediatr 114 2016 4 10.3389/fped.2016.00114
38 Nayak M.K. Panda S.K. Panda S.S. Rath S. Ghosh A. Mohakud N.K. Neonatal outcomes of pregnant women with COVID-19 in a developing country setup Pediatr Neonatol 62 2021 499 505 10.1016/j.pedneo.2021.05.004 34147430
39 Düppers A.L. Bohnhorst B. Bültmann E. Schulz T. Higgins-Wood L. von Kaisenberg C.S. Severe fetal brain damage subsequent to acute maternal hypoxemic deterioration in COVID-19 Ultrasound Obstet Gynecol 58 2021 490 491 10.1002/uog.23744 34319630
40 Archuleta C. Wade C. Micetic B. Tian A. Mody K. Maternal COVID-19 infection and possible associated adverse neurological fetal outcomes, two case reports Am J Perinatol 39 2022 1292 1298 10.1055/a-1704-1929 34814196
41 Wainstock T. Yoles I. Sergienko R. Sheiner E. Prenatal maternal COVID-19 vaccination and pregnancy outcomes Vaccine 39 2021 6037 6040 10.1016/j.vaccine.2021.09.012 34531079
42 Blakeway H. Prasad S. Kalafat E. Heath P.T. Ladhani S.N. Le Doare K. COVID-19 vaccination during pregnancy: coverage and safety Am J Obstet Gynecol 226 2022 236 10.1016/j.ajog.2021.08.007 e1-4
43 Haas E.J. Angulo F.J. McLaughlin J.M. Anis E. Singer S.R. Khan F. Impact and effectiveness of mRNA BNT162b2 vaccine against SARS-CoV-2 infections and COVID-19 cases, hospitalisations, and deaths following a nationwide vaccination campaign in Israel: an observational study using national surveillance data Lancet 397 2021 1819 1829 10.1016/S0140-6736(21)00947-8 33964222
| 0 | PMC9741496 | NO-CC CODE | 2022-12-14 23:29:55 | no | Pediatr Neurol. 2022 Dec 10; doi: 10.1016/j.pediatrneurol.2022.12.003 | utf-8 | Pediatr Neurol | 2,022 | 10.1016/j.pediatrneurol.2022.12.003 | oa_other |
==== Front
Environ Adv
Environ Adv
Environmental Advances
2666-7657
Elsevier Ltd
S2666-7657(22)00163-6
10.1016/j.envadv.2022.100328
100328
Article
Impact assessment of COVID-19 global pandemic on water, environment, and humans
Raza Taqi a⁎
Shehzad Muhammad b
Abbas Mazahir c
Eash Neal S. a
Jatav Hanuman Singh de
Sillanpaa Mika f
Flynn Trevan g
a Department of Biosystems Engineering & Soil Science, University of Tennessee, USA
b Government College University Lahore, Pakistan
c Department of Bioscience, University of Wah Cantt, Quaid Avenue, Wah Cantt 47040, Pakistan
d Department of Soil Science and Agricultural Chemistry, Sri Karan Narendra Agriculture University, Rajasthan 303329, India
e Department of Soil Science and Agricultural Chemistry, Institute of Agricultural Sciences, Banaras Hindu University, Varanasi 221005, India
f Department of Chemical Engineering, School of Mining, Metallurgy and Chemical Engineering, University of Johannesburg, P.O. Box 17011, Doornfontein 2028, South Africa
g Department of Horticulture and Natural Resources, University of Bonn, Germany
⁎ Corresponding author.
10 12 2022
4 2023
10 12 2022
11 100328100328
16 7 2022
15 11 2022
4 12 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.
One of the most significant threats to global health since the Second World War is the COVID-19 pandemic. Due to COVID-19 widespread social, environmental, economic, and health concerns. Other unfavourable factors also emerged, including increased trash brought on by high consumption of packaged foods, takeout meals, packaging from online shopping, and the one-time use of plastic products. Due to labour shortages and residents staying at home during mandatory lockdowns, city municipal administrations' collection and recycling capacities have decreased, frequently damaging the environment (air, water, and soil) and ecological and human systems. The COVID-19 challenges are more pronounced in unofficial settlements of developing nations, particularly for developing nations of the world, as their fundamental necessities, such as air quality, water quality, trash collection, sanitation, and home security, are either non-existent or difficult to obtain. According to reports, during the pandemic's peak days (20 August 2021 (741 K cases), 8 million tonnes of plastic garbage were created globally, and 25 thousand tonnes of this waste found its way into the ocean. This thorough analysis attempts to assess the indirect effects of COVID-19 on the environment, human systems, and water quality that pose dangers to people and potential remedies. Strong national initiatives could facilitate international efforts to attain environmental sustainability goals. Significant policies should be formulated like good quality air, pollution reduction, waste management, better sanitation system, and personal hygiene. This review paper also elaborated that further investigations are needed to investigate the magnitude of impact and other related factors for enhancement of human understanding of ecosystem to manage the water, environment and human encounter problems during epidemics/pandemics in near future.
Keywords
COVID-19
Environmental concern
Water and air pollution
Environmental imbalance
Health risks
==== Body
pmc1 Introduction
The historical background of human coronaviruses started in 1965 when Tyrrell and Bynoe discovered a virus called B-814 (Tyrrell and Bynoe, 1965). The virus was present in human embryonic tracheal organ cultures taken from an adult's respiratory tract with common cold symptoms. Hence, it was not surprising when a new, Severe Acute Respiratory Syndrome (SARS) appeared in the form of the coronavirus from southern China and spread all around the world with haste in 2002 and 2003 (Drosten et al., 2003; Ksiazek et al., 2003; Peiris et al., 2003). SARS was reported in twenty-nine countries in South America, North America, Asia, and Europe during the 2002-2003 outbreaks. Global virus strain outbreaks from the past 50 years are shown in Table 1 . When a person is in close proximity (within 1 m) to someone who is coughing or sneezing, they are in danger of having their mucosae (mouth and nose) or conjunctiva (eyes) exposed to potentially infectious respiratory droplets. Fomites in the area surrounding the afflicted person may also transmit the disease. As a result, the COVID-19 virus can spread through direct contact with infected people and indirect contact with nearby surfaces or. Droplet transmission refers to the presence of microbes within droplet nuclei, typically particles with a diameter of less than 5 µm. On the other hand, airborne transmission implies that there are microbes in the air that can be transported more than 1 m and infect others (Shah et al., 2022).Table 1 Global virus strain outbreaks from the past 50 years.
Table 1Strain/Virus Year Patient Illness Sources
B814 1960 Youth Common cold Tyrrell and Bynoe (1965)
229E 1962 Young adult Minor URI (Upper Respiratory infection) Hamre and Procknow (1966)
OC43 1966 Young adult URI McIntosh et al. (1967b)
692 1966 29 year, male URI Kapikian et al. (1973)
NL63 2004 7-month old child conjunctivitis and bronchiolitis Fouchier et al. (2004); van der et al. (2004)
HKUI 2005 71-year-old man pneumonia Woo et al. (2005)
SARS-CoV 2003 Humans Respiratory infection Rota et al. (2003)
SARS-CoV-2 2019 More adult than children Respiratory infection WHO (2020)
SARS Variant
Alpha B.1.1.7 Sep, 2020 Variation in Patients Respiratory infection WHO (2020)
Beta B.1.351 May, 2020 Variation in Patients Respiratory infection WHO (2020)
Gamma P.1 Nov, 2020 Variation in Patients Respiratory infection WHO (2020)
Delta B.1.617.2 Oct, 2020 Adult Respiratory infection WHO (2020)
Omicron B.1..1.529 Nov, 2021 Child Respiratory infection WHO (2020)
Lambda C.37 Dec, 2020 Older Respiratory infection WHO (2020)
Mu B.1.621 Jan, 2021 Variation in Patients Respiratory infection WHO (2021)
Data Updated on 6th July 2022 Source: https://covid19.who.int/
Initially, the novel coronavirus known as SARS-CoV-2 was identified in Wuhan, China, and dangerously spread in the large region of Hubei and then across China. At an early stage, the world was informed about the first cases; however, many countries did not expect the immediate health risk the virus presented. At the end of January 2020, a novel disease caused by SARS-CoV-2 was named COVID-19. Later, the World Health Organization (WHO) confirmed that the disease was a public health emergency (Ali et al., 2020) and characterized it as a contagious pandemic on March 11, 2020 (Green, 2020). Globally, 205 million cases and 4.3 million deaths due to COVID-19 were reported on August 10, 2021 (STATISTA 2021). A new variant of the SARS-COVID was reported to the WHO from South Africa on 24th November 2021. Another recent variant was the Delta variant (Omicron), officially designated as B.1.1.529, whose specimen first collected on 9th November 2022 (Fidan et al., 2022). Data on the spread of COVID-19 has been gathered from different countries worldwide. The data showed that the United States had the most significant number of infections, followed by India, Brazil, etc. The latest global statistics on the COVID-19 pandemic can be seen in the supplementary file (Organization, 2020). Individuals with COVID-19 have many symptoms, ranging from mild indications to deadly diseases. One to two days after contact with the virus, symptoms may appear. Individuals with these indications, such as chills or fever, shortness of breath, dry cough, body or muscle aches, fatigue, sore throat, headache, the loss of smell or taste, runny nose or congestion, diarrhoea, vomiting or nausea, may have COVID-19 (Figs. 1 and 2 ).Fig. 1 Systematic representation of Coronavirus (COVID-19) infection symptoms of different organs in the human body.
Fig 1
Fig. 2 Stages of Coronavirus (COVID-19) infection symptoms in humans.
Fig 2
Starting in China, the virus spread to Thailand, Iran, South Korea, and Japan and then into Europe through air travel (Cereda et al., 2020). Initially, the most affected country in Europe was Italy, especially the region of Lombardy, and the cases started to multiply without much knowledge of what was present. Later, the outbreak entered the America (Cereda et al., 2020). Due to air travel, countries are interconnected and depend on different regions to conduct business. Hence, the virus spread rapidly around the world. The United States' response to the COVID-19 pandemic has caused several unexpected economic and market effects (Nilashi et al., 2020). According to a New York Times report (Gössling et al., 2020). For example, daily life has been disturbed for many months, and a significant portion of the population has more health issues (FITBIT, 2020). Strict action by the government made restrictions such as banning movement (e.g., travel restrictions, curfews, etc.), and the number of daily life opportunities declined (e.g., stores closed, restaurants closed, health/fitness centres closed, etc.). Prices of goods increased or became difficult to obtain (e.g., food). The combined effect increased the negative impact on the environment, daily life (e.g., food security and livelihood), water quality, and the economy at a global level. Social distance and especially hygiene were marked as a standard operating procedure (SOP) for rapid protection against COVID-19 by the WHO. It has brought a number of changes to the environment in different countries. It has been recorded that air pollution has been reduced due to a significant reduction in toxic gas emissions (NOx, CO, and other gases) in most countries with restrictions (Nilashi et al., 2020; Wang et al., 2020). The pandemic of COVID-19 has also indirectly impacted animal life, especially wildlife. Considering direct or indirect effects, actions are needed to control the impact of COVID-19. Lal et al. (2020a) reported that essential factors such as air pollution, temperature, and humidity are directly affected by the expansion of COVID-19 (Lal et al., 2020).
This study hypothesizes that the COVID-19 pandemic significantly impacts water quality, the environment, land pollution, the sanitation system, and human lives differently. One of the most important factors is the unawareness or lack of information on the impact of COVID-19. Thus, there is a need to study the critical impacts of COVID-19 in detail to generate fundamental knowledge of the factors that affect environmental sustainability in different ways. Finally, the findings of this review paper will help decision-makers and policymakers count the challenges and develop new regulations to achieve environmental sustainability.
2 Methodology
The methodology of this study was to summarise data related to the environment (air, soil, and water), human systems, and COVID-19 and to develop the links between them. The data used for this review has been collected from different sources that provided global information related to COVID-19. The essential literature sites used were Scopus, Web of Science, Google Scholar, Research Gate, Science Direct, official updated data by the World Metrology Organization (WMO), and the WHO. The collected data comprised observations, research articles, viewpoints, and information obtained during the pandemic.
3 World region economic cost situation
Table 2 shows that European regions have the maximum number of infected individuals and the highest death rate per 100,000 individuals, followed by the Americas, the Eastern Mediterranean region, and so on (WHO, 2020). At first, the discernment was that COVID-19 would be limited to China, but later on it spread worldwide via individual movements. Financial crises occurred because individuals were advised to remain at home. The seriousness was felt in different economic zones, with travel bans influencing the sports industry, aviation industry, agriculture and mass-crowd prohibition influencing the entertainment and events industries (Larry Elliot, 2020). Amidst the worldwide turbulence, in a primary evaluation, the International Monetary Fund (IMF) anticipated China's economic growth rate to decrease by 0.4% when China's development targeted a 5.6% growth rate. Hence, the worldwide economy decreased by 0.1% (Aguiar et al., 2019). The Chinese economy represents around 16% of the world's Gross Domestic Product (GDP), and it is the most prominent exchange companion of most African nations and Southeast Asia (Aguiar et al., 2019). The Organization for Economic Co-operation and Development (OECD) estimated the reduction in economic development rates as follows: China will grow at a rate of 4.9% rather than 5.7%, Europe will grow at a rate of 0.8% rather than 1.1%, and the rest of the world will grow at a rate of 2.4% rather than 2.9%, with global GDP falling by 0.41% in the first quarter of 2020. The United Nations Conference on Trade and Development (UNCTAD) stated a downward trend of 5% to 15% in foreign direct investments. On March 23, 2020, the IMF announced that since the beginning of the crisis, investors had pulled back $83 billion from emerging markets (Chepeliev, 2020). The coronavirus occurrence has led numerous governments to compel limitations on unnecessary journeys. Even certain countries set an entire travel prohibition on a wide range of internal and external travel, shutting down all airports in the country. At the peak of the coronavirus outbreak, most aircraft flew approximately devoid of passengers because of mass travel boycotts and prohibition (Jagannathan et al., 2013). The travel limitations forced by governments afterward provoked a lessening enthusiasm for a wide range of travelwhich compelled some companies to briefly suspend activities such as Line of Travel (Rolland et al., 2012), Polish Airlines, Air Baltic, Scandinavian and La Compagnie Airlines. According to the International Air Transport Association (IATA), such travel restrictions cost the tourism industry over $200 billion globally, and excluding other income losses for the industry, cost the flying industry $113 billion. US airline companies received $50 billion in bailout money to combat such losses. The Global Business Travel Association (GTBA) estimated that the commercial travel industry would lose $820 billion in revenue due to the coronavirus pandemic.Table 2 COVID-19 global pandemic statistics showing the total number of cases, cases per 100,000 people, total deaths, deaths per 100,000 people and major transmission types at the regional level according to the World Health Organization (WHO).
Table 2Region Total cumulative cases Cumulative cases per 100,000 people Total cumulative deaths Cumulative deaths per 100,000 people Major transmission type
Africa 9,138,808 32,456.9 78,140 469.54 Community transmission
Americas 163,881,638 26218.1 1008832 2127 Community transmission; clusters of cases
Eastern Mediterranean 7,564,704 59,129.59 158,961 651.09 Community transmission
Europe 230,320,699 351,594.09 969,779 6,346.4 Community transmission; clusters of cases
South-East Asia 58,692,921 8,110.85 219,797 63.83 Community transmission; clusters of cases
Western Pacific 64,851,400 19,050.85 31,685 202.19 Sporadic cases; clusters of cases
Others 745 0 13 0
Source: (WHO, 06 July 2022) https://covid19.who.int/table.
The COVID-19 outbreak has also tested the resilience of the agricultural division. Globally, the closing of restaurants and hotels has declined the consumption of agricultural commodities by 20% (Bhosale, 2020). Upon interaction with speculated carriers of the virus, recommendations on self-quarantine will most likely influence the number of available inspectors and distribution staff critical to ensuring affirmation and transportation of products. Besides this, markets have also shuttered down the trading of commodities, drastically influencing The American Veterinary Medical Association (AVMA) has conveyed anxiety over low levels of animal pharmaceuticals for a couple of large prescription suppliers (Jagannathan et al., 2013). Several hotels in the United Kingdom, the United States, and some European countries announced a temporary exercise halt. It was estimated that 24.3 million jobs will be lost, including 3.9 million in the US alone, because of the reduction in hotel guests during the outbreak. Due to disruption due to self-isolation policies and supply chains, importation problems and staffing insufficiencies were the main issues for businesses (Barro, 2015). The chemical industry is estimated to decrease its global production by 1.2%, and the economy has gone toward a drastic loss (Bentolila et al., 2018). Central chemical manufacturing enterprises like Baden Aniline and Soda Factory (BASF) have been required to defer their production, adding to a slowdown in expected development (Bezemer, 2011). The February 2020 announcement on the Chinese significant workshop action and market manufacturing index that the Purchasing Managers Index (PMI) recorded a minimum at that time (El-Erian, 2020). "China's industrial economy was influenced by an outbreak a month ago," said Zhengsheng Zhong, principal economist at the group Centre for Evidence-based Medicine (CEBM) (Horowit, 2020). While the Chinese labor force is returning to work, the PMIs across East Asia have shown sharp decreases, especially in Taiwan, Japan, Vietnam, and South Korea (Wang et al., 2020). The sports industry was also seriously affected throughout the outbreak. Many other important markets, for instance, oil, coal, and sustainable energy also significantly influenced by the pandemic. The coronavirus occurrence also influenced the supply chain of pharmaceuticals. Before the coronavirus outbreak, 60% of active pharmaceutical components were made in China, and the pandemic caused severe supply chain problems. Therefore, China closed many of its workshops that produce drugs. Hence, every economic sector has been directly and indirectly affected due to the COVID-19 pandemic.
4 Socioeconomic and environmental aspects of COVID-19
Socioeconomic factors have shown that the effect of COVID-19 is not the same for everyone (Collivignarelli et al., 2020). It is difficult to know why COVID-19 varies from person to person and among different socioeconomic populations. The most critical elements affecting an individual's socioeconomic status are their educational level, population density (rural and urban settings), lifestyle, and the size of houses and number of members in a household. Unfortunately, it was found that the spread of cases was higher in poor and less developed regions (Messner, 2020). The current literature on COVID-19 shows that in populations with a lower socioeconomic status, COVID-19 cases are higher with higher infection rates than in populations in higher income areas. The New York City research report showed that less developed residential areas have a significantly higher infection rate than other areas of the same city.
5 COVID-19 and the global environment
The pandemic has significantly impacted the environment over a short duration. Almost all countries limited international flights to combat the spread of the virus. It has been stated that the resources of a country play an essential role in implementing preventive or control measures (Zhu et al., 2020). However, the situation is worse in developing countries, where the population density is usually high. In such countries, social distancing is minimized due to huge gatherings and a lack of public healthcare services, which cause high infection rates. Similarly, worldwide institutions should have followed strict SOPs to reduce infections and minimize the 2nd and 3rd waves of COVID-19. Historical analysis shows that the 2nd wave of COVID-19 was more prominent and lethal than the 1st wave, even though cases were reduced. The pandemic has changed the world's social life and economic status immensely. Every country has implemented different strategies to combat this pandemic (Debiec-Andrzejewska et al., 2020). However, a short-term shutdown of industry, business, and transportation has reduced the emission of greenhouse gases (GHG) compared to previous years. Air quality is a crucial element of a healthy life; however, about 91% of the world's population lives in places where the low air quality surpasses the allowable limit (Zhang et al., 2017). The report showed that the most influential countries belong to Asia, Africa, and a part of Europe, with an average air quality of 8% (WHO, 2021). It has been reported that GHGs dropped by 40% and 50% in China and New York City, respectively, with a global drop of 25% in GHGs during the 3-month peak of COVID-19 (Gautam, 2020). In another study in Salé (North-Western Morocco), NO2 concentrations decreased by 96% (Ocak and Turalioglu, 2008).
Environmental factors incorporate air pollution, noise pollution, water pollution, depletion of the ozone layer, climate change, groundwater level depletion, arsenic contamination, biodiversity, ecosystem change, etc. (Bremer et al., 2019). Due to the increasing GHG emissions (CO2, N2O, CH4, etc.), global warming occurs, and these gases are also called the driving forces of air pollution (Lyu et al., 2016). The Air Quality Index (AQI) and health breakpoints for different pollutants, along with possible impacts, are presented in Tables 3 and 4 . The presence of NO2, PM10, and SO2 in urban environments causes air pollution and is an indirect cause of severe health issues such as cardiovascular and respiratory diseases, lung cancer, and hypertension (Koken et al., 2003; Le Tertre et al., 2002). These pollutants emerge from anthropogenic sources including suspended particles from wheel turbulence, street traffic, and industrial activities (He et al., 2019; Thorpe and Harrison, 2008). According to studies, coal combustion produces high air pollutants such as particulate matter (PM), NO2, CO2, SO2, and heavy metals (Munawer, 2018). Additionally, the buildup of these contaminants in the water and air causes a hazard to ecological systems (Raza et al., 2022).Table 3 Air Quality Index (AQI) categories and health breakpoints for different pollutant. Pollutants include particulate matter < 10 μm (PM10) and < 2.5 μm (PM2.5), nitrogen dioxide (NO2), ozone(O3), carbon monoxide (CO), sulfur dioxide (SO2), ammonia (NH3) and lead (Pb).
Table 3Image, table 3
*CO in mg/m3 and other pollutants in µg/m3; 2 hourly average values for PM10, PM2.5, NO2, SO2, NH3 and Pb; 8-hourly values for CO and O3; (The Colour Green to Red shows increasing toxicity level).
Table 4 Air Quality Index (AQI) and possible health impacts of air standards on human health.
Table 4Image, table 4
(The Colour Green to Red shows increasing toxicity level).
5.1 Nitrogen dioxide (NO2)
Nitrogen dioxide is a significant source of pollution, which is not only injurious to human health-but also significantly affects biodiversity (Reale et al., 2019; Verburg and Osseweijer, 2019). Long-term exposure to high NO2 concentrations disturbs the respiratory system and causes many other harmful lung diseases, asthma, skin problems, etc. (Arden et al., 2004). Short-term exposure to high NO2 concentrations can cause respiratory indications and intensify respiratory maladies. Additionally, it has been recorded that short and long-term exposure increased mortality rates. It has also been stated that about 2.6 million people are affected by bad air quality (Cohen et al., 2017). Nitrogen dioxide and nitrogen monoxide (NO) cause smog and acid rain, damaging the environment (Akimoto, 2003). Nitrogen dioxide is known as a NO marker and is utilized to evaluate stages of environmental pollution (Wang et al., 2018). According to an Environmental Site Assessment (ESA) report, NO2 was reduced by 20-30% in France, Italy, and Spain during lockdown measures. It was also noted that NO2 reduction was higher in Asian countries than in European countries during the lockdown period. Asian countries have dealt with the pandemic uniquely, and Asian countries adopted travel restrictions and lockdown measures much earlier than the rest of the world. It is thought that cases and death rates were lower in Asian countries due to early restrictions compared to European countries. The positive air quality during the pandemic and unwanted lockdown is an indication for individuals and researchers/scientists to improve the quality of the environment by reducing NO2 levels in the air (Zhou et al., 2020).
5.2 Particulate matter (PMx)
Air is considered a medium to transfer viruses, bacteria and other organisms into the environment (Piazzalunga-Expert, 2020; Zhou et al., 2020). Recently, it has been reported that the presence of PM in the air may increase the spread of COVID-19 (Barakat et al., 2020). Particulate matter is an air-suspended mixture of liquid and solid particles which are hazardous to one's health. Their size varies from particle to particle based on the diameter of the matter. The most notable of these sizes are PM10 and PM2.5, with diameters of less than 10 and 2.5 micrometers, respectively (Khan et al., 2019; Zoran et al., 2019). These fine microparticles are categorized as toxic elements for human health and are produced in combustion engines. Particulate matter is small enough to be inhaled and causes severe health-related problems, such as asthma and other lethal effects (Zhang et al., 2019). Particulate matter's toxicity multiplies when it is attached or absorbed to a surface. Worldwide, the relationship between PM2.5 and health is well known. These effects include contagious and chronic respiratory maladies, neurocognitive diseases, cardiovascular diseases, and pregnancy risks (Brook et al., 2004; Hart et al., 2015; Kandari and Kumar, 2021).
China, the UK, Brazil, Italy, and the USA are the most affected countries by COVID-19, and there is a strong link between the level of cases and air pollution. With this background, several studies have been conducted to find the correlation between COVID-19 and air pollution. A positive correlation exists between air quality and COVID-19 infections in Italy. However, this correlation varies from country to country (Comunian et al., 2020). It was concluded that the poor quality of air and the higher PM2.5 caused a high mortality rate. Other studies in Italy found the same relationship (Conticini et al., 2020; Fattorini and Regoli, 2020). It has been recorded that the concentration of PM2.5 and PM10 increased substantially compared to the last four years and reached a chronic level in Northern Italy, where the death rate from COVID-19 also increased. It can be concluded from this paper that air quality was correlated with COVID-19 cases and long-term exposure to such poor air quality increased the spread of the virus and hence, confirmed cases. In the USA, a cross-sectional examination by Kandari and Kumar (2021) showed that an exposure of 1 g m3 PM2.5 for longstanding contact is related to an 8% rise in the mortality rate of COVID-19 patients (Kandari and Kumar, 2021). As a result, increased exposure to PM2.5 increased COVID-19 deaths because it increased PM concentration by 11, causing mortality. Another study reported that within a few days after COVID-19 countermeasures in Salé, Morocco, PM10 concentrations were reduced by 49% (Ocak and Turalioglu, 2008). Mandal and Pal (2020) observed a decline in PM10 concentrations from 189-278 g m3 to 50-60 g m3 in Eastern India (Mandal and Pal, 2020). It is reported that there was a notable decrease of 85.1% in PM2.5 in the most polluted city (Ghaziabad) in India during the three months before restrictions were put in place (Lokhandwala and Gautam, 2020). A decline in PM2.5 and PM10 was generally much lower in urban Europe (8%) than in Wuhan (42%) in both relative change and magnitude (Mahato et al., 2020). During COVID-19 restrictions, PM2.5 concentrations were reduced by 17% across Europe. Lastly, a study by Pansini and Fornacca (2021) found that the USA, China, UK, Germany, France, Spain, Iran, and Italy air quality index showed a statistically positive and significant relationship between COVID-19 infection and the level of pollution in the air. The presence of PM indicates the quality of air and related health issues for future strategies.
5.3 Sulfur dioxide (SO2)
Sulfur dioxide is an important air pollutant linked with coal, oil, and chemical emissions, and it is the main indication for the formation of different hazardous particles in the atmosphere. These particles are continuously rising due to human activities which are abundant in urban areas (Kulmala et al., 2004). A study in Salé, Mo, reported that SO2 concentrations were reduced by 49% during three months under Covid-19 restrictions. During January, February, and March of 2020, based on data collected by Xu et al. (2020) from the three cities, Wuhan, Enshi, and Jingmen in China, the average SO2 concentration decreased by 45%, 30%, 33% compared to the same months from 2017 to 2019 (Yang et al., 2018). The US and Spain experienced a similar outcome. In Korea and Germany, SO2 concentrations were reduced by about 20 to 30% compared to the same month in 2019 (Doumbia et al., 2021).
5.4 Carbon monoxide and dioxide (CO/CO2)
Carbon monoxide is one of the significant markers of air pollutants created by incomplete combustion, such as automobile exhaust and fuel combustion. Thus, carbon emissions are divided into two categories; one is direct sources due to energy use like combustion, and the second is indirect sources due to the usage of non-energy products (Guo et al., 2018). Indirect sources play a more prominent role in carbon emissions than direct sources (Clarke et al., 2017; Geng et al., 2011). High CO concentrations represent a significant risk to human health and can rapidly cause hypoxia in humans, prompting giddiness and even death (Li et al., 2017). Based on the data collected by Xu et al. (2020) from 2017-2019 in the three cities of Wuhan, Enshi, and Jingmen, China (January, February, and March 2020), the average CO concentration was reduced by 28.8%, 20.3%, and 27.9%, respectively. Carbon dioxide occurs naturally in the atmosphere, with its primary anthropic drivers being deforestation and fossil fuel burning, such as coal and oil. Carbon dioxide harms air pollution and the greenhouse effect, trapping heat in the atmosphere. According to Le Quéré et al. (2020), global emissions of CO2 were reduced by 17% during early April 2020, the peak three months of COVID-19 as compared to April 2019 levels, which might have been caused by the reduction in transportation like vehicle and flight shutdowns (Le Quéré et al., 2020). The are many direct and indirect methods, but the Logarithmic Mean Divisia Index (LMDI) is one of the best and most popular methods to measure carbon emissions (Zhang et al., 2020). Thus, carbon emissions can be measured easily during certain times, like in a pandemic.
5.5 Ozone (O3)
Ozone (O3) is an essential helpful greenhouse gas produced from anthropogenic activities (Such as power plants, fossil fuel emissions, and industrial exhaust), which has increased the level of O3 in the atmosphere (Logan et al., 1981; Ryerson et al., 2001). High air humidity, solar radiation, expanded Volatile Organic Compounds (VOCs), and NOx in the environment stimulate the photochemical reaction and increase O3 levels. Based on the data collected by Xu et al. (2020) from the three cities, Wuhan, Jingmen, and Enshi, China, for three months (January, February, and March 2020), the average O3 concentration was elevated by 12.7%, 14.3%, and 11.6% compared to 2017-2019, respectively. According to Sicard et al. (2020), during the lockdown in 2020, the daily O3 level was raised by 36% in Wuhan, 27% in Turin, 14% in Rome, 24% in Nice, and 2.4% in Valencia as compared to 2017-2019 period (Sicard et al., 2020). The lockdown impact on the production of O3 was 38% higher in Wuhan and 10% higher in Southern Europe than the weekend effect.
As indicated in the previous paragraph, O3 levels are firmly linked to VOCs and NO2. When the concentration of NOx is low, NOx supports the O3 concentration, and the formation of VOCs has little impact on O3. NOx concentrations negatively correlate with O3 production when low VOC concentrations (Chameides et al., 1992). In photochemical reactions, NO2 acts as a precursor under adequate solar radiation power and is first dissociated into O3 and NO. Overall, urban O3 and NOx have negative association features, mainly in winter. This is because in summer, because of the extreme sun-based radiation and the predominant photochemical responses, the environment is increasingly suitable for the build-up of O3. The photochemical response during winter is moderately low. Therefore, under higher concentrations of NO2, a distinct range of O3 is vital to building up. During peak days of COVID-19, a lower concentration of NO2 causes more formation of O3, which cannot change into another element (Biswas et al., 2019).
5.6 Water pollution
Water plays a fundamental role in daily life and is considered essential for all life (Postel et al., 1996). Drinking water sometimes contains numerous chemical, biological, or physical impurities, which cause chronic human issues or even death. Water pollution occurs when microorganisms, poisonous synthetic compounds from industries, and local waste interact with water bodies, overflow, or filter into freshwater or groundwater assets. As reported by WHO in 2017, almost two billion people drink contaminated water worldwide. Different types of diseases spread, such as cholera, polio, dysentery, and typhoid, from contaminated water (Chen et al., 2019). The estimated death toll is about 485.000 yearly from diarrheal diseases alone, as reported by the WHO (Liu et al., 2018). Additionally, around 5.000.000 child deaths happen in developing countries due to polluted drinking water supplies (Holgate, 2000). Almost 60% of infant deaths occur due to water-related diarrhea in Pakistan alone, the most elevated proportion in Asia (Yousafzai et al., 2020). The rise of urban areas and industrialization put enormous pressure on water resources, and the release of wastewater into natural water resources decreases the quality of surface and groundwater. Thus, to understand water pollution during the peak days of the lockdown, scientists and researchers need to assess the impact of COVID-19’s short- and long-term impact on the hydrosphere, such as groundwater, oceans, rivers, and lakes.
During the COVID-19 lockdown, most industries were closed, or if open, their manufacturing demand was reduced due to a decrease in demand and the economy. Thus, lockdown may save our surface water resources like lakes and rivers, as contamination of water resources has decreased. During a lockdown, the most affected industries included plastic, crude oil, wastewater disposal, and heavy metals (Häder et al., 2020). The news media in Italy reported that the Grand Canal turned into clear water during the lockdown and aquatic life reappeared and flourished (Yunus et al., 2020). Due to the lockdown, 22 drains that disposed of sewage into the Ganga River were sealed, thereby making the Ganga River cleaner. Specialists and researchers revealed that the water of the Ganga, from Devprayag to Harki Paudi, fell beneath the 'A' category considered by the Central Pollution Control Board (CPCB). This suggests that water is suitable for drinking in 2020 (Mani, 2020). The biochemical oxygen level fell under 3 mg/liter, which indicates good water quality (Gautam et al., 2020). An examination by Yunus et al. (2020) reported that during the lockdown period, the Suspended Particulate Matter (SPM) concentration was reduced by 15.9% on average (range: -10.3% to 36.4%, up to 8 mg/l decreases) in the freshwater Indian Vembanad Lake compared with the pre-lockdown period (Yunus et al., 2020). Compared to previous years, the rate of SPM declined by 34% in April 2020. To stop the spread of coronavirus through the wastewater, China has developed wastewater treatment plants to reinforce its decontamination (fundamentally through expanded chlorine usage). However, there is no proof of the existence of the coronavirus in wastewater or drinking water (La Rosa et al., 2020). Nevertheless, the abundance of chlorine in the water could harm human and animal health (Zambrano-Monserrate et al., 2020).
5.7 Noise pollution
Environmental noise is one of the leading causes of environmental distress, health problems, and changes in the ecosystem's natural conditions (Zambrano-Monserrate and Ruano, 2019). Noise pollution is an unwanted sound due to anthropogenic activities, either commercial or industrial activities. Introduction to submerged noise pollution from transportation is known to cause a variety of harmful effects in invertebrates, including the brittle star (Amphiura filiformis), the decapod (Nephrops norvegicus), clams (Ruditapes philippinarum), fish, and marine mammals (whales). These effects include the disruption of the behavior (Weilgart, 2018; Wisniewska et al., 2018), augmented level of physiological stress (Debusschere et al., 2016; Rolland et al., 2012), and masking of acoustic communication (Putland et al., 2018; Stanley et al., 2017). High noise levels can also cause cardiovascular impacts in people and elevated rates of Coronary Heart Disease (Brunekreef, 2006; Münzel et al., 2018). Due to restrictions, transportation from small to large vehicles was reduced significantly. Meanwhile, the activities at the commercial level were entirely stopped. Most of the road networks were found to be predominantly empty, therefore was no noise of vehicle engines, no honking, no commercial and sporting events, no loud echo of speakers, and no noise from factories (Espejo et al., 2020). Globally, shipping, imports, and exports were significantly reduced. Due to COVID-19, the load of isolating restrictions by many governments made individuals stay at home. Therefore, shipping and public and private transport were reduced significantly. Almost wholly, business activities were also halted (Wang et al., 2020).
6 Human health affected by COVID-19 pandemic
The COVID-19 pandemic severely threatens human health worldwide by causing anxiety, mental stress, depression, and many other negative behaviors (Shigemura et al., 2020). Globally, all countries were affected by it; however, there is minimal information and data on human health except for COVID-19-infected cases. Mental health is one of the major issues that was more pronounced during the COVID-19 pandemic. Human mental performance is strongly influenced due to serious human tragedies and natural disasters (Ćosić et al., 2020). Besides the pandemic, health issues cause enormous economic loss. It has been reported that16.4 trillion dollars will be spent globally on mental health care from 2010 to 2030 (Mari and Oquendo, 2020). It is also reported that many human stresses are highly associated with large-scale disasters like pandemics. The pandemic of COVID-19 has significantly impacted human health and created several health and mental complications, such as uncertainty in body performance and enormous unemployment (Inkster, 2021; Reeves et al., 2014). According to a report, when asked by the health minister of Afghanistan about the emotional condition during the Afghan war in 2001, he said that about 60% of people suffered mental and health issues (Ćosić et al., 2012). The FAO stated that almost 37 million people in Europe generally suffer from anxiety and 44 million from depression (Dubey, 2020). Almost every person is unexpectedly disturbed with negative emotions, emotional traumas, and other kinds of mental and health disordered due to the COVID-19 pandemic. Confirmed COVID-19 cases suggest that almost 200 countries will face severe health challenges (STATISTA). Among the challenges, unemployment and health were in most of the report. According to a report, almost 3.0 million people claimed unemployment in the USA only during March 2020. The published report estimated that GDP declined in developing and developed industrial countries and benchmarked for a world with rates of 2.5%, 1.8%, and 4%, respectively, due to the COVID-19 outbreak (Maliszewska et al., 2020). Besides spreading the virus, treatments and the development of vaccines were addressed widely; however, there is a need for awareness of the long term effect on health and mental issues in global communities.
Due to economic uncertainty and public health, it is well known that mental health is seriously affected due to the threat of COVID-19. A report by the WHO addressed that anxiety and stress are highly associated with this pandemic (WHO, 2020). Reports have also stated that the rate of other activities such as level of loneliness, abuse of alcohol and drugs and depression have increased in the affected individual. Similarly, many other concerns like fear, negative social behavior, health consciousness, and medical checkups increased significantly (Shigemura et al., 2020). It is reported that almost 32% of adults in China worried and were stressed continuously about COVID-19, and almost 14% were severely affected by mental health issues during March 2020 (Hamel et al., 2020). Therefore, taking action and creating awareness among societies is indispensable to reducing the mental damage due to fear of the COVID-19 pandemic. Meanwhile, there is a need to find a way in national and international policies to address the health concern due to pandemics and other disasters. Omicron, Delta, and Gamma variants have been associated with the same symptoms developed by patients in the early COVID-19 variant. However, Omicron is less severe than any other strain as it does not cause severe pneumonia and alveoli destruction. Meanwhile, very few people are associated with losing taste or smell (Monajjemi et al., 2022).
6.1 Medical waste due to COVID-19
This review identifies the relationship between solid wastes generated during the COVID-19 pandemic. Furthermore, these investigations could be used to evaluate the quantity and variation in waste production during the COVID-19 pandemic. This work can also provide the basis of more research on the sustainable management of waste during pandemics. The pandemic of COVID-19 has created a number of environmental challenges such as biomedical waste and municipal waste. The North American Association for Solid Waste described how COVID-19 lockdown has increased the sources and volume of waste to reduce disease exposure. Similarly, in Hubei, China, according to a press release on March 11, 2020, municipal waste was reduced by almost 30% and medical waste increased by 370% during the peak days of COVID-19 (Kleme et al., 2020). It has been stated that generated medical waste accounts for more than 80% of infectious waste disposed of by municipalities (WHO, 2020).
Among the number of unexpected impacts of COVID-19, management of municipal waste has become a potential concern due to its increasing mass (Smart Waste report, European Union, 2020). To protect themselves from COVID-19 infection, individuals are competing to purchase face masks, particularly medical masks (Secon, 2020). It has prompted a clinical mask scarcity for individuals in need and may bring about a scarcity of clinical masks around the globe (Miller, 2020). As reported in previous studies (Johnson and Morawska, 2009), the mouth is the key source of a respiratory virus that spreads into the air by coughing and speech. Additionally, it has described that common cloth (70% cotton and 30% polyester) and towel (100% cotton) masks exhibited 40–60% filtration efficacy for particles (Rengasamy et al., 2010). For the safe disposal of face masks and other clinical waste, individuals need guidelines and protection for safe disposal (Phan and Ching, 2020). The workers responsible for collecting trash often dump the material in the open environment, which creates danger and an unfavorable environment. Employees or individuals who are responsible for the cleanliness of urban communities need to perform their duties regularly, which reduces the spread of the virus by waste. Masks are manufactured with liquid resistant plastic building materials and are long-lived. Consequently, they are castoffs and end up in landfills or the ocean. Aside from clinical masks, hand sanitizer, gloves, and tissue paper have been increasingly used and build up clinical waste in nature. It was reported that a 100 m stretch of beach in the Soko islands, Hong Kong has a massive buildup of masks, as reported by Ocean Asia. Due to the COVID-19 outbreak, 7 million individuals unexpectedly began wearing single-use gloves, one or several masks every day, hand sanitizers, and thus, and the quantity of trash delivered was significant. During the outbreak, an average of 240 metric tons of medical waste was produced by hospitals in Wuhan per day when compared with their past average of fewer than 50 metric tons. There has also been an upsurge in trash from personal protective equipment (PPE) such as gloves and masks in other countries such as the USA (Zambrano-Monserrate et al., 2020). Waste generation during the COVID-19 global pandemic has seen a sharp increase due to huge safety measures and daily consumption (Fig. 3 ). It has reported that plastic use in COVID has increased (Nghiem et al., 2020). The management of clinical waste could be a major problem as coronavirus is rapidly spreading around the world. Waste management companies and medical health organizations have developed different strategies for the purification of coronavirus from waste streams. Recycling of waste has consistently been a key environmental issue and is important for countries (Liu et al., 2020). Recycling is a common and compelling approach to stopping pollution, conserving natural resources and saving energy (Ma et al., 2019). As specialists have been worried about the danger of spreading COVID-19 in recycling hubs, countries like the USA have quit recycling programs in a portion of their urban areas. Management of waste has been restricted in mainly developed countries. Furthermore, the industry has reduced the production of dispensable bags, and on the other hand, single-use plastics can even hold bacteria and viruses (Bir, 2020). A possible waste management strategy for citizens to overcome these issues with environmental concerns depicted in Fig. 4 . Thus, waste management, either municipal or clinical, is a very important part of health services and it needs to be highly considered even before the pandemic. Management of this waste reduces the risk associated with waste handling and transmission among people.Fig. 3 Potential waste generation during the COVID- 19 global pandemic.
Fig 3
Fig. 4 Waste management strategies for a green and clean environment to be adopted by populations in developing country.
Fig 4
6.2 Agricultural and food systems
Food is a basic necessity of health and countries try to ensure the security of food during a time of difficulty (Raheel et al., 2021) as is the case in the pandemic (Mwalupaso et al., 2019; Skaf et al., 2019). Similarly, water security is also a critical issue during times of difficulties and its security is necessary to satisfy the standard of a healthy life (Cai et al., 2020). Influence on food security and disruption to food systems is of immediate concern. Globally, food supply and revenue collection have been disturbed profoundly. There was extensive mass media attention given to the unexpected decrease in food security because of market disruptions, closures or lessened institution capacity ultimately vegetables, milk and fruits wastage, etc. (Torero, 2020). A number of restrictions have firmly been recommended from the 1st day of the pandemic. Most countries banned the travelling, import and export of goods at international and national levels to control the spread of the virus, which abruptly disturbed the socio-economic status of many countries. Meanwhile, it also influenced the agro-food system all over the world. The indirect impact of COVID-19 is well known on the daily life system. The demands for food and its services at commercial and restaurant levels decreased many times due to restrictions in the food processing system, unavailability of labour and storage problems at the farm and industry levels. Quarantine strictly affected the availability of labour for agriculture like sowing or maturity/ harvesting stages of fruits, grains and vegetables. During the peak days of COVID-19, the impact of food issues become more prominent and expanded from agricultural production (Torero, 2020). There are a number of reasons behind the food insecurity during the pandemic and the most notable were the loss of jobs and workers’ income, which reduced the ability to purchase food supplies. Similarly, quarantine bounded to stay at home reduces selling and buying capacities of food items. It has also reduced the gathering activities of people to go out for fast food and hence, the food market is affected. The market of food was severely affected as the central distribution system of many supermarkets was hindered due to an imbalance in supply and demand for food items. The stock of milk, perishable fruit and vegetable was wasted due to transportation issues from the place of production to the local market.
The second most important factor that impacted the agro-food system was the availability of labour (Jiang et al., 2017). Because of isolation measures and workforce loss from COVID-19 deaths and severe illness, labour has been unexpectedly limited in many regions. In global work movements and worker programs, there have been substantial limitations that are dangerous to the production of agriculture in some zones or that have caused bottlenecks. The COVID-19 outbreak is influencing global relations a long way past the agri-food division's work power. All the agricultural practices either livestock/agricultural production or planting/harvesting are laborious. The availability of labour has affected all these services and therefore, the food and agricultural system of most countries has been disrupted. Similarly, the border closure banned the import and export of agricultural commodities across the world, which is another factor that reinforces the impact of COVID-19 on the agro-food system. Thirdly, the resilience of the food system considers the important domain that impacted it during the time of the crisis. Most importantly, connectivity among the agro-food system was disturbed and ultimately the global agro-food market and trade were affected intensively (Laborde, 2020). As the global trade (Import and export) depended on the connectivity of borders, due to the shutdown of ports and commercial flights, the activity of agricultural and food goods was disrupted (Ivanov, 2020; Laborde, 2020). It incorporates the declarations of export constraints across numerous countries that restrict worldwide agri-food exchange and access to the market (Laborde, 2020). Internationally, the agri-food zone is vastly linked. Ports that reduce or shut down movements, for agricultural goods massively decreased freight volume on business flights and other wide international supply chain distractions because of the COVID-19 emergency (Ivanov, 2020) can restrict basic access to agrarian efforts and markets. These undermined the impact of COVID-19 dramatically influencing the global food chain system for an unknown period. This may harmfully influence agrarian output for present and upcoming periods.
6.3 Forest sector
Forest protection is an important strategy for mitigation of climate change (Foley et al., 2011) and is always acknowledged because it sequesters a significant amount of carbon (Elisa et al., 2020; Pan et al., 2011). Furthermore, forests are biologically important and eco-friendly for nature and offer many significant services to maintain the ecosystem (Syed et al., 2022). Like other industries, COVID-19 has destroyed the woodland-related industry as a sharp drop in imports and exports of wood products occurred all over the globe. Worldwide interest in wood and wood items, including wood furniture, tropical timber and graphic paper has collapsed. Meanwhile, woodland-connected industries have not been capable to carry on working at full capacity. On the opposite side, there have been steady or even expanded requests for other woodland-based items, including wooden pallets, wrapping materials and tissue for masks and lavatory paper. At the start of the outbreak, requests for toilet paper increased many folds around the world and in some European countries, it increased by nearly 200% per week. The predictable development of the e-commerce industry is probably getting attention for wrapping materials. Despite the capability of the sector to endorse employment and development, the tenacious work shortages have deteriorated by the outbreak. Worldwide, numerous occupations have been forgotten and a lot more are still in danger, as enterprises have confronted difficulties in holding their labour force around the world and meeting financial obligations, leaving people jobless or furloughed. Steady advancement has been made to date to uplift women by supporting their contribution to legitimate and sustainable fuelwood and charcoal production. However, the COVID-19 outbreak is assessed to get expanding trouble on forest assets through unlawful production of charcoal where living dependent on lawful exercises is forfeited in favour of rapid economic gains (Organisation, 2020). The irresistible increase in the population of humans leads to deforestation for industrial means and land for grazing or agriculture (Afelt et al., 2018). Afforestation is an important element of sustainability and its changes during the days of social transformation like a pandemic. Evidence reported that during COVID-19, the illegal clearing of forests and threats to the ecosystem have increased. Global Land Analysis and Discovery (GLAD) has reported that almost 9,583 km2 of deforestation around the tropical during peak months of lockdown to control the spread of the virus (COVID-19), which was double that of 2019 (Brancalion et al., 2020). Deforestation is also associated with various types of fowls and diseases such as Zoonotic a bat-borne viral outbursts (Olivero et al., 2017; Smith and Olesen, 2010).
Lastly, it has reported that in deforestation regions, outbreaks of viral diseases spread at a higher rate (Shah et al., 2019). It is stated that COVID-19 proposed many negative impacts on food security, increased deforestation and many other zoonotic diseases. In the pandemic scenario, it will be very challenging for governments to save lives and forests in tropical areas and support communities serving on a cash economy (Ferrante and Fearnside, 2020) and such a situation gives birth to a new pandemic (Everard et al., 2020).
6.4 Biodiversity
The pandemic resulted in many extreme challenges that threaten biodiversity. The richness of a different animal and plant species in a habitat are referred to as biodiversity (Verma and Prakash, 2020). Nature always develops and promotes balance among the entire living organism by providing a natural environment. All unexpected environmental threats are associated with humans in some way. COVID-19 is also considered as a result of an imbalance in biodiversity such as bat populations. It is noted this time that there is a relationship exists between pandemic (Verma and Prakash, 2020) and biodiversity. It is observed that a number of diseases are associated with animals and birds. With regards to COVID-19, bats and rats got major attention. Due to fear of COVID-19, humans disturbed the natural environment of biodiversity. It is cited that for healthy growth of humans and the economy, a healthy environment is a critically important (Chakraborty and Maity, 2020).
Due to the pandemic, the number of anthropogenic activities were reduced. When human activity decreases due to quarantine, the surrounding biodiversity increases and the surrounding environment emerges as green and clean (Verma and Prakash, 2020). It was also recorded that the water quality improved and aquatic organisms, especially the fish population, flourished during the days of lockdown (Cooke et al., 2021; Verma and Prakash, 2020). COVID-19 outbreak can cause serious consequences on biodiversity and protection upshots. This virus arose because of wildlife misuse (Zhou et al., 2020) and with environmental degradation, the hazard of new diseases rises (Keesing et al., 2010). Previous events for instance pandemics, financial crises and wars have also activated measurable environmental changes (Pongratz et al., 2011; Sayer et al., 2012). There may be problems but on a recent indication, practical conservation seems to be ongoing in several spots. There have even been episodic information on diminished human stresses on wild species. Drops in visitor numbers triggered by travel limitations and park closures have decreased pressures on sensitive animals and crushing weight on popular trails in protected areas. In protected areas, conservation develops a lot of its community support from the wild nature availability but for sensitive species, decreased human weight in the most well-known parks will be acceptable. We have likewise observed reports of wild species wandering into urban and rural regions, including beaches and parks, where they have not been seen for a long time, as traffic and other human movement decays. In zones where protected areas remain open and travel is still possible, vegetation has often significantly expanded reflecting extensive feeling that movement in a natural location is both a mental and physical remedy to the pressure of the outbreak (Corlett et al., 2020).
In coastal areas, beaches are one of the most significant natural resources (Zambrano-Monserrate and Ruano, 2019). They offer services (sand, land, tourism and recreation) that are important to the coastal communities’ survival and hold essential qualities that must be secure from overutilization (Lucrezi et al., 2016). However, individuals have made numerous beaches contamination problems in the world by non-responsible use (Partelow et al., 2015). Social distancing actions because of the COVID-19 disease results in the tourist's lack that has instigated a prominent alteration in the look of many beaches of the world. Such seashores like those of Salinas (Ecuador), Acapulco (Mexico) or Barcelona (Spain), currently look cleaner and with precious stone clear waters.
6.5 Soil health
This article demonstrates the importance of soil restoration and management to mitigate COVID-19 adverse impacts through improving soil health and improving the production system. Removing the communication gap among policymakers, soil scientists, and people through distance learning may be helpful for the restoration of COVID-19 impacts.
The pandemic of COVID-19 has globally disturbed the food system and increased food insecurity. The lockdown has significantly increased the price of commodities like wheat and rice, which increased by 8 and 25% during March 2019, respectively (Torero, 2020). The pandemic influenced fruit, vegetables, crops, legumes and oil commodities (Lal et al., 2020). Perishable food items were significantly affected due to a shortage of labour. Food security is drastically affected due to the challenges of transportation and the irregular supply of food items to be stored (Deaton and Deaton, 2020). Keeping in view the global agriculture importance, soil plays a very important role in crop production, and serves as a key condition for crop production and resilience during times of crisis. Soil management played a very important role in achieving yield and soil health sustainability during the COVID-19 pandemic.
During the peak of the pandemic, the demand for food reduced and production increased. This imbalance created a surplus, which led to the disposal and wastage of food due to the reduction in demand. It is also recorded that in the USA during March-May 2020, meat consumption declined and the bulk of meat products expired. Ultimately, the market decided to dump the waste into the soil and disturbed the soil health (Lal et al., 2020). Similarly, a reduction in the use of perishable foods like tomatoes and potatoes created waste in an open environment (Lal et al., 2020). During the peak of COVID-19, the amount of milk that was bought dropped significantly, and millions of tons of milk were thrown away daily. Different types of waste produce pollution and toxicity for the soil and disturb soil physiochemical and biological properties and therefore, soil health is significantly affected (FAO, 2015). Similarly, the leaching of heavy metals (Raza et al., 2021) and toxic substances from the disposal of medical waste, detergents, and chemicals used during COVID-19 has increased the toxic impact on soil health.
The functionality and restoration of soil quality could be accelerated by the application of modern techniques that improve the local system of food production and improve the quality of the ecosystem to recover the pandemic (Moran et al., 2020). Soil can be used for the disposal of waste but safe disposal should be kept in mind to maintain soil health. Viruses may be transfer of different processes occurring in soil and therefore, there is a need to understand microorganism movement through the different processes such as pedogenic processes occurring in the soil. The understanding of human and COVID-19 interconnectivity is very important to comprehend but the functionality and health of soil cannot minimized (Heffron et al., 2021).
7 Summary
The COVID-19 pandemic is spreading rapidly and consistently across the globe and there is an immediate misfortune to the worldwide economy. Climate change is one of the most significant and crucial difficulties of the 21st century. Despite many endeavors to re-establish nature during the most recent decades, humans could just push a couple of strides ahead. COVID-19 created both positive and negative backhanded impacts on the environment. The current study shows that population movements may cause the spread of the virus (COVID-19) nationally and internationally. Evaluation of the literature showed that a particular matter (PM) also plays a role in the viral spread. Globally, roads are the primary means of transportation and movement, and in many countries, unsafe roads lead to several injuries and deaths. It has been discovered that the amount and frequency of road traffic varied from country to country during the lockdown. In many countries, the traffic level has reached zero, while road traffic has not changed in others. However, compared with 2018, road accidents, deaths and injuries decreased by half during the lockdown. Similarly, handling municipal and medical waste during the pandemic is very important to assess. There is a need to evaluate management services and personal safety during the lockdown because poor sanitation also responds to COVID-19. There is an utmost need for urgent environmental investigations because the surveillance of the COVID-19 virus is high in wastewater. Thus, there is a need for the treatment of wastewater to reduce the abundance of the virus in water.
On the other hand, there is evidence that the spread of the virus has been reducing air (NOx, PMX and SOX), water and noise pollution and most likely is even saving lives due to a cleaner environment. However, the O3 concentration in the atmosphere has been increasing due to low NOx. The chlorine excess in the water could damage individuals' health. Climate characteristics assumed a vital role in the infection during the initial phase, the temperate regions being more susceptible than the dry or tropical areas. Covid-19′s influence on disrupting the food systems and security is also of immediate concern. The pandemic has the potential to activate significant impacts on biodiversity and preservation. National governments and intergovernmental organizations should implement clear policies to protect biodiversity. Medical waste has not been addressed by pollution control under environmental pollution control authorities. This is an excellent hazard for human beings and the environment, and there is a need to take simultaneous action to mitigate the environment. Medium and longer-term planning require for how the economy is re-stabilized and re-empowered after this emergency. Thus, there is a need for immediate action to evaluate the social activities, environmental impacts and economic status during public health tragedies.
Ethical approval
All the authors have agreed to submit it.
Consent to participate
Before the submission of the paper all the authors have given consent to publish.
Consent to publish
All the authors have given consent to publish.
Availability of data and materials
The information is a compilation from different databases.
Disclosure statement
No potential conflict of interest was reported by the authors.
CRediT authorship contribution statement
Taqi Raza: Conceptualization, Writing – original draft. Muhammad Shehzad: Methodology, Validation. Mazahir Abbas: Methodology, Writing – review & editing. Neal S. Eash: Writing – review & editing. Hanuman Singh Jatav: Data curation, Writing – original draft. Mika Sillanpaa: Writing – review & editing. Trevan Flynn: Data curation, Writing – original draft.
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.
==== Refs
Reference
Afelt A. Frutos R. Devaux C. Bats, coronaviruses, and deforestation: toward the emergence of novel infectious diseases? Front. Microbiol. 9 2018 702 10.3389/fmicb.2018.00702 29696007
Akimoto H. Global air quality and pollution Science 302 2003 1716 1719 10.1126/science.1092666 14657488
Bremer S. Schneider P. Glavovic B. Climate change and amplified representations of natural hazards in institutional cultures Oxford Res. Encycl. Nat. Hazard Sci. 2019 10.1093/acrefore/9780199389407.013.354
Fidan O. Mujwar S. Kciuk M. Discovery of adapalene and dihydrotachysterol as antiviral agents for the Omicron variant of SARS-CoV-2 through computational drug repurposing Mol. Divers. 2022 1 13 https://link.springer.com/article/10.1007/s11030-022-10440-6
Aguiar A. Chepeliev M. Corong E. McDougall R. van der Mensbrugghe D. The GTAP data base: version 10 J. Glob. Econ. Anal. 4 1 2019 1 27 10.21642/JGEA.040101AF
Ali S.A. Baloch M. Ahmed N. Ali A.A. Iqbal A. The outbreak of coronavirus disease 2019 (COVID-19)—an emerging global health threat J. Infect. Public Health 13 4 2020 644 646 10.1016/j.jiph.2020.02.033 32199792
Arden C. Iii P. Burnett T. Thurston D. Michael S. Thun J. Calle E.E. Krewski P.D. Godleski J. Cardiovascular mortality and long-term exposure to particulate air pollution epidemiological evidence of general pathophysiological pathways of disease Circulation 109 1 2004 71 77 10.1161/01.CIR.0000108927.80044.7F 14676145
Barakat T. Muylkens B. Su B.-L. Is particulate matter of air pollution a vector of COVID-19 pandemic? Matter 3 4 2020 977 980 10.1016/j.matt.2020.09.014 33047101
Barro R.J. Convergence and modernisation Econ. J. 125 585 2015 911 942 10.1111/ecoj.12247
Bentolila S. Jansen M. Jiménez G. When credit dries up: job losses in the great recession J. Eur. Econ. Assoc. 16 3 2018 650 695 10.1093/jeea/jvx021
Bezemer D.J. The credit crisis and recession as a paradigm test J. Econ. Issues 45 1 2011 1 18 10.2753/JEI0021-3624450101
Bir, B. 2020. Single-use items not safest option amid COVID-19. Coronavirus leads to rise in all sorts of plastic bags as well as including single-use items, says environmentalists. Retrieved from https://www.aa.com.tr/en/health/single-use-items-not-safest-option-amid-COVID-19/1787067 (Accessed date, 26.).
Biswas M.S. Ghude S.D. Gurnale D. Prabhakaran T. Mahajan A.S. Simultaneous observations of nitrogen dioxide, formaldehyde and ozone in the indo-gangetic plain Aerosol Air Qual. Res. 19 8 2019 1749 1764 10.4209/aaqr.2018.12.0484
Bhosale J. Prices of Agricultural Commodities Drop 20% Post COVID-19 Outbreak 19 2020 The Economic Times
Brancalion P.H. Broadbent E.N. de-Miguel S. Cardil A. Rosa M.R. Almeida C.T. Almeida D.R. Chakravarty S. Zhou M. Gamarra J.G. Emerging threats linking tropical deforestation and the COVID-19 pandemic Perspect. Ecol. Conserv. 18 4 2020 243 246 10.1016/j.pecon.2020.09.006 33020748
Brook R.D. Franklin B. Cascio W. Hong Y. Howard G. Lipsett M. Luepker R. Mittleman M. Samet J. Smith S.C. Jr Air pollution and cardiovascular disease: a statement for healthcare professionals from the expert panel on population and prevention science of the American Heart Association Circulation. 109 21 2004 2655 2671 10.1161/01.CIR.0000128587.30041.C8 15173049
Brunekreef B. Traffic and the heart Eur. Heart J. 27 22 2006 2621 2622 10.1093/eurheartj/ehl319 17035254
Cai J. He Y. Xie R. Liu Y. A footprint-based water security assessment: an analysis of Hunan province in China J. Clean. Prod. 245 2020 118485 10.1016/j.jclepro.2019.118485
Cereda, D., Tirani, M., Rovida, F., Demicheli, V., Ajelli, M., Poletti, P., Merler, S., 2020. The early phase of the COVID-19 outbreak in Lombardy, Italy. arXiv preprint arXiv:2003.09320. 10.48550/arXiv.2003.09320.
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
Chameides W. Fehsenfeld F. Rodgers M. Cardelino C. Martinez J. Parrish D. Lonneman W. Lawson D. Rasmussen R.A. Zimmerman P. Greenberg J. Mlddleton P. Wang T. Ozone precursor relationships in the ambient atmosphere J. Geophys. Res. 97 1992 6037 10.1029/91JD03014
Chen B. Wang M. Duan M. Ma X. Hong J. Xie F. Zhang R. Li X. In search of key: protecting human health and the ecosystem from water pollution in China J. Clean. Prod. 228 2019 101 111 10.1016/j.jclepro.2019.04.228
Chepeliev M. Development of the Non-CO2 GHG Emissions Database for the GTAP Data Base Version 10A (No. 5993) 2020 Center for Global Trade Analysis, Department of Agricultural Economics, Purdue University 10.1007/s10640-013-9664-9
Clarke J. Heinonen J. Ottelin J. Emissions in a decarbonised economy? Global lessons from a carbon footprint analysis of Iceland J. Clean. Prod. 166 2017 1175 1186
Cohen A.J. Brauer M. Burnett R. Anderson H.R. Frostad J. Estep K. Balakrishnan K. Brunekreef B. Dandona L. Dandona R. Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: an analysis of data from the Global Burden of Diseases Study 2015 Lancet N. Am. Ed. 389 10082 2017 1907 1918 10.1016/S0140-6736(17)30505-6
Collivignarelli M.C. Abbà A. Bertanza G. Pedrazzani R. Ricciardi P. Miino M.C. Lockdown for CoViD-2019 in Milan: what are the effects on air quality? Sci. Total Environ. 732 2020 139280 10.1016/j.scitotenv.2020.139280
Comunian S. Dongo D. Milani C. Palestini P. Air pollution and COVID-19: the role of particulate matter in the spread and increase of COVID-19′s morbidity and mortality Int. J. Environ. Res. Public Health 17 12 2020 4487 10.3390/ijerph17124487 32580440
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
Cooke S.J. Twardek W.M. Lynch A.J. Cowx I.G. Olden J.D. Funge-Smith S. Lorenzen K. Arlinghaus R. Chen Y. Weyl O.L. A global perspective on the influence of the COVID-19 pandemic on freshwater fish biodiversity Biol. Conserv. 253 2021 108932 10.1016/j.biocon.2020.108932
Corlett R.T. Primack R.B. Devictor V. Maas B. Goswami V.R. Bates A.E. Koh L.P. Regan T.J. Loyola R. Pakeman R.J. Impacts of the coronavirus pandemic on biodiversity conservation Biol. Conserv. 246 2020 108571 10.1016/j.biocon.2020.108571
Ćosić K. Popović S. Šarlija M. Kesedžić I. Impact of human disasters and COVID-19 pandemic on mental health: potential of digital psychiatry Psychiatr. Danub. 32 1 2020 25 31 10.24869/psyd.2020.25 32303026
Ćosić K. Srbljinović A. Popović S. Wiederhold B.K. Wiederhold M.D. Emotionally based strategic communications and societal stress-related disorders Cyberpsychol. Behav. Soc. Netw. 15 11 2012 597 603 10.1089/cyber.2012.0410 23078339
Deaton B.J. Deaton B.J. Food security and Canada's agricultural system challenged by COVID-19 Can. J. Agric. Econ. 68 2 2020 143 149 10.1111/cjag.12227 /Revue canadienne d'agroeconomie
Debiec-Andrzejewska K. Krucon T. Piatkowska K. Drewniak L. Enhancing the plants growth and arsenic uptake from soil using arsenite-oxidizing bacteria Environ. Pollut. 264 2020 114692 10.1016/j.envpol.2020.114692
Debusschere E. Hostens K. Adriaens D. Ampe B. Botteldooren D. De Boeck G. De Muynck A. Sinha A.K. Vandendriessche S. Van Hoorebeke L. Acoustic stress responses in juvenile sea bass Dicentrarchus labrax induced by offshore pile driving Environ. Pollut. 208 2016 747 757 10.1016/j.envpol.2015.10.055 26561450
Doumbia T. Granier C. Elguindi N. Bouarar I. Darras S. Brasseur G. Gaubert B. Liu Y. Shi X. Stavrakou T. Changes in global air pollutant emissions during the COVID-19 pandemic: a dataset for atmospheric chemistry modeling Earth Syst. Sci. Data Discuss. 2021 1 26 10.5194/essd-13-4191-2021
Drosten C. Günther S. Preiser W. Van Der Werf S. Brodt H.-R. Becker S. Rabenau H. Panning M. Kolesnikova L. Fouchier R.A. Identification of a novel coronavirus in patients with severe acute respiratory syndrome N. Engl. J. Med. 348 20 2003 1967 1976 10.1056/NEJMoa030747 12690091
Dubey, A.K., 2020. Impact of Coronavirus (COVID-19) pandemic on personal relationship of human being.
El-Erian M. The coming coronavirus recession and the uncharted territory beyond Foreign Aff. 2020 10.5194/essd-2020-348 Accessed 27
Elisa P. Alessandro P. Andrea A. Silvia B. Mathis P. Dominik P. Manuela R. Francesca T. Voglar G.E. Tine G. Environmental and climate change impacts of eighteen biomass-based plants in the alpine region: a comparative analysis J. Clean. Prod. 242 2020 118449 10.1016/j.jclepro.2019.118449
Espejo W. Celis J.E. Chiang G. Bahamonde P. Environment and COVID-19: pollutants, impacts, dissemination, management and recommendations for facing future epidemic threats Sci. Total Environ. 747 2020 141314 10.1016/j.scitotenv.2020.141314
Everard M. Johnston P. Santillo D. Staddon C. The role of ecosystems in mitigation and management of COVID-19 and other zoonoses Environ. Sci. Policy 111 2020 7 17 10.1016/j.envsci.2020.05.017 32501392
FAO I. Status of the World's Soil Resources (SWSR)–technical summary 2015 Food and Agriculture Organization of the United Nations and Intergovernmental Technical Panel on Soils Rome, Italy
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
Ferrante L. Fearnside P.M. Protect Indigenous peoples from COVID-19 Science 368 6488 2020 251 10.1126/science.abc0073 251
FITBIT, B., 2020. FITBIT news: the impact of coronavirus on global activity [Internet].
Foley J. Ramankutty N. Brauman K. Cassidy E. Gerber J. Johnston M. Zaks D. Solutions for a cultivated planet Nature 478 2011 7369 10.1038/nature10452
Gautam M. Kumari S. Gautam S. Singh R.K. Kureel R. The novel coronavirus disease-COVID-19: pandemic and its impact on environment Curr. J. Appl. Sci. Technol. 39 17 2020 13 21
Gautam S. COVID-19: air pollution remains low as people stay at home Air Qual. Atmos. Health 13 2020 853 857 10.1007/s11869-020-00842-6 32837609
Geng Y. Tian M. Zhu Q. Zhang J. Peng C. Quantification of provincial-level carbon emissions from energy consumption in China Renew. Sustain. Energy Rev. 15 8 2011 3658 3668
Gössling S. Scott D. Hall C.M. Pandemics, tourism and global change: a rapid assessment of COVID-19 J. Sustain. Tour. 29 1 2020 1 20 10.1080/09669582.2020.1758708
Green M.S. Did the hesitancy in declaring COVID-19 a pandemic reflect a need to redefine the term? Lancet N. Am. Ed. 395 10229 2020 1034 1035 10.1016/S0140-6736(20)30630-9
Guo J. Zhang Y.J. Zhang K.B. The key sectors for energy conservation and carbon emissions reduction in China: evidence from the input-output method J. Clean. Prod. 179 2018 180 190 10.1016/j.jclepro.2018.01.080
Häder D.-P. Banaszak A.T. Villafañe V.E. Narvarte M.A. González R.A. Helbling E.W. Anthropogenic pollution of aquatic ecosystems: emerging problems with global implications Sci. Total Environ. 713 2020 136586 10.1016/S0140-6736(20)30630-9
Hamel L. Lopes L. Muñana C. Kates J. Michaud J. Brodie M. KFF Coronavirus Poll: March 2020 2020 Kaiser Family Foundation 10.1007/s11126-020-09836-0
Monajjemi, M., Kandemirli, F., Mollaamin, F., & Küçük, Ö., 2022. An overview on nature function in relation with spread of Omicron-COVID-19: where will the next pandemic begin and why the amazon forest offers troubling clues. 10.33263/BRIAC133.225.
Hart J.E. Liao X. Hong B. Puett R.C. Yanosky J.D. Suh H. Kioumourtzoglou M.-A. Spiegelman D. Laden F. The association of long-term exposure to PM 2.5 on all-cause mortality in the nurses’ health study and the impact of measurement-error correction Environ. Health 14 1 2015 1 9 10.1186/s12940-015-0027-6 25564290
He G. Zhao Y. Wang L. Jiang S. Zhu Y. China's food security challenge: effects of food habit changes on requirements for arable land and water J. Clean. Prod. 229 2019 739 750 10.1016/j.jclepro.2019.05.053
Heffron R.J. Körner M.-F. Schöpf M. Wagner J. Weibelzahl M. The role of flexibility in the light of the COVID-19 pandemic and beyond: contributing to a sustainable and resilient energy future in Europe Renew. Sustain. Energy Rev. 140 2021 110743 10.1016/j.rser.2021.110743
Holgate G. Water quality: DETR consultation on new regulations for drinking water Environ. Waste Manag. 3 3 2000 105 112 10.1155/2017/7908183
Horowit J. The Global Coronavirus Recession is Beginning 2020 CNN. Media Report
Inkster B. Early warning signs of a mental health tsunami: a coordinated response to gather initial data insights from multiple digital services providers Front. Digit. Health 2 2021 64 10.3389/fdgth.2020.578902
Ivanov D. Predicting the impacts of epidemic outbreaks on global supply chains: a simulation-based analysis of the COVID-19/SARS-CoV2 case Transp. Res. E 2020 10.1016/j.cub.2017.05.064
Jagannathan R. Kapoor M. Schaumburg E. Causes of the great recession of 2007–2009: the financial crisis was the symptom not the disease! J. Financ. Intermediation 22 1 2013 4 29 10.1016/j.jfi.2012.06.002
Jiang G. Ma W. Dingyang Z. Qinglei Z. Ruijuan Z. Agglomeration or dispersion? Industrial land-use pattern and its impacts in rural areas from China's township and village enterprises perspective J. Clean. Prod. 159 2017 207 219 10.1016/j.jclepro.2017.04.152
Johnson G. Morawska L. J. Aerosol Med. Pulm. Drug Deliv. 22 2009 229 237 19415984
Kandari R. Kumar A. COVID-19 pandemic lockdown: effects on the air quality of South Asia Environ. Sustain. 2021 1 7 10.1007/s42398-020-00154-6
Keesing F. Belden L.K. Daszak P. Dobson A. Harvell C.D. Holt R.D. Hudson P. Jolles A. Jones K.E. Mitchell C.E. Impacts of biodiversity on the emergence and transmission of infectious diseases Nature 468 7324 2010 647 652 10.1038/nature09575 21124449
Khan M.F. Hamid A.H. Bari M.A. Tajudin A.B.A. Latif M.T. Nadzir M.S.M. Yusoff M.F. Airborne particles in the city center of Kuala Lumpur: Origin, potential driving factors, and deposition flux in human respiratory airways Sci. Total Environ. 650 2019 1195 1206 10.1016/j.scitotenv.2018.09.072 30308807
Klemeš J.J. Van Fan Y. 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 109883 10.1016/j.rser.2020.109883
Koken P.J. Piver W.T. Ye F. Elixhauser A. Olsen L.M. Portier C.J. Temperature, air pollution, and hospitalization for cardiovascular diseases among elderly people in Denver Environ. Health Perspect. 111 10 2003 1312 1317 10.1289/ehp.5957 12896852
Ksiazek T.G. Erdman D. Goldsmith C.S. Zaki S.R. Peret T. Emery S. Tong S. Urbani C. Comer J.A. Lim W. A novel coronavirus associated with severe acute respiratory syndrome N. Engl. J. Med. 348 20 2003 1953 1966 10.1056/NEJMoa030781 12690092
Kulmala M. Vehkamäki H. Petäjä T. Dal Maso M. Lauri A. Kerminen V.M. McMurry P.H. Formation and growth rates of ultrafine atmospheric particles: a review of observations J. Aerosol Sci. 35 2 2004 143 176 10.1016/j.jaerosci.2003.10.003
La Rosa G. Bonadonna L. Lucentini L. Kenmoe S. Suffredini E. Coronavirus in water environments: occurrence, persistence and concentration methods-A scoping review Water Res. 115899 2020 10.1016/j.watres.2020.115899
Laborde, D., 2020. International food policy research institute food export restrictions during the COVID-19 crisis. 10.1126/science.abc4765.
Lal P. Kumar A. Kumar S. Kumari S. Saikia P. Dayanandan A. Adhikari D. Khan M. 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
Lal R. Brevik E.C. Dawson L. Field D. Glaser B. Hartemink A.E. Hatano R. Lascelles B. Monger C. Scholten T. Managing soils for recovering from the COVID-19 pandemic Soil Syst. 4 3 2020 46 10.3390/soilsystems4030046
Larry Elliot L. Prepare For the Coronavirus Global Recession 2020 The Guardian. Media report
Logan J.A. Prather M.J. Wofsy S.C. McElroy M.B. Tropospheric chemistry: a global perspective J. Geophys. Res. Oceans. 86 C8 1981 7210 7254
Le Quéré C. Jackson R.B. Jones M.W. Smith A.J. Abernethy S. Andrew R.M. De-Gol A.J. Willis D.R. Shan Y. Canadell J.G. Temporary reduction in daily global CO 2 emissions during the COVID-19 forced confinement Nat. Clim. Change 10 7 2020 647 653 10.1038/s41558-020-0797-x
Le Tertre A. Medina S. Samoli E. Forsberg B. Michelozzi P. Boumghar A. Vonk J. Bellini A. Atkinson R. Ayres J. Short-term effects of particulate air pollution on cardiovascular diseases in eight European cities J. Epidemiol. Commun. Health 56 10 2002 773 779 10.1136/jech.56.10.773
Li T. Shen H. Yuan Q. Zhang X. Zhang L. Estimating ground-level PM2. 5 by fusing satellite and station observations: a geo-intelligent deep learning approach Geophys. Res. Lett. 44 23 2017 11 985
Liu M. Tan S. Zhang M. He G. Chen Z. Fu Z. Luan C. Waste paper recycling decision system based on material flow analysis and life cycle assessment: a case study of waste paper recycling from China J. Environ. Manag. 255 2020 109859 10.1016/j.jenvman.2019.109859
Liu W. Yang H. Liu Y. Kummu M. Hoekstra A.Y. Liu J. Schulin R. Water resources conservation and nitrogen pollution reduction under global food trade and agricultural intensification Sci. Total Environ. 633 2018 1591 1601 10.1016/j.scitotenv.2018.03.306 29758909
Lokhandwala S. Gautam P. Indirect impact of COVID-19 on environment: a brief study in Indian context Environ. Res. 188 2020 109807 10.1016/j.envres.2020.109807
Lucrezi S. Saayman M. Van der Merwe P. An assessment tool for sandy beaches: a case study for integrating beach description, human dimension, and economic factors to identify priority management issues Ocean Coast. Manag. 121 2016 1 22 10.1016/j.ocecoaman.2015.12.003
Lyu W. Li Y. Guan D. Zhao H. Zhang Q. Liu Z. Driving forces of Chinese primary air pollution emissions: an index decomposition analysis J. Clean. Prod. 133 2016 136 144 10.1016/j.jclepro.2016.04.093
Ma B. Li X. Jiang Z. Jiang J. Recycle more, waste more? When recycling efforts increase resource consumption J. Clean. Prod. 206 2019 870 877 10.1016/j.jclepro.2018.09.063
Mahato S. Pal S. Ghosh K.G. Effect of lockdown amid COVID-19 pandemic on air quality of the megacity Delhi, India Sci. Total Environ. 730 2020 139086 10.1016/j.scitotenv.2020.139086
Maliszewska, M., Mattoo, A., Van Der Mensbrugghe, D., 2020. The potential impact of COVID-19 on GDP and trade: a preliminary assessment. World Bank Policy Research Working Paper(9211).
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
Mani, K., 2020. The lockdown cleaned the Ganga more than ‘Namami Gange'ever did. Link: https://bit.ly/3mY6IVY.
Mari J.D.J. Oquendo M.A. Mental health consequences of COVID-19: the next global pandemic Trends Psychiatry Psychother. 42 3 2020 219 220 10.1590/2237-6089-2020-0081 32844977
Messner, W., (2020). The institutional and cultural context of cross-national variation in COVID-19 outbreaks. Medrxiv. 10.1101/2020.03.30.20047589.
Moran D. Cossar F. Merkle M. Alexander P. UK food system resilience tested by COVID-19 Nat. Food 1 5 2020 242 10.1038/s43016-020-0082-1 242 32395719
Munawer M.E. Human health and environmental impacts of coal combustion and post-combustion wastes J. Sustain. Min. 17 2 2018 87 96 10.1016/j.jsm.2017.12.007
Münzel T. Schmidt F.P. Steven S. Herzog J. Daiber A. Sørensen M. Environmental noise and the cardiovascular system J. Am. Coll. Cardiol. 71 6 2018 688 697 https://www.jacc.org/doi/abs/10.1016/j.jacc.2017.12.015 29420965
Miller G.A. Buck C.R. Kang C.S. Aviles J.M. Younggren B.N. Osborn S. Keay C.R. COVID-19 in Seattle—early lessons learned J. Am. Coll. Emerg. Phys. Open 1 2 2020 85 91
Mwalupaso G.E. Korotoumou M. Eshetie A.M. Alavo J.-P.E. Tian X. Recuperating dynamism in agriculture through adoption of sustainable agricultural technology-Implications for cleaner production J. Clean. Prod. 232 2019 639 647 10.1016/j.jclepro.2019.05.366
Nghiem L.D. Morgan B. Donner E. Short M.D. The COVID-19 pandemic: considerations for the waste and wastewater services sector Case Stud. Chem. Environ. Eng. 1 2020 100006 10.1016/j.cscee.2020.100006
Nilashi M. Asadi S. Abumalloh R.A. Samad S. Ibrahim O. Intelligent recommender systems in the COVID-19 outbreak: the case of wearable healthcare devices J. Soft Comput. Decis. Support Syst. 7 4 2020 8 12 http://jscdss.com/index.php/files/article/view/233
Ocak S. Turalioglu F.S. Effect of meteorology on the atmospheric concentrations of traffic-related pollutants in Erzurum, Turkey Environ. Appl. Sci. 3 5 2008 325 335 10.18502/japh.v5i1.2856
Olivero J. Fa J.E. Real R. Márquez A.L. Farfán M.A. Vargas J.M. Gaveau D. Salim M.A. Park D. Suter J. Recent loss of closed forests is associated with Ebola virus disease outbreaks Sci. Rep. 7 1 2017 1 9 10.1111/conl.12813 28127051
Organisation I.L. COVID-19 crisis and the informal economy: immediate responses and policy challenges ILO Brief 2020 10.1177/00194662211023847
WHO, 2020. Coronavirus disease (COVID-19): situation report, 166. 10.1016/j.scitotenv.2020.139051.
Pan Y. Birdsey R.A. Fang J. Houghton R. Kauppi P.E. Kurz W.A. Phillips O.L. Shvidenko A. Lewis S.L. Canadell J.G. A large and persistent carbon sink in the world's forests Science 333 6045 2011 988 993 10.1126/science.1201609 21764754
Pansini R. Fornacca D. Early spread of COVID-19 in the air-polluted regions of eight severely affected countries Atmosphere 12 6 2021 795 10.3390/atmos12060795
Partelow S. von Wehrden H. Horn O. Pollution exposure on marine protected areas: a global assessment Mar. Pollut. Bull. 100 1 2015 352 358 10.1016/j.marpolbul.2015.08.026 26330016
Peiris J. Lai S. Poon L. Guan Y. Yam L. Lim W. Nicholls J. Yee W. Yan W. Cheung M. Coronavirus as a possible cause of severe acute respiratory syndrome Lancet N. Am. Ed. 361 9366 2003 1319 1325 10.1016/S0140-6736(03)13077-2
Phan T.L. Ching C.T.-S. A reusable mask for coronavirus disease 2019 (COVID-19) Arch. Med. Res. 51 5 2020 455 457 10.1016/j.arcmed.2020.04.001 32448491
Piazzalunga-Expert A. Evaluation of the potential relationship between particulate matter (PM) pollution and COVID-19 infection spread in Italy Mimeo 2020 10.1016/j.envres.2021.110898
Pongratz J. Caldeira K. Reick C.H. Claussen M. Coupled climate–carbon simulations indicate minor global effects of wars and epidemics on atmospheric CO2 between AD 800 and 1850 Holocene 21 5 2011 843 851 10.1177/0959683610386981
Postel S.L. Daily G.C. Ehrlich P.R. Human appropriation of renewable fresh water Science 271 5250 1996 785 788 10.1126/science.271.5250.785
Putland R.L. Merchant N.D. Farcas A. Radford C.A. Vessel noise cuts down communication space for vocalizing fish and marine mammals Glob. Change Biol. 24 4 2018 1708 1721 10.1111/gcb.13996
Raheel A. Ali S.Z. Waris M. Basharat M. Basheer A. Ullah M.A. Hyder S.I. Raza T. Post-harvest methodology and technology for horticultural products in agricultural commercial areas of Pakistan Pak. J. Agric. Res. 34 4 2021 700 10.17582/journal.pjar/2021/34.4.700.705
Raza T. Shehzad M. Qadir M.F. Kareem H.A. Eash N.S. Sillanpaa M. Hakeem K. Indirect effects of COVID-19 on water quality Water-Energy Nexus 2022 10.1016/j.wen.2022.10.001
Raza T. Qureshi K.N. Imran S. Eash N.S. Bortone I. Associated health risks from heavy metal-laden effluent into point drainage channels in Faisalabad Pakistan 34 3 2021 487 494 10.17582/journal.pjar/2021/34.3.487.494
Reale R. Magro T.C. Ribas L.C. Biodiversity conservation actions as a tool to improve the management of sustainable corporations and their needs ecosystem services J. Clean. Prod. 219 2019 1 10 10.1016/j.jclepro.2019.02.039
Reeves A. McKee M. Stuckler D. Economic suicides in the great recession in Europe and North America Br. J. Psychiatry 205 3 2014 246 247 10.1192/bjp.bp.114.144766 24925987
Rengasamy S. Eimer B. Shaffer R.E. Simple respiratory protection—evaluation of the filtration performance of cloth masks and common fabric materials against 20–1000 nm size particles Ann. Occup. Hyg. 54 7 2010 789 798 10.1093/annhyg/meq044 20584862
Rolland R.M. Parks S.E. Hunt K.E. Castellote M. Corkeron P.J. Nowacek D.P. Wasser S.K. Kraus S.D. Evidence that ship noise increases stress in right whales Proc. R. Soc. B Biol. Sci. 279 1737 2012 2363 2368 10.1098/rspb.2011.2429
Ryerson T.B. Trainer M. Holloway J.S. Parrish D.D. Huey L.G. Sueper D.T. Kuster W.C. Observations of ozone formation in power plant plumes and implications for ozone control strategies Science 292 5517 2001 719 723 11326097
Sayer J. Endamana D. Ruiz-Perez M. Boedhihartono A.K. Nzooh Z. Eyebe A. Awono A. Usongo L. Global financial crisis impacts forest conservation in Cameroon Int. For. Rev. 14 1 2012 90 98 https://link.springer.com/article/10.1007/s10640-020-00480-7
Secon H. People are racing to buy face masks amid the coronavirus outbreak, but they probably won't protect you from illness Bus. Insider 2020 Available online https://www.businessinsider.com/wuhancoronavirus-face-masks-not-entirely-effective$-$2020$-$1 (accessed on 27 August 2020)
Shah S. Gwee S.X.W. Ng J.Q.X. Lau N. Koh J. Pang J. Wastewater surveillance to infer COVID-19 transmission: a systematic review Sci. Total Environ. 804 2022 150060 10.1016/j.scitotenv.2021.150060
Shah H.A. Huxley P. Elmes J. Murray K.A. Agricultural land-uses consistently exacerbate infectious disease risks in Southeast Asia Nat. Commun. 10 1 2019 1 13 10.1038/s41467-019-12333-z 30602773
Shigemura J. Ursano R.J. Morganstein J.C. Kurosawa M. Benedek D.M. Public responses to the novel 2019 coronavirus (2019-nCoV) in Japan: Mental health consequences and target populations Psychiatry Clin. Neurosci. 74 4 2020 281 10.1111/pcn.12988 32034840
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 139542 10.1016/j.scitotenv.2020.139542
Skaf L. Buonocore E. Dumontet S. Capone R. Franzese P.P. Food security and sustainable agriculture in Lebanon: an environmental accounting framework J. Clean. Prod. 209 2019 1025 1032 10.1016/j.jclepro.2018.10.301
Smith P. Olesen J.E. Synergies between the mitigation of, and adaptation to, climate change in agriculture J. Agric. Sci. 148 5 2010 543 552 10.1017/S0021859610000341
Stanley J.A. Van Parijs S.M. Hatch L.T. Underwater sound from vessel traffic reduces the effective communication range in Atlantic cod and haddock Sci. Rep. 7 1 2017 1 12 10.1038/s41598-017-14743-9 28127051
Syed A. Raza T. Bhatti T.T. Eash N.S. Climate impacts on the agricultural sector of Pakistan: risks and solutions Environ. Chall. 6 2022 100433
Thorpe A. Harrison R.M. Sources and properties of non-exhaust particulate matter from road traffic: a review Sci. Total Environ. 400 1-3 2008 270 282 10.1016/j.scitotenv.2008.06.007 18635248
Torero M. Without food, there can be no exit from the pandemic Nature 580 2020 588 589 10.1038/d41586-020-01181-3 32327745
Tyrrell D.A.J. Bynoe M.L. Cultivation of a novel type of common-cold virus in organ cultures Br. Med. J. 1 5448 1965 1467 10.1136/bmj.1.5448.1467 (Clin. Res. Ed) 14288084
Verburg R.W. Osseweijer F. A framework to estimate biodiversity loss and associated costs due to nitrogen emissions from single power plants J. Clean. Prod. 239 2019 117953 10.1016/j.jclepro.2019.117953
Verma A. Prakash S. Impact of covid-19 on environment and society J. Glob. Biosci. 9 5 2020 7352 7363 10.1016/j.jclepro.2020.122834
Wang C. Horby P.W. Hayden F.G. Gao G.F. A novel coronavirus outbreak of global health concern Lancet N. Am. Ed. 395 10223 2020 470 473 10.1016/S0140-6736(20)30185-9
Wang J. Qiu Y. He S. Liu N. Xiao C. Liu L. Investigating the driving forces of NOx generation from energy consumption in China J. Clean. Prod. 184 2018 836 846 10.1016/j.jclepro.2018.02.305
Weilgart, L., 2018. The impact of ocean noise pollution on fish and invertebrates. Report for OceanCare, Switzerland. https://www.oceancare.org/wp,content/uploads/2017/10/OceanNoise_FishInvertebrates_May2018.pdf.
Wisniewska D.M. Johnson M. Teilmann J. Siebert U. Galatius A. Dietz R. Madsen P.T. High rates of vessel noise disrupt foraging in wild harbour porpoises (Phocoena phocoena) Proc. R. Soc. B Biol. Sci. 285 1872 2018 20172314 10.1098/rspb.2017.2314
Xu K. Cui K. Young L.H. Wang Y.F. Hsieh Y.K. Wan S. Zhang J. Air quality index, indicatory air pollutants and impact of COVID-19 event on the air quality near central China Aerosol Air Qual. Res. 20 2020 10.4209/aaqr.2020.04.0139
Yang J. Liu H. Sun J. Evaluation and application of an online coupled modeling system to assess the interaction between urban vegetation and air quality Aerosol Air Qual. Res. 18 3 2018 693 710 10.4209/aaqr.2017.06.0199 + ap691-692
Yousafzai M.T. Nawaz M. Xin C. Tsai S.-B. Lee C.-H. Sustainability of waste picker sustainopreneurs in Pakistan's informal solid waste management system for cleaner production J. Clean. Prod. 267 2020 121913 10.1016/j.jclepro.2020.121913
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. Does environmental noise affect housing rental prices in developing countries? Evidence from Ecuador Land Use Policy 87 2019 104059 10.1016/j.landusepol.2019.104059
Zambrano-Monserrate M.A. Ruano M.A. Sanchez-Alcalde L. Indirect effects of COVID-19 on the environment Sci. Total Environ. 728 2020 138813 10.1016/j.scitotenv.2020.138813
Zhang C. Wu Z. Li J.-W. Zhao H. Wang G.-Q. Cytokine release syndrome in severe COVID-19: interleukin-6 receptor antagonist tocilizumab may be the key to reduce mortality Int. J. Antimicrob. Agents 55 5 2020 105954 10.1016/j.ijantimicag.2020.105954
Zhang Q. Jiang X. Tong D. Davis S.J. Zhao H. Geng G. Feng T. Zheng B. Lu Z. Streets D.G. Transboundary health impacts of transported global air pollution and international trade Nature 543 7647 2017 705 709 10.1038/nature21712 28358094
Zhang Y. Shuai C. Bian J. Chen X. Wu Y. Shen L. Socioeconomic factors of PM2. 5 sconcentrations in 152 Chinese cities: decomposition analysis using LMDI J. Ceaner Prod. 218 2019 96 107 10.1016/j.jclepro.2019.01.322
Zhou F. Yu T. Du R. Fan G. Liu Y. Liu Z. Xiang J. Wang Y. Song B. Gu X. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study Lancet N. Am. Ed. 395 10229 2020 1054 1062 10.1016/S0140-6736(20)30566-3
Zhou, P., Yang, X.-L., Wang, X.-G., Hu, B., Zhang, L., Zhang, W., Si, H.-R., Zhu, Y., Li, B., Huang, C.-L., 2020. Discovery of a novel coronavirus associated with the recent pneumonia outbreak in humans and its potential bat origin. bioRxiv.10.1101/2020.01.22.914952.
Zhu Y. Xie J. Huang F. Cao L. The mediating effect of air quality on the association between human mobility and COVID-19 infection in China Environ. Res. 189 2020 109911 10.1016/j.envres.2020.109911
Zoran M.A. Savastru R.S. Savastru D.M. Penache M.C.V. Temporal trends of carbon monoxide (CO) and radon (222 Rn) tracers of urban air pollution J. Radioanal. Nucl. Chem. 320 1 2019 55 70 10.1007/s10967-019-06443-7
| 0 | PMC9741497 | NO-CC CODE | 2022-12-15 23:17:48 | no | Environ Adv. 2023 Apr 10; 11:100328 | utf-8 | Environ Adv | 2,022 | 10.1016/j.envadv.2022.100328 | oa_other |
==== Front
Spat Spatiotemporal Epidemiol
Spat Spatiotemporal Epidemiol
Spatial and Spatio-Temporal Epidemiology
1877-5845
1877-5853
The Authors. Published by Elsevier Ltd.
S1877-5845(22)00083-1
10.1016/j.sste.2022.100560
100560
Article
Fine-scale variation in the effect of national border on COVID-19 spread: A case study of the Saxon-Czech border region
Mertel Adam 1⁎
Vyskočil Jiří 1
Schüler Lennart 12
Schlechte-Wełnicz Weronika 1
Calabrese Justin M. 1345
1 Center for Advanced Systems Understanding (CASUS), Untermarkt 20, 02826, Görlitz, Germany
2 Department of Computational Hydrosystems, Helmholtz Centre for Environmental Research (UFZ), Permoserstr. 15, 04318 Leipzig, Germany
3 Helmholtz-Zentrum Dresden Rossendorf (HZDR), Bautzner Landstraße 400, 01328, Dresden, Germany
5 Department of Biology, University of Maryland, College Park, MD 20742, USA
4 Department of Ecological Modelling, Helmholtz Centre for Environmental Research (UFZ), Permoserstr. 15, 04318 Leipzig, Germany
⁎ Corresponding author: Dr. Adam Mertel. Center for Advanced Systems Understandings, Germany
11 12 2022
11 12 2022
10056025 3 2022
14 9 2022
7 12 2022
© 2022 The Authors. 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.
The global extent and temporally asynchronous pattern of COVID-19 spread have repeatedly highlighted the role of international borders in the fight against the pandemic. Additionally, the deluge of high resolution, spatially referenced epidemiological data generated by the pandemic provides new opportunities to study disease transmission at heretofore inaccessible scales. Existing studies of cross-border infection fluxes, for both COVID-19 and other diseases, have largely focused on characterizing overall border effects. Here, we couple fine-scale incidence data with localized regression models to quantify spatial variation in the inhibitory effect of an international border. We take as a case study the border region between the German state of Saxony and the neighboring regions in northwestern Czechia, where municipality-level COVID-19 incidence data are available on both sides of the border. Consistent with past studies, we find an overall inhibitory effect of the border, but with a clear asymmetry, where the inhibitory effect is stronger from Saxony to Czechia than vice versa. Furthermore, we identify marked spatial variation along the border in the degree to which disease spread was inhibited. In particular, the area around Löbau in Saxony appears to have been a hotspot for cross-border disease transmission. The ability to identify infection flux hotspots along international borders may help to tailor monitoring programs and response measures to more effectively limit disease spread.
Keywords
COVID-19
spatial epidemiological modeling
regression model
border effect
spatio-temporal data analysis
==== Body
pmc1 Introduction
The COVID-19 pandemic is an event of unprecedented magnitude in modern world history and has consequently set off a flurry of research activity across the health-related sciences. These efforts have generated an equally unprecedented deluge of spatially referenced epidemiological data. High resolution epidemiological data have the potential to reveal disease transmission processes on heretofore inaccessible spatial scales, but significant challenges in data harmonization and modeling remain. The global extent and temporally asynchronous nature of the COVID-19 spread have repeatedly put the spotlight on international borders and the policies governing them. Research focusing on cross-border infection fluxes can thus deepen our understanding of COVID-19 spread dynamics while also serving to inform border-related policy decisions.
Several papers have studied COVID-19 border dynamics for different regions with a range of methodologies. For example, Grimmé et al. (2021) used the Endemic-Epidemic framework (Bekker-Nielsen Dunbar and Held, 2020) to model the effect of policies on Swiss-Italian borders. (Eckardt et al., 2020)) applied a Bayesian spatio-temporal Poisson model to study the role of border controls in the Schengen Area, and (Hossain et al., 2020) used a metapopulation model for estimating the effect of travel restrictions on the COVID-19 outbreak. Additionally, Laroze et al. (2021) constructed a multi-source spatial model on top of the pre-crisis commute to work to study the effects of borders between regions within the same country, and (Han et al., 2021) explored the balance between domestic and imported cases in Chinese cities. While all of these studies suggested a significant effect of border presence and border control regime on the spread of COVID-19, all focused on determining the overall effect of the focal borders.
Most of the border-focused studies mentioned in the previous paragraph are based on the spatial level of whole countries, regions, or spatially separated metropolitan areas. More generally, most spatio-temporal COVID-19 models are applied either on the level of whole countries or on medium-sized provinces and regions due to data limitations. In contrast, we developed a model of COVID-19 spread on the very detailed level of individual municipalities. While several studies consider this level of granularity for data analysis and modeling (e.g., (Arauzo-Carod, 2021, Cole et al., 2020, Neyens et al., 2020, Schuler et al., 2021)), research combining datasets on this scale from more than one country is still exceedingly rare even it may provide an unique value for studying the spatiotemporal trends in detail. Furthermore, while governing policy may be uniform across an entire border, variation in geomorphology, population distribution, transportation infrastructure, socio-economical aspects, history and other factors along a border may generate spatial variation in the inhibitory effect of the border on disease spread. To our knowledge, this type of fine-scale variation has not been studied in the context of COVID-19 spread. We therefore aim to quantify cross-border infection fluxes at the scale of individual municipalities in the Saxon-Czech border region.
Our approach is based on the simple assumption that the virus travels mainly with humans. Therefore, COVID-19 may be transmitted more easily between two localities with a higher intensity of mobility connecting them than between localities with lower mobility. Several approaches are widely used in epidemiology to estimate mobility. Probably the most common is the gravity model, which has many possible adaptations and extensions (Yan and Zhou, 2019) and has been a mainstay of spatial COVID-19 modeling (e.g., (Chen et al., 2021, Ezzat et al., 2021, Werner PA 2021, Zhu et al., 2021)).
The basic gravity model estimates the mobility between two places primarily from two variables - their separation distance and the population size of both places. The gravity model can be further extended to include border effects. This idea was applied in the context of various topics, e.g., international to domestic trade (McCallum, 1995), the cross-border flows of immigration (Lewer and Van den Berg, 2008), air traffic (Becker et al., 2018), or mergers and acquisitions (Wong, 2008). Even more relevant to this study is the paper of (Kramer et al., 2016), in which the authors extended the gravity model with a border effect to study the international transmission of Ebola virus. This study concluded that the presence of the national border and its closure may have played a significant role in the transmission of the virus.
Here, we propose a set of local multiple regressions–inspired by the principles of the gravity model–that allow us to quantify fine-scale spatial variation in the effect of the border. Consistent with other studies, we find that the Saxon-Czech border has an overall inhibitory effect on COVID-19 spread. However, we also identify both between-country asymmetry and substantial fine-scale variation in the strength of the border effect. In particular, the border in the region around Löbau in Saxony appears to have had a weaker inhibitory effect than other areas along the Saxon-Czech border. We conclude by discussing how high resolution border studies can potentially pinpoint hotspots for disease spread where additional border monitoring and controls may be warranted.
1.1 Data
This section describes the data sources used in this study and the first preprocessing to clean, integrate, and precalculate the Czech and German data. For Germany, we work with the dataset "Gemeindegrenzen 2018 mit Einwohnerzahl" (GeoBasis-DE et al., 2020) for the geometries and population sizes on the municipality (“Gemeinde”) level. Czech population numbers on the municipality level ("obec") were taken from the Czech Statistical Agency (Czech Statistical Office 2021), while the geometries were obtained from RÚIAN (Czech Office for Surveying, Mapping and Cadastre 2021). We further joined the population size numbers with their associated geometries and merged the datasets from both countries. To keep the same geometry detail on both sides of the borders, we applied the Douglas-Peucker (Douglas and Peucker, 1973) simplification algorithm implemented in the Python library TopoJSON (mattijn 2021).
Due to the geographical focus of this study, the number of infections in single municipalities was obtained only for Saxony on the German side. These data were published on the official website of Saxonian goverment coronavirus.sachsen.de (Sächsische Staatsregierung 2021).These numbers are taken from the reports of doctors and laboratories that are responsible for PCR tests, the assignment to the municipalities is done based on the information from patients. Unfortunately, the website does not provide any historical data directly in a tabular format or via an API service, just a table with data from the last seven days. We therefore set up a CRON job to visit the website every day and scrape the contents in the form of a .csv table. On the Czech side, we worked with the daily updated dataset from the Czech Ministry of Health (Komenda et al., 2021), which provide municipality-level values localised based on the usual presence of the patients. We also reclassified both datasets to weekly granularity to avoid problems with weekly cyclicity (i.e., pronounced decreases in cases reported on weekends, national holidays, etc.). For this study, on the Czech side, we used only data for the three regions in Czechia that share a border with Saxony - Liberec, Ústí nad Labem, and Karlovy Vary. The entire region defined for the case study is shown in Figure 1 . The preprocessed and cleaned dataset is available on the Zenodo server ((Mertel, 2022): 2).Figure 1 Overview map shows the location of the case study area and the extent of two selected bordering regions - the federal state Saxony in Germany, and the regions of Liberec, Ústí nad Labem and Karlovy Vary in Czechia.
Figure 1
After data cleaning and harmonization, we obtained a dataset of 1116 municipalities (almost 6 million inhabitants), with 415 located in Saxony (4,077,937 inhabitants), and 701 in Czechia (1,810,283 inhabitants). The municipalities in Czechia are smaller on average, which reflects a slightly different administrative organization. The timeframe of the case study ranges from the 31st of January to the 13th of June 2021. This timeframe was selected due to the availability of the data on both sides of the border, but also to capture the wave of high case numbers in late winter and spring 2021. In Figure 2 , we can see the weekly case numbers in the study area in each municipality. This series of maps shows the municipalities with increased cases each week of the time series. Here we can follow the general spatio-temporal trends of the spread. At the very beginning of the timeframe, the situation was considerably worse in the western part of Czechia, particularly around the towns of Karlovy Vary and Cheb. During February, the wave of high numbers came to all Czech regions, while from March, the places with increases in case numbers are more dispersed and mostly located within the regions around Děčín, Ústí nad Labem, and Chomutov. On the German side, the first more significant and spatially concentrated wave of increased cases appeared in the middle of March in the region of western Saxony, around Plauen and Zwickau. The following week, we can see an increase also in north-western Saxony (Leipzig region), followed by the central (Dresden, Chemnitz) and then eastern (Görlitz, Bautzen, Zittau)regions of Saxony. Figure 3 shows these values grouped based on the administrative regionalization on a timeline. This graph show a substantial difference between the trends of the German and Czech regions. While the numbers in the Czech regions tend to rise at the beginning of the timeframe (February), German regions have a wave of high numbers in the middle of the timeframe (late March, April). Even though the trends within each country are similar, considerable differences also exist among neighboring regions. Although the Czech regions are considerably smaller in population, the number of cases is similar, implying a higher incidence on the Czech side.Figure 2 Weekly increases in the number of cases on the level of municipalities.
Figure 2
Figure 3 Temporal patterns of the weekly cases in every region (14 "okres" in CZ and 12 "Kreise" in DE).
Figure 3
Most of the above-cited spatial epidemiology papers use the Euclidian distance between location centroids as the measure of distance. However, this method is problematic on the granular spatial scale of our study. Instead, the connectivity between municipalities is mostly defined by the infrastructure that is determined by social (e.g., culture, work/school commuting, local subsidies) and physical (e.g., mountainous areas, rivers) factors. To account for variation in these structural factors, we estimate “temporal distances” between pairs of municipalities via the typical car-based driving time between them.
To calculate the temporal distances between municipalities, we firstly needed to extract the point representing the real central place of each municipality. This was done by the search engine Nominativ (osm-search 2021), which offers a free API service to geocode places by their names. For every request, it returns the point coordinates that represent the geographical unit. For municipalities, the API returns mostly the point of the central square or the centroid of the built-up areas. This point was then used as a starting point for the calculation of isochrones. For this step, we used the Openroute API (GIScience Research Group and HeiGIT 2021) that returned the areas accessible by car within a specific time limit. We set the upper value to 1 hour and worked with 10 six minute intervals. Then, the temporal distance was calculated by iterating through all pairs of municipalities and checking into which isochrone interval of the first municipality the centroid of the second municipality falls. This measure of distance between two municipalities accounts for the transportation networks connecting them and is thus more informative for our purposes than the simple Euclidean distance.
We performed several exploratory analyses before implementing our regression models. Here, we quantified the overall patterns in the temporal trends, the spatial co-occurrence of the case number increases, and the effect of the spatial and temporal autocorrelation. We conclude that the pandemic situation in each country developed somewhat independently, with indications of a significant but heterogeneous spatial effect of the national border. Case numbers in municipalities within a short driving distance (10 minutes and less) are much more strongly correlated than those separated by longer distances. Additionally, municipalities' case numbers are, in general, more correlated at shorter time lags, with the lowest variance in correlations occurring at time lags of 1 to 2 weeks.
1.2 Model
To quantify the local effect of the national border on the spread of COVID-19, we constructed a local model, which was further applied to each municipality in the vicinity of the national border. Inspired by the gravity model, each local model quantifies the contributions of neighboring municipalities to an index of “virus import potential” for the focal municipality as a function of population sizes, inter-municipality temporal distances, and presence/absence of the border between each pair of municipalities.
1.2.1 Time gap
In the whole model, we work with the assumption that the high number of cases in municipality b in time t−x may predict the number of cases in the neighboring municipality a in time t. The x is then the time lag that defines the time that the virus needs to being imported between municipalities. We chose the x to be one week based on the common understanding of the virus spread and the time-lagged correlation analysis, which can be seen in Figure 4 . While the median correlation values change little across various time lags, we note a difference in the distribution of values within the higher part of the scale (75th percentile). This refers to the situations when not all municipalities have similar trends in case numbers, but the correlation is present much more often when the temporal lag of zero or one week and the municipalities lie within one country. Maximum values of the correlation for each municipality show a significant decrease in the time lag of one week, while all the other time lags retain similar distributions. More interesting is the comparison of the variance in the neighborhood municipality-municipality correlations. In this case, the time lags of one and two weeks have the lowest numbers; slightly higher numbers are reached when we consider the municipalities on the other side of the border. This analysis may indicate overall spatio-temporal trends of the virus spread. While the median values do not change significantly over time, the distributions of the medians and the values of the maximum correlations reveal that the similarity within the neighborhood is higher when there is no time lag. In contrast, the variance analysis revealed the highest consistency in correlation values when the time lag is one or two weeks. Additionally, correlations are generally higher when we consider only municipalities from one country, which again highlights the effect of the border.Figure 4 Correlation values of the municipality-municipality temporal trends considering the neighborhood defined by one hour of driving distance. The dark blue color stands for all combinations of municipalities in one country, the light blue color for all municipality-municipality combinations.
Figure 4
1.2.2 Neighborhood
The computation of the model starts by iterating through the list of all municipalities, and for each focal municipality a, we define a neighborhood of municipalities {b, c, ..., f} , which are reachable from a in less than forty minutes of driving time in a car. Further, this neighborhood of municipality a is noted as Θa. This value is supported by additional travel data for this geographical region (Bundesagentur für Arbeit 2020, Centrum dopravního výzkumu, v. v. i 2019, infas Institut für angewandte and Sozialwissenschaft GmbH 2017). These sources show that the largest distance people usually travel to work, school, or shopping is not more than 20 kilometers / 30 minutes. But there are types of travel that often occur over longer distances and times, such as tourism, family visits, and business-related journeys that may take even more than one hour / 100 kilometers. We therefore chose forty minutes as a conservative value that would comfortably accommodate most normal travel. An analysis of the effect of different travel distances to the correlation of new case values is displayed in Figure 5 . Here we notice a strong similarity of case numbers across municipalities within the same country. Such correlations tend to be significant when the municipalities are within 20 minutes of each other, implying strong spatial autocorrelation. A similar relationship of autocorrelation in case numbers at short temporal distances is not present for pairs of municipalities separated by the national border. This finding suggests a strong effect of the national border on the spread of the virus.Figure 5 Spatial autocorrelation graph. The dependence of the municipality-municipality correlations in the case numbers on the temporal distance. The chart highlights the similarity between two municipalities (quantified by the correlation number) on their temporal distance.
Figure 5
1.2.3 Virus import potential
The "virus import potential" is a number that represents the potential that a virus is imported from a neighborhood municipalities Θa. While the spatial covariances may differ in various periods of the pandemic (Schuler et al., 2021), we focused on the weeks when the number of cases in the municipality was increasing because. For each of the increasing weeks t, we looked at the situation in the neighborhood to identify the places, where the increase may have been spread from. We looked at the number of cases within the municipalities in the Θa at t-1, one week before the increase in a. Then, we relativized the numbers for Θa to the sum of all cases in the neighborhood, so we ended with virus import potential values scaled to [0, 1]. Formally noted, the import potential from b to a, in time t, Iab(t), is:Iab(t)=Nb(t−1)∑i∈ΘaNi(t−1),
where Nb(t) is the number of cases in b in time t.
We calculate the import potential from a to b over the whole study period, Iab, as the average of all weeks, when the number of cases in a increased. While not every lagged co-variate in cases increase may indicate a high potential of import, we use weights to incorporate principles of virus transmission dynamics. This can be written as:Iab=∑t=1npwa(t)*Iab(t)∑t=1npwa(t)
Where np is the number of weeks, when the case numbers in a are increasing. The weight of import potential for municipality a in time t, wa(t) has three components:wa(t)=wΘa(t)*wRa(t)*wΔa(t)
(i) The first weight, neighborhood effect in municipality a in week t, wΘa(t), compares the epidemiological situation in the municipality a to the situation in Θa . We assume that a high relative number of cases in the Θa compared to the a may indicate a higher potential of the import from these municipalities. The wN a(t) is calculated as:wΘa(t)=1−(3σa+Za(t)/6σa)−(3σn+Zn(t)/6σn),
where Za(t) is the z-score of the case number in a in week t. n indicates the sum of cases in Θa , Zn(t) is then the z-score of this value in week t, and σa/σn is the standard deviation of the case values in a, respectively Θa. The resulting values will be in between 0 and 2. Value 2 indicates that the import potential in week t is high, value of 0 represents the situation that a is, compared to Θa, completely saturated and there is a low possibility of import.
(ii) wRa(t) is the second weight used for the import potential calculation and represents the impact of instantaneous reproduction number Ra(t) of the municipality a in week t. Including this weight, we incorporate the factor of local transmission dynamics in a itself. We assume that if Ra(t) is high, then the dynamics in municipality a may explain the increase more probably that the import from Θa . Ra(t) was estimated using the epyestim Python package (Cori et al., 2013). For the calculation, we interpolated daily values of the municipalities case numbers, and used the window of 14 days to ensure a smooth result in the smaller municipalities. wRa(t) can be then easily derived from Ra(t) as:wRa(t)=1/Ra(t)
(iii) The third weight is the coefficient of the increase magnitude wΔa(t). For any week t, this value is calculated as the increase a in week t divided by a mean case increase in municipality a for the study period, 〈Δa(t)〉.wΔa(t)=Δa(t)/〈Δa(t)〉
1.2.4 Regression model
For each municipality a, we constructed a local beta regression model that is based on the characteristics of the relations of a to the municipalities in the neighborhood Θa. As a dependent value for the model, we calculated the possibility of importing the virus from municipalities in Θa to a so that they could be used as the dependent variable in a beta (multiple) regression model. The linear predictor of the mean of our beta regression model, which is linked to the (0, 1) index of virus import via the logit link function, can be written asIab=A+wCa*Cab+wPa*Pab+wBa*Bab,
where Iab stands for the virus import potential from municipality b to municipality a, and A is an intercept. Ca, Ba, and Pa are the three predictors of the model, and wCa, wBa and wPa represent their regression coefficients for the municipality a, which we try to estimate with this model. In this study, we assume that both the predictors Ca,Ba,Pa, and the regression coefficients wC, wB, and wP of a single municipality are constant over the time span.
C stands for closeness, which was calculated as the inverse value of the relative driving distance. The higher the closeness value, the faster it was possible to reach the focal municipality from the second municipality by car. The Cab can be calculated as:Cab=Dab/DΘ,
where Dab is the temporal distance between a, and b. DΘ is the temporal distance that we assigned for defining the neighborhood, which is for 40 minutes in this study.
P is the factor of the municipality's population size, that compares the population in b to the sum of populations in Θa. The calculation of this predictor can be written as:Pab=Pb/∑i∈ΘaPi
B is the third predictor variable that represents the presence of the border. Here, the Bab value is 0 if both a, and b are in the same country, value of 1 then a For example, when a was in Saxony, we assigned 0 to all the Saxonian municipalities and 1 to all Czech municipalities, and vice versa when a was in Czechia.
For each municipality a, we calculated a model consisting of the dependent variable and all three predictor variables for each municipality in the neighborhood Θa for the mean of the response, and a separate model consisting of just the border term for the precision of the response. We first standardized predictors of relative municipality size and closeness variables via z-scores, and then estimated the beta models via maximum likelihood. These analyses were implemented in R (R Core Team 2000) and the betareg package (Zeileis et al., 2016). For a visual guide of the whole procedure, please refer to Figure 6 .Figure 6 Visual overview of the model construction framework.
Figure 6
The spatial pattern of wB is of particular interest to us. Specifically, spatial variation in this coefficient allows us to quantify and visualize spatial variation in the border effect. We then modeled the precision parameter of the beta regression as a linear function (identity link) of border presence, wBΦ * Bab, which improved both model fit and convergence relative to a constant precision model.
2 Results
The model coefficients (wD, wB, and wP) are estimated based on the three predictor variables for each municipality, from which the wB, the effect of the border, was the focal point of our study. The model summary is displayed in Figure 7 , a more detailed picture of spatial distribution of the border effect weight is displayed on map in Figure 8 . This map shows that within most of the municipalities in our study area, the border strongly inhibits virus spread (pink to magenta colors), but with pronounced the spatial variation in the strength of this inhibitory effect. In particular, the national border appears to have a less inhibitory effect (greener colors) in the area around the town of Löbau in eastern Saxony.Figure 7 Regression models result summary for the municipalities close to the national border.
Figure 7:
Figure 8 Map of municipalities in the neighborhood of the Saxon-Czech border. The circle size represents the population, and the color depicts the estimated wB, which represents the potential of the border to inhibit the spread of COVID-19. The green color stands for a less inhibitory effect of the border, while magenta tones represent municipalities where the border strongly inhibited disease spread. The width of the circle stroke represents the p-value of the border parameter of the municipality. The location of the municipalities Löbau and Šluknov is highlighted.
Figure 8
In total, 365 local models were constructed, one for each municipality that fulfilled the condition to have at least five municipalities in the neighborhood that lay on the other side of the border. Only three of those models did not converge, and was therefore removed from the analysis. Across all the remaining 362 local models, the median r2 value is ∼0.42. Our methodology was not capable of identifying statistically significant results (p < 0.05) mostly for the municipalities located on the outskirts of the study area, which may have been influenced by regions not considered in this study. From the summary numbers in Figure 7, we can conclude that the population size has the highest impact on the import potential. This result was expected, while compared to other studies, we did not use relativized incidence numbers, mainly because of the high diversity of municipalities, including very small ones. On the other hand, the closeness predictor has only a small impact on the model results, and the p-value is lower than 0.05 only in the case of 81 municipalities out of 362. While the analysis in Figure 5 suggested the impact of the autocorrelation on the temporal distances between municipalities, we assume that the regions close to the national border may have different dynamics. Results of the border predictor highlight marked asymmetry on opposite sides of the national border, with a higher potential of virus import from Czechia to Saxony than vice versa. This may be explained by the overall spatio-temporal trend of the spread over the timeframe of our study, in which the Northern Czech Republic (and specifically the province of Cheb) was reported as a European hotspot at the very beginning of the year 2021, followed by high incidence values in Saxony in the following months (March, April). Another explanation may be the effect of a different quality of COVID-19 tracking, healthcare quality, and different internal regulations in both countries.
The border region near the towns of Löbau and Bautzen in eastern Saxony stands out as a hotspot where increases in focal municipalities are more strongly linked to case number increases on the Czech side of the border. There are also single municipalities in the other parts of both Saxony and Czechia that show weaker inhibition (e.g., Marienberg, Johanngeorgenstadt, Neuhausen, and Dubí located in the central part of the study area). Outside of these exceptions, the overall effect of the border on disease transmission is mixed or strongly inhibitory.
To validate the results of the model, we did two post-hoc analyses. The first check is based on a comparison of the average of all weeks when cases in the focal municipality increased weighted by the magnitudes of the increases (referred to as central increase week further in the study). This number represents the differential timing of the waves in the two countries as an alternative way to look at the degree of cross-border coupling. The spatial distribution of the central increase week is shown on the left map of Figure 9 . In this picture, we can identify the spatial discontinuity created by the national border. This effect is least pronounced at two places - around the town of Šluknov (including the German neighborhoods of Löbau and Sebnitz) where the COVID-19 wave came later (during mid-March) compared to other Czech regions, and the westernmost part of Saxony, where the case numbers peaked relatively early compared to the rest of Saxony. The right map then shows the neighborhood variance (within 40 minutes of driving distance) of this value. Because the central increase week values form, in general, two spatial clusters - one for each country, the places with high values of the variance are then concentrated around the national border. One exception to this trend is located around the towns of Löbau (Germany) and Šluknov (Czechia), which again indicates a less inhibitory effect of the national border on the virus spread in this area. This result is consistent with the outcome of the beta regression analysis, which shows the highest potential for cross-border virus import exactly in this area.Figure 9 Weighted central increase week (left) and the neighborhood variance in central increase week (right).
Figure 9
The second post-analysis to validate the results of the model is based on the idea of the Granger causality metric (Granger, 1969, Romero García et al., 2021), which tests the causation of one time series by another time series considering a time lag. For each municipality m, we constructed two microregions. The first one consisted of the municipalities falling into a driving distance limit of 20 minutes (half of the threshold used for the model) from m and the same country as m, and the second consisted of the municipalities in the driving distance of 40 minutes from m but located on the other side of the border. We summed weekly case numbers within both microregions and calculated the change. These values were further used to perform the Granger causality test. Figure 10 shows the map of Granger causality p-values for the time lag of one week. This map has a similar spatial pattern than the map of border effect weight (Figure 8). This map shows the Löbau as one the hotspots of a proved granger causality, which is in accordance with the results of the model.Figure 10 A map showing a granger causality test p-value considering the temporal lag of 1 week. The granger causality values were calculated by comparing the case values in the neighboring municipalities in the same country to the case values in neighboring municipalities on the other side of the border. A blue circle depict municipalities with a granger causality test p-value lower than 0.2.
Figure 10
3 Discussion
In this paper, we proposed a regression-based methodology to estimate spatial variation in the effect of a national border on disease spread. This calculation does not directly measure the real import of the virus. Instead, it is a reasonable approximation of the possibility to explain an increase in the municipality a with the number of cases in municipality b, one week before this increase. Our approach leverages fine-scale, spatially-referenced case count data to estimate a location-specific index of virus import potential. We then demonstrated the power of our technique on a municipality-level case study of COVID-19 spread in the Saxon-Czech border region. Consistent with other studies of border effects, we found an overall inhibitory effect of the national border on disease spread between the two regions. Further, the border inhibition is stronger in the direction of Saxony to Czechia than vice versa, which may have been a result of the different epidemiological dynamics in the two countries. While the incidence in Czech municipalities peaked earlier than in Saxony, a cross-border commuter from Czechia would be more likely to get infected in Czechia and carry it over to Saxony than the other way around. Importantly, our approach also highlighted pronounced spatial variation in the inhibitory effect of the border, particularly in the region around Löbau in Saxony. Specifically, the border provided much weaker inhibition of disease spread in this region from Czechia to Saxony compared to most other areas along the border. This hotspot was clearly identified by our local regression models, and this central result was also verified by post-hoc analysis focusing on differences in the timing of the outbreaks on either side of the border.
In the hotspot around Löbau, the timing of the localized case count peak was much closer to that of the neighboring region in Czechia (neighborhood of town Šluknov) than it was to other accessible municipalities in Saxony. This finding suggests a pronounced, cross-border coupling of epidemic dynamics in this region, which our regression models clearly identified. In contrast, the central region of the border displayed a stronger inhibitory effect, with the disease dynamics between neighboring regions of Saxony and Czechia more clearly decoupled.
We suspect that these localized differences in border effect, particularly those between the eastern (Löbau) and central border regions, are driven by differential cross-border traffic patterns. The restrictions for the travel between two countries during the study period were regularly updated, mainly in the Czechia - Saxony direction. But during the whole time period, travel was highly restricted with the number of exceptions valid for crossing in one or both directions. First of all, commuters traveling for work from the Czech Republic to Saxony, on a daily or weekly basis, are essential to several sectors of the Saxonian economy, including healthcare, tourism, transport, and manufacturing (Sujata et al., 2020). Some of these workers had exceptions even during the hardest lockdown, and some were pushed to minimize border crossings ( Sachsen.de , 2021a; Sachsen.de , 2021b). The state authorities announced mandatory testing in January 2021 for all cross-border workers ( Sachsen.de , 2021c), which increased the case numbers on the German side. On the other hand, people from Saxony used to go shopping in Czechia regularly due to the lower prices of some commodities (e.g., gas, groceries, and some services), and the closure of the border would sharply reduce such traffic. Physical conditions also modulate cross-border mobility. For example, the Ore Mountains (Erzgebirge / Krušné Hory) along the central part of the Saxon-Czech border form a natural barrier for many cross-border activities. The distribution of cities and areas of higher population density between sides of the borders can also have an effect. Specifically, the central region of the border features larger cities on the Czech side than on the German side. When coupled with the mountainous terrain along the border and sparser road networks, this difference in population density may reduce border crossings both for Czechs (less need to commute for work) and for Germans (less convenient to commute for commodities and leisure). In contrast, the opposite factors would likely contribute to increased border traffic in the region around Löbau. While our results are consistent with this hypothesis, explicit data on cross-border movements, which were unavailable to us, would be required to confirm this interpretation fully.
A limitation of our study is the lack of direct consideration of some of the surrounding regions, which may have influenced COVID-19 dynamics near the Saxon-Czech border, notably Poland to the far east and the German states of Bavaria and Thuringia to the west. The unavailability of municipality-scale data from these regions precluded their consideration in our analysis. This issue manifests itself in the lower pseudo R2 values for the models and higher p-values on the border term that tend to occur on the outer edges of our study area. This is specifically relevant for the regions of Görlitz in the east and Cheb / Plauen in the west. Outside of these outlying regions, we are confident that our analysis accurately quantifies disease import potential from surrounding regions and provides reliable results. This robustness is further evidenced by the agreement between our core model-based analysis and the confirmatory analyses based on timing differences and on Granger causality.
Our index of virus import is based on a comparison of the case increase in a focal municipality with the relative number of cases in the neighboring municipalities one week earlier. Calculation of this index therefore requires determining appropriate values for neighborhood size, time lag, and relative weightings of contributions with the neighborhood. While our selection of these parameters was guided by literature review and knowledge gained via exploratory analysis, we tested the robustness of our results by various driving distance thresholds, temporal lags, and weighting techniques. Similarly, we considered a range of options for dealing with exact 0’s and 1’s in our index of virus import. Collectively, these alternative modeling assumptions do slightly affect our numerical results; however, these variations changed neither our qualitative results nor the inferences we drew from our analysis. Finally, we urge caution in interpreting patterns in our index of virus import. While this metric identifies areas of the higher potential for cross-border disease transfer, it does not directly measure such transfer. In other words, the high estimated potential for cross-border transmission does not guarantee that such transmission actually happened.
We have developed a method for quantifying fine-scale variation in the effect of a border on between-country (or between-region) disease transmission potential. A key advantage of our approach is that it does not require data on cross-border movements of individuals, which are generally difficult to obtain and come with substantial privacy concerns. Furthermore, our approach is able to detect potential hotspots for cross-border disease transmission, which might then be flagged for additional monitoring efforts or stricter border measures in the future. While this type of analysis is retrospective, it could potentially be automated such that, as new data become available, it provides near real-time results on patterns of cross-border disease transmission. Such an automated system could help authorities decrease the time needed to adjust border policies to changing epidemic conditions. A related issue concerns determining the extent to which such cross-border transmission hot spots are consistent over time. The multiple waves of COVID-19 that have occurred worldwide should provide ample opportunities to address this and related questions in future studies.
4 Uncited References
GeoBasis-DE, Statistisches Bundesamt 2018, Grimée et al., 2021, Laroze et al., 2021, Sachsen.de 2021a, Sachsen.de 2021b, Sachsen.de 2021c, Berry and Ayers, 2006
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.
Acknowledgments
This work was partially funded by the Center of Advanced Systems Understanding (CASUS), which is financed by Germany's Federal Ministry of Education and Research (BMBF) and by the Saxon Ministry for Science, Culture, and Tourism (SMWK) with tax funds on the basis of the budget approved by the Saxon State Parliament.
==== Refs
References
Arauzo-Carod J-M A first insight about spatial dimension of COVID-19: analysis at municipality level Journal of Public Health 43 1 2021 98 106 10.1093/pubmed/fdaa140 32808010
Becker K Terekhov I Gollnick V A global gravity model for air passenger demand between city pairs and future interurban air mobility markets identification 2018 Aviation Technology, Integration, and Operations Conference 2018 American Institute of Aeronautics and Astronautics Atlanta, Georgia 10.2514/6.2018-2885 25 June 2018
Bekker-Nielsen Dunbar M and Held L (2020) Endemic-epidemic framework used in Covid-19 modelling : discussion on the paper by Nunes, Caetano, Antunes and Dias. Instituto Nacional de Estatística. DOI: 10.5167/UZH-198715.
Berry DA Ayers GD Symmetrized Percent Change for Treatment Comparisons The American Statistician 60 1 2006 27 31 10.1198/000313006X90684
Bundesagentur für Arbeit (2020) Regionale Mobilität - Statistik der Bundesagentur für Arbeit. Available at: https://statistik.arbeitsagentur.de/DE/Navigation/Statistiken/Themen-im-Fokus/Regionale-Mobilitaet/Regionale-Mobilitaet-Nav.html (accessed 20 December 2021).
Centrum dopravního výzkumu, v. v. i. (2019) První celostátní průzkum dopravního chování. Available at: https://www.ceskovpohybu.cz/#tl (accessed 19 December 2021).
Chen Y Li Y Feng S Gravitational scaling analysis on spatial diffusion of COVID-19 in Hubei Province, China PLOS ONE 16 6 2021 10.1371/journal.pone.0252889 Public Library of Science: e0252889
Cole MA Ozgen C Strobl E Air Pollution Exposure and Covid-19 in Dutch Municipalities Environmental and Resource Economics 76 4 2020 581 610 10.1007/s10640-020-00491-4 32836849
Cori A Ferguson NM Fraser C A New Framework and Software to Estimate Time-Varying Reproduction Numbers During Epidemics American Journal of Epidemiology 178 9 2013 1505 1512 10.1093/aje/kwt133 24043437
Czech Office for Surveying, Mapping and Cadastre (2021) Chosen RÚIAN data distributed by municipalities in the SHP format. Available at: https://geoportal.cuzk.cz/(S(fbm4po4b4wmxhybo1qmqynnb))/Default.aspx?lng=EN&mode=TextMeta&side=dsady_RUIAN_vse&metadataID=CZ-00025712-CUZK_SERIES-MD_RUIAN-OBCE-SHP (accessed 18 December 2021).
Czech Statistical Office (2021) Population of Municipalities - 1 January 2021. Available at: https://www.czso.cz/csu/czso/population-of-municipalities-1-january-2021 (accessed 17 December 2021).
Douglas DH Peucker TK Algorithms for the reduction of the number of points required to represent a digitized line or its caricature. Cartographica the international journal for geographic information and geovisualization 10(2) 1973 University of Toronto Press 112 122
Eckardt M Kappner K Wolf N Covid-19 Across European Regions: The Role of Border Controls ID 3688126 SSRN Scholarly Paper 2020 Social Science Research Network Rochester, NY 1 AugustAvailable at https://papers.ssrn.com/abstract=3688126 accessed 9 September 2021
Ezzat D Hassanien AE Ella HA An optimized deep learning architecture for the diagnosis of COVID-19 disease based on gravitational search optimization Applied Soft Computing 98 2021 106742 10.1016/j.asoc.2020.106742
GeoBasis-DE, Statistisches Bundesamt (Destatis), and BKG (2020) Gemeindegrenzen 2018 mit Einwohnerzahl. Available at: https://hub.arcgis.com/datasets/916dcfd2fe54465baebcd5fdf9f6ac94_0/geoservice (accessed 17 December 2021).
GIScience Research Group and HeiGIT (2021) Openrouteservice. Java. GIScience Research Group and HeiGIT. Available at: https://github.com/GIScience/openrouteservice (accessed 30 December 2021).
Granger CWJ Investigating Causal Relations by Econometric Models and Cross-spectral Methods Econometrica 37 3 1969 424 438 10.2307/1912791 [Wiley, Econometric Society]
Grimée M Dunbar MB-N Hofmann F Modelling the effect of a border closure between Switzerland and Italy on the spatiotemporal spread of COVID-19 in Switzerland 20 May Epidemiology 2021 10.1101/2021.05.19.21257329
Han X Xu Y Fan L Quantifying COVID-19 importation risk in a dynamic network of domestic cities and international countries Proceedings of the National Academy of Sciences 118 31 2021 10.1073/pnas.2100201118 National Academy of Sciences
Hossain MP Junus A Zhu X The effects of border control and quarantine measures on the spread of COVID-19 Epidemics 32 2020 100397 10.1016/j.epidem.2020.100397
infas Institut für angewandte and Sozialwissenschaft GmbH (2017) Mobilität in Deutschland. Available at: http://www.mobilitaet-in-deutschland.de/publikationen2017.html (accessed 20 December 2021).
Komenda M, Panoška P, Bulhart V, et al. (2021) COVID‑19: Přehled aktuální situace v ČR. Onemocnění aktuálně. Available at: https://onemocneni-aktualne.mzcr.cz/api/v2/covid-19 (accessed 18 December 2021).
Kramer AM Pulliam JT Alexander LW Spatial spread of the West Africa Ebola epidemic Royal Society Open Science 3 8 2016 160294 10.1098/rsos.160294
Laroze D Neumayer E Plümper T COVID-19 does not stop at open borders: Spatial contagion among local authority districts during England's first wave Social Science & Medicine 270 2021 113655 10.1016/j.socscimed.2020.113655
Lewer JJ Van den Berg H A gravity model of immigration Economics letters 99 1 2008 164 167
mattijn (2021) Topojson. Python. Available at: https://github.com/mattijn/topojson (accessed 18 December 2021).
McCallum J National Borders Matter: Canada-U.S. Regional Trade Patterns The American Economic Review 85 3 1995 615 623 American Economic Association
Mertel A Where2Test Saxony-Czechia COVID-19 new cases dataset Zenodo 2022 10.5281/zenodo.6328304 DOI
Neyens T Faes C Vranckx M Can COVID-19 symptoms as reported in a large-scale online survey be used to optimise spatial predictions of COVID-19 incidence risk in Belgium? Spatial and Spatio-temporal Epidemiology 35 2020 100379 10.1016/j.sste.2020.100379
osm-search (2021) Nominatim. Available at: https://nominatim.openstreetmap.org/ui/search.html (accessed 30 December 2021).
R Core Team (2000) R language definition. Vienna, Austria: R foundation for statistical computing.
Romero García C Iftimi A Briz-Redón Á Trends in Incidence and Transmission Patterns of COVID-19 in Valencia, Spain JAMA Network Open 4 6 2021 e2113818 10.1001/jamanetworkopen.2021.13818
Sachsen.de (2021a) Ausweitung der Übernachtungszuschüsse für tschechische Pendler aller Branchen. Available at: https://www.medienservice.sachsen.de/medien/news/247147 (accessed 20 February 2022).
Sachsen.de (2021b) Sachsen passt Quarantäne-Verordnung an: Ausnahmen für Berufs-Pendler aus Mutationsgebieten festgelegt. Available at: https://www.medienservice.sachsen.de/medien/news/247142 (accessed 20 February 2022).
Sachsen.de (2021c) Testpflicht für Grenzpendler ab 18. Januar 2021. Available at: https://www.medienservice.sachsen.de/medien/news/245443 (accessed 20 February 2022).
Sächsische Staatsregierung (2021) Infektionsfälle in Sachsen - sachsen.de. Available at: https://www.coronavirus.sachsen.de/infektionsfaelle-in-sachsen-4151.html (accessed 18 December 2021).
Schuler L, Calabrese JM and Attinger S (2021) Data Driven High Resolution Modeling and Spatial Analyses of the COVID-19 Pandemic in Germany. medRxiv: 21.
Sujata U, Weyh A and Zillmann M (2020) Kurzstudie zur Bedeutung von Grenzpendelnden für den sächsischen Arbeitsmarkt. IAB-Regional Sachsen.
Werner PA Tracing and Modeling of the COVID-19 Pandemic Infections in Poland Using Spatial Interactions Models Gervasi O Murgante B Misra S Computational Science and Its Applications – ICCSA 2021 2021 Springer International Publishing Cham 641 657 10.1007/978-3-030-86979-3_45 2021Lecture Notes in Computer Science
Wong W-K Comparing the fit of the gravity model for different cross-border flows Economics Letters 99 3 2008 474 477 10.1016/j.econlet.2007.09.018
Yan X-Y Zhou T Destination choice game: A spatial interaction theory on human mobility Scientific Reports 9 1 2019 9466 10.1038/s41598-019-46026-w 31263166
Zeileis A Cribari-Neto F Gruen B Package ‘betareg R package 3 2 2016
Zhu D Ye X Manson S Revealing the spatial shifting pattern of COVID-19 pandemic in the United States Scientific Reports 11 1 2021 8396 10.1038/s41598-021-87902-8 33875751
| 0 | PMC9741554 | NO-CC CODE | 2022-12-14 23:43:06 | no | Spat Spatiotemporal Epidemiol. 2022 Dec 11;:100560 | utf-8 | Spat Spatiotemporal Epidemiol | 2,022 | 10.1016/j.sste.2022.100560 | 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)00699-5
10.1016/j.jinf.2022.12.008
Letter to the Editor
Letter to the Editor
Changes of tuberculosis infection in mainland China before and after the COVID-19 pandemic
Xu Jie 1
Wang Yujia 2
Liu Fang 1⁎⁎
Yang Haiyan 1⁎
1 Department of Epidemiology, School of Public Health, Zhengzhou University, Zhengzhou 450001, China
2 School of Nursing, Hebi Polytechnic, Hebi 458030, China
⁎ Corresponding author: Dr. Haiyan Yang, School of Public Health, Department of Epidemiology, Zhengzhou University, No. 100 of Science avenue, Zhengzhou 450001, China, Phone: 86-371-67781248, Fax: 86-371-67781248
⁎⁎ Corresponding author: Fang Liu, Department of Epidemiology, School of Public Health, Zhengzhou University, No. 100 of Science Avenue, Zhengzhou 450001, China
11 12 2022
11 12 2022
8 12 2022
© 2022 The British Infection Association. Published by 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.
Keywords
COVID-19 pandemic
tuberculosis
incidence
==== Body
pmcDear editor:
Lai et al. (1) and Komiya et al. (2) sequentially reported in this journal the changes in tuberculosis (TB) infection before and after the COVID-19 pandemic in Taiwan and Japan. The results of both studies showed that the number of patients newly diagnosed with TB has significantly decreased since the COVID-19 outbreak (1, 2). Lai et al. suggested that interventions regarding COVID-19 in infection control such as city lockdown, wearing masks, etc. have a positive impact on TB incidence. Komiya et al. considered that the statistical decrease in TB incidence after the COVID-19 outbreak was likely influenced by the decrease in the number of Mycobacterium tuberculosis (M. tuberculosis) tests and might not reflect the true incidence of TB in Japan.
TB is a seasonal infectious disease caused by M. tuberculosis infection. It is currently the second leading cause of death from a single infectious disease after COVID-19 (3, 4, 5, 6). China is the third largest country in terms of the estimated number of TB cases (7.4% of global infections) (4). Unlike other countries, the Chinese government has been implementing strict infection control measures for respiratory diseases since January 2020 until now. However, no study is available on changes in TB infection before and after the COVID-19 pandemic in mainland China. To fill the gap, we therefore evaluated the trends in TB infection before and after the COVID-19 pandemic, which may help guide the implementation of clinical prevention strategies.
National surveillance data on TB infection provided by the Chinese Center for Disease Control and Prevention in mainland China were collected from January 1, 2017 to August 31, 2022. The results showed that there was a significant seasonality in the number of TB infection, in which the positive detection rate was highest in March and lowest in February (Fig. 1 A). In addition, there is a decreasing trend in the number of TB infection over the last 5 years (Fig. 1A). To discuss whether the COVID-19 outbreak has decreased the inflection of the TB, we removed the seasonal component, a redundant component usually caused by the weather and holidays that occurred similarly every year, to get a much more reliable result. The seasonal component is calculated by averaging the same month of every year, so it's a function of the month, independent of the year.N(m)=1#ofyears∑yN(y,m)
Fig. 1 The monthly new cases of TB before and after the COVID-19 outbreak. A) the raw monthly new cases of TB (solid line) and the seasonal component (dash line) in the upper panel, and the residual after removing the seasonal component in the lower panel. B) the residual distributions before (black) and after (red) the outbreak. They illustrated that the outbreak decreases the number of new cases significantly (P-value < 105 from the student's t-test on the two distributions in Figure B).
Fig 1:
In this function, the N (y, m) is the number of new cases in the year (y) and the month (m). And the adjusted series will be N(y,m)−S(m) . After the seasonal adjustment, a t-test was used to test the difference between the adjusted number of new cases before and after the COVID-19 outbreak. Results from the analysis of the number of positive TB infections revealed a significant difference in the number of TB infections before and after the COVID-19 outbreak in China (Fig. 1A and 1B; P < 0.0001). Additionally, we also observed that the adjusted value of TB infections had a local minimal value in February 2020 and April 2022, respectively, which coincided with the time of household isolation implemented by the Chinese government through a society-wide lockdown from January 24, 2020 to May 2020 and from April to May 2022. The main cause might be that the government imposed the lockdowns in many areas during these two time periods and restricted access to hospitals, especially for people with non-emergency symptoms. In this case, the physicians were deprived of the opportunity to suspect TB infection and to test for M. tuberculosis. It implies an increase in the number of undiagnosed, untreated TB patients, which will lead to an increase in the number of deaths due to TB and a potential risk of community transmission. Besides, absence of immune stimulation due to decreased TB diagnoses and relative decrease in vaccine uptake may lead to "immune debt", which may negatively impact and potentially make the population more vulnerable in the next epidemic season (7).
Otherwise, the COVID-19 pandemic brought about lifestyle changes, including encouragement to wear masks, wash hands, maintain social distance, and suspend mass gatherings. These preventive measures against transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) may have contributed to the decline in many types of infectious diseases, especially respiratory infectious diseases. This short-term positive impact is welcome. Whereas, the majority of TB observed worldwide is secondary (reactivated) TB, which usually develops after decades due to reactivation of latent TB infection (8, 9). Thus, infection control measures enacted after the COVID-19 outbreak to prevent the spread of SARS-CoV-2 would not be expected to show a real impact on the number of TB infections until decades later. Nevertheless, the COVID-19 pandemic influenced the epidemiological trend of TB infections in mainland China. It may be mainly attributed to a series of strict measures taken during the COVID-19 pandemic, as well as the increased awareness of self-protection and protective measures taken by people, which also reduced the chance of TB infection. The global pandemic is still ongoing and the situation of epidemic prevention and control is still severe, despite the fact that COVID-19 still occurs in some cities in China and is generally under control, so the long-term epidemiological trend of TB deserves our continuous attention.
In conclusion, there has been a devastating impact on access to TB diagnosis, treatment, and the burden of TB due to the COVID-19 pandemic. Intensive efforts are urgently needed to mitigate and reverse the negative impact of the COVID-19 pandemic on TB.
Author contributions
Haiyan Yang and Fang Liu designed the study. Jie Xu and Yujia Wang conducted data collection. Jie Xu conducted statistical analyses. Jie Xu wrote the manuscript. All the authors approved the final version of the manuscript.
Data availability statement
The data that support the findings of this study are included in this article and available from the corresponding author upon reasonable requests.
Declaration of Competing Interest
The authors declare that they have no any potential conflict of interest regarding this submitted manuscript.
Funding
This study was supported by the grant from National Natural Science Foundation of China (No. 82273696). The funder has no role in the data collection, data analysis, preparation of manuscript and decision to submission.
==== Refs
References
1 Lai CC Yu WL. The COVID-19 pandemic and tuberculosis in Taiwan J Infect 81 2 2020 Aug e159 ee61 PubMed PMID: 32534000. Pubmed Central PMCID: PMC7286835. Epub 2020/06/14. eng 32534000
2 Komiya K Yamasue M Takahashi O Hiramatsu K Kadota JI Kato S. The COVID-19 pandemic and the true incidence of Tuberculosis in Japan J Infect 81 3 2020 Sep e24 ee5 PubMed PMID: 32650109. Pubmed Central PMCID: PMC7338857. Epub 2020/07/11. eng
3 Luies L du Preez I. The Echo of Pulmonary Tuberculosis: Mechanisms of Clinical Symptoms and Other Disease-Induced Systemic Complications Clin Microbiol Rev 33 4 2020 Sep 16 PubMed PMID: 32611585. Pubmed Central PMCID: PMC7331478. Epub 2020/07/03. eng
4 Jeremiah C Petersen E Nantanda R Mungai BN Migliori GB Amanullah F The WHO Global Tuberculosis 2021 Report - not so good news and turning the tide back to End TB Int J Infect Dis 2022 Mar 20 PubMed PMID: 35321845. Pubmed Central PMCID: PMC8934249. Epub 2022/03/25. eng
5 Dong Y Wang L Burgner DP Miller JE Song Y Ren X Infectious diseases in children and adolescents in China: analysis of national surveillance data from 2008 to 2017 Bmj 369 2020 Apr 2 m1043 PubMed PMID: 32241761. Pubmed Central PMCID: PMC7114954 at http://www.icmje.org/coi_disclosure.pdf and declare: support from the China Ministry of Science and Technology, National Natural Science Foundation of China, China Scholarship Council, and Peking University Health Science Centre for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work. Epub 2020/04/04. eng 32241761
6 Thorpe LE Frieden TR Laserson KF Wells C Khatri GR. Seasonality of tuberculosis in India: is it real and what does it tell us? Lancet 364 9445 2004 Oct 1613 1614 30-Nov 5PubMed PMID: 15519633. Epub 2004/11/03. eng 15519633
7 Cohen R Ashman M Taha MK Varon E Angoulvant F Levy C Pediatric Infectious Disease Group (GPIP) position paper on the immune debt of the COVID-19 pandemic in childhood, how can we fill the immunity gap? Infectious diseases now 51 5 2021 Aug 418 423 PubMed PMID: 33991720. Pubmed Central PMCID: PMC8114587. Epub 2021/05/16. eng 33991720
8 Mori T Leung CC. Tuberculosis in the global aging population Infect Dis Clin North Am 24 3 2010 Sep 751 768 PubMed PMID: 20674802. Epub 2010/08/03. eng 20674802
9 Furin J Cox H Pai M. Tuberculosis. Lancet. 393 10181 2019 Apr 20 1642 1656 PMID: 30904262. Epub 2019/03/25. eng 30904262
| 36516903 | PMC9741555 | NO-CC CODE | 2022-12-14 23:29:55 | no | J Infect. 2022 Dec 11; doi: 10.1016/j.jinf.2022.12.008 | utf-8 | J Infect | 2,022 | 10.1016/j.jinf.2022.12.008 | oa_other |
==== Front
Clinical Immunology Communications
2772-6134
2772-6134
The Authors. Published by Elsevier Inc.
S2772-6134(22)00035-X
10.1016/j.clicom.2022.12.001
Article
Partial recovery of SARS-CoV-2 immunity after booster vaccination in renal transplant recipients
Urra J.M. ab⁎
Castro P. c
Jiménez N. a
Moral E. c
Vozmediano C. bc
a Immunology, Hospital General Universitario de Ciudad Real, Ciudad Real, Spain
b Facultad de Medicina de Ciudad Real, Universidad de Castilla La Mancha (UCLM), Spain
c Nephrology, Hospital General Universitario de Ciudad Real, Ciudad Real, Spain
⁎ Corresponding autor at: Immunology, Hospital General Universitario de Ciudad Real, c/Obispo Torija s/n, Ciudad Real 13005, Spain.
11 12 2022
12 2023
11 12 2022
3 15
11 11 2022
24 11 2022
10 12 2022
© 2022 The Authors. 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.
The pandemic caused by the SARS-CoV-2 coronavirus has been especially detrimental to patients with end-stage renal disease. History with other vaccines suggests that patients with renal disease may not respond adequately to the SARS-CoV-2 vaccine. The aim of this study is to evaluate the immunity to SARS-CoV-2 mRNA vaccines in renal patients. Post SARS-CoV-2 vaccination first, and after the booster dose, antibodies and cellular immunity were studied in patients on hemodialysis (N = 20), peritoneal dialysis (N = 10) and renal transplantation (N = 10). After the two doses of vaccine, there was an effective immunity in dialysis patients, with 100% seroconversion and 87% detection of cellular immunity (85% in hemodialysis and 90% in peritoneal dialysis). In contrast, in renal transplant recipients there was only 50% seroconversion and cellular immunity was detected in 30% of patients. After the booster dose, all dialysis patients achieved a cellular and antibody immunity, whereas in transplant patients, despite improvement, 20% did not produce antibodies and in 37.5% cellular immunity could not be detected. The mRNA vaccine plus booster performs excellently in dialysis patients, whereas in kidney transplant recipients, despite the booster, complete immunization is not achieved.
Keywords
SARS-CoV-2 vaccine
End stage kidney disease
Hemodialysis
Peritoneal dialysis
Renal transplant recipient
Abbreviations
HD, hemodialysis
CAPD, peritoneal dialysis
KTR, kidney transplant recipient
ESKD, end stage kidney disease
RBD, receptor binding domain
==== Body
pmc1 Introduction
The pandemic caused by the SARS-CoV-2 coronavirus has been especially deleterious in patients with end stage kidney disease (ESKD) because of its severity and high mortality [1]. Patients on dialysis maintenance and those who have received a renal transplant have been particularly affected. Multiple comorbidities can explain the severity of COVID-19 in patients with renal involvement, such as age, metabolic alterations in these patients, as well as the immunosuppression originated, either by pharmacological therapy or dialysis related. Given this situation, international clinical guidelines have recommended urgent vaccination against SARS-CoV-2 in ESKD patients [2]. Although large-scale clinical trials have demonstrated the high efficacy of the vaccines developed to date, in patients with certain comorbidities this efficacy is seriously impaired [3]. ESKD patients may not develop a good response to vaccination not even after a previous infection with SARS-CoV-2. Initial studies described that hemodialysis patients have a lower response to SARS-CoV-2 vaccines in terms of antibody production [4]. Experiences in vaccinations against influenza [5] and hepatitis B [6,7] have shown a decreased immunity in kidney transplant patients and hemodialysis patients compared to the general population. The evaluation of the immunity to the vaccine in terms of antibody production determines its efficacy, however, in a population characterized by its low rate of seroconversion compared to the non-immunosuppressed population, the determination of the cellular immunity may be relevant. Two mRNA vaccines have been approved for use: mRNA-1273 (Moderna) and BNT162b2 (Pfizer-BioNTech). Both consist of lipid-encapsulated nanoparticles containing the mRNA encoding the prefusion-stabilized full-length SARS-CoV-2 spike. Extensive trials have confirmed 95% efficacy against COVID-19, largely by preventing hospital admissions and mortality from the disease.
The aim of this study is to evaluate the response to SARS-CoV-2 mRNA vaccines in ESKD patients undergoing two dialysis modalities (hemodialysis and peritoneal dialysis) or transplanted patients and to compare it with a healthy population. Both the humoral immunity through antibody quantification and the cellular immunity through IFN-γ production were studied. Vaccine response was analyzed after the two initial doses of mRNA vaccine and the effect of a third booster dose in non-responder patients was evaluated.
2 Material and methods
2.1 Study subjects
To evaluate the humoral and cellular immunity to the SARS-CoV-2 vaccine, 30 patients with ESKD treated at the Hospital General Universitario de Ciudad Real undergoing dialysis maintenance and vaccinated with two initial doses of the mRNA-1273 vaccine (Moderna) were included in the study. Twenty of them underwent hemodialysis (HD) and the other ten underwent continuous ambulatory peritoneal dialysis (CAPD). Simultaneously, ten kidney transplant recipients (KTR) vaccinated with two doses of BNT162b2 (Pfizer-BioNTech) were included in the study. The mean time of transplantation was 16.6 years (range 1-23). Regarding their immunosuppressive treatments, 9 of them were treated with mycophenolate mofetil, 7 with tacrolimus, 3 with sirolimus, 3 with prednisone and 1 with cyclosporine. The results were compared with a healthy control population composed of ten healthy individuals who had received two doses of BNT162b2 vaccine (Pfizer-BioNTech). Vaccine allocation was performed following the instructions of the national health authorities. Vaccination was performed following the first and second dose interval recommended by the manufacturers, 21 days for the BNT162b2 vaccine and 28 days for the mRNA-1273 vaccine. A summary of the demographic characteristics of the different study groups are presented in Table 1 .Table 1 Demographic characteristics and time elapsed since the second dose of vaccine of the patients enrolled in the study classified according to the clinical group to which they belong.
Table 1 N AGE (Years) SEX (F/M) TIME FROM 2° DOSE (Days)
HEALTH 10 45.8 (28 - 62) 6/4 209,6 (200-220)
HD 20 53.6 (34 - 65) 9/11 23,7 (22-26)
CAPD 10 56.3 (33 - 72) 7/3 23,9 (22-26)
KTR 10 59.7 (37 - 70) 4/6 108,7 (90-122)
2.2 Humoral immunity assessment
Post-vaccination, the humoral immunity was identified by quantifying antibodies against the spike protein of SARS-CoV-2 virus. For this purpose, the ARCHITECT IgG II Quant test (Abbott) was used. This test measures serum IgG class antibodies against the receptor binding domain (RBD) of the S1 subunit of the virus spike using chemiluminescent microparticle immunoassay (CMIA) technology. The resulting chemiluminescent reaction is recorded as arbitrary units per milliliter (AU/mL). There is a direct relationship between the amount of IgG antibody to SARS-CoV-2 in the sample and the AU/mL detected by the optics of the system. The detection range of the system is 5.7 to 40000 AU/mL. The presence of specific antibodies is assumed for values above 50 AU/mL. According to the manufacturer's specifications, the test has a sensitivity of 97.56% and a specificity of 99.6%.
2.3 Cellular immunity assessment
For the evaluation of the cellular immunity against the vaccine, an INF-γ release assay (IGRA) previously accepted [8] was used (Euroimmun). For this purpose, a tube of whole blood anticoagulated with lithium heparin was drawn. Heparinized whole blood samples were incubated in three test tubes, a positive control nonspecifically stimulated with a mitogen, a negative control, and a third coated with SARS-CoV-2 S1 antigen. Samples were incubated at 37 °C for 24 h and then centrifuged for 2 min at 8,000 rpm. The plasma supernatant was then analyzed for released INF-γ using ELISA plates provided by the manufacturer. The IFN-γ concentration in mU/mL is calculated from the absorbances obtained in a calibrator curve. A specific cellular response against SARS-CoV-2 was considered when the stimulation index (SI), defined as the ratio of IFN-γ detected between the sample stimulated with the S1 antigen and the negative control, was superior to 2.
2.4 Statistical analysis
Quantitative variables are expressed as the mean with standard deviation, while qualitative variables, such as the presence or absence of response to the vaccine, are reported as absolute numbers and relative frequencies. Differences in qualitative variables are evaluated using the Chi-square test or Fisher's exact test when any of the expected values is less than five or the variable does not have a normal distribution. Comparison of quantitative variables, if made between two groups, was performed with the Mann-Whitney U test, and for analyses involving more than two groups, the Kruskal-Wally's test was used. The relationship between two variables was examined by simple linear regression analysis. For all tests a P < 0.05 was considered statistically significant. We performed the analyses with the statistical software IBM SPSS Statistics version 26, and the graphs were generated with GraphPad Prism version 8 (GraphPad Software).
2.5 Study approval
All patients were informed about the study and gave written informed consent for their participation. The study protocol was approved by the ethics committee of the Gerencia Integrada de Ciudad Real (B-127) and was carried out in accordance with its guidelines. The study was conducted in accordance with the Declaration of Helsinki of the World Medical Association.
3 Results
3.1 Post-vaccination immunity
After the two doses of SARS-CoV-2 vaccine, different behavior was observed among the different groups analyzed in the study. Most notably, KTR patients had a very limited response to the vaccine, both in humoral and cellular immunity. No antibodies to SARS-CoV-2 protein S were detected in 5 of the 10 patients and no cellular immunity was observed in 7 of the 10 patients studied. As shown in Table 2 , HD and CAPD patients behaved similarly to the healthy group and the presence of antibodies was detected in all of them. However, in the cellular immunity studies, 3 of the HD patients and one of the CAPD patients did not reach the defined stimulation index (SI) threshold. As expected, all healthy controls had a demonstrable cellular immunity in terms of SI > 2.Table 2 Humoral and cellular response of the groups studied. Antibodies are expressed in AU/mL and cellular immunity as Stimulation Index (SI), both as mean ± standard deviation. Number of patients and in parentheses the percentage with an effective humoral (antibody titer >50 AU/mL) and cellular (Stimulation Index > 2) immunity.
Table 2 Antibodies AU/mL Antibodies > 50 AU/mL SI SI> 2
HD 17634 ± 3488 20/20 (100 %) 9.2 ± 2.1 17/20 (85 %)
CAPD 6254 ± 3858 10/10 (100 %) 9.4 ± 1.8 9/10 (90 %)
KTR 501 ± 301 5/10 (50 %) 2.4 ± 0.6 3/10 (30 %)
HEALTHY 1528 ± 275 10/10 (100 %) 7.8 ± 3.4 10/10 (100 %)
Mean antibody level was significantly lower in KTR patients (501 ± 301 AU/mL) compared to the rest of the groups analyzed, as shown in Fig. 1 A. Healthy patients had lower antibody (1528 ± 275 AU/mL) than HD and CAPD patients, which could be explained by the longer time elapsed since the last vaccine dose (24 days vs. 210). However, there were significant differences (p = 0.03) between HD (17634 ± 3488) and CAPD (6254 ± 3858 AU/mL) patients, despite the times since vaccination were similar. In contrast to the level of antibodies, as shown in Fig. 2 B, the SI of the cellular immunity was similar between HD (9.2 ± 2.1) and CAPD (9.4 ± 1.8) patients and the healthy group (7.8 ± 3.4). The cellular immunity seems to be less influenced by the time elapsed since vaccination. As well as the humoral immunity, patients with KTR have the lowest SI values (2.4 ± 0.6). Two patients in the HD group and one healthy individual had high SI values above 30, which in the case of the two HD patients corresponded to the maximum antibody detection value in the test. All were found to have been previously infected with SARS-CoV-2.Fig. 1 The figure shows the antibody levels in the different groups studied expressed in AU/mL units (A), as well as the cellular immunity determined as SI, defined as the proportion of IFN production in the sample stimulated with the S1 fraction of the virus compared to the sample without antigen (B). AU/mL values greater than 50 and SI values greater than 2 are considered positive. Values below the established detection limits are shown as empty plots.
Fig 1
Fig. 2 Mean results of antibody quantity in AU/mL units (A) and cellular immunity in SI (B) measured in KTR patients before and after the booster dose of SARS-CoV-2 vaccine. AU/mL values above 50 and SI values above 2 are considered positive. Values below the established detection limits are shown as empty plots.
Fig 2
No correlation was found between the results of humoral and cellular immunity after vaccination with age, sex, time on dialysis, time since the last transplant, number of previous transplants and immunosuppressive treatment. Nor was any relationship found between vaccine response and the HLA antigens.
3.2 Post-vaccine booster
After the booster dose, 8 KTR patients were studied (the other two could not receive the booster dose because they had COVID-19), along with the three HD patients and one CAPD who did not previously generate a cellular response after the two-dose vaccination schedule. After the booster dose, all four dialysis patients had a high cellular immunity (3.3 - 5.8) as well as a high detectable antibody titer (29102 – 40000 AU/mL).
Regarding the behavior of KTR patients after the booster dose, there was an increase in both, the number of patients with detectable antibodies and their titer. Patients in whom antibodies were detectable increased non-significantly, from 5/10 (50%) to 6/8 (80%); p = 0.17, as shown in Fig. 2A. The mean antibody titer was significantly increased from 501 ± 301 to 4862 ± 3231 AU/mL; p = 0.03. On the other hand, cellular immunity had a very discrete increase from 2.41 ± 0.6 to 2.54 ± 0.7 as shown in Fig. 2B, however the number of patients in whom a immunity was detected had an increase from 3/10 (30%) to 5 /8 (62.6%) although these differences were not significant (p = 0.09).
Finally, Fig. 3 shows individually the variations after the booster dose in humoral and cellular immunity of the 8 KTR patients. As can be seen, practically all patients had increases, although in some of them they were not sufficient to be above the detection limit. After the booster dose, in none of the patients did the antibodies or cellular immunity fall below the established detection limits.Fig. 3 Individual variations for antibody quantity (A) and cellular immunity (B) in KTR patients before and after the booster dose. AU/mL values greater than 50 and SI values greater than 2 are considered positive.
Fig 3
4 Discussion
The objective of this study was to evaluate the immunity to two doses of mRNA vaccines against SARS-CoV-2 in ESKD patients, taking into account the excellent results that these vaccines have shown in the general population [9,10] and that this novel design had never been used. It is well known that in ESKD patients the immunity to vaccines is generally poorer than in the general population [11], although mRNA-based vaccines had not been included until the COVID-19 pandemic and their behavior in immunosuppressed patients is unknown. Here we report anti-SARS-CoV-2 antibody and T-cell immunity after two doses of mRNA vaccine in a cohort of KTR and hemodialysis patients. We found that 100% of chronic dialysis patients developed a substantial humoral immunity after the second dose of the vaccine. The first conclusion of our work is the excellent antibody immunity generated by dialysis patients after the vaccination schedule. Emerging humoral response data from other hemodialysis cohorts have shown comparable seroconversion results ranging from 95% to 100% [12], [13], [14], [15], [16]. This new mRNA-based technology in vaccine development proves to be effective even in dialysis patients. A characteristic of our study is that, unlike others, the healthy population presented a lower level of antibodies than the dialysis patients. Unfortunately, in our cohort there was an important difference in the time elapsed since the last dose of vaccine, which was much longer in the healthy control group. Numerous studies have reported a rapid decline in the amount of antibody over time, and this affects both the general population and dialysis patients [15,17]. Because of the difference in time since vaccination, the amount of antibody between dialysis patients and healthy patients is not properly comparable in our study. The greatest differences in the humoral immunity were observed in the KTR group, in which both seroconversion and the level of antibodies were clearly different and lower than in the rest of the study groups. In KTR patients, it is to be expected that vaccine responses are impaired, originated by immunosuppressive treatments. Use of immunosuppressive therapy predicts fewer seroconverted patients and significantly lower anti-S-IgG levels. The results observed in our study are in agreement with other similar studies [16,18]. The performance of the KTR group in cellular immunity was even poorer with detection in only 30% of patients and with practically undetectable IS in several patients. These results are similar to those described in other studies [19,20] and very different from those observed in the general population both in our work and in other series [21]. The cellular immunity in dialysis patients, like the humoral immunity, was consistent, although in a small proportion of patients it was not detected. The cellular immunity elicited by the vaccine is lower than the humoral immunity and is greatly enhanced by COVID-19 as previously described [22]. In our study the four patients in whom no cellular response was detected had values very close to the limit of detection. In fact, after the booster dose, all four showed a strong cellular and humoral immunity, which is evidence that the vaccine booster was very effective in dialysis patients. The same conclusion has been obtained from other series of patients by increasing antibodies such as cell-mediated immunity [23,24]. In our study and in others [24] we observed that cellular immunity is maintained over time and not decreases rapidly as occurs with antibodies. In patients under dialysis, the increased immunity caused by the booster dose even protects against non-vaccine variants such as Omicron [25]. Similar to our study, it has been described that the effect of the booster dose in KTR patients, although it clearly boosts immunity, unlike in dialysis patients, is not effective in most of the patients and some do not develop an acceptable threshold of antibodies or cellular immunity [26,27]. All patients had increased antibody and cell-mediated immunity, but in a proportion of patients this effect was not sufficient, and an effective immune response was not obtained. The immunosuppressive treatments clearly influence and prevent an adequate immunity, these drugs even in non-transplanted patients impair the response to the vaccine [28]. All transplant patients received the same vaccination and booster regimen with BNT-162b2, according to the instructions of the Spanish health authorities. Therefore, in our study it is not possible to compare with other vaccination schemes. In any case, in immunosuppressed patients it would be interesting to compare other vaccination regimens with SARS-CoV-2 RNA vaccines to optimize the response given the poor immunization observed in our study and others. What does seem obvious is that KTR patients are strong candidates to receive new doses of the vaccine, including periodic vaccinations, to obtain a more robust immunity.
In conclusion, the booster dose of the SARS-CoV-2 vaccine is very effective in patients undergoing dialysis, but in renal transplant recipients, despite increasing immunity, the protection provided is not completely effective and some of them are still exposed to infection.
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.
Acknowledgments
We thank Euroimmun Spain for the technical assistance and for providing us with the IGRA assay, and the laboratory technicians for their collaboration.
==== Refs
References
1 Jager K.J. Kramer A. Chesnaye N.C. Couchoud C. Sánchez-Álvarez J.E. Garneata L. Collart F. Hemmelder M.H. Ambühl P. Kerschbaum J. Legeai C. Del Pino Y Pino M.D. Mircescu G. Mazzoleni L. Hoekstra T. Winzeler R. Mayer G. Stel V.S. Wanner C. Zoccali C. Massy Z.A. Kidney Int. 98 2020 1540 1548 10.1016/j.kint.2020.09.006 32979369
2 Combe C. Kirsch A.H. Alfano G. Luyckx V.A. Shroff R. Kanbay M. van der Sande F. Basile C. Nephrol. Dial. Transplant. 36 2021 571 574 10.1093/ndt/gfab007 33475137
3 Dagan N. Barda N. Kepten E. Miron O. Perchik S. Katz M.A. Hernán M.A. Lipsitch M. Reis B. Balicer R.D. N. Engl. J. Med. 384 2021 1412 1423 10.1056/NEJMoa2101765 33626250
4 Simon B. Rubey H. Treipl A. Gromann M. Hemedi B. Zehetmayer S. Kirsch B. Nephrol. Dial. Transplant. 36 2021 1709 1716 10.1093/ndt/gfab179 33999200
5 Broeders N.E. Hombrouck A. Lemy A. Wissing K.M. Racapé J. Gastaldello K. Massart A. Van Gucht S. Weichselbaum L. De Mul A. Brochier B. Thomas I. Abramowicz D. Clin. J. Am. Soc. Nephrol. CJASN. 6 2011 2573 2578 10.2215/CJN.04670511 21921153
6 Elhanan E. Boaz M. Schwartz I. Schwartz D. Chernin G. Soetendorp H. Gal Oz A. Agbaria A. Weinstein T. Clin. Exp. Nephrol. 22 2018 151 158 10.1007/s10157-017-1416-7 28456864
7 Peces R. de la Torre M. Alcázar R. Urra J.M. Am. J. Kidney Dis. Off. J. Natl. Kidney Found. 29 1997 239 245 10.1016/s0272-6386(97)90036-6
8 Renaudineau Y. Abravanel F. Izopet J. Bost C. Treiner E. Congy N. Blancher A. Clin. Immunol. 237 2022 108979 10.1016/j.clim.2022.108979
9 Jackson L.A. Anderson E.J. Rouphael N.G. Roberts P.C. Makhene M. Coler R.N. McCullough M.P. Chappell J.D. Denison M.R. Stevens L.J. Pruijssers A.J. McDermott A. Flach B. Doria-Rose N.A. Corbett K.S. Morabito K.M. O'Dell S. Schmidt S.D. Swanson P.A. Padilla M. Mascola J.R. Neuzil K.M. Bennett H. Sun W. Peters E. Makowski M. Albert J. Cross K. Buchanan W. Pikaart-Tautges R. Ledgerwood J.E. Graham B.S. Beigel J.H. N. Engl. J. Med. 383 2020 1920 1931 10.1056/NEJMoa2022483 32663912
10 Walsh E.E. Frenck R.W. Falsey A.R. Kitchin N. Absalon J. Gurtman A. Lockhart S. Neuzil K. Mulligan M.J. Bailey R. Swanson K.A. Li P. Koury K. Kalina W. Cooper D. Fontes-Garfias C. Shi P.Y. Türeci Ö. Tompkins K.R. Lyke K.E. Raabe V. Dormitzer P.R. Jansen K.U. Şahin U. Gruber W.C. N. Engl. J. Med. 383 2020 2439 2450 10.1056/NEJMoa2027906 33053279
11 Chang Y.T. Guo C.Y. Tsai M.S. Cheng Y.Y. Lin M.T. Chen C.H. Shen D. Wang J.R. Sung J.M. Vaccine 30 2012 5009 5018 10.1016/j.vaccine.2012.05.016 22658967
12 Grupper A. Sharon N. Finn T. Cohen R. Israel M. Agbaria A. Rechavi Y. Schwartz I.F. Schwartz D. Lellouch Y. Shashar M. Clin. J. Am. Soc. Nephrol. 16 2021 1037 1042 10.2215/CJN.03500321 33824157
13 Frantzen L. Cavaillé G. Thibeaut S. El-Haik Y. Nephrol. Dial. Transplant. 36 2021 1756 1757 10.1093/ndt/gfab165 34450646
14 Ikizler T.A. Coates P.T. Rovin B.H. Ronco P. Kidney Int. 99 2021 1275 1279 10.1016/j.kint.2021.04.007 33931226
15 Murt A. Dinc H.O. Altiparmak M.R. Yalin S.F. Yadigar S. Parmaksiz E. Kocazeybek B. Pekpak M. Ataman M.R. Nephron 2022 1 5 10.1159/000524658
16 Sanders J.-S.F. Bemelman F.J. Messchendorp A.L. Baan C.C. van Baarle D. van Binnendijk R. Diavatopoulos D.A. Frölke S.C. Geers D. GeurtsvanKessel C.H. den Hartog G. van der Heiden M. Imhof C. Kho M.M.L. Koopmans M.P.G. Malahe S.R.K. Mattheussens W.B. van der Molen R. van Mourik D. Remmerswaal E.B.M. Rots N. Vart P. de Vries R.D. Gansevoort R.T. Hilbrands L.B. Reinders M.E.J. R. Collaborators* Transplantation 106 2022 821 834 10.1097/TP.0000000000003983 34753894
17 Naaber P. Tserel L. Kangro K. Sepp E. Jürjenson V. Adamson A. Haljasmägi L. Rumm A.P. Maruste R. Kärner J. Gerhold J.M. Planken A. Ustav M. Kisand K. Peterson P. Lancet Reg. Health Eur. 10 2021 100208 10.1016/j.lanepe.2021.100208
18 Benotmane I. Gautier-Vargas G. Cognard N. Olagne J. Heibel F. Braun-Parvez L. Martzloff J. Perrin P. Moulin B. Fafi-Kremer S. Caillard S. Kidney Int. 99 2021 1498 1500 10.1016/j.kint.2021.04.005 33887315
19 Broseta J.J. Rodríguez-Espinosa D. Rodríguez N. del M. Mosquera M. Marcos M.Á. Egri N. Pascal M. Soruco E. Bedini J.L. Bayés B. Maduell F. Am. J. Kidney Dis. 78 2021 571 581 10.1053/j.ajkd.2021.06.002 34174364
20 Bertrand D. Hamzaoui M. Lemée V. Lamulle J. Hanoy M. Laurent C. Lebourg L. Etienne I. Lemoine M. Roy F.L. Nezam D. Plantier J.-C. Boyer O. Guerrot D. Candon S. J. Am. Soc. Nephrol. 32 2021 2147 2152 10.1681/ASN.2021040480 34112706
21 Sahin U. Muik A. Derhovanessian E. Vogler I. Kranz L.M. Vormehr M. Baum A. Pascal K. Quandt J. Maurus D. Brachtendorf S. Lörks V. Sikorski J. Hilker R. Becker D. Eller A.K. Grützner J. Boesler C. Rosenbaum C. Kühnle M.C. Luxemburger U. Kemmer-Brück A. Langer D. Bexon M. Bolte S. Karikó K. Palanche T. Fischer B. Schultz A. Shi P.Y. Fontes-Garfias C. Perez J.L. Swanson K.A. Loschko J. Scully I.L. Cutler M. Kalina W. Kyratsous C.A. Cooper D. Dormitzer P.R. Jansen K.U. Türeci Ö. Nature 586 2020 594 599 10.1038/s41586-020-2814-7 32998157
22 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. Clin. Immunol. 236 2022 108961 10.1016/j.clim.2022.108961
23 Bensouna I. Caudwell V. Kubab S. Acquaviva S. Pardon A. Vittoz N. Bozman D.F. Hanafi L. Faucon A.L. Housset P. Am. J. Kidney Dis. Off. J. Natl. Kidney Found. 79 2022 185 192 10.1053/j.ajkd.2021.08.005 e1
24 Karakizlis H. Agarwal V. Aly M. Strecker K. Csala B. Esso I. Chen J. Nahrgang C. Wolter M. Slanina H. Schüttler C.G. Jessen S. Ronco C. Seeger W. Weimer R. Sester M. Birk H.-W. Husain-Syed F. J. Nephrol. 2022 10.1007/s40620-022-01371-4
25 Herman-Edelstein M. Ben-Dor N. Agur T. Guetta T. Raiter A. Meisel E. Alkeesh W. Ori Y. Rozen-Zvi B. Zingerman B. Vaccines 10 2022 967 10.3390/vaccines10060967 35746575
26 Stumpf J. Tonnus W. Paliege A. Rettig R. Steglich A. Gembardt F. Kessel F. Kröger H. Arndt P. Sradnick J. Frank K. Tonn T. Hugo C. Transplantation 105 2021 e267 e269 10.1097/TP.0000000000003903 34342963
27 Kamar N. Abravanel F. Marion O. Couat C. Izopet J. Del Bello A. N. Engl. J. Med. 385 2021 661 662 10.1056/NEJMc2108861 34161700
28 Magliulo D. Wade S.D. Kyttaris V.C. Clin. Immunol. 234 2022 108897 10.1016/j.clim.2021.108897
| 0 | PMC9741556 | NO-CC CODE | 2022-12-15 23:16:04 | no | 2023 Dec 11; 3:1-5 | utf-8 | null | null | null | oa_other |
==== Front
Chaos, Solitons & Fractals: X
2590-0544
2590-0544
The Author(s). Published by Elsevier Ltd.
S2590-0544(22)00019-7
10.1016/j.csfx.2022.100090
100090
Article
Cost-effectivness of a Mathematical Modeling with optimal control Approach of Spread of COVID-19 Pandemic: a case study in Peru
Kouidere Abdelfatah ⁎a
Balatif Omar b
Rachik Mostafa a
a LAMS, Department of Mathematics and Computer Science, Faculty of Sciences Ben M’Sik, Hassan II University of Casablanca, Morocco
b Laboratory of Dynamical Systems, Mathematical Engineering Team (INMA), Department of Mathematics, Faculty of Sciences El Jadida, Chouaib Doukkali University, El Jadida, Morocco
⁎ Corresponding author.
11 12 2022
11 12 2022
10009015 1 2021
5 11 2022
5 12 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.
COVID-19 pandemic affects 213 countries and regions around the world. Which the number of people infected with the virus exceeded 26 millions infected and more than 870 thousand deaths until september 04,2020, in the world, and Peru among the countries most affected by this pandemic. So we proposed a mathematical model describes the dynamics of spread of the COVID-19 pandemic in Peru. The optimal control strategy based on the model is proposed, and several reasonable and suitable control strategies are suggested to the prevention and reduce the spread COVID-19 virus, by conducting awareness campaigns and quarantine with treatment. coronavirus 2019 (COVID-19). Pontryagin’s maximum principle is used to characterize the optimal controls and the optimality system is solved by an iterative method. Finally, some numerical simulations are performed to verify the theoretical analysis using Matlab.
Keywords
Optimal control
SARS-COV-2
Mathematical modeling
COVID-19
==== Body
pmc1 Introduction
On March 11, 2020, the World Health Organization (WHO) announced that COVID-19 had become a global pandemic. This virus caused terror in the world. It first appeared in Wuhan, China, in December 2019. In January 2020, Chinese authorities informed the World Health Organization of a rapidly spreading virus. He named several names like COVID-19, SARS-COV-19 and NCOV-19[18].
COVID-19 virus belongs to respiratory virus strains. The most prominent of these are acute respiratory infections (SARS) and respiratory syndrome in the Middle East (MERS) [1], [2], [3], which are similar in early symptoms to symptoms of seasonal influenza, but differ in their incubation period and the severity of their aggression.
The COVID-19 virus is transmitted in many ways from an infected person to another person, the most prominent of which is touching an infected person or by repelling after sneezing or coughing. It is transmitted to the second person either by touching the mouth, nose or eye after contact with the infected person. It then moves to the larynx, then to the lung, and then develops into millions of viruses, thereby destroying the respiratory system. The severity of the COVID-19 virus lies in the period of its incubation, since the person does not show signs or symptoms of the virus, for a period ranging between 2 to 14 days and on average between 4 to 5 days. This facilitates the spread of the virus. Then, the patient develops mild symptoms similar to the seasonal cold. Then it is relatively severe with a feeling of severe fatigue and severe pain in the joints. And great difficulty breathing. Whereas in this period the most significant damage was caused to the respiratory system that leads to the collapse of the respiratory system. And death after that.
COVID-19 virus is classified as the most deadly and dangerous compared to all other corona viruses. As of Friday, September 04, 2020, 12:31 GMT, the number of people infected with the virus was 26506434, and 873882 people had died, and 18686193 had recovered. whereas the number of people with severe acute respiratory syndrome (SARS) [20] reached about 8098 people and the death of about 774 people, which also started from Asia and China specifically in 2002 and which researchers suggested transmission from Bats to humans, as for the Middle East Syndrome (MERS) virus, 858 people died out of the total number of infected people, who numbered about 2494 people since its appearance in 2012[4], [5], [6], which appeared initially in the Kingdom of Saudi Arabia.
The pandemic was COVID-19. Serious negative impact on the global economy. As the virus tightened its grip on the ground. Close several international companies, stop all types of connectors, stop commercial traffic.
The category of people with chronic diseases is most affected by the COVID-19 virus. Especially the category of diabetics, hypertension, chronic diseases, and asthma diseases. As a virus does not neglect them much until they die.
Globally, the number of infected people. In USA, about the infected reached 6336000 and the deaths reached 191114. In Brazil, 4046150 people were infected and 124729 died. In India, about 3940131 people were infected and 68598 died. According to the last stats declared on September 04, 2020, 12:31 GMT. The epidemic has moved from China to Europe and to Peru and now to South America, which has become the main epidemic area according to the World Health Organization.
The first infection with the virus was recorded in Peru on March 6,2020, while the first death occurred on the 20 of the same month.. The spread of the Covid-19 virus was initially stable. Then in recent weeks, where the spread of the virus has experienced tremendous acceleration, 3 months after the registration of the first case, approximately 670145 infected, 29405 deaths. Peru has become the fifth country in the world in terms of infected.
Where he recorded in State of Arequipa about 30997 infected, followed by State of Piura about 27490 infected, and in State of Callao Region about 26939 infected [30].
In last days. A large number of mathematical models have been developed to simulate, analyse and understand the corona virus, in a related research work, Delfim F.M. Torres et al [7] proposed a Mathematical modeling of COVID-19 transmission dynamics with a case study of Wuhan, and Zhi-Qiang Xia et al [8] Early dynamics of transmission and control of COVID-19: a mathematical modelling study. Choi SC et al [21] Estimating the reproductive number and the outbreak size of Novel Coronavirus disease (COVID-19) using mathematical model in the Republic of Korea. Also, many researches have focused on this topic and other related topics [16], [17], [22], [26], [27] and other topics [30], [31], [32].
So, in this paper we sutdy the spread of Covid-19 pandemic in Peru. By propsing a mthemetical model SEAICCWHR that describes the dynamic of transmission of this disease.
In this paper, we present the proposed mathematical model and we give some basic properties of the model in Section 2. In Section 3, we present the optimal control problem for the proposed model where we give some results concerning the existence and characterization of the optimal controls using Pontryagin’s maximum principle. Numerical simulations through MATLAB are given in section 4. Finally we conclude the paper in section 5.
2 Mathematical model
We consider a mathematical model SEAICCWHR that described the dyanamic of transmission of COVID-19 in a given population. We divide the population denoted by N into eight compartments.
2.1 Description of the Model
The graphical representation of the proposed model is shown in Figure 1 .Fig. 1 Population compartments
Fig. 1
Compartment(S) is representing the number of susceptible. The compartment S is increasing by Λ (denote the incidence of susceptible). This compartment is decreasing by the amount μ (natural mortality), β1SAN (The number of people who were infected with the virus by contact with the infected and asymptomatics people) and β2SIN (The number of people who were infected with the virus by contacting with the infected and symptomatics people).(1) dS(t)dt=Λ−μS(t)−β1S(t)A(t)N−β2S(t)I(t)N
Compartment(E) is representing the number of exposed. The compartment E is increasing by amounts β1SAN and β2SIN. This compartment is decreasing by μ (natural mortality) and also by α1E (the number of exposed become asymptomatic and infectious) and also by α2E (the number of people become infectious and symptomatic).(2) dE(t)dt=β1S(t)A(t)N+β2S(t)I(t)N−(μ+α1+α2)E(t)
Compartment(A) is representing the number of individual infected with asymptomatic. The compartment I is increasing α1E. This compartment is decreasing by μ(natural mortality) and decreasing by χA (The number of infected and asymptomatics people become infected and symptomatics) and also decreasing by θ1I (The number of infected and symptoms who have been under lockdown).(3) dA(t)dt=α1E(t)−(χ+μ)A(t)
Compartment(I) is representing the number of infected and symptomatics. The compartment I is increasing by α2E and also increasing by χA. This compartment is decreasing by μ (natural mortality) and decreasing by θI(The number of infected and symptomatics who have been under lockdown) and also decreasing by (1−ϱ)γI (The number of people become infected with complications and those without chronic disease) and also by ϱγI (The number of people become infected with complications and those with chronic disease)(4) dI(t)dt=α2E(t)+χA(t)−(γ+θ2+μ)I(t)
Compartment (C) is representing the number of infected with complications and those without chronic diseases. The compartment C is increasing by (1−ϱ)γI a. This compartment is decreasing by η1C (The number of people with serious complications who have been under lockdown) and decreasing δ1C (mortality rate due to complications) and also decreasing by μC(natural mortality).(5) dC(t)dt=(1−ϱ)γI(t)−(η1+μ+δ1)C(t)
Compartment (CW) is representing the number of infected with complications and those with chronic diseases. The compartment CW is increasing (1−ϱ)γI a. This compartment is decreasing by η2CW (The number of people with serious complications and chronic diseases who have been under lockdown) and decreasing δ2CW (mortality rate due to complications and with chronic diseases) and also decreasing by μCW(natural mortality).(6) dCW(t)dt=τγ1I(t)+ϱγ2I(t)−(η2+μ+δ2)CW(t)
Compartment (H) is representing the number of people who have been under lockdown in hospitals with follow-up and health monitoring. The compartment H is increasing by θ1Aand θ2I and also by η1C and by η2CW. This compartment is also decreasing by σH (The rate of people who recovered from the virus) and decreasing by δ3H (The number of people who died in hospitals) and also decreasing by μH(natural mortality)(7) dH(t)dt=θ1A(t)+θ2I(t)+η1C(t)+η2CW(t)−(μ+σ+δ3)H(t)
Compartment (R) is representing the number of recovered. The compartment H is increasing by σH. This compartment is decreasing by μR(natural mortality)(8) dR(t)dt=σH(t)−μR(t)
Hence, we present the COVID-2019 mathematical model is governed by the following system of differential equation :(9) {dS(t)dt=Λ−μS(t)−β1S(t)A(t)N−β2S(t)I(t)NdE(t)dt=β1S(t)A(t)N+β2S(t)I(t)N−(μ+α1+α2)E(t)dA(t)dt=α1E(t)−(θ1+χ+μ)A(t)dI(t)dt=α2E(t)+χA(t)−(γ+θ2+μ)I(t)dC(t)dt=(1−ϱ)γI(t)−(η1+μ+δ1)C(t)dCW(t)dt=ϱγI(t)−(η2+μ+δ2)CW(t)dH(t)dt=θ1I(t)+θ2I(t)+η1C(t)+η2CW(t)−(μ+σ+δ3)H(t)dR(t)dt=σH(t)−μR(t)
where S(0)≥0, E(0)≥0, A(0)≥0, I(0)≥0, C(0)≥0, CW(0)≥0, H(0)≥0 and R(0)≥0 are the initial state.
Figure 2 shows that, if Peru does not apply preventive precautions to reduce the spread of the COVID-19 pandemic, the number of COVID-19 infected cases is likely to reach millions.Fig. 2 Accumulated cases of COVID-19 in Peru after 4 months.
Fig. 2
2.2 Model Basic Properties
2.2.1 Positivity of Solutions
Theorem 1 If S(0)≥0, E(0)≥0,A(0)≥0, I(0)≥0, C(0)≥0, CW(0)≥0, H(0)≥0 and R(0)≥0, the solutions S(t), E(t), A(t), I(t), C(t), CW(t), H(t) and R(t) of system (9) are positive for all t≥0 .
Proof dS(t)dt=Λ−μS(t)−β1S(t)A(t)N−β2S(t)I(t)N≥−μS(t)−β1S(t)A(t)N−β2S(t)I(t)N
dS(t)dt+(μ+β1E(t)N+β2I(t)N)S(t)≥0
where F(t)=μ+β1E(t)N+β2I(t)N. The both sides in last inequality are multiplied by exp(∫0tF(s)ds).
We obtainexp(∫0tF(s)ds).dS(t)dt+F(t)exp(∫0tF(s)ds).S(t)≥0
then ddt(S(t)exp(∫0tF(s)ds))≥0
Integrating this inequality from 0 to t gives:∫0tdds(S(s)exp(∫0t(μ+β1E(s)N+β2I(s)N)ds))ds≥0
thenS(t)≥S(0)exp(∫0t(μ+β1E(s)N+β2I(s)N)ds)
⟹S(t)>0
.
Similarly, we prove that E(t)≥0,A(t)≥0, I(t)≥0, C(t)≥0, CW(t)≥0, H(t)≥0 and R(t)≥0. □
2.2.2 Boudedness of the solutions
Theorem 2 The set
Ω={(S,E,A,I,C,CW,H,R)∈R+8/0≤S+E+A+I+C+CW+H+R≤Λμ} positively invariant under system (9) with initial conditions, S(t)≥0, E(t)≥0,A(t)≥0, I(t)≥0, C(t)≥0, CW(t)≥0, H(t)≥0 and R(t)≥0 .
Proof Also, one assumes that:dNdt=Λ−μN−δ1C≤Λ−μN⟹N(t)≤Λμ+N(0)e−μt⟹N(t)≤Λμ
If we take limit t→∞thenN(t)≤Λμ □
2.2.3 Exictence of solutions
Theorem 3 The system (9) that satisfies a given initial condition
(S(0),E(0),A(0),I(0).C(0),CW(0),H(0),R(0)) has a unique solution.
Proof Let X=(S(t)E(t)A(t)I(t)C(t)CW(t)H(t)R(t)) and φ(X)=(dS(t)dtdE(t)dtdA(t)dtdI(t)dtdC(t)dtdCW(t)dtdH(t)dtdR(t)dt) so the system (9) is rewritten in the following form: φ(X)=AX+B(X) whereA=(−μ00000000A10000000α1A2000000α2χA30000000(1−ϱ)γA4000000ϱγ0A50000θ1θ2η1η2A60000000σ−μ)
Where A1=−( μ +α1+α2), A2=−(θ1+χ+μ), A3=−(γ+θ2+μ), A4=−(η1+μ+δ1), A5=−(η2+μ+δ2), A6=−(μ+σ+δ3) andB(X)=(Λ−β1S(t)A(t)N−β2S(t)I(t)Nβ1S(t)A(t)N+β2S(t)I(t)N000000)
|B(X1)−B(X2)|≤M..∥X1−X2∥ then ∥φ(X1)−φ(X2)∥≤V.∥X1−X2∥Where V=max(M,∥A∥)<∞, and M is constant.
Thus, it follows that the function φ is uniformly Lipschitz continuous, and the restriction on S(t)≥0, E(t)≥0,A(t)≥0, I(t)≥0, C(t)≥0, CW(t)≥0, H(t)≥0 and R(t)≥0, we see that a solution of the system (9)exists [18]. □
3 The Controlled Mathematical model
As of today 31 August 2020, there is no cure or vaccine for the disease, so we suggest the following strategies: there are two controls u(t) and v(t).The first control can be interpreted as the rate to be subjected to sensitation and prevention. So, we note that the aim of u is to diagnosed and awerness program to susceptible people at time t. The second control can be interpreted as quarantine and health monitoring, so, we note that v(I(t)+C(t)+CW(t)) is the number of indivduals people with the diseasein quarantine and health monitoring.
Hence, we present the following controlled system of differential equation :(10) {dS(t)dt=Λ−μS(t)−β1(1−u(t))S(t)A(t)N−β2(1−u(t))S(t)I(t)NdE(t)dt=β1(1−u(t))S(t)A(t)N+β2(1−u(t))S(t)I(t)N−(μ+α1+α2)E(t)dA(t)dt=α1E(t)−(χ+μ)I(t)dI(t)dt=α2E(t)+χA(t)−(γ2+θ2+μ)I(t)−v(t)I(t)dC(t)dt=(1−ϱ)γI(t)−(η1+μ+δ1)C(t)−v(t)C(t)dCW(t)dt=ϱγI(t)−(η2+μ+δ2)CW(t)−v(t)CW(t)dH(t)dt=θ1A(t)+θ2I(t)+η1C(t)+η2CW(t)−(μ+σ+δ3)H(t)+v(I(t)+C(t)+CW(t))dR(t)dt=σH(t)−μR(t)
3.1 The optimal control problem
The problem is to minimize the objective functional(11) J(u,v)=I(T)+C(T)+CW(T)+∫0T[I(t)+I(t)+C(t)+CW(t)+A12u2(t)+A22v2(t)]dt
Where A1>0, A2>0 are the cost coefficients. They are selected to weigh the relative importance of u(t) and v(t) at time t, T is the final time.
In other words, we seek the optimal controls u* and v* such that(12) J(u*,v*)=minu,v∈UJ(u,v)
Where U is the set of admissible controls defined by(13) U={u,v/0≤u(t)≤1and0≤v(t)1,t∈[0,T]}
3.2 The optimal control: existence and characterization
3.2.1 Existence of an Optimal Control
We first show the existence of solutions of the system (10), thereafter we will prove the existence of optimal control([13]).Theorem 4 Consider the control problem with system(10). There exists an optimal control(u*,v*)∈U3such thatJ(u*,v*)=minu,v∈UJ(u,v)
Proof The existence of the optimal control can be obtained using a result by Fleming and Rishel [12], checking the following steps :• It follows that the set of controls and corresponding state variables is nonempty using in Boyce and DiPrima [15]
To prove that the set of controls and the corresponding state variables is nonempty, we will use a simplified version of an existence result [15]. Let Xi′=FXi(t;X1,X2,⋯,X8) with i=1;⋯;8 where (X1,X2,⋯,X8)=(S,E,A,I,C,CW,H,R) where X1,⋯,X7 and X8 form the right-hand side of the system of equations (10). Let u,v and w for some constants and since all parameters are constants and X1,⋯,X7 and X8 are continuous, then FS,FE,FA,FI;FC,FCW,FH and FR are also continuous. Additionally, the partial derivatives ∂FXi∂Xi with i=1;⋯;8 are all continuous. therefore, there exists a unique solution (S,E,A,I,C,CW,H,R) that satisfies the initial conditions. therefore, the set of controls and the corresponding state variables is nonempty and condition 1 is satisfied.
• J(u,v)=I(T)+C(T)+CW(T)+∫0T[I(t)+C(t)+CW(t)+A12u2(t)+A22v2(t)]dt is convex in U.
• The control space U={(u,v)/(u,v) is measurable,0≤u(t)≤1 and 0≤v(t)≤1,t∈[0,T]}. is convex and closed by definition.
So, take any controls u,v∈U and λ∈[0,1]. then 0≤λu+(1−λ)v
Additionally, we observe that λu≤λ and (1−λ)v≤(1−λ), then λu+(1−λ)v≤λ+(1−λ)=1.
Hence, 0≤λu+(1−λ)v≤1, for all u,v∈U and λ∈[0,1].
• All the right hand sides of equations of system are continuous, bounded above by a sum of bounded control and state, and can be written as a linear function of u and v with coefficients depending on time and state.
From the system of differential equations (10),dNdt=Λ−μN
limt→∞supN(t)≤Λμ
Therefore, all solutions of the model (10) are bounded. So, there exist positive constants Z1,⋯,Z7 and Z8 such that ∀t∈[0,T]:S(t)≤Z1,E(t)≤Z2A(t)≤Z3,I(t)≤Z4,C(t)≤Z5,CW(t)≤Z6,H(t)≤Z7andR(t)≤Z8
We consider{FS=S.(t)≤ΛFE=E(t)≤β1(1−u(t))S(t)Z3(t)N+β2(1−u(t))S(t)Z4(t)NFA=A.(t)≤α1E(t)−(χ+μ)Z4(t)FI=I.(t)≤α2E(t)+χA(t)FC=C.(t)≤(1−ϱ)γI(t)−v(t)Z5FCW=CW.(t)≤ϱγI(t)−v(t)Z6(t)FH=H(t)≤θ1A(t)+θ2I(t)+η1C(t)+η2CW(t)+v(Z3+Z4+Z5+Z6)FR=R(t)≤σH(t)
So, we can rewrite system (10) in matrix form asF(t;S,E,A,I,C,CW,H,R)≤Λ+AX(t)−BU(t)
whereF(t;S,E,A,I,C,H,R)=[FSFEFAFIFCFCWFHFR]TΛ=[Λ0000000]TX(t)=[SEAICCWHR]TU(t)=[uv]T
A=[00000000000000000000000000000000000(1−ϱ)γ0000000ϱγ000000θ1θ2η1η200000000σ0]
B(X)=[−β1SAN−β2SIN0β1SAN+β2SIN0000I0C0CW0I+C+CW00]
It gives a linear function of control vector and state variable vector. therefore, we can writeF(t;S,E,A,I,C,CW,H,R)≤∥Λ∥+∥A1∥∥X(t)∥+∥A2∥∥U(t)∥≤φ+ψ(∥X(t)∥+∥U(t)∥)
where φ ≤ ∥Λ∥ and ψ=max(∥A1∥,∥A2∥).
Hence, we see the right-hand side is bounded by a sum of state and control vectors.
• The integrand in the objective functional I(t)+C(t)+CW(t)+A12u2(t)+A22v2(t) is clearly convex on U
It rest to show that there exists constants ζ1,ζ2,ζ3>0,
and ζsuch that I(t)+C(t)+CW(t)+A12u2(t)+A22v2(t) satisfiesI(t)+C(t)+CW(t)+A2u2(t)+B2v2(t)≥ζ1+ζ2|u|ζ+ζ3|v|ζ
The state variables being bounded, let ζ1=inft∈[0,T](I(t)+C(t)+CW(t)),ζ2=A12,ζ3=A22 and ζ=2 then it follows that:I(t)+C(t)+CW(t)+A12u2(t)+A22v2(t)≥ζ1+ζ2|u|ζ+ζ3|v|ζ
Then from Fleming and Rishel [12] we conclude that there exists an optimal control.
□
3.2.2 Characterization of the optimal control
In order to derive the necessary conditions for the optimal control, we apply Pontryagin’s maximum [11] principle to the Hamiltonian H∧ at time t defined by(14) H∧(t)=I(t)+C(t)+CW(t)+A12u2(t)+A22v2(t)+∑i=18λi(t)fi(S,A,I,C,H,R)
where fi, is the right side of the difference equation of the ith state variable.Theorem 5 Given the optimal controls(u*,v*)and the solutionsS*,E*,A*,I*C*,CW*,H*andR*of the corresponding state system(10), there exists adjoint variablesλ1,⋯,λ7andλ8satisfying:λ1′=λ1(μ+β1(1−u(t))A(t)N+β2(1−u(t))I(t)N−λ2(β1(1−u(t))A(t)N+β2(1−u(t))I(t)N)
λ2′=−λ2(−(μ+α1+α2))−λ3α1−λ4α2
λ3′=λ1β1(1−u(t))S(t)N−λ2β1(1−u(t))S(t)N+λ3(γ1+θ1+χ+μ+v(t))−λ4χ−λ7(θ1+v(t))
λ4′=−1+λ1β2(1−u(t))S(t)N−λ2β2(1−u(t))S(t)N+λ4(γ+θ2+μ+v(t))−λ5(1−ϱ)γ−λ6ϱγ−λ7(θ2+v(t))
λ5′=−1+λ5(η1+μ+δ1+v(t))−λ7(η1+v(t))
λ6′=−1+λ6(η2+μ+δ2+v(t))−λ7(η2+v(t))
λ7′=λ7(μ+σ+δ3)−λ8σ
λ8′=λ8μ
With the transversality conditions at timeT:λ1(T)=0,λ2(T)=0,λ3(T)=0,λ4(T)=−1,λ5(T)=−1,λ6(T)=−1,λ7(T)=0andλ8(T)=0.
Furthermore, fort∈[0,T], the optimal controlsu*andv*are given by(15) u*=min(1,max(0,(λ2−λ1)A(β1S(t)A(t)N+β2S(t)I(t)N)))
(16) v*=min(1,max(0,(λ4I(t)+λ5C(t)+λ6CW(t)−λ5(I(t)+C(t)+CW(t))B))
Proof The Hamiltonian H∧ is defined as follows:
H∧(t)=I(t)+C(t)+CW(t)+A12u2(t)+A22v2(t)+∑i=18λi(t)fi(S,E,A,I,C,CW,H,R) where :
f1(S,E,A,I,C,CW,H,R)=Λ−μS(t)−β1(1−u(t))S(t)A(t)N−β2(1−u(t))S(t)I(t)N
f2(S,E,A,I,C,CW,H,R))=β1(1−u(t))S(t)A(t)N+β2(1−u(t))S(t)I(t)N
−(μ+α1+α2)E(t)
f3(S,E,A,I,C,CW,H,R)=α1E(t)−(χ+μ)A(t)
f4(S,E,A,I,C,CW,H,R)=α2E(t)+χA(t)−(γ+θ2+μ)I(t)−v(t)I(t)
f5(S,E,A,I,C,CW,H,R)=(1−ϱ)γI(t)−(η1+μ+δ1)C(t)−v(t)C(t)
f6(S,E,A,I,C,CW,H,R)=ϱγI(t)−(η2+μ+δ2)CW(t)−v(t)CW(t)
f7(S,E,A,I,C,CW,H,R)=θ1I(t)+θ2I(t)+η1C(t)+η2CW(t)−(μ+σ+δ3)H(t)+v(I(t)+C(t)+CW(t))
f8(S,E,A,I,C,CW,H,R)=σH(t)−μR(t)
For t∈[0,T], the adjoint equations and transversality conditions can be obtained by using Pontryagin’s maximum principle[5], [10], [11] such thatλ1′=−∂H∧∂S=λ1(μ+β1(1−u(t))A(t)N+β2(1−u(t))I(t)N)−λ2(β1(1−u(t))A(t)N+β2(1−u(t))I(t)N)
λ2′=−∂H∧∂E=−λ2(−(μ+α1+α2))−λ3α1−λ4α2
λ3′=−∂H∧∂A=−1+λ1β1(1−u(t))S(t)N−λ2β1(1−u(t))S(t)N+λ3(θ1+χ+μ+v(t))−λ4χ−λ7(θ1)
λ4′=−∂H∧∂I=−1+λ1β2(1−u(t))S(t)N−λ2β2(1−u(t))S(t)N+λ4(γ+θ2+μ+v(t))−λ5(1−ϱ)γ−λ6ϱγ−λ7(θ2+v(t))
λ5′=−∂H∧∂C=−1+λ5(η1+μ+δ1+v(t))−λ7(η1+v(t))
λ6′=−∂H∧∂CW=−1+λ6(η2+μ+δ2+v(t))−λ7(η2+v(t))
λ7′=−∂H∧∂H=λ7(μ+σ+δ3)−λ8σ
λ8′=−∂H∧∂R=λ8μ
With the transversality conditions at time T : λ1(T)=0,λ2(T)=0, λ3(T)=0,λ4(T)=−1, λ5(T)=−1,λ6(T)=−1, λ7(T)=0 and λ8(T)=0.
For t∈[0,T], the optimal controls u*and v* can be solved from the optimality condition,
We have−∂H∧∂u=−A1u(t)+(λ2−λ1)(β1S(t)A(t)N+β2S(t)I(t)N)=0−∂H∧∂v=−A2v(t)λ4I(t)+λ5C(t)+λ6CW(t)−λ7(I(t)+C(t)+CW(t))=0
we aveu(t)=(λ2−λ1)A(β1S(t)A(t)N+β2S(t)I(t)N)
v(t)=(λ4I(t)+λ5C(t)+λ6CW(t)−λ7(I(t)+C(t)+CW(t))B
By the bounds in U of the controls, it is easy to obtain u* and v* are given in ((15), (16),) the form of system (10). □
4 Numerical simulation
In this section, we present the results obtained by numerically solving the optimality system. In our control problem, we have initial conditions for the state variables and terminal conditions for the adjoints. That is, the optimality system is a two-point boundary value problem with separated boundary conditions at times step i=0 and i=T. We solve the optimality system by an iterative method with forward solving of the state system followed by backward solving of the adjoint system. We start with an initial guess for the controls at the first iteration and then before the next iteration, we update the controls by using the characterization. We continue until convergence of successive iterates is achieved. A code is written and compiled in Matlab using the following data.
Different simulations can be carried out using various values of parameters. In the present numerical approach:
Since control and state functions are on different scales, the weight constant value is chosen as follows: A1=10000, A2=10000,
Based on Tables 1 and 2 , and from Figure 3 . We note that when no proactive precaution is applied, the number of people infected with the COVID-19 virus will rise very significantly and will exceed millions in Peru, and may even reach a billion infected people IN Peru, which will result in an increase in the number of deaths in Peru due to the COVID-19 pandemic.Table 1 Parameter values used in numerical simulation
Table 1Paramter Description value in d−1
Λ Denote the incidence of susceptible. 13.106
β1 The rate of people who were infected by contact with infected and asympt 0.25
β2 The rate of people who were infected by contact with infected and sympt 0.2
α1 The rate of exposed become infected and asymptomatic 0.25
α2 The rate of exposed become infected and symptomatic 0.45
θ1 The number of infected and asymptomatic who have been under lockdown 0.1
θ2 The number of infected and symptomatic who have been under lockdown 0.2
δ1 Mortality rate due to serious complications 0.1
δ2 mortality rate due to serious complications and with chronic diseases 0.4
δ3 The rate of people who died under quarantine in hospitals. 0.1
μ Natural mortality 0.02
η1 The rate of infected with serious compl without chronic diseasesunder lockdown 0.08
η2 The rate of infected with serious compl with chronic diseases under lockdown 0.08
σ The rate of people who recovered from the virus and decreasing 0.08
γ The rate of inf become with serious compl and those without and with chronic disease 0.2
χ The number of infected and asympt become with infected and sympt 0.1
Table 2 Population values used in numerical simulation
Table 2Popoluation Description Value
S(0) The number of population in Peru 3.31×108
E(0) The number of exposed indivduals 15×106
A(0) The number of infected and asymptomatics people 6×106
I(0) The number of infected and symptomatics peoples 5,89×106
C(0) The number of Infected with serious complications and those without chronic diseases 10000
CW(0) The number of Infected with serious complications and those with chronic diseases 3×106
H(0) The number of Infected who have been under lockdown in hospitals 3×106
R(0) The number of Recovered people 3,17×106
* After last statistics of WHO on Friday September 04, 2020, 12:44 GMT
Table 3 Total costs and total averted infections for strategies 1–3
Table 3Strategy Total averted infections (TA) Total cost (TC)
3 4.85×106 3.05×106
2 4.82×106 1.57×106
1 4.52×106 5.92×105
Fig. 3 The evolution of the infected without controls
Fig. 3
The proposed control strategy in this work helps to achieve several objectives.
4.1 Strategy A: Sensitization and prevention
We use only the optimal control u(t).
This strategy aims to increase the number of people protected from COVID-19 in Peru. Through Figure 4 , we note a decrease in the number of exposed, which led to a decrease in the number of infecteds, which will lead to a decrease in infected with serious complications in Peru. And after implementing this strategy aimed at implementing awareness-raising campaigns for all American citizens, to make them aware of the seriousness of the COVID-19 virus, through the media and others, and to take preventive measures to avoid the wounded and the hand-wash regularly, especially after sneezing, wear face masks, which reduces the spread of the COVID-19 pandemic in Peru.Fig. 4 The evolution of the number of infected and those with complication with control u(t)
Fig. 4
Note: To date there is no vaccine or treatment for this disease.
4.2 Strategy B: Quarantine
We use only the optimal control v(t)
According to Figure 5 , we notice a decrease in the number of infected in Peru, which will lead to a decrease in the number of infected with complications of both types. This is due to the adoption of a quarantine strategy for all American people who suffer from symptoms and complications within hospitals and psychological follow-up in addition to quarantine, in addition to preventing entry and exit to and from areas of COVID-19. The diagnostic and monitoring strategy has been adopted, and this appears in the diagnosis of all cases of American people with the potential virus, especially the family, relatives and neighbors of the person who was diagnosed, as well as preventive precautions such as monitoring all means of transportation such as the metro, trains, airports. The main objective of this procedure is to reduce the spread of the disease in Peru. This explains the primary role of quarantine and treatment to reduce the spread of the Covid-19 pandemicFig. 5 Evolution of the number of people infected without and with complications with control v(t) in Peru
Fig. 5
4.3 Strategy C: Quarantine and awarness program
we combine the optimal control u(t) and v(t).
In this strategy, the two optimal controls u(t) and v(t) are activated at the same time in order to improve the numerical results of case 1.
Figure 6: Evolution of the number of people infected without and with complications with controls u(t) and v(t)
Starting from Figure 6 , we notice a significant decrease in the number of people infected with the two types of S and L. The number of people infected with the virus also decreased with complications, whether ordinary or with chronic disease. This is due to the implementation of strict precautionary measures such as quarantine with treatment, and the strategy relies on awareness campaigns in Peru.Fig. 6 Evolution of the number of people infected without and with complications with control v(t)
Fig. 6
4.4 Cost-effectiveness analysis
In this section, we analyze the profitability of the previous three scenarios and strategies by comparing these two control strategies to determine the most profitable strategy. Following the method applied in several studies [8], we assess costs using the differential cost-effectiveness (ICER). This ratio used compare the differences between costs and health outcomes of two competing intervention strategies.
The ICER is defined as the quotient of the difference in costs in strategies i and j, by the difference in infected averted in strategies i and j (i,j∈{1,2,3}).
Given two competing strategies (1) and (2), where strategy (2) has higher effectiveness than strategy (1) (TA(2)>TA(1)), the ICER values are calculated as follow:ICER(1)=TC(1)TA(1)
ICER(2)=TC(2)−TC(1)TA(2)−TA(1)
Where the total costs (TC) and the total cases averted (TA) are defined, during a given period for strategy i for i=1,2,3 by:TC(i)=∫0T(β1u(t)S(t)A(t)N+β2u(t)S(t)I(t)N+v(t)(I(t)+C(t)+CW(t)))dt
TA(i)=∫0T(I(t)+C(t)+CW(t)))−(I*(t)+C*(t)+CW*(t)))dt
while (I*(t)+C*(t)+CW*(t)) is the optimal solution associated to the optimal control (u*,v*). Using the simulation results and we ranked, in the Table 2 our control strategies in order of increased numbers of averted infections.
Strategy I is compared with strategy J with respect to increased effectiveness, in reference to Table 2. So:ICER(2)=TC(2)TA(2)=1.57×1064.82×106=0.032
ICER(3)=TC(3)−TC(2)TA(3)−TA(2)=3.05×106−1.57×1064.85×106−4.82×106=4.93
Since ICER(2)<ICER(3), then strategy 3 is less effective than strategy 1. Therefore, strategy 2 is excluded from the set of alternatives.
Next, strategy 2 is compared to strategy 1. The ICER values for strategy 2 and strategy 1 are calculated below:ICER(2)=TC(2)TA(2)=1.57×1064.82×106=0.032
ICER(1)=TC(1)−TC(2)TA(1)−TA(2)=5.92×105−1.57×1064.52×106−4.82×106=32.6
Since ICER(2)<ICER(1), then strategy 1 is less effective than strategy 2. Therefore, strategy 2 is excluded from the set of alternatives.
Therefore, the conclusion is that strategy 1 (awarness campaigns to protect potential individuals infected with the virus, prevent contact with people infected with COVID-19 and with hospital quarantine for the infected) is the most effective strategy as previously mentioned by the proportions.
5 Conclusion
In this paper, we introduced a mathematical model of transmission of COVID-19 virus in Peru. in order to minimize the number of infected and infected with serious complications and number of people in guarantine. We also introduced two controls which, respectively, represent sensitization, prevention, treatment and psychological support with follow-up. We also studied the optimal control. that if preventive and proactive measures are implemented, such as awareness-raising and quarantine campaigns in all overe Peru, the spread of the COVID-19 epidemic will be reduced, thus the number of people infected with the virus and the number of deaths will be reduced. We applied the results of the control theory and we managed to obtain the characterizations of the optimal controls. The numerical simulation of the obtained results showed the effectiveness of the proposed control strategies. In future we will study the spread of this disease in time and place, and with ABC fractional differential.
Uncited References
[9], [14], [19], [23], [24], [25], [28], [29]
Declaration of Competing Interest
All authors declare no conflict of interest.
Data Availability
Data will be made available on request.
==== Refs
References
1 World Health Organization.who.int/csr/don/12-january-2020-novel-coronavirus-china.
2 World Health Organization.who.int/emergencies/diseases/novel-coronavirus-2019.
3 World Health Organization.who.int/health-topics/coronavirus/laboratory-diagnostics-for-novel-coronavirus.
4 Middle east respiratory syndrome coronavirus (MERS-cov) – the kingdom of saudi arabia, 18 december. 2019. Who.int/csr/don/05-december-2019-mers-saudi-arabia/.
5 Middle east respiratory syndrome coronavirus (MERS-cov) summary of current situation 21 july. 2017.
6 Middle east respiratory syndrome coronavirus (MERS-cov) – qatar, 26 december. 2019. www.who.int/csr/don/26-december-2019-mers-qatar.
7 Ndairou F. Area I. Mathematical modeling of COVID-19 transmission dynamics with a case study of wuhan Chaos, Solitons and Fractals 135 2020 109846 10.1016/j.chaos.2020.109846 32341628
8 Fanelli D. Piazza F. Analysis and forecast of COVID-19 spreading in china, italy and france Chaos, Solitons and Fractals 134 2020 109761 10.1016/j.chaos.2020.109761 32308258
9 Khajji B. Kada D. Balatif O. A multi-region discrete time mathematical modeling of the dynamics of covid-19 virus propagation using optimal control J Appl Math Comput 2020 10.1007/s12190-020-01354-3
10 Pontryagin L.S. Boltyanskii V.G. Gamkrelidze R.V. Mishchenko E.F. The Mathematical Theory of Optimal Processes 1962 Wiley New York, NY, Peru
11 Fleming W.H. Rishel R.W. Deterministic and Stochastic Optimal Control 1975 Springer New York, NY, Peru
12 Novel coronavirus (2019-ncov). https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200315-sitrep-55-COVID-19.pdf.sfvrsn=33daa5cb_8.
13 Novel coronavirus (2019-ncov) technical guidance: Early investigations. Who.int/emergencies/diseases/novel-coronavirus-2019/technical-guidance/early-investigations.
14 Boyce W.E. DiPrima R.C. Elementary Differential Equations and Boundary Value Problems 2009 John Wiley & Sons New York
15 Wang H. Wang Z. Dong Y. Phase-adjusted estimation of the number of coronavirus disease 2019 cases in wuhan, china Cell Discov 6 2020 10 10.1038/s41421-020-0148-0 32133152
16 Al-Asuoad N. Rong L. Alaswad S. Shillo M. Mathematical model and simulations of MERS outbreak: Predictions and implications for control measures 5 2 2016
17 Birkhoff G. Rota G.C. Ordinary Differential Equations, 4th edition 1989 JohnWiley & Sons New York
18 The food and agriculture organization of the united nations (FAO). Coronavirus disease (COVID-19) outbreak, http://www.fao.org/2019-ncov/en/.
19 https://infographics.channelnewsasia.com/COVID-19/map.html.
20 World health organization. SARS Virus. https://www.who.int/ith/diseases/sars/en/.
21 Choi S.C., Ki M.. Estimating the reproductive number and the outbreak size of novel coronavirus disease (COVID-19) using mathematical model in republic of korea. PMID32164053. 10.4178/epih.e2020011
22 Rodriguez-Morales A.J. Bonilla-Aldana D.K. Tiwari R. Sah R. Rabaan A.A. Dhama K. COVID-19, an emerging coronavirus infection: Current scenario and recent developments – an overview J Pure Appl Microbiol 14 1 2020
23 Tang X., Wu C., Li X., Song Y., Yao X., Wu X., Duan Y., Zhang H., Wang Y., Qian Z., Cui J., Lu J.. On the origin and continuing evolution of SARS-cov-2. National Science ReviewNwaa036. 10.1093/nsr/nwaa036
24 World health organization. https://www.who.int/emergencies/diseases/novel-coronavirus-2019(03/04/2020).
25 Xia S. Liu M. Wang C. Inhibition of SARS-cov-2 (previously 2019-ncov) infection by a highly potent pan-coronavirus fusion inhibitor targeting its spike protein that harbors a high capacity to mediate membrane fusion Cell Res 30 2020 343 355 10.1038/s41422-020-0305-x 32231345
26 Guo Y. Cao Q. Hong Z. The origin, transmission and clinical therapies on coronavirus disease 2019 (COVID-19) outbreak - an update on the status Military Med Res 7 2020 11 10.1186/s40779-020-00240-0
27 Kouidere A. Khajji B. Bhih A.E. Balatif O. Rachik M. A mathematical modeling with optimal control strategy of transmission of COVID-19 pandemic virus Commun Math Biol Neurosci 2020 2020
Article ID 24
28 https://www.gob.pe/institucion/minsa/noticias/299532-minsa-casos-confirmados-por-coronavirus-covid-19-ascienden-a-670-145-en-el-peru-comunicado-n-228.
29 Kouidere A. Kada D. Balatif O. Rachik M. Naim M. Optimal control approach of a mathematical modeling with multiple delays of the negative impact of delays in applying preventive precautions against the spread of the COVID-19 pandemic with a case study of brazil and cost-effectiveness Chaos, Solitons and Fractals 2020 110438 33519112
30 Kouidere A. Balatif O. Rachik M. Analysis and optimal control of a mathematical modeling of the spread of african swine fever virus with a case study of south korea and cost-effectiveness Chaos, Solitons and Fractals 146 2021 110867
May 1
31 Kouidere A. Youssoufi L.E. Ferjouchia H. Balatif O. Rachik M. Optimal control of mathematical modeling of the spread of the COVID-19 pandemic with highlighting the negative impact of quarantine on diabetics people with cost-effectiveness Chaos, Solitons and Fractals 145 2021 110777 33613000
Apr 1
32 Baleanu D. Abadi M.H. Jajarmi A. Vahid K.Z. Nieto J.J. A new comparative study on the general fractional model of COVID-19 with isolation and quarantine effects Alexandria Engineering Journal 61 6 2022 4779 4791
| 0 | PMC9741557 | NO-CC CODE | 2022-12-14 23:29:55 | no | 2022 Dec 11;:100090 | utf-8 | null | null | null | oa_other |
==== Front
J Parasit Dis
J Parasit Dis
Journal of Parasitic Diseases: Official Organ of the Indian Society for Parasitology
0971-7196
0975-0703
Springer India New Delhi
1556
10.1007/s12639-022-01556-5
Original Article
Toxoplasmosis and symptoms severity in patients with COVID-19 in referral centers in Northern Iran
Geraili Ali 1
Badirzadeh Alireza 1
Sadeghi Maryam 1
Mousavi Seyed Mahmoud 2
Mousavi Parisa 3
Shahmoradi Zabihollah 34
Hosseini Sayed-Mohsen 5
http://orcid.org/0000-0002-3733-1220
Hejazi Seyed Hossein [email protected]
23
http://orcid.org/0000-0001-5489-823X
Rafiei-Sefiddashti Raheleh [email protected]
1
1 grid.411746.1 0000 0004 4911 7066 Department of Parasitology and Mycology, School of Medicine, Iran University of Medical Sciences, Tehran, Iran
2 grid.411036.1 0000 0001 1498 685X Department of Parasitology and Mycology, School of Medicine, Isfahan University of Medical Sciences, Isfahan, Iran
3 grid.411036.1 0000 0001 1498 685X Skin Diseases and Leishmaniasis Research Center, Department of Parasitology and Mycolog, Isfahan University of Medical Sciences, Isfahan, Iran
4 grid.411036.1 0000 0001 1498 685X Department of Dermatology, Al-Zahra Hospital, Isfahan University of Medical Sciences, Isfahan, Iran
5 grid.411036.1 0000 0001 1498 685X Department of Epidemiology and Biostatistics, School of Public Health, Isfahan University of Medical Sciences, Isfahan, Iran
11 12 2022
17
4 9 2022
23 11 2022
© Indian Society for Parasitology 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.
Toxoplasmosis has been categorized as one of the long-lasting protozoan parasitic infections. It affects almost one-third of the world’s population. In recent years, several documented studies have elucidated that infected individuals have a remarkably higher incidence of distinct health problems and show various adverse effects. In the PCR-positive COVID-19 patients in Gonbad-e-Kavus, Kalaleh, and Minoodasht counties in the northern part of Iran from June 2021 to December 2021, we sought to investigate any potential relationships between the severity of COVID-19 symptoms and acute and latent toxoplasmosis caused by Toxoplasma gondii (T. gondii). Whole blood samples of 161 COVID-19 patients with positive PCR. The samples were centrifuged to separate serum and screened for two important antibodies against T. gondii (IgM and IgG) by using ELISA kits for human anti-T. gondii IgM and IgG. Anti-T. gondii IgM and IgG antibodies were detected in 8/161 (5.0%) and 42/161 (26.1%) COVID-19 patients, respectively. No significant relationships were found between Toxoplasma IgM and IgG results with clinical signs, age, sex, contact with animals, comorbidities, and also the mortality rate of people with COVID-19. These findings showed that acute and latent toxoplasmosis infections are common among patients with COVID-19; however, no significant associations were found between toxoplasma infections and the symptoms of COVID-19. Therefore, toxoplasmosis is not considered a risk factor for COVID-19.
Keywords
Toxoplasmosis
Toxoplasma gondii, COVID-19
Symptoms
Cat
Golestan province
Iran
http://dx.doi.org/10.13039/501100003970 Isfahan University of Medical Sciences 53462 Hejazi Seyed Hossein
==== Body
pmcIntroduction
Toxoplasma gondii (T. gondii), an obligate intracellular protozoan parasite that infects a wide range of species including birds and mammals but notably felids (cats), is the cause of the neglected parasitic illness toxoplasmosis (Zarean et al. 2017). Depending on social and cultural habits, geographical factors, climate and transmission routes, the Seroprevalence of toxoplasmosis varies widely between different countries (10–80%) (Zarean et al. 2017). On a global scale, T. gondii infection is thought to affect more than a billion people (Firouzeh et al. 2021). Ingesting cat feces-contaminated food or eating tissue cysts from raw or undercooked meat are the two major ways that people get sick, so that is the second most important food-borne parasitic disease in Europe (Firouzeh et al. 2021). In those with immune deficiencies or immune breakdown, this protozoan is regarded as an opportunistic and potentially fatal parasite (Sina et al. 2021). In addition, in immunocompromised patients it leads to severe clinical signs such as Toxoplasma encephalitis (TE), which can be fatal in severe cases (Sina et al. 2021) and also this disease is associated with several brain related disorders in both mothers and newborn (Fallahi et al. 2018) and furthermore, it should not be neglected as a cause of retinochoroiditis (Norouzi et al. 2016).
Severe acute respiratory syndrome may be caused by SARS-CoV-2. It is mostly spread by droplets and has caused major public health complications (Moriyama et al. 2020). The studies of the early accumulation of virus in the upper respiratory tract indicate that they attack and kill goblet cells and ciliated cells in the lungs. Then, these dead cells enter the lungs, the lungs gradually become blocked, and the person develops pneumonia (Moriyama et al. 2020).
SARS-CoV-2 lifecycle is similar to SARS-CoV as the virus binds to Angiotensin-converting enzyme 2 (ACE2) receptor on the surface of the host cell membrane via spike proteins (S-protein) and begins its infectious lifecycle (Senapati et al. 2021). Some people with coronavirus are asymptomatic, but in others, it can cause mild to severe pneumonia as well as fever, cough, anosmia, and ageusia. COVID-19 disease requires special care in the elderly, heart problems, hyperglycemia, and corticosteroid consumption people (Struyf et al. 2021), and they may have a high risk of death in these situations (Li et al. 2020).
Experimental evidence has shown that some parasitic infections can have immunogenic potential in preventing some diseases. In a retrospective case-control study, participants with a history of previous cutaneous leishmaniasis scars were significantly prevented from COVID-19 complications and mortality (Bamorovat et al. 2021). Coronavirus 2019 also has elucidated a statistically lower incidence in malaria-endemic areas due to the possible interaction via HLA-A ∗ 02: 01 and activation of CD8 + T-cell. The result of the common protection of the immunodominant epitope between SARS-CoV-2 (N and open reading frame (ORF) 1ab) and Plasmodium falciparum thrombospondin-related anonymous protein (TRAP) could provide a low incidence of COVID-19 in these areas (Iesa et al. 2020).
Some studies have shown that T. gondii demonstrates antiviral effects both directly and indirectly by causing the secretion of Dense Granule Protein-7 (GRA-7) in its host cells (Fekadu et al. 2010; Flegr 2013; Weeratunga et al. 2017). According to different studies, this intracellular parasite inhibited virus replication (Fekadu et al. 2010; Flegr 2013; Jankowiak et al. 2020a; Weeratunga et al. 2017). Both in vivo and in vitro studies have confirmed this evidence has been proven both in in vitro and in vivo experiments against herpes simplex virus, influenza A virus, coxsackie virus, and Indiana vesiculovirus (Jankowiak et al. 2020a) . Generally, GRA–7 has shown immune-stimulant effects and a wide range of antiviral effects through interferons signaling type I (Weeratunga et al. 2017). Moreover, apicoplast proteins are known to have immunogenic activities (Can et al. 2020). The immune response to Toxoplasma is a combination of TH1 cellular response and humoral immune response or specific anti-Toxoplasma antibodies (Munoz et al. 2011).
To date, an experiment has demonstrated a negative relationship between the incidence of coronavirus disease and distinct infections caused by parasites (Jankowiak et al. 2020a). It is noteworthy that COVID-19 and T. gondii infection can activate host innate immune responses via the same pathway. Indeed, in both intracellular pathogens, infected host toll-like receptors (TLR), such as TLR 2, TLR4, and TLR7, are strongly activated through the canonical or signaling pathway. Accordingly, some stimulated cytokines in toxoplasmosis patients may worsen the severity of coronavirus disease (Jankowiak et al. 2020a; Flegr 2013). Therefore, we hypothesized that Toxoplasma infection maybe has a possible relationship with COVID-19 in hospitalized patients. In this regard, the current experiment was mainly aimed to determine whether Toxoplasma infection has any positive/negative effects on the severity of COVID-19 symptoms or not. Hence, serological tests were used to detect T. gondii infection in 161 PCR-positive COVID-19 hospitalized individuals in the Gonbad-e-Kavus, Kalaleh, and Minoodasht counties in the northern part of Iran from Jun 2021 to December 2021.
Materials and methods
Ethics approval
The present study was conducted in accordance with the principles of the Helsinki Declaration. The ethical approval of this cross-sectional study was obtained from the Research Ethics Committees of Isfahan University of Medical Sciences, School of Medicine, Isfahan, Iran (IR.MUI.MED.REC.1400.049 to Dr. Seyed Hossein Hejazi). Before starting patient recruitment, the Institutional Review Board (IRB) of Isfahan University of Medical Sciences, Isfahan, Iran reviewed and approved our application for the current research projects. In the present study, adult participants and at least one parent or guardian of the child participants signed the written informed consent.
Study area and participants
From Jun 2021 to December 2021, we examined 161 PCR-positive COVID-19 individuals at referral hospitals in three countries, including Payambar-E Azam Hospital in Gonbad-e-Kavous county, Rasool Akram Hospital in Kalaleh county, and Fatemeh Zahra Hospital in Minoodasht county Hospitals, which located in the Golestan province in the northern part of Iran (Fig. 1).
Fig. 1 Map of Gonbad-e-Kavus, Kalaleh, and Minoodasht counties which are located in the Golestan Province in Northern Iran, showing the study area and sampling site from PCR-positive COVID-19 individuals in the referral centers (Created by Arc GIS version 10.2)
Sample collection
Having positive COVID19 PCR was the inclusion criteria in the current study. Blood specimens were taken from COVID-19 individual with positive PCR in Eppendorf tubes and, after 20 min, centrifuged for 4 min at 3000 rpm. The serum was then separated from the plasma and kept in the freezer at − 20 °C. Finally, all taken sera were submitted to the Department of Parasitology and Mycology, School of Medicine, Iran University of Medical Sciences, Tehran, Iran, for further examination.
Sociodemographic and clinical evaluation
By using a questionnaire checklist, sociodemographic data such as age, gender, clinical data including hospitalization or non-hospitalization due to COVID19, duration of symptoms, length of hospital stay, place of residence, history, different signs and symptoms, occupation, use of a ventilator, and type of treatment were documented. All extra data was taken from the patient’s physicians and documented records of medical examination.
Serological analysis
The serum samples were screened for two main anti-T. gondii antibodies, including IgM and IgG, by using a human Anti-T. gondii IgM and IgG by ELISA kit (Vircell, Granada, Spain), following the manufacturer’s protocol. For all samples, the optical density was determined at the wavelength of 450 nm, and both antibodies’ titers equal to or above 1.1 were considered positive. High IgM titer is an indication of active or acute toxoplasmosis, but increased IgG is used for showing latent toxoplasmosis. The tests had a sensitivity of 99.9% and specificity of 100%.
Statistical analysis
SPSS software (version 19; SPSS Inc., Chicago, IL, USA) was applied to the statistical analysis of the data. Kolmogorov–Smirnov (KS) test was used to examine the normal distribution of the quantitative variables. Frequency, median and interquartile ranges (IQR) were used to describe quantitative variables. The Chi-square and student’s t-tests were utilized to investigate the relationship between qualitative variables. Using Kruskal-Wallis test, the age variable was compared in all tests. P-values below 0.05 were considered statistically significant.
Results
In the present study, 161 PCR-positive COVID-19 patients were studied in the referral centers of Gonbad-e-Kavus, Kalaleh, and Minoodasht counties in the northern part of Iran (Fig. 1). Among the participants, 81 (50.3%) were female, and 80 (49.7%) were male. The mean age of participants was 50.0 (23%). Table 1 presents the rate of infection with T. gondii in COVID-19 patients. Complete sociodemographic, clinical data and serological analysis are shown in Tables 2 and 3. The analyzed data of the independent t-test elucidated that the mean age of participants was not significantly different among both men and women (P > 0.05). Among 161 PCR-positive COVID-19 individuals, 39 (24.2%), 41 (25.5%), and 60 (37.3%) had a fever above 38 °C, cough, and dyspnea, respectively, which showed the severe form of COVID-19. Our finding has shown no significant correlation between both gender and the severity of COVID-19 symptoms (P > 0.05) (Tables 2 and 3).
Table 1 The prevalence of Toxoplasma gondii infection in COVID-19 patients
Toxoplasma gondii IgM Toxoplasma gondii IgG
Positive Borderline Negative Positive Borderline Negative
N (%) 8 (5.0) 6 (3.7) 147 (91.3) 42 (26.1) 5 (3.1) 114 (70.8)
Table 2 Clinical features in COVID-19 positives and results of Toxoplasma IgM test
Variable All patients N (%) Toxoplasma IgM (Positive) N (%) Toxoplasma IgM (Borderline) N (%) Toxoplasma IgM (Negative) N (%) P value
Sociodemographic features Female 81 (50.3) 5 (62.5) 5 (83.3) 71 (48.3) 0.189
Male 80 (49.7) 3 (37.5) 1 (16.7) 76 (51.7)
Age [median (IQR)] 50.0 (23) 44.5 (26) 49.5 (24) 51.0 (23) 0.464
Contact with animals 44 (27.3) 44 (50.0) 3 (50.0) 37 (25.2) 0.138
No contact with animals 117 (72.7) 44 (50.0) 3 (50.0) 110 (74.8)
Clinical symptoms and signs Fever > 38 °C 39 (24.2) 1 (12.5) 2 (33.3) 36 (24.5) 0.645
Cough 41 (25.5) 3 (37.5) 3 (50.0) 35 (23.8) 0.256
Other 17 (10.6) 0 (0.0) 0 (0.0) 17 (11.6) 0.405
Shortness of breath 60 (37.3) 4 (50.0) 1 (16.7) 55 (37.4) 0.439
Non-specific symptoms Body pain 21 (13.0) 0 (0.0) 3 (50.0) 18 (12.2) 0.014
Headache 18 (11.2) 1 (12.5) 0 (0.0) 17 (11.6) 0.673
Trembling 13 (8.1) 0 (0.0) 0 (0.0) 13 (8.8) 0.510
Diarrhea 5 (3.1) 0 (0.0) 0 (0.0) 5 (3.4) 0.782
Comorbidities Hypertension 20 (12.4) 0 (0.0) 1 (16.7) 19 (12.9) 0.530
Chronic heart disease 25 (15.5) 0 (0.0) 1 (16.7) 24 (16.3) 0.641
Chronic pulmonary disease 13 (8.1) 1 (12.5) 1 (16.7) 11 (7.5) 0.645
No spleen and chronic hepatic disease 6 (3.7) 0 (0.0) 0 (0.0) 6 (4.1) 0.743
Diabetes 12 (7.5) 2 (25.0) 1 (16.7) 9 (6.1) 0.096
Patient condition Death 113 (70.6) 3 (37.5) 3 (50.0) 107 (73.3) 0.051
Recovery (alive) 47 (29.4) 5 (62.5) 3 (50.0) 39 (26.7)
Table 3 Clinical features in COVID-19 positives and results of Toxoplasma IgG test
Characteristic All patients
N (%) Toxoplasma IgG (Positive) N (%) Toxoplasma IgG (Borderline) N (%) Toxoplasma IgG (Negative) N (%) P value
Sociodemographic features Female 81 (50.3) 22 (52.4) 0 (0.0) 59 (51.8) 0.073
Male 80 (49.7) 20 (47.6) 5 (100.0) 55 (48.2)
Age [median (IQR)] 50.0 (23) 51.5 (22) 53.0 (19) 48.5 (24) 0.639
Contact with animals 44 (27.3) 40 (95.2) 1 (20.0) 3 (2.6) < 0.001
No contact with animals 117 (72.7) 2 (4.8) 4 (80.0) 111 (97.4)
Clinical symptoms and signs Fever > 38 °C 39 (24.2) 9 (21.4) 0 (0.0) 30 (26.3) 0.359
Cough 41 (25.5) 10 (23.8) 0 (0.0) 31 (27.2) 0.378
Other 17 (10.6) 3 (7.1) 2 (40.0) 12 (10.5) 0.078
Shortness of breath 60 (37.3) 20 (47.6) 3 (60.0) 37 (32.5) 0.125
Non-specific symptoms Body pain 21 (13.0) 7 (16.7) 0 (0.0) 14 (12.3) 0.523
Headache 18 (11.2) 3 (7.1) 0 (0.0) 15 (13.2) 0.413
Trembling 13 (8.1) 5 (11.9) 0 (0.0) 8 (7.0) 0.487
Diarrhea 5 (3.1) 1 (2.4) 0 (0.0) 4 (3.5) 0.863
Comorbidities Hypertension 20 (12.4) 6 (14.3) 0 (0.0) 14 (12.3) 0.655
Chronic heart disease 25 (15.5) 4 (9.5) 2 (40.0) 19 (16.7) 0.169
Chronic pulmonary disease 13 (8.1) 2 (4.8) 0 (0.0) 11 (9.6) 0.487
No spleen and chronic hepatic disease 6 (3.7) 1 (2.4) 0 (0.0) 5 (4.4) 0.762
Diabetes 12 (7.5) 4 (9.5) 1 (20.0) 7 (6.1) 0.430
Patient condition Death 113 (70.6) 25 (59.5) 4 (80.0) 84 (74.3) 0.178
Recovery (alive) 47 (29.4) 17 (40.5) 1 (20.0) 29 (25.4)
Anti-T. gondii IgM and IgG antibodies were identified in 8/161 (5.0%) and 42/161 (26.1%) COVID-19 patients, respectively. No significant relationships were found between Toxoplasma IgM and IgG results with clinical signs, age, sex, and comorbidities and also the mortality rate of due to COVID-19. However, in non-specific symptoms, the results showed that body pain was significantly different in individuals of different Toxoplasma IgM groups (P > 0.05) (Table 2). Accordingly, out of 147 people with Toxoplasma IgM negative test, 18 patients (12.2%) reported physical pain, which was 0% and 50% for those with a positive test and borderline test, respectively. The level of T. gondii IgG antibody was higher in COVID-19 patients in comparison with IgM antibody (Table 1). The results of the Toxoplasma IgG test revealed a significant relationship between animal contacts and Toxoplasma IgG test results (P > 0.05). Accordingly, 95.2% of those with a positive IgG test had a history of contact with animals, which was 20% and 2.6% in borderline individuals and those with a positive test, respectively (Table 3).
Discussion
Patients with COVID-19 who suffer from underlying diseases or receiving immunosuppressive therapy are at increased risk for opportunistic infections, in which prevention, screening, early diagnosis, and treatment is recommended (Abdoli et al. 2021; Mewara et al. 2021). According to the literature review, the association between toxoplasmosis and the severity of COVID-19 is not well-known. A study demonstrated a significant negative correlated variation between toxoplasmosis and COVID-19 (Jankowiak et al. 2020a).
We evaluated 161 PCR-positive COVID-19 individuals for acute and latent (chronic) toxoplasmosis by using anti-T. gondii antibodies (IgM and IgG). The current study demonstrated that 8/161 (5.0%) and 42/161 (26.1%) of the patients with COVID-19 were positive for anti-T. gondii antibodies (IgM and IgG). The study revealed a relatively high prevalence of latent toxoplasmosis in these patients; however, this high prevalence was not statistically significant Although in areas with high temperature, humidity and precipitation, the prevalence rates of toxoplasmosis are higher (Rostami et al. 2021). Several reports in the other parts of Iran showed reduced levels of T. gondii seroprevalence in different diseases (Ghaffari et al. 2021; Kalantari et al. 2018), which is in agreement with the current experiment.
In a study in Iran, the rate of latent IgG positive toxoplasmosis was reported at about 39% (Daryani et al. 2014). In another study in Golestan province, the rate of chronic Toxoplasmosis during pregnancy was reported to be 39.8% (Sharbatkhori et al. 2014). Nairi et al. in 2020 reported a global rate of antibodies against Toxoplasma of 33% and 43% in women who had recently had an abortion and those with previous abortion, respectively (Nayeri et al. 2020). Najm et al. in 2020 reported positive abortion between rubella virus infection and T. gondii (Najm et al. 2020). In Egypt, the rate of parasitic infections was 68.8%, in patients with COVID-19 and the most common parasite was T. gondii, with a prevalence of 22.4% (Abdel-Hamed et al. 2021).
An essential covariant factor in the toxoplasmosis epidemiology was recorded as age. Various experiments have revealed the increased rate of seroprevalence for T. gondii infection with aging, and the highest levels were reported above the age of 50 (Kalantari et al. 2018). In our study, the average age of COVID-19 patients was 50 years which is similar to the mentioned experiments. However, our study elucidated that the rate of seropositivity for both acute and latent Toxoplasma infection did not remarkably increase by age. Furthermore, there were no significant associations between the participants’ gender and the rate of seroprevalence regarding acute and latent Toxoplasma infection. These findings are not consistent with other documented studies that report a significant relationship between the seropositivity for T. gondii, gender, and age (Achaw et al. 2019, Ghaffari et al. 2021). This is possibly due to variations in the type and nature of the study population and its related regions (Ghaffari et al. 2021).
In the current study, we have confirmed no statistical relationship between COVID-19 outcomes and/or severity and acute/latent Toxoplasma infection, which has been demonstrated in preceding studies (Ghaffari et al. 2021; Montazeri et al. 2022). However, our results do not support those of Jankowiak et al. (2020a) which they found a significant negative correlated variation between toxoplasmosis and COVID-19.
In some viral infections such as AIDS, Toxoplasma latent tissue cysts containing bradyzoites show a higher risk of being reactivated by a decrease in CD4 + T cells and often present as Toxoplasma encephalitis. In these cases, reactivation is a possible scenario in patients with COVID-19 experiencing progressive lymphopenia but may not be considered due to the characteristics of Toxoplasma encephalitis, including neuropsychiatric symptoms including seizure, and altered sensorium, are also seen patients with COVID-19 (Roe 2021). In several studies, researchers discussed that various parasitic co-infections might be indicate the severity of COVID-19 (Ghaffari et al. 2021; Montazeri et al. 2022; Wolday et al. 2021). This may be due to host immune responses regarding T-helper 1 (TH1), which induce higher levels of IFN-ɣ and causes intense host tissue damage (Montazeri et al. 2022). Several animal co-infection in vivo experiments elucidated positive/negative impacts on the immunity against viral infections. In this regard, further studies should be conducted on COVID-19 severity and various parasitic co-infection is needed.
The main limitation of the present study could be related to the kind of study in which, due to the COVID-19 pandemic, we could not take samples from healthy individuals in the same period for a case-control study. It is recommended that ongoing experiments be carried out as multicenter studies in distinct geographic regions of the country to make more detailed health decisions about the relationship between acute and latent toxoplasmosis and COVID-19 severity. In addition, to precisely analyze the relationship between toxoplasmosis and the severity of COVID-19 symptoms, more comprehensive studies with higher sample sizes are required.
Conclusion
To our knowledge, the current study is a fundamental experiment that scrutinized the seroprevalence of active and latent toxoplasmosis in patients with COVID-19 and investigated their relationship with the severity of COVID-19. These findings showed that acute and latent toxoplasmosis infections are prevalent amongst COVID-19 patients; however, no significant and direct association was seen between toxoplasma infections and COVID-19 severity. Therefore, toxoplasmosis is not considered a risk factor for COVID-19.
Acknowledgements
We thank Dr. Zahra Asadgol (Health Deputy of Iran University of Medical Sciences, Tehran, Iran) for technical assistance in GIS map design.
Authors’ contribution
AG, AB: Formal analysis, investigation, conceptualization, methodology, validation, software. MS, SMM, PM: Investigation, methodology, software, data curation, writing- original draft preparation. ZS, SMH: Visualization, validation, writing - review and editing. RRS, SHH: Formal analysis, investigation, conceptualization, methodology, validation, software, resources, supervision, writing- original draft preparation, writing - review and editing.
Funding
Isfahan University of Medical Sciences funded the present study, with grant number 53462 to Dr. Seyed Hossein Hejazi. The funders were not involved in any steps of designing the study, collecting and analyzing data, publication decision, or manuscript preparation.
Data Availability
The corresponding author can provide the datasets collected during this study upon reasonable request.
Declarations
Conflict of interest
The authors declare no conflict of interests.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Seyed Hossein Hejazi and Raheleh Rafiei-Sefiddashti have contributed equally to this work.
==== Refs
References
Abdel-Hamed EF Ibrahim MN Mostafa NE Moawad HS Elgammal NE Darwiesh EM El-Rafey DS ElBadawy NE Al-Khoufi EA Hindawi SI Role of interferon gamma in SARS-CoV-2-positive patients with parasitic infections Gut Pathogens 2021 13 1 1 7 10.1186/s13099-021-00427-3 33436066
Abdoli A, Falahi S, Kenarkoohi A (2021) COVID-19-associated opportunistic infections: a snapshot on the current reports. Clin Exp Med, pp 1–20
Achaw B Tesfa H Zeleke AJ Worku L Addisu A Yigzaw N Tegegne Y Sero-prevalence of Toxoplasma gondii and associated risk factors among psychiatric outpatients attending University of Gondar Hospital, Northwest Ethiopia BMC Infect Dis 2019 19 1 1 8 10.1186/s12879-019-4234-6 30606108
Bamorovat M, Sharifi I, Aflatoonian MR, Karamoozian A, Tahmouresi A, Jafarzadeh A, Heshmatkhah A, Sharifi F, Salarkia E, Khaleghi T (2021) Prophylactic effect of cutaneous leishmaniasis against COVID-19: a case-control field assessment.International J Infect Dis
Can H Alak SE Köseoğlu AE Döşkaya M Ün C Do Toxoplasma gondii apicoplast proteins have antigenic potential? An in silico study Comput Biol Chem 2020 84 107158 10.1016/j.compbiolchem.2019.107158 31810853
Daryani A Sarvi S Aarabi M Mizani A Ahmadpour E Shokri A Rahimi M-T Sharif M Seroprevalence of Toxoplasma gondii in the iranian general population: a systematic review and meta-analysis Acta Trop 2014 137 185 194 10.1016/j.actatropica.2014.05.015 24887263
Fallahi S Rostami A Shiadeh MN Behniafar H Paktinat S An updated literature review on maternal-fetal and reproductive disorders of Toxoplasma gondii infection J Gynecol Obstet Hum Reprod 2018 47 3 133 140 10.1016/j.jogoh.2017.12.003 29229361
Fekadu A Shibre T Cleare AJ Toxoplasmosis as a cause for behaviour disorders-overview of evidence and mechanisms Folia Parasitol 2010 57 2 105 10.14411/fp.2010.013
Firouzeh N Ziaali N Sheibani V Doustimotlagh AH Afgar A Zamanpour M Keshavarz H Shojaee S Shafiei R Esmaeilpour K Chronic Toxoplasma gondii infection potentiates Parkinson’s disease course in mice model Iran J Parasitol 2021 16 4 527 35082880
Flegr J How and why Toxoplasma makes us crazy Trends Parasitol 2013 29 4 156 163 10.1016/j.pt.2013.01.007 23433494
Ghaffari S, Kalantari N, Gorgani-Firouzjaee T, Bayani M, Jalali F, Daroonkola MA (2021) Is COVID-19 associated with latent toxoplasmosis? Environ Sci Pollut Res, pp 1–5
Iesa MA Osman ME Hassan MA Dirar AI Abuzeid N Mancuso JJ Pandey R Mohammed AA Borad MJ Babiker HM SARS-CoV-2 and Plasmodium falciparum common immunodominant regions may explain low COVID-19 incidence in the malaria-endemic belt New Microb New Infect 2020 38 100817 10.1016/j.nmni.2020.100817
Jankowiak Ł Rozsa L Tryjanowski P Møller AP A negative covariation between toxoplasmosis and CoVID-19 with alternative interpretations Sci Rep 2020 10 1 1 7 10.1038/s41598-020-69351-x 31913322
Kalantari N Sheikhansari M Ghaffari S Alipour J Gorgani-Firouzjaee T Tamadoni A Bayani M Seroprevalence and molecular detection of Toxoplasma gondii in young healthy blood donors in Northern Iran Trop Biomed 2018 35 1017 1027 33601849
Li X Xu S Yu M Wang K Tao Y Zhou Y Shi J Zhou M Wu B Yang Z Zhang C Yue J Zhang Z Renz H Liu X Xie J Xie M Zhao J Risk factors for severity and mortality in adult COVID-19 inpatients in Wuhan J Allergy Clin Immunol 2020 146 1 110 118 10.1016/j.jaci.2020.04.006 32294485
Mewara A Sahni N Jain A Considering opportunistic parasitic infections in COVID-19 policies and recommendations Trans R Soc Trop Med Hyg 2021 115 11 1345 1347 10.1093/trstmh/trab142 34480170
Montazeri M, Nakhaei M, Fakhar M, Pazoki H, Pagheh AS, Nazar E, Zakariaei Z, Mirzaeian H, Sharifpour A, Banimostafavi ES (2022) Exploring the association between latent Toxoplasma gondii infection and COVID-19 in hospitalized patients: first registry-based study. Acta Parasitol, pp 1–8
Moriyama M, Hugentobler WJ, Iwasaki A (2020) Seasonality of respiratory viral infections. Annu Rev Virol 7(1), https://www.annualreviews.org/doi/pdf/10.1146/annurev-virology-012420-022445
Munoz M Liesenfeld O Heimesaat MM Immunology of Toxoplasma gondii Immunol Rev 2011 240 1 269 285 10.1111/j.1600-065X.2010.00992.x 21349099
Najm MA Razaak HAA Fahad H Detect the infection with Rubella virus and toxoplasmosis in pregnancy causes suffering from early abortion by using real time PCR Medico Legal Update 2020 20 1 397 402
Nayeri T, Sarvi S, Moosazadeh M, Amouei A, Hosseininejad Z, Daryani A (2020) The global seroprevalence of anti-Toxoplasma gondii antibodies in women who had spontaneous abortion: a systematic review and meta-analysis. PLoS Negl Trop Dis 14(3), e0008103
Norouzi M Tabaei SJS Niyyati M Saber V Behniafar H Genotyping of Toxoplasma gondii strains isolated from patients with ocular toxoplasmosis in Iran Iran J Parasitol 2016 11 3 316 28127336
Roe K The symptoms and clinical manifestations observed in COVID-19 Patients/Long COVID-19 symptoms that parallel Toxoplasma gondii infections J Neuroimmune Pharmacol 2021 16 3 513 516 10.1007/s11481-021-09997-0 34050501
Rostami A Riahi SM Esfandyari S Habibpour H Mollalo A Mirzapour A Behniafar H MohammadiMoghadam S Kyvanani NA Aghaei S Geo-climatic factors and prevalence of chronic toxoplasmosis in pregnant women: a meta-analysis and meta-regression Environ Pollut 2021 288 117790 10.1016/j.envpol.2021.117790 34329056
Senapati S Banerjee P Bhagavatula S Kushwaha PP Kumar S Contributions of human ACE2 and TMPRSS2 in determining host–pathogen interaction of COVID-19 J Genet 2021 100 1 1 16 10.1007/s12041-021-01262-w
Sharbatkhori M Moghaddam YD Pagheh AS Mohammadi R Mofidi HH Shojaee S Seroprevalence of Toxoplasma gondii infections in pregnant women in Gorgan city, Golestan province, Northern Iran-2012 Iran J Parasitol 2014 9 2 181 25848383
Sina S Mohammad JM Reza S Anita M Soudabeh E Hadi M Determination of parasitic burden in the brain tissue of infected mice in acute toxoplasmosis after treatment by fluconazole combined with sulfadiazine and pyrimethamine Eur J Med Res 2021 26 1 1 8 10.1186/s40001-021-00537-3 33388089
Struyf T, Deeks JJ, Dinnes J, Takwoingi Y, Davenport C, Leeflang MM, Spijker R, Hooft L, Emperador D, Domen J (2021) Signs and symptoms to determine if a patient presenting in primary care or hospital outpatient settings has COVID-19. Cochrane Database Syst Rev; 2(2):CD013665
Weeratunga P Herath TU Kim T-H Lee H-C Kim J-H Lee B-H Lee E-S Chathuranga K Chathuranga WG Yang C-S Dense granule Protein-7 (GRA-7) of Toxoplasma gondii inhibits viral replication in vitro and in vivo J Microbiol 2017 55 11 909 917 10.1007/s12275-017-7392-5 29076073
Wolday D Tasew G Amogne W UrbanUrban B Schallig HD Harris V de Wit TFR Interrogating the impact of intestinal parasite-microbiome on pathogenesis of COVID-19 in Sub-Saharan Africa Front Microbiol 2021 12 614522 10.3389/fmicb.2021.614522 33935986
Zarean M Shafiei R Gholami M Fata A Rahmati Balaghaleh M Karimi A Tehranian F Hasani A Akhavan A Seroprevalence of Anti–Toxoplasma gondii antibodies in healthy voluntary blood donors from Mashhad City, Iran Arch Iran Med 2017 20 7 441 445 28745905
| 0 | PMC9741704 | NO-CC CODE | 2022-12-14 23:29:57 | no | J Parasit Dis. 2022 Dec 11;:1-7 | utf-8 | J Parasit Dis | 2,022 | 10.1007/s12639-022-01556-5 | 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
36502476
24577
10.1007/s11356-022-24577-2
Research Article
A study on the response of carbon emission rights price to energy price macroeconomy and weather conditions
Shi Changfeng [email protected]
Zeng Qingshun [email protected]
Zhi Jiaqi [email protected]
Na Xiaohong [email protected]
Cheng Shufang [email protected]
grid.257065.3 0000 0004 1760 3465 Business School of Hohai University, Changzhou, 213022 China
Responsible Editor: Roula Inglesi-Lotz
11 12 2022
116
25 7 2022
30 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.
China’s carbon emission trading market has gradually attracted worldwide attention. In this paper, a structural VAR model and Shenzhen, a typical city in China, are selected to study the dynamic relationships between China’s carbon emission rights price, energy prices, macroeconomic level, and weather conditions. Shanghai crude oil futures, the first crude oil futures contract in China, is used to describe changes in oil market as a substitute for Daqing crude oil price. The results show that the price of carbon emission rights is mainly affected by its own historical price; and the price of carbon emission rights is positively correlated with crude oil price and natural gas price, but negatively correlated with coal price; the change of macroeconomic level will still have a relatively large impact on carbon emission rights price in the current stage of economic development in China, but this impact is not significant; The impact of weather conditions on the price of carbon emission rights is not obvious. It is found that the launch of the national unified carbon market has indeed achieved certain results, but the situation that China’s carbon market is still in its infancy has not been changed; further efforts are needed.
Keywords
China’s carbon emission rights price
Shanghai crude oil futures
Impulse response functions
Variance decomposition
SVAR model
==== Body
pmcIntroduction
In the past few decades, with the rapid economic development, carbon emission has gradually attracted worldwide attention (Alberola et al. 2008; Benz and Trück 2009; Li et al. 2014; Zeng et al. 2017). The Kyoto Protocol, an international agreement, set targets to limit carbon emissions from 37 industrialized nations and the European Union. In addition, different targets were set for the 37 industrialized countries in accordance with the principle of general but differentiated responsibilities, but not for countries with low economic level like China and India. Although China is not obliged to observe these conventions, it has spared no effort to minimize its own greenhouse gas emissions to slow down the deterioration of the global climate. China overtook the USA as the world’s biggest emitter of CO2 in 2006. In 2020, China’s carbon emissions were nearly 9.9 billion tons, accounting for about 30% of the total global carbon emissions (Lin and Xu 2021). In order to achieve the strategic goal of carbon peak and carbon neutrality in an orderly manner, the Chinese government has implemented a number of governance policies and measures to reduce CO2 emissions, including adjusting industrial structure, increasing investment in research and development of emission reduction technologies, and expanding the proportion of clean energy use. However, the implementation results of these measures are not ideal (Xu et al. 2021). The reason is that China’s energy consumption structure dominated by fossil energy such as coal, crude oil, and natural gas cannot change significantly in the short-term, so CO2 emissions cannot be significantly reduced (Wang et al. 2017). Establishing carbon emission trading market has been regarded as an effective way for major energy consuming countries in the world to control CO2 emissions (Wang et al. 2018), and this approach has been also proved applicable to China (Cui et al. 2014; Ji et al. 2021; Lin and Wesseh 2020; Tang et al. 2015).
Carbon emission pricing and trading is a key feature of the global regulatory response to climate change. The CO2 emissions (or equivalent greenhouse gas) of enterprises are regulated by governments, and different enterprises have different emission reduction capacity and emission demand, which makes carbon emission rights a tradable commodity. For demanders and suppliers in the carbon trading market, it is important to understand how the dynamics of the carbon emission rights price (carbon price) are affected by shocks caused by changes in weather conditions, energy prices, and economic activities with the aim of minimizing trading risk, and this study helps to explain this problem.
When China first tried to establish a carbon trading market, differentiated carbon allowance allocation policies were implemented for carbon trading pilots in different provinces, which provided evidence for operational comparisons and laid the foundation for establishing a unified national carbon trading market with the best carbon reduction effect (Nie et al. 2019; Shi et al. 2022). However, China’s carbon emission trading market started late. Compared with the international carbon emission trading market, China’s carbon emission trading market has major problems such as imperfect trading mechanism and unreasonable market pricing (Weng and Xu 2018). Some current studies have proved that the construction of China’s carbon trading market is helpful to reduce regional carbon emissions and achieve the expected effect of carbon emission constraints (Chang et al. 2019; Song et al. 2019), but few studies have analyzed the exact interaction between carbon price, energy prices, macroeconomic level, and weather conditions in China. This research gap hinders our understanding of the influencing factors of carbon price, prevents decision makers and investors from making reasonable judgments on the fluctuations of carbon price, and makes policy making and carbon asset investment full of risks and uncertainties.
In this study, the Shenzhen Carbon Emissions Trading Market, which has the most extensive carbon trading price record in China (Cong and Lo 2017) was selected as the research object. A SVAR model with impulse response function and variance decomposition technology was employed to analyze how the carbon emission rights price dynamics in Shenzhen carbon emission trading market was affected by the impact of shocks from energy prices, economic activities, and weather conditions. The main contribution of this study is that we introduce the price of China’s first crude oil futures contract into the influencing factor system. In addition to innovation of the selection of influencing variables of carbon price, we also analyze the hysteresis effect and variance contribution rate of each influencing factor on the fluctuation of carbon price from the perspective of time.
The rest of this article is arranged as follows: the “Literature review” section is a literature review, the “Methodology” section is a theoretical analysis and introduction to research methods, the “Results and discussion” includes a description of variable information and an analysis of empirical results, and finally comes the conclusion.
Literature review
With the development of the international carbon market and many countries becoming active participants in carbon trading, more and more studies have investigated the impact and determinants of carbon emission trading price. Some earlier studies modeled carbon prices in time series. For example, Paolella and Taschini (2008) conducted an econometric analysis on the return rate of spot price of emission allowance in European carbon emission rights trading market based on mixed-normal GARCH model. Seifert et al. (2008) developed a theoretical stochastic equilibrium model to reflect the stylized characteristics of the EU ETS, and analyzed the dynamics of CO2 spot prices affected by this. Daskalakis et al. (2009) established a pricing model to study the three main emission allowance markets under the EU ETS: Powernext, Nord Pool, and ECX, and found the jump and non-stationarity of the price of carbon emission rights. Chevallier (2010) modeled and analyzed the risk premium of spot and futures prices of CO2 allowance. These studies focus on the analysis of the change and fluctuation characteristics of carbon price, but pay little attention to the influencing factors of carbon price.
Subsequently, many scholars have carried out supplementary studies on the influencing factors of carbon price. Some studies have analyzed the influence mechanism of economic activities and macro policies on carbon price. For example, Guðbrandsdóttir and Haraldsson (2011) analyzed the impact of the UK energy market and global stock index on the price of carbon credits in the EU Emissions Trading System (EU ETS) and found that the relationship between EU quotas (EUAs) and the UK electricity market was not significant. Chevallier (2011a, b) used industrial production as a proxy for economic activities and proved that carbon price and industrial production changed in the same direction during economic upswing. Fan and Todorova (2017) used stock futures index as proxy economic factors and found that there was a common integration between carbon price and economic indicators. Jiménez-Rodríguez (2019) found the potential relationship between the stock market and the EU ETS by studying the causal relationship between the common factors calculated from the main European stock market indexes and the EU ETS prices. Tan et al. (2020) analyzed the directional connectivity in the “carbon-energy-finance” system through variance decomposition, and concluded that the carbon market was closely related to the stock and non-energy commodity markets, but not significantly related to the bond market. Zheng et al. (2021) took China’s carbon emission trading market as an example to study the long-term asymmetric impact of oil shocks on prices of carbon allowance. Brauneis et al. (2013) used firm-level data to prove that the uncertainty of long-term climate policy was the main factor driving the change of carbon price. Koch et al. (2014) studied whether the use of renewable energy policies and international credit directly led to the decline of EUA price during the economic recession from 2008 to 2013, and found that the changes of solar power generation and economic activities could explain the changes of EUA price to a certain extent. Tu and Mo (2017) explored the interaction between renewable energy policies and carbon prices by constructing a partial equilibrium model. Gugler et al. (2021) compared and analyzed the differential characteristics of carbon prices under different climate change policies in Germany and the United Kingdom, and their research conclusions confirmed the impact of macro policies on carbon prices.
However, the research on the influencing factors of carbon price is not limited to economic activities and macro policies. There are also many studies devoted to analyzing the impact of energy prices and weather conditions on carbon prices at the micro level (Tian and Yang 2020), because the production or use of energy will produce a large amount of carbon emissions, and weather conditions will indirectly affect carbon emissions by affecting the use of energy (Ike et al. 2020; Laha and Chakraborty 2017; Moz-Christofoletti and Pereda 2021). For example, Aatola et al. (2013) proved that there is no obvious causal relationship between carbon price and electricity price. Hammoudeh et al. (2014a) analyzed the impact of the prices of crude oil, natural gas, coal and electricity on the carbon emission allowance price in the United States, and found that the prices of crude oil, natural gas, and coal had a negative impact on the carbon price, while the impact of electricity price was not significant enough. Ji et al. (2018) found that the carbon market is integrated and cross-correlated with other energy markets. Wang and Guo (2018) reported significant spillover effects of natural gas on the carbon market. Zhao et al. (2018) and Tan et al. (2020) studied the predictive ability of crude oil, natural gas, and coal prices on European carbon prices, which further proved the impact of energy prices on carbon prices. Meier and Rehdanz (2010) found in their research that changes in weather conditions would indirectly affect carbon emissions by affecting the heating demand of British households, but they did not explicitly state that weather conditions would have an impact on carbon prices. In a follow-up study, using temperature as a proxy for weather conditions, Batten et al. (2021) demonstrated that unexpected changes in weather conditions can have an impact on carbon prices.
There are a wide range of factors affecting the price fluctuation of carbon emission trading. Some scholars have used VAR model and its improved form to explore the interaction between carbon emission trading and other factors. For example, Chevallier (2011b) used two-regime Markov-Switching VAR models to establish the relationship between carbon price return and industrial production, and found the lagged impact of macroeconomic activities on carbon price. Da Silva et al. (2016) used a VECM model and found that the EU ETS had a positive long-term effect on the aggregated power sector stock Market return in Phase II. Gong et al. (2021) used The TVP-VAR-SV Model and Impulse response function to analyze the intensity and direction of time-varying spillovers between the carbon market and the fossil energy market in the European Union. Lin and Chen (2019) employed VAR-DCC-GARCH and VAR-Bekk-Agarch models, finding that there was a significant time-varying correlation between carbon emission trading market, coal market, and stock market. Compared with VAR model, SVAR model better captures the instantaneous structural relationship between variables in the model system, and is able to introduce the structural relationship between economic variables and financial theory into VAR model. At present, many studies have applied SVAR model to the study of price influencing factors. For example, Liu et al. (2016) used the SVAR model to study the response of stock returns of Chinese listed companies to oil price shocks basing on the oil industry chain. Adedeji et al. (2021) used SVAR model to study the dynamic impact of COVID-19 on the decline of oil price. Ederington et al. (2021) used the SVAR model to prove that spot gasoline prices and real heating oil prices do not respond to structural oil supply shocks.
The existing studies of carbon price provide important experience for this paper, but they still have some limitations. Current research on the exact interactions between carbon prices, energy prices, macroeconomic level, and weather conditions in China is limited. In addition, for the crude oil price in China’s energy prices, previous studies tend to choose Daqing crude oil price to measure. However, we believe that Shanghai Crude oil Futures, China’s first crude oil futures contract officially listed in Shanghai International Exchange (INE) in March 2018, can better reflect the oil supply and demand relationship in China (Ji and Zhang 2019; Yang et al. 2020; Zhang et al. 2021). Using a SVAR model with impulse response function and variance decomposition techniques, as well as time series data from March 26, 2018, to February 28, 2022, we studied the influencing factors of carbon emission rights prices in China, including energy prices, macroeconomic level, and weather conditions. We also explained the extent to which these factors affect the price of carbon emission rights, which will provide a decision-making reference for national policymakers and carbon asset investors.
Methodology
Influencing mechanism analysis
Weather conditions, energy market, and economic activities are closely related to carbon price. Therefore, this paper mainly selects the influencing factors of carbon prices from the above three aspects.
The impact of energy prices on carbon price is direct. The industrial development is closely related to energy sector, especially heavy industry, which is highly dependent on fuels such as coal, oil, and natural gas, and is a major source of carbon emissions. The carbon emission coefficient of coal is much higher than that of oil and gas. In the absence of rigid demand for energy, when the price of coal rises and the prices of natural gas and oil remain unchanged, the consumption of coal will decrease, the consumption of natural gas and oil will increase, and then, the demand for carbon emission rights will decrease, eventually leading to a decline in price of carbon emission rights (Lin and Xu 2020). However, due to the fact that enterprises have rigid demands for a specific energy source (Fathollahzadeh and Tabares-Velasco 2021), different enterprises have different fossil energy consumption structures, and there are difficulties in equipment replacement or capital investment when shifting from the use of one energy source to the use of another energy source. When the price of clean fossil energy such as natural gas rises, some enterprises may reduce their production to reduce the use of clean fossil energy instead of directly using low-priced non-clean fossil energy, which will also reduce the demand for carbon emission rights and eventually lead to a decline in the price of carbon emission rights.
Changes in macroeconomic level also have a direct impact on carbon price (Creti et al. 2012). Chevallier (2009) believed that macroeconomic shocks affect the relative demand of commodities, which in turn affects prices. The national macroeconomic level will directly affect the income level and consumption ability of consumers, and then affect the relative demand of consumers for goods. In times of economic prosperity, when market demand rises, enterprises will expand production to bring market supply and market demand to a new equilibrium point, as a result of which pollutants and greenhouse gas emissions will also increase. Correspondingly, the government will intervene in the market for the consideration of sustainable economic development and set limits on the carbon emissions of enterprises to achieve emission reduction targets. Then, enterprises will trade carbon emission rights through the carbon trading market to obtain a higher emission allowance, and the price of carbon emission rights will rise accordingly. However, Chan (2020b) obtained the opposite conclusion using dynamic stochastic general equilibrium model.
Changes in weather conditions have an indirect effect on carbon prices. In the short-term, the abnormal fluctuation of energy prices is affected by the abnormal change of weather (Bredin and Muckley 2011). For example, in cold weather in winter, there is a demand for heating through electricity or other energy sources; In hot weather in summer, there is a demand for cooling through electricity or other energy sources (Zhu et al. 2019). These demands will lead to a seasonal peak in energy consumption, resulting in an increase in carbon emissions, followed by an increase in the demand for carbon emission rights, and ultimately affect the price of carbon emission rights (Fig. 1).Fig. 1 Graph of theoretical analysis
SVAR model
In this paper, structural vector auto regression (SVAR) model, impulse response function, and variance decomposition analysis are used to test the influencing factors of carbon price, including energy prices, economic factors, and weather conditions.
Sims (1980) first introduced VAR model to analyze the dynamic influence of multiple related economic indicators and disturbance terms on variable system. VAR model optimizes the auto regressive model which uses only one variable. It adopts the form of multi-equation simultaneous, and is not based on economic theory. In each equation of the model, regression is performed on the lag of all the endogenous variables to estimate the dynamic relationships between them under the circumstances of no prior constraints. However, VAR model cannot reflect the current variable relationship hidden in the structure of the error term, which makes this model lack the current value of the endogenous variable. Therefore, we employ a more advanced SVAR model that contains the current relationships between variables to analyze the impact of energy prices, macroeconomic development, and weather conditions on the price of carbon emission rights.
Economic constraints are applied to the SVAR model to mark structural formulas, and additional restrictions are imposed in order to determine the structure of the model. SVAR model requires short-term constraints: For models with n variables, n (n − 1)/2 constraints are required to determine the impact of structural shocks. The SVAR model with t endogenous variables and a lag order of i can be expressed as Eq. (1).1 Byt=β+∑i=1mAiyt-i+μt
where B matrix is a sparse matrix with main diagonal element of 1, β and μt are k-order constant matrix and error matrix respectively, and Ai is the coefficient matrix of lag term. Yt is the M × 1 vector containing M variables, and μt is the disturbance term, which allows the existence of contemporaneous correlation. If B is an invertible matrix, Eq. (1) can be simplified as Eq. (2).2 yt=β+∑i=1mB-1Aiyt-i+B-1μt
The B−1 matrix in Eq. (2) is a 7 × 7 structure coefficient matrix. SVAR model mainly bases on economic theory and mutual logical relationship among variables to determine the current relationship between these seven variables. εt=B-1μt, and it can be expressed as follows:3 εtεtTEMεtINDεtHSεtGASεtCOALεtOILεtSZA=1000000α21100000α31α3210000α41α42α431000α51α52α53α54100α61α62α63α64α6510α71α72α73α74α75α761μtTEMμtINDμtHSμtGASμtCOALμtOILμtSZA
Results and discussion
Information of data
In this paper, we use data of all variables on a daily basis to study the influencing factors of China’s carbon emission price. The data period is from March 26, 2018, to February 28, 2022, and a total of 840 data points were obtained after removing some invalid values. In this paper, the average values of two adjacent days were used to explain the missing data.
The specific description of relevant explanatory variables and explained variables in this paper are as follows: (1) the price of carbon emission rights. Shenzhen has the earliest carbon market with complete carbon price data. Therefore, this paper uses the daily closing price of Shenzhen carbon emission rights to measure the carbon price (yuan/ton). (2) Energy prices. The price of Shanghai crude oil futures (yuan/barrel) is used to replace Daqing oil price. To response to the unavailability of daily price of coal and natural gas, the coal price is replaced by the thermal coal futures settlement price (yuan/ton), and the natural gas price is replaced by the benchmark price of the LNG (yuan/ton). (3) Macroeconomic level. The macroeconomic level is measured by Shanghai Shenzhen 300 index-HS300 and China Securities Industrial Index-IND (points). (4) Weather conditions. Due to the lack of daily mean temperature data in Shenzhen, the daily mean temperature of Shenzhen is estimated by the average of daily maximum and minimum temperature (Celsius).The original data of all variables are from Wind database, and the specific variables are shown in Table 1.Table 1 Information of explained and explanatory variables
Category Variable Unit Data source
Explained variable Shenzhen’s carbon emission rights price SZA Yuan/ton Wind
Explanatory variables Energy prices OIL Yuan/barrel
COAL Yuan/ton
GAS Yuan/ton
Macroeconomic level HS300 Point
IND Point
Weather conditions TEM Celsius
Descriptive statistical analysis of data
Figure 2 shows the fluctuations of the price of carbon emission rights in Shenzhen from March 26, 2018, to February 28, 2022. Overall, from the beginning of 2018 to the end of 2021, the price of carbon emission rights showed a fluctuating decline. The situation shows that the development of clean energy and research on emission reduction technologies in China had achieved certain results from 2018 to 2021.Fig. 2 The closing prices of Shenzhen’s carbon emission rights
However, the low price of carbon emission rights should also attract the attention of the government and other relevant departments. The low price of carbon emission rights will discourage the enthusiasm of enterprises in production and emission reduction, which is not conducive to the long-term stable operation of the carbon market. Locally, the price of carbon emission rights fluctuated around 32 yuan/ton from March 2018 to September 2018, 20 yuan/ton from December 2018 to June 2019, and 10 yuan/ton from July 2019 to April 2020. One possible reason for this pattern is the positive progress made in China’s 13th Five-Year Plan for ecological and environmental protection and new breakthroughs in energy-saving and emission reduction technologies. From May 2021 to September 2020, the price of carbon emission rights rose to about 30 yuan/ton. Then, from October 2020 to February 2021, the market of carbon emission rights price fluctuated sharply, with a maximum difference of about 50 yuan. From March 2021 to June 2021, the price of carbon emission rights gradually returned to the normal level with little fluctuation. From July 2021 to early 2022, as the unified national carbon emission trading market officially opened online trading, the price of carbon emission rights gradually recovered to a reasonable range.
Figures 3, 4, and 5 show the development trend of various influencing factors of carbon emission rights price from 2018 to 2022. Crude oil, coal, and LNG prices continue to rise, coal prices are very close to LNG prices, and there was a price peak in the last quarter of 2021. The reason why the crude oil price curve is below the coal and LNG price curve is that the price of crude oil is calculated in yuan/barrel. HS300 and IND index increased significantly without drastic fluctuations during the period, indicating that China’s macroeconomic level showed a steady growth trend within the research range. Shenzhen, a typical city in southern China, has a relatively stable temperature change, with an average daily temperature range of only about 20 °C.Fig. 3 Development trend of influencing factors (OIL, COAL, GAS)
Fig. 4 Development trend of influencing factors (HS300, IND)
Fig. 5 Development trend of influencing factors (TEM)
Table 2 shows the descriptive statistical results for each time series. According to the Jarque–Bera test, the null hypothesis of normal distribution is rejected by all the time series at a significant level.Table 2 Descriptive statistical results
Statistic SZA Oil Coal Gas HS300 IND TEM
Mean 18.73213 418.6224 656.4635 656.1766 4225.808 3098.697 24.42143
Median 15.37000 440.0000 605.3000 605.4500 4002.948 2881.664 25.50000
Maximum 56.24000 609.7000 1835.600 1908.200 5625.923 4440.652 31.50000
Minimum 3.030000 202.3000 479.8000 481.2000 2964.842 2161.618 6.500000
Std. Dev. 11.15570 87.89805 154.8006 154.1636 690.4976 624.7964 4.853984
Skewness 0.666356 − 0.632411 3.579536 3.574961 0.068742 0.618746 − 0.699798
Kurtosis 2.477422 2.499057 21.41211 21.48788 1.744476 2.086094 2.787198
Jarque–Bera 71.72239 64.77516 13626.52 13752.31 55.83347 82.83133 70.14546
P-value 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
Test of model establishment
In order to ensure the validity of the model and avoid the “pseudo-regression phenomenon,” this paper first processed the logarithm of the time series data, and then tested the stability of the data.
Dickey-Fuller (DF) test is one typical method for testing the stability of time series. It is based on the first-order auto regressive AR (1), assuming that the random disturbance term is white noise and has no auto correlation. However, most of the time series in economic phenomena do not satisfy this assumption. When the random disturbance term has auto correlation or the time series are generated by the higher-order auto regressive process, DF test is obviously not applicable.
In this paper, the augmented Dickey-Fuller (ADF) test is used to test the stability of time series data to overcome the limitations of DF test.
The results of ADF test are shown in Table 3. The SZA, OIL, and COAL data series are stationary and they are time series of order 0, I(0). The original data series of HS300, IND, and TEM all have a unit root. Their 1st difference can pass the test and they are the time series of I(1). Therefore, we use the 1st difference sequence of HS300, IND, and TEM to construct the VAR model. The estimated results of the VAR model are shown in the appendix Table 7.Table 3 Results of the ADF unit root test
Variable Model 5% Statistics P-value Result
SZA CT − 3.415213 − 3.785677 0.0000 Stationary
OIL NCNT − 1.941201 − 0.318223 0.7773 Non-stationary
COAL CT − 3.415239 − 4.432398 0.0020 Stationary
GAS CT − 4.488563 − 4.488563 0.0016 Stationary
HS300 NCNT − 1.941201 0.230472 0.7529 Non-stationary
IND NCNT − 1.941201 0.562828 0.8377 Non-stationary
TEM NCNT − 1.941204 − 0.642864 0.4388 Non-stationary
DOIL CT − 3.415187 − 29.57162 0.0000 Stationary
DHS300 CT − 3.415187 − 29.42084 0.0000 Stationary
DIND CT − 3.415187 − 28.68824 0.0000 Stationary
DTEM CT − 3.415226 − 16.87310 0.0000 Stationary
CT, constant and trend; CNT, constant and no trend; NCNT, no constant and no trend
To determine the appropriate lag structure of VAR model, relevant tests are carried out in this paper. The specific lag order results are shown in Table 4. According to AIC and SC, the lag order of VAR model in this paper is determined to be 7. AR root test results of VAR model are shown in Fig. 6. The model is composed of 7 variables. AR root test shows that all characteristic roots are in the unit circle, indicating that the 7-variable VAR model has good stability. Therefore, these seven variables can be selected to construct the SVAR model.Table 4 Lag order selection
Lag order LR FPE AIC SC HQ
0 NA 2.03e + 17 59.71586 59.75602 59.73127
1 4256.997 1.22e + 15 54.59907 54.92037* 54.72234
2 283.7722 9.63e + 14 54.36636 54.96880 54.59751
3 201.3986 8.44e + 14 54.23366 55.11724 54.57268*
4 118.0328 8.19e + 14 54.20400 55.36872 54.65088
5 164.2409 7.49e + 14 54.11414 55.56000 54.66889
6 138.3796 7.06e + 14 54.05564 55.78264 54.71826
7 115.1560 6.86e + 14* 54.02565* 56.03379 54.79614
8 81.56819* 6.95e + 14 54.03825 56.32753 54.91661
*Indicates lag order selected by the criterion
LR, sequential modified LR test statistic (each test at 5% level)
FPE, final prediction error
AIC, Akaike information criterion
SC, Schwarz information criterion
HQ, Hannan-Quinn information criterion
Fig. 6 Graphic results of stability test
Granger Causality Tests are mainly used to test whether an endogenous variable can be treated as an exogenous variable. For each equation in the SVAR model, the lag term of each endogenous variable (excluding its own lag term) was output combined with significant χ2 (Wald) statistics, and the combined significant χ2 statistics of all lagged endogenous variables were listed in the last line (ALL).
The specific results were shown in Table 5. The observation results show that SZA, COAL, and GAS pass the test and reject the null hypothesis, indicating that these three variables are endogenous variables.Table 5 Results of the Granger Causality Test
Dependent variable Chi-sq df Prob.
SZA 22.42402 12 0.0330
DOIL 10.90157 12 0.5371
COAL 49.83702 12 0.0000
GAS 260.6710 12 0.0000
DHS300 3.462511 12 0.9913
DIND 8.317088 12 0.7599
DTEM 15.63936 12 0.2083
Estimated SVAR model
The B−1 matrix in the “Methodology” section represents the short-term relationship between variables, and its estimated results are shown in Table 6.Table 6 Results of structural shock from B−1
Dependent variable: Disturbance term
DOIL COAL GAS DHS300 DIND DTEM SZA
DOIL 1.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
COAL 0.072395 1.000000 0.000000 0.000000 0.000000 0.000000 0.000000
(0.1888)
GAS − 0.067853 − 0.183926 1.000000 0.000000 0.000000 0.000000 0.000000
(0.5806) (0.0171)
DHS300 − 0.013033 0.021517 − 0.723534 1.000000 0.000000 0.000000 0.000000
(0.6620) (0.2522) (0.0000)
DIND 0.368571 − 0.840377 − 0.267494 0.253693 1.000000 0.000000 0.000000
(0.1886) (0.0000) (0.2819) (0.4361)
DTEM − 0.063478 0.038280 0.076502 − 0.135789 − 0.682317 1.000000 0.000000
(0.6075) (0.6276) (0.4850) (0.3441) (0.0000)
SZA 0.004051 − 0.006809 0.006292 − 0.009530 0.000529 0.009929 1.000000
(0.6696) (0.2613) (0.4547) (0.3875) (0.8067) (0.7270)
P-values are given in parentheses
The effects of COAL on GAS, COAL on DIND, GAS on DHS300, and DIND on DTEM are significant at 5% probability level. From the perspective of energy prices, crude oil futures price and natural gas benchmark price are positively correlated with carbon emission trading price, and a 1% change in them will increase the carbon emission price by 0.41% and 0.63% on average, respectively. Coal, crude oil, and natural gas are to some extent substitutes for each other, and the carbon emission coefficient of coal is higher than that of crude oil and natural gas. When the price of crude oil and natural gas rises, manufacturing enterprises tend to use more (relatively cheap) coal for the sake of reducing production cost, which will increase the carbon emissions of enterprises, increase the demand for carbon emission rights, and eventually the market demand becomes greater than the market supply, leading to an increase in carbon price; there is a negative correlation between coal and carbon emission trading price, and a 1% change in coal price will lead to an average decrease of 0.68% in carbon emission trading price, which is consistent with the conclusion of previous studies (Hammoudeh et al. 2014b; Lovcha et al. 2022). Coal as a non-clean energy, the inverse relationship between its price and carbon price has been confirmed by a large number of literature, and our empirical result is consistent with them. However, the absolute value of the estimated results of the coal impact coefficient in this paper is much smaller than the existing literature on the EU carbon trading market, which may be due to the fact that China’s current energy consumption structure is still dominated by coal compared with the EU, and there is a certain rigidity in the consumption of coal by enterprises, which makes the impact of rising coal prices on carbon prices weakened (Duan et al. 2021). Both crude oil and natural gas have less absolute impact on the price of carbon rights than coal, which is closely related to the current situation in which most Chinese enterprises still use coal as the main power energy (Zeng et al. 2021). From the perspective of macroeconomic level, the 1% change of the HS300 index and China Securities Industrial Index make the price of carbon emission rights fall by 0.95% and 0.05%, respectively. This result indicates that the change of macroeconomic level will still have a relatively large impact on carbon price in the current stage of economic development in China, but this impact is not significant, which means that the sensitivity of carbon price to the change of macroeconomic level is gradually decreasing. In terms of weather conditions, Shenzhen is located in the southern part of China, where the temperature varies a little during the year, so it is also consistent with the actual situation that the weather conditions have little impact on the price of carbon emission rights.
Results of impulse response functions
This paper estimates the impulse response function of the SVAR model. Within a 95% confidence interval, a standard deviation shock is used to obtain the response of the variable. Figure 7 shows the response of Shenzhen carbon emission rights price (SZA) to Shanghai crude oil futures price shock, thermal coal futures settlement price shock, liquefied natural gas (LNG) benchmark price shock, HS300 index shock, China Securities Industrial Index shock, Shenzhen daily average temperature shock, and the shock from itself.Fig. 7 Outcome of impulse response of Shenzhen carbon prices
In Fig. 7, the vertical axis represents the response of Shenzhen carbon emission rights price (yuan/ton), and the lag phase of impact is represented by the horizontal axis, with an interval of 1 day. The solid line indicates the response degree of Shenzhen carbon emission rights price to the impact of crude oil price, coal price, LNG price, HS300 index, China Securities Industrial Index and its own price. The dotted line indicates the magnitude of the deviation between positive and negative responses, and it is twice the standard deviation. It can be seen from the results that the response of carbon emission rights prices is most affected by itself, while other variables have relatively little impact. The first figure in Fig. 7 shows the response of Shenzhen carbon emission rights price to a standard deviation shock of its own. A positive standard deviation change in the price of carbon emission rights in Shenzhen itself will result in an initial increase of about 6% for the SZA itself, a rapid decline to 3% after 2 days, a rebound to 4% on around day 5, and then decrease at a slower rate to 0 and remain unchanged after 20 days. The second figure shows that a standard shock of crude oil prices can cause carbon emission rights prices to grow negatively at first, but will grow to 0.2% after about 2 days, then fall below 0 on around day 7, and continue until day 50. This result shows that the response of carbon emission rights prices to the shock of crude oil prices is relatively rapid, and it takes only a week or so to return to stability. The third figure shows that a standard shock of coal prices will cause the price of carbon rights to rise to 0.2% on around day 2, and then slowly rise to − 0.2% after falling to − 0.5% on about day 20. Compared with crude oil, carbon emission rights prices have responded more violently to coal price shock, and it takes longer to return to stability. The fourth figure shows that a standard shock of LNG prices will initially lead to a negative growth of about − 0.4% in carbon emission rights, but this negative growth will end on about day 10 and then level off. Carbon emission rights prices response to LNG price shocks in a similar way to crude oil, but it takes longer to go back to normal. The fifth, sixth, and seventh figure show how carbon emission rights prices response to HS300, China Securities Industrial Index and daily average temperature, respectively. Among them, the response of the carbon emission rights price to the HS300 index is similar to that of the crude oil price, and the response to the China Securities Industrial Index is to increase to 0.3% on about day 2, then decline to − 0.6% quickly, and return to stable to 0 on about day 20, the response of carbon emission rights price to the daily average temperature shock is the least obvious, and the carbon emission rights price quickly return to stability on around day 5.
Results of variance decomposition
The objective of this paper is to analyze the influencing factors of the price of carbon emission rights in Shenzhen. Therefore, variance decomposition analysis is only performed on the price of carbon emission rights in Shenzhen, and the results are shown in Fig. 8. According to the graphical results, the influencing degree of each variable on the price of carbon emission rights in Shenzhen is explained. The contribution of each structural impacting factor to the explained variable can be determined, and then the importance of different structural shocks can also be assessed. In Fig. 8, the horizontal axis represents the lag phase, 1 day is a cycle. The vertical axis represents the contribution rate of crude oil price, coal price, LNG price, HS300 index, China Securities Industrial Index, daily average temperature, and the price itself to Shenzhen carbon emission rights price.Fig. 8 Outcome of variance decomposition of Shenzhen carbon prices
The variance contribution rate of carbon emission rights prices to itself has continued to decline from 100% initially to 94% in Phase 50. This result shows that the price of carbon emission rights in Shenzhen is mainly affected by its own historical price, but this impact is gradually weakening. The variance contribution rate of crude oil price to the price of carbon emission rights in Shenzhen increases rapidly for about 3 periods and then begins to grow at a stable rate after a mild decline, indicating that the impact of Shanghai crude oil futures contract prices on carbon emission rights prices becomes steady and rising after a short period of adjustment, which is consistent with the results of existing research on Shanghai crude oil futures contracts. After the variance contribution rate of coal price to the price of carbon emissions rights rising from the initial 0 to 5% in Phase 50, its rate of rise begins to slow down. Comparing the price of coal with the price of crude oil, although the influence of both on the price of carbon emission rights are rising, the price of crude oil is faster. The variance contribution rate of LNG price to the price of carbon emission rights begins to stabilize after rising to about 0.15% in Phase 5. As for the macroeconomic level, HS300 index and China Securities Industrial Index have basically the same pattern of variance contribution to the price of carbon emission rights, but the China Securities Industrial Index rises faster and eventually stabilizes at a higher level. The highest variance contribution rate of weather conditions to the price of carbon emission rights is only about 0.04%, which indicates that the impact of changes in weather conditions on the price of carbon emission rights in Shenzhen can be nearly ignored.
Variance decomposition analysis describes the dynamic characteristics of the model. The variance contribution rate of coal and crude oil price to carbon emission rights price are on the rise. Coal and crude oil are the main energy in our actual business and production activities, as a result of which they play important roles in determining the price of carbon emission rights in Shenzhen. Shanghai crude oil futures have just started, and the impact mechanism of its price fluctuation on the price of carbon emission rights need to be supplemented by follow-up research. According to the existing research results, the price of carbon emission rights in Shenzhen is mainly affected by its own historical price, with an influence of more than 90%, and the impact of other variables on the price of carbon emission rights in Shenzhen is relatively indistinctive.
Conclusion
In this paper, we investigate the factors that influence Shenzhen carbon emission rights price. Consistent with the results of most existing studies, the empirical results of this paper show that the price of carbon emission rights in Shenzhen is negatively correlated with the price of coal, and positively correlated with the price of crude oil and natural gas. Because coal is a non-clean energy and its carbon emission coefficient is much higher than that of crude oil and natural gas. When the price of coal rises, manufacturing enterprises will reduce the use of coal, thus reducing the emission of greenhouse gases and the demand for carbon emission rights. Finally, the price of carbon emission rights will fall accordingly. Crude oil, natural gas, and coal are substitutes for each other to some extent. When the prices of crude oil and natural gas rise, the use of coal will increase due to the substitution effect, which will increase the emission of greenhouse gases and expand the demand for carbon emission rights. Finally, the price of carbon emission rights will also have a rise. The change of macroeconomic level will still have a great impact on carbon price in the current stage of economic development in China, but this impact is weakening with the continuous healthy development of China’s economy. This paper is different from existing studies, because the empirical results of this paper show that the change of weather conditions has little impact on the price of carbon emission rights in Shenzhen, which we believe is related to the small fluctuation of temperature in Shenzhen during the year.
The influence of coal price and crude oil price on the price of carbon emission rights in Shenzhen is gradually increasing, which is closely related to China’s current energy consumption structure. The reason why the variance contribution rate of coal price is higher than that of crude oil price is that the Shanghai crude oil futures contract selected in this paper has developed late and is still in its initial stage, and its impact on the price of carbon emission rights still needs follow-up research. A 1% change in the price of coal, natural gas, and crude oil will lead to a 0.68%, 0.63%, and 0.41% change in the price of carbon emission rights in Shenzhen, respectively. Among the energy prices, coal price has a higher impact on carbon emission rights price than natural gas and crude oil, but the results of variance decomposition analysis show that the growth rate of variance contribution rate of crude oil price to carbon emission rights price has exceeded that of coal price. Therefore, investors and government decision-makers should also pay more attention to the changes of Shanghai crude oil futures contract price while paying close attention to the changes of coal market price and natural gas market price. Although the variance contribution rate of macroeconomic level to the price of carbon emission rights is relatively small, China is currently in an important transition period from rapid economic development to high-quality development; it is still necessary to attach importance to the healthy development of domestic economy and guard against potential economic crisis.
According to the above research results, we also find that in China’s carbon emission trading market represented by the Shenzhen carbon emission trading market, problems such as the level of marketization is not high, the market liquidity is insufficient, the time lag in the process of market information transmission have been improved to a certain extent, and the ability of the China’s carbon market to regulate price fluctuations has been enhanced, which shows that the launch of the national unified carbon market has indeed achieved certain results, but the situation that China’s carbon market is still in its infancy has not been changed.
In conclusion, we put forward the following policy implications. Firstly, China’s carbon trading market still needs to further improve the information disclosure system and mechanism, strengthen market information transparency, and improve market liquidity, because this is the key to promoting enterprises to actively participate in carbon emission trading. Secondly, more efforts should be put into improving the carbon emission pricing mechanism and reducing market risks, because the price fluctuation in China’s carbon trading market is relatively large, which increases the carbon trading costs and risks of enterprises. Finally, a sound legal system is necessary for the orderly implementation and healthy development of carbon emission trading market. China should improve relevant policies and regulatory systems according to the actual situation of building a unified national carbon market, and learn from the successful experience of other countries in the development of carbon markets.
Appendix
See Table 7.
Table 7 The estimated results of the VAR model
SZA DOIL COAL GAS DHS300 DIND DTEM
SZA(− 1) 0.412177 0.007423 − 0.094997 − 0.069136 0.189030 − 0.018575 − 0.012775
(0.03152) (0.05017) (0.11199) (0.08537) (0.25892) (0.20897) (0.00863)
[13.0764] [0.14795] [− 0.84826] [− 0.80986] [0.73008] [− 0.08889] [− 1.48023]
SZA(− 2) 0.419726 0.054907 0.154789 0.099751 − 0.132614 − 0.047417 0.006436
(0.03141) (0.05000) (0.11161) (0.08508) (0.25803) (0.20825) (0.00860)
[13.3615] [1.09812] [1.38691] [1.17249] [− 0.51394] [− 0.22769] [0.74834]
DOIL(− 1) − 0.011961 − 0.022545 0.124416 0.108944 − 0.128117 − 0.202179 0.012120
(0.02226) (0.03542) (0.07907) (0.06027) (0.18281) (0.14754) (0.00609)
[− 0.53744] [− 0.63641] [1.57346] [1.80745] [− 0.70081] [− 1.37030] [1.98903]
DOIL(− 2) 0.015688 − 0.037366 0.130953 0.108888 0.155900 0.269475 0.003082
(0.02261) (0.03599) (0.08033) (0.06123) (0.18571) (0.14988) (0.00619)
[0.69389] [− 1.03834] [1.63029] [1.77834] [0.83948] [1.79791] [0.49796]
COAL(− 1) 0.047325 − 0.020305 1.276693 1.304833 0.053080 − 0.246108 − 0.007542
(0.03106) (0.04944) (0.11035) (0.08411) (0.25512) (0.20590) (0.00850)
[1.52378] [− 0.41075] [11.5699] [15.5126] [0.20806] [− 1.19528] [− 0.88687]
COAL(− 2) − 0.011657 0.006130 0.449437 0.387876 − 0.122521 − 0.229791 0.016300
(0.03393) (0.05400) (0.12055) (0.09189) (0.27870) (0.22493) (0.00929)
[− 0.34358] [0.11351] [3.72836] [4.22113] [− 0.43962] [− 1.02160] [1.75466]
GAS(− 1) − 0.054685 0.021256 − 0.173557 − 0.229104 − 0.085776 0.325978 0.004509
(0.04051) (0.06449) (0.14394) (0.10972) (0.33279) (0.26859) (0.01109)
[− 1.34977] [0.32961] [− 1.20574] [− 2.08800] [− 0.25775] [1.21367] [0.40649]
GAS(− 2) 0.015917 − 0.003427 − 0.572951 − 0.480780 0.148590 0.155079 − 0.013967
(0.02771) (0.04411) (0.09847) (0.07506) (0.22765) (0.18373) (0.00759)
[0.57431] [− 0.07768] [− 5.81874] [− 6.40536] [0.65270] [0.84404] [− 1.84065]
DHS300(− 1) − 0.011239 − 0.010086 − 0.037882 − 0.029267 0.023779 0.042558 − 0.004276
(0.00795) (0.01265) (0.02823) (0.02152) (0.06527) (0.05268) (0.00218)
[− 1.41438] [− 0.79747] [− 1.34182] [− 1.35995] [0.36432] [0.80788] [− 1.96519]
DHS300(− 2) 0.025784 0.028044 0.002538 0.003633 − 0.043898 0.006865 0.000260
(0.00795) (0.01266) (0.02826) (0.02154) (0.06534) (0.05274) (0.00218)
[3.24136] [2.21487] [0.08981] [0.16862] [− 0.67182] [0.13018] [0.11958]
DIND(− 1) 0.013960 0.014047 0.016812 0.014816 − 0.051491 − 0.024955 0.005060
(0.00978) (0.01557) (0.03474) (0.02649) (0.08033) (0.06483) (0.00268)
[1.42748] [0.90241] [0.48388] [0.55943] [− 0.64101] [− 0.38492] [1.88986]
DIND(− 2) − 0.029594 − 0.024283 − 0.021967 − 0.020253 0.015667 − 0.039682 − 0.000170
(0.00977) (0.01555) (0.03470) (0.02645) (0.08023) (0.06475) (0.00267)
[− 3.02983] [− 1.56190] [− 0.63301] [− 0.76561] [0.19526] [− 0.61280] [− 0.06354]
DTEM(− 1) 0.070052 − 0.024820 0.125962 0.124813 0.458023 0.124120 − 0.117196
(0.12644) (0.20126) (0.44923) (0.34244) (1.03861) (0.83824) (0.03462)
[0.55403] [− 0.12333] [0.28040] [0.36448] [0.44100] [0.14807] [− 3.38531]
DTEM(− 2) − 0.069704 − 0.081670 − 0.962206 − 0.787469 0.423833 0.095546 − 0.154763
(0.12597) (0.20050) (0.44754) (0.34115) (1.03471) (0.83509) (0.03449)
[− 0.55336] [− 0.40733] [− 2.14998] [− 2.30827] [0.40962] [0.11441] [− 4.48732]
C 5.116327 − 3.317423 12.27511 10.54131 4.246933 − 0.977025 0.567026
(1.23659) (1.96829) (4.39346) (3.34904) (10.1576) (8.19796) (0.33857)
[4.13746] [− 1.68543] [2.79395] [3.14756] [0.41810] [− 0.11918] [1.67475]
R-squared 0.621158 0.014230 0.975415 0.985628 0.005446 0.010855 0.051766
Adj. R-squared 0.614675 − 0.002641 0.974995 0.985382 − 0.011576 − 0.006074 0.035537
Sum sq. residual 39025.42 98873.47 492620.3 286246.6 2633175 1715187 2925.535
S.E. equation 6.907123 10.99419 24.54028 18.70654 56.73659 45.79089 1.891150
F-statistic 95.80102 0.843465 2318.193 4007.142 0.319947 0.641185 3.189756
Log likelihood − 2784.224 − 3171.414 − 3840.271 − 3614.159 − 4538.411 − 4359.870 − 1705.183
Akaike AIC 6.720826 7.650454 9.256352 8.713467 10.93256 10.50389 4.130091
Schwarz SC 6.805911 7.735539 9.341437 8.798552 11.01764 10.58897 4.215175
Mean dependent 18.63103 0.240576 656.8128 656.6205 0.852229 1.055931 − 0.001801
S.D. dependent 11.12714 10.97970 155.1893 154.7236 56.41103 45.65244 1.925676
Standard errors in () and t-statistics in []
Author contribution
All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Qingshun Zeng, Jiaqi Zhi, and Changfeng Shi. The first draft of the manuscript was written by Xiaohong Na and Shufang Cheng and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Data Availability
The authors can provide relevant information upon request.
Declarations
Ethical approval
The authors declare that this manuscript has followed the ethical responsibilities of authors mentioned in the guidance.
Consent to participate
All authors of this manuscript consent to participate in this research.
Consent for publication
All authors of this manuscript consent to publish this work under the Environmental Science and Pollution Research.
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
Aatola P Ollikainen M Toppinen A Price determination in the EU ETS market: theory and econometric analysis with market fundamentals Energ Econ 2013 36 380 395 10.1016/j.eneco.2012.09.009
Adedeji AN, Ahmed FF, Adam SU (2021) Examining the dynamic effect of COVID-19 pandemic on dwindling oil prices using structural vector autoregressive model. Energy 230
Alberola E Chevallier J Chèze B Price drivers and structural breaks in European carbon prices 2005–2007 Energ Policy 2008 36 787 797 10.1016/j.enpol.2007.10.029
Batten JA Maddox GE Young MR Does weather, or energy prices, affect carbon prices? Energ Econ 2021 96 105016 10.1016/j.eneco.2020.105016
Benz E Trück S Modeling the price dynamics of CO2 emission allowances Energ Econ 2009 31 4 15 10.1016/j.eneco.2008.07.003
Brauneis A Mestel R Palan S Inducing low-carbon investment in the electric power industry through a price floor for emissions trading Energ Policy 2013 53 190 204 10.1016/j.enpol.2012.10.048
Bredin D Muckley C An emerging equilibrium in the EU emissions trading scheme Energ Econ 2011 33 353 362 10.1016/j.eneco.2010.06.009
Chan YT Collaborative optimal carbon tax rate under economic and energy price shocks: a dynamic stochastic general equilibrium model approach J Clean Prod 2020 256 120452 10.1016/j.jclepro.2020.120452
Chang C Mai T McAleer M Establishing national carbon emission prices for China Renew Sustain Energy Rev 2019 106 1 16 10.1016/j.rser.2019.01.063
Chevallier J Carbon futures and macroeconomic risk factors: a view from the EU ETS Energ Econ 2009 31 614 625 10.1016/j.eneco.2009.02.008
Chevallier J Modelling risk premia in CO2 allowances spot and futures prices Econ Model 2010 27 717 729 10.1016/j.econmod.2010.01.012
Chevallier J A model of carbon price interactions with macroeconomic and energy dynamics Energ Econ 2011 33 1295 1312 10.1016/j.eneco.2011.07.012
Chevallier J Evaluating the carbon-macroeconomy relationship: evidence from threshold vector error-correction and Markov-switching VAR models Econ Model 2011 28 2634 2656 10.1016/j.econmod.2011.08.003
Cong R Lo AY Emission trading and carbon market performance in Shenzhen, China Appl Energ 2017 193 414 425 10.1016/j.apenergy.2017.02.037
Creti A Jouvet P Mignon V Carbon price drivers: Phase I versus Phase II equilibrium? Energ Econ 2012 34 327 334 10.1016/j.eneco.2011.11.001
Cui L Fan Y Zhu L Bi Q How will the emissions trading scheme save cost for achieving China’s 2020 carbon intensity reduction target? Appl Energ 2014 136 1043 1052 10.1016/j.apenergy.2014.05.021
Da Silva PP Moreno B Figueiredo NC Firm-specific impacts of CO2 prices on the stock market value of the Spanish power industry Energ Policy 2016 94 492 501 10.1016/j.enpol.2016.01.005
Daskalakis G Psychoyios D Markellos RN Modeling CO2 emission allowance prices and derivatives: evidence from the European trading scheme J Bank Financ 2009 33 1230 1241 10.1016/j.jbankfin.2009.01.001
Duan K Ren X Shi Y Mishra T Yan C The marginal impacts of energy prices on carbon price variations: evidence from a quantile-on-quantile approach Energ Econ 2021 95 105131 10.1016/j.eneco.2021.105131
Ederington LH, Fernando CS, Lee TK, Linn SC, Zhang HM (2021) The relation between petroleum product prices and crude oil prices. Energ Econ 94
Fan JH Todorova N Dynamics of China’s carbon prices in the pilot trading phase Appl Energ 2017 208 1452 1467 10.1016/j.apenergy.2017.09.007
Fathollahzadeh MH, Tabares-Velasco PC (2021) Electric demand minimization of existing district chiller plants with rigid or flexible thermal demand. Appl Energ 289
Gong X Shi R Xu J Lin B Analyzing spillover effects between carbon and fossil energy markets from a time-varying perspective Appl Energ 2021 285 116384 10.1016/j.apenergy.2020.116384
Guðbrandsdóttir HN Haraldsson HÓ Predicting the price of EU ETS carbon credits Syst Eng Procedia 2011 1 481 489 10.1016/j.sepro.2011.08.070
Gugler K, Haxhimusa A, Liebensteiner M (2021) Effectiveness of climate policies: carbon pricing vs. subsidizing renewables. J Environ Econ Manag 106, 102405
Hammoudeh S Nguyen DK Sousa RM Energy prices and CO2 emission allowance prices: a quantile regression approach Energ Policy 2014 70 201 206 10.1016/j.enpol.2014.03.026
Hammoudeh S Nguyen DK Sousa RM What explain the short-term dynamics of the prices of CO 2 emissions? Energ Econ 2014 46 122 135 10.1016/j.eneco.2014.07.020
Ike GN, Usman O, Alola AA, Sarkodie SA (2020) Environmental quality effects of income, energy prices and trade: the role of renewable energy consumption in G-7 countries. Sci Total Environ 721
Ji C Hu Y Tang B Qu S Price drivers in the carbon emissions trading scheme: evidence from Chinese emissions trading scheme pilots J Clean Prod 2021 278 123469 10.1016/j.jclepro.2020.123469
Ji Q Zhang D China’s crude oil futures: introduction and some stylized facts Financ Res Lett 2019 28 376 380 10.1016/j.frl.2018.06.005
Ji Q Zhang D Geng J Information linkage, dynamic spillovers in prices and volatility between the carbon and energy markets J Clean Prod 2018 198 972 978 10.1016/j.jclepro.2018.07.126
Jiménez-Rodríguez R What happens to the relationship between EU allowances prices and stock market indices in Europe? Energ Econ 2019 81 13 24 10.1016/j.eneco.2019.03.002
Koch N Fuss S Grosjean G Edenhofer O Causes of the EU ETS price drop: recession, CDM, renewable policies or a bit of everything?—New evidence Energ Policy 2014 73 676 685 10.1016/j.enpol.2014.06.024
Laha P Chakraborty B Energy model – a tool for preventing energy dysfunction Renew Sustain Energy Rev 2017 73 95 114 10.1016/j.rser.2017.01.106
Li JF Wang X Zhang YX Kou Q The economic impact of carbon pricing with regulated electricity prices in China—an application of a computable general equilibrium approach Energ Policy 2014 75 46 56 10.1016/j.enpol.2014.07.021
Lin B Chen Y Dynamic linkages and spillover effects between CET market, coal market and stock market of new energy companies: a case of Beijing CET market in China Energy 2019 172 1198 1210 10.1016/j.energy.2019.02.029
Lin B Wesseh PK On the economics of carbon pricing: insights from econometric modeling with industry-level data Energ Econ 2020 86 104678 10.1016/j.eneco.2020.104678
Lin B Xu B Effective ways to reduce CO2 emissions from China’s heavy industry? Evidence from semiparametric regression models Energ Econ 2020 92 104974 10.1016/j.eneco.2020.104974
Lin B Xu B A non-parametric analysis of the driving factors of China’s carbon prices Energ Econ 2021 104 105684 10.1016/j.eneco.2021.105684
Liu L Wang YD Wu CF Wu WF Disentangling the determinants of real oil prices Energ Econ 2016 56 363 373 10.1016/j.eneco.2016.04.003
Lovcha Y Perez-Laborda A Sikora I The determinants of CO2 prices in the EU emission trading system Appl Energ 2022 305 117903 10.1016/j.apenergy.2021.117903
Meier H Rehdanz K Determinants of residential space heating expenditures in Great Britain Energ Econ 2010 32 949 959 10.1016/j.eneco.2009.11.008
Moz-Christofoletti MA, Pereda PC (2021) Distributional welfare and emission effects of energy tax policies in Brazil. Energ Econ 104
Nie PY, Chen ZR, Wang C, Chen XL (2019) Innovation analysis under trading energy efficiency. Energy 186
Paolella MS Taschini L An econometric analysis of emission allowance prices J Bank Financ 2008 32 2022 2032 10.1016/j.jbankfin.2007.09.024
Seifert J Uhrig-Homburg M Wagner M Dynamic behavior of CO2 spot prices J Environ Econ Manag 2008 56 180 194 10.1016/j.jeem.2008.03.003
Shi B Li N Gao Q Li G Market incentives, carbon quota allocation and carbon emission reduction: evidence from China’s carbon trading pilot policy J Environ Manage 2022 319 115650 10.1016/j.jenvman.2022.115650 35820308
Song Y Liu T Ye B Zhu Y Li Y Song X Improving the liquidity of China’s carbon market: Insight from the effect of carbon price transmission under the policy release J Clean Prod 2019 239 118049 10.1016/j.jclepro.2019.118049
Tan X Sirichand K Vivian A Wang X How connected is the carbon market to energy and financial markets? A systematic analysis of spillovers and dynamics Energ Econ 2020 90 104870 10.1016/j.eneco.2020.104870
Tang L Wu J Yu L Bao Q Carbon emissions trading scheme exploration in China: a multi-agent-based model Energ Policy 2015 81 152 169 10.1016/j.enpol.2015.02.032
Tian YX Yang XM Asymmetric effects of industrial energy prices on carbon productivity Environ Sci Pollut R 2020 27 42133 42149 10.1007/s11356-020-10204-5
Tu Q Mo JL Coordinating carbon pricing policy and renewable energy policy with a case study in China Comput Ind Eng 2017 113 294 304 10.1016/j.cie.2017.09.026
Wang J, Lv K, Bian Y, Cheng Y (2017) Energy efficiency and marginal carbon dioxide emission abatement cost in urban China. energ policy 105, 246–255
Wang Q Hang Y Su B Zhou P Contributions to sector-level carbon intensity change: an integrated decomposition analysis Energ Econ 2018 70 12 25 10.1016/j.eneco.2017.12.014
Wang YD Guo ZY The dynamic spillover between carbon and energy markets: new evidence Energy 2018 149 24 33 10.1016/j.energy.2018.01.145
Weng Q Xu H A review of China’s carbon trading market Renew Sustain Energy Rev 2018 91 613 619 10.1016/j.rser.2018.04.026
Xu B Luo Y Xu R Chen J Exploring the driving forces of distributed energy resources in China: using a semiparametric regression model Energy 2021 236 121452 10.1016/j.energy.2021.121452
Yang C Lv F Fang L Shang X The pricing efficiency of crude oil futures in the Shanghai International Exchange Financ Res Lett 2020 36 101329 10.1016/j.frl.2019.101329
Zeng S Nan X Liu C Chen J The response of the Beijing carbon emissions allowance price (BJC) to macroeconomic and energy price indices Energ Policy 2017 106 111 121 10.1016/j.enpol.2017.03.046
Zeng S, Su B, Zhang ML, Gao Y, Liu J, Luo S, Tao QM (2021) Analysis and forecast of China’s energy consumption structure. Energ Policy 159
Zhang Q Di P Farnoosh A Study on the impacts of Shanghai crude oil futures on global oil market and oil industry based on VECM and DAG models Energy 2021 223 120050 10.1016/j.energy.2021.120050
Zhao X Han M Ding L Kang W Usefulness of economic and energy data at different frequencies for carbon price forecasting in the EU ETS Appl Energ 2018 216 132 141 10.1016/j.apenergy.2018.02.003
Zheng Y Zhou M Wen F Asymmetric effects of oil shocks on carbon allowance price: evidence from China Energ Econ 2021 97 105183 10.1016/j.eneco.2021.105183
Zhu B Ye S Han D Wang P He K Wei Y Xie R A multiscale analysis for carbon price drivers Energ Econ 2019 78 202 216 10.1016/j.eneco.2018.11.007
| 36502476 | PMC9741706 | NO-CC CODE | 2022-12-15 23:15:21 | no | Environ Sci Pollut Res Int. 2022 Dec 11;:1-16 | utf-8 | Environ Sci Pollut Res Int | 2,022 | 10.1007/s11356-022-24577-2 | oa_other |
==== Front
Indian J Otolaryngol Head Neck Surg
Indian J Otolaryngol Head Neck Surg
Indian Journal of Otolaryngology and Head & Neck Surgery
2231-3796
0973-7707
Springer India New Delhi
3299
10.1007/s12070-022-03299-4
Original Article
Should Silicone Lacrimal Stenting be a Better Choice for Primary Endoscopic Powered Dacryocystorhinostomy?
http://orcid.org/0000-0003-3977-3230
Dutta Mainak [email protected]
1
Ghatak Soumya 2
Bandyopadhyay Samrat 3
1 grid.413204.0 0000 0004 1768 2335 Department of Otorhinolaryngology and Head-Neck Surgery, Medical College and Hospital, Kolkata, West Bengal India
2 Department of Otorhinolaryngology and Head-Neck Surgery, Coochbehar Government Medical College and Hospital, Coochbehar, West Bengal India
3 Department of Otorhinolaryngology and Head-Neck Surgery, Narayana Superspeciality Hospital, Howrah, West Bengal India
11 12 2022
17
9 2 2022
21 11 2022
© Association of Otolaryngologists of India 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.
Objective: To compare endoscopic dacryocystorhinostomy (EnDCR) with and without silicone lacrimal stenting through subjective (patients’) and objective (surgeons’) outcome parameters. Methodology: Following defined selection criteria, EnDCR was performed on patients with primary chronic dacryocystitis with post-saccal stenosis. Every alternate patient had silicone lacrimal stenting (group A: no stenting; group B: with stenting); stents were removed at three months. At six months (minimum follow-up period), patients’ responses on symptom relief (through a five-point score) and naso-endoscopic evaluation (visualization of rhinostome; presence of granulations and synechiae; lacrimal drainage patency by estimating methylene blue flow pattern) were compared between the groups. Results: Each group had 20 patients. There was no statistically significant difference in group-wise follow-up periods. Five-point score at six months revealed 85% and 95% of patients in groups A and B, respectively, experienced “success”; among them, 60% and 75% were “symptom-free”. The majority (75%) in group B experienced no discomfort from stenting. Naso-endoscopy revealed 80% patients in group A and 65% in group B had well-delineated rhinostome, albeit with granulations in 25% and 50%, respectively. Spontaneous dye flow was achieved, respectively, in 75% and 90%. The difference in none of the subjective and endoscopic parameters achieved statistical significance. None had synechia; fibrosis was seen in the four patients with no dye flow even with pressure/massaging. Conclusion: There was no statistically significant difference in EnDCR with and without silicone lacrimal stenting in the overall outcome of symptomatic improvement and endoscopic assessment of the surgical site.
Keywords
Endoscopic dacryocystorhinostomy
Lacrimal stenting
Silicone stent
Rhinostome
Lacrimal drainage
==== Body
pmcLevel of Evidence
2b.
Introduction
Endoscopic dacryocystorhinostomy (EnDCR) is the gold standard operation for chronic dacryocystitis with post-saccal stenosis. The procedure, however, is not without complications, which are predominantly due to the surgical technique itself. These include failure to preserve the medial wall of the sac and leaving bare bone around the rhinostome leading to tissue trauma, poor healing, granulations, neo-osteogenesis, bone remodeling, and synechia [1, 2]. In fact, stenosis of the rhinostome is a major complication that can follow both endoscopic and external approaches, necessitating revision surgery. Among the modifications suggested to reduce the incidence and extent of rhinostome stenosis, bicanalicular silicone lacrimal stent is one of the most widely practised techniques. In recent times, there has been a considerable body of work on EnDCR with stenting [3–7], although admittedly, proper indication and long-term results of stent placement require further, periodic evaluation. The present study investigates, through a rigorously carried out methodology, the outcome of EnDCR with silicone stenting in terms of success and complications, and compares the data with those of EnDCR without stenting. The study thereby re-explores the rationality and suitability of placing stent in EnDCR in a primary, uncomplicated clinical setting.
Materials and Methods
Study Set-Up
This comparative study was conducted in a tertiary care teaching institute during January 2019 to June 2020. Patients with epiphora were subjected to probe test and lacrimal syringing. Those with a hard stop (probe test) and with slow/delayed regurgitation from the upper punctum (lacrimal syringing), signifying primary chronic dacryocystitis with post-saccal stenosis were considered. The patients were the primary attendee in the department of Otorhinolaryngology and Head-Neck surgery, but those who opted for EnDCR were also referred from the department of Ophthalmology. They were recruited for surgery during January-December 2019, and were followed up for a minimum period of six months, up to June 2020.
Patients with lid problems (ectropion, entropion, lid laxity, ptosis, lagophthalmos), history of ipsilateral DCR, acute dacryocystitis [with lacrimal sac abscess/pyomucocele, skin inflammation, draining fistula in the lacrimal region], predominantly purulent regurgitate on syringing, punctal stenosis, symptoms suggesting malignancy or mass lesion (neoplastic/granulomatous), post-traumatic bone deformity, pre-existing primary bone diseases affecting the nose and orbit, co-existent systemic morbidities like diabetes mellitus and auto-immune endarteritis, and those with clinical features suggesting primary or secondary atrophic rhinitis, acute/acute-on-chronic rhinosinusitis and allergic rhinitis were excluded from the study. This list also included patients lost to follow-up. Those presenting with bilateral chronic dacryocystitis were operated only on the side with dominant symptoms, or according to their choice when the distress was comparable on both sides.
The study was approved by the Institutional Ethical Committee. Informed consent in writing was obtained from each patient prior to his/her inclusion in the study. Investigations and interventions were strictly according to the principles stated in the declaration of Helsinki 1964 and its subsequent revisions.
Prior to surgery, all patients were prescribed co-amoxiclav in appropriate dosage for seven days, and topical antibiotic medication in the affected eye. All surgeries and subsequent evaluations at follow-up were performed by the authors themselves as a team who followed identical surgical principles and follow-up protocol. The authors had comparable experience in EnDCR, and had been practising the procedure for at least eight years.
Surgical Procedure
All patients were operated under general anesthesia. Adequate nasal decongestion was achieved with cotton pledgets soaked with diluted 1:1000 adrenaline (3 ml adrenaline in 30 ml normal saline). The endoscopic surgery console system consisted of 4 mm 0 and 30 degree endoscopes (Karl Storz SE & Co. KG; Tuttlingen, Germany) and three-chip camera with high-definition monitor system (Stryker; Kalamazoo, Michigan, USA). Septal deviation significant enough needing correction, and the anatomic disposition of the middle turbinate, axilla and agger nasi were noted.
A posterior-based mucosal flap was raised following the surgical steps and dimensions described by Wormald PJ [2]. The lacrimal bone was identified at its junction with the frontonasal process of maxilla, and was peeled off. The exposed frontonasal process was next completely removed, including the area where it articulated with the skull-base, by powered drilling (endo-nasal drill bur for Straightshot® M5 microdebrider handpiece and Integrated Power Console System; Medtronic; Minneapolis, Minnesota, USA). The agger nasi cell, when present and large enough, was opened up. This was irrespective of the patients’ symptomatology. The reasons for opening up of the agger cell was because when large, it caused hindrance to the endoscopic vision, and also because agger signified pneumatization of the lacrimal sac bed (frontonasal process of maxilla and lacrimal bone), and the purpose of drilling was to expose the entire medial wall of the sac completely, without any bony overhang. The medial wall of the lacrimal sac was next infiltrated with diluted adrenaline solution before incising. The lower puncta was then dilated, and a Bowman’s probe (Karl Storz SE & Co. KG; Tuttlingen, Germany) was inserted to tent the medial wall of the lacrimal sac. With a 30º endoscope, the medial wall of the lacrimal sac was incised with a disposable keratome in a horizontal H-shaped manner, with the vertical limb of the incision made at the peak of the tent. The anterior and posterior flaps thus created were excised with 45º upturned Blakesley forceps (Karl Storz SE & Co. KG; Tuttlingen, Germany), and the lateral wall of the lacrimal sac was flushed with the lateral nasal wall. The cavity of the lacrimal sac was irrigated with normal saline; this was followed by lacrimal syringing whereby free flow of normal saline was ensured.
Every alternate patient was chosen for silicone stenting (group B). Those who were not, were considered as group A. In group B, one end of the silicone tube (Medtronic, Jacksonville, Florida, USA) was inserted through the upper punctum and the other end through the lower one. Both ends were then taken outside the nasal cavity and a knot was applied which was pushed back close to the lateral wall of the lacrimal sac. Several such knots (4–6 in number) were made, care being taken not to make them too tight.
Finally, the posterior-based mucosal flap so long tucked between the septum and middle turbinate was retrieved, cut into two incomplete halves, and were trimmed appropriately to cover any bare bone. Antibiotic-impregnated Merocel® nasal pack (Medtronic; Minneapolis, Minnesota, USA), was inserted and kept in-situ for 24 h.
Follow-Up and Subsequent Evaluation
During the follow-up period, gentle digital massaging was advised (repetitive circular movements with the pulp of the index finger placed over the lacrimal fossa, medial to the medial canthus), 10 rounds at a time, four times/day, for three months. Irrigation with isotonic normal saline was also advised for a minimum of three weeks. Minimum follow-up period for each patient was six months, although naso-endoscopy and subjective point-based evaluation were performed periodically at one, three and six months, and subsequently bi-annually for patients within the scope of the study period. Stents were removed at three months. No lacrimal syringing was attempted in either group.
Assessment at six months (the minimum follow-up period) was considered for the final outcome analysis for both objective and subjective evaluation. A subjective evaluation in a five-point scale was made depending upon the patients’ responses regarding the outcome of the surgery, that is, relief from epiphora. Patients’ responses varied from “symptom-free”, “significantly improved”, “slightly improved”, “no improvement”, and “worse”. The five-point score was constructed in the same order, the highest point assigned to “symptom-free” (score 5), and the least point to “worse” (score 1). Scores 3–5 were considered as “success”, and 1–2 as “failure”. Besides, an objective assessment was also carried out based on the endoscopic findings at six months. Observations were made on: (a) whether the rhinostome could be clearly seen; (b) presence of granulations and/or synechiae at the rhinostome site; and (c) whether the lacrimal drainage was patent, by estimating the flow pattern of methylene blue at syringing (spontaneous free flow, flow only on pressing or massaging the lacrimal sac region, or no flow). An objective checklist was maintained during the endoscopic evaluation, and appropriate responses were recorded.
Outcome of both the subjective and objective analyses were compared. Additionally, any outcome or complication exclusive to stenting, like opened knots, granulations at the opening of the common canaliculus etc., were specifically noted.
Statistical Analysis
The variations were analyzed as percentage of the two groups. Student’s t-test was used to compare the mean of the groups like age, sex and follow-up period. Comparisons of the subjective and objective outcome between the groups were performed by chi-square test. P values < 0.05 were considered statistically significant. All calculations were carried out in multiple Windows Excel spreadsheets (Microsoft Corporation; Redmond, Washington, USA), using SPSS (Statistical Package for Social Sciences) software version 22 (IBM Corporation; Armonk, New York, USA).
A level of evidence of 2b has been assigned to this study, following guidelines provided by the Oxford Centre of Evidence-based Medicine [8].
Results
Patients were included in an alternative manner such that there were 20 patients each in group A (no stent given) and group B (stenting done). The study cohort overall had a mean age of 44.25 ± 16.44 years (range: 18–73 years). The mean age was 42.25 ± 16.63 years (range: 18–72 years) for group A, and 46.25 ± 16.42 years (range: 19–73 years) for group B. The difference was not statistically significant (p = 0.765; student’s t-test; degrees of freedom [DF] = 3.8). There was an evident female preponderance overall (67.5%) and intra-group (70% in group A; 65% in group B), the difference not being statistically significant (p = 0.114; Pearson’s χ2 test, DF = 1). Irrespective laterality of involvement, 60% patients had dacryocystitis predominantly on the right that actually required surgical intervention. The difference between the groups in this regard (65% and 55% in groups A and B, respectively) was not statistically significant (p = 0.417; Pearson’s χ2 test, DF = 1).
Naso-endoscopy performed immediately prior to surgery revealed ipsilateral deviated nasal septum in three patients (two in group A; one in group B) that ultimately required limited endoscopic septoplasty for optimum surgical exposure. The agger nasi was prominent in 23 patients (10 in group A; 13 in group B) such that it needed to be opened up (reasons stated earlier) during drilling of the lateral nasal wall bone forming bed of the lacrimal fossa.
The mean follow-up was for 11.20 ± 3.28 months (range: 6–18 months). There was no statistically significant difference between the groups with mean follow-up of 11.34 ± 3.12 months (range: 8–18 months) in group A, and 11.05 ± 3.52 months (range: 6–18 months) in group B (p = 0.19; student’s t-test, DF = 3.8). Analysis of the five-point subjective outcome score at six months (the minimum follow-up period) revealed that 85% and 95% patients in group A and B, respectively, experienced “success”. However, the difference was not statistically significant (p = 0.277; Pearson’s χ2 test with Yate’s correction, DF = 1). Among those experiencing “success”, 60% in group A and 75% in group B declared themselves symptom-free (score 1). Majority (75%) in group B had no discomfort from the stent like irritation, eyelid swelling, pain, etc. The subjective outcome at six months follow-up is summarized in Table 1.
Table 1 Subjective evaluation at the end of six months post-surgery (n = 40)
Group A (without stent) (n = 20) Group B (with stent) (n = 20) p-value
Subjective Evaluation
• Symptom free (score 1)
• Significant improvement (score 2)
• Slight improvement (score 3)
• Same (score 4)
• Worse (score 5)
12 (60%)
4 (20%)
1 (5%)
2 (10%)
1 (5%)
15 (75%)
4 (20%)
0
1 (5%)
0
0.256 *
Outcome
• Success
• Failure
17 (85%)
3 (15%)
19 (95%)
1 (5%)
0.277**
Discomfort from stent (irritation, eyelid swelling, pain, etc.)†
• Yes
• No
-
-
5 (25%)
15 (75%)
-
* Pearson’s χ2 test, degrees of freedom (DF) = 1
** Pearson’s χ2 test with Yate’s correction, DF = 1
† Information obtained at six months on recollection of symptoms prior to and/or at three months post-surgery, when the stents were removed
Endoscopic evaluation at six months post-surgery revealed that 80% patients in group A and 65% in group B had well-delineated rhinostome, although granulations were there, respectively, in 25% and 50%. These observations did not achieve statistical significance. Spontaneous dye flow was achieved in 75% patients in group A and in 90% in group B; however, the difference was not statistically significant (p = 0.652; Pearson’s χ2 test with Yate’s correction, DF = 1). No synechia was encountered in any group, although fibrosis was seen in all four patients irrespective of the groups who had no dye flow even with pressure/massaging. Results of endoscopic evaluation at six months follow-up are summarized in Table 2.
Table 2 Endoscopic evaluation at the end of six months post-surgery (n = 40)
Group A (without stent) (n = 20) Group B (with stent) (n = 20) p-value
Rhinostome
• Visible
• Invisible
16 (80%)
4 (20%)
13 (65%)
7 (35%)
1.536*
Granulations
• Present
• Absent
5 (25%)
15 (75%)
10 (50%)
10 (50%)
2.667*
Methylene blue flow
• Spontaneous
• With pressure/massaging
• No flow
15 (75%)
2 (10%)
3 (15%)
18 (90%)
1 (5%)
1 (5%)
0.652**
Knot opened - 1 (5%) -
* Pearson’s χ2 test, degrees of freedom (DF) = 1
** Pearson’s χ2 test with Yate’s correction, DF = 1
Discussion
Silicone stents have been proved to be useful in EnDCR in selected indications, like in revision, infected lacrimal system (lacrimal mucopyocele, granulations),5 and also in primary surgery [1, 6, 7]. It keeps the distal common canalicular outflow patent by preventing stenosis and synechia [1]. Recent researches on animal model have introduced more biocompatible and biodegradable materials (e.g., polylactic acid-polyprolactone-polyethylene glycol complexes) for lacrimal stent that minimize cicatrization of the canaliculi and subsequent stenosis [9]. On the other hand, a “tight” common canaliculus on lacrimal probing is considered one of the indications for stenting [10]. The stent in such a situation would keep the tightened valve of Rosenmüller guarding the distal common canalicular opening dilated. There are counter-opinions as well; several studies reported favorable outcomes in EnDCR without stenting [11, 12], and authors have reported in comparative analyses statistically significant success rates with EnDCR without stenting [13]. Bicanalicular stenting has often been implicated for causing post-operative bleed and eyelid complications like irritation, swelling, pain etc. [1]. Furthermore, prolonged duration of stenting results in microbial growth that might predispose to stenosis [4, 13], although a recent study revealed that appropriate antimicrobial therapy positively correlates with the patency of the lacrimal system in the long term [9]. Therefore it is apparent that the indications for lacrimal stenting and its short and long-term outcomes are essentially heterogeneous and are yet to be standardized.
The present study utilizes this scope for a re-assessment and investigates the effects of silicone stenting during EnDCR on subsequent lacrimal patency. The methodology adopted here is reproducible and transparent, and expresses rigor in terms of study design construct and evaluation of results. The study deals with two defined patient groups who are homogeneous regarding their age-sex composition, and in their clinical profile. Although chronic dacryocystitis has been noted to be commonly involving the left side and female sex [14], the results in our study hinted otherwise, might be due to a comparatively lower sample strength. The differences in laterality and sex predilection however did not attain statistical significance. The study adopted surgical principles and follow-up protocol which were uniform for both groups, universally accepted, evidence-based, and reproducible. Although follow-up endoscopy was carried out periodically, the study took the liberty to consider evaluation at six months as the final outcome for analysis and interpretation. This ensured uniformity in the follow-up period, and also provided optimum time for healing, and resolution of granulation tissue, if any, following surgery and removal of stent. Moreover, the endoscopic surgeons were in similar positions of their learning curves for the procedure concerned. The study setting further ensured that the operative field remained comparable in the two groups because there were no or minimal infection (mucopurulence) in the lacrimal environment. This eliminated any possible bias in patient selection that could influence the outcome regarding healing. The study principally aimed at finding out whether lacrimal stenting actually had any benefit in the endoscopic surgical management of chronic dacryocystitis in primary, uncomplicated events, and therefore, logical patient selection through a strict set of exclusion criteria was an important determinant for its execution.
Our study revealed no statistically significant differences in the outcome parameters in patients undergoing EnDCR with and without lacrimal stenting, although the values numerically favored the former group (Tables 1 and 2). The comparable outcome can be attributed to the surgical procedure itself. The aim of surgery was to expose the lateral wall of the lacrimal sac in its entirety, following drilling off of large agger nasi cell and excision of the medial wall of the sac so that its lateral wall gets flushed with the lateral nasal wall, without any bony overhang (especially in the superior aspect) guarding the medial opening of the common canaliculus. Also, as stated previously, the surgical technique was uniform in both groups. Digital lacrimal massaging in the follow-up period might be another factor for such comparable results and an overall favorable outcome.
That limited granulations were observed in the post-operative endoscopic observations around the rhinostome in both groups A and B (25% and 50% respectively; not statistically significant; Table 2) could be explained by the mucosa-sacrificing technique adopted, potentially leaving minimal bare bone to heal by secondary intention. Granulations were relatively more in group B. Stenting itself, and microbial growth following its long-term stay might play additional role in it. However, since there was no statistically significant difference between the groups regarding the occurrence of granulations, this study cannot conclusively prove that patients undergoing stenting had a predilection for granulation tissue formation.
It should be noted that none was subjected to lacrimal syringing in the follow-up period. We in our institute prefer routine digital massage of the lacrimal fossa over periodic lacrimal syringing after surgery. In our experience, we find this practice obviates potential injury to the puncta and canaliculi, and subsequent stenosis that often follow repeated, too often syringing. It can be speculated that, given the favorable outcomes, religious digital lacrimal massaging as advised was a practical alternative for syringing and was sufficient to maintain lacrimal patency.
Nevertheless, in spite of the uniform and meticulous surgical techniques and post-operative care, four patients experienced failure (persistence, worsening, or return of symptoms) due to fibrosis. Fibrosis generally relates to stenting, but was found to be numerically more in group A. Because difference in outcome in the two groups did not attain statistical significance, it cannot be concluded from the present study that not providing the stent would result in a better outcome. Interestingly, at six months follow-up, endoscopic evaluation could not localize the rhinostome in 20% and 35% of subjects in group A and B respectively (Table 2). Except for the four patients who had persistence, worsening, or return of symptoms due to fibrosis, the remaining patients in this cohort where the rhinostome could not be localized did not experience “failure” (Table 1). This suggests that non-visualization of the rhinostome could be because of healthy mucosalization of the lateral wall of the lacrimal sac that did not occlude the common canalicular opening, and does not always suggest or predict failure.
There are several limitations of the present study. First, the number of patients were limited, and the elaborate and strict exclusion criteria adopted might be responsible for this. However, the study set-up being a tertiary-care teaching institute, and given the relatively short study duration, the sample strength is not negligible from the perspective of determining statistical strength and significance. Also, the rigid exclusion criteria stress on the homogeneity of the study population at the cost of the generalizability of the study outcomes. Nevertheless, the fact that there were no statistically significant differences in the outcome parameters between the groups might ultimately be related to the limited sample strength.
Second, the subjects were grouped on an alternative basis; thus there was a palpable chance factor in having the groups age and sex-matched because the methodology adopted no randomization technique. The chance factor might also play its role in not having synechia in the limited number of patients requiring septal correction prior to EnDCR. A larger sample size could have resulted in a different outcome.
Third, although the outcome instruments as described are being used in our institute as part of the follow-up assessment protocol that provide us with consistent results, there is need for their statistical validation, especially for the point-based subjective evaluation. In this context, it should be acknowledged that the minimum follow-up duration should have been extended to bring reliability to the study and to ensure reproducibility to the results. However, the ongoing coronavirus disease pandemic forced us to limit the minimum follow-up duration. Nevertheless, the results obtained are from a rigorously conducted methodology, hence they firmly show a definite trend.
Fourth, the outcome opening up of agger nasi could have had on rhinostome patency and formation of granulations was not assessed. Agger was considered here only for anatomic and surgical interest when it caused visual obstruction and/or reduced the endoscopic working space, and the effect of agger mucosa on epithelialization around the rhinostome was not assessed.
Finally, evaluation of lacrimal patency distal to the puncta during the follow-up period relied only on the mechanical property of luminal fluid transport, and did not consider integrity of the lacrimal pump system. With no provision for lacrimal scintigraphy in the present study set-up, it would not be prudent to conclude that fibrosis and/or granulations following EnDCR with/without lacrimal stenting were the only reasons for the seven patients (five in group A and two in group B; the difference not statistically significant) not achieving spontaneous dye flow on syringing.
It needs to be stressed that the present study might not be specific and universal about answering the question “whom to stent”, but through an elaborate and rigorous exclusion criteria, it has attempted to explore how useful lacrimal stenting would be in patients undergoing EnDCR in a given setting of primary, uncomplicated chronic dacryocystitis. For analyzing the applicability of lacrimal setting in specific situations like mucopurulence, revision, associated comorbidities etc., other study design instruments like prospective, randomized control trials are required in a problem-specific manner. The present study essentially and deliberately deals with a niche population that is tailor-made to be well-defined and homogeneous, and therefore, the resultant restricted sample size is effectively one of its major strengths. The usefulness of lacrimal stenting has been judged here in a neutral, unbiased, homogeneous study setting, and the subsequent outcomes would add to the existing body of literature by virtue of its consistency and reproducibility.
Conclusion
Notwithstanding the numerical edge and the apparent clinical benefits of EnDCR with lacrimal stenting, this study did not reveal any statistically significant difference between EnDCR with and without stenting in terms of overall outcome of surgery (success/failure; and patency of the lacrimal system), subjective improvement of the symptom spectrum, and post-operative endoscopic assessment of the surgical site (visibility of the rhinostome and presence of granulations).
Declarations
Conflict of Interest
None.
Financial Disclosure
None declared.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Kim DH Kim SI Jin HJ Kim S Hwang SH The clinical efficacy of silicone stents for endoscopic dacryocystorhinostomy: a Meta-analysis Clin Exp Otorhinolaryngol 2018 11 3 151 157 10.21053/ceo.2017.01781 29590744
2. Wormald PJ (2018) Powered endoscopic dacryocystorhinostomy. Endoscopic sinus surgery: anatomy, three-dimensional reconstruction, and surgical technique, 4th edn. Delhi:Thieme, pp 161–176
3. Unlu HH Gunhan K Baser EF Songu M Long-term results in endoscopic dacryocystorhinostomy: is intubation really required? Otolaryngol Head Neck Surg 2009 140 4 589 595 10.1016/j.otohns.2008.12.056 19328352
4. Goel R Nagpal S Kamal S Kumar S Mishra B Loomba PS Study of microbial growth on silicone tubes after transcanalicular laser-assisted dacryocystorhinostomy and correlation with patency Nepal J Ophthalmol 2016 8 16 119 127 10.3126/nepjoph.v8i2.16992 28478465
5. Lin L, Yang L, Jin X, Zhao Y, Fan F Management lacrimal sac abscesses using lacrimal probe and crawford silicon tube. BMC Ophthalmol 2016;Nov 30;16(1):211. 10.1186/s12886-016-0378-y
6. Al-Qahtani AS Primary endoscopic dacryocystorhinostomy with or without silicone tubing: a prospective randomized study Am J Rhinol Allergy 2012 26 4 332 334 10.2500/ajra.2012.26.3789 22732136
7. Feng YF Cai JQ Zhang JY Han XH A meta-analysis of primary dacryocystorhinostomy with and without silicone intubation Can J Ophthalmol 2011 46 6 521 527 10.1016/j.jcjo.2011.09.008 22153640
8. https://www.cebm.net/2009/06/oxford-centre-evidence-based-medicine-levels-evidence-march-2009/ [accessed on: September 4, 2021]
9. Zhan X Guo X Liu R Hu W Zhang L Xiang N Intervention using a novel biodegradable hollow stent containing polylactic acid-polyprolactone-polyethylene glycol complexes against lacrimal duct obstruction disease PLoS ONE 2017 12 6 e0178679 10.1371/journal.pone.0178679 28570687
10. Callejas CA Tewfik MA Wormald PJ Powered endoscopic dacryocystorhinostomy with selective stenting Laryngoscope 2010 120 7 1449 1452 10.1002/lary.20916 20564733
11. Smirnov G Tuomilehto H Terasvirta M Nuutinen J Seppa J Silicone tubing is not necessary after primary endoscopic dacryocystorhinostomy: a prospective randomized study Am J Rhinol 2008 22 2 214 217 10.2500/ajr.2008.22.3132 18416983
12. Ciğer E Balci MK Arslanoğlu S Eren E Endoscopic-powered dacryocystorhinostomy without stenting: long-term outcomes of 120 procedures Am J Rhinol Allergy 2018 32 4 303 309 10.1177/1945892418773638 29745245
13. Samimi DB Ediriwickrema LS Bielory BP Miller D Lee W Johnson TE Microbiology and Biofilm Trends of Silicone Lacrimal Implants: comparing infected Versus routinely removed stents Ophthalmic Plast Reconstr Surg 2016 32 6 452 457 10.1097/IOP.0000000000000590 26588208
14. https://emedicine.medscape.com/article/1210688-overview#showall (accessed on: November 15, 2021)
| 0 | PMC9741707 | NO-CC CODE | 2022-12-14 23:29:57 | no | Indian J Otolaryngol Head Neck Surg. 2022 Dec 11;:1-7 | utf-8 | Indian J Otolaryngol Head Neck Surg | 2,022 | 10.1007/s12070-022-03299-4 | oa_other |
==== Front
J Happiness Stud
J Happiness Stud
Journal of Happiness Studies
1389-4978
1573-7780
Springer Netherlands Dordrecht
609
10.1007/s10902-022-00609-z
Research Paper
Well-being Effects of Natural Disasters: Evidence from China’s Wenchuan Earthquake
Wang Zou 1
http://orcid.org/0000-0002-3456-3157
Wang Fei [email protected]
2
1 grid.24539.39 0000 0004 0368 8103 Institute of Regional and Urban Economics, School of Applied Economics, Renmin University of China, Beijing, China
2 grid.24539.39 0000 0004 0368 8103 School of Labor and Human Resources, Renmin University of China, Beijing, China
11 12 2022
125
30 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.
This study finds that the Wenchuan earthquake in 2008, one of China’s most catastrophic earthquakes, substantially decreased victims’ subjective well-being even after incorporating the offsetting effects of post-disaster relief programs. This net well-being impact lasted for nearly 10 years and was on average equivalent to a loss of 67% of the average equivalized household income. Although the post-disaster measures largely restored income, health, and employment, they failed to prevent well-being losses due to family dissolution, as reflected in the higher rates of divorce and widowhood after the earthquake. We find that rural populations, older adults, the less educated, and residents without social insurance were more vulnerable to the earthquake shock. This study uses six waves of a nationally representative dataset of China and a difference-in-differences approach to identify the short- and long-term causal well-being effects of the Wenchuan earthquake. Deeper analyses on mechanisms and heterogeneity suggest that post-disaster policies should focus more on aspects beyond economic factors and on the well-being of disadvantaged populations in particular.
Supplementary Information
The online version contains supplementary material available at 10.1007/s10902-022-00609-z.
Keywords
Subjective well-being
Natural disasters
Wenchuan earthquake
Difference-in-differences
http://dx.doi.org/10.13039/501100004260 Renmin University of China 21XNLG03 Wang Fei
==== Body
pmcIntroduction
Natural disasters, including meteorological, geological, and biological disasters, often induce threats to economic development (Gignoux & Menéndez, 2016; Strobl, 2012), human capital (Caruso, 2017; Caruso & Miller, 2015), and subjective well-being (SWB; Carroll et al., 2009; Luechinger & Raschky, 2009). The well-being cost of natural disasters could even be comparable to the financial losses caused by disasters (Carroll et al., 2009; Luechinger & Raschky, 2009; Ohtake, et al., 2016; Rehdanz et al., 2015). This study quantifies the well-being effect of the 2008 Wenchuan earthquake, China’s most devastating natural disaster in decades, and deciphers the mechanisms through which the disaster impacted victims’ well-being. This study’s findings can help policymakers evaluate the welfare losses associated with natural disasters and develop targeted measures to mitigate damage to public well-being, particularly in the developing world, where disasters may be more destructive and disaster relief is relatively insufficient.
A branch of research has focused on the well-being impact of natural disasters, such as droughts (Carroll et al., 2009), floods (Luechinger & Raschky, 2009), and hurricanes (Kimball et al., 2006; LaJoie et al., 2010). People facing these disasters could be prepared based on available forecasts and seasonal experience, or may have enough time to conduct urgent adjustments (Berlemann, 2016). In contrast, an earthquake shock is unpredictable and random, and may lead to greater hardship. Some studies have focused on Japan, where earthquakes occur frequently. Yamamura (2012) examines changes in victims’ SWB over time after the 1995 Hanshin-Awaji earthquake. Rehdanz et al. (2015) estimate the loss in SWB caused by the 2011 Fukushima disaster, which included an earthquake, a tsunami, and a nuclear accident. A few studies examine the short-term impact of the Wenchuan earthquake on well-being. Xin et al. (2009) and Tao et al. (2009) assess local victims’ SWB 1 to 3 months after the Wenchuan earthquake. Li et al. (2009) measure the SWB of local residents periodically in the 4th, 6th and 8th month after the Wenchuan earthquake. All the studies confirm the considerable negative effects of earthquakes on SWB, at least in the short term.
Studies usually track victims’ SWB over time (Li et al., 2009; Tao et al., 2009; Xin et al., 2009) or compare victims and non-victims’ SWB at a certain point in time (Yamamura, 2012) to estimate the well-being effect of earthquakes. Despite the randomness of earthquakes, the time-series change or cross-sectional comparison of SWB may have generated biased estimates. On the one hand, temporal changes in victims’ SWB may have resulted from sources other than earthquakes. On the other hand, people choosing to reside in an earthquake-prone region may have different SWB-related characteristics from those living in other areas. Rehdanz et al. (2015) handle this issue by using a difference-in-differences (DID) strategy to identify the causal effects of the 2011 Fukushima disaster on SWB based on variations in pre- and post-disaster data and distance from the epicenter. As their dataset contains only one wave before the disaster, the critical parallel-trend assumption of DID is not formally tested.
A mechanism analysis is crucial as it could enrich our understanding of the channels through which disasters impact well-being and direct policymakers to take effective preventive and remedial measures toward disasters; however, the literature contains very few analyses and discussions on the mechanisms. Carroll et al. (2009) indirectly conclude that an Australian drought may affect SWB through agricultural production and related economic factors because only a spring drought has a significant impact on the SWB of rural residents. By comparing the results with potential channel variables being controlled to those not being controlled, Rehdanz et al. (2015) indirectly infer that disasters do not seem to affect SWB through health, income, or employment.
To fill these gaps in the literature, our study makes two contributions. First, we employ a DID approach to identify the causal effects of the Wenchuan earthquake on SWB. This approach is widely used to estimate the impact of a shock that affects some people over a period of time but not others. The causal effect can be estimated by specifying the multiple dimensions of the shock, provided that SWB in the affected areas shares the same trend as SWB in the unaffected areas had the shock never happened. One dimension of the DID strategy is timing—before versus after the year of the earthquake (2008); the other dimension is location—unaffected provinces versus Sichuan province, that is, the location of the earthquake’s epicenter in Wenchuan county. The strategy is implemented with a pooled individual-level cross-sectional dataset that is nationally representative and covers two waves before the earthquake and four waves after the earthquake in 2008. The rich dataset allows us to test the parallel-trend assumption formally and explore the long-run well-being effect, which has been largely ignored by other studies on the Wenchuan earthquake. Next, we delve into a concrete mechanism analysis. Following the paradigm of Cantril (1965) and Easterlin (2010), we investigate five likely channels through which disasters impact SWB, including income, health, family, jobs, and social capital.
We find that, compared to residents in the unaffected areas, Sichuan residents, specifically rural populations, have experienced a significant decline in SWB since the Wenchuan earthquake and that this negative effect was maintained for nearly 10 years. The quantified compensation variation of well-being loss is 67% of the average equivalized household income—similar to the amount estimated in the Fukushima earthquake. Regarding the mechanisms, economic status (income and jobs) and physical health may not play critical roles, probably due to the offsetting effect of the post-disaster relief and recovery plans. Family destruction caused by the earthquake, which cannot be mitigated through outside help, appears to be a major channel through which SWB is reduced.
This paper is structured as follows. Section 2 offers an overall review of the Wenchuan earthquake, the damage caused by the earthquake, and the reconstruction process. Our theoretical framework is also included in this section. Section 3 introduces the data and the empirical strategy. Section 4 presents the results. Section 5 reports the quantified income equivalence of well-being losses. Section 6 provides concluding observations.
Background and Theory
Wenchuan Earthquake
A catastrophic M8 earthquake hit Sichuan Province in China on May 12, 2008. It is named after its epicenter, Wenchuan county (Fig. 1). China’s National Earthquake Relief Headquarter (2008) reported that 46 million residents were affected and approximately 1.5 million victims were housed in temporary settlements until September 2008. Wenchuan earthquake ultimately left 87,476 people either dead or missing, and 374,643 injured (International Disaster Database, 2008).Fig. 1 Sichuan province and its neighboring provinces.
Source Own representation. Wenchuan earthquake mostly impacted the Sichuan province, where the epicenter was located. The provinces indicated by the hatched pattern—Shaanxi, Gansu, and Chongqing—were partially affected by the earthquake. The 10 provinces in gray, located in central and western China, are at a similar level of development as Sichuan
In addition to the numerous casualties, livelihoods were severely impacted, particularly in the rural areas. Many towns and villages, such as Beichuan county, were almost razed to the ground. An estimated 200 million square meters of rural housing was seriously damaged (State Council of China, 2008). The destruction of agricultural infrastructure impacted over one million farmers and destroyed their source of income. Thousands of mu1 of farmland were destroyed, more than 50,000 greenhouses collapsed, and over 20,000 units of agricultural machinery were damaged (China’s Ministry of Agriculture and Rural Affairs, 2008). According to Chen and Hu (2009), an additional 370,000 individuals were left unemployed and 51,000 families were left with no source of income. Non-agricultural residents temporarily lost their jobs. Farmers who had been deprived of their land accounted for the greatest proportion of the unemployed, and agricultural activity could only be resumed once production infrastructure had been restored.
An emergency aid program was launched immediately. The Chinese government promptly established an earthquake relief command center and initiated a first-level natural disaster response to link emergency systems at a national, provincial, city, and county level, which avoided potential deficiencies and inappropriate use of local resources due to unclear rescue tasks and fragmented management (Deng et al., 2010). Medical staff immediately began to rescue the injured, and 85% of the medical institutions in the affected areas started to accept victims within 30 min of the earthquake (Zhang et al., 2012). On day one, local health workers from Sichuan and around 16,000 health workers nationwide were deployed to 21 heavily affected counties, 446 towns, and 4185 village health units (Wu et al., 2008); this was accompanied by the joint rescue forces of the army and non-governmental organizations. Emergency rescue efforts continued for half a month: 94% of the total fatalities and 96% of the injured were found during the first two weeks (Zhang et al., 2012).
A grand plan for recovery and reconstruction followed. Governments at all levels had invested 67.2 billion yuan in earthquake relief funds by September 2008 (China’s Ministry of Finance, 2008). The Overall Plan for Post-Wenchuan Earthquake Recovery and Reconstruction was issued in September for 51 seriously affected counties2 where a 1 trillion yuan fund was allocated for population relocation, housing provision, rural rebuilding, industry reconstruction, infrastructure renewal, and ecological restoration. The intensive recovery projects also stimulated economic performance. By 2017, the total local gross domestic product (GDP), local GDP per capita, and household income of the 39 most seriously hit areas in Sichuan had increased to approximately three times their pre-disaster levels (Sichuan Province’s Bureau of Statistics, 2018).
Theoretical Framework
Understanding what elements prominently determine well-being is a premise for understanding the mechanisms through which earthquakes impact SWB. Cantril (1965) has shown that, despite different development levels and cultures, residents from 14 countries share similar considerations for happiness, including economic, health, family, work, and social aspects. Easterlin (2010, pp. 171–172) proposes a framework where life happiness is the net outcome of satisfaction in multiple life domains, reflecting both objective circumstances and subjective aspirations, and the four most influential domains are material conditions, family life, health, and work. In accordance with the two frameworks, five essential life domains—income, health, family, jobs, and social capital—are treated as the major determinants of well-being.
Life domains can be influenced simultaneously by the shock of an earthquake and the post-disaster recovery. Earthquakes have a significant impact on victims’ income, work situations, health status, and family structure. Life satisfaction may be negatively affected by loss of property, deteriorating living conditions, injuries and trauma, and family dissolution (Kun et al., 2010). Victims may also experience social disconnection due to the loss of acquaintances or social tension due to the competition for scarce resources in the reconstruction stage (Li et al., 2015; Okuyama & Inaba, 2017; Wang et al., 2000). Post-disaster relief and restoration measures may to some extent offset the negative consequences of earthquakes, and this might be a reason why, for example, Rehdanz et al. (2015) find that the Fukushima disaster had not affected SWB through health, income, or employment. Nevertheless, recovery policies may not fully mitigate family dissolution caused by a disaster; moreover, post-earthquake social networks may be strengthened because survivors need to cooperate to get over difficulties (Zhao, 2013).
Equation (1) summarizes our theoretical framework. H represents a happiness function, W is an earthquake indicator, D is the objective outcome of various life domains, and A indicates a list of individual attributes. An earthquake, along with individual characteristics, reshapes objective outcomes of life domains and further influences well-being. Concurrently, individual characteristics appear in the happiness function to approximate subjective aspirations in order to capture that the association between well-being and objective outcomes of life domains could vary by individual characteristics. 1 HW,A=HDW,A,A
The empirical analysis will first spotlight a reduced form equation of well-being on the Wenchuan earthquake and individual attributes, and then test, based on Eq. (1), through which life domains the earthquake influenced happiness.
Data and Methodology
Data Description
We use data from the China General Social Survey (CGSS), one of China’s earliest nationally representative continuous surveys jointly launched by the Renmin University of China and Hong Kong University of Science and Technology. The CGSS has completed 10 waves (2003, 2005, 2006, 2008, 2010, 2011, 2012, 2013, 2015 and 2017) that cover at least 100 counties/districts nationwide and accumulate a rich pooled cross-sectional dataset. Data at individual, family, community, and society levels are comprehensively collected to reflect social trends.
Our sample comprises two pre-earthquake waves (2005 and 2006) and four post-earthquake years (2010, 2012, 2015, and 2017). Data from the years 2003, 2008, 2011, and 2013 are excluded because the 2003 wave covers only urban areas, the 2008 and 2011 waves have insufficient observations for Sichuan, and the 2013 wave disproportionally oversamples younger birth cohorts in Sichuan. Other than Sichuan, observations from 13 provinces in central and western China (Fig. 1) are used to construct the comparison group. Five provinces in the region (Xinjiang, Tibet, Qinghai, Ningxia, and Hainan) are removed because they are not included in all the waves of our sample. The exclusion of the five provinces should not affect our sample’s representativeness as they account for a small proportion of China’s total population. The CGSS only investigates adults (18 years old and above). We also exclude individuals who failed to report SWB and key characteristics, including gender, ethnicity, age, education, and location of residence. The full sample size for the main analysis is 32,205 adults with about 500–600 observations in Sichuan and approximately 4,300–5,000 observations in other provinces for each year.
SWB, the dependent variable, is derived from a self-reported happiness question in which respondents were asked “Overall, do you think your life is happy.” Answers are based on five response options ranging from “very unhappy” to “very happy.” We measure SWB on a 1 to 5 scale with a larger number representing greater happiness. Self-reported happiness has been proven as a valid measurement (Helliwell et al., 2012) and been widely used in studies, including studies in China (e.g., Morgan & Wang, 2019).
Life domains, the mechanism variables through which the earthquake may have affected well-being, are measured as follows. Income is measured by the equivalized household income which divides the household annual total income by the equivalent household size. The household total income is adjusted to the level in 2005 using annual province-based consumer price indices. The equivalent household size assigns 1 to the major respondent, 0.5 to other family members aged 14 and above, and 0.3 to members aged below 14, according to the modified OECD equivalence scale. The income variable is in the form of natural logarithm. Employment status is a dummy variable with 1 indicating employed and 0 unemployed and those not in the labor markets. Health is a self-evaluated perception of health rated from 1 (very unhealthy) to 5 (very healthy). Another category, “extremely healthy,” appears in the 2005 wave and is merged with “very healthy.” Family dissolution is measured by two indicators. The divorce dummy takes 1 for divorcees and 0 otherwise; the widowhood indicator assigns 1 to widows and 0 to others. Bjørnskov (2006) points out that social networks, social trust, and social norms are essential facets of social capital. Due to limitations in data availability, we only measure social network based on a survey question that records “frequency of contact with your friends/relatives,” where responses range from 1 (very rare) to 5 (very close). The social capital variable is available for all years except 2006.
Some key personal characteristics that may affect SWB are also considered (Dolan et al., 2008), including gender, age, ethnicity (Han or non-Han), education level (primary school and below, middle school, or high school and above), and location of residence (urban or rural). A few additional individual variables (hukou status,3 party membership, and household size) and provincial features (population size, gender and age composition, and education structure, collected from China Statistical Yearbooks) are included for the robustness checks (see Table A1 in the online appendix for variable descriptions). Table 1 presents the mean, standard deviation, minimum, maximum, and total observations for the dependent variable, mechanism variables, and major control variables of our analysis.Table 1 Descriptive statistics of the main variables
Variables Mean S.D. Min. Max. N
Dependent variable
SWB (1–5)a 3.61 0.86 1 5 32,205
Mechanism variables
Income (natural logarithm of equivalized household income, yuan)b 8.92 1.11 1.64 15.80 28,920
Health (1–5, “very unhealthy” to “very healthy”)c 3.51 1.14 1 5 28,800
Divorce (1 = divorced)d 0.02 0.14 0 1 32,201
Widow (1 = widowed)e 0.08 0.27 0 1 32,201
Job (1 = employed)f 0.65 0.48 0 1 32,139
Social (1–5, “very rare” to “very close”)g 2.51 1.11 1 5 27,147
Control variables
Gender (1 = male)h 0.48 0.50 0 1 32,205
Agei 48.00 15.68 18 103 32,205
Han (1 = Han ethnicity)j 0.90 0.30 0 1 32,205
Middle (1 = middle school)k 0.29 0.46 0 1 32,205
High (1 = high school and above)l 0.25 0.43 0 1 32,205
Urban (1 = urban residents)m 0.47 0.50 0 1 32,205
aThe dependent variable SWB is self-reported happiness ranging from 1–5. For mechanism variables, bIncome is the natural logarithm of the equivalized household income adjusted to the level in the year 2005 based on provinces’ annual consumer price index (CPI); cHealth is a self-evaluated health status (1–5); dDivorce is 1 for divorcees and 0 otherwise; eWidow is 1 for widows and 0 otherwise; fJob is 1 for the employed and 0 for the unemployed and those not in the labor markets; gSocial refers to the frequency of contact with friends and relatives (1–5). For major individual controls, hGender is 1 for male; iAge confirms adult respondents in CGSS; jHan ethnicity; kAttainment of secondary-school education level; lEducational level of high-school or above; and mLiving in an urban area. All data are from the CGSS
Empirical Strategy
We adopt a DID approach to identify the causal effect of the Wenchuan earthquake on well-being. Differentiating post-earthquake and pre-earthquake happiness in Sichuan, the affected province (treatment group) yields a combination of the well-being effect of the earthquake and the natural evolvement of happiness. Using the same procedure for unaffected provinces (control group), only the natural change in SWB is obtained. If the natural shift in happiness in the treatment group is the same as that of the control group when there were no earthquakes—often called the parallel-trend assumption—any additional difference between the two temporal differences would reveal the causal effect of the Wenchuan earthquake on SWB. The DID method allows for unobserved determinants of happiness in different levels in the two groups, and rules out all unobservables that confound the causal effect as long as the parallel-trend assumption holds. The DID approach is implemented with a linear model as set out in Eq. (2), which represents a reduced form model of Eq. (1).2 SWBipt=βSichuanp×Postt+θSichuanp+ρPostt+Xiptγ+uipt=βQuakept+μp+λt+Xiptγ+∈ipt
SWBipt is an individual i’s happiness in the province p in the year t. Sichuanp is a binary variable for Sichuan province, and Postt is a binary variable indicating years after 2008. Xipt is a set of control variables for personal characteristics, including gender, age and squared age, ethnicity, level of education, and the urban indicator for the baseline model and adding more individual and provincial characteristics for the robustness examination. The first half of Eq. (2) presents a basic form of DID with covariates, and the coefficient β, essentially the (conditional) double-difference of SWB, identifies the well-being effect of the earthquake. In the second half of Eq. (2), we define Quakept=Sichuanp×Postt as the indicator for the Wenchuan earthquake. We further extend the DID form to a more general model of two-way fixed effects (TWFE), where μp and λt are province fixed effects and year fixed effects intended to capture location-specific and time-specific characteristics, respectively. The TWFE model, usually regarded as a generalized DID setting, is preferred to the traditional DID form in our analysis because the former is able to capture unobservables more comprehensively.4uipt and ϵipt are the error terms in the two forms. For convenience, we still call the TWFE setting as a DID approach.
Although the DID method allows the treatment and the control groups to have different underlying levels of SWB, the control group should be properly selected so that the two groups are similar enough to support the parallel-trend assumption. That is why we exclude provinces in northeastern, eastern, and southern China as they are far from Sichuan and their well-being trends may differ from the treatment group. Thus, the control group includes only 13 provinces (Fig. 1). For the main analysis, we regard partially affected provinces (Shaanxi, Gansu, and Chongqing, the provinces indicated by the hatched pattern in Fig. 1) as part of the control group. We examine the robustness of the estimates by excluding the three provinces from the sample, or by treating them as a second treatment group.
Other than incorporating the partially affected provinces in the control group, there are at least two reasons our estimates would tend to be within the lower bounds of the true effects. First, changes in population composition and distribution due to the earthquake may lead to an underestimation of the well-being effect. Residents with a lower socioeconomic status (SES) are more likely to live in poorly constructed houses in a fragile geological environment and are therefore more likely to perish in an earthquake. If so, post-earthquake observations in the treatment group tend to indicate higher SES and life satisfaction compared to the pre-earthquake population. Moreover, surviving victims may migrate to unaffected provinces and shrink the well-being gap between the treatment and control groups. During the post-disaster state-led relocation project, most of the 1.5 million victims who were resettled, moved to towns and cities adjacent to their own jurisdictions in Sichuan (State Council of China, 2008; Ge et al., 2010). Bound by the confidentiality rules of the CGSS, we are not able to identify the cities or counties where the interviewees reside; however, using the whole Sichuan province as the treatment group to a large extent avoids the issue created by population relocation.5
Second, spillover and crowding-out effects from Sichuan to other provinces may underestimate the well-being effect of the earthquake. On the one hand, the collapse of Sichuan’s economic system may have negatively impacted the lives and well-being of residents in provinces with close economic ties to Sichuan. On the other hand, massive relief and recovery resources pouring into Sichuan may have crowded out financial support to other provinces and consequently impaired residents’ well-being there.
To summarize, given the miscellaneous confounders, what the study actually derives is the lower bounds of the well-being effects of the Wenchuan earthquake. All the estimates deserve careful interpretation. Nevertheless, as long as we find negative effects, which is the case in our study, we can confirm the destructive nature of the earthquake.
Following standard procedures, we test the parallel-trend assumption with Eq. (3).3 SWBipt=∑t=2005,10,12,15,17βtWavet×Sichuanp+μp+λt+Xiptγ+∈ipt
Wavet is a dummy variable representing whether individuals are in Wavet. Wave2006×Sichuanp is dropped as a reference, and β2005 should be zero if the parallel-trend assumption holds. In practice, a statistically insignificant estimate for β2005 would endorse the assumption. Moreover, β2010 to β2017 estimate the well-being effects of the earthquake in each post-earthquake wave when the parallel-trend assumption is valid, so we are able to examine fully how the effects change in the short- and long-term.
Results
Main Results
Based on Eq. (2), Table 2 shows the ordinary least squares (OLS) estimates of the well-being effects of the Wenchuan earthquake for the full sample and all the rural and urban subsamples. Column (1) displays a negative and statistically significant correlation between well-being and the earthquake after controlling for the year and province fixed effects (FEs).Table 2 Effects of the earthquake on well-being, OLS estimation
Dependent variable: Happiness (1–5)a Full (1) Full (2) Rural (3) Rural (4) Urban (5) Urban (6)
Quakeb − 0.085*** (0.027) − 0.096*** (0.026) − 0.156*** (0.033) − 0.153*** (0.035) − 0.014 (0.028) − 0.023 (0.027)
Male − 0.034*** (0.009) − 0.040** (0.017) − 0.030** (0.013)
Age − 0.015*** (0.003) − 0.009*** (0.003) − 0.021*** (0.003)
Squared age/1000 0.180*** (0.031) 0.122*** (0.034) 0.244*** (0.034)
Han ethnicity − 0.110*** (0.031) − 0.126** (0.044) − 0.045 (0.030)
Middle school 0.163*** (0.015) 0.175*** (0.019) 0.147*** (0.028)
High school and above 0.310*** (0.019) 0.314*** (0.034) 0.300*** (0.031)
Urban 0.004 (0.017)
Year FE Y Y Y Y Y Y
Province FE Y Y Y Y Y Y
Adjusted R2 0.060 0.081 0.069 0.085 0.052 0.077
N1 3677 3677 1986 1986 1691 1691
N 32,205 32,205 17,047 17,047 15,158 15,158
aThe dependent variable is self-reported happiness (1–5); bQuake is 1 for Sichuan after 2008 and 0 otherwise. Major individual controls are included in columns (2), (4) and (6), and year and province fixed effects (FEs) are included in all the columns. N and N1 represent the sample size and number of observations from Sichuan province, respectively. OLS estimation is applied and standard errors, clustered at a province level, are shown in parentheses. *** significance at the 1% level, ** significance at the 5% level
Individual control variables, including male respondents, age and its square, Han ethnicity, middle and high school indicators (primary school and below is the base category), and an urban dummy are added to column (2). The earthquake is associated with a 0.096 reduction in SWB on a scale of 1–5. By turning off the “Quake” indicator in column (2), we can predict that, without the earthquake, the average level of happiness in Sichuan province after 2008 would have been 3.594, increasing by 0.035 from the predicted level before 2008 (3.559), while the earthquake diminished happiness by 0.096, nearly three times the counterfactual increment of well-being. We will further quantify the loss in well-being at an income level in the next section. Our findings on the coefficients of the covariates are consistent with the literature: male respondents and the ethnic majority seem less happy; higher levels of education contribute to greater happiness; and the SWB–age relationship shows a typical U-shaped pattern.
The results summarized in the last four columns imply that rural residents suffered far more from the earthquake than urban populations in terms of well-being, and the overall decrease in SWB after the earthquake is primarily driven by the rural subsample. This is probably because the most stricken areas were in rural counties and because of the greater vulnerability of those living in a rural environment due to fragile housing and underdeveloped infrastructure. For the rest of the study, we will focus on the analysis of the full sample and the rural subsample, and use columns (2) and (4) as our baseline models.
A crucial assumption for DID is the parallel-trend assumption of well-being for the treatment and the control groups before the earthquake’s shock. Taking 2006 as the reference year, Fig. 2 shows the estimates for βt, based on Eq. (3), in the other years with 95% confidence intervals, for both the full sample and the rural subsample. The regression coefficients are shown in Table A2 of the online appendix.Fig. 2 Parallel-trend tests for happiness. Both figures plot estimates for βt from Eq. (3) with 95% confidence intervals, with the covariates being identical to columns (2) and (4) of Table 2, respectively
Both subfigures present statistically insignificant coefficients for the year 2005, implying that the parallel-trend assumption has not been violated at the 5% level.6 Coefficients for 2010, 2012, 2015, and 2017 represent the short- and long-term effects of the earthquake on well-being. Immediate post-disaster measures were sufficiently effective to offset the well-being impact of the earthquake so that a relatively smaller net effect could be observed in 2010. Temporary post-disaster aid gradually faded away while some of the consequences of the earthquake either remained or became even worse (e.g., badly injured family members passed away after several years.); consequently, the net effects in 2012 and 2015 were stronger. Eventually SWB started to recover in 2017. Once again, the well-being net effects for rural residents are more salient across all the years, either because of the greater impact of the earthquake or due to comparatively inadequate post-earthquake recovery measures.
Robustness Analysis
Table 3 shows the robustness of our baseline results. The first set of the robustness analysis, shown in Panel A, focuses on Sichuan’s adjacent provinces (Shaanxi, Gansu, and Chongqing), which were only partially affected by the earthquake. In the baseline models of Table 2, Sichuan is the treatment group and the three provinces along with the others belong to the control group. In columns (1) and (3), we exclude the three adjacent provinces from the analysis and the estimates remain similar. In columns (2) and (4), Sichuan is the first treatment group, the adjacent provinces form a second treatment group, and the rest serve as the control group, so that we can separate the treatment effect of the earthquake in Sichuan from that in the adjacent provinces.Table 3 Robustness analysis
Panel A. Alternative treatment and control groups
Dependent variable: Happiness (1–5) Adjacent provinces excluded Full (1) Adjacent provinces as another treatment Full (2) Adjacent provinces excluded Rural (3) Adjacent provinces as another treatment Rural (4)
Quake (Sichuan) − 0.093*** (0.023) − 0.094*** (0.023) − 0.146*** (0.035) − 0.146*** (0.035)
Quake (adjacent provinces) 0.012 (0.107) 0.049 (0.122)
Baseline covariates Y Y Y Y
Year FE Y Y Y Y
Province FE Y Y Y Y
Adjusted R2 0.086 0.081 0.092 0.085
N1 3677 3677 1986 1986
N2 4835 2748
N 27,370 32,205 14,299 17,047
Panel B. Alternative model specifications
Dependent variable: Happiness (1–5) Additional covariates Full (1) Ordered probit Full (2) Additional covariates Rural (3) Ordered probit Rural (4)
Quake − 0.079** (0.027) − 0.113** (0.042)
Average partial effects of the quake on the probability of happiness being equal to
1 0.003*** (0.000) 0.005*** (0.001)
2 0.012*** (0.003) 0.020*** (0.004)
3 0.027*** (0.006) 0.044*** (0.010)
4 − 0.015*** (0.003) -0.031*** (0.005)
5 − 0.026*** (0.006) -0.039*** (0.009)
Baseline covariates Y Y Y Y
Year FE Y Y Y Y
Province FE Y Y Y Y
Adjusted R2 0.093 0.081 0.097 0.136
N1 3672 3677 1984 1984
N 32,127 32,205 17,006 17,047
The dependent variable is self-reported happiness (1–5). In Panel A, columns (1) and (3) exclude Shaanxi, Gansu, and Chongqing; columns (2) and (4) treat the three provinces as a second treatment group. In Panel B, columns (1) and (3) add covariates, including a hukou dummy, a membership of the Communist Party dummy, household size, provincial population size, provincial cohort of high school students, provincial sex ratio, and old dependency ratio; columns (2) and (4) estimate ordered probit models and present average partial effects of the earthquake on the probability of happiness at specific levels. All the columns include baseline covariates, year fixed effects (FEs), and province FEs as in columns (2) and (4) of Table 2. N, N1, and N2 represent sample size, the number of Sichuan observations, and the number of observations from Shaanxi, Gansu, and Chongqing, respectively. Pseudo-R-squared is reported for ordered probit models. Standard errors, clustered at the province level, are shown in parentheses. *** significance at the 1% level, ** significance at the 5% level
The two coefficients in column (2) and (4) of Panel A are estimates of β1 and β2 in Eq. (4), where Adjacentp is a binary indicator for the three adjacent provinces. The well-being effects of the earthquake in Sichuan are similar to the baseline estimates, while the net impact of the earthquake in Sichuan’s adjacent provinces is statistically insignificant.4 SWBipt=β1Postt×Sichuanp+β2Postt×Adjacentp+μp+λt+Xiptγ+∈ipt
Panel B of Table 3 examines how the results vary according to the model’s specifications. Columns (1) and (3) incorporate additional individual and provincial covariates, including hukou status, membership of the Communist Party, household size, provincial population size, provincial cohort of high school students, provincial sex ratio, and population dependency ratio. The additional covariates are not controlled for in the baseline models because they are not sufficiently exogenous, but adding these covariates could help to eliminate more confounding effects. The coefficients in columns (1) and (3) are slightly weaker but remain robust.
Due to the ordinal nature of SWB, we estimate an ordered probit model and calculate the average partial effects of the earthquake on the probability of happiness being at specific levels, following Puhani’s (2012) framework of DID in non-linear models. Columns (2) and (4) show that the earthquake decreases the probability of being happy (4) and very happy (5), and the estimates are in line with earlier studies (Carroll et al., 2009; Luechinger & Raschky, 2009; Rehdanz et al., 2015). Our results therefore remain robust after a series of examinations.
Mechanism Analysis
Given the detrimental effects of the Wenchuan earthquake on the well-being of victims in Sichuan, we further explore through which channels the earthquake exerted these adverse shocks. Based on our theoretical framework, we probe five mechanisms: income, health, family, jobs, and social capital. By replacing the dependent variable of the baseline models with each mechanism variable, we estimate the function DW,A in Eq. (1) and determine to what extent the earthquake affected various life domains. Panels A and B of Table 4 summarize the mechanisms analyses for the full sample and the rural subsample, respectively. The dependent variables are measures of life domains whose definition and descriptive statistics are described in the data description section. When divorce and widowhood are mechanism variables, we restrict the zero-valued respondents of the two dummies to the currently married or those cohabitating with unmarried partners, to highlight life transitions from being married to divorce or widowhood in the wake of the earthquake. The rest of the regressions are based on all individuals with available information on the mechanism variables. The independent variables are identical to those in the baseline models of Table 2. Table 4 is based on the linear models. The results remain robust if probit models are estimated for divorce, widowhood, and employment status, and ordered probit models are used for health and social capital.Table 4 Mechanisms analysis
Dependent variable: Incomea (1) Healthb (2) Divorcec (3) Widowd (4) Jobe (5) Socialf (6)
Panel A. Full sample
Quake 0.077 (0.048) 0.090 (0.066) 0.015*** (0.005) 0.020*** (0.006) − 0.052** (0.019) 0.179** (0.064)
Baseline covariates Y Y Y Y Y Y
Year FE Y Y Y Y Y Y
Province FE Y Y Y Y Y Y
Adjusted R2 0.348 0.208 0.022 0.253 0.226 0.328
N1 3454 3224 3011 3294 3672 3023
N 28,920 28,800 23,915 26,037 32,139 24,415
Panel B. Rural subsample
Quake 0.017 (0.069) 0.112 (0.095) 0.017*** (0.003) − 0.034 (0.026) 0.203*** (0.066)
Baseline covariates Y Y Y Y Y
Year FE Y Y Y Y Y
Province FE Y Y Y Y Y
Adjusted R2 0.253 0.198 0.012 0.243 0.395
N1 1880 1693 1653 1982 1599
N 15,414 15,167 14,513 17,001 11,546
Measures of the five life domains serve as dependent variables in columns (1)–(6): aIncome is the natural logarithm of the equivalized household income adjusted to the level in the year 2005 based on provinces’ annual consumer price index (CPI); bHealth is self-evaluated health status (1–5); cDivorce is 1 for divorcees and 0 for those currently married or cohabitating with a partner; dWidow is 1 for widows and 0 for those currently married or cohabitating with a partner; eJob is 1 for the employed and 0 for the unemployed and those not in the labor markets; and fSocial refers to frequency of contact with friends and relatives (1–5). Independent variables are identical to the baseline models of Table 2. N and N1 represent the sample size and number of observations from Sichuan province, respectively. Standard errors, clustered at the province level, are shown in parentheses. *** significance at the 1% level, ** significance at the 5% level. FE: Fixed effect
The parallel-trend assumption also needs to hold for the mechanisms analyses so that we are able to interpret all the coefficients in Table 4 as the causal effects of the earthquake on the mechanism variables. In Panel A, columns (1), (2), and (5) satisfy the parallel-trend assumption with the baseline model setting of Table 2. Pre-earthquake trends of divorce and widowhood are not parallel in the baseline setting. We manage to satisfy the assumption by matching each Sichuan individual in a year with similar non-Sichuan residents of the same year before implementing the DID regressions, usually called a DID matching strategy (Smith & Todd, 2005), and report the results in columns (3) and (4).7 In Panel B, columns (1) to (3) and column (5) satisfy the parallel-trend assumption with the baseline setting, and the mechanism variable “Widow” is excluded as the assumption is violated even with the DID matching strategy. Social capital is unavailable in 2006 so the parallel-trend assumption cannot be tested. We apply the DID matching strategy to columns (6) of Panel A and Panel B to make the coefficients as causal as possible.
Income and Employment
The statistically insignificant coefficients in columns (1) of Panels A and B imply that, despite the earthquake’s substantial negative shocks to income, post-disaster recovery policies may have been effective in offsetting these influences. For example, the 18 counties most damaged were paired with 18 of the richest provinces, where the latter were obliged to transfer at least 1% of their previous year’s revenue to the matched counties for three consecutive years (2008–2010; Bulte et al., 2018).
Moreover, the Sichuan government swiftly created over 800,000 jobs to mitigate the shock to employment. Figure 3a presents the parallel-trend test and post-earthquake dynamics of employment based on column (5) of Panel A in Table 4. The statistically insignificant effects in 2010 and 2012 imply the post-disaster policy has essentially undone the short-term impact on jobs, but many informal and temporary jobs may have disappeared in the long run so that the victims suffer from negative shocks on employment in 2015 before they could fully recover with formal jobs in 2017. Column (5) of Panel B implies post-earthquake aid may be stronger in rural areas so that the net effect on rural employment is statistically insignificant. The living allowances and housing subsidies (including loans) granted directly to rural victims, roughly estimated, exceeded 100 billion yuan (Lu et al., 2014). The whole investment package was designed to have a positive impact on employment and income, and limit the depressive effect of the destruction caused by the earthquake (Dunford & Li, 2011).Fig. 3 Parallel-trend tests for mechanism variables. All figures plot estimates of the year-by-year effects of earthquake on mechanism variables, relative to the base year of 2006, with 95% confidence intervals, and covariates being identical to columns (2) and (4) of Table 2, respectively
Another factor that played a role in rural areas is strategic behavior adjustment. In a study of droughts in Australia, Carroll et al. (2009) argue that when the situation were to worsen, there would be out-migration from affected regions, and within the affected regions both the composition of industries and alternative (agricultural) production methods would change. This phenomenon was also observed after the Wenchuan earthquake. Lu et al. (2014) find that due to the loss of agricultural income after the earthquake, farmers migrated to earn income from non-agricultural jobs, which could, to a large extent, offset the income and employment shocks from the earthquake.
The significant employment effect of the earthquake in the full sample could explain why happiness declines, while the insignificant coefficient for the rural subsample is inconsistent with the fact that rural victims have suffered greater well-being losses from the earthquake. Accordingly, we take a cautious attitude towards this result and employment status will not be regarded as a crucial mechanism in the study.
Health
The insignificant effects of the earthquake on health in columns (2) of Panels A and B need to be interpreted with caution. It is possible that emergency medical rescue efforts and post-earthquake treatments (Wu et al., 2008) counteracted the earthquake’s harmful impact on health, but their impact may be understated for a number of reasons. First, it is inevitable in a huge earthquake that some residents may be omitted from a survey because, for instance, they were badly injured, died, or moved to other provinces because of the trauma (Yamamura, 2012). Chen et al. (2009) investigate 1,243 victims in five of the worst hit counties, where the disabled and incapacitated accounted for about 10% of the victims. Second, the health question of the CGSS was changed in the 2010 wave to “How do you experience your body’s health” instead of asking respondents to “Evaluate your health status,” which may have induced people to focus only on physical rather than mental conditions.
Divorce and Widowhood
Columns (3) and (4) of Panel A and column (3) of Panel B show that the earthquake significantly increased the probability of divorce and of being widowed. Becker et al. (1977) indicate that unanticipated events may destabilize marriage. This study, among others (e.g., Martin & Bumpass, 1989; White, 1990), also reveals that the likelihood of divorce is negatively associated with SES. It is possible that the earthquake reduced family members’ SES and contributed to higher divorce rates before their SES could be substantially restored. In addition, the earthquake killed people and resulted in widowed families, and badly injured family members were more likely to pass away a few years later, which would have exacerbated the problem of widowhood. Figures 3b–d present the parallel-trend tests and post-earthquake dynamics of divorce and widowhood based on columns (3) and (4) of Panel A and column (3) of Panel B in Table 4. The effects of the earthquake on divorce and widowhood are most salient in 2012 and 2015, implying that the issues of divorce and widowhood may have emerged gradually after the earthquake, and are consistent with the happiness dip in Fig. 2.
The significantly higher rates of divorce and widowhood caused by the earthquake explain why a loss in well-being could still be detected despite compensation for loss of post-earthquake income and employment. As Easterlin (2010) points out, the formation and dissolution of unions are essential to determine satisfaction with family life. Losing a job or other sources of income could be temporary and repaired by outside forces or personal adjustment, while being separated from life partners may cause long-term trauma that public policies can hardly address. Lucas et al. (2003) report that adaptation to widowhood may take eight years on average; this is consistent with what Figs. 2 and 3c show in the 2017 wave.
Social Capital
Post-earthquake recovery of social capital could even surpass the damages. The positive and significant estimates in columns (6) of Panel A and Panel B indicate that the victims in Sichuan had stronger social ties with friends and relatives than they did before the Wenchuan earthquake. As a result, social capital is unlikely to be a channel through which the earthquake deteriorates happiness.
Not only that, the effect of social capital on SWB remains ambiguous. Yamamura et al. (2015) find that the positive relation between social trust and happiness was strengthened after the Great East Japan Earthquake, which mitigated the disaster shock. Social capital contributes to post-disaster recovery with some evidence from the Wenchuan earthquake as well (Zhao, 2013). However, a few empirical studies (Chen & Meng, 2010; Xin et al., 2009) fail to find significant benefits from social networks, probably because social networks are informal and fragile, especially in the case of victims in rural areas. Our sample presents positive correlations between social capital and happiness (Table A3).
Classic mediation analysis further compares the coefficient of the earthquake with controlled mediators to the coefficient without mediators, but the estimators are usually inconsistent due to the mediators’ endogeneity. We follow the standard procedure of mechanism analysis in the literature examining the causal effect of an event, and verify the linkages between the earthquake and multiple mechanisms empirically (Table 4) while building the bridges between the mechanisms and SWB theoretically (Sect. 2.2). We supplement the evidence in Table A3 by showing how the mechanisms relate to SWB in our sample. The results confirm that all the mechanisms are significantly connected to SWB at the 1% level except that being employed in the rural subsample is significantly correlated with happiness at the 10% level. To summarize, evidence supports that divorce and widowhood may be crucial mechanisms through which the earthquake has affected happiness.
Heterogeneity Analysis
The well-being of populations with different characteristics may be affected differently by the earthquake. We use Eq. (5) to examine how the effects of the Wenchuan earthquake on SWB vary by specific individual characteristics Zipt, which include age, education, and having a basic pension or medical insurance. δ measures the heterogeneous effects of the earthquake on well-being. Xipt contains the baseline covariates as well as Zipt. For the model’s completeness, we also include the interactions between each pair of the variables among Sichuanp, Postt, and Zipt.5 SWBipt=δQuakept×Zipt+βQuakept+θSichuanp×Zipt+ρPostt×Zipt+μp+λt+Xiptγ+∈ipt
Table 5 summarizes the estimates of δ, for both the full sample and the rural subsample. Columns (1) and (2) show that older adults and less educated respondents are more vulnerable to the shock of the earthquake (statistically insignificant for the rural subsample), which is consistent to findings that populations with disadvantaged SES were more vulnerable in terms of SWB during China’s economic transitions in the late 1990s and early 2000s (Easterlin et al., 2012, 2017; Morgan & Wang, 2019). In a study on floods, Luechinger and Raschky (2009) find that compulsory insurance could significantly mitigate the negative impact on SWB. As columns (3) and (4) illustrate, participation in basic pension or medical insurance programs could mitigate the well-being damages caused by the earthquake, except that the estimate in column (3) of Panel A is only statistically significant at the margin (p value = 0.117).Table 5 Heterogeneity analysis
Dependent variable: Happiness (1–5) (1) (2) (3) (4)
Panel A. Full sample
Age*quake − 0.003** (0.001)
Middle school*quake 0.056 (0.036)
High school*quake 0.112** (0.044)
Pension*quake 0.062 (0.037)
Medical insurance*quake 0.294*** (0.041)
Baseline covariates Y Y Y Y
Year FE Y Y Y Y
Province FE Y Y Y Y
Adjusted R2 0.082 0.082 0.059 0.061
N1 3677 3677 2920 2965
N 32,205 32,205 27,152 27,525
Panel B. Rural subsample
Age*Quake − 0.001 (0.002)
Middle school*quake 0.054 (0.055)
High school*quake 0.006 (0.068)
Pension*quake 0.314*** (0.099)
Medical insurance*quake 0.624*** (0.065)
Baseline covariates Y Y Y Y
Year FE Y Y Y Y
Province FE Y Y Y Y
Adjusted R2 0.085 0.085 0.048 0.050
N1 1986 1986 1321 1350
N 17,047 17,047 12,891 13,102
The dependent variable is self-reported happiness (1–5). The quake indicator is interacted with age, education dummies, a pension dummy, and a medical insurance dummy in columns (1) to (4), respectively. Columns (3) and (4) further control for the pension and medical insurance dummies. All the columns include interactions between each pair of the variables among Sichuan dummy, a post-earthquake dummy, and the characteristics used for the heterogeneity analysis. N and N1 represent sample size and the number of observations from Sichuan province, respectively. Standard errors, clustered at a province level, are shown in parentheses. *** significance at the 1% level, ** significance at the 5% level. FE: Fixed effect
Quantifying Loss in Well-Being
When evaluating the total cost of a natural disaster, people tend to gauge it based on the physical damage and neglect the loss in well-being. A natural disaster could be viewed as “public bads” for which there is no tradable market or price, which challenges traditional valuation methods (e.g., hedonic or contingent valuation methods). However, we have the option of employing a life satisfaction analysis (LSA) to assess the intangible costs with our happiness data. Luechinger and Raschky (2009) compare several valuation methods and highlight the advantages of LSA.
Under the hypothesis of rational economic agents, maximal utility is maintained when MUincΔInc=MUquakeΔQuake, other determinants of utility being constant. MUinc and MUquake are the marginal utilities of income and earthquake, respectively. Empirically, utility could be proxied by SWB, as Kahneman et al. (1997) indicate that life satisfaction could be regarded as a well-defined form of experienced utility. The two marginal utilities can thus be obtained by regressing SWB on income and earthquake, conditional on other covariates. We can further acquire the willingness to pay (WTP) or the marginal rate of substitution between income and earthquake (Eq. (6)) to suggest how much income a person would be willing to sacrifice to avoid an earthquake, or, in another words, how much income would compensate a person for experiencing a disaster in terms of SWB.6 WTPMRS=ΔIncΔQuake=MUquakeMUinc
Table 6 estimates marginal utilities with respect to the earthquake and equivalized household income for the full sample and the rural subsample. The income variable is included in the form of natural logarithm to alleviate the influence of outliers and capture the nonlinear relations between SWB and income. Coefficients in column (1) imply that MUinc=0.156/Income. The post-earthquake average equivalized household income (adjusted to the year 2005) of Sichuan residents is 17,259 yuan, and then the overall WTP is approximately 11,617 yuan (0.105*17,259/0.156), which amounts to 67% of the average equivalized household income. The post-earthquake average equivalized household income (adjusted to the year 2005) of Sichuan rural residents is 10,197 yuan, and similarly, the rural WTP is about 9,581 yuan (0.140*10,197/0.149), which is about 94% of the rural average equivalized household income.Table 6 Quantifying the well-being loss with income
Dependent variable: Happiness (1–5)a Full (1) Rural (2)
Quake − 0.105*** (0.028) − 0.140*** (0.037)
Income (natural logarithm of equivalized household income, yuan)b 0.156*** (0.008) 0.149*** (0.011)
Baseline covariatesc Y Y
Year FE Y Y
Province FE Y Y
Adjusted R2 0.110 0.113
N1 3,454 1,880
N 28,920 15,414
aHappiness, the dependent variable, is self-reported happiness (1–5); bIncome is the natural logarithm of the equivalized household income adjusted to the level in the year 2005 based on provinces’ annual consumer price index (CPI); and cOther covariates are identical to the baseline models of Table 2. N and N1 represent sample size and the number of observations from Sichuan province, respectively. Standard errors, clustered at a province level, are shown in parentheses. *** significance at the 1% level. FE: Fixed effect
The WTPs computed for the Wenchuan earthquake are much larger than those for other disasters in the literature. Carroll et al. (2009) state that spring droughts in rural Australia cause an average loss of 38% in annual household income. In a study on floods in Europe, Luechinger and Raschky (2009) conclude that people are willing to spend 24% of their average annual household income to avoid a particular flood. For the Fukushima earthquake, a similar case to the one covered by our study, Rehdanz et al. (2015) find that a person living within a 150–300 km range of the epicenter experiences a reduction in happiness worth 70% of their household’s average annual income, while for a person living within 150 km of the event, it is up to 240%, similar to our estimates.
Conclusions
By utilizing six waves of CGSS data and the DID method, this study is among the first to identify the causal well-being effect of China’s 2008 Wenchuan earthquake. We explored its long-term influence as well as potential mechanisms to quantify the equivalent monetary loss. We detected a significant reduction in SWB among Sichuan residents compared to their neighboring counterparts. This depressive effect was sustained for nearly 10 years, despite the finding that SWB tends to return to the pre-earthquake level in the long-term. We gauged the integrated net outcome of the earthquake per se along with the relief and recovery programs after the disaster, and found that the effect remains robust after a series of checks. As for impact channels, losses concerning the economic aspects (income and jobs) as well as the health of residents appeared to be largely compensated owing to enormous investment in post-earthquake reconstruction. Social capital was strengthened after the earthquake so that SWB benefited from interpersonal connections.
The significantly higher rates for divorce and widowhood could explain why Sichuan residents continued to suffer a deterioration in life satisfaction, particularly once the relief programs were phased out in the medium term. Catastrophic natural disasters, such as huge earthquakes, impact victims’ well-being through more hidden channels, including, but not limited to, family structure (e.g., family dissolution and less parenting of children because adults have to work outside). While a typical reconstruction program mainly focuses on economic recovery, these “forgotten” corners often fail to attract the attention of policymakers. Yet this aspect of well-being loss should not be disregarded, for the results of our analysis estimate its monetary equivalence to be 67% of the average equivalized household income. Due to the limitation of our datasets, crucial life domains through which the earthquake could affect happiness may not be measured accurately and comprehensively. Further research is needed to examine more potential mechanisms.
Prominent inequality was found in the well-being impact of the earthquake. Rural populations were more severely hit, such that the income equivalent loss in SWB was as high as 94% of the rural equivalized household income. Older adults, the less educated, and those without social insurance are more vulnerable to earthquake shocks. Post-disaster policies should focus on disadvantaged groups to help them recover as much as possible.
In the context of the coronavirus disease 2019 (COVID-19) pandemic, our study concerning the well-being effect of natural disasters sheds some light on the design of practical programs for disaster prevention and recovery. The content of relief efforts and their cost–benefit analyses should not simply concentrate on economic aspects. It is also necessary to deliberate further on what public policies can do for affected populations—often permanently scarred—particularly when intensive financial aid is phased out. Variations in human welfare due to natural disasters always call for deeper academic research. At a grassroots level, long-term assistance with more comprehensive humanitarian features during the post-disaster stage should be elaborated, especially for those from a disadvantaged socioeconomic background.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 27 KB)
Funding
This work was supported by the “Thematic Research Project on China’s Income Distribution” (No. 21XNLG03) with funding from the research fund of Renmin University of China.
Declarations
Conflicts of interest
The authors declare that they have no conflict of interest.
Ethical Approval
This research was conducted in accordance with the ethical guidelines of the local university and with the Ethical Standards of the 1964 Declaration of Helsinki.
Informed Consent
Informed consent was obtained from all participants.
1 Mu is a commonly used area unit for measuring agricultural land in China: 1000 mu is approximately 0.67 km2.
2 Based on the intensity of the earthquake and the extent of the damage, official assessments classified 237 counties into three types: severely affected areas (10), strongly affected areas (41), and moderately affected areas (186). A full list is available at http://www.gov.cn/zwgk/2008-07/22/content_1052835.htm.
3 Hukou is a China-specific household registration and population management system. A person is registered as either an agricultural or non-agricultural resident.
4 In a two-group and two-period setting, TWFE is equivalent to DID. When the treatment is binary and the design is staggered, as in our case, the TWFE coefficient β is a weighted summation of multiple two-group two-period DID estimates (De Chaisemartin & d’Haultfoeuille, 2022). Recent studies have found that TWFE models may suffer from estimation biases caused by “forbidden comparison.” However, in a scenario with binary treatment, staggered design, and unique treatment timing, TWFE is exempt from the problem (De Chaisemartin & d’Haultfoeuille, 2022). Our case satisfies all the conditions and therefore TWFE is a reasonable choice.
5 Although 141 out of the 181 counties in Sichuan were affected by the earthquake, there were also a number of unaffected residents in the western part of Sichuan that could not be ignored. Consequently, regarding the entire Sichuan as the treatment group may still have understated the earthquake’s effect.
6 The p value for the 2005 coefficient in the rural subsample is 0.072, indicating that the parallel-trend assumption is marginally violated. At the 10% level, the upward pre-trend indicates that the rural SWB in Sichuan province would have grown faster than that of other provinces had the earthquake not happened. As a result, the dynamic effects of the earthquake on happiness in Fig. 2b are likely the lower bounds of the true shocks, implying that the earthquake may have impacted the rural SWB more heavily than what the subfigure has revealed.
7 We calculate the propensity scores of being treated (being in Sichuan province) for each individual in each year, using a logit model on gender, age, squared age, ethnicity, education, and place of residence. Then we match each Sichuan individual in a year with non-Sichuan residents of the same year having similar propensity scores using the method of kernel matching. We further confirm that the parallel-trend assumption still holds for happiness with the DID matching strategy, and the post-earthquake dynamics of happiness are similar to those presented in Fig. 2.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Becker GS Landes EM Michael RT An Economic Analysis of Marital Instability Journal of Political Economy 1977 85 6 1141 1187 10.1086/260631
Berlemann M Does hurricane risk affect individual well-being? Empirical evidence on the indirect effects of natural disasters Ecological Economics 2016 124 99 113 10.1016/j.ecolecon.2016.01.020
Bjørnskov C The multiple facets of social capital European Journal of Political Economy 2006 22 1 22 40 10.1016/j.ejpoleco.2005.05.006
Bulte E Xu L Zhang X Post-disaster aid and development of the manufacturing sector: Lessons from a natural experiment in China European Economic Review 2018 101 441 458 10.1016/j.euroecorev.2017.10.019
Cantril, H. (1965). Pattern of human concerns.
Carroll N Frijters P Shields MA Quantifying the costs of drought: New evidence from life satisfaction data Journal of Population Economics 2009 22 2 445 461 10.1007/s00148-007-0174-3
Caruso G Miller S Long run effects and intergenerational transmission of natural disasters: A case study on the 1970 Ancash Earthquake Journal of Development Economics 2015 117 134 150 10.1016/j.jdeveco.2015.07.012
Caruso The legacy of natural disasters: The intergenerational impact of 100 years of disasters in Latin America Journal of Development Economics 2017 127 209 233 10.1016/j.jdeveco.2017.03.007
Chen S Hu AG Study on the impact of the May 12 Wenchuan earthquake on the affected residents Academia 2009 02 7 20
Chen S Meng Q A study on human capital, social capital and disaster recovery of the affected residents-exemple from the recovery from Wenchuan earthquake Economist 2010 5 94 101
Chen S Meng QG Hu AG An Empirical Study on the main needs and related characteristics of people affected by the Wenchuan earthquake Academia 2009 05 17 29
China's Ministry of Finance (2008). Retrieved January 10, 2022, from http://www.china.com.cn/zhibo/zhuanti/2008-09/05/content_16392500.htm
China's National Earthquake Relief Headquarter (2008). Retrieved January 10, 2022, from http://www.china.com.cn/zhibo/zhuanti/2008-09/05/content_16392500.htm
China's Ministry of Agriculture and Rural Affairs (2008). Retrieved January 10, 2022, from http://www.agri.cn/
De Chaisemartin, C., & d'Haultfoeuille, X. (2022). Two-way fixed effects and differences-in-differences with heterogeneous treatment effects: A survey (No. w29691). National Bureau of Economic Research.
Deng S Zheng S Shi Y Applying lessons from China’s Wenchuan earthquake to medical rescue following the Yushu earthquake Journal of Evidence-Based Medicine 2010 3 2 62 64 10.1111/j.1756-5391.2010.01084.x 21349046
Dolan P Peasgood T White M Do we really know what makes us happy? A review of the economic literature on the factors associated with subjective well-being Journal of Economic Psychology 2008 29 1 94 122 10.1016/j.joep.2007.09.001
Dunford M Li L Earthquake reconstruction in Wenchuan: Assessing the state overall plan and addressing the ‘forgotten phase’ Applied Geography 2011 31 3 998 1009 10.1016/j.apgeog.2011.01.001
Easterlin RA Happiness, growth, and the life cycle 2010 Oxford University Press
Easterlin RA Morgan R Switek M Wang F China’s life satisfaction, 1990–2010 Proceedings of the National Academy of Sciences 2012 109 25 9775 9780 10.1073/pnas.1205672109
Easterlin RA Wang F Wang S Helliwell J Layard R Sachs J De Neve J Huang H Wang S Growth and happiness in China, 1990–2015 World happiness report 2017 2017 Sustainable Development Solutions Network 48 83
EM-DAT Public, the International Disaster Database, https://public.emdat.be/
Ge Y Gu Y Deng W Evaluating China’s national post-disaster plans: The 2008 Wenchuan earthquake’s recovery and reconstruction planning International Journal of Disaster Risk Science 2010 1 2 17 27
Gignoux J Menéndez M Benefit in the wake of disaster: Long-run effects of earthquakes on welfare in rural Indonesia Journal of Development Economics 2016 118 26 44 10.1016/j.jdeveco.2015.08.004
Helliwell, J. F., Layard, R., & Sachs, J. (2012). World happiness report [2012].
Kahneman D Wakker PP Sarin R Back to Bentham? Explorations of experienced utility The Quarterly Journal of Economics 1997 112 2 375 406 10.1162/003355397555235
Kimball, M., Levy, H., Ohtake, F., & Tsutsui, Y. (2006). Unhappiness after hurricane Katrina. National Bureau of Economic Research.
Kun P Wang Z Chen X Le H Gong X Zhang L Yao L Public health status and influence factors after 2008 Wenchuan earthquake among survivors in Sichuan province, China: Cross-sectional trial Public Health 2010 124 10 573 580 10.1016/j.puhe.2010.06.001 20833404
LaJoie AS Sprang G McKinney WP Long-term effects of Hurricane Katrina on the psychological well-being of evacuees Disasters 2010 34 4 1031 1044 10.1111/j.1467-7717.2010.01181.x 20572850
Li H-F Kuang W-H Yang H-Q Subjective well-being in elderly survivors after the 5· 12 Wenchuan earthquake Chinese Journal of Gerontology 2009 9 42
Li L Ruijie Li Yandong Z Transformation of natural disaster and rural social relations: A case study of a village in Wenchuan earthquake-affected area (J) Social Science Front 2015 1 190 200
Lu JL Feng SZ Ai CR The impact of natural disasters and government assistance on Farmers' income and consumption: Experience from the Wenchuan earthquake Economics (quarterly) 2014 13 02 745 766
Lucas RE Clark AE Georgellis Y Diener E Reexamining adaptation and the set point model of happiness: Reactions to changes in marital status Journal of Personality and Social Psychology 2003 84 3 527 10.1037/0022-3514.84.3.527 12635914
Luechinger S Raschky PA Valuing flood disasters using the life satisfaction approach Journal of Public Economics 2009 93 3 620 633 10.1016/j.jpubeco.2008.10.003
Martin TC Bumpass L Recent trends in marital disruption Demography 1989 26 1 37 51 10.2307/2061492 2737357
Morgan R Wang F Well-being in transition: Life satisfaction in urban China from 2002 to 2012 Journal of Happiness Studies 2019 20 8 2609 2629 10.1007/s10902-018-0061-5
Ohtake F Yamada K Yamane S Appraising unhappiness in the wake of the Great East Japan earthquake The Japanese Economic Review 2016 67 4 403 417 10.1111/jere.12099
Okuyama N Inaba Y Influence of natural disasters on social engagement and post-disaster well-being: The case of the Great East Japan Earthquake Japan and the World Economy 2017 44 1 13 10.1016/j.japwor.2017.10.001
Puhani PA The treatment effect, the cross difference, and the interaction term in nonlinear “difference-in-differences” models Economics Letters 2012 115 1 85 87 10.1016/j.econlet.2011.11.025
Rehdanz K Welsch H Narita D Okubo T Well-being effects of a major natural disaster: The case of Fukushima Journal of Economic Behavior and Organization 2015 116 500 517 10.1016/j.jebo.2015.05.014
Sichuan Province's Bureau of Statistics (2018). Retrieved January 10, 2022, from https://www.sc.gov.cn/10462/10464/10797/2018/5/8/10450386.shtml
Smith JA Todd PE Does matching overcome LaLonde's critique of nonexperimental estimators? Journal of Econometrics 2005 125 1–2 305 353 10.1016/j.jeconom.2004.04.011
Strobl E The economic growth impact of natural disasters in developing countries: Evidence from hurricane strikes in the Central American and Caribbean regions Journal of Development Economics 2012 97 1 130 141 10.1016/j.jdeveco.2010.12.002
Tao S Wang F Yan XU Jian LI Fang LUO Sheng-Nan Z The change of teacher’s subjective well-being after the Wenchuan earthquake and mediator analysis Advances in Psychological Science 2009 17 588
The State Council of China (2018). Planning group of post-Wenchuan earthquake recovery and reconstruction. 2008. The Overall Planning for Post-Wenchuan Earthquake Recovery and Reconstruction. Beijing, China. Available at http://www.gov.cn/gongbao/content/2008/content_1120416.htm
Wang X Gao L Zhang H Zhao C Shen Y Shinfuku N Post-earthquake quality of life and psychological well-being: Longitudinal evaluation in a rural community sample in northern China Psychiatry and Clinical Neurosciences 2000 54 4 427 433 10.1046/j.1440-1819.2000.00732.x 10997859
White LK Determinants of divorce: A review of research in the eighties Journal of Marriage and the Family 1990 52 4 904 10.2307/353309
Wu XP Fang G Tang XF Dong YC Zhang JX Lin L Epidemic prevention after Wenchuan earthquake in Sichuan province Chinese Journal of Evidence-Based Medicine 2008 8 803 809
Xin J-L Wu S-T Wu K-K Wang W-Z Zhang J-X Social support system of people in Sichuan earthquake area and its relationship with subject well-being Advances in Psychological Science 2009 17 03 532 536
Yamamura, E. (2012). Natural disasters and their long-term effect on happiness: The case of the great Hanshin-Awaji earthquake. MPRA Paper 37505, University Library of Munich, Germany.
Yamamura E Tsutsui Y Yamane C Yamane S Powdthavee N Trust and happiness: Comparative study before and after the Great East Japan Earthquake Social Indicators Research 2015 123 3 919 935 10.1007/s11205-014-0767-7
Zhang L Liu X Li Y Liu Y Liu Z Lin J Liang W Emergency medical rescue efforts after a major earthquake: Lessons from the 2008 Wenchuan earthquake The Lancet 2012 379 9818 853 861 10.1016/S0140-6736(11)61876-X
Zhao, Y. (2013). Social networks and reduction of risk in disasters: An example of the Wenchuan Earthquake. Economic stress, human capital, and families in Asia (pp.171–182). Springer.
| 0 | PMC9741708 | NO-CC CODE | 2022-12-14 23:29:57 | no | J Happiness Stud. 2022 Dec 11;:1-25 | utf-8 | J Happiness Stud | 2,022 | 10.1007/s10902-022-00609-z | oa_other |
==== Front
Early Child Educ J
Early Child Educ J
Early Childhood Education Journal
1082-3301
1573-1707
Springer Netherlands Dordrecht
1431
10.1007/s10643-022-01431-1
Article
Assessment of Parent–Teacher Relationships in Early Childhood Education Programs During the COVID-19 Pandemic
http://orcid.org/0000-0001-6943-9175
Keengwe Grace [email protected]
1
Onchwari Ariri [email protected]
2
1 grid.266862.e 0000 0004 1936 8163 Department of Teaching, Leadership and Professional Practice, University of North Dakota, Grand Forks, ND USA
2 grid.266744.5 0000 0000 9540 9781 Department of Education, University of Minnesota Duluth, Duluth, MN 55803 USA
11 12 2022
113
16 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.
Relationships between families and schools are important in the educational experiences of young children. However, the COVID-19 pandemic that began in 2019 and spread rapidly around the world disrupted many families, teachers, early childhood programs, and other child-support institutions. There is much to be learned on how this pandemic specifically affected parent–teacher relationships. This study examined whether parent, teacher and other program characteristics had an impact on early childhood parents’ ratings of the quality of their relationships with teachers. Results suggest that parent’s education, income, age of child, location of the center and distance learning availability, impacted how parents perceived their relationships with teachers. Supporting parents’ home environments may be one effective strategy for promoting positive relationships between parents and teachers in times of challenges.
Keywords
Parents
Early childhood teachers
Parent–teacher relationships
Childcare centers
Early childhood
COVID-19
==== Body
pmcIntroduction and Review of Literature
It is an established fact that parent–teacher relationships are a critical and integral part of early childhood education. The benefits of a strong parent–teacher relationship on children’s learning and development are numerous and long lasting (Lang et al., 2013). During a crisis this relationship becomes more significant, as the parent–teacher duo often scrambles to pool resources at their disposal together to support children. During the COVID-19 pandemic, all humanity was put in survival mode with people leaning into their own silos for self-protection. Many support systems, including parent–teacher relationships, were thrown into disarray (Annat & Gassman-Pines, 2020; Karpman et al., 2020). The impact of COVID-19 is still felt and remains under investigation for years to come. This research seeks to specifically investigate the pandemic’s impact on the parent–teacher relationship.
The COVID-19 pandemic had a colossal impact on all spheres of human life at micro and macro levels. The initial impact on the early childhood field has already been explored and documented by a growing number of studies (for e.g., Wei et al., 2021; Hae Min et al., 2021). All aspects of children’s development were impacted. For instance, children’s physical activity, nutrition (Lafave et al., 2021), socio-emotional wellbeing, and cognitive stimulation (Egan et al., 2021) were adversely compromised. Egan et al. (2021) also reported increased maladaptive behaviors like tantrums, clinginess, and boredom. Even though the government’s responses to the crisis created by COVID-19 to support young children was commendable, it was far from adequate (Hae Min et al., 2021).
Following the onset of the pandemic’s lockdown, it took a few months for programs to enter into an adaptation mode (McKenna et al., 2021). Programs were compelled to shift their resources from the regular programing to the safety of children and families. A national survey of over 5000 childcare providers conducted by NAEYC in July 2020 further demonstrates the dilemmas programs were thrown into. For instance, NAEYC found that 40% of childcare providers reported that they would close their programs permanently if they did not receive additional public assistance. Also, many of the programs that were open reported enrollment going down by an average of 67%. A whopping 70% of the programs experienced substantial extra costs in areas such as staff, cleaning supplies, and personal protection equipment (NAEYC, 2020). Delivery of education programming was a challenge. Communication with all families, access to technology e.g., devices and internet, and assessing and screening children, were impossible tasks to perform adequately (McKenna et al., 2021). The pandemic inadvertently challenged programs to rethink delivery of educational experiences and professional development trainings to support teachers to be effective in delivering early childhood programming remotely (Crawford et al., 2021).
The pandemic’s specific impact on teachers was vast. NAEYC (2020) found that very many teachers lost their jobs while others’ working hours were reduced. Teachers had challenges in delivering play-based developmentally appropriate educational experiences online, supporting parents to engage their children, or even contacting parents all together. They were also faced with limited teacher-collaboration opportunities due to reduced staff. All these challenges created considerable stress and anxiety in them. They reported feeling sad, restless, worthless, nervous, hopeless and overall decreased levels of wellbeing (Swigonskil et al., 2021). Eadie et al. (2021) found that stronger professional wellbeing, especially among senior or more experienced staff, was associated with less conflicts in educator-child relationships and lowered risk of staff turnover more. Ironically, educators working remotely were more likely to report a lower level of stress than those working with children at childcare centers (Elek & Church, 2021).
Meeting the needs of children with special needs demanded a higher level of engagement from the teachers. They (teachers) spent more time providing support to families and improvising ways to assess children (Steed & Leech, 2021). To add more stress on teachers, there were high turnover rates among child protective service workers. This represented both the loss of important support services and the development of a working relationship that could support the process of noticing and identifying children in need (Toros et al., 2021).
Preservice teacher preparation programs were also impacted. Opportunities for field-based learning had to be improvised, reduced, or eliminated altogether. This led to feelings of inadequacy among preservice teachers as they entered the workforce (Callaway-Cole & Kimble, 2021).
Swigonskil et al. (2021) reference the age-old societal issue of underpaying early childhood educators prior to the pandemic. Teacher pay went to newer lows, making it almost impossible for teachers to afford basic needs like food, utilities and rent and/or mortgage payments.
Despite all these challenges. Wei et al. (2021) reported that many early childhood teachers remained optimistic about their occupation during the heights of the pandemic period. Elek and Church (2021) also found that the existence of some organizational structures supporting professional wellbeing also mitigated the impact of the pandemic on some early childhood professionals. Some of the research done on remote learning delivery during the pandemic (e.g., McKenna et al., 2021) found that even with the limited guidance educators got from administrators, most adapted successfully to remote instruction, with confidence of educators growing over time.
Like teachers, parents were faced with more responsibilities for their children. For the most part, teachers relied upon families to deliver educational activities. The online platform teachers had to use to deliver educational materials to parents was uncharted and intimidating territory (Crawford et al., 2021). Families were also adjusting to public health measures and other changes in their work while supporting all their children’s learning from home. They also had to adjust to changes in routines that required a great deal of time and effort. It was difficult for families to find time to educate their children at home (Kimil et al., 2021). Consequently, parents’ depression went up and their quality of co-parenting declined (Feinberg et al., 2021).
Educators estimated that the parents who used their emergency childcare service presented either high (37.7%) or average (32.2%) levels of wellbeing. The factors that educators identified as facilitating their interactions with families included parental recognition of their work (11.68%) and having direct contact with them (12.62%) (Bigras et al., 2021).
Family income and education continues to be a strong variable in determining children’s developmental outcomes across the Prek-16 spectrum. In pre-pandemic research, Iruka et al. (2011) found that parent–teacher relationships of kindergarten children were stronger for high income families than those of the low-income. Some studies have found that this disadvantage was amplified during COVID-19 pandemics. For instance, Kimil et al. (2021) found that parents in the low income and low education spectrums were less likely to have educational resources and learning activities occuring at home.
Additionally, during the pandemic families with low income were more impacted in areas of employment like job loss, reduced pay, and non-remote working as many were over-represented in the essential workers category (Annat & Gassman-Pines, 2020; Karpman et al., 2020). This further amplified challenges in childcare responsibilities given that schools were closed (Kerr et al., 2021). Due to limited resources in these households, there was difficulty in arranging child-care, leading to many parents being forced to stay at home from work to take care of children (Karpman et al., 2020). Further, most of them were less likely to receive COVID-19 governmental support (Annat & Gassman-Pines, 2020), adding to more stress in such a time of uncertainty.
Families or homes that had pre-pandemic chaos such as destabilized routines, disorganization, instability, noise, and crowding, and low income experienced elevated depression levels during the pandemic (Johnson et al., 2022).
McKenna et al. (2021) found that many early childhood administrators made gains in various family support areas including regular communications, accessibility of technology tools and internet, assessments, and screening of children, etc. However, Mckenna et al.’s study did not examine socio-economic status or parent education as variables.
One of the variables the current study is examining is parent–teacher relationship differences with regards to the age of the children during the COVID-19 pandemic (infants, toddlers, and preschoolers). Pre-pandemic literature is clear about the fact that, due to young children’s limited self-advocacy and communication skills, reliance on the teacher–child relationship is arguably more critical (Lang et al., 2016). Unfortunately, most research on parent–teacher relationships has focused on older children (Bernhard et al., 1988; Nzinga-Johnson et al., 2009). Further, there is even much less research focusing on parent–teacher relationships of infants and toddlers (Lang et al., 2016). This study therefore also examines how parent–teacher relationships of the various ages of children within 0–5 years, were impacted by COVID-19.
The Family Provider Teacher Relationship Quality Conceptual Model FPTRQ
This study follows the Family Provider Teacher Relationship Quality Conceptual Model FPTRQ (Forry et al., 2012) as its conceptual framework. The FPTRQ places family sensitive care, family support, and parental involvement/engagement at the core of positive relationships. These conceptual perspectives share an ecological view of child development and wellbeing that views parents, families, communities, and programs as interdependent (Forry et al., 2012) with the goal of improving child outcomes. Second, the model argues that even though the three groups/pathways vary by model, all the three are focused on improving outcomes across all domains.
The FPTRQ perspective posits that empowering families and facilitating parental well-being will enhance child development. That children’s learning will be enhanced when focus is placed on supporting parent’s active and direct engagement in their children’s development across all domains. Third, the framework reiterates that all the three perspectives in addition to the focus being on child outcomes, do also impact family-related outcomes such as family self-sufficiency, well-being, and efficacy, and recognize that social and peer support are key elements hypothesized to shape and enhance the ways families care for their children (Forry et al., 2012). The overarching idea of the FPTRQ model is that children are impacted by the programs i.e., teachers and classrooms they attend and by the families and home environment they belong to. Teachers have the role of making sure that these entities interact well to lead to positive children and family outcomes. For example, knowledge of parenting styles can help teachers and parents work together to better serve the child in behavior management. The model is made up of three constructs that were obtained during the development Family Provider Teacher Relationship Quality (FPTRQ) assessment tool that are effective for family teacher’s relationships: family specific knowledge which is knowledge about families, practices with families, and attitudes about families (Porter et al., 2015). All the constructs other than the knowledge have several elements (Porter et al., 2011; Forry et al., 2012).
Family Specific Knowledge
This is defined as the knowledge teachers generally have about the children and families. It includes knowing and understanding children’s family cultures, the contexts they live in, the situations that affect them, knowledge of community resources that are helpful to families, and understanding of families’ abilities, needs and goals. The thinking behind this is that there should be a reciprocal relationship where teachers gather family child -specific knowledge from families and offer families relevant information about themselves. The programs allow providers to be responsive to families and serve as resources to families (Porter et al., 2011).
Teacher Practices
This component examines how teachers collaborate and engage families in their programs. Teacher practices believed to facilitate high quality family provider relationships should have mastery of both relational and goal-oriented skills (Forry et al., 2012). Teachers that have solid relational skills develop positive reciprocal communications, are engaging, sensitive, flexible, and act in responsive ways (Forry et al., 2012). Crucial goal-oriented skills include advocating for families, teaching families to advocate for themselves, connecting families to resources and, families and parents working collaboratively in decision making (Forry et al., 2012, Porter et al., 2011). The teacher practice component is broken down into 4 sub scales; (1) collaboration which examines how teachers collaborate and engage families in the program; (2) responsiveness that assess how teachers are engaging, sensitive and responsive to families’ needs and goals; (3) communication that looks at how teachers promote 2-way communication while respecting parents preferences and also considering their own personal boundaries, and; (4) family focused concerns which assess if teachers communicate with families in ways that show interest; and understanding contexts looks at teacher’s appreciation of broader contexts of situations of families and how those contribute to their development (Porter et al. 2011).
Attitudes
These component beliefs that teachers’ caring, and commitments are to be reflected through the way they show sensitivity in meeting the needs of the children, parents, and families they are working with (Poter et al., 2011). Forry et al. (2012) have identified four attitudes as reinforcers of high quality family provider relationships: respect which examines teachers valuing of children and families- for example do they engage in activities that are welcoming, accepting, nondiscriminatory and non-judgmental; commitment observes how sensitive teachers are to the needs of children, parents and families; openness to change focuses on how willing teachers are to alter their typical practices so as to be sensitive to individual children, parents or family needs; and understanding contexts examines teachers understanding and appreciation of the impact of the different contexts on children and families (Forry et al., 2012).
The model has been used to look at how programs completed the tasks in each of these constructs during COVID-19. Teachers were working with parents in a different environment as most programs went online synchronously and/or asynchronously or closed. During pre-COVID-19, teachers were expected to meet expectations in the FPTRQ measures. For example, teachers were expected to meet with families to discuss children’s learning. At these meetings they shared information about their children’s day at school, resources on parenting or activities for parents to complete with their children at home. The model’s constructs help us understand how teachers were able to develop relationships with parents during these challenging times, and also demonstrates the effectiveness of the model during unique times.
Methods
This study aimed at examining the relationships between teachers and parents of young children in early childhood programs during COVID-19 pandemic. A survey approach was used to gather parent views. Quantitative approaches like surveys are often used to reveal how existing situations look while being objective (Karasar, 1999). This was seen as the appropriate method of data collection given the situation with the COVID-19 pandemic and the limitations of not being able to directly have access to parents like during the typical times that you can drop a questionnaire to go home with them. Two research questions were investigated:How do early childhood education parents perceive parent–teacher relationship practices in their programs during COVID-19?
Were there differences in parent perceptions of parent–teacher relationships based on parent income, education, age of child, distance learning option and program location during COVID-19?
Sample and Sampling
Data was collected from parents with children enrolled in early childhood programs in a midwestern U.S. state. The early childhood programs included in the study comprises Head Start programs, family childcare centers, and private center-based programs. To determine the programs to participate in the study, the Department of Public Instruction (DPI) directors of early learning programs, and Head Start programs in the state were contacted. The directors provided website links with the list of individual Head Start programs and the names of their respective directors. Emails of the Head Start directors were copied and sent the invitation to participate. For all the other early childhood programs in the state, there contact information were searched on the internet, and the information about the study and the Qualtrics survey link sent to the directors to forward to the parents in their programs.
A total of 83 parents completed the survey (Table 1). Parents in the study reported having more girls (52.4%) than boys (42.9%) in their families. It was also determined that 50% of the parents were White, and 27.4% were Native Americans. The majority of the programs were in the urban areas (53.2%) compared to rural areas (46.8%). Most of the parents who participated had children enrolled in private childcare 73.8% compared to those in Head Start 26.3%. 64.3% of the programs had a distance learning option during COVID-19 pandemic. Also 34.5% of the parents had stayed with the same teacher for more than 2 years. Only 8% of the parents had worked with the teacher for less than 6 months. over half (63%) of them had worked with the teachers more than 1 year. 76.2% of the parents had either some college/associate degree or bachelor’s/graduate education. Over 50% of the parents were earning more than $35,000 a year.Table 1 Demographics of parents and children included in the study N = 83
Variables Variable
Number of children N % Family income N %
Boys 36 42.9 0–$34,999 14 16.7
Girls 44 52.4 35,000–44,999 20. 23.8
45,000–54,999 19 22.6
55,000–75,000 27. 32.1
Race
White 42 50
Black/African American 8 9.5
American Indian or Alaska Native 23 27.4
Asian Indian 3 3.6 Program location
Chinese 1 1.2 Rural 37. 46.8
Vietnamese 1 1.2 Urban 42. 53.2
Other Asian 1 1.2 Type of program
Filipino 1 1.2 Head start 21 26.3
Education level Private childcare centers 59 73.8
GED/no high school diploma 16 19
Some college/associate degree 39 46.4
Bachelors/graduate degree 25 29.8 Language used at home
English 51. 60.7
Spanish 18 21.4
Parent experience with provider/teacher English & Spanish 9 10.7
Less than 6 months 8 9.5 English & another language 1. 1.2
6 months–1 year 19 22.6
1–2 years 24 28.6 Age of child
2+ years 29 34.5 0–2 years old 28 33.3
Hispanic or Latino origin 3–4 years old 52 61.9
Yes 26 31
No 53 63.1
No. of program closed during COVID 77
Distance learning option
Yes 54 64.3
No 16 19.0
Data Collection Tool
Parent–teacher relationship quality was measured by the Family and Provider Teacher Relationship Quality (FPTRQ) measure (Porter et al., 2015). This measure asks questions about how parents work with their childcare provider/teacher. It was developed to be used within a wide range of early childhood programs. It has been field tested across race, ethnicity, and class (Kim et al., 2014). The measure consists of 23 items (short form) asking questions related to three broad constructs: teacher’s knowledge of parents they work with, practices they use with parents, and their attitudes towards parents. Means were calculated for each construct as well as for each subscale. The items were scored on a 4-point likert scale with a few items using “yes” or “no” responses. Each construct has a total of 8 various subscales: Family specific knowledge consists of 15 items and has a Cronbach alpha of 0.9, col4. It measures a teacher's knowledge of the families they work with. Collaboration has 11 items with a Cronbach alpha of 0.92, responsiveness has also 11 items with a Cronbach alpha of 0.91, communication consists of 8 items with a 0 0.91, alpha and family focused concern has 3 items with a Cronbach alpha of 0.75. Commitment has 9 items with an alpha of 0.90, understanding context has 4 items with an alpha of 0.97t and respect has 4 3 items with an alpha of 0.83. Mean scores for each construct and total scores are obtained by adding individual items in the subscales after reverse coding the negatively worded items as described by the scoring manual (Alison & Gaviria-Loaiza, 2021). When the short measure was tested, the items were found to have high item response rates and the Cronbach’s alphas for each construct were above 0.6. An example of an item from the family specific knowledge construct is, “How comfortable would or do you feel sharing with your childcare provider your marital status”. An example from the practices construct is, “How often have you met with or talked to your childcare provider or teacher about what to expect at each stage of learning.” And finally, an example from the attitude is “my childcare provider/teacher judges my family because of our culture and values”. During the development of the FPTRQ tool the reliability of all items in the 8 subscales measured showed an acceptable, and mostly good or excellent reliability of above 0.70 (Porter et al., 2015).
Program Closure, Distance Learning and Program Location Variables
These three variables were examined on how they impacted parent–teacher relationships. We investigated the impact of program closure during the COVID-19 pandemic on parent–teacher relationships. In typical settings before COVID-19, programs were building parent–teacher relationships through their daily contacts with parents. The study also explored how distance learning impacted the development of parent–teacher relationships. Distance learning was an important variable to include since programs were either operating remotely, closed, or continued services via distance learning. Third, program location was necessary to investigate because some studies on COVID-19 indicated that rural programs were most affected by the pandemic. The study sought to investigate if there were any differences in parent–teacher relationships based on where the program was located.
Data was collected in the Fall of 2020 after most programs were open in one way or another. The COVID-19 pandemic started in fall 2019 and the shutdown of most programs occurred in spring & fall of 2020. Collecting the data in fall of 2020 captured parent responses better as they had just started getting back to school.
Data Analysis
Descriptive statistics were used to examine the means and frequencies of the parent–teacher relationships constructs and subscales. Independent t tests and one way ANOVA were used to calculate mean differences. It is advised that although the total score of the constructs provides a broad overview of the quality of the teacher/provider and parent relationship, it is better to use subscale scores to correctly identify potential areas for professional development and training (Porter et al., 2015).
Results
Overview
The field test data obtained during the development of the FPTRQ tool provided mean scores that were obtained for each construct and subscales that were used to compare with the scores of parents in this study. This field test data was obtained by assessing parent teacher relationships across different child care centers nationaly. This data was used to determine if the average construct and subscale scores were higher or lower than those obtained in the field study (Porter et al., 2015). All means in all the constructs and subscales were lower than those of the field test parents (Table 2). Parents had a total FPTRQ mean score of 181.32, compared to the field test parents 229.7. The mean differences between parents in the field test and the parents of the study in attitudes construct showed a bigger difference of 19.39, compared to the other constructs, practices 18.69, and family specific 11.2. Parents in the study had a mean score of 48.31 compared to parents in the field test 67.7 in the attitudes construct. Mean differences between the maximum score and the score of parents in the study show that the practices component had the biggest difference of 66. Also comparing the maximum mean differences of the field test parents with those in the study, the field test mean differences are way low (see Table 2). Suggesting that parents in the field test group had higher ratings of parent–teacher relationships than those who participated in this study in all the subscales. In the commitment subscale, parents frequently agreed that teachers shared information with them about their child’s day, parents noted that their teachers were caring, understanding, flexible, dependable, trustworthy, respectful, available, were trustworthy and had their children’s best interest. Parents highly believed that the teachers could maintain a safe environment. However, in the communication subscale, 36 parents out of the 83 reported never or rarely did the teachers suggest activities for them and their child to do at home during COVID-19 pandemic. There were 33 parents who reported that teachers never or rarely offered ideas or suggestions about parenting. Meanwhile, 38 parents said that never or rarely did the teachers provide them with opportunities to make decisions about their child’s care and education. There were 33 parents who reported that never or rarely were they given the opportunity to give feedback on the teacher’s performances. And, 25 said that never or rarely did their teachers work with them to develop strategies they could use at home to support their child’s learning and development during COVID-19.Table 2 Total family provider relationship quality subscale means scores
Variable N Min Max M SD Field M Differences
Max.-P Max.-F F-M
Total FPTRQ 80 139.0 230 181.32 48/28.07 229. 748.7 0.348 4
Family knowledge 79 15 60 41.5 7.7 52.6 18.5 7.4 11.2
Construct practices 73 62 128 90.7 14.6 109.4 66 18.6 18.7
Collaboration 80 19 44 30.0 5.3 37.1 14 6.9 7.1
Responsiveness 73 19 44 31.7 5.9 38.7 12.33 5.3 7
Communication 80 16 32 21.2 3.5 23.3 10.8 8.7 2.1
Family focused 77 4 12 7.7 1.78 10.2 4.3 1.8 2.5
Construct attitudes 73 36 68 48.3 5.3 67.7 19.4 0.3 19.4
Commitment 75 16 36 25.4 5 34.0 10.6 2 8.6
Understanding
Context 76 4 20 10.89 4.27 15 9.1 5 4.1
Respect 79 7 18 12.2 2.8 18 5.8 0 5.8
Min minimum mean, Max. maximum mean, SP-M parents in study mean, F. field test parents mean, M mean
There were significant differences in ratings of parents with a GED or less and those with an associate or some college education ANOVA (F (2,77) = 8.026, p = 0.000. A Tukey post hoc test revealed that the parent–teacher relationships ratings were statistically significantly lower for parents with a GED education or less (Table 3) than those of parents with some or associate degree in all the subscales except in the family focused concern, understanding context and respect. No significant difference in ratings for the parents with a bachelors or graduate degree were found and those with an associate or some college degree. Mean differences (Table 3) show that parents with a GED or less education had lower mean scores in all the dimensions of parent–teacher relationships compared to the other groups of parents. The lower educated parents felt that their parent–teacher relationships were not strong compared to the more educated parents (Table 4).Table 3 ANOVA results on parent–teacher relationship (PTR) according to the parent’s education level
PTR dimensions Variable N SD F P M
FPTRQ total GED & less 16 0.47 8.02 0.001 19.8
Some or assoc 39 2.8 22.5
Bachelors/grad 25 1.9 21
Family specific GED & less 16 1.2 6.07 0.004 36.4
Knowledge Some or assoc 39 7.6 44
Bachelors/grad 25 8.7 40.6
Collaboration GED & less 16 1.3 4.01 0.022 27.4
Some or assoc 39 5.3 31.5
Bachelors/grad 25 6.3 29.2
Responsiveness GED & less 16 0.65 4.22 0.019 28.0
Some or assoc 39 6.6 33.1
Bachelors/grad 25 5.9 31.9
Communication GED & less 16 1.4 5.66 0.005 18.6
Some or assoc 39 3.5 22.0
Bach/grad 25 3.9 21.4
Family focused GED & less 16 0.44 1.26 0.288 7.0
Some or assoc 39 1.9 7.8
Bach/grad 25 2.0 7.8
Commitment GED & less 16 1.2 4.13 0.017 22.4
Some or assoc 39 5.2 26.6
Bach/grad 25 5.4 25.5
Understanding cont. Ged & less 16 4.8 0.46 0.630 9.9
Some or assoc 39 4.4 11.1
Bach/grad 25 4.2 11.1
Respect GED & less 16 0.79 3.07 0.055 13.6
Some or assoc 39 2.9 15.6
Bach/grad 25 3.1 15.08
Table 4 ANOVA results on parent–teacher relationship according to the parent’s income level
FPTRQ dimension Variable N SD F P M
Family specific 0–34,999 14 0.29 5.240 0.002 46.6
35,000–44,999 20 0.18 36.6
45,000–54,999 19 0.32 43.0
55,000–75,000 27 0.29 41.5
Collaboration 0–34,999 14 0.72 2.416 0.073 33.1
35,000–44,999 20 0.44 28.7
45,000–54,999 19 0.55 28.7
55,000–75,000 27 0.23 30.3
Responsiveness 0–34,999 14 0.74 2.021 0.179 33.0
35,000–44,999 20 0.38 29.8
45,000–54,999 19 0.64 30.8
55,000–75,000 27 0.31 32.6
Communication 0–34,999 14 0.50 1.614 0.193 22.3
35,000–44,999 20 0.43 19.8
45,000–54,999 19 0.46 21.1
55,000–75,000 27 0.41 21.6
Family focused concerns 0–34,999 14 0.68 6.43 0.010 7.3
35,000–44,999 20 0.39 6.8
45,000–54,999 19 0.62 7.6
55,000–75,000 27 0.66 8.5
Commitment 0–34,999 14 0.68 3.01 0.039 28.5
35,000–44,999 20 0.43 23.6
45,000–54,999 19 0.65 25.5
55,000–75,000 27 0.34 24.8
Understanding contexts 0–34,999 14 0.88 17.28 0.000 6.7
35,000–44,999 20 0.63 9.9
45,000–54,999 19 0.99 10.7
55,000–75,000 27 0.87 14.2
Respect 0–34,999 14 0.78 10.97 0.000 9.9
35,000–44,999 20 0.62 13.3
45,000–54,999 19 0.67 10.7
55,000–75,000 27 0.47 13.4
Total FPTRQ 0–34,999 14 0.29 7.106 0.000 21.4
35,000–44,999 20 0.18 20.3
45,000–54,999 19 0.32 21.2
55,000–75,000 27 0.29 23.2
There was a significant difference in the parents’ ratings across the subscales according to the family income variable F (3.76) = 7.106, p = 0.000. Differences were seen in family specific knowledge (p = 0.002), family focused concerns (p = 0.001), commitment (p = 0.032), understanding context (p = 0.000) and respect (0.002). An examination of the result of the Tukey test to determine among which groups the differences were, there were differences in scores between the parents with an income of between 35,000 and 44,999 and those earning 55,000 and above and those earning between 45,000–54,999 and 55,000 and above. (p = 0.000). Therefore, in general parents with an income between 35,000 to 54,999 scored differently from those parents with an income from 55,000 and above. The means show that parents with the lower income rated understanding contexts (M = 6.7) and respect (M = 9.9) lower than those with higher incomes. These parents felt that their teachers judged them more based on their faith and religion, cultural values, race and ethnicity and financial situations than the parents with the higher income. However, in the family specific (M = 46.4) subscale and commitment (M = 28.5), parents with the lowest income scored the highest means. These parents also felt that in family focused concerns (M = 8.5), understanding contexts (M = 14.2) and respect (M = 15.4), parents with the highest income scored the highest means. The income breakdowns used in the development of the FPTRQ tool were used.
Parent–Teacher Relationship Based on the Child’s Age
Parents of older children, 3–5 years, had significantly higher parent–teacher relationship scores t (78) = − 4.12, p = 0.000) than those of the younger children (0–2 years old). These was seen in communication (22.01 ± 3.2), t (78) = − 3.132, p = 0.002, family focused concerns (8.1 + 1.7), t (78) = − 3.278, p = 0.002, and understanding context (11.97 ± 4.4) t (78) = − 3.049, p = 0.003 than parents with younger children, communication (19.5 ± 3.6), family focused concerns (6.8 ± 1.6), and understanding context (9.0 ± 3.3). During the pandemic, parents of younger children felt that teachers did not promote 2-way communication that considered the parents’ preferences while considering their own personal boundaries. They also felt that teachers did not address parents’ family-focused concerns in ways that showed interest in meeting their needs. Finally, parents of younger children felt that teachers did not show understanding and appreciation of the broader contexts of situations the families were going through and how that contributed to their development. That is, they felt they were judged based on their religious beliefs, culture, values, race & ethnicity, and their financial situations. They had lower mean scores in these subscales.
Parents with children in programs at the rural locations reported significantly higher parent -teacher relationship ratings on family specific knowledge (M = 43.4) collaboration (M = 31), responsiveness (M = 33.3), communication (M = 22.1), and commitment (27.2) (Table 5). They felt that teachers knew them and their families well, collaborated with them and engaged them in the program. Specifically, they engaged parents in appropriate communication and were sensitive to the needs of parents and their children compared to parents in the urban programs. However, they felt teachers did not understand the context they were living in; they judged them based on their faith and religion, culture and values, race, and financial situations. They did not show appreciation of the broader contexts of the experiences they were having and did not feel respected from the way teachers treated their children and families. Parents in the urban programs had statistically significant higher parent–teacher ratings in understanding context (M = 12.4) and respect (M = 13.6). They felt teachers showed appreciation of the broader contexts of the experiences they were having and how it impacted their development. They also felt teachers in the urban areas were respectful based on how they valued their children and families. They were welcoming, accepting, nondiscriminatory and non-judgmental compared to how the parents in the rural areas felt. There were no significant differences in the family focused concerns. Both parents from the urban and rural areas did not differ in the way they felt teachers communicated with them.Table 5 Independent t test results on parent–teacher relationships and according to the social context variable rural/urban
Variable M SD T p
N Rural = 37
N Urban = 42
Family specific knowledge 1.99 76 0.000
Rural 43.4 9.9
Urban 39.2 4.9
Total FPTRQ practices 2.04 66 0.003
Rural 187.2 17.5
Urban 176.11 2
Collaboration 1.41 77 0.000
Rural 31.0 6.8
Urban 29.2 3.5
Responsiveness 2.42 71 0.000
Rural 33.3 7.5
Urban 30.1 3.4
Communication 2.06 77 0.004
Rural 22.0 4.2
Urban 20.4 2.7
Family focused concerns − 0.356 74 0.453
Rural 7.6 1.7
Urban 7.7 1.8
Commitment 3.30 72 0.000
Rural 27.2 6.2
Urban 23.6 2.2
Total attitudes − 1.86 0.888
Rural 47.2 5.6
Urban 49.5 4.8
Commitment 3.30 77 0.002
Rural 7.62 1.7
Urban 7.76 1.8
Understanding context − 3.49 73 0.000
Rural 9.2 4.7
Urban 12.4 3.2
Respect − 5.81 76 000
Rural 10.5 3.2
Urban 13.6 1.2
Parent–Teacher Relationships Based on Mode of Learning (Distance or in Person)
Parents with children in programs that provided distance learning option had statistically significantly higher parent–teacher relationship rating scores (22.1 ± 2.5) compared to parents in programs that did not provide a distance learning option (19.7 ± 1.2), t (68) = 3.68, p = 0.000. Specific differences were noted on family specific knowledge (42.2 ± 6.3), collaboration (29.5 + 1.8) responsiveness (30.9 ± 4.9), family focused (7.9 ± 1.9), and understanding context (12.4 + 0.006) for families with distance learning option compared to families without the option (36.5 ± 3.0). (28.3 ± 2.6), (7.0 ± 0.00), and (9.8 ± 1.9). Communication, commitment, and respect were not statistically significant. Commitment looks at how sensitive teachers are to the needs of children, parents, and families; how they are open to change, how willing they are to alter their typical practices to be sensitive to individual child, parent, or family needs. Communication looks at how teachers promote 2-way communication while respecting parents’ preferences and considering their own personal boundaries. Since programs were closed most of these services may have not seamlessly continued. Respect measures how teachers value child & family by being welcoming, accepting, nondiscriminatory & non judge mental. Again, it’s presumed that this was not going on during COVID-19 pandemic as there was a lot of fluctuations in centers opening or closing. The distance learning option helped parents feel connected to the programs in most of the parent–teacher relationship areas i.e., they felt teachers knew who they were and their children, they were more responsible in meeting their needs, teachers collaborated with them about their child, teachers were also focused on family needs and were committed to supporting their child and family. Also, teachers understood what they were going through and the contexts they were living in.
Discussions and Conclusions
It is clear from the findings that the quality of parent–teacher relationships during COVID-19 was lower compared to those who had participated in the field test study that analyzed parent teacher relationships across different early childhood programs nationally. There were big differences in all the 8 subscales beginning from family specific knowledge to respect. However, during COVID-19 parents agreed that they were comfortable sharing a variety of information with their teachers such as what activities they were doing outside school to support their child’s learning, their household schedules, and their family culture and values. Also, several parents agreed that they had met with their teachers during the year to discuss their child’s abilities. It looked like even during COVID-19, parents were in contact with their teachers. This finding is in line with NAEYC’s survey (2020) which suggested that programs created avenues to reach parents. However, there were parents with a mean score of 15 out of 60 on family specific knowledge subscale. This could imply that some parents did not feel their teachers connected with them sufficiently during this pandemic period. The mean scores of the individual items in the family specific knowledge were all 2.5 and higher on a scale of 1–4). Items on sharing information about their spouse, family life, and their religion and faith scored the least compared to families feeling that they were very comfortable sharing information about their culture and family values (M = 2.98), sharing what they do outside school to support their child’s learning (M = 2.95), and sharing about their household schedule (M = 2.91). It can be said that during COVID-19 teacher connections with parents to know them better was not disrupted. Programs stepped up and continued to provide these connections.
Additionally, in collaboration and responsiveness practices, some parents had a mean score of 19 out of a 44 possible score. Both items scored a mean of 30. The highest ranked areas where parents felt teachers collaborated well with them during COVID-19 were on meeting and talking with parents about their child’s abilities (M = 3:1), their child’s learning (M = 3.0), the goals they had for their child (M = 3.0). The lowest mean item on this subscale was meeting and talking with parents on what they do outside education and care setting to support their child’s learning (M = 2.5). These two items had the biggest difference in means, followed by communication, where also some parents had a minimum mean of 16 out of a 32 possible score. In understanding context, we had some parents scoring teachers with a mean of 4 out of a possible 20; which was very low. Overall, the practices component that requires parents to work with families suffered in COVID-19 for some parents. However, most parents agreed that teachers were respectful and committed to their work.
Parents with the least education (GED or less) reported low scores in parent–teacher relationships, specially when compared to the other groups of parents with either some college or an associate degree, and those with a bachelors or graduate degree. This group of parents scored less in all the subscales during COVID-19. Examining other factors contributing to this in parents with low level education is necessary to better serve them during challenging times. Some studies on parent involvement among immigrants have found that often the most parent volunteers in school were those with less education (Peña, 2000). In this study, 29 of the participants were native Americans. Also, it has been found that while most of the parents that volunteer in schools are less educated it does not translate to building better relationships with their teachers. Other studies have statistically found parent education as one of the family demographic variables significantly affecting parent involvement (Fantuzzo et al., 2000; Manz et al., 2004). Even if these parents had low means in all subscales, it was evident that the means of the parents with either a bachelors or graduate degree and those with some or associate degree were not significantly different. Both groups scored low compared to the field test parents, suggesting that COVID-19 impacted the parent–teacher relationships for most groups of parents. It is also possible that the less educated parents had less knowledge of ways to reach their childcare providers or that they lacked strategies for surviving during the pandemic. Developing tools for this group of parents may be helpful during pandemics e.g., stress reductions tools, time management, defense mechanism tools, etc. The higher educated parents may have found other ways to meet their needs that could be investigated and possibly adapted to support other parents. Also, a look at program characteristics would help shade light into these findings.
Another finding was that parent income affected how teachers rated the family specific knowledge, family focused concerns, commitment, and understanding contexts during COVID-19. In all these three, parents with an income between 35,000 and 44,999 had the lowest mean scores. Generally, parents with the second lower income had lower means scores than the high-income parents. It is possible that higher income parents in the study had advantages of being more knowledgeable, committed to their children’s learning and were more likely to understand situations better than less educated parents. Parents of lower income agreed that teachers knew them and their families and were committed to them. However, they didn’t feel the same about teachers’ understanding of their contexts and being respectful to them. These group of parents felt that teachers judged their families based on their faith and religion, their culture and values, their race and ethnicity, and their financial situation compared to the parents who were earning between 45,0000 and 54,999. This could be areas for teachers to work on in their programs especially for these group of parents and looking more into why the parents felt this way during COVID-19.
Parents of older children, 3–5 years, had statistically significantly higher parent–teacher relationship scores than those of the younger children (0–2 years old) in communication, family focused concerns and understanding context. During the pandemic, parents of younger children felt that teachers did not promote 2-way communication that considered their preferences and personal boundaries. They also felt that teachers did not address parent’s family-focused concerns in ways that showed interest in meeting their needs. Finally, parents of younger children felt that teachers did not show understanding and appreciation of the broader contexts of situations the families were going through and how that contributed to their development. The COVID-19 pandemic had a big impact on parents and teachers in this area.
Parents living in rural areas had higher mean ratings for parent–teacher relationships in family specific knowledge collaboration, responsiveness, communication, and commitment. Parents in the urban programs rated respect and understanding context as the highest mean scores. They felt teachers collaborated with them and engaged them in the program, were committed to working with them to improve their children’s learning, and were responsive in meeting their needs. Many schools or programs closed during the COVID-19 pandemic, however data reports that closures in rural areas were even more than in urban areas (Lee et al., 2021). This census report found that more closures happened along the west coast and northeast coast which experienced relatively high rates of closure. This study was in a midwestern state. It is important to note that even before the pandemic there was a shortage of childcare programs in rural areas due to the decline in licensed family and home-based providers. Quality is often likely to be an issue (Center for Rural Development and Policy, 2017). Thus, assessing parent–teacher relationships in already under qualified programs was a big call. Most of these programs fall under the home/family childcares and private programs. This data looked at the parents who came from rural areas and who also were in this kind of program. The data revealed that parents in the private/family/home child care programs and in rural areas had much lower ratings compared to those in Head Start. Thus, there is a need to advocate for more quality programs in rural areas.
Parents that had the distance learning option reported higher ratings on parent–teacher relationships. The distance learning option during COVID-19 helped parents feel connected to the programs in most of the parent–teacher relationship areas. They felt teachers knew who they were and their children, they were more responsible in meeting their needs, teachers communicated to them about their child, teachers were also focused on family needs and were committed to supporting their child. Family and teachers understood what they were going through and the contexts within which they lived. A survey of preschool parents on distance learning by Stites et al. (2021) found that parents reported engaging more in literacy-based activities and receiving few opportunities for social emotional development when the distance learning option was used. Parents were not supportive of the amount of time involved in assisting children with distance learning opportunities and preferred activities that took less time, allowing for social interaction with other children. There is still more to learn on the aspects of distance learning that were effective during COVID-19 and how those impacted parent–teacher relationships. But from this study the distance learning option gave parents a connecting place for learning but not for the social needs of their children.
Limitations, Implications and Recommendations
This is a self-report study of early childhood parents’ perceptions of parent–teacher relationships in their programs and therefore bears limitations of that methodology. Additionally, the study investigated a wide range of early childhood programs. Early childhood programs are very diverse with some not having an inbuilt parent–teacher relationship program such as those in Head Start. This study did not explore program differences as a variable. A study that investigates that variable might give further useful insights into other factors affecting parent–teacher relationships. This study collected data from only parent–teacher relationships. Looking at other sources of data, for instance, teacher reports of the same could also contribute to a strong understanding of parent–teacher relationships during the pandemic. Data on student outcomes or parent outcomes would also be helpful for understanding if the benefits of parent–teacher relationships have an impact on children and parent outcomes.
The other limitation of the study is related to the differences in the times programs were shutdown due the pandemic, as well as the nature of the program clientele in these programs. The COVID-19 pandemic started Fall 2019 and most programs started closing in the mid-2020 spring semester when it was clear the world was going through a pandemic. Most ECE programs experienced this effect from spring 2020 when most campuses and schools closed too. The NAEYC (2020) survey that makes its monthly updated database on childcare closures to the public shows that two-thirds of childcare centers closed in April 2020, while one-third remained closed in April 2021. The data also points out that non-White families were more likely to be exposed to childcare closures than White families. Such findings have greater implications on the widening inequalities in access to childcare and the pace of labor market recovery after the pandemic subsides. Since there was no specific time that all programs closed, it is difficult to capture data across these programs.
Declarations
Competing Interests
There are no financial or non-financial interests that are directly or indirectly related to this work submitted for publication.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Alison, H. & Gaviria-Loaiza, J. (2021). Predictors of head start teachers perceived quality of relationships with families. NHSA Dialog, The research-to-practice journal for the early childhood field. Volume healthy children, healthy adults, in an unhealthy Time, 24, 1.
Annat, E. O., & Gassman-Pines, A. (2020). Snapshot of the COVID crisis impact on working families. Retrieved March 20, 2022, from https://econofact.org/snapshot-of-the-covid-crisis-impact-on-working-families
Bernhard JK Lefebvre ML Kilbride KM Chud G Lange R Troubled relationships in early childhood education; parent-teacher interactions in ethnoculturally diverse childcare setting Early Education & Development 1988 9 5 28 10.1207/s15566935eed0901_1
Bigras N Lemay L Lehrer J Charron A Duvali S Robert-Mazaye C Laurin E Early childhood educators’ perceptions of their emotional state, relationships with parents, challenges, and opportunities during the early stage of the pandemic Early Childhood Education 2021 49 775 787 10.1007/s10643-021-01224-y
Callaway-Cole L Kimble A Maintaining professional standards in early childhood teacher preparation; evaluating adaptations to fieldwork-based experiences during COVID-19 Early Childhood Education Journal 2021 49 841 853 10.1007/s10643-021-01227-9 34177245
Center for Rural Development and Policy (2017). Child care solutions with initiative foundation. Available online https://www.ruralmn.org/podcast/season-3-episode-1-childcare-solutions-with-initiative-foundation/
Crawford A Vaughan KA Guttentag CL Varghese C Yoonkyung O Zucker T “Doing what I can, but I got no magic wand:” A snapshot of early childhood educator experiences and efforts to ensure quality during the COVID-19 pandemic Early Childhood Education Journal 2021 49 829 840 10.1007/s10643-021-01215-z 34092993
Current Population Reports, P70-135. U.S. Census Bureau.
Den Berg MV Subjective social mobility: Definitions and expectations of ‘moving up’ of poor Moroccan women in the Netherlands International Sociology 2011 26 503 523 10.1177/0268580910393042
Eadie P Levickis P Murray L Page J Elek C Church A Early childhood educators’ wellbeing during the COVID-19 pandemic Early Childhood Education Journal 2021 49 903 913 10.1007/s10643-021-01203-3 33994770
Egan SM Pope J Moloney M Hoynel C Beatty C Missing early education and care during the pandemic; the socio-emotional impact of the COVID-19 crisis on young children Early Childhood Education 2021 49 925 934 10.1007/s10643-021-01193-2
Elek and Church Early Childhood Educators’ wellbeing during the COVID-19 Pandemic Early Childhood Education Journal 2021 49 903 913 10.1007/s10643-021-01203-3 33994770
Fantuzzo J Tighe E Childs S Family Involvement Questionnaire: A multivariate assessment of family participation in early childhood education Journal of Educational Psychology 2000 92 367 376 10.1037/0022-0663.92.2.367
Feinberg, M. E., A Mogle, J., Lee, J., Tornello, S. L., Hostetler, M. L., Cifelli, J. A. Bai, S., & Hostez, E. (2021). Impact of the COVID-19 pandemic on parent, child, and family functioning. Family Process. 10.1111/famp.12649
Forry N Bromer J Chrisler A Rothenberg L Simkin S Daneri P Quality of family-provider relationships: Review of conceptual and empirical literature of family-provider relationships 2012 Child
Hae Min Y Yu Jin C Hyun Jeong K Jin HK Jee Hyun B A mixed-methods study of early childhood education and care in South Korea: Policies and practices during COVID-19 Early Childhood Education Journal 2021 49 6 1141 1154 10.1007/s10643-021-01239-5 34404971
Hooper, A., & Gavaria-Loaiza, J. (2021). Predictors of head start teachers’ perceived quality of relationships. Healthy children, healthy adults, in an unhealthy time, 24, https://journals.charlotte.edu/dialog/article/view/1054
Iruka IU Winn DC Kingsley SJ Orthodoxou YJ Links between parent teacher relationships and kindergartners’ social skills; do child ethnicity and family income matter? The Elementary School Journal 2011 111 3 387 409 10.1086/657652
Johnson AD Martin A Partika A Philips DA Castle S Chaos during the COVID-19 outbreak of household chaos among low-income families during a pandemic Family Relations 2022 71 18 28 10.1111/fare.12597 34898781
Karasar N Bilimsel aras turma yontemi 1999
Karpman, M., Zuckerman, S., Gonzalez, D., & Kenney, G. (2020, April 28). The COVID-19 pandemic is straining families' abilities to afford basic needs. Urban Institute. Retrieved from https://www.urban.org/research/publication/covid-19-pandemic-straining-families-abilities-afford-basic-needs
Kerr ML Rasmussen HF Fanning KA Braaten SM Parenting during COVID-19: A study of parents’ experiences across gender and income levels Family Relations 2021 70 1327 1342 10.1111/fare.12571 34548726
Kim, K., Porter, T., Atkinson, V., Guzman, L., Rui, N., Ramos, M., et al. (2014). Family and provider/Teacher relationship quality: User’s manual. OPRE Report 2014–65. Washington, DC: Office of planning, research and evaluation, administration for children and families, U.S. Department of Health and Human Services.
Kimil JH Araya M Hailu BH Rose PM Woldehanna T The implications of COVID-19 for early childhood education in Ethiopia; perspectives from parents and caregivers Early Childhood Education 2021 49 855 867 10.1007/s10643-021-01214-0
Lafave L Webster AD McConnell C Impact of COVID-19 on early childhoodeducator’s perspectives Early Childhood Education 2021 49 935 945 10.1007/s10643-021-01195-0
Lang SN Tolbert AR Schoppe-Sullivan SJ Bonomi AE A cocaring framework for infants and toddlers; applying a model of coparenting to parent-teacher relationships Early Childhood Research Quarterly 2016 34 40 52 10.1016/j.ecresq.2015.08.004
Lang SN Schoppe-Sullivan SJ Kotila LE Kamp Dush CM Daily parenting engagement among new mothers and fathers: The role of romantic attachment in dual-earner families Journal of Family Psychology 2013 27 862 872 10.1037/a0034510 24127790
Lee E Parolin Z The care burden during COVID-19: A national database of child care closures in the United States Socius 2021 7 23780231211032028 10.1177/23780231211032028
Manz PH Fantuzzo JW Power T Multidimensional assessment of family involvement among urban elementary students Journal of School Psycology 2004 42 461 475 10.1016/j.jsp.2004.08.002
McKenna M Soto-Boykin X Cheng K Haynes E Osorio A Altshuler J Initial development of a national survey on remote learning in early childhood during COVID-19. Establishing content validity and reporting successes and barriers Early Childhood Education 2021 49 815 827 10.1007/s10643-021-01216-y
National Association for the Education of Young Children. (2020). Developmentally appropriate practice (DAP). Position statement. Retrieved from: https://www.naeyc.org/resources/position-statements/dap/contents
Nzinga-Johnson S Baker JA Aupperlee J Teacher-parent relationships and school involvement among racially and educationally diverse parents of kindergartners Elementary School Journal 2009 110 81 91 10.1086/598844
Peña DC Parent Involvement: Influencing Factors and Implications The Journal of Educational Research 2000 94 42 54 10.1080/00220670009598741
Porter, T., Bromer, J., & Moodie, S. (2011). Quality rating and improvement systems (QRIS) and Family-sensitive caregiving in early care and education arrangements: promising directions and challenges, research-to-policy, research-to practice brief OPRE 2011-11d. Washington, DC: Office of planning, research and evaluation, administration for children and families, U.S. Department of Health and Human Services.
Porter, T., Bromer, J., & Forry, N. (2015). Assessing Quality in Family and Provider/Teacher Relationships: Using the Family and Provider Teacher Relationship Quality (FPTRQ) Measures in Conjunction with Strengthening Families and the Head Start Parent, Family and Community Engagement Frameworks and their Self-Assessment Tools. OPRE Report 2015-56. Office of Planning, Research and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Services.
Steed EA Leech N Shifting to remote learning during COVID-19; differences for early childhood and early childhood special education teachers Early Childhood Education Journal 2021 49 789 798 10.1007/s10643-021-01218-w 34131379
Stites ML Sonneschein S Galczyk SH Preschool parents’ views of distance learning during COVID-19 Early Education and Development 2021 32 923 939 10.1080/10409289.2021.1930936
Swigonskil NL James B Wynns W Casavan K Physical, mental, and financial stress: impacts of COVID-19 on early childhood educators Early Childhood Journal 2021 49 799 806 10.1007/s10643-021-01223-z
Toros K Tart K Falch-Eriksen A Collaboration of Child Protective services and early childhood educators.; enhancing the well-being of children in need Early Childhood Education 2021 49 995 1006 10.1007/s10643-020-01149-y
Wei Y Wang L Tan L Li Q Zhou D Occupational commitment of Chinese kindergarten teachers during the COVID-19 pandemic: Predictions of anti-epidemic action, income reduction, and career confidence Early Childhood Education Journal 2021 49 1031 1045 10.1007/s10643-021-01232-y 34248326
| 0 | PMC9741752 | NO-CC CODE | 2022-12-14 23:29:57 | no | Early Child Educ J. 2022 Dec 11;:1-13 | utf-8 | Early Child Educ J | 2,022 | 10.1007/s10643-022-01431-1 | oa_other |
==== Front
TechTrends
TechTrends
Techtrends
8756-3894
1559-7075
Springer US New York
818
10.1007/s11528-022-00818-6
Column: Graduate Member Musings
Instructional Design: A Workforce Perspective for 2023
Luchs Christopher [email protected]
grid.261368.8 0000 0001 2164 3177 Old Dominion University, Norfolk, VA USA
11 12 2022
13
23 11 2022
© Association for Educational Communications & Technology 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.
Instructional design like many fields is experiencing the effects of workforce pressures and struggles to hire, retain, and promote employees. This article looks at several global trends affecting the industry, industry forecasts and trends, and ends with a look at instructional design and recent trends in online learning executive positions.
==== Body
pmcLike my colleague, Melissa Jones who related being a graduate student and professional in the last Graduate Member Musing (Jones, 2022), I am mid-career with over 15 years in higher education. I’ve served in many positions such as faculty, adjunct, coordinator, director, and associate dean in the career technical education side of academics bouncing between transfer and workforce development. As I have taken graduate and doctoral courses, I am not looking at how to be an instructional designer; instead, I am looking at how to develop and retain talent in design, technology, and teaching. A few workforce trends affecting all industries including educational technology and instructional design are the Great Resignation, Quiet Quitting, and employee protest over workloads, required overtime, and working conditions.
The Great Resignation was first identified and brought to the public’s attention in 2021 as over 47 million employees in the United States (US) quit their jobs (Fuller & Kerr, 2022). However, Fuller and Kerr (2022) noted that overall resignations had been climbing since 2009. In 2009, approximately 1.4% of the US workforce left their jobs, this rate continued to climb over the next decade to approximately 2.3% of the US workforce in 2019 (Fuller & Kerr, 2022, Fig. 1). In 2020, the rate decreased to approximately 2.1% during the first year of COVID-19 and then increased to approximately 2.5% in 2021 (Fuller & Kerr, 2022, Fig. 1). The causes of these declines were attributed to five factors dubbed the “five r’s”: retirement, relocation, reconsideration, reshuffling, and reluctance (Fuller & Kerr, 2022). Most of these factors are straight-forward in their meaning. The two factors needing clarification are reconsideration and reshuffling. Reconsideration was the term used by Fuller and Kerr (2022) to include personal reflection during the pandemic on what one values and has led to a possible disconnect between employer demands and employee’s tolerances for these demands (Reconsideration section, para. 3). The reconsideration factor played a major role in the greater exits of women and younger age groups (Fuller & Kerr, 2022, Reconsideration section, para 1–2). Reshuffling refers to a worker moving between and among jobs within a given industry (Fuller & Kerr, 2022, Reshuffling section, para. 2).
The advent of employee resistance to employer demands has played out domestically in the US via the unionizing of service workers at Starbucks and Amazon. This has also spread internationally in various fields, like the anti-996 movement in China. 996 refers to a common employer requirement for employees to work from 9am to 9 pm for six days a week (Chappell, 2021). Beginning in 2019 the anti-996 campaign did not show up outside of technology worker forums until 2021(Wang, 2020; Yang, 2019; Huang, 2019). In 2021, Chinese media highlighted the practice of 996 after a young woman died after several long shifts at an e-commerce startup. During the subsequent investigation, her co-workers stated they regularly worked more than 300 h each month despite China’s official limit of no more than 36 h of overtime work (Chappell, 2021). Additional worker deaths contributed to protests and the formation of the 996.ICU Github site. 996.ICU refers to the saying “Work by 996, sick in ICU”, which Chinese developers coined meaning that if you follow the 996 work schedule, you are risking ending up in the Intensive Care Unit (Github.com, 2022). The 996 practice was not limited to technology companies as many service industries also unofficially implemented to increase productivity, lower costs, and cover for sick workers.
Education has not been exempted from the effect of the Great Resignation, Quiet Quitting, or staff or student protests over institutional demands. Recently, the University of California recognized graduate student researchers as employees. This allows these researchers to join the United Auto Workers (UAW) 2865 union which supports over 19,000 tutors, readers, graduate student instructors and teaching assistants at the University of California (UAW 2865, 2022). UAW 2865 highlights their work to increase the wages of unionized student-workers by 30% and recapturing over $8 million in back-pay damages and misclassification settlements since their 2018 contract (UAW 2865, 2022).
Instructional Design as a Department of Labor Category
No workforce analysis discussion would be complete without visiting the US Department of Labor (USDL) and the Occupational Information Network (O*NET) for information about the future of the field. O*NET was developed under the sponsorship of the U.S. Department of Labor’s Employment and Training Administration via a grant from the North Carolina Department of Commerce (O*Net, 2022a). Both O*Net and the USDL use Standard Occupational Classification (SOC) codes to parse out various jobs and industries. Instructional design falls under the SOC code: 25–9031.00 for Instructional Coordinators. This category includes the following job titles: Curriculum and Instruction Director, Curriculum Coordinator, Curriculum Director, Curriculum Specialist, Education Specialist, Instructional Designer, Instructional Systems Specialist, Instructional Technologist, Learning Development Specialist, and Program Administrator (O*NET, 2022b).
Nationally, instructional coordinators earn a median wage of $63,740 annually and represent 205,700 employees (O*Net, 2022b). These positions are mostly concentrated in the following industries: Educational Support Services (7.34% of industry employment); Junior Colleges (1.22% of industry employment); Elementary and Secondary Schools (1.08% of industry employment); Colleges, Universities, and Professional Schools (0.99% of industry employment); and State Government, excluding schools and hospitals (0.34% of industry employment) (U.S. Bureau of Labor Statistics, 2022). This sector is also anticipated to grow up to 7% from 2021 to 2031 (O*Net, 2022b).
O*NET (2022b) found that educational requirements for instructional coordinator positions varied. Sixty percent of their respondents indicated that a master’s degree was required, 25% of respondents indicated a bachelor’s degree was required, and 11% percent of respondents indicated that a post-master’s certificate was required. Additionally, many required over five years of experience in related areas (O*NET, 2022b).
Instructional Design as a Gatekeeper Degree
Another interesting trend that emerged during the pandemic was the Instructional Design degree becoming a required degree in many online learning administrative jobs. Many job descriptions are now listing a master’s degree in a field related to Library or Information Science, Instructional Technology, Education, or Instructional Design in their minimum qualifications.
As I close out this musing, I’m left with many questions still unanswered. Is there a parallel between instructional design and information science when it comes to job performance? Do our instructional design programs properly prepare students to take on academic leadership roles such as Deans or Directors of Online Learning? How do we build pathways to help our students gain five years of previous experience required? Is an instructional design degree correlated with being a better online learning administrator? Do our Educational services, colleges, and secondary schools offer growth opportunities for our students once they attain their jobs. I’d love to continue these discussions in AECT.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Chappell, B. (2021). Employer’s can’t require people to work 72 hours a week, China’s high court says. National Public Radio. Retrieved November 13, 2022, from https://www.npr.org/2021/08/30/1032458104/12-hour-6-day-996-work-schedule-illegal-china-deaths-tech-industry
Fuller, J., & Kerr, W. (2022). The Great Resignation didn’t start with the pandemic. Harvard Business Review. Retrieved November 12, 2022, from https://hbr.org/2022/03/the-great-resignation-didnt-start-with-the-pandemic
Github.com. (2022). 996.ICU. Retrieved November 13, 2022, from https://github.com/996icu/996.ICU
Huang, Z. (2019). No sleep, no sex, no life: tech workers in China's Silicon Valley face burnout before they reach 30. South China Morning Post. Retrieved November 13, 2022, from https://www.scmp.com/tech/apps-social/article/3002533/no-sleep-no-sex-nolife-tech-workers-chinas-silicon-valley-face
Jones MK Centering Teaching and Learning Centers in Instructional Systems Design Conversations TechTrends 2022 66 903 904 10.1007/s11528-022-00786-x
O*NET. (2022a). About O*NET. https://www.onetcenter.org/overview.html. Accessed 13 Nov 2022
O*NET. (2022b). Instructional Coordinators: 25–9031.00. O*NET OnLine. Retrieved November 13, 2022, from https://www.onetonline.org/link/summary/25-9031.00
UAW 2865. (2022). About UAW 2865. Retrieved November 13, 2022, from https://uaw2865.org/about-our-union/
U.S. Bureau of Labor Statistics. (2022). Occupational employment and wages, May 2021: 25–9031 instructional coordinators. Retrieved November 12, 2022, from https://www.bls.gov/oes/current/oes259031.htm
Wang JJ How managers use culture and controls to impose a '996' work regime in China that constitutes modern slavery Accounting & Finance. 2020 60 4 4331 4359 10.1111/acfi.12682
Yang, Y. (2019, April 3). China tech worker protest against long working hours goes viral. Financial Times. Retrieved November 13, 2022, from https://www.ft.com/content/72754638-55d1-11e9-91f9-b6515a54c5b1
| 0 | PMC9741753 | NO-CC CODE | 2022-12-14 23:29:57 | no | TechTrends. 2022 Dec 11;:1-3 | utf-8 | TechTrends | 2,022 | 10.1007/s11528-022-00818-6 | oa_other |
==== Front
Adv Ther
Adv Ther
Advances in Therapy
0741-238X
1865-8652
Springer Healthcare Cheshire
36502449
2360
10.1007/s12325-022-02360-6
Original Research
PRESTO 2: An International Survey to Evaluate Patients’ Injection Experiences with the Latest Devices/Formulations of Long-Acting Somatostatin Analog Therapies for Neuroendocrine Tumors or Acromegaly
http://orcid.org/0000-0001-6404-991X
O’Toole Dermot [email protected]
1
http://orcid.org/0000-0003-1613-3919
Kunz Pamela L. 2
http://orcid.org/0000-0001-7052-6436
Webb Susan M. 3
http://orcid.org/0000-0002-0586-5586
Goldstein Grace 4
http://orcid.org/0000-0003-0407-953X
Khawaja Sheila 5
http://orcid.org/0000-0002-4135-9226
McDonnell Mark 6
http://orcid.org/0000-0001-8208-6490
Boiziau Sandra 7
http://orcid.org/0000-0002-6813-0579
Gueguen Delphine 7
http://orcid.org/0000-0001-6248-8986
Houchard Aude 7
http://orcid.org/0000-0002-6114-5786
Ribeiro-Oliveira Antonio Jr 8
http://orcid.org/0000-0001-6767-7313
Prebtani Ally 9
1 grid.412751.4 0000 0001 0315 8143 Neuro Endocrine Tumours-ENETS Centre of Excellence, St. Vincent’s University Hospital, Dublin, Ireland
2 grid.47100.32 0000000419368710 Yale School of Medicine and Yale Cancer Center, New Haven, CT USA
3 grid.7080.f 0000 0001 2296 0625 Departamento de Medicina/Endocrinología, Hospital Sant Pau, IIB-Sant Pau, Universitat Autònoma de Barcelona, CIBERER U747 Barcelona, Spain
4 grid.478600.d 0000 0004 5906 2249 Carcinoid Cancer Foundation, Mount Kisco, NY USA
5 World Alliance of Pituitary Organizations, Zeeland, The Netherlands
6 NET Patient Network Ireland, Dublin, Ireland
7 grid.476474.2 0000 0001 1957 4504 Ipsen, Boulogne-Billancourt, France
8 grid.423023.4 0000 0004 6011 1247 Ipsen, Cambridge, MA USA
9 grid.25073.33 0000 0004 1936 8227 Faculty of Health Sciences, McMaster University, Hamilton, ON Canada
11 12 2022
120
12 9 2022
14 10 2022
© The Author(s), under exclusive licence to Springer Healthcare Ltd., part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Introduction
Real-world data evaluating patients’ injection experiences using the latest devices/formulations of the long-acting (LA) somatostatin analogs (SSAs) lanreotide Autogel/Depot (LAN; Somatuline®) and octreotide LA release (OCT; Sandostatin®) are limited.
Methods
PRESTO 2 was a 2020/2021 e-survey comparing injection experience of adults with neuroendocrine tumors (NETs) or acromegaly treated with LAN prefilled syringe versus OCT syringe for > 3 months in Canada, Ireland, the UK and the USA (planned sample size, 304). Primary endpoint: the proportion of patients with injection-site pain lasting > 2 days after their most recent injection, analyzed using a multivariate logistic regression model. Secondary endpoints included interference with daily life due to injection-site pain and technical injection problems in patients with current SSA use for ≥ 6 months.
Results
There were 304 respondents (acromegaly, n = 85; NETs, n = 219; LAN, n = 168; OCT, n = 136; 69.2% female; mean age, 59.6 years). Fewer patients had injection-site pain lasting > 2 days after the most recent injection with LAN (6.0%) than OCT (22.8%); the odds of pain lasting > 2 days were significantly lower for LAN than OCT, adjusted for disease subgroup and occurrence of injection-site reactions (odds ratio [95% confidence interval]: 0.13 [0.06–0.30]; p < 0.0001). Injection-site pain interfered with daily life “a little bit” or “quite a bit” in 37.2% and 3.8% (LAN) versus 52.5% and 7.5% (OCT) of patients, respectively. Among patients with ≥ 6 months’ experience with current SSA (92.4% of patients), technical injection problems never occurred in 76.8% (LAN) and 42.9% (OCT) of patients.
Conclusions
Compared with OCT, significantly fewer patients using LAN had injection-site pain lasting > 2 days after their most recent injection. Also, fewer LAN-treated patients experienced technical problems during injection. These findings demonstrate the importance of injection modality for overall LA SSA injection experience for patients with acromegaly or NETs.
Graphical abstract
Plain Language Summary
Patients with neuroendocrine tumors or acromegaly often receive long-term monthly treatment with somatostatin analogs. These injectable drugs stop the body from making an excess of certain hormones. Understanding patients’ experiences of these injections helps to provide better care. The PRESTO 2 online study surveyed 304 patients in Canada, Ireland, the UK and the USA with neuroendocrine tumors or acromegaly who were being treated with a somatostatin analog, either lanreotide Autogel/Depot (LAN) or octreotide long-acting release (OCT). The survey asked about injection experience, including injection-site pain lasting > 2 days and how it affected patients’ lives, anxiety before injections and technical problems during injections (like syringe blockages). The survey showed fewer patients receiving LAN than OCT had injection-site pain that lasted > 2 days, and fewer said that the pain interfered with their daily lives. There were fewer technical injection problems with LAN than with OCT. However, more patients receiving LAN than OCT felt anxious before their injection. In some countries (including Canada, Ireland and the UK, but not the USA), the patient (or family member/friend) can inject LAN if they are on a stable dose, their doctor agrees, and they received training. A nurse/doctor must inject OCT. In PRESTO 2, about 40% of non-US patients who were eligible injected themselves (or were helped by a family member/friend). This may explain why more patients reported anxiety in the LAN group. PRESTO 2 provides important insights into patients’ experiences of receiving somatostatin analogs and helps identify areas for improving patient care.
Keywords
Acromegaly
Injection
Neuroendocrine tumors
Pain
Patient experience
Somatostatin analog
http://dx.doi.org/10.13039/501100014382 Ipsen
==== Body
pmcKey Summary Points
Why carry out this study?
Understanding patients’ experience of treatment is important in delivering the best care possible
The PRESTO 2 survey asked patients with acromegaly or neuroendocrine tumors (NETs) about their experience of long-acting somatostatin analogues (LA SSA: lanreotide [LAN] or octreotide [OCT]) injections), using questions including their evaluation of pain at the injection site lasting > 2 days, the effects on their daily lives, their anxiety before injections and any technical problems that occurred during injections
What was learned from the study?
Compared with those patients receiving OCT injections, significantly fewer patients receiving LAN had injection-site pain lasting > 2 days after their most recent injection
Fewer technical problems were reported with the injections received by LAN-treated patients than by those receiving OCT
The findings of the PRESTO 2 survey demonstrate the importance of injection modality in the treatment experience of patients receiving LA SSAs for acromegaly or NETs
Digital Features
This article is published with digital features, including a graphical abstract, to facilitate understanding of the article. To view digital features for this article go to 10.6084/m9.figshare.21333102.
Introduction
Injectable long-acting (LA) somatostatin analogs (SSAs) are often used long term to treat acromegaly and advanced, unresectable, moderately-to-well-differentiated neuroendocrine tumors (NETs) [1, 2]. Acromegaly is usually caused by excess production of growth hormone as a result of a benign tumor (adenoma) in the pituitary gland [3–5]. NETs are a heterogeneous group of slow-growing neoplasms that occur in diffuse neuroendocrine cells and that are accompanied, in some patients, by carcinoid syndrome [6, 7]. There are currently two proprietary (branded) injectable LA SSAs available for first-line treatment of NETs and acromegaly: Somatuline® (lanreotide) Autogel/Depot (Ipsen)—herein referred to as LAN—and Sandostatin® (octreotide) Long-Acting Release (Novartis)—herein referred to as OCT. LAN is provided as a ready-to-use prefilled syringe for administration by deep subcutaneous injection once every 28 days [8, 9]. OCT is provided in single-use kits with a vial containing octreotide powder and a syringe containing the diluent; after reconstitution, it is administered by intramuscular injection once every 28 days [10–12]. Nonproprietary (generic) LA SSAs are now available in some markets, although generic lanreotide was not available at the time of this study.
Evaluation of treatment success has traditionally focused on clinician assessments of efficacy or effectiveness and safety, but there is a growing recognition of the key role that patient experience has in overall treatment outcomes [13, 14]. Positive patient perspectives and acceptance of treatment can contribute to patient empowerment, helping patients to self-manage their condition, increase treatment adherence and persistence, and improve treatment outcomes [14]. In the case of chronic conditions requiring long-term injectable drugs, the injection experience is an important part of the overall patient reality. To maximize treatment adherence, patients should ideally have access to effective, tolerable treatments that are easy and efficient to deliver.
In 2019, Cella et al. conducted a systematic literature review to investigate which characteristics, beyond therapeutic efficacy, influenced the patient and healthcare professional (HCP) experience of LAN and OCT for the treatment of NETs or acromegaly. The review included 18 studies that reported patients’ perspectives of LAN and of OCT treatment and concluded that factors underlying the patient treatment experience included injection-associated pain, time and convenience, technical problems and emotional aspects/anxiety. Most of these studies compared LAN with OCT, and results generally favored LAN [15], but the LAN syringe has since been redesigned and the diluent for OCT has been reformulated, limiting the generalizability of these findings to current treatment options.
The LAN syringe was redesigned (v3.0) in collaboration with patients, caregivers and HCPs [16] to make the device more user-friendly and to improve its overall ergonomics. The redesigned syringe has been approved in many countries and has largely replaced the former LAN syringe. For OCT, the diluent used in the syringe was reformulated to facilitate its preparation and administration, resulting in fewer administration problems than with the previous formulation, improved ease of use and overall increased nurse satisfaction [17]. The PRESTO simulated-use study, conducted in 2019, found that nurses preferred the redesigned LAN syringe to the OCT syringe with the reformulated diluent, particularly in terms of the faster speed of administration and greater confidence that the LAN syringe will not become clogged [18]. However, no studies have explored the real-world preference or opinions of patients using the latest proprietary LAN and OCT devices and formulations.
PRESTO 2 was an online, international patient survey conducted to compare the injection experiences of patients receiving SSA therapy via the current LAN device versus the new OCT formulation for the management of NETs or acromegaly.
Methods
Study Design
PRESTO 2 was an online, English-language, patient-experience survey conducted in Canada, Ireland, the UK and the USA in adults with NETs or acromegaly who had received treatment with LAN or OCT. The survey was launched on 5 August 2020, put on hold between 31 August 2020 and 8 July 2021 on the Sponsor’s decision, to assess and review the internal processes applicable to this type of project and ended on 1 September 2021.
The survey design was based on concepts identified from the previous preference studies [18–20] and a gap analysis of the published literature, as well as input from an advisory board consisting of an expert on patient-reported outcomes (PROs), physicians and nurses. Before the survey was launched, validity of the questionnaire was tested through a cognitive debriefing with patients with NETs (n = 3) and acromegaly (n = 2).
Participants
Participants were adults (aged ≥ 18 years) who had been receiving LA SSA treatment with either LAN or OCT for NETs or acromegaly for at least the previous 3 months. Those who had received both were asked to complete the survey based on the SSA currently being used. Participants treated with LAN were included irrespective of the mode of injection (administered by an HCP or independently [self-injection or by a caregiver] in countries where independent injection was approved). Self-injection is not indicated in the label for patients receiving LAN in the USA, or for those receiving OCT in any country; these participants were therefore not asked any of the questions on independent injection.
Recruitment
Participants were recruited electronically in collaboration with patient association groups and advocacy organizations from selected English-speaking countries (Acromegaly Canada, Acromegaly Community, the Canadian Neuroendocrine Tumour Society, the Carcinoid Cancer Foundation, NET Patient Network, the Neuroendocrine Tumor Research Foundation, the Pituitary Network Alliance and the World Alliance of Pituitary Organizations) and Carenity (a social network for people living with chronic conditions), which shared a link to the online survey with their members. Respondents were asked to enter their initials, date of birth and the name of their treating center (so that quality and plausibility checks could be made). Data were pseudo-anonymized before analysis to protect participants’ confidentiality.
Primary and Secondary Endpoints
The endpoints were chosen based on those identified as most relevant to patients’ injection experiences with SSAs in the systematic literature review by Cella et al. [15]. The primary endpoint of PRESTO 2 was the proportion of participants with injection-site pain lasting > 2 days after their most recent injection of LAN or OCT in the Survey population (which included all participants who started the survey with no data quality concerns; this was irrespective of whether they had NETs or acromegaly). Secondary endpoints related to participants’ most recent injection experience were as follows: perceived pain during injection, duration of perceived pain at the injection site, extent of interference with daily life from pain at the injection site, pre-injection anxiety and occurrence of reactions (i.e., nodules, bruising or swelling) at the injection site. Secondary endpoints related to participants’ overall injection experience were as follows: frequency, details and consequences of technical injection problems, frequency of pain at the injection site and most burdensome injection-related issues. An additional secondary endpoint was the most important benefits of independent injection (non-US participants in the LAN group only). It was also planned to evaluate patient preference for LAN or OCT based on the overall injection experience and comparative experience of their current SSA with their previous SSA. However, because of the small sample size (and imbalances in sample sizes between groups), as well as the potential for recall bias, patient preference data are not presented. Safety was not in the scope of this study. Because this survey collected retrospective data on participants’ general experience regarding very well known, non-serious, listed adverse drug reactions (ADRs) for LAN, i.e., pain and local reactions at the injection site, the collection and reporting of adverse events, special situations and/or product complaints followed regulations related to spontaneous cases. In case of ADRs, special situations and/or product complaints, participants were advised to report these to National Health Authorities or their local Ipsen contact as mentioned in the informed consent form.
Statistics and Prespecified Analyses
It was assumed that the proportions of participants with pain lasting > 2 days would be approximately 25% for LAN and 40% for OCT. This assumption was based on the results of a previous study in patients with acromegaly (N = 195) in which the odds ratio (OR) of no pain days after LAN or OCT injection was 2 (74% and 57% of patients receiving LAN and OCT respectively, had no pain) [21]. Using a chi-squared test with a 5% two-sided significance level and 80% power, it was estimated that 152 participants should be enrolled in each treatment group. Based on this sample size calculation, the target sample size was 304 participants. A minimum of 76 participants in each disease cohort was to be recruited.
For the primary endpoint, the proportions of participants, and the 95% Wilson score confidence intervals (CIs) were estimated for the Survey population. The relationships between the primary endpoint and potentially relevant participant or treatment characteristics were investigated by univariate logistic regression. The potential factors investigated (country, disease [acromegaly versus NET], treatment [LAN versus OCT], age, sex [male, female], BMI, time since initiation of treatment [3–6 months vs. > 6 months to < 1 year vs. > 1 year to < 3 years vs. ≥ 3 years] and occurrence of reaction at injection side [yes vs. no]) were selected because they were perceived by the study team to have the potential to influence injection experience. Variables that were significant at a 20% level in the univariate analysis were assessed to identify potential collinearity using the Spearman’s correlation coefficient, Fisher’s exact test and Wilcoxon-Mann-Whitney tests, as appropriate. Noncollinear factors with a p value < 20% in the univariate context were included in a multivariate logistic regression model, with treatment (LAN/OCT) and disease (NETs/acromegaly) parameters forced into the model. ORs and 95% CIs were provided for all parameters included in the final model.
Secondary endpoints were evaluated for the Survey population and by treatment (LAN/OCT) and disease (NETs/acromegaly) subgroups. Endpoints relating to overall injection experience (rather than the most recent injection) were evaluated only in participants with at least 6 months’ experience with their current SSA treatment (the Overall experience population). All secondary endpoint analyses were descriptive; no formal statistical tests were performed.
Post hoc Analyses
Details of post hoc analyses are given in the supplementary materials. Briefly, analyses were conducted to evaluate injection-site pain according to disease type, the extent to which certain factors were associated with anxiety before injection, and pre-injection anxiety in those whose injections were administered by HCPs versus those whose injections were administered independently (non-US LAN group only).
Ethics
The survey was reviewed and approved by the Western Instructional Review Board (IRB protocol number 20202117). Ethical committee review was not required in Ireland or the UK. The survey was conducted in compliance with relevant regulations pertaining to the ethical conduct of patient surveys, including with the Declaration of Helsinki [22] and with the International Ethical Guidelines for Epidemiological Studies published by the Council for International Organizations of Medical Sciences [23]. All participants had to provide consent in an electronic format before starting the survey. The study was initiated and sponsored by Ipsen.
Results
Participant Disposition
Of 538 patients screened, 304 completed the survey with no data quality concerns (Survey population, Fig. 1); 168 (55.3%) were receiving LAN and 136 (44.7%) were receiving OCT (four patients were receiving generic OCT); 219 (72.0%) had NETs and 85 (28.0%) had acromegaly. Most participants (n = 281; 92.4%) had at least 6 months’ experience with their current SSA treatment (Overall experience population); 51.3% had been receiving their current SSA for at least 3 years. Forty-nine participants (29.2%) in the LAN group had previously received OCT, and 18 in the OCT group (13.2%) had previously received LAN. Of the 102 non-US participants in the LAN group, 40 (39.2%) had their most recent injection administered independently (i.e., by the patient or a partner/relative) (Independent injection population).Fig. 1 Patient disposition. aAll participants from the screened population who met the inclusion/exclusion criteria and started to complete the survey. bAll enrolled participants who completed the survey with no data quality concerns and irrespective of whether they had NETs or acromegaly; c132 patients (97.1%) had received the branded formulation, and 4 patients (2.9%) had received the generic formulation. dParticipants with at least 6 months’ experience of their current SSA treatment; e122 participants (96.8%) had received the branded formulation, and 4 participants had received (3.2%) the generic formulation. fParticipants who had been treated with both LAN and OCT. gAll participants received the branded formulation. hExcluded participants from the USA as independent injection of lanreotide autogel is not approved in the USA. LAN lanreotide autogel/depot, NET neuroendocrine tumor, OCT octreotide long-acting release
In the Survey population, mean age was 59.6 years and most participants (209/302; 69.2%) were female (Table 1A). Participants in the acromegaly subgroup were younger than those with NETs: mean age was 52.4 years in the acromegaly group and 64.2 years in the NETs group (Table 1B). In both disease subgroups, there were more female participants in the LAN group than in the OCT group (Table 1B). Among those with NETs, a high proportion had carcinoid syndrome (62.3% in the LAN group and 72.2% in the OCT group). Disease characteristics are summarized in Supplementary Table S1, stratified by treatment.
Primary Endpoint: Injection-Site Pain Lasting More Than 2 Days
Fewer patients receiving LAN than OCT had injection-site pain lasting > 2 days after the most recent injection (6.0% [95% CI 3.3–10.6%] vs. 22.8% [95% CI 16.5–30.5%]) (Fig. 2; Table 2). In the multivariate analysis, the odds of experiencing injection-site pain lasting > 2 days were significantly lower for LAN than for OCT when controlled for disease group and occurrence of injection-site reactions (OR [95% CI] 0.13 [0.06–0.30]; p < 0.0001 [country of residence, age, sex, body mass index and duration of treatment were not significant in the univariate analysis]). The odds of injection-site reactions were significantly higher in participants who experienced injection-site pain lasting > 2 days than in those who did not (OR [95% CI] 9.09 [4.00–20.65]; p < 0.0001) (Table 2).Fig. 2 Proportion of patients with pain at injection site lasting > 2 days after most recent injection (Survey populationa). aAll participants who completed the survey with no data quality concerns and irrespective of whether they had NETs or acromegaly. CI confidence interval, LAN lanreotide autogel/depot, NET neuroendocrine tumor, OCT octreotide long-acting release
Table 1 Patient and treatment characteristics in (A) Survey populationa and (B) disease subgroups stratified by treatment
(A) Characteristic Survey population, N = 304
LAN (n = 168) OCT (n = 136) All (N = 304)
Country of residence, n (%)
Canada 44 (26.2) 39 (28.7) 83 (27.3)
Ireland 28 (16.7) 13 (9.6) 41 (13.5)
UK 30 (17.9) 10 (7.4) 40 (13.2)
USA 66 (39.3) 74 (54.4) 140 (46.1)
Sex, n (%)
Male 40 (23.8) 51 (38.1) 91 (30.1)
Female 128 (76.2) 81 (60.4) 209 (69.2)
Nonbinary 0 2 (1.5) 2 (0.7)
Missing 0 2 (1.5) 2 (0.7)
Age, mean (SD), years 59.4 (10.7) 59.9 (11.5) 59.6 (11.1)
BMI category, n (%)
Underweight (< 18.5 kg/m2) 4 (2.5) 2 (1.5) 6 (2.1)
Normal weight
(18.5 to < 25.0 kg/m2)
40 (25.3) 29 (22.1) 69 (23.9)
Overweight (25 to < 30 kg/m2) 52 (32.9) 40 (30.5) 92 (31.8)
Obese (≥ 30 kg/m2) 62 (39.2) 60 (45.8) 122 (42.2)
Missing 10 5 15
Duration of treatment with LAN or OCT, n (%)
3 to 6 months 13 (7.7) 10 (7.4) 23 (7.6)
> 6 months to < 1 year 23 (13.7) 10 (7.4) 33 (10.9)
≥ 1 to < 3 years 59 (35.1) 33 (24.3) 92 (30.3)
≥ 3 years 73 (43.5) 83 (61.0) 156 (51.3)
(B) Characteristic Population with NETs,
n = 219 Population with acromegaly,
n = 85
LAN
(n = 122) OCT
(n = 97) LAN
(n = 46) OCT
(n = 39)
Country of residence, n (%)
Canada 35 (28.7) 28 (28.9) 9 (19.6) 11 (28.2)
Ireland 25 (20.5) 10 (10.3) 3 (6.5) 3 (7.7)
UK 15 (12.3) 2 (2.1) 15 (32.6) 8 (20.5)
USA 47 (38.5) 57 (58.8) 19 (41.3) 17 (43.6)
Sex, n (%)
Male 38 (31.1) 40 (41.7) 2 (4.3) 11 (28.9)
Female 84 (68.9) 55 (57.3) 44 (95.7) 26 (68.4)
Nonbinary 0 1 (1.0) 0 1 (2.6)
Missing 0 1 (1.0) 0 1 (2.6)
Age, mean (SD), years 61.9 (9.5) 63.2 (9.2) 52.9 (11.2) 51.8 (12.8)
BMI category, n (%)
Underweight (< 18.5 kg/m2) 4 (3.5) 2 (2.1) 0 0
Normal weight
(18.5 to < 25.0 kg/m2)
34 (29.8) 26 (27.4) 6 (13.6) 3 (8.3)
Overweight
(25 to < 30 kg/m2)
35 (30.7) 27 (28.4) 17 (38.6) 13 (36.1)
Obese
(≥ 30 kg/m2)
41 (36.0) 40 (42.1) 21 (47.7) 20 (55.6)
Missing 8 (6.6) 2 (2.1) 2 (4.3) 3 (8.3)
Work status, n (%)
In full or part-time work 36 (29.8) 32 (33.3) 26 (56.5) 23 (59.0)
Retired from work 55 (45.5) 48 (50.0) 6 (13.0) 8 (20.5)
Unable to work due to my disease 25 (20.7) 15 (15.6) 5 (10.9) 7 (17.9)
Not working (other reasons) 4 (3.3) 1 (1.0) 7 (15.2) 0
Other 1 (0.8) 0 2 (4.3) 1 (2.6)
Missing 1 1 0 0
Duration of treatment with LAN or OCT, n (%)
3 to 6 months 11 (9.0) 6 (6.2) 2 (4.3) 4 (10.3)
> 6 months to < 1 year 16 (13.1) 7 (7.2) 7 (15.2) 3 (7.7)
≥ 1 to < 3 years 49 (40.2) 23 (23.7) 10 (21.7) 10 (25.6)
≥ 3 years 46 (37.7) 61 (62.9) 27 (58.7) 22 (56.4)
BMI body mass index, LAN lanreotide autogel/depot, NET neuroendocrine tumor, OCT octreotide long-acting release, SD standard deviation
aAll participants who completed the survey with no data quality concerns and irrespective of whether they had NETs or acromegaly
Table 2 Primary endpoint analysis (Survey population)a
Survey population, N = 304
LAN
(n = 168) OCT
(n = 136) All
(n = 304)
Participants with perceived pain at injection site with a duration of > 2 days, n (%) [95% CI]
Yes 10 (6.0)
[3.3–10.6]
31 (22.8)
[16.5–30.5]
41 (13.5)
[10.1–17.8]
No 158 (94.0)
[89.4–96.7]
105 (77.2)
[69.5–83.5]
263 (86.5)
[82.2–89.9]
Multivariate logistic regressionb of influential participant characteristics, OR (95% CI) [p value]
LAN vs. OCT 0.13 (0.06–0.30)
[< 0.0001]
Acromegaly vs. NETs 0.77 (0.32–1.83)
[0.5518]
Injection-site reaction occurrence vs. none 9.09 (4.00–20.65)
[< 0.0001]
BMI body mass index, CI confidence interval, LAN lanreotide autogel/depot, NET neuroendocrine tumor, OCT octreotide long-acting release, OR odds ratio
aAll participants who completed the survey with no data quality concerns and irrespective of whether they had NETs or acromegaly
bCharacteristics included in the univariate analysis were country of residence (Canada, Ireland, UK, USA); disease (NETs, acromegaly); treatment (LAN, OCT); age (continuous—rather than patient age being categorized according to specific age brackets [e.g., 45–55 years], i.e., a categorical variable—all patient ages were included in the model); sex (male, female); BMI (< 18.5, 18.5 to < 25, 25 to < 30, ≥ 30 kg/m2); time since treatment initiation (3–6 months, > 6 months to < 1 year, 1 year to < 3 years, ≥ 3 years); occurrence of injection-site reactions (yes, no)
In post hoc subgroup analyses, the proportion of participants with injection-site pain lasting > 2 days after their most recent injection was lower with LAN than with OCT for both NETs and acromegaly subgroups (Fig. 2).
Secondary Endpoints
Pain, Interference with Daily Living, Local Site Reactions and Anxiety Associated with Patients’ Most Recent Injection Experience
In the Survey population, 46.4% of patients receiving LAN had injection-site pain of any duration after their most recent injection versus 58.8% with OCT (Table 3). Among those who experienced injection-site pain, 33.3% of patients receiving LAN had pain only during the injection vs. 7.5% receiving OCT.Table 3 Secondary endpoints relating to most recent injection experience (Survey population)a
Endpoint Survey population, N = 304
LAN
(n = 168) OCT
(n = 136) All
(n = 304)
Occurrence of injection-site pain, n (%) [95% CI] 78 (46.4)
[39.1–54.0]
80 (58.8)
[50.4–66.7]
158 (52.0)
[46.4–57.5]
Perceived pain intensity during injection, median (95% CI)b 4.0
(3.0–5.0)
4.0
(4.0–5.0)
4.0
(4.0–5.0)
Duration of perceived pain at injection site,b n (%) [95% CI]
Only during the injection 26 (33.3)
[23.9–44.4]
6 (7.5)
[3.5–15.4]
32 (20.3)
[14.7–27.2]
< 1 day 26 (33.3)
[23.9–44.4]
17 (21.3)
[13.7–31.4]
43 (27.2)
[20.9–34.6]
1–2 days 16 (20.5)
[13.0–30.8]
26 (32.5)
[23.2–43.4]
42 (26.6)
[20.3–34.0]
> 2 days 10 (12.8)
[7.1–22.0]
31 (38.8)
[28.8–49.7]
41 (25.9)
[19.7–33.3]
Extent of interference with daily life from perceived pain at injection site,b n (%) [95% CI]
Not at all 46 (59.0)
[47.9–69.2]
32 (40.0)
[30.0–51.0]
78 (49.4)
[41.7–57.1]
A little bit 29 (37.2)
[27.3–48.3]
42 (52.5)
[41.7–63.1]
71 (44.9)
[37.4–52.7]
Quite a bit 3 (3.8)
[1.3–10.7]
6 (7.5)
[3.5–15.4]
9 (5.7)
[3.0–10.5]
Very much 0 (0) [0.0–0.0] 0 (0) [0.0–0.0] 0 (0) [0.0–0.0]
Occurrence of reactions at injection site,c n (%) [95% CI] 76 (45.2)
[37.9–52.8]
45 (33.1)
[25.7–41.4]
121 (39.8)
[34.5–45.4]
Occurrence of pre-injection anxiety, n (%)d 100 (59.5) 59 (43.7) 159 (52.5)
CI confidence interval, LAN lanreotide autogel/depot, NET neuroendocrine tumor, OCT octreotide long-acting release
aAll participants who completed the survey with no data quality concerns and irrespective of whether they had NETs or acromegaly
bIn participants who declared injection-site pain with intensity > 0 (pain intensity was rated on a scale of 0 to 10, where higher scores represent greater pain intensity)
cNodules, bruising or swelling
dPercentages based on number of patients with data available (168 in the LAN group and 135 in the OCT group)
Among participants who experienced pain, median (95% CI) pain intensity (on a scale of 0 to 10, where higher scores represent greater intensity) was similar with LAN and OCT: 4.00 (3.00–5.00) and 4.00 (4.00–5.00)], respectively. Nevertheless, a higher proportion of patients receiving LAN than OCT reported no interference with their daily life resulting from injection-site pain from their latest injection (59.0% vs. 40.0%) (Table 3). By contrast, more patients receiving LAN than OCT reported injection-site reactions (45.2% vs. 33.1%) (Table 3). Similar results on occurrence of injection-site pain, interference with daily life and injection-site reactions were observed in the NET and acromegaly subgroups (Supplementary Table S2).
Anxiety before their most recent injection was reported by 59.5% of patients receiving LAN and 43.7% receiving OCT (Table 3). Corresponding values for pre-injection anxiety in the population with NETs were 54.9% (LAN) and 38.5% (OCT) and in the population with acromegaly were 71.7% (LAN) and 56.4% (OCT) (Supplementary Table S2).
In the post hoc multivariate analysis, the odds of pre-injection anxiety were higher for LAN than for OCT (OR [95% CI] 2.40 [1.40–4.11]; p = 0.0014). Older age decreased the odds of pre-injection anxiety (OR [95% CI]: 0.97 [0.95–1.00]; p = 0.0462), while mild and severe (vs. no) injection-site pain increased the odds (OR [95% CI] 2.19 [1.23–3.90]; p = 0.0074 and 9.47 [3.74–23.98]; p < 0.0001, respectively).
In the post hoc analysis of data from the LAN group, the proportion of participants reporting anxiety before their most recent injection was higher among those receiving injections by independent administration (72.5%) than among those receiving the injection from an HCP (55.5%).
Pain Associated with Injections Based on Participants’ Overall Injection Experience
Participants with at least 6 months’ experience with their current SSA treatment (the Overall experience population) also answered questions about their overall experience with SSA injections.
In the Overall experience population, 21 patients (13.5%) in the LAN group and 9 patients (7.1%) in the OCT group had never experienced injection-site pain. Overall, 42 patients (27.1%) in the LAN group and 54 patients (42.9%) in the OCT group reported injection-site pain most of the time (Fig. 3A and Supplementary Table S3).Fig. 3 Injection-site pain (a), injection-related issues (b) and technical injection problems (c, d) associated with LA SSA injection (Overall experience populationa). aParticipants with at least 6 months’ experience with their current SSA treatment. bParticipants were asked which symptoms/problems they had experienced—data are for those who experienced at least one issue (note that participants could choose more than one issue). cIn participants who experienced technical problems (36 participants in the LAN group and 72 in the OCT group). CI confidence interval, LA long-acting, LAN lanreotide autogel/depot, OCT octreotide long-acting release, SSA somatostatin analog
Issues Associated with Injections Based on Participants’ Overall Injection Experience
Among all participants who reported experiencing at least one issue with their injections (89.0% in the LAN group and 94.4% in the OCT group), the most common (with multiple selections allowed) was pain during injection (61.9%), followed by pain after injection (51.6%). Pain after injection was more frequent with OCT than with LAN (Fig. 3B). Other issues (see Fig. 3B) were selected by similar proportions of patients in both treatment groups, though with a tendency for higher percentages for pain during injection and anxiety before injection with LAN than with OCT (pain, 64.5% vs. 58.7%; anxiety, 28.4% vs. 19.8%). The higher frequency of anxiety before injection with LAN (than with OCT) was driven by the participants with acromegaly, among whom 43.2% receiving LAN reported anxiety compared with 22.9% receiving OCT. Within the NETs treatment group, rates of anxiety were similar for the LAN and OCT groups (22.5% and 18.7%, respectively) (Supplementary Table S3).
The most burdensome issues included local site reactions (LAN, 20.2%; OCT, 29.6%), pain during injection (LAN, 34.0%; OCT, 12.3%), pain after injection (LAN, 13.8%; OCT, 33.3%) and anxiety before injection (LAN, 18.1%; OCT, 3.7%) (Supplementary Table S3; results were generally consistent in the disease subgroups). Anxiety was more commonly reported as the most burdensome issue in those treated with LAN by patients with acromegaly (LAN, 25.0%; OCT, 4.8%) than by those with NETs (LAN, 14.5%; OCT; 3.3%).
Technical Injection Problems Based on Participants’ Overall Injection Experience
Participants with at least 6 months’ experience with their current SSA treatment reported that technical injection problems occurred less frequently with LAN than with OCT (Fig. 3C, D, and Supplementary Table S3). The syringe became blocked either before (20 participants) or during (27 participants) injections in participants in the OCT group compared with two before injection and one during injection in the LAN group. The most frequent consequence of the technical injection problems was needle replacement (meaning that the person performing the injection had to replace the needle on the syringe before the dose could be given), which occurred with 36 participants in the OCT group, but no participants in the LAN group. The whole syringe had to be replaced before the dose could be administered for 17 participants receiving OCT versus two receiving LAN. The patient’s appointment had to be rescheduled because of technical injection problems for four participants receiving OCT group and one receiving LAN.
Most Important Benefits of Independent Injection
Among the 102 non-US participants (i.e., those potentially eligible for independent injection), 40 had their most recent LAN injection administered by independent injection (NETs, n = 29; acromegaly, n = 11). The most common reasons for choosing independent injection were ‘It gave me flexibility’ (80.0%), ‘It saved time’ (70.0%), ‘The injection is easy to do’ (60.0%) and ‘It helped me feel more independent’ (45.0%). Results were similar in the NET and acromegaly subgroups, except for ‘The injection is easy to do’ (NETs, 69.0%; acromegaly, 36.4%). However, due to the small number of participants in the acromegaly subgroup, it is not possible to determine whether this difference is meaningful; further research is needed.
Discussion
Patients’ experiences of their disease management have a critical role in their adherence to treatment, and in their overall outcomes, particularly in chronic conditions [13, 14]. PRESTO 2 is the first international online patient survey comparing the latest LAN and OCT devices/formulations in patients with NETs or acromegaly. In the primary analysis, the likelihood of patients experiencing pain lasting > 2 days was significantly lower with LAN than with OCT when controlled for disease group and occurrence of injection-site reactions. These results were supported by the secondary analyses, which showed that fewer patients treated with LAN than with OCT experienced injection-site pain of any duration and fewer LAN-treated patients reported that pain after injection was one of the most burdensome issues. Also, fewer patients receiving LAN than OCT reported pain “most of the time.”
Patients reported that the injection-related issues associated with the greatest burden were pain after injection and local injection-site reactions (both more common with OCT than with LAN) as well as pain during injection and pre-injection anxiety (the latter more common with LAN than with OCT). Importantly, from the patient’s perspective, fewer participants receiving LAN than OCT reported that injection-site pain interfered with their daily living, possibly because of the smaller proportion of patients with injection-site pain lasting > 2 days in the LAN (vs. OCT) group, and relatively few patients in either treatment group reported substantial interference with daily living as a result of injection-site pain.
Pre-injection anxiety was assessed via several secondary endpoints, and the results suggested that more patients receiving LAN than OCT experienced pre-injection anxiety. These results contrast with those of a previous study in patients with NETs that reported lower rates of pre-injection anxiety with LAN than with OCT [24]. The subgroup and post hoc analyses of PRESTO 2 provide some insight into the potential drivers of anxiety: pre-injection anxiety was more common in the acromegaly (vs. NETs) disease group and in patients receiving LAN via independent (vs. HCP-administered) injections. LAN, but not OCT, is approved for independent administration, which may in part explain why pre-injection anxiety was higher in the LAN group. Greater training and education of patients/caregivers conducting independent LAN injections may help to mitigate these issues.
Compared with the LAN group, more than twice as many participants receiving OCT reported technical injection problems. These results are clinically relevant and may indicate that further improvements in the devices are warranted or that improved education around the injection process is required. As well as affecting patients, technical problems may have a negative impact on healthcare costs if they lead to syringe/device replacement and rescheduling of appointments. In the PRESTO simulated-use study, 97.8% of nurses preferred the LAN syringe to the OCT syringe, and one of the attributes most frequently rated as being the most important was confidence that the syringe would not become clogged [18].
Almost 40% of eligible (i.e., non-US) participants received independent LAN injection. The most common reasons participants gave for choosing independent injection were that it gave them flexibility, saved time, was easy to do and helped them feel more independent. These findings indicate that independent injection is a valuable treatment option for a substantial number of patients. However, the findings also indicate a potential need to increase patients’ confidence and level of training on non-HCP injection to reduce or avoid pre-injection anxiety and/or pain during injection. Fewer participants with acromegaly than with NETs specified that the reason for choosing independent injection was because the injection was easy to do; this finding may reflect the physical difficulties associated with acromegaly (i.e., larger fingers and hands, making it more difficult to hold the syringe).
Strengths and Limitations
PRESTO 2 is the first international, English-language, online survey in patients receiving current LA SSA therapies and captures direct insights from many patients in Ireland, North America and the UK. Several earlier studies have compared patients’ experiences of LAN and OCT treatment for NETs and/or acromegaly [15]; however, none of these older studies were in patients treated with the current LAN syringe. A more recent study (a prospective, single-blind, single-center “crossover” study in 51 patients with NETs) found that mean pain scores over the first three injections were similar for LAN and OCT but, unlike PRESTO 2, pain was only assessed immediately after the injection—delayed pain was not measured [25]. PRESTO 2 assessed factors identified as being important to the patient’s experiences of LAN and OCT treatment, i.e., injection pain, injection time and convenience, technical problems and emotional aspects/anxiety [15]. Furthermore, an expert in PROs tested the validity of the questionnaire through a cognitive debriefing with a small group of patients (n = 5), ensuring its robustness before use in the survey.
Another strength of PRESTO 2 was the use of patient groups and networks to help ensure the enrollment of relevant individuals. The survey also included data capture fields to validate participants as genuine recipients of SSA therapy; as such, the characteristics of the study population were generally consistent with those expected for patient populations with NETs or acromegaly. One exception is the overrepresentation of female participants, though this is typical of surveys [26]; also, sex did not have a significant effect on the primary endpoint in the logistic regression modeling, suggesting that the overrepresentation of women is not likely to have affected the findings. Finally, a minority of patients recorded ‘do not know’ answers, which implies a high-quality survey design and supports the value of including cognitive debriefings with patients with NETs or acromegaly as part of the design process.
Limitations include possible selection bias in view of recruitment of participants via patient organizations and potential recall bias, particularly related to overall injection experience. The use of an online survey may also be a source of bias in the form of variation in participation due to demographic differences (sex, age) and/or level of education. Also, the study was conducted in a small number of English-speaking countries and in English language only, limiting the ethnic diversity of the participants. The small size of the acromegaly subgroup limits the ability to draw strong disease-specific conclusions. Also, the impact (on patients’ responses/experiences) of conducting the survey during the COVID-19 pandemic is unknown. Results for LAN are specific to the current syringe design and cannot be extrapolated to earlier versions of the syringe or to generic products. Likewise, the vast majority of patients in the OCT group were being treated with proprietary OCT. In patients with NETs, there were some differences in baseline characteristics between the LAN and OCT groups, but these differences were not unexpected given the variations in the licensed indications for LAN and OCT and the dates of product approval.
Conclusions
In this survey, LAN was associated with advantages relative to OCT in terms of less prolonged pain at the injection site, less interference of injection-site pain with everyday life and fewer occurrences of technical problems. More patients receiving LAN than OCT reported anxiety before injection, although these findings may reflect the fact that LAN, but not OCT, can be administered by independent injection. Overall, the results demonstrate the importance of mode of injection and device for the injection experience of patients with acromegaly or NETs using LA SSAs. The results also highlight the impact of SSA therapy on patients’ everyday lives and the need to consider and improve this aspect of patient management.
Acknowledgements
The authors thank all patients who took part in the survey as well as the patient association groups and advocacy organizations (Acromegaly Canada, Acromegaly Community, the Canadian Neuroendocrine Tumour Society, the Carcinoid Cancer Foundation, NET Patient Network, the Neuroendocrine Tumor Research Foundation, the Pituitary Network Alliance, the World Alliance of Pituitary Organizations) and Carenity for assisting with recruitment. The authors also thank Professor David Cella, PhD, Director, Institute for Public Health and Medicine—Center for Patient-Centered Outcomes, Northwestern University Feinberg School of Medicine, Chicago, IL, USA) for his expert input on the survey design and for overseeing the validity testing of the questionnaire.
Funding
This study was sponsored by Ipsen. The sponsor was involved in the design of the study, analysis and interpretation of the data, and review and funding of the manuscript and its publication costs, including the journal’s Rapid Service Fees.
Medical Writing, Editorial and Other Assistance
The authors thank Nicky French (PhD) and Tamzin Gristwood (PhD) of Oxford PharmaGenesis, Oxford, UK, who provided medical writing and editorial support, which was sponsored by Ipsen in accordance with Good Publication Practice (GPP3) guidelines.
Author Contributions
Substantial contributions to study conception/design, or acquisition/analysis/interpretation of data: all authors. Drafting of the publication or revising it critically for important intellectual content: all authors. Final approval of the publication: all authors.
Disclosures
D. O’Toole: no conflict of interest to declare. P.L. Kunz: participated in advisory boards for Amgen, Crinetics, Genentech, HutchMed, Ipsen, Natera, Novartis (Advanced Accelerator Applications) and RayzeBio and received research funding from Novartis (Advanced Accelerator Applications). S.M. Webb: was a consultant for, and received honoraria from, Ipsen, Novartis and Recordati. G. Goldstein: was a consultant for Advanced Accelerator Applications and Ipsen. S. Khawaja: has nothing to disclose. M. McDonnell: undertakes voluntary work for NET Patient Network, which receives funding from Ipsen, Novartis and Pfizer for patient initiatives. S. Boiziau, D. Gueguen, A. Houchard and A. Ribeiro-Oliveira Jr. are employees of Ipsen: ARO Jr has stocks from Ipsen. A. Prebtani: acts as a consultant and speaker for, and conducts research on behalf of, Ipsen.
Compliance with Ethics Guidelines
The survey was reviewed and approved by the Western Instructional Review Board (IRB protocol number 20202117). Ethical committee review was not required in Ireland or the UK. The survey was conducted in compliance with relevant regulations pertaining to the ethical conduct of patient surveys, including with the Declaration of Helsinki [22] and with the International Ethical Guidelines for Epidemiological Studies published by the Council for International Organizations of Medical Sciences [23]. All participants had to provide consent in an electronic format before starting the survey.
Prior Presentation
Parts of this manuscript, including presentations of information in Figs. 2 and 3 were previously presented at the Endocrine Society Annual Conference (ENDO 2022), June 11–14, 2022, Atlanta USA, the European Congress of Endocrinology (ECE 2022), 21–24 May 2022, Milan, Italy and at the Annual Conference of the European Neuroendocrine Tumor Society (ENETS) 2022, 10–11 March 2022, Barcelona, Spain.
Data Availability
Qualified researchers may request access to patient-level study data that underlie the results reported in this publication. Additional relevant study documents, including the clinical study report, study protocol with any amendments, annotated case report form, statistical analysis plan and dataset specifications may also be made available. Patient level data will be anonymized, and study documents will be redacted to protect the privacy of study participants. Where applicable, data from eligible studies are available 6 months after the studied medicine and indication have been approved in the US and EU or after the primary manuscript describing the results has been accepted for publication, whichever is later. Further details on Ipsen's sharing criteria, eligible studies and process for sharing are available here (https://vivli.org/members/ourmembers/). Any requests should be submitted to www.vivli.org for assessment by an independent scientific review board.
==== Refs
References
1. Caplin ME Pavel M Ćwikła JB Phan AT Raderer M Sedláčková E Lanreotide in metastatic enteropancreatic neuroendocrine tumors N Engl J Med 2014 371 3 224 233 10.1056/NEJMoa1316158 25014687
2. Strosberg J Kvols L Antiproliferative effect of somatostatin analogs in gastroenteropancreatic neuroendocrine tumors World J Gastroenterol 2010 16 24 2963 2970 10.3748/wjg.v16.i24.2963 20572298
3. Melmed S Medical progress: acromegaly N Engl J Med 2006 355 24 2558 2573 10.1056/NEJMra062453 17167139
4. Melmed S Bronstein MD Chanson P Klibanski A Casanueva FF Wass JAH A consensus statement on acromegaly therapeutic outcomes Nat Rev Endocrinol 2018 14 9 552 561 10.1038/s41574-018-0058-5 30050156
5. Vilar L Vilar CF Lyra R Lyra R Naves LA Acromegaly: clinical features at diagnosis Pituitary 2017 20 1 22 32 10.1007/s11102-016-0772-8 27812777
6. Grozinsky-Glasberg S Davar J Hofland J Dobson R Prasad V Pascher A European neuroendocrine tumor society (ENETS) 2022 guidance paper for carcinoid syndrome and carcinoid heart disease J Neuroendocrinol 2022 2 e13146
7. Shah MH, Goldner WS, Benson AB, Bergsland E, Blaszkowsky LS, Brock P, et al. Neuroendocrine and Adrenal Tumors, Version 2.2021, NCCN Clinical Practice Guidelines in Oncology. Journal of the National Comprehensive Cancer Network : JNCCN. 2021;19(7):839–68.
8. Ipsen Ltd. Summary of product characteristics, Somatuline Autogel 60 mg, 90 mg, 120 mg, solution for injection in a prefilled syringe. European Medicines Compendium 2021 [Available from: https://www.medicines.org.uk/emc/product/4808/smpc.
9. Ipsen Pharma. Prescribing Information, Somatuline®Depot (lanreotide) injection for subcutaneous use 04/2019 [Available from: https://www.accessdata.fda.gov/drugsatfda_docs/label/2019/022074s024lbl.pdf.
10. Novartis Pharmaceuticals Corporation e. Prescribing information, Sandostatin LAR® Depot (octreotide acetate for injectable suspension) 03/2021 [Available from: https://www.accessdata.fda.gov/drugsatfda_docs/label/2021/021008s041lbl.pdf.
11. Novartis Pharmaceuticals UK Ltd. Prescribing information, Sandostatin LAR 10 mg powder and solvent for suspension for injection. European Medicines Agency 2018 [Available from: https://www.medicines.org.uk/emc/product/1038/smpc.
12. Teva UK Limited. Prescribing information, Olatuton 30 mg Powder and Solvent for Prolonged-release Suspension for Injection. European Medicines Agency 2018 [Available from: https://www.medicines.org.uk/emc/product/10841/smpc.
13. Forestier B Anthoine E Reguiai Z Fohrer C Blanchin M A systematic review of dimensions evaluating patient experience in chronic illness Health Qual Life Outcomes 2019 17 1 19 10.1186/s12955-019-1084-2 30665417
14. Glasgow RE Wagner EH Schaefer J Mahoney LD Reid RJ Greene SM Development and validation of the patient assessment of chronic illness care (PACIC) Med Care 2005 43 5 436 444 10.1097/01.mlr.0000160375.47920.8c 15838407
15. Cella D Evans J Feuilly M Neggers S Van Genechten D Herman J Patient and healthcare provider perspectives of first-generation somatostatin analogs in the management of neuroendocrine tumors and acromegaly: a systematic literature review Adv Ther 2021 38 2 969 993 10.1007/s12325-020-01600-x 33432541
16. Adelman DT Van Genechten D Megret CM Truong Thanh XT Hand P Martin WA Co-creation of a lanreotide autogel/depot syringe for the treatment of acromegaly and neuroendocrine tumours through collaborative human factor studies Adv Ther 2019 36 12 3409 3423 10.1007/s12325-019-01112-3 31612358
17. Delemer B Nguyen-Tan-Hon T Coriat R Smith D Schillo F Raingeard I Evaluation of nurses' and patients' overall satisfaction with new and previous formulations of octreotide long-acting release (sandostatin LAR(®)): a French observational study Adv Ther 2020 37 9 3901 3915 10.1007/s12325-020-01429-4 32683667
18. Adelman D Truong Thanh XM Feuilly M Houchard A Cella D Evaluation of nurse preferences between the lanreotide autogel new syringe and the octreotide long-acting release syringe: an international simulated-use study (PRESTO) Adv Ther 2020 37 4 1608 1619 10.1007/s12325-020-01255-8 32157626
19. Adelman DT Burgess A Davies PR Evaluation of long-acting somatostatin analog injection devices by nurses: a quantitative study Medical devices (Auckland, NZ) 2012 5 103 109
20. Ryan P Phan AT Adelman DT Iwasaki M Neuroendocrine tumors and lanreotide depot: clinical considerations and nurse and patient preferences Clin J Oncol Nurs 2016 20 6 E139 E146 10.1188/16.CJON.E139-E146 27857269
21. Strasburger CJ Karavitaki N Störmann S Trainer PJ Kreitschmann-Andermahr I Droste M Patient-reported outcomes of parenteral somatostatin analogue injections in 195 patients with acromegaly Eur J Endocrinol 2016 174 3 355 362 10.1530/EJE-15-1042 26744896
22. World Medical Association World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects JAMA 2013 310 20 2191 2194 10.1001/jama.2013.281053 24141714
23. Council for International Organizations of Medical Sciences. International ethical guidelines for epidemiological studies. Available from: https://cioms.ch (Accesed 20 December 2021) 2009.
24. Ström T Kozlovacki G Myrenfors P Almquist M Patient and nurse experience of using somatostatin analogues to treat gastroenteropancreatic neuroendocrine tumors: results of the somatostatin treatment experience trial (STREET) Patient Prefer Adherence 2019 13 1799 1807 10.2147/PPA.S213472 31695341
25. Raj N Cruz E O'Shaughnessy S Calderon C Chou JF Capanu M A randomized trial evaluating patient experience and preference between octreotide long-acting release and lanreotide for treatment of well-differentiated neuroendocrine tumors JCO Oncol Pract. 2022 2 200055
26. Smith G. Does gender influence online survey participation?: A record-linkage analysis of university faculty online survey response behavior. ERIC Document Reproduction Service No. ED 501717 2008.
| 36502449 | PMC9741754 | NO-CC CODE | 2022-12-15 23:15:16 | no | Adv Ther. 2022 Dec 11;:1-20 | utf-8 | Adv Ther | 2,022 | 10.1007/s12325-022-02360-6 | oa_other |
==== Front
Indian J Orthop
Indian J Orthop
Indian Journal of Orthopaedics
0019-5413
1998-3727
Springer India New Delhi
777
10.1007/s43465-022-00777-3
Original Article
Evaluation of Anthropometric Measurements of the Aspect Ratio of Knee in Indian Population and its Correlation with the Sizing of Current Knee Arthroplasty System
Mukhopadhaya John [email protected]
Kashani Andalib [email protected]
Kumar Nishikant [email protected]
http://orcid.org/0000-0002-3329-8355
Bhadani Janki S. [email protected]
Department of Orthopaedics, Paras HMRI Hospital, Patna, Bihar 800014 India
11 12 2022
17
14 8 2022
7 11 2022
© Indian Orthopaedics 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
Most of the commercially available TKR implants are designed for western populations, which are known to have larger build and stature compared to Asian counterparts often leading to mismatch between resected bony surfaces and implant components. There is paucity of morphometric data of distal femur and proximal tibia in the Indian population. Thus, it becomes important to obtain anthropometric data to achieve the best stability and long-term success of implant.
Materials and Methods
Intraoperative morphological measurements of 100 knees (59 female and 41 males) were done using vernier calliper during TKR. The anteroposterior (AP) and mediolateral (ML) dimensions of cross-section of the femur and tibia were noted before bony resection. The aspect ratios were calculated and compared with that of implant used (DePuy, Stryker, Maxx).
Results
We have found that Indian males have larger dimensions of distal femur as well as proximal tibia than females. There exists some degree of mismatch in patients’ dimensions and the sizes of all the three commercially available implant system as well their aspect ratios.
Conclusion
Specific designing of implants with dimensions in accordance with the morphometric measurements of Indian population should be done. Also gender specific implant designing should be done.
Keywords
TKR
Aspect ratio
Mismatch
Anthropometric measurement
Calliper
Arthroplasty
DePuy
Stryker
Maxx
==== Body
pmcIntroduction
In advanced knee osteoarthritis, total knee replacement provides a safe, well-tolerated, and cost-effective treatment [1]. It is preferably recommended for those for whom other modalities of management are contraindicated or have failed [2]. TKR is a precise surgical procedure. The long-term success of the TKR depends largely on accurate bone cutting, adequate soft tissue balancing, restoring the mechanical alignment and prosthesis selection that best matches the sizes of the resected surfaces of the distal femur and proximal tibia [3–5]. The component overhang may interfere with soft tissue balance which in turn can lead to post-operative knee pain and decreased range of motion. The undersized tibial component can result in subsidence and loosening which leads to instability, malalignment of the joint, and finally early TKR failure [6–8]. A properly matched prosthesis can provide the optimum coverage over the resected surfaces and avoid soft tissue abrasion. Most of the commercially available TKR implants are designed for western populations, which are known to have larger build and stature compared to Asian counterparts [6, 9]. This often leads to a mismatch between resected bony surfaces and implant components. There is a paucity of morphometric data on the distal femur and proximal tibia in the Indian population. Thus, it becomes important to obtain anthropometric data to achieve the best stability and long-term success of the implant. The study was conducted in a tertiary centre of eastern India to examine the distal femoral and proximal tibial morphometry of Indian patients, and to evaluate anatomical differences between males and females. Data obtained in this study will provide key information for designing optimal size-matched femoral and tibial components for total knee arthroplasty for the Indian population.
Materials and Methods
Under prospective observational study a total of 100 knee replacements in 81 patients between December 2017 to June 2021 were analysed. The patients were evaluated with no anatomical abnormality, knee injury and other knee pathologies. The DePuy (PFC, Sigma), Stryker (Scorpio, NRG), Maxx (Freedom) PS semi-constrained, fixed bearing systems were used. Standard surgical procedures were adhered to, midline longitudinal skin incision with medial para-patellar approach were used. Bicompartmental knee replacement was done, resurfacing of the patella was done. After soft tissue release all the visible osteophytes were removed. The measurement of distal femur included the most proximal part of the junction between bone and articular surface. The anterior–posterior (AP) length was measured using vernier- calliper based on the most anterior and most posterior points. The medial–lateral (ML) width was measured from the most medial and most lateral point along the trans-epicondylar axis. The measurement of proximal tibia was taken similarly before the bone cuts. The tibial ML dimension was measured as the longest medio-lateral line of the proximal tibia. The tibial AP dimension was taken as the length of the line drawn passing through the mid-point of and perpendicular to the ML line. The tibial and femoral aspect ratio were calculated as ML/AP X 100.
Statistical Analysis
The initial data thus obtained were captured into the customized proforma and then transferred to Microsoft excel. The data were analysed using online statistical software, the mean standard deviation was calculated. Student ‘t’ test was used to compare the mean values between the two groups. Final data are presented in the form of tables and graphs. Using Minitab tool scatter plots of data were drawn, the linear regression lines of different sets of data were drawn and compared. Thus, gender mismatch and also the mismatch between patients’ morphological data and that of the implant’s sizes used were calculated and appreciated.
Results
In our study, out of 100 knees in 81 patients, 59 (59%) were females and 41 (41%) were males. Out of 100 cases, we used DePuy (PFC, Sigma) in 61 cases, Stryker (Scorpio, NRG) in 34 cases and Maxx (Freedom) in 5 cases. Following observations were made regarding the intraoperative morphometric measurements of distal femur and proximal tibia using vernier calliper. The data were summarised as the mean and standard deviation and the mismatch calculated.
The mean AP as well as ML dimension of the proximal tibia as well as for distal femur was more in males, and these are statistically significant (P < 0.001). There is no statistically significant difference between the tibial aspect ratio as well as femoral aspect ratio of males and females. (Table 1).Table 1 Dimensions of proximal tibia and distal femur of our study and their statistical significance
Male Female Total P value
Number 59 41 100
The mean AP dimensions (In mm) of the proximal tibia (minimum–maximum) 51.731 (41–60) 47.152 (40–58) 49.03 (40–60) < 0.001
The mean ML dimensions (In mm) of the proximal tibia, (minimum–maximum) 78.39 (66–91) 73.067 (64–91) 75.25 (64–91) < 0.001
Tibial aspect ratio 1.524 (1.29–1.82) 1.558 (1.37–1.73) 1.544 (1.29–1.82) > 0.05
The mean AP dimensions (In mm) of the distal femur, (minimum–maximum) 65.073 (56–74) 58.796 (51–70) 61.37 (51–74) < 0.001
The mean ML dimensions (In mm) of the distal femur, (minimum–maximum) 84.536 (73–95) 77.271 (70–87) 80.25 (70–95) < 0.001
Femoral aspect ratio 1.3 (1.18–1.42) 1.316 (1.13–1.48) 1.309 (1.13–1.48) > 0.05
Mismatch in tibial dimension vs implant used were studied and found that the mean mismatch for the tibial AP as well as ML dimension is minimum for DePuy and maximum for Stryker. The mean tibial aspect ratio mismatch is minimum for DePuy and maximum for Maxx. The mean femoral AP dimension mismatch is minimum for Maxx and maximum for Stryker. The mean femoral ML dimension mismatch is minimum for Stryker and maximum for Maxx the mean femoral aspect ratio mismatch is minimum for Stryker and maximum for Maxx, and it is statistically significant. (Table 2).Table 2 Mismatch (implant wise) of tibial and femoral side of our study and their statistical significance
Implant (N) Mismatch (implant wise) tibial side Mismatch (implant wise) femoral side
(a) Mean tibia-AP mismatch (b) Mean tibia-ML mismatch (c) Mean tibial AR mismatch (a) Mean femoral-AP mismatch (b) Mean femoral-ML mismatch (c) Mean femoral AR mismatch
Depuy (61) 3.885 (− 1 to 14) 7.508 (2 to 20) 2.786 (− 5 to 12) 17.728 (8 to 32) 0.04 (− 0.2 to 0.31) 0.236 (0.08 to 0.41)
Stryker (34) 6 (− 10 to 14) 10.97 (− 3 to 16) 2.764 (− 10 to 11) 15.058 (4 to 23) 0.05 (− 0.12 to 0.3) 0.19 (0.03 to 0.29)
Maxx (5) 5.4(2 to 8) 10 (8 to 13) 2.2 (1 to 4) 19.8 (16–23) 0.104 (− 0.07 to 0.13) 0.292 (0.2 to 0.35)
Total (100) 16.65 (4–32) 0.04 (− 0.2 to 0.31) 0.223 (0.03 to 0.41)
P Value > 0.05 < 0.001 > 0.05 < 0.05 > 0.05 < 0.001
Femoral ML Vs AP measurement (mm) for 100 knees shows the mismatch between the patient’s morphological data and the dimensions of all the prosthetic systems used (Fig. 1a). Femoral ML vs AP measurement for 59 female patients showing the prosthetic designs tended to be smaller than the morphological data in the ML dimension for a given AP measurement Femoral ML vs AP measurement for 41 male patients showed the prosthetic designs tended to be smaller than the morphological data in the ML dimension for a given AP measurement. Femoral aspect ratio for the morphological data for females showed a higher ratio for smaller knees and proportionally lower ratios for larger knees (Fig. 1c). But the implants have a relatively constant aspect ratio. While the femoral aspect ratio for the morphological data for males showed a higher ratio for smaller knees and a proportionally lower ratio for larger knees (as the AP dimension increased). But the implants have a constant aspect ratio, in case of DePuy the ratio actually increased.Fig. 1 Scatter plots of the observations tibial side: a The tibial medial–lateral (ML) measurement versus the anteroposterior (AP) measurements (mm) for 100 knees show a close approximation of the implant size to the morphological data. Note: approximation less profound for Maxx b The tibial aspect ratio for the morphological data for males showed a higher ratio for smaller knees and a proportionally lower ratio for larger knees. But the Depuy and Stryker showed a constant aspect ratio as the AP dimension increased. c The tibial aspect ratio for the morphological data for females showed a higher ratio for smaller knees and proportionally lower ratios for larger knees. It shows as the AP dimension increased the aspect ratio decreased. But the DePuy shows a constant aspect ratio and Maxx showed an increasing aspect ratio as the AP dimension increased
Tibial ML measurement versus AP measurements (mm) for 100 knees show a close approximation of the implant size to the morphological data (approximation less profound for Maxx). Tibial ML measurements versus AP measurements for 59 female knees show a close approximation of the implant size to the morphological data for larger knees. The smaller knees had too small implant size for a given AP dimensions. Tibial ML versus AP measurements of 41 male knees show a close approximation of the implant size to the morphological data for larger knees. The small knees had too small implant sizes in AP and ML dimensions. Tibial aspect ratio for the morphological data for females showed a higher ratio for smaller knees and proportionally lower ratios for larger knees (Fig. 2a). It shows as the AP dimension increased the aspect ratio decreased. But the DePuy shows a constant aspect ratio and Maxx showed an increasing aspect ratio as the AP dimension increased. Tibial aspect ratio for the morphological data for males showed a higher ratio for smaller knees and a proportionally lower ratio for larger knees (Fig. 2b). But the DePuy and Stryker showed a constant aspect ratio as the AP dimension increased.Fig. 2 Scatter plots of the observations femoral side: the best fit lines were calculated a Femur ML Vs AP measurement (mm) for 100 knees shows the mismatch between the patient’s morphological data and the dimensions of all the prosthetic systems used. b The femoral aspect ratio for the morphological data for males showed a higher ratio for smaller knees and a proportionally lower ratio for larger knees (as the AP dimension increased). But the implants have a constant aspect ratio, in case of DePuy the ratio actually increased c The femoral aspect ratio for the morphological data for females showed a higher ratio for smaller knees and proportionally lower ratios for larger knees. But the implants have a relatively constant aspect ratio
This study investigated the anthropometric measurements of the aspect ratio of the knee in Indian population and its correlation with the sizing of current knee arthroplasty system. The data indicated that male patients have larger distal femoral and proximal tibial dimensions as compared to females and difference is found to be maximum in ML dimensions. There is no statistically significant difference between the tibial and femoral aspect ratio of male and female. There exists certain degree of mismatch between the patient’s morphological data and dimensions of implant used in all three implant systems. The mean mismatch for tibial AP and ML dimension is minimum for DePuy and maximum for Stryker and it is statistically significant. The mean tibial aspect ratio mismatch is minimum for DePuy and maximum for Maxx. The mean femoral AP dimension mismatch is minimum for Maxx and maximum for Stryker. The mean femoral ML dimension mismatch is minimum for Stryker and maximum for Maxx. The mean femoral aspect ratio mismatch is minimum for Stryker and maximum for Maxx and it is statistically significant. For better understandings of the mismatch trends and the disparity in aspect ratios we have drawn scatter plots and following observations were made.
The scatter plot for femoral ML vs AP measurement for both female and male patients showed the prosthetic designs tended to be smaller than the morphological data in the mediolateral dimension for a given anteroposterior measurement. The scatter plot for femoral aspect ratio Vs AP dimension of the morphological data for males and females showed a higher ratio for smaller knees and a proportionally lower ratios for larger knees. But the implants have a relatively constant aspect ratio (in case of DePuy the ratio actually increased in males). The scatter plot of tibial medial–lateral (ML) measurement versus the anteroposterior (AP) measurements (mm) for 100 knees showed close approximation of the implant size to the morphological data. The approximation is less profound for Maxx. The scatter plot for tibial ML measurements versus AP measurements for both male and female knees showed close approximation of the implant size to the morphological data for larger knees but the smaller knees had too small implant size for a given AP dimensions (or ML dimensions for males.). The scatter plot of tibial aspect ratio for the morphological data for males and females showed a higher ratio for smaller knees and a proportionally lower ratios for larger knees. It shows as the AP dimension increased the aspect ratio decreased. But the DePuy and Stryker shows a constant aspect ratio and Maxx showed an increasing aspect ratio as the AP dimension increased.
Discussion
We have compared our results with previous similar studies [10–20]. There are multiple similar studies done to find mean femoral AP and ML and aspect ratio in Indian, American, Chinese, Malaysian and Thai populations. (Table 3) [10, 12, 14–18, 20]. Mean femoral AP and ML and aspect ratio of our study is comparable to that of Yue et al [12] in Chinese population. Similarly, the AP and ML tibial dimension and the aspect ratio were studied in Indian, Japanese, Korean, Thai, and, Chinese population. [11–16]. The findings of our study are comparable that of Shah et al [14] over Indian Population. The mismatch between the morphological dimension of our patients and the size of implants is more for femoral ML dimensions. Similar findings were also observed by Thilak et al [17] and Hitt et al [6]. Uehera et al [11] had found tibial AP mismatch in his study.Table 3 Comparison of dimensions of proximal tibia and distal femur of our study with previous literature
Series Year Study population No. of knees Mean AP diameter (in mm) Mean ML diameter (in mm) Aspect ratio
(A) Comparison of proximal tibial dimensions
Uehara et al. [11] 2002 Japanese 100 M 54.1 ± 3.0
F 49.2 ± 2.9
M 77.9 ± 4.1
F 69.5 ± 3.4
–
Kwak et al. [13] 2007 Korean 200 M 48.2 ± 3.3
F 43.2 ± 2.3
M 76.1 ± 4.0
F 67.6 ± 3.1
–
Chaichankul et al. [15] 2009 Thai 200 M 50.15 ± 3.1
F 43.23 ± 2.6
M 74.44 ± 3.4
F 64.95 ± 3.5
–
Yue et al. [12] 2011 Chinese 40 M 41.5 ± 2.1
F 37.3 ± 2.8
M 75.2 ± 3.6
F 66.2 ± 2.1
M 1.82 ± 0.07
F 1.78 ± 0.10
Caucasian 36 M 45.0 ± 2.8
F 39.3 ± 2.6
M 78.7 ± 5.4
F69.0 ± 4.2
M1.75 ± 0.11
F1.76 ± 0.08
Li et al. [16] 2014 Chinese 148 M 49.6 ± 2.4
F 44.2 ± 2.3
M 77.4 ± 3.3
F 69.1 ± 2.8
M 1.56 ± 0.07
F 1.56 ± 0.06
Caucasians 127 M 49.5 ± 2.9
F 45.2 ± 2.3
M 79.4 ± 4.3
F 70.2 ± 2.7
M 1.61 ± 0.08
F 1.54 ± 0.07
Shah et al. [14] 2014 Indian 66 M 53.9 ± 2.8
F 48.5 ± 3.9
M 77.5 ± 4.0
F 70.0 ± 3.4
M 1.44 ± 0.05
F 1.45 ± 0.09
Agrawal et al. [22] 2017 Indian 60 M48.9 ± 2.6
F 43.1 ± 3.0
M 74.7 ± 3.6
F65.8 ± 3.9
M1.52 ± 0.08
F 1.52 ± 0.08
Our study 2021 Indian 100 M 51.73 ± 3.9
F 47.15 ± 4.2
M 78.39 ± 4.4
F 73.06 ± 3.6
M 1.52 ± 0.1
F 1.55 ± 0.9
(B) Comparison of distal femoral dimensions
Mensch et al. [18] 1975 American 30 – M 82.1 ± 4.7
F 69.9 ± 2.6
–
Berger et al. [20] 1993 American 75 M 68.1 ± 4.6
F 60.2 ± 2.0
M 85.6 ± 5.1
F 75.4 ± 2.3
–
Vaidya et al. [10] 2000 Indian 47 M 57.55 ± 3.3
F 56.48 ± 1.7
M 68.58 ± 3.2
F 61.98 ± 3.6
–
Chaichankul et al. [15] 2009 Thai 200 M 48.55 ± 3.7
F 43.32 ± 3.7
M 70.15 ± 3.9
F 64.06 ± 6.3
–
Ewe et al. [21] 2009 Malaysian 69 M 64.55 ± 6.1
F 59.14 ± 4.2
M 72.45 ± 5.6
F 63.83 ± 3.8
M 1.13 ± 0.06
F 1.08 ± 0.07
Yue et al. [12] 2011 Chinese 40 M 65.0 ± 2.8
F 58.8 ± 2.5
M 82.6 ± 3.6
F 72.8 ± 2.6
M 1.27 ± 0.03
F 1.24 ± 0.04
Caucasian 36 M 67.5 ± 3.6
F 59.7 ± 2.6
M 86.0 ± 5.6
F76.4 ± 4.0
M 1.28 ± 0.07
F 1.28 ± 0.06
Li et al. [16] 2014 Chinese 148 M 56.5 ± 2.5
F 52.8 ± 2.6
M 72.7 ± 3.8
F 64.4 ± 2.6
M 1.29 ± 0.04
F 1.22 ± 0.05
Caucasians 127 M 59.6 ± 3.2
F 55.4 ± 2.8
M 74.6 ± 3.9
F 65.4 ± 1.4
M 1.25 ± 0.05
F 1.18 ± 0.05
Shah et al [13] 2014 Indian 66 M 65.6 ± 3.8
F59.8 ± 4.3
M71.5 ± 2.5
F65.1 ± 3.1
M1.09 ± 0.04
F1.09 ± 0.05
Thilak et al. [14] 2016 Indian 150 M 66.91 ± 3.3
F 59.38 ± 4.1
M 78.55 ± 4.8
F 66.88 ± 4.7
Agrawal et al. [22] 2017 Indian 60 M 62.7 ± 3.2
F 59.4 ± 4.1
M 76.4 ± 3.2
F 67.4 ± 3.7
M 1.21 ± 0.4
F 1.13 ± 0.5
Current study 2021 Indian 100 M 65.07 ± 3.6
F 58.79 ± 7.9
M 84.53 ± 4.8
F 77.27 ± 4.05
M 1.30 ± 0.6
F 1.31 ± 0.6
Thus, various studies have proved that there exists a wide variation in anatomical dimensions of different races and gender. Most of the prosthesis available for knee and hip replacements in India are either imported or are copy of the imported ones. There have been few data on the Indian anatomy. This study proves the lack of concurrence in available implants and the need of designing an implant which suits the anatomy of Indian population.
One of the limitations of our study is that it is a calliper based morphometric measurements while most of the studies used 3D CT based measurements like the one did by Agrawal et.al. [22] in Indian population. Larger sample size and better investigative tools (3D CT, MRI) needed for data collection on larger scale for recommendation of specific implant designing’s for Indian population for TKR. The number of patients /elective surgeries were significantly reduced due to COVID-19 pandemic which coincided with our study period. There is a lack of control group in our study. More such studies are needed in various subsets of Indian population keeping in mind presence of various ethnic races in our country Future studies should compare these anatomic knee data with the TKA dimensions used in surgeries.
Conclusion
Proximal tibial and distal femoral dimensions of male and female vary in Indians. Indian males have larger dimensions than females, but there is no statistically significant difference between the aspect ratio of males and females. There exists some degree of mismatch in patients’ dimensions and the sizes of all the three commercially available implant system which are designed on the basis of western population’s dimensions. The aspect ratio of none of the three commercially available implant system matches with that of the aspect ratio of the Indian patients. Aspect ratio of the patients are higher for smaller knees and lower for larger knees but the aspect ratio the implants either remains constant or increases for larger knees. Based on our study we can conclude that specific designing of implants with dimensions in accordance with the morphometric measurements of Indian population should be done. We also need to have gender specific implant designing. The implant designing should be such that the aspect ratio of implant should match with the aspect ratio of the patients.
Acknowledgements
None.
Funding
None.
Declarations
Conflict of Interest
We the authors declare that we have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Ethical Approval
Yes.
Informed Consent
For this type of study informed consent is not required.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Chen AT Bronsther CI Stanley EE Paltiel AD Sullivan JK Collins JE Neogi T Katz JN Losina E The value of total knee replacement in patients with knee osteoarthritis and a body mass index of 40 kg/m2 or greater: a cost-effectiveness analysis Annals of Internal Medicine 2021 174 6 747 757 10.7326/M20-4722 33750190
2. Varacallo M Luo TD Johanson NA Total Hip Arthroplasty Techniques StatPearls [Internet], Treasure Island (FL) 2022 StatPearls Pusblishing
3. Insall JN Dorr LD Scott RD Scott WN Rationale of the knee society clinical rating system Clin Orthopaedics Related Res 1989 248 13 14 10.1097/00003086-198911000-00004
4. Westrich GH Haas SB Insall JN Frachie A Resection specimen analysis of proximal tibial anatomy based on 100 total knee arthroplasty specimens The Journal of Arthroplsty 1995 10 47 51 10.1016/S0883-5403(06)80064-1
5. Pathak SK Sethi M Salunke AA Thivari P Gautam RK Anjum R Chawla J Sharma A Is flexion gap rectangular in native Indian knees? Results of an MRI study Indian Journal of Orthopedics 2021 55 5 1127 1134 10.1007/s43465-021-00418-1
6. Hitt K Shurman JR II Greene K McCarthy J Moskal J Hoeman T Anthropometric measurements of the human knee: correlation to the sizing of current knee arthroplasty systems Journal of Bone and Joint Surgery 2003 85-A Suppl 4 115 122 10.2106/00004623-200300004-00015
7. Westrich GH Agulnick MA Laskin RS Current analysis of tibial coverage in total knee arthroplasty The Knee 1997 4 87 91 10.1016/S0968-0160(96)00243-8
8. Stulberg BN Dombrowski RM Froimson M Easley K Computed tomography analysis of proximal tibial coverage Clinical Orthopaedics Related Reserch 1995 311 148 156
9. Mahfouz M Abdel Ftah Bowers Scuderi EELSG Three-dimensional morphology of the knee reveals ethnic differences Clinical Orthopaedics and Related Research 2012 470 172 185 10.1007/s11999-011-2089-2 21948324
10. Vaidya SV Ranawat CS Aroojis A Laud NS Anthropometric measurements to design total knee prostheses for the Indian population Journal of Arthroplasty 2000 15 79 85 10.1016/S0883-5403(00)91285-3 10654467
11. Uehara K Kadoya Y Kobayashi A Ohashi H Yamano Y Anthropometry of the proximal tibia to design a total knee prosthesis for the Japanese population Journal of Arthroplasty 2002 17 1028 1032 10.1054/arth.2002.35790 12478514
12. Yue B Varadarajan KM Ai S Tang T Rubash HE Li G Differences of knee anthropometry between Chinese and white men and women Journal of Arthroplasty 2011 26 124 130 10.1016/j.arth.2009.11.020
13. Kwak DS Sibin S Patinharayil G Morphometry of the proximal tibia to design the tibial component of total knee arthroplasty for the Korean population The Knee 2007 14 295 300 10.1016/j.knee.2007.05.004 17600719
14. Shah DS Ghyar R Ravi B Shetty V (2013) 3D morphological study of the Indian arthritic knee: comparison with other ethnic groups and conformity of current TKA implant Open Journal of Rheumatology and Autoimmune Diseases 2013 3 263 269 10.4236/ojra.2013.34041
15. Chaichankul C Tanavalee A Itiravivong P Anthropometric measurements of knee joints in Thai population: correlation to the sizing of current knee prostheses The Knee 2011 18 1 5 24 10.1016/j.knee.2009.12.005 20133135
16. Li P Tsai T-Y JingSheng Li Yu Zhang Y-M Kwon HE Rubash GL Morphological measurement of the knee: race and sex effects Acta Orthopaedica Belgica 2014 80 260 268 25090801
17. Thilak J George MJ Patient implant dimension mismatch in total knee arthroplasty: is it worth worrying? An Indian scenario Indian J Orthop 2016 50 512 517 10.4103/0019-5413.189618 27746494
18. Mensch JS Amstutz HC Knee morphology as a guide to knee replacement Clinical Orthopaedics and Related Research 1975 112 231 241
19. Kim TK Phillips M Bhandari M Watson J Malhotra R What differences in morphologic features of the knee exist among patients of various races? a systematic review Clinical Orthopaedics and Related Research 2017 475 1 170 182 10.1007/s11999-016-5097-4 27704318
20. Berger RA Rubash HE Seel MJ Thompson WH Crossett LS Determining the rotational alignment of the femoral component in total knee arthroplasty using the epicondylar axis Clinical Orthopaedics and Related Research 1993 286 40 47 10.1097/00003086-199301000-00008
21. Ewe TW Ang EKC Ng WM An Analysis of the relationship between the morphometry of the distal femur, and total knee arthroplasty implant design Malaysian Orthopaedic Journal 2009 3 2 24 28 10.5704/MOJ.0911.005
22. Agrawal RRK Dalei TR Reddy AVG Computer tomography 3D reconstruction-based study of knee anthropometry of Indian population: comparison with other ethnic groups and current TKA implants Int J Orthop Sci 2017 3 4 242 249 10.22271/ortho.2017.v3.i4d.34
| 0 | PMC9741755 | NO-CC CODE | 2022-12-14 23:29:57 | no | Indian J Orthop. 2022 Dec 11;:1-7 | utf-8 | Indian J Orthop | 2,022 | 10.1007/s43465-022-00777-3 | oa_other |
==== Front
School Ment Health
School Ment Health
School Mental Health
1866-2625
1866-2633
Springer US New York
9560
10.1007/s12310-022-09560-z
Original Paper
Application of a Model of Workforce Resilience to the Education Workforce: Expanding Opportunities for Support
http://orcid.org/0000-0003-2637-3748
Prout Joanna T. [email protected]
1
Moffa Kathryn 2
Bohnenkamp Jill 1
Cunningham Dana L. 1
Robinson Perrin J. 1
Hoover Sharon A. 1
1 grid.411024.2 0000 0001 2175 4264 Division of Child and Adolescent Psychiatry, Department of Psychiatry, National Center for School Mental Health, University of Maryland School of Medicine, 737 West Lombard Street, 4th Floor, Baltimore, MD 21210 USA
2 grid.2515.3 0000 0004 0378 8438 Department of Psychiatry and Behavioral Sciences, Boston Children’s Hospital Neighborhood Partnerships, Boston Children’s Hospital, Boston, MA USA
11 12 2022
114
22 11 2022
© This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The current study analyzed 502 responses from members of the education workforce on the Resilience at Work (RAW) scale and other measures of health and job satisfaction as part of an initiative offering training and technical assistance to support student and staff well-being. A latent profile analysis using scores on components of the RAW identified three resilience profiles: lower, moderate, and higher capacities for resilience. Profiles were differentiated across components related to resilience capacity including alignment of work and personal values, level of social support, and ability to manage stress. Differences between profiles were observed across days of poor physical health, days of poor mental health, days of activity restriction, general health rating, and domains of burnout, compassion satisfaction, and secondary traumatic stress. These findings reinforce calls to support the education workforce through changes that allow access to meaningful work, an evaluation of demands including workload, relevant training on emotional wellness, positive experiences, connections with others, and stress management.
Keywords
Educators
Resilience
Well-being
Burnout
Secondary traumatic stress
http://dx.doi.org/10.13039/100005977 Kaiser Permanente
==== Body
pmcIntroduction
Even prior to the COVID-19 pandemic and racial reckoning related to the murder of George Floyd and many others in recent years, members of the education workforce in the USA reported higher levels of stress and poorer physical and mental health than other professionals (American Federation of Teachers, 2015, 2017; Jarvis, 2002). Recent work indicates that many are reaching a breaking point and are considering or actually leaving the profession (Steiner & Woo, 2021). While interventions that help educators manage negative emotions have a role in addressing this problem (Hagermoser Sanetti et al., 2020), additional supports and resources are necessary to support them in balancing the multiple demands of their profession (Farley & Chamberlain, 2021). By looking at the association of a range of factors related to capacity for resilience, physical and mental health, and professional quality of life, this study of the education workforce including K-12 teachers, administrators, and other school staff supports work calling for structural change to support the education workforce in the face of overwhelming demands. We connect our findings to an existing literature that recommends supporting professionals working within schools using a strengths-based systemic approach that integrates insights from veterans in the field.
Educator Stress and Associated Problems
Educator stress and associated health problems, burnout, and attrition are long-standing concerns (Boström et al., 2020; Gray et al., 2017; Ingersoll, 2001; Johnson et al., 2005; Kacey, 2004; Kyriacou, 2001; Saloviita & Pakarinen, 2021). Even before the COVID-19 pandemic, working within education, whether as a teacher, school leader, or support staff member, was highly stressful (Camacho et al., 2018; Jarvis, 2002; Kyriacou, 2001; Mahfouz et al., 2019). The pandemic added demands including implementing COVID-19 mitigation strategies, supporting students who had lost instructional time, assuming the work of absent colleagues, and supporting the mental health of students and staff experiencing intense stress (Diliberti et al., 2021). In January 2022, 78% of teachers reported frequent job-related stress, about twice that of the general working population (Steiner et al., 2022). Similarly, surveys of school leaders during the pandemic indicated high levels of stress, anxiety, and overwhelm (Brackett et al., 2020). School staff working to support the mental health of students and families experienced heightened stress due to frequent crises, increased needs, and decreased resources (Villares et al., 2022). While some stress is inevitable, experiencing extended periods of heightened stress without access to adequate resources can be associated with poor physical and mental health (Farley & Chamberlain, 2021; Prilleltensky et al., 2016).
Extended periods of stress and limited support have been associated with burnout, a complex syndrome with symptoms in three areas: (1) extreme exhaustion, (2) feelings of cynicism or detachment, and (3) low sense of accomplishment or effectiveness (Bakker & Demerouti, 2007; Maslach & Leiter, 2016). Burnout has been linked to mental health difficulties such as anxiety and depression as well as physical health problems such as heart disease and type 2 diabetes (Morse et al., 2012; Salvagioni et al., 2017). Like other helping professionals, school staff may be vulnerable to burnout, as they are responsible for ensuring the emotional well-being of others, often putting their needs ahead of their own. Recent surveys and meta-analyses indicate that burnout is a key factor in attrition, or decision to leave their careers, for education staff, and that it may be becoming more prevalent over time (GBAO, 2022; Hanover Research, 2022; Madigan & Kim, 2021).
Like stress and burnout, workforce attrition in education is a long-term problem. Prior to the COVID-19 pandemic, teachers in the USA had an annual attrition rate of approximately 8% with an estimated 40% of teachers quitting within the first four years of work (Carver-Thomas & Darling-Hammond, 2017). Although there is variation by locale, rates of public-school teacher attrition have increased slightly during the pandemic (Barnum, 2022). Results of a poll conducted in June 2022 indicated that 65% of public-school leaders were concerned about filling vacant staff positions (School Staffing Shortages: Report from the January School Pulse Panel, 2022). Further, recent surveys indicate that about 33% of remaining educators are thinking of exiting in the next two years and nearly 55% are thinking of exiting sooner than initially planned (GBAO, 2022; Hanover Research, 2022). Attrition rates for school leaders are high, with about 18% of principals leaving their position each year (Levin et al., 2019). School superintendents have similarly high attrition rates at around 13% annually (Schwartz & Diliberti, 2022). Turnover rates for mental health therapists in all settings are high at 30–60% annually, and therapists working within schools face additional pressures that may increase attrition (Adams et al., 2019).
Models of job satisfaction emphasize that workers must have adequate resources to meet demands (Bakker & Demerouti, 2007), or excessive stress, burnout, and attrition will occur. The demands placed on people working within education, including work overload, time pressure, lack of social support, and difficult-to-achieve goals are well-documented (Hakanen et al., 2006; Maas et al., 2021; Prilleltensky et al., 2016; Skaalvik & Skaalvik, 2011, 2018). While demands placed on the education workforce are high, resources are limited (Camacho et al., 2018; Kyriacou, 2001; Skaalvik & Skaalvik, 2018), putting them in a position where burnout is likely. Autonomy is important for job satisfaction (Kengatharan, 2020; Worth & Van den Brande, 2020), yet it is limited in education by policies promoting high-stakes testing and curriculum restrictions (Farley & Chamberlain, 2021). The COVID-19 pandemic and racial reckoning brought additional stress for the education workforce, and recent polling indicates that many are leaving, or considering leaving, the field (Chan et al., 2021; Steiner & Woo, 2021). To address this issue, we must identify resources to support the education workforce without placing additional burden on school staff (Madigan & Kim, 2021; Steiner et al., 2022).
Supporting the Education Workforce Through Adversity
Supporting students and families during times of adversity and trauma may increase education staff’s vulnerability to burnout and secondary traumatic stress, factors which can lead to attrition (Christian-Brandt et al., 2020; Hydon et al., 2015; Miller & Flint-Stipp, 2019). Trauma or adversity have been common experiences for children in the USA for many years (Finkelhor, 2020; Sacks & Murphey, 2018). More than two-thirds of children report at least one traumatic event or instance of threatened death, serious injury, or sexual violence, by age 16 (Copeland et al., 2007). While most children who experience a traumatic event do not develop symptoms of post-traumatic stress disorder, those who do are at risk for problems related to learning and education, poor physical and mental health, and involvement with the justice system (Grummitt et al., 2021; Larson et al., 2017; Mandavia et al., 2016; Saunders & Adams, 2014). Many of the children who do not experience a traumatic event may be exposed to toxic stress, which differs from trauma in that it is a period of sustained high-intensity adversity rather than a discrete life-threatening event (Garner et al., 2012; Shonkoff et al., 2012). Although it is difficult to estimate the number of children exposed to toxic stress, recent work in the USA indicates that almost 60% of adults report experiencing at least one adverse childhood experience and about 20% report experiencing three or more (Giano et al., 2020). Characteristics such as minoritized race or ethnicity, lower income or education level, and being a woman are associated with increased risk of experiencing trauma and adversity (Giano et al., 2020).
Some sources indicate the mental health of children and adolescents has declined over recent decades (Collishaw & Sellers, 2020; Twenge et al., 2019), and that stressors during the pandemic, such as loss of social support and routine, have exacerbated this problem (Samji et al., 2022; Zolopa et al., 2022). Currently, youth mental health needs outweigh resources, and shifting to prevention and early intervention is a frequently proposed solution (Kollins, 2022). Educators, who are often the first point of contact for youth experiencing mental health difficulties, can feel unprepared and overwhelmed by student and family mental health needs, an experience that may contribute to emotional exhaustion and burnout (Alisic, 2012; Berger et al., 2021; Temkin et al., 2020). Despite the awareness of this risk, there is limited knowledge of how to support non-clinical staff in helping children and families who have experienced trauma and adversity. Models explaining how helping professionals like therapists maintain personal well-being emphasize the importance of balancing energy-depleting experiences with renewing experiences that build compassion satisfaction, a sense of fulfillment gained from work as a helper (Ludick & Figley, 2017). Due to their role in supporting youth, these processes may be applicable to education staff as well (Kangas-Dick & O’Shaughnessy, 2020; Skaalvik & Skaalvik, 2018), although recent research indicates that processes related to trauma exposure and coping may be unique to educators (Fleckman et al., 2022).
The high prevalence of trauma and toxic stress in children means that education staff will likely support students coping with these adversities on a regular basis (Copeland et al., 2007; Finkelhor, 2020). Secondary traumatic stress, or the experience of post-traumatic stress symptoms such as anxiety, nightmares, and withdrawal, can result from hearing about students’ traumas (Baird & Kracen, 2006), and many school staff members receive limited preparation for dealing with this experience (Caringi et al., 2015; Christian-Brandt et al., 2020; Rankin, 2022; Simon et al., 2022). Without appropriate supports, secondary traumatic stress symptoms can contribute to poor outcomes for both students and school staff (Caringi et al., 2015; Rankin, 2022). This study will tie together education staff’s ratings of compassion satisfaction, secondary traumatic stress, and burnout with factors related to resilience to inform efforts to better support school teams in navigating stress.
Resilience
When adversity is unavoidable, interest often turns to learning how others have persevered or even thrived despite hardships. Many models of resilience, the ability to withstand and “bounce back” from stress, emphasize that it is not a fixed ability within the individual but a function of the interaction between the person and environmental resources (Masten, 2015). The multitude of factors that influence education staffs’ abilities to persevere include policy, social-cultural environment, and relationships with leaders, peers, and students (Gu & Day, 2013). Most models emphasize an interactional approach to resilience where intrapersonal (e.g., sense of vocation or purpose, ability to use problem solving and emotion regulation strategies) and contextual factors (e.g., level of social support) work in combination to influence peoples’ capacities to manage stressors and thrive in their work (Mansfield et al., 2016). This work emphasizes complexity and the interaction of systems that influence the well-being of school staff while providing multiple points for intervention.
Current Status of Supports for Education Staff
Many interventions for education staff focus on teaching them to manage their anxiety and distress via cognitive techniques and other stress management strategies (Hagermoser Sanetti et al., 2020; Jennings et al., 2017; Klingbeil & Renshaw, 2018). However, the current state of crisis, with many school staff members feeling overwhelmed and burnt out, brings to the forefront long-standing calls to support them not solely through helping them manage emotions but by capitalizing on strengths and renewing experiences, enhancing social support, and making systemic changes. The current study will support this purpose by examining the initial application of a general model of workforce resilience to a sample of education staff. Latent profile analysis (LPA) was used to explore possible groupings of education staff based on capacity for resilience. Since this is the first time latent profiles have been explored among education staff using this model of workforce resilience, emerging groupings were validated with meaningful concurrent measures of physical and mental health and professional quality of life. Implications of findings for education staff, the workforce, and school systems are presented.
Method
Participants
Participants (N = 502) included K-12 teachers, administrators, and other school staff at 20 public schools across the USA participating in an initiative to provide training and technical assistance to enhance trauma-responsive practices and improve student and staff well-being. Schools volunteered to receive these resources. As shown in Table 1, 75.9% (n = 388) of respondents identified as teachers, 53.2% (n = 272) worked in the schools for over 11 years, and 52.1% (n = 266) had earned a master’s degree. Further, 58.9% (n = 301) of respondents identified as women, 44.8% as White, 21.7% as Hispanic or Latinx, and 19.6% as Black or African American. Respondents were from 20 different schools representing many regions across the USA and were employed at a variety of school levels (seven elementary schools, nine middle/intermediate schools, four high schools). In 17 out of 20 schools, at least 90% of students identified as ethnic or racial minorities.Table 1 Demographic information for survey respondents
n Percent
Role
Teacher 381 75.9
Administrator 30 6.0
Healthcare provider 5 1.0
Student support staff 38 7.6
Write-in 9 1.8
Missing 39 7.8
Years in schools
< 2 years 40 8.0
2–5 years 72 14.3
6–10 years 82 16.3
> 11 years 271 54.0
Missing 37 7.4
Highest level of education
< High school diploma 2 0.4
High school diploma 4 0.8
Associate degree 4 0.8
Bachelor’s degree 155 30.9
Master’s degree 263 52.4
Doctoral degree 11 2.2
Write-in 25 5.0
Missing 38 7.6
Gender
Woman 297 59.2
Man 147 29.3
Non-binary/third gender 2 0.4
Prefer not to say 19 3.8
Missing 37 7.4
Race
Black/African American 98 19.5
American Indian 8 1.6
Asian 29 5.8
Hawaiian 2 0.4
White 225 44.8
Multi-racial 22 4.4
Write-in 66 13.1
Missing 17 3.4
Ethnicity
Hispanic 110 21.9
Not Hispanic 346 68.6
Missing 46 9.2
Procedures
Staff members were asked to complete an online survey that included the Resilience at Work Scale (Winwood et al., 2013), the Healthy Days module of the Health-Related Quality of Life scale (Centers for Disease Control & Prevention, 2020), and the Professional Quality of Life scale (Stamm, 2002) during the autumn of 2018 prior to provision of additional resources. Demographic information including role at the school, years of experience, race, ethnicity, and gender were also collected. Since surveys were designed to be anonymous, no information that could be used to identify a respondent was collected. Completion of the survey was voluntary; participants were informed that their decision to complete or not complete any question would not impact their employment or any other benefits they might receive. Survey distribution and analyses were approved as part of a larger evaluation project by the university IRB. Participants were not compensated for their completion of the survey.
Measures
Resilience at Work Scale
The Resilience at Work (RAW) scale (Winwood et al., 2013) is a 25-item measure of workplace resilience that is broadly applicable to all fields of employment. The RAW scale provides a total score as well as scores for the seven domains of this model: (1) Living Authentically: living by personal values, having good awareness and ability to regulate emotion, using personal strengths; (2) Finding Your Calling: seeking work that has a purpose, aligning work with core values and beliefs; (3) Maintaining Perspective: capacity to reframe setbacks, maintain solution focus, manage negativity; (4) Managing Stress: use of work and life routines to manage stressors and maintain a healthy balance with adequate time for relaxation; (5) Interacting Cooperatively: seeking and providing support in the workplace, (6) Staying Healthy: physical fitness and healthy diet, and (7) Building Networks: having others to provide support both within and outside of work. Figure 1 provides a visual of the domains of the RAW. Respondents are asked to rate their level of agreement with statements about themselves and their work (e.g., “I have important core values that I hold fast to in my work-life,” “I have friends at work I can rely on to support me when I need it”) on a 7-point Likert-type scale with 0 being “strongly disagree” and 6 being “strongly agree.” Initial analysis of the RAW scale showed an overall Cronbach’s α of 0.84 (Winwood et al., 2013), and it demonstrated adequate internal consistency in this sample with scale values ranging from α = 0.60 to α = 0.83 (see Table 2). Further testing has shown the RAW scale to have high reliability and has found RAW scale scores to be positively associated with work engagement (Malik & Garg, 2018).Fig. 1 Resilience at work model
Table 2 Descriptive statistics of RAW scale subdomains and cross-sectional outcomes
Scale Range M SD N of items Cronbach’s α
RAW scales
Living authentically 0–6 4.92 0.72 4 0.71
Finding your calling 0–6 4.88 0.91 4 0.81
Maintaining perspective 0–6 3.58 1.01 4 0.60
Managing stress 0–6 4.19 1.13 4 0.81
Interacting cooperatively 0–6 4.77 0.87 3 0.60
Staying healthy 0–6 4.22 1.27 3 0.83
Building networks 0–6 4.84 0.93 3 0.66
Healthy days measure
General health rating 1–5 2.54 0.88 1 –
Days poor physical health 0–30 4.27 6.49 1 –
Days poor mental health 0–30 5.47 6.94 1 –
Days activity restriction 0–30 3.74 5.97 1 –
Pro-QOL
Compassion satisfaction 10–50 40.85 6.27 10 0.93
Burnout 10–43 22.76 5.75 10 0.82
Secondary traumatic stress 10–50 22.09 6.08 10 0.83
Professional Quality of Life Scale (Pro-QOL)
The Pro-QOL (Stamm, 2002) is a 30-item self-report measure of factors related to potential positive and negative impacts of working as a helper. The Pro-QOL has three subscales: Compassion Satisfaction, Burnout, and Secondary Traumatic Stress. When used with educators, the Compassion Satisfaction subscale measures pleasure derived from helping students to learn, positive feelings about colleagues and being part of a team, and ability to contribute to the greater good of society. The Burnout subscale measures feelings of depression or hopelessness associated with having a high workload without adequate support. The Secondary Traumatic Stress subscale measures fear and anxiety-related symptoms that may be related to working with a population that has experienced trauma or chronic stress. Each subscale can be calculated as a continuous score or categorized into low, average, or high using cutoff scores. Respondents are asked to rate how frequently they have had different thoughts or experiences (e.g., “My work makes me feel satisfied,” “I feel worn out because of my work as a teacher”) on a 5-point Likert-type scale with 1 being “never” and 5 being “very often.” The Pro-QOL is a well-established measure that showed strong internal consistency in our sample, with Cronbach’s alphas ranging from 0.82 to 0.93 (see Table 2).
Healthy Days Measure (HDM)
The standard four-item set of Healthy Days questions, developed by the Centers for Disease Control and Prevention for use in national surveys beginning in 1993, asks respondents to provide: (1) a rating of their overall level of health on a five-point scale from excellent to poor, (2) the number of days out of the past 30 when the respondent experienced poor physical health, (3) the number of days out of the past 30 when the respondent experienced poor mental health, and (4) the number of days out of the past 30 when poor physical or mental health kept the respondent from completing their usual activities. In an analysis of 42,000 adults across 13 states, an average of 3.1 days of poor physical health, 2.8 days of poor mental health, and 1.7 days of activity limitation were reported across genders (CDC, 2000).
Data Analysis Plan
Preliminary Analyses
Descriptive statistics including mean and standard deviation were calculated for all measures. To examine internal consistency, Cronbach’s alpha was calculated for all scales containing more than one item. To examine validity, correlations between measures were calculated. As different groups of educators by role (teacher, administrator, other school staff) were included in the sample, we conducted analysis of variance (ANOVA) to examine between-group differences in scores on included scales.
Latent Profile Analysis
LPA with full information maximum likelihood (FIML) estimation using Mplus version 8.0 (Muthén & Muthén, 1998) was conducted to explore underlying latent class structure based on educators’ resilience at work on the seven domains specified by the RAW Scale. A class-invariant diagonal structure, the default in Mplus, was considered with covariances between indicators fixed to zero within classes and variances constrained to be equal across groups. First, a one-class model was fit to the sample. The number of classes was subsequently increased until indicators of absolute, relative, and substantive fit supported a given solution.
To assess fit between latent class models, Bayesian Information Criterion (BIC), Adjusted BIC, Bayes Factor (BF), Lo-Mendell-Rubin likelihood ratio test (LMR-LRT) and Bootstrap Likelihood Ratio Test (BLRT) were considered to compare K-classes to K-1-classes. The lowest values for the BIC and ABIC may suggest the best fitting model, although “elbow” plots of BIC and ABIC values were also examined, as diminishing benefit was observed in increasing class number (Masyn, 2013). Significant p-values for the LMR-LRT and BLRT suggest that the model with K-classes (i.e., alternative model) fit the data significantly better than the model with K-1 classes (i.e., null model). Finally, theoretical rationale for a given latent class structure was considered, as determining the best model fit based only on fit statistics negates the substantive interpretation of the model and its usefulness within the context of educators’ capacity for resilience. Following class enumeration and selection of possible models, entropy, and average latent class probabilities were also considered, with a recommended entropy value of 0.80 or above.
LPA with Distal Outcomes
Once an unconditional model was chosen, the researchers specified a conditional model with proximal distal outcomes to examine the relation of latent profiles to well-validated measures of well-being. The proximal distal outcomes of interest were educators’ ratings of their overall health, physical health, and mental health from the HDM, as well as self-rating of compassion satisfaction, burnout, and secondary traumatic stress from the Pro-QOL. A BCH modified 3-step approach was employed to ensure that latent profiles did not shift in size (Asparouhov & Muthén, 2014; Masyn, 2013). Meaningful differences between latent profiles across these outcomes provided evidence that emerging profiles were meaningful and different across constructs of interest.
Results
Preliminary Analyses
Means, standard deviations, and Cronbach’s alpha for each scale aligned with previous research (see Table 2). Correlation coefficients between scales were in the expected directions, with negative outcomes such as poorer health being associated with higher levels of burnout and secondary traumatic stress and lower levels of compassion satisfaction (see Table 3). The Pro-QOL and many RAW scales showed associations in the expected directions, with scores associated with higher resilience also being associated with higher compassion satisfaction and lower secondary traumatic stress and burnout. ANOVA of scores by role showed no significant differences by group across the HDM, Pro-QOL, or RAW subscales.Table 3 Correlations for study scales
Scale 1 2 3 4 5 6 7 8 9 10 11 12 13 14
RAW scales
1. Living authentically –
2. Finding your calling .58** –
3. Maintaining perspective .45** .36** –
4. Managing stress .58** .46** .53** –
5. Interacting cooperatively .55** .42** .31** .34** –
6. Staying healthy .38** .27** .20** .46** .17** –
7. Building networks .46** .58** .21** .39** .46** .25** –
Healthy days measure
8. General health Rating − 0.05 − 0.01 − 0.02 − 0.08 0.02 0.00 0.03 –
9. Days poor physical health − 0.05 − 0.02 − 0.06 − 0.06 − 0.05 − 0.02 0 .70** –
10. Days poor mental health − 0.08 − 0.06 − 0.08 − 0.08 0 − 0.05 − 0.02 .66** .67** –
11. Days activity restriction − .24** − .25** − .30** − .35** − .18** − .23** − .15** .32** − 0.011 0.06 –
Pro-QOL
12. Compassion satisfaction .63** .67** .40** .49** .49** .31** .41** − .30** − .06** − .11* − .29** –
13. Burnout − .54** − .55** − .59** − .56** − .38** − .27** − .35** .31** − 0.01 0.09 .42** − .72** –
14. Secondary traumatic stress − .22** − .20** − .48** − .25** − .12* 0.01 − 0.01 0.06 − 0.01 − 0.01 .31** − .25** .58** –
*p < .05, **p < .01
Latent Profile Analysis
Average responses on the RAW domains and cross-sectional physical and mental health outcomes are shown in Table 2. The elbow plot of BIC and size-adjusted BIC values supported a three-class solution, and the AIC supported a three- and five-class solution. The adjusted LMRT p-value and BF p-value supported a two-class solution. Upon further examination, it was determined that the addition of the fourth and fifth classes explained variance in an unreliably small proportion of the sample (1.6%). Based on the elbow plot of BIC values and substantive interpretation for a one- through six-class solution (see Table 4), a three-class solution was chosen as the best fitting model. Entropy for the three-class solution was 0.79.Table 4 Fit Statistics for one- through six-class solutions
Model (K-class) LL npar AIC CAIC BIC saBIC AWE LRTS Adj LMR p-value Bootstrapped (BLRT) p-value BF (K, K + 1) cmP (K) SIC Exp(SIC-max)
1-class − 4855.01 14 9738.02 9811.08 70,266.89 9752.65 9926.14 686.03 < .001 < .001 0.00 0.00 − 35,133 0
2-class − 4512.00 22 9068.00 9182.81 65,414.24 8969.33 9363.61 285.55 0.11 < .001 0.00 0.00 − 32,707 0
3-class − 4369.23 30 8798.45 8955.01 63,452.49 8663.90 9201.57 5.45 0.27 < .001 11,722,753,721,072.70 0.00 − 31,726 0
4-class − 4366.50 38 8809.00 9007.31 63,512.68 8849.37 9319.62 197.85 0.12 < .001 0.00 0.00 − 31,756 0
5-class − 4267.58 46 8627.16 8867.21 62,183.99 8420.85 9245.27 65.62 0.10 < .001 416,665,259.00 0.00 − 31,092 0
6-class − 4234.77 54 8577.54 8859.34 8859.34 8335.35 9303.15 8469.54 0.81 < .001 0.00 1.00 − 4429.7 1
The profile plots of the estimated mean values for each domain of resilience at work are displayed in Fig. 2 (0 = Strongly Disagree to 6 = Strongly Agree). Based on the pattern of mean scores on the domains of workplace resilience, the following labels were given for each class: Higher Capacity for Resilience (i.e., High; 46.4%), with the highest means across all seven domains of resilience, Moderate Capacity for Resilience (i.e., Moderate; 45%), with the second highest means across all seven domains, and Lower Capacity for Resilience (i.e., Low; 8.6%), with the lowest means across the seven domains. Highest average scores for the Higher and Moderate Capacity groups were in the domains of Living Authentically, Finding Your Calling, Interacting Cooperatively, and Building Networks. Of note, the pattern of responses of those falling in the Lower Capacity group was similar except for lower scores in Finding Your Calling. Most educators fell into the Higher or Moderate Capacity for Resilience classes. The probability of educators being classified into each capacity for resilience class, given their classification into one resilience class, is presented in Table 5. The values of the bolded cells presented on the diagonal for each class indicate the probability that members of a given class would be classified into that class. The classification for each resilience capacity class was high (> 0.80).Fig. 2 Graph of component scores by resilience capacity profile
Table 5 Average probability of most likely latent class membership by latent resilience class
Class Higher (1) Moderate (2) Lower (3)
1 0.923 0.077 0.000
2 0.012 0.893 0.095
3 0.000 0.096 0.904
Differences in Physical and Mental Health Outcomes Based on LPA Classification
A modified BCH 3-step approach was utilized to examine relations between cross-sectional (i.e., proximal) outcomes of physical and mental health and educators’ workplace resilience. To test which physical and mental health outcomes differed in their mean outcome scores, a Wald Test was carried out on all between-class comparisons. Table 6 shows that there were significant differences between all classes across all outcomes in the expected direction. The Higher Capacity for Resilience group, on average, reported the fewest days of poor physical health (M = 2.98, SE = 0.41) and mental health (M = 3.09, SE = 0.40) per month, “very good” general health (M = 2.22, SE = 0.07), significantly less burnout (M = 19.07, SE = 0.34) and secondary traumatic stress (M = 10.63, SE = 0.44) than the other two groups, and significantly higher compassion satisfaction (M = 45.06, SE = 0.31) than both groups. Additionally, the Moderate Capacity for Resilience group indicated, on average, approximately five days of poor physical health per month (M = 4.77, SE = 0.52), six days of poor mental health per month (M = 6.20, SE = 0.53), “good” general health (M = 2.76, SE = 0.06), significantly more burnout (M = 24.66, SE = 0.36) and secondary traumatic stress (M = 22.87, SE = 0.43) and significantly less compassion satisfaction (M = 38.62, SE = 0.41) than the Higher group. Finally, the Lower Capacity for Resilience group, on average, indicated approximately eight days of poor physical health (M = 8.49, SE = 1.41) and almost 14 days of poor mental health (M = 13.95, SE = 1.49), “good” general health (M = 3.18, SE = 0.17), significantly more burnout (M = 32.39, SE = 0.81) and secondary traumatic stress (M = 25.72, SE = 1.28), and significantly less compassion satisfaction (M = 30.28, SE = 1.06) than the Moderate and Higher groups.Table 6 Resilience profiles’ relations to physical and mental health and professional quality of life
Resiliency Profile Days activity restriction Days poor mental health Days poor physical health Health rating Burnout Compassion satisfaction Secondary traumatic stress
M(SE) M(SE) M(SE) M(SE) M(SE) M(SE) M(SE)
Lower N = 43 9.28 (1.36)* 13.95 (1.49)* 8.49 (1.41)* Good 3.18 (.17)* 32.39 (.81)* 30.28 (1.06)* 25.72 (1.28)*
Moderate N = 226 4.47 (.49)* 6.20 (.53)* 4.77 (.52)* Good 2.76 (.06)* 24.66 (.36)* 38.62 (.41)* 22.87 (.43)*
Higher N = 233 1.88 (.31)* 3.09 (.40)* 2.98 (.41)* Very Good 2.22 (.06)* 19.07 (.34)* 45.06 (.31)* 20.63 (.44)*
*Indicates significantly different from the other two classes at p < .05
Discussion
The current study adds to our understanding of how the education workforce’s capacity for resilience, which is closely tied to health, exists within a complex system of resources and demands. The health of staff working within schools varies with factors associated with professional well-being, including compassion satisfaction, burnout, and secondary traumatic stress. The link between burnout, secondary traumatic stress, and poorer health is important to consider as educators are currently tasked with supporting students and families struggling with trauma and adversity while also coping with their own stressors. Perhaps the most intriguing finding is that resilience capacity groups varied across all factors on the RAW scale, including those related to context versus those solely related to ability to manage stress. This finding indicates that systemic changes related to decreasing demands and increasing resources are necessary and align with calls to broaden educator support through systemic changes (Kangas-Dick & O’Shaughnessy, 2020; Skaalvik & Skaalvik, 2018).
The association between resilience capacity group and scores on the professional quality of life scales—Compassion Satisfaction, Burnout, and Secondary Traumatic Stress—support the idea that efforts to help the education workforce should follow best practices used for people whose work involves high levels of empathy and exposure to trauma, such as frontline workers and mental health clinicians. The association of higher levels of compassion satisfaction with higher capacity for resilience is consistent with the idea that access to the emotional rewards of working in a helping profession is vital to maintain energy and motivation (Ludick & Figley, 2017; Santoro, 2011, 2018, 2019). Insights from the field have emphasized that being able to see the rewards of helping is key to maintaining energy and motivation.
The differentiation across groups in secondary traumatic stress scores warns that the emotional impact of working with students and families who have experienced trauma may contribute to educator burnout and attrition, and that this factor should be examined in further detail (Caringi et al., 2015). As school staff are key supports in the emotional lives of children, it is unsurprising that their well-being can suffer when children suffer. As so many communities experience trauma and adversity and trauma-informed care becomes more common in schools, we must consider how to support those professionals who are engaged in this work within schools (Christian-Brandt et al., 2020). In alignment with prior research, we found that school staff with lower levels of burnout and secondary traumatic stress had higher levels of compassion satisfaction and resources such as social support. This finding supports the idea that community and sense of self-efficacy can buffer against the impact of stress (Berger et al., 2021; Miller & Flint-Stipp, 2019). It is imperative that educators receive training that normalizes the experiences of sadness, anger, and other emotions when working with highly stressed children and families. Furthermore, the need to access resources to maintain well-being should be normalized, and easy access to these resources should be supported.
The fact that resilience capacity groups showed differentiation across all domains on the RAW scale, not solely those related to stress management, aligns with calls to consider support for school staff resilience from a broad perspective. The differentiation of scores on the Finding Your Calling and Living Authentically domains (see Fig. 2) supports the importance of alignment of personal values and work tasks in determining capacity for resilience, with school staff in the Higher and Moderate classes having substantially higher average scores when compared to school staff in the Lower Capacity class. Similar differences were seen in scores on domains related to social connectedness including Interacting Cooperatively and Building Networks, indicating the importance of interpersonal support. These results suggest that while helping the education workforce cope with stress is one point of intervention, a multi-pronged approach that uses all available resources to support the education workforce may be most effective. Also, acknowledging that external factors also play an important role in one’s success in carrying out a task validates the experiences of the education workforce and reduces the sense that they are being blamed for being unable to cope with unreasonable demands.
Fortunately, there is a wealth of information that can guide efforts to support morale and well-being for the education workforce (Casely-Hayford et al., 2022; Chiong et al., 2017; Glazer, 2018). Reports from veteran teachers that remained in challenging positions for years are key sources of guidance (Bullough & Hall‐Kenyon, 2011; Sell, 2019). These reports indicate that having a sense of meaning in their work and feeling connected with others were key factors in being able to remain in their roles. This work indicates that one area of change might be to ensure that the daily tasks of the education workforce align with their mission and individual sense of purpose. That is, tapping into the reasons they entered the profession (e.g., connecting with students, using creativity to teach science) and then allowing their daily activities to reflect this “calling” may help decrease burnout and attrition within the education workforce (Santoro, 2018). This need may be particularly urgent as school staff are faced with highly stressed students and families while themselves coping with stressors related to disruptions from COVID-19 and intense social unrest.
Another resource and strength to build on to support the education workforce is their connections with others including students and families, peers, and school leadership. Building relationships with students and watching children learn and grow is a key source of joy for many educators (Chiong et al., 2017), but the demands of intensive curricula and pressures of testing can reduce or eliminate their ability to access these rewards. Reductions in high-pressure testing, decreased staff-to-student ratios, and an emphasis on flexibility and relationship building are changes that would align with this need (Maas et al., 2021). In addition to relationships with students, the education workforce needs to have a community of support among their peers and support from leadership that has similar values and enthusiasm for the work of supporting students and families (Chiong et al., 2017). The creation of time and spaces for the education workforce to connect with their colleagues, receive support, and celebrate success is essential for maintaining well-being and is often minimized due to time pressures. Many educators report feelings of isolation, as they remain in classrooms with few other adults for most of their working hours (Drago‐Severson & Pinto, 2006; Hadar & Brody, 2010). This problem has been exacerbated by the COVID-19 pandemic, which has led to educators conducting classes via virtual platforms and staffing shortages that have increased workload. Increasing staff ratios to allow staff to take breaks and connect with others would be beneficial in decreasing daily stress and improving long-term morale and retention.
Limitations and Future Directions
Next steps might include adapting these suggestions into changes and supports that can be integrated within the education setting, advocating for their use, testing how they work in “real world” situations, and further tailoring them with user feedback and results showing their impact. Intensive collaboration with educators will be essential in moving forward in this area. It is clearly time to move beyond suggesting stress management interventions to people with strength and purpose who are coping with unreasonable demands. Examining the feasibility of and actual impact of supportive changes and interventions using techniques such as those developed in implementation science will be key to ensuring sustainable and cost-effective change (Moir, 2018).
There is building evidence that racially minoritized educators and school staff experience stressors related to systemic racism embedded in the USA’s public education system and are more likely than white educators to consider leaving the field (Grooms et al., 2021; Mahatmya et al., 2021; Mason-Williams et al., 2022). There is an urgency to address this issue, as education professionals working with racially minoritized students are already disproportionately white (Race and Ethnicity of Public-School Teachers and Their Students, 2020). The lack of racially minoritized educators due to racialized school climate limits the ability of school systems to develop culturally sensitive and equitable practices. The burden for creating spaces within schools that racially minoritized professionals can thrive in must be placed on the system rather than expecting them to adapt to hostile or unaffirming spaces (Grooms et al., 2021; Mahatmya et al., 2021).
The results of the current study are limited by several factors, including lack of information regarding the supports available at the school level. To examine the impact of individual and systemic factors within the education workforce, it is necessary to include measures of system-level variables such as school climate and the level of support staff receive in areas such as opportunities for formal or informal mentorship, level of collaboration between staff and leadership in decision making, and other factors related to job satisfaction such as compensation, ability to use sick and vacation leave, and workload. Future research can build on these findings by collecting comprehensive data regarding both individual- and context-level factors related to resilience in teachers and school staff. Further, future research should explicitly consider the impact of systemic racism on burnout for racially minoritized school staff and test the impact of changes to reduce discrimination. For example, Mahatmya et al. (2021) and colleagues found that racially minoritized educators in environments less open to discussing racial conflict reported higher levels of burnout when compared to those in more open environments. It is essential that research about how to best support workplace resilience be led by racially minoritized education staff.
Additionally, some fit statistics supported different latent profile solutions than the three-class solution ultimately chosen based on fit statistics and substantive interpretation. It is possible that the RAW Scale was not developed with the intention of classification or differentiation, as it was used in the current study. Future studies should examine the use of LPA with the RAW Scale with a larger, diverse group of education staff to better assess the fit of a three-class solution. Upon further evidence of a class solution with other samples, educators and school systems would benefit from research that explores how these groups differ based on several factors, including role at school, primary versus secondary school, and school resources and supports provided. Research into these factors will help to tailor needed policy, systemic change, and interventions for educators.
Acknowledgements
We would like to acknowledge the work of the Los Angeles Education Partnership and the schools and educators that participated in RISE.
Funding
This work was supported by a grant from Kaiser Permanente Thriving Schools to the University of Maryland School of Medicine, Division of Child and Adolescent Psychiatry, National Center for School Mental Health.
Declarations
Conflict of interest
Authors report no competing interests related to this publication. All authors materially participated in research and/or article preparation. All authors have approved the final article.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Adams DR Williams NJ Becker-Haimes EM Skriner L Shaffer L DeWitt K Neimark G Jones DT Beidas RS Therapist financial strain and turnover: Interactions with system-level implementation of evidence-based practices Administration and Policy in Mental Health 2019 46 6 713 723 10.1007/s10488-019-00949-8 31203492
Alisic E Teachers’ perspectives on providing support to children after trauma: A qualitative study School Psychology Quarterly: THe Official Journal of the Division of School Psychology 2012 27 1 51 59 10.1037/a0028590
American Federation of Teachers. (2015). Quality of Worklife Survey.
American Federation of Teachers. (2017). 2017 Educator Quality of Work Life Survey. https://www.aft.org/sites/default/files/2017_eqwl_survey_web.pdf
Asparouhov, T., & Muthén, B. (2014). Auxiliary variables in mixture modeling: Using the BCH method in Mplus to estimate a distal outcome model and an arbitrary secondary model. https://www.semanticscholar.org/paper/Auxiliary-Variables-in-Mixture-Modeling%3A-Using-the-Asparouhov-Muth%C3%A9n/d07ec12834242eb9ae651267e18ecf1d5cd5f5e0
Baird K Kracen AC Vicarious traumatization and secondary traumatic stress: A research synthesis Counselling Psychology Quarterly 2006 19 2 181 188 10.1080/09515070600811899
Bakker AB Demerouti E The job demands-resources model: State of the art Journal of Managerial Psychology 2007 22 3 309 328 10.1108/02683940710733115
Barnum M Uptick but no exodus: Despite stress, most teachers stay put 2022 Chalkbeat
Berger E Chionh N Miko A School leaders’ experiences of dealing with students exposed to domestic violence Journal of Family Violence 2021 10.1007/s10896-021-00310-4
Boström M Björklund C Bergström G Nybergh L Schäfer Elinder L Stigmar K Wåhlin C Jensen I Kwak L Health and work environment among female and male Swedish elementary school teachers—A cross-sectional study International Journal of Environmental Research and Public Health 2020 17 1 227 10.3390/ijerph17010227
Bullough RV Hall-Kenyon KM The call to teach and teacher hopefulness Teacher Development 2011 15 2 127 140 10.1080/13664530.2011.571488
Camacho DA Vera E Scardamalia K Phalen PL What are urban teachers thinking and feeling? Psychology in the Schools 2018 55 9 1133 1150 10.1002/pits.22176
Caringi JC Stanick C Trautman A Crosby L Devlin M Adams S Secondary traumatic stress in public school teachers: Contributing and mitigating factors Advances in School Mental Health Promotion 2015 8 4 244 256 10.1080/1754730X.2015.1080123
Carver-Thomas D Darling-Hammond LD Teacher turnover: Why it matters and what we can do about it 2017 Learning Policy Institute
Casely-Hayford J Björklund C Bergström G Lindqvist P Kwak L What makes teachers stay? A cross-sectional exploration of the individual and contextual factors associated with teacher retention in Sweden Teaching and Teacher Education 2022 113 103664 10.1016/j.tate.2022.103664
Centers for Disease Control and Prevention. (2020). Measuring Healthy Days. https://www.cdc.gov/hrqol/pdfs/mhd.pdf
Chan M Sharkey JD Lawrie SI Arch DAN Nylund-Gibson K Elementary school teacher well-being and supportive measures amid COVID-19: An exploratory study School Psychology 2021 10.1037/spq0000441
Chiong C Menzies L Parameshwaran M Why do long-serving teachers stay in the teaching profession? Analysing the motivations of teachers with 10 or more years’ experience in England British Educational Research Journal 2017 43 6 1083 1110 10.1002/berj.3302
Christian-Brandt AS Santacrose DE Barnett ML In the trauma-informed care trenches: Teacher compassion satisfaction, secondary traumatic stress, burnout, and intent to leave education within underserved elementary schools Child Abuse & Neglect 2020 110 104437 10.1016/j.chiabu.2020.104437 32151429
Collishaw S Sellers R Taylor E Verhulst F Wong JCM Yoshida K Trends in child and adolescent mental health prevalence, outcomes, and inequalities Mental health and illness of children and adolescents 2020 Springer 63 73
Copeland WE Keeler G Angold A Costello EJ Traumatic events and posttraumatic stress in childhood Archives of General Psychiatry 2007 64 5 577 584 10.1001/archpsyc.64.5.577 17485609
Diliberti MK Schwartz HL Grant D Stress topped the reasons why public school teachers quit, even before COVID-19 2021 RAND Corporation
Drago-Severson E Pinto KC School leadership for reducing teacher isolation: Drawing from the well of human resources International Journal of Leadership in Education 2006 9 2 129 155 10.1080/13603120500508080
Farley AN Chamberlain LM The teachers are not alright: A call for research and policy on teacher stress and well-being The New Educator 2021 17 3 305 323 10.1080/1547688X.2021.1939918
Finkelhor D Trends in adverse childhood experiences (ACEs) in the United States Child Abuse & Neglect 2020 108 104641 10.1016/j.chiabu.2020.104641 32739600
Fleckman JM Petrovic L Simon K Peele H Baker CN Overstreet S Compassion satisfaction, secondary traumatic stress, and burnout: A mixed methods analysis in a sample of public-school educators working in marginalized communities School Mental Health: A Multidisciplinary Research and Practice Journal 2022 10.1007/s12310-022-09515-4
Garner AS Shonkoff JP Committee on Psychosocial Aspects of Child and Family Health, Committee on Early Childhood, Adoption, and Dependent Care, & Section on Developmental and Behavioral Pediatrics Early childhood adversity, toxic stress, and the role of the pediatrician: Translating developmental science into lifelong health Pediatrics 2012 129 1 e224 231 10.1542/peds.2011-2662 22201148
GBAO. (2022). Poll results: Stress and burnout pose threat of educator shortage. https://www.nea.org/sites/default/files/2022-02/NEA%20Member%20COVID-19%20Survey%20Summary.pdf
Giano Z Wheeler DL Hubach RD The frequencies and disparities of adverse childhood experiences in the U.S. BMC Public Health 2020 20 1 1327 10.1186/s12889-020-09411-z 32907569
Glazer J Leaving lessons: Learning from the exit decisions of experienced teachers Teachers and Teaching: Theory and Practice 2018 24 1 50 62 10.1080/13540602.2017.1383238
Gray C Wilcox G Nordstokke D Teacher mental health, school climate, inclusive education and student learning: A review Canadian Psychology 2017 58 3 203 210 10.1037/cap0000117
Grooms AA Mahatmya D Johnson ET The retention of educators of color amidst institutionalized racism Educational Policy 2021 35 2 180 212 10.1177/0895904820986765
Grummitt LR Kreski NT Kim SG Platt J Keyes KM McLaughlin KA Association of childhood adversity with morbidity and mortality in US adults: A systematic review JAMA Pediatrics 2021 175 12 1269 1278 10.1001/jamapediatrics.2021.2320 34605870
Gu Q Day C Challenges to teacher resilience: Conditions count British Educational Research Journal 2013 39 1 22 44 10.1080/01411926.2011.623152
Hadar L Brody D From isolation to symphonic harmony: Building a professional development community among teacher educators Teaching and Teacher Education 2010 26 8 1641 1651 10.1016/j.tate.2010.06.015
Hagermoser Sanetti LM Boyle AM Magrath E Cascio A Moore E Intervening to decrease teacher stress: A review of current research and new directions Contemporary School Psychology 2020 10.1007/s40688-020-00285-x
Hakanen JJ Bakker AB Schaufeli WB Burnout and work engagement among teachers Journal of School Psychology 2006 43 6 495 513 10.1016/j.jsp.2005.11.001
Hanover Research. (2022). 2022 Trends in K-12 Research. https://insights.hanoverresearch.com/hubfs/2022-Trends-in-K-12-Education.pdf
Hydon S Wong M Langley AK Stein BD Kataoka SH Preventing secondary traumatic stress in educators Child and Adolescent Psychiatric Clinics 2015 24 2 319 333 10.1016/j.chc.2014.11.003
Ingersoll RM Teacher turnover and teacher shortages: An organizational analysis American Educational Research Journal 2001 38 3 499 534 10.3102/00028312038003499
Jarvis M Teacher stress: A critical review of recent findings and suggestions for future research Stress News 2002 14 1 7
Jennings PA Brown JL Frank JL Doyle S Oh Y Davis R Rasheed D DeWeese A DeMauro AA Cham H Greenberg MT Impacts of the CARE for teachers program on teachers’ social and emotional competence and classroom interactions Journal of Educational Psychology 2017 109 7 1010 1028 10.1037/edu0000187
Johnson S Cooper C Cartwright S Donald I Taylor P Millet C The experience of work-related stress across occupations Journal of Managerial Psychology 2005 20 2 178 187 10.1108/02683940510579803
Kacey G Chronic teacher turnover in urban elementary schools Education Policy Analysis Archives 2004 10.14507/epaa.v12n42.2004
Kangas-Dick K O’Shaughnessy E Interventions that promote resilience among teachers: A systematic review of the literature International Journal of School & Educational Psychology 2020 8 2 131 146 10.1080/21683603.2020.1734125
Kengatharan N The effects of teacher autonomy, student behavior and student engagement on teacher job satisfaction Educational Sciences: Theory & Practice 2020 20 4 1 15 10.12738/jestp.2020.4.001
Klingbeil DA Renshaw TL Mindfulness-based interventions for teachers: A meta-analysis of the emerging evidence base School Psychology Quarterly: THe Official Journal of the Division of School Psychology, American Psychological Association 2018 33 4 501 511 10.1037/spq0000291 30507235
Kollins SH Editorial: The gathering storm: A US perspective on the scientific response to the COVID-19 child and adolescent mental health crisis Journal of Child Psychology and Psychiatry, and Allied Disciplines 2022 63 2 129 131 10.1111/jcpp.13576 35165904
Kyriacou C Teacher stress: Directions for future research Educational Review 2001 53 1 27 35 10.1080/00131910120033628
Larson S Chapman S Spetz J Brindis CD Chronic childhood trauma, mental health, academic achievement, and school-based health center mental health services The Journal of School Health 2017 87 9 675 686 10.1111/josh.12541 28766317
Levin, S., Bradley, K., & Scott, C. (2019). Principal Turnover: Insights from Current Principals. Learning Policy Institute, National Association of Secondary School Principals. https://learningpolicyinstitute.org/sites/default/files/product-files/NASSP_LPI_Insights_Principals_BRIEF.pdf
Ludick M Figley CR Toward a mechanism for secondary trauma induction and reduction: Reimagining a theory of secondary traumatic stress Traumatology 2017 23 1 112 123 10.1037/trm0000096
Maas J Schoch S Scholz U Rackow P Schüler J Wegner M Keller R Teachers’ perceived time pressure, emotional exhaustion and the role of social support from the school principal Social Psychology of Education 2021 24 2 441 464 10.1007/s11218-020-09605-8 34776753
Madigan DJ Kim LE Towards an understanding of teacher attrition: A meta-analysis of burnout, job satisfaction, and teachers’ intentions to quit Teaching and Teacher Education 2021 105 103425 10.1016/j.tate.2021.103425
Mahatmya D Grooms A Kim J McGinnis D Johnson E Burnout and race-related stress among BIPOC women K–12 educators Journal of Education Human Resources 2021 10.3138/jehr-2021-0007
Mahfouz, J., Greenberg, M., & Rodriguez, A. (2019). Principals’ Social and Emotional Competence: A Key Factor for Creating Caring Schools. 1–14. https://www.researchgate.net/publication/337001604_Principals'_Social_and_Emotional_Competence_A_Key_Factor_for_Creating_Caring_Schools
Malik P Garg P Psychometric testing of the resilience at work scale using Indian sample Vikalpa 2018 43 2 77 91 10.1177/0256090918773922
Mandavia A Robinson GGN Bradley B Ressler KJ Powers A Exposure to childhood abuse and later substance use: Indirect effects of emotion dysregulation and exposure to trauma Journal of Traumatic Stress 2016 29 5 422 429 10.1002/jts.22131 27622844
Mansfield CF Beltman S Broadley T Weatherby-Fell N Building resilience in teacher education: An evidenced informed framework Teaching and Teacher Education 2016 54 77 87 10.1016/j.tate.2015.11.016
Maslach C Leiter MP Understanding the burnout experience: Recent research and its implications for psychiatry World Psychiatry 2016 15 2 103 111 10.1002/wps.20311 27265691
Mason-Williams L Bettini E Morris Mathews H Boveda M Rodgers W Disparities in teachers’ access to schools’ collective social assets based on role, race, and poverty Remedial and Special Education 2022 10.1177/07419325211068170
Masten AS Pathways to integrated resilience science Psychological Inquiry 2015 26 2 187 196 10.1080/1047840X.2015.1012041
Masyn KE Latent class analysis and finite mixture modeling Oxford Handbook of Quantitative Methods in Psychology 2013 10.1093/oxfordhb/9780199934898.013.0025
Miller K Flint-Stipp K Preservice teacher burnout: Secondary trauma and self-care issues in teacher education Issues in Teacher Education 2019 28 2 28 45
Moir T Why is implementation science important for intervention design and evaluation within educational settings? Frontiers in Education 2018 10.3389/feduc.2018.00061
Morse G Salyers MP Rollins AL Monroe-Devita M Pfahler C Burnout in mental health services: A review of the problem and its remediation Administration and Policy in Mental Health and Mental Health Services Research 2012 39 5 341 352 10.1007/s10488-011-0352-1 21533847
Muthén, L. K., & Muthén, B. O. (1998). Mplus User’s Guide: Eigth Edition. https://www.statmodel.com/download/usersguide/MplusUserGuideVer_8.pdf
Prilleltensky I Neff M Bessell A Teacher stress: What it is, why it’s important, how it can be alleviated Theory into Practice 2016 55 2 104 111 10.1080/00405841.2016.1148986
Race and Ethnicity of Public School Teachers and Their Students. (2020). https://nces.ed.gov/pubs2020/2020103/index.asp
Rankin B An overview of research on secondary traumatic stress in K-12 teaching: What we know and what we still need to learn The Educational Forum 2022 86 2 138 150 10.1080/00131725.2020.1860172
Sacks, V., & Murphey, D. (2018). The prevalence of adverse childhood experiences, nationally, by state, and by race or ethnicity. https://www.childtrends.org/publications/prevalence-adverse-childhood-experiences-nationally-state-race-ethnicity
Saloviita T Pakarinen E Teacher burnout explained: Teacher-, student-, and organisation-level variables Teaching and Teacher Education 2021 97 103221 10.1016/j.tate.2020.103221
Salvagioni DAJ Melanda FN Mesas AE González AD Gabani FL de Andrade SM Physical, psychological and occupational consequences of job burnout: A systematic review of prospective studies PLoS ONE 2017 12 10 e0185781 10.1371/journal.pone.0185781 28977041
Samji H Wu J Ladak A Vossen C Stewart E Dove N Long D Snell G Review: Mental health impacts of the COVID-19 pandemic on children and youth: A systematic review Child and Adolescent Mental Health 2022 27 2 173 189 10.1111/camh.12501 34455683
Santoro DA Good teaching in difficult times: Demoralization in the pursuit of good work American Journal of Education 2011 118 1 1 23 10.1086/662010
Santoro DA Demoralized: Why teachers leave the profession they love and how they can stay 2018 Harvard Education Press
Santoro DA The problem with stories about teacher “burnout” Phi Delta Kappan 2019 101 4 26 33 10.1177/0031721719892971
Saunders BE Adams ZW Epidemiology of traumatic experiences in childhood Child and Adolescent Psychiatric Clinics of North America 2014 23 2 167 184 10.1016/j.chc.2013.12.003 24656575
School Pulse Panel. (2022). Institute of Education Sciences. https://ies.ed.gov/schoolsurvey/spp/
Schwartz HL Diliberti MK State of the superintendent: High job satisfaction and a projected normal turnover rate: Selected findings from the fifth American school district panel survey 2022 RAND Corporation
Sell CR Rinke CR Mawhinney L What it takes to stay: Three stories of teacher retention Opportunities and challenges in teacher recruitment and retention: Teachers’ voices across the pipeline 2019 Information Age Publishing 93 119
Shonkoff JP Garner AS Siegel BS Dobbins MI Earls MF Garner AS McGuinn L Pascoe J Wood DL Pediatrics 2012 129 1 e232 e246 10.1542/peds.2011-2663 22201156
Simon K Petrovic L Baker C Overstreet S An examination of the associations among teacher secondary traumatic stress, teacher–student relationship quality, and student socio-emotional functioning School Mental Health 2022 14 213 224 10.1007/s12310-022-09507-4
Skaalvik EM Skaalvik S Teacher job satisfaction and motivation to leave the teaching profession: Relations with school context, feeling of belonging, and emotional exhaustion Teaching and Teacher Education 2011 27 6 1029 1038 10.1016/j.tate.2011.04.001
Skaalvik EM Skaalvik S Job demands and job resources as predictors of teacher motivation and well-being Social Psychology of Education 2018 21 5 1251 1275 10.1007/s11218-018-9464-8
Stamm, B. H. (2002). Measuring compassion satisfaction as well as fatigue: Developmental history of the Compassion Satisfaction and Fatigue Test. In Treating compassion fatigue (pp. 107–119). Brunner-Routledge.
Steiner ED Doan S Woo A Gittens AD Lawrence RA Berdie L Wolfe RL Greer L Schwartz HL Restoring teacher and principal well-being is an essential step for rebuilding schools: Findings from the state of the American teacher and state of the American principal surveys 2022 RAND Corporation
Steiner ED Woo A Job-related stress threatens the teacher supply: Key findings from the 2021 state of the U.S. Teacher Survey 2021 RAND Corporation
Temkin D Harper K Stratford B Sacks V Rodriguez Y Bartlett JD Moving policy toward a whole school, whole community, whole child approach to support children who have experienced trauma The Journal of School Health 2020 90 12 940 947 10.1111/josh.12957 33184886
Twenge JM Cooper AB Joiner TE Duffy ME Binau SG Age, period, and cohort trends in mood disorder indicators and suicide-related outcomes in a nationally representative dataset, 2005–2017 Journal of Abnormal Psychology 2019 128 3 185 199 10.1037/abn0000410 30869927
Villares E Starrett A Limberg D Exploring school counseling during the first wave of COVID-19 Journal of Counseling & Development 2022 100 4 386 398 10.1002/jcad.12433
Winwood PC Colon R McEwen K A practical measure of workplace resilience: Developing the resilience at work scale Journal of Occupational and Environmental Medicine 2013 55 10 1205 1212 10.1097/JOM.0b013e3182a2a60a 24064782
Worth, J., & Van den Brande, J. (2020). Teacher autonomy: How does it relate to job satisfaction and retention? In National Foundation for Educational Research. National Foundation for Educational Research. https://eric.ed.gov/?id=ED604418
Zolopa C Burack JA O’Connor RM Corran C Lai J Bomfim E DeGrace S Dumont J Larney S Wendt DC Changes in youth mental health, psychological wellbeing, and substance use during the COVID-19 pandemic: A rapid review Adolescent Research Review 2022 7 2 161 177 10.1007/s40894-022-00185-6 35252542
| 0 | PMC9741756 | NO-CC CODE | 2022-12-14 23:29:57 | no | School Ment Health. 2022 Dec 11;:1-14 | utf-8 | School Ment Health | 2,022 | 10.1007/s12310-022-09560-z | 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
36502482
24648
10.1007/s11356-022-24648-4
Applications of Emerging Green Technologies for Efficient Valorization of Agro-Industrial Waste: A Roadmap Towards Sustainable Environment and Circular Economy
Performance and emission analysis of blends of bio-oil obtained by catalytic pyrolysis of Argemone mexicana seeds with diesel in a CI engine
Pandey Satya Prakash [email protected]
1
Upadhyay Rakesh [email protected]
1
Prakash Ramakrishnan [email protected]
2
http://orcid.org/0000-0001-8943-3459
Kumar Sachin [email protected]
13
1 grid.448765.c 0000 0004 1764 7388 Department of Energy Engineering, Central University of Jharkhand, Ranchi, India
2 grid.412813.d 0000 0001 0687 4946 Department of Thermal and Energy Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu India
3 grid.448765.c 0000 0004 1764 7388 Centre of Excellence – Green and Efficient Energy Technology (CoE-GEET), CUJ, Ranchi, India
Responsible Editor: Philippe Garrigues
11 12 2022
114
17 9 2022
4 12 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.
Because of their adaptability and user-friendliness, internal combustion engines are widely used for a variety of purposes, including the generation of consistent power as well as transportation. As a result, the question of whether or not these engines can make use of bio-oils is an important one today. Bio-oils derived from biomass pyrolysis differ significantly from those derived from petroleum-based fuels and biodiesel. They could, however, be valuable alternatives to fossil fuels in order to attain a carbon–neutral future. Bio-oil obtained from catalytic pyrolysis of Argemone Mexicana seed using titanium oxide (TiO2) nanoparticles was employed in IC engine to check its suitability as an alternative fuel. Engine performance analysis was conducted at B5, B10, B15, B20, B25, and B30 blend for different parameters such as brake thermal efficiency, exhaust gas temperature, and brake-specific fuel consumption with change in engine load. Emission analysis was also carried out for carbon monoxide, hydrocarbon, nitrogen oxides, and carbon dioxide emissions. It was found that B30 blend resembled best performance and the bio-oil produced from catalytic pyrolysis of Argemone mexicana seed can be utilised as a substitute of fossil fuels in IC engine.
Graphical Abstract
Keywords
Argemone mexicana seeds
TiO2 nanoparticles
Performance and emission analysis
CI engine
Renewable fuel
==== Body
pmcIntroduction
According to British Petroleum’s (BP) statistical analysis, the overall R/P ratio for crude oil is predicted to be 49.9. Therefore, based on the resource (R) that is currently accessible and the pace of production (P), the world’s crude oil reserves will run out within the next 50 years (BP Statistical Review of World Energy 2020). Therefore, it would be unavoidable for its supply to run out. Additionally, burning fossil fuels releases greenhouse gases (GHGs) into the atmosphere, which pollutes it and causes global warming (Perera et al. 2019). The modern world is seriously threatened by global warming. The Lancet paper states that the global temperature has risen by more than 1.2 °C since the pre-industrial era, which has serious health consequences and impacts global food security. It should be noted that the five hottest years to date have all occurred between 2015 and 2020 as a result of global warming. Additionally, over the past 20 years, there has been a 53.7% increase in heat-related mortality. This has exacerbated the labour dilemma, particularly in nations like India and Indonesia. The emissions from factories and automobiles are the main causes of this accelerated global warming and air pollution. This is seen by the 8% drop in GHGs brought on by the COVID-19 pandemic, which stopped the majority of industry and transportation activities (Watts et al. 2021).
The globe currently has a high demand for energy and is suffering from the effects of global warming as a result of the increase in population. Petroleum fuel consumption is also rising daily in the meantime. The production of energy has significantly expanded in recent years in order to balance the demand for fossil fuels and to limit the use of renewable sources (Souza et al. 2016). By providing additional financing for research activities, which addresses fuel demand and pollution management, the government also promotes the use of alternative fuel in place of fossil fuel. Many research studies suggest that bio-oil can be produced from renewable sources using thermochemical and chemical conversion methods. Due to the many components present, the physical characteristics of the bio-biodiesel differ. For the manufacture of biofuel, primarily non-edible, oil-bearing seeds and other organic biomass wastes have been utilised (Kyari 2008, Prakash et al. 2013) because there is fierce rivalry for edible oil-bearing seeds. The most common procedure of chemical and thermochemical conversion is the production of biodiesel through thermal cracking. Transesterification and esterification are the two most frequent chemical processes used to create biodiesel, and they depend on the amount of free fatty acids present in the source oil (Madras et al. 2004). In order to create ester and glycerol from triglyceride (raw oil), these two chemical processes alter the structure of the compounds. However, the transesterification process uses triglyceride and alcohol (butanol, ethanol, propanol, and methanol) with a catalyst (NaOH or KOH), whereas the esterification process uses concentrated H2SO4 to reduce the excess of the majority of research findings indicated that non-edible seeds. Baiju et al. (2009) conducted tests to determine the properties of Karanja oil’s ethyl and methyl ester.
Both esters can be used as fuel for CI engines; however, the methyl ester is more effective and emits fewer pollutants than the ethyl ester. The amount of biodiesel produced depends on a number of factors, including the molar ratio of the alcohol and oil, reaction temperature, stirring rate, and reaction time. The esterification process and butanol were used to turn mahua into biodiesel (Sathish Kumar and Krupa Vara Prasad 2019). The best circumstances were discovered to be a molar ratio of 6:1, 1.5% (w/w) KOH, 80 °C of temperature, 90 min of residence time, and 500 rpm of stirring rate for the formation of 94.8% of butyl ester. It goes on to describe how different processing variables impact the biodiesel yield. Thermal cracking or pyrolysis, a developing process for the creation of alternative fuels, produces bio-oil from any kind of biomass. The main end products are char, gaseous compounds, and bio-oil since the bio-oil comprises complex mixtures of water and numerous organic molecules such sugars, alcohols, acids, esters, and hydrocarbons (Shadangi and Mohanty 2014a, b, Sharuddin et al. 2016). The fuel attributes of bio-oil have been improved through the use of upgrading procedures in order to be used for better utilisation in CI engines because they are unfavourable for alternative fuel. Ghodke et al. upgraded pyrolytic oil from wood by the esterification process and stated that the pH valve and kinematic viscosity were decreased by mixing the bio-oil with alcohol, which can be used as fuel (Ghodke et al. 2015). The jatropha methyl ester (JME) and tyre pyrolysis oil (TPO) were utilised by Sharma et al. as fuel blends in ratios ranging from 10 to 50% on a volume basis (Sharma and Murugan 2013). The results indicate that the blend 20% performs better than other blends. It demonstrated that employing several types of biomass-derived pyrolysis oil from julifora seeds, and coconut shells significantly reduced NOx (Masimalai and Kuppusamy 2015). By using bio-based materials like biomass–related solid wastes and non-edible seeds as well as bio-based inputs for fuel production, Tenzin Tseten et al. discussed the use of bio-fuel and its manufacturing method. According to the survey, both bio-oil and biodiesel are used in CI engines. When compared to neat biodiesel, the biofuel (biodiesel and bio-oil) performed better and produced fewer emissions at all loads (Tseten and Murthy 2014). Because of its high surface area, robust contact with the metal support, chemical stability, and acid–base property, titanium dioxide has demonstrated a great potential for use as a heterogeneous catalyst in biodiesel production (Carlucci et al. 2019). However, the application of titanium dioxide as a catalyst in pyrolysis of biomass is very limited. To the best of our knowledge, this is the first study to report the engine performance and emission analyses using the bio-oil obtained from catalytic pyrolysis of Argemone mexicana seed. In the present study, titanium oxide nanocatalyst was synthesised through sol–gel method and used in pyrolysis of Argemone mexicana seeds. The obtained catalytic bio-oil was characterised for physical properties, GC–MS, FTIR, NMR and CHNS. The blends prepared from catalytic bio-oil and diesel were investigated in a test engine for the suitability of bio-oil as an alternative fuel.
Materials and methods
Raw material preparation and characterisation
Fruits of the Argemone mexicana species were gathered at the Rajhara village’s agricultural region in the Palamau district of Jharkhand, India. It has geocoordinate of latitude 23.36° N and longitude 85.33° E. There are several seeds inside every fruit. Argemone mexicana’s fruits are covered in thorns; hence, some of the thorns are also picked along with the seeds. Thorns were removed using conventional techniques, and seeds were sun-dried for 8 h. The seeds were ground into a powder and then heated in a hot air oven for 24 h at 60 °C. Using the CHNS-392 (LECO) Elemental Analyzer, powdered seeds were further analysed in order to estimate their amounts of carbon, hydrogen, nitrogen, sulphur, and oxygen. Raw materials were subjected to thermogravimetric analysis (TGA) using a TGA 4000 system, which operates at 100–240 V and 50–60 Hz (Perkin Elmer). For the TGA experiment, 10.45 mg of the powdered seeds was used. Our earlier work contains the results of the in-depth characterisation of Argemone mexicana seeds (Pandey and Kumar 2020).
Synthesis and characterisation of titanium oxide nanoparticles
Titanium oxide nanoparticles were synthesised through sol–gel procedure as shown in Fig. 1.Fig. 1 Scheme for synthesis of titanium dioxide nanoparticles
The sol–gel method was used to produce titanium oxide nanoparticles. Titanium iso-propoxide and 2-propanol were mixed at a ratio of 1:10, respectively. The mixture was mixed continuously for 24 h following the addition of 0.5 ml of water drop by drop after 15 min. After then, the gel was let to sit for 12 h. The final product underwent a 12-h drying process at 80 °C. The powder was shaped into particles and then heated to 650 °C in air at a pace of 40° per minute to create crystalline nanoparticles.
Experimental procedure for catalytic pyrolysis and characterisation of obtained bio-oil
As shown in our earlier work, the experiment was conducted in a reactor-furnace system with temperature regulation provided by a PID controller (Pandey et al. 2020). For the pyrolysis experiments, a reactor containing synthesised TiO2 nanoparticles with a catalyst concentration of 5wt%, 10wt%, 15wt%, and 20wt% was heated to a constant temperature of 550 °C. After each trial, the reactor was allowed to cool to ambient temperature and the leftovers were saved as bio-char. Bio-oil and bio-char yields were derived on a weight-percentage basis, whereas the yield of non-condensable gas was determined using a difference formula. The catalytic oil’s composition was analysed by Fourier transform infrared spectroscopy (FTIR) of Jasco FT/IR-300 E Model to determine functional groups present. Using gas chromatography–mass spectrometry (model number 450-GC, 240-MS; make: M/s Varian) and a capillary column with dimensions of 50 m 0.2 mm (0.33 m film thickness), the presence of organic components in the bio-oil was determined. The HNMR spectra were analysed using a 300 MHz Varian Universal INOVA-300 spectrometer by mixing the nanocatalyst sample with the necessary amount of tert-butanol in D2O and then recording the chemical shifts relative to the tert-butanol CH3 signal, which was set up at 1.20 ppm. A transmitter was used to dampen the water resonance by presaturation, and the deconvolution of the resulting spectrum with lorentzian peaks was used to examine the peak regions. The density, viscosity, specific gravity, sulphur content, cetane index, Conradson carbon residue, pour point, flash point, pH, cloud point, fire point, and boiling point of catalytic bio-oil were determined to ascertain its suitability as an alternative feedstock. A greater heating value of catalytic bio-oil was established using a bomb calorimeter (model: C 2000 basic IKA-bomb calorimeter, Germany). While density was measured with an Anton Paar density metre, calorific value was calculated using an oxygen bomb calorimeter (Parr 6100, Parr Instruments, Molin, IL, USA) (DMA 4500 M, Ashland, VA, USA). The specific gravity was determined with the aid of a Fisherbrand 150-mm-long precision specific gravity hydrometer (Fisher Scientific, Pittsburgh, PA). Combustion was performed in a tubular furnace at a high enough temperature for the LECO TruSpec S 630–100-700 elemental analyser to accurately measure the sulphur content. The ASTM D 86 standard method using the three-distillation process procedure was used to calculate the cetane index. The ASTM D 189, D 93, and D 97 standards were used to calculate the Conradson carbon residue, flash point, and pour point, respectively (Pandey and Kumar 2020).
Experimental set up of CI engine and test detail
Figure 2 demonstrates the schematic representation of the experimental setup. The electrical dynamometer was coupled with the engine for providing load to the engine. The air box along with the U-tube manometer was attached for the engine intake. The U-tube manometer was used for the measurement of the air consumption inside the engine. A burette equipped with a three-way cock was used for the measurement of the fuel consumption. One end of the cock is connected with fuel line to the fuel tank while the second and the third end of the same cock was attached to the burette and the fuel supply system respectively. The stopwatch was used to measure the fuel flow rate on a volumetric basis. A thermocouple (Chromel alumel) and the digital temperature indicator were attached for the exhaust gas temperature measurement.Fig. 2 Schematic diagram of experimental setup for CI engine
The engine exhaust emissions were analysed through the AVL DiGas analyser. The AVL 437 c diesel smoke metre was applied to measure the smoke density. The exhaust gas sample from the engine was received with the help of the probe. The rated engine speed having 1500 rpm was used to conduct the experiments. Catalytic bio-oil and the diesel blends of 5%, 10%, 15%, 20%, 25%, and 30% were applied for testing in the engine. The blends used for the experiment are named as B5, B10, B15, B20, B25, and B30 in which the numbers denote the percentage of catalytic bio-oil used for blending with the diesel. Due to excess knocking effect in the engine that might be caused by longer ignition delay which restricted the use of higher blends, each experiment was done by starting the engine with blends only while diesel was used during the end of the experiment to flush out the blended oil from the injection system and the fuel line.
Results and discussions
Influence of catalyst concentration on the amount of product produced
Figure 3 shows the yield of bio-oil, non-condensable gas, and bio-char in catalytic pyrolysis at different concentrations of the catalyst (5%, 10%, 15%, and 20%). When comparing thermal pyrolysis of Argemone mexicana seed to its treatment with nano TiO2, a notable shift in the distribution of the final products is seen (Pandey and Kumar 2020).Fig. 3 Figure demonstrating effect of catalyst concentration on product yield
When thermal pyrolysis was performed at an ideal temperature of 550 °C, a yield of 52% bio-oil was achieved. The yield of bio-oil increased by 5.75%, 12.98%, 22.29%, and 20.76% when the concentration of the catalyst was 5, 10, 15, and 20%, respectively. With a 5 wt% catalyst ratio, bio-oil production increased to 57.75%. When using a catalyst at 15 wt%, it could grow by as much as 74.29%. Titanium oxide nanoparticles speed up the chemical breakdown, leading to a higher bio-yield (Li et al. 2021). Nanocrystalline zeolite applications have shown significant feedstock conversion. Nanocrystalline zeolite applications have shown significant feedstock conversion (Mohan et al. 2022). In addition, the release of gases was decreased up to the catalyst loading of 15% and then increase for higher percentage of catalyst. When the catalyst loading was increased to 15wt% from 5wt%, the gas yield dropped drastically by 44%. It can also be seen that when nanocatalyst was added, the percentage of oxygen dropped from 14.462% to 9.58%. As catalyst concentration was increased, the yield of bio-char decreased, with the lowest yield obtained at 15% catalyst concentration (12.57% non-condensable gas and 13.14% bio-char).
Characterisation of liquid product
FTIR
Figure 4 illustrates the study of the bio-oil produced by thermal and catalytic processes using Fourier transform infrared spectroscopy. It shows the presence of different functional groups present in the bio-oils. A broad peak at the wavelength 3384 cm-1 corresponds to O-Hstr in case of catalytic pyrolysis oil which confirms the presence of alcoholic/phenolic compounds in it. The same signal may be due to N-Hstr of amides. The presence of a signal at 1200 cm−1 corresponds to C-Ostr further supports the presence of phenolic compounds in catalytic pyrolysis oil and is missing in thermal pyrolysis oil. Again, the presence of a weak signal at 1060 cm−1 corresponds to C-Ostr for alcohols. These peaks for the oil from thermal pyrolysis are not prominent which signifies the presence of very small/absence of alcohols. The signals at 1200 cm-1and1060cm-1 may also correspond to C-Nstr vibration. The strong peak obtained in the wavelength range of 2924 cm−1 and 2857 cm−1 is because of the C-Hstr. The peaks at 1715 cm-1 and 1644 cm-1 relate to C = Ostr, which confirms the presence of carboxylic acids, amides, and carbonyl compounds in the oil. The band at 735 cm-1 refers to N-Hbending of amides. The peak at 1644 cm−1 may also correspond to C = Cstr. Similarly, a weak signal at 2330 cm−1 corresponds to the C≡Cstr. From the above analysis of the FTIR plot, it can be analysed that the bio-oil sample may contain different organic compounds like saturated and unsaturated hydrocarbons, phenols, alcohols, carboxylic acid, and amides, which could be further confirmed by GC–MS results.Fig. 4 FTIR analysis of Argemone mexicana catalytic bio-oil
GC–MS
The GC–MS analysis was used to perform a complete compositional study of thermal and catalytic bio-oil samples produced under ideal conditions, and the comparative results are summarised in Table 1. There are 83 separate components in the pyrolytic oil that is produced during catalytic pyrolysis against the 54 components in thermal pyrolysis, consisting of alkanes, alkenes, aromatics, alcohols, phenols, carboxylic acids, ketones, ethers, esters, nitriles, amines, and amides.Table 1 GC–MS compositional analysis of obtained bio-oil
S.No. Nature of compounds Concentration obtained from catalytic pyrolysis (%) Concentration obtained from thermal pyrolysis (%) (Pandey and Kumar 2020)
1. Saturated hydrocarbons 8.73 2.75
2. Unsaturated hydrocarbons 16.99 26.49
3. Cycloalkanes/aromatics 4.34 0.38
4. Carboxylic acids 41.12 41.46
5. Carbonyl compounds 0.91 0.3
6. Alcoholic 0.5 1.9
7. Nitriles 14.46 13.05
8. Amides 10.04 11.37
9. Phenolic 2.55 0
10. Ether/esters 0.36 2.3
This indicates the complex nature of the oil. Unsaturated hydrocarbons, carboxylic acids, nitriles, and amides are the main substances found in both the thermal and catalytic pyrolysis oil. The effect of nano titanium catalyst facilitates the cracking of the biomass to simpler components, which is reflected in Table 1. Three significant changes are detected in the catalytic pyrolysis, as observed from Table 2.Table 2 Distribution of products as per the carbon number
S.No Compound C number Concentration obtained from catalytic pyrolysis (%) Concentration obtained from thermal pyrolysis (%) (Pandey and Kumar 2020)
1. C6-C8 5.17 0
2. C9-C11 5.66 3.06
3. C12-C14 13.88 21.05
4. C15-C17 32.84 23.61
5. C18-C20 41.96 40.58
6. > C21 0.49 11.7
Firstly, there is formation of lighter fractions containing C6–C8 components (5.1%) in presence of catalysts, which are not present at all in thermal process. Further, an increase in the concentration of C9-C11 components and drastic decrease in concentration > C21 components are observed in presence of the catalyst. Secondly, significant increase of saturated hydrocarbons and decrease of unsaturated hydrocarbons are observed in presence of catalyst. Thirdly, the concentration of carboxylic acids, amides, and nitriles is unaffected by the addition of catalyst. The high surface area of the nanocatalyst must be facilitating the cracking of the biomass into simpler products. In addition, it also facilitates hydrogenation of the unsaturated products obtained, thus shows an increase in saturated hydrocarbons.
For the purpose of improving thermal comfort, hexadecane has been used commercially in the production of thermo-regulated fabrics for garments (Salaün et al. 2010). It is also used in the mass production of personal heating and cooling appliances, where it has been incorporated into textiles (Scaringe et al. 1990), thermal insulation purposes in protective textiles as pillowcase, duvets, blankets and mattresses in household textiles (Nelson 2002). Phenyloctanoic acid is used in development of intracytoplasmic lipid inclusions (Alvarez et al. 1996). Tetradecene is widely employed in hydroformylation and copolymerisation whereas dodecyne (Haumann et al. 2002; Kotzabasakis et al. 2009) is utilised in transition-metal catalysed silylmetalation of acetylenes (Hayami et al. 1983). Heptadecanoic acid, hexadecanol, and heptadecanenitrile are utilised in binary Langmuir films (Williams et al. 1993) whereas octadecadienoic acid found its utility in chemoenzymatic synthesis of coriolic acid (Martini et al. 1994). Oleanitrile is employed in nutraceutical applications (Fondevila et al. 2019) and the endogenous lipid that induces sleep is hydrolysed by oleamide hydrolase, which is inhibited by 9-octadecenamide (Patterson et al. 1996).
1HNMR analysis
Figure 5 illustrates the distribution of hydrogen in the bio-oil 1 HNMR spectrum. Based on the chemical shifts of particular proton types, 1 HNMR spectra can be split into four main regions: aromatic, olefinic, aliphatic, and transient. For the compounds produced by catalytic breakdown, there are more protons connected to hydroxyl groups or ring-joined methylene. It is seen that there were no aromatic compounds found between the range of 4.0–3.3 ppm. In the bio-oil produced by catalytic pyrolysis, protons were connected to CH3 or upwards through an aromatic ring in about 7.6% of the cases. This shows that the huge ratio of aliphatic structure unit dominated the bio-oil obtained through catalytic pyrolysis.Fig. 5 1HNMR spectrum of catalytic bio-oil
Physical properties
Table 3 displays the characteristics of the thermal and catalytic fuel produced under ideal pyrolysis circumstances. When comparing the results of Table 3, it clearly demonstrates how catalysis affects the physical characteristics of bio-oil.Table 3 Physical properties of obtained catalytic bio-oil
Physical properties Argemone Mexicana seed catalytic pyrolysis oil Argemone Mexicana seed thermal pyrolysis oil (Pandey and Kumar 2020) Diesel (Pandey and Kumar 2020)
Appearance Dark black Dark black Little brown with green tint
Odour Smoky smell Smoky smell Aromatic
Density @15 °C 931.1 kg/m3 932 kg/m3 832 kg/m3
Specific gravity@ 30 °C 0.9323 0.925 0.82
Viscosity @ 40 °C 19.44 30.30cst 32.78cst
Viscosity @ 100 °C 16.69 8.28cst 1.21cst
Conradson carbon residue 2.78% 3.19% -
Flash point 65 °C 70 °C 52°–96 °C
Pour point − 12 °C − 10 °C − 20 °C
Gross calorific value 34.67 MJ/kg 27.39 MJ/kg 41.86 MJ/kg
Sulphur content 0.18% 0.30% 0.05%
pH 5.23 5.47 6–8
Boiling point - 132 °C 180–360 °C
Fire point 66 °C 80 °C 62–106 °C
Cetane index 25.49 27.20 40–60
The bio-oil obtained from both the processes is dark in colour with smoky odour. The density and specific gravity of the oil have not changed significantly but the viscosity of the oil is reduced drastically from 30.30cst to 19.44cst when nanocatalyst is used. The decrease in viscosity can be explained due to the formation of simpler components due to cracking and more water as the catalyst facilitates dehydration. This is also confirmed by the studies undertaken by the previous works which report that there is always an increase in water formation in catalytic pyrolysis (Shadangi and Mohanty 2014a, b). The catalyst used in this work, i.e. titanium oxide, enhanced content of water in the obtained bio-oil through pyrolysis which also significantly reduced the viscosity. A significant enhancement in the calorific value of the oil obtained in catalytic pyrolysis is observed when compared with AM seed thermal pyrolysis oil as well as diesel fuel. This could be explained by the oil’s rising carbon, hydrogen, and falling oxygen content. During the catalytic-pyrolysis, the molecules of the oxygen react with molecules of hydrogen to produce water, which might decrease the viscosity and content of oxygen. In addition to this, because there are more carbon and hydrogen atoms present in the catalytic bio-oil than in the thermal bio-oil, it has a higher heating value. A decrease in flash point, fire point, pour point, and cetane index of the oil is found by pyrolysis of the biomass in presence of nano TiO2. The fire point and flash point are in the range of diesel. The pour point and cetane number of the AM seed catalytic pyrolysis oil were improved when comparing with diesel. The pour point is very useful for the evaluation of the quality of the fuel and depends mainly on the chemical structure and the component composition of the bio-oil and it also determines the fuel fluidity at low temperatures.
CHNS analysis
The percentage of carbon, hydrogen, nitrogen, sulphur, and oxygen in the Argemone mexicana catalytic bio-oil is shown in Table 4.Table 4 Elemental analysis of Argemone mexicana catalytic bio-oil
S.No. Element Amount (%)
1. Carbon 71.24
2. Hydrogen 11.45
3. Nitrogen 7.69
4. Sulphur 0.04
5. Oxygen 9.58
The percentage amount content of carbon and hydrogen demonstrates good fuel combustion properties. Lower content of sulphur demonstrates good characteristics of emission. Content of carbon in the bio-oil was observed to be 71.24% by weight while the content of hydrogen was found to be 11.45% by weight. Additionally, contents of nitrogen and sulphur obtained were 7.69% and 0.04% by weight respectively with content of oxygen found to be 9.58% by weight. In order to substitute the obtained bio-oil with liquid fuel, the molar ratio of O/C must be less and the molar ratio of H/C must be higher (Zhang et al. 2021). It can be determined from the table that the molar ratio of O/C is 0.1344 while the molar ratio of H/C was found to be 0.1607 which lays among the likes of light and heavy petroleum oils, demonstrating that biofuel is a better replacement for the standard fuels. Additionally, the sulphur proportion was found to be quite low which shows the lesser possibility to pollute the environment.
Performance and emission analysis of catalytic bio-oil in a test engine
Brake-specific fuel consumption
Brake-specific fuel consumption represents charge efficiency of the engine indicating how effectively an engine fuel is changed into work done. Figure 6 shows the variations in brake-specific fuel consumption (BSFC) with respect to engine loads for different blends. It was observed that brake-specific fuel consumption for higher blends were more than that for lower blends due to their low calorific values (Dhar and Agarwal 2014; Salam and Verma 2019). At the beginning of the engine load at 25–50%, the BSFC decreases linearly in a small quantity. However, as the load on the engine increases to 50%, the BSFC decreases gradually 0.3542 0.75 kg/kwh, 0.383 kg/kwh, 0.386 kg/kwh, 0.3826 kg/kwh, 0.3733 kg/kwh, and 0.378 kg/kwh for B5, B10, B15, B20, B25, and B30 blend respectively. At 75% load, BSFC values were 0.289 kg/kwh, 0.28 kg/kwh, 0.283 kg/kwh, 0.2822 kg/kwh, 0.296 kg/kwh, and 0.289 kg/kwh for B5, B10, B15, B20, B25, and B30 blend respectively. At full load, brake-specific fuel consumption was 0.239 kg/kwh, 0.25 kg/kwh, 0.2687 kg/kwh, 0.2648 kg/kwh, 0.2563 kg/kwh, and 0.251 kg/kwh for B5, B10, B15, B20, B25, and B30 blend respectively. Increasing proportion of bio-oil in the blend would increase the calorific value of the fuel (Agarwal et al. 2015, Agarwal and Dhar 2013). It is also clear that the B15 blend has the highest BSFC at the full load (100%). It can also be seen that the B30 blend has the lowest BSFC at full load capacity.Fig. 6 Brake-specific fuel consumption of catalytic bio-oil with respect to engine load
Brake thermal efficiency
The brake thermal efficiency, expressed as the ratio of work developed divided by the heating value of the fuel, is an essential parameter for analysing to how well an engine utilises chemical energy of the fuel to produce work. Figure 7 shows the brake thermal efficiency at varying loads from 25% to full load at 100%.Fig. 7 Brake thermal efficiency of catalytic bio-oil with respect to engine load
BTE first increased with the increase in the engine load for all the tested fuels and then gradually decreased after 75% load. BTEs for B5, B10, B15, B20, B25, and B30 were about 34.69%, 31.24%, 32.03%, 32.36%, 34.85%, and 35.02%, respectively. The brake thermal efficiencies with 30% blends were higher compared to those with lower blend at 100% load because of lower viscosity of the blend leading to increase in the atomisation and vaporisation of the bio-oil. It is clear that the B30 blend of the catalytic bio-oil has the highest efficiency. Also, it is evident that the B10 blend has the lowest brake thermal efficiency at full load which is due to higher viscosity of the bio-oil and higher latent heat of vaporisation required for the blend (Banapurmath et al. 2008, Qi et al. 2010); however, when the load is around 60–80%, the B10 blend exhibited enhanced efficiency which decreases at higher loads. This proves that when the load is full, the B25 and B30 blend can be efficient, however, as the load increases; the B30 blend proves to be the most efficient. Higher viscosity, density, and evaporation property of bio-oil would lead in formation of large droplet throughout fuel atomisation, which results in the evolution of irregular mixture of fuel and air. It can be concluded that as the load increases, the specific fuel consumption decreases as a result of the highest brake thermal efficiency being at the top for the full load. It is also clear from the above explanation and study that the brake thermal efficiency is inversely proportional to the brake-specific fuel consumption.
Carbon monoxide emission
Carbon monoxide (CO) is generally evolved in the combustion chamber as consequence of incomplete fuel combustion in the absence of oxygen. At a higher load of bio-oil/diesel blends, a significant decrease in CO emissions for B10 blend can be observed in Fig. 8.Fig. 8 Carbon monoxide emission of catalytic bio-oil with respect to engine load
For all the blends except B10 blend, carbon monoxide emissions first drastically decrease up to 50% load and then gradually decrease beyond 50% load. Gradual decrease may be caused by excessive amount of oxygen available inside the combustion chamber. This process in turn leads to oxidation of CO to CO2. The peak CO emission concentrations found for bio-oil/diesel blend B5 at L1, L2, L3, and L4 of diesel fuel, were 0.38%, 0.23%, 0.21%, and 0.2% by volume, respectively. For Argemone mexicana bio-oil/diesel blend B10, the maximum amounts of CO at L1, L2, L3, and L4, were found to be 0.4%, 0.39%, 0.38%, and 0.36% by volume, respectively. For blend B15, the maximum amounts of CO at L1, L2, L3, and L4, were found to be 0.47%, 0.23%, 0.11%, and 0.1% by volume, respectively. For blend B20, the maximum amounts of CO at L1, L2, L3, and L4, were found to be 0.32%, 0.24%, 0.31%, and 0.91% by volume, respectively. Similarly, for blend B25, the maximum amounts of CO at L1, L2, L3, and L4, were found to be 0.57%, 0.12%, 0.11%, and 0.098% by volume, respectively, while for blend B30, the maximum amounts of CO at L1, L2, L3, and L4, were found to be 0.43%, 0.24%, 0.11%, and 0.091% by volume, respectively. Here, L1, L2, L3, and L4 refer to 25%, 50%, 75%, and full load respectively. Further, higher CO emissions at low load conditions are due to incomplete bulk gas reactions. It may be possible that large amounts of oxygen available in bio-oil/diesel blends support complete effective combustion in the engine cylinder and thereby help in reduction of carbon monoxide formation (Yusaf et al. 2013; Tutak et al. 2017; Datta et al. 2014).
Carbon dioxide emission
Carbon dioxide (CO2) is a hazardous chemical compound that contributes to the greenhouse effect (Osman et al. 2021). Higher emissions of carbon dioxide symbolise complete fuel combustion in the combustion chamber. It can be observed from Fig. 9 that B10 blend exhibited higher carbon dioxide emission than B5 blend due to higher oxygen concentration in the combustion chamber whereas emission decreased with increase in bio-oil proportion in the blend for B15, B20, B25, and B30 blend.Fig. 9 Carbon dioxide emission of catalytic bio-oil with respect to engine load
Carbon monoxide gets oxidised to carbon dioxide due to presence of more amount of oxygen supporting combustion process. In addition, CO2 emissions rose as the load for biodiesel/oil blends. Compared to B10, B15, B20, B25, and B30 bio-oil blends, B10 was shown to emit greater carbon dioxide at higher load (L4). At full load condition, the carbon dioxide emissions for B5, B10, B15, B20, B25, and B30, were found to be about 1.11%, 1.17%, 0.91%, 0.72%, 0.68%, and 0.45%, by volume, respectively. The oxygen content of the bio-oil is responsible for production of leaner mixture. CO2 emissions increased due to more temperatures at higher loads, thus improving the combustion phenomenon in combustion chamber.
Exhaust gas temperature
The variation of exhaust gas temperature with applied load for different blends is shown in Fig. 10. The result indicates that the exhaust gas temperature increases with increase in proportion of bio-oil in the blend. The highest temperature obtained is 220 ℃ for B30 blend whereas the maximum temperature is only 212 ℃, 205 ℃, 202 ℃, 198 ℃, and 195 ℃ for the blend B25, B20, B15, B10, and B5 respectively. The decreased calorific value of blended fuel is the cause of the decrease in exhaust gas temperature that occurs with a decrease in the amount of bio-oil. Higher proportion may be due to reduced exhaust loss (Hebbal et al. 2006). Exhaust gas temperature increases with respect to engine load due to the fact that additional amount of fuel was required for generating extra power by the engine (Hebbal et al. 2006; Lucas et al. 2008). Lower thermal efficiency of engine leads to enhanced exhaust gas temperature as reduced proportion of energy available in the fuel is transformed into work (Nagarajan et al. 2002).Fig. 10 Plot of exhaust gas temperature with change in engine load
It is clear from Fig. 10 that the B30 blend has the highest exhaust gas temperature which indicates that the heat recovery is the best in the B30 blend. Also, it indicates that the heat transfer rate of the B30 blend is the best among all the selected blends. It is also an indication that the fuel/air ratio provided to the engine is perfect for the combustion and as the heat transfer rate is also good which leaves no or less fuel unburned in the engine.
Hydrocarbon emission
Unburned hydrocarbon is an essential parameter to analyse combustion inefficiency. Unburned hydrocarbon indicates unburnt fuel. The term “hydrocarbon” commonly refers to both solid hydrocarbons found in particulate matter and organic molecules in the gaseous state (Walendziewski 2002). Low exhaust temperatures, lean fuel air mixture regions, and surplus air may survive and escape into the exhaust at reduced loads. In the case of waste plastic oil, the level of unburned hydrocarbon is lower at lower load ranges due to charge uniformity and better oxygen availability, but it is larger at higher load ranges due to more fuel being admitted (Mani et al. 2009).
Figure 11 displays the variation of unburned hydrocarbon with load for the tested fuels. As the load increases, the hydrocarbon emission decreases. In the case of the B5 blend, HC varies from 112 ppm at low load to 110 ppm at full load, while in the case of the B10 mix, it varies from 121.34 ppm at low load to 118.23 ppm at full load. Similar variations are shown for B15 mix, B20 blend, B25 blend, and B30 blend, with 128.45 ppm at low load and 119.95 ppm at full load, 131.94 ppm at low load and 128.75 ppm at full load, and 138.12 ppm at full load and 147.23 ppm at low load and 139.46 ppm at full load, respectively. More fuel supplied at low load is responsible for more hydrocarbon emission. Emission of unburned hydrocarbon is less at lighter loads owing to charge homogeneity and more availability of oxygen, whereas emission increases at higher loads due to more quantity of fuel admitted (Das et al. 2022).Fig. 11 Amount of hydrocarbon emission with engine load
Oxides of nitrogen emission
Combustion zones having high temperatures and excess air exhibit nitrogen oxide emissions (Osman 2020). The amount of NOx produced in any area is dependent on the amount of oxygen available along with peak temperature and temperature gradient obtained in the initial stage of combustion. The surplus air–fuel ratio at the highest combustion efficiency was found to produce the highest NOx emissions in the exhaust gas. Because there is enough air in the cylinder, the combustion process in CI engines often takes place at high temperature and high-pressure conditions, which is what produces NOx (Tutak et al. 2016; Jamrozik et al. 2017).
Figure 12 shows varied NOx emission concentrations measured in ppm under various engine conditions and engine load. The NOx emissions from the engine rose for various bio-oil/diesel blends as the engine load increased. Peak NOx concentrations for B5, B10, B15, B20, B25, and B30 were 200, 220, 80, 50, 40, and 20 ppm, respectively, at a 50% engine load. Peak NOx concentrations for B5, B10, B15, B20, B25, and B30 were 201, 210, 100, 55, 48, and 25 ppm, respectively, at an engine load of 80%. It can be observed from the figure that NOx emissions increased with increase in load for all fuel blends due to injection of more fuel in the cylinder leading to increase in peak cylinder pressure, temperature, and oxygen availability. Further, results reveal that bio-oil mixed with diesel leads to increase in NOx emissions due to the higher oxygen contents in the structure of these fuels reacting with the nitrogen molecules present in the air contributing to the nitrogen oxide production.Fig. 12 Number of oxides of nitrogen emission with engine load
Oxygen
Figure 13 shows how the oxygen content in the exhaust varies in relation to engine load. The graph shows that oxygen concentration in the exhaust decreases as engine load increases, which may be due to titanium oxide’s catalytic function, which encourages the efficient use of oxygen present in a new charge (Patnaik et al. 2017).Fig. 13 Amount of oxygen emission with engine load
Oxygen in exhaust varies from 20.27% (vol.) at low load to 19.8% (vol.) at full load for B5 blend, and it varies from 20.18% at low load to 19.77% at full load for B10 blend. Similarly, for B15 blend, it varies from 20.12% at low load to 19.74% at full load; for B20 blend, it varies from 19.95 to 19.68%; for B25 blend, it varies from 19.91 to 19.42%; and for B30 blend, it varies from 19.82 at low load to 19.31% at full load. However, the B30 blend has proved to be the best for the engine performance as during the initial less loads for the engine, the oxygen release was more than during the full load of the engine. The lesser oxygen release after the combustion proves that the backfiring will not take place during the combustion process because the combustion feed will not flow to the exhaust where a possibility of backfire is confirmed. This would lead to enhancement of the overall efficiency of the engine with least pollution and full combustion of the fuel present in the chamber.
Conclusion
In this study, the experiments were conducted using diesel, bio-oil obtained by catalytic pyrolysis of Argemone mexicana seed and their blends under different loads to investigate the impact of bio-oil on the performance and emission characteristics of compression ignition engine at varying load conditions. Based on the experimental results, the following major conclusions have been drawn:
Brake-specific fuel consumption for higher blends was found more than that for lower blends. Brake thermal efficiency first increased with the increase in the engine load for all the tested fuels and then gradually decreased after 75% load. For all the blends except B10 blend, carbon monoxide emissions first drastically decrease up to 50% load and then gradually decrease beyond 50% load. At full load condition, B10 was found to emit more carbon dioxide, as compared to that of B10, B15, B20, B25, and B30 bio-oil blends. Exhaust gas temperature was found to be 177 °C, 179 °C, 177 °C, 181 °C, 185 °C, and 175 °C for B5, B10, B15, B20, B25, and B30 blend respectively. B30 blend was found to be providing least harmful emission of Sox, NOx etc. as well as lowest brake-specific fuel consumption. The experimental output of the present research can be successfully employed for design of an industrial model for bio-fuel generation from Argemone Mexicana seed.
Acknowledgements
The Ministry of New and Renewable Energy (MNRE), New Delhi, India, and the Centre of Excellence-Green and Efficient Energy Technology (COE-GEET), Central University of Jharkhand, Ranchi, India, have both provided financial support to the authors for use in various research initiatives.
Author contribution
Satya Prakash Pandey: conceptualisation, formal analysis, investigation, methodology, software, visualisation, data curation, writing—original draft. Rakesh Upadhyay: data curation, supervision, writing—review and editing. Ramakrishnan Prakash: data curation, supervision, writing—review and editing. Sachin Kumar: project administration, supervision, validation, resources, writing—review and editing.
Data availability
All data generated or analysed during this study are included in this article.
Declarations
Ethics approval
Not applicable.
Consent to participate
Not applicable.
Consent to publish
Not applicable.
Competing interests
The authors declare no competing interests.
Highlights
• First study concerning the use of titanium oxide (TiO2) as a catalyst in the pyrolysis process.
• Characterisation of bio-oil obtained by catalytic pyrolysis of Argemone mexicana non-edible oilseed.
• BSFC and BTE analysis to investigate the performance of CI engine with blend of pyrolysis oil and diesel fuel.
• CO, CO2, HC, NOx, and EGT analysis to explore the emissions of CI engine with blend of pyrolysis oil and diesel fuel.
• Suitability of obtained catalytic pyrolysis oil as renewable fuel for substitute of fossil fuels.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Agarwal AK Dhar A Experimental investigations of performance, emission and combustion characteristics of Karanja oil blends fuelled DICI engine Renew Energy 2013 52 283 291 10.1016/j.renene.2012.10.015
Agarwal AK Dhar A Gupta JG Kim WI Choi K Lee CS Park S Effect of fuel injection pressure and injection timing of Karanja biodiesel blends on fuel spray, engine performance, emissions and combustion characteristics Energy Convers Manage 2015 91 302 314 10.1016/j.enconman.2014.12.004
Alvarez HM Mayer F Fabritius D Steinbüchel A Formation of intracytoplasmic lipid inclusions by Rhodococcus opacus strain PD630 Arch Microbiol 1996 165 6 377 386 10.1007/s002030050341 8661931
Baiju B Naik MK Das LM A comparative evaluation of compression ignition engine characteristics using methyl and ethyl esters of Karanja oil Renew Energy 2009 34 6 1616 1621 10.1016/j.renene.2008.11.020
Banapurmath NR Tewari PG Hosmath RS Performance and emission characteristics of a DI compression ignition engine operated on Honge, Jatropha and sesame oil methyl esters Renew Energy 2008 33 9 1982 1988 10.1016/j.renene.2007.11.012
BP Statistical Review of World Energy (2020). https://www.bp.com/en/global/corporate/energy-economics/statistical-review-of-world-energy.html. Accessed 9 Dec 2022
Carlucci C Degennaro L Luisi R Titanium dioxide as a catalyst in biodiesel production Catalysts 2019 9 1 75 10.3390/catal9010075
Das AK Sahu SK Panda AK Current status and prospects of alternate liquid transportation fuels in compression ignition engines: a critical review Renew Sustain Energy Rev 2022 161 112358 10.1016/j.rser.2022.112358
Datta A Dutta S Mandal BK Effect of methanol addition to diesel on the performance and emission characteristics of a CI engine J Basic Appl Eng Res 2014 1 3 8 13
Dhar A Agarwal AK Performance, emissions and combustion characteristics of Karanja biodiesel in a transportation engine Fuel 2014 119 70 80 10.1016/j.fuel.2013.11.002
Fondevila MP Manzano PI Choez IA Chemical analysis of Passiflora edulis fo. Flavicarpa leaves and its nutraceutical applications Pharmacol Online 2019 2 322 328
Ghodke P Ganesh A Mahajani S Stabilization of fast pyrolysis oil derived from wood through esterification Int J Chem Reactor Eng 2015 13 3 323 334 10.1515/ijcre-2014-0102
Haumann M Yildiz H Koch H Schomäcker R Hydroformylation of 7-tetradecene using Rh-TPPTS in a microemulsion Appl Catal A 2002 236 1–2 173 178 10.1016/S0926-860X(02)00284-3
Hayami H Sato M Kanemoto S Morizawa Y Oshima K Nozaki H Transition-metal-catalyzed silylmetalation of acetylenes and its application to the stereoselective synthesis of steroidal side chains J Am Chem Soc 1983 105 13 4491 4492 10.1021/ja00351a069
Hebbal OD Reddy KV Rajagopal K Performance characteristics of a diesel engine with deccan hemp oil Fuel 2006 85 14–15 2187 2194 10.1016/j.fuel.2006.03.011
Jamrozik A Tutak W Pyrc M Sobiepański M Effect of diesel-biodiesel-ethanol blend on combustion, performance, and emissions characteristics on a direct injection diesel engine Thermal Science 2017 21 1 Part B 591 604 10.2298/TSCI160913275J
Kotzabasakis V Petzetakis N Pitsikalis M Hadjichristidis N Lohse DJ Copolymerization of tetradecene-1 and octene-1 with silyl-protected 10-undecen-1-ol using a Cs-symmetry hafnium metallocene catalyst. A route to functionalized poly (α-olefins) J Polymer Sci Part A: Polymer Chem 2009 47 3 876 886 10.1016/S0926-860X(02)00284-3
Kyari MZ Extraction and characterization of seed oil Int Agrophys 2008 22 1 139 142
Li Y, Shaheen SM, Rinklebe J, Ma NL, Yang Y, Ashraf MA, ..., Peng WX (2021) Pyrolysis of Aesculus chinensis Bunge Seed with Fe2O3/NiO as nanocatalysts for the production of bio-oil material. J Hazard Mater 416:126012. 10.1016/j.jhazmat.2021.126012
Lucas N Bienaime C Belloy C Queneudec M Silvestre F Nava-Saucedo JE Polymer biodegradation: mechanisms and estimation techniques–a review Chemosphere 2008 73 4 429 442 10.1016/j.chemosphere.2008.06.064 18723204
Madras G Kolluru C Kumar R Synthesis of biodiesel in supercritical fluids Fuel 2004 83 14–15 2029 2033 10.1016/j.fuel.2004.03.014
Mani M Subash C Nagarajan G Performance, emission and combustion characteristics of a DI diesel engine using waste plastic oil Appl Therm Eng 2009 29 13 2738 2744 10.1016/j.applthermaleng.2009.01.007
Mohan I, Panda AK, Volli V, Kumar S (2022) An insight on upgrading of biomass pyrolysis products and utilization: current status and future prospect of biomass in India. Biomass Convers Biorefin 1-19. 10.1007/s13399-022-02833-2
Martini D Iacazio G Ferrand D Buono G Triantaphylides C Optimization of large scale preparation of 13-(S)-hydroperoxy-9Z, 11E-octadecadienoic acid using Soybean lipoxygenase. Application to the chemoenzymatic synthesis of (+)-coriolic acid Biocatalysis 1994 11 1 47 63 10.3109/10242429409034376
Masimalai S Kuppusamy V A comprehensive assessment on performance behavior of a CI engine using bio oil emulsions (PJSO10, KSO10 and CSO10) as fuels J Mech Sci Technol 2015 29 10 4491 4498 10.1007/s12206-015-0948-6
Nagarajan G Rao AN Renganarayanan S Emission and performance characteristics of neat ethanol fuelled Dl diesel engine Int J Ambient Energy 2002 23 3 149 158 10.1080/01430750.2002.9674883
Nelson G Application of microencapsulation in textiles Int J Pharm 2002 242 1–2 55 62 10.1016/S0378-5173(02)00141-2 12176225
Osman AI Mass spectrometry study of lignocellulosic biomass combustion and pyrolysis with NOx removal Renew Energy 2020 146 484 496 10.1016/j.renene.2019.06.155
Osman AI Hefny M Abdel Maksoud MIA Elgarahy AM Rooney DW Recent advances in carbon capture storage and utilisation technologies: a review Environ Chem Lett 2021 19 2 797 849 10.1007/s10311-020-01133-3
Prakash R Singh RK Murugan S Use of biodiesel and bio-oil emulsions as an alternative fuel for direct injection diesel engine Waste Biomass Valorization 2013 4 3 475 484 10.1007/s12649-012-9182-y
Pandey SP Kumar S Valorisation of argemone mexicana seeds to renewable fuels by thermochemical conversion process J Environ Chem Eng 2020 8 5 104271 10.1016/j.jece.2020.104271
Patnaik PP Jena SP Acharya SK Das HC Effect of FeCl3 and diethyl ether as additives on compression ignition engine emissions Sustain Environ Res 2017 27 3 154 161 10.1016/j.serj.2017.01.002
Patterson JE Ollmann IR Cravatt BF Boger DL Wong CH Lerner RA Inhibition of oleamide hydrolase catalyzed hydrolysis of the endogenous sleep-inducing lipid cis-9-octadecenamide J Am Chem Soc 1996 118 25 5938 5945 10.1021/ja954064z
Perera F Ashrafi A Kinney P Mills D Towards a fuller assessment of benefits to children’s health of reducing air pollution and mitigating climate change due to fossil fuel combustion Environ Res 2019 172 55 72 10.1016/j.envres.2018.12.016 30771627
Qi DH Chen H Geng LM Bian YZ Experimental studies on the combustion characteristics and performance of a direct injection engine fueled with biodiesel/diesel blends Energy Convers Manage 2010 51 12 2985 2992 10.1016/j.enconman.2010.06.042
Salam S Verma TN Appending empirical modelling to numerical solution for behaviour characterisation of microalgae biodiesel Energy Convers Manage 2019 180 496 510 10.1016/j.enconman.2018.11.014
Salaün F Devaux E Bourbigot S Rumeau P Thermoregulating response of cotton fabric containing microencapsulated phase change materials Thermochim Acta 2010 506 1–2 82 93 10.1016/j.tca.2010.04.020
Sathish Kumar R Krupa Vara Prasad A Environment friendly butyl ester biodiesel production from mahua oil: optimization and characterization SN Appl Sci 2019 1 8 1 10 10.1007/s42452-019-0913-6
Scaringe RP Buckman JA Grzyll LR Mahefkey ET Leland JE Heat-pump-augmented spacecraft heat-rejection systems J Spacecr Rocket 1990 27 3 318 323 10.2514/3.26142
Shadangi KP Mohanty K Comparison of yield and fuel properties of thermal and catalytic Mahua seed pyrolytic oil Fuel 2014 117 372 380 10.1016/j.fuel.2013.09.001
Shadangi KP Mohanty K Production and characterization of pyrolytic oil by catalytic pyrolysis of Niger seed Fuel 2014 126 109 115 10.1016/j.fuel.2014.02.035
Sharma A Murugan S Investigation on the behaviour of a DI diesel engine fueled with Jatropha Methyl Ester (JME) and Tyre Pyrolysis Oil (TPO) blends Fuel 2013 108 699 708 10.1016/j.fuel.2012.12.042
Sharuddin SDA Abnisa F Daud WMAW Aroua MK A review on pyrolysis of plastic wastes Energy Convers Manage 2016 115 308 326 10.1016/j.enconman.2016.02.037
Souza GK Scheufele FB Pasa TLB Arroyo PA Pereira NC Synthesis of ethyl esters from crude macauba oil (Acrocomia aculeata) for biodiesel production Fuel 2016 165 360 366 10.1016/j.fuel.2015.10.068
Tseten T Murthy TPK Advances and biotechnological applications in biofuel production: a review Open J Renew Sustain Energy 2014 1 2 29 34
Tutak W Jamrozik A Pyrc M Sobiepański M Investigation on combustion process and emissions characteristic in direct injection diesel engine powered by wet ethanol using blend mode Fuel Process Technol 2016 149 86 95 10.1016/j.fuproc.2016.04.009
Tutak W Jamrozik A Pyrc M Sobiepański M A comparative study of co-combustion process of diesel-ethanol and biodiesel-ethanol blends in the direct injection diesel engine Appl Therm Eng 2017 117 155 163 10.1016/j.applthermaleng.2017.02.029
Walendziewski J Engine fuel derived from waste plastics by thermal treatment Fuel 2002 81 4 473 481 10.1016/S0016-2361(01)00118-1
Watts N, Amann M, Arnell N, Ayeb-Karlsson S, Beagley J, Belesova K, ..., Costello A (2021) The 2020 report of the Lancet Countdown on health and climate change: responding to converging crises. Lancet 397(10269):129-170. 10.1016/S0140-6736(20)32290-X
Williams H Armitage DA Linford RG Da Silva AMG Binary Langmuir films involving 1-heptadecanoic acid, 1-hexadecanol, and 1-heptadecanenitrile J Colloid Interface Sci 1993 156 2 438 445 10.1006/jcis.1993.1136
Yusaf T Hamawand I Baker P Najafi G The effect of methanol-diesel blended ratio on CI engine performance Int J Automot Mech Eng 2013 8 1 1385 1395 10.15282/ijame.8.2013.26.0114
Zhang C, Chao L, Zhang Z, Zhang L, Li Q, Fan H, ..., Hu X (2021). Pyrolysis of cellulose: evolution of functionalities and structure of bio-char versus temperature. Renew Sustain Energy Rev 135:110416. 10.1016/j.rser.2020.110416
| 36502482 | PMC9741757 | NO-CC CODE | 2022-12-15 23:15:19 | no | Environ Sci Pollut Res Int. 2022 Dec 11;:1-14 | utf-8 | Environ Sci Pollut Res Int | 2,022 | 10.1007/s11356-022-24648-4 | oa_other |
==== Front
Int J Infect Dis
Int J Infect Dis
International Journal of Infectious Diseases
1201-9712
1878-3511
The Author(s). Published by Elsevier Ltd on behalf of International Society for Infectious Diseases.
S1201-9712(22)00646-4
10.1016/j.ijid.2022.12.009
Article
SARS-CoV-2 Specific T-cell and Humoral Immunity in HIV-infected and -uninfected Individuals in an African Population: A Prospective Cohort Study
Ngalamika Owen 12⁎
Lidenge Salum J. 34
Mukasine Marie Claire 2
Kawimbe Musonda 2
Kamanzi Patrick 1
Ngowi John R. 3
Mwaiselage Julius 34
Tso For Yue 5
1 Dermatology and Venereology Division, Department of Medicine, University Teaching Hospital, University of Zambia School of Medicine, Lusaka, Zambia
2 HHV-8 Molecular Virology Laboratory, University Teaching Hospital, Lusaka, Zambia
3 Ocean Road Cancer Institute, Dar-es-Salam, Tanzania
4 Muhimbili University of Health and Allied Sciences, Dar-es-Salam, Tanzania
5 Department of Interdisciplinary Oncology, and the Stanley S Scott Cancer Center, Louisiana State University Health Sciences Center, New Orleans, LA, USA
⁎ Corresponding Author: Owen Ngalamika; University of Zambia; Tell: +260961406928
11 12 2022
11 12 2022
7 6 2022
7 11 2022
6 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.
Objective
: To longitudinally compare SARS-CoV-2-specific T-cell and humoral immune responses between convalescent HIV+ and HIV- individuals.
Methods
: We conducted enzyme-linked Immunospots to determine SARS-CoV-2-specific T-cell responses to spike and NMO (nucleocapsid, membrane protein, and other open reading frame proteins), while an immunofluorescence assay was used to determine the humoral responses. Participants were sampled at baseline and after 8 weeks of follow-up.
Results
: HIV- individuals had significantly more T-cell responses to NMO and spike than HIV+ individuals at baseline, (p=0.026) and (p=0.029) respectively. At follow-up, T-cell responses to NMO and spike in HIV+ individuals increased to levels comparable with HIV- individuals. T-cell responses in the HIV- group significantly decreased from baseline levels at the time of follow-up, spike (p=0.011) and NMO (p=0.014). A significantly higher number of individuals in the HIV+ group had an increase in T-cell responses to spike (p=0.01) and NMO (p=0.026) during the follow-up period compared to the HIV- group. Anti-spike and anti-nucleocapsid antibody titers were high (1:1280) and not significantly different between HIV- and HIV+ individuals at baseline. A significant decrease in anti-nucleocapsid titer was observed in the HIV- (p=0.0001) and the HIV+ (p=0.001) groups at follow-up. SARS-CoV-2 vaccination was more effective in boosting the T-cell than antibody responses shortly after infection.
Conclusion
: There is an impairment of SARS-CoV-2-specific T-cell immunity in HIV-infected individuals with advanced immunosuppression. SARS-CoV-2-specific T-cell immune responses may be delayed in HIV-infected individuals, even in the ones on antiretroviral therapy. There is no difference in SARS-CoV-2-specific humoral immunity between HIV- and HIV+ individuals.
Keywords
SARS-CoV-2
T-cell responses
humoral immunity
HIV
Vaccination
==== Body
pmcIntroduction
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a single-stranded, encapsulated RNA virus that causes coronavirus disease 2019 (COVID-19).(Pal et al., 2020) SARS-CoV-2 has so far caused millions of infections and deaths since its identification in 2019.(Bobrovitz et al., 2021) There have been several waves of infections, with some countries experiencing up to 4 waves and increased transmissibility of mutated virus with each successive wave.(Abdalla et al., 2021, Fisayo and Tsukagoshi, 2021)
Several populations are at a higher risk of severe COVID-19 disease and death due to many comorbidities that may compromise their ability to fight the infection. Among these vulnerable populations are the elderly, individuals with HIV/AIDS who have high HIV viral loads and low CD4 counts, individuals with cancer, and individuals with other comorbidities including diabetes, hypertension, and obesity.(Danwang et al., 2022, Liang et al., 2020, Ng et al., 2021, Peron and Nakaya, 2020) It has been reported by a study conducted in Spain that SARS-CoV-2 seroprevalence was much higher among people living with HIV than the general population.(Berenguer et al., 2021) Also, other studies have reported that HIV-infected individuals who are not on antiretroviral therapy (ART)-naïve are at a higher risk of COVID-19 in-hospital mortality compared to HIV-uninfected individuals and HIV-infected individuals on ART.(Ambrosioni et al., 2021, Davies, 2020, Jassat et al., 2021)
HIV infects CD4+ T lymphocytes and results in a progressive depletion of this cell population and ultimately leads to an impairment of cell-mediated immunity.(Okoye and Picker, 2013) This then increases the risk of opportunistic infections and cancer.(Scanga et al., 2000) There are several subsets of CD4+ T cells, some of which are helper T cells that are required for survival of memory CD8+ T cells during viral infections.(Novy et al., 2007) CD8+ T cells, also known as cytotoxic T lymphocytes, mediate adaptive immunity and are important for killing cancerous cells and viral-infected cells.(Chiang et al., 2013) Importance of T cell immunity in SARS-CoV-2 infection has been reported, where elderly COVID-19 patients above 80 years old have been observed to have diminished CD8 T cell responses which could explain the more frequent severity of COVID-19 in the elderly.(Westmeier et al., 2020) It is therefore apparent that both CD4+ and CD8+ T cell-mediated immunity, in addition to humoral and innate immunity, are a critical part of the host responses against SARS-CoV-2 infection.
Most immunological studies during the COVID-19 pandemic have focused on humoral immunity against SARS-CoV-2,(Lee and Oh, 2021, Zheng et al., 2021) since antibodies are especially important in neutralizing viruses and preventing infection, especially after vaccination.(Jiang et al., 2020, Ju et al., 2020) However, when prevention of infection fails, as in breakthrough infections, T-cell immunity will be key in recognizing and killing infected cells.(Hacisuleyman et al., 2021) Studies on SARS-CoV-2-specific T-cell immunity have been reported, mostly in developed countries, and those conducted in Africa have not focused on the effect of HIV infection.(Bjorkander et al., 2022, Kojima and Klausner, 2022, Le Bert et al., 2020, Riou et al., 2020) In a study conducted in South Africa, it was observed that depletion of CD4+ T cells by HIV infection was associated with suboptimal SARS-CoV-2-specific T-cell and humoral responses, including a decrease in polyfunctionality of SARS-CoV-2-specific T cells.(Riou et al., 2021) A cross-sectional study on SARS-CoV-2-specific humoral and T-cell immune responses done in the United Kingdom reported that at 5-7 months post-infection with SARS-CoV-2, the immune responses were comparable between HIV+ and HIV- individuals, and that the T-cell responses were dominated by CD4+ T cells.(Alrubayyi et al., 2021) In this study, we longitudinally investigated SARS-CoV-2-specific T-cell and humoral immunity in HIV+ individuals who had recovered from COVID-19, and compared their immune response to HIV- individuals who had also recovered from COVID-19.
Methods
Study design and participants
We conducted a prospective cohort study of individuals who had previously been infected with SARS-CoV-2. Study participants were HIV+ and HIV- adult males and females who had been diagnosed with SARS-CoV-2 infection by RT-PCR in the past 12 weeks. We obtained informed consent from the study participants, then collected clinical, sociodemographic, and HIV status information using the study questionnaire, followed by collection of venous whole blood for subsequent laboratory assays. At the time of recruitment, the participants had recovered from COVID-19. Participants were mainly recruited at the outpatient clinics within the University Teaching Hospital (UTH) in Lusaka (Zambia) during their follow-up visits post-hospitalization for COVID-19 or post-infection with SARS-CoV-2. Follow-up of the study participants was done approximately 8 weeks after recruitment. Ethical approval to conduct this study was obtained from the University of Zambia Biomedical Ethics Research Committee (REF. No. 019-017-18) and from the National Health Research Committee (Ref No: NHRA00001/17/09/2021).
Sampling of study participants
We collected 16mls venous whole blood at both baseline and follow-up visits. About 0.1ml of the blood was used for CD4 counting, while 0.3ml was used for T-cell immunophenotyping by flow cytometry, and then the rest of the blood was subjected to centrifugation to separate plasma. About 1ml of the plasma was used for determining HIV viral load in HIV-infected participants while the rest was stored at -80°C for subsequent immunofluorence assay to detect and titer anti-SARS-CoV-2 antibodies. After plasma separation, the remaining cellular component was then subjected to density gradient centrifugation to isolate peripheral blood mononuclear cells (PBMC) which were used for the IFNγ-Enzyme-linked immunospot (IFNγ-ELISpot) assays on the same day of collection.
CD4 counts and HIV viral loads
CD4 counts were determined using the BD TriTest kit (BD biosciences) according to the manufacturer's protocol, on a BD FACSCalibur (BD Biosciences). HIV viral loads were measured using the Aptima HIV-1 Quant Dx Assay kit on a Hologic Panther (Hologic) according to the manufacturer's protocol.
ELISpot peptide pools and assay
The peptides including SARS-CoV-2 S1 scanning pool and SARS-CoV-2 NMO defined peptide pool were obtained from MABTECH AB (Sweden). The SARS-CoV-2 S1 scanning pool contains 166 15-mer peptides which overlap with 11 amino acids and covers the S1 subunit of the spike protein. The SARS-CoV-2 NMO defined peptide pool contains 101 peptides derived from the nucleoprotein (N), membrane protein (M), and open reading frame proteins (ORF) including ORF1, non-structural protein (nsp) 3, ORF-3a, ORF-7a, and ORF8. These peptides are specific for different human leukocyte antigens (HLAs) under both classes 1 and 2.
All the SARS-CoV-2 proteins whose peptides were used in our study have many important functions. The S1 subunit of the spike protein contains the receptor binding domain (RBD) and is important for viral entry into host cells by binding to the host angiotensin-converting enzyme 2 (ACE2) receptor.(Xia, 2021) The nucleocapsid is an RNA-binding protein that is important for viral packaging, and has been observed to be a strong enhancer of virion quality and infectivity.(Mishra et al., 2021) The membrane protein also plays an important role in viral assembly by interacting with all the other structural proteins.(Boson et al., 2021) The nsp3 is the largest SARS-CoV-2 protein that contains a macrodomain that suppresses the host interferon (IFN) response.(Russo et al., 2021) ORF-3a is a protein involved in viral replication, assembly, and release.(Azad and Khan, 2021) ORF-7a and ORF8 are accessory proteins that have been shown to antagonize IFN-1 signalling and therefore impair the host immune response.(Flower et al., 2021, Redondo et al., 2021) Refer to supplementary information for the ELISpot assay.
Immunofluorescence assay
We performed an immunofluorescence assay to detect the presence and titer of antibodies against SARS-CoV-2 spike and nucleocapsid proteins as described previously.(Tso et al., 2021) Transfected HEK-293T cells (ATCC, USA) that expressed either the spike (Addgene, USA) or nucleocapsid (Sini Biological, USA) proteins of SARS-CoV-2 were fixed and seeded onto 12-well polytetrafluoroethylene (PTFE) printed slides (Electron Microscopy Sciences, USA). After a 1:20 dilution in 1X PBS with 0.1% Tween-20, the plasma was added to the corresponding wells on the slide and incubated for 1 hour at 37°C. The slides were then washed and incubated with secondary mouse monoclonal anti-human IgG antibody (ATCC, USA), with subsequent removal of unbound antibodies and incubation with tertiary CY2-conjugated donkey anti-mouse IgG (Jackson Immuno Research Laboratories, USA). This was then followed by counterstaining the cells with Evans blue dye. For the titration, serial dilutions on positive samples were done beginning from 1:20 until a negative reading. Reading of the slides was done by three independent readers on a Nikon E400 fluorescence microscope to determine positive or negative signals, and only harmonious results from at least two independent readers were reported as the outcome.
Flow cytometry
Whole blood was stained for 30 minutes with CD3-APC-Vio770, CD4-PE, CD8-PE-Vio770, CD45RO-APC, and CD197-CCR7-FITC, and the flow cytometry was carried out using a 6-color BD FACSVerse instrument. The antibodies were obtained from Miltenyi Biotec (Germany) and BD Biosciences (Belgium). Fluorescence-minus-one (FMO) controls were used to identify and gate cell populations. Gating strategies used are shown in Supplementary Figure 2. Analysis of the flow data was done using Flow Jo version 10 (TreeStar, Ashland, OR).
Statistical analysis
Descriptive statistics were used to analyze baseline characteristics. Continuous variables are presented as median and interquartile range for skewed data or mean and standard deviation for normally distributed data, while categorical or binary variables are presented as percentages. Comparison of continuous variables between the two groups was done using the Wilcoxon rank-sum test for skewed data, and student t test for normally distributed data. The Shapiro wilk test was used to test for normality. Within-group comparisons of baseline and follow-up values was done using the Wilcoxon matched-pairs signed-rank test. Determination of correlation between continuous variables was done using the Spearman's rank correlation. We also used the Fisher's exact test to determine whether there was any significant difference in binary variables between the two groups. A p value of less than 0.05 was considered statistically significant. Stata version 17 (StataCorp LLC, USA) was used for all statistical analyses. Prism 9 (GraphPad Software) was used to generate the figures.
Results
Baseline characteristics
We recruited 85 study participants who previously had PCR-confirmed SARS-CoV-2 infection. Forty-six (54.1%) of the participants were HIV- and 39 (45.9%) were HIV+. The median time since SARS-CoV-2 diagnosis for the HIV- group was 10.5 weeks in the HIV- group while it was 10 weeks in the HIV+ group. The proportion of hospitalizations was higher among HIV+ than HIV- individuals (p=0.012) (Table 1 ). The proportions of fully vaccinated individuals were not significantly different between the two groups at baseline (p=0.38). The median HIV viral load was undetectable in HIV+ individuals, and most (94.3%) but not all were on ART at baseline. Median CD4 T cell counts were above 450 cells/µl in both groups, but the counts were much higher in HIV-uninfected individuals (705[556-905] vs. 475[359-738]; p=0.001) as expected (Table 1). The rest of the baseline characteristics including age, gender, body mass index, and presence of comorbid conditions were not significantly different between the two groups (Table 1).Table 1 Baseline characteristics of study participants
Table 1 HIV-
(N=46) HIV+
(N=39) p-value
Age (Years) 48.7(±14.7) 53(±10.3) 0.13
Males 18(39.1%) 15(38.5%) 1.00
BMI (kg/m2) 28.9[25-32.7] 27.3[26.2-30.5] 0.91
CD4 count (cells/µl) 705[556-905] 475[359-738] 0.001
HIV viral load (copies/ml) N/A 0[0-0] N/A
On ART N/A 33(94.2%) N/A
Fully vaccinated 23(50%) 15(38.5%) 0.38
Days since Covid19 diagnosis 74[56-77] 70[56-91] 0.49
Hospitalized for COVID19 24(52.2%) 31(79.5%) 0.012
Comorbidities (Diabetes, Hypertension, and Obesity) 35(76.1%) 22(56.4%) 0.07
BMI: Body Mass Index; ART: Antiretroviral Therapy; N/A: Not Applicable. Among HIV-infected individuals, the total number with ART status information was 35 out of 39. All continuous variables are presented as median and interquartile ranges.
Comparison of T-cell responses between HIV+ and HIV- individuals
At both baseline and follow-up, 97%-100% of individuals in both groups had detectable responses (>55SFU/million cells) to spike and/or NMO (nucleoprotein, membrane, other open reading frame proteins) peptide pools. At baseline, HIV- individuals had a significantly higher number of spot-forming units (SFU) towards the S1 subunit of the spike and NMO peptide pools compared to HIV+ individuals (Figure 1 A). At follow-up, there was no significant difference in T-cell responses towards both spike and NMO peptide pools between the two groups (Figure 1B). After 8 weeks of follow-up, there was a significant decrease in T-cell responses towards the NMO peptide pool in the HIV- group while a significant increase towards the NMO peptide pool was observed in the HIV+ group (Figure 1C). Also, a statistically significant decrease in T-cell responses towards the spike peptide pool was observed after 8 weeks of follow-up in the HIV- group, while the T-cell responses in the HIV+ group increased but this was not statistically significant (Figure 1D).Figure 1 SARS-CoV-2- Specific T-cell Responses. a At baseline, HIV- individuals had significantly higher number of T-cell responses towards both spike S1 and NMO peptide pools compared to HIV+ individuals as indicated by the higher number of spot-forming units (SFU) per million cells. b At follow-up, there was an increase in T-cell responses among HIV+ individuals to levels that were not significantly different from that of the HIV- group. c HIV- individuals had a significant decrease in number of T-cell responses towards the NMO peptide pool after 8 weeks of follow-up while HIV+ individuals had a significant increase in number of T-cell responses; d HIV- individuals had a significant decrease in number of T-cell responses towards the S1 subunit of spike after 8 weeks of follow-up, while no significant difference between baseline and follow-up was observed in the HIV+ group.
Figure 1
Thirty-four percent (10/29) of HIV- individuals had an increase in T cell responses against NMO while 71.4% (15/21) of HIV+ individuals had an increase in T cell responses against NMO during the follow-up period. Among the HIV- individuals 31% (14/45) had an increase in anti-spike T cell responses, while 55.3% (21/38) of HIV+ individuals had an increase in spike T cell responses during the follow up period. HIV+ individuals had a statistically higher proportion of individuals that experienced an increase in T cell responses against NMO (p=0.026) and spike (p=0.01) during the follow-up period. Comparison of clinical and demographic characteristics of HIV+ individuals who had an increase in spike and NMO T cell responses to those who had a decrease or no increase in the T cell responses showed that higher CD4 counts were associated with an increase in T cell counts during the follow-up period (Supplementary Table 1).
Comparison of humoral responses between HIV+ and HIV- individuals
Both groups had 97.7% of individuals with detectable antibody responses at baseline, while 100% of the participants in both groups had detectable antibody responses at follow-up. At baseline, there was no significant difference in anti-spike and anti-nucleocapsid antibody titers between the HIV- and HIV+ groups (Figure 2 A). At follow-up, anti-spike antibody titers were not significantly different between the two groups, while anti-nucleocapsid antibody titers were significantly higher in the HIV+ group compared to the HIV- group (Figure 2B). There was a significant decrease in anti-spike antibody titers in the HIV- group after 8 weeks of follow-up, while no significant difference between baseline and follow-up anti-spike antibody titers was observed in the HIV+ group (Figure 2C). There was a significant decrease in anti-nucleocapsid antibody titers in both groups after 8 weeks of follow-up (Figure 2D).Figure 2 SARS-CoV-2-Specific Antibody Responses. a No significant difference in anti-spike and anti-nucleocapsid antibody titers between HIV- and HIV+ groups at baseline. The antibody titers were high in both groups. The median dilutions for both anti-spike and anti-nucleocapsid antibodies in both groups was 1:1280; b At follow-up, the median antibody titer against spike (1:1280) was higher than that against nucleocapsid protein (1:640) in both groups. There were no significant differences in ati-spike antibody titers between the two groups at follow-up, while the anti-nucleocapsid antibody titer was significantly higher in the HIV+ group as indicated by the distribution of more individuals with higher titers compared to the HIV- group; c There was a significant decrease in anti-spike antibody titers in the HIV- group at follow-up, while no significant change between baseline and follow-up titers was observed in the HIV+ group; d There was a statistically significant decrease in anti-nucleocapsid antibody titers after 8 weeks of follow-up in both the HIV- and HIV+ groups. The Titer on Y-axis indicate the number of dilutions.
Figure 2
Correlation of T-cell and humoral responses with CD4 T-cell counts in the HIV+ group
The HIV+ individuals were further categorized into those with CD4 counts below 200 cells/µl and those with CD4 counts ≥200 cells/µl. Individuals with CD4 counts below 200 cells/µl had significantly lower T-cell responses towards the S1 subunit of spike protein compared to those with counts ≥200 cells/µl at baseline (Figure 3 A). Also, at the time of follow-up, the T-cell responses towards both peptide pools were significantly higher among individuals with higher CD4 T cell counts than individuals with low CD4 T cell counts (Figure 3B). There was no significant correlation between CD4 counts and T-cell responses to spike (rho=0.05; p=0.75) and NMO (rho=0.20; 0.38) at baseline. However, there were statistically significant positive correlations between CD4 counts and T-cell responses to spike (rho=0.59; p=0.0001) and NMO (rho=0.51; p=0.001) at follow-up. Anti-spike and anti-nucleocapsid antibody titers were not significantly different between the two groups at both baseline and follow-up (Figures 3C and 3D). There were no significant correlations between CD4 counts and anti-spike or anti-nucleocapsid titers at both baseline and follow-up, (rho=0.12; p=0.46, or rho=0.19; p=0.27) and (rho=0.31; p=0.08, or rho=0.03; p=0.88) respectively.Figure 3 SARS-CoV-2-Specific T-cell and Antibody Responses by CD4 Counts in HIV+ Group. a A sub-analysis by CD4 counts was done in the HIV+ group. There was no significant difference in T-cell responses towards the NMO peptide pool at baseline between individuals with low CD4 counts (<200cells/µl) and those with high CD4 counts (≥200cells/µl). T-cell responses towards spike S1 subunit were significantly higher at baseline in individuals with higher CD4 counts (≥200cells/µl) compared to those with low CD4 counts (<200cells/µl); b At follow-up, T-cell responses towards the NMO and spike peptide pools were significantly more in individuals with CD4 counts ≥200cells/µl than those with lower CD4 counts; c No significant difference in anti-spike and anti-nucleocapsid antibody titers at baseline between HIV+ individuals with high and low CD4 counts; d No significant difference in anti-spike and anti-nucleocapsid antibody titers at follow-up between HIV+ individuals with high and low CD4 counts. At baseline, a few participants did not have their cells stimulated with NMO due to unavailability of the peptide pool at the beginning of the study. Therefore, the number of participants analyzed for the NMO peptide pool is slightly lower than those analyzed for the spike peptide pool at baseline (Figure 3a). Figure 3b shows a slight increase in the total number of individuals with higher CD4 counts and a slight decrease in those with lower CD4 counts at follow up because of an individual who had an improvement in the CD4 count during the follow up period.
Figure 3
Among individuals with low CD4 counts <200cells//µl, the majority (83.3%) had been hospitalized at the time of diagnosis, 16.7% of individuals had breakthrough infections after vaccination, and 33.3% had other comorbid conditions. Only 16.7% of individuals were vaccinated at baseline, while 66.7% of individuals were vaccinated at time of follow up. However, a comparison of spike T cell responses between baseline and follow up was not statistically significant (324SFU[132-408] vs. 270SFU[84-580]; p=0.46).
Effect of vaccination on T-cell and humoral responses to Spike
At baseline, there was no significant difference in the proportion of individuals between the two groups who were vaccinated (52.2% in HIV- and 56.4% in HIV+; p=0.88) or fully vaccinated against SARS-CoV-2 (Table 1 and Supplementary Table 2). Also, some of the participants got vaccinated during the follow-up period. However, there were no significant differences in SARS-CoV-2 vaccination status at baseline and during follow-up between participants in the HIV- and HIV+ groups. The participants all received either of two viral vector COVID-19 vaccines (Johnson & Johnson or Oxford AstraZeneca) which were the only vaccines available in the country at the time of recruitment and follow-up. There was no significant difference in the type of vaccine received between the two groups. However, the HIV- participants had a significantly shorter time from the last SARS-CoV-2 vaccine dose to recruitment into the study compared to the HIV+ participants (p=0.013, Supplementary Table 2). Since the study participants were previously infected with SARS-CoV-2, full vaccination was considered approximately 2 weeks after the single dose of Johnson & Johnson vaccine and at least 3 weeks after the first dose for the Oxford AstraZeneca vaccine (Hall et al., 2022, Mazzoni et al., 2021, Sasikala et al., 2021). Based on vaccination history, only 2/46(4.35%) of individuals in the HIV- group and 3/39(7.7%) of individuals in the HIV+ group had breakthrough infections, which was not significantly different between the two groups (p=0.66). The majority of the participants had no reported or confirmed re-infection during the follow-up period.
At baseline, about 47% of the participants were not vaccinated, approximately 8.2% were partially vaccinated, while 44.7% were fully vaccinated. At follow-up, about 23.5% of the participants were not vaccinated, about 5.8% were partially vaccinated, and about 70.6% were fully vaccinated. Since there were very few partially vaccinated individuals, they were combined with unvaccinated individuals for statistical analysis purposes. Overall, fully vaccinated individuals had significantly higher number of T-cell responses to spike S1 subunit than unvaccinated/partially vaccinated individuals at baseline (Figure 4 A). At follow-up, there was a trend towards higher number of T-cell responses in fully vaccinated compared to unvaccinated/partially vaccinated individuals, but this was not statistically significantly different possibly due to insufficient number of individuals in the unvaccinated/partially vaccinated group to make statistical comparisons (Figure 4B). There were high numbers of T-cell responses to spike in fully vaccinated individuals than unvaccinated/partially vaccinated individuals at baseline and follow up in the HIV- group, although this was not statistically significant, possibly because the study was not powered enough to address this subgroup analysis (Table 2 ). While T-cell responses in fully vaccinated HIV+ individuals at baseline were not significantly different from those in unvaccinated/partially vaccinated HIV+ individuals, these responses were significantly higher in fully vaccinated individuals at the time of follow-up (Table 2).Figure 4 Effect of Vaccination on T-cell and Antibody Responses. a Fully vaccinated individuals had significantly higher number of spike-specific T-cell responses than unvaccinated/partially vaccinated individuals at baseline; b Fully vaccinated individuals had marginally significant higher number of spike-specific T-cell responses that unvaccinated/partially vaccinated individuals at follow-up; c No significant difference in anti-spike antibody titers between fully vaccinated and unvaccinated/partially vaccinated individuals at baseline; d No significant difference in anti-spike antibody titers between fully vaccinated and unvaccinated/partially vaccinated individuals at follow-up.
Figure 4
Table 2 SARS-CoV-2 Specific T Cell and Antibody Responses to Spike in HIV- and HIV+ Individuals by Vaccination Status
Table 2 HIV- HIV+
Fully
Vaccinated Unvaccinated/
Partially vaccinated p value Fully
Vaccinated Unvaccinated/
Partially vaccinated p value
T Cell Responses
Baseline 1344 SFU
[412-2100] 772 SFU
[408-1264] 0.084 820 SFU
[440-1736] 406 SFU
[170-1248] 0.075
Follow-up 718 SFU
[344-1190] 548 SFU
[212-1132] 0.58 1000 SFU
[536-1684] 604 SFU
[124-908] 0.015
Antibody Responses (Titers)
Baseline 1:2560
[1:1280-1:10240] 1:1280
[1:320-1:2560] 0.01 1:1280
[1:1280-1:5120 1:3840
[1:1280-1:5120] 0.87
Follow-up 1:1280
[1:320-1:2560] 1:1280
[1:320-1:1280] 0.69 1:1280
[1:1280-1:2560] 1:1280
[1:1280-1:5120] 0.84
SFU: Spot-forming units
There were no statistically significant differences in median anti-spike antibody titers at both baseline and follow-up between fully vaccinated and unvaccinated/partially vaccinated individuals (Figures 4C and 4D). However, when a subgroup analysis was done, fully vaccinated HIV- individuals had higher anti-spike antibody titers than unvaccinated/partially vaccinated HIV- individuals at baseline, while no significant difference in anti-spike antibody titers was observed between fully vaccinated and unvaccinated/partially vaccinated HIV+ individuals (Table 2). At time of follow-up, there were no significant differences in anti-spike antibody titers between fully vaccinated and unvaccinated/partially vaccinated individuals in both HIV- and HIV+ groups (Table 2).
Changes in CD4 Count, HIV Viral Load, and COVID-19 Vaccination Status During Follow-up
There were no statistically significant differences between baseline and follow-up CD4 counts and HIV viral loads among the HIV+ individuals, (475 cells/µl [359-738] vs. 550 cells/µl [346-762]; p=0.20) and (0 copies/ml [0-0] vs. 0 copies/ml [0-0]; p=0.77) respectively. The full vaccination status in the HIV- group increased from 50% at baseline to 71.7% at time of follow-up, while the full vaccination status in the HIV+ group increased from 38.5% at baseline to 69.2% at time of follow-up. There were no significant differences in vaccination status between the HIV+ and HIV- groups at both baseline (p=0.38) and at time of follow-up (p=0.82).
CD4 and CD8 T-cell Immunophenotypes
To determine whether HIV infection affects the proportions of circulating T-cell subsets in COVID-19 convalescent individuals, we compared proportions of the naïve, central memory, effector memory, and effector CD4+ and CD8+ T cells between HIV+ and HIV- individuals, and the changes between the groups after 8 weeks of follow-up. We observed that the proportions of T cell subsets were not significantly different at baseline and at time of follow-up between the HIV+ and HIV- groups. However, there was a significant change in the proportion of the subsets within the groups at follow-up. For CD4+ T cells, the proportion of naïve cells in the whole blood of both HIV+ and HIV- groups significantly increased from baseline levels after 8 weeks of follow-up, (26.9%[16.8-32.4]vs. 29.8%18.3-41.2]; p=0.0001) and (24.4%[14.8-33.9] vs. 28.7%[21.6-38.1]; p=0.0001) respectively (Supplementary Table 3). On the other hand, the proportion of CD4 effector memory cells significantly reduced from baseline levels after 8 weeks of follow-up in both HIV+ and HIV- groups, (24.2%[11.8-34.9] vs. 15%[10.5-22.6]; p=0.001) and (17.8%[12.3-38.8] vs. 15.9%[10.8-20.9]; p=0.002) respectively. In addition, there was a significant increase in proportion of CD4 effector cells in the HIV+ group at follow-up (0.5%[0.1-1.3] vs. 0.7%[0.5-1.9]; p=0.015), but not in the HIV- group. For CD8+ T cells, the proportion of central memory cells significantly increased from baseline levels in both HIV+ and HIV- groups, (2%[1.5-4.6] vs. 5.3%[2.8-79.5]; p=0.0001) and (2.7%[1.2-4.5] vs. 5.6%[2.9-13.5]; p=0.0001) respectively (Supplementary Table 4). The proportion of effector cells significantly reduced in both HIV+ and HIV- groups, (37.8%[17.8-55.9] vs. 30.5%[2.3-50.1]; p=0.003) and (38.3%[27.3-58] vs. 31.3%[13.1-43.5]; p=0.001) respectively, while the proportion of effector memory cells significantly reduced in the HIV+ group (11.5%[6.9-18.4] vs. 7.4%[4.3-13.4]; p=0.01) with a statistically insignificant reduction in the HIV- group. We also compared the proportion of CD4 (25.1%[12.7-33.9] vs. 24.2%[9.5-54.8]; p=0.76) and CD8 (12.5%[6.9-18.2] vs. 8%[7.7-28]; p=0.96) effector memory T cells at baseline between HIV+ individuals with high CD4 counts to those with lower CD4 counts, and observed no significant difference in the proportions which could be due to the low sample size for the sub-analysis.
Discussion
Cell-mediated and humoral immunity are two major aspects of adaptive immunity that play a critical role against SARS-CoV-2 infection. Hence, our study focused on SARS-CoV-2 specific T-cell and antibody responses in HIV+ versus HIV- individuals in Africa, who were previously infected with SARS-CoV-2 and had recovered from COVID-19. We also looked at how this immunity changed over an 8-week follow-up period.
We observed that at 10 weeks post-infection with SARS-CoV-2, SARS-CoV-2-specific T-cell responses are present in both HIV+ and HIV- individuals. However, they are significantly greater in the HIV- than the HIV+ individuals. There was a significantly shorter time from last SARS-CoV-2 vaccine dose to recruitment in HIV- compared to HIV+ participants, possibly explaining the observed differential T-cell responses to spike. Nevertheless, the T cell responses to NMO were also significantly higher in the HIV- group at baseline, and only about half of the entire study population was vaccinated at baseline. This observation changed after an eight-week follow-up period, where the SARS-CoV-2-specific T-cell responses in the HIV+ group significantly improved to match the levels observed in the HIV- group. It is possible that, this improved response in the HIV+ group over the follow-up period could be due to ongoing/prolonged asymptomatic infection and/or improved vaccination status over time. Our findings are similar to a previous study where an observation was made that both humoral and T-cell responses are comparable in HIV-uninfected individuals and HIV-infected individuals who are on ART and have suppressed HIV viral loads.(Alrubayyi et al., 2021) However, our data suggests that HIV+ individuals mount a less robust SARS-CoV-2-specific T-cell response both at baseline and follow-up, but that these can be boosted by vaccination.
It is interesting to note that the T-cell responses in our cohort of HIV+ and HIV- individuals were still present at almost 4.5 months post the initial infection, since our baseline analysis was at about 10 weeks after infection. These responses increased in the HIV+ group to levels comparable with that of the HIV- group during the 8-week follow-up period. This is similar to a previous study by Zuo et al. where it was observed that functional SARS-CoV-2 specific T-cell responses were maintained at 6 months following primary infection (Zuo et al., 2021). In another study, it was observed that adaptive cellular immunity was durable at 7 months after primary infection (Gurevich et al., 2021). A study by Lu et al. observed that the T-cell responses were still detectable after a longer period of 12 months after SARS-CoV-2 infection, and that individuals with severe illness had a higher frequency of SARS-CoV-2-specific T cells and antibodies (Lu et al., 2021). In our study, attempts to categorize immune responses based on the initial COVID-19 severity was not done as some participants had missing clinical information for the time they were acutely infected because the study recruited participants after hospital discharge or recovery from acute infection. On the other hand, hospital admission which may be associated with disease severity, was not associated with differences in the immune responses between the two groups. The slow development of T cell responses in HIV+ individuals over time could also be due to more prolonged asymptomatic infection or re-infection during the follow up period. Nevertheless, based on our studies and that of others, T-cell and humoral immunity appears to be present for at least several months after primary infection in both HIV+ and HIV- individuals. However, at least in the HIV- population, these responses seem to weaken over time, necessitating the need for SARS-CoV-2 vaccination and/or boosters to prevent re-infections and COVID-19 disease. Our study is unique from the other studies in that we investigated T-cell and humoral responses in people living with HIV, and how these responses change over time in comparison to HIV- individuals.
We observed that humoral immune responses against the two SARS-CoV-2 proteins (spike and nucleocapsid) were detectable in most participants in both HIV- and HIV+ groups at baseline and follow-up. The antibody titers were high and not significantly different between the two groups at baseline. However, at follow-up, the anti-nucleocapsid antibody titers were significantly higher in the HIV+ group than the HIV- group. Our observation is similar to other studies that have reported that SARS-CoV-2 seroconversion is similar between people living with HIV who are on ART and HIV-uninfected individuals (Alrubayyi et al., 2021, Yamamoto et al., 2021). After 8 weeks of follow-up, the antibody responses were still detectable in all the participants including those that were undetectable at baseline. The one individual who had undetectable levels of anti-SARS-CoV-2 antibody at baseline and detectable titers at follow up may represent a case of some individuals with weak/delayed initial responses, vaccine response, or reinfection during the study period. However, there was a significant decrease in titers against both spike and nucleocapsid proteins in both groups, except against spike in the HIV+ group. Other studies have also reported that antibody titers against SARS-CoV-2 decrease after a few months but are still detectable several months after the infection (Yamayoshi et al., 2021). Our findings and that of others indicate that SARS-CoV-2 vaccination or boosters may be very important to maintain the immunity and prevent re-infections.
Most of our HIV+ study participants had suppressed and/or undetectable HIV viral loads at baseline, and 33 (94.2%) were on ART. Despite this, we observed that HIV+ individuals with CD4 counts below 200 cells/µl had significantly weaker T-cell responses than those with higher CD4 counts at both baseline and after 8 weeks of follow-up. This is in line with other studies that have reported that severe immunosuppression, with CD4 counts below 200 cells/µl, is associated with severe COVID19 disease (Hoffmann et al., 2021). On the other hand, no significant difference was observed in antibody titers against spike and nucleocapsid between HIV-infected individuals with higher CD4 and low CD4 counts. Our observation may be due to insufficient numbers to make comparisons for this sub-group analysis or may be due to a better humoral than T-cell immunity against SARS-CoV-2 in severely immunosuppressed HIV infected individuals. Furthermore, the higher proportion of hospitalizations in the HIV+ than the HIV- populations may be due to our observed less robust initial T-cell responses to SARS-CoV-2 in the HIV+ population.
Fully vaccinated individuals in the entire combined cohort compared to unvaccinated and partially vaccinated individuals had significantly higher number of T-cell responses to spike at baseline, and marginally significantly higher number of T-cell responses at follow-up as expected due to the vaccine stimulatory effect. Subgroup analyses by HIV status showed that the T-cell responses against spike were much higher in the fully vaccinated individuals at both time points, with responses in the fully vaccinated HIV+ group being significantly higher at time of follow-up as the number of fully vaccinated participants increased over time. There was no significant difference in anti-spike antibody titers between fully vaccinated and unvaccinated/partially vaccinated individuals at both baseline and follow-up. This could be due to durable responses provided by SARS-CoV-2 primary infection as all these individuals were previously infected by SARS-CoV-2. However, on subgroup analysis, fully vaccinated individuals among the HIV- group had significantly higher anti-spike antibody titers than unvaccinated/partially vaccinated individuals. Again, this could be due to the marginal boosting effect of SARS-CoV-2 vaccine, although both groups had high titers at baseline. Our observations are similar to previous studies that have observed that people living with HIV develop high anti-spike antibody titers which are similar to individuals without HIV infection (Ruddy et al., 2021). Our findings may suggest that vaccination in individuals recently infected with SARS-CoV-2 may be more beneficial in improving T-cell immunity. The antibody titers were very high in both groups after infection with SARS-CoV-2 reaching several thousand-fold-dilution, and hence vaccination did not seem to have a profound effect on the titers which may have affected our comparisons. However, considering the gradual decrease in antibody titers including T-cell immunity over time as observed in our study and by others (Mazzoni et al., 2021), vaccination to boost immunity is absolutely necessary to prevent re-infections and severe COVID-19 in both HIV+ and HIV- individuals. Also, as noted from our study, the higher anti-spike antibody titers at follow-up compared to anti-nucleocapsid antibody titers may be explained by the effect of vaccination as the majority of the participants in both groups were vaccinated at follow-up.
We observed significant changes in circulating proportions of CD4+ and CD8+ T-cell phenotypes after 8 weeks of follow-up. During this period, the proportions of naïve CD4+ T cells increased from baseline levels, indicating a normalization of circulating CD4+ T-cell subsets after recovery from COVID-19 disease (Hanna et al., 2021, Kalpakci et al., 2020). This increase in naïve CD4+ T cells was accompanied by an observed decrease in effector memory cells over time, reflecting a recovery of CD4+ T-cell subsets towards normal after infection (Bordoni et al., 2019). The proportion of CD4+ effector T cells in HIV+ individuals was observed to significantly increase at follow-up compared to baseline levels, but the proportions were very low compared to other subsets. This could be associated with the observed delayed SARS-CoV-2-specific T-cell responses observed in the HIV+ group.
The proportion of CD8+ central memory T cells significantly increased while CD8+ effector T cells significantly reduced in both HIV+ and HIV- individuals after 8 weeks of follow-up. These findings are in agreement with studies that have shown that during acute viral infection, naïve CD8+ T cells differentiate into effector cells which kill target cells.(Cui and Kaech, 2010) Upon clearance of the infection, a proportion of these cells differentiate into memory cells that protect the host during re-infection.(Rha and Shin, 2021) Hence both CD4+ and CD8+ T-cell subsets may be useful in the monitoring of disease progression and recovery from COVID-19 disease in both HIV+ and HIV- individuals.
Study limitations
The overall small sample size might have limited our ability to elicit some of the differences that might exist between subgroups. As immune responses are impacted by vaccination status, timing, types of vaccine used, the lack of a standardized cohort with a uniform vaccination profile might have confounded some of our findings. However, we observed no difference in type of vaccine received by both groups, and the proportion of vaccinated individuals. Our assessment of immune responses may also be confounded by the lack of data on infection status of the participants during the follow up period as some responses could be due to ongoing/prolonged SARS-CoV-2 infection, or asymptomatic re-infection during the study period. Although most of the participants received SARS-CoV-2 vaccination during follow up, we cannot rule out the role of re-infection in the observed responses. Another study limitation is on the lack of clinical detailed information on severity of COVID-19 disease and SARS-CoV-2 viral load where it could have been useful to compare immune responses in different COVID-19 stages and viral burden. However, our study provides important findings on presence of SARS-CoV-2-specific T-cell and antibody responses in both HIV+ and in HIV- individuals approximately 4.5 months after recovery from acute infection.
Conclusion
SARS-CoV-2-specific T-cell and antibody responses are present in both HIV+ and in HIV- individuals after recovery from acute infection. These responses are generally high and still detectable 4.5 months after initial infection. HIV+ participants have less robust T-cell responses both at baseline and follow-up, but these are boosted by vaccination. Severe immunosuppression among HIV-infected individuals mostly affects T-cell and not antibody responses. SARS-CoV-2-specific T-cell responses may be delayed in HIV-infected individuals despite them being on antiretroviral therapy. T-cell responses in the HIV- individuals and antibody responses in both HIV- and HIV+ individuals decrease over time, which could increase the risk of re-infection and severe disease, and necessitates vaccine boosters.
Funding
Research reported in this publication was supported by the Fogarty International Center and National Cancer Institute of the National Institutes of Health under Award Numbers K43TW011095 and U54CA221204 to ON and K43TW011418 to SJL. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Data availability
All data generated or analysed during this study are available from the corresponding author on reasonable request.
Contributors
O.N.: performed and supervised the laboratory experiments on T-cell work, drafted the manuscript, performed statistical analyses, wrote, and edited the manuscript. S.L.: performed and supervised the laboratory experiments on humoral responses work, drafted the manuscript, performed statistical analyses, wrote, and edited the manuscript, J.R.N., J.M. and F.Y.T.: performed laboratory experiments on humoral response. M.C.M.: performed laboratory experiments on T-cell functional assays. M.K.: performed laboratory experiments on flow cytometry. P.K.: recruited, followed up the study participants and collected specimens. All authors reviewed and edited the manuscript.
Declaration of Interests
The authors declare no competing interests.
Acknowledgments
We thank all the participants for consenting to participate in this study.
References
Abdalla S, Bakhshwin D, Shirbeeny W, Bakhshwin A, Bahabri F, Bakhshwin A, et al. Successive waves of COVID 19: confinement effects on virus-prevalence with a mathematical model. Eur J Med Res 2021;26(1):128.
Alrubayyi A, Gea-Mallorqui E, Touizer E, Hameiri-Bowen D, Kopycinski J, Charlton B, et al. Characterization of humoral and SARS-CoV-2 specific T cell responses in people living with HIV. Nat Commun 2021;12(1):5839.
Ambrosioni J, Blanco JL, Reyes-Uruena JM, Davies MA, Sued O, Marcos MA, et al. Overview of SARS-CoV-2 infection in adults living with HIV. Lancet HIV 2021;8(5):e294-e305.
Azad GK, Khan PK. Variations in Orf3a protein of SARS-CoV-2 alter its structure and function. Biochem Biophys Rep 2021;26:100933.
Berenguer J, Diez C, Martin-Vicente M, Mican R, Perez-Elias MJ, Garcia-Fraile LJ, et al. Prevalence and factors associated with SARS-CoV-2 seropositivity in the Spanish HIV Research Network Cohort. Clin Microbiol Infect 2021;27(11):1678-84.
Bjorkander S, Du L, Zuo F, Ekstrom S, Wang Y, Wan H, et al. SARS-CoV-2-specific B- and T-cell immunity in a population-based study of young Swedish adults. J Allergy Clin Immunol 2022;149(1):65-75 e8.
Bobrovitz N, Arora RK, Cao C, Boucher E, Liu M, Donnici C, et al. Global seroprevalence of SARS-CoV-2 antibodies: A systematic review and meta-analysis. PLoS One 2021;16(6):e0252617.
Bordoni V, Brando B, Piselli P, Forini O, Perna FE, Atripaldi U, et al. Naive/Effector CD4 T cell ratio as a useful predictive marker of immune reconstitution in late presenter HIV patients: A multicenter study. PLoS One 2019;14(12):e0225415.
Boson B, Legros V, Zhou B, Siret E, Mathieu C, Cosset FL, et al. The SARS-CoV-2 envelope and membrane proteins modulate maturation and retention of the spike protein, allowing assembly of virus-like particles. J Biol Chem 2021;296:100111.
Chiang SC, Theorell J, Entesarian M, Meeths M, Mastafa M, Al-Herz W, et al. Comparison of primary human cytotoxic T-cell and natural killer cell responses reveal similar molecular requirements for lytic granule exocytosis but differences in cytokine production. Blood 2013;121(8):1345-56.
Cui W, Kaech SM. Generation of effector CD8+ T cells and their conversion to memory T cells. Immunol Rev 2010;236:151-66.
Danwang C, Noubiap JJ, Robert A, Yombi JC. Outcomes of patients with HIV and COVID-19 co-infection: a systematic review and meta-analysis. AIDS Res Ther 2022;19(1):3.
Davies MA. HIV and risk of COVID-19 death: a population cohort study from the Western Cape Province, South Africa. medRxiv 2020.
Fisayo T, Tsukagoshi S. Three waves of the COVID-19 pandemic. Postgrad Med J 2021;97(1147):332.
Flower TG, Buffalo CZ, Hooy RM, Allaire M, Ren X, Hurley JH. Structure of SARS-CoV-2 ORF8, a rapidly evolving immune evasion protein. Proc Natl Acad Sci U S A 2021;118(2).
Gurevich M, Zilkha-Falb R, Sonis P, Magalashvili D, Menascu S, Flechter S, et al. SARS-COV-2 Memory B and T Cells Profile in Mild COVID-19 Convalescents subjects. Int J Infect Dis 2021.
Hacisuleyman E, Hale C, Saito Y, Blachere NE, Bergh M, Conlon EG, et al. Vaccine Breakthrough Infections with SARS-CoV-2 Variants. N Engl J Med 2021;384(23):2212-8.
Hall V, Foulkes S, Insalata F, Kirwan P, Saei A, Atti A, et al. Protection against SARS-CoV-2 after Covid-19 Vaccination and Previous Infection. N Engl J Med 2022;386(13):1207-20.
Hanna SJ, Codd AS, Gea-Mallorqui E, Scourfield DO, Richter FC, Ladell K, et al. T cell phenotypes in COVID-19 - a living review. Oxf Open Immunol 2021;2(1):iqaa007.
Hoffmann C, Casado JL, Harter G, Vizcarra P, Moreno A, Cattaneo D, et al. Immune deficiency is a risk factor for severe COVID-19 in people living with HIV. HIV Med 2021;22(5):372-8.
Jassat W, Cohen C, Tempia S, Masha M, Goldstein S, Kufa T, et al. Risk factors for COVID-19-related in-hospital mortality in a high HIV and tuberculosis prevalence setting in South Africa: a cohort study. Lancet HIV 2021;8(9):e554-e67.
Jiang S, Hillyer C, Du L. Neutralizing Antibodies against SARS-CoV-2 and Other Human Coronaviruses. Trends Immunol 2020;41(5):355-9.
Ju B, Zhang Q, Ge J, Wang R, Sun J, Ge X, et al. Human neutralizing antibodies elicited by SARS-CoV-2 infection. Nature 2020;584(7819):115-9.
Kalpakci Y, Hacibekiroglu T, Trak G, Karacaer C, Demirci T, Kocayigit H, et al. Comparative evaluation of memory T cells in COVID-19 patients and the predictive role of CD4+CD8+ double positive T lymphocytes as a new marker. Rev Assoc Med Bras (1992) 2020;66(12):1666-72.
Kojima N, Klausner JD. Protective immunity after recovery from SARS-CoV-2 infection. Lancet Infect Dis 2022;22(1):12-4.
Le Bert N, Tan AT, Kunasegaran K, Tham CYL, Hafezi M, Chia A, et al. SARS-CoV-2-specific T cell immunity in cases of COVID-19 and SARS, and uninfected controls. Nature 2020;584(7821):457-62.
Lee E, Oh JE. Humoral Immunity against SARS-CoV-2 and the Impact on COVID-19 Pathogenesis. Mol Cells 2021;44(6):392-400.
Liang W, Guan W, Chen R, Wang W, Li J, Xu K, et al. Cancer patients in SARS-CoV-2 infection: a nationwide analysis in China. Lancet Oncol 2020;21(3):335-7.
Lu Z, Laing ED, Pena DaMata J, Pohida K, Tso MS, Samuels EC, et al. Durability of SARS-CoV-2-Specific T-Cell Responses at 12 Months Postinfection. J Infect Dis 2021;224(12):2010-9.
Mazzoni A, Di Lauria N, Maggi L, Salvati L, Vanni A, Capone M, et al. First-dose mRNA vaccination is sufficient to reactivate immunological memory to SARS-CoV-2 in subjects who have recovered from COVID-19. J Clin Invest 2021;131(12).
Mishra T, Sreepadmanabh M, Ramdas P, Sahu AK, Kumar A, Chande A. SARS CoV-2 Nucleoprotein Enhances the Infectivity of Lentiviral Spike Particles. Front Cell Infect Microbiol 2021;11:663688.
Ng WH, Tipih T, Makoah NA, Vermeulen JG, Goedhals D, Sempa JB, et al. Comorbidities in SARS-CoV-2 Patients: a Systematic Review and Meta-Analysis. mBio 2021;12(1).
Novy P, Quigley M, Huang X, Yang Y. CD4 T cells are required for CD8 T cell survival during both primary and memory recall responses. J Immunol 2007;179(12):8243-51.
Okoye AA, Picker LJ. CD4(+) T-cell depletion in HIV infection: mechanisms of immunological failure. Immunol Rev 2013;254(1):54-64.
Pal M, Berhanu G, Desalegn C, Kandi V. Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2): An Update. Cureus 2020;12(3):e7423.
Peron JPS, Nakaya H. Susceptibility of the Elderly to SARS-CoV-2 Infection: ACE-2 Overexpression, Shedding, and Antibody-dependent Enhancement (ADE). Clinics (Sao Paulo) 2020;75:e1912.
Redondo N, Zaldivar-Lopez S, Garrido JJ, Montoya M. SARS-CoV-2 Accessory Proteins in Viral Pathogenesis: Knowns and Unknowns. Front Immunol 2021;12:708264.
Rha MS, Shin EC. Activation or exhaustion of CD8(+) T cells in patients with COVID-19. Cell Mol Immunol 2021;18(10):2325-33.
Riou C, du Bruyn E, Stek C, Daroowala R, Goliath RT, Abrahams F, et al. Relationship of SARS-CoV-2-specific CD4 response to COVID-19 severity and impact of HIV-1 and tuberculosis coinfection. J Clin Invest 2021;131(12).
Riou C, Schafer G, du Bruyn E, Goliath RT, Stek C, Mou H, et al. Rapid, simplified whole blood-based multiparameter assay to quantify and phenotype SARS-CoV-2 specific T cells. medRxiv 2020.
Ruddy JA, Boyarsky BJ, Bailey JR, Karaba AH, Garonzik-Wang JM, Segev DL, et al. Safety and antibody response to two-dose SARS-CoV-2 messenger RNA vaccination in persons with HIV. AIDS 2021;35(14):2399-401.
Russo LC, Tomasin R, Matos IA, Manucci AC, Sowa ST, Dale K, et al. The SARS-CoV-2 Nsp3 macrodomain reverses PARP9/DTX3L-dependent ADP-ribosylation induced by interferon signaling. J Biol Chem 2021;297(3):101041.
Sasikala M, Shashidhar J, Deepika G, Ravikanth V, Krishna VV, Sadhana Y, et al. Immunological memory and neutralizing activity to a single dose of COVID-19 vaccine in previously infected individuals. Int J Infect Dis 2021;108:183-6.
Scanga CA, Mohan VP, Yu K, Joseph H, Tanaka K, Chan J, et al. Depletion of CD4(+) T cells causes reactivation of murine persistent tuberculosis despite continued expression of interferon gamma and nitric oxide synthase 2. J Exp Med 2000;192(3):347-58.
Tso FY, Lidenge SJ, Pena PB, Clegg AA, Ngowi JR, Mwaiselage J, et al. High prevalence of pre-existing serological cross-reactivity against severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) in sub-Saharan Africa. Int J Infect Dis 2021;102:577-83.
Westmeier J, Paniskaki K, Karakose Z, Werner T, Sutter K, Dolff S, et al. Impaired Cytotoxic CD8(+) T Cell Response in Elderly COVID-19 Patients. mBio 2020;11(5).
Xia X. Domains and Functions of Spike Protein in Sars-Cov-2 in the Context of Vaccine Design. Viruses 2021;13(1).
Yamamoto S, Saito M, Nagai E, Toriuchi K, Nagai H, Yotsuyanagi H, et al. Antibody response to SARS-CoV-2 in people living with HIV. J Microbiol Immunol Infect 2021;54(1):144-6.
Yamayoshi S, Yasuhara A, Ito M, Akasaka O, Nakamura M, Nakachi I, et al. Antibody titers against SARS-CoV-2 decline, but do not disappear for several months. EClinicalMedicine 2021;32:100734.
Zheng J, Deng Y, Zhao Z, Mao B, Lu M, Lin Y, et al. Characterization of SARS-CoV-2-specific humoral immunity and its potential applications and therapeutic prospects. Cell Mol Immunol 2021.
Zuo J, Dowell AC, Pearce H, Verma K, Long HM, Begum J, et al. Robust SARS-CoV-2-specific T cell immunity is maintained at 6 months following primary infection. Nat Immunol 2021;22(5):620-6.
Appendix Supplementary materials
Image, application 1
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ijid.2022.12.009.
| 36516914 | PMC9741763 | NO-CC CODE | 2022-12-14 23:29:57 | no | Int J Infect Dis. 2022 Dec 11; doi: 10.1016/j.ijid.2022.12.009 | utf-8 | Int J Infect Dis | 2,022 | 10.1016/j.ijid.2022.12.009 | oa_other |
==== Front
Disabil Health J
Disabil Health J
Disability and Health Journal
1936-6574
1876-7583
Elsevier Inc.
S1936-6574(22)00191-1
10.1016/j.dhjo.2022.101428
101428
Brief Report
Challenges experienced by U.S. K-12 public schools in serving students with special education needs or underlying health conditions during the COVID-19 pandemic and strategies for improved accessibility
Spencer Patricia PhD 1∗#
Timpe Zach PhD 2
Verlenden Jorge PhD 3
Rasberry Catherine N. PhD 3
Moore Shamia MPH 1
Yeargin-Allsopp Marshalyn MD 4
Claussen Angelika H. PhD 4
Lee Sarah PhD 5
Murray Colleen DrPH, MPH 2
Tripathi Tasneem DrPH, MPH 2
Conklin Sarah PhD 2
Iachan Ronaldo PhD 2
McConnell Luke MS 2
Deng Xiaoyi MS 2
Pampati Sanjana MPH 3
1 Oak Ridge Associated Universities
2 ICF
3 CDC Division of Adolescent and School Health, National Center for HIV, Viral Hepatitis, STD, and TB Prevention
4 CDC Division of Human Development and Disability, National Center on Birth Defects and Developmental Disabilities
5 CDC Division of Population Health, National Center for Chronic Disease Prevention and Health Promotion
∗ Corresponding author. , PhD, Division of Adolescent and School Health, Centers for Disease Control and Prevention, 1600 Clifton Road, MS US8-1, Atlanta, GA 30329. Tel: 512-5866747;
# Present address: Division of Reproductive Health, Centers for Disease Control and Prevention, 4770 Buford Highway, Chamblee, GA 30341.
11 12 2022
11 12 2022
10142814 9 2022
29 11 2022
6 12 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.
Background
Students with special education needs or underlying health conditions have been disproportionately impacted (e.g., by reduced access to services) throughout the COVID-19 pandemic.
Objective
This study describes challenges reported by schools in providing services and supports to students with special education needs or underlying health conditions and describes schools’ use of accessible communication strategies for COVID-19 prevention.
Methods
This study analyzes survey data from a nationally representative sample of U.S. K-12 public schools (n = 420, February-March 2022). Weighted prevalence estimates of challenges in serving students with special education needs or underlying health conditions and use of accessible communication strategies are presented. Differences by school locale (city/suburb vs. town/rural) are examined using chi-square tests.
Results
The two most frequently reported school-based challenges were staff shortages (51.3%) and student compliance with prevention strategies (32.4%), and the two most frequently reported home-based challenges were the lack of learning partners at home (25.5%) and lack of digital literacy among students’ families (21.4%). A minority of schools reported using accessible communications strategies for COVID-19 prevention efforts, such as low-literacy materials (7.3%) and transcripts that accompany podcasts or videos (6.7%). Town/rural schools were more likely to report non-existent or insufficient access to the internet at home and less likely to report use of certain accessible communication than city/suburb schools.
Conclusion
Schools might need additional supports to address challenges in serving students with special education needs or with underlying health conditions and improve use of accessible communication strategies for COVID-19 and other infectious disease prevention.
Keywords
COVID-19
disabilities
K-12
accessibility
challenges
==== Body
pmcFunding source:
This study is funded, in part, by task order 75D30121F10577 from the Centers for Disease Control and Prevention to ICF.
Introduction
The COVID-19 pandemic impacted the lives of students through disruptions to typical in-person school and household routines. Approximately 15% of public school students received special education services in 2020-2021 and over 40% of school-children and adolescents have at least one chronic health condition, such as asthma or obesity.1 , 2 The Individuals with Disabilities Education Act (IDEA) in K-12 public schools addresses nine categories of disabilities, with specific learning disabilities, speech or language impairment, and other health impairments (e.g., limited strength, vitality, or alertness due to chronic health problems such as a heart condition or epilepsy) being the most prevalent.1 As a result of the COVID-19 pandemic, students with disabilities or with underlying health conditions have experienced numerous challenges, including reduced or loss of services and supports, issues with access to remote learning, and lack of home-based support.3, 4, 5 Although virtual learning afforded students with health needs and disabilities the continuity of an education during the pandemic, the quality of virtual supports may have been reduced by lack of resources and insufficient training and technology access.6 Through lessons learned from the COVID-19 pandemic, school districts have an opportunity to improve virtual supports by investing time and resources into planning, training special education staff in the delivery of virtual based supports, and enhancing technology access, to benefit these populations now and during future public health emergencies.
In August 2022, the U.S. Centers for Disease Control and Prevention (CDC) released updated guidance for schools to promote safe, in-person instruction and recommended COVID-19 prevention strategies based on local context.7 This guidance included special considerations for people at risk of severe illness from COVID-19. Students with certain disabilities or underlying health conditions might be at increased risk of severe COVID-19 illness or might experience unique challenges to using prevention strategies. For example, children with autism or hearing loss might have difficulty adhering to mask requirements, and children with intellectual and developmental disabilities and children with medical complexity are at increased risk of severe COVID-19 outcomes.8 , 9 Special considerations apply to these populations to promote safe inclusion in schools and ensure continuity of education and services. Such considerations emphasize modified strategies to meet the needs of students receiving special education services and protections for students at risk of severe COVID-19 illness or with family members with similar risks.
Guidance also underscores the importance of communicating about COVID-19 prevention using accessible communication strategies.7 For example, schools have been encouraged to share information about vaccines tailored to those with limited English proficiency (LEP) and those with disabilities that could require modified formats to support accessibility (e.g., print or electronic material written in plain language and translated into multiple languages; materials accessible to screen readers). However, little is known about the challenges schools experienced during the pandemic in serving students with disabilities or with underlying health conditions or schools’ uptake of accessible communication strategies in their COVID-19 prevention efforts.
Accordingly, using data from a nationally representative sample of U.S. K-12 public schools, this study aims to describe 1) challenges in serving students in school or home settings with disabilities or underlying health conditions; and 2) use of accessible communication strategies to convey COVID-19 prevention efforts.
Methods
Data
CDC initiated the National School COVID-19 Prevention Study (NSCPS) to better understand schools’ experiences throughout the COVID-19 pandemic.10 NSCPS involved five survey waves administered June 2021–May 2022 to a nationally representative sample of K-12 public schools. The sampling frame (i.e., a list of all public schools in the 50 states and the District of Columbia) used Common Core Data from the National Center for Education Statistics (NCES) and the MDR database.11 Excluded from the sampling frame were schools that were private, alternative, run by the U.S. Department of Defense, had fewer than 30 students, or provided services to a “pull-out” population in another eligible school.
The sample was stratified based on school level (elementary, middle, high), NCES school locale (city, suburb, town, rural), and region (Northeast, South, Midwest, West). For each survey wave, we invited the entire sample of 1602 schools, excluding schools which explicitly refused participation. Schools that did not participate in one survey wave were still eligible to participate in subsequent waves. For each wave, participants were contacted by phone calls and emails, and those who agreed to participate were emailed a unique survey link to complete the survey online. Respondents were school principals or designees familiar with schools’ COVID-19 policies and practices (e.g., assistant principals, school nurses). Data for this study came from Wave 4 (February 14–March 27, 2022, n=420, response rate = 26%), and included a one-time module assessing challenges schools experienced in serving students who receive special education services or have underlying health conditions, as well as accessible communication strategies schools implemented. Respondents were primarily principals (n=307), followed by school nurses (n=53), and other school-level designees (n=39). This study is part of a larger data collection effort to characterize schools’ responses to the COVID-19 pandemic; additional details about the NSCPS can be found elsewhere.10 , 12
Measures
We assessed challenges that schools experienced in providing services and related supports or accommodations since the start of the 2021-2022 school year to students who receive special education services or those with underlying health conditions as a proxy to identify students who might face barriers to implementing prevention strategies or who are at increased risk of severe COVID-19 illness.13 We used two “mark all that apply” survey questions listing potential challenges. We assessed five school-based challenges (barriers presented by the school’s physical infrastructure; lack of time to prepare and implement preventive measures; students’ difficulties following COVID-19 prevention strategies; staff difficulties following COVID-19 prevention strategies; and staff shortages) and five home-based challenges (non-existent or insufficient home internet access; lack of learning partners or coaches; lack of digital literacy among students’ families; difficulty in providing families with technology support; and difficulty in communicating with families). Additionally, schools reported accessible communication strategies used for COVID-19 prevention, such as visual messaging, translation into other needed languages, and auditory messages with a separate “mark all that apply” survey question. See Appendix 1 for exact question wording and response options. The NCES school locale, an urban-rural school-level classification scheme, was linked to survey data.14 Due to small sample sizes, we combined “city” and “suburb” locales and “town” and “rural” locales.
Statistical analyses
Analyses used survey weights that accounted for survey nonresponse and the design strata. We present weighted prevalences and 95% confidence intervals (CI) of challenges and the strategies among K-12 public schools, and by locale, suppressing estimates with a relative standard error ≥30%. Chi-square tests were used to test for differences in the prevalence of challenges and strategies by locale. P-values < .05 were considered statistically significant. Analyses were conducted using R (version 4.1.2; R Foundation).
Results
The participating sample was diverse in terms of school level (elementary = 234, middle = 100, high = 86), region (Northeast = 67, Midwest = 125, South = 131, West = 97), and school locale (city = 81, suburb = 129, town = 58, and rural = 125). Weighted prevalences of school- and home-based challenges and use of accessible communication strategies among K-12 public schools are presented in Table 1 . For school-based challenges in providing services and related supports or accommodations to students receiving special education or with underlying health conditions, over half of schools reported staff shortages (51.3%) and almost a third reported difficulty in having students follow COVID-19 prevention strategies (32.4%). Other less frequent challenges included barriers presented by physical infrastructure (e.g., insufficient space to support podding or cohorting) (15.6%), lack of time to prepare and implement prevention strategies (7.2%), and difficulty in having staff follow COVID-19 prevention strategies (7.1%).Table 1 School- and home-based challenges in providing services to students who receive special education services or students with underlying health conditions and accessible communication strategies used in COVID-19 prevention among K-12 public schools – National School COVID-19 Prevention Study, United States, February-March 2022
Table 1Experience or strategy % (95% CI)a
School-based challengesb
Barriers presented by the school’s physical infrastructure 15.6 (12.1 - 20.0)
Lack of time to prepare and implement COVID-19 preventive Measures 7.2 (4.8 - 10.6)
Difficulty in having students follow COVID-19 prevention strategies 32.4 (27.6 - 37.4)
Difficulty in having school and support staff follow COVID-19 prevention strategies 7.1 (4.8 - 10.5)
Staff shortages 51.3 (45.9 - 56.6)
None 28.7 (24.3 - 33.6)
Don’t know 9.0 (6.5 – 12.4)
Home-based challengesc
Non-existent or insufficient access to the internet at home 18.8 (15.0 - 23.3)
Lack of learning partners or coaches at home 25.5 (21.1 - 30.4)
Lack of digital literacy among students’ families 21.4 (17.5 - 25.9)
Difficulty in providing families with technology support 13.0 (9.7 - 17.1)
Difficulty in communicating with families 14.8 (11.5 - 18.9)
None 42.7 (37.4 – 48.1)
Don’t know 16.5 (13.0 – 20.7)
Accessible communication strategies used in COVID-19 preventiond
American Sign Language 3.9 (2.2 - 6.9)
Translation into the main languages used by families in schools 48.4 (43.4 - 53.4)
Auditory messages 42.2 (37.3 - 47.3)
Low-literacy materials 7.3 (4.9 - 10.7)
Visual messaging 45.8 (40.4 - 51.2)
Image descriptors (or “alt-text”) 12.9 (9.6 - 17.2)
Captions to videos on school webpages or social media sites 12.7 (9.7 - 16.5)
Transcripts that accompany podcasts or videos 6.7 (4.4 - 10.1)
None 14.2 (10.8 – 18.3)
Don’t know 13.1 (9.9 – 17.1)
Note: CI = Confidence interval
a Weighted percentages and 95% CIs are presented. The prevalence of responses where the relative standard error was >30% were not presented. For school-based challenges this included: “Lack of personal protective equipment (e.g., masks, face shields)”; “Lack of suitable computer hardware and software technology to support students’ remote learning (e.g., assistive technology tools, adaptive technology tools)”. For accessible communication strategies, this included “Braille”.
b Question asked, “Since the start of the 2021-2022 school year, which of the following school-based challenges is your school experiencing in providing services and related supports or accommodations to students who receive special education services or students with underlying health conditions that place them at greater risk of severe illness and complications from COVID-19 (e.g., asthma, immunosuppression)?”
c Question asked, “Since the start of the 2021-2022 school year, which of the following challenges related to students’ home environment is your school experiencing in providing services and related supports oraccommodations to students who receive special education services or students with underlying health conditions that place them at greater risk of severe illness and complications from COVID-19 (e.g., asthma, immunosuppression)?”
d Question asked, “Since the start of the COVID-19 pandemic, what types of accessible communication strategies has your school used with students or their families to prevent the spread of COVID-19?”
For home-based challenges, over a fifth of schools reported lack of learning partners or coaches (25.5%) and lack of digital literacy among students’ families (21.4%). Additional home-based challenges include non-existent or insufficient home internet access (18.8%), difficulty in communication with families (14.8%), and difficulty in providing families with technology support (13.0%). Finally, 16.5% of respondents reported lack of knowledge about home-based challenges experienced by these students.
The three most frequently reported accessible communication strategies for COVID-19 prevention were translation into other languages (48.4%), visual messaging (45.8%), and auditory messages (42.2%). The three least frequently reported were American Sign Language (3.9%), transcripts that accompany podcasts or videos (6.7%), and low-literacy materials (7.3%).
Table 2 presents challenges and strategies by school locale. Compared to city/suburb schools, town/rural schools were more likely to experience non-existent or insufficient home internet access (14.8% vs. 23.9%, p=0.04). No other challenges significantly differed by school locale. In terms of accessible communication strategies, town/rural schools were less likely than city/suburb schools to report translation into other languages (60.1% vs. 34.8%, p<0.001) and visual messaging (54.2% vs. 35.0%, p=0.001). Compared to city/suburb schools, town/rural schools were more likely to lack accessible communication strategies (8.3% vs. 22.1%, p<0.001).Table 2 School and home-based challenges in providing services to students who receive special education services or students with underlying health conditions and accessible communication strategies used in COVID-19 prevention by NCES school locale – National School COVID-19 Prevention Study, United States, February-March 2022
Table 2 NCES school localea p-valuec
City/Suburb
% (95% CI)b Town/Rural
% (95% CI)b
School-based challengesd
Barriers presented by the school’s physical Infrastructure 16.9 (12.0 - 23.2) 13.5 (8.6 - 20.6) 0.42
Lack of time to prepare and implement preventive Measures 8.2 (4.9 - 13.4) ‒e
Difficulty in having students follow COVID-19 prevention strategies 32.4 (26.1 - 39.4) 33.3 (25.6 - 42.0) 0.87
Difficulty in having school and support staff follow COVID-19 prevention strategies 6.7 (3.8 - 11.6) 7.0 (3.9 - 12.3) 0.92
Staff shortages 55.0 (47.3 - 62.5) 49.9 (41.6 - 58.2) 0.37
Lack of software and hardware technology ‒e ‒e
None 25.0 (19.3 – 31.7) 30.4 (23.1 – 38.8) 0.29
Don’t Know 9.5 (6.1 – 14.5) 8.6 (5.0 – 14.5) 0.78
Home-based challengesf
Non-existent or insufficient access to the internet at Home 14.8 (10.1 - 21.0) 23.9 (17.6 - 31.6) 0.04
Lack of learning partners or coaches at home 23.9 (18.1 - 30.9) 28.1 (21.0 - 36.5) 0.41
Lack of digital literacy among students’ families 22.8 (17.7 - 28.8) 16.3 (10.9 - 23.6) 0.14
Difficulty in providing families with technology Support 12.7 (8.4 - 18.9) 13.7 (8.8 - 20.7) 0.81
Difficulty in communicating with families 16.5 (11.7 - 22.7) 11.9 (7.6 - 18.2) 0.23
None 45.3 (38.1 52.7) 40.6 (32.3 – 49.5) 0.42
Don’t Know 15.9 (11.4 – 21.7) 16.8 (11.5 – 24.0) 0.82
Accessible communication strategies used in COVID-19 preventiong
Translation into the main languages used by families in schools 60.1 (53.0 - 66.7) 34.8 (27.7 - 42.7) <0.001
Auditory messages 45.5 (38.8 - 52.4) 39.8 (31.8 - 48.4) 0.30
Low literacy materials ‒e 11.5 (7.0 - 18.3)
Visual messaging 54.2 (46.3 - 61.8) 35.0 (27.5 - 43.3) 0.001
Image descriptors (or “alt-text”) 16.3 (11.2 - 23.0) ‒e
Captions to videos on school webpages or social media sites 14.1 (9.8 - 19.8) 12.0 (7.6 - 18.5) 0.57
Transcripts that accompany podcasts or videos 8.3 (4.9 - 13.9) ‒e
None 8.3 (5.1 – 13.3) 22.1 (15.5 – 30.3) <0.001
Don’t Know 11.4 (7.4 – 17.2) 13.4 (8.7 – 20.1) 0.59
NOTE: CI = Confidence interval; NCES = National Center for Education Statistics
a School locale was categorized based on the NCES locale classification scheme into two categories: City/Suburb or Town/Rural.
b Weighted percentages and 95% CIs are presented. The prevalence of responses where the relative standard error was >30% for both categories of NCES locale are not presented. For school-based challenges this included: “Lack of personal protective equipment (e.g., masks, face shields)”; “Lack of suitable computer hardware and software technology to support students’ remote learning (e.g., assistive technology tools, adaptive technology tools)”. For accessible communication strategies, this included “American Sign Language” and “Braille”.
c Chi-square tests were used to identify differences in prevalence of experiences and strategies by NCES locale.
d Question asked, “Since the start of the 2021-2022 school year, which of the following school-based challenges is your school experiencing in providing services and related supports or accommodations to students who receive special education services or students with underlying health conditions that place them at greater risk of severe illness and complications from COVID-19 (e.g., asthma, immunosuppression)?”
e Estimate was suppressed as the relative standard error was ≥30%.
f Question asked, “Since the start of the 2021-2022 school year, which of the following challenges related to students’ home environment is your school experiencing in providing services and related supports oraccommodations to students who receive special education services or students with underlying health conditions that place them at greater risk of severe illness and complications from COVID-19 (e.g., asthma, immunosuppression)?”
g Question asked, “Since the start of the COVID-19 pandemic, what types of accessible communication strategies has your school used with students or their families to prevent the spread of COVID-19?”
Discussion
This study provides insight on home- and school-based challenges schools experienced in serving students who receive special education services or with underlying health conditions using nationally representative data on K-12 public schools. The most frequently reported school-based challenge was staff shortages; this was the only challenge in this study reported by more than half of the schools and aligns with research highlighting staff shortage problems in school settings.15 , 16 Students who receive special education services or have underlying health conditions, in particular, might require dedicated support from staff (e.g., paraprofessionals who provide 1-1 support) and might be more acutely impacted by staff shortages.17 , 18 Staff shortages might have also contributed to schools’ inability to provide adequate virtual supports and accommodations.3 A frequently reported home-based challenge, and the only challenge in the study to differ significantly by school locale, was non-existent or insufficient home internet access. Schools often rely on web-based platforms to deliver communication and educational materials, and many relied on remote instruction during the pandemic. However, the lack of stable internet access might have posed a barrier to maintaining continuity in communication and meeting the needs of these student populations.19 As previously documented for low SES communities, our findings suggest town/rural schools experienced greater challenges in internet access.20, 21, 22, 23, 24 As schools continue to navigate the COVID-19 pandemic, and prepare for local outbreaks of other infectious diseases, addressing internet and technology barriers with Wi-Fi hotspots and other equipment to support student populations is important. Moreover, schools might need more staff and resources to address challenges reported by populations that receive special education services or have underlying health conditions, such as lack of learning partners and digital literacy among students’ families. Concerningly, almost 17% of schools reported that they did not know what home-based challenges were experienced by these populations, underscoring the importance of communication between students, staff, and families.25, 26, 27
Clear and timely communication between schools and families is critical during public health emergencies. Information on how to prevent the spread of COVID-19 and other diseases is key in educating and empowering all students and their families. Although schools reported several different communication strategies to share COVID-19 information, the prevalence of schools using accessible communication strategies was low overall. For example, only about 7% of schools reported using low-literacy materials in communicating about COVID-19. Several resources and guidance documents are available to help make health communication materials easy to understand for different audiences.28, 29, 30 Schools can also use the CDC Clear Communication Index to assess the effectiveness of their health communication materials.31 This index aims to assist in the development of clear messages that aid public understanding. Additionally, translation of materials into other languages can help ensure key information reaches all students and families. Community partners who provide support (e.g., community health partners, faith-based organizations) and understand the beliefs and practices within the community can be included when developing communication plans for schools.32
Results also suggest underuse of accessible communication strategies such as American Sign Language and captioning which could have benefitted students and families. For students with disabilities or underlying health conditions at increased risk or with additional barriers to using prevention strategies, this communication is needed to make individualized determinations and ensure equal access to health and educational services.33 Although our study did not provide data on students’ specific accessibility related needs, given that 15% of the student population has a documented disability, it is likely that some of these strategies could have helped bridge potential communication gaps.1 Almost one third of schools reported challenges in having these student populations follow COVID-19 prevention strategies, although we are not able to discern whether this was due to low overall compliance or disability-related challenges. Nonetheless, this underscores the importance of clear and accessible health communication messages. Given that town/rural schools were less likely to use accessible communication strategies compared to city/suburb schools, town/rural schools might need additional supports in developing and using such strategies. Additional training on developing materials that incorporate accessible communication strategies may be needed.19 , 34 State and local education agencies and teachers can collaborate to develop training on accessible communication strategies, digital tools and evidence-based instructional methods such as Universal Design for Learning (UDL) that promote accessibility for all learners.6 , 35, 36, 37
This study has several limitations. First, the data represented only the perspective of school administrators and might be influenced by social desirability bias or limits on respondents’ knowledge of the assessed challenges and strategies. Second, the survey response rate was low (26%) but our analyses incorporated weights which accounted for nonresponse. Third, our sample was small. Fourth, we did not collect data on the distribution of disabilities and underlying health conditions among the students and families in sampled schools, nor on the specific special education services that these schools delivered, prohibiting us from examining differences by these school characteristics. Fifth, the school- and home- based challenges and accessible strategies assessed in the survey were not exhaustive.
Conclusions
Schools experienced numerous school- and home-based challenges in serving students who receive special education services or who have underlying health conditions during the COVID-19 pandemic. Key school-based challenges included staff shortages and difficulty in having students comply with COVID-19 prevention strategies that were implemented, while key home-based challenges included insufficient internet access, lack of learning partners, and low digital literacy among families. The benefits of in-person learning for students with disabilities or those with underlying health conditions extend beyond academics to include meals and extra-curricular activities.38 However, these school- and home-based challenges underscore the importance of improving internet access and bolstering staffing support in schools, as well as improving communication with families to better support students at home when remote learning is utilized. Additionally, promoting the use of accessible communication strategies for COVID-19 and other infectious disease prevention is important, particularly in town/rural communities where low uptake of accessible communication strategies indicates need for more supports and resources to meet student needs. Improving accessibility of communication might increase reach of important messages during public health emergencies and support positive health and educational outcomes for students by enhancing access to materials and the usability of critical information provided.
Conflict of interest:
The authors declare no conflict of interest.
Appendix A Supplementary data
The following is the Supplementary data to this article:
Acknowledgements
The authors would like to thank the school staff for their participation in the study and willingness to provide insights on COVID-19 prevention. Additionally, we acknowledge April Carswell, James Demery, Cherrelle Dorleans, Adrian King, Leah Powell, Lynnea Roberts, India Rose, Syreeta Skelton-Wilson, Lorin Stewart, Dana Keener Mast, Lucas Godoy Garraza, Nicole Gonzalez, Christine Walrath, Lisa Barrios, Leah Robin, Carmen Ashley, Seraphine Pitt-Barnes, Michelle Carman-McClanahan, Nancy Brener, Marci Hertz, and the rest of the National School COVID-19 Prevention Study team.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.dhjo.2022.101428.
==== Refs
References
1 National Center for Education Statistics.Students with Disabilities.Available from, https://nces.ed.gov/programs/coe/indicator/cgg/students-with-disabilities
2 Health NSoCs.NSCH 2018-2019: Number of Current or Lifelong Health Conditions, Nationwide, Age in 3 Groups. February 24, 2021.www.childhealthdata.org
3 Averett K.H. Remote learning, COVID-19, and children with disabilities AERA Open 7 2021 23328584211058471 10.1177/23328584211058471
4 Harris B. McClain M.B. O'Leary S. Shahidullah J.D. Implications of COVID-19 on School Services for Children with Disabilities: Opportunities for Interagency Collaboration J Dev Behav Pediatr 42 2021 236 239 10.1097/DBP.0000000000000921 33596007
5 Houtrow A. Harris D. Molinero A. Levin-Decanini T. Robichaud C. Children with disabilities in the United States and the COVID-19 pandemic Journal of Pediatric Rehabilitation Medicine 13 2020 415 424 10.3233/prm-200769 33185616
6 Rice M.F. Special education teachers' use of technologies during the COVID-19 era (Spring 2020-Fall 2021) TechTrends 66 2022 310 326 10.1007/s11528-022-00700-5 35098256
7 Centers for Disease Control and Prevention.Operational Guidance for K-12 Schools and Early Care and Education Programs to Support Safe In-Person Learning. August 18, 2022.Available from, https://public4.pagefreezer.com/browse/CDC%20Covid%20Pages/12-08-2022T12:32/https://www.cdc.gov/coronavirus/2019-ncov/community/schools-childcare/k-12-childcare-guidance.html
8 Sherby M.R. Kalb L.G. Coller R.J. Supporting COVID-19 School Safety for Children With Disabilities and Medical Complexity Pediatrics 149 2022 10.1542/peds.2021-054268H
9 Schafer E.C. Dunn A. Lavi A. Educational Challenges During the Pandemic for Students Who Have Hearing Loss Lang Speech Hear Serv Sch 52 2021 889 898 10.1044/2021_lshss-21-00027 34185568
10 Centers for Disease Control and Prevention.National School COVID-19 Prevention Study (NSCPS). July 27, 2022.Available from, https://www.cdc.gov/healthyyouth/data/nscps/index.htm
11 Market Data Retrieval. 2022.https://mdreducation.com/
12 Pampati S. Rasberry C.N. McConnell L. Ventilation Improvement Strategies Among K-12 Public Schools - The National School COVID-19 Prevention Study, United States, February 14-March 27, 2022 MMWR Morb Mortal Wkly Rep 71 2022 770 775 10.15585/mmwr.mm7123e2 35679198
13 Centers for Disease Control and Prevention.People of any age with underlying medical conditions.https://stacks.cdc.gov/view/cdc/89841
14 National Center for Education Statistics.Education Demographic and Geographic Estimates - Locale Classifications. July 27, 2022.Available from, https://nces.ed.gov/programs/edge/Geographic/LocaleBoundaries
15 Institute of Education Sciences.2022 School Pulse Panel. July 27, 2022.Available from, https://ies.ed.gov/schoolsurvey/spp/
16 Rosenberg M. Mason-Williams L. Kimmel L. Sindelar P.T. Addressing Teacher Shortages in the COVID-19 Landscape: Viewing Teacher Candidates as Assets Excelsior: Leadership in Teaching and Learning 13 2021 86 95 10.14305/jn.19440413.2021.13.2.01
17 Hurwitz S. Garman-McClaine B. Carlock K. Special education for students with autism during the COVID-19 pandemic: "Each day brings new challenges Autism 26 2022 889 899 10.1177/13623613211035935 34344221
18 Sonnenschein S. Stites M.L. Grossman J.A. Galczyk S.H. This will likely affect his entire life": Parents' views of special education services during COVID-19 Int J Educ Res 112 2022 101941 10.1016/j.ijer.2022.101941
19 Smith C. Challenges and opportunities for teaching students with disabilities during the COVID-19 pandemic International Journal of Multidisciplinary Perspectives in Higher Education 5 2021 167 173 10.32674/jimphe.v5i1.2619
20 Jameson J.M. Stegenga S.M. Ryan J. Green A. Free Appropriate Public Education in the time of COVID-19 Rural Special Education Quarterly 39 2020 181 192 10.1177/8756870520959659
21 Kormos E.M. The Unseen Digital Divide: Urban, Suburban, and Rural Teacher Use and Perceptions of Web-Based Classroom Technologies Computers in the Schools 35 2018 19 31 10.1080/07380569.2018.1429168
22 Lai J. Widmar N.O. Revisiting the Digital Divide in the COVID-19 Era Appl Econ Perspect Policy 43 2021 458 464 10.1002/aepp.13104 33230409
23 Francom G.M. Lee S.J. Pinkney H. Technologies, Challenges and Needs of K-12 Teachers in the Transition to Distance Learning during the COVID-19 Pandemic TechTrends 65 2021 589 601 10.1007/s11528-021-00625-5 34223560
24 Reddick C.G. Enriquez R. Harris R.J. Sharma B. Determinants of broadband access and affordability: An analysis of a community survey on the digital divide Cities 106 2020 102904 10.1016/j.cities.2020.102904
25 Lipkin M. Crepeau-Hobson F. The impact of the COVID-19 school closures on families with children with disabilities: A qualitative analysis Psychol Sch 2022 10.1002/pits.22706
26 Klosky J.V. Gazmararian J.A. Casimir O. Blake S.C. Effects of remote education during the COVID-19 pandemic on young children's learning and academic behavior in Georgia: Perceptions of parents and school administrators Journal of School Health 92 2022 656 664 10.1111/josh.13185 35411613
27 Neece C. McIntyre L.L. Fenning R. Examining the impact of COVID-19 in ethnically diverse families with young children with intellectual and developmental disabilities J Intellect Disabil Res 64 2020 739 749 32808424
28 Centers for Disease Control and Prevention.Health Equity Considerations for Developing Public Heath Communications. July 27, 2022.Available from, https://www.cdc.gov/healthcommunication/Comm_Dev.html
29 National Institutes of Health.Clear Communication. July 28, 2022.Available from, https://www.nih.gov/institutes-nih/nih-office-director/office-communications-public-liaison/clear-communication/clear-simple
30 plainlanguage.gov.Plain Language Guidelines. July 27, 2022.Available from, https://www.plainlanguage.gov/
31 Centers for Disease Control and Prevention.The CDC Clear Communication Index. July 27, 2022.Available from, https://www.cdc.gov/ccindex/index.html
32 Centers for Disease Control and Prevention.Access and Functional Needs Toolkit: Integrating a Community Partner Network to Inform Risk Communication Strategies. July 27, 2022.Available from, https://www.cdc.gov/cpr/readiness/afntoolkit.htm
33 McDevitt S.E. Mello M.P. From Crisis to Opportunity: Family Partnerships with Special Education Preservice Techers in Remote Practicum during the COVID-19 School Closures School Community Journal 31 2021 325 346
34 Clausen J.M. Bunte B. Robertson E.T. Professional Development to Improve Communication and Reduce the Homework Gap in Grades 7-12 during COVID-19 Transition to Remote Learning Journal of Technology and Teacher Education 28 2020 443 451
35 Basham J.D. Blackorby J. Marino M.T. Opportunity in crisis: The role of Universal Design for Learning in educational redesign Learning Disabilities: A Contemporary Journal 18 2020 71 91
36 Huck C. Zhang J. Effects of the COVID-19 pandemic on K-12 education: A systematic literature review Educational Research and Developmental Journal 24 2021 53 84
37 Barbour M.K. Introducing a special collection of papers on k-12 online learning and continuity of Instruction after emergency remote teaching TechTrends 66 2022 298 300 10.1007/s11528-022-00712-1 35229086
38 Brandenburg J.E. Holman L.K. Apkon S.D. Houtrow A.J. Rinaldi R. Sholas M.G. School reopening during COVID-19 pandemic: Considering students with disabilities Journal of Pediatric Rehabilitation Medicine 13 2020 425 431 10.3233/PRM-200789 33136082
| 0 | PMC9741764 | NO-CC CODE | 2022-12-14 23:29:57 | no | Disabil Health J. 2022 Dec 11;:101428 | utf-8 | Disabil Health J | 2,022 | 10.1016/j.dhjo.2022.101428 | oa_other |
==== Front
Ageing Res Rev
Ageing Res Rev
Ageing Research Reviews
1568-1637
1872-9649
Elsevier B.V.
S1568-1637(22)00260-4
10.1016/j.arr.2022.101818
101818
Review
Immunosenescence and inflamm-ageing in COVID-19
Zinatizadeh Mohammad Reza ab
Zarandi Peyman Kheirandish ab
Ghiasi Mohsen c
Kooshki Hamid d
Mohammadi Mozafar e
Amani Jafar f
Rezaei Nima ghi⁎
a Department of Biology, Science and Research Branch, Islamic Azad University, Tehran, Iran
b Cancer Biology Signaling Pathway Interest Group (CBSPIG), Universal Scientific Education and Research Network (USERN), Tehran, Iran
c Department of Cellular and Molecular Biology, Faculty of Advanced Science and Technology, Tehran Medical Sciences, Islamic Azad University, Tehran, Iran
d Nanobiotechnology Research Center, Baqiyatallah University of Medical Sciences, Tehran, Iran
e Applied Biotechnology Research Centre, Baqiyatallah University of Medical Sciences, Tehran, Iran
f Applied Microbiology Research Center, Systems Biology and Poisonings Institute, Baqiyatallah University of Medical Sciences, Tehran, Iran
g Research Center for Immunodeficiencies, Children’s Medical Center, Tehran University of Medical Sciences, Tehran, Iran
h Department of Immunology, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran
i Network of Immunity in Infection, Malignancy and Autoimmunity (NIIMA), Universal Scientific Education and Research Network (USERN), Tehran, Iran
⁎ Correspondence to: Children's Medical Center Hospital, Dr. Qarib St, Keshavarz Blvd, Tehran 14194, Iran.
11 12 2022
2 2023
11 12 2022
84 101818101818
28 4 2022
4 11 2022
8 12 2022
© 2022 Elsevier B.V. All rights reserved.
2022
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The destructive effects of coronavirus disease 2019 (COVID-19) on the elderly and people with cardiovascular disease have been proven. New findings shed light on the role of aging pathways on life span and health age. New therapies that focus on aging-related pathways may positively impact the treatment of this acute respiratory infection. Using new therapies that boost the level of the immune system can support the elderly with co-morbidities against the acute form of COVID-19. This article discusses the effect of the aging immune system against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and the pathways affecting this severity of infection.
Abbreviations
COVID-19, Coronavirus disease 2019
SARS-CoV-2, Severe acute respiratory syndrome coronavirus 2
JAK, Janus kinase
STAT, Signal transducer and activator of transcription
MyD88, Myeloid differentiation factor 88
NF-κB, Nuclear factor kappa B
PRRs, Pattern recognition receptors
IFN, Interferon
ACE2, Angiotensin-converting enzyme 2
ARDS, Acute respiratory distress syndrome
IL, Interleukin
TNF-α, Tumor necrosis factor-alpha
MAPK, Mitogen-activated protein kinase
G-CSF, Granulocyte colony-stimulating factor
GM, Granulocyte-macrophage
CXCL, C–X–C motif chemokine ligand
NOD, Nucleotide-binding oligomerization domain
NLRs, NOD-like receptors
NLRP3, NLRs-pyrin domain-containing protein 3
DAMPs, Damage-associated molecular patterns
ROS, Reactive oxygen species
ATP, Adenosine triphosphate
TLR, Toll-like receptor
SASP, Senescence-associated secretory phenotype
GFs, Growth factors
MMPs, Matrix metalloproteinases
PAMPs, Pathogen-associated molecular patterns
TMPRSS2, Transmembrane protease serine 2
pDCs, Plasmacytoid dendritic cells
IRF7, Interferon regulatory factor 7
RIG-I, Retinoic acid-inducible gene-I
MAVS, Mitochondrial antiviral signaling
MHC, Major histocompatibility complex
APCs, Antigen-presenting cells
CCR7, C-C chemokine receptor 7
HLA, Human leukocyte antigen
CoVs, Coronaviruses
TCRs, T-cell receptors
CMV, Cytomegalovirus
IgM, Immunoglobulin M
RBD, Receptor-binding domain
GC, Germinal center
BCR, B-cell receptor
PD-1, programmed cell death protein-1
Tim-3, T-cell immunoglobulin mucin 3
AA, Amino acid
IDO, Indoleamin 2,3 dioxygenase
GCN-2, General control non-derepressible-2
Eif-2, Eukaryotic initiationfactor-2
APOA1, Apolipoprotein A1
APOM, Apolipoprotein M
S1P, Sphingosine-1-phosphate
S1PL, S1P lyase
Spns2, Spinster homology 2
LDH-A, Lactat dehydrogenase A
GLUT-1, Glucose transporter 1
PHD, Prolyl hydroxylase domain
HIF, Hypoxia-inducible factor
SLOs, Secondary lymphoid organs
FAO, Fatty acid oxidation
TCAs, Tricarboxylic acids
cPLA2, Cytosolic phospholipase A2
LPS, Lipopolysaccharid
SDH, Succinate dehydrogenase
RET, Reverse electron transport
mtROS, Mitochondrial ROS
PFKFB3, 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 PFKFB3
Keywords
Immunosenescence
Inflamm-ageing
SARS-CoV-2
COVID-19
==== Body
pmc1 Introduction
As people get older, their immunity to pathogens decreases, and this lack has been proven in experiments. Studies show that aging reduces the effectiveness of the immune system against pathogens. In older people, the body shows reduce tolerance to the pathogen, and the effectiveness of related vaccines decreases (Goodwin et al., 2006, Hainz et al., 2005, Kaml et al., 2006, Melegaro and Edmunds, 2004, Wolters et al., 2003, Yoshikawa, 2000, Zinatizadeh et al., 2022). So it is clear that the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has affected the most vulnerable and older people. Since the coronavirus disease 2019 (COVID-19) outbreak, age has been proposed and proven to be the most critical factor influencing the severity of this disease (Wu et al., 2020). The death toll from the infection of SARS-CoV-2 is high in the elderly (Bajgain et al., 2021, Santesmasses et al., 2020, Zheng et al., 2020c). The incidence of SARS-CoV-2 infection and mortality in the elderly is high worldwide. In the United States alone, the incidence of hospitalization and death from COVID-19 is about 1300 folds larger in the 65–74 age group and 8700 folds greater in the 85-year-old than in the 5–17 age group (Gold et al., 2020). Also, a study in Iran showed that most cases of acute infections were in the age group of 50–60 years. And mortality is significantly higher in the elderly (Nikpouraghdam et al., 2020). In addition, studies in Saudi Arabia (Ibrahim et al., 2021), Iraq (Al-Mosawi, 2021), China (Leung, 2020), and Brazil (Machado et al., 2020) have also shown that age has a special role in mortality. Other diseases of the elderly, such as cardiovascular problems and diabetes, also play a role in the increasing vulnerability of this group to COVID-19. However, it should be noted that age quantity is still the most critical factor influencing the severity of this disease (Bajgain et al., 2021, Ho et al., 2020).
This article has tried to investigate the role of the age factor on the severity of COVID-19 disease, its host action and responses during infection, and physiological and immunological mechanisms against the pathogen.
2 Dysregulation of immune system on ageing
The main functions of the immune system in responding to a viral infection include the following: A) Initiation of a local inflammatory response to activate immune cells; B) Destroying the cells involved in the virus; C) Activation of the adaptive immune response. As they grow older, the immune system will not be able to process these functions properly. The phenomenon of inflamm-ageing, which represents the attendance of systemic inflammatory mediators in the body of the elderly, causes more turbulences to the immune system and strengthens many persistent diseases of them. This phenomenon further causes the weakening of the immune system in people with increasing age, also known as immunosenescence (Ferrucci and Fabbri, 2018). In response to internal and external physiological stresses, inflammation occurs in the body and is usually caused by aging and senescence of immune cells. These stressors can explain the difference in inflammatory biological ages (Alpert et al., 2019, Furman et al., 2019).
Many inflammatory intermediates are generated by immune cells and can affect their function. Excessive release of proinflammatory cytokines can make it difficult for immune cells to signal. Such problems can be noted in the Janus kinase (JAK)-signal transducer and activator of transcription (STAT) pathway, myeloid differentiation factor 88 (MyD88), nuclear factor kappa B (NF-κB), and inflammasomes, resulting in high basal activation but impaired immune cell response to more cytokine and pattern recognition receptors (PRRs) excitation (Rea et al., 2018, Shen-Orr et al., 2016). It can be found that many immune cells respond less to external and internal stresses as we age (Chougnet et al., 2015, Cumberbatch et al., 2002, Manser and Uhrberg, 2016, Metcalf et al., 2015, Metcalf et al., 2017, Rea et al., 2018, Van Duin et al., 2007, Zacca et al., 2015). This may explain why mediated chronic systemic inflammation is associated with optimal vaccine responses in younger and older (Fourati et al., 2016, McDade et al., 2011, Verschoor et al., 2017). Immunosenescence with symptoms such as impaired antigen exposure and naive priming of T cells, tendency to segregation myeloid lineage in the bone marrow, delayed type I interferon (IFN) response, decreased cytotoxic function of CD8+ T cells, reduced phagocytic function for many types of the innate immune cell, a confined naive T cell, and B cell repertoire and ruinous generation of highly agitated antibodies ( Table 1) (Derhovanessian et al., 2008, Weiskopf et al., 2009). This incapability in the immune system makes older people more defenseless to viral diseases such as SARS-CoV and SARS-CoV-2. The effect of aging on immune system strength has been widely investigated (Akha, 2018, Boraschi and Italiani, 2014, Nikolich-Žugich, 2018, Nikolich-Zugich et al., 2020, Weiskopf et al., 2009). This paper focuses on the effects of age on immune system response to COVID-19.Table 1 Immunosenescence and its negative consequences in elderly COVID-19 patients.
Table 1Immunosenescence characteristics Immune response consequences References
Monocytes/macrophages Innate Altered TLR expressionDecreased phagocytic abilityDecreased antigen presentationIncreased cytokine production (Zheng et al., 2020b)
Dendritic cells Innate Reduced of pDCs (Márquez et al., 2020, Zingaropoli et al., 2021)
Neutrophils Innate Altered TLR expressionVariations in oxidative stress pathwaysInflammatory activityRaised cellular module and SASPDecreased autophagy and DNA damage (Bajaj et al., 2021b)
NK cells Innate Prevention of upregulation HLA Class IAltered NKG2A expression (Wen et al., 2020, Wilk et al., 2020, Witkowski et al., 2022, Yoo et al., 2021)
B cells Adaptive Decreased antibody titers (Singh et al., 2021)
T cells Adaptive Decreased numbers of T cellDecreased number of regulatory T cellsScarcity of naive T cellsDisrupted of antigen-specific responsesUncoordinated adaptive immune responses (Cunha et al., 2020, Diao et al., 2020, Moderbacher et al., 2020)
3 How immunosenescence and inflamm-ageing may contribute to severity of COVID-19
3.1 Age-related alterations in ACE2 receptor
The angiotensin-converting enzyme 2 (ACE2) cellular receptor plays an essential role in primary inflammatory procedures through the renin-angiotensin-aldosterone signaling pathway. It is involved in processing the functions of the innate immune system. As mentioned earlier, the onset of an inflammatory response leads to the activation of immune cells and the recruitment of these cells. Angiotensin II is converted by ACE2 to angiotensin 1–7. Signaling of the angiotensin II in vascular cells causes a proinflammatory condition (Wang et al., 2014, Zarandi et al., 2021). Animal examines performed on mice suffering from acute respiratory distress syndrome (ARDS) have shown that ACE2 has an anti-inflammatory effect and prevents critical lung damage (Imai et al., 2005). On the other hand, angiotensin 1–7 has been shown to reduce the generation of proinflammatory cytokines such as interleukin 6 (IL-6), tumor necrosis factor-alpha (TNF-α), and IL-8 by inhibiting the signaling pathway of P38 mitogen-activated protein kinase (MAPK)-NF-kB. In addition, it upregulates the expression of the anti-inflammatory cytokine IL-10 (Yu et al., 2018). During the initial response to SARS-CoV-2, the raised presence of angiotensin II, possibly due to overproduction of TNF-α and activation of local macrophages, has a significant effect on the promotion of severe COVID-19 related cytokine release (Banu et al., 2020). Animal studies have shown a decrease in ACE2 expression in the lungs of older mice contrasted to younger mice (Xudong et al., 2006). In addition, a study found a reduction in ACE2 mRNA in several tissues in older humans (Chen et al., 2020a). Downregulation of ACE2 levels has been observed in people with cardiovascular diseases and diabetics, which is also associated with the severity of COVID-19 (Tikellis and Thomas, 2012). Indeed, ACE2 expression is inversely correlated with the severity of COVID-19 (Chen et al., 2020a), which can be attributed to reduced disease tolerance (Wanhella and Fernandez-Patron, 2022). Disease tolerance declines with increased age and as a result of deteriorated immunity and may affect older individuals infected by SARS-CoV-2. Tolerance may become impaired with age due to declining tissue maintenance and capacity to repair damaged cells (Medzhitov et al., 2012). Although decreased ACE2 levels may be due to minor SARS-CoV-2 attack on host cells, excessive reduced levels of ACE2 may intensify proinflammatory response that leads to cytokine storm, severe lung damage, and ARDS ( Fig. 1). The high levels of angiotensin II shown in the plasma of patients with severe COVID-19 also support this hypothesis (Liu et al., 2020b). It should be noted that in studies on old and young rhesus macaques with SARS-CoV-2 infection, proinflammatory cytokines levels were higher in older macaques and emphasized the hypothesis that disease tolerance in older individuals is reduced (Rosa et al., 2021).Fig. 1 The presumptive pattern of immunosenescence and inflamm-ageing in COVID-19.
Fig. 1
3.2 Uninhibited inflammatory responses with age
Lung tissue damage results from uninhibited inflammatory responses to SARS-CoV-2 infection (Fig. 1). In patients with severe COVID-19, high levels of pro-inflammatory cytokines and chemokines, including IL-6, IL-1β, IL-2, IL-8, IL-17, granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage (GM)-CSF, C–X–C motif chemokine ligand 10 (CXCL10), CCL2, CCL3, and TNF is observed (Dinnon et al., 2020, Qin et al., 2020, Xu et al., 2020b, Zarandi et al., 2021). Similar cytokines, such as IL-6, IL-1α, IL-1β, TNF-α, and the chemokine CCL2, were also related to the senescent host response to SARS-CoV-2 in experiments on mice (Dinnon et al., 2020, Leist et al., 2020). Many levels of these cytokines have been shown to rise in the elderly due to inflamm-ageing. Research has shown that an increase in IL-6 is a negative consequence of COVID-19, a parameter of aging (Johnson, 2006). The IL-6 transcription factor activates the NF-κB pathway. This transcription factor has a significant role in regulating many pro-inflammatory genes (Brasier, 2010). Aging, NF-κB signaling, and inflammation are closely related (Salminen et al., 2008, Zinatizadeh et al., 2021). On the other hand, an anti-IL-6 receptor antibody has been shown to reduce the severity of COVID-19 in a group of elderly patients and points to the role of some cytokines in disease intensity (Guaraldi et al., 2020). Although IL-6 may play an important role in both local and systemic inflammation, it is unlikely to be the primary cause of inflamm-ageing.
The critical role of cytokines IL-1β and IL-18 in inflamm-ageing has also been proven (Rodrigues et al., 2020). These cytokines, which belong to the nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs)-pyrin domain-containing protein 3 (NLRP3), are involved in the pathology of aging-related diseases (Youm et al., 2013). SARS-CoV-2 activates the NLRP3 inflammasome and IL-1β and IL-18 levels in the elderly patient and increases inflammation and the severity of COVID-19 by potentiating pyroptosis and releasing damage-associated molecular patterns (DAMPs) (Junqueira et al., 2021, Rodrigues et al., 2021). Increasing reactive oxygen species (ROS) levels in the aging process leads to the activation of NLRP3 inflammation and intensifies the COVID-19 process in the elderly patient (Mishra et al., 2021, Tay et al., 2020, Zheng et al., 2020b). It should be noted that metformin reduces the severity of the disease by reducing the levels of adenosine triphosphate (ATP) and mitochondrial ROS levels that can stimulate NLRP3-mediated IL-1β generation (Xian et al., 2021). On the other hand, results using anti-IL-1 receptor agents have been shown to reduce COVID-19 fatalities and suggest NLRP3 inflammasome as the target of COVID-19 (Barkas et al., 2021, Shah, 2020).
On the other hand, an increase in NF-κB signaling has been shown in aging. Immune cells derived from an inflamm-ageing conditions have decreased response to severe ex vivo stimuli. However, monocytes emerge to be having an abnormal degree of responsiveness (Sayed et al., 2021). In a clinical study, analysis of monocytes in elderly COVID-19 patients cleared that they were enhanced for the IFN-γ response and TNF-α, IL-1β, and CXCL8 expression (Zheng et al., 2020b). Senescent monocytes were also enhanced for wide pathways related to toll-like receptor (TLR) signaling, oxidative stress, MAPK, and NF-κB, as well as the cyclin-dependent kinase inhibitor (p21), made this hypothesis that the existence of inflammatory aged or senescent cells with a senescence-associated secretory phenotype (SASP) was one activity for severe inflammatory in these cells in COVID-19 (Zheng et al., 2020b). Aging cells remain and cumulate in aging, indicating an extraordinary secretory phenotype defined by the existence of inflammatory cytokines. This SASP is defined by generating cytokines, chemokines, growth factors (GFs), matrix metalloproteinases (MMPs), fibronectin, and ROS (Coppé et al., 2010). A recent study found that exposure of senescent cell lines to pathogen-associated molecular patterns (PAMPs) and the SARS-CoV-2 spike glycoprotein results in considerably higher SASP synthesis and expression of the viral entrance genes ACE2 and transmembrane protease serine 2 (TMPRSS2) (Camell et al., 2021).
Some reduced responses are seen in the neutrophils of older people, including reduced bactericidal activity, reduced respiratory burst, and reduced neutrophil extracellular trap formation (Ortmann and Kolaczkowska, 2018). However, aberrant colonization and increased degranulation are also seen in the neutrophils of these people, which shows the defect of the neutrophils of the elderly in fighting pathogens and producing much inflammatory (Sapey et al., 2014). On the other hand, high levels of IL-6, which is caused by a pathogen or inflamm-ageing, lead to the long-term duration of neutrophils by decreasing apoptosis and indicate poor clinical outputs (Asensi et al., 2004, Liu et al., 2020a).
Therefore, the function of innate immune cells and inflammatory response to SARS-CoV-2 are not regulated in the elderly. Studies have shown a relationship between neutrophils and monocytes in the blood and severe COVID-19 (Kuri-Cervantes et al., 2020, Lucas et al., 2020). Based on the information, it can be concluded that these changes are partly due to the activation of age-related vital TLR, variations in oxidative stress pathways, inflammatory activity, raised cellular module and SASP, as well as other pathways like decreased autophagy and DNA damage with age (Bajaj et al., 2021a). Therefore, it can be concluded that SARS-CoV-2 infection, along with inflammation, worsens the consequences of COVID-19, and further studies are needed to eliminate the relationship between these two options.
3.3 Metabolites reprogramming
Numerous investigations have shown the point of creating between intracellular metabolism and inflammation. The field of immunometabolism concentrates on the shifts in intracellular metabolism that follow immune cell activation and regulate immune cell activities. Immune cells undergo substantial metabolite reprogramming during activation to meet the dramatic demand for energy and sustain immune cell capabilities such as strong cytokine generation, fast proliferation, and migratory processes (O'Neill et al., 2016). The information required to establish the metabolic effects of COVID-19 infection is currently being compiled. A number of studies have investigated the metabolic profiles of COVID-19 patients (Gardinassi et al., 2020).
Based on the most recent studies conducted on the effect of COVID-19 on metabolism, infected patients may have elevated blood glucose and fatty acid levels, as well as aberrations in amino acid (AA) metabolism. The gene expression encoding tryptophan metabolic enzymes, such as kynurenine and indoleamine 2, 3 dioxygenase (IDO), has been shown to have increased. (Blasco et al., 2020, Thomas et al., 2020b). The elevated amount of these metabolites in the circulation of severe COVID-19 patients has been linked to a reduction in tryptophan (Chen and Guillemin, 2009, Guillemin et al., 2003). Due to the decreased supply of this essential AA that activates general control non-derepressible-2 (GCN-2) or eukaryotic initiation factor-2 (eIF-2) stress kinase, tryptophan catabolism inhibits T cell proliferation and stimulates their anergy in severe COVID-19 patients, as illustrated by the reduces in various circulating T cells (Mellor and Munn, 2003, Moffett and Namboodiri, 2003, Munn et al., 2005). The elevated levels of kynurenic acid or kynurenate in the circulating suppress the pro-inflammatory action of monocytes and macrophages (Moroni et al., 2007, Sekkaï et al., 1997). For instance, in patients with severe COVID-19, circulating monocytes and macrophages do not contribute to the formation of the cytokine storm. Furthermore, tryptophan depletion in circulating DCs, monocytes, and macrophages increases GCN-2 activation, which boosts the production of anti-inflammatory cytokines (IL-10 and TGF-β) by phosphorylating eIF-2 at serine 51 and rendering it passive (Munn et al., 2005, Ravishankar et al., 2015, Sorgdrager et al., 2019). As a consequence of modified immunometabolism, circulating tryptophan metabolites inhibit circulating monocytes/macrophages, DCs, and T cells in severe COVID-19 patients.
Further intermediates involved in the metabolism of arginine, aspartate, tyrosine, and lysine may be changed in infected people (Blasco et al., 2020, Moolamalla et al., 2021, Thomas et al., 2020b). Serum from severe COVID-19 individuals demonstrates a reduction in apolipoprotein A1 (APOA1) and apolipoprotein M (APOM) (Shen et al., 2020). Patients with COVID-19 experience downregulation of sphingolipids as well. Glycerophospholipid levels in the serum are continuously reduced by SARS-CoV-2 infection and reach dangerously low levels in COVID-19 patients. In individuals with severe COVID-19, choline and its derivatives declined as well, although phosphocholine levels increased (Shen et al., 2020). These case studies demonstrate that many aspects of healthy metabolism can be disturbed as a result of sickness, making it difficult for any therapy used to restore metabolic equilibrium.
Sphingosine-1-phosphate (S1P) levels in plasma drop in COVID-19 patients but the rise in those who are convalescing (Song et al., 2020). The largest concentrations of S1P are seen in human red blood cells and platelets because these cells lack the pyridoxal phosphate-dependent S1P lyase (S1PL) needed to break down S1P into 2 hexadecenal and ethanolamine phosphate (Ito et al., 2007, Saba and Hla, 2004). So, under typical circumstances, hematopoietic and vascular endothelial cells expressing the ABC and spinster homolog 2 (Spns2) transporters carry S1P out of these cells and into the circulation (Fukuhara et al., 2012, Kim et al., 2009, Venkataraman et al., 2008). Under homeostasis, the mature T cell and B cell egress from the thymus and bone marrow to the circulation and the secondary lymphoid organs (SLOs) is promoted by the circulating S1P produced from vascular endothelial cells (Fukuhara et al., 2012, Nijnik et al., 2012). Although animals missing the Spns2 transporter exhibit a significant reduction in S1P in the lymph, the loss of circulating lymphocytes is caused by a slight reduction in S1P in the plasma (Mendoza et al., 2012). Therefore, a decline in the level of circulating S1P in COVID-19 patients reflects a reduction in the number of T cells in the blood, but an increase in the inflamed organ, which is more pronounced in patients with severe COVID-19. SARS-CoV2 infection reduces the level of circulating S1P in COVID-19 patients by infecting RBCs, platelets (thrombocytopenia), and vascular endothelial cells (Cavezzi et al., 2020, Lippi et al., 2020, Thomas et al., 2020a, Xu et al., 2020a).
In COVID-19 patients, levels of several acylcarnitines, including palmitoylcarnitine, stearoylcarnitine, and oleoylcarnitine, fell. Reduced levels of acylcarnitines in the blood may signify that the entry of fatty acids into the mitochondria for β-oxidation or fatty acid oxidation (FAO) has been reduced (Song et al., 2020). Citrate, succinate, and other tricarboxylic acids (TCAs) cycle metabolites typically decline in COVID-19 patients, and this decline increases with infection severity. Therefore, a diminished metabolic response to the decreased lung functions and blood oxygen to a lower dependency on oxygen for cellular energy production may be indicated by decreased TCA cycle metabolites in the circulation in severe COVID-19 patients (Song et al., 2020).
Increased cytosolic phospholipase A2 (cPLA2) activity in severe COVID-19 may worsen lung inflammation (Bhowmick et al., 2017, Song et al., 2020). The increase in PLA2 results in an increase in the levels of circulating glycerophospholipids (Song et al., 2020). In severe COVID-19 patients, there is a strong correlation between the rise in plasma sphingolipid GM3 enriched exosomes and the fall in T cell and CD4+ T cell counts (Song et al., 2020).
Inflammatory cytokines and chemokines are produced when acute macrophage activation occurs under inflammatory conditions (IFN- γ or lipopolysaccharide (LPS) stimulation). This is done by starting glycogenesis (glycogen synthesis), which produces glucose-6-phosphate for glycolysis to aggravate it further (Ma et al., 2020). In macrophages isolated from the lungs of severe COVID-19 patients, high glucose levels promote the multiplication of the SARS-CoV-2 and the production of pro-inflammatory cytokines, leading to cytokine storms (Codo et al., 2020). This happens as a result of the SARS-CoV-2 infection's overexpression of several glycolysis-related genes in these macrophages (Codo et al., 2020). Thus, only SARS-CoV-2 infection causes elevated glycolysis in human macrophages isolated from COVID-19 patients with severe lung disease. Compared to OXPHOS, which provides frequent energy for macrophages during homeostasis to transcribe and translate pro-inflammatory genes, glycolysis is induced in SPP1+ monocyte-derived macrophages that have become infiltrated in the lungs of severe COVID-19 patients. Additionally, these included increased transcriptional activity for enzymes such as lactate dehydrogenase A (LDH-A), pyruvate kinase M2, glucose transporter 1 (GLUT-1), 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3) (Bost et al., 2020, Codo et al., 2020). The elevation of the pro-inflammatory macrophage transcriptome for different cytokines and chemokines (TNF-, IL-6, IL-12A, IL-23A, and CCL3, CCL4, CCL5, CCL20, CCL23, CXCL2, CXCL11) isolated from sepsis patients is attributed to the HIF-1α overexpression (Kumar, 2018, Shalova et al., 2015). Immune cells' levels of LDH rise as COVID-19 severity increases. Even so, there is no discernible difference between COVID-19 patients and controls in the plasma lactate level (Song et al., 2020). The Krebs or TCA cycle intermediates (succinate and citrate) and the TCA cycle-derived itaconate accumulate as a result of enhanced glycolysis in pro-inflammatory macrophages, which also controls the expression of inflammatory genes (Ryan and O'Neill, 2020, Seim et al., 2019). The direct inhibition of prolyl hydroxylase domain (PHD) activity by succinate transfer to the cytosol stabilizes the HIF-1α and increases glycolysis (Codo et al., 2020, Ryan and O'Neill, 2020, Tannahill et al., 2013). Complex II, also known as succinate dehydrogenase (SDH), oxidizes succinate in the mitochondria, triggering reverse electron transport (RET), which encourages the production of mitochondrial ROS (mtROS), which suppresses PHD and stabilizes HIF-1α and glycolysis (Codo et al., 2020, Ryan and O'Neill, 2020, Tannahill et al., 2013). Thus, SARS-CoV-2 replication in infected macrophages depends critically on succinate oxidation.
3.4 The relationship between aging cells and thrombosis
In addition to the relationship between aging cells and the severity of inflammation in COVID-19, there is another link between them: thrombosis. Aging cells have a paracrine pro-coagulation effect (Wiley et al., 2019). In experiments performed on mice, infusion of a cellular aging reporter by doxorubicin remedy, p16–3MR, was related to notably minor bleeding time, higher platelet numbers, more activated platelets, and higher serum thrombopoietin (Wiley et al., 2019). The mission of aging cells in vivo turned these pro-thrombotic phenotypes. Stable isotope labeling by amino acids in cell culture analysis of aging cells displayed the further activity of human platelets stimulate by aging cell supernatants (Wiley et al., 2019).
Severe COVID-19 cases are related to agents such as venous thromboembolism and arterial thrombosis (Chen et al., 2020b, Cui et al., 2020, Fox et al., 2020, Guan et al., 2020, Klok et al., 2020, Lodigiani et al., 2020, Tang et al., 2020, Wu et al., 2020, Xu et al., 2020b, Zhou et al., 2020a). More precisely, factor CD142 (a blood coagulation stimulant) is expressed by proinflammatory cytokines on some of the immune cells. Supplement activity, neutrophil extracellular traps, and lung hypoxia further boost pro-thrombotic conditions, linking thrombosis and the innate immune system (Price et al., 2020). Generation of viral-induced pro-thrombotic auto-antibodies versus phospholipids and phospholipid-binding proteins is also seen within severe SARS-CoV-2 infection (Zuo et al., 2020). Recently, a study has shown that patients with critical COVID-19 pneumonia produce autoantibodies against type I INF. The response of extrafollicular B cell causes the production of these autoantibodies, which in turn cause thrombosis (Knight et al., 2021). Therefore, it can be concluded that the uncontrolled inflammatory response to SARS-CoV-2 in thrombotic-prone elderly can intensify the coagulation resulting from COVID-19.
3.5 Delayed response of type I IFN with ageing
IFN deficiency has been seen in a patient's serum with acute COVID-19 (Hadjadj et al., 2020). In addition, experiments on elderly macaques with SARS-CoV-2 infection also showed fewer type I IFN and Notch signaling pathways in the lung contrasted to young macaques (Rosa et al., 2021). The reduction in these responses includes a decrease in the number of plasmacytoid dendritic cells (pDCs), potent producer of IFN-α, that have also been seen in a patient with acute COVID-19 (Zhou et al., 2020b). It is noteworthy that only ten percent of patients with acute COVID-19 have type I anti-IFNs antibodies (Bastard et al., 2020). Dysfunction of genes associated with the TLR3 and interferon regulatory factor 7 (IRF7) pathways involved in inducing and amplifying type I IFNs response has also been observed in patients with acute infection (Zhang et al., 2020b). They can effectively prevent the spread of viral infection. They play a vital role in the optimal activity of macrophages, the presentation of antigens by DCs, and the growth of effective antiviral T cell responses (McNab et al., 2015).
Experiments have shown the critical effect of type I IFN responses on the adjustment of monocytes and neutrophils following SARS-CoV-2 infection. In experiments on patients with mild COVID-19, more classical monocytes were observed in the blood, indicating the early and passing effect of a type I IFN. Nevertheless, on the other hand, monocytes and neutrophils in the patient with acute COVID-19 showed more genes involved in NF-κB signaling and the generation of ROS or nitric oxide synthase during the disease. During viral infections, non-classical and intermediate proinflammatory subsets of monocytes develop (Wong et al., 2012). Therefore, they also spread during SARS-CoV-2 infection (Zhang et al., 2020a, Zhou et al., 2020c). Of course, the dysregulated strong response of these cells also helps the inflammatory conditions seen in the severe pathophysiology of COVID-19. In addition, non-classical monocytes also generate type I IFN (IFN-α) in response to TLR3 (Boyette et al., 2017). Therefore, the kinetics of non-classical monocyte dysregulation should be noted as severe consequences. Early monocyte deficiency decreases type I IFN response in patients with acute COVID-19 (Schulte-Schrepping et al., 2020, Silvin et al., 2020).
The aging process delays type I IFN responses, which has also been seen in the SARS-CoV infection. Although the effect of aging on susceptibility to viral infection is not well understood, it causes damage to the early and peripheral retinoic acid-inducible gene I (RIG-I) signaling pathways that rein the expression of many types of I IFN genes. This reduces the generation of I IFNs in people over 65 years of age and thus disrupts their body's antiviral responses (Molony et al., 2017). During SARS-CoV infection, the type I IFN signaling proteins downstream of RIG-I more decreases because of mitochondrial dysfunction associated with the mitochondrial antiviral signaling (MAVS) (Shi et al., 2014). In the elderly, factors such as basal mitochondrial dysfunction decreased TNF receptor-related agent adapter protein, and phosphorylated IRF3 amount associated with RIG-I signaling increases vulnerability to RIG-I deficiency (Feng et al., 2021, Molony et al., 2017). It should be noted that RIG-I can inhibit the proliferation of SARS-CoV-2 in human lung cells, although this function does not need I IFN signaling strength (Yamada et al., 2021). On the other hand, a decrease in the sum number of pDCs has been seen in the elderly. In addition, a decrease in TLR7 and I IFNs production ability has been observed in these elderlies (Feng et al., 2021, Jing et al., 2009, Pérez-Cabezas et al., 2007, Shodell and Siegal, 2002). Although there are many similarities between reduced IFN responses in patients with acute COVID-19 and the elderly, more research is needed to understand their relationship clearly.
3.6 Age-associated modifications to antigen presentation
With proper activation of T lymphocytes, innate immunity plays an essential role in generating adaptive immune responses. In order to efficiently prim T cells, innate immune cells have two functions: (1) to present antigen through the molecules of major histocompatibility complex (MHC) co-stimulatory receptors through a cell-cell interplay between T cells and antigen-presenting cells (APCs) and (2) generate appropriate cytokines to slant the separation of CD4+ T cells into a particular responsibility for the invasive pathogen. Disruption of each APC function has adverse effects on the adaptive immune responses. Like DCs and monocytes, APCs showed a shortage of antigen presentation in people with severe COVID-19. COVID-19-induced DC experiments showed the lowest expression of the chemokine receptors CD80, CD86, C-C chemokine receptor 7 (CCR7), and human leukocyte antigen (HLA)-DR (Giamarellos-Bourboulis et al., 2020, Schulte-Schrepping et al., 2020, Silvin et al., 2020, Zhou et al., 2020c). Disorders in antigen presentation occur with getting old. The number of monocytes also changes during this process, and the accumulation of non-classical monocytes during this period causes a downregulating of HLA-DR (Seidler et al., 2010). In mice, one study revealed a decrease in MHC class II, CD40, and CD86 levels in old DC subsets following activity by TLR agonists. Of course, there are contradictory observations, and further studies are needed (Wong and Goldstein, 2013). In infection of SARS-CoV, older mice lung DCs showed a ruined strength to immigrate to the draining lymph node, which hurt following T cell priming (Zhao et al., 2011). Raised amount of prostaglandin D2 in the lungs of older mice causes this immigration problem, which in turn leads to a decrease in CCR7 surface expression in DCs (Zhao et al., 2011). In turn, the aging factor disrupts T-cell antigen and antigen presentation, so it can be concluded that with viral infections that impair these functions, older people are more prone to dysfunctional adaptive responses of the immune system.
3.7 Dysregulation of B and T lymphocytes during ageing
3.7.1 B lymphocytes
Following the SARS-CoV-2 infection, B cells generate the amount of immunoglobulin M (IgM), IgG, and IgA antibodies for SARS-CoV-2 up to 1 week after symptoms and up to 2 weeks, most patients seroconverted to IgG and IgM. By identifying SARS-CoV-2-particular follicular T helper cells in the circulation, this response begins and plays a role in generating T cell-dependent antibodies. Counteracting antibodies of the SARS-CoV-2 receptor-binding domain (RBD) have been demonstrated in mice and humans, and translocation of these antibodies in experiments has decreased the severity of SARS-CoV-2 disease (Alsoussi et al., 2020, Zost et al., 2020). However, serum samples of a patient showing improvement have shown positive effects (Devarasetti et al., 2021). Determining the innate role of antibodies during COVID-19 requires further investigation.
The results of many experiments have shown a relationship between greater IgG antibody titers and disease severity (Moderbacher et al., 2020). The ambivalent role of antibodies within viral infections can clear the different results: Although counteracting antibodies are generally effective in combating and eliminating the viral agent, the initial production or pre-existence of pre-circulating non-counteracting antibodies can conduce to the antibody-mediated enhancement of viral entry and infuses severe inflammatory response. Here, an antibody-related issue exacerbates the disease. However, the exact role of antibodies in exacerbating SARS-CoV-2 infection has not yet been established. In addition, antibodies such as IgG with afucosylated IgG Fc series may exacerbate SARS-CoV-2 and further damage due to pro-inflammatory activity via FcγRIIIa (Larsen et al., 2021).
The effectiveness of the humoral immune response also decreases with aging, and older B cells show less power to sustain with physical mutation, which reduces the generation of counteracting antibody titers in the elderly (Frasca et al., 2017). Experiments on animals have shown that the IgG amount of viral S, protein-particular at the onset of acute SARS-CoV-2 infection, is less in older macaques than in young macaques; therefore, the establishment of antibody titers against viral infection decreases with aging (Singh et al., 2021).
On the other hand, B cell accumulation acquires unique attributes with aging. Another study in mice has shown that TLR7 responses elevate these cells (Cancro, 2020). These cells reside in the late human memory segment and release inflammatory mediators such as TNF-α, IL-6, and IL-8, thus having a role in autoimmune disease and recently in COVID-19 (Cancro, 2020, Frasca, 2018, Woodruff et al., 2020). In severe COVID-19 patients, the development of such cells has been observed to induce CD11c and T-bet expression in a skewed extra-follicular B cell response (Tay et al., 2020).
Favorable to the predominant extra-follicular B cell response, experiments have confirmed defective germinal center (GC) responses in patients with acute SARS-CoV-2 infection in the secondary lymphatic organs. Decreased GC establishment has also been reported in the results of one of these experiments and occurs when there are many amounts of TNF-α in lymph node follicles. TNF-α production also raises during aging and may play a role in phenotype (Bruunsgaard et al., 2003, Fagiolo et al., 1993, Penninx et al., 2004). Other aging problems also affect GC establishment. The expression of CD40 ligand, a co-stimulatory molecule essential for well T-cell and B-cell interplay, reduces with the increasing age of CD4+ T cells. In addition, lymphopenia in the elderly and patients with acute COVID-19 could mention by reducing access to CD4+ T cells to employ with B cells. Since the likelihood of T-cell and B-cell interaction is considerably decreased, the loss of a clonal diversity of TCR and B-cell receptor (BCR) in the elderly exacerbates this problem. Lack of longanimity and the appearance of auto-antibodies are other difficulties associated with aging. According to research on SARS-CoV-2, the lethality of this agent, especially in men, has been linked to the existence of auto-antibodies particular to the type I IFNs (Bastard et al., 2020).
Although these auto-antibodies have been found at different ages, their existence has been seen more in those patients over 65 years old. Based on these findings, it can be concluded that the effect of aging on the production of these antibodies and other auto-antibodies that are effective in exacerbating COVID-19 should be carefully investigated. On the other hand, the production of auto-antibodies, especially in men, requires more research.
3.7.2 T lymphocytes
One week after SARS-CoV-2 infection, adaptive responses are activated, and CD4+ and CD8+ T cells appear (Moderbacher et al., 2020). T helper 1 (TH1) cells respond to SARS-CoV-2 by producing IFN-γ and IL-2, among other cytokines, including TNF-α (Braun et al., 2020, Grifoni et al., 2020, Kaneko et al., 2020, Moderbacher et al., 2020, Zhou et al., 2020c). It should be noted that raised chemokines like CXCL9 and CXCL10 in the blood of COVID-19 patients have a role in the recruiting and/or differentiating naive T cells into TH1 cells. Experiments have shown that CD4+ and CD8+ T cells have protective effects on former coronaviruses (CoVs) (Zhao et al., 2010). To date, the role of T cells' response to SARS-CoV-2 infection has been demonstrated, although cytokine storms in these cells lead to the dysfunctional activity of monocytes (Moderbacher et al., 2020, Nelde et al., 2021, Peng et al., 2020). Considering the role of the T cell in fighting SARS-CoV-2 infection and its changes in aging, it is necessary to investigate the effect of biological changes of this cell to fight against this infection.
3.7.2.1 Decrease in clonal lymphocyte
To have a strong immune system, it is necessary to have a varied set of T-cell receptors (TCRs), but with increasing age, the diversity of CD4+ and CD8+ naive TCR species decreases (Britanova et al., 2014). A detailed study of the factors affecting TCR variability in adults is still ongoing. The process of thymic shrinkage in old age prevents the expansion of new T cells in the elderly and reduces the TCR set versatility. Experimental evidence has shown that with the introduction of naive, newly produced T cells into the old T cell set, the effectiveness of these cells decreases, and they produce fewer efficient memory cells. Of course, the problem of these young T cells, particularly CD8+ T cells, is not the cause of all the problems. Modified homeostatic proliferation in humans and the silencing preservation of T cell inactivity stillness are essential for a stable, naive peripheral T cell compartment. These two agents play an essential role in maintaining the diversity of naive T cells during aging (Goronzy and Weyand, 2019).
On the other hand, chronic antigen motivation may lead to developing a polyclonal T cell memory repository that reduces the existence of naive polyclonal T cells. In severe COVID-19, less TCR variability was observed against SARS-CoV-2 epitopes (Nelde et al., 2021, Peng et al., 2020). Compared with patients with moderate COVID-19, patients with severe the disease have a weaker T cell response to the N-terminal segment of the SARS-CoV-2 S protein (Braun et al., 2020). Low T cell frequency is also directly related to the effects of severe COVID-19 (Moderbacher et al., 2020). Therefore, the relationship between reduced TCR in the elderly and worsening consequences of SARS-CoV-2 infection can be understood.
3.7.2.2 Lymphopenia
Lymphopenia is one of the factors influencing the severity of COVID-19 and a decrease in the total number of peripheral T cells in the blood (Huang and Pranata, 2020). Experiments have shown the role of blood lymphocyte percentage as a criterion for diagnosing mild, severe disease and choosing the appropriate treatment process for the condition (Tan et al., 2020). Lymphopenia during COVID-19 can have a more adverse effect on the elderly and exacerbate poor T-cell responses by reducing the number of T cells. COVID-19-related lymphopenia resulted from increased T cell migration to infection sites. However, experiments have shown that lymphocyte migration is not the only factor affecting the lymphopenia related to COVID-19. In addition, a study has shown that the whole number of CD8+ T cells in the tissue of patients with mild COVID-19 is higher than in patients with acute symptoms (Liao et al., 2020).
Experimental findings confirming that T-cell immigration to the lungs alone does not cause blood lymphopenia in patients with severe COVID-19, as showed that lung macrophages in patients with severe COVID-19 mainly use chemokines to use inflammatory monocytes and neutrophils express, while lung macrophages in patient of mild COVID-19 express higher amount T cell-activating chemokines. Other factors that may reduce T cells in SARS-CoV-2 infection include the direct involvement of these cells in the virus or the cell death increase in response to antigen and the release of cytokines (Merad and Martin, 2020). In these experiments, dependence on the amount of cytokines IL-6, IL-10, or TNF-α and blood lymphopenia was seen in disease, and IL-6 receptor antagonist tocilizumab raised the issue of lymphocytes in the blood of COVID-19 patients (Diao et al., 2020, Giamarellos-Bourboulis et al., 2020, Wan et al., 2020). In addition, there is a direct relationship between lymphopenia and the age of the COVID-19 patient. Notably, in patients over 60, the whole number of blood T cells reaches its lowest point (Diao et al., 2020). Findings on the influenza virus have shown a decrease in the multiplication of old CD8 + T cells and its negative effect on the fight against the viral agent in older patients (Gruver et al., 2007).
3.7.2.3 T cell exhaustion
The effectiveness of T cell activity reduces with age, and on the other hand, the exhaustion of these cells also raises. So far, conflicting observations about the activity of these cells after SARS-CoV-2 infection have been seen. Weakness of essential activities of CD4+ T cells, such as production of IFN-γ, IL-2, and/or TNF-α, has been observed in patients with acute conditions. However, other experiments have not observed this weakness (Mazzoni et al., 2020, Zheng et al., 2020a). Decreased cytotoxicity of CD8+ and production of cytokines at CD8+ T cells activity in acute conditions have been observed, while the inverse of that has been observed in other experiments; it should be noted that in some experiments, no variation was seen in the function of these cells (Moderbacher et al., 2020, Peng et al., 2020, Zheng et al., 2020a). The variation in cytokine sampling time seems to be the reason for these different results. Due to the negative role of aging in the adaptive immune process, in our view, in the elderly, the effectiveness of T cell activity decreases and leads to a decrease in cytokine production and/or cytotoxicity, resulting in a weak fight against the virus disease. Experiments have identified the existence of IFN-γ-producing CD8+ T cells in the acute stages of the disease as a sign of the moderate COVID-19 effects (Moderbacher et al., 2020). In addition, in acute cases of COVID-19, higher expression of exhaustion indicators by CD4+ and CD8+ T cells has been observed (Diao et al., 2020). Recent findings have indicated the upregulation of exhaustion indicators such as programmed cell death protein-1 (PD-1) and T-cell immunoglobulin mucin 3 (Tim-3) by SARS-CoV-2 infection (Diao et al., 2020, Wu et al., 2021). However, these indicators may be signs of activation, not function exhaustion (Rha et al., 2021).
3.7.2.4 Pre-existing SARS-CoV-2 epitopes in unexposed individuals
CD4+ T cells that fight SARS-CoV-2 infection have already been seen in people who did not expose to the disease (Grifoni et al., 2020). In vitro studies have shown that T cells that fight SARS-CoV-2 infection, which already exists in the body, are mainly from the memory T cells (Mateus et al., 2020). Thus, many people have memory T cells that can cross-recognize SARS-CoV-2 epitopes (Braun et al., 2020). Other experiments have suggested a role for the absence of these cells in the TCR set in the exacerbation of COVID-19 (Nelde et al., 2021). The results of these experiments still need further investigation. For example, it should be investigated to what extent older adults have been at risk for CoV infections and whether this led to memory T cell responses. The absence of TCR diversity as we age can prevent memory T cells from proliferating in the previous face of CoVs. It may be due to inflating memory that memory T cells of the elderly may not be well stored in the peripheral repertoire (Klenerman and Oxenius, 2016).
A correlation has been found between the severity of COVID-19 and the occurrence of cytomegalovirus (CMV), an early pathogen related to memory inflation. However, more research is needed on the impact of formation, function, and longevity of new T cells and the CMV and other latent viruses on each other (Shrock et al., 2020). CD8+ T cells may also become active as aging and may be isolated without the presence of their homogeneous antigen and in reacting to IL-15 signaling. Although the function of CD8+ T cells has been extensively investigated in mice, it has been proven to be directly related to the age of humans (White et al., 2016). Cytokine-activated CD8+ T cells have an instinctive function and can induce cytotoxic effects without the presence of homogeneous antigen within viral infections. Aside from the antiviral benefits of this, an unrestrained function can cause harm to the host (Kim and Shin, 2019). The exact role of these cells in the effects of COVID-19 requires investigation.
4 Conclusions
Abnormal pathological inflammatory responses cause devastating consequences of SARS-CoV-2 infection and cause uncontrolled local tissue damage, vascular leakage, systemic cytokine storm, and thrombosis. NF-κB signaling causes inflammation and thus leads to an inadequate response to immune cells to severe activation, although monocyte subsets are an exception to this rule. Loss of early type I IFN responses during acute COVID-19, together with JAK-STAT hypo-responsiveness of old immune cells, leads to rapid viral duplication and exacerbates vulnerability to SARS-CoV-2 attack in the elderly. When the aging immune system overcomes the initial signaling and peak viral attack, many pro-inflammatory cytokines are released, which may cause tissue damage and vascular permeability, leading to the continued release of innate inflammatory pathological responses. Excess priming of TLR4 due to secondary bacterial infections after the viral attack is also involved in creating this peak. It should be noted that weak immune responses of the elderly to SARS-CoV-2 can be due to ineffective T cell priming, loss of naive T cell diversity, decreased antibody maturity, and/or impaired memory in these individuals. With its unregulated activity and without the aid of an antigen-specific immune response, the innate immune system exacerbates the complications of COVID-19.
The role of aging in the worsening of SARS-CoV-2 symptoms is one of the hottest topics. Most experiments have investigated the role of aging in people's blood factors, and there is still no substantiated research on tissue information. In the information obtained from the study of the immune system of mice and the role of aging in it, differences have been observed from the information obtained from human experiments. In addition, immune cell subsets are also not always fully maintained between these two experimental species. These contradictions necessitate careful consideration from one model to another. The investigations performed should integrate all the information obtained and provide comprehensive results.
Platforms of mRNA COVID-19 vaccines have proven highly effective in the elderly. The Pfizer vaccine has a success rate of 95% in people over 65 years old, and the Moderna vaccine has a success rate of 86% in this age group. Recently, 17 days after the second dose injection, an experiment showed fewer anti-SARS-CoV-2 antibodies in people over 80 than in those under 60. Determining the lasting efficacy of vaccines provided to the elderly requires further investigation. The long-term effects of SARS-CoV-2 infection have not been studied. The development of a strong immune system against COVID-19 is the main focus of today's research, which continues in the hope of increasing the health age of individuals.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
CRediT authorship contribution statement
Mohammad Reza Zinatizadeh: Conceptualization, Writing – original draft preparation, Data curation. Peyman Kheirandish Zarandi: Writing – original draft preparation, Data curation. Mohsen Ghiasi: Data curation and figure. Hamid Kooshki: Writing – review & editing. Mozafar Mohammadi: Writing – review & editing. Jafar Amani: Writing – review & editing. Nima Rezaei: Project administration, Supervision, Conceptualization, Writing – review & editing. All authors contributed to the paper and approved the submitted version.
Conflict of interest
The authors report no conflicts of interest in this work.
Acknowledgments
None declared.
==== Refs
References
Akha A.A.S. Aging and the immune system: an overview J. Immunol. Methods 463 2018 21 26 30114401
Al-Mosawi A.J. The pattern of covid-19 disease in Iraq during the year 2020 Sch. Int J. Anat. Physiol. 4 2021 127 134
Alpert A. Pickman Y. Leipold M. Rosenberg-Hasson Y. Ji X. Gaujoux R. Rabani H. Starosvetsky E. Kveler K. Schaffert S. A clinically meaningful metric of immune age derived from high-dimensional longitudinal monitoring Nat. Med. 25 2019 487 495 30842675
Alsoussi W.B. Turner J.S. Case J.B. Zhao H. Schmitz A.J. Zhou J.Q. Chen R.E. Lei T. Rizk A.A. McIntire K.M. A potently neutralizing antibody protects mice against SARS-CoV-2 infection J. Immunol. 205 2020 915 922 32591393
Asensi V. Valle E. Meana A. Fierer J. Celada A. Alvarez V. Paz J. Coto E. Carton J.A. Maradona J.A. In vivo interleukin-6 protects neutrophils from apoptosis in osteomyelitis Infect. Immun. 72 2004 3823 3828 15213123
Bajaj V. Gadi N. Spihlman A.P. Wu S.C. Choi C.H. Moulton V.R. Aging, immunity, and COVID-19: how age influences the host immune response to coronavirus infections? Front. Physiol. 2021 1793
Bajaj V. Gadi N. Spihlman A.P. Wu S.C. Choi C.H. Moulton V.R. Aging, immunity, and COVID-19: how age influences the host immune response to coronavirus infections? Front. Physiol. 11 2021 571416
Bajgain K.T. Badal S. Bajgain B.B. Santana M.J. Prevalence of comorbidities among individuals with COVID-19: A rapid review of current literature Am. J. Infect. Control 49 2021 238 246 32659414
Banu N. Panikar S.S. Leal L.R. Leal A.R. Protective role of ACE2 and its downregulation in SARS-CoV-2 infection leading to macrophage activation syndrome: therapeutic implications Life Sci. 256 2020 117905
Barkas F. Filippas-Ntekouan S. Kosmidou M. Liberopoulos E. Liontos A. Milionis H. Anakinra in hospitalized non-intubated patients with coronavirus disease 2019: a systematic review and meta-analysis Rheumatology 60 2021 5527 5537 33999135
Bastard P. Rosen L.B. Zhang Q. Michailidis E. Hoffmann H.-H. Zhang Y. Dorgham K. Philippot Q. Rosain J. Béziat V. Autoantibodies against type I IFNs in patients with life-threatening COVID-19 Science 370 2020 eabd4585
Bhowmick R. Clark S. Bonventre J.V. Leong J.M. McCormick B.A. Cytosolic phospholipase A2α promotes pulmonary inflammation and systemic disease during Streptococcus pneumoniae infection Infect. Immun. 85 2017 e00280 00217 28808157
Blasco H. Bessy C. Plantier L. Lefevre A. Piver E. Bernard L. Marlet J. Stefic K. Benz-de Bretagne I. Cannet P. The specific metabolome profiling of patients infected by SARS-COV-2 supports the key role of tryptophan-nicotinamide pathway and cytosine metabolism Sci. Rep. 10 2020 1 12 31913322
Boraschi D. Italiani P. Immunosenescence and vaccine failure in the elderly: strategies for improving response Immunol. Lett. 162 2014 346 353 24960535
Bost P. Giladi A. Liu Y. Bendjelal Y. Xu G. David E. Blecher-Gonen R. Cohen M. Medaglia C. Li H. Host-viral infection maps reveal signatures of severe COVID-19 patients Cell 181 2020 1475 1488 e1412 32479746
Boyette L.B. Macedo C. Hadi K. Elinoff B.D. Walters J.T. Ramaswami B. Chalasani G. Taboas J.M. Lakkis F.G. Metes D.M. Phenotype, function, and differentiation potential of human monocyte subsets PLoS One 12 2017 e0176460
Brasier A.R. The nuclear factor-κB–interleukin-6 signalling pathway mediating vascular inflammation Cardiovasc. Res. 86 2010 211 218 20202975
Braun J. Loyal L. Frentsch M. Wendisch D. Georg P. Kurth F. Hippenstiel S. Dingeldey M. Kruse B. Fauchere F. SARS-CoV-2-reactive T cells in healthy donors and patients with COVID-19 Nature 587 2020 270 274 32726801
Britanova O.V. Putintseva E.V. Shugay M. Merzlyak E.M. Turchaninova M.A. Staroverov D.B. Bolotin D.A. Lukyanov S. Bogdanova E.A. Mamedov I.Z. Age-related decrease in TCR repertoire diversity measured with deep and normalized sequence profiling J. Immunol. 192 2014 2689 2698 24510963
Bruunsgaard H. Andersen-Ranberg K. vB Hjelmborg J. Pedersen B.K. Jeune B. Elevated levels of tumor necrosis factor alpha and mortality in centenarians Am. J. Med. 115 2003 278 283 12967692
Camell C.D. Yousefzadeh M.J. Zhu Y. Prata L.G.L. Huggins M.A. Pierson M. Zhang L. O’Kelly R.D. Pirtskhalava T. Xun P. Senolytics reduce coronavirus-related mortality in old mice Science 373 2021 eabe4832
Cancro M.P. Age-associated B cells Annu. Rev. Immunol. 38 2020 315 340 31986068
Cavezzi A. Troiani E. Corrao S. COVID-19: hemoglobin, iron, and hypoxia beyond inflammation. A narrative review Clin. Pract. 10 2020 1271 32509258
Chen J. Jiang Q. Xia X. Liu K. Yu Z. Tao W. Gong W. Han J.D.J. Individual variation of the SARS‐CoV‐2 receptor ACE2 gene expression and regulation Aging Cell 19 2020 e13168
Chen N. Zhou M. Dong X. Qu J. Gong F. Han Y. Qiu Y. Wang J. Liu Y. Wei Y. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study Lancet 395 2020 507 513 32007143
Chen Y. Guillemin G.J. Kynurenine pathway metabolites in humans: disease and healthy states. International journal of tryptophan research 2 IJTR 2009 S2097
Chougnet C.A. Thacker R.I. Shehata H.M. Hennies C.M. Lehn M.A. Lages C.S. Janssen E.M. Loss of phagocytic and antigen cross-presenting capacity in aging dendritic cells is associated with mitochondrial dysfunction J. Immunol. 195 2015 2624 2632 26246142
Codo A.C. Davanzo G.G. de Brito Monteiro L. de Souza G.F. Muraro S.P. Virgilio-da-Silva J.V. Prodonoff J.S. Carregari V.C. de Biagi Junior C.A.O. Crunfli F. Elevated glucose levels favor SARS-CoV-2 infection and monocyte response through a HIF-1α/glycolysis-dependent axis Cell Metab. 32 2020 437 446 e435 32697943
Coppé J.-P. Desprez P.-Y. Krtolica A. Campisi J. The senescence-associated secretory phenotype: the dark side of tumor suppression Annu. Rev. Pathol.: Mech. Dis. 5 2010 99 118
Cui S. Chen S. Li X. Liu S. Wang F. Prevalence of venous thromboembolism in patients with severe novel coronavirus pneumonia J. Thromb. Haemost. 18 2020 1421 1424 32271988
Cumberbatch M. Dearman R.J. Kimber I. Influence of ageing on Langerhans cell migration in mice: identification of a putative deficiency of epidermal interleukin‐1β Immunology 105 2002 466 477 11985667
Cunha L.L. Perazzio S.F. Azzi J. Cravedi P. Riella L.V. Remodeling of the immune response with aging: immunosenescence and its potential impact on COVID-19 immune response Front. Immunol. 2020 1748 32849623
Derhovanessian E. Solana R. Larbi A. Pawelec G. Immunity, ageing and cancer Immun. Ageing 5 2008 1 16 18267021
Devarasetti P.K. Rajasekhar L. Baisya R. Sreejitha K. Vardhan Y.K. A review of COVID-19 convalescent plasma use in COVID-19 with focus on proof of efficacy Immunol. Res. 69 2021 18 25 33492637
Diao B. Wang C. Tan Y. Chen X. Liu Y. Ning L. Chen L. Li M. Liu Y. Wang G. Reduction and functional exhaustion of T cells in patients with coronavirus disease 2019 (COVID-19) Front. Immunol. 2020 827 32425950
Dinnon K.H. Leist S.R. Schäfer A. Edwards C.E. Martinez D.R. Montgomery S.A. West A. Yount B.L. Hou Y.J. Adams L.E. A mouse-adapted model of SARS-CoV-2 to test COVID-19 countermeasures Nature 586 2020 560 566 32854108
Fagiolo U. Cossarizza A. Scala E. Fanales‐Belasio E. Ortolani C. Cozzi E. Monti D. Franceschi C. Paganelli R. Increased cytokine production in mononuclear cells of healthy elderly people Eur. J. Immunol. 23 1993 2375 2378 8370415
Feng E. Balint E. Poznanski S.M. Ashkar A.A. Loeb M. Aging and interferons: impacts on inflammation and viral disease outcomes Cells 10 2021 708 33806810
Ferrucci L. Fabbri E. Inflammageing: chronic inflammation in ageing, cardiovascular disease, and frailty Nat. Rev. Cardiol. 15 2018 505 522 30065258
Fourati S. Cristescu R. Loboda A. Talla A. Filali A. Railkar R. Schaeffer A.K. Favre D. Gagnon D. Peretz Y. Pre-vaccination inflammation and B-cell signalling predict age-related hyporesponse to hepatitis B vaccination Nat. Commun. 7 2016 1 12
Fox S.E. Akmatbekov A. Harbert J.L. Li G. Brown J.Q. Vander Heide R.S. Pulmonary and cardiac pathology in African American patients with COVID-19: an autopsy series from New Orleans Lancet Respir. Med. 8 2020 681 686 32473124
Frasca D. Senescent B cells in aging and age-related diseases: their role in the regulation of antibody responses Exp. Gerontol. 107 2018 55 58 28687479
Frasca D. Diaz A. Romero M. Blomberg B.B. Human peripheral late/exhausted memory B cells express a senescent-associated secretory phenotype and preferentially utilize metabolic signaling pathways Exp. Gerontol. 87 2017 113 120 27931848
Fukuhara S. Simmons S. Kawamura S. Inoue A. Orba Y. Tokudome T. Sunden Y. Arai Y. Moriwaki K. Ishida J. The sphingosine-1-phosphate transporter Spns2 expressed on endothelial cells regulates lymphocyte trafficking in mice J. Clin. Investig. 122 2012 1416 1426 22406534
Furman D. Campisi J. Verdin E. Carrera-Bastos P. Targ S. Franceschi C. Ferrucci L. Gilroy D.W. Fasano A. Miller G.W. Chronic inflammation in the etiology of disease across the life span Nat. Med. 25 2019 1822 1832 31806905
Gardinassi L.G. Souza C.O. Sales-Campos H. Fonseca S.G. Immune and metabolic signatures of COVID-19 revealed by transcriptomics data reuse Front. Immunol. 11 2020 1636 32670298
Giamarellos-Bourboulis E.J. Netea M.G. Rovina N. Akinosoglou K. Antoniadou A. Antonakos N. Damoraki G. Gkavogianni T. Adami M.-E. Katsaounou P. Complex immune dysregulation in COVID-19 patients with severe respiratory failure Cell Host Microbe 27 2020 992 1000 e1003 32320677
Gold J.A. Rossen L.M. Ahmad F.B. Sutton P. Li Z. Salvatore P.P. Coyle J.P. DeCuir J. Baack B.N. Durant T.M. Race, ethnicity, and age trends in persons who died from COVID-19—United States, May–August 2020 Morb. Mortal. Wkly. Rep. 2020 1517
Goodwin K. Viboud C. Simonsen L. Antibody response to influenza vaccination in the elderly: a quantitative review Vaccine 24 2006 1159 1169 16213065
Goronzy J.J. Weyand C.M. Mechanisms underlying T cell ageing Nat. Rev. Immunol. 19 2019 573 583 31186548
Grifoni A. Weiskopf D. Ramirez S.I. Mateus J. Dan J.M. Moderbacher C.R. Rawlings S.A. Sutherland A. Premkumar L. Jadi R.S. Targets of T cell responses to SARS-CoV-2 coronavirus in humans with COVID-19 disease and unexposed individuals Cell 181 1489–1501 2020 e1415
Gruver A. Hudson L. Sempowski G. Immunosenescence of ageing J. Pathol.: A J. Pathol. Soc. Gt. Br. Irel. 211 2007 144 156
Guan W.-j Ni Z.-y Hu Y. Liang W.-h Ou C.-q He J.-x Liu L. Shan H. Lei C.-l Hui D.S. Clinical characteristics of coronavirus disease 2019 in China New Engl. J. Med. 382 2020 1708 1720 32109013
Guaraldi G. Meschiari M. Cozzi-Lepri A. Milic J. Tonelli R. Menozzi M. Franceschini E. Cuomo G. Orlando G. Borghi V. Tocilizumab in patients with severe COVID-19: a retrospective cohort study Lancet Rheumatol. 2 2020 e474 e484 32835257
Guillemin G.J. Smith D.G. Smythe G.A. Armati P.J. Brew G.J. Expression of the kynurenine pathway enzymes in human microglia and macrophages Developments in Tryptophan and Serotonin Metabolism 2003 Springer 105 112
Hadjadj J. Yatim N. Barnabei L. Corneau A. Boussier J. Smith N. Péré H. Charbit B. Bondet V. Chenevier-Gobeaux C. Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients Science 369 2020 718 724 32661059
Hainz U. Jenewein B. Asch E. Pfeiffer K.-P. Berger P. Grubeck-Loebenstein B. Insufficient protection for healthy elderly adults by tetanus and TBE vaccines Vaccine 23 2005 3232 3235 15837226
Ho F.K. Petermann-Rocha F. Gray S.R. Jani B.D. Katikireddi S.V. Niedzwiedz C.L. Foster H. Hastie C.E. Mackay D.F. Gill J.M. Is older age associated with COVID-19 mortality in the absence of other risk factors? General population cohort study of 470,034 participants PLoS One 15 2020 e0241824
Huang I. Pranata R. Lymphopenia in severe coronavirus disease-2019 (COVID-19): systematic review and meta-analysis J. Intensive Care 8 2020 1 10 31908779
Ibrahim M.E. Al-Aklobi O.S. Abomughaid M.M. Al-Ghamdi M.A. Epidemiological, clinical, and laboratory findings for patients of different age groups with confirmed coronavirus disease 2019 (COVID-19) in a hospital in Saudi Arabia PLoS One 16 2021 e0250955
Imai Y. Kuba K. Rao S. Huan Y. Guo F. Guan B. Yang P. Sarao R. Wada T. Leong-Poi H. Angiotensin-converting enzyme 2 protects from severe acute lung failure Nature 436 2005 112 116 16001071
Ito K. Anada Y. Tani M. Ikeda M. Sano T. Kihara A. Igarashi Y. Lack of sphingosine 1-phosphate-degrading enzymes in erythrocytes Biochem. Biophys. Res. Commun. 357 2007 212 217 17418101
Jing Y. Shaheen E. Drake R.R. Chen N. Gravenstein S. Deng Y. Aging is associated with a numerical and functional decline in plasmacytoid dendritic cells, whereas myeloid dendritic cells are relatively unaltered in human peripheral blood Hum. Immunol. 70 2009 777 784 19596035
Johnson T.E. Recent results: biomarkers of aging Exp. Gerontol. 41 2006 1243 1246 17071038
Junqueira C. Crespo Â. Ranjbar S. Lewandrowski M. Ingber J. de Lacerda L.B. Parry B. Ravid S. Clark S. Ho F. SARS-CoV-2 Infects Blood Monocytes to Activate NLRP3 and AIM2 Inflammasomes, Pyroptosis and Cytokine Release 2021 Research Square
Kaml M. Weiskirchner I. Keller M. Luft T. Hoster E. Hasford J. Young L. Bartlett B. Neuner C. Fischer K.-H. Booster vaccination in the elderly: their success depends on the vaccine type applied earlier in life as well as on pre-vaccination antibody titers Vaccine 24 2006 6808 6811 16872725
Kaneko N. Kuo H.-H. Boucau J. Farmer J.R. Allard-Chamard H. Mahajan V.S. Piechocka-Trocha A. Lefteri K. Osborn M. Bals J. Loss of Bcl-6-expressing T follicular helper cells and germinal centers in COVID-19 Cell 183 2020 143 157 e113 32877699
Kim R.H. Takabe K. Milstien S. Spiegel S. Export and functions of sphingosine-1-phosphate Biochim. Et Biophys. Acta (BBA)-Mol. Cell Biol. Lipids 1791 2009 692 696
Kim T.-S. Shin E.-C. The activation of bystander CD8+ T cells and their roles in viral infection Exp. Mol. Med. 51 1–9 2019
Klenerman P. Oxenius A. T cell responses to cytomegalovirus Nat. Rev. Immunol. 16 2016 367 377 27108521
Klok F. Kruip M. Van der Meer N. Arbous M. Gommers D. Kant K. Kaptein F. van Paassen J. Stals M. Huisman M. Incidence of thrombotic complications in critically ill ICU patients with COVID-19 Thromb. Res. 191 2020 145 147 32291094
Knight J.S. Caricchio R. Casanova J.-L. Combes A.J. Diamond B. Fox S.E. Hanauer D.A. James J.A. Kanthi Y. Ladd V. The intersection of COVID-19 and autoimmunity J. Clin. Investig. 2021 131
Kumar V. Targeting macrophage immunometabolism: dawn in the darkness of sepsis Int. Immunopharmacol. 58 2018 173 185 29625385
Kuri-Cervantes L. Pampena M.B. Meng W. Rosenfeld A.M. Ittner C.A. Weisman A.R. Agyekum R.S. Mathew D. Baxter A.E. Vella L.A. Comprehensive mapping of immune perturbations associated with severe COVID-19 Sci. Immunol. 5 2020 eabd7114
Larsen M.D. de Graaf E.L. Sonneveld M.E. Plomp H.R. Nouta J. Hoepel W. Chen H.-J. Linty F. Visser R. Brinkhaus M. Afucosylated IgG characterizes enveloped viral responses and correlates with COVID-19 severity Science 371 2021 eabc8378
Leist S.R. Dinnon K.H. III Schäfer A. Longping V.T. Okuda K. Hou Y.J. West A. Edwards C.E. Sanders W. Fritch E.J. A mouse-adapted SARS-CoV-2 induces acute lung injury and mortality in standard laboratory mice Cell 183 2020 1070 1085 e1012 33031744
Leung C. Risk factors for predicting mortality in elderly patients with COVID-19: a review of clinical data in China Mech. Ageing Dev. 188 2020 111255
Liao M. Liu Y. Yuan J. Wen Y. Xu G. Zhao J. Cheng L. Li J. Wang X. Wang F. Single-cell landscape of bronchoalveolar immune cells in patients with COVID-19 Nat. Med. 26 2020 842 844 32398875
Lippi G. Plebani M. Henry B.M. Thrombocytopenia is associated with severe coronavirus disease 2019 (COVID-19) infections: a meta-analysis Clin. Chim. Acta 506 2020 145 148 32178975
Liu Y. Du X. Chen J. Jin Y. Peng L. Wang H.H. Luo M. Chen L. Zhao Y. Neutrophil-to-lymphocyte ratio as an independent risk factor for mortality in hospitalized patients with COVID-19 J. Infect. 81 2020 e6 e12
Liu Y. Yang Y. Zhang C. Huang F. Wang F. Yuan J. Wang Z. Li J. Li J. Feng C. Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads and lung injury Sci. China Life Sci. 63 2020 364 374 32048163
Lodigiani C. Iapichino G. Carenzo L. Cecconi M. Ferrazzi P. Sebastian T. Kucher N. Studt J.-D. Sacco C. Bertuzzi A. Venous and arterial thromboembolic complications in COVID-19 patients admitted to an academic hospital in Milan, Italy Thromb. Res. 191 2020 9 14 32353746
Lucas C. Wong P. Klein J. Castro T.B. Silva J. Sundaram M. Ellingson M.K. Mao T. Oh J.E. Israelow B. Longitudinal analyses reveal immunological misfiring in severe COVID-19 Nature 584 2020 463 469 32717743
Ma J. Wei K. Liu J. Tang K. Zhang H. Zhu L. Chen J. Li F. Xu P. Liu J. Glycogen metabolism regulates macrophage-mediated acute inflammatory responses Nat. Commun. 11 2020 1 16 31911652
Machado C.J. Pereira C.Cd.A. Viana B.D.M. Oliveira G.L. Melo D.C. Carvalho J.F.M.G.D. Moraes F.L.D. Moraes E.N.D. Estimates of the impact of COVID-19 on mortality of institutionalized elderly in Brazil Cienc. Saude Coletiva 25 2020 3437 3444
Manser A.R. Uhrberg M. Age-related changes in natural killer cell repertoires: impact on NK cell function and immune surveillance Cancer Immunol. Immunother. 65 2016 417 426 26288343
Márquez E.J. Trowbridge J. Kuchel G.A. Banchereau J. Ucar D. The lethal sex gap: COVID-19 Immun. Ageing 17 2020 1 8 31911808
Mateus J. Grifoni A. Tarke A. Sidney J. Ramirez S.I. Dan J.M. Burger Z.C. Rawlings S.A. Smith D.M. Phillips E. Selective and cross-reactive SARS-CoV-2 T cell epitopes in unexposed humans Science 370 2020 89 94 32753554
Mazzoni A. Salvati L. Maggi L. Capone M. Vanni A. Spinicci M. Mencarini J. Caporale R. Peruzzi B. Antonelli A. Impaired immune cell cytotoxicity in severe COVID-19 is IL-6 dependent J. Clin. Investig. 130 2020 4694 4703 32463803
McDade T.W. Adair L. Feranil A.B. Kuzawa C. Positive antibody response to vaccination in adolescence predicts lower C‐reactive protein concentration in young adulthood in the Philippines Am. J. Hum. Biol. 23 2011 313 318 21484910
McNab F. Mayer-Barber K. Sher A. Wack A. O'garra A. Type I interferons in infectious disease Nat. Rev. Immunol. 15 2015 87 103 25614319
Medzhitov R. Schneider D.S. Soares M.P. Disease tolerance as a defense strategy Science 335 2012 936 941 22363001
Melegaro A. Edmunds W.J. The 23-valent pneumococcal polysaccharide vaccine. Part I. Efficacy of PPV in the elderly: a comparison of meta-analyses Eur. J. Epidemiol. 19 2004 353 363 15180106
Mellor A.L. Munn D.H. Tryptophan catabolism and regulation of adaptive immunity J. Immunol. 170 2003 5809 5813 12794104
Mendoza A. Bréart B. Ramos-Perez W.D. Pitt L.A. Gobert M. Sunkara M. Lafaille J.J. Morris A.J. Schwab S.R. The transporter Spns2 is required for secretion of lymph but not plasma sphingosine-1-phosphate Cell Rep. 2 2012 1104 1110 23103166
Merad M. Martin J.C. Pathological inflammation in patients with COVID-19: a key role for monocytes and macrophages Nat. Rev. Immunol. 20 2020 355 362 32376901
Metcalf T.U. Cubas R.A. Ghneim K. Cartwright M.J. Grevenynghe J.V. Richner J.M. Olagnier D.P. Wilkinson P.A. Cameron M.J. Park B.S. Global analyses revealed age‐related alterations in innate immune responses after stimulation of pathogen recognition receptors Aging Cell 14 2015 421 432 25728020
Metcalf T.U. Wilkinson P.A. Cameron M.J. Ghneim K. Chiang C. Wertheimer A.M. Hiscott J.B. Nikolich-Zugich J. Haddad E.K. Human monocyte subsets are transcriptionally and functionally altered in aging in response to pattern recognition receptor agonists J. Immunol. 199 2017 1405 1417 28696254
Mishra S.R. Mahapatra K.K. Behera B.P. Patra S. Bhol C.S. Panigrahi D.P. Praharaj P.P. Singh A. Patil S. Dhiman R. Mitochondrial dysfunction as a driver of NLRP3 inflammasome activation and its modulation through mitophagy for potential therapeutics Int. J. Biochem. Cell Biol. 136 2021 106013
Moderbacher C.R. Ramirez S.I. Dan J.M. Grifoni A. Hastie K.M. Weiskopf D. Belanger S. Abbott R.K. Kim C. Choi J. Antigen-specific adaptive immunity to SARS-CoV-2 in acute COVID-19 and associations with age and disease severity Cell 183 2020 996 1012 e1019 33010815
Moffett J.R. Namboodiri M.A. Tryptophan and the immune response Immunol. Cell Biol. 81 2003 247 265 12848846
Molony R.D. Nguyen J.T. Kong Y. Montgomery R.R. Shaw A.C. Iwasaki A. Aging impairs both primary and secondary RIG-I signaling for interferon induction in human monocytes Sci. Signal. 10 2017 eaan2392
Moolamalla S. Balasubramanian R. Chauhan R. Priyakumar U.D. Vinod P. Host metabolic reprogramming in response to SARS-CoV-2 infection: a systems biology approach Microb. Pathog. 158 2021 105114
Moroni F. Fossati S. Chiarugi A. Cozzi A. Kynurenic acid actions in brain and periphery International Congress Series 2007 Elsevier 305 313
Munn D.H. Sharma M.D. Baban B. Harding H.P. Zhang Y. Ron D. Mellor A.L. GCN2 kinase in T cells mediates proliferative arrest and anergy induction in response to indoleamine 2, 3-dioxygenase Immunity 22 2005 633 642 15894280
Nelde A. Bilich T. Heitmann J.S. Maringer Y. Salih H.R. Roerden M. Lübke M. Bauer J. Rieth J. Wacker M. SARS-CoV-2-derived peptides define heterologous and COVID-19-induced T cell recognition Nat. Immunol. 22 2021 74 85 32999467
Nijnik A. Clare S. Hale C. Chen J. Raisen C. Mottram L. Lucas M. Estabel J. Ryder E. Adissu H. The role of sphingosine-1-phosphate transporter Spns2 in immune system function J. Immunol. 189 2012 102 111 22664872
Nikolich-Zugich J. Knox K.S. Rios C.T. Natt B. Bhattacharya D. Fain M.J. SARS-CoV-2 and COVID-19 in older adults: what we may expect regarding pathogenesis, immune responses, and outcomes Geroscience 42 2020 505 514 32274617
Nikolich-Žugich J. The twilight of immunity: emerging concepts in aging of the immune system Nat. Immunol. 19 2018 10 19 29242543
Nikpouraghdam M. Farahani A.J. Alishiri G. Heydari S. Ebrahimnia M. Samadinia H. Sepandi M. Jafari N.J. Izadi M. Qazvini A. Epidemiological characteristics of coronavirus disease 2019 (COVID-19) patients in IRAN: a single center study J. Clin. Virol. 127 2020 104378
O'Neill L.A. Kishton R.J. Rathmell J. A guide to immunometabolism for immunologists Nat. Rev. Immunol. 16 2016 553 565 27396447
Ortmann W. Kolaczkowska E. Age is the work of art? Impact of neutrophil and organism age on neutrophil extracellular trap formation Cell Tissue Res. 371 2018 473 488 29250748
Peng Y. Mentzer A.J. Liu G. Yao X. Yin Z. Dong D. Dejnirattisai W. Rostron T. Supasa P. Liu C. Broad and strong memory CD4+ and CD8+ T cells induced by SARS-CoV-2 in UK convalescent individuals following COVID-19 Nat. Immunol. 21 2020 1336 1345 32887977
Penninx B.W. Kritchevsky S.B. Newman A.B. Nicklas B.J. Simonsick E.M. Rubin S. Nevitt M. Visser M. Harris T. Pahor M. Inflammatory markers and incident mobility limitation in the elderly J. Am. Geriatr. Soc. 52 2004 1105 1113 15209648
Pérez-Cabezas B. Naranjo-Gómez M. Fernández M.A. Grífols J.R. Pujol-Borrell R. Borràs F.E. Reduced numbers of plasmacytoid dendritic cells in aged blood donors Exp. Gerontol. 42 2007 1033 1038 17606348
Price L.C. McCabe C. Garfield B. Wort S.J. Thrombosis and COVID-19 Pneumonia: The Clot Thickens! 2020 Eur Respiratory Soc.
Qin C. Zhou L. Hu Z. Zhang S. Yang S. Tao Y. Xie C. Ma K. Shang K. Wang W. Dysregulation of immune response in patients with coronavirus 2019 (COVID-19) in Wuhan, China Clin. Infect. Dis. 71 2020 762 768 32161940
Ravishankar B. Liu H. Shinde R. Chaudhary K. Xiao W. Bradley J. Koritzinsky M. Madaio M.P. McGaha T.L. The amino acid sensor GCN2 inhibits inflammatory responses to apoptotic cells promoting tolerance and suppressing systemic autoimmunity Proc. Natl. Acad. Sci. U.S.A. 112 2015 10774 10779 26261340
Rea I.M. Gibson D.S. McGilligan V. McNerlan S.E. Alexander H.D. Ross O.A. Age and age-related diseases: role of inflammation triggers and cytokines Front. Immunol. 2018 586 29686666
Rha M.-S. Jeong H.W. Ko J.-H. Choi S.J. Seo I.-H. Lee J.S. Sa M. Kim A.R. Joo E.-J. Ahn J.Y. PD-1-expressing SARS-CoV-2-specific CD8+ T cells are not exhausted, but functional in patients with COVID-19 Immunity 54 2021 44 52 e43 33338412
Rodrigues T.S. de Sá K.S. Ishimoto A.Y. Becerra A. Oliveira S. Almeida L. Gonçalves A.V. Perucello D.B. Andrade W.A. Castro R. Inflammasome activation in COVID-19 patients medRxiv 2020
Rodrigues T.S. de Sá K.S. Ishimoto A.Y. Becerra A. Oliveira S. Almeida L. Gonçalves A.V. Perucello D.B. Andrade W.A. Castro R. Inflammasomes are activated in response to SARS-CoV-2 infection and are associated with COVID-19 severity in patients J. Exp. Med. 2021 218
Rosa B.A. Ahmed M. Singh D.K. Choreño-Parra J.A. Cole J. Jiménez-Álvarez L.A. Rodríguez-Reyna T.S. Singh B. Gonzalez O. Carrion R. IFN signaling and neutrophil degranulation transcriptional signatures are induced during SARS-CoV-2 infection Commun. Biol. 4 2021 1 14 33398033
Ryan D.G. O'Neill L.A. Krebs cycle reborn in macrophage immunometabolism Annu. Rev. Immunol. 38 2020 289 313 31986069
Saba J.D. Hla T. Point-counterpoint of sphingosine 1-phosphate metabolism Circ. Res. 94 2004 724 734 15059942
Salminen A. Huuskonen J. Ojala J. Kauppinen A. Kaarniranta K. Suuronen T. Activation of innate immunity system during aging: NF-kB signaling is the molecular culprit of inflamm-aging Ageing Res. Rev. 7 2008 83 105 17964225
Santesmasses D. Castro J.P. Zenin A.A. Shindyapina A.V. Gerashchenko M.V. Zhang B. Kerepesi C. Yim S.H. Fedichev P.O. Gladyshev V.N. COVID‐19 is an emergent disease of aging Aging Cell 19 2020 e13230
Sapey E. Greenwood H. Walton G. Mann E. Love A. Aaronson N. Insall R.H. Stockley R.A. Lord J.M. Phosphoinositide 3-kinase inhibition restores neutrophil accuracy in the elderly: toward targeted treatments for immunosenescence Blood J. Am. Soc. Hematol. 123 2014 239 248
Sayed N. Huang Y. Nguyen K. Krejciova-Rajaniemi Z. Grawe A.P. Gao T. Tibshirani R. Hastie T. Alpert A. Cui L. An inflammatory aging clock (iAge) based on deep learning tracks multimorbidity, immunosenescence, frailty and cardiovascular aging Nat. Aging 1 2021 598 615 34888528
Schulte-Schrepping J. Reusch N. Paclik D. Baßler K. Schlickeiser S. Zhang B. Krämer B. Krammer T. Brumhard S. Bonaguro L. Severe COVID-19 is marked by a dysregulated myeloid cell compartment Cell 182 2020 1419 1440 e1423 32810438
Seidler S. Zimmermann H.W. Bartneck M. Trautwein C. Tacke F. Age-dependent alterations of monocyte subsets and monocyte-related chemokine pathways in healthy adults BMC Immunol. 11 2010 1 11 20064252
Seim G.L. Britt E.C. John S.V. Yeo F.J. Johnson A.R. Eisenstein R.S. Pagliarini D.J. Fan J. Two-stage metabolic remodelling in macrophages in response to lipopolysaccharide and interferon-γ stimulation Nat. Metab. 1 2019 731 742 32259027
Sekkaï D. Guittet O. Lemaire G. Tenu J.-P. Lepoivre M. Inhibition of nitric oxide synthase expression and activity in macrophages by 3-hydroxyanthranilic acid, a tryptophan metabolite Arch. Biochem. Biophys. 340 1997 117 123 9126284
Shah A. Novel coronavirus-induced NLRP3 inflammasome activation: a potential drug target in the treatment of COVID-19 Front. Immunol. 11 2020 1021 32574259
Shalova I.N. Lim J.Y. Chittezhath M. Zinkernagel A.S. Beasley F. Hernández-Jiménez E. Toledano V. Cubillos-Zapata C. Rapisarda A. Chen J. Human monocytes undergo functional re-programming during sepsis mediated by hypoxia-inducible factor-1α Immunity 42 2015 484 498 25746953
Shen B. Yi X. Sun Y. Bi X. Du J. Zhang C. Quan S. Zhang F. Sun R. Qian L. Proteomic and metabolomic characterization of COVID-19 patient sera Cell 182 2020 59 72 e15 32492406
Shen-Orr S.S. Furman D. Kidd B.A. Hadad F. Lovelace P. Huang Y.-W. Rosenberg-Hasson Y. Mackey S. Grisar F.A.G. Pickman Y. Defective signaling in the JAK-STAT pathway tracks with chronic inflammation and cardiovascular risk in aging humans Cell Syst. 3 2016 374 384 e374 27746093
Shi C.-S. Qi H.-Y. Boularan C. Huang N.-N. Abu-Asab M. Shelhamer J.H. Kehrl J.H. SARS-coronavirus open reading frame-9b suppresses innate immunity by targeting mitochondria and the MAVS/TRAF3/TRAF6 signalosome J. Immunol. 193 2014 3080 3089 25135833
Shodell M. Siegal F. Circulating, interferon‐producing plasmacytoid dendritic cells decline during human ageing Scand. J. Immunol. 56 2002 518 521 12410802
Shrock E. Fujimura E. Kula T. Timms R.T. Lee I.-H. Leng Y. Robinson M.L. Sie B.M. Li M.Z. Chen Y. Viral epitope profiling of COVID-19 patients reveals cross-reactivity and correlates of severity Science 370 2020 eabd4250
Silvin A. Chapuis N. Dunsmore G. Goubet A.-G. Dubuisson A. Derosa L. Almire C. Hénon C. Kosmider O. Droin N. Elevated calprotectin and abnormal myeloid cell subsets discriminate severe from mild COVID-19 Cell 182 2020 1401 1418 e1418 32810439
Singh D.K. Singh B. Ganatra S.R. Gazi M. Cole J. Thippeshappa R. Alfson K.J. Clemmons E. Gonzalez O. Escobedo R. Responses to acute infection with SARS-CoV-2 in the lungs of rhesus macaques, baboons and marmosets Nat. Microbiol. 6 2021 73 86 33340034
Song J.-W. Lam S.M. Fan X. Cao W.-J. Wang S.-Y. Tian H. Chua G.H. Zhang C. Meng F.-P. Xu Z. Omics-driven systems interrogation of metabolic dysregulation in COVID-19 pathogenesis Cell Metab. 32 2020 188 202 e185 32610096
Sorgdrager F.J. Naudé P.J. Kema I.P. Nollen E.A. Deyn P.P.D. Tryptophan metabolism in inflammaging: from biomarker to therapeutic target Front. Immunol. 10 2019 2565 31736978
Tan L. Wang Q. Zhang D. Ding J. Huang Q. Tang Y.-Q. Wang Q. Miao H. Lymphopenia predicts disease severity of COVID-19: a descriptive and predictive study Signal Transduct. Target. Ther. 5 2020 1 3 32296011
Tang N. Li D. Wang X. Sun Z. Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia J. Thromb. Haemost. 18 2020 844 847 32073213
Tannahill G. Curtis A. Adamik J. Palsson-McDermott E. McGettrick A. Goel G. Frezza C. Bernard N. Kelly B. Foley N. Succinate is an inflammatory signal that induces IL-1β through HIF-1α Nature 496 2013 238 242 23535595
Tay M.Z. Poh C.M. Rénia L. MacAry P.A. Ng L.F. The trinity of COVID-19: immunity, inflammation and intervention Nat. Rev. Immunol. 20 2020 363 374 32346093
Thomas T. Stefanoni D. Dzieciatkowska M. Issaian A. Nemkov T. Hill R.C. Francis R.O. Hudson K.E. Buehler P.W. Zimring J.C. Evidence of structural protein damage and membrane lipid remodeling in red blood cells from COVID-19 patients J. Proteome Res. 19 2020 4455 4469 33103907
Thomas T. Stefanoni D. Reisz J.A. Nemkov T. Bertolone L. Francis R.O. Hudson K.E. Zimring J.C. Hansen K.C. Hod E.A. COVID-19 infection alters kynurenine and fatty acid metabolism, correlating with IL-6 levels and renal status JCI Insight 2020 5
Tikellis C. Thomas M. Angiotensin-converting enzyme 2 (ACE2) is a key modulator of the renin angiotensin system in health and disease Int. J. Pept. 2012 2012
Van Duin D. Mohanty S. Thomas V. Ginter S. Montgomery R.R. Fikrig E. Allore H.G. Medzhitov R. Shaw A.C. Age-associated defect in human TLR-1/2 function J. Immunol. 178 2007 970 975 17202359
Venkataraman K. Lee Y.-M. Michaud J. Thangada S. Ai Y. Bonkovsky H.L. Parikh N.S. Habrukowich C. Hla T. Vascular endothelium as a contributor of plasma sphingosine 1-phosphate Circ. Res. 102 2008 669 676 18258856
Verschoor C.P. Lelic A. Parsons R. Evelegh C. Bramson J.L. Johnstone J. Loeb M.B. Bowdish D.M. Serum C-reactive protein and congestive heart failure as significant predictors of herpes zoster vaccine response in elderly nursing home residents J. Infect. Dis. 216 2017 191 197 28838148
Wan S. Yi Q. Fan S. Lv J. Zhang X. Guo L. Lang C. Xiao Q. Xiao K. Yi Z. Characteristics of lymphocyte subsets and cytokines in peripheral blood of 123 hospitalized patients with 2019 novel coronavirus pneumonia (NCP) MedRxiv 2020
Wang M. Jiang L. Monticone R.E. Lakatta E.G. Proinflammation: the key to arterial aging Trends Endocrinol. Metab. 25 2014 72 79 24365513
Wanhella K.J. Fernandez-Patron C. Biomarkers of ageing and frailty may predict COVID-19 severity Ageing Res. Rev. 73 2022 101513
Weiskopf D. Weinberger B. Grubeck‐Loebenstein B. The aging of the immune system Transpl. Int. 22 2009 1041 1050 19624493
Wen W. Su W. Tang H. Le W. Zhang X. Zheng Y. Liu X. Xie L. Li J. Ye J. Immune cell profiling of COVID-19 patients in the recovery stage by single-cell sequencing Cell Discov. 6 2020 31 32377375
White J.T. Cross E.W. Burchill M.A. Danhorn T. McCarter M.D. Rosen H.R. O’Connor B. Kedl R.M. Virtual memory T cells develop and mediate bystander protective immunity in an IL-15-dependent manner Nat. Commun. 7 2016 1 13
Wiley C.D. Liu S. Limbad C. Zawadzka A.M. Beck J. Demaria M. Artwood R. Alimirah F. Lopez-Dominguez J.-A. Kuehnemann C. SILAC analysis reveals increased secretion of hemostasis-related factors by senescent cells Cell Rep. 28 2019 3329 3337 e3325. 31553904
Wilk A.J. Rustagi A. Zhao N.Q. Roque J. Martínez-Colón G.J. McKechnie J.L. Ivison G.T. Ranganath T. Vergara R. Hollis T. A single-cell atlas of the peripheral immune response in patients with severe COVID-19 Nat. Med. 26 2020 1070 1076 32514174
Witkowski J.M. Fulop T. Bryl E. Immunosenescence and COVID-19 Mech. Ageing Dev. 204 2022 111672
Wolters B. Junge U. Dziuba S. Roggendorf M. Immunogenicity of combined hepatitis A and B vaccine in elderly persons Vaccine 21 2003 3623 3628 12922091
Wong C. Goldstein D.R. Impact of aging on antigen presentation cell function of dendritic cells Curr. Opin. Immunol. 25 2013 535 541 23806201
Wong K.L. Yeap W.H. Tai J.J.Y. Ong S.M. Dang T.M. Wong S.C. The three human monocyte subsets: implications for health and disease Immunol. Res. 53 2012 41 57 22430559
Woodruff M.C. Ramonell R.P. Nguyen D.C. Cashman K.S. Saini A.S. Haddad N.S. Ley A.M. Kyu S. Howell J.C. Ozturk T. Extrafollicular B cell responses correlate with neutralizing antibodies and morbidity in COVID-19 Nat. Immunol. 21 2020 1506 1516 33028979
Wu C. Chen X. Cai Y. Zhou X. Xu S. Huang H. Zhang L. Zhou X. Du C. Zhang Y. Risk factors associated with acute respiratory distress syndrome and death in patients with coronavirus disease 2019 pneumonia in Wuhan China JAMA Intern. Med. 180 2020 934 943 32167524
Wu P. Chen D. Ding W. Wu P. Hou H. Bai Y. Zhou Y. Li K. Xiang S. Liu P. The trans-omics landscape of COVID-19 Nat. Commun. 12 2021 1 16 33397941
Xian H. Liu Y. Nilsson A.R. Gatchalian R. Crother T.R. Tourtellotte W.G. Zhang Y. Aleman-Muench G.R. Lewis G. Chen W. Metformin inhibition of mitochondrial ATP and DNA synthesis abrogates NLRP3 inflammasome activation and pulmonary inflammation Immunity 54 2021 1463 1477 e1411 34115964
Xu P. Zhou Q. Xu J. Mechanism of thrombocytopenia in COVID-19 patients Ann. Hematol. 99 2020 1205 1208 32296910
Xu Z. Shi L. Wang Y. Zhang J. Huang L. Zhang C. Liu S. Zhao P. Liu H. Zhu L. Pathological findings of COVID-19 associated with acute respiratory distress syndrome Lancet Respir. Med. 8 2020 420 422 32085846
Xudong X. Junzhu C. Xingxiang W. Furong Z. Yanrong L. Age-and gender-related difference of ACE2 expression in rat lung Life Sci. 78 2006 2166 2171 16303146
Yamada T. Sato S. Sotoyama Y. Orba Y. Sawa H. Yamauchi H. Sasaki M. Takaoka A. RIG-I triggers a signaling-abortive anti-SARS-CoV-2 defense in human lung cells Nat. Immunol. 22 2021 820 828 33976430
Yoo J.-S. Sasaki M. Cho S.X. Kasuga Y. Zhu B. Ouda R. Orba Y. de Figueiredo P. Sawa H. Kobayashi K.S. SARS-CoV-2 inhibits induction of the MHC class I pathway by targeting the STAT1-IRF1-NLRC5 axis Nat. Commun. 12 2021 1 17 33397941
Yoshikawa T.T. Epidemiology and unique aspects of aging and infectious diseases Clin. Infect. Dis. 30 2000 931 933 10880303
Youm Y.-H. Grant R.W. McCabe L.R. Albarado D.C. Nguyen K.Y. Ravussin A. Pistell P. Newman S. Carter R. Laque A. Canonical Nlrp3 inflammasome links systemic low-grade inflammation to functional decline in aging Cell Metab. 18 2013 519 532 24093676
Yu X. Cui L. Hou F. Liu X. Wang Y. Wen Y. Chi C. Li C. Liu R. Yin C. Angiotensin-converting enzyme 2-angiotensin (1-7)-Mas axis prevents pancreatic acinar cell inflammatory response via inhibition of the p38 mitogen-activated protein kinase/nuclear factor-κB pathway Int. J. Mol. Med. 41 2018 409 420 29138810
Zacca E.R. Crespo M.I. Acland R.P. Roselli E. Núñez N.G. Maccioni M. Maletto B.A. Pistoresi-Palencia M.C. Morón G. Aging impairs the ability of conventional dendritic cells to cross-prime CD8+ T cells upon stimulation with a TLR7 ligand PLoS One 10 2015 e0140672
Zarandi P.K. Zinatizadeh M.R. Zinatizadeh M. Yousefi M.H. Rezaei N. SARS-CoV-2: from the pathogenesis to potential anti-viral treatments Biomed. Pharmacother. 137 2021 111352
Zhang D. Guo R. Lei L. Liu H. Wang Y. Wang Y. Qian H. Dai T. Zhang T. Lai Y. COVID‐19 infection induces readily detectable morphologic and inflammation‐related phenotypic changes in peripheral blood monocytes J. Leukoc. Biol. 2020
Zhang Q. Bastard P. Liu Z. Pen J.L. Moncada-Velez M. Chen J. Ogishi M. Sabli I.K.D. Hodeib S. Korol C. Rosain J. Bilguvar K. Ye J. Bolze A. Bigio B. Yang R. Arias A.A. Zhou Q. Zhang Y. Onodi F. Korniotis S. Karpf L. Philippot Q. Chbihi M. Bonnet-Madin L. Dorgham K. Smith N. Schneider W.M. Razooky B.S. Hoffmann H.-H. Michailidis E. Moens L. Han J.E. Lorenzo L. Bizien L. Meade P. Neehus A.-L. Ugurbil A.C. Corneau A. Kerner G. Zhang P. Rapaport F. Seeleuthner Y. Manry J. Masson C. Schmitt Y. Schlüter A. Voyer T.L. Khan T. Li J. Fellay J. Roussel L. Shahrooei M. Alosaimi M.F. Mansouri D. Al-Saud H. Al-Mulla F. Almourfi F. Al-Muhsen S.Z. Alsohime F. Turki S.A. Hasanato R. Beek Dvd Biondi A. Bettini L.R. D’Angio’ M. Bonfanti P. Imberti L. Sottini A. Paghera S. Quiros-Roldan E. Rossi C. Oler A.J. Tompkins M.F. Alba C. Vandernoot I. Goffard J.-C. Smits G. Migeotte I. Haerynck F. Soler-Palacin P. Martin-Nalda A. Colobran R. Morange P.-E. Keles S. Çölkesen F. Ozcelik T. Yasar K.K. Senoglu S. Karabela Ş.N. Rodríguez-Gallego C. Novelli G. Hraiech S. Tandjaoui-Lambiotte Y. Duval X. Laouénan C. Snow A.L. Dalgard C.L. Milner J.D. Vinh D.C. Mogensen T.H. Marr N. Spaan A.N. Boisson B. Boisson-Dupuis S. Bustamante J. Puel A. Ciancanelli M.J. Meyts I. Maniatis T. Soumelis V. Amara A. Nussenzweig M. García-Sastre A. Krammer F. Pujol A. Duffy D. Lifton R.P. Zhang S.-Y. Gorochov G. Béziat V. Jouanguy E. Sancho-Shimizu V. Rice C.M. Abel L. Notarangelo L.D. Cobat A. Su H.C. Casanova J.-L. Foti G. Bellani G. Citerio G. Contro E. Pesci A. Valsecchi M.G. Cazzaniga M. Abad J. Aguilera-Albesa S. Akcan O.M. Darazam I.A. Aldave J.C. Ramos M.A. Nadji S.A. Alkan G. Allardet-Servent J. Allende L.M. Alsina L. Alyanakian M.-A. Amador-Borrero B. Amoura Z. Antolí A. Arslan S. Assant S. Auguet T. Azot A. Bajolle F. Baldolli A. Ballester M. Feldman H.B. Barrou B. Beurton A. Bilbao A. Blanchard-Rohner G. Blanco I. Blandinières A. Blazquez-Gamero D. Bloomfield M. Bolivar-Prados M. Borie R. Bosteels C. Bousfiha A.A. Bouvattier C. Boyarchuk O. Bueno M.R.P. Bustamante J. Agra J.J.C. Calimli S. Capra R. Carrabba M. Casasnovas C. Caseris M. Castelle M. Castelli F. Vera M.Cd Castro M.V. Catherinot E. Chalumeau M. Charbit B. Cheng M.P. Clavé P. Clotet B. Codina A. Colkesen F. Çölkesen F. Colobran R. Comarmond C. Dalmau D. Darley D.R. Dauby N. Dauger S. Pontual Ld Dehban A. Delplancq G. Demoule A. Diehl J.-L. Dobbelaere S. Durand S. Eldars W. Elgamal M. Elnagdy M.H. Emiroglu M. Erdeniz E.H. Aytekin S.E. Euvrard R. Evcen R. Fabio G. Faivre L. Falck A. Fartoukh M. Faure M. Arquero M.F. Flores C. Francois B. Fumadó V. Fusco F. Solis B.G. Gaussem P. Gil-Herrera J. Gilardin L. Alarcon M.G. Girona-Alarcón M. Goffard J.-C. Gok F. González-Montelongo R. Guerder A. Gul Y. Guner S.N. Gut M. Hadjadj J. Haerynck F. Halwani R. Hammarström L. Hatipoglu N. Hernandez-Brito E. Heijmans C. Holanda-Peña M.S. Horcajada J.P. Hoste L. Hoste E. Hraiech S. Humbert L. Iglesias A.D. Íñigo-Campos A. Jamme M. Arranz M.J. Jordan I. Jorens P. Kanat F. Kapakli H. Kara I. Karbuz A. Yasar K.K. Keles S. Demirkol Y.K. Klocperk A. Król Z.J. Kuentz P. Kwan Y.W.M. Lagier J.-C. Lambrecht B.N. Lau Y.-L. Bourgeois F.L. Leo Y.-S. Lopez R.L. Leung D. Levin M. Levy M. Lévy R. Li Z. Linglart A. Loeys B. Lorenzo-Salazar J.M. Louapre C. Lubetzki C. Luyt C.-E. Lye D.C. Mansouri D. Marjani M. Pereira J.M. Martin A. Pueyo D.M. Martinez-Picado J. Marzana I. Mathian A. Matos L.R.B. Matthews G.V. Mayaux J. Mège J.-L. Melki I. Meritet J.-F. Metin O. Meyts I. Mezidi M. Migeotte I. Millereux M. Mirault T. Mircher C. Mirsaeidi M. Melián A.M. Martinez A.M. Morange P. Mordacq C. Morelle G. Mouly S. Muñoz-Barrera A. Naesens L. Nafati C. Neves J.F. Ng L.F.P. Medina Y.N. Cuadros E.N. Ocejo-Vinyals J.G. Orbak Z. Oualha M. Özçelik T. Pan-Hammarström Q. Parizot C. Pascreau T. Paz-Artal E. Pellegrini S. Diego R.Pd Philippe A. Philippot Q. Planas-Serra L. Ploin D. Poissy J. Poncelet G. Pouletty M. Quentric P. Raoult D. Rebillat A.-S. Reisli I. Ricart P. Richard J.-C. Rivet N. Rivière J.G. Blanch G.R. Rodrigo C. Rodriguez-Gallego C. Rodríguez-Palmero A. Romero C.S. Rothenbuhler A. Rozenberg F. Prado M.Y.Rd Riera J.S. Sanchez O. Sánchez-Ramón S. Schluter A. Schmidt M. Schweitzer C.E. Scolari F. Sediva A. Seijo L.M. Sene D. Senoglu S. Seppänen M.R.J. Ilovich A.S. Shahrooei M. Slabbynck H. Smadja D.M. Sobh A. Moreno X.S. Solé-Violán J. Soler C. Soler-Palacín P. Stepanovskiy Y. Stoclin A. Taccone F. Tandjaoui-Lambiotte Y. Taupin J.-L. Tavernier S.J. Terrier B. Thumerelle C. Tomasoni G. Toubiana J. Alvarez J.T. Trouillet-Assant S. Troya J. Tucci A. Ursini M.V. Uzunhan Y. Vabres P. Valencia-Ramos J. Braeckel E.V. Velde S.Vd Rym A.M.V.D. Praet J.V. Vandernoot I. Vatansev H. Vélez-Santamaria V. Viel S. Vilain C. Vilaire M.E. Vincent A. Voiriot G. Vuotto F. Yosunkaya A. Young B.E. Yucel F. Zannad F. Zatz M. Belot A. Bole-Feysot C. Lyonnet S. Masson C. Nitschke P. Pouliet A. Schmitt Y. Tores F. Zarhrate M. Abel L. Andrejak C. Angoulvant F. Bachelet D. Basmaci R. Behillil S. Beluze M. Benkerrou D. Bhavsar K. Bompart F. Bouadma L. Bouscambert M. Caralp M. Cervantes-Gonzalez M. Chair A. Coelho A. Couffignal C. Couffin-Cadiergues S. D’Ortenzio E. Silveira C.D. Debray M.-P. Deplanque D. Descamps D. Desvallées M. Diallo A. Diouf A. Dorival C. Dubos F. Duval X. Eloy P. Enouf V.V. Esperou H. Esposito-Farese M. Etienne M. Ettalhaoui N. Gault N. Gaymard A. Ghosn J. Gigante T. Gorenne I. Guedj J. Hoctin A. Hoffmann I. Jaafoura S. Kafif O. Kaguelidou F. Kali S. Khalil A. Khan C. Laouénan C. Laribi S. Le M. Hingrat Q.L. Mestre S.L. Nagard H.L. Lescure F.-X. Lévy Y. Levy-Marchal C. Lina B. Lingas G. Lucet J.C. Malvy D. Mambert M. Mentré F. Mercier N. Meziane A. Mouquet H. Mullaert J. Neant N. Noret M. Pages J. Papadopoulos A. Paul C. Peiffer-Smadja N. Petrov-Sanchez V. Peytavin G. Picone O. Puéchal O. Rosa-Calatrava M. Rossignol B. Rossignol P. Roy C. Schneider M. Semaille C. Mohammed N.S. Tagherset L. Tardivon C. Tellier M.-C. Téoulé F. Terrier O. Timsit J.-F. Trioux T. Tual C. Tubiana S. Werf Svd Vanel N. Veislinger A. Visseaux B. Wiedemann A. Yazdanpanah Y. Alavoine L. Amat K.K.A. Behillil S. Bielicki J. Bruijning P. Burdet C. Caumes E. Charpentier C. Coignard B. Costa Y. Couffin-Cadiergues S. Damond F. Dechanet A. Delmas C. Descamps D. Duval X. Ecobichon J.-L. Enouf V. Espérou H. Frezouls W. Houhou N. Ilic-Habensus E. Kafif O. Kikoine J. Hingrat Q.L. Lebeaux D. Leclercq A. Lehacaut J. Letrou S. Lina B. Lucet J.-C. Malvy D. Manchon P. Mandic M. Meghadecha M. Motiejunaite J. Nouroudine M. Piquard V. Postolache A. Quintin C. Rexach J. Roufai L. Terzian Z. Thy M. Tubiana S. Werf Svd Vignali V. Visseaux B. Yazdanpanah Y. Agtmael Mv Algera A.G. Baarle Fv Bax D. Beudel M. Bogaard H.J. Bomers M. Bos L. Botta M. Brabander Jd Bree Gd Brouwer M.C. Bruin Sd Bugiani M. Bulle E. Chouchane O. Cloherty A. Elbers P. Fleuren L. Geerlings S. Geerts B. Geijtenbeek T. Girbes A. Goorhuis B. Grobusch M.P. Hafkamp F. Hagens L. Hamann J. Harris V. Hemke R. Hermans S.M. Heunks L. Hollmann M.W. Horn J. Hovius J.W. Jong M.Dd Koning R. Mourik Nv Nellen J. Paulus F. Peters E. Poll Tvd Preckel B. Prins J.M. Raasveld J. Reijnders T. Schinkel M. Schultz M.J. Schuurman A. Sigaloff K. Smit M. Stijnis C.S. Stilma W. Teunissen C. Thoral P. Tsonas A. Valk Mvd Veelo D. Vlaar A.P.J. Vries Hd Vugt Mv Wiersinga W.J. Wouters D. Zwinderman A.H. Beek Dvd Abel L. Aiuti A. Muhsen S.A. Al-Mulla F. Anderson M.S. Arias A.A. Feldman H.B. Bogunovic D. Bolze A. Bondarenko A. Bousfiha A.A. Brodin P. Bryceson Y. Bustamante C.D. Butte M. Casari G. Chakravorty S. Christodoulou J. Cirulli E. Condino-Neto A. Cooper M.A. Dalgard C.L. David A. DeRisi J.L. Desai M. Drolet B.A. Espinosa S. Fellay J. Flores C. Franco J.L. Gregersen P.K. Haerynck F. Hagin D. Halwani R. Heath J. Henrickson S.E. Hsieh E. Imai K. Itan Y. Karamitros T. Kisand K. Ku C.-L. Lau Y.-L. Ling Y. Lucas C.L. Maniatis T. Mansouri D. Marodi L. Meyts I. Milner J. Mironska K. Mogensen T. Morio T. Ng L.F.P. Notarangelo L.D. Novelli A. Novelli G. O’Farrelly C. Okada S. Ozcelik T. Diego R.Pd Planas A.M. Prando C. Pujol A. Quintana-Murci L. Renia L. Renieri A. Rodríguez-Gallego C. Sancho-Shimizu V. Sankaran V. Barrett K.S. Shahrooei M. Snow A. Soler-Palacín P. Spaan A.N. Tangye S. Turvey S. Uddin F. Uddin M.J. Beek Dvd Vazquez S.E. Vinh D.C. Bernuth Hv Washington N. Zawadzki P. Su H.C. Casanova J.-L. Jing H. Tung W. Luthers C.R. Bauman B.M. Shafer S. Zheng L. Zhang Z. Kubo S. Chauvin S.D. Meguro K. Shaw E. Lenardo M. Lack J. Karlins E. Hupalo D.M. Rosenberger J. Sukumar G. Wilkerson M.D. Zhang X. Inborn errors of type I IFN immunity in patients with life-threatening COVID-19 Science 370 2020 eabd4570
Zhao J. Zhao J. Perlman S. T cell responses are required for protection from clinical disease and for virus clearance in severe acute respiratory syndrome coronavirus-infected mice J. Virol. 84 2010 9318 9325 20610717
Zhao J. Zhao J. Legge K. Perlman S. Age-related increases in PGD 2 expression impair respiratory DC migration, resulting in diminished T cell responses upon respiratory virus infection in mice J. Clin. Investig. 121 2011 4921 4930 22105170
Zheng H.-Y. Zhang M. Yang C.-X. Zhang N. Wang X.-C. Yang X.-P. Dong X.-Q. Zheng Y.-T. Elevated exhaustion levels and reduced functional diversity of T cells in peripheral blood may predict severe progression in COVID-19 patients Cell. Mol. Immunol. 17 2020 541 543 32203186
Zheng Y. Liu X. Le W. Xie L. Li H. Wen W. Wang S. Ma S. Huang Z. Ye J. A human circulating immune cell landscape in aging and COVID-19 Protein Cell 11 2020 740 770 32780218
Zheng Z. Peng F. Xu B. Zhao J. Liu H. Peng J. Li Q. Jiang C. Zhou Y. Liu S. Risk factors of critical & mortal COVID-19 cases: A systematic literature review and meta-analysis J. Infect. 81 2020 e16 e25
Zhou F. Yu T. Du R. Fan G. Liu Y. Liu Z. Xiang J. Wang Y. Song B. Gu X. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study Lancet 395 2020 1054 1062 32171076
Zhou R. To K.K.-W. Wong Y.-C. Liu L. Zhou B. Li X. Huang H. Mo Y. Luk T.-Y. Lau T.T.-K. Acute SARS-CoV-2 infection impairs dendritic cell and T cell responses Immunity 53 2020 864 877 e865 32791036
Zhou Y. Fu B. Zheng X. Wang D. Zhao C. Qi Y. Sun R. Tian Z. Xu X. Wei H. Pathogenic T-cells and inflammatory monocytes incite inflammatory storms in severe COVID-19 patients Natl. Sci. Rev. 7 2020 998 1002 34676125
Zinatizadeh M.R. Schock B. Chalbatani G.M. Zarandi P.K. Jalali S.A. Miri S.R. The Nuclear Factor Kappa B (NF-kB) signaling in cancer development and immune Dis. Genes Dis. 8 2021 287 297 33997176
Zinatizadeh M.R. Zarandi P.K. Zinatizadeh M. Yousefi M.H. Amani J. Rezaei N. Efficacy of mRNA, adenoviral vector, and perfusion protein COVID-19 vaccines Biomed. Pharmacother. 146 2022 112527
Zingaropoli M.A. Nijhawan P. Carraro A. Pasculli P. Zuccala P. Perri V. Marocco R. Kertusha B. Siccardi G. Del Borgo C. Increased sCD163 and sCD14 plasmatic levels and depletion of peripheral blood pro-inflammatory monocytes, myeloid and plasmacytoid dendritic cells in patients with severe COVID-19 pneumonia Front. Immunol. 12 2021 627548
Zost S.J. Gilchuk P. Case J.B. Binshtein E. Chen R.E. Nkolola J.P. Schäfer A. Reidy J.X. Trivette A. Nargi R.S. Potently neutralizing and protective human antibodies against SARS-CoV-2 Nature 584 2020 443 449 32668443
Zuo Y. Estes S.K. Ali R.A. Gandhi A.A. Yalavarthi S. Shi H. Sule G. Gockman K. Madison J.A. Zuo M. Prothrombotic autoantibodies in serum from patients hospitalized with COVID-19 Sci. Transl. Med. 12 2020 eabd3876
| 36516928 | PMC9741765 | NO-CC CODE | 2022-12-14 23:45:33 | no | Ageing Res Rev. 2023 Feb 11; 84:101818 | utf-8 | Ageing Res Rev | 2,022 | 10.1016/j.arr.2022.101818 | oa_other |
==== Front
J Obstet Gynaecol India
J Obstet Gynaecol India
Journal of Obstetrics and Gynaecology of India
0971-9202
0975-6434
Springer India New Delhi
1727
10.1007/s13224-022-01727-7
Original Article
Effectiveness of Tibolone in Relieving Postmenopausal Symptoms for a Short-Term Period in Indian Women
http://orcid.org/0000-0002-7816-3980
Malik Renuka [email protected]
Meghana Reddy P.
grid.414117.6 0000 0004 1767 6509 Department of OB-GYN, ABVIMS & Dr RML Hospital, New Delhi, India
11 12 2022
16
24 1 2022
30 10 2022
© Federation of Obstetric & Gynecological Societies of India 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
Tibolone is an alternative to conventional estrogen and progesterone in relieving post-menopausal symptoms in Indian women.
Material and Methods
A prospective short-term observational study was done at a tertiary care teaching hospital in New Delhi from November 2019 to September 2021. Fifty-three women, less than 60 years of age, presenting with moderate to severe intensity of menopausal symptoms as assessed by measuring menopausal rating score (MRS > 8) were enrolled and given Tibolone 2.5 mg daily for 3 months. Improvements in symptoms were seen at 1 month and 3 months. Side effects were also noted.
Results
Marked improvement was seen as reduction in scores of psychological, somatic and genitourinary symptoms was noted. The psychological symptoms reduced from 8.92 ± 1.959 to 2.905 ± 1.042, the somatic symptoms decreased from 8.33 ± 2.299 to 3.4 ± 1.167, and genitourinary symptoms decreased from 3.64 ± 1.42 to 2.150 ± 0.948 after 3 months of treatment with Tibolone. Only 3 patients (5.6%) experienced vaginal spotting with no major side effects.
Conclusions
Tibolone is a highly effective and well accepted drug to reduce moderate to severe menopausal symptoms, especially psychological symptoms including depression.
Keywords
Tibolone
Menopausal symptoms
Hot flushes
Hormone replacement therapy
Genitourinary symptoms of menopause
Psychological symptoms at menopause
Somatic symptoms at menopause
==== Body
pmcIntroduction
Menopause is a naturally occurring event in all women after the reproductive age. The menopausal symptoms adversely affect the quality of life in menopausal women. Hormonal replacement therapy helps to alleviate the bothersome symptoms. Tibolone, which is a STEAR (Selective Tissue Estrogenic Activity Regulator), is used for the treatment of women with menopausal symptoms. It possesses estrogenic, progestogenic, and androgenic properties [1]. This study was done to see the effectiveness of Tibolone in alleviating post-menopausal symptoms in Indian women.
Materials and Methods
A short-term prospective observational study was conducted in the department of obstetrics and gynecology, from November 2019 to September 2021. Inclusion criteria were women with age less than 60 years having moderate to severe intensity of menopausal symptoms as assessed by Menopausal Rating Score (MRS > 8). The exclusion criteria were postmenopausal women with endometrial thickness greater than 4 mm, women with breast cancer or family history of breast cancer, severe medical disorders (cardiac, liver, renal diseases), undiagnosed vaginal bleeding, history of psychiatric illness diagnosed before menopause, and women with an active or past history of thromboembolic disorders.
A total of 150 patients attending menopausal OPD were screened for enrollment, which included a detailed history, examination, and ultrasound evaluation for endometrial thickness. Out of these, 76 patients who met the inclusion criteria were enrolled in the study. Enrolled patients were given Tibolone 2.5 mg once daily for a period of 3 months and were assessed for improvement in menopausal symptoms objectively by MRS questionnaire at 1 month and 3 months after taking Tibolone. Side effects at 15 days, 1 month, and 3 months were noted.
Observations and Results
Patients presenting to menopause OPD and gynecology OPD were assessed for the severity of menopausal symptoms. Patients meeting exclusion criteria and those with mild severity scores (MRS score < 8) were excluded from the study. Seventy-six patients were enrolled in this study, and among them, 23 patients (30.2%) were lost to follow-up (due to the Covid-19 pandemic, which occurred during the duration of our study period) and were not included in the study. Fifty-three patients completed the 3 month treatment with Tibolone.
Demographic Variables of the Study Population
The age of patients in the study population ranged from 39 to 59 years, with a mean age of 51.2 years. Thirty-seven patients (69.8%) were in the age group of 50–60 years, 15 patients (28.3%) were in the age group of 40–50 years, and only one patient (1.8%) was in the age group of 30–40 years. The age at menopause in the study population ranged from 36 to 54 years. The mean age at menopause was 46.79 years. There was only one patient among the study population, with premature ovarian failure with menopause at 36 years. The duration of menopause ranged from 1 to 14 years in the study population, with a mean duration of menopause of 4.6 years with a SD of 3.05 years. Thirty-eight patients (71.69%) had their menopause for duration of 1–6 years, 13 patients (24.52%) had their menopause for a duration of 7–12 years, and 2 patients (3.7%) had their menopause for a duration of 13–18 years. In the study population, 48 patients (90.5%) had natural menopause, while 5 patients (9.4%) had surgical menopause.
Among the study population, 33 patients (62%) belonged to the lower middle class. Only 1 patient (1.8%) belonged to the lower class, 4 patients (7.5%) belonged to the upper lower class, 12 patients (23%) belonged to upper middle class and 3 patients (5.6%) belonged to upper class.
Effect of Tibolone on Severity of Menopausal Symptoms
At enrollment, 11 patients (20.75%) had a total MRS score in the moderate severity range (9–15) and 42 patients (79.42%) had a total MRS score in the severe range (> 15). There was a marked reduction in the severity of symptoms after 1 month and 3 months of taking Tibolone (Fig. 1). No patient reported severe intensity of symptoms at the end of 3 months of taking Tibolone. The mean of total MRS score at enrollment was 21 with SD of 4.5. This reduced it to 13.98 with a SD of 2.196 after 1 month of taking Tibolone. After 3 months of taking Tibolone, the mean of the total score further decreased to 8.47 with a SD of 2.09 (Fig. 2). This difference was statistically significant with a p value of < 0.00001 both at 1 month and 3 months of taking Tibolone.Fig. 1 Decreasing severity of menopausal symptoms in the study population at enrollment and at 1 month and at 3 months of treatment with Tibolone
Fig. 2 Comparison of total MRS scores at enrollment, at 1 month and at 3 months of Tibolone
Effect of Tibolone on Psychological Symptoms
The psychological symptoms at enrollment ranged from 5 to 12 with a mean score of 8.92 and a SD of 1.95. This was decreased to 5.622 with a SD of 1.259 after 1 month of taking Tibolone and further to 2.905 with a SD of 1.042 after 3 months of taking Tibolone (Fig. 3). This improvement was found to be highly statistically significant with a p value of < 0.00001 both at 1 month and 3 months of taking Tibolone. 7 patients (13.2%), 18 patients (33.9%), and 23 patients (43.3%) who reported very severe, severe and moderate intensity of depression, respectively, at enrollment, had significant improvement in their symptoms with none of the patients reporting severe or very severe intensity, while only 1 patient (1.8%) reported moderate intensity of depression at 3 months of taking Tibolone.Fig. 3 Improvement of psychological subscore at 1 month and at 3 months of treatment with Tibolone
Effect of Tibolone on Somatic Symptoms
At enrollment, the somatic subscore ranged from 4 to 12 with a mean score of 8.33 with a SD of 2.29, which reduced to 5.4 with a SD of 1.39 after 1 month of taking Tibolone and further to 3.4 with a SD of 1.16 after 3 months of taking Tibolone (Fig. 4). This difference was found to be statistically significant with a p value of < 0.00001, both at 1 month and 3 months of treatment with Tibolone.Fig. 4 Improvement of Somatic symptom score at 1 month and at 3 months of Tibolone
At enrollment, 63% of patients reported hot flushes of varying intensity. Fifteen patients (28.3%), 13 patients (24.5%), 4 patients (7.54%) and 1 patient (1.8%) reported very severe, severe, moderate and mild intensity of hot flushes, respectively. These patients had improvement in their symptom intensity after 1 month of taking Tibolone, with 4 patients (7.54%) reporting complete resolution of hot flushes and none of the patients reported very severe intensity of hot flushes. Only 3 patients (5.6%) reported severe intensity. After 3 months of taking Tibolone, a total of 11 patients (20.75%) reported having complete resolution of hot flushes and none of the patients had severe or very severe intensity of hot flushes (Fig. 5). Fig. 5 Improvement in severity of hot flushes in study population at enrolment and after 1 month and 3 months of treatment with Tibolone
Effect of Tibolone on Genitourinary Symptoms
The Genitourinary subscore at enrollment ranged from 1–9 with a mean score of 3.64 and a SD of 1.42. This decreased to 2.867 with a SD of 0.832 after 1 month of taking Tibolone and further decreased to 2.150 with a SD of 0.948 after 3 months of taking Tibolone (Fig. 6).Fig. 6 Improvement of genitourinary symptoms at 1 month and at 3 months of treatment with Tibolone
Effect of Tibolone on Endometrial Thickness
The endometrial thickness at enrollment was 2.6 ± 0.53 mm. It was reassessed after 3 months of treatment with Tibolone, and it was observed that the mean value decreased to 2.19 mm with a SD of 0.44 mm. The difference in the mean values was found to be significant (p < 0.00001).
Side Effects with Tibolone
The patients were assessed for side effects at 15 days, 1 month, and 3 months after starting treatment with Tibolone. Only 3 patients (5.6%) were reported to have vaginal spotting, 2 of them in the initial 15 days of taking Tibolone. 3rd patient reported the same after 1 month of taking Tibolone. All were isolated incidents that resolved spontaneously and were not reported to be bothersome by the patients. None of the patients in the study reported having breast tenderness, calf pain or other symptoms of DVT at 15 days, 1 month, or at 3 months of taking Tibolone (Table 1).Table 1 Side effects experienced by patients on treatment with Tibolone
Side effect 15 Days
n (%) 1 Month
n (%) 3 Months
n (%)
Vaginal spotting 2 (3.7%) 1 (1.8%) 0 (0%)
Nausea 7 (13.2%) 2 (3.7%) 0 (0%)
Vomiting 1 (1.8%) 0 (0%) 0 (0%)
Breast Tenderness 0 (0%) 0 (0%) 0 (0%)
DVT 0 (0%) 0 (0%) 0 (0%)
DVT Deep vein Thrombosis
Effect on LFT and Lipid Profile
Lipid profile and liver function tests were assessed at enrollment, and after 3 months of taking Tibolone, no significant statistical difference was observed. After 3 months of treatment, the mean total bilirubin decreased to 0.635 mg/dl, the mean SGOT increased, and the SGPT value decreased. These differences were not significant statistically (p = 0.176, 0.412, 0.17, respectively). The effect on the lipid profile showed no statistical difference. An increase in total VLDL cholesterol was seen from a mean of 33 mg/dl to33.4 mg/dl, but was not statistically different (p = 0.502). There was a non-significant reduction in total cholesterol, HDL, LDL, and TG with p values of 0.09, 0.25, 0.21 and 0.31, respectively.
Discussion
There has been a resurgence of interest in treatment for postmenopausal symptoms, which can be distressing and reduce the quality of life. A relook at WHI study and as per recommendations of international menopausal society in 2016 [2, 3], selection of patients less than 60 years of age reduces the adverse effects of hormonal replacement therapy. Tibolone can be a safer alternative to the conventional estrogen and progesterone treatments as its pharmacodynamic profile is significantly different. It is synthetic steroid which is a Selective Tissue Estrogen Activity Regulator (STEAR). Upon oral ingestion, it is converted into 3 α hydroxyl Tibolone, 3 β hydroxy Tibolone and δ4 isomer. The hydroxy metabolites are estrogenic and have a positive effect on vagina and bone homeostasis. The delta isomer is responsible for both the progestogenic as well as androgenic properties of Tibolone. The progestogenic action is responsible for the protective effect of Tibolone in endometrium, while the androgenic action is responsible for the reduction in vasomotor symptoms of hot flushes and enhancing the sexual well being [4, 5]. The major advantage of Tibolone and its hydroxy metabolites is their lack of estrogenic activity in breast tissue, mostly by inhibiting the enzyme estrogen sulfatase and partly by increasing the enzyme activity of sulfatotransferase thereby producing the inactive metabolite-estrogen sulfate and effectively decreasing the active estrogen-estradiol thus decreasing incidence of breast cancer.
In our study population, 33 patients (62%) presented with symptoms of hot flushes. Indian studies find hot flushes less common and less bothersome. An incidence of 36.7% was found in rural India [6] and 34% in urban India [7]. Western countries report a higher incidence (80% to 88%) of postmenopausal hot flushes [8, 9]. At the end of three months of taking Tibolone, patients reported significant improvement in the hot flushes as there were no patients who reported very severe or severe intensity of hot flushes. Only 8 patients (15%) reported moderate hot flushes; 14 patients reported only mild hot flushes; and 31 patients (58.4%) reported complete resolution of hot flushes. This effect of Tibolone was similar to the Cochrane review and other studies, which spanned from three months to three years [1, 4, 5, 8, 9].
The mean total MRS score decreased to 8.47 ± 2.09 after 3 months of treatment with Tibolone. Somatic, psychological, and genitourinary symptoms significantly improved from the baseline, with maximum reduction in symptoms seen in the psychological subscale, followed by somatic and then genitourinary symptoms.
As seen in our study, Tibolone is a well-tolerated drug. In our study, only 5.6% (n = 3) reported a single self-limiting episode of vaginal spotting. These results were lower when compared to the THEBES trial, in which vaginal bleeding was 14.5%. The Cochrane review also reported an incidence of 31–44% [1]. This could be explained as our inclusion criteria in patients was endometrial thickness less than 4 mm, unlike other studies. No significant change in endometrial thickness with Tibolone was noted in our study, which is similar to that reported in the OPAL study [10]. During treatment with Tibolone, no effect on liver function tests or lipid profile was noted in our study. This effect on lipid profile differs from the results of a meta-analysis in which they reported a decrease in total cholesterol, HDL-C and triglyceride levels [11–13]. However, they had included studies in which Tibolone was given for a longer duration in contrast to our study where Tibolone was given only for a period of 3 months. Tibolone given for longer period has been shown to cause elevated liver enzymes [14, 15].
There are few studies using Tibolone for relief of menopausal symptoms in India. Authors found a short trial of three months of Tibolone helps menopausal patients tide over, offering a marked reduction in symptoms and an improved quality of life with no side effects, but patients need reassessment for prolonged use. The ease of drug usage, daily dosing compared to sequential estrogen and progesterone in patients with an intact uterus is an advantage for patient compliance. The short coming of our study was the short duration of 3 months. As the symptoms are seen in the transition phase too, Tibolone can be given for short term, though in our study we restricted its use in menopausal patients. The effect of Tibolone on bone health was not assessed because of the short-term nature of our study. The beneficial effect is seen in prevention, although Tibolone is not recommended for treatment of osteopenia under current guidelines. Tibolone is also a drug of choice in patients who had endometriosis [16] although no patient in our study had endometriosis. A short-term treatment is highly effective in reducing distressing symptoms of menopause, improving quality of life with no side effects. Patients should be re-assessed after three months and re-evaluated for long-term treatment in case of recurrence symptoms.
Conclusion
Tibolone is a highly effective and well accepted drug to reduce moderate to severe menopausal symptoms, especially psychological symptoms including depression. The maximum reduction was seen in the psychological symptoms, followed by somatic and genitourinary symptoms. A short three-month treatment relieves most of the symptoms without any side effects; long-term treatment needs further monitoring.
Funding
None.
Declarations
Conflict of Interest
None.
Ethical Approval
Ethical approval was taken from Institutional Ethical committee-FP. No TP(MD/MS)()/IEC/ABVIMS/RMLH736/19.
Ethical Standards
This study was performed in line with the principles of the Declaration of Helsinki.
Informed Consent
Written informed consent was taken from all patients for participation and publication.
Dr. Renuka Malik is working as a consultant OB-GYN in ABVIMS and Dr. RML Hospital, New Delhi. She is an alumni of Lady Hardinge Hospital, New Delhi. Her areas of interest include high-risk obstetrics and menopause. She runs the menopause clinic at her institution. Dr. P. Meghana Reddy is Resident, OB-GYN, ABVIMS & Dr. RML Hospital, New Delhi.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Giulio F Enrica P Susanna M Short-term and long term effects of tibolone in postmenopausal women, Cochrane review Cochrane Database Syst Rev 2016 10 CD008536 27733017
2. Baber RJ Panay N Fenton A IMS Writing Group IMS Recommendations on women’s midlife health and menopause hormone therapy Climacteric 2016 19 2 109 150 10.31019/13697137.2015.1129166 26872610
3. Amy V Rebecca D Jennifer K Menopause: a global perspective and clinical guide for practice Clin Obstet Gynecol 2021 64 3 528 554 10.1097/GRF.0000000000000639 34323232
4. Bansal R Aggarwal N Menopausal hot flashes: a concise review J Mid Life Health 2019 10 1 6 13 10.4103/jmh.JMH_7_19
5. Jane FM Davis SR A practitioner’s toolkit for managing the menopause Climacteric 2014 17 5 564 579 10.3109/13697137.2014.929651 24998761
6. Meenakshi K Komal S Priyanka C Seema V Pankaj K Tarun S Prevalence of menopausal symptoms and its effect on quality of life among rural middle aged women (40–60 years) of Haryana, India [internet].org Int J Appl Basic Med Res 2020 10 3 183 188 10.4103/ijabmr.IJABMR_428_19 33088741
7. Malik R Pokeria C Singh S Correlation of menopausal symptoms with serum estradiol: a study in urban Indian postmenopausal women J Obstet Gynecol India 2021 10.1007/s13224-021-01518-6
8. Huang KE Baber R Pacific A Tibolone Consensus Group Updated clinical recommendations for the use of tibolone in Asian women Climacteric 2010 13 4 317 327 10.3109/13697131003681458 20443720
9. Landgren MB Helmond FA Engelen S Tibolone relieves climacteric symptoms in highly symptomatic women with at least seven hot flushes and sweats per day Maturitas 2005 50 3 222 230 10.1016/j.maturitas.2004.06.001 15734603
10. Changu LV Zhang W Tan X Shang X Găman MA Salem H The effect of Tibolone treatment on lipid profile in women: a systematic review and dose–response meta-analysis of randomized controlled trials Pharmacol Res 2021 169 105612 10.1016/j.phrs.2021.105612 33865986
11. Langer RD Landgren BM Rymer J Helmond FA OPAL Investigators Effects of Tibolone and continuous combined conjugated equine estrogen/medroxyprogesterone acetate on the endometrium and vaginal bleeding: results of the OPAL study Am J Obstet Gynecol 2006 195 5 1320 1327 10.1016/j.ajog.2006.03.045 16875644
12. Nieciecka A Kędziora-Kornatowska K Janiszewska M Tibolone among drugs in the therapy of postmenopausal women Med Res J 2021 6 140 146 10.5603/MRJ.a2021.0014
13. Kotani K Sahebkar A Serban C Andrica F Toth PP Jones SR Lipid and blood pressure meta-analysis collaboration (LBPMC) group. Tibolone decreases lipoprotein (a) levels in postmenopausal women Atherosclerosis 2015 242 1 87 96 10.1016/j.atherosclerosis.2015.06.056 26186655
14. Fait T Menopause hormone therapy: latest developments and clinical practice Drugs Context 2019 2 8 212551 10.7573/dic.212551
15. LiverTox. Clinical and research information on drug-induced liver injury [internet]. Bethesda: National Institute of Diabetes and Digestive and Kidney Diseases; 2012-. Tibolone. [Updated Sep 2 2020].
16. Gemmell LC Webster KE Kirtley S Vincent K Zondervan KT Becker CM The management of menopause in women with a history of endometriosis: a systematic review Hum Reprod Update 2017 23 4 481 500 10.1093/humrupd/dmx011 28498913
| 0 | PMC9741860 | NO-CC CODE | 2022-12-14 23:29:58 | no | J Obstet Gynaecol India. 2022 Dec 11;:1-6 | utf-8 | J Obstet Gynaecol India | 2,022 | 10.1007/s13224-022-01727-7 | oa_other |
==== Front
3 Biotech
3 Biotech
3 Biotech
2190-572X
2190-5738
Springer International Publishing Cham
3416
10.1007/s13205-022-03416-8
Review Article
Susceptibility of SARS Coronavirus-2 infection in domestic and wild animals: a systematic review
Rao Sudhanarayani S. 1
http://orcid.org/0000-0003-1846-800X
Parthasarathy Krupakar [email protected]
1
Sounderrajan Vignesh 1
Neelagandan K. 2
Anbazhagan Pradeep 3
Chandramouli Vaishnavi 3
1 grid.412427.6 0000 0004 1761 0622 Centre for Drug Discovery and Development, Sathyabama Institute of Science and Technology, Chennai, 600119 India
2 grid.475408.a 0000 0004 4905 7710 Centre for Chemical Biology and Therapeutics, Institute for Stem Cell Science and Regenerative Medicine, Bengaluru, India
3 Advanced Institute for Wildlife Conservation, Vandalur, Chennai, India
11 12 2022
1 2023
13 1 527 8 2022
26 11 2022
© King Abdulaziz City for Science and Technology 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.
Animals and viruses have constantly been co-evolving under natural circumstances and pandemic like situations. They harbour harmful viruses which can spread easily. In the recent times we have seen pandemic like situations being created as a result of the spread of deadly and fatal viruses. Coronaviruses (CoVs) are one of the wellrecognized groups of viruses. There are four known genera of Coronavirus family namely, alpha (α), beta (β), gamma (γ), and delta (δ). Animals have been infected with CoVs belonging to all four genera. In the last few decades the world has witnessed an emergence of severe acute respiratory syndromes which had created a pandemic like situation such as SARS CoV, MERS-CoV. We are currently in another pandemic like situation created due to the uncontrolled spread of a similar coronavirus namely SARSCoV-2. These findings are based on a small number of animals and do not indicate whether animals can transmit disease to humans. Several mammals, including cats, dogs, bank voles, ferrets, fruit bats, hamsters, mink, pigs, rabbits, racoon dogs, and white-tailed deer, have been found to be infected naturally by the virus. Certain laboratory discoveries revealed that animals such as cats, ferrets, fruit bats, hamsters, racoon dogs, and white-tailed deer can spread the illness to other animals of the same species. This review article gives insights on the current knowledge about SARS-CoV-2 infection and development in animals on the farm and in domestic community and their impact on society.
Keywords
Coronaviruses
MERS
SARS CoV-2
Pandemic
Transmission
Animal models
http://dx.doi.org/10.13039/501100001411 Indian Council of Medical Research VIR/COVID19/33/2021/ECD1). Parthasarathy Krupakar issue-copyright-statement© King Abdulaziz City for Science and Technology 2023
==== Body
pmcIntroduction
The Coronaviridae family of viruses have a single-stranded, positive-sense RNA genome. They have been dormant in the environment for a long time and are popular and extensively researched family of viral pathogens in veterinary science. Every pet breed, companion animal, and wild animal has certainly experienced at least one viral infection from this family at a certain point in its life. Coronaviruses (CoVs) are majorly classified into four genera such as alpha, beta, gamma, and delta coronavirus (Perlman 2018). There was a severe outbreak of SARS two decades ago in the major parts of China (Nova 2021), which proved to be the discovery of the natural reservoir of bats family. Since then, there has been a close association as there was high similarity between the bat coronavirus with SARS-CoV, and now, the ongoing SARS-CoV-2 (Li et al. 2005), Zhou et al. (2020) reported a close association of the bat origin of SARS-COV-2 as there was 96% whole-genome sequence identity between the RaTG13 (SARS r-CoV) and SARS-CoV-2. The 800 amino acids containing the ACE-2 receptor’s first description appeared in 2000.The ACE-2 receptor is present in a wide range of animal species and tissues, such as the heart, kidneys, testes, large intestines, and kidneys. The viral spike’s receptor-binding domain RBD interacts with the ACE-2 receptor (Saravanan et al. 2022). Protein sequences from several species were analyzed to find out the similarity in the ACE-2 receptor region which plays a pivotal role in binding to the spike protein thus giving the virus an entry into the human host. There is evidence about 5 critical hotspots of ACE-2 (Position 31, 53, 38, 82, and 353) to the SARS-CoV-2 and for cross-species transmission. Sequence analysis of ACE-2 shows that the crab-eating macaque and chimpanzee share identical amino acids with humans in all the five hotspot areas proving to be ideal candidates for being animal models. Other animals like cattle and pig show four sites which are identical with a single variation sites. Ferrets and dogs possess three identical sites with two different sites. Bats and mice have only two similar amino acid with humans (Sharun et al. 2020).
The most variable region of the coronavirus genome is the receptor-binding domain (RBD) found in the spike protein (Zhou et al. 2020). It has been demonstrated that six RBD amino acids are essential for SARS-CoV-like viral host range determination and binding to ACE2 receptors (Wan et al. 2020). When compared to SARS-CoV, the main mutations include L455, F486, Q493, S494, N501, and Y505 which were found on SARS-CoV-2 (Wan et al. 2020). The crown-shaped spikes on the surfaces of coronaviruses give it the name. They are divided into the alpha, beta, gamma, and delta subgroups which are commonly found coronaviruses such as 229E (alpha coronavirus), NL63 (alpha coronavirus), OC43 (beta coronavirus), HKU1 (beta coronavirus). The viruses known to affect humans are SARS-CoV, MERS-CoV, and SARS-CoV-2 (Lim et al. 2016).
Middle East Respiratory Syndrome (MERS-CoV) was a highly contagious viral disease. Bats are thought to be the natural hosts for coronaviruses. In numerous Middle-eastern nations, camels have been shown to carry different strains of this virus. Nearly, 90% of camels and calves tested positive for MERS-CoV antibodies in Saudi Arabia, according to serological tests. According to a report, human MERS-CoV sequences and camel MERS-CoV sequences are closely linked (Alnuqaydan et al. 2021). The dipeptidyl peptidase DPP4 (also known as CD 26), which was discovered to be the host cell receptor for MERS-CoV, was one of two essential components needed for MERS-CoV to adhere to or fuse with host cells in order to begin infection. DPP4 is only expressed in specific animal species, which accounts for the limitation of the MERS-species CoV’s tropism (Alnuqaydan et al. 2021).
The H5N1 strain of the avian flu first appeared in farm-grown geese which then went on to spread in chickens. The most dangerous types of bird flu that can be spread from birds to people include H5N1 and H7N9, which have killed thousands of people in China and across the globe. People who come into contact with the feathers, flesh, or droppings of infected birds were at high risk of contracting the virus (PETA U. Coronavirus, Swine Flu, n.d. XXXX). The current pandemic, driven by the severe acute respiratory syndrome (SARS) coronavirus-2, is having a devastating effect on society and the planetary health and also serves as an indicator of multiple dangers that may rapidly evolve to cause decline of human and animal populations.
We are in a fierce battle to lessen the effects of the pandemic on human health and wellbeing. In order to survive this socioeconomic disaster, required steps must be taken everywhere to flatten the pandemic curve that SARS-CoV-2 has caused. Additionally, it has the capacity to infect most mammalian species and induce serious respiratory distress. It has been documented that humans can transmit the disease to canines and minks. In numerous nations, there has been intraspecific transmission between minks that has been closely observed and documented (WHO-FAO-WOAH 2021).
Potential source of transmission
Researchers have traced the SARS-CoV-2 outbreak back to a first identified case in massive market at Wuhan that retailed live wild animals among other goods and put forward the hypothesis that most possible reason is spillover of SARS-CoV-2 virus to humans from animals sold at the market (Maxmen 2022; Gao et al. 2022). The SARS-CoV-2 samples collected from the Wuhan market environment and infected people during December 2019 and January 2020 for genetic analysis showed many of the positive cases were found in the samples obtained from regions where live animals were sold at the wet market. The SARS-CoV-2-specific antibodies was found in all blood samples collected from the infected people in the market and from the animals for sale in the market (Gao et al. 2022). Researchers believed that the reservoir may very well be squat dog-like species, particularly raccoon dogs that were sold in the live market and had positive cases. The animals harboring the virus before transmitting it to humans were still stable(Worobey et al. 2022). Five samples tested positive, which were from single live animal shop where animals were housed and moved in metal cages, carts, and a machine used for removing bird feathers. It is believed they could have contracted the virus when they were stored in the caged containers in the same place.
The speed at which the SARS-CoV-2 is creating a pandemic is alarming. Most of the countries are in the fourth wave of the pandemic. There are high chances of SARS-CoV-2 re-infection. The infected population often produces a very high viral load; this increases the chances of spillover to other animal species such as pet animals, farm animals, and wild animals which inhabit regions close to human settlements. It is reported that the likelihood of the viral amplification in pigs does not have high occurrences, and similar instances must be monitored very cautiously in order to prevent any spillovers (Opriessnig and Huang 2020). The Centre for Disease Control and Prevention (CDC), USA, has strongly advised that all laboratory-confirmed COVID-19 cases (RT-PCR positive) limit their interaction with their companion pet animals in light of the potential for SARS-CoV-2 transmission between humans and animals (Mallapaty 2020). The Coronaviruses infection initially has cold-like symptoms which further worsens into respiratory difficulties, chest congestion in humans, while other viruses can cause illness in certain types of animals such as bats, camel, and cattle.
Some coronaviruses, which belong to canine and feline coronaviruses family, infect only animals and not humans. In an extensive study, 1914 serum samples were collected from 35 animal species with symptoms and suspected of SARS-CoV-2 infection and were subjected to ELISA with double antigen sandwich that detects SARS-CoV-2-specific antibodies. The findings indicated the absence of SARS-CoV-2-specific antibodies, which ruled out the notion that an animal species could serve as an intermediate transmission host for SARS-CoV-2 infection (Deng et al. 2020). The spread of the virus raised many concerns by the pet owners, which led them to a state of anxiety and fear which became one of the major reason for pets being abandoned and rendered homeless in many cities. This has created a huge impact on the welfare of the animals.
Genetic diversity of SARS-CoV-2
Coronaviruses are enclosed as positive-sense single-stranded RNA viruses that are members of the family Coronaviridae, sub-class Orthocoronavirinae, and are named after their surface proteins with a crown-like structure. They are divided into four genera such as alpha coronavirus, beta coronavirus, delta coronavirus, and gamma coronavirus. The first two class of the virus majorly cause infection in mammals where as in birds the source of infection is gamma coronaviruses. Delta coronaviruses infect both mammals and birds.
SARS-CoV-2 is an enveloped single-stranded RNA virus which has a genomic size of 30,000 nucleotides and eleven open reading frames (ORFs) that encode for 29 proteins. 16 non-structural proteins are encoded by the first 21,552 nucleotides of the genome, which make up ORF1ab (nsp1–nsp16). The structural proteins are spike (S), envelope (E), matrix (M), and nucleocapsid (N), and nine auxiliary proteins are encoded in the final part of the genome ORF3a, ORF3b, ORF6, ORF7a, ORF7b, ORF8b, ORF9b, ORF9c, and ORF10 (Helmy et al. 2020; Bai et al. 2022).
The SARS-CoV-2 has evolved over the period of time in multiple locations and has been classified by the World Health Organization under specific categories such as variant of interest, variant of concern, and variant under monitoring. This classification is based on the severity and progression of the SARS-CoV-2 variants. There have been several variants that had been circulating around the globe and causing loss of human and animals. There are five major variants of concern such as alpha (B.1.1.7), beta (B.1.351), gamma (P.1), delta (B.1.617.2), and Omicron (BA.1). The Omicron variant shows diversity as it has been found in multiple locations in the world; they have been sub-classified as Omicron (BA.2), Omicron (BA.4), Omicron (BA.5), Omicron (BA.2.12.1), Omicron (BA.2.75). Multiple SARS-CoV-2 variations have been identified, with a few classified as variants of concern (VOCs) owing to their public health implications. VOCs have been linked to increased transmissibility or virulence, decreased neutralization by antibodies that are produced by natural infection or vaccination, the capacity to elude detection, and a reduction in therapeutic or vaccine efficiency (Aleem et al. 2021).
All five of the reported VOCs such as have multiple mutations in the receptor-binding domain and the N-terminal domain of the spike protein, with the exception of the delta variant, which has a N501Y mutation on the RBD which has resulted in the increased affinity of the interaction between the spike protein and ACE 2 receptors, enhancing viral attachment and subsequent entry into host cells. All the major mutations encoding the structural proteins of SARS -CoV-2 and its variants are given in the following Table 1. The WHO has presently identified eight variants of interest (VOIs) since the beginning of the pandemic, which include epsilon (B.1.427 and B.1.429), zeta (P.2), eta (B.1.525), theta (P.3), iota (B.1.526), kappa (B.1.617.1), lambda (C.37), and mu (B.1.621). VOIs are identified as genetic variants with particular genetic markers that have been linked to changes that may result in increased transmissibility or virulence and decrease in the neutralization capability of antibodies generated through natural infection or due to vaccination. Table 1 Structural protein mutations (Spike (S), envelope (E), membrane, (M), and nucleocapsid (N)) of different variants of SARS-CoV-2 isolated from humans
Name of the variant Pango Lineage Origin Mutations on the Structural protein References
Spike Nucleocapsid Membrane Envelope
Alpha B.1.1.7 Dec 2020
England
N501Y, A570D, P681H, T716I, S928A, D1118H D3L, S235F – – Aleem et al. (2021)
Beta B.1.351 Dec 2020
South Africa
N501Y, E484K, L18F, K417N
D80A
T205I – P71L Tegally et al. (2021)
Gamma P.1 Jan 2021
Brazil
N501Y, E484K, D614G, H655Y P80R, R203K
G204R
– – Amanat et al. (2021)
Delta B.1.617.2 2021
India
T19R, G142D, R158G, L452R, T478K, D614G, P681R, D950N D63G, R203M
D377Y
I82T – Kannan et al. (2021)
Omicron BA.1 Nov 2021
South Africa
A67V, T95I, Y145D, L212I, G339D, S371L, S373P,S375F, K417N, N440K, G446S, S477N, T478K, E484A, Q493R, G496S, Q498R, N501Y, Y505H, T547K, D614G, H655Y, N679K, P681H, N764K, D796Y, N856K, Q954H, N969K P13L, R203K, G204R D3G, Q19E, A63T T91 Zhang et al. (2022)
Animal-to-human transmission
A lethal zoonotic virus known as SARS-CoV-2 is thought to have spread from one of the animal species to humans after infecting and spreading among other animals. A team of researchers used whole-genome sequencing to demonstrate that in the early stages of the 2021 (March) pandemic, SARS-CoV-2 infections were widespread among mink farms in the southeast of the Netherlands. 68% of mink farm employees screened positive for the infection or have had antibodies against SARS-CoV-2. The human COVID-19 patients that had viruses with the D614G mutation were what started these sizable clusters of infection. Sequencing has since demonstrated that mink-to-human transfer also took place (Munnink et al. 2021).
Another incident of animal-to-human transfer being suspected included hamsters at a pet store. A team of researchers published the results of their epidemiological and viral analyses of a COVID-19 epidemic linked to a pet shop that may have involved interspecies transmission. The 3 vaccine recipients in this cluster were found to have contracted the most closely related virus strain of the SARS-CoV-2, the delta variant AY.127, according to the epidemiological research. The hamsters at the warehouse that were harboring SARS-CoV-2 were not visited by the afflicted patient. Therefore, it seemed most likely that the patient contracted the virus at the pet store, either by close contact with sick animals or by being exposed to contaminated surroundings (Chan et al. 2020).
A case was reported in Thailand, where there was a spread of the virus from a cat to human as the lady veterinarian who had handled the cat showed severe symptoms. The cat's nasal swab’s comparatively low RT-PCR cycle thresholds indicate that the virus load was high and contagious (Sila 2022).
The Ohio area underwent a thorough research in the winter of 2021, testing white-tailed deer (O. virginianus) at random locations in nine distinct locations for the presence of SARS-CoV-2 and its variations. Out of these, four areas where further sampling was conducted to indicate it as a significant deer cluster were where the highest prevalence estimates of SARS-CoV-2 (B.1.596 viruses) were discovered. The sequencing unmistakably demonstrated that a transfer from deer to deer had occurred. A variety of mutations, including one in the receptor-binding motif, were found in white-tailed deer but only at extremely low frequency in humans. Such alterations could be enhanced in a fresh reservoirs host with significant infection rates and other evolutionary constraints (Hale et al. 2022). A schematic representation of the possible route of transmission of SARS-CoV-2 starting from animal as a primary host to the intermediate and then jumping to humans is illustrated in Fig 1. The animals infected with the virus were studied; their samples were sequenced, and the data are submitted to GISAID databank. The following table gives an overview about the mutations found in SARS-CoV-2 isolated from animal sources across the globe. This data have been retrieved and analyzed for mutations from the GISAID databank. Fig. 1 Above image shows the possible route of transmission of SARS-CoV-2 from animals starting from primary host, then to intermediate and finally to humans and animals. The image was produced using BioRender.com
Structural protein mutations in SARS-CoV-2 spike, envelope, membrane, and nucleocapsid isolated from humans for different variants at multiple time points as given in Table1. The most frequently occurring mutations were selected, compared, and analyzed with the SARS-CoV-2 viral isolates found in domestic and wild animals. Based on the analysis, we have summarized the presence and absence of the mutations in animals in Table 2.Table 2 Structural protein mutations (Spike (S), envelope (E), membrane, (M) and nucleocapsid (N)) of different variants of SARS-CoV-2 isolated from wild and domestic animals
Natural infection of SARS-CoV-2 in animals
The SARS-CoV-2 has four structural proteins, namely spike (S), membrane (M), envelope (E), and nucleocapsid (N) which collectively constitute the structure of the virion. The virion binds to the host cell receptor with the help of S1 subunit of spike protein. This interaction between the spike protein of SARS-CoV-2 and the host receptor determines the host species range and also the tissue tropism of the virus. The alpha coronavirus uses the aminopeptidase N as its receptor for interaction, whereas ACE-2 is used as a host receptor by SARS-CoV and HCoV NL63. Another group of viruses such as MHV utilizes the CEACAM1 to enter the human cells. MERS-CoV majorly binds to the dipeptidyl peptidase4 (DPP4) to enter the human cells (Hamming et al. 2004). The angiotensin-converting enzyme 2 (ACE2) in human binds to the spike protein of the SARS-CoV-2 virus (Luan et al. 2020). There have been reported evidences of transmission of SARS-CoV-2 between other domestic animals such as cats, dogs, minks, hamsters, and ferrets and some wild animals such as lions and tigers. The key ACE-2 residue plays a very important role in recognizing the spike/S1 subunit protein; it was analyzed in order to identify all the possible potential host ranges of SARS-CoV-2. The SARS-CoV-2 virus was inoculated in numerous animal model species for an experimental examination, and the results revealed that ferrets and cats are extremely vulnerable to the virus whereas dogs are relatively less susceptible. The researchers discovered that pigs, chickens, and ducks are either not at all or only marginally infected with SARS-CoV-2 (Shi et al. 2020a, b).
SARS-CoV-2 in apes
The studies conclude that, the Hominidae family of wild great apes is extremely vulnerable to a variety of viral infections that affect humans. The diseases that constantly endanger humans also represent a serious threat to global efforts to save wild animal populations. The traits of ACE2 indicate that primates, especially Old-World species, will be very susceptible to SARS-CoV-2. A study proves that there is a naturally acquired infection in captive gorillas. Direct and indirect contact of humans to primates leads to the covid infection. The interaction with humans is rising constantly due to deforestation, rehabilitation, research, and tourism. Most of the primates’ conservation habitat is geographically placed near highly populated urban areas; hence, there is a sharp rise in the susceptibility toward SARS-CoV-2 in primates. Great apes Western low land gorillas are infected with SARS-CoV-2. Infected animals had visible symptoms of coughing, nasal discharge, and some old apes’ showed pneumonia like symptoms, which reduced after two or more days. The severity of infection in wild gorillas was very high and unpredictable. Strict precautions must be taken by the all people who come in close contact with primates such as tourists, researchers, and conservation workers. Other apes belonging to different families such as captive bonobos (Pan paniscus) and orangutans (Pongo sp.) are also vulnerable to such viral infections; therefore, administration of vaccines is of prime importance (Delahay et al. 2021).
SARS-CoV-2 in dogs
In the past, when the pandemic had hit in 2003 due to the emergence of SARS-CoV, it was detected in domestic dogs (Group 2003). In the current pandemic, there has been multiple cases about the susceptibility of domestic pet mammals to SARS-CoV-2. The testing is done using RT PCR method. In a research study from Hong Kong, it was discovered that 2 out of 15 dogs from residents that had confirmed human cases of COVID-19 were found to be infected with SARS-CoV-2. In a second investigation, the virus was isolated from nasal and oral swabs, and the 2.5 year-old male German shepherd was confirmed to be positive for SARS-CoV-2 RNA on two separate occasions. Plaque-reduction-neutralization assays were used to evaluate the dog’s antibody responses. The genetic material of the viruses obtained from the two dogs matched those of the corresponding human cases. These indications reflect the instances of SARS-CoV-2 transmission from humans to animals (Goumenou et al. 2020).
Dogs may have served as intermediary hosts in the spread of SARS-CoV-2 among people in Italy, according to a group of researchers. This assumption was supported by the finding that, despite the implementation of stringent laws and regulations in Italy, the number of cases increased rapidly. There is one dog for every six people living in Italy, the likelihood of both human-to-human and animal-to-human transmission. A case was studied in the United States Department of Agriculture’s (USDA) National Veterinary Services Laboratories (NVSL), where a pet dog (German shepherd) had showed signs of respiratory distress. Swab samples from two dogs were taken after its owner tested positive. One of the dogs showed symptoms whereas the other did not have symptoms. Antibodies were also detected in both samples. This showed transmission from human to animals. SARS-CoV-2 infections have been reported in a small number of animals across the world, mostly in animals that are in close contact with an infected person.
The RBD of dog ACE-2 complex and the RBD of human ACE-2 complex have several structural similarities. The two complexes’ interaction interfaces, however, are marginally dissimilar. The RBD/dACE-2 interface has significantly fewer contact atoms, residues, and hydrogen bonds than the RBD/hACE-2 interface, which accounts for the 6.65 times lower binding efficiency of dACE-2 for RBD than that of hACE-2.
SARS-CoV-2 in cats
Considering that cats were previously known to have been infected with severe acute respiratory syndrome coronavirus SARS-CoV during the 2003 outbreak; their sensitivity to severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) was predicted to be more transmissible. SARS-CoV-2 has been discovered to replicate in domestic cats, and experimental cat-to-cat viral transmission has been noted. Infection from COVID-19-positive pet owners via human-to-cat SARS-CoV-2 transmission has been confirmed in reports from various countries. The disease manifested in varying degrees of severity in the affected cats (Halfmann et al. 2020). In New York, two pet cats tested positive. One of the pet owners was a symptomatic virus carrier. Both animals had a slight respiratory infection that caused breathing difficulties. There may have been a human-to-animal transfer or an animal-to-human transmission, according to reports. A study was conducted for the alleged infection that 22 pet cats may have had. The owners themselves were thought to have contracted the disease or to be the source of it. Rectal, nasopharyngeal, and biological fluids were used to collect the swabs for the RT-qPCR test, which looked for 2 SARS-CoV-2 genes. Antibodies to the SARS-CoV-2 were found, according to a serological investigation. Additionally, it was revealed that the SARS-CoV-2 genome sequence analysis belonged to clade a2a, which shared similarities with the human SARS-CoV-2 samples isolated from France (Sailleau et al. 2020).
SARS-CoV-2 in minks
First confirmed case of SARS-CoV-2 was reported in Netherlands inside a mink farm in the first wave of the SARS-COV-2 outbreak. The neighboring farm had a probable suspected case; both farms exhibited respiratory difficulties, which led to a greater death rate. The farm mink may have been infected after being exposed to a virus-infested droplet (Oreshkova 2020). A farm worker was suspected to have contracted SARS-CoV-2 from mink raised for food. Four distinct mutations were found in the S gene that gave birth to the mink adapted variant (G75V, M177T, Y453F, and C1247F) in an isolate taken from a farm worker who tested positive for SARS-CoV-2. The additional alterations also included the Y453F mutation, which had been previously noted to have emerged in mink during sequential passages, and a unique mutation that truncates ORF 7b at location L22 and is absent in any other global SARS-CoV-2 isolate.
The virus’s apparent adaptation to the mink host led to the emerging variation, with some point mutations being repaired early during the mink’s dissemination and further modifications accumulating over time (Bayarri-Olmos et al. 2021; Rabalski et al. 2022). Scientists examined over 91 mink samples for SARS-CoV-2 infection. It was validated that 15 animals screened tested positive for SARS-CoV-2 combining detection technique such as RT- PCR, antigen-based detection, and NGS technique. On sequencing the whole viral genomes, researchers validated this discovery that there was presence of virus variation with sporadic mutations on the protein sequence such as G75V and C1247F which coded for spike protein (Rabalski et al. 2021).
SARS-CoV-2 in Asiatic lion
In one of the biggest zoos in Asia, the Nehru zoological park in Hyderabad, eight of the Asiatic lions were found to be positive against SARS-CoV-2 during the second wave of covid-19 which was at its peak during April 2021–June 2021, and more than two dozen of the zoo staffers had been tested positive at the time. This reported case was one of a kind to prove that there is human-to-animal transmission. There was an initial suspect when one of the caretakers found that these big cats developed symptoms such as dry cough, nasal discharge, and loss of appetite. These alarming symptoms were soon addressed by the in-house veterinary officials who took the oropharyngeal swab samples for further testing to test and find out that they were positive for SARS-CoV-2.
In Nahargarh Biological Park, India, one of the lions was displaying mild clinical symptoms, leading to discovery of an additional case of the feline SARS-CoV-2 infection. Testing revealed a positive result for SARS-CoV-2 employing primers and probes targeting RdRP, the E gene, and RT-PCR results were verified by Sanger sequencing of the S, N, and E genes. A further 14 days of isolation and quarantine followed. Testing of the entire zoo crew indicated that the veterinarian who was caring for the sick lions also experienced clinical symptoms (Karikalan et al. 2021). A similar case of feline infection was reported in Singapore zoo, where four endangered Asiatic lions had started to show symptoms such as coughing and sneezing after they had contact with one of the zoo keeper who was found to be infected. They were immediately isolated and quarantined in a separate den. They remained bright and active all the while and showed signs of improvement.
Two Asiatic lions from Vandalur Zoo in Chennai died due to severe infection of SARS-CoV-2. Nine of the lions were tested during the time. Of the two, one was 9 year-old lioness, and the other was 12 year-old male. This test report was issued by NISHAD in Bhopal. The lion had been kept in the intensive care unit and was receiving treatment since a long time. These two cases were reported of feline deaths in India (Mishra et al. 2021).
SARS-CoV-2 in white tailed deer
The ACE2 receptor, which is the primary entry point for the viral invasion, is shared by humans, reindeer, and Pere David’s deer (Damas et al. 2020).
SARS-CoV-2 was also reported in several deer spp. The contagious virus was discovered in deer excrement. RT-PCR was used to analyze the samples, which were tested from nasal and rectal swabs. In several cases, the body temperature was temporarily increased, but there are no evident lesions that were visibly seen. Acute alveolar damage-related lesions were seen histologically, that resembled infections in humans, but no viral RNA was discovered, suggesting the pathogen had already been wiped off (Michelitsch et al. 2021).
The main risk that is associated with the ongoing pandemic caused by SARS-CoV-2 outbreak is the transmission of the virus from human to human. It is true to believe that the persisting evidences show and support that the plausibility of the deadly SARS-CoV-2 spilling over to the undiscovered new hosts through medium involving fecal shedding by infected humans which reaches the natural aquatic environment through the waste water treatment system (Franklin and Bevins 2020).
SARS-CoV-2 in avian species
A group of researchers studied 5 different species of poultry such as chicken (Gallus gallus domesticus), turkey (Meleagris gallopavo), quail (Coturnix japonica), Pekin ducks (Anas platyrhinchos domesticus), and white Chinese geese (Anser cygnoides). They challenged 10 birds from each species with SARS-CoV-2 virus obtained from the Biodefense and Emerging Infectious Resources Repository, USA. The swab samples particularly the oropharyngeal and cloacal were collected and analyzed for the presence of the virus at different time intervals. Their serum was also analyzed for antibody levels and the virus neutralization. The results revealed absence of viral antibodies even at a window period of 14 days post the challenge. These experiments clearly state that the virus did not replicate in any of the avian species (Suarez et al. 2020).
Challenges in animal model
The search for a preclinical animal model that accurately replicates the severe and fatal type of human COVID-19 is one of the major limitations in the advancement of SARS-CoV-2 infection models. It will be an advantage for several groups of researchers working on animal models. Furthermore, it would offer a method for evaluating the transition from a moderate to a severe illness, which may help to identify disease and its biomarkers. Additionally, it would broaden the scope of currently existing animal models for testing vaccinations and treatments, which may lead to the development of critically required medical countermeasures during rescue situations.
There are still a number of unsolved concerns about how SARS-CoV-2 is spread. Are there any other possible transmission routes, such as mother-to-child or fecal contamination. Due to the disease’s unexpected outcomes, it has been challenging to comprehend the pathophysiology of the condition. The disease is more infectious than MERS and SARS for what reasons? Why are older people and others who have underlying medical conditions more vulnerable (Rothan 2020). Does the host’s sex or genetic composition also affects pathogenicity? More study is required, and these issues should be examined using physiologically plausible animal models. Most of the SARS-CoV-2 comorbidities and co-infections remain unknown. Concerns exist regarding why some drugs that have been repurposed for COVID-19 therapy are effective for some infected people. As a result, in vivo research using the most effective animal models is necessary to gather preliminary information. The wide variation in human genetic make-up makes it difficult to understand the workings of processes.
Animal models such ferrets, mice, and hamsters can be used to study a wide range of experiments including the mechanistic of action of antivirals, the efficacy and safety of vaccines, and the impact of comorbidities on the development of COVID-19 (Olwenyi et al. 2020). As a result, in vivo research using the most effective animal models is necessary to gather first information. The wide variation in human genetic make-up makes it difficult to understand the workings of processes. Animal models such ferrets, mice, and hamsters can be used to study a variety of topics, including the mechanism of action of antivirals, the efficacy and safety of vaccines, and the impact of comorbidities on the development of COVID-19. No one animal model is likely to be able to handle all the issues associated to translating from humans because of inherent differences in the development and physiology of the organism (Pandey et al. 2021). Furthermore, there is a great deal of heterogeneity among the animals, including variances in their biology, genetics, and level of ACE2 receptor expression, all of which might impact the infection rate. Understanding SARS-CoV-2 pathophysiology requires the use of primary cell lines, organoids, and in vitro models (Leist et al. 2020). The creation of humanized mouse models with tissues similar to those of humans may also pave the door for fast testing of antivirals and vaccines.
To further prove the safety and effectiveness of potential therapies and vaccines, researchers can use a number of small and large animal models to study critical aspects of SARS-CoV-2 including pathogenesis, transmission, and host reactions to SARS-CoV-2. In order to measure the pathogenic pathways and prospective treatments, it is essential for this effort to advance investigations on several animal models of COVID-19. To this purpose, research for SARS-CoV-2 has made good use of a variety of animal species, such as mouse, great apes, hamsters, ferrets, and cats. A wide variety of experimental animal model for SARS-CoV-2 infection and their application is widely discussed in Fig. 2Fig. 2 Schematic representation of experimental animal model for SARS-CoV-2 with their application in research and development
Mouse models
For several viral experiments, the mouse model has indeed been widely employed. The hepatitis B virus (HBV), hepatitis C virus (HCV), cytomegalovirus (CMV), Zika virus, and other viruses have all been studied using it as the best small animal model. The mouse model is appropriate for extensive research on viruses, both for pathogenesis and antiviral treatments, due to its economic cost, compact size, simple operation, and excellent repeatability. Mice are easier to handle as compared to non-human primates in a laboratory with a higher level of biosafety because of their comparatively small size and lack of operational challenges. Importantly, mice may be easily genetically altered for accurate study. For investigations on viral infection and transmission limitation, viral pathogenesis, and antiviral immunity, a large number of genetically altered mice are available. Furthermore, it is possible to research viral pathogenesis and host immune responses using modern mouse immunological tools.
Several viral investigations, such as those involving the Middle East respiratory syndrome coronavirus and SARS-CoVs, have made good use of the mouse (Mus musculus) model (Bi et al. 2021). Since SARS-CoV-2 penetrates host cells primarily through the human angiotensin-converting enzyme 2 (ACE2) receptor rather than the mouse ACE2 receptor, the main challenge to SARS-CoV-2 infection in mice is a lack of appropriate receptors (Bi et al. 2021). The strategies implemented for using mouse as an animal model will involve alterations in the animals to express human ACE-2. To introduce hACE-2 in mice, CRISPR/Cas9 knock-in technology or recombinant plasmids are used to construct permanent genetic alterations. The human lung ciliated epithelial cell-specific HFH4/FOXJ1 promoter, the human epithelial cell cytokeratin-18 (K18) promoter, and the mouse ACE2 (mACE2) promoter all regulate the production of hACE-2 (Zhao et al. 2020a, b). By employing reverse genetics or serial passaging, viruses can be modified to directly infect wide-type mice. The SCID mice were surgically implanted with human fetal lung tissue under their dorsal skin to create a SARS-CoV-2 xenograft model. When infected with SARS-CoV-2, the human lung xenografts exhibited high viral replication with spreading to the entire lung tissue and generated mature structures that nearly resemble the normal human lung (Fu et al. 2021).
Systems based on adenovirus vectors are advantageous for usage with transgenic or factor-deficient mice because they may be utilized to incorporate human receptors into mouse genomes. hDDP4 has been inserted into WT mice using replication-defective adenovirus vectors, making them vulnerable to MERS-CoV infection. Mice that were infected developed pneumonia with significant pulmonary immune cell infiltration and viral clearance by 6 days post-infection (dpi) although they suffered less damage than mice that had completely hDPP4 transgenic bodies (Zhao et al. 2014).
Syrian hamster models
In order to study respiratory viruses, hamsters are extensively employed. The SARS-CoV-2 spike protein is strongly bound by the hamster ACE2 protein, which facilitates entrance. In order to explore the pathophysiology of SARS-CoV-2 infections, hamsters are a prospective infection model (Shou et al. 2021). This model is effective for researching infectious biology, including post-bacterial, viral, and parasitic diseases, because the immunological reactions of Syrian hamsters to pathogenic organisms are comparable to those of humans. It also helps in evaluating the effectiveness and inter-connections of prescribed drugs and vaccines for those pathogens (Miao et al. 2019). The ACE 2 receptor is reported to interact with the most epitopic area of SARS-CoV-2 structural spike glycoprotein. In a study, scientists found that Syrian hamsters are susceptible to SARS-CoV-2 infection (Chan et al. 2020). Syrian hamsters had pathological symptoms after infections that were comparable to those in humans. Focal, diffuse alveolar damage, hyaline membrane development, and mononuclear cell infiltration were all visible in the lung tissue. The Syrian hamster may develop lesions in the kidney, adrenal gland, ovary, spleen, lymph nodes, and other numerous organs depending on the severity of the coronavirus-2 infections. One of these, focal to multifocal inflammation, was seen in the adrenal gland (Song et al. 2021). Hamsters were immune from secondary infection by neutralizing antibodies produced due to the initial SARS-CoV-2 infection. Likewise, passive serum transfer protected the naive hamsters from viral lung replication (Imai et al. 2020).
Ferret models
Numerous human respiratory viruses, including the influenza virus, syncytial virus, para influenza virus, and coronavirus, can infect ferrets. Due to the existence of viral receptors and the similarities in their respiratory tract structure to that of humans, ferrets can mimic the clinical signs of viral infections. In ferrets’ tracheobronchial sub-mucosal glands, type-II pneumocytes and serous epithelial cells primarily express ACE-2. Only, two amino acids differentiate the ferret ACE-2 domain from its human counterpart domain when it comes to binding the spike protein of the SARS-CoV-2 virus.
Ferrets (Mustela putorius furo) have indeed been found to be an extremely useful model for evaluating the virulence and propagation of human respiratory viruses such as influenza and respiratory syncytial virus. It is not unusual, then, that the ferret model has been examined for research into the pathophysiology of SARS-CoV-2 spread. Despite the use of various SARS-CoV-2 isolates, the results were remarkably comparable across all laboratories (Shi et al. 2020a, b). In laboratory conditions, ferrets can easily spread viruses to uninfected ferrets. Transfer between experimentally infected ferrets to naive cage mates happened efficiently, while transmission between ferrets exposed to companion ferrets separated by steel grids did not. These investigations demonstrated the possibility of SARS-CoV-2 airborne transmission and recommended that future research on transmission may benefit from using the ferret model (Richard et al. 2020).
Tree shrew models
An animal resembling rat and squirrel-like is known to be Chinese tree shrew (Tupaia belangeri chinensis) which shares the genetic affinity with primates. Tree shrews have been a new experimental animal that has been successfully tested as a substitute to primates in several researches and drug safety trials which majorly focused on research related to hepatitis C and B viruses. Tree shrews have the benefits of being small in size, low feeding costs, and short reproductive cycles (Fan et al. 2013).
In an experimental context, SARS-CoV-2 was delivered to males as well as female tree shrews of different ages, ranging from 6 months to 7 years. After inoculation, the majority of animals, particularly females, showed an elevation in body temperature but no clinical signs or obvious lesions. Particularly, in younger animals, viral RNA has been discovered in blood samples, nasal samples, and also in throat swabs for up to 12 days. Compared to the existing animal models, the tree shrew is a little less vulnerable to SARS-CoV-2 infection and might not be an appropriate animal for COVID-19-related research. However, as an asymptomatic carrier, the tree shrew may serve as a significant intermediate host for the SARS-CoV-2 virus (Zhao et al. 2020a, b).
Pig models
Pigs are frequently employed in research due to their anatomical, genetic, physiologic, and immunological parallels to humans. Indeed, pig experiments are more reliable than rodent experiments to indicate therapeutic and preventive treatments for humans (Meurens et al. 2012). Piglets can be used as model animals as they were injected with the severe acute respiratory syndrome coronavirus-2 by a variety of methods, including intranasal, intratracheal, intramuscular, and intravenous inoculation. Seroconversion was reported in pigs inoculated parenterally, despite the fact that piglets were not sensitive to SARS-CoV-2 and lacked lesions or viral RNA in tissues/swabs intramuscularly or intravenously (Vergara-Alert et al. 2021). Another group of scientists challenged chickens, turkeys, ducks, quail, and geese with severe acute respiratory syndrome coronavirus-2 or Middle East respiratory syndrome coronavirus, observed no disease and detected no virus replication and no serum antibodies. It was concluded that poultry is unlikely to serve a role in maintenance of either virus.
Dog model
A group of scientists examined at SARS-CoV-2 replication and transmission in canines. Five three-month-old beagles were intra-nasally immunized and kept in a room with two other beagles that were not immunized. Each beagle’s oropharyngeal and rectal swabs were obtained every other day for virus quantification in Vero E6 cells and viral RNA detection. On days 2 and 6, viral RNA was found in the rectal swabs of two virus-inoculated dogs, but no viral RNA was found in any organs or tissues removed from this animal. On the 14th day following infection, serum from all dogs was taken for ELISA antibody detection. These findings suggest that dogs are not very susceptible to SARS-CoV-2 (Shi et al. 2020a, b).
Cat model
Wild cats have been shown to carry the SARS-CoV-2 infection and show detectable symptoms (Mallapaty 2020). Cats are sensitive to experimental infection and have found to shed the viral particle through the nasal turbinates, soft palates, tonsils, tracheas, lungs, and small intestines. Limited shedding of virus via that route was found to be in the tissues except intestine or feces (Rudd et al. 2021; Johansen et al. 2020). Viral RNA had been eliminated from the lungs but managed to remain in other tissues after 6 days post-infection. Although juvenile cats had longer viral RNA shed in lung tissue, their upper respiratory tract had lower levels of viral. In line with this, infected wild cats displayed varying degrees of respiratory problems.
Non-human primates model
Rhesus macaques, cynomolgus monkeys, common marmosets, and African green monkeys, sometimes known as Chlorocebus aethiops, are non-human primates that have been used to study the pathogenesis of SARS-CoV-2 and test therapeutic strategies (Zhao et al. 2022). A study was conducted in which young and old cynomolgus macaques were exposed to the SARS-CoV-2 strain using a combination of routes of administration such as intratracheal and intranasal routes. Except for one old macaque that experienced nasal discharge after the 14th day following vaccination, no other macaques specifically showed any overt clinical signs or weight loss. Early in the infection, the virus was mostly discharged from the throat and nose. It is noteworthy that on the fourteenth day, a macaque shed virus RNA in rectal swabs. During the autopsies of four macaques, SARS-CoV-2 RNA was discovered in several tissues from the tracheobronchial lymph nodes, ileum, and respiratory tract (Rockx et al. 2020). When the SARS-CoV-2 virus was delivered to rhesus macaques in a study, it was shown that the animals exhibited high viral loads both in the upper and lower tracts and showed humoral and cellular immune responses. Most of the symptoms showed signs of viral pneumonia. A set of animals were re-exposed to SARS-CoV-2 after the initial viral clearance, and they had lower median viral loads in both bronchoalveolar lavage and nasal mucosa compared to before the initial infection (Chandrashekar et al. 2020).
Another study was done on rhesus macaques by exposing to SARS-CoV-2 developed a respiratory infection. Radiographs of the lungs of infected animals revealed pulmonary infiltrates, changes in respiratory pattern, alterations in piloerection, lack of appetite, hunchback, pale skin, dehydration, and weight loss. The macaque model essentially recapitulates the pathological hallmarks of COVID-19, as demonstrated by the fact that three out of four animals experienced mild to severe interstitial pneumonia with symptoms such as thickening of the alveolar septa, alveolar edema, and hyaline membranes (Munster et al. 2020). The vaccinations appeared to protect non-human primates because when the vaccinated animals contracted viral SARS-CoV-2, sampling of their bronchoalveolar lavage as well as nasal mucosa showed lower virus levels. Furthermore, the human monoclonal antibody CB6 protected rhesus macaques infected with SARS-CoV-2 by lowering viral titters and preventing pathological lung damage, suggesting that CB6 might be used as a COVID-19 therapy (Shi et al. 2020a, b).
According to recent research, African green monkeys (AGMs), another non-human primate, experienced strong SARS-CoV-2 replication and acquired a severe respiratory illness. SARS-CoV-2 shedding from the pulmonary and gastrointestinal tracts was also observed in AGMs, which may closely resemble the cases in humans. In conclusion, the evaluation of COVID-19 therapeutics and vaccines may have another alternative to using AGMs as animal models (Woolsey et al. 2021) (Hartman et al. 2020).
Apart from ethical implications, the main concern is eventually the choice of animal model to be used in the research for SARS-CoV-2.However, the local infrastructure and resources have a significant influence on this choice. Small animal models like hamsters, ferrets, or mice are the sole practical experimental strategy for many researchers.
Vaccination for animals
SARS-CoV-2 variations are now being studied for their impact on the effectiveness of existing vaccinations, treatments, and diagnostics. The SARS-CoV-2 virus has been associated with cross-species jumping in animals, raising zoonotic concerns about the possibility of reintroduction into human populations by interspecies transmission between people and animals. Domesticated animals such as cats, ferrets, and dogs, captive animals such as lion, tiger, puma, gorilla, and leopard, and wild and farmed minks have already been reported to have SARS-CoV-2 infections. The emergence of SARS-CoV-2 further into feral population and its subsequent transmission to wildlife may be prevented if domestic animals were vaccinated. Although it might appear unreasonable from a public health point of view to vaccinate sensitive and susceptible animal species such as dogs, minks, and cats, the complete eradication of SARS-CoV2 will still demand control over the transmission across all vulnerable animal species. This is needed to avoid SARS-CoV-2 from re-emerging in the future.
Vaccinations of domestic animals such as cats can help reduce the spread of SARS-CoV-2 to feral cats and subsequent transmission to wildlife. A coordinated effort has been displayed between Applied DNA Sciences (United States) and Evvi Vax (Italy) in actively developing a vaccine name Linear DNA™, a potential COVID-19 vaccine candidate to be used for cats. This vaccine has successfully acquired approval from the United States Department of Agriculture (USDA) to proceed to clinical studies in domestic felines and further assess its safety and immunogenicity (Brook 2020). Canines, lions, leopards, mice, and rabbits can all be treated with a new vaccine developed in India called Ancovax. The inactivated vaccine is developed using the highly epitopic component of the delta variant of SARS-CoV-2. Alhydrogel is also used as an adjuvant to strengthen the immunological response (Dutt 2022). A Russian-based vaccine called Carnivac Cov, especially designed for animals such as cats, dogs, minks, and foxes, has proven to be effective, and it is also said to be working against the mutated SARS-CoV-2 and its variants. It is also reported to prevent mutations in animals (Tétrault-Farber and Vasilyeva 2021).
Current challenges
In order to understand the etiology of acute infectious diseases and create vaccines and medications, animal models are vital. Animal Biosafety Level 3 (ABSL-3) facilities must be used for any animal investigations containing risk Grade 3 agents, including such SARS-CoV, HIV, Mtb, and H7N9 (Guo et al. 2019). Animal Biosafety Level 3 is the pre-requisite for working with the viruses and pathogens belonging to Risk group 3 micro-organisms (DBT 2018). The Risk Group 3 viruses involve SARS-CoV-2, MERS-CoV-2, Monkey Pox Virus, HIV, Swine Flu Virus, and West Nile Virus. Higher BSL-3 precautions are necessary in scenarios involving virus growth, zoonosis probable aerosol exposure, handling diverse and highly transmissible variants, involving laboratory animals (Yeh et al. 2021). In Nashik, Maharashtra, India, the nation's first mobile Biosafety Level 3 containment laboratory was unveiled. The purpose of the mobile laboratory is to look into viral illnesses that are very contagious and have the potential to be fatal to people. It is important to set up various BSL labs in order to be prepared for any pandemic like situations.
Discussion
This paper offers a summary of the present understanding of the connection between SARS-CoV-2 infections in animals. Based on a common assumption, SARS-CoV-2 started in bats and spread to humans through an unidentified intermediary host. Although it has not been clear about the primary source contributing to the SARS-CoV-2 pandemic, a number of animal species either naturally contracted the virus after coming into touch with an infected person or as a result of experimental infection. It has been suggested that a number of animal models might be used to evaluate potential SARS-CoV-2 vaccines or assess the effectiveness and safety of antiviral drugs.
The animal model proves to be an excellent idea to test the nature and function of the virus infection. The examples of the animal models include mice, hamster models, cat, ferret, and primate models. A well strategized plan should be emphasized which collaborates the interdisciplinary cooperation among several medical and research fields such as public health, environmental sciences, veterinary medicine, and social sciences which are the prime areas where strict and efficient measures should be followed to mitigate the spread of the virus. Combined approaches of classical epidemiology and contemporary biomedicine have helped us discover a great deal about SARS-CoV-2 in the recent year. It is astonishing how quickly the novel synthetic recombinant vaccines against SARS-CoV-2 are being developed, and recent breakthroughs may open the door for other RNA-based vaccinations in the future.
The new recombinant vaccines include the viral protein's mRNA and trigger a cascade that results in the production of potent, neutralizing antibodies, in contrast to conventional vaccinations, which typically provide inactivated viral proteins to create an immunizations process. An inactivated SARS-CoV-2 delta vaccine for animals is the Ancovax Vaccine for Equines Both the Delta and Omicron Variants of SARS-CoV-2 are neutralized by the immunity brought on by Ancovax. The vaccine includes Alhydrogel as an adjuvant and inactivated SARS-CoV-2 (Delta) antigen. Russian scientists have developed a vaccine to counteract the danger of animal-to-human transfer and the emergence of mutant versions. The Karnivak-Kov vaccination strategy has been successfully launched. The new vaccines appear to have excellent translation effectiveness even in light of the many SARS-CoV-2 mutations that are now evolving. Thus, they may be especially helpful for senior citizens who are maintaining companion pet animals. Despite the minimal risk shown by experimental and epidemiological evidence, a generalized warning against the possible pet-to-human transmission is however required while the probability of SARS-CoV-2 transmission between dogs and people is especially low. The most exposed and vulnerable members of the population in many nations begin to receive protection through vaccination programs that are successful. To prevent interaction between sick humans and animals, it is advised to practice basic hygiene precautions.
It has been suggested that a number of animal models might be used to evaluate potential SARS-CoV-2 vaccines or determine the efficacy and safety of antiviral medications. To mitigate the virus from re-emerging in the future, it is important to vaccinate vulnerable animal species like cats, minks, and great apes. This may appear preposterous, but it is the only way to successfully eradicate SARS-CoV-2 is by limiting the transmissions in all vulnerable animal species. A required vaccination passport should also be implemented in order to transfer domestic animals as well as wild and captive animals across international boundaries.
To mitigate the spread of SARS-CoV-2 across wild and domesticated animal species, a stringent transit restrictions must be implemented across international boundaries, and passive or active surveillance frameworks for domesticated, confined, and wild animals must be developed. An attempt to stop the spread of SARS-CoV-2 to local domestic and wild populations of animals and its ensuing re-emergence will be strengthened by such a move.
Conclusion and future prospects
It is very interesting to note that there is a lot of similarity that exists between SARS-CoV-2 and SARS-CoV in several aspects such as the susceptibility of cats and ferrets, transmission of infection to cage mates, and resistance of chickens and pigs to infection. This has given a new path for researchers where one can understand the nature of infection in the animal and wildlife. In order to conserve nature and to protect the wildlife species, it is very important to understand the new emerging viruses and its pathogenesis. It is very important to establish BSL-3 facilities at every regional level which will help in undertaking research related to Risk Group 3 and Risk Group 4 viruses and pathogens in animals. Investigations undertaken on different types of animal model also give us more clarity regarding the same. Large-level vaccination drive should be undertaken against most neglected diseases which can pose a threat in future. It is very important to break the chain of human-to-animal transmission and the animal-to-human transmission; this will only be possible if a large population of mankind and wildlife are vaccinated against a wide range of infectious viruses. Also, the wildlife and zoo authorities must always keep an eye on the infections that the inhabited animals may potentially carry. The petting zoos and animal park that allow the human visitors to interact with animals must be thoroughly screened. This can help in increasing the life expectancy of the wildlife and will also prevent them from becoming endangered or extinct.
Acknowledgements
The authors would like to acknowledge Indian Council of Medical Research (ICMR) (VIR/COVID19/33/2021/ECD-1) for funding.
Data availability
Data of the manuscript will be provided on request.
Declarations
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
==== Refs
References
Aleem A, Bari Akbar Samad A, Slenker AK (2021) Emerging variants of SARS-CoV-2 and novel therapeutics against coronavirus (COVID-19). StatPearls. http://www.ncbi.nlm.nih.gov/pubmed/34033342
Alnuqaydan AM Almutary AG Sukamaran A Yang BTW Lee XT Lim WX Ng YM Middle east respiratory syndrome (MERS) virus—pathophysiological axis and the current treatment strategies AAPS PharmSciTech 2021 10.1208/s12249-021-02062-2
Amanat F Thapa M Lei T Sayed SM Ahmed DC Adelsberg JM Carreno SS The plasmablast response to SARS-CoV-2 MRNA vaccination is dominated by non-neutralizing antibodies that target both the NTD and the RBD MedRxiv 2021 10.1101/2021.03.07.21253098
Bai C Zhong Q Gao GF Overview of SARS-CoV-2 genome-encoded proteins Sci China Life Sci 2022 65 2 280 294 10.1007/s11427-021-1964-4 34387838
Bayarri-Olmos R Rosbjerg A Johnsen LB Helgstrand C Bak-Thomsen T Garred P Skjoedt MO The SARS-CoV-2 Y453F mink variant displays a pronounced increase in ACE-2 affinity but does not challenge antibody neutralization J Biol Chem 2021 296 100536 10.1016/j.jbc.2021.100536 33716040
Bi Z Hong W Yang J Shuaiyao Lu Peng X Animal models for SARS-CoV-2 infection and pathology MedComm 2021 2 4 548 568 10.1002/mco2.98 34909757
Brook S (2020) Applied DNA, EvviVax & GVS receive regulatory approval to begin veterinary clinical trial for lead linear DNA COVID-19 vaccine candidate. http://pharmabiz.com/NewsDetails.aspx?aid=133853&sid=2
Chan JF Woo AJ Zhang SY Poon VKM Chan CCS Lee ACY Chan WM Simulation of the clinical and pathological manifestations of coronavirus disease 2019 (COVID-19 Clin Infect Dis 2020 71 9 2428 2446 32215622
Chandrashekar A Liu J Martino AJ McMahan K Mercad NB Peter L Tostanosk LH SARS-CoV-2 infection protects against rechallenge in rhesus macaques Science 2020 369 6505 812 817 10.1126/science.abc4776 32434946
Damas J Hughes GM Keough KC Painter CA Persky NS Corbo M Hiller M Broad host range of SARS-CoV-2 predicted by comparative and structural analysis of ACE2 in vertebrates Proc Natl Acad Sci USA 2020 117 36 22311 22322 10.1073/pnas.2010146117 32826334
DBT (2018) Regulations and guidelines on biosafety of recombinant DNA Research & Biocontainment 2017.” http://www.dbtindia.nic.in/wp-content/uploads/Regulations-Guidelines-for-Reocminant-DNA-Research-and-Biocontainment-2017.pdf
Delahay RJ de la Fuente J Smith GC Sharun K Snary EL Flores Girón L Nziza J Assessing the risks of SARS-CoV-2 in wildlife One Health Outlook 2021 10.1186/s42522-021-00039-6
Deng J Jin Y Liu Y Sun J Hao L Bai J Huang T Lin D Jin Y Tian K Serological survey of SARS-CoV-2 for experimental, domestic, companion and wild animals excludes intermediate hosts of 35 different species of animals Transbound Emerg Dis 2020 67 4 1745 1749 10.1111/tbed.13577 32303108
Dutt A (2022) India’s first covid-19 vaccine for animals : why the need was felt. The Indian Express June 11, 2.
Fan Yu Huang ZY Cao CC Chen CS Chen YX Fan DD He J Genome of the Chinese tree shrew Nat Commun 2013 10.1038/ncomms2416
Franklin AB Bevins SN Spillover of SARS-CoV-2 into novel wild hosts in North America: a conceptual model for perpetuation of the pathogen Sci Total Environ 2020 733 139358 10.1016/j.scitotenv.2020.139358 32416535
Fu W, Wang W, Yuan L, Lin Y, Huang X, Chen R, Cai M (2021) Theranostics A SCID mouse-human lung xenograft model of SARS-CoV-2 infection. Theranostics 11(13):6607. 10.7150/thno.58321
Gao G, Liu W, Wong G, Wang J, Wang F, Li M (2022) Surveillance of SARS-CoV-2 in the environment and animal samples of the Huanan seafood market. Res Square. 10.21203/rs.3.rs-1370392/v1
Goumenou M Spandidos DA Tsatsakis A Possibility of transmission through dogs being a contributing factor to the extreme Covid-19 outbreak in North Italy Mol Med Rep 2020 21 6 2293 2295 10.3892/mmr.2020.11037 32236594
Guo M Wang Y Liu J Huang Z Li X Biosafety and data quality considerations for animal experiments with highly infectious agents at ABSL-3 facilities J Biosaf Biosecurity 2019 1 1 50 55 10.1016/j.jobb.2018.12.011
Hale VL Dennis PM McBride DS Nolting JM Madden C Huey D Ehrlich M SARS-CoV-2 infection in free-ranging white-tailed deer Nature 2022 602 7897 481 486 10.1038/s41586-021-04353-x 34942632
Halfmann PJ Masato Hatta DVM Chiba S Tadashi Maemura DVM Fan S Transmission of SARS-CoV-2 in domestic cats New England J Med 2020 383 6 590 92 10.1056/nejmc2011400 32402155
Hamming I Timens W Bulthuis MLC Lely AT Navis GJ van Goor H Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A first step in understanding SARS pathogenesis J Pathol 2004 203 2 631 637 10.1002/path.1570 15141377
Hartman AL Nambulli S McMillen CM White AG Tilston-Lunel NL Albe JR Cottle E SARS-CoV-2 infection of African green monkeys results in mild respiratory disease discernible by PET/CT imaging and shedding of infectious virus from both respiratory and gastrointestinal tracts PLoS Pathog 2020 16 9 1 24 10.1371/journal.ppat.1008903
Helmy YA Fawzy M Elaswad A Sobieh A Kenney SP Shehata AA The COVID-19 pandemic: a comprehensive review of taxonomy, genetics, epidemiology, diagnosis, treatment, and control J Clin Med 2020 10.3390/jcm9041225
Imai M, Iwatsuki-Horimoto K, Hatta M, Loeber S, Halfmann PJ, Nakajima N, Watanabe T et al (2020) Syrian hamsters as a small animal model for SARS-CoV-2 infection and countermeasure development. Proc Nat Acad Sci U.S.A. 117(28):16587–95. 10.1073/pnas.2009799117
Johansen MD Irving A Montagutelli X Tate MD Rudloff I Nold MF Hansbro NG Animal and translational models of SARS-CoV-2 infection and COVID-19 Mucosal Immunol 2020 13 6 877 891 10.1038/s41385-020-00340-z 32820248
Kannan SR Spratt AN Cohen AR Hasan Naqvi S Chand HS Quinn TP Lorson CL Byrareddy SN Singh K Evolutionary analysis of the delta and delta plus variants of the SARS-CoV-2 viruses J Autoimmun 2021 124 1 5 10.1016/j.jaut.2021.102715
Karikalan M Chander V Mahajan S Deol P Agrawal RK Nandi S Rai SK Natural infection of delta mutant of SARS-CoV-2 in Asiatic lions of India Transbound Emerg Dis 2021 10.1111/tbed.14290
Leist SR Schäfer A Martinez DR Cell and animal models of SARS-CoV-2 pathogenesis and immunity DMM Dis Models Mech 2020 10.1242/dmm.046581
Li W Shi Z Meng Yu Ren W Smith C Epstein JH Wang H Bats are natural reservoirs of SARS-like coronaviruses Science 2005 310 5748 676 679 10.1126/science.1118391 16195424
Lim Y Ng Y Tam J Liu D Human coronaviruses: a review of virus-host interactions Diseases 2016 4 4 26 10.3390/diseases4030026 28933406
Luan J Yue Lu Jin X Zhang L Spike protein recognition of mammalian ACE2 predicts the host range and an optimized ACE2 for SARS-CoV-2 infection Biochem Biophys Res Commun 2020 526 1 165 169 10.1016/j.bbrc.2020.03.047 32201080
Mallapaty S Coronavirus can infect cats—dogs, not so much Nature 2020 10.1038/d41586-020-00984-8
Maxmen A Wuhan market was epicentre of pandemic’s Nature 2022 603 15 16 10.1038/d41586-022-00584-8 35228730
Meurens F Summerfield A Nauwynck H Saif L Gerdts V The pig: a model for human infectious diseases Trends Microbiol 2012 20 1 50 57 10.1016/j.tim.2011.11.002 22153753
Miao J Chard LS Wang Z Wang Y Syrian hamster as an animal model for the study on infectious diseases Front Immunol 2019 10 1 12 10.3389/fimmu.2019.02329 30723466
Michelitsch A Wernike K Ulrich L Mettenleiter TC Beer M SARS-CoV-2 in animals: from potential hosts to animal models Adv Virus Res 2021 110 59 102 10.1016/bs.aivir.2021.03.004 34353482
Mishra A Kumar N Bhatia S Aasdev A Kanniappan S Sekhar AT Gopinadhan A Sars-Cov-2 delta variant among Asiatic Lions, India Emerg Infect Dis 2021 27 10 2723 2725 10.3201/eid2710.211500 34545805
Munnink BB Oude RS Sikkema DF Nieuwenhuijse RJ Molenaar EM Molenkamp R Van Der Spek A Transmission of SARS-CoV-2 on mink farms between humans and mink and back to humans Science 2021 371 6525 172 177 10.1126/science.abe5901 33172935
Munster VJ Feldmann F Williamson BN van Doremalen N Pérez-Pérez L Schulz J Meade-White K Respiratory disease in rhesus macaques inoculated with SARS-CoV-2 Nature 2020 585 7824 268 272 10.1038/s41586-020-2324-7 32396922
Nova N Cross-species transmission of coronaviruses in humans and domestic mammals, what are the ecological mechanisms driving transmission, spillover, and disease emergence? Front Public Health 2021 9 1 11 10.3389/fpubh.2021.717941
Olwenyi OA, Ravi Dyavar S, Acharya A, Podany AT, Fletcher CV, Ng CL, Patrick Reid S, Byrareddy SN (2020) Immuno-epidemiology and pathophysiology of coronavirus disease 2019 (COVID-19). J Mol Med 98(10):1369–1383
Opriessnig T Huang YW Coronavirus disease 2019 (COVID-19) outbreak: could pigs be vectors for human infections? Xenotransplantation 2020 27 2 1 3 10.1111/xen.12591
Oreshkova N SARS-CoV-2 infection in farmed minks Euro Surveill 2020 25 23 1 7 10.2807/1560-7917.ES.2020.25.23.2001005
Pandey K Acharya A Mohan M Ng CL Reid SP Byrareddy SN Animal models for SARS-CoV-2 research: a comprehensive literature review Transbound Emerg Dis 2021 68 4 1868 1885 10.1111/tbed.13907 33128861
PETA U. Coronavirus, Swine Flu, SARS: Major Diseases Linked to Eating Animals. n.d. “Swine Flu, SARS : Major Diseases Linked to Eatin Animals
Perlman S, Fehr AR (2018) Coronaviruses: an overview of their replication and pathogenesis. Coronaviruses: Methods Protoc, Methods Mol Biol, Springer vol.1282, 10.1007/978-1-4939-2438-7_1
Rabalski L, Kosinski M, Mazur-panasiuk N, Szewczyk B, Bienkowska-szewczyk K, Kant R, Sironen T, Pyrc K, Grzybek M (2022) Zoonotic spill-over of SARS-CoV-2 : mink-adapted virus in humans. Clinical Microbiology an Infectio 1:28(3):451
Rabalski L Kosinski M Smura T Aaltonen K Kant R Sironen T Szewczyk B Grzybek M Severe acute respiratory syndrome coronavirus 2 in farmed mink (Neovison Vison), Poland Emerg Infect Dis 2021 27 9 2333 2339 10.3201/eid2709.210286 34423763
Richard M Kok A de Meulder D Bestebroer TM Lamers MM Okba NMA Fentener van Vlissingen M SARS-CoV-2 is transmitted via contact and via the air between ferrets Nat Commun 2020 11 1 1 6 10.1038/s41467-020-17367-2 31911652
Rockx B, Kuiken T, Herfst S, Bestebroer T, Lamers MM, Oude Munnink BB, De Meulder D, et al (2020) Comparative pathogenesis of COVID-19, MERS, and SARS in a nonhuman primate model. Science 368(6494):1012–15. 10.1126/science.abb7314
Rothan HA, Acharya A, Reid SP, Kumar M, Byrareddy SN (2020) Molecular aspects of COVID-19 differential pathogenesis. Pathogens 9(7):538. 10.3390/pathogens9070538
Rudd JM Selvan MT Cowan S Kao YF Midkiff CC Narayanan S Ramachandran A Ritchey JW Miller CA Clinical and histopathologic features of a feline SARS-Cov-2 infection model are analogous to acute Covid-19 in humans Viruses 2021 13 8 1 16 10.3390/v13081550
Sailleau C Dumarest M Vanhomwegen J Delaplace M Caro V Kwasiborski A Hourdel V First detection and genome sequencing of SARS-CoV-2 in an infected cat in France Transbound Emerg Dis 2020 67 6 2324 2328 10.1111/tbed.13659 32500944
Saravanan KM Zhang H Senthil R Vijayakumar KK Sounderrajan V Wei Y Shakila H Structural basis for the inhibition of SARS-CoV2 main protease by indian medicinal plant-derived antiviral compounds J Biomol Struct Dyn 2022 40 5 1970 1978 10.1080/07391102.2020.1834457 33073712
Severe Acute Respiratory Syndrome (SARS) Epidemiology Working Group (2003) SARS Report WHO. https://apps.who.int/iris/handle/10665/70863
Sharun K Tiwari R Patel SK Karthik K Yatoo MI Malik YS Singh KP Coronavirus disease 2019 (COVID-19) in domestic animals and wildlife: advances and prospects in the development of animal models for vaccine and therapeutic research Hum Vaccin Immunother 2020 16 12 3043 3054 10.1080/21645515.2020.1807802 32915100
Shi J Wen Z Zhong G Yang H Wang C Huang B Liu R Susceptibility of ferrets, cats, dogs, and other domesticated animals to SARS-coronavirus 2 Science 2020 368 6494 1016 1020 10.1126/science.abb7015 32269068
Shi R Shan C Duan X Chen Z Liu P Song J Song T A human neutralizing antibody targets the receptor-binding site of SARS-CoV-2 Nature 2020 584 7819 120 124 10.1038/s41586-020-2381-y 32454512
Shou S Liu M Yang Y Kang N Song Y Tan D Liu N Wang F Liu J Xie Y Animal models for COVID-19: hamsters, mouse, ferret, mink, tree shrew, and non-human primates Front Microbiol 2021 10.3389/fmicb.2021.626553
Sila T (2022) Suspected cat-to-human transmission of SARS-CoV-2. Thailand, July–September 2021. Emerg Infect Dis 28(7):1485–1488
Song Z, Bao L, Yu P, Qi F, Gong S, Wang J, Zhao B et al (2021) SARS-CoV-2 Causes a systemically multiple organs damages and dissemination in hamsters. Front Microbiol 11:1–15. 10.3389/fmicb.2020.618891
Suarez DL, Pantin-jackwood MJ, Swayne DE, Lee SA, Deblois SM (2020) In Poultry 26(12):3074–3076
Tegally H Wilkinson E Giovanetti M Iranzadeh A Fonseca V Giandhari J Doolabh D Detection of a SARS-CoV-2 variant of concern in South Africa Nature 2021 592 7854 438 443 10.1038/s41586-021-03402-9 33690265
Tétrault-Farber G, Vasilyeva M (2021) Russia registers world’s first COVID-19 vaccine for animals. Reuters [(accessed on 15 May 2021)]. 2021. “Russia Registers ’ World’s First’ COVID Vaccine for Animals,” no. March: 2021–22
Vergara-Alert J, Rodon J, Carrillo J, Te N, Izquierdo-Useros N, Luisa Rodríguez de la Concepción M, Ávila-Nieto C et al (2021) Pigs are not susceptible to SARS-CoV-2 infection but are a model for viral immunogenicity studies. Transbound Emerg Dis 68 (4): 1721–25. 10.1111/tbed.13861
Wan Y Shang J Graham R Baric RS Li F Receptor recognition by the novel coronavirus from wuhan: an analysis based on decade-long structural studies of SARS coronavirus J Virol 2020 10.1128/jvi.00127-20
WHO-FAO-WOAH (2021) SARS-CoV-2 in animals used in fur farming: GLEWS + Risk Assessment,” no. January.
Woolsey C Borisevich V Prasad AN Agans KN Deer DJ Dobias NS Heymann JC Establishment of an African green monkey model for COVID-19 and protection against re-infection Nat Immunol 2021 22 1 86 98 10.1038/s41590-020-00835-8 33235385
Worobey M, Levy JI, Malpica Serrano LM, Crits-christoph A, Pekar JE, Goldstein SA, Rasmussen AL, Kraemer MUG, Newman C, Koopmans MPG (2022) The Huanan market was the epicenter of SARS-CoV-2 emergence. Zenodo
Yeh KB Tabynov K Parekh FK Mombo I Parker K Tabynov K Bradrick SS Significance of high-containment biological laboratories performing work during the COVID-19 pandemic: biosafety level-3 and -4 labs Front Bioeng Biotechnol 2021 9 1 9 10.3389/fbioe.2021.720315
Zhang J Cai Y Lavine CL Jianming Lu Xiao T Chen B Zhang J Article structural and functional impact by SARS-CoV-2 omicron spike mutations Ll Ll structural and functional impact by SARS-CoV-2 omicron spike mutations Cell Rep 2022 39 4 110729 10.1016/j.celrep.2022.110729 35452593
Zhao J Li K Wohlford-Lenane C Agnihothram SS Fett C Zhao J Gale MJ Rapid generation of a mouse model for middle east respiratory syndrome Proc Natl Acad Sci USA 2014 111 13 4970 4975 10.1073/pnas.1323279111 24599590
Zhao L, Liu P, Zhao L, Ye F, Wang H, Zhou W, Zhu N et al (2020a) The pathogenicity of SARS-CoV-2 in HACE2 transgenic mice. 10.1038/s41586-020-2312-y
Zhao Y Wang J Kuang D Jingwen Xu Yang M Ma C Zhao S Susceptibility of tree shrew to SARS-CoV-2 infection Sci Rep 2020 10 1 1 9 10.1038/s41598-020-72563-w 31913322
Zhao S Fan J Liu E Animal models for COVID-19 therapeutic development: where we are and where we need to go Front Microbiol 2022 13 2 5 10.3389/fmicb.2022.907406
Zhou P Yang XL Wang XG Ben Hu Zhang L Zhang W Si HR A pneumonia outbreak associated with a new coronavirus of probable bat origin Nature 2020 579 7798 270 273 10.1038/s41586-020-2012-7 32015507
| 36514483 | PMC9741861 | NO-CC CODE | 2022-12-14 23:29:59 | no | 3 Biotech. 2023 Jan 11; 13(1):5 | utf-8 | 3 Biotech | 2,022 | 10.1007/s13205-022-03416-8 | oa_other |
==== Front
Asia Pacific Educ. Rev.
Asia Pacific Education Review
1598-1037
1876-407X
Springer Netherlands Dordrecht
9813
10.1007/s12564-022-09813-1
Article
Practice of leadership competencies by a principal: case study of a public experimental school in Taiwan
http://orcid.org/0000-0001-6930-8319
Chen Chien-Chih [email protected]
grid.445072.0 0000 0001 0049 2445 Department of Educational Management, National Taipei University of Education, Taipei City, Taiwan
11 12 2022
112
23 4 2022
28 11 2022
29 11 2022
© Education Research Institute, Seoul National 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.
Experimental education in Taiwan developed rapidly since the promulgation of the Three Acts Governing Experimental Education in 2014, after which public experimental schools were established in response to local educational needs. The majority of restructuring cases have involved small rural schools faced with high drop-out rates and staff retrenchment plus merger (SRM). Indigenous schools also initiated experimental education by introducing teaching modules on cultural responses to develop ethnic education that better reflects local culture. In metropolitan areas, experimental schools were established through restructuring for educational innovation or for coping with the pressure of competition from neighboring schools. The number of experimental schools and enrolled students has increased yearly, which reveals the expectations of parents about diversified education and their right to make educational choices. The school selected for this study is an elementary school in a metropolitan area that experienced an SRM crisis and was restructured as a public experimental school in 2019. The study examined the methods through which the principal transformed leadership competencies to leadership strategies. The school overcame the competition and SRM crises and even achieved full-capacity enrollment in Taipei city. Information was collected from the relevant literature and through interviews with the principal, a head of department, teachers, a parent, and an administrative official of the Bureau of Education. The study analyzed the leadership competencies and corresponding strategies of the principal and the future prospects of the school. The findings provided principals of experimental schools with a basis for sustained improvement in leadership competencies and recommendations for the future development of school affairs.
Keywords
Alternative education
Experimental education
Leadership competencies of principals
The publicly owned and publicly run school
http://dx.doi.org/10.13039/100007225 Ministry of Science and Technology 110-2410-H-152 -030 -SSS Chen Chien-Chih
==== Body
pmcThe policy of compulsory education, which was implemented in Taiwan prior to the 1990s, endows the state with the right to selection, which includes the introduction of a standardized academic system, compulsory enrollment for all students, and discouragement of private education (Chin, 2014). However, liberalization and democratization were introduced as part of educational reform with the lifting of the martial law in 1987. Thus, education was no longer a mere political tool of the government, and diverse opportunities became available for parents in selecting a suitable educational environment and method for their children. This initiative led to the development of experimental education in Taiwan, which non-governmental forces that promoted experimental schools outside the mainstream educational system introduced in 1989. In recent years, parents made educational choices in Taiwan and became more attentive to school-based education. Moreover, the Ministry of Education (MOE) promulgated the Three Acts Governing Experimental Education (hereinafter referred to as the Three Acts) in 2014, which led to an annual increase in the number of experimental schools (MOE, 2014a; Taiwan Experimental Education Center, 2021).
Experimental education is synonymous with the internationally accepted alternative education. Alternative schools are those outside the mainstream educational system in the United States and Europe. In terms of characteristics, the major commonality between experimental and alternative education is that they are non-traditional and differ from public schools (Conley, 2010). After the adoption of the Three Acts in 2014, alternative education became nearly equivalent to experimental education in terms of legal basis; thus, it deviated from its previous connotation of non-mainstream education and became a part of the educational landscape in Taiwan (Lee & Cheng, 2019).
Experimental education in Taiwan is expanding in terms of the number of schools (from 11 [eight publicly owned and publicly run and three publicly owned and privately run] in 2015 to 114 [91 publicly owned and publicly run; 15 publicly owned and privately run; and 8 privately owned and privately run] in 2021). The total number of students is also increasing annually. Analysis of the literature reveals that several types of public schools in Taiwan offer experimental education. One type comprises small rural schools that face the crises of high drop-out rates and staff retrenchment plus merger (SRM). Specifically, in elementary and secondary schools in Taiwan, if a school has one class without students, then teacher vacancies will decrease; consequently, these teachers will be transferred from their current schools to that one school. In addition, if the total number of students in a school is less than 50, then the local education authority will consider merging the school with neighboring ones or encourage the school to implement mixed-age class teaching to facilitate the appropriate allocation and application of educational resources (MOE, 2017).
Another type includes indigenous schools, which introduced teaching modules on cultural responses to develop ethnic education that better reflects the local culture. The third type is found in metropolitan areas, where the establishment of experimental schools that involve the restructuring of the educational set-up was based on educational innovation to cope with the pressure of competition from neighboring schools (Hsieh et al., 2021; Taiwan Experimental Education Center, 2021). In the past, experimental education in Taiwan mainly aimed to solve the educational problems faced by the disadvantaged. At the time, experimental schools were predominantly located in rural areas (Chin & Chuang, 2019; Hsieh et al., 2021; Lin & Chen, 2020).
To a large extent, experimental schools operate very differently from those in the past. To ensure the complete implementation of experimental education, restructuring the system, administrative operation, type of organization, curriculum, pedagogy, and community participation is first necessary by modifying outdated practices. The process of transformation will be met with challenges, and principals will need to lead their teams to breakthroughs (Chin & Chuang, 2019). Thus, the leadership of the school principal plays a critical role in this process (Tseng et al., 2021; Wang et al., 2016), because they are important helmspersons in school management and considerably impact its performance. Leadership is the ability of an individual to motivate, influence, and enable other people to contribute to the success of an organization. A principal communicates to set directions, inspire, and motivate teams (Grace, 2005). Thus, the principals of experimental schools should carefully evaluate the external environment, effectively utilize internal and external talents and resources to enhance administrative efficiency, and emphasize the flexibility in the organization.
A review of the literature revealed that the aspects of experimental education that studies currently highlight include curriculum development, the professional training of teachers, the developmental process of school restructuring, and policy evaluation. In contrast, relatively few studies focus on the leadership of principals in experimental schools (Chen, 2020; Hsieh et al., 2021). Research on principals emphasizes their leadership strategies, actions, and characteristics; however, those that discuss their leadership competencies are lacking. Competency refers to the behavioral characteristics of individuals who exhibit high levels of performance in their profession and encompasses personal thoughts, values and judgments, behavioral representations, personal beliefs, and internal and external images. In other words, a difference exists in competency between individuals with high and average levels of performance (Boyatzis, 2008; McClelland, 1973). Leadership competencies are reflected in follow-up strategies adopted by individuals and require continuous reflection and evaluation (Lamb, 2014; Saruchera, 2022). In the school, the leadership competencies of principals will influence their strategies and implementation actions.
At present, studies on the leadership of principals of experimental schools in Taiwan are insufficient (Chang, 2021; Chen, 2020; Lin & Chen, 2020; Tseng et al., 2021). The tacit knowledge of principals about school management can only be ascertained through the evaluation of experimental education. If the explicit behaviors and intrinsic characteristics of principals are properly recorded, then the results can be effectively used for the promotion of their leadership. Hence, this study aims to examine the leadership strategies, actions, and corresponding competencies of principals, which, thereby, portray a comprehensive image of their leadership.
Given that operational characteristics and the conduct of experiments demonstrate the significance of experimental schools (Leithwood et al., 2020; Liang & Fan, 2020; Liu, 2018), scholarly attention focused on the effective management of these schools by principals. This ideal can be achieved through the development of leadership competencies and by mobilizing parents, external resources, teaching teams, and the staff of the Bureau of Education (BOE) prior to the final execution of an educational plan based on its goals and established guidelines.
Publicly owned and -run experimental schools (POESs) constitute the largest proportion of experimental schools in Taiwan. The school selected for this study was an elementary school in a metropolitan area that faced an SRM crisis prior to being restructured as a POES in 2019. The emphasis was on the process through which the principal led the school team in facing the competitive pressure from neighboring schools and addressing the SRM crisis through the promotion of leadership competencies, which were then transformed into corresponding leadership strategies that led the school to full-capacity enrollment. The selected school was formally evaluated regarding its provisions for experimental education and received a Teaching Excellence award in 2022. In addition, the Taipei city governance certified it as a high-quality school in the innovative and experimental category in 2020 and recommended that the school should participate in the competition for the 2021 National Excellence in Teaching Award. Initially, the school had less than 10 students, but its student intake subsequently increased, and full-capacity enrollment was attained. Although this success may be attributed to other factors, such as the emphasis of parents on experimental education and effective teaching by teachers, the effectiveness of the leadership and management of the principal is one of the major factors of success that requires exploration. Therefore, this study aimed to analyze the development of the leadership competencies and strategies adopted by the principal to address challenges during the development of the POES.
The research questions are as follows:i. What were the leadership competencies of the school principal and how were they implemented?
ii. What were the challenges faced by the school principal while managing the POES? Which strategies were adopted to resolve these challenges?
Connotations of the leadership competencies of principals and related research
Spencer and Spencer (1993) proposed that competency1 refers to the basic underlying characteristics of individuals that are not only related to their duties at work but can also be used to understand their actual or expected responses. In turn, these characteristics influence their behavior and performance, which enables employees to effectively complete their tasks and achieve the desired outputs. In terms of connotations, Cofsky (1993) argued that certain characteristics possessed by an individual (e.g., knowledge and skills) do not necessarily guarantee high levels of work performance. In fact, superior performance is dependent on underlying traits such as aptitude, attribute, motivation, concept of self, and values. These traits can be used to predict high-performance outputs when properly combined within the appropriate context. In terms of application, Parry (1998) argued that competency influences the performance of an individual at work. It is measured by a set of acceptable standards and is enhanced through training and development. Additionally, Grzesik and Piwowar-Sulej (2018) suggested that future research should examine competencies and leadership styles, which occur across types of project-oriented organizations.
Several studies focused on the competencies of school principals. For example, Choi et al. (2022) suggested that principals should implement foundational and task-focused competencies to enhance the learning of students. Foundational competency includes cognitive, influencing, managing, and personal characteristics. In contrast, task-focused competency covers teacher, curriculum, school climate, and environmental elements. Hellriegel et al. (2005) defined competency as the combination of knowledge, behaviors, skills, and attitudes that help improve personal work performance. Moreover, competency refers to excellent technical abilities, the reliable execution of processes, and positive relationships with various stakeholders. The authors further proposed that managers should respond to the management style of the new era and achieve a balanced development of six competencies, namely, strategic operation, management planning and execution, communication, team operation, self-management, and global cognition. After a further examination of the competencies to be developed, Chen and Lin (2019) and Hsiao et al. (2010) put forward an additional seventh competency, namely, innovative integration and marketing, which may include the elements proposed by Choi et al. (2022).
Information compiled from the relevant literature and studies indicates that the leadership of principals influences the effectiveness of schools, pedagogy of teachers, learning of students, and community relations. The current discussions on the leadership and business strategies of principals are diverse and include vision planning, resource integration, environmental assessment, leadership in curriculum and teaching, professional development of teachers, and division of labor within teams. However, a system for structural analysis is lacking. When individuals exhibit high levels of job performance and related behavioral characteristics, their competencies include personal thoughts, values, moral judgment, behavioral representations, personal beliefs, and external images. These concepts can serve as the basis for an integrated analysis and for the management of the strategies of schools as an extension of leadership competencies, which are clear criteria for evaluating school leadership (Boyatzis, 2008; Chen & Lin, 2019; Hsiao et al., 2010; Lamb, 2014; Spencer & Spencer, 1993).
In light of this discussion, the seven leadership competencies and their elements can be described as follows:i. Strategic operation This includes three elements, namely, general education industry, internal organization of the school, and the advancement of school policies.
ii. Management planning and execution This includes information management, problem solving, and planning organizational proposals.
iii. Communication This comprises formal communication, informal communication, and negotiation.
iv. Team operation This includes operating similar to a professional team, creating a supportive environment, and maintaining the team’s momentum.
v. Self-management This includes sincerity and moral leadership, autonomy and adaptability, work–family life balance, and self-awareness and improvement.
vi. Global cognition Its two elements are perception and understanding of the local culture and broad cultural outlook and sensitivity.
vii. Innovative integration and marketing The three elements are innovative management of school affairs, cross-domain integrated management, and marketing of the school.
Methodology
The study adopted the interview (semi-structured) and textual analysis methods to elucidate the promotion and application of leadership competencies by the school principal. First, the researcher shortlisted schools in six metropolitan areas in Taiwan2 and those that underwent restructuring as POESs in recent years. Schools that may have faced SRM initially but later achieved full-capacity enrollment after restructuring were selected for the case study. The final school was selected in consultation with administrative officials of the BOE. Internal interviews were then conducted with the school principal, the head of the department who promoted the restructuring of the POES, two teachers, and a parent; an external interview was conducted with the administrative official of the BOE.
The selected school was established in 1956. The total number of students decreased to 70 in 2016 with less than 10 students enrolled in grade 1, which indicated an SRM crisis. The current principal was transferred to the school of the same year. After three years of preparation (2016–2018), the school was restructured as a POES in 2019. The curriculum was divided into four semesters, and the core concepts experiential learning and habits demonstrating high efficiency levels were introduced. Immediately after the school was restructured, it achieved full-capacity enrollment. The curriculum included learning of Chinese, English, and Mathematics as well as four sets of learning periods with the contents of the curriculum grouped into the following themes: speed of wind, sentiments of farming, rippling of water, and passing of clouds. The MOE also stipulated a curriculum3 (MOE, 2014b). Apart from those assigned for learning Chinese, English, and Mathematics, curriculum hours were devoted to experimental education.
The study conducted semi-structured interviews to collect information on the development of the leadership competencies of the principal. The interview codes for the school principal, the head of the department, two teachers, one parent, and one BOE administrative official are P, D, T1, T2, F, and E, respectively. In the following abbreviations, which were enclosed in parentheses, of the interview transcript, I denotes the interview method, and P, D, T1, T2, F, and E pertain to the codes of the interviewees. The interviews were conducted from March 3 to May 6, 2021. Each interview was conducted based on an interview outline and lasted for approximately 2 h.
Additionally, the study employed relevant documents, such as the experimental education plan of the school, and submission documents pertaining to participation in the awards for high-quality schools and teaching excellence, as supporting materials. The researcher further conducted thematic coding analysis to fulfill the requirements stipulated by Denzin (1994), Yazan (2015), and Yin (2009), that is, the qualitative research text becomes a meaningful discourse only after the researcher has interpreted and summarized it. In the concept review stage, the researcher coded similar data into the same concept and checked all concepts repeatedly until no new one could be iterated. In the category development stage, similar concepts were then integrated into a category. This type of analysis was based on an iteration process (Cohen et al., 2011), where the content of the qualitative research underwent an interpretation process to express the meaning (Denzin, 1994). Finally, all data were integrated and compared with the literature to describe and identify the concepts, categories, and themes relevant to the objective of the research. Additionally, the relevant documents were used to confirm the certainty of the interview contents. For example, when the principal described the curriculum contents as being grouped into four themes, namely, speed of wind, sentiments of farming, rippling of water, and passing of clouds, we checked for evidence of the implementation of such content in the education plan of the school by examining the relevant documents in an effort to address the abovementioned research questions.
Prior to the interviews, the researcher briefed the participants about the details of the plan and the interview site and ensured that they read the informed consent form and the outline of the interview prior to providing informed consent. The objective of the research was fully explained, and the participants were assured that their privacy would be protected and that the results would be accurately presented. Hence, the study ensured compliance with research ethics (Cohen et al., 2011). The study adopted triangulation for the compilation and coding of the collected data (Lincoln & Guba, 1985, p. 283). Specifically, the triangulation of the participants (including the school principal, head of department, two teachers, the parent, and the BOE administrative official) served as the basis for the study.
Research findings
Comparing the results of the interviews with the school documents and literature review, the study found that the selected principal demonstrated the seven leadership competencies. The study presents these leadership competencies and their resulting strategies to answer research question 1 and provides excerpts from each interviewee.
“Experiential learning” and “habits demonstrating high efficiency levels” as the basis for strategic operations and leadership planning
Competency in strategic operations includes two elements, namely, understanding the school’s internal organization and the general industry and advancing school policies. In the process of promoting experimental education, the principal of the selected school adopted experiential learning and habits demonstrating high efficiency levels as the planning directions of his strategic leadership. These elements form the main pillar of the experimental education plan of the school and, at the same time, serve as a crucial support for transacting the school curriculum and for administrative operations. Simultaneously, the school paid attention to experiential learning while developing the aforementioned curriculum on the basis of the four themes.4 Moreover, the teachers set goals given the learning outcomes of the students upon completion of the curriculum. The learning process primarily involves experiential learning activities, as diversified evaluation is conducted upon curriculum completion. Evidently, the school adopted experiential learning as its core axis, which permeates various aspects of the school that range from curriculum design to student learning and diversified evaluation.Consensus is very important for restructured experimental schools. First of all, the principals must reach a consensus with their school colleagues and evaluate whether the teaching team is willing to work to attain the goals. The main goal of the school is experiential learning and to allow the children develop through experiential curriculum. The behavior and mode of functioning of the teachers and the students are cultivated to be highly efficient and habitual through the seven habits course. (I_E)
Our experimental education places strong emphasis on experiential learning. It is believed that children can learn through their senses and through actual experiences. On this basis, our experimental education curriculum is developed. (I_D)
Habits demonstrating high efficiency levels are based on the seven habits listed by the Paradigm International Educational Organization. The program combines teaching by words and example and teaching based on context along with the teaching system of the school. The curriculum stipulated by the school is implemented at all grades for three periods per week to cultivate habits demonstrating high efficiency levels in teachers and students, which have also become the common vocabulary among parents, teachers, and students. Thus, a principal becomes a person with a high level of self-reflection through the cultivation of habits demonstrating high efficiency levels.It was found that the teachers’ effectiveness improved significantly after the seven habits were introduced when the school was restructured. Ever since I was transferred here a few years ago, we hardly have to work overtime for the school’s restructuring for experimental education, or when we participated in competitive programs such as “high-quality schools” and “activating teaching.” The reason why a culture of working overtime does not exist here is due to our time management habits and team performance. (I_P)
The main goal of the school’s experimental education is experiential learning and development of habits demonstrating high efficiency levels. To achieve this purpose, the school system was restructured to include four semesters, which makes the effectiveness of the main goal more apparent. The children’s learning can be more focused after the restructuring of the curriculum, because it is expected that they may have been adapted to experiential learning during the four semesters. After the ten-week course, there is a short vacation which allows them to practice the lessons learned in class and share these with their family members. In the process, they become more immersed in what they have learned previously. (I_T1)
“Consensus at meetings” and “full participation of the staff” as the basis for planning and executing management strategies
Competency in management planning and implementation involves information management, problem solving, and planning of organizational proposals. The principal could effectively implement his plans and maintain progress, obtain the resources required to complete tasks, and exercise authority in designating and prioritizing the order of tasks. Planning and implementation of management strategies was based on consensus established at meetings and with the full participation of the staff. In addition, the principal maintains a record of the internal and external resources of the school in the past, evaluates the capacities of teachers, and promotes administrative leadership and the transaction of the curriculum. Therefore, the theme-based lesson plans of the school can be implemented even before the restructuring. With the joint participation of teachers, the curriculum that was initially implemented became the predecessor of subsequent lesson plans for experimental education, which, thereby, reduced resistance due to the restructuring.After being appointed as the principal, I spent my first year dialoguing with the community members, parents, and teachers to create a good organizational atmosphere. Sufficient discussions were held when the school administrative meeting was held a year later, and a consensus about school’ goals was reached quickly. Of course, we have included new content in the curriculum before drafting the experimental plan to restructure the school as an experimental school. The curriculum that was followed at that time was the original one that was eventually transformed into the school’s current four sets of theme-based lesson plans. (I_P)
All the teachers of the school jointly participated in the restructuring. Based on the themes of wind, water, clouds, and farming, we were similarly divided into the four groups. Members of each group had to cooperate to prepare the curriculum contents of the assigned theme. While formulating groups, the attributes and expertise of each group member were considered. After the groups had progressed to a certain level with respect to curriculum development, we made presentations at a common school meeting. (I_T2)
“Team awareness” and “providing care for growth” as principles for communication
Competent communication includes three elements, namely, informal communication, formal communication, and negotiation. The principal adopted the principles of team awareness and providing care for growth. His informal communication channel focused on two-way communication, that is, empathizing with the feelings of others and establishing interpersonal relationships. Formal communication was held through meetings, which enabled the faculty and staff to understand the current situation, openly discuss and address problems, and formulate strategies for restructuring the school via discussions and by arriving at a consensus. Indeed, restructuring the school to implement experimental education enabled it to attain full-capacity enrollment while improving the learning effectiveness of the students, cultivating a culture of organizational cooperation, and boosting the morale of teachers to a large extent.Considering that harmony and unity among its people are critical for an experimental school, I spent a year getting to know all my colleagues. I managed to understand their basic details, additional expertise, living patterns, family status, and personal values through multiple informal conversations. (I_P)
I realized that cohesion among the teachers has strengthened after the restructuring. All of us are willing to spend time teaching, preparing, and discussing together, leading to a deeper understanding of the school curriculum. The practical activities included in the curriculum contents allow the students to achieve mastery through a comprehensive study of the subjects, after which they gained better abilities at integrating information. (I_T2)
“Team momentum” and “encouraging multiple expertise” for team operation
Competency in team operation includes three elements, namely, creating a supportive environment, operating similar to a professional team, and maintaining team momentum. The principal used team momentum and encouraging multiple expertise to create a supportive team environment, which is characterized by efficiency that fosters identity and shares praises and rewards. In the staff, he also nurtured a sense of identity with the school, which facilitated their growth. He frequently assisted teachers in locating resources and was happy to recommend their participation in competitions for special and excellent teacher awards organized by the MOE and BOE, which, thereby, boosted team morale and reinforced their identity with the school. Given the emphasis of the school on experiential learning and its initiative for the development of the four sets of theme-based lesson plans, the principal expects the teaching team to possess diverse expertise, which is reflected in his leadership strategies.We have participated in several competitive programs that examine the ways schools are run using objective external indicators, such as the city’s program for high-quality schools. In addition to team performance, I also assist teachers in applying for recognition as special and excellent teachers. Even though I am unable to increase their salary, their inner drive will be strengthened when they receive such honors. In turn, they will be more enthusiastic to contribute toward the development of the school curriculum. (I_P)
The capacities, including multiple areas of expertise, are highly valued when implementing the school’s curriculum. Because our curriculum is closely related to many life skills, I have to tap the teachers’ talent. In my opinion, the teachers definitely do not doubt their own professionalism. However, I have found that enthusiasm is the key. In other words, the teachers must be willing to initiate change in the experimental school. (I_P)
“Team appreciation” and “responsibility-bearing culture” as principles for self-management
The four elements of competency that pertain to self-management include sincerity and moral leadership, autonomy and adaptability, work–family life balance, and self-awareness and improvement. The principal focused on team appreciation and responsibility-bearing culture as the criteria for self-management. For this reason, the principal hoped that his enthusiasm for education can influence the school team and encourage them to complete their tasks well and wholeheartedly and to focus on the administrative and teaching processes.The principal is a person of great vision. He is very caring and thoughtful, and is also very respectful to the children, parents, and teachers. (I_F)
The principal emphasizes a responsibility-bearing culture. Although he does not impose very strict restrictions on us, he hopes that we take initiative to accept and complete our own tasks. He also trusts the professionalism of our administrative colleagues and provides us with the authority to decide how we should carry out our tasks. (I_D)
However, the great importance attached by the principal to the harmony of the team and the enthusiasm of the teachers led to the frequent necessity for him to wait patiently for the teachers to change. This scenario led the parents to hold dissenting opinions.The principal is very respectful of the teachers of his school and is unwilling to force change upon them. He is always of the opinion that when he takes the lead in making the change, the teachers will follow suit after observing him for a while. However, from the parents’ perspective, our child will only be with the school for six short years. I want the principal to be more proactive in the process of urging the teachers to change. However, he keeps hoping that we will give those teachers and him more time and promises to do his best in leading the teachers to change. (I_F)
“Deep rootedness” and “themed-based lesson plans” for global cognition
Competency in global cognition comprises two elements, namely, perception and understanding of local culture and broad cultural outlook and sensitivity. The school uses deep rootedness5 and themed-based lesson plans to help the teaching staff realize the impact of globalization and related policies on the school and to discover other means of strategizing to meet the goal of globalization through the localized theme-based curriculum. As the school is located in a metropolitan area, it has many international students. During class discussions, international students share experiences from their countries and even lead group members in sharing-and-exchange sessions with teachers and students from other countries via the Internet. Actual international exchange and learning activities are also conducted. The subject school was supposed to go to Australia for a subsidized exchange program in 2021, but it was disrupted due to the worsening COVID-19 situation. Nevertheless, exchanges and interactions with foreign teachers and students are conducted via video conferencing due to the adaptability of the school to the pandemic situation. This mode of interaction and exchange may become a major one in the future.In recent years, bilingual lesson plans for the smaller curriculum modules have been developed, such as pizza making, beekeeping, and cycling. This gives the students opportunities to learn and communicate in international languages to connect with the rest of the world. Bilingual presentations to introduce these lesson plans are also made during visits by foreign guests. This is in line with Taiwan’s national policy Bilingual 2030. (I_P)
This school has many international students with different nationalities. One of its students from Lithuania led his group during an interaction with the teachers and students of a Lithuanian school. Such globalized interactive teaching can successfully be conducted once the Internet and related soft- and hard-ware environments are properly set up. (I_T2)
“Active cooperation” and “word-of-mouth marketing” as the main axis for innovative integration and marketing
Competency in innovative integration and marketing includes the marketing of the school, cross-domain integrated management, and innovative management of the affairs of the school. The principal focuses on active cooperation and word-of-mouth marketing as the main axis. He is diligent in visiting communities and cooperating with the BOE in its activities for policy promotion, such that community members, the BOE, and parents acquired positive impressions of him. The head of department of the school further added that “parents comprise the most substantial marketing channel, that is, through their word-of-mouth.” The parent interviewees express that parents of students in the metropolitan schools of the city have several ideas with regard to school management. If they approve of a school, then they are likely to serve as the main channel of the publicity of the school. The administrative official of the BOE also mentions that the school, after being restructured as a POES, has attracted the attention of parents from neighboring counties and cities, who drop their children off at school as they commute to work daily. One of the driving forces behind this phenomenon is the word-of-mouth marketing performed by the parents.Our school does its best to cooperate with BOE’s policies, such as the adoption of school dogs and cats, so as to develop the school’s character. We also assist the BOE whenever it holds external expositions. Therefore, when the media requires life stories on adopted animals, the BOE always recommends our school for the interviews. For a community-based school like ours, interactions with the community are vital. (I_P)
The best form of marketing for a school is running it well. I rarely issue any press releases on my own initiative since the restructuring. Nevertheless, I am overwhelmed by the number of interviews and feature photoshoots being scheduled. So, it is back to what I said at the beginning, “running a school well is the best form of marketing.” The BOE, parents, and the community will help you in its promotion. (I_P)
Predicaments and actions
The important mechanisms that facilitate the successful operation of a POES and the management of experimental education include authorization by government units, innovations in the transaction of the curriculum, and cooperation among teaching teams. Predicaments will occur when any of the aforementioned problems emerge (e.g., non-cooperation with the BOE and disagreements among administrators, teachers, and parents) during a principal’s leadership of the school. One of the predicaments faced by the subject principal is competition and cooperation with the BOE over the existing operations. Given the public status of the school, the principal eventually opted to comply with the policies of the BOE.The principal is a person who fully respects the system. He is running a public elementary school after all, so he will not be the person who confronts and challenges the existing system. This is because he feels his responsibility toward the BOE. (I_F)
I had a meeting with my school colleagues last year. At that time, the BOE wanted to promote long-term care education and hoped that schools at all levels would incorporate it into the curriculum structure. However, not enough opportunities are available to the children to be in contact with their grandparents, keeping in mind their family organization. Being a public elementary school, we eventually had to include the concept of long-term care education into the curriculum structure and submit it to the BOE for vetting and archival for future reference. (I_D)
The majority of the school teachers expressed enthusiasm and willingness to provide assistance in exploring the curriculum and to aim toward the development of the school. However, as an experimental school, a major proportion of the curriculum requires continuous revision and innovation. This notion implies inevitable disagreements among the administrators, teachers, and parents. For example, a few teachers felt that certain portions of the experimental curriculum were unrelated to teaching. Hence, they were less willing to participate in a number of enrichment classes. The parents wished that the teachers of the school would take the initiative to learn; as such, they suggested that the principal should be more forceful in guiding and controlling these teachers. However, under the team appreciation strategy adopted by the principal as a result of self-management competency, he hoped that these teachers will eventually be motivated to learn after seeing other team members and he also take the initiative to keep on learning.There are many curriculum activities that are being conducted in the school, such as cycling, river tracing, and observing organisms in the water. I know that some of the teachers are not in support of organizing activities outside the school premises. However, as a parent, I wonder how the teachers are going to teach our children if they themselves lack the drive to learn. I often suggest to the principal that he should be stricter in exercising control over the teachers during their learning process. However, he prefers to wait and let those teachers engage in learning spontaneously and proactively. For him, that would be the best situation to learn. (I_F)
Another issue that attracted social concern is the selection and allocation of teachers on the basis of the existing screening process. The reason is that the selected candidates do not possess the required competency to engage in experimental education. Regarding the screening of teachers, the principal conducts robust discussions with his team regarding the approach to be adopted and considers whether or not the school should screen teachers who could engage in experimental education or whether or not to entrust the process to the BOE. In the end, the decision was to entrust standardized screening to the BOE. Unfortunately, the BOE does not conduct specially tailored interviews for experimental schools. Consequently, a few selected teachers of the school lack the desired competency. The leadership approach of the principal is to find the means to include these teachers in the team. The alternative is to wait for the expiration of the term of appointment of these teachers prior to being transferred to another school.Some of the teachers allocated to the school through the screening process are suitable for our experimental curriculum. For those who are less suitable, I hope to motivate them to learn proactively so that they can be a part of the school team as soon as possible. (I_P)
Recruitment is one of the biggest issues faced by the school. This is because regular teachers of public schools are allocated through the screening exercise organized by the respective country’s or city’s BOE. However, teachers who pass the screening are not necessarily suitable for this vibrant and active school. The first remark made by some teachers upon their arrival is that they have never been under the sun or participated in outdoor sports. It will take time to get accustomed to the functioning of the school. (I_D)
Another issue is the alignment between the theme-based curriculum formulated by the school and the major learning points of the 2019 curriculum promoted in Taiwan, which is a topic frequently discussed between teachers and parents during their communication. The school curriculum includes concepts pertaining to nature, society, and other fields. Nevertheless, the parents are also concerned that the students will not have learned the concepts that are taught in regular schools when they graduate. Hence, the principal holds a Parents’ Day every semester to communicate the curriculum and operation of the school with the parents. He also expects that the four sets of theme-based lesson plans have horizontal linkages in addition to vertical linkages, which can be perfectly integrated. In the subject school, vertical linkages represent the curriculum connection from grades 1 to 6, and horizontal linkages denote the curriculum connection among the four lesson plans, Chinese, English, and Mathematics in the same grade.The process of vertically developing these four sets of theme-based lesson plans (speed of wind, sentiments of farming, rippling of water, and passing of clouds) is nearly complete, and they are aligned with the learning goals of the 2019 curriculum. We have explained it in detail to the parents. However, I also hope that the horizontal linkages of the curriculum can be completed and integrated. (I_P)
Finally, the school has begun implementing relevant theme-based lesson plans prior to the appointment of the current principal. Two years have passed since the school was transformed into an experimental one with gradual development and improvement in the system and the curriculum. In addition, the teachers have identified suitable teaching methods for the experimental education curriculum and have gained familiarity with regard to their execution. However, the outside world has expected the school team to continuously innovate and improve, which inevitably exerts additional pressure. Despite this situation, the principal has upheld a positive attitude and continued to share details with the team through the competency of communication. Additionally, he has continued to communicate with the team through formal and informal meetings on various issues, such as bilingual education, in the future and bridging the current learning of the students in school and their future learning in high school. This perspective has reduced the resistance of the team members to continuous innovation.Although the plan for experimental education has been completed, it can still be revised. In fact, I think that it should be modified from time to time. The curriculum should generally proceed in a better direction while being refined and improved. Innovate and innovate in the process of getting better and break through the original framework to strive for greater achievements. (I_E)
From restructuring as an experimental school to the present, the teachers feel that they have identified teaching methods which they are familiar with. Hence, it will be slightly difficult if the curriculum or teaching method needs to be modified at this time. Sometimes, teachers generate self-doubt due to the doubts and challenges from other schools’ teachers or parents. (I_T1)
Discussion
In summary, the study addressed the research questions through data analysis and reflection on the narratives. The leadership competencies of the principal include strategic operation, management planning and execution, communication, team operation, self-management, global cognition, and innovative integration and marketing. When managing the POES, the principal also implemented leadership strategies and actions according to the abovementioned leadership competencies. For example, experiential learning and habits demonstrating high efficiency levels were strategies for implementing the strategic operation competency. Other strategies and actions can be identified through the abovementioned findings. Research question 1 can be answered using this point.
In reality, the principal converted his leadership competencies into strategies and overcame challenges while managing the POES due to the competition and cooperation with the BOE over the existing operations. Given the public status of the school, the principal eventually opted to comply with the policies of the BOE. Disagreements also emerged among the administrators, teachers, and parents. The strategy adopted by the principal was to promote other enthusiastic teachers in the hopes of inspiring them. Another problem was the selection and allocation of teachers; a few of them lacked the desired competency, because they were jointly recruited by the BOE. Thus, the principal was required to formulate a plan to integrate these teachers into the team.
The differences in the curriculum of the school and those of other schools in 2019 was also a problem. Hence, the principal organized a Parents’ Day to communicate with the parents and planned the four sets of theme-based lesson plans to include vertical and horizontal linkages. However, the team members continued to face pressure with respect to curriculum innovation due to external concerns pertaining to the school. Principals communicated actively through formal and informal meetings to reduce the pressure on the teachers in relation to curriculum innovation. Through this initiative, the study addressed research question 2 “What were the challenges faced by the school principal while managing the POES? Which strategies were adopted to resolve the challenges?”.
Conclusion and future prospects
The leadership competencies of the school principal included strategic operation, management planning and execution, communication, team operation, self-management, global cognition, and innovative integration and marketing. Chen and Lin (2019) and Hsiao et al. (2010) also cited these leadership competencies. In the current study, these competencies were practiced through the leadership strategies and actions of the principal. The major challenges faced by the school principal in the management of the POES were due to the competition and cooperation with the BOE, curriculum differences with other schools, and pressure on the team members, among others. The principal was required to determine corresponding strategies to resolve the challenges.
The leadership of the principal indicated that strategic operation was the top priority during the restructuring of the school as an experimental school; Chin and Chuang (2019) also proposed this concept. In the school, the principal used experiential learning and habits demonstrating high efficiency levels as promotion strategies in alignment with the spirit of the experimental education plan of the school. The study then combined the management planning and execution competency with the leadership strategies of the principal to conduct a critical inspection of the performance and effectiveness of the school in terms of management. The results can also be viewed from the perspective of the development of the school of the theme-based lesson plans and the efforts exerted by the teachers.
The students engaged in in-depth learning through experiential operations, which led to the transfer of knowledge (Chen, 2020). In turn, this process facilitates the development of wisdom, which influences their learning in other topics. However, a few aspects of experimental education continue to require attention, including the differences between the old curriculum and the revised 2019 curriculum stipulated for regular schools.
However, the school may not have effectively ensured that the community is aware of the main objectives and characteristics of experimental education and if it had not implemented a four-semester system. Moreover, it would have merely used the curriculum stipulated for regular schools for its experimental lesson plans or adopted the standard teaching materials prepared by the MOE for learning pertaining to various fields. These assumptions are a major point on which to ponder for the future restructuring of schools that provide experimental education.
The BOEs of various counties or cities are likely to relax their restrictions on POES whenever possible. However, whether the extent and scope of relaxation matches the development of schools is another issue. Furthermore, as previously discussed, the principal opted to entrust the BOE of the city to screen teachers for various reasons. Consequently, becoming accustomed to the new system was a lengthy process for a few of the selected teachers. This issue can be addressed effectively if experimental schools formally select their teachers.
At present, the main focus of many experimental schools in Taiwan is curriculum experimentation. For this reason, the examination of other cross-domain systems, such as the school administration, is lacking. If experimentation is conducted at various levels, then more distinct differences will become apparent between the old and the new 2019 curriculum. Thus, the future development of experimental education in Taiwan can focus on the school system or operation. This perspective may enrich the types and models of the development of experimental education in the country.
Finally, the study revealed the manner in which a principal from a POES developed and practiced leadership competencies and implemented strategies for solving challenges while managing the school. The findings can inform and contribute to the development of effective methods for examining school principals and their leadership competencies. Additionally, leadership competencies in POESs can be viewed as a theoretical framework, and its indicators can be developed. Thus, future studies may examine a theoretical framework with leadership competencies as the independent and other related dependent variables.
Acknowledgements
This work was supported by National Science and Technology Council in Taiwan (110-2410-H-152 -030 -SSS).
Declarations
Conflict of interest
This work was supported by one organization and has been presented in the title page.
Ethical approval
This work was approved by the Research Ethics Committee of National Taiwan University (202107ES020) and prior written informed consent to participate in the study was obtained from each subject.
1 Spencer and Spencer (1993) chose the term competence in their paper’s title. However, more researchers chose competency and competencies, and indicate that competence, competency, and competencies have the same meaning. Therefore, the author has used competency and competencies in this research.
2 In Taiwan, a population of more than 1.25 million is located in metropolitan areas, called special municipalities. Since 2022, Taipei city, New Taipei city, Taoyuan city, Taichung city, Tainan city, and Kaohsiung city have been included.
3 Taiwan implemented a new curriculum in 2019, which is divided into two parts for elementary schools, and is mandated by the MOE and stipulated by the school. The former is under standardized planning by the state and aims to cultivate the basic academic abilities of students; the latter is independently planned and developed by the schools.
4 The four sets of theme-based curriculum plans of the subject school are speed of wind, sentiments of farming, rippling of water, and passing of clouds. Science, social studies, life curriculum, arts, and humanities are included in these studies.
5 Deep rootedness denotes letting students, teachers, and parents understand and learn the local culture.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Boyatzis RE Competencies in the 21st century Journal of Management Development 2008 27 1 5 12 10.1108/02621710810840730
Chang CH A study of teacher’s difficulty in the public schools changing to experimental education schools: Qualitative meta-analysis research School Administrators 2021 131 118 137 10.6423/HHHC.202101_(131).0007
Chen, C. C. (2020). A case study of the remote principal’s management strategies in one private management of public management experimental school [Paper presentation] East Asian Principalship International Conference (Vol. 2020).
Chen CC Lin KP A case study of the president’s leadership in one private southern university of technology Journal of Higher Education 2019 14 1 41 79 10.3966/199254762019061401002
Chin JMC Chuang CP Experimental education in Taiwan: An analysis of policies and problems Journal of Education and Research 2019 299 55 74 10.3966/168063602019030299004
Chin MC A study on the outcomes of charter schools in the United States Educational Resources and Research 2014 115 169 192 10.6724/ERR.201411_(115).0007
Cofsky KM Critical keys to competency-based pay Compensation and Benefits Review 1993 25 6 46 52 10.1177/088636879302500609
Cohen L Manion L Morrison K Research methods in education 2011 7 Routledge
Conley, B. E. (2010). Alternative schools: A reference handbook. http://ebooks.abc-clio.com/reader.aspx?isbn=9781576074411&id=ALTSCLE-h2-.ABC-CLIO
Choi Y Lee S Oh H From the best practices of successful school leaders: Developing and validating the principal competencies inventory in South Korea Educational Management Administration and Leadership 2022 10.1177/17411432221082914
Denzin NK Denzin KD Lincoln YS The art and politics of interpretation Handbook of qualitative research 1994 1 Sage 500 515
Grace G School leadership: Beyond education management 2005 Routledge
Grzesik K Piwowar-Sulej K Project managers’ competencies and leadership styles from the perspective of organizations functioning in Poland Journal of Entrepreneurship, Management and Innovation 2018 14 3 35 60 10.7341/20181432
Hellriegel D Jackson SE Slocum JW Management: A competency based approach 2005 10 South-Western
Hsiao HC Wu CF Huang TC Developing a leadership competency model through outstanding leadership award of vocational school principals by the job competence assessment method Educational Policy Forum 2010 13 4 33 67
Hsieh CC Tseng HK Chen RJC Transformation from traditional schools to alternative schools: Curriculum leadership of the principals of Taiwanese aborigines Asia Pacific Education Review 2021 22 1 53 66 10.1007/s12564-020-09663-9
Lamb, L. D. (2014). An investigation of teacher, principal, and superintendent perceptions on the ability of the national framework for principal evaluations to measure principals’ leadership competencies [Unpublished Doctoral Dissertation]. Missouri Baptist University.
Lee CN Cheng TL A case study of a Taiwanese experimental school female founder’s life: Li Ya-Qing, founder of the seedling experimental elementary school Research in Applied Psychology 2019 70 279 316
Leithwood K Harris A Hopkins D Seven strong claims about successful school leadership revisited School Leadership and Management 2020 40 1 5 22 10.1080/13632434.2019.1596077
Liang JJ Fan CW A study of connotations and developmental implementations of quality assurance for school-based experimental education in Taiwan Journal of Education and Research 2020 315 41 56 10.3966/168063602020070315003
Lin KP Chen CC A study of the processes of teachers’ professional development: The case of an elementary school changing from a charter school to an experimental education school Educational Policy Forum 2020 23 1 93 123 10.3966/156082982020022301004
Lincoln YS Guba EG Naturalistic inquiry 1985 Sage
Liu CN Implementation of school-style experimental education in remote-area schools: From turning point to obstacle and way out of practice-take the Taitung County as example Journal of Educational Research 2018 287 52 66 10.3966/168063602018030287004
McClelland DC Testing for competence rather than for “intelligence” American Psychologist 1973 28 1 1 14 10.1037/h0034092 4684069
Ministry of Education. (2014a). Enforcement act for school-based experiment education. https://law.moj.gov.tw/ENG/LawClass/LawAll.aspx?pcode=H0070060
Ministry of Education. (2014b). Curriculum guideline of 12-year basic education. https://www.naer.edu.tw/upload/1/16/doc/1325/%E5%8D%81%E4%BA%8C%E5%B9%B4%E5%9C%8B%E6%95%99%E8%AA%B2%E7%A8%8B%E7%B6%B1%E8%A6%81%E7%B8%BD%E7%B6%B1(%E8%8B%B1%E8%AD%AF%E7%89%88).pdf
Ministry of Education. (2017). Guidelines for merger or closure of public elementary and secondary schools. https://law.moj.gov.tw/LawClass/LawAll.aspx?pcode=H0070070
Parry SB Just what is a competency? and why should you care? Training 1998 6 58 64
Saruchera, F. (2022). Effective agile project leadership through competency-based self-reflection. In Advances in logistics, operations, and management science (pp. 89–110). IGI Global. 10.4018/978-1-7998-7872-8.ch006
Spencer LM Spencer SM Competence at work: Models for superior performance 1993 Wiley
Taiwan Experimental Education Center. (2021). Handbook for the conducting of experimental education. https://teec.nccu.edu.tw/tw/news/show.php?num=441
Tseng HK Hsieh CC Chen RJC Practice and reflection of principals’ transformative curriculum leadership in a Taiwan KIST school KEDI Journal of Educational Policy 2021 18 2 65 83
Wang LH Gurr D Drysdale L Successful school leadership: Case studies of four Singapore primary schools Journal of Educational Administration 2016 54 3 270 287 10.1108/JEA-03-2015-0022
Yazan B Three approaches to case study methods in education Merriam: Yin, and Yazan, B. Three Approaches to Case Study Methods in Education: Yin, Merriam, and Stake Qualitative Report 2015 20 2 134 152 10.46743/2160-3715/2015.2102
Yin RK Case study research: Design and methods 2009 4 Sage
| 0 | PMC9741918 | NO-CC CODE | 2022-12-14 23:29:58 | no | Asia Pacific Educ. Rev.. 2022 Dec 11;:1-12 | utf-8 | null | null | null | oa_other |
==== Front
J Clin Immunol
J Clin Immunol
Journal of Clinical Immunology
0271-9142
1573-2592
Springer US New York
1409
10.1007/s10875-022-01409-z
Letter to Editor
Through Education and Collaboration to Better Care for Primary Imunodeficiencies in Albania and Kosovo
http://orcid.org/0000-0002-9654-4103
Ismaili-Jaha Vlora [email protected]
1
Kuli-Lito Gjeorgjina [email protected]
2
Spahiu-Konjusha Shqipe 1
Baloku Arbana 1
Maródi László 34
1 grid.412416.4 0000 0004 4647 7277 Department of Pediatrics, University Clinical Center of Kosova, Prishtina, Kosovo Albania
2 grid.412765.3 0000 0004 8358 0804 Department of Pediatrics, University Hospital “Mother Theresa”, Tirana, Albania
3 grid.11804.3c 0000 0001 0942 9821 PID Clinical Unit and Laboratory, Department of Dermatology, Venereology and Dermatooncology, Semmelweis University, Mária Út 41, Budapest, 1085 Hungary
4 grid.134907.8 0000 0001 2166 1519 St. Giles Laboratory of Human Genetics of Infectious Diseases, Rockefeller University, 1230 York Ave, New York, NY 10065 USA
11 12 2022
15
17 6 2022
16 11 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Keywords
Primary imunodeficiencies
J Project
Kosovo
Albania
==== Body
pmcPrimary immunodeficiency diseases (PIDs) also referred to as inborn errors of immunity (IEI) have long been neglected as medical conditions but are now recognized as a worldwide health problem. The rapid progress of research in this field has widened the gap between cutting-edge medical care in more developed western countries and the lack of appropriate diagnosis and treatment of these conditions in most countries, especially in those with poor socioeconomic conditions [1, 2]. Being in development, both Albania and especially Kosovo are good examples. The J Project physician education and clinical research collaboration program, however, resulted in a remarkable change in PID education and treatment in our countries over the past years and opened the way to the development of this previously neglected area of clinical medicine.
The J Project Albania (JPA)
Albania joined the J Project (JP) in 2009, and the very first meeting was organized in 2010 which was attended by 120 participants (Photo 1). At this inauguration meeting, we hosted the Albanian Minister of Health. Since then, JP meeting was held every year at different locations, mostly in Tirana, Albania. At each of these meetings, participation was growing both in number and diversity including pediatricians, immunologists, general practitioners, and other health professionals.Photo 1 Fist J Project meeting in Albania, year 2009
The JPA established the National Registry of Patients with IEI, and currently, the registry includes 262 patients. Of these, 132 belong to the primary immunodeficiency diseases (Table 1). All patients, except one case with CVID, are diagnosed below the age of 18. Ten children are diagnosed during the first year of life. Until now, there is no any screening program for immunodeficiency or De George syndrome.Table 1 Patients with IEI according to IUIS classification in Albania
Inborn errors of immunity according to the IUIS classification (until December 2021) Abs. number
1. Immunodeficiency affecting cellular and humoral immunity 6
2. combined immunodeficiencies with associated or syndromic features 17
3. Predominantly antibody deficiencies 73
4. Diseases of immune dysregulation 5
5. Congenital defects of phagocyte number or function 4*
6. Defects in intrinsic and innate immunity 4
7. Autoinflammatory disorders 12
8. Complement deficiencies 2
9. Bone marrow failure 5
10. Phenicopies of inborn errors of immunity 1
Undefined PID 3
Total registered patients 132
Prevalence of PID per 105 inhabitants 2.86**
*This number doesn’t include 130 patients with Cystic fibrosis
**Selective Iga excluded from the total number of Pid patients
The immunologic service in Albania is part of the University Hospital Center of Tirana “Mother Theresa,” and it did clinical consultations of the patients in collaboration with the specialists of pediatric infectious diseases, hematology, pulmology, or other services where the patients are diagnosed. There is an immunologic lab, mainly for research work directed by the Academy of Science. The immunologic laboratory investigations are performed by the Central Laboratory of UHC “Mother Theresa.” Genetic investigations were done by joint projects or agreements between UHC and the other hospitals. Some of the patients are diagnosed in “Gaslini” hospital in Genoa, Italy, and the Acibadem hospital in Istanbul, Turkey, and mainly, the genetic examinations are analyzed at CENTOGENE, Rostock, Germany.
The only data that we have for vaccination complications are two cases. One with Bruton who developed vAPP (vaccine-associated poliomyelitis) after vaccination with OPV at the age of 4 months. The other child of 5 months old developed a fulminate fasciitis after BCG, in terrene of SCID (diagnosed post mortem).
The only center involved in the PID registry is the University Hospital Center”Mother Theresa” in Tirana, with its Department of Pediatrics and Central Laboratories.
The J Project Kosovo (JPK)
Primary immunodeficiency diseases have been brought to the attention of the medical authorities in Kosovo only in the past few years. For those familiar with the socioeconomic and medical care conditions in Kosovo, this is not surprising. With 31% of the population being 0–18 years old and with an average age of 29.5 years, Kosovo has the youngest population in Europe, but it also has the poorest child health indicators in the region. This country has the highest infant mortality rate in Europe with 15 deaths per 1000 live births which is nearly two times higher among children from Roma, Ashkali, and Egyptian communities reaching 26 per 1000 live births [3]. Child malnutrition continues to be a concern in Kosovo, and striking inequalities exist for children in the poorest households. Nearly 23% of children live in poverty, and 7% live in extreme poverty. Only one in three children between 6 and 36 months of age receives the minimum acceptable diet. Only 73% of children under 2 years of age are fully immunized. When it comes to the Roma, Ashkali, and Egyptian communities, the rate goes down to only 38%. Only 20% of household members rely on clean fuels and technology for cooking, heating, and lighting. Nearly 90% of households have access to clean drinking water, but 23% of the population still consumes water tainted with E. coli. Therefore, infective diarrhea and pneumonia continue to be the most common childhood diseases [3]. Kosovo also has a very low budget for health. With about $100 per capita per year, it is probably the lowest-income country in Europe. Importantly, almost half of this budget goes to salaries.
IEIs and other rare diseases still fall out of the attention of the health government and remain the problem for the patients and their families itself and the medical community. Although frequent infections, anemia, thrombocytopenia, inflammations, digestive tract problems, growth delay, and autoimmune diseases are indicative of IEIs, it was not possible before to confirm with laboratory tests. Therefore, the entire treatment was insufficient and focused on the treatment of complications.
While a specialized diagnostic lab for IEI is still missing in Kosovo, recent progress is promising, largely because of the implementation of the JP. From 2018 when a small group of enthusiastic doctors within the Pediatric Clinic became a member of the JP community, JPK has transformed into a movement numbering more than one hundred doctors, mostly pediatricians, family medicine specialists, and other concerned specialists. By promoting IEI diagnostics through research and collaboration, JPK has held annual meetings to discuss contemporary diagnosis and treatment for these diseases. Apart from four large annual events, two additional, smaller meetings were held in various regions to foster inclusiveness. One of these meetings was held during the COVID-19 pandemic in 2021. Hundreds of pediatricians, general physicians, immunologists, and other specialists attended these meetings, among them a number of colleagues from neighboring Albania and the Republic of North Macedonia, and on one occasion from the USA. As a result, the number of patients diagnosed with IEIs has increased (Fig. 1). While in 2018, only five patients were diagnosed with IEI, in 2022 we registered 17 patients. Over the years, the number of genetically confirmed diseases has also increased up to nine (Table 2).Fig. 1 a Increase in the number of patients through the years in Albania. b Increase in the number of patients through the years in Kosovo
Table 2 Patients with IEI according to IUIS classification in Kosovo
Country Center Reported by Inborn errors of immunity according to the IUIS classification ® All Population Pts/103
1 2 3 4 5 6 7 8 9 10 UD
Kosovo Prishtina Ismaili-Jaha V 1 4 8 0 4 0 0 0 0 0 0 17 1.935.000 0.88
JPK has been working closely with the Association of Kosovo Pediatricians’ Society and the Kosovo Association for Rare Diseases. Together, we managed to influence the policymakers within the Ministry of Health of the Republic of Kosovo to include the drugs for the treatment of IEI at the list of essential drugs and to establish a registry of patients. Since then, we have registered 17 patients (9 boys (52%) and 8 girls (48%). One of these patients who had—Telangiectasia passed away recently at the age of 7. The most common subcategory of IEI according to the IUIS classification was type 3 error representing primarily antibody deficiencies (8/17 or 47.06%). Type 2 and type 4 subcategories represented similar number of patients (4/14 each or 23.53% each). With a population of 1.935 million, Kosovo has a prevalence of confirmed IEIs of 0.88 (Table 2). The low prevalence of the IEI in Kosovo probably does not reflect the real frequency of these diseases in the country, but rather the need for more work to increase awareness. Additionally, the Ministry of Health has recognized the specialty of clinical immunology and has established a separate department to diagnose and treat immunodeficiency diseases within the University Clinical Center of Kosovo, the largest medical institution in the country. The Kosovo Government also covers the expenses of genetic testing. The JPK was supported from the Medical Chamber of Kosovo that accredited all JP programs. The JP program encouraged the families of the patients with IEI to organize themselves, and they are now important advocates within the country.
Advances have also been made in the treatment of patients with IEI. Two patients are receiving immunoglobulins, one of them with X-linked agamaglobulinemia and another with common variable immunodeficiency. One of the patients was treated with bone marrow transplantation in Turkey.
Abstract and Further Challenges
Both countries largely benefited from the association with the JP and should continue to actively participate in this program. On the national level, it increased awareness on PIDs and improved diagnosis and treatment. It also helped to shape national health politics and establish registries and subsidy medical treatment for the patients. The JP has also built cross-society cooperation and involved patients and the IEI community in this action. On the regional level, it helped both Kosovo and the Albania to build strong alliances with JP in the Balkans and coordinate professional collaboration. Members from our societies attended international meetings held in Western European countries, exchange ideas and experiences, and plan future activities. On the international level, we attended several JP meetings and participated at the biannual meetings of the European Society for Immunodeficiency. Several important documents were adopted and then implemented at these meetings. One of these documents, that advocates for free access to genetic testing for all patients, is the Antalya Declaration. We hope to help us to diagnose more patients and offer them up-to-date treatment earlier [4]. This will also help our countries to overcome the biggest difficulties, insufficient testing infrastructure, and lack of financial resources. Future challenges include raising awareness on IEIs, mobilizing health professionals, and increasing patient groups’ activity. We are devoted and eager to participate more actively in IEI-focused clinical research locally and in collaboration with foreign groups.
Acknowledgements
We thank the patients and parents of the patients for their helpful collaboration and trust. We thank the healthcare professionals, especially clinicians and laboratory experts who supported the JP centers.
Author Contribution
VIJ and GKL designed this study and wrote the paper. All co-authors read the paper and approved its content.
Data Availability
Data used in this study are available on request.
Declarations
Ethics Approval
This publication was approved by the Research Ethics Committees of Kosovo.
Consent to Participate
All authors approved to participate in this study.
Consent for Publication
All authors approved the publication of the manuscript.
Conflicts 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. Maródi L Casanova JL Primary immunodeficiency diseases: the J Project Lancet 2009 373 2179 2181 10.1016/S0140-6736(09)61171-5 19560586
2. Pac M, Casanova JL, Reisli I, Maródi L. Advances in primary immunodeficiency in Central-Eastern Europe. Frontiers in immunology. 2021;12.
3. Multiple Indicator Cluster SurveyKosovo. MICS 2020 key findings in snapshots | UNICEF Kosovo Programme. 2020.
4. Maródi L; J Project Study Group. The Konya Declaration for patients with primary immunodeficiencies. J Clin Immunol. 2020;40(5):770–773. 10.1007/s10875-020-00797-4.
5. Albanian Demographic and Health Survey. Albania demographic and health survey 2017–18 (dhsprogram.com). 2018.
| 36504257 | PMC9741942 | NO-CC CODE | 2022-12-14 23:29:58 | no | J Clin Immunol. 2022 Dec 11;:1-5 | utf-8 | J Clin Immunol | 2,022 | 10.1007/s10875-022-01409-z | oa_other |
==== Front
Med Care
Med Care
MLR
Medical Care
0025-7079
1537-1948
Lippincott Williams & Wilkins Hagerstown, MD
36477619
10.1097/MLR.0000000000001788
00008
3
Original Articles
Changes in Health Care Access by Race, Income, and Medicaid Expansion During the COVID-19 Pandemic
Auty Samantha G. MS *[email protected]
Aswani Monica S. PhD †[email protected]
Wahbi Rafik N. MPH [email protected]
‡
http://orcid.org/0000-0002-8304-8602
Griffith Kevin N. PhD §[email protected]
* Department of Health Law, Policy and Management, Boston University School of Public Health, Boston, MA
† Department of Health Services Administration, University of Alabama at Birmingham, Birmingham, AL
‡ Department of Community Health Sciences, University of California, Los Angeles, Los Angeles, CA
§ Department of Health Policy, Vanderbilt University School of Medicine, Nashville, TN
Correspondence to: Kevin Griffith, PhD, 2525 West End Avenue, Suite 1204, Nashville, TN, 37203. E-mail: [email protected].
1 2023
1 11 2022
1 11 2022
61 1 4549
Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.
2023
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Background:
The intersecting crises of the COVID-19 pandemic, job losses, and concomitant loss of employer-sponsored health insurance may have disproportionately affected health care access within minorized and lower-socioeconomic status communities.
Objective:
To describe changes in access to care during the COVID-19 pandemic, stratified by race/ethnicity, household income, and state Medicaid expansion status.
Research Design:
We used interrupted time series and difference-in-differences regression models, controlling for respondent characteristics and preexisting trends.
Subjects:
Data were extracted for all adults aged 18–64 surveyed in the 2015-2020 Behavioral Risk Factor Surveillance System (N=1,731,699) from all 50 states and the District of Columbia.
Measures:
Our outcomes included indicators for whether respondents had any health insurance coverage or avoided seeking care because of cost within the prior year. The primary exposure was the onset of the COVID-19 pandemic in the United States in March 2020.
Results:
The pandemic was associated with a 1.2 percentage point (pp) decline in uninsurance for Medicaid expansion states (95% CI, −1.8, −0.6); these reductions were concentrated among respondents who were Black, multiracial, or low income. The rates of uninsurance were generally stable in nonexpansion states. The rates of avoided care because of cost fell by 3.5 pp in Medicaid expansion states (95% CI, −3.9, −3.1), and by 3.6 pp (95% CI, 4.3–2.9) in nonexpansion states. These declines were concentrated among respondents who were Hispanic, Other Race, or low income.
Conclusions:
Our findings reinforce the value of Medicaid expansion as one tool to improve access to health insurance and care for marginalized and vulnerable populations.
Key Words:
access to care
Medicaid
health disparities
COVID-19
SDCT
OPEN-ACCESSTRUE
==== Body
pmcThe sudden and unforeseen onset of the COVID-19 pandemic had a profound impact on the US economy and delivery of health services. State and local policy responses to the pandemic frequently included extensive mobility restrictions,1 and hospitals intentionally stalled most routine and elective procedures.2 Although these policies mitigated the spread of COVID-19, they also decreased access to routine, urgent, and emergency care. Likewise, the fear of exposure to COVID-19 led many individuals to delay care.2 The pandemic also triggered a recession with an unemployment rate of almost 15% in April 2020—the highest level since the Great Depression—and a corresponding rise in uninsurance because of losses in employer-sponsored health insurance coverage (ESHI).3 Marginalized racial and ethnic groups were disproportionately impacted by job losses, especially among women.4 Racial and sex disparities in unemployment rates have yet to recover with potential long-term consequences on access to health services and health outcomes.4
The Affordable Care Act extended the reach of the US safety net primarily through the implementation of state-based Marketplaces for individual insurance coverage and expanded income eligibility for Medicaid coverage, which has been adopted by 38 states and the District of Columbia as of July 2022.5 These reforms reduced income-based and race-based disparities in health care access,6,7 and Medicaid expansion may have facilitated health insurance coverage and subsequent access to care for those who lost ESHI during the COVID-19 pandemic. Prior research found that Medicaid enrollment increased during the COVID-19 pandemic,8 particularly among those earning between 138% and 400% of the federal poverty level (FPL).9 However, these prior works relied on an experimental survey with an approximate 7% response rate. Other studies found changes in unemployment were only weakly10 or negatively11 associated with Medicaid enrollment growth.
The ability of the US health care safety net to respond to race-based/income-based disparities in health care access during the pandemic remains unclear. The intersecting crises of the pandemic, unemployment, and uninsurance may have exacerbated preexisting disparities in insurance coverage and access to care among minoritized groups.12 Medicaid expansion may have mitigated increased inequities through policies that have previously promoted insurance coverage and access to care. Accordingly, the objectives of this study were to assess if Medicaid expansion status was associated with changes in insurance status and self-reported access to care during the pandemic by race/ethnicity and household income as a percentage of the FPL.
METHODS
Data and Measures
We extracted data for all US adults aged 18–64 from the 2015 to 2020 Behavioral Risk Factor Surveillance System (BRFSS), a large national phone based survey conducted by states in partnership with the Centers for Disease Control and Prevention.13 Despite the onset of the COVID-19 pandemic, the response rate was 47.4% for the 2020 survey, similar to that achieved in 2019 (49.4%).14 The BRFSS incorporates survey weights to ensure the weighted sample matches the sociodemographic composition (ie, age, race/ethnicity, education level, marital status, and home ownership) both nationally and for individual states.14 Data on state Medicaid expansion status were obtained from the Kaiser Family Foundation.5
Primary Outcomes
Our outcomes included 2 binary measures of self-reported access to health care. The first asked respondents “Do you have any kind of health care coverage, including health insurance, prepaid plans such as HMOs, or government plans such as Medicare, or Indian Health Service?” The second asked “Was there a time in the past 12 months when you needed to see a doctor but could not because of cost?” BRFSS does not collect data on insurance type (ie, Medicaid, ESHI, etc.) or other reasons why respondents may have avoided care.
Primary Exposure
The primary exposure was a dummy variable taking on a value of 1 if the survey was completed between March and December 2020, zero otherwise.
Covariates
Other individual-level covariates included age, race/ethnicity, sex, marital status, presence of children in the household, home ownership, educational attainment, income, veteran status, employment status, and household size. Respondents’ race/ethnicity were categorized as non-Hispanic White, non-Hispanic Black, non-Hispanic multiple races, non-Hispanic other race, or Hispanic. Income measurement in the BRFSS is imprecise. Thus, after the methodology outlined by Sommers et al,15 we imputed household income as a percentage of the FPL, which was then categorized into 3 groups: <138%, 138%–400%, and >400% FPL. We also included a dummy variable taking on a value of 1 if the respondent resided in a state that implemented the Affordable Care Act’s Medicaid expansion by December 31, 2020, zero otherwise.5
Analytic Approach
First, we used interrupted time series models to estimate overall changes in health care access by state Medicaid expansion status during March to December 2020, controlling for prepandemic outcome levels and respondent characteristics. Next, we estimated difference-in-differences (DID) regression models to identify whether there were varying changes in outcomes between expansion and nonexpansion states after the onset of the COVID-19 pandemic. We estimated both overall models and stratified our analyses by respondents’ race/ethnicity, household income, and state Medicaid expansion status. Given our large sample size, all regressions were estimated as linear probability models using ordinary least squares.6,16,17 All models included the aforementioned covariates with state and year fixed effects and standard errors clustered at the state level. We did not include employment status or COVID-19 burden (eg, infections, hospitalizations, stay-at-home orders) as these may be on the causal pathway between exposure and outcomes. For more details including regression specifications, please see the Appendix, Supplemental Digital Content 1, http://links.lww.com/MLR/C554.
RESULTS
Our analytic sample included 1,731,699 respondents; sample characteristics are presented were similar before and during the COVID-19 pandemic (Table 1). Before the pandemic (January 2019 to February 2020), uninsurance was higher among respondents residing in Medicaid nonexpansion states (21.1%) compared with those in expansion states (12.9%) (Table 2). Avoidance of care due to cost was also higher in Medicaid nonexpansion (19.0%) than expansion states (13.7%) (Table 3). There were no significant differences in prepandemic trends between respondents in Medicaid expansion and nonexpansion states for rates of uninsurance and care avoidance, either overall or when stratified by race/ethnicity and income group (Appendix A1, Supplemental Digital Content 1, http://links.lww.com/MLR/C554 & A2, Supplemental Digital Content 1, http://links.lww.com/MLR/C554).
TABLE 1 Characteristics of the Study Sample From 2015 to 2019 and 2020
2015-2019 (N=1,508,630) 2020 (N=223,069)
Variable Raw % Weighted %* Raw % Weighted %*
Female 53.9 50.2 52.5 50.3
Married 55.9 51.9 54.4 51.2
Children in household 38.0 44.0 38.6 42.8
Age group
18–24 9.0 15.9 10.0 15.8
25–29 7.6 10.3 8.2 10.3
30–34 8.5 11.7 9.1 11.8
35–39 9.1 10.1 10.0 10.2
40–44 9.1 10.3 10.0 10.3
45–49 10.5 9.2 10.4 9.1
50–54 13.2 11.3 12.2 11.2
55–59 15.7 10.4 14.1 10.4
60–64 17.4 10.7 16.0 10.8
Race/ethnicity
White 73.3 60.2 71.9 58.4
Black 9.0 12.6 8.4 12.4
Multiracial 2.3 1.5 2.5 1.4
Hispanic 9.7 18.0 10.5 19.3
Other 5.7 7.8 6.7 8.5
College graduate 38.1 27.9 39.7 30.2
Household income group
<138% FPL 23.1 26.6 23.6 25.7
138%–400% FPL 43.2 42.1 43.7 43.4
>400% FPL 33.7 31.3 32.7 30.9
Employment status
Employed 5.6 6.4 8.4 9.8
Unemployed 68.5 68.0 69.1 67.0
Not in labor force 25.9 25.6 22.4 23.2
Homeownership status
Renter 35.4 38.3 37.1 38.5
Homeowner 64.6 61.7 62.9 61.5
* Data are adjusted for BRFSS sampling weights.
FPL indicates federal poverty level.
Source: Author’s analysis of data from the 2015 to 2020 Behavioral Risk Factor Surveillance System (BRFSS).
TABLE 2 Changes in Uninsurance During the COVID-19 Pandemic by Race, Income, and State Medicaid Expansion Status
Unadjusted rates in 2019† Adjusted changes 2019–2020‡
Expansion status (%) Expansion status
Yes No Yes No Difference§
Overall 12.9 21.1 −1.2 (−1.8, −0.6)*** −0.2 (−0.7, 0.3) −1.0 (−1.7, −0.3)**
Racial category‖
Black 13.8 21.6 −1.6 (−2.7, −0.4)** −0.1 (−1.5, 1.2) −1.4 (−3.1, 0.2)
Other race 29.1 41.5 −2.3 (−4.2, −0.5)* 1.8 (0.2, 3.5)* −4.2 (−6.6, −1.8)***
Multiple race 11.3 18.8 −3.1 (−4.5, −1.6)*** −3.7 (−5.7, −1.7)*** 0.5 (−1.7, 2.7)
Hispanic 10.3 18.5 −0.9 (−2.5, 0.7) 2.7 (−1.9, 7.4) −3.6 (−8.0, 0.7)
White 8.8 14.7 −0.5 (−0.9, −0.1)** −0.5 (−1.5, 0.4) 0.0 (−0.9, 0.9)
Household income¶
<138% FPL 22.4 36.0 −1.9 (−3.2, −0.6)** 0.6 (−1.0, 2.3) −2.1 (−4.1, −0.1)*
138%–400% FPL 12.5 18.2 −1.2 (−2.0, −0.4)** 0.5 (−0.7, 1.7) −1.3 (−2.6, −0.1)*
>400% FPL 5.8 9.8 −0.3 (−1.0, 0.5) 0.1 (−0.9, 1.2) −0.3 (−1.4, 0.8)
The exhibit displays changes in the percentage of noninstitutionalized US adults aged 18–64 years who reported that they lacked insurance coverage. “Expansion states” are those that expanded eligibility for Medicaid by the end of 2020. Standard errors are adjusted for clustering at the state level.
* P<0.05.
** P<0.01.
*** P<0.001.
† Weighted means using BRFSS sampling weights.
‡ Regression estimates adjusted for covariates described in the text. Numbers in parentheses represent 95% CIs.
§ Difference between expansion and nonexpansion states in changes over time, adjusted for covariates.
‖ BRFSS 5-level race/ethnicity category.
¶ Income group is defined as an imputed percentage of federal poverty level (FPL); see the Appendix for more details.
Source: Authors analysis of data from the Behavioral Risk Factor Surveillance System (BRFSS) 2015–2020. Data for years 2015–2018 were included to control for baseline trends.
TABLE 3 Changes in Avoided Care Because of Cost During the COVID-19 Pandemic by Race, Income, and State Medicaid Expansion Status
Unadjusted rates in 2019† Adjusted changes 2019-2020‡
Expansion status Expansion status
Yes No Yes No Difference§
Overall 13.7 19.0 −3.5 (−3.9, −3.1)*** −3.6 (−4.3, −2.9)*** 0.1 (−0.6, 0.9)
Racial category‖
Black 15.4 21.2 −3.4 (−4.6, −2.1)*** −3.1 (−6.0, −0.2)* −0.2 (−3.3, 2.7)
Other race 20.8 25.9 −5.1 (−6.1, −4.1)*** −5.7 (−6.7, −4.6)*** 0.5 (−0.9, 2.0)
Multiple race 11.4 17.3 −2.6 (−3.6, −1.5)*** −2.3 (−5.7, 0.9) −0.2 (−3.1, 2.8)
Hispanic 16.6 23.9 −5.5 (−8.6, −2.5)*** −1.5 (−7.2, 4.2) −4.0 (−9.7, 1.7)
White 11.8 16.2 −3.1 (−3.5, −2.8)*** −3.2 (−4.0, −2.4)*** 0.0 (−0.01, 0.1)
Household income¶
<138% FPL 21.5 31.1 −6.1 (−6.9, −5.2)*** −5.7 (−7.3, −4.2)*** −0.2 (−1.9, 1.5)
138%–400% FPL 13.9 17.3 −3.5 (−4.0, −2.9)*** −3.3 (−4.1, −2.6)*** 0.0 (−0.9, 0.9)
>400% FPL 7.1 8.7 −1.4 (−1.9, −0.9)*** −0.8 (−1.8, 0.1) −0.5 (−1.5, 0.4)
* P<0.05.
** P<0.01.
*** P<0.001.
† The exhibit displays changes in the percentage of noninstitutionalized US adults aged 18–64 years who reported that they avoided medical care because of cost. “Expansion states” are those that expanded eligibility for Medicaid by the end of 2020. Standard errors are adjusted for clustering at the state level.Weighted means using BRFSS sampling weights.
‡ Regression estimates adjusted for covariates described in the text. Numbers in parentheses represent 95% CIs.
§ Difference between expansion and nonexpansion states in changes over time, adjusted for covariates.
‖ BRFSS 5-level race/ethnicity category.
¶ Income group is defined as an imputed percentage of federal poverty level (FPL); see the Appendix for more details.
Source: Authors analysis of data from the Behavioral Risk Factor Surveillance System (BRFSS) 2015–2020. Data for years 2015–2018 were included to control for baseline trends.
Adjusted Changes in Expansion States
Respondents in Medicaid expansion states experienced a 1.2 percentage point (pp) decrease in uninsurance during the COVID-19 pandemic (95%: CI, −1.8, −0.6). The largest declines were observed in multiple race (−3.1 pp, 95% CI, −4.5 pp, −1.6), Other race (−2.3 pp, 95% CI, −4.2, −0.5), and in households with income below 138% (−1.9 pp, 95% CI, −3.2, −0.6) and 400% (−1.2 pp, 95% CI, −2.0, −0.4) FPL. Stratified results also indicate coverage gains for both White and Hispanic respondents in households earning <138% FPL (Appendix A3, Supplemental Digital Content 1, http://links.lww.com/MLR/C554).
In Medicaid expansion states, the rates of avoided care because of cost decreased by 3.5 pp overall (95% CI −3.9, −3.1); all race and income strata experienced declines in 2020 (Appendix A4, Supplemental Digital Content 1, http://links.lww.com/MLR/C554). The largest declines were observed in other race (−5.1 pp, 95% CI −6.1, −4.1), Hispanic (−5.5 pp, 95% CI −8.6, −2.5), and households earning <138% FPL (−6.1 pp, 95% CI −6.9, −5.2).
Adjusted Changes in Nonexpansion States
The pandemic was not associated with significant changes in uninsurance rates for Medicaid nonexpansion states. However, uninsurance decreased by 3.7 pp for multiple race (95% CI −5.7, −1.7) and increased by 1.8 pp for other race respondents (95% CI, 0.2, 3.5). Stratification revealed that Hispanic respondents in households earning <138% of the FPL experienced a significant increase in uninsurance (9.2 pp, 95% CI, 1.1, 17.3), whereas White respondents earning <138% FPL experienced a significant decrease (−1.8 pp, 95% CI, −3.1, −0.4).
Overall, nonexpansion states experienced a 3.6 pp decline in avoided care because of cost (95% CI, −4.3, −2.9). The observed decline in avoided because of cost was primarily driven by Black (−3.1pp, 95% CI, −6.0, −0.2), other race (−5.7 pp, 95% CI, −6.7, −4.6), and White (−3.2 pp, 95% CI −4.0, −2.4) respondents. Households earning <138% (−5.7 pp, 95% CI, −7.3, −4.2) and between 138% and 400% (−3.3 pp, −4.1, −2.6) also experienced declines in avoided care.
Difference-in-Differences Estimates
In DID models, the COVID-19 pandemic was associated with a 1.0 pp decrease in uninsurance in Medicaid expansion relative to nonexpansion states (95% CI, −1.7, −0.3). In stratified models, the pandemic was associated with decreases in uninsurance among multiracial respondents (−4.2 pp, 95% CI, −6.6, −1.8 and in households earning <138% FPL (−2.1 pp, 95% CI, −4.1, −0.1) or between 138% and 400% FPL (−1.3 pp, 95% CI, −2.6, −0.1). There were no difference changes in avoided care because of cost between Medicaid expansion and nonexpansion states, either overall or in stratified models.
DISCUSSION
In this cross-sectional study, we observed slight declines in the national rates of uninsurance and avoided care because of cost during the COVID-19 pandemic. Our DID results also suggest that if all states had implemented the Medicaid expansion, an additional 1.0 million adults would have access to insurance coverage.18 At a population level, a 3.5% reduction in avoided care equates to 9 million fewer US adults experiencing that outcome in 2020 compared with previous years. Given that the BRFSS question on avoided care specifically focuses on cost; it is possible and perhaps likely that pandemic fears displaced cost as the preeminent reason to skip care during the pandemic. Future researchers using the BRFSS to study health care access should account for this peculiar pandemic effect.
The COVID-19 pandemic resulted in an economic recession and subsequently, an estimated 7.3 million people lost employer-sponsored insurance coverage.19 Despite these losses, we observed a slight decline in the national uninsurance rate during 2020. Although we cannot identify coverage source in BRFSS, our results also comport with estimates from the National Health Interview Survey,20 the Current Population Survey,21 the Survey of Income and Program Participation,22 and the Health Reform Monitoring Survey,11 which all suggest that gains in public coverage may have offset losses in private coverage.8,23 Thus, we have good reason to believe uninsurance and lack of health care access was prevented for some individuals because of Medicaid expansion. All states participated in the “maintenance of effort” provisions in the Coronavirus Aid, Relief, and Economic Security Act, which offered an enhanced federal match rate to states that facilitated continuous Medicaid coverage for beneficiaries. Consequently, the increase in Medicaid enrollment is attributable to greater continuity of existing coverage in addition to an influx of new enrollees.24
We also found uninsurance rates declined for many marginalized groups in Medicaid expansion states, whereas rates in nonexpansion states remained stable or worsened slightly. Racial/ethnic minority groups face a disproportionately high risk of COVID-19 exposure and increased risk of COVID-19-related disease or death.25,26 Health insurance coverage is critical to improve access to health services, and Medicaid expansion facilitated access to care during the pandemic for these vulnerable groups.
Limitations
This study has multiple limitations that warrant consideration. First, the BRFSS is a telephone-based survey and is thus prone to selection and nonresponse bias. However, the BRFSS provides complex weighing to mitigate these biases, and both response rates and respondent characteristics were similar before and during the pandemic. Second, the BRFSS only collects data on type of health insurance in a small number of states. Thus, we cannot estimate changes in insurance status by type. Relatedly, the BRFSS does not collect data on avoidance of care for reasons other than cost. The documented decline in avoidance of care because of cost may reflect national changes in the primary reasons for avoiding care (ie, fear of COVID-19, policies to reduce spread of COVID-19). This study focuses on national changes in insurance coverage and access to care during the COVID-19 pandemic, which may mask state-level differences in these outcomes that may stem from differences in Medicaid policies or population characteristics (eg, the proportion employed in health care or other essential roles). Finally, the observational nature of the study design limits any causal conclusions; our results should be interpreted as associations.
CONCLUSIONS
Taken together, our findings suggest modest improvements in overall insurance coverage within Medicaid expansion states during the COVID-19 pandemic. Our results also comport with prior work suggesting a shift toward public coverage may have blunted pandemic-related losses of employer-sponsored health insurance.23,24 Moreover, these results reinforce the value of Medicaid expansion as one tool to improve access to health insurance and care for marginalized and vulnerable populations, even during intersecting public health and economic crises.
Supplementary Material
SUPPLEMENTARY MATERIAL
Ms S.G.A. was supported by the National Institute of Drug Abuse [T32-DA041898-0SZA3] and Dr K.N.G. was supported by the Agency for Healthcare Research and Quality [K12-HS026395].
The authors declare no conflict of interest.
Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's website, www.lww-medicalcare.com.
==== Refs
REFERENCES
1 Feyman Y Bor J Raifman J . Effectiveness of COVID-19 shelter-in-place orders varied by state. PLoS One. 2020;15 :e0245008.33382849
2 Lawrence E . Nearly half of Americans delayed medical care due to pandemic. Kaiser Family Foundation. 2020.
3 Dorn S The COVID-19 pandemic and resulting economic crash have caused the greatest health insurance losses in American history. Families USA. 2020. Available at: https://familiesusa.org/resources/the-covid-19-pandemic-and-resulting-economic-crash-have-caused-the-greatest-health-insurance-losses-in-american-history/. Accessed August 30th, 2022.
4 Couch KA Fairlie RW Xu H . Early evidence of the impacts of COVID-19 on minority unemployment. J Public Econ. 2020;192 :104287.32952224
5 Status of State Action on the Medicaid Expansion Decision. Kaiser Family Foundation.2021. Available at: https://www.kff.org/health-reform/state-indicator/state-activity-around-expanding-medicaid-under-the-affordable-care-act. Accessed August 30, 2022.
6 Griffith K Evans L Bor J . The Affordable Care Act reduced socioeconomic disparities in health care access. Health Aff (Millwood). 2017;36 :1503–1510.
7 Buchmueller TC Levy HG . The ACA’s impact on racial and ethnic disparities in health insurance coverage and access to care: an examination of how the insurance coverage expansions of the Affordable Care Act have affected disparities related to race and ethnicity. Health Aff (Millwood). 2020;39 :395–402.32119625
8 Khorrami P Sommers BD . Changes in US Medicaid enrollment during the COVID-19 pandemic. JAMA Netw Open. 2021;4 :e219463–e219463.33950210
9 Benitez J Dubay L . COVID-19-related unemployment and health insurance coverage in medicaid expansion and non-expansion states. Health Serv Res. 2021;56 :20–20.
10 Clemens J Ippolito B Veuger S . Medicaid and fiscal federalism during the COVID‐19 pandemic. Public Budg Finance. 2021;41 :94–109.
11 Karpman M Zuckerman S . The Uninsurance Rate Held Steady During the Pandemic as Public Coverage Increased. Urban Institute; 2021.
12 Beech BM Ford C Thorpe RJ Jr . Poverty, racism, and the public health crisis in America. Front Public Health. 2021;1274 :699049.
13 Pierannunzi C Hu SS Balluz L . A systematic review of publications assessing reliability and validity of the Behavioral Risk Factor Surveillance System (BRFSS), 2004–2011. BMC Med Res Methodol. 2013;13 :1–14.23297754
14 Control CfD, Prevention. Behavioral Risk Factor Surveillance System, 2020. 2020.
15 Sommers BD Gunja MZ Finegold K . Changes in self-reported insurance coverage, access to care, and health under the Affordable Care Act. JAMA. 2015;314 :366–374.26219054
16 Lumley T Diehr P Emerson S . The importance of the normality assumption in large public health data sets. Annu Rev Public Health. 2002;23 :151–169.11910059
17 Griffith KN Jones DK Bor JH . Changes In Health Insurance Coverage, Access To Care, And Income-Based Disparities Among US Adults, 2011–17: Assessing access to care among US non-elderly adults before and after Trump administration policies that may have affected the Affordable Care Act’s effectiveness. Health Aff (Millwood). 2020;39 :319–326.32011953
18 U.S. Census Bureau. County Population by Characteristics: 2010-2018. US Census Bureau; 2020.
19 Banthin J Simpson M Buettgens M . Changes in Health Insurance Coverage Due to the COVID-19 Recession: Preliminary Estimates Using Microsimulation. Urban Institute; 2020:1–9.
20 Cohen RA Terlizzi EP Cha AE , Health insurance coverage: early release of estimates from the National Health Interview Survey, January–June 2020. 2021.
21 Ruhter J Conmy AB Chu R , Tracking health insurance coverage in 2020-2021. Washington, DC: US Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, Office of Health Policy. 2021.
22 US Department of Health Human Services. Office of the Assistant Secretary for Planning and Evaluation. Unwinding the Medicaid Continuous Enrollment Provision: Projected Enrollment Effects and Policy Approaches. 2022.
23 Corallo B Rudowitz R . Analysis of recent national trends in Medicaid and CHIP enrollment. Henry J Kaiser Family Foundation; 2021.
24 Dague L Badaracco N DeLeire T , Trends in medicaid enrollment and disenrollment during the early phase of the COVID-19 pandemic in Wisconsin. Paper presented at: JAMA Health Forum. 2022.
25 Gemelas J Davison J . Inequities in employment by race, ethnicity, and sector during COVID-19. J Racial Ethn Health Disparities. 2022;9 :350–355.33452573
26 Anyamele OD McFarland SM Fiakofi K . The disparities on loss of employment income by US households during the COVID-19 pandemic. JERP. 2022;5 :115–133.
| 36477619 | PMC9741953 | NO-CC CODE | 2022-12-14 23:29:58 | no | Med Care. 2023 Jan 1; 61(1):45-49 | utf-8 | Med Care | 2,022 | 10.1097/MLR.0000000000001788 | oa_other |
==== Front
J Vet Behav
J Vet Behav
Journal of Veterinary Behavior
1558-7878
1878-7517
Elsevier Inc.
S1558-7878(22)00151-4
10.1016/j.jveb.2022.12.001
Article
IMPACTS OF THE COVID-19 PANDEMIC ON THE BEHAVIOR AND PHYSICAL HEALTH OF DOGS IN RIO DE JANEIRO STATE: REFLECTIONS ON THE QUALITY OF LIFE OF DOGS AND THEIR OWNERS
Ribeiro Luana de Sousa a⁎
Soares Guilherme Marques b
Arnold Emmanuel c
Nobre e Castro Maria Cristina d
a Residency Program in the Professional Health Area in Veterinary Medicine, Universidade Fluminense Federal, Niterói, Rio de Janeiro, Brazil
b School of Veterinary Medicine, Universidade Santa Úrsula, Rio de Janeiro, Rio de Janeiro, Brazil
c Department of Animal Science, School of Veterinary and Animal Science, Universidade Federal de Goiás, Goiânia, Goiás, Brazil
d Department of Pathology and Veterinary Clinic, School of Veterinary Medicine, Universidade Federal Fluminense, Niterói, Rio de Janeiro, Brazil
⁎ Corresponding author: Luana de Sousa Ribeiro, DVM, MS, Residency in medical clinic of dogs and cats in Hospital Professor Firmino Marsico Filho, Universidade Fluminense Federal, Niterói, Rio de Janeiro, Brasil.
11 12 2022
11 12 2022
11 3 2021
4 12 2022
5 12 2022
© 2022 Elsevier Inc. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The COVID-19 pandemic changed the routines of people, consequently changing the daily lives of their pets. Behavioral and emotional changes caused by the stress resulting from restrictions of social isolation and their consequences in the human-animal relationship have been discussed. However, there are still no studies that identify the factors that affect behavior and which are the most susceptible groups. The purpose of this study is to identify behavioral and emotional changes on dogs during the COVID-19 pandemic and their effects on the quality of life of animals and their owners. The methodology used was online questionnaires, which were posted on social networks aimed to dog owners in Rio de Janeiro state, Brazil. The results showed that age, sex, dog size, type of home, and restrictions imposed differently affected the type of behavioral change. However, the most frequent type of change was the worsening of previous conditions.
Altered behaviors directly interfered in the lives of owners and their pets, as owners managed the situation and sought information without guidance from a veterinarian, with the possibly consequence of worsening the situation in the future. Veterinarians should actively investigate behavioral changes that have occurred through anamnesis to avoid abandonment and instability in the human-animal relationship.
Keywords
Abandonment
canine
behavior
epidemiology
SARSCoV-2
==== Body
pmcIntroduction
The COVID-19 pandemic has had repercussions in several areas, from health to economy. The influence on mental health and well-being has been one of the discussion points in the scientific community (Fiorillo and Gorwood, 2020; Homes et al., 2020). Social isolation combined with excessive information on the SARS-CoV-2 virus has had deleterious effect on people who may or may not have had pre-existing psychological diseases (Fiorillo and Gorwood, 2020). During psychotherapy, it's clear that animals have been used to relieve the anxiety generated by the pandemic (Hargrave et al., 2020; Vicent et al., 2020), with dogs and cats present in most households and being considered family members.
Sudden changes in routine may affect both humans and pets (Vicent et al., 2020), increasing stress in animals, which are vulnerable to significant behavioral changes if managed incorrectly (Hargrave et al., 2020). Generally, the classes of behavioral complaints most frequently reported by owners in consultations are aggressiveness, destructive behavior, and urination/defecation in inappropriate places (Fatjó et al., 2006; Soares et al., 2010), which may lead to euthanasia or animal abandonment (Fatjó et al., 2006; Soares et al., 2010; Hargrave et al., 2020; Vicent et al., 2020). The increase in numbers of stray animals and the risk of attacks against humans are serious public health problems (Slater, 2001; Fatjó et al., 2006). Changes in the routine of owners, and consequently of their animals, during the COVID-19 pandemic may increase the prevalence of the problems previously mentioned.
The importance of the veterinarian in this context is to intervene and advise on mental and physical diseases (Hargrave et al., 2020). Face-to-face consultations during the pandemic were challenging (Vicent et al., 2020), so epidemiological surveys are essential for the development of prevention and treatment strategies (Fatjó et al., 2006). These allow for a longer observation time to detect subtle abnormalities or behavioral changes that were not previously observed by the owners (Vicent et al., 2020). The use of questionnaires aimed at pet owners has been a useful tool to detect behavior abnormalities (Fatjó et al., 2006).
This study identifies the relationship between social isolation and the behavioral and physical health of pets, as well as the life quality of animals during COVID-19 pandemic. We also identified the extent to which dogs were important for their owners and the effects of the dogs’ behavioral changes on owners’ routine. The hypotheses are: 1) Change in routine affects the behavior of dogs, consequently interfering in the quality of life of owners; 2) Behavioral changes depend on age, lifestyle, and restrictions imposed.
Materials and Methods
The study was conducted with dog owners from Rio de Janeiro state, Brazil. First, a pilot questionnaire whose purpose was to validate the clarity, relevance, and time to complete the questionnaire elaborated on Google Forms® on behavioral changes of pets during the COVID-19 pandemic (Appendix 1) was conducted. Ten individuals who were not veterinarians were selected to answer the questionnaire elaborated. This group subsequently answered some questions on the understanding of questions and time spent to answer them (Appendix 2). After necessary adaptations, the questionnaires were disseminated through social networks (WhatsApp, Twitter, Facebook, and Instagram) and 300 responses were collected during the period from July to September 2020. The project was exempted from evaluation by the Human Research Ethics Committee (CONEP) through opinion 4,131,826.
The questionnaire was divided into four sections. The first section consisted of questions on the characterization of the environment where the animal lived and changes in the routine of owners after COVID-19 restrictions, which directly affected the life of the pet. The second section consisted of questions with the objective of characterizing the animal and identifying chronic diseases. Animal age groups were classified as puppies (≤ 1 year old), young adults (1 to 4 years old), adults (5 to 7 years old), and elderly (≥ 8 years old). Emotional predisposition was evaluated through the PANAS method adapted by Savalli et al. (2019), in which affect is classified as negative (frightened, anxious, fearful or phobic, and indifferent to the environment and routine changes) or positive (adaptation to unknown environments, insistent on playing, little interest in the environment, has a lot of energy and/or is lazy, persistence in disobeying, and being noisy). The third section consisted of an investigation into changes in the routine and behavior of animals after the COVID-19 pandemic restrictions, using a list of behaviors adapted from Fatjó et al. (2006) and Soares et al. (2010). In the last section, effects on the physical health of the pet and on the quality of life of the pet and owner were analyzed.
To evaluate behavioral changes and the perception of owners about these changes, 95% confidence intervals were used, and Spearman's correlation coefficients were applied to ordinal and quantitative variables. The relationship between qualitative data was statistically evaluated Using Chi-square tests, considering an a priori 0.05 probability significance level for closed questions with only one answer. Information about breed and dog size were divided into mixed-breed and pure-breed, and small (< 15Kg), medium (15-25 Kg), and large (> 25Kg), respectively. The software R (R Core Team, 2020) was used for statistical analysis. Questions with open answers or those that could indicate more than one option were evaluated descriptively.
Results
The study received 321 answers, of which 21 were excluded as the respondent was under 18 years old or did not agree with the study (n = 2); did not reside in the Rio de Janeiro state (n = 8); did not inform the city of residence for confirmation of the state (n = 1); and answered the questionnaire more than once (n = 5). Another five answers were removed after four owners indicated which dog they felt closest to. A total of 300 questionnaires were evaluated.
Demographic data
Of 300 respondents, 91% (273/300) lived in the metropolitan region of Rio de Janeiro state, where the ratio of respondents who lived in an apartment (54%, 162/300) or house (42%, 126/300) was similar. Three or more people per household was the most frequent answer and 14% (42/300) of households had children (Table 1 ). Regarding respondents, 85% (255/300) were women and 45% (135/300) were between 25 and 44 years old.Table 1 Relative frequency of the total number of inhabitants and children by type of home of the population under study.
Table 1Type of home Apartment (54%) House (42%) Others (4%)
Inhabitants/home 0 1 2 ≥3 0 1 2 ≥3 0 1 2 ≥3
Total (%) 13 31 56 2 22 76 9 27 64
Children (%) 86 14 0 0 84 16 0 0 100 0 0 0
A similar proportion of females (52%, 156/300) and males (48%, 144/300) was observed in the dogs evaluated, of which 69% (207/300) and 50% (150/300) were neutered, respectively. Age groups were also equally represented, except for under one year old. Data on breed, size, origin, and environment of the household in which these animals live are shown in Table 2 , 5% (15/300) of the respondents didn't answer about the breed, size and/or origin.Table 2 Relative frequency of size of pure-breed and mixed-breed dogs in relation to their origin and household environment.
Table 2Household environment Internal (48%) External (10%) Both (42%) Total
PURE-BREED (%)
HB R S K HB R A C HB R S K
Small 9 0 1 10 1 0 0 0 10 2 1 6 40
Medium 1 0 0 1 0.3 0 0 0.3 2 0 0 0.3 5
Large 1 0 0 0.3 2 0 0 0 1 0.3 0.3 1 6
MIXED-BREED (%)
HB R S K HB R S K HB R S K
Small 2 6 6 0 1 1 0.3 0 3 7 2 0 28
Medium 0.3 4 0.3 0.3 1 1 1 0.3 1 3 1 0.3 14
Large 0 0.3 0.3 0 0 0 0 0 1 0.3 0 0.3 2
Total 13 10 8 12 5 2 1 1 18 13 4 8
Origin of the dogs evaluated - homebred (HB); rescued from the street (R); shelter (S); or kennel (K).
Influence of environment on dog behavior during the COVID-19 pandemic
Type of home was related to hyperactivity (p = 0.01). All dogs that already had hyperactivity and remained hyperactive were living in apartments. On the other hand, the house dogs were the most prevalent in the group that began to exhibit hyperactivity, but these dogs comprised a minority (8.8%, 11/125) of those living in houses. The environment where the animal lived in (internal or external), the number of inhabitants in the household, and the presence of children did not influence the behavioral change of the dog population evaluated.
Characteristics and emotional predispositions of dogs related to behavioral changes
In the evaluation of factors intrinsic to the animal regarding behavioral changes, it was observed that intact (p <0.01) males (p <0.01) were more prevalent in the group of dogs that began to urinate in inappropriate places, and in the group where there was a change in the frequency of this behavioral problem (60%, 180/300). Females expressed “fear of people” more intensely (p = 0.01), when this condition previously existed (89%, 14/157). Adult and elderly dogs were the ones that most exhibited changes in urination behavior (p = 0.03) and defecation (p = 0.02) in inappropriate places. This increase in house soiling occurred in 64.5% (20/31) of these animals for urination and 70% (17/24) for defecation. Young and elderly dogs presented destructive behaviors (p <0.01) and hyperactivity (p = 0.01), respectively, more frequently than other groups. Approximately 75% of small-sized dogs, regardless of being of mixed-breed or pure-breed, showed changes (increase, decrease or begin) in hyperactivity (p = 0.03) or vocalization (p = 0.03). These two behaviors (p = 0.01 and 0.04, respectively) were less observed in animals with chronic diseases, however this group represented only 24.6% (74/300) of the population studied.
In the evaluation of emotional predispositions, the mean frequency of negative and positive affects was similar, but not significantly correlated with the behavioral changes of the dogs evaluated.
Investigation of changes in the routine and behavior of dogs after COVID-19 pandemic restrictions
With respect to changes in routine after the beginning of the COVID-19 pandemic (Table 3 ), dogs spent less time alone since in 46% (138/300) of homes there was always someone present and in 49% (147/300) of homes the time people spent at home increased. In addition, the human being who felt closest to the dog was usually always present. Most respondents were at home almost all of the time, 67% (201/300) of owners claimed to be staying only at home and 21% (63/300) leaving only to work, with reduced workload. Changes in people's routine did not correlate with any specific reported behavior.Table 3 Relative frequency of changes in routine presented by families and particularly by owners and their dogs during the COVID-19 pandemic in the population evaluated.
Table 3CHANGE IN ROUTINE (%) Family Time people spend at home
Decreased Unchanged Increased Constant
1 4 49 46
Owner Unchanged Goes out only to work Remains only at home
WL ↓ Normal WL WL ↑
5 2 5 21 67
Dog* Unchanged Walks Time of contact with
Ø ↓ ↑ Cohabiting humans Non-cohabiting humans Non-cohabiting animals
6 15 34 14 79 47 28
Workload (WL); decreased (↓); increased (↑); and suspended (Ø). *It was possible to select more than one alternative.
The most notable change in the routine of dogs (Table 3) was the intensified coexistence with people who lived in the same home. In the owner comments at the end of the questionnaire, the relationship between increased time spent together and the joy and affection of these individuals, whether human or dogs, was frequently identified. However, signs of anxiety including vocalization when alone, agitation to go for walks, not leaving the side of his owner and increased solicitation of affection and attention were the most prevalent reported behaviors. Owners reported the fear that both humans and dogs, but especially dogs, would suffer when the routine returned to normal. In addition, the opportunity for socialization and environmental enrichment decreased with more restricted walks and reduced contact with other animals and non-cohabiting people.
The mean response rate for questions about each behavior after the beginning of COVID-19 restrictions was 96% (288/300) (Table 4 ). The most frequent statistically significant response (Figure 1 ) for behavior studied was the lack of observation of the behavior. Just for excessive vocalization the most prevalent response was the previous existence without changes in frequency, although not differing statistically from lack of behavior observation.Table 4 Behavioral problems of dogs (n = 300) and changes that occurred during restrictions of the COVID-19 pandemic according to the perception of owners.
Table 4Categories of behavioral problems Behavior characteristics after COVID-19 (%)
Never observed Existed before isolation Developed
Improved Unchanged Worsened
Aggressiveness 60 5 19 9 3
Compulsive behavior 67 3 14 8 4
Destructive behavior 50 9 27 7 3
Defecation in inappropriate places 62 5 15 8 6
Hyperactivity 43 7 27 15 4
Fear of noises 47 4 32 10 2
Fear of other animals 68 3 19 4 1
Fear of people 70 3 15 5 1
Urination in inappropriate places 51 6 21 10 8
Excessive vocalization 32 5 40 16 4
Figure 1 Confidence intervals of frequencies of behavioral changes in dogs (n = 300) after the COVID-19 pandemic in Rio de Janeiro state.
Intervals that are not overlapped are significantly different (p <0.05). A) Already existed and remained equal; B) Already existed and the intensity increased; C) Already existed and the intensity decreased; D) Behavior developed during the COVID-19 pandemic; E) Not observed.
Figure 1
Out of 300 questionnaires, 260 (87%) showed a change in frequency (increased or decreased), or development of behaviors, for at least 1 behavior during the pandemic. Data analysis (Figure 2, Figure 3 ) showed that 74.2% (193/260) of dogs showed up to five altered behaviors and 10% (26/260) demonstrated some change for all behaviors questioned in the study. The most frequent behaviors that were altered (either increased or decreased) excessive vocalization (71.1%, 185/260), followed by hyperactivity (56.1%, 146/260), and fear of noises (53%, 138/260). The most frequent type of change was the worsening of pre-existing conditions, except for destructive behavior, which decreased in intensity, but was not statistically significant (Figure 3 - Confidence intervals, p<0.05). A statistically significant difference in the frequency of behavioral change was observed only for compulsive behavior, hyperactivity, fear of noise, and excessive vocalization.Figure 2 Quantification of altered behavior in dogs (n = 260) in Rio de Janeiro state.
Figure 2
Figure 3 Confidence intervals of frequencies of dogs with behavioral alterations caused by the COVID-19 pandemic (n = 260).
Intervals that are not overlapped are significantly different (p <0.05). B) Already existed and the intensity increased; C) Already existed and the intensity decreased; D) Behavior developed during the COVID-19 pandemic.
Figure 3
Effects on the physical health of pets and on the life quality of pets and their owners
When the development of diseases during the COVID-19 period was observed, approximately 32% (95/300) of respondents reported that their dog had some physical alteration. Of these, 47.3% (45/95) mentioned variation in weight, in which 37 dogs gained weight (38.9%), two had an increase in weight with subsequent decrease (2.1%), and six lost weight (6.3%).
The need to take the dog to the veterinarian's office was reported for 120 respondents (40%), showing a positive and statistically significant correlation (Chi-square tests, p < 0.001) with the presence of physical alterations reported by the owners. The primary reason for seeking veterinary care was that vaccines were due, and only one owner indicated weight-control follow-up. Among owners who mentioned behavioral problems, only an increase in libido was considered a reason for veterinary consultation.
An average 33% (99/300) of respondents answered each item. The most frequent response was being indifferent to the behaviors listed. However, for compulsive behavior, defecation or urination in inappropriate places, and excessive vocalization, no statistical difference was observed between being indifferent and having a negative effect on the owner's life (Figure 4 ). Of all respondents, 97% (291/300) considered the dog to be a companion and 66% (198/300) considered dogs as a way to relax. However, 64% (192/300) of respondents stated that the dog was a distraction and 10% (30/300) considered dogs as an extra job.Figure 4 Confidence intervals of frequencies regarding the repercussion of behavioral changes in dogs caused by the COVID-19 pandemic in the lives of owners.
Intervals that are not overlapped are significantly different (p <0.05).
Figure 4
Owners managed behavioral changes in their dogs by reprimanding them in 55% (165/300) and implementing measures of environmental enrichment in 53% (159/300) of the cases. Other owners sought help with an acquaintance (1%, 3/300), trainer (4%, 12/300), general practice veterinarian (8.3%, 26/300) or veterinarian specialized in animal behavior (1%, 3/300). Others also mentioned increasing walking time, daycare centers, internet research, and reinforcement of behavior, with more attention and affection.
Discussion
Concerns, difficulties, and stressors of pet owners during the COVID-19 pandemic were mostly related to their animals and were reported more often than difficulties related to the person and household (Applebaum et al., 2020). In the present study, the main concern reported by owners was the potential suffering of both humans and dogs when the time of coexistence decreases after the return to normal activities at the end of the pandemic. This response was also frequent in a questionnaire conducted in England (Ratschen et al., 2020), and was likely associated with the fear of the dogs developing separation anxiety syndrome (Applebaum et al., 2020; Hargrave, 2020), since some dogs already vocalized when being alone, as was also observed by Applebaum et al. (2020).
On the other hand, the need for constant attention from animals, caused by home confinement, is a limiting factor for work at home and other activities (Applebaum et al., 2020). Thus, the prevalence of responses in the questionnaire indicating signs of attention-seeking and anxiety in dogs may explain why more than half of respondents considered the dog to be a distraction. In the United States of America, this was the most highlighted stressor (Applebaum et al., 2020). In Spain, where the main behavior observed in the dogs was the demand for excessive attention, 5.8% of participants considered that the relationship with their dog worsened during isolation (Bowen et al., 2020). Therefore, intervention of a veterinarian is essential to prevent abandonment caused by behavioral problems (Hargrave, 2020), especially given the reported favorable scenario, in which the perception of owners in the population studied regarding their animals is mostly of being a companion and a way to relax. Dogs were also considered emotional supporters during restrictions and anxieties resulting from COVID-19 (Bowen et al., 2020; Ratschen et al., 2020). The window of opportunity to offer help and guidance is restricted since dog owners relinquish dogs with behavioral problems sooner than they relinquish cats with behavioral problems (Salmam et al., 2010).
Yet owners do not consider these behavioral changes to be a reason for veterinary consultation and often sought help from inadequate sources, leading to incorrect management and reinforcement of negative behaviors. This pattern may result in the worsening of behavioral changes over time (Bowen et al., 2020). In addition, punishment can also cause incidents of aggression against humans (Hargrave, 2020), which is a public health problem and a burden on the health system (Slater, 2001; Fatjó et al., 2006).
Altered behavior was not observed by the owners for most of the dogs of the population under study. When it was observed, it had worsened from a previous (pre-existing) condition. Our analysis was conducted in the short term (seven months of COVID-19 pandemic), which may have favored observation of the worsening of pre-existing alterations rather than the appearance of new alterations (Applebaum et al., 2020; Bowen et al., 2020).
Some changes may become evident only when the routine returns to normal (Applebaum et al., 2020; Bowen et al., 2020). However, as active search is necessary during anamnesis, the identification of the current context helps to identify the most frequent problems, as well as predisposing factors. The present survey found that compulsiveness, excessive vocalization, hyperactivity, and fear of noise were statistically significant changed in the population under study. These results corroborate the study conducted in Spain, in which excessive vocalization and fear of noise also worsened with the pandemic situation (Bowen et al., 2020). These changes are among the ten main reasons for abandonment (Salmam et al., 2010), suggesting that professional evaluation and intervention are warranted.
Among the changes observed, only compulsiveness and urination and defecation in inappropriate places were considered negative by the owners. These are common reasons to relinquish pets (Salmam et al., 2010). Bowen et al. reported that, regardless of type of behavioral change, the chance that animals exhibiting problem behaviors were considered a difficulty to the owner increased 1.9-fold (Bowen et al., 2020).
The main change in the routine of the dogs evaluated was the greater time spent in contact with people, which differs from other studies that showed the main change was a decrease in the time or frequency of walks (Applebaum et al., 2020; Bowen et al., 2020). Constant contact with humans was associated with greater joy for the owners and affection from their animals. On the other hand, this isolation with restricted contact with other people and animals may have masked the fear of people and animals, seeing that these two were the least frequently observed behaviors, with 30% (90/300) and 32% (96/300) of frequency, respectively. Combined with the reduction or absence of walks, these concerns may be impairing socialization and environmental enrichment, which may lead to future behavioral problems.
Hyperactivity was most observed in dogs living in apartments and was pre-existing. Dogs which began to exhibit this behavior mostly lived in houses. Possibly as COVID-19 restrictions prevent access to essential elements of well-being (Hargrave, 2020), leading to insufficient stimuli and environmental enrichments (Applebaumet al., 2020), the number of people at home and the presence of children did not interfere in the behavior of pets, corroborating the literature (Bowen et al., 2020). However, it was noticed that 86% (258/300) of the houses had no child, and that even with this scenario there was a report that a dog was aggressive against a child for not having its space respected. This was observed in another study, in which even without statistical significance, there was a report that children did not know how to deal with animals (Applebaum et al., 2020). Thus, veterinarians must be kind enough to instruct adult on how to manage this situation in order to avoid damage to any individual, such as accidents followed by abandonment.
Small dogs were the most predisposed to hyperactivity and excessive vocalization, probably for increased anxiety and fear, as owners of small dogs tend to have less time for consistent interaction, training, and games, associated with higher frequency of punishments (Arhant et al., 2010). In contrast, the hyperactivity and excessive vocalization were less observed in animals with chronic diseases, which are related to lower levels of activity (Fatjó and Bowen et al., 2020). Interestingly, walks also decreased excessive vocalization (Bowen et al., 2020), which highlights the importance of correct management with environmental enrichment for the treatment and prevention of behavioral problems.
Adult and elderly animals were the ones that most exhibited changes in the behavior of urinating and defecation in inappropriate places, although this behavior had been expected especially in puppies before the age of maturity and without training (Martínez et al., 2011). One explanation for this result is related to behavioral changes and cognitive decline during aging (Chapagain et al., 2020) associated with a reduced flexibility (Wallis et al., 2016) that predisposes them to emotional distress, caused by even a mild social challenge (Mongillo et al., 2013). These changes may be clinical signs of the cognitive dysfunction syndrome in aged dogs (Dewey et al., 2019). Stress can disturb dogs’ abilities to inhibit elimination (Chung et al., 2016), so social isolation may have been the stressor. More males may have marked associated with perceived stress (Martínez et al., 2011).
Elderly dogs and young adults experienced changes in destructive behavior and hyperactivity. These may be associated to an absence of daily dog social exposure (Chung et al., 2016), or to separation anxiety syndrome (Soares et al., 2009). Female dogs worsened with respect to pre-existing fear of people, and female dogs have been reported to have more phobias (Bamberger et al., 2006).
Access to the veterinarian is one of the concerns of owners during the COVID-19 pandemic (Applebaum et al., 2020; Bowen et al., 2020). Weight gain was the most reported physical condition and may have been associated with an increase in snacks and decreased physical activity (Applebaum et al., 2020). Obesity has been reported as a pandemic-related concern of owners (Bowen et al., 2020). However, like behavioral changes, this condition is not considered a reason for consultation.
Limitations and future research
Although Brazil presented non-standardization of restrictive measures against COVID-19 (Ortega and Orsini, 2020), when the change in the routine of families during the research period was evaluated, social distancing and isolation were observed in the population under study. In addition, social isolation was positively associated with the routine of owners, who reported being alone at home or leaving only to work with reduced workload.
Another issue is that 85% of the participants were women, who are 1.72-fold more likely to have the pet as an emotional support (Bowen et al., 2020). This population may not represent that as a whole, although this sample profile is frequent (Bowen et al., 2020; Ratschen et al., 2020), as women are more likely to answer online questionnaires (Smith, 2008). The elderly people, like men, were underrepresented, and the elderly may not have had access to the internet (Deursen and Helpsper, 2015). Future research needs to consider adaptations for the inclusion of these individuals, especially in the context of a pandemic in which distancing measures recommended for prevention require this methodology, as used in studies on COVID-19 (Applebaum et al., 2020; Bowen et al., 2020; Ratschen et al., 2020).
The results shown in this study reflect the perception of owners, not necessarily indicating the presence of behavioral disorders, which can only be diagnosed by veterinarians during consultation. However, these data are important for veterinarians to recognize behavioral changes and their effects on the quality of life of owners and their dogs. Finally, only domestic dogs were evaluated, which does not necessarily reflect behavioral changes and effects on the quality of life of other species or even of their owners. Therefore, even if dogs are currently suffering more with restrictions (Bowen et al., 2020), the recognition of behavioral changes and their effects on other species must be evaluated in future studies, especially as the relationship between man and animal does not depend on species (Ratschen et al., 2020).
Conclusion
Restrictions of the COVID-19 pandemic affected the behavior and emotions of dogs. Restrictions affected the quality of life of these animals and their owners by different degrees. The type of behavioral change depends on age, sex, dog size, type of home, and restrictions imposed, with the most prevalent changes being the worsening of pre-existing of compulsiveness, excessive vocalization, hyperactivity, and fear of noises. The data collected provide substantial information for the veterinarian to actively intervene in the behavioral and physical consequences of dogs resulting from the COVID-19 pandemic. The need for measures to prevent future abandonment and the development of aggressiveness, which are public health problems, is highlighted.
Uncited References
Fatjó and Bowen, 2020, McGreevy et al., 2018, Salvin et al., 2011, Zhang and Song, 2020, Brown et al., 2020, Cory, 2013
Authorship Statement
The idea for the paper was conceived by Luana de Sousa Ribeiro and Maria Cristina Nobre e Castro. The experiments were designed and analyzed by all authors. The experiments were performed, and the paper was written by Luana de Sousa Ribeiro.
Acknowledgements
Luana S. Ribeiro would like to thank the Ministry of Education for providing the scholarship for veterinary medicine residency, which allowed this research. The authors also declare no conflict of interest.
==== Refs
References
Applebaum J.W. Tomlinson C.A. Matijczak A. McDonald S.E. Zsembik B.A. The concerns, difficulties, and stressors of caring for pets during COVID-19: Results from a large survey of U.S. pet owners Animals 10 1882 2020 1 14 10.3390/ani10101882
Arhant C. Bubna-Littitz H. Bartels A. Futschik A. Troxler J. Behaviour of smaller and larger dogs: effects of training methods, inconsistency of owner behaviour and level of engagement in activities with the dog Appl. Anim. Behav. Sci. 123 2010 131 142 10.1016/j.applanim.2010.01.003
Bamberger M. Houpt K.A. Signalment factors, comorbidity, and trends in behavior diagnoses in dogs: 1,644 cases (1991–2001) J. Am. Vet. Med. Assoc. 229 10 2006 1591 1601 17107314
Bowen J. Garcia E. Darder P. Arguelles J. Fatjo J. The effects of the Spanish COVID-19 lockdown on people, their pets, and the human-animal bond J. Vet. Behav. 40 2020 75 91 10.1016/j.jveb.2020.05.013 32837452
Brown E.E. Kumar S. Rajji T.K. Pollock B.G. Mulsant B.H. Anticipating and mitigating the impact of the COVID-19 pandemic on Alzheimer's disease and related dementias Am. J. Geriatr. Psychiatry. 28 10 2020 712 721 10.1016/j.jagp.2020.04.010 32331845
Chapagain D. Wallis L.J. Range F. Affenzeller N. Serra J. Virányi1 Z. Behavioural and cognitive changes in aged pet dogs: no effects of an enriched diet and lifelong training PLoSONE 15 9 2020 1 32 10.1371/journal.pone.0238517
Chung T. Park C. Kwon Y. Yeon S. Prevalence of canine behavior problems related to dog-human relationship in South Korea J. Vet. Behav. 11 2016 26 30 10.1016/j.jveb.2015.10.003
Cory J. Identification and management of cognitive decline in companion animals and the comparisons with Alzheimer disease: A review J. Vet. Behav. 8 2013 291 301 10.1016/j.jveb.2012.08.001
Deursen A.J. Helpsper E.J. A nuanced understanding of Internet use and non-use among the elderly Eur. J. Commun. 30 2 2015 171 187 10.1177/0267323115578059
Dewey C.W. Davies E.S. Wakshlag J.J. Canine cognitive dysfunction: pathophysiology, diagnosis, and treatment Vet. Clin. North. Am.: Small Anim. Pract. 49 2019 477 499 10.1016/j.cvsm.2019.01.013 30846383
Fatjó J. Ruiz-De-La-Torre J.L. Manteca X. The epidemiology of behavioural problems in dogs and cats: a survey of veterinary practitioners Anim. Welf. 15 2006 19 185
Fatjó J. Bowen J. Making the case for multi-axis assessment of behavioural problems Animals 10 383 2020 1 17 10.3390/ani100303832020
Fiorillo A. Gorwood P. The consequences of the COVID-19 pandemic on mental health and implications for clinical practice Eur. Psychiatry. 63 1 2020 1 2 10.1192/j.eurpsy.2020.35
Hargrave C. COVID-19: implications of self‑isolation and social distancing for the emotional and behavioural health of dogs Companion Anim 25 4 2020 1 8 10.12968/coan.2020.0032
Holmes E.A. O'Connor R.C. Perry V.H. Tracey I. Wessely S. Arseneault L. Ballard C. Christensen H. Silver R.C. Everall I. Ford T. John A. Kabir T. King K. Madan I. Michie S. Przybylski A.K. Shafran R.. Sweeney A. Worthman C.M. Yardley L. Cowan K. Cope C. Hotopf M. Bullmore E. Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science Lancet Psychiatry 2020 1 14 10.1016/S2215-0366(20)30168-1 31860445
Martínez A.G. Pernas G.S. Casalta J.D. Rey M.L.S. Palomino L.F.C. Risk factors associated with behavioral problems in dogs J. Vet. Behav. 6 2011 225 231 10.1016/j.jveb.2011.01.006
McGreevy P.D. Wilson B. Starling M.J. Serpell J.A. Behavioural risks in male dogs with minimal lifetime exposure to gonadal hormones may complicate population-control benefits of desexing PLoS ONE 13 5 2018 1 14 10.1371/journal.pone.0196284 2018
Mongillo P. Pitteri E. Carnier P. Gabai G. Adamelli S. Marinelli S. Does the attachment system towards owners change in aged dogs? Physiol. Behav. 120 2013 64 69 10.1016/j.physbeh.2013.07.011 23911691
Ortega F. Orsini M. Governing COVID-19 without government in Brazil: ignorance, neoliberal authoritarianism, and the collapse of public health leadership Glob. Public Health. 15 9 2020 1257 1277 10.1080/17441692.2020.1795223 32663117
Ratschen E. Shoesmith E. Shahab L. Silva K. Kale D. Toner P. Reeve C. Mills D.S. Human-animal relationships and interactions during the COVID-19 lockdown phase in the UK: investigating links with mental health and loneliness PLoS ONE 15 9 2020 1 17 10.1371/journal.pone.0239397
Salman D. Hutchison J. Ruch-Gallie R. Kogan L. New J.C. Jr. Kass P.H. Scarlett J.M. Behavioral reasons for relinquishment of dogs and cats to 12 shelters J. Appl. Anim. Welf. Sci. 3 2 2010 93 106 10.1207/S15327604JAWS0302_2
Salvin H.E. McGreevy P.D. Sachdev P.S. Valenzuela M.J. The canine cognitive dysfuncion rating scale (CCDR): a data-driven and ecologically relevant assessment tool Vet. J. 188 2011 331 336 10.1016/j.tvjl.2010.05.014 20542455
Slater M.R. The role of veterinary epidemiology in the study of free-roaming dogs and cats Prev. Vet. Med. 48 2001 273 286 11259820
Smith G. Does gender influence online survey participation?: a record-linkage analysis of university faculty online survey response behavior 2008 ERIC Document Reproduction Service 1 21
Soares G.M. Souza-Dantas L.M. D'Almeida J.M. Paixão R.L. Epidemiology of dogs behavioral problems in Brazil: a survey between small animals veterinary practitioners Cienc. Rural. 40 4 2010 873 879
Vicent A. Mamzer H. Ng Z. Farkas K.J. People and their pets in the times of the COVID-19 pandemic Society Register 4 3 2020 111 128 10.14746/sr.2020.4.3.06
Wallis L.J. Virányi Z. Müller C.A. Aging effects on discrimination learning, logical reasoning and memory in pet dogs AGE 38 6 2016 1 18 10.1007/s11357-015-9866-x 26695510
Zhang Q. Song W. The challenges of the COVID-19 pandemic: approaches for the elderly and those with Alzheimer's disease MedComm 2020 1 5 10.1002/mco2.4
| 0 | PMC9741960 | NO-CC CODE | 2022-12-14 23:29:58 | no | J Vet Behav. 2022 Dec 11; doi: 10.1016/j.jveb.2022.12.001 | utf-8 | J Vet Behav | 2,022 | 10.1016/j.jveb.2022.12.001 | oa_other |
==== Front
Sci Afr
Sci Afr
Scientific African
2468-2276
The Author(s). Published by Elsevier B.V. on behalf of African Institute of Mathematical Sciences / Next Einstein Initiative.
S2468-2276(22)00406-9
10.1016/j.sciaf.2022.e01502
e01502
Article
Platform for hands-on remote labs based on the ESP32 and NOD-red
ABEKIRI Najib a⁎
RACHDY Azzedine a
AJAAMOUM Mohammed a
NASSIRI Boujemaa b
ELMAHNI Lahoussine a
OUBAIL Youssef a
a Ibn Zohr University Higher School of Technology, Agadir, Morocco
b Polytechnic School of Agadir, Morocco
⁎ Corresponding author.
11 12 2022
3 2023
11 12 2022
19 e01502e01502
21 6 2022
14 11 2022
10 12 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.
Not only in Morocco, throughout the walks of the world covid 19 pandemics has seriously questioned policymakers from different sectors. Think-tank in the educational sector notably higher education addressed by such a wide range of challenges brought about by covid 19. The characteristic concern that educationalists in Moroccan universities have to reconsider in this pandemic period should not be beyond rethinking new pedagogical alternatives including approaches, methods, techniques and didactic materials which can successfully assist practioners of the teaching and learning process to keep up with the current alterations. Practical work (PW) is an indispensable type of teaching in scientific and technical training and meets a real complementary need through real, remote or virtual laboratories. Students can consolidate what they have learnt and develop analytical skills by comparing experimental results with those obtained during the manipulation. In this context, the Laboratory of Engineering Sciences and Energy Management (LASIME) at the Superior School of Technology of Agadir has developed a low-cost platform called LABERSIME installed in the cloud (LMS, IDE) and equipped with an embedded system to drive real laboratory equipment and perform experiments qualitatively more efficient than those in face-to-face mode. The ultimate goal is to stimulate self-learning motivation in students through a creative approach.
Keywords
Remote laboratory
IoT
e-learning
Remote control
Remote monitoring
Open-source hardware
Editor: DR B Gyampoh
==== Body
pmcIntroduction and literature review
Higher education has always been at the heart of all debates on education throughout the world, and is at the centre of several problems and debates [1]: massification, democracy, etc. It is increasingly playing a role as an essential lever for development, especially in a knowledge-based economy based on research and innovation [2]. It has been transformed over the last two decades under the effect of economic and financial globalisation and also the arrival of new technological advances, and we are already beginning to wonder about its probable continuity without ICT. Indeed, over the last twenty-five years, many countries have developed mechanisms to guarantee the quality of university education based on new innovative technologies [3]
In this context, Moroccan universities are faced with several challenges and need to rethink teaching/learning methods to encourage greater student participation and promote innovative teaching practices. Given the growth in enrolment and the limited means of supervision, such an evolution leads to simultaneously take up pedagogical, technological and organisational challenges [4]. The Moroccan university must therefore succeed in facing up to the structural upheaval of the new modes of access, creation and dissemination of knowledge. It must think of following the train of such an evolution which is part of a partial transition towards the "all digital". The main objective of such actions is the development of new forms of active and collaborative learning based on nomadic and ubiquitous course tools [5].
In general, practical work (PW) is considered an essential part of teaching and learning science, this position has been confirmed by international researchers, teachers and curricula [6]. Furthermore, it is considered that students generally seem to enjoy practical work and this increases their motivation to study science [6]. Indeed, many researchers have found that science learning and understanding is enhanced when students are engaged in hands-on experiments in a science laboratory. In this sense, the sociologist Edgar Morin (1984) has even stated that "Science is a mode of knowledge based on the dialogue between theories and observed or experimental data". The results of science are therefore not to be taught as data to be believed but as results produced by experimental methods. Experimental know-how can only be acquired by students if they have manipulated during practical work sessions.
In science faculties, practical work should be carried out in small groups of students, to allow the theory learned in class to be put into practice through experiments [7]. Practical work stimulates students' curiosity by allowing them to observe and ask questions. They also help to develop a spirit of initiative and above all a critical mind when it comes to analysing and interpreting results. Indeed, the experimental approach helps, on the one hand, to master the concepts that manage the functioning of a device and, on the other hand, to articulate the experimental practices in order to achieve the appropriation of knowledge qualified as theoretical. However, the sessions of the TP in the Moroccan establishments are impacted by the massification. To the problem of the increase in the number of enrolled students is added the insufficiency of scientific equipment: materials that must be monitored throughout the experiments to avoid any damage. This situation has created some management complexities for both teachers and students. At the level of the teachers, the supervision rate increases as the number of enrolled students increases, we are talking about a number that varies from three to four trinomials per teacher, given the volume of time per session reserved for practical work, the time distribution is hardly sufficient to cover and achieve all the objectives with efficiency.
Faced with these imperatives, students are now required to deepen their understanding of theoretical concepts and to acquire technical know-how: to master the theoretical aspects related to the manipulation, to handle the material correctly, to read and interpret the results and finally to extract constructive conclusions. Achieving these various learning and behavioural skills cannot be achieved entirely and solely during the sessions reserved for practical work under the conditions mentioned above. Furthermore, the growing interest in the use of technology in higher education cannot be ignored, as today's students think and process information in a fundamentally different way from their predecessors.
Science and technology are based on the application of theoretical knowledge through the manipulation of objects and instruments. Practical work (P.W.) requires a special time and place. Due to the high enrolment in higher education and limited resources (number of rooms, equipment and human resources) [8], the acquisition of knowledge in the face-to-face mode is rather delicate. However, new strategies have been developed to motivate students to learn electronics and electrical engineering as one cannot limit oneself to classical classroom-based methods. In addition, the open-source hardware movement has gained popularity with the emergence of low-cost, high-performance technologies such as Arduino and Raspberry Pi, due to the community of manufacturers who actively share their creations for study, modification and application in educational projects [9].
In order to face these challenges, different remote laboratories have been developed for several years. Among these labs, we can distinguish between remote labs and purely virtual labs. And as we know, setting up a remote lab is more expensive and complex than a virtual lab [10] but on the other hand and pedagogically speaking, the added value in learning via a remote lab is more beneficial than the virtual lab [11] because the latter is only a computer model of approximations of reality via modelling to do simulations.
Remote laboratories offer advantages, especially in terms of reduced investment costs for equipment and instruments. Remote laboratories offer a number of advantages [12], not least in terms of reduced investment costs for equipment and instruments. One equipment can be used by many students and maintenance costs can also be shared if the equipment is used by different universities. Students who work and study part-time may also find remote labs useful in balancing their commitments. Using a smartphone or tablet, the student can continue learning from the classroom at home with flexible scheduling [13].
In recent years, remote laboratories have been deployed [14,15] as remote learning solutions. For example, the European Go-Lab initiative aims to enhance the classroom experience by using inquiry-based learning [3] through physical and virtual online manipulations [16] in universities.- Laborem: developed at the University Institute of Bayonne, a platform installed in localhost via a Raspberry Pi board used as a server containing an open-source software (Moodle LMS) and an electronic board to ensure communication between the server and the equipment used in the practical work. LPyScada makes the control and data acquisition [17].
- IoT System for Remote Practical Works: the IoT system for remote practical works is mainly based on the Red Pitaya card. The remote communication is done by the RIP protocol for the control and acquisition of measured data, with the use of an electronic card so the teachers can choose the practical works for students [18].
- IoT4SMEs: based on MATLAB, integrated via ViSH is a social and collaborative platform based on creating and sharing virtual learning resources [19].
ThingSpeak, a publicly accessible cloud-based data collector. This open-source application provides various services such as storage and retrieval of data from IoT sources via HTTP Protocol, and monitoring and analysis of data based on MATLAB numerical computation software [20].
- RemotElecLab: This is a newer remote lab platform for students to experiment with electrical and electronic circuits. This solution was created after studying the shortcomings of current remote lab solutions for similar tests in order to improve them, especially by using generic equipment. It is accessible through a generic interface in dependable on the circuit under test [21].
- Another example is the LabsLand1 network, a commercial venture that aspires to construct a broad worldwide network of remote laboratories from various supplying institutions, allowing schools and universities to access them via a single and trustworthy platform [22,23]. These endeavours can build upon the numerous breakthroughs in the literature about remote laboratories. Today, remote laboratories may be entirely web-based, making them accessible to the majority of users [24]
- Similarly, iSES2 has established its own online teaching system [25] that contains a number of remote laboratories with particularly specialized physics applications. Among these are the ability to validate the Heisenberg uncertainty principle, perform the Franck-Hertz experiment, and do microcontroller experiments [26]
- The GOLDi3 provides access to a remote laboratory with multiple setup choices. Here, different control units (a CPLD, an FPGA, a microcontroller, an FSM, etc.) and physical systems can be selected. This provides the user with a great level of flexibility in executing various experiments [27]
- The University of Lisbon has developed e-lab, its own remote laboratory initiative. It is an online platform that allows access to remote physics and chemistry laboratories. This laboratory allows working with AC and DC electrical panels, as well as engineering laboratories for programming microcontrollers using Arduino uno boards [28].
As described above, remote labs can guarantee the same pedagogical objectives as classical labs [29,16] as long as they maintain a certain quality of service, but when remote labs are down, have connection problems or are not able to support user management, it is not possible to ensure an adequate quality of service and a better user experience, which may lead users to abandon remote labs in favour of other solutions such as virtual labs or classical practical labs,
In addition, the implementation of the above-mentioned projects is complicated for teachers and requires the mastery of the technology and the architecture adopted so that they can use it and postulate new manipulations.
To solve this problem, the Laboratory of Engineering Sciences and Energy Management (LASIME) at the Superior School of Technology of Agadir has developed a low-cost platform called LABERSIME that is installed in the cloud (LMS, IDE) as a complementary solution with a simple architecture (experimentation based on the ESP32 microcontroller and other electronic components) designed to reduce costs and share hardware components in an optimal way with the platform users. The ultimate objective is to encourage students' self-learning motivation through a creative method.
To this end, the contributions of this work are:1) The description of a set of remote laboratories created as part of the set of technological solutions previously presented above.
2) The description of a remote laboratory, which is based on the IOT architecture, first tested in the practical work of photovoltaic systems for the students of the first year of electrical engineering at the Agadir Higher School of Technology.
3) Conclusions, based on the previous quantitative analysis, taking into account mainly the availability and quality of service provided by the laboratory from the user's point of view.
This article is organised as follows: Sect.2 presents materials and methods of the remote laboratory solution. Then Sect.3 describes the experimentation and implementation of the Platform. Finally, Sect.4 presents the main results summarised and the conclusion.
Materials and methods
We propose in this context, the design and the realization of a platform for the remote practical works for the students of electrical engineering. Our proposal consists in reinforcing the preliminary preparation of the manipulations outside the face-to-face sessions. The evaluation-test of our approach is carried out initially for the practical work of "characteristics of the photovoltaic cells" for the students registered in first year of electrical engineering at the higher school of technology of Agadir, the objectives of this self-training are the following three points:• Knowledge and prerequisites: students are invited to review basic scientific concepts, definitions, units, models.
• Working methodology: it concerns the reasoned use of laws and formulas, calculation of uncertainties, exploitation of curves, reasoning methods.
• Experimental methodology: it targets practical know-how concerning the use of measuring devices, the graphic representation of measurements.
The platform consists of a human-machine interface via LMS that allows students and teachers to access the laboratories remotely via the MQTT protocol [30], a motherboard based on an ESP32 microcontroller, connected instruments and input/output devices to manipulate the targeted systems. Users of the platform can control the equipment through pre-actuators that distribute the energy required for their operation. The feedback chain ensures that the commands issued are executed or not in an interactive process. In addition, electronic components (ADC's) have been integrated to allow the acquisition of data from different sensors.
The software and hardware used are open-source, and have been developed to meet the growing demand from electronics developers. Students and teachers find it easy to use, adaptable to different scenarios depending on the description of the manipulation.
The architecture of LABERSIME is presented in Fig. 1 . A cloud-based VPS is used as a server containing the LMS Chamilo and Node-RED, the motherboard has different input/output ports to connect all the necessary devices (components, measuring devices, power supplies, oscilloscopes).Fig. 1 Platform architecture.
Fig. 1
Hardware parts
For the hardware we have chosen components and equipment that are easy to integrate in the systems as shown below:
As shown in the figure above, hardware composes of the following: • ESP32 Microcontroller:
Fig. 2 Schematic view of the hardware components connected to the microcontroller.
Fig. 2
ESP 32, which is a low-cost microcontroller that replaces the previous version ESP8266, is at the heart of the architecture of hardware. Along with integrated Wi-Fi, ESP 32 offers a variety of functionalities compatible with a wide range of sensors for a different use, including IoT and streaming its specification. It as well offers programmers a powerful toolkit [31] with dual cores,240 MHz, 520 Kbyte SRAM and peripherals including: I2C, DAC, ADC, I2S, SPI, UART, 34 physical GPIO pins (Fig. 3 ). The power management unit and the lower power controller enable the ESP32 to run lower than 1mA in deep sleep mode. These advantages make the ESP32 a better device for low power applications.• ESP32-CAM
Fig. 3 ESP32 board.
Fig. 3
The ESP32-CAM is a low-cost card that includes an ESP32 microcontroller and an OV2640 camera. One of the most interesting features of the ESP32 is the ability to communicate by Wi-Fi, an obvious use of the ESP-32 CAM is to transmit live video images via Wi-Fi (surveillance camera, etc.). In addition to an OV2640 camera, the module is equipped with a micro-SD card reader that can be used to store images or video sequences [32].• INA219
INA219 is a High Side DC Current Sensor Breakout circuit for measuring the power consumption The INA 219 can measure DC current up to 26V / 3.2A. It is equipped with an I2C bus, which makes it very easy to retrieve measurements using an MCU (Arduino, ESP8266, ESP32) or a Raspberry Pi [33].• MOSFET D4184
The component can be used in applications such as controlling the output of DC electrical equipment, DC motors, light bulbs, LED lights, DC motors, micro-pumps, solenoid valves, etc. The input voltage (VIN) is 5V to 36V DC and the rated output current (ILoad) is 15A with a maximum power of 400W. The trigger control signal is 3.3 to 20V DC and the standard operating frequency is 20Khz [34].• INSTRUSTAR ISDS210A Oscilloscope:
An oscilloscope is used to visualize electrical signals during manipulations [35]. The device connected to our platform is the digital model attached to the USB port (INSTRUSTAR ISDS210A) which stores the received signals and displays them on a screen for consulting and studying them remotely via the RDP protocol.
Software Parts
Built on a Node.js framwork, Node-RED is a stream-based programming interface. Input and output nodes allow scenarios of any kind to be built graphically, providing a JavaScript function editor for further customisation. As illustrated in Fig. 4 , the Node-RED transforms messages acquired from UC ESP32 into JSON format via MQTT Protocol, Mosquitto acts as a broker that runs with the Node-RED on the Cloud.• Chamilo LMS
Fig. 4 Software Architecture.
Fig. 4
Chamilo is platform known for its ease to use (3 times less learning time than Moodle) [36] and its light weigh. It is also one of the three most popular open-source educational platforms globally. Chamilo is an intuitive and comprehensive platform for managing courses and learning content. Each course has a minimum of 24 tools to choose from, which you can select with a mouse click. This educational platform meets the needs of all types of structures (both public and private), from the autonomous user to large companies. Many features are available, such as Course tools (courses, courses, announcements, class diary, attendance sheets, synchronous and asynchronous communication, glossary, survey, forum, documents, groups, chat, wiki), and numerous monitoring tools (reports, statistics).• Influxdata
The choice of the right type of database is crucial to managing IoT data efficiently [37]. For this purpose, we used InfluxDB, which is known for its reliability and is generally used for real-time applications.• Micopython
MicroPython is a Python interpreter optimised for microcontrollers such as ESP32 boards with thread support. Script [38] can be upload directly on ESP32 board. By simply flash the microcontroller board with firmware available on micropython.org. An IDE software (e.g. Thonny IDE which embeds the latest version of Python) [39] with built-in communication to interact with REPL.
The algorithm for controlling actuators and the acquisition of the sensors is shown in Fig. 5 , the flowcharts illustrate the operations executed at the Cloud, from the initialization of the server until the end of the processing of the received data and the control of the pre-actuators (Fig. 6 ).Fig. 5 Algorithm for sensor acquisition.
Fig. 5
Fig. 6 Algorithm for actuator control.
Fig. 6
Experimentation and implementation of the IoT platform
By and large, to get students exposed to experiments in the platform, professors are required to consider the following phases respectively. First of all, the availability of not only adequate but also motivating and resourceful activities should be within reach. So doing would not take place beyond lecturers’ engagement via preparing beforehand so that learners can be familiar with new forms of learning. If successfully learners get accessed to suggested materials in advance, they can have sufficient time to work at ease and get prepared to deal with proposed activities. The social network function provided by Chamilo comes as supporting equivalent LMS that provides learners to exchange and interact with each other via discussions, debates and forums. However, having effectively created such kind of experiments is conditioned by having an administrator role in LABERSIME platform; not all professors can do the task if not subjected to such a role. With so they can modify existing resources or create new ones, the administrator can also consult the list of users or add a new group of students.
The pedagogical approach-based
The process of teaching and learning does only rely on competence- input- but also performance- output-; the two necessitate an effective method which cannot be segmented from technique as well. Within a huge number of approaches, I thus opt for applying the so-called ludic approach, it best works to improve, immerse and motivate students while learning process [40] The choice is not out of the ashes, rather it keeps learners within the process and then become into the centre of the teaching and learning. Learning in an atmosphere which makes learners feel their sense of being human is thus given by ludic method, as it leaves them learn at ease and additionally enjoy. Like other approaches, ludic method suggests the following stages:- Pre-requisites: learners are asked about their prior knowledge.
- New theories: learners get exposed to new input.
- Quizzes: learners sit for quizzes as forms of assessment.
- Remote experiments: learners get engaged in practical works.
- Final exam: learners are in the process of evaluation; be it pertaining to one module.
In each activity time allocated should be respected, as the number of allowed attempts should not exceeded. An adopted scenario is described in Fig. 7 below:Fig. 7 Pedagogical scenario describes the learning sequences via LABERSIME.
Fig. 7
Interaction interface
Fig. 8 shows the client interface screen when exposed to a sequence (Remote Laboratory).Fig. 8 Client interface (LMS + Interface for interaction with the remote laboratory).
Fig. 8
The menu describes the learning scenario, the student may or may not be allowed to access to a course. These conditions depend on the student's results and course choices.
The central area is for manipulation and access to the results of the experiment. In this dynamic area, the user can interact with the remote laboratory by getting position in the queue, setting the parameters of the experiment, and view the equipment located in the laboratory, participate in the PW by making measurements and export graphs to enrich report of his experiment. subscription to a topic (sensor/MQTT/publish) [41] on the MQTT broker is done by Node-RED to publishes data received from ESP32. We design the flow to manage and process the device's sensor data. Example of the split method to separates a received string into an ordered list of substrings, each string represents a particular value of a sensor.
A flowing design for handling INA219 data is shown in Fig. 9 . First, a subscription node subscribes to the topic (MQTT/publish). This node will get data from the MQTT broker whenever the device publishes it from the INA219 Breakout [42]. Then the acquired data will be displayed on the gauge and graph node. Finally, a dashboard will be shown as a user interface and present the physical quantities measured (voltage, current, power).Fig. 9 Node-RED's Overall Flow Management Design.
Fig. 9
The same principle is used to control the pin status of the ESP32, so relays can be activated and deactivated to control equipment or to ensure connectivity between electronic components.
Results and discussion
This work's outcome is provided in two parts. The first stage is to determine the final user interface as well as the technology that will be utilised to undertake the distant practical work. The platform is then evaluated in the second stage, which involves remote manipulation a photovoltaic panel (Fig. 10 ) and compare the same practical manipulations that are performed in the classical system under the same conditions.Fig. 10 Remote hands-on experience of photovoltaic cell characterization.
Fig. 10
The user interface
To test the platform, we proposed two practical works on operational amplifiers and photovoltaic panels for the students in the first year of electrical engineering. The first practical work aims to study some assemblies in a non-linear regime. The objective of the second one is to plot the current-voltage characteristic of a photovoltaic cell, and determine the characteristic quantities of short-circuit current ICC, open-circuit voltage VCO and maximum power, and calculate the efficiency of the photovoltaic panel.
After completing the required lab exercises, and work has successfully been validated, students can perform commands, record measurements, plots, and integrate results in their reports. The last developed graphical interface is shown in Fig. 11 .Fig. 11 Dashboard and control panel for remote photovoltaic cell characterization hand-on.
Fig. 11
In addition, the LMS allows tracking of each student: number of connections, time spent, number of attempts, progress, level of difficulty taken, and answers to tests.
Most connections take place the day before; the deadline for accessing the remote laboratory. To this end, it requires a solid platform to manage multiple connections.
Student follow-up and satisfaction survey
In order to demonstrate the usefulness of the remote laboratory, from a technical and pedagogical point of view, we tested the platform with 40 students of the first year of electrical engineering during photovoltaic cell characterisation experiments.
The students who completed the previous experiment filled in questionnaires that allowed us to identify important indicators: remote control of real physical systems, space-time constraints, collaborative work, autonomy, online activity and prerequisites. All these indicators allowed us to evaluate the efficiency of the platform and to improve it for future activities so that the manipulations can be more motivating and significant.
Likert scales were used for questions concerning the platform [43], its usefulness, its functionality, the time it takes, and whether students would use it again or not.
The analysis of the data in Fig. 12 reveals some essential points. According to the responses to the survey, most students favor choosing between the first and second options to answer each question in both practical systems. The results of the practical lab are always higher than those of the remote lab, which leads to two conclusions: firstly, the practical lab is the most favored by students; secondly, the slight difference in the choice of the remote lab and the practical lab shows that the platform has succeeded in transmitting the same knowledge at a distance and that it can be adapted to unforeseen cases, such as the case of covid'19, to avoid interruptions in the students' practical learning. From the two questions presented in Fig. 12, it is clear that practical distance work saves time for students.Fig. 12 Data from the online surveys of students on the LIKERT scale.
Fig. 12
Conclusion
Based on experimental survey undertaken, The LABERSIME board is mainly used for remote communication and control using the MQTT protocol for measurements and data acquisition.
The implementation of this remote experimentation has helped to increase students’ motivation. If this positive point was noted thanks to our monitoring, we are not claiming that the best solution for practical work is an entirely remote solution. Indeed, the LABERSIME platform comes as a complement to the practical work done in classes. LABERSIME adds possible functionalities since the evolution of ICT that allows a differentiated pedagogy more focused on each student. Moreover, this prototype is used as a demonstrator during the open days of the IUT of Agadir and seems to be of a great significance to high school students as well. This contributes to an increase in the attractiveness of our STEM courses.
In brief if any objective conclusion can be drawn it as follows: our attempt cannot be considered neither perfect nor final. It is a rather a starting point which paves the way for us to think of developing it in the forthcoming works. the ultimate goals are in a nutshell to generalize it by and large in wider scale on the one hand, and on the other hand to stimulate professors to get low-cost device into use and finally make the manipulations easy to deploy via a simple architecture.
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 Khan M.A. COVID-19’s Impact on higher education: a rapid review of early reactive literature Educ. Sci. 11 2021 421 10.3390/educsci11080421
2 S. Ouahib, K. El Kharki, R. Bendaoud, D. Burgos, K. Berrada, Open educational resources as a global solution for wider class courses, in: 2022: pp. 31–48. doi:10.1007/978-981-19-5137-4_4.
3 Khan P.A. Johl S.K. Akhtar S. Asif M. Salameh A.A. Kanesan T. Open innovation of institutional investors and higher education system in creating open approach for SDG-4 quality education: a conceptual review J. Open Innov. Technol. Mark. Complexity 8 2022 49 10.3390/joitmc8010049
4 H. Akhasbi, N. Belghini, B. Riyami, SELF-DETERMINED LEARNER MOTIVATION EVALUATION IN DISTANCE LEARNING CONTEXT USING WEB 2.0 TOOLS, DURING THE COVID-19 PANDEMIC PERIOD, in: 2021: pp. 2143–2151. doi:10.21125/inted.2021.0466.
5 H. Akhasbi, N. Belghini, B. Riyami, O. Cherrak, H. Bouassam, Moroccan higher education at confinement and post confinement period: review on the experience, in: 2022: pp. 130–164. doi:10.1007/978-3-031-14756-2_8.
6 Juuti K. Lavonen J. Salonen V. Salmela-Aro K. Schneider B. Krajcik J. A teacher–researcher partnership for professional learning: Co-Designing project-based learning units to increase student engagement in science classes J. Sci. Teacher Educ. 32 2021 625 641 10.1080/1046560X.2021.1872207
7 Haug B.S. Mork S.M. Taking 21st century skills from vision to classroom: what teachers highlight as supportive professional development in the light of new demands from educational reforms Teach. Teach. Educ. 100 2021 103286 10.1016/j.tate.2021.103286
8 Pilotti M.A.E. El-Moussa O.J. Abdelsalam H.M. Measuring the impact of the pandemic on female and male students’ learning in a society in transition: a must for sustainable education Sustainability 14 2022 3148 10.3390/su14063148
9 Fernandez-Pacheco A. Martin S. Castro M. Implementation of an arduino remote laboratory with raspberry Pi 2019 IEEE Global Engineering Education Conference (EDUCON) 2019 IEEE 1415 1418 10.1109/EDUCON.2019.8725030
10 Kalúz M. Čirka Ľ. Valo R. Fikar M. ArPi Lab: a low-cost remote laboratory for control education IFAC Proc. Vol. 47 2014 9057 9062 10.3182/20140824-6-ZA-1003.00963
11 Nedic Z. Machotka J. Nafalski A. Remote laboratories versus virtual and real laboratories 33rd Annual Frontiers in Education 2003 IEEE 10.1109/FIE.2003.1263343 n.d.: p. T3E_1-T3E_6
12 Oellermann M. Jolles J.W. Ortiz D. Seabra R. Wenzel T. Wilson H. Tanner R.L. Open hardware in science: the benefits of open electronics Integr. Comp. Biol. 62 2022 1061 1075 10.1093/icb/icac043 35595471
13 Mathrani A. Sarvesh T. Umer R. Digital divide framework: online learning in developing countries during the COVID-19 lockdown Globalisation Soc. Educ. 20 2022 625 640 10.1080/14767724.2021.1981253
14 v. Nickerson J. Corter J.E. Esche S.K. Chassapis C. A model for evaluating the effectiveness of remote engineering laboratories and simulations in education Comput. Educ. 49 2007 708 725 10.1016/j.compedu.2005.11.019
15 Valentin L.-P. Michel Christine Serna Adurey Tabard Aurélien Challenges and opportunities for multi-device management in classrooms ACM Trans. Computer-Hum. Interact. 2022 10.1145/3519025
16 Honig C.D. Sutton C.C. Bacal D.M. Off-campus but hands-on: mail out practicals with synchronous online activities during COVID-19 Educ. Chem. Eng. 39 2022 84 93 10.1016/j.ece.2022.03.004
17 Lavayssière C. Larroque B. Luthon F. Laborem Box: a scalable and open source platform to design remote lab experiments in electronics HardwareX 11 2022 e00301 10.1016/j.ohx.2022.e00301 35509900
18 Moulay Taj A. Chacon Sombria J. Gaga A. Abouhilal A. Malaoui A. Conception and implementation of an IoT system for remote practical works in open access university’s electronic laboratories Int. J. Online Biomed. Eng. (IJOE) 17 2021 19 10.3991/ijoe.v17i02.19755
19 Leisenberg M. Stepponat M. Internet of things remote labs: experiences with data analysis experiments for students education 2019 IEEE Global Engineering Education Conference (EDUCON) 2019 IEEE 22 27 10.1109/EDUCON.2019.8725070
20 N. Penchalaiah, J.N. Emmanuel, S.S. Kamal, C.V. Lakshmi Narayana, IoT based smart farming using thingspeak and MATLAB, in: 2021: pp. 1273–1295. doi:10.1007/978-981-15-7961-5_117.
21 Slamnik-Kriještorac N. de Resende H.C.C. Marquez-Barja J.M. Practical teaching of distributed systems Proceedings of the 6th EAI International Conference on Smart Objects and Technologies for Social Good 2020 ACM New York, NY, USA 120 125 10.1145/3411170.3411230
22 Sanusi I.T. Olaleye S.A. Agbo F.J. Chiu T.K.F. The role of learners’ competencies in artificial intelligence education Comput. Educ. Artif. Intell. 3 2022 100098 10.1016/j.caeai.2022.100098
23 R. Ryu, S. Yeom, D. Herbert, J. Dermoudy, An adaptive biometric authentication system for online learning environments across multiple devices, in: 2022: pp. 375–378. doi:10.1007/978-3-031-11647-6_73.
24 Xie C. Li C. Sung S. Jiang R. Engaging students in distance learning of science with remote labs 2.0 IEEE Trans. Learn. Technol. 15 2022 15 31 10.1109/TLT.2022.3153005
25 Ozer T. Agir I. Henry C.S. Low-cost internet of things (IoT)-enabled a wireless wearable device for detecting potassium ions at the point of care Sens. Actuators B Chem. 365 2022 131961 10.1016/j.snb.2022.131961
26 Yang H. Tong J.P. Xu D.Y. Sui C.H. Design of universal experimental interface based on STM32 Adv. Mat. Res. 562–564 2012 1615 1618 10.4028/www.scientific.net/AMR.562-564.1615
27 K. Henke, J. Nau, R.N. Bock, H.-D. Wuttke, A hybrid online laboratory for basic STEM education, in: 2021: pp. 29–39. doi:10.1007/978-981-16-2834-4_3.
28 Cardoso R. Santos M. Fernandes H. Negirneac V. Santos A.M. A MOOC with an e-lab remotely controlled experiment: student behaviour analysis 2019 5th Experiment International Conference (Exp.at’19) 2019 IEEE 52 57 10.1109/EXPAT.2019.8876472
29 Marks B. Thomas J. Adoption of virtual reality technology in higher education: an evaluation of five teaching semesters in a purpose-designed laboratory Educ. Inf. Technol. (Dordr) 27 2022 1287 1305 10.1007/s10639-021-10653-6 34257511
30 A. Shehab, A. Ismail, L. Osman, M. Elhoseny, I.M. El-Henawy, Quantified self using IoT wearable devices, in: 2018: pp. 820–831. doi:10.1007/978-3-319-64861-3_77.
31 Babiuch M. Foltynek P. Smutny P. Using the ESP32 microcontroller for data processing 2019 20th International Carpathian Control Conference (ICCC) 2019 IEEE 1 6 10.1109/CarpathianCC.2019.8765944
32 Kanakaraja P. Syam Sundar P. Vaishnavi N. Gopal Krishna Reddy S. Sai Manikanta G. IoT enabled advanced forest fire detecting and monitoring on Ubidots platform Mater. Today Proc. 46 2021 3907 3914 10.1016/j.matpr.2021.02.343
33 Ji Q. Wang X.V. Wang L. Feng L. Online reinforcement learning for the shape morphing adaptive control of 4D printed shape memory polymer Control Eng. Pract. 126 2022 105257 10.1016/j.conengprac.2022.105257
34 Nahrendra I.M.A. Adiprawita W. Susanto A. Hadinata S.K. Engine control unit for 2 stroke internal combustion engine on medium altitude unmanned aerial vehicle 2018 International Symposium on Electronics and Smart Devices (ISESD) 2018 IEEE 1 6 10.1109/ISESD.2018.8605492
35 Aung S.Thu Marakhtanov M.K. Khokhlov Y.A. Low-frequency plasma oscillations in transient conditions of an accelerator with an anode layer J Phys Conf Ser 2021 IOP Publishing Ltd 012033 10.1088/1742-6596/1799/1/012033
36 Astriawati F. Djukri Developing chamilo-based E-learning in environmental change material to enhance students’ scientific literacy skills J Phys Conf Ser Vo. 1397 2019 012049 10.1088/1742-6596/1397/1/012049
37 Bromig L. Leiter D. Mardale A.-V. von den Eichen N. Bieringer E. Weuster-Botz D. The SiLA 2 manager for rapid device integration and workflow automation SoftwareX 17 2022 100991 10.1016/j.softx.2022.100991
38 Wang L. Li Y. Zhang H. Han Q. Chen L. An efficient control-flow based obfuscator for micropython bytecode 2021 7th International Symposium on System and Software Reliability (ISSSR) 2021 IEEE 54 63 10.1109/ISSSR53171.2021.00028
39 Keesey R. LeSuer R. Schrier J. Sidekick: a low-cost open-source 3D-printed liquid dispensing robot HardwareX 12 2022 e00319 10.1016/j.ohx.2022.e00319 35677813
40 Letowski B. Lavayssière C. Larroque B. Schröder M. Luthon F. A fully open source remote laboratory for practical learning Electronics (Basel) 9 2020 1832 10.3390/electronics9111832
41 Ihirwe F. Indamutsa A. di Ruscio D. Mazzini S. Pierantonio A. Cloud-based modeling in IoT domain: a survey, open challenges and opportunities 2021 ACM/IEEE International Conference on Model Driven Engineering Languages and Systems Companion (MODELS-C) 2021 IEEE 73 82 10.1109/MODELS-C53483.2021.00018
42 Ramadhan J.M. Mardiati R. Haq I.N. IoT monitoring system for solar power plant based on MQTT publisher/subscriber protocol 2021 7th International Conference on Wireless and Telematics (ICWT) 2021 IEEE 1 6 10.1109/ICWT52862.2021.9678503
43 Murano D. Lipnevich A.A. Walton K.E. Burrus J. Way J.D. Anguiano-Carrasco C. Measuring social and emotional skills in elementary students: development of self-report Likert, situational judgment test, and forced choice items Pers. Individ. Dif. 169 2021 110012 10.1016/j.paid.2020.110012
| 0 | PMC9741961 | NO-CC CODE | 2022-12-16 23:18:10 | no | Sci Afr. 2023 Mar 11; 19:e01502 | utf-8 | Sci Afr | 2,022 | 10.1016/j.sciaf.2022.e01502 | oa_other |
==== Front
Child Abuse Negl
Child Abuse Negl
Child Abuse & Neglect
0145-2134
1873-7757
Elsevier Ltd.
S0145-2134(21)00512-3
10.1016/j.chiabu.2021.105443
105443
Article
Impact of COVID-19 pandemic on child abuse and neglect: A cross-sectional study in a French Child Advocacy Center
Massiot L. a
Launay E. bc⁎
Fleury J. ab
Poullaouec C. ab
Lemesle M. a
Guen C. Gras-le bc
Vabres N. ab
a Child Advocacy Center, Pediatric Department, Nantes University Hospital, 7 quai Moncousu, 44093 Nantes, France
b Pediatric Department, Nantes University Hospital, 7 quai Moncousu, 44093 Nantes, France
c Clinical Investigation Center 0004, Nantes University Hospital, 7 quai Moncousu, 44093 Nantes, France
⁎ Corresponding author at: Service de pédiatrie générale, 7 quai moncousu, 44000 Nantes, France.
13 12 2021
8 2022
13 12 2021
130 105443105443
24 8 2021
5 12 2021
7 12 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.
Objective
This study aimed to describe the impact of the first COVID-19 lockdown in France on the activity of a Child Advocacy Center.
Methods
This cross-sectional, observational study included all children involved in the activity of the CAC during the first lockdown, from March 16 to May 10, 2020 and the next 3 months and the corresponding periods in 2018 and 2019. Cases were considered severe when a hospitalization, social alert and/or judicial report to the prosecutor was decided.
Results
Data for 1583 children were analyzed. During the lockdown, the global center activity decreased with 26.4 consultations per 100.000 children in 2018, 46 in 2019 and 20.7 in 2020 (p < 0.001). Judicial activity decreased (forensic examinations and child forensic interview recordings), whereas assessment consultations increased. Cases were more severe during the lockdown than in 2019 and 2018 (12.3, 9.4 and 6.04/100.000 children, respectively, p < 0.0001). The global activity of the center increased in the 3 months after the lockdown as compared with during the lockdown (38.2/100.000 versus 20.7/100.000, respectively, p < 0.001) but did not differ from activity in 2018 and 2019. Severe cases were more frequent in the 3 months after the lockdown than the previous years (13.7/100.000 in 2020, 9.62 in 2019 and 8.17 in 2018, p = 0.0002).
Conclusion
The CAC activity decreased during the lockdown in France but the increase in incidence of severe abuse cases during the lockdown and the next 3 months confirm the need for optimal screening, care and support of child abuse and neglect victims even in the context of health crisis.
Keywords
COVID-19
Lockdown
Child abuse and neglect
Child advocacy center
==== Body
pmc1 Introduction
As compared with adult epidemiologic information, the proportion of children infected by the SARS-CoV-2 virus has been lower since the beginning of the pandemic, with also very mild morbidity and quasi-null mortality (Santé Publique France, 2021). However, although initially underestimated, the psychosocial impact and negative effects on the mental and social health of the pediatric population, including adolescents, are now reported by many countries (Adams, 2020; Bryant, Oo, & Damian, 2020; Cluver et al., 2020; Ghosh, Dubey, Chatterjee, & Dubey, 2020; Green, 2020; Marques, de Moraes, Hasselmann, Deslandes, & Reichenheim, 2020; Meherali et al., 2021; UNICEF, 2020). The first lockdown led to the closure of schools, daycare centers, preschool nurseries, and sports clubs, and stopping any outside school activities. Children and their parents increased their outdoor activities, parents were teleworking, social interactions were modified, dietary and sleeping habits were altered, and sometimes, the influence of the media was harmful with discordant information and the presentation of an uncertain future. This abrupt disruption of daily routines has led to psychological distress for many people and has put fragile psychosocial situations under pressure, especially with household crowding associated with job loss and economic insecurity. The exceptional situation of the lockdown has also been associated with increased use of the national hotline for child abuse and neglect (GIP Enfance en Danger, 2020.).
Furthermore, the lockdown decreased access to professionals who usually detect suggestive signs and situations of child abuse and neglect, such as teachers, health professionals, social and youth workers (Caron, Plancq, Tourneux, Gouron, & Klein, 2020). During the first lockdown, visits to pediatric emergency departments and hospital admissions decreased sharply (Angoulvant et al., 2020; Swedo, 2020). This situation may have also led to reduced opportunities for screening for child abuse (Martins-Filho, Damascena, Lage, & Sposato, 2020).
The aim of this study was to evaluate the quantitative and qualitative impacts of the first lockdown in France on the activity of a regional child advocacy center. Our main hypothesis was based on the clinical perceptions of the staff during the lockdown suggesting fewer consultations for abuse combined with increased severity of cases in a pediatric population that differed from usual observations.
2 Patients and methods
2.1 Study design and setting
This single-center, observational, cross-sectional study was conducted at Nantes University Hospital and was based on the registry of the Child Advocacy Center (CAC), a pediatric hospital mobile team specialized in child abuse and neglect that has 2 missions: diagnostic and forensic support. These missions are accomplished by the following: 1) assessment consultations (i.e., medico-psychological consultations dedicated to early detection and diagnosis of child abuse and neglect after a referral by a healthcare or socio-educative professional or by the family); 2) forensic examinations (i.e., medical and psychological consultations requested by law enforcement) and 3) child forensic interview recordings (i.e., by judicial investigators in a dedicated room).
We compared weeks 12 to 19 in 2018, 2019 and 2020 (corresponding to the first lockdown in France from March 16 to May 10, 2020). We also studied the post-lockdown evolution by comparing the lockdown period to the post-lockdown period in 2020 (comparing weeks 12–19 to 20–33) and also the corresponding weeks (weeks 20–33) in 2018 and 2019.
2.2 Patient selection and data collection
We included all consultations for a child ≤15 years old by the CAC team for weeks 12 to 33 in 2018, 2019 and 2020. When a consultation was dedicated to siblings, we considered each child as having an individual consultation. If the same patient was seen multiple times, every consultation was analyzed individually. Consultations were identified in the routine registry systematically completed after each consultation.
We analyzed the global activity of the CAC quantitatively (number of consultations by the CAC for children ≤15 years old between weeks 12–19 and 20–33 in 2020 and compared to 2018 and 2019) but also qualitatively by measuring the severity of cases of child abuse and neglect comparing weeks 12–19 and 20–33 in 2020 to those in 2018 and 2019. A case was considered “severe” if at least one of the three following criteria was met at the time of the consultation by the CAC: a hospitalization decision, a social alert to child protective services, or a judicial report to the prosecutor. We described these severity criteria quantitatively (incidence rate) and qualitatively (proportion of total consultations).
The following data were analyzed for all consultations: type of consultations, hospitalization, types of alerts (judicial reports to the prosecutor or social alerts to the district child protection services). For assessing consultations and forensic examinations (but not child interview forensic recordings), we also analyzed the child's age at the time of the consultation, origin of the referral, type of violence suspected (psychological, physical, sexual, domestic, neglect), violence inside or outside the family, alleged perpetrator (age, status and relationship to the victim) and hospitalization. All these data are routinely available in the administrative database.
2.3 Statistical methods
Qualitative data are described with percentages. Percentages were compared by chi-square test or Fisher's exact test. To analyze trends in incidence rates and proportions, we used the chi-square test and tested the departure from linearity trend. Quantitative data were compared by Student t-test or Mann Whitney test depending on the distribution. We approximated incidence rates and their trends for consultations, hospitalizations and severe cases per 100.000 children ≤15 years old per month (number of events during n weeks ∗ 4 ∗ 100.000/number of exposed children ≤15 years old during n weeks ∗ n). For this estimation, we took into account the total number of children ≤15 years old in one administrative district (Loire-Atlantique) (https://atlas.loire-atlantique.fr/). The data for this population were available for the year 2017 (269.519 children ≤15 years old), with an estimated increase of 1.23% per year between 2013 and 2018. We then approximated the population of children ≤15 years old for 2018 (272.834), 2019 (276.190) and 2020 (279.587) and arbitrarily considered the population stable during the years. p < 0.05 was considered significant, but for post-hoc comparisons after global comparison for categorical variables with more than 2 classes, we reduced the p-value to <0.01. We used STATA v13 for analyses with the p trend module (Patrick Royston, 2014. “PTREND: Stata module for trend analysis for proportions,” Statistical Software Components S426101, Boston College Department of Economics).
2.4 Ethics
This study was approved by the local ethics committee (Groupe Nantais d'Ethique en Santé). Data were extracted from the epidemiologic database of the CAC and de-identified.
3 Results
3.1 Description of the population
A total of 1583 children were included in the study. Clinical features of the children seen in an assessment consultation or a forensic examination are summarized in Table 1. Data for weeks 20–33 in 2018 and 2019 are in Table S1.
3.1.1 Lockdown
The age classes of children seen during the lockdown did not differ from those for the same weeks in 2018 and 2019 (Table 1). The proportion of children seen in the CAC after an intra-hospital health caregiver request was higher during the lockdown as compared with the previous years (p = 0.002 vs 2018 and p < 0.001 vs 2019, post-hoc analysis) and forensic examination requests were lower (p = 0.003 vs 2018 and p < 0.001 vs 2019, post-hoc analysis). We also observed fewer sexual abuse cases in 2020 than 2019 (20% vs 38%, p = 0.002, post-hoc analysis); fewer violence cases were perpetrated outside the family as compared with 2018 and 2019 (5% vs 41% and 31%, p < 0.001 for each post-hoc analysis) and the alleged perpetrator was more often an adult family member: the father (67% vs 31% and 39%, p < 0.001 for each post-hoc analysis) or the mother (49% vs 20% in 2018, p < 0.001, and 29% in 2019, p = 0.001).Table 1 Clinical features of pediatric patients seen in assessment consultations and forensic examinations in the Nantes Child Advocacy Center during weeks 12 to 19 in 2018, 2019 and 2020 and weeks 20 to 33 in 2020.
Table 1 Weeks 12–19, 2018 (N = 100) n (%) Weeks 12–19, 2019 (N = 173) n (%) Weeks 12–19, 2020 (lockdown) (N = 101) n (%) Weeks 20–33, 2020 (post-lockdown) (N = 257) n (%) p-Valuea p-Valueb
Child age (years) Median [IQR] 8.6 [4–13.5] 8 [4–12.1] 7 [3–11.9] 9 [5.5–13] 0.17 0.003
<2 13 (13) 26 (15) 19 (19) 22 (9) 0.621 0.019
2–11 47 (47) 82 (47) 51 (50) 137 (53)
>11 40 (40) 65 (38) 31 (31) 98 (38)
Origin of the referral Forensic examinations 70 (70) 128 (74) 50 (50) 174 (68) 0.001 0.007
Intra-hospital health caregiver request 12 (12) 20 (12) 30 (30) 48 (19)
Extra-hospital health caregiver request 10 (10) 14 (8) 13 (13) 27 (11)
Family request 8 (8) 11 (6) 8 (8) 8 (3)
Type of child abuse and neglect
Physical Yes 57 (57) 110 (64) 62 (61) 151 (59) 0.56 0.65
No 43 (43) 63 (36) 38 (39) 106 (41)
Sexual Yes 31 (31) 65 (38) 20 (20) 68 (26) 0.008 0.188
No 69 (69) 108 (62) 81 (80) 189 (74)
Psychological Yes 28 (28) 45 (26) 34 (34) 121 (47) 0.40 0.021
No
Severe neglect Yes 17 (17) 24 (14) 26 (26) 45 (18) 0.045 0.079
No
Domestic violence Yes 26 (26) 23 (13) 19 (19) 72 (28) 0.032 0.072
No
Nature of the abuse related to the family Inside family 56 (56) 101 (58) 91 (90) 202 (79) <0.001 0.008
Outside family 41 (41) 54 (31) 5 (5) 45 (18)
Unknown 3 (3) 18 (10) 5 (5) 10 (4)
Age of the abuse alleged perpetrator (years) ≥18 72 (72) 137 (79) 94 (93) 202 (79) <0.001 0.004
<18 26 (26) 34 (20) 2 (2) 41 (16)
Unknown 2 (2) 2 (1) 5 (5) 14 (6)
Relationship with the victim
Father involved Yes 31 (31) 67 (39) 68 (67) 127 (49) <0.001 0.02
No 69 (69) 106 (61) 33 (33) 130 (51)
Mother involved Yes 20 (20) 50 (29) 49 (49) 81 (32) <0.001 0.003
No 80 (80) 123 (71) 52 (51) 176 (68)
a calculated by chi-square or Fisher exact test, comparing weeks 12–19 in 2018, 2019 and 2020.
b calculated by chi-square or Fisher exact test, comparing weeks 12–19 (lockdown) with weeks 20–33 (post lockdown) in 2020.
3.1.2 Post-lockdown
As compared with the lockdown period, in the post-lockdown period, the children seen were older (9 years [interquartile range 5.5–13] vs 7 years [3–11.9], p = 0.003) and more often referred by forensic services (174/257 [68%] vs 50/101 [50%], p = 0.007). The proportion of violence outside the family or perpetrated by minors was higher (Table 1).
As compared with 2018 and 2019, during the post-lockdown period, the children seen had similar clinical features except for 2: 1) more psychological abuse cases and 2) more violence perpetrated inside the family (Table S1).
3.2 Quantitative activity (incidence rates)
3.2.1 Lockdown
The analysis of the global activity of the CAC during the lockdown is described in Fig. 1. The CAC experienced globally less activity during the lockdown in terms of decreased number of forensic examination requests and child forensic interview recordings. The incidence of consultations was 26.4, 46 and 20.7 per 100.000 children ≤15 years old living in Loire-Atlantique and per month in 2018, 2019 and 2020, respectively (p < 0.001, chi-square test). However, the assessment consultations increased, with 5.5, 8.33 and 9.12 routine examinations per 100.000 children ≤15 years old living in Loire-Atlantique and per month in 2018, 2019 and 2020, respectively (p = 0.03, chi-square test for trend; no deviance from linear trend). The incidence of hospitalizations increased between 2018 and 2020, with 2.7, 5.1 and 6.8 hospitalizations in Nantes University Hospital per 100.000 children ≤15 years old and per month (p = 0.02, chi-square test for trend; no deviance from linear trend). The incidence of severe cases referred to the CAC increased, with 6.04, 9.4 and 12.3 cases per 100.000 children ≤15 years old living in Loire-Atlantique and per month (p = 0.006, chi-square test for trend; no deviance from linear trend) (Table S2).
Fig. 1 Activity of the child advocacy center (CAC) expressed as number of events per 100.000 children ≤ 15 years old living in Loire-Atlantique per month during weeks 12 to 19 and 20 to 33 in 2018, 2019 and 2020. Severe case = hospitalization and/or social alert to child protective services and/or a judicial report to the prosecutor.
*p < 0.01, chi-square test for trend (significant increase from 2018 to 2020, weeks 12–19); **p < 0.01, chi-square test (significant difference between the 3 years, weeks 12–19); NS, p > 0.05 (non-significant difference between 3 groups for weeks 12–19 or weeks 20–33); #, p < 0.01, chi-square test (significant difference between post-lockdown to lockdown in 2020); 0, p > 0.05 chi-square test (non-significant difference between post-lockdown to lockdown in 2020).
Fig. 1
3.2.2 Post-lockdown
In the post-lockdown period, the CAC global activity increased, with 38.2 per 100.000 children ≤15 years old living in Loire-Atlantique and per month as compared with 20.7 during the lockdown (p < 0.001) (Table S1). However, no catch-up was observed, and the global activity post-lockdown did not differ from that observed during the same weeks in 2018 and 2019, with 33.9 and 39.9 consultations per 100.000 children ≤15 years old living in Loire-Atlantique, respectively (p = 0.08). We also observed a significant increase in severe cases, with 8.17, 9.62 and 13.7 children being victims of severe abuse per 100.000 children ≤15 years old living in Loire-Atlantique in 2018, 2019 and 2020, respectively (p = 0.0002; no deviance from linear trend).
3.3 Qualitative activity (distribution)
3.3.1 Lockdown
The analysis of the severity of the cases seen in assessment consultations and forensic examinations during the lockdown is summarized in Fig. 2 . During the lockdown, we found relatively more hospitalization decisions in 2020 than 2018 and 2019 (38/101 [38%] vs 15/100 [15%] and 28/173 [16%], p = 0.003, chi-square test for trend; deviance from linear trend), more social alerts to child protective services (25/101 [25%] vs 13/100 [13%] and 24/173 [14%], p = 0.02, chi-square test for trend; no deviance from linear trends), more judicial reports to the prosecutor (33/101 [33%] vs 15/100 [15%] and 20/173 [12%], p = 0.00, chi-square test for trend; deviance from linear trend), and relatively more severe cases (69/101 [68%] vs 33/100 [33%] and 52/173 [30%], p < 0.000, chi-square test for trend; deviance from linear trend).Fig. 2 Proportion of children with severe cases among the children cared for by the Child Advocacy Center team during the lockdown and post lockdown period and the corresponding period (weeks 12–19 and 20–33) in 2018 and 2019.
Severe case = hospitalization and/or social alert to child protective services and/or a judicial report to the prosecutor.
*p < 0.05, chi-square test for trend (significant increase from 2018 to 2020, weeks 12–19); **p < 0.05, chi-square test (significant increase between from 2018 to 2020, weeks 20–23); NS, p > 0.05 (non-significant difference between the 3 years, weeks 20–33); #, p < 0.01, chi-square test (significant difference between post-lockdown to lockdown in 2020); 0, p > 0.05, chi-square test (non-significant difference between post-lockdown to lockdown in 202.
Fig. 2
3.3.2 Post-lockdown
In the post-lockdown period, we observed relatively more hospitalization decisions in 2020 than 2018 and 2019 (61/257 [24%]) vs 37/222 [17%] and 45/273 [16%]), p = 0.04, chi-square test for trend; no deviance from linear trend), more social alerts to child protective services (63/257 [25%] vs 34/222 [15%] and 50/273 [18%], p = 0.01, chi-square test for trend; no deviance from linear trends), and relatively more severe cases (134/257 [52%] vs 78/22 [35%] and 93/273 [34%], p = 0.0001, chi-square test for trend; deviance from linear trend). Among all consultations, the proportion of judicial reports to the prosecutor did not differ significantly in 2020 as compared with 2018 and 2019 (39/257 [15%] vs 29/222 [13%] and 33/273 [12%], p = 0.57, chi-square test) (Fig. 2).
As compared with the lockdown, in the post-lockdown period, the proportion of hospitalizations was reduced (38% vs 24%, p = 0.008) as were judicial reports to the prosecutor (33% vs 15%, p < 0.0001) and severe abuse (68% vs 52%, p = 0.0006) (Fig. 2).
4 Discussion
According to our clinical impression, the lockdown in France increased intra-familial violence and the severity of child abuse and neglect. Our hypothesis for the effect of the first lockdown on the activity of the CAC in Nantes University Hospital was confirmed and also reported in many countries (Barboza, Schiamberg, & Pachl, 2021; Kovler et al., 2021). Indeed, the global activity of the CAC decreased during the lockdown, mainly related to a decrease in judicial activity. However, the number of screening consultations increased, despite a sharp decline in visits to pediatric emergency departments, which suggests the central role of the pediatric hospital mobile teams specialized in child abuse and neglect. Indeed, we may have expected that the increase observed the previous year would have been slowed by the lockdown as it was for other pediatric consultations, but the increase was still in line with the previous year, as shown by a significant increase without deviation to linearity. We also found increased severity of abuse cases during the lockdown and the next 3 months (hospitalizations and judicial reports to the prosecutor). Here again the increase in incidence of hospitalizations or judicial reports was not deviant to a linear trend, but the increase in proportion was higher (deviant to a linear trend). This observation gave the impression that there were more severe cases due to the lockdown when in fact the increase in severe cases observed the previous year was just not slowed by the lockdown as was observed for the other reasons for consultation. Moreover, although hospitalisations and judicial reports did not significantly increase in the post-lockdown period as compared with the previous year, the number of social alerts to child protective services significantly increased in comparison to previous years and to the lockdown period. This finding could reflect a “catch-up” of the less severe cases not seen during the lockdown. A recent study based on a national administrative database did not show an absolute increase in hospitalizations related to child abuse, unlike in the present study, but the authors observed a relative increase, which is consistent with our results (Loiseau et al., 2021). We hypothesize that the identification of abused children by using ICD-10 codes may be less sensitive in an administrative database than in a database completed by clinicians of the CAC, with a consequent under-evaluation of the absolute number of hospitalizations.
The present study has several limitations. First it is a monocentric study, and results may not be generalized to other administrative districts in France. Nevertheless, if the incidence and detection of child abuse differs among areas, the changes in activity may be less susceptible to inter-regional variations because the whole French territory was under the same restrictions during the lockdown. Second, we were not able to assess the real incidence rates in Loire-Atlantique because we studied the consultations in only the Nantes CAC (without the other smaller center). Nevertheless, the population flow between centers did not change during the studied period; hence, the changes in incidence reflect the changes in activity, and the chi-square test results for trend are relevant. Moreover, the most severe cases requiring admission to intensive care units were all admitted to the CAC university hospital. Finally, we observed a significant increase in all factors, as assessed by the chi-square test for trend, but this trend was assessed at only 3 times (2018–2019 and 2020) and for most, the result did not deviate from the linear trend. Therefore, we cannot confirm that this increase was due to the COVID-19 crisis and not preexistent trends. Time-series analyses would be relevant to analyze more precisely the effect of the pandemic and could be conducted in light of this work.
Consequently, the COVID-19 healthcare crisis should continue to have a deleterious impact on child abuse and neglect screening, diagnosis and quality care. This study also shows that distressed and abusive families continued to visit the public hospital when few childcare professionals were available, as other professionals are able to recognise signs and symptoms that indicate child abuse and neglect (Caron et al., 2020; Martins-Filho et al., 2020). During the lockdown, pediatric emergency departments and hospitalization services continued to welcome all children and their families, especially children at risk, around the clock for diagnosis, care and protection. As recommended by the French Society of Medico-legal Paediatrics (Balençon, Avenard, Delacourt, Gras-Le Guen, & Hedouin, 2020), there is a need to consolidate and further develop pediatric hospital mobile teams specializing in child abuse and neglect throughout the country, in accordance with the national plan to improve the early detection and diagnosis of child violence (Ministère des solidarités et de la santé, 2019; Martinkevich et al., 2020).
5 Conclusion
This study points out that despite children's health not being directly affected by the severity of the COVID-19 pandemic, the disease had a major indirect negative impact and emphasizes the importance of keeping care available for them (Tsur & Abu-Raiya, 2020). The upcoming sanitary measures must weigh benefits and risks for children's health at a time when it is also important to keep their needs at the forefront (Bérubé et al., 2020). Children must have access to health and social services, daycare centers and schools. Some preventive and intervention targets should be implemented as soon as possible for families most at risk of violence (Brown, Doom, Lechuga-Peña, Watamura, & Koppels, 2020; Griffith, 2020; Herrenkohl, Scott, Higgins, Klika, & Lonne, 2021; Katz et al., 2021).
The following are the supplementary data related to this article.Table S1
Clinical features of patients seen in routine examinations and forensic examinations between week 20 and week 33 in 2018, 2019 and 2020.
Table S1
Table S2
Incidence expressed as number of event per 100.000 children aged 15 years or less living in Loire Atlantique per month.
Table S2
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors declare that they have no competing interest.
==== Refs
References
Adams C. Is a secondary pandemic on its way? Retrieved February 26, 2021, from 2020 Institute of Health Visiting https://ihv.org.uk/news-and-views/voices/is-a-secondary-pandemic-on-its-way/
Angoulvant F. Ouldali N. Yang D.D. Filser M. Gajdos V. Rybak A. Guedj R. Soussan-Banini V. Basmaci R. Lefevre-Utile A. Brun-Ney D. Beaujouan L. Skurnik D. COVID-19 pandemic: Impact caused by school closure and national lockdown on pediatric visits and admissions for viral and non-viral infections, a time series analysis Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America 2020 10.1093/cid/ciaa710
Balençon M. Avenard G. Delacourt C. Gras-Le Guen C. Hedouin V. Pédiatrie médico-légale: Mineurs en danger: du dépistage à l’expertise pour un parcours spécialisé protégé 2020
Barboza G.E. Schiamberg L.B. Pachl L. A spatiotemporal analysis of the impact of COVID-19 on child abuse and neglect in the city of Los Angeles, California Child Abuse & Neglect 116 2021 104740 10.1016/j.chiabu.2020.104740
Bérubé A. Clément M.-È. Lafantaisie V. LeBlanc A. Baron M. Picher G. Turgeon J. Ruiz-Casares M. Lacharité C. How societal responses to COVID-19 could contribute to child neglect Child Abuse & Neglect 104761 2020 10.1016/j.chiabu.2020.104761
Brown S.M. Doom J.R. Lechuga-Peña S. Watamura S.E. Koppels T. Stress and parenting during the global COVID-19 pandemic Child Abuse & Neglect 110 2020 104699 10.1016/j.chiabu.2020.104699
Bryant D.J. Oo M. Damian A.J. The rise of adverse childhood experiences during the COVID-19 pandemic Psychological Trauma: Theory, Research, Practice and Policy 12 S1 2020 S193 S194 10.1037/tra0000711 32551773
Caron F. Plancq M.-C. Tourneux P. Gouron R. Klein C. Was child abuse underdetected during the COVID-19 lockdown? Archives de Pédiatrie 27 7 2020 399 400 10.1016/j.arcped.2020.07.010 32807622
Cluver L. Lachman J.M. Sherr L. Wessels I. Krug E. Rakotomalala S. Blight S. Hillis S. Bachman G. Green O. Butchart A. Tomlinson M. Ward C.L. Doubt J. McDonald K. Parenting in a time of COVID-19 Lancet (London, England) 395 10231 2020 e64 10.1016/S0140-6736(20)30736-4
Ghosh R. Dubey M.J. Chatterjee S. Dubey S. Impact of COVID -19 on children: Special focus on the psychosocial aspect Minerva Pediatrica 72 3 2020 226 235 10.23736/S0026-4946.20.05887-9 32613821
GIP Enfance en Danger Bilan relatif à l’activité du 119. Période de confinement du 18 mars au 10 mai https://www.giped.gouv.fr/pdf/Etude-confinement-119-20mai2020.pdf 2020
Green P. Risks to children and young people during covid-19 pandemic BMJ (Clinical Research Ed.) 369 2020 m1669 10.1136/bmj.m1669
Griffith A.K. Parental burnout and child maltreatment during the COVID-19 pandemic Journal of Family Violence 2020 10.1007/s10896-020-00172-2
Herrenkohl T.I. Scott D. Higgins D.J. Klika J.B. Lonne B. How COVID-19 is placing vulnerable children at risk and why we need a different approach to child welfare Child Maltreatment 26 1 2021 9 16 10.1177/1077559520963916 33025825
Katz C. Priolo Filho S.R. Korbin J. Bérubé A. Fouché A. Haffejee S. Kaawa-Mafigiri D. Maguire-Jack K. Muñoz P. Spilsbury J. Tarabulsy G. Tiwari A. Thembekile Levine D. Truter E. Varela N. Child maltreatment in the time of the COVID-19 pandemic: A proposed global framework on research, policy and practice Child Abuse & Neglect 116 Pt 2 2021 104824 10.1016/j.chiabu.2020.104824
Kovler M.L. Ziegfeld S. Ryan L.M. Goldstein M.A. Gardner R. Garcia A.V. Nasr I.W. Increased proportion of physical child abuse injuries at a level I pediatric trauma center during the Covid-19 pandemic Child Abuse & Neglect 116 2021 104756 10.1016/j.chiabu.2020.104756
Loiseau M. Cottenet J. Bechraoui-Quantin S. Gilard-Pioc S. Mikaeloff Y. Jollant F. François-Purssell I. Jud A. Quantin C. Physical abuse of young children during the COVID-19 pandemic: Alarming increase in the relative frequency of hospitalizations during the lockdown period Child Abuse & Neglect 122 2021 105299 10.1016/j.chiabu.2021.105299
Marques E.S. de Moraes C.L. Hasselmann M.H. Deslandes S.F. Reichenheim M.E. Violence against women, children, and adolescents during the COVID-19 pandemic: Overview, contributing factors, and mitigating measures Cadernos De Saude Publica 36 4 2020 e00074420 10.1590/0102-311X00074420
Martinkevich P. Larsen L.L. Græsholt-Knudsen T. Hesthaven G. Hellfritzsch M.B. Petersen K.K. Møller-Madsen B. Rölfing J.D. Physical child abuse demands increased awareness during health and socioeconomic crises like COVID-19 Acta Orthopaedica 91 5 2020 527 533 10.1080/17453674.2020.1782012 32573297
Martins-Filho P.R. Damascena N.P. Lage R.C. Sposato K.B. Decrease in child abuse notifications during COVID-19 outbreak: A reason for worry or celebration? Journal of Paediatrics and Child Health 2020 10.1111/jpc.15213. doi:10.1111/jpc.15213
Meherali S. Punjani N. Louie-Poon S. Rahim K.A. Das J. Salam R. Lassi Z. Mental Health of Children and Adolescents Amidst COVID 19 and Past Pandemics: A Rapid Systematic Review 2021 10.20944/preprints202103.0149.v1
Ministère des solidarités et de la santé Pacte pour l’enfance. Synthèse du plan violence November https://solidarites-sante.gouv.fr/IMG/pdf/synthese_du_plan_violences.pdf 2019
Royston P. Statistical Software Components S426101, Boston College Department of Economics PTREND: Stata module for trend analysis for proportions 2014 https://econpapers.repec.org/software/bocbocode/s426101.htm
Santé Publique France COVID-19: Point épidémiologique du 5 aout Retrieved August 5, 2021, from https://www.santepubliquefrance.fr/maladies-et-traumatismes/maladies-et-infections-respiratoires/infection-a-coronavirus/documents/bulletin-national/covid-19-point-epidemiologique-du-5-aout-2021 2021
Swedo E. Trends in U.S. Emergency Department visits related to suspected or confirmed child abuse and neglect among children and adolescents aged 18 Years before and during the COVID-19 pandemic—United States, January 2019–September 2020 MMWR. Morbidity and Mortality Weekly Report 69 2020 10.15585/mmwr.mm6949a1
Tsur N. Abu-Raiya H. COVID-19-related fear and stress among individuals who experienced child abuse: The mediating effect of complex posttraumatic stress disorder Child Abuse & Neglect 110 2020 104694 10.1016/j.chiabu.2020.104694
UNICEF Geneva Palais briefing note on the impact of COVID-19 on children 2020 UNICEF https://www.unicef.org/press-releases/geneva-palais-briefing-note-impact-covid-19-children
| 34952733 | PMC9741962 | NO-CC CODE | 2022-12-14 23:29:58 | no | Child Abuse Negl. 2022 Aug 13; 130:105443 | utf-8 | Child Abuse Negl | 2,021 | 10.1016/j.chiabu.2021.105443 | oa_other |
==== Front
J Nerv Ment Dis
J Nerv Ment Dis
JNMD
The Journal of Nervous and Mental Disease
0022-3018
1539-736X
Lippincott Williams & Wilkins
35764594
JNMD_220265
10.1097/NMD.0000000000001560
00006
3
Original Articles
COVID-19 Survivors' Intensive Care Unit Experiences and Their Possible Effects on Mental Health
A Qualitative Study
Telatar Tahsin Gökhan MD ∗
Telatar Ayça MD [email protected]
†
Hocaoğlu Çiçek MD [email protected]
‡
Hızal Ayşe MD [email protected]
§
Sakın Mustafa MD [email protected]
†
Üner Sarp MD, PhD [email protected]
∥
∗ Department of Public Health, Faculty of Medicine, Recep Tayyip Erdoğan University
† Department of Anesthesiology and Reanimation, Rize State Hospital
‡ Department of Psychiatry, Faculty of Medicine, Recep Tayyip Erdoğan University
§ Department of Anesthesiology and Reanimation, Training and Research Hospital, Rize
∥ Department of Public Health, Faculty of Medicine, Lokman Hekim University, Ankara, Turkey.
Send reprint requests to Tahsin Gökhan Telatar, MD, Recep Tayyip Erdogan Universitesi, Tip Fakultesi, Dekanlik Binasi, Oda No:316, Islampasa Mh., 53200, Merkez, Rize, Turkey. E-mail: [email protected].
12 2022
27 6 2022
27 6 2022
210 12 925929
Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.
2022
Wolters Kluwer Health, Inc. All rights reserved.
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Abstract
It is known that being hospitalized in the intensive care unit (ICU) for any reason is a risk factor for future psychiatric problems. This qualitative study aims to identify the experiences of coronavirus disease 2019 (COVID-19) ICU survivors and provide insights for relevant mental health problems after being discharged. Participants were COVID-19 patients discharged from ICUs of a secondary care hospital. The experiences of 21 ICU survivors were evaluated using Colaizzi's 7-step approach, which were determined by the purposeful sampling method. There were three themes generated from the interviews as “emotions on COVID-19 diagnosis,” “feelings about ICU stay and health care providers,” and “life in the shadow of COVID-19.” Two subthemes for every single theme were generated, and a total of 19 codes were extracted. It is essential to understand the individual's unique experiences in designing preventive interventions and apply individual preventive mental health interventions during ICU stay.
Key Words
COVID-19
intensive care units
survivors
experiences
phenomenological approach
SDCT
==== Body
pmcCoronavirus disease 2019 (COVID-19), caused by SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), has become a devastating and unprecedented global public health problem since the 21st century (World Health Organization, 2021). According to Johns Hopkins University Coronavirus Resource Center, total cases exceeded 300 million with 5.5 million deaths globally as January 2022 (Johns Hopkins University Coronavirus Resource Center, 2022). Although there are no global intensive care unit (ICU) admission rates due to COVID-19, the global shortage of ICU beds for COVID-19 patients reveals the high incidence of ICU admission (Ladds et al., 2020). Although it varies periodically and from country to country, at the beginning of January 2022 in the United States, France, and United Kingdom, the weekly new hospital admissions for COVID-19 per million were 298, 165, and 217, respectively. In the same countries at the same period, the number of COVID-19 patients in the ICU per million were 58.9, 53.0, and 12.7, respectively (Ritchie et al., 2020).
ICU stay may have psychological effects, including depression, anxiety, and posttraumatic stress disorder (PTSD) regardless of the underlying disease (Pattison, 2005). The depressive symptoms occur in almost 30% of patients after discharge from ICU (Wang et al., 2017). Four of every five ICU survivors experience new or worsened psychological or psychical problems due to post–intensive care syndrome (Needham et al., 2012).
Emerging disease outbreaks are known to cause severe psychological effects on patients (Kim et al., 2018; Kisely et al., 2020). There are still many unknown situations regarding the COVID-19 clinic. The unknown course of the COVID-19 raises concerns about the future that causes fear (Olufadewa et al., 2020). Although it commonly affects the respiratory system, many studies suggest that COVID-19 affects many systems including mental health (Dubey et al., 2020). Mental health problems including anxiety, depression, sleep disorders, and PTSD have been reported to be related with COVID-19 (Choi et al., 2020; Wang et al., 2020; Yang et al., 2020). A large-scale retrospective cohort study revealed that COVID-19 patients who were admitted to the ICU had higher hazards ratios for any psychiatric diagnoses compared with those who were not admitted to the ICU (Taquet et al., 2021).
Everyone has unique responses to stressful life events, and their coping methods differ. Environmental, biological, and psychological factors, life experiences, personality traits, and cultural factors increase this diversity (Schneiderman et al., 2005). Qualitative studies are successful research methods in revealing these differences. A qualitative study design would provide detailed insight into the quality of health services in the ICU and the patients' recovery process, which are usually missed in quantitative studies (Teti et al., 2020). Restructuring the health services provided in ICUs under suitable conditions during the COVID-19 pandemic would be possible with understanding the experiences gained from qualitative studies (Bavel et al., 2020; Kürtüncü et al., 2021; Olsen et al., 2017; Shah, 2020).
This study aims to understand the subjective experience of COVID-19 patients during their ICU stay to provide fundamental data for managing the ICU survivors' future mental health problems.
METHODS
Study Design
This qualitative research was carried out according to the Standards for Reporting Qualitative Research (O'Brien et al., 2014). All stages of the study and liabilities are clearly explained to the participants before asking for their consent. The Ethical Board of Recep Tayyip Erdogan University, Faculty of Medicine, approved the study on December 23, 2020 (registration number 2020/246). The phenomenological approach allows researchers to analyze people's unique experiences through dialogue, especially in health sciences (Berry et al., 2006). Colaizzi's 7-step approach was used to ensure the reliability of the patient experiences (Edward and Welch, 2011). The researchers consisted of two public health specialists, one psychiatrist, and three anesthesiologists. All researchers had previous experiences with qualitative research design: the public health specialists and the psychiatrist mainly designed the study, whereas the anesthesiologists conducted the interviews. During the processing of data, all researchers contributed equally. All anesthesiologists had more than 10 years of ICU experience, which dramatically improved the quality of interviews and none of them have treated the participants.
All interviews were conducted during January and February of 2021. A purposeful sampling method with criterion strategy was used, and the study was held with 21 participants. This method ensures the selection of possible information-rich cases (Palinkas et al., 2015). The inclusion criteria were as follows: having positive polymerase chain reaction test result for COVID-19 before ICU admission, being treated at least 72 hours in the ICU because of COVID-19, being discharged from the ICU to home or a general ward, being older than 18 years, and giving consent for participation to study. Exclusion criteria were being treated by any of the researchers of this study, having a cognitive or mental condition hindering participation, and not consenting to participate. After every five interviews, the researchers carefully evaluated the content to specify the sample size saturation. Potential themes and codes were defined until no themes or codes could be generated in five consecutive interviews.
Data Collection
Data were collected by both face-to-face interviews and phone calls. The patients who were discharged from the ICU and continued their treatment in a general ward in the same hospital were interviewed directly. Other participants were interviewed by phone calls. Patients' medical records were obtained from their hospital files, including their accompanying chronic conditions, duration of ICU stay, and presence of intubation. The data gathering form had four main topics: general opinions about the COVID-19 pandemic, family and social support, ICU period, and future predictions. Researchers conducted the interviews with unconditional acceptance, used active listening techniques to avoid bias, and only made clarifications if requested by the participant. All interviews were recorded with audiorecorders and then converted to text. The interview questions and interview protocol have been provided as a Supplement File (Supplemental Digital Content 1, http://links.lww.com/JNMD/A157).
Data Processing and Analysis
Following the Colaizzi's 7-step approach, all interviews are transcribed from audio recordings. Then they were entered into the QDA Miner Lite v2.0.8 software. Then significant statements were extracted from participants' narratives, and those statements made up the study's themes and subthemes. The researchers carefully evaluated each statement, and general restatements were generated, which formed the codes for further analysis. After completing generating codes, every code is assigned to relevant themes and subthemes. Researchers developed a detailed description regarding all themes and condensed it into a fundamental structure. The last step of Colaizzi's method requires returning the results to the participants for validation. Still, this stage is heavily criticized because of the different perspectives of the researchers and the participants (Giorgi, 2006). Considering Merleau-Ponty's approach, which states “the experiencer is not necessarily the best judge of the meaning of his or her experience,” that step was skipped (Merleau-Ponty, 1964).
RESULTS
Of the participants, 61.9% (13) were male, and 90.5% (19) were married. The mean age was 54.05 ± 11.85, and 47.6% (10) had a high school degree or higher (Table 1). The average ICU stay was 15.95 ± 9.35 days, and 5 (23.8%) of the participants were intubated. Thirteen patients (61.9%) were interviewed in the hospital at the general wards, and the rest of the interviews were held on the phone. After analyzing the data, 3 themes, 6 subthemes, and 18 codes were generated (Table 2). The extracted three themes were “emotions on COVID-19 diagnosis,” “feelings about ICU stay and health care providers,” and “life in the shadow of COVID-19.”
TABLE 1 Sociodemographic Characteristics of the Participants
Characteristic Frequency or Mean Percentage or Standard Deviation
Age 54.05 11.85
Sex
Male 13 61.9
Female 8 38.1
Marital status
Married 19 90.5
Widow or divorced 2 9.5
Education status
Literate 4 19.0
Primary school 4 19.0
Secondary school 3 14.3
High school 4 19.0
University 3 14.3
MSc, PhD, etc. 3 14.3
TABLE 2 Distribution of Themes, Subthemes, and Codes Generated From the Interviews
Theme Subtheme Code Within Codes, n (%) Within Cases, n (%)
Emotions on COVID-19 diagnosis Toward others Anger 12 (5.4) 10 (47.6)
Support 13 (5.8) 11 (52.4)
Concern about family 6 (2.7) 6 (28.6)
Toward themselves Sadness 9 (4) 7 (33.3)
Regret 14 (6.3) 12 (57.1)
Physical disturbances 13 (5.8) 11 (52.4)
Feelings about ICU stay and health care providers Positive feelings Gratefulness 19 (8.5) 14 (66.7)
Relief 16 (7.1) 16 (76.2)
Feeling safe 1 (0.4) 1 (4.8)
Negative feelings Dissatisfaction 1 (0.4) 1 (4.8)
Fear and anxiety 33 (14.7) 17 (81.0)
Loneliness 5 (2.2) 5 (23.8)
Being uncomfortable 6 (2.7) 6 (28.6)
Life in the shadow of COVID-19 General effects Seeking information 13 (5.8) 13 (61.9)
Bewildering 14 (6.3) 9 (42.9)
Resignation 19 (8.5) 10 (47.6)
Future alterations Changing lifestyle 11 (4.9) 11 (52.4)
Deliberation 5 (2.2) 4 (19.0)
Expectation 14 (6.3) 13 (61.9)
Theme 1: Emotions on COVID-19 Diagnosis
Participants' feelings toward themselves and other people due to their experiences during the COVID-19 pandemic and related to their diseases were evaluated in this section. Those emotions or feelings were evaluated in two subthemes as feelings toward other people and themselves.
Feelings Toward Others
The illness process caused the participants to develop various feelings toward other people. These feelings include blaming and angering other people, concerns about family members, and being grateful to people for their support during their illness. Some of the quotes from the participants are as follows.
“There is no fear bigger than death. I scared so much, not for myself but my family, especially for my children. I couldn't stand the thought of leaving them alone just because of a small virus” (male, 61 years old).
“I'm angry with my neighbor. I told them not to visit us, but they didn't listen. They came to our hour house like nothing was going on. They never cared about the outbreak. Finally, I got the disease from them. I don't think that I'm going to forgive them ever” (female, 50 years old).
“My family did everything they could during my disease. They prayed for me, they spend everything they have. I owe them big time” (male, 63 years old).
Feelings Toward Themselves
Regret was the most common feeling among the participants. Many of them regret their attitudes and decisions before the disease. Sadness and physical disturbances were also other feelings of the participants toward themselves. Some of the quotes from the participants regarding those feelings are as follows.
“I blame myself. It's my fault that I relocated during the pandemics. I underestimated the outbreak, I regret so much for my decisions. I literally made myself sick” (male, 44 years old).
“It makes you feel sad about you and your family. My children were crying, they were asking Allah to save their father. I can't tell enough how that made me sad” (male, 54 years old).
“I was ready to give everything I have just for a deep breath. It was like drowning” (male, 43 years old).
Theme 2: Feelings About ICU Stay and Health Care Providers
Although five of the participants were intubated, all participants had time in the ICU to be aware of their surroundings. ICU experiences of the participants included positive and negative components.
Positive Feelings
The gratitude to the ICU and its staff, who helped them hold on to life, and the comfort of receiving treatment were examples of positive experiences. Some of the participants stated that receiving treatment at the ICU made them feel safe. Some of the quotes from the participants regarding those experiences are as follows.
“God bless all the white angels working at the ICU. I'm ashamed to say, but they even cleaned my diaper. I owe my life to them” (female, 46 years old).
“Words are not enough to describe the joy of being able to breathe again. Everything was tough, I would never want anyone to have these troubles, but the comfort of being able to breathe in at the end of all difficulties has no description” (female, 49 years old).
“They said that I should get treatment, but there is no room at ICU. At that moment, I realized the seriousness of the situation. I thought I was going to die, and there was nothing to do. When a place was opened in ICU, and they put me in there, I was relieved. I felt safe. I knew I wouldn't die anymore because they would be able to do all the necessary interventions for me at ICU. I trusted them so much that I didn't want to leave the ICU when I learned that I would recover and be discharged. I was afraid of getting worse again. I wanted to stay a little longer” (male, 46 years old).
Negative Feelings
Fear of death, dissatisfaction with the health service received, loneliness, and discomfort constitute the foremost negative experiences. Some of the quotes from the participants regarding those experiences are as follows.
“The only thing I can think about in the ICU was death. One day, three patients died in front of my eyes. I never thought that death was so close” (female, 28 years old).
“The sound of the breathing machine was like a nightmare. As if I still hear it” (male, 61 years old).
“I don't think the doctors looking after me were competent enough. There could have been more experienced staff. They didn't treat me nice” (male, 42 years old).
Theme 3: Life in the Shadow of COVID-19
The effects of the COVID-19 pandemic on the lives of the participants and their future thoughts under the influence of the pandemic were evaluated in this theme. Two subthemes were generated under this theme, which are general effects and future alterations.
General Effects
The pandemic has had many effects on people's lives. Participants made an effort to obtain information about the disease during this process, but in some cases, they could not fully understand what they lived through and its causes. Although some participants tended to accept their experiences, they stated that they were desperate in the face of what happened. Some of the quotes from the participants regarding those effects on their lives are as follows.
“I was obsessed with the news. I was watching every news about COVID, trying to get information about the disease and its course from every source I could find” (male, 71 years old).
“I was very careful about hygiene. I was constantly changing my mask. I was applying every precaution suggested to us. Despite all these measures, I have no idea how and where I got the disease. Many people become ill and survive the disease outpatient. Why did I suffer so much? I cannot understand any of these” (male, 63 years old).
“If anything happens, it happens. It is not possible to avoid fate. It was my destiny to catch this disease and experience these troubles” (female, 28 years old).
Future Alterations
It was observed that the COVID-19 disease and the ICU experiences of the participants caused changes and expectations regarding the rest of their lives. These changes mainly focus on lifestyles and future decisions. Some of the quotes from the participants regarding those changes and expectations are as follows.
“Whether the epidemic is over or not, I will never accept guests at home again. I will not come closer than 2 meters to anyone” (female, 72 years old).
“I stopped working as a taxi driver, now I'll protect myself. Making money is not more important than living” (male, 68 years old).
“Nobody should think that nothing will happen to them. Even if they are not thinking of themselves, they should be thinking of others. They do not know what the infected person and his family are going through. Everyone should do their part. I wish everyone followed the rules” (female, 56 years old).
DISCUSSION
This qualitative study of 21 COVID-19 ICU survivors has identified important findings regarding the experiences of those participants. It has been shown that patients who have survived through ICU stay have experiences that change their lifestyles and shape their expectations for themselves and their relatives. Intense anxiety, fear, regret, and helplessness were the main negative feelings. In addition, it was observed that the experiences of the patients during the treatment process varied widely. Although some patients were very grateful to the health personnel who provide health services to them, others stated that they were not satisfied with the service they received. These findings provide a framework for understanding the postdischarge life, expectations, and possible associated health problems of COVID-19 patients who have experienced ICU.
There are very few previously published studies about the ICU experiences of COVID-19 survivors, and most of them are conducted with quantitative methods (Berends et al., 2021; Tingey et al., 2020). One of the recent qualitative studies conducted with COVID-19 survivors revealed similar themes with our study as “living in limbo” and “psychological distress behind the wall” (Moradi et al., 2020). That study's findings especially imply the disturbance caused by uncertainty and mental health impairments due to COVID-19. Similarly, our findings reveal that uncertainty is a significant source of disturbance among ICU survivors. Also, the high frequency of codes such as anxiety, fear, regret, and anger suggests possible future mental health impairment risks for those individuals. Physical disturbances such as fatigue and breathlessness are also accompanied by psychological disturbances such as PTSD and decreased quality of life after discharge from ICU in COVID-19 patients (Halpin et al., 2021; Hosey and Needham, 2020). It has been shown by some qualitative studies that developing the capacity to cope with stress by supporting the positive emotions of COVID-19 patients is crucial for promoting their mental health (Olufadewa et al., 2020).
Existing studies are primarily cross-sectional and indicate that various mental health problems, especially PTSD, may be seen in individuals after ICU discharge (Pattison, 2005). However, these studies cannot reveal the reasons for the vulnerability of individuals. In our study, individual risk factors that may cause possible emotional distress in the future were determined. Each individual may have their own concerns. For example, an issue that is very important for one person may not be a priority for another person. Individual risk factors should be determined to prevent psychological problems in the post-ICU discharge period, and individual psychosocial support should be planned.
The negative experiences of COVID-19 survivors discharged from the ICU are not only caused by psychological factors. It has been observed that environmental stimuli they are exposed to, especially during ICU hospitalization, are also a significant stress factor. Events such as the constant alarm and other sounds heard from the devices, or the death of another patient lying next to them in front of their eyes and another patient being brought to the same bed after a short time, have a traumatic effect on patients (Gültekin et al., 2018). When these effects are combined with loneliness, they form the basis for chronic conditions that may occur in the future. For example, being exposed to the physical discomfort listed previously or being connected to a mechanical ventilator during ICU treatment increases the risk of delirium (Brown et al., 2020).
The qualitative approach provides a thorough evaluation of the experiences and potential challenges for the future life of COVID-19 ICU survivors. The most important aspect of the study is that it revealed the leading causes of possible mental health conditions after discharge. Each patient has a unique inner world, and the preventive interventions should be designed individually for the patients.
The main limitation of this study is the unknown psychiatric illness status of the patients after ICU discharge. The study was conducted with patients treated and discharged from the ICU due to COVID-19 after a certain period and did not include long-term follow-up results. It is thought that cohort-style studies that include quantitative items will help to evaluate the long-term results. The small sample size of the study prevents the results to be generalized for the general population, but the findings provide important clues for future studies. In this study, which included qualitative evaluations, more detailed sociodemographic information about the participants could not be presented and analyzes evaluating the relationships between factors and outcomes could not be performed.
CONCLUSIONS
The qualitative approach provides a thorough evaluation of the experiences and potential challenges for the future life of COVID-19 ICU survivors. As a result of this study, it can be stated that people who are hospitalized in intensive care due to COVID-19 are among the risky groups for mental symptoms. Therefore, effective and individualized psychosocial interventions are required for these patients. It is crucial to apply individualized psychiatric treatment approaches before discharge, during treatment in the ICU, in terms of postdischarge follow-up.
DISCLOSURE
No funding was received for this study.
The study was conducted according to acceptable research standards, including having obtained informed consent of participants. The study was approved by Ethical Board of Recep Tayyip Erdogan University, Faculty of Medicine with registration number 2020/246 on December 23, 2020.
All authors have read and approved the submitted manuscript.
The authors declare no conflict of interest.
T.G.T., Ç.H., and S.Ü. helped the design and planning of the study, analysis and interpretation of the data, and drafting and revising the final version of the manuscript. A.T., A.H., and M.S. helped the design and planning of the study, collecting the data and conducting the interviews, analysis, and interpretation of the data, and drafting and revising the final version of the manuscript.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.jonmd.com).
==== Refs
REFERENCES
Berends K Claus L De Waele E Crunelle CL Matthys F Vanderbruggen N (2021) Experiences of COVID-ICU-survivors: Mixed-methods study of psychological consequences by written survey [in Dutch]. Tijdschr Psychiatr. 63 :324–330.34043219
Berry DL Maliski SL Ellis WJ (2006) Qualitative research techniques. In Penson DF Wei JT (Eds), Clinical research methods for surgeons (pp 297–310). Totowa, NJ: Humana Press.
Brown KN Soo A Faris P Patten SB Fiest KM Stelfox HT (2020) Association between delirium in the intensive care unit and subsequent neuropsychiatric disorders. Crit Care. 24 :476.32736572
Choi EPH Hui BPH Wan EYF (2020) Depression and anxiety in Hong Kong during COVID-19. Int J Environ Res Public Health. 17 :3740.32466251
Dubey S Biswas P Ghosh R Chatterjee S Dubey MJ Chatterjee S Lahiri D Lavie CJ (2020) Psychosocial impact of COVID-19. Diabetes Metab Syndr Clin Res Rev. 14 :779–788.
Edward K-L Welch T (2011) The extension of Colaizzi's method of phenomenological enquiry. Contemp Nurse. 39 :163–171.22551429
Giorgi A (2006) Concerning variations in the application of the phenomenological method. Humanist Psychol. 34 :305–319.
Gültekin Y Özçelik Z Akıncı SB Yorgancı HK (2018) Evaluation of stressors in intensive care units. Turk J Surg. 34 :5–8.29756097
Halpin SJ McIvor C Whyatt G Adams A Harvey O McLean L Walshaw C Kemp S Corrado J Singh R Collins T O'Connor RJ Sivan M (2021) Postdischarge symptoms and rehabilitation needs in survivors of COVID-19 infection: A cross-sectional evaluation. J Med Virol. 93 :1013–1022.32729939
Hosey MM Needham DM (2020) Survivorship after COVID-19 ICU stay. Nat Rev Dis Primers. 6 :60.32669623
Johns Hopkins University Coronavirus Resource Center (2022) COVID-19 dashboard. Available at: https://coronavirus.jhu.edu/map.html. Accessed February 10, 2022.
Kim H-C Yoo S-Y Lee B-H Lee SH Shin H-S (2018) Psychiatric findings in suspected and confirmed Middle East respiratory syndrome patients quarantined in hospital: A retrospective chart analysis. Psychiatry Investig. 15 :355–360.
Kisely S Warren N McMahon L Dalais C Henry I Siskind D (2020) Occurrence, prevention, and management of the psychological effects of emerging virus outbreaks on healthcare workers: Rapid review and meta-analysis. BMJ (Clin Res Ed). 369 :m1642.
Kürtüncü M Kurt A Arslan N (2021) The experiences of COVID-19 patients in intensive care units: A qualitative study. Omega (Westport). 302228211024120 . doi:10.1177/00302228211024120.
Ladds E Rushforth A Wieringa S Taylor S Rayner C Husain L Greenhalgh T (2020) Persistent symptoms after Covid-19: Qualitative study of 114 “long covid” patients and draft quality principles for services. BMC Health Serv Res. 20 :1144.33342437
Merleau-Ponty M (1964) The primacy of perception. In Edie J (Ed), Phenomenology and the sciences of man (pp 43–95). Evanston, IL: Northwestern University Press.
Moradi Y Mollazadeh F Karimi P Hosseingholipour K Baghaei R (2020) Psychological disturbances of survivors throughout COVID-19 crisis: A qualitative study. BMC Psychiatry. 20 :594.33334331
Needham DM Davidson J Cohen H Hopkins RO Weinert C Wunsch H Zawistowski C Bemis-Dougherty A Berney SC Bienvenu OJ Brady SL Brodsky MB Denehy L Elliott D Flatley C Harabin AL Jones C Louis D Meltzer W Muldoon SR Palmer JB Perme C Robinson M Schmidt DM Scruth E Spill GR Storey CP Render M Votto J Harvey MA Harvey MA (2012) Improving long-term outcomes after discharge from intensive care unit: Report from a stakeholders' conference. Crit Care Med. 40 :502–509.21946660
O'Brien BC Harris IB Beckman TJ Reed DA Cook DA (2014) Standards for reporting qualitative research: A synthesis of recommendations. Acad Med. 89 :1245–1251.24979285
Olsen KD Nester M Hansen BS (2017) Evaluating the past to improve the future—A qualitative study of ICU patients' experiences. Intensive Crit Care Nurs. 43 :61–67.28899584
Olufadewa II Adesina MA Oladokun B Baru A Oladele RI Iyanda TO Ajibade OJ Abudu F (2020) “I was scared I might die alone”: A qualitative study on the physiological and psychological experience of COVID-19 survivors and the quality of care received at health facilities. Int J Travel Med Global Health. 8 :51–57.
Palinkas LA Horwitz SM Green CA Wisdom JP Duan N Hoagwood K (2015) Purposeful sampling for qualitative data collection and analysis in mixed method implementation research. Adm Policy Ment Health. 42 :533–544.24193818
Pattison N (2005) Psychological implications of admission to critical care. Br J Nurs (Mark Allen Publishing). 14 :708–714.
Ritchie H Mathieu E Rodés-Guirao L Appel C Giattino C Ortiz-Ospina E Hassel J Macdonald B Beltekian D Roser M (2020) Coronavirus (COVID-19) hospitalizations. Available at: https://ourworldindata.org/covid-hospitalizations#citation. Accessed February 10, 2022.
Schneiderman N Ironson G Siegel SD (2005) Stress and health: Psychological, behavioral, and biological determinants. Annu Rev Clin Psychol. 1 :607–628.17716101
Shah H (2020) Global problems need social science. Nature. 577 :295.31942060
Taquet M Geddes JR Husain M Luciano S Harrison PJ (2021) 6-month neurological and psychiatric outcomes in 236 379 survivors of COVID-19: A retrospective cohort study using electronic health records. Lancet Psychiatry. 8 :416–427.33836148
Teti M Schatz E Liebenberg L (2020) Methods in the time of COVID-19: The vital role of qualitative inquiries. Int J Qual Methods. 19 :160940692092096.
Tingey JL Bentley JA Hosey MM (2020) COVID-19: Understanding and mitigating trauma in ICU survivors. Psychol Trauma: Theory Res Pract Pol. 12 (S1 ):S100–S104.
Van Bavel JJ Baicker K Boggio PS Capraro V Cichocka A Cikara M Crockett MJ Crum AJ Douglas KM Druckman JN Drury J Dube O Ellemers N Finkel EJ Fowler JH Gelfand M Han S Haslam SA Jetten J Kitayama S Mobbs D Napper LE Packer DJ Pennycook G Peters E Petty RE Rand DG Reicher SD Schnall S Shariff A Skitka LJ Smith SS Sunstein CR Tabri N Tucker JA van der Linden S van Lange P Weeden KA Wohl MJA Zaki J Willer R (2020) Using social and behavioural science to support COVID-19 pandemic response. Nat Hum Behav. 4 :460–471.32355299
Wang C Pan R Wan X Tan Y Xu L Ho CS Ho RC (2020) 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 :1729.32155789
Wang S Mosher C Gao S Kirk K Lasiter S Khan S Kheir YN Boustani M Khan B (2017) Antidepressant use and depressive symptoms in intensive care unit survivors. J Hosp Med. 12 :731–734.28914277
World Health Organization (2021) World Health Organization. Weekly epidemiological update—2 May 2021. Geneva: World Health Organization.
Yang Y Li W Zhang Q Zhang L Cheung T Xiang Y-T (2020) Mental health services for older adults in China during the COVID-19 outbreak. Lancet Psychiatry. 7 :e19.32085843
| 35764594 | PMC9741998 | NO-CC CODE | 2022-12-15 00:01:29 | no | J Nerv Ment Dis. 2022 Dec 27; 210(12):925-929 | utf-8 | J Nerv Ment Dis | 2,022 | 10.1097/NMD.0000000000001560 | oa_other |
==== Front
J Nerv Ment Dis
J Nerv Ment Dis
JNMD
The Journal of Nervous and Mental Disease
0022-3018
1539-736X
Lippincott Williams & Wilkins
35687723
JNMD_220248
10.1097/NMD.0000000000001543
00003
3
Original Articles
SARS-CoV–Related Pandemic Outbreaks and Mental Disorder Risk
Deng Xiangling MD [email protected]
∗
He Mengyang MD [email protected]
∗
Zhang Jinhe MD [email protected]
†
Huang Jinchang PhD [email protected]
‡
Luo Minjing MD [email protected]
∗
Zhang Zhixin PhD [email protected]
§
Niu Wenquan PhD ∥
∗ Graduate School, Beijing University of Chinese Medicine
† Peking University Hui Long Guan Clinical Medical School, Beijing Hui Long Guan Hospital
‡ Beijing University of Chinese Medicine Third Affiliated Hospital
§ International Medical Department, China-Japan Friendship Hospital
∥ Institute of Clinical Medical Sciences, China-Japan Friendship Hospital, Beijing, China.
Send reprint requests to Wenquan Niu, PhD, Institute of Clinical Medical Sciences, China-Japan Friendship Hospital, No. 2 Yinghua East St, Chao Yang District, Beijing 100029, China. E-mail: [email protected]; or Zhixin Zhang, PhD, Institute of Clinical Medical Sciences, China-Japan Friendship Hospital, No. 2 Yinghua East St, Chao Yang District, Beijing 100029, China. E-mail: [email protected].
12 2022
18 5 2022
18 5 2022
210 12 900911
Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.
2022
Wolters Kluwer Health, Inc. All rights reserved.
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Abstract
This study aimed to quantify the association between exposure to pandemic outbreaks and psychological health via a comprehensive meta-analysis. Literature retrieval, study selection, and data extraction were completed independently and in duplicate. Effect-size estimates were expressed as odds ratio (OR) with 95% confidence interval (CI). Data from 22 articles, involving 40,900 persons, were meta-analyzed. Overall analyses revealed a significant association of exposing to SARS-CoV–related pandemics with human mental health (OR, 1.32; 95% CI, 1.24–1.40; p < 0.001). Subgroup analyses showed that anxiety (OR, 1.37; 95% CI, 1.19–1.58; p < 0.001), depression (OR, 1.28; 95% CI, 1.15–1.42; p < 0.001), posttraumatic stress (OR, 1.36; 95% CI, 1.17–1.58; p < 0.001), and psychological distress (OR, 1.25; 95% CI, 1.11–1.40; p < 0.001) were all obviously related to pandemic diseases. In the context of infectious disease outbreaks, the mental health of general populations is clearly vulnerable. Therefore, all of us, especially health care workers, need special attention and psychological counseling to overcome pandemic together.
Key Words
COVID-19
SARS
pandemic
mental health
psychological distress
SDCT
==== Body
pmcGrowing epidemiologic data have demonstrated that pandemics can cause a broad spectrum of issues involving both physical and mental health in human beings (Holmes et al., 2020; Norris et al., 2002). In particular, pandemics, such as severe acute respiratory syndrome (SARS) in 2003, influenza virus with the H1N1 subtype in 2009, Middle East respiratory syndrome (MERS) in 2012, Ebola virus in 2014, and coronavirus disease 2019 (COVID-19), during the past two decades are highly contagious and have caused heavy public health burdens regionally and globally (Fisman and Laupland, 2009; Klenk, 2014; Lam et al., 2004; Li et al., 2020b). More recently, the terrible COVID-19 have caused 42,055,863 confirmed cases, including 1,141,567 deaths as of October 24, 2020, and these numbers keep rising at an alarming rate. An equally serious problem is the profound consequence on spiritual damage to mankind, as researchers have proposed that patients who survived severe and life-threatening illnesses are at an enhanced risk of experiencing mental disorder. Factors such as long duration of quarantine, fears for infection, inadequate information, stigma, or financial loss were found to be more or less related to negative psychological impact (Brooks et al., 2020). In 2003, SARS outbreak had caused 50% of recovered patients who remained anxious and 29% of health care workers who experienced probable emotional distress (Nickell et al., 2004; Tsang et al., 2004). As demonstrated in the recent meta-analysis by Krishnamoorthy and colleagues, COVID-19 pandemic raised stress disorder by 40%, anxiety by 30%, burnout by 28%, depression by 24%, and posttraumatic stress disorder by 13% (Krishnamoorthy et al., 2020). Yet, the magnitude of the association between pandemic and mental disorder is still an open question, due to the sustainable skyrocketed growth of confirmed cases and death for COVID-19. Meanwhile, the pooled prevalence rate of psychological morbidities could not intuitively represent the association between the outbreak of pandemic and mental health.
To quantify the association between exposure to pandemic outbreaks and psychological health, we synthesized the results of cross-section studies in medical literature through a comprehensive meta-analysis.
METHODS
The performance of meta-analysis adhered to the guidelines in the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement (Page et al., 2021). The PRISMA checklist is presented in Supplementary Table 1 (Supplemental Digital Content 1, http://links.lww.com/JNMD/A147).
This study is a meta-analysis on published studies, and so ethical approval and informed consent are not needed.
Search Strategy
Literature search was conducted by scanning PubMed, EMBASE, and Web of Science databases as of August 4, 2020. The following medical topic terms are used: (obsession compulsion*) OR (depression) OR (depressive symptom*) OR (anxiety disorder*) OR (neurotic anxiety state) OR (hostility) OR (phobic disorder) OR (phobic*) OR (paranoid disorder) OR (paranoi*) OR (suicide*) OR (mental health*) OR (mental*) OR (psychiatric disorder) OR (psychiatric*) OR (psycho*) [Title/Abstract] AND (Acute Respiratory Syndrome Virus*) OR (SARS-Related*) OR (SARS-CoV) OR (Urbani SARS-Associated Coronavirus) OR (Influenza in Bird) OR (Avian Flu*) OR (H1N1 Virus*) OR (Swine-Origin Influenza A H1N1 Virus) OR (novel coronavirus vaccine) OR (coronavirus disease 2019 vaccine*) OR (2019 novel coronavirus vaccine) OR (SARS 2 vaccine*) OR (Wuhan coronavirus vaccine) OR (Zika*) OR (Middle East Respiratory Syndrome*) [Title/Abstract].
The reference lists of major retrieved articles were also manually searched to avoid potential missing hits.
Search process was independently conducted by two investigators (X.D. and M.H.) using the same medical topic terms. All references retrieved were combined, and duplicates were removed.
Inclusion/Exclusion Criteria
Our analysis was restricted to articles that met the following criteria: (1) study participants, aged ≥18 years old; (2) end points, related mental disorder; (3) study design, cross-sectional or cohort studies; (4) baseline exposure, different kinds of exposure to pandemic diseases; and (5) odds ratio (OR) as effect-size estimate. Articles were excluded if they involved study participants with serious diseases or if they are case reports or case series, editorials, and narrative reviews.
Data Extraction
Two investigators (M.H. and X.D.) independently extracted data from each qualified article, including first author, year of publication, country where study was conducted, sample size, sex, baseline age, study type, the type of exposed infectious disease, the method of assessing mental health, the type of psychological-related questionnaire, effect estimation, and severity of exposure to pandemic diseases, if available. The divergence was resolved through joint reevaluation of original articles, and if necessary, by a third author (W.N.).
Quality Assessment
The quality of all eligible studies was assessed using the 11-item checklist, which was recommended by Agency for Healthcare Research and Quality (AHRQ) (Rostom et al., 2004). The item would be scored “0” if the answer was “no” or “unclear,” whereas “1” represented the answer “yes.” Article quality was assessed to three different grades: low quality (0–3); moderate quality (4–7); and high quality (8–11). Differences in article quality were discussed to reach a final agreement.
Statistical Analyses
Data management and analysis were handled using the STATA software version 14.1 for Windows (Stata Corp, College Station, TX). Effect-size estimates were expressed as OR with 95% confidence interval (CI). Pooled effect-size estimates were derived under the random-effects Mantel-Haenszel model, irrespective of the magnitude of between-study heterogeneity.
The inconsistency index (I2), which represents the percent of diversity that is due to heterogeneity rather than chance, is used to quantify between-study heterogeneity. The I2 greater than 50% denotes significant heterogeneity, and a higher percentage indicates a higher degree of heterogeneity. To account for possible sources of between-study heterogeneity from clinical and methodological aspects, a large number of prespecified subgroup analyses were done according to major exposure subjects, the level of development of the countries, the different pandemic diseases, and the different kinds of exposure respectively.
The probability of publication bias was evaluated by both Begg's funnel plots and Egger's regression asymmetry tests at a significance level of 10%. The trim-and-fill method was used to estimate the number of theoretically missing studies.
RESULTS
Eligible Studies
After searching prespecified public databases using predefined medical subject terms, a total of 1073 articles were initially identified, and 22 of them were qualified for analysis, including 40,900 study participants (Abdessater et al., 2020; Cao et al., 2020; Chan and Huak, 2004; Choi et al., 2020; Gómez-Salgado et al., 2020; Gualano et al., 2020; Guo et al., 2020; Huang and Zhao, 2020; Lai et al., 2020; Leung et al., 2005; Li et al., 2020a; Lu et al., 2020; Peng et al., 2010; Rossi et al., 2020; Shacham et al., 2020; Sim et al., 2004; Tam et al., 2004; Verma et al., 2004; Wu et al., 2009; Xiao et al., 2020; Yang et al., 2020; Ying et al., 2020). The detailed selection process is schematized in Figure 1.
FIGURE 1 Flowchart of records retrieved, screened, and included in this meta-analysis.
Study Characteristics
Table 1 shows the baseline characteristics of studies that respectively recorded OR in this meta-analysis. Because of the lack of data on other pandemics, the present analysis was only restricted to SARS and COVID-19.
TABLE 1 The Baseline Characteristics of All Involved Studies in This Meta-analysis
First Author Year Country Gender Age, y Exposure Subjects Sample Size Study Type Infectious Disease Method Type of Questionnaire Assessment of Different Kinds of Exposures Mental Disorder Effect Size 95% LL 95% UL
Chan 2004 Singapore All >18 Health care workers 661 Cross-section study SARS Questionnaire GHQ Direct contact with suspected patients Psychological distress 1.60 1.10 2.50
Chan 2004 Singapore All >18 Health care workers 661 Cross-section study SARS Questionnaire GHQ Direct contact with suspected patients Psychological distress 1.40 1.02 2.00
Tam 2004 China All 33 Health care workers 652 Cross-section study SARS Questionnaire GHQ Direct contact with suspected patients Psychological distress 1.00 0.73 1.38
Verma 2004 Singapore All >18 Health care workers 721 Cross-section study SARS Questionnaire GHQ Direct contact with suspected patients Psychological distress 2.90 1.30 6.30
Sim 2004 Singapore All >18 Health care workers 277 Cross-section study SARS Questionnaire GHQ Direct contact with suspected patients Psychological distress 0.82 0.39 1.72
Sim 2004 Singapore All >18 Health care workers 277 Cross-section study SARS Questionnaire GHQ Direct contact with suspected patients Psychological distress 0.51 0.17 1.50
Leung 2005 China All >18 General population 480 Longitudinal study SARS Random-digit dialing STAI Exposed to epidemics Anxiety 2.63 1.43 4.84
Leung 2005 China All >18 General population 480 Longitudinal study SARS Random-digit dialing STAI Exposed to epidemics Anxiety 2.95 1.56 5.60
Leung 2005 China All >18 General population 272 Longitudinal study SARS Random-digit dialing STAI Exposed to epidemics Anxiety 0.87 0.42 1.81
Leung 2005 China All >18 General population 272 Longitudinal study SARS Random-digit dialing STAI Exposed to epidemics Anxiety 1.04 0.43 2.51
Wu 2009 China All >18 Health care workers 549 Cross-section study SARS Questionnaire IES-R Exposed to epidemics Posttraumatic stress 3.47 1.90 6.20
Wu 2009 China All >18 Health care workers 549 Cross-section study SARS Questionnaire IES-R Relatives or friends infected Posttraumatic stress 3.74 1.80 7.60
Peng 2010 China All ≥18 Health care workers 1278 Cross-section study SARS Random-digit dialing BSRS Perceived epidemics as serious Psychological distress 1.27 0.54 2.98
Cao 2020 China All ≥18 Students 7143 Cross-section study COVID-19 Questionnaire GAD Relatives or friends infected Anxiety 3.01 2.38 3.80
Choi 2020 China All ≥18 General population 500 Cross-section study COVID-19 Questionnaire PHQ Perceived epidemics as serious Depression 1.86 1.37 2.52
Choi 2020 China All ≥18 General population 500 Cross-section study COVID-19 Questionnaire GAD Perceived epidemics as serious Anxiety 1.73 1.25 2.40
Gualano 2020 Italy All ≥18 General population 1515 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 0.99 0.73 1.34
Gualano 2020 Italy All ≥18 General population 1515 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.41 1.04 1.89
Gómez-Salgado 2020 Spain All ≥18 General population 4180 Cross-section study COVID-19 Questionnaire GHQ Direct contact with suspected patients Psychological distress 1.24 1.03 1.50
Gómez-Salgado 2020 Spain All ≥18 General population 4180 Cross-section study COVID-19 Questionnaire GHQ Direct contact with suspected patients Psychological distress 1.20 0.98 1.46
Gómez-Salgado 2020 Spain All ≥18 General population 4180 Cross-section study COVID-19 Questionnaire GHQ Direct contact with suspected patients Psychological distress 1.26 1.05 1.50
Gómez-Salgado 2020 Spain All ≥18 General population 4180 Cross-section study COVID-19 Questionnaire GHQ Relatives or friends infected Psychological distress 1.08 0.88 1.32
Gómez-Salgado 2020 Spain All ≥18 General population 4180 Cross-section study COVID-19 Questionnaire GHQ Relatives or friends infected Psychological distress 1.11 0.69 1.80
Guo 2020 China All ≥18 General population 2441 Cross-section study COVID-19 Questionnaire PTSD Direct contact with suspected patients Posttraumatic stress 1.21 0.92 1.60
Guo 2020 China All ≥18 General population 2441 Cross-section study COVID-19 Questionnaire PTSD Direct contact with suspected patients Depression 1.39 1.08 1.80
Huang 2020 China All ≥18 General population 7236 Cross-section study COVID-19 Online survey GAD Direct contact with suspected patients Anxiety 1.30 0.82 2.08
Huang 2020 China All ≥18 General population 7236 Cross-section study COVID-19 Online survey CES-D Direct contact with suspected patients Depression 1.02 0.58 1.81
Huang 2020 China All ≥18 General population 7236 Cross-section study COVID-19 Online survey GAD Perceived epidemics as serious Anxiety 0.80 0.38 1.69
Huang 2020 China All ≥18 General population 7236 Cross-section study COVID-19 Online survey CES-D Perceived epidemics as serious Depression 1.12 0.42 3.02
Lai 2020 China All 26–40 Health care workers 1257 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 1.52 1.11 2.09
Lai 2020 China All 26–40 Health care workers 1257 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.57 1.22 2.02
Lai 2020 China All 26–40 Health care workers 1257 Cross-section study COVID-19 Questionnaire IES-R Direct contact with suspected patients Psychological distress 1.60 1.25 2.04
Li 2020 China Female ≥19 Health care workers 5317 Cross-section study COVID-19 Questionnaire PHQ Relatives or friends infected Depression 1.39 1.16 1.66
Li 2020 China Female ≥19 Health care workers 5317 Cross-section study COVID-19 Questionnaire GAD Relatives or friends infected Anxiety 1.32 0.94 1.87
Lu 2020 China All ≥18 Health care workers 2299 Cross-section study COVID-19 Questionnaire NRS Direct contact with suspected patients Fear 1.30 0.99 1.73
Lu 2020 China All ≥18 Health care workers 2299 Cross-section study COVID-19 Questionnaire NRS Direct contact with suspected patients Fear 1.41 1.03 1.93
Lu 2020 China All ≥18 Health care workers 2299 Cross-section study COVID-19 Questionnaire HAMA Direct contact with suspected patients Anxiety 1.31 0.89 1.92
Lu 2020 China All ≥18 Health care workers 2299 Cross-section study COVID-19 Questionnaire HAMA Direct contact with suspected patients Anxiety 2.06 1.35 3.15
Lu 2020 China All ≥18 Health care workers 2299 Cross-section study COVID-19 Questionnaire HAMD Direct contact with suspected patients Depression 1.39 0.80 2.43
Lu 2020 China All ≥18 Health care workers 2299 Cross-section study COVID-19 Questionnaire HAMD Direct contact with suspected patients Depression 2.02 1.10 3.69
Xiao 2020 China All ≥17 Students 933 Cross-section study COVID-19 Questionnaire GAD Perceived epidemics as serious Anxiety 1.07 0.90 1.27
Xiao 2020 China All ≥17 Students 933 Cross-section study COVID-19 Questionnaire GAD Perceived epidemics as serious Anxiety 1.01 0.90 1.13
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GPS-PTSD Direct contact with suspected patients Posttraumatic stress 1.37 1.05 1.80
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 1.04 0.76 1.42
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.13 0.80 1.59
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PSS Direct contact with suspected patients Posttraumatic stress 1.16 0.84 1.60
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GPS-PTSD Direct contact with suspected patients Posttraumatic stress 1.12 0.81 1.55
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 1.36 0.95 1.96
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.09 0.74 1.61
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PSS Direct contact with suspected patients Posttraumatic stress 0.74 0.51 1.08
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GPS-PTSD Direct contact with suspected patients Posttraumatic stress 1.20 0.86 1.67
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 0.71 0.48 1.05
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 0.96 0.64 1.44
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PSS Direct contact with suspected patients Posttraumatic stress 0.75 0.51 1.11
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GPS-PTSD Direct contact with suspected patients Posttraumatic stress 1.75 1.03 2.97
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 0.98 0.53 1.82
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.05 0.53 2.08
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PSS Direct contact with suspected patients Posttraumatic stress 1.18 0.66 2.11
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GPS-PTSD Direct contact with suspected patients Posttraumatic stress 1.54 1.10 2.16
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 1.39 0.95 2.03
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.18 0.78 1.77
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PSS Direct contact with suspected patients Posttraumatic stress 1.93 1.30 2.85
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GPS-PTSD Direct contact with suspected patients Posttraumatic stress 1.59 1.21 2.09
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 1.38 1.00 1.90
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.19 0.85 1.67
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PSS Direct contact with suspected patients Posttraumatic stress 1.66 1.19 2.32
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GPS-PTSD Direct contact with suspected patients Posttraumatic stress 1.23 0.93 1.62
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 1.54 1.11 2.14
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.14 0.81 1.62
Rossi 2020 Italy All >18 Health care workers 1379 Cross-section study COVID-19 Questionnaire PSS Direct contact with suspected patients Posttraumatic stress 1.01 0.73 1.41
Shacham 2020 Israel All 24–74 Health care workers 338 Cross-section study COVID-19 Online survey K6 Direct contact with suspected patients Psychological distress 0.91 0.64 1.28
Shacham 2020 Israel All 24–74 Health care workers 338 Cross-section study COVID-19 Online survey K6 Direct contact with suspected patients Psychological distress 2.11 1.24 3.60
Abdessater 2020 French All 29.5 Health care workers 275 Cross-section study COVID-19 Questionnaire GHQ Direct contact with suspected patients Posttraumatic stress 1.85 0.98 3.59
Yang 2020 South Korea All 20–50 Health care workers 65 Cross-section study COVID-19 Questionnaire PHQ Relatives or friends infected Depression 1.52 0.26 20.90
Yang 2020 South Korea All 20–50 Health care workers 65 Cross-section study COVID-19 Questionnaire GAD Relatives or friends infected Anxiety 0.68 0.12 9.19
Ying 2020 China All 37 General population 285 Cross-section study COVID-19 Questionnaire PHQ Direct contact with suspected patients Depression 1.04 0.77 1.42
Ying 2020 China All 37 General population 406 Cross-section study COVID-19 Questionnaire GAD Direct contact with suspected patients Anxiety 1.41 1.05 1.89
Ying 2020 China All 37 General population 47 Cross-section study COVID-19 Questionnaire PHQ Relatives or friends infected Depression 1.20 0.68 2.07
Ying 2020 China All 37 General population 70 Cross-section study COVID-19 Questionnaire GAD Relatives or friends infected Anxiety 1.43 0.84 2.42
Ying 2020 China All 37 General population 112 Cross-section study COVID-19 Questionnaire PHQ Perceived epidemics as serious Depression 0.89 0.51 1.52
Ying 2020 China All 37 General population 112 Cross-section study COVID-19 Questionnaire GAD Perceived epidemics as serious Anxiety 1.15 0.68 1.94
Ying 2020 China All 37 General population 212 Cross-section study COVID-19 Questionnaire PHQ Perceived epidemics as serious Depression 1.54 1.01 2.35
Ying 2020 China All 37 General population 212 Cross-section study COVID-19 Questionnaire GAD Perceived epidemics as serious Anxiety 1.95 1.28 2.96
LL, lower limit; UL, upper limit; GHQ, General Health Questionnaire; PHQ, Patient Health Questionnaire; GAD, Generalized Anxiety Disorder scale; STAI, State Trait Anxiety Inventory; PTSD, posttraumatic stress disorder; IES-R, Impact of Event Scale–Revised; BSRS, five-item Brief Symptom Rating Scale; NRS, Numeric Rating Scale; HAMA, Hamilton Anxiety Scale; HAMD, Hamilton Depression Scale; GPS, Global Psychotrauma Scale; PSS, Perceived Stress Scale; K6, Kessler's K6.
Of 22 eligible articles, eight looked at the relationship between pandemic disease and psychological distress (Chan and Huak, 2004; Gómez-Salgado et al., 2020; Lai et al., 2020; Peng et al., 2010; Shacham et al., 2020; Sim et al., 2004; Tam et al., 2004; Verma et al., 2004), 11 articles related to anxiety (Cao et al., 2020; Choi et al., 2020; Gualano et al., 2020; Huang and Zhao, 2020; Lai et al., 2020; Leung et al., 2005; Li et al., 2020a; Lu et al., 2020; Rossi et al., 2020; Xiao et al., 2020; Yang et al., 2020), nine involved in depression (Choi et al., 2020; Gualano et al., 2020; Guo et al., 2020; Huang and Zhao, 2020; Lai et al., 2020; Li et al., 2020a; Lu et al., 2020; Rossi et al., 2020; Yang et al., 2020), and four studies connected to posttraumatic stress (Abdessater et al., 2020; Guo et al., 2020; Rossi et al., 2020; Wu et al., 2009). In terms of study subjects, 14 concerned the mental health of the health care workers (Abdessater et al., 2020; Chan and Huak, 2004; Gualano et al., 2020; Lai et al., 2020; Li et al., 2020a; Lu et al., 2020; Peng et al., 2010; Rossi et al., 2020; Shacham et al., 2020; Sim et al., 2004; Tam et al., 2004; Verma et al., 2004; Wu et al., 2009; Yang et al., 2020), six adopted general population (Choi et al., 2020; Gómez-Salgado et al., 2020; Guo et al., 2020; Huang and Zhao, 2020; Leung et al., 2005; Ying et al., 2020), and two pointed to college students (Cao et al., 2020; Xiao et al., 2020). Meanwhile, seven studies assessed SARS (Chan and Huak, 2004; Leung et al., 2005; Peng et al., 2010; Sim et al., 2004; Tam et al., 2004; Verma et al., 2004; Wu et al., 2009), and the other 15 regarded COVID-19 (Abdessater et al., 2020; Cao et al., 2020; Choi et al., 2020; Gómez-Salgado et al., 2020; Gualano et al., 2020; Guo et al., 2020; Huang and Zhao, 2020; Lai et al., 2020; Li et al., 2020a; Lu et al., 2020; Rossi et al., 2020; Shacham et al., 2020; Xiao et al., 2020; Yang et al., 2020; Ying et al., 2020). On the basis of different countries, all studies were divided into developed countries (Abdessater et al., 2020; Chan and Huak, 2004; Gómez-Salgado et al., 2020; Gualano et al., 2020; Rossi et al., 2020; Shacham et al., 2020; Sim et al., 2004; Verma et al., 2004; Yang et al., 2020) and developing countries (Cao et al., 2020; Choi et al., 2020; Guo et al., 2020; Huang and Zhao, 2020; Lai et al., 2020; Leung et al., 2005; Li et al., 2020a; Lu et al., 2020; Peng et al., 2010; Tam et al., 2004; Wu et al., 2009; Xiao et al., 2020; Ying et al., 2020). Assessment tools used in the studies were various questionnaires; more details were presented in Table 1. As to the assessment of different kinds of exposures, we excluded the diverse ways of expression from the initial studies and concluded them to similar expression of unity. For example, we treated “worried about being infected by COVID-19” and “concerns about the COVID-19 pandemic,” “time to think about COVID-19 per day (hours) (2–3 hours),” and “time to think about COVID-19 per day (hours) (>3 hours)” as “perceived pandemic as serious,” and viewed the expressions such as “working frontline,” “high-risk contact,” and “low-risk contact” in health care workers as “direct contact with suspect or probable patients.” All of the different kinds of exposures were seen as the psychological impact of the pandemic.
Quality Assessment
Table 2 shows the quality assessment of all eligible articles by using the AHRQ cross-sectional study evaluation criteria. The average total score was 5.55 (range, 3 to 8).
TABLE 2 AHRQ Cross-Sectional Study Evaluation Criteria
Year First Author 1 2 3 4 5 6 7 8 9 10 11 Total
2004 Chan 1 1 1 0 0 0 0 1 0 0 0 4
2004 Sim 1 1 1 0 0 0 0 1 0 1 0 5
2004 Tam 1 1 1 0 0 0 0 1 0 1 0 5
2004 Verma 1 1 1 0 0 0 1 1 1 1 0 7
2005 Leung 1 1 1 1 0 0 0 1 0 1 1 7
2009 Wu 1 1 1 0 0 0 0 1 0 0 0 4
2010 Peng 1 1 1 1 0 0 1 1 1 1 0 8
2020 Abdessater 1 1 1 0 0 0 0 1 0 1 0 5
2020 Cao 1 1 1 0 0 0 0 1 0 1 0 5
2020 Choi 1 1 1 1 0 0 1 1 0 1 0 7
2020 Gómez-Salgado 1 1 1 0 0 0 1 1 1 1 0 7
2020 Gualano 1 1 1 1 0 0 1 0 1 1 0 7
2020 Guo 1 1 1 1 0 0 1 1 1 1 0 8
2020 Huang 1 1 1 1 0 0 1 1 0 1 0 7
2020 Lai 1 1 1 1 0 0 0 1 0 1 0 6
2020 Li 1 1 1 0 0 0 0 1 0 1 0 5
2020 Lu 1 1 1 0 0 0 1 1 0 1 0 6
2020 Rossi 1 1 1 1 0 0 1 1 0 1 0 7
2020 Shacham 1 1 1 0 0 0 0 1 0 0 0 4
2020 Xiao 1 1 1 0 0 0 0 1 0 1 0 5
2020 Yang, S 1 1 1 0 0 0 0 0 0 0 0 3
2020 Yang, Y 1 1 1 1 0 0 1 0 1 1 0 7
Notes: 1) define the source of information (survey, record review); 2) list inclusion and exclusion criteria for exposed and unexposed subjects (cases and controls) or refer to previous publications; 3) indicate period used for identifying patients; 4) indicate whether or not subjects were consecutive if not population-based; 5) indicate if evaluators of subjective components of study were masked to other aspects of the status of the participants; 6) describe any assessments undertaken for quality assurance purposes (e.g., test/retest of primary outcome measurements); 7) explain any patient exclusions from analysis; 8) describe how confounding was assessed and/or controlled; 9) if applicable, explain how missing data were handled in the analysis; 10) summarize patient response rates and completeness of data collection; 11) clarify what follow-up, if any, was expected and the percentage of patients for which incomplete data or follow-up was obtained.
Overall Analyses
A statistically significant association of exposing to pandemic with humans' mental health was observed based on the overall analysis (OR, 1.32; 95% CI, 1.24–1.40; p < 0.001; I2, 62.0%) (Table 3), which was calculated by random-effects model (p < 0.05) with between-study heterogeneity. Four groups of mental disorders were analyzed separately, including anxiety (OR, 1.37; 95% CI, 1.19–1.58; p < 0.001; I2, 75.7%), depression (OR, 1.28; 95% CI, 1.15–1.42; p < 0.001; I2, 35.2%), posttraumatic stress (OR, 1.36; 95% CI, 1.17–1.58; p < 0.001; I2, 66.6%), and psychological distress (OR, 1.25; 95% CI, 1.11–1.40; p < 0.001; I2, 44.2%) (Fig. 2).
TABLE 3 Overall and Subgroup Analyses on the Association of Exposure to Pandemics With Human Mental Health
Group No. Qualified Studies All Types Anxiety Depression Posttraumatic Stress Psychological Distress
OR (95% CI); P I 2 OR (95% CI); P I 2 OR (95% CI); P I 2 OR (95% CI); P I 2 OR (95% CI); P I 2
Overall analyses
Mental disorders 83/27/21/18/15 1.32 (1.24–1.40); <0.001 62% 1.37 (1.19–1.58); <0.001 75.7% 1.28 (1.15–1.42); <0.001 35.2% 1.36 (1.17–1.58); <0.001 66.6% 1.25 (1.11–1.40); <0.001 44.2%
Subgroup analyses
By source of participants
General population 27/12/9/*/15 1.30 (1.19–1.42); <0.001 36.0% 1.50 (1.26–1.78); <0.001 35.5% 1.24 (1.04–1.48); 0.017 42.1% 1.21 (0.92–1.60); 0.177 * 1.20 (1.09–1.31); <0.001 0.0%
Health care workers 53/12/12/7/10 1.31 (1.22–1.41); <0.001 48.9% 1.27 (1.13–1.43); <0.001 7.7% 1.30 (1.14–1.49); <0.001 34.6% 1.38 (1.17–1.62); <0.001 68.4% 1.32 (1.05–1.65); 0.019 59.1%
Students 3/3/*/* 1.47 (0.83–2.60); 0.186 97.1% 1.47 (0.83–2.60); 0.186 97.1% * * * * * *
By country development
Developing countries 38 1.46 (1.30–1.63); <0.001 73.3% * * * * * * * *
Developed countries 45 1.22 (1.14–1.30); <0.001 36.2% * * * * * * * *
By type of pandemics
SARS 13/4/*/2/7 1.61 (1.20–2.17); 0.002 69.9% 1.70 (0.92–3.14); <0.091 66.9% * * 3.58 (2.27–5.65); <0.001 0.0% 1.26 (0.94–1.68); 0.118 49.7%
COVID-19 70/23/21/16/8 1.29 (1.21–1.38); <0.001 59.6% 1.34 (1.16–1.54); <0.001 76.5% 1.28 (1.15–1.42); <0.001 35.2% 1.27 (1.11–1.44); <0.001 52.8% 1.24 (1.10–1.40); 0.001 46.8%
By kinds of exposure
Direct contact with suspect patients 57 1.27 (1.20–1.34); <0.001 37.1% * * * * * * * *
Exposed to epidemic 5 1.99 (1.17–3.39); 0.012 67.7% * * * * * * * *
Relatives or friends infected 10 1.54 (1.13–2.09); 0.006 83.4% * * * * * * * *
Perceived epidemics as serious 11 1.29 (1.07–1.56); 0.007 67.7% * * * * * * * *
*Data are unavailable.
FIGURE 2 Forest plot presenting the association between the emerging outbreak of infection disease with common mental disorders among people in studies providing OR and 95% CI.
Cumulative and Influential Analyses
In cumulative analysis, the first two published studies by Sim and Tam made in 2004 concluded the pandemic as a protective factor for human' mental disorder; since then, other studies all got completely opposite conclusions consistently, and the trend tended to stabilize. The influential analyses revealed no significant impact of any single studies on overall effect-size estimates.
Publication Bias
Figure 3 shows the Begg's and filled funnel plots on the association of pandemic disease with mental health. The overall analysis of pandemic disease revealed that no publication bias relies on Egger's test (p = 0.14). Similarly, there were evidences of symmetry of study effects in terms of anxiety (p = 0.23), depression (p = 0.47), posttraumatic stress (p = 0.11), and psychological distress (p = 0.84).
FIGURE 3 The Begg's and filled funnel plots of the association of pandemic disease with mental health.
Further investigation using the “trim-and-fill” method produced that there was one theoretically missing study aforementioned required for a further test of symmetry in overall analysis. Meanwhile, three missing studies were requested for both comparisons to make the Begg's funnel plots symmetrical in posttraumatic stress. After adjusting, the ORs for influencing mental health and getting posttraumatic stress were 1.31 (95% CI, 1.23–1.40; p < 0.001) and 1.24 (95% CI, 1.05–1.46; p = 0.01), respectively. Concerning the group of having anxiety, depression, and psychological distress, they did not produce any correction to the original estimates.
Subgroup Analyses
Between-study heterogeneity existed in the overall analysis for exposing to pandemic associated with human's mental disorder (I2 = 62.0%). A series of prespecified subgroup analysis were done to explore possible sources of between-study heterogeneity (Table 3). By major exposure subjects, in general population, there was a remarkably significant association of exposing to pandemic with mental disorder (OR, 1.30; 95% CI, 1.19–1.42; p < 0.001), including anxiety (OR, 1.50; 95% CI, 1.26–1.78; p < 0.001), depression (OR, 1.24; 95% CI, 1.04–1.48; p = 0.017), posttraumatic stress (OR, 1.21; 95% CI, 0.92–1.60; p = 0.177), and psychological distress (OR, 1.20; 95% CI, 1.09–1.31; p < 0.001). In health care workers, synthetic analysis demonstrated the risk magnitude, and the OR was 1.31 (95% CI, 1.22–1.41; p < 0.001), containing anxiety (OR, 1.27; 95% CI, 1.13–1.43; p < 0.001), depression (OR, 1.30; 95% CI, 1.14–1.49; p < 0.001), posttraumatic stress (OR, 1.38; 95% CI, 1.17–1.62; p < 0.001), and psychological distress (OR, 1.32; 95% CI, 1.05–1.65; p = 0.019). At the same time, no detectable significance was observed in students, although the OR was 1.47 (95% CI, 0.83–2.60; p = 0.186).
By different levels of development, developing countries were illustrated to have a high risk of experiencing mental disorder (OR, 1.46; 95% CI, 1.30–1.63; p < 0.001), and developed countries satisfied this relationship as well (OR, 1.22; 95% CI, 1.14–1.30; p < 0.001).
In terms of different infectious diseases, there was a remarkably significant association of SARS with mental disorder (OR, 1.61; 95% CI, 1.20–2.17; p = 0.002), and it was the same in COVID-19 (OR, 1.29; 95% CI, 1.21–1.38; p < 0.001).
With regard to different kinds of exposure, the OR of subjects who directly contacted with suspected patients was 1.27 (95% CI, 1.20–1.34; p < 0.001), that of those who were exposed to pandemic was 1.99 (95% CI, 1.17–3.39; p = 0.012), and that of those whose relatives or friends were infected was 1.54 (95% CI, 1.13–2.09; p = 0.006). Meanwhile, those who perceived a pandemic as serious also experienced high risk of mental disorder (OR, 1.29; 95% CI, 1.07–1.56; p = 0.007).
DISCUSSION
To the best of our knowledge, this is thus far the most comprehensive meta-analysis that has explored the relationship between pandemic exposure and mental disorder by using the OR as the analytical indicator. Our key findings indicated that people who experienced a pandemic could increase approximately 1.32 times the risk of having psychological problems, including 1.37 times the risk of getting anxiety, 1.28 times the risk of getting depression, 1.36 times the risk of having posttraumatic stress, and 1.25 times the risk of having psychological distress. Moreover, our subsidiary analyses demonstrated that different major exposure subjects, countries' level of development, infectious disease, and kinds of exposure were possible sources of between-study heterogeneity. Comprehensive evaluation made by our study focused on the magnitude of the outbreak of pandemic, and we found the related risk factors including contacted with suspect patients directly, exposed to the pandemic, whose relatives or friends infected, and perceived pandemic as serious. We highlighted the importance and the necessity for a sustained, efficient mental health care delivery along with a pandemic.
Pandemics, with filled unpredictability and uncertainty, had posed numerous and unprecedented challenges and threats worldwide during each outbreak. Although much of the early scholarly work had focused on intensive, emergency, and primary care, a long time of searching precautionary measures had proposed that the role of mental health clinicians was key on multiple levels. A recent review concluded that negative psychological effects caused by pandemic were widespread and stressors were also diversified, including longer quarantine duration, infection fears, frustration, boredom, inadequate supplies, inadequate information, financial loss, and stigma (Brooks et al., 2020). Simultaneously, evidence showed that these maladaptive reactions can be long-lasting, as the study illustrated that mental disorder, such as posttraumatic and depressive symptoms, and general psychological distress were reported after periods ranging from 6 months up to 3 years after the pandemic outbreak (Liu et al., 2012; Maunder et al., 2006). Long-term behavioral changes such as vigilant handwashing and avoidance of crowds were described in a qualitative study among general population; for some, the return to normality was delayed by many months (Cava et al., 2005). The same is true of health care workers. Researchers found that alcohol abuse or dependency symptoms were positively related to quarantined health care workers even 3 years after the SARS outbreak (Wu et al., 2008).
In this sense, public health should not only increase mental health literacy but provide clear and concise information about infection rates and risk of infection to reduce uncertainty. Our meta-analysis was based on researches involved in people at high or low risk of exposure to a pandemic, with such OR to describe the risk of getting mental disorder, not relied on the prevalence of stress, anxiety, and depression in the general population after the pandemic. We got the statistically significant association of exposing to pandemic with humans' mental health, especially for health care workers. Studies investigated the possible work-related features connected to mental health outcomes, including working in high-risk units, contacting affected patients directly, being quarantined, having relatives or friends get infected, sharper at disease-related risk perception (Preti et al., 2020). These were consistent with our causal analysis.
There were several limitations in this study. First of all, this meta-analysis only included SARS and COVID-19; influenza caused by the virus subtype H1N1, MERS, and Ebola virus disease were not concluded, which limited the representativeness of this study. Second, the degree of exposures was ranked from least to most in the original articles, with different assessments of exposures; the present secondary analysis could not avoid the fuzzy definition of control groups. Meanwhile, all of the studies in our analysis were cross-section studies, which could reflect the psychological state of the population over a period. Hence, more studies on account of a longer and more forward-looking period observation can be helpful in further identification of mental health status.
CONCLUSIONS
In the context of infectious disease outbreaks, the mental health of general populations is clearly vulnerable. Both the general population and health care workers all experienced a high risk of developing mental health problems. Therefore, all of us need urgent attention and psychological counseling to overcome pandemic together.
DISCLOSURE
The authors declare that they have no conflict of interest.
Ethics approval and consent to participate were received by each involved study in this meta-analysis.
The data sets used and/or analyzed during the current meta-analysis are available from the corresponding author upon reasonable request.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
All authors read and approved the final manuscript before submission. W.N. and Z.Z. conceived and designed the experiments. X.D., J.H., and M.H. performed the experiments. X.D., M.H., J.Z., and W.N. analyzed the data. X.D., M.H., J.Z., M.L., and Z.Z. contributed the materials/analysis tools. X.D. and W.N. wrote the article.
X.D., M.H., and J.Z. shared first authors.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.jonmd.com).
==== Refs
REFERENCES
Abdessater M Rouprêt M Misrai V Matillon X Gondran-Tellier B Freton L Vallée M Dominique I Felber M Khene ZE Fortier E Lannes F Michiels C Grevez T Szabla N Boustany J Bardet F Kaulanjan K Seizilles de Mazancourt E Ploussard G Pinar U Pradere B (2020) COVID19 pandemic impacts on anxiety of French urologist in training: Outcomes from a national survey. Prog Urol. 30 :448–455.32376208
Brooks SK Webster RK Smith LE Woodland L Wessely S Greenberg N Rubin GJ (2020) The psychological impact of quarantine and how to reduce it: Rapid review of the evidence. Lancet. 395 :912–920.32112714
Cao W Fang Z Hou G Han M Xu X Dong J Zheng J (2020) The psychological impact of the COVID-19 epidemic on college students in China. Psychiatry Res. 287 :112934.32229390
Cava MA Fay KE Beanlands HJ McCay EA Wignall R (2005) The experience of quarantine for individuals affected by SARS in Toronto. Public Health Nurs. 22 :398–406.16229732
Chan AO Huak CY (2004) Psychological impact of the 2003 severe acute respiratory syndrome outbreak on health care workers in a medium size regional general hospital in Singapore. Occup Med (Lond). 54 :190–196.15133143
Choi EPH Hui BPH Wan EYF (2020) Depression and anxiety in Hong Kong during COVID-19. Int J Environ Res Public Health. 17 :3740.32466251
Fisman DN Laupland KB (2009) Influenza mixes its pitches: Lessons learned to date from the influenza a (H1N1) pandemic. Can J Infect Dis Med Microbiol. 20 :89–91.20808467
Gómez-Salgado J Andres-Villas M Dominguez-Salas S Diaz-Milanes D Ruiz-Frutos C (2020) Related health factors of psychological distress during the COVID-19 pandemic in Spain. Int J Environ Res Public Health. 17 :3947.32498401
Gualano MR Lo Moro G Voglino G Bert F Siliquini R (2020) Effects of Covid-19 lockdown on mental health and sleep disturbances in Italy. Int J Environ Res Public Health. 17 :4779.32630821
Guo J Feng XL Wang XH van IMH (2020) Coping with COVID-19: Exposure to COVID-19 and negative impact on livelihood predict elevated mental health problems in Chinese adults. Int J Environ Res Public Health. 17 :3857.32485859
Holmes EA O'Connor RC Perry VH Tracey I Wessely S Arseneault L Ballard C Christensen H Cohen Silver R Everall I Ford T John A Kabir T King K Madan I Michie S Przybylski AK Shafran R Sweeney A Worthman CM Yardley L Cowan K Cope C Hotopf M Bullmore E (2020) Multidisciplinary research priorities for the COVID-19 pandemic: A call for action for mental health science. Lancet Psychiatry. 7 :547–560.32304649
Huang Y Zhao N (2020) Generalized anxiety disorder, depressive symptoms and sleep quality during COVID-19 outbreak in China: A web-based cross-sectional survey. Psychiatry Res. 288 :112954.32325383
Klenk HD (2014) Lessons to be learned from the ebolavirus outbreak in West Africa. Emerg Microbes Infect. 3 :e61.26038754
Krishnamoorthy Y Nagarajan R Saya GK Menon V (2020) Prevalence of psychological morbidities among general population, healthcare workers and COVID-19 patients amidst the COVID-19 pandemic: A systematic review and meta-analysis. Psychiatry Res. 293 :113382.32829073
Lai J Ma S Wang Y Cai Z Hu J Wei N Wu J Du H Chen T Li R Tan H Kang L Yao L Huang M Wang H Wang G Liu Z Hu S (2020) Factors associated with mental health outcomes among health care workers exposed to coronavirus disease 2019. JAMA Netw Open. 3 :e203976.32202646
Lam CW Chan MH Wong CK (2004) Severe acute respiratory syndrome: Clinical and laboratory manifestations. Clin Biochem Rev. 25 :121–132.18458712
Leung GM Ho LM Chan SK Ho SY Bacon-Shone J Choy RY Hedley AJ Lam TH Fielding R (2005) Longitudinal assessment of community psychobehavioral responses during and after the 2003 outbreak of severe acute respiratory syndrome in Hong Kong. Clin Infect Dis. 40 :1713–1720.15909256
Li G Miao J Wang H Xu S Sun W Fan Y Zhang C Zhu S Zhu Z Wang W (2020a) Psychological impact on women health workers involved in COVID-19 outbreak in Wuhan: A cross-sectional study. J Neurol Neurosurg Psychiatry. 91 :895–897.32366684
Li Q Guan X Wu P Wang X Zhou L Tong Y Ren R Leung KSM Lau EHY Wong JY Xing X Xiang N Wu Y Li C Chen Q Li D Liu T Zhao J Liu M Tu W Chen C Jin L Yang R Wang Q Zhou S Wang R Liu H Luo Y Liu Y Shao G Li H Tao Z Yang Y Deng Z Liu B Ma Z Zhang Y Shi G Lam TTY Wu JT Gao GF Cowling BJ Yang B Leung GM Feng Z (2020b) Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia. N Engl J Med. 382 :1199–1207.31995857
Liu X Kakade M Fuller CJ Fan B Fang Y Kong J Guan Z Wu P (2012) Depression after exposure to stressful events: Lessons learned from the severe acute respiratory syndrome epidemic. Compr Psychiatry. 53 :15–23.21489421
Lu W Wang H Lin Y Li L (2020) Psychological status of medical workforce during the COVID-19 pandemic: A cross-sectional study. Psychiatry Res. 288 :112936.32276196
Maunder RG Lancee WJ Balderson KE Bennett JP Borgundvaag B Evans S Fernandes CM Goldbloom DS Gupta M Hunter JJ McGillis Hall L Nagle LM Pain C Peczeniuk SS Raymond G Read N Rourke SB Steinberg RJ Stewart TE VanDeVelde-Coke S Veldhorst GG Wasylenki DA (2006) Long-term psychological and occupational effects of providing hospital healthcare during SARS outbreak. Emerg Infect Dis. 12 :1924–1932.17326946
Nickell LA Crighton EJ Tracy CS Al-Enazy H Bolaji Y Hanjrah S Hussain A Makhlouf S Upshur RE (2004) Psychosocial effects of SARS on hospital staff: Survey of a large tertiary care institution. CMAJ. 170 :793–798.14993174
Norris FH Friedman MJ Watson PJ Byrne CM Diaz E Kaniasty K (2002) 60,000 disaster victims speak: Part I. An empirical review of the empirical literature, 1981–2001. Psychiatry. 65 :207–239.12405079
Page MJ McKenzie JE Bossuyt PM Boutron I Hoffmann TC Mulrow CD Shamseer L Tetzlaff JM Akl EA Brennan SE Chou R Glanville J Grimshaw JM Hrobjartsson A Lalu MM Li T Loder EW Mayo-Wilson E McDonald S McGuinness LA Stewart LA Thomas J Tricco AC Welch VA Whiting P Moher D (2021) The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ. 372 :n71.33782057
Peng EY Lee MB Tsai ST Yang CC Morisky DE Tsai LT Weng YL Lyu SY (2010) Population-based post-crisis psychological distress: An example from the SARS outbreak in Taiwan. J Formos Med Assoc. 109 :524–532.20654792
Preti E Di Mattei V Perego G Ferrari F Mazzetti M Taranto P Di Pierro R Madeddu F Calati R (2020) The psychological impact of epidemic and pandemic outbreaks on healthcare workers: Rapid review of the evidence. Curr Psychiatry Rep. 22 :43.32651717
Rossi R Socci V Pacitti F Di Lorenzo G Di Marco A Siracusano A Rossi A (2020) Mental health outcomes among frontline and second-line health care workers during the coronavirus disease 2019 (COVID-19) pandemic in Italy. JAMA Netw Open. 3 :e2010185.32463467
Rostom A Dubé C Cranney A Saloojee N Sy R Garritty C Sampson M Zhang L Yazdi F Mamaladze V Pan I McNeil J Moher D Mack D Patel D (2004) Celiac disease. Rockville, MD: Agency for Healthcare Research and Quality (US) Evidence Reports/Technology Assessments, No. 104. Available at: https://www.ncbi.nlm.nih.gov/books/NBK35149/. Accessed September 15, 2020.
Shacham M Hamama-Raz Y Kolerman R Mijiritsky O Ben-Ezra M Mijiritsky E (2020) COVID-19 factors and psychological factors associated with elevated psychological distress among dentists and dental hygienists in Israel. Int J Environ Res Public Health. 17 :2900.32331401
Sim K Chong PN Chan YH Soon WS (2004) Severe acute respiratory syndrome-related psychiatric and posttraumatic morbidities and coping responses in medical staff within a primary health care setting in Singapore. J Clin Psychiatry. 65 :1120–1127.15323599
Tam CW Pang EP Lam LC Chiu HF (2004) Severe acute respiratory syndrome (SARS) in Hong Kong in 2003: Stress and psychological impact among frontline healthcare workers. Psychol Med. 34 :1197–1204.15697046
Tsang HW Scudds RJ Chan EY (2004) Psychosocial impact of SARS. Emerg Infect Dis. 10 :1326–1327.15338536
Verma S Mythily S Chan YH Deslypere JP Teo EK Chong SA (2004) Post-SARS psychological morbidity and stigma among general practitioners and traditional Chinese medicine practitioners in Singapore. Ann Acad Med Singapore. 33 :743–748.15608831
Wu P Fang Y Guan Z Fan B Kong J Yao Z Liu X Fuller CJ Susser E Lu J Hoven CW (2009) The psychological impact of the SARS epidemic on hospital employees in China: Exposure, risk perception, and altruistic acceptance of risk. Can J Psychiatry. 54 :302–311.19497162
Wu P Liu X Fang Y Fan B Fuller CJ Guan Z Yao Z Kong J Lu J Litvak IJ (2008) Alcohol abuse/dependence symptoms among hospital employees exposed to a SARS outbreak. Alcohol Alcoholism (Oxford, Oxfordshire). 43 :706–712.
Xiao H Shu W Li M Li Z Tao F Wu X Yu Y Meng H Vermund SH Hu Y (2020) Social distancing among medical students during the 2019 coronavirus disease pandemic in China: Disease awareness, anxiety disorder, depression, and behavioral activities. Int J Environ Res Public Health. 17 :5047.32674285
Yang S Kwak SG Ko EJ Chang MC (2020) The mental health burden of the COVID-19 pandemic on physical therapists. Int J Environ Res Public Health. 17 :3723.32466164
Ying Y Ruan L Kong F Zhu B Ji Y Lou Z (2020) Mental health status among family members of health care workers in Ningbo, China, during the coronavirus disease 2019 (COVID-19) outbreak: A cross-sectional study. BMC Psychiatry. 20 :379.32680478
| 35687723 | PMC9741999 | NO-CC CODE | 2022-12-15 00:01:29 | no | J Nerv Ment Dis. 2022 Dec 18; 210(12):900-911 | utf-8 | J Nerv Ment Dis | 2,022 | 10.1097/NMD.0000000000001543 | oa_other |
==== Front
J Nerv Ment Dis
J Nerv Ment Dis
JNMD
The Journal of Nervous and Mental Disease
0022-3018
1539-736X
Lippincott Williams & Wilkins
36206312
JNMD_210172
10.1097/NMD.0000000000001464
00004
3
Original Articles
Evaluation of Early Cognitive Functions in Patients With COVID-19 Infection
Yildirim Gulfem PhD
Chest Disease Department, Medicana Hospital, KTO Karatay University School of Medicine, Konya, Turkey.
Send reprint requests to Gulfem Yildirim, PhD, Chest Disease Department, Medicana Hospital, KTO Karatay University School of Medicine, Musalla Bagları Mah. Gurz St, No. 1 Selcuklu, Konya, Turkey. E-mail: [email protected].
12 2022
7 10 2022
7 10 2022
210 12 912914
Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.
2022
Wolters Kluwer Health, Inc. All rights reserved.
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Abstract
In December 2019, some pneumonia cases emerged in Wuhan, China. It was named as Coronavirus 2019 (COVID-19) by the World Health Organization.
Patients developed anxiety and sleep problems after treatment. Patients with confirmed COVID-19 (n = 57) admitted to our study and whose treatment was completed 3 months ago were included in the study. This is a case-control study, and 22 patients included the control group. We found statistical significance between the average score of Beck anxiety and Pittsburgh Sleep Quality Index (p = 0.03, p = 0.01).
In our study, we investigated the psychological conditions that may occur in the postacute COVID-19 period. We recommend that patients should be directed to appropriate clinics for rehabilitation. Clinicians must be aware that prompt and correct diagnosis with careful management is essential for recovery.
Key Words
COVID-19
cognitive functions
depression
==== Body
pmcIn December 2019, some unknown pneumonia cases emerged in Wuhan, Hubei, China. Patients' clinical presentations were resembling viral pneumonia. It was named as coronavirus 2019 (COVID-19) by the World Health Organization (WHO) (Lai et al., 2020). The worldwide spread of COVID-19 led the WHO to declare the COVID-19 pandemic on March 11, 2020.
Patients diagnosed with this disease must be treated in isolation. Furthermore, many patients developed anxiety and sleep problems after isolation treatment. Developing anxiety caused decreasing immunity and other physiological events (Rajeswari and SanjeevaReddy, 2019). Studies suggest that patients with COVID-19 experience depression, anxiety, and insomnia (Rogers et al., 2020). Except for isolation treatment, the uncertainty of the future, stigma, and traumatic memories of severe illness experienced by patients during COVID-19 are significant psychological stressors of psychopathological outcome (Brooks et al., 2020; Carvalho et al., 2020).
On the other hand, coronaviruses can affect the central nervous system (CNS) directly or indirectly via an immune response (Wu et al., 2020). Postmortem studies showed that coronaviruses are neurotropic and can cause neurological diseases (Desforges et al., 2019). We consider that postacute COVID-19 patients are likely to have highly psychiatric conditions. In our study, we evaluated factors such as anxiety, depression, and sleep quality disorder in COVID-19 patients, as well as their relationship with indirect cognition, and directly examine the relationship of COVID-19 with cognition.
METHODS
Design and Participants
This study is approved by the local ethics committee.
Patients with confirmed COVID-19 (n = 57) were admitted to Medicana Hospital, and whose treatment was completed 3 months ago were included in the study between November 2020 and March 2021. This study is a case-control study, and 22 patients were included the control group. Patients were informed about the study, and written informed consent was obtained from all patients.
Data Collection and Statistical Analysis
The study population included patients aged 18 or older with a positive diagnosis of COVID-19. Thirty-one patients were men, and 26 patients were women. For COVID-19–positive diagnosis, we used PCR-positive. Sociodemographic and clinical data were collected using a data extraction form, including age, sex, psychiatric history, and baseline inflammatory markers. Baseline inflammatory markers during acute COVID-19 were extracted from electronic documents charts: C-reactive protein (CRP), neutrophil and lymphocyte counts, d-dimer levels, and ferritin levels.
Current sleep quality and psychological assessments were measured using the following questionnaires: Beck Anxiety Inventory (a total score of 0–7 is interpreted as a “minimal” level of anxiety, 8–15 as “mild,” 16–25 as “moderate,” and 26–63 as “severe”), Beck Depression Inventory (scores from 0 through 9 indicate no or minimal depression; scores from 10 through 18 indicate mild to moderate depression; scores from 19 through 29 indicate moderate to severe depression; and scores from 30 through 63 indicate severe depression), Pittsburgh Sleep Quality Index (score of five or more indicates poor sleep quality; the higher the score, the worse the quality), and Montreal Cognitive Assessment (cutoff point is 26).
Statistical analyses were performed using SPSS software (Version 22 for Windows; SPSS Inc, Chicago, IL). Descriptive analysis was conducted on the basic characteristics of residents by frequency and composition ratio, and compared their differences between basic situation and mental health statuses among various demographics by Mann-Whitney U-test and correlation analysis. The data were tested for normal distribution. A value of p < 0.05 (two-tailed) was considered statistically significant.
RESULTS
All patients completed the study without data leakage. The demographic and clinical characteristics of the 57 patients (53.94 years; 31 males and 26 females) and 22 healthy controls (54.27 years; 9 males and 13 females) are outlined in Table 1. When the demographic data of the patients were evaluated, we found no significant difference between the control group and the patient groups in terms of age, sex, smoking, and education.
TABLE 1 Demographic Features of Patients
Control Group, n = 22 COVID-19 Patients, n = 57 p
Age 54.273 ± 13.932 53.946 ± 16.662 0.662
Sex 9 female, 13 male 26 female, 31 male 0.409
Smoking 5/22 14/57 0.733
Education 9.364 ± 4.685 9.947 ± 4.505 0.618
In the patient group, B-anxiety and B-depression mean scores were 12.78 ± 10.12 and 10.87 ± 10.67, respectively. The average score of Beck depression in the two groups was not statistically significant (p = 0.18). We found a statistical significance between the average score of Beck anxiety and Pittsburgh Sleep Quality Index (p = 0.03, p = 0.01). No significant difference was observed between the groups in terms of cognition (p = 0.908). Our test results are outlined in Table 2. No significant difference was observed when the correlation analysis of Montreal Cognitive Assessment (MoCA) value with inflammatory markers was performed in patients (p > 0.05 for all inflammatory markers).
TABLE 2 COVID-19 Infection's Effect on Sleep, Mood, and Cognition
Control Group COVID-19 Patients p
Beck Anxiety Inventory 7.409 ± 4.2612 12.789 ± 10.1291 0.031
Beck Depression Inventory 6.864 ± 4.6115 10.877 ± 10.6772 0.183
Pittsburgh Sleep Quality Index 3.909 ± 2.1137 6.561 ± 4.3178 0.010
MoCA 19.091 ± 5.0982 18.930 ± 6.2873 0.908
DISCUSSION
In our study, we evaluated anxiety, depression, sleep disorders, and cognitive dysfunctions in patients with postacute COVID-19 patients. As a result, we found a statistically significant increase in anxiety and sleep disorders compared with the normal population. In our study, we found no significant change in cognitive dysfunction between the control and the patient groups. Considering that COVID-19 has systemic effects rather than a disease that only involves the lungs, we think that patients may have anxiety and sleep disorders, and it will be beneficial to arrange their treatment in this direction in the early period.
Studies have shown that coronavirus diseases not only affect physical health but also human psychology. For example, during the SARS pandemic, anxiety and depression increased and sleep was affected in the general population. Some previous studies showed that respiratory infections and tuberculosis assumed disability in daily life, respectively (GBD 2017 Disease and Injury Incidence and Prevalence Collaborators, 2018; Murray et al., 2018).
In patients with COVID-19, older age and severe disease were major risk factors of disability and mortality (Zhou et al., 2020a). However, anxiety and sleep disturbances that develop in patients cause prolonged hospital stay and negatively affect mortality. In a meta-analysis, it was found that many COVID-19 patients experience depression, anxiety, and/or sleep disturbances. The cause of these psychiatric disorders is multifactorial. One major factor is a lack of contact with families (Dubey et al., 2020). Others are isolation and stigma of disease. These problems continue in the posttreatment period (Guan et al., 2020; Zhou et al., 2020a). Therefore, we investigated the psychological conditions that may continue in post-COVID period COVID-19 patients. We found a significant increase in sleep disturbance and anxiety in the follow-up after COVID-19.
Clinical symptoms are associated with inflammatory factor storms and elevated CRP, ferritin, d-dimer, fibrinogen, and lactate dehydrogenase levels in patients with COVID-19 (Zeng et al., 2020). In our study, we found that the aforementioned markers increased in the acute phase in most of our patients. COVID-19 changes peripheral immune cells and elicits a cascade of immune responses, finally resulting in a damaging cytokine storm (Yang et al., 2020). Several studies have shown that a critical role of inflammation was involved in the pathological process of mild cognitive impairment (Shen et al., 2020). However, we evaluated cognitive functions with the MOCA test in the post-COVID period, and we did not find any significant cognitive dysfunction in the patients. Also, we did not find a relationship between inflammatory markers and MOCA test.
Anxiety and depression may have a negative impact on patient immune system response (Leonard, 2001). Patients with depression may also have negative attitudes toward antiviral therapy, which may reduce their treatment adherence and recovery. The prevalence of depression among patients with SARS was 18%, 1 month after their discharge (Wu et al., 2005), and 15.6% at 30-month assessment after SARS outbreak (Mak et al., 2009). A new study in an isolation unit of a general hospital that examined the mental health status of 106 patients with COVID-19 early stage of the outbreak found a prevalence of depression of 49.06% (Zhao et al., 2020). In our study, we did not find a significant increase in depression in the postacute period in patients with COVID-19.
Cognitive dysfunction in patients with viral infection has been commonly reported in prior studies. Moreover, viral infection that involves the CNS and cardiopulmonary failure may be associated with neurologic sequelae, delayed neurodevelopment, and reduced cognitive functioning (Chang et al., 2007). In a recent study, it was found that there is a significant correlation between continuous attention function changes at presentation and CRP levels in COVID-19 patients (Zhou et al., 2020b).
Previous studies on mental health in the COVID-19 pandemic have been conducted in the general population and health professionals but not in the patient population. Our study is different from the others in terms of evaluating anxiety, depression, and sleep disorders in postacute stage COVID-19 patients. In our study, we found that the prevalence of anxiety in outpatients with COVID-19 was 48%, which is a substantial increase compared with prepandemic anxiety levels. On the other hand, we used MoCA test to evaluate the cognitive functions of the patients. Although no significant loss of cognition was observed in the post-COVID period, the mean MoCA level was found to be low because elderly patients were included in our control and patient groups.
CONCLUSIONS
Following the acute phase, health care services are planning rehabilitation strategies. Postacute COVID-19 refers to persistent symptoms 3 weeks after COVID-19 infection, whereas “chronic COVID” describes symptoms lasting more than 12 weeks (Halpin et al., 2021). COVID-19 will cause longer-term mental and physical health problems, work disability, and reduced quality of life in survivors. Because of that, the number of patients needing follow-up care and rehabilitation due to COVID-19 will be unprecedented. In our study, we investigated the psychological conditions that may occur in the postacute COVID-19 period. According to the results of our study, we recommend that patients should be directed to appropriate clinics for rehabilitation. Clinicians must be aware that prompt and correct diagnosis with careful management is essential for recovery.
ACKNOWLEDGMENT
I would like to express my gratitude to Dr. Faik Ilik for his help.
DISCLOSURE
The study is approved by the Ethics Committee of KTO Karatay University, Medical Faculty.
The author declares no conflict of interest.
This publication has not been presented or published anywhere else.
==== Refs
REFERENCES
Brooks SK Webster RK Smith LE Woodland L Wessely S Greenberg N Rubin GJ (2020) The psychological impact of quarantine and how to reduce it: Rapid review of the evidence. Lancet. 395 :912–920.32112714
Carvalho PMM Moreira MM de Oliveira MNA Landim JMM Neto MLR (2020) The psychiatric impact of the novel coronavirus outbreak. Psychiatry Res. 286 :112902.32146248
Chang LY Huang LM Gau SS Wu YY Hsia SH Fan TY Lin KL Huang YC Lu CY Lin TY (2007) Neurodevelopment and cognition in children after enterovirus 71 infection. N Engl J Med. 356 :1226–1234.17377160
Desforges M Le Coupanec A Dubeau P Bourgouin A Lajoie L Dubé M Talbot PJ (2019) Human coronaviruses and other respiratory viruses: Underestimated opportunistic pathogens of the central nervous system? Viruses. 12 :14.31861926
Dubey S Biswas P Ghosh R Chatterjee S Dubey MJ Chatterjee S Lahiri D Lavie CJ (2020) Psychosocial impact of COVID-19. Diabetes Metab Syndr. 14 :779–788.32526627
GBD 2017 Disease and Injury Incidence and Prevalence Collaborators (2018) Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: A systematic analysis for the Global Burden of Disease Study 2017. Lancet. 392 :1789–1858.30496104
Guan WJ Liang WH Zhao Y Liang HR Chen ZS Li YM Liu XQ Chen RC Tang CL Wang T Ou CQ Li L Chen PY Sang L Wang W Li JF Li CC Ou LM Cheng B Xiong S Ni ZY Xiang J Hu Y Liu L Shan H Lei CL Peng YX Wei L Liu Y Hu YH Peng P Wang JM Liu JY Chen Z Li G Zheng ZJ Qiu SQ Luo J Ye CJ Zhu SY Cheng LL Ye F Li SY Zheng JP Zhang NF Zhong NS He JX , China Medical Treatment Expert Group for COVID-19 (2020) Comorbidity and its impact on 1590 patients with COVID-19 in China: A nationwide analysis. Eur Respir J. 55 :2000547.32217650
Halpin S O'Connor R Sivan M (2021) Long COVID and chronic COVID syndromes. J Med Virol. 93 :1242–1243.33034893
Lai CC Shih TP Ko WC Tang HJ Hsueh PR (2020) 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 :105924.32081636
Leonard BE (2001) The immune system, depression and the action of antidepressants. Prog Neuropsychopharmacol Biol Psychiatry. 25 :767–780.11383977
Mak IW Chu CM Pan PC Yiu MG Chan VL (2009) Long-term psychiatric morbidities among SARS survivors. Gen Hosp Psychiatry. 31 :318–326.19555791
Murray CJL Kyu HH Abate D Kalkidan HA Abay SM Abbafati C Abbasi N Abbastabar H Ebro JA Abdelalim A Abdollahpour I Daba AK (2018) Global, regional, and national disability-adjusted life-years (DALYs) for 359 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990–2017: A systematic analysis for the Global Burden of Disease Study 2017. Lancet. 392 :1859–1922.30415748
Rajeswari S SanjeevaReddy N (2019) Efficacy of progressive muscle relaxation on pregnancy outcome among anxious Indian primi mothers. Iran J Nurs Midwifery Res. 25 :23–30.31956594
Rogers JP Chesney E Oliver D Pollak TA McGuire P Fusar-Poli P Zandi MS Lewis G David AS (2020) 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 :611–627.32437679
Shen K Yang Y Wang T Zhao D Jiang Y Jin R Zheng Y Xu B Xie Z Lin L Shang Y Lu X Shu S Bai Y Deng J Lu M Ye L Wang X Wang Y Gao L , China National Clinical Research Center for Respiratory Diseases; National Center for Children's Health, Beijing, China; Group of Respirology, Chinese Pediatric Society, Chinese Medical Association; Chinese Medical Doctor Association Committee on Respirology Pediatrics; China Medicine Education Association Committee on Pediatrics; Chinese Research Hospital Association Committee on Pediatrics; Chinese Non-government Medical Institutions Association Committee on Pediatrics; China Association of Traditional Chinese Medicine, Committee on Children's Health and Medicine Research; China News of Drug Information Association, Committee on Children's Safety Medication; Global Pediatric Pulmonology Alliance (2020) Diagnosis, treatment, and prevention of 2019 novel coronavirus infection in children: Experts' consensus statement. World J Pediatr. 16 :223–231.32034659
Wu KK Chan SK Ma TM (2005) Posttraumatic stress, anxiety, and depression in survivors of severe acute respiratory syndrome (SARS). J Trauma Stress. 18 :39–42.16281194
Wu Y Xu X Chen Z Duan J Hashimoto K Yang L Liu C Yang C (2020) Nervous system involvement after infection with COVID-19 and other coronaviruses. Brain Behav Immun. 87 :18–22.32240762
Yang L Liu S Liu J Zhang Z Wan X Huang B Chen Y Zhang Y (2020) COVID-19: Immunopathogenesis and Immunotherapeutics. Signal Transduct Target Ther. 5 :128.32712629
Zeng F Huang Y Guo Y Yin M Chen X Xiao L Deng G (2020) Association of inflammatory markers with the severity of COVID-19: A meta-analysis. Int J Infect Dis. 96 :467–474.32425643
Zhao Q Hu C Feng R Yang Y (2020) Investigation of the mental health of patients with novel coronavirus pneumonia [in Chinese]. Chin J Neurol. 53 :E003.
Zhou F Yu T Du R Fan G Liu Y Liu Z Xiang J Wang Y Song B Gu X Guan L Wei Y Li H Wu X Xu J Tu S Zhang Y Chen H Cao B (2020a) Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: A retrospective cohort study. Lancet. 395 :1054–1062.32171076
Zhou H Lu S Chen J Wei N Wang D Lyu H Shi C Hu S (2020b) The landscape of cognitive function in recovered COVID-19 patients. J Psychiatr Res. 129 :98–102.32912598
| 36206312 | PMC9742000 | NO-CC CODE | 2022-12-15 00:01:29 | no | J Nerv Ment Dis. 2022 Dec 7; 210(12):912-914 | utf-8 | J Nerv Ment Dis | 2,022 | 10.1097/NMD.0000000000001464 | oa_other |
==== Front
J Nerv Ment Dis
J Nerv Ment Dis
JNMD
The Journal of Nervous and Mental Disease
0022-3018
1539-736X
Lippincott Williams & Wilkins
35764593
JNMD_220262
10.1097/NMD.0000000000001557
00009
3
Original Articles
Trait Versus State Predictors of Emotional Distress Symptoms
The Role of the Big-5 Personality Traits, Metacognitive Beliefs, and Strategies
Nordahl Henrik PhD ∗
Ebrahimi Omid V. Cand.Psych [email protected]
†‡
Hoffart Asle PhD [email protected]
†‡
Johnson Sverre Urnes PhD [email protected]
†‡
∗ Department of Psychology, Norwegian University of Science and Technology, Trondheim
† Department of Psychology, University of Oslo, Oslo
‡ Research Institute, Modum Bad Psychiatric Hospital, Vikersund, Norway.
Send reprint requests to Henrik Nordahl, PhD, Department of Psychology, Norwegian University of Science and Technology, Dragvoll, 7491 Trondheim, Norway. E-mail: [email protected].
12 2022
21 6 2022
21 6 2022
210 12 943950
Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.
2022
Wolters Kluwer Health, Inc. All rights reserved.
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Abstract
To enhance formulation and interventions for emotional distress symptoms, research should aim to identify factors that contribute to distress and disorder. One way to formulate emotional distress symptoms is to view them as state manifestations of underlying personality traits. However, the metacognitive model suggests that emotional distress is maintained by metacognitive strategies directed by underlying metacognitive beliefs. The aim of the present study was therefore to evaluate the role of these factors as predictors of anxiety and depression symptoms in a cross-sectional sample of 4936 participants collected during the COVID-19 pandemic. Personality traits (especially neuroticism) were linked to anxiety and depression, but metacognitive beliefs and strategies accounted for additional variance. Among the predictors, metacognitive strategies accounted for the most variance in symptoms. Furthermore, we evaluated two statistical models based on personality traits versus metacognitions and found that the latter provided the best fit. Thus, these findings indicate that emotional distress symptoms are maintained by metacognitive strategies that are better accounted for by metacognitions compared with personality traits. Theoretical and clinical implications of these findings are discussed.
Key Words
Anxiety
depression
personality
Big-5
metacognitive beliefs
==== Body
pmcCommon mental disorders are highly prevalent in the general population (Kessler et al., 2005), and as many as one third of adults in Western countries report elevated emotional distress symptoms such as anxiety and depression (Haller et al., 2014; Shim et al., 2011; Steel et al., 2014). These symptoms frequently co-occur (Jacobson and Newman, 2017) and are associated with substantially reduced quality of life for the individual (Saarni et al., 2007) as well as with enormous societal costs (Greenberg and Birnbaum, 2005; Mathers and Loncar, 2006). Even mild symptoms can be debilitating (Chachamovich et al., 2008; Haller et al., 2014), and there is an elevated risk that mild depression and anxiety symptoms can later develop into clinical disorders (Bosman et al., 2019; Pietrzak et al., 2013). Recently, the COVID-19 pandemic has led to dramatic increases in emotional distress symptoms in the general population (Daly et al., 2020; Ebrahimi et al., 2021; Ettman et al., 2020), causing concern that it will lead to a mental health crisis in the years to come (Holmes et al., 2020). Therefore, the need to identify which factors contribute to emotional distress symptoms with an aim to inform effective interventions is as relevant as ever.
One way to formulate emotional distress symptoms is to view them as state manifestations of underlying personality traits. The Five-Factor Model (Costa and McCrae, 1985) is a frequently used framework to describe universal personality dimensions founded on five broad trait dimensions (“Big-5”): neuroticism, extraversion, openness, agreeableness, and conscientiousness. These traits have been linked to psychopathology (Kotov et al., 2010), and in particular, neuroticism is viewed as a trait that accounts for psychological vulnerability (Naragon-Gainey and Watson, 2018) and can account for individual differences in affect in response to extreme environmental contexts. In line with this notion, recent studies show that people high on neuroticism react to the pandemic with stronger negative affect (Aschwanden et al., 2020; Kroencke et al., 2020; Lee and Crunk, 2022). Furthermore, lower extraversion, lower openness, lower agreeableness, and lower conscientiousness have been found to be related to higher distress levels (Nikčević and Spada, 2020).
The metacognitive model of psychological disorders (Wells and Matthews, 1994) offers an alternative formulation of emotional distress symptoms. In this perspective, emotional distress is maintained by a negative thinking style called the cognitive attentional syndrome (CAS) consisting of worry/rumination, threat monitoring, and unhelpful coping strategies (Wells, 2009). The CAS is further linked to underlying metacognition, for example, in the form of metacognitive beliefs (e.g., “I cannot stop worrying”). In this perspective, the CAS is considered the more proximal cause of distress, whereas metacognitions direct the CAS and can also be identified as markers of psychological vulnerability in the absence of CAS activation (Wells, 2019). Beliefs about the uncontrollability and dangerousness of cognition are central to the model as these beliefs compromise choice of effective self-regulation strategies and can lead to negative interpretations of cognition itself (Wells, 2009).
A challenge with formulating emotional distress symptoms within a personality framework is that personality dispositions do not yield useful information on the underlying mechanisms of how traits are connected to maladaptive self-regulation and negative outcomes that limit the application in clinical settings (Claridge and Davis, 2001). However, the metacognitive framework holds the advantage of specifying mechanisms underlying distress and interventions that can effectively reduce negative outcomes (Wells, 2009).
The aim of the present study was therefore to explore personality traits and metacognitive beliefs and strategies as predictors of anxiety and depression symptoms while controlling for demographic variables because female sex, younger age, and lower education levels are associated with elevated distress levels (Ettman et al., 2020; Ebrahimi et al., 2021). The study was conducted during the COVID-19 pandemic, which can be considered an extreme stressor likely to impact on trait (e.g., personality traits and metacognitions) and state factors (e.g., metacognitive strategies) linked to emotional distress symptoms. Our hypotheses were as follows: 1) personality traits, and in particular neuroticism, will be significantly associated with anxiety and depression symptoms; and 2) positive metacognitive beliefs, negative metacognitive beliefs, and metacognitive strategies will be significantly and positively associated with anxiety and depression symptoms and explain additional variance in symptoms when controlling for demographics and personality traits. In addition, we expected that metacognitive strategies will be stronger associated with anxiety and depression than personality traits and metacognitive beliefs as they are considered the more proximal influence on distress according to metacognitive theory (Wells and Matthews, 1994).
We also set out to evaluate the absolute and relative fit of a basic metacognitive model specified with metacognitive beliefs leading to metacognitive strategies (i.e., CAS), which further lead to emotional distress symptoms. The aim here was to further evaluate the appropriateness of formulating emotional distress symptoms and its hypothesized maintenance factors in the metacognitive model (Wells, 2009). To strengthen this evaluation, we also evaluated the fit of a comparison model where the metacognitive factor in the first model was replaced with personality traits. Our research question was if these models fit the data and if the metacognitive model fitted better than the comparison.
METHODS
Participants and Procedure
The present study is part of the Norwegian COVID-19, Mental Health, and Adherence Project, a survey-based project aiming to evaluate mental health consequences and associated factors related to the COVID-19 pandemic in Norway. The project was approved by the Regional Committee for Medical and Health Research Ethics (reference number: 125510) and registered with the Norwegian Centre for Research Data (reference number: 802810). Invitation to take part in the survey was distributed on national, regional, and local information platforms in addition to dissemination of the online survey to a random selection of Norwegian adults through a Facebook Business algorithm. Eligible participants were adults (18 years and older) currently residing in Norway. The procedure is carefully explained elsewhere (Ebrahimi et al., 2021), with a total of 10,061 participants in the first data collection.
The present study use data from the second data collection (survey-based project in the section described above), where data on personality were gathered at the same time with emotional distress and metacognitions. Following the guidelines of Strengthening the Reporting of Observational Studies in Epidemiology (Von Elm et al., 2007), and the health estimate reporting standards laid out in the GATHER statement (Stevens et al., 2016), this study was preregistered before any data collection, which can be derived from ClinicalTrials.gov (identifier NCT04444505). The data were collected between June 22 and July 13, 2020, a period approximately 1 week after the major viral mitigation protocols were lifted in Norway and where the national viral mitigation protocols and guidelines were held constant.
A total of 4936 participants took part in the study, and the sample characteristics are presented in Table 1. In the total sample, the mean age was 38.93 (SD = 13.75), 3911 (79.2%) were female, and 901 (18.3%) self-reported that they were currently diagnosed with a mental disorder.
TABLE 1 Sample Characteristics (n = 4936)
Age, mean (SD) 38.93 (13.75)
Sex
Female, n (%) 3911 (79.2)
Male, n (%) 1010 (20.5)
Other, n (%) 15 (0.3)
Current self-reported mental disorder
Yes, n (%) 901 (18.3)
No, n (%) 4035 (81.7)
Civil status
Married or in a civil partnership, n (%) 2337 (47.35)
Single/divorced, n (%) 2599 (52.65)
Education
Did not complete junior high school, n (%) 6 (0.12)
Completed junior high school, n (%) 186 (3.77)
Completed high school, n (%) 741 (15.01)
Currently studying at a university, n (%) 779 (15.78)
Completed university degree, n (%) 3224 (65.32)
Ethnicity
Native, n (%) 4542 (92.0)
Nonnative, n (%) 394 (18.0)
n, number; SD, standard deviation.
Measures
The Generalized Anxiety Disorder 7 (GAD-7; Spitzer et al., 2006) is a 7-item self-report scale assessing severity of anxiety symptoms during the past 2 weeks. The instrument has shown excellent internal consistency with an alpha of 0.92 (Löwe et al., 2008). It was developed as a screener for GAD in primary care setting but is increasingly used as a measure for anxiety in general (Johnson et al., 2019; Magnúsdóttir et al., 2022) and in anxiety disorder research (Dear et al., 2011), thus the term “anxiety symptoms” is used in this manuscript when referring to the GAD-7. In the present study, the internal consistency was good, with a Cronbach's alpha of 0.90.
The Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001) is a 9-item self-report scale assessing the severity of depression symptoms during the past 2 weeks. The instrument has shown good internal consistency with Cronbach’s alpha from 0.80 and above (Kroenke et al., 2010). In the present study, the internal consistency was good, with a Cronbach's alpha of 0.91.
The Big Five Inventory 10 (Rammstedt and John, 2007) is a 10-item self-report scale assessing the Big-5 personality dimensions using two items for each dimension. The psychometric properties of the BFI-10 have been reported as good (Rammstedt and John, 2007). No Cronbach’s alpha was calculated for the factors as they are indicated by two items each.
The CAS-1 (Wells, 2009) is a 16-item self-report tool developed to assess metacognitive beliefs (positive and negative) and strategies (worry/rumination, threat monitoring, coping strategies). It has a three-factor solution with good psychometric properties (Nordahl and Wells, 2019). In the present study, the internal consistency was 0.63 for positive beliefs, 0.71 for negative beliefs, and 0.91 for strategies.
Overview of Statistical Analyses
Using IBM SPSS Statistics version 27, bivariate correlations were used to explore the basic associations among the variables, and two hierarchical linear regressions were used to assess the additional contribution of personality traits and metacognitive factors in explaining variance in anxiety and depression when controlling for demographics. The metacognitive factors were entered on separate steps with the aim to evaluate any additional contribution from them as metacognitive theory distinguishes among these components and suggest a causal sequence for them (Wells, 2009).
In a secondary set of analyses, we evaluate the model fit of a metacognitive model consisting of and distinguishing among metacognitive beliefs, metacognitive strategies, and emotional distress symptoms in line with metacognitive theory (Wells, 2009). This model was specified by a latent distress factor consisting of the observed variables GAD-7 total score and PHQ-9 total score; a latent CAS factor consisting of the observed variables worry/rumination (CAS-1 item 1), threat monitoring (CAS-1 item 2), and maladaptive coping strategies (CAS-1 total score of scale 3); and a latent metacognitive factor. To create the metacognitive belief factor, we used the four items assessing negative metacognitive beliefs from the CAS-1 measure as negative metacognitive beliefs is the most important factor underlying CAS strategies according to metacognitive theory (Wells, 2009) and because earlier studies have indicated that negative but not positive metacognitive beliefs account for unique variance in symptoms when also controlling metacognitive strategies (Nordahl and Wells, 2019). In addition, we evaluated a second model where personality traits (the four with the strongest association to symptoms) constituted a latent personality factor that replaced the latent metacognitive factor in the first model, keeping the number of indicators and all other variables and paths unchanged. The aims here were to evaluate if both models with the same number of variables and identical paths fit the data and to evaluate if one model fitted the data better than the other.
The SEM analyses were conducted using IBM SPSS AMOS Graphics version 26. Four commonly recommended fit statistics were used to evaluate the models: the comparative fit index (CFI), the Tucker-Lewis index (TLI), the standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA). The CFI and TLI should be above 0.95, the RMSEA should be below or close to 0.06, the upper limit of the 90% RMSEA confidence interval (CI) should not exceed 0.10 to represent a good fit, and the SRMR should be less than 0.08 (Browne and Cudeck, 1993; Hu and Bentler, 1999). We used the Akaike Information Criterion (AIC; Akaike, 1974) to compare the fit of the two nonnested models. When both models have the same number of indicators, latent variables, and paths, the model with the lowest AIC value is considered to fit best to the data.
RESULTS
Correlational Analyses
The mean scores of the variables and the bivariate correlations among them are presented in Table 2. Anxiety and depression showed the strongest association with neuroticism among the personality traits, and with metacognitive strategies among the metacognitive factors.
TABLE 2 Bivariate Correlations Between the Variables With Mean Values and Standard Deviations (n = 4936)
2 3 4 5 6 7 8 9 10 Mean (SD)
1 GAD 0.805** 0.534** −0.167** 0.038** −0.233** −0.095** 0.277** 0.514** 0.804** 4.66 (4.37)
2 PHQ 0.464** −0.231** 0.002 −0.232** −0.180** 0.257** 0.494** 0.757** 6.63 (5.66)
3 N −0.277** 0.033 −0.280** −0.152** 0.205** 0.402** 0.534** 5.54 (2.22)
4 E 0.079** 0.258** 0.155** −0.113** −0.163** −0.193** 6.92 (2.20)
5 O 0.017 0.038** 0.027 −0.009 0.043** 7.16 (2.02)
6 A 0.174** −0.173** −0.231** −0.233** 7.49 (1.57)
7 C −0.065** −0.133** −0.136** 7.76 (1.60)
8 PMB 0.497** 0.362** 143.16 (74.73)
9 NMB 0.567** 112.43 (84.65)
10 CAS 13.35 (11.94)
GAD, anxiety symptoms; PHQ, depression symptoms; N, neuroticism; E, extraversion; O, openness; A, agreeableness; C, conscientiousness; PMB, positive metacognitive beliefs; NMB, negative metacognitive beliefs; CAS, metacognitive strategies.
**p < 0.01.
Linear Regression Analyses
When anxiety symptoms were used as the dependent variable, sex, age, and education level were entered together in step 1 and accounted for 7.3% of the variance. On the second step, personality traits entered as a block accounted for an additional 24.0% of the variance in anxiety symptoms. Among the personality traits, neuroticism, openness, and agreeableness accounted for individual variance. On the third step, positive metacognitive beliefs accounted for an additional 2.2% of the variance, and on the fourth step, negative metacognitive beliefs accounted for an additional 6.8% of the variance. In the fifth and final step, metacognitive strategies accounted for an additional 26.6% of the variance. In the final model when the overlap among all the predictors were controlled, lower education level, higher neuroticism, higher extraversion, lower agreeableness, higher conscientiousness, lower levels of positive metacognitive beliefs, higher levels of negative metacognitive beliefs, and higher levels of metacognitive strategies all accounted for unique variance in (higher) anxiety symptoms. Among the predictors in the final model, metacognitive strategies were by far the strongest predictor of anxiety. In sum, the predictors accounted for 66.9% of the variance in anxiety symptoms. The regression summary statistics are presented in Table 3.
TABLE 3 Hierarchical Linear Regression With Anxiety Symptoms as the Dependent, and Sex/Age/Education Level, Personality Traits, and Metacognitive Beliefs and Strategies as Predictors (n = 4963)
Anxiety–GAD-7 F Change R2 Change r t
1 129.801 0.073**
Sex −0.070 −5.142**
Age −0.200 −14.583**
Education −0.112 −8.137**
2 343.951 0.240**
Sex −0.016 −1.351
Age −0.099 −8.398**
Education −0.073 −6.223**
N 0.412 34.925**
E −0.013 −1.066
O 0.053 4.522**
A −0.089 −7.511**
C 0.008 0.673
3 163.130 0.022**
Sex −0.026 −2.232*
Age −0.075 −6.423**
Education −0.077 −6.642**
N 0.392 33.723**
E −0.006 −0.481
O 0.046 3.967**
A −0.071 −6.138**
C 0.008 0.658
Positive MB 0.148 12.772**
4 558.911 0.068**
Sex −0.018 −1.618
Age −0.050 −4.512**
Education −0.051 −4.655**
N 0.308 27.946**
E −0.001 −0.065
O 0.047 4.298**
A −0.049 −4.460**
C 0.020 1.844
Positive MB 0.018 1.619
Negative MB 0.260 23.641**
5 3945.435 0.266**
Sex 0.007 0.881
Age −0.007 −0.865
Education −0.027 −3.265**
N 0.109 13.227**
E 0.018 2.141*
O 0.010 1.183
A −0.028 −3.430**
C 0.033 4.014**
Positive MB −0.041 −4.952**
Negative MB 0.062 7.509**
CAS 0.515 62.813**
r, semipartial (part) correlation; N, neuroticism; E, extraversion; O, openness; A, agreeableness; C, conscientiousness; MB, metacognitive beliefs; CAS, metacognitive strategies.
*p < 0.05, **p < 0.01.
When depression symptoms were used as the dependent variable, sex, age, and education level were entered together in step 1 and accounted for 7.8% of the variance, and all variables accounted for unique variance. In the second step, personality traits entered as a block accounted for an additional 19.1% of the variance in depression symptoms. In the third step, positive metacognitive beliefs accounted for an additional 1.9% of the variance, and on the fourth step, negative metacognitive beliefs accounted for an additional 6.8% of the variance. In the fifth and final step, metacognitive strategies accounted for an additional 24.3% of the variance. In the final model when the overlap among all the predictors was controlled, lower age, lower education level, higher neuroticism, lower extraversion, lower agreeableness, lower conscientiousness, lower levels of positive metacognitive beliefs, higher levels of negative metacognitive beliefs, and higher levels of metacognitive strategies all accounted for unique variance in (higher) depression symptoms. Metacognitive strategies were by far the strongest correlate of depression in the final model. In sum, the predictors accounted for 59.9% of the variance in depression symptoms. The regression summary statistics are presented in Table 4.
TABLE 4 Hierarchical Linear Regression With Depression Symptoms as the Dependent, and Sex/Age/Education Level, Personality Traits, and Metacognitive Beliefs and Strategies as Predictors (n = 4963)
Depression–PHQ-9 F Change R2 Change r t
1 139.727 0.078**
Sex −0.059 −4.334**
Age −0.193 −14.134**
Education −0.145 −10.576**
2 257.909 0.191**
Sex −0.033 −2.712**
Age −0.104 −8.534**
Education −0.104 −8.570**
N 0.314 25.795**
E −0.086 −7.101**
O 0.029 2.416*
A −0.082 −6.708**
C −0.078 −6.436**
3 128.670 0.019**
Sex −0.042 −3.506**
Age −0.081 −6.755**
Education −0.108 −8.962**
N 0.296 24.591**
E −0.080 −6.650**
O 0.023 1.890
A −0.066 −5.467**
C −0.079 −6.543**
Positive MB 0.136 11.343**
4 516.708 0.068**
Sex −0.034 −2.976**
Age −0.056 −4.919**
Education −0.082 −7.144**
N 0.215 18.811**
E −0.075 −6.563**
O 0.024 2.095*
A −0.044 −3.808**
C −0.066 −5.763**
Positive MB 0.007 0.620
Negative MB 0.260 22.731**
5 2973.641 0.243**
Sex −0.010 −1.116
Age −0.016 −1.720*
Education −0.058 −6.454**
N 0.030 3.294**
E −0.058 −6.371**
O −0.012 −1.322
A −0.024 −2.605**
C −0.054 −5.958**
Positive MB −0.049 −5.382**
Negative MB 0.069 7.681**
CAS 0.493 54.531**
r, semipartial (part) correlation; N, neuroticism; E, extraversion; O, openness; A, agreeableness; C, conscientiousness; MB, metacognitive beliefs; CAS, metacognitive strategies.
*p < 0.05; **p < 0.01.
Secondary Analyses
Structural equation modeling was conducted to test the fit of a metacognitive model. The model is presented in Figure 1 and showed the following fit indices: χ2(25) = 547.207, p < 0.001, CFI = 0.980, TLI = 0.971, RMSEA = 0.065 (90% CI, 0.060–0.070), SRMR = 0.031, indicating a good model fit to the data. The AIC value for the model was 605.207.
FIGURE 1 Structural equation model of the metacognitive model (n = 4936). Circles represent latent variables, and rectangles represent observed variables (indicators). CAS1, worry/rumination; CAS2, threat monitoring; CAS3, maladaptive coping strategies; CAS4A-D, negative metacognitive beliefs.
The comparison model presented in Figure 2 where personality traits (all but openness) were used to specify a latent “personality” factor with the means to compare the metacognitive model showed the following fit indices: χ2(25) = 716.080, p < 0.001, CFI = 0.971, TLI = 0.958, RMSEA = 0.075 (90% CI, 0.070–0.080), SRMR = 0.034, also indicating a good model fit to the data. The AIC value for the model was 774.198. The personality model showed a higher AIC value compared with the metacognitive model, indicating that the latter provided a better fit to the data.
FIGURE 2 Structural equation model of the comparison model that included personality traits (n = 4936). Circles represent latent variables, and rectangles represent observed variables (indicators). CAS1, worry/rumination; CAS2, threat monitoring; CAS3, maladaptive coping strategies; N, neuroticism; E, extraversion; A, agreeableness; C, conscientiousness.
DISCUSSION
In the present study, we set out to test the relative contribution of personality traits and metacognitive beliefs and strategies to anxiety and depression symptoms in a large sample during the COVID-19 pandemic in Norway.
In terms of mean scores on anxiety and depression (mean = 6.63 for depression and 4.66 for anxiety), the current sample scored on average higher relative to a representative Scandinavian sample assessed before the pandemic (PHQ-9 = 3.59 [CI, 3.36–3.81] and GAD-7 = 3.70 [CI, 3.44–3.96]; Johansson et al., 2013), probably due to the threats associated with the ongoing pandemic and the nonpharmacological strategies aimed at impeding viral transmission chains (e.g., Campion et al., 2020). In terms of mean scores on metacognitive beliefs and strategies assessed with the CAS-1 (mean = 143.16 for positive beliefs, 112.43 for negative beliefs, and 13.35 for strategies), there is to the authors' knowledge only one Scandinavian prepandemic study that can be used as a reference point. Nordahl and Wells (2019) used convenience sampling with restricted knowledge about the participants and reported mean scores as follows (n = 773): 130.89 for positive beliefs, 146.98 for negative beliefs, and 21.10 for strategies. Face value comparison between the current sample and this one indicates that the current sample report lower mean scores for negative metacognitive beliefs and strategies, but higher scores for positive metacognitive beliefs. However, normative data for the CAS-1 have not been established so caution is warranted in comparing the samples on these scales.
All the personality traits and the metacognitive factors were significantly correlated with anxiety symptoms, but anxiety showed weak associations with extraversion, openness, agreeableness, conscientiousness, and positive metacognitive beliefs. Furthermore, all personality traits and metacognitive factors were significantly correlated with depression symptoms except for openness, and depression showed weak correlations with extraversion, agreeableness, conscientiousness, and positive metacognitive beliefs.
In the regression models, we observed that female sex, lower age, and lower education level were unique and significant correlates of higher anxiety and depression symptoms. Personality traits entered as a block significantly accounted for variance in anxiety and depression symptoms, and neuroticism, openness, and agreeableness were unique and independent predictors of anxiety symptoms, whereas all the personality traits were unique and independent predictors of depression symptoms. On top of the demographics and personality traits, positive metacognitive beliefs accounted for additional variance in both types of distress, and negative metacognitive beliefs accounted for additional variance on top of positive metacognitive beliefs. Here we observed that entering negative metacognitive beliefs to the models lead positive metacognitive beliefs to be nonsignificant as an individual predictor of both anxiety and depression symptoms. Metacognitive strategies were entered last in both models and accounted for a substantial part of additional variance in both anxiety and depression symptoms on top of all the other predictors.
In the final equation when anxiety symptoms were used as the dependent variable, education level, neuroticism, extraversion, agreeableness, conscientiousness, positive metacognitive beliefs, negative metacognitive beliefs, and metacognitive strategies accounted for independent and unique variance. Education level, agreeableness, and positive metacognitive beliefs were negatively associated, whereas the others were positively associated with anxiety symptoms when the overlap among all the predictors was controlled. When depression symptoms were used as the dependent variable, age, education level, neuroticism, extraversion, agreeableness, conscientiousness, positive metacognitive beliefs, negative metacognitive beliefs, and metacognitive strategies accounted for independent and unique variance in the final equation. Age, education level, extraversion, agreeableness, conscientiousness, and positive metacognitive beliefs were negatively associated, whereas the others were positively associated with depression symptoms when the overlap among all the predictors was controlled.
In line with previous studies, female sex, lower age, and lower education level were associated with higher distress levels (Daly et al., 2020; Ettman et al., 2020). Among these predictors, education levels showed a robust association with distress, whereas the association between symptoms and sex/age in large was accounted for by personality traits and metacognitive factors. Furthermore, neuroticism was the strongest predictor of anxiety and depression symptoms among the personality traits, and as expected, we observed that higher neuroticism was associated with higher distress levels. This observation is also in line with previous studies indicating that neuroticism is an important vulnerability factor underlying emotional distress (Kroencke et al., 2020; Lee and Crunk, 2022) and with studies indicating a role for other personality traits as well but to a lesser degree (Nikčević and Spada, 2020). Moreover, all the metacognitive factors accounted for additional variance on top of demographics and personality traits when entered on consecutive steps, and as expected, metacognitive strategies showed the strongest link with emotional distress symptoms. Adding negative metacognitive beliefs to the model lead positive metacognitive beliefs to be nonsignificant as a predictor, whereas negative metacognitive beliefs remained significant as a unique predictor when metacognitive strategies entered the model. In line with metacognitive theory (Wells, 2009), these observations suggest that the CAS and negative metacognitive beliefs are the most important influences on anxiety and depression symptoms among the metacognitive factors and are consistent with an earlier study reporting the same finding when social anxiety was used as the dependent variable (Nordahl and Wells, 2019). However, we observed that adding metacognitive strategies to the regression models in the final steps led positive metacognitive beliefs to become a significant and negative predictor of anxiety and depression. According to metacognitive theory (Wells, 2009), positive metacognitive beliefs are less relevant to distress and disorder compared with metacognitive strategies and negative metacognitive beliefs but should also be positively related to symptoms. In line with theory (Wells, 2009), positive metacognitive beliefs showed significant and positive bivariate correlations to anxiety and depression and was also a significant and positive predictor of distress before controlling negative metacognitive beliefs, indicating that our observations here are a result of interactions among the variables in the final equation and the outcomes in the two models as positive metacognitive beliefs and metacognitive strategies only show a moderate positive association, indicating that collinearity cannot account for this observation.
Our results from the regressions further suggest that metacognitive beliefs account for variance in emotional distress above personality traits. This finding is in line with a study showing that metacognitive beliefs accounted for variance in health anxiety above neuroticism (Bailey and Wells, 2013). In addition to previous research, we demonstrate that metacognitive strategies account for anxiety and depression symptoms even when controlling for personality traits and metacognitive beliefs, and that this state variable is by far the strongest associate of distress. This finding is in line with the metacognitive model of psychological disorder where metacognitive strategies is a transdiagnostic factor underlying different types and is considered the more proximal cause of distress (Wells and Matthews, 1994).
In the secondary analyses where we tested the structural relationships and the model fit of a basic metacognitive model, we found that the data fitted the model reasonably well, even when comparing it to an alternative model where personality was used as the underlying factor. The personality model also provided good model fit, but accounted for a slightly lower part of the variance in metacognitive strategies, and the overall model fit was not as good as for the metacognitive model. Thus, the metacognitive model provided the best fit to the data among these two.
It could be that dysfunctional metacognition can be considered lower-order dispositions and a function of higher-order personality traits. However, we did not test such a model as it is not consistent with the metacognitive theory where dysfunctional metacognition is considered the underlying mechanism in distress and vulnerability. Research indicates that there is a bidirectional relationship between metacognitions and “trait-anxiety” (closely linked to neuroticism) (Nordahl et al., 2019), that metacognition moderate the effect of emotional reactivity on anxiety (Clauss et al., 2020), and that metacognitive change is associated with changes in Big-5 personality traits (Kennair et al., 2021). Thus, it is an option that metacognition can be an underlying mechanism in vulnerability attributed to personality traits in personality theory. In addition, the concerns about personality dispositions not providing a framework of how they relate in a useful way to maladaptive self-regulation still stands (Claridge and Davis, 2001), whereas the metacognitive approach holds the advantage providing a model for these associations and a manual on how they can be effectively modified (Wells, 2009).
The implication from the present results is that emotional distress is most strongly related to metacognitive strategies, but also to personality traits (specifically neuroticism) and negative metacognitive beliefs. Formulations and interventions should therefore focus on metacognitive strategies as maintenance factors of distress, which from a statistical point of view based on the results from the current study can be seen as a result of personality traits and/or metacognitive beliefs. From a theoretical point of view, the metacognitive model holds the advantage of specifying metacognitive beliefs as a mechanism underlying the CAS and provides a means to which these factors can be effectively formulated within metacognitive theory and treated with metacognitive therapy (Wells, 2009), which have shown to effectively reduce emotional distress and maladaptive metacognitions (Normann and Morina, 2018).
There are several limitations that must be considered. Causal inferences cannot be made based on cross-sectional data. All variables were assessed with self-report, and we used cost-beneficial assessment tools. Some of the significant associations between the predictors and the outcome variables were of weak strength, and it is possible that a large sample size accounts for the significant relationships observed in these cases. These limitations show that there is a need to be cautious in generalizing from these findings. Further studies should assess the theorized relations between factors in longitudinal data, which may enable interference about the temporal and reciprocal relations among them, and there is a need to explore if personality traits and/or metacognitive factors play a role in the transition from distress to disorder. There is also a need to clarify the temporal and reciprocal relationships between personality traits (e.g., neuroticism) and metacognition with an aim to improve formulation of psychological vulnerability and preventive interventions.
In conclusion, the present study indicates a role for personality traits and metacognitive beliefs and strategies in emotional distress symptoms, with the strongest contribution from maladaptive metacognitive strategies. These can effectively be formulated and targeted within the metacognitive approach.
DISCLOSURE
The project was approved by the Regional Committee for Medical and Health Research Ethics (reference number: 125510) and registered with the Norwegian Centre for Research Data (reference number: 802810). Signed informed consent was obtained from all study participants.
The authors declare no conflict of interest.
==== Refs
REFERENCES
Akaike H (1974) A new look at the statistical model identification. IEEE Trans Automatic Control. 19 :716–723.
Aschwanden D Strickhouser JE Sesker AA Lee JH Luchetti M Stephan Y Sutin AR Terracciano A (2020) Psychological and behavioural responses to coronavirus disease 2019: The role of personality. Eur J Pers. 35 :51–56.
Bailey R Wells A (2013) Does metacognition make a unique contribution to health anxiety when controlling for neuroticism, illness cognition, and somatosensory amplification? J Cogn Psychother. 27 :327–337.32759104
Bosman RC Ten Have M de Graaf R Muntingh AD van Balkom AJ Batelaan NM (2019) Prevalence and course of subthreshold anxiety disorder in the general population: A three-year follow-up study. J Affect Disord. 247 :105–113.30660019
Browne MW Cudeck R (1993) Alternative ways if assessing model fit. In Bollen KA Long JS (Eds), Testing structural equation models (pp 136–162). Newbury Park, CA: Sage.
Campion J Javed A Sartorius N Marmot M (2020) Addressing the public mental health challenge of COVID-19. Lancet Psychiatry. 7 :657–659.32531299
Chachamovich E Fleck M Laidlaw K Power M (2008) Impact of major depression and subsyndromal symptoms on quality of life and attitudes toward aging in an international sample of older adults. Gerontologist. 48 :593–602.18981276
Claridge G Davis C (2001) What's the use of neuroticism? Personal Individ Differ. 31 :383–400.
Clauss K Bardeen JR Thomas K Benfer N (2020) The interactive effect of emotional reactivity and maladaptive metacognitive beliefs on anxiety. Cognit Emot. 34 :393–401.31159645
Costa PT McCrae RR (1985) The NEO personality inventory. Odessa, FL: Psychological Assessment Resources.
Daly M Sutin AR Robinson E (2020) Longitudinal changes in mental health and the COVID-19 pandemic: Evidence from the UK household longitudinal study. Psychol Med. 1–10. doi:10.1017/S0033291720004432.
Dear BF Titov N Sunderland M McMillan D Anderson T Lorian C Robinson E (2011) Psychometric comparison of the Generalized Anxiety Disorder Scale-7 and the Penn State Worry Questionnaire for measuring response during treatment of generalised anxiety disorder. Cogn Behav Ther. 40 :216–227.21770844
Ebrahimi OV Hoffart A Johnson SU (2021) Physical distancing and mental health during the COVID-19 pandemic: Factors associated with psychological symptoms and adherence to pandemic mitigation strategies. Clin Psychol Sci. 9 :489–506.
Ettman CK Abdalla SM Cohen GH Sampson L Vivier PM Galea S (2020) Prevalence of depression symptoms in US adults before and during the COVID-19 pandemic. JAMA Netw Open. 3 :e2019686.32876685
Greenberg PE Birnbaum HG (2005) The economic burden of depression in the US: Societal and patient perspectives. Expert Opin Pharmacother. 6 :369–376.15794728
Haller H Cramer H Lauche R Gass F Dobos GJ (2014) The prevalence and burden of subthreshold generalized anxiety disorder: A systematic review. BMC Psychiatry. 14 :128.24886240
Holmes EA O'Connor RC Perry VH Tracey I Wessely S Arseneault L Bullmore E (2020) Multidisciplinary research priorities for the COVID-19 pandemic: A call for action for mental health science. Lancet Psychiatry. 7 :547–560.32304649
Hu LT Bentler PM (1999) Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Model Multidiscip J. 6 :1–55.
Jacobson NC Newman MG (2017) Anxiety and depression as bidirectional risk factors for one another: A meta-analysis of longitudinal studies. Psychol Bull. 143 :1155–1200.28805400
Johansson R Carlbring P Heedman Å Paxling B Andersson G (2013) Depression, anxiety and their comorbidity in the Swedish general population: Point prevalence and the effect on health-related quality of life. PeerJ. 1 :e98.23862109
Johnson SU Ulvenes PG Øktedalen T Hoffart A (2019) Psychometric properties of the general anxiety disorder 7-item (GAD-7) scale in a heterogeneous psychiatric sample. Front Psychol. 10 :1713.31447721
Kennair LEO Solem S Hagen R Havnen A Nysæter TE Hjemdal O (2021) Change in personality traits and facets (revised NEO personality inventory) following metacognitive therapy or cognitive behaviour therapy for generalized anxiety disorder: Results from a randomized controlled trial. Clin Psychol Psychother. 28 :872–881.33338315
Kessler RC Berglund P Demler O Jin R Merikangas KR Walters EE (2005) Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 62 :593–602.15939837
Kotov R Gamez W Schmidt F Watson D (2010) Linking “big” personality traits to anxiety, depressive, and substance use disorders: A meta-analysis. Psychol Bull. 136 :768–821.20804236
Kroencke L Geukes K Utesch T Kuper N Back MD (2020) Neuroticism and emotional risk during the COVID-19 pandemic. J Res Pers. 89 :104038.33071370
Kroenke K Spitzer RL Williams JB (2001) The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med. 16 :606–613.11556941
Kroenke K Spitzer RL Williams JB Löwe B (2010) The patient health questionnaire somatic, anxiety, and depressive symptom scales: A systematic review. Gen Hosp Psychiatry. 32 :345–359.20633738
Lee SA Crunk EA (2022) Fear and psychopathology during the COVID-19 crisis: Neuroticism, hypochondriasis, reassurance-seeking, and coronaphobia as fear factors. Omega (Westport). 85 :843–496.
Löwe B Decker O Müller S Brähler E Schellberg D Herzog W Herzberg PY (2008) Validation and standardization of the generalized anxiety disorder screener (GAD-7) in the general population. Med Care. 46 :266–274.18388841
Magnúsdóttir I Lovik A Unnarsdóttir AB McCartney D Ask H Kõiv K Christoffersen LAN Johnson SU Hauksdóttir A Fawns-Ritchie C Helenius D González-Hijón J Lu L Ebrahimi OV Hoffart A Porteous DJ Fang F Jakobsdóttir J Lehto K Andreassen OA Pedersen OBV Aspelund T Valdimarsdóttir UA (2022) Acute COVID-19 severity and mental health morbidity trajectories in patient populations of six nations: An observational study. Lancet Public Health. 7 :e406–e416.35298894
Mathers CD Loncar D (2006) Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 3 :e442.17132052
Naragon-Gainey K Watson D (2018) What lies beyond neuroticism? An examination of the unique contributions of social-cognitive vulnerabilities to internalizing disorders. Assessment. 25 :143–158.27411679
Nikčević AV Spada MM (2020) The COVID-19 anxiety syndrome scale: Development and psychometric properties. Psychiatry Res. 292 :113322.32736267
Nordahl H Hjemdal O Hagen R Nordahl HM Wells A (2019) What lies beneath trait-anxiety? Testing the self-regulatory executive function model of vulnerability. Front Psychol. 10 :122.30804834
Nordahl H Wells A (2019) Measuring the cognitive attentional syndrome associated with emotional distress: Psychometric properties of the CAS-1. Int J Cogn Ther. 12 :292–306.
Normann N Morina N (2018) The efficacy of metacognitive therapy: A systematic review and Meta-analysis. Front Psychol. 9 :2211.30487770
Pietrzak RH Kinley J Afifi TO Enns MW Fawcett J Sareen J (2013) Subsyndromal depression in the United States: Prevalence, course, and risk for incident psychiatric outcomes. Psychol Med. 43 :1401–1414.23111093
Rammstedt B John OP (2007) Measuring personality in one minute or less: A 10-item short version of the big five inventory in English and German. J Res Pers. 41 :203–212.
Saarni SI Suvisaari J Sintonen H Pirkola S Koskinen S Aromaa A Lönnqvist J (2007) Impact of psychiatric disorders on health-related quality of life: General population survey. Br J Psychiatry. 190 :326–332.17401039
Shim RS Baltrus P Ye J Rust G (2011) Prevalence, treatment, and control of depressive symptoms in the United States: Results from the National Health and Nutrition Examination Survey (NHANES), 2005–2008. J Am Board Fam Med. 24 :33–38.21209342
Spitzer RL Kroenke K Williams JB Löwe B (2006) A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch Intern Med. 166 :1092–1097.16717171
Steel Z Marnane C Iranpour C Chey T Jackson JW Patel V Silove D (2014) The global prevalence of common mental disorders: A systematic review and meta-analysis 1980–2013. Int J Epidemiol. 43 :476–493.24648481
Stevens GA Alkema L Black RE Boerma JT Collins GS Ezzati M , GATHER Working Group (2016) Guidelines for accurate and transparent health estimates reporting: The GATHER statement. PLoS Med. 13 :e1002116.27504831
Von Elm E Altman DG Egger M Pocock SJ Gøtzsche PC Vandenbroucke JP (2007) The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Ann Intern Med. 147 :573–577.17938396
Wells A (2009) Metacognitive therapy for anxiety and depression. New York: Guilford Press.
Wells A (2019) Breaking the cybernetic code: Understanding and treating the human metacognitive control system to enhance mental health. Front Psychol. 10 :2621.31920769
Wells A Matthews G (1994) Attention and emotion: A clinical perspective. Hove: Erlbaum.
| 35764593 | PMC9742004 | NO-CC CODE | 2022-12-15 00:01:29 | no | J Nerv Ment Dis. 2022 Dec 21; 210(12):943-950 | utf-8 | J Nerv Ment Dis | 2,022 | 10.1097/NMD.0000000000001557 | oa_other |
==== Front
Paediatr Drugs
Paediatr Drugs
Paediatric Drugs
1174-5878
1179-2019
Springer International Publishing Cham
552
10.1007/s40272-022-00552-9
Acknowledgement to Referees
Acknowledgement to Referees
12 12 2022
14
© Springer Nature Switzerland AG 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
==== Body
pmcDear Reader
Welcome to the first edition of Pediatric Drugs for 2023.
As we embark on the 25th year of the journal, we wish to reflect on the very successful year that was 2022, and to thank all who have contributed their valuable time and effort to Pediatric Drugs over the past 12 months.
There were 55 articles published in issues in 2022; the most popular of these, in terms of downloads from SpringerLink, were:
Zahn J, Eberl S, Rödle W, et al. Metamizole Use in Children: Analysis of Drug Utilisation and Adverse Drug Reactions at a German University Hospital between 2015 and 2020. Pediatr Drugs 2022; 24: 45–56.
Islabão AG, Trinidade VC, da Mota LM et al. Managing Antiphospholipid Syndrome in Children and Adolescents: Current and Future Prospects. Pediatr Drugs 2022; 24: 13–27.
Eichenfield LF, Stripling S, Fung S, et al. Recent Developments and Advances in Atopic Dermatitis: A Focus on Epidemiology, Pathophysiology, and Treatment in the Pediatric Setting. Pediatr Drugs 2022; 24: 293–305.
Gupta A, Cripe TP, et al. Immunotherapies for Pediatric Solid Tumors: A Targeted Update. Pediatr Drugs 2022; 24: 1–12.
Ziesenitz VC, Welzel T, van Dyk M, et al. Efficacy and Safety of NSAIDs in Infants: A Comprehensive Review of the Literature of the Past 20 Years. Pediatr Drugs 2022; 24: 603–655.
Wong CK, Low MC, Kwok AC, et al. Slower Recovery with Early Lopinavir/Ritonavir use in Pediatric COVID-19 Patients: A Retrospective Observational Study. Pediatr Drugs 2022; 24: 269–280.
Magnolo N, Kingo K, Laquer V, et al. Efficacy of Secukinumab Across Subgroups and Overall Safety in Pediatric Patients with Moderate to Severe Plaque Psoriasis: Week 52 Results from a Phase III Randomized Study. Pediatr Drugs 2022; 24: 377–387.
Freriksen JJ, van der Zanden TM, Holsappel IG, et al. Best Evidence-Based Dosing Recommendations for Dexmedetomidine for Premedication and Procedural Sedation in Pediatrics: Outcome of a Risk-Benefit Analysis By the Dutch Pediatric Formulary. Pediatr Drugs 2022; 24: 247–257.
Masi, G, Pfanner C, Liboni F, et al. Acute Tolerability of Methylphenidate in Treatment-Naïve Children with ADHD: An Analysis of Naturalistically Collected Data from Clinical Practice. Pediatr Drugs 2022; 24: 147–154.
Luger TA, Hebert AA, Zaenglein AL, et al. Subgroup Analysis of Crisaborole for Mild-to-Moderate Atopic Dermatitis in Children Aged 2 to <18 Years. Pediatr Drugs 2022; 24: 175–183.
The high quality of content published in Pediatric Drugs has been reflected in the most recent Journal Impact Factor of 3.930 (placing the journal in the first quartile of the Pediatrics category) and CiteScore™ of 5.4.
We thank the authors who have contributed articles to Pediatric Drugs over the course of 2022. The skill and dedication of all authors are critical to the continued success of the journal. The quality of published articles is also testament to the significant efforts of the peer reviewers, whose commitment ensures that the journal’s content is held to the highest possible standard. We would like to thank the following individuals who acted as reviewers for Pediatric Drugs in the last 12 months:
Robina Aerts, Belgium
Masashi Akiyama, Japan
Daoud Al-Badriyeh, Qatar
Doralina L. Anghelescu, USA
Jack E. Ansell, USA
Roberto Antonucci, Italy
Ann Lisa Arulappen, Malaysia
Michael Auerbach, USA
Yaron Avitzur, Canada
Stephen Balevic, USA
Mariana Baserga, USA
Raman Baweja, USA
Anton M. Bennett, USA
Osvaldo Borrelli, UK
Sasigarn Bowden, USA
Finola Bruins, The Netherlands
Claudio Bruno, Italy
Matteo Bruschettini, Sweden
Reiner Buchhorn, Germany
Gilbert J. Burckart, USA
Luis Bustos Fernandez, Argentina
Florencia Carbone, Belgium
Fabianne A. Carlesse, Brazil
Waldemar A. Carlo, USA
Lucinda Carr, UK
Elio Castagnola, Italy
Jennifer C. Cather, USA
Manuela Cerbone, UK
Godfrey Chan, Hong Kong, China
Mary Chandran, USA
Joyce Chen, Taiwan, Republic of China
Dinesh K. Chirla, India
Julia Chisholm, UK
Imti Choonara, UK
Samuele Cortese, UK
Damien C. Croteau-Chonka, USA
Mihaela Damian, USA
Dominique Darmaun, France
Jennifer Day, USA
Lissy de Ridder, The Netherlands
Vito Di Lernia, Italy
Carmel Doyle, Ireland
Prakash Dubey, India
Diana Dubrall, Germany
Jan Dudley, UK
Friedrich Ebinger, Germany
Hirotoshi Echizen, Japan
Graham Emslie, USA
Johanna C. Escher, Canada
Susanna Every-Palmer, New Zealand
Yamen Ezaizi, USA
Laurent Favie, The Netherlands
Istvan Fedor, Hungary
Steven R. Feldman, USA
Regina Folster-Holst, Germany
Ted Chun Tat Fong, Hong Kong, China
Fabrizio Franceschini, Italy
Wayne Furman, USA
Alexander S. Gallus, Australia
Justin Godown, USA
Jennifer L. Goldman, USA
Kenneth R. Goldschneider, USA
Rachel G. Greenberg, USA
Teri Greiling, USA
Cecilia Grinsvall, Sweden
Jian-Long Guan, China
Elizabeth A. Hall, USA
Angela Hassiotis, UK
Jason A. Hayes, Canada
Mingyan Hei, China
Paula E. Heikkilä, Finland
Maria Teresa Ferreira Herdeiro, Portugal
Suzanna Hirsch, USA
Alexis E. Horace, USA
Yi-Wei Huang, Taiwan, Republic of China
Selene Huerta-Olvera, Mexico
Tim Hundscheid, The Netherlands
Anna L. Hurley-Wallace, UK
Giuseppe Indolfi, Italy
Evelyne Jacqz Aigrain, France
Naveen Jain, India
Laura L. James, USA
Zainub Jooma, South Africa
Stuart S. Kaufman, USA
Steven J. Kindel, USA
Sébastien Kindt, Belgium
Sarah Kittel-Schneider, Germany
Hannu Kokki, Finland
Maria Kourti, Greece
Usha Krishnan, Australia
Abbot Laptook, USA
Helen Laycock, UK
Zhiping Li, China
Christoph Licht, Canada
Marijn Lijffijt, USA
Miquéias Lopes-Pacheco, Portugal
Iftekhar Mahmood, USA
Mohamed Mahmoud, USA
Suvankar Majumdar, USA
Kenneth K.C. Man, UK
David Mandelbaum, USA
Anna Mania, Poland
Anshu Marathe, USA
Arthur J. Matas, USA
Sujeev Mathur, UK
Tess McPherson, UK
Kevin Meesters, UK
Jondavid Menteer, USA
Robin Michelet, Germany
Gregorio P. Milani, Italy
Jeffrey A. Mills, USA
Wendy G. Mitchell, USA
Engy Mogahed, Egypt
Sue Moloney, Australia
Maria C. Mondardini, Italy
Joaquim Monteiro, Portugal
Jaume Mora, Spain
Raysa Morales-Demori, USA
Frank Moriarty, Ireland
James Morse, New Zealand
Laura Moschino, Italy
Katalin Müller, Hungary
Simon Nadel, UK
Michael Narvey, Canada
Charlotte Naslund-Koch, Denmark
Miguel Nogueira, Portugal
Allison Norton, USA
Seigo Okada, Japan
Christine Olbjørn, Norway
Maryam Oskoui, Canada
Belén Pérez Solans, USA
Clare E. Pain, UK
Lilian Palma, Brazil
Michael Perch, Denmark
B. Ryan Phelps, USA
Maria Pokorska-Śpiewak, Poland
Andi A.W. Ramlan, Indonesia
Laura B. Ramsey, USA
David M. Reith, New Zealand
Michael J. Rieder, Canada
Claudio Romano, Italy
Jumpei Saito, Japan
Ryuichi Sakate, Japan
Neha Santucci, USA
Miguel Saps, USA
Markus Schmugge, Switzerland
Siti Maisharah Sheikh Ghadzi, Malaysia
Dror Shouval, Israel
Tor T. Shwayder, USA
Jose Maria Sistac Ballarín, Spain
Ajoke Sobanjo ter Meulen, USA
Daniela C. Souza, Brazil
Monika Sparber-Sauer, Germany
Alessandro Squizzato, Italy
Ashley Stark, USA
Blanka Stiburkova, Czech Republic
Matthew L. Stoll, USA
Michael V. Tatley, New Zealand
Celine Thibault, Canada
Paul S. Thornton, USA
Caroline Tianeze de Castro, Brazil
Paolo Tomasi, The Netherlands
Rosa Torres, Spain
Van L. Tran, USA
Helmut Trimmel, Austria
Ivana Trivic, Croatia
Andrea Trombetta, Italy
Daniel S. Tsze, USA
Stefano Tumini, Italy
Tim Ulinski, France
Benjamin Ungar, USA
Wendy van der Wekken-Pas, The Netherlands
Steven Van Laecke, Belgium
David van Mater, USA
Benedetto Vitiello, Italy
Yukihiro Wada, Japan
Charlotte M. Walter, USA
Forrest Williamson, USA
Stefan Willmann, Germany
Marco Yamazaki‐Nakashimada, Mexico
Jingjing Ye, USA
And, of course, I am also very grateful to the members of the journal’s Honorary Editorial Board, who have acted as peer reviewers and authors, and have provided guidance on journal content, policy and processes:
Karel Allegaert, Leuven, Belgium
Marina Aloi, Rome, Italy
John Berkenbosch, Louisville, KY, USA
Frank M. C. Besag, London, UK
Stan L. Block, Bardstown, KY, USA
Andrew Bush, London, UK
Rolando Cimaz, Florence, Italy
Saskia de Wildt, Rotterdam, the Netherlands
Kevin J. Downes, Philadelphia, PA, USA
L. Lee Dupuis, Toronto, ON, Canada
Geneviѐve Durrieu, Toulouse, France
Brian Eley, Cape Town, South Africa
Susanna Esposito, Parma, Italy
Carmen Ferrajolo, Napoli, Italy
Robert L. Findling, Baltimore, MD, USA
Dominic Fitzgerald, Westmead, Australia
Teresa Giani, Milan, Italy
Jennifer E. Girotto, Hartford, CT, USA
David A. Gremse, Mobile, AL, USA
Andreas H. Groll, Münster, Germany
Scott A. Halperin, Nova Scotia, Canada
Gregory L. Holmes, Burlington, VT, USA
Daniel B. Horton, New Brunswick, NJ, USA
Gregory L. Kearns, Little Rock, AR, USA
Nai Ming Lai, Kuala Lumpur, Malaysia
Lynne Levitsky, Boston, MA, USA
Stuart M. MacLeod, Vancouver, Canada
Christoph Male, Vienna, Austria
Antje Neubert, Erlangen, Germany
Terry Nolan, Parkville, Australia
Michael E. Pichichero, Rochester, NY, USA
Michele Ramien, Calgary, Canada
Emmanuel Roilides, Thessaloniki, Greece
Catherine M. Sherwin, Salt Lake City, UT, USA
Robert Sidbury, Seattle, WA, USA,
William M. Splinter, Ottawa, Canada
John van den Anker, Washington, DC, USA
Diana A. van Riet-Nales, Utrecht, the Netherlands
Karen D. Wagner, Galveston, TX, USA
Hervé Walti, Paris, France
Ian C. Wong, London, UK
C. Michel Zwaan, Rotterdam, the Netherlands
We hope that you have found the articles published throughout 2022 to be both interesting and informative. The editors have appreciated the high quality of content contributed to the journal this year and look forward to keeping you up to date with topical issues in the pediatric pharmacotherapy field in 2023.
With best wishes from the staff of Pediatric Drugs.
Rod
| 0 | PMC9742009 | NO-CC CODE | 2022-12-14 23:29:58 | no | Paediatr Drugs. 2022 Dec 12;:1-4 | utf-8 | Paediatr Drugs | 2,022 | 10.1007/s40272-022-00552-9 | oa_other |
==== Front
Soc Netw Anal Min
Soc Netw Anal Min
Social Network Analysis and Mining
1869-5450
1869-5469
Springer Vienna Vienna
998
10.1007/s13278-022-00998-2
Original Article
A reliable sentiment analysis for classification of tweets in social networks
AminiMotlagh Masoud 1
http://orcid.org/0000-0002-6042-0993
Shahhoseini HadiShahriar [email protected]
1
Fatehi Nina 2
1 grid.411748.f 0000 0001 0387 0587 School of Electrical Engineering, Iran University of Science and Technology, Tehran, Iran
2 grid.254444.7 0000 0001 1456 7807 Department of Electrical and Computer Engineering, Wayne State University, Detroit, USA
12 12 2022
2023
13 1 727 11 2021
5 11 2022
6 11 2022
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
In modern society, the use of social networks is more than ever and they have become the most popular medium for daily communications. Twitter is a social network where users are able to share their daily emotions and opinions with tweets. Sentiment analysis is a method to identify these emotions and determine whether a text is positive, negative, or neutral. In this article, we apply four widely used data mining classifiers, namely K-nearest neighbor, decision tree, support vector machine, and naive Bayes, to analyze the sentiment of the tweets. The analysis is performed on two datasets: first, a dataset with two classes (positive and negative) and then a three-class dataset (positive, negative and neutral). Furthermore, we utilize two ensemble methods to decrease variance and bias of the learning algorithms and subsequently increase the reliability. Also, we have divided the dataset into two parts: training set and testing set with different percentages of data to show the best train–test split ratio. Our results show that support vector machine demonstrates better outcomes compared to other algorithms, showing an improvement of 3.53% on dataset with two-class data and 7.41% on dataset with three-class data in accuracy rate compared to other algorithms. The experiments show that the accuracy of single classifiers slightly outperforms that of ensemble methods; however, they propose more reliable learning models. Results also demonstrate that using 50% of the dataset as training data has almost the same results as 70%, while using tenfold cross-validation can reach better results.
Keywords
Social networks analysis
Sentiment analysis
Data mining
Text mining
issue-copyright-statement© Springer-Verlag GmbH Austria, part of Springer Nature 2023
==== Body
pmcIntroduction
Social networks (SNs) are becoming increasingly popular platforms among people all across the world, and nowadays they are utilized even more than ever. With the growth of SNs like Twitter and increasing their popularity, people share more personal emotions and opinions about various issues in such networks. This rapid growth of SNs, combined with the accessibility of a large amount of data on a multitude of topics, provides a great research potential for a wide range of applications, such as customer analysis, product analysis, sector analysis and digital marketing (Bhatnagar and Choubey 2021; Fatehi, et al. 2022). In addition, identifying users' polarities and mining their opinions shared in various areas, especially SNs, have become one of the most popular and useful research fields. Social media platforms are able to build rich profiles from the online presence of users by tracking activities such as participation, messaging, and Web site visits (Cui, et al. 2020). By an increased growth in the number of users in the SNs and subsequently exponential rise in the interactions between them, large volumes of user-generated content are produced. It is difficult to analyze all these data since most of the social media data are unstructured and dynamic data which frequently alters. Social network analysis provides innovative techniques to analyze interactions among entities by emphasizing on social relationships (Kumar and Sinha 2021). Nowadays, analyzing SNs with data mining and machine learning algorithms has become a must-have strategy for obtaining useful data. Data mining is the process of extracting and identifying useful patterns and relationships from piles of data sets that may lead to the extraction of new information by using data analysis tools (Keyvanpour, et al. 2020).
Among different SNs, twitter is one of the most studied SNs for social networks' research. Twitter is a SN that enables users to share their daily emotions and opinions. It is considered a convenient platform for users to share personal messages, pictures, and videos. One of the main advantages of platforms like twitter is that users are organized in these platforms, making this possible to investigate groups of people or communities who are united by common interests, rather than individual profiles. Furthermore, this is possible through extensive use of hashtags, mentions, and retweets that form a complex network, which can provide us with a rich source of data to analysis. Twitter is known to be a novel source of data for those studying attitudes, beliefs, and behaviors of consumers and opinion makers (Islam, et al. 2020; Kwak and Grable 2021).
Among all various forms of communications, text messages are considered one of the most conspicuous forms, since users can express their opinions and emotions on various and diverse topics using text. Text mining is the process of exploring and transforming unstructured text data into structured data to find meaningful insights. It is defined as a multi-purpose research method to study a wide range of issues by systematically and objectively identifying characteristics of large sample data. Text mining is a sub-field of data mining and an extension of classical data mining methods, which can be applied when making sophisticated formulations using text classification and clustering procedures (Yang, et al. 2021). Hossny, et al. 2020 listed the key challenges for analyzing the text on Twitter including the tweet’s length, frequent use of abbreviations, misspelled words and acronyms, transliterating non-English words using Roman scripts, ambiguous semantics and synonyms.
Information in several social media platforms, like blogs, reviewing SNs, and Twitter, is being processed for extracting people’s opinions about a particular product, organization, or situation. The attitude and feelings comprise an essential part in evaluating the behavior of an individual that is known as sentiments. These sentiments can further be analyzed using a field of study, known as sentiment analysis (SA) (Singh, et al. 2021). SA belongs to the area of natural language processing (NLP) (Chen, et al. 2020) and it has been an active research topic in NLP, which is a cognitive computing study of people’s opinions, sentiments, emotions, appraisals, and attitudes toward entities such as products, services, organizations, individuals, issues, events, topics, and their attributes (Dai, et al. 2021). Also, it aims to analyze and extract knowledge from the subjective information published on the Internet (Basiri, et al. 2021). Sentiment analysis of user-generated data is very useful to know the opinion of the crowd. Two main approaches for sentiment analysis of text documents are described in the literature, specifically approaches based on machine learning and approaches based on symbolic techniques. Symbolic techniques use lexicons and other linguistic resources to determine the sentiment of a given text. Some research has used machine learning for classifying the sentiment of a given text, sometimes following the approach of most symbolic techniques and seeking to identify positive, negative and neutral categories, but sometimes also considering other sentiment categories such as anger, joy and sadness (Moutidis and Williams 2020).
The SA plays significant role in many domain by extracting the people’s emotions which then assist business industry to be developed accordingly. In this study, we investigate the performance of different ML models to analyze the sentiment of two real datasets.
So, the contributions in this paper are summarized as follows:We generate and preprocess two real datasets extracted with Twitter Application Programming Interface (API)—binomial and polynomial—to investigate the sentiment analysis. Binomial dataset incorporating two polarities of positive and negative which is the typical dataset used in the literature, polynomial dataset, however, includes three positive, negative, and neutral polarities.
We investigate the performance of sentiment classification in terms of accuracy /AUC and accuracy/kappa for four classifiers on both binomial and polynomial datasets, respectively.
To increase the reliability of SA and reduce variance and bias of learning models, we apply ensemble methods on both the binomial and polynomial datasets and then report the accuracy values for these methods.
To find out the best train–test split ratio in addition to K-fold cross-validation, we divide the dataset into two parts: training set and testing set with different percentages of data.
The rest of this paper is structured as follows: Sect. 2 reviews some of the related works in the literature. A description of the methodology that includes data collection, preprocessing for sentiment analysis, sentiment detection, and classification modelling is presented in Sect. 3. The results are presented and discussed in Sect. 4, and eventually, the conclusion is detailed in Sect. 5.
Related work
Researchers in the field of sentiment analysis have been mostly used supervised machine learning algorithm for primary classification, such as the work done by Chauhan et al. (2020). Furthermore, many of the recent studies use Twitter as the primary source of data (Al-Laith, et al. (2021), Yadav, et al. (2021)).
Henríquez and Ruz (2018) used a non-iterative deep random vectorial functional link called D-RVFL. They analyzed two different datasets. Dataset 1 contains a collection of 10,000 tweets from the Catalan referendum of 2017 and dataset 2 contains a collection of 2187 tweets from the Chilean earthquake of 2010. They consider the datasets as a two-class classification problem with the labels of positive and negative. By the use of D-RVFL, results show the best performance compared to SVM, random forest, and RVFL.
Ankit and Saleena (2018) proposed an ensemble classification system formed by different learners, such as naive Bayes, random forest, SVMs, and logistic regression classifiers. Their system employs two algorithms: the first algorithm calculates a positive and a negative score for the tweet, and the second algorithm utilizes these scores to predict the sentiment of that tweet. Furthermore, the dataset consists of 43,532 negative and 56,457 positive tweets.
Symeonidis et al. (2018) evaluated the preprocessing techniques on their resulting classification accuracy and the number of features they produce. However, this paper worked on lemmatization, removing numbers, and replacing contractions techniques, while the detection accuracy is low. For this task, they used four classification algorithms named logistic regression, Bernoulli Naive Bayes, linear SVC, and convolutional neural networks on two datasets with the classes of positive, negative, and neutral.
Sailunaz and Alhajj (2019), used a dataset to detect sentiment and emotion from tweets and their replies and measured the influence scores of users based on various user-based and tweet-based parameters. The dataset also includes replies to tweets, and the paper introduces agreement score, sentiment score and emotion score of replies in influence score calculation.
Ruz, et al. (2020), reviewed five classifiers and assessed their performances on two Twitter datasets of two different critical events. Their datasets were Spanish, and they concluded that there is no difference between the behavior of support vector machine (SVM) and random forest in English and Spanish. In order to automatically control the number of edges supported by the training examples in the Bayesian network classifier, they adopt a Bayes factor approach, yielding more realistic networks.
Wang et al. (2021) proposed a system for general population sentiment monitoring from a social media stream (Twitter), through comprehensive multilevel filters, and improved latent Dirichlet allocation (LDA) method for sentiment classification. They reached an accuracy of 68% for general sentiment analysis using real-world content. Also, they used a dataset with three categories (positive, negative, and neutral) and a dataset with four categories (positive, negative, neutral and junk).
Ali et al. (2021) utilized the bilingual (English and Urdu) data from Twitter and NEWS websites to do the sentiment and emotional classification using machine learning and deep learning models. Kaur and Sharma (2020) used API to collect beneficial-related corona virus tweets and then categorized them in three groups (positive, negative, and neutral) to investigate the feeling of people about the COVID-19 pandemic. Nuser et al. (2022) proposed an unsupervised learning framework based on serial ensemble of some hierarchical clustering methods for sentiment analysis on a binomial dataset collected from Twitter.
Machuca et al. (2021) used English COVID-19 pandemic tweets to do the sentiment analysis using a logistic regression algorithm on a binomial dataset including positive and negative labels.
In Table 1, we present a review of the state-of-the-art and their reported accuracy for the sentiment classification with data structures of binomial (positive and negative) and polynomial (positive, negative, and neutral).Table 1 Comparison of sentiment analisys approches
Paper Dataset structure Reported accuracy (%)
Henríquez and Ruz (2018) Binomial 82.90
Ankit and Saleena (2018) Binomial 75.81
Symeonidis, et al. (2018) Polynomial 67.30
Sailunaz, et al. (2019) Polynomial 66.86
Ruz, et al. (2020) Binomial 81.20
Wang, et al. (2021) Polynomial 68.00
Al-Laith, et al (2021) Polynomial 69.40
Nuser, et al. (2022) Binomial 73.75
Ali, et al. (2021) Polynomial 80.00
Machuca, et al. (2021) Binomial 78.50
Methodology
This section introduces our research framework in four phases: data collection, preprocessing, sentiment detection, and classification modeling (Fig. 1).Fig. 1 Overview of proposed sentiment classification workflow
Data collection
Twitter is among the most popular social networking platforms nowadays. It provides its users with a platform to share their daily lives with other users and express their opinions about different national, international issues from various perspectives. Every user can write a short text called tweet with a maximum length of 140 characters. These opinions and comments can be used to raise public awareness to help the government and enterprises understand the views of the public. Twitter also can be used to predict event trends. Therefore, tweets are an important resource to study public awareness.
Researchers and practitioners can access Twitter data using Twitter API. Search and streaming APIs allow them to collect Twitter data using different types of queries, including keywords and user profiles, which has offered them opportunities to access the data needed to analyze challenging problems in diverse domains. Thus, many researchers and practitioners have begun to focus on Twitter data mining to obtain more research value and business value from this research (Li et al. 2019).
For our experiments, in order to collect tweets, we selected a few recent events and issues; search keywords about corona virus like #covid-19, #coronavirus. For our experiments, in order to collect tweets, we selected a few recent events and issues; search keywords about corona virus like #covid-19, #coronavirus, #covid19vaccine, etc. A total of 14,000 tweets were extracted using Twitter API. 6980 of which were written in English; therefore, we picked these tweets. These tweets were sentences; consequently, we had to preprocess these sentences and convert them to a set of words. Then, these words were classified to be understood by the classifier. In the following sections, we elaborate the mentioned procedure.
Preprocessing
Tweets are sometimes not in a usable format, for instances they include characters, symbols or emoticons. Therefore, we need to format them in an appropriate usable form to be able to extract meaningful opinions from them. As a first step in preprocessing, most (if not all) studies apply tokenization. Tokenization is a task for separating the full text string into a list of separate words. Tokenization is defined as a kind of lexical analysis that breaks a stream of text up into words, phrases, symbols, or other meaningful elements called tokens. At its core, the process of tokenization is a standard method for further natural language processing (NLP) transformation in preprocessing (Symeonidis, et al. 2018). For the preprocessing steps, various methods have been proposed and can be applied for data cleaning. Following are the steps in the data preprocessing that we used in this article:All non-English tweets are eliminated.
User names preceded by ‘@’ and external links are omitted.
All hashtags (only the # symbol) are removed.
Stop-words or useless words are removed from the tweet.
All emoticons were removed (i.e.,:-),:-( etc.).
All the tweets were converted to lower case to make the dataset uniform.
Detection
Each tweet should be labeled with sentiment with three possible values: negative, neutral, or positive. The first step to label the tweets is to apply unsupervised methods due to the large dataset we have. For this purpose, we used the TextBlob library in the python programming language to label tweets. This library assigns each tweet a number between − 1 and + 1 (-1 is the most negative and + 1 is the most positive value). Then, we double-checked the labels manually. Tweets between [− 1, − 0.1], [− 0.1, + 0.1] and [+ 0.1, + 1] were labeled negative, neutral, and positive, respectively. Figure 2 illustrates the results from the sentiment analysis. Also, the number of tweets in each class is shown in Table 2. We have a total of 6980 tweets: 977 of which are negative, 3689 of which are neutral and positive tweets are 2314.Fig. 2 Sentiment proportion of dataset
Table 2 Dataset structure
Number of tweets in dataset
Positive 2314
Neutral 3689
Negative 977
Total 6980
Classification modelling
For our experiment and in order to make a comparative analysis, we employed four classifiers, which are the most widely used classifiers for sentiment analysis, namely (1) K-nearest neighbor (KNN), (2) decision tree (DT), (3) support vector machine (SVM), (4) Naive Bayes (NB), and also two ensemble methods including voting and bagging.
K-nearest neighbor
The logic behind KNN classification is that we expect a test sample X to have the same label as the training sample located in the local region surrounding X denoting by K. Training a KNN classifier simply consists of determining K. KNN simply memorizes all samples in the training set and then compares the test sample with them.
Decision tree
The decision tree is a particularly efficient method of producing classifiers from data. It is a tree-like collection of nodes intended to create a decision on values affiliation to a class or an estimate of a numerical target value. Each node represents a splitting rule for one specific attribute. For classification, this rule separates values belonging to different classes. The building of new nodes is repeated until the stopping criteria are met. A prediction for the class label attribute is determined depending on the majority of examples which reached this leaf during generation.
Support vector machine
An SVM is a supervised learning algorithm creating learning functions from a set of labeled training data. Support vector machine solves the traditional text categorization problem effectively. The main principle of SVMs is to determine a linear separator that separates different classes in the search space with a maximum distance. SVM’s classification function is based on the concept of decision planes that define decision boundaries between classes of samples. The main idea is that the decision boundary should be as far away as possible from the data points of both classes. There is only one that maximizes the margin.
Naive Bayes
The naive Bayesian method is one of the most widely used methods for text data classification. The naive Bayesian is a simple probabilistic classifier that uses the concept of mixture models to perform classification. The mixture model relies on the assumption that each of the predefined classes is one of the components of the mixture itself. The components of the mixture model denote the probability of belongingness of any term to the particular component. Naive Bayes classifier uses the concept of Bayes theorem and finds the maximum prospect of the probability of any word fitting to a particular given or predefined class. This algorithm assumes that the elements in the dataset are independent from each other and their occurrences in different datasets indicate their relevance to certain data attributes (Desai and Mehta 2016). This method is a high-bias, low-variance classifier, and it can build a good model even with a small data set. Typical use cases involve text categorization, including spam detection, sentiment analysis, and recommender systems.
Ensemble methods
Ensemble methods are learning algorithms which by try to improve the predicted performance by employing a set of learning algorithms. They reduce bias and variance of the model and so are more reliable compared to the single classifier (Dietterich 2000). The voting method can be used with different combination sets of the classifiers; therefore, we applied the voting method with the combination set of all classifiers to get the maximum value for accuracy. We also used the bagging method with DT (generally this amalgamation has shown a better performance) and bagging with SVM, KNN, and NB.
Evaluation metric
Accuracy
accuracy=TP+TNTP+TN+FP+FN
TP, TN, FP, and FN are the number of true positive, true negative, false positive, and false negative.
AUC
The area under the curve (AUC) is the measure of the ability of a classifier to distinguish between classes and is used as a summary of the receiver operator characteristic (ROC) curve. The higher the AUC, the better the performance of the model at distinguishing between the positive and negative classes.
Kappa
Kappa is a metric that provides a comparison between observed accuracy and expected accuracy.
To start the classification, we divided the dataset into a training set and a testing set with different percentages of data. Common ratios used are 70% or 60% of the dataset for training and 30% or 40% for testing. In our experiment, we used different train–test split percentage, which are 10% to 70%. Continuing the classification, we also used K-fold cross-validation (K-FCV) with K = 10 to generate the training set and the testing set and compare the results with above-mentioned split ratios.
In this paper, first, the above-mentioned classifiers were applied to a dataset with just negative and positive tweets (binomial), and then, the classifiers were applied to a dataset including negative, positive, and neutral tweets (polynomial).
Result analysis
This section gives an overview of the accuracy rates of different trained classifiers. All the calculations are done in the RapidMiner Studio application.
Table 3 shows the predicted accuracy of all classifiers when the tweets are binomial. Our results in Table 3 demonstrate that K-FCV with k = 10 has the highest accuracy rate, except DT, besides the accuracy when we use the train–test split procedure. SVM with 86.42% in single methods and voting with 86.75% in ensemble methods has the best accuracy rates. In Table 4, we can see the differences between the accuracy rates. In most algorithms, there is some decrease in accuracy rate when we used 60% of the dataset for training data. Also, this decrease can be seen when 40% of the dataset is used for training in some methods. Furthermore, in all methods when the ratio is 20%, there is the most increase in accuracy rate in comparison with the ratio of 10%. NB algorithm with + 9.15% and bagging with NB with + 9.62% have the most variation in accuracy rate from 10 to 70% train–test split percentages of the dataset.Table 3 Sentiment accuracy comparison on binomial dataset
Algorithm Train–test split percentage 10-FCV
10% 20% 30% 40% 50% 60% 70%
KNN 73.26 76.45 78.86 78.98 82.13 81.47 82.37 82.89
DT 74.34 75.96 76.65 76.34 77.26 76.23 77.61 76.39
SVM 76.47 78.58 80.90 83.28 83.65 84.13 85.21 86.42
NB 71.54 76.79 77.30 79.03 80.67 80.49 81.05 81.43
Voting 76.23 80.14 82.16 83.89 85.35 85.35 85.71 86.75
Bagging (KNN) 73.63 76.95 78.82 79.38 81.95 82.08 82.57 82.86
Bagging (DT) 75.15 76.34 77.34 76.75 78.05 76.84 78.32 76.85
Bagging (SVM) 76.33 78.05 80.86 82.57 83.53 83.90 85.31 86.08
Bagging (NB) 71.64 76.98 77.21 79.28 80.67 80.64 81.26 81.46
Table 4 Sentiment accuracy differences on binomial dataset
Algorithm Variation between train–test split percentages
10–20% 20–30% 30–40% 40–50% 50–60% 60–70%
KNN + 3.19 + 2.41 + 0.12 + 3.15 − 0.66 + 0.90
DT + 1.62 + 0.69 − 0.31 + 0.92 − 1.03 + 1.38
SVM + 2.11 + 2.32 + 2.38 + 0.37 + 0.48 + 1.08
NB + 5.25 + 0.51 + 1.73 + 1.64 -0.18 + 0.56
Voting + 3.19 + 2.02 + 1.73 + 1.46 0.00 + 0.36
Bagging (KNN) + 3.32 + 1.87 + 0.56 + 2.57 + 0.13 + 0.49
Bagging (DT) + 1.19 + 1.00 − 0.59 + 1.30 − 1.21 + 1.48
Bagging (SVM) + 1.72 + 2.81 + 1.71 + 0.96 + 0.37 + 1.41
Bagging (NB) + 5.34 + 0.23 + 2.07 + 1.39 − 0.03 + 0.62
Table 5 shows the predicted AUC for binomial dataset. SVM and bagging with SVM have the best values. We can also see that the 10-FCV has better results than the split procedure. From Table 6, the results show that there is some reduction when we use 60% of the dataset for training data than 50%. An increase in AUC from 10 to 20% of the dataset is more than other ratios.Table 5 Sentiment AUC comparison on binomial dataset
Algorithm Train–test split percentage 10-FCV
10% 20% 30% 40% 50% 60% 70%
KNN 0.749 0.800 0.828 0.845 0.863 0.871 0.868 0.876
DT 0.579 0.579 0.610 0.604 0.619 0.601 0.625 0.604
SVM 0.793 0.847 0.878 0.897 0.917 0.913 0.929 0.932
NB 0.495 0.550 0.556 0.608 0.643 0.637 0.655 0.601
Voting 0.598 0.670 0.704 0.731 0.779 0.745 0.761 0.794
Bagging (KNN) 0.741 0.792 0.825 0.839 0.861 0.865 0.861 0.877
Bagging (DT) 0.618 0.637 0.641 0.624 0.652 0.651 0.647 0.638
Bagging (SVM) 0.795 0.849 0.879 0.898 0.918 0.917 0.929 0.934
Bagging (NB) 0.706 0.753 0.768 0.787 0.821 0.813 0.817 0.824
Table 6 Sentiment AUC differences on binomial dataset
Algorithm Variation Between Train–Test Split Percentages
10–20% 20–30% 30–40% 40–50% 50–60% 60–70%
KNN + 0.051 + 0.028 + 0.017 + 0.180 + 0.008 -0.003
DT 0.000 + 0.031 − 0.006 + 0.015 − 0.018 + 0.024
SVM + 0.054 + 0.031 + 0.019 + 0.020 − 0.004 + 0.016
NB + 0.055 + 0.006 + 0.052 + 0.035 − 0.006 + 0.018
Voting + 0.072 + 0.034 + 0.027 + 0.048 − 0.034 + 0.016
Bagging (KNN) + 0.051 + 0.033 + 0.014 + 0.022 + 0.004 − 0.004
Bagging (DT) + 0.019 + 0.004 − 0.017 + 0.028 − 0.001 − 0.004
Bagging (SVM) + 0.054 + 0.030 + 0.019 + 0.020 − 0.001 + 0.012
Bagging (NB) + 0.047 + 0.015 + 0.019 + 0.034 − 0.008 + 0.004
The classification continued with the polynomial dataset. So we applied classifiers to the dataset with three categories including positive, negative, and neutral tweets. Tables 7, 8, 9, 10 show the comparison between classifiers in terms of accuracy and kappa metrics when the tweets are polynomial. According to Tables 7, 8, 9, 10, there is some reduction in accuracy and kappa rates when we use 60% of the dataset for training data than 50% in most classifiers, and in some cases we have just a little increase in the accuracy and kappa rates. SVM and bagging with SVM have better results compared to other classifiers. SVM with an accuracy of 73.91% is the better choice for polynomial classification. However, the bagging with SVM is a more reliable model compared to SVM, employing the ensemble method. This technique makes the learning model more reliable by reducing variance and bias. Tables 7 and 10 show that the most positive variation has happened from 10 to 20% of the dataset in both accuracy and kappa terms.Table 7 Sentiment accuracy comparison on polynomial dataset
Algorithm Train–test split percentage 10-FCV
10% 20% 30% 40% 50% 60% 70%
KNN 57.02 59.55 61.46 62.79 63.72 64.09 64.95 66.50
DT 54.08 54.94 54.89 54.72 55.47 55.25 55.73 55.49
SVM 61.73 65.29 67.56 69.09 70.69 71.00 71.97 73.91
NB 54.14 57.27 57.90 58.90 60.69 60.08 60.89 61.69
Voting 58.48 61.19 62.98 64.60 65.93 65.81 66.76 68.30
Bagging (KNN) 56.37 59.06 60.52 62.17 63.32 63.62 64.37 66.54
Bagging (DT) 54.36 54.96 54.56 55.22 55.47 55.25 55.87 55.56
Bagging (SVM) 61.72 65.28 67.56 68.98 70.72 70.96 71.97 73.87
Bagging (NB) 54.17 57.18 58.00 58.97 60.74 60.08 61.03 61.92
Table 8 Sentiment accuracy differences on polynomial dataset
Algorithm Variation between train–test split percentages
10–20% 20–30% 30–40% 40–50% 50–60% 60–70%
KNN + 2.53 + 1.91 + 1.33 + 0.93 + 0.37 + 0.86
DT + 0.86 − 0.05 − 0.17 + 0.75 − 0.22 + 0.48
SVM + 3.56 + 2.27 + 1.53 + 1.60 + 0.31 + 0.97
NB + 3.13 + 0.63 + 1.00 + 1.79 − 0.61 + 0.81
Voting + 2.71 + 1.79 + 1.62 + 1.33 − 0.12 + 0.95
Bagging (KNN) + 2.69 + 1.46 + 1.65 + 1.15 + 0.30 + 0.75
Bagging (DT) + 0.60 − 0.40 + 0.66 + 0.25 − 0.22 + 0.62
Bagging (SVM) + 3.56 + 2.28 + 1.42 + 1.74 + 0.24 + 1.01
Bagging (NB) + 3.01 + 0.82 + 0.97 + 1.77 -0.66 + 0.95
Table 9 Sentiment Kappa comparison on polynomial dataset
Algorithm Train–test split percentage 10-FCV
10% 20% 30% 40% 50% 60% 70%
KNN 0.108 0.173 0.221 0.253 0.275 0.284 0.306 0.341
DT 0.042 0.064 0.063 0.058 0.077 0.070 0.083 0.077
SVM 0.225 0.310 0.363 0.398 0.433 0.441 0.463 0.504
NB 0.247 0.298 0.315 0.335 0.369 0.362 0.377 0.399
Voting 0.150 0.218 0.261 0.300 0.330 0.328 0.351 0.384
Bagging (KNN) 0.090 0.160 0.196 0.237 0.265 0.272 0.292 0.343
Bagging (DT) 0.051 0.066 0.053 0.073 0.077 0.070 0.087 0.079
Bagging (SVM) 0.225 0.310 0.363 0.396 0.434 0.440 0.463 0.503
Bagging (NB) 0.247 0.296 0.316 0.336 0.370 0.362 0.379 0.398
Table 10 Sentiment Kappa differences on polynomial dataset
Algorithm Variation between train–test split percentages
10–20% 20–30% 30–40% 40–50% 50–60% 60–70%
KNN + 0.065 + 0.048 + 0.032 + 0.022 + 0.009 + 0.022
DT + 0.022 − 0.001 − 0.005 + 0.019 − 0.007 + 0.013
SVM + 0.055 + 0.053 + 0.035 + 0.035 + 0.008 + 0.022
NB + 0.051 + 0.017 + 0.020 + 0.034 − 0.007 + 0.015
Voting + 0.068 + 0.043 + 0.039 + 0.030 − 0.002 + 0.023
Bagging (KNN) + 0.070 + 0.063 + 0.041 + 0.028 + 0.007 + 0.020
Bagging (DT) + 0.015 -0.013 + 0.020 + 0.004 -0.007 + 0.017
Bagging (SVM) + 0.085 + 0.053 + 0.033 + 0.038 + 0.006 + 0.023
Bagging (NB) + 0.049 + 0.020 + 0.020 + 0.034 − 0.008 + 0.017
From the results of accuracy and AUC on the binomial dataset (Tables 3, 4, 5, 6) and the results of accuracy and kappa on the polynomial dataset (Tables 7, 8, 9, 10), we can observe that SVM and bagging with SVM have better results compared to other classifiers. However, the accuracy of polynomial classification is less than binomial. The reason of over-performing of SVM can be the fact the text data have a sparse nature. In such type of data, there are few irrelevant features that tend to have a correlation with each other. This leads those features to turn into some distinct categories, which can be separated by linear separators. Also, we can see most of the classifiers in 50% train–test split percentage have almost the same results as 70% in accuracy (Figs. 3 and 4), AUC and kappa rates, while using 10-FCV can reach better results.Fig. 3 Classification accuracy on binomial dataset
Fig. 4 Classification accuracy on polynomial dataset
We also compared the performance of SVM, when 10-FCV is imposed, with state of the art presented in Table 1. The results showed that overall accuracy has improved at least 3.52% and 5.91% on binomial and polynomial datasets, respectively. This improvement can be a result of using the training and testing data divided through the K-fold cross-validation method.
Conclusion
In this paper, we aimed to analyze the sentiment of social media data, specifically Twitter, using both single classifiers and ensemble models combined with single classifiers on two datasets including binomial (positive and negative) and polynomial (positive, negative, and neutral) datasets.
From the results, we observed that data mining is a good choice for sentiment prediction since the accuracy rates are relatively high values. We also reviewed four classifiers, including SVM, K-nearest neighbor, decision tree and naive Bayes and two bagging ensemble methods.
From the results, we concluded that among single classifiers and their combination with the ensemble methods, SVM reached 3.53% and 7.41% over performances on binomial and polynomial datasets, respectively. Although ensemble methods do not show over performance compared to single methods, they are able to decrease the bias or variance of the learning models and also decrease the generalization error. Therefore, there is a trade-off between reliability of the algorithm and accuracy.
Our results show that using 50% of the dataset as training data has almost the same results as 70%; however, using 10-FCV has better results. This conclusion can be seen both in the accuracy and AUC rates in the binomial dataset and accuracy and kappa rates in the polynomial dataset.
In future studies, we will apply other ensemble methods, such as boosting and stacking combined with other classifiers, along with single classifiers. Furthermore, we will attempt to improve our dataset by selecting other keywords including both negative and positive sentiments and increasing the size of the dataset by extracting more tweets.
Authors' contributions
All authors have contributed in this research.
Funding
The authors received no funding to conduct the research.
Declarations
Conflict of interest
The authors declare that they have no competing interests.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Ali MZ, Javed K, Tariq A (2021) Sentiment and emotion classification of epidemic related bilingual data from social media. arXiv preprint arXiv:2105.01468
Al-Laith A Shahbaz M Alaskar HF Rehmat A Arasencorpus: a semi-supervised approach for sentiment annotation of a large arabic text corpus Appl Sci 2021 11 5 2434 10.3390/app11052434
Ankit Saleena N An ensemble classification system for Twitter sentiment analysis Procedia Comput Sci 2018 132 937 946 10.1016/j.procs.2018.05.109
Basiri ME Nemati S Abdar M Cambria E Rajendra AU ABCDM: an attention-based bidirectional CNN-RNN deep model for sentiment analysis Futur Gener Comput Syst 2021 115 279 294 10.1016/j.future.2020.08.005
Bhatnagar S Choubey N Making sense of tweets using sentiment analysis on closely related topics Soc Netw Anal Min 2021 11 44 10.1007/s13278-021-00752-0 33968279
Chauhan UA Afzal MT Shahid A Abdar M Basiri ME Zhou X A comprehensive analysis of adverb types for mining user sentiments on amazon product reviews World Wide Web 2020 23 3 1811 1829 10.1007/s11280-020-00785-z
Chen J Hossain MS Zhang H Analyzing the sentiment correlation between regular tweets and retweets Soc Netw Anal Min 2020 10 13 10.1007/s13278-020-0624-4
Cui R Agrawal G Ramnath R Tweets can tell: activity recognition using hybrid gated recurrent neural networks Soc Netw Anal Min 2020 10 16 10.1007/s13278-020-0628-0
Dai Y Liu J Zhang J Fu H Xu Z Unsupervised Sentiment Analysis by Transferring Multi-source Knowledge Cogn Comput 2021 10.1007/s12559-020-09792-8
Desai M, Mehta MA (2016) Techniques for sentiment analysis of Twitter data: A comprehensive survey. In: 2016 International Conference on Computing, Communication and Automation (ICCCA). 149–154 10.1109/CCAA.2016.7813707
Dietterich TG, (2000) Ensemble methods in machine learning. In: Multiple Classifier Systems. MCS 2000. Lecture Notes in Computer Science. 1857, 1–15. 10.1007/3-540-45014-9_1
Fatehi N Shahhoseini HS Wei J Chang CT An automata algorithm for generating trusted graphs in online social networks Appl Soft Comput 2022 118 108475 10.1016/j.asoc.2022.108475
Henríquez PA, Ruz GA (2018) Twitter Sentiment Classification Based on Deep Random Vector Functional Link. In: 2018 International Joint Conference on Neural Networks (IJCNN), 1–6 10.1109/IJCNN.2018.8489703
Hossny AH Mitchell L Lothian N Osborne G Feature selection methods for event detection in Twitter: a text mining approach Soc Netw Anal Min 2020 10 61 10.1007/s13278-020-00658-3
Islam MR Liu S Wang X Xu G Deep learning for misinformation detection on online social networks: a survey and new perspectives Soc Netw Anal Min 2020 10 82 10.1007/s13278-020-00696-x 33014173
Kaur C, Sharma A, (2020). Twitter sentiment analysis on coronavirus using textblob (No. 2974). EasyChair.
Keyvanpour M Karimi Zandian Z Heidarypanah M OMLML: a helpful opinion mining method based on lexicon and machine learning in social networks Soc Netw Anal Min 2020 10 10 10.1007/s13278-019-0622-6
Kumar P Sinha A Information diffusion modeling and analysis for socially interacting networks Soc Netw Anal Min 2021 11 11 10.1007/s13278-020-00719-7 33456625
Kwak EJ Grable JE Conceptualizing the use of the term financial risk by non-academics and academics using twitter messages and science direct paper abstracts Soc Netw Anal Min 2021 11 6 10.1007/s13278-020-00709-9 33425061
Li X Xie Q Jiang J Zhou Y Huang L Identifying and monitoring the development trends of emerging technologies using patent analysis and Twitter data mining: The case of perovskite solar cell technology Technol Forecast Soc Chang 2019 146 687 705 10.1016/j.techfore.2018.06.004
Machuca CR Gallardo C Toasa RM 2021, Twitter sentiment analysis on coronavirus: Machine learning approach J Phys Conf Series 1828 1 012104
Moutidis I Williams HTP Good and bad events: combining network-based event detection with sentiment analysis Soc Netw Anal Min 2020 10 64 10.1007/s13278-020-00681-4
Ruz GA Henríquez PA Mascareño A Sentiment analysis of Twitter data during critical events through Bayesian networks classifiers Futur Gener Comput Syst 2020 106 92 104 10.1016/j.future.2020.01.005
Sailunaz K Alhajj R Emotion and sentiment analysis from Twitter text J Comput Sci 2019 36 101003 10.1016/j.jocs.2019.05.009
Singh M Jakhar AK Pandey S Sentiment analysis on the impact of coronavirus in social life using the BERT model Soc Netw Anal Min 2021 11 33 10.1007/s13278-021-00737-z 33758630
Symeonidis S Effrosynidis D Arampatzis A A comparative evaluation of pre-processing techniques and their interactions for twitter sentiment analysis Expert Syst Appl 2018 110 298 310 10.1016/j.eswa.2018.06.022
Wang D Al-Rubaie A Hirsch B Pole GC National happiness index monitoring using Twitter for bilanguages Soc Netw Anal Min 2021 11 24 10.1007/s13278-021-00728-0
Yadav N Kudale O Rao A Gupta S Shitole A Hemanth J Bestak R Chen JI-Z Twitter sentiment analysis using supervised machine learning Intelligent Data Communication Technologies and Internet of Things 2021 Singapore Springer 631 642
Yang Y Hsu JH Löfgren K CHo W Cross-platform comparison of framed topics in Twitter and Weibo: machine learning approaches to social media text mining Soc Netw Anal Min 2021 11 75 10.1007/s13278-021-00772-w
Nuser M, Alsukhni E, Saifan A, Khasawneh R, Ukkaz D, (2022) Sentiment analysis of COVID-19 vaccine with deep learning. J Theor Appl Inf Technol. 100(12):4513-4521.
| 0 | PMC9742011 | NO-CC CODE | 2022-12-14 23:29:58 | no | Soc Netw Anal Min. 2023 Dec 12; 13(1):7 | utf-8 | Soc Netw Anal Min | 2,022 | 10.1007/s13278-022-00998-2 | oa_other |
==== Front
J Fam Violence
J Fam Violence
Journal of Family Violence
0885-7482
1573-2851
Springer US New York
485
10.1007/s10896-022-00485-4
Review Article
The Trinidad and Tobago Covid-19 Domestic Violence Victimization and Perpetration Study
http://orcid.org/0000-0002-5164-5103
Wallace Wendell C. [email protected]
1
County Keel [email protected]
2
Mason Russel [email protected]
1
Humphrey April [email protected]
1
1 grid.430529.9 Department of Behavioural Sciences, The University of the West Indies, Carmody Road, St. Augustine, Trinidad and Tobago
2 St. Augustine, Trinidad and Tobago
12 12 2022
112
1 2 2022
4 12 2022
5 12 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Purpose
While there is available scholarship in the Global North on DV victimization and perpetration during the COVID-19 pandemic, there is a dearth of similar scholarship in the Global South. With this in mind, the Trinidad and Tobago COVID-19 Domestic Violence Victimization and Perpetration study was conducted in an attempt to fill that void.
Method
An online questionnaire containing a qualitative component was used to gather data aimed at determining whether DV victimization and perpetration had increased during the COVID-19 pandemic as well as possible contributory factors. Participants were 602 married or cohabiting adult citizens in Trinidad and Tobago.
Results
The results indicated that overall there was an increase in DV perpetration (13%) as well as an increase in DV victimization (16%) among the sampled population. The results also indicated that males (17%) and females (13%) in the sample engaged in increased levels of DV perpetration, while males (25%) and females (12%) were victims of increased DV victimization. Six themes emanated from the qualitative component of the study, namely: isolation/Covid-19 restrictions, lack of assistance for victims, male fear of reporting DV, work as a safe space, mental health effects and job loss.
Conclusion
The findings revealed increased DV perpetration and victimization among the study’s participants. These findings have implications for policymakers in Trinidad and Tobago.
Keywords
Domestic violence
Victimization
Perpetration
COVID-19
Trinidad and Tobago
==== Body
pmcOn 11 March 2020, the World Health Organization (WHO) formally declared the coronavirus (COVID-19) a global health pandemic (WHO, 2020) and called on states to take urgent measures to tackle it. In light of the call by the WHO (2020) and subsequent governmental action (stay-at-home orders, lockdowns, and restrictions of movement) many aspects of pre-Covid-19 life were upended and this included family life. As a result of the outbreak of Covid-19 disease and the subsequent confinement measures aimed at its containment, specialists and advocates worldwide raised concerns regarding increases in DV perpetration and victimization (Kourti et al., 2021; United Nations, 2020). For example, Boserup, McKenney, and Elkbuli (2020) and Evans, Lindauer, and Farrell (2020) submitted that life during the Covid-19 pandemic has the potential to exacerbate the environmental milieus of victims of DV, while Dlamini (2020) argued that the challenges brought about by the Covid-19 pandemic has exacerbated pre-existing toxic social norms which underpin DV.
The views espoused above on DV are premised on the notion that the Covid-19 pandemic has the potential to mimic patterns of past pandemics, where social isolation, associated stress and family conflicts were strongly connected to increases in DV (Piquero et al., 2020) as well as increased risk of familial violence (Gama et al., 2021). The possibility of exacerbated familial and intimate violence are also premised on the notion that victims of DV would not be able to safely contact the police for assistance or receive support from family and friends as stay-at-home orders meant that many victims of DV were now required to stay at home with their abusers and could not distance themselves from their abusers as was possible in the pre-Covid-19 era (Bradbury-Jones & Isham, 2020; Buttell & Ferreira, 2020; Evans, Lindauer, & Farrell, 2020; Kofman & Garfin, 2020; Plasilova et al., 2021). Some scholars view Covid-19 as a shadow pandemic (Davies, Guenfoud, & Jovanovic, 2021; UN Women, 2020), a parallel pandemic of DV (Gama et al., 2021), a pandemic within a pandemic (Evans, Lindauer, & Farrell, 2020) as well as the implacable enemy of DV (Wallace, 2021), hence its omnipresence in the Covid-19 environment.
While extensive research on DV during the present pandemic is scarce (Sharma & Borah, 2020), there is an emerging body of scholarship on the phenomenon. Not surprisingly, the majority of scholarship emanates from the Global North (Bradbury-Jones & Isham 2020; Davies, Guenfoud, & Jovanovic, 2021; Dlamini, 2020; Gama et al., 2021; Kourti et al., 2021; Leslie & Wilson, 2020; Pina, 2021; Sharma & Borah, 2020; van Gelder et al., 2020) to the exclusion of traditionally un- and under-researched jurisdictions, for example, the Caribbean. While this is no fault of researchers in the Global North, the exclusion of, and lack of scholarship on DV perpetration and victimization in the Caribbean during the Covid-19 pandemic inevitably leads to ‘universalizing’ the issue or speciously thinking that problems manifest themselves in the same manner everywhere, while disregarding local features [and cultures] of other approaches (Milan & Treré, 2019).
By mid-March 2020, the Government of Trinidad and Tobago had implemented the Public Health Regulations (2020) to enforce stay at home orders, closure of non-essential businesses, restrict the movement of non-essential personnel, reduced capacity of public transportation vehicles and public gatherings. Added to this, the Public Health Regulations 2020 oversaw the closure of Trinidad and Tobago’s borders and made the wearing of face masks mandatory. A State of Emergency with a stipulated curfew was also declared on May 15th 2021 and this lasted until November 17th 2021. In Trinidad and Tobago, a major facet of pre-Covid-19 life that was altered post Covid-19, was that family members were coerced into spending time together in environments that were often volatile. While this situation is not unique to Trinidad and Tobago, the extremely outgoing culture associated with life on the island (beach-going, carnival, hiking, nature adventures and a host of outdoor activities surrounding religious festivals) was impacted. In the pre-Covid-19 era, such volatile environments could have been avoided, for example, by participating in a plethora of outdoor activities mentioned above or by attending work at the office. However, during the Covid-19 pandemic, outdoor activities were restricted, the office was located at home and this meant little to no escape.
Despite the possibility of increased DV victimization and perpetration as a result of the Covid-19 pandemic, there is little research on the phenomenon in Trinidad and Tobago’s context on the issue under inquiry. In order to negate the universalizing of increased DV during the Covid-19 pandemic, this research focuses on DV in Trinidad and Tobago during the Covid-19 pandemic. Further, there is no doubt that DV is a global Public Health concern that affects millions of individuals regardless of age, race, religion, ethnicity, economic status, sexual orientation, or educational background (ACOG, 2012; Bair-Merritt, 2010). Additionally, DV is not the sole domain of any single jurisdiction and Trinidad and Tobago is not exempted from its clutches. Therefore, this adds to the need for DV to be examined in Trinidad and Tobago’s context.
In Trinidad and Tobago, DV refers to, physical, sexual, emotional or psychological or financial abuse committed by a person against a spouse, child, any other person who is a member of the household or dependant (Domestic Violence Act, 2020) and it is in this context that DV is used in this article. The current study was guided by four research questions (RQ), namely:
RQ1: Did individuals engage in increased domestic violence perpetration in Trinidad and Tobago during the Covid-19 pandemic?
RQ2: Did individuals become victims of increased domestic violence in Trinidad and Tobago during the Covid-19 pandemic?
RQ3: Did fear of violence in intimate relationships increase in Trinidad and Tobago during the Covid-19 pandemic?
RQ4: Were increases in domestic violence perpetration/victimization/fear of victimization a result of family isolation due to the Covid-19 pandemic in Trinidad and Tobago?
Covid-19 and Domestic Violence
Acts of violence in domestic settings are most common, most difficult to monitor and manage, are of great concern and have received considerable amounts of attention in academia, policymaking and legislative agendas globally (Plasilova et al., 2021; Stoianova, Ostrovska, & Tripulskyir, 2020) and the Covid-19 pandemic and attendant lockdown reminded the world of this. For instance, Sharma & Borah (2020), point out that the United States has enacted lockdowns regionally to contain the spread of the virus, while in Italy, citizens were required to stay indoors under stay-at-home orders indefinitely and that lockdowns are a global phenomenon. While social distancing and self-quarantining are viewed as the best ways to protect the general population from the Covid-19 disease, it is argued that this new normal has led to alarming increases in DV (Ali & Khalid, 2021; Hansen & Lory, 2020).
There is both existing and emergent data alluding to increases in DV whenever families are confined or are required to spend more time together, for example, holidays or when children are not attending school (Joshi & Sorenson, 2010; Vazquez, Stohr, & Purkiss, 2005) in similar conditions to existing pandemic conditions. Other research suggest that lockdown measures are increasing the incidence of DV and not only in number, but also in severity (Sharma & Borah, 2020). In a similar vein, van Gelder et al., (2020) posit that cases of DV have increased significantly during the Covid-19 pandemic as physical isolation is now a government-sanctioned approach. It should be noted that physical isolation is a long existing and prominent tool utilized by abusers to distance victims from their support networks (Coohey, 2007; Menjívar & Salcido, 2002) and the Covid-19 pandemic appears to be working in their favor due to stay-at-home orders and restrictions on movement.
Preliminary data emanating from other parts of the world paint a picture of increased DV during the Covid-19 pandemic, for example, Italy (Bellizzi et al., 2020), Mexico, Brazil (Bettinger-Lopez & Bro, 2020), the United Kingdom (UK) (Davidge, 2020), and Victoria, Australian (Pfitzner et al., 2020), with rates of DV recorded in China increasing by 50%, Colombia, by 79%, and Tunisia by 400% (Mlambo-Ngcuka, 2020). Davies et al. (2021) point out that in the U.S., the National Commission on COVID-19 and Criminal Justice reported an 8.1% increase in DV incidents after lockdown orders, while in an editorial, the New York Times (2020) cogitates that domestic abuse is flourishing in the conditions created by the pandemic and acting like an opportunistic infection. Similarly, Thomas (2020) points out that in the United States, the National Domestic Violence Hotline reported a significant surge in calls from DV victims. In a similar vein, McCrary & Sanga (2020) reported that DV increased by 12% on average in their study conducted 14 large U.S. cities. Quite interestingly, using data from Dallas, Texas to examine the extent the lockdown on increases in DV, Piquero et al. (2020) found evidence for a short-term spike in the two weeks immediately after the lockdown was instituted.
In another study conducted by Leslie and Wilson (2020) in 14 large US cities before and after social distancing began, the researchers compared DV calls for service before and after social distancing began. The results of Leslie and Wilson’s (2020) study indicated that the pandemic led to a 7.5% increase in calls for service during March, April, and May of 2020 and that the biggest increase came during the first five weeks after widespread social distancing began, when DV calls increased by 9.7%. Similarly, Campbell (2020) reports that in China, DV is reported to have tripled during their shelter in-place mandate, France witnessed a 30% increase in DV reports, Brazil estimates that DV reports have increased by 40–50%, and Italy has indicated that reports of DV are on the rise as a result of the Covid-19 pandemic. In other parts of the world, the prevalence of DV appears to have increased as DV calls in Argentina have increased 25% since their March 20, 2020 lockdown, while there is a 30% increase in calls to DV helplines in Cyprus and a 33% increase in Singapore (UN Women, 2020).
Scholars have also indicated increased DV victimization and perpetration as a result of the Covid-19 pandemic. For example, in the US, DV victimization increased from 20 to 30% soon after the lockdowns (Kofman & Garfin 2020; Piquero et al., 2020), while domestic and sexual abuse in Spain increased by 20%, DV increased by 30% in Cyprus and by 25% in the United Kingdom (Bradbury & Isham, 2020). According to the World Health Organization (WHO) (2020), global reports from China, the United Kingdom, the United States of America, and other countries suggest a significant increase in DV cases related to the COVID-19 pandemic. Instructively, the literature and data presented above on DV are snapshots of the Western world, are not representative of occurrences in the Caribbean and cannot be ‘universalized’ to the region.
At this juncture, it should be noted that developed countries such as the UK, US, China, France, Cyprus, and Spain have well-established systems for monitoring and reporting DV. However, the United Nations Office on Drugs and Crime (UNODC) (2020) points out in developing countries such systems are nonexistent or inefficient. It is against this background that Ali and Khalid (2021) cogitate that the increase in DV cases during the COVID-19 pandemic has not been adequately addressed in developing countries. The position alluded to by Ali and Khalid (2021) is applicable to Trinidad and Tobago ‘as empirical research on pandemic related DV is lacking’ (Gama et al., 2021).
The Trinidad and Tobago COVID-19 Domestic Violence Victimization and Perpetration (C-19 DVAP) study is undergirded by the opportunity theory as it aids our understanding of increases in the commission of crimes (including DV) during a pandemic. The opportunity theory suggest that pandemic lockdown measures can activate causal mechanisms for an increase in crimes, with some types being more likely to increase and others being more likely to decrease (Eisner & Nivette, 2020), due to the restrictions being imposed on personal mobility and social interactions. The opportunity theory also argues that there is a tendency for property crimes to be reduced when opportunities for the commission of those acts are reduced, however, interpersonal crimes, for example DV, tend to increase.
Methods and Materials
A self-administered questionnaire was utilized to collect data for this study. This approach was utilized as questionnaires are a dependable and rapid technique for gathering information from a large number of participants in an effective and timely manner. Further, questionnaires serve to reduce the ‘social desirability bias’ or the tendency for survey participants to respond favorably whenever researchers are facing them (Tourangeau & Yan, 2007). The use of questionnaires is also critical in large projects such as this study with several complicated objectives and when time is a key constraint (Bell, 2005; Greenfield, 2002; Silverman, 2004). The current research utilized an online questionnaire to garner the public’s views of DV perpetration, victimization and fear of victimization in Trinidad and Tobago during the Covid-19 pandemic. This approach was used as the Covid-19 pandemic is characterized by many mutually reinforcing obstacles that render populations increasingly vulnerable as personal health, wellbeing, and the physical environment are affected (Lupton, 2021).
As the researchers were conducting research under pandemic conditions (Lupton, 2021), they opted to use a non-contact online questionnaire in order to protect the research participants and the researchers. Questions asked to the participants included, but were not limited to: (1) Were you a victim of domestic violence during the Covid-19 pandemic? (2) Did you commit any acts of domestic violence during the Covid-19 pandemic? (3) Did your fear of being a victim of increased domestic violence during the Covid-19 pandemic? (4) What do you think led to your increased domestic violence perpetration/victimization/fear of victimization? and (5) Were the changes in your domestic violence perpetration/victimization/fear of victimization due to family isolation? The study’s methods was reviewed and approved by the Ethics Committee of the Association of Caribbean Criminal Justice Practitioners (ACCJP) as the study was conceptualized and facilitated by that organization.
Instrumentation
The questionnaire used in this study is an adapted variant of a previously validated instrument used by Mumford & Rothman (2020) in their research project ‘Cyber-Abuse Research Initiative’ (CARI). The instrument was modified to reflect the realities of demographic characteristics of Trinidad and Tobago’s population as demographic questions related to ethnicity were removed from Mumford and Rothman’s (2020) questionnaire and replaced with ethnicities related to Trinidad and Tobago’s population. Further, two closed ended questions and one open-ended question were added to Mumford and Rothman’s (2020) questionnaire in an attempt to garner context-based information (for example, area of residence). Despite the modified nature of the instrument, validity and reliability were not affected as the changes were minor. The questionnaire consisted of four sections and nineteen questions. The questionnaire gathered data on participants’ demographics, DV victimization and perpetration during Covid-19, fear of domestic violence victimization during Covid-19 and rationale for victimization, perpetration and fear. An open-ended question was enmeshed in the questionnaire to gather data on factors that contributed to DV during the pandemic. Questions in sections two and three were anchored on a five-point Likert scale.
In seeking to determine whether there were increases in DV perpetration/victimization/fear of victimization during the Covid-19 pandemic in Trinidad and Tobago, baselines of no prior DV victimization/perpetration were established for the participants as they were asked to indicate that they had no previous victimization and/or perpetration experiences. Importantly, the participants were under no obligation to answer the question on their previous victimization experiences. Further, the self-administered nature of the instrument served to diminish the potential for ‘social desirability bias’ or the inclination by survey participants to falsely report a favourable state of affairs when in the presence of a researcher who is collecting data in person or by telephone (Tourangeau & Yan, 2007).
Data Collection
The inclusion criteria for this study were: (1) participants must be eighteen years and older, (2) be citizens of Trinidad and Tobago, (3) be involved in an intimate relationship for any period during the Covid-19 pandemic, and (4) no prior DV perpetration/victimization. Participants were invited to complete the anonymous online survey on their own time based solely on their willingness to participate. Advertisements promoting the study were placed on university and college websites, Facebook, Government websites, and personal websites (LinkedIn). The survey was hosted on Google Forms and invitations to participants were circulated using the Meta Platforms, Incoroprarion, and WhatsApp platforms utilizing a clickable link.
Issues of anonymity, consent, confidentiality, and participant safety were addressed in the questionnaire’s introduction. Participants were asked to complete the instrument at a time and location that is devoid of their spouse’s presence and to not complete the instrument if they felt threatened in any way. Confidentially was guaranteed to the participants as their real names and/or identifying features are not included in the article. A unique Internet Protocol (IP) address was utilized to indicate each submission and avoid multiple individual responses. The questionnaire was available for completion over a fourteen-day period (November 26, 2021 to December 10, 2021) at the back end of the pandemic when the government of Trinidad and Tobago were considering lifting existing pandemic restrictions. With this in mind, the researchers thought it best to conduct the research in a short period when pandemic conditions were prevailing, rather than run the risk of conducting the research under pandemic and non-pandemic conditions.
Data Analysis
The data were analyzed using Statistical Package for Social Sciences (SPSS version 28). Descriptive data are used to present the results by way of frequencies and cross-tabulations. Frequency distribution is used to analyze the demographic data in Section one of the questionnaire to produce frequency counts, percentages, and cumulative percentages for those values in that section (Norušis, 1993). Cross-tabulations are used to examine the relationship between variables in sections two and three of the instrument and this assisted the researchers to make informed research decisions by recognizing relationships between the study’s parameters (Garson, 2013).
To complement the quantitative analyses, the researchers utilized thematic analysis as espoused by Braun and Clarke (2006) to analyze the narratives of the participants in Sect. 4 of the instrument. This approach sought to understand the phenomenon and identify which events led to specific consequences, and this helped to enhance the quantitative data (Tracy, 2019) as well as to evaluate patterns (themes) within the data. Of the 602 participants, 33 persons completed the qualitative component of the instrument. The general terms that emanated from the narratives of the participants were coded into themes (Liamputtong, 2020), their frequency counted and themes organized according to patterns developed from an inductive analysis of the data (Punch, 2014).
The Participants
A total of six hundred and two (n = 602) participants completed the survey instrument. The participants included 408 females (68%) and 185 males (31%). Non-binary and individuals who preferred not to disclose their gender accounted for 1% of the study’s population (n = 9). In terms of age distribution, 43 participants (7%) were aged 18–25, 139 (23%) participants were aged 26–34, 179 participants (30%) were aged 35–44, 163 participants (27%) were aged 45–54, and 77 participants (13%) were above the age of 55. When distilled by location, the greatest number of participants resided in East Trinidad (35%), while the least number of participants were resident in West Trinidad (9%) and Tobago (9%). Persons residing in South Trinidad accounted for 20%, North Trinidad 14%, Central Trinidad 13%, and Tobago 9% of the sample. A great majority of the participants identified their socio-economic status as middle class (430 participants or 71%), 17% (n = 101) identified themselves as belonging to the upper class, and 12% (n = 75) indicated that they belonged to the lower socio-economic class.
Table 1 Participants’ demographics
SAMPLE CHARACTERISTICS
N %
GENDER
Male
Female
Undisclosed
185 31
408 68
9 1
AGE
18–25
26–34
35–44
45–54
55+
43 7
139 23
179 30
163 27
77 13
GEOGRAPHIC LOCATION
East Trinidad
West Trinidad
North Trinidad
South Trinidad
Central Trinidad
Tobago
35
9
14
20
13
9
SOCIO-ECONOMIC STATUS
Upper Class
Middle Class
Lower Class
97 16
430 71
75 13
* - Indicates Missing data. Source Fieldwork, 2022.
Results
The quantitative component of the instrument was utilized to answer the four research questions, while the qualitative component elicited themes from the participant’s narratives on DV perpetration and victimization during the Covid-19 pandemic in Trinidad and Tobago.
Answering the Research Questions
In answering the first research question (Did individuals engage in increased domestic violence perpetration in Trinidad and Tobago during the Covid-19 pandemic?), the data indicated that 13% of the total number of participants reported an increase in their perpetration of DV during the pandemic. When disaggregated by gender, 17% of the male participants reported that they engaged in acts of DV towards their intimate partners during the pandemic, while 13% of the female participants indicated that they engaged in acts of DV towards their intimate partners during the pandemic.
As it relates to the second research question (Did individuals become victims of increased domestic violence in Trinidad and Tobago during the Covid-19 pandemic?), 16% of the participants indicated that they experienced an increase in DV victimization during the Covid-19 pandemic. When distilled by gender, the data revealed that 25% of the males were victims of increased DV victimization, while 12% of the females in the sample suffered increased DV victimization. Answers to RQ3 (Did fear of violence in intimate relationships increase in Trinidad and Tobago during the Covid-19 pandemic?) revealed that ten per cent of the participants (5% male and 5% female) indicated that they became increasingly fearful of the occurrence of DV within their intimate relationships during the Covid-19 pandemic in Trinidad and Tobago.
Research question number four sought to determine whether increases in DV perpetration/victimization/fear of victimization during the Covid-19 pandemic in Trinidad and Tobago was as a result of family isolation? Thirteen per cent of the male participants indicated that the increase in their perpetration of DV was the result of family isolation during the pandemic, while 16% of the female participants indicated that their increase in DV victimization was as a result of family isolation during the Covid-19 facilitated pandemic. On the other hand, 10% of the participants (5% male and 5% female) indicated that they became increasingly fearful of the occurrence of violence within their intimate relationships during the Covid-19 pandemic in Trinidad and Tobago. While stress, fear of engaging in erratic behaviors, job loss, and fear of the unknown were put forward as reasons for the heightened fear of DV victimization during the Covid-19 pandemic in Trinidad and Tobago, being isolated with a spouse and family was proffered as the main reason for fear of increased DV victimization among the surveyed population.
Themes
This section presents the qualitative results that emanated from the data collection instrument as thirty-three participants completed this section. Of the thirty-three individuals who completed the qualitative component of the instrument, twelve were perpetrators (solely), fifteen were victims (solely), four were perpetrators/victims, and two were neither perpetrators nor victims. Six themes emanated from the narratives of the participants who completed this component of the instrument. In descending order of prevalence, the themes are: (1) isolation/Covid-19 restrictions, (2) lack of assistance for victims, (3) male fear of reporting DV, (4) work as a safe space, (5) mental health effects, and (6) job loss. The themes are discussed in the following paragraphs, however, due to constraints of space, applicable participant quotes are used to illuminate the voice of the study’s participants (Wallace et al., 2021).
Theme 1 – Isolation/Covid-19 Restrictions
Thirteen individuals indicated that for them, isolation and Covid-19 restrictions impacted DV perpetration and victimization during the pandemic. For as a male victim, aged 35–44 years, residing in North Trinidad explained:During the pandemic there was a rise in domestic violence, as partners were not able to either get away from the abuse or the abuse was made easier as the restrictions on movement did not allow them to escape. This was my experience as a victim.
In a similar vein, a female victim aged 34–44, from Central Trinidad submitted:Whenever persons are isolated together for long periods without an outlet for personal space or time alone there will be a tendency for levels of frustration and irritability to become unbearable. I have had two childhood male friends express to me how heavy the weight of their families seems to them. One expressed frustration of their constant presence and thoughts of starting over with someone new; while another expressed it ‘like chains dragging me down’.
Theme 2 - Lack of Assistance for Victims
Eight of the study’s participants opined that lack of assistance for victims of DV was prevalent during the pandemic and this impacted DV perpetration and victimization. According to a female victim, aged 45–54 years, residing in East Trinidad:Domestic violence is real, and not much was done during the pandemic to provide housing support to victims and children as well as perpetrators in Trinidad and Tobago which is heartrending.
On the other hand, a female, aged 35–44 years, residing in North Trinidad stated:I work[ed] during the COVID assisting domestic violence victims and referring them to shelters and different agencies and there are so many times this couldn’t be done because of the lack of safe homes for victims of domestic violence.
Theme 3 - Male Fear of Reporting DV
The theme male fear of reporting emanated from the narratives of four participants. A male participant (perpetrator/victim), aged 45–54, residing in Tobago submitted the following:The time at home with my partner grew frustrating over time and things got a bit heated. At times, we both got physical towards each other, but I reminded myself that no one would believe me if I reported that she victimized me, especially if I reported it to the police.
Similarly, a female victim, aged 18–25, residing in South Trinidad stated:Domestic violence against men has not been given the attention it deserves as this is important. Toxic masculinity has caused men to shy away from taking further action, especially during the pandemic. This is also an issue that should be address through the DV support programmes.
A male, 35–44 (perpetrator/victim), residing in Tobago submitted the following:My female partner became increasingly physically and emotionally aggressive. But you know what, I didn’t report [anything] because society will laugh at me, both the police, my friends and women. Plus, when I look at academia for help, as they should be a bit more understanding, and I see Dr. X1 talking that is only women victims of domestic violence in the homes in TT [Trinidad and Tobago], I shake my head in disgust. Nobody sees the male side of the coin. It’s as though we cannot be victims of domestic violence. That’s why we do not report it [DV], pandemic or not.
Theme 4 - Work as a Safe Space
Three of the study’s participants used the term work as a safe space as they sought to explain DV perpetration and victimization during the Covid-19 enforced pandemic in Trinidad and Tobago. The following view was submitted by one participant (non-victim/non-perpetrator):Because of being a health centre worker, due to long hours and being short staffed which meant I had to do extras, I was not home to experience sexual abuse. What I experienced was emotional and physical abuse when I got home. Work was my safe place. We ended the relationship within two weeks of the pandemic. We were engaged to be married. (female, aged 35–44, residing in West Trinidad).
According to a female, aged 35–44, residing in South Trinidad:Covid-19 forced many couples to remain together for period longer than usual. Resultant from this was the unfortunate release of aggression on partners. The time that they would normally be apart was now spent together. Many women who because of their domestic situations relished the time apart for their safety, had to now live in fear of what could happen next.
One participant submitted the following:The Covid-19 pandemic increased my stress and negative feelings (in some instances) towards my partner. Being at work was a solace to me in the pre-Covid-19 era as I was very occupied with work, late meetings and other physical engagements, but the pandemic forced me to be at home with my partner and not having space to blow off steam increased my stress level and some domestic violence tendencies. (male, perpetrator/victim) aged 45–54, residing in East Trinidad).
Theme 5 - Mental Health Effects
Mental health effects emerged as another theme from the narratives of three participants, with the most succinct being proffered by a female participant who submitted:Just being at home all the time, I can’t exercise, I can’t go to the gym, I can’t be outside at certain hours just going for a drive, plus I had the children home, like forever with online school, and then constant talk from my husband, who we were already not seeing eye to eye. It was mentally taxing. I felt like I was going mad. (female, aged 35–44, South Trinidad, non-perpetrator/non-victim).
Theme 6 – Job Loss
Two of the study’s participants indicated that loss of jobs impacted DV perpetration and victimization during the pandemic and that this was their lived experience. This participant stated:I lost my job transporting children to school. Mentally, this was really tough on me to sit and look at my wife go to work and become the sole breadwinner. It drove me up a wall, negatively impacted my mental health and at times, I became very hostile and verbally abusive to my wife. (Male participant, perpetrator, aged 45–54 from East Trinidad).
Fig. 1 Schematic Representation of Themes by Prevalence.
Discussion
The finding of a 13% increase in DV perpetration among those who responded to the survey was understandable, yet surprising to the research team. While the researchers expected an increase in DV perpetration as persons were ‘locked away’ together during the pandemic, the surprise was premised on the notion espoused in the literature of massive increases in DV, for example, by 50%, Colombia, by 79%, and Tunisia by 400% (Mlambo-Ngcuka, 2020). Further, while the literature as well as the opportunity theory which undergirds this study spoke to potential increases, the majority of research on increases in DV perpetration spoke to ‘drastic’, ‘massive’ and ‘alarming’ increases (Ali & Khalid, 2021; Hansen & Lory, 2020; Mlambo-Ngcuka, 2020), however, this was not the case with Trinidad and Tobago as the increase among the individuals who completed the instrument was 13%. Though the increased in DV perpetration was not as striking as those in similar studies, such as, Ali and Khalid (2021), Hansen & Lory (2020) and Mlambo-Ngcuka (2020), the increase is aligned with the narrative of the opportunity theory which submit that pandemics creates the opportunity for interpersonal violence between intimate partners to increase.
The data also indicated that 17% of the male participants who responded to the survey conducted increased acts of DV towards their intimate partner or spouse during the pandemic. This increased DV perpetration by males is consistent with previous research which pointed out males are generally viewed as perpetrators of DV and females as victims (Dutton & White, 2013; Verma & Raina, 2014). On the other hand, 13% of the female participants who completed the instrument indicated that they conducted acts of DV towards their intimate partners. This result is consistent with research indicating that women are increasingly becoming perpetrators of DV and that female perpetrators and male victims are not unusual (See Nesset et al., 2021 for support).
Overall, 16% of the sample in this study indicated that they experienced an increase in DV victimization within their intimate relationships during the Covid-19 facilitated pandemic in Trinidad and Tobago. In a similar vein to the increase in perpetration, this finding of increased DV victimization is slightly similar, yet a bit dissimilar from global research. The similarity is based on increased DV victimization being a commonality in the majority of studies conducted on DV during the pandemic (Ali & Khalid, 2021; Hansen & Lory, 2020; van Gelder et al., 2020). Concomitantly, the dissimilarity is premised on the notion that the increased DV victimization in the C-19 DVAP study was not as drastic or alarming as reported in other studies (Ali & Khalid, 2021; Bellizzi et al., 2020; Pfitzner et al., 2020).
When disaggregated by gender, the data revealed that 25% of the males in this study were victims of increased DV victimization during the pandemic, while 12% of the females in the sample suffered from increased DV victimization. This finding is both interesting and surprising given that the academic literature indicate that women are generally victims of DV (See for example, European Institute for Gender Equality 2012). As it relates to the current study, this finding can be explained from the perspective of male fear of reporting their DV victimization to state agencies in Trinidad and Tobago (Wallace et al., 2019). The fact that the data collection was conducted online might have accounted for this unexpected high rate of males reporting DV victimization as the males might have felt comfortable in sharing their experiences of DV victimization in an anonymous forum with little possibility of being judged by the person receiving the data.
The finding of increased male victimization within the sample is inconsistent with previous research on DV which reported males as predominantly perpetrators of DV, IPV, and abuse in intimate and familial settings and females as victims (Dutton & White, 2013; Verma & Raina, 2014). The finding of this study indicating more male than female victims of increased DV victimization during the pandemic is supported by the work of Nesset et al. (2021) who found that while there was an increase in intimate partner violence (IPV) in Norway during the Covid-19 lockdown, there were significantly more male victims. Notably, the finding of increased interpersonal violence suffered by male participants in this study is aligned to the opportunity theory.
10% of the study’s participants (5% male and 5% female) indicated that they became increasingly fearful that violence would occur within their intimate relationships during the Covid-19 pandemic. While fear of victimization has been studied widely, research and scholarship on the association between fear of victimization in the context of DV is a neglected one. However, it is argued that there is an association between the fear of DV victimization and intimate relationships especially in times of social inequality (Iglesias, Cardoso, & Sousa, 2019), such as pandemics. With the pronouncement of Iglesias, Cardoso, and Sousa (2019) in mind, it is hardly surprising that some male and female participants in this study indicated that they became fearful of violent victimization (DV) occurring in their relationships during the Covid-19 pandemic.
The survey findings showed minor increases in DV victimization and this may suggest something about DV and perpetration and victimization in Trinidad and Tobago overall. However, the survey findings elicited from adults who responded to the survey in Trinidad and Tobago is incongruent with other research findings measuring the same phenomenon (increases in DV perpetration and victimization) (Bettinger-Lopez & Bro, 2020; Davidge, 2020; Hansen & Lory, 2020; Mlambo-Ngcuka, 2020; Pfitzner et al., 2020). Though surprising, the results of this study lend support to the notion that researchers must avoid ‘universalizing’ problems, despite this being a tempting phenomenon in the era of global uncertainty. The increase in DV perpetration and victimization found in the sample of persons who completed the survey in Trinidad and Tobago may be the result of differences in culture whereby in different cultures, people react differently and develop different coping strategies and resilience during times of crises. Instructively, the results of the study in Trinidad and Tobago is largely inconsistent with other research which spoke to massive increases in DV and is consistent with the pronouncement of Amahazion (2021, p. 1) who points out that “While disasters can harm all, they do not impact or affect all people equally or in the same ways.”
The findings of this study which reported minor increases in DV victimization, perpetration and fear of victimization among the sampled population of individuals in intimate relationships in Trinidad and Tobago is also not singular. For example, some researchers have reported small increases in DV perpetration (Leslie & Wilson, 2020), conflicting effects (Miller et al., 2020) and even reductions in DV generally (Bettinger-Lopez & Bro, 2020; Mlambo-Ngcuka, 2020; Øverlien 2020; Silverio-Murillo, Balmori de la Miyar, & Hoehn-Velasco, 2020). This reduction was also reported by some UK-based services (Davidge, 2020; SafeLives, 2020). However, the results of the current research effort should be interpreted with caution, as without a population-based survey, it is impossible to determine a country’s incidence of DV overall and this has been a key global challenge for most countries during the Covid-19 pandemic.
The findings of the qualitative component of the study indicated that there was no equal spread of contributory factors for increased DV perpetration and victimization for individuals in intimate relationships in Trinidad and Tobago. For example, the majority of the participants in the qualitative component of the study indicated that for them, isolation/Covid-19 restriction was the major contributing factor for increased DV victimization and perpetration. This position is hardly surprising as the literature espouse similar sentiments (See Bilyeau, 2020; Dekel & Abrahams, 2021 for support).
In the context of contributory factors for enhanced DV victimization and perpetration during the pandemic, isolation/Covid-19 restrictions were major contributing factors. Male fear of reporting DV, work as a safe space, mental health effects and job loss all emanated from the participant’s narratives as contributory factors to increased DV victimization and perpetration. The aforementioned contributory factors mirror the work of Capaldi et al. (2012) who found that financial stress has been linked to perpetration of partner abuse and Sharma & Borah (2020) who pointed out that layoffs and loss of income are driving up the incidence of DV during the Covid-19 pandemic. Similarly, there is support for the aforementioned findings from Wallace et al. (2019) on male fear of reporting DV victimization, Campbell (2020) on work as a safe space, and Dekel and Abrahams (2021) on mental health and DV as factors that contribute to elevated levels of DV victimization.
Study’s Limitations
This study is subject to a number of limitations, including the short time frame that the researchers allocated for data collection (two weeks) and the method of collecting the data. The data collection component of the study focused on gathering data within a 14-day period and this posed a challenge as several individuals contacted the researchers indicating an interest in participating in the study after the data collection phase was ended. Another limitation of this study was that the data were collected online and it is important to note that victims and potential victims of DV are often technologically-isolated. Further, the online data collection approach might pose a threat to the external validity of this study as the sample for this study was limited to individuals with computers and mobile phones as well as an understanding of the online survey environment (See Khubchandani et al., 2021 for support). Importantly, the researchers sought to offset the technologically-isolated nature of DV victims as participants were instructed to select a safe period when their spouse was absent to complete the instrument. The non-random dissemination of the instrument and the challenges inherent to self-reporting of DV are other limitations of this study.
The limitations mentioned above may preclude generalization of the study’s findings to the larger population in Trinidad and Tobago. However, despite the limitations, this study appears to be the first in the Caribbean to conduct an evaluation of DV perpetration, victimization and fear of victimization during the Covid-19 lockdown and the study is therefore of much utility to policy makers in the Caribbean as it can be used as a starting point for further research as well as to implement policies to deal with similar occurrences in the future.
Conclusion
The attention to potential increases in DV perpetration and victimization is quite natural given the significant economic and social costs associated with the phenomenon (Fearon & Hoeffler, 2014). This attention is also as a result of the potential for increased DV as a result of the Covid-19 pandemic as espoused by the opportunity theory. This article focused on factors associated with increases in DV perpetration and victimization among persons in intimate relationships in Trinidad and Tobago. The authors of this paper presents and contextualizes findings from the first nationwide survey that measured DV perpetration and victimization in Trinidad and Tobago during the Covid-19 pandemic.
Though hampered by limitations, the study’s findings indicated an overall increase of 13% in DV perpetration and a 16% increase in victimization, with a slightly higher percentage of male perpetrators (17%) when compared to female perpetrators (13%) among persons who completed the instrument. The data also revealed that during the Covid-19 pandemic, males were almost 50% more likely to become victims of DV than females. Quite notably, 5% of both males and females reported increased fear of violence in their relationships during the pandemic. Six contributory factors for increased DV perpetration and victimization were elicited from the narratives of the participants (isolation/Covid-19 restrictions, lack of assistance for victims, job loss, male fear of reporting DV, work as a safe space, and mental health effects).
To conclude, while comprehensive Caribbean data are not yet available on increases in DV perpetration and victimization in intimate settings during the Covid-19 pandemic, this paper contributes to the emerging body of global and regional scholarship on the impact of the Covid-19 pandemic on DV by reporting specifically on the DV perpetration and victimization experiences in Trinidad and Tobago. Importantly, the authors view this article as a call to action for Caribbean policymakers to broaden their knowledge base as understandings of DV perpetration and victimization during the Covid-19 pandemic will be useful in managing similar events in the future as the qualitative data indicated a lack of shelters for victims of DV.
Acknowledgements
The authors acknowledge the Association of Caribbean Criminal Justice Practitioners for conceptualizing, supporting and facilitating this study.
Declarations
Conflict of Interest
The authors declare that they have no conflict of interest.
1 The correct name of the individual is withheld and Dr. X used as a pseudonym.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Ali R Khalid A COVID-19 and domestic violence in Pakistan: an analysis of the media perspective Journal of International Women’s Studies 2021 22 12 98 114
Amahazion FF Introduction to Journal of International Women’s Studies Special Issue COVID-19: the impact on women and girls around the World Journal of International Women’s Studies 2021 22 12 1 6
American College of Obstetricians and Gynecologists (ACOG) Intimate Partner violence American College of Obstetricians and Gynecologists Committee Opinion 2012 518 1 6
Bair-Merritt MH Intimate Partner violence Pediatrics in Review 2010 31 4 145 150 10.1542/pir.31-4-145 20360408
Bell J Doing your Research Project 2005 Berkshire Open University Press/McGraw-Hill Education
Bellizzi S Nivoli A Lorettu L Farina G Ramses M Ronzoni AR Violence against women in Italy during the COVID-19 pandemic International Journal of Gynecology and Obstetrics 2020 150 2 258 259 10.1002/ijgo.13270 32533860
Bettinger-Lopez C Bro A A double pandemic: domestic violence in the age of COVID-19 Domestic Violence Report 2020 25 5 85 87
Bilyeau, N., & Says UN Group. (2020). COVID-19 Stokes ‘Shadow Pandemic’ of Violence Against Women,. The Crime Report. https://thecrimereport.org/2020/04/07/covid-19-stokes-shadow-pandemic-of-violence-against-women-says-un-group/
Boserup B McKenney M Elkbuli A Alarming trends in US domestic violence during the COVID-19 pandemic American Journal of Emergency Medicine 2020 38 12 2753 2755 10.1016/j.ajem.2020.04.077 32402499
Bradbury-Jones C Isham L The Pandemic Paradox: the Consequences of COVID-19 on domestic violence Journal of Clinical Nursing 2020 29 13–14 2047 2049 10.1111/jocn.15296 32281158
Braun V Clarke V Using thematic analysis in psychology Qualitative Research in Psychology 2006 3 2 77 101 10.1191/1478088706qp063oa
Buttell F Ferreira RJ The hidden disaster of COVID-19: intimate Partner violence Psychological Trauma: Theory Research Practice and Policy 2020 12 1 197 198 10.1037/tra0000646
Campbell AM An increasing risk of family violence during the Covid-19 pandemic: strengthening community collaborations to save lives Forensic Science International Reports 2020 2 100089 10.1016/j.fsir.2020.100089
Capaldi DM Knoble NB Shortt JW Kim HK A systematic review of risk factors for intimate partner violence Partner Abuse 2012 3 2 231 280 10.1891/1946-6560.3.2.231 22754606
Coohey C The relationship between mothers’ social networks and severe domestic violence: a test of the social isolation hypothesis Violence and Victims 2007 22 4 503 512 10.1891/088667007781554008 17691556
Davidge, S. (2020). A perfect storm: the impact of the COVID-19 pandemic on domestic abuse survivors and the services supporting them. Women’s Aid.
Davies, G., Guenfoud, I., & Jovanovic, D. (March 20 2021). ‘Shadow Pandemic’: Domestic abuse reports soar during Europe’s coronavirus crisis. https://abcnews.go.com/International/shadow-pandemic-domestic-abuse-reports-soar-europes-coronavirus/story?id=76325386
Dekel B Abrahams N ‘I will rather be killed by corona than by him… Experiences of abused women seeking shelter during South Africa’s COVID-19 lockdown Plos One 2021 16 10 e0259275 10.1371/journal.pone.0259275 34710174
Dlamini NJ Gender-based violence, Twin Pandemic to COVID-19 Critical Sociology 2020 47 4–5 583 590 10.1177/0896920520975465
Domestic Violence Act (2020). Ministry of Legal Affairs, Port-of-Spain, Trinidad and Tobago.
Dutton DG White KR Male victims of domestic violence New Male Studies: An International Journal 2013 2 1 5 17
European Institute for Gender Equality. (2012). Review of the implementation of the Beijing platform for action in the EU member States: violence against women – victim support. European Union.
Evans ML Lindauer M Farrell ME A pandemic within a pandemic – intimate Partner Violence during Covid-19 New England Journal of Medicine 2020 383 24 2302 2304 10.1056/NEJMp2024046 32937063
Fearon, J., & Hoeffler, A. (2014). Benefits and costs of the conflict and violence targets for the Post-2015 Development Agenda (pp. 1–65). Conflict and Violence Assessment Paper, Copenhagen Consensus Center.
Gama, A., Pedro, A. R., de Carvalho, A. E., Duarte, V., Quintas, J., Matias, A., Keygnaert, I., & Dias, S. (2021). Domestic Violence During The COVID-19 Pandemic in Portugal”. Portuguese Journal of Public Health, 38(1), 32–40. doi:10.1159/000514341.
Garson, D. (2013). Crosstabulation. Statistical Associates Publishers.
Greenfield, T. (2002). Research methods for postgraduates. Arnold.
Hansen, J. A., & Lory, G. L. (2020). Rural Victimization and Policing during the COVID-19 Pandemic. American journal of criminal justice: AJCJ, 1–12. Advance online publication. 10.1007/s12103-020-09554-0
Iglesias C Cardoso C Sousa P Macrosystem and Women’s fear of crime in intimate relations: an Integrated Model International Annals of Criminology 2019 57 1–2 172 197 10.1017/cri.2020.10
Joshi M Sorenson SB Intimate Partner violence at the scene: incident characteristics and implications for Public Health Surveillance Evaluation Review 2010 34 2 116 136 10.1177/0193841X09360323 20233999
Khubchandani J Sharma S Webb FJ Wiblishauser MJ Bowman SL Post lockdown depression and anxiety in the USA during the COVID-19 pandemic Journal of Public Health 2021 43 2 246 253 10.1093/pubmed/fdaa250 33426559
Kofman BY Garfin R Home is not always a haven: the domestic violence Crisis amid the COVID-19 pandemic Psychological Trauma: Theory Research Practice and Policy 2020 12 S1 S199 S201 10.1037/tra0000866 32478558
Kourti A Stavridou A Panagouli E Psaltopoulou T Spiliopoulou C Tsolia M Sergentanis TN Tsitsika A Domestic violence during the COVID-19 pandemic: a systematic review Trauma Violence Abuse 2021 10.1177/15248380211038690
Liamputtong, P. (2020). Qualitative research methods. 5th edition. Oxford University Press.
Leslie E Wilson R Sheltering in place and domestic violence: evidence from calls for service during COVID- 19 Journal of Public Economics 2020 189 104241 10.1016/j.jpubeco.2020.104241 32834179
Lupton, D. (2021). Doing Fieldwork in a Pandemic. https://www.researchgate.net/publication/352373476_Doing_Fieldwork_in_a_Pandemic
McCrary, J., & Sanga, S. (2020). The Impact of the Coronavirus Lockdown on Domestic Violence.American Law and Economics Review, 1–27.
Menjívar C Salcido O Immigrant women and domestic violence: common Experiences in different countries Gender & Society 2002 16 6 898 920 10.1177/089124302237894
Milan S Treré E Big Data from the South(s): Beyond Data Universalism Television & New Media 2019 20 4 319 335 10.1177/1527476419837739
Miller, A. R., Segal, C., & Spencer, M. K. (2020). Effects of the COVID-19 Pandemic on Domestic Violence in Los Angeles. NBER Working Paper No. 28068, 1–33.
Mlambo-Ngcuka, P. (2020). Violence against women and girls: The shadow pandemic. UN Women, 6 April. https://www.unwomen.org/en/news/stories/2020/4/statement-ed-phumzile-violence-against-womenduring-pandemic
Mumford, E., & Rothman, E. (2020). Cyber-Abuse Research Initiative (CARI). https://www.nlm.nih.gov/dr2/Impact_of_COVID-19_social_distancing_on_domestic_abuse-national_study.pdf
Norušis, M. J. (1993). The SPSS Guide to Data Analysis. SPSS Incorporated.
Nesset MB Gudde CB Mentzoni GE Palmstierna T Intimate partner violence during COVID-19 lockdown in Norway: the increase of police reports Bmc Public Health 2021 21 1 2292 10.1186/s12889-021-12408-x 34915874
New York Times (April 6, 2020). A New Covid-19 Crisis: Domestic Abuse Rises Worldwide. https://www.nytimes.com/2020/04/06/world/coronavirus-domestic-violence.html
Øverlien C The COVID-19 pandemic and its impact on children in domestic violence refuges Child Abuse Review 2020 29 4 379 386 10.1002/car.2650 32904999
Pfitzner, N., Fitz-Gibbon, K., & True, J. (2020). Responding to the ‘shadow pandemic’: practitioner views on the nature of and responses to violence against women in Victoria, Australia during the COVID-19 restrictions. Monash Gender and Family Violence Prevention Centre, Monash University.
Pina, A. (2021). Research to investigate prevalence of Technology Facilitated Domestic Abuse. iCSS Newsletter, Spring, 2021, 1–8. University of Kent. https://cyber.kent.ac.uk/newsletters/iCSS-NL-Spring2021.pdf
Piquero AR Riddell JR Bishopp SA Narvey C Reid JA Piquero NL Staying home, staying safe? A short-term analysis of COVID-19 on Dallas domestic violence American Journal of Criminal Justice 2020 14 1 35 10.1007/s12103-020-09531-7
Plasilova L Hula M Krejcova L Klapilova K The COVID-19 pandemic and intimate Partner violence against women in the Czech Republic: incidence and Associated factors International Journal of Environmental Research and Public Health 2021 18 19 1 14 10.3390/ijerph181910502
Punch, K. F. (2014). Introduction to social research: Qualitative and quantitative approaches. 3rd edition. SAGE.
SafeLives (2020). Domestic abuse frontline service COVID-19 survey results. https://safelives.org.uk/practitionersurvey-results-covid-19
Sharma, A., & Borah, S. B. (2020). Covid-19 and domestic violence: an indirect path to Social and Economic Crisis. Journal of Family Violence, 1–7. 10.1007/s10896-020-00188-8.
Silverman, D. (2004). Qualitative research: theory, Method and Practice (2nd ed.). Sage Publication.
Silverio-Murillo, A., de la Balmori, J. R., & Hoehn-Velasco, L. (2020). Families under Confinement: COVID-19, Domestic Violence, and Alcohol Consumption. SSRN Scholar Paper ID 3688384. Social Science Research Network. 10.2139/ssrn.3688384.
Stoianova T Ostrovska L Tripulskyir G Covid-19: pandemic of domestic violence Revista de Derecho 2020 9 2 111 136 10.31207/ih.v9i2.246
Thomas, J. (April 16, 2020). Domestic violence calls increase locally, nationwide during COVID- 19 pandemic. https://www.kalb.com/content/news/Domestic-violence-calls-increase-locally-nationwide-during-COVID-19-pandemic%2D%2D569689371.html
Tourangeau R Yan T Sensitive questions in surveys Psychological Bulletin 2007 133 5 859 883 10.1037/0033-2909.133.5.859 17723033
Tracy, S. J. (2019). Qualitative research methods: collecting evidence, crafting analysis, communicating impact. Wiley-Blackwell.
United Nations (2020). Addressing the impact of the COVID-19 pandemic on violence against women and girls. https://www.un.org/en/addressing-impact-covid-19-pandemic-violence-against-women-and-girls
United Nations Office on Drugs and Crime. (UNODC). (2020). Gender and pandemic: urgent call for action. Advocacy brief 4. Government of Punjab.
UN Women (2020). COVID-19 and ending violence against women and girls. Gender-Based Violence. 1. https://digitalcommons.wcl.american.edu/wlpviolence/1
van Gelder, N., Peterman, A., Potts, A., O’Donnell, M., Thompson, K., Shah, N., & Oertelt- Prigione, S. (2020). Gender and COVID-19 working group. COVID-19: reducing the risk of infection might increase the risk of intimate partner violence. EClinicalMedicine, 21. 10.1016/j.eclinm.2020.100348.
Vazquez SP Stohr MK Purkiss M Intimate Partner Violence incidence and characteristics: Idaho NIBRS 1995 to 2001 data Criminal Justice Policy Review 2005 16 1 99 114 10.1177/0887403404267771
Verma A Raina A Transcending the stereotypes of gender: a portrayal of emotional violence experienced by males in select novels International Journal of English: Literature Language & Skills 2014 3 2 20 26
Wallace WC Gibson C Gordon N Lakhan R Mahabir J Seetahal C Domestic violence: intimate Partner Violence victimization non-reporting to the police in Trinidad and Tobago Justice Policy Journal 2019 16 1 1 30
Wallace WC Criminal Justice Systems (GOVT 2011) lecture. The University of the West Indies 2021 St. Augustine Trinidad and Tobago
Wallace WC Harry A Ramdass R Salina S Transitioning to Online Teaching, Learning, and Assessment in the COVID-19 era: understanding Student and Faculty Perspectives The UWI Quality Education Forum 2021 25 51 77
WHO (2020). WHO Director-General’s Opening Remarks at the Media Briefing on COVID-19. 11 March 2020. World Health Organization. https://www.who.int/director-general/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19---11-march-2020
| 0 | PMC9742019 | NO-CC CODE | 2022-12-14 23:29:58 | no | J Fam Violence. 2022 Dec 12;:1-12 | utf-8 | J Fam Violence | 2,022 | 10.1007/s10896-022-00485-4 | oa_other |
==== Front
Adv Neurodev Disord
Adv Neurodev Disord
Advances in Neurodevelopmental Disorders
2366-7532
2366-7540
Springer International Publishing Cham
310
10.1007/s41252-022-00310-5
Original Paper
Facilitators and Barriers to Physical Activity Involvement as Described by Autistic Youth with Mild Intellectual Disability
http://orcid.org/0000-0002-0648-1752
Boucher Troy Q. [email protected]
McIntyre Cassia L.
Iarocci Grace
grid.61971.38 0000 0004 1936 7494 Department of Psychology, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia V5A 1S6 Canada
12 12 2022
113
28 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.
Objectives
Physical activity involvement among autistic youth and youth with an intellectual disability is significantly lower than the general population. Few studies have included youth with comorbid diagnoses of ASD and intellectual disability. Fewer studies collect information from the youth themselves. This study examined barriers and facilitators to physical activity in autistic youth with mild intellectual disability using semi-structured interviews with youth and through caregiver reports.
Methods
Fourteen caregivers and their children ages 8 to 16 years old participated. Caregivers completed a questionnaire about their thoughts on their child’s physical activity while their children completed the semi-structured interview. A descriptive phenomenological approach was followed.
Results
Four themes were inductively identified: intrapersonal barriers (factors that are within the person which impede physical activity involvement, such as exhibiting challenging behaviors that inhibit engagement), interpersonal barriers (factors external to the person, such as lack of community support), intrapersonal facilitators (factors within the person that enhance physical activity involvement, such as being intrinsically motivated to improve skills), and interpersonal facilitators (external factors such as supports from caregivers).
Conclusions
Interviewing youth is important to capture a holistic picture of factors influencing physical activity. Future research may focus on implementing and assessing the efficacy of strategies to address the barriers facing youth diagnosed with ASD and intellectual disability.
Supplementary Information
The online version contains supplementary material available at 10.1007/s41252-022-00310-5.
Keywords
Intellectual disability
Autism spectrum disorder
Physical activity
Semi-structured interview
Phenomenology
http://dx.doi.org/10.13039/100001346 Special Olympics Youth Innovation Grant Youth Engagement Grant Boucher Troy Q.
==== Body
pmcYouths with intellectual disabilities have below-average cognitive functioning as well as poor adaptive behavior that impacts their communication, social, and life skills (American Psychiatric Association, 2022). They often face challenges maintaining a healthy lifestyle and good quality of life. For example, youth with an intellectual disability engages in significantly lower levels of physical activity compared to youth in the general population (Case et al., 2020; Stanish et al., 2016). This is concerning given the high prevalence of health challenges in people with intellectual disabilities, such as being overweight and obese, and the health consequences of high blood pressure, high cholesterol, insulin resistance and type 2 diabetes, sleep apnea, asthma, joint and musculoskeletal problems, gastrointestinal problems, and social and psychological difficulties (Cooper et al., 2015). Youth with intellectual disability may also miss out on the many known benefits of physical activity, such as improved mental health and increased self-efficacy (Ginis et al., 2021; Kapsal et al., 2019). Understanding why physical activity rates are comparatively lower in this population is critical in order to improve their physical health and well-being.
Youth diagnosed with autism spectrum disorder (ASD) encounters similar health challenges to those diagnosed with intellectual disability. Compared to non-autistic children, autistic children are more likely to be overweight and obese and engage in less physical activity (McCoy & Morgan, 2020; Rech et al., 2022). Autistic children are also less motivated to engage in physical activities than their non-autistic peers (Scharoun et al., 2017). Difficulty understanding the rules associated with the activity, difficulties with social interaction and lack of interest, motor skill deficits, and fear of injury or feeling unsafe are possible barriers to participation in physical activity by autistic youth (Arnell et al., 2020; Healy et al., 2013).
Conversely, there are several facilitators to physical activity. In a study of Special Olympics athletes with intellectual disabilities aged 11 to 22 years old, Weiss et al. (2020) identified that athlete retention in sport programs was related to the degree of parental support such as providing transportation and encouragement to their child. The frequency of participation in sport also predicted retention, which may be a factor influencing the positive relationships formed by athletes, a sense of belongingness, and positive experiences in the program by athletes. Other factors identified in the literature that influence or encourage physical activity in autistic people or people diagnosed with intellectual disabilities include encouragement by friends and parents (Brown et al., 2020), freedom to choose their physical activities (Arnell et al., 2018; Bossink et al., 2017), and being rewarded or given praise (Obrusnikova & Miccinello, 2012).
Few studies have focused on the perceived facilitators and barriers of physical activity specifically for children diagnosed with both ASD and intellectual disability. Tint et al. (2017) examined participation in community activities by autistic people with an intellectual disabilities ranging from 11 to 22 years old. Autistic participants with intellectual disabilities engaged in significantly fewer community events than youth with intellectual disability only, such as community events, organizations, clubs, and neighborhood outings. Caregivers of autistic participants with intellectual disabilities reported their environment as less supportive for their child relative to caregivers of participants with intellectual disabilities only. Caregivers also perceived the social demands of activities and relationships with peers to be a barrier for their autistic child with intellectual disability, more so than for participants with intellectual disability only. The authors highlight the need to examine the specific challenges of youth diagnosed with both ASD and intellectual disability given the differences in perceived barriers and rates of participation compared to those with intellectual disability only.
More research is needed to identify the specific barriers and facilitators to physical activity involvement for autistic youth with intellectual disability that specifically address the perceptions and experiences of the youth themselves (Chown et al., 2017). Autistic children and children with intellectual disabilities are under-included as direct research participants, with the majority of studies focusing on informant-report questionnaires to characterize and examine their behaviors (Bölte, 2014). However, the subjective experiences of autistic individuals or those diagnosed with intellectual disability are lost with the heavy emphasis on informant-report data, as caregivers do not necessarily have insight into their child’s thought processes, unspoken opinions, or their behaviors engaged in isolation. The aim of the present study was to explore the barriers and facilitators to physical activity as reported directly by the youth themselves, supplemented by parent-report questionnaire data.
Method
Participants
Fourteen caregivers and their children ages 8 to 16 years old who were diagnosed with both intellectual disability and ASD participated in this study. Participants were recruited through a day camp for youth diagnosed with intellectual disability which was advertised through Facebook, local ASD service provider websites, the research team’s laboratory website, and through word of mouth. The camp was held during a weekend in February 2020 prior to the start of social distancing measures in Canada as a result of the COVID-19 pandemic. Inclusion criteria for the study required that the child could speak in English and had a diagnosis of both ASD and intellectual disability. Two youths participated in the camp but were excluded from the study as they did not have a diagnosis of intellectual disability. Thus, fourteen youths (mage = 12.23, SDage = 2.51, 13 males) and their caregivers were included in the analysis. Participant demographic information can be found in Table 1.Table 1 Demographic information
Variable Mean (SD)
Age 12.23 (2.51)
WASI-II FSIQ-4 65.64 (7.78)
Ethnicity
Caucasian 10
Asian 2
Indigenous 2
School type
Public school 10
Private school 2
Specialty school 2
Annual family income
$20,000 to $49,999 1
$50,000 to $79,999 4
$80,000 to $109,000 6
Greater than $110,000 3
Abbreviations: WASI-II FSIQ-4 Wechsler Abbreviated Scale of Intelligence, 2nd Edition, Full Scale IQ-4
N = 14
Diagnosis of ASD was confirmed through the presentation of the child’s diagnostic report or a government funding eligibility report. A diagnosis of ASD in the Canadian province of British Columbia involves a standardized diagnostic procedure using the Autism Diagnostic Observation Schedule (ADOS) and the Autism Diagnostic Interview-Revised (ADI-R) conducted by a trained clinician. The presence of intellectual disability was confirmed through the caregivers’ confirmation of their child’s diagnostic report, which was based on the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (APA, 1994) or Fifth Edition (APA, 2013). All youth were diagnosed with intellectual disability, mild severity with an IQ ranging between 53 and 70 on their diagnostic report.
Procedures
Four interviewers, including the second author, were trained by the first author to administer the interview protocol. Interviewers had previous experience working with youth with developmental disabilities in clinical, academic, or research settings and had prior experience conducting semi-structured interviews with non-clinical adult populations. Interviewers received 2 h of training from the first author to refine their skills in conducting semi-structured interviews using the interview protocol, which included instruction on how to re-word questions when the participant did not understand, when and how to use the specified prompts, and when and how to prompt participants using the Picture Exchange Communication System (PECS; Bondy & Frost, 1994). Interviewers were also instructed on how to adhere to ethical principles of qualitative data collection with vulnerable populations. As part of this training, interviewers completed interactive exercises provided by Bhattacharya (2017) to develop an understanding of their positionality as non-disabled researchers and to role-play administering the semi-structured interview. Following their first and second interviews, interviewers reviewed their recordings with the first author to ensure that the interview protocol was followed and to evaluate the prompts and probes used by the interviewer. The interviewers were then provided feedback to incorporate on their next interviews if needed. Across all interviews, interviewers were found to have asked each question on the protocol and were identified by the first author to have correctly prompted the participant as needed and correctly followed procedures for re-wording questions and using the PECS cards.
Interviews were conducted in a private room at a table, with the child sitting either across from or to the side of the interviewer. When assent was granted, the interviewer began recording using a Philips Voice Tracer DVT1150 audio recorder placed between the child and the interviewer. For all youth, the interview ended only after all questions were asked and an answer was provided. Interviews lasted between 6 and 21 min. When the interview was completed, the youth then completed the WASI-II before participating in various non-research activities. Caregivers completed a demographic questionnaire, the Health & Wellness and Risk and Protection Survey (HWREPS; McFee, 2019), and the additional physical activity preference questions while their child participated in the day camp. Youth were compensated with a t-shirt and were provided with food, and caregivers were reimbursed for parking or public transit costs.
Measures
Intelligence Quotient
The Wechsler Abbreviated Scale of Intelligence, Second Edition (WASI-II; Wechsler, 2011) was administered with all youth to assess their intellectual ability. The WASI-II provides an accurate estimate of cognitive ability comparable to other measures of intelligence, developed for quick administration with individuals ages 6 to 90 years old. The Full-Scale IQ-4 (FSIQ-4) is derived from administering all four subtests: block design, vocabulary, matrix reasoning, and similarities. The WASI-II was administered by trained graduate students in the Clinical Psychology program at Simon Fraser University.
Facilitators and Barriers to Physical Activity as Perceived by Caregivers
The Health & Wellness and Risk Engagement & Protection Survey (McFee, 2019) was administered to all caregivers. The HWREPS is a 35-item measure designed to assess caregiver reports of their child’s physical and recreational activity, non-therapeutic electronic use, sleep quality and schedule, diet, barriers to physical activity, and injury history. Additionally, the HWREPS assesses caregivers’ self-reported beliefs about protecting their child from injuries, the protective measures they take when their child participates in physical activity, and their overall thoughts about their child’s physical activity and risk of injury. Six items from the HWREPS that ask caregivers to report on their child’s activity preferences were included in the present analyses; these questions utilize a 7-point Likert scale, ranging from (1) very strongly disagree to (7) very strongly agree. The HWREPS has not yet been validated for use with individuals diagnosed with both ASD and intellectual disability but has been previously used in research with autistic youth between the ages of 6 and 12 years old (McFee, 2019).
An additional 14 questions were drafted by the research team and community consultants following a literature review on barriers and facilitators to physical activity in ASD and intellectual disability (Bossink et al., 2017; Healy et al., 2013; Obrusnikova & Cavalier, 2011; Obrusnikova & Miccinello, 2012; Stanish et al., 2015, 2016; van Schijndel-Speet et al., 2014). The team of community consultants was comprised of two care providers for people with intellectual disabilities, two coaches of inclusive sport programs for people with developmental disabilities, and two adults with intellectual disabilities. In total, 14 themes of the barriers and facilitators to physical activity were identified from the literature that guided the creation of the 14 questions: transportation, availability of programs, awareness of available programs, staff training, cost, age and developmental appropriateness of activities, and various characteristics of individuals that influence physical activity involvement (e.g., physical health problems, behavioral difficulties, social ability, need for routine). The consultants provided feedback on the wording and any potential challenges with interpretation of the questions. The questions were revised according to the consultants’ feedback, sent back to the consultants, then finalized. The questions ask caregivers about their thoughts on factors that may influence their child’s participation in physical activities. Caregivers endorsed the extent to which they agree with various statements on a 7-point Likert scale, ranging from (1) very strongly disagree to (7) very strongly agree.
Facilitators and Barriers to Physical Activity as Perceived by Children with ASD and Intellectual Disability
A semi-structured interview protocol was adapted from van Schijndel-Speet et al. (2014) for use with youth in the present study (see Supplementary Material 1). The interview protocol was evaluated by all members of the research and consultation team. This evaluation process helped identify how questions may be perceived, processed, and responded to by the youth participants so that questions were worded clearly and succinctly and with appropriate prompts.
Youth were read the assent script by an interviewer which asked if they would be willing to talk about the types of sports and activities that they do. After assent was granted, the interview began with rapport-building questions asking about the child’s name, age, and grade, what they had for breakfast, and their favorite and least favorite foods. The interviewer then asked the 15 questions regarding physical activities that the child engages in, the types of activities that they do and do not like, who they do these activities with, and the things that prevent or stop them from doing physical activities. Each question had a series of prompts that researchers asked if the child did not provide a response. For example, the question “How do you get to school?” would be prompted with up to three additional questions: “Do you walk to school?,” “Do you drive in a car to school?.” and “Do you take a bus to school?”.
The child was prompted using a series of cards from PECS on select questions if they had no response or had difficulty articulating their answer. If the child was not verbally answering questions, the interviewer directed the child to the PECS cards corresponding to specific questions. For questions that had an associated PECS deck, the interviewer presented the picture cards to the child and asked the question again. If the child indicated an answer, the interviewer said the name of the card aloud, so the answer was recorded on the audio recording device. If no answer was provided, the interviewer re-asked the question and then answered it themselves in order to provide the participant with an example. For example, “What is your favorite food? My favorite food is cheeseburgers (point to cheeseburger card) because I think they taste really good.” If the child still did not provide a response, it was re-phrased using yes/no language if possible, and the interviewer continued to the next question. At the beginning of the interview, the interviewer presented a PECS card titled “I don’t know” and informed the child that they could use this card if they did not know an answer, although none of the participants used this card during the interview. In total, four participants used the PECS cards along with verbal responses to answer a portion of the rapport-building questions (n = 4), questions asking what activity preferences (n = 4), and to identify physical complaints they have when physically active (n = 1).
Data Analyses
A descriptive phenomenological approach was used to examine the phenomenon of physical activity directly from the autistic children with intellectual disability and their caregivers. Descriptive phenomenology enables the analysis and understanding of the phenomenon through an investigation of the subjective everyday experiences of participants (Giorgi & Giorgi, 2003; Sundler et al., 2019). This is achieved through open-ended interview questions that are non-leading which relate to the phenomenon of interest, with follow-up questions to clarify the context and meaning derived from their experiences (Giorgi, 1997). The interviews were transcribed verbatim by the first two authors and four research assistants. All identifying information was removed and replaced with descriptive terms, such as “(child’s mother).” The transcriptions were reviewed for accuracy two times by other transcribers, and then a final time by the first author. If any audio was unintelligible and could not be deciphered by the transcribers or by caregivers upon follow-up, it was omitted from the analysis. The transcriptions were then coded using NVivo 12 software. The first and second authors reviewed each of the transcripts several times with an open mind in order to familiarize themselves with the data before identifying codes or marking meanings within the transcript (Sundler et al., 2019). After the authors were familiar with the data, significant statements relevant to the phenomenon of interest were extracted; meaning was identified within these statements regarding whether the segment broadly fit the category of “facilitator” or “barrier.” Facilitators were operationalized as anything that encourages, promotes, enables, or motivates the child’s physical activity, or creates opportunities to engage in physical activity. Barriers were operationalized as anything that prevents, impedes, discourages, or otherwise negatively impacts the child’s physical activity or their ability to engage in physical activity.
Next, segments within the broad “facilitator” and “barrier” clusters were reviewed with greater detail and specificity to better represent the meaning of each segment. Seventeen specific codes were formulated following a review of the significant segments of transcripts to better describe the essence of the child’s experience. The meaning of each code was discussed by the authors. In evaluating the meaning of the codes, four themes were inductively identified that grouped together clusters of codes: within-person barriers (intrapersonal), between-person barriers (interpersonal), within-person facilitators (intrapersonal), and between-person facilitators (interpersonal). The authors reviewed significant segments a final time and assigned specific codes, then matched these codes to a theme; there was 100% inter-rater agreement for this final pass (Table 2).Table 2 Themes derived from the semi-structured interview with youth
Theme Definition Codes n
Within-person Barriers (intrapersonal) Barriers that are intrinsic or internal to the person, such as their own cognitions, their own attitudes, behaviors, and perceptions Physical complaints and injury 3
Dislike sports 6
Fear of injury 3
Child’s own challenging behaviors 2
Negative past experiences 4
Between-person Barriers (interpersonal) Barriers to PA that are caused or made worse by factors that are external to the participant, or feasibly outside of the participant's locus of control Lack of available resources 7
Friends (barrier) 6
Lack of support (rules, physical assistance) 4
Within-person facilitators (intrapersonal) Facilitators that are intrinsic or internal to the participant, such as their own cognitions, their own attitudes, behaviors, and perceptions Intrinsic motivation to improve 5
Positive attitudes about PA 8
“Feels good” 2
Between-person facilitators (interpersonal) Facilitators to PA that are external to the participant, or feasibly outside of the participant's locus of control Available resources in the community 5
School-scheduled PA 13
Family members actively supporting 4
Family members create opportunities 7
Friends (facilitator) 8
Support and encouragement 5
Abbreviations: PA, physical activity
N = 14
Trustworthiness is the extent to which there is confidence in the design, analysis, and presentation of qualitative data in a study (Lincoln & Guba, 1985). To obtain trustworthiness of the research study and data, the criteria of credibility, dependability, confirmability, and authenticity were examined (Schwandt et al., 2007). Following the completion of the study, caregivers were contacted with a summary of findings to confirm if these had resonated with their experiences. The findings were also presented to the local chapter of Special Olympics and to community stakeholders (coaches, staff, and parents) (see Table 3 for a summary of the trustworthiness criteria).Table 3 Trustworthiness
Criterion Method Description
Credibility Methods triangulation Several questions about attitudes, behaviors, facilitators, and barriers of physical activity engagement were asked to both caregivers and youth participants
Member checking Caregivers were contacted with a summary of findings of the current study to confirm if these resonated with their experiences. Findings were presented to the local chapter of Special Olympics and community stakeholders (coaches and staff, parents) as a formal test of credibility
Dependability and confirmability Audit trail Kept a dated log describing all steps in study design, data collection, entry, and analysis, including the creation and revision of the codes & themes and all modifications to definitions
Authenticity Catalytic authenticity The results enable community stakeholders (coaches and local organizations) to act when presented with the findings and begin addressing the barriers experienced by youth and their families in accessing physical activity
Fairness The perceptions of both youth and their caregivers are considered with the understanding that their experiences with the current issue may be different from one another
Results
Descriptive Statistics
The statistical mean and standard deviation were calculated for the WASI-II and caregiver-report questionnaires. The youth had an average FSIQ-4 of 65.64 on the WASI-II (SD = 10.73), ranging from 51 to 75 (see Table 1). Caregivers stated on the HWREPS that their child had difficulties with coordination or motor skills which made it difficult for them to engage in physical activities (m = 5.27, SD = 1.44; see Table 4). Caregivers frequently endorsed statements that their child would engage in more physical activities if they (the caregiver) were more aware of programs available for their child (m = 5.93, SD = 1.16) if the staff was trained to work with individuals with disabilities (m = 5.93, SD = 1.28), and if more programs existed specifically for individuals with disabilities (m = 5.47, SD = 1.92; see Table 5). When asked about the types of physical activities that their child prefers, caregivers said that their child prefers activities that involve an aspect of fun (m = 6.40, SD = 0.910), can be part of a routine (m = 5.80, SD = 1.37), and involve rewards such as prizes or celebrations (m = 5.80, SD = 1.57; see Table 6).Table 4 HWREPS-child factors
Item Mean (SD) Minimum and maximum
1. My child dislikes team sports 4.00 (2.39) 1 to 7
2. My child dislikes individual physical activities 3.33 (2.19) 1 to 7
3. My child has difficulties with coordination or motor skills 5.27 (1.44) 2 to 7
4. My child has behaviors that make participation difficult 3.60 (1.92) 1 to 7
5. My child is aware of the benefits of eating healthy snacks and having a well-balanced diet 5.47 (1.72) 2 to 7
6. My child is aware of the benefits of participating in physical activities 5.40 (1.77) 2 to 7
Abbreviation: HWREPS, Health and Wellness & Risk and Protection Survey
N = 14
Table 5 Caregiver response items—“The person under my care would participate in more physical activities if…”
Item Mean (SD) Minimum and maximum
1. More transportation was available 4.07 (2.25) 1 to 7
2. There were more programs available for individuals with disabilities 5.47 (1.92) 1 to 7
3. We were more aware of what programs were available 5.93 (1.16) 3 to 7
4. Staff were trained to work with individuals with disabilities 5.93 (1.28) 4 to 7
5. They did not have physical health problems 3.80 (1.70) 1 to 7
6. They had fewer behavioral difficulties 4.13 (1.92) 1 to 7
7. There were more age-appropriate options 4.80 (2.04) 1 to 7
8. There were more affordable options 5.40 (2.17) 1 to 7
N = 14
Table 6 Caregiver response items—“The person under my care prefers physical activities that…”
Item Mean (SD) Minimum and maximum
1. They can do by themselves 5.53 (1.60) 1 to 7
2. Involve social interaction with peers 5.33 (1.60) 2 to 7
3. Involve spending time with family 5.20 (1.01) 4 to 7
4. Involve an aspect of fun 6.40 (.910) 4 to 7
5. Involve rewards, such as prizes or celebrations 5.80 (1.57) 3 to 7
6. Can be part of a routine 5.80 (1.37) 3 to 7
N = 14
Interview
Physical Activity Descriptors
The most common activity among youth was soccer, with 10 out of 14 participants either playing soccer with a community team, at recess or school breaks, or with friends. Other common activities included running games (e.g., tag), floor or ice hockey, and basketball. Seven youths said that their favorite activity is sports or some other physical activity; three of these participants indicated that soccer was their favorite activity using the corresponding PECS card. One child said that their favorite physical activity was walking their dogs, and another said that gardening was their favorite, both of which they considered to be physical activities. The other seven participants reported that video games or watching television were their favorite activities, with one participant indicating their preference for video games using the corresponding PECS card. Five participants were identified by caregivers as being currently or previously enrolled in sports programming from Special Olympics.
Within-Person Facilitators (Intrapersonal)
Eight of the fourteen youths reported that they generally enjoy playing sports or engaging in physical activities. Each of these participants reported that they would be willing to play more sports and do more physical activity, such as trying different types of sports that they had not been able to play yet. Four of the remaining youth were content with their level of physical activity, whereas two wanted to do less physical activity each week. Although not every child enjoyed every sport or physical activity that they engaged in, seven children named specific physical activities that they looked forward to playing each week, either with friends, family, or with peers in the community.
Two youths reported that they liked engaging in physical activity because it made them feel good, even if they feel tired or sore. Five youths stated that they were motivated to do physical activity alone because they really liked the activity and wanted to improve their skills. One 9-year-old participant said that they liked being physically active, but that they did not like doing it with friends:Interviewer: What do you do at recess or lunch?
Participant: Run outside.
Interviewer: Do you play with friends at recess or lunch?
Participant: [yelling] No I play by myself, I like it that way!
Between-Person Facilitators (Interpersonal)
The child’s family was stated as a significant facilitator of physical activity. Nine youths stated that their parents or a close family member had either engaged in the physical activity with them or provided support during the activity so that they could perform the activities. Three youths stated that their parents were the ones who taught them the rules or helped develop their skills so that they could play with their friends or play on a sports team in the community. One child said that their older brother helped them practice, whereas another stated that their neighbor served this role. A younger participant, age 8 received help from his father and sister:Participant: I get the ball when I’m goalie but I’m still horrible at goalie.
Interviewer: Hmm. Well if you keep doing it, practice will make you better. If you want to be better.
Participant: Yeah I do. Me and my sister and my dad practice and I get better.
Two youths said that they only liked a sport because a parent did the activity with them:Interviewer: Do you like swimming?
Participant: I don’t actually like it, but…
Interviewer: You don’t actually like it?
Participant: I like it when my parents do that with me. But that’s it.
Two youths also reported that their parents controlled the amount of electronic use that they were allowed after school and on the weekends to encourage greater physical activity. One participant, age 9, stated that they had a system whereby they had to earn electronic use after doing physical activities in the home or community:Interviewer: What are the things you do when you’re not at school?
Participant: Sports, like right now… but I want to play [video] games.
Interviewer: Do you watch a lot of TV, play Minecraft on the weekends too?
Participant: Uh, I don’t get to play Minecraft yet until I earn points for Minecraft, so… I need to do sports. Yeah.
One of the most common ways that family facilitated physical activity was by walking with the child to and from school. Three children remarked that this was some of the only physical activities that they would do during the week because they did not like sports or playing with others at recess. However, this also meant that for several months during the year, the weather would not permit walking to and from school, and several children mentioned that bad weather limits their ability to do physical activities outside with their family.
Several youths reported that many non-family members, such as coaches and teachers, provided support when they engaged in a physical activity in the community. When asked if they would like more help from other people in learning the rules or engaging in an activity, ten participants said that they have sufficient support already, or that they have someone that they can ask for help when needed. A common theme was liking having their teachers support them and encourage them to engage in physical activities during class:Interviewer: Do, uh, any people help you with doing these [gym] activities?
Participant: Yeah.
Interviewer: Yeah? Who helps you?
Participant: The teachers.
Interviewer: The teachers at your school?
Participant: Yeah! They say I – I try my best.
The majority of youth said that they engaged in scheduled physical education classes at school. The number of days that the child had gym class varied between 2 and 5 days per week. Opinions about whether or not they liked gym class varied among respondents, but they ultimately ended up engaging in gym class when they had it. School is also the place where many youths were able to engage in physical activities with their friends. Younger participants commonly reported playing active games at recess such as tag or soccer. Seven youths said that doing physical activities and sports with their friends at recess or during school breaks was fun. Older participants less frequently endorsed engaging in physical activities during their breaks primarily because they do not get recess in high school. However, two youths in high school prioritized playing basketball at lunch breaks as a way to hang out with their friends.
Within-Person Barriers (Intrapersonal)
At the beginning of the interview, each child was asked to tell the interviewer their favorite activity. Half of the youth told the interviewer that video games or watching television were their favorite activities, two of which indicated that they would rather play video games than do physical activities. Two youths stated that they did not like doing sports where other people have to help them, so they avoided engaging in most sports. Six youths plainly stated that they did not like sports at all. One child specifically mentioned that video games prevented him from doing more physical activity:Interviewer: So what are some things that stop you from doing physical activities?
Participant: Having a bad addiction to video games, apparently.
Interviewer: Yeah? You’re distracted by video games?
Participant: Yeah. Definitely.
When asked what activities they do on the weekend, 10 of the 14 participants said they mostly play games and watch TV. One child said that they loved the weekend particularly because they could play on the computer:Interviewer: What do you do on the weekend?
Participant: On the weekends I stay home, I watch videos on the computer.
Interviewer: And play video games on the computer?
Participant: Yeah, I do. I watch YouTube and I enjoy Saturday morning a lot, it’s my favourite morning.
Interviewer: Do you go outside and do things outside on the weekends?
Participant: No, not really.
Two children explained that their behavior prevented them from playing a sport. One child said that he would continuously pick up the soccer ball during the game because he thought the rules were “stupid.” Another child mentioned that he is prevented from playing soccer because he would bite and hurt other players:Participant: I actually kind of like soccer.
Interviewer: Yeah?
Participant: But people keep stopping me from doing it. I can actually, I can beat them and bite them and kick them.
Interviewer: Oh. But our mouths are not for hurting, right, and our feet aren’t for hurting. We don’t wanna hurt other people, that’s not part of the game.
Participant: The biting is
Interviewer: Noooo. No it’s not!
Participant: Yeah it is.
Several youths mentioned that they had negative experiences playing sports or had a fear of getting injured, which diminished their desire to engage in physical activities. Two youths stated that they completely avoided playing sports that are physical, namely lacrosse and rugby, even though their friends were playing them. Four youths explained that they had a bad experience playing a sport and have stayed away from that sport completely as a result. One older child recalled several negative experiences in gym class that led him to stop doing all sports entirely:Interviewer: Do you like to play sports?
Participant: No. When I went to my first school and did gym, I got hit in the face with the ball. I chipped a tooth when I did volleyball one time. I got hit in the face. And then one time at gym we were doing, like, running, back and forth, and I crashed into one of the guys and I got a little bit bruised. So that’s why I’m scared I’m gonna fall and get hurt ‘cause I don’t like doing sports.
Interviewer: So you had some bad experiences?
Participant: And then I don’t do sports ever again.
When asked what things stopped them from doing more physical activity, three youths worried that they would get injured:Interviewer: What sports do you not like?
Participant: Soc- like I don’t like soccer, I don’t. I get worried that I get kicked in the head. It scares me.
Others stated that they felt too much pain when doing physical activity, thus they disliked sports and being physically active. Two youths said that they experienced headaches when playing sports, and one child used the PECS cards to indicate that their eyes and back started to hurt too much when they swim or do physical activities outside.
Between-Person Barriers (Interpersonal)
While having friends was a facilitator for some youths, friends had also interfered with some participants’ ability to engage in physical activity. Three older participants reported that other activities with friends got in the way of doing physical activities, such as listening to music and playing electronic games together. Three younger participants stated that their friends chose activities at lunch that they did not like, so they do not get to play the games or sports that they want:Interviewer: So you like playing the tag game with [name of friend] at recess?
Participant: They say it’s fun but it actually is not fun. [Name of friends] always play that game. I want to play soccer.
Interviewer: Do you guys play soccer at recess time?
Participant: Noooo. I always wanted to go [play] soccer.
Three youths stated that they did not have friends to do activities with, so they spent their recesses or breaks alone. Two of these respondents said that they instead used the computer inside or played on their phones. The other respondent said that he did not really have friends, and when asked what he does during recess, he said that he did nothing but sit and wait for recess to finish.
In total, four youths said that they want more people to help them play sports and that this is a reason why they did not do more physical activity:Interviewer: Do people help you play these games?
Participant: No.
Interviewer: No? So you usually just do it alone?
Participant: <Exhales>. I need to know the game first. And I never know how to play. I need to know the game so [I] can play with them.
Interviewer: So do these people help you?
Participant: No.
Only three youths indicated that they had regular, scheduled physical activities during the weekend. One child said that the recreation center in their community had closed, so they no longer had an accessible place to swim. Another child said that the swimming program for children with disabilities that they had been attending stopped suddenly due to a lack of instructors and that his parents did not know when another program would restart. Three youths each said that they desired to play a specific sport, but their parents had not yet found a team that they could be on.
Discussion
The current study examined the facilitators and barriers to engaging in physical activity for autistic youth with mild intellectual disability between the ages of 8 and 16 years old. Caregivers provided their views on their child’s physical activity through a questionnaire, while youth participants completed a semi-structured interview with researchers. Several facilitators to physical activity involvement were reported by the youth themselves. Having a personal interest in sports and enjoying physical activities were common facilitators, with ten of the 14 youth indicating that they have the desire to do more physical activity. Youth reported that their parents were the primary facilitator of physical activities in the community, helping to provide support when needed. The majority of youth reported that they engaged in physical activity at school, primarily through gym class and/or recess for younger participants. Similar to the support provided by parents, teachers were commonly reported by youth as being motivators and facilitators of physical activity at school. Many youths also reported that they would play sports or other physical activities with friends at school and that these activities were very enjoyable.
The youths’ friends were identified as either a facilitator or barrier, as they would influence the types of activities the child would engage in. For a few children, the activities with friends were sedentary such as listening to music or playing video games. Other children reported that they did not have friends, and as a result, spent their free time alone and sedentary. Similarly, many children reported interest in video games and television, preferring to engage in these activities over physical activities. Injury or fear of injury was a barrier, where previous negative experiences decreased their desire to engage in sports in the future. However, for those who wanted to engage in more sports, several youths reported being unable to find a team or not having access to a place to play.
The questionnaire completed by caregivers suggests a desire for knowledge of available programs for their child to participate in, and that staff be specifically trained and capable of working with their youth with disabilities. Caregivers also reported that their child prefers activities that involve an aspect of fun and involve rewards and activities that can be a part of their routine. Most caregivers reported that their child was aware of the benefits of participating in physical activities and of eating healthy.
The desire for staff to have better training to work with their autistic child with intellectual disability is a consistent finding throughout previous research (Nichols et al., 2019; Stanish et al., 2016). Caregivers reported that their child had motor difficulties that impeded their ability to engage in physical activity. Poor motor control can be present in autistic children with low and high intellectual ability (Kaur et al., 2018). Many parents report that their child is unable to keep up with other children while participating in physical activity due to poor coordination or motor ability (Arnell et al., 2020; Obrusnikova & Miccinello, 2012). Staff who have training and experience working with autistic children and youth diagnosed with intellectual disability can help develop the individual’s skills in a safe and engaging way which may help to reduce this as a barrier to physical activity involvement. Several youth participants reported that this role is currently being filled by their parents or family members, so having access to trained personnel who can facilitate skill development and physical activity can also help reduce the need for parent involvement in this area.
The idea that the child is rewarded for engaging in physical activity was brought up by both caregivers and youth in the current study. One child said that they are rewarded with access to their video games after completing a certain amount of physical activity, and caregivers agreed that their child prefers activities that involve some sort of reward or positive consequence. These findings are supported by Healy and Marchand (2020) who identified that weekly rewards for engaging in a parent-mediated physical activity intervention were highly motivating for autistic youth. Using rewards and similar extrinsic motivators can be helpful for youth who show a lack of intrinsic interest or motivation to engage in physical activity. However, self-determined and intrinsic motivation, rather than extrinsic motivation, is associated with greater and sustained physical activity behaviors (Owen et al., 2014). When activity is intrinsically motivated, the person typically engages in the activity for longer periods of time and with greater effort than when it is extrinsically motivated (Deci & Ryan, 2010). In addition, a lack of self-efficacy has been linked to low motivation for engaging in physical activity in youth with intellectual disability, and the importance of social support for building competency in fitness domains has been emphasized as a critical factor for increasing their motivation and physical activity involvement (Hutzler & Korsensky, 2010). This is an important consideration for increasing the physical activity rates of autistic youth with intellectual disability as several participants in the current study and in previous studies have shown a significant lack of interest in physical activities (Bossink et al., 2017; Stanish et al., 2016). Efforts to captivate the interests of autistic youth with intellectual disabilities such as tailoring activities to their interests, as well as efforts to increase their self-efficacy by ensuring that adequate social supports are in place, may ultimately help enhance physical activity involvement and intrinsic motivation to engage in physical activity.
A few caregivers indicated that their child’s behaviors impeded their ability to engage in physical activities. Two youths also explained how their behaviors prevented them from playing the sport that they liked, specifically by not following the rules of the game or by harming other players. For some, physical activities may help reduce negative behaviors. For example, a previous exercise-based intervention targeting emotion regulation in autistic children was shown to be efficacious, resulting in the reduction of behavioral problems for those in the exercise intervention compared to a comparison group (Tse, 2020). In addition, a systematic review by Bremer et al. (2016) identified that exercise interventions in martial arts training and horseback riding showed moderate to large effects in reducing socio-emotional and stereotypic behavioral challenges. However, coaches, trainers, or teachers who are not trained or experienced in managing children’s challenging and disruptive behaviors will likely be unable to use physical activity to improve behavioral outcomes. Thus, specific programming with trained staff may be necessary to engage youth who exhibit challenging behaviors.
The interviews also revealed that youths were fearful of injury or had previous negative experiences with physical activity that resulted in injury. Fear of injury has previously been identified as a barrier for adults with intellectual disabilities (van Schijndel-Speet et al., 2014). Autistic adolescents have previously reported greater fear of getting hurt during physical activities compared to non-autistic adolescents, which prevents them from participating in sports and other physical activities (Stanish et al., 2015). These fears of injury, or prior experiences of getting injured, can negatively impact the likelihood of participating in physical activities. Addressing these fears is one important step in encouraging physical activity in youth and adults who have substantial injury-related anxiety. Providing an environment that is free from risks is not entirely feasible, particularly if physicality is instrumental to the sport as is the case with football or rugby. However, other activities may be modified to accommodate those who have fears or have been injured, such as playing flag football or disallowing slide tackles in soccer.
Several of the within-person barriers identified in the study are addressed in community programs by using behavioral principles for working with youth with developmental disabilities. Non-profit organizations such as the Canucks Autism Network and Jumpstart teach coaches behavioral principles such as positive reinforcement and goal setting to use with autistic athletes (Canucks Autism Network, 2022; Jumpstart, 2022). These strategies are used to improve motivation to engage in physical activity while developing skills that increase the likelihood of future sports involvement (Arthur-Banning & Windbiel, 2022; Luiselli, 2016). Between-person barriers, such as a lack of programming for children with disabilities, are also addressed by community organizations such as Special Olympics. For example, the FUNdamentals program offered by Special Olympics Canada is an introductory and family-centered program to develop basic movement skills through basic sport skills for youth ages 7 to 12 years old with an intellectual disability (Special Olympics Canada, 2022). These skills can be practiced and developed under the supervision of trained coaches and staff with the goal of transitioning into other programs offered by Special Olympics or in the athlete’s community. Such accommodations can be made to make the activity more inclusive, which may generate an interest in that activity and ease injury-related worry or fear.
A strength of this study was the collection of information from autistic youth with intellectual disability directly, in addition to information from caregivers. Interview formats which consider the communication styles of participants provide a feasible way to obtain information and opinions from the youth themselves (Courchesne et al., 2022), which lends valuable insight into how they perceive the challenges or helpful factors to their physical activity. Compared to self-report questionnaires, interviews require a reduced cognitive demand in reporting answers while also allowing the opportunity for the participant to ask questions and for the researcher to prompt or clarify the participant’s responses. Additionally, relying solely on information obtained from caregivers may provide an incomplete picture of the ways that autistic youth with intellectual disability view and engage in physical activity across various contexts. For example, caregivers may have limited knowledge on the types of activities that their child engages in during recess or breaks at school.
Another strength of the study was looking at the population of youth diagnosed with both ASD and mild intellectual disability. Compared to children diagnosed with ASD or intellectual disability alone, autistic youth with intellectual disability report greater difficulty participating in leisure activities, activities in the home, with friends, and participating in classroom learning (Hilton et al., 2021) and are under-represented in research (Russell et al., 2019). Another strength of the study was working with consultants to create questions on the parent-report measure and to refine the interview protocol. Accessing the expertise of consultants is necessary to develop study materials that are accessible to participants and to meaningfully capture their lived experiences. While the ideal of participatory and inclusive research is to include consultants and community stakeholders in study design, implementation, and management (Bigby et al., 2014), the current study is a positive example of involving community stakeholders in research that has potential implications for that population.
Limitations and Future Research
A limitation of the current study is having only one female youth as the experiences of autistic girls with intellectual disabilities may be different from those of boys. Another limitation is the bias in the self-selection of participants to the study. The study was developed to be included in a free day camp involving physical and creative activities and to develop health literacy for young people with intellectual disabilities. Although many efforts were made to reduce barriers to involvement in the current study (i.e., free participation, compensation for travel, trained and competent supervision, complementary food and beverages), individuals who face more extreme barriers to participation (e.g., physical mobility issues, poverty) may not have been able to attend. Similarly, caregivers and youth who are interested in physical activities may be more likely to sign up for our day camp than those who do not value physical activity. Another limitation arises in the use of semi-structured interviews. Semi-structured interviews may influence or contaminate participant responses through potentially leading, suggestive, direct, or repeated questions (Newton, 2010). The validity and reliability of the information obtained in the interviews are limited by the flexibility in question order, wording, and prompting between and within interviews which makes it difficult to make comparisons across participants.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (PDF 161 KB)
Acknowledgements
The authors thank the consultants for their time, effort, and expertise, and the research assistants at the Autism and Developmental Disabilities Lab for assisting with data collection and transcription.
Author Contribution
TQB: designed and executed the study, led data analyses, and wrote the paper. CLM: collaborated with designing the study, assisted with data analyses, and assisted with writing the paper. GI: collaborated with the design of the study and edited the final manuscript. All authors approved the final version of the manuscript for submission.
Funding
Special Olympics International Youth Innovation Grant and Special Olympics British Columbia Youth Engagement Grant to Troy Q. Boucher.
Data Availability
Due to the identifying nature of the qualitative data in this study, supporting data is not available.
Declarations
Research Involving Human Participants and/or Animals
The current study was approved by the Office of Research Ethics at Simon Fraser University (protocol number 2019s0458). The study was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki.
Informed Consent
Informed consent was collected from all legal guardians for their participation and for their child’s participation in the current study. Verbal assent was collected from all child participants prior to the semi-structured interview and prior to the brief measure of cognitive ability. The author affirms that the legal guardians provided informed consent for the publication of de-identified interview segments conducted with their child.
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
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). American Psychiatric Association Publishing.
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). American Psychiatric Association Publishing.
American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev). American Psychiatric Association Publishing.
Arnell S Jerlinder K Lundqvist LO Perceptions of physical activity participation among adolescents with autism spectrum disorders: A conceptual model of conditional participation Journal of Autism and Developmental Disorders 2018 48 5 1792 1802 10.1007/s10803-017-3436-2 29236210
Arnell S Jerlinder K Lundqvist LO Parents’ perceptions and concerns about physical activity participation among adolescents with autism spectrum disorder Autism 2020 24 8 2243 2255 10.1177/1362361320942092 32713182
Arthur-Banning S Windbiel K Recreational programming for children with autism: Using applied behavior analysis values to encourage participation Journal of Physical Education, Recreation & Dance 2022 93 9 7 14 10.1080/07303084.2022.2120124
Bhattacharya, K. (2017). Fundamentals of qualitative research: A practical guide. Routledge.
Bigby C Frawley P Ramcharan P Conceptualizing inclusive research with people with intellectual disability Journal of Applied Research in Intellectual Disabilities 2014 27 1 3 12 10.1111/jar.12083 24390972
Bölte S The power of words: Is qualitative research as important as quantitative research in the study of autism? Autism 2014 18 2 67 68 10.1177/1362361313517367 24596959
Bondy AS Frost LA The picture exchange communication system Focus on Autistic Behavior 1994 9 3 1 19 10.1177/108835769400900301
Bossink LW van der Putten AA Vlaskamp C Understanding low levels of physical activity in people with intellectual disabilities: A systematic review to identify barriers and facilitators Research in Developmental Disabilities 2017 68 95 110 10.1016/j.ridd.2017.06.008 28750208
Bremer E Crozier M Lloyd M A systematic review of the behavioural outcomes following exercise interventions for children and youth with autism spectrum disorder Autism 2016 20 8 899 915 10.1177/1362361315616002 26823546
Brown DM Arbour-Nicitopoulos KP Martin Ginis KA Latimer-Cheung AE Bassett-Gunter RL Examining the relationship between parent physical activity support behaviour and physical activity among children and youth with autism spectrum disorder Autism 2020 24 7 1783 1794 10.1177/1362361320922658 32476438
Canucks Autism Network (2022). Training & try-it workshops. https://www.canucksautism.ca/training-community-engagement/sport/
Case L Ross S Yun J Physical activity guideline compliance among a national sample of children with various developmental disabilities Disability and Health Journal 2020 13 2 1 7 10.1016/j.dhjo.2019.100881
Chown N Robinson J Beardon L Downing J Hughes L Leatherland J Fox K Hickman L MacGregor D Improving research about us, with us: A draft framework for inclusive autism research Disability & Society 2017 32 5 720 734 10.1080/09687599.2017.1320273
Cooper SA McLean G Guthrie B McConnachie A Mercer S Sullivan F Morrison J Multiple physical and mental health comorbidity in adults with intellectual disabilities: Population-based cross-sectional analysis BMC Family Practice 2015 16 1 11 10.1186/s12875-015-0329-3 25608667
Courchesne V Tesfaye R Mirenda P Nicholas D Mitchell W Singh I Zwaigenbaum L Elsabbagh M Autism voices: A novel method to access first-person perspective of autistic youth Autism 2022 26 5 1123 1136 10.1177/13623613211042128 34482746
Deci EL Ryan RM Intrinsic motivation 2010 Plenum Press
Ginis KAM van der Ploeg HP Foster C Lai B McBride CB Ng K Pratt M Shirazipour CH Smith B Vásquez PM Heath GW Participation of people living with disabilities in physical activity: A global perspective The Lancet 2021 398 10298 443 455 10.1016/S0140-6736(21)01164-8
Giorgi A The theory, practice, and evaluation of the phenomenological method as a qualitative research procedure Journal of Phenomenological Psychology 1997 28 2 235 260 10.1163/156916297X00103
Giorgi, A. P., & Giorgi, B. M. (2003). The descriptive phenomenological psychological method. In P. M. Camic, J. E. Rhodes, & L. Yardley (Eds.), Qualitative research in psychology: Expanding perspectives in methodology and design (pp. 243–273). American Psychological Association. 10.1037/10595-013
Healy S Marchand G The feasibility of Project CHASE: A Facebook-delivered, parent-mediated physical activity intervention for children with autism International Journal of Disability, Development and Education 2020 67 2 225 242 10.1080/1034912X.2019.1597968
Healy S Msetfi R Gallagher S ‘Happy and a bit nervous’: The experiences of children with autism in physical education British Journal of Learning Disabilities 2013 41 3 222 228 10.1111/bld.12053
Hilton CL Ratcliff K Hong I Participation difficulties in autism spectrum disorders and intellectual disabilities: Findings from the 2011 survey of pathway to diagnosis and services Journal of Autism and Developmental Disorders 2021 51 4 1210 1223 10.1007/s10803-020-04591-4 32648148
Hutzler Y Korsensky O Motivational correlates of physical activity in persons with an intellectual disability: A systematic literature review Journal of Intellectual Disability Research 2010 54 9 767 786 10.1111/j.1365-2788.2010.01313x 20712695
Jumpstart (2022). Coaching kids of all abilities. https://jumpstart.canadiantire.ca/pages/coaching-kids-of-all-abilities
Kapsal NJ Dicke T Morin AJ Vasconcellos D Maïano C Lee J Lonsdale C Effects of physical activity on the physical and psychosocial health of youth with intellectual disabilities: A systematic review and meta-analysis Journal of Physical Activity and Health 2019 16 12 1187 1195 10.1123/jpah.2018-0675 31586434
Kaur M Srinivasan SM Bhat AN Comparing motor performance, praxis, coordination, and interpersonal synchrony between children with and without autism spectrum disorder (ASD) Research in Developmental Disabilities 2018 72 79 95 10.1016/j.ridd.2017.10.025 29121516
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. SAGE.
Luiselli, J. K. (2016). Increasing and maintaining exercise and physical activity. In J. K. Luiselli (Ed.), Behavioral health promotion and intervention in intellectual and developmental disabilities (pp. 73–93). Springer. 10.1007/978-3-319-27297-9
McCoy SM Morgan K Obesity, physical activity, and sedentary behaviors in adolescents with autism spectrum disorder compared with typically developing peers Autism 2020 24 2 387 399 10.1177/1362361319861579 31364386
McFee, E.M. (2019). An exploratory analysis of parental factors related to physical activity in children living with autism. [Master’s thesis, University of British Columbia (Okanagan)]. University of British Columbia Library. 10.14288/1.0379612
Newton N The use of semi-structured interviews in qualitative research: Strengths and weaknesses Exploring Qualitative Methods 2010 1 1 1 11
Nichols C Block ME Bishop JC McIntire B Physical activity in young adults with autism spectrum disorder: Parental perceptions of barriers and facilitators Autism 2019 23 6 1398 1407 10.1177/1362361318810221 30486668
Obrusnikova I Cavalier AR Perceived barriers and facilitators of participation in after-school physical activity by children with autism spectrum disorders Journal of Developmental and Physical Disabilities 2011 23 3 195 211 10.1007/s10882-010-9215-z
Obrusnikova I Miccinello DL Parent perceptions of factors influencing after-school physical activity of children with autism spectrum disorders Adapted Physical Activity Quarterly 2012 29 1 63 80 10.1123/apaq.29.1.63 22190053
Owen KB Smith J Lubans DR Ng JY Lonsdale C Self-determined motivation and physical activity in children and adolescents: A systematic review and meta-analysis Preventive Medicine 2014 67 270 279 10.1016/j.ypmed.2014.07.033 25073077
Rech JP Irwin JM Rosen AB Baldwin J Schenkelberg M Comparison of physical activity between children with and without autism spectrum disorder: A systematic review and meta-analysis Adapted Physical Activity Quarterly 2022 39 4 456 481 10.1123/apaq.2021-0152 35405634
Russell G Mandy W Elliott D White R Pittwood T Ford T Selection bias on intellectual ability in autism research: A cross-sectional review and meta-analysis Molecular Autism 2019 10 1 10 10.1186/s13229-019-0260-x 30647876
Scharoun, S. M., Wright, K. T., Robertson-Wilson, J. E., Fletcher, P. C., & Bryden, P. J. (2017). Physical activity in individuals with autism spectrum disorders (ASD): A review. In M. Fitzgerald & J. Yip (Eds.), Autism: Paradigms, recent research and clinical applications (pp. 301–331). InTech Open. 10.5772/66680
Schwandt TA Lincoln YS Guba EG Judging interpretations: But is it rigorous? Trustworthiness and authenticity in naturalistic evaluation New Directions for Evaluation 2007 114 11 25 10.1002/ev.223
Special Olympics Canada. (2022). FUNdamentals. https://www.specialolympics.ca/programs-gamesprograms/fundamentals
Stanish HI Curtin C Must A Phillips S Maslin M Bandini LG Enjoyment, barriers, and beliefs about physical activity in adolescents with and without autism spectrum disorder Adapted Physical Activity Quarterly 2015 32 4 302 317 10.1123/APAQ.2015-0038 26485735
Stanish HI Curtin C Must A Phillips S Maslin M Bandini LG Physical activity enjoyment, perceived barriers, and beliefs among adolescents with and without intellectual disabilities Journal of Physical Activity and Health 2016 13 1 102 110 10.1123/jpah.2014-0548 25830443
Sundler AJ Lindberg E Nilsson C Palmér L Qualitative thematic analysis based on descriptive phenomenology Nursing Open 2019 6 3 733 739 10.1002/nop2.275 31367394
Tint A Maughan AL Weiss JA Community participation of youth with intellectual disability and autism spectrum disorder Journal of Intellectual Disability Research 2017 61 2 168 180 10.1111/jir.12311 27492816
Tse AC Brief report: Impact of a physical exercise intervention on emotion regulation and behavioral functioning in children with autism spectrum disorder Journal of Autism and Developmental Disorders 2020 50 11 4191 4198 10.1007/s10803-020-04418-2 32130593
van Schijndel-Speet M Evenhuis HM van Wijck R van Empelen P Echteld MA Facilitators and barriers to physical activity as perceived by older adults with intellectual disability Intellectual and Developmental Disabilities 2014 52 3 175 186 10.1352/1934-9556-52.3.175 24937743
Wechsler, D. (2011). Wechsler abbreviated scale of intelligence – Second edition (WASI-II). Psychological Corporation.
Weiss JA Robinson S Harlow M Mosher A Fraser-Thomas J Balogh R Lunsky Y Individual and contextual predictors of retention in Special Olympics for youth with intellectual disability: Who stays involved? Journal of Intellectual Disability Research 2020 64 7 512 523 10.1111/jir.12731 32390189
| 0 | PMC9742026 | NO-CC CODE | 2022-12-14 23:29:59 | no | Adv Neurodev Disord. 2022 Dec 12;:1-13 | utf-8 | Adv Neurodev Disord | 2,022 | 10.1007/s41252-022-00310-5 | oa_other |
==== Front
Drugs Ther Perspect
Drugs Ther Perspect
Drugs & Therapy Perspectives
1172-0360
1179-1977
Springer International Publishing Cham
975
10.1007/s40267-022-00975-x
Practical Issues and Updates
Antidepressants with anti-inflammatory properties may be useful in long COVID depression
Fenton Caroline
http://orcid.org/0000-0002-6519-7831
Lee Arnold [email protected]
grid.420067.7 0000 0004 0372 1209 Springer Nature, Private Bag 65901, Mairangi Bay, Auckland, 0754 New Zealand
12 12 2022
16
4 12 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.
Long COVID, which is characterised by the presence of persistent symptoms following a COVID infection, may also cause long COVID depression (LCD). Although the risk factors for LCD are not directly characterised, prior mental health visits were associated with an increased risk for long COVID, whereas antidepressant use was not. Current evidence suggests that pro-inflammatory factors in the brain are linked to LCD, thus anti-inflammatory agents may be useful in mitigating LCD. Limited evidence suggests that selective serotonin reuptake inhibitors may also be effective in the treatment of LCD.
==== Body
pmcLong COVID is vaguely defined, but well recognised
Long COVID (LC) may occur in adults after an infection with SARS-CoV-2, the virus that causes COVID-19 (referred to here as COVID) [1]. Estimated prevalence varies from 10–20% [2] to 50% in adults [1]. The prevalence of LC and the often profound effect on people’s quality of life highlight the need to identify effective treatments [1], but even defining LC has been difficult [2, 3].
According to the definition set by the WHO, LC usually occurs 3 months after a probable or confirmed SARS-CoV-2 infection. LC is typically accompanied by the presence of new or persistent symptoms that affect everyday functioning and cannot be explained by an alternative diagnosis, which last ≥ 2 months [2]. The US definition simply includes signs and symptoms occurring ≥ 4 weeks after a COVID infection; the UK splits LC into symptomatic COVID lasting for 4–12 weeks and post-COVID-19 (PC) syndrome beyond this timeframe [2, 3].
The signs and symptoms, as well as the patterns of relapse and remission with LC are heterogeneous, and most organ systems may be potentially affected [2, 3]. However, fatigue, shortness of breath, and cognitive dysfunction are among the most commonly reported symptoms [2], with fatigue being an important factor associated with depression [1].
Depressive symptoms are clinically significant in 30–40% of patients 1–12 months after a COVID infection [1]. While also heterogenous and variably defined, LC depression (LCD) shares features of major depressive disorder (MDD), including abnormal brain imaging, symptoms such as fatigue [1], distress, apathy, insomnia and cognitive impairment, and negative thinking styles [1, 4]. Depression is also part of the neurological manifestations of COVID (neuro-COVID or nCOVID) and may incorporate conditions as diverse as post-traumatic stress disorder and new or worsening Alzheimer’s disease [4].
This article summarises early data on the epidemiology, pathophysiology and potential treatments of LCD, as reviewed by Mazza et al [1]. Other large studies including LCD among broader LC symptoms were also referenced [5, 6]. The data presented in this article are current as of October 2022.
A large study has indicated potential risk factors for long COVID
Risk factors for LC in general were identified in a US Veterans Affairs database study including ≈ 200,000 people who were infected with COVID [6]. The risk for LC correlated with higher Charlson Comorbidity Index scores (a weighted measure of comorbidity, with higher values indicating greater severity); underlying asthma, diabetes, chronic kidney disease, chronic obstructive pulmonary disease, congestive heart failure and other cardiovascular disease were significant risk factors [adjusted odds ratios (ORs) of 1.07–1.42, with the ranges of the 95% CIs all excluding 1] [6].
More severe COVID infections may be associated with greater risk for LC [6]. The presence of symptomatic illness [OR 1.46 (95% CI 1.42–1.52) in people with 1 or 2 symptoms; higher risk in people with ≥ 3 symptoms], hospitalisation [OR 2.6 (95% CI 2.51–2.69)] and mechanical ventilation [OR 2.5 (95% CI 2.26–2.69)] were risk factors for LC [6].
Regarding depression-related risk factors, an increased risk of LC was observed in people who made 7–19 mental health visits [OR 1.05 (95% CI 1.01–1.09)] or ≥ 20 visits [OR 1.16 (95% CI 1.11–1.21)] in 2 years; but not in people who made ≤ 6 visits [OR 1.02 (95% CI 0.98–1.05)] [6]. However, the effects of comorbid depression were not directly assessed in this study. The presence of fatigue was also a risk factor for LC [OR 1.51 (95% CI 1.44–1.59)], but antidepressant use did not appear to increase the risk of LC [OR 1.02 (0.99–1.05)] [6].
A lower risk for LC was observed in people who received two doses of vaccines versus those receiving no vaccination [OR 0.78 (95% CI 0.68–0.90)], in contrast to people who received one dose of vaccine [OR 1.03 (95% CI 0.95–1.12)] [6].
Long COVID depression may be common...
The prevalence of clinically-relevant depressive symptoms in people who were infected with COVID varies from approximately 20–50% [1, 3]. The use of different screening instruments may have influenced estimates in meta-analyses [1]; the most commonly used tool was the Patient-Health-Questionnaire-9, which identified depressive symptoms in about half of all patients (47–52%), versus ≈ 20–22% if the Hospital Anxiety and Depression scale or Symptom Checklist-90 were used [7]. In two analyses, mild, moderate and severe LCD occurred in ≈ 30%, ≈ 15% and ≈ 8% of all people who were infected with COVID [1].
It may be difficult to distinguish LCD from depression caused by broader COVID-related stress and social isolation [8] and a full discussion of those issues is outside the scope of this article. However, based on a cohort in one study published in early 2021 [9], LCD represented an increase in the incidence of depression [1]. Approximately 42% of COVID-infected patients, versus 32% of the general population and 31% of health workers developed depressive symptoms [9].
...especially in lonely people after severe COVID
A history of severe COVID infection consistently increased the risk for LCD in published studies, with sociodemographic risk factors also evident [1, 6, 10]. In meta-analyses, women and patients with severe COVID were more prone to LCD [1]. Other studies highlighted COVID-related social isolation [4], low or high education and wide exposure to social media as additional risk factors [10].
Biomarkers may indicate the likelihood of developing LCD [10, 11]. Consistent with a link to severe COVID (as is the case for all CNS complications [4]), high levels of cortisol, C-reactive protein (CRP) and interleukin (IL)-1β increase depression risks [10]; while other biomarkers are under investigation [1, 4, 12]. A study preprint reports that current HIV infection or transient reactivation of Epstein Barr virus (EBV) during acute COVID increased LC symptoms (including fatigue and cognitive dysfunction) [11], while prior cytomegalovirus infection was protective against LC symptoms [11].
Inflammation and depression may be related
A known, but complex two-way association between systemic inflammation and nitro-oxidative stress, and mood disorders, poor cognitive functioning and fatigue is supported by early COVID data [1, 13]. Long-term depression is not unique to COVID, as infection with SARS, influenza and EBV are all known risks for persistent mental health issues [8, 14].
Aside from numerous social stressors, the persistence of both the virus and inflammation, cerebrovascular damage including strokes, disrupted brain glucose metabolism, hypoxia and hypoxia-induced mitochondrial damage are all possible underlying mechanisms for LCD [1, 8]. In terms of biomarkers, investigations have linked CRP, IL-6, the neutrophil/lymphocyte ratio, the neutrophil count and the systemic inflammation index with LCD [1]. Sphingomyelinase activity (Tables 1, 2) correlates with inflammation and COVID severity [15].Table 1 Initial data on selective serotonin reuptake inhibitors based on a review by Mazza et al. [1]
Examples of drugs Citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine and sertraline
Potential mechanisms of action Fluvoxamine, sertraline, fluoxetine and citalopram activate S1R, which ↓ pro-inflammatory cytokine production, especially ILs [1, 12]
Fluvoxamine ↓ platelet activation and aggregation via ↓ serotonin loading into platelets, which may ↓ thrombosis and ↓ elevated serotonin levels that may result in acute respiratory distress
Fluoxetine, sertraline and paroxetine and, likely, citalopram, escitalopram, fluvoxamine ↓ viral entry via FIASMA activity [15]
Fluvoxamine and fluoxetine accumulate in lysosomes, ↓ lysosome-mediated viral exit
Several SSRIs will inhibit CYP1A2, ↑ plasma melatonin levels, which is an antioxidant and ↓ inflammation; fluoxetine also has antioxidant properties [12]
Clinical data In three trials in acute COVID, fluvoxamine ↓ clinical deterioration and hospitalisation, and in one randomised trial, ↓ deaths by 90%
Fluvoxamine ↓ mortality in comparison with matched controls in pts admitted to the ICU [12]
Fluoxetine and fluvoxamine (but not other SSRIs) ↓ all-cause mortality in pts with COVID in a large cohort analysis of health records
Pre-COVID SSRI exposure dose-dependently ↓ ED visits or hospital admissions within 30 d of a COVID infection in a large retrospective cohort study including all antidepressants [15]
Pre-COVID SSRI exposure at ≥ 20 mg/d fluoxetine equivalent (but not lower doses) ↓ in-hospital 28 d mortality in a large cohort study [20]
ILs interleukins, ED emergency department, FIASMA functional inhibition of acid sphingomyelinase, pts patients, S1R sigma-1 receptor, ↓ decrease(s/ed), ↑ increase(s)
Table 2 Initial data on pharmacological therapies for acute COVID that reduce the incidence of long COVID depression based on a review by Mazza et al. [1]
Studied drugs and proposed mechanisms of action Clinical outcomes
Antidepressants
Bupropion is a non-SSRI dopamine and noradrenaline reuptake inhibitor with possible S1R agonist activity (↓ inflammation) [15] At all doses, exposure to bupropion prior to COVID infection ↓ ED visits or hospital admissions within 30 d of an infection [15]
Tricyclic antidepressants (amitriptyline, clomipramine, desipramine, doxepin, imipramine, nortriptyline) and venlafaxine (an SNRI) all have in vitro FIASMA activity [15] Pre-COVID use dose-dependently ↓ ED visits or hospital admissions within 30 d of COVID infection, and ↓ intubation and death [15]
Pre-COVID exposure to FIASMA-active agents, especially tricyclics (vs SSRIs) ↓ in-hospital mortality in large cohort study [20]
SSRIs, SNRIs and tricyclic antidepressants ↓ inflammation and viral activity ↓ risk of intubation or death, especially with escitalopram, fluoxetine and venlafaxine [12]
Melatonin-related therapies
Melatonin ↑ haem oxygenase-1 and ↓ inflammation ↓ acute COVID susceptibility, symptoms and severity (hospitalisation and length of stay)
Melatonin receptor agonist agomelatine ↓ IL and CRP levels in the plasma and the brain, and has antioxidant and antiapoptotic properties No studies are available, but agomelatine is approved in major depressive disorder
Miscellaneous
Lithium ↓ inflammation, ↓ CRP and neutrophil/lymphocyte ratio and ↑ T-cell differentiation and proliferation; ↑ axial diffusivity that is ↓ in COVID (thus ↑ white matter integrity) ↓ inflammation and immune response in 6 pts with severe COVID; in follow-up trial of 30 pts, ↓ hospital stay to 6.5 vs 12.4 d for non-recipients [21] ↓ neurological effects 1 mo post-discharge (40 vs 73% of pts) [21]
May have direct anti-viral effect via competition with magnesium
IL-1 and IL-6 receptor antagonists anakinra and tocilizumab
↓ systemic inflammation index
Protects against LCD, with 3.6% of hospitalised recipients vs 16.4% of non-recipients having LCD 1 or 3 mo post-discharge
IL interleukin, CRP C-reactive protein, ED emergency department, FIASMA functional inhibition of acid sphingomyelinase, LCD long COVID depression, pts patients, S1R sigma-1 receptor, SNRI serotonin and noradrenaline reuptake inhibitor, ↓ decrease(s/ed), ↑ increase(s/ed)
Based on imaging, structural changes in the brain associated with LCD are similar to those sometimes seen in patients with MDD [1]. A small study revealed decreased grey and white matter in patients with LCD, 3 months following a COVID infection [16].
SARS-Cov-2 infects and inflames the brain...
SARS-CoV-2 affects the brain via systemic inflammation, appearing to preferentially target neurones [1, 4], and may be a neurotropic virus, acting via direct CNS infection [4]. While even a mild infection may prompt a damaging systemic response [1], severe inflammation and consequent hypercoagulability are most associated with COVID-related cerebrovascular disease [4].
SARS-CoV-2 infection leading to severe inflammation harms the brain in many ways [1, 4, 8]. For example, it increases the permeability of the blood-brain barrier (BBB), allowing cytokines to enter the brain [1]. Although cytokines are involved in learning and memory, abnormally high levels can reduce the plasticity of brain synapses, thereby decreasing learning and memory [17]; these changes likely underlie many aspects of depression [1].
Inflammation may also cause microglia and astrocytes to be abnormally and sometimes persistently activated [1, 4]. These cells normally support brain homeostasis and regulate immune responses [1, 4]. Activated microglia produce cytokines such as ILs and tumour necrosis factor (TNF)-⍺ [1], as well as reactive oxygen and nitroxygen species (ROS and NOS) [1]. Additionally, severe inflammation also causes hyperferritinaemia (elevated blood iron levels), which leads to the production of ROS [4]. Brain cells are especially prone to oxidative stress, and stress biomarkers have been associated with MDD [1].
... with specific depression-related effects
Depression, similar to other psychological disorders, is associated with a dysregulated immune response, which is amplified by reduced serotonin and increased cortisol levels [12]. The production, release and reabsorption of neurotransmitters, including serotonin, dopamine and glutamate, are sensitive to inflammation [1]. The availability of serotonin in particular can decrease due to the abnormal diversion of tryptophan (a precursor) [1], especially in women [13]. The beneficial effects of serotonin against viral inflammation (in addition to their effects on depression) includes the promotion of natural killer cells, reduced production of proinflammatory cytokines and inhibiting the cell entry of other viruses (HIV and dengue virus) [12].
In the short-term, the hypopituitary-adrenal axis tends to dampen the immune response via the production and release of glucocorticoids (e.g. cortisol) as a negative feedback mechanism [1]. However, over the long term, high glucocorticoid levels may result in resistance to this mechanism, thereby promoting inflammation. This may increase cortisol levels, which can precipitate depressive symptoms in vulnerable patients [1].
Treat depression with established options
A published trial in 60 patients with established LCD reported a potential benefit with selective serotonin reuptake inhibitors (SSRIs) [5], which have anti-inflammatory, anti-thrombotic and anti-viral properties [18]. SSRIs significantly improved LCD in 89 and 95% of patients with and without previous psychiatric history, respectively [5]. Furthermore, improvements in depression-related symptoms were reported with the mean Hamilton Depression Rating Scale score decreasing from 23.3 to 6.7 over 4 weeks.
Transcranial direct current stimulation was also helpful in one case study [1] and post-infection vaccination reduced depressive symptoms and other symptoms associated with LC in a large (n > 28,000) observational cohort study in the UK [5]:The first dose of vaccine led to average 13% decrease in the odds of any LC symptoms and a 12% decrease for activity-limiting LC symptoms; however, improvements may or may not have been sustained
The second dose (a median 72 days later) led to a further immediate 9% decrease in the odds of any LC symptoms with a 0.8% decrease for each follow-up week
Additionally, the second dose led to an immediate 9% decrease in the odds of activity-limiting LC symptoms with a 0.5% decrease for each follow-up week
Following the second dose, the symptoms with the largest decrease (9–10%) were fatigue, headache and trouble sleeping
Trials for other options for the management of LCD are urgently needed [1]. Many have been discussed in the literature (e.g. alternative medicine, oxytocin and non-pharmacological support), but no trial data are available [1]. Trials in LC have been initiated, including a trial investigating temelimab (an antibody that targets human endogenous retroviral envelope protein HERV-W-Env) in patients with post-COVID neuropsychiatric symptoms [19].
While referral to psychological or psychiatric services is suggested in some guidelines [3], this may be impractical due to the high demand for these services [1].
Reducing the severity of COVID infections may reduce the risk of long COVID depression
Consistent with findings linking severe COVID infections to LCD, several studies confirm that pre-empting severe illness both before and during infection decreases the incidence of LCD [1, 13]. Agents with anti-inflammatory, anti-thrombotic and/or anti-viral properties are all helpful, with many readily available to outpatients via primary care providers (Tables 1, 2) [1].
The antidepressants in Tables 1 and 2 with COVID-moderating effects, can also reduce stress and anxiety [12] and some have other known or emerging beneficial actions [1, 12]. SSRIs may minimise the severity of COVID infections and/or depression mediated by increased cortisol, while clomipramine has anti-inflammatory properties and can easily cross the BBB [1, 12]. Melatonin in depression can re-establish normal circadian rhythm, which can be dysregulated with depression [1].
Agents with functional inhibition of acid sphingomyelinase (FIASMA) activity have recently been highlighted, as their ability to block the activity of acid sphingomyelinase can reduce viral cell entry and inflammation [15, 20]. Many antidepressants which are used for depression have been identified as posessing FIASMA activity, which may contribute to their efficacy in LCD (Tables 1, 2).
Drugs with other mechanisms of action have been reported to have potential benefits during COVID infections. The serotonin antagonist ondansetron improved outcomes in hospitalised patients, as it counteracts elevated serotonin levels that occur during COVID infections [15]. Casirivimab is a monoclonal antibody against SARS-CoV-2 that blocks viral attachment to ACE-2 receptors to prevent cell entry [4]. Metformin, aside from its benefits for the management of diabetes, has anti-inflammatory, anti-thrombotic and cardioprotective effects [8]; it is readily available and may be useful in both acute and long COVID [8].
Vaccination, as well as the administration of antidepressants or other neurological agents prior to infection may all prevent COVID hospitalisations and reduce LC (Tables 1, 2) [5, 8, 15, 20]. COVID vaccinations lessen disease severity and, in one phone application study, reduced the incidence of LC by 50% [5].
Two recent retrospective analyses of large health databases confirm that many antidepressants were protective against more severe illness (Tables 1, 2) [15, 20]. Anti-depressants with FIASMA activity appear to be particularly effective (Tables 1, 2) [15, 20], in addition to bupropion in one study [15].
Clinical trials a long way from treatments
Current clinical trials are focusing on risk factors and pathophysiology for LCD, fatigue and/or cognitive symptoms, rather than treatment [12]. Vortioxetine, a multimodal antidepressant, is being evaluated for cognitive deficits during or after COVID infection, but no other studies for the treatment of LCD were identified [12]. Potential research topics in this field include the investigation of the link between severe COVID infections and depression [1], the efficacy of phytochemicals, the relevance of FIASMA activity and examining pathways that are implicated in various neurodegenerative conditions, and potentially, depression [4, 8].
Take home messages
LCD is a vaguely defined, but common, condition that is associated with depressive symptoms lasting 1–12 months after a COVID infection
Although the risk factors for LCD are not directly known, the presence of comorbidities, greater disease severity and more frequent mental health visits were associated with LC
Reducing the severity of COVID infections, via pharmacotherapy or prior vaccination may also decrease the risk of LCD
Current data suggest pro-inflammatory cytokines and conditions may play a role in LCD, thus anti-inflammatory agents may be useful
SSRIs typically used for the treatment of depression may also be effective in LCD, owing to their anti-inflammatory, anti-thrombotic and anti-viral properties; though more trials are needed in this therapeutic area
Declarations
Funding
The preparation of this review was not supported by any external funding.
Authorship and conflict of interest
C. Fenton, a contracted writer for Adis International Ltd/Springer Nature, and A. Lee, a salaried employee of Adis International Ltd/Springer Nature, declare no relevant conflicts of interest. All authors contributed to the review and are responsible for the article content.
Ethics approval, Consent to participate, Consent for publication, Availability of data and material, Code availability
Not applicable.
==== Refs
References
1. Mazza MG Palladini M Poletti S Post-COVID-19 depressive symptoms: epidemiology, pathophysiology, and pharmacological treatment CNS Drugs 2022 36 7 681 702 10.1007/s40263-022-00931-3 35727534
2. Soriano JB, Allan M, Alsokhn C, et al. A clinical case definition of post COVID-19 condition by a Delphi consensus. 2021. https://www.who.int/publications/i/item/WHO-2019-nCoV-Post_COVID-19_condition-Clinical_case_definition-2021.1. Accessed 7 Oct 2022.
3. National Institute for Health and Care Excellence (NICE). COVID-19 rapid guideline: managing the longterm effects of COVID-19 2022. https://www.nice.org.uk/guidance/ng188/resources/covid19-rapid-guideline-managing-the-longterm-effects-of-covid19-pdf-51035515742. Accessed 7 Oct 2022.
4. Chandra A Johri A A peek into Pandora's box: COVID-19 and neurodegeneration Brain Sci 2022 12 2 190 10.3390/brainsci12020190 35203953
5. Ayoubkhani D Bermingham C Pouwels KB Trajectory of long COVID symptoms after COVID-19 vaccination: community based cohort study BMJ 2022 377 e069676 10.1136/bmj-2021-069676 35584816
6. Ioannou GN Baraff A Fox A Rates and factors associated with documentation of diagnostic codes for long COVID in the national Veterans Affairs health care system JAMA Netw Open 2022 5 7 e2224359 10.1001/jamanetworkopen.2022.24359 35904783
7. Deng J Zhou F Hou W The prevalence of depression, anxiety, and sleep disturbances in COVID-19 patients: a meta-analysis Ann N Y Acad Sci 2021 1486 1 90 111 10.1111/nyas.14506 33009668
8. Tang SW Leonard BE Helmeste DM Long COVID, neuropsychiatric disorders, psychotropics, present and future Acta Neuropsychiatrica. 2022 34 109 126 10.1017/neu.2022.6 35144718
9. Wu T Jia X Shi H Prevalence of mental health problems during the COVID-19 pandemic: a systematic review and meta-analysis J Affect Disord 2021 15 281 91 98 10.1016/j.jad.2020.11.117
10. Dong F Liu HL Dai N A living systematic review of the psychological problems in people suffering from COVID-19 J Affect Disord 2021 1 292 172 188 10.1016/j.jad.2021.05.060
11. Peluso MJ, Deveau TM, Munter SE, et al. Impact of pre-existing chronic viral infection and reactivation on the development of long COVID. medRxiv [preprint]. 2022. 10.1101/2022.06.21.22276660.
12. Foletto VS da Rosa TF Serafin MB Selective serotonin reuptake inhibitor (SSRI) antidepressants reduce COVID-19 infection: prospects for use Eur J Clin Pharmacol 2022 78 10 1601 1611 10.1007/s00228-022-03372-5 35943535
13. Maes M Al-Rubaye HT Almulla AF Lowered quality of life in long COVID is predicted by affective symptoms, chronic fatigue syndrome, inflammation and neuroimmunotoxic pathways Int J Environ Res Public Health 2022 19 16 10362 10.3390/ijerph191610362 36011997
14. Vindegaard N Petersen LV Lyng-Rasmussen BI Infectious mononucleosis as a risk factor for depression: a nationwide cohort study Brain Behav Immun 2021 94 259 265 10.1016/j.bbi.2021.01.035 33571632
15. Fritz BA Hoertel N Lenze EJ Association between antidepressant use and ED or hospital visits in outpatients with SARS-CoV-2 Transl Psychiatry 2022 12 341 10.1038/s41398-022-02109-3 35995770
16. Benedetti F Palladini M Paolini M Brain correlates of depression, post-traumatic distress, and inflammatory biomarkers in COVID-19 survivors: a multimodal magnetic resonance imaging study Brain Behav Immun Health. 2021 18 100387 10.1016/j.bbih.2021.100387 34746876
17. Bourgognon JM Cavanagh J The role of cytokines in modulating learning and memory and brain plasticity Brain Neurosci Adv. 2020 2020 4 2398212820979802
18. Mazza MG Zanardi R Palladini M Rapid response to selective serotonin reuptake inhibitors in post-COVID depression Eur Neuropsychopharmacol 2022 54 1 6 10.1016/j.euroneuro.2021.09.009 34634679
19. ClinicalTrials.gov. Temelimab as a disease modifying therapy in patients with neuropsychiatric symptoms in post-COVID 19 or PASC syndrome. 2022. https://www.clinicaltrials.gov/ct2/show/NCT05497089. Accessed 23 Nov 2022.
20. Hoertel N Sanchez-Rico M Kornhuber J Antidepressant use and its association with 28-day mortality in inpatients with SARS-CoV-2: support for the FIASMA model against COVID-19 J Clin Med 2022 2022 11 5882 10.3390/jcm11195882
21. Spuch C Lopez-Garcia M Rivera-Baltanas T Efficacy and safety of lithium treatment in SARS-CoV-2 infected patients Front Pharmacol 2022 13 850583 10.3389/fphar.2022.850583 35496309
| 0 | PMC9742030 | NO-CC CODE | 2022-12-14 23:29:59 | no | Drugs Ther Perspect. 2022 Dec 12;:1-6 | utf-8 | Drugs Ther Perspect | 2,022 | 10.1007/s40267-022-00975-x | 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
24433
10.1007/s11356-022-24433-3
Research Article
Associations of exposures to air pollution and greenness with mortality in a newly treated tuberculosis cohort
Wang Xin-Qiang 1
Zhang Kang-Di 1
Yu Wen-Jie 1
Zhao Jia-Wen 1
Huang Kai 2
Hu Cheng-Yang 3
Zhang Xiu-Jun 1
http://orcid.org/0000-0002-0903-6937
Kan Xiao-Hong [email protected]
14
1 grid.186775.a 0000 0000 9490 772X Department of Epidemiology and Biostatistics, School of Public Health, Anhui Medical University, 81 Meishan Road, Hefei, 230032 China
2 grid.452696.a 0000 0004 7533 3408 The Second Affiliated Hospital of Anhui Medical University, 678 Furong Road, Hefei, 230601 China
3 grid.186775.a 0000 0000 9490 772X Department of Humanistic Medicine, School of Humanistic Medicine, Anhui Medical University, 81 Meishan Road, Hefei, 230032 China
4 grid.186775.a 0000 0000 9490 772X Clinical College of Chest, Anhui Chest Hospital, Anhui Medical University, 397 Jixi Road, Hefei, 230022 China
Responsible Editor: Lotfi Aleya
12 12 2022
114
27 6 2022
23 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.
Some previous studies had linked air pollutants and greenness to the risk of death from tuberculosis (TB). Only a few studies had examined the effect of particulate matter (PM2.5) on the mortality of TB, and few studies had assessed the impact and interaction of multiple air pollutants and greenness on the mortality of newly treated TB patients. The study included 29,519 newly treated TB patients from three cities in Anhui province. We collected meteorological data and five pollutants data from The National Meteorological Science Center and air quality monitoring stations. Greenness data were generated by remote sensing inversion of medium-resolution satellite images. We geocoded each patient based on the residential address to calculate the average exposure to air pollutants and the average greenness exposure for each patient during treatment. The Cox proportional risk regression model was used to evaluate the effects of air pollutants and greenness on mortality in newly treated tuberculosis patients. Our results found that the higher the concentration of air pollutants in the living environment of newly treated TB patients, the greater the risk of death: HR 1.135 (95% CI: 1.123–1.147) and HR 1.333 (95% CI: 1.296–1.370) per 10 μg/m3 of PM2.5 and SO2, respectively. Greenness reduced the mortality among newly treated TB patients: HR for NDVI exposure 0.936 (95% CI: 0.925–0.947), HR for NDVI_250m exposure 0.927 (95% CI: 0.916–0.938), and HR for NDVI_500m exposure 0.919 (95% CI: 0.908–0.931). Stratifying the cohort by median greenness exposure, HRs for air pollutants were lower in the high greenness exposure group. Mortality in newly treated TB patients is influenced by air pollutants and greenness. Higher green exposure can mitigate the effects of air pollution. Improving air quality may help reduce mortality among newly treated TB patients.
Supplementary Information
The online version contains supplementary material available at 10.1007/s11356-022-24433-3.
Keywords
Ambient air pollutants
Greenness
Mortality
Newly treated patients with tuberculosis
Cohort study
Multi-city
==== Body
pmcIntroduction
Tuberculosis (TB) is a chronic infectious disease caused by mycobacterium tuberculosis (MTB), which is one of the oldest diseases in the world. Mycobacterium tuberculosis can invade various organs of the human body, but mainly invades the lung. Globally, TB is the 13th leading cause of death and the second leading infectious killer after COVID-19 (ahead of HIV/AIDS). In 2020, 1.5 million people died from TB (including 214,000 people living with HIV). In 2020, an estimated 10 million people worldwide are living with TB, including 5.6 million men, 3.3 million women, and 1.1 million children. TB is present in all countries and age groups. China has the second-highest number of TB patients in the world, accounting for 8.5% of the global total. The World Health Organization (WHO) estimates that an uncured active TB patient can infect 10 to 15 people in a year through close contact (WHO 2021).
Many previous studies have shown that air pollution plays a role in the development of TB (Kim 2014; Rivas-Santiago et al. 2015; Li et al. 2019). A meta-analysis has shown that long-term exposure to biomass smoke and ambient air pollution in adults is associated with an increased risk of respiratory infections such as chronic obstructive pulmonary disease and tuberculosis (Kc et al. 2018). Tobacco smoke consists of a large number of compounds that are also present in air pollutants. Many studies have shown that smoking tobacco can lead to an increased risk of contracting TB, developing active TB, and even worse treatment outcomes such as increased mortality (Lin et al. 2007; Slama et al. 2007; Jee et al. 2009; Horne et al. 2012; Maciel et al. 2013). A series of time-series studies have shown a link between air pollution and TB risk with mixed results. The relationship between six common air pollutants and the risk of tuberculosis outpatient visits was examined in a time-series study in Fuyang, China. The results showed that exposure to fine particulate matter PM2.5 (particulate matter < 2.5 μm in aerodynamic diameter), fine particulate matter PM10 (particulate matter < 10 μm in aerodynamic diameter), ozone (O3), nitrogen dioxide (NO2), and carbon monoxide (CO) increased the risk of outpatient visits, and only sulfur dioxide (SO2) had a certain protective effect (Wang et al. 2022a, b). Another study (Huang et al. 2020) also found that exposure to air pollutants PM2.5, NO2, SO2, and O3 was associated with the risk of TB outpatient visits, and a subgroup analysis showed that seasonal variations may have an even greater impact on the risk. A link between air pollutants and TB has also been found in several multi-city studies. A multi-city study in China found that short-term exposure to SO2 reduced the risk of TB, and that exposure had a greater impact on farmers and workers (Wang et al. 2022a, b). A study about the impact of particulate air pollution on TB in seven major Korean cities found that one standard deviation increase in PM10 exposure over 6 years was associated with a 1.2 times higher TB notification rate (Kim et al. 2020). A Chinese study in Beijing and Hong Kong found a positive correlation between outdoor PM2.5 exposure and seasonal changes in tuberculosis incidence (You et al. 2016). Although many studies have examined the relationship between air pollution and TB, most focused on the incidence of TB and the risk of TB hospital admission. Only a few studies have looked at the effect of ambient air pollution exposure on TB treatment outcomes. A cohort study of tuberculosis in China (Peng et al. 2017) found that long-term exposure to PM2.5 was significantly associated with tuberculosis mortality (aHR = 1.46, 95% CI: 1.15–1.85). Another TB cohort study found that traffic-related air pollution was associated with increased all-cause mortality among active TB patients being treated in California, USA (Blount et al. 2017). The effect of air pollution on TB mortality needs further study.
Environmental green exposure as a potential natural indicator is considered to have three general functions, which are reducing the harm from exposure to noise, high temperature, and air pollution; improving resilience; and promoting healthy activities (Markevych et al. 2017). Several studies have shown an association between exposure to green space and respiratory mortality (James et al. 2016; Crouse et al. 2017; Ji et al. 2019). Green exposure may contribute to the therapeutic rehabilitation of TB patients. A cohort study on the effect of green exposure on all-cause mortality in patients with multidrug-resistant tuberculosis (MDR-TB) treatment in Zhejiang, China, found that green exposure reduced mortality in patients with MDR-TB in areas with lower nighttime light (Ge et al. 2021a, b). A population-based cohort study in Zhejiang and Ningxia, China, finally found that PM2.5 exposure increased mortality among patients with multidrug-resistant tuberculosis, but patients could benefit from green exposure by reducing the effects of PM2.5 (Ge et al. 2021a, b). Although several studies have assessed the effects of greenness exposure, no studies have examined whether green space exposure affects treatment outcomes in newly treated TB patients.
As one of the most populous countries in the world, China is also a country with a high burden of TB, with a rising incidence and mortality of TB. Better treatment and management of newly treated TB patients to reduce secondary transmission and relapse can effectively reduce the TB burden. Based on a cohort of newly treated TB patients in Anhui province, we aimed to investigate the combined effects of air pollutants and greenness on newly treated TB patients.
Materials and methods
Study location
Anhui province is located in eastern China, between 114°54′–119°37′ east longitude and 29°41′–34°38′ north latitude. It covers an area of 140,100 km2. According to the latest census data, Anhui province has 61.3 million permanent residents, ranking among the top 10 in China. Anhui province spans the Yangtze River, Huai River, and Xin’an River, with the Huai River basin covering 67,000 km2, the Yangtze River basin 66,000 km2, and the Xin’an River basin 60,500 km2. The Yangtze River flows through central and southern Anhui with a total length of 416 km, while the Huai River flows through northern Anhui with a total length of 430 km. Anhui is located in the transition region of warm temperate zone, with the Huai River as the dividing line. The north of Anhui is warm temperate semi-humid monsoon climate, and the south is subtropical humid monsoon climate. The main characteristics are mild climate, plenty of sunshine, obvious monsoon, and four distinct seasons. The annual average temperature of the province is 14–16 °C, with a north–south difference of about 2 °C. The annual average sunshine is 1800–2500 h, average frost-free period 200–250 days, and average precipitation 800–1600 mm. This study takes the Huai River and the Yangtze River as the boundary, and selects Huainan, Hefei, and Huangshan as the research areas. Cement industry is one of the most important sources of industrial pollutant discharge in Anhui province. According to estimates, in 2018, SO2 and NO2 emissions from Anhui’s cement industry were 13,000 tons and 140,000 tons, respectively, accounting for 10.7% and 45.9% of the total emissions from major industries in the province, respectively. Nitrogen oxides are precursors to the formation of PM2.5 and ozone and have a direct impact on air quality. At the same time, automobile exhaust is one of the main sources of air pollutants. According to the report of the Anhui Statistical Yearbook 2022, Hefei has 61.13 million permanent residents and 59.39% of the urban population; Huainan has a permanent population of 3.04 million, with 61.91% of the urban population; the permanent population of Huangshan city is 1.332 million, and the urban population accounts for 59.25%. The number of days with air quality reaching or better than grade II is 86% in Hefei, 74.8% in Huainan, and 99.7 in Huangshan city, indicating that the air quality of Huangshan city is better. The number of medical and health institutions in the three cities is as follows: Hefei city 3543, Huainan city 1657, Huangshan city 1108.
Object of the study
The object of the study was a newly treated patient, which was first discovered and did not receive any anti-tuberculosis drug treatment, or after the discovery of irregular, unreasonable treatment, but the course of treatment does not exceed 1 month.
Study cohort
All newly treated TB patients reported from January 1, 2015, to December 31, 2020, in the three cities were eligible for inclusion in this study. These cases were reported by the Centers for Disease Control and Prevention in the three cities, and the data were recorded and collected by the provincial Tuberculosis Prevention and Control Institute. Immigrant patients were excluded from the study because they might return home for treatment and we could not collect information on where they lived and when they were treated. We also excluded patients with incomplete information, such as no residential address and incomplete treatment date records.
Environmental exposure assessment
Assessment of exposure to air pollutants
Our air pollutants’ concentration data came from the China Air Quality Online Monitoring and Analysis platform (https://www.aqistudy.cn/historydata/). We collected the data of the three cities from January 1, 2015, to December 31, 2020, including PM10, SO2, PM2.5, NO2 (μg/m3, 24 h), and O3 (μg/m3, 8 h). We geocoded the detailed address of each patient and assigned them exposure estimates based on data from contaminant sites.
Assessment of exposure to greenness
In this study, the normalized difference vegetation index (NDVI) was used as the evaluation index of green exposure. Geocoding was carried out through address information to obtain the latitude and longitude coordinates. The 2015–2020 NDVI monthly distribution data product of Anhui province is generated and processed by remote sensing inversion on the basis of medium-resolution satellite images, with a spatial resolution of 250 m. We combined cloud-free images from each summer to create an image covering the entire province of Anhui and selected a map of green space density in our study area (Fig. 1). The range of NDVI value is between − 1 and 1, and the negative value indicates that the land cover is water, cloud, etc., while 0 stands for bare earth, rock, etc. A positive value represents healthy vegetation, and a larger value represents a higher greenness (Klompmaker et al. 2019). We set the negative value of NDVI to 0. We calculated the mean monthly NDVI exposure values for patients during treatment and the mean monthly exposure allocated within the 250-m and 500-m buffer zones of residential addresses. The calculation formula of NDVI is as follows:Fig. 1 Characteristics of greenness in the study area from 2015 to 2020. A is 2015, B is 2016, C is 2017, D is 2018, E is 2019, and F is 2020, and (a) is Huainan, (b) is Hefei, and (c) is Huangshan
NDVI=(NIR-R)/(NIR+R)
NIR is the pixel value of the infrared band, and R is the pixel value of the red band (Li et al. 2022).
All pollutant exposure matching, NDVI calculation, and geographical distribution processing were performed on ArcGIS software (version 10.8).
Individual and environmental covariates
We collected three types of potential covariates: basic individual characteristics, patient type and treatment management style, and environmental factors. Individual characteristics of patients include sex, age, and occupation. The patient type includes the source of patients, diagnostic classification of patients, whether patients are severe, and whether patients are detainees. The environmental factors include working environment, region, mean temperature (MT), and relative humidity (RH). Meteorological data (MT and RH) were obtained from the National Center for Meteorological Science (http://data.cma.cn/), and we estimated the mean temperature and mean relative humidity of the patients during treatment using geocoded residential addresses.
Statistical analysis
The number and proportion of all newly treated TB patients were stratified according to individual characteristics, source of patients, environmental factors, and so on. A simple descriptive statistical analysis was performed on environmental exposure variables including air pollutants, greenness exposure, and meteorological data. The correlations among air pollutant individual exposure concentrations and environmental exposure were described by the Spearman correlation coefficient and scatter plot.
We used Cox proportional risk regression models to assess the relationship between persistent exposure to air pollutants and green exposure and mortality. Covariates with p < 0.2 in the univariate Cox model were included in the multivariate Cox proportional risk regression model (Ge et al. 2021a, b). HRs for environmental exposure variables are reported as fixed increments of 10 μg/m3 for air pollutants and fixed increments of per unit for greenness exposure. We conducted three kinds of modeling for the multivariate Cox proportional risk regression model: Model 1, adjusted for the individual characteristics of the patients including sex, age, and occupation; Model 2, Model 1, plus patient type (source of patients, diagnostic classification of patients, whether patients are severe, and whether patients are detainees) and treatment management style; Model 3, Model 2, plus environmental factors including working environment, region, mean temperature, and relative humidity. We evaluate the models’ proportional risk assumptions by visually examining fractional residuals corresponding to the event time. Based on the main model (Model 3), we stratified the entire cohort according to the median greenness exposure to investigate whether the impact of air pollution exposure on the treatment outcome of newly treated TB patients varies with the change of greenness exposure.
Three sensitivity analyses were performed: (1) Stratified analyses of the entire cohort were performed to explore the correction effects of age, sex, and region in Model 3. (2) NO2 exposure was adjusted on the basis of Model 3, and the potential impact between exposures was assessed through the bi-exposure models. (3) Patients enrolled in the cohort after the outbreak of COVID-19 were excluded.
Finally, we used restricted cubic splines (RCS) to determine the exposure–response relationship between environmental exposure and mortality. All analyses were performed through the “survival” package in R software (version 4.0.0).
Results
We followed 29,519 newly treated TB patients between January 1, 2015, and December 31, 2020. There were 17,747 person-years used in the analysis, with a mean follow-up of 219 days, and 369 people died during treatment (1.25%).
Descriptive results
Table 1 and Table 2 provide the results of the simple descriptive statistical analysis of air pollutants, greenness exposures, meteorological indicators, and newly treated TB patients. The average patient exposure concentration of PM2.5 was 48.03 μg/m3 (12.61–103.35 μg/m3), PM10 was 76.85 μg/m3 (25.03–142.22 μg/m3), SO2 was 11.28 μg/m3 (4.05–37.88 μg/m3), O3 was 111.91 mg/m3 (40.43–173.54 μg/m3), and NO2 was 34.85 μg/m3 (7.65–80.75 μg/m3). The average concentrations of PM2.5 and PM10 were between the level 1 and level 2 daily concentration limits for ambient air pollutants in the national environmental quality standards. The average concentration of O3 is between the 8-h average level 1 concentration limit and level 2 concentration limit. The average concentrations of SO2 and NO2 were below the average daily level 1 concentration limit. The minimum and maximum NDVI were 0.011 and 0.836, respectively. The minimum value of NDVI_250m was 0.148, and the maximum was 0.872. The minimum and maximum NDVI_500m were 0.011 and 0.910, respectively. The minimum value of the mean temperature was 3.99 °C, and the maximum was 28.20 °C. The minimum value of the relative humidity was 65.84%, and the maximum was 86.92%.Table 1 Environmental exposures of study population
Variables Mean ± SD Centiles
Min 25% 50% 75% Max IQR
Atmospheric pollutants (μg/m3)
PM2.5 48.03 ± 15.44 12.61 36.39 47.73 58.93 103.35 22.54
PM10 76.85 ± 20.03 25.03 63.39 76.28 90.23 142.22 26.84
SO2 11.28 ± 4.73 4.05 6.98 10.73 14.10 37.88 7.12
O3 111.91 ± 24.49 40.43 95.86 112.21 128.28 173.54 32.42
NO2 34.85 ± 13.35 7.65 23.60 33.70 44.71 80.75 21.11
Greenness exposure
NDVI 0.413 ± 0.131 0.011 0.305 0.398 0.512 0.836 0.207
NDVI_250m 0.416 ± 0.130 0.148 0.310 0.402 0.513 0.872 0.203
NDVI_500m 0.424 ± 0.131 0.011 0.316 0.412 0.523 0.910 0.207
Meteorology measure
Mean temperature (°C) 17.12 ± 4.48 3.99 14.12 16.96 20.57 28.20 6.45
Relative humidity (%) 76.57 ± 3.01 65.84 74.57 76.63 78.41 86.92 3.84
PM2.5, particulate matter < 2.5 μm in aerodynamic diameter; PM10, particulate matter < 10 μm in aerodynamic diameter; SO2, sulfur dioxide; O3, ozone; NO2, nitrogen dioxide; NDVI, normalized difference vegetation index; NDVI_250m, normalized difference vegetation index in the 250-m buffer; NDVI_500m, normalized difference vegetation index in the 500-m buffer; Min, minimum; Max, maximum; IQR, interquartile range
Table 2 Individual baseline characteristics for the entire cohort and stratified by greenbelt exposure
Number (%) NDVI, n (%) NDVI_250m, n (%) NDVI_500m, n (%)
Low exposure a High exposure b Low exposure a High exposure b Low exposure a High exposure b p Value
TB case
Total 29,519 (100) 14,778 (49.9) 14,741 (50.1) 14,769 (50.0) 14,750 (50.0) 14,763 (50.0) 14,756 (50.0)
Stratified by sex 0.062
Male 20,937 (70.9) 10,190 (68.9) 10,747 (72.9) 10,167 (68.8) 10,770 (73.0) 10,184 (69.0) 10,753 (72.9)
Female 8582 (29.1) 4588 (31.1) 3994 (27.1) 4602 (31.2) 3980 (27.0) 4579 (31.0) 4003 (27.1)
Stratified by age 0.150
0–64 years 21,311 (72.2) 11,609 (78.6) 9702 (65.8) 11,628 (78.7) 9683 (65.6) 11,636 (78.8) 9675 (65.6)
≥ 65 years 8280 (27.8) 3169 (21.4) 5037 (34.2) 3141 (21.3) 5067 (34.4) 3127 (21.2) 5081 (34.4)
Stratified by working environment 0.000
Indoor 17,872 (60.5) 6265 (42.3) 11,607 (78.7) 6215 (42.1) 11,657 (79.0) 6171 (41.8) 11,701 (79.3)
Outdoor 11,647 (39.5) 8513 (57.7) 3134 (21.3) 8554 (57.9) 3093 (21.0) 8592 (58.2) 3055 (20.7)
Stratified by occupation 0.000
Intelligent-intensive 5214 (17.7) 3663 (24.7) 1551 (10.5) 3669 (24.8) 1545 (10.5) 3679 (24.9) 1535 (10.4)
Labor-intensive 24,305 (82.3) 11,115 (75.3) 13,910 (89.5) 11,100 (75.2) 13,205 (89.5) 11,084 (75.1) 13,221 (89.6)
Stratified by region 0.000
Urban 9211 (31.2) 8039 (54.3) 1172 (8.0) 8095 (54.8) 1116 (7.6) 8234 (55.8) 977 (6.6)
Rural 20,308 (68.8) 6739 (45.7) 13,569 (92.0) 6674 (45.2) 13,634 (92.4) 6529 (44.2) 13,779 (93.4)
Stratified by city 0.479
Hefei 18,087 (61.3) 10,186 (68.9) 7901 (53.6) 10,296 (69.7) 7791 (52.8) 10,337 (70.0) 7750 (52.5)
Huainan 8288 (28.1) 3880 (26.2) 4408 (29.9) 3882 (26.3) 4406 (29.9) 3853 (26.1) 4435 (30.1)
Huangshan 3144 (10.7) 712 (4.9) 2432 (16.5) 591 (4.0) 2553 (17.3) 573 (3.9) 2571 (17.4)
Detainee c 0.177
Yes 426 (1.4) 221 (1.4) 205 (1.4) 236 (1.6) 190 (1.3) 233 (1.6) 193 (1.3)
No 29,093 (98.6) 14,557 (98.6) 14,536 (98.6) 14,533 (98.4) 14,560 (98.7) 14,530 (98.4) 14,563 (98.7)
Source of TB 0.000
Health check or contact check 305 (1.0) 193 (1.3) 112 (0.8) 189 (1.3) 116 (0.8) 195 (1.3) 110 (0.7)
Clinical consultation 4260 (14.4) 1816 (12.2) 2444 (16.6) 1836 (12.4) 2424 (16.4) 1826 (12.4) 2434 (16.5)
Transfer treatment 24,954 (84.5) 12,769 (96.5) 12,185 (82.6) 12,744 (86.3) 12,210 (82.8) 12,742 (86.3) 12,212 (82.8)
Diagnostic classification 0.000
I 5 (0.0) 2 (0.0) 3 (0.0) 2 (0.0) 3 (0.0) 2 (0.0) 3 (0.0)
II 146 (0.5) 67 (0.5) 79 (0.5) 65 (0.4) 81 (0.5) 67 (0.5) 79 (0.5)
III 27,253 (92.3) 13,692 (92.6) 13,564 (92.1) 13,684 (92.7) 13,569 (92.0) 13,696 (92.8) 13,557 (91.9)
IV 1795 (6.1) 890 (6.0) 905 (6.1) 885 (6.0) 910 (6.2) 877 (5.9) 918 (6.2)
V 320 (1.1) 127 (0.9) 193 (1.3) 133 (0.9) 187 (1.3) 121 (0.8) 199 (1.3)
Sever case 0.007
Yes 540 (1.8) 375 (1.5) 165 (1.1) 372 (2.5) 168 (1.1) 380 (2.6) 160 (1.1)
No 28,979 (98.2) 14,403 (98.5) 14,576 (98.9) 14,397 (97.5) 14,582 (98.9) 14,383 (97.4) 14,596 (98.9)
Treatment management style c 0.004
The entire supervision 20,450 (69.3) 11,457 (77.5) 9000 (61.1) 11,474 (77.7) 8983 (60.9) 11,531 (78.1) 8926 (60.5)
Intensive supervision 8965 (30.4) 3292 (22.3) 5673 (38.5) 3273 (22.2) 5692 (38.6) 3206 (21.7) 5759 (39.0)
Self-administrative treatment 97 (0.3) 29 (0.2) 68 (0.4) 22 (0.1) 75 (0.5) 26 (0.2) 71 (0.5)
aLow exposure is defined as from the minimum to the median
bHigh exposure is defined as from the median to maximum
cTreatment management style: In order to ensure that patients can adhere to regular medication and complete the prescribed course of treatment, effective management measures must be taken to ensure that patients can complete the course of treatment
NDVI, normalized difference vegetation index; NDVI_250m, normalized difference vegetation index in the 250-m buffer; NDVI_500m, normalized difference vegetation index in the 500-m buffer
A total of 29,519 newly treated TB patients were included in this cohort, including 20,937 males and 8582 females. 72.2% of patients were younger than 65 years old. 60.5% of patients worked indoors. 24,305 patients worked in labor-intensive jobs. A total of 9211 patients lived in urban areas and 20,308 in rural areas. Hefei had the highest number of patients, while Huangshan had the least. A minority (1.4%) of patients were detainees. 84.5% of patients were on transfer treatment. 92.3% of patients were III TB patients. A total of 540 patients were seriously ill with TB. Among all patients, 20,450 patients received entire supervision, 8965 patients received intensive supervision, and 97 patients were self-medicated. The results of the entire cohort stratified by green exposure are shown in Table 2. As shown in Fig. 2, we found significant correlations between the two particulate pollutants (|r|= 0.90) and between the three greenness exposure indicators.Fig. 2 The Spearman rank correlation coefficients and scatter plot between air pollutant individual exposure concentrations and environmental exposure. PM2.5, particulate matter < 2.5 μm in aerodynamic diameter; PM10, particulate matter < 10 μm in aerodynamic diameter; SO2, sulfur dioxide; O3, ozone; NO2, nitrogen dioxide; NDVI, normalized difference vegetation index; NDVI_250m, normalized difference vegetation index in the 250-m buffer; NDVI_500m, normalized difference vegetation index in the 500-m buffer; MT, mean temperature; RH, relative humidity. *rs > 0.7
Effects of air pollutants and greenness on mortality in newly treated TB patients
We used Cox proportional risk regression models and fitted three models to assess the impact of air pollutants and greenness exposure on mortality (see the “Materials and methods” section for details). Overall, an increase of 10 μg/m3 in the exposure concentration of various air pollutants had a positive and statistically significant effect on mortality. The results from Model 3 found that exposure to PM2.5 was associated with increased mortality (HR = 1.135, 95% CI: 1.123–1.147). The HR for PM10 exposure is 1.182 (95% CI: 1.173–1.147), the HR for O3 exposure is 1.034 (95% CI: 1.028–1.040), and the HR for SO2 exposure is 1.333 (95% CI: 1.296–1.370). We observed that greenness exposure reduced mortality. The results from Model 3 found that exposure to NDVI associated with a reduction in mortality (HR = 0.936, 95% CI: 0.925–0.947). The HR for NDVI_250m exposure is 0.927 (95% CI: 0.916–0.938), and the HR for NDVI_500m exposure is 0.919 (95% CI: 0.908–0.931). The results of each model are shown in Table 3.Table 3 Hazard ratios and 95% CI of per 10 µg/m3 increase in air pollutant exposure and per 0.1 µg/m3 increase in greenness exposure associated with mortality among newly treated patients with tuberculosis in full cohort
Environmental exposure Model 1 Model 2 Model 3
HR LCI UCI HR LCI UCI HR LCI UCI
PM2.5 1.138 1.129 1.147 1.149 1.140 1.158 1.135 1.123 1.147
PM10 1.165 1.158 1.172 1.175 1.167 1.182 1.182 1.173 1.192
O3 1.043 1.037 1.048 1.050 1.045 1.056 1.034 1.028 1.040
SO2 1.551 1.512 1.591 1.517 1.479 1.556 1.333 1.296 1.370
NDVI 0.941 0.932 0.950 0.921 0.912 0.930 0.936 0.925 0.947
NDVI_250m 0.935 0.926 0.943 0.913 0.904 0.922 0.927 0.916 0.938
NDVI_500m 0.932 0.923 0.940 0.909 0.900 0.918 0.919 0.908 0.931
Model 1, adjusted for the individual characteristics of the patients including sex, age, and occupation; Model 2, Model 1, plus patient type (source of patients, diagnostic classification of patients, whether patients are severe, and whether patients are detainees) and treatment management style; Model 3, Model 2, plus environmental factors including working environment, region, mean temperature, and relative humidity
PM2.5, particulate matter < 2.5 μm in aerodynamic diameter; PM10, particulate matter < 10 μm in aerodynamic diameter; SO2, sulfur dioxide; O3, ozone; NDVI, normalized difference vegetation index; NDVI_250m, normalized difference vegetation index in the 250-m buffer; NDVI_500m, normalized difference vegetation index in the 500-m buffer; HR, hazard ratios; UCI, upper confidence interval; LCI, lower confidence interval
Effects of air pollution on mortality in newly treated TB patients stratified by greenness exposure across the full cohort
We stratified the entire cohort according to greenness exposure and used Model 3 to assess whether the impact of exposure to various air pollutants on mortality was affected by different greenness. From Fig. 3 and Supplementary Table S1, we can see that the impact of air pollutants on mortality was low during high green exposure. For example, after stratification of the cohort by median NDVI, the HRs of PM2.5 exposure on mortality were 1.139 (95% CI: 1.122–1.155) and 1.121 (95% CI: 1.103–1.140) in the low exposure group and the high exposure group, respectively. After stratification of the cohort by median NDVI_250m, the HRs of SO2 exposure on mortality were 1.352 (95% CI: 1.300–1.406) and 1.308 (95% CI: 1.256–1.362) in the low exposure group and the high exposure group, respectively.Fig. 3 Hazard ratios and 95% CI of per 10 µg/m3 increase in air pollutant exposure associated with mortality among newly treated patients with tuberculosis in the full cohort after being stratified by greenness exposure. Cox models are adjusted for covariates: sex, age, occupation, source of patients, diagnostic classification of patients, whether patients are severe, whether patients are detainees, treatment management style, working environment, region, mean temperature, and relative humidity. PM2.5, particulate matter < 2.5 μm in aerodynamic diameter; PM10, particulate matter < 10 μm in aerodynamic diameter; SO2, sulfur dioxide; O3, ozone; NDVI, normalized difference vegetation index; NDVI_250m, normalized difference vegetation index in the 250-m buffer; NDVI_500m, normalized difference vegetation index in the 500-m buffer. aLow exposure is defined as from the minimum to the median. bHigh exposure is defined as from the median to maximum
Sensitivity analysis
Effects of air pollutants and greenness on mortality in newly treated TB patients in different sex, age, and region groups
We stratified the entire cohort by age, sex, and region, and then analyzed the independent effects of air pollutant exposure and greenness exposure on mortality in different subgroups. We found that exposure to air pollutants was more harmful for patients living in urban areas, elderly patients (age ≥ 65 years old), and men, with the exception of SO2. Living in areas with higher greenness exposure may be more protective for male patients, younger patients (age < 65 years old), and patients living in urban areas (Fig. 4; Supplementary Table S2).Fig. 4 Hazard ratios and 95% CI of per 10 µg/m3 increase in air pollutant exposure and per 0.1 µg/m3 increase in greenness exposure associated with mortality among newly treated patients with tuberculosis stratified by sex, age, and region. Cox models are adjusted for covariates: sex, age, occupation, source of patients, diagnostic classification of patients, whether patients are severe, whether patients are detainees, treatment management style, working environment, region, mean temperature, and relative humidity. PM2.5, particulate matter < 2.5 μm in aerodynamic diameter; PM10, particulate matter < 10 μm in aerodynamic diameter; SO2, sulfur dioxide; O3, ozone; NDVI, normalized difference vegetation index; NDVI_250m, normalized difference vegetation index in the 250-m buffer; NDVI_500m, normalized difference vegetation index in the 500-m buffer
Bi-exposure models
Based on Model 3, NO2 exposure was added to fit the bi-exposure model. We found that there was no significant change in the results before and after NO2 inclusion (Table 3; Supplementary Fig. S1 and Supplementary Table S3).
Excluding patients after the outbreak of COVID-19
In Supplementary Table S4, we found that the results of the three models did not change substantially after removing the patients from the cohort after the outbreak of COVID-19, indicating that the outbreak of COVID-19 had no significant impact on the results of the study.
Exposure–response curves
As shown in Supplementary Fig. S2 and Supplementary Fig. S3, the exposure–response curves suggested a non-linear association between environmental exposure (air pollutants and greenness) and mortality in newly treated TB patients. Greenness exposure was negatively associated with the risk of death in newly treated TB patients, while air pollutants were positively associated with the risk of death.
Discussion
To our knowledge, this is the first study on the relationship between environmental exposure (air pollutants and greenness) and mortality in newly treated TB patients in China. Our study found that air pollutant exposure and greenness exposure were associated with mortality during treatment in newly treated TB patients. Exposure to air pollutants increased mortality in newly treated TB patients, while greenness exposure was associated with reduced mortality in newly treated TB patients. Higher levels of greenness exposure may reduce the adverse effects of air pollutants on TB patients. Our results suggest that overall environmental exposure has a greater impact on male patients and urban patients.
After adjusting for confounding factors such as general individual characteristics, patient types, average temperature, and average relative humidity, we found that all four pollutants had a positive impact on mortality in newly treated TB patients. The results showed that the mortality increased by 13.5% for every 10 μg/m3 increase in PM2.5 exposure, 18.2% for every 10 μg/m3 increase in PM10 exposure, 3.4% for every 10 μg/m3 increase in O3 exposure, and 33.3% for every 10 μg/m3 increase in SO2 exposure.
Some previous studies had also shown a positive impact of air pollutants on TB mortality. A multi-city time-series study in Shandong, China, found that short-term exposure (≤ 30 days) to SO2, CO, and PM2.5 had a significant impact on TB mortality (Liu et al. 2021). A cohort study of drug-resistant TB in China found that every 10 μg/m3 increase in PM2.5 was associated with a 70.2% increase in mortality among drug-resistant patients (Ge et al. 2021a, b). In another tuberculosis cohort study in China, it was found that with each 2.06 μg/m3 increase in PM2.5 exposure, the total mortality of patients increased by 30%, and the tuberculosis mortality increased by 46%. Long-term exposure to PM2.5 had a positive effect on the mortality of patients with tuberculosis (Peng et al. 2017). Previous studies had explored how PM2.5 damages human systems and organs and accelerates the progression of TB (Nel 2005; D’Amato et al. 2010; Ibironke et al. 2019; Popovic et al. 2019). PM2.5 can damage the immune function of the human respiratory system, because PM2.5 can reduce the function of macrophages, cause stress response and oxidation, and reduce the immunity of the lungs. PM2.5 can also damage airway epithelial cells and reduce the body’s immunity to mycobacterium tuberculosis. PM2.5 can also carry some external impurities into the lungs, such as the alveoli and bronchi, which can damage the body’s immune function (Bauer et al. 2012; Laumbach and Kipen 2012; Bai et al. 2018; Cheng et al. 2019; Torres et al. 2019). Some studies suggest that PM10 has a similar damaging effect to PM2.5 (Kim et al. 2020; Pompilio and Di Bonaventura 2020; Xiang et al. 2021). As for the harmful effect of SO2, some studies have shown that exposure to SO2 can reduce and hinder the production and release of tumor necrosis factor-α (TNF-α) in vivo. As TNF-α can inhibit mycobacterium tuberculosis in vivo by controlling the formation of granuloma, SO2 can hinder the resistance of the human body to tuberculosis branch rods (Knorst et al. 1996; Mohan et al. 2001). A laboratory study had shown that O3 can damage lung function, affect gas exchange between the lungs and the outside world, and increase airway inflammation (Smith et al. 2016).
This study found that greenness exposure reduced mortality in newly treated TB patients. We found that exposure to NDVI was associated with a reduction in mortality (HR = 0.936), and we calculated greenness exposures for the 250-m and 500-m buffers, the results still show the protective effect. The protective effect of greenness exposure on tuberculosis was consistent with previous studies. A cohort study in China (Ge et al. 2021a, b) found that greenness exposure within a 500-m buffer was stratified by quartile, with patients in the remaining quartile having a lower risk of death than those in the lowest quartile and HR values of less than 1. In the past, people were treated and prevented by breathing fresh air, and some patients were even sent to the mountains for fresh air. One study found that this method of breathing fresh air could reduce the risk of death from tuberculosis (McCarthy 2001). Spending more time in nature is good for human health, and greenness may help TB patients recover better, as living in greener areas may expose them to more fresh air. In addition, studies have shown that higher levels of greenness may be beneficial to mental health. A study conducted in Australia found that people who thought their neighborhoods were greener reported better mental health (Sugiyama et al. 2008). Some studies suggested that greenness can reduce stress and air pollution, which is good for human health. Green can absorb more particle pollutants in the air, and the surface of plants can also absorb some gaseous pollutants, effectively reducing the content of pollutants in the air (Chen et al. 2016; Markevych et al. 2017; Ji et al. 2020). Our study stratified the full cohort according to the median greenness exposure and found that greenness exposure significantly reduced the harmful effects of air pollution, which was consistent with previous studies. A cohort study of MDR-TB in China found that greater greenness exposure resulted in an average 0.188–0.194 reduction in HR associated with PM2.5.
Our data suggested that exposure to greenness and air pollutants had a greater impact on male patients and patients living in urban areas, a conclusion supported by previous studies. A Canadian cohort study found that greenness exposure was more protective against non-accidental death in men than in women (Crouse et al. 2017). A Chinese study also found a greater protective effect of greenness exposure for younger patients (< 60 years old) and patients living in urban areas (Ge et al. 2021a, b). A cohort study in China found that older people living in urban areas tended to be more likely to benefit from a greener environment (Ji et al. 2019). Compared with rural areas, pollutant concentrations were much higher in urban areas, which may be the reason why air pollutants had a greater impact on urban patients. The greater impact of environmental exposure on men may be due to their greater exposure to the outdoors than women and the fact that more men have unhealthy habits such as smoking and drinking. In addition, physiological differences in the airways between men and women may also play a role (Sopori et al. 1998). Our study also found that pollutant exposure had a greater impact on the elderly, which may be due to their decreased body defense ability and lower ability to resist the harm of pollutant exposure than the young (Zhang et al. 2019; Huang et al. 2020).
Our study had some strengths: First, our study was one of the few to evaluate the effects of both air pollutant exposure and greenness exposure on TB, and we were the first cohort study of newly treated TB patients Secondly, our study included all the newly treated TB patients in the three cities from 2015 to 2020, and the three cities were distributed in central Anhui, northern Anhui, and southern Anhui, respectively, with good regional heterogeneity, which made our study to have good representativeness. Finally, our study explored the effects of multiple pollutants. Compared with previous studies, our contents were more abundant and the bi-exposure model was fitted to make our research results more accurate. But our study still had some limitations. First, we used contaminant data from monitoring sites to match patients, which may affect the authenticity of exposure. Due to the very similar local meteorological conditions and geographical location of each city, the heterogeneity of individual exposure was very low, so there was little difference between the underestimated or overestimated value and the real exposure value. Second, we used NDVI and two buffer zones to evaluate greenness exposure, missing the information of vegetation type, so we could only do general greenness exposure, which could not reflect the type and quality of greenness exposure. Finally, smoking status, socioeconomic status, and exposure to indoor pollutants were not included in the study, although we adjusted for multiple covariates. Future studies should consider incorporating these confounding factors.
Conclusions
In this study, air pollutants were positively correlated with mortality in newly treated TB patients, while greenness exposure was negatively correlated with mortality in newly treated TB patients. The association between air pollutants and mortality in newly treated TB patients decreased with increased exposure to greenness space. Overall, air pollutants and greenness had a greater impact on male patients and urban patients. Greenness was more protective for younger patients, while air pollutants were more harmful for older people. Increasing green vegetation and reducing emissions of air pollutants may help reduce mortality among newly treated TB patients.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 128 kb)
Supplementary file2 (DOCX 144 kb)
Supplementary file3 (DOCX 182 kb)
Supplementary file4 (DOCX 18 kb)
Supplementary file5 (DOCX 19 kb)
Supplementary file6 (DOCX 16 kb)
Supplementary file7 (DOCX 18 kb)
Acknowledgements
We would like to acknowledge the School of Public Health, Anhui Medical University, and Anhui Chest Hospital for their support to this project.
Author contribution
All authors contributed to the study conception and design. Xin-Qiang Wang and Xiao-Hong Kan collaboratively designed the study, both making substantial intellectual contribution. Xin-Qiang Wang, Kang-Di Zhang, and Wen-Jie Yu analyzed the data and drafted the manuscript. Xiao-Hong Kan and Xiu-Jun Zhang revised the manuscript. Jia-Wen Zhao, Cheng-Yang Hu, and Kai Huang contributed in collecting the data. Xin-Qiang Wang, Kang-Di Zhang, and Wen-Jie Yu contributed equally. All authors read and approved the final manuscript.
Funding
This study was supported by the National Natural Science Foundation of China (82073565), the National Key Project for Infectious Disease (2018ZX10722301-001–004), and the Major National Science and Technology Projects during the 12th Five-Year Plan period (2013ZX10003008-001–003).
Data availability
The datasets analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval
This study was approved by the Anhui Medical University Ethics Committee. All patient information included in the study was unidentified and anonymous.
Consent to participate
Not applicable.
Consent for publication
Not applicable.
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.
Xin-Qiang Wang, Kang-Di Zhang, and Wen-Jie Yu contributed equally to this article.
==== Refs
References
Bai L Su X Zhao D Zhang Y Cheng Q Zhang H Wang S Xie M Su H Exposure to traffic-related air pollution and acute bronchitis in children: season and age as modifiers J Epidemiol Community Health 2018 72 5 426 433 10.1136/jech-2017-209948 29440305
Bauer RN, Diaz-Sanchez D, Jaspers I (2012) Effects of air pollutants on innate immunity: the role of Toll-like receptors and nucleotide-binding oligomerization domain-like receptors. J Allergy Clin Immunol. 129 (1), 14–24, 25–26. 10.1016/j.jaci.2011.11.004
Blount RJ Pascopella L Catanzaro DG Barry PM English PB Segal MR Flood J Meltzer D Jones B Balmes J Nahid P Traffic-related air pollution and all-cause mortality during tuberculosis treatment in California Environ Health Persp 2017 125 9 97026 10.1289/ehp1699
Chen L Liu C Zou R Yang M Zhang Z Experimental examination of effectiveness of vegetation as bio-filter of particulate matters in the urban environment Environ Pollut 2016 208 Pt A 198 208 10.1016/j.envpol.2015.09.006 26385643
Cheng H, Zhu F, Lei R, Shen C, Liu J, Yang M, Ding R, Cao J (2019) Associations of ambient PM2.5 and O3 with cardiovascular mortality: a time-series study in Hefei, China. Int J Biometeorol 63 (10), 1437–1447. 10.1007/s00484-019-01766-2.
Crouse DL Pinault L Balram A Hystad P Peters PA Chen H van Donkelaar A Martin RV Menard R Robichaud A Villeneuve PJ Urban greenness and mortality in Canada’s largest cities: a national cohort study Lancet Planet Health 2017 1 7 e289 e297 10.1016/s2542-5196(17)30118-3 29851627
D’Amato G, Cecchi L, D’Amato M, Liccardi G (2010) Urban air pollution and climate change as environmental risk factors of respiratory allergy: an update. J Investig Allergol Clin Immunol. 20 (2), 95–102, 102
Ge E Gao J Ren Z Liu X Luo M Zhong J Fei F Chen B Wang X Wei X Peng Y Greenness exposure and all-cause mortality during multi-drug resistant tuberculosis treatment: a population-based cohort study Sci Total Environ 2021 771 145422 10.1016/j.scitotenv.2021.145422 33548711
Ge E, Gao J, Wei X, Ren Z, Wei J, Liu X, Wang X, Zhong J, Lu J, Tian X, Fei F, Chen B, Wang X, Peng Y, Luo M, Lei J (2021b) Effect modification of greenness on PM2.5 associated all-cause mortality in a multidrug-resistant tuberculosis cohort. Thorax. 2020–216819. 10.1136/thoraxjnl-2020-216819.
Horne DJ, Campo M, Ortiz JR, Oren E, Arentz M, Crothers K, Narita M (2012) Association between smoking and latent tuberculosis in the U.S. population: an analysis of the National Health and Nutrition Examination Survey. Plos One. 7 (11), e49050. 10.1371/journal.pone.0049050.
Huang K Ding K Yang XJ Hu CY Jiang W Hua XG Liu J Cao JY Zhang T Kan XH Zhang XJ Association between short-term exposure to ambient air pollutants and the risk of tuberculosis outpatient visits: a time-series study in Hefei China Environ Res 2020 184 109343 10.1016/j.envres.2020.109343 32192989
Ibironke O, Carranza C, Sarkar S, Torres M, Choi H T, Nwoko J, Black K, Quintana-Belmares R, Osornio-Vargas A, Ohman-Strickland P, Schwander S (2019) Urban air pollution particulates suppress human T-cell responses to mycobacterium tuberculosis. Int J Environ Res Public Health. 16 (21). 10.3390/ijerph16214112.
James P Hart JE Banay RF Laden F Exposure to greenness and mortality in a nationwide prospective cohort study of women Environ Health Perspect 2016 124 9 1344 1352 10.1289/ehp.1510363 27074702
Jee SH Golub JE Jo J Park IS Ohrr H Samet JM Smoking and risk of tuberculosis incidence, mortality, and recurrence in South Korean men and women Am J Epidemiol 2009 170 12 1478 1485 10.1093/aje/kwp308 19917554
Ji JS Zhu A Bai C Wu CD Yan L Tang S Zeng Y James P Residential greenness and mortality in oldest-old women and men in China: a longitudinal cohort study Lancet Planet Health 2019 3 1 e17 e25 10.1016/s2542-5196(18)30264-x 30654864
Ji JS Zhu A Lv Y Shi X Interaction between residential greenness and air pollution mortality: analysis of the Chinese Longitudinal Healthy Longevity Survey Lancet Planet Health 2020 4 3 e107 e115 10.1016/s2542-5196(20)30027-9. 32220672
Kc R Shukla SD Gautam SS Hansbro PM O’Toole RF The role of environmental exposure to non-cigarette smoke in lung disease Clin Transl Med 2018 7 1 39 10.1186/s40169-018-0217-2 30515602
Kim H Yu S Choi H Effects of particulate air pollution on tuberculosis development in seven major cities of Korea from 2010 to 2016: methodological considerations involving long-term exposure and time lag Epidemiol Health 2020 42 e2020012 10.4178/epih.e2020012 32164052
Kim J Is ambient air pollution another risk factor of tuberculosis? Korean J Intern Med 2014 29 2 170 172 10.3904/kjim.2014.29.2.170 24648798
Klompmaker JO Janssen N Bloemsma LD Gehring U Wijga AH van den Brink C Lebret E Brunekreef B Hoek G Associations of combined exposures to surrounding green, air pollution, and road traffic noise with cardiometabolic diseases Environ Health Perspect 2019 127 8 87003 10.1289/ehp3857 31393793
Knorst MM Kienast K Muller-Quernheim J Ferlinz R Effect of sulfur dioxide on cytokine production of human alveolar macrophages in vitro Arch Environ Health 1996 51 2 150 156 10.1080/00039896.1996.9936009 8638967
Laumbach RJ, Kipen HM (2012) Respiratory health effects of air pollution: update on biomass smoke and traffic pollution. J Allergy Clin Immunol. 129 (1), 3–11, 12–13. 10.1016/j.jaci.2011.11.021
Li G Zhang H Hu M He J Yang W Zhao H Zhu Z Zhu J Huang F Associations of combined exposures to ambient temperature, air pollution, and green space with hypertension in rural areas of Anhui province, China: a cross-sectional study Environ Res 2022 204 Pt D 112370 10.1016/j.envres.2021.112370 34780789
Li Z Mao X Liu Q Song H Ji Y Xu D Qiu B Tian D Wang J Long-term effect of exposure to ambient air pollution on the risk of active tuberculosis Int J Infect Dis 2019 87 177 184 10.1016/j.ijid.2019.07.027 31374344
Lin HH Ezzati M Murray M Tobacco smoke, indoor air pollution and tuberculosis: a systematic review and meta-analysis Plos Med 2007 4 1 e20 10.1371/journal.pmed.0040020 17227135
Liu Y Zhao S Li Y Song W Yu C Gao L Ran J He D Li H Effect of ambient air pollution on tuberculosis risks and mortality in Shandong, China: a multi-city modeling study of the short- and long-term effects of pollutants Environ Sci Pollut R 2021 28 22 27757 27768 10.1007/s11356-021-12621-6
Maciel EL Brioschi AP Peres RL Guidoni LM Ribeiro FK Hadad DJ Vinhas SA Zandonade E Palaci M Dietze R Johnson JL Smoking and 2-month culture conversion during anti-tuberculosis treatment Int J Tuberc Lung Dis 2013 17 2 225 228 10.5588/ijtld.12.0426 23317958
Markevych I Schoierer J Hartig T Chudnovsky A Hystad P Dzhambov AM de Vries S Triguero-Mas M Brauer M Nieuwenhuijsen MJ Lupp G Richardson EA Astell-Burt T Dimitrova D Feng X Sadeh M Standl M Heinrich J Fuertes E Exploring pathways linking greenspace to health: theoretical and methodological guidance Environ Res 2017 158 301 317 10.1016/j.envres.2017.06.028 28672128
McCarthy OR The key to the sanatoria J R Soc Med 2001 94 8 413 417 10.1177/014107680109400813 11461990
Mohan VP Scanga CA Yu K Scott HM Tanaka KE Tsang E Tsai MM Flynn JL Chan J Effects of tumor necrosis factor alpha on host immune response in chronic persistent tuberculosis: possible role for limiting pathology Infect Immun 2001 69 3 1847 1855 10.1128/iai.69.3.1847-1855.2001 11179363
Nel A (2005) Atmosphere. Air pollution-related illness: effects of particles. Science 308 (5723), 804–806. 10.1126/science.1108752
Peng Z Liu C Xu B Kan H Wang W Long-term exposure to ambient air pollution and mortality in a Chinese tuberculosis cohort Sci Total Environ 2017 580 1483 1488 10.1016/j.scitotenv.2016.12.128 28038878
Pompilio A Di Bonaventura G Ambient air pollution and respiratory bacterial infections, a troubling association: epidemiology, underlying mechanisms, and future challenges Crit Rev Microbiol 2020 46 5 600 630 10.1080/1040841x.2020.1816894 33059504
Popovic I Soares MR Ge E Marks GB Dong GH Wei X Knibbs LD A systematic literature review and critical appraisal of epidemiological studies on outdoor air pollution and tuberculosis outcomes Environ Res 2019 170 33 45 10.1016/j.envres.2018.12.011 30557690
Rivas-Santiago CE Sarkar S Cantarella PT Osornio-Vargas A Quintana-Belmares R Meng Q Kirn TJ Ohman SP Chow JC Watson JG Torres M Schwander S Air pollution particulate matter alters antimycobacterial respiratory epithelium innate immunity Infect Immun 2015 83 6 2507 2517 10.1128/iai.03018-14 25847963
Slama K Chiang CY Enarson DA Hassmiller K Fanning A Gupta P Ray C Tobacco and tuberculosis: a qualitative systematic review and meta-analysis Int J Tuberc Lung Dis 2007 11 10 1049 1061 17945060
Smith GS Van Den Eeden SK Garcia C Shan J Baxter R Herring AH Richardson DB Van Rie A Emch M Gammon MD Air pollution and pulmonary tuberculosis: a nested case-control study among members of a Northern California health plan Environ Health Perspect 2016 124 6 761 768 10.1289/ehp.1408166 26859438
Sopori ML Kozak W Savage SM Geng Y Kluger MJ Nicotine-induced modulation of T Cell function. Implications for inflammation and infection Adv Exp Med Biol 1998 437 279 289 10.1007/978-1-4615-5347-2_31 9666281
Sugiyama T Leslie E Giles-Corti B Owen N Associations of neighbourhood greenness with physical and mental health: do walking, social coherence and local social interaction explain the relationships? J Epidemiol Community Health 2008 62 5 e9 10.1136/jech.2007.064287 18431834
Torres M Carranza C Sarkar S Gonzalez Y Osornio VA Black K Meng Q Quintana-Belmares R Hernandez M Angeles GJ Paramo-Figueroa VH Iniguez-Garcia MA Flores JL Zhang JJ Gardner CR Ohman-Strickland P Schwander S Urban airborne particle exposure impairs human lung and blood mycobacterium tuberculosis immunity Thorax 2019 74 7 675 683 10.1136/thoraxjnl-2018-212529 31036772
Wang XQ Li YQ Hu CY Huang K Ding K Yang XJ Cheng X Zhang KD Yu WJ Wang J Zhang YZ Ding ZT Zhang XJ Kan XH Short-term effect of ambient air pollutant change on the risk of tuberculosis outpatient visits: a time-series study in Fuyang China Environ Sci Pollut Res Int 2022 29 20 30656 30672 10.1007/s11356-021-17323-7 34993790
Wang XQ, Zhao JW, Zhang KD, Yu WJ, Wang J, Li Y Q, Cheng X, Li ZH, Mao YC, Hu CY, Huang K, Ding K, Yang XJ, Chen SS, Zhang XJ, Kan XH (2022b) Short-term effect of sulfur dioxide (SO2) change on the risk of tuberculosis outpatient visits in 16 cities of Anhui Province, China: the first multi-city study to explore differences in occupational patients. Environ Sci Pollut Res Int. 10.1007/s11356-022-19438-x
Xiang K Xu Z Hu Y He Y Dan Y Wu Q Fang X Pan H Association between ambient air pollution and tuberculosis risk: a systematic review and meta-analysis Chemosphere 2021 277 130342 10.1016/j.chemosphere.2021.130342 33794431
You S Tong YW Neoh KG Dai Y Wang CH On the association between outdoor PM2.5 concentration and the seasonality of tuberculosis for Beijing and Hong Kong Environ Pollut 2016 218 1170 1179 10.1016/j.envpol.2016.08.071 27595179
Zhang Y Ni H Bai L Cheng Q Zhang H Wang S Xie M Zhao D Su H The short-term association between air pollution and childhood asthma hospital admissions in urban areas of Hefei city in China: a time-series study Environ Res 2019 169 510 516 10.1016/j.envres.2018.11.043 30544078
| 36504301 | PMC9742034 | NO-CC CODE | 2022-12-14 23:29:59 | no | Environ Sci Pollut Res Int. 2022 Dec 12;:1-14 | utf-8 | Environ Sci Pollut Res Int | 2,022 | 10.1007/s11356-022-24433-3 | oa_other |
==== Front
Educ Inf Technol (Dordr)
Educ Inf Technol (Dordr)
Education and Information Technologies
1360-2357
1573-7608
Springer US New York
11499
10.1007/s10639-022-11499-2
Article
Longitudinal study of teacher acceptance of mobile virtual labs
http://orcid.org/0000-0003-2035-3439
Kolil Vysakh Kani [email protected]
1
Achuthan Krishnashree [email protected]
12
1 grid.411370.0 0000 0000 9081 2061 Center for Cybersecurity Systems and Networks, Amrita Vishwa Vidyapeetham, Amritapuri, Kollam, 690525 Kerala India
2 grid.411370.0 0000 0000 9081 2061 Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Amritapuri, Kollam, 690525 Kerala India
12 12 2022
134
10 7 2022
28 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.
Synthesizing the advancements in technology with classroom practices depends considerably on teachers acceptance of such internet and communication technology (ICT) tools. Adequate teacher training and upgrading of their IT skills are not prioritized in developing economies leading to poor adoption of emerging technology assisted pedagogic interventions. This paper investigated the underlying characteristics of teachers acceptance of mobile friendly virtual laboratories (M-VLs) as part of a longitudinal study conducted over 5 years covering both pre-pandemic and pandemic periods. Systematic analysis of quantitative data from 650 chemistry teachers was carried out. Viewing through the theoretical lens of Unified Theory of Acceptance and Use of Technology (UTAUT2) theory, the effects of performance expectancy (PE), effort expectancy (EE), social influence (SI), facilitating conditions (FC), hedonic motivation (HM) and habit (HA) on the behavioral intention (BI) and use behavior (UB) were scrutinized. Structural Equation Modeling (SEM) analysis revealed that PE, SI, and HA are the considerable predictors of the BI to use M-VLs and HA is the predictor of UB. The present study found HM influencing teacher’s BI and UB before COVID-19. However during COVID-19 the FC influenced usage. Moreover, we found that the technology training focused on enhancing knowledge, skill and, access leads to teachers’ are critical to empowering teachers and causing wider adoption.
Keywords
Virtual labs
UTAUT2 model
Mobile technology adoption
User intention
https://doi.org/10.13039/501100004541 Ministry of Education, India
==== Body
pmcIntroduction
Teachers in science education play a monumental role not only in imparting of scientific skills amongst students but also invoking inquisitiveness, engagement and perseverance required for excellence (Fauth et al., 2019; Ben Ouahi et al., 2022). Training students to enhance their grasping of fundamental concepts and application of concepts to real-life scenarios helps with retention of knowledge in them. Approaches to teaching vary and depend directly on teacher beliefs. While some advocate teacher-centered approach (Alger, 2009), others are grounded on using learner-centered methods (Kim et al., 2013). Ultimately adapting teaching approaches to enhance student understanding and performance determines the extent of success teachers achieve in imparting knowledge, facilitating understanding and empowering students towards developing expertise.
Traditional classroom methods that requires chalk and board or teacher-led presentations are still common practices in many parts of the world. However since the turn of the century, the use of ICT in education has been widely endorsed towards achieving scalability and sustainability (Raja & Priya Lakshmi, 2022). The use of innovative technology integrated practices in teaching learning processes are multi-fold such as enhancing visualization, immersion and knowledge transfer (Kochkorbaevna & Gulomova, 2022; Abduraxmanova, 2022).
Yet surprisingly, the diffusion of ICT innovations into teaching practices have been slow and gradual. The determinants include apprehension to technology, lack of infrastructure and hesitancy to change long accustomed teaching and adequate training to comfortably use constantly evolving technologies. On the other end of spectrum, constraints on adoption of innovations are induced by challenges in their respective institutions in terms of prioritization of innovative practices with or without the right infrastructure. Irrespective of individual circumstances, literature overwhelmingly supports the use of technology in teaching-practices towards positive student outcomes. Social, behavioral and academic factors have been extensively studied that suggest teachers may fall broadly in these categories: those that are open to considering innovations but are either early adopters, late adopters or laggards and those that are critical to use of technology (Achuthan et al., 2020).
COVID-19, a health-induced pandemic had a colossal impact globally, overturning almost every aspect of day-to-day lives. In the field of education, institutions and all its stakeholders were restrained significantly and forced to adopt online communication and teaching activities. Recent studies conducted during-COVID-19 showed that due to the a shift from “face-to-face” teaching to exclusively online teaching that required use of either synchronous or asynchronous methods significantly affected, the perception of their psychological well-being and management of smart working modes (Varghese & Musthafa, 2022; El-Soussi, 2022; Li & Yu, 2022). In most cases, teachers have had no specific training to supplement their digital technology based teaching methods (Schleicher, 2020). It is commonplace for academicians to integrate use of online resources into their various teaching-learning activities (Mishra et al., 2020; Bordoloi et al., 2021). However, laboratory experimentation or development of laboratory skills through a virtual platform is uncommon. The exceptions are few institutes that have developed and/or mandate use of virtual resources to complement their laboratory learning (Achuthan et al., 2020; 2021). Needless to say, the majority of teacher perception towards any platform impacting laboratory skills can be poor due to the significant shift away from the norm of associating laboratory learning with physical laboratories only. The pandemic forced teachers and students to embrace online resources much more than ever before.
While virtual platforms are not meant to be substitutes for physical laboratories, they are successfully utilized to execute distance learning and supplement curricular learning. Virtual laboratories (VLs) offer a paradigm shift in laboratory education practices and have shown promise in enhancing conceptual understanding, self-efficacy, and practice of laboratory protocols online (Kolil et al., 2020; Achuthan et al., 2018a). The early adopters of this technology often integrated VLs as pre-laboratory formalities to maximize learning outcomes among students. VLs are learner-centric by design and support inquiry-based learning (Hilfert-Rüppell et al., 2013), adapt to diverse learners (Blumer & Beck, 2019) and impact cognitive and effective aspects of learning (Aikens, 2020; Brownell et al., 2012; Gormally et al., 2009; Harrison et al., 2011; Hofferber et al., 2016). Understanding the driving factors of teacher acceptance of virtual laboratories is not only critical in situations such as the pandemic but also an important direction for teachers to consider use of ICT in laboratory teaching.
While COVID-19 has had a positive impact on adoption of technologies in general, such adoption may remain temporal unless a deeper engagement and appreciation on the use and benefit of technologies are tangibly witnessed and capitalized by teachers. There is a need to theoretically study the socio-contextual factors involved in the prediction and acceptance of technologies amongst teachers. Key contributions of this work include cross comparison of teacher behavior’s in adoption of VLs both pre-pandemic, and during the pandemic. The challenges to their wide adoption were also investigated. This work analyzes the critical factors such as performance expectancy (PE), effort expectancy (EE), social influence (SI), hedonic motivation (HM), facilitating conditions (FC) and habit (HA) based on UTAUT2 (Unified Theory of Acceptance and Use of Technology-2) theory that triggered VL adoption over five years including the two years of the pandemic with the following research questions: What is the overall level of acceptance of VL among teachers?
To what extent is the research model valid in explaining the acceptance
What are the differences in teacher beliefs prior to and during pandemic?
To what extent does the model predicts acceptance of VL
Additionally, the role of teacher training precipitated by the absence of institutional support services for VL adoption and the usage experiences of teachers is also investigated.
A theory-based approach to modeling acceptance of virtual laboratories amongst teachers has not been studied before and is the core contribution of our work. Most studies focusing on teacher adoption of technologies have used the technology acceptance model (TAM). However, the recent introduction of the UTAUT2 theory is more appropriate as it builds on several important theories such as the theory of reasoned action, the theory of planned behavior, diffusion of innovations theory, social cognitive theory, model of personal computer use and the motivational model (Chatterjee & Bolar, 2019). In fact, the UTAUT2 model evidences accurate prediction of behavioral intentions in the context of mobile learning technologies and their acceptance across multiple studies (Arain et al., 2019; Nikolopoulou et al., 2020; Ameri et al., 2020). Recent studies have utilized UTAUT2 in contexts such as acceptance of MOOCs (Tseng et al., 2022) that showed the most influential aspects that were responsible for increased adoption were facilitating conditions and behavioral intention. Pandemic-based studies on teacher acceptances of mobile technologies (Dahri et al., 2021), emphasized the importance of integrating technology-focused training as part of the continuous professional development of teachers. While these studies have important implications, most of the prior studies on teacher acceptances have been cross-sectional. Additional novelty of the work described here includes presenting a longitudinal study that spans the pandemic period. Furthermore, this study also delves into overcoming one of the critical barriers in acceptance of technology i.e. lack of training and characterizes its influence in enhancing acceptance and adoption.
Mobile-learning
The penetration rate of smartphone in developing countries has been exponential over the past decade. Mobile-learning has significant potential to motivate students to learn as these devices are part and parcel of most students’ daily lives and are available to them 24 hours a day. However, youth often use mobile phones to communicate with peers, post status on social media and likely to engage in counterproductive activities (Li et al., 2022; Fredrick et al., 2022). They are often distracted due to constant flooding of alerts, group chats or news. This in itself often causes institutions to ban mobile phones during instructional hours (Selwyn & Aagaard, 2021; Dontre, 2021; Kolhar et al., 2021). It is unsurprising that teachers do not consider use of mobile phones in mainstream education as effective due to a number of fears associated with them (O’Bannon & Thomas, 2015). These observations however do not undermine the importance of mobile apps and interfaces in education. Numerous studies exemplify its use in formal and informal contexts. Previous studies claim mobile technologies augment knowledge transfer and retention (Zydney & Warner, 2016) while other studies (Jeno et al., 2019) elaborate facilitation of collaboration and interactions that ensure student engagement. In order to ensure students embrace such technologies, teacher play a dominant role in technology integration. Studies show teacher’s beliefs and attitudes guide their action (Kalogiannakis & Papadakis, 2019; Khlaif, 2018). Literature is sparse on how teacher beliefs translate to development of skills and sustained adoption. During the pandemic, mobile phones served as lifelines to stay connected. They were critical to systematizing educational activities for both students and teachers. However, their improper use post pandemic could retreat teachers from encouraging mobile platforms for educational assignments and wane away their interest to pursue newer pedagogical m-learning interventions. This is the first of its kind study to characterize and predict teacher’s intention to use of mobile interfaces towards digital laboratory instruction.
Mobile virtual laboratories
Experimental laboratories hosted online that are either simulation based, animation based on remote triggered (Granjo & Rasteiro, 2020; Liu et al., 2022) offers students significant flexibility to access and perform multiple trials of experiments at their convenience in comparison to physical laboratories. They are by design student-centric and facilitate individualized learning of experimental concepts. The interactive animations and simulations emulate experimentation in a physical laboratory enhancing the visualization and retention of experimental procedures. As part of this work, the underlying platform for virtual laboratories was built using HTML5 making them suitable to run across a variety of mobile devices. Henceforth, we refer to this platform as mobile virtual labs (M-VL). Demand for M-VL over the past couple of years has skyrocketed in India due to the pandemic; whereby roughly 320 million learners were affected by the closure of schools and colleges (Jena, 2020).
While several higher educational institutions (HEIs) managed to ensure the continuity of teaching-learning using ICT tools such as M-VL, majority of the institutions were either oblivious or less receptive to the myriad choices (Lederman, 2019). This exacerbated the already compromised state of education during the pandemic. As indicated in the literature (Kaliisa et al., 2022), teachers play a crucial role in integrating pedagogic interventions resulting in successful student adoption of technologies. Prior study summarizes factors (Khlaif, 2018) influencing teachers’ attitudes towards VL adoption, namely, prior experience with mobile technologies, access to technical support, instructional assistance, and infrastructure. Needless to say, the pandemic had a debilitating impact on most of these factors that would have otherwise been available to teachers at their respective institutes. The motivation for this study was to look at possibilities that transmute these constraints. Comprehensive studies characterizing the transformation of teacher attitude and beliefs induced by the pandemic of online laboratories are absent. The role of teacher training and assessing the overall acceptance of late adopters or laggards of technology towards sustainable teaching practices was studied as part of this study.
Educational outcomes form M-VLs
Three distinct educational outcomes (Fig. 1) that are embedded in the design of M-VLs of value to both teachers and students include: Knowledge Enhancement: The literature supports that use of virtual laboratories in classrooms can improve conceptual understanding (Gunawan et al., 2018; Darby-White et al., 2019) about the theory behind experiments and provide better understanding about the procedural steps of the experiment for students (Adam et al., 2020; Kolil et al., 2020). The modality of usages vary in how they are integrated within the learning environment. Some mandate for VLs be used as practice platforms prior to performing real experimentation and while others use VLs to supplement theoretical courses. Previous studies suggest that the visualization plays an important role in mitigating the alternate conception students have about scientific concepts (Achuthan et al., 2018a). Additionally protocols associated with experimental study are demonstrable and aid students with rehearsing them prior to their entry into physical laboratories. As an example, extension studies have shown performing online experiments imparted sufficient practical knowledge about handling of hazardous materials, reading safety labels and guidelines about their safe storage (Artdej, 2012). Other studies elaborate the scientific skills such as observation, data analysis that are exercised through VLs (Hodges et al., 2018). Thus the primary education outcome of VLs is the imparting and enhancement of knowledge that occurs amongst students through perusal of primary and supplemental information (Brüggemann & Bizer, 2016). Teachers can further enhance the experience by giving them additional opportunities to apply the learning with additional assignments.
Repetitive Experimental Practice: Educational institutions often group students to perform experiments in their physical laboratories. Primary causes for this include: limited instrumentation and infrastructure coupled with large volume of students. This results in partial experience students gain in doing experiments inside the laboratory. Additionally, when students are grouped together, there is a natural tendency for some to be active participants while others remain passive. Thus the extent of hands-on knowledge gained by passive students remain questionable. In physical laboratory, there is also a need to demonstrate the performance of the experiment before students can execute them. All of these limitations that are part of traditional laboratories can be addressed using M-VLs. Through repetitive performance of online experiments, the meta cognitive abilities are enhanced amongst students (Yusuf & Widyaningsih, 2020; Achuthan et al., 2017). This results in reinforcing the details of experimental knowledge and protocols far more efficiently. This also avoids the need for demonstration of experiments in physical labs. Doing VLs have shown to significantly reduce the experimental time students spend in physical laboratories (Bolkas et al., 2022; Domínguez Alfaro et al., 2022).
Assessment of Learning: Assessing learning helps every teacher strategize their pedagogic approaches to learning and maximize students’ application-oriented conceptual understanding. In the present scenario of M-VLs, a teacher is provided with three assessment tools. 1) pre-assessment questions or self-evaluation done prior to running the online simulator, 2) post-assessment questions or assignments after performing the online experiments, and 3) VL-learning management system (VL-LMS) (Achuthan et al., 2021) that monitors and manages the student usage with specific metrics such as time spent on an experiment, the attempts made to answer questions and so on. The pre-assessment multiple-choice questions gauge the understanding of the underlying theory of experiments. The post-assessment questions are more related to practical applications of experimental concepts. These questions also demonstrate critical thinking and problem-solving ability. By automating and using these assessments, teachers can readily gauge an overall understanding of students’ knowledge about specific and relevant topics related to the experiments.
Fig. 1 Educational outcomes of mobile virtual labs
Theoretical framework and research hypotheses
The acceptance of technology is measured based on the models and frameworks that have been developed to explain the adoption of new technologies (Davis, 1985; 1989; Davis et al., 1989). One of the prevailing models for user acceptance is the Technology Acceptance Model (TAM) (Venkatesh et al., 2003; Bagozzi et al., 1992). As an adaption of the Theory of Reasoned Action (TRA) (Fishbein & Ajzen, 1977), TAM is a generic model developed particularly for explaining computer acceptance (Davis, 1989). Initially, TAM and its derivations (Venkatesh & Bala, 2008; Venkatesh & Davis, 2000) were research tools that were used to investigate end users’ acceptance of technological innovations. However, it does not incorporate individual’s perspectives and that attitude towards usage and behavior (Chao, 2019). Later, it became clear that TAM was only able to predict technology acceptance in 40% of instances (Venkatesh & Davis, 2000). These drawbacks of TAM eventually led to the development of the Unified Theory of Acceptance and Use of Technology (UTAUT) (Venkatesh et al., 2003).
The UTAUT model proposes four main constructs. They are performance expectancy (PE), effort expectancy (EE), social influence (SI) and facilitating conditions (FC). With regard to mobile learning, UTAUT is considered most appropriate to study technology use and acceptance. Later in 2012, the model was modified to UTAUT2 by the addition of another two constructs to further enhance the accuracy technology acceptance and usage. The additional factors included hedonic motivation (HM) and habit (HA) (Venkatesh et al., 2012). These factors influenced both the behavioral intention (BI) and usage behavior (UB), that are ostensibly critical to M-VLs as well. Literature has not delved into usage of UTAUT2 in the context of M-VLs. Thus the current study gains relevance in investigating factors influencing teacher acceptance and use of M-VLs based on the UTAUT2 towards sustained adoption (Fig. 2). Fig. 2 Mobile Virtual Lab Research Model: UTAUT2 adapted from Venkatesh et al. (2012)
Performance expectancy (PE)
It is defined as the extent to which individuals believe that they can fulfill tasks quickly using the technology (Venkatesh et al., 2003). Studies showed that PE will have a positive influence on users’ behavioral intention (BI) such as in the case of using OTT video streaming platforms (Malewar and Bajaj, 2020),or mobile phones (Nikolopoulou et al., 2020), animation and storytelling and so on Suki and Suki (2017). PE predicts that teachers’ use of mobile technology makes daily teaching smoother and can increase work productivity (Ismail et al., 2022). In post-pandemic teaching, reliance on the internet increased. Another study posited that when teachers believe that mobile internet is useful for educational purposes, and enhances achievements and productivity, their intention to use it increases (Nikolopoulou et al., 2021a). In this study, PE for teachers was measured in terms of usefulness in teaching, efficiency of task completion, improvement in quality of teaching and degree of employ ability. Consistent with the literature, this study therefore proposes that: H1: Performance expectancy has an effect on teachers intention to use mobile VL as a teaching tool.
Effort expectancy (EE)
It is characterized as the extent of ease with which users are using learning technology (Venkatesh et al., 2003). User-friendly technology is easy for users to accept and implement. Past studies corroborate that learners have shown preference for technologies that are user friendly, versatile and efficient. Research has shown that EE will have a positive impact on BI, PE (Davis, 1989) and users’ intention to use technologies (Giesing & et al., 2005). Teachers’ convenience in using mobile devices is a powerful way to enable them to adopt innovative and technological learning (Ismail et al., 2022). Past findings show that effort expectancy significantly affects teachers’ behavior intention to use mobile technology (Thomas et al., 2013; Nikolopoulou et al., 2021a; Hu et al., 2020). When teachers believe that it is easy to use mobile internet in their teaching practices (i.e. with not much difficulty, effort, or complexity), their intentions are likelier to be more positive. The greater the EE, the faster the expected rate of mobile internet and technology adoption (Nikolopoulou et al., 2021a). In this study, EE was measured in terms of ease of teaching in relation to physical labs, improvement in skill level in using M-VL to teach, duration of time taken to teach using M-VL and quality of resource packages available in M-VL. In accordance with this reasoning, the following hypothesis is postulated: H2: Effort expectancy has an effect on teachers intention to use mobile VL as a teaching tool.
Social influence (SI)
It is described as the extent to which one perceives it important that others agree he or she should use a new technology (Venkatesh et al., 2012). It was shown that this perception had a positive relationship with the BI to use a technology (AbuShanab et al., 2010; Eckhardt et al., 2009; San Martín & Herrero, 2012; Venkatesh et al., 2003). When teachers believe that important persons (e.g., stakeholders, school leaders, colleagues) influence their use of mobile technology, they are more likely to adopt mobile internet for educational purposes if the leadership also views mobile VL as an investment. Teachers acknowledge that other people have a role to play in the process of their mobile internet adoption (Nikolopoulou et al., 2021a) For example, if institutional leadership initiated digital leadership and increased the speed of transition to using VL (Alammary et al., 2021), teachers are more likely to follow their example and the perceived value of an educational resource like VL will increase, paving way for its adoption. In this study, social influence on teachers was characterized by actual usage of mobile VL by colleagues and friends, opinions of valued colleagues and friends, support from leadership bodies in institutions and student preferences. This study proposes that: H3: Social influence has an effect on teachers intention to use mobile VL as a teaching tool.
Facilitating conditions (FC)
This was defined as the degree to which a user believes that an organizational and technical structure exists to support the use of the technology (Venkatesh et al., 2012). It is shown that users will have high tendency to use a particular technology if there is a support system in place to troubleshoot or guide their use. AbuShanab et al. (2010), Eckhardt et al. (2009), and San Martín and Herrero (2012) FC has direct effect on use behavior (Venkatesh et al., 2003). Technical (e.g., restricted internet access and limited mobile technology infrastructure) and organizational problems (e.g., lack of support personnel) could possibly inhibit teachers’ intentions and use of mobile internet in teaching. When teachers believe there is sufficient school support for mobile VL implementation, their intentions to adopt-use it will be stronger (Nikolopoulou et al., 2021a, b). Pre-pandemic, institution based facilitating conditions might have exerted a stronger impetus on teachers by provided conducive external environment with infrastructure and support. Post-pandemic, however, facilitating conditions remain relevant in context of work-from-home but are still significant. For example, the same issues which existed previously; redefining the teacher’s role within the learning process (Kalimullina et al., 2021; Noble, 1998) have to be addressed now to facilitate smooth adoption and usage of mobile VL by teachers. A recent study argues that crises like COVID-19 themselves are facilitating conditions that effect teachers’ efforts toward adoption of innovation in online teaching. It is achieved through the improvement of personal teaching efficacy and ICT efficacy (Yu et al., 2021). In this study, we characterized facilitating conditions as the infrastructure to use VL, knowledge to use VL, and availability of technical assistance to troubleshoot while using VL. From this discussion, the following hypotheses are formulated: H4a: Facilitating conditions has an effect on teachers intention to use mobile VL as a teaching tool.
H4b: Facilitating conditions has an effect on teachers use of M-VL as a teaching tool.
Hedonic motivation (HM)
It is defined as the fun or pleasure derived from using a technology (Brown & Venkatesh, 2005). HM shows positive relationship with technology adoption (Yang, 2013). Hedonic motivation for teachers to use mobile VL has to be significantly stronger than the pressures of smart working due to the pandemic; in order to have a tangible effect on teachers who have very stressful jobs compared to other professions (Mari et al., 2021). We hypothesize that hedonic motivation has an effect on teacher’s intention to use mobile VL as it is practically status quo to teach using it. Naturally, it inevitably increases usage and habit development in its widespread adoption. Hedonic motivation is primarily derived from usage and has been cited as a strong predictor for teachers to adopt VL. The enjoyment derived from using mobile VL during course instruction, also making lessons enjoyable for students and reinforcing their engagement, is expected to positively affect teachers’ intention to adopt mobile internet in their teaching (Nikolopoulou et al., 2021a). Hedonic motivation in this study is measured by process enjoyment, reduction of workload and aesthetic appeal of mobile VL. From this discussion, the following hypotheses are formulated: H5a: Hedonic motivation to use mobile VL has an effect teachers intention to use mobile VL as a teaching tool.
H5b: Hedonic motivation has an effect on teachers use of mobile VL as a teaching tool.
Habit (HA)
According to the literature, habit is a repeated action that can sometimes occur unconsciously and is shaped by experiences, knowledge as well as skills learned over time (Limayem et al., 2007; Venkatesh et al., 2012). Studies confirm a positive association exists among habit, behavioral intention, and adoption. Teachers’ prior experience and habit of using mobile internet in daily life possibly stimulates its adoption and continuous use for educational purposes (Nikolopoulou et al., 2021a). Teachers developed the habit of using mobile VL during the pandemic even if it was intermittent or non-existent before. With habitual use over the prolonged pandemic, effort expectancy also improves as usage becomes easier. While adoption of technology forced by the pandemic can wane away, consistent practice in using the technology as teaching aids is the only way to sustained use of VL even after the pandemic. From this discussion, the following hypotheses are formulated: H6a: Habit to use M-VL has an effect teachers intention to use mobile VL as a teaching tool.
H6b: Habit has an effect on teachers use of mobile VL as a teaching tool.
Behavioral intention (BI)
Behavioral intention is a function of both attitudes towards and subjective norms regarding target behavior. Behavioral intention predicts actual behavior (Pickett et al., 2012). The willingness of users to use technology can be ascertained from their behavioral intention. In this study, BI is the extent to which teachers intend and continue to use mobile VLs in their teaching. Hence, the following hypothesis is offered: H7: Behavioral intention to use mobile VL has an effect on teachers use behavior.
Use behavior (UB)
The use behavior represents the actual behavior from the person. Studies show that there exists a positive relationship between past behavior and future behavior (Venkatesh & Davis, 2000) and behavioral intention is a valid predictor of actual usage behavior. A previous study concluded that teachers display a positive view towards learning new techniques and of using instructional technology for their teaching (Peluchette & Rust, 2005). Existing research on user behavior indicates that there was a positive correlation between behavioral intention and users behavior in many models of technology acceptance theories (Venkatesh et al., 2003). In this study, UB is the extent to which teachers use their mobile VLs in their teaching (as learning tools).
Methods
Participants
A five year investigation that encompassed teacher adoption of technology prior to and during the pandemic between 2017 and 2021 from 115 higher educational institutes from India is included in this study. Theory based comparisons were made between institutions who had teachers that were early adopters of VL versus those that used VL only due to the pandemic. Teachers specialized in teaching chemistry were invited to share their experience with chemistry VLs as part of this work. The overwhelming response from 650 teachers from 115 institutions were recorded and analyzed. The institutions and teachers were divided into two groups. Group I had 58 institutions, whose teachers were familiar with the VL platform and used them as supplemental aids to their teaching practices. The transformation in the behavioral characteristics of these teachers prior to (Group IA) and during the pandemic (Group IB) were addressed in this study. Group II had 57 institutions that had adopted VLs just during the pandemic and had no prior to pandemic experience with VLs.
Table 1 shows the selected demographics of the respondents. This study included 60.8% female participants and 39.2% male participants. The average age of the participants were 34 and 22.5% of them have 1 year of VL experiences, 52.6% have 2 years and 24.9% have 3 years and VL experience. Approximately 16% (Group I participants) and 15% (Group II participants) had PhD degree and 78% had only masters degree.
We chose chemistry teachers (650) from virtual lab nodal centers for the study (Achuthan et al., 2020). The data were collected through online survey and site workshops. Table 1 Demographic profile of the participants
Category Classification Group I Group II
N = 325 % N = 325 %
Gender Female 207 64% 188 58%
Male 118 36% 137 42%
Age < 25 19 6% 21 6%
25-30 52 16% 43 13%
31-35 168 52% 183 56%
> 35 86 26% 78 24%
Virtual Lab experience 1 year 32 10% 114 35%
2 years 210 65% 132 41%
3 years & above 83 26% 79 24%
Qualification PhD 52 16% 48 15%
Mphil. 18 6% 23 7%
Masters 255 78% 254 78%
# Institutes adopted VLs N = 58 N = 57
2017 18 31%
2018 15 26%
2019 25 43%
2020 26 46%
2021 31 54%
Study design
The study design included two parts. The first part of the study analyzed the multi- year (5-year) data of chemistry teachers’ acceptance of M-VLs (Fig. 3a) by comparing Group I and Group II. The acceptance of M-VLs was measured using the UTAUT2 constructs such as performance expectancy (PE), effort expectancy (EE), social influence (SI), facilitating conditions (FC), hedonic motivation (HM), habit (HA), behavioral intention (BI) and use behavior (UB). The data from 2017-2019 were categorized as Group I (pre-COVID19), and 2020-2021 were categorized as either Group IB (i.e. same cohort as Group I during the pandemic) or Group II (new adopters during the pandemic).
The second part of the study describes the teacher training pedagogy of M-VLs that included descriptions and use of various M-VL features that aid laboratory education (Fig. 3b). M-VL training is given to teachers through online and offline workshops. The training output was focused on 1) technology knowledge, 2) technology skill, and 3) technology access. Teachers’ technology understanding and capacity to use various technologies, technological tools, and associated resources in their teaching is referred to as technological knowledge. Technology skills describes teachers’ ability to interact, use the technology and assign tasks to students using technologies and other associated technologies. Technology access describes teachers’ understanding of the methods for gaining access to various features associated with the technologies. The training includes an introductory session, a hands-on session in a computer laboratory, and followed by a feedback session. Trained teachers can use the M-VL platform for 1) course content (theory, procedure, self-evaluation, assignments, animation, simulation, and videos), 2) communication with students, and 3) student evaluations. Fig. 3 Study design of the study. (a) Comparison of teachers adoption factors pre and during COVID-19. (b) Features of mobile virtual labs (M-VL) and teacher training pedagogy
The user interface of a calorimetry virtual lab experiment is shown in Fig. 4a, which allows the user to perform the experiment at any time and observe the temperature changes. The five-year (2017-2021) usage data of M-VLs (Fig. 4b) analyzed as part of this study suggest the increased demand of M-VLs over the years especially during COVID-19 (2020 and 2021). The usage data were collected from M-VLs server and Google Analytics. The retention of users over the years and the increased number of new users over the years are shown in Fig. 4c. Approximately, 40% of the users are returned into the M-VLs every years and approximately 60% of new users are using M-VLs every year. Fig. 4 Mobile virtual Lab user interface and users. (a) Mobile virtual lab simulation interface of Calorimetry experiment, (b) percentage growth of mobile users and (c) new and returning users of mobile users during 2017-2021
Questionnaire design and data collection
The study was undertaken within the context of chemistry teachers’ adoption of mobile virtual lab tool in to their teaching. The research model is shown in Fig. 2. Quantitative methodology was adopted for the study and the data collected using online survey technique to empirically validate the research hypothesis. The data were collected from the teachers belonging to 115 higher education institutes in India on the basis of their willingness and availability via survey method. The survey item divided into two sections. The first section collected the demographic information and VL usage experience (5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree)) of the participants and the second section collected responses for M-VL UTAUT2 model constructs. These constructs include performance expectancy (PE), effort expectancy (EE), social influence (SI), facilitating conditions (FC), hedonic motivation (HM), habit (HA), behavioral intention (BI), and user behavior (UB). The measurement scale of this study referred to the classical scale of Venkatesh et al. (2003). In view of the above analysis, the questionnaire uses a 5-point Likert-scale ranging from 1 (strongly disagree) to 5 (strongly agree). Other than wording modifications to fit the specific technology studied in this research, no changes were made to the user acceptance scale. Survey questionnaire adapted from Venkatesh et al. (2003). The questionnaire of the study consists of 28 items which were contextualized and modified.
Data analysis
Data analysis consisted of two parts. The first part explored the reliability and exploratory factor analysis of the survey instrument and the second part focused on the structural equation model assessment. Data collected was entered into SPSS (version 20) for the analysis of descriptive and inferential statistics and AMOS 23.0 statistical software was used for the structural equation modeling (SEM). The UTAUT2 constructs like performance expectancy, effort expectancy, social influence, facilitating conditions, hedonic motivation and habit were considered as the independent variables in this study. The dependent variable consists of behavioral intention and use behavior.
Results and discussion
Assessment of proposed virtual lab model
The descriptive statistics of the virtual lab adoption using UTAUT2 constructs were shown in Table 2. The mean values of all indicators are between 5.0 and 7.0. This implies that most of the teachers’ responses were either completely agree, moderately agree or somewhat agree. Results from Table 2 also pointed out that most of the respondents concur with the statements in the questionnaire. The reliability co-efficient (α) of all the indicators were above 0.85, indicating a high internal consistency and quality for all scales. The overall reliability co-efficient (α) of the questionnaire was tabulated in the Table 3 and reflects high internal consistency (> 0.70). Table 2 Survey instrument and responses of participants
Indicator Group I Group II
Mean SD Mean SD
PE1: I find that using VL aids in teaching. 3.9 1.63 3.8 1.07
PE2: Using VL in science classrooms enables me to accomplish tasks quickly. 4.0 1.54 4.1 0.90
PE3: Using VL for teaching science improves the quality of my teaching. 4.1 1.53 4.3 0.81
PE4: Using VL for teaching increases my employment opportunities. 3.9 1.59 4.3 0.84
EE1: Teaching students to perform experiments in virtual labs is easy compared to doing so in real labs. 3.7 1.67 3.7 1.11
EE2: It is easy for me to become skillful at utilizing VL to teach. 3.7 1.69 3.6 1.29
EE3: I can learn to teach via VL in a short time. 3.6 1.75 3.8 1.02
EE4: VL’s extensive resource package allows me to channel effort to address student laboratory needs instead. 3.6 1.69 4.1 0.91
SI1: I would use VL if my friends and colleagues used them. 3.8 1.70 4.0 0.89
S12: My colleagues who are important to me think that I should use VL. 3.8 1.67 4.2 0.84
SI3: The University has been helpful with enabling me to learn how to use VL as a teaching tool. 3.8 1.66 4.2 0.88
SI4: The students prefer to learn via technology like VL rather than traditional laboratory teaching method. 3.8 1.67 4.3 0.72
FC1: I have the infrastructure to use VL. 3.6 1.73 4.1 0.91
FC2: I have the necessary knowledge to use VL. 3.7 1.73 4.0 0.96
FC3: When I need help to use VL, technical assistance is available. 3.6 1.73 4.1 0.92
BI1: I intend to utilize VL as a teaching tool. 3.9 1.65 4.1 0.87
BI2:I think VL will be well-received by students. 3.9 1.63 4.2 0.85
BI3: I plan to incorporate VL into the science syllabus. 3.9 1.63 4.4 0.72
HM1: Using VL is an enjoyable process. 4.0 1.56 4.4 0.68
HM2: VL reduces my workload. 3.9 1.63 4.2 0.82
HM3: VL is an aesthetically pleasing way to learn how to perform experiments. 4.0 1.59 4.3 0.79
HA1: I turn to VL first when I am looking for solutions to problems in laboratory experiments. 3.8 1.70 4.2 0.76
HA2: When I want to learn proper methodology of performing experiments I regularly refer to VL to corroborate my understanding. 3.7 1.72 4.3 0.76
HA3: I routinely obtain teaching materials from VL. 3.7 1.67 4.4 0.66
HA4: Using VL has become natural to me. 3.7 1.73 4.3 0.78
UB1: I intend to continue using VL in the future. 3.9 1.61 4.2 0.74
UB2: I prefer using VL to perform laboratory experiments. 4.0 1.62 3.8 0.94
UB3: Most of laboratory classes rely on VL for holistic learning. 4.0 1.61 4.3 0.70
PE: performance expectancy; EE: effort expectancy; SI: social influence
FC: facilitating conditions; BI: behavioral intention; HM: hedonic motivation
HA: habit; UB: use behavior
Table 3 Reliability test results of the instrument
Constructs Cronbach’s Alpha (α) No. of Items AVE Composite reliability
Performance Expectancy (PE) 0.907 4 0.712 0.908
Effort Expectancy (EE) 0.907 4 0.732 0.916
Social Influence (SI) 0.902 4 0.697 0.902
Facilitating Condition (FC) 0.846 3 0.675 0.862
Behavioral Intention (BI) 0.872 3 0.747 0.899
Hedonic Motivation (HM) 0.876 3 0.739 0.895
Habit (HA) 0.897 4 0.681 0.895
Use Behavior (UB) 0.890 3 0.627 0.834
Overall 0.800 28
AVE = average variance extracted
Table 4 shows the results of the principle component analysis (PCA) with Varimax rotation describing the items associated with every factors components (PE, EE, SI, FC, BI, HM, HA and UB). The final loading and cross-loadings table displays adequate convergent and discriminant validity. The average variance extracted (AVE) and composite reliability of the constructs were tabulated in the Table 3 and theresults suggest higher convergent validity (composite reliability > 0.70 and AVE > 0.50) (Hair et al., 2006) of the constructs. The discriminant validity is assessed by comparing the square roots of AVE to the inter-factor correlations between constructs and the results suggest that the AVE is higher than the squared inter-scale correlations of the constructs (Fornell and Larcker, 1981) (Table 5). Thus the discriminant validity is supported. A few items had to be dropped from the initial solution due to low loadings and high cross-loadings. The cumulative variance explained by the eight factors was 79.610% and lowest loading in the final model was 0.735 for use behavior (UB2) construct. The Kaiser-Mayer-Olkin (KMO) measure of sample adequacy was identified as 0.849. Table 4 Factor loadings and cross-loadings of the survey instrument
Item Factor
PE EE SI FC BI HM HA UB
PE1 .889 .042 .075 − 9.11 −.035 .016 .020 −.101
PE2 .808 .014 .043 −.199 −.101 −.034 .077 −.236
PE3 .821 .051 .027 −.234 −.068 −.017 .045 −.179
PE4 .856 .037 −.025 −.121 −.027 .041 −.010 −.163
EE1 .033 .855 .137 −.049 −.013 .027 .193 −.079
EE2 .080 .866 .110 .051 .061 −.011 .163 .022
EE3 −.027 .862 .180 −.089 −.047 .032 .123 −.102
EE4 .054 .839 .171 .148 .075 −.050 .147 .116
SI1 −.018 .178 .837 .023 .114 .087 .216 −.055
SI2 .012 .138 .837 .020 .162 .073 .150 .041
SI3 .108 .151 .827 .032 .126 .098 .123 .047
SI4 .003 .144 .839 −.015 .079 .135 .138 .117
FC1 −.204 .037 −.005 .847 −.028 .032 .028 −.239
FC2 −.229 −.014 .055 .813 −.055 −.015 .077 −.236
FC3 −.261 .031 .021 .804 −.091 .003 .029 −.220
BI1 .001 −.005 .228 −.071 .872 .002 .128 −.044
BI2 −.119 .019 .087 .029 .875 .039 .172 −.007
BI3 −.079 .049 .132 −.111 .846 −.038 .177 −.072
HM1 −.004 −.017 .112 .007 .019 .868 .088 −.030
HM2 −.039 .003 .109 −.061 −.023 .868 .111 −.114
HM3 .058 .012 .113 .073 .008 .843 .113 .176
HA1 .042 .154 .117 .097 .134 .149 .820 .041
HA2 −.014 .162 .227 .036 .162 .137 .810 .056
HA3 .034 .160 .131 −.084 .137 .026 .850 −.018
HA4 .048 .175 .166 .079 .101 .071 .821 .060
UB1 −.264 −.081 .049 −.279 −.061 .001 .037 .806
UB2 −.374 .003 .072 −.297 −.070 −.043 .102 .735
UB3 −.252 .026 .064 −.273 −.036 .055 .032 .832
Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization
Values in the bold (values above 0.5) are representing those that conform to that category
Table 5 Discriminant validity (Fornell and Larcker Criterion) of the instrument
Constructs PE EE SI FC BI HM HA UB
PE 0.844
EE 0.087 0.856
SI 0.053 0.365 0.835
FC 0.014 0.042 0.257 0.822
BI − 0.095 0.100 0.320 0.034 0.864
HM − 0.352 0.040 0.025 0.024 − 0.061 0.860
HA 0.052 0.383 0.412 0.244 0.340 0.059 0.825
UB − 0.402 − 0.026 0.085 0.028 − 0.031 − 0.402 0.068 0.792
PE: performance expectancy; EE: effort expectancy; SI: social influence
FC: facilitating conditions; BI: behavioral intention; HM: hedonic motivation
HA: habit; UB: use behavior
Values in the bold representing the square root of average variance extracted
Table 6 shows the list of model fit indices and their corresponding threshold values. From the results it is clear that virtual lab model fit indices fall within acceptable regions. The confirmatory factor analysis (CFA) shows the absolute fit indices AGFI [adjusted goodness-of-fit index] = 0.893, RMSEA [root mean square error of approximation] = 0.035, NFI [normed fit index] = 0.927, CFI [comparative fit index] = 0.978 and ratio of χ2 (480.683) to degrees of freedom (df = 325) = 1.387 had better values than the thresholds. Hence the result revealed that the measurement model fit with the data collected. Table 6 Model fit indices (CFA)
Fit Indices Threshold values Model fit Indices Ref.
χ2/df < 3 1.387 (Bagozzi & Yi, 1988)
AGFI > 0.8 0.893 (Chau & Hu, 2001)
RMSEA < 0.08 0.035 (Browne et al., 1993)
NFI > 0.9 0.927 (Chin & Todd, 1995; Hair, 1998)
CFI > 0.9 0.978 (Bagozzi & Yi, 1988)
Structural equation modeling
The UTAUT2 model used in this work served as the foundation to map the relationship between the various exogenous and endogenous variables in VL. The model was derived using the statistical software Analysis of Moments Structures (AMOS) 23.0. The model investigated most influential factors that drive user acceptance behavior. We tested the model with a data set of 325 samples each for Group I and II. The relationship between the exogenous and endogenous variables are summarized in Table 7: Table 7 Relationship between exogenous and endogenous variables
Group I Group II
Hypothesis Path Estimate p Information Estimate p Information
H1 BI <— PE − 0.556 *** Supported 0.212 *** Supported
H2 BI <— EE − 0.058 0.084 Not Supported − 0.106 0.064 Not Supported
H3 BI <— SI 0.179 *** Supported 0.219 *** Supported
H4a BI <— FC − 0.105 0.128 Not Supported 0.273 *** Supported
H5a BI <— HM − 0.339 *** Supported − 0.086 0.075 Not Supported
H6a BI <— HA 0.409 *** Supported − 0.325 *** Supported
H7 UB <— BI 1.000 *** Supported 1.000 *** Supported
H4b UB <— FC 0.08 0.403 Not Supported 0.287 *** Supported
H5b UB <— HM − 0.339 *** Supported − 0.14 0.172 Not Supported
H6b UB <— HA − 0.318 *** Supported − 0.412 *** Supported
Estimate = regression coefficient; p = significant level, 0.05
Factors influencing teachers acceptance of M-VL
The p value of PE (Table 7) is less than the threshold (p< 0.05) which indicates a significant effect of PE on BI, and thus we accept the hypothesis H1. The result suggests that performance expectancy negatively correlates to teachers’ behavioral intention to incorporate M-VLs into their teaching before COVID-19; but positively correlates to teachers behavioral intention to integrate M-VLs into their teaching during COVID-19. The disparity in PE’s effect on BI is explained by increased reliance on technology for teaching and heightened emphasis on self-learning during the pandemic. It has been well established that VL’s truly interactive learning environment has the power to improve the learning process and skill acquisition (Kolil et al., 2020; Achuthan et al., 2018b). M-VL particularly shortens learning curves in STEM subjects like chemistry as students may use M-VL to experiment with real-world phenomena using literally their finger tips. In addition, M-VL could affect students’ spatial ability for visualization and spatial arrangement, critical to experimental skills. Past literature proposes mobile technologies as cheaper alternatives for educational institutes too. In spite of all these, teachers frequently encounter issues concerning student engagement with teaching materials (Manzoor et al., 2021) and BI to use M-VL may not have been strong in the past. COVID-19 has been a strong catalyst for change. Past studies have had mixed results with some showing that performance expectancy is not significantly related to teachers’ behavioral intention to use mobile technology (Testa & Tawfik, 2017) while others show that performance expectancy is integral to teacher’s intention to use mobile VL as a teaching tool (Hu et al., 2020). More recently, UTAUT2 was used in a few studies globally, showing that PE significantly affects teachers’ behavioral intentions to use mobile devices and mobile technologies in the curriculum (Adov et al., 2020; Kim & Lee, 2020; Omar et al., 2019; Wong et al., 2020).
Effort expectancy has insignificant effect on BI (Group I: p = 0.084; Group II: p = 0.064) and the hypothesis H2 is rejected (Table 7). Based on this result, we conclude that EE is not a valid predictor in determining BI. These results corroborate with prior work that found EE to have no statistically significant correlation to BI (Raman et al., 2014; Verkijika, 2018; Herrero et al., 2017; Lallmahomed et al., 2017; Morosan & DeFranco, 2016; Oliveira et al., 2016). This contradicts the UTAUT model where EE is thought to be a significant influence on BI (Venkatesh et al., 2003; Attuquayefio & Addo, 2014). While this may seem non-intuitive since teachers’ usage of M-VL increased by default during the pandemic, it is also indicative of all teachers not solely relying on M-VL to teach chemistry. Globally and within India, the literature confirms that only dire need and compulsion from management forced teachers to adopt M-VL during the COVID-19 pandemic (Selvaraj et al., 2021). In India, schools faced closure for a prolonged periods of close to two years. Some literature argues that being forced to adopt a teaching method reduces enthusiasm for use of it. During the pandemic, teachers also felt burdened by compounded issues such as technical issues that they could not resolve by themselves, slow internet, change in classroom dynamics (Joshi et al., 2020; Kruszewska et al., 2022; Selvaraj et al., 2021), lesser controls in the classroom and other unprecedented issues that made adopting M-VL less smooth that intended. Thus, effort expectancy might have worsened and was insignificant on teacher’s BI to use M-VL.
Social influence has a significant positive effect on behavioral intention (Group I: 0.179, p< 0.05; Group II: 0.219, p< 0.05) and we accept the hypothesis H3 (Table 7). And we conclude that SI is a strong predictor in determining BI before and during-COVID19 lockdown. The result of SI was supported by other studies as well Nandwani and Khan (2016), Venkatesh et al. (2012), Raman and Don (2013), Teo et al. (2008), Teo (2009), Looney et al. (2004), and Farahat (2012). Pre-pandemic, using a device to learn or teach might have been perceived negatively by those with traditional views (Kalimullina et al., 2021). However, post-pandemic, many institutions have adopted a new, positive perspective (Ngugi and Goosen, 2021; Jena, 2020). Studies report that post-COVID-19, attitudes toward e-learning, self-efficacy and perceived reliability have significant positive effects on behavioral intention to adopt e-learning or mobile technologies such as VL (Alammary et al., 2021). School leaders around the world have contributed to positive attitudes; bolstering teacher self-efficacy and perceived reliability of M-VL to support educational needs by normalizing shifting of classes to online platforms like Zoom and Google Meet (Osman, 2020; Shrestha et al., 2022; Setyawan et al., 2020; Pratama et al., 2020). Along with these synchronous modes of teaching, asynchronous self-learning platforms like edX and Coursera have also seen an increase in independent enrollments (Shah, 2020) encouraged by leadership.
Facilitating conditions (FC) have insignificant effect on behavioral intention to use M-VL (p = 0.128) and use behavior (p = 0.403) before COVID-19 (Table 7). Thus we reject the hypotheses H4a and H4b. The result of FC with BI was supported by Oechslein et al. (2014) but contradicts the results of students’ intention to use the available ICT for learning and research intention (Attuquayefio and Addo, 2014). But FC shows a positive significant influence on the BI (0.273, p< 0.05) and UB (0.287, p < 0.05) during COVID19 lockdown and we accept the hypotheses H4a and H4b during COVID-19. This is indicative of the fact that educational institutes such as in Group II encouraged and facilitated use of M-VL only during the pandemic.
Hedonic motivation had a significant negative effect on behavioral intention to use VL (-0.339, p< 0.05) and use behavior (-0.339, p < 0.05) before COVID19 and was not a significant predictor (BI: p = 0.075, ; UB: (p = 0.172)) during-COVID19 (Table 7). Thus we accept the hypotheses H5a and H5b before COVID19 and reject H5a and H5b during COVID19. The pandemic stressors that could inhibit hedonic motivation for teachers include lack of human interaction, overwhelmed and unprepared parents being unable to support their children with mobile VL (Petrie et al., 2020); needing to maintain a positive relationship in the context of collaborative learning (Petrie et al., 2020) and lastly concerns about the possible gap in student achievement between high and low socioeconomic status families due to differences in living space or time availability (Bol, 2020). These insurmountable issues can dampen their enthusiasm to use M-VL. Many studies report that teaching processes have gotten more arduous during the pandemic and no amount of aesthetic appeal or fun elements to using M-VL can compensate for such experiences.
Habit has a significant positive effect on behavioral intention (p< 0.05) and significant negative effect on use behavior (p < 0.05) before COVID19 and significant negative effect on BI (-0.325, p< 0.05) and UB (-0.412, p < 0.05) during COVID19 (Table 7). Thus we accept the hypotheses H6a and H6b. The impact of HA on BI was supported by Oechslein et al. (2014) and Landis et al. (1978). Like previous results behavioral intention to use M-VLs shows a significant positive impact on teachers use behavior (p< 0.05) before and during COVID19 (Miltgen et al., 2013; Oliveira et al., 2016). Thus we confirm the hypothesis H7.
Habit was the strongest determinant of behavioral intention to use VL, followed by social influence. From our results we observe that there is initial reluctance on the part of teachers to adopt a new technology despite it contributing to significant improvements in teaching. This observation can be attributed to the relationship between behavioral intention and effort expectancy among teachers. It can also be linked to perceived usefulness of M-VL by teachers (Akram et al., 2021). However, once COVID-19 struck, it left teachers with no other choice but to adapt to usage of M-VL. Soon, by repetition, teachers found that perceived usefulness increased significantly with usage as they habitually turned to M-VL first when faced with experimental problems, clarifications of experimental methodologies and need of teaching resources. While hedonic motivation remained low, new necessity induced routines were forged during the pandemic, paving way to increased BI to use M-VL and higher reliance on M-VL as a teaching tool. The habit factor has been identified as a major influence on behavioral intention (Venkatesh et al., 2012; 2016). Past literature reports that habit and hedonic motivation have a non-linear relationship and can be the strongest predictors of BI to use new technology. In fact, over time, habit could increase hedonism for users who do not enjoy the usage of the technology. When teachers are firmly convinced of the positive experiences from M-VL, it is easier for them to accept the utilization of the technology without coercion (Kumar and Mohite, 2018).
Further, it has been observed that the willingness to adopt newer technologies is influenced by gender and age (Venkatesh et al., 2003; 2012). Traditional barriers to technology integration (such as educational level and experience, school teachers gender and age, their experience with technology in educational settings, and their views and attitudes toward technology and its use, can influence the integration of technology into the classroom environment) hinder the integration of mobile devices into the classroom (Teo et al., 2015). One can intuit that younger teachers are more welcoming of newer technologies for teaching. Up to 78% of our participants in this study were between the age range of 31 to 35 and more malleable to change. In that sense, older teachers’ ingrained or traditional teaching habits could work against the adoption of VL. Interestingly, habit remained the strongest predictor of BI to use M-VL in this study. Also, age did not negatively influence adoption and teachers of all ages developed new teaching habits during the pandemic, aside from younger teachers adapting well to M-VLs. There was no gender bias traditionally seen with technology adoption observed in this study. Close to 64% of our participants were women and they were equally inclined to adopt technology as male participants (Table 1). This result could also be attributed to the pandemic necessitating action.
The adoption of VL by teachers was also dependent on referral by their colleagues or higher authorities, implying strong SI. It shows that society, and assistance from VL experts motivates teachers and strengthens their intention to use M-VL. In the case of VL adoption by teachers, the study identified two key categories that VL providers must focus on. Both these categories are environmental factors (SI and HA) (Linder et al., 2022). While habit strengthens only with use, social influence is a significant external factor that predicts BI and is within the domain of influence of the teacher community. However, past studies report that teachers feel adequately motivated only when stakeholders such as their management and leadership along with colleagues and parents are appreciative and equally invested in the same technology (Aliyyah et al., 2020; Liu et al., 2016; Ottenbreit-Leftwich et al., 2010).
Effect of virtual lab training on teachers
One of the key reasons to delve into training pedagogy was to understand its influence on easier adoption during the pandemic. Teachers have been forced to expand their use of digital tools in their classrooms in order to prepare schools during the COVID-19 lockdown (Mishra et al., 2020; Naik et al., 2021). The adoption of any technology can be short-lived if teachers are not fully convinced of their value. Thus it becomes imperative to chronicle the training outcomes to influence sustain adoption of technologies. Prior research points to the need for a pedagogical strategy requiring teachers to integrate digital skills into their teaching practice (Aslan & Zhu, 2016; Tammaro & D’Alessio, 2016). Helping teachers to understand a technology plays a vital role in its adoption. Literature suggested that the training could fulfill teachers’ needs for ICT support, particularly for the teachers in the high confidence group (Pongsakdi et al., 2021). Another study on teachers highlighted the need for constant ICT technologies training to achieve sustainability goals and technology adaptation (Rangel-Pérez et al., 2021). One of the key issues experienced by teachers when handling distance learning classes during the pandemic relate to limited training in remote instruction (Assunção Flores & Gago, 2020) and the limited availability of resources to teach (Schleicher, 2020).
The number of M-VL teacher training sessions from 2017 to 2021 and the number of teachers who participated each year are shown in Fig. 5. In person training sessions were done prior to the pandemic and online training sessions were conducted during the pandemic. Approximately 278 and 543 training sessions were conducted yearly in-person and online respectively during the five year period. These sessions cumulatively had 2378 to 3874 teachers participating during in-person and online training sessions respectively. The data suggests a steeper increase in the average percentage growth rate of teachers’ participation in the training session 10.70% during 2020-2021 compared to 6.30% between 2017-2019. Fig. 5 Mobile virtual lab training sessions, number of teachers participated in each session and the number of institutes adopted M-VL over the past years
The training focused on upskilling teachers in three categories: course content, student evaluation, and communication with student. The usage analytics of study groups (IA, IB, and II) provides insights into the the average number of times teachers used individual features of VLs. These are reported in Table 8. The analytics suggests that the groups started using M-VL before the training (Group IA) had used an average number of approximately 3.47 ± 1.60 animations and 10.77 ± 4.21 simulations per semester for their teaching. Group IA teachers used M-VL mainly for explaining the concepts to students in their teaching. After attending the training sessions, the usage of M-VL increased amongst Group IA teacher cohort as well. During the pandemic the average number of experiment sessions per semester were 32.17 and 32.83 by Group IB and Group II respectively. The average number of communication with students were increased from 5.74 ± 3.70 to 18.67 ± 4.06 after the training sessions indicative of stronger engagement between teachers and students on the M-VL platform. Table 8 Mobile virtual lab usage statistics of teachers before and after training
Category Item Average number/semester
Group IA Group IB Group II
Course content 1 Animations 3.47 ± 1.69 9.58 ± 2.18 8.52 ± 2.80
2 Simulations 10.77 ± 4.21 14.49 ± 3.84 13.34 ± 4.60
Student evaluation 1 Assignments 1.10 14.07 13.51
2 Experiment sessions 2.51 32.17 32.83
Communication with students Communications per experiment 3.76 ± 2.59 6.10 ± 2.48 6.79 ± 3.68
Teachers’ feedback was collected before and after the training sessions, and their responses were associated to technology knowledge, technology skill, and technology access (Fig. 6). The responses suggest the crucial role of teacher training towards technology adoption. More than 70% (strongly agreed and agreed) of the participants were able to gain adequate knowledge, skills, and access to increase their use of M-VL functions effectively in their classroom through the M-VL training. The fact that teachers could use M-VL for evaluating students, communicating with students, and edit course content on-demand played a significant role in their adoption. Group IA participants are early adopters of M-VL and they started using the platform before attending the training sessions. Even though most of the participants in Group IA, have better (89%) knowledge about the different features of M-VL, they did not have enough skills (75.27%) and understanding of access to use those features in their classroom (94.04%). After attending the training sessions Group IB’s skill and their understanding of how to access the M-VL in their classroom improved significantly. Classroom management is challenging, and it’s intimidating for teachers to build lesson plans around tools they aren’t comfortable using. Training help teachers to understand how to access to VL-LMS platform and students’ assignment responses, and get skills to group students into different groups, assigning tasks to each group or specific student, using simulations or animations in classrooms, etc. Our study confirms that enough understanding about the M-VL allows teachers to effectively use the platform in their classrooms. Fig. 6 Self-reported feedback of teachers before and after the training sessions
The M-VL usage experiences of Group I and Group II were assessed by collecting teachers’ feedback online (Fig. 7). The user feedback form was provided for each experiment in the M-VL and feedback was collected on the complexity of 1) the content (theoretical notes and procedural notes) provided on the M-VL, 2) the simulated experiments, 3) the animated experiments, 4) VL-LMS platform and 5) student evaluations through the M-VL platform. The responses suggest that both Group I and II were similar in their M-VL usage experiences and more than 80% of participants selected either agree or strongly agree. Fig. 7 M-VL usage experiences of teachers
Conclusion
Using ICT tools in higher education in developing countries like India remains challenging. The successful adoption of mobile virtual labs in teaching depends on teachers’ engagement. This study has compared the factors that affect teachers’ successful adoption of mobile virtual labs using the UTAUT2 model before and during the COVID-19 pandemic. The conceptual model was empirically validated using SEM and IBM’s AMOS software. Five were accepted out of the proposed seven hypotheses, while two were rejected. Social influence and habit directly affect behavioral intention before COVID-19, which suggests that SI and HA were the strongest determinants of mobile virtual lab adoption among teachers before the pandemic. Thus, it is important for virtual lab content developers to ensure that the contents are up-to-date, easy to understand, user-friendly, and synchronous on mobile devices. However, when comparing teachers’ adoption factors during the pandemic, a drastic change was observed in the factor HA. During the pandemic forced situations, more than their habit, teachers are giving more priority to facilitating conditions and performance expectancy. Before the pandemic, performance expectancy and hedonic motivation negatively correlated with the teachers’ behavioral intention. Effort expectancy and facilitating conditions did not significantly affect teachers’ behavioral intention to adopt VL before the pandemic. In summary, the availability of high-quality learning materials and user-friendly operating interfaces are critical measures for promoting the adoption of virtual labs. Conducting workshops and faculty development programs for teachers helps to increase social influence. Hands-on sessions help to increase technology knowledge and skills, and access. Training, in turn, influences the habit of the users to adopt M-VL.
Limitations and future perspective
Although this study focuses on identifying factors that affect users’ acceptance behavior to mobile VLs, it is not without limitations. Firstly, key moderators such as gender, experience, and age were only peripherally assessed in this study. These will be examined more deeply in our further studies. Secondly, it would be interesting to cross compare platform features to technology adoption in the virtual laboratory space. Additionally contextual and cultural determinants across teachers from various disciplines would comprehensively cover the barriers for adoption of various types of virtual laboratories as well.
Acknowledgements
This work derives direction and ideas from the Chancellor of Amrita Vishwa Vidyapeetham, Sri Mata Amritanandamayi Devi. Authors would like to thank Ms. Smitha S. Murali, Ms. Sharanya Muthupalani, Ms. Parvathy S. U., Mr. Saneesh P. F. and Mr. Stevan Humphreys for their contribution to the study. Authors also acknowledge the use of Amrita Virtual Labs (https://vlab.amrita.edu) and Amrita Technology Enabling Center (TEC) (https://amritatec.in).
Funding
This work was funded by Virtual Labs project, NMEICT, Ministry of Education, Government of India.
Availability of data and materials
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Abduraxmanova SA Individualization of professional education process on the basis of digital technologies World Bulletin of Social Sciences 2022 8 65 67
AbuShanab E Pearson JM Setterstrom AJ Internet banking and customers’ acceptance in Jordan: The unified model’s perspective Communications of the Association for Information Systems 2010 26 1 23
Achuthan K Francis SP Diwakar S Augmented reflective learning and knowledge retention perceived among students in classrooms involving virtual laboratories Education and Information Technologies 2017 22 6 2825 2855 10.1007/s10639-017-9626-x
Achuthan K Kolil VK Diwakar S Using virtual laboratories in chemistry classrooms as interactive tools towards modifying alternate conceptions in molecular symmetry Education and Information Technologies 2018 23 6 2499 2515 10.1007/s10639-018-9727-1
Achuthan K Kolil VK Diwakar S Using virtual laboratories in chemistry classrooms as interactive tools towards modifying alternate conceptions in molecular symmetry Education and Information Technologies 2018 23 6 2499 2515 10.1007/s10639-018-9727-1
Achuthan, K., Nedungadi, P., Kolil, V., & et al. (2020). Innovation adoption and diffusion of virtual laboratories. International Journal of Online and Biomedical Engineering.
Achuthan K Raghavan D Shankar B Impact of remote experimentation, interactivity and platform effectiveness on laboratory learning outcomes International Journal of Educational Technology in Higher Education 2021 18 1 1 24 10.1186/s41239-021-00272-z
Adam A Lutfiyah A Mubarok H The use of virtual measuring toolkit toward students’ cognitive and procedural knowledge achievements Journal of Physics: Conference Series 2020 1491 1 012,017
Adov L Pedaste M Leijen Ä Does it have to be easy, useful, or do we need something else? stem teachers’ attitudes towards mobile device use in teaching Technology, Pedagogy and Education 2020 29 4 511 526 10.1080/1475939X.2020.1785928
Aikens ML Meeting the needs of a changing landscape: Advances and challenges in undergraduate biology education Bulletin of Mathematical Biology 2020 82 5 1 20 10.1007/s11538-020-00739-6
Akram H Aslam S Saleem A The challenges of online teaching in COVID-19 pandemic: a case study of public universities in Karachi, Pakistan Journal of Information Technology Education Research 2021 20 263 10.28945/4784
Alammary, A., Alshaikh, M., & Alhogail, A. (2021). The impact of the COVID-19 pandemic on the adoption of e-learning among academics in Saudi Arabia. Behaviour & Information Technology, 1–23.
Alger CL Secondary teachers’ conceptual metaphors of teaching and learning: Changes over the career span Teaching and Teacher Education 2009 25 5 743 751 10.1016/j.tate.2008.10.004
Aliyyah RR Rachmadtullah R Samsudin A The perceptions of primary school teachers of online learning during the COVID-19 pandemic period: A case study in Indonesia Journal of Ethnic and Cultural Studies 2020 7 2 90 109
Ameri A Khajouei R Ameri A Acceptance of a mobile-based educational application (LabSafety) by pharmacy students: An application of the UTAUT2 model Education and Information Technologies 2020 25 1 419 435 10.1007/s10639-019-09965-5
Arain AA Hussain Z Rizvi WH Extending UTAUT2 toward acceptance of mobile learning in the context of higher education Universal Access in the Information Society 2019 18 3 659 673 10.1007/s10209-019-00685-8
Artdej R Investigating undergraduate students’ scientific understanding of laboratory safety Procedia-Social and Behavioral Sciences 2012 46 5058 5062 10.1016/j.sbspro.2012.06.385
Aslan A Zhu C Influencing factors and integration of ICT into teaching practices of pre-service and starting teachers International Journal of Research in Education and Science 2016 2 2 359 370 10.21890/ijres.81048
Assunção Flores M Gago M Teacher education in times of COVID-19 pandemic in Portugal: national, institutional and pedagogical responses Journal of Education for Teaching 2020 46 4 507 516 10.1080/02607476.2020.1799709
Attuquayefio, S., & Addo, H. (2014). Using the UTAUT model to analyze students’ ICT adoption. International Journal of Education and Development using ICT, 10(3).
Bagozzi RP Yi Y On the evaluation of structural equation models Journal of the Academy of Marketing Science 1988 16 1 74 94 10.1007/BF02723327
Bagozzi RP Davis FD Warshaw PR Development and test of a theory of technological learning and usage Human Relations 1992 45 7 659 686 10.1177/001872679204500702
Ben Ouahi, M., Lamri, D., Hassouni, T., & et al. (2022). Science teachers’ views on the use and effectiveness of interactive simulations in science teaching and learning. International Journal of Instruction, 15(1).
Blumer LS Beck CW Laboratory courses with guided-inquiry modules improve scientific reasoning and experimental design skills for the least-prepared undergraduate students CBE—Life Sciences Education 2019 18 1 ar2 10.1187/cbe.18-08-0152 30681903
Bol, T. (2020). Inequality in homeschooling during the corona crisis in the netherlands. first results from the liss panel. SocArXiv.
Bolkas D Chiampi JD Fioti J First assessment results of surveying engineering labs in immersive and interactive virtual reality Journal of Surveying Engineering 2022 148 1 04021,028 10.1061/(ASCE)SU.1943-5428.0000388
Bordoloi, R., Das, P., & Das, K. (2021). Perception towards online/blended learning at the time of COVID-19 pandemic: an academic analytics in the Indian context. Asian Association of Open Universities Journal.
Brown SA Venkatesh V A model of adoption of technology in the household: A baseline model test and extension incorporating household life cycle Management Information Systems Quarterly 2005 29 3 11 10.2307/25148690
Browne MW Cudeck R Alternative ways of assessing model fit Sage Focus Editions 1993 154 136 136
Brownell, S.E., Kloser, M.J., Fukami, T., & et al. (2012). Undergraduate biology lab courses: Comparing the impact of traditionally based” cookbook” and authentic research-based courses on student lab experiences. Journal of College Science Teaching, 41(4).
Brüggemann J Bizer K Laboratory experiments in innovation research: A methodological overview and a review of the current literature Journal of Innovation and Entrepreneurship 2016 5 1 1 13 10.1186/s13731-016-0053-9
Chao CM Factors determining the behavioral intention to use mobile learning: An application and extension of the UTAUT model Frontiers in Psychology 2019 10 1652 10.3389/fpsyg.2019.01652 31379679
Chatterjee D Bolar K Determinants of mobile wallet intentions to use: The mental cost perspective International Journal of Human–Computer Interaction 2019 35 10 859 869 10.1080/10447318.2018.1505697
Chau PY Hu PJH Information technology acceptance by individual professionals: A model comparison approach Decision Sciences 2001 32 4 699 719 10.1111/j.1540-5915.2001.tb00978.x
Chin, W.W., & Todd, P.A. (1995). On the use, usefulness, and ease of use of structural equation modeling in MIS research: A note of caution. MIS Quarterly, 237–246.
Dahri NA Vighio MS Bather JD Factors influencing the acceptance of mobile collaborative learning for the continuous professional development of teachers Sustainability 2021 13 23 13,222 10.3390/su132313222
Darby-White T Wicker S Diack M Evaluating the effectiveness of virtual chemistry laboratory (VCL) in enhancing conceptual understanding: Using VCL as pre-laboratory assignment Journal of Computers in Mathematics and Science Teaching 2019 38 1 31 48
Davis, F.D. (1985). A technology acceptance model for empirically testing new end-user information systems: Theory and results. PhD thesis, Massachusetts Institute of Technology.
Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 319–340.
Davis FD Bagozzi RP Warshaw PR User acceptance of computer technology: A comparison of two theoretical models Management Science 1989 35 8 982 1003 10.1287/mnsc.35.8.982
Domínguez Alfaro JL Gantois S Blattgerste J Mobile augmented reality laboratory for learning acid–base titration Journal of Chemical Education 2022 99 2 531 537 10.1021/acs.jchemed.1c00894
Dontre AJ The influence of technology on academic distraction: A review Human Behavior and Emerging Technologies 2021 3 3 379 390 10.1002/hbe2.229
Eckhardt A Laumer S Weitzel T Who influences whom? analyzing workplace referents’ social influence on it adoption and non-adoption Journal of Information Technology 2009 24 1 11 24 10.1057/jit.2008.31
El-Soussi A The shift from face-to-face to online teaching due to COVID-19: Its impact on higher education faculty’s professional identity International Journal of Educational Research Open 2022 3 100,139 10.1016/j.ijedro.2022.100139
Farahat T Applying the technology acceptance model to online learning in the Egyptian universities Procedia-Social and Behavioral Sciences 2012 64 95 104 10.1016/j.sbspro.2012.11.012
Fauth B Decristan J Decker AT The effects of teacher competence on student outcomes in elementary science education: The mediating role of teaching quality Teaching and Teacher Education 2019 86 102,882 10.1016/j.tate.2019.102882
Fishbein, M., & Ajzen, I. (1977). Belief, attitude, intention, and behavior: An introduction to theory and research. Philosophy and Rhetoric, 10(2).
Fornell C Larcker DF Evaluating structural equation models with unobservable variables and measurement error Journal of Marketing Research 1981 18 1 39 50 10.1177/002224378101800104
Fredrick, S.S., Nickerson, A.B., & Livingston, J.A. (2022). Adolescent social media use: Pitfalls and promises in relation to cybervictimization, friend support, and depressive symptoms. Journal of Youth and Adolescence, 1–16.
Giesing, I., et al. (2005). User perceptions related to identification through biometrics within electronic business. PhD thesis, University of Pretoria.
Gormally C Brickman P Hallar B Effects of inquiry-based learning on students’ science literacy skills and confidence International Journal for the Scholarship of Teaching and Learning 2009 3 2 n2 10.20429/ijsotl.2009.030216
Granjo JF Rasteiro MG Enhancing the autonomy of students in chemical engineering education with LABVIRTUAL platform Education for Chemical Engineers 2020 31 21 28 10.1016/j.ece.2020.03.002
Gunawan G Nisrina N Suranti N Virtual laboratory to improve students’ conceptual understanding in physics learning Journal of Physics: Conference Series 2018 1108 1 012,049
Hair J Multivariate data analysis 1998 USA Prentice Hall
Hair JF Black WC Babin BJ Multivariate data analysis 6th edition. Pearson Prentice Hall New Jersey humans: Critique and reformulation Journal of Abnormal Psychology 2006 87 49 74
Harrison M Dunbar D Ratmansky L Classroom-based science research at the introductory level: Changes in career choices and attitude CBE—Life Sciences Education 2011 10 3 279 286 10.1187/cbe.10-12-0151 21885824
Herrero Á San Martín H Explaining the adoption of social networks sites for sharing user-generated content: A revision of the UTAUT2 Computers in Human Behavior 2017 71 209 217 10.1016/j.chb.2017.02.007
Hilfert-Rüppell D Looß M Klingenberg K Scientific reasoning of prospective science teachers in designing a biological experiment Lehrerbildung auf dem Prüfstand 2013 6 2 135 154
Hodges GW Wang L Lee J An exploratory study of blending the virtual world and the laboratory experience in secondary chemistry classrooms Computers & Education 2018 122 179 193 10.1016/j.compedu.2018.03.003
Hofferber N Basten M Großmann N The effects of autonomy-supportive and controlling teaching behaviour in biology lessons with primary and secondary experiences on students’ intrinsic motivation and flow-experience International Journal of Science Education 2016 38 13 2114 2132 10.1080/09500693.2016.1229074
Hu S Laxman K Lee K Exploring factors affecting academics’ adoption of emerging mobile technologies-an extended UTAUT perspective Education and Information Technologies 2020 25 5 4615 4635 10.1007/s10639-020-10171-x
Ismail SN Omar MN Don Y Teachers’ acceptance of mobile technology use towards innovative teaching in malaysian secondary schools International Journal of Evaluation and Research in Education 2022 11 1 120 127
Jena, P.K. (2020). Impact of COVID-19 on higher education in India. International Journal of Advanced Education and Research (IJAER), 5.
Jeno LM Adachi PJ Grytnes JA The effects of m-learning on motivation, achievement and well-being: A self-determination theory approach British Journal of Educational Technology 2019 50 2 669 683 10.1111/bjet.12657
Joshi, A., Vinay, M., & Bhaskar, P. (2020). Impact of coronavirus pandemic on the Indian education sector: perspectives of teachers on online teaching and assessments. Interactive Technology and Smart Education.
Kaliisa R Kluge A Mørch AI Overcoming challenges to the adoption of learning analytics at the practitioner level: A critical analysis of 18 learning analytics frameworks Scandinavian Journal of Educational Research 2022 66 3 367 381 10.1080/00313831.2020.1869082
Kalimullina O Tarman B Stepanova I Education in the context of digitalization and culture: Evolution of the teacher’s role, pre-pandemic overview Journal of Ethnic and Cultural Studies 2021 8 1 226 238 10.29333/ejecs/629
Kalogiannakis M Papadakis S Evaluating pre-service kindergarten teachers’ intention to adopt and use tablets into teaching practice for natural sciences International Journal of Mobile Learning and Organisation 2019 13 1 113 127 10.1504/IJMLO.2019.096479
Khlaif Z Teachers’ perceptions of factors affecting their adoption and acceptance of mobile technology in K-12 settings Computers in the Schools 2018 35 1 49 67 10.1080/07380569.2018.1428001
Kim C Kim MK Lee C Teacher beliefs and technology integration Teaching and Teacher Education 2013 29 76 85 10.1016/j.tate.2012.08.005
Kim, J., & Lee, K.S.S. (2020). Conceptual model to predict Filipino teachers’ adoption of ICT-based instruction in class: using the UTAUT model. Asia Pacific Journal of Education, 1–15.
Kochkorbaevna KB Gulomova O Technologies for teaching students to think independently in the process of analyzing literary texts based on an innovative approach International Journal Of Culture And Modernity 2022 13 115 120
Kolhar M Kazi RNA Alameen A Effect of social media use on learning, social interactions, and sleep duration among university students Saudi Journal of Biological Sciences 2021 28 4 2216 2222 10.1016/j.sjbs.2021.01.010 33911938
Kolil VK Muthupalani S Achuthan K Virtual experimental platforms in chemistry laboratory education and its impact on experimental self-efficacy International Journal of Educational Technology in Higher Education 2020 17 1 1 22 10.1186/s41239-020-00204-3
Kruszewska A Nazaruk S Szewczyk K Polish teachers of early education in the face of distance learning during the COVID-19 pandemic–the difficulties experienced and suggestions for the future Education 3-13 2022 50 3 304 315 10.1080/03004279.2020.1849346
Kumar BA Mohite P Usability of mobile learning applications: A systematic literature review Journal of Computers in Education 2018 5 1 1 17 10.1007/s40692-017-0093-6
Lallmahomed MZ Lallmahomed N Lallmahomed GM Factors influencing the adoption of e-Government services in Mauritius Telematics and Informatics 2017 34 4 57 72 10.1016/j.tele.2017.01.003
Landis D Triandis HC Adamopoulos J Habit and behavioral intentions as predictors of social behavior The Journal of Social Psychology 1978 106 2 227 237 10.1080/00224545.1978.9924174 28135539
Lederman, D. (2019). Professors’ slow, steady acceptance of online learning: A survey. Inside Higher Ed.
Li M Yu Z Teachers’ satisfaction, role, and digital literacy during the COVID-19 pandemic Sustainability 2022 14 3 1121 10.3390/su14031121
Li, Y., Li, J., Fan, Q., & et al. (2022). Cybercrime’s tendencies of the teenagers in the COVID-19 era: assessing the influence of mobile games, social networks and religious attitudes. Kybernetes.
Limayem, M., Hirt, S.G., & Cheung, C.M. (2007). How habit limits the predictive power of intention: The case of information systems continuance. MIS Quarterly, 705–737.
Linder N Giusti M Samuelsson K Pro-environmental habits: An underexplored research agenda in sustainability science Ambio 2022 51 3 546 556 10.1007/s13280-021-01619-6 34519955
Liu L Ling Y Gao Q Supporting students’ inquiry in accurate precipitation titration conditions with a virtual laboratory tool as learning scaffold Education for Chemical Engineers 2022 38 78 85 10.1016/j.ece.2021.11.001
Liu, W.C., Wang, J.C.K., & Ryan, R.M. (2016). Understanding motivation in education: Theoretical and practical considerations. In W.C. Liu, J.C.K. Wang, & R.M. Ryan (Eds.). Building autonomous learners: Perspectives from research and practice using self-determination theory (pp. 1–7). Singapore: Springer Singapore.
Looney, C.A., Valacich, J.S., & Akbulut, A.Y. (2004). Online investment self-efficacy: Development and initial test of an instrument to assess perceived online investing abilities. In Null (p. 70184b). IEEE.
Malewar, S., & Bajaj, S. (2020). Acceptance of OTT video streaming platforms in India during COVID-19: Extending UTAUT2 with content availability. Journal of Content Community and Communication, 89–106.
Manzoor SR Mohd-Isa WN Dollmat KS Post-pandemic e-learning: A pre-protocol to assess the integration of mobile VR and its effect on VARK learning styles F1000Research 2021 10 1106 1106 10.12688/f1000research.73311.1 35646326
Mari E Lausi G Fraschetti A Teaching during the pandemic: A comparison in psychological wellbeing among smart working professions Sustainability 2021 13 9 4850 10.3390/su13094850
Miltgen CL Popovič A Oliveira T Determinants of end-user acceptance of biometrics: Integrating the “big 3” of technology acceptance with privacy context Decision Support Systems 2013 56 103 114 10.1016/j.dss.2013.05.010
Mishra L Gupta T Shree A Online teaching-learning in higher education during lockdown period of COVID-19 pandemic International Journal of Educational Research Open 2020 1 100,012 10.1016/j.ijedro.2020.100012
Morosan C DeFranco A It’s about time: Revisiting UTAUT2 to examine consumers’ intentions to use NFC mobile payments in hotels International Journal of Hospitality Management 2016 53 17 29 10.1016/j.ijhm.2015.11.003
Naik, G.L., Deshpande, M., Shivananda, D., & et al. (2021). Online teaching and learning of higher education in India during COVID-19 emergency lockdown. Pedagogical Research, 6(1).
Nandwani S Khan S Teachers’ intention towards the usage of technology: An investigation using UTAUT model Journal of Education & Social Sciences 2016 4 2 95 111 10.20547/jess0421604202
Ngugi, J.K., & Goosen, L. (2021). Innovation, entrepreneurship, and sustainability for ICT students towards the post-COVID-19 era. In C Luisa Cagica, L Reis, & C Silveira (Eds.) Handbook Of Research On Entrepreneurship, Innovation, Sustainability, and ICTs in the Post-COVID-19 Era (pp. 110–131). USA: IGI Global.
Nikolopoulou K Gialamas V Lavidas K Acceptance of mobile phone by university students for their studies: An investigation applying UTAUT2 model Education and Information Technologies 2020 25 5 4139 4155 10.1007/s10639-020-10157-9
Nikolopoulou K Gialamas V Lavidas K Habit, hedonic motivation, performance expectancy and technological pedagogical knowledge affect teachers’ intention to use mobile internet Computers and Education Open 2021 2 100,041 10.1016/j.caeo.2021.100041
Nikolopoulou K Gialamas V Lavidas K Teachers’ readiness to adopt mobile learning in classrooms: A study in Greece Technology, Knowledge and Learning 2021 26 1 53 77 10.1007/s10758-020-09453-7
Noble DF Digital diploma mills: The automation of higher education Science as Culture 1998 7 3 355 368 10.1080/09505439809526510
O’Bannon BW Thomas KM Mobile phones in the classroom: Preservice teachers answer the call Computers & Education 2015 85 110 122 10.1016/j.compedu.2015.02.010
Oechslein, O., Fleischmann, M., & Hess, T. (2014). An application of UTAUT2 on social recommender systems: Incorporating social information for performance expectancy. In 2014 47th Hawaii international conference on system sciences (HICSS) (pp. 3297–3306). IEEE.
Oliveira T Thomas M Baptista G Mobile payment: Understanding the determinants of customer adoption and intention to recommend the technology Computers in Human Behavior 2016 61 404 414 10.1016/j.chb.2016.03.030
Omar M Ismail S Kasim A The influence of mobile technology adoption among secondary school teachers using the UTAUT 2 model International Journal of Recent Technology and Engineering 2019 8 4 3827 3831
Osman ME Global impact of COVID-19 on education systems: The emergency remote teaching at Sultan Qaboos University Journal of Education for Teaching 2020 46 4 463 471 10.1080/02607476.2020.1802583
Ottenbreit-Leftwich AT Glazewski KD Newby TJ Teacher value beliefs associated with using technology: Addressing professional and student needs Computers & Education 2010 55 3 1321 1335 10.1016/j.compedu.2010.06.002
Peluchette JV Rust KA Technology use in the classroom: Preferences of management faculty members Journal of Education for Business 2005 80 4 200 205 10.3200/JOEB.80.4.200-205
Petrie, C., Aladin, K., Ranjan, P., & et al. (2020). Spotlight: Quality education for all during COVID-19 crisis. https://hundred.org/en/research.
Pickett LL Ginsburg HJ Mendez RV Ajzen’s theory of planned behavior as it relates to eating disorders and body satisfaction North American Journal of Psychology 2012 14 2 339
Pongsakdi N Kortelainen A Veermans M The impact of digital pedagogy training on in-service teachers’ attitudes towards digital technologies Education and Information Technologies 2021 26 5 5041 5054 10.1007/s10639-021-10439-w
Pratama H Azman MNA Kassymova GK The trend in using online meeting applications for learning during the period of pandemic COVID-19: A literature review Journal of Innovation in Educational and Cultural Research 2020 1 2 58 68 10.46843/jiecr.v1i2.15
Raja, M., & Priya Lakshmi, G. (2022). Using virtual reality and augmented reality with ICT tools for enhancing quality in the changing academic environment in COVID-19 pandemic: An empirical study. In Technologies, artificial intelligence and the future of learning post-COVID-19 (pp. 467–482). Cham: Springer.
Raman A Don Y Preservice teachers’ acceptance of learning management software: An application of the UTAUT2 model International Education Studies 2013 6 7 157 164 10.5539/ies.v6n7p157
Raman A Don Y Khalid R Usage of learning management system (Moodle) among postgraduate students: UTAUT model Asian Social Science 2014 10 14 186 10.5539/ass.v10n14p186
Rangel-Pérez C Gato-Bermúdez MJ Musicco-Nombela D The massive implementation of ICT in universities and its implications for ensuring SDG 4: Challenges and difficulties for professors Sustainability 2021 13 22 12,871 10.3390/su132212871
San Martín H Herrero Á Influence of the user’s psychological factors on the online purchase intention in rural tourism: Integrating innovativeness to the UTAUT framework Tourism Management 2012 33 2 341 350 10.1016/j.tourman.2011.04.003
Schleicher, A. (2020). The impact of COVID-19 on education: Insights from education at a glance 2020. Retrieved from OECD ORG website: https://www.oecdorg/education/the-impact-of-covid-19-on-education-insights-education-at-a-glance-2020.pdf.
Selvaraj A Radhin V Nithin K Effect of pandemic based online education on teaching and learning system International Journal of Educational Development 2021 85 102,444 10.1016/j.ijedudev.2021.102444
Selwyn N Aagaard J Banning mobile phones from classrooms—an opportunity to advance understandings of technology addiction, distraction and cyberbullying British Journal of Educational Technology 2021 52 1 8 19 10.1111/bjet.12943
Setyawan A Aznam N Hw P Effects of the Google Meet assisted method of learning on building student knowledge and learning outcomes Universal Journal of Educational Research 2020 8 9 3924 3936 10.13189/ujer.2020.080917
Shah, D. (2020). Highlights from Coursera partners conference 2020. The Report by Class Central, 10(3371382.3377437).
Shrestha S Haque S Dawadi S Preparations for and practices of online education during the COVID-19 pandemic: A study of Bangladesh and Nepal Education and Information Technologies 2022 27 1 243 265 10.1007/s10639-021-10659-0 34341654
Suki NM Suki NM Determining students’ behavioural intention to use animation and storytelling applying the UTAUT model: The moderating roles of gender and experience level The International Journal of Management Education 2017 15 3 528 538 10.1016/j.ijme.2017.10.002
Tammaro R D’Alessio A Teacher training and digital competence: A pedagogical recommendation International Journal of Digital Literacy and Digital Competence (IJDLDC) 2016 7 2 1 10 10.4018/IJDLDC.2016040101
Teo T Modelling technology acceptance in education: A study of pre-service teachers Computers & Education 2009 52 2 302 312 10.1016/j.compedu.2008.08.006
Teo T Lee CB Chai CS Understanding pre-service teachers’ computer attitudes: applying and extending the technology acceptance model Journal of Computer Assisted Learning 2008 24 2 128 143 10.1111/j.1365-2729.2007.00247.x
Teo, T., Fan, X., & Du, J. (2015). Technology acceptance among pre-service teachers: Does gender matter? Australasian Journal of Educational Technology, 31(3).
Testa N Tawfik A Mobile, but are we better? understanding teacher’s perception of a mobile technology integration using the unified theory of acceptance and use of technology (UTAUT) framework Journal of Formative Design in Learning 2017 1 2 73 83 10.1007/s41686-017-0010-4
Thomas, T., Singh, L., & Gaffar, K. (2013). The utility of the UTAUT model in explaining mobile learning adoption in higher education in Guyana. International Journal of Education and Development using ICT, 9(3).
Tseng TH Lin S Wang YS Investigating teachers’ adoption of MOOCs: the perspective of UTAUT2 Interactive Learning Environments 2022 30 4 635 650 10.1080/10494820.2019.1674888
Varghese J Musthafa MMA Online teaching: Perceptions, challenges and opportunities during and beyond COVID-19 pandemic among secondary school teachers Indonesian Journal of Contemporary Education 2022 4 1 1 7 10.33122/ijoce.v4i1.19
Venkatesh V Bala H Technology acceptance model 3 and a research agenda on interventions Decision Sciences 2008 39 2 273 315 10.1111/j.1540-5915.2008.00192.x
Venkatesh V Davis FD A theoretical extension of the technology acceptance model: Four longitudinal field studies Management Science 2000 46 2 186 204 10.1287/mnsc.46.2.186.11926
Venkatesh, V., Morris, M.G., Davis, G.B., & et al. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 425–478.
Venkatesh, V., Thong, J.Y., & Xu, X. (2012). Consumer acceptance and use of information technology: Extending the unified theory of acceptance and use of technology. MIS Quarterly, 157–178.
Venkatesh V Thong JY Xu X Unified theory of acceptance and use of technology: A synthesis and the road ahead Journal of the Association for Information Systems 2016 17 5 328 376 10.17705/1jais.00428
Verkijika, S.F. (2018). Factors influencing the adoption of mobile commerce applications in Cameroon. Telematics and Informatics.
Wong SM Leong CM Puah CH Mobile internet adoption in Malaysian suburbs: The moderating effect of gender Asian Journal of Business Research 2020 9 3 90 114 10.14707/ajbr.190069
Yang S Understanding undergraduate students’ adoption of mobile learning model: A perspective of the extended UTAUT2 Journal of Convergence Information Technology 2013 8 10 969 10.4156/jcit.vol8.issue10.118
Yu H Liu P Huang X Teacher online informal learning as a means to innovative teaching during home quarantine in the COVID-19 pandemic Frontiers in Psychology 2021 12 2480
Yusuf, I., & Widyaningsih, S.W. (2020). Implementing e-Learning-based virtual laboratory media to students’ metacognitive skills. International Journal of Emerging Technologies in Learning, 15(5).
Zydney JM Warner Z Mobile apps for science learning: Review of research Computers & Education 2016 94 1 17 10.1016/j.compedu.2015.11.001
| 0 | PMC9742037 | NO-CC CODE | 2022-12-14 23:29:59 | no | Educ Inf Technol (Dordr). 2022 Dec 12;:1-34 | utf-8 | Educ Inf Technol (Dordr) | 2,022 | 10.1007/s10639-022-11499-2 | oa_other |
==== Front
Educ Inf Technol (Dordr)
Educ Inf Technol (Dordr)
Education and Information Technologies
1360-2357
1573-7608
Springer US New York
11479
10.1007/s10639-022-11479-6
Article
Towards intelligent E-learning systems
http://orcid.org/0000-0002-8245-2355
Liu Mengchi [email protected]
1
Yu Dongmei [email protected]
2
1 grid.263785.d 0000 0004 0368 7397 Guangzhou Key Laboratory of Big Data and Intelligent Education School of Computer Science South China Normal University Guangzhou, Guangdong, 510631 China
2 grid.450322.2 0000 0004 1804 0174 Shanghai Astronomical Observatory Chinese Academy of Sciences, Shanghai, 200030 China
12 12 2022
132
8 7 2022
16 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 prevalence of e-learning systems has made educational resources more accessible, interactive and effective to learners without the geographic and temporal boundaries. However, as the number of users increases and the volume of data grows, current e-learning systems face some technical and pedagogical challenges. This paper provides a comprehensive review on the efforts of applying new information and communication technologies to improve e-learning services. We first systematically investigate current e-learning systems in terms of their classification, architecture, functions, challenges, and current trends. We then present a general architecture for big data based e-learning systems to meet the ever-growing demand for e-learning. We also describe how to use data generated in big data based e-learning systems to support more flexible and customized course delivery and personalized learning.
Keywords
E-learning system
Big data
Smart learning
Personalized earning
General Research Plan61672389 Liu Mengchi Guangzhou Key Laboratory of Big Data and Intelligent Education201905010009 Liu Mengchi
==== Body
pmcIntroduction
The fast development of e-learning systems has radically transformed the way in which learning resources are imparted to students. They make educational resources more accessible, interactive and effective to learners without the geographic and temporal boundaries. E-learning (Alonso et al., 2005) has been defined as the use of information and communication technologies to improve the quality of learning by enabling access to resources and services, as well as remote exchange and collaboration. Some synonymous terms including open learning, distance learning, web-based education and online learning have been alternatively used over the past decades. In general, it is considered as an educational process that enables transfer of knowledge and skills flexibly to a large number of recipients at various times and locations. The combination of education and technologies provides a new way for people to learn in the era of information and communicationtechnology.
However, it has become challenging to provide excellent e-learning services as modern e-learning applications are increasingly data-intensive due to the following reasons: the number of learners and courseworks increases dramatically as e-learning applications get more and more popular, especially during the covid pandemic;
different roles generate a huge amount of interactive information when posting or exchanging messages;
the diversity of resources make each type of information isolated;
a variety of personal information and sensitive data need related access control and security policies;
the gigantic amounts of data need to be stored and managed properly.
As a result, e-learning systems need to evolve to provide smart services. In this context, intelligent technologies have been gradually used to collect, preprocess, analyze, store, and visualize huge amount of data from various learning sources. They are utilized to eliminate noise and extract valuable information to improve the effectiveness of e-learning. Also, they enable learning resources to be tailored for each individual learner according to the contents and learners’ interactive behaviors (Gamalel-Din, 2010; Kumar et al., 2016; Hu et al., 2020). In other words, precise customization and personalization of knowledge or services should be provided to each individual learner accordingly. Consequently, smart e-learning systems make it possible for educational organizations to offer improved teaching and learning deliveries, thus deal with the challenges in current e-learning services.
This paper mainly reviews the studies that apply various technologies to e-learning systems and hence provide personalized and precise teaching and learning services. We first systematically investigate the status of current e-learning systems in terms of their features, classifications, architecture, challenges and trends. Then we present big data based e-learning systems for more flexible course delivery and personalized learning, and also show how data can be processed to facilitate learning and teaching. Finally we discuss the beneficial effects of integrating big data technology with e-learning systems.
E-learning system classifications
E-learning systems facilitate the planning, management, and delivery of content for e-Learning. Based on the target users and the cost, we classify current e-learning systems into two kinds: Massive Open Online Course (MOOC) platforms and Learning Management Systems (LMSs). Firstly, MOOC platforms are open to a large number of individuals who are intended to learn. Even though some courses are produced by certain universities, they are not limited to student in post-secondary institutions. They can be accessed by people regardless of their location, culture, nationality, and any other criteria. On the other hand, LMSs are usually implemented for post-secondary institutions. Thus, they are not by default open to the general public, only a certain group of people can have access to it. Secondly, most MOOCs are free of cost or cost little so that individuals can afford, while the cost for LMS is higher based on the number of users and usually borne by post-secondary institutions rather than individuals.
Massive open online courses (MOOCs) platforms
MOOCs are online courses open to unlimited participants that are offered by many universities and institutions on the web. The term was firstly coined in 2008 by Canadian researchers Dave Cormier and Brian Alexander (Goldie, 2016). In fall of 2011, Stanford University offered the first MOOC course, which was originally registered by more than 160,000 learners from all over the world, and eventually about 20,000 of them completed it. Then, three significant MOOCs platforms Udacity1, edX2, and Coursera3 were developed in 2012 and used to offer MOOCs for free. Since then, the number of available MOOCs and MOOC learners increased dramatically, with more online platforms available. By the end of 2020, about 950 universities around the world launched 16,300 MOOCs with 180 million MOOC learners (Shah, 2020).
Generally, there are two kinds of MOOCs based on different learning theories: cMOOCs (the connectivist MOOCs) and xMOOCs (extended MOOCs) (Alonso et al., 2005). cMOOCs mainly emphasize connection and promote interaction by digital tools like blogs, learning communities and social media platform. Learners can also create and generate knowledge by themselves. xMOOCs are based on traditional university courses by focusing more on the content and deliver knowledge in small units so that the number of students can be increased significantly. Currently, the most popular and influential MOOCs providers are Coursera, edX and Udacity.
In addition, various national online platforms have emerged in a number of countries (Shah, 2021), including FutureLearn in Great Britain4, XuetangX in China5, France Université Numérique (FUN) in France6, OpenHPI in German7, EduOpen in Italy8, SWAYAM in India9, gacco in Japan10, ThaiMOOC in Thailand11, the National Open Education Platform (NOEP) in Russia12, etc. In summary, these MOOCs are open, participatory and distributed (Baturay, 2015). They have the potential to disrupt the traditional education due to their easy accessibility and free or low-cost content delivery, especially considering educational credentials including micro-credentials, specializations or degrees from accredited institutions (Pickard et al., 2018).
Learning management systems (LMSs)
LMSs are e-learning systems for hosting, assigning, managing, reporting and evaluating e-learning courses. Many postsecondary education institutions adopt LMSs as critical educational tools to support course management and to foster interaction among students, teachers and content resources. They are also used to identify training and learning gaps, implementing a wide range of pedagogical methods to promote education process. LMSs are generally classified as commercial and non-commercial systems. Commercial LMSs like WebCT, Blackboard, D2L Brightspace have been frequently and very successfully used in the past decades. They are basically easy to deploy and use, and technical support services are provided without additional costs. However, some non-commercial LMSs such as Moodle13, Canvas14, Open edX15 and Sakai 16 also become popular recently. The open source feature of non-commercial LMSs makes them attractive since they are easy to obtain, as many are free, especially those that provide a basic level of service. They also provide more flexible and scalable architecture to meet users’ needs. Generally, the most successful LMSs in North America are the Big Four (Hill, 2019): Blackboard, Canvas, D2L Brightspace and Moodle.
Architecture of current E-learning systems
Many conventional frameworks have been used to create and improve e-learning system effectiveness. One is based on service-oriented architectures (SOA) that allow to easily extend the capabilities and functionalities of the system by dynamically adding services. For example, Fajar et al. (2018) present a SOA system architecture and reference for an e-learning system, which consists of six components: data layer, resource layer, application layer, business process layer, presentation layer and governance layer. The data of each layer are treated as the service in the SOA system, which makes it more reusable, flexible and accessible to extended tools. This method allows business and wider society to improve e-learning and offer affordable education. Similarly, (González et al., 2009) extend existing e-learning systems to external mobile scenarios based on SOA as well. The architecture ensures the independence of e-learning systems, mobile applications and external applications, and provides a reliable data exchange and interoperability between them. Furthermore, (Kappe & Scerbakov, 2017) present an innovative object-oriented architecture for the implementation of e-learning systems on a single software platform to meet the requirements of various e-learning scenarios. Abstract data objects (ADOs) that encapsulate private memory together with some methods are widely used as the main components of an functional objects like courses, announcements, curriculum and so on. The implementation shows that this architecture is highly modular since documents and objects can be created independently but also re-used through a flexible nesting or containment mechanism.
Recently, the availability of high speed networks, low-cost computers and storage devices has resulted in the significant advances in the cloud computing technology, which is the on-demand usage of a network of remote servers hosted on the internet to store, manage, and process data, rather than on one or more local servers. (Riahi, 2015) reviewed recent cloud-based systems and proposed an e-learning cloud architecture, which includes hardware resource layer, software resource layer, resources management layer, service layer and business application layer. They also conclude several advantages of cloud based e-learning like low cost, improved performance and compatibility, information security and benefits for both students and students. Other researches (El Mhouti et al., 2018; Masud & Huang, 2012; Riahi, 2015; Sun et al., 2015; Chao et al., 2015; Hendradi et al., 2020; Rani et al., 2015) also focus on combining e-learning systems with cloud computing. There are two advantages in doing so. Firstly, it is easy to create and maintain, and the investment cost can be reduced significantly using the pay-as-you-go method. Also, it allows to scale the services according to the need. (Sun et al., 2015) introduce a cloud-based virtual learning environment called MLaaS, which aims to provide adaptive micro learning contents and customized learning route for every single learner. Education data mining scheme is used to discover features of learning resources and understand learners’ behaviors. In addition, (Chao et al., 2015) propose a cloud-based ecosystem called CLEM for teachers and learners. Their implementation shows that the cloud-based platform gathers heterogeneous and distributed devices in a common pool that makes computational resources more accessible and sharable. Furthermore, (Jeong et al., 2013) introduce a private-cloud-based e-learning system with six components: a private cloud platform, an XML based common file format, an authoring tool, a content viewer, an inference engine and a security system. By using these components, it can deliver and share various types of educational resources effectively. However, according to some literatures (El Mhouti et al., 2018; Laisheng & Zhengxia, 2011), the challenges of cloud-based e-learning system are mainly related to cloud privacy, security and confidence. At the same time, these concerns also provide opportunities for e-learning promotion and development in cloud computing environment.
Based on previous researches, we conclude a general framework for current e-learning systems in Fig. 1, which basically consists of three logical layers contributing to better teaching and learning effectiveness: presentation layer, e-learning system layer and data layer. Fig. 1 A General Framework for E-learning Systems
The presentation layer
The presentation layer focuses on the human computer interaction by providing accessible user interface and learning resources to end users. It aims at improving the usability, accessibility, credibility and the user experience of the learning ecosystems (García-Holgado & García-Peñalvo, 2018). Firstly, it provides a unified interface for all the services or functionalities provided by the lower layers and hide system complexity from users. Users can utilize this interface to construct and control the contents of e-learning systems. The feedback from the e-learning system layer is delivered through this interface. Secondly, due to the prevalence of various mobile devices (e.g. mobile phones, laptops, tablets and other portable devices), the presentation layer should ensure that e-learning systems support mobile learning paradigm (Schuck et al., 2017). In other words, e-learning systems are adaptable to distinct screen sizes, which allows learners to gain any information flexibly. Also, mobile learning has been proved to be able to improve student participation and engagement during learning process, while learners have high levels of motivation and satisfaction (Cheng et al., 2015). It also has a positive influence on learners’ academic performance (Han & Shin, 2016). Therefore, it is necessary to use proper front-end techniques such as HTML, XHTML, CSS, and JavaScripts to support mobile learning, in which presented pages can be rendered properly on a browser to meet the compatibility requirements of devices.
The E-learning system layer
The e-learning system layer aims at synthesizing educational resources by way of various functions such as course enrolment and management, user profile and activities, teaching or learning assessment and feedback, user communication or collaboration and so forth. It can also be an integration of related components which support instructional model or learning model (Lu et al., 2015). Users are able to choose the components to satisfy the different needs for teaching and learning. For most MOOCs and LMSs, this layer plays an important role between the presentation layer and the database layer. Learning and teaching information including users profile, learning resources, teaching and learning activities is collected and passed through e-learning system layer. It is also a teaching and learning platform that enable each learner to access specific education resources flexibly.
The database layer
The database layer hosts data generated by using e-learning systems. It is the critical place where education data is collected, stored and used. Due to the individual differences, collecting the massive data and retaining the diversity and dynamic features is very important. Additionally, all the collected data need to be stored until their use. Alternatively, some processed results may be put to use immediately, while most of them will serve some purposes later on. The main benefit is that it enables the collected or processed data to be accessed and retrieved easily.
Usually, existing solutions for e-leaning storage mainly rely on relational database, such as Mysql (Wangmo & Ivanova, 2017) and Oracle (Datta & Bhattacharyya, 2018). Moodle’s database is typically MySQL or Postgres, and can also be Microsoft SQL Server or Oracle. Sakai and Blackboard can be deployed in a SQL or Oracle environment as well.
Also, NoSQL databases are increasingly used for large sets of distributed data due to flexible and scale-out architecture. They work as a complementary technology for the relational databases system and are suitable for distributed applications with the demand of high data scalability and availability (Davoudian et al., 2018). For example, MongoDB is choosed by Open edX for storing large files which are text files, PDFs, audio/video clips, etc.
Additionally, distributed storage technology is increasingly used to replace traditional local storage. Some run on top of file systems while others work as standalone systems. For example, (Zhang et al., 2020) use distributed storage technologies for experimental education systems. Specifically, the interplanetary file system (IPFS), an external storage server and external cloud storage are combined for storage management. Among them, IPFS determines the overall performance of the storage module and contributes to system reliability and flexibility. Additionally, a file table is defined to manage each learning content such as documents, video and problem books in a distributed database. (Kawato et al., 2020) create an e-learning system by implementing Apache Cassandra, which is an open-source distributed database system to handle large volumes of data. By combining distributed hash tables (DHT), which hold information of the connected computer nodes, it is able to share various education resources spanning multiple servers. (Otoo-Arthur & van Zyl, 2020) present a framework on a distributed and parallel computing environment to provide new value to teaching and learning process.
Moreover, cloud storage as a large scale distributed storage paradigm is also used to education system, in which learning resources is stored on remote storage systems. Compared with traditional storage ways, it has many advantages in terms of scalability, flexibility, safety, ease of use and cost saving. (Sun et al., 2015) deploy Mobile MOOC learning on the Amazon EC2, and Amazon S3 is considered as the MOOC learner and course data storage because of its robustness and mature disaster recovery mechanisms. (Jeong et al., 2013) propose a content-oriented smart education system based on a small-scale, private cloud. A common file format based on XML are defined as a means of representing data and meta - data. The Document Type Definition (DTD) and the eXtensible Style sheet Language (XSL) are used to described the schema and styles for the XML document structure seperately, which enables the same content can be viewed on multiple devices. Furthermore, (Rani et al., 2015) deployed e-learning system on remote cloud host, where all required learning resources are stored. They also build a simple MySOL on the cloud host for authentication of the system. By doing so, an expanded and secure environment is built to raise e-learning system.
Functions of E-learning systems
The functions of e-learning systems depend on its potential usage such as system scale requirement, organizational objectives, online training strategy and desired pedagogical outcomes. (Cavus & Zabadi, 2014) summarize that standard LMSs should have various tools for e-learning systems. They compare six popular open source LMSs in terms of video services, discussion forum, file exchange, email, realtime chat and so forth, and discover that communication tools provided by Moodle and ATutor are efficient, but it is not easy to obtain information on Claroline and Sakai due to their complex webpages. Similarly, (Chung et al., 2013) suggest that LMSs should have five components: transmitting course content, creating a discussion, evaluating students, evaluating courses and instructors, and creating computer-based instruction. However, most of the existing e-learning systems do not contain all the features in a single system. So, we highlight the general function components (Fig. 2) that most e-learning systems have to support teaching and learning process. Fig. 2 Functions of E-learning Systems
System administration
This module includes a full range of functions for the management and configuration of system parameters and attributes in terms of users, courses grades, appearance reports and so forth. It covers components such as user authentication, user management, roles and permission management, customizable preferences, log and report management, calendar and appointment scheduler. User management: For e-learning users, email or mobile phone-based self-registration authentication method is commonly adopted to fulfill user authentication. Also, a category hierarchy is usually built to organize users from different organizations. Common operations such as adding, deleting, modifying and querying related to user management are supported. Normally, there are several categories of e-learning users including administrators, instructors, teaching assistants (TAs) and learners. The administrators set up and configure the system. The instructors prepare the lessons and access the learners’ progress. TAs assist instructors with instructional responsibilities. Learners are anyone interested in learning and being educated in the courses.
Roles and permissions: E-learning systems usually support several standard user roles and has the potential to create an unlimited number of additional roles. Therefore, it is necessary to control users’ access rights only to the information they need and to prevent them from accessing information that does not pertain to them. For example, role based access control (RBAC) is used to control client access and consents in DidaTec LMSs platform (Laura et al., 2018). Another example is that the Access Control List (ACL) is used to maintain the user information and their permissions, while group key is utilized to secure course materials and to ensure that only approved participants have access to it (Kanimozhi et al., 2019).
Customizable preferences: Personalized setting for user profiles and system preferences such as privacy, design and layout of the websites is allowed to enhance users’ experience.
Log and report management: Event log analysis is displayed through graphical user interface to assist teaching and learning. Also, analysis reports are available and exported to help administrators and teachers to make decisions based on the statistical results.
Calendar: It displays a consolidated view of all the course-related events by day, week, or month in the e-learning system. It allows users to view the available learning programs or courses with specific due dates. Also, the calendar usually automatically synchronizes with other teaching or learning activities such as syllabus, assignments, tests, and grades. In the case that users create, change, or delete the date of an activity in the LMS, the change will show up in calendar and vice versa. Finally, systems such as Moodle and Coursera also allow users to export calendar, so they may be imported into other calendar programs, as a backup or to create a copy. Conversely, other agendas can also be imported to calendars in LMSs to facilitate time management in some university (Mei, 2016).
Appointment scheduler: It helps teachers schedule appointments with their students. Teachers specify time slots and locations for online or offline activities and students choose for their attendance.
Course management
It is a basic but most important component of an e-learning system (Cavus & Zabadi, 2014), which mainly refers to create, organize and deliver various coursework. Most LMSs allow users to add course material from various sources in different formats such as text, graphics, audio, video and so forth. Platforms like Moodle, Open edX allow to use the SCORM (Shareable Content Object Reference Model) standard for its online courses. The benefit is that it provides a standardized course model that supports the reusability of learning objects. For example, multiple individual lessons can be stringed together into a complete course. Participants are also encouraged to have more interactivity within e-learning systems. With the proper authoring tool, they can create their own courses and eliminate the need to outsource course development. Similarly, (Gamalel-Din, 2010) tailor course materials by drawing multimedia Learning Objects(LO) from LO repositories, which are composed of small granular multimedia objects. This idea helps teachers to find the best available assets and LOs for course design. Students are also able to get tailored learning strategy based on their abilities and previous knowledge. Basically, the specific functions of course management are as follows. Participants management: It enables an administrator or teacher to easily enrol, view, filter, edit and delete participants for each course, and also group participants or invite learners. It is also a centralized place where teachers are able to trace student’s attendance, increase student enrollment and avoid high drop-out rates during the course. By comparing user activity and identifying attendance trends, regular attendance of all students is recorded and ready for further analysis. Since student attendance is strongly linked to learning outcome, it is also necessary for teacher to give a warning to those with poor attendance during online learning.
Contents management: Contents are organized in descriptive categories so that users can easily find their desired resources. Both static contents and interactive resources are delivered according to students needs. Some contents might be made either for a restricted audience or for a wider population, either as a free offering or as paid courses. Generally, a LMS allows course creators to freely structure their e-learning offerings in a manner that best fits their purposes and requirements. Also, instructors can trace the progress of each course and adjust their pedagogical strategy accordingly.
Gradebook: It is a central location where teachers can manage grades for courses and track student activities relative to gradable items. It plays an critical role in performance monitoring and feedback seeking associated with self-paced learning practices.
Exercise and Assessment
This module utilizes some testing and evaluation capabilities to monitor, track and evaluate the effectiveness of the e-learning process. Most e-learning systems support learning assessments periodically and some of them even support the teachers to identify gaps or intervene when necessary. Generally, a broad range of e-learning assessment methods are considered in terms of learners’ progress and performance. Some offer built-in auto-graded evaluation tools (Baturay, 2015), such as quizzes, tests, assignment, group exercises, examinations and surveys so that both instructors and students can track the learning performance in gradebook easily. Some even have diagnostic assessments to evaluate the level of knowledge of learners and assign suitable level to them. Furthermore, peer assessment (Lynda et al., 2017) is widely used in MOOC platforms which involves learners in grading and giving feedback from the work of their peers. It is also recognized as one important feature that affects the effectiveness of e-learning systems. For example, Coursera has regarded peer review as a scalable and sustainable way to guide students in assessing each other’s job as well as providing feedback. Lastly, evaluation reports used to assess e-learning are generated to query and display data in graphs and charts, allowing users to easily spot teaching or learning trends or issues. Additionally, this report should reflect the user performance on both individual and group level from multiple perspectives and monitor if the learners achieve their required objectives. Generally, the following activities are normally used to perform assessment and feedback. Assignment: It allows learners to submit their work online and teachers to grade and give response. Teachers are allowed to select excellent assignments to share with all students enrolled in the courses.
Test and examination: It is necessary to assess course quality and learning outcomes. Teachers are allowed to create quizzes that are made up of a wide range of questions derived from a question bank. This enables a question to be re-used in different quizzes and facilitate the teaching process. Examinations are also conducted online to assess student performance. Furthermore, a remote proctor or students’ webcam can be used to monitor the student’s activities and the surroundings during the examination, which is an effective solution to maintain academic integrity for e-learning examination.
Survey: It is used to help teachers to gather information from students and reflect on their own teaching. It is also used to identify certain trends that may be happening among course participants.
Workshop: It allows the learners to perform peer assessment activity according to teacher’s guidance.
Collaboration and communication
Effective collaboration and communication help the transfer, sharing and co-construction of knowledge as well as the sharing of experiences in teacher-student and peer-to-peer relationships (Chiu & Hew, 2018). There are various communication tools in existing e-learning systems that encourage participants to support each other in the learning process. Live chat & video: Live chat is an instant messaging application that allows users to discuss in real time while they participate in the teaching and learning process. It usually supports features such as real time chat monitor, chat history, file sharing and so forth. A platform-independent and web-based instant messaging can be embedded to support convenient communication among users. The chat tool (Bagarukayo et al., 2014; Carmona et al., 2008) can be integrated as a synchronous, live communication way to aid interaction and collaboration among e-learning users. It allows students and teachers to interact in real time, such as conducting group discussions or study sessions effectively. Live video tools such as Microsoft Teams, Skype, Zoom and Tencent Meeting are increasing used as communication tools for e-learning especially during COVID-19 pandemic. According to (Alameri et al., 2020), 80.7% of participants agree that Moodle, Microsoft teams and Zoom platforms enhance the communication between teachers and students in higher education. Learners are easy to concentrate on classes by constructing visual presentations. Most students believe Moodle, Microsoft teams and Zoom platforms are critical for them to handle learning process and they will be an indispensable part for online learning.
Forum: It provides space where students and teachers can discuss a specific topic or a group of topics to exchange their ideas. Three types of forums can be built: public forum, course forum and class forum. A three-level forum can contribute significantly to successful collaboration and community building in an online environment. Forums are usually integrated into the e-learning systems (Kakasevski et al., 2008; Baturay, 2015) . It helps learners to exchange their ideas and knowledge effectively so that they are not constrained in a passive role but can instead help each other and engage in active ways. Also, forums allow instructors to post course-related questions that can be accessed and discussed by learners. Then the extensions of questions and ideas for interaction are available regardless of whether the instructor is available. Furthermore, forums can also be split into several subforums to provide specific discussions. For example, Coursera provides a default partition of subforums, which includes study groups, general discussions, lectures, assignments, logistics and feedback (Rossi & Gnawali, 2014). The instructors are also allowed to customize the subforums flexibly.
Email: The traditional email has been widely used by Moodle, Blackboard and Open edX and other LMSs (Kakasevski et al., 2008; Bagarukayo et al., 2014). It supports instructors to send email to individual learner or a group of students in the course without launching a separate email program.
Notification: It alerts users about events or activities update in the system.
Others
E-commerce: E-commerce exists in some e-learning systems especially MOOC platforms. It provides sophisticated business transaction functionalities such as payment processing, shopping cart, and customer analytics capabilities. In MOOC platforms, to complete a course or learning module, users need to provide their user profiles and make a one-time payment or agree to a monthly subscription. To earn a degree or a certificate, users also need to pay tuition fees accordingly. E-commerce integrated LMSs usually allow learners to carry out all their transactions starting from registration to making the payment through a single portal, which in turn helps improve user experience.
Challenges of E-learning systems
E-learning systems have profoundly changed the traditional methods of teaching and learning by offering enhanced access to information and interactive resources at all levels of education. They are a supplementary offer to traditional education and to some extent have the ability to substitute it. Despite the advantages it offers, there are still some pedagogical and technical problems that need to be addressed. (Moubayed et al., 2018) analysize several challenges from different aspects, which includes transmission/delivery, personalization, enabling technologies, collaborative/cooperative learning facilitation, and evaluation & assessment. (Islam et al., 2015) also discuss some challenges existing in the success of e-learning, which are mainly related to technology, learning style, training and management. During the Covid-19 pandemic, e-learning faces more challenges as a massive adoption of online education. (Hamdan et al., 2020) analyze several challenges and obstacles including the lace of access to ICT tools, the adequate training for teachers using technological devices, the limited budget for digital devices and poor e-learning environment. (Oryakhail et al., 2021) investigate barriers that hinder the implementation of e-learning in Afghanistan Higher Education. Their research shows specific challenges related to students, lectures, infrastructure and university management.
One major concern for e-learning systems is to use new pedagogy and cognitive approaches to achieve efficient transmission and delivery of e-learning system resources. Since e-learning is quite different from face-to-face education, the courses have to be adapted more attractive or interactive for students, which could be a challenge for teachers who have been used to traditional teaching. E-learning systems require a different approach to pedagogy instead of simply uploading large amounts of resources onto the e-learning systems. (Bari et al., 2018) state that there is no adequate design strategies adapted to the e-learning process and the evaluation of its success implementation. (Andersson, 2008; Moubayed et al., 2018) also discover that some hands-on courses conducted through face-to-face teaching can be difficult to carry out on e-learning systems so that the students cannot fully grasp the content as they learn from traditional classroom-based training. For example, practical lessons or laboratory work are difficult to be conducted on e-learning system (Karjo et al., 2021). Another major concern of e-learning is human resistance, which refers to lack of motivation for both students and teachers. For students, the lack of learning motivation and persistence has been research widely. Since e-learning is self-regulated learning, unmotivated learners may get behind without adequate supervision and guidance. Some statistics show that there is a high dropout rate on MOOCs platforms, which means the majority of students who signed up the course in the beginning could not finish in the end. For example, The Open University found out that only 6.5% of those enrolled students complete the course (Jordan, 2014). The number of enrolments decreases over time and is strongly linked with the duration of the course. As a result, e-learning ends up with a high dropout rates and low effectiveness. To overcome this hurdle, it’s important to stimulate the deep motivations that drive the learners to study or induce them to drop out based on data analysis methods. Also, some kind of useful interventions like self-regulated learning can be delivered to potentially prevent learners’ dropout behavior (Min & Nasir, 2020). Based on (Hapsari et al., 2021), the heavy workloads and more time requirement for teachers have been a challenge affecting the adoption of e-learning. E-learning acceptability is important to the success of e-learning (Hapsari et al., 2021). If teachers have confidence in e-learning and are willing to master both technical and conceptual issues, it will be easy to achieve e-learning success.
Also, with the rapid development of technologies, e-learning systems grow dramatically in terms of the services offered and the available contents generated. Therefore, ensuring that the e-learning system has the means to adapt to the evolving scalability and robustness needs is particularly crucial. According to (Hapsari et al., 2021; Oryakhail et al., 2021; Hamdan et al., 2020; Karjo et al., 2021), lack of reliable internet connection has been a major barrier among e-learning challenges. Also, the lack of infrastructure capability is a common problem faced by both teachers and students in developing countries. If e-learning infrastructure fail to handle requests from thousands of users simultaneously, the system timeout or latency will definitely lead to the interruption of e-learning. Thus, we need to consider how to optimize various hardware or software resources to meet the storage and network requirements as well (El Mhouti et al., 2018). Technologies like cloud computing (Riahi, 2015; El Mhouti et al., 2018; Masud & Huang, 2012; Sun et al., 2015; Chao et al., 2015) have been introduced to provide efficient scalable architecture for e-learning systems.
Moreover, discovering useful information that can be utilized to help teachers determine proper pedagogical strategies and achieve better learning outcomes is also difficult in an e-learning environment (Islam et al., 2015). However, using big data based statistical and mathematical procedure to identify and extract valuable knowledge from large data source is a feasible solution to solve the problem related to “information overload” (Brajkovic et al., 2018).
Furthermore, compared with traditional classroom, it is quite difficult for teachers to track or monitor student progress in e-learning system. AI provides a solution to this problem (Klašnja-Milićević & Ivanović, 2021). It allows teachers to monitor or assess student progress timely. If there is a problem with student performance, AI can be used to alert teachers and assist students based on their strengths and weaknesses.
Lastly, several social challenges faced by e-learning cannot be ignored. Firstly, the cost for e-learning is an issue. (Hamdan et al., 2020) find financial cost for students from low-income families might prevent them from online education. Students need more financial support to purchase computers or stable internet connection services. Secondly, cyber security and privacy is another social challenge facing e-learning. For example, live video applications like Zoom and Microsoft Teams have end-to-end encryption for videos or calls, which ensures the content is encrypted before it’s sent and decrypted only by the intended recipient. However, for most e-learning systems, cyber security and privacy is an optional function which might place the systems and information at risk. Thus, it’s critical to choose the reliable e-learning system and tighten up the security and privacy of online education.
Current trends
Modern e-learning has evolved as a multi-disciplinary process including pedagogy, psychology, various aspects of computer science and many other fields of engineering. Both pedagogy and technology factors have been considered in these trends. Concepts like blended learning, adaptive learning have been introduced to change the traditional in-class education into competency-based education. Also, the wide use of e-learning systems has resulted in huge amount of data generated. By applying data-intensive approaches to educational resources, we can get better understanding of learners, educational settings, and the education results and then improve the teaching and learning process.
Blended learning
Blended learning (or hybrid learning) is evolved from the original computer-based learning environment. It combines the benefits of classroom learning with the advantages of e-learning to ensure an effective learning environment. In other words, learning activities take place inside and outside the classroom. Especially during the COVID-19 pandemic, some or even all classroom teaching are replaced with online teaching (Müller & Mildenberger, 2021; Prahmana et al., 2021) and evaluating the effectiveness of blended learning have been studied widely. Normally, there are many education and technology elements that can be incorporated in learning and teaching processes based on different learning purpose. According to (Valiathan, 2002) and (Prahmana et al., 2021), there are three blended learning models: skill-driven, attitude-driven and competency-driven. —Skilled-driven model—: It aims at providing students specific knowledge and skills, while teachers give feedback and guidance.
—Attitude-driven model—: It enables learners to gain new attitude and behaviors, and interaction and collaboration between learners and teachers plays an important role.
—Competency-driven model—: It aims at transferring learners tacit knowledge by observing and interacting with experts on the job.
Unlike traditional education where the classroom focus is on the teacher, blended learning allows the use of digital texts and tools, and the students become the protagonist of their own learning process, constructing their own knowledge together with the teachers. This mix between classroom learning and e-learning facilitates the students to carry out a more direct and flexible learning style that matches students’ diverse needs.
Adaptive learning
Traditionally, a standard e-learning system does not consider individual differences of learners and treat all learners equally. However, adaptive learning or personalized learning aims to tailor massive information to them based on their features, preferences, background and learning behaviors (Aroyo et al., 2006; Gomede et al., 2021; Mavroudi et al., 2018). In order to do so, adaptive learning basically utilizes a data-driven method to identify the students’ needs faster, and therefore enable the delivery of personalized learning at scale. It also needs differentiated teaching strategies and smart feedback to build learner skills (Sonwalkar, 2013). For instance, algorithms are used to evaluate students’ current learning conditions using online tests, thus adapted modules will be provided to identify their learning gap and improve learning outcomes. According to (Onah & Sinclair, 2015) , the most common adaptive method for course development have the following aspects: the adaptive hypermedia information retrieval system; adaptive annotation system; adaptive recommendation system; adaptive web navigation; adaptive feedback. (Oxman et al., 2014) conclude that at least three components are needed for an adaptive system. a content model to structure the provided contents,
a learner model to understand learners abilities, and
an instructional model to match the content with learners in a dynamic and personalized way.
Generally, adaptive learning can support adjustments in faculty role, allow creative teaching methods, and facilitate learning process in multiple ways.
Educational data mining (EDM) and learning analytics (LA)
EDM refers to developing technical methods to explore different kinds of educational datasets for the purpose of better understanding of learners and educational settings (Mohamad & Tasir, 2013). LA is a closely related endeavor to EDM, and mainly emphasizes on the process of collecting, measuring, analyzing and reporting data about learners and their contexts, in order to understand and optimize learning and the environments in which it takes place (Knobbout & van der Stappen, 2018). However, there are several differences between them according to (Siemens & Baker, 2012). EDM focuses much more on automated adaptation and discovery, whereas LA is mainly in support of teachers and learners judgment. Technically, EDM focus on using typical data mining methods to assist learning process analysis including commonly used techniques like classification, clustering and association rule mining (Mavroudi et al., 2018). In addition to these methods, LA may also take advantage of methods like statistical analysis, Social Network Analysis (SNA) and visualization tools to enable users to gain an overview of the learning results. They establish an ecosystem to reshape the existing models of education and provide new solutions to facilitate the teaching and learning process.
In response to these trends, the focus has been shifted from traditional e-learning toward smart e-learning by integrating big data technology within e-learning paradigm (Kumar et al., 2016) or embedding students’ cognitive model and theories into intelligent learning environments (Gamalel-Din, 2010).
Big data based E-learning systems
Generally, technologies like data science and artificial intelligence (AI) are driving blended learning and adaptive learning effectively. However, combining these technologies with e-learning is still in the early stages. Where and how to place these technologies in e-learning need to be exploited to achieve smart blended learning and adaptive learning. Many researches have been conducted during the past few years. We present a general architecture of big data based e-learning in Fig. 3. It combines essential technologies contributing to collect, aggregate, preprocess, analyze, store and manage big data in e-learning systems. These logical components perform specific functions to enable readers to understand the lifecycle of transforming the different e-learning data into valuable teaching and learning guidance through state-of-the-art technologies (Davoudian & Liu, 2020). Fig. 3 The Architecture of Big Data Based E-learning Systems
Firstly, data in e-learning systems come from different sources with various formats and different granularity. Generally, there are four types of data sources: learning resources, user information, user behavior/activities, and collaboration information. Specifically, learning resource can be audios, videos, slides, texts, presentations or any other kind of documents or files that are used to create course content and pedagogical resources. Another data source comes from users’ demographic information including their gender, age, education, occupation, experiences, and prior knowledge etc. The third type of data source includes users’ behavior and interactions with the system (e.g., log in/out activities, visited contents, quizzes, tests, assignments). The last type of data source comes from the users’ cooperation via emails, instant communication tools, forums and so on.
Taking into account the heterogeneous and hierarchical nature of data sources, it becomes essential to determine data structures and formats that reflect an event (Romero & Ventura, 2013; Otoo-Arthur & van Zyl, 2020). There are three kinds of data: structured, semi-structured or unstructured. Structured data are organized through predefined structures and are stored as records in tables with predefined columns in relational database systems such as MySQL, Oracle, SQL Server and PostgreSQL. Semistructured data have internal structures but are not organized through predefined structures. They are represented in CSV, XML, JSON and other markup languages and are stored in NoSQL stores such as Redis, MongoDB, Cassandra and HBase (Davoudian et al., 2018). Unstructured data have no predefined structures and are stored as text files, emails, videos, audios, web pages, etc. For e-learning data analysis, these different types of data need to be considered together.
Data acquisition
It is the very first step for e-learning data processing. It collects data generated from distributed information sources with various frequencies, sizes, and formats (Wang et al., 2018). It is important that the collected data align with the research questions to be solved in the system. The commonly used open protocols for data acquisition are Advanced Message Queuing Protocol (AMQP) and Java Message Service(JMS) (Lyko et al., 2016). The former is an open source standard for asynchronous messaging between applications considering security, reliability, and performance, whereas the latter allows programs to access system’s messages easily. Regarding common techniques for data acquisition, several tools can be used to collect and aggregate multiple datasets effectively based on different data sources. For instance, Apache Flume can transport large amounts of streaming data such as log files into a centralized store like HDFS at a higher speed, or turn it into a producer of Kafka (Landset et al., 2015). Additionally, Apache Sqoop is designed to transfer batch data between Apache Hadoop and structured database, and dump structured data into HDFS (Geng et al., 2019; Dahdouh et al., 2018).
Data preprocessing
It involves converting original data into an understandable format by using multiple data mining methods or techniques. The collected data are often incomplete, inconsistent, noisy, superfluous and containing errors, which can contribute to incorrect or misleading data analysis and consequently make data analysis slow and inaccurate. So it is not directly applicable to start a data mining process. Alternatively, for further processing and evaluation, the raw data need be converted into correct and helpful information. Generally, data preprocessing includes the following techniques. Data cleaning. It is utilized to discovery missing values, highlight errors, identify outliers, smooth the noisy data, get rid of duplicates, convert improperly formatted or address the inconsistencies in the data to improve data accuracy and quality.
Data integration. It is used to combine data with different representations from multiple sources and also detect and resolve data conflicts.
Data transformation. It converts the original data from one format to another by using normalization, aggregation and generalization and includes steps such as interpretation, pre-translation data quality check, data translation and post-translation data quality check.
Data reduction. It is used to reduce the amount of data required for analysis, through methods such as dimensionality reduction, compression, deduplication, numerosity reduction and so on.
Data discretization. It involves the process of partitioning continuous attributes to several discretized values to prepare datasets well for some mining algorithms such as decision tree that assume discrete values.
Data analysis
It is utilized to process various types of data and perform proper analyses for teaching and learning assistance and improvements. Generally, the common analysis methods can be carried out in realtime, offline and hybrid way. Realtime analysis involves a continual input, analyze and output of data that need to be processed within a short period of time. Parallel processing and memory-based processing are two general methods for realtime analysis (Wang et al., 2018). Tools like Apache Storm, Spark, Flink, S4 (da Silva & et al, 2016), Kafka, SAP Hana (Chandio et al., 2015) are used to deal with realtime data. Offline analysis is utilized to analyze historical data for the purpose of identifing patterns in the environment without high requirement on response time. Most traditional e-learning systems utilized offline analysis method to handle and analyse the learning data in batch. The typical examples include components of Hadoop framework. Specifically, Spark and Hadoop MapReduce are commonly used tools to deal with batch processing. Additonally, a hybrid computation model can be used to deal with massive data by combing the advantages of both batch and realtime analysis.
Several representative data analysis techniques are introduced below to analyze huge data and mining valuable information concealed in the raw data sets. Cluster analysis. It involves grouping same or similar records into clusters. For example, Coursera utilizes t-distributed stochastic neighbour embedding (t-SNE) algorithm to group courses into categories. This method makes categorization scheme simpler with less redundant. Clusters can also be developed in educational information at several distinct grain sizes. Student activities can be combined to explore behavior patterns (Baker & Inventado, 2014). Similarly, teachers or students can be grouped together to discovery similarities or differences among them.
Classification. It is used to classify a value into a specific category in which all the items have very similar or the same characteristics. Classification has been commonly utilized in e-learning to predict the performance of learners (Ahmed & Elaraby, 2014; Baker & Inventado, 2014; Rasheed & Wahid, 2021). For example, the possibility of a student to pass the exam (Rustia et al., 2018) or the students who tend to dropout (Pal, 2012) can be predicted based on students and behavior characteristics.
Correlation analysis. Its goal is to find correlations between variables or attributes. This may be an attempt to determine which variables are most closely related to a single variable of specific concern (Sachin & Vijay, 2012). For example, correlation analysis tools can be used to identify cheat behaviors based on knowledge test and exercise tasks databases (Teodorov et al., 2011).
Regression analysis. It is a predictive modelling techinique, which usually uses statistical method to predict one variable by examining the relationships among a series of variables. Usually, a variable is identified as the predicted variable and a set of other variables as the predictors (Angeli et al., 2017). For example, regression analysis is applied to predict which metrics help explain bad examination results (Feng et al., 2005) in the smart e-learning system.
Data storage
Its main objective is to guarantee the efficient storage of raw data, processed data, analyzied data, and serve data with various types throughout the lifecycle of big data architecture. Since the data are produced at ever-increasing velocity, they must be gathered and stored at low price. Distributed file systems (DFS) are commonly adopted to keep low storage cost and ensure data availability and reliability for data analytics. The typical file systems include GFS, HDFS, TFS and FastTFS by Taobao, Microsoft Cosmos, and Facebook Haystack (Chandio et al., 2015). Other alternatives for distributed data storage are also available that either operate on top of the file storage system or run as standalone devices. Traditional relational databases like Oracle and Postgres are widely used to store structured data. Techniques including replication, caching, horizontal or vertical scaling can be used to deal with the huge volume of data. Non-relational databases known as NoSQL (Not only SQL) stores, are appropriate for intelligent applications as they support multiple data structures (Landset et al., 2015; Davoudian et al., 2018). The types of NoSQL databases usually include key-value storage like Redis, document storage like MongoDB, column-oriented storage like HBase and Graph-based storage like Titan, Neo4J and OrientDB.
Data governance
It refers to a collection of practices and processes about security, integrity, usability and availability of the data employed in an application. It builds rules and guidelines for data quality management across the organization, and allocates responsibilities to each role it defines (Wende & Otto, 2007). It also implements access control and other data security measures, capture the meta data of datasets to support security efforts and facilitates end-user data consumption. Generally, data governance usually includes master data management, meta data management, data quality management, data lifecycle management, data security and privacy management. Master data management. Master data is data most critical to an organization’s operations and analytics. Master data management is a technology-enabled discipline which aims to control master data assets and ensure their consistency and accuracy (Allen & Cervo, 2015). It can help to improve data quality and facilitate computing in multiple system architectures, platforms and applications.
Meta data management. Meta data is the information that describes the semantics of other data. It specifically identifies the attributes, properties and tags that describe and classify information. Meta data management is a critical component for data governance practice since data are scattered in various formats and coming from many sources in the big data environment.
Data quality management. It plans, uses, organizes and disposes data in a quality-oriented manner for the purpose of improving decision making and business services (Weber et al., 2009). It is one of the most significant fields since unqualified data ultimately have enormous adverse effects on data analysis. For example, if we intend to accurately predict student behavior, high quality data are required to develop prediction models.
Data lifecycle management. It refers to decisions that define the definition, production, retention and retirement of data (Khatri & Brown, 2010). In other words, it comprehensively monitors and tracks the life process of data presentation, which paves the way for policies on sensitive data protection and security control.
Data security and privacy management. For the learning process, there are very few researches specifically on data security and privacy, even though both play a significant role in e-learning. Usually, we need measures to implement users authentication and privacy protection, exam protection for data integrity, courseware copyright protection, floor control security for synchronized communication activities (Eibl, 2009) and so forth. For example, student information that will not be used for analysis purpose should be excluded in data acquisition stage to maintain the confidentiality of individual information. Also, personal privacy data need to be protected and encrypted during the lifecycle of data analysis process. Therefore, considering the security and privacy factors while developing e-learning applications can ensure a reliable, efficient learning environment for teaching and learning activities.
Data application
It refers to the application domain that EDM/LA technologies can be adopted in e-learning systems. (Fauvel et al., 2018) reviewed innovative AI techniques that affects MOOCs education. They focused on the researches about student learning behaviors, engagement, and learning performance by constructing intelligent and personalized learning tracks. They categorized the research into three areas including learner modeling, learning experience improvement and learner assessment. Similar, (Bakhshinategh et al., 2018) reviewed current applications that have been used for EDM and classified them into multiple groups and subgroups. Generally, the major goals of EDM are as follows (Vora & Iyer, 2018): (1) providing feedback to support instructors; (2) detecting student behavior; (3) predicting student’s performance; (4) recommendations for students; (5) constructing courseware; (6) planning and scheduling. According to these researches, we review the literatures which focus on the specific learning/teaching purposes including behavior analysis and prediction, recommendation system, personalized learning and multidimensional assessment below.
Behavior analysis and prediction
With big data analytic techniques, instructors can monitor and analyze various online activities accurately such as how long the learners take to answer a question or submit an assignment, how much time they spend in a course, which questions they skip on the test, which part of knowledge they are interested most and so forth. They could also discover various factors that influence students performance and predict the future trends based on these explorations. Additionally, disruptive behaviors include low engagement, excessive lateness, high dropping out rate, cheating on assignments and tests, low learning effectiveness, derogatory comments in online discussion or email can also be found. Since instructors and learners do not have face-to-face interaction, disruptive student behaviors existing among traditional e-learning education environment cannot be disclosed immediately. Big data technology can detect and determine those unacceptable behaviors, while adhering to procedures for reporting disruptive incidents. Then, proper interventions aiming to stop those behaviors or motivate weaker students can be conducted accordingly to mitigate negative factors in e-learning environment. Generally, behavior analysis and prediction mainly focuses on aspects like student learning motivation, engagement, participation, dropout rate, performance success and so on. Several examples are listed in Table 1. Table 1 Behavior analysis and prediction
Topic Methods Reference
User satisfaction, student motivation and engagement, learning performance Gamification techniques (Lisitsyna et al., 2015; Urh et al., 2015; Muntean, 2011)
Correlation between students’ involvement in asking questions and final performance Data mining & text mining (He, 2013)
Relationship between achievement objectives and help-seeking strategies Clustering (Vaessen et al., 2014)
Student behavioral structure, predict possibility to obtain a certificate Latent Dirichlet Allocation (Seaton & Chuang, 2015)
Patterns of similar user behaviors Clustering (Talavera & Gaudioso, 2004)
Students’ learning pace and progress Data clean and analyze (Hadavand & Leek, 2018)
Predict & evaluate low-engagement Classification (Hussain et al., 2018)
Dropout prediction Classification (Liang et al., 2016; Feng et al., 2019; Xing & Du, 2019)
Detect learning style Classification (Rasheed & Wahid, 2021)
Recommendation system
Learners are overwhelmed with the large numbers of learning resources available online. It is becoming more and more difficult for them to select suitable learning materials in e-learning environments. Recommendation systems provide an effective approach to solve this issue by assisting learners to discover appropriate learning contents and improve learning outcomes (Fauvel et al., 2018). Basically, there are four strategies commonly used in recommendation system: collaborative filtering (CF), matrix and tensor factorization, content-based (CB) techniques and association rule mining (Klašnja-Milićević et al., 2017; Ibrahim et al., 2020). Also, many researchers adopt a hybrid recommendation approach by combining the advantages of above-mentioned strategies to promote the quality of recommendations. Some recommendation system examples are shown in Table 2. Table 2 Recommendation system
Topic Methods Reference
Dissimilarity among courses Content-based (Hou et al., 2018)
Relationships between courses activities Association rules (Dahdouh et al., 2019)
Recommend learning resources Context-based method, sequential pattern mining, CF algorithms (Tarus et al., 2018)
Distribute tailored learning contents Association rules, content filtering, collaborative filtering (Xiao et al., 2018)
Recommend learning resources Cluster-based hybrid algorithm (Bhaskaran et al., 2021)
Personalized learning
It refers to a customized way that adapts to the learners’ personalized requirements in e-learning systems. Personalized learning is critical since numerous learners with various background and needs will get involved in e-learning systems. The predefined procedure of learning resources to be followed by the students in a course cannot meet all learners’ particular objectives. According to students learning pattern, individual preferences, and knowledge states (Jeong et al., 2013), personalized learning content can be provided to each individual learner accordingly. Learners are able to reduce the time spent on finding proper contents and get personalized service and meaningful learning experience. Additionally, it can assist them to access the individualized sequence of resources produced and adapted to what they need, rather than following the predefined learning route. Also, the identification of targeted content can satisfy teaching needs among large, heterogeneous and complicated resources. Some examples of personalized learning can be found in Table 3. Table 3 Personalized learning
Topic Methods Reference
Produce tailored learning paths AI planning techniques (Garrido et al., 2016)
Recognize learning style and adjust learning content ThoTh Lab platform (Deng et al., 2018)
Personalized curriculum sequencing with matched difficulties Genetic algorithm (Huang et al., 2007)
Multidimensional assessment
Student learning evaluation and assessment is a key characteristic for e-learning systems. However, it is a challenging task to conduct self-sustainable or personalized evaluation to suit the learner population. Therefore, various dimensions need to be adopted to assess the efficiency of e-learning systems. Table 4 provides some examples for multidimensional assessment. Table 4 Multidimensional assessment
Topic Methods Reference
Autonomous assessment of assignments Web-based system (Le Ru et al., 2015)
Evaluate assignment Formative assessments (Klimova, 2015)
Peer assessment Clustering (Lynda et al., 2017)
Learning performance assessment Classification (Cheng et al., 2011)
Decision making
Decision making is an essential part of e-learning systems. The participants and related stakeholders need to make appropriate decisions to adjust the teaching/learning method or the program process. According to (Galvis, 2018), many contextual factors have an influence on an institution’s decisions. Users are able to establish decision processes that convert education data into actionable insight that can help improve learning performance (Picciano, 2012). Basically, they identify problems and spot opportunities for positive change. Consequently, they can draw precise conclusions and make a better decision by examining and analyzing data consecutively. Additionally, e-learning systems provide possibility to improve decision making capability based on extensive data analysis. Table 5 gives some examples for decision making. Table 5 Decision making
Topic Methods Reference
Precision teaching interventions Data-driven decision model (Zhu et al., 2016)
Provide knowledge for decision-maker Knowledge-based data prototype (Abu-Naser et al., 2011)
Decisions for mobile learning Task-interaction model (Fulantelli et al., 2015)
Conclusion
E-learning systems have been increasingly used to provide efficient learning services, especially after the declaration of the global COVID-19 pandemic by the World Health Organization in mid-March 2020. A lot of post-secondary institutions have introduced e-learning systems alongside online courses.
In this paper, we provide comprehensive review on the efforts of applying new information and communication technologies to improve e-learning services. We first systematically investigate current e-learning systems in terms of their classification, architecture, functions, challenges, and current trends. We then present a general architecture for big data based e-learning systems to meet the ever-growing demand for e-learning. We also describe how to use data generated in big data based e-learning systems to support more flexible and customized course delivery and personalized learning.
Based on the general architecture presented, we have systematically implemented a novel big data based e-learning system called Weblearn17 that has been used by several universities in China for e-learning since 2021.
We are now working on data preprocessing, data analysis, and data application as shown in Fig. 3 in order to provide customized course delivery and personalized learning.
Acknowledgements
The authors would like to thank the anonymous reviewers for their critical reading of the article and their valuable feedbacks, which have substantially helped to improve the quality and accuracy of this article.
Funding
This work was partly supported by Guangzhou Key Laboratory of Big Data and Intelligent Education (No. 2015010009) and National Natural Science Foundation of China (No. 61672389)
1 https://www.udecity.com
2 https://www.edx.org
3 https://www.coursera.org
4 https://www.futurelearn.com
5 https://www.xuetangx.com
6 https://www.fun-mooc.fr
7 https://open.hpi.de
8 https://learn.eduopen.org
9 https://swayam.gov.in
10 https://gacco.org
11 https://thaimooc.org
12 https://openedu.ru/university/hse
13 https://moodle.org
14 https://www.instructure.com
15 https://open.edx.org
16 https://www.sakailms.org
17 http://www.weblearn.cn/
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Abu-Naser S Al-Masri A Sultan YA Zaqout I A prototype decision support system for optimizing the effectiveness of elearning in educational institutions Int. J. Data Mining Knowledge Management Process (IJDKP) 2011 1 1 13 10.5121/ijdkp.2011.1401
Ahmed ABED Elaraby IS Data mining: A prediction for student’s performance using classification method World Journal Computer Applied Technology 2014 2 2 43 47
Alameri, J., Masadeh, R., Hamadallah, E., Bani, H., & Fakhouri, H. (2020). Students’ perceptions of e-learning platforms (moodle, microsoft teams and zoom platforms) in the university of jordan education and its relation to self-study and academic achievement during covid-19 pandemic, vol. 11:13.
Allen, M., & Cervo, D. (2015). Multi-domain master data management: Advanced MDM and data governance in practice.
Alonso F López G Manrique D Viñes JM An instructional model for web-based e-learning education with a blended learning process approach British Journal of Educational Technology 2005 36 2 217 235 10.1111/j.1467-8535.2005.00454.x
Andersson A Seven major challenges for e-learning in developing countries: Case study ebit, sri lanka International Journal of Education and Development using ICT 2008 4 3 45 62
Angeli C Howard SK Ma J Yang J Kirschner PA Data mining in educational technology classroom research: Can it make a contribution? Computers Education 2017 113 226 242 10.1016/j.compedu.2017.05.021
Aroyo L Dolog P Houben G-J Kravcik M Naeve A Nilsson M Wild F Interoperability in personalized adaptive learning Journal of Educational Technology Society 2006 9 2 4 18
Bagarukayo, E., Mpangwire, V., & Kalema, B.M. (2014). Integrating a chatting tool into a learning management system. In 3Rd international conference on informatics engineering and information science (ICIEIS2014), pp. 271–275.
Baker, R.S., & Inventado, P.S. (2014). Educational data mining and learning analytics. In Learning analytics, pp. 61–75.
Bakhshinategh B Zaiane OR ElAtia S Ipperciel D Educational data mining applications and tasks: A survey of the last 10 years Education Information Technologies 2018 23 1 537 553 10.1007/s10639-017-9616-z
Bari M Djouab R Hoa CP Elearning current situation and emerging challenges PEOPLE: International Journal of Social Sciences 2018 4 2
Baturay MH An overview of the world of moocs Procedia - Social and Behavioral Sciences 2015 174 427 433 10.1016/j.sbspro.2015.01.685
Bhaskaran S Marappan R Santhi B Design and analysis of a cluster-based intelligent hybrid recommendation system for e-learning applications Mathematics 2021 9 2 197 10.3390/math9020197
Brajkovic, E, Rakic, K, & Kraljevic, G. (2018). Application of data mining in e-leaming systems. In 2018 17th international symposium on INFOTEH-JAHORINA, INFOTEH 2018 - proceedings 2018-Janua(March), pp 1–5. 10.1109/INFOTEH.2018.8345536.
Carmona, A.B.C., Muñoz, C. M., & Vargas, A.C. (2008). Virtual whiteboard and chat for a learning management system. In E-learning, pp. 263–270.
Cavus, N., & Zabadi, T. (2014). A comparison of open source learning management systems. In Procedia - social and behavioral sciences, vol. 143, pp. 521–526. 10.1016/j.sbspro.2014.07.430. 3rd cyprus international conference on educational research, CY-ICER 2014, 30 January–2014, 1 February, Lefkosa, North Cyprus.
Chandio, A.A., Tziritas, N., & Xu, C.-Z. (2015). Big-data processing techniques and their challenges in transport domain ZTE Communications, vol. 1(010).
Chao K-M James AE Nanos AG Chen J-H Stan S-D Muntean I Figliolini G Rea P Bouzgarrou CB Vitliemov P Cloud e-learning for mechatronics: Clem Future Generation Computer Systems 2015 48 46 59 10.1016/j.future.2014.10.033
Cheng C-H Chen T-L Wei L-Y Chen J-S A new e-learning achievement evaluation model based on rbf-nn and similarity filter Neural Computing and Applications 2011 20 5 659 669 10.1007/s00521-009-0280-0
Cheng P-H Yang Y-TC Chang S-HG Kuo F-RR 5e mobile inquiry learning approach for enhancing learning motivation and scientific inquiry ability of university students IEEE Transactions on Education 2015 59 2 147 153 10.1109/TE.2015.2467352
Chiu, T.K, & Hew, T.K. (2018). Factors influencing peer learning and performance in mooc asynchronous online discussion forum. Australasian Journal of Educational Technology, vol. 34(4), 10.14742/ajet.3240.
Chung C-H Pasquini LA Koh CE Web-based learning management system considerations for higher education Learning and Performance Quarterly 2013 1 4 24 37
Seaton, D.T, & Chuang, I. (2015). Probabilistic use cases: Discovering behavioral patterns for predicting certification. In Proceedings of the second (2015) ACM conference on learning scale. LS ’15, pp. 141–148. ACM, 10.1145/2724660.2724662.
da Silva VA Strategies for big data analytics through lambda architectures in volatile environments IFAC-PapersOnLine 2016 49 30 114 119 10.1016/j.ifacol.2016.11.138
Dahdouh K Dakkak A Oughdir L Messaoudi F Big data for online learning systems Education and Information Technologies 2018 23 6 2783 2800 10.1007/s10639-018-9741-3
Dahdouh K Oughdir L Dakkak A Ibriz A Building an e-learning recommender system using association rules techniques and r environment International Journal of Information Science and Technology 2019 3 2 11 18
Datta, S., & Bhattacharyya, S. (2018). Simple spirited scalable e-learning system. In 2018 IEEE 5th international congress on information science and technology (CiSt), pp. 368–373. 10.1109/CIST.2018.8596631.
Davoudian A Liu M Big data systems: A software engineering perspective ACM Computing Surveys 2020 53 5 110 111039 10.1145/3408314
Davoudian A Chen L Liu M A survey on nosql stores ACM Computing Surveys 2018 51 2 40 14043
Deng, Y., Lu, D., Chung, C.-J., Huang, D., & Zeng, Z. (2018). Personalized learning in a virtual hands-on lab platform for computer science education. In 2018 IEEE frontiers in education conference (FIE), pp. 1–8. IEEE. 10.1109/FIE.2018.8659291.
Eibl, C.J. (2009). Privacy and confidentiality in e-learning systems. In 2009 fourth international conference on internet and web applications and services, pp. 638–642. 10.1109/ICIW.2009.102.
El Mhouti A Erradi M Nasseh A Using cloud computing services in e-learning process: Benefits and challenges Education and Information Technologies 2018 23 2 893 909 10.1007/s10639-017-9642-x
Fajar AN Nurcahyo A Sriratnasari SR Soa system architecture for interconected modern higher education in indonesia Procedia Computer Science 2018 135 354 360 10.1016/j.procs.2018.08.184
Fauvel, S., Yu, H., Miao, C., Cui, L., Song, H., Zhang, L., Li, X., & Leung, C. (2018). Artificial intelligence powered moocs: A brief survey. In 2018 IEEE international conference on agents (ICA), pp. 56–61. 10.1109/AGENTS.2018.8460059.
Feng, M., Heffernan, N.T., & Koedinger, K.R. (2005). Looking for sources of error in predicting student’s knowledge. In Educational data mining: Papers from the 2005 AAAI workshop, pp. 54–61.
Feng, W., Tang, J., & Liu, T.X. (2019). Understanding dropouts in moocs. In AAAI 2019.
Fulantelli G Taibi D Arrigo M A framework to support educational decision making in mobile learning Computers in Human Behavior 2015 47 50 59 10.1016/j.chb.2014.05.045
Galvis ÁH Supporting decision-making processes on blended learning in higher education: Literature and good practices review International Journal of Educational Technology in Higher Education 2018 15 1 25 10.1186/s41239-018-0106-1
Gamalel-Din SA Smart e-learning: A greater perspective; from the fourth to the fifth generation e-learning Egyptian Informatics Journal 2010 11 1 39 48 10.1016/j.eij.2010.06.006
García-Holgado, A., & García-Peñalvo, F. J. (2018). Human interaction in learning ecosystems based on open source solutions. In P. Zaphiris A. Ioannou (Eds.) Learning and collaboration technologies. Design, development and technological innovation, pp. 218–232. Springer.
Garrido A Morales L Serina I On the use of case-based planning for e-learning personalization Expert Systems with Applications 2016 60 1 15 10.1016/j.eswa.2016.04.030
Geng D Zhang C Xia C Xia X Liu Q Fu X Big data-based improved data acquisition and storage system for designing industrial data platform IEEE Access 2019 7 44574 44582 10.1109/ACCESS.2019.2909060
Goldie JGS Connectivism: A knowledge learning theory for the digital age? Medical teacher 2016 38 10 1064 1069 10.3109/0142159X.2016.1173661 27128290
Gomede, E., Barros, R.M.D., & Mendes, L.D.S. (2021). Deep auto encoders to adaptive e-learning recommender system - sciencedirect. Computers and education: Artificial intelligence.
González, M.C., García-Peñalvo, F., Guerrero, M.C., & Forment, M.A. (2009). Back and forth: From the lms to the mobile device. a soa approach. In Proceedings of the IADIS international conference mobile learning 2009 (Barcelona, Spain, 26-28 February, 2009), pp. 114–120. IADIS Press, Portugal.
Hadavand, A., & Leek, J. (2018). crsra: A package for cleaning and analyzing coursera research export data. bioRxiv, pp. 275537, 10.1101/275537.
Hamdan, M., Jaidin, J.H., Fithriyah, M., & Anshari, M. (2020). E-learning in time of covid-19 pandemic: Challenges & experiences. In 2020 sixth international conference on e-learning (econf), pp. 12–16. IEEE.
Han I Shin WS The use of a mobile learning management system and academic achievement of online students Computers & Education 2016 102 79 89 10.1016/j.compedu.2016.07.003
Hapsari, I.N., Parhusip, A.R., Wahyu, S., Sutanto, I., Firmansyah, G., & Rosyid, A. (2021). E-learing service issues and challenges: An exploratory study. In 2021 1st international conference on computer science and artificial intelligence (ICCSAI), vol. 1, pp. 196–201. IEEE.
He W Examining students’ online interaction in a live video streaming environment using data mining and text mining Computers in Human Behavior 2013 29 1 90 102 10.1016/j.chb.2012.07.020
Hendradi P Abd Ghani MK Mahfuzah S Yudatama U Prabowo NA Widyanto RA Artificial intelligence influence in education 4.0 to architecture cloud based e-learning system International Journal of Artificial Intelligence Research 2020 4 1 30 38 10.29099/ijair.v4i1.109
Hill, P. (2019). State of higher Ed LMS market for US and Canada: 2018 Year-End Edition. https://eliterate.us/state-higher-ed-lms-market-us-canada-end-2018.
Hou, Y., Zhou, P., Xu, J., & Wu, D.O. (2018). Course recommendation of mooc with big data support: A contextual online learning approach. In IEEE INFOCOM 2018-IEEE conference on computer communications workshops (INFOCOM WKSHPS), pp. 106–111. 10.1109/INFCOMW.2018.8406936.
Hu, H., Zhang, G., Gao, W., & Wang, M. (2020). Big data analytics for mooc video watching behavior based on spark. Neural Computing and Applications, vol. 32, 10.1007/s00521-018-03983-z.
Huang M-J Huang H-S Chen M-Y Constructing a personalized e-learning system based on genetic algorithm and case-based reasoning approach Expert Systems with Applications 2007 33 3 551 564 10.1016/j.eswa.2006.05.019
Hussain, M., Zhu, W., Zhang, W., & Abidi, S.M.R. (2018). Student engagement predictions in an e-learning system and their impact on student course assessment scores. Computational Intelligence and Neuroscience, vol. 2018, 10.1155/2018/6347186.
Ibrahim TS Saleh AI Elgaml N Abdelsalam MM A fog based recommendation system for promoting the performance of e-learning environments Computers & Electrical Engineering 2020 87 106791 10.1016/j.compeleceng.2020.106791
Islam N Beer M Slack F E-learning challenges faced by academics in higher education Journal of Education and Training Studies 2015 3 5 102 112 10.11114/jets.v3i5.947
Jeong J-S Kim M Yoo K-H A content oriented smart education system based on cloud computing International Journal of Multimedia and Ubiquitous Engineering 2013 8 6 313 328 10.14257/ijmue.2013.8.6.31
Jordan, K. (2014). Initial trends in enrolment and completion of massive open online courses. The International Review of Research in Open and Distributed Learning, vol. 15(1), 10.19173/irrodl.v15i1.1651.
Kakasevski, G., Mihajlov, M., Arsenovski, S., & Chungurski, S. (2008). Evaluating usability in learning management system moodle. In ITI 2008-30th international conference on information technology interfaces, pp. 613–618. IEEE, 10.1109/ITI.2008.4588480.
Kanimozhi, S., Kannan, A., Suganya Devi, K., & Selvamani, K. (2019). Secure cloud-based e-learning system with access control and group key mechanism. Concurrency and computation: Practice and experience, 4841, 10.1002/cpe.4841.
Kappe, F., & Scerbakov, N. (2017). Object-oriented architecture of a modern learning management system. In EDULEARN17 proceedings. 9th international conference on education and new learning technologies, pp. 4910–4916. 10.21125/edulearn.2017.2101.
Karjo, C.H., Andreani, W., Herawati, A., Ying, Y., Yasyfin, A.P., & Marie, K. (2021). Teachers’ challenges and needs in e-learning environment. In 2021 international seminar on application for technology of information and communication (iSemantic), pp. 405–409. IEEE.
Kawato, T., Higashino, M., Takahashi, K., & Kawamura, T. (2020). Proposal of distributed e-learning system using idle resources. In 2020 5Th international conference on computer and communication systems (ICCCS).
Khatri V Brown CV Designing data governance Communications of the ACM 2010 53 1 148 152 10.1145/1629175.1629210
Klašnja-Milićević, A., & Ivanović, M. (2021). E-learning personalization systems and sustainable education MDPI.
Klašnja-Milićević, A., Vesin, B., Ivanović, M., Budimac, Z., & Jain, L.C. (2017). Recommender systems in e-learning environments. In E-learning systems, pp. 51–75.
Klimova, B. (2015). Assessment in smart learning environment – a case study approach. In V.L. Uskov, R.J. Howlett, & L.C. Jain (Eds.) Smart Education and Smart e-Learning, pp. 15–24. Springer.
Knobbout, J., & van der Stappen, E. (2018). Where is the learning in learning analytics? In V. Pammer-Schindler, M. Pérez-sanagustín, H. Drachsler, R. Elferink, & M Scheffel (Eds.) Learning, lifelong technology-enhanced, pp. 88–100. Springer.
Kumar, P., Malali, P., & Noronha, H. (2016). Big data integration for transition from e-learning to smart learning framework. In 2016 3rd MEC international conference on big data and smart city (ICBDSC), pp. 1–4. IEEE, 10.1109/ICBDSC.2016.7460379.
Laisheng, X., & Zhengxia, W. (2011). Cloud computing: A new business paradigm for e-learning. In 2011 third international conference on measuring technology and mechatronics automation, vol. 1, pp. 716–719. 10.1109/ICMTMA.2011.181.
Landset S Khoshgoftaar TM Richter AN Hasanin T A survey of open source tools for machine learning with big data in the hadoop ecosystem Journal of Big Data 2015 2 1 24 10.1186/s40537-015-0032-1
Laura, G., Bogdan, O., Aurelia, C., & Serban, M. (2018). University learning management system based on office 365. In 2018 17th international conference on information technology based higher education and training (ITHET),pp. 1–6. IEEE, Olhao, Portugal, 10.1109/ITHET.2018.8424799.
Le Ru, Y., Aron, M., Gerval, J.-P., & Napoleon, T. (2015). Tests generation oriented web-based automatic assessment of programming assignments. In V.L. Uskov, R.J. Howlett, & L.C. Jain (Eds.) Smart education and smart e-learning, pp. 117–127. Springer.
Liang, J., Li, C., & Zheng, L. (2016). Machine learning application in moocs: Dropout prediction. In 2016 11th International conference on computer science & education (ICCSE), pp. 52–57, 10.1109/ICCSE.2016.7581554.
Lisitsyna, L.S., Pershin, A.A., & Kazakov, M.A. (2015). Game mechanics used for achieving better results of massive online courses. In V.L. Uskov, R.J. Howlett, & L.C Jain (Eds.) Education, smart e-learning, smart, pp. 183-192. Springer.
Lu P Cong X Zhou D E-learning-oriented software architecture design and case study International Journal of Emerging Technologies in Learning (iJET) 2015 10 4 59 65 10.3991/ijet.v10i4.4698
Lyko, K., Nitzschke, M., & Ngonga Ngomo, A.-C. (2016). Big data acquisition. In New horizons for a data-driven economy, pp. 39–61.
Lynda, H., Farida, B.-D., Tassadit, B., & Samia, L. (2017). Peer assessment in moocs based on learners’ profiles clustering. In 2017 8th international conference on information technology (ICIT), pp. 532–536. IEEE, 10.1109/ICITECH.2017.8080054.
Masud MAH Huang X An e-learning system architecture based on cloud computing System 2012 10 11 255 259
Mavroudi A Giannakos M Krogstie J Supporting adaptive learning pathways through the use of learning analytics: Developments, challenges and future opportunities Interactive Learning Environments 2018 26 2 206 220 10.1080/10494820.2017.1292531
Mei J Learning management system calendar reminders and effects on time management and academic performance International Research and Review 2016 6 1 29 45
Min H Nasir M Self-regulated learning in a massive open online course: a review of literature European Journal of Interactive Multimedia and Education 2020 1 2 02007 10.30935/ejimed/8403
Mohamad SK Tasir Z Educational data mining: A review Procedia-Social and Behavioral Sciences 2013 97 320 324 10.1016/j.sbspro.2013.10.240
Moubayed A Injadat M Nassif AB Lutfiyya H Shami A E-learning: Challenges and research opportunities using machine learning & data analytics IEEE Access 2018 6 39117 39138 10.1109/ACCESS.2018.2851790
Muntean, C.I. (2011). Raising engagement in e-learning through gamification. In Proc. 6th international conference on virtual learning ICVL, vol. 1, pp. 323–329. Romania.
Müller C Mildenberger T Facilitating flexible learning by replacing classroom time with an online learning environment: A systematic review of blended learning in higher education Educational Research Review 2021 34 100394 10.1016/j.edurev.2021.100394
Onah, D.F., & Sinclair, J. (2015). Massive open online courses: an adaptive learning framework. In 9Th international technology, education and development conference, pp. 2–4.
Oryakhail, M.S., Saay, S., & Nasery, H. (2021). Challenges in the implementation of e-learning in afghanistan higher education. In 2021 international conference advancement in data science, e-learning and information systems (ICADEIS), pp. 1–6. IEEE.
Otoo-Arthur, D., & van Zyl, T. (2020). A scalable heterogeneous big data framework for e-learning systems. In 2020 international conference on artificial intelligence, big data, computing and data communication systems (icABCD), pp. 1–15. IEEE,. 10.1109/icABCD49160.2020.9183863.
Oxman, S., Wong, W., & Innovations, D. (2014). White paper: Adaptive learning systems. Integrated Education Solutions, pp. 6–7.
Pal, S. (2012). Mining educational data using classification to decrease dropout rate of students. arXive:1206.3078.
Picciano AG The evolution of big data and learning analytics in american higher education Journal of Asynchronous Learning Networks 2012 16 3 9 20
Pickard, L., Shah, D., & De Simone, J.J. (2018). Mapping microcredentials across mooc platforms. In 2018 learning with MOOCS (LWMOOCS), pp. 17–21. 10.1109/LWMOOCS.2018.8534617.
Prahmana RCI Hartanto D Kusumaningtyas DA Ali RM Muchlas: Community radio-based blended learning model: A promising learning model in remote area during pandemic era Heliyon 2021 7 7 07511 10.1016/j.heliyon.2021.e07511
Rani M Nayak R Vyas OP An ontology-based adaptive personalized e-learning system, assisted by software agents on cloud storage Knowledge-Based Systems 2015 90 33 48 10.1016/j.knosys.2015.10.002
Rasheed, F., & Wahid, A. (2021). Learning style detection in e-learning systems using machine learning techniques. Expert Systems with Applications (3):114774.
Riahi G E-learning systems based on cloud computing: A review Procedia Computer Science 2015 62 352 359 10.1016/j.procs.2015.08.415
Romero C Ventura S Data mining in education Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery 2013 3 1 12 27
Rossi, L.A., & Gnawali, O. (2014). Language independent analysis and classification of discussion threads in coursera mooc forums. In Proceedings of the 2014 IEEE 15th international conference on information reuse and integration (IEEE IRI 2014), pp. 654–661. IEEE. 10.1109/IRI.2014.7051952.
Rustia, R.A., Cruz, M.M.A., Burac, M.A.P., & Palaoag, T.D. (2018). Predicting student’s board examination performance using classification algorithms. In Proceedings of the 2018 7th international conference on software and computer applications, pp. 233–237. ACM. 10.1145/3185089.3185101.
Sachin, R.B., & Vijay, M.S. (2012). A survey and future vision of data mining in educational field. In 2012 second international conference on advanced computing & communication technologies, pp. 96–100. IEEE. 10.1109/ACCT.2012.14.
Schuck S Kearney M Burden K Exploring mobile learning in the third space Technology, Pedagogy and Education 2017 26 2 121 137 10.1080/1475939X.2016.1230555
Shah, D. (2020). By The Numbers: MOOCs in 2020. https://www.classcentral.com/report/mooc-stats-2020.
Shah, D. (2021). Massive List of MOOC providers around the world, https://www.classcentral.com/report/mooc-providers-list.
Siemens, G., & Baker, R.S.J.d. (2012). Learning analytics and educational data mining: Towards communication and collaboration. In Proceedings of the 2Nd international conference on learning analytics and knowledge. LAK ’12, pp. 252–254. ACM, New York. 10.1145/2330601.2330661.
Sonwalkar, N. (2013). The first adaptive mooc: A case study on pedagogy framework and scalable cloud architecture—part i. In MOOCs Forum, vol. 1, pp. 22–29. Mary Ann Liebert. 10.1089/mooc.2013.0007.
Sun G Cui T Yong J Shen J Chen S Mlaas: a cloud-based system for delivering adaptive micro learning in mobile mooc learning IEEE Transactions on Services Computing 2015 11 2 292 305 10.1109/TSC.2015.2473854
Talavera, L., & Gaudioso, E. (2004). Mining student data to characterize similar behavior groups in unstructured collaboration spaces. In Workshop on artificial intelligence in CSCL. 16th European conference on artificial intelligence, pp. 17–23.
Tarus JK Niu Z Kalui D A hybrid recommender system for e-learning based on context awareness and sequential pattern mining Soft Computing 2018 22 8 2449 2461 10.1007/s00500-017-2720-6
Teodorov, G., Kir, O., & Zheliazkova, I. (2011). Correlation analysis of educational data mining by means a postprocessor’s tool.
Urh M Vukovic G Jereb E Pintar R The model for introduction of gamification into e-learning in higher education Procedia - Social and Behavioral Sciences 2015 197 388 397 10.1016/j.sbspro.2015.07.154
Vaessen BE Prins FJ Jeuring J University students’ achievement goals and help-seeking strategies in an intelligent tutoring system Computers & Education 2014 72 196 208 10.1016/j.compedu.2013.11.001
Valiathan P Blended learning models Learning circuits 2002 3 8 50 59
Vora DR Iyer K Edm–survey of performance factors and algorithms applied International Journal of Engineering & Technology 2018 7 2.6 93 97 10.14419/ijet.v7i2.6.10074
Wang Y Kung L Byrd TA Big data analytics: Understanding its capabilities and potential benefits for healthcare organizations Technological Forecasting and Social Change 2018 126 3 13 10.1016/j.techfore.2015.12.019
Wangmo, C., & Ivanova, G. (2017). Analysis and design of e-learning system prototype for lifelong learning in bhutan. Applied Researches in Technics, Technologies and Education, vol. 5, 10.15547/artte.2017.04.007.
Weber K Otto B Österle H One size does not fit all—a contingency approach to data governance Journal of Data and Information Quality (JDIQ) 2009 1 1 4
Wende, K., & Otto, B. (2007). A contingency approach to data governance. In M.A. Robert, R. O’Hare, M.L. Markus, & B. Klein (Eds.) Proceedings of 12th international conference on information quality, Cambridge, pp. 163–176. https://www.alexandria.unisg.ch/213308/.
Xiao J Wang M Jiang B Li J A personalized recommendation system with combinational algorithm for online learning Journal of Ambient Intelligence and Humanized Computing 2018 9 3 667 677 10.1007/s12652-017-0466-8
Xing W Du D Dropout prediction in moocs: Using deep learning for personalized intervention Journal of Educational Computing Research 2019 57 3 547 570 10.1177/0735633118757015
Zhang H Zhang J Yin X Zhou K Pan Z Cloud-to-end rendering and storage management for virtual reality in experimental education Virtual Reality & Intelligent Hardware 2020 2 4 368 380 10.1016/j.vrih.2020.07.001
Zhu Z-T Yu M-H Riezebos P A research framework of smart education Smart Learning Environments 2016 3 1 4 10.1186/s40561-016-0026-2
| 0 | PMC9742041 | NO-CC CODE | 2022-12-14 23:29:59 | no | Educ Inf Technol (Dordr). 2022 Dec 12;:1-32 | utf-8 | Educ Inf Technol (Dordr) | 2,022 | 10.1007/s10639-022-11479-6 | 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
36504305
24022
10.1007/s11356-022-24022-4
Review Article
Challenges in developing strategies for the valorization of lignin—a major pollutant of the paper mill industry
http://orcid.org/0000-0002-1070-1367
Mandal Dalia Dasgupta [email protected]
[email protected]
1
Singh Gaurav [email protected]
1
Majumdar Subhasree [email protected]
12
Chanda Protik 1
1 grid.444419.8 0000 0004 1767 0991 Department of Biotechnology, National Institute of Technology Durgapur, Mahatma Gandhi Avenue, Durgapur, 713209 West Bengal India
2 Department of Zoology, Sonamukhi College, Sonamukhi, Bankura, 722207 West Bengal India
Responsible Editor: Guilherme L. Dotto
12 12 2022
122
3 8 2022
1 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.
Apart from protecting the environment from undesired waste impacts, wastewater treatment is a crucial platform for recovery. The exploitation of suitable technology to transform the wastes from pulp and paper industries (PPI) to value-added products is vital from an environmental and socio-economic point of view that will impact everyday life. As the volume and complexity of wastewater increase in a rapidly urbanizing world, the challenge of maintaining efficient wastewater treatment in a cost-effective and environmentally friendly manner must be met. In addition to producing treated water, the wastewater treatment plant (WWTP) has a large amount of paper mill sludge (PMS) daily. Sludge management and disposal are significant problems associated with wastewater treatment plants. Applying the biorefinery concept is necessary for PPI from an environmental point of view and because of the piles of valuables contained therein in the form of waste. This will provide a renewable source for producing valuables and bio-energy and aid in making the overall process more economical and environmentally sustainable. Therefore, it is compulsory to continue inquiry on different applications of wastes, with proper justification of the environmental and economic factors. This review discusses current trends and challenges in wastewater management and the bio-valorization of paper mills. Lignin has been highlighted as a critical component for generating valuables, and its recovery prospects from solid and liquid PPI waste have been suggested.
Graphical abstract
Keywords
Biorefinery approach
Carbon dots
Lignin nanoparticles
Pulp and paper industry wastewater
Paper mill sludge (PMS)
Microbial treatment
==== Body
pmcIntroduction
Manufacturing enterprises have a significant role in strengthening the financial system of any nation. During the manufacturing process of different products, these industries yield harmful waste in the form of gasses, solids, and liquids which deteriorate the quality of the environment and directly impact human health (Murillo-Luna et al. 2011). As industrial development began in the late twentieth century, industrial wastewater generation increased significantly. Industrial wastewater contains many new organic compounds generated yearly due to industrial activities. Technological changes in the manufacturing unit also change the combination of discharged and, in turn, wastewater characteristics (Tchobanoglous et al. 1991). Several compounds generated from industrial processes are strenuous and expensive to treat by traditional wastewater processes.
Interestingly, the wastewater also represents a repository of valuable by-products. However, the trend toward converting waste into valuable materials is not practiced appreciably in developing countries (Bajwa et al. 2019). Looking at the global scenario of environmental consciousness, it has become essential for all industries to use each component holistically and follow a biorefinery approach. The conventional use of physical and chemical methods can lead to rising energy utilization and the generation of secondary pollutants, for which biological methods are coming up as a choice.
Among the various industrial sectors, the pulp and paper industry has gained the pace of development. Several governments of leading nations have recognized it as a high-priority industrial sector (Kamali and Khodaparast 2015). It is known that this industry, when processing lignocellulosic biomass and chemicals, requires a large volume of water at each processing stage and generates enormous amounts of pulp mill effluent (liquid waste) and sludge (solid waste) (Patel et al. 2017). Therefore, the disposal of both parts are the two main environmental issues. Most paper pulp mill effluent is characterized by alkaline pH, dark color, high chemical oxygen demand (COD), and suspended solids that result in acute and chronic toxicity to aquatic biota (Patel et al. 2017; Majumdar et al. 2019b, a).
The complex and recalcitrant lignin with chlorinated lignin derivatives present in the paper pulp mill effluent is the major pollutant and causative agent for genetic mutations in exposed organisms (Lindholm-Lehto et al. 2015). Lignin imparts dark brown color, which severely impedes the natural process of photosynthesis by obstructing sunlight’s penetration into the water. Various treatment methods based on physical (adsorption, microfiltration, and photoionization) and chemical (sedimentation, coagulation, oxidation, and ozonation) principles are used (Kamali and Khodaparast 2015). However, these methods have serious drawbacks, including generating secondary pollution with costly mitigation processes. It seems that if this wastewater is used for the extraction of lignin, it cannot only help in the reduction but also in the production of valuable lignin derivatives. Now-a-days lignin can be converted into lignin nanoparticles and even carbon dots (H. Liu et al. 2018a, b; T. Liu et al. 2020). Recently these compounds have been found to indicate several biomedical and industrial applications, such as drug delivery systems, imaging dye, and many more (Cheng et al. 2020; Rai et al. 2017).
This review appraises the possibilities of producing valuables from both solid and liquid waste of the pulp and paper industry for various purposes with the overall utilization of the waste materials using biological routes. This review addresses the strategies developed firstly for the effluent valorization and second for the use of paper mill sludge (PMS), the solid waste of the same industry as a storehouse of essential commodities, majorly through the utilization of lignin.
Paper and pulp industry: the need of developing countries/threat and benefit
Paper plays a vital role in the evolution of our civilization, and it is tough to imagine modern life without paper. Although the SARS-CoV-2 pandemic has severely affected the world, the global consumption of paper and cardboard in 2020 is still 399 million metric tons and is expected to increase to 461 million metric tons by the end of the next decade (Tiseo 2021). India ranks 20th in world paper production, with China, the USA, and Japan leading worldwide (Malaviya and Rathore 2007b). The pulp and paper industries are among the most environmentally concerned, especially in countries like India (Bajpai 2015a, b). The annual paper production in India is around 10.11 million tons, estimated to be 2.6% of the world’s overall production (Working group report 2011).
Furthermore, India possesses one of the fastest-growing pulp and paper markets, with an annual growth rate of > 10% per capita consumption (Bajpai 2015a, b). In India, 60–70% of raw materials used for papermaking are hardwood, bamboo fibers, and agro-wastes, whereas 30–40% are from recycled sources. Generally, the demand for paper and board is growing annually by 5–6% globally (Bhatnagar 2015). The economic profits of the pulp and paper industry have made it a necessary industrial sector globally. However, the industry is among the five major pollution contributors to precious water bodies (Pokhrel and Viraraghavan 2004). It releases millions of tons of contaminated effluent into the aquatic resources yearly (Rigol et al. 2003). Every year approximately 100 million kg of harmful contaminants are generated from the paper and pulp industry (Gopal et al. 2019).
As per our country’s scenario, India’s paper manufacturing units produce 11 million tons per annum and generate 40–50 kg of sludge (dry) per tonne of paper (Working Group report of the 12th Fifth Year Plan). The abundant amount of paper mill sludge (PMS) is mainly considered waste and usually disposed of or burnt off, emitting greenhouse gasses (Fang et al. 2017). Disposing of or using sludge has long been challenging for the pulp and paper industry (Geng et al. 2007). The impetus toward resource utilization and reduced landfills has emphasized the research into the constructive use of PMS (Deeba et al. 2016). The abundantly available PMS has been explored for incorporation into building materials, cement, and ceramics (Coutts 2005; Frías et al. 2015) attributed to its compositional feature of higher cellulose content imparting mechanical strength. The lignocellulosic composition of PMS reminds us of the generation of valuables from the same via utilization of components that will be again a proper biorefinery strategy for the pulp and paper industry.
The primary processes in the paper and pulp industries are debarking, pulping, separating the pulp from cooking liquor, bleaching, stock preparation, and producing final paper products (Bajpai 2015a, b). The primary and foremost aim lies in the effective removal of the lignin seal to expose the cellulose structure for the conversion and formation of valuable products. Now-a-days there is a wide range of conventional physicochemical technologies used for the delignification of lignocellulose biomass. A modern paper mill can operate using less than 6 m3 fresh water per ton of paper. In contrast, older mills may require more water, even 60 m3 per ton of paper (Mänttäri et al. 2004). More than 42% of the wastewater produced in the world is generated by the pulp and paper industry (Pokhrel and Viraraghavan 2004). It poses a significant threat to the environment and ignites the necessity of wastewater management. Moreover, the enormous amount of sludge generated as a result of the water treatment process is a significant source of soil pollution, leading to further alarming situations (Deeba et al. 2016). Except for a small percentage of the waste paper used in paper recycling, the rest is dumped or incinerated without any economic importance. In addition, till now not a single satisfactory waste recycling system is implemented, although this leading industrial sector enormously contributes to the country’s economy. A biorefinery approach for this industry is the need of the hour. Wastewater and PMS, if utilized for the generation of valuables, can further contribute to the economy and reduce environmental hazards to a great extent.
Paper and pulp mill wastes: characteristics and potential hazards
Pulp and paper mill wastes are broadly classified into liquid and solid wastes each having its uniqueness. A dark-colored alkaline waste stream produced by paper mills, referred to as “black liquor,” contains inorganic compounds and wood waste (Chandra et al. 2011). Kraft lignin and phenol derivatives are the primary environmental pollutants in black liquor produced by pulp and paper mills (Raj et al. 2014; Hooda et al. 2015). It is characterized by color, very high biological oxygen demand (BOD), and chemical oxygen demand (COD) due to lignin and its derivatives (Tiku et al. 2010). The pH ranges from 6.1 to 8.3, with the low pH being due to the result of the metabolism of fungal species and the bio-acid metabolism of the native microflora (Kesalkar et al. 2012). Total dissolved solids concentration in the wastewater generally ranges from 395 to 2500 mg/l, and COD varies from 480 to 4450 mg/l (Kumar et al. 2015a, b). In addition to the natural polymers (tannins, lignin), it also contains xenobiotic components produced or used during manufacturing process (chlorinated lignins, chlorinated resin acids, chlorinated phenols, chlorinated polyaromatic hydrocarbons compounds) (Chandra and Singh 2012) and high sodium concentration (Kesalkar et al. 2012). In addition to that, recycled wastewater contains 2,4,7,9-tetramethyl-S-decyne-4,7-diol surfactant used in paints and printing inks (Kamali and Khodaparast 2015) and non-biodegradable organic substances including metal, sand, glass, and plastic components (Buyukkamaci and Koken 2010).
This wastewater, mainly discharged into the environment without adequate treatment, poses severe threats to aquatic, animal, and plant life. Large-size paper mills are generally located near perennial aquatic bodies, and the extracted alkali-soluble lignin fragments and their derivatives are discharged during the process (Chhonkar et al. 2000). Black liquor disposal generates a high concentration of persistent organic pollutants (POPs) as effluent, which harms aquatic flora and fauna (Kortekaas et al. 1998; Nie et al. 2013). According to numerous studies (Owens et al. 1994; Vass et al. 1996; Schnell et al. 2000; Lindström-Seppä et al. 1998; Leppänen and Oikari 1999; Johnsen et al. 1998; Ericson and Larsson 2000), adverse effects can include respiratory stress, mixed oxygenase activity, toxicity and mutagenicity, liver damage, or genotoxicity. They can also cause health hazards such as diarrhea, vomiting, headaches, nausea, and eye irritation in children and employees (Mandal 1996). Lignin and its derivatives released during delignification impart a dark brown color that makes it difficult for sunlight to penetrate the water (Karrasch et al. 2006). It is responsible for extreme adverse effects on aquatic flora and fauna and ultimately worsens human health through biomagnification in the food chain.
In addition, the solid waste generated from paper mill sludge (PMS) has potential threats. It is produced during deinking, pulping processes, and wastewater treatment. As studied, the extent and consistency of the solid waste rely on the raw resources used, the application of techniques, grade, and quality of the paper to be accomplished. Solid waste is generally moist and contains significant ash (Monte et al. 2009). Various gasses such as nitrogen oxides, hydrogen sulfides, sodium sulfides, methyl mercaptan, chlorine dioxides, and sulfur compounds are released from solid waste. These gasses and solid waste are responsible for chronic disorders, such as headaches and nausea. A trace of metals is also released into the water bodies as effluents. These have many unenthusiastic environmental and communal impacts and cause climate alteration (Sarma 2014). Sludge generation is mainly after primary and secondary wastewater treatment. Apart from these many other pollutants, such as grit bark from sorting and screening raw wood material, ash from coal-fired boilers and power generation systems, and lime sludge from a chemical recovery section of mills are produced (Sumathi and Hung 2006).
Conventional treatment strategies for paper pulp industry wastes
Physico-chemical treatment of papermill effluent
The effluent must be treated internally to reduce COD, BOD, and color. Treatment includes physical, chemical, physicochemical, and biological processes (Tchobanoglous et al. 1991). Different physical, chemical, physicochemical, and electrochemical measures have been utilized to mitigate various industries’ effluents. Suspended solids are removed in physical treatment, and wastewater is homogenized to regulate fluctuating flows (Buyukkamaci and Koken 2010). The wastewater is then chemically treated using coagulation, precipitation, adsorption, and coagulation techniques. Adsorbents such as chitosan, rice hull ash containing silica, rice hull ash without silica, ferric sulfate, and poly aluminum chloride have been used to remove pollutants from wastewater by several researchers (Buyukkamaci and Koken 2010; Kamali and Khodaparast 2015; Kumar et al. 2016). Advanced oxidation processes (AOPs) and membrane technologies have gained significant importance among the various methods, thus being highlighted in this review.
Advanced oxidation processes
AOP plays a significant role in paper and pulp waste treatment, specifically, using ozone and Fenton’s reaction. Advanced oxidation processes (AOPs) are widely employed chemical treatments to deal with refractory organic pollutants (Babuponnusami and Muthukumar 2012b). The oxidation of complex compounds is carried out by non-selective hydroxyl radicals produced from Fenton’s reagent, generated through a cascade of complex reactions (Babuponnusami and Muthukumar 2012a, b; Hussain et al. 2013). Fenton reagent’s application had been proven to reduce COD value by 95% in industrial effluents containing phenolics, cyanides at optimum temperature, pH, FeSO4 concentration, and H2O2 concentration (Amat et al. 2005; Sevimli 2005). Mandal et al. (2010) showed that when used in combination with Thiobacillus ferrooxidans, H2O2 concentration can be minimized and thus the cost. It is noticed that though effective, AOPs are quite cost-intensive processes, leading to the generation of undesirable by-products (Chaparro and Rueda-Bayona 2020; Liu et al. 2020).
Similarly, ozone oxidation can help in the oxidation of compounds in two ways: directly by reacting with dissolved substances or indirectly by hydroxyl radicals generated during the decomposition process (Amat et al. 2005). Ozonation proved effective in oxidizing chemicals in paper and pulp industry effluents such as catechol, vanillin, phenol, syringaldehyde, trichlorophenol, chlorophenol, eugenol, guaiacol, and derivatives of cinnamic acid. When assisted with UV rays at pH 6, this ozone can be removed by approximately 40% COD (Amat et al. 2005; Sevimli 2005). Though effective, the significant drawback is because of its short life, and the requirement of continuous ozonation makes this process costly.
Membrane technologies
Membrane technology has emerged as the preferred option to recover water from various wastewater streams for reuse (Asano and Cotruvo 2004). It has been proven that the membrane filtration technique has been found beneficial in treating the paper and mill industry. Microfiltration, followed by ultrafiltration and reverse osmosis, can provide high water recovery rates and excellent chemical composition (Pizzichini et al. 2005). Zhang (2012a) has shown that using a composite flocculant before reverse osmosis can reduce COD by up to 75%. Progress in wastewater treatment processes is obligatory for converting treated wastewater to re-usable for industrial, agricultural, and domestic purposes. However, such systems suffer serious drawbacks with technical and economic constraints mainly related to the retentate disposal and the membrane susceptibility to fouling (Obotey Ezugbe and Rathilal 2020).
Treatment of PMS: the solid waste
Several strategies have been taken to explore technologies that can help increase the digestibility of PMS in industries for its management. Among these, the various pretreatment technologies prior to anaerobic digestion have become more prominent (Meyer and Edwards 2014). These are broadly divided into physical, chemical, and biological methods (Fig. 1), as discussed below.Fig. 1 Various pretreatment methods of PMS (physical, chemical, and biological) in pulp and paper industries (Veluchamy and Kalamdhad 2017b, c, d, a)
One of the most common modes of PMS pretreatment in industries is physical pretreatment, which consists of mechanical and thermal pretreatment. The main objective behind mechanical pretreatment is to disrupt cellulosic fraction crystallinity and reduce the degree of polymerization, making PMS more susceptible to enzymatic hydrolysis (Veluchamy and Kalamdhad 2017a, b, c, d). However, mechanical pretreatment is energy-consuming, high cost, and time-consuming. This strategy is lesser effective because it does not remove the lignin content from PMS biomass (Zheng et al. 2014). In thermal pretreatment, holocellulose hydrolysis can be achieved by different modes of heating over some time with rapid decompression (Veluchamy and Kalamdhad 2017a, b, c, d). Thermal pretreatment is highly effective in increasing the accessible surface area of cellulosic moiety and enhancing the susceptibility of cellulose to microbes and enzymes. However, the optimum parameters and magnitude of enhancement have been found to vary considerably (Wood et al. 2009). Among the other pretreatment methods, chemical pretreatment has gained significant attention. Paper industries use this pretreatment for PMS to delignify it to manufacture cellulose-based commodities or bioethanol and biogas. Various chemical methods are in use (Fig. 1), like acid and alkaline (thermal) hydrolysis, ozonation, and advanced oxidation methods (Veluchamy and Kalamdhad 2017a, b, c, d). These methods are applied for the pretreatment of PMS in paper industries for the removal of lignin and exposure of cellulosic fraction for the generation of valuables. Figure 2 represents the generation and various management methods of PMS in industries. However, environmental hazards generated through these processes remain the foremost hindrance in applying chemical pretreatment methods.Fig. 2 Generation and management strategies of paper mill sludge in pulp and paper industries (Veluchamy and Kalamdhad 2017b, c, d, a)
Need for biological treatments for paper industry wastewater and sludge
The physico-chemical techniques have severe hindrances, such as massive generation of sludge, huge capital investment, formation of secondary by-products, and dependence on energy sources limits its implementation (Banat et al. 1996). Due to the zero discharged policy, the industries involve innovative eco-friendly, sustainable, and efficient treatment practices. Nowadays, microbial technology is a much-admired option to treat wastewater (Chen et al. 2008). As studied, it is considered effective in combating lignin and other xenobiotic components found in industrial effluents (Rai et al. 2005). Various fungal species including basidiomycetes and ascomycetes along with bacterial strains are considered nature’s original recyclers that could convert toxic organic compounds into harmless products. Exploration of the microbial diversity, particularly from the contaminated areas, offers a valuable platform for searching for organisms that can degrade a broad spectrum of pollutants in paper effluent (Kumar et al. 2017; Majumdar et al. 2019a, b; Beheraa et al. 2019). Biological processes have been noticed to be beneficial with respect to generation of reduced harmful by products and sludge. It is interesting that these processes not only require minimum fresh water and energy but it can also produce useful products such as pigments and enzymes, phenolics byproducts (Banat et al. 1996; Rai et al. 2005). The effective function of microbial process has been found to be associated with their inductive metabolic activity. Some bacterial and fungal species can convert hazardous pollutants into non-toxic end products (Priyadaarshinee et al. 2016; Majumdar et al. 2020a, b, c). In the bioremediation process, the potentiality of these microbes to decolorize, degrade, and detoxify lignin and other xenobiotics is harnessed. Microorganism-based remediation approaches have meticulously been scrutinized as potential tools to remove and degrade diverse kinds of hazardous environmental contaminants. Table 1 represents the various microbial species and their lignin degradation potentials.Table 1 Various microorganisms and their lignin degradation potentials: (a) fungal species and (b) bacterial strains
(a)
Sl. no Fungal species Source of lignin % lignin degradation Incubation time Reference
1 Merulius aureus syn Phlebia sp. and Fusarium sambucinum Fuckel MTC3788 Paper and pulp mill effluent 79% 4 days Malaviya and Rathore (2007a)
2 Emericella nidulans var. nidulans Pulp and paper mill effluent 37% 1 day Singhal and Thakur (2009a)
3 Cryptococcus sp. Pulp and paper mill effluent 35–40% 1 day Singhal and Thakur (2009b)
4 Paraconiothyrium variabile Damm Decay wood of Salix matsudana Koidz 22.99% 40 days Gao et al. (2011)
5 Phanerochaete chyrosporium Rice straw 64.25% 28 days Zhang et al. (2012a, b)
6 Trabulsiella guamensis Alkyl lignin 66% 28 days Sumathi and Hung (2006)
7 Aspergillus flavus; Emericella nidulans Alkali lignin 19–41.6% 21 days Barapatre and Jha (2017)
8 Ganoderma lucidum BEOFB 435 Wheat straw 58.5% 21 days Ćilerdžić et al. (2017)
9 Ceriporiopsis subvermispora Wheat straw 66% 49 days Erven et al. (2018)
10 Myrothecium verrucaria Corn stover 36.73 96 h Su et al. (2018)
11 Shinella sp., Cupriavidus sp., and Bosea sp. Wooden antiques 54% 2 days Zhang et al. (2021)
12 Trametes hirsuta BYL-3, Trametes versicolor BYL-7, and Trametes hirsuta BYL-8 Rice straw 39.7% 10 days Wang et al. (2022)
(b)
Sl. no Bacterial species Source of lignin % lignin degradation Incubation time Reference
1 Bacillus sp. Paper and pulp mill effluent 53% 6 days Raj et al. (2007)
2 Paenibacillus sp. strain LD1 Paper and pulp mill effluent 54% 6 days A. Raj et al. (2014)
3 Bacillus magnetrium and Pseudomonas plecoglossicida Black liquor 82% 7 days R. Paliwal et al. (2015)
4 Bacillus subtilis and Klebsiella pneumonia Kraft lignin 58% 6 days S. Yadav and Chandra (2015)
5 Serratia liquefaciens Paper and pulp mill effluent 58% 6 days I. Haq et al. (2016)
6 Bacillus subtilis, Bacillus endo-phyticus, Bacillus sp. Paper and pulp mill effluent 40.19% 2 days Ojha and Tiwari (2016)
7 Bacillus ligniniphilus L1 Alkaline lignin 38.9% 7 days Zhu et al. (2017)
8 Planococcus sp. TRC1 Paper mill sludge 74% 2.5 days Majumdar et al. (2019b, a)
9 Serratia marcescens NITDPER1 Paper mill sludge 54% 3 days Majumdar et al. (2020a, b, c)
10 Bacillus flexus RMWW II Alkali lignin 85% 6 days Kumar et al. (2019)
11 Bacillus amyloliquefaciens SL-7 Tobacco straw 28.55% 15 days Mei et al. (2020)
12 Bacillus altitudinis SL7 Paper and pulp effluent 44% 5 days Khan et al. (2022)
Role of fungal species
In literature, many fungi have been widely used in biological processes to treat wastewater from the pulp and paper industry. Fungi degrade lignin and other phenolic compounds by producing enzymes such as lignin peroxidase laccase and manganese peroxidase. The widely used fungal species are Merulius aureus (Malaviya and Rathore 2007a), Fusarium sambucinum (Malaviya and Rathore 2007a), Rhizopus oryzae (Nagarathnamma and Bajpai 1999), Phanerochaete chrysosporium (Zhang et al. 2012a, b), Trametes pubescens (González et al. 2010), and Aspergillus niger (Liu et al. 2011). Phanerochaete chyrosporium and some brown rot fungi are predominantly used to treat paper and pulp mill effluent. Specifically Pleurotus sajor caju and Rhizopus oryzae species can reduce COD by 74–81% from effluents of a bleached kraft Eucalyptus globulus after 10 days of incubation (Freitas et al. 2009). Singhal and Thakur (2009a) studied the fungal species Emericella nidulans var. nidulans which was found to degrade 37% of lignin at optimum temperature (30–35 °C), pH (5), rpm (125), and inoculation size (7.5%) after 24 h. A consortium of Merulius aureus syn Phlebia sp. and Fusarium sambucinum Fuckel MTC3788 was studied to treat the paper mill effluent by P. Malaviya et al. (2007a) that resulted in a reduction of color, lignin, and COD by 78.6%, 79%, and 89.4% within 4 days. About 50–53% reduction in color, 35–40% lignin concentration at optimum temperature (30–35 °C), pH (5), rpm (125), and inoculation size (7.5%), after 24 h was shown by Cryptococcus sp. as studied by A. Singhal et al. (2009b). Effectively 61.9% reduction in lignin has been shown by Aspergillus flavus strain F10 (Barapatre and Jha 2016). A consortium of Nigrospora sp. LDF00204 and Curvularia lunata LDF21 showed 80% reduction in COD and 76.1% in lignin concentration (Rajwar et al. 2017). Bjerkandera adusta and Phanerochaete chyrosporium showed the ability to completely delignify paper mill effluent within 8 days (Costa et al. 2017). Ganoderma lucidum BEOFB 435, Ganoderma applanatum BEOFB 411, Pleurotus eryngii HAI 711, Pleurotus pulmonarius HAI 573, Pleurotus eryngii, Pleurotus ostreatus, and Pleurotus pulmonarius have been reported to have excellent delignification capacity due to their ability to produce ligninolytic enzymes (Stajić et al. 2006; Ćilerdžić et al. 2016, 2017). Few ascomycetes such as Podospora anserina, Chrysonilia sitophila, Chaetomium globosum, Ustulina deusta, and Alternaria alternata have also been reported to have delignification potentiality (Ferraz and Durán 1995; Sigoillot et al. 2012; Ferrari et al. 2021). However, it has been found from several studies that fungal treatment has a few limitations, including the requirement for exogenous carbon sources, adjustment of low pH, the requirement for hydraulic retention time, and penetration of mycelium into substrates. R. Priyadarshinee et al. (2016) critically reviewed these limitations and concluded that these are creating hindrances in the commercial application of fungi in the treatment of effluents.
Bacterial treatment strategies
Bacterial strains have emerged as prospective substitutes after researching the drawbacks of using fungal species in this wastewater treatment. According to research, Pseudomonas, Nocardia, and Corynebacterium strains could grow fast on phenols linked to lignin but could not break down lignin (Janshekar et al. 1982). However, several research studies have reported on certain bacteria’s delignification abilities, such as Pseudomonas, Arthrobacterium, and Xanthomonas. According to Raj et al. (2007), bacteria may absorb K.L. from paper mill effluents as the only carbon source and decompose it up to 98% of the same. Using three bacterial isolates, Bacillus subtilis, Bacillus endo-phyticus, and Bacillus sp., Ojha and Tiwari (2016) studied the degradation of lignin in paper industry wastewater. Their results showed 17.8%, 17.9%, and 16.87% delignification, respectively. Interestingly, when a consortium made up of these bacterial strains was applied, lignin degradation increased to 40.19%. With the use of Bacillus megaterium, Bacillus subtilis, Bacillus sp., and Pseudomonas aeruginosa in batch study experiments, lignin breakdown and color reduction from effluent from paper and pulp mills have been accomplished up to 50–97% (Raj et al., 2007; Tiku et al. 2010; Tyagi et al. 2014). Bacteria can break down only monomeric lignin, according to several findings (Priyadarshinee et al. 2016; Majumdar et al. 2020a, b, c), and only a small number of strains have been reported that can break down the complex derivatives of lignin produced during pulping operations (Chandra and Bharagava 2013; Priyadarshinee et al. 2016). Using Serratia liquefaciens and Paenibacillus sp., it has been demonstrated that lignin peroxidase and laccase induction may also be used to bioremediate paper and pulp effluent (Raj et al. 2014; Haq et al. 2016).
Bacterial treatment is gaining more interest as several bacterial strains can survive during a broad range of temperature, pH, and in the presence of toxic components. The bacterial growth rate is also faster than fungal. In the case of bacterial treatment, there is no additional cost for removing mycelium (Priyadarshinee et al. 2016). Bacterial species such as Bacillus sp. CS1 (Chang et al. 2014), Sphingobacterium sp. (Wang et al. 2018), Comamonas sp. B9 (Zhang et al. 2012a, b), Pandoraea sp. ISTKB (Kumar et al. 2015a, b), Pandoraea sp. B-6 (Fang et al. 2018), and Novosphingobium sp. B-7 (Chen et al. 2012) is gaining popularity in lignin degradation over fungal species. Paenibacillus sp. strain LD1 could reduce COD by 78% and lignin by 54%, as studied by Raj et al. (2014). Serratia liquefaciens remarkably reduce COD by 85% and lignin by 58% from paper mill effluent, as reported by Haq et al. (2016). Consortium culture of Bacillus magnetrium and Pseudomonas plecoglossicida immobilized on corn cob agricultural residue showed remarkable efficiency in biodegradation of black liquor by mitigating 79 and 82% of COD and lignin, respectively, after 168 h of bacterial treatment (Paliwal et al. 2015). Pseudomonas holds a significant position in wastewater remediation processes due to its remarkable biodegradation capability of several persistent xenobiotic compounds (Zhu et al. 2015). Recently Paenibacillus gluanolyticus showed the potential to deconstruct lignin found in black liquor into value-added products like ethanol, succinic, lactic, propanoic, and malonic acid (Mathews et al. 2016). Among other bacterial species, Bacillus flexus RMWWII effectively reduced the alkali lignin concentration found in the effluent (Kumar et al. 2019). Bacillus cereus AKRC03 could decolorize and degrade 78.67% of hazardous organic pollutants of paper mill effluent (Kumar and Chandra 2021). Recent studies by Khan et al. (2022) showed that Bacillus altitudinis SL7 could degrade 44% lignin if a load of lignin in the effluent is below 3 g/l.
Apart from the various bacterial species studied, our laboratory found a less explored bacterial species from the genus Planococcus showing delignification capacity. The specific role of Planococcus sp. TRC1 for the abatement of paper mill effluent has been investigated (Majumdar et al. 2019b, a). The results revealed a reduction in 85% COD, 74% lignin, 81% phenol, and 96% color after 60 h of treatment. In this unique strategy, the bacterial strain was immobilized on the surface of paper mill sludge beads and used in a fluidized bed reactor to detoxify paper industry wastewater. The use of Planococcus in diverse biotechnological applications is steadily gaining momentum (Behera et al. 2010; Satpute et al. 2010; Engelhardt et al. 2001). Li et al. (2006) reported 100% removal of 1 mM toluene after 32 h of incubation by Planococcus spp. under hypersaline conditions. The catabolic ability of Planococcus sp. for the biodegradation of aromatic hydrocarbons under extreme conditions might be attributed to the presence of essential genes which can translate the stress response proteins, as revealed during decoding the first whole-genome sequence of Planococcus antarcticus DSM (Margolles et al. in 2012). Although several investigations on the degradation of complex hydrocarbons by Planococcus species have been performed (Satpute et al. 2010; Engelhardt et al. 2001), seldom the ligninolytic activity of this potential species has been explored; however, KL degradation efficiency of this genus has been studied by our group.
Biological treatment of PMS
Looking at environmental safety and effectiveness, biological treatment is gaining importance for the solid waste also. With delignifying microorganisms like ascomycetes, basidiomycetes, bacteria, and enzymes, deconstruction of PMS is possible (Zeng et al. 2012). As a result, delignified PMS achieves more exposure for enzyme digestion of cellulosic fraction (Karlsson et al. 2011). It has been found that brown rot and soft rot fungi attack the cellulose material of lignocellulose and can only impart a minor impact on the lignin fraction (Priyadarshinee et al. 2016). So comparatively, white rot fungi has become the microbe of choice, which degrade the lignin component more actively than cellulose (Zeng et al. 2012). In this context, bacterial strains are not way behind; strains of Planococcus sp. TRC1, Pseudomonas fluorescens NITDPY, and Serratia marcescens NITDPER1 have shown the potential of selective lignin deconstruction from biomasses of kraft pulp and PMS, respectively, which can be significant candidates of future choice in pulp and paper industries (Priyadarshinee et al. 2015; Majumdar et al. 2020b). Bacillus sp. IITRDVM-5 showed tremendous potential for remediation of PMS (Sonkar et al. 2021). Biological pretreatment has several advantages like no chemical requirement, lesser energy requirement, and is environment-friendly.
Waste treatment and recovery of valuables: a key platform
Apart from protecting the environment from undesired waste impacts, wastewater treatment is a crucial platform for valuables recovery. There is demand for some new generation technology for wastewater treatment plants, where energy, organics, and other resources can be recuperated as valuable by-products instead of being wastefully destroyed. Recycling waste will significantly support the scheme of product diversity and the power of economic growth (Landrigan and Fuller 2015). Currently, researchers and industries are focusing on the integrated biorefinery approach to protect the environment and dig out cost–benefit from this waste. Figure 3 represents the biorefinery approach for pulp and paper mill waste. In industry, two routes can be implemented for the biorefinery approach. The first one is the presence of integrated biorefinery machinery at the same establishment of the industry. The second route involves collecting effluent from industries at a definite place and extracting valuables via various approaches. The second model may create a new industrial sector for the usage of waste generated by other industries. However, in the case of the second model, the cost of transportation may increase capital expenditure (Sailwal et al. 2020).Fig. 3 Extraction and utilization of lignin from pulp and paper mill industries
Although the composition of PMS greatly varies among paper mills depending on the raw materials used, it is majorly lignocellulosic biomass (Sonowal et al. 2014). Various studies have revealed the potential applications of PMS in numerous ways, including the production of biogas (Lopes et al. 2018), pyrolysis (Cho et al. 2017), ethanol, compost production (Boshoff et al. 2016; Hazarika et al. 2017), material preparation (Goel and Kalamdhad 2017), and vermicomposting (Negi and Suthar 2018). Apart from these functions, recently, Majumdar et al. reported unique properties of PMS such as acting as a carrier matrix for bacterial species in solid state fermentation (Majumdar et al. 2019a, b), a substrate for bacterial pigment production (Majumdar et al. 2020a, 2020b), and source of cellulose nanocrystals through bacterial delignification (Majumdar et al. 2020b). Due to its lignin content, it can be thought of as a source of lignin-based valuables such as lignin nanoparticles, as a binder for particleboard and similar laminated or composite wood products, as a soil conditioner, as a filler or an active ingredient of phenolic resins, and as an adhesive for linoleum, source of vanillin (synthetic vanilla) and dimethyl sulfoxide. Similarly, it can act as a low-cost source of cellulose-based valuables like cellulose microcrystals, carbon dots, and carbon nanotubes that have gained immense industrial importance recently (Veluchamy and Kalamdhad 2017a, b, c, d). Table 2 represents the different valuable compounds generated from PMS from industrial perspectives through chemical, physical, or biological treatment methods.Table 2 Valuables generated from PMS through various treatment methods
Treatment method Valuable generated Uses References
Chemical Activated carbon and bioactive adsorbent Solvent recovery, gas refining, air purification, exhaust desulfurization, and deodorization processes Khalili et al. (2002)
Chemical Manufacture of fired bricks High durability bricks Goel and Kalamdhad (2017)
Mechanical (microfluidization) Cellulose nanofibrils Reinforcing nanofillers, biomedicine, optoelectronic devices, sensors, and energy storage devices Du et al. (2020)
Chemical Sintered ceramics Flame tubes, firing trays etc Asquini et al. (2008)
Supercritical water treatment Energy generation Unconventional energy source Zhang et al. (2010)
Chemical Wood plastic composite Composite material Soucy et al. (2014)
Biological Bioethanol Biofuel Kang et al. (2010)
Biological Beta-carotene pigment Drug, food colorant Majumdar et al. (2020a)
Biological Prodigiosin pigment and cellulose nanocrystals Drug, food colorant, drug delivery system, filler Majumdar et al. (2020b)
Biological Biogas Energy generation Priadi et al. (2014)
Biological Vermicompost Green fertilizer Negi and Suthar (2018)
Lignin: the untapped wealth from paper pulp wastewater and sludge
Wastewater and PMS from paper industries contain lignin, hemicellulose, phenol, and some common xenobiotics (Strassberger et al. 2014). Usually, cellulose and hemicellulose are widely used to produce sugar, paper, and biofuels, while lignin is yielded from these industries (Cao et al. 2018). As a result, lignin remains the most underrated component of lignocellulosic biomass due to its high calorific value and use as an energy provider in the pulp and paper industry (Banu et al. 2022). Figure 4 represents the various valuables that can be generated from lignin isolated from wastewater or sludge. Traditionally, lignin is used for the pulp and paper industry’s bioenergy generation for the mill. It is the primary producer of lignin to generate energy for running the mill and produce paper and additional lignin. Kraft lignin (KL) is the primary energy source in kraft mills, whereas lignosulfonates are commercialized for various applications (Dessbesell et al. 2020). However, recently greater interest of scientists and researchers in lignin because of its excellent properties has paved the way for discoveries and innovations. Extensive studies are being done, and more new and novel techniques are being developed. However, it is scarce compared to other biopolymers. One of the main obstacles to developing novel products from lignin is its complex structure and valorization methods. The recalcitrant nature of lignin is a more significant caveat to producing valuable and novel products (Domínguez-Robles et al. 2020; Banu et al. 2022). Table 3 represents the various valuable metabolites generated from diverse lignin sources through biological methods. By valorizing this technical lignin, they can be used for many different purposes (Banu et al. 2022).Fig. 4 Utilization of lignin extracted from pulp and paper mill waste into various valuables
Table 3 Various valuable metabolites generated from diverse lignin sources through biological methods
Degradation method Bacterial/fungal sp. used Material used Metabolites formed Reference
Bacterial degradation Paenibacillus sp., Aneurinibacillus aneurinilyticus, and Bacillus sp. Kraft lignin Acetic acid, gallic acid, propanoic acid, guaiacol, valeric acid, ferulic acid, t-cinnamic acid, bis(2-ethylhexyl) phthalate A. Raj et al. (2007)
Bacterial treatment Novosphingobium sp. B-7 Bamboo slips Octadecanoic acid, hexadecenoic acid, 3-methoxy-4,β-dihydroxy benzene propanoic acid, vanillic acid, butylated hydroxytoluene, salicylic acid, acetylsalicylic acid, lactic acid, 3-methyl-2-butanol, 3,5-dimethyl benzaldehyde Y. Chen et al. (2012)
Bacterial and fungal C. versicolor, T. gallica, Mycobacterium sp., Streptomyces sp. Kraft lignin Apocynin, vanillic acid, guaiacol F. Asina et al. (2016)
Bacterial and fungal degradation Mixture of various strains including Ascomycota, Glomeromycota, Proteobacteria, Bacteroidetes, Actinobacteria Oil palm empty fruit bunches Catechol, benzoic acid, apocynin, vanillin Tahir et al. (2019)
Bacterial treatment Streptomyces sp. S6 Kraft lignin Triphenylphosphine oxide, bis(2-ethylhexyl) phthalate, butyl acetate, 3-methyl-butanoic acid, 2-methoxyphenol (guaiacol) Riyadi et al. (2020)
Bacterial treatment Bacillus cereus Paper and pulp wastewater Methyl isoleucine, hexadecane, pregna-3,5-dien-20a-ol, O-trimethylsilyl, 4 methyl-10-(1-phenyl-ethylidine)-2,4,6-triaza-tricyclo [5.2.1.0(2,6)] decane R. Kumar et al. (2022)
Utilization of black liquor for recovery of lignin
Lignin isolation from spent or black liquor is a process that has been in operation since the beginning of the twentieth century. The black liquor that contains approximately 35–45% lignin can be utilized to produce energy and several chemical commodities (Zhu et al. 2015). It can be utilized for various value-added product generations, but its use is still limited. Extraction of lignin from black liquor can be a tedious job. However, researchers have come up with various chemical procedures such as acid precipitation, ultrafiltration, and ion exchange to extract lignin from black liquor (Zhu et al. 2015; Oliveira et al. 2020).
The purified lignin can be directly converted into activated carbon, phenol, bio-oil, aromatic compounds, and other exceptional value-added products (Cao et al. 2018). There are different uses of black liquor. Roy et al. (2019) reported that when it is heated to 400–900 °C in anoxic conditions, it can be converted into biofuel, syngas, and biochar. Syngas and biofuels can be utilized to produce energy. In contrast, biochar can be used as biofertilizers, adsorbing agents, sequestering environmental CO2, and immobilizing heavy metals such as arsenic and cadmium in the environment (Roy et al. 2019; Sailwal et al. 2020). Recently Paenibacillus gluanolyticus showed the potential to deconstruct lignin found in black liquor into different value-added products like ethanol, succinic, lactic, propanoic, and malonic acid (Mathews et al. 2016).
Usage of lignin in specialized valuables
The lignin valorization rate of annually produced lignin is less than 2%, proposing the need for technological improvement to get benefit from lignin as a versatile feedstock (Dessbesell et al. 2020; Cao et al. 2018). In recent years, efficient utilization of lignin has attracted wide attention. Studies indicate advancements in two significant lignin methodologies utilization: catalytic degradation into aromatics and thermochemical treatment for carbon material production. Hydrogenolysis, direct pyrolysis, hydrothermal liquefaction, and hydrothermal carbonization of lignin are some of the processes that are being explored (Dessbesell et al. 2020; Cao et al. 2018). Various value-added products could be produced from lignin, such as biomedical materials, copolymer composites, resins, antioxidants, adsorbents, catalysts and catalyst support, and carbon electrodes (Kai et al. 2016; Chen et al. 2018). In addition, many biodegradable polymers like polylactic acid (PLA), polyhydroxyalkanoates, polyhydroxybutyrate and non-biodegradable polyolefin (PL), polyurethane (PU), polyester (PE), polyol (PO), and polyamide (PA) are produced using lignin (Banu et al. 2022). Also, many other value-added chemicals and materials like important intermediate for organic synthesis—ethylbenzene, cholagogue drug material—hydroxyl acetophenone, and carbon-based supercapacitor and catalyst are made from lignin (Cao et al. 2018).
In biomedical and pharmaceuticals, lignin is a relatively new biopolymer. However, recent studies have revealed its potential for being one of the most helpful biopolymers. It has proven to elicit many health benefits due to its specific properties like antimicrobial, anti-inflammatory, anti-carcinogenic, prebiotic, antioxidant (Ayyachamy et al. 2013; Cheng et al. 2020), susceptibility to enzymatic degradation (Culebras et al. 2021), and lower cytotoxicity. These properties make it one of the potential resources for developing many novel systems for biomedical and pharmaceutical applications. In recent times, some of the most sought-after developments in these regards are hydrogels, drug delivery system platforms, nanoparticles, nanotubes, oral solid dosage forms, sunscreen lotions, wound dressing, wound healing material, pharmaceutical excipients, tissue engineering materials, gene delivery systems (Domínguez-Robles et al. 2020), bionic sensor or materials, and UV blocking material (Culebras, et al. 2021). Lignin is widely studied and used in various industries (Haq et al. 2020), such as fuels, agriculture, and construction materials (Ayyachamy et al. 2013). Interestingly, lignin is used for energy generation as an alternative fuel for power and heat generation in industries, as surfactants, resins, bonding agents, and dispersants for ceramics and concrete, polyurethane (PU) foams, vanillin production, and animal feed additive (Dessbesell et al. 2020; Cao et al. 2018). Apart from these direct uses of lignin, two major lignin-derived components of enormous industrial importance, lignin nanoparticles and carbon dots can be considered as next-generation valuables from waste.
Lignin nanoparticles
One of the most lucrative products derived using lignin in today’s biomedical world is lignin nanoparticles. Biopolymeric nanoparticles generally have favorable dispersibility, high specific surface area, flexible molecular design (Dai et al. 2017), high drug loading capacity, prolonged half-life in the bloodstream, and sustained drug release behavior (Liu et al. 2018a, b). They can optimize the pharmacokinetics and pharmacodynamics of drugs (Dai et al. 2017). Bio-based nanomaterials and nanoparticles most commonly used as drug delivery systems are polymer–drug conjugates and liposomes (Cheng et al. 2020). LNPs have improved antioxidant and UV blocking properties compared to macromolecular lignin particles (Domínguez-Robles et al. 2020). Lignin nanoparticles have great potential as nano and micro-carriers of soluble drugs, pesticides, genes, and food additives (Venkatesan et al. 2009; Tsuji 2001; Green et al. 2006; Desai et al. 2005). It can be used as a surfactant in Pickering emulsions. Reports showed that lignin nanoparticles of about 320 nm made from kraft lignin have effectively stabilized hexadecane droplets in aqueous emulsions (Nypelö et al. 2015). Lignin has remarkable free radical scavenging properties, reducing oxygen radicals; therefore, it can be a powerful antioxidant (Lu et al. 1998). Moreover, Zimniewska et al. (2008) showed that fabrics treated with nanostructured lignin possess excellent UV resistance.
Nanoparticles synthesized from biopolymer have been widely used in drug delivery systems or RNA-interfering effectors (Yearla and Padmasree 2016). Lignin nanoparticles effectively improve membranes’ mechanical properties as cellulose triacetate (Dia et al. 2017). Dai et al. (2017) reported efficient delivery of resveratrol. Besides, they contain different monomers with an antibacterial activity that cause cell membrane damage and bacterial cell lysis and can even act as a natural biocide (Cazacu et al. 2013). Extraction of lignin from pulp and paper industry wastewater can lead to a significant source for producing these wonder molecules that cannot only pave the way toward a waste-to-wealth approach but also be an excellent tool for mitigating toxic wastewater from paper industries.
In recent decades due to the extensive research and developments in nanoscience and nanotechnological studies, their implication in the development of nanomedicine and pharmaceuticals has just exploded. It has become one of the hot topics in the research community. Many lignin-based novel nanomaterials and biopolymers have already been developed. It is sometimes regarded as the most up-and-coming technology in the future as a way to overcome medical problems (Cheng et al. 2020) due to the magnetic properties of polymer-based nanoparticles. Biopolymeric nanoparticles generally have favorable dispersibility, high specific surface area, flexible molecular design (Dai et al. 2017), high drug loading capacity, prolonged half-life in the bloodstream, and sustained drug release behavior (Liu et al. 2018a, b). They can optimize the pharmacokinetics and pharmacodynamics of drugs (Dai et al. 2017). Lignin has other use as well due to its excellent properties. Some of these less explored fields are the use of lignin as a pharmaceutical excipient and other materials. Gao et al. (2021) created lignin-based fluoroquinolone antibiotics removing adsorbents by using actinia-shaped lignin-based adsorbents (LNAEs), using lignin as core, and designing a tentacle of grafted poly-acrylic acid (PAA). They were used to adsorb ofloxacin and ciprofloxacin from water. One such sunscreen was developed by Lee et al. (2020). Light-colored lignin (CEL) was extracted from rice husks by cellulolytic enzyme treatment and solvent extraction then spherical nanoparticles (CEL-NP) were prepared using a solvent shifting method. They exhibited higher SPF and UVA PF. Other uses of lignin-based biopolymers include gene delivery systems (Domínguez-Robles et al. 2020) and other precursor molecules. Table 4 lists various lignin-based products with biomedical and pharmaceutical applications.Table 4 Lignin-based pharmaceuticals and biomedical techniques and products
Technique/methodology Type Mechanism and findings References
Hydrogel-based drug delivery system for controlled release of paracetamol Drug delivery system Hydrogel prepared using crosslinking poly(ethylene) glycol diglycidyl ether (PEGDGE) for delivering paracetamol. It is found that greater the crosslinking density greater the hydrogel swelling and more hydrophilic groups yielding more control over delivery and pore size. Lignin:PEGDGE 1:1 ratio showed higher drug diffusion for paracetamol Culebras et al. (2021)
Hydrogel-based drug (curcumin) delivery system Drug delivery system Hydrogels were prepared from lignin by combining lignin with poly(ethylene glycol)/PEG and poly(methyl vinyl ether-co-maleic acid) by esterification reaction
Curcumin was used as model drug for controlled delivery for up to 4 days
Larrañeta et al. (2018)
Stretchable, conductive, self-healing, and ultraviolet-blocking Fe-SL-g PAA hydrogels Potential bioelectric and bionic sensor Fe-SL-g-polyacrylic acid (PAA) hydrogel was developed in rapid timescale from sulfonated lignin (SL) and Fe3 + at 20 °C by dynamic redox reactions with APS Wang et al. (2020)
Lignin-based self-assembling Drug delivery system Lignin-based self-assembling nanomicelle was prepared by synthesizing macromonomer Cheng et al. (2020)
Bio-based nanomaterials and nanoparticles most commonly used as drug delivery systems are polymer–drug conjugates and liposomes (Cheng et al. 2020). LNPs have improved antioxidant and UV blocking properties compared to macromolecular lignin particles (Domínguez-Robles et al. 2020). Cheng et al. (2020) developed intelligent lignin-based self-assembling nanomicelles for oral drug delivery. The nanomicelles were pH-responsive and cell culture showed the ability to inhibit the survival of human colon cancer cells HT-29 while proliferating human bone marrow stromal cells hBMSC and mouse embryonic fibroblast cells NIH-3T3. Ibuprofen (IBU) was used as a model drug. The lignin nanoparticles have great potential in cancer treatment and drug delivery. About 40– 60% of the new drugs are generally hydrophobic (Domínguez-Robles et al. 2019), whereas drug carriers like lignin hydrogels are bad at loading and delivering hydrophobic drugs. This problem is mainly solved by the nanoparticles and nanomicelles developed using lignin and other amphiphilic polymers (Cheng et al. 2020).
Recently three-dimensional-printed biomaterials have become a popular study topic for developing excellent biomedical and pharmaceutical products. 3d-printed wound dressing is one of them. Lignin and antibiotic (tetracycline) are combined with poly(lactic acid) (PLA) pellets, coated with castor oil and polyhydroxybutyrate (PHB) to create such materials which can be used as 3d printing biocomposite filaments (Domínguez-Robles et al. 2020; Yu and Kim 2020). As a result, wound dressing mesh could be developed using 3d printing technology with the extra benefit of customization in size and shape. Another potential use of 3d-printed biomaterials is the fabrication of precise scaffolds for biomedical and tissue engineering applications. Zhang et al. (2020) developed spherical colloidal lignin particles and cellulose nanofibril-alginate hydrogel-based 3d-printed scaffold by combining cellulose nanofibril hydrogels with alginate through crosslinking in the presence of Ca2 + ions. These particles are known as spherical colloidal lignin particles (CLPs). These CLPs were combined with CNF-alginate to create cellulose nanofiber (CNF)-alginate-CLP nanocomposite scaffolds. As noticed, they were found to be biocompatible and thus can be used to print scaffolds for soft-tissue engineering and regenerative-medicine applications.
Carbon dots
Nowadays, a new type of carbon-based nanomaterial, carbon dots (CDs), has attracted broad research interest because of their various physicochemical properties and favorable features like good biocompatibility, unique optical properties, low cost, eco-friendliness, abundant functional groups (e.g., amino, hydroxyl, carboxyl), high stability, and electron mobility (Chen et al. 2016). Carbon dots (CDs) are a type of fluorescent carbon nanomaterial with several distinct characteristics, including tunable emission, good biocompatibility, low cost, and ease of manufacture. They have been used in cellular imaging, sensors, drug delivery, solar cells, and catalysis, among other things (Lim et al. 2015; Wang et al. 2017). They are very flexible materials and can be quickly fixed in sturdy holders for chemical and biochemical testing (Raj et al. 2007).
Sun et al. (2021) synthesized hydrogel containing lignin-based carbon dots. Liu et al. (2020) showed the formation of luminescent transparent wood based on lignin-derived carbon dots as a building material for dual-channel, real-time, and visual detection of formaldehyde gas. Rai et al. (2017) and Myint et al. (2018) prepared lignin-derived carbon dots with excellent bioimaging properties. Researchers have revealed the synthesis of carbon dot (CD), an exciting type of carbon nanoparticle with high luminescence by hydrothermal treatment of lignin in the presence of H2O2 that can be useful as bio-imaging sensors for bioanalytical diagnostics (Chen et al. 2016). CD can be utilized for various other applications, such as oxygen reduction reaction (Mohideen et al. 2020), as the catalyst for green oxidation of phenol (Pirsaheb et al. 2018), pH sensor, moisture sensor, solvatochromism, and solid-state multicolor lighting (Moniruzzaman and Kim 2019). But such application of lignin for the production of CD is still scarce in the literature, and the raw materials used for carbon dots preparation are restricted to lignocellulosic biomasses and other carbon-based natural substances. In this regard, lignin extracted from pulp and paper industry wastewater may be a valuable source for the same generation. It will reduce the cost of producing these CDs and lead to unleashing a new niche for utilizing lignin from waste.
Utilization of PMS through microbial treatment for valuable compounds
Microbes’ role in valorizing waste has always been critical to industries. These organisms not only play a crucial role in biodegradation but also are environmentally safe and acceptability is higher. Several studies have utilized these organisms to valorize PMS and generate valuable compounds. Figure 5 represents the various strategies of microbial treatment for the generation of valuables from PMS. These involve technologies like anaerobic digestion for the generation of biogas. Priadi et al. (2014) reported that anaerobic digestion uses diverse groups of natural bacteria for biodegradation and aids in producing renewable energy from discarded organic material-rich waste. The biogas generated through anaerobic digestion could be used for heating the digester or lighting in the industry. But, anaerobic digestion of PMS is still in its infancy as compared to wastewater (Meyer and Edwards 2014). This is because of the recalcitrant nature of the lignocellulose content of PMS, making the hydrolysis critical. In a study by Lin et al. (2017), it was found that methane yield was increased 1.4-fold by pretreating PMS feedstock with the active bacterial consortium, and the maximum methane yield of 429.19 ml/g/S was noted.Fig. 5 Strategies of microbial treatment for the generation of valuables from PMS
Some recent studies have shown great prospects among other microbe-based treatments of PMS. Looking at the nutritive composition of PMS, it has been used as a growth substrate for microbes. Majumdar et al. (2020a) revealed that Planococcus sp. TRC1 could grow on PMS in solid-state fermentation (SSF), giving celebrated productivity of the pharmaceutically important compound β-carotene. This study revealed an 84% cost reduction of β-carotene from the current market price when PMS was used as the substrate. Besides that, the residual PMS biomass after bacterial pigment production was bioconverted into cellulose-rich biomass due to the promising lignin-degrading capacity of the microbe. Such cellulose-rich biomasses hold excellent prospects when industrial importance is concerned.
In another study by Majumdar et al. (2020b), PMS was dual valorized through the action of bacterial isolate Serratia marcescens NITDPER1. This isolate produced anticancer, antibiotic, and anti-inflammatory compound prodigiosin in SSF, utilizing PMS as substrate. The residual PMS biomass was a potential source of cellulose nanocrystals with excellent thermal stability. The products generated through the processes were confirmed to be non-toxic and safe for industrial use. Such studies hold great potential in presenting PMS as the next-generation lignocellulosic biomass for cellulose and lignin-based valuables through microbial action.
Another essential utilization of PMS through microbial action has been observed during vermicomposting. In association with earthworm species, the addition of microbial species Oligoporus placenta in PMS has shown remarkable results in decreasing total organic carbon, C/N ratio, and cellulose but an increase in total Kjeldahl nitrogen, total phosphorus, total potassium, and pH during vermicomposting. This study by Negi and Suthar (2018) revealed that fungal inoculation during vermicomposting was effective in decomposing cellulose-rich PMS and aided in better quality compost generation.
In the studies of Kang et al. (2010), two types of paper mill sludges, primary and recycle sludge, were used as feedstock for bioconversion to ethanol. The study revealed that commercial cellulase was inefficient because of interference with ash. Simultaneous saccharification and co-fermentation (SSCF) strategies were used involving cellulase and recombinant Escherichia coli (ATCC-55124), and simultaneous saccharification and SSF were followed using cellulase and Saccharomyces cerevisiae (ATCC-200062). Celebrated ethanol yields of 75–81% were achieved from the SSCF. The work of Niju and Vijayan (2020) can also support such studies, where PMS has been demonstrated as a potential feedstock for microbial ethanol production. The various values generated from PMS through various treatment methods are represented in Table 2. These findings highlight the role microbes possess in the valorization of waste PMS for the benefit of society. Besides these, cellulolytic bacteria or enzymatic treatment may also be beneficial in extracting the lignin fraction from PMS for valorization. However, such studies are scarce in literature but hold excellent prospects for future research on PMS valorization.
Challenges, prospects, and concluding remarks
The paper industries are growing daily, and their demand is constantly increasing. Although not severe, due to the excessive amount of waste, these industries’ impact on the environment is very alarming. These industries produce modified lignin, the main component of wastewater, which places the local environment vulnerable to contamination. Biological processes are beneficial because treatment must be environmentally friendly. Waste disposal is a high cost for industries and a significant concern. The significant drivers for waste valorization from the pulp and paper industry are economy and environmental expertise and industries pushing to recover and regain all these substances because 50 and 100% of lost waste resources are contained in wastewater. Recovering valuables from this waste can offset industry costs and reduce the pollution load on the local environment. This holistic approach to extracting wealth from waste must correctly identify the problems that exist in the industry. It is challenging to find out economically viable solutions for optimizing material recovery and production. Environmental impact assessment and industry life cycle assessment will help provide a new way to develop better waste utilization processes while minimizing the environmental burden. Considering the previous report, the industry’s design should adopt an interdisciplinary approach, considering the theory of circular economy, green chemistry, and industrial ecology.
Acknowledgements
The authors convey their sincere gratitude to the Dept. of Biotechnology and Dept. of Chemical Engineering for providing the necessary support to the authors in conducting waste valorization-related research carried out to date. The authors thank the Central library of NIT Durgapur for providing necessary support in the literature review.
Author contribution
• Dalia Dasgupta Mandal—overall drafting and critical revision, editing of the article.
• Gaurav Singh—contributed in literature survey on paper mill effluent based related scientific information and manuscript writing
• Subhasree Majumdar—contributed in literature survey on paper mill sludge related scientific information and manuscript writing
• Protik Chanda—contributed in literature survey in specific area and collection of data
Data availability
NA.
Declarations
Ethical approval
Ethical approval is not applicable.
Consent to participate
Not applicable.
Consent for publication
All authors have agreed to give their consent to publish this review article in your journal.
Competing interests
The authors declare no competing interests.
Highlights
• Biorefinery approach for valorization of paper mill wastes.
• Conventional treatment strategies.
• Need for microbial treatment of paper mill wastewater.
• Biological methods of solid waste utilization.
• Usage of lignin into specialized valuables.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Amat AM Arques A Miranda MA López F Use of ozone and/or U.V. in the treatment of effluents from board paper industry Chemosphere 2005 60 8 1111 1117 10.1016/j.chemosphere.2004.12.062 15993159
Asano T Cotruvo JA Groundwater recharge with reclaimed municipal wastewater: health and regulatory considerations Water Res 2004 38 8 1941 1951 10.1016/j.watres.2004.01.023 15087175
Asina F Brzonova I Voeller K Kozliak E Kubátová A Yao B Ji Y Biodegradation of lignin by fungi, bacteria and laccases Biores Technol 2016 220 414 424 10.1016/j.biortech.2016.08.016
Asquini L Furlani E Bruckner S Maschio S Production and characterization of sintered ceramics from paper mill sludge and glass cullet Chemosphere 2008 71 1 83 89 10.1016/j.chemosphere.2007.10.037 18061240
Ayyachamy M Cliffe FE Coyne JM Collier J Tuohy MG Lignin: untapped biopolymers in biomass conversion technologies Biomass Conversion and Biorefinery 2013 3 3 255 269 10.1007/s13399-013-0084-4
Babuponnusami A Muthukumar K Removal of phenol by heterogenous photo electro Fenton-like process using nano-zero valent iron Sep Purif Technol 2012 98 130 135 10.1016/j.seppur.2012.04.034
Babuponnusami A Muthukumar K Advanced oxidation of phenol: a comparison between Fenton, electro-Fenton, sono-electro-Fenton and photo-electro-Fenton processes Chem Eng J 2012 183 1 9 10.1016/j.cej.2011.12.010
Bajpai P Management of pulp and paper mill waste 2015 Switzerland Springer International Publishing
Bajpai P Management of pulp and paper mill waste 2015 Switzerland Springer International Publishing
Bajwa DS Pourhashem G Ullah AH Bajwa SG A concise review of current lignin production, applications, products and their environmental impact Ind Crops Prod 2019 139 111526 10.1016/j.indcrop.2019.111526
Banat IM Nigam P Singh D Marchant R Microbial decolorization of textile-dyecontaining effluents: a review Biores Technol 1996 58 3 217 227 10.1016/S0960-8524(96)00113-7
Banu JR Kumar G Tyagi VK Bajhaiya AK Gugulothu P Gunasekaran M Biohydrogen production from waste activated sludge through thermochemical mechanical pretreatment Biores Technol 2022 358 127301 10.1016/j.biortech.2022.127301
Barapatre A Jha H Decolourization and biological treatment of pulp and paper mill effluent by lignin-degrading fungus Aspergillus flavus strain F10 Int J Curr Microbiol App Sci 2016 5 5 19 32 10.20546/ijcmas.2016.505.003
Barapatre A Jha H Degradation of alkali lignin by two ascomycetes and free radical scavenging activity of the products Biocatal Biotransform 2017 35 4 269 286 10.1080/10242422.2017.1327953
Behera M Jana PS More TT Ghangrekar MM Rice mill wastewater treatment in microbial fuel cells fabricated using proton exchange membrane and earthen pot at different pH Bioelectrochemistry 2010 79 2 228 233 10.1016/j.bioelechem.2010.06.002 20615762
Beheraa M Paula I Paula SS Mandalb T Mandala DD Simultaneous o-cresol degradation and biosurfactant production by indigenous bacterial monoculture: kinetics and genotoxic risk assessment Environment 2019 31 32
Bhatnagar A Assessment of physico-chemical characteristics of paper industry effluents Rasayan J Chem 2015 8 1 143 145
Boshoff S Gottumukkala LD Van Rensburg E Görgens J Paper sludge (PS) to bioethanol: evaluation of virgin and recycle mill sludge for low enzyme, high-solids fermentation Biores Technol 2016 203 103 111 10.1016/j.biortech.2015.12.028
Buyukkamaci N Koken E Economic evaluation of alternative wastewater treatment plant options for pulp and paper industry Sci Total Environ 2010 408 24 6070 6078 10.1016/j.scitotenv.2010.08.045 20870270
Cao L Iris KM Liu Y Ruan X Tsang DC Hunt AJ Ok YS Song H Zhang S Lignin valorization for the production of renewable chemicals: state-of-the-art review and future prospects Biores Technol 2018 269 465 475 10.1016/j.biortech.2018.08.065
Cazacu G, Capraru M, Popa VI (2013) Advances concerning lignin utilization in new materials. In Advances in natural polymers (pp. 255–312). Springer, Berlin, Heidelberg
Chandra R Bharagava RN Bacterial degradation of synthetic and kraft lignin by axenic and mixed culture and their metabolic products J Environ Biol 2013 34 6 991 24555327
Chandra R Singh R Decolourisation and detoxification of rayon grade pulp paper mill effluent by mixed bacterial culture isolated from pulp paper mill effluent polluted site Biochem Eng J 2012 61 49 58 10.1016/j.bej.2011.12.004
Chandra R Abhishek A Sankhwar M Bacterial decolorization and detoxification of black liquor from rayon grade pulp manufacturing paper industry and detection of their metabolic products Biores Technol 2011 102 11 6429 6436 10.1016/j.biortech.2011.03.048
Chang YC Choi D Takamizawa K Kikuchi S Isolation of Bacillus sp. strains capable of decomposing alkali lignin and their application in combination with lactic acid bacteria for enhancing cellulase performance Biores Technol 2014 152 429 436 10.1016/j.biortech.2013.11.032
Chaparro TR Rueda-Bayona JG Ecotoxicity and genetic toxicity data from a pulp mill bleaching effluent treated with anaerobic digestion and advanced oxidation process (AOP) Data Brief 2020 29 105141 10.1016/j.dib.2020.105141 32021888
Chen Y Cheng JJ Creamer KS Inhibition of anaerobic digestion process: a review Biores Technol 2008 99 10 4044 4064 10.1016/j.biortech.2007.01.057
Chen Y Chai L Tang C Yang Z Zheng Y Shi Y Zhang H Kraft lignin biodegradation by Novosphingobium sp. B-7 and analysis of the degradation process Biores Technol 2012 123 682 685 10.1016/j.biortech.2012.07.028
Chen W Hu C Yang Y Cui J Liu Y Rapid synthesis of carbon dots by hydrothermal treatment of lignin Materials 2016 9 3 184 10.3390/ma9030184 28773309
Chen Y Fan D Han Y Lyu S Lu Y Li G Jiang F Wang S Effect of high residual lignin on the properties of cellulose nanofibrils/films Cellulose 2018 25 11 6421 6431 10.1007/s10570-018-2006-x
Cheng L Deng B Luo W Nie S Liu X Yin Y Liu S Wu Z Zhan P Zhang L Chen J pH-responsive lignin-based nanomicelles for oral drug delivery J Agric Food Chem 2020 68 18 5249 5258 10.1021/acs.jafc.9b08171 32286845
Chhonkar PK, Datta SP, Joshi HC, Pathak H (2000) Impact of industrial effluents on soil health and agriculture-Indian experience: part I-distillery and paper mill effluents
Cho DW Kwon G Yoon K Tsang YF Ok YS Kwon EE Song H Simultaneous production of syngas and magnetic biochar via pyrolysis of paper mill sludge using CO2 as reaction medium Energy Convers Manage 2017 145 1 9 10.1016/j.enconman.2017.04.095
Ćilerdžić J Stajić M Vukojević J Degradation of wheat straw and oak sawdust by Ganoderma applanatum Int Biodeterior Biodegradation 2016 114 39 44 10.1016/j.ibiod.2016.05.024
Ćilerdžić J Galić M Vukojević J Brčeski I Stajić M Potential of selected fungal species to degrade wheat straw, the most abundant plant raw material in Europe BMC Plant Biol 2017 17 2 75 81 28410577
Costa S Dedola DG Pellizzari S Blo R Rugiero I Pedrini P Tamburini E Lignin biodegradation in pulp-and-paper mill wastewater by selected white rot fungi Water 2017 9 12 935 10.3390/w9120935
Coutts RSP A review of Australian research into natural fibre cement composites Cem Concr Compos 2005 27 518 526 10.1016/j.cemconcomp.2004.09.003
Culebras M Pishnamazi M Walker GM Collins MN Facile tailoring of structures for controlled release of paracetamol from sustainable lignin derived platforms Molecules 2021 26 6 1593 10.3390/molecules26061593 33805704
Dai L Liu R Hu LQ Zou ZF Si CL Lignin nanoparticle as a novel green carrier for the efficient delivery of resveratrol ACS Sustain Chem Eng 2017 5 9 8241 8249 10.1021/acssuschemeng.7b01903
Deeba F Pruthi V Negi SY Converting paper mill sludge into neutral lipids by oleaginous yeast Cryptococcus vishniaccii for biodiesel production Bioresour Technol 2016 213 214 222 10.1016/j.biortech.2016.02.105
Desai KGH, Jin Park H (2005) Recent developments in microencapsulation of food ingredients. Dry Technol 23:1361–1394
Dessbesell L Paleologou M Leitch M Pulkki R Xu CC Global lignin supply overview and kraft lignin potential as an alternative for petroleum-based polymers Renew Sustain Energy Rev 2020 123 109768 10.1016/j.rser.2020.109768
Domínguez-Robles J Martin NK Fong ML Stewart SA Irwin NJ Rial-Hermida MI Donnelly RF Larrañeta E Antioxidant PLA composites containing lignin for 3D printing applications: a potential material for healthcare applications Pharmaceutics 2019 11 4 165 10.3390/pharmaceutics11040165 30987304
Domínguez-Robles J Cárcamo-Martínez Á Stewart SA Donnelly RF Larrañeta E Borrega M Lignin for pharmaceutical and biomedical applications–could this become a reality? Sustain Chem Pharm 2020 18 100320 10.1016/j.scp.2020.100320
Du H Parit M Wu M Che X Wang Y Zhang M Wang R Zhang X Jiang Z Li B Sustainable valorization of paper mill sludge into cellulose nanofibrils and cellulose nanopaper J Hazard Mater 2020 400 123106 10.1016/j.jhazmat.2020.123106 32580093
Engelhardt MA Daly K Swannell RPJ Head IM Isolation and characterization of a novel hydrocarbon-degrading, Gram-positive bacterium, isolated from intertidal beach sediment, and description of Planococcus alkanoclasticus sp. nov J Appl Microbiol 2001 90 2 237 247 10.1046/j.1365-2672.2001.01241.x 11168727
Ericson G, Larsson Å (2000) DNA adducts in perch (Perca fluviatilis) living in coastal water polluted with bleached pulp mill effluents. Ecotoxicol Environ Saf46(2):167–173
Fang S Yu Z Lin Y Lin Y Fan Y Liao Y Ma X A study on experimental characteristic of co-pyrolysis of municipal solid waste and paper mill sludge with additives App Therm Eng 2017 111 292 300 10.1016/j.applthermaleng.2016.09.102
Fang X Li Q Lin Y Lin X Dai Y Guo Z Pan D Screening of a microbial consortium for selective degradation of lignin from tree trimmings Biores Technol 2018 254 247 255 10.1016/j.biortech.2018.01.058
Ferrari R, Gautier V, Silar P (2021) Lignin degradation by ascomycetes. In Advances in botanical research (Vol. 99, pp. 77–113). Academic Press
Ferraz A Durán N Lignin degradation during softwood decaying by the ascomycete Chrysonilia sitophila Biodegradation 1995 6 4 265 274 10.1007/BF00695257
Freitas AC Ferreira F Costa AM Pereira R Antunes SC Gonçalves F Rocha-Santos TAP Diniz MS Castro L Peres I Duarte AC Biological treatment of the effluent from a bleached kraft pulp mill using basidiomycete and zygomycete fungi Sci Total Environ 2009 407 10 3282 3289 10.1016/j.scitotenv.2009.01.054 19269018
Frías M Rodríguez O Sánchez de Rojas MI Paper sludge, an environmentally sound alternative source of MK-based cementitious materials A Review Constr Build Mater 2015 74 37 48 10.1016/j.conbuildmat.2014.10.007
Gao H Wang Y Zhang W Wang W Mu Z Isolation, identification and application in lignin degradation of an ascomycete GHJ-4 Afr J Biotech 2011 10 20 4166 4174
Gao B Chang Q Cai J Xi Z Li A Yang H Removal of fluoroquinolone antibiotics using actinia-shaped lignin-based adsorbents: role of the length and distribution of branched-chains J Hazard Mater 2021 403 123603 10.1016/j.jhazmat.2020.123603 32777749
Geng X, Zhang SY, Deng J (2007) Characteristics of paper mill sludge and its utilization for the manufacture of medium density fiberboard. Wood Fiber Sci:345–351
Goel G Kalamdhad AS An investigation on use of paper mill sludge in brick manufacturing Constr Build Mater 2017 148 334 343 10.1016/j.conbuildmat.2017.05.087
González LF Sarria V Sánchez OF Degradation of chlorophenols by sequential biological-advanced oxidative process using Trametes pubescens and TiO2/U.V Bioresource Technology 2010 101 10 3493 3499 10.1016/j.biortech.2009.12.130 20097065
Gopal PM Sivaram NM Barik D Paper industry wastes and energy generation from wastes In energy from toxic organic waste for heat and power generation 2019 Woodhead Publishing 83 97
Green DW Mann S Oreffo ROC Mineralized polysaccharide capsules as biomimetic microenvironments for cell, gene and growth factor delivery in tissue engineering Soft Matter 2006 2 732 737 10.1039/b604786f 32680213
Haq I Kumar S Kumari V Singh SK Raj A Evaluation of bioremediation potentiality of ligninolytic Serratia liquefaciens for detoxification of pulp and paper mill effluent J Hazard Mater 2016 305 190 199 10.1016/j.jhazmat.2015.11.046 26686478
Haq I Mazumder P Kalamdhad AS Recent advances in removal of lignin from paper industry wastewater and its industrial applications–a review Bioresour Technol 2020 312 123636 10.1016/j.biortech.2020.123636 32527619
Hazarika J Ghosh U Kalamdhad AS Khwairakpam M Singh J Transformation of elemental toxic metals into immobile fractions in paper mill sludge through rotary drum composting Ecol Eng 2017 101 185 192 10.1016/j.ecoleng.2017.02.005
Hooda R Bhardwaj NK Singh P Screening and identification of ligninolytic bacteria for the treatment of pulp and paper mill effluent Water Air Soil Pollut 2015 226 9 1 11 10.1007/s11270-015-2535-y
Hussain S Shaikh S Farooqui M COD reduction of waste water streams of active pharmaceutical ingredient–atenolol manufacturing unit by advanced oxidation-Fenton process J Saudi Chem Soc 2013 17 2 199 202 10.1016/j.jscs.2011.03.006
Janshekar H Brown C Haltmeier T Leisola M Fiechter A Bioalteration of kraft pine lignin by Phanerochaete chrysosporium Arch Microbiol 1982 132 1 14 21 10.1007/BF00690810
Johnsen K Tana J Lehtinen KJ Stuthridge T Mattsson K Hemming J Carlberg GE Experimental field exposure of brown trout to river water receiving effluent from an integrated newsprint mill Ecotoxicol Environ Saf 1998 40 3 184 193 10.1006/eesa.1998.1683 9679680
Kai D Chua YK Jiang L Owh C Chan SY Loh XJ Dual functional anti-oxidant and SPF enhancing lignin-based copolymers as additives for personal and healthcare products RSC Adv 2016 6 89 86420 86427 10.1039/C6RA21433A
Kamali M Khodaparast Z Review on recent developments on pulp and paper mill wastewater treatment Ecotoxicol Environ Saf 2015 114 326 342 10.1016/j.ecoenv.2014.05.005 24953005
Kang L Wang W Lee YY Bioconversion of kraft paper mill sludges to ethanol by SSF and SSCF Appl Biochem Biotechnol 2010 161 1 53 66 10.1007/s12010-009-8893-4 20099047
Karlsson A Truong XB Gustavsson J Svensson BH Nilsson F Ejlertsson J Anaerobic treatment of activated sludge from Swedish pulp and paper mills–biogas production potential and limitations Environ Technol 2011 32 14 1559 1571 10.1080/09593330.2010.543932 22329147
Karrasch B Parra O Cid H Mehrens M Pacheco P Urrutia R Valdovinos C Zaror C Effects of pulp and paper mill effluents on the microplankton and microbial self-purification capabilities of the Biobio River Chile Sci Total Environ 2006 359 1–3 194 208 10.1016/j.scitotenv.2005.03.029 15923023
Kesalkar VP Khedikar IP Sudame AM Physico-chemical characteristics of wastewater from paper industry Int J Eng Res Appl 2012 2 4 137 143
Khalili NR Vyas JD Weangkaew W Westfall SJ Parulekar SJ Sherwood R Synthesis and characterization of activated carbon and bioactive adsorbent produced from paper mill sludge Sep Purif Technol 2002 26 2–3 295 304 10.1016/S1383-5866(01)00184-8
Khan SI Zarin A Ahmed S Hasan F Belduz AO Çanakçi S Khan S Badshah M Farman M Shah AA Degradation of lignin by Bacillus altitudinis SL7 isolated from pulp and paper mill effluent Water Sci Technol 2022 85 1 420 432 10.2166/wst.2021.610 35050893
Kortekaas S Vidal G Yan-Ling H Lettinga G Field JA Anaerobic-aerobic treatment of toxic pulping black liquor with upfront effluent recirculation J Ferment Bioeng 1998 86 1 97 110 10.1016/S0922-338X(98)80041-X
Kumar A Chandra R Biodegradation and toxicity reduction of pulp paper mill wastewater by isolated laccase producing Bacillus cereus AKRC03 Cleaner Engineering and Technology 2021 4 100193 10.1016/j.clet.2021.100193
Kumar A Singha S Dasgupta D Datta S Mandal T Simultaneous recovery of silica and treatment of rice mill wastewater using rice husk ash: an economic approach Ecol Eng 2015 84 29 37 10.1016/j.ecoleng.2015.07.010
Kumar M Singh J Singh MK Singhal A Thakur IS Investigating the degradation process of kraft lignin by β-proteobacterium, Pandoraea sp ISTKB Environmental Science and Pollution Research 2015 22 20 15690 15702 10.1007/s11356-015-4771-5 26018290
Kumar A Priyadarshinee R Roy A Dasgupta D Mandal T Current techniques in rice mill effluent treatment: emerging opportunities for waste reuse and waste-to-energy conversion Chemosphere 2016 164 404 412 10.1016/j.chemosphere.2016.08.118 27596828
Kumar A Roy A Priyadarshinee R Sengupta B Malaviya A Dasguptamandal D Mandal T Economic and sustainable management of wastes from rice industry: combating the potential threats Environ Sci Pollut Res 2017 24 34 26279 26296 10.1007/s11356-017-0293-7
Kumar A Priyadarshinee R Singha S Sengupta B Roy A Dasgupta D Mandal T Biodegradation of alkali lignin by Bacillus flexus RMWW II: analyzing performance for abatement of rice mill wastewater Water Sci Technol 2019 80 9 1623 1632 10.2166/wst.2020.005 32039894
Kumar R Singh A Maurya A Yadav P Yadav A Chowdhary P Raj A Effective bioremediation of pulp and paper mill wastewater using Bacillus cereus as a possible kraft lignin-degrading bacterium Biores Technol 2022 352 127076 10.1016/j.biortech.2022.127076
Landrigan PJ Fuller R Global health and environmental pollution Int J Public Health 2015 60 7 761 762 10.1007/s00038-015-0706-7 26135237
Larrañeta E Imízcoz M Toh JX Irwin NJ Ripolin A Perminova A Domínguez-Robles J Rodríguez A Donnelly RF Synthesis and characterization of lignin hydrogels for potential applications as drug eluting antimicrobial coatings for medical materials ACS Sustain Chem Eng 2018 6 7 9037 9046 10.1021/acssuschemeng.8b01371 30023141
Lee SC Yoo E Lee SH Won K Preparation and application of light-colored lignin nanoparticles for broad-spectrum sunscreens Polymers 2020 12 3 699 10.3390/polym12030699 32245174
Leppänen H Oikari A Occurrence of retene and resin acids in sediments and fish bile from a lake receiving pulp and paper mill effluents Environmental Toxicology and Chemistry: an International Journal 1999 18 7 1498 1505 10.1002/etc.5620180723
Li H Liu YH Luo N Zhang XY Luan TG Hu JM Wang ZY Wu PC Chen MJ Lu JQ Biodegradation of benzene and its derivatives by a psychrotolerant and moderately haloalkaliphilic Planococcus sp. strain ZD22 Res Microbiol 2006 157 7 629 636 10.1016/j.resmic.2006.01.002 16815683
Lim SY Shen W Gao Z Carbon quantum dots and their applications Chem Soc Rev 2015 44 1 362 381 10.1039/C4CS00269E 25316556
Lin Y Liang J Zeng C Wang D Lin H Anaerobic digestion of pulp and paper mill sludge pretreated by microbial consortium OEM1 with simultaneous degradation of lignocellulose and chlorophenols Renewable Energy 2017 108 108 115 10.1016/j.renene.2017.02.049
Lindholm-Lehto PC Knuutinen JS Ahkola HS Herve SH Refractory organic pollutants and toxicity in pulp and paper mill wastewaters Environ Sci Pollut Res 2015 22 9 6473 6499 10.1007/s11356-015-4163-x
Lindström-Seppä P Huuskonen S Kotelevtsev S Mikkelson P Räänen T Stepanova L Hänninen O Toxicity and mutagenicity of waste waters from Baikalsk Pulp and Paper Mill: evaluation of pollutant contamination in Lake Baikal Mar Environ Res 1998 46 1–5 273 277 10.1016/S0141-1136(97)00074-3
Liu T Hu H He Z Ni Y Treatment of poplar alkaline peroxide mechanical pulping (APMP) effluent with Aspergillus niger Biores Technol 2011 102 15 7361 7365 10.1016/j.biortech.2011.04.043
Liu H Li Z Sun Y Geng X Hu Y Meng H Ge J Qu L Synthesis of luminescent carbon dots with ultrahigh quantum yield and inherent folate receptor-positive cancer cell targetability Sci Rep 2018 8 1 1 8 29311619
Liu ZH Hao N Shinde S Olson ML Bhagia S Dunlap JR Kao KC Kang X Ragauskas AJ Yuan JS Codesign of combinatorial organosolv pretreatment (COP) and lignin nanoparticles (LNPs) in biorefineries ACS Sustain Chem Eng 2018 7 2 2634 2647 10.1021/acssuschemeng.8b05715
Liu H Luo J Shukla P Effluents detoxification from pulp and paper industry using microbial engineering and advanced oxidation techniques J Hazard Mater 2020 398 122998 10.1016/j.jhazmat.2020.122998 32502804
Lopes ADCP Silva CM Rosa AP de Ávila Rodrigues F Biogas production from thermophilic anaerobic digestion of kraft pulp mill sludge Renewable Energy 2018 124 40 49 10.1016/j.renene.2017.08.044
Lu F-J Chu L-H Gau R-J Free radical-scavenging properties of lignin Nutr Cancer 1998 30 31 38 10.1080/01635589809514637 9507510
Majumdar S Priyadarshinee R Kumar A Mandal T Mandal DD Exploring Planococcus sp. TRC1, a bacterial isolate, for carotenoid pigment production and detoxification of paper mill effluent in immobilized fluidized bed reactor J Clean Prod 2019 211 1389 1402 10.1016/j.jclepro.2018.11.157
Majumdar S Priyadarshinee R Kumar A Mandal T Dasgupta Mandal D Exploring Planococcus sp. TRC1 for carotenoid pigment production and detoxification of paper mill effluent in immobilized fluidized bed reactor J Clean Prod 2019 211 1389 1402 10.1016/j.jclepro.2018.11.157
Majumdar S Paul I Dey S Dutta S Mandal T Mandal DD Biotransformation of paper mill sludge by Serratia marcescens NITDPER1 for prodigiosin and cellulose nanocrystals: a strategic valorization approach Biochem Eng J 2020 164 107766 10.1016/j.bej.2020.107766
Majumdar S Paul I Dey S Dutta S Mandal T Dasgupta Mandal D Biotransformation of paper mill sludge by Serratia marcescens NITDPER1 for prodigiosin and cellulose nanocrystals: a strategic valorization approach Biochem Eng J 2020 164 107766 10.1016/j.bej.2020.107766
Majumdar S Mandal T Dasgupta Mandal D Production kinetics of β-carotene from Planococcus sp. TRC1 with concomitant bioconversion of industrial solid waste into crystalline cellulose rich biomass Process Biochem 2020 92 202 213 10.1016/j.procbio.2020.01.012
Malaviya P Rathore VS Bioremediation of pulp and paper mill effluent by a novel fungal consortium isolated from polluted soil Bioresource Technology 2007 98 18 3647 3651 10.1016/j.biortech.2006.11.021 17208440
Malaviya P Rathore VS Seasonal variations in different physico-chemical parameters of the effluents of Century Pulp and Paper Mill, Lal Kuan Uttarakhand J Environ Biol 2007 28 2 219 224 17915754
Mandal TN Studies in physico-chemical and biological characteristics of pulp and paper mill effluent and its impact on human beings J Freshw Biol 1996 8 4 191 196
Mandal T Maity S Dasgupta D Datta S Advanced oxidation process and biotreatment: their roles in combined industrial wastewater treatment Desalination 2010 250 1 87 94 10.1016/j.desal.2009.04.012
Mänttäri M Pekuri T Nyström M NF270, a new membrane having promising characteristics and being suitable for treatment of dilute effluents from the paper industry J Membr Sci 2004 242 1–2 107 116 10.1016/j.memsci.2003.08.032
Margolles A, Gueimonde M, Sánchez B (2012) Genome sequence of the Antarctic psychrophile bacterium Planococcus antarcticus DSM 14505
Mathews SL Grunden AM Pawlak J Degradation of lignocellulose and lignin by Paenibacillus glucanolyticus Int Biodeterior Biodegradation 2016 110 79 86 10.1016/j.ibiod.2016.02.012
Mei J Shen X Gang L Xu H Wu F Sheng L A novel lignin degradation bacteria-Bacillus amyloliquefaciens SL-7 used to degrade straw lignin efficiently Biores Technol 2020 310 123445 10.1016/j.biortech.2020.123445
Meyer T Edwards EA Anaerobic digestion of pulp and paper mill wastewater and sludge Water Res 2014 65 321 349 10.1016/j.watres.2014.07.022 25150519
Mohideen MM Liu Y Ramakrishna S Recent progress of carbon dots and carbon nanotubes applied in oxygen reduction reaction of fuel cell for transportation Appl Energy 2020 257 114027 10.1016/j.apenergy.2019.114027
Moniruzzaman M Kim J N-doped carbon dots with tunable emission for multifaceted application: solvatochromism, moisture sensing, pH sensing, and solid state multicolor lighting Sens Actuators, B Chem 2019 295 12 21 10.1016/j.snb.2019.05.035
Monte MC Fuente E Blanco A Negro C Waste management from pulp and paper production in the European Union Waste Manag 2009 29 293 308 10.1016/j.wasman.2008.02.002 18406123
Murillo-Luna JL Garcés-Ayerbe C Rivera-Torres P Barriers to the adoption of proactive environmental strategies J Clean Prod 2011 19 13 1417 1425 10.1016/j.jclepro.2011.05.005
Myint AA Rhim WK Nam JM Kim J Lee YW Water-soluble, lignin-derived carbon dots with high fluorescent emissions and their applications in bioimaging J Ind Eng Chem 2018 25 66 387 395 10.1016/j.jiec.2018.06.005
Nagarathnamma R Bajpai P Decolorization and detoxification of extraction-stage effluent from chlorine bleaching of kraft pulp by Rhizopus oryzae Appl Environ Microbiol 1999 65 3 1078 1082 10.1128/AEM.65.3.1078-1082.1999 10049866
Negi R Suthar S Degradation of paper mill wastewater sludge and cow dung by brown-rot fungi Oligoporus placenta and earthworm (Eisenia fetida) during vermicomposting J Clean Prod 2018 201 842 852 10.1016/j.jclepro.2018.08.068
Nie S, Wu Z, Liu J, Liu X, Qin C, Song H, Wang S (2013) Optimization of AOX formation during the first chlorine dioxide bleaching stage (D0) of soda AQ bagasse pulp. Appita: Technology, Innovation, Manufacturing, Environment 66(4): 306–312
Niju S, Vijayan V (2020) Paper mill sludge as a potential feedstock for microbial ethanol production. In Microbial strategies for techno-economic biofuel production (pp. 35–57). Springer, Singapore.
Nypelö TE Carrillo CA Rojas OJ Lignin supracolloids synthesized from (W/O) microemulsions: use in the interfacial stabilization of Pickering systems and organic carriers for silver metal Soft Matter 2015 11 2046 2054 10.1039/C4SM02851A 25629687
Obotey Ezugbe E Rathilal S Membrane technologies in wastewater treatment: a review Membranes 2020 10 5 89 10.3390/membranes10050089 32365810
Ojha AK, Tiwari M (2016) Lignin decolorization and degradation of pulp and paper mill effluent by ligninolytic bacteria. Iranian (Iranica). J Energy Environ 7(3):282–293
Oliveira DM Hoshino ÉP Mota TR Marchiosi R Ferrarese-Filho O dos Santos WD Modulation of cellulase activity by lignin-related compounds Bioresource Technology Reports 2020 10 100390 10.1016/j.biteb.2020.100390
Owens JW Swanson SM Birkholz DA Environmental monitoring of bleached kraft pulp mill chlorophenolic compounds in a northern Canadian river system Chemosphere 1994 29 1 89 109 10.1016/0045-6535(94)90093-0 8044637
Paliwal R Uniyal S Rai JPN Evaluating the potential of immobilized bacterial consortium for black liquor biodegradation Environ Sci Pollut Res 2015 22 9 6842 6853 10.1007/s11356-014-3872-x
Patel SB Mehta A Solanki HA Physicochemical analysis of treated industrial effluent collected from Ahmedabad Mega pipeline J Environ Anal Toxicol 2017 7 5 1 6 10.4172/2161-0525.1000497
Pirsaheb M Moradi S Shahlaei M Farhadian N Application of carbon dots as efficient catalyst for the green oxidation of phenol: kinetic study of the degradation and optimization using response surface methodology J Hazard Mater 2018 353 444 453 10.1016/j.jhazmat.2018.04.038 29704796
Pizzichini M Russo C Di Meo C Purification of pulp and paper wastewater, with membrane technology, for water reuse in a closed loop Desalination 2005 178 1–3 351 359 10.1016/j.desal.2004.11.045
Priadi C Wulandari D Rahmatika I Moersidik SS Biogas production in the anaerobic digestion of paper sludge APCBEE Proc 2014 9 65 69 10.1016/j.apcbee.2014.01.012
Pokhrel D Viraraghavan T Treatment of pulp and paper mill wastewater—a review Sci Total Environ 2004 333 1–3 37 58 10.1016/j.scitotenv.2004.05.017 15364518
Priyadarshinee R Kumar A Mandal T Dasgupta Mandal D Improving the perspective of raw eucalyptus craft pulp for industrial applications through autochthonous bacteria mediated delignification Ind Crop Prod 2015 74 293 303 10.1016/j.indcrop.2015.05.023
Priyadarshinee R Kumar A Mandal T Dasguptamandal D Unleashing the potential of ligninolytic bacterial contributions towards pulp and paper industry: key challenges and new insights Environ Sci Pollut Res 2016 23 23 23349 23368 10.1007/s11356-016-7633-x
Rai HS Bhattacharyya MS Singh J Bansal TK Vats P Banerjee UC Removal of dyes from the effluent of textile and dyestuff manufacturing industry: a review of emerging techniques with reference to biological treatment Crit Rev Environ Sci Technol 2005 35 3 219 238 10.1080/10643380590917932
Rai S Singh BK Bhartiya P Singh A Kumar H Dutta PK Mehrotra GK Lignin derived reduced fluorescence carbon dots with theranostic approaches: nano-drug-carrier and bioimaging J Lumin 2017 1 190 492 503 10.1016/j.jlumin.2017.06.008
Raj A Reddy MK Chandra R Identification of low molecular weight aromatic compounds by gas chromatography–mass spectrometry (GC–MS) from Kraft lignin degradation by three Bacillus sp Int Biodeterior Biodegradation 2007 59 4 292 296 10.1016/j.ibiod.2006.09.006
Raj A Kumar S Haq I Singh SK Bioremediation and toxicity reduction in pulp and paper mill effluent by newly isolated ligninolytic Paenibacillus sp Ecol Eng 2014 71 355 362 10.1016/j.ecoleng.2014.07.002
Rajwar D Paliwal R Rai JPN Biodegradation of pulp and paper mill effluent by co-culturing ascomycetous fungi in repeated batch process Environ Monit Assess 2017 189 9 1 16 10.1007/s10661-017-6198-6
Rigol A Lacorte S Barceló D Sample handling and analytical protocols for analysis of resin acids in process waters and effluents from pulp and paper mills TrAC, Trends Anal Chem 2003 22 10 738 749 10.1016/S0165-9936(03)01011-2
Riyadi FA Tahir AA Yusof N Sabri NSA Noor MJMM Akhir FN Othman NA Zakaria Z Hara H Enzymatic and genetic characterization of lignin depolymerization by Streptomyces sp. S6 isolated from a tropical environment Sci Rep 2020 10 1 1 9 10.1038/s41598-020-64817-4 31913322
Roy S Kumar U Bhattacharyya P Synthesis and characterization of exfoliated biochar from four agricultural feedstock Environ Sci Pollut Res 2019 26 7 7272 7276 10.1007/s11356-018-04117-7
Sailwal M, Banerjee A, Bhaskar T, Ghosh D (2020) Integrated biorefinery concept for Indian paper and pulp industry. In Waste biorefinery (pp. 631–658). Elsevier
Sarma BK (2014) Impact of paper industry on environment: a case study of the Nagaon paper mill. In Research article presented at the International Conference on Trends in Economics, Humanities and Management (ICTEHM’14), Pattaya (Thailand)
Satpute SK Banat IM Dhakephalkar PK Banpurkar AG Chopade BA Biosurfactants, bioemulsifiers and exopolysaccharides from marine microorganisms Biotechnol Adv 2010 28 4 436 450 10.1016/j.biotechadv.2010.02.006 20172021
Schnell A Hodson PV Steel P Melcer H Carey JH Enhanced biological treatment of bleached kraft mill effluents—II. Reduction of mixed function oxygenase (MFO) induction in fish Water Res 2000 34 2 501 509 10.1016/S0043-1354(99)00161-X
Sevimli MF (2005) Post-treatment of pulp and paper industry wastewater by advanced oxidation processes. Ozone: Sci Eng 27(1):37–43
Sigoillot JC Berrin JG Bey M Lesage-Meessen L Levasseur A Lomascolo A Record E Uzan-Boukhris E Fungal strategies for lignin degradation Adv Bot Res 2012 61 263 308 10.1016/B978-0-12-416023-1.00008-2
Singhal A Thakur IS Decolourization and detoxification of pulp and paper mill effluent by Cryptococcus sp Biochem Eng J 2009 46 1 21 27 10.1016/j.bej.2009.04.007
Singhal A Thakur IS Decolourization and detoxification of pulp and paper mill effluent by Emericella nidulans var. nidulans J Hazard Mater 2009 171 1–3 619 625 10.1016/j.jhazmat.2009.06.041 19586717
Sonkar M Kumar V Dutt D A novel sequence batch treatment of wastewater using Bacillus sp. IITRDVM-5 mixing with paper mill and sewage sludge powders Environmental Technology & Innovation 2021 21 101288 10.1016/j.eti.2020.101288
Sonowal P Khwairakpam M Kalamdhad AS Vermicomposting of solid pulp and paper mill sludge (SPPMS) using Eudrilus eugeniae earthworm Int J Environ Sci 2014 5 3 502
Soucy J Koubaa A Migneault S Riedl B The potential of paper mill sludge for wood–plastic composites Ind Crop Prod 2014 54 248 256 10.1016/j.indcrop.2014.01.013
Stajić M Persky L Friesem D Hadar Y Wasser SP Nevo E Vukojević J Effect of different carbon and nitrogen sources on laccase and peroxidases production by selected Pleurotus species Enzyme Microb Technol 2006 38 1–2 65 73 10.1016/j.enzmictec.2005.03.026
Strassberger Z Tanase S Rothenberg G The pros and cons of lignin valorisation in an integrated biorefinery RSC Adv 2014 4 48 25310 25318 10.1039/C4RA04747H
Sumathi S Hung YT Sumathi S Treatment of pulp and paper mill wastes Waste treatment in the process industries 2006 United States CRC Press-Taylor & Francis Group 453 497
Su Y Yu X Sun Y Wang G Chen H Chen G An efficient strategy for enhancing enzymatic saccharification with delignified fungus Myrothecium verrucaria and solid acid Ind Crop Prod 2018 121 396 404 10.1016/j.indcrop.2018.05.034
Sun Y Zhou Q Sheng X Li S Tong Y Guo J Zhou B Zhao J Liu M Li Z Li Y Highly selective fluorescence sensor sensing benzo [a] pyrene in water utilizing carbon dots derived from 4-carboxyphenylboronic acid Chemosphere 2021 282 131127 10.1016/j.chemosphere.2021.131127 34119727
Tchobanoglous G Burton FL Stensel HD Wastewater Engineering Management 1991 7 1 4
Tahir AA, Barnoh NFM, Yusof N, Said NNM, Utsumi M, Yen AM, Hashim H, Noor MJMM, Akhir FNM, Mohamad SE, Sugiura N (2019) Microbial diversity in decaying oil palm empty fruit bunches (OPEFB) and isolation of lignin-degrading bacteria from a tropical environment. Microbes Environ: ME18117
Tiku DK Kumar A Chaturvedi R Makhijani SD Manoharan A Kumar R Holistic bioremediation of pulp mill effluents using autochthonous bacteria Int Biodeterior Biodegradation 2010 64 3 173 183 10.1016/j.ibiod.2010.01.001
Tiseo I (2021) Paper demand globally 2020–2030 | Statista. [online] Statista. Available at: <https://www.statista.com/statistics/1089078/demand-paper-globally-until-2030/> [Accessed 20 August 2021].
Tsuji K Microencapsulation of pesticides and their improved handling safety J Microencapsul 2001 18 137 147 10.1080/026520401750063856 11253931
Tyagi S Kumar V Singh J Teotia P Bisht S Sharma S Bioremediation of pulp and paper mill effluent by dominant aboriginal microbes and their consortium Int J Environ Res 2014 8 3 561 568
van Erven G Nayan N Sonnenberg AS Hendriks WH Cone JW Kabel MA Mechanistic insight in the selective delignification of wheat straw by three white-rot fungal species through quantitative 13C-IS py-GC–MS and whole cell wall HSQC NMR Biotechnol Biofuels 2018 11 1 1 16 29321810
Vass KK Mukhopadhyay MK Mitra K Joshi HC Respiratory stresses in fishes exposed to paper and pulp wastewater Environ Ecol 1996 14 4 895 897
Veluchamy C Kalamdhad AS Influence of pretreatment techniques on anaerobic digestion of pulp and paper mill sludge: a review Bioresour Technol 2017 245 1206 1219 10.1016/j.biortech.2017.08.179 28893499
Veluchamy C Kalamdhad AS Biochemical methane potential test for pulp and paper mill sludge with different food/microorganisms ratios and its kinetics Int Biodeterior Biodegrad 2017 117 197 204 10.1016/j.ibiod.2017.01.005
Veluchamy C Kalamdhad AS Enhanced methane production and its kinetics model of thermally pretreated lignocellulose waste material Bioresour Technol 2017 241 1 9 10.1016/j.biortech.2017.05.068 28549251
Veluchamy C Kalamdhad AS A mass diffusion model on the effect of moisture content for solid state-anaerobic digestion J Clean Prod 2017 162 371 379 10.1016/j.jclepro.2017.06.099
Venkatesan P Manavalan R Valliappan K Microencapsulateation: a vital technique in novel drug delivery system J Pharm Sci Res 2009 1 26 35
Wang R Lu KQ Tang ZR Xu YJ Recent progress in carbon quantum dots: synthesis, properties and applications in photocatalysis J Mater Chem A 2017 5 8 3717 3734 10.1039/C6TA08660H
Wang J Liang J Gao S Biodegradation of lignin monomers vanillic, p-coumaric, and syringic acid by the bacterial strain Sphingobacterium Sp HY-h Current Microbiology 2018 75 9 1156 1164 10.1007/s00284-018-1504-2 29750329
Wang Q Pan X Lin C Ma X Cao S Ni Y Ultrafast gelling using sulfonated lignin-Fe3+ chelates to produce dynamic crosslinked hydrogel/coating with charming stretchable, conductive, selfhealing, and ultraviolet-blocking properties Chem Eng J 2020 396 125341 10.1016/j.cej.2020.125341
Wang J Li L Xu H Zhang Y Liu Y Zhang F Shen G Yan L Wang W Tang H Qiu H Construction of a fungal consortium for effective degradation of rice straw lignin and potential application in bio-pulping Bioresour Technol 2022 344 126168 10.1016/j.biortech.2021.126168 34737050
Wood N Tran H Master E Pretreatment of pulp mill secondary sludge for high-rate anaerobic conversion to biogas Bioresour Technol 2009 100 23 5729 5735 10.1016/j.biortech.2009.06.062 19615891
Working Group Report on Pulp & Paper Sector for 12th Five Year Plan, India, PCo (Ed.), 2011
Yadav S Chandra R Syntrophic co-culture of Bacillus subtilis and Klebsiella pneumonia for degradation of Kraft lignin discharged from rayon grade pulp industry J Environ Sci 2015 33 229 238 10.1016/j.jes.2015.01.018
Yearla SR Padmasree K Preparation and characterisation of lignin nanoparticles: evaluation of their potential as antioxidants and U.V. protectants Journal of Experimental Nanoscience 2016 11 4 289 302 10.1080/17458080.2015.1055842
Yu O Kim KH Lignin to materials: a focused review on recent novel lignin applications Appl Sci 2020 10 13 4626 10.3390/app10134626
Zeng Y Yang X Yu H Zhang X Ma F The delignification effects of white-rot fungal pretreatment on thermal characteristics of moso bamboo Biores Technol 2012 114 437 442 10.1016/j.biortech.2011.10.036
Zhang L Xu CC Champagne P Energy recovery from secondary pulp/paper-mill sludge and sewage sludge with supercritical water treatment Biores Technol 2010 101 8 2713 2721 10.1016/j.biortech.2009.11.106
Zhang C Chen J Wen Z Alternative policy assessment for water pollution control in China’s pulp and paper industry Resour Conserv Recycl 2012 66 15 26 10.1016/j.resconrec.2012.06.004
Zhang W Ren X Lei Q Wang L Screening and comparison of lignin degradation microbial consortia from wooden antiques Molecules 2021 26 10 2862.F 10.3390/molecules26102862 34066118
Zhang X, Fu J, Paulo AC, Yu C, Guebitz GM (2012b) Bioprocessing of bamboo materials.
Zhang X Morits M Jonkergouw C Ora A Valle-Delgado JJ Farooq M Ajdary R Huan S Linder M Rojas O Sipponen MH Three-dimensional printed cell culture model based on spherical colloidal lignin particles and cellulose nanofibril-alginate hydrogel Biomacromolecules 2020 21 5 1875 1885 10.1021/acs.biomac.9b01745 31992046
Zheng Y Zhao J Xu F Li Y Pretreatment of lignocellulosic biomass for enhanced biogas production Prog Energy Combust Sci 2014 42 35 53 10.1016/j.pecs.2014.01.001
Zhu J Rong Y Yang J Zhou X Xu Y Zhang L Chen J Yong Q Yu S Integrated production of xylonic acid and bioethanol from acid-catalyzed steam-exploded corn stover Appl Biochem Biotechnol 2015 176 5 1370 1381 10.1007/s12010-015-1651-x 25947618
Zhu D Zhang P Xie C Zhang W Sun J Qian WJ Yang B Biodegradation of alkaline lignin by Bacillus ligniniphilus L1 Biotechnol Biofuels 2017 10 1 1 14 10.1186/s13068-017-0735-y 28053662
Zimniewska M, Kozłowski R, Batog J (2008) Nanolignin modified linen fabric as a multifunctional product. Mol. Cryst. Liq. Cryst, 484:43/(409)–50/(416)
| 36504305 | PMC9742045 | NO-CC CODE | 2022-12-14 23:45:33 | no | Environ Sci Pollut Res Int. 2022 Dec 12;:1-22 | utf-8 | Environ Sci Pollut Res Int | 2,022 | 10.1007/s11356-022-24022-4 | oa_other |
==== Front
J Affect Disord
J Affect Disord
Journal of Affective Disorders
0165-0327
1573-2517
Elsevier B.V.
S0165-0327(22)01375-1
10.1016/j.jad.2022.12.015
Research Paper
Telehealth treatment of patients with major depressive disorder during the COVID-19 pandemic: Comparative safety, patient satisfaction, and effectiveness to prepandemic in-person treatment
Zimmerman Mark ⁎
D'Avanzato Catherine
King Brittany T.
Department of Psychiatry and Human Behavior, Brown Medical School, Rhode Island Hospital, Providence, RI, United States
⁎ Corresponding author at: 146 West River Street; Providence, RI 02904, United States.
12 12 2022
15 2 2023
12 12 2022
323 624630
31 7 2022
2 12 2022
5 12 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
The COVID-19 pandemic impelled a transition from in-person to telehealth psychiatric treatment. There are no studies of partial hospital telehealth treatment for major depressive disorder (MDD). In the present report from the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project, we compared the effectiveness of partial hospital care of patients with MDD treated virtually versus in-person.
Methods
Outcome was compared in 294 patients who were treated virtually from May 2020 to December 2021 to 542 patients who were treated in the in-person partial program in the 2 years prior to the pandemic. Patients completed self-administered measures of patient satisfaction, symptoms, coping ability, functioning, and general well-being.
Results
In both the in-person and telehealth groups, patients with MDD were highly satisfied with treatment and reported a significant reduction in symptoms from admission to discharge. Both groups also reported a significant improvement in positive mental health, general well-being, coping ability, and functioning. A large effect size of treatment was found in both treatment groups. Contrary to our hypothesis, the small differences in outcome favored the telehealth-treated patients. The length of stay and the likelihood of staying in treatment until completion were significantly greater in the virtually treated patients.
Limitations
The treatment groups were ascertained sequentially, and telehealth treatment was initiated after the COVID-19 pandemic began. Outcome assessment was limited to a self-administered questionnaire.
Conclusions
In an intensive acute care setting, delivering treatment to patients with MDD using a virtual, telehealth platform was as effective as treating patients in-person.
Keywords
Depression
Treatment
Telehealth
Virtual
Partial hospital
==== Body
pmc1 Introduction
Depression is one of the leading causes of psychosocial morbidity worldwide (Liu et al., 2020) and is responsible for excess mortality (Laursen et al., 2016). Depression is one of the most frequently treated disorders in primary care (Finley et al., 2018), and the most frequently diagnosed disorder treated in outpatient psychiatric practice (Zimmerman et al., 2008).
The COVID-19 pandemic has been depressogenic (Bueno-Notivol et al., 2020; Jia et al., 2022; Kessler et al., 2022; Morin et al., 2021; Salari et al., 2020; Stephenson et al., 2022). Major disruptions in lifestyle due to social isolation, job loss, financial strain, and deaths of neighbors, family and friends are potential contributors to the increased levels of depression due to the pandemic. As one of the core elements of psychotherapeutic approaches towards treating depression is behavioral activation and increased social contact (Forbes, 2020; Nagy et al., 2020), the psychosocial limitations imposed by COVID-19 might make it more difficult to treat depression during the pandemic.
The pandemic prompted recommendations for social distancing and other safety measures resulting in a rapid transition from in-person to telehealth behavioral health visits (Montoya et al., 2022; Wright and Caudill, 2020). Even before the pandemic, telehealth services for mental health treatment had already been recognized as a cost-effective way to increase accessibility to evidence-based treatments (Gros et al., 2013; Ralston et al., 2019). Reviews of the research literature suggest that telehealth treatment is generally acceptable, feasible, and comparable to in-person mental health services in improving symptoms of psychiatric disorders (Drago et al., 2016; Shigekawa et al., 2018). In addition to similar rates of improved clinical outcomes, an equally strong therapeutic alliance can be developed with telehealth visits as with in-person therapy (Simpson and Reid, 2014).
It is uncertain how long the COVID-19 pandemic will last. It is also uncertain what role telehealth treatment will continue to play in the delivery of ambulatory behavioral health treatment. While some states have mandated an expansion of telehealth services and required private payers to continue to reimburse telehealth services at the same level as in-person treatment, other states have already rescinded, or allowed to expire, emergency orders that required equivalent telehealth reimbursements. The ongoing and future reimbursement for telehealth services is likely to depend, in part, on research determining whether telehealth treatment is as safe and effective as in-person treatment.
Several studies have found that the treatment of depression with synchronous telehealth methods to be equally effective as in-person treatment (Choi et al., 2012; Egede et al., 2015; Luxton et al., 2016; Mohr et al., 2012; Ruskin et al., 2004). There are, however, several limitations to this literature. Many studies excluded patients with suicidal ideation or recent suicide attempts (Choi et al., 2012; Egede et al., 2015; Luxton et al., 2016; Mohr et al., 2012). Some studies limited the age range of the patients (Choi et al., 2012; Egede et al., 2015; Luxton et al., 2016). One study was of elderly, low income, home bound patients in which the in-person treatment was delivered in the patients' home, and the telehealth treatment began with a single in-person session (Choi et al., 2014). Three studies were conducted in Veterans Affairs medical centers; therefore, the sample composition was predominantly male (Egede et al., 2015; Luxton et al., 2016; Ruskin et al., 2004). All studies provided protocol-driven, manualized psychotherapy with a fixed number of sessions; medication, if prescribed, was at a stable dose before study entry and was not changed during the study. Thus, the studies to date comparing telehealth and in-person treatment of depression deviate in many ways from how depression is treated in usual clinical practice.
Furthermore, all research comparing the effectiveness of synchronous telehealth and in-person treatment of depression has been conducted in the context of outpatient, individual treatment settings. There is a paucity of research, in general, assessing the comparative efficacy of telehealth and in-person delivery in partial hospital and other intensive treatment settings. In partial hospital and intensive outpatient settings the level of severity and the risk of self-harm and suicidal behavior is generally greater than in outpatient practice thereby raising concerns as to whether telehealth treatment could be provided while maintaining patient safety. This is particularly important in the treatment of patients with depression where suicidal behavior is a concern.
Because of the COVID-19 pandemic, most treatment in ambulatory behavioral health settings has transitioned to a virtual format due to public health recommendations and legal guidelines for social distancing (Lewnard and Lo, 2020; Wright and Caudill, 2020). While this has impelled clinicians across settings to quickly adapt and make significant changes to the structure of their service delivery, partial hospital programs (PHPs) and intensive outpatient treatment programs confronted distinct concerns and obstacles (Hom et al., 2020; Inchausti et al., 2020). For instance, in working virtually with acutely and severely ill psychiatric patients who require a higher level of care than usual outpatient treatment, enhanced appropriate risk management is critical. Furthermore, for group therapy-based programs, additional considerations regarding privacy, confidentiality, and technological limitations are needed.
In the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project we previously examined the effectiveness of our in-person partial hospital treatment program utilizing an Acceptance and Commitment Therapy (ACT) treatment model, a well-established, inherently transdiagnostic, behavior therapy (Morgan et al., 2020). ACT is a third wave behavioral therapy treatment that has been demonstrated to be of benefit for patients with depression (Coto-Lesmes et al., 2020; Twohig and Levin, 2017; Washburn et al., 2021). In our transition to a completely telehealth-based program as a response to the COVID-19 pandemic, we continued to evaluate the effectiveness of treatment in our PHP in a diagnostically heterogeneous sample (Zimmerman et al., 2021). We did not alter the inclusion and exclusion criteria to the program upon transitioning from in-person to telehealth treatment. Certainly, we were concerned about patient safety, and we instituted procedures in the telehealth program to minimize risk. The addition of risk management strategies, while essential, can be perceived as intrusive and burdensome and reduce patient satisfaction.
In the present report from the MIDAS project, we focus on patients with major depressive disorder (MDD) who were treated in our PHP program. We compare the safety, effectiveness, and patient satisfaction with PHP services delivered via telehealth to in-person PHP treatment provided to patients treated prior to the COVID-19 outbreak. Because of the depressogenic nature of the COVID-19 epidemic we speculated that treatment might be less effective. If telehealth treatment during the pandemic was not demonstrably different in effectiveness as in-person treatment delivered prior to the pandemic, this would strengthen the evidence base for the safety and effectiveness of telehealth treatment of depression.
2 Methods
2.1 Setting
The Rhode Island Hospital Adult PHP (RIH PHP) is an acute care setting serving a range of presenting concerns referred from various clinical settings. A multidisciplinary team of psychiatrists, psychologists, clinical social workers, postdoctoral fellows, and doctoral level graduate student therapists delivered the treatment. All intake assessments, individual therapy and psychiatry visits, and group therapy sessions were conducted virtually using HIPAA-compliant real-time audio and visual computer-based communication using the Zoom virtual platform, business account version.
A minority of the patients in the PHP were interviewed by a diagnostic rater who administered the Structured Clinical Interview for DSM-IV (SCID) (First et al., 1997) and the borderline personality disorder section of the Structured Interview for DSM-IV Personality (SIDP-IV) (Pfohl et al., 1997). Most patients who presented for treatment were not evaluated with the semi-structured interviews because of a lack of available interviewers but were instead diagnosed by board-certified psychiatrists.
We previously described in detail our adaption of the program from in-person treatment to a virtual format (Zimmerman et al., 2021). In brief, the inclusion and exclusion criteria for admission to the PHP were not changed when we transitioned from the in-person to the virtual program. The therapeutic orientation of the RIH PHP is based on ACT and related evidence-based psychotherapy techniques (e.g. CBT, dialectical behavior therapy) delivered consistent with ACT principles (Morgan et al., 2020). The length of treatment is flexible, based on patients' symptoms, functioning and engagement in treatment. Patients meet with a therapist and psychiatrist daily or nearly every day for individual sessions, as well as attend multiple group therapy sessions. The content and structure of the telehealth group therapy sessions was consistent with the in-person treatment (Morgan et al., 2020).
All elements of PHP treatment, including the intake assessment, individual therapy, psychiatrist meeting, and group therapy sessions were conducted virtually using real-time audio and visual computer-based communication using Zoom as a platform for telehealth. Additional safety procedures were implemented to address the unique challenges of delivering treatment via telehealth in an acute care setting in which patients frequently present with safety concerns, including suicidal ideation, self-injurious behavior, and aggressive ideation and behavior. To address the challenge of tracking patient attendance and location in a virtual program, a daily check-in procedure was implemented in which patients initiate a Zoom call with support and administrative staff who record their attendance. A program attendance record is sent to the full multidisciplinary team along with daily updated email and physical address information for each patient. As patients are not physically present in the program and thus cannot be accompanied throughout the process of transferring to inpatient or emergency room care, this information was necessary to send emergency support to patients' residences in instances of worsening suicidal ideation or safety concerns. Patients were also required to identify an emergency contact support person and to submit release of information paperwork upon beginning the program.
Additionally, in the virtual program, a therapist assumed a clinical and technical oversight role. This person monitored all group sessions and was available at all times throughout program hours to respond by phone or Zoom visit to urgent patient needs, including needs for urgent clinical support outside of individual sessions, assistance with troubleshooting technical problems, and assistance in connecting patients to their individual providers.
2.2 Measures
After completion of the initial evaluation by the psychiatrist, the patients were asked to complete the Clinically Useful Patient Satisfaction Scale (CUPSS) (Zimmerman et al., 2017). The partial hospital version of the CUPSS includes an item assessing overall global satisfaction with the initial evaluation (Please rate your overall level of satisfaction with your initial visit with your doctor. 0 = not at all satisfied; 1 = slightly satisfied; 2 = moderately satisfied; 3 = very satisfied; 4 = extremely satisfied) and an item assessing expectation of improvement in the program (After the evaluation I was more hopeful I would get better. 0 = definitely not; 1 = probably not; 2 = not sure; 3 = probably yes; 4 = definitely yes). On the day of discharge from the program, the patients completed a satisfaction scale on which they rated their overall satisfaction with treatment (Please rate your overall level of satisfaction with the program: 0 = not at all satisfied; 1 = slightly satisfied; 2 = moderately satisfied; 3 = very satisfied; 4 = extremely satisfied), whether they would recommend the program to a friend or family, and their overall level of improvement (Compared to how you were feeling when you first started the program, at the time of ending do you feel: 0 = no better; 1 = slightly better; 2 = moderately better; 3 = a lot better; 4 = very much better).
The primary outcome measure in the present study was the depression subscale of a modified version of the Remission from Depression Questionnaire (RDQ-M) (Zimmerman et al., 2014). In contrast to most measures that assess only symptom presence during the past week or two, the RDQ assesses a broader array of features reported by patients as relevant to determining treatment outcome–symptoms, functioning, coping ability/stress tolerance, positive mental health, and general well-being/life satisfaction. The domains covered on the RDQ were based on a literature review, our previous study of patients' ratings of the relative importance of 16 factors in determining remission from depression (Zimmerman et al., 2006), and two focus groups. The depression subscale includes 14 items assessing the DSM-5 symptom criteria of MDD. As previously reported, the depression subscale had high internal consistency and test-retest reliability (Zimmerman et al., 2013). We modified the RDQ to accommodate use with patients with varied diagnoses as well as patients with multiple psychiatric disorders. Nineteen items were added to the original 41-item scale. The modified 60-item measure included 25 symptom items, 5 coping ability/stress tolerance items, 12 positive mental health items, 10 functioning items and 8 general well-being/life-satisfaction items. The time frame of the measure is the past week. The items are rated on a 3-point rating scale (not at all or rarely true; sometimes true; often or almost always true). The items are scored 0, 1, and 2 with higher scores indicating more severe symptomatology, better coping ability, more positive mental health, better functioning, and greater well-being. The internal consistency (Cronbach's alpha) of the RDQ-M subscales at discharge in PHP patients was high in both the in-person and telehealth samples (symptom scale 0.94 and 0.94; coping/stress tolerance subscale 0.75 and 0.74; positive mental health subscale 0.93 and 0.94; functioning subscale 0.86 and 0.88; well-being/life satisfaction; subscale 0.92 and 0.93).
As part of our usual clinical procedure the patients were asked to complete the RDQ-M at admission and discharge. During the in-person program, the RDQ-M was handed to the patients by their treating clinicians. In the virtual program, patients were sent links to complete the scales online. The Rhode Island Hospital institutional review committee approved the research protocol, and all patients provided informed consent to allow us to use the information collected for research purposes. Consent in the in-person program was obtained on a paper consent form, whereas in the virtual program it was obtained on an electronically signed form.
2.3 Data analysis
We used t-tests to compare the telehealth and in-person groups on continuously distributed variables and chi-square statistics to compare categorical variables. For each of the RDQ-M subscales, paired t-tests were used to compare follow-up scores to baseline values, and effect sizes (Cohen's d) were computed. Consistent with prior recommendations, an effect size of 0.2 was considered small, 0.5 medium, and 0.8 large (Cohen, 1988). Pre-post change scores were used to compare the amount of change from admission to discharge on the RDQ-M subscales between the in-person and telehealth groups.
3 Results
3.1 Patient characteristics
During an 18-month period from May 1, 2020, to December 1, 2021, 294 patients with a principal diagnosis of MDD were treated for the first time in the RIH PHP. Patients who had been previously treated in the program were not included in our analyses. The sample included patients who dropped out during the course of treatment as this was one of the outcomes of interest.
During the 18 months from May 1, 2018, to December 1, 2019, 542 patients were treated for the first time in the program and had a principal diagnosis of MDD. The in-person and telehealth-treated groups were similar in age, gender, race, and marital status (Table 1 ). Significantly more patients in the telehealth group had graduated from a 4-year college.Table 1 Demographic characteristics of partial hospital patients with major depressive disorder treated in-person or in a telehealth format.
Table 1 In-Person
(n = 542) Telehealth (n = 294) χ2 p level
Gender, % (n): 1.95 n.s.
Male 27.5 (149) 23.3 (68)
Female 69.4 (376) 74.0 (216)
Transgender or non-binary 3.1 (17) 2.7 (8)
Race, % (n): 6.80 n.s.
White 66.9 (362) 71.5 (208)
Hispanic 13.1 (71) 14.4 (42)
Black 7.8 (42) 7.2 (21)
Asian 3.5 (19) 1.7 (5)
Other 8.7 (47) 5.1 (15)
Education, % (n): 15.96 .001
Less than high school graduate 7.0 (38) 2.7 (7)
High school diploma or GED 60.2 (325) 51.7 (135)
4-year college degree 32.8 (177) 45.6 (119)
Marital status, % (n): 5.09 n.s.
Married 23.8 (129) 29.5 (86)
Living together 13.5 (73) 12.3 (36)
Widowed 3.7 (20) 2.1 (6)
Separated 3.1 (17) 2.1 (6)
Divorced 14.4 (78) 14.4 (42)
Never married 41.4 (224) 39.7 (116)
Agea, M (SD): 39.07 (15.26) 38.27 (14.17) t = 0.732 .47
Data missing Gender: 2 telehealth; Race: 1 in-person, 3 telehealth; Education: 2 in-person, 33 telehealth; Marital Status: 1 in-person, 2 telehealth.
n.s. indicates not significant.
a Age was compared by t-test.
About half of the patients in the telehealth and in-person groups were referred to the PHP by outpatient mental health clinicians (54.8 % vs. 47.8 %, X 2 = 3.71, NS). Less than 20 % of the patients in the telehealth and in-person cohorts were referred from inpatient psychiatric units (12.9 % vs. 18.1 %, X 2 = 3.72, NS), and >10 % were referred from emergency services (11.6 % vs. 16.4 %, X 2 = 3.58, NS).
The patients treated by telehealth were twice as likely to have been interviewed with the SCID and SIDP-IV (62.3 % vs 33.1 %, X 2 = 64.68, p < .01). When examining diagnostic frequencies, we limited our analyses to the patients interviewed with semi-structured interviews. Most patients were diagnosed with a comorbid disorder. The mean number of current disorders was significantly higher in the telehealth group (3.6 ± 1.5 vs. 3.2 ± 1.7, t = −2.12, p < .05). The patients in the telehealth cohort were significantly more often diagnosed with generalized anxiety disorder and borderline personality disorder (Table 2 ).Table 2 Current diagnoses of partial hospital patients treated in-person or in a telehealth format.
Table 2 In-person (n = 187) Telehealth (n = 185) χ2 p level
Anxiety disorders, % (n):
Panic disorder 11.2 (21) 12.4 (23) 0.13 .72
Panic disorder with agoraphobia 12.3 (23) 11.9 (22) 0.02 .90
Agoraphobia without panic 2.7 (5) 1.1 (2) 1.28 .26
Social anxiety disorder 31.0 (58) 27.0 (50) 0.79 .40
Specific phobia 5.3 (10) 5.4 (10) 0.00 .98
Posttraumatic stress disorder 25.7 (48) 33.5 (62) 2.75 .10
Generalized anxiety disorder 53.5 (100) 64.3 (119) 4.52 .03
Obsessive-compulsive disorder 5.3 (10) 5.9 (11) 0.06 .80
Body dysmorphic disorder 3.2 (6) 2.7 (5) 0.08 .77
Other anxiety disorder 11.2 (21) 15.1 (28) 1.24 .27
Substance use disorders, % (n):
Alcohol abuse/dependence 10.7 (20) 10.8 (20) 0.00 .97
Drug abuse/dependence 7.5 (14) 13.5 (25) 3.60 .06
Borderline personality disorder, % (n): 14.4 (27) 25.9 (48) 7.65 .006
Any eating disorder, % (n): 8.6 (16) 9.7 (18) 0.15 .70
Any somatoform disorder, % (n): 0.5 (1) 1.1 (2) 0.35 .56
Any impulse control disorder, % (n)a 9.6 (18) 10.3 (19) 0.04 .84
Any adjustment disorder, % (n): 1.6 (3) 1.1 (2) 0.19 .66
3.2 Patient satisfaction
The completion rate on the CUPSS after the initial evaluation by the psychiatrist was significantly lower in the telehealth cohort (55.1 % vs. 76.4 %, X 2 = 40.29, p < .01). We compared the demographic and diagnostic characteristics of the patients who did and did not complete the CUPSS. In the telehealth sample, those who did not complete the initial satisfaction scale were more likely to have a diagnosis of alcohol use disorder (16.7 % vs. 6.5 %, X 2 = 4.80, p < .05). In the in-person sample individuals who did not complete the initial satisfaction scale were more likely to have a diagnosis of drug use disorder (14.9 % vs. 5.0 %, X 2 = 4.97, p < .05). There were no other significant differences between the groups who did and did not complete the CUPSS.
Significantly more patients in the in-person sample indicated that they were very or extremely satisfied with the initial evaluation (87.9 % vs. 76.6 %, X 2 = 13.18, p < .01). The majority of patients in both the telehealth and the in-person groups were hopeful that they would get better (82.4 % vs. 76.4 %, X 2 = 1.61, NS). At the end of treatment, about 90 % of the patients in the telehealth and in-person groups reported being very or extremely satisfied with their treatment (87.7 % vs. 90.5 %, X 2 = 0.89, NS). More than 90 % of the patients treated in both formats indicated that they would recommend the treatment program to a friend or family member (93.9 % vs. 97.8 %, X 2 = 0.63, NS).
3.3 Program completion
The average number of days attending the program was significantly higher in the telehealth program (14.5 ± 8.2 vs. 8.8 ± 4.9, t = 12.25, p < .01). When limiting this analysis to patients who had completed treatment, the average length of stay remained significantly greater in the telehealth program (16.4 ± 7.9 vs. 10.7 ± 4.4, t = 11.03, p < .01). The average number of days missed while in treatment was low in both the telehealth and in-person programs (1.8 ± 2.5 vs. 1.6 ± 1.9, t = 0.92, NS).
A significantly higher proportion of patients completed treatment in the telehealth program (77.5 % vs. 67.2 %, X 2 = 9.57, p < .01). Consistent with this, a significantly higher percentage of patients were discharged from the in-person program due to nonattendance (13.4 % vs. 7.3 %, X 2 = 7.0, p < .01). Few patients in the telehealth and in-person programs required hospitalization (1.0 % vs. 2.3 %, X 2 = 1.56, NS). Likewise, only a small number of patients receiving telehealth and in-person treatment discontinued treatment because of dissatisfaction with the program (3.5 % vs. 3.2 %, X 2 = 0.04, NS). No patients attempted or completed suicide during their treatment in the program.
3.4 Treatment outcome
A significantly higher percentage of in-person patients completed the RDQ-M at admission (89.9 % vs. 75.8 %, X 2 = 29.67, p < .01) and at discharge (64.9 % vs. 55.2 %, X 2 = 7.65, p < .01). Complete outcome data was available for a higher percentage of in-person patients (61.8 % vs. 42.8 %, X 2 = 28.13, p < .01). In the telehealth group, patients with complete outcome data were more likely to have social anxiety disorder (36.9 % vs. 18.8 %, X 2 = 7.61, p < .01).
There were no significant differences between groups on the RDQ-M subscales at admission. For both the telehealth and in-person programs the patients significantly improved from admission to discharge on each of the RDQ-M subscales, with large effect sizes found for most of the subscales (Table 3 ). Change scores from admission to discharge were significantly greater in the telehealth group for the depression (−10.0 ± 6.8 vs. -8.3 ± 6.1, t = 2.74, p < .01) and anxiety (−3.8 ± 3.2 vs. -3.2 ± 3.0, t = 2.0, p < .05) subscales. On the discharge satisfaction scale, the majority of patients in the telehealth and in-person groups indicated that they were a lot or very much better at discharge (73.6 % vs. 75.0 %, X 2 = 0.80, NS).Table 3 Admission and discharge scores on Remission from Depression Questionnaire Modified (RDQ-M) subscales for patients with major depressive disorder treated in the partial hospital in-person or via telehealth.
Table 3RDQ-M subscale Admission
M (SD) Discharge
M (SD) Paired t-test Effect size (Cohen's d)
In-person group (n = 338)
Total symptoms subscale 31.7 (8.1) 17.6 (10.5) t = 24.8, p < .01 1.35
Depression 18.2 (4.5) 10.0 (6.0) t = 24.9, p < .01 1.35
Anxiety 7.4 (2.5) 4.2 (2.8) t = 19.7, p < .01 1.07
Anger 3.0 (2.0) 1.3 (1.6) t = 16.2, p < .01 0.88
Physical pain 3.2 (2.0) 2.1 (2.0) t = 9.6, p < .01 0.52
Positive mental health 5.9 (5.0) 13.2 (6.2) t = −20.4, p < .01 1.11
Functioning 7.6 (4.1) 12.5 (4.8) t = −17.1, p < .01 0.93
Coping skills 2.8 (2.1) 5.6 (2.5) t = −18.2, p < .01 1.00
Well-being 3.3 (3.3) 8.5 (4.5) t = −20.0, p < .01 1.08
Telehealth group (n = 127)
Symptoms 33.0 (7.7) 16.0 (10.6) t = 16.0, p < .01 1.41
Depression 18.9 (4.1) 8.8 (5.9) t = 16.7, p < .01 1.48
Anxiety 7.6 (2.4) 3.7 (2.8) t = 13.3, p < .01 1.18
Anger 3.3 (1.9) 1.4 (1.8) t = 10.0, p < .01 0.88
Physical pain 3.2 (1.9) 2.1 (1.8) t = 6.5, p < .01 0.58
Positive mental health 5.1 (3.7) 13.3 (6.2) t = −13.1, p < .01 1.17
Functioning 7.8 (4.0) 12.6 (4.7) t = −10.3, p < .01 0.91
Coping skills 2.8 (2.0) 6.2 (2.5) t = −11.6, p < .01 1.06
Well-being 3.1 (2.9) 8.7 (4.5) t = −13.5, p < .01 1.19
There were no significant differences between groups for scores on the RDQ-M subscales at admission.
Change scores from admission to discharge were significantly greater in the telehealth group for depression (10.0 ± 6.8 vs. 8.3 ± 6.1, t = 2.74, p = 0. < 01) and anxiety (3.8 ± 3.2 vs. 3.2 ± 3.0, t = 2.0, p = 0. < 05) subscales.
We also examined the 2 items on the RDQ-M that assessed suicidal ideation. From admission to discharge there was a significant reduction in the percentage of patients reporting death wishes (telehealth: 66.1 % vs. 14.2 %, X 2 = 7.50, p < .01; in-person: 60.4 % vs. 29.0 %, X 2 = 50.0, p < .001) and suicidal ideation during the past week (telehealth: 41.7 % vs. 10.2 %, X 2 = 11.0, p < .001; in-person: 47.3 % vs. 19.2 %, X 2 = 44.86, p < .001). At admission there was no difference between the telehealth and in-person groups in the frequency of patients reporting death wishes or suicidal ideation. At discharge half as many patients in the telehealth group continued to report death wishes (14.2 % vs. 29.0 %, X 2 = 10.83, p < .001) and suicidal ideation (10.2 % vs. 19.2 %, X 2 = 5.35, p < .05).
4 Discussion
In an intensive acute care setting consisting of daily group and individual therapy sessions as well medication treatment, providing care to patients with MDD using a virtual, telehealth platform was as effective as treating patients in-person. For both methods of delivering treatment, patients were largely satisfied with the initial diagnostic evaluation. Though fewer patients were satisfied with initial evaluation when it was conducted virtually, equal numbers were hopeful at admission that treatment would be beneficial. In both the in-person and telehealth groups there was a significant reduction in depressive symptoms and suicidality from admission to discharge, a significant reduction in symptoms of anxiety, anger, and pain, as well as improvement in functioning, coping ability, positive mental health, and general well-being. A large effect size of treatment was found in both treatment groups. Contrary to our hypothesis, the small differences in outcome favored the telehealth-treated patients. The length of stay and likelihood of staying in treatment until completion was greater in the telehealth treated patients.
In advance of the transition to the telehealth platform, our group discussed concerns about treating patients with MDD because of the associated suicide risk. In general, to qualify for partial hospital level of care patients need to be significantly functionally impaired, have failed to progress in outpatient treatment, and/or be at risk for self-harm. More than one-quarter of the patients were referred from inpatient units or emergency rooms. Thus, patients in a partial hospital program tend to be more severely and chronically ill than patients treated as outpatients. As we described in the Methods, we adopted procedural safeguards to reduce risk and attend to crises should they arise during the treatment day. Despite the severity and acuity of the patients' symptoms, no patient attempted suicide during the study. Moreover, the safeguards that were implemented did not reduce overall patient satisfaction with treatment.
Many studies have demonstrated that treatment delivered with synchronous audio and video transmission is as effective as in-person treatment; however, little research has examined telehealth treatment for patients requiring PHP level of care. Moreover, no prior study of PHP treatment has examined the acceptability, safety, and effectiveness of telehealth to treat patients with MDD. Unlike outpatient telehealth studies of individual therapy of MDD, many of the patients in the PHP reported suicidal ideation at admission to the program because the presence of suicidal ideation did not exclude patients from treatment. Precautions were taken to ensure that emergencies could be addressed in the virtually treated patients. Because a PHP is essentially an outpatient treatment setting, albeit more intensive than usual outpatient treatment in terms of the frequency of visits (5 days per week) and the duration of each visit (6 h per day), it is routine to assess risk and conduct safety planning interventions. We did not refuse admission of suicidal patients to our PHP, whether conducted virtually or in-person, unless a high level of intent was judged to be present whereupon the patient was referred for inpatient care. In fact, a small percentage of patients in both treatment formats were referred for inpatient admission though there was no significant difference between the formats in this regard. Likewise, there was no significant difference between the formats in the percentage of patients reporting suicidal thoughts at admission.
The present study was not a randomized controlled study comparing in-person and virtual PHP treatment. We transitioned to the virtual platform because of the COVID-19 pandemic, and we therefore examined the effectiveness of treatment in sequentially recruited cohorts. The only variable we controlled for was the time of year the patient was admitted to the PHP. Fortunately, there were few differences between the patient groups in demographic characteristics, comorbid psychiatric diagnoses, and baseline scores on the outcome measure. Of course, a randomized, controlled trial is the gold standard clinical trial design; however, it would be very costly to do such a study because it would require the doubling of clinical staff needed to run two simultaneous PHPs.
We were more successful collecting data when the patients were treated in-person. Handing paper-and-pencil questionnaires directly to patients by the treating clinician likely enhanced completion rates when compared to sending electronic links to surveys to be completed by patients online. We observed few differences in the demographic and clinical characteristics of the patients who did and did not complete the various measures. Social anxiety disorder was associated with the completion of outcome scales in the telehealth but not the in-person group. We are not aware of prior studies suggesting that social anxiety impacts the collection of outcome data.
In contrast to the lower level of cooperation in completing the questionnaires by the patients treated via telehealth, treatment participation was higher in telehealth-treated patients. Consistent with research in outpatient mental health clinics which found a lower “no show” rate for telehealth visits during the pandemic compared to in-person visits scheduled before the pandemic (Mishkind et al., 2021) we found that the patients in the virtual program more frequently completed treatment and more patients were discharged from in-person treatment due to nonattendance. We would hypothesize that nonattendance was greater in the in-person program because of transportation issues and difficulty waking up and being sufficiently motivated to arrive on time to the program.
Patients in the virtual program were treated for more days. The longer duration of treatment and greater completion rate in the telehealth group may have been artefacts of the COVID-19 pandemic. COVID-19 inspired social distancing recommendations increased social isolation and for some patients, attendance in the PHP was a primary source of social engagement. This may have increased some patients' desire to stay in the virtual program for a longer amount of time. The pandemic affected some patients' job status, with some having been furloughed or laid off—these patients were less pressured to be discharged in order to return to work. Early in the pandemic, health insurance companies' utilization review procedures were suspended thereby reducing pressure to discharge patients sooner than clinicians would have liked. An indirect contributor to the longer duration of treatment and greater treatment completion rate in the telehealth patients was the elimination of patient travel. Clinicians may have been more hesitant to discharge patients with some suicidal ideation in the telehealth program and thus kept them longer until the suicidal ideation more completely resolved. Patients treated virtually were diagnosed with more comorbid diagnoses and significantly more often diagnosed with borderline personality disorder and perhaps this resulted in longer duration of treatment because of lower effectiveness of treatment. Post hoc analyses did not find that in the pre-COVID in-person sample the number of diagnoses was associated with treatment completion. Finally, it is also possible that the response to treatment was slower with the telehealth format, and this resulted in a longer duration of treatment.
A limitation of the study is that outcome was only evaluated with self-administered questionnaires, and we did not include clinician rating scales. However, previous research from the MIDAS project found that the effect size of treatment was similar when based on self-report scales and clinician-administered measures (Zimmerman et al., 2018).
A limitation of comparing treatment effectiveness in sequentially treated cohorts is that circumstances unrelated to treatment efficacy could impact treatment outcomes. We adopted the telehealth format out of necessity due to the COVID-19 pandemic. Our uncertainty about the effectiveness and safety of virtual treatment during the pandemic motivated the study. The cohorts thus differ in two ways—how treatment was delivered and the altered social zeitgeist due to the pandemic. The pandemic has had a negative impact on the mental health of the general population (Bueno-Notivol et al., 2020; Jia et al., 2022; Kessler et al., 2022; Morin et al., 2021; Salari et al., 2020; Stephenson et al., 2022), as well as psychiatric patients (Dalkner et al., 2022; Fleischmann et al., 2021; Lewis et al., 2022). We speculated that the impact of the pandemic on social engagement, employment, education, and parental responsibilities would impede achieving positive treatment outcomes in the telehealth group. Fortunately, there was no evidence of inferior outcome in the telehealth cohort despite being treated while dealing with the COVID-19 pandemic. Nonetheless, to enhance confidence that a telehealth PHP is as effective as in-person treatment for MDD it will be important to compare treatment formats when pandemic-related issues have subsided.
Finally, a few words about the future. We are hopeful that pandemic will end in the not-too-distant future. However, we are unsure of how low a level of community spread will be required before returning to an in-person program largely based on group therapy. Several patients whom we have treated virtually have commented that they would not have presented for in-person treatment even if there was no pandemic. Some of these patients had medical illnesses that made in-person treatment attendance more difficult to manage. For some patients, limited transportation options made in-person treatment more difficult. Thus, we hope that telehealth partial hospital treatment is here to stay. Of course, decisions about how care is delivered in the future likely will be determined by insurance reimbursement. Hopefully, regulations will be adopted requiring equal compensation for treatment (Zimmerman, 2022). In the absence of such regulations, we fear that insurance companies will eliminate coverage for telehealth treatment thereby reducing access. If this occurs, it will be done despite considerable evidence that telehealth behavioral treatment is as effective as in-person care.
Role of funding source
None.
CRediT authorship contribution statement
Mark Zimmerman designed the study, wrote the first draft of the manuscript, and directed the data analysis. Catherine D'Avanzato contributed to the writing of the first draft and reviewed the draft of the entire manuscript and provided feedback that was incorporated into the final submission. Brittany King managed and analyzed the data, conducted the evaluations, and reviewed the draft of the manuscript and provided feedback to Dr. Zimmerman that was incorporated into the final submission.
Conflict of interest
None.
Acknowledgments
None.
==== Refs
References
Bueno-Notivol J. Gracia-García P. Olaya B. Lasheras I. López-Antón R. Santabárbara J. Prevalence of depression during the COVID-19 outbreak: a meta-analysis of community-based studies Int. J. Clin. Health Psychol. 21 2020 1 11
Choi I. Zou J. Titov N. Dear B.F. Li S. Johnston L. Andrews G. Hunt C. Culturally attuned internet treatment for depression amongst Chinese Australians: a randomised controlled trial J. Affect. Disord. 136 2012 459 468 22177742
Choi N.G. Hegel M.T. Marti N. Marinucci M.L. Sirrianni L. Bruce M.L. Telehealth problem-solving therapy for depressed low-income homebound older adults Am. J. Geriatr. Psychiatry 22 2014 263 271 23567376
Cohen J. Statistical Power Analysis for the Behavioral Sciences 2nd ed. 1988 Lawrence Erlbaum Associates Hillsdale, NJ
Coto-Lesmes R. Fernandez-Rodriguez C. Gonzalez-Fernandez S. Acceptance and commitment therapy in group format for anxiety and depression.A systematic review J. Affect. Disord. 263 2020 107 120 31818766
Dalkner N. Ratzenhofer M. Fleischmann E. Fellendorf F.T. Bengesser S. Birner A. Maget A. Grossschadl K. Lenger M. Platzer M. Queissner R. Schonthaler E. Tmava-Berisha A. Berndt C. Martini J. Bauer M. Sperling J.D. Vinberg M. Reininghaus E.Z. Psychological and behavioral response on the COVID-19 pandemic in individuals with bipolar disorder: a multicenter study Psychiatry Res. 310 2022 114451
Drago A. Winding T.N. Antypa N. Videoconferencing in psychiatry, a meta-analysis of assessment and treatment Eur. Psychiatry 36 2016 29 37 27311105
Egede L.E. Acierno R. Knapp R.G. Lejuez C. Hernandez-Tejada M. Payne E.H. Frueh B.C. Psychotherapy for depression in older veterans via telemedicine: a randomised, open-label, non-inferiority trial Lancet Psychiatry 2 2015 693 701 26249300
Finley C.R. Chan D.S. Garrison S. Korownyk C. Kolber M.R. Campbell S. Eurich D.T. Lindblad A.J. Vandermeer B. Allan G.M. What are the most common conditions in primary care? Systematic review Can. Fam. Physician 64 2018 832 840 30429181
First M.B. Spitzer R.L. Williams J.B.W. Gibbon M. Structured Clinical Interview for DSM-IV (SCID) 1997 American Psychiatric Association Washington, D.C.
Fleischmann E. Dalkner N. Fellendorf F.T. Reininghaus E.Z. Psychological impact of the COVID-19 pandemic on individuals with serious mental disorders: a systematic review of the literature World J. Psychiatry 11 2021 1387 1406 35070784
Forbes C.N. New directions in behavioral activation: using findings from basic science and translational neuroscience to inform the exploration of potential mechanisms of change Clin. Psychol. Rev. 79 2020 101860
Gros D.F. Morland L.A. Greene C.J. Acierno R. Strachan M. Egede L.E. Tuerk P.W. Myrick H. Frueh B.C. Delivery of evidence-based psychotherapy via video telehealth J. Psychopathol. Behav. Assess. 35 2013 506 521
Hom M.A. Weiss R.B. Millman Z.B. Christensen K. Lewis E.J. Cho S. Yoon S. Meyer N.A. Kosiba J.D. Shavit E. Schrock M.D. Levendusky P.G. Björgvinsson T. Development of a virtual partial hospital program for an acute psychiatric population: lessons learned and future directions for telepsychotherapy J. Psychother. Integr. 30 2020 366 382
Inchausti F. Macbeth A. Hasson-Ohayon I. Dimaggio G. Telepsychotherapy in the age of COVID-19: a commentary J. Psychother. Integr. 30 2020 394 405
Jia R. Ayling K. Chalder T. Massey A. Gasteiger N. Broadbent E. Coupland C. Vedhara K. The prevalence, incidence, prognosis and risk factors for symptoms of depression and anxiety in a UK cohort during the COVID-19 pandemic BJPsych Open 8 2022 e64
Kessler R.C. Chiu W.T. Hwang I.H. Puac-Polanco V. Sampson N.A. Ziobrowski H.N. Zaslavsky A.M. Changes in prevalence of mental illness among US adults during compared with before the COVID-19 pandemic Psychiatr. Clin. North Am. 45 2022 1 28 35219431
Laursen T.M. Musliner K.L. Benros M.E. Vestergaard M. Munk-Olsen T. Mortality and life expectancy in persons with severe unipolar depression J. Affect. Disord. 193 2016 203 207 26773921
Lewis K.J.S. Lewis C. Roberts A. Richards N.A. Evison C. Pearce H.A. Lloyd K. Meudell A. Edwards B.M. Robinson C.A. Poole R. John A. Bisson J.I. Jones I. The effect of the COVID-19 pandemic on mental health in individuals with pre-existing mental illness BJPsych Open 8 2022 e59
Lewnard J.A. Lo N.C. Scientific and ethical basis for social-distancing interventions against COVID-19 Lancet Infect. Dis. 20 2020 631 633 32213329
Liu Q. He H. Yang J. Feng X. Zhao F. Lyu J. Changes in the global burden of depression from 1990 to 2017: findings from the global burden of disease study J. Psychiatr. Res. 126 2020 134 140 31439359
Luxton D.D. Pruitt L.D. Wagner A. Smolenski D.J. Jenkins-Guarnieri M.A. Gahm G. Home-based telebehavioral health for U.S. military personnel and veterans with depression: a randomized controlled trial J. Consult. Clin. Psychol. 84 2016 923 934 27599225
Mishkind M.C. Shore J.H. Bishop K. D'amato K. Brame A. Thomas M. Schneck C.D. Rapid conversion to telemental health services in response to COVID-19: experiences of two outpatient mental health clinics Telemed. J. e-Health 27 2021 778 784 33393857
Mohr D.C. Ho J. Duffecy J. Reifler D. Sokol L. Burns M.N. Jin L. Siddique J. Effect of telephone-administered vs face-to-face cognitive behavioral therapy on adherence to therapy and depression outcomes among primary care patients: a randomized trial JAMA 307 2012 2278 2285 22706833
Montoya M.I. Kogan C.S. Rebello T.J. Sadowska K. Garcia-Pacheco J.A. Khoury B. Kulygina M. Matsumoto C. Robles R. Huang J. Andrews H.F. Ayuso-Mateos J.L. Denny K. Gaebel W. Gureje O. Kanba S. Mare K. Medina-Mora M.E. Pike K.M. Roberts M.C. Sharan P. Stein D.J. Scott Stroup T. Zhao M. Reed G.M. An international survey examining the impact of the COVID-19 pandemic on telehealth use among mental health professionals J. Psychiatr. Res. 148 2022 188 196 35131587
Morgan T. Dalrymple K. D'avanzato C. Zimage S. Balling C. Ward M. Zimmerman M. Conducting outcomes research in a clinical practice setting: the effectiveness and acceptability of acceptance and commitment therapy (ACT) in a partial hospital program Behav. Ther. 52 2020 272 285 33622499
Morin C.M. Bjorvatn B. Chung F. Holzinger B. Partinen M. Penzel T. Ivers H. Wing Y.K. Chan N.Y. Merikanto I. Mota-Rolim S. Macedo T. De Gennaro L. Leger D. Dauvilliers Y. Plazzi G. Nadorff M.R. Bolstad C.J. Sieminski M. Benedict C. Cedernaes J. Inoue Y. Han F. Espie C.A. Insomnia, anxiety, and depression during the COVID-19 pandemic: an international collaborative study Sleep Med. 87 2021 38 45 34508986
Nagy G.A. Cernasov P. Pisoni A. Walsh E. Dichter G.S. Smoski M.J. Reward network modulation as a mechanism of change in behavioral activation Behav. Modif. 44 2020 186 213 30317863
Pfohl B. Blum N. Zimmerman M. Structured Interview for DSM-IV Personality 1997 American Psychiatric Press, Inc. Washington, DC
Ralston A.L. Andrews A.R. Hope D.A. Fulfilling the promise of mental health technology to reduce public health disparities: review and research agenda Clin. Psychol. 26 2019 1 14
Ruskin P.E. Silver-Aylaian M. Kling M.A. Reed S.A. Bradham D.D. Hebel J.R. Barrett D. Knowles F. III Hauser P. Treatment outcomes in depression: comparison of remote treatment through telepsychiatry to in-person treatment Am. J. Psychiatry 161 2004 1471 1476 15285975
Salari N. Hosseinian-Far A. Jalali R. Vaisi-Raygani A. Rasoulpoor S. Mohammadi M. Rasoulpoor S. Khaledi-Paveh B. Prevalence of stress, anxiety, depression among the general population during the COVID-19 pandemic: a systematic review and meta-analysis Global Health 16 2020 57 32631403
Shigekawa E. Fix M. Corbett G. Roby D.H. Coffman J. The current state of telehealth evidence: a rapid review Health Aff. (Millwood) 37 2018 1975 1982 30633674
Simpson S.G. Reid C.L. Therapeutic alliance in videoconferencing psychotherapy: a review Aust. J. Rural Health 22 2014 280 299 25495622
Stephenson E. O'Neill B. Kalia S. Ji C. Crampton N. Butt D.A. Tu K. Effects of COVID-19 pandemic on anxiety and depression in primary care: a retrospective cohort study J. Affect. Disord. 303 2022 216 222 35139415
Twohig M.P. Levin M.E. Acceptance and commitment therapy as a treatment for anxiety and depression: a review Psychiatr. Clin. N. Am. 40 2017 751 770
Washburn M. Yu M. Rubin A. Zhou S. Web-based acceptance and commitment therapy (ACT) for symptoms of anxiety and depression: within-group effect size benchmarks as tools for clinical practice J. Telemed. Telecare 27 2021 314 322 33966524
Wright J.H. Caudill R. Remote treatment delivery in response to the COVID-19 pandemic Psychother. Psychosom. 89 2020 130 132 32213775
Zimmerman M. Should blind psychiatrists be paid less? J. Clin. Psychiatry 83 2022 14354
Zimmerman M. Mcglinchey J. Posternak M. Friedman M. Attiullah N. Boerescu D. How should remission from depression be defined? The depressed patient's perspective Am. J. Psychiatry 163 2006 148 150 16390903
Zimmerman M. Mcglinchey J.B. Chelminski I. Young D. Diagnostic co-morbidity in 2300 psychiatric out-patients presenting for treatment evaluated with a semi-structured diagnostic interview Psychol. Med. 38 2008 199 210 17949515
Zimmerman M. Martinez J. Attiullah N. Friedman M. Toba C. Boerescu D. Ragheb M. A new type of scale for determining remission from depression: the remission from depression questionnaire J. Psychiatr. Res. 47 2013 78 82 23102820
Zimmerman M. Martinez J.H. Attiullah N. Friedman M. Toba C. Boerescu D.A. The remission from depression questionnaire as an outcome measure in the treatment of depression Depress. Anxiety 31 2014 533 538 24115164
Zimmerman M. Gazarian D. Multach M.D. Attiullah N. Benoff T. Boerescu D.A. Friedman M.A. Mehring L.B. Moon S. Patel S. Guzman Holst C. A clinically useful self-report measure of psychiatric patients' satisfaction with the initial evaluation Psychiatry Res. 252 2017 38 44 28242516
Zimmerman M. Walsh E. Friedman M. Boerescu D.A. Attiullah N. Are self-report scales as effective as clinician rating scales in measuring treatment response in routine clinical practice? J. Affect. Disord. 225 2018 449 452 28858659
Zimmerman M. Terrill D. D'avanzato C. Tirpak J.W. Telehealth treatment of patients in an intensive acute care psychiatric setting during the COVID-19 pandemic: comparative safety and effectiveness to in-person treatment J. Clin. Psychiatry 82 2021
| 36521663 | PMC9742046 | NO-CC CODE | 2022-12-15 23:17:45 | no | J Affect Disord. 2023 Feb 15; 323:624-630 | utf-8 | J Affect Disord | 2,022 | 10.1016/j.jad.2022.12.015 | oa_other |
==== Front
Psychiatry Res
Psychiatry Res
Psychiatry Research
0165-1781
1872-7123
Elsevier B.V.
S0165-1781(22)00600-X
10.1016/j.psychres.2022.115009
115009
Letter to the Editor
COVID-19 induced psychosis: A case report
Ellini Sana 12⁎
Romdhane Imen Ben 12
Bougacha Dhouha 12
Abassi Ameni 12
Cheour Majda 12
Damak Rahma 12
1 Department of adult psychiatry "Ibn Omrane" in The University psychiatric Hospital Razi in Manouba. Tunis. Tunisia
2 Faculty of Medicine of Tunis. University of Tunis El Manar. Tunis. Tunisia
⁎ Corresponding author.
12 12 2022
12 12 2022
11500920 6 2021
9 12 2022
10 12 2022
© 2022 Elsevier B.V. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Keywords
Coronavirus
COVID-19
psychosis
Tunisia
==== Body
pmcDear Editor,
Since the corona virus disease was declared a pandemic in March 2020, more disparate clinical phenotypes are reported. The pandemic situation is considered a major psychosocial stressor due to the radical changes it brought to the general population such as social isolation, economic crises and the fear of exposure and death. Whereas the mental health challenges caused by the COVID-19 are established, the virus-specific neuropsychiatric effects on both short and long terms are yet to be elucidated. Nowadays, very few papers on acute psychiatric manifestations attributed to the COVID-19 infection are emerging. Here, we report the case of an acute psychotic episode in a 31-year-old single male from Tunisia, with no family history of mental illness. He was diagnosed with mild intellectual disability related to DSM-5 criteria and for whom the intelligence quotient was <69 according to the Weshler intelligence scale. He did not smoke and he did not consume alcohol or illicit substances. He had one prior psychiatric hospitalization in 2012 for reactive behavioral disorders related to character disorders, treated with chlorpromazine 50 mg per day. The patient had self-discontinued his medication in 2015 and remained stable for six years. On May 4th, 2021, he was admitted to the psychiatric inpatient's unit for severe agitation and aggression. His family noted an acute, rapidly progressive change in his behavior characterized by extreme anxiety, weird behavior, incoherent speech, persecutory thought patterns, food refusal and decreased sleep four days preceding the hospitalization. At the time of his initial psychiatric assessment, the patient was agitated, suspicious and extremely anxious. He had disorganized behavior and discourse. His speech was marked by perseverations and conveyed religious and paranoid delusions as well as thoughts of guilt and divine punishment. He believed that someone would kill his parents and was seeking absolution by food refusal. He also had auditory hallucinations with negative content and terrifying visual hallucinations including zoopsia. He was fully oriented and did not seem to have neither attention nor memory impairments. We did not observe characteristic fluctuations in the level of alertness, apart from a discreet worsening of the symptoms at night. His physical examination was normal, apart from a tachycardia. His blood pressure was 100/60 mmHg. His body temperature was 36,9°C and respiratory rate was 15 breaths/min. His peripheral oxygen saturation was 99%. Initial routine laboratory tests revealed elevated peripheral markers of inflammation, rhabdomyolysis, and dehydration. Full blood count showed a mild leukocytosis (11.7 × 103 /μL, 74% neutrophils). The C - reactive protein (CRP) was elevated to 175 mg/L (reference range <10 mg/L). The creatine phosphokinase (CK) was 7265 UI/L (reference range <220 UI/L). The aspartate aminotransferase (AST) was 116 UI/L (reference range <40 UI/L). Initial laboratory workup was also notable for hypernatremia of 152 mmol/L and elevated blood urea nitrogen (BUN) of 10.4 mmol/L (reference range <7.5 mmol/L). Creatinine clearance was normal (83.7 ml/min). Also, we carried out a cerebral scanner which was without anomaly and an Electro-encephalogram which did not detect any anomalies.
Given these laboratory abnormalities, a COVID-19 infection was suspected. The patient was tested for SARS-CoV-2 with nasopharyngeal swab and the PCR test returned positive. Interestingly, he reported no respiratory or gastrointestinal symptoms. Also he had neither anosmia nor ageusia having appeared before or during the hospitalization. He was put in isolation and initially treated with three daily intramuscular injections of 1 mg of clonazepam for acute anxiolysis with the introduction of risperidone at a rate of 2 mg per day. In addition, due to signs of dehydration in this non cooperating patient, intravenous rehydration for three days was performed. As he remained asymptomatic, he did not need any specific therapy for his COVID-19 infection.
The evolution of the psychiatric symptoms was favorable with disappearance of anxiety and the improvement of psychomotor agitation and resumption of a normal diet on the 5th day of his hospitalization. On the 7th day, there was a complete cleaning of the symptoms marked by a distancing of delusional words and a resumption of an organized behavior and a coherent thought brought out through his logical discourse. During the follow-up, we observed a temporal link between the resolution of the psychiatric symptoms and the improvement of the biological disturbances concerning rhabdomyolysis and the peripheral inflammatory markers. The duration of the period of illness did not exceed fifteen days in total. The patient was discharged on risperidone at the same initial prescribed dose (2 mg per day) but with a reduction in the doses of clonazepam which were planned to be gradually stopped at the post treatment period.
So, the case we have presented is testimony to the existence of direct links between COVID-19 and psychosis. The diagnosis of psychotic disorder induced by an organic affection which is COVID-19 was retained according to the criteria of the DSM-5. Although, it is always possible that psychiatric disorders are due to physical problems such as dehydration and inflammation or to resulting psychological reactions and not to COVID-19 per se, the hypothesis of direct links is very likely. The clinical cases published on this subject to date remain rare. The vast majority of cases reported were mainly triggered by psychosocial stressors related to COVID pandemic (Huarcaya-Victoria et al., 2020). Nevertheless, our patient had never reported COVID preoccupations or concerns. In a series of ten Spanish patients diagnosed with a recent onset psychotic episode linked to COVID-19 (Parra et al., 2020), only one patient who had a schizoid personality had no physical symptoms pointing to this infection with COVID-19. Also, he had no associated biological signs. But in his case, his behavioral disorders could be secondary on the one hand to the infection by COVID-19, and on the other hand to the significant stress felt due to a severe form of the COVID-19 disease in his parents. Our case is surprisingly like those recently described by Ferrando in New York, who reported three recent cases of paranoid psychosis in asymptomatic COVID-19 patients (Ferrando et al., 2020). All three patients were incidentally found to have positive SARS-CoV-2 tests, and elevated peripheral inflammatory markers. Similarly, a case of a 52-year-old man with a medical history of obstructive sleep apnea, without a significant psychiatric history, who exhibited disorganized behavior and paranoia leading to a suicide attempt was reported (Chacko et al., 2020). This patient had high inflammatory markers without any clinical signs in favor of COVID-19 disease.
Also, a case of a schizophreniform disorder secondary to SARS CoV-2 infection has been published (DeLisi, L. E., 2021). In this research study, the author has emphasized the viral hypothesis at the origin of schizophrenia. According to this author, the viral hypothesis of schizophrenia has persisted for decades even though no specific virus has been implicated, and the current pandemic that is spreading the corona virus SARS CoV-2 worldwide is now showing anecdotal evidence of newly developed psychoses after viral exposure, involving neuronal inflammation in crucial areas of the brain that could trigger psychotic symptoms.
Through the presentation of our case, and the other studies the direct implication of COVID-19 infection in the emergence of psychosis seems to be relevant. We conclude that recent psychosis in a patient without obvious precipitating factors should suggest COVID-19 infection in such a pandemic context. Further research is recommended to study the direct links between COVID19 and psychosis.
Ethical considerations
Informed consent was obtained from the post-recovery patient and his family for case presentation
Conflict of interest
The authors declare: no conflict of interest.
Declarations of interest: none
==== Refs
References
Huarcaya-Victoria J. Herrera D. Castillo C. Psychosis in a patient with anxiety related to COVID-19: a case report Psychiatry research 289 2020 113052
Parra A Juanes A Losada CP Álvarez-Sesmero S Santana VD Martí I Psychotic symptoms in COVID-19 patients. A retrospective descriptive study Psychiatry Res 291 2020 113254 10.1016/j.psychres.2020.113254
Ferrando S.J. Klepacz L. Lynch S. Tavakkoli M. Dornbush R. Baharani R. Bartell A. COVID-19 psychosis: a potential new neuropsychiatric condition triggered by novel coronavirus infection and the inflammatory response? Psychosomatics 61 5 2020 551 32593479
Chacko M. Job A. Caston F. George P. Yacoub A. Cáceda R. COVID-19-induced psychosis and suicidal behavior: case report SN comprehensive clinical medicine 2 11 2020 2391 2395 33015547
DeLisi L.E. A commentary revisiting the viral hypothesis of schizophrenia: Onset of a schizophreniform disorder subsequent to SARS CoV-2 infection Psychiatry Research 295 2021 113573 10.1016/j.psychres.2020.113573
| 0 | PMC9742060 | NO-CC CODE | 2022-12-14 23:54:17 | no | Psychiatry Res. 2022 Dec 12;:115009 | utf-8 | Psychiatry Res | 2,022 | 10.1016/j.psychres.2022.115009 | 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)00698-3
10.1016/j.jinf.2022.12.006
Letter to the Editor
Changes of coagulase-negative Staphylococci infections in children before and after the COVID-19 pandemic in Zhengzhou, China
Li Lifeng 1⁎
Ma Jiayue 1
Guo Pengbo 1
Gao Kaijie 2
Yang Junmei 2
Sun Huiqing 1⁎⁎
1 Henan International Joint Laboratory of Children's Infectious Diseases, Department of Neonatology, Children's Hospital Affiliated to Zhengzhou University, Henan Children's Hospital, Zhengzhou Children's Hospital, Zhengzhou, China
2 Zhengzhou Key Laboratory of Children's Infection and Immunity, Department of Laboratory Medicine, Children's Hospital Affiliated to Zhengzhou University, Henan Children's Hospital, Zhengzhou Children's Hospital, Zhengzhou, China
⁎ Corresponding author: Dr. Lifeng Li, Henan International Joint Laboratory of Children's Infectious Diseases, Neonatology, Children's Hospital Affiliated to Zhengzhou University, Zhengzhou, China, Phone: +86-373-85515773, Fax: +86-373-85515773
⁎⁎ Corresponding author: (Huiqing Sun)
12 12 2022
12 12 2022
8 12 2022
© 2022 The British Infection Association. Published by 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.
==== Body
pmcDear editor,
We read with great interest the findings of Li et al.1 and Zhou et al.2 published in the Journal of infection, which reported a decline in Streptococcus pneumoniae and Haemophilus influenzae infections in children under the influence of COVID-19 pandemic. Li et al.3 found that COVID-19 had a greater impact on Escherichia coli isolated from children with respiratory infections than children with digestive system infections. Decreased carbapenemase-producing Enterobacteriaceae incidence after COVID-19 pandemic was also reported by Duverger et al.4. However, there is no report focusing on the prevalence of coagulase-negative Staphylococci infections in children before and after COVID-19.
Coagulase-positive Staphylococci (CoPS) and coagulase-negative Staphylococci (CoNS) are two major types of the genus Staphylococci, which has become a serious challenge to the global public health system due to increasing antibiotic resistance5. For a long time, studies have mainly focused on CoPS such as Staphylococcus aureus, while few research has focused on CoNS. People start to pay attention to CoNS with the increasing detection rates of CoNS in clinical practice e.g., blood cultures or other sterile samples5 , 6. CoNS consists of a heterogeneous and ever-increasing group of species that colonize the skin and mucosa of their hosts, which could also cause invasive diseases including bloodstream infection, urinary tract infection, and endocarditis6. Common CoNS species include Staphylococcus epidermidis, Staphylococcus haemolyticus, Staphylococcus capitis, Staphylococcus hominis, etc7. The COVID-19 pandemic has significantly changed human lifestyles worldwide. Infection prevention and control measures (hand hygiene, wearing masks and limitation of gatherings) were taken to control the spread of SARS-CoV-2. These management measures could also reduce the transmission of many other pathogens1, 2, 3, 4 , 8. Therefore, analysis of the prevalence trend of CoNS infection in children before and after the COVID-19 pandemic would be important for the prevention and control of related infectious diseases.
To evaluate the impact of COVID-19 on the epidemiological characteristics of CoNS infection in children, laboratory data were collected and analyzed for children with blood culture records in Henan Children's hospital from January 2018 to November 2022. The total numbers of blood culture cases, the total positive numbers and the total positive rates were analyzed (Fig. 1 ). A total of 98, 245 children were included (n=22, 196 in 2018, n=23, 047 in 2019, n=15, 216 in 2020, n=19,735 in 2021 and n=18, 051 in 2022) (Fig. 1A). The total number of blood culture positive cases was 1, 410 (n=421 in 2018, n=353 in 2019, n=223 in 2020, n=195 in 2021 and n=218 in 2022). Significant decrease in the total cases was found in 2020 (the first year of COVID-19 pandemic), which increased in 2021 and 2022, whereas the cases were still lower than that in 2018 and 2019. The positive cases and the total positive rates showed a decreasing trend from 2018 to 2021 (Fig. 1B). In 2022, there was a slight increase in the positive rates, but it was still lower than that before COVID-19. These results indicated significant impact of COVID-19 pandemic on the epidemiological characteristics of children with bloodstream infection.Fig. 1 The total number of cases and the number of positive cases (A), and the positive rates (B) in children with blood cultures from 2018-2022.
Fig 1
Then, the culture-positive cases were analyzed to reveal the pathogen distribution with special focus on the epidemic trend of CoNS infection in children before and after the COVID-19 epidemic. The pathogens mainly included S. epidermidis, S. hominis, S. capitis, S. haemolyticus, S. aureus, E. coli, Enterococcus faecium, S. pneumoniae, Klebsiella pneumoniae, and Pseudomonas aeruginosa (Table 1 ). These pathogens accounted for more than 55% of the total pathogens detected. Among these pathogens, S. epidermidis and S. hominis accounted for a higher proportion. The positive cases (108 in 2018 and 15 in 2022) and the positive rates (25.65% in 2018 and 6.88% in 2022) of S. epidermidis showed continuous decrease before and after the COVID-19 pandemic (Fig. 2 A). The positive cases (76 in 2018 and 16 in 2021) and the positive rates (18.05% in 2018 and 8.21% in 2022) of S. hominis decreased from 2018 to 2021, which increased slightly in 2022 (28 and 12.84%) (Fig. 2B). The proportion of S. haemolyticus and S. capitis was low, but reduced infection rates were indicated in 2020. E. coli also showed decreased positive rates from 2019 to 2022. The other pathogen showed no significant increase or decrease in the years analyzed. The main strains of CoNS were analyzed together (Fig. 2C), and the positive rates decreased from 2018 to 2022. In addition, all Staphylococcus detected were analyzed, which showed a decline trend from 2018 to 2021, but increased slightly in 2022 (Fig. 2D). According to the above analysis, significant effect of COVID-19 on the prevalence of CoNS infections especially S. epidermidis and S. hominis were indicated.Table 1 The pathogen distribution in culture-positive children from 2018 to 2022.
Table 1Pathogen 2018 (n=421) 2019 (n=353) 2020 (n=223) 2021 (n=195) 2022 (n=218)
Staphylococcus epidermidis 108 (25.65%) 65 (18.41%) 34 (15.25%) 18 (9.23%) 15 (6.88%)
Staphylococcus hominis 76 (18.05%) 53 (15.01%) 28 (12.56%) 16 (8.21%) 28 (12.84%)
Staphylococcus haemolyticus 16 (3.80%) 11 (3.12%) 3 (1.35%) 8 (4.10%) 4 (1.83%)
Staphylococcus capitis 5 (1.19%) 7 (1.98%) 4 (1.79%) 6 (3.08%) 3 (1.38%)
Staphylococcus aureus 16 (3.80%) 27 (7.65%) 26 (11.66%) 18 (9.23%) 31 (14.22%)
Streptococcus pneumoniae 25 (5.94%) 23 (6.52%) 9 (4.04%) 16 (8.21%) 11 (5.05%)
Enterococcus faecium 9 (2.14%) 9 (2.55%) 6 (2.69%) 7 (3.59%) 6 (2.75%)
Pseudomonas aeruginosa 8 (1.90%) 7 (1.98%) 2 (0.90%) 5 (2.56%) 5 (2.29%)
Escherichia coli 21 (4.99%) 42 (11.90%) 24 (10.76%) 16 (8.21%) 16 (7.34%)
Other pathogens 137 (32.54%) 109 (30.88%) 87 (39.01%) 85 (43.59%) 99 (45.41%)
Fig. 2 The number of cases and the positive rates of Staphylococcus epidermidis (A), Staphylococcus hominis (B), main coagulase-negative staphylococci (C), and total strains of the genus staphylococci (D) isolated from the blood of children from 2018-2022.
Fig 2
In summary, our data indicated that the epidemiological characteristics of children with blood infections including total culture cases, the culture-positive cases, the positive rates and pathogen distribution were affected by COVID-19. The prevalence of CoNS infections especially S. epidermidis and S. hominis decreased gradually during the COVID-19 pandemic. As the COVID-19 pandemic eased, the infection of some pathogens may recover or increase. Hence, continuous surveillance with large samples is required to obtain the epidemiological features and provide evidence for the prevention and control of related infectious diseases.
Declaration of Competing Interest
The authors declare no conflict of interests.
Acknowledgments
This work was supported by grants from the National Natural Science Foundation of China (31900116) and the Medical Science and Technology Projects of Henan Province (LHGJ20190955).
==== Refs
References
1 Li Y Guo Y Duan Y. Changes in Streptococcus pneumoniae infection in children before and after the COVID-19 pandemic in Zhengzhou, China J Infect 85 3 2022 e80 ee1 35659542
2 Zhou J Zhao P Nie M Gao K Yang J Sun J. Changes of Haemophilus influenzae infection in children before and after the COVID-19 pandemic, Henan, China J Infect 2022
3 Li Liping Song Chunlan Li Peng Li Y. Changes of Escherichia coli infection in children before and after the COVID-19 pandemic in Zhengzhou, China Journal of Infection 2022
4 Duverger C Monteil C Souyri V Fournier S. Decrease of carbapenemase-producing Enterobacteriaceae incidence during the first year of the COVID-19 pandemic J Infect 85 1 2022 90 122
5 Michels R Last K Becker SL Papan C. Update on Coagulase-Negative Staphylococci-What the Clinician Should Know Microorganisms 9 4 2021
6 Becker K Both A Weißelberg S Heilmann C Rohde H. Emergence of coagulase-negative staphylococci Expert Rev Anti Infect Ther 18 4 2020 349 366 32056452
7 Ye Y Tian Y Kong Y Ma J Shi G. Trends of Antimicrobial Susceptibility in Clinically Significant Coagulase-Negative Staphylococci Isolated from Cerebrospinal Fluid Cultures in Neurosurgical Adults: a Nine-Year Analysis Microbiol Spectr 10 1 2022 e0146221
8 Meyer Sauteur PM Beeton ML Uldum SA Mycoplasma pneumoniae detections before and during the COVID-19 pandemic: results of a global survey, 2017 to 2021 Euro Surveill 27 19 2022
| 36521565 | PMC9742061 | NO-CC CODE | 2022-12-14 23:29:59 | no | J Infect. 2022 Dec 12; doi: 10.1016/j.jinf.2022.12.006 | utf-8 | J Infect | 2,022 | 10.1016/j.jinf.2022.12.006 | oa_other |
==== Front
Radiol Case Rep
Radiol Case Rep
Radiology Case Reports
1930-0433
The Authors. Published by Elsevier Inc. on behalf of University of Washington.
S1930-0433(22)01014-7
10.1016/j.radcr.2022.11.030
Case Report
The unpredictability of labile blood pressure: Afferent baroreflex failure in a critical patient with multiple thyroid surgeries and COVID-19 infection
Mahmoud Anas MD a
Kania Brooke DO a⁎
Geris Shady DO a
Akroush Wadah MD a
Manickam Rajapriya MD b
Azzam Moh'd Hazem K MD b
a Department of Medicine, St. Joseph's University Medical Center, 703 Main St, Paterson, NJ 07503, USA
b Department of Critical Care, St. Joseph's University Medical Center, 703 Main St, Paterson, NJ 07503, USA
⁎ Corresponding author.
12 12 2022
2 2023
12 12 2022
18 2 715718
29 10 2022
9 11 2022
11 11 2022
© 2022 The Authors. Published by Elsevier Inc. on behalf of University of Washington.
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 carotid sinus-arterial baroreflex is essential in maintaining blood pressure (BP) regulation. Afferent baroreflex failure (ABF) can present with labile changes in BP within seconds and can be secondary to neck surgery or radiation. We present here the first case, to our knowledge, of ABF precipitated by thyroidectomy, in a patient with active COVID-19 pneumonia, causing difficult control of severely labile BP in a critical care unit. Contributing factors included her critical illness state with upregulation of IL-6 leading to pituitary-adrenal axis alteration, her thyroidectomy further exacerbating autonomic dysfunction, as well as downregulation of ACE2 via COVID-19 infection. Management was achieved with a combination of midodrine and clonidine catered to specific BP thresholds. Additional research with a multidisciplinary approach is warranted to fully optimize the treatment of ABF in patients with neck surgery and or inflammatory conditions such as COVID-19.
Keywords
Afferent baroreflex failure
COVID-19 pneumonia
Head and neck surgery
Thyroidectomy
Labile blood pressure
==== Body
pmcIntroduction
Autonomic dysfunction is often multifactorial and can be difficult to manage given the hurdles in pinpointing the underlying cause. In a normal state, blood pressure (BP) is regulated via the arterial baroreflex (ABF), where sympathetic activity is downregulated and parasympathetic activity is upregulated when BP is elevated, and vice versa [1]. During critical states, IL-6, IL-8, and ACTH may be upregulated, leading to further ABF dysfunction [2,3]. Although COVID-19 has been a relatively new viral state with further research warranted, utilizing data from coronaviruses, we can observe how the coronavirus downregulates the ACE2 enzyme, which may explain why patients with COVID-19 have demonstrated autonomic dysregulation [4]. In addition, head and neck surgeries or radiation to this area can lead to carotid sinus nerve damage [5]. Here we present a 74-year-old female who underwent recent thyroid surgery and presented with acute hypoxic respiratory failure in the setting of COVID-19, who developed ABF likely contributed by neck surgery, critical illness, and COVID-19 infection.
Case presentation
A 74-year-old female with a history of chronic obstructive pulmonary disease (COPD) and a thyroid mass status post an open left hemithyroidectomy & isthmusectomy, partial right thyroidectomy with drain placement presented to the Emergency Department (ED) with dyspnea and hypoxia. The patient was found to have COVID-19 pneumonia (PNA) and superimposed bacterial PNA. COVID-19 variant and vaccination status was unknown. She was immediately intubated and admitted to the intensive care unit (ICU).
Due to improved alertness and breathing, an extubation trial was done on day 2 but was unsuccessful due to a neck mass compressing the trachea, and during extubation, the patient began to develop stridor, desaturate, and was reintubated. Computed tomography (CT) head and neck showed a markedly enlarged thyroid with left tracheal deviation and the patient underwent complete thyroidectomy the following day (Fig. 1 ). Pathology of the left thyroid mass was consistent with Hashimoto's Thyroiditis, with prominent sclerosis and nodular hurthle cell metaplasia/hyperplasia. Pathology of the right thyroid was consistent with scattered lymphocytic infiltrates with focal reactive germinal centers, and prominent Hurthle cell changes with nodular Hurthle cell hyperplasia, which was associated with stromal sclerosis. On the 4th day following surgery, the patient desaturated on pressure regulated volume control (PRVC), and a chest X-ray (CXR) demonstrated a new consolidation, and the PNA panel was positive for Klebsiella pneumoniae (Fig. 2 ).Fig. 1 CT Head and Neck demonstrating a markedly enlarged thyroid with left tracheal deviation.
Fig 1
Fig. 2 Chest X-Ray demonstrating a new consolidation.
Fig 2
The patient's blood pressure (BP) began to fluctuate from 80’s/40’s-260’s/190’s mmHg. Titrating pressors were not effective in controlling her volatile BP. Clonidine was started to control hypertensive urgencies, but severe subsequent hypotensive episodes made it difficult to continue. A trial of fentanyl drip did not add a benefit either. Adequate BP control was finally achieved by administering clonidine only when systolic BP (SBP) reached above 180 mmHg and midodrine when SBP reached below 80 mmHg. The patient's blood pressure fluctuated for 10 days total. The patient's blood pressure was stabilized on the 10th day of admission and subsequently the patient was transferred to the medical floors and discharged 10 days thereafter. The patient has been following up in Pulmonology and Endocrinology clinics outpatient.
Discussion
The carotid arterial baroreflex plays a significant role in the regulation of blood pressure, where carotid sinus stretch receptors have the ability to sense alterations in blood pressure, and afferent nerve signals are distributed to the brain stem leading to adjustments to both parasympathetic and sympathetic mechanisms [1]. When the blood pressure is increased, the arterial baroreflex leads to an increase in parasympathetic activity and a decrease in sympathetic activity [1]. Alternatively, when the blood pressure is decreased, the baroreflex will upregulate sympathetic activity and downregulate parasympathetic activity [1]. The arterial baroreflex mechanism becomes crucial in critically ill patients, in order to respond and maintain vital stability. Unfortunately, with our patient's critical presentation, normal baroreflex regulation was not achieved.
When determining the pathophysiology of ABF, the key is to obtain a thorough history, as this can lead to hints regarding the etiology [6]. Critically ill patients with poor prognoses have demonstrated higher ACTH levels with a longer cortisol release, with elevated IL-8 and IL-6 concentrations, concluding potential destructive pituitary-adrenal axis response in the setting of inflammation [7]. IL-6 in particular can manifest following hypoxic conditions, and in diabetic patients, IL-6 upregulation has been associated with an increased risk of cardiovascular disease and autonomic dysfunction [2,3]. In our patient's case, she presented with acute hypoxic respiratory failure warranting intubation and therefore was proposed to have increased ACTH, IL-8, and IL-6 causing subsequent pituitary-adrenal axis derangement.
ABF can manifest secondary to carotid sinus nerve damage following neck surgery or radiation [5]. In patients who develop ABF in the setting of recent neck surgery or neck irradiation, acute manifestations include the development of hypertension, with subacute manifestations leading to the development of orthostatic tachycardia and labile blood pressure [5]. Typically, the patients who undergo neck surgery tend to develop ABF much more rapidly in comparison to those who underwent neck irradiation [6]. In our patient's case, she developed new-onset ABF shortly following undergoing neck surgery, making her surgical intervention a likely cause for her labile blood pressure.
Patients with COVID-19 infection have demonstrated autonomic dysfunction, with research continuing to investigate the etiology of this relationship. Given prior studies investigating the SARS epidemic in the early 2000s, coronaviruses have been shown to bind to and downregulate the type-2 angiotensin-converting enzyme (ACE2), which helps both regulate angiotensin II and upregulate the inflammatory processes [8]. With the downregulation of ACE2, the renin-angiotensin-aldosterone-system (RAAS) is upregulated, further exacerbating lung edema [4]. In general, infectious processes can precipitate autonomic dysfunction [9]. With COVID-19, autoimmunity, hypovolemia, and involvement of the brainstem can offer explanations for the development of this process [9]. In our patient, on top of her thyroidectomy, her COVID-19 infection may have further contributed to her autonomic dysfunction. Although, it is difficult to discern which was the predominant contributing factor as further research is warranted on the relationship between COVID-19 and autonomic dysfunction.
The diagnosis of ABF remains ill-defined; with limited research available to guide definitive management. Ultimately, first-line treatment of ABF should be focused on non-pharmacological management such as preventative measures to avoid exacerbation of symptoms, elevating the head of the bed, ensuring adequate fluid and electrolyte balance, and physical therapy to prevent deconditioning [9]. When patients are unresponsive to non-pharmacological management, medications can be considered with a regimen catered towards the hemodynamics of the patient [9]. For instance, patients presenting with hypotension may require volume resuscitation or vasopressors, whereas patients presenting with more predominant tachycardia may benefit from beta-blockers [9]. Recent literature has demonstrated benefits of utilizing intravenous metoprolol when COVID-19 patients develop acute respiratory distress syndrome (ARDS) where patients required the ventilator for less time and had a reduction of inflammation of their alveoli [9]. In certain cases of POTS and AD in COVID-19, there has been an improvement in symptoms with the use of B-blockers, fludrocortisone, desmopressin, normal saline, octreotide, methylphenidate, droxidopa, midodrine, methyldopa, and clonidine [9,10]. Ultimately, in our patient's case, adequate BP control was achieved after trial and error, where specific treatment (clonidine and midodrine) was administered depending on blood pressure parameters to maintain a steady state.
Conclusion
As with our patient's case, autonomic dysfunction can be multifactorial in origin and requires a thorough history to determine the underlying cause. Our patient presented following head and neck surgery, was critically ill, requiring intubation, and was found to be COVID-19 positive. These factors above likely all played a part in her autonomic dysfunction, which required multiple trials of medications to achieve BP stability. Treatment must be catered to each patient's case and specific holding parameters and treatment thresholds may be necessary when developing a therapeutic plan for these patients. In conclusion, additional research on optimal management of ABF in COVID-19 patients is warranted, especially those with recent head and neck surgeries or critically ill states.
Patient consent
Informed consent for publication of their case was obtained from the patient.
Appendix Supplementary materials
Image, application 1
Acknowledgments
The authors would like to thank the patient and their family for allowing us to share this case with our colleagues. We would also like to thank the Critical Care team for their assistance in diagnosing and managing the patient's disease.
This case report was previously presented at CHEST 2022 as a rapid oral presentation.
Competing Interests: The authors report no conflict of interest. Ethical review is not necessary, because this is a case report. 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.radcr.2022.11.030.
==== Refs
References
1 Biaggioni I Shibao CA Jordan J. Evaluation and diagnosis of afferent baroreflex failure Hypertension 79 1 2022 57 59 10.1161/HYPERTENSIONAHA.121.18372 34878899
2 Leyfman Y Erick TK Reddy SS Galwankar S Nanayakkara PWB Di Somma S Potential immunotherapeutic targets for hypoxia due to COVI-Flu Shock 54 4 2020 438 450 10.1097/SHK.0000000000001627 32649367
3 Shinohara T Takahashi N Yufu K Anan F Kakuma T Hara M Role of interleukin-6 levels in cardiovascular autonomic dysfunction in type 2 diabetic patients Eur J Nucl Med Mol Imaging 35 9 2008 1616 1623 10.1007/s00259-008-0809-y 18449539
4 Kuba K Imai Y Ohto-Nakanishi T Penninger JM. Trilogy of ACE2: a peptidase in the renin-angiotensin system, a SARS receptor, and a partner for amino acid transporters Pharmacol Ther 128 1 2010 119 128 10.1016/j.pharmthera.2010.06.003 20599443
5 Sharabi Y Dendi R Holmes C Goldstein DS. Baroreflex failure as a late sequela of neck irradiation Hypertension 42 1 2003 110 116 10.1161/01.HYP.0000077441.45309.08 12782644
6 Heusser K Tank J Luft FC Jordan J. Baroreflex failure Hypertension 45 5 2005 834 839 10.1161/01.HYP.0000160355.93303.72 15837841
7 Goldstein DS. The extended autonomic system, dyshomeostasis, and COVID-19 Clin Auton Res 30 4 2020 299 315 10.1007/s10286-020-00714-0 32700055
8 Del Rio R Marcus NJ Inestrosa NC. Potential role of autonomic dysfunction in Covid-19 morbidity and mortality Front Physiol 11 2020 561749 10.3389/fphys.2020.561749
9 Bisaccia G Ricci F Recce V Serio A Iannetti G Chahal AA Post-acute sequelae of COVID-19 and cardiovascular autonomic dysfunction: what do we know? J Cardiovasc Dev Dis 8 11 2021 156 10.3390/jcdd8110156 34821709
10 Dani M Dirksen A Taraborrelli P Torocastro M Panagopoulos D Sutton R Autonomic dysfunction in 'long COVID': rationale, physiology and management strategies Clin Med (Lond) 21 1 2021 e63 e67 10.7861/clinmed.2020-0896 33243837
| 0 | PMC9742062 | NO-CC CODE | 2022-12-14 23:29:59 | no | Radiol Case Rep. 2023 Feb 12; 18(2):715-718 | utf-8 | Radiol Case Rep | 2,022 | 10.1016/j.radcr.2022.11.030 | oa_other |
==== Front
SSM Popul Health
SSM Popul Health
SSM - Population Health
2352-8273
The Authors. Published by Elsevier Ltd.
S2352-8273(22)00294-4
10.1016/j.ssmph.2022.101315
101315
Article
Trajectories of psychological distress over multiple COVID-19 lockdowns in Australia
Botha Ferdi a
Morris Richard W. b
Butterworth Peter c
Glozier Nick d∗
a Melbourne Institute: Applied Economic & Social Research, The University of Melbourne, & ARC Centre of Excellence for Children and Families Over the Life Course, Australia
b Central Clinical School, Faculty of Medicine and Health, University of Sydney, & School of Psychology, Faculty of Science, University of Sydney, & ARC Centre of Excellence for Children and Families Over the Life Course, Australia
c Melbourne Institute: Applied Economic & Social Research, The University of Melbourne, & National Centre for Epidemiology and Population Health, The Australian National University, Australia
d Central Clinical School, Faculty of Medicine and Health, University of Sydney, & ARC Centre of Excellence for Children and Families Over the Life Course, Australia
∗ Corresponding author.
12 12 2022
3 2023
12 12 2022
21 101315101315
18 9 2022
7 12 2022
8 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.
The impact of the global COVID-19 pandemic, including the indirect effect of policy responses, on psychological distress has been the subject of much research. However, there has been little consideration of how the prevalence of psychological distress changed with the duration and repetition of lockdowns, or the rate of resolution of psychological distress once lockdowns ended. This study describes the trajectories of psychological distress over multiple lockdowns during the first two years of the pandemic across five Australian states for the period May 2020 to December 2021 and examines whether psychological distress trajectories varied as a function of time spent in lockdown, or time since lockdown ended.
A total of N = 574,306 Australian adults completed Facebook surveys over 611 days (on average 940 participants per day). Trajectories of psychological distress (depression and anxiety) were regressed on lockdown duration and time since lockdown ended. Random effects reflecting the duration of each lockdown were included to account for varying effects on psychological distress associated with lockdown length.
The prevalence of psychological distress was higher during periods of lockdown, more so for longer lockdowns relative to shorter lockdowns. Psychological distress increased rapidly over the first ten weeks of lockdowns spanning at least twelve weeks, though less rapidly for short lockdowns of three weeks or less. Psychological distress levels tended to stabilise, or even decrease, after ten consecutive weeks of lockdown. After lockdown restrictions were lifted, psychological distress rapidly subsided but did not return to pre-lockdown levels within four weeks, although continued to decline afterwards.
In Australia short lockdowns of pre-announced durations were associated with slower rises in psychological distress. Lockdowns may have left some temporary residual population effect, but we cannot discern whether this reflects longer term trends in increasing psychological distress. However, the findings do re-emphasise the resilience of individuals to major life stressors.
Keywords
Depression
Anxiety
Mental health
Trajectories
COVID-19
Australia
==== Body
pmc1 Introduction
Both the COVID-19 pandemic itself and the government policy responses introduced to limit the spread of the virus (such as ‘lockdowns’) were associated with significant life stressors including illness and bereavement, isolation and loneliness, loss of employment, and economic uncertainty (Hertz-Palmor et al., 2021; Wu, Yao, Deng, Marsiglia, & Duan, 2021). Many studies examined population mental health during the initial phase of the pandemic, drawing upon data from representative cohorts that surveyed the same individuals prior to, and then at multiple time points during the pandemic. These studies showed that, after an initial deterioration in population mental health at the onset of the pandemic, there was evidence of recovery (Daly & Robinson, 2021; Pierce et al., 2020). A subsequent meta-analysis of longitudinal studies conducted in Europe and North America in the first year of the pandemic, 2020, confirmed that lockdown had a negative impact on population levels psychological distress, although the effect size was small (Robinson, Sutin, Daly, & Jones, 2022).
Research examining the effect of lockdowns on population mental health face several challenges. Worsening mental health over time may reflect the effect of COVID-19 lockdowns or may be the continuation of longer-term population trends (e.g., Butterworth, Watson, & Wooden, 2020; Twenge & Cooper, 2020). Studies drawing on longitudinal cohort studies with multiple pre-COVID measurement occasions have adjusted for underlying trends, and still found small but significant worsening of mental health during the pandemic (Pierce et al., 2020). Another challenge is to disentangle the mental health consequences of lockdowns from the direct effects of the pandemic (such as fear of catching the virus, Chandola, Kumari, Booker, & Benzeval, 2020). Studies using quasi-experimental designs (e.g., difference-in-difference) to contrast the mental health of people in areas that were and were not exposed to lockdowns (e.g., Butterworth, Schurer, Trinh, Vera-Toscano, & Wooden, 2022; Serrano-Alarcón, Kentikelenis, Mckee, & Stuckler, 2022) have demonstrated that lockdowns had a modest negative effect on overall mental health.
Although the existing research provides important insights into the average effects of lockdown restrictions on mental health, it usually only includes a limited number of observations during COVID-19. As a result of this limited temporal resolution, such studies provide no information about two key issues, namely (i) how mental health changes over the course of lockdown, and (ii) how quickly mental health recovers after lockdowns are lifted. Lockdowns may have cumulative effects on psychological distress; depression and anxiety may continue to rise with longer lockdowns, or there may be patterns of stabilisation or improvement that cannot be detected from a single COVID measurement occasion. Even surveys that have included multiple measures during the pandemic still may not include adequate measurement occasions to capture the variability in lockdown duration for individuals in different locations. As a result, estimates relying on such comparisons will fail to detect acute or transient changes in psychological distress and so may underestimate the total, or maximal, effect of lockdown.
Another body of research conducted during the COVID-19 pandemic has adopted a surveillance or monitoring approach (e.g., Botha, Butterworth, & Wilkins, 2022; Fancourt, Steptoe, & Bu, 2021). These studies have recruited representative cross-sectional samples at regular occasions throughout the pandemic (as regularly as weekly), using consistent measures over time. This approach provides an opportunity to examine how population mental health has changed over time, and in relation to changes in lockdown status. For example, the COVID-19 Behaviour Tracker Global Survey drew fortnightly representative surveys from existing online panels in 15 different countries for more than 12 months. Linking this data to details of the stringency of local policy responses (Oxford COVID-19 Government Response Tracker) showed how population mental health was associated with policy stringency: mental health was worse at the times policies were strictest (Aknin et al., 2022).
This impact of the pandemic and lockdowns is unsurprising as negative life events are well known to be associated with adverse impacts on mental wellbeing and increased psychological distress (Frijters, Johnston, & Shields, 2011; Jeong et al., 2016). However, the impact of significant life stressors is often transient, with mental health and subjective wellbeing recovering to baseline levels for most individuals (Kettlewell et al., 2020). Similar transient effects have also been found in pre-COVID research of the mental health effects of quarantine (Jeong et al., 2016). Understanding whether any lockdown mental health impact is transient and the trajectory of this may help determine policy and service responses to future pandemics.
Over the course of the pandemic, different Australian states had very different lockdown experiences. All states experienced multiple lockdowns, although the duration of lockdowns varied considerably: sometimes lasting for as little as a week or, as in the case of Melbourne in the state of Victoria (VIC) extending for up to 4 months. Victoria experienced the most severe and extended lockdown restrictions of any region in the world in 2020 of 112 consecutive days, while New South Wales (NSW) experienced a 106-day lockdown in the second half of 2021. Given Australia's aggressive suppression strategy, there were also many shorter lockdowns (introduced at the first sign of community transmission) with the duration of these lockdowns often pre-determined and announced by policy makers. Studies with high frequency data collection (e.g., daily or weekly) are required to evaluate how psychological distress is affected by different lockdown durations.
The aim of the present study is to investigate the mental health effects of COVID-19 policy responses based on daily data from five Australian states. We examined 1) trajectories of change, 2) recovery during and after lockdowns, and 3) the effect on mental health of lockdown duration and the number of lockdowns experienced. The main source of data used in this paper was the Global COVID-19 Trends and Impact Survey (UMD Global CTIS), which recruited a new random sample of Facebook users each day, stratified by country and region, and assessed COVID symptoms, depression, anxiety, and financial stress among other items (Astley et al., 2021; Kreuter et al., 2020). As such, it provides the type of high frequency dataset needed to estimate the changes in psychological distress that occurred within lockdown, as well as after the restrictions lifted, for each Australian state throughout the pandemic.
2 Material and methods
2.1 Sampling method
We use data from the UMD Global CTIS, which was a partnership between the University of Maryland and Facebook. Facebook users were invited to take off-platform surveys of COVID-19-related symptoms beginning April 23rd, 2020. Approved by the UMD IRB (1,587,016–10), the survey and sampling strategy was designed by the University of Maryland Joint Program in Survey Methodology. The survey participants provided their written informed consent to UMD to participate in the survey. The anonymised data was obtained for this paper under a data use agreement (DUA) between the University of Maryland and the University of Sydney. Full details of the methods of the stratified survey collection are described in Kreuter et al. (2020). Briefly, every day a unique random sample of Facebook users over 18 years old (stratified by region) was invited to consent and participate via an invitation at the top of their Facebook News Feed (i.e., a repeated cross-sectional survey design). Probability sampling on existing attributes (e.g., region, age, gender) based on internal Facebook data was used to ensure new users were included each wave and to reduce survey fatigue. In low density geographical regions, users may be sampled again once a month (in high density regions users are sampled every 2–6 months), but survey responses cannot be linked longitudinally.
Participants reported on their COVID-19 symptoms, psychological distress, and financial concerns (the complete list of survey variables is available at https://gisumd.github.io/COVID-19-API-Documentation/docs/indicators/indicators. Survey weights were developed from the United Nations Population Division 2019 World Population Projections for age and gender, and used to minimize errors of representation, including coverage, sampling and non-response bias in each geographic region. The resulting weighted estimates aim to represent the general population of adults in each state rather than Facebook users per se. More details on the sampling frame, non-response modelling to reduce nonresponse and coverage bias, and post-stratification to represent the general adult population are available at https://covidmap.umd.edu/document/css_methods_brief.pdf.
The average daily sample size for each State, stratified by age and gender is shown in Supplementary methods in the Appendix (Table A1). Tasmania, Northern Territory, and the Australian Capital Territory were not included in the analysis as the average number of users sampled in those regions was less than 20 percent of the average sample size in the other states, and the sampling dates did not extend over the same period.
2.2 Psychological distress
Psychological distress was measured by two items on depression and anxiety taken from the Kessler-10 (K10) (Kessler et al., 2003):“During the last 7 days, how often did you feel so depressed that nothing could cheer you up?” (5 = “All of the time”, 4 = “most of the time”, 3 = “some of the time”, 2 = “a little of the time”, 1 = “none of the time”).
“During the last 7 days, how often did you feel so nervous that nothing could calm you down?” (5 = “All of the time”, 4 = “most of the time”, 3 = “some of the time”, 2 = “a little of the time”, 1 = “none of the time”).
The K10 is a widely used instrument in Australia, in both epidemiology and clinical reporting. The administration of the K10 to monitor mental health outcomes is mandated for patients of public mental health services in the Australian State of NSW (Andrews & Slade, 2001; Hickie, Andrews, & Davenport, 2002), and evidence of internal consistency (α = 0.93), test-retest reliability (ICC = 0.86, r = 0.76), factorial validity, convergent and discriminant validity, and treatment sensitivity has been provided (Berle et al., 2010; Kessler et al., 2002; Merson, Newby, Shires, Millard, & Mahoney, 2021; Slade, Grove, & Burgess, 2011; Sunderland, Mahoney, & Andrews, 2012a, 2012b). The psychometric properties of the instrument are invariant across the adult lifespan (Sunderland, Hobbs, Anderson, & Andrews, 2012a, 2012b). We report the population weighted proportions of adults responding “most” or “all of the time” to each question as the population prevalence of depression and anxiety, respectively.
2.3 Lockdown dates
After an initial national lockdown from the end of March 2020 to mid-May 2020, Australia successfully reduced COVID-19 cases to negligible levels (as few as 3 new cases a day according to the 7-day trailing average, www.covidlive.com.au). Australian data in the UMD Global CTIS is only available from early May 2020, towards the end of the first national lockdown. Subsequently different states in Australia underwent distinct episodes of lockdowns of varying length over 2020 and 2021 (see Table 1 ). By the end of 2021, Melbourne (VIC) and Sydney (NSW) had experienced 272 and 150 days of lockdown respectively, while Queensland (QLD) (and the rest of Australia) had remained relatively free of restrictions. This makes NSW and VIC a good case-study to examine the impact of extended lockdowns on the prevalence of depression and anxiety in the population, relative to its temporal trend as well as by comparisons with the rest of Australia. The lockdown dates and durations for each State are summarised in Table 1.Table 1 Lockdown characteristics, by state.
Table 1State Lockdown Start End Duration
Victoria (VIC) 1 2020-03-31 2020-05-12 42 days
2 2020-07-09 2020-10-28 111 days
3 2021-02-12 2021-02-17 5 days
4 2021-05-27 2021-06-10 14 days
5 2021-07-15 2021-07-27 12 days
6 2021-08-05 2021-10-22 78 days
New South Wales (NSW) 1 2020-03-31 2020-05-15 45 days
2 2020-12-17 2021-01-09 23 days
3 2021-06-26 2021-10-10 106 days
Queensland (QLD) 1 2020-03-31 2020-05-02 32 days
2 2021-01-08 2021-01-11 3 days
3 2021-03-29 2021-04-01 3 days
4 2021-06-29 2021-07-03 4 days
5 2021-07-31 2021-08-08 8 days
South Australia (SA) 1 2020-03-31 2020-05-11 41 days
2 2020-11-19 2020-11-22 3 days
3 2021-07-21 2021-07-28 7 days
Western Australia (WA) 1 2020-03-23 2020-04-27 35 days
2 2021-01-31 2021-02-05 5 days
3 2021-04-24 2021-04-27 3 days
4 2021-06-29 2021-07-03 4 days
Note: Lockdown dates were sourced from State Premier announcements and news reports, and curated by Anthony Macali.
Note we excluded the initial national lockdown in the analyses below, as data collection only commenced towards the end of the first lockdown period. Furthermore, the second lockdown listed for NSW was restricted to a single local government area (LGA), representing fewer than 65,000 people (less than 0.8 percentage points of the NSW population), so was excluded from the analyses.
2.4 Modelling
The outcome variables were the daily prevalence of depression and anxiety. For duration of lockdown the main explanatory variable was the cumulative number of weeks spent to date in the current lockdown (“week”). To estimate the post-lockdown trajectory, we used the number of weeks in the post-lockdown period since the most recent lockdown ended, top-coded as a maximum of 5 weeks (“postweek”). To capture the non-linear trajectory of weekly changes in depression and anxiety with lockdown duration, the cumulative lockdown week (or post lockdown week) was modelled with cubic regression splines, using the mgcv package (version 1.8) by Wood et al. (Wood, 2004; Wood, Pya, & Säfken, 2016) running in R (version 4.1, R Core Team, 2013). We included varying coefficients (i.e., random effects) for the total duration of each lockdown to correctly account for variations in trajectory due to the total length of each lockdown (“duration”). We present the predicted population-level estimates of prevalence as a function of time in lockdown. The marginal effects of each lockdown duration were calculated to allow comparison between lockdown trajectories with different durations.
The linear effect of time since the start of the pandemic (“month”) was entered into each model to control for trends in levels of psychological distress over the pandemic (Butterworth et al., 2022). State fixed effects were included to capture average differences between regions in Australia (“State”) and the potential cumulative effect of new lockdowns (“number”) in each state.
Formally the daily prevalence of each outcome (pr) was modelled for each i=1...I days of the pandemic for each j=1...J State (NSW, VIC, QLD, SA, WA) as:prij=β0+β1(monthi)+β2[j](Statej)+β3[j](numberi×Statej)+f1(weeki)+f1[k](weeki,durationk)+εif1[k]∼f1b1[k](durationk)b1[k]∼N(0,σduration2)εi∼N(0,σε2)
Where prij is the population estimate of daily prevalence of psychological distress in each State, β1 is the underlying trend in levels of psychological distress in Australia, β2 is the fixed estimate for average differences in psychological distress in each State over the pandemic, β3 is the trend in psychological distress over different lockdown numbers in each State, f1 is a smooth function(s) for the non-linear trend in psychological distress over lockdown weeks, and b1 is the random effect (slope) of lockdown duration for each k=1...K durations.
f1 is a penalized cubic regression spline of the form:f(x)=β0+β1(x)+β2(x)2+β3(x)3+βp(x−τp)3
With equally spaced knots τ1<...<τP for p=1...P over the range of x.
The marginal effects of each lockdown duration and number were calculated by holding all other effects constant. We obtained the change in y (i.e., prevalence) at different values of x (lockdown duration or number) by using the delta method. In brief, y was estimated at given values of x (holding all other x′ constant). For each x we determined the change in y by Δ=(f(x+h)−f(x))/h, where h is an arbitrary value less than the range of x. The results express the average change in y at each x, marginalising over all other x′.
In sensitivity analyses we also included the number of daily new infections and daily financial concerns as confounding variables (see Appendix Section 2). Appropriate model diagnostic information is available in Appendix Section 3.
3 Results
The demographic features of Facebook users in our sample of 574,306 who responded to either the depression or anxiety item between April 2020 and December 2021 are shown in Table 2 .Table 2 Summary statistics, overall and by state.
Table 2Characteristic Australia
N = 5,743,061 VIC
N = 1,594,651 NSW
N = 1,446,051 QLD
N = 1,280,281 SA
N = 632,831 WA
N = 789,251
Gender
Female 348,842 (61%) 95,083 (60%) 86,968 (60%) 79,591 (62%) 39,602 (63%) 47,598 (60%)
Male 225,464 (39%) 64,382 (40%) 57,637 (40%) 48,437 (38%) 23,681 (37%) 31,327 (40%)
Age
18-24 60,800 (11%) 17,927 (11%) 15,878 (11%) 13,989 (11%) 5949 (9.4%) 7057 (8.9%)
25-44 217,775 (38%) 64,480 (40%) 57,974 (40%) 48,544 (38%) 19,394 (31%) 27,383 (35%)
45-64 203,992 (36%) 54,320 (34%) 49,133 (34%) 45,348 (35%) 25,260 (40%) 29,931 (38%)
65+ 91,739 (16%) 22,738 (14%) 21,620 (15%) 20,147 (16%) 12,680 (20%) 14,554 (18%)
Region
City 287,802 (51%) 77,990 (50%) 66,058 (47%) 63,066 (50%) 34,149 (55%) 46,539 (60%)
Town 182,543 (33%) 52,848 (34%) 50,782 (36%) 41,921 (34%) 16,624 (27%) 20,368 (26%)
Rural 89,470 (16%) 24,432 (16%) 24,094 (17%) 19,976 (16%) 10,889 (18%) 10,079 (13%)
Fear of infection 191,607 (43%) 60,749 (48%) 49,565 (46%) 39,668 (40%) 18,857 (38%) 22,768 (37%)
Financial concerns 144,856 (25%) 42,148 (26%) 37,656 (26%) 32,724 (26%) 14,867 (23%) 17,461 (22%)
Depressed 36,795 (6.4%) 12,019 (7.5%) 9494 (6.6%) 7669 (6.0%) 3744 (5.9%) 3869 (4.9%)
Anxious 24,057 (4.2%) 7683 (4.8%) 6101 (4.2%) 5268 (4.1%) 2508 (4.0%) 2497 (3.2%)
1n (%). Tasmania, Australian Capital Territory, Northern Territory were excluded due to small sample size.
Most responses were from females (61%) and/or adults aged 25–64 (74%), living in a city or town (84%), and over half came from VIC or NSW (53%). Compared to Australian 2021 census data (see Table A2), females are relatively oversampled, along with middle-aged adults (45–64). The survey weights provided by UMD Global CTIS were included in all models to adjust for disproportionate sampling over age groups and gender. Fear of infection was highest in VIC (48%) and lowest in WA (37%), X2 (4, N = 574,306) = 3774, p < .0001. Financial concerns were similar in VIC, NSW and QLD (26%) but lower in SA and WA, X2 (4, N = 574,306) = 685, p < .0001. The proportion of people feeling depressed was highest in VIC (7.5%), followed by NSW (6.6%), X2 (4, N = 574,306) = 706, p < .0001. Anxiety prevalence was also highest in VIC (4.8%) relative to other states, X2 (4, N = 574,306) = 374, p < .0001.
3.1 Depression
The daily prevalence of depression, as estimated by a weighted proportion of the population, is shown in Fig. 1 for each state over the course of the COVID-19 pandemic. A linear regression of depression prevalence on time (month) revealed the average monthly increase in prevalence over the pandemic was positive and significant in each state: β [95%CI] = 0.11 [0.1, 0.12], 0.1 [0.09, 0.11], 0.08 [0.07, 0.1], 0.16 [0.15, 0.17], 0.04 [0.03, 0.06] percentage points for NSW, QLD, SA, VIC, WA, respectively. Lockdown had a significant effect on increasing the prevalence of depression in each State except WA: β [95%CI] = 2.49 [2.27, 2.7], 0.46 [0.13, 0.8], 0.65 [0.01, 1.28], 2.36 [2.2, 2.53], 0.18 [−0.38, 0.74] percentage points for NSW, QLD, SA, VIC, WA, respectively.Fig. 1 Daily population prevalence of depression (±95% CI) by state and year.Note: Daily population weighted estimates of depression prevalence (±95%CI shaded) in each Australian State between the end of April 2020 and December 2021. Grey shaded regions indicate lockdown periods in each State.
Fig. 1
Fig. 2 reports the estimated trajectories for depression by lockdown duration. The estimated depression prevalence increased week on week over a lockdown before peaking in week 10 (adjusted R-squared = 76%, F = 55.46, p < .001). Including the random effect of lockdown duration in the lockdown model explained only an additional 0.2 percentage points of variance (adjusted R-squared = 76.2%, F = 29.32, p < .001).Fig. 2 Effect of lockdown on population prevalence of depression (±95% CI). Note: Figure shows the estimated effect of lockdown on the population prevalence of depression as a function of time (weeks). In the lockdown period (left), the Week 0 estimate represents the average prevalence in the period immediately prior to lockdown (left). In the post lockdown period (right), Week 0 represents the average depression during the prior lockdown and the Week 5 estimate represents the average prevalence after the 4th week post-lockdown period and before the next lockdown.
Fig. 2
The smooth trajectories in the post lockdown period shows the prevalence of depression lowered rapidly in the initial two weeks after lockdown on average, before approaching stable levels (adjusted R-squared = 72%, F = 22.34, p < .001). Including the random effect of lockdown duration in the post-lockdown model explained an additional 2 percentage points of variance (adjusted R-squared = 74%, F = 217.37, p < .001). The prevalence fell to 6 percent by 4 weeks post-lockdown regardless of lockdown duration. Note however the week 5 post lockdown estimate falls below the week 4 post lockdown estimate, indicating depression levels still had further to fall after four weeks.
The varying effect of lockdown duration plotted in Fig. 2 shows that short lockdowns (1–3 weeks) tended to have less impact on prevalence levels over the same initial period of a lockdown than longer lockdowns. The marginal effect of lockdown duration (over all states and lockdown numbers) is compared in Table 3 .Table 3 Marginal effect of lockdown duration on depression.
Table 3lockdown length (weeks) marginal delta (%) SE lower 95%CI upper 95%CI
1 0.36 0.05 0.27 0.46
2 0.38 0.05 0.29 0.47
3 0.40 0.04 0.31 0.48
12 0.55 0.04 0.48 0.63
16 0.62 0.04 0.54 0.71
17 0.64 0.04 0.56 0.73
Table 3 shows the marginal effect of lockdown duration on prevalence of depression is greater for longer lockdowns than shorter lockdowns. Depression prevalence increases by between 0.36 and 0.4 percentage points per week over short lockdowns, but a faster increase of 0.55–0.64 percentage points occurs for longer lockdowns. Comparison of the 95% confidence intervals shows the rate of increase in each case is significantly greater than the short lockdowns.
The number of new lockdowns (“number”) had different and even opposite cumulative effects on depression prevalence in some states.
The marginal estimates (Table 4 ) show the percent prevalence of depression increases by 0.45–0.465 percentage points with each additional lockdown in VIC and NSW, was also positive in QLD but in SA ranged around zero and was negative in WA. Comparison of the 95% confidence intervals shows the cumulative effect of each additional lockdown was significantly greater in VIC and NSW than any of the other states.Table 4 Marginal effect of lockdown number on depression.
Table 4State marginal delta (%) SE lower 95%CI upper 95%CI
Victoria 0.45 0.03 0.39 0.51
New South Wales 0.46 0.06 0.35 0.58
Queensland 0.08 0.03 0.01 0.15
South Australia −0.07 0.10 −0.27 0.13
Western Australia −0.22 0.05 −0.33 −0.12
3.2 Anxiety
Fig. 3 shows the daily prevalence of anxiety, as a weighted proportion of the population, for each state over the course of the COVID-19 pandemic. Changes in anxiety prevalence using this one question were much more modest during lockdown periods as compared to depression.Fig. 3 Daily population prevalence of anxiety (±95% CI) by state and year. Note: Daily population weighted estimates of anxiety prevalence (±95% CI shaded) in each Australian State between the end of April 2020 and December 2021. Grey shaded regions indicate lockdown periods in each State.
Fig. 3
A linear regression of anxiety prevalence on time (month) by lockdown revealed the average monthly increase in prevalence over the pandemic was positive and significant in each state: β [95%CI] = 0.11 [0.1, 0.12], 0.07 [0.06, 0.08], 0.08 [0.07, 0.09], 0.14 [0.13, 0.14], 0.05 [0.03, 0.06] percentage points for NSW, QLD, SA, VIC, WA, respectively. There were some differences between states in the effect of lockdown on anxiety with lockdown increasing anxiety in each state except QLD and WA β [95%CI] = 1.03 [0.88, 1.19], −0.38 [−0.65, −0.1], 0.89 [0.37, 1.42], 1.01 [0.89, 1.12], 0.11 [−0.4, 0.62] percentage points for NSW, QLD, SA, VIC, WA.
The estimated trajectories in Fig. 4 show similar patterns for the different lockdown durations, to the differences we observed for depression. The model estimates show anxiety prevalence increased rapidly week-on-week for the first 5 weeks, before falling after 10 weeks (adjusted R-squared = 65%, F = 33.53, p < .001). Adding the random effect of lockdown duration explained an additional 0.1 percentage points of variance (F = 10.01, p < .001).Fig. 4 Effect of lockdown on population prevalence of anxiety (±95%CI). Note: Figure shows the estimated effect of lockdown on the population prevalence of anxiety as a function of time (weeks). In the lockdown period (left), the Week 0 estimate represents the average prevalence in the period immediately prior to lockdown. In the post-lockdown period (right), Week 0 represents the average anxiety during the prior lockdown and the Week 5 estimate represents the average prevalence after the 4th week post-lockdown period and before the next lockdown.
Fig. 4
The estimates of the post lockdown period further show the prevalence of anxiety declining over the four-week period modelled here (adjusted R-squared = 61.6%, F = 6.20, p < .001), and continuing to decline after this four-week period as indicated by the post-lockdown Week 5 estimate. Adding the random effect of lockdown duration explained an additional 1.1 percentage points of variance (F = 85.71, p < .001).
The varying effect of lockdown duration plotted in Fig. 4 shows that short lockdowns (1–3 weeks) tended to have less impact on anxiety prevalence over the same period than longer lockdowns, but this effect was not significantly different in the estimated marginal effects of different lockdown durations in Table 5 .Table 5 Marginal effect of lockdown duration on anxiety.
Table 5lockdown length (weeks) marginal delta (%) SE lower 95%CI upper 95%CI
1 0.09 0.04 0.02 0.17
2 0.10 0.04 0.03 0.17
3 0.11 0.04 0.04 0.18
12 0.18 0.03 0.11 0.24
16 0.21 0.04 0.14 0.28
17 0.22 0.04 0.15 0.29
As with depression, the cumulative number of lockdowns had differential effects on anxiety prevalence (Table 6 ). Anxiety increased with each additional lockdown in VIC and NSW, whereas for the other states anxiety either decreased (QLD, WA) or there was no evidence of change (SA).Table 6 Marginal effect of lockdown number on anxiety.
Table 6State marginal delta (%) SE lower 95%CI upper 95%CI
Victoria 0.20 0.03 0.15 0.25
New South Wales 0.11 0.05 0.02 0.21
Queensland −0.12 0.03 −0.18 −0.07
South Australia −0.04 0.08 −0.21 0.12
Western Australia −0.22 0.04 −0.31 −0.13
4 Discussion
This study is, to our knowledge, the first to describe how psychological distress rose and fell across multiple lockdowns, and assess whether mental wellbeing recovered following lockdowns.
4.1 Key results
Using high temporal resolution daily survey data from Australian respondents, we found that the prevalence of psychological distress tended to increase over the course of the pandemic in almost all states. The consistent increase we demonstrated was not reported in a meta-analysis by Robinson et al. (2022), in which most studies which found that symptoms of psychological distress tended to decline over the pandemic in European and North American countries after an initial rise, as health and wellbeing improved after an initial adverse response. Our data did not capture the immediate pandemic period and so there may also have been an initial increase in psychological distress that we may have not measured. However, this general trend may not represent a COVID-19 effect as there were increasing baseline rates of psychological distress in the community reported in the years prior to the pandemic (see Butterworth et al., 2020). This shifting baseline contaminates estimates of the average lockdown effect in other studies relying on comparisons between two timepoints unless carefully controlled. However, by considering the long-term trend in daily psychological distress enabled us to distinguish the temporal effects of lockdown and its alleviation.
Our results suggest that psychological distress, primarily depression and to a lesser extent anxiety, increased over lockdown periods, with lockdowns of 12 weeks or more producing a more rapid increase than shorter lockdowns. Lockdowns of one, two or three weeks had little to no impact on psychological distress prevalence, potentially because their short and limited duration was often communicated to the public prior to their imposition We found that the effect of lockdowns on psychological distress was not permanent, with the levels of psychological distress prevalence declining to near, but still slightly elevated, pre-lockdown levels within four weeks following the end of lockdown, and continued to decline over the subsequent post-lockdown period.
The results from this study are therefore consistent with previous work demonstrating poorer mental health during lockdown. However, our findings suggest that this adverse mental health effect was likely only experienced in the case of lockdowns lasting more than three weeks, and most of the increase in psychological distress was transitory once lockdowns ended. However, a residual effect of lockdown may remain. We also found (see Appendix Section 2) that at least part of the association between lockdown duration and psychological distress was mediated by financial concerns.
4.2 Strengths
A key strength of this study is that the data came from five States in one country where there is a relatively homogeneous health and social care systems, and social and population structures who experienced similar Federal economic responses to the pandemic. The state variation in lockdown timing and duration enables some of these effects to be at least partially controlled, and the specific effects of lockdown be more evident. However, the degree of restriction within each lockdown varied although we could discern no pattern of association between this and psychological distress. The temporal resolution of these data is the only available that we are aware of that can address our trajectory questions. Although the samples are large it is very likely that some respondents responded multiple times and these people will contribute more to the findings.
4.3 Limitations
Despite considerable effort in survey design, including stratified probabilistic sampling over a large sampling frame and adjustments to population controls including age group, region and gender, sampling bias may still be present in this dataset. In particular, as an internet-based social media user group, the survey is a nonrandom sample of the population in the countries/territories covered by UMD-CTIS. As such, any stable, time-invariant differences between people may contribute to bias in the current results, such as unadjusted differences in education or rurality (Bradley et al., 2021; but see Astley et al., 2021). The effect of such bias is likely to be restricted to the estimation of the intercept in the models used here, and not interact with the trajectories or slopes over time. Of more concern would be the presence of time-varying differences. For example, working people might have elevated levels of psychological distress and only find time to respond during lockdown and not after work responsibilities have resumed. Such time-varying effects on the outcome could bias the trajectories of psychological distress. That said, the representativeness of longitudinal studies, particularly over a long period of time with frequent follow-up, would also be affected by selection and attrition.
4.4 Generalizability and interpretation
Overall, these high temporal resolution data from a very large sample, although limited to only two psychological distress questions, provide a guide to how psychological distress rises and falls in the population over the course of repeated lockdowns. These findings may be useful for public health communication, and, assuming that demand for mental health services follows the same pattern, for policy makers and clinicians. They also remind us how resilient people are in general to major life stressors (Kettlewell et al., 2020), an observation often missing from the social discourse. An interesting avenue for future research is to examine the potential value of collecting related data from social media outlets to understand the consequences of government policy in large populations. Another possibility is to examine different sub-populations, such as age groups and gender, to study whether lockdown duration impacted the psychological distress trajectories of these groups in different ways.
Ethical statement
The UMD Global CTIS was a partnership between the University of Maryland and Facebook. The survey and sampling strategy was designed by the University of Maryland Joint Program in Survey Methodology, and full details of the methods of the stratified survey collection are described in Kreuter et al. (2020). The UMD Global CTIS data collection was approved by the UMD IRB (1,587,016–10).
Author statement
Ferdi Botha: Conceptualization, Methodology, Writing – Original Draft, Writing – Review & Editing, Project administration, Funding acquisition. Peter Butterworth: Conceptualization, Methodology, Writing – Original Draft, Writing – Review & Editing. Richard W. Morris: Conceptualization, Methodology, Software, Formal analysis, Data Curation, Writing – Original Draft, Writing – Review & Editing, Visualization, Project administration, Funding acquisition. Nick Glozier: Conceptualization, Methodology, Writing – Original Draft, Funding acquisition.
Declaration of competing interest
None.
1.0 Supplementary methods
Daily sampling rate
The daily sampling rate of Facebook users who responded to either the depression or anxiety item between April 2020 and December 2021 is shown in Table A1 for each state, stratified by age group and gender:Table A1 Total respondents and average per day.
Table A1State Age Total respondents (%) Average respondents per day
Females Males Females Males
Victoria 18–24 11,665 (65%) 6262 (35%) 19 10
25–44 39,705 (62%) 24,775 (38%) 65 41
45–64 31,856 (59%) 22,464 (41%) 52 37
65+ 11,857 (52%) 10,881 (48%) 19 18
New South Wales 18–24 10,464 (66%) 5414 (34%) 17 9
25–44 35,830 (62%) 22,144 (38%) 59 36
45–64 29,227 (59%) 19,906 (41%) 48 33
65+ 11,447 (53%) 10,173 (47%) 19 17
Queensland 18–24 9721 (69%) 4268 (31%) 16 7
25–44 31,656 (65%) 16,888 (35%) 52 28
45–64 27,720 (61%) 17,628 (39%) 45 29
65+ 10,494 (52%) 9653 (48%) 17 16
South Australia 18–24 3978 (67%) 1971 (33%) 7 4
25–44 12,547 (65%) 6847 (35%) 21 11
45–64 16,214 (64%) 9046 (36%) 27 15
65+ 6863 (54%) 5817 (46%) 11 10
Western Australia 18–24 4594 (65%) 2463 (35%) 8 4
25–44 17,431 (64%) 9952 (36%) 29 16
45–64 18,063 (60%) 11,868 (40%) 30 19
65+ 7510 (52%) 7044 (48%) 12 12
Australian demographics (age and sex)
Table A2 shows age and sex distributions from the 2021 national Census for Australia and each state, for comparison with the demographic features of Facebook users in our sample (Table 2).Table A2 National Census data 2021
Table A2Characteristic Australia N = 25,422,788 VIC N = 6,503,491 NSW N = 8,072,163 QLD N = 5,156,138 SA N = 1,781,516 WA N = 2,660,026
Gender
Females 12,877,635 (51%) 3,302,528 (51%) 4,087,995 (51%) 2,615,736 (51%) 902,924 (51%) 1,337,171 (50%)
Males 12,545,154 (49%) 3,200,963 (49%) 3,984,166 (49%) 2,540,404 (49%) 878,592 (49%) 1,322,855 (50%)
Age
18-24 2,150,360 (11%) 554,388 (11%) 674,712 (11%) 444,934 (11%) 148,273 (10%) 217,581 (11%)
25-44 7,112,430 (36%) 1,894,231 (37%) 2,245,196 (36%) 1,389,541 (35%) 461,947 (33%) 751,844 (37%)
45-64 6,256,922 (31%) 1,573,432 (31%) 1,978,732 (31%) 1,289,431 (32%) 451,788 (32%) 661,700 (32%)
65+ 4,378,094 (22%) 1,092,833 (21%) 1,424,141 (23%) 875,603 (22%) 356,325 (25%) 428,619 (21%)
*n (%).
2.0 Supplementary results
In follow-up sensitivity analyses we included the number of daily new infections and daily financial concerns as explanatory variables.
Each variable was added to the model as a penalized cubic regression spline, and we calculated the marginal effects of each lockdown duration as before. The results, when compared to the marginal effects in the main report, indicate the mediating effect of each variable (financial concerns or new infections) on the effect of lockdown duration.
2.1 Financial concern
Financial concerns during lockdown have been shown to be a significant mediating factor of psychological distress for various disadvantaged groups (Botha et al., 2022). Financial concern was measured by a single item:“How worried are you about your household’s finances in the next month?” (Very, somewhat, not too worried, not at all)
We included the weighted percentage of people reporting they were “very” or “somewhat” worried about their household finances as representing financial concern.Table A3 Marginal effect of lockdown length on depression after including financial concerns
Table A3lockdown length (weeks) marginal delta (%) SE lower 95%CI upper 95%CI
1 0.08 0.05 −0.01 0.17
2 0.10 0.04 0.01 0.19
3 0.12 0.04 0.04 0.20
12 0.28 0.03 0.21 0.34
16 0.35 0.04 0.27 0.42
17 0.36 0.04 0.29 0.44
Adding financial concerns improved the fit with depression prevalence (explained deviance increased from 76.1% to 79.1%). In comparison to Table 3 (Marginal effect of lockdown duration on depression), the marginal effect of lockdown duration on depression prevalence was reduced by approximately 0.28 percentage points after including financial concerns. The remaining effect was indistinguishable from zero for short duration lockdowns (e.g., 1 week), and the effect at longer durations is almost half that observed without financial concerns (e.g., 0.36 v 0.64 during 17-week lockdown). Nevertheless, lockdowns of 2-week or longer tended to increase depression prevalence, with or without accounting for financial concerns.Table A4 Marginal effect of lockdown length on anxiety after including financial concerns
Table A4lockdown length (weeks) marginal delta (%) SE lower 95%CI upper 95%CI
1 −0.12 0.04 −0.20 −0.04
2 −0.11 0.04 −0.19 −0.04
3 −0.11 0.04 −0.18 −0.03
12 −0.04 0.04 −0.11 0.03
16 −0.02 0.04 −0.09 0.06
17 −0.01 0.04 −0.09 0.07
Adding financial concerns improved the fit with anxiety prevalence (explained deviance increased from 65.2% to 69.5%). Comparing the marginal effect of lockdown duration on anxiety prevalence with and without financial concerns in the model (e.g., Table 4. Marginal effect of lockdown duration on anxiety), reveals financial concerns substantially mediated anxiety over varying lockdown durations. After accounting for financial concerns, short lockdowns up to 3-weeks tended to reduce anxiety, while the effect of longer lockdowns was indistinguishable from zero.
2.2 Fear of infection
Other factors such as daily media reports of the rate of new infections, or announcement of temporary changes in government support could also drive changes in psychological distress on a daily basis. Even people who have not directly experienced pandemic-related stressors such as infection, bereavement or job loss can nevertheless be negatively affected by the fear of experiencing them, often fueled by exposure to a continuous deluge of negative media coverage of the spreading infection rates in the community (Bower et al., 2021; Digby, Winton-Brown, Finlayson, Dobson, & Bucknall, 2021; Garfin, Silver, & Holman, 2020). In Australia, the number of new infections was reported daily at official government press conferences, which were widely reported and tracked in the media. Thus like lockdown, their impact on psychological distress may be transient and shortlived – difficult to detect without daily measurements of both. However controlling for daily changes in the salient influence of infection rate when estimating trends in distress has not been widely done.
Although the UMD Global CTIS included a single item measuring fear of infection, the responses were only collected between May 1st, 2020 and May 20th, 2021, which excludes the extended lockdown period in NSW. Because of the restricted availability of this item, we adopted another measure as a proxy for fear of infection. Daily case numbers (new infections) reported by each State Government for the entire pandemic period were collected and curated by Anthony Macali at covidlive. com.au, and downloaded from www.covidlive.com.au/covid-live.csv on the 01-15-2022.
The correlation between (log) daily cases and responses to the fear of infection item over 2020 in VIC (i.e., the time period both were available over an extended lockdown period) was Pearson ρ= 0.931.Table A5 Marginal effect of lockdown length on depression after including new infections
Table A5lockdown length (weeks) marginal delta (%) SE lower 95%CI upper 95%CI
1 0.35 0.05 0.25 0.44
2 0.36 0.05 0.27 0.46
3 0.38 0.05 0.29 0.47
12 0.51 0.04 0.43 0.59
16 0.57 0.05 0.48 0.66
17 0.58 0.05 0.49 0.68
Adding new infections improved the fit with depression prevalence (explained deviance increased from 76.1% to 76.7%). The marginal effect of lockdown duration on depression prevalence was very similar with and without new infections in the model (compare to Table 3. Marginal effect of lockdown duration on depression).Table A6 Marginal effect of lockdown length on anxiety after including new infections
Table A6lockdown length (weeks) marginal delta (%) SE lower 95%CI upper 95%CI
1 0.09 0.04 0.02 0.17
2 0.10 0.04 0.03 0.17
3 0.11 0.03 0.05 0.18
12 0.21 0.03 0.16 0.26
16 0.25 0.03 0.19 0.31
17 0.26 0.03 0.20 0.32
Adding new infections improved the fit with anxiety prevalence (explained deviance increased from 65.2% to 66.1%). The marginal effect of lockdown duration on anxiety prevalence was very similar with and without new infections in the model (compare to Table 5. Marginal effect of lockdown duration on anxiety).
Overall, the results including mediating variables indicated almost all the effect of lockdown on anxiety was mediated by financial concerns, as well as a substantial portion of the effect of lockdown on depression. Indeed, once financial concerns were explained, short lockdowns tended to decrease anxiety prevalence and had little further impact on depression. By contrast, new infections (a proxy for fear of infection) had little mediating impact on depression or anxiety.
Model diagnostics
Model fits were checked and assessed for oversmoothing, as well as violation of the distributional assumptions.
The k-index represents the adequacy of the basis dimension for the fit (Wood, 2017; section 5.9). The further below 1, the more likely there is a missed pattern left in the residuals. The k-index for the random effect of lockdown duration were below 1 in the models of depression (k-index = 0.5) and anxiety (k-index = 0.52).
Four residual plots were also inspected for each fit, with plots of deviance residuals against approximate theoretical quantiles of the deviance residual distribution according to the fitted model. The Q-Q plot (top left) indicated some deviation in the tails of the distribution from normal, however there were no identifiable pattern in the residual vs predicted scatterplot, and the histogram of residuals was approximately normal.
Effect of lockdown duration on depression model.Image 1
Effect of lockdown duration on anxiety model.Image 2
Data availability
Data will be made available on request.
Acknowledgement
We are grateful to three anonymous referees for helpful comments and suggestions. This research was supported by the 10.13039/100015539 Australian Government through the Australian Research Council's Centre of Excellence for Children and Families over the Life Course (Project ID CE200100025).
==== Refs
References
Aknin L.B. Andretti B. Goldszmidt R. Helliwell J.F. Petherick A. De Neve J.-E. Policy stringency and mental health during the COVID-19 pandemic: A longitudinal analysis of data from 15 countries The Lancet Public Health 7 5 2022 e417 e426 10.1016/S2468-2667(22)00060-3 35461592
Andrews G. Slade T. Interpreting scores on the Kessler psychological distress scale (K10) Australian & New Zealand Journal of Public Health 25 6 2001 494 497 10.1111/j.1467-842X.2001.tb00310.x 11824981
Astley C.M. Tuli G. Mc Cord K.A. Cohn E.L. Rader B. Varrelman T.J. Global monitoring of the impact of the COVID-19 pandemic through online surveys sampled from the Facebook user base Proceedings of the National Academy of Sciences 118 51 2021 e2111455118 10.1073/pnas.2111455118
Berle D. Starcevic V. Milicevic D. Moses K. Hannan A. Sammut P. The factor structure of the Kessler-10 questionnaire in a treatment-seeking sample The Journal of Nervous and Mental Disease 198 9 2010 660 664 10.1097/NMD.0b013e3181ef1f16 20823728
Botha F. Butterworth P. Wilkins R. Protecting mental health during periods of financial stress: Evidence from the Australian Coronavirus Supplement income support payment Social Science & Medicine 115158 2022 10.1016/j.socscimed.2022.115158
Bower M. Smout S. Ellsmore S. Donohoe-Bales A. Sivaprakash P. Lim C. COVID-19 and Australia's mental health: An overview of academic literature, policy documents, lived experience accounts, media and community reports. The University of Sydney; Australia's Mental Health Think Tank Retrieved from https://ses.library.usyd.edu.au/bitstream/handle/2123/26777/AustraliasMentalHealthThinkTank-EvidenceSummary-COVID-MentalHealth.pdf 2021
Bradley V.C. Kuriwaki S. Isakov M. Sejdinovic D. Meng X.-L. Flaxman S. Unrepresentative big surveys significantly overestimated US vaccine uptake Nature 600 7890 2021 695 700 10.1038/s41586-021-04198-4 34880504
Butterworth P. Schurer S. Trinh T.-A. Vera-Toscano E. Wooden M. Effect of lockdown on mental health in Australia: Evidence from a natural experiment analysing a longitudinal probability sample survey The Lancet Public Health 2022 10.1016/S2468-2667(22)00082-2
Butterworth P. Watson N. Wooden M. Trends in the prevalence of psychological distress over time: Comparing results from longitudinal and repeated cross-sectional surveys Frontiers in Psychiatry 2020 1345 10.3389/fpsyt.2020.595696
Chandola T. Kumari M. Booker C.L. Benzeval M. The mental health impact of COVID-19 and lockdown-related stressors among adults in the UK Psychological Medicine 2020 1 10 10.1017/S0033291720005048
Daly M. Robinson E. Psychological distress and adaptation to the COVID-19 crisis in the United States Journal of Psychiatric Research 136 2021 603 609 10.1016/j.jpsychires.2020.10.035 33138985
Digby R. Winton-Brown T. Finlayson F. Dobson H. Bucknall T. Hospital staff well-being during the first wave of COVID-19: Staff perspectives International Journal of Mental Health Nursing 30 2 2021 440 450 10.1111/inm.12804 33098222
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 The Lancet Psychiatry 8 2 2021 141 149 10.1016/S2215-0366(20)30482-X 33308420
Frijters P. Johnston D.W. Shields M.A. Life satisfaction dynamics with quarterly life event data The Scandinavian Journal of Economics 113 1 2011 190 211 10.1111/j.1467-9442.2010.01638.x
Garfin D.R. Silver R.C. Holman E.A. The novel coronavirus (COVID-2019) outbreak: Amplification of public health consequences by media exposure Health Psychology 39 5 2020 355 10.1037/hea0000875 32202824
Hertz-Palmor N. Moore T.M. Gothelf D. DiDomenico G.E. Dekel I. Greenberg D.M. Association among income loss, financial strain and depressive symptoms during COVID-19: Evidence from two longitudinal studies Journal of Affective Disorders 291 2021 1 8 10.1016/j.jad.2021.04.054 34022550
Hickie I.B. Andrews G. Davenport T.A. Measuring outcomes in patients with depression or anxiety: An essential part of clinical practice Medical Journal of Australia 177 4 2002 205 207 10.5694/j.1326-5377.2003.tb05056.x 12175326
Jeong H. Yim H.W. Song Y.-J. Ki M. Min J.-A. Cho J. Mental health status of people isolated due to Middle East Respiratory Syndrome Epidemiology and Health 38 2016 10.4178/epih.e2016048
Kessler R.C. Andrews G. Colpe L.J. Hiripi E. Mroczek D.K. Normand S.-L. Short screening scales to monitor population prevalences and trends in non-specific psychological distress Psychological Medicine 32 6 2002 959 976 10.1017/S0033291702006074 12214795
Kessler R.C. Barker P.R. Colpe L.J. Epstein J.F. Gfroerer J.C. Hiripi E. Screening for serious mental illness in the general population Archives of General Psychiatry 60 2 2003 184 189 10.1001/archpsyc.60.2.184 12578436
Kettlewell N. Morris R.W. Ho N. Cobb-Clark D.A. Cripps S. Glozier N. The differential impact of major life events on cognitive and affective wellbeing SSM-Population Health 10 2020 100533 10.1016/j.ssmph.2019.100533
Kreuter F. Barkay N. Bilinski A. Bradford A. Chiu S. Eliat R. Partnering with a global platform to inform research and public policy making Survey Research Methods 14 2020 159 163 10.18148/srm/2020.v14i2.7761
Merson F. Newby J. Shires A. Millard M. Mahoney A. The temporal stability of the Kessler psychological distress scale Australian Psychologist 56 1 2021 38 45 10.1080/00050067.2021.1893603
Pierce M. Hope H. Ford T. Hatch S. Hotopf M. John A. Mental health before and during the COVID-19 pandemic: A longitudinal probability sample survey of the UK population The Lancet Psychiatry 7 10 2020 883 892 10.1016/S2215-0366(20)30308-4 32707037
R Core Team R: A language and environment for statistical computing 2013 R Foundation for Statistical Computing Vienna, Austria Retrieved from http://www.R-project.org/
Robinson E. Sutin A.R. Daly M. Jones A. A systematic review and meta-analysis of longitudinal cohort studies comparing mental health before versus during the COVID-19 pandemic in 2020 Journal of Affective Disorders 296 2022 567 576 10.1016/j.jad.2021.09.098 34600966
Serrano-Alarcón M. Kentikelenis A. Mckee M. Stuckler D. Impact of COVID-19 lockdowns on mental health: Evidence from a quasi-natural experiment in England and Scotland Health Economics 31 2 2022 284 296 10.1002/hec.4453 34773325
Slade T. Grove R. Burgess P. Kessler psychological distress scale: Normative data from the 2007 Australian national survey of mental health and wellbeing Australian and New Zealand Journal of Psychiatry 45 4 2011 308 316 10.3109/00048674.2010.543 21332432
Sunderland M. Hobbs M.J. Anderson T.M. Andrews G. Psychological distress across the lifespan: Examining age-related item bias in the Kessler 6 psychological distress scale International Psychogeriatrics 24 2 2012 231 242 10.1017/S1041610211001852 21933463
Sunderland M. Mahoney A. Andrews G. Investigating the factor structure of the Kessler psychological distress scale in community and clinical samples of the Australian population Journal of Psychopathology and Behavioral Assessment 34 2 2012 253 259 10.1007/s10862-012-9276-7
Twenge J.M. Cooper A.B. The expanding class divide in happiness in the United States, 1972–2016. Emotion 2020 10.1037/emo0000774
Wood S.N. Stable and efficient multiple smoothing parameter estimation for generalized additive models Journal of the American Statistical Association 99 467 2004 673 686 10.1198/016214504000000980
Wood S.N. Generalized additive models 2017 Chapman and Hall Boca Raton, FL
Wood S.N. Pya N. Säfken B. Smoothing parameter and model selection for general smooth models Journal of the American Statistical Association 111 516 2016 1548 1563 10.1080/01621459.2016.1180986
Wu S. Yao M. Deng C. Marsiglia F.F. Duan W. Social isolation and anxiety disorder during the COVID-19 pandemic and lockdown in China Journal of Affective Disorders 294 2021 10 16 10.1016/j.jad.2021.06.067 34256180
| 0 | PMC9742066 | NO-CC CODE | 2022-12-14 23:45:34 | no | SSM Popul Health. 2023 Mar 12; 21:101315 | utf-8 | SSM Popul Health | 2,022 | 10.1016/j.ssmph.2022.101315 | oa_other |
==== Front
Comput Ind Eng
Comput Ind Eng
Computers & Industrial Engineering
0360-8352
1879-0550
Published by Elsevier Ltd.
S0360-8352(22)00881-6
10.1016/j.cie.2022.108893
108893
Article
Elective surgery scheduling under uncertainty in demand for intensive care unit and inpatient beds during the epidemic outbreaks
Dai Zongli a
Perera Sandun b
Wang Jian-Jun a⁎
Kumar Mangla Sachin c
Li Guo def
a School of Economics and Management, Dalian University of Technology, Dalian 116024, China
b College of Business, The University of Nevada, Reno, NV 89557, USA
c Jindal Global Business School, O P Jindal Global University, Sonepat, Haryana, India
d School of Management and Economics, Beijing Institute of Technology, China
e Center for Energy and Environmental Policy Research, Beijing Institute of Technology, China
f Sustainable Development Research Institute for Economy and Society of Beijing, China
⁎ Corresponding author.
12 12 2022
12 12 2022
1088936 5 2022
28 11 2022
6 12 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.
Amid the epidemic outbreaks such as COVID-19, a large number of patients occupy inpatient and intensive care unit (ICU) beds, thereby making the availability of beds uncertain and scarce. Thus, elective surgery scheduling not only needs to deal with the uncertainty of the surgery duration and length of stay in the ward, but also the uncertainty in demand for ICU and inpatient beds. We model this surgery scheduling problem with uncertainty and propose an effective algorithm that minimizes the operating room overtime cost, bed shortage cost, and patient waiting cost. Our model is developed using fuzzy sets whereas the proposed algorithm is based on the differential evolution algorithm and heuristic rules. We set up experiments based on data and expert experience respectively. A comparison between the fuzzy model and the crisp (non-fuzzy) model proves the usefulness of the fuzzy model when the data is not sufficient or available. We further compare the proposed model and algorithm with several extant models and algorithms, and demonstrate the computational efficacy, robustness, and adaptability of the proposed framework.
Keywords
Healthcare operations
surgery scheduling
decision making under uncertainty
fuzzy theory
COVID-19
==== Body
pmc1 Introduction
Elective surgeries contribute to a substantial portion of hospital revenue and healthcare systems have faced an unprecedented financial crisis by delaying elective surgeries due to the COVID-19 pandemic (Best et al., 2020, Kliff, 2020). Uncertainty is one of the most critical factors leading to the delay of elective surgery, and the COVID-19 pandemic aggravates this uncertainty. For example, surgery duration has increased significantly and become more unpredictable in the COVID-19 era as surgical teams have to wear additional personal protective equipment before surgery and there is a postoperative disinfection procedure in operating rooms. In addition, the deterioration of the patient’s condition, caused by a delay in surgery, increases the uncertainty of surgery duration (SD), length of the stay in the ward (LOSW), length of the stay in the intensive care unit (LOSI), and intensive care unit demand (ICD) of elective patients(Dai, Wang, & Shi, 2022). In addition to demand, the supply of beds for elective patients became uncertain because they can be preempted by non-elective patients at any time. For example, COVID-19 patients occupy most intensive care unit (ICU) beds and nursing staff, and their daily demand is uncertain, resulting in the shortage and uncertainty of ICU bed capacity available to elective patients (ICC) 1.
There are many methods to deal with uncertainty, and stochastic optimization is considered to be one of the widely used methods. Stochastic optimization assumes that uncertain parameters follow a known distribution or fitted distribution of actual data (Bovim et al., 2020, Zhang et al., 2019). However, it is difficult to accurately know the distribution in practice (Shehadeh, 2022). In addition, based on our interviews 2, sometimes the hospital administrators do not have sufficient data. In particular, the historical data cannot reflect the increase in SD, LOSW, and LOSI caused by the COVID-19 pandemic, and there is no data about ICC and ICD. Fuzzy optimization can reduce the dependence on data by making use of expert knowledge. At present, it has been widely used in various fields, including project selection (Singh, Rathi, Antony, & Garza-Reyes, 2022) and supply chain management (Cao et al., 2021, Gabriel et al., 2021). Considering these scenarios with inadequate data, we propose a fuzzy scheduling method based on expert estimation. The surgeon team estimate SD, LOSW, LOSI, ICD, and ICC based on their experience and the patient’s condition.
As far as we know, there is no accurate method to solve a fuzzy model directly. Most studies first transform the fuzzy model into a non-fuzzy model and then use solvers or heuristic algorithms to solve the problem as shown in Figure 1 (Abdullah and Abdolrazzagh-Nezhad, 2014, Gonzalez-Rodriguez et al., 2008). However, the transformation process may lead to the loss of decision information. Therefore, we propose a transformation process to reduce the loss of decision information and improve the performance of the fuzzy model. On the other hand, for large-scale problems, accurate methods are often difficult to solve the model in an acceptable time because surgical scheduling is a complex combinatorial optimization problem. Therefore, we propose a hybrid heuristic algorithm to solve the transformation model for large-scale problems.Figure 1 Surgery scheduling problem and optimization framework.
Concisely, this paper considers the scheduling of elective surgeries with uncertainty. We capture the uncertainty in SD, LOSW, LOSI, ICC, and ICD using fuzzy numbers and sets. We develop a fuzzy model to deal with the uncertainty and insufficient data caused by the COVID-19 pandemic. To solve the fuzzy model, we first transform the fuzzy model into a tractable mixed-integer programming (MIP) model, and then propose a hybrid heuristic algorithm for the large-scale problems to reduce the solving time. The experiment shows that the fuzzy model based on expert experience can effectively deal with scheduling problems with insufficient data. In addition, the proposed model has excellent adaptability to the uncertainty caused by the COVID-19 pandemic. Finally, the experience also proves that the transformation model that is easy to solve can provide an accurate solution for small-scale problems. At the same time, the proposed hybrid heuristic algorithm can obtain a satisfactory solution in a reasonable time for large-scale problems.
In summary, our contributions are as follows. First, we use expert knowledge to deal with insufficient data on SD, LOSW, LOSI, ICC, and ICD by using fuzzy numbers and fuzzy sets. Second, we model a surgery scheduling problem with uncertainty incorporating the challenges brought by COVID-19. Third, we present an approach to solve the proposed fuzzy model. Specifically, we first transform the fuzzy model into a tractable mixed-integer programming (MIP) model and then propose a hybrid heuristic algorithm for the large-scale problems to reduce the solving time.
The remainder of this paper is organized as follows. Following the literature review in Section 2, we developed a fuzzy model in Section 3 and solve it in Section 4. Computational experiments are provided in Section 5. Section 6 concludes the paper with discussions and remarks.
2 Related Literature
We review the related literature on elective surgery scheduling in this section. Our thorough survey 3 clearly indicated that there are only a handful of studies that focus on surgery ‘scheduling’ problems within a framework that facilitates uncertainty and scarcity driven by epidemic outbreaks such as COVID-19. Nevertheless, many researchers focus on the impact of the COVID-19 epidemic on elective surgeries (cf. Best et al., 2020, Beninato et al., 2021; Nguyen et al., 2021; Norris et al., 2022) and demonstrate the impact of the COVID-19 epidemic on surgical scheduling, in general, and elective surgeries, in particular. The COVID-19 pandemic has impacted the scheduling of elective surgeries in the following ways: First, historical data is simply unavailable to produce reasonable forecasts for scheduling; for example, complicated preoperative preparation procedures, postoperative disinfection, and reduction in the number of staff have led to changes in the distribution of surgery duration, and/or the mean and variance of the surgery duration. Secondly, the availability of ICU and inpatient beds has become uncertain and scarce due to increased and highly volatile emergency COVID-19 demand. While the extant literature mainly identifies these characteristics, we focus on modeling and solving the resulting problems. The following subsections respectively discuss the surgery scheduling literature and uncertainty therein.
2.1 Surgery Scheduling
Surgical suites typically operate following either an open or block scheduling policy (Freeman et al., 2016, Miao and Wang, 2021). The open scheduling policy means that a surgeon can choose any working day to process a case. In addition, for the block scheduling policy, surgeon or surgeon groups are assigned to a period during which they can schedule their surgical cases. These time blocks are owned by surgeons and reserved in advance. Even if some time blocks are not used, they cannot be released during the planning period. Since patients in the same block must be scheduled for surgery on the same day, this may not be conducive to the flexible allocation and full utilization of ICU beds and inpatient beds, so we adopted an open scheduling policy.
Since the upstream stage consists of more expensive resources of hospitals, most studies focus on improving the utilization of upstream resources. For example, Li et al. (2016) propose a rescheduling method to improve the utilization of related resources in the OR. Batun et al. (2011) consider OR as a bottleneck resource and propose a stochastic mixed-integer programming model to minimize the total expected cost. Similarly, Roshanaei et al. (2017) consider the scarcity of OR and suggest a joint operation scheduling method based on multiple hospitals. These studies usually treat different hospital units in isolation. In contrast, our research treats surgery scheduling as a coordinated process between the upstream and downstream. Specifically, we consider the utilization efficiency of OR and the availability of ICU and ward.
Surgery scheduling under limited downstream resources has also been investigated. For example, Min and Yih (2010) propose a stochastic mixed-integer programming model for the shortage of downstream surgical intensive care unit (SICU) beds. Zhang et al. (2019) study a two-level optimization model considering the capacity limitation of the downstream SICU for the problem of elective surgery planning in a single department. These studies attempt to address the adverse effects of ICU capacity constraints on upstream OR utilization, whereas our study also highlights the importance of wards.
To compare the existing work on surgery scheduling and to highlight our contribution, we have summarized the relevant literature in Table 1 . Generally, the focus of research has been on the OR unit, while only several researchers have incorporated the ICU and ward and studied all three at the same time. Moreover, the objectives of surgical scheduling have varied across operating room overtime, patient waiting for costs, and extra ICU beds.Table 1 Related literature on surgery scheduling.
Paper Objective function Focused unit Planning horizon
Denton et al., (2010) OR overtime cost, OR open cost OR Intra-day
Lee et al. (2014) Completion time, waiting time OR, Post-anesthesia CU Intra-day
Min and Yih (2010) Patient costs, expected overtime costs OR, SICU Intra-day
Gul et al. (2015) Expected OR overtime, waiting and cancellation costs OR week
Jebali and Diabat (2015) Patient-related cost, expected OR Utilization cost, penalty cost for exceeding ICU capacity OR, ICU Intra-day
Freeman et al. (2016) The surgery revenue, costs for overtime and tardiness OR Intra-day
Neyshabouri and Berg (2017) Cost of patient priority and waiting time, overtime cost, cost of lack of SICU capacity OR, ICU week
Kumar et al. (2018) The LOSI of scheduled patients, the LOSI of canceled patients OR, ICU week
Eun et al. (2019) Patient health condition and total overtime OR Intra-day
Behmanesh et al. (2019) Makespan and the unscheduled surgical cases OR Intra-day
Zhang et al. (2019) Waiting cost, surgery cost, overuse of ORs, inadequate SICU beds and OR open cost OR, ICU week
Wang et al. (2019) Operational costs, including the fixed costs for Opening ORs and the expected penalty costs of overtime OR Intra-day
Bovim et al. (2020) Number of patients scheduled, cancellations, and resting in wards not designated OR, Ward week
Wang et al. (2020) The expected value of average recovery completion time for all patients OR Intra-day
This paper Waiting cost, OR overtime cost, extra ICU bed, extra inpatient bed OR, ICU, Ward week
Note: Different studies have used different terms for beds that exceed the capacity of the ICU. Commonly used terms include the exceeding ICU capacity, lack of SICU capacity, the extra beds acquired in the ward, and inadequate SICU beds. We use the term ‘extra’ ICU bed and ‘extra’ inpatient bed.
2.2 Uncertainty in Surgery Scheduling
The uncertainty in the scheduling of elective surgery mainly includes the SD, LOSI, LOSW, ICD, and ICC. The upstream stage mainly involves surgery duration, which can be subdivided into a pre-operative holding unit (PHU) duration, surgery duration, and post-anesthesia care unit (PACU) duration. Most of the scheduling studies focus on the uncertainty of surgery duration, as shown in Table 2 below. For example, Eun et al. (2019) use a stochastic mixed-integer program to optimize the assignment of surgeries. Furthermore, Neyshabouri and Berg, 2017, Schiele et al., 2021, and Zhang et al. (2019) consider the uncertainty of SD and LOSI. Only very few studies incorporate the uncertainty of LOSW (cf. Moosavi et al. 2020; Bovim et al., 2020, Schiele et al., 2021). As seen in Table 2 , none of the existing studies consider the uncertainty of ICU demand and capacity. This is because, under normal circumstances, hospitals have sufficient ICU beds reserved for elective surgeries, thereby the uncertainties therein do not play a major role in scheduling decisions. However, when a large number of ICU beds are occupied by emergency patients (due to unprecedented events such as the COVID-19 pandemic), the availability of ICU beds becomes limited and uncertain. Thus, the hospital can no longer reserve ICU beds for each elective surgery patient, and it has to carefully consider the need for ICU beds for elective patients. In addition, since ICU beds are shared by elective and emergency patients, the uncertainty in available ICU capacity becomes very critical in this case.Table 2 Related literature on surgery scheduling considering uncertain factors.
Literature SD LOSI LOSW ICD ICC Method
Denton et al. (2010) ✓ ✕ ✕ ✕ ✕ RO
Min and Yih (2010) ✓ ✓ ✕ ✕ ✕ SO
Lee et al. (2014) ✓ ✕ ✕ ✕ ✕ FO
Gul et al. (2015) ✓ ✕ ✕ ✕ ✕ SO
Jebali and Diabat (2015) ✓ ✓ ✕ ✕ ✕ SO
Freeman et al. (2016) ✓ ✕ ✕ ✕ ✕ SO
Neyshabouri and Berg (2017) ✓ ✓ ✕ ✕ ✕ RO
Kumar et al. (2018) ✕ ✓ ✕ ✕ ✕ SO
Eun et al. (2019) ✓ ✕ ✕ ✕ ✕ SO
Behmanesh et al. (2019) ✓ ✕ ✕ ✕ ✕ FO
Zhang et al. (2019) ✓ ✓ ✕ ✕ ✕ SO
Wang et al. (2019) ✓ ✕ ✕ ✕ ✕ RO
Bovim et al. (2020) ✓ ✕ ✓ ✕ ✕ SO
Zhang et al. (2020) ✓ ✓ ✕ ✕ ✕ SO
Wang et al. (2020) ✓ ✕ ✕ ✕ ✕ FO
This paper ✓ ✓ ✓ ✓ ✓ FO
Note: SO - Stochastic Optimization, RO - Robust Optimization. FO - Fuzzy Optimization
There exist other stochastic, robust, and fuzzy optimization models that deal with the uncertainty of surgery scheduling. For example, Kumar et al. (2018) propose a stochastic mixed-integer programming model to capture the uncertainty of LOSI when downstream capacity is limited. Min and Yih (2010) establish a stochastic compensation model and apply the sample average approximation algorithm to solve it. These authors apply stochastic optimization, and they need to assume that distributions of the parameters are known. In contrast, our model does not need to make such assumptions as we rely on expert opinion; thus, our model does not depend on historical data.
Denton et al. (2010) compare stochastic optimization (SO) and robust optimization (RO) models for the uncertainty of surgery duration. Their results show that RO performs better in situations where information about parameter distribution is limited. Neyshabouri and Berg (2017) employ a two-stage RO model to deal with the uncertainty in LOSI and LOSW. While we employ a fuzzy model to address the inadequacy of data in this paper, robust optimization has also been employed in the literature (Denton et al., 2010, Neyshabouri and Berg, 2017, Wang et al., 2019). However, for our surgery scheduling problem, since each surgery operation has a unique complexity and features, surgeons need to make a specific judgment and estimate for each patient based on their own experience, knowledge, and patient's physical condition(Chung et al., 2022, Moreno and Blanco, 2018) ; for example, the SD for cataract surgery was significantly influenced by anesthesia type, surgeon grade, high case complexity, pupil size, pupil expander use/type, CTR use, and intraoperative complications(Nderitu & Ursell, 2019). Thus, each estimate by the surgeons is not an exact value but a fuzzy interval with some uncertainty, which is difficult to describe by the uncertain sets in robust optimization. These models apply RO, which is suitable for situations with limited parameter distribution information. In comparison, we obtain the uncertain parameters from expert estimates that reflect the heterogeneity of patients, thereby reducing the complexity of the model. Thus far, there is limited literature on surgery scheduling with fuzzy theory. For example, Lee and Yih (2014) study a fuzzy model with the uncertainty of PACU duration. Behmanesh and Zandieh (2019) use a fuzzy optimization model based on multi-objective for the uncertainty of PHU time, SD, and PACU time. These studies focus on the intra-day surgery scheduling and highlight the uncertainty of the upstream (i.e., SD). In contrast, we consider the uncertainty in SD, LOSW, and ICD, and use a multi-day scheduling scheme.
3 Fuzzy Surgery Scheduling Model with Uncertain
In order to clearly represent the modeling process, we first developed a crisp surgery scheduling model by assuming that the surgical scheduling environment is certain. In subsection 3.2, we further developed a fuzzy model considering the uncertainty of the scheduling environment, i.e., fuzziness.
3.1 The Crisp Surgery Scheduling Model
In this subsection, we present a crisp (non-fuzzy) model (CM) for our elective surgery scheduling problem. The goal is to schedule patients optimally for the surgery when the capacity of the OR, ICU, and ward are all limited. Before each planning horizon (week), all patients stay on a waiting list and the hospital needs to optimally 4 select some patients from this list for treatment due to the capacity limit; the remaining patients will be considered during the next planning stage, i.e., in the subsequent week. Following the standard surgical practice, we assume that each patient is pre-assigned to a surgery team based on his/her primary surgeon and current needs, and this information is available at the time of the patient selection. Also, every patient has a latest surgery date before which his/her surgery must be completed. Therefore, patients are heterogeneous in terms of the surgery time requirement. Note that, whenever the capacity is limited, some patients may decide to leave the waiting list and look for another hospital, if they can’t get a timely appointment. After surgery, some patients will be discharged directly, some patients will enter the ICU, and the remaining patients will enter the ward to recover until they are discharged. Figure 2 below illustrates the flow of patients during the surgery procedure. It should be observed that ICU and ward can admit patients externally and thus, the ICU capacity is shared among the patients from the OR and the direct ICU inpatients whereas the ward capacity is shared among the patients from OR, ICU, and the direct inpatients. As we discussed earlier, during epidemic outbreaks such as COVID-19, the external demand for ICU and ward beds increases rapidly. Therefore, the analysis of the problem under stringent ICU and ward capacities would be of particular interest in this study.Figure 2 The flow of patients during the surgery procedure.
We now turn attention to the key determinants of our cost minimization objective function in the formulation. Unlike emergency surgeries, elective surgeries are less sensitive to the date of surgery. Nevertheless, due to health risks associated with waiting, elective surgeries can’t be postponed indefinitely. Therefore, our objective function incorporates the patients’ waiting cost for surgery. Waiting costs refer to the health risks and loss of satisfaction caused by waiting, which can be described as a loss of patient and social productivity due to treatment delays (Gerchak et al., 1996; Ayvaz-Cavdaroglu, 2010). However, choosing too many patients for surgery to reduce patient waiting times often results in OR overtime and overload of the ward which is costly. Thus, there exists an interesting tradeoff between the waiting cost and OR overtime and bed shortage costs. Consequently, our objective of this paper is to jointly reduce patient waiting costs, OR overtime costs, and both ICU and inpatient bed shortage costs by aptly scheduling elective surgeries.
To present the formulation of our problem, we first introduce the following notations in Table 3 .Table 3 Notation for indices, parameters, and decision variables.
Indices
i Patient index; i=1,2,3,...,N, where N indicates the number of elective surgeries on the waiting list.
s Surgeon index; s=1,2,3,...,S, where S indicates the number of surgeons.
j OR index; j=1,2,3,...,J, where J indicates the number of ORs.
d Surgery date index; d=1,2,3,...,D/D', where D indicates the number of days in the current planning horizon, and D' is a dummy day to accommodate excessive demand.
e Date index of discharge from ward; e=1,2,3,...,D,...,D+QW, where QW indicates the maximum LOSW of patients in Ward.
r Date index of patient leave ICU; r=1,2,3,...,D,...,D+QU, where QU indicates the maximum LOSI of patients in ICU.
Parameters
HA Index set of surgery date; HA=1,2,3,...,D'.
HE Index set of discharge date; HE=1,2,3,...,D,...,D+QW.
HU Index set of the date that a patient leaves ICU; HU=1,2,3,...,D,...,D+QU.
HI Index set of patients;HI=1,2,3,...,N.
BW Number of available inpatient beds at the beginning of the current planning horizon.
BU Number of available ICU beds at the beginning of the current planning horizon.
ujO Unit overtime cost per operating room ORj.
uW Unit cost per an optional extra inpatient bed.
uU Unit cost per extra ICU bed.
uiT Unit waiting cost of patient i.
MO Upper-bound for daily overtime hours of each OR.
MB Upper-bound on extra inpatient beds for each day.
MU Upper-bound on extra ICU beds for each day.
YisN×S Surgeon-patient matrix, Yis=1, if the surgeon s is the attending surgeon of patient i; otherwise Yis=0.
Zi Type of patient; Zi= 1 if a patient is an inpatient; otherwise Zi= 0.
βsdS×D Availability of surgical team; βsd= 1 if surgeon s is available on day d, and otherwise βsd= 0.
Tdj Open duration of operation room ORj on day d.
LiS Surgery duration (SD) of patient i.
LiW Length of stay in the ward (LOSW) of patient i.
LiU Length of stay in ICU (LOSI) of patient i.
DiU Type of patient; DiU=1 if a patient is admitted to ICU after surgery; otherwise DiU=0.
RdU The number of released ICU beds on day d.
RdW Number of released inpatient beds on day d.
Ksd Maximum working time of surgeons s on day d.
Duei The latest date of the surgery of patient i. To keep the patient healthy, each patient has a due date by which the operation must be completed. The due date reflects the heterogeneity in the relative urgency and severity of the patient's condition.
WiB Total waiting days of patient i before the beginning of the current planning horizon.
θ Penalty coefficient of waiting time for patients deferred to the next planning horizon.
Decision Variables
Xidjs Binary variable; Xidjs= 1, if patient i is assigned to ORj, surgeon s, on day d; otherwise Xidjs= 0.
Nie Binary variable; Nie= 1, if patient i is discharged from the ward on day e; otherwise Nie= 0.
NirU Binary variable; NirU = 1, if patient i is discharged from ICU on day r; otherwise NirU= 0.
ΔdjO Total overtime of the ORj on day d.
ΔdW The number of extra beds in the ward used on day d
ΔdU The number of extra ICU beds used on day d.
ΔiT Total waiting days of patient i.
Note: The superscripts of all parameters are only used to distinguish symbols and have no specific meaning.
Some model parameters, such as the patient and attending surgeon team match YisN×S, the surgeon's working date βsdS×D, the latest operation date Duei for each patient, and the surgical patient's demand for an inpatient bed Zi, are known. In addition, some parameters such as MO, MB, Ksd, WiB, and θ are usually set by decision-makers. Finally, we refer to a bed in the ward as the inpatient bed, and a bed in the ICU as the ICU bed; the two are collectively called ‘bed’.
With these notations, we can formulate our joint scheduling problem with cost minimizing objective as follows:(1) Min∑i=1NuiTΔiT+∑d=1D∑j=1JujOΔdjO+∑d=1DuWΔdW+∑d=1DuUΔdU
The first term in the objective function (1) represents the patients’ waiting cost whereas the second, third, and fourth terms respectively represent the overtime cost of the OR, and the costs of the extra inpatient and ICU beds. As we have defined in Table 3, corresponding u in each cost term of the objective function represents the unit cost associated with the respective decision variable.
Some model parameters, such as the patient and attending surgeon team match YisN×S, the surgeon's working date βsdS×D, the latest operation date Duei for each patient, and the surgical patient's demand for an inpatient bed Zi, are known. In addition, some parameters such as MO, MB, Ksd, WiB, and θ are usually set by decision-makers. Finally, we refer to a bed in the ward as the inpatient bed, and a bed in the ICU as the ICU bed; the two are collectively called ‘bed’.
With these notations, we can formulate our joint scheduling problem with cost minimizing objective as follows:(2) Min∑i=1NuiTΔiT+∑d=1D∑j=1JujOΔdjO+∑d=1DuWΔdW+∑d=1DuUΔdU
The first term in the objective function (1) represents the patients’ waiting cost whereas the second, third, and fourth terms respectively represent the overtime cost of the OR, and the costs of the extra inpatient and ICU beds. As we have defined in Table 3, corresponding u in each cost term of the objective function represents the unit cost associated with the respective decision variable.
Next, we introduce related sets of constraints and their interpretations.(3) ΔiT=∑d=1D∑j=1J∑s=1SHdAXidjs+WiB∑d=1D∑j=1J∑s=1SXidjs+θ·D·1-∑d=1D∑j=1J∑s=1SXidjs,∀i,
(4) ΔdjO⩾∑i=1N∑s=1SLiSXidjs-Tdj,∀j,∀d,
(5) ΔdjO⩾0,∀j,∀d,
(6) ΔdjO⩽MO,∀j,∀d.
Equation (2) represents the waiting time of each patient. In particular, this constraint includes the waiting time of the current planning horizon as well as the waiting time before the start of the planning horizon. Among the patients with the same waiting unit cost, the patient with a long waiting time gets the priority. Due to the limited capacity of the hospital, not all patients on the waiting list can be served in the current planning horizon, and thus, the surgeries of some patients will be postponed to the next planning horizon; the third term on the right-hand side of Equation (2) captures the postponement penalty cost associated with these postponed patients. Since the total overtime of ORj on day d should be non-negative, bounded, and larger than or equal to the excessive usage of the OR over the regular opening duration, we need constraints (3)-(5). Observe that, in our objective function (1), the second term on OR overtime can be represented as ∑d=1D∑j=1JujOmax∑i=1N∑s=1SLiSXidjs-Tdj,0 without the constraints (3)-(4). However, the introduction of an auxiliary variable ΔdjO gives us a linear model. The same remark applies for ΔdW and ΔdU in the model. Nevertheless, in the fuzzy model in Subsection 3.2, we will use the original form with max(x,0) for convenience.(7) ∑d=1D∑j=1J∑s=1SHdAXidjs+LiU∑d=1D∑j=1J∑s=1SXidjsDiU·Zi=∑r=1D+QUHeENirU,∀i,
(8) ∑r=1D+QUNirU⩽1,∀i,
(9) ∑d=1D∑j=1J∑s=1SHdAXidjs+∑d=1D∑j=1J∑s=1SXidjsLiW·1-DiU+DiUZi=∑e=1D+QWHeENie,∀i,
(10) ∑e=1D+QWNie⩽1,∀i.
The discharge day of an ICU patient can be calculated by adding LOSI to the operation day; Equations (6)-(7) serve this purpose. Equations (8)-(9) calculate the patient’s expected discharge date, which is the initial date of operation plus the number of days in the hospital 5. For patients who enter the ICU before the current planning horizon, if they leave the ICU during the current horizon, they will directly enter the ward. If a patient does not leave the hospital in the current planning horizon, then that patient will still be assigned to a bed in the next planning horizon until the discharge date. For patients scheduled in the current planning horizon, the day of leaving the ICU and the ward can be estimated based on their LOSI and LOSW. While variables Nie and NirU can be viewed as redundant variables, these variables not only help to understand the model, but also help the computations of the number of patients discharged each day.(11) ∑d′=1d∑i=1N∑j=1J∑s=1SZiXid′js-∑e=1d∑i=1NNie-BW-∑d′=1dRd′W+∑d′=1dRd′U+∑r=1d∑i=1NNirU⩽ΔdW,∀d,
(12) ΔdW⩾0,∀d,
(13) ΔdW⩽MB,∀d.
Constraints (10)-(12) handle the extra inpatient beds. The patients who enter the ward include patients leaving the operating room (see process P3 in Figure 2), as well as patients who enter the ICU and then leave the ICU in the current planning horizon (process P6 in Figure 2), and also include patients who enter the ICU before current planning horizon and leave the ICU in the current planning horizon. The patients leaving the ward include those who entered the ward during the current horizon and those before. Constraints (12) assure that the shortage of inpatient beds per day is less than an upper bound.(14) ∑i=1I∑j=1J∑d′=1d∑s=1SXid′jsDiU-BU-∑d′=1dRd′U-∑r=1d∑i=1NNirU⩽ΔdU,∀d,
(15) ΔdU⩾0,∀d,
(16) ΔdU⩽MU,∀d.
Equations (13)-(15) represent extra ICU beds, i.e., the number of beds beyond the capacity. The availability of ICU beds includes the initially available ICU beds and the released ICU beds. These released ICU beds refer to patients who entered the ICU before the current decision period and left the ICU during the current decision period. Constraints (15) assure that the shortage of ICU beds per day is less than an upper limit.(17) ∑j=1J∑s=1SXidjs⩽∑s=1Sβsd,∀i,
(18) ∑j=1J∑i=1NXidjsLiSβsd⩽Ksd,∀s,∀d,
(19) ∑d=1Duei∑j=1J∑s=1SXidjs=1,∀i,
(20) ∑d=1D′∑j=1J∑s=1SXidjs=1,∀i,
(21) ∑d=1D′∑j=1JXidjs=Yis,∀i,s.
In order to guarantee the availability of the assigned surgeon for each patient on the day of surgery, we have added the constraint (16). Constraints (17) respectively ensure that the daily working hours of the surgical team. Since patients must be admitted before their latest surgery date and each patient can only be assigned/discharged once, we employ constraints (18)-(19). Constraint (19) ensures that unplanned patients in the current planning horizon will be assigned to a dummy date, meaning that some patients are postponed to the next planning horizon due to capacity constraints. Constraint (20) indicates that each patient is assigned to one surgeon only.
Although the solutions to the above scheduling problem would provide the directions on the type of patients that deserves a priority in admission, which OR should be assigned to these patients, the optimal time for surgeries, etc., we present a more robust version (with uncertainty) of the above deterministic problem next.
3.2 Fuzzy Model
In subsection 3.1, we assume that the surgical scheduling environment is certain, but in fact, the surgical scheduling environment is fuzzy. Therefore, in this subsection, we express the parameters in the surgical scheduling model as fuzzy numbers and fuzzy sets. Specifically, in the crisp model, we assume that all parameters, such as SD, LOSW, LOSI, ICD, and ICC, are known. In practice, decision-makers cannot obtain this information in advance. Although hospitals have historical data on SD, LOSW, and ICD, each patient presents a unique case with different characteristics and ICD dynamics changes due to external factors. For example, the demand for ICU beds has surged and become very uncertain due to the COVID-19 pandemic. In this subsection, we propose an approach based on expert estimates to handle the uncertainties. We represent uncertain parameters as fuzzy sets and fuzzy numbers. One of the key advantages of the fuzzy model is that the parameters can be determined either from expert estimates or a small amount of data (Yao and Lin, 2002). Subsection 3.2 introduces the fuzzy representation of uncertain parameters and presents a comprehensive description of the fuzzy model. The literature on decision-making in the fuzzy environment is rich(Bastos et al., 2019, Bellman and Zadeh, 1970) .
Fuzzy models have been widely employed in hospital environments. In particular, triangular fuzzy numbers (TFNs) have been used for uncertain parameters in surgery scheduling studies (cf. Lee et al., 2014; Behmanesh et al., 2019; Wang et al., 2020). Similarly, we employ TFNs to capture uncertainty in our model (The definition of triangle fuzzy number can be seen in Appendix C.2). Specifically, we represent the patients’ SD by the TFN, L∼iS=(LiSl,LiSm,LiSr), where LiSl, LiSm, and LiSr respectively denote the most optimistic, plausible, and pessimistic values of surgery duration of patient i; see Figure 3 . Also, LOSI and LOSW can be represented by fuzzy numbers L∼iW=(LiWl,LiWm,LiWr) and L∼iU=(LiUl,LiUm,LiUr). It should be noted that we denote fuzzy numbers using ‘tilde’ throughout this paper.Fig 4. Figure 3 Patients’ SD represented as a triangular fuzzy number L∼iS.
Figure 4 Heuristics for OR allocation.
Since some patients have entered the ward (or ICU) before the current planning horizon and may leave the ward (or ICU) in the current planning horizon, it is necessary to estimate the ICU or inpatient beds that may be released in the current planning horizon. Specifically, the number of released ICU beds on day d can be expressed as a TFN, R∼dU, and the number of released inpatient beds on day d can be expressed as R∼dW. For patients whose discharge date exceeds the planned period, the hospital needs to evaluate the possibility of leaving the ward (or ICU) at the beginning of the next decision-making period. This approach ensures the adaptability of the model for highly uncertain environments with smaller decision-making periods.
Usually, hospitals cannot directly observe the patient’s type, DiU. However, hospitals can form a fuzzy set F∼ through expert estimation and then derive DiU using a transformation of F∼. Specifically, experts (or surgeons) can assess whether each patient needs an ICU bed after the surgery and establish a fuzzy set F∼ after evaluating the current condition and the entire medical history of the patients. Let P=p1,p2,⋯,pn be a set of elective patients on the waiting list. Then, the fuzzy set of patients entering the ICU after surgery is as follows:F∼=p1,μ1,p2,μ2,...,pk,μk,...,pn,μn
where μk is the membership degree of the patient pk belonging to F∼, and μk∈[0,1]. The membership degree indicates the degree to which each element in the set belongs to the set 6 and Bellman and Zadeh (1970) proposed this idea in their seminal paper; please see Appendix C.1 for a complete definition. In this paper, the fuzzy set F∼ reflects the attending surgeon's uncertainty about whether a patient needs ICU care after a surgery.
Nevertheless, hospital managers still need to convert fuzzy sets into definite sets according to their attitudes toward risks in order to create surgical schedules. Specifically, let A=Fλ denote the λ-cut set of F∼, where λ-cut set of a fuzzy number F∼ is defined as Fλ=x∈Ω|μF∼≥λ. Here, λ represents the risk coefficient of the decision-maker's assessment of ICU needs. Then, the decision-maker gets a definite set A of ICU requirements which determines DiU through the following relationships:∑i=1NHiIDiU=Av,∀v=1,2,...,|A|,
DiU⩽1,∀i.
Based on the crisp model and the operation rules of triangular fuzzy numbers 7, we can establish the fuzzy model (FM) below. Note that constraints (23), (28), and (29) jointly represent constraints (3)-(5), (10)-(12), and (13)-(15) in the original formulation, respectively.
Min ∑i=1NuiTΔiT+∑d=1D∑j=1JujOΔ∼djO+∑d=1DuWΔ∼dW+∑d=1DuUΔ∼dU (22)
Subject to:(23) ΔiT=∑d=1D∑j=1J∑s=1SHdAXidjs+WiB∑d=1D∑j=1J∑s=1SXidjs+θ·D·1-∑d=1D∑j=1J∑s=1SXidjs,∀i,
(24) Δ∼djO=max∑i=1N∑s=1SXidjsL∼iS-Tdj,0⩽MO,∀j,∀d,
(25) ∑d=1D∑j=1J∑s=1SXidjsHdA+L∼iU∑d=1D∑j=1J∑s=1SXidjsDiU·Zi=∑r=1D+QUHeENirU,∀i,
(26) ∑r=1D+QUNirU⩽1,∀i,
(27) ∑d=1D∑j=1J∑s=1SXidjsHdA+∑d=1D∑j=1J∑s=1SXidjsL∼iW·1-DiU+DiUZi=∑e=1D+QWHeENie,∀i,
(28) ∑e=1D+QWNie⩽1,∀i,
(29) Δ∼dW=max∑d′=1d∑i=1N∑j=1J∑s=1SZiXid′js-∑e=1d∑i=1NNie-B∼W-∑d′=1dR∼d′W+∑d′=1dR∼d′U+∑r=1d∑i=1NNirU,0⩽MB∀d,
(30) Δ∼dU=max∑i=1I∑j=1J∑d′=1d∑s=1SXid′jsDiU-∑r=1d∑i=1NNirU-B∼U-∑d′=1dR∼d′U,0⩽MU,∀d,
(31) ∑j=1J∑s=1SXidjs⩽∑s=1Sβsd,∀i,∀d=1,...,D,
(32) ∑j=1J∑i=1NXidjsL∼iSβsd⩽Ksd,∀s,∀d=1,...,D,
(33) ∑d=1Duei∑j=1J∑s=1SXidjs=1,∀i,
(34) ∑d=1D′∑j=1J∑s=1SXidjs=1,∀i,
(35) ∑d=1D′∑j=1JXidjs=Yis,∀i,s,
(36) ∑i=1NHiIDiU=Av,∀v=1,2,...,|A|,
(37) DiU⩽1,∀i=1,2,...,N.
4 Solution Approach
Given that the fuzzy model cannot be solved directly by a solver, we first transform the fuzzy model into a tractable mixed-integer programming (MIP) model, and then propose a hybrid heuristic algorithm for the large-scale problems to reduce the solving time. The solver can be applied to small-scale problems, such as small hospitals or single departments, because its solution process is stable, fast, and convenient. The hybrid heuristic algorithm can be applied to large-scale problems, such as large hospitals, because its solution can be achieved in an acceptable time. Specifically, since fuzzy models can’t be solved directly using commercially available solvers (e.g., AIMMS, CPLEX, Gurobi), we provide two approaches. In subsection 4.1, we transform the fuzzy model (FM) into an equivalent crisp model (FECM) and use a solver (CPLEX) to solve it. This transformation is required as the fuzzy models cannot be easily converted to an equivalent MIP model that can be solved using a commercial solver such as CPLEX; moreover, this approach is also useful when employing general meta-heuristics such as GA and DE. In our second approach in subsection 4.2, we provide a heuristic algorithm based on an evolutionary algorithm and employ the proposed algorithm to obtain a satisfactory solution in a short time. The advantage of the transformation approach is that the resulting FECM can be solved directly to get accurate results utilizing an existing solver. However, the solution time will increase exponentially with the solution scale as surgical scheduling is an NP-hard problem (Denton et al., 2010, Gul et al., 2015). In addition, the process of transforming the FM into a FECM increases the complexity of the model. Nevertheless, this approach works well when 1) the planning horizon is long, 2) the problem scale is not too large, and 3) the model solution time requirement is low and produces very accurate results. The heuristic approach overcomes most of the disadvantages of the first method. In particular, the heuristic approach suits scenarios where the problem is large in scale and time-sensitive. This approach can be very useful in highly uncertain environments as shorter planning horizons and faster solution times are essential in such environments in order to enhance the adaptability of the models.
4.1 Transformation of the Fuzzy Model
We employ a transformation method proposed by Jiménez et al.(2007) to transform FM into a FECM. In our approach, we first transform FM into a FECM and then further linearize it.
First, we define the transformation approach and provide some background information. Consider the following fuzzy model:Minz=c∼Tx
subjecttox∈N(A∼,b∼)=x∈Rna∼ix⩾b∼i,i=1,2,...,m,x⩾0
where c∼=(c∼1,c∼2,...,c∼n), A∼=a∼ijm×n, b∼=(b∼1,b∼1,...,b∼n) represent, respectively, fuzzy parameters involved in the objective function and constraints. Then, the equivalent crisp (non-fuzzy) model can be written as follows:MinEV(c∼)x
(38) subjectto[(1-α)E2ai+αE1ai]x⩾αE2bi+(1-α)E1bi,i=1,2,...,m,x⩾0,α∈[0,1]
where α represents the degree that, at least, all the constraints are satisfied; that is, α is the feasibility degree of a decision x; the expected value of a fuzzy number, EV(c∼), is the halfway point of its expected interval, EI(c∼) (Heilpern, 1992). Specifically, we have(39) EV(c∼i)=E1ci+E2ci2
(40) EI(c∼i)=E1ci,E2ci
where c∼i=cil,cim,cir,E1ci=12cil+cim and E2ci=12cim+cir.
When (37) is a less than or equal (≤) type constraint, the crisp constraint can be written as follows:[(1-α)E1ai+αE2ai]x⩽αE1bi+(1-α)E2bi,i=1,2,...,m,x⩾0,α∈[0,1]
Using the above specifications, the fuzzy model (FM) in Subsection 3.2 can be transformed into the FECM below; note that constraints in Equations (22), (25), (27), (30), and (32)-(36) are crisp (non-fuzzy) constraints, and therefore, we don’t have to transform them. Thus, we don’t repeat those constraints here.(41) Min∑i=1NuiTΔiT+∑d=1D∑j=1JujOEVΔ∼djO+∑d=1DuWEVΔ∼dW+∑d=1DuUEVΔ∼dU
subject to:(42) EVΔ∼djO=EVmax∑i=1N∑s=1SXidjsL∼iS-Tdj,0,∀j,∀d,
(43) ∑i=1N∑s=1SXidjs1-αLiSl+LiSm2+αLiSm+LiSr2-Tdj⩽MO,∀j,∀d,
(44) ∑d=1D∑j=1J∑s=1SXidjsHdA+12LiUl+LiUm2+12LiUm+LiUr2∑d=1D∑j=1JXidjDiU·Zi=∑r=1D+QUHeENirU,∀i,
(45) ∑d=1D∑j=1J∑s=1SXidjsZiHdA+∑d=1D∑j=1J∑s=1SXidjsZi12LiWl+LiWm2+12LiWm+LiWr2·1-DiU+DiU=∑e=1D+QWHeENie,∀i,
(46) EVΔ∼dW=EVmax∑d′=1d∑i=1N∑j=1J∑s=1SZiXid′js-∑e=1d∑i=1NNie-B∼W-∑d′=1dR∼d′W+∑d′=1dR∼d′U+∑r=1d∑i=1NNirU,0,∀d,
(47) ∑d′=1d∑i=1N∑j=1J∑s=1SZiXid′js-∑e=1d∑i=1NNie-αBWl+BWm2+1-αBWm+BWr2-∑d′=1dαRd′Wl+Rd′Wm2+1-αRd′Wm+Rd′Wr2+∑d′=1d1-αRd′Ul+Rd′Um2+αRd′Um+Rd′Ur2+∑r=1d∑i=1NNirU⩽MB,∀d,
(48) EVΔ∼dU=EVmax∑i=1I∑j=1J∑d′=1d∑s=1SXid′jsDiU-∑r=1d∑i=1NNirU-B∼U-∑d′=1dR∼d′U,0,∀d,
(49) ∑i=1I∑j=1J∑d′=1d∑s=1SXid′jsDiU-∑r=1d∑i=1NNirU-αBUl+BUm2+1-αBUm+BUr2-∑d′=1dαRd′Ul+Rd′Um2+1-αRd′Um+Rd′Ur2⩽MU,∀d,
(50) ∑j=1J∑i=1N∑s=1SXidjs1-αLiSl+LiSm2+αLiSm+LiSr2βsd⩽Ksd,∀s,∀d.
In our formulation above, the objective and several constraints are not linear as the terms with EV(∙) are not linear. For example, EV(Δ∼djO) given in Equation (41) is not linear. In order to linearize this, we introduce a crisp auxiliary variable ΦdjO, and change the related constraints as follows:(51) ΦdjO⩾∑i=1N∑s=1SXidjsL∼iS-Tdj,∀j,∀d,
ΦdjO⩾0,∀j,∀d.
Then, we also have that EVΔ∼djO=EVΦdjO=ΦdjO,∀j,∀d. Moreover, constraint (50) can be expressed as:ΦdjO⩾∑i=1N∑s=1SXidjs1-αLiSl+LiSm2+αLiSm+LiSr2-Tdj,∀j,∀d.
Similarly, we can introduce crisp auxiliary variables ΦdW and ΦdU to linearize Equations (45), (47).ΦdW⩾∑d′=1d∑i=1N∑j=1J∑s=1SZiXid′js-∑e=1d∑i=1NNie-αBWl+BWm2+1-αBWm+BWr2-∑d′=1dαRd′Wl+Rd′Wm2+1-αRd′Wm+Rd′Wr2+∑d′=1d1-αRd′Ul+Rd′Um2+αRd′Um+Rd′Ur2+∑r=1d∑i=1NNirU,∀d,
ΦdW⩾0,∀d,
ΦdU⩾∑i=1I∑j=1J∑d′=1d∑s=1SXid′jsDiU-∑r=1d∑i=1NNirU-αBUl+BUm2+1-αBUm+BUr2-∑d′=1dαRd′Ul+Rd′Um2+1-αRd′Um+Rd′Ur2,∀d,
ΦdU⩾0,∀d.
A complete formulation of FECM after linearization is provided in Appendix D.
4.2 DE-OR Algorithm
We propose a heuristic rule called OR-heuristic and combine it with the Differential Evolution (DE) algorithm to obtain a hybrid genetic algorithm which we call the ‘DE-OR’ algorithm.
4.2.1 Framework of DE-OR
Code: The chromosome I consists of two substrings γ and β, that represent the patient's OR allocation and surgery date allocation, i.e.,I=γβ=o1,...,oNd1,...,dN,dN+ 1
where dN+1 is a dummy date, indicating that the patient has not been selected for surgery in the current decision period. For example, the OR of patient 1 is o1, and the operation date of patient 5 is d5.
Population Initialization: To improve the efficiency of the algorithm, we need to generate as many feasible solutions as possible. Here, we adopt a semi-random method based on heuristic rules. Specifically, we develop an OR-heuristic and then use it to generate the substring γ of the chromosome I. A semi-random method is used to produce the substring β, taking into account the latest start date of the patient and the availability of the surgeon on a given day. The specific process is detailed in Algorithm 1.Algorithm 1:Population Initialization
Input The latest date of the patient's operation H, the available date of the attending surgeon P, planning horizon T. m←0While m not in P do
for each h∈H doif h≤T thenm←rand1,helsem←rand1,T+1
n=OR_heuristic(m)
Output [m,n]
Fitness Function: Fitness function is defined asH(s)=O(s), ifsis feasible,C,otherwise,
where O(s) is the objective function and C is a constraint penalty.
Crossover and Mutation: The algorithm generates the substring β based on crossover and mutation, then generates γ based on the OR-heuristic and β. The ‘crossover and mutation’ process is not completely random since we ensure that the feasible solution is still a feasible solution after each operation.
Population Diversification: We remove duplicate individuals from the population after each iteration, and then use Algorithm 1 to regenerate it.
More details on algorithms similar to the one above can be found in Storn and Price (1997).
4.2.2 OR-heuristic
The algorithm includes an assignment operator and an adjust operator. To illustrate the process of the algorithm, we first present an example with 11 patients, 2 ORs with a maximum individual capacity of 45; see Figure 4.
Assignment Operator: In the first step, the operator groups the patients according to the admission date, i.e., Day 1 and Day 2. For each group, the operator sorts the patients in descending order according to the surgery duration. Each patient is inserted sequentially into the freest OR until all patients obtain their OR; here, the patients are assigned to the OR with the highest availability first.
Adjust Operator: For each day, the operator calculates the surgery duration of each OR, and then selects the OR with the largest overtime (OR1 for Day 1) and the OR with the idlest time (OR2 for Day 1). Since the capacity of each OR is 45 units of time, the operator can calculate the timeout of OR1, which is 46-45=1, and the idle time of OR2, which is 45-39=6. The patient with the shortest surgery duration in OR2 is p1, and its surgery duration is 19. Using this, a threshold L=19+1+6 will be calculated. In the interval [19,L], select the patient p9 with the largest surgery duration. Finally, exchange the ORs of patients p1 and p9. Repeat the above process until the patients on other days are also treated.
The algorithm process of OR assignment is summarized below.Algorithm 2:OR_Assignment
Input patient set P, patient's surgery duration S, patient's surgery date D, planning horizon T, operation room set R,Pr, patients belonging to the operating room rThe P is divided into sub-sets P1, P2, PT+1, according to the surgery date D. for l←1 to T+1 doPl←Sort in descending order (Pl)for each p∈Pl dofor each r∈R doCr←Calculate the remaining capacity of operation room rr←argmaxCrOp←rCr←Cr-Sp If minCr<0 and maxCr>0 thenrmin←argminCrrmax←argmaxCrpm←Find the patient with the smallest surgery time in PrmaxY←minCr+maxCr+SpmPY←Patients whose surgery duration in Prmin is less than the threshold Ypn←The patient with the longest surgery duration in PYL←OpnOpn←OpmOpm←L
Output O
We present several computational experiments to illustrate the efficacy of the DE-OR algorithm in the next section.
5 Computational Experiments
For the contributions mentioned in this paper, we have carried out the verification one by one in the experimental part. The first part is the effectiveness of the fuzzy model, the second part is the verification of the proposed methods for resource shortage and uncertainty in the COVID-19 environment, and the last part is the effectiveness of the transformation method and heuristic algorithm. Specifically, in this section, we design three sets of experiments. First, the robustness and adaptability of the fuzzy model are tested in the first experiment when the data is limited, and the parameter distributions are unknown. The second experiment studies the adaptability of the fuzzy model for extreme circumstances with resource shortages and uncertain capacities such as COVID-19 environment. At last, the experiment is designed to verify the efficacy of the hybrid (DE-OR) algorithm and FECM.
Experimental Design:
Our experiments are designed using two decision-making environments in terms of the availability of data and information on underlying parameter distributions. In the first environment, it is assumed that neither data nor information on the underlying parameter distribution is available. We call this the ‘no decision-making information (No-Info)’ environment. The second environment assumes either data or information on the underlying parameter distribution is available. Thus, we call this the ‘given decision-making information (Given-Info)’ environment. Figure 5 below depicts the two decision-making environments and data generation processes under each approach.Figure 5 No-Info and Given-Info decision-making environments.
For the No-Info environment, it is required to obtain fuzzy numbers through expert (surgeon) estimates in order to solve the fuzzy model. However, since fuzzy data is not readily available, Behmanesh and Zandieh (2019) replicate the surgeon’s process of generating fuzzy data (by using crisp data) and generate fuzzy processing times t-u,t,t+u, where u and v were randomly approximated between intervals of 1% to 30% of the deterministic duration t. This approach has also been employed and validated recently by Wang et al. (2020), and thereby, we also adopt the same approach in this paper. In the Given-Info environment, since either fuzzy data or the underlying parameter distribution is available, we can simply utilize the fuzzy model or available stochastic optimization models to solve the problem. In particular, our numerical experiment adopts the data configuration proposed by Min and Yih (2010), where surgical patients are considered under 9 groups, namely, ENT, OBGYN, ORTHO, NEURO, GEN, OPHTH, VASCULAR, CARDIAC, and UROLOGY. Table B.1 in Appendix B provides a summary of descriptive statistics for these 9 groups. Moreover, to test the performance of the model, we divide the history records into two subsets. The first subset is called the training set, which serves as the input to the model, and the second subset is called the test set, which is used for performance evaluation. Finally, our algorithm was programmed in Python, and executed on a 3.7 GHz Intel Core i7 CPU computer with 16 GB memory using solver CPLEX within the Python docplex package for the MIP model.
Regarding the impact of the number of operations on the model, we tested the performance under different numbers of operating rooms. Although we assume operating rooms are homogeneous for simplicity, our model can be adapted to situations with different operating room types. Since the surgeons have different skills or experience, the same operation could have different surgery duration under different surgeons; therefore, we express the surgery duration of each patient as a different fuzzy number. Moreover, it is assumed that the ratio of surgeons to patients is 1:4. In addition, since the priority of the patient is related to the patient’s condition, we set different unit waiting costs to capture this important heterogeneity in our model; nevertheless, within each class and feasible range, waiting costs are generated randomly between two bounds. The differences in the surgical needs are reflected in the matching of patients and surgeons, as well as differences in surgery duration. We consider non-elective surgery as an external uncertainty and express it as a fuzzy number based on expert experience. Moreover, we do not consider temporary cancellations of patients as cancellations are rare.
Different studies in the literature have set the unit overtime cost according to the actual operating conditions of the related hospital. For example, Gul et al. (2015) set the operation unit overtime cost to $13/min. The main part of the downstream hospitalization cost is the cost of care. Izady and Israa (2021) use the cost of human resources to represent the cost of care. Following their approach, considering the human resource cost of the actual hospital, we set the bed shortage cost to $100. It may not be possible to know the true value of the patient waiting cost as it is highly dependent on individuals. In this study, based on the information that we gathered during our interviews with the hospital officials, we set the unit waiting cost of patients to be a randomly generated number between $70 to $80. Several other studies have employed similar estimates in the literature; for example, Gul et al. (2015) regard waiting costs as a punishment and employ a spectrum of daily waiting costs between two bounds. 8 Table 4 summarizes our model parameters.Table 5. Table 4 Model parameters.
Planning horizon (D) one week
Unit overtime cost per operating room (ujO). Rand10,16
Unit cost per optional extra inpatient bed (uW). 100
Unit cost per optional extra ICU bed (uU). 500
Unit waiting cost of patient i (uiT). Rand70,80
Upper-bound for daily overtime hours of each OR (MO). 3 hours
Upper-bound on extra inpatient beds for each day (MB). 2
Upper-bound on extra ICU beds for each day (MU). 2
Open duration of ORj on day d (Tdj). 8 hours
Maximum working time of surgeons s on day d (Ksd). 11 hours
The latest date of the surgery of patient i (Duei). Rand1,2D
Total waiting days of patient i before the beginning of the current planning horizon (WiB). Rand0,D
Penalty coefficient of waiting time for patients deferred to the next planning horizon (θ). 1-10
The feasibility of decision vector (α). 0.6
Risk coefficient of the decision-maker's assessment of ICU needs (λ). 0.6
Note: For a<b, Randa,b denotes a randomly generated number from the uniform distribution over (a,b).
Table 5 Simulation results of different models.
Problem NP LS US CNE MODE FECM
ob cons ob cons ob cons ob cons ob cons
1 18 9680 3.40 36837 0.72 14657 0.72 14980 0.57 19746 0.35
2 40 181099 54.48 65611 0.00 37389 0.76 36751 0.75 23786 0.04
3 40 156040 19.32 64729 0.00 37031 0.86 36616 0.86 24190 0.12
4 40 149329 6.98 71244 0.00 37218 1.10 36413 1.10 27142 0.28
5 48 220909 80.69 101706 0.00 48174 2.50 47067 1.00 45891 0.46
6 48 197943 50.64 99693 0.00 45787 0.87 46570 0.92 44902 0.62
7 70 282255 36.08 161099 0.00 66490 2.21 65218 15.96 73562 3.64
8 70 321212 217.06 148628 0.00 64339 1.24 67221 1.88 61376 1.03
9 80 376338 582.96 146433 0.00 78151 2.26 80024 12.22 53096 1.18
10 80 320126 216.28 141464 0.00 77179 9.32 74284 1.42 53523 3.72
11 90 421228 180.66 194405 0.00 92297 3.65 87300 2.02 80405 0.82
12 90 392636 177.58 184652 0.00 93593 3.00 89831 2.37 76180 0.80
13 100 417802 85.32 174112 0.00 98398 2.06 101064 4.04 72653 1.52
14 100 408013 62.93 171656 0.24 101501 2.10 103263 3.60 73222 3.02
15 110 413234 178.20 166377 0.00 108109 28.16 109685 24.94 70319 0.80
16 110 415357 211.67 161410 0.00 104783 6.71 102108 7.26 67775 4.14
17 120 454517 108.00 212156 0.02 122262 7.80 125478 5.06 84835 2.42
18 120 480787 192.48 201256 0.00 126226 3.50 126174 6.75 86661 3.22
19 150 532588 100.98 281970 0.02 158976 7.42 155041 7.68 121713 4.96
20 150 482209 87.21 176132 0.02 147631 19.94 154943 12.49 89149 0.57
Average 331665 132.65 148078 0.05 83010 5.31 83002 5.64 62506 1.69
5.1 Robustness and Adaptability of the Fuzzy Model
Mulvey et al. (1995) describe robustness as the robustness of the model as well as the solution. A robust model should be less sensitive against modifications while the number of violated constraints should be minimized under a more robust solution; a model that possesses both of these characteristics can be thought of as an adaptable model. This notion of robustness has also been employed in Gorissen et al. (2015) as a performance measure of robustness. Therefore, we use the following formulas to describe the performance and robustness of our model:ob:=1N×∑n=1Nobn;
cons:=1N×∑n=1Nconsnumn,
where n is the problem index in the simulation experiment, obn is the objective value under the solution in the simulation experiment, and consnumn represents the number of times wherein the current scheduling solution exceeds the maximum capacity limit in the simulation environment. For example, consider a scheduling plan derived from the model for a 5-day working schedule with 2 operating rooms. For this plan, if a realization in the simulation environment produces OR overtime numbers of {190, 170, 160, 120, 170, 150, 120, 110, 130, 170}, then the number of times the OR capacity exceeds the limit (MO=180) is 1. When the number of extra inpatient beds per day is {0,3,1,2,3}, the number of times it exceeds the capacity limit (MB=2) is 2. Moreover, if the number of extra ICU beds per day is given by {1,1,0,2,3}, then the number of times it exceeds the capacity limit (MU=2) is 1. Thus, the total number of violations is consnumn=1+2+1=4. Consequently, cons reflects the adaptability of the solution scheme and represents the robustness of the model.
We next evaluate the performance and robustness of the fuzzy model by comparing it with several crisp (non-fuzzy) models. Following Lee et al. (2014), we set the modal point (MODE), center (CNE), lower bound (LS), and upper bound (US) of the support set of a TFN as the parameter values of the crisp models as shown in Figure 6 . In Figure 6, m represents the degree of membership, and x represents the elements in the TFN. It must be noted that all the models are solved using the DE-OR algorithm since it is suitable for both the fuzzy and crisp models.Fig 7. Figure 6 Parameters of the crisp models.
Figure 7 Simulation results of different models.
As shown in Table 5 and Figure 7, not only FECM method consistently produces the lowest ob values but also generates very robust solutions as indicated by cons. While US is the only model that outperforms the FECM model in terms of ‘cons’, this model produces inferior ob values to the FECM model. Consequently, the scheduling scheme of FECM can adapt to an uncertain environment with high accuracy.
Stochastic optimization is a widely proven method that can effectively deal with uncertain environments. To prove the applicability of the fuzzy model, we compare it with the classical stochastic model. In addition, the comparison is to show that the fuzzy model can deal with the surgery schedule problems in environment both with data and without data. We compare the performance of the fuzzy model with a well-known stochastic model (SM) proposed by Min and Yih (2010) in this subsection. Min and Yih (2010) use the sample average approximation (SAA) to solve their model. For fairness of comparison, we only consider the uncertainty in SD, LOSW, and LOSI, while all other parameters are assumed to be known. Moreover, to highlight the benefits of the fuzzy model, we select small-scale problems in our experiments. Finally, note that an SM requires an environment with known data while the fuzzy model requires fuzzy data. Thus, we use the given data to generate fuzzy data by representing the patients’ SD, LOSW, and LOSI as TFN L∼i(.)=(Li(.)l,Li(.)m,Li(.)r), where Li(.)l, Li(.)m, and Li(.)r respectively represent the lowest value, mean value, and maximum value of the respective parameter distribution.
As shown in Table 6 , the crisp model (CM) has relatively higher ob and cons, indicating that the CM solution leads to higher costs and the CM does not adapt to uncertain environments easily. While the SM is very robust and associated with smaller objective values for some test problems, when the number of patients exceeds a certain threshold, the solution can be very long under the SM. Moreover, we observe that the performance of FECM and SM are similar although the solution time under FECM is much shorter than that under SM. Finally, as the sample size increases, both the SAA-based SM and the FECM usually cannot produce optimal results within a reasonable time frame whereas the DE-OR-based FECM (FECM-I) method produces a satisfactory solution in a quick time.Table 6 Comparison of CM, SM and FECM.
Problem NP CM SM FECM FECM-I
ob cons T ob cons T ob cons T ob cons T
1 16 12995 0.21 0.11 9197 0.17 200 9253 2.24 0.12 8980 0.01 173
2 18 12240 0.18 0.08 8246 0.02 137 8118 0.01 0.08 8007 0 170
3 20 12142 0.15 0.23 9518 0.04 454 9835 0.1 0.11 9386 0.01 176
4 25 20846 0.22 0.45 15163 0.03 >36k 14868 0.02 0.5 14783 0.02 181
5 30 25340 2.63 0.14 14523 0.21 >36k 14726 0.07 1.84 15094 0.03 188
6 35 36783 0.73 2.47 23806 0.15 >36k 22164 0.06 8.55 21049 0.04 195
7 40 37949 0.8 >36k 25267 0.17 >36k 24335 0.04 >36k 24752 0.08 199
8 48 60732 1.81 >36k 84733 0.03 >36k 55285 1.3 >36k 61390 3.81 212
9 60 61629 1.26 >36k 44509 0.28 >36k 41819 0.26 >36k 42734 0.9 311
Note: CM and FECM are solved by CPLEX; SM is solved by SAA; FECM-I is the model FECM solved by DE-OR.
According to the above experimental results, the adaptability of the fuzzy model to the uncertain environment is not much different from that of the stochastic optimization model. In comparison with the stochastic optimization model, the fuzzy model has a lower solution time complexity and requires a shorter solution time, thereby increasing the responsiveness of the model in an uncertain environment. Finally, it must be reiterated that fuzzy models are mainly suitable for uncertain environments with unknown parameter distributions. One of the main advantages of the fuzzy approach over stochastic programming is that the uncertain parameters do not have to follow any statistical distribution.
5.2 Adaptability of the Fuzzy Model for Uncertainty during COVID-19 Pandemic
In an uncertain environment, not only the patient’s SD, LOSW, and LOSI are uncertain, but also the available resources are highly uncertain. In particular, amid the pandemic outbreaks such as COVID-19, ICU resources become scarce. In this subsection, we focus on the adaptability of models when ICC and ICD are uncertain. To reduce the interference of the uncertainty of SD, LOSW, and LOSI in our findings, it is assumed that these parameters are given. Moreover, two separate sets of experiments are designed to study the effects due to uncertainty in each ICD and ICC. Specifically, Experiments 1 and 2 respectively consider the uncertainty in ICC and ICD while keeping the other component fixed.
In Experiment 1, we select the traditional ICU bed assessment policy for comparison, where the decision-makers use a certain percentage of surgical patients as the estimated number of patients who need ICU care after the surgery; these percentages are denoted by RT in Table 7 . Moreover, we set NP=70 and NR=4. For the estimation of the capacity parameters of the fuzzy model, we use the following three levels: L1=(6,12,15), L2=(16,20,30), and L3=(30,35,40). In the experiment, different from the input data of the model earlier, the simulation data simulates the ICC when the actual plan is executed, whereas the earlier model estimates ICC before the plan is implemented. For our simulation data, ICC is randomly generated from the support of the fuzzy number (between LS and US) in the simulation environment; thus, we denote it by S B, and consider the following three value ranges: S1=[6,15], S2=[16,30], and S3=[30,40]. The results of the experiment are given in Table 7 below, where BU represents the estimated number of available ICU beds in CM, B∼U represents the estimated fuzzy number of available ICU beds in FECM, WC denotes the waiting cost of all patients, IC denotes the extra ICU bed cost of all patients, and TC is calculated by adding WC and IC. WC and IC are the sub-objectives of the proposed model. We chose these evaluation criteria since the uncertainty of ICC and ICD mainly affects the admission, i.e., patient waiting time and extra ICU bed demand.Table 7 Performance of FECM and CM under the uncertainty in ICC.
RT BU B∼U SB CM FECM
WC IC TC WC IC TC
0.1 7 L1 S1 33700 745 34445 32940 821 33761
0.1 7 L2 S2 35000 0 35000 28500 0 28605
0.1 7 L3 S3 34040 0 34040 27500 0 27515
0.3 21 L1 S1 27640 10914 38599 32940 3030 35970
0.3 21 L2 S2 31320 0 31380 31860 0 31905
0.3 21 L3 S3 27940 0 28000 28220 0 28220
0.5 35 L1 S1 29040 9286 38431 35280 855 36135
0.5 35 L2 S2 27680 0 27695 28240 0 28240
0.5 35 L3 S3 29900 0 29900 30380 0 30380
0.7 49 L1 S1 30900 3014 34064 36680 203 36883
0.7 49 L2 S2 29300 2 29302 29860 0 29860
0.7 49 L3 S3 27640 0 27640 28280 0 28280
0.9 63 L1 S1 29480 3155 32725 34860 67 34927
0.9 63 L2 S2 28500 0 28710 28880 0 28880
0.9 63 L3 S3 30020 0 30155 30000 0 30000
Firstly, it is evident from our experiment that pessimistic ICC estimates by hospital managers can lead to a reduction in the number of scheduled elective surgeries, and thereby reduce or eliminate the cost of extra ICU beds; e.g., RT=0.1, SB=S3, WC=34,040, and IC=0 in Table 7. In contrast, more optimistic estimates by the managers can lead to a higher number of scheduled surgeries; while this can reduce the patient waiting costs, it will lead to a serious shortage of ICU beds and increase the cost therein, e.g., RT=0.9, SB=S1, WC=29,480, and IC=3155.
Figure 8 compares the FECM method with the CM under small (S1), medium (S2), and large (S3) scale scenarios. As seen in Figure 8, when the value of RT is small, i.e., approximately lower than 0.3, FECM clearly dominates CM while it is mostly very comparable with CM for other values of RT. Note that when RT>0.3, the cost of CM and FECM looks similar because we have not calculated the fixed cost of beds. In fact, when RT>0.3, the CM model needs to reserve a large number of bed resources, which is difficult to achieve in the resource shortage environment.Figure 8 The total cost under models FECM and CM.
In Experiment 2, the ICD for CM is a randomly generated 0-1 sequence. Specifically, we first generate a random set ‘SS’ between [0,1]. For each element of SS, if it is greater than λ, then it is rounded up to 1, otherwise, it is rounded down to 0, then we get 0-1 set S. The ICD of the fuzzy model is crisp set S. The simulations are then carried forward using SS=SS∗(1±0.3) in each step. Finally, the λ defined in subsection 3.2 represents the decision maker's assessment of ICD in FECM.
Figure 9 and Table 8 depict TC, IC, and WC as functions of λ. As seen in Figure 9, when the hospital managers are more pessimistic as characterized by a smaller value of λ, it is less likely to exceed the ICU capacity. At the same time, the patients’ waiting costs are higher as expected. Interestingly, the total cost stays flat as a function of λ; thus, the model is not affected by the risk preference of the decision-maker and can be considered fairly robust. In addition, we found that only when the decision-maker is extremely conservative or extremely optimistic, FECM performs slightly worse than CM. Thus, our results suggest that a more moderate approach might be the best approach with fuzzy data.Figure 9 Costs as a function of the preference coefficient (λ) of the decision-maker.
Table 8 Performance of FECM and CM under the ICD uncertainty.
λ CM FECM
WC IC TC WC IC TC
0.1 29400 4107 33507 34100 692 34792
0.3 31080 3746 34826 33280 164 33444
0.5 32620 6827 39447 33400 1363 34763
0.7 32100 3961 36061 31880 5292 37172
0.9 30560 3072 33632 30160 3132 33292
5.3 Efficacy of the Solution Methods
In this section, we conduct extensive numerical experiments to demonstrate the efficacy of the transformation method and heuristic algorithm for the elective surgery scheduling problem.
To further demonstrate the effectiveness of the transformation method, we compare the proposed crisp model (FECM) with other models by the same solver to avoid interference from the heuristic algorithm. We select Rodriguez’s and Niu’s models (Abdullah and Abdolrazzagh-Nezhad, 2014, Niu et al., 2008), two transformation models (named FRM and FNM in this paper), as our comparison models. In addition, the model MODE is also a comparison model. The experimental results are shown in Figure 10 . For small-scale test problems 1-6, all the models can be solved within 800 seconds, but for large-scale problem 7, the solution times of all the models increase exponentially. Therefore, there is no significant difference in computational complexity between FECM and FRM, MODE for small-scale problems. On the other hand, FECM has a lower ob and cons, indicating that it has better environmental adaptability. In addition, we found that the cons of FECM are stable and do not change with the increase of the problem size, which shows that it has excellent robustness. On the contrary, MODE, FRM and FNM fluctuate considerably.Figure 10 The comparison of model FECM and other transformation models.
The model MODE is a simplification of the fuzzy model according to fuzzy theory. For example, without loss of generality, the surgery duration of the fuzzy model is L∼iS=(LiSl,LiSm,LiSr), while that of MODE is LiSm by assuming LiSl=LiSm=LiSr. Therefore, the complexity of MODE is less than that of the fuzzy model. The experimental results show that there is no significant difference in the computational complexity between MODE and FECM, which indirectly proves that the complexity of FECM is not greater than that of the original fuzzy model.
For DE-OR algorithm, we have selected the following representative algorithms for comparison:• CPLEX: CPLEX optimization software package can be used to solve integer linear programming models CM and FECM and obtain accurate solutions for small-scale problems. However, this process could be time-consuming and may not produce optimal solutions for large-scale problems (Zhou, Geng, Jiang, & Wang, 2018)
• GA: This is a standard genetic algorithm (Bonabeau et al., 1999), except that it uses the same initial population method as in DE-OR (to generate many feasible solutions); all other algorithms use the classic tournament selection, two-point crossover, and insertion mutation.
• PSO-GA:Niu et al. (2008) proposed this hybrid particle swarm PSO-GA algorithm to solve the job-shop scheduling problem. Considering the similarity between surgery scheduling and job shop scheduling, we applied it to our problem for comparison.
• DE: This is a standard Differential Evolution algorithm (Storn and Price, 1997), except that it uses the same initial population method as in DE-OR (to generate many feasible solutions).
• FDE: This is an improved Differential Evolution algorithm (Tsafarakis, Zervoudakis, Andronikidis, & Altsitsiadis, 2020).
• GA-VNS: This is a hybrid algorithm based on GA and variable neighborhood search (Wang et al., 2021).
Relative Percentage Deviation (RPD) has been used as a general performance indicator to evaluate the optimization effect of the algorithm. Since each algorithm needs n runs, following Wang et al. (2018), we use the Average Relative Percentage Deviation (ARPD) defined below as the performance evaluation indicator:ARPD=1n∑l=1nRlk-RbestRbest×100
where Rk represents the target value of algorithm k, Rbest represents the target value of the optimal solution, and n represents the number of runs.
In our experiment, test problems 1-6 consider small-scale cases whereas problems 7-13 focus on large-scale cases. Table 9 and Figure 11 present our experimental results for the small-scale cases, where NP, BU, BW, and NR respectively represent the number of the patients, the available number of ICU beds, the available number of inpatient beds, and the number of ORs; T and OB respectively denote the time to convergence in seconds and objective function value. In this case, CPLEX produces the optimal results (Rbest) and thus, we can calculate the ARPD under each algorithm. As we noted above, one contribution of this study is to transform the fuzzy model into a simpler MIP model that can be handled directly using the commercial solver CPLEX, and this approach is also useful when employing general meta-heuristics such as GA and DE. As can be seen from Table 9, for small scales, our model can be solved quickly via CPLEX. In comparison, the meta-heuristic algorithm shows better stability. Moreover, compared with the classical heuristic algorithms GA, DE, and PSO-GA, the accuracy and speed of DE-OR are exceptional. By comparing the values of ‘ARPD’ and ‘T’ in Table 9 and Figure 11, it is easily seen that the proposed DE-OR algorithm outperforms both classical heuristic algorithms GA and DE as well as the hybrid PSO-GA algorithm based on particle swarm and genetic algorithm. It should also be noted that the DE-OR algorithm produces solutions relatively quickly compared to PSO-GA, GA, and DE algorithms.Table 9 Comparison of experiment results under different algorithms (Small-scale case).
Problem NP BU BW NR CPLEX PSO-GA GA DE DE-OR
OB T ARPD T ARPD T ARPD T ARPD T
1 18 1 6 2 13075 0.16 0.38 291 22.56 309 22.56 289 0 281
2 40 4 40 2 13901 12 4.86 333 6.29 370 4.55 334 0.89 324
3 40 4 35 2 13950 2 7.81 336 4.3 372 3.76 334 1.19 324
4 40 4 30 2 14025 8 6.95 336 4.91 371 4.81 335 2.14 324
5 48 4 45 2 18375 323 5.25 348 5.27 387 5.14 349 1.77 340
6 48 4 40 2 18551 2450 7.75 349 6.39 387 5.12 349 1.88 340
Figure 11 Comparison of experiment results under different algorithms (Small-scale case).
In order to minimize the selection bias in our comparison, we chose the most advanced ones such as FDE and GA-VNS. Furthermore, to prove the effectiveness of the algorithm in large-scale scenarios, we set up large-scale test cases. For the large-scale cases, CPLEX does not converge to the optimal solution in a reasonable amount of time even after increasing the value of the relative MIP gap tolerance to 0.1 (see values of ‘T’ under CPLEX in Table 10 ). Note that DE-OR has a significant advantage over solver (CPLEX) in terms of solution time when the problem scale is large and there is a shortage of resources such as ICU and inpatient beds (BU and BW). As we can see, the solution time in the CPLEX solver easily exceeds 10 hours and this is not acceptable for a time-sensitive process such as surgery scheduling. In contrast, the DE-OR algorithm produces a satisfactory solution within 0-0.3 hours even for large-scale cases. Finally, as shown in Figure 12 , the DE-OR algorithm clearly dominates all other algorithms by producing a higher value for the ARPD. (SEE Table A1. )Table 10 Comparison of experiment results under different algorithms (Large-scale case). *Note: The test probems 7-13 are executed using 0.2, 0.11, 0, 0.14, 0.23, 0.19, and 0.33, respectively, as the value of the relative MIP gap tolerance.
Problem NP BU BW NR CPLEX FDE GA-VNS DE-OR
OB T* ARPD T ARPD T ARPD T
7 70 7 65 3 28900 >36k 96.64 732 18.43 616 3.3 548
8 70 7 55 3 28868 >36k 40.85 749 4.72 614 1.89 560
9 80 8 70 4 28725 >36k 34.73 891 4.18 715 1.31 669
10 80 8 60 4 28575 >36k 43.18 875 3.94 705 2.1 650
11 90 9 80 4 33075 >36k 66.29 1048 6.55 866 1.81 781
12 90 9 70 4 33375 >36k 109.13 1036 6.25 863 2.02 776
13 100 10 70 5 37425 >36k 67.46 1334 5.81 1102 1.6 981
Figure 12 Comparison of experiment results under different algorithms (Large-scale case).
Table A1 List of Abbreviations and Acronyms
DE Differential Evolution
DE-OR Hybrid Genetic (algorithm)
FM Fuzzy model
CM Crisp(non-fuzzy) model
FECM Equivalent crisp(non-fuzzy) model
ICU Intensive Care Unit
SD Surgery Duration
ICD ICU Demands
ICC Available ICU capacity
LOSI Length of Stay (in ICU)
LOSW Length of Stay (in the ward)
LS Lower Bound (crisp model)
US Upper bound (crisp model)
CNE Center (crisp model)
MODE Modal Point (crisp model)
OR Operating Room
PACU Post-Anesthesia Care Unit
PHU Pre-operative Holding Unit
RO Robust Optimization
SO Stochastic Optimization
TFN Triangular Fuzzy Number
Table B1 The structure of test problems
Surgical group SD (minute) LOSI (day) LOSW (day) Observations
Mean Standarddeviation Mean Standarddeviation Mean Standarddeviation
ENT 74 37 0.1 0.1 3 1 788
OBGYN 86 40 2 2 2 2 342
ORTHO 107 44 1.5 1.5 1 2 859
NEURO 160 77 2 2 2 2 186
GEN 93 49 0.05 0.05 3 1 817
OPHTH 38 19 0.05 0.05 4 1 110
VASCULAR 120 61 3.5 3.5 5 2 303
CARDIAC 240 103 2 2 2 2 90
UROLOGY 64 52 0.8 0.8 6 1 198
Note:Min and Yih (2010) also generated SD and LOSI data using this approach.
6 Discussion and Concluding Remarks
We have studied the scheduling problem of elective surgeries as an uncertain system where the downstream includes both the ICU and ward. The proposed DE-OR algorithm is proven to be computationally efficient and more effective than the existing algorithms for the research problem in this paper. For example, the DE-OR algorithm outperforms both classical heuristic algorithms GA and DE as well as the hybrid PSO-GA algorithm based on particle swarm and genetic algorithms; moreover, when the problem scale is large and there is a shortage of resources such as ICU and inpatient beds, the DE-OR algorithm clearly dominates popular algorithms of CPLEX in terms of the solution time. The robustness and adaptability of the proposed fuzzy model and its solution via the algorithm DE-OR is shown to be superior to extant crisp heuristic methods such as modal point (MODE), center (CNE), lower bound (LS), upper bound (US). We have further shown that the solution time under FECM is much shorter than that under the SAA-based SM although the performance of FECM and SM are similar; moreover, our experiments show that while both the SAA-based SM and the FECM take a very long time to produce optimal results as the sample size increases, the optimal solution to the proposed FECM method can be computed relatively quickly using the DE-OR algorithm. Finally, we have also demonstrated the adaptability of the fuzzy model to scenarios with uncertainty in ICU demand and capacity as well as shortages in resources amid the pandemic such as COVID-19. The fuzzy method is based on expert experience and inference to describe the uncertain information of the surgical process. This method can be very useful when scheduling with uncertainties if historical data is insufficient or the existing medical environment undergoes major changes. Moreover, when an environment changes rapidly and the dynamic of the associated processes switches regimes, historical data can no longer predict future trends, and fuzzy models can be a solution for such scenarios.
In practical application, the solution time by the solver may be too long for large-scale problems. To make up for the defect of this practical application, we propose the algorithm DE-OR for our research problem, which can meet some environments with high requirements on scheduling time. The performance of the algorithm (DE-OR) is compared with different heuristic algorithms, yet DE-OR is highly customized for the specific model in the paper. Our future research will focus on the algorithm in more detail to further assess its superiority.
We have considered the initial state at the beginning of each planning horizon such as the initial waiting time of the patient and the availability of inpatient and ICU beds at the beginning and during a planning horizon. Thus, our formulation also facilitates replanning as the problem can be reformulated at any time for a given set of initial conditions. However, when the disturbances of the initial schedule get too large, the replanning problem needs a more rigorous treatment. While modeling the replanning horizon problem is beyond the scope of this paper, it will be an interesting and challenging problem for future research. Moreover, it should be noted that the hospital profit analysis may also include other costs such as the cost of empty wards and the cost of patients, surgeons, staff, operating room, etc. Although we have focused on fulfilling the patients’ surgery requirements while avoiding the risk of overloading due to insufficient resources, it will be interesting to see how one can incorporate other financial performance measures associated with surgeries in this setting as future research.
Data Availability
The datasets generated during and/or analyzed during the current study are available from the online materials alongside with this article or the corresponding author on reasonable request.
Uncited references
Ayvaz-Cavdaroglu and Huh, 2010, Beninato et al., 2022, Dai et al., 2022, Nguyen et al., 2022, Norris et al., 2021.
CRediT authorship contribution statement
Zongli Dai: Investigation, Methodology, Software, Visualization, Writing – original draft. Sandun Perera: Investigation, Supervision, Writing – review & editing. Jian-Jun Wang: Conceptualization, Methodology, Funding acquisition, Project administration, Resources, Supervision, Writing – review & editing. Sachin Kumar Mangla: Writing – review & editing. Guo Li: Methodology, Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A .
Appendix B .
Appendix C .1: Definition of a fuzzy set
Let U be a classical set of objects, called the universe, whose generic elements are denoted by x, i.e., U=x. A fuzzy set A in U is characterized by a membership function μA(x), which associates each element in U with a real number in [0,1]. A fuzzy set Ais usually denoted by the set of pairsA=x,μAx,x∈U
For example, for an ordinary set A,μAx=1,ifx∈A0,ifx∉A
When U is a finite set such that U=x1,x2,⋯,xm, a fuzzy set A in U can be represented asA=∑i=1nxi/μAxi
Note that symbols ‘Σ’ and symbol ‘/’ are not traditional summation and division symbols here.
Appendix C .2: Definition of triangle fuzzy number
If T=aL,aM,aR, where 0<aL⩽aM⩽aR, T is called a triangular fuzzy number, and its membership function can be expressed as:μA(x)=x-aLaM-aL,ifaL<x<aMx-aRaM-aR,ifaM<x<aR0,otherwise
Appendix D .
Min∑i=1NuiTΔiT+∑d=1D∑j=1JujOΦdjO+∑d=1DuWΦdW+∑d=1DuUΦdU,
ΔiT=∑d=1D∑j=1J∑s=1SHdAXidjs+WiB∑d=1D∑j=1J∑s=1SXidjs+θ·D·1-∑d=1D∑j=1J∑s=1SXidjs,∀i,
ΦdjO⩾∑i=1N∑s=1SXidjs1-αLiSl+LiSm2+αLiSm+LiSr2-Tdj,∀j,∀d,
ΦdjO⩾0,∀j,∀d,
ΦdjO⩽MO,∀j,∀d,
∑d=1D∑j=1J∑s=1SXidjsHdA+12LiUl+LiUm2+12LiUm+LiUr2∑d=1D∑j=1JXidjDiU·Zi=∑r=1D+QUHeENirU,∀i,
∑r=1D+QUNirU⩽1,∀i,
∑d=1D∑j=1J∑s=1SXidjsZiHdA+∑d=1D∑j=1J∑s=1SXidjsZi12LiWl+LiWm2+12LiWm+LiWr2·1-DiU+DiU=∑e=1D+QWHeENie,∀i,
∑e=1D+QWNie⩽1,∀i,
ΦdW⩾∑d′=1d∑i=1N∑j=1J∑s=1SZiXid′js-∑e=1d∑i=1NNie-αBWl+BWm2+1-αBWm+BWr2-∑d′=1dαRd′Wl+Rd′Wm2+1-αRd′Wm+Rd′Wr2+∑d′=1d1-αRd′Ul+Rd′Um2+αRd′Um+Rd′Ur2+∑r=1d∑i=1NNirU,∀d,
ΦdW⩾0,∀d,
ΦdW⩽MB,∀d,
ΦdU⩾∑i=1I∑j=1J∑d′=1d∑s=1SXid′jsDiU-∑r=1d∑i=1NNirU-αBUl+BUm2+1-αBUm+BUr2-∑d′=1dαRd′Ul+Rd′Um2+1-αRd′Um+Rd′Ur2,∀d,
ΦdU⩾0,∀d,
ΦdU⩽MU,∀d,
∑j=1J∑s=1SXidjs⩽∑s=1Sβsd,∀i,∀d,
∑j=1J∑i=1N∑s=1SXidjs1-αLiSl+LiSm2+αLiSm+LiSr2βsd⩽Ksd,∀s,∀d,
∑d=1Duei∑j=1J∑s=1SXidjs=1,∀i,
∑d=1D′∑j=1J∑s=1SXidjs=1,∀i,
∑d=1D′∑j=1JXidjs=Yis,∀i,s,
∑i=1NHiIDiU=Av,∀v=1,2,...,|A|,
DiU⩽1,∀i.
Data availability
Data will be made available on request.
Acknowledgments
The authors are indebted to the Editor in Chief, Dr. Yasser Dessouky, for his precious time, the Area Editor and the two anonymous reviewers for their constructive comments and helpful suggestions that helped improve this paper considerably.
1 All abbreviations and Acronyms see Appendix A.
2 The authors conducted a field study at a hospital in Liaoning Province, China, where they interviewed doctors, operating room nurses, ICU nurses, ward nurses, surgical scheduling staff, and hospital administrators.
3 In our survey, we searched Web of Science databases for articles belonging to the area of operations research and management science from 2010 to 2021. The search term includes a combination of the following words: surgical scheduling, COVID-19, resource shortage, uncertainty.
4 It must be noted that the job selection, “patient selection” by the hospital decision-makers, will be done optimally in our setting.
5 Before surgery, each patient will be assigned a bed; so, every non-outpatient surgery patient will occupy at least one bed for a day. Patients who enter the ICU directly after surgery occupy a bed for one day to improve the utilization of hospital beds, after entering the ICU, the beds will be allocated to other patients.
6 The membership degree is different from the binary logic of traditional sets where it is only indicates whether an element belongs to the set or not.
7 For any two TFNs L∼1=L1l,L1m,L1r and L∼2=L2l,L2m,L2r, the addition, subtraction, and scalar multiplication can be defined as L∼1+L∼2=L1l+L2l,L1m+L2m,L1r+L2r;L∼1-L∼2=L1l-L2r,L1m-L2m,L1r-L2l;and λL∼1=λL1l,λL1m,λL1r for λ>0, λL∼1=λL1r,λL1m,λL1l for λ<0.
8 While the cost calculations in Gul et al. (2015) show that elective surgery patients are time sensitive, Zhang et al. (2019) assume that patients scheduled for surgery during the planning horizon will not incur waiting costs; however, the authors introduce a higher waiting cost when the waiting time is longer than the planning horizon.
==== Refs
References
Abdullah S. Abdolrazzagh-Nezhad M. Fuzzy job-shop scheduling problems: A review Information Sciences 278 2014 380 407 10.1016/j.ins.2014.03.060
Ayvaz-Cavdaroglu N. Huh W. Allocation of hospital capacity to multiple types of patients Journal of Revenue and Pricing Management 9 2010 386 398 10.1057/rpm.2010.30
Bastos L.S.L. Marchesi J.F. Hamacher S. Fleck J.L. A mixed integer programming approach to the patient admission scheduling problem European Journal of Operational Research 273 3 2019 831 840 10.1016/j.ejor.2018.09.003
Batun S. Denton B.T. Huschka T.R. Schaefer A.J. Operating room pooling and parallel surgery processing under uncertainty INFORMS Journal on Computing 23 2 2011 220 237 10.1287/ijoc.1100.0396
Behmanesh R. Zandieh M. Surgical case scheduling problem with fuzzy surgery time : An advanced bi-objective ant system approach Knowledge-Based Systems 186 2019 104913 10.1016/j.knosys.2019.104913
Bellman R. Zadeh L.A. Decision-making in fuzzy environment Management Science 17 4 1970 B141 B164 10.1287/mnsc.17.4.B141
Beninato T. Laird A.M. Graves C.E. Drake F.T. Alhefdhi A. Lee J.A. …Lubitz C.C. Impact of the COVID-19 pandemic on the practice of endocrine surgery The American Journal of Surgery 2022 10.1016/j.amjsurg.2021.07.009
Best M.J. Aziz K.T. McFarland E.G. Anderson G.F. Srikumaran U. Economic implications of decreased elective orthopaedic and musculoskeletal surgery volume during the coronavirus disease 2019 pandemic International Orthopaedics 44 11 2020 2221 2228 10.1007/s00264-020-04713-8 32681371
Bonabeau, E., Dorigo, M., & Theraulaz, G. (1999). Swarm intelligence: from natural to artificial systems. In New York,United States:Oxford University Press. https://doi.org/10.1093/oso/9780195131581.001.0001
Bovim T.R. Christiansen M. Gullhav A.N. Range T.M. Hellemo L. Stochastic master surgery scheduling European Journal of Operational Research 285 2 2020 695 711 10.1016/j.ejor.2020.02.001
Cao C. Liu Y. Tang O. Gao X. A fuzzy bi-level optimization model for multi-period post-disaster relief distribution in sustainable humanitarian supply chains International Journal of Production Economics 235 2021 108081 10.1016/j.ijpe.2021.108081
Chung W.H. Mihara Y. Chiu C.K. Hasan M.S. Chan C.Y.W. Kwan M.K. Factors affecting operation duration in posterior spinal fusion (psf) using dual attending surgeon strategy among lenke 1 and 2 adolescent idiopathic scoliosis (AIS) patients Clinical Spine Surgery 35 1 2022 18 23 33979103
Dai Z. Wang J.J. Shi J.(Junmin) How does the hospital make a safe and stable elective surgery plan during COVID-19 pandemic? Computers and Industrial Engineering 169 May 2022 108210 10.1016/j.cie.2022.108210
Denton B.T. Miller A.J. Balasubramanian H.J. Huschka T.R. Denton B.T. Miller A.J. Optimal allocation of surgery blocks to operating rooms under uncertainty Operations Research 58 4 2010 802 816 10.1287/opre.1090.0791
Eun J. Kim S.-P. Yih Y. Tiwari V. Scheduling elective surgery patients considering time-dependent health urgency : Modeling and solution approaches Omega 86 2019 137 153 10.1016/j.omega.2018.07.007
Freeman N.K. Melouk S.H. Mittenthal J. A scenario-based approach for operating theater scheduling under uncertainty Manufacturing and Service Operations Management 18 2 2016 245 261 10.1287/msom.2015.0557
Gabriel L. Marcelloni F. Cecílio M. Cesar L. Carpinetti R. Exploring the relations between supply chain performance and organizational culture : A fuzzy grey group decision model International Journal of Production Economics 233 2021 108023 10.1016/j.ijpe.2020.108023
Gerchak Y. Gupta D. Henig M. Reservation planning for elective surgery under uncertain demand for emergency surgery Management Science 42 3 1996 321 334 10.1287/mnsc.42.3.321
Gonzalez-Rodriguez I. Puente J. Vela C.R. Varela R. Semantics of schedules for the fuzzy job-shop problem IEEE Transactions on Systems, Man, and Cybernetics 38 3 2008 655 666 10.1109/TSMCA.2008.918603
Gorissen B.L. Yanikoğlu I. den Hertog D. A practical guide to robust optimization Omega 53 2015 124 137 10.1016/j.omega.2014.12.006
Gul S. Denton B.T. Fowler J.W. Denton B.T. Fowler J.W. A progressive hedging approach for surgery planning under uncertainty INFORMS Journal on Computing 27 4 2015 755 772
Heilpern S. The expected value of a fuzzy number Fuzzy Sets and Systems 47 1 1992 81 86 10.1016/0165-0114(92)90062-9
Izady N. Israa M. A clustered overflow configuration of inpatient beds in hospitals Manufacturing & Service Operations Management 23 1 2021 139 154 10.1287/msom.2019.0820
Jebali A. Diabat A. A stochastic model for operating room planning under capacity constraints International Journal of Production Research 53 2015 1 19 10.1080/00207543.2015.1033500
Jiménez M. Arenas M. Bilbao A. Rodríguez M.V. Linear programming with fuzzy parameters: An interactive method resolution European Journal of Operational Research 177 3 2007 1599 1609 10.1016/j.ejor.2005.10.002
Kliff S. Hospitals Knew How to Make Money 2020 Then Coronavirus Happened The New York Times Retrieved from https://www.nytimes.com/2020/05/15/us/hospitals-revenue-coronavirus.html
Kumar A. Costa A.M. Fackrell M. Taylor P.G. A sequential stochastic mixed integer programming model for tactical master surgery scheduling European Journal of Operational Research 270 2 2018 734 746 10.1016/j.ejor.2018.04.007
Lee S. Yih Y. Reducing patient-flow delays in surgical suites through determining start-times of surgical cases European Journal of Operational Research 238 2 2014 620 629 10.1016/j.ejor.2014.03.043
Li F. Gupta D. Potthoff S. Improving operating room schedules Health Care Management Science 19 3 2016 261 278 10.1007/s10729-015-9318-2 25687390
Miao H. Wang J.J. Scheduling elective and emergency surgeries at shared operating rooms with emergency uncertainty and waiting time limit Computers and Industrial Engineering 160 June 2021 107551 10.1016/j.cie.2021.107551
Min D. Yih Y. Scheduling elective surgery under uncertainty and downstream capacity constraints European Journal of Operational Research 206 3 2010 642 652 10.1016/j.ejor.2010.03.014
Moreno M.S. Blanco A.M. A fuzzy programming approach for the multi-objective patient appointment scheduling problem under uncertainty in a large hospital Computers and Industrial Engineering 123 2018 33 41 10.1016/j.cie.2018.06.013
Mulvey J.M. Vanderbei R.J. Zenios S.A. Robust optimization of large-scale systems Operations Research 43 2 1995 264 281 10.1287/opre.43.2.264
Nderitu P. Ursell P. Factors affecting cataract surgery operating time among trainees and consultants Journal of Cataract & Refractive Surgery 45 6 2019 816 822 10.1016/j.jcrs.2019.01.002 30879720
Neyshabouri S. Berg B.P. Two-stage robust optimization approach to elective surgery and downstream capacity planning European Journal of Operational Research 260 1 2017 21 40 10.1016/j.ejor.2016.11.043
Nguyen T.C. Thourani V.H. Nissen A.P. Habib R.H. Dearani J.A. Ropski A. …Badhwar V. The effect of covid-19 on adult cardiac surgery in the united states in 717 103 patients The Annals of Thoracic Surgery 113 3 2022 738 746 10.1016/j.athoracsur.2021.07.015 34343473
Niu Q. Jiao B. Gu X. Particle swarm optimization combined with genetic operators for job shop scheduling problem with fuzzy processing time Applied Mathematics and Computation 205 1 2008 148 158 10.1016/j.amc.2008.05.086
Norris Z.A. Sissman E. O’Connell B.K. Mottole N.A. Patel H. Balouch E. …Fischer C.R. COVID-19 pandemic and elective spinal surgery cancelations – what happens to the patients? The Spine Journal 21 12 2021 2003 2009 10.1016/j.spinee.2021.07.019 34339887
Roshanaei V. Luong C. Aleman D.M. Urbach D.R. Collaborative operating room planning and scheduling INFORMS Journal on Computing 29 3 2017 558 580 10.1287/ijoc.2017.0745
Schiele J. Koperna T. Brunner J.O. Predicting intensive care unit bed occupancy for integrated operating room scheduling via neural networks Naval Research Logistics 68 1 2021 65 88 10.1002/nav.21929
Shehadeh, K. S. (2022). Data-driven distributionally robust surgery planning in flexible operating rooms over a Wasserstein ambiguity. Computers and Operations Research, 146(August 2021), 105927. https://doi.org/10.1016/j.cor.2022.105927
Singh M. Rathi R. Antony J. Garza-Reyes J.A. Lean six sigma project selection in a manufacturing environment using hybrid methodology based on intuitionistic fuzzy MADM approach IEEE Transactions on Engineering Management 2022 10.1109/TEM.2021.3049877
Storn R. Price K. Differential evolution - a simple and efficient heuristic for global optimization over continuous spaces Journal of Global Optimization 11 1997 341 359 10.1023/A:1008202821328
Tsafarakis S. Zervoudakis K. Andronikidis A. Altsitsiadis E. Fuzzy self-tuning differential evolution for optimal product line design European Journal of Operational Research 287 3 2020 1161 1169 10.1016/j.ejor.2020.05.018
Wang Y. Zhang Y. Tang J. A distributionally robust optimization approach for surgery block allocation European Journal of Operational Research 273 2 2019 740 753 10.1016/j.ejor.2018.08.037
Wang K. Qin H. Huang Y. Luo M.W. Zhou L. Surgery scheduling in outpatient procedure centre with re-entrant patient flow and fuzzy service times Omega 102 2021 102350 10.1016/j.omega.2020.102350
Wang K. Luo H. Liu F. Yue X. Permutation Flow Shop Scheduling With Batch Delivery to Multiple Customers in Supply Chains IEEE Transactions on Systems, Man, and Cybernetics: Systems 48 10 2018 1 12
Yao J. Lin F. Constructing a fuzzy flow-shop sequencing model based on statistical data Internat. J. Approx. Reason 29 2002 215 234
Zhang J. Dridi M. El Moudni A. A two-level optimization model for elective surgery scheduling with downstream capacity constraints European Journal of Operational Research 276 2 2019 602 613 10.1016/j.ejor.2019.01.036
Zhang J. Dridi M. Moudni A.E. Column-generation-based heuristic approaches to stochastic surgery scheduling with downstream capacity constraints International Journal of Production Economics 229 2020 107764 10.1016/j.ijpe.2020.107764
Zhou L. Geng N. Jiang Z. Wang X. Multi-objective capacity allocation of hospital wards combining revenue and equity Omega 81 2018 220 233 10.1016/j.omega.2017.11.005
| 0 | PMC9742073 | NO-CC CODE | 2022-12-14 23:29:59 | no | Comput Ind Eng. 2022 Dec 12;:108893 | utf-8 | Comput Ind Eng | 2,022 | 10.1016/j.cie.2022.108893 | oa_other |
==== Front
101262796
32819
J Expo Sci Environ Epidemiol
J Expo Sci Environ Epidemiol
Journal of exposure science & environmental epidemiology
1559-0631
1559-064X
35710593
10.1038/s41370-022-00451-8
nihpa1812226
Article
Environmental Mixtures and Breast Cancer: Identifying Co-Exposure Patterns between Understudied vs Breast Cancer-Associated Chemicals using Chemical Inventory Informatics
Koval Lauren E. 12
Dionisio Kathie L. 3
Friedman Katie Paul 4
Isaacs Kristin K. 4
Rager Julia E. 125*
1 Department of Environmental Sciences and Engineering, Gillings School of Global Public Health, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA
2 The Institute for Environmental Health Solutions, Gillings School of Global Public Health, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA
3 Immediate Office of the Assistant Administrator, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC
4 Center for Computational Toxicology and Exposure, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC
5 Curriculum in Toxicology and Environmental Medicine, School of Medicine, University of North Carolina, Chapel Hill, North Carolina, USA
Author Contributions
LEK, KLD, KPM, KKI, and JER were responsible for the overall research design.
LEK and JER were responsible for data analyses.
LEK, KLD, KPM, and KKI were responsible for data extraction and organization from utilized databases.
LEK and JER were responsible for manuscript drafting.
All study coauthors reviewed manuscript text and provided scientific feedback.
* Corresponding author: Julia E. Rager, The University of North Carolina at Chapel Hill, 135 Dauer Drive, Chapel Hill, NC, 27599. phone: 919-966-4410; [email protected]
3 6 2022
11 2022
16 6 2022
16 12 2022
32 6 794807
http://www.nature.com/authors/editorial_policies/license.html#terms Users may view, print, copy, and download text and data-mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use:http://www.nature.com/authors/editorial_policies/license.html#terms
Background:
Although evidence linking environmental chemicals to breast cancer is growing, mixtures-based exposure evaluations are lacking.
Objective:
This study aimed to identify environmental chemicals in use inventories that co-occur and share properties with chemicals that have association with breast cancer, highlighting exposure combinations that may alter disease risk.
Methods:
The occurrence of chemicals within chemical use categories was characterized using the Chemical and Products Database. Co-exposure patterns were evaluated for chemicals that have an association with breast cancer (BC), no known association (NBC), and understudied chemicals (UC) identified through query of the Silent Spring Institute’s Mammary Carcinogens Review Database and the U.S. Environmental Protection Agency’s Toxicity Reference Database. UCs were ranked based on structure and physicochemical similarities and co-occurrence patterns with BCs within environmentally relevant exposure sources.
Results:
A total of 6,793 chemicals had data available for exposure source occurrence analyses. 50 top-ranking UCs spanning five clusters of co-occurring chemicals were prioritized, based on shared properties with co-occuring BCs, including chemicals used in food production and consumer/personal care products, as well as potential endocrine system modulators.
Significance:
Results highlight important co-exposure conditions that are likely prevalent within our everyday environments that warrant further evaluation for possible breast cancer risk.
Cancer
Co-exposures
Environmental Chemicals
Mixtures
Informatics
ExpoCast
==== Body
pmcINTRODUCTION
Globally, breast cancer is the most commonly diagnosed cancer and the leading cause of cancer-related death in women [1]. Genetic risk factors are estimated to contribute to only 5–10% of breast cancer cases [2,3], leaving a substantial portion of cases attributable to other risk factors, including environmental exposures. The link between environmental chemicals and breast cancer has specifically been identified as a research priority by multiple organizations worldwide, including The Institute of Medicine of the National Academies and the World Cancer Research Fund [4,5]. Determining exposures in the environment that can impact breast cancer will build an evidence base needed to better identify sources of exposure that should be reduced and/or eliminated, with the goal of reducing the global burden of environmentally influenced disease.
Environmental chemical exposures have been previously related to breast cancer. Because breast cancer etiology is highly intertwined with reproductive status, including serum hormone levels of estrogen and progesterone, chemicals that modulate signaling relevant to estrogen/progesterone levels have been linked to this disease outcome [6–8]. These chemicals include endocrine modulating chemicals, such as bisphenol A, parabens, phthalates, and polybrominated diphenyl ethers [6–8]. Other environmentally relevant chemicals that have been linked to increased risk of breast cancer include air pollutants (e.g., polycyclic aromatic hydrocarbons [PAHs]), dioxins, metals (e.g., cadmium and lead), industrial chemicals (e.g., benzene, ethylene oxide, and 1,3-butadiene), perfluoroalkyl substances (e.g., perfluorooctanoic acid [PFOA] and perfluorooctane sulfonate [PFOS]), and pesticides [6–8]. These chemicals have been largely evaluated based on individual exposure conditions due to a lack of information regarding co-occurrence of environmental chemicals, i.e., mixtures, with breast cancer incidence. This is a critical research gap, as humans are commonly exposed to multiple potentially harmful chemicals at a time [9], and thus potential joint toxicities resulting from co-occurring chemicals remain understudied in relation to breast cancer incidence. At the same time, the design of such studies remains difficult, as data are lacking that describe which chemicals commonly occur in our everyday environment as mixtures that may have associations with breast cancer incidence.
Humans are exposed to chemicals that originate from a variety of sources in their everyday environments. Sources include industrial, agricultural, and consumer uses of chemicals. Human exposure can occur via contact with a chemical source directly (e.g., a consumer product) or via contact with contaminated environmental media (e.g., air, soil, food, house dust). Characterization of the uses associated with the thousands of chemicals in commerce is needed to identify and prioritize chemicals and chemical combinations according to their exposure potential and ultimate human health impacts. To address this research goal, the curation of use information in chemical inventories has expanded in recent years, providing the foundation for much needed evaluation of understudied chemicals in the environment. For example, the U.S. Environmental Protection Agency’s (U.S. EPA’s) Chemical and Product Categories database (CPCat) was organized to capture data on consumer exposure pathways and patterns of chemical use in the environment [10]. This database has since been expanded and combined with additional product and chemical use data to form the Chemicals and Products Database (CPDat) [11]. These databases serve as an important foundation for a better understanding of the landscape of human exposures, and ongoing curation efforts will further increase their utility.
As described within this study, we have recently expanded data within CPDat to include additional source documents and improved descriptions of chemical use terms. We leveraged this unique resource to identify understudied chemicals in the environment, with respect to their association with breast cancer risk, that co-occur alongside chemicals shown to have an association with breast cancer. These exposure combinations could modify breast cancer risk. Understudied chemicals were then evaluated for physicochemical and structural similarities to chemicals associated with breast cancer. Resulting chemicals that likely co-occurred with, and showed similar chemical properties to, breast cancer carcinogens were identified as priority contaminants for further study to inform possible breast cancer risk, individually, and as mixtures.
METHODS
Exposure Source Categories
This study relied on the general chemical use data contained within CPDat [11], which collates and curates data obtained from federal and state reports, academic journal articles, and publications from international government agencies. Resulting chemical records span worldwide chemical use and product information inventories, chemical safety guideline sheets, food inventories, pesticide use information, and water and soil contamination data. To provide examples of such inventories, these include (listed respectively): the Danish Environmental Protection Agency’s Surveys on Chemicals in Consumer Products, Washington State’s Children’s Safe Products Act, USDA Annual Reports on Agricultural Chemical Usage for Field Crops and Fruit, State of Arizona’s Reported Pesticide Use Within Arizona, and the Minnesota Department of Agriculture’s Annual Water Quality Monitoring Report, among many others. CPDat aggregates chemical use information from hundreds of such documents into one organized dataset, harmonized to general descriptors that impart information on how the chemical is used, according to the original source. Here, we updated the CPDat descriptors to better align them with newly developed consumer product [12] and functional use categories [13]. These newly developed “CPDat Chemical List Presence keywords” are provided publicly in current versions of CPDat (EPA 2020a) and are further described in Supplemental Methods. A key feature of this updated dataset is the additional harmonization of the various chemical identifiers to Distributed Structure-Searchable Toxicity Database Substance Identifiers (DTXSID) to remove redundancies in chemical annotations, yielding unique substance identifiers [14,15].
The current study used Version 2 of the CPDat Chemical List Presence dataset [16]. These data included 73,573 chemical records derived from 1,543 chemical use reports (covering the years 1987–2019), reflecting 20,530 unique DTXSIDs and 129 unique use keywords, with multiple keywords (reflecting multiple uses) often assigned to a single chemical record. For the purposes of the current evaluation, related keywords were grouped into 32 broader “exposure source categories.” These exposure source categories served as higher-level descriptors of chemical use information designed to aid in user interpretation while allowing more effective chemical use clustering in proceeding analyses (multiple categories could still result for a single chemical). Select keywords or terms that did not impart information descriptive of an existing specific exposure source (e.g., “nondetect”, “prohibited”, “restricted”) were excluded from this analysis.
Association with Breast Cancer
This study focused on chemicals and their co-occurrence patterns in relation to breast cancer. Chemicals were grouped into categories based upon query of two databases: Silent Spring Institute’s Mammary Carcinogens Review Database [17,18] and U.S. EPA’s Toxicological Reference Database (ToxRefDB) [19]. Data from both human and animal studies were leveraged here, as there is evidence to support genetic similarities between rodent mammary tumors and breast carcinogenesis pathways in humans [20,21]. The Mammary Carcinogens Review Database includes information on 216 chemicals and breast cancer-relevant findings aggregated across findings from the International Agency for Research on Cancer Monographs, Carcinogenic Potency Database, U.S. National Toxicology Program (NTP), NTP 11th Report on Carcinogens, and the Chemical Carcinogenesis Research Information System Database. Thus, this database draws findings from both human and animal studies. The ToxRefDB represents one of the largest publicly available databases of curated in vivo study results, consisting largely of data from animal studies performed in accordance with or similar to U.S. Environmental Protection Agency Health Effects Series Guideline Studies from pharmaceutical, agrochemical, and other industrial chemicals [19]. This database was used to identify chemicals for which repeat dose studies in adults evaluated potential mammary gland changes in phenotypes.
Chemicals were organized into the following three categories: (1) Breast cancer chemicals (BCs) associated with breast cancer in humans and/or mammary gland cancer in animals; (2) Non-breast cancer chemicals (NBCs) that have been tested but not found to cause mammary gland carcinogenicity in animals and currently have no known association with breast cancer; (3) Understudied chemicals (UCs) that remain understudied in relation to human breast cancer and/or mammary gland phenotype changes in animals.
To identify BCs, both the Mammary Carcinogens Review Database and ToxRefDB were queried. All 216 chemicals currently included in the Mammary Carcinogens Review Database were included as BCs, determined by Silent Spring Institute as chemicals that met at least one of the following criteria: (i) reported through IARC Monograph summaries to increase mammary gland tumors; (ii) included in the Carcinogenic Potency Database with at least one study that reported an increase in mammary gland tumors; (iii) reported on the NTP website as “chemicals associated with site-specific tumor induction in mammary gland,” (drawn from the collection of NTP Technical Reports) as well as chemicals that increased mammary tumors in the NTP Study Reports Collection: Abstracts and Target sites in two-year studies; (iv) reported in the NTP 11th Report on Carcinogens as associated with increased mammary tumors; and (v) reported through the Chemical Carcinogenesis Research Information System Database as having positive results in the “Carcinogenicity Studies” section after filtering for “mammary” [17]. All chemicals were linked to DTXSIDs and corresponding chemical names through a batch search (using CASRN) on the CompTox Chemicals Dashboard [22]. ToxRefDB was queried as an additional database to identify BCs, selecting chemicals that had recorded instances of a treatment-related mammary gland change that was cancer-related in adult animals exposed in subchronic or chronic studies [19]. These effects were either micro- or macroscopic pathology, and included findings described as fibroadenoma or fibroma, adenocarcinoma or adenoma, carcinoma, and mixed tumor (malignant or not otherwise specified).
To identify NBCs, ToxRefDB was queried for chemicals that were tested for mammary gland changes in pathology, focusing on mouse and rat species to parallel the ToxRefDB BC query results. Records were confined to chronic two-year animal bioassays, to focus on exposure designs that included sufficient time for cancer endpoints to develop. Chemicals that were included in this list, but were not identified as BCs, were then defined as the list of NBCs for the purposes of the current investigation. As the Mammary Carcinogens Review Database lacked parallel “negative” data, the identification of NBCs was based solely on animal data contained in ToxRefDB; though the final NBC list notably contained chemicals that were not present in the Mammary Carcinogens Review Database, supporting their lack of currently known association with breast cancer.
To identify UCs, chemicals within the CPDat Chemical List Presence dataset that were not identified as BCs or NBCs were then designated as the UCs. These UCs represent chemicals that have yet to be tested in relation to breast cancer / changes in mammary gland phenotypes, or generally lacked information surrounding their potential carcinogenicity, that also have exposure source information relevant to human environmental exposure conditions.
Chemical Exposure Patterns
The chemicals represented in CPDat were hierarchically clustered based on their associated exposure source category linkages to characterize potential chemical co-occurrence patterns. We have demonstrated success with this approach in identifying trends in both chemical use and molecular response signatures [23–26] and have used it here to identify co-occurrence between BCs and UCs that could potentially alter disease risk.
CPDat data were first filtered for unique combinations of chemicals and exposure source categories. Additional filtering to include only those chemicals present in at least two categories to allow analysis of potential environmental co-occurrence patterns resulted in 6,793 unique chemicals. A summary table of the resulting chemicals and exposure source categories was produced, containing values of 1, indicating at least one association between the chemical and the exposure category, and values of 0, indicating no association. A distance matrix was derived from this summary table using the Jaccard distance, the complement of the Jaccard similarity, through the ‘vegan’ package (v2.5–7) in R (v4.0.3). The Jaccard similarity for two sets is defined as the size of intersection divided by the size of the union of the sets, thus the Jaccard distance can be calculated as follows (Eqn. 1): Eqn 1. D(A,B)=1−|A∩B||A∪B|=1−Number of exposure source categories in common between chemicals A and B Total number of exposure source categories associated with chemicals A and/or B
For this equation, D(A,B) is the Jaccard distance for sets A and B [27], here, representing a pair of chemicals (chemical A and chemical B). For the purposes of this evaluation, the magnitudes of the intersection and union reflect the number of associated exposure source categories common between chemicals and the total number of exposure source categories associated with either of the chemicals, respectively. This metric was used here to gauge the similarities between exposure source categories associated with each possible pair of chemicals. The resulting distance ranged from 0 (low dissimilarity [i.e., high similarity]) to 1 (high dissimilarity [i.e., low similarity]). To determine the optimal number of chemical clusters, the within cluster sum of squares and average silhouette width [28] (both measures of within-cluster compactness) were calculated and visualized for 1≤ k ≤100 using the fviz_nbclust function in the R package ‘factoextra’ (v1.0.7). The minimum number of clusters that produced a reasonable cluster compactness, while allowing for interpretable clusters, was selected as optimal. These clustering methods were selected due to their recognized utility in the field of exposure science [29–31] and efficiency in terms of computing time which was required for our large dataset.The data were grouped into the selected optimal number of clusters through hierarchical clustering using the diana function in the R package ‘cluster’ (v2.1.1). The same methods were also employed to derive clusters of similar exposure source categories, with the goal of enhancing interpretability of chemical use patterns. A heatmap was produced to visualize the resulting chemical clusters through the pheatmap function in the R package ‘pheatmap’ (v1.0.12). The heatmap was color-coded such that a different color indicated an association between an exposure source category and BCs, NBCs, or UCs, allowing for visual results interpretation.
Identifying Structural Features that are Enriched in BCs vs NBCs
Chemical structural features are known to influence (i) whether or not a chemical elicits toxicity and ultimately causes disease; and (ii) how a chemical elicits toxicity (i.e., the underlying etiology of a chemical-induced disease outcome). Structural features are thus the primary data used to inform quantitative structure-activity relationship (QSAR) / read-across toxicity predictions, which are entirely based upon in silico approaches [32]. Here, structural feature data were first evaluated comparing BCs to NBCs to identify which structural attributes are more commonly abundant. Then, this information was used to identify UCs that contain these breast cancer-associated features and should be prioritized for further evaluation.
Structural feature descriptors (i.e., atom, bond, chain, and ring types) were acquired as ToxPrint fingerprint data, which contain information on whether a specific structural feature was present (1) or absent (0) [33], downloaded from the CompTox Chemicals Dashboard. This batch search yielded ToxPrint fingerprint data with 729 available chemical structural features. Features specifically enriched in BCs were identified using a method previously published [34] to test whether a specific structural attribute was present in BCs at a rate higher than would occur by chance, in comparison to NBCs. The significance of this association was indicated by a one-sided Fisher’s exact p-value ≤ 0.05 and an odds ratio ≥ 3. There was also a requirement for at least three BCs to contain the feature. The converse was also evaluated, whereby absence of structural features was analyzed for enrichment in BCs vs NBCs, using parallel filters. Calculations were carried out in R using the ‘tidyverse’ (v1.3.0) and ‘janitor’ (v2.1.0) packages.
Assessing Physicochemical Property Similarity across Chemicals
Physicochemical properties were evaluated here to further inform which chemicals may induce toxicity similar to chemicals known to cause breast cancer, based on similar etiologies, as we have previously published evidence supporting the utility of physicochemical properties in computational-based predictions of in vivo toxicology [24,35]. Furthermore, physicochemical properties play a critical role in chemical fate and transport within the environment, and thus further inform occurrence patterns of chemicals across environmental exposure sources [36]. Here, physicochemical property data were evaluated amongst BCs, and then compared to UCs, to identify pairs of co-occurring BCs and UCs in the environment that display similar physicochemical properties. These chemicals may impart similar toxicity and environmental fate/transport patterns and should likely be prioritized for further evaluation.
Physicochemical property data for all evaluated chemicals were obtained from the CompTox Chemicals Dashboard [22]. A batch search was performed for the DTXSIDs of interest, and OPEn structure-activity Relationship App (OPERA) predictions for physicochemical properties and environmental fate endpoints were downloaded [37]. These included atmospheric hydroxylation rate, bioconcentration factor, biodegradability half-life, boiling point, fish biotransformation half-life, Henry’s Law constant, melting point, octanol/air partition coefficient, octanol/water partition coefficient, soil adsorption coefficient, vapor pressure, and water solubility. Organized data were then z-score normalized by property within each cluster. The degree of physicochemical property similarity between pairs of chemicals (i.e., UC-BC pairs) based on scaled properties was evaluated using the Spearman Rank Correlation test (cor.test ‘stats’ [v4.0.3]). The resulting correlation coefficients (R) and p-value distributions were evaluated across all possible UC-BC chemical pairs per cluster, and the highest correlation result for each UC from all BC pairings was selected to inform the final chemical prioritization.
Prioritizing Chemical Mixtures for Breast Cancer Evaluations
Information on chemical exposure sources, structural similarities to BCs, and physicochemical similarities to BCs were integrated to identify UCs that co-occur alongside chemicals associated with elevated breast cancer risk, and from a mixtures-based exposure and toxicity standpoint, should be prioritized for further evaluation. First, a series of filters were applied at the cluster-level, to prioritize clusters that included chemicals most likely to be prevalent in everyday environments and that showed some of the highest structural and physicochemical property similarities to chemicals associated with breast cancer. This set of filters specifically included the following criteria: (1) Clusters were required to include at least one UC that had high structural similarity to BCs, defined as containing at least four structural features that were statistically enriched within BCs. (2) Clusters were required to include at least one UC that was physicochemically similar to co-occurring BCs, defined as having physicochemical properties features that were correlated at R ≥ 0.8 to a BC in the same cluster. Structural and physicochemical property similarity cut-offs were selected based review of the resulting distribution of enriched features/correlations, as described in the results. (3) Clusters were required to include chemicals that map to exposure source categories that were relevant to multiple common environmental exposures, such as those pertaining to personal care, the household environment, food, and water. These filters resulted in prioritized clusters of co-occurring chemicals in the environment that were selected for further examination in this analysis.
A ranking scheme was then applied at the individual chemical level (Figure 1), focusing on chemicals within each of the prioritized clusters. This ranking scheme was based on a score that combined information on chemical structural similarity and chemical physicochemical property similarity to BCs. This score parallels other algorithms used in chemical prioritization efforts [38–40]. For the structural similarity component, each UC in the cluster was given a structural similarity score (SS) based on enriched structural features (EFs) calculated as follows (Eqn. 2): Eqn 2. SSuc=EFuc−EFcluster, minEFcluster, max−EFcluster, min
Here, EFuc reflects the number of enriched structural features present in the UC for which a score is being calculated. EFcluster,min and EFcluster,max are, respectively, the minimum and maximum number of enriched features identified in any UC within the cluster under evaluation. Therefore, if the number of enriched features present in the UC under consideration equaled the minimum for the entire cluster, the SS for the UC would be 0, indicating low concern (i.e., low priority). Conversley, if the number of enriched features in the UC under consideration equaled the maximum for the cluster, the SS for the UC would be 1, indicating an elevated concern (i.e., high priority). A similar physicochemical similarity (PS) score was calculated for each UC in a cluster based on the highest property correlation value of the UC with a BC. Since a correlation value was calculated for each UC-BC pair in a cluster, there were often multiple correlation values for each UC. As the most conservative approach, we selected the highest correlation value (R) for each UC, and these values informed the PS scoring for the cluster, calculated as follows (Eqn. 3): Eqn 3. PSuc=Ruc−Rcluster, minRcluster,max−Rcluster,min
Here, Ruc reflects the highest correlation value for the UC for which a score is being calculated. Rcluster,min and Rcluster,max are, respectively, the minimum and maximum of the highest selected correlation values of any UC within the cluster under evaluation. Similarly to the SS, the PS could range from 0–1, with 0 indicating the lowest concern (i.e., low priority) and 1 indicating the highest concern (i.e., high priority) based on degree of physicochemical correlation to a BC. The structural and physicochemical similarity scores were then summed to provide an overall score (OS) for each UC within each previously identified cluster of interest. The overall scores were then used to inform the final UC ranking in each of the clusters of interest. Top-ranking UCs were also reported alongside their co-occurring BCs to provide examples of high priority mixtures likely occurring in our everyday environment that require further evaluation for putative relationships to breast cancer risk. All calculations were carried out in R using base statistical packages, ‘tidyverse’ (v1.3.0), and ‘janitor’ packages (v2.1.0).
RESULTS
Exposure Source Categories for Describing Human Chemical Use Patterns
Patterns of human exposure to chemicals in the environment were evaluated using chemical use inventory information organized within CPDat. The updated CPDat Chemical List Presence dataset contained 140 unique keywords; these were mapped to 32 unique exposure source categories (Figure 2; and Table S1 available at [41]). After the filtering described in the Methods, a final list of 6,793 chemicals was carried forward in the current analysis (Table S2 available at [41]). The final exposure source categories notably captured those that are relevant to environmental exposure sources that humans experience in their everyday environment, including sources from arts and crafts / office supplies, building materials, children’s products and toys, cleaning products, electronics, furniture, general consumer products, household care and cleaning products, personal care, and other common sources of exposure.
Categorizing Chemicals based on Association to Breast Cancer
Chemicals with associated chemical use information were binned into categories of BCs, NBCs, and UCs to describe their current known (or unknown) association to breast cancer risk. The Mammary Carcinogens Review Database contained 216 chemicals, 208 of which had a CASRN and 199 mapped to DTXSIDs, which were carried forward in the analysis (Table S3 available at [41]). Within ToxRefDB, a total of 53 unique chemicals were identified to show instances of causing mammary gland cancer-related effects, spanning results from mouse and rat chronic and subchronic bioassays (Table S4 available at [41]). In total, 228 unique chemicals were identified between the two sources as being associated with breast cancer, including human breast cancer and/or animal cancer-related mammary gland changes. These chemicals thus represented the full list of identified BCs (Table S5 available at [41]). Also within ToxRefDB, 535 unique chemicals were identified to show instances of being tested, in general, for any mammary gland cancer-related effects from mouse and rat chronic bioassays (Table S6 available at [41]). Of these 535 unique chemicals, 53 had been identified as having an association with mammary gland cancer-related effects, and therefore were already classified as BCs. As a result, 482 remaining chemicals were identified as NBCs (Table S5 available at [41]), representing chemicals that currently lack a known association with breast cancer. Mapping these chemicals to those with the required chemical use information in CPDat resulted in the following counts of chemicals in each category: 78 BCs, 409 NBCs, and 6,306 UCs (Table S5 available at [41]).
Groups of Co-Occurring Chemicals based on Chemical Use Patterns
Chemicals were evaluated for common exposure source patterns through clustering algorithms based on a Jaccard distance metric. Selection of the number of clusters (k) was determined by examination of the reduction in the proportion of within cluster variance (or sum of squares) compared to total variance and average silhouette of the resulting distance matrix (Supplemental Figure S1A). Based on the results, an optimal k = 19 clusters was selected to minimize compactness (variability) within the clusters, while grouping data into interpretable cluster assignments. Exposure source categories were similarly grouped into k=12 clusters (Supplemental Figure S1B). These resulting clusters inform the identification of 19 groups of chemicals that likely co-occur within the environment based on chemical use patterns (Figure 3). These chemical clusters include a variety of different total chemical counts, ranging between 5 and 3,175 (mean = 358). Chemical clusters also contain a wide range of BC/NBC/UC distributions, with some clusters containing 0 BCs (e.g., cluster 14) and others containing up to 15% BCs (e.g., cluster 15). Other clusters represent largely untested chemicals with up to 99% UCs. Specific clusters of interest are further detailed below.
Structural Features Enriched in BCs
Chemical structural feature data were compared between BCs vs NBCs with the goal of identifying feature attributes that are significantly enriched within chemicals associated with carcinogenic changes in mammary glands in animals and/or breast cancer in humans. Here, ToxPrint data were organized across 78 BCs and 393 NBCs with available feature data (Table S7 available at [41]). 390 features were present in at least one BC or NBC (614 with UCs were included). Enrichment analyses identified 26 structural features that were more commonly present in BCs vs NBCs (Table 1), highlighting chemotypes that could be evaluated further in future studies for mechanistic involvement in potential cancer etiology. Enrichment analyses also identified 14 features that were more commonly absent in BCs vs NBCs (Table S8 available at [41]). Note that this does not provide evidence that the 14 features are more commonly present in NBCs. These features were not considered in further characterization steps or used to prioritize UCs. The maximum number of enriched features within an individual UC was 8, out of the possible 26 features (Figure S2). Also, the overall distribution of feature presence was right-skewed, where many UCs had either 0 or 1 enriched features. This information was then carried forward in the chemical ranking step, in which UCs were evaluated for the presence of the 26 structural features associated with BCs, as detailed in the chemical prioritization results.
Physicochemical Property Similarity Results
Physicochemical property data were used to compare property similarities between individual chemical pairings within clusters of chemicals likely co-occurring due to chemical use patterns. A total of 4,251 chemicals (78 BCs, 380 NBCs, and 3,793 UCs) had physicochemical data available (Table S9 available at [41]). Correlating physicochemical properties between pairs of chemicals identified some chemicals that were highly correlated, while others were not. To provide example data distributions of the correlation results, all correlation values across UC-BC pairs in each cluster were combined and visualized (Figure S3). These results show that there is a wide distribution of physicochemical property similarities between UC-BC pairs. These findings were then used to inform the chemical ranking, in which UCs with physicochemical properties that were highly correlated to BCs were ranked highly in the overall mixtures-based prioritization, as detailed in the next section.
Chemical Prioritization Results for Mixtures Evaluation in Breast Cancer Studies
UCs were characterized in this study to yield a list of top-ranking chemicals to test individually and in combination with co-occuring BCs in the context of mixtures-based cancer evaluations. Chemicals were first prioritized by cluster, according to groups of chemicals that likely co-occur in the environment due to similar chemical use patterns. Clusters of interest were specifically selected by first applying a set of quantitative filters, that required that the cluster include at least one UC that was highly structurally similar to BCs, defined as containing at least four structural features that were statistically enriched within BCs. Second, clusters were required to include at least one UC that was physicochemically similar to a co-occurring BC, as defined as having physicochemical features that were correlated at R ≥ 0.8 to a BC in the same cluster. These two filters were applied to prioritize UCs that could be involved in similar breast cancer initiation/propagation pathways and that may display similar fate and transport properties within the environment. This filtering strategy resulted in a list of ten chemical clusters of interest to further evaluate: clusters 1,2,3,4,5,6,7,8,9, and 11.
These ten chemical clusters were then investigated further, with a focus on those that included chemicals present in several exposure source categories relevant to the environment. This last prioritization filter yielded five top-ranking chemical clusters; namely, clusters 1, 4, 5, 6 and 9 (Figure 3). UCs within these clusters were then ranked based upon overall score (OS), representing a combination of physicochemical and structural similarities to chemicals that have an association with breast cancer. It is important to note that these scores were produced by comparing chemicals within clusters, and should thus be interpreted on a per-cluster basis, as opposed to a per-chemical basis across various clusters. 50 top-ranking UCs were summarized in Table 2 and Figure 4, with all results detailed in Table S10 (available at [41]). Notable data trends per-cluster are summarized below:
Cluster 1 results summary:
Cluster 1 notably contained the largest number of chemicals, including inert ingredients, pesticides, and chemicals labelled for nonfood use. Nonetheless, there are some instances of cluster 1 chemicals being detected in food or food contact substances. Notable top-ranking chemicals in this cluster included dyes, such as C.I. acid blue 9 and methyl blue, which co-occur alongside the known BC, benzyl violet 4B. Interestingly, benzyl violet 4B was used as a food and cosmetics additive in the U.S. until its delisting by the U.S. Federal Drug Administration in 1977 [42], though it may still be used in other countries. Regardless, the potential for co-occurrence of two potentially similar dyes, or one of these dyes with chemicals of similar properties within similar exposure sources remains a concern. The other primary co-occuring BC in this cluster was 2-amino-5-azotoluene.
Cluster 4 results summary:
Cluster 4 contained chemicals included in personal care products, children’s products and toys, and some other product-relevant categories, with three BCs involved in pharmaceuticals. Chemicals in this cluster included those that co-occurred alongside BCs 17β-estradiol (also known as estradiol), estrone, and progesterone, representing chemicals shown to have notable associations with breast cancer in humans [7,8,43]. Noteworthy UCs in this cluster included 2-nitro-5-glyceryl methylaniline, 4-(3-((9,10-Dihydro-4-hydroxy-9,10-dioxo-1-anthracenyl)amino)prop-yl)-4-methylmorpholinium methyl sulfate, acid green 22, and HC red 3. The observed exposure source categories in cluster 4 included dyes, including those used in hair products and cosmetics.
Cluster 5 results summary:
Cluster 5 contained chemicals with exposure sources related to pesticides, food, and drinking water. This cluster included top-ranking chemicals containing many structural features that were enriched amongst BCs. Specifically, all ten top-ranking chemicals contained six structural features that were statistically enriched within BCs, with eight being the highest number found in any UC as a point of comparison. Additionally, the range of correlations, 0.7–1, indicates high physicochemical similarity to BCs for nine of the ten top-ranking UCs. It is also worth noting co-occurrence with the seven unique BCs 1,2-dichloropropane, ametryn, etoxazole, oryzalin, prosulfuron, quercetin, and tribenuron-methyl, indicating that this cluster contained a relatively high number of chemicals associated with breast cancer. A few of the top-ranking chemicals in this cluster included 1,3-dichloropropene, ethametsulfuron, methoprotryne, secbumeton, and (Z)-dichloropropene. 1,3-dichloropropene is of particular interest as it has recently been identified by Cardona and Rudel as one of 35 pesticides of concern regarding effects to the mammary gland [44]. This work reviewed EPA pesticide Registration Eligibility Decisions and examined whether mammary tumors were considered in corresponding carcinogenicity classifications. Additionally, mechanistic data were evaluated for biological activity relevant to in vivo outcomes. BCs ametryn and oryzalin were also included amongst these pesticides.
Cluster 6 results summary:
Cluster 6 showed patterns similar to cluster 5 in that it contained chemicals with exposure sources related to pesticides, food, and drinking water; in addition it contained the exposure sources of groundwater and surface water. Top-ranking UCs in cluster 6, which included cyanazine, deethylatrazine, and deisopropylatrazine, displayed co-occurrence with the seven BCs. These BCs included 1,2-dibromo-3-chloropropane, azoxystrobin, and triasulfuron. Additional BCs atrazine, dichlorvos, diuron, and simazine were notably among the 35 pesticides of concern identified by Cardona and Rudel [44]. The number of enriched structural features of top-ranking chemicals in the cluster ranged from two to seven and the physicochemical correlations ranged from 0.789–0.992.
Cluster 9 results summary:
Cluster 9 contained chemicals involved in a few different exposure source categories, particularly those relevant to consumer products and various interior / household sources of exposure through product use categories typically used in the indoor environment. This cluster contained chemicals that co-occurred with the BC, isoeugenol. An example chemical to highlight is 2-methoxy-4-vinylphenol, which like other top-ranking chemicals in the cluster, may be used as a flavoring or fragrance additive.
DISCUSSION
This study aimed to characterize combinations of chemicals that likely occur in our everyday environment which may be impacting risk of acquiring breast cancer. We used informatics-based approaches to identify novel understudied chemicals that may co-occur with chemicals associated with cancer risk. Clustering-based analyses of the 6,793 chemicals with use information (78 BCs; 409 NBCs; 6,306 UCs) yielded 19 clusters of chemicals that represent likely patterns of co-exposure conditions that humans may experience. These results are of high translational relevance, as humans are commonly exposed to multiple chemicals and other stressors in their everyday environments [9]. Our finding that understudied chemicals in the environment share chemical property similarities with cancer-associated chemicals is novel, and particularly impactful when chemical combinations are prioritized based on likely co-occurrence in everyday exposure scenarios. We further identified 50 top-ranking chemicals in five clusters of environmental relevance, ranked based on co-occurrence exposure patterns and predicted carcinogenicity from structural and physicochemical property similarities to BCs. Findings from this study yielded a novel list of understudied chemicals that warrant further testing through toxicological and human biomonitoring studies, individually and in combination with co-occuring BCs in the context of environmental mixtures.
One of the prioritized chemical clusters (cluster 4) contained chemicals present in personal care products and children’s products and toys co-occurring with the compounds, 17β-estradiol (also known as estradiol), estrone, and progesterone. Estradiol, estrone, and progesterone are major steroid hormones largely produced by ovaries in females that regulate female reproductive cycling, as well as a variety of cell processes and functions [43,45,46]. Alterations in the levels of these hormones and associated changes in the estrogen and progesterone receptor pathways have known implications in breast cancer etiology, often serving as targets in therapeutic intervention [46,47]. The observed exposure patterns in cluster 4 were interesting for two distinct reasons. First, these chemicals largely encompass dyes, including dyes used in hair products and cosmetics (e.g., 4-(3-((9,10-dihydro-4-hydroxy-9,10-dioxo-1-anthracenyl)amino)prop-yl)-4-methylmorpholinium methyl sulfate, acid green 22, and HC red 3). The potential co-occurrence of these chemicals in personal care products of this nature is concerning, as subsets of the population are likely experiencing combined exposures to these specific chemicals. Therefore, if these chemicals induce similar toxicological changes (e.g., changes in estrogen or progesterone receptor-related signaling), their combined exposures may alter associated disease risk. Second, the UCs in this cluster largely emphasized co-occurrence patterns with estradiol, estrone, and progesterone within the exposure source category of ‘personal care’. However, these data suggest that these UCs may also be present within exposure sources, particularly drinking water, surface water, and wastewater, that perhaps have yet to be evaluated for the presence of these chemicals. Future biomonitoring efforts could focus on the potential presence of these understudied chemicals in other exposure media.
Two separate priority chemical clusters (cluster 5 and cluster 6) included chemicals with exposure sources of pesticides, food, and water. The observed exposure patterns in clusters 5 and 6 were interesting for the following three reasons: First, there are a high proportion of BCs contained in each of these clusters, highlighting important chemicals that already show an association with breast cancer and co-occur across environmental exposure sources. In cluster 5 some of the chemicals that induced cancer-related changes in mammary tissue in animals and/or were associated with breast cancer in humans included ametryn, prosulfuron, and tribenuron-methyl. In cluster 6 these chemicals included atrazine, azoxystrobin, and simazine. These chemicals, in themselves, warrant further investigation, as potential contributors towards human disease based on cumulative exposure impacts. Second, the top-ranking UCs, within these clusters had many structural features that were statistically enriched in BCs and showed distinctly high physicochemical correlations with BCs in the cluster. Specific chemicals with high structural and physicochemical similarities included 1,3-dichloropropene, ethametsulfuron, and secbumeton (cluster 5) and cyanazine, deethylatrazine, and deisopropylatrazine (cluster 6). Third, these chemicals showed co-occurrence patterns largely involving pesticides, food, and water, indicating that humans may experience combined exposures to these chemicals via ingestion. This common exposure route may also exacerbate the effects of these chemicals on the human body, given that they may hit similar initial target tissues.
Exposure sources for the remaining two priority clusters (1 and 9) spanned pesticide and food related categories (cluster 1) and indoor environment and household categories (cluster 9). Cluster 9 contained chemicals that may co-occur via multiple exposure routes. However, more data are needed surrounding the presence of these chemicals in specific exposure scenarios and environmental media to further develop future hypotheses to test these chemicals. Chemicals in cluster 1 included dyes (C.I. acid blue 9 and methyl blue) and chemicals in cluster 9 included flavorants and fragrances (e.g., 2-methoxy-4-vinylphenol). Thus, such chemicals could be present in multiple types of products within household environments.
This study advances knowledge surrounding new exposure patterns relevant to environmental mixtures and their potential relevance to breast cancer, though it is notable that future research could further enhance this topic of investigation. The focus here was on identifying chemical co-occurrence based on use pattern alone, without the explicit additional consideration of fate and transport or exposure mechanism. Although some chemicals may co-occur with BCs in various exposure sources, their properties may render them more or less likely to result in an ultimate exposure. These considerations were addressed in a manner herein via the prioritization based on property similarity. In the future, studies could expand on these findings by leveraging additional exposure-relevant resources, such as biological or environmental monitoring data (e.g., National Health and Nutrition Examination Survey [NHANES] [48] or the National Water Information System [49], and exposure models that incorporate mechanistic information (e.g., the consumer and ambient models included in The Systematic Empirical Evaluation of Models [SEEM] framework) [50].
In this work, the Mammary Carcinogens Review Database and ToxRefDB were used to identify chemicals with associations to breast cancer. These databases consist of findings from both human and animal studies, with the broad assumption that rodent models of carcinogenesis in the mammary gland are indicative of human breast carcinogenesis. This assumption also contributed to the identification of potential NBCs, representing a list of chemicals needed to identify which chemical structures were enriched amongst BCs in comparison to NBCs. As breast cancer is a multi-etiological disease, identification of chemicals associated with increased rates of breast cancer is challenging due to lack of comprehensive datasets and current consensus supporting breast cancer classifications across the wide landscape of chemicals. As data on this topic continues to expand, these classifications will continue to improve. For instance, the final list of NBCs included chemicals without evidence of causing mammary gland tumors in animal models and chemicals that were not present in the Mammary Carcinogens Review Database. This list therefore represents chemicals that currently lack a known association with breast cancer (and thus have a lower prioritization in terms of breast cancer concern), though these associations may change over time as studies continue to develop. Additional health databases could be queried to inform the delineation of BCs, NBCs, and UCs, such as more exhaustive literature reviews and/or text mining approaches [51]. Other informatics approaches could be leveraged as these efforts continue to expand, including frequent itemset mining [52] and machine learning/predictive modeling approaches [35,53,54]. Additionally, leveraging the increasing chemical annotation in resources like PubMed could expand upon this study’s findings; for instance, normalized pointwise mutual information approaches could be expanded to find chemical:gene:disease associations in the published literature [55]. The use of animal model data as an indicator of putative breast cancer risk avoids imputing relationships from the published literature on chemical:gene and gene:disease associations, but as confidence and data increase in these chemical:gene and gene:disease associations, this could revolutionize how the risk of exposure to mixtures is evaluated.
In conclusion, this study set out to identify which understudied chemicals in our everyday environment co-occur alongside BCs according to chemical usage, and also show structural and physicochemical similarities to BCs. The resulting chemicals represent those that are of high interest in the designing of future epidemiological and toxicological investigations for understanding the effects of exposure to individual chemicals and mixtures. We specifically highlighted 50 top-ranking chemicals that remain understudied in relation to their putative breast cancer risk. These chemicals on their own may warrant further investigation, and when co-occuring with BCs, may represent high priority mixtures in the environment that have the potential to impact breast cancer risk. Though there is recent momentum surrounding the evaluation of chemical mixtures in the environment [56–58], it is imperative that this focus continues to expand across exposure and health science fields.
Supplementary Material
1
Acknowledgements
The research described in this manuscript has been reviewed by the Center for Computational Toxicology and Exposure, U.S. EPA, and approved for publication. Approval does not signify that contents necessarily reflect the views and policies of the agency, nor does the mention of trade names or commercial products constitute endorsement or recommendation for use. The authors would like to thank Drs. Peter Egeghy and Chris Corton for providing internal technical review of this manuscript.
Funding
This study was supported by the Institute for Environmental Health Solutions (IEHS) at the Gillings School of Global Public Health, RFA-18-01, ‘Identifying solutions that optimize the health of cancer survivors’, and through the National Institutes of Health (NIH) from the National Institute of Environmental Health Sciences, including grant funds (P42ES031007). Support was also provided by the Intramural Research Program of the Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina.
Data Availability
All data used for these analyses are publicly available, either through CPDat [16] ToxRefDB [19], or the CompTox Chemicals Dashboard [22]. Script associated with these analyses are publicly available through the Ragerlab Github repository [59]. Data that were combined and analyzed in generating results for this specific study are provided as supplemental material (Supplemental Tables S1–S10, provided online through the Ragerlab-Dataverse repository [41]).
Figure 1. Schematic for ranking of UCs based potential for influencing breast cancer risk.
This analysis resulted in the ranking of UCs within specific chemical clusters based on use co-occurrence, structural similarity, and physicochemical property similarity to BCs. Chemical data were analyzed separately, here on a per-cluster basis, for clusters 1, 4, 5, 6, and 9. These clusters were selected based on environmental relevancy as well as other requirements described within Methods (see section Prioritizing Chemical Mixtures for Breast Cancer Evaluations).
Figure 2. Translating chemical use inventory data to inform human exposure patterning.
Groups A-I illustrate the identified clusters of exposure source categories.
Figure 3. Clusters of chemicals arranged by human use patterns.
Each row reflects a chemical while each column reflects an exposure source category. An association between an exposure source category and a chemical is shown as yellow (UCs), blue (NBCs), or red (BCs). Grey indicates chemicals that were not present in a particular exposure source category. Chemical clusters 1, 4, 5, 6 and 9 were prioritized for further characterization.
Figure 4. The ten UCs with the highest Overall Scores in clusters 1, 4, 5, 6 and 9 along with the most similar BCs (based on physicochemical property correlations) in each respective cluster.
Table 1. ToxPrint chemotypes identified as being enriched in BCs.
Enrichment statistics are shown here, including the odds ratio and p-value for each structure in relation to its occurrence in chemicals associated with breast cancer. The number of true positives is also listed, indicating the number of BCs that contain each chemotype. Chemotypes reflect general structural fragments, which detail information surrounding each chemical’s atom, bond, chain, and ring types, as well as group-level information when available [60].
Chemotype Odds Ratio Fisher p-value True Positives
ring.fused_steroid_generic_.5_6_6_6. 50.51 5.42E-07 9
ring.fused_.6_6._tetralin 26.56 5.82E-04 5
chain.alkeneLinear_diene_1_3.butene 15.54 1.55E-02 3
bond.CX_halide_alkyl.X_ethyl 12.71 2.11E-04 7
bond.CX_halide_alkyl.Cl_ethyl 10.75 9.59E-04 6
bond.C.N_imine_C.connect_H_gt_0. 7.77 3.41E-02 3
ring. hetero_.3._Z_generic 7.77 3.41E-02 3
chain.aromaticAlkene_Ph.C2_acyclic_generic 7.6 1.05E-03 7
chain.alkeneLinear_mono.ene_ehtylene_terminal 5.47 1.68E-03 8
bond.CN_amine_sec.NH_aromatic_aliphatic 5.41 3.41E-03 7
bond.CX_halide_alkenyl.Cl_dichloro_.1_1.. 5.23 2.88E-02 4
chain.aromaticAlkene_Ph.C2 5.23 2.88E-02 4
chain.alkaneCyclic_pentyl_C5 4.51 2.15E-03 9
ring.hetero_.6._N_triazine_.1_3_5.. 4.36 4.29E-03 8
bond.CX_halide_alkenyl.X_dihalo_.1_1.. 4.18 4.54E-02 4
chain.alkaneCyclic_hexyl_C6 3.85 1.79E-03 11
bond.CN_amine_sec.NH_alkyl 3.76 1.26E-02 7
bond.CX_halide_alkenyl.X_acyclic_generic 3.54 2.45E-02 6
ring.hetero_.6._Z_1_3_5. 3.52 6.77E-03 9
bond.CX_halide_alkyl.X_ethyl_generic 3.41 1.79E-02 7
bond.CX_halide_alkyl.X_primary 3.41 1.79E-02 7
bond.CN_amine_aromatic_generic 3.24 6.15E-04 17
bond.CN_amine_pri.NH2_aromatic 3.18 3.41E-02 6
bond.CN_amine_sec.NH_aromatic 3.12 2.45E-02 7
chain.alkeneLinear_mono.ene_ethylene_generic 3.11 1.17E-03 16
bond.C.N_imine_N.connect_noZ. *INF 4.40E-03 3
* INF refers to infinite. This occurs when the odds ratio could not be calculated for a chemotype due the absence of the chemotype in all NBCs.
Table 2. Top 50 ranking UCs, spanning ten in each of the prioritized chemical clusters, representing groups of chemicals that likely co-occur in the environment due to common chemical use patterns.
Chemicals are arranged, per cluster, by overall score, which is a composite score based on physicochemical and structural similarity to the listed BC. Each UC is listed alongside the BC that displayed the most similar physicochemical properties per cluster (i.e., highest correlated BC).
UC DTXSID UC Chemical Name Highest Correlated BC DTXSID Highest Correlated BC Chemical Name Spearman Rank Correlation Spearman Rank p- value Physicoche mical Similarity Score Total Enriched Features Structural Similarity Score Overall Score
Cluster 1
DTXSID90889705 Methyl Blue DTXSID7021441 Benzyl Violet 4B 0.993 <0.001 0.992 5 0.714 1.706
DTXSID2020189 FD&C Blue No. 1 DTXSID7021441 Benzyl Violet 4B 1 <0.001 1 4 0.571 1.571
DTXSID4034310 C.I. Acid Blue 9 DTXSID7021441 Benzyl Violet 4B 1 <0.001 1 4 0.571 1.571
DTXSID601015325 C.I. Acid Blue 9, aluminum salt (3:2) DTXSID7021441 Benzyl Violet 4B 1 <0.001 1 4 0.571 1.571
DTXSID10925937 Sodium 4-[[4-(diethylamino)phe nyl][4-(diethyliminio)cycl ohexa-2,5-dien-1-ylidene]methyl]na phthalene-2,7-disulfonate DTXSID7021441 Benzyl Violet 4B 0.993 <0.001 0.992 4 0.571 1.564
DTXSID3020673 FD&C Green No. 3 DTXSID7021441 Benzyl Violet 4B 0.993 <0.001 0.992 4 0.571 1.564
DTXSID3026065 Sulfan blue DTXSID7021441 Benzyl Violet 4B 0.979 <0.001 0.977 4 0.571 1.548
DTXSID0029264 C.I. Fluorescent brightening agent 28 DTXSID1020069 2-Amino-5-azotoluene 0.594 4.58E-02 0.547 7 1 1.547
DTXSID2027757 Disodium 4,4’-bis-(2-sulfostyryl)biphen yl DTXSID1020069 2-Amino-5- azotoluene 0.594 4.58E-02 0.547 7 1 1.547
DTXSID5038888 Basic Blue 7 DTXSID7021441 Benzyl Violet 4B 0.545 7.07E-02 0.492 7 1 1.492
Cluster 4
DTXSID40974175 Acid green 22 DTXSID4022367 Estrone 0.671 2.04E-02 0.801 4 0.8 1.601
DTXSID1036541 Pregnenolone DTXSID3022370 Progesterone 0.972 <0.001 1 3 0.6 1.6
DTXSID00873918 Acid Blue 3 DTXSID4022367 Estrone 0.629 3.24E-02 0.773 4 0.8 1.573
DTXSID80192119 4-(3-((9,10-Dihydro-4-hydroxy-9,10-dioxo-1-anthracenyl)amin o)prop- yl)-4-methyl morpholinium methyl sulfate DTXSID0020573 17beta-Estradiol 0.629 3.24E-02 0.773 4 0.8 1.573
DTXSID2021236 HC Red 3 DTXSID0020573 17beta-Estradiol 0.035 9.21E-01 0.38 5 1 1.38
DTXSID90179613 2-(4-Amino-3-nitroanilino)ethanol DTXSID0020573 17beta-Estradiol 0.035 9.21E-01 0.38 5 1 1.38
DTXSID9044532 D&C Blue No. 9 DTXSID0020573 17beta-Estradiol 0.846 <0.001 0.917 2 0.4 1.317
DTXSID60874042 Basic Blue 99 DTXSID4022367 Estrone 0.238 4.57E-01 0.514 4 0.8 1.314
DTXSID90885262 2-Propen-1-one, 1-[4-[[6-O-(6-deoxy-.alpha.-L-mannopyranosyl)-.beta.-D-glucopyranosyl]ox y]-2-hydroxy-6-methoxyphenyl]-3-(3-hydroxy-4-methoxyphenyl)-, (2E)- DTXSID4022367 Estrone 0.469 1.27E-01 0.667 3 0.6 1.267
DTXSID50868556 2-Nitro-5-glyceryl methylaniline DTXSID0020573;
DTXSID4022367 17beta-Estradiol; Estrone 0.154 6.35E-01 0.458 4 0.8 1.258
Cluster 5
DTXSID8037594 Secbumeton DTXSID1023869 Ametryn 1 <0.001 1 6 0.857 1.857
DTXSID90869542 Ethametsulfuron DTXSID9034868;
DTXSID8024101 Prosulfuron; Tribenuron-methyl 0.993 <0.001 0.991 6 0.857 1.848
DTXSID1022057 1,3-Dichloropropene DTXSID0020448 1,2-Dichloropropan e 0.979 <0.001 0.972 6 0.857 1.829
DTXSID1032305 (Z)-Dichloropropene DTXSID0020448 1,2-Dichloropropan e 0.979 <0.001 0.972 6 0.857 1.829
DTXSID2040286 Methoprotryne DTXSID1023869 Ametryn 0.972 <0.001 0.963 6 0.857 1.82
DTXSID3024318 Terbutryn DTXSID8034586 Etoxazole 0.972 <0.001 0.963 6 0.857 1.82
DTXSID2042488 Trietazine DTXSID8034586 Etoxazole 0.951 <0.001 0.935 6 0.857 1.792
DTXSID3032416 Cybutryne DTXSID8034586;
DTXSID1023869 Etoxazole; Ametryn 0.951 <0.001 0.935 6 0.857 1.792
DTXSID3041615 Aziprotryne DTXSID8024238 Oryzalin 0.937 <0.001 0.917 6 0.857 1.774
DTXSID0058223 Indaziflam DTXSID4021218 Quercetin 0.797 3.16E-03 0.731 7 1 1.731
Cluster 6
DTXSID5037494 Deethylatrazine DTXSID9020112 Atrazine 0.888 <0.001 0.886 7 1 1.886
DTXSID0037495 Deisopropylatrazi ne DTXSID9020112 Atrazine 0.867 <0.001 0.862 7 1 1.862
DTXSID1023990 Cyanazine DTXSID0032520 Azoxystrobin 0.832 1.44E-03 0.821 6 0.857 1.678
DTXSID1037806 6-Chloro-1,3,5- triazine-2,4-diamine DTXSID4021268 Simazine 0.979 <0.001 0.992 4 0.571 1.563
DTXSID30886374 Cyclopropanecarb oxylic acid, 3-(2,2-dichloroethenyl)- 2,2-dimethyl-, methyl ester, (1R,3R)-rel- DTXSID3020413 1,2-Dibromo-3-chloropropane 0.965 <0.001 0.976 4 0.571 1.547
DTXSID90886375 Cyclopropanecarb oxylic acid, 3-(2,2-dichloroethenyl)- 2,2-dimethyl-, methyl ester, (1R,3S)-rel- DTXSID3020413 1,2-Dibromo-3-chloropropane 0.965 <0.001 0.976 4 0.571 1.547
DTXSID5024344 Tri-allate DTXSID3020413 1,2-Dibromo-3-chloropropane 0.804 2.75E-03 0.789 4 0.571 1.36
DTXSID001017911 Desulfinylfipronil amide DTXSID0020446 Diuron 0.972 <0.001 0.984 2 0.286 1.269
DTXSID1024124 Thifensulfuron methyl DTXSID0024345 Triasulfuron 0.958 <0.001 0.967 2 0.286 1.253
DTXSID7027833 2-Ethyl-6-methylaniline DTXSID5020449 Dichlorvos 0.951 <0.001 0.959 2 0.286 1.245
Cluster 9
DTXSID7052529 2-Methoxy-4-vinylphenol DTXSID7022413 Isoeugenol 0.832 1.44E-03 0.868 4 1 1.868
DTXSID5021625 p-Isopropenylaceto phenone DTXSID7022413 Isoeugenol 0.944 <0.001 0.96 3 0.75 1.71
DTXSID40110056 Cinnamic acid DTXSID7022413 Isoeugenol 0.769 5.25E-03 0.816 3 0.75 1.566
DTXSID7047647 3-Phenyl-2-propen-1-yl 3-phenylacrylate DTXSID7022413 Isoeugenol 0.755 6.60E-03 0.805 3 0.75 1.555
DTXSID4025587 2-Methylcinnamicaldehyde DTXSID7022413 Isoeugenol 0.748 7.35E-03 0.799 3 0.75 1.549
DTXSID201016569 3,5-Dimethoxy-4-hydroxycinnamald ehyde DTXSID7022413 Isoeugenol 0.734 9.05E-03 0.787 3 0.75 1.537
DTXSID3052143 dl-Borneol DTXSID7022413 Isoeugenol 0.979 <0.001 0.989 2 0.5 1.489
DTXSID40905045 (+)-trans-4-Thujanol DTXSID7022413 Isoeugenol 0.979 <0.001 0.989 2 0.5 1.489
DTXSID9021841 N-Methylaniline DTXSID7022413 Isoeugenol 0.364 2.46E-01 0.483 4 1 1.483
DTXSID9027520 Hexa(methoxymethyl)melamine DTXSID7022413 Isoeugenol 0.615 3.73E-02 0.69 3 0.75 1.44
IMPACT STATEMENT
Most environmental studies on breast cancer have focused on evaluating relationships between individual, well-known chemicals and breast cancer risk. This study set out to expand this research field by identifying understudied chemicals and mixtures that may occur in everyday environments due to their patterns of commercial use. Analyses focused on those that co-occur alongside chemicals associated with breast cancer, based upon in silico chemical database querying and analysis. Particularly in instances when understudied chemicals share physicochemical properties and structural features with carcinogens, these chemical mixtures represent conditions that should be studied in future clinical, epidemiological, and toxicological studies.
Conflict of interest
The authors declare no competing interests.
==== Refs
REFERENCES
1. IBCERCC. Breast Cancer and the Environment: Prioritizing Prevention. Report of the Interagency Breast Cancer and Environmental Research Coordinating Committee (IBCERCC) 2013 [cited 2021 Jun 1]. Available from: https://www.niehs.nih.gov/about/assets/docs/ibcercc_full_508.pdf.
2. Campeau PM , Foulkes WD , Tischkowitz MD . Hereditary breast cancer: new genetic developments, new therapeutic avenues. Hum Genet. 2008;124 (1 ):31–42.18575892
3. Apostolou P , Fostira F . Hereditary breast cancer: the era of new susceptibility genes. Biomed Res Int. 2013;2013 :747318.23586058
4. IOM. Breast Cancer and the Environment: A Life Course Approach: The Institute of Medicine (IOM) of the National Academies; 2012 [cited 2021 Jun 1]. Available from: https://www.nap.edu/catalog/13263/breast-cancer-and-the-environment-a-life-course-approach.
5. WCRF. Diet, nutrition, physical activity and breast cancer: World Cancer Research Fund (WCRF); 2018 [cited 2021 Jun 1]. Available from: https://www.wcrf.org/wp-content/uploads/2021/02/Breast-cancer-report.pdf.
6. Hiatt RA , Brody JG . Environmental Determinants of Breast Cancer. Annu Rev Public Health. 2018;39 :113–33.29328875
7. Gray JM , Rasanayagam S , Engel C , Rizzo J . State of the evidence 2017: an update on the connection between breast cancer and the environment. Environ Health. 2017;16 (1 ):94.28865460
8. Rodgers KM , Udesky JO , Rudel RA , Brody JG . Environmental chemicals and breast cancer: An updated review of epidemiological literature informed by biological mechanisms. Environ Res. 2018;160 :152–82.28987728
9. Carlin DJ , Rider CV , Woychik R , Birnbaum LS . Unraveling the health effects of environmental mixtures: an NIEHS priority. Environ Health Perspect. 2013;121 (1 ):A6–8.23409283
10. Dionisio KL , Frame AM , Goldsmith MR , Wambaugh JF , Liddell A , Cathey T , Exploring consumer exposure pathways and patterns of use for chemicals in the environment. Toxicol Rep. 2015;2 :228–37.28962356
11. Dionisio KL , Phillips K , Price PS , Grulke CM , Williams A , Biryol D , The Chemical and Products Database, a resource for exposure-relevant data on chemicals in consumer products. Sci Data. 2018;5 :180125.29989593
12. Isaacs KK , Dionisio K , Phillips K , Bevington C , Egeghy P , Price PS . Establishing a system of consumer product use categories to support rapid modeling of human exposure. J Expo Sci Environ Epidemiol. 2020;30 (1 ):171–83.31712628
13. OECD. Internationally harmonised functional product and article use categories ENV/JM/MONO(2017)14. Organisation for Economic Co-operation and Development (OECD). Organisation for Economic Co-operation and Development, 2017.
14. Williams AJ , Grulke CM , Edwards J , McEachran AD , Mansouri K , Baker NC , The CompTox Chemistry Dashboard: a community data resource for environmental chemistry. J Cheminform. 2017;9 (1 ):61.29185060
15. Grulke CM , Williams AJ , Thillanadarajah I , Richard AM . EPA’s DSSTox database: History of development of a curated chemistry resource supporting computational toxicology research. Comput Toxicol. 2019;12 .
16. EPA U The Chemical and Products Database (CPDat) MySQL Data File 2020 [cited 2020 September 23]; Version 2:[Available from: https://epa.figshare.com/articles/dataset/The_Chemical_and_Products_Database_CPDat_MySQL_Data_File/5352997/2.
17. SSI. Mammary Carcinogens Review Database. Silent Spring Institute (SSI) 2021 [cited 2021 Nov 1]. Available from: http://sciencereview.silentspring.org/mamm_about.cfm.
18. Rudel RA , Attfield KR , Schifano JN , Brody JG . Chemicals causing mammary gland tumors in animals signal new directions for epidemiology, chemicals testing, and risk assessment for breast cancer prevention. Cancer. 2007;109 (12 Suppl ):2635-66.17503434
19. Watford S , Ly Pham L , Wignall J , Shin R , Martin MT , Friedman KP . ToxRefDB version 2.0: Improved utility for predictive and retrospective toxicology analyses. Reprod Toxicol. 2019;89 :145–58.31340180
20. Harvey JB , Hong HH , Bhusari S , Ton TV , Wang Y , Foley JF , F344/NTac Rats Chronically Exposed to Bromodichloroacetic Acid Develop Mammary Adenocarcinomas With Mixed Luminal/Basal Phenotype and Tgfbeta Dysregulation. Vet Pathol. 2016;53 (1 ):170–81.25732176
21. Dunnick JK , Elwell MR , Huff J , Barrett JC . Chemically induced mammary gland cancer in the National Toxicology Program’s carcinogenesis bioassay. Carcinogenesis. 1995;16 (2 ):173–9.7859345
22. EPA U CompTox Chemicals Dashboard Batch Search 2020 [cited 2021]; 3.5:[Available from: https://comptox.epa.gov/dashboard/dsstoxdb/batch_search.
23. Rager JE , Clark J , Eaves LA , Avula V , Niehoff NM , Kim YH , Mixtures modeling identifies chemical inducers versus repressors of toxicity associated with wildfire smoke. Sci Total Environ. 2021;775 :145759.33611182
24. Klaren WD , Ring C , Harris MA , Thompson CM , Borghoff S , Sipes NS , Identifying Attributes That Influence In Vitro-to-In Vivo Concordance by Comparing In Vitro Tox21 Bioactivity Versus In Vivo DrugMatrix Transcriptomic Responses Across 130 Chemicals. Toxicol Sci. 2019;167 (1 ):157–71.30202884
25. Rager JE , Suh M , Chappell GA , Thompson CM , Proctor DM . Review of transcriptomic responses to hexavalent chromium exposure in lung cells supports a role of epigenetic mediators in carcinogenesis. Toxicol Lett. 2019;305 :40–50.30690063
26. Phillips KA , Wambaugh JF , Grulke CM , Dionisio KL , Isaacs KK . High-throughput screening of chemicals as functional substitutes using structure-based classification models. Green Chem. 2017;19 (4 ):1063–74.30505234
27. Leydesdorff L On the normalization and visualization of author co-citation data: Salton’s Cosine versus the Jaccard index. Journal of the American Society for Information Science & Technology. 2008;59 (1 ):77–85.
28. Rousseeuw PJ . Silhouettes: a graphical aid to the interpretation and validation of cluster analysis Journal of Computational and Applied Mathematics. 1987;20 :53–65.
29. Krishna S , Berridge B , Kleinstreuer N . High-Throughput Screening to Identify Chemical Cardiotoxic Potential. Chem Res Toxicol. 2021;34 (2 ):566–83.33346635
30. Lowe CN , Phillips KA , Favela KA , Yau AY , Wambaugh JF , Sobus JR , Chemical Characterization of Recycled Consumer Products Using Suspect Screening Analysis. Environ Sci Technol. 2021;55 (16 ):11375–87.34347456
31. Beckers LM , Busch W , Krauss M , Schulze T , Brack W . Characterization and risk assessment of seasonal and weather dynamics in organic pollutant mixtures from discharge of a separate sewer system. Water Res. 2018;135 :122–33.29466716
32. Patlewicz G , Ball N , Booth ED , Hulzebos E , Zvinavashe E , Hennes C . Use of category approaches, read-across and (Q)SAR: general considerations. Regul Toxicol Pharmacol. 2013;67 (1 ):1–12.23764304
33. Yang C , Tarkhov A , Marusczyk J , Bienfait B , Gasteiger J , Kleinoeder T , New publicly available chemical query language, CSRML, to support chemotype representations for application to data mining and modeling. J Chem Inf Model. 2015;55 (3 ):510–28.25647539
34. Wang J , Hallinger DR , Murr AS , Buckalew AR , Lougee RR , Richard AM , High-throughput screening and chemotype-enrichment analysis of ToxCast phase II chemicals evaluated for human sodium-iodide symporter (NIS) inhibition. Environ Int. 2019;126 :377–86.30826616
35. Ring C , Sipes NS , Hsieh JH , Carberry C , Koval LE , Klaren WD , Predictive modeling of biological responses in the rat liver using in vitro Tox21 bioactivity: Benefits from high-throughput toxicokinetics. Comput Toxicol. 2021;18 .
36. Zhang Z , Wang S , Li L . Emerging investigator series: the role of chemical properties in human exposure to environmental chemicals. Environ Sci Process Impacts. 2021.
37. Mansouri K , Grulke CM , Judson RS , Williams AJ . OPERA models for predicting physicochemical properties and environmental fate endpoints. J Cheminform. 2018;10 (1 ):10.29520515
38. Rager JE , Strynar MJ , Liang S , McMahen RL , Richard AM , Grulke CM , Linking high resolution mass spectrometry data with exposure and toxicity forecasts to advance high-throughput environmental monitoring. Environ Int. 2016;88 :269–80.26812473
39. Auerbach S , Filer D , Reif D , Walker V , Holloway AC , Schlezinger J , Prioritizing Environmental Chemicals for Obesity and Diabetes Outcomes Research: A Screening Approach Using ToxCast High-Throughput Data. Environ Health Perspect. 2016;124 (8 ):1141–54.26978842
40. Reif DM , Martin MT , Tan SW , Houck KA , Judson RS , Richard AM , Endocrine profiling and prioritization of environmental chemicals using ToxCast data. Environ Health Perspect. 2010;118 (12 ):1714–20.20826373
41. Koval LE , Dionisio KL , Friedman KP , Isaacs KK , Rager JE . Dataset for Environmental Mixtures and Breast Cancer: Identifying Co-Exposure Patterns between Understudied vs Breast Cancer-Associated Chemicals using Chemical Inventory Informatics 2022 [cited 2022 May 27]. Available from: 10.15139/S3/UMPCKW.
42. CFR- Code of Federal Regulations Title 21. Sect. 81.10 (1977).
43. Samavat H , Kurzer MS . Estrogen metabolism and breast cancer. Cancer Lett. 2015;356 (2 Pt A ):231–43.24784887
44. Cardona B , Rudel RA . US EPA’s regulatory pesticide evaluations need clearer guidelines for considering mammary gland tumors and other mammary gland effects. Mol Cell Endocrinol. 2020;518 :110927.32645345
45. Stillwell W An Introduction to Biological Membranes Composition, Structure and Function. 2nd edition ed: Elsevier Science; 2016 June 30.
46. Trabert B , Sherman ME , Kannan N , Stanczyk FZ . Progesterone and Breast Cancer. Endocr Rev. 2020;41 (2 ).
47. Kulkoyluoglu-Cotul E , Arca A , Madak-Erdogan Z . Crosstalk between Estrogen Signaling and Breast Cancer Metabolism. Trends Endocrinol Metab. 2019;30 (1 ):25–38.30471920
48. CDC. Fourth National Report on Human Exposure to Environmental Chemicals. 2021.
49. USGS. USGS Water Data for USA 2021. Available from: https://waterdata.usgs.gov/nwis?
50. Ring CL , Arnot JA , Bennett DH , Egeghy PP , Fantke P , Huang L , Consensus Modeling of Median Chemical Intake for the U.S. Population Based on Predictions of Exposure Pathways. Environ Sci Technol. 2019;53 (2 ):719–32.30516957
51. Baker N , Knudsen T , Williams A . Abstract Sifter: a comprehensive front-end system to PubMed. F1000Res. 2017;6 .
52. Kapraun DF , Wambaugh JF , Ring CL , Tornero-Velez R , Setzer RW . A Method for Identifying Prevalent Chemical Combinations in the U.S. Population. Environ Health Perspect. 2017;125 (8 ):087017.28858827
53. Clark J , Avula V , Ring C , Eaves LA , Howard T , Santos HP , Comparing the Predictivity of Human Placental Gene, microRNA, and CpG Methylation Signatures in Relation to Perinatal Outcomes. Toxicol Sci. 2021.
54. Wambaugh JF , Wang A , Dionisio KL , Frame A , Egeghy P , Judson R , High throughput heuristics for prioritizing human exposure to environmental chemicals. Environ Sci Technol. 2014;48 (21 ):12760–7.25343693
55. Watford SM , Grashow RG , De La Rosa VY , Rudel RA , Friedman KP , Martin MT . Novel application of normalized pointwise mutual information (NPMI) to mine biomedical literature for gene sets associated with disease: use case in breast carcinogenesis. Comput Toxicol. 2018;7 :46–57.32274464
56. Taylor KW , Joubert BR , Braun JM , Dilworth C , Gennings C , Hauser R , Statistical Approaches for Assessing Health Effects of Environmental Chemical Mixtures in Epidemiology: Lessons from an Innovative Workshop. Environ Health Perspect. 2016;124 (12 ):A227–A9.27905274
child lead exposure for the plaintiffs in a public nuisance case related to childhood lead poisoning. None of these activities were directly related to the present study. The other authors declare they have no actual or potential competing financial interests.
57. Drakvik E , Altenburger R , Aoki Y , Backhaus T , Bahadori T , Barouki R , Statement on advancing the assessment of chemical mixtures and their risks for human health and the environment. Environ Int. 2020;134 :105267.31704565
58. Rider CV , McHale CM , Webster TF , Lowe L , Goodson WH 3rd , La Merrill MA , Using the Key Characteristics of Carcinogens to Develop Research on Chemical Mixtures and Cancer. Environ Health Perspect. 2021;129 (3 ):35003.33784186
59. Ragerlab. Ragerlab Github 2021 [cited 2021]. Available from: https://github.com/Ragerlab.
60. ToxPrint. ToxPrint: Altamira LLC; 2021 [cited 2021 August, 6]. Available from: https://toxprint.org.
| 35710593 | PMC9742149 | NO-CC CODE | 2022-12-16 23:23:54 | no | J Expo Sci Environ Epidemiol. 2022 Nov 16; 32(6):794-807 | utf-8 | J Expo Sci Environ Epidemiol | 2,022 | 10.1038/s41370-022-00451-8 | oa_other |
==== Front
Immunity
Immunity
Immunity
1074-7613
1097-4180
The Authors. Published by Elsevier Inc.
S1074-7613(22)00641-0
10.1016/j.immuni.2022.12.005
Article
Immunoglobulin germline gene polymorphisms influence the function of SARS-CoV-2 neutralizing antibodies
Pushparaj Pradeepa 1
Nicoletto Andrea 1∗
Sheward Daniel J. 1∗
Das Hrishikesh 2
Castro Dopico Xaquin 1
Perez Vidakovics Laura 1
Hanke Leo 1
Chernyshev Mark 1
Narang Sanjana 1
Kim Sungyong 1
Fischbach Julian 1
Ekström Simon 3
McInerney Gerald 1
Hällberg B. Martin 2
Murrell Ben 1
Corcoran Martin 1
Karlsson Hedestam Gunilla B. 13#
1 Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, SE-171 77 Stockholm, Sweden
2 Department of Cell and Molecular Biology, Karolinska Institutet, SE-171 77 Stockholm, Sweden
3 Department of Biomedical Engineering, Lund University, SE-221 84 Lund, Sweden
# Corresponding author: Gunilla B. Karlsson Hedestam, Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, Box 280, S-171 77 Stockholm, Sweden.
∗ Equal contribution
12 12 2022
12 12 2022
28 5 2022
23 9 2022
7 12 2022
© 2022 The Authors. 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.
The human immunoglobulin heavy chain (IGH) locus is exceptionally polymorphic, with high levels of allelic and structural variation. Thus, germline IGH genotypes are personal, which may influence responses to infection and vaccination. For an improved understanding of inter-individual differences in antibody responses, we isolated SARS-CoV-2 spike-specific monoclonal antibodies from convalescent health care workers, focusing on the IGHV1-69 gene, which has the highest level of allelic variation of all IGHV genes. The IGHV1-69∗20-using CAB-I47 antibody and two similar antibodies isolated from an independent donor were critically dependent on allele usage. Neutralization was retained when reverting the V region to the germline IGHV1-69∗20 allele but lost when reverting to other IGHV1-69 alleles. Structural data confirmed that two germline-encoded polymorphisms, R50 and F55, in the IGHV1-69 gene were required for high affinity receptor binding domain interaction. These results demonstrate that single polymorphisms in IGH genes can influence the function of SARS-CoV-2 neutralizing antibodies.
Graphical abstract
The genes and alleles of the antigen receptor loci are highly variable between individuals, which may affect the quality of the immune response to different pathogens. Here, Pushparaj et. al. use immunoglobulin genotyping and monoclonal antibody engineering to illustrate how heritable differences in such genes can modulate anti-SARS-CoV-2 antibody function.
Keywords
SARS-CoV-2
immunoglobulin allelic diversity
copy number variation
single nucleotide polymorphism
neutralizing antibody
==== Body
pmc3Lead contact
| 0 | PMC9742198 | NO-CC CODE | 2022-12-15 23:16:09 | no | Immunity. 2022 Dec 12; doi: 10.1016/j.immuni.2022.12.005 | utf-8 | Immunity | 2,022 | 10.1016/j.immuni.2022.12.005 | oa_other |
==== Front
J Am Med Dir Assoc
J Am Med Dir Assoc
Journal of the American Medical Directors Association
1525-8610
1538-9375
Published by Elsevier Inc. on behalf of AMDA -- The Society for Post-Acute and Long-Term Care Medicine.
S1525-8610(22)00969-0
10.1016/j.jamda.2022.12.008
Original Studies
Impact of COVID-19 on long-term care service utilization of older home-dwelling adults in Japan
Ishii Shinya PhD 1∗
Tanabe Kazutaka PhD 1
Ishimaru Bunji PhD 1
Kitahara Kanako PhD 1
1 Division of the Health for the Elderly, Health and Welfare Bureau for the Elderly, Ministry of Health, Labour and Welfare, Tokyo, Japan
∗ Correspondence: Dr. Shinya Ishii, Division of the Health for the Elderly, Health and Welfare Bureau for the Elderly, Ministry of Health, Labour and Welfare, Japan, 1-2-2 Kasumigaseki, Chuo-ku, Tokyo 100-8916. Phone (03)5253-1111.
12 12 2022
12 12 2022
5 7 2022
3 11 2022
4 12 2022
© 2022 Published by Elsevier Inc. on behalf of AMDA -- The Society for Post-Acute and Long-Term Care Medicine.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objectives
The COVID-19 outbreak severely affected long-term care (LTC) service provision. This study aimed to quantitatively evaluate its impact on the utilization of LTC services by older home-dwelling adults and identify its associated factors.
Design
A retrospective repeated cross-sectional study
Setting and Participants
Data from a nationwide LTC Insurance Comprehensive Database comprising monthly claims from January 2019 to September 2020
Methods
Interrupted time series analyses and segmented negative binomial regression were employed to examine changes in use for each of the 15 LTC services. Results of the analyses were synthesized using random effect meta-analysis in three service types (home-visit, commuting, and short-stay services).
Results
LTC service use declined in April 2020 when the state of emergency (SOE) was declared, followed by a gradual recovery in June after the SOE was lifted. There was a significant association between decline in LTC service use and SOE while the association between LTC service use and the status of the infection spread was limited. Service type was associated with changes in service utilization, with a more precipitous decline in commuting and short-stay services than in home-visiting services during the SOE. Service use by those with dementia was higher than that by those without dementia, particularly in commuting and short-stay services, partially cancelling out the decline in service use that occurred during the SOE.
Conclusions and Implications
There was a significant decline in LTC service utilization during the SOE. The decline varied depending on service types and the dementia severity of service users. These findings would help LTC professionals identify vulnerable groups and guide future plans geared toward effective infection prevention while alleviating unfavorable impacts by infection prevention measures. Future studies are required to examine the effects of the LTC service decline on older adults.
Key words
COVID-19
long-term care service
dementia
==== Body
pmcBrief summary: The study found that long term care service use declined significantly during the state of emergency declared to control the spread of COVID-19 and the decline varied depending on service types and dementia severity of service users.
Funding:
This research did not receive any funding from agencies in the public, commercial, or not-for-profit sectors.
Data availability
The database is strictly regulated for its use, but it is available for academic researchers as per reasonable and eligible application to the Ministry of Health, Labour and Welfare and after deliberation and approval by a third-party committee. (https://www.mhlw.go.jp/content/12301000/000342386.pdf).
| 0 | PMC9742200 | NO-CC CODE | 2022-12-14 23:53:20 | no | J Am Med Dir Assoc. 2022 Dec 12; doi: 10.1016/j.jamda.2022.12.008 | utf-8 | J Am Med Dir Assoc | 2,022 | 10.1016/j.jamda.2022.12.008 | oa_other |
==== Front
Cell Rep
Cell Rep
Cell Reports
2211-1247
The Author(s).
S2211-1247(22)01791-0
10.1016/j.celrep.2022.111892
111892
Article
SARS-CoV-2 escapes direct NK cell killing through Nsp1-mediated downregulation of ligands for NKG2D
Lee Madeline J. 12
Leong Michelle W. 2
Rustagi Arjun 2
Beck Aimee 2
Zeng Leiping 3
Holmes Susan 4
Qi Lei S. 356
Blish Catherine A. 267∗
1 Stanford Immunology Program, Stanford University School of Medicine, Stanford, CA, 94305, United States of America
2 Department of Medicine, Stanford University School of Medicine, Stanford, CA, 94305, United States of America
3 Department of Bioengineering, Stanford University, Stanford, CA, 94305, United States of America
4 Department of Statistics, Stanford University, Stanford, CA, 94305, United States of America
5 Sarafan Chem-H, Stanford University, Stanford, CA, 94305, United States of America
6 Chan Zuckerberg Biohub, San Francisco, CA, 94157, United States of America
7 Stanford Medical Scientist Training Program, Stanford University School of Medicine, Stanford, CA, 94305, United States of America
∗ Lead Contact:
12 12 2022
12 12 2022
11189230 6 2022
9 11 2022
5 12 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.
Natural killer (NK) cells are cytotoxic effector cells that target and lyse virally-infected cells; many viruses therefore encode mechanisms to escape such NK cell killing. Here, we interrogate the ability of SARS-CoV-2 to modulate NK cell recognition and lysis of infected cells. We find that NK cells exhibit poor cytotoxic responses against SARS-CoV-2-infected targets, preferentially killing uninfected bystander cells. We demonstrate that this escape is driven by downregulation of ligands for the activating receptor NKG2D (“NKG2D-L”). Indeed, early in viral infection, prior to NKG2D-L downregulation, NK cells are able to target and kill infected cells; however, this ability is lost as viral proteins are expressed. Finally, we find that SARS-CoV-2 non-structural protein 1 (Nsp1) mediates downregulation of NKG2D-L and that Nsp1 alone is sufficient to confer resistance to NK cell killing. Collectively, our work demonstrates that SARS-CoV-2 evades direct NK cell cytotoxicity and describes a mechanism by which this occurs.
Graphical abstract
Natural killer cells are designed to kill virally infected cells. Lee et al. demonstrate that healthy NK cells are unable to efficiently lyse SARS-CoV-2-infected cells, likely due to their loss of the ligands for the activating receptor NKG2D. These effects are mediated by the viral protein Nsp1.
Keywords
COVID-19
SARS-CoV-2
natural killer cells
NKG2D
Nsp1
immune escape
==== Body
pmc
| 0 | PMC9742201 | NO-CC CODE | 2022-12-14 23:29:59 | no | Cell Rep. 2022 Dec 12;:111892 | utf-8 | Cell Rep | 2,022 | 10.1016/j.celrep.2022.111892 | oa_other |
==== Front
Intensive Crit Care Nurs
Intensive Crit Care Nurs
Intensive & Critical Care Nursing
0964-3397
1532-4036
Elsevier Ltd.
S0964-3397(22)00177-X
10.1016/j.iccn.2022.103374
103374
Correspondence
Quantifying oxygen supply and demand during the COVID-19 pandemic: An integrated health system perspective
White Heath D ab
Danesh Valerie cd⁎
Ogola Gerald O e
Jimenez Edgar J fb
Arroliga Alejandro C gh
a Pulmonary, Critical Care and Sleep Medicine, Baylor Scott & White Health, Temple, TX, United States
b College of Medicine, Texas A&M University, College Station, TX, United States
c Center for Applied Health Research, Baylor Scott & White Research Institute, Dallas, TX, United States
d School of Nursing, University of Texas at Austin, Austin, TX, United States
e Biostatistics, Baylor Scott & White Research Institute, Dallas, TX, United States
f Baylor Scott & White Health, Dallas, TX, United States
g Pulmonary and Critical Care Medicine, Baylor Scott & White Health, Dallas, TX, United States
h Baylor College of Medicine, Houston, TX, United States
⁎ Corresponding author.
12 12 2022
12 12 2022
1033746 12 2022
9 12 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.
==== Body
pmcDear Editor,
In 2020, widespread attention was turned to evaluate the supply and demand for mechanical ventilation in combination with critical care workforce staffing strategies to accommodate surges in COVID-19 critical illness (Tsai et al., 2022). Beyond critical care demand, oxygen therapy is a predominant first-line intervention for hypoxemic respiratory failure across a broad range of settings, including delivery via nasal cannula. Thus, we examine longitudinal oxygen utilization during the first 24 months of the COVID-19 pandemic to quantify oxygen consumption increases inclusive of high-flow nasal cannula (HFNC) use for patients hospitalized with COVID-19 in a large and diverse integrated health system.
Methods
Data were drawn from a prospective, cross-sectional, observational study conducted in collaboration with the Society of Critical Care Medicine Discovery Viral Infection and Respiratory Illness Universal Study (VIRUS): COVID-19 Registry. We analyzed the volume of oxygen used, types of oxygen support, and the volume of hospitalized patients with and without COVID-19 over a 2-year period (April 2020 and March 2022) in a 4,457-bed integrated health system of 26 adult acute care hospitals serving rural, suburban and urban populations in the Southern region of the United States. The VIRUS registry was approved by the Baylor Scott & White Research Institute Institutional Review Board. We followed the STROBE reporting guideline.
Results
Oxygen consumption peaks coincided with peaks in COVID-related hospitalization, including up to a 134% increase in volume of oxygen used (Figure 1a ). The predominant type of oxygen support for patients with COVID-related hospitalizations was delivered via HFNC (Figure 1b). High levels of oxygen consumption coincide with high levels of HFNC use for patients hospitalized with COVID-19. Reciprocal changes in HFNC use during non-peak periods of COVID-19 hospitalizations coincided with lower volumes of oxygen utilization. Widespread adoption of HFNC within an integrated health system setting is discernible within the first 6 months. Furthermore, the dual increase in hospitalization volumes due to COVID-19 and oxygen flow rates for HFNC of 30-60 L/min (compared with 20-30L/min for noninvasive and invasive mechanical ventilation), illustrate a sustained high demand for medical grade oxygen.Figure 1a Oxygen utilization and hospitalization volumes (April 2020 – March 2022) Figure 1b. Oxygen utilization and oxygen support types for COVID-19 hospitalizations (April 2020 – March 2022) Abbreviations: O2 = CuFt = cubic feet; O2 = oxygen; IMV = invasive mechanical ventilation; NIV = non-invasive mechanical ventilation; HFNC = high-flow nasal cannula
Discussion
Oxygen therapy is a potentially scarce resource in the context of a long-term respiratory pandemic. While oxygen scarcity is a known challenge in low-resource settings (Fowler et al., 2008), the COVID-19 pandemic has introduced the relevance of oxygen conservation strategies at a global level to consider disaster management beyond short-term access problems to oxygen supply associated with natural disasters (e.g., hurricanes) (Blakeman and Branson, 2013). Similar to Days of Cash On Hand (DCOH) metrics, oxygen supply affects the solvency and viability of healthcare delivery, with some hospitals reporting less than a 48-hour supply of oxygen in reserve(Parkinson, 2021). Further, the demand for high-flow oxygen therapy has been established as a non-invasive intervention for respiratory failure as early concerns related to efficacy and aerosolization were tempered by favorable outcomes. The combination of a respiratory virus pandemic with clinical practice pattern changes creates a sustained change in demand for medical oxygen with increases in average and peak utilization.
Ethical Statement
The VIRUS registry was approved by the Baylor Scott & White Research Institute Institutional Review Board. The approval number is #020-119. The ClinicalTrials.gov identifier for the VIRUS registry is NCT04323787. The reporting of this study confirms to the STROBE statement.
Author Contributions:
Concept and design: Arroliga, White, Danesh, Jimenez.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of manuscript: White, Danesh, Arroliga, Jimenez.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: All authors.
Obtained funding: Danesh.
Administrative, technical, or material support: Danesh, Arroliga.
Supervision: Danesh, Arroliga.
Name and location of the institution where the study was performed:
Baylor Scott & White Health, Texas, United States
Name, date and location of any meeting or forum where research data were previously presented, and who presented:
The full-length Research Letter and Figure 1a have not been previously presented. An abbreviated meeting abstract version with Figure 1b has been submitted to the American Thoracic Society (ATS) 2023 meeting for consideration. Status as of November 21, 2022: Under review with ATS meeting abstract reviewers for decisions in mid-January 2023 for meeting abstract publication in May 2023.
Sources of financial support:
This work was partially funded by the Cardiovascular Research Review Committee of the Baylor Healthcare System Foundation, Society of Critical Care Medicine, and the Gordon and Betty Moore Foundation.
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
Tsai TC, Orav EJ, Jha AK, Figueroa JF. National Estimates of Increase in US Mechanical Ventilator Supply During the COVID-19 Pandemic. JAMA Network Open. 2022;5(8):e2224853.
Fowler R.A. Adhikari N.K.J. Bhagwanjee S. Clinical review: critical care in the global context–disparities in burden of illness, access, and economics Critical Care. 12 5 2008 225 19014409
Blakeman T.C. Branson R.D. Oxygen Supplies in Disaster Management Respiratory Care. 58 1 2013 173 183 23271827
Parkinson N. Oxygen supplies tight in Southern US states In. Gas World Vol 2022 2021
| 0 | PMC9742205 | NO-CC CODE | 2022-12-14 23:29:59 | no | Intensive Crit Care Nurs. 2022 Dec 12;:103374 | utf-8 | Intensive Crit Care Nurs | 2,022 | 10.1016/j.iccn.2022.103374 | oa_other |
==== Front
Rev Esp Med Nucl Imagen Mol
Rev Esp Med Nucl Imagen Mol
Revista Espanola De Medicina Nuclear E Imagen Molecular
2253-654X
2253-8070
Sociedad Española de Medicina Nuclear e Imagen Molecular. Published by Elsevier España, S.L.U.
S2253-654X(22)00188-3
10.1016/j.remn.2022.11.002
Article
Valoración del tromboembolismo pulmonar relacionado con infección SARS-CoV-2 activa en pacientes embarazadas
Assessment of pulmonary embolism related to active SARS-CoV-2 infection in pregnant women.Moreno-Ballesteros Ana 1⁎
Rebollo-Aguirre Ángel C. 1
Bolívar-Roldán Isabel 2
Busquier Teresa 3
Mora Elena Sanchez-de 1
Jimenez-Heffernan Amelia 1
1 Unidad de Medicina Nuclear. Hospital Universitario Juan Ramón Jiménez. Ronda Norte, s/n, 21005, Huelva, España
2 Unidad de Medicina Nuclear. Hospital Universitario Virgen Macarena. Avenida Dr Fedriani nº3, 41009, Sevilla, España
3 Unidad de Radiodiagnóstico. Hospital Universitario Virgen Macarena. Avenida Dr Fedriani nº3, 41009, Sevilla, España
⁎ Autor de correspondencia: Unidad de Medicina Nuclear, Hospital Universitario Juan Ramón Jiménez, Ronda Norte, s/n, 21005, Huelva, España
12 12 2022
12 12 2022
2 11 2022
20 11 2022
© 2022 Sociedad Española de Medicina Nuclear e Imagen Molecular. Published by Elsevier España, S.L.U. 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.
Objetivo: Analizar la muestra de pacientes embarazadas a las que se le realizó una gammagrafía de perfusión pulmonar para descartar la sospecha de tromboembolismo pulmonar (TEP) durante el ingreso en nuestro centro por infección aguda por COVID-19.
Material y métodos: A todas las pacientes (n = 5) se les realizó una gammagrafía SPECT con dosis reducida (111 MBq) de 99mTc-macroagregados de albúmina. Las imágenes obtenidas se interpretaron comparando los hallazgos con la imagen radiológica según criterios PISAPED.
Resultados: De las 5 pacientes, tan sólo en una se diagnosticó TEP. En dos pacientes los hallazgos patológicos de la gammagrafía fueron atribuibles a alteraciones radiológicas por neumonía COVID-19, y otras dos mostraron una perfusión pulmonar normal.
Conclusión: Dado lo inespecífico de las manifestaciones clínicas y valores del dímero-D dentro de la COVID-19, así como su similitud con los de TEP, la gammagrafía de perfusión pulmonar, por su alta sensibilidad y menor irradiación que la TC, tiene un papel crucial en el despistaje de TEP en estas pacientes. Los resultados obtenidos son de especial relevancia, a pesar del número limitado de pacientes, dada la ausencia de publicaciones científicas en este grupo de pacientes dentro de la situación excepcional por la pandemia COVID-19.
Aim: To analyze the sample of pregnant patients who underwent pulmonary perfusion scintigraphy to rule out the pulmonary embolism (PE) suspicion during the acute COVID-19 infection hospitalization period in our hospital.
Material and methods: SPECT scintigraphy with a reduced dose (111 MBq) of 99mTc-macroaggregated albumin was performed in all of the patients (n=5). The obtained images were interpreted by comparing the findings with the radiological images according to the PISAPED criteria.
Results: Only one of the 5 patients was diagnosed with PE. Two patients obtained pathological findings of the scintigraphy attributable to radiological alterations due to COVID-19 pneumonia, and the other two had normal pulmonary perfussion.
Conclusion: Given the non-specific features of the clinical manifestations and D-dimer values in COVID-19, as well as their similarity to those of PE, the pulmonary perfusion scintigraphy plays a crucial role in the screening of PE in these patients due to its high sensitivity and lower irradiation compared to CT. Despite the limited number of patients, the results obtained have special relevance related to the absence of scientific publications on this group of patients within the context of COVID-19 pandemic exceptional situation.
Palabras clave
tromboembolismo pulmonar
embarazo
gammagrafía
SPECT
COVID-19
SARS-CoV-2
Keywords
pulmonary embolism
pregnancy
scintigraphy
SPECT
COVID-19
SARS-CoV-2
==== Body
pmc
| 0 | PMC9742206 | NO-CC CODE | 2022-12-14 23:29:59 | no | Rev Esp Med Nucl Imagen Mol. 2022 Dec 12; doi: 10.1016/j.remn.2022.11.002 | utf-8 | Rev Esp Med Nucl Imagen Mol | 2,022 | 10.1016/j.remn.2022.11.002 | oa_other |
==== Front
Arthrosc Sports Med Rehabil
Arthrosc Sports Med Rehabil
Arthroscopy, Sports Medicine, and Rehabilitation
2666-061X
Published by Elsevier Inc. on behalf of the Arthroscopy Association of North America.
S2666-061X(22)00198-5
10.1016/j.asmr.2022.11.026
Article
Injury rates remained elevated in the second National Football League season after the onset of the COVID-19 pandemic
Platt Brooks MD ∗
Abed Varag BS ∗
Khalily Camille MD ∗
Sullivan Breanna BA ∗
Skinner Matthew BS ∗
Jacobs Cale PhD ∗
Johnson Darren MD ∗
Stone Austin V. MD PhD ∗†
∗ Department of Orthopaedic Surgery and Sports Medicine, University of Kentucky
† Address correspondence to: Austin V. Stone, MD PhD, Department of Orthopaedic Surgery and Sports Medicine, 2195 Harrodsburg Rd, Lexington, KY 40504, University of Kentucky. . Phone: (859) 218-3065
12 12 2022
12 12 2022
16 8 2022
8 11 2022
25 11 2022
© 2022 Published by Elsevier Inc. on behalf of the Arthroscopy Association of North America.
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.
Purpose
The purpose of this study is to compare the injury incidence of the 2018-2019 and 2020 National Football League (NFL) seasons with the 2021 season.
Methods
Publicly released NFL weekly injury reports were queried to identify players listed as “out” or placed on injured reserve (IR) for at least one game in the 2018-2021 seasons. Injuries were then categorized into upper extremity, lower extremity, spine/core, and head. Incidence per 1,000 athlete exposures were calculated for each season and proportions of injuries by position were calculated separately for the 2018-2019, 2020, and 2021 cohorts. Incidence rate ratios (IRR) were used to compare injury rates.
Results
Overall injury incidence in the 2021 NFL season increased compared to the pre-COVID-19 seasons (2018-2019) in all anatomical zones except for the upper extremity. [28.70 vs. 23.09 per 1,000 exposures, IRR 1.24 (95% CI: 1.14-1.36); p< 0.001]. The injury rate remained elevated and further increased in 2021 compared to the 2020 season for all anatomical zones other than the spine/core [28.70 vs. 21.64 per 1,000 exposures, IRR 1.33 (1.19-1.47); p< 0.001]. No significant difference existed during the early season (weeks 1-4); however, injury rates after week 4 increased in 2021 compared to both the 2018-2019 and 2020 seasons.
Conclusion
The injury incidence in the 2021 season remained elevated and increased further compared to both the 2018-2019 and 2020 seasons. Traumatic injuries resulting in missed games increased despite return to a more traditional season since the beginning of the COVID-19 pandemic. The injury rates significantly increased in mid- to late season.
Level of Evidence
III, cross-sectional study
==== Body
pmcIRB: This project did not require review by the institutional review board.
| 0 | PMC9742207 | NO-CC CODE | 2022-12-14 23:29:59 | no | Arthrosc Sports Med Rehabil. 2022 Dec 12; doi: 10.1016/j.asmr.2022.11.026 | utf-8 | Arthrosc Sports Med Rehabil | 2,022 | 10.1016/j.asmr.2022.11.026 | oa_other |
==== Front
Vaccine
Vaccine
Vaccine
0264-410X
1873-2518
Elsevier Science
S0264-410X(22)01527-4
10.1016/j.vaccine.2022.12.013
Article
Patient flow time data of COVID-19 vaccination clinics in 23 sites, United States, April and May 2021
Cho Bo-Hyun a⁎
Athar Heba M. a
Bates Laurel G. b
Yarnoff Benjamin O. b
Harris LaTreace Q. a
Washington Michael L. a
Jones-Jack Nkenge H. a
Pike Jamison J. a
a Centers for Disease Control and Prevention, Atlanta, GA, United States
b RTI International, Research Triangle Park, NC, United States
⁎ Corresponding author at: Health Economist, Centers for Disease Control and Prevention, 1600 Clifton Rd, NE Atlanta, GA 30329, United States.
12 12 2022
12 12 2022
3 8 2022
10 11 2022
6 12 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.
Introduction
Public health department (PHD) led COVID-19 vaccination clinics can be a critical component of pandemic response as they facilitate high volume of vaccination. However, few patient-time analyses examining patient throughput at mass vaccination clinics with unique COVID-19 vaccination challenges have been published.
Methods
During April and May of 2021, 521 patients in 23 COVID-19 vaccination sites counties of 6 states were followed to measure the time spent from entry to vaccination. The total time was summarized and tabulated by clinic characteristics. A multivariate linear regression analysis was conducted to evaluate the association between vaccination clinic settings and patient waiting times in the clinic.
Results
The average time a patient spent in the clinic from entry to vaccination was 9 min 5 s (range: 02:00–23:39). Longer patient flow times were observed in clinics with higher numbers of doses administered, 6 or fewer vaccinators, walk-in patients accepted, dedicated services for people with disabilities, and drive-through clinics. The multivariate linear regression showed that longer patient waiting times were significantly associated with the number of vaccine doses administered, dedicated services for people with disabilities, the availability of more than one brand of vaccine, and rurality.
Conclusions
Given the standardized procedures outlined by immunization guidelines, reducing the wait time is critical in lowering the patient flow time by relieving the bottleneck effect in the clinic. Our study suggests enhancing the efficiency of PHD-led vaccination clinics by preparing vaccinators to provide vaccines with proper and timely support such as training or delivering necessary supplies and paperwork to the vaccinators. In addition, patient wait time can be spent answering questions about vaccination or reviewing educational materials on other public health services.
Keywords
COVID-19 vaccine
Mass vaccination
Patient time
Throughput time
Public health emergency
==== Body
pmc1 Introduction
Public health department (PHD) led vaccination clinics are designed to administer vaccines to large groups of individuals as rapidly and safely as possible. These PHD programs must accurately and efficiently store, distribute, allocate, and administer vaccine and accompanying supplies, while also monitoring vaccine uptake and safety [1]. To achieve the rapid vaccination in a setting of 1) limited supplies of vaccines, 2) medical resources with competing priorities and 3) vaccination priorities by age or high-risk status, large scale vaccination clinics were crucial in expediting an efficient roll-out of the Coronavirus disease 2019 (COVID-19) vaccines [2]. Although PHD-led vaccination clinics are not a novel concept [3], [4], the COVID-19 pandemic presented unique operational challenges to the implementation of such clinics. For example, the mRNA COVID-19 vaccines present logistical challenges related to the ultra-cold chain requirements for distribution and storage and proper preparation before administration. There is also the need for maintaining a physical distance in the clinic for infection control, as well as a need for verifying patient eligibility and vaccination status via immunization information system reporting for safe vaccine administration.
One of the primary goals of PHD-led vaccination clinics is to meet the high demand for vaccination by having a high throughput of patients. Clinic planners draw on a variety of resources to design and run clinics to achieve this goal. In the United States, the Centers for Disease Control and Prevention (CDC) provides guidance on how to plan and prepare for these types of clinics [1]. To adhere to local population needs, clinic preparation decisions are made at the state and local levels with guidance from the CDC [5]. Some PHD-led public health clinics benefited from the innovation and expertise from private-sector partnerships. For example, Washington State partnered with Starbucks, Microsoft, and Costco to enhance operations and build a mock vaccination site to test flow and identify bottlenecks [2].
Limited publications have highlighted the success of vaccination clinic sites [6], [7]. A recent study reviewed previous experiences of PHD-led vaccination centers (primally influenza) to answer questions on how to organize a PHD-led vaccination center during the COVID-19 pandemic, highlighting the most important organizational aspects that should be considered while planning [8]. Although these resources can be beneficial to planners, none of them provide quantitative evidence of specific characteristics across multiple clinics that foster high throughput.
Our study aims to fill this gap by collecting data on the time spent at clinics for 521 patients in 23 COVID-19 vaccination clinics during April and May of 2021, to examine the association between clinic throughput time and the characteristics of clinics. This information will be useful for public health planners and policymakers to consider as they create plans and infrastructure to respond to future pandemics.
2 Methods
2.1 Data collection
Data collection relied on PHD’s willingness and availability of the clinic sites. Twenty-four state and local public health departments that were supporting COVID-19 vaccination clinics were invited to participate in this study based on their interests and willingness. Of these, 17 counties in 6 states (Georgia, Illinois, Michigan, Minnesota, Nevada, and Washington) agreed to participate and support data collection from large scale COVID-19 vaccination clinics run by their respective public health agencies. Data from these clinics were collected during in-person visits to these clinics during April and May of 2021. In some counties, clinics operated in multiple locations or focused on different populations at the same location. We followed individual patients from a distance to record the time spent at each station from the time of entry to post-vaccination observation using a work study app [9]. The application, designed to collect, record and export the time data by each activity for multiple subjects, was loaded onto tablets and used for data collection. Since we did not engage in direct contact with the patients, demographic, socioeconomic or clinical data for the patients were not collected. As the post-vaccination observation time varied depending on a patient’s medical conditions and is beyond the scope of the study, it was not collected, either.
While there was variation in clinic flow across the sampled clinics, all clinics were generally divided into 4 stations: 1) check-in, where the patient was greeted and temperature was taken; 2) documentation, where patient identification was confirmed, vaccination card was provided, and the second dose appointment was scheduled; 3) vaccination, where patients were vaccinated; and 4) post vaccination observation, where patients typically waited for 15 to 30 min after the vaccination and potentially had the opportunity to ask questions regarding the vaccination or second dose appointment. The patient flow time includes time when the patient arrived at the welcome station where staff directed the patients to lines for check-in stations. At each station, the time recording started upon arrival or stopped upon departure. Time spent moving between stations and for waiting to be called was recorded as waiting time. As this study focused on what could be completely controlled by the clinic, the patient flow from outside prior to entering the clinic was not captured. For example, people arriving in large mass outside the clinic in a short time (i.e. bus load) or before the clinic opens will likely increase wait time before entering the clinic. Such is almost uncontrollable and likely not modifiable by the clinic, thus was beyond our scope. Information from clinic managers on clinic operations including clinic hours, number and types of COVID-19 vaccines offered, number of vaccination stations, and number of staff was gathered. We also integrated state and county characteristics such as county population, National Center for Health Statistics (NCHS) urban–rural designation, medically underserved areas, healthcare professional shortage area indicator, states’ COVID-19 vaccination eligible population, and other public health measures [10], [11], [12], [13], [14]. This study was reviewed by the RTI Institutional Review Board and determined not to be human subjects research and had the public health emergency waiver for data collection at the clinics.
2.2 Data analysis
Patient flow time was calculated from check-in to vaccination using the collected time data using STATA (Release 16) [15]. Time at each station and waiting time between stations was included to reflect the flow efficiency of the vaccination clinic. We considered the data as repeated measure of the clinic performance. Thus, mean, median, and range of patient flow time was examined across the clinics. A multivariate linear regression model was conducted against the mean patient waiting time for each clinic to evaluate influential factors controlling the clinic settings such as staffing and patient throughput with the state fixed effects and the nested effects of clinic and county. All statistical data analyses were performed using SAS Enterprise Guide (Ver. 7.1) [16].
3 Results
We visited 23 community-based clinics held in 17 different counties in 6 states. All 23 vaccination clinics were open to anyone aged 16 years and over. Sixteen of the 23 clinics were held in medically underserved areas with limited access to primary care, as designated by Health Resources & Services Administration (HRSA)[14]. Six clinics were held in rural areas. During the site visits, clinic time was measured for 529 patients. Data for this analysis was used from 521 patients, after reviewing the data and notes from the field to exclude outliers with extreme values, possibly due to errors during data collection or missing values due to lost contact with the patients. Clinics on the day of the site visit were open for an average of 6.6 h (range 2 – 8.5) and 931 patients were vaccinated on average each day (range 150 – 2,607) (Table 1 ). The average number of total staff was 61, including 21 medical staff and 15 vaccinators, on average. The number of staff increased with the size of the clinics. Fourteen of the 23 clinics offered only one brand of mRNA COVID-19 vaccine – either Pfizer-BioNTech or Moderna. Clinics offering more than one brand of vaccines deployed 21 more staff members on average, including 10 more medical staff on average. There were 9 appointment-only based clinics, and 14 clinics that accepted walk-ins in addition to appointments. On average, the appointment-only clinics were open 1.3 h longer with 32 more staff to vaccinate more than 1,000 people on average per clinic. Twelve clinics had dedicated accommodations for patients with mobility challenges such as dedicated line for people with disabilities. The clinics with dedicated lines for people with disabilities utilized 8 more staff, on average. Drive-through clinics vaccinated more than 1,000 patients with fewer medical staff (=16), yet with more supporting staff (=44) compared to non-drive-through clinics with 22 medical staff and 39 supporting staff. Eighteen clinics were administering the first vaccine dose, which appeared to need more staff in total (66 vs 43). Eleven clinics used quick-response (QR) codes for appointments and registration. For the 12 clinics that did not use QR codes for appointments, the average operation hours (=7 h) and the number of total staff (=75) were higher than those clinics that used QR codes (6.2 h and 46 total staff members, respectively).Table 1 Characteristics of Public Health Department-led COVID-19 Vaccination Clinics in 23 sites, United States, April and May 2021.
Characteristics Number of Clinics Average Operating Hours Average Number of
Patients with Appointment Average Doses Administered Average Number of
All Staff Average Number of
Vaccinators Average Number of
Medical Staff
Overall 23 6.6 941 931 61 15 21
Clinic Hours
< 5 h 6 3.9 376 364 19 7 9
5 ∼ 9 h 17 7.6 1,140 1,132 76 19 25
Clinic size by patient volume on day of visit
Super-Large (>1200) 7 7.6 1,836 1,835 122 32 42
Large (600–1200) 6 7.1 852 818 53 11 13
Medium (350–599) 5 6.0 516 494 24 6 13
Small (<3 5 0) 5 5.2 218 239 22 7 10
Clinic size by number of vaccinators
Large (≥11) 9 7.7 1,653 1,650 109 28 36
Medium (7 ∼ 10) 6 6.0 585 562 41 9 13
Small (1 ∼ 6) 8 5.9 406 400 22 6 10
Different COVID-19 vaccines offered
Yes 9 6.8 918 904 74 20 27
No 14 6.5 955 949 53 13 17
Appointment requirement
Appointment Only 9 7.4 1,122 1,091 80 15 22
Appointment and Walk-ins 14 6.1 824 829 48 15 20
Dedicated lines for people with disabilities
Yes 12 6.9 984 910 64 14 21
No 11 6.3 893 955 57 17 21
Drive-through
Yes 5 6.8 988 1,031 60 13 16
No 18 6.6 928 904 61 16 22
Second-dose only
Yes 5 6.9 1,090 1,057 43 14 19
No 18 6.5 899 896 66 16 21
QR code use for registration
Yes 11 6.2 824 805 46 11 16
No 12 7 1,048 1,047 75 19 25
The mean time for a patient to pass through the clinic, from the welcome station at entry to the completion of vaccination, was 9 min and 5 s (9:05) (Table 2 ). Compared to the average time (12:18) for the small-size clinics (<350 patients per day), larger clinics had lower per patient flow times, particularly the large-size clinics (600–1,200 patients per day). Differences in mean patient flow time compared the small-size clinics were statistically significant. However, the super-large-size clinics (>1,200 patients per day) resulted in higher patient flow times (10 min) than medium or large-size clinics (9:29 and 6:19, respectively). We also found higher and statistically significant average patient flow time for the clinics accepting no-appointment compared to by-appointment (9:50 vs 7:25), the drive-through clinics compared to non-drive-through clinics (12:42 vs 8:40), first-dose clinics (10:18 vs 6:11), and the clinics with dedicated lines or services for people with disabilities (10:05 vs 8:22). The higher average patient flow time was also observed in the clinics offering more than one COVID-19 vaccine brand, and non-QR code clinics, but the differences were not statistically significant.Table 2 Descriptive Statistics of Patient Flow time Spent in Clinic at 23 COVID-19 Public Health Department led Vaccination Sites, United States, April and May 2021.
Characteristics Number of Clinics Number of Observations Time from Entrance to Vaccination (in minutes/seconds)
Mean Median Minimum Maximum
Overall 23 521 09:05 07:22 02:00 29:39
Clinic Hours
< 5 h 6 112 09:47 02:33 02:00 29:17
5 ∼ 9 h (referent) 17 409 08:54 01:49 02:20 29:39
Clinic size by patient volume on day of visit
Super-Large (>1200) 7 162 10:00* 08:17 02:29 29:17
Large (600–1200) 6 166 06:19* 05:13 02:20 27:30
Medium (350–599) 5 110 09:29* 08:10 02:00 29:26
Small (<3 5 0) (Referent) 5 83 12:18 09:27 02:17 29:39
Clinic size by number of vaccinators
Large (≥11) 9 221 08:39* 07:02 02:29 29:17
Medium (7 ∼ 10) 6 146 07:22 05:46 02:17 29:17
Small (1 ∼ 6) (Referent) 8 154 11:20 10:01 02:00 29:39
Different COVID-19 vaccines offered
Yes 9 238 09:29 08:01 02:00 28:21
No (referent) 14 283 08:45 06:34 02:17 29:39
Appointment requirement
Appointment Only 9 163 07:25* 05:11 02:26 28:31
Appointment and Walk-ins (referent) 14 358 09:50 08:09 02:00 29:39
Dedicated lines for people with disabilities
Yes 12 218 10:05* 08:44 02:26 29:17
No (referent) 11 303 08:22 06:40 02:00 29:39
Drive-through
Yes 5 54 12:42* 10:25 02:26 29:17
No (referent) 18 467 08:40 07:11 02:00 29:39
Second-dose only
Yes 5 154 06:11* 04:58 02:20 21:58
No (referent) 18 367 10:18 08:20 02:00 29:39
QR code use for registration
Yes 11 206 08:55 06:54 02:26 29:39
No (referent) 12 315 09:11 07:29 02:00 28:31
*Statistically significant at 5% level.
Factors associated with the patient waiting time in the clinic were evaluated using a multivariate regression model (Table 3 ). The clinic characteristics associated with more waiting time (i.e., positive parameter estimates) were the number of total vaccinations given, the presence of dedicated lines for people with disabilities, and the rural area indicator. Vaccination-related characteristics such as more than one brand of COVID-19 vaccines offered, and clinics where the vaccine was drawn by the vaccinator were associated with longer patient waiting time. The presence of an on-site translator was associated with shorter patient waiting times.Table 3 Factors associated with per patient waiting time in the clinics - A linear regression model estimation.
Variables Estimate Standard Error Pr>|t|
Number of Vaccinated Doses administered 0.003 0.001 0.031*
Percentage of Patients Getting their First Dose 0.916 1.603 0.593
Appointment requirement
Appointment Only 3.036 1.244 0.059
Appointment and Walk-ins Referent
Drive-through Clinic
Yes −1.124 1.143 0.371
No Referent
QR code use for registration
Yes −0.017 0.831 0.985
No Referent
Dedicated lines for people with disabilities
Yes 1.826 0.701 0.048*
No Referent
Temperature Check
Yes −1.930 2.156 0.412
No Referent
ID Check
Yes 0.536 2.027 0.802
No Referent
Translators on site
Yes −7.171 1.928 0.014*
No Referent
Different COVID-19 vaccines offered
Yes 1.411 0.467 0.029*
No Referent
Vaccine drawn by vaccinators
Yes 15.341 1.898 0.001*
No Referent
Number of vaccinators
Large (≥11) −2.462 2.552 0.379
Medium (7 ∼ 10) 1.028 1.187 0.426
Small (1 ∼ 6) Referent
Urbanicity**
Rural 4.197 0.557 0.001*
Urban Referent
*Statistically significant at 5% level; ** National Center for Health Statistics Urban-Rural Indicator was used.
4 Discussion
During the COVID-19 pandemic, PHD-led COVID-19 vaccination clinics served as convenient locations where community residents could easily access a COVID-19 vaccine regardless of health insurance coverage [17]. Patient flow time at vaccination clinics is an important piece of information in evaluating a clinic’s efficiency at achieving the primary goal of vaccinating as many eligible patients as possible with the available resources. However, while there have been a few studies on cost and capacities of PHD-led vaccination clinics for the preparedness exercise operations or 2019 pandemic influenza responses [4], [18], [19], [20], [21], [22], [23], [24], to our best knowledge, there have been few studies to measure the patient-time for vaccination during an ongoing public health emergency.
Before the COVID-19 pandemic, literature on PHD-led vaccination programs and clinics has been reported as a part of preparedness exercises or for the 2009 H1N1 pandemic influenza response. While most reports presented the vaccination doses per hours in PHD-led vaccination settings, to our knowledge, there are no reports on patient flow times based on direct observation [25]. Although the number of doses per hour per vaccinator or clinic is an important piece of information in planning and evaluating PHD-led vaccination clinics, patient flow time and patient waiting times capture another aspect of efficiency. For example, the total number of doses administered depends on the number of patients coming for vaccination during clinic operation hours and numbers of vaccinators and support staff. If the patient turnout were less than expected, then total number of doses per hour per clinic may underestimate the efficiency and capability of the clinics. In this sense, the patient throughput time may capture the actual time spent in the clinics to measure the efficiency of the clinic operations with a certain level of staffing. Furthermore, patient waiting times reflect not only the efficiency of the clinic layouts and staffing but also impacts patient satisfaction while in the clinics [26], [27].
This study documents that a patient spends approximately 9 min on average getting vaccinated at PHD-led vaccination clinics with a large range anywhere from 2 min to 29 min. Clinic size measured by the number of patients was associated with patient flow time in the clinics with the patient flow time decreasing by the patient volume up the point where the clinics serving 1,200 patients. Also, average patient flow times varied significantly by clinic characteristics such as by-appointment requirement for vaccination, dedicated lines or services for people with mobility challenges, or drive-through vaccination. Patient-time was the highest in small clinics with 6 or fewer vaccinators, and lower in clinics with 7 or more vaccinators. Longer patient waiting times were significantly associated with rural areas.
A notable finding is that regression analysis of patient waiting time showed that such variations were most significantly associated with clinic characteristics related to the vaccination station, such as whether the vaccine was drawn by vaccinators, or more than one COVID-19 vaccine was offered. This may imply that in order for patient flow control measures such as QR codes to be effective, vaccination stations could be managed efficiently to reduce the patient waiting times. Indeed, at the time of data collection, most clinics were just opened to the populations eligible for COVID-19 vaccination, which requires more time and staff experience. Over time, the use of QR codes may improve efficiency as more patients become familiar with them. However, vaccination stations must be staffed with medical staff with the appropriate qualifications, trainings and/or supervision. It is critical to identify key vaccine providers and prepare them for efficient implementation of PHD-led vaccination by periodic training and preparedness exercises. Furthermore, we observed many occasions that the drawing of vaccine was conducted in a separate area by other medical personnel in spite of the CDC guidance recommending that the same vaccinator to draw up the vaccine before vaccination [28]. Further studies are warranted to evaluate the impact of such practice on the mass vaccination operations in terms of safety as well as efficiency.
Presence of translators on site was associated with lower waiting time. Through translators, clinics can offer patients necessary help to streamline the vaccination process. While most COVID-19 related information is available in multiple languages [29], [30], staff who are proficient in health communication in languages other than English are critical for patient education, clinic efficiency, and reducing barriers to vaccination.
The majority of COVID-19 vaccination clinics were run by appointments in 30-minute intervals to check vaccine supply and prepare the doses, as well as to enable social distancing. However, the regression analysis found that a prior appointment requirement for vaccination was one factor associated with longer patient waiting time, which may have been caused by simultaneous arrival of patients around their appointment time or patients arriving early. It is not necessarily contradictory that appointment-only clinics had a shorter patient time while the prior appointment is a factor associated with the longer waiting time because other factors might contributed to reducing the overall patient flow time (such as shorter process time, the clinics operations, or fewer patients for vaccination). In addition, other circumstances such as temporary holdup for staff shift change or vaccine lot switches may have caused delays. Because some wait time is generally unavoidable, patients can be otherwise occupied at clinics while they wait. Wait time can be spent answering questions about vaccination or reviewing educational materials on other public health services[27]. To improve vaccination output and to reduce patient waiting time, reviews of staffing needs and further evaluations of resource allocation plans as well as patient flow could be considered[31].
4.1 Limitations
This study is subject to several limitations. First, due to convenience sampling and small sample size, the results may not be generalizable. Second, to minimize contact with patients and maintain privacy, patient characteristics such as disability status or limited English proficiency were not collected for the analysis. As a result, we could not consider any individual effect on longer or shorter patient flow time. Third, COVID-19 vaccination demand had begun to decrease towards the end of the study. Therefore, clinic operations before or after the study period may have differed from what was observed during the study. Fourth, a few clinics we visited were planning to offer the Janssen vaccine, targeting populations that may face barriers to returning for a second dose, such as migrant farm workers, those experiencing homelessness, or those living in remote areas. However, at the time of our site visits, almost all vaccination clinics were only providing the two-dose mRNA vaccines due to the temporary pause of the Janssen COVID-19 vaccine after rare reports of thrombosis with thrombocytopenia syndrome among vaccine recipients [32]. Therefore, our results may not be directly comparable to the time studies conducted with Janssen COVID-19 vaccination clinics or one-dose seasonal or pandemic influenza vaccination sites.
5 Conclusions
In order to vaccinate as many people as possible in a short period of time, efficient flow in vaccination clinics is crucial. Given the standardized procedures outlined by CDC guidelines and state health departments, reducing wait time is critical to lowering the patient flow time by relieving bottleneck effects in the clinic. While this study was conducted with a relatively small sample, it could still provide baseline estimates of patient flow time spent in PHD-led vaccination clinics. This study suggests enhancing the efficiency of the PHD-led vaccination clinics by preparing more vaccinators to provide vaccines with proper and timely support such as training or delivering necessary supplies and paperwork to the vaccinators. When planning mass vaccination clinics, strategic staffing and resource allocation would be crucial to address the needs of the community of focus as well as to build the capacity of public health preparedness.
6 Disclaimer
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: ‘Benjamin O Yarnoff and Laurel G Bates reports financial support was provided by Centers for Disease Control and Prevention.’
Data availability
Data will be made available on request.
Acknowledgements
The authors would like to thank participating jurisdictions’ immunization information system (IIS) agencies and participating vaccination clinics as well as public health departments for the opportunities to collect the time data used in this study. Also, we are grateful for the insightful comments and suggestions made by Sam Graitcer, James Tseryuan Lee, Erin Kennedy, Cindy Weinbaum, Lynn Gibbs-Scharf and Georgina Peacock. In addition, we thank Zohra Tayebali for her assistance with data management and analysis.
Financial Support
Data collection was conducted in collaboration with RTI International, under contract 200-2013-M-53964B to the Centers for Disease Control and Prevention.
==== Refs
References
1 Centers for Disease Control and Prevention. Guidance for Planning Vaccination Clinics Held at Satellite, Temporary, or Off-Site Locations. In: National Center for Immunization and Respiratory Diseases, editor; 2020.
2 Goralnick E. Kaufmann C. Gawande A.A. Mass-vaccination sites — an essential innovation to curb the Covid-19 pandemic N Engl J Med 384 2021 e67 33691058
3 Porter D. Hall M. Hartl B. Raevsky C. Peacock R. Kraker D. Local health department 2009 H1N1 influenza vaccination clinics—CDC staffing model comparison and other best practices J Public Health Manag Pract 17 2011 530 533 21964365
4 Klaiman T, Oʼconnell K, Stoto M. Local health department public vaccination clinic success during 2009 pH1N1. J Public Health Manag Pract. 2013;19:E20–6.
5 Fitzgerald T.J. Moulia D.L. Graitcer S.B. Vagi S.J. Dopson S.A. 2015 pandemic influenza readiness assessment among US public health emergency preparedness awardees Am J Public Health 107 2017 S177 S179 28892450
6 Moyce S. Ruff J. Galloway A. Shannon S. Implementation of a COVID-19 mass vaccination clinic to college students in Montana Am J Public Health 111 2021 1776 1779 34499538
7 Andrade J. Slaby M. DeAngelis J. Connors J. Truong J. Ciaramella C. Implementation of a pharmacist-led COVID-19 vaccination clinic at a community teaching hospital Am J Health Syst Pharm 78 2021 1038 1042 33772261
8 Gianfredi V. Pennisi F. Lume A. Ricciardi G.E. Minerva M. Riccò M. Challenges and opportunities of mass vaccination centers in COVID-19 times: a rapid review of literature Vaccines 9 2021
9 Tomlinson B. Work Study for Industrial/Production Engineers. 2.0 ed2020.
10 Health Resources and Services Administration. Area Health Resources Files - 2019-2020 County Level Data. July 20, 2020 ed2019-20.
11 Centers for Disease Control and Prevention. U.S. State and Territorial Public Mask Mandates From April 10, 2020 through August 15, 2021 by County by Day. Policy Surveillance. September 10, 2021 ed. Atlanta, GA: Environmental Public Health Tracking; 2021.
12 US Census Bureau. County Population Totals: 2010-2020. Annual Resident Population Estimates for States and Counties. May 2021 ed2021.
13 Foundation KF. State COVID-19 Data and Policy Actions. 2021.
14 Health Resources & Services Administration. Shortage Areas - Medically Underserved Areas/Populations (MUA/P). 2021 ed2020.
15 Corp S. STATA Statistical Software (Release 19) 2019 StataCorp LLC College Station, TX
16 SAS Institute Inc. SAS Enterprise Guide (Ver. 7.1). Cary, NC, USA.
17 Prosser L.A. O’Brien M.A. Molinari N.-A.-M. Hohman K.H. Nichol K.L. Messonnier M.L. Non-traditional settings for influenza vaccination of adults Pharmacoeconomics 26 2008 163 178 18198935
18 Georgia's East Metro Health District (EMHD). Mall Clinics Provide H1N1 Vaccine to a Diverse Population. Center for Infectious Disease Research and Policy; 2009.
19 Kansagra S.M. McGinty M.D. Morgenthau B.M. Marquez M.L. Rosselli-Fraschilla A. Zucker J.R. Cost comparison of 2 mass vaccination campaigns against influenza A H1N1 in New York City Am J Public Health 102 2012 1378 1383 22676501
20 Saha S. Dean B. Teutsch S. Borse R. Meltzer M. Bagwell D. Efficiency of points of dispensing for influenza A(H1N1)pdm09 vaccination, Los Angeles County, California, USA, 2009 Emerg Infect Disease J 20 2014 590
21 Banks L.L. Crandall C. Esquibel L. Throughput times for adults and children during two drive-through influenza vaccination clinics Disaster Med Public Health Prep 7 2013 175 181 24618169
22 Pérez Velasco R. Praditsitthikorn N. Wichmann K. Mohara A. Kotirum S. Tantivess S. Systematic review of economic evaluations of preparedness strategies and interventions against influenza pandemics PLoS One 7 2012 e30333 22393352
23 Swift M.D. Aliyu M.H. Byrne D.W. Qian K. McGown P. Kinman P.O. Emergency preparedness in the workplace: the flulapalooza model for mass vaccination Am J Public Health 107 2017 S168 S176 28892449
24 Yarnoff B, Pike J, Athar H, Bates L, Tayebali Z, Harris L, et al. Assessment of the costs of implementing COVID-19 vaccination clinics in 34 sites, United States, March 2021. J Public Health Manage Practice. Forthcoming.
25 Yarnoff B. Bodhaine S. Cohen E. Buck P.O. Time and cost of administering COVID-19 mRNA vaccines in the United States Hum Vaccin Immunother 17 2021 3871 3875 34613860
26 Chadha M, Dunne A, Rodriguez NC, Rahul;. COVID-19 and clinic workflow optimization using lean six sigma. Am J Managed Care 2021;27.
27 Thompson D.A. Yarnold P.R. Williams D.R. Adams S.L. Effects of actual waiting time, perceived waiting time, information delivery, and expressive quality on patient satisfaction in the emergency department Ann Emerg Med 28 1996 657 665 8953956
28 Kroger A, Bahta L, Hunter P. General Best Practice Guidelines for Immunization: Best Practices Guidance of the Advisory Committee on Immunization Practices (ACIP). Updated March 15, 2022.
29 Food and Drug Administration. Multilingual COVID-19 Resources. 2021.
30 Centers for Disease Control and Prevention. CDC Resources in Languages Other than English. 2022.
31 Centers for Disease Control and Prevention What to consider when planning to operate a COVID-19 vaccine clinic 2021 Centers for Disease Control and Prevention Atlanta, GA
32 MacNeil J.R. Su J.R. Broder K.R. Guh A.Y. Gargano J.W. Wallace M. Updated recommendations from the advisory committee on immunization practices for use of the Janssen (Johnson & Johnson) COVID-19 vaccine after reports of thrombosis with thrombocytopenia syndrome among vaccine recipients - United States, April 2021 MMWR Morb Mortal Wkly Rep 70 2021 651 656 33914723
| 0 | PMC9742208 | NO-CC CODE | 2022-12-14 23:54:01 | no | Vaccine. 2022 Dec 12; doi: 10.1016/j.vaccine.2022.12.013 | utf-8 | Vaccine | 2,022 | 10.1016/j.vaccine.2022.12.013 | oa_other |
==== Front
Journal of Hospitality and Tourism Management
1447-6770
1447-6770
The Authors.
S1447-6770(22)00197-8
10.1016/j.jhtm.2022.12.010
Article
Losing talent due to COVID-19: The roles of anger and fear on industry turnover intentions
Popa Iuliana a
Lee Lindsey b
Yu Heyao c
Madera Juan M. a∗
a Conrad N. Hilton College of Global Hospitality Leadership, University of Houston, 4450 University Drive, Room 235, Houston, TX, 77204-3028, United States
b School of Sport, Tourism and Hospitality Management, Temple University, 1810 North 13th Street, Philadelphia, PA, 19122, USA
c School of Hospitality Management, Pennsylvania State University 230 Mateer Building, University Park, PA, 16802, United States
∗ Corresponding author.
12 12 2022
3 2023
12 12 2022
54 119127
31 5 2022
7 11 2022
9 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.
Early in the COVID-19 pandemic, the US hospitality industry workforce experienced significant job loss via furloughs and job eliminations. Over a year later, the American hospitality industry is now facing a labor shortage. However, there is a dearth of literature explaining why the hospitality industry's response due to a mega-event, like the pandemic, can motivate employees to leave the hospitality industry. Instead, theory and research have primarily focused on organizations as the focal point for understanding turnover, while neglecting the industry. Using the affect theory of social exchange, this paper examined how anger and fear related to job status changes (i.e., being furloughed or laid-off) due to the pandemic, influence intentions to leave the industry. Study 1 used a survey of management-level employees, whereas Study 2 used an experiment to test the proposed model. Both studies showed that employees who lost their job due to the pandemic felt more anger and fear than those still employed. However, mediation analyses revealed anger, but not fear, as the primary driver of industry turnover intentions. These results highlight a potentially problematic trend. Should skilled hospitality workers switch industries due to job loss amidst an industry-wide negative event, it may become difficult for hospitality businesses to find qualified employees once the industry recovers and rehiring begins.
Keywords
COVID-19
Industry turnover intensions
Negative emotions
Job loss
Turnover
==== Body
pmc1 Introduction
Since March of 2020, the US hospitality industry workforce, particularly in the lodging and food and beverage sectors, has experienced significant furloughs and job loss due to the COVID-19 pandemic (Gursoy & Chi, 2020; U.S. Bureau of Labor Statistics, 2020). Travel and dining out declined greatly, placing a financial strain on the industry and threatening the job security of millions (Nicola et al., 2020). While many companies from other industries have adapted their operating style to accommodate social distancing measures, hospitality companies face greater challenges in adapting to remote work and service delivery methods (Baum et al., 2020). In the lodging sector, approximately 1.6 million workers have lost their jobs, with an average vacancy of eight out of ten hotel rooms during 2020 (American Hotel & Lodging Association, 2020). The food and beverage sector was especially hard hit early on, with two-thirds of employees losing their jobs during the first two months of the pandemic, amounting to over 8 million jobs lost (National Restaurant Association, 2020).
With the rollout of vaccinations, the hospitality industry is slowly regaining thousands of jobs; but the industry is now facing a labor shortage (Picchi, 2021). Employers are struggling to attract talent—both who were furloughed or laid-off due to the pandemic—back to the industry despite offering signing bonuses and competitive benefits. For example, Taco Bell is offering paid family leave to company store managers and Jimmy Johns and other fast-food restaurants are offering $250 signing bonuses (Haddon, 2021). However, many hospitality industry employees have left for jobs in other industries, particularly ones that were less affected by the pandemic and were able to adjust to social distancing and remote work (Wiener-Bronner, 2021; Yu et al., 2021). Although our current study focuses on the US workforce, these trends have also been found in Australia (Powell, 2022), Asia (TTG Asia, 2021), and Europe (Diazgranados, 2021).
Despite these realities, there is a dearth of literature surrounding the mechanisms that explain why the industry's response to a mega-event, such as the COVID-19 pandemic, can motivate employees to leave the hospitality industry. Event-system theory (Morgeson et al., 2015) states that mega-events are unexpected, external disasters that require organizational actions. Accordingly, mega-events can have a negative impact on employees' emotions, attitudes, and behaviors via organizations' actions, such as layoffs and furloughs. While research shows that employment decisions (e.g., layoffs and furloughs) do indeed elicit negative emotions (e.g., Grandey et al., 2021; Huffman et al., 2022; Shepherd & Williams, 2018), a question remains of whether these emotions trigger negative attitudes toward not only the organization that made them unemployed, but to the industry they work in. In other words, although it is expected that employees will experience negative emotions when they are made unemployed, research has primarily focused on how employees experience negative emotions within the context of the specific organization that made them unemployed. However, it is not clear whether these emotions will also be felt targeting the industry they work in. The hospitality industry's response to the COVID-19 pandemic provides a unique context to examine this research gap.
First, the hospitality industry was hit the hardest by the pandemic. Approximately two-thirds of jobs lost in the US were from the hospitality industry (U.S. Bureau of Labor Statistics, 2020), thereby placing a negative spotlight on the hospitality industry as an employer. Second, the hospitality industry relies on a pipeline of talent with industry-specific education, competencies, and skills that are transferable across organizations within the industry, thereby requiring talent to remain within the industry (King et al., 2021). Third, talent in the industry is ingrained in the industry. Specifically, the hospitality industry commonly requires a diversity of frontline, operational experience, across organizations within the industry for supervisory and management positions (Suh et al., 2012). This necessitates that talent stays within the industry and the industry needs to retain talent with the relevant skills and experiences to maintain a pipeline for management positions. Thus, hospitality organizations have invested time and resources on training management-level employees, who likewise, have invested their time working in the industry. It is therefore important to understand why hospitality employees, particularly those in management roles, leave the industry during the COVID-19 pandemic.
To address this gap in the literature, our study is guided by the affect theory of social exchange (Lawler, 2018), which provides insight on why hospitality employees might leave the industry when they experience furloughs or job loss during the pandemic. This theory contends that exchanges at work almost invariably produce positive or negative emotions. In the context of this study, employees who were furloughed or laid-off will most likely feel stronger negative emotions than employees who were able to work during the pandemic. Although the affect theory of social exchange (Lawler, 2018) limits its scope within organizations, the current research expands on this theory by examining how in exchange for being laid-off or furloughed, and subsequently feeling negative emotions to this action, employees will be motivated to leave the industry. We test our hypotheses by leveraging a combination of methods and samples. Study 1 uses a survey with management-level employees, whereas Study 2 uses an experiment to further test our proposed model using a sample of aspiring entrants into hospitality management (i.e., management in training sample of hospitality management students).
The current study makes several contributions to the literature on understanding why hospitality employees leave the industry. First, event-system theory (Morgeson et al., 2015) limits its focus on organization-focused attitudes and behaviors, despite the fact that industries can face mega-events that negatively impact them more than other industries, such as the COVID-19 pandemic having the greatest impact on the hospitality industry in regard to furloughs and job loss (U.S. Bureau of Labor Statistics, 2020). We expand on this theory by showing that the industry can also be the target of negative emotions related to a mega-event. Likewise, we expand on the theory of social exchange (Lawler, 2018) by showing that negative emotions felt by decisions made by organizations can also motivate employees to make career changes, such as leaving the industry they work in.
Second, and related to the first point, is that much of the literature on why employees, including managers, leave the hospitality industry has focused on personal attributes, job attributes, and organizational-specific attitudes, such as job satisfaction, internal motivation, perceived organizational support, and psychological contracts (e.g., Ann & Blum, 2020; Blomme et al., 2010; Brown et al., 2015; Guchait et al., 2015). In other words, the literature has primarily focused on the organization as the foci for understanding turnover, while neglecting the industry. For example, in their critique of the literature, King et al. (2021) argued that the difficulties the hospitality industry faces in attracting and retaining talent, particularly during the pandemic, “also seem to reside at the industry level, and not solely at the organizational level where most research is focused” (p. 252). Therefore, the current paper focuses on industry turnover.
2 Theoretical background and hypotheses
2.1 Job status and feeling anger and fear
The current paper uses the affect theory of social exchange (Lawler, 2001) as the guiding framework. This theory argues that emotions are produced when two or more entities exchange valued outcomes (e.g., payment, rewards, or goods for work). Entities can include people and/or social units, such as organizations they work for. Accordingly, after exchanges, employees process information and interpret intentions from their organization, and emotionally respond to an exchange. The emotional reactions can involve positive or negative emotions, depending on the exchange (e.g., positive emotions like pride after a promotion or negative emotions like fear after a furlough). Per this theory, these emotional reactions also lead to attributions, in which emotions are linked to people or social unites. For example, employees who feel negative emotions, like fear after a furlough, will then attribute these emotions to their organizations. Thus, this theory provides an important lens through which to understand how the hospitality industry response to the COVID-19 pandemic can drive industry turnover.
One of the broadest effects of the COVID-19 pandemic on employees’ jobs is their job status. In the early months of the pandemic, many jobs were changed to a remote format, hours were reduced, job tasks were modified, or they were eliminated temporarily through furloughs or indefinitely through layoffs (Brynjolfsson et al., 2020). Current literature further indicates that COVID-19 has created an environment where hospitality employees who experienced job insecurity had lower work motivation (Bajrami et al., 2021) and those who are high in career adaptability are particularly prone to developing high industry turnover intentions in situations of low supervisor support (Lee et al., 2021), indicating a high-stakes situation for the hospitality industry in terms of talent loss.
In addition, as reported by Wong et al. (2021, p. 102798), employee “job satisfaction, organizational commitment, job performance, subjective well-being, and prosocial behavior had each significantly decreased after the pandemic took hold, whereas turnover intention was significantly higher after COVID-19 had become quite prevalent,” demonstrating a widespread negative impact of COVID-19 on employee well-being, with negative implications for the hospitality industry in turn. Further contributing to these indications of COVID-19 as a deeply negative, impactful crisis for the hospitality industry, a study by Yan et al. (2021) has found that lower job satisfaction strengthens the relationship between employee perceptions of COVID-19 risk perception and experiencing depressive symptoms.
Thus, given the unprecedented hospitality industry job loss and rapid change in operations due to the pandemic (US Bureau of Labor Statistics, 2020), employees are prone to experiencing significant levels of negative emotions (Mimoun et al., 2020). Although there are a number of negative emotions that employees can feel, Lebel's (2017) model of negative emotion regulation points to fear and anger as particularly relevant for the current study. Whether laid off, furloughed, or still employed, hospitality employees are facing substantial uncertainty, which can result in negative emotions such as fear and anger. For example, managers still employed might feel fear from the uncertainty of working during a pandemic, or anger with the direction or lack of guidance provided by their organization (Chen & Eyoun, 2021; Guzzo et al., 2021).
Anger is defined as feeling indignation with desires to redress a perceived wrongdoing or violation of a perceived contract (Greenbaum et al., 2020); whereas fear is defined as feeling uncertainty and a lack of control or efficacy (Osborne et al., 2012). Anger and fear are both negatively valanced, high-arousal, discrete emotions and have been found to have strong counterproductive effects. For instance, aggression and other deviant behavior are common responses to anger (Fox & Spector, 1999), whereas silence on important matters is often a result of fear (Kish-Gephart et al., 2009).
Furthermore, anger and fear are often felt when losing important resources, such as one's job, whether it is through layoffs or furloughs (Pugh et al., 2003; Osborne et al., 2012). As suggested by affect theory of social exchange (Lawler, 2018), discrete emotions, such as anger and fear, are felt as a result of the decisions organizations make that have implications to one's well-being. For example, one's employment is an exchange for the time and effort one gives to their organization. When an organization makes an employee unemployed, whether from a layoff or furlough, the employee might see this as a violation of the time and effort they gave their organization.
In addition, losing one's job is a form of relative deprivation and can lead to viewing one's financial circumstances as undeserved and worse in comparison to coworkers who remained employed, which often results in feeling negative emotions such as anger and fear (Smith & Pettigrew, 2014). One's job is not only related to financial resources, but also to one's identity, job satisfaction, sense of accomplishment, and affiliation (Miscenko & Day, 2016). Therefore, losing one's job not only deprives oneself of financial resources, but also of resources that one gains through work. Such deprivation can lead to feelings of anger, due to a perceived wrongdoing or violation of trust on behalf of their organization, and fear, due to the uncertainty for one's financial future. Thus, managers who lost their jobs due to the pandemic are more likely to feel anger and fear relative to managers still employed in the hospitality industry.H1a Unemployed and furloughed managers will feel greater anger than managers working during the pandemic.
H1b Unemployed and furloughed managers will feel greater fear than managers working during the pandemic.
2.2 Hospitality industry turnover: the mediating effects of anger and fear
The affect theory of social exchange (Lawler, 2018) suggests that negative emotions felt due to an exchange—being unemployed by one's organization—can decrease future exchanges or terminate a relationship with an organization. The current paper advances this theory by suggesting that the employees who feel negative emotions, like anger or fear, might attribute these emotions to working in an industry that was vulnerable to the pandemic. In other words, equally important is that people can look beyond one's organization as the cause of the negative emotions. For example, managers who were laid-off or furloughed by their hospitality organizations are not only likely to feel anger and fear, but also to attribute these emotions to working in the hospitality industry. In response to losing one's job, managers might terminate their relationship with working in the hospitality industry due to the anger and fear elicited by the unemployment status. Thus, the current study examined anger and fear as mediators of the relationship between job status (i.e., employed or unemployed) and hospitality industry turnover intentions.
It is important to note here that while anger and fear may sometimes be perceived as passing emotions, they are emotional experiences that can influence our perceptions and behaviors in the long-term. Specifically, research shows that discrete emotions, such as anger and fear, influence how people appraise situations, which then influences how people judge and plan for future events (Han et al., 2007; Lerner & Tiedens, 2006). For instance, discrete emotions have been linked to helping behavior, sabotage, pro-environmental behavior, organizational commitment, and absenteeism, suggesting that discrete emotions influence long-term behaviors and decisions at work (Conroy et al., 2017; Rubino et al., 2013).
Therefore, considering the important effects anger and fear have on future behavior, it is important to examine how unemployment related to the COVID-19 pandemic is affecting hospitality employees’ industry career change intentions. Industry turnover refers to a change in an occupation that is not part of a normal career evolution (McGinley et al., 2014). For instance, a job change from hotel front desk agent to front desk supervisor is a natural career progression, whereas a change from hotel front desk agent to real estate agent is considered industry turnover. Although industry turnover intentions and organizational turnover in the hospitality industry have been the focus of previous research (Brown et al., 2015), a gap exists. Past research largely focused on the factors that contribute to hospitality industry turnover during normal times and has identified causes such as working environment and career progression (Haldorai et al., 2019). Considering the global reach and magnitude of the COVID-19 pandemic, this presents a unique chance to gain insight into the effects of such a crisis on hospitality industry employee turnover intentions.
Anger and fear, are negative emotions, frequently associated with counterproductive results and often act as motivators that lead individuals to taking proactive measures to remedy these feelings (Lebel, 2017). For the purposes of this study, proactive behavior can be defined as an action that is geared towards a future goal (Parker et al., 2010), which can include becoming disengaged at work or looking for alternative employment opportunities (Osborne et al., 2012). Both anger and fear have been associated with fight and flight responses, motivating individuals to be proactive in their approach (Lebel, 2017). Specifically, anger is associated with high certainty situations and motivates the one experiencing it to take corrective action, or a “fight” response (Crisp et al., 2007). Meanwhile, fear arises from a sense of uncertainty surrounding a negative event, which in turn leads to a “flight” response – a proactive effort to cut off the source of fear (Dasborough et al., 2020). In both cases, these two negative emotions incite proactive behavior, since they motivate the individual experiencing them to assess their situation and address it (Lazarus & Folkman, 1984).
In the current context—that is, the pandemic's negative effect on the hospitality industry—the anger and fear managers feel can be directed towards the industry. In other words, because the hospitality industry suffered the most furloughs and job losses during the pandemic, industry turnover intention can serve as a ‘fight and flight’ response triggered by anger and fear, respectively. Industry turnover intention can be a corrective action triggered by employees' anger because they believe working in the hospitality industry is at fault. In addition, industry turnover can be a flight response triggered by fear because employees' perceive the unstable nature of the hospitality industry as a source of their fear. Therefore, as suggested by the affect theory of social exchange (Lawler, 2018), these negative emotions might drive talent to eliminate future exchanges with the source of the negative emotions. Thus, managers who were laid-off or furloughed by their hospitality organization might attribute their anger and fear to working in the hospitality industry and therefore are motivated to leave the industry. The conceptual model is illustrated in Fig. 1 .H2a Anger will mediate the relationship between employee status (still employed or furloughed/unemployed) and career change intentions.
H2b Fear will mediate the relationship between employee status (still employed or furloughed/unemployed) and career change intentions.
Fig. 1 Conceptual model.
Note. *Employment status = still employed or furloughed/unemployed during the pandemic.
Fig. 1
3 Methodology: study 1
3.1 Data collection
Data was collected using an online survey distributed through the Amazon Mechanical Turk (or MTurk) platform. MTurk is a website that helps match workers on the platform to tasks requiring human intelligence, such as research surveys. Human intelligence task (HIT) requesters can set demographic parameters that individuals must meet in order to complete the requested HIT, such as age, HIT-performance history (past task approval rate), and location data. The data was collected in June and July of 2020, after the first wave of shutdowns.
Eligibility criteria for participants in this study included employment in the hospitality industry in management-level positions who (1) have worked in the industry for at least one year and (2) are still employed in the industry or have been laid off or furloughed due to COVID-19 within the three months of the quarantine mandates (e.g., workplace restriction orders) executed in March 2020. In addition, participants were required to be US residents and be at least 18 years or older. Those who met the criteria were asked to complete the survey. To further ensure quality responses, the survey included various “attention-check” questions throughout, such as, “Select ‘Strongly disagree’.” These questions served to drop inattentive respondents from the survey before completion. In addition, only participants with HIT approval rates above 98% and over 5000 submitted HITs were qualified. All data collection for this study was remote, and respondents were paid $1.50 for successfully completing the survey.
3.2 Participants
A total of 350 participants were surveyed, but 24 respondents were not used due to incomplete or inaccurate responses, resulting in 326 participants. In regard to employment status, 53% were currently working, and 38% were furloughed, and 9% were laid off as a result of COVID-19 workplace restriction orders; 44% were from hotel and lodging and 56% were from the food and beverage operations. The majority were White (57%), male (76%), paid a salary (89%), and had a 4-year college degree (64%) or professional degree (25%). The majority were between 25 and 34 (58%) or 35 to 44 (22%) years old. Lastly, the participants reported an average of 7.03 (SD = 6.30) years working in the hospitality industry.
3.3 Measures
Employment status. Respondents were asked to specify the effects that the pandemic had on their job, such as whether they were laid off, furloughed, or were still employed. Participants who were laid off or furloughed were coded as “unemployed” (47%) and all others were coded as still “employed” (53%). A planned contrast showed no significant differences in fear [t(323) = 0.68, p = 0.49], anger [t(323) = −0.39, p = 0.69], and industry turnover intentions [t(323) = −1.54, p = 0.12] between the ‘laid off’ and ‘furloughed’ participants, thereby providing evidence for combining both groups as “unemployed.”
Anger and fear. Fear and anger were measured using the anger and fear subscale of Izard's (1991) Differential Emotion Scale III (DES III). This was a 5-point Likert scale with answer choices ranging from “very much” to “not at all.” Participants were asked to select responses on this scale in reference to their experiences of fear, anger, and related emotions, upon learning about the effects of COVID-19 on their job. The items for fear were “scared,” “fearful,” and “afraid,” (Cronbach's alpha = 0.77) and the items for anger were “enraged,” “angry,” and “mad” (Cronbach's alpha = 0.74).
Industry turnover intention. Items by McGinley and Mattila (2020) were used to measure hospitality industry turnover intentions. Some example items include: “You think a lot about leaving the industry,” and “You are actively searching for an alternative to this industry.” Participants were asked to respond to these questions using a 5-point Likert scale, with answer choices that ranged from “strongly disagree” to “strongly agree” (Cronbach's alpha = 0.83).
Control variables. The control variables used were age and type of operation (hotel/lodging or restaurant). We controlled for age because age is negatively related to career change intentions (e.g., Carless & Arnup, 2011). In addition, we also controlled for whether the participants worked in food and beverage or hotel and lodging to control for any potential idiosyncratic differences in turnover between these two types of operations.
4 Results
4.1 Psychometric analyses
A confirmatory factor analysis (CFA) and the heterotrait-monotrait ratio (HTMT) of the correlations were used to examine the psychometric properties of the measures. A CFA of a three-factor model with fear, anger, and industry turnover intentions demonstrated adequate fit: χ2 = 39.86, df = 24, NFI = 0.97, IFI = 0.98, CFI = 0.98; RMSEA = 0.045. All loadings were statistically significant and were larger than 0.50 (they varied from 0.66 to 0.84), and the composite reliabilities were greater than the 0.70 threshold, indicating convergent validity (Hair et al., 2010). This model was compared to a one-factor-model, which demonstrated poor fit: χ2 = 398.86, df = 27, NFI = 0.62, IFI = 0.63, CFI = 0.63; RMSEA = 0.21.
As shown in Table 1 , the average variance extracted (AVE) for each measure was greater than the 0.50 cutoff (Bagozzi & Yi, 1988). In addition, the maximum shared variance (MSV) for each construct was less than the AVEs and square root of the AVEs for each measure were greater than the correlations among the measures, thereby demonstrating discriminant validity (Fornell & Larcker, 1981). Lastly, the HTMT of the correlations was used to further assess discriminant validity (Henseler et al., 2015). The results of the HTMT showed that the values ranged from 0.24 to 0.69, which were less the 0.85 threshold (Kline, 2011).Table 1 Validity results for the measures for Study 1.
Table 1 Means (SD) CR AVE MSV Fear Anger Turnover Intentions
Fear 3.53 (0.91) 0.75 0.51 0.48 0.71
Anger 3.34 (1.01) 0.75 0.51 0.48 0.69* 0.71
Turnover intentions 3.49 (1.05) 0.83 0.62 0.26 0.23* 0.51* 0.79
Note. CR = composite reliability; AVE = average variance extracted; MSV = maximum shared variance. The square root of the AVEs are in bold.
*p < 0.01.
4.2 Test of hypotheses
The hypotheses were tested using PROCESS Model 4 with two parallel mediators and a bootstrap function extracting 5,000 samples for the analysis (95% confidence interval [CI]) was used to test the conceptual model (Hayes, 2017). The results showed that unemployed participants felt more anger than employed participants (β = 0.27, p = 0.02; CI = [0.05, 0.49]). Similarly, unemployed participants felt more fear than employed participants (β = 0.10, p = 0.03; CI = [0.02, 0.42]), thereby supporting H1a and H1b.
The results for the indirect effect showed that the relationship between employment status (unemployed vs. employed) and industry turnover intentions was mediated by anger (effect = 0.11; CI = [0.02, 0.21]), supporting H2a. However, fear did not mediate the relationship between employment status (unemployed vs. employed) and industry turnover intentions (effect = −0.01; CI = [-0.05, 0.03]), which did not support H2b. Table 2 shows the results of the main and indirect effects.Table 2 Direct and indirect effect for Study 1.
Table 2 Effect SE LLCI ULCI
H1a Employment status → anger 0.27 0.11 0.05 0.49
H1b Employment status → fear 0.10 0.10 0.02 0.42
Indirect effects BootSE LLCI ULCI
H2a Employment status → anger → industry turnover intentions 0.11 0.05 0.02 0.21
H2b Employment status → fear → industry turnover intentions −0.01 0.02 −0.05 0.03
Note. SE = standard error; LLCI = lower limit confidence interval; ULCI = upper limit confidence interval.
4.3 Discussion: study 1
Study 1 examined management-level employees’ emotional reactions to how the pandemic has affected their job status, such as being laid-off or furloughed, and influences their intentions to leave the industry. The results showed that unemployed participants felt more anger and fear than employed participants. The results for the indirect effect showed that the relationship between employment status (unemployed vs. employed) and industry turnover intentions was mediated by anger but not fear.
Despite using a sample of managers with industry experience, which is a strength, Study 1 is a cross-sectional survey. Therefore, an experiment was used in Study 2 to not only address the limitations of Study 1 but also examine the causal effect of why job loss due to mega-events, such as the COVID-19 pandemic, can motivate employees to leave the hospitality industry. Specifically, in Study 2 we used an experiment to further test our proposed model with a sample of hospitality management students with hospitality work experience. Using hospitality management students with work experience in the industry allows us to further test the model with a sample who are currently investing in a hospitality degree and working in the industry, with the goal of attaining management-level positions in the hospitality industry.
5 Methodology: study 2
5.1 Sample
The target sample for Study 2 was aspiring entrants into hospitality management, therefore, we targeted hospitality management students with work experience in the industry. Hospitality management students with work experience in the industry are an appropriate sample considering that the current study is focusing on industry turnover intentions and hospitality management students are important stakeholders of the industry because they are part of a significant pipeline of hospitality management talent (King et al., 2021). This sample additionally represents stakeholders who are currently investing in a hospitality degree and working in the industry, with the goal of attaining management-level positions in the hospitality industry. The data was collected in the fall of 2021, September and October, after the first wave of available vaccinations.
A total of 150 senior-level students currently majoring in the hospitality industry with at least one year of work experience were contacted via email to participate in a study about their career interest. Of these, 104 (39% men, 61% women) completed the study. Of those currently working (81%), the majority, 69%, had an hourly, non-supervisor job and 31% had a supervisor/management level job. They had an average age of 24.35 (SD = 7.7) and an average of 4.92 (SD = 5.39) years of work experience in the industry. Most identified as Caucasian (43.4%), Asian (24.2%), Latinx (20.2%), African-American/Black (6.1%), and multiracial (6.1%).
5.2 Design and procedure
A 2-group (employee status: still employed or job eliminated) between-subjects experimental design was used. Participants read a scenario in which their job was either eliminated or not, adapted from the scenario used by Guzzo et al. (2021). They were instructed to imagine that they are attending a shift meeting with their manager at a company similar to their current or past job. For the eliminated job condition (n = 48), the participants read “You are informed that the company will be eliminating job positions due to the COVID-19 pandemic. You are told that your job at this company has been affected. Although some jobs were saved, your job has been eliminated due to the pandemic.” For the not eliminated job condition (n = 52), the participants read “You are informed that the company will be eliminating job positions due to the COVID-19 pandemic. You are told that your job at this company has not been affected. Although some jobs were eliminated due to the pandemic, you will continue to be employed despite the pandemic.” After reading the manipulated statements, the participants completed the measures.
5.3 Measures
Anger and fear. The same measures for anger (Cronbach's alpha = 0.96) and fear (Cronbach's alpha = 0.93) from Study 1 were used.
Industry turnover intention. The same measure from Study 1 was used (Cronbach's alpha = 0.86).
Control variables. We controlled for current hours working and whether respondents were furloughed or lost their job due to COVID-19 in the past. We controlled for these in case currently either working in the industry during the pandemic and/or losing a job due to the pandemic had an effect on their reactions to the manipulated scenarios.
Manipulation check. The participants were asked to rate the extent to which their job was affected by the pandemic described in the scenario using a 5-point Likert-type scale from “not at all” to “very much.” As expected, the participants’ rating was higher in the condition in which their job was eliminated (M = 4.14, SD = 1.39) than when it was not eliminated (M = 2.62, SD = 1.14), F(1, 98) = 21.61, p = 0.001.
6 Results
6.1 Psychometric analyses
A CFA and the HTMT ratio of the correlations were used to examine the psychometric properties of the measures. A CFA of a three-factor model with fear, anger, and industry turnover intentions demonstrated adequate fit: χ2 = 42.32, df = 24, NFI = 0.96, IFI = 0.98, CFI = 0.98; RMSEA = 0.09. All loadings were statistically significant and were larger than 0.50 (they varied from 0.65 to 0.96), and the composite reliabilities were greater than the 0.70 threshold, indicating convergent validity (Hair et al., 2010). This model was compared to a one-factor-model, which demonstrated poor fit: χ2 = 352.92, df = 27, NFI = 0.63, IFI = 0.65, CFI = 0.65; RMSEA = 0.34.
As shown in Table 3 , the average variance extracted (AVE) for each measure was greater than the 0.50 cutoff (Bagozzi & Yi, 1988). In addition, the maximum shared variance (MSV) for each construct was less than the AVEs and square root of the AVEs for each measure were greater than the correlations among the measures, thereby demonstrating discriminant validity (Fornell & Larcker, 1981). Lastly, the HTMT of the correlations was used to further assess discriminant validity (Henseler et al., 2015). The results of the HTMT showed that the values ranged from 0.50 to 0.69, which were less the 0.85 threshold (Kline, 2011).Table 3 Validity results for the measures for Study 1.
Table 3 Means (SD) CR AVE MSV Fear Anger Turnover Intentions
Fear 2.85 (1.10) 0.93 0.83 0.35 0.91
Anger 2.15 (1.28) 0.96 0.90 0.47 0.59* 0.95
Turnover intentions 2.82 (1.17) 0.86 0.69 0.47 0.50* 0.69* 0.83
Note. CR = composite reliability; AVE = average variance extracted; MSV = maximum shared variance. The square root of the AVEs are in bold.
*p < 0.01.
6.2 Test of hypotheses
The hypotheses were tested using PROCESS Model 4 with two parallel mediators, the control variables, and a bootstrap function extracting 5,000 samples for the analysis (95% confidence interval [CI]) (Hayes, 2017). The results showed that the participants who read that their job was eliminated due to the pandemic felt more anger than the participants who read that their job was not eliminated (β = 1.51, p = 0.001; CI = [1.10, 1.93]). Similarly, the participants who read that their job was eliminated due to the pandemic felt more anger than the participants who read that their job was not eliminated (β = 0.56, p = 0.01; CI = [0.13, 1.01]), thereby supporting H1a and H1b.
The results for the indirect effect showed that the relationship between employment status (job eliminated vs. employed) and industry turnover intentions was mediated by anger (effect = 0.65; CI = [0.30, 1.04]), supporting H2a. However, fear did not mediate the relationship between employment status (job eliminated vs. employed) and industry turnover intentions (effect = 0.11; CI = [-0.01, 0.26]), which did not support H2b. Table 4 shows the results of the main and indirect effects.Table 4 Direct and indirect effect for Study 2.
Table 4 Effect SE LLCI ULCI
H1a Employment status → anger 0.27 0.11 0.05 0.49
H1b Employment status → fear 0.10 0.10 0.02 0.42
Indirect effects BootSE LLCI ULCI
H2a Employment status → anger → industry turnover intentions 0.11 0.05 0.02 0.21
H2b Employment status → fear → industry turnover intentions −0.01 0.02 −0.05 0.03
Note. SE = standard error; LLCI = lower limit confidence interval; ULCI = upper limit confidence interval.
7 Discussion
Theory and research have primarily focused on organizations as the target for understanding turnover, while neglecting the industry. To address this limitation in the literature, the current paper used the hospitality industry as the foci for understanding turnover. Using two different samples (management-level employees in Study 1 and hospitality management students with hospitality work experience in Study 2) and methodology (survey for Study 1 and experiment for Study 2), the results showed being made unemployed due to the pandemic resulted in more anger and fear than being employed. Both studies, however, converged to show that the relationship between employment status (unemployed vs. employed) and industry turnover intentions was mediated by anger but not fear. These findings provide implications to explain why the industry's response to a mega-event, such as the COVID-19 pandemic, can motivate employees to leave the hospitality industry.
7.1 Theoretical implications
The current paper offers several theoretical implications. First, it advances the event-system theory (Morgeson et al., 2015) by showing how an industry's response to a mega-event, such as the COVID-19 pandemic, has a direct influence on employees' attitudes toward the industry. Specifically, event-system theory (Morgeson et al., 2015) focuses on how major events can affect the organization and how the actions taken by the organization can affect employee attitudes toward their organization. In addition, much of the literature on understanding why employees, including managers, leave the hospitality industry has focused on personal attributes, job attributes, and organizational-specific attributes (e.g., Ann & Blum, 2020; Blomme et al., 2010; Brown et al., 2015; Guchait et al., 2015). By using the affect theory of social exchange (Lawler, 2018) as a framework, however, the current paper showed how emotional reactions related to job status (i.e., being laid-off or furloughed) due to the pandemic influences intentions to leave the industry. Thus, an industry's response to a mega-event, such as the pandemic, can elicit negative emotions, such as anger and fear, which then motivate employees to leave the industry.
Second, the current study also has important implications for Lebel's (2017) model of fear and anger regulation. Lebel contends that fear and anger can act as motivators to lead individuals to take proactive measures to remedy these feelings, such as intentions to leave an industry. However, across both studies, anger, but not fear, was the primary driver of industry turnover intentions. One possible explanation is that anger is associated with high certainty situations and motivates a “fight” response (Novaco, 2016), whereas fear is associated with the “flight” response (Ohman & Wiens, 2003). While both emotions motivate a proactive effort to cut off the source of fear (Lebel, 2017), anger might motivate action to leave, whereas fear might motivate emotional withdrawal (Kish-Gephart et al., 2009). For example, people who feel higher intensity of anger than fear are more confident in their actions and more likely to take risks than people who feel higher intensity of fear than anger (Lerner & Keltner, 2001). These opposing views in risk appraisal from feeling anger or fear might explain why feeling anger was a stronger motivator of intentions to leave the industry. Because Study 1 had a sample of managers and Study 2 had a sample of hospitality management students with work experience, both samples had employees who have invested resources in working in the industry. Therefore, leaving the industry is a risk, which research has shown to be motivated by feelings of anger rather than feelings of fear.
Third, the current paper also builds on the hospitality turnover literature by demonstrating why job loss due to mega-events, such as the COVID-19 pandemic, can motivate employees to leave the hospitality industry (e.g., Akkermans et al., 2020; Bufquin et al., 2021; Chen & Chen, 2021; Yu et al., 2021). For example, a recent study showed how stress related to the COVID-19 pandemic can subsequently generate negative employee attitudes towards the hospitality industry (e.g., Yu et al., 2021). Similarly, Chen and Chen (2021) found that the stress related to job loss due to the pandemic was also related to impaired well-being and higher intentions to leave the hospitality industry. Unlike Yu et al. (2021) and Chen and Chen (2021), the current studies compared those who were made unemployed and those who remained employed during the pandemic in order to understand how discrete emotional reactions—anger and fear—affect their intentions to leave the industry. Specifically, it was the employees who lost their job due to the pandemic who felt more anger and fear than those still employed, and these emotions were related to industry turnover intentions.
Bufquin et al. (2021) also compared those who were made unemployed or remained employed during the pandemic but found that employees still working during the pandemic experienced higher levels of psychological distress, drug, and alcohol use than furloughed employees, all of which were related to industry turnover intentions. One reason for these contrasting findings is that sample from Bufquin et al. (2021) used non-management, hourly restaurant employees, a sample that research has shown to have high tendencies of substance abuse (Hight & Park, 2018; Kitterlin et al., 2015). The current research used management level employees (Study 1) and employed hospitality management students with goals of attaining management level careers (Study 2) —two samples that have invested resources in hospitality industry careers. Thus, remaining employed is consistent with their career aspirations, whereas those who lost their jobs were faced with a situation that was detrimental to their career aspirations.
Given that the hospitality industry was one of the industries most affected by COVID-19, negative attitudes among employees could mean a significant loss of talent as workers and job seekers consider switching industries for greater stability. Those who were furloughed or laid off due to COVID-19 felt greater levels of subjective anger than their counterparts who were still employed during the pandemic. As normal as these reactions to job and income loss may appear, the industry-wide circumstances that led to such widespread furloughs and lay-offs could result in subsequent actions taken by former employees to avoid a reoccurrence of such events; in this case, potentially leaving the hospitality industry altogether. Therefore, hospitality industry employees who lost their jobs due to COVID-19 may not return to work in the industry at all, even as conditions improve.
7.2 Practical implications
The current study additionally provides important practical implications. The results revealed that intentions to leave the industry were related to the anger felt from being unemployed from the hospitality industry. This is a cause for concern, because the hospitality industry relies on a pipeline of professionals who have skills and knowledge that are industry-specific and transferable across organizations within an industry (Baum, 2019). In addition, the hospitality industry relies on a pipeline of talent from which management positions are often filled by employees who have frontline work experience in the industry. Therefore, the hospitality industry may suffer from a loss of talent if industry-related negative work events, such as the pandemic, negatively influence employee attitudes and behaviors toward the industry. As such, these results point to the importance for hospitality organizations to develop strategies that can attenuate employees’ feelings of anger related to being unemployed. For example, hospitality organizations currently recruiting talent might use messages that build trust and/or address how they will avoid layoffs and furloughs in the future (Guzzo et al., 2021).
In addition, the results suggest that the pandemic is a problem for the industry as a whole, rather than a problem for individual organizations within the industry. This suggests that the hospitality industry must rebuild trust among its talent by communicating what was learned for addressing future events and how it plans to recover. This requires trade associations, such as the National Restaurant Association and the American Hotel &Lodging Association, and industry partners, such as university programs, to focus on recovery efforts (King et al., 2021). For example, industry trade associations and major chain corporations could partner to communicate plans for recovering lost jobs and plans to address future pandemics and other negative work events that can affect the industry.
It is also important for hospitality organizations to consider how they may prevent or at least reduce anger-driven industry turnover intentions to begin with, should the industry be faced with similar future crises. Contingency plans such as offering employees continuing benefits, alternative work arrangements, and/or training programs, may go a long way towards attenuating negative emotions towards the organization and hospitality industry. In addition, it is critical that hospitality organizations maintain clear and consistent communication with their employees regarding the situation that they are facing in times of uncertainty, as well as how they plan to address these challenges. These factors, if combined with an emphasis on integrity and employee welfare, could demonstrate to hospitality employees that despite the difficulties faced by their industry, the organizations they work for value them, and are prepared to protect employee interests in the face of hardships (Guzzo et al., 2021).
7.3 Limitations and future research
The main limitation of this study is the fact that the data was collected using samples from one country, the United States. Although similar labor shortages in the hospitality industry have been found on other parts of the world, like Australia (Powell, 2022), Asia (TTG Asia, 2021), and Europe (Diazgranados, 2021), future research can examine these relationships in other cultures, wherein different emotions might predict different outcomes (e.g., Luo et al., 2019). Another limitation is that in both current studies, intentions to leave the industry were used. Future studies should be conducted to establish whether the negative emotions that respondents report do, in fact, result in real industry turnover among those individuals. It is possible, for instance, that individuals who were laid off or furloughed only temporarily experienced greater negative feelings towards the hospitality industry, and their desire to leave the industry altogether decreased over time. An additional limitation is the fact that the data for this study only examines the attitudes of hospitality industry employees. Comparing this with the attitudes of current and laid off or furloughed employees from other industries affected by COVID-19 could provide important context for hospitality industry leaders as well as future researchers.
Despite these limitations, across two studies, we found that hospitality talent who experienced a change in job status due to the COVID-19 pandemic, either in the form of lay-offs or furloughs, consequently felt higher levels of subjective anger and fear than their counterparts who were still employed. In turn, higher levels of anger, but not fear, were related to higher industry turnover intentions. These results highlight a potentially problematic trend. Should skilled hospitality workers switch industries due to job loss amidst an industry-wide negative event, it may become difficult for hospitality businesses to find qualified employees once the industry recovers and rehiring begins. For this reason, it is important that companies within the hospitality industry carefully consider how to approach staffing cuts during difficult times, as talent that is let go may not return.
==== Refs
References
Akkermans J. Richardson J. Kraimer M.L. The Covid-19 crisis as a career shock: Implications for careers and vocational behavior Journal of Vocational Behavior 119 2020 1 5
American HotelLodging Association COVID-19's impact on the hotel industry Retrieved from https://www.ahla.com/covid-19s-impact-hotel-industry 2020
Ann S. Blum S.C. Motivating senior employees in the hospitality industry International Journal of Contemporary Hospitality Management 32 1 2020 324 346
Asia T.T.G. IHG embraces flexible working to tackle hospitality labour shortage. TTG Asia https://www.ttgasia.com/2021/09/23/ihg-embraces-flexible-working-to-tackle-hospitality-labour-shortage/ 2021
Bagozzi R.P. Yi Y. On the evaluation of structural equation models Journal of the Academy of Marketing Science 16 1 1988 74 94
Bajrami D.D. Terzić A. Petrović M.D. Radovanović M. Tretiakova T.N. Hadoud A. Will we have the same employees in hospitality after all? The impact of COVID-19 on employees' work attitudes and turnover intentions International Journal of Hospitality Management 94 2021 102754
Baum T. Does the hospitality industry need or deserve talent? International Journal of Contemporary Hospitality Management 31 10 2019 3823 3837
Baum T. Mooney S.K. Robinson R.N. Solnet D. COVID-19's impact on the hospitality workforce – new crisis or amplification of the norm? International Journal of Contemporary Hospitality Management 32 9 2020 2813 2829
Blomme R.J. Van Rheede A. Tromp D.M. The use of the psychological contract to explain turnover intentions in the hospitality industry: A research study on the impact of gender on the turnover intentions of highly educated employees The International Journal of Human Resource Management 21 1 2010 144 162
Brown E.A. Thomas N.J. Bosselman R.H. Are they leaving or staying: A qualitative analysis of turnover issues for generation Y hospitality employees with a hospitality education International Journal of Hospitality Management 46 2015 130 137
Brynjolfsson E. Horton J.J. Ozimek A. Rock D. Sharma G. TuYe H.Y. COVID-19 and remote work: An early look at US data (No. w27344) 2020 National Bureau of Economic Research
Bufquin D. Park J.Y. Back R.M. de Souza Meira J.V. Hight S.K. Employee work status, mental health, substance use, and career turnover intentions: An examination of restaurant employees during COVID-19 International Journal of Hospitality Management 93 2021 102764
Carless S.A. Arnup J.L. A longitudinal study of the determinants and outcomes of career change Journal of Vocational Behavior 78 1 2011 80 91
Chen C.C. Chen M.H. Well-being and career change intention: COVID-19's impact on unemployed and furloughed hospitality workers International Journal of Contemporary Hospitality Management 33 8 2021 2500 2520
Chen H. Eyoun K. Do mindfulness and perceived organizational support work? Fear of COVID-19 on restaurant frontline employees' job insecurity and emotional exhaustion International Journal of Hospitality Management 94 2021 102850
Conroy S.A. Becker W.J. Menges J.I. The meaning of my feelings depends on who I am: Work-related identifications shape emotion effects in organizations Academy of Management Journal 60 3 2017 1071 1093
Crisp R.J. Heuston S. Farr M.J. Turner R.N. Seeing red or feeling blue: Differentiated intergroup emotions and ingroup identification in soccer fans Group Processes & Intergroup Relations 10 2007 9 26
Dasborough M.T. Hannah S.T. Zhu W. The generation and function of moral emotions in teams: An integrative review Journal of Applied Psychology 105 5 2020 433 31512899
Diazgranados D. Labor shortages in European hospitality industry threaten hopes of tourism boost. Medium https://danilodiazgranados.medium.com/labor-shortages-in-european-hospitality-industry-threaten-hopes-of-tourism-boost-7211f1369b7d 2021
Fornell C. Larcker D.F. Evaluating structural equation models with unobservable variables and measurement error Journal of Marketing Research 18 1 1981 39 50
Fox S. Spector P.E. A model of work frustration–aggression Journal of Organizational Behavior 20 6 1999 915 931
Grandey A.A. Sayre G.M. French K.A. A blessing and a curse”: Work loss during coronavirus lockdown on short-term health changes via threat and recovery Journal of Occupational Health Psychology 26 4 2021 261 275 34292019
Greenbaum R. Bonner J. Gray T. Mawritz M. Moral emotions: A review and research agenda for management scholarship Journal of Organizational Behavior 41 2 2020 95 114
Guchait P. Cho S. Meurs J.A. Psychological contracts, perceived organizational and supervisor support: Investigating the impact on intent to leave among hospitality employees in India Journal of Human Resources in Hospitality & Tourism 14 3 2015 290 315
Gursoy D. Chi C.G. Effects of COVID-19 pandemic on hospitality industry: Review of the current situations and a research agenda Journal of Hospitality Marketing & Management 29 5 2020 527 529
Guzzo R.F. Wang X. Madera J.M. Abbott J. Organizational trust in times of COVID-19: Hospitality employees' affective responses to managers' communication International Journal of Hospitality Management 93 2021 102778
Haddon H. Restaurants serve up signing bonuses, higher pay to win back workers Retrieved from https://www.wsj.com/articles/restaurants-serve-up-signing-bonuses-higher-pay-to-win-back-workers-11619359201 2021
Hair J.F. Black W.C. Babin B.J. Anderson R.E. Multivariate data analysis: A global perspective 2010 Pearson New York, NY
Haldorai K. Kim W.G. Pillai S.G. Park T.E. Balasubramanian K. Factors affecting hotel employees' attrition and turnover: Application of pull-push-mooring framework International Journal of Hospitality Management 83 2019 46 55
Han S. Lerner J.S. Keltner D. Feelings and consumer decision making: The appraisal-tendency framework Journal of Consumer Psychology 17 3 2007 158 168
Hayes A.F. Introduction to mediation, moderation, and conditional process analysis: A regression-based approach 2017 Guilford publications
Henseler J. Ringle C.M. Sarstedt M. A new criterion for assessing discriminant validity in variance-based structural equation modeling Journal of the Academy of Marketing Science 43 1 2015 115 135
Hight S.K. Park J.Y. Substance use for restaurant servers: Causes and effects International Journal of Hospitality Management 68 2018 68 79
Huffman A.H. Albritton M.D. Matthews R.A. Muse L.A. Howes S.S. Managing furloughs: How furlough policy and perceptions of fairness impact turnover intentions over time International Journal of Human Resource Management 33 14 2022 2801 2828
Izard C.E. The psychology of emotions 1991 Plenum Press New York
King C. Madera J.M. Lee L. Murillo E. Baum T. Solnet D. Reimagining attraction and retention of hospitality management talent–A multilevel identity perspective Journal of Business Research 136 2021 251 262
Kish-Gephart J.J. Detert J.R. Treviño L.K. Edmondson A.C. Silenced by fear: The nature, sources, and consequences of fear at work Research in Organizational Behavior 29 2009 163 193
Kitterlin M. Moll L. Moreno G. Foodservice employee substance abuse: Is anyone getting the message? International Journal of Contemporary Hospitality Management 27 5 2015 810 826
Kline R.B. Principles and practice of structural equation modeling 2011 Guilford Press New York
Lawler E.J. An affect theory of social exchange American Journal of Sociology 107 2 2001 321 352
Lawler E.J. The affect theory of social exchange Burke P.J. Contemporary social psychological theories 2018 Stanford University Press 224 248
Lazarus R.S. Folkman S. Stress appraisal and coping 1984 Springer New York
Lebel R.D. Moving beyond fight and flight: A contingent model of how the emotional regulation of anger and fear sparks proactivity Academy of Management Review 42 2 2017 190 206
Lee P.C. Xu S.T. Yang W. Is career adaptability a double-edged sword? The impact of work social support and career adaptability on turnover intentions during the COVID-19 pandemic International Journal of Hospitality Management 94 2021 102875
Lerner J.S. Keltner D. Fear, anger, and risk Journal of Personality and Social Psychology 81 1 2001 146 159 11474720
Lerner J.S. Tiedens L.Z. Portrait of the angry decision maker: How appraisal tendencies shape anger's influence on cognition Journal of Behavioral Decision Making 19 2 2006 115 137
Luo A. Guchait P. Lee L. Madera J.M. Transformational leadership and service recovery performance: The mediating effect of emotional labor and the influence of culture International Journal of Hospitality Management 77 2019 31 39
McGinley S. Mattila A.S. Overcoming job insecurity: Examining grit as a predictor Cornell Hospitality Quarterly 61 2 2020 199 212
McGinley S. O'Neill J. Damaske S. Mattila A.S. A grounded theory approach to developing a career change model in hospitality International Journal of Hospitality Management 38 2014 89 98
Mimoun E. Ben Ari A. Margalit D. Psychological aspects of employment instability during the COVID-19 pandemic Psychological Trauma: Theory, Research, Practice, and Policy 12 S1 2020 S183 32538650
Miscenko D. Day D.V. Identity and identification at work Organizational Psychology Review 6 3 2016 215 247
Morgeson F.P. Mitchell T.R. Liu D. Event system theory: An event-oriented approach to the organizational sciences Academy of Management Review 40 4 2015 515 537
National Restaurant Association Coronavirus information and resources Retrieved from https://restaurant.org/Covid19 2020
Nicola M. Alsafi Z. Sohrabi C. Kerwan A. Al-Jabir A. Iosifidis C. …Agha R. The socio-economic implications of the coronavirus pandemic (COVID-19): A review International Journal of Surgery 78 2020 185 193 32305533
Novaco R.W. Anger Stress: Concepts, cognition, emotion, and behavior 2016 Academic Press 285 292
Ohman A. Wiens S. On the automaticity of autonomic responses in emotion: An evolutionary perspective Davidson R.J. Scherer K.R. Goldsmith H.H. Handbook of affective sciences 2003 Oxford University Press New York 256 275 Organizational Psychology and Organizational Behavior, 7, 451-477
Osborne D. Smith H.J. Huo Y.J. More than a feeling: Discrete emotions mediate the relationship between relative deprivation and reactions to workplace furloughs Personality and Social Psychology Bulletin 38 5 2012 628 641 22215700
Parker S.K. Bindl U.K. Strauss K. Making things happen: A model of proactive motivation Journal of Management 36 4 2010 827 856
Picchi A. Restaurants face new COVID-19 challenge: Finding enough workers Retrieved from https://www.cbsnews.com/news/restaurants-covid-19-worker-shortage/ 2021
Powell D. Demand for hospitality staff hits two-year high as union warns on exploitation Sydney Morning Herald 2022 https://www.smh.com.au/business/companies/demand-for-hospitality-staff-hits-two-year-high-as-union-warns-on-exploitation-20220317-p5a5ja.html#:~:text=Applicants%20for%20hospitality%20jobs%20are,lowest%20level%20since%20mid%2D2008
Pugh S.D. Skarlicki D.P. Passell B.S. After the fall: Layoff victims' trust and cynicism in re‐employment Journal of Occupational and Organizational Psychology 76 2 2003 201 212
Rubino C. Wilkin C.L. Malka A. Under pressure: Examining the mediating role of discrete emotions between job conditions and well-being The role of emotion and emotion regulation in job stress and well being 2013 Emerald Group Publishing Limited
Shepherd D.A. Williams T.A. Hitting rock bottom after job loss: Bouncing back to create a new positive work identity Academy of Management Review 43 1 2018 28 49
Smith H.J. Pettigrew T.F. The subjective interpretation of inequality: A model of the relative deprivation experience Social and Personality Psychology Compass 8 12 2014 755 765
Suh E. West J.J. Shin J. Important competency requirements for managers in the hospitality industry Journal of Hospitality, Leisure, Sports and Tourism Education 11 2 2012 101 112
U.S. Bureau of Labor Statistics Impact of the coronavirus (COVID-19) pandemic on the employment situation for august 2020 U.S. Bureau of Labor Statistics, U.S. Bureau of Labor Statistics 2020 2020 www.bls.gov/covid19/employment-situation-covid19-faq-may-2020.htm
Wiener-Bronner D. A perfect storm': These restaurants survived the pandemic. Now they can't find workers Retrieved from https://www.cnn.com/2021/04/21/business/restaurant-labor-shortage/index.html 2021
Wong A.K.F. Kim S.S. Kim J. Han H. How the COVID-19 pandemic affected hotel Employee stress: Employee perceptions of occupational stressors and their consequences International Journal of Hospitality Management 93 2021 102798
Yan J. Kim S. Zhang S.X. Foo M.D. Alvarez-Risco A. Del-Aguila-Arcentales S. Yáñez J.A. Hospitality workers' COVID-19 risk perception and depression: A contingent model based on transactional theory of stress model International Journal of Hospitality Management 95 2021 102935
Yu H. Lee L. Popa I. Madera J.M. Should I leave this industry? The role of stress and negative emotions in response to an industry negative work event International Journal of Hospitality Management 94 2021 102843
| 0 | PMC9742209 | NO-CC CODE | 2022-12-14 23:53:28 | no | 2023 Mar 12; 54:119-127 | utf-8 | null | null | null | oa_other |
==== Front
Int Dent J
Int Dent J
International Dental Journal
0020-6539
1875-595X
The Authors. Published by Elsevier Inc. on behalf of FDI World Dental Federation.
S0020-6539(22)00279-9
10.1016/j.identj.2022.12.002
Letter to the Editor
COVID-19 Vaccine-Induced Immune Response in Oral Fluids and Serum.: correspondence
Sookaromdee Pathum 1⁎
Wiwanitkit Viroj 2
1 Private Academic Consultant, Bangkok Thailand
2 Honorary professor, Dr DY Patil Vidhyapeeth, Pune, India
⁎ Correspondence: Pathum Sookaromdee, Private Academic Consultant, Bangkok Thailand
12 12 2022
12 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.
==== Body
pmcDear Editor, we would like to share ideas “BNT162b2 mRNA Vaccine-Induced Immune Response in Oral Fluids and Serum.1.” Seneviratne et al. examined the immune responses in the blood of 50 healthy healthcare professionals following two doses of the intramuscular Pfizer/BioNTech-BNT162b2 vaccination and compared them to anti-spike protein antibodies in gingival crevicular fluid (GCF) and saliva1. Seneviratne et al. came to the conclusion that, under certain constraints, GCF would be a less intrusive alternative to serum and a better choice than saliva to detect antibody responses by the current COVID-19 vaccines1. Further study is required to determine the suitability of GCF for community immunological surveillance for vaccinations, according to Seneviratne et al.1. Because COVID-19 is a prevalent clinical entity2 that cannot be easily determined from a clinical history taking, it is crucial to rule out any prior asymptomatic episodes. Genetic variables play a role in the immunological response to the COVID-19 vaccination3. There ought to be more prospective research that can control for confounding variables.
Ethical disclosure
No conflict of interest and no fund is disclosed
Conflict of interest
None
==== Refs
References
1 Seneviratne CJ Balan P de Alwis R Udawatte NS Herath T Toh JZN Tin GB Ooi EE Hong JLG Ying JSX. BNT162b2 mRNA Vaccine-Induced Immune Response in Oral Fluids and Serum Int Dent J 2022 Nov 16 10.1016/j.identj.2022.09.005 S0020-6539(22)00225-8Online ahead of print
2 Joob B Wiwanitkit V. Letter to the Editor: Coronavirus Disease 2019 (COVID-19), Infectivity, and the Incubation Period J Prev Med Public Health 53 2 2020 Mar 70 32268458
3 Chen DP Wen YH Lin WT Hsu FP. Association between the side effect induced by COVID-19 vaccines and the immune regulatory gene polymorphism Front Immunol 13 2022 Oct 26 941497
| 0 | PMC9742210 | NO-CC CODE | 2022-12-14 23:29:59 | no | Int Dent J. 2022 Dec 12; doi: 10.1016/j.identj.2022.12.002 | utf-8 | Int Dent J | 2,022 | 10.1016/j.identj.2022.12.002 | oa_other |
==== Front
Transp Res Part A Policy Pract
Transp Res Part A Policy Pract
Transportation Research. Part A, Policy and Practice
0965-8564
1879-2375
The Author(s). Published by Elsevier Ltd.
S0965-8564(22)00312-3
10.1016/j.tra.2022.103561
103561
Article
Willingness to pay for COVID-19 mitigation measures in public transport and paratransit in low-income countries
Bwambale Andrew a
Uzondu Chinebuli b
Islam Mohaimanul c
Rahman Farzana c
Batool Zahara d
Mukwaya Paul e
Wadud Zia f⁎
a College of Engineering Design, Art and Technology, Makerere University, P.O. Box 7062, Kampala, Uganda
b Transport Management Technology, Federal University of Technology Owerri, PMB 1526 Owerri, Imo State, Nigeria
c Department of Civil Engineering, University of Asia Pacific, Dhaka 1205, Bangladesh
d Institute for Transport Studies and School of Chemical and Process Engineering, University of Leeds, Leeds LS2 9JT
e Department of Geography, Geo-informatics and Climatic Sciences, Makerere University, P.O. Box 7062, Kampala Uganda
f Institute for Transport Studies and School of Chemical and Process Engineering, University of Leeds, Leeds LS2 9JT
⁎ Corresponding author.
12 12 2022
12 12 2022
10356128 2 2022
27 11 2022
3 12 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 order to combat the spread of COVID-19, various measures were taken in most countries to make public transit and paratransit safer. These additional measures, which include restrictions on number of passengers, provision of hand sanitisers and face coverings, and more frequent cleaning add to the costs of operations or reduce profitability. The resulting financial pressure on the transport operators raises an important question on who pays for these additional measures. In most countries, this has been covered by one-time government bailouts to operators or strategies to increase fare, the latter of which directly affects the users. However, even without these interventions, there could be a demand and as such willingness to pay (WTP) for some of these intervention measures from the consumers concerned about safety. Knowing such WTP will not only help operators set their fare, but also help the governments decide the appropriate bailout needed. This paper addresses the issue by estimating the user’s willingness to pay for selected COVID-19 mitigation measures in public transport and paratransit (motorcycle taxis) using survey data collected from two cities in low-income countries as case studies – Kampala, Uganda and Dhaka, Bangladesh. For public transport, these measures are - (1) social distancing (passenger loading at half capacity), and (2) mandatory hand sanitisation and increased cleaning of surfaces, while for paratransit, they are - (1) provision of a transparent shield between the rider and the passenger, and (2) provision of cleaned helmets at the start of each trip. The study analyses stated preference data using the utility maximisation framework and finds that the implementation or provision of COVID-19 mitigation measures improves the attractiveness of the associated public transport or paratransit alternatives, and transport users make trade-offs between safety and cost when making travel decisions. We find positive willingness to pay for all four mitigation measures, suggesting potential existence of a market for these measures. We also find that the typical mode choice factors such as costs, travel time and convenience became less important during the pandemic and the safety measures became more important considerations.
Keywords
COVID-19
Public transport
Paratransit
Willingness to pay
Social distancing
Risk mitigation
==== Body
pmc1 Introduction
The outbreak of COVID-19 has substantially affected travel behaviour, transport operations and associated policies in many countries across the world. As of now, the pandemic has taken over 6.56 Million lives, and infected over 626 Million people globally (Worldometer, 2022). The disease is known to spread through physical proximity via respiratory droplets of infected persons (WHO, 2020). As such, various countries responded by implementing unprecedented measures to minimise social proximity including isolation of arriving international passengers, introduction of local travel restrictions, and closure of schools, workplaces, shopping centres, and other public places, effectively imposing lockdowns at regional and/or national levels (De Vos, 2020). The lockdown measures were supplemented with additional preventive measures including social distancing, facemask wearing, hand sanitisation, and temperature screening.
However, such lockdowns could not be sustained for long in most countries due to the associated economic challenges and had to be lifted or eased, despite the potential risk of resurgence of new COVID-19 infections (Hargreaves, 2020). This has come with a raft of new measures affecting various sectors, including transport. The interventions in the transport sector have mainly affected public transport and paratransit, with some investments in walking and cycling provisions. Social distancing (passenger loading at reduced capacity), regular cleaning of vehicle interior surfaces, passenger hand sanitisation, compulsory wearing of face masks, and temperature screening have become requirements for public transport operation in various countries (MOH, 2020, De Vos, 2020, Tirachini and Cats, 2020, Dzisi and Dei, 2020, Wadud et al., 2022). Similarly, for paratransit (motorcycle taxis), where it is impossible to achieve distancing between the driver and passenger(s) due to the small size of the vehicles, the measures introduced include regular sanitisation of motorcycle surfaces, compulsory wearing of helmets by riders 1, and face masks by passengers (Otto, 2020). These new standards coupled with increased operating costs and reduced passenger demand have resulted in a financial crisis for public transport and paratransit operators, with many of them turning to their governments for bailouts (Tirachini and Cats, 2020, De Vos, 2020). While bailouts have been offered in many developed countries, this has not been done in most low-income countries where governments have limited or no capacity to subsidise transport services, which are often unorganized and loosely regulated. Consequently, most of the additional operating costs associated with COVID-19 mitigation have been transferred to transport users with little regard to their sensitivities. This could potentially lead to pressure from lobby groups and low compliance from users and operators, thereby jeopardising the progress made in making public transport and paratransit safe. This is a cause of concern especially in low-income countries where the rates of COVID-19 vaccination are still very low - currently standing at 9.5% for the first dose (Our World in Data, 2022), and mitigiating the spread through transport related measures remains important.
The pandemic has also inspired innovation in both technological (Peters, et al., 2021) and non-technological areas (Athumani, 2020). In particular, installing transparent shields between riders and passengers on motorcycle taxis was either discussed or implemented as a means to reducing transmission in a few low-income countries, such as Uganda, Philippines or Indonesia. Yet some policymakers have argued that helmets may have more benefits compared to shields (Athumani, 2020). Regardless of the risk mitigation approaches, it is likely that the cost of adding the shields or cleaning the helmets will be transferred to the users, especially in the largely unorganized profit-driven motorcycle taxi sector in these countries.
The recent emergence of the Omicron variant suggests that it is still important for the policymakers and regulators to remain vigilant and open to new interventions to reduce transmission. For example, Bangladesh had recently proposed half-capacity public transport again (New Age, 2022), although this was later overturned due to pressure from transport lobby ((Adhikary, 2022)). Under the current circumstances, where most of the COVID-19 mitigation costs are being transferred to transport users, it is important to understand the users’ willingness to pay (WTP) for COVID-19 mitigation to help guide policymakers on possible interventions to sustain compliance to the set guidelines for transport.
Given this context, this study uses Stated Preference (SP) data and discrete choice methods to estimate users’ WTP for selected COVID-19 mitigation measures in public transport (buses and minibuses) and paratransit (motorcycle taxis) using two cities in low-income countries, Kampala (Uganda) and Dhaka (Bangladesh), as case studies. The study estimates the WTP for social distancing (half capacity) and sanitisation (of both hands and surfaces) in public transport, as well as, for provision of a shield (between the rider and passenger) and of a cleaned/disinfected helmet at the start of each trip in paratransit. Despite some literature on WTP for various COVID-19 mitigation measures in transport (Arunwuttipong et al., 2021, Thombre and Agarwal, 2021, Awad-Núñez et al., 2021), they did not investigate the relative importance attached to the different mitigation measures in monetary terms, a gap addressed in this study. There is also a lack of any studies on users’ attitudes and willingness to pay for safety interventions in motorcycle taxis. In particular, we estimate the WTP for safety barriers on motorcycle taxis to test the feasibility of this innovation.
The rest of the paper is organised as follows, Section 2 presents a brief review of the literature, while Section 3 presents the survey design and data used in the study. Section 4 presents the econometric modelling framework, while Section 5 presents descriptive insights from the data, modelling results, and willingness to pay estimates. Section 6 draws conclusions.
2 Literature review
2.1 The impacts of COVID-19 on public transport
Studies conducted in various countries have reported that the COVID-19 pandemic has substantially affected travel behaviour as most policies to mitigate the spread of the virus have restricted non-essential activities outside home (Arellana et al., 2020, Beck and Hensher, 2020, Beck et al., 2020, Ssali, 2020, Kitara and Ikoona, 2020). This has led to significant reductions in the number of trips, with public transport experiencing the greatest decline in ridership, mainly due to the high risk of transmission associated with mass travel (Teixeira and Lopes, 2020, Bucsky, 2020, de Haas et al., 2020, Aloi et al., 2020, Reza et al., 2020). Many countries have seen increased reliance on private car travel (Campisi et al., 2020, Abdullah et al., 2020, Fatmi, 2020, Das et al., 2021) and active travel (Zafri et al., 2021, Zhang et al., 2021, Budd and Ison, 2020, De Vos, 2020) as means to avoid proximity to others while making essential trips.
Despite the rise in active travel, public transport and paratransit system remains crucial to the functioning of large cities in the developing and emerging countries – and the lockdown measures could only be temporary. There were clear preferences among respondents in India to continue using their preferred pre-COVID modes during the pandemic (Bhaduri, et al., 2020), and for low-income countries, these are mainly public transport and paratransit modes. However, the need to maintain safe public transport operations during the pandemic has neccessiated the introduction of risk mitigation measures in public transport sector, as described earlier. These mitigation measures have come with cost implications for the operators which were either supported by government subsidies (Tirachini and Cats, 2020, De Vos, 2020) or increased fare for passengers. The latter case is more common in most low-income countries (Mogaji, 2020, Porter et al., 2021), and requires further understanding of the passengers’ willingness to pay for these increased costs.
2.2 The impacts of COVID-19 on paratransit (motorcycle taxis)
Motorcycle taxis are a popular mode of transport in many low-income countries (Ehebrecht et al., 2018, Starkey, 2016, Wadud, 2020). This is largely attributed to the absence of reliable alternative public transport modes and the ability of motorcycle taxis to manoeuvre around traffic jams and provide fast door-to-door services, enabling quick access to essential social services and economic opportunities (Olvera, et al., 2012). The outbreak of COVID-19 has had significant negative impacts on motorcycle taxi transport in many countries, especially during lockdown periods, where passenger transport services were generally restricted and goods delivery services were promoted instead (Peters, et al., 2021). This led to an overall reduction in the number of trips, negatively affecting the revenues and livelihoods of operators, too (Batool, et al., 2022). To reverse the situation, different countries have introduced measures to ensure that motorcycle taxis get back to normal and safe operations. For example, in Uganda, the measures include regular sanitisation of motorcycle surfaces, carrying of strictly one passenger, compulsory wearing of helmets by riders, and face masks by passengers (Otto, 2020). However, some of these measures have led to fare increases. Despite this increase, it is expected that motorcycle taxis will continue to dominate the urban transport market in most sub-Saharan and south-east Asian countries. Indeed, results from a worldwide expert survey show substantial mode shifts from public transport to motorcycles in various countries during the pandemic, with the shifts being much higher in India and other Asian countries (Zhang, et al., 2021).
The pandemic has also increased smartphone and mobile internet penetration among motorcycle taxi operators, who are increasingly relying on text messages, ride-hailing apps and e-payments to arrange trips (Peters, et al., 2021). Indigeneous innovations were also observed, of which safety barriers between motorcycle rider and passenger is a key one. These barriers have been mandated by the government in Philippines (DILG, 2020), and piloted in Indonesia by ridesourcing operator Grab (Ardiansyah & Purnomo, 2020). On the other hand, they were discouraged by authorities in Uganda (Athumani, 2020), citing lack of evidence about their effectiveness in mitigating the risks. Computational Fluid Dynamics (CFD) modelling shows that these shields can indeed reduce the risk of COVID-19 airborne transmission, thus allowing for safe travel (Hetherington, et al., 2021). However, it is not known whether the users are willing to pay extra (through increased fares) for such an innovation. This could dictate whether the market may have a role in providing such measures, or government mandates are required to encourage their use.
2.3 Previous studies on willingness to pay for COVID-19 mitigation measures
Following the outbreak of COVID-19, some studies have investigated the WTP for improved safety (reduced transmission risk) during travel. Awad-Núñez, et al. (2021) use Heckman’s choice modelling framework to estimate the probabilities of accepting specific post-COVID-19 WTP levels (in terms of percentages above the current fare) linked to selected sanitisation measures and increased service frequency (to avoid crowding) in public transport and shared mobility services in Spain, however, they do not report the disaggregated WTP estimates for each mitigation measure in monetary terms. Thombre & Agarwal (2021) use descriptive statistics to establish the post-COVID-19 WTP for safer, faster, cleaner, comfortable and resilient public transport services in India in terms of percentages above the base fare. Similarly Arunwuttipong, et al. (2021) use descriptive statistics to estimate the mean and median WTP estimates for enhanced disinfection in public transport in Thailand. Away from transport, Oreffice & Quintana-Domeque (2021) estimate the general WTP for COVID-19 protective gear and how this is influenced by information on the rates of infection and death. However, to the best of our knowledge, no previous study has attempted to understand the underlying preferences and relative importance attached to the different mitigation measures in monetary terms.
3 Study design and data
3.1 Brief context of the case study cities
3.1.1 Kampala
Kampala is the capital and largest city of Uganda. The city has a multi-modal transport system comprising private cars, 14-seater small buses (matatus), buses, motorcycle taxis (boda bodas), cycling and walking. Like in many African cities, there is no formal public transport system in Kampala. Commuters rely on informal systems dominated by matatus (46%) and boda bodas (32%) followed by cars (19%), then buses (2%) (KCCA, 2018). Following the first confirmed case of COVID-19 in Uganda on 21 March 2020, transport was one of the most hit sectors, with public transport and motorcycle taxis being suspended and restrictions being placed on private vehicle travel on 25 March 2020. The resumption of transport operations came nearly two and a half months later with new restrictions and regulations including carrying passengers at half capacity for public transport, limiting the number of private vehicle occupants to three (including the driver), compulsory sanitisation of hands and surfaces, temperature monitoring, and compulsory wearing of facemasks. It is worth noting that despite the resumption of transport operations in June 2020, services would occasionally be suspended following new waves of the pandemic in the country. The COVID-19 transport regulations remained in force and were only relaxed in April 2022 with increased vaccination drive (MOH, 2022).
3.1.1.1 Dhaka
Dhaka is the capital and largest city of Bangladesh. It is also among the megacities of the world, with a population of 16.8 Million (Demographia, 2021), and the most densely populated city in the world. Despite its size, there are no rail based mass rapid transit in Dhaka, although one is currently under construction and is expected to open soon. The city’s transportation needs are currently fulfilled by road transport vehicles like cars, buses, paratransit (laguna, autorickshaws, pedal rickshaws, motorcycle taxis), motorcycles, bicycles and walking. According to RSTP (2015), more than 60% of the travellers use public transport for their journey to work. There were two lockdowns in Dhaka in response to COVID-19 that affected the public transport sector. All public transport activities (including trains, flights) were closed between March 28 and June 1 in 2020, when buses were allowed to operate at half capacity (with accompanying fare increases). Motorcycle taxis were banned for nearly 5 months. The second ban on public transport started on 5 April 2021, but the decision was quickly overturned to allow buses and trains to run at half capacities in 11 large cities. Nationwide, buses resumed operations much later on 11 August 2021, but only half of the buses were allowed. A third wave of restrictions in January 2022 on public transport was overturned almost as soon as it started due to the pressure from the strong transport lobby.
3.2 Survey design
The survey questionnaire was divided into three parts. The first part of the survey collected information on the respondent’s travel behaviour before and during the pandemic, as well as their attitudes towards the COVID-19 mitigation measures in public transport and paratransit. The second part of the survey was a stated choice experiment only for respondents who frequently used either public transport or paratransit before the pandemic, and the experiments differed depending on the dominant mode category. The third part of the survey collected information on the demographics of the respondents, including their gender, age, level of education, employment status, income and parenthood.
Respondents were presented with hypothetical choices requiring trading between the transport fare and selected COVID-19 mitigation attributes. The travel time attribute was not considered as the study focussed on choice between same mode alternatives, only with differences in the levels of implementation of the COVID-19 mitigation measures, and in all cases, these have the same travel time. This precludes the ability to capture the possibility that those in longer journeys could have a higher WTP to mitigate risks, as the exposure risks could be higher in those trips.
The mitigation measures considered are those where the cost is initially incurred by the operator and recovered through user fares. For public transport (buses and minibuses), the measures considered were (1) social distancing (passenger loading at half capacity), and (2) mandatory hand sanitisation plus increased cleaning/disinfection frequency of surfaces, while for paratransit (motorcycle taxis), the measures considered were (1) provision of a transparent shield between the rider and the passenger, and (2) provision of cleaned/disinfected helmets at the start of each trip. 2 Each choice task was composed of two unlabelled (same mode) alternatives, each with three attributes, and distinguished by the attribute levels. The attribute levels used in the stated choice experiments were reviewed and discussed with stakeholders to ensure that they are realistic in each city. The fare attribute levels were respondent-specific and were generated based on the respondent’s reported pre-COVID fare for a typical commuting or education trip. The excerpts of the choice questions for public transport and paratransit are presented in Table 1, Table 2 , respectively. Given that the shields for motorcycle taxis were not a familiar concept, images were shown to the repondents to familiarize them with the shields, before the actual choice questions were asked.Table 1 Sample choice question for public transport
Service attributes Option 1 Option 2
Social distancing (half capacity) Not implemented Implemented
Mandatory hand sanitising and increased cleaning/disinfection frequency Implemented Not implemented
Fare, UGX 4500 3750
Please select your best option y□ y□
Table 2 Sample choice question for paratransit
Service attributes Option 1 Option 2
Transparent shield between the rider and the passenger Not installed Installed
Helmets are provided and cleaned/disinfected at the start of each trip Not provided Provided
Fare, UGX 3000 3750
Please select your best option y⊠ y□
The “idefix” R package was used to generate efficient designs of the choice experiments (Traets & Gil, 2020). The choice tasks were designed to encourage trading between attributes, for example, alternatives, where COVID-19 mitigation measures were being implemented, were designed to be more expensive and paired with cheaper alternatives where COVID-19 mitigation measures were not being implemented. It may be noted that the generation of efficient stated choice experiments requires knowledge on the priors of the parameter estimates (Bliemer & Collins, 2016). This study being the first of its kind, we could not rely on literature sources to determine the priors. We, therefore, organised pilot experiments in both Kampala and Dhaka and used the results of the pilot to generate the final efficient designs. The final designs were composed of 60 choice tasks, divided into 20 blocks, each with 3 choice tasks. Each respondent was presented with three choice tasks from one of the 20 blocks. A respondent could either participate in the public transport or the paratransit stated choice experiment depending on their respective mode usage frequencies.
3.3 Data collection
The surveys were conducted in Kampala (for Uganda) and Dhaka (for Bangladesh) between August and September 2021, immediately after the second wave of the pandemic. 3 Although the same survey questionnaire was used in both Kampala and Dhaka, there were slight differences in the tool itself and the data collection approaches. In Kampala, respondents were interviewed at their homes or places of work, while in Dhaka, they were intercepted and interviewed at bus stops or informal motorcycle taxi stands. Respondents were asked to provide consent before being interviewed, and the typical survey length was 15 – 20 minutes. Ethical approval of the design was obtained from the University of Leeds prior to the start of actual data collection. All health and safety precautions were maintained to keep both the surveyor and the respondents safe during the data collection process.
Table 3 presents the summary statistics of the survey data. Modal share and characteristics of users for different modes are not available in either cities. As such we included some socio-demographics from the national population. From Table 3, it appears that the Kampala sample is reasonably representative. The large gender imbalance in the Dhaka sample is due to the very low proportion of female users in public transport, and an even lower proportion in motorcycle taxis; anecdotal evidence suggests they are likely representative of the respective modes, too. Similarly, motorcycle taxis are more popular among the younger population in Dhaka, hence the larger share of that group in the sample. It is possible that young people were over-represented and older group under-represented in the public transport user sample in Dhaka.Table 3 Summary statistics of the survey data in comparison with national data (PopulationPyramid.net, 2020)
Demographic attribute Kampala, Uganda Dhaka, Bangladesh
Public Transport Para-transit National Population Public Transport Para-transit National Population
Total sample size (No. of respondents) 695 674 576 637
Average fare (Standard deviation) in Current USD* 0.54(0.34) 0.87 (0.51) 0.27(0.20) 1.63 (0.60)
Proportion of respondents by gender (%)
Male 35.3 48.7 49.3 89.4 94.7 50.6
Female 64.7 51.3 50.7 10.6 5.3 49.4
Proportion of respondents by parenthood (%)
With children 78.4 65.6 35.9 27.5
Without children 21.6 34.4 64.1 72.5
Proportion of respondents by age (%)
18 - 29 years 39.4 50.3 40.5 64.6 71.9 28.3
30 - 39 years 39.4 33.2 26.2 21.4 19.5 25.2
40 - 49 years 14.0 11.9 16.1 11.6 7.7 19.7
50+ years 7.2 4.6 17.1 2.4 0.9 26.8
* The calculation is based in central bank exchange rates of 03 December 2021 for Uganda (Bank of Uganda, 2021) and 02 December 2021 for Bangladesh (Bangladesh Bank, 2021).
4 Modelling framework
4.1 Utility model
The modelling in this study is based on the random utility theory (Marschak, 1960). Let Unik be the utility of individual n derived from choosing alternative i in choice situation k. This can be expressed as:
Unik=Vnik+εnik where k=1,2,⋯,K, i=1,2⋯,J, and n=1,2,⋯,N (1)
Where Vnik is the observed utility of alternative i for individual n in choice situation k and εnik is the random (unobserved) component of utility. Vnik has been expressed as a function of the travel cost (fare) and the COVID-19 mitigation attributes as follows:For public transport (buses and minibuses)
(2) Vnik=βFAREFAREnik+βSDSDnik+βSNTSNTnik
For paratransit (motorcycle taxis)
(3) Vnik=βFAREFAREnik+βSHIELDSHnik+βHLMHLMnik
Where FAREnik is the fare of alternative i for individual n in choice situation k, SDnik, SNTnik, SHnik, and HLMnik are dummy variables indicating the implementation status/provision of social distancing (passenger loading at half capacity), better sanitisation measures, the transparent rider/passenger shield, and a cleaned/disinfected helmet, respectively, in alternative i for individual n in choice situation k. A dummy variable is assigned a value of 1 if the associated mitigation measure is being implemented; otherwise, it is assigned a value of 0. The β s are the model parameters to be estimated. These were interacted with the socio-demographics of the respondents as discussed later in Section 5 to identify any potential heterogeneity in preferences.
Assuming that the error terms (εnik) are distributed independently and identically across alternatives and individuals using a type I extreme value distribution and considering that the choices follow the unordered response mechanism, the choice probabilities can be calculated using the multinomial logit (MNL) model (McFadden, 1974).(4) Pnik=expVnik∑j∈CnkexpVnjk
Where Pnik is the probability of individual n choosing alternative i in choice situation k, and Cnk is the choice set of individual n in choice situation k.
Given the choice probabilities, the model parameters can be determined by maximising the log-likelihood function. However, in this case, we have several choice tasks (choice situations) per respondent, therefore, we need to capture the panel effect (that there are several observations per respondent) in the log-likelihood function as shown below.(5) LLβ=∑n∑k∑iZniklnPnik
Where dummy variable znik=1 if individual n makes chooses alternative i in choice situation k, otherwise znik=0.
4.2 Willingness to Pay (WTP) estimation
WTP is the maximum price a consumer is willing to pay for a product or a service (Stobierski, 2020). The metric represents an upper threshold beyond which consumers will not pay a higher price and is widely used for transport policymaking. In the context of this study, WTP values help in quantifying the maximum price users would be willing to pay for the different COVID-19 mitigation measures in monetary terms (i.e. the monetary benefits attached to the mitigation measures). The WTP values for particular mitigation measures can be estimated as ratios of the partial derivatives of the applicable systematic utility functions with respect to the mitigation measure and the fare as shown in Equations (6), (7), (8), (9):For social distancing (half capacity) in public transport
(6) WTPSD=∂Vnik/∂SDnik∂Vnik/∂FAREnik=βSDβFARE
For better sanitisation measures in public transport
(7) WTPSNT=∂Vnik/∂SNTnik∂Vnik/∂FAREnik=βSNTβFARE
For the rider/passenger shield on motorcycle taxis
(8) WTPSHIELD=∂Vnik/∂SHnik∂Vnik/∂FAREnik=βSHIELDβFARE
For provision of cleaned/disinfected helmets on motorcycle taxis
(9) WTPHLM=∂Vnik/∂HLMnik∂Vnik/∂FAREnik=βHLMβFARE
5 Results and discussion
5.1 Preliminary insights from the data
5.1.1 Factors in travel choice decisions
During the surveys, respondents were asked to mention the most important factor affecting their travel choice decisions before and during the COVID-19 pandemic. Figure 1 summarises the responses to this question in Kampala and Dhaka. In both cities, travel time and cost were the most important factors before the pandemic, followed by convenience and general cleanliness. However, since the pandemic started, the importance of travel time and cost has diminished in both cities, more so in Dhaka. Social distancing and the wearing of face masks by all travellers have emerged as very important mode choice factors during the pandemic.Figure 1 Distribution of the most important factors affecting travel choice decisions before and during the COVID-19 pandemic
5.1.2 Attitudes towards the motorcycle taxi shield
The surveys also captured the attitudes of the respondents towards the novel concept of using shields between motorcycle driver and passengers to mitigate COVID-19 risk exposure for the passengers. Figure 2 summarises the attitudes towards the shields in Kampala and Dhaka. In both cities, majority of the respondents agree that motorcycle travel in the current circumstances presents a high risk of COVID-19 transmission and that the installation of a shield between the rider and passenger is likely to minimise the risk of transmission. However, there seems to be consensus in both cities that the shield has the potential to harbour micro-organisms and can be a source of new infections. It should be noted that although the shield offers susbtantial reduction in airborne COVID-19 exposure for motorcycle taxi passengers (Hetherington, et al., 2021), it does not offer absolute protection (e.g. via the fomite route), and there would be a need to sensitize the public about this. This also means that the shields would have to be sanitised regularly if installed on motorcycle taxis, which would have an impact on the operating costs.Figure 2 Attitudes towards motorcycle taxi shield
Finally, there is a general concern in both cities, more so in Dhaka, that the shield could compromise the overall safety of motorcycle taxis as it might cause potential aerodynamic issues and uncomfortable seating positions. In general, residents from Kampala expressed stronger preferences compared to those from Dhaka, possibly due to cultural differences.
5.2 Model specification
The key variables included in the utility models for public transport and paratransit have been presented in Equations (2), (3), respectively. Several interactions of each of these variables with the socio-demographics of the respondents have been tested to understand how sensitivities vary across demographic groups. The demographic attributes collected during the survey include the gender, age, level of education, employment status, income and parenthood status of the respondents.
All the variables in the utility models for public transport and paratransit have been successfully interacted with each of the collected demographic attributes, except income, in both cities. This is likely because majority of the respondents were not comfortable with disclosing their income. At this level of demographic interaction, all the model results are satisfactory, with the expected parameter signs and acceptable levels of statistical significance. An attempt was made to interact the utility model parameters with combinations of demographic attributes (e.g. age - gender combinations); however, there were no significant gains in model fit.
5.3 Model results
This section presents the results of the best comparable models between the two cities. For public transport, these are the models with parenthood interactions (i.e. individuals with or without children); while for paratransit, the models include age-group interactions. 4
5.3.1 Public transport (Buses and Minibuses)
Table 4 presents the estimation results of the public transport models for both cities showing that all the parameter estimates, and the overall models are statistically significant at the 95% level of confidence. A statistically significant positive parameter sign indicates that the presence of or an increase in the value of the variable increases the probability of choosing an alternative, while a negative parameter sign indicates the opposite effect.Table 4 Estimation results of the public transport model
Variable Kampala, Uganda Dhaka, Bangladesh
Parameter t-stat Parameter t-stat
Social distancing (half capacity)
Individuals with children 1.1532 13.07 1.1260 7.47
Individuals without children 0.8542 4.90 1.6070 13.71
Mandatory hand sanitisation plus increased cleaning/disinfection frequency of surfaces
Individuals with children 0.2941 4.31 0.3306 3.10
Individuals without children 0.3815 3.12 0.3386 4.30
Fare*
Individuals with children -0.1932 -3.52 -0.0261 -3.41
Individuals without children -0.4057 -3.03 -0.0477 -6.21
Measures of fit in estimation
Number of observations 2080 1720
Number of decision makers 695 576
LL(0) -1441.75 -1192.21
LL(F) -1250.22 -960.11
Number of parameters 6 6
ρadj2 w.r.t LL(0) 0.1287 0.1896
LR w.r.t LL(0) 383.06 464.20
p-value of LR <0.0001 <0.0001
* For Uganda, the fare is in UGX – Uganda Shillings (x1000), while for Bangladesh, it is in BDT – Bangladeshi Taka
From Table 4, it is observed that the social distancing and the sanitisation parameters have positive signs, while the fare parameters have negative signs in both cities. This implies that keeping all other factors constant, the implementation of social distancing and sanitisation measures improves the attractiveness of a public transport alternative when compared to alternatives where such mitigation measures are not being implemented. An increase in fare reduces the attractiveness of a public transport mode alternative, as is expected. This shows that public transport users make a trade-off between safety and cost. The results also suggest that unsustainable fare increments for COVID-19 mitigation can reduce the net benefit from safety improvements, thereby prompting users to settle for cheaper and unsafe alternatives, which would compromise the gains made in fighting the pandemic.
Table 4 also shows that the social distancing parameters generally have a higher magnitude compared to the sanitisation parameters. This indicates that public transport users in both cities are more concerned about changes in the social distancing measures than they are to changes in the sanitisation measures. Individuals with children are generally more sensitive to COVID-19 mitigation measures and less sensitive to changes in the fare when compared to individuals without children. This could be explained by the higher degrees of responsibility that individuals with children have, which makes them avoid risk and value their safety more (Eibach & Mock, 2011).
5.3.1.1 Paratransit (Motorcycle taxis)
Table 5 presents the estimation results of the paratransit models for both cities showing that most of the parameter estimates and the overall models are statistically significant at the 95% level of confidence.Table 5 Estimation results of the paratransit model
Variable Kampala, Uganda Dhaka, Bangladesh
Parameter t-stat Parameter t-stat
Provision of a transparent shield between the rider and the passenger
18 - 29 years 0.7267 7.62 1.0276 10.96
30 - 39 years 0.6360 5.66 1.4013 8.33
40 - 49 years 0.6406 3.95 0.8372 3.23
50+ years 1.3408 4.04 1.8314 3.05
Provision of a cleaned/disinfected helmet at the start of each trip
18 - 29 years 0.1841 2.28 0.7015 9.67
30 - 39 years 0.1226 1.31 0.9377 6.04
40 - 49 years 0.5529 3.62 0.7025 3.48
50+ years 1.1513 3.79 1.3937 3.21
Fare*
18 - 29 years -0.3490 -6.83 -0.0208 -13.10
30 - 39 years -0.2349 -3.61 -0.0250 -7.94
40 - 49 years -0.2462 -3.28 -0.0082 -3.03
50+ years -0.9047 -4.70 -0.0208 -1.87
Measures of fit in estimation
Number of observations 2013 1905
Number of decision makers 674 637
LL(0) -1395.31 -1320.45
LL(F) -1285.42 -996.48
Number of parameters 12 12
ρadj2 w.r.t LL(0) 0.0702 0.2363
LR w.r.t LL(0) 219.78 647.94
p-value of LR <0.0001 <0.0001
* For Uganda, the fare is in UGX (x1000), while for Bangladesh, it is in BDT
The shield and helmet cleaning parameters both have positive signs, while the fare parameters have negative signs in both cities. This implies that keeping all other factors constant, the provision of a transparent shield between the rider and the passenger and a cleaned/disinfected helmet at the start of each trip would improve the attractiveness of motorcycle taxis compared to those that do not offer such provisions. As before, an increase in the fare reduces the attractiveness of the alternative.
The parameters associated with the shield generally have a higher magnitude compared to those associated with cleaned helmets for individuals within the same age group. This shows that motorcycle taxi users in both cities would be more sensitive to changes in the provision of the shields than they would be to changes in the provision of cleaned/disinfected helmets. Suspicions about whether cleaning operations would actually take place in reality as stated in the choice experiment could cause the observed lower preference for the helmet cleaning option.
Although there are no clear patterns of how preferences vary across age groups, it can be seen that individuals who are 50 years and above would generally be more sensitive to changes in the provision of transparent rider/passenger shields and cleaned/disinfected helmets compared to younger individuals. In the case of Uganda, the same group are also the most sensitive to changes in fare.
5.4 Willingness to pay
As mentioned earlier, WTP values help in quantifying the monetary benefits attached to the different COVID-19 mitigation measures. These values have been estimated using formulae described earlier in Section 4.2.
This section reports the WTP estimates derived from the model results presented in Section 5.3 above. We report the estimates in the local currencies of each city, as well as their equivalent values in current United States Dollars (USD) to facilitate comparison between the cities.
5.4.1 Public transport (Buses and Minibuses)
Table 6 presents the WTP estimates for COVID-19 mitigation in public transport in Uganda and Bangladesh. From the table, it is observed that the WTP for social distancing measures is higher than that for sanitisation measures in both cities. In addition, it is observed that individuals with children generally have a higher WTP compared to individuals without children. As explained earlier, this could be attributed to the relatively higher degree of risk aversion behaviour among parents (Eibach & Mock, 2011).Table 6 Willingness to pay estimates for COVID-19 mitigation in public transport
Mitigation measure Kampala, Uganda Dhaka, Bangladesh
UGX Current USD* BDT Current USD*
Social distancing (half capacity)
Individuals with children 5,950 1.67 43.17 0.50
Individuals without children 2,100 0.59 33.72 0.39
All (average) 4,700 1.32 36.80 0.43
Mandatory hand sanitisation plus increased cleaning/disinfection frequency of surfaces
Individuals with children 1,500 0.42 12.68 0.15
Individuals without children 950 0.27 7.10 0.08
All (average) 1,350 0.38 8.61 0.10
* The calculation is based in central bank exchange rates of 03 December 2021 for Uganda (Bank of Uganda, 2021) and 02 December 2021 for Bangladesh (Bangladesh Bank, 2021).
We also observe that the WTP (in current USD) for COVID-19 mitigation in public transport is generally higher in Uganda compared to Bangladesh, despite the former’s lower GDP per capita compared to the latter (The World Bank, 2020). This could be attributed to the Government of Uganda’s widely publicised campaigns and activism against COVID-19. Also, it may be noted that public transport in Kampala is dominated by small buses (matatus), while in Dhaka, it is dominated by large buses, where seats are more spaced out compared to the matatus. Therefore, it is likely that public transport users in Kampala are more conscious and concerned about the spread of COVID-19 infection than those in Dhaka.
It may be noted that the average WTP for social distancing is more than 100% of the current average public transport fares in both cities, while the average WTP for sanitisation measures is approximately 70% and 36% of the current average public transport fares in Kampala and Dhaka, respectively. This again indicates potentially higher concerns about safety in Kampala. It may be noted that in order to accommodate the loss of revenue from the capacity restrictions in public transport, fares were increased by 100% in Kampala and 60% in Dhaka. Therefore, the increases in fare attributed to the COVID-19 social distancing requirements imposed by the governments of both countries were within the maximum threshold users were willing to pay for safe travel in the context of the pandemic.
5.4.1.1 Paratransit (Motorcycle taxis)
Table 7 presents the WTP estimates for COVID-19 mitigation in paratransit in Kampala and Dhaka. It is observed that the WTP for shields between riders and passengers is higher than that for cleaned/disinfected helmets in both cities. As can be expected from the parameters estimates earlier, the WTP in both cities generally increases with age up to a certain age group (40 - 49 years), after which it starts to fall.Table 7 Willingness to pay estimates for COVID-19 mitigation in paratransit
Mitigation measure Kampala, Uganda Dhaka, Bangladesh
UGX Current USD BDT Current USD
Provision of a transparent shield between the rider and the passenger
18 - 29 years 2,100 0.59 49.29 0.57
30 - 39 years 2,700 0.76 56.12 0.65
40 - 49 years 2,600 0.73 101.83 1.19
50+ years 1,500 0.42 88.22 1.03
All (average) 2,200 0.62 53.67 0.63
Provision of a cleaned/disinfected helmet at the start of each trip
18 - 29 years 550 0.15 33.65 0.39
30 - 39 years 500 0.14 37.55 0.44
40 - 49 years 2,250 0.63 85.45 1.00
50+ years 1,250 0.35 67.14 0.78
All (average) 800 0.22 36.40 0.42
* The calculation is based in central bank exchange rates of 03 December 2021 for Uganda (Bank of Uganda, 2021) and 02 December 2021 for Bangladesh (Bangladesh Bank, 2021).
Although the WTP trends between the two cities differ across age groups, the average WTP for the rider/passenger shield is almost the same in the two cities. Respondents were – on average – willing to pay around UGX 2,200 (approx. USD 0.62) per trip in Uganda and BDT 54 (approx. USD 0.63) per trip in Bangladesh for the shields. Although the WTP values (in current USD) are similar in the two cities, the average WTP for the safety shields is approximately 71% and 38% of the current average motorcycle taxi fares in Kampala and Dhaka, respectively. This indicates a relatively higher value attached to safety provisions in Uganda.
On the other hand, the average absolute WTP for a cleaned/disinfected helmet is much lower in Kampala compared to Dhaka. However, the average WTP for cleaned/disinfected helmets is approximately 26% of the current average motorcycle taxi fares in both cities. This indicates a similar level of valuation attached to the safety provided by clean helmets.
6 Conclusions
In this paper, we investigated if there is a potential market for safety measures in the public transport and paratransit sectors and estimated the passengers’ willingness to pay for such mitigation measures. Distancing (or capacity) measures and provision of sanitization were considered for public transport and frequent cleaning of helmets was considered for motorcycle taxis. In addition, the WTP for a novel COVID19 mitigation measure in motorcycle taxis – a barrier between the rider and the passenger – was estimated.
Our questionnaire survey results show that there have been substantial changes in the relative importance of the factors for mode choice in both Kampala and Dhaka. For example, typical mode choice factors such as costs, travel time and convenience became less important during the pandemic and COVID-19 mitigation and safety measures became important considerations. Choice experimentation shows that the implementation or provision of COVID-19 mitigation measures clearly improves the attractiveness of the associated public transport or paratransit alternatives. These changes in preferences will likely hold not only for Bangladesh and Uganda, but also other countries during times of other similar virus-borne pandemics. How much of these shifts are permanent is an important area for future exploration.
Public transport and paratransit users in both cities are willing to pay extra for all the four COVID-19 mitigation measures considered in this study. In particular, the average willingness to pay for social distancing was more than 100% of the current average public transport fares in both cities. Given that this particular measure was implemented along with 100% and 60% increases in actual fares in Kampala and Dhaka, respectively, our results indicate that the fare increase was well within the users’ valuation of the provision. The relatively muted public discontent on fare increases in both cities possibly reflect this.
Increases in public transport fare is not popular whether in a developed or a developing country. The impact of the additional fare on the disadvantaged sections of the society cannot be captured by the WTP metrics, and there is a need to understand the welfare and transport poverty impacts of such increases during the times of emergencies, when income potential for low-income group is already substantially diminished. It may well be possible that the relatively high WTP is a result of the non-discretionary nature of travel during the pandemic.
On the other hand, without the fare increase, transport operators – which act purely on the basis of profit maximization – could refuse to provide the necessary precautionary measures on-board increasing the risks further. Indeed bus operators in Bangladesh forced the Bangladesh government to backtrack on its decision to run buses at reduced capacity without additional fare increases during the third wave of the pandemic (Adhikary, 2022). Such WTP estimates will likely be useful in such negotiations between the government and the operators, especially to understand the balance between government support and fare increases to recover the lost revenue for public transport operators, especially in Western countries, which provided large subsidies to private operators during COVID-19. WTP estimates also provide a maximum acceptable level of transport fares during a pandemic – any increase beyond these could result in covert practices of using unsafe modes by passengers, especially in the context of low income countries where monitoring and enforcement practices can be poor.
Adding shields or barriers between the rider and the passenger on motorcycle taxis was viewed positively by more than two-thirds of the respondents in both cities, although there were some concerns about aerodynamic safety and potential for infections via the fomite route. Choice modelling shows that respondents were – on average – willing to pay around UGX 2,200 (approx. USD 0.62) per trip in Kampala and BDT 54 (approx. USD 0.63) per trip in Dhaka for the shields. However, considering the lower average motorcycle taxi fare in Kampala, this shows relatively higher valuation for the shields in Kampala.
Our results show that there is a potential demand for safety as an attribute in the public transport and paratransit sectors, and transport operators could potentially pass on the costs of COVID-19 mitigation provisions to the passengers and defray some or all of the costs. At a time when governments in low-income countries are averse to impose further lockdowns, transport operators may be able to compete on the basis of safety features and partially mitigate the policy gaps.
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
We acknowledge the support from UK Research and Innovation as part of the Global Challenges Research Fund (GCRF) and Newton Fund Agile Response Call to address COVID-19 (No. EP/V043226/1). We are also grateful to the two anonymous reviewers for suggestions to improve the work.
Contributions
AB designed the choice experiments, estimated the models and led the writing. ZW conceived the project, obtained funding, and checked the model results. PM, FR, MI, CU collected the data. All contributed to the wider survey design and writing.
1 The term rider is often used to refer to a motorcycle taxi driver in Bangladesh and Uganda.
2 Increased ventilation or filtration rate in buses is not considered as the buses are almost always naturally ventilated and have windows open due to hot and humid conditions.
3 The survey design relied on retrospective questions to compare the pre-pandemic behaviour, which could be associated with recall issues. However given we had asked about the most frequent trip (and not infrequent ones), and the reasonable time span between the pre-pandemic period and the data collection, recall accuracy should not be a critical issue (Hipp, et al., 2020).
4 The results for the other models specified with gender, age, level of education and employment status interactions are available in supporting information.
==== Refs
References
Abdullah M. Dias C. Muley D. Shahin M. Exploring the impacts of COVID-19 on travel behavior and mode preferences Transportation research interdisciplinary perspectives 8 2020 100255
Adhikary, T. S., 2022. Buses at half capacity: Transport lobby prevails again. [Online] Available at: https://www.thedailystar.net/news/bangladesh/transport/news/buses-half-capacity-transport-lobby-prevails-again-2941236 Accessed February 2022].
Aloi A. Effects of the COVID-19 lockdown on urban mobility: Empirical evidence from the city of Santander (Spain) Sustainability 12 9 2020 3870
Ardiansyah, T. & Purnomo, J., 2020. Jakarta's motorcycle taxis get backpack shield for virus battle. [Online] Available at: [Accessed 17 January 2022].
Arellana, J., Marquez, L. & Cantillo, V., 2020. COVID-19 Outbreak in Colombia: An Analysis of Its Impacts on Transport Systems. Journal of Advanced Transportation, Volume 2020.
Arunwuttipong A. INTENTION TO USE AND WILLINGNESS TO PAY PREMIUM FOR ENHANCED DISINFECTION OF PUBLIC TRANSPORTATION Journal of Management Information and Decision Sciences 2021 1 8
Athumani, H., 2020. Ugandan Creates COVID Shield for Motorcycles. [Online] Available at: https://www.voanews.com/a/africa_ugandan-creates-covid-shield-motorcycles/6191949.html [Accessed 26 November 2021].
Awad-Núñez S. Post-COVID-19 travel behaviour patterns: impact on the willingness to pay of users of public transport and shared mobility services in Spain European Transport Research Review 13 1 2021 1 18
Bangladesh Bank, 2021. Exchange Rate of Taka. [Online] Available at: https://www.bb.org.bd/econdata/exchangerate.php [Accessed 04 December 2021].
Bank of Uganda, 2021. Exchange Rates. [Online] Available at: https://www.bou.or.ug/bou/bouwebsite/ExchangeRates/index.html [Accessed 04 December 2021]
Batool Z. The Effects of COVID19 on Public and Para-Transit Drivers in Developing Countries – A Case Study of Dhaka and Owerri Working Paper. 2022
Beck M.J. Hensher D.A. Insights into the impact of COVID-19 on household travel and activities in Australia – The early days of easing restrictions Transport Policy 99 2020 95 119 32836998
Beck M.J. Hensher D.A. Wei E. Slowly coming out of COVID-19 restrictions in Australia: Implications for working from home and commuting trips by car and public transport Journal of Transport Geography 88 2020 2020 102846
Bhaduri E. Modelling the effects of COVID-19 on travel mode choice behaviour in India Transportation research interdisciplinary perspectives 8 2020 100273
Bliemer M.C. Collins A.T. On determining priors for the generation of efficient stated choice experimental designs Journal of Choice Modelling 21 2016 10 14
Bucsky P. Modal share changes due to COVID-19: The case of Budapest Transportation Research Interdisciplinary Perspectives 8 2020 100141
Budd L. Ison S. Responsible Transport: A post-COVID agenda for transport policy and practice Transportation Research Interdisciplinary Perspectives 6 2020 2020 100151
Campisi T. The impact of COVID-19 pandemic on the resilience of sustainable mobility in Sicily Sustainability 12 21 2020 8829
Das S. A radical modal shift from public to private transport mode Transport Policy 109 2021 1 11
de Haas M. Faber R. Hamersma M. How COVID-19 and the Dutch ‘intelligent lockdown’change activities, work and travel behaviour: Evidence from longitudinal data in the Netherlands Transportation Research Interdisciplinary Perspectives 6 2020 100150
De Vos J. The effect of COVID-19 and subsequent social distancing on travel behavior Transportation Research Interdisciplinary Perspectives 5 2020 2020 100121
DILG, 2020. DILG: Another motorcycle barrier design for back-riding couples approved by NTF. [Online] Available at: https://www.dilg.gov.ph/news/DILG-Another-motorcycle-barrier-design-for-back-riding-couples-approved-by-NTF/NC-2020-1241 [Accessed 17 January 2022].
Dzisi E.K.J. Dei O. Adherence to social distancing and wearing of masks within public transportation during the COVID 19 pandemic Transportation Research Interdisciplinary Perspectives 7 2020 2020 100191
Ehebrecht D. Heinrichs D. Lenz B. Motorcycle-taxis in sub-Saharan Africa: Current knowledge, implications for the debate on “informal” transport and research needs Journal of transport geography 69 2018 242 256
Eibach R.P. Mock S.E. The vigilant parent: Parental role salience affects parents' risk perceptions, risk-aversion, and trust in strangers Journal of Experimental Social Psychology 47 3 2011 694 697
Fatmi M.R. COVID-19 impact on urban mobility Journal of Urban Management 9 3 2020 270 275
Hargreaves J.R. Lifting lockdown policies: A critical moment for COVID-19 stigma Global Public Health 15 12 2020 1917 1923 32985374
Hetherington R. Exposure risk analysis of COVID-19 for a ride-sharing motorbike taxi Physics of Fluids 33 11 2021 113319
Hipp L. Bünning M. Munnes S. Sauermann A. Problems and pitfalls of retrospective survey questions in COVID-19 studies Survey Research Methods 14 2 2020 109 114
KCCA, 2018. Public Transport a Challenge: Does KCCA Care?. [Online] Available at: https://www.kcca.go.ug/news/308/#.Yz6UzXZBxPY [Accessed 06 October 2022]
Kitara D.L. Ikoona E.N. COVID-19 pandemic, Uganda’s story The Pan African Medical Journal 35 Suppl 2 2020
Marschak, J., 1960. Binary Choice Contsraints on Random Utility Indicators. In: K. Arrow, ed. Stanford Symposium on Mathematical Methods in the Social Science. California: Stanford University Press
McFadden, D., 1974. Conditional logit analysis of qualitative choice behaviour. In: P. Zarembka, ed. Frontiers in Econometrics. New York: Academic Press, pp. 105-142
Mogaji E. Impact of COVID-19 on transportation in Lagos, Nigeria. Transportation Research Interdisciplinary Perspectives 6 2020 100154
MOH, 2020. Guidelines for use of public transportation during the COVID-19 pandemic. [Online] Available at: https://covid19.gou.go.ug/uploads/document_repository/authors/ministry_of_health/document/GUIDELINES_FOR_USE_OF_PUBLIC_TRANSPORTATION_DURING_THE_COVID-19_PANDEMIC.pdf [Accessed 26 November 2021].
MOH, 2022. The COVID-19 Situation Update and Adjustments on Travel Restrictions in Uganda, Kampala: Ministry of Health
New Age, 2022. Buses to run at half capacity across Bangladesh from Saturday. [Online] Available at: https://www.newagebd.net/article/159786/buses-to-run-at-half-capacity-across-bangladesh-from-saturday [Accessed 15 February 2022].
Olvera, L. D., Plat, D., Pochet, P. & Maïdadi, S., 2012. Motorbike taxis in the“ transport crisis” of West and Central African cities. EchoGéo, Volume 20.
Oreffice S. Quintana-Domeque C. COVID-19 Information, Demand and Willingness to Pay for Protective Gear in the UK Studies in Microeconomics 9 2 2021 180 195
Otto, A., 2020. Gov't Issues SOPs for Boda Bodas. [Online] Available at: https://ugandaradionetwork.com/story/govt-issues-sops-for-boda-bodas-operation [Accessed 26 November 2021].
Our World in Data, 2022. Coronavirus (COVID-19) Vaccinations. [Online] Available at: https://ourworldindata.org/covid-vaccinations [Accessed 17 January 2022].
Peters, K. et al., 2021. Final Report: Accelerating COVID-19 related ‘best practice’ in the urban motorcycle taxi sector in SubSaharan Africa, London: High Volume Transport Applied Research Programme, IMC Worldwide Ltd.
PopulationPyramid.net, 2020. Population Pyramids of the World from 1950 to 2100, Uganda. [Online] Available at: https://www.populationpyramid.net/uganda/2020/ [Accessed 06 October 2022].
Porter G. Women’s mobility and transport in the peripheries of three African cities: Reflecting on early impacts of COVID-19 Transport policy 110 2021 181 190 35721253
Reza F. Kawsar E. Mahabub T.B. COVID-19 impact and responses: Bangladesh 2020 Build Bangladesh s.l.
Ssali S.N. Gender, Economic Precarity and Uganda Government’s COVID-19 Response African Journal of Governance & Development 9 1.1 2020 287 308
Starkey, P., 2016. The benefits and challenges of increasing motorcycle use for rural access. Mombasa, Kenya, International Conference on Transport and Road Research.
Stobierski, T., 2020. WILLINGNESS TO PAY: WHAT IT IS & HOW TO CALCULATE. [Online] Available at: https://online.hbs.edu/blog/post/willingness-to-pay [Accessed 17 January 2022].
Teixeira J.F. Lopes M. The link between bike sharing and subway use during the COVID-19 pandemic: The case-study of New York's Citi Bike Transportation Research Interdisciplinary Perspectives 6 2020 100166
The World Bank, 2020. GDP per capita, PPP (current international $) - Bangladesh, Uganda. [Online] Available at: https://data.worldbank.org/indicator/NY.GDP.PCAP.PP.CD?locations=BD-UG [Accessed 05 December 2021].
Thombre A. Agarwal A. A paradigm shift in urban mobility: policy insights from travel before and after COVID-19 to seize the opportunity Transport Policy 110 2021 335 353
Tirachini A. Cats O. COVID-19 and Public Transportation: Current Assessment, Prospects, and Research Needs Journal of Public Transportation 22 1 2020 1 21 36118518
Traets, F. & Gil, D., 2020. Package ‘idefix’: Efficient Designs for Discrete Choice Experiments. [Online] Available at: https://cran.r-project.org/web/packages/idefix/idefix.pdf [Accessed 2021 May 03].
Wadud Z. The effects of e-ridehailing on motorcycle ownership in an emerging-country megacity Transportation research part A: policy and practice 137 2020 301 312 32508400
Wadud Z. Rahman S.M. Enam A. Face mask mandates and risk compensation: an analysis of mobility data during the COVID-19 pandemic in Bangladesh BMJ global health 7 1 2022 e006803
WHO, 2020. Q&A: How is COVID-19 transmitted?. [Online] Available at: https://www.who.int/vietnam/news/detail/14-07-2020-q-a-how-is-covid-19-transmitted [Accessed 25 November 2021].
Worldometer, 2022. COVID-19 CORONAVIRUS PANDEMIC. [Online] Available at: https://www.worldometers.info/coronavirus/ [Accessed 11 October 2022].
Zafri N.M. Khan A. Jamal S. Alam B.M. Impacts of the COVID-19 Pandemic on Active Travel Mode Choice in Bangladesh: A Study from the Perspective of Sustainability and New Normal Situation Sustainability 13 12 2021 6975
Zhang J. Hayashi Y. Frank L.D. COVID-19 and transport: Findings from a world-wide expert survey Transport policy 103 2021 68 85 33519127
| 0 | PMC9742211 | NO-CC CODE | 2022-12-14 23:29:59 | no | Transp Res Part A Policy Pract. 2023 Jan 12; 167:103561 | utf-8 | Transp Res Part A Policy Pract | 2,022 | 10.1016/j.tra.2022.103561 | oa_other |
==== Front
Aust Crit Care
Aust Crit Care
Australian Critical Care
1036-7314
1036-7314
Australian College of Critical Care Nurses Ltd. Published by Elsevier Ltd.
S1036-7314(22)00246-6
10.1016/j.aucc.2022.12.001
Research Paper
Nursing workforce deployment and ICU strain during the COVID-19 pandemic in Victoria, Australia
Topple Michelle RN a∗1
Jaspers Rose RN, MN (Crit Care Nsg) b1
Watterson Jason RN, PhD cf
McClure Jason MB ChB, MRCP, FRCA, FCICM, PGD Engineering def
Rosenow Melissa d
Pollock Wendy RN, PhD b2
Pilcher David MBBS FCICM efg2
a Department of Intensive Care, Austin Health, 145 Studley Rd, Heidelberg, Melbourne, Victoria, Australia
b School of Nursing and Midwifery, Monash University, Wellington Rd, Clayton, Melbourne, Victoria, Australia
c Department of Intensive Care, Peninsula Health, 2 Hastings Rd, Frankston, Victoria, Australia
d Adult Retrieval Victoria, 61-75 Brady St, South Melbourne, Victoria, Australia
e Department of Intensive Care, Alfred Hospital, 55 Commercial Rd, Melbourne, Victoria, Australia
f School of Public Health and Preventive Medicine, Monash University, 553 St Kilda Rd, Melbourne, Victoria, Australia
g Australian and New Zealand Intensive Care Society Centre for Outcome and Resource Evaluation, 1/277 Camberwell Rd, Camberwell, Melbourne, Victoria, Australia
∗ Corresponding author.
1 joint first authors.
2 joint last authors.
12 12 2022
12 12 2022
5 6 2022
29 11 2022
5 12 2022
© 2022 Australian College of Critical Care Nurses Ltd. Published by Elsevier Ltd. All rights reserved.
2022
Australian College of Critical Care Nurses 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
critical care nursing
intensive care unit
staffing levels
workforce
workload management
==== Body
pmcCRediT authorship contribution statement
MT - Conceptualisation; Investigation; Methodology; Project administration; Validation; Data entry; Writing - original draft; Writing - review & editing .
RJ - Conceptualisation; Investigation; Methodology; Project administration; Validation; Writing - original draft; Writing - review & editing.
JW - Conceptualisation; Investigation; Methodology; Project administration; Validation; Data entry; Writing - original draft; Writing - review & editing.
JMcC - Data curation, Writing - review & editing.
MR - Data curation, Writing - review & editing.
WP - Conceptualisation; Investigation; Methodology; Project administration; Validation; Writing - original draft; Writing - review & editing.
DP - Conceptualisation; Data curation; Formal analysis; Investigation; Methodology; Project administration; Validation; Visualisation; Writing - original draft; Writing - review & editing.
Introduction
The coronavirus (COVID-19) pandemic has had an unprecedented effect on the number of severely ill patients admitted to intensive care units (ICUs) globally. Consequently, surge models evolved to expand ICU capacity including space, beds, equipment and deployment of staff.[ [1], [2], [3], [4], [5], [6], [7] ] Early analyses in Australia suggested that treble the number of ICU beds might be required to accommodate the predicted surge in the critically ill population, but doing so would necessitate a 269% increase in registered nurses over the baseline.[ 8 ] In April 2020, to monitor ICU demand and capacity, the Critical Health Resource Information System (CHRIS) was developed as a collaboration between The Australian and New Zealand Intensive Care Society (ANZICS) and Ambulance Victoria, and funded by The Australian Government Department of Health.[ 9 ] In May 2020, Safer Care Victoria developed the ‘Coronavirus (COVID-19) Intensive Care Unit Surge Workforce Models Of Care Delivery’ guideline to address structured ICU workforce surge planning, which were updated in September 2021 (V2).[ 10 ]
ICU workforce shortages, particularly nursing, provided an ongoing discussion as a potential key limiter to ICU capacity, however workforce data were not collected in CHRIS. Consequently, through collaboration and consultation with the Victorian ICU Nurse Unit Managers Community of Practice, nursing workforce data were added in November 2021, reflecting the categories of surge workforce as outlined by the Safer Care Victoria Guidelines (V2).[ 10 ] The extent to which workforce redeployment and workforce expansion occurred, and the potential effect on delivery of 1:1 care for critically ill patients within the ICU, has not been described. This is important as there is substantial evidence that the number of critical care nurses impacts on patient outcomes, nurses and health services.11 , 12 For example, low levels of critical care nurses are associated with increased patient mortality, increased nosocomial infections, and higher hospital costs.[ 12 ] Additionally, poor ICU nurse staffing levels can affect nurse wellbeing, including increasing burnout, depersonalisation and emotional exhaustion.[ 13 ] Thus we sought to quantify the critical care nursing deficit associated with the pandemic surge-related increase in ICU capacity.(see Figs. 1 and 2 )Figure 1 Study period (1st December 2021 to 11th April 2022) where nursing skill-mix information was available, shown against overall number of patients in all 45 active Victorian ICUs, open ICU beds and numbers of COVID-19 patients in ICU between 1st Jun 2021 and 11th April 2022.
Explanatory footnote: The trendline for mean daily occupied ICU beds in all Victorian hospitals is shown by the blue dotted line and represents all ICU equivalent patients (i.e. all patients requiring 1:1 nursing + (0.5 x the number of patients requiring 1:2 nursing within ICU).
Figure 1
Figure 2 Nursing skill-mix distribution at study ICUs between 1st December 2021 to 11th April 2022.
Explanatory footnote: Over the study period which began during a period of high demand for ICU beds due to the COVID-19 pandemic, there was a decline in the number and proportion of the ICU nursing workforce who were redeployed nurses from areas outside ICU (light purple) or staff without critical care experience (dark purple).
Figure 2
Objectives
Our aim was firstly, to describe the extent and manner by which the increased demand for ICU care during the COVID-19 pandemic was met by ICU nursing workforce expansion in late 2021 and early 2022 in Victoria, secondly, to quantify provision of redeployed non-critical care nursing staff working in ICUs and thirdly, to attempt to identify factors associated with insufficient provision of critical care nursing staff.
Methods
Design and Setting
We conducted a retrospective cohort study at Victorian public and private, adult and paediatric hospital ICUs between 1st December 2021 and 11th April 2022. This study period was after the peak of the delta wave (peak in mid-October 2021 of 2,257 new community cases reported in Victoria) and during the omicron wave (peak in mid-January 2022 of 51,144 new community cases reported in Victoria).[ 14 ] Associated hospital admission data for all of 2021 and up to the end of the study period is in Figure One.
Data source
All data were extracted from CHRIS. ICU staff entered ‘snapshot’ aggregated census information about local critical care resources, demand and activity, between two and four times per day. No individual patient or outcome data were entered. ICU Nurse Unit Managers at all Victorian hospitals were also invited to voluntarily collect additional information about nurse staffing skill-mix at least once and up to four times per day. The number of staff providing direct patient care at the timepoint considered was entered. ICU nursing skill-mix was categorised into four groups as defined by the Safer Care Victoria guidelines V2.[ 10 ] • Group One: experienced critical care nursing staff including clinical nurse specialists, associate nurse unit managers, post-graduate critical care nurses and nurses with five plus years of current/continuous ICU experience
• Group Two: early career critical care nursing staff including foundation year/transition to ICU speciality nurses, 2021 post-graduate ICU students (employment study model), nurses with critical care experience not normally working in ICU pre-pandemic
• Group Three: redeployed nursing staff with no ICU experience (novice to ICU)
• Group Four: registered undergraduate students of nursing, enrolled nurses and allied health staff providing direct patient care
The mean daily values for each nursing skill-mix group were extracted. For the purposes of the study, Groups One and Two were considered as critical care registered nurses (CCRN). Groups Three and Four were considered as non-critical care registered nurses (non-CCRN). Both these groups reflect ICU nursing expansion as Group Three were redeployed nursing staff novice to ICU and Group Four were staff that would not normally be responsible for direct patient care of an allocated ICU patient.
The following data were also extracted as daily mean values for each hospital: number of patients requiring 1:1 critical care nursing and invasive mechanical ventilation,[ 15 ] the number high-dependency patients requiring 1:2 critical care nursing,[ 15 ] the number of patients with COVID-19 within each ICU and in other wards outside of ICU, the daily number of open ICU beds, the baseline number of open ICU beds prior to the pandemic, numbers of critical care staff unavailable due to COVID-19 illness/exposure and the ICU Activity Index. The ICU Activity Index is a calculated ‘score’ which indicates the acuity level within each ICU. A value over two represents a very busy ICU which is potentially under strain (Pilcher et al., 2021).[ 9 ] Daily mean occupancy was calculated as [total number of 1:1 nursed patients + (0.5 x 1:2 high dependency patients)] divided by total daily number of open ICU beds.
Outcomes
The primary outcome was ‘insufficient ICU skill-mix’. This outcome was met whenever a site had more patients needing 1:1 critical care nursing care than the mean daily number of CCRN staff. For instance, a site which had 10 patients who required 1:1 critical care nursing and only had nine CCRNs available, would be considered as having ‘insufficient skill-mix’ on that day. The number of high dependency patients requiring 1:2 nursing within the ICU was not considered in this calculation.
Two secondary outcomes were considered. The secondary outcome of ‘insufficient ventilated skill-mix’ was met whenever a site had more patients needing invasive mechanical ventilation than the number of CCRN staff. The secondary outcome of ‘reduced ICU nursing ratio’ was met whenever there were more ‘ICU equivalent’ patients than total number of bedside nursing staff. For instance, a site with 10 patients who required 1:1 critical care nursing and six high-dependency patients who required 1:2 nursing (i.e. 13 ‘ICU equivalents’) and a total of 12 or fewer nursing staff involved in direct patient care (from all four groups) would be classified as having ‘reduced ICU nursing ratio’ for that day.
Sample size
A convenience sample based on the number of hospitals who contributed data for more than two months during the five-month study period was chosen.
Statistical analysis
Data are presented as percentage (number/proportion) for categorical data or median (inter-quartile range) for continuous data. All continuous data were assessed for normality and found to be non-parametrically distributed. Chi square, Wilcoxon rank sum and Kruskal-Wallis tests were used to compare groups depending on type of data and number of groups examined. After assessing for co-linearity, mixed effects logistic regression was used to determine variables independently associated with the primary outcome, with site entered as a random effect with a random intercept per facility applied. A P value of <0.05 was considered statistically significant. All data were analysed using Stata version 16.1 College Station, Texas. No imputation for missing data was performed.
Following the main analysis which assessed the impact of observed redeployment of staff into ICUs during the study period, a counterfactual analysis was conducted to assess hypothetical nursing ratios that would have been present in the absence of any redeployed non-critical care trained staff. This analysis re-calculated nursing ratios assuming the daily number of critical care trained staff (Groups One and Two) and number of patients present within each ICU remained unchanged but there were zero redeployed staff present from Groups Three and Four. This was in recognition of the fact that without redeployment of nursing staff within the ICU to meet the additional demand for ICU beds, nursing ratios would have been even lower than measured.
Subgroups
Outcomes for all ICUs for the whole study period are reported, and for two time periods: peak (1st December 2021 to 19th February 2022) and post-peak (20th February to 11th April 2022, after daily Victorian ICU director cluster demand meetings were ceased and switched to twice weekly) phases. Outcomes are also reported by hospital type. Hospitals were classified into one of five groups as tertiary, metropolitan, rural/regional, private or paediatric.
Funding and ethics
The study was self-funded by the researchers and was approved as a low-risk project by the Human Research and Ethics Committee of the Alfred Hospital (HREC 246/22).
Results
All 47 ICUs in Victoria contributed daily data about ICU resources and activity. Of these, 32 hospitals also provided information on nursing skill-mix. Eight were excluded because of contributing fewer than two months staffing data each, leaving 24 participating ICUs over the 132-day study period. These 24 participating ICUs represent 71% (344/486) of all the baseline ICU beds in Victoria. In total, nursing skill-mix data was available for 86% (2725/3168) of all days. At least 20 of the 24 hospitals contributed nursing skill-mix data on 89% (118/132) of the days, but there were only four days when all 24 hospitals contributed data. Appendix Table One shows the basic characteristics and number of ICU beds at study hospitals. All hospital types were represented.
ICU capacity, patient needs and staffing data were highly variable (Figure One; Table One). In summary, the median number of open ICU beds was 8, with a maximum of 65; the median daily number of ventilated patients was 2.5, with a maximum of 46.7, and the median daily number of COVID-19 patients in each ICU was 1, with a maximum of 32.3. The total daily number of patients and open ICU beds in all Victorian ICUs between 1st June 2021 and 11th April 2022, and the study period (1st December 2021 to 11th April 2022) over which nursing skill-mix data were available for contributing hospitals are shown in Figure One. The daily number of all 1:1 nursed, ventilated and COVID-19 patients in the 24 study hospital ICUs during the study period are shown in Appendix Figure One.
Table One shows a comparison of days when there was insufficient ICU skill-mix compared to days when this was not present. Insufficient ICU skill-mix occurred more commonly when there were more ICU beds open, lower numbers of vacant ICU beds, higher occupancy and ICU activity, more patients requiring 1:1 nursing, more ventilated patients, more COVID-19 patients and lower numbers of high dependency patients (see Table 1 ).Table 1 Staffing and activity characteristics of ICUs on days with and without insufficient ICU skill-mix (more patients needing 1:1 critical care nursing than available CCRNs: primary study outcome).
Table 1Primary outcome: Insufficient ICU skill-mix All days No Yes P value
Days with more patients needing 1:1 nursing than CCRNs, % (number) - 89.7% (2445/2725) 10.3% (280/2725) N/A
Secondary outcome: Insufficient ventilated skill-mix 0.7% (19/2725) 0% (0/2445) 6.8% (19/280) N/A
Secondary outcome: Reduced ICU nursing ratio 3% (82/2725) 1.8% (43/2445) 13.9% (39/280) <0.001
Mean daily CCRNs per site 8 (5-17) [1, 70] 8 (5-17) [2, 70] 8 (4-23) [1, 58] 0.33
Group 1 (CCRN nurses - experienced) 7 (4-13) [1, 48] 7 (4-13) [1, 48] 6 (3-19) [1, 43] 0.98
Group 2 (CCRN nurses - early career) 2 (1-4) [0, 33] 2 (1-4) [0, 33] 2 (1-5) [0, 30] 0.62
Mean daily non-CCRNs per site 1 (0-2) [0, 31] 0 (0-2) [0, 20] 4 (2-7) [0, 31] <0.001
Group 3 (non-CCRN redeployed nurses) 0 (0-2) [0, 31] 0 (0-2) [0, 20] 4 (2-7) [0, 31] <0.001
Group 4 (RUSONs, enrolled nurses & allied health) 0 (0-0) [0, 4] 0 (0-0) [0, 4] 0 (0-0) [0, 2] <0.001
Critical care staff unavailable due to COVID-19 illness or furlough 1.6 (0-4) [0, 131.8] 1.5 (0-4) [0, 131.8] 2 (0-5) [0, 61] 0.63
Open available ICU beds 8 (5.5-15.5) [0, 65.3] 8 (5.5-15) [0, 63.8] 11.8 (5-27.2) [4, 65.3] 0.001
ICU beds open over 'business as usual' -2 (-4-0.3) [-23, 19.3] -2 (-4-0) [-23, 17.8] 1 (-2-2.6) [-18, 19.3] <0.001
Vacant ICU beds 1.8 (0.8-3) [0, 12.7] 2 (1-3) [0.1, 12.7] 1 (0.2-1.8) [0, 8.3] <0.001
ICU occupancy 83% (64-93.5) [0, 200] 81% (60-92) [0, 200] 96% (88.6-100) [66.7, 150] <0.001
Activity Index 1.1 (0.6-1.5) [0, 2.8] 1 (0.5-1.5) [0, 2.8] 1.6 (1.2-1.9) [0.7, 2.6] <0.001
Days with Activity Index > 2.0, % (number) 6.5% (178/2725) 5.1% (125/2445) 18.9% (53/280) <0.001
'ICU equivalent' patients 6.5 (3.5-14) [0, 63.8] 6 (3.3-13.3) [0, 62] 10.8 (5-26) [2.7, 63.8] <0.001
Patients in ICU needing 1:1 critical care nursing 5 (2-13.3) [0, 62.3] 4.5 (1.5-12.8) [0, 60] 9.8 (5-24.8) [2.5, 62.3] <0.001
Ventilated patients 2.5 (0.7-7) [0, 46.7] 2 (0.5-6.5) [0, 43.8] 6 (2-16.2) [0, 46.7] <0.001
COVID-19 patients in each ICU 1 (0-3) [0, 32.3] 0.5 (0-2) [0, 31] 3 (1-9) [0, 32.3] <0.001
HDU patients (1:2 nursing) in each ICU 2.3 (0.7-4.3) [0, 14] 2.5 (1-4.5) [0, 14] 0 (0-2) [0, 12] <0.001
COVID-19 patients on the hospital ward 6 (0-16.8) [0, 173] 6 (0-16) [0, 173] 10.1 (0-23) [0, 119.5] 0.10
Data reported are median (interquartile range) [minimum, maximum] daily values per site (unless otherwise stated).
ICU – Intensive Care Unit, HDU – High Dependency Unit, CCRN – Critical Care Registered Nurse, COVID-19 – Corona Virus Disease 2019, RUSON – Registered undergraduate student of nursing.
Overall, insufficient ICU skill-mix occurred on 10.3% (280/2725) days at 66.7% (16/24) of ICUs, most commonly during the peak phase from December to mid-February (Table 2 ). Insufficient ventilated skill-mix occurred on 0.7% (19/2725) days overall at 37.5% (9/24) of ICUs and was more common during the peak phase. Reduced nursing ratios occurred on 3% (82/2725) days at 75% (18/24) of ICUs but were equally common in both phases. Counterfactual analysis suggested that had there been no redeployment of nursing staff to ICU, reduced nursing ratios would have occurred on 15.2% (415/2725) days at 91.7% (22/24) ICUs and more commonly during the peak phase (see Table 3 ).Table 2 Outcomes for the whole study period (1st December 2021 to 11th April 2022) and for peak (1st December 2021 to 19th February 2022) and post-peak (20th February to 11th April 2022) phases, showing the number of days on which the outcome was observed and the number of ICUs where this outcome occurred.
Table 2 Days with this outcome ICUs with this outcome P value for
Outcomes Time Period Proportion Number Proportion Number difference between phases
Primary outcome: Insufficient ICU skill-mix
Whole study period (1st December 2021 to 11th April 2022) 10.3% 280/2725 66.7% 16/24
Peak phase (1st December 2021 to 19th February 2022) 14.3% 242/1690 62.5% 15/24 <0.001
Post peak phase (20th February 2022 to 11th April 2022) 3.7% 38/1035 45.8% 11/24
Secondary outcome: Insufficient ventilated skill-mix
Whole study period (1st December 2021 to 11th April 2022) 0.7% 19/2725 37.5% 9/24
Peak phase (1st December 2021 to 19th February 2022) 1.1% 18/1690 33.3% 8/24 0.003
Post peak phase (20th February 2022 to 11th April 2022) 0.1% 1/1035 4.2% 1/24
Secondary outcome: Reduced ICU nursing ratio
Whole study period (1st December 2021 to 11th April 2022) 3.0% 82/2725 75% 18/24
Peak phase (1st December 2021 to 19th February 2022) 3.0% 51/1690 70.8% 17/24 0.86
Post peak phase (20th February 2022 to 11th April 2022) 3.0% 31/1035 45.8% 11/24
Counterfactual analysis: Days with reduced ICU nursing ratio if there had been no redeployment of nurses into ICU
Whole study period (1st December 2021 to 11th April 2022) 15.2% 415/2725 91.7% 22/24
Peak phase (1st December 2021 to 19th February 2022) 19.9% 336/1690 91.7% 22/24 <0.001
Post peak phase (20th February 2022 to 11th April 2022) 7.6% 79/1035 70.8% 17/24
ICU = Intensive Care Unit.
Insufficient ICU skill-mix = Days with more patients needing 1:1 critical care nursing than available Critical Care Registered Nurses (CCRNs), % (number).
Insufficient ventilated skill-mix = Days with more ventilated patients than available Critical Care Registered Nurses (CCRNs), % (number).
Reduced ICU nursing ratio = Days with less than 1:1 overall critical care nursing ratio, % (number).
Table 3 Mixed effects multivariable logistic regression (with site as a random effect). Factors associated with having more patients needing 1:1 critical care nursing than CCRNs.
Table 3 Odds Ratio (95% CI) P value
Hospital type
Tertiary Reference group
Metropolitan 7.40 (0.34 - 160.8) 0.20
Rural/regional 0.02 (0.00 - 0.80) 0.038
Private 1.12 (0.06 - 22.48) 0.94
Daily number of critical care staff unavailable due to COVID-19 illness or furlough 1.02 (1.00 - 1.04) 0.044
Number of baseline 'business as usual' ICU beds 0.94 (0.82 - 1.08) 0.39
Daily ICU beds open over 'business as usual' 1.28 (1.16 - 1.41) <0.001
Daily ICU occupancy (%) 1.08 (1.06 - 1.09) <0.001
Daily number of ventilated patients 1.05 (0.97 - 1.13) 0.23
Daily number of COVID-19 patients in each ICU 1.13 (1.07 - 1.20) <0.001
Daily number of COVID-19 patients on the hospital ward 0.98 (0.97 - 0.99) 0.002
ICU – Intensive Care Unit.
COVID-19 – Corona Virus Disease 2019.
Table Three shows that after adjusting for confounders, days with insufficient ICU skill-mix were more likely to occur when there were more additional ICU beds open over the ‘business-as-usual’ number, when there was higher occupancy, higher numbers of COVID-19 patients within the ICU. The strongest factor was when additional ICU beds were open over the ‘business as usual’ number. Insufficient ICU skill-mix was less likely to occur in rural/regional sites and when there were relatively higher numbers of COVID-19 patients on the general wards outside of ICU.
Figure Two shows a declining daily number and proportion of non-CCRNs over the study period. Appendix Figure Two shows a progressive decline in the daily number and proportion of ICUs which had insufficient ICU skill-mix. The 10th of December was the day with the highest proportion of the workforce who were non-CCRNs (27%) and the greatest proportion/number of ICUs with insufficient ICU skill-mix (40%, 8/20). During the peak phase, there were fewer CCRNs, more non-CCRNs, higher occupancy and activity levels, and higher numbers of ventilated and COVID-19 patients than during the post-peak phase (Appendix Table Two). Insufficient ICU skill-mix was most commonly observed in tertiary and metropolitan ICUs and was rare in rural/regional, private and paediatric ICUs (Appendix Table Three).
Discussion
This study provides evidence for the first time, on the critical care nursing deficit created by pandemic surge models. Insufficient ICU skill-mix was present on over one in ten days during the recent pandemic wave December 2021-April 2022. There was a reduced ICU nursing ratio on many occasions, when there were insufficient numbers of nurses to care for patients in ICU according to the nurse patient ratio standards and demonstrated an insufficient population of skilled CCRNs in Victoria to provide care as the population and acuity of ICU patients increased. The redeployed nurses filled an important gap in meeting patient care needs and minimising reduced ratios, however ICU nurse staffing gaps remained during the study period.
Adapted models of care delivery in ICUs were common across the world during the COVID-19. 4 , 16 , 17 In countries which experienced a high COVID-19 disease burden, ICU capacity was rapidly doubled or even quadrupled.18 , 19 Whilst there are large numbers of published studies reporting on increasing ICU capacity in response to COVID-1918 , 19 and non-ICU staff education preparation and deployment,[ 20 ] there are few on the critical care nursing deficit created by pandemic surge staffing models. A single paper reported the association with critical care staffing and patient mortality during the COVID-19 pandemic, demonstrating that ICU nurse and medical staffing were important determinants of mortality in ICU.[ 21 ] This is consistent with pre-pandemic data, that patient mortality is associated with staff resources and workload in the ICU.[ 22 ] However, the nuances of what constitutes an “ICU nurse” when referring to ICU staffing and the implications for patient outcomes is poorly understood.[ 23 ] The clinical relevance for patient outcomes resulting from a deficit of critical care registered nurses in ICU requires further exploration.
Direct ‘one to one’ (1:1) nursing care is required for ICU patients who receive invasive therapies such as mechanical ventilation, renal replacement therapy, high dose vasoactive infusions or extra-corporeal membrane oxygenation.[ 15 ] This is also required for other patients with complex or high-risk needs such as potential airway compromise or severe delirium. In Australia, the bedside ICU nurse titrates and manages all equipment and medication associated with patient care. Consequently, ICU nursing is a postgraduate specialty qualification, not replaceable by pandemic ‘upskilling’ education programs. ICU care delivered by clinicians without adequate expertise, qualifications, training or support, may result in adverse outcomes.4 , 24 Patients in ICU with COVID-19 require more nursing time than non-COVID-19 patients[ 25 ] and pandemic staffing models have contributed to missed nursing care.[ 26 ] Excess mortality for patients with COVID-19 admitted to Australian ICUs during the peak phases of the pandemic in late 2021 has been reported.[ 27 ]
There was wide variation of ICU capacity, with some operating with fewer ICU beds than usual and others well over their usual bed number. This likely reflects both the Victorian Department of Health’s policy of 'streaming’ COVID-19 patients to specified hospitals and also reductions in elective surgery particularly at private hospitals. Insufficient ICU skill-mix occurred in over two-thirds of ICUs. Insufficient ICU skill-mix was more likely to occur when there were more additional ICU beds open over the ‘business-as-usual’ number, when there was higher occupancy, and with higher numbers of COVID-19 patients within the ICU. The increased demand for critical care support meant that patients with COVID-19 had to be admitted to ICU and cared for with whatever resources were available. Redeployment of nursing staff with or without critical care experience from outside the ICU was required in response to this increased demand. The ability to match provision of critical care trained nursing staff to the increased demand was likely limited by a fixed available number of critical care trained staff, organisational logistics of redeployment of staff both within and between hospitals and also by staff who were unavailable for work due to either having COVID-19 or being furloughed due to COVID-19 exposure.
Most ICUs experienced an occasion when they did not have enough nurses for recognised nurse:patient ratio standards.[ 11 ] Notably, if redeployed staff were not utilised during the study period, there would have been more occasions of reduced nurse patient ratios. As a response to an extreme health system emergency, redeployed staff were an important strategy to meet patient care needs. Of note, other countries have reported chronic ICU nurse under-staffing prior to the pandemic, which the pandemic has exacerbated.4 , 28 , 29 The ICU nurse vacancy rate across Australia prior to the pandemic was over 6%[ 27 ] so likewise, the Australian ICU nursing workforce was already depleted. Detailed information about the ICU nursing workforce has not been regularly captured before and continued collection of such data has the potential to provide valuable information regarding the ICU nursing workforce during regular operations and during other extraordinary events e.g. disasters.
There is a risk that pandemic-inspired new models of care will remain.[ 30 ] However, these new models place unreasonable burden on experienced critical care nurses and may contribute to burnout, disengagement,[ 31 ] and job dissatisfaction (intent to leave).[ 13 ] Our findings also highlight that quantifying the gap in service provision (critical care nurses to ICU patients) is highly relevant when considering future demands on ICU services and sustainability of ICU nurse staffing.
There are multiple reports of poor staff wellbeing associated with working during the pandemic, including in Australia.[ 32 ] As the pandemic surge demand peaks and troughs, there is a need for restorative care for staff whenever an opportunity arises. The Safer Care Victoria guidelines incorporate a ‘stand down and recovery’ phase, where healthcare services are expected to provide opportunities for staff to have personal leave and rest, and access to psychological supports and counselling.[ 10 ] Internationally, it is well recognised that nurses working to meet the pandemic surge require organisational support to assist their recovery.[ 33 ] It is unclear how the negative impacts on nurses from delivering healthcare in response to the pandemic will be addressed in Victoria.[ 24 ]
Strengths and limitations
CHRIS had good capability to allow quick initiation and collection of additional nursing staff data items. Whilst data entry for nursing staff items was voluntary, familiarity with using CHRIS and with the defined surge workforce groupings, provided high quality, consistent, generalisable and representative data across all care settings. However, although we report (Appendix Tables One and Three) and also adjusted for different hospital characteristics (Table Three), our study includes data from 24 out of a possible 47 ICUs, and thus may not be representative of all ICUs. There may be a bias related to capacity to provide data, for example, busier units may not have contributed data.
The ICU nursing staff data collection began after the pandemic surge associated with the October 2021 Delta wave, so it is possible that the worst of the ICU nursing deficit was not captured. Data gathered about nursing workforce was specific to nurses ‘directly involved in patient care’ only. CHRIS did not capture other nursing roles and resources like educators, managers, equipment nurses and outreach team members who are usually supernumerary to the nurse:patient ratio numbers and who are a vital support resource each shift. Thus, deficits in ICU nursing staff reported in our study refer only to bedside nursing and not the whole ICU. Numbers of supernumerary staff supporting nursing skill mix or redeployed to replace bedside nursing shortages were unknown. Furthermore, we have no data on overtime worked. Anecdotally, we are aware that double-shifts were common in ICUs during the study period to meet direct patient care needs. Additionally, as defined by the Safer Care Victoria guidelines, Group One critical care registered nurses included experienced ICU nurses who do not possess a postgraduate critical care nursing qualification. This limits our ability to interpret our findings as evidence suggests that a minimum proportion of postgraduate qualified ICU nurses providing direct care is of relevance.[ 11 ]
Furthermore, the timing of data entry on nursing skill-mix varied and the mean daily values may not account for variation over the day e.g. unknown if staffing worse at night than day. This also means some variation in interpretation for ICUs that have hybrid staffing models e.g. both 12-hr shifts and 8/8/10hr shift options. Due to the lack of data collection on supernumerary staff, we are unable to comment on how non-CCRNs were supported or supervised as they cared for ICU patients. We are also unable to report on the impact of pandemic staffing models on patient outcomes or nurse outcomes e.g. burnout.
Future research
We have demonstrated that it is feasible to collect high quality data regarding the ICU nursing workforce. Detailed information about the ICU nursing workforce has not been regularly captured before and has the potential to provide valuable information post-pandemic. Future studies should capture categories of nursing workforce as detailed in the ACCCN Workforce Standards,[ 11 ] in addition to ICU medical and allied health workforce data.
The COVID pandemic put ICU staffing in the spotlight and exposed the critical dependence on having an available, highly skilled ICU nursing workforce. Routinely collected and access to detailed data on the ICU nursing workforce will be important as health services recover from the previous waves of the pandemic, manage COVID patients ongoing, and deal with a predicted increase in influenza presentations and increase elective/emergency surgery demands. The creation of an ICU staffing dashboard is feasible[ 34 ] and would enable benchmarking across the sector. There is a pressing need to research the impact of ICU nursing skillmix on patient outcomes and ICU workforce factors on nurse outcomes. Furthermore, research on how to create and maintain a potential flexible redeployment workforce, and the implications for patient care and ICU staffing sustainability, is warranted, if redeployment of non-ICU staff is to become part of the health system response to extreme events.
Conclusion
This study has demonstrated that redeployment of workforce was required to ensure adequate patient care and accommodate the increased burden in Victorian ICUs during the COVID-19 pandemic. Despite this, at times, some ICUs had insufficient staff to cope with the number and acuity of patients. There is concern that the pandemic model of care is not sustainable in the long-term because of the burden on staff and further research is needed to examine the impact of ICU nursing models of care on patient outcomes and on nurse outcomes.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Appendix A Supplementary data
The following are the Supplementary data to this article:Appendix Table 1
Baseline characteristics of Intensive Care Units - number of 'business as usual' ICU beds per site as of July 2021
Appendix Table 2 - Staffing and activity characteristics of Intensive Care Units (ICUs) during peak phase (1st December 2021 to 19th February 2022) and post-peak phase (20th February 2022 to 11th April 2022)
Appendix Table 3 – Staffing and Intensive Care Unit (ICU) activity characteristics each day at different hospital types
Appendix Table 1
Appendix Figure 1 Daily number of total, 1:1 nursed, ventilated and COVID-19 patients in the 24 study hospital ICUs
Explanatory footnote: This shows the number of COVID-19 patients (blue) in study hospital ICUs plotted against the numbers of overall ventilated (dark green), not ventilated but 1:1 nursed (mid-green) and HDU 1:2 nursed patients (light green). As COVID-19 numbers declined in the later phase, the total number of patients requiring 1:1 ICU nursing care also declined and the number of HDU 1:2 patients increased as more ICU beds became available
Appendix Figure 1
Appendix Figure 2 Daily proportion and number of Intensive Care Units (ICUs) which reported more patients requiring 1:1 nursing than the number of critical care nursing staff working in the ICU
Explanatory footnote: The daily proportion and number of ICUs which reported having more patients receiving 1:1 nursing than they had critical care nursing staff declined progressively over the study period
Appendix Figure 2
Acknowledgements
The authors would like to thank all ICU staff at the following hospitals who were included in this study: The Alfred Hospital, Angliss Hospital, The Austin Hospital, Ballarat Health Services, Bendigo Hospital, Dandenong Hospital, Epworth Freemasons, Epworth Geelong, Epworth Richmond, Frankston Hospital, University Hospital Geelong, Latrobe Regional Hospital, Maroondah Hospital, Mildura Base Hospital, Monash Children's Hospital, Monash Medical Centre Clayton, Northeast Health Wangaratta, The Northern Hospital, Peninsula Private Hospital, The Royal Children’s Hospital, The Royal Melbourne Hospital, St John of God Geelong Hospital, St Vincent’s Hospital, Sunshine Hospital.
The authors gratefully acknowledge the contribution and work of the following Nurse Unit Managers without whom this study would not have been possible:
Tania Birthisel, Clare Kitch, Michelle Topple, Courtney Rowe, Penny Spencer, Dacielle Johnson, Monique Sammut, Vanessa Sawyer, Stuart Shakespeare, Jason Watterson, Donna Robertson, Bec Wittmer, Carol McKenzie, Sue Hale, Diana Sarraj, Adrienne Pendry, Juliana Sheridan, Narkitaa Van Ekeren, Sarah Edwards, Ashley Doherty, Michelle Spence, Sharnie McAuliffe, Philippe Thomas, Sam Angiolella
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.aucc.2022.12.001.
==== Refs
References
1 Bhatla, A. and K.L. Ryskina. Hospital and ICU patient volume per physician at peak of COVID pandemic: State-level estimates. in Healthcare. 2020. Elsevier.
2 Burau V. Health system resilience and health workforce capacities: Comparing health system responses during the COVID‐19 pandemic in six European countries The International Journal of Health Planning and Management 2022
3 Doyle J. Mobilising a workforce to combat COVID-19: an account, reflections, and lessons learned Journal of the Intensive Care Society 2020 1751143720971540
4 Endacott R. How COVID‐19 has affected staffing models in intensive care: A qualitative study examining alternative staffing models (SEISMIC) Journal of advanced nursing 78 4 2022 1075 1088 34779532
5 Goh K.J. Preparing your intensive care unit for the COVID-19 pandemic: practical considerations and strategies Critical Care 24 1 2020 1 12 31898531
6 Gupta N. Health workforce surge capacity during the COVID‐19 pandemic and other global respiratory disease outbreaks: A systematic review of health system requirements and responses The International Journal of Health Planning and Management 36 S1 2021 26 41
7 Marshall A.P. A critical care pandemic staffing framework in Australia Australian Critical Care 34 2 2021 123 131 33039301
8 Litton E. Surge capacity of intensive care units in case of acute increase in demand caused by COVID‐19 in Australia Medical Journal of Australia 212 10 2020 463 467 32306408
9 Pilcher D. A national system for monitoring intensive care unit demand and capacity: the Critical Health Resources Information System (CHRIS) The Medical Journal of Australia 214 7 2021 297 33774832
10 Department of Health Victoria, Coronavirus (COVID-19) Intensive Care Unit surge workforce models of care delivery.
11 Chamberlain D. Pollock W. Fulbrook P. ACCCN workforce standards for intensive care nursing: systematic and evidence review, development, and appraisal Australian Critical Care 31 5 2018 292 302 29246795
12 Rae P.J. Outcomes sensitive to critical care nurse staffing levels: A systematic review Intensive and Critical Care Nursing 67 2021 103110
13 Bae S.H. Intensive care nurse staffing and nurse outcomes: a systematic review Nursing in Critical Care 26 6 2021 457 466 33403791
14 Government of Victoria. Victorian COVID-19 data. 2022 18 October 2022]; Available from: https://www.coronavirus.vic.gov.au/victorian-coronavirus-covid-19-data.
15 Australian College of Critical Care Nurses Workforce Standards for Intensive Care Nursing. 2016, ACCCN Ltd: Melbourne.
16 Al Mutair A. Nursing surge capacity strategies for management of critically ill adults with COVID-19 Nursing Reports 10 1 2020 23 32 34968261
17 Kerlin M.P. Actions taken by US hospitals to prepare for increased demand for intensive care during the first wave of COVID-19: a national survey Chest 160 2 2021 519 528 33716038
18 Carenzo L. Hospital surge capacity in a tertiary emergency referral centre during the COVID‐19 outbreak in Italy Anaesthesia 75 7 2020 928 934 32246838
19 Chew M.S. A descriptive study of the surge response and outcomes of ICU patients with COVID‐19 during first wave in Nordic countries Acta Anaesthesiologica Scandinavica 66 1 2022 56 64 34570897
20 Camilleri M. Covid-19 ICU remote-learning course (CIRLC): rapid ICU remote training for frontline health professionals during the COVID-19 pandemic in the UK Journal of the Intensive Care Society 2020 1751143720972630
21 Xi, J., et al., COVID-19 mortality in ICUs associated with critical care staffing. Burns & Trauma, 2021. 9.
22 Neuraz A. Patient mortality is associated with staff resources and workload in the ICU: a multicenter observational study Critical care medicine 43 8 2015 1587 1594 25867907
23 Pattison, N., An ever‐thorny issue: Defining key elements of critical care nursing and its relation to staffing. 2021, Wiley Online Library. p. 421-424.
24 Riddell K. The context, contribution and consequences of addressing the COVID‐19 pandemic: A qualitative exploration of executive nurses' perspectives Journal of Advanced Nursing 2022
25 Lucchini A. Nursing Activities Score is increased in COVID-19 patients Intensive & critical care nursing 59 2020 102876
26 Bergman L. Registered nurses' experiences of working in the intensive care unit during the COVID‐19 pandemic Nursing in critical care 26 6 2021 467 475 33973304
27 Australian and New Zealand Intensive Care Society Centre for Outcome and Resource Evaluation, Report on COVID-19 admissions to Intensive Care in Australia 01 January 2021 - 31 December 2021. 2022, The Australian and New Zealand Intensive Care Society (ANZICS).
28 Endacott R. The organisation of nurse staffing in intensive care units: a qualitative study Journal of Nursing Management 2022
29 Lasater K.B. Chronic hospital nurse understaffing meets COVID-19: an observational study BMJ Quality & Safety 30 8 2021 639 647
30 Arabi Y.M. How the COVID-19 pandemic will change the future of critical care Intensive care medicine 47 3 2021 282 291 33616696
31 Baez‐Leon C. A qualitative study on a novel peer collaboration care programme during the first COVID‐19 outbreak: A SWOT analysis Nursing Open 9 1 2022 765 774 34773372
32 Hammond N.E. Impact of the coronavirus disease 2019 pandemic on critical care healthcare workers' depression, anxiety, and stress levels Australian Critical Care 34 2 2021 146 154 33632606
33 Clark S.E. Chisnall G. Vindrola-Padros C. A systematic review of de-escalation strategies for redeployed staff and repurposed facilities in COVID-19 intensive care units (ICUs) during the pandemic EClinicalMedicine 44 2022 101286
34 Davidson B. Requirements for a Bespoke Intensive Care Unit Dashboard in Response to the COVID-19 Pandemic: Semistructured Interview Study JMIR Human Factors 9 2 2022 e30523
| 0 | PMC9742212 | NO-CC CODE | 2022-12-14 23:29:59 | no | Aust Crit Care. 2022 Dec 12; doi: 10.1016/j.aucc.2022.12.001 | utf-8 | Aust Crit Care | 2,022 | 10.1016/j.aucc.2022.12.001 | oa_other |
==== Front
Transp Res Part A Policy Pract
Transp Res Part A Policy Pract
Transportation Research. Part A, Policy and Practice
0965-8564
1879-2375
The Author(s). Published by Elsevier Ltd.
S0965-8564(22)00311-1
10.1016/j.tra.2022.103560
103560
Article
Exploring attitude-behaviour dynamics during COVID-19: How fear of infection and working from home influence train use and the attitude toward this mode
Kroesen Maarten a
De Vos Jonas b
Le Huyen T.K. c
Ton Danique d
a Faculty of Technology, Policy and Management, Delft University of Technology, the Netherlands
b Bartlett School of Planning, University College London, UK
c Department of Geography, The Ohio State University, USA
d Station Division, Netherlands Railways (NS), the Netherlands
12 12 2022
1 2023
12 12 2022
167 103560103560
19 4 2022
11 11 2022
3 12 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.
Research on the relationships between travel-related attitudes and travel behaviour has recently been reinvigorated by new theorizing as well as new empirical models. While traditional theories assume a rather static role of attitudes, i.e. acting as stable predispositions that cause behaviours in a unidirectional manner, recent models assume that attitudes and behaviours mutually influence each other over time. This study aims at better understanding attitude-behaviour dynamics by capitalizing on the circumstances presented by the ongoing COVID-19 pandemic. It assesses how the fear of COVID-19 infection and (the attitude towards) working-from-home influence train use as well as train use attitudes. To explore the (within-person) reciprocal relationships between these variables, random-intercept cross-lagged panel models were estimated using a 4-wave longitudinal dataset collected during the COVID-19 pandemic from a large panel of train travellers in the Netherlands. The results indicate that train use and the attitude towards train use reciprocally influence each other. Those with stronger fears of infection in one wave tend to use the train less in a subsequent wave, but higher use of the train in one wave also reduces the fear of infection in the next. We also found that working from home (WFH) and travelling by train operate as substitutes for one another. Moreover, people who work from home frequently become more fearful of infection. All the findings are consistent with cognitive dissonance theory that people develop attitudes that align with their behaviours. The paper concludes with several policy implications related to changing attitudes and promoting train use.
Keywords
COVID-19 infection fear
Train use
Attitude towards train use
Working from home
Panel data
Cross-lagged panel model
==== Body
pmc1 Introduction
Research on the relationships between travel-related attitudes and travel behaviour has recently been reinvigorated by new theorizing (De Vos and Singleton, 2020, De Vos et al., 2021, Kroesen and Chorus, 2020, Van Wee et al., 2019) as well as new empirical models (Kroesen et al., 2017, Olde Kalter et al., 2021). While traditional theories and models (e.g. the Theory of Planned Behaviour) assume a rather static role of attitudes, i.e. acting as stable predispositions that cause behaviours in a unidirectional manner, recent theories and models assume that attitudes can change and that attitudes and behaviours mutually influence each other over time (De Vos et al., 2021, Kroesen et al., 2017). Classic social-psychological theories such as cognitive dissonance theory support such a dynamic perspective, i.e. to resolve dissonance between attitudes and behaviours, behaviours may be aligned with attitudes or attitudes may be aligned with behaviours. The recently introduced travel mode choice cycle proposed by De Vos et al. (2021), which, among others, draws from cognitive dissonance theory, also explicitly supports bidirectional effects between behaviours and attitudes, via travel satisfaction.
In line with the goal to better understand attitude-behaviour dynamics, Van Wee et al. (2019) have proposed a conceptual model of attitude change. Based on the earlier work of Eagly and Chaiken (1993), they identify-three interrelated mechanisms that may drive attitude change, namely cognitive, affective and behavioural mechanisms. In short, attitudes may change because a person is exposed to new knowledge, or because the person experiences new emotions or feelings and/or because the person engages in new behaviours. Often, these processes are interrelated. For example, a person who learns that cycling is good for health, may then start cycling and then experience that cycling is fun. In turn, this affective outcome may reinforce the behaviour and/or the cognitive belief. This argument has also recently been confirmed empirically by Kroesen and Chorus (2020) who show that cognitive and affective attitudes and behaviours influence each other in dynamic psychological networks.
Capitalizing on the circumstances presented by the ongoing COVID-19 pandemic, this research empirically follows up on the model proposed by Van Wee et al. (2019) and explores how the fear of COVID-19 infection and (the attitude towards) working-from-home influence train use as well as train use attitudes. The fear of infection can be seen as an entirely new element that has entered the ‘attitude-behaviour domain’ of train travel. As such, by studying its effects, a better understanding of the (general) processes of attitude-behaviour dynamics may arise. For example, it may be expected that people with stronger fears of getting infected will directly reduce their travel by train to reduce infection risks, but they may also revise their attitude toward travelling by train accordingly (i.e. they start disliking travel by train because of the fear). In addition, reverse effects may also exist. For example, if a person has a favourable travel experience by train (during the pandemic), e.g. because the train was not perceived as (too) crowded, the person may adjust his/her fear of infection downwards (or upwards, if the experience was bad). Furthermore, the attitude may also affect the fear of infection. For example, a person who really likes to travel by train may suppress and/or downplay his/her fears. By assessing which of these effects actually occur, new theoretical insights on how the fear of infection as a new disposition would inform the dynamics in this particular attitude-behaviour domain.
In addition to the fear of infection, working from home (WFH) and the attitude towards this behaviour also present new elements that may influence train use and/or the attitude towards train use. While WFH has been around before COVID-19, it has received a strong boost, partly due to the travel restrictions imposed by national governments (e.g., Mohammadian et al., 2022; Ton et al., 2022, Wang et al., 2022). Here, it may be expected that the extent to which a person works from home directly functions as a substitute for travel by train. In addition, also the attitude toward WFH may play a role. As shown by Rubin et al., 2020, Beck et al., 2020, experiences with WFH differ greatly among people and these may also determine the current and future use of the train (Ton et al., 2022a). Finally, reverse effects are also plausible for these relationships. For example, people who continue to travel by train or (still) like to travel by train may choose to work less from home and/or reduce their (positive) evaluation of it.
The objective of this study is to analyse how the fear of infection and (attitude towards) WFH influence the use of the train and the attitude toward the use of this mode. In addition to the theoretical contributions mentioned above, knowledge of these effects arguably has relevant practical implications. For example, should the fear of infection only influence the behaviour and not the attitude, it may be expected that travel behaviours, such as train use, will resume faster to pre-COVID-19 levels when the pandemic is over, or under control. However, should the fear of infection also influence the attitude towards train use, the current decline in train use will likely last longer.
To assess how the fear of infection and (attitude towards) WFH influence train use and the attitude towards train use as well as possible reverse effects, we estimate (two) random-intercept cross-lagged panel models (RI-CLPM) (Hamaker et al., 2015). This model is ideally suited to assess ‘within-person’ reciprocal effects between multiple variables over time. Data to estimate the model come from a large panel of train travellers in the Netherlands. In this panel a survey was administrated at four time points after the onset of the COVID-19 pandemic.
The structure of this paper is as follows. We will first briefly review the growing body of literature on attitude change in travel behaviour research and several empirical studies that have explored attitude changes in response to the COVID-19 crisis. Next, we will present the results of two models, one for the full sample to explore relationships between the (perceived) risk of infection, train use and train use attitude, and another model for the subsample of people who were able to work from home to additionally explore relationships with WFH and the attitude toward this behaviour. In discussing the results, we provide possible explanations for observed effects. The conclusion summarises the main findings and discusses implications for policy.
2 Background
2.1 Attitude change: Theory
Van Wee et al. (2019) recently presented a conceptual model of attitude change, in which attitude change is assumed to originate from three interrelated processes, namely cognitive, affective and behavioural processes. In turn, these processes may be activated by external triggers, which are categorized into personal, social and environmental triggers. According to Van Wee et al. (2019), personal triggers relate to an actor’s own information and experiences, social triggers to influences from the actor's network (e.g. family, friends or colleagues) and environmental triggers to all other possible triggers from the surrounding environment, for example, changes in the residential environment or in the transport system. Similar to the three mechanisms of attitude change, the triggers may also be interrelated. For example, the birth of a child can be seen as both a personal and a social trigger, as it affects the actor directly, but also indirectly through his or her network.
Within the model of Van Wee et al. (2019), the COVID-19 pandemic can clearly be identified as an environmental trigger, which may have led to attitude changes in various ways. Because train travel implies being in physical proximity to other people, the fear of infection (as a new disposition) represents one such way. This new disposition may trigger cognitive, affective and behavioural processes leading to attitude change. On a cognitive level, people who are fearful of getting infected likely take this new information into account when deciding to ride public transport modes vis-à-vis private travel modes (e.g. car and bicycle) or not travelling at all (e.g. by WFH). On an affective level, the fear of getting infected may negatively influence a person’s preference toward travelling by public transport modes and/or positively influence preferences toward travelling by private modes or WFH. And on a behavioural level, the actual decision to travel with modes other than those previously used (before COVID-19) or to WFH may have led to new experiences and thereby attitude change.
Although most social-psychological theories such as the Theory of Planned Behaviour (Ajzen, 1991), and the Theory of Interpersonal Behaviour (Triandis, 1977) suggest that attitudes are important predictors of behavioural intention, some also indicate that behaviour may affect attitudes. The relatively old Theory of Cognitive Dissonance (Festinger, 1957) and Balance Theory (Heider, 1958) indicate that attitudes may change, especially to reduce inconsistency between attitudes and behaviour. In case of such an inconsistency (or dissonance) people may either change their behaviour, or (in case changing behaviour is perceived impossible) change their attitudes. In terms of travel, this would mean that people who (are forced to) travel with an undesired travel mode will try to change their behaviour (i.e., use another mode) or justify their choice by improving their attitude toward the chosen mode (Kroesen et al., 2017, De Vos and Singleton, 2020).
Some travel behaviour studies have found bidirectional relationships between travel attitudes and travel behaviour (e.g., Dobson et al., 1978, Tardiff, 1977), while some of them even indicate that people are more likely to change their travel attitudes than their travel behaviour (e.g., Golob, 2001, Kroesen et al., 2017, McCarthy et al., 2021), indicating that travel behaviour may often be hard to change and that some travellers may be captive travellers (e.g., captive public transport users). Recently, De Vos et al. (2021) have introduced the travel mode choice cycle indicating that attitudes can influence behaviour through desires and intentions, but that behaviour can also affect attitudes, through satisfaction. The former effects have been found by some studies testing social-psychological theories (such as the Theory of Planned Behaviour) in a travel behaviour context (e.g., Bamberg et al., 2003). The latter effects have, for instance, been found by Fujii and Kitamura (2003), who observed that attitudes toward bus use improved after a switch from car use to bus use, partly because people started appreciating the positive aspects of public transport after using it.
2.2 Empirical studies of behavioural and attitudinal changes in response to COVID-19
Various empirical studies have examined the effects of the COVID-19 pandemic on travel preferences and behaviour. Empirical research has found substantial changes in travel frequency and trip purpose during the pandemic (Capasso da Silva et al., 2021, De Haas et al., 2020, Kar et al., 2021, Salon et al., 2021). Specifically, trip frequency and distances travelled decreased, mostly due to travel restrictions such as lockdown and stay-at-home orders in most countries. As most commuting halts due to COVID-19, WFH has become popular, as well as recreation and leisure trips such as visiting parks and exercise loop trips (e.g., running around the neighbourhood) (De Haas et al., 2020, Hook et al., 2021, Galleguillos-Torres et al., 2022, Kar et al., 2021, Salon et al., 2021).
In addition to changes in trip frequency and distances travelled, mode switch is also common. Many studies have found a significant drop in public transport ridership and an increase in private automobile and active transport mode use (Abdullah et al., 2021, De Haas et al., 2020, Salon et al., 2021, Shamshiripour et al., 2020), where sometimes a car is purchased especially to replace public transport trips (van Hagen et al., 2021). Unlike changes in travel distance, trip frequency, and trip purpose which are likely introduced by travel restrictions (e.g., lockdown), a mode switch and increased WFH levels are likely due to health concerns and fears of infection as sharing space with other people can be perceived as unsafe during this time (Capasso da Silva et al., 2021).
A few stated-preference studies also examine the possibility that concerns around COVID-19 and on-going changes in travel will persist after the pandemic is over (Mirtich et al., 2021, Salon et al., 2021, Shamshiripour et al., 2020). It is implied that attitude toward travel and WFH is fairly stable for up to a year (Mirtich et al., 2021), suggesting that travel adjustments during the pandemic may not be long-lasting. Salon et al. (2021) found that shifting to active transport and shifting away from public transport are likely to stick post-COVID-19, while Mohammadian et al. (2022) – using a 4-wave panel – mainly indicates that public transport will struggle to attract passengers post-COVID-19. It is noteworthy that these studies used US-based samples, which may be different from other cities/countries in the world.
2.3 Synthesis and research focus
To summarise, the COVID-19 pandemic can be considered an environmental trigger that has resulted in attitude and behaviour changes in various ways. In this study we focus in particular on how the fear of infection and (the attitude toward) WFH influence the (attitude toward the) use of the train. The fear of infection is an entirely new element that can trigger affective, cognitive and behavioural processes that influence (the attitude toward) train travel. In addition, WFH has received a strong boost due to travel restrictions and thus may influence the (attitude toward) train travel, especially considering that relatively many train travellers are highly educated and often have the ability to WFH. Finally, in line with recent theorising and empirical research in the field of travel behaviour we assume that the fear of infection and (the attitude toward) WFH may vice-versa be influenced by the (attitude toward) train use. Hence, all possible bidirectional effects will be explored.
3 Method
3.1 Data collection
To gain insights into passenger behaviour during and after the pandemic, a longitudinal survey is organised by NS (Dutch Railways) and Delft University of Technology with the goal to capture behaviour, attitudes and intentions regarding travel behaviour in general and train use in particular. The participants of the survey are part of the existing panel of NS. This panel represents all train travellers in the Netherlands and participation is voluntary. The total panel encompasses more than 80,000 members and members can receive invitations for a variety of research initiatives related to train travel. Since it was expected that behaviour, attitudes and intentions would change during the pandemic, multiple surveys were planned and held. Beforehand it was unknown how many surveys would be distributed in total. The argument was that when drastic changes in the measures happened, it would possibly lead to behavioural changes, therefore a new survey would be distributed.
The first survey was distributed among all panel members and about 46,000 respondents completed the survey (roughly 57 % response rate). This survey aimed at capturing respondents’ behaviour in April 2020 during the “intelligent lockdown”. In this period, train travelling was only allowed for people working in essential sectors. 96 % of the respondents agreed to participate in a longitudinal study to monitor trends and changes. In this survey, the pre-COVID-19 situation was also captured by asking about the travel behaviour in February 2020. In June 2020 (end of lockdown, but still many limitations), September 2020 (more working allowed in the office), December 2020 (second COVID-19 wave and news about a vaccine), April 2021 (restrictions are relaxed), September 2021 (restrictions are relaxed further), and April 2022 (no restrictions), follow-up surveys were held. In addition, more in-depth questions were asked to focus on specific topics, for instance WFH (see Ton et al., 2022a), vehicle purchase behaviour and home relocation behaviour. Every survey was filled in by at least 18,000 respondents, but a decline in response can be seen over time.
To check for bias and self-selection among the internal NS panel, an external panel was approached in parallel, where a sample representative for the Dutch train traveling population was invited (1,500 respondents) to verify the behaviour, attitudes and intentions of the internal panel members. Furthermore, each of the waves of data collected on the internal panel, is weighted against the Dutch train traveling population. These two panels showed similar behaviour, attitudes and intentions; hence, we conclude that the internal panel can be considered representative of the train traveling population (Ton et al., 2022b).
3.2 Data description
For this study the first four waves of data are used (April, June, September, and December 2020). As the goal is to capture the within-person changes over time, it is essential that each individual has participated in all waves. This also means that the resulting dataset is not weighted against the Dutch train traveling population and is therefore not necessarily representative. A total of 14,760 respondents participated in each wave and are used in the analysis. Of those respondents, 3,587 could work from home. Table 1 shows the socio-demographic characteristics of both samples (full and WFH sample).Table 1 Sample distributions of socio-demographic characteristics of the full sample and WFH sample.
Variable Categories Full sample (%) WFH Sample (%)
Gender Male 48.1 55.3
Female 49.7 42.2
Other / missing 2.1 2.6
Age 18–34 years 5.1 8.8
35–44 years 5.2 11.6
45–54 years 10.5 21.8
55–64 years 24.9 38.9
65–74 years 38.8 12.4
75 years and older 10.8 1.7
Missing 4.6 4.8
Level of education Intermediate secondary education 9.8 1.6
Higher secondary education 10 6.6
Intermediate vocational education 12.4 5.6
Higher vocational education (college) 33.7 30.6
University 29.0 50.7
Missing or other 5.2 4.7
Occupational status Paid employment 35.9 73.4
Freelancer or self-employed 4.4 11.6
Attends school or is studying 2.2 2.3
Takes care of the housekeeping 1.8 0.0
Pensioner 46.1 6.7
Missing or other 9.5 6.0
Household composition Alone 34.0 27.2
With partner 49.2 40.3
With partner and child(ren) 10.3 23.8
Missing or other 6.5 8.7
There are quite some differences between the full sample and the WFH sample, most of these are attributed to the large share of pensioners (46.1 %) in the full sample, who no longer perform paid work. This affects age distributions and household composition. Furthermore, the share of highly educated respondents in the WFH sample is much larger compared to the full sample (50.7 % versus 29 %). This is mostly because highly educated respondents often hold white-collar office jobs that facilitate WFH (Ton et al., 2022a). In addition, there are fewer women in the WFH sample (compared to the general population). This is in part due to the lower employment rate among women in the Netherlands. In Q2 of 2022, 76 % of the (Dutch) men were employed compared to 68 % of the women (Statistics Netherlands, 2022a). In addition, the average commuting distance of women (19.1 km) is lower than that of men (24.8 km) (Statistics Netherlands, 2022b). Given that the train only becomes popular for longer commute distances (>30 km) it seems plausible that men are overrepresented among the population of employed people travelling by train.
For this study, the questions related to train use frequency, WFH frequency, attitude toward train use, attitude toward WFH, and fear of infection are most relevant. These questions were included in each wave, with the first survey also reflecting the pre-COVID-19 situation (only travel behaviour). As no information was available before the pandemic, this information serves as the reference point for changes occurring during the pandemic. To ensure good quality of the information on travel behaviour, the respondents were asked about their travel behaviour in the past week. Regarding train use, the respondents were asked how often they used the train in the past week (0–7 days). For WFH, an ordinal scale was used, ranging from none (1), 1 day (2), 2–3 days (3) and 4 or more days (3). For the attitudes toward train use (“I enjoy travelling by train”) and WFH (“I enjoy working from home”) 5-point Likert scales were used with answer categories ranging from strongly disagree (1) to strongly agree (5). The fear of infection was captured using the following statement: “I am afraid to become infected by the coronavirus”. This was also measured on a 5-point Likert scale from strongly disagree (1) to strongly agree (5).
Table 2 shows the means and standard deviations of the train use, train attitude and fear of infection variables for the full sample. The results show that train use decreased significantly during the pandemic and has not yet reached its reference point. Furthermore, both the attitude toward train use and the fear of infection vary slightly over time and are seemingly correlated with the number of positive COVID-19 cases and the strictness of the measures imposed by the government.Table 2 Means and standard deviations of the dependent variables (full sample).
Train use (days/week) Attitude toward train use (1–5) Fear of infection (1–5)
Wave Mean SD Mean SD Mean SD
0 (pre-COVID-19) 1.3 1.7
1 (April - lockdown) 0.2 0.7 2.8 1.5 3.3 1.1
2 (June – end of lock down, with restrictions) 0.4 0.9 3.1 1.3 3.0 1.0
3 (September – some office working allowed) 0.7 1.2 3.3 1.2 3.2 1.0
4 (December – second wave of COVID-19) 0.5 1.0 3.1 1.3 3.4 1.1
Table 3 shows the means and standard deviations of all the relevant variables for the WFH sample. Also here, several differences can be seen compared to the full sample. Train use before COVID-19 is higher for the WFH sample. These respondents were mostly commuting by train to work, hence their weekly train use is higher compared to the full sample where many incidental social and recreational trips are also made. Furthermore, the attitude toward the train is slightly lower for the WFH sample, and fear of infection is also slightly lower. The attitude toward WFH is generally very positive among the WFH sample and the frequency is high. In addition, it is also interesting to see that the attitude towards WFH has become more positive over time.Table 3 Means and standard deviations of the dependent variables (WFH sample).
Train use (days/week) Attitude toward train use (1–5) Fear of infection (1–5) WFH (1–4) Attitude toward WFH (1–5)
Wave Mean SD Mean SD Mean SD Mean SD Mean SD
0 (pre-COVID-19) 2.4 1.9 2.8 1.3
1 (April 2020) 0.1 0.5 2.7 1.5 3.1 1.1 3.7 0.5 3.5 1.0
2 (June 2020) 0.4 0.8 3.0 1.3 2.9 1.0 3.6 0.6 3.7 1.0
3 (September 2020) 0.7 1.1 3.2 1.2 3.1 1.0 3.4 0.7 3.7 1.0
4 (December 2020) 0.4 0.8 3.0 1.3 3.2 1.1 3.6 0.6 3.8 1.0
3.3 Statistical model
To test the bidirectional relationships between train use, the attitude toward train use, the fear of infection, and (the attitude toward) WFH, two Random-Intercept Cross-lagged Panel Models (RI-CLPM) were specified. In the following, the first model will be briefly introduced (for a full description of the RI-CLPM we refer to Hamaker et al. (2015)).
Fig. 1 presents the structure of the first RI-CLPM that was specified and estimated in this study. For each observed variable (in rectangles) a respective latent variable is specified. The paths linking these latent variables to the observed ones are set to 1. In addition, temporal means are included for each respective point in time. As such, the latent variables effectively capture respondents’ temporal deviations from the time-varying group means, thereby accounting for population-wide structural change in the variables of interest. Note that such structural changes are indeed present (see Table 2, Table 3).Fig. 1 A 4-wave 3-variable Random Intercept Cross-Lagged Panel Model (Model 1).
Next, it is assumed that the (mean-centred) latent variables influence their future counterparts (autoregressive effects) as well as each of the other variables (cross-lagged effects) over time. The cross-lagged effects are of main interest as they indicate to which extent causal effects exist and in which directions.
Finally, the error terms of the latent variables on each occasion are allowed to correlate (not shown) as well as exogenous latent variables at wave 1. These (dynamic) correlations account for possible synchronous effects between both variables as well as the effects of (unmodelled) time-varying factors between the three variables.
So far, the above description captures the traditional CLPM. An important limitation of this model, as argued by Hamaker et al. (2015), is that, while the CLPM is able to capture temporal stability, it does not account for stable individual differences that endure over time (at least for the periods typically considered in panel studies). Indeed, this is a problematic assumption, since behavioural and attitudinal variables are generally characterized by stable individual differences (Hamaker et al., 2015), e.g. due to the presence of habits.
These stable differences may be accounted for by introducing three additional latent variables, the so-called random intercepts. To capture the notion that they have a constant (time-independent) ‘trait-like’ influence on the observed outcomes, the paths linking these variables to the observed variables are set to 1. Essentially, since the random intercepts capture variation between persons, stable ‘between-person’ variation is factored out. This has two benefits, namely (1) the stability/cross-lagged relations now capture ‘within-person’ carry-over effects from one occasion to the next, i.e. the level at which the causal processes are assumed to operate, and (2) all time-constant variables that may influence the three dependent variables are controlled for. This also means that it is not vital to include socio-demographic variables, which are generally (very) inert, as confounders in the model. Finally, the three random intercepts are allowed to correlate. These correlations indicate the extent to which the variables of interest are correlated at the ‘between-person’ level due to other factors than the assumed causal effects that operate between the three variables at the ‘within-person’ level.
To estimate model 1, data from the full sample are used. The second model (model 2) is specified as similar to model 1, but considers two additional variables, namely WFH and the attitude toward WFH. This model thus includes five autoregressive effects and twenty (5x4) cross-lagged relationships. This model is estimated using data from the WFH sample. For both models, it is assumed that the effects are stable over time, hence equality constraints are imposed on the same effects across each wave-pair.
4 Results
The models were estimated using Mplus applying the robust maximum likelihood estimator to account for the fact that the data are not normally distributed. Table 4 presents the model fit of the CLPM (a) and the RI-CLPM (b) of models 1 and 2. For both models a large improvement in model fit can be observed indicating that the random intercepts indeed capture stable between-person individual differences. The RI-CLPMs fit well in terms of conventionally used relative fit indices (Hu and Bentler, 1999).Table 5 presents the (standardized) parameter estimates of model 1 (the full sample).1 Turning first to the autoregressive effects, which can be interpreted as ‘within-person’ carry-over effects of the same variable from one wave to the next, the results indicate that train use at t-1 has the largest effect on its respective counterpart at t (0.328), followed by the fear of infection (0.144) and the attitude toward train use (0.117). This means that, on top of the overall stability in - for example - the use of the train, if a person has a higher (or lower) use of the train than his/her expected score in a certain wave (following from the random intercept for train use), he or she will also have a ‘higher (lower) than expected’ score in the subsequent wave. The presence and significance of these effects thus indicate that there are ‘within-person’ processes at work that enable these intrapersonal carryover effects. For example, for train use, it may be speculated that the experience of using the train leads a person to use the train again (in the next wave).Table 4 Model fit.
Model χ2 df p-value RMSEAa CFIb SRMRc
Model 1a (CLPM) 10437.8 46 0.000 0.124 0.781 0.124
Model 1b (RI-CLPM) 1690.4 40 0.000 0.053 0.965 0.041
Model 2a (CLPM) 3170.4 120 0.000 0.084 0.870 0.068
Model 2b (RI-CLPM) 709.4 105 0.000 0.040 0.974 0.039
a Root Mean Square Error of Approximation (<0.06 indicates good fit. Hu and Bentler (1999)).
b Comparative Fit Index (>0.95 indicates good fit. Hu and Bentler (1999)).
c Standardized Root Mean squared Residual (<0.08 indicates good fit. Hu and Bentler (1999)).
Table 5 Standardized parameter estimates of model 1 (full sample).a
Autoregressive effects
Effect Est. p-value
Train use (t-1) → train use (t) 0.328 0.000
Attitude toward train use (t-1) → Attitude toward train use (t) 0.117 0.000
Fear of infection (t-1) → Fear of infection (t) 0.144 0.000
Cross-lagged effects
(within-person) Correlation RIs
(between-person)
Effect Est. p-value Est. p-value
Train use (t-1) → Attitude toward train use (t) 0.108 0.000 0.264 0.000
Attitude toward train use (t-1) → Train use (t) 0.058 0.000
Train use (t-1) → Fear of infection (t) −0.071 0.000 −0.186 0.000
Fear of infection (t-1) → Train use (t) −0.048 0.000
Attitude toward train use (t-1) → Fear of infection (t) −0.013 0.073 −0.377 0.000
Fear of infection (t-1) → Attitude toward train use (t) −0.013 0.109
a The presented values in the table show the means of the standardized estimates across all waves. Note that, while the unstandardized estimates are equal across all waves, the standardized estimates differ slightly from wave to wave due to the time-varying variances of the variables.
The cross-lagged effects, which are of main interest, indicate the extent to which the variables influence each other over time. Similar to the autoregressive effects, the estimates can be interpreted as within-person carry-over effects from one occasion to the next, but now from one variable to the other. With respect to the relation between train use and the attitude toward train use, the results indicate that the effect of behaviour on attitude is approximately twice as large (0.108) as vice versa (0.058), though both effects are statistically significant. This result is consistent with findings of previous panel studies, which have consistently shown that the effect of behaviour on later attitudes is stronger than the other way around (Kroesen et al., 2017, Olde Kalter et al., 2021). Significant bidirectional effects also exist between train use and the fear of infection: those with stronger fears of infection in one wave tend to use the train less in a subsequent wave (-0.048). Meanwhile, the use of the train reduces the fear of infection over time (-0.071). Here, in line with the expectation formulated in the introduction, it seems that people’s experiences of travelling by train generally lead them to become less fearful of being infected. Finally, with regards to the relation between the fear of infection and the attitude toward travel by train, both effects are also negative but not statistically significant. Hence, while the fear of infection does lead people to travel less by train, it does not directly alter people’s attitude toward travelling by train (although the attitude is indirectly affected via the reduced use of the train).
Finally, the correlations between the random intercepts capture the associations between the stable individual factors that are assumed to exist for each variable in the model. For the relationships between train use and the attitude toward the train, as well as between train use and the fear of infection, these correlations (0.264 and −0.186, respectively) have the same signs and (relative) sizes as the within-person effects. Hence, at the between-person level, people who use the train more often have more positive attitudes toward train travel and lower fears of infection. It is likely that a part of these between-persons correlations has resulted from the (accumulation of) within-person effects. But the correlations may also (partly) be the result of other stable individual differences. For example, older people may generally use the train less and be more fearful of infection.
Interestingly, while the within-person effects were not significant between the attitude toward train use and the fear of infection, at the between-person level, the respective individual factors are strongly correlated (-0.377). As the within-person effects are insignificant here, this correlation should be entirely attributed to stable between-person differences. It may be speculated that certain stable personality traits inform both the fear of infection as well as the attitude toward train travel. For example, a general dislike of being in physical proximity to other people may be responsible. The fact that for this relationship the between-person correlation is large (in fact the largest of all three pairs) whereas the within-person effects are non-significant also illustrates the importance of discriminating these effects in the first place. Based on the (cross-sectional) correlation (Table 4) one might conclude that the fear of infection leads to a dislike of travelling by train (or vice versa) at the within-person level. The present results, however, disconfirm this conclusion and indicate that the association is entirely due to other stable differences across people.
Table 6 shows the estimates of model 2, in which two additional concepts have been added, namely WFH and the attitude toward WFH, yielding five autoregressive effects, twenty cross-lagged effects and ten correlations between the (five) random intercepts. The estimates associated with (the attitude toward) train use and the fear of infection are overall similar to those in model 1, indicating that being employed and being able to WFH, does not moderate these effects.Table 6 Standardized parameter estimates of model 2 (WFH sample).a
Autoregressive effects
Effect Est. p-value
Train use (t-1) -> train use (t) 0.271 0.000
Attitude toward train use (t-1) -> Attitude toward train use (t) 0.153 0.000
Fear of infection (t-1) -> Fear of infection (t) 0.139 0.000
WFH (t-1) -> WFH (t) 0.155 0.000
Attitude toward WFH (t-1) -> Attitude toward WFH (t) 0.121 0.000
Cross-lagged effects
(within-person) Correlation RIs
(between-person)
Effect Est. p-value Est. p-value
Train use (t-1) -> Attitude toward train use (t) 0.160 0.000 0.247 0.000
Attitude toward train use (t-1) -> Train use (t) 0.100 0.000
Train use (t-1) -> Fear of infection (t) −0.055 0.000 −0.187 0.000
Fear of infection (t-1) -> Train use (t) −0.046 0.001
Attitude toward train use (t-1) -> Fear of infection (t) −0.006 0.722 −0.392 0.000
Fear of infection (t-1) -> Attitude toward train use (t) −0.025 0.102
Train use (t-1) -> WFH (t) −0.062 0.000 −0.387 0.000
WFH (t-1) -> Train use (t) −0.035 0.006
Train use (t-1) -> Attitude toward WFH (t) −0.059 0.000 −0.120 0.000
Attitude toward WFH (t-1) -> Train use (t) −0.058 0.000
Attitude toward train use (t-1) -> WFH (t) 0.003 0.843 −0.146 0.000
WFH (t-1) -> Attitude toward train use (t) 0.015 0.277
Attitude toward train use (t-1) -> Attitude toward WFH (t) −0.052 0.003 −0.049 0.118
Attitude toward WFH (t-1) -> Attitude toward train use (t) −0.045 0.003
Fear of infection (t-1) -> WFH (t) 0.038 0.015 0.010 0.675
WFH (t-1) -> Fear of infection (t) 0.043 0.004
Fear of infection (t-1) -> Attitude toward WFH (t) 0.045 0.004 0.158 0.000
Attitude toward WFH (t-1) -> Fear of infection (t) 0.017 0.271
WFH (t-1) -> Attitude toward WFH (t) 0.031 0.064 0.192 0.000
Attitude toward WFH -> WFH (t) 0.012 0.491
a The presented values show the means of the standardized estimates across all waves. Note that, while the unstandardized estimates are equal across all waves, the standardized estimates differ slightly from wave to wave due to the time-varying variances of the variables.
The autoregressive effects of WFH and the attitude toward WFH are positive and similar in size as those of the fear of infection and the attitude toward train use. Again, it may be speculated that the experiences gained during WFH lead people to continue this behaviour at the next point in time (in addition to an overall stable habitual effect as captured by the random intercept).
Train use negatively influences both WFH frequency (-0.062) and the attitude toward WFH (-0.059), and these variables in turn also influence train use negatively (respectively −0.035 and −0.058). Regarding the relationship between WFH and train use, these negative reciprocal influences can be interpreted as substitution effects, where an increase in one results in a decline in the other (and vice versa). It is interesting to see that train use not only affects WFH but also the attitude toward WFH. Hence, people who (continue to) travel by train - over time - develop more negative attitudes toward WFH. Since frequently travelling by train results in lower WFH frequencies, people may develop more negative WFH attitudes in order to justify the choice to travel, in line with the cognitive dissonance theory. This pattern does not exist for the relationship between WFH and the attitude toward train use; no significant effects are found in either direction. At the level of attitudes, however, again significant bidirectional effects exist: the attitude toward train use negatively affects the attitude toward WFH (-0.052) and vice versa (-0.045).
The fear of infection positively influences WFH frequency (0.038), and those who work from home also become more fearful of infection over time (0.043). This pattern is consistent with the negative reciprocal effects between train use and the fear of infection. It seems the fear of infection is strengthened when people are not exposed to other people and reduced when people are. Overall, these findings support cognitive dissonance theory: people align their attitudes and behaviours to avoid contradictions between attitudes and behaviours (over time). For example, travelling by train and being (very) fearful of infection is a combination that would lead to a strong psychological tension, which can be reduced by either travelling less or becoming less fearful of infection. Similarly, people who WFH can ‘allow’ themselves to become (very) fearful of infection, as they are not exposed to actual risks. In this case, a position of being fearful does not lead to tensions with actual behaviour.
Finally, it is interesting to see that WFH and the attitude toward WFH do not strongly influence each other; the effects are not significant in either direction. This is surprising given that attitudes directed toward specific behaviours generally correlate quite strongly with the behaviour in question, and contradicts existing COVID-19 studies (e.g., Lee & De Vos, 2022). A plausible explanation is that WFH is often involuntary. Due to the restrictions imposed by the government, many people who do not like WFH were forced to do so. Similarly, there may be many people who want to work from home but their job does not allow it. Having these groups forced into these respective combinations suppresses the correlation between WFH and the attitude toward WFH. In this case, the ‘inconsistent’ combinations arguably do not result in a psychological tension, since the inconsistency (between attitude and behaviour) is not due to an individual choice but due to restrictions imposed by the government. Hence, there is no strong need to either adjust the behaviour or the attitude, which may have given rise to the small (insignificant) effects.
The correlations between the random intercepts again provide some interesting additional insights. For example, a large negative correlation exists between train use and WFH (-0.387). This means that in addition to the negative within-person reciprocal effects, there is a correlation across people in the general tendency to work from home and the general tendency to travel by train. This signifies that there are different groups that either tend to work from home or travel by train, this is also found by Ton et al. (2022a). Also a substantial correlation exists (0.192) between the attitude toward WFH and WFH frequency, even though the within-person effects were insignificant for this relationship. Again, this correlation has likely arisen due to other factors at the between-person level.
5 Conclusion and discussion
This study examined the effects of fear of infection with COVID-19 and WFH on the attitude toward train use and train use itself during the COVID-19 pandemic in the Netherlands by using a longitudinal dataset collected among train travellers during four periods during the pandemic (April 2020, June 2020, September 2020, and December 2020). This study contributes to the understanding of travel behaviour during COVID-19 and its implications for public transport planning beyond the pandemic. Our study provides a unique perspective and evidence for the attitude-behaviour dynamics, both within- and between-person levels, during an unprecedented social and public health disruption. We found that, at the within-person level, the attitude toward train travel was not affected by the fear of infection, but was negatively affected by the reduced train travel. At the between-person level, however, the fear of infection and attitude toward the train are strongly correlated. It seems plausible that similar psychological traits inform both, such as the dislike of being in physical proximity to other people. We also found that WFH clearly substitutes travel by train and vice versa. People who (continue) to travel by train become less fearful of infection while people who WFH become more fearful of infection.
Overall, our study confirms cognitive dissonance theory, i.e. people mutually adjust their attitudes and behaviors over time. This finding invalidates theories that assume that attitudes act as (stable) precedents of behavior and favors theories that do account for reciprocal effects, such as the travel mode choice cycle introduced by De Vos et al. (2021) or theoretical framework of Van Wee et al. (2019). In addition, whereas previous studies examined bidirectional effects between attitudes and behaviors with respect to the same mode (see e.g. Kroesen et al., 2017, Olde Kalter et al., 2021), the present study shows that people align their attitudes and behaviors across multiple domains. For example, people who travel less by train become more fearful of infection, or: people who travel more by train develop a more negative attitude towards WFH. Hence, the tendency to reach/maintain cognitive consistency is not confined to (same) attitude-behavior pairs. At the level of behaviors, our findings also provide evidence of substitution between train travel and WFH. Such substitution effects thus operate in tandem with tendencies to reach cognitive consistency, adding to the complexity of attitude-behavior dynamics.
This study offers several implications for urban planning and transportation practice. First, the plunge in train ridership during COVID-19 might be temporary and may recover after the pandemic. While the fear of infection is found to negatively affect train use, we found no evidence that the fear of infection alters the attitude toward train travel. This indicates that people may resume train travel again when the pandemic and the fear of infection are over. Furthermore, as train use has a strong negative effect on the fear of infection, current train users are consequently likely to continue train use after the pandemic.
Second, our results imply that the fear of infection might have driven some people to WFH. It is therefore important to nudge these people to travel by train again and disincentivise them from switching to car use when commuting resumes after the pandemic. Since using the train negatively affects fear, giving people temporal incentives to travel by train (e.g., a one-month free public transport pass) may reduce their fear, resulting in more train use in the future. Doing so may also improve train attitudes, which can then become stronger than feelings of fear. Previous studies have found that temporary incentives for using public transport can increase public transport ridership, but also improve attitudes and satisfaction levels (e.g., Abou-Zeid and Ben-Akiva, 2012, Fujii and Kitamura, 2003).
Third, it is unclear whether WFH will continue to affect train ridership post-pandemic. Insofar as a long-term negative effect on the attitude towards train use indeed exists, our results suggest this is due to the reduction in train travel and not due to the fear of infection. The fact that WFH does not affect the attitude toward WFH and vice versa may be attributed to the partly involuntary nature of WFH during COVID-19. Specifically, some people who WFH actually prefer not to do so, while some people who cannot WFH would actually want to do so. Also the insignificant bidirectional links between WFH and train attitudes suggest that some people may be forced to WFH, which is likely the case during the pandemic. In case WFH would be a free choice, it could be expected that those who WFH are those with especially negative attitudes towards train travel. Since train commutes often have relatively long commute durations, WFH may result in a considerable gain of time, which especially is of interest for those disliking the time on a train. Finally, the attitude toward WFH has become more positive over time, not so much due to the increased WFH, but due to reduced train travel. Due to this positive attitude, it is likely that a considerable portion of people will continue to work from home.
On a more general note, this study revealed that WFH and train travel act as clear substitutes for one another. Both are typically identified as ‘desirable’ behaviours from a policy perspective as they are associated with less environmental impacts compared to car use (typically the dominant alternative). Our results suggest that policy makers seeking to stimulate travel by train or WFH may inadvertently also discourage the other desirable behaviour (i.e. WFH or travel by train). Knowledge of such side-effects is crucial when trying to optimise the transport system as a whole.
This study has several limitations pertaining to data collection and measurement. First, our study relies on a set of single measures for train use, attitude toward train use, WFH, and fear of COVID-19 infection, therefore measurement errors are not accounted for in the analysis. However, the presented results are likely on the conservative side because, if we controlled for measurement errors (i.e. if multiple items were available), the estimates would have been larger than reported presently, as (random) measurement errors in the dependent variable typically attenuate estimates downwards. In addition, standard errors would have become smaller, leading to more significant effects. Nonetheless, future studies should employ multi-item measurement and account for measurement errors to have a better estimate of the true effects.
Secondly, the data for this study was collected during the COVID-19 pandemic, a period with an unusually high level of WFH and fear of infection. Although of course the pandemic was one of the reasons for analysing links between train use, WFH and fear of infection, performing similar analyses with data from a non-pandemic time (e.g., post-COVID-19), or a combination of stated and revealed preference data would be of interest. Doing so may better indicate, for instance, how attitudes toward train use influence WFH (attitudes) and whether the current travel adjustments would linger beyond the pandemic. In addition, once WHF becomes completely voluntary, it can be expected that the effects between WFH and the attitude towards WFH would also become stronger. These research directions may be explored when new waves of the panel become available (at the time of writing this article, the panel consists of 7 waves, but even in the last wave restrictions existed).
Thirdly, the current study focuses only on train travel while dismissing other travel modes. Future studies should include multiple travel modes to explore a full range of effects of COVID-19 on mode shift, public transport ridership, and travel rebound effects of WFH. Specifically, the fear of infection may actually have increased car use and the use of active modes, and attitudes toward these modes. In a post-pandemic world, it is also possible that WFH may not only result in fewer commute trips, but also in more frequent leisure and shopping trips due to the saved time of not commuting (i.e., rebound effect). These leisure/shopping trips are often shorter than commute trips (which are often covered by train) and may be taken by active modes or car.
Fourthly, even though the present study has established significant (within-person) effects between various concepts, i.e. the fear of infection, (the attitude towards) travel by train, and (the attitude towards) WFH, a qualitative research approach would be required to uncover the causal mechanisms and processes behind the specific observed effects. Such an approach would be better suited to empirically validate the model of attitude change as proposed by Van Wee et al. (2019) and assess which particular mechanism (affective, cognitive, behavioural or combination) indeed underlies the observed effects.
Finally, it would be worthwhile to consider other statistical models to explore the panel dataset. In this regard, a specific direction would be to estimate a latent class trajectory model (Muthén and Muthén, 2000). Such a model would be able to reveal the various latent trajectories in the dependent variables (WFH, train use and the attitudinal variables) over the considered time period. The trajectories thus revealed would show to what extent initial reactions echo through in later waves, and whether there are groups that (quickly or slowly) revert to per-COVID-19 behavioural/attitudinal patterns or groups that have durably changed their behaviour/attitudes.
To conclude, there is much scope to increase our understanding of attitude-behavior dynamics. A main challenge will be to connect (new) dynamic theories of travel behaviour (change) with suitable modelling frameworks. We look forward to addressing this challenge and welcome other researchers to join us in this effort.
CRediT authorship contribution statement
Maarten Kroesen: Conceptualization, Methodology, Formal analysis, Writing – original draft. Jonas De Vos: Writing – original draft, Writing – review & editing. Huyen T.K. Le: Writing – original draft, Writing – review & editing. Danique Ton: Writing – original draft, Writing – review & editing, Investigation, Data curation.
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.
Acknowledgments
We would like to thank Herman Dirkzwager for helping out with exploring the set-up of the initial models.
1 Note that the model 1 effects are not (yet) controlled for working from home (which model 2 does). To assess whether the differences in the effects between model 1 and 2 are due to this correction for confounding or due to the difference in sample composition, we estimated model 1 (with the 3 variables) using the data from model 2, i.e. the subsample of employed people who can work from home. The estimated effects remain close to the estimated coefficients of model 2, indicating that the differences in the estimates between model 1 and 2 are primarily due to the difference in sample composition (rather than the presence of WFH as additional control variable).
==== Refs
References
Abdullah M. Ali N. Hussain S.A. Aslam A.B. Javid M.A. Measuring changes in travel behaviour pattern due to COVID-19 in a developing country: A case study of Pakistan Transp. Policy 108 2021 21 33
Abou-Zeid M. Ben-Akiva M. Travel mode switching: comparison of findings from two public transportation experiments Transp. Policy 24 2012 48 59
Ajzen I. The theory of planned behaviour Org. Behaviour Human Decision Processes 50 1991 179 211
Bamberg S. Ajzen I. Schmidt P. Choice of travel mode in the theory of planned behaviour: The roles of past behaviour, habit, and reasoned action Basic Appl. Soc. Psychol. 25 2003 175 187
Beck M.J. Hensher D.A. Wei E. Slowly coming out of COVID-19 restrictions in Australia: Implications for working from home and commuting trips by car and public transport J. Transp. Geogr. 88 2020 102846
Capasso da Silva D. Khoeini S. Salon D. Conway M.W. Chauhan R.S. Pendyala R.M. Shamshiripour A. Rahimi E. Magassy T. Mohammadian A. Derrible S. How Are Attitudes Toward COVID-19 Associated with Traveller Behaviour During the Pandemic? Findings 2021 10.32866/001c.24389
De Haas M. Faber R. Hamersma M. How COVID-19 and the Dutch ‘intelligent lockdown’ change activities, work and travel behaviour: Evidence from longitudinal data in the Netherlands Transport. Res. Interdisciplinary Perspect. 6 2020 100150
De Vos J. Singleton P.A. Travel and cognitive dissonance Transp. Res. A Policy Pract. 138 2020 525 536
De Vos J. Singleton P.A. Gärling T. From attitude to satisfaction: introducing the travel mode choice cycle Transp. Rev. 2021 1 18
Dobson R. Dunbar F. Smith C.J. Reibstein D. Lovelock C. Structural models for the analysis of traveller attitude behaviour relationships Transportation 7 1978 351 363
Eagly, A.H., Chaiken, S., 1993. The psychology of attitudes. Harcourt brace Jovanovich college publishers.
Festinger, L., 1957. Theory of Cognitive Dissonance. Stanford University Press.
Fujii S. Kitamura R. What does a one-month free bus ticket do to habitual drivers? Transportation 30 2003 81 95
Galleguillos-Torres M. Brouillet C. Molloy J. Axhausen K. Zani D. Van Strien M. Grêt-Regamey A. Do We Have Enough Recreational Spaces during Pandemics? An Answer Based on the Analysis of Individual Mobility Patterns in Switzerland Landsc. Urban Plan. 221 2022 104373
Golob T.F. Joint models of attitudes and behaviour in evaluation of the San Diego I-15 congestion pricing project Transp. Res. A 35 2001 495 514
Hamaker E.L. Kuiper R.M. Grasman R.P. A critique of the cross-lagged panel model Psychol. Methods 20 1 2015 102 25822208
Heider, F., 1958. The psychology of interpersonal relations. Lawrence Erlbaum Associates.
Hook H. De Vos J. Van Acker V. Witlox F. Does undirected travel compensate for reduced directed travel during lockdown? Transport. Lett. 13 5–6 2021 414 420
Kar A. Le H.T.K. Miller H.J. What Is Essential Travel? Socioeconomic Differences in Travel Demand in Columbus, Ohio, during the COVID-19 Lockdown Ann. Am. Assoc. Geogr. 2021 1 24
Kroesen M. Chorus C. A new perspective on the role of attitudes in explaining travel behaviour: A psychological network model Transp. Res. A Policy Pract. 133 2020 82 94
Kroesen M. Handy S. Chorus C. Do attitudes cause behaviour or vice versa? An alternative conceptualization of the attitude-behaviour relationship in travel behaviour modelling Transp. Res. A Policy Pract. 101 2017 190 202
Lee Y. De Vos J. Who would continue working remotely in Hong Kong as the pandemic progresses? Transp. Res. D 2022 submitted for publication
McCarthy L. Delbosc A. Kroesen M. de Haas M. Travel attitudes or behaviours: Which one changes when they conflict? Transportation 2021 10.1007/s11116-021-10236-x
Mirtich, L., Conway, M.W., Salon, D., Kedron, P., Chauhan, R.S., Derrible, S., Khoeini, S., Mohammadian, A., Rahimi, E., Pendyala, R., 2021. How Stable Are Transport-Related Attitudes over Time? Findings, June. https://doi.org/10.32866/001c.24556.
Mohammadian, A., Javadinasr, M., Mohammadi, M., Rahimi, E., Derrible, S., Conway, M., Salon, D., Pendyala, R., 2022. The enduring effects of the pandemic on travel behaviour. Paper presented at the Transportation Research Board (TRB) 101st Annual Meeting.
Muthén B. Muthén L.K. Integrating person-centered and variable-centered analyses: Growth mixture modeling with latent trajectory classes Alcohol. Clin. Exp. Res. 24 6 2000 882 891 10888079
Olde Kalter M.J. Puello L.L.P. Geurs K.T. Exploring the relationship between life events. mode preferences and mode use of young adults: A 3-year cross-lagged panel analysis in the Netherlands Travel Behav. Soc. 24 2021 195 204
Rubin O. Nikolaeva A. Nello-Deakin S. te Brömmelstroet M. What can we learn from the COVID-19 pandemic about how people experience working from home and commuting Centre for Urban Studies, University of Amsterdam 1 9 2020
Salon, D., Conway, M.W., Capasso da Silva, D., Chauhan, R.S., Derrible, S., Mohammadian, A., Khoeini, S., Parker, N., Mirtich, L., Shamshiripour, A., Rahimi, E., & Pendyala, R.M. (2021). The Potential Stickiness of Pandemic-Induced Behavior Changes in the United States. Proceedings of the National Academy of Sciences. 118(27). e2106499118.
Shamshiripour A. Rahimi E. Shabanpour R. Mohammadian A. How is COVID-19 reshaping activity-travel behaviour? Evidence from a comprehensive survey in Chicago Transport. Res. Interdisciplinary Perspect. 7 2020 100216
Statistics Netherlands (CBS), 2022a Arbeidsparticipatie naar leeftijd en geslacht. Accessible at: https://www.cbs.nl/nl-nl/visualisaties/dashboard-arbeidsmarkt/werkenden/arbeidsparticipatie-naar-leeftijd-en-geslacht#:∼:text=Tussen%20het%20tweede%20kwartaal%20van,jonge%20mannen%20tot%2025%20jaar.
Statistics Netherlands (CBS), 2022b. Banen van werknemers naar woon- en werkregio. Accessible at: https://opendata.cbs.nl/statline/#/CBS/nl/dataset/83628NED/table.
Tardiff T.J. Causal inferences involving transportation attitudes and behaviour Transp. Res. 11 1977 397 404
Ton D. Arendsen K. de Bruyn M. Severens V. van Hagen M. van Oort N. Duives D. Teleworking during COVID-19 in the Netherlands: Understanding behaviour, attitudes, and future intentions of train travelers Transp. Res. A 159 2022 55 73
Ton, D., de Bruyn, M., van Hagen, M., Duives, D., van Oort, N., 2022b. Monitoring the impact of COVID-19 on the travel behaviour of train travelers in the Netherlands. Proceedings of the 12th International Conference on Transport Survey Methods.
Triandis, H.C., 1977. Interpersonal behaviour. Brooks/Cole Publishing Company.
Van Hagen M. de Bruyn M. Ton D. Severens V. Duives D. van Oort N. Train traveller behaviour during and after Covid: insights of a longitudinal survey of Dutch train passengers BIVEC/GIBET Transport Res. Days 2021 1 12
Van Wee B. De Vos J. Maat K. Impacts of the built environment and travel behaviour on attitudes: Theories underpinning the reverse causality hypothesis J. Transp. Geogr. 80 2019 102540
Wang, X., Kim, S.H., Mokhtarian, P.L., 2022. Teleworking Behavior Pre-, During, and Expected Post-COVID: Identification and Empirical Description of Trajectory Types. Paper presented at the Transportation Research Board (TRB) 101st Annual Meeting.
| 0 | PMC9742216 | NO-CC CODE | 2022-12-14 23:45:31 | no | Transp Res Part A Policy Pract. 2023 Jan 12; 167:103560 | utf-8 | Transp Res Part A Policy Pract | 2,022 | 10.1016/j.tra.2022.103560 | oa_other |
==== Front
J Transp Geogr
J Transp Geogr
Journal of Transport Geography
0966-6923
1873-1236
Elsevier Ltd.
S0966-6923(22)00233-2
10.1016/j.jtrangeo.2022.103510
103510
Article
The role of neighbourhood design in cycling activity during COVID-19: An exploration of the Melbourne experience
Naseri Mahsa a⁎
Delbosc Alexa a
Kamruzzaman Liton b
a Department of Civil Engineering, Monash University, Melbourne, Australia
b Monash Art Design and Architecture, Monash University, Melbourne, Australia
⁎ Corresponding author at: Room 107, Building 36, Clayton Campus, Monash university, VIC 3800, Australia.
12 12 2022
1 2023
12 12 2022
106 103510103510
11 6 2022
9 11 2022
7 12 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
COVID-19 restrictions imposed significant changes on human mobility patterns, with some studies finding significant increases or decreases in cycling. However, to date there is little understanding on how the neighbourhood-level built environment influenced cycling behaviour during the COVID-19 restrictions. As different neighbourhood have different built environment characteristics, it is possible that cycling trends varied across different built environment settings. We aimed to answer this question by examining recreational cycling during different stages of lockdown in Melbourne, Australia. We compared self-reported recreational cycling frequency (weekly) data from 1344 respondents between pre-COVID and two different stages in lockdown. We tested whether the built environment of their residential neighbourhood and different sociodemographic characteristics influenced leisure cycling rates and whether the effect of these factors varied between different stages of COVID-19 restriction. We found that cycling declined significantly during the two stages of COVID-19 lockdown. Cycling infrastructure density and connectivity are two built environment factors that had a significant effect on limiting the decline in leisure cycling during the pandemic. Furthermore, men and younger people had higher cycling rates in comparison to other groups, suggesting that restrictions on indoor activities and travel limits were not enough to encourage women or older people to cycle more during the pandemic.
Keywords
Built environment
Recreational cycling
Equity
COVID-19
Mobility pattern
Socio-demographic characteristics
Connectivity
Cycling infrastructure
Gender
Age
==== Body
pmc1 Introduction
There is a close relationship between the built environment and active transport (Sallis et al., 2018). Exposure to a supportive built environment such as green space can affect people's attitudes toward active travel and consequently their travel behaviour (Barajas, 2019; Cheng et al., 2019). The built environment and socio-demographics are the two important factors determining cycling behaviour. Research has shown that these two factors often have interdependencies. For example, low-income people living in high-density areas are less likely to cycle compared to higher-income people for commute purpose (Tortosa et al., 2021; Hino et al., 2014).
During COVID-19, many cities implemented significant restrictions on travel distances or reasons to leave the home. These restrictions limited people's exposure to the built environment, but they also offered opportunities for many people to explore their local surroundings. In addition, the closure of gyms and indoor recreation in many cities meant that, for many people, cycling was one of the few remaining sources of physical activity. In the first year of the pandemic, many cities reported significant increases in cycling rates (Schweizer et al., 2021; Fuller et al., 2021) and bicycle purchases(Buehler and Pucher, 2021). As a result, it is possible that the established relationships between the built environment and cycling behaviour did not hold during the pandemic. It is even possible that the role of demographics changed during the height of the pandemic; for example, perhaps women were more willing to take up cycling during the pandemic because car traffic levels were much lower.
In Melbourne, Australia, residents experienced some of the longest and toughest lockdowns in the world (263 days in total). Two significant ‘lock-down’ periods in 2020 restricted travel to within 5 km of people's homes. Although some cities, including Melbourne, implemented temporary or permanent cycling infrastructure as an immediate response to COVID-19 (Kraus and Koch, 2021; Dunning and Nurse, 2020), residents in most cases had to cycle within their immediate neighbourhood due to travel restrictions. The regulatory changes in Melbourne highlighted the role of the local built environment on cycling activity. In ‘normal’ times this relationship may be somewhat clouded as people can travel to distant locations to enjoy a walk or bicycle ride. Given that a recreational cycling trip length is usually longer than cycling for transport (Xing et al., 2010), the 5 km travel restriction experienced in Melbourne in 2020 is likely to have affected cyclist's usual behaviour and provides a unique opportunity to explore the role of neighbourhood built environment on cycling for exercise or leisure.
Hence there is scope to explore whether the built environment interacted with demographics factors to influence change in leisure cycling during the 2020 COVID-19 lockdowns. This study has the following objectives:1) To examine changes in leisure cycling during COVID-19 and how these relate to the local built environment.
2) To determine whether built environment effects varied between different demographic groups during COVID-19.
Research on those two main objectives provide insight on the interaction between leisure cycling, the built environment and socio demographic characteristics.
2 Literature review
2.1 COVID-19 and cycling
On January 2020 the World Health Organization (WHO) declared an emergency in regards to COVID-19. Consequent responses such as social distancing, movement restrictions, work from home and travel bans were introduced in different countries and communities. People's movement and public activities were greatly restricted, causing mobility disruption and modal shift (Barbieri et al., 2020). As a consequence of these restrictions, the amount and pattern of travel changed. This change in mobility pattern is not identical for all transport modes (Bucsky, 2020) and for all neighbourhoods. Road networks, which are dominated by cars and play a major role in urban mobility, were empty due to travel restrictions, working from home, e-learning and reduced number of public activities (De Vos, 2020).
Among other modes of transport, cycling received significant attention both as a mode for transport and leisure activity. Cycling was popular alternative transport modes during the peak of COVID-19 because it provided both exercise and mobility without close exposure to other people who might spread the virus (Budi et al., 2021). In response, many cities reallocated urban spaces for cycling and improved their active transport infrastructure in the hope that increased active travel would be maintained as COVID-19 receded (Combs and Pardo, 2021; Shamshiripour et al., 2020).
It should be noted that there is a disagreement over whether cycling increased or decreased during the pandemic. Some studies found a increases in cycling rates (Buehler and Pucher, 2021; Venter et al., 2020) and Strava data in Germany showed cycling in urban public green spaces increased (Schweizer et al., 2021). However other studies found a decrease in cycling rates (Patterson et al., 2021). Clearly there is still much unknown about the interaction between COVID-19 and cycling, and the built environment is likely one key to this relationship.
2.2 Built environment and cycling
The built environment is the human-made surroundings that support human needs. A long history of research explores how built environment characteristics affect travel behaviour in general and cycling in particular (Ewing and Cervero, 2010). This history suggests characteristics of the built environment such as density (Cervero and Duncan, 2003), green space percentage (Pasha et al., 2016), land-use diversity (Cervero et al., 2019), residential area percentage (Sun et al., 2017) and street connectivity (Cerin et al., 2017) may impact people's cycling behaviour by providing more or fewer opportunities to safely reach local destinations such as shops or parks.
Higher residential density is associated with increasing rates of cycling, at least for transportation (as opposed to recreation) (Beenackers et al., 2012; Saelens et al., 2003). Parks and green space in urban environments provide contact with nature and may increase recreational cycling (Mäki-Opas et al., 2016; Beenackers et al., 2012). Land-use mix (LUM) quantifies the heterogeneity of the built environment and has been shown to be associated with physical activity. More diverse land-use may provide a broader range of destinations within close proximity, which in turn is associated with more active travel trips in a neighbourhood (Saelens et al., 2003). Connectivity, defined as the density of routes (sidewalk/ bike lanes) and connections between them, improves the directness of travel between two points (Jia et al., 2019). Previous studies have demonstrated positive association between street connectivity with physical activity and active transport (Yang et al., 2019). Finally, a higher density of cycling infrastructure facilitates opportunities for cycling (Mölenberg et al., 2019; Hull and O’holleran, 2014).
However, as yet very little is known about whether differences in the local built environment played a role in differences in people's cycling response to COVID-19. Since there are different built environment characteristics in each neighbourhood, it is possible that variations in the response to COVID-19 restrictions may be associated with some differences across different neighbourhoods. One study of active travel in Shiraz, Iran found people's self-reported perception of the quality of urban vegetation, safety of intersections, and cycling infrastructure played an important role in active travel during COVID-19 (Shaer et al., 2021). A study in the UK found that, although leisure cycling increased on ‘standard’ roads during lockdowns, the increases were much greater increases in safe cycling lanes and off-road paths (Hong et al., 2020).
This leads us to explore the effects of the built environment on leisure cycling behaviour before and during the pandemic. The effect of the built environment on cycling is not necessarily constant over time. It could be that the effect of some characteristics may be stronger or weaker during the pandemic compared to before. For example, diverse land-use could encourage more leisure cycling if interesting destinations are closer to home. But during the pandemic, when non-essential businesses were closed, diverse land-use may no longer entice people to ride in their local neighbourhoods.
2.3 Socio-demographics and cycling
Socio-demographic characteristics may have an effect on leisure cycling, since people with different characteristics are likely to have different preferences or concerns. Moreover, it is possible that people with different sociodemographic characteristics have different access to cycling infrastructure or other encouraging built environment elements.
Income, education, immigration status and cultural norms can all influence preferences and attitudes toward cycling(Barajas, 2020; Barajas, 2019). But the most noticeable examples of demographic influences are the role of gender and age. In most countries, men are more likely to ride a bicycle than women (Wittmann et al., 2015; Prati et al., 2019; Heesch et al., 2012). This is primarily because on average, women are more likely to prefer separated cycling infrastructure because it increases their sense of safety (Aldred et al., 2017). Therefore, it is possible men and women living in the same neighbourhood have equal access to cycling infrastructure, but in practice women's participation in cycling is less than men because of the difference in sense of safety.
Age is another significant factor in studies of cycling. In many countries cycling is more likely to occur among young rather than older adults (Bourke et al., 2019; Zander et al., 2013; Grudgings et al., 2021). Aldred et al. (2016) examined places where cycling rates were increasing to determine whether older adults and women were also increasing their cycling rates. Their results found no improvement in representation among older people or women. A possible reason for this lack of diversification is lack of change in the environment which did not reflect the needs of female or older cyclists.
In addition, people with different sociodemographic characteristics may be exposed to different types of built environments and cycling infrastructure. Income, education and employment status are three closely related sociodemographic characteristics related to uneven distribution of cycling infrastructure. In many cities, low-income people are forced to live far from the city centre and they experience long commutes between their home and workplace which cause reduction of cycling rates as a commute mode. Moreover, many analyses have found that low-income people have poor access to cycling infrastructure whereas more affluent areas have better access to cycling infrastructure (Winters et al., 2018; Pistoll and Goodman, 2014; Flanagan et al., 2016; Braun et al., 2019).
It may be that COVID-19 created a situation that overcame some of these barriers to cycling for low-income people, women or older adults. For example, because car traffic was extremely low during COVID lockdowns, perhaps women or older adults felt safer riding a bicycle than they might have before lockdowns and increase their cycling rate (Fischer et al., 2022). However, at least one study suggests that COVID-19 has not changed these relationships; a study in the United States found that for most older adults, their cycling habits did not change during COVID and that older adults with a higher income are more likely to report an increase in cycling during COVID-19 (Gladwin and Duncan, 2022).
3 Data sources
3.1 Study area
Melbourne is the capital and most-populous city of the Australian state of Victoria. We chose Melbourne as a case study location because it experienced among the world's longest and most restrictive lockdowns (263 days of lockdown in total). The first lockdown restrictions were progressively implemented by Australian government in mid-late March 2020 to restrict citizens' movements and reduce their opportunities to gather with other people outside their household. It lasted until mid-May 2020. The second wave of infections emerged in Victoria during May and June, which led to a second lockdown in late June which eventually lasted almost four months. In these lockdown periods, a 5-km travel restriction forced people to stay in their neighbourhood and use their local infrastructure for exercise and leisure. Schools were closed for much of these periods and any non-essential workers were forced to work from home.
These two lockdowns in Melbourne substantially disrupted normal urban life. However, they also provided an opportunity to conduct a case study to evaluate changes in active travel behaviour across three stages (pre-COVID, first-, and second-lockdowns) and how these changes relate to the built environment in which people lived.
3.2 Survey data
We used two different datasets to compare recreational cycling behaviour, which is done for purpose of pleasure not reaching to a specific destination, in different built environments. First, we used a self-reported questionnaire survey data that measured travel behaviour in different stages of lockdown. The survey was conducted as part of the ‘C-19 Long Term Transport Impact Study’ in Melbourne (Currie et al., 2021). This questionnaire gathered a wide range of information about respondent travel behaviour, commute mode, changes to working from home, attitudes and socio-demographics. Although the survey gathered a range of travel behaviour information from the respondents, we only explored their self-reported recreational cycling data in this paper.
The questionnaire survey was conducted through an online panel survey company. The survey was run in two different waves; the first lag of data covered the first lockdown in Melbourne and the second one gathered after the second lockdown and latest travel restrictions were announced. As we are interested in whether recreational cycling behaviour changed during the two different lockdown periods, we analysed data from the second wave of this dataset (sample size 1341 responses), which ran from 16 July to 8 August 2020. The characteristics of the sample are presented in Table 1 . It shows that the survey sample was over-representative of females but the income and age characteristics were broadly similar to the Melbourne average.Table 1 Sociodemographic characteristics of sample.
Table 1Socio demographic characteristic Sample Percentage Melbourne average (ABS 2016)
Gender Male 37.9% 49.0%
Female 62.0% 51.0%
Income Low (0-$530 per week) 34.0% 38%
Medium ($530–$1870 per week) 48.7% 42.2%
High (> $1870 per week) 17.3% 19.8%
Age Youth (15–24) 11.8% 13.4%
Adult (25–64) 72.0% 54.3%
Older adult (>65) 16.1% 14.1%
Employment Employed 55.4% 88.6%
Non-worker 36.2% 6.8%
Not in labour force 8.4% 4.6%
Education Non-university degree 40.5% 72.5%
Higher education 59.5% 27.5%
Respondents were asked to indicate how often they cycled for recreational purposes in three different stages of the lockdown: Stage 1- pre-COVID, Stage 2- during first lockdown, and Stage 3- during second lockdown. Note that the two lockdown periods were during autumn and winter of 2020 when cycling rates are historically lower (Pazdan, 2020).
Responses were recorded on a seven-point scale:0. Didn't do this
1. 1 time a week
2. 2 times a week
3. 3 times a week
4. 4 times a week
5. 5 times a week
6. >5 times a week
Note that because these cycling behaviours were measured retrospectively, changes between ‘pre-COVID’ and the second wave of COVID rely on respondent recall.
3.3 Built environment data
The second set of data set that we used in this research is open spatial data to quantify built environment characteristics. We linked the cycling questions and socio-demographics of respondents to the built environment of the residential location reported by respondents. As we only knew the suburb of the respondents, we had to use the spatial resolution of SA2 statistical area for this analysis, which matches closely to the suburb boundary.
Based on the literature, we calculated built environment indicators including residential density (Forsyth et al., 2007), green space area (Watts et al., 2013), and land-use mix (Cervero and Kockelman, 1997), street connectivity (Bentley et al., 2018), and cycling infrastructure coverage (Buehler and Dill, 2016).
To quantify the built environment characteristics, baseline data were downloaded from Victorian Government open data and GIS methods are utilised to form indicators. For dwelling numbers in each neighbourhood, the 2016 census data were used. Green space percentage was calculated based on the total area of parklands within each neighbourhood. Residential area percentage was determined based on the total area of residential blocks in each neighbourhood, and residential density was based on the number of dwellings per Km2 in each neighbourhood.
Land-use diversity index was measured by simpson's diversity index (Eq. (1)) in which a is the total area of specific land-use in a neighbourhood and A is the total area of all land-use categories within the neighbourhood. This indicator value ranges from 0 to 1; and a higher value indicates more diverse land-use pattern (Kamruzzaman and Hine, 2013).(1) Land−usediversity=1−∑aA2
Cycling infrastructure data were downloaded from Vicroads and AURIN and by applying GIS methods, we calculated the length of cycling infrastructure in each neighbourhood. In order to have a normalized comparision among different neighbourhoods, we divided total length of bike lanes (km) by neighbourhood area (km2). For connectivity, we extract the number of 3 (or more) way intersections and divided by neighbourhood area (km2).
In order to conduct a comparative analysis between built environment characteristics and cycling dynamic, we considered quartiles of each built environment characteristic. The distribution of built environment variables is provided in Table 2 and the spatial distribution of them is provided in Fig. 1 .Table 2 Built environment measurement method and their quartile measurement.
Table 2Measure Measurement method Lowest quartile Second quartile Third quartile Highest quartile
Residential Area percentage Area of residential blocks divided by neighbourhood area (Km2) 0–34.47 34.47–58.89 58.89–75.68 75.68–97.66
Residential density Number of dwellings divided by neighbourhood area (Km2) 0–581.79 581.79–918.825 918.825–2075.586 2075.586–6874.92
Green space percentage Area of parklands divided by neighbourhood area (Km2) 0–6.09 6.09–11.38 11.38–18.38 18.38–66.31
Land use diversity Simpson diversity index 0–0.38 0.38–0.52 0.52–0.68 0.68–2.29
Cycling infrastructure Bicycle lane length (km) divided by neighbourhood area (Km2) 0–0.71 0.71–1.66 1.66–3.24 3.24–7.09
Connectivity Number of intersections divided by neighbourhood area (Km2) 2.36–54.07 54.07–65.85 65.85–105.42 105.42–225.03
Fig. 1 Spatial distribution of different built environment characteristics.
Fig. 1
4 Analytical methods
In this study we used two different analyses to understand the cycling trend during different stages of lockdown. First, we used descriptive statistics to analyse the distribution of leisure cycling across the 3 different stages of COVID-19. To determine if there was a statistically significant change in cycling frequency between the three time periods (before-COVID, during-COVID and between-waves), we used Cochran's Q test, which is an extension of McNemar test based on Chi square distribution for testing differences between repeated data. Then, we compared the mean cycling rates across different built environment and socio-demographic characteristics.
In the second stage of the results, we used multivariate models to isolate the association between the built environment, demographics and time period on cycling rates. Because we asked respondents to report their cycling frequency for three different time periods, we needed to use repeated-measures modelling. After checking the cycling data distribution, we found that our data set has a negative-binomial distribution. Therefore, we used multilevel modelling to analyse repeated measures data with a negative-binomial distribution. As our explanatory variables (built environment and socio-demographics) do not change during the study time, we consider them as fixed effects in all models. This method gives us the flexibility to investigate potential interactions between different variables.
In order to test the main effects of COVID stage, built environment and demographics, as well as their potential interactions, we formulated a series of six models. Models 1 and 2 only considered the main effects, testing whether demographics improve the model fit. Model 3 explores the research aim of whether or not the effect of built environment varied depending on COVID stage, by adding a series of interactions between the two variables. Model 4 checks whether the influence of demographics varies based on COVID stage through interaction effects. And finally, Models 5 and 6 test whether the influence of the built environment interacts with participants' gender.
In order to interpret the effects of the models, we graphed out the estimated marginal means from the model results, using a contrast test to determine whether there were statistically significant differences between groups.
5 Descriptive results
5.1 Leisure cycling dynamics and the built environment
In this study, leisure cycling rates declined during COVID-19. Table 3 shows that overall, leisure cycling declined during the first and second lockdown stages. The changes in mean cycling level between different stages were statistically significant (F(2,4020) = 7.467, p < 0.01) based on the multilevel modelling results by considering time in the model. Similarly, the percentage of people who choose to ride a bike as their leisure physical activity dropped from 24.5% before COVID-19 to 21.0% in the first lockdown and 16.8% in the second lockdown. Based on the Cochran's Q test results for cycled / did not cycle, the changes in cycling level during different stages of lockdown is statistically significant (χ2(2) = 72.518, p < 0.01).Table 3 Leisure cycling distribution during the pandemic in Melbourne.
Table 3 Before COVID-19 First lockdown Second lockdown
Didn't do this 75.5% 79.0% 83.2%
1 per week 9.3% 8.4% 6.8%
2 per week 5.5% 4.7% 3.2%
3 per week 4.5% 3.8% 2.8%
4 per week 2.3% 1.9% 1.9%
5 per week 1.4% 1.1% 1.1%
5+ per week 1.4% 1.1% 1.1%
Total cycled 24.5% 21.0% 16.8%
Mean 0.59 0.49 0.41
Std. Deviation 1.27 1.16 1.11
Note: Change in mean and percentage who cycled were both statistically significant.
We now examine whether the built environment influenced leisure cycling rates, perhaps ‘protecting’ against the impact of COVID-19. Fig. 2 shows the changes in leisure cycling during different lockdown stages by built environment characteristics. As we would expect, people living in neighbourhoods with higher cycling infrastructure density were more likely to cycle pre-COVID and also during the first lockdown. However, by the second lockdown these rates had dropped to be similar in all areas.Fig. 2 Dynamics of leisure cycling by considering built environment characteristic.
Fig. 2
Unexpectedly, people who live in the lowest quartile of green space had the highest cycling level before COVID-19 but their cycling level dropped significantly in lockdown. This may be because places with lower green space also tended to have higher cycling infrastructure density (see Fig. 1). As with cycling infrastructure, by the second lockdown cycling rates are similar in all areas.
Land-use diversity does not appear to interact with cycling rates either before or during COVID. In contrast, street connectivity was clearly associated with higher cycling rates pre-COVID. Although cycling rates reduced, they remained higher among highly connected neighbourhoods.
The effect of residential land use varies depending on whether you looked at residential density or residential area percentage. In higher residential density areas, we saw higher cycling levels pre-pandemic. Although they dropped during the COVID-19 lockdowns, the rate remained higher high-density areas compared to low-density areas. In contrast, people who lived in neighbourhoods with the lowest residential percentage had the highest level of cycling but this dropped off to the lowest levels during COVID. An area with lowest residential percentage means that these are non-residential in nature (e.g. predominantly commercial, industrial), which remained closed during the lockdowns providing little opportunities for cycle.
5.2 Leisure cycling dynamics and socio-demographic characteristics
Demographic characteristics play an important role in decisions to cycle, and they may influence whether someone persists with cycling during the pandemic. For this reason, we compared key demographic variables across the pandemic periods.
Fig. 3 shows the changes in leisure cycling during lockdown stages based on socio-demographic characteristics. Similar to past research, people were more likely to cycle for leisure pre-COVID if they were male, a young adult, employed, medium income and with a university degree. Cycling rates fell for all demographic groups during COVID and at similar rates.Fig. 3 Dynamics of leisure cycling by considering socio-demographic characteristics.
Fig. 3
6 Multivariate results
In the next step we assess the interaction of built environment and socio-demographic characteristics on leisure cycling during pandemic stages.
From the descriptive analysis, it appears that some demographic and built environment characteristics such as gender and residential area percentage may be influencing cycling rates before and during COVID. In order to explore the combined influence of built environment and socio-demographic characteristics, we conducted a series of multilevel models for repeated-measure data to examine the association between built environment characteristics, socio-demographics and cycling levels over time.
6.1 Fixed effects models
In this section, we interpret the initial fixed effects F-tests of Models 1 to 6 (in Table 4 ), to determine whether each variable is a significant contribution to the overall model. The individual coefficients are interpreted for the two best-fitting models (Models 2 and 6) using estimated marginal means. The full set of coefficient results are provided in Appendix A.Table 4 Fixed effects F-tests between built environment characteristics, socio-demographics, and leisure cycling.
Table 4 Model 1
Base model: built environment Model 2
Base model: BE + demographics Model 3
BE and time interaction Model 4
Demographics and time interaction Model 5
BE and gender interaction Model 6
Gender and residential percentage interaction
COVID stage 7.12* 6.57* 6.73* 2.16 7.58* 6.90*
Residential area percentage 2.67* 1.24 1.40 1.22 1.19 1.34
Residential density 2.59* 2.13 2.23 2.12 2.61* 2.05
Green space percentage 2.20 1.65 1.74 1.62 1.37 1.39
Land-use diversity 1.04 1.70 1.82 1.68 1.69 1.81
Cycling infrastructure density 5.66* 4.26* 4.41* 4.28* 5.21* 4.61*
Connectivity 4.17* 2.21 2.18* 2.22 3.36* 2.59*
Gender – 88.13* 88.39* 87.19* 77.10* 84.98*
Income – 2.83 2.83 2.79 2.77 2.39
Age – 10.89* 10.68* 10.93* 12.24* 10.95*
Employment – 0.35 0.40 0.42 0.31 0.48
Education – 2.07 1.99 1.92 3.87* 1.75
Residential area percentage * COVID stage – – 0.60 – – –
Residential density * COVID stage – – 0.33 – – –
Green space percentage * COVID stage – – 0.26 – – –
Land-use diversity * COVID stage – – 0.27 – – –
Cycling infrastructure density * COVID stage – – 0.51 – – –
Connectivity * COVID stage – – 0.38 – – –
Gender * COVID stage – – – 0.16 – –
Income * COVID stage – – – 0.29 – –
Age * COVID stage – – – 0.07 – –
Employment * COVID stage – – – 0.16 – –
Education * COVID stage – – – 0.38 – –
Residential area percentage * Gender – – – – 3.06* 3.49*
Residential density * Gender – – – – 0.17 –
Green space percentage * Gender – – – – 2.45 –
Land-use diversity * Gender – – – – 0.26 –
Cycling infrastructure density * Gender – – – – 0.68 –
Connectivity * Gender – – – – 2.48 –
Model fit statistics
AIC 6874.05 6754.95 6813.95 6783.78 6762.59 6749.47
BIC 7012.40 6943.48 7227.50 7072.48 7063.80 6956.80
Log likelihood 6829.80 6694.48 6679.71 6690.69 6665.41 6682.91
The first two models are ‘base models’ that test the effects of built environment and demographics on leisure cycling rates over time. In this and every model except model 4, the impact of COVID stage was significant, reflecting the decrease in leisure cycling across the two waves of COVID. In Model 1, residential area percentage, residential density, cycling infrastructure density, and connectivity have significant effects on cycling rates. In Model 2, including demographics reduced the number of significant built environment variables to only include cycling infrastructure density, as well age and gender. Because including demographics improved the model fit (AIC reduced by 119 and BIC reduced by 69), they were included as control variables in the remaining models.
One of our research questions was to explore whether built environment characteristics might ‘protect against’ reductions in cycling during COVID. For this reason, in Model 3 we included the interaction between built environment parameters and time. None of the interactions were statistically significant, suggesting that the effect of built environment on leisure cycling was the same regardless of the time period.
On a similar vein, we tested whether the effect of demographic characteristics varied depending on the time period, because people with different demographic characteristics may have different responses to COVID-19 related restrictions (Model 4). Results of this model shows there is not any significant interaction between demographic characteristics and time on leisure cycling.
Finally, because of the well-documented relationship between gender and cycling, we considered a detailed look at the interaction between gender and built environment variables (Model 5). Based on this model's results, residential area percentage has a significant interaction with gender. Finally, to isolate this interaction effect, we ran a final model (Model 6). In this model the interaction of gender and residential area percentage is statistically significant. Moreover, in this model connectivity, cycling infrastructure density, residential density, and age have significant effect on leisure cycling. This model has the best fit based on AIC, but a slightly poorer fit than Model 2 using BIC (6956.8 vs 6943.5); this is not surprising as BIC penalises additional parameters more strongly than AIC.
In the next section we interpret the estimated marginal means for the two best-fitting models, Model 2 (Base model with built environment and demographics) and Model 6 (Built environment, demographics and gender by residential area percentage).
6.2 Estimated marginal means
Fig. 4 shows the estimated marginal means of the significant effects from Model 2 (base model of built environment and demographics) and their contrast (shown by an asterisk). The significant variables were COVID stage (Fig. 4-a), cycling infrastructure density (Fig. 4-b), gender (Fig. 4-c) and age (Fig. 4-d). Echoing Table 3, over time we saw a declining trend in leisure cycling, although only the difference between pre-COVID and the second lockdown was statistically significant.Fig. 4 Estimated marginal means for significant fixed effects in Model 2.
Note: a ^ indicates the baseline condition and an asterisk indicates this value is statistically significantly different.
Fig. 4
Cycling infrastructure density has a significant effect on leisure cycling, but the estimated marginal means show that the relationship is not intuitive: the highest quartile of infrastructure density had the lowest leisure cycling rates. This could be because the higher leisure cycling rates in the descriptive comparison (Fig. 2) was masking other factors that could be associated with higher leisure cycling, such as a higher concentration of young adults in places with more cycling infrastructure. The demographic results show that, as found in the descriptive results, men are more likely to cycle for leisure than women and both adults and young adults are more likely to cycle than older adults.
Fig. 5 shows the estimated marginal means for the significant variables of Model 6 and their contrast (shown by a star). The significant variables from Model 2 were also significant in this model, but it also includes connectivity and the interaction between gender and residential area percentage.Fig. 5 Estimated marginal means for significant fixed effects in Model 6.
Note: a ^ indicates the baseline condition and an asterisk indicates this value is statistically significantly different.
Fig. 5
As in Model 2, in this model leisure cycling declined over time but only the difference between pre-COVID and the second lockdown were significant (Fig. 5-a). Also similar to Model 2, leisure cycling was lowest in locations with the highest cycling infrastructure density (Fig. 5-b). People who lived in more connected neighbourhoods cycled more but the contrast test did not find statistically significant differences between individual categories of connectivity (Fig. 5-c). Age was the only demographic characteristics that had a significant effect on the leisure cycling dynamic in Model 6. Leisure cycling had a declining trend by increasing age and the difference between older people cycling level and young people is statistically significant (Fig. 5-d).
Finally, the interaction between gender and residential area percentage was statistically significant. Among women, living in locations with higher residential percentage was associated with more leisure cycling. In contrast, men showed the opposite trend with the lowest leisure cycling in areas with the highest residential percentage. This could suggest that women are responding to the built environment differently to men (Fig. 5- e).
7 Discussion
This paper had two research objectives. In response to the first objective, we found a declining trend for leisure cycling in our survey sample. This was somewhat unexpected as recent studies found increases in non-commute cycling during COVID-19 (Hong et al., 2020), although other studies showed a decrease in cycling rates (Patterson et al., 2021). This study could be reflecting a real reduction in leisure cycling, or it could be due to recall bias among survey participants. It may be, for example, the commuter cyclists are over-estimating how much they cycled before COVID even though we asked them to only report on cycling for leisure.
Moreover, we found that leisure cycling rates were only sensitive to cycling infrastructure density and, to some extent, connectivity. People who lived in neighbourhoods with greater connectivity cycled more, suggesting that greater intersection density encourages leisure cycling. However, the effect of cycling infrastructure density was non-intuitive. In the descriptive comparisons, higher infrastructure density was associated with more leisure cycling, especially pre-COVID. But in the models that control for demographics and other variables, higher infrastructure density was associated with lower cycling. This could be because our built environment measures were not spatially sensitive enough as we could only link survey participants to an SA2 census boundary (a neighbourhood of around 10,000 people). This could also explain why we did not see more significant associations between the built environment and leisure cycling. Another possible explanation could be concentration of bike lanes and infrastructure in Melbourne's inner-city area, where there is less green space and other characteristics that encourage people to cycle for recreation rather than commute.
Interestingly, we did not find any interaction over time between cycling and built environment characteristics. The effects of infrastructure were the same whether we were predicting cycling before COVID-19 or during the first or second lock-down. This is indirect contrast to a paper in the UK that found a greater influence of cycling infrastructure during COVID-19 (Hong et al., 2020).
Finally, we wanted to examine the relationship between leisure cycling and socio-demographic characteristics. As expected from the literature, younger adults and men were more likely to cycle than older adults and women. This did not vary over time; for example, women and older people did not suddenly take up cycling more during COVID-19. We did, however, find that women were more likely to cycle if they lived in more residential neighbourhoods, whereas men were not influenced by the residential area of their home location.
These findings suggest that although planners have tried to make cycling more accessible by adding new cycling infrastructure during the pandemic, there is still a significant gender and age gap in cycling. Being forced to exercise within 5 km of their home, at a time when car traffic was negligible, was not enough incentive to encourage these groups to cycle more, at least not in this study sample. It suggests that more needs to be done to attract these groups to cycling and reflect their needs in future planning.
8 Conclusion
Cycling is potential means of physical activity which can contribute to people's health. It can be considered to be a potential low-cost strategy to improve communities' health and wellbeing. The link between physical activity and health, physical and mental, has long been known (Warburton and Bredin, 2016; Biddle et al., 2019). Supportive infrastructure plays an important role for people with inadequate access to private recreational facilities or limited mobility such as low-income people, youth, and women. These vulnerable groups are already at greater risk of physical inactivity; therefore, it is important to consider them in neighbourhoods planning. This need was recognized by the United Nations who have codified the aim of providing “universal access to safe, inclusive and accessible, green and public spaces, particularly for women and children, older persons and persons with disabilities” in their Sustainable Development Goals (United Nations General Assembly, 2015).
Although the travel restrictions of COVID-19 are an extremely unique circumstance that is unlikely to be repeated, the findings from this paper are likely to be relevant into the future. Many of the most vulnerable people in society – teenagers, older adults, people living in poverty and people who cannot drive – are more likely to be dependent on their local neighbourhood to provide access to safe cycling options. As cycling barriers and incentives are a multidimensional phenomenon, we may be better able to plan for the travel needs of the most vulnerable if we acknowledge the interdependency of socio-economic characteristics and neighbourhood built environment.
CRediT authorship contribution statement
Mahsa Naseri: Conceptualization, Methodology, Writing – original draft, Visualization. Alexa Delbosc: Conceptualization, Methodology, Writing – review & editing. Liton Kamruzzaman: Conceptualization, Methodology, Writing – review & editing.
Declaration of Competing Interest
None.
Appendix A Fixed effect coefficients for Models 1 through 6
Model 1-Base model (built environment) coefficients
Unlabelled TableModel Term Coefficient Std. Error t Sig. Model Term Coefficient Std. Error t Sig.
Intercept −0.91 0.23 −3.89 0.00 Green space percentage = Highest quartile 0b
COVID stage = Pre COVID 0.37 0.10 3.77 0.00 Residential area percentage = Lowest quartile 0.03 0.19 0.15 0.88
COVID stage = Lockdown 1 0.18 0.10 1.82 0.07 Residential area percentage = Second quartile 0.37 0.20 1.82 0.07
COVID stage = Lockdown 2 0b Residential area percentage = Third quartile 0.26 0.18 1.50 0.13
Connectivity = Lowest quartile −0.60 0.24 −2.49 0.01 Residential area percentage = Highest quartile 0b
Connectivity = Second quartile −0.34 0.21 −1.64 0.10 LUD = Lowest quartile −0.20 0.18 −1.11 0.27
Connectivity = Third quartile −0.02 0.18 −0.13 0.89 Land-use diversity = Second quartile −0.13 0.16 −0.81 0.42
Connectivity = Highest quartile 0b Land-use diversity = Third quartile −0.20 0.12 −1.65 0.10
Cycling infrastructure density = Lowest quartile 0.67 0.20 3.27 0.00 Land-use diversity = Highest quartile 0b
Cycling infrastructure density = Second quartile 0.35 0.20 1.76 0.08 Residential density = Lowest quartile −0.30 0.29 −1.03 0.31
Cycling infrastructure density = Third quartile 0.49 0.16 3.13 0.00 Residential density = Second quartile −0.30 0.22 −1.38 0.17
Cycling infrastructure density = Highest quartile 0b Residential density = Third quartile 0.09 0.19 0.47 0.64
Green space percentage = Lowest quartile 0.04 0.15 0.26 0.79 Residential density = Highest quartile 0b
Green space percentage = Second quartile −0.26 0.15 −1.80 0.07 NegativeBinomial 5.16 0.27
Green space percentage = Third quartile −0.18 0.12 −1.55 0.12
Model 2-Base model (BE + demographics) Coefficients
Unlabelled TableModel Term Coefficient Std. Error t Sig. Model Term Coefficient Std. Error t Sig.
Intercept −0.028 0.3449 −0.081 0.935 Land-use diversity = Second quartile −0.177 0.1747 −1.014 0.311
COVID stage = Pre COVID 0.394 0.1091 3.612 0.000 Land-use diversity = Third quartile −0.293 0.1377 −2.124 0.034
COVID stage = Lockdown 1 0.179 0.1101 1.623 0.105 Land-use diversity = Highest quartile 0b
COVID stage = Lockdown 2 0b Residential density = Lowest quartile −0.230 0.3188 −0.720 0.471
Connectivity = Lowest quartile −0.456 0.2687 −1.697 0.090 Residential density = Second quartile −0.234 0.2449 −0.955 0.340
Connectivity = Second quartile −0.384 0.2327 −1.651 0.099 Residential density = Third quartile 0.160 0.2131 0.752 0.452
Connectivity = Third quartile −0.072 0.2042 −0.352 0.725 Residential density = Highest quartile 0b
Connectivity = Highest quartile 0b Gender = Female −0.873 0.0930 −9.388 0.000
Cycling infrastructure densityy = Lowest quartile 0.648 0.2254 2.875 0.004 Gender = Male 0b
Cycling infrastructure densityy = Second quartile 0.363 0.2174 1.669 0.095 Age = adult −0.234 0.1524 −1.532 0.126
Cycling infrastructure densityy = Third quartile 0.501 0.1727 2.902 0.004 Age = older adult −0.902 0.2055 −4.389 0.000
Cycling infrastructure densityy = Highest quartile 0b Age = youth 0b
Green space percentage = Lowest quartile −0.070 0.1629 −0.430 0.667 Education = Higher education 0.145 0.1010 1.438 0.151
Green space percentage = Second quartile −0.325 0.1608 −2.020 0.043 Education = Non-university degree 0b
Green space percentage = Third quartile −0.173 0.1312 −1.319 0.187 Employment = Employed 0.083 0.1856 0.447 0.655
Green space percentage = Highest quartile 0b Employment = Non-worker −0.016 0.1927 −0.081 0.935
Residential area percentage = Lowest quartile −0.214 0.2150 −0.993 0.321 Employment = Not in labor force 0b
Residential area percentage = Second quartile 0.031 0.2264 0.137 0.891 Income = High −0.304 0.1282 −2.375 0.018
Residential area percentage = Third quartile 0.062 0.1986 0.311 0.756 Income = Low −0.032 0.1175 −0.271 0.787
Residential area percentage = Highest quartile 0b Income = Medium 0b
Land-use diversity = Lowest quartile −0.278 0.2027 −1.373 0.170 NegativeBinomial 4.383 0.2795
Model 3 (BE and time interaction) coefficients
Unlabelled TableModel Term Coefficient Std. Error t Sig. Model Term Coefficient Std. Error t Sig.
Intercept −0.02 0.51 −0.05 0.96 [Cycling infrastructure density = Lowest quartile]*[COVID stage = Lockdown 1] −0.30 0.57 −0.52 0.60
COVID stage = Pre COVID 0.34 0.62 0.56 0.58 [Cycling infrastructure density = Second quartile]*[COVID stage = Lockdown 1] −0.16 0.55 −0.30 0.76
COVID stage = Lockdown 1 0.19 0.63 0.30 0.77 [Cycling infrastructure density = Third quartile]*[COVID stage = Lockdown 1] −0.22 0.44 −0.49 0.62
COVID stage = Lockdown 2 0b [Cycling infrastructure density = Highest quartile]*[COVID stage = Lockdown 1] 0b
Connectivity = Lowest quartile −0.32 0.49 −0.65 0.52 [Cycling infrastructure density = Lowest quartile]*[COVID stage = Lockdown 2] 0b
Connectivity = Second quartile −0.53 0.43 −1.23 0.22 [Cycling infrastructure density = Second quartile]*[COVID stage = Lockdown 2] 0b
Connectivity = Third quartile −0.14 0.38 −0.37 0.71 [Cycling infrastructure density = Third quartile]*[COVID stage = Lockdown 2] 0b
Connectivity = Highest quartile 0b [Cycling infrastructure density = Highest quartile]*[COVID stage = Lockdown 2] 0b
Cycling infrastructure density = Lowest quartile 1.04 0.40 2.58 0.01 [Green space percentage = Pre COVID]*[COVID stage = Pre COVID] 0.31 0.40 0.77 0.44
Cycling infrastructure density = Second quartile 0.58 0.39 1.50 0.13 [Green space percentage = Second quartile]*[COVID stage = Pre COVID] 0.06 0.39 0.16 0.87
Cycling infrastructure density = Third quartile 0.69 0.31 2.20 0.03 [Green space percentage = Third quartile]*[COVID stage = Pre COVID] 0.25 0.32 0.78 0.43
Cycling infrastructure density = Highest quartile 0b [Green space percentage = Highest quartile]*[COVID stage = Pre COVID] 0b
Green space percentage = Lowest quartile −0.23 0.29 −0.78 0.44 [Green space percentage = Lowest quartile]*[COVID stage = Lockdown 1] 0.18 0.41 0.45 0.66
Green space percentage = Second quartile −0.42 0.29 −1.46 0.14 [Green space percentage = Second quartile]*[COVID stage = Lockdown 1] 0.19 0.40 0.48 0.63
Green space percentage = Third quartile −0.28 0.23 −1.22 0.22 [Green space percentage = Third quartile]*[COVID stage = Lockdown 1] 0.08 0.33 0.25 0.80
Green space percentage = Highest quartile 0b [Green space percentage = Highest quartile]*[COVID stage = Lockdown 1] 0b
Residential area percentage = Lowest quartile −0.44 0.39 −1.12 0.26 [Green space percentage = Lowest quartile]*[COVID stage = Lockdown 2] 0b
Residential area percentage = Second quartile 0.10 0.40 0.25 0.81 [Green space percentage = Second quartile]*[COVID stage = Lockdown 2] 0b
Residential area percentage = Third quartile 0.05 0.35 0.15 0.88 [Green space percentage = Third quartile]*[COVID stage = Lockdown 2] 0b
Residential area percentage = Highest quartile 0b [Green space percentage = Highest quartile]*[COVID stage = Lockdown 2] 0b
Land-use diversity = Lowest quartile −0.38 0.36 −1.05 0.29 [Residential area percentage = Lowest quartile]*[COVID stage = Pre COVID] 0.40 0.52 0.78 0.44
Land-use diversity = Second quartile −0.18 0.31 −0.57 0.57 [Residential area percentage = Second quartile]*[COVID stage = Pre COVID] −0.22 0.54 −0.40 0.69
Land-use diversity = Third quartile −0.38 0.25 −1.56 0.12 [Residential area percentage = Third quartile]*[COVID stage = Pre COVID] 0.07 0.48 0.14 0.88
Land-use diversity = Highest quartile 0b [Residential area percentage = Highest quartile]*[COVID stage = Pre COVID] 0b
Residential density = Lowest quartile −0.44 0.57 −0.76 0.45 [Residential area percentage = Lowest quartile]*[COVID stage = Lockdown 1] 0.21 0.54 0.38 0.70
Residential density = Second quartile −0.30 0.44 −0.68 0.50 [Residential area percentage = Second quartile]*[COVID stage = Lockdown 1] 0.02 0.56 0.03 0.98
Residential density = Third quartile 0.28 0.38 0.72 0.47 [Residential area percentage = Third quartile]*[COVID stage = Lockdown 1] −0.03 0.49 −0.07 0.94
Residential density = Highest quartile 0b [Residential area percentage = Highest quartile]*[COVID stage = Lockdown 1] 0b
Gender = Female −0.88 0.09 −9.40 0.00 [Residential area percentage = Lowest quartile]*[COVID stage = Lockdown 2] 0b
Gender = Male 0b [Residential area percentage = Second quartile]*[COVID stage = Lockdown 2] 0b
Age = adult −0.23 0.15 −1.50 0.13 [Residential area percentage = Third quartile]*[COVID stage = Lockdown 2] 0b
Age = older adult −0.90 0.21 −4.34 0.00 [Residential area percentage = Highest quartile]*[COVID stage = Lockdown 2] 0b
Age = youth 0b [Land-use diversity = Lowest quartile]*[COVID stage = Lowest quartile] 0.16 0.50 0.32 0.75
Education = Higher education 0.14 0.10 1.41 0.16 [Land-use diversity = Second quartile]*[COVID stage = Pre COVID] −0.21 0.43 −0.49 0.63
Education = Non university degree 0b [Land-use diversity = Third quartile]*[COVID stage = Pre COVID] 0.06 0.34 0.18 0.86
Employment = Employed 0.09 0.19 0.49 0.62 [Land-use diversity = Highest quartile]*[COVID stage = Pre COVID] 0b
Employment = Non worker −0.01 0.19 −0.07 0.95 [Land-use diversity = Lowest quartile]*[COVID stage = Lockdown 1] 0.11 0.51 0.22 0.83
Employment = Not in labor force 0b [Land-use diversity = Second quartile]*[COVID stage = Lockdown 1] 0.19 0.43 0.44 0.66
Income = High −0.30 0.13 −2.37 0.02 [Land-use diversity = Third quartile]*[COVID stage = Lockdown 1] 0.18 0.34 0.54 0.59
Income = Low −0.02 0.12 −0.18 0.86 [Land-use diversity = Highest quartile]*[COVID stage = Lockdown 1] 0b
Income = Medium 0b [Land-use diversity = Lowest quartile]*[COVID stage = Lockdown 2] 0b
[Connectivity = Lowest quartile]*[COVID stage = Pre COVID] −0.32 0.66 −0.48 0.63 [Land-use diversity = Second quartile]*[COVID stage = Lockdown 2] 0b
[Connectivity = Second quartile]*[COVID stage = Pre COVID] 0.30 0.57 0.53 0.59 [Land-use diversity = Third quartile]*[COVID stage = Lockdown 2] 0b
[Connectivity = Third quartile]*[COVID stage = Pre COVID] 0.10 0.50 0.20 0.84 [Land-use diversity = Highest quartile]*[COVID stage = Lockdown 2] 0b
[Connectivity = Highest quartile]*[COVID stage = Pre COVID] 0b [Residential density = Lowest quartile]*[COVID stage = Lowest quartile] 0.68 0.77 0.88 0.38
[Connectivity = Lowest quartile]*[COVID stage = Lockdown 1] −0.07 0.68 −0.10 0.92 [Residential density = Second quartile]*[COVID stage = Pre COVID] 0.31 0.59 0.53 0.60
[Connectivity = Second quartile]*[COVID stage = Lockdown 1] 0.12 0.60 0.20 0.84 [Residential density = Third quartile]*[COVID stage = Pre COVID] 0.00 0.51 0.00 1.00
[Connectivity = Third quartile]*[COVID stage = Lockdown 1] 0.11 0.53 0.20 0.84 [Residential density = Highest quartile]*[COVID stage = Pre COVID] 0b
[Connectivity = Highest quartile]*[COVID stage = Lockdown 1] 0b [Residential density = Lowest quartile]*[COVID stage = Lockdown 1] −0.08 0.81 −0.10 0.92
[Connectivity = Lowest quartile]*[COVID stage = Lockdown 2] 0b [Residential density = Second quartile]*[COVID stage = Lockdown 1] −0.13 0.63 −0.21 0.84
[Connectivity = Second quartile]*[COVID stage = Lockdown 2] 0b [Residential density = Third quartile]*[COVID stage = Lockdown 1] −0.33 0.54 −0.62 0.54
[Connectivity = Third quartile]*[COVID stage = Lockdown 2] 0b [Residential density = Highest quartile]*[COVID stage = Lockdown 1] 0b
[Connectivity = Highest quartile]*[COVID stage = Lockdown 2] 0b [Residential density = Lowest quartile]*[COVID stage = Lockdown 2] 0b
[Cycling infrastructure density = Lowest quartile]*[COVID stage = Pre COVID] −0.86 0.55 −1.55 0.12 [Residential density = Second quartile]*[COVID stage = Lockdown 2] 0b
[Cycling infrastructure density = Second quartile]*[COVID stage = Pre COVID] −0.50 0.53 −0.95 0.34 [Residential density = Third quartile]*[COVID stage = Lockdown 2] 0b
[Cycling infrastructure density = Third quartile]*[COVID stage = Pre COVID] −0.32 0.42 −0.75 0.45 [Residential density = Highest quartile]*[COVID stage = Lockdown 2] 0b
[Cycling infrastructure density = Highest quartile]*[COVID stage = Pre COVID] 0b NegativeBinomial 4.31 0.28
Model 4 (Demographics and time interaction) coefficients
Unlabelled TableModel Term Coefficient Std. Error t Sig. Model Term Coefficient Std. Error t Sig.
Intercept −0.09 0.48 −0.19 0.85 Income = Medium 0b
COVID stage = Pre COVID 0.40 0.54 0.74 0.46 [Gender = Female]*[COVID stage = Pre COVID] 0.07 0.23 0.31 0.76
COVID stage = Lockdown 1 0.36 0.55 0.66 0.51 [Gender = Male]*[COVID stage = Pre COVID] 0b
COVID stage = Lockdown 2 0b [Gender = Female]*[COVID stage = Lockdown 1] −0.05 0.23 −0.23 0.82
Connectivity = Lowest quartile −0.45 0.27 −1.68 0.09 [Gender = Male]*[COVID stage = Lockdown 1] 0b
Connectivity = Second quartile −0.39 0.23 −1.66 0.10 [Gender = Female]*[COVID stage = Lockdown 2] 0b
Connectivity = Third quartile −0.07 0.20 −0.35 0.72 [Gender = Male]*[COVID stage = Lockdown 2] 0b
Connectivity = Highest quartile 0b [Age = adult]*[COVID stage = Pre COVID] 0.08 0.36 0.21 0.84
Cycling infrastructure density = Lowest quartile 0.65 0.23 2.86 0.00 [Age = older adult]*[COVID stage = Pre COVID] 0.25 0.49 0.50 0.62
Cycling infrastructure density = Second quartile 0.36 0.22 1.65 0.10 [Age = youth]*[COVID stage = Pre COVID] 0b
Cycling infrastructure density = Third quartile 0.50 0.17 2.91 0.00 [Age = adult]*[COVID stage = Lockdown 1] 0.02 0.37 0.06 0.95
Cycling infrastructure density = Highest quartile 0b [Age = older adult]*[COVID stage = Lockdown 1] 0.15 0.51 0.29 0.77
Green space percentage = Lowest quartile −0.08 0.16 −0.46 0.64 [Age = youth]*[COVID stage = Lockdown 1] 0b
Green space percentage = Second quartile −0.33 0.16 −2.02 0.04 [Age = adult]*[COVID stage = Lockdown 2] 0b
Green space percentage = Third quartile −0.17 0.13 −1.30 0.19 [Age = older adult]*[COVID stage = Lockdown 2] 0b
Green space percentage = Highest quartile 0b [Age = youth]*[COVID stage = Lockdown 2] 0b
Residential area percentage = Lowest quartile −0.21 0.22 −0.99 0.32 [Education = Higher education]*[COVID stage = Pre COVID] 0.17 0.24 0.71 0.48
Residential area percentage = Second quartile 0.03 0.23 0.14 0.89 [Education = Non university degree]*[COVID stage = Pre COVID] 0b
Residential area percentage = Third quartile 0.06 0.20 0.29 0.78 [Education = Higher education]*[COVID stage = Lockdown 1] −0.02 0.24 −0.06 0.95
Residential area percentage = Highest quartile 0b [Education = Non university degree]*[COVID stage = Lockdown 1] 0b
Land-use diversity = Lowest quartile −0.28 0.20 −1.36 0.17 [Education = Higher education]*[COVID stage = Lockdown 2] 0b
Land-use diversity = Second quartile −0.17 0.18 −0.98 0.33 [Education = Non university degree]*[COVID stage = Lockdown 2] 0b
Land-use diversity = Third quartile −0.29 0.14 −2.11 0.04 [Employment = Employed]*[COVID stage = Pre COVID] −0.20 0.45 −0.45 0.65
Land-use diversity = Highest quartile 0b [Employment = Non worker]*[COVID stage = Pre COVID] −0.04 0.47 −0.09 0.93
Residential density = Lowest quartile −0.23 0.32 −0.73 0.47 [Employment = Not in labor force]*[COVID stage = Pre COVID] 0b
Residential density = Second quartile −0.23 0.25 −0.94 0.35 [Employment = Employed]*[COVID stage = Lockdown 1] −0.08 0.46 −0.17 0.87
Residential density = Third quartile 0.16 0.21 0.76 0.45 [Employment = Non worker]*[COVID stage = Lockdown 1] −0.12 0.48 −0.26 0.80
Residential density = Highest quartile 0b [Employment = Not in labor force]*[COVID stage = Lockdown 1] 0b
Gender = Female −0.88 0.17 −5.29 0.00 [Employment = Employed]*[COVID stage = Lockdown 2] 0b
Gender = Male 0b [Employment = Non worker]*[COVID stage = Lockdown 2] 0b
Age = adult −0.27 0.27 −1.02 0.31 [Employment = Not in labor force]*[COVID stage = Lockdown 2] 0b
Age = older adult −1.05 0.36 −2.86 0.00 [Income = High]*[COVID stage = Pre COVID] −0.26 0.31 −0.86 0.39
Age = youth 0b [Income = Low]*[COVID stage = Pre COVID] −0.17 0.28 −0.60 0.55
Education = Higher education 0.09 0.18 0.50 0.62 [Income = Medium]*[COVID stage = Pre COVID] 0b
Education = Non university degree 0b [Income = High]*[COVID stage = Lockdown 1] −0.26 0.31 −0.84 0.40
Employment = Employed 0.18 0.34 0.54 0.59 [Income = Low]*[COVID stage = Lockdown 1] −0.14 0.29 −0.48 0.63
Employment = Non worker 0.04 0.35 0.10 0.92 [Income = Medium]*[COVID stage = Lockdown 1] 0b
Employment = Not in labor force 0b [Income = High]*[COVID stage = Lockdown 2] 0b
Income = High −0.13 0.22 −0.57 0.57 [Income = Low]*[COVID stage = Lockdown 2] 0b
Income = Low 0.08 0.21 0.38 0.70 [Income = Medium]*[COVID stage = Lockdown 2] 0b
NegativeBinomial 4.36 0.28
Model 5 (BE and gender interaction) coefficients
Unlabelled TableModel Term Coefficient Std. Error t Sig. Model Term Coefficient Std. Error t Sig.
Intercept −0.71 0.43 −1.66 0.10 [Gender = Female]*[Connectivity = Third quartile] −0.15 0.41 −0.36 0.72
COVID stage = Pre COVID 0.41 0.11 3.88 0.00 [Gender = Male]*[Connectivity = Third quartile] 0b
COVID stage = Lock down 1 0.19 0.11 1.74 0.08 [Gender = Female]*[Connectivity = Highest quartile] 0b
COVID stage = Lockdown 2 0b [Gender = Male]*[Connectivity = Highest quartile] 0b
Connectivity = Lowest quartile −0.95 0.42 −2.27 0.02 [Gender = Female]*[Cycling infrastructure density = Lowest quartile] −0.63 0.46 −1.38 0.17
Connectivity = Second quartile −0.75 0.36 −2.11 0.04 [Gender = Male]*[Cycling infrastructure density = Lowest quartile] 0b
Connectivity = Third quartile −0.03 0.31 −0.11 0.91 [Gender = Female]*[Cycling infrastructure density = Second quartile] −0.42 0.45 −0.93 0.35
Connectivity = Highest quartile 0b [Gender = Male]*[Cycling infrastructure density = Second quartile] 0b
Cycling infrastructure density = Lowest quartile 1.06 0.35 3.05 0.00 [Gender = Female]*[Cycling infrastructure density = Third quartile] −0.32 0.35 −0.92 0.36
Cycling infrastructure density = Second quartile 0.64 0.35 1.86 0.06 [Gender = Male]*[Cycling infrastructure density = Third quartile] 0b
Cycling infrastructure density = Third quartile 0.71 0.25 2.81 0.01 [Gender = Female]*[Cycling infrastructure density = Highest quartile] 0b
Cycling infrastructure density = Highest quartile 0b [Gender = Male]*[Cycling infrastructure density = Highest quartile] 0b
Green space percentage = Lowest quartile −0.37 0.25 −1.52 0.13 [Gender = Female]*[Green space percentage = Lowest quartile] 0.52 0.33 1.61 0.11
Green space percentage = Second quartile −0.28 0.25 −1.10 0.27 [Gender = Male]*[Green space percentage = Lowest quartile] 0b
Green space percentage = Third quartile −0.47 0.20 −2.41 0.02 [Gender = Female]*[Green space percentage = Second quartile] −0.01 0.33 −0.04 0.97
Green space percentage = Highest quartile 0b [Gender = Male]*[Green space percentage = Second quartile] 0b
Residential area percentage = Lowest quartile 0.40 0.33 1.23 0.22 [Gender = Female]*[Green space percentage = Third quartile] 0.52 0.26 2.01 0.04
Residential area percentage = Second quartile 0.84 0.34 2.49 0.01 [Gender = Male]*[Green space percentage = Third quartile] 0b
Residential area percentage = Third quartile 0.66 0.31 2.17 0.03 [Gender = Female]*[Green space percentage = Highest quartile] 0b
Residential area percentage = Highest quartile 0b [Gender = Male]*[Green space percentage = Highest quartile] 0b
Land-use diversity = Lowest quartile −0.16 0.30 −0.54 0.59 [Gender = Female]*[Residential area percentage = Lowest quartile] −0.92 0.43 −2.16 0.03
Land-use diversity = Second quartile −0.09 0.27 −0.35 0.72 [Gender = Male]*[Residential area percentage = Lowest quartile] 0b
Land-use diversity = Third quartile −0.31 0.20 −1.56 0.12 [Gender = Female]*[Residential area percentage = Second quartile] −1.34 0.45 −2.99 0.00
Land-use diversity = Highest quartile 0b [Gender = Male]*[Residential area percentage = Second quartile] 0b
Residential density = Lowest quartile 0.02 0.48 0.03 0.97 [Gender = Female]*[Residential area percentage = Third quartile] −0.93 0.39 −2.36 0.02
Residential density = Second quartile −0.14 0.36 −0.38 0.71 [Gender = Male]*[Residential area percentage = Third quartile] 0b
Residential density = Third quartile 0.32 0.33 0.99 0.32 [Gender = Female]*[Residential area percentage = Highest quartile] 0b
Residential density = Highest quartile 0b [Gender = Male]*[Residential area percentage = Highest quartile] 0b
Gender = Female 0.06 0.50 0.12 0.90 [Gender = Female]*[Land-use diversity = Lowest quartile] −0.24 0.40 −0.59 0.55
Gender = Male 0b [Gender = Male]*[Land-use diversity = Lowest quartile] 0b
Age = adult −0.22 0.15 −1.47 0.14 [Gender = Female]*[Land-use diversity = Second quartile] −0.19 0.35 −0.53 0.59
Age = older adult −0.93 0.20 −4.61 0.00 [Gender = Male]*[Land-use diversity = Second quartile] 0b
Age = youth 0b [Gender = Female]*[Land-use diversity = Third quartile] 0.06 0.27 0.21 0.84
Education = Higher education 0.20 0.10 1.97 0.05 [Gender = Male]*[Land-use diversity = Third quartile] 0b
Education = Non university degree 0b [Gender = Female]*[Land-use diversity = Highest quartile] 0b
Employment = Employed 0.09 0.18 0.51 0.61 [Gender = Male]*[Land-use diversity = Highest quartile] 0b
Employment = Non worker 0.01 0.19 0.03 0.98 [Gender = Female]*[Residential density = Lowest quartile] −0.36 0.64 −0.55 0.58
Employment = Not in labor force 0b [Gender = Male]*[Residential density = Lowest quartile] 0b
Income = High −0.30 0.13 −2.32 0.02 [Gender = Female]*[Residential density = Second quartile] −0.14 0.49 −0.30 0.77
Income = Low 0.00 0.12 −0.02 0.99 [Gender = Male]*[Residential density = Second quartile] 0b
Income = Medium 0b [Gender = Female]*[Residential density = Third quartile] −0.21 0.42 −0.49 0.62
[Gender = Female]*[Connectivity = Lowest quartile] 0.72 0.55 1.32 0.19 [Gender = Male]*[Residential density = Third quartile] 0b
[Gender = Male]*[Connectivity = Lowest quartile] 0b [Gender = Female]*[Residential density = Highest quartile] 0b
[Gender = Female]*[Connectivity = Second quartile] 0.54 0.47 1.14 0.25 [Gender = Male]*[Residential density = Highest quartile] 0b
[Gender = Male]*[Connectivity = Second quartile] 0b NegativeBinomial 4.27 0.27
Model 6 (Gender and residential percentage interaction) coefficients
Unlabelled TableModel Term Coefficient Std. Error t Sig. Model Term Coefficient Std. Error t Sig.
Intercept −0.32 0.36 −0.89 0.38 Green space percentage = Lowest quartile −0.06 0.16 −0.40 0.69
COVID stage = Pre COVID 0.40 0.11 3.70 0.00 Green space percentage = Second quartile −0.29 0.16 −1.82 0.07
COVID stage = Lockdown 1 0.18 0.11 1.65 0.10 Green space percentage = Third quartile −0.17 0.13 −1.32 0.19
COVID stage = Lockdown 2 0b Green space percentage = Highest quartile 0b
Gender = Female −0.39 0.19 −2.05 0.04 Residentail area percentage = Lowest quartile 0.10 0.26 0.37 0.71
Gender = Male 0b Residentail area percentage = Second quartile 0.52 0.27 1.93 0.05
Age = adult −0.24 0.15 −1.58 0.11 Residentail area percentage = Third quartile 0.43 0.25 1.73 0.08
Age = older adult −0.90 0.20 −4.42 0.00 Residentail area percentage = Highest quartile 0b
Age = youth 0b Land-use diversity = Lowest quartile −0.28 0.20 −1.38 0.17
Education = Higher education 0.13 0.10 1.32 0.19 Land-use diversity = Second quartile −0.17 0.17 −1.00 0.32
Education = Non university degree 0b Land-use diversity = Third quartile −0.30 0.14 −2.19 0.03
Employment = Employed 0.06 0.18 0.34 0.74 Land-use diversity = Highest quartile 0b
Employment = Non worker −0.06 0.19 −0.30 0.76 Residential density = Lowest quartile −0.17 0.32 −0.53 0.60
Employment = Not in labor force 0b Residential density = Second quartile −0.18 0.24 −0.74 0.46
Income = High −0.28 0.13 −2.19 0.03 Residential density = Third quartile 0.20 0.21 0.92 0.36
Income = Low −0.05 0.12 −0.39 0.69 Residential density = Highest quartile 0b
Income = Medium 0b [Gender = Female]*[Residentail area percentage = Lowest quartile] −0.47 0.26 −1.81 0.07
Connectivity = Lowest quartile −0.53 0.27 −1.98 0.05 [Gender = Female]*[Residentail area percentage = Second quartile] −0.81 0.25 −3.16 0.00
Connectivity = Second quartile −0.43 0.23 −1.87 0.06 [Gender = Female]*[Residentail area percentage = Third quartile] −0.57 0.26 −2.22 0.03
Connectivity = Third quartile −0.11 0.20 −0.55 0.58 [Gender = Female]*[Residentail area percentage = Highest quartile] 0b
Connectivity = Highest quartile 0b [Gender = Male]*[Residentail area percentage = Lowest quartile] 0b
Cycling infrastructure density = Lowest quartile 0.66 0.22 2.95 0.00 [Gender = Male]*[Residentail area percentage = Second quartile] 0b
Cycling infrastructure density = Second quartile 0.35 0.22 1.60 0.11 [Gender = Male]*[Residentail area percentage = Third quartile] 0b
Cycling infrastructure density = Third quartile 0.50 0.17 2.90 0.00 [Gender = Male]*[Residentail area percentage = Highest quartile] 0b
Cycling infrastructure density = Highest quartile 0b NegativeBinomial 4.34 0.27
Data availability
The authors do not have permission to share data.
Acknowledgements
The authors would like to thank Graham Currie and his team for providing access to the ‘C-19 Long Term Transport Impact Study’ and thank Taru Jain for her guidance to decode the data.
==== Refs
Reference
Aldred R. Woodcock J. Goodman A. Does more cycling mean more diversity in cycling? Transp. Rev. 36 2016 28 44
Aldred R. Elliott B. Woodcock J. Goodman A. Cycling provision separated from motor traffic: a systematic review exploring whether stated preferences vary by gender and age Transp. Rev. 37 2017 29 55 28190905
Barajas J.M. Perceptions, people, and places: influences on cycling for latino immigrants and implications for equity J. Plan. Educ. Res. 0739456X19864714 2019
Barajas J.M. Supplemental infrastructure: how community networks and immigrant identity influence cycling Transportation 47 2020 1251 1274
Barbieri D.M. Lou B. Passavanti M. Hui C. Lessa D.A. Maharaj B. Banerjee A. Wang F. Chang K. Naik B. A survey dataset to evaluate the changes in mobility and transportation due to Covid-19 travel restrictions in Australia, Brazil, China, Ghana, India, Iran, Italy, Norway, South Africa, United States Data Brief 33 2020 106459
Beenackers M.A. Foster S. Kamphuis C.B. Titze S. Divitini M. Knuiman M. Van Lenthe F.J. Giles-Corti B. Taking up cycling after residential relocation: built environment factors Am. J. Prev. Med. 42 2012 610 615 22608378
Bentley R. Blakely T. Kavanagh A. Aitken Z. King T. Mcelwee P. Giles-Corti B. Turrell G. A longitudinal study examining changes in street connectivity, land use, and density of dwellings and walking for transport in Brisbane, Australia Environ. Health Perspect. 126 2018 057003
Biddle S.J. Ciaccioni S. Thomas G. Vergeer I. Physical activity and mental health in children and adolescents: An updated review of reviews and an analysis of causality Psychol. Sport Exerc. 42 2019 146 155
Bourke M. Craike M. Hilland T.A. Moderating effect of gender on the associations of perceived attributes of the neighbourhood environment and social norms on transport cycling behaviours J. Transp. Health 13 2019 63 71
Braun L.M. Rodriguez D.A. Gordon-Larsen P. Social (in) equity in access to cycling infrastructure: Cross-sectional associations between bike lanes and area-level sociodemographic characteristics in 22 large Us cities J. Transp. Geogr. 80 2019 102544
Bucsky P. Modal share changes due to Covid-19: The case of Budapest Transp. Res. Interdiscip. Perspect. 8 2020 100141
Budi D.R. Widyaningsih R. Nur L. Agustan B. Dwi D. Qohhar W. Asnaldi A. Cycling during Covid-19 pandemic: sports or lifestyle? Int. J. Human Move. Sports Sci. 9 2021 765 771
Buehler R. Dill J. Bikeway networks: A review of effects on cycling Transp. Rev. 36 2016 9 27
Buehler R. Pucher J. COVID-19 Impacts on Cycling, 2019–2020 2021 Taylor & Francis
Cerin E. Nathan A. Van Cauwenberg J. Barnett D.W. Barnett A. The neighbourhood physical environment and active travel in older adults: a systematic review and meta-analysis Int. J. Behav. Nutr. Phys. Act. 14 2017 1 23 28057008
Cervero R. Duncan M. Walking, bicycling, and urban landscapes: evidence from the San Francisco Bay Area Am. J. Public Health 93 2003 1478 1483 12948966
Cervero R. Kockelman K. Travel demand and the 3Ds: density, diversity, and design Transp. Res. Part D: Transp. Environ. 2 1997 199 219
Cervero R. Denman S. Jin Y. Network design, built and natural environments, and bicycle commuting: Evidence from British cities and towns Transp. Policy 74 2019 153 164
Cheng L. Chen X. Yang S. Cao Z. De Vos J. Witlox F. Active travel for active ageing in China: the role of built environment J. Transp. Geogr. 76 2019 142 152
Combs T.S. Pardo C.F. Shifting streets Covid-19 mobility data: findings from a global dataset and a research agenda for transport planning and policy Transp. Res. Interdiscip. Perspect. 9 2021 100322
Currie G. Jain T. Aston L. Reynolds J. Long Term Impacts of COVID-19 on Melbourne Research 2021
De Vos J. The effect of COVID-19 and subsequent social distancing on travel behavior Transp. Res. Interdiscip. Perspect. 100121 2020
Dunning R. Nurse A. The surprising availability of cycling and walking infrastructure through COVID-19 Town Plan. Rev. 2020 1 7
Ewing R. Cervero R. Travel and the built environment: a meta-analysis J. Am. Plan. Assoc. 76 2010 265 294
Fischer J. Nelson T. Winters M. Riding through the pandemic: using Strava data to monitor the impacts of Covid-19 on spatial patterns of bicycling Transp. Res. Interdiscip. Perspect. 15 2022 100667
Flanagan E. Lachapelle U. El-Geneidy A. Riding tandem: does cycling infrastructure investment mirror gentrification and privilege in Portland, OR and Chicago, IL? Res. Transp. Econ. 60 2016 14 24
Forsyth A. Oakes J.M. Schmitz K.H. Hearst M. Does residential density increase walking and other physical activity? Urban Stud. 44 2007 679 697
Fuller G. Mcguinness K. Waitt G. Buchanan I. Lea T. The reactivated bike: Self-reported cycling activity during the 2020 Covid-19 pandemic in Australia Transp. Res. Interdiscip. Persp. 10 2021 100377
Gladwin K. Duncan M. Covid-19’s impact on older adults’ cycling behaviors in a small, auto-centric urban area Transp. Res. Interdiscip. Perspect. 100675 2022
Grudgings N. Hughes S. Hagen-Zanker A. The comparison and interaction of age and gender effects on cycling mode-share: an analysis of commuting in England and Wales J. Transp. Health 20 2021 101004
Heesch K.C. Sahlqvist S. Garrard J. Gender differences in recreational and transport cycling: a cross-sectional mixed-methods comparison of cycling patterns, motivators, and constraints Int. J. Behav. Nutr. Phys. Act. 9 2012 1 12 22233712
Hino A.A. Reis R.S. Sarmiento O.L. Parra D.C. Brownson R.C. Built environment and physical activity for transportation in adults from Curitiba, Brazil J. Urban Health 91 2014 446 462 24096625
Hong J. Mcarthur D. Raturi V. Did safe cycling infrastructure still matter during a covid-19 lockdown? Sustainability 12 2020 8672
Hull A. O’holleran C. Bicycle infrastructure: can good design encourage cycling? Urban Plan. Transp. Res. 2 2014 369 406
Jia P. Zou Y. Wu Z. Zhang D. Wu T. Smith M. Xiao Q. Street connectivity, physical activity, and childhood obesity: a systematic review and meta-analysis Obes. Rev. 22 Suppl 1(Suppl 1) 2019 e12943 10.1111/obr.12943 Epub 2019 Sep 10. PMID: 31507068; PMCID: PMC7988624 31507068
Kamruzzaman M. Hine J. Self-proxy agreement and weekly school travel behaviour in a sectarian divided society J. Transp. Geogr. 29 2013 74 85
Kraus S. Koch N. Provisional COVID-19 infrastructure induces large, rapid increases in cycling Proc. Natl. Acad. Sci. 118 2021
Mäki-Opas T.E. Borodulin K. Valkeinen H. Stenholm S. Kunst A.E. Abel T. Härkänen T. Kopperoinen L. Itkonen P. Prättälä R. The contribution of travel-related urban zones, cycling and pedestrian networks and green space to commuting physical activity among adults–a cross-sectional population-based study using geographical information systems BMC Public Health 16 2016 1 14 26728978
Mölenberg F.J. Panter J. Burdorf A. Van Lenthe F.J. A systematic review of the effect of infrastructural interventions to promote cycling: strengthening causal inference from observational data Int. J. Behav. Nutr. Phys. Act. 16 2019 1 31 30606197
Pasha M. Rifaat S. Tay R. De Barros A. Urban design and planning influences on the share of trips taken by cycling J. Urban Des. 21 2016 471 480
Patterson R. Ogilvie D. Panter J. Travel Levels Before and After COVID-19 Control Measures in Cambridge, UK. Findings, June 2021
Pazdan S. The impact of weather on bicycle risk exposure Arch. Transp. 56 2020
Pistoll C. Goodman A. The link between socioeconomic position, access to cycling infrastructure and cycling participation rates: An ecological study in Melbourne, Australia J. Transp. Health 1 2014 251 259
Prati G. Fraboni F. De Angelis M. Pietrantoni L. Johnson D. Shires J. Gender differences in cycling patterns and attitudes towards cycling in a sample of European regular cyclists J. Transp. Geogr. 78 2019 1 7
Saelens B.E. Sallis J.F. Frank L.D. Environmental correlates of walking and cycling: findings from the transportation, urban design, and planning literatures Ann. Behav. Med. 25 2003 80 91 12704009
Sallis J.F. Conway T.L. Cain K.L. Carlson J.A. Frank L.D. Kerr J. Glanz K. Chapman J.E. Saelens B.E. Neighborhood built environment and socioeconomic status in relation to physical activity, sedentary behavior, and weight status of adolescents Prev. Med. 110 2018 47 54 29432790
Schweizer A.-M. Leiderer A. Mitterwallner V. Walentowitz A. Mathes G.H. Steinbauer M.J. Outdoor cycling activity affected by COVID-19 related epidemic-control-decisions PLoS One 16 2021 e0249268
Shaer A. Rezaei M. Rahimi B.M. Shaer F. Examining the associations Between perceived built environment and active travel, before and after the COVID-19 outbreak in Shiraz city, Iran Cities 115 2021 103255
Shamshiripour A. Rahimi E. Shabanpour R. Mohammadian A.K. How is COVID-19 reshaping activity-travel behavior? Evidence from a comprehensive survey in Chicago Transp. Res. Interdiscip. Perspect. 7 2020 100216
Sun Y. Du Y. Wang Y. Zhuang L. Examining associations of environmental characteristics with recreational cycling behaviour by street-level Strava data Int. J. Environ. Res. Public Health 14 2017 644 28617345
Tortosa E.V. Lovelace R. Heinen E. Mann R.P. Cycling behaviour and socioeconomic disadvantage: An investigation based on the English National Travel Survey Transp. Res. A Policy Pract. 152 2021 173 185
United Nations General Assembly Transforming our World: The 2030 Agenda for Sustainable Development 2015
Venter Z.S. Barton D.N. Gundersen V. Figari H. Nowell M. Urban nature in a time of crisis: recreational use of green space increases during the COVID-19 outbreak in Oslo, Norway Environ. Res. Lett. 15 2020 104075
Warburton D.E. Bredin S.S. Reflections on physical activity and health: what should we recommend? Can. J. Cardiol. 32 2016 495 504 26995692
Watts P. Phillips G. Petticrew M. Hayes R. Bottomley C. Yu G. Schmidt E. Tobi P. Moore D. Frostick C. Physical activity in deprived communities in London: examining individual and neighbourhood-level factors PLoS One 8 2013 e69472
Winters M. Fischer J. Nelson T. Fuller D. Whitehurst D.G. Equity in spatial access to bicycling infrastructure in mid-sized Canadian cities Transp. Res. Rec. 2672 2018 24 32
Wittmann K. Savan B. Ledsham T. Liu G. Lay J. Cycling to high school in Toronto, Ontario, Canada: Exploration of school travel patterns and attitudes by gender Transp. Res. Rec. 2500 2015 9 16
Xing Y. Handy S.L. Mokhtarian P.L. Factors associated with proportions and miles of bicycling for transportation and recreation in six small US cities Transp. Res. Part D: Transp. Environ. 15 2010 73 81
Yang Y. Wu X. Zhou P. Gou Z. Lu Y. Towards a cycling-friendly city: An updated review of the associations between built environment and cycling behaviors (2007–2017) J. Transp. Health 14 2019 100613
Zander A. Passmore E. Mason C. Rissel C. Joy, exercise, enjoyment, getting out: a qualitative study of older people's experience of cycling in Sydney, Australia J. Environ. Public Health 2013 2013
| 0 | PMC9742218 | NO-CC CODE | 2022-12-14 23:53:26 | no | J Transp Geogr. 2023 Jan 12; 106:103510 | utf-8 | J Transp Geogr | 2,022 | 10.1016/j.jtrangeo.2022.103510 | oa_other |
==== Front
Int Dent J
Int Dent J
International Dental Journal
0020-6539
1875-595X
Published by Elsevier Inc. on behalf of FDI World Dental Federation.
S0020-6539(22)00280-5
10.1016/j.identj.2022.12.003
Scientific Research Report
Socio-behavioural factors associated with child oral health during COVID-19
Gudipaneni Ravi Kumar 1⁎
Alruwaili Mohammed Farhan O 2
Ganji Kiran Kumar 3
Karobari Mohmed Isaqali 4
Kulkarni Sachin 56#
Metta Kiran Kumar 7$
Assiry Ali A 8
Israelsson Nicholas 9^
Bawazir Omar A 10
1 Department of Preventive Dentistry, Pediatric dentistry Division, College of Dentistry, Jouf University, Sakaka, Al Jouf, Saudi Arabia
2 College of Dentistry, Jouf University, Al Jouf, Sakaka, Saudi Arabia
3 Department of Preventive Dentistry, Periodontics Division, College of Dentistry, Jouf University, Al Jouf, Sakaka, Saudi Arabia
4 Center for Transdisciplinary Research (CFTR), Saveetha Dental College & Hospitals, Saveetha Institute of Medical and Technical Sciences University, Chennai 600077, Tamil Nadu, India
5 School of Dentistry and Oral Health, University of Adelaide, Adelaide, SA
6 Griffith University, Australia
7 Department of Conservative Dental Sciences, Ibn Sina National College For Medical Studies, Jeddah, Saudi Arabia
8 Preventive Dental Science Department, Faculty of Dentistry, Najran University, Najran 55461, Kingdom of Saudi Arabia
9 SA dental Service, Adelaide, South Australia
10 Department of Pediatric Dentistry and Orthodontics, College of Dentistry, King Saud University, Riyadh, Saudi Arabia
⁎ Corresponding author: Ravi Kumar Gudipaneni, Assistant Professor, Department of Preventive Dentistry, Pediatric dentistry Division, College of Dentistry, Jouf University, Sakaka, Al Jouf, Saudi Arabia. Mobile: +966540684272
# Mobile: +61430218069
$ Mobile: +966532569519
^ Mobile: +61 430218069
12 12 2022
12 12 2022
© 2022 Published by Elsevier Inc. on behalf of FDI World Dental Federation.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objectives
To identify the socio behavioural factors that influenced children's oral health during the coronavirus outbreak.
Methods
The online cross-sectional study was conducted in Al Jouf Province, the northern region of Saudi Arabia. A total of 960 parents of children aged 5 to 14 years were invited by multistage stratified random sampling. Descriptive, multinomial and multiple logistic regression analyses were performed to estimate odds ratios and determine the relationship between independent and dependent variables. P < 0.05 was considered statistically significant.
Results
Of the 960 participants, 693 (72.1%) reported that their child had one or more untreated dental decay. The children of uneducated parents were 1.6-fold more likely to have one or more untreated dental decay (adjusted odds ratio [AOR]: 1.66, 95% CI 0.74–3.73; P < 0.001). The children of unemployed parents were 4.3-fold more likely to have a financial burden for child dental visit (AOR: 4.34, 95% CI 2.73–6.89; P < 0.001). The parents from the rural area were 26.3-fold more likely to have more than 2 years since their last dental visit (AOR: 26.34, 95% CI 7.48–92.79; P < 0.001). Nursery-level children were 5.4-fold more likely to need immediate care (AOR: 5.38, 95% CI 3.01–9.60; P < 0.001).
Conclusion
The present study demonstrated a very high prevalence of one or more untreated dental decay. Children of rural areas, uneducated, unemployed, widow/divorced, low and middle-income parents and nursery school children were linked with poorly predicted outcomes of child oral health during the pandemic.
Keywords
COVID-19 outbreak
self-reported child oral health
self-perceived need of dental care
dental visiting pattern
oral health care affordability
==== Body
pmcIntroduction
The coronavirus disease 2019 (COVID-19) outbreak affects healthcare workers and the general population. Its economic and mental health impact continues to unfold, which has been acknowledged.1 In addition, the COVID-19 crisis has heightened fear and ambiguity, placing a burden on the oral health care system and resources.2 It poses significant concerns to children's oral health which can affect adults and children differently.3 , 4 The ‘World Health Organization Emergency Committee’ declared a worldwide health emergency on January 30, 2020, and COVID-19 was considered as a pandemic in March 2020, putting the whole world on lockdown, thereby limiting access to dental emergencies.5 Visiting dental offices has become risky for children and adults, and providing oral health care during the COVID-19 pandemic placed dental practitioners at a higher risk of contracting the virus.6 , 7
Unprecedented actions to combat the spread of coronavirus in Saudi Arabia (KSA) have been considered, such as the suspension of communal transportation, closure of public areas and isolation of infected and suspicious cases. Furthermore, dental services in KSA have been halted from March to August 2020. Consequently, access to dental services for children throughout the nation has been temporarily shut down and limited to only emergency dental procedures across the country.8 Although dental clinic closure is a necessary precaution, it may have a negative impact on child oral health from low-income and disadvantaged communities.9
Prior to the pandemic, Saudi children were reported to have poor oral health and a heavy burden of dental caries, with high levels of untreated caries (80%)10 and only 18.9% of children attended routine dental visits.11 During the pandemic, children did not have access to routine, and follow-up dental visits had a significant impact on child oral health. Routine, non-urgent follow-up visits are required for child's optimal oral health, which enable early diagnosis and administration of preventative oral health care services.12 , 13 , 14 Studies have shown that the level of maternal education, income and preventative mother–child counselling are the primary factors influencing how often should a child attend routine dental check-ups.15 Another study has reported that child routine dental visits are dependent on their insurance coverage, affordability, child's age, parental attitude and access to dental treatments.16 Although the situation is constantly changing, most of the dental facilities have resumed, providing routine or non-urgent dental care with limited understanding of the impact of COVID-19 on child oral health.
The effects of COVID-19 exposed and aggravated the depth and breadth of social and economic inequality. Hence, the present study designed to identify the socio behavioural factors that influence the oral health of children aged 5 to 14, measured by using a self-reported questionnaire reported by parents during the COVID-19 outbreak. Self-reported questionnaires can obtain data economically and safely without the risk of contracting the virus during the pandemic. Furthermore, this type of questionnaires could be used to collect information for obtaining a quick overview of health care during the pandemic.17 However, literature on the socio-behavioural factors influencing child oral health during the pandemic is limited. Thus, evaluating socio-behavioural factors associated with children's oral health is important because such factors can be used in formulating policies for improving child oral health, promoting oral health care and providing a snapshot view to the regulatory authorities for decision-making and resource allocation during the pandemic. Therefore, this study aimed to determine the socio-behavioural variables that affect child oral health by investigating the self-reported oral health of children (presence of one or more untreated dental caries), self-perceived need of care based on parental perspective or experience, pattern of child dental visits and affordability in accessing dental care amongst a sample obtained from Al Jouf Province, northern region of KSA.
Methods
Study design and setting
The present cross-sectional study was conducted in the province of Al Jouf, in the northern part of KSA. The parents/guardians of school children aged 5–14 years were invited to participate in this study on behalf of their children. The Local Committee of Bioethics, Jouf University, KSA, granted the ethical permission for this study (04-02-42). The data were collected using google forms from January to June 2021. The study was conducted following STROBE guidelines.
Conceptual framework
The present study was designed on the basis of Andersen's behavioural model to measure the outcome variables.18 The three primary components of Andersen's model were ‘predisposing’, ‘enabling’ and ‘need factors’.19 The predisposing factors included ‘demographic and social characteristics’. Age, gender, race/ethnicity and social characteristics such as participants’ places of residence and employment, educational attainment and marital status were considered as demographic and social characteristics in this study. The enabling factors used in this study included financial considerations that affect individual's ability to utilise health care services.19 , 20 With regard to need factors, included perceived need for oral health care services amongst children based on parents’ view and experience.19 , 20
Sample size calculation
In the present study, sample size calculation was estimated with open Epi software.17 Considering that the population size was limited (1,000,000), the expected response rate would be 50%; the margin of error was 5%, and the design effect was 2. The minimum estimated sample size was 768 participants. A total of 960 parents or guardians of schoolchildren were included in the final sample for data analysis to reflect the Al Jouf population.
Study population and sampling technique
The parents of school children were invited by multistage stratified random sampling method. Parents of school children aged 5 to 14 years were invited to be a part of this study through school social media. The Al Jouf province has four districts: Sakaka, Qurayyat, Duma Al Jandal and Tabarjal. Each district has been divided into two sections: urban and rural. In the first stage, the districts were selected as the first strata, and under each stratum (district), one urban and one rural area was chosen randomly using a simple random sampling technique. The list of registered schools from each district's selected urban and rural areas was obtained and numbered sequentially in the second stage. From the list of schools, one male and one female school were chosen by a simple random sampling method using a random number generated by a computer. In the third stage, participants were chosen randomly from a list of students from each class. The schools that refused to participate in the study were replaced by another school from the same stratum. The inclusion criterion of this study was Saudi parents/caretakers of school children aged 5-14-years. Parents who were not interested in participating in this study and failed to provide written informed consent were excluded from this study.
Data collection
Data were collected by using a self-reported online questionnaire distributed to the parents through school's official social media. The closed questionnaire items used in this study were written initially in English and then translated into Arabic by a native Arabic bilingual translator. The test–retest reliability (Cronbach's alpha) of the questionnaire was assessed using 10% of the responses, which was found to be 0.86. Parents/guardians were asked to fill out a closed-ended questionnaire about three basic sections. The first section includes demographic details of participants that include; gender, nationality, age, parental education, marital status, occupation, and current residential location. The second section assessed socioeconomic status (SES) of families based on the family income; if income per month less than 3,000 SAR (low economic status); between 3,000 and 10,000 SAR (middle economic status), and more than 10,000 SAR (high economic status).18 The third section focused on oral health behavioural characteristics that include; self-reported child oral health, self-perceived need of oral care, pattern of dental visits, and oral health care affordability. The questionnaire items of behavioural characteristics were illustrated in Table 1 .Table 1 Demographic, sociobehavioural characteristics of study participants
Table 1Characteristics Variables Number of participants N (%)
Gender of the parent Male 454 (47.3%)
Female 506 (52.7%)
Parents age groups 25-30 - years 148 (15.4%)
30-35- years 211 (22.0%)
35-40- years 227 (23.6%)
> 40 years 374 (38.9%)
Parental educational level Uneducated 76 (7.9%)
Primary school 103 (10.7%)
Secondary school 232 (24.1%)
Graduation and above 549 (57.1%)
Parental marital status Married 836 (87.0%)
Widowed/ Divorced/Separated 124 (12.9%)
Parental occupation Unemployed 171 (17.8%)
Government employee 592 (61.6%)
Private employee 81 (8.4%)
Self-employed/ own business 116 (12.1%)
Place of residence Rural 663 (69.0%)
Urban 297 (30.9%)
Level of child study Nursery School 580 (60.4%)
Primary school 380 (39.5%)
Socio economic status Low (≤ 3,000 Saudi Riyals) 307 (31.9%)
Middle (> 3,000 to < 10,000 Saudi Riyals) 322 (33.5%)
High (≥ 10,000 Saudi Riyals) 331 (34.4%)
Self-reported child oral health Presence of one or more untreated decayed teeth 693 (72.1%)
Self-perceived need of oral care Need of immediate care (dental pain management, extractions, dental trauma) 160 (16.6%)
Preventive/routine dental services (Scaling and polishing, topical fluorides, fissure sealants) 547 (56.9%)
Advanced treatment options (Restorations, root canal treatments, crowns, orthodontic procedure, and others) 253 (26.3%)
Pattern of dental visit i) less than 6 months ago 475 (49.4%)
ii) 6 months to less than 1 years ago 165 (17.2%)
iii) 1 year to less than 2 years ago 221 (23.0%)
iv) 2 years or more years ago 99 (10.3%)
Oral health care affordability Dental visit for the last 6 months was a financial burden 377 (39.2%)
Statistical analysis
Statistical analyses performed included descriptive statistics, binary logistic regression and multinomial logistic regression. The binary logistic regression and multinomial logistic regression analyses were performed to determine the variables that are associated with the dependent variables at two and three levels, respectively. Regression analyses presented the crude odds ratio, adjusted odds ratio (AOR) and their 95% CI.
In the binary logistic regression analysis, all variables with P-value < 0.25 were considered important and included in the multiple logistic regression analysis. The forward and backward selection methods were used in multiple logistic regression to retain the final variables with P-values < 0.05. Furthermore, the fitness of the final model was evaluated using the Hosmer–Lemeshow test and receiver operating characteristic curve.
The multinomial logistic regression analysis, the full model, including all the independent variables, was run, and the variables with higher P-values were removed manually until a parsimonious model was obtained. All data analyses were performed with SPSS version 24 (IBM Corp., Armonk, NY, USA). P < 0.05 was considered statistically significant.
Results
The demographic and socio behavioural details of the study participants are shown in Table 1. In the current study, 506 (52.7%) of the participants were mothers. The majority of respondents (663, 69.0%) were from rural areas. Of the 960 participants, 693 (72.1%) parents reported that their children have one or more than one untreated dental decay, and 377 (39.3%) respondents reported that dental visits in the last 6 months were a financial burden for the family.
Self-reported child oral health
Table 2 illustrates the multiple logistic regression model in determining the variables linked to self-reported child oral health with one or more untreated dental decay. Children of uneducated parents were 1.6-fold more likely to have one or more untreated dental caries (AOR: 1.66, 95% CI 0.74–3.73; P < 0.001). The children of unemployed parents were 2.1-fold more likely to have one or more untreated dental caries (AOR: 2.11 95% CI 1.24–3.58; P = 0.006). However, children of parents with middle SES were 3.1-fold more likely to have one or more untreated dental caries (AOR: 3.18, 95% CI 1.96–5.17: P < 0.001). Nursery school children were 2.3-fold more likely to have one or more untreated dental caries (AOR: 2.27, 95% CI 1.55–3.35; P < 0.001).Table 2 Multiple logistic regression model of self-reported one or more untreated dental caries
Table 2:Characteristics COR (95% CI) p-value AOR (95% CI) p-value
Gender of the parent Male 0.54 (1.09, 0.82) 0.538 - -
Female (ref) 1 - - -
Parents age groups 25-30 (ref) 1 - 1 -
30-35 0.92 (0.58, 1.45) 0.713 0.72 (0.41, 1.24) 0.234
35-40 1.43 (0.88, 2.31) 0.146 1.26 (0.69, 2.29) 0.449
> 40 years 1.02 (0.67, 1.56) 0.919 1.87 (1.12, 3.12) 0.017
Parental education level Graduation and above (ref) 1 - 1 -
High school 1.04 (0.73, 1.47) 0.848 0.65 (0.41, 1.02) 0.063
Primary school 0.44 (0.29, 0.68) < 0.001 0.28 (0.15, 0.51) < 0.001
Uneducated 2.18 (1.12, 4.25) 0.022 1.66 (0.74, 3.73) 0.218
Marital status Married (ref) 1 - 1 -
Widowed/ Divorced/Separated 0.39 (0.27, 0.58) < 0.001 0.31 (0.19, 0.50) < 0.001
Parental occupation Unemployed 1.45 (0.99, 2.13) 0.059 2.11 (1.24, 3.58) 0.006
Self-employed 2.21 (1.33, 3.65) 0.002 3.22 (1.75, 5.92) < 0.001
Government employee (ref) 1 - 1 -
Private employee 3.90 (1.91, 7.96) < 0.001 3.41 (1.30, 9.00) 0.013
Level of child study Nursery School 1.72 (1.29, 2.29) < 0.001 2.27 (1.55, 3.35) < 0.001
Primary school (ref) 1 - 1 -
Place of residence Rural 0.61 (0.44, 0.84) 0.002 0.54 (0.37, 0.81) 0.003
Urban (ref) 1 - 1 -
Socio economic status Low (≤ 3,000 Saudi Riyals) 1.09 (0.78, 1.52) 0.612 1.02 (0.59, 1.77) 0.935
Middle (> 3,000 to < 10,000 Saudi Riyals) 2.52 (1.74, 3.64) < 0.001 3.18 (1.96, 5.17) < 0.001
High (≥ 10,000 Saudi Riyals) (ref) 1 - 1 -
Self-perceived need for child dental care
Table 3 shows the multinomial logistic regression model to evaluate the variables associated with self-perceived need for child dental care during the COVID 19 outbreak. Parents with primary educational level children were 2.6-fold more likely to need immediate oral care (AOR: 2.63, 95% CI 1.27–5.44; P = 0.009). Children of widowed/divorced parents were 6.2-fold more likely to need immediate dental care (AOR: 6.16, 95% CI 3.14–12.07; P < 0.001). In addition, children of unemployed parents were 4.9-fold more likely to need advanced care (AOR: 4.91, 95% CI 3.00–8.06; P < 0.001). Children of self-employed parents were 11.9-fold more likely to need immediate care (AOR: 11.94, 95% CI 5.65–25.23; P < 0.001). Children of privately employed parents were 21.2-fold more likely to need immediate care (AOR: 21.23, 95% CI 9.70–46.49; P < 0.001). The parents of nursery-level children were 5.4-fold more likely to need immediate care (AOR: 5.38, 95% CI 3.01–9.60; P < 0.001).Table 3 Multinomial logistic regression model for self-perceived need for dental care during the COVID 19 outbreak
Table 3Characteristics Need immediate care AOR (95% CI) p-value Need advance care AOR (95% CI) p-value
Gender of the parent Male 8.39 (4.63, 15.19) 0.018 1.57 (1.10, 2.26) 0.014
Female (ref) 1 - 1 -
Parents age groups 25-30 (ref) 1 - 1 -
30-35 - - - -
35-40 - - - -
> 40 years - - - -
Parental Education level Graduation and above (ref) 1 - 1 -
High school 0.72 (0.40, 1.28) 0.260 0.55 (0.26, 1.15) 0.111
Primary school 2.63 (1.27, 5.44) 0.009 1.05 (0.57, 1.93) 0.888
Uneducated 0.31 (0.14, 0.72) 0.006 0.90 (0.58, 1.39) 0.623
Marital status Married (ref) 1 - 1 -
Widowed/ Divorced/Separated 6.16 (3.14, 12.07) < 0.001 1.83 (1.08, 3.10) 0.024
Parental occupation Unemployed 5.49 (2.84, 10.64) < 0.001 4.91 (3.00, 8.06) < 0.001
Self-employed 11.94 (5.65, 25.23) < 0.001 0.90 (0.48, 1.70) 0.753
Government employed (ref) 1 - 1 -
Private employee 21..23 (9.70, 46.49) < 0.001 1.03 (0.46, 2.27) 0.952
Level of child study Nursery School 5.38 (3.01, 9.60) < 0.001 0.97 (0.66, 1.40) 0.850
Primary school (ref) 1 - 1 -
Place of residence Rural 0.29 (0.18, 0.47) < 0.001 1.10 (0.76, 1.59) 0.619
Urban (ref) 1 - 1 -
Socio economic status Low (≤ 3,000 Saudi Riyals) 11.37 (4.93, 26.26) < 0.001 1.34 (0.78, 2.32) 0.293
Middle (> 3,000 to < 10,000 Saudi Riyals) 5.44 (2.54, 11.63) < 0.001 1.11 (0.73, 1.70) 0.629
High (≥ 10,000 Saudi Riyals) (ref) 1 - 1 -
Pattern of child dental visits
Table 4 illustrates a multinomial logistic regression model to evaluate the variables associated with the pattern of child dental visits during the COVID-19 outbreak. Children of widowed/divorced parents were 6.0-fold more likely to have more than 2 years since their last dental visit (AOR: 5.99, 95% CI 2.89–12.40; P < 0.001). Children of unemployed parents were 2.2-fold more likely to have a dental visit within 6 months to 1 year (AOR: 2.21, 95% CI 1.35–3.60; P = 0.002). Moreover, children of self-employed parents were 12.8-fold more likely to have more than 2 years since their last dental visit compared with government-employed parents (AOR: 12.79, 95% CI 5.55–29.48; P < 0.001). Parents from rural areas were 26.3-fold more likely to have more than 2 years since their last dental visit for their children (AOR: 26.34, 95% CI 7.48–45.79; P < 0.001).Table 4 Multinomial Logistic regression model for pattern of dental visit during the COVID 19 outbreak
Table 4Characteristics 6m – 1 year AOR (95% CI) p-value 1 – 2 years AOR (95% CI) p-value More than 2 years (AOR 95% CI) p-value
Gender of the parent Male 0.37 (0.24, 0.56) < 0.001 1.05 (0.72, 1.53) 0.815 2.35 (1.35, 4.11) 0.003
Female (ref) 1 - 1 - 1 -
Parents age groups 25-30 (ref) 1 - 1 - 1 -
30-35 1.61 (0.80, 3.24) 0.181 0.53 (0.30, 0.91) 0.023 0.67 (0.31, 1.46) 0.315
35-40 1.56 (0.79, 3.11) 0.201 0.77 (0.45, 1.33) 0.346 0.61 (0.20, 1.81) 0.372
> 40 years 1.18 (0.61, 2.28) 0.631 0.36 (0.22, 0.61) < 0.001 1.39 (0.60, 3.23) 0.439
Parental education level Graduation and above (ref) 1 - - - - -
High school - - - - - -
Primary school - - - - - -
Uneducated - - - - - -
Marital status Married (ref) 1 1 1
Widowed/ Divorced/Separated 0.87 (0.49, 1.55) 0.634 1.55 (0.88, 2.74) 0.129 5.99 (2.89, 12.40) < 0.001
Parental occupation Unemployed (ref) 2.21 (1.35, 3.60) 0.002 0.29 (0.16, 0.51) < 0.001 3.06 (1.70, 5.52) < 0.001
Self-employed 0.42 (0.22, 0.81) 0.009 0.04 (0.01, 0.14) < 0.001 12.79 (5.55, 29.48) < 0.001
Government employed 1 1 1
Private employee 0.97 (0.52, 1.79) 0.916 0.14 (0.06, 0.33) < 0.001 0.80 (0.27, 2.40) 0.688
Level of child study Nursery School 0.26 (0.17, 0.39) < 0.001 0.95 (0.64, 1.41) 0.803 1.54 (0.78, 3.05) 0.210
Primary school (ref) 1 - 1 - 1 -
Place of residence Rural 1.42 (0.88, 2.29) 0.153 0.67 (0.46, 0.98) 0.041 26.34 (7.48, 45.79) < 0.001
Urban (ref) 1 - 1 - 1 -
Socio economic status Low (≤ 3,000 Saudi Riyals) - - - - - -
Middle (> 3,000 to < 10,000 Saudi Riyals) - - - - - -
High (≥ 10,000 Saudi Riyals) (ref) - - - - - -
Oral health care affordability
Last 6 months of child dental visits caused a financial burden for families
Table 5 illustrates a multiple logistic regression model in determining the variables associated with oral health care affordability. Children of parents with high school educational background were 1.2-fold more likely to have a financial burden than graduate parents (P = 0.408). Children of widowed/divorced parents were 1.7-fold more likely to have financial burden (AOR: 1.68, 95% CI 1.06–2.66; P = 0.028). Similarly, children of unemployed parents were 4.3-fold more likely to have a financial burden (AOR: 4.34, 95% CI 2.73–6.89; P < 0.001). Consequently, children of privately employed parents were 15.8-fold more likely to have a financial burden (AOR: 15.81, 95% CI 7.72–22.37; P < 0.001). Children of parents with low SES were 1.09-fold more likely to have a financial burden (AOR: 1.09, 95% CI 0.68–1.75; P = 0.715).Table 5 Multiple logistic regression model of dental care affordability during the COVID 19 outbreak
Table 5Characteristics Financial burden for last 6-month dental visit
COR (95% CI) p-value AOR (95% CI) p-value
Gender of the parent Male 0.63 (0.48, 0.82) 0.001 0.67 (0.48, 0.94) 0.020
Female (ref) 1 - 1 -
Parents age groups 25-30 (ref) 1 - 1 -
30-35 3.41 (2.13, 5.45) < 0.001 3.60 (2.03, 6.36) < 0.001
35-40 2.30 (1.43, 3.72) 0.001 1.90 (1.08, 3.35) 0.026
> 40 years 2.51 (1.61, 3.91) < 0.001 3.40 (2.01, 5.75) < 0.001
Parental educational level Graduation and above (ref) 1 - 1 -
High school 1.30 (0.95, 1.77) 0.099 1.18 (0.80, 1.75) 0.408
Primary school 0.52 (0.32, 0.85) 0.008 0.50 (0.27, 0.92) 0.025
Uneducated 1.81 (1.12, 2.93) 0.016 0.59 (0.31, 1.11) 0.099
Marital status Married (ref) 1 - 1 -
Widowed/ Divorced/Separated 1.85 (1.27, 2.70) 0.001 1.68 (1.06, 2.66) 0.028
Parental occupation Unemployed 3.41 (2.40, 4.85) < 0.001 4.34 (2.73, 6.89) < 0.001
Self-employed 1.43 (0.94, 2.16) 0.094 2.29 (1.40, 3.76) 0.001
Government employee (ref) 1 - 1 -
Private employee 7.39 (4.33, 12.62) < 0.001 15.81 (7.72, 22.37) < 0.001
Level of child study Nursery School 0.93 (0.72, 1.22) 0.610 - -
Primary school (ref) 1 - - -
Rural 0.92 (0.69, 1.21) 0.533 - -
Urban (ref) 1 - - -
Socio economic status Low (≤ 3,000 Saudi Riyals) 2.21 (1.61, 3.04) < 0.001 1.09 (0.68, 1.75) 0.715
Middle (> 3,000 to < 10,000 Saudi Riyals) 0.76 (0.55, 1.06) 0.111 0.56 (0.37, 0.84) 0.005
High (≥ 10,000 Saudi Riyals) (ref) 1 - - -
Discussion
The present study evaluated the socio-behavioural variables linked with child oral health during the COVID-19 outbreak in Al Jouf, Northern Province of KSA. Given the feasibility and practical limits of the pandemic, self-reported data were used in this study. In this study, 72.1% of parents reported that their children have one or more untreated dental decay. During the pandemic, self-reported dental needs have increased in the Saudi population.21 Children of widowed/divorced parents, unemployed parents and parents with nursery-school-level children were 6.2-fold, 5.5-fold and 5.4-fold more likely to need immediate dental care for dental pain management, extractions and dental trauma in the current study. A study investigated the need for emergency dental treatment in the Saudi population during the COVID-19 outbreak and found that some people were afraid to seek dental treatment.21 By contrast, some people were eager to attend non-essential and aesthetic concerns.22 In many urgent dental situations, most people selected teleconsultations to address dental problems at home instead of going to dentist's office.22 Furthermore, age, gender, socioeconomic factors, accessibility to health care and perceptions of service quality influenced health-seeking behaviour during the pandemic.22
In the current study, privately employed parents were 15.8-fold, and un-employed parents were 4.3-fold more likely to have a financial burden for the last child dental visit. This result could be due to the COVID-19 pandemic, thereby affecting the labour market and economy throughout the world. COVID-19 had a significant influence on the Saudi economy in 2020, creating a major spike in the cost of living and worsening living conditions for wide segments of the Saudi populace.23 Furthermore, private and public sectors experienced negative growth rates of 10.1% and 3.5%, respectively, whereas the unemployment rate rose to 15.4% in KSA.24 Based on a study from the United States, COVID-19 exacerbates child oral health disparities the most for disadvantaged children.9 Unaffordable dental treatment has been demonstrated to have a negative impact on person's health and general well-being.25
This study found that parents from rural areas, self-employed parents and widowed/divorced parents were 26.3-fold, 12.8-fold and 6.0-fold more likely to have more than 2 years of without seeing a dentist for their children, respectively. The oral health care services in KSA were underutilised well before the pandemic.26 Based on the American Academy of Pediatric Dentistry, children should visit a dentist regularly, with the frequency of visits primarily determined by the risk of disease or child's individualised needs.27 Children in Riyadh, KSA, were reported to have used low rates of emergency dental treatment during the outbreak.28 The ministry of health in KSA provides free oral health care to Saudi population. However, in the Saudi population, non-regular dental visits have a significant rate, which has been related to dental anxiety, a lack of parental motivation and inadequate oral health literacy.10 , 29 , 30 In addition, parents’ education and awareness play an important role in determining whether to take their children to the dentist for disease or preventative treatment.31 Therefore, changing this behaviour is critical, and further studies should be planned to assess the barriers in utilising oral care services in the Saudi population.
Children in the present study have a higher prevalence of self-reported dental caries and a higher unmet need for dental care. Before the pandemic, the management of children's oral health remained a serious concern, although the Saudi population have free access to dental care.30 The leading causes of disparities in dental treatment utilisation in North Africa and the Middle East include the lack of education, a low income and health insurance.32 , 33 In addressing the compromised oral health care delivery system during the current pandemic, oral health care providers and policymakers should integrate tele dentistry into routine dental practice.34 , 35
This study has few limitations. The cross-sectional nature of the study causes difficulty in making causal assumptions, and the data were collected through a self-reported online questionnaire. The self-reported questionnaires are at risk of distortion and bias, such as inaccurate recall and bias due to social desirability.17 Moreover, establishing whether financial constraints play a role in poor predicted oral health outcomes is difficult.36 Children of uneducated, unemployed, widowed/divorced, low-income parents have been connected to poor oral health outcomes, affordability, dental visit patterns, and self-perceived dental need. Oral health professionals and policy makers should identify the socio socio-behavioural characteristics linked to children's oral health needs during the coronavirus occurrence and plan to increase the utilization of preventive oral health services and promote preventive health-seeking behaviours among children and their families to prevent the onset of oral disease and to intervene in a timely manner during a pandemic. The policymakers and oral health care providers adopt the policies to provide primary preventive care and enhance patient education by using technology such as teledentistry. Further longitudinal studies should be conducted to determine the socio-behavioural barriers that affect child oral health in handling future epidemic or pandemic emergencies. The present study recommends that oral health literacy programmes could shift parent's attitude on availing dental care and encourage early preventive oral care for children, particularly those at high risk for oral disease and those from low-income families.
Conclusion
Children in this study had a higher prevalence of self-reported dental caries. Children of rural areas, uneducated, unemployed, widowed/divorced and low and middle-income parents and nursery school children have been associated with poor predicted oral health outcomes during the pandemic.
AUTHORS’ CONTRIBUTIONS
The study was designed and planned by RKG, MFA, and KKG carried out the study. RKG and KKG contributed with the statistical analyses. RKG, KKG, and MIK assisted in the interpretation of the findings. RKG, KKG, MIK, and SK were contributed to writing the first and final drafts of the manuscript. NI, AAB KKM, OAB and AAB were contributed in major revision of manuscript according to reviewer's comments. All authors contributed to critical review of the manuscript.
DATA AVAILABILITY STATEMENT
The datasets generated and/or analyzed during the current study are not publicly available due to limitations of ethical approval involving the patient data and anonymity but are available from the corresponding author on reasonable request.
FUNDING
This work was supported by Deanship of Scientific Research at Jouf University through research grant no (DSR-2021-01-0116).
Declaration of Competing Interest
The authors declare no potential conflict of interest for this article.
Appendix Supplementary materials
Image, image 1
Image, image 2
ACKNOWLEDGEMENTS
The authors extend their appreciation to the Deanship of Scientific Research at Jouf University for funding this work through research grant no (DSR-2021-01-0116).
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.identj.2022.12.003.
==== Refs
REFERENCES
1 Luzzi V Ierardo G Bossù M Polimeni A. Paediatric Oral Health during and after the COVID-19 Pandemic Int J Paediatr Dent 31 1 2021 Apr 1 20 26 33012056
2 Fofana NK Latif F Sarfraz S Bilal Bashir MF Komal B Fear and agony of the pandemic leading to stress and mental illness: An emerging crisis in the novel coronavirus (COVID-19) outbreak Psychiatry Research 291 2020 113230 Elsevier Ireland Ltd
3 Peng X Xu X Li Y Cheng L Zhou X Ren B. Transmission routes of 2019-nCoV and controls in dental practice Int J Oral Sci 12 1 2020 Mar 3 1 6 31900382
4 Izzetti R Nisi M Gabriele M Graziani F. COVID-19 Transmission in Dental Practice: Brief Review of Preventive Measures in Italy J Dent Res 99 9 2020 Aug 1 1030 1038 32302257
5 World Health Organization. Considerations for the provision of essential oral health services in the context of COVID-19 [Internet]. Vol. 3 August, Interim Guidance. 2020. p. 1–5. Available from: https://www.who.int/publications/i/item/who-2019-nCoV-oral-health-2020.1
6 Volgenant CMC de Soet JJ. Cross-transmission in the Dental Office: Does This Make You Ill? Curr Oral Heal Reports 5 4 2018 Dec 25 221 228
7 Mallineni SK Innes NP Raggio DP Araujo MP Robertson MD Jayaraman J. Coronavirus disease (COVID-19): Characteristics in children and considerations for dentists providing their care International Journal of Paediatric Dentistry 30 2020 245 250 Blackwell Publishing Ltd 32250505
8 Al-Khalifa KS Bakhurji E Halawany HS Alabdurubalnabi EM Nasser WW Shetty AC Pattern of dental needs and advice on Twitter during the COVID-19 pandemic in Saudi Arabia BMC Oral Health 21 1 2021
9 Kalash DA. How COVID-19 deepens child oral health inequities J Am Dent Assoc 151 9 2020 643 645 32854864
10 Al Agili DE Farsi NJ Need for dental care drives utilisation of dental services among children in Saudi Arabia Int Dent J 70 3 2020 Jun 8 183 192 31912900
11 Nazir MA. Predictors of routine dental check-up among Male adolescents in Saudi Arabia Acta Stomatol Croat 53 3 2019 255 263 31749457
12 Thomson WM Williams SM Broadbent JM Poulton R Locker D. Long-term dental visiting patterns and adult oral health J Dent Res 89 3 2010 307 311 Mar 20093674
13 Crocombe LA Broadbent JM Thomson WM Brennan DS Poulton R. Impact of dental visiting trajectory patterns on clinical oral health and oral health-related quality of life J Public Health Dent 72 1 2012 36 44 Dec 22316176
14 Improving Access to Oral Healthcare: Adopted by the FDI General Assembly: 27–29 September 2021, Sydney, Australia Int Dent J 72 1 2022 2 3 Feb 1 35074203
15 Camargo MBJ Barros AJD Frazão P Matijasevich A Santos IS Peres MA Predictors of dental visits for routine check-ups and for the resolution of problems among preschool children Rev Saude Publica 46 1 2012 87 97 Feb 22218761
16 John JR Mannan H Nargundkar S D'Souza M Do LG Arora A Predictors of dental visits among primary school children in the rural Australian community of Lithgow BMC Health Serv Res 17 1 2017 Apr 11 1 10 28049468
17 Choi BCK Pak AWP. A catalog of biases in questionnaires Prev Chronic Dis 2 1 2005
18 Andersen RM, Davidson PL. Improving Access to Care in America: Individual and Contextual Indicators. In: Changing the US Health Care System: Key Issues in Health Services Policy and Management. 2007. p. 3–31.
19 Babitsch B Gohl D von Lengerke T. Das Verhaltensmodell der Inanspruchnahme gesundheitsbezogener Versorgung von Andersen re-revisited: Ein systematischer Review von Studien zwischen 1998-2011 GMS Psycho-Social-Medicine 9 2012
20 Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav 36 1 1995 1 10 7738325
21 Al-Khalifa KS Bakhurji E Halawany HS Alabdurubalnabi EM Nasser WW Shetty AC Correction to: Pattern of dental needs and advice on Twitter during the COVID-19 pandemic in Saudi Arabia (BMC Oral Health, (2021), 21, 1, (456), 10.1186/s12903-021-01825-4) BMC Oral Health 21 1 2021
22 Meisha DE Alsolami AM Alharbi GM. Social determinants of seeking emergency and routine dental care in Saudi Arabia during the COVID-19 pandemic BMC Oral Health 21 1 2021 212 33902535
23 United Nations Diagnostics paper on Socio-Economic impact of COVID-19 in the Kingdom of Saudi Arabia (KSA) UN in KSA [Internet]. 2020 [cited 2022 Mar 3]. Available from: https://saudiarabia.un.org/en/105031-diagnostics-paper-socio-economic-impact-covid-19-kingdom-saudi-arabia
24 Islam MM Alharthi M. Impact of COVID-19 on the Quality of Life of Households in Saudi Arabia Int J Environ Res Public Health 19 3 2022 Jan 29 1538 35162560
25 da Fonseca MA Avenetti D. Social Determinants of Pediatric Oral Health 61 2017 Dental Clinics of North America. W.B. Saunders 519 532
26 AlHumaid J Tantawi M El AlAgl A Kayal S Suwaiyan Z Al Al-Ansari A. Dental Visit Patterns and Oral Health Outcomes in Saudi Children Saudi J Med Med Sci 6 2 2018 89 30787827
27 American Academy of Pediatric Dentistic Best Practices: Examination, Prevention, Guidance/Counseling and Treatment the Reference Manual of Pediatric Dentistry 241 Ref Man Pediatr Dent 1 2018 11
28 Alzahrani SB Alrusayes AA Alfraih YK Aldossary MS. Characteristics of paediatric dental emergencies during the COVID-19 pandemic in Riyadh City, Saudi Arabia Eur J Paediatr Dent 22 2 2021 95 97 34237997
29 Baghdadi ZD. Managing dental caries in children in Saudi Arabia Int Dent J 61 2 2011 Apr 101 108 21554279
30 Alshahrani A Raheel S. Health-care System and Accessibility of Dental Services in Kingdom of Saudi Arabia: An Update J Int Oral Heal 8 8 2016 883 887
31 Council I of M and NR Improving access to oral health care for vulnerable and underserved populations Improv Access to Oral Heal Care Vulnerable Underserved Popul 2012 1 279 Jul 13
32 Reda SF Reda SM Murray Thomson W Schwendicke F Inequality in Utilization of Dental Services: A Systematic Review and Meta-analysis Am J Public Health 108 2 2018 Feb 1 e1 e7
33 Bourgeois DM Llodra JC. Global burden of dental condition among children in nine countries participating in an international oral health promotion programme, 2012-2013 Int Dent J 64 Suppl 2 2014 27 34 25209648
34 Ghai S. Teledentistry during COVID-19 pandemic Diabetes Metab Syndr Clin Res Rev 14 5 2020 Sep 1 933 935
35 Kumar S. The Role of Dentists and Members of the Dental Team During Infectious Diseases Outbreaks: Adopted by the FDI General Assembly: 27-29 September 2021, Sydney, Australia Int Dent J 72 1 2022 19 21 Feb 1 35074202
36 Thompson B Cooney P Lawrence H Ravaghi V Quiñonez C. The potential oral health impact of cost barriers to dental care: Findings from a Canadian population-based study BMC Oral Health 14 1 2014 1 10 Jun 25 24383547
| 0 | PMC9742219 | NO-CC CODE | 2022-12-14 23:29:59 | no | Int Dent J. 2022 Dec 12; doi: 10.1016/j.identj.2022.12.003 | utf-8 | Int Dent J | 2,022 | 10.1016/j.identj.2022.12.003 | oa_other |
==== Front
Conscious Cogn
Conscious Cogn
Consciousness and Cognition
1053-8100
1090-2376
The Author(s). Published by Elsevier Inc.
S1053-8100(22)00186-6
10.1016/j.concog.2022.103454
103454
Article
The era of our lives: The memory of Korsakoff patients for the first Covid-19 pandemic lockdown in the Netherlands
Herrmann Dianne ⁎
Oudman Erik
Postma Albert
Helmholtz Institute, Experimental Psychology, Utrecht University, Utrecht, the Netherlands
Lelie Care Group, Slingedael Korsakoff Center, Slinge, 901, 3086 EZ Rotterdam, the Netherlands
⁎ Corresponding author at: Dianne Herrmann, Utrecht University, Heidelberglaan 1, 3584 CS Utrecht, The Netherlands.
12 12 2022
1 2023
12 12 2022
107 103454103454
14 7 2022
2 12 2022
5 12 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.
Memories for worldwide and emotional events (such as 9/11) are more vividly relived and recalled than memories for everyday events. Previous studies have shown that flashbulb memories of a single event enhanced the memory strength in severe amnesia. It is currently unknown whether macro-events that stretch out over longer periods of time (weeks, months) strengthen memory even further. Our aim was therefore to investigate to what extent patients with severe amnesia, due to Korsakoff’s syndrome (KS), were able to relive the first Covid-19 lockdown in the Netherlands, and whether experienced emotions enhanced reliving of the participants. We included 22 KS patients and 24 age-, education-, and gender-matched healthy controls. Covid-19 related memories were assessed by measures of autobiographical memory specificity, phenomenological reliving, emotional intensity and semantic-and episodic knowledge about the first lockdown in March 2020 – May 2020 in the Netherlands. Although amnesia patients remembered significantly fewer autobiographical details regarding the Covid-19 lockdown than healthy controls, one fourth of the KS patients recalled specific events. Amnesia patients reported levels of emotional intensity equivalent to those in the control group. Stronger autobiographical reliving was associated with higher emotional intensity. Both amnesia patients and healthy controls had higher recall of episodic than semantic lockdown related information. In conclusion, results demonstrate that information for macro-events can still be memorized and relived, most specifically when emotional valence is high, even by highly amnestic patients.
Keywords
Autobiographical memory
Korsakoff’s syndrome
Phenomenological reliving
Emotion
Covid-19
==== Body
pmc1 Introduction
Korsakoff’s syndrome (KS) is a neuropsychiatric syndrome caused by thiamine deficiency and concomitant alcoholism leading to lesions in the diencephalon. KS is characterized by severe anterograde and retrograde amnesia (Arts, Walvoort, Kessels, 2017). Memory impairments in patients with KS are associated with a difficulty in retrieving contextual information, and a difficulty in remembering where, when and how a memory was acquired, typically regarded as a deficit in episodic memory. Memory deficits are also present in their autobiographical memory (El Haj & Nandrino, 2017).
In spite of the severe memory impairments found in KS patients, not all events are forgotten. Multiple studies have found a relationship in amnesic patients between emotionally charging of events and the ability to recall them. Emotionally charged memories are often assessed in experimental setups by asking details on flashbulb memories. Flashbulb memories are detailed and vivid autobiographical memories for surprising and emotionally arousing, public events (El Haj et al., 2016a), for example memories for the terrorist attacks on 9/11/2001. More than half of the KS patients remembered 9/11, and some had strong flashbulb memories regarding the event (Candel et al., 2003). Moreover, emotions seems to enhance the autobiographical memory through activating the ability to remember thoughts and feelings associated with the original 9/11 event. This could imply that emotion plays an important role in the recall of autobiographical memories in KS patients (El Haj & Nandrino, 2017). Moreover, in a recent case study by Gandolphe & El Haj (2018) a male patient with KS also had vivid emotional recall of the terrorist attack in Paris 2015, although memory for the semantic contents of the event was not fully intact (see Table 1 ).Table 1 Demographic Characteristics of KS Patients (KS) and Healthy Controls (HC).
KS (n = 22) HC (n = 24) Statistics
Gender (m: f)a 15: 7 14: 10 U = 292.00, z = 0.68, p =.500
Age (M, SD)b 65.23 (8.39) 59.12 (13.83) t(44) = 1.83, p =.076
Education level (M, SD)c
TICS-m 4.23 (1.48)27.27
(5.20) 4.50 (1.10)36.88
(1.45) U = 293.00, z = 0.67, p = 0.480
U = 528.00, z = 5.83, p <.001
dCognitive skills (TICS-m).
a Gender ratio male: female.
b Age in years.
c Educational level was assessed in seven categories: one, primary school; seven, academic degree (Verhage, 1964).
Amnesic patients show higher recall of memories for emotional and arousing events regarding episodic details compared to semantic information, but their amnesia is often more of an episodic nature than a semantic. A possible explanation for this apparent contradiction could be that for an emotional event outside the laboratory with widespread media coverage, such as the September 11th attacks, cognitive deficits can be partially overcome (Budson et al., 2004). Memories for high emotional and public events as such may elicit improved recall of episodic and autobiographical details in amnesic patients (Candel et al., 2003, Budson et al., 2004, El Haj et al., 2016, El Haj et al., 2016, Luchetti and Sutin, 2017, El Haj and Nandrino, 2017, El Haj and Nandrino, 2018).
One development that has shocked our public and personal lives at an universal scale is the emergence of the Covid-19 pandemic in 2019. The social restrictions have had a substantial impact on inpatient and outpatient diagnosed with amnesia, with cancelations or postponement of outpatient visits starting at the beginning of March 2020 in The Netherlands and other European countries (). The Covid-19 era does not concern a singular event but rather a large period of time, or what Gardner (2009) calls a macro event. It is important to examine whether KS patients living as inpatients in a specialized facility, can remember the Covid-19 pandemic due to this macro event being important, emotional, public and having widespread media-coverage.
Research on the connections between memory, emotion and worldwide events, especially macro events, is very scarce in patients with KS. Therefore, the current study focused on whether and how KS patients remember the first Covid-19 lockdown in the Netherlands in March 2020. In particular we attempted to investigate various aspects of declarative memory for the Covid-19 era (autobiographical, semantic and episodic memory). Additionally, we examined to what extent KS patients could recall phenomenological characteristics of a memory related this lockdown to measure the amount of autobiographical reliving. Qualitative analyses were used to express the differences in recalled information regarding the first Covid-19 Lockdown.
2 Methods
2.1 Participants
In this study, 22 patients diagnosed with KS participated. They were all inpatients of Korsakoff Center ‘Slingedael’, Rotterdam, The Netherlands; a specialized long-term care facility for KS patients. All patients fulfilled the criteria for KS described by Kopelman (2002) and the DSM-5 criteria for alcohol-induced major neurocognitive disorder of the amnestic-confabulatory type (APA, 2013). The patients were in the chronic, amnestic stage of KS. None of them were in the Wernicke Encephalopathy phase. All patients were at least sober for six months, and testing was conducted after at least six months post-Wernicke Encephalopathy; to assure a stable cognitive state. KS patients and healthy controls had a comparable age, gender and educational level. Also, 24, gender-, age-, and education-matched healthy controls participated in the study. General cognitive functioning, as indicated on the TICS-M was lower in KS patients than healthy control subjects (see Table 4). There were no statistically significant group differences regarding age, sex and education level between KS patients and matched healthy controls. Level of education was assessed according to the Dutch education classification system by Verhage. It consists out of seven categories,‘1′ being the lowest (primary school) and ‘7′ the highest (academic degree) (Bouma et al., 2012).
Participants did not receive financial compensation for their participation. Informed consent was obtained for all participants. Ethical approval was obtained by the Ethics Review board of the Faculty of Social and Behavioral Sciences of Utrecht University. Data collection took place in December of 2020.
2.1.1 Procedure
The KS patients were individually interviewed at the KS care facility. Informed consent was obtained at the start of the interview. The interview was conducted inside a testing room without distractions.. The control group was individually recruited trough Facebook and through personal connections. The controls and patients participated voluntarily. The controls were interviewed by telephone due to the Covid-19 restrictions from the government at that moment. The informed consent was read aloud and obtained orally. All participants had to respond to the questions asked by the experimenter. The memories were written down word for word by the experimenter. In both controls and patients, scaled questions (see Table 2 and Table 3 ) were circled by the experimenter. Open questions (see Table 4 ) were written down by the experimenter. If an answer was vague or needed more elaboration, the experimenter asked the participant to explain or clarify their answers. Further instructions were provided if the participant experienced any difficulties with the subjective ratings or questions.Table 2 Variables of Phenomenological Reliving.
Variable Description of Rating Scale
Reliving I feel as though I am reliving the original event
Back in Time As I remember the event, I feel that I travel back to the time when it happened
Remember/know I can actually remember the event rather than knowing that it happened
Real/imagine I believe the event in my memory really occurred in the way I remember it and that I have not imagined or fabricated anything that did not occur
See As I remember the event, I can see it in my mind
Hear As I remember the event, I can hear it in my mind
Talk As I remember the event, I or other people are talking
Emotion As I remember the event, I can feel the emotions I felt then
Importance This memory is significant for my life
Impact The event from the memory has made a lot of impact
Rehearsal Since it happened, I have thought about this event
Rehearsal Since it happened, I have talked about this event with others
News rehearsal Since it happened, I have watched the news about Covid-19
Place I can recall the place where it occurred
Time I can recall the time (day/month/year) when it occurred
Own perspective I remember the event from my own perspective
Other perspective I remember the event from another perspective
Table 3 Variables Used to Measure Emotion.
1. How sad do you feel when you remember the event?
2. How angry do you feel when you remember the event?
3. How much fear do you feel when you remember the event?
4. How happy do you feel when you remember the event?
Table 4 Episodic and Semantic Knowledge Variables about the first lockdown in March 2020.
Semantic Knowledge
1. In which month did the first quarantine period start?
2. Who was the minister that spoke about the lockdown during the press-conference?
3. What happened with sports during this quarantine period?
4. Which product got massively hoarded during the first quarantine period?
5. At what time did that press conference start?
Episodic Knowledge
1. Where were you (on March 12, 2020) when the minister told that the Netherlands was about to go into lockdown?
2. What where you doing when you heard the news about the lockdown?
3. With whom where you when the heard the news?
4. How did you hear the news about the lockdown (radio, television, etc.)?
5. How where you feeling when you heard the news about the lockdown?
2.1.2 Design
The total interview took about 20 min to complete and consisted out of 40 questions. Question 1–17 measured the amount of phenomenological reliving, question 18–22 measured emotion valence, question 23–30 measured the episodic memory, question 31–38 measured semantic memory and question 39–40 measured emotion for the first lockdown. Questions were assessed using a directive interview technique. The initial responses of the patients and controls were written down. There was a small delay between every question because the answers had to be written down.
The sub-questionnaires in this interview examined to what extent KS patients were able to remember the first Covid-19 lockdown by measuring different types of the declarative memory. These measured the autobiographical memory, episodic memory and semantic memory. Autobiographical memory specificity was assessed through the Autobiographical Memory Interview (AMI), to measure if KS patients are able to recall a specific memory related to the Covid-19 lockdown. Phenomenological reliving of that memory was assessed through a questionnaire based on the Autobiographical Memory Questionnaire (AMQ), to measure how many details KS patients could recall of that memory. Semantic and episodic memory was examined through questions based on the studies of Candel et al., 2003, Budson et al., 2004, measuring personal and factual knowledge about the first Covid-19 lockdown. Together they will provide a deeper understanding of the memory for the Covid-19 era in KS patients.
2.1.3 Materials
2.1.3.1 Cognitive impairment
The Modified Telephone Interview for Cognitive Status (TICS-M) was used to screen cognitive deficits of the participants. The TICS-M has high sensitivity and specificity and is a reliable and valid method for detecting cognitive impairment (Fong et al., 2009, Seo et al., 2011). The TICS-M has a cut off score of < 34 for Mild Cognitive Impairment and < 28 for dementia. A specific advantage of the TICS-M over regular screening methods was that the.
Dutch Covid-19 measures could be more easily managed (1.5 m distance) than in case of other screening methods.
2.1.4 Autobiographical memory specificity
To examine autobiographical memory, participants were asked to describe a personal memory of an event related to the first Covid-19 lockdown within a time frame of 2 min. This time frame was chosen to avoid redundancy or distractibility and is found sufficient for autobiographical recall in prior research regarding individuals with Alzheimer’s disease (El Haj & Nandrino, 2018). Participants had to specify their memory as much as possible to see if Korsakoff patients could still generate specific autobiographical memories in relation to this macro-event. Memory specificity was rated on a scale of 0 to 4 based on the AMI (Kopelman, 1994, Levine et al., 2002, El Haj and Nandrino, 2018). These four points measure to what extend both groups can recall specific memories. Four points were given if the event of the memory was specific, vivid, provided details and lasted no longer than 24 h. 3 points were assigned if the event of the memory lasted less than 24 h and situated in time and space without details. 2 points were given if the memory was a macro or repeated event but contained details. 1 point was given if the memory was a macro or repeated event without any details. 0 points were assigned when there was no memory at all. In earlier research this interview had a high reliability between 0.83 and 0.86 and is considered to be a valid instrument for measuring episodic memory richness (Kopelman, 1994, Levine et al., 2002).
2.1.5 Phenomenological reliving
Phenomenological reliving of an autobiographical memory related to the first Covid-19 lockdown was assessed through an adapted version of the AMQ (Rubin et al., 2003, El Haj and Nandrino, 2017). Questions about space, importance, rehearsal and perspective were added to provide more knowledge about these components of reliving (see Table 2). The Cronbach’s alpha showed a reliability of 0.86, indicating high internal consistency. The translation of the AMQ into the Dutch language was done by a professional translator of the Utrecht University to achieve higher translation validity. Validity of the ratings were established by asking participants to provide specific information about that memory if not given during the AMQ. Questions were scored on a seven-point Likert scale from 1 (not at all), 3 (vague), 5 (clear) tot 7 (as clearly as if the event was happening now). A seven-point Likert scale was chosen to measure the participants true subjective evaluation (Finstad, 2010). A high score indicated a high level of phenomenological reliving for a memory, thus, indicating higher reliving of the a memory related to the first Covid-19 lockdown (see Table 5 ).Table 5 Frequencies of Memory Specificity from KS Patients (KS) and Healthy Controls (HC).
KS (n = 22) HC (n = 24)
No memory or general information 2
Repeated or extended event without details 8
A repeated event that situated in a time and space with details 6 2
A specific event (<24 h) without details 1 6
A specific event (<24 h) that was enriched with details 5 16
2.1.6 Emotion
We added four questions about emotion related to the first Covid-19 lockdown to the interview. Although the AMI has one question about emotion in it, we wanted to get more insight into memory and emotion related to the first Covid-19 lockdown in the Netherlands and know what type of emotion predicts higher recall. Emotional intensity of a memory related to the first wave of Covid-19 was measured through four questions on a seven-point Likert scale (see Table 3). The following emotions were included: sad, anger, happy, fear. These four emotions had been chosen given their generality. Emotion for the first lockdown in March 2020 was assessed with an open question followed by a seven-point Likert scale.
2.1.7 Semantic content and episodic context memory
Semantic and episodic memory regarding the first lockdown in March 2020, due to the Covid-19 pandemic, were assessed through 10 questions (see Table 4). The period of the first lockdown was chosen due to the amount of impact this period had on our society. These questions were all based on recall of events. The maximum score for semantic and episodic memory are each 10 points, indicating high recall of knowledge about the first lockdown.
2.1.8 Confabulations
KS patients are known to confabulate, since confabulations are a diagnostic criterium for KS. Confabulations are exaggerated, fantasized or untrue stories that KS patients use to fill the holes they cannot remember. These confabulations do not happen on purpose, but are the result of specific brain damage). Four control questions, based on the Dalla Barba (Dalla Barba et al., 2020), were included in this questionnaire to control for the severity of confabulations; namely:1. What was the color of the shirt of the sign language interpreter during the Covid-19 press conferences?
2. Who succeeded the king when he was infected with Covid-19?
3. Can you remember what you did March 12th?
4. What did you do when the air raid siren went off, because of the Covid-19 lockdown?
The right answer to these questions would be: ‘I don’t know’ or ‘did that really happen?” If KS patients respond with concrete answers to these questions, it could be likely they were confabulating. Furthermore, these questions gave an indication to whether the patients confabulated on the other questions as well. Scoring went be as follows: ‘0′ for a wrong answer, ‘1′ for a vague or almost correct answer, ‘2′ for a right or precise answer and ‘c’ for confabulations. Results of KS patients scoring more than two ‘c’s had to be investigated thoroughly. If the answers to those and other questions were strongly of confabulatory nature, the participant was removed, because of the confabulations influencing the recall of the events too much. We discuss the extent of confabulations in severe KS in our discussion section.
2.1.9 Data analysis
All test scores were compared between KS patients and the controls. Memories were transformed into specificity scores and internal and external details. Internal details are details related to the episodic memory (the main event itself, time, place, perception, thoughts/emotions). External details are related to semantic memory or other events (external events, semantic information, repetition and other statements). Each part of a sentence was transformed into points following the description of Levine et al. (2002) and the categories were each transformed into a sum of scores per participant. Higher scores of internal details indicate higher episodic re-experiencing of a memory (Levine et al., 2002). A 2 × 2 mixed model ANOVA measured the differences between the two groups and amount of internal and external details, with details (internal and external) as the within-subjects variable and group (KS group and controls) the between-subjects factor. The assumptions for a 2x2 mixed model ANOVA were not violated. Due to violation of the assumption of normality and the assumption of homogeneity of variances, the non-parametric Mann-Whitney U test was applied to measure the differences between the groups on memory specificity and memory details. All scoring was performed by the experimenter. An independent rater, blind to diagnosis and study hypotheses, co-scored all transcripts. Interrater reliability was established using the intraclass correlation coefficient (absolute agreement) and was high across both conditions (internal details, r = 0.76, p <.001; external details, r = 0.87, p <.001) The scores reported in this paper are the means of the scores given by the experimenter and by the independent observer.
Emotional intensity was measured by the sum of scores of the four emotions (emotional intensity overall). Mann-Whitney U tests were then used to compare the scores on emotional intensity overall and the emotions separately between the two groups. A Kendell’s tau-b correlation was used to examine the relation between phenomenological reliving and emotion.
Episodic and semantic knowledge scores were composed into a sum of scores for both knowledge types after the correction for possible confabulations. A Mann-Whitney U test was applied to measure the differences in knowledge scores between the participant groups. A Kendell’s tau-b correlation was used to examine the relations between episodic/semantic knowledge and emotion for the first lockdown; exposure to news and episodic/semantic knowledge. More Kendell’s tau-b correlations were used to examine cognitive skills and memory specificity; cognitive skills and phenomenological reliving. A bivariate Pearson’s correlation coefficient was used to measure the relation between cognitive skills and phenomenological reliving. All correlations were two-tailed.
3 Results
3.1 Severe confabulations
Four questions were included in the interview to assess confabulation tendencies (see Methods). Two KS patients were removed from further analysis by reason of severe confabulations (possibly indicating psychosis/ severe lack of reality filtering), as indicated by three or four severe confabulations on the four questions (see also the discussion section). Answers on these control questions and other questions of the semantic/episodic knowledge questionnaire were abnormal and out of context as well. Examples of such answers were:Patient X: I do not remember what I had to do” / “I was watching the outside world.”
Patient Z: “I was eating those chips, you know the orange ones. They are called Cheeto’s.” / “I remember it was announced on Monday when the alarm always goes off, Corona was announced. Every first Monday of the month”.
3.1.1 Autobiographical recall and phenomenological reliving
KS patients (Mdn = 2.00) showed significantly lower scores on memory specificity than the controls (Mdn = 4.00), (U = 414.00, z = 3.58, p <.001, r = 0.53), suggesting that KS patients have a less specific autobiographical memory for a memory related to the first wave of the Covid-19 pandemic (see Table 2). Furthermore, KS patients (M = 4.11, SD = 1.08) recalled less phenomenological characteristics than the controls (M = 5.51, SD = 0.67) This difference (23.90, 95 % CI [–33.17, −14.62]), was significant (t(44) = -5.23, p <.001, d = 1.56) and represented a large effect, suggesting that KS patients have reduced memory reliving for a memory related to the Covid-19 pandemic than the controls.
3.2 Internal and external details
There was a significant main effect of recalled internal and external details (F(1, 44) = 29.03, p <.001, ηp2= 0.61) with more internal details recalled than external details in both participant groups. There was another significant main effect of group on the overall amount of recalled details (F(1, 44) = 14.43, p <.001, ηp2= 0.25) with the control group reporting more details overall. There was no interaction of recalled details for a memory related to the first wave and group (F(1, 44) = 1.59, p =.215, ηp2= 0.04). Suggesting that the number of recalled details did not depend on the participant group.
A follow-up analysis on internal details indicated that healthy controls (M = 7.96, SD = 2.42) recalled more internal details than the KS group (M = 5.64, SD = 2.99). This difference (-2.32, BCa 95 % CI [-3.926, -0.718]), was significant (t(44) = -2.92, p <.01, d = 0.85). Both groups preferred event details over other internal details (see Fig. 1 ) and there was a significant difference in the amount of recalled event details between the KS group and control group (U = 364.50, z = 2.27, p <.05, r = 0.33). More significant differences were found for time (U = 407.00, z = 3.59, p <.001, r = 0.53) and place (U = 172.50, z = -2.27, p <.05, r = -0.33). KS patients scored lower on time and higher on place than the control group (see Fig. 1). No significant differences were found for perception (U = 316.50, z = 1.55, p =.121, r = 0.23) and thought/emotion either (U = 330.50, z = 1.51, p =.131, r = 0.22), indicating that both groups recalled the same amount of these details (see Fig. 1).Fig. 1 Mean Scores of Generated Internal and External Details for a Memory Related to the First Wave of Covid-19 by KS Patients and Controls. Note. Internal details: Event; Time; Place; Perc = perception; T/Em = thought/emotion. External details: Ext = external event; Sem = semantic; Rep = repetition; Other.
A follow-up analysis on external details indicated that healthy controls (Mdn = 3.50) recalled more external details than the KS group (Mdn = 2.50). This difference was significant (U = 352.50, z = 1.98, p <.05, r = 0.29). Out of the external details, semantic information was the most recalled external detail in both groups. Significant differences between KS patients and the control group were found for semantic information (U = 350.50, z = 2.02, p <.05, r = 0.30) and external events (U = 347.50, z = 2.10, p <.05, r = 0.31.). The control group reported higher scores on both details (see Fig. 1). No significant differences were found between the groups for repetition (U = 233.00, z = -0.95, p =.34, r = -0.14) and other (U = 252.50, z = -0.29, p =.77, r = −0.04). Indicating that both groups did not differ on these recalled details (see Fig. 1). KS patients (M = 0.69, SD = 0.25) and the controls (M = 0.71, SD = 0.16) did not significantly differ in internal-to-total detail ratios (t(44) = -0.38, p =.706, d = 0.10). These results suggest that both groups generated a similar amount of internal and external details. Both groups scored close to 1, indicating episodic re-experiencing in both groups.
3.3 Phenomenological reliving
KS patients scored quite high on the individual questions realness (M = 5.82, SD = 1.56), belief (M = 6.73, SD = 0.77), visual imagery (M = 4.73, SD = 2.05), spatial imagery (M = 5.77, S = 1.77) and perspective (M = 6.64, SD = 1.18), indicating high reliving of these phenomenological characteristics. Furthermore, no significant differences were found between KS patients and the control group on realness (U = 289.00, z = 1.13, p =.261, r = 0.17), importance (U = 321.50, z = 1.29, p =.196, r = 0.10) and perspective (U = 246.00, z = -0.68, p =.499, r = 0.19), indicating that both groups had similar scores on these questions.
There was a positive and strong correlation between cognitive skills and memory specificity (τ(46) = 0.54, p <.001). Cognitive skills had a strong and positive relation for phenomenological characteristics as well (r(44) = 0.71, p <.001). These correlations suggest that higher cognitive skills result in higher recall of specific memories and higher reliving.
3.4 Emotion and memory
Correlations were statistically non-significant for the following emotions: anger (τ = 0.07, p =.563), happy (τ = 0.07, p =.575) and fear (τ = 0.16, p =.168)., suggesting that there were no relations between these emotions and phenomenological characteristics of a memory related to the Covid-19 pandemic. The correlation between sad and phenomenological characteristics was positive and moderate (τ(46) = 0.34, p <.01). Also, a significant moderate and positive relation was found between emotional intensity overall for a memory related to the first wave and phenomenological characteristics (τ(46) = 0.32, p <.01), suggesting that more emotional intensity overall and the emotion sad lead to higher reliving of a memory related to the first wave of the pandemic.
There was a significant, positive and moderate correlation between internal details and emotional intensity overall (τ(46) =.28p <.01), suggesting that when participants reported more internal details of a memory related to the first wave, they felt stronger emotions to that memory. However, there was no significant correlation between external details and emotional intensity overall (τ(46) = 0.20, p =.069), suggesting that external details does not have a relation with emotional intensity overall.
3.4.1 Semantic versus episodic knowledge
The total of correct answers of episodic knowledge in KS patients (Mdn = 7.00) was significantly lower than the control group (Mdn = 9.50), (U = 467.00, z = 4.55, p <.001, r = 0.67). Also, KS patients (Mdn = 5.00) scored significantly lower for semantic knowledge than the control group (Mdn = 10.00), (U = 441.00, z = 3.99, p <.001, r = 0.59). These results suggest that KS patients are impaired with respect to episodic and semantic memory compared to the controls for the first lockdown. The difference of accurately recalled information between the two memory types was the largest in the KS group.
The relation between exposure to news about the Covid-19 pandemic and semantic knowledge was positive (τ(46) = 0.33, p <.01). Suggesting that more rehearsal of information about the Covid-19 pandemic leads to higher recall of semantic knowledge.
3.4.2 Emotional intensity
Overall, KS patients reported, slightly, lower scores of emotion for a memory related to the first wave of the pandemic than the control group (see Fig. 2 ). When looking at the emotions separately, we neither observed significant differences between the two groups for sad (U = 334.00, z = 1.76, p =.079, r = 0.26), happy (U = 290.00, z = 0.68, p =.500, r = 0.10), angry (U = 288.50, z = 0.60, p =.551, r = 0.09) and fear (U = 310.00, z = 1.14, p =.256, r = 0.17). These results suggest that KS patients showed a similar ratings of emotion to their memory as the control group. However, the scores of emotional intensity overall (sum of emotional intensity) for a memory related to the first wave significantly differed between the KS group (Mdn = 7.50) and the control group (Mdn = 10), (U = 365.50, z = 2.24, p <.05, r = 0.33). Suggesting that KS patients scored lower emotional intensity overall than the control group. As for the emotional intensity for the first lockdown itself scores did significantly not differ between KS patients patients (Mdn = 5.00) and the control group (Mdn = 5.00), (U = 310.50, z = 1.05, p =.295, r = 0.15). Suggesting that KS patients had a normal emotional response for the first lockdown (see Fig. 2).Fig. 2 Mean Scores Emotional Intensity of KS Patients and Controls. Note. Mean scores with standard deviation of emotion on a 1 (low)- to 7 (high)-point scale: sad, anger, fear, happy, Em. Lckdown = emotion for the first lockdown in March 2020. Error bars are standard error of the mean.
4 Discussion
The aim of this study was to investigate whether patients with severe amnesia due to KS could remember the first Covid-19 lockdown in The Netherlands, and whether they had comparable emotions and reliving of memory recall as healthy controls. While KS patients had worse memory recall regarding the Covid-19 measures, many patients still had vivid reliving of memories despite the severe amnesia. Semantic information was also relatively preserved in KS. There were no differences between KS patients and healthy controls on impact, realness and reversed perspective which are important factors for reliving an event (Greenberg, & Knowlton, 2014). In patients, memory specificity was lower than healthy controls.
4.1 Memory specificity and reliving
Lower specificity of memory implies that KS patients are less able to recall a specific and detailed memory related to the first Covid-19 lockdown. Although, five of the twenty KS patients described a detailed, specific episodic memory; twenty patients were not able to describe with such detail. Memory for a worldwide, public and emotional event may be more enriched with episodic details than regular events. KS patients with more intact cognitive abilities were able to retrieve more specific memories and recalled more details, as indicated by a correlation between the cognitive screening instrument TICS-M and the Covid-19 memory recall. Our results are in line with earlier findings of Kessels & Kopelman, 2012; and the two studies performed by El Haj and colleagues (El Haj and Nandrino, 2017, El Haj and Nandrino, 2018) showing that KS patients still have spared abilities to recall emotional events. In contrast to these studies, here one-fourth of all patients had normal memory specificity, with both earlier studies finding generally compromised memory scores without preservations. Importantly, our results and that of El Haj & Nandrino (2018) highlight that patients diagnosed with KS still can have a strong subjective experience regarding the past, particularly when emotional valence is high. In our study, another one-fourth described a memory that lasted longer than 24 h but still contained details and situated in time and space. This result could be explained by the possible influence of the emotional impact of the global pandemic, and also the effects of the strong impact on memory preservation in severe amnesia patients. It might be that the Covid-19 lockdown was of more emotional relevance for the patients than 9/11 or the Paris terrorist attacks.The latter two were events that could be considered to be more removed from one daily’ lives.
4.2 Semantic and episodic memory
KS patients scored relatively well on the semantic questions, with all patients scoring at least one point on semantic recall about the first lockdown in March 2020. This is consistent with several earlier studies indicating preserved flashbulb memory in KS (Tulving et al., 1991, Waring et al., 2014). In line with the study of Candel et al., 2003, Janssen et al., 2022 we found that although KS patients scored lower than the healthy control group on semantic information of a specific event, the KS patients still had some preserved ability to remember the event. The results of our study add to the indication that KS patients can recall information without very specific cues for emotional and public events that are often repeated in the news following a media event such as terrorist attacks. The KS patients were also able to learn new information independently regarding the Covid-19 lockdown.
Another finding was that KS patients were able to retrieve more episodic than semantic knowledge about the lockdown. Interestingly, both groups remembered more episodic than semantic knowledge and this difference was the largest in KS patients. However, KS patients still performed worse than the control group. This result is in line with other studies (Mori et al., 1999, Budson et al., 2004, Waring et al., 2014) who affirm greater contribution of emotional-based networks like the amygdala when storing and retrieving episodic information for high emotional and worldwide events. The relative sparing of the amygdala within KS may account for the retained emotional memory in KS patients (Candel et al., 2003, Gandolphe and El Haj, 2018).
4.3 Emotion and memory
Emotion plays a pivotal part in the autobiographical and episodic memory system. Emotions activate thoughts and feelings associated with an event, which facilitates the retrieval of emotional memories (El Haj & Nandrino, 2017). Our results showed that the emotional intensity for an autobiographical memory was related to higher reliving in KS patients and healthy controls. Moreover, sadness was related to a stronger sense of reliving as well. These findings are in line with several studies (Candel et al., 2003, Budson et al., 2004, El Haj et al., 2016, El Haj et al., 2016, Luchetti and Sutin, 2017) stating higher reliving for emotional memories. Higher emotional intensity for a memory related to the first wave of Covid-19, led to more reliving of that memory. In addition, higher recall of internal details was related to higher emotional intensity which indicated that emotion is related to the active recall of episodic memories. The results show that despite the introspection difficulties and confabulations, patients with KS can experience some authentic subjective experience of the past when the emotional intensity of that memory is high (El Haj & Nandrino, 2018). This research offers an alternate view on the general consideration that KS patients suffer a diminished subjective experience. KS patients could benefit from highly emotional authentic subjective experiences (El Haj & Nandrino, 2018).
KS patients and the control group reported similar levels of emotions individually for a memory related to the first lockdown. We observed similar results on emotion for the first Covid-19 lockdown, suggesting that KS patients have preserved emotional feelings toward this worldwide and public event and their memories (Gandolphe & El Haj, 2018). The case study by Gandolphe & El Haj (2018) on the Paris attacks reported a comparable finding in one KS patient. These results confirm that for a daily life affecting period, extending beyond the laboratory, such as the Covid-19 pandemic, deficits in emotional processing can be overcome in KS patients. This result provides new knowledge about the emotional intensity of memories of KS patients having the same emotional response as controls for this worldwide event.
4.4 Memory details
One of the contributions of the present study was to acknowledge that such phenomena still hold for emotionally challenging events, such as the first wave of covid-19. However, the generated internal details show that KS patients are still able to re-experience an autobiographical memory, but this re-experience is still more reduced than in healthy controls. Better episodic re-experiencing of a memory was correlated with more internal than external details. In our study, KS patients provided less internal details to their memory related to the first Covid-19 lockdown than the control group. Murphy et al. (2008) found that KS patients and patients with other cognitive impairments recall more external details than the controls for a memory in the past year. In contrast, the internal to total ratio score was quite low in our patients. A possible explanation for the lack of difference in external details between the two groups in our study could be that the control group reported more specific memories containing more details. The control group shared more details along with details non-related to the main event of their memory, whereas the KS group kept close to the main event and describing less enriched memories.
4.5 Confabulations
In our study-two patients with severe confabulations were excluded from further analysis. One reason to do so is that we could not separate actual memories of the Covid-19 lockdown from confabulations based on the regular answers. Because the two patients systematically reported confabulations on the provoked-confabulations questions based on the Dalla Barba list, we excluded them from further analysis (Dalla Barba et al., 2020). There is no consensus on indexing the severity of confabulations within a clinical context objectively, because some patients do not confabulate on the Dalla Barba questionnaire, but confabulate based on proxy-based reports (Rensen, Oudman, Oosterman, Kessels, 2020).
5 Conclusion
To conclude, despite the lower memory scores than the controls, KS patients were still able to recall and relive events from the first wave of the Covid-19 pandemic in March 2020 – May 2020. Some of these memories were enriched with internal memory details stating higher episodic re-experiencing for this worldwide and emotional event. Higher emotional intensity was related to a more thorough level of autobiographical reliving in both KS patients and the controls. Furthermore, the amount of episodic information was more spared than semantic information of this major event, suggesting that KS patients are able to retrieve and learn new episodic and semantic information independently. In addition, new semantic information can be learned by the help of accidental rehearsal and emotional load, however the amount of recall is also related to spared cognitive and memory functioning. This study adds new evidence that despite the severe and clear memory impairments, episodic and semantic information about major, worldwide events can still be memorized in KS patients and that associated emotions may corroborate the recall of the events.
Authorship contributions
E.D. Herrmann: Conceptualization, Methodology, Visualization, Investigation, Data curation, Writing- Original draft preparation, Formal analysis. A. Postma: Visualization, Conceptualization, Validation, Writing - Review & Editing, Supervision. E. Oudman: Supervision, Conceptualization, Resources, Visualization, Writing - Review & Editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
The data that has been used is confidential.
Acknowledgements
We are grateful Julie Janssen for her help in scoring the transcripts.
==== Refs
References
American Psychiatric Association (2013). Diagnostics and statistical manual of mental disorders (5th ed.). Washington, DC.
Arts N.J.M. Walvoort S.J.W. Kessels R.P.C. Korsakoff’s syndrome: A critical review Neuropsychiatric disease and treatment 13 2017 2875 2890 10.2147/NDT.S130078 29225466
Bouma A. Mulder J. Lindeboom J. Schmand B. Handboek neuropspychologische diagnostiek 2de ed. 2012 Pearson Education Benelux B.V
Budson A.E. Simons J.S. Waring J.D. Sullivan A.L. Hussoin T. Schacter D.L. Memory for the September 11, 2001, Terrorist attacks one year later in patients with Alzheimers Disease, patients with Mild Cognitive Impairment, and healthy older a dults Cortex 43 7 2004 875 888 10.1037/08944105.18.2.315
Candel I. Jelicic M. Merckelbach H. Wester A. KS patients’ memories of September 11, 2001 The Journal of Nervous and Mental Disease 191 4 2003 262 265 10.1080/13803390490919335 12695739
Dalla Barba G. Brazzarola M. Marangoni S. La Corte V. Screening for confabulations with the confabulation screen Neuropsychological Rehabilitation 30 2020 116 129 10.1080/09602011.2018.1464475 29688124
El Haj M. Gandolphe M.C. Wawrziczny E. Antoine P. Flashbulb memories of Paris attacks: Recall of these events and subjective reliving of these memories in a case with Alzheimer disease Medicine 95 46 2016 1 5 10.1097/MD.0000000000005448
El Haj M.E. Kapogiannis D. Antoine P. Phenomenological reliving and Visual imagery during autobiographical recall in Alzheimer’s Disease Journal of Alzheimer’s Disease 52 2016 421 431 10.3233/JAD-151122
El Haj M.E. Nandrino J.L. Phenomenological characteristics of autobiographical memory in KS’s syndrome Consciousness and Cognition 55 2017 188 196 10.1016/j.concog.2017.08.011 28886469
El Haj M.E. Nandrino J.L. Seeing life through rose-colored spectacles: Autobiographical memory as experienced in KS’s syndrome Consciousness and Cognition 60 2018 9 16 10.1016/j.concog.2018.02.007 29501971
Finstad K. Response interpolation and scale sensitivity: Evidence against 5-point scales Journal of Usability Studies 5 3 2010 104 110
Fong T.G. Fearing M.A. Jones R.N. Shi P. Marcantonio E.R. Rudolph J.L. …Inouye S.K. Telephone Interview for Cognitive Status: Creating a Crosswalk with the Mini-Mental State Examination Alzheimer’s & Dementia 5 6 2009 492 497 10.1016/j.jalz.2009.02.007
Gandolphe M.C. El Haj M. Flashbulb memories for the Paris attacks in KS’s syndrome: A case study Archives of Clinical Psychiatry 45 2 2018 49 50 10.1590/0101-60830000000155
Gardner R.M. A comparison of event analysis and multilinear events sequencing techniques for reconstructing unique phenomena Journal of the Association for Crime Scene Reconstruction 16 2009 1 9
Greenberg D.L. Knowlton B.J. The role of visual imagery in autobiographical memory Memory and Cognition 42 2014 922 934 10.3758/s13421-014 0402-5 24554279
Janssen J. Oudman E. Irish M. Postma A. Exploring episodic and semantic contributions to past and future thinking performance in Korsakof’s syndrome Memory & Cognition 50 2022 630 640 10.3758/s13421-021-01262-2 35084717
Kessels R.C.P. Kopelman M.D. Context memory in KS’s Syndrome Neuropsychology Review 22 2012 117 131 10.1007/s11065-012-9202-5 22580849
Kopelman M.D. The Autobiographical Memory Interview (AMI) in organic and psychogenic amnesia Memory 2 1994 211 235 10.1080/09658219408258945 7584292
Kopelman M.D. Disorders of memory Brain 125 2002 2152 2190 10.1093/brain/awf229 12244076
Levine B. Svoboda E. Hay J.F. Wincour G. Moscovitch M. Aging and autobiographical memory: Dissociating episodic from semantic retrieval Psychology and Aging 17 2002 677 689 10.1037/0882-7974.17.4.677 12507363
Luchetti M. Sutin A.R. Age differences in autobiographical memory across the adult lifespan: Older adults report stronger phenomenology Memory 26 1 2017 117 130 10.1080/09658211.2017.1335326 28585461
Mori E. Ikeda M. Hirono N. Kitagaki H. Imamura T. Shimomura T. Amygdalar volume and emotional memory in Alzheimer’s Disease The American Journal of Psychiatry 156 1999 216 222 10.1176/ajp.156.2.216 9989557
Murphy K.J. Troyer A.K. Levine B. Moscovitch M. Episodic, but not semantic, autobiographical memory is reduced in amnestic mild cognitive impairment Neuropsychologia 46 13 2008 3116 3123 10.1016/j.neuropsychologia.2008.07.004 18675285
Rensen Y.C.M. Oudman E. Oosterman J.M. Kessels R.P.C. Confabulations in alcoholic Korsakoff’s Syndrome: A factor analysis of the Nijmegen-Venray Confabulation List Assessment 18 2020 1545 1555 10.1177/1073191119899476
Rubin D.C. Schrauf R.W. Greenberg D.L. Belief and recollection of autobiographical memories Memory and Cognition 31 6 2003 887 901 10.3758/BF03196443 14651297
Seo, E. H., Lee, D. Y., Kim, S. G., Kim, K. W., Kim, D. H., Kim, B. J., Kim, M., Kim, S. Y., Kim, Y. H., Kim, J., Kim, J. W., Moon, S. W., Park, J. H., Ryu, S., Yoon, J. C., Lee, N. J., Lee, C. U., J, J. H., Choo, I. H., & Woo, J. I. (2011). Validity of the telephone interview for cognitive status (TICS) and modified TICS (TICSm) for mild cognitive impairment (MCI) and dementia screening. Archives of Gerontology and Geriatrics, 52, 26-30. 10.1016/j.archger.2010.04.008.
Tulving E. Hayman C.A.G. Macdonald C.A. Long-lasting perceptual priming and semantic learning in amnesia: A case experiment Journal of Experimental Psychology: Learning, Memory, and Cognition 17 1991 595 617 10.1037/0278 -7393.17.4.595 1832430
Waring J.D. Seiger A.N. Solomon P.R. Budson A.E. Kesinger E.A. Memory for the 2008 presidential election in healthy ageing and mild cognitive impairment Cognition and Emotion 28 2014 1407 1421 10.1080/02699931.2014.886558 24533684
| 0 | PMC9742220 | NO-CC CODE | 2022-12-14 23:53:26 | no | Conscious Cogn. 2023 Jan 12; 107:103454 | utf-8 | Conscious Cogn | 2,022 | 10.1016/j.concog.2022.103454 | oa_other |
==== Front
Cytotherapy
Cytotherapy
Cytotherapy
1465-3249
1477-2566
Published by Elsevier Inc. on behalf of International Society for Cell & Gene Therapy.
S1465-3249(22)01053-2
10.1016/j.jcyt.2022.12.001
Full-Length Article
Donor selection for adoptive cell therapy with CD45RA− memory T cells for COVID-19 patients, and dexamethasone and interleukin-15 effect on the phenotype, proliferation and interferon gamma release
Al-Akioui-Sanz Karima 1
Pascual-Miguel Bárbara 1
Díaz-Almirón Mariana 2
Mestre-Durán Carmen 1
Navarro-Zapata Alfonso 1
Clares-Villa Laura 1
Martín-Cortázar Carla 1
Vicario José Luis 3
Moreno Miguel Ángel 3
Balas Antonio 3
De Paz Raquel 4
Minguillón Jordi 1
Pérez-Martínez Antonio *156
Ferreras Cristina *#1
1 IdiPAZ, Hospital La Paz Institute for Health Research, La Paz University Hospital, Madrid, Spain
2 Biostatistics Department, La Paz University Hospital, Madrid, Spain
3 Histocompatibility Unit. Transfusion Center of Madrid. Madrid, Spain
4 Cell Therapy Unit, Hematology Department, La Paz University Hospital, Madrid, Spain
5 Pediatric Hemato-oncology Department, La Paz University Hospital, Madrid, Spain
6 Faculty of Medicine Autonomous University of Madrid, Madrid, Spain
# Corresponding author:
⁎ These authors contributed equally
12 12 2022
12 12 2022
13 6 2022
6 12 2022
© 2022 Published by Elsevier Inc. on behalf of International Society for Cell & Gene Therapy.
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.
We have previously demonstrated the safety and feasibility of adoptive cell therapy with CD45RA− memory T cells containing SARS-CoV-2-specific T cells for COVID-19 patients from an unvaccinated donor who was chosen based on human leukocyte antigen compatibility and cellular response. In this study, we examined the durability of cellular and humoral immunity within CD45RA− memory T cells and the effect of dexamethasone, the current standard of care treatment and interleukin-15, a cytokine critically involved in T-cell maintenance and survival. We performed a longitudinal analysis from previously SARS-CoV-2 infected and infection-naïve individuals covering 21 months from infection and 10 months after full vaccination with the BNT162b2 Pfizer/BioNTech vaccine. We observed that cellular responses are maintained over time. Humoral responses increased after vaccination but were gradually lost. Our results suggest that the best donors for adoptive cell therapy would be recovered individuals and two months after vaccination although further studies with larger cohorts would be needed to confirm this finding. Besides, dexamethasone did not alter cell functionality or proliferation of CD45RA− T cells at a concentration used in the clinical practice, and interleukin-15 increased the memory T-cell activation state, regulatory T cell expression, and interferon gamma release.
Graphical Abstract
Image, graphical abstract
Keywords
CD45RA− memory T cells
cellular immunity
humoral immunity
COVID-19 treatment
==== Body
pmcIntroduction
A novel coronavirus named SARS-CoV-2 was identified in early 2020, causing a global coronavirus disease pandemic (COVID-19). Infected patients experience a wide range of symptoms from asymptomatic to severe. Although vaccines have dramatically decreased infection rates and the number of deaths and hospitalizations, they are not 100% effective and immunity is gradually lost. 1 Vaccines are also not completely effective at preventing coronavirus variants of concern (VOCs).2 , 3
Immune dysregulation is related to disease severity. The immune response includes humoral and cellular responses with an adaptive immune response providing long-term protection. Plasma cells are an essential part of humoral immunity, secreting neutralizing antibodies and blocking viruses from entering cells. Cellular immunity (such as the T-cell adaptive immune response) is needed to control and eliminate SARS-CoV-2 infection.[4], [5], [6] Antibodies can protect against infection and reduce disease severity during the first months post-vaccination or infection; however, memory T cells confer long-term protection even in cases where the humoral response is poor.7 , 8 Infected cells are eliminated by T cells, specifically cytotoxic T cells and T-helper cells, coordinate the long-term immune reaction, collaborating in creating long-living plasma cells.[9], [10], [11] Besides, T-cell responses to SARS-CoV-2 can confer protection to SARS-CoV-2 variants due to the T cells’ ability to recognize parts of the virus less susceptible to mutational pressure.4 , [12], [13], [14], [15], [16], [17] The entire immune system needs to respond to SARS-CoV-2 to efficiently eliminate the infection and avoid negative clinical outcomes.
Lymphopenia is a biomarker of disease severity, and profound lymphocytopenia correlates with severe clinical COVID-19 outcomes, while resolution of lymphopenia correlates with recovery.[18], [19], [20], [21], [22], [23] The mechanisms underlying lymphocytopenia are unclear, but impaired lymphocyte proliferation, lymphocyte apoptosis induction, bone marrow impairment and tissue redistribution have been suggested.4 , 20 , 24
We and other authors have hypothesized that passive adoptive cellular immunotherapy with various T-cell subsets and/or natural killer (NK) cells might provide an effective mechanism to control COVID-19 infection due to their antiviral properties.[25], [26], [27] We have previously shown how to detect, isolate and produce, at a large clinical scale, CD45RA− memory T cells containing SARS-CoV-2-specific T cells from COVID-19 convalescent donors.25 In the first-in-human clinical trial with adoptive cell therapy using allogeneic, off-the-shelf CD45RA− memory T cells from a convalescent donor, we demonstrated the safety and feasibility of treating moderate/severe COVID-19 hospitalized patients.26 The infusion of memory T cells to hospitalized patients with lymphopenia could help normalize T-cell counts and clear SARS-CoV-2 infection more efficiently than through the patient's immune system. We are currently assessing the efficacy of this treatment in a phase II clinical trial.
Several factors can alter the functionality of the infused CD45RA− memory T cells. The efficacy of adoptive T-cell therapy can be impaired by dexamethasone, a synthetic glucocorticoid widely used for treating human inflammatory diseases.28 A short course of dexamethasone is the current standard of care (SoC) treatment for COVID-19 hospitalized patients,29 and high doses of dexamethasone have been shown to have a negative impact on proliferation and cytokine production by T cells.[30], [31], [32]
In addition, we have previously shown that interleukin 15 (IL-15) which is critically involved in the maintenance and survival of memory and naïve CD8+ T cells and NK cells33 , 34, is producing an activating phenotype in the CD45RA− memory T cells of convalescent donors25 We hypothesize that the stimulation of CD45RA− T cells with IL-15 overnight (O/N) will maintain T-cell functionality even in the presence of dexamethasone, making this approach a potential advanced cell therapy for COVID-19 patients.
This study seeks to determine several aspects of CD45RA− memory T cells for its use as an adoptive therapy for COVID-19 patients. On one hand, we have studied the durability and level of cellular immunity within CD45RA− memory T cells and the changes with vaccination, virus exposure and time. We therefore, performed a longitudinal exploratory analysis of the SARS-CoV-2 specific humoral and cellular immunity within memory CD45RA− T cells in SARS-CoV-2 naive and previously infected individuals at different time points before and after two doses of the BNT162b2 BioNTech/Pfizer vaccine. On the other hand, we explored the effect of the current SoC treatment on the adoptive cell therapy based on CD45RA− memory T cells and how this therapy could be improved to increase efficiency. For that, we evaluated the effect of dexamethasone and IL-15 on the proliferation, phenotype and functionality of SARS-CoV-2-specific CD45RA− memory T cells.
Material and Methods
Study participants
The study included 6 COVID-19-recovered donors and 4 healthy controls (Table S1), all of whom were healthcare workers. All of the recovered donors were COVID-19-positive between March and April 2020 and had mild disease.25 All donors tested negative for SARS-CoV-2, and none of the recovered donors experienced reinfections. The healthy donors had not been exposed to COVID-19 patients and tested negative for anti-SARS-CoV-2 antibodies in June 2020. All participants granted their written consent, and the study was approved by the Hospital Institution Review Board (IRB number: 254/20). All participants were immunized with 2 doses of mRNA Pfizer vaccine between the end of January 2021 and the beginning of February 2021.
Cell processing
Paired plasma and peripheral blood mononuclear cells (PBMCs) samples were collected from all individuals at various time points, as well as buffy coats when required (Figure 1 ). Briefly, PBMCs were isolated from peripheral blood by density gradient centrifugation using Ficoll-Paque (GE Healthcare, Chicago, IL, United States). Cells were preserved in 90% fetal bovine serum (Sigma-Aldrich, MO) and 10% dimethyl sulfoxide and stored in liquid nitrogen. Plasma was isolated from the blood collection after centrifugation and stored at −80°C.Figure 1 Time point scheme for blood collection in recovered and control individuals. T1: A mean of 13 days after clearing COVID-19 infection, T2: 9 months after clearing COVID-19 infection, T3: 11 days after full BNT162b2 Pfizer/BioNTech vaccination T4: 65 days after full BNT162b2 Pfizer/BioNTech vaccination, T5: 9-10 months after full BNT162b2 Pfizer/BioNTech vaccination
Figure 1
Proliferation assay of SARS-CoV-2-specific memory T cells in the presence of dexamethasone and IL-15
PBMCs from buffy coats were obtained from the Centro de Transfusion of Community of Madrid. The SARS-COV-2 infection or vaccination status was not assessed. PBMCs from buffy coats were thawed and then incubated with or without IL-15 (50 ng/mL Miltenyi Biotec, Germany) overnight at 37°C in TexMACS™ Medium (Miltenyi Biotec, Germany) supplemented with 10% Human Serum Type AB (Sigma Aldrich) and 100 U/mL penicillin - 100 μg/mL streptomycin sulphate (Sigma Aldrich). The cells were then washed and approximately 2 million cells per condition were resuspended in 1 mL of supplemented medium and 100 µL of a CFSE stock dilution (final concentration 5 µM) (CFSE, CellTrace™ CFSE Cell Proliferation Kit, Invitrogen). The cells were then incubated for 10 min in the dark and washed 3 times. Lastly, the cells were cultured in a P24 plate in 1 mL of supplemented medium with 1.5% v/v PHA (Phytohemagglutinin M form, Thermofisher) plus 0.05 µg/mL of mouse anti-human CD3 (Clone OKT3; BD Biosciences) and 5 µg/mL of CD28/CD49d Purified (Clone L293 L25 RUO GMP,BD Biosciences). In addition, various concentrations of dexamethasone (Kern Pharma 4 mg/mL solution for injection EFG) (0, 10−7 M, 10−6 M and 10−5 M) were added to the supplemented medium. PBMCs without IL-15 and without dexamethasone were employed as the negative control (Figure S1). The dexamethasone doses were calculated considering that the patients were administered 6 mg/day of the glucocorticoid, and the volume of the drug distribution was calculated according to DrugBank's online dexamethasone information page (https://go.drugbank.com/drugs/DB01234). According to this data and previous publications35 , 36, 6 mg of dexamethasone administered to an adult of approximately 60–70 kg corresponds to 1–2 × 10−7 M.
The cells were incubated at 37°C for 72 h, and a proliferation assay was conducted to determine whether high doses of the drug decrease their division or proliferation index. In addition to CFSE, cells were stained with the following cell surface antibodies to define different subpopulations: CD27 APC (BD Pharmingen), CD3 Viogreen (Miltenyi Biotec), CD4 PE-Cy7 (BD Pharmingen), CD8 APC-Cy7 (BD Pharmingen), L/D 7AAD (BD Pharmingen), CD45RA Alexa Fluor 700 (BD Pharmingen), CD127 PE-CF594 (BD Horizon) and CD25 BV421 (BD Horizon). Cell acquisition was then performed using a Navios cytometer (Beckman Coulter), acquiring a mean of 200,000 cells. The analysis was performed using FlowJo 10.7.1 software (FlowJo LLC).
Proliferation cells’ capacity under different conditions was compared using two parameters: the proliferation index, which is the total number of divisions divided by the number of cells that went into division; and the relative geometric mean fluorescent intensity (GeoMFI), expressed as ratio GeoMFI /(GeoMFI of the negative control) (cells without dexamethasone and without IL-15).
Detection of SARS-CoV-2-specific memory T cells by interferon-gamma assay
The assay was performed as previously described.25 Briefly, the peptide pools were short 15-mer peptide pools with 11 amino acid overlaps that can bind MHC class I and II complexes and are therefore able to stimulate CD4+ and CD8+ T cells. The peptide pools cover the immunodominant sequence domains of the surface glycoprotein S, the complete sequence of the nucleocapsid phosphoprotein N and the membrane glycoprotein M (GenBank MN908947.3, Protein QHD43416.1, Protein QHD43423.2, Protein QHD43419.1; Miltenyi Biotec, Germany). The cells were rested O/N at 37°C and after 5 h stimulation with individual SARS-CoV-2 peptide pools or combination of peptide pools (M, N, S), the cells were labeled with interferon-gamma (IFN-γ) catch reagent (Human IFN-γ Secretion Assay-Detection Kit, Miltenyi Biotec), and the cell surface-bound IFN-γ was targeted using the IFN-γ PE antibody. Background subtraction was performed from parallel unstimulated cultures. A positive sample included the following acceptance criteria: 0.1% of IFN-γ+ cells out of the total cell population with a minimum of 150,000 events analyzed, at least twice the number of IFN-γ+ cells in the sample than in the negative control, and a positive control based on plate-bound cells stimulated with mouse anti-human CD3 and co-stimulated with purified CD28/CD49d. Basal IFN-γ production by PBMCs was included as a background control in the absence of stimulation and co-stimulation.
The experiments with and without IL-15 incubation O/N and with different concentrations of dexamethasone were performed as described above. After incubation, cells where stained using the following fluorochrome-conjugated anti-human surface antibodies: CD45RA FITC, CD27 APC, CD3 VioGreen, CD4 PECy7, CD8 APC Cy7, L/D 7AAD and IFN-γ PE. Cell acquisition was then performed using a Navios cytometer (Beckman Coulter), acquiring a mean of 200,000 cells. The analysis was performed using FlowJo 10.7.1 software (FlowJo LLC).
Phenotype of memory T cells containing SARS-CoV-2-specific T cells determined by flow cytometry assay
The phenotype assay was performed as previously described.25 Briefly, we stained the cell surface for 20 min at 4°C using the following fluorochrome-conjugated antihuman antibodies: CD45RA FITC, CD27 APC, CD3 VioGreen, CD4 PECy7, CD8 APC Cy7 and L/D 7AAD. We employed other antibodies for specific cell populations: CD25 BV421 (BD Horizon) and CD127 PE-CF594 (BD Horizon) for regulatory T cells (Treg); HLA-DR BV421 (BD Pharmingen), CD69 PE (Miltenyi Biotec) and CD25 BV421 (BD Horizon) for activation makers; CD279 (PD1) AF700 (BioLegend) and NKG2A BV421 (BD Optibuild) for exhaustion markers; and, CD103 BV421 (BD Horizon) and CCR7 PE-CF594 (BD Horizon) for chemokine receptor and integrin markers.
The experiments with and without IL-15 incubation O/N and with different concentrations of dexamethasone were performed as described above. After the staining, cell acquisition was performed using a Navios cytometer (Beckman Coulter), acquiring a mean of 200,000 cells. The analysis was performed using FlowJo 10.7.1 software (FlowJo LLC).
Analysis of humoral responses
For the detection of the receptor-binding domain (RBD), spike 1 (S1), spike 2 (S2) and nucleocapsid (N) antibodies, we used the BioPlex 2200 SARS-CoV-2 IgG panel. Briefly, K3 EDTA plasma samples from patients were mixed with fluoromagnetic dyed beads, each coated independently with the 4 different antigens, and processed in the BioPlex 2200 system, as per the manufacturer's instructions (Bio-Rad Laboratories, CA, USA). Samples were first tested undiluted and, for results above 100 U/mL, serially diluted to determine the exact antibody level. Final results are shown in units (World Health Organization [WHO] BAU/mL) by transforming U/mL with the first WHO international standard for binding activity of anti-SARS-CoV-2 immunoglobulins to RBD, S1 and nucleocapsid proteins.
We also employed a test designed to mimic plasma neutralization for the interaction of RBD SARS-CoV-2 and human angiotensin-converting enzyme (ACE2) by competitive chemiluminescence immunoassay (Snibe, Shenzhen, China). Briefly, donor samples, buffer, magnetic microbeads coated with hACE2, and SARS-CoV-2-RBD ABEI-labelled antigen were mixed. In the incubation, immunoglobulins of the sample compete with RBD-ACE2 binding, and the RBD label is recognized by chemiluminescence reagents after applying a magnetic device and washing process. All steps were performed in a Maglumi 2000 system. Samples were first tested undiluted and, for results above 30 µg/mL, serially diluted to determine the exact neutralizing capacity. The final results are shown in units (WHO IU/mL) by transforming µg/mL with the WHO standard (20/136).
Statistical analysis
Given the nature of the correlated data, lymphocyte averages were estimated using least squares means with a linear mixed-effects model employing the restricted maximum likelihood method. Time, controls, recovered donors, and the interaction were included as the fixed effect, and the intercepts were included as the random effect. The least squares means were estimated and compared, and the Bonferroni adjustment was considered for multiple comparisons. The global effect of each variable in the model was assessed with the type 3 test. All statistical tests were considered bilateral, with statistically significant values having a p<0.05. The data was analyzed with the statistical program SAS Enterprise Guide 8.2 (SAS Institute Inc., Cary, NC, USA).
To analyze the differences in proliferation, functionality (IFN-γ production) and phenotype of PBMCs due to exposure to increasing concentrations of dexamethasone in different cell subpopulations, we performed a one-way or two-way ANOVA test. To analyze the differences due to the absence or presence of IL-15, we performed a Wilcoxon matches-pairs signed rank test for CFSE experiments and a Friedman Test with paired data with Dunn's multiple comparisons test in all other cases. We employed GraphPad Prism (GraphPad v.8.0 Software, San Diego, CA, United States) for the analysis. P-values <0.05 were considered statistically significant. The quantitative variables are expressed as mean ± standard error, while the qualitative variables are expressed as percentages (%). For the exploratory analysis, Spearman's correlation coefficient was calculated to evaluate the linear association between quantitative variables; ±0.8-1 correlation values were considered a strong association.
RESULTS
Participants and sample collection
Ten participants were administered 2 doses of the BNT162b2 Pfizer/BioNTech vaccine between January 10 and February 9, 2021 at La Paz University Hospital, Madrid (Spain). Six individuals were previously infected,25 and 4 were defined as SARS-CoV-2 naïve (Table S1). The mean age was 37 years (range 23–41 years) for the recovered donors and 29.7 years (range 26–34 years) for the infection-naïve controls. Seven donors were women and 3 were men. We obtained PBMCs at 5 time points, a mean of 13 days after the COVID-19 recovered individuals tested negative (by PCR) for SARS-CoV-2 (T1); between 9 and 10 months after T1, which was just before the first dose of the vaccine (T2); 11 days after the second dose of the mRNA vaccine (T3), 65 days after the second dose of the mRNA vaccine (T4), and between 9 and 10 months after the second dose (T5) (Figure 1). The data at T1 for all participants have been already published by Ferreras C et al 25. For one of the control donors, we collected data only at time points 1 and 2 because the donor was infected after the first dose of the vaccine. For one of the recovered donors, we did not collect data at the last time point (T5) because the donor was taking anakinra to treat myocarditis.
Quantification of memory T cells specific for SARS-CoV-2 immunity
We first analyzed the changes in the percentage of T-cell subsets at the various time points and the presence of SARS-CoV-2-specific T cells within the CD45RA− memory T-cell subpopulation by the release of IFN-γ upon in vitro stimulation of PBMCs to the individual or combination of peptide pools (Figure S2). As expected, there were no changes in the various T-cell subpopulations (Table S2). In general, there were no statistically significant differences in cellular response (Figure 2 , Table S2, and Figure S3).Figure 2 Detection of SARS-CoV-2 specific T cells within the CD45RA- memory T cells and subsets. IFN-g was measured by flow cytometry after exposure to the single SARS-CoV-2 peptides (M, N, and S) and the peptide pools (PepX3). 1,2,3,4, and 5 correspond to the different time points described in Figure 1. Independent data for each naïve control is shown by the blue dots and independent data for each recovered donor is shown by the red dots. Mean and SEM are shown. *p<0.05
Figure 2
For the N single SARS-CoV-2 peptide in the recovered donors, the highest cellular immune response occurred after COVID-19 infection. In the CD45RA− T-cell population, the mean IFN-γ expression decreased from T2 to T3 (p=0.0382) and from the T2 to T4 time points (p= 0.0311). There were no statistically significant differences in the other CD45RA− subsets: CD4+ and CD8+, central memory T cells (TCM) (CD4+CD27+ and CD8+CD27+) and effector memory T cells (TEM) (CD4+CD27− and CD8+CD27−) subpopulations (Figure 2 and Figure S3).
For the M and S SARS-CoV-2 individual peptide pools or the combination of peptide pools, SARS-CoV-2-specific memory T-cell responses were maintained over time. Interestingly, there was no increase in cellular responses after full mRNA immunization in the recovered individuals.
We detected no cellular immune response in the control participants before the vaccination (Figure 2 , Table S2 and Figure S3). However, we observed induction of SARS-CoV-2 T-cell responses to the S peptide pool at an early time point after mRNA immunization (T3), and the response was maintained at time points T4 and T5 within all CD45RA− subpopulations (Figure 2 , and Figure S3). Although the sample size was small, this was expected because the S protein is the target for the mRNA vaccine. We also observed that mRNA vaccination of COVID-19-naïve participants induced an immune cellular response to the S peptide pools similar to the memory T-cell responses from recovered donors at the same time point. Surprisingly, infection-naïve individuals showed cellular responses to the M and N SARS-COV-2 peptide pools 10 months after the second dose of the mRNA vaccine.
When examining both groups, we observed a tendency toward higher cellular immune responses in the recovered individuals than in the control group (Figure 2 and Figure S3).
In summary, memory T-cell responses tended to be higher in the recovered individuals at all time points due to a pre-existing memory T-cell population formed after SARS-CoV-2 infection, this cellular response that was maintained even 10 months after mRNA immunization in both groups (recovered and infection-naïve donors). In addition, COVID-19 infection produced memory T-cells responses to the combination of SARS-CoV-2 peptide pools; while in our control group, the vaccine was able to only produce Spike-specific SARS-CoV-2 memory T-cell responses.
Humoral response
We analyzed the antibody titers to nucleocapsid (N) representing prior SARS-CoV-2 infection, RBD, S1 and S2 domains of the Spike, representing a response to either previous infection or vaccination by mRNA vaccine, and levels of the neutralizing antibody (NAB). We observed antibody production in both groups (recovered and control donors) (Figure 3 ). As expected, antibody production was observed in previously SARS-CoV-2-infected individuals at T1 but was not detected in COVID-19-naïve participants.In the recovered individuals, vaccination dramatically increased antibody titers for RBD, S1, S2 and NAB. For the N antibody, the highest value was obtained after COVID-19 infection (T1), decreasing at T2 (Figure 3 A). Vaccination did not boost the humoral immune response, with values decreasing at T3, T4, and T5 with respect to T1. There was also a significant decrease from T4 to T5, with values below T1.
For the other antibodies tested (RBD, S1, S2, and NAB), we observed vaccine-induced antibody responses with a markedly increased antibody production at T3 that gradually decreased at T4 and T5. For S2, the values at T5 were similar to the pre-vaccination values (T1) (Figures 3 B, 3 C, 3 D, and 3 E). In the COVID-19-naïve participants, the vaccine did not elicit IgG N antibody responses as expected at any tested time point (T3, T4, and T5). For RBD, S1, S2, and NAB, early antibody production was observed at T3, with the response decreasing over time.
Values after vaccination (T3–T5) were significantly higher in the recovered individuals than in the infection-naïve controls at all time points, with a decrease in antibody titers over time. Interestingly, the antibody titers for the control individuals after full immunization (T3 and T4) seemed to be much higher than in the recovered donors after COVID-19 infection (T1 and T2; RBD p<0.0001; S1 p<0.0001; S2 p=0.0175; NAB p<0.0001) (Figure 3 B, 3 C, 3 D, and 3 E).
Figure 3 Antibody titers for A) Nucleocapsid, B) RBD, C) S1, D) S2 and E) NAB at different time points in recovered and control individuals. 1,2,3,4, and 5 correspond to the different time points described in Figure 1. Independent data for each naïve control is shown by the blue dots and independent data for each recovered donor is shown by the red dots. p values different from p<0.0001 are shown in the figure. *p value in recovered donors, *p value in control donors, *p value between both groups. *p<0.05; **p<0.01; ***p <0.001; **** p<0.0001. Mean and SEM are shown
Figure 3
These results showed that the BNT162b2 mRNA vaccine induced significantly higher antibody titers in the recovered donors compared with the infection-naïve individuals at all time points. However, those values decreased gradually over time in both groups, with values at 10 months after immunization resembling pre-vaccination values.
Correlation between humoral and cellular response in the recovered donors
To assess the correlation between cellular and humoral immunity over time, we performed an exploratory analysis of non-parametric Spearman's rho correlation of humoral and cellular responses in the recovered donors. We compared the humoral parameters in plasma with the cellular response to the combination of SARS-CoV-2 peptide pools (Figure 4 ) at the different time points.Figure 4 Cellular and humoral immunity correlation responses in recovered donors. The nonparametric Spearman test was used for correlation analysis. A matrix heatmap shows the correlation between cellular and immune humoral responses. Cellular responses were obtained upon SARS-CoV-2 peptide pools stimulation. The Spearman's correlation matrices show the relationship between N, RBD, S1, S2 and NAB antibodies and the different CD45RA- cell subsets. T1, T2, T3, T4, and T5 correspond to the time points shown in Figure 1. Values inside the squares indicate the correlation coefficient. All data points are shown. CM: Central Memory, EM: Effector Memory. Blue color means a positive correlation. Red color means a negative correlation. The intensity of the color indicates how strong this correlation is.
Figure 4
At T1, after the COVID-19 infection was cleared, our analysis showed a pattern of positive cellular and humoral correlation in most of the CD45RA− subsets for N, RBD, S1 and NAB antibodies. In contrast, the data 9 months after the infection and early after vaccination (T2 and T3) displayed a negative cellular and humoral correlation. However, at T4 a tendency to a positive correlation was again observed for all the studied antibodies that although less pronounced was maintained at T5. Interestingly, we observed a similar pattern at T5 to the one after COVID-19 infection (Figure 4). These data suggest that after SARS-CoV-2 infection antigen-specific T-cell responses induced upon peptide pool stimulation correlated with humoral responses. This positive correlation is lost over time and is only recovered and maintained long term after a complete vaccination regimen with BNT162b2 vaccination.
Effects of dexamethasone and IL-15 on SARS-CoV-2-specific CD45RA− memory T cells
We analyzed the effect of dexamethasone on the CD45RA− memory T-cell subsets and the proliferation, functionality and phenotype of SARS-CoV-2-specific CD45RA− memory T cells.
No significant differences were observed in proliferation at the studied dexamethasone concentrations in the absence or presence of IL-15 in the live cell's gate or the CD45RA− memory T cells (Figure 5 A and 5 B). Also, the geometric mean fluorescent intensity did not show any differences (Figure 5 C).Figure 5 Effect of dexamethasone and IL-15 on total PBMCs and CD45RA− memory T cells proliferation. PBMCs were cultured in presence or absence of IL15 (50 ng/ml) O/N. Then cells were washed and labeled with CFSE and incubated for 72h with different concentrations of dexamethasone (0, 10−7 M DEX, 10−6 M DEX and 10−5 M) for 72 h. A) Proliferation Index of total live cells and CD45RA− memory T cells. The PI is calculated as total number of divisions divided by the number of cells that went into division. One-way Annova was performed to analyze differences due to different concentrations of dexamethasone previously cultured with IL-15 O/N. Mean +/- SEM, N=5-9 independent donors. B) Representative determination of live cells and CD45RA− memory T cells proliferation by flow cytometry with or without IL-15 and in presence of increasing concentrations of dexamethasone C) Geometric mean fluorescence intensity (GeoMFI), expressed as ratio GeoMFI /(GeoMFI of the negative control) (cells without dexamethasone and without IL-15). Mean +/- SEM of the ratios are represented.
Figure 5:
We observed no changes in the percentage of each T-cell subset in the presence of dexamethasone after an O/N or 72-h incubation regardless a IL-15 pre-incubation (Table 1 A and 1 B). In the absence of a previous IL-15 incubation, an O/N incubation with dexamethasone did not change the frequency of Treg and only the highest concentration showed a tendency to decrease Treg expression (11.3 vs 9.8, 8.2, and 6.9) (Table 1 A). This effect, although no statistically significant, was more pronounce after 72 h incubation with dexamethasone (19.3 vs 11, 6.4, 3.3) (Table 1 B). On the contrary, in all conditions studied, IL-15 increased the expression of Treg and dexamethasone did not have any effect with values similar to the ones obtained with no dexamethasone (Table 1 A and 1 B). These changes in Treg expression are due to an activation rather than an increase in cell proliferation, since no changes were observed in proliferation index (data not shown). In general, at 72 h the cells previously exposed to IL-15 showed a lower Treg induction regardless the incubation with dexamethasone comparing with the same dexamethasone concentration after an O/N incubation. Together these findings indicate that IL-15 alone could increase the activation state of Tregs, while dexamethasone had no effect on this activation.Table 1 Effect of dexamethasone on the percentage of CD45RA− memory T cell subpopulations after an A) O/N and B) 72 h incubation with PBMCs previously treated O/N with and without IL-15. CM: central memory, EM: effector memory Treg: regulatory T-cells. The experiment has been designed following the scheme in Figure S1. ** p˂0.005. ***p˂0.0001. Mean and SEM is shown. N=3-4
Table 1A
WO DEX O/N 10−7 M O/N 10−6 M O/N 10−5 M O/N
-IL15 +IL15 -IL15 +IL15 -IL15 +IL15 -IL15 +IL15
CD45RA− % CD45RA− 25.2 18.6 30.1 21.6 30.8 21.4 33.0 28.1
% CD3+ 88.8 90.3 89.8 88.0 89.2 86.8 90.5 86.4
% CD4+ 76.8 79.2 75.8 81.3 77.5 82.6 78.6 84.1
% CD8+ 17.1 19.5 18.5 17.4 16.9 16.1 15.3 14.6
CD45RA−CD4+ CD27+ (CM) 80.4 80.5 80.9 77.7 80.4 78.4 78.0 74.7
CD27−(EM) 18.7 19.5 18.5 22.3 21.8 21.5 20.2 25.3
CD127lowCD25+ (Treg) 11.3 40.7*** 9.8 42.0** 8.2 40.1*** 6.9 36.1***
CD45RA−CD8+ CD27+ (CM) 74.1 74.2 75.3 72.8 75.6 73.7 69.4 72.3
CD27−(EM) 24.8 25.8 23.8 27.2 23.8 26.3 24.2 27.7
B
WO DEX 72H 10−7 M 72H 10−6 M 72H 10−5 M 72H
-IL15 +IL15 -IL15 +IL15 -IL15 +IL15 -IL15 +IL15
CD45RA− % CD45RA− 19.1 20.1 21.8 22.8 23.5 23.4 23.5 25.3
% CD3+ 95.7 93.5 85.1 92.1 91.9 90.9 91.8 91.8
% CD4+ 88.7 77.5 84.0 79.7 85.2 78.5 84.5 80.2
% CD8+ 10.4 20.9 12.9 19.0 12.9 20.3 13.2 18.6
CD45RA−CD4+ CD27+ (CM) 71.0 75.7 73.0 76.9 76.9 82.0 76.9 81.2
CD27−(EM) 29.0 24.3 27.0 23.1 23.1 17.9 23.1 18.8
CD127lowCD25+ (Treg) 19.3 26.4 11.0 25.9*** 6.4 27.9*** 3.3 27.7***
CD45RA−CD8+ CD27+ (CM) 82.5 72.9 79.7 74.6 79.7 76.6 79.2 75.7
CD27−(EM) 17.5 26.9 20.3 25.4 20.3 23.4 20.8 24.3
We then analyzed IFN-γ production after exposure to the combination of SARS-CoV-2 peptide pools in 3 separate situations: first, IFN-γ production after an O/N incubation of IL-15; second, IFN-γ production after an O/N incubation with dexamethasone; and third, the effect of dexamethasone on IFN-γ production after with a previous incubation with IL-15 (Figure 6 ). Figure 6 A shows that when the cells were previously incubated with IL-15 in the absence of dexamethasone there was a tendency to increase the percentage of IFN-γ in all CD45RA− memory T-cell subsets studied (Figure 6 A black and grey bars). After an O/N incubation with dexamethasone, we observed no changes in IFN-γ release in the different T-cell subpopulations at the dexamethasone concentrations resembling those in clinical practice (Figure 6 A red bars). Although higher concentrations tended to decrease IFN-γ values, there were no statistically significant differences (Figure 6 A dark green and dark blue bars). When the cells were cultured with dexamethasone but previously exposed to IL-15, we observed IFN-γ values similar to the values with no dexamethasone and no previous IL-15 pre-incubation (Figure 6 A light bars: red, green and blue).Figure 6 Effect of IL-15 and dexamethasone on the IFN-g production after co-culture with the combination of the SARS-CoV-2 peptide pools. IFN-g production of SARS-CoV-2 specific CD45RA− memory T cells after co-culture with the combination of peptide pools in the presence or absence of IL-15 O/N and A) dexamethasone O/N or B) for 72 h. Several subsets are represented: CD45RA−CD3+, CD45RA−CD3+CD8+, CD45RA−CD3+CD4+. Black bars represent cells cultured in the absence of dexamethasone, dark color bars represent cells incubated with the different concentrations of dexamethasone and no IL-15 previous incubation, light color bars represent cells cultures with the different concentrations of dexamethasone previously exposed to IL-15. A Two-way ANOVA test was performed *p<0.05; **p<0.01; ***p <0.001; **** p<0.0001 N=4 . All donors had a previous COVID-19 infection. Mean and ±SEM.
Figure 6
We then analyzed the same effect after a 72 h incubation. Figure 6 B shows a statistically significant increase in IFN-γ values when the cells were previously incubated with IL-15 in the absence of dexamethasone (Figure 6 B black and grey bars). At 72 h incubation with dexamethasone alone tended to decrease IFN-γ release at all the concentrations used and in all the CD45RA− memory T-cell subsets, and this effect was more pronounced at higher concentrations (Figure 6 B dark bars: red, green and blue), specifically in the CD45RA−CD3+CD8+ subset at 10−6 and 10−5 M (p= 0.01). When the cells were cultured with dexamethasone for 72 h and previously exposed to IL-15, we observed IFN-γ values comparable to the values with no dexamethasone and no previous IL-15 pre-incubation, similar to data for an O/N incubation (Figure 6 B light red, green and blue).
These results suggest that incubation of memory T cells with dexamethasone concentrations used in clinical practice for hospitalized COVID-19 patients did not affect the functionality of SARS-CoV-2-specific CD45RA− memory T cells based on IFN-γ release. Moreover, a previous incubation of CD45RA− memory T cells with IL-15 counteracted the effect of higher dexamethasone concentrations when co-cultured with the cells for 72 h.
We then performed a Friedman test to analyze the global effect of IL-15 regardless of the dexamethasone concentration at both time points, showing that the increased IFN-γ release was due to IL-15 incubation. We observed statistically significant differences in the following subsets: CD45RA− (p=0.0068 O/N, p= 0.013 at 72 h), CD45RA−CD3+ (p=0.0152 O/N, p=0.024 at 72 h) and CD45RA−CD3+CD4+ (p=0.033 O/N, p=0.019 at 72 h), with no differences in the CD45RA−CD3+CD8+ subsets.
Lastly, we analyzed the changes in phenotype induced by dexamethasone and IL-15. We studied a panel of activation, exhaustion, and homing cell surface markers (Table S3 and Figure S4, S5 and S6).
The cells cultured with IL-15 showed a 2.2-fold and 5.0-fold increase in the activation markers CD69 and CD25high, respectively, in the CD45RA− memory T-cell subpopulation. A dexamethasone O/N incubation did not have any effect on this activation (Figure S4A). However, the CD25high expression showed a tendency to decrease after 72 h with dexamethasone dose-dependently in the absence of a previous incubation with IL-15 (Figure S4B). Also, no changes were observed in the presence of dexamethasone after O/N or 72 h incubation in the activation marker HLA-DR, exhaustion (NKG2A, PD1) or homing markers (CD103, CCR7) (Figures S4, S5, S6, and Table S3). To analyze the global effect of IL-15, we performed a Friedman test and observed statistically significant differences for the activation markers HLA-DR, CD25high and CD69 in the CD45RA−, CD45RA−CD3+, CD45RA−CD3+CD4+, CD45RA−CD3+CD8+ cell subsets after an O/N incubation with dexamethasone and in CD45RA−CD3+, CD45RA−CD3+CD4+, CD45RA−CD3+CD8+, after a 72 h incubation with dexamethasone (Table S4). There were also no significant changes in the percentage of the different various T-cell subsets (data not shown). In general, IL-15 increased the activation state of cells, while the exhaustion or homing markers remain unchanged. Dexamethasone did not alter the expression of these markers on the CD45RA− cell surface subsets.
Discussion
Vaccination has dramatically decreased the number of deaths and hospitalizations due to SARS-CoV-2 infection; however, specific and effective therapeutic anti-viral therapies for COVID-19 are lacking. New COVID-19 pandemic waves and the emergence of SARS-CoV-2 VOCs with greater transmissibility risk, mortality rates, and ability to evade previously acquired immunity37 have shown that the pandemic has not yet ended and will be around for a long time.
There have been two very well-differentiated time points in the SARS-CoV-2 pandemic: before and after vaccination, with widely studied cellular and humoral responses.4 , 15 We and other authors have shown the antiviral properties of CD45RA− memory T cells in hematopoietic stem cell transplantation and COVID-19 settings.25 , [38], [39], [40], [41] We have demonstrated the safety and feasibility of adoptive cell therapy using CD45RA− memory T cells containing SARS-CoV-2 specific T cells in hospitalized COVID-19 patients. In this approach, we pursued to increase the pool of lymphocytes in patients with lymphopenia, exerting an antiviral effect on the coronavirus and protecting the patient from other viruses that the donor encountered.25 In this trial, the patients were administered memory T cells from a convalescent and unvaccinated donor,26 who was chosen based on human leukocyte antigen compatibility and cellular response to SARS-CoV-2 peptide pools. Based on the results presented here, the donor's CD45RA− memory T cells had a cellular response, and a positive cellular and humoral correlation; however, the antibody titers were low because the donor had not been administered any vaccine (T1).
The present study sought to determine on the one hand, who would be the best donor candidate for adoptive cell therapy for COVID-19 patients in the current scenario of infections and vaccinations. To this end, we analyzed the SARS-CoV-2-specific T-cell responses within the CD45RA− memory T-cell subpopulation and subsets at different time points, covering a 21 months after SARS-CoV-2 infection in the recovered donors and between 9 and 10 months after a full BNT162b2 BioNTech/Pfizer vaccination in both groups. We also analyzed the changes in humoral response by measuring antibody titers for 5 specific antibodies. Besides we studied the effect of dexamethasone and the T cell activator IL-15 on these cells to analyze the effect of these compounds on the living drug.
Our data show a tendency toward a higher cellular response in recovered individuals at all time points that is maintained over time. Humoral responses are boosted after vaccination, with higher titers in the recovered individuals than in the controls; however, these responses are gradually lost.
T-cell responses against SARS-CoV-2 have been extensively studied, showing responses against almost all proteins of the viral proteome,17 , 42 with spike-specific T-cell responses dominated by CD4+ T cells.4 Our product (CD45RA−) is mainly composed of CD4+ T cells.25 Several authors have shown the effects of the first and second immunization on cellular and humoral responses and that CD4+ T cells are primed by the vaccine.5 CD4+ T cells are required to promote B-cell antibody production, enhance and maintain the responses of CD8+ T cells, orchestrate immune responses against a wide variety of pathogenic microorganisms, and contribute to viral elimination by a direct cytotoxic effect on virus-infected cells.[43], [44], [45], [46], [47] In convalescent donors, SARS-CoV-2-specific CD4+ T cells produce a Th1 response, include cells with lymphoid and tissue-homing potential, are long-lived, and are capable of proliferation.48 CD45RA− memory T cells are composed of TCM or TEM cell subpopulations. The TCM cells can home in secondary lymphoid tissues, proliferate and create a new round of effector T cells49, whereas TEM subsets are the first responders to infection, with a quick and strong response to pathogens and can home in on peripheral lymphoid tissues.50 Unlike a number of authors, we observed no difference in the CD4+ TEM and TCM responses after vaccination, although this could be due to our small sample size.
Current vaccines are targeting the S protein because it has shown high antigenicity and the ability to induce robust immune responses.51 , 52 Our data are in agreement with those indicating that the mRNA vaccine generates Spike-specific memory CD4+ and CD8+ T-cell responses.5 , [53], [54], [55] In COVID-19-naïve individuals, we observe Spike-specific CD45RA− memory T-cell responses as early as 10 days after the second dose, as has been reported in higher cohorts.56 , 57 Recovered donors retained SARS-CoV-2 specific T cells within the CD45RA− memory T cells that recognize different parts of the coronavirus, the M, N, and S peptide pools. In line with previous publications, our data suggest that vaccine immunization does not increase the cellular responses in recovered individuals, and that pre-existing immunity due to infection is maintained over time.58
Antibody weaning has been observed 10 months after infection in non-immunized individuals59 , 60 and after mRNA immunization. As others have published we have observed that antibody levels peak after immunization in previously SARS-CoV-2-infected individuals58 , 61 , 62; after that, antibody production decrease dramatically over time,60 , [63], [64], [65], [66] being more profound in infection-naïve donors than in recovered individuals. Consistent with published reports, we observe no N antibody titers for SARS-CoV-2-naïve participants, the level for recovered donors peak after infection, and the values are not increased by immunization with mRNA.[67], [68], [69]
A number of COVID-19-naïve individuals show cellular immunity for the N and S peptides at T5. Although this might be consistent with a breakthrough infection, we would expect an increase in antibody titers after infection, which is not the case here. One possible explanation is that these individuals have been infected with the Omicron variant but did not experience any coronavirus-related symptoms.68 , 70
Recent studies have reported that the Omicron variant can evade SARS-CoV-2-specific and neutralizing antibodies.[71], [72], [73]
T-cell responses elicit a similar response to SARS-CoV-2 VOCs either by infection or vaccination13 , 74 and that most T-cell epitopes are not affected by mutations,17 making SARS-CoV-2 evade cellular immunity unlikely.
Nevertheless, the antibody response is lower for certain variants including Omicron; considering that antibody titers decrease over time, they might not be protective for infections.13 , 68 , 74 In this scenario, protection might be conferred by cellular immunity due to a pre-existing memory of VOCs conferred after SARS-CoV-2 infection75 but further experiments would confirm this hypothesis.
Coordination between humoral and cellular immune responses is necessary to eliminate SARS-CoV-2 infection and it is related to milder disease.[76], [77], [78] , 5 Our study suggests that a positive correlation of immune responses is achieved after COVID-19 infection and at least 2 months after 2 doses of mRNA vaccine, with higher levels of antibodies correlating with higher levels of memory T cells responding to the infection. We can hypothesize that recurrent immunizations will maintain this pattern, although further studies are needed.
We have also addressed how current treatment can affect adoptive cell therapy with CD45RA− memory T cells. We have found that dexamethasone does not affect the proliferation, phenotype and functionality of theCD45RA− memory T cells. The previous incubation with IL-15 positively affects the release of IFN-γ when the cells are co-cultured with the combination of peptide pools without increasing cell exhaustion.
In addition, we observe an increase in the activation markers after IL-15 cell incubation that is maintained in the presence of dexamethasone. It is already known that IL-15 is a Treg inducer.79 We have detected an increase in Treg induction with IL-15 but more importantly that induction is maintained in the presence of dexamethasone. This increase is likely not due to an increase in the percentage of Tregs but rather to their activation state, given that we observe no increase in the proliferation index. Further studies on FoxP380 and CTLA-4 expression, along with certain functionality assays, would help confirm our results. In COVID-19 patients, a decrease in the number of Tregs has been associated with a poorer prognosis; therefore, keeping T cell numbers and better functionality for the existing Tregs would result in a better prognosis.81 , 82
In summary, our results suggest that the best donors for adoptive cell therapy for COVID-19 patients would be immunized individuals who recovered from COVID-19 with mild disease and ideally 2 months after immunization. We show that the cellular response is maintained over a year post-SARS-CoV-2 infection and 2 months after full immunization, and that it seems there is a positive cellular and humoral correlation, with high antibody titers (T4). Also, dexamethasone does not affect the proliferation, phenotype and functionality of CD45RA− memory T cells at the concentration resembling the one employed in clinical practice for these patients, with IL-15 showing a positive effect on SARS-CoV2-specific CD45RA− T cells phenotype and IFN-γ release. These data are supported with the results of in our phase I clinical trial and the preliminary phase II trial (unpublished data), where donor microchimerism was detected for at least 3 weeks26. Our data indicate that the development of a biobank of living drugs with adoptive cell therapy is feasible as a treatment strategy for COVID-19 patients and future viral pandemics.
Limitations
Our findings are limited by the small sample size. Larger studies are needed to confirm our findings, especially for the cellular response.
Moreover, our study is biased toward young individuals, and our control cohort consisted exclusively of young women. A number of reports have observed that age and gender affect the cellular response after mRNA immunization.83 However, other reports have found no differences in these variables.57 , 84
Another limitation for humoral responses is related to the study time points. We are lacking time points which would have captured the decay in antibody titers more accurately 85.. Based on our results and those of other authors, cellular responses do not change and humoral responses gradually decrease,84 widening the range for our donor selection to 2 months and 6 months after full immunization.
In our study, all recovered donors had mild disease. We do not know how cellular responses will behave in individuals who have experienced moderate/severe disease.86 Wang Z et al. showed that asymptomatic and symptomatic COVID-19 patients have similar levels of SARS-CoV-2-specific T-cells,46 supporting the idea that any recovered immunized individual would be a good donor for our product, an important consideration given that most COVID-19 infections are asymptomatic,87 even more so after vaccination.[88], [89], [90]
We used IFN-γ as a marker to detect functional activity for cellular immunity. IFN-γ is a marker of Th1, and the Th1 subset coordinates the cell-mediated response, which is essential in macrophages, cytotoxic T cells and, NK activation via IL-2 and IFN-γ. However, we cannot rule out the possibility that other CD4+ T-cell subsets are activated upon antigen-specific encounters.5 , 91 Other authors have used different sets of markers to detect cellular immune response depending on the study's cell subset.5 , 56 , 92
We studied T cells from peripheral blood, when memory T cells in peripheral blood are only a small subset of the body's memory T cells. As for other memory T cells from other viruses, the formation of memory T cells can occur at distinct sites preferentially maintained at the sites of initial effect T-cell recruitment.49
Future studies are needed to better define the optimal donors for adoptive cell therapy in COVID-19 patients.
Funding
This study has been supported by CRIS Cancer Foundation and IMMUNOVACTER-CM from the European Regional Development Fund – Resources REACT-EU. K.A-A.S is funded by project DECODE-19 (Combinatorial Cell Therapy for high mortality risk Covid-19 patients, 500 ICI21/00016).
Author Contributions
Methodology, K.A-A.S, C.F, B.P-M, A.P.M.; Research, K.A-A.S, C.F., A.P.M.; Biostatistics, K.A-A.S, M.D-A; writing of the original draft C.F.; writing, reviewing and editing, all authors; resources, A.P.M.; supervision, A.P.M, C.F.
Declaration of Interest
AP-M and CF filed patent EP20382850 on Memory T cells as adoptive cell therapy for viral diseases.All other authors declare no competing interest
Appendix Supplementary materials
Image, application 1
Image, application 2
Image, application 3
Image, application 4
Image, application 5
Acknowledgements
We would like to thank all of the donors who participated in this study.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jcyt.2022.12.001.
==== Refs
References
1 Israel A. Elapsed time since BNT162b2 vaccine and risk of SARS-CoV-2 infection in a large cohort medRxiv Prepr. Serv. Heal. Sci. 2021 1 7 10.1101/2021.08.03.21261496
2 Wilder-Smith A. What is the vaccine effect on reducing transmission in the context of the SARS-CoV-2 delta variant? Lancet Infect. Dis. 380 2021 4 5
3 Singanayagam A. Community transmission and viral load kinetics of the SARS-CoV-2 delta (B . 1 . 617 2) variant in vaccinated and unvaccinated individuals in the UK : a prospective, longitudinal, cohort study 3099 2021
4 Moss P. Review Article | SERIES The T cell immune response against SARS-CoV-2 23 2022
5 Painter M.M. Rapid induction of antigen-specific CD4+ T cells is associated with coordinated humoral and cellular immunity to SARS-CoV-2 mRNA vaccination Immunity 54 2021 2133 2142 e3 34453880
6 Zheng M. Functional exhaustion of antiviral lymphocytes in COVID-19 patients Cell. Mol. Immunol. 17 2020 533 535 32203188
7 Zabalza A. Humoral and Cellular Responses to SARS-CoV-2 in Convalescent COVID-19 Patients With Multiple Sclerosis 0 2022 1 13
8 Riise, J. et al. treated patients with lymphoma develop strong CD8 cell responses following COVID- 19 vaccination. 697–708 (2022) doi:10.1111/bjh.18149.
9 Radbruch A. Chang H.D. A long-term perspective on immunity to COVID Nature 595 2021 359 360 34127832
10 Feng C. Protective humoral and cellular immune responses to SARS-CoV-2 persist up to 1 year after recovery Nat. Commun. 12 2021 1 7 33397941
11 Cox, R. J. & Brokstad, K. A. COmmenT Not just antibodies : B cells and T cells mediate immunity to COVID-19. Nat. Rev. Immunol.1–2 doi:10.1038/s41577-020-00436-4.
12 Lu Z. Durability of SARS-CoV-2 – Specific T-Cell Responses at 12 Months Postinfection 224 2021
13 Jordan S.C. T cell immune responses to SARS-CoV-2 and variants of concern (Alpha and Delta) in infected and vaccinated individuals Cell. Mol. Immunol. 18 2021 2554 2556 34531555
14 Marco L. De Assessment of T-cell Reactivity to the SARS-CoV-2 Omicron Variant by Immunized Individuals 5 2022 1 8
15 Kedzierska K. Thomas P.G. Count on us : T cells in SARS-CoV-2 infection and vaccination Cell Reports Med 3 2022 100562
16 Vardhana S. Understanding T-cell responses to COVID-19 is essential for informing public health strategies 1303 2022 1 9
17 Tarke A. Comprehensive analysis of T cell immunodominance and immunoprevalence of SARS-CoV-2 epitopes in COVID-19 cases Cell Reports Med 2 2021 100204
18 Qin C. Dysregulation of Immune Response in Patients with COVID-19 in Wuhan, China SSRN Electron. J 2020 10.2139/ssrn.3541136
19 Zhao Q. Lymphopenia is associated with severe coronavirus disease 2019 (COVID-19) infections: A systemic review and meta-analysis Int. J. Infect. Dis. 96 2020 131 135 32376308
20 Jafarzadeh A. Jafarzadeh S. Nozari P. Mokhtari P. Nemati M. Lymphopenia an important immunological abnormality in patients with COVID-19: Possible mechanisms Scand. J. Immunol. 93 2021 1 16
21 Odak I. Reappearance of effector T cells is associated with recovery from COVID-19 EBioMedicine 57 2020
22 Lee J. Park S.S. Kim T.Y. Lee D.G. Kim D.W. Lymphopenia as a biological predictor of outcomes in COVID-19 patients: A nationwide cohort study Cancers (Basel) 13 2021 1 15
23 Garbo R. Valent F. Gigli G.L. Valente M. Pre-Existing Lymphopenia Increases the Risk of Hospitalization and Death after SARS-CoV-2 Infection Infect. Dis. Rep. 14 2022 20 25 35076526
24 Xiang Q. SARS-CoV-2 Induces Lymphocytopenia by Promoting Inflammation and Decimates Secondary Lymphoid Organs Front. Immunol. 12 2021 1 13
25 Ferreras C. SARS-CoV-2-Specific Memory T Lymphocytes From COVID-19 Convalescent Donors: Identification, Biobanking, and Large-Scale Production for Adoptive Cell Therapy Front. Cell Dev. Biol 9 2021 1 12
26 Pérez-Martínez A. Phase I dose-escalation single centre clinical trial to evaluate the safety of infusion of memory T cells as adoptive therapy in COVID-19 (RELEASE) EClinicalMedicine 39 2021
27 Herrera L. Identifying SARS-CoV-2 ‘memory’ NK cells from COVID-19 convalescent donors for adoptive cell therapy Immunology 165 2022 234 249 34775592
28 Torres K.C.L. Norepinephrine, dopamine and dexamethasone modulate discrete leukocyte subpopulations and cytokine profiles from human PBMC J. Neuroimmunol. 166 2005 144 157 16026859
29 RECOVERY Collaborative GroupHorby P Dexamethasone in Hospitalized Patients with Covid-19 N Engl J Med 384 8 2021 Feb 25 693 704 10.1056/NEJMoa2021436 32678530
30 Gutsol A.A. Sokhonevich N.A. Seledtsov V.I. Litvinova L.S Dexamethasone effects on activation and proliferation of immune memory T cells Bull. Exp. Biol. Med. 155 2013 474 476 24143372
31 De Antonio, S.R. Effects of dexamethasone on lymphocyte proliferation and cytokine production in rheumatoid arthritis J. Rheumatol. 29 2002 46 51 11824970
32 Giles A.J. Dexamethasone-induced immunosuppression: Mechanisms and implications for immunotherapy J. Immunother. Cancer 6 2018 1 13 29298730
33 Schluns K.S. Lefrançois L. Cytokine control of memory T-cell development and survival Nat. Rev. Immunol. 3 2003 269 279 12669018
34 Arneja A. Johnson H. Gabrovsek L. Lauffenburger D.A. White F.M. Qualitatively Different T Cell Phenotypic Responses to IL-2 versus IL-15 Are Unified by Identical Dependences on Receptor Signal Strength and Duration J. Immunol. 192 2014 123 135 24298013
35 Knijff E.M. A relative resistance of T cells to dexamethasone in bipolar disorder Bipolar Disord 8 2006 740 750 17156159
36 Lowy M.T. Reder A.T. Gormley G.J. Meltzer H.Y. Comparison of in vivo and in vitro glucocorticoid sensitivity in depression: Relationship to the dexamethasone suppression test Biol. Psychiatry 24 1988 619 630 3262379
37 Fontanet A. SARS-CoV-2 variants and ending the COVID-19 pandemic Lancet 397 2021 952 954 33581803
38 Maschan M. Low-dose donor memory T-cell infusion after TCR alpha/beta depleted unrelated and haploidentical transplantation: Results of a pilot trial Bone Marrow Transplant 53 2018 264 273 29269793
39 Blagov S. T-cell tracking, safety, and effect of low-dose donor memory T-cell infusions after αβ T cell-depleted hematopoietic stem cell transplantation Bone Marrow Transplant 56 2021 900 908 33203952
40 Gasior Kabat M Selective T-cell depletion targeting CD45RA as a novel approach for HLA-mismatched hematopoietic stem cell transplantation in pediatric nonmalignant hematological diseases Int J Hematol 114 1 2021 Jul 116 123 10.1007/s12185-021-03138-2 33772729
41 Gasior M. The role of early natural killer cell adoptive infusion before engraftment in protecting against human herpesvirus-6B encephalitis after naïve T-cell-depleted allogeneic stem cell transplantation Transfusion 61 2021 1505 1517 33713461
42 Grifoni A. Targets of T Cell Responses to SARS-CoV-2 Coronavirus in Humans with COVID-19 Disease and Unexposed Individuals Cell 181 2020 1489 1501 .e15 32473127
43 Swain S.L. McKinstry K.K. Strutt T.M. Expanding roles for CD4 + T cells in immunity to viruses Nat. Rev. Immunol. 12 2012 136 148 22266691
44 Luckheeram R.V. Zhou R. Verma A.D. Xia B. CD4 +T cells: Differentiation and functions Clin. Dev. Immunol. 2012 2012
45 Dong C. Flavell R.A. Cell fate decision: T-helper 1 and subsets in immune responses Arthritis Res 2 2000 179 188 11094427
46 Wang Z. Exposure to SARS-CoV-2 generates T-cell memory in the absence of a detectable viral infection Nat. Commun. 12 2021 6 13 33397903
47 Ahrends T. CD4+ T cell help creates memory CD8+ T cells with innate and help-independent recall capacities Nat. Commun. 10 2019 1 13 30602773
48 Peng Y. Broad and strong memory CD4+ and CD8+ T cells induced by SARS-CoV-2 in UK convalescent individuals following COVID-19 Nat. Immunol. 21 2020 1336 1345 32887977
49 Farber DL Yudanin NA Restifo NP. Human memory T cells: generation, compartmentalization and homeostasis Nat Rev Immunol 14 1 2014 Jan 24 35 10.1038/nri3567 24336101
50 Mahnke Y.D. Brodie T.M. Sallusto F. Roederer M. Lugli E. The who's who of T-cell differentiation: Human memory T-cell subsets Eur. J. Immunol. 43 2013 2797 2809 24258910
51 Martínez-Flores D. SARS-CoV-2 Vaccines Based on the Spike Glycoprotein and Implications of New Viral Variants Front. Immunol. 12 2021
52 Glycoprotein C.-S. Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein Cell 181 2020 281 292 e6 32155444
53 Oberhardt V. Rapid and stable mobilization of CD8+ T cells by SARS-CoV-2 mRNA vaccine Nature 597 2021 268 273 34320609
54 Tarke A. Impact of SARS-CoV-2 variants on the total CD4+ and CD8+ T cell reactivity in infected or vaccinated individuals Cell Reports Med 2 2021 100355
55 Mateus J. Low-dose mRNA-1273 COVID-19 vaccine generates durable memory enhanced by cross-reactive T cells Science 374 80- 2021
56 Guerrera G. BNT162b2 vaccination induces durable SARS-CoV-2 specific T cells with a stem cell memory phenotype 5344 2021 1 18
57 Goel R.R. Distinct antibody and memory B cell responses in SARSCoV-2 naïve and recovered individuals following mRNA vaccination Sci. Immunol. 6 2021 1 19
58 Goel R.R. mRNA vaccines induce durable immune memory to SARS-CoV-2 and variants of concern Science 0829 80- 2021 1 28
59 Long Q.X. Clinical and immunological assessment of asymptomatic SARS-CoV-2 infections Nat. Med. 26 2020 1200 1204 32555424
60 Schiffner J. Long-Term Course of Humoral and Cellular Immune Responses in Outpatients After SARS-CoV-2 Infection Front. Public Heal. 9 2021 1 11
61 Prendecki M. Humoral and T-cell responses to SARS-CoV-2 vaccination in patients receiving immunosuppression Ann. Rheum. Dis. 80 2021 1322 1329 34362747
62 Krammer F. Antibody Responses in Seropositive Persons after a Single Dose of SARS-CoV-2 mRNA Vaccine N. Engl. J. Med. 384 2021 1372 1374 33691060
63 Dan J.M. Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection Science 371 80- 2021
64 Turner J.S. SARS-CoV-2 mRNA vaccines induce persistent human germinal centre responses Nature 596 2021 109 113 34182569
65 Esparcia-Pinedo L. CD4+ T Cell Immune Specificity Changes After Vaccination in Healthy And COVID-19 Convalescent Subjects Front. Immunol. 12 2022 1 9
66 De Giorgi V. Naturally Acquired SARS-CoV-2 Immunity Persists for up to 11 Months following Infection J. Infect. Dis. 224 2021 1294 1304 34089610
67 Mazzoni A. First-dose mRNA vaccination is sufficient to reactivate immunological memory to SARS-CoV-2 in subjects who have recovered from COVID-19 First-dose mRNA vaccination is sufficient to reactivate immunological memory to SARS-CoV-2 in subjects who have recover 131 2021
68 Evans J.P. Neutralizing antibody responses elicited by SARS-CoV-2 mRNA vaccination wane over time and are boosted by breakthrough infection Sci. Transl. Med. 14 2022 1 13
69 Ebinger J.E. Antibody responses to the BNT162b2 mRNA vaccine in individuals previously infected with SARS-CoV-2 Nat. Med. 27 2021 981 984 33795870
70 Naranbhai V. T cell reactivity to the SARS-CoV-2 Omicron variant is preserved in most but not all individuals Cell 185 2022 1041 1051 .e6 35202566
71 Dejnirattisai W. SARS-CoV-2 Omicron-B.1.1.529 leads to widespread escape from neutralizing antibody responses Cell 185 2022 467 484 e15 35081335
72 Yu X. Reduced sensitivity of SARS-CoV-2 Omicron variant to antibody neutralization elicited by booster vaccination Cell Discov 8 2022 8 11 35102135
73 Delamater P.L. Elimination of nonmedical immunization exemptions in California and school-entry vaccine status Pediatrics 143 2019 2021 2023
74 Geers D. SARS-CoV-2 variants of concern partially escape humoral but not T-cell responses in COVID-19 convalescent donors and vaccinees Sci. Immunol. 6 2021 1 15
75 Cele S. Omicron extensively but incompletely escapes Pfizer BNT162b2 neutralization Nature 602 2022 654 656 35016196
76 Agrati C. Coordinate induction of humoral and spike specific t-cell response in a cohort of italian health care workers receiving bnt162b2 mrna vaccine Microorganisms 9 2021
77 Rydyznski Moderbacher C. Antigen-Specific Adaptive Immunity to SARS-CoV-2 in Acute COVID-19 and Associations with Age and Disease Severity Cell 183 2020 996 1012 e19 33010815
78 Ni L. Detection of SARS-CoV-2-Specific Humoral and Cellular Immunity in COVID-19 Convalescent Individuals Immunity 52 2020 971 977 e3 32413330
79 Imamichi H Sereti I Lane HC. IL-15 acts as a potent inducer of CD4(+)CD25(hi) cells expressing FOXP3 Eur J Immunol 38 6 2008 Jun 1621 1630 10.1002/eji.200737607 18493981
80 Wang Y. The role of CD4 + FoxP3 + regulatory T cells in the immunopathogenesis of COVID-19 : implications for treatment 17 2021
81 Wang H. Regulatory T cells in COVID-19 Aging Dis 12 2021 1545 1553 34631206
82 Caldrer S. Regulatory T Cells as Predictors of Clinical Course in Hospitalised COVID-19 Patients Front. Immunol. 12 2021 1 11
83 Flanagan K.L. Fink A.L. Plebanski M. Klein S.L Sex and gender differences in the outcomes of vaccination over the life course Annu. Rev. Cell Dev. Biol. 33 2017 577 599 28992436
84 Guerrera G. BNT162b2 vaccination induces durable SARS-CoV-2 specific T cells with a stem cell memory phenotype Sci. Immunol. 2021 10.1126/sciimmunol.abl5344
85 Levin E.G. Waning Immune Humoral Response to BNT162b2 Covid-19 Vaccine over 6 Months N. Engl. J. Med. 2021 1 11 10.1056/nejmoa2114583
86 Papayanni P.G. Vaccinated and Convalescent Donor-Derived Severe Acute Respiratory Syndrome Coronavirus 2-Specific T Cells as Adoptive Immunotherapy for High-Risk Coronavirus Disease 2019 Patients Clin. Infect. Dis. 73 2021 2073 2082 33905481
87 Kimball A. Asymptomatic and Presymptomatic SARS-CoV-2 Infections in Residents of a Long-Term Care Skilled Nursing Facility Morb. Mortal. Wkly. Rep. Summ. CDC 69 2020 377 381
88 Tang L Hijano DR Gaur AH Asymptomatic and Symptomatic SARS-CoV-2 Infections After BNT162b2 Vaccination in a Routinely Screened Workforce JAMA 325 24 2021 2500 2502 10.1001/jama.2021.656 33956050
89 Gohil S.K. Asymptomatic and Symptomatic COVID-19 Infections among Health Care Personnel before and after Vaccination JAMA Netw. Open 4 2021 8 11
90 Shah ASV Effect of Vaccination on Transmission of SARS-CoV-2 N Engl J Med 385 18 2021 Oct 28 1718 1720 10.1056/NEJMc2106757 34496200
91 Gil-Etayo F.J. T-Helper Cell Subset Response Is a Determining Factor in COVID-19 Progression Front. Cell. Infect. Microbiol. 11 2021 1 11
92 Payne R.P. Immunogenicity of standard and extended dosing intervals of BNT162b2 mRNA vaccine Cell 184 2021 5699 5714 e11 34735795
| 0 | PMC9742221 | NO-CC CODE | 2022-12-14 23:29:59 | no | Cytotherapy. 2022 Dec 12; doi: 10.1016/j.jcyt.2022.12.001 | utf-8 | Cytotherapy | 2,022 | 10.1016/j.jcyt.2022.12.001 | oa_other |
==== Front
Biomed Signal Process Control
Biomed Signal Process Control
Biomedical Signal Processing and Control
1746-8094
1746-8094
Elsevier Ltd.
S1746-8094(22)00953-3
10.1016/j.bspc.2022.104499
104499
Article
COVID-19 and human development: An approach for classification of HDI with deep CNN
Kavuran Gürkan a⁎
Gökhan Şeyma b
Yeroğlu Celaleddin c
a Department of Electrical and Electronics Engineering, Faculty of Engineering and Natural Sciences, Malatya Turgut Ozal University, Malatya, Turkey
b TUBITAK BILGEM Software Test and Quality Evaluation Laboratory, Turkey
c Department of Computer Engineering, Faculty of Engineering, Inonu University, Malatya, Turkey
⁎ Corresponding author.
12 12 2022
3 2023
12 12 2022
81 104499104499
21 4 2022
18 10 2022
1 12 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.
The measures taken during the pandemic have had lasting effects on people's lives and perceptions of the ability of national and multilateral institutions to drive human development. Policies that changed people's behavior were at the heart of containing the spread of the virus. As a result, it has become a systemic human development crisis affecting health, the economy, education, social life, and accumulated gains. This study shows how the relationship of the Human Development Index (HDI), which has combined effects on health, education, and the economy, should be considered in the context of pandemic factors. First, COVID-19 data of the countries received from a public and credible source were extracted and organized into an acceptable structure. Then, we applied statistical feature selection to determine which variables are closely related to HDI and enabled the Deep Convolutional Neural Network (DCNN) model to give more accurate results. The Continuous Wavelet Transform (CWT) and scalogram methods were used for the time-series data visualization. Three different images of each country are combined into a single image to penetrate each other for ease of processing. These images were made suitable for the input of the ResNet-50 network, which is a pre-trained DCNN model, by going through various preprocessing processes. After the training and validation processes, the feature vectors in the fc1000 layer of the network were drawn and given to the Support Vector Machine Classifier (SVMC) input. We achieved total performance metrics of specificity (88.2%), sensitivity (96.5%), precision (99%), F1 Score (94.9%) and MCC (85.9%).
Keywords
Human Development Index
Deep learning
COVID-19
Continuous wavelet transform
Artificial intelligence
Classification
==== Body
pmc1 Introduction
The disease called COVID-19, caused by the new type of coronavirus that emerged in Wuhan, China, has adversely affected the world in many ways. A pandemic was declared by the World Health Organization (WHO) on March 11, 2020. As of December 1, 2021, the total number of cases worldwide was 262 million 178 thousand 403, while the total deaths were 5 million 215 thousand 745. Turkey ranks sixth in the world with 8 million 795 thousand 588 cases. While the first dose of vaccine was administered to 56 million 38 thousand 139 people, the second dose was administered to 50 million 750 thousand 190 people. Turkey became the 6th country with the highest number of vaccinations after the United States of America, India, and Brazil [1], [2], [3].
It has been clearly stated that the COVID-19 pandemic is more than a health emergency faced and is a systemic crisis affecting the world economy [4]. In many parts of the world, the pandemic is causing a crisis in human development. Today's conditions are comparable to deprivation levels last observed in the mid-1980 s in some dimensions of human development. However, the crisis is having a significant impact on all aspects of human growth, including income, health, and education. Although it is known that the pandemic has indirect effects on the increasing gender-based violence, it has not yet been thoroughly documented. Human capital is a critical driver of long-term prosperity, poverty reduction, and prosperous societies [5]. It should be connected with education and training and all types of activities that boost labor productivity and raise future income levels. In this regard, health-related expenditures are also seen as human capital investments [6]. Models made utilizing artificial intelligence began to make early diagnosis and treatment of patients easier at the outset of these investments.
The fact that the studies on artificial intelligence (AI) are applicable in many fields of science today is an indication that it has a practical and robust theoretical background [7], [8]. AI can assist us in combating this virus by recommending population screening, medical assistance, notification, and infection management [9]. In recent studies, the effectiveness of AI applications in the diagnosis of COVID-19 pneumonia has been investigated, and it has been determined to have great diagnostic performance in this field [10], [11], [12], [13], [14]. There are many studies in the literature about the combination of AI and medical applications related to COVID-19. For example, researchers at Mount Sinai have created an AI system to help diagnose COVID-19 patients using an algorithm trained on more than 900 Computed Tomography (CT) lung scans from Chinese patients [15]. Another AI system, developed by researchers at MIT, can detect COVID-19 based on the sound of a patient's coughing [16]. It has been shown that the dl-CRC model created using the DARI algorithm and the CNN model can be accurately detected by X-ray [17]. Using CNN and ConvLSTM as two deep learning methods, it has been shown that COVID-19 and normal cases can be tested on both CT and X-ray images. An efficient diagnosis can be made to detect COVID-19, among other related infections [18]. The CNN-based transfer learning-BiLSTM hybrid construct has been shown to be highly effective for diagnosing COVID-19. Since the proposed study automatically performs the segmentation process, it provides high classification accuracy and convenience [19]. Yang and colleagues analyzed 152 CT images using CNN and demonstrated an accuracy of 0.89 in diagnosing COVID-19 [20]. Ahamed et al. developed an extended ResNet50V2 based deep learning model where finetuning has been performed to detect and diagnose COVID-19 infected cases [21]. Kavuran et al. evaluated a fully automatic deep-learning system for the diagnosis of COVID-19 using thoracic computed tomography in [22]. They reached an overall accuracy of 97.7 % with the optimized features from concatenated layers.
Within the scope of this study, a machine learning-based decision support system has been proposed in order to find the relationship between the HDI of countries and the COVID-19 data of that country. These time-varying data basically include approximately 35 variables, such as the number of daily cases, the number of tests, the death rate, the vaccination rate, and sociodemographic information. First of all, statistical feature selection was applied to determine which of these variables are closely related to HDI and enable the DCNN model to give more accurate results. According to the ANOVA f-value, the study was started by considering the variable with the highest three scores for each country. For the visualization of the time series data, the Continuous Wavelet Transform (CWT) method and the scalogram method were used. Three different images of each country are combined into a single image to penetrate each other for ease of processing. These images were made suitable for the input of the ResNet-50 network, which is a pre-trained DCNN model, by going through various preprocessing processes. The weighting coefficients in the learnable convolution layer (fc1000) of the network have been revised, and the classification performance has been increased. After the training and validation processes, the feature vectors in the fc1000 layer of the network were drawn and given to the SVMC input, which is another machine learning-based classifier. Based on the confusion matrix and Receiver Operating Characteristic (ROC) curves, system performance was examined using metrics such as accuracy, sensitivity, specificity, sensitivity, F1, and Matthew Correlation Coefficient (MCC).
In the second part of the study, the steps of the proposed method will be explained. While the experimental results are given in Chapter 3, the results and discussion are given in Chapter 4.
2 Methodology
In this section, the steps of the proposed method will be explained in sub-headings. The study can be examined under three main topics. In the first stage, data cleaning processes were carried out by obtaining the publicly available data set. Thus, an operable form was created from the raw data. Since the number of features that make the data meaningful is high, elimination has been made. Attribute selection helps us remove the less important parts of the data and shorten the training time. For data labeling, the three selected attribute vectors were associated with the HDI of the related country. The “new_tests” vector, which is the third component of the obtained package, was scanned according to the sliding window method with a length of 50 units, and its standard deviation was taken. Inner Penetrated Images (IPI) were created by composing three separate scalogram images into each other for each class. Thus, a single image containing the structural features of all three images was obtained. In the next stage, the pre-trained ResNet-50 network, which is used as a deep convolutional neural network, was fed with the obtained IPI images. During the learning period, the number of images was increased at this step by using image augmentation methods online. The feature vectors from the modified fc1000 fully connected layer of the ResNet-50 were normalized to zero mean to be given to the SVMC input. In the last stage, deep feature vectors drawn from the trained network are reserved for 75 % training and 25 % validation. Evaluation metrics are presented with the help of the ROC curve and complexity matrix. The schematic diagram of the study is given in Fig. 1 .Fig. 1 Schematic diagram of the study.
2.1 Preprocess data for domain-specific deep learning
2.1.1 Data description and feature selection
The dataset used in this study was obtained by reporting data on confirmed cases and deaths, managed by a team at the Johns Hopkins University Center for Systems Science and Engineering (CSSE). Since January 22, 2020, data on confirmed cases and deaths for all countries has been updated several times daily. This data is obtained from national and local agencies of governments, and a complete list of data sources for each country is published on the Johns Hopkins GitHub site [23]. Fig. 2 . shows cumulative confirmed cases of COVID-19 as of January 11, 2022. The number of confirmed cases is lower than the actual number of infections due to limited testing.Fig. 2 Cumulative confirmed COVID-19 cases, January 11, 2022 [23].
In the dataset consisting of a total of 187 countries, we have taken the time-dependent changes between February 14, 2020, and August 10, 2021, given in Table 2 . These form the predictors' vectors that we will use in the study. We chose HDI as the output vector in order to classify the relationship of the feature vectors of each country with the HDI, which is a metric that measures how far people have come in their development in three fundamental characteristics of humans: living a long and healthy life, having access to knowledge, and having a reasonable quality of living. The HDI is calculated as the geometric mean of the three-dimensional index [24],(1) HDI=IHealth·IEducation·IIncome1/3
Table 1 Cutoff points on the HDI for grouping countries [24].
Very high human development (VH) 0.800 and above
High human development (H) 0.700–0.799
Medium human development (M) 0.550–0.699
Low human development (L) Below 0.550
Table 2 List of the predictors for COVID-19.
ID Predictors ID Predictors
1. Total cases 19. Weekly icu admissions per million
2. New cases 20. Weekly hospital admissions
3. New cases smoothed 21. Weekly hospital admissions per million
4. Total deaths 22. New tests
5. New deaths 23. Total tests
6. New deaths smoothed 24. Total tests per thousand
7. Total cases per million 25. New tests per thousand
8. New cases per million 26. New tests smoothed
9. New cases smoothed per million 27. New tests smoothed per thousand
10. Total deaths per million 28. Positive rate
11. New deaths per million 29. Tests per case
12. New deaths smoothed per million 30. Tests units
13. Reproduction rate 31. Total vaccinations
14. Icu patients 32. People vaccinated
15. Icu patients per million 33. People fully vaccinated
16. Hosp patients 34. New vaccinations
17. Hosp patients per million 35. New vaccinations smoothed
18. Weekly icu admissions
In the 2014 Human Development Report, a system of predetermined cutoff criteria for the four categories of human development accomplishments was implemented. This Report retains the same HDI cutoff parameters for categorizing nations announced in the 2014 Report as in Table 1.
Thus, we aimed to reveal the structure of the link between COVID-19 data and HDI. The unstable change and instability in COVID-19 data do not constitute a clear boundary between countries in terms of HDI. Therefore, M and L classes are grouped under L, and H and VH classes are under H. We eliminated some of the 35 predictor vectors considered using the univariate feature selection method for the deep CNN structure to give more accurate results.
When preparing a big dataset for training, selecting the best features is critical. It assists us in removing less essential sections of the data and reducing training time. It can be viewed as a phase in deep CNN preprocessing. Univariate feature selection chooses the best features based on statistical testing. To expose feature selection routines, we used Python Scikit-learn modules. For the samples in Table 1, we compute the ANOVA F-value [25]. The modules remove all but the highest-scoring features. These objects accept a scoring function as an input and return univariate scores and p-values. This method can be used on sample sets for feature selection and dimensionality reduction to either increase deep CNN accuracy scores or improve their performance on very high-dimensional datasets. It provides the optimal class for extracting the most useful features from a given dataset. This technique chooses the features based on the k highest score. The method can be used with both classification and regression data. The feature importance graph is given in Fig. 3 . The top 4 feature vectors with the best score are ID_26 (new tests smoothed), ID_22 (new tests), ID_4 (total deaths), and ID_1 (total cases), respectively. One of these two vectors was chosen since ID_26 is a smoothed version of ID_22. In comparison to raw data, smoothed data eliminates reporting irregularities and provides a more accurate view of time. Data at the state level is noisy at the raw data level, making it challenging to detect trends. Thus, a feature vector package with an average length of 3x550 for each country was obtained.Fig. 3 Feature importance graph.
It is necessary to highlight some features in the signal structure that changes depending on time. This situation provides positive benefits in the performance of the classifier [26], [27]. Some statistical methods are available for the analysis and interpretation of these properties. In order to choose the appropriate statistical method for data analysis, it is necessary to know the assumptions and conditions of statistical methods. One of these methods is the standard deviation. The standard deviation is a statistical analysis approach for determining the spread of data around the mean. When dealing with a high standard deviation, this refers to data that is spread out from the mean over a vast area. Similarly, a low deviation shows that most data agree with the mean and can also be referred to as a cluster's expected value. The standard deviation is mostly used to determine the distribution of data points. The “new_tests” vector, which is the third component of the obtained data package, was scanned according to the sliding window method with a length of 50 units, and its standard deviation was taken. The formulation of the standard deviation is given in Equation (2).(2) Standard=(1N-1∑i=1N(X(i)-X¯)2)12
Time-Frequency Representations.
In many engineering applications, time–frequency analysis is an effective tool for the in-depth investigation of signal properties, separating input signals into their time-varying spectral components. The Continuous Wavelet Transform (CWT) has played an essential role in analyzing time–frequency data. This section will discuss the CWT method for 2D time–frequency representation of time series COVID-19 data. CWT is a method that allows us to analyze time scale features (new tests, total deaths, total cases) in COVID-19 data at different times and scales. The timescale feature in CWT allows obtaining high time–frequency resolution, which helps identify sub-features of COVID-19 data. Wavelet analysis is commonly used to extract local characteristics, as it can provide an in-depth perspective of the temporal and frequency fluctuation of the examined signals. The CWT of time series COVID-19 data is represented as:(3) C(a,b,f(t),ψ(t))=1a∫-∞∞f(t)ψ∗(t-ba)·dt
where, ψ(t) is the mother wavelet, a is the scale, C(a,b) is the CWT coefficients, and b is the time shift. Deviations of cumulative total cases, cumulative total deaths, and daily new COVID-19 tests against to days for Albania and the Democratic Republic of Congo are given in Fig. 4 .Fig. 4 Deviation of cumulative total cases, cumulative total deaths, and daily new COVID-19 tests against days for Albania and the Democratic Republic of Congo.
The scalogram aids in visualizing rapid changes in the signal by utilizing the coefficients acquired after applying the continuous wavelet transform to the signal. The CWT scalogram is a two-dimensional drawing that depicts time on the horizontal axis and scales on the vertical axis. Fig. 5 illustrates the scalogram of ID_22 (new tests), ID_4 (total deaths), and ID_1 (total cases) for both High HDI and Low HDI classes. The time–frequency representations of the scalograms are sent into the input of deep CNN structures that will be utilized to classify. Filter banks contribute to the extraction of frequency content by passing scalogram images obtained with COVID-19 data through various filters. IPI were created by composing three separate images into each other for each class to eliminate the processing complexity and make the training-validation data distribution correctly. Thus, a single image containing the structural features of all three images was obtained. Fig. 4 depicts the scalogram and IPI images of High-Low HDI.Fig. 5 An example of Scalogram and IPI images of High-Low HDI.
2.2 Deep CNN and SVM model
This section presents the methodological content of the proposed method. Deep learning is a method of developing mathematical models using artificial intelligence without predetermining the signal or image attributes. Throughout reality, neural networks automatically extract and define features in the deep learning process. The output of each previous layer is sent as input to the next layer, and then events “L” or “H” are formed. Hidden layers process significant qualities, which are subsequently supplied to each of the deep levels. Fig. 6 shows the Resnet-50 structure, with input features traveling through 50 hidden layers before reaching the modified fully connected layer and SVMC, which are classed as “L” or “H.”Fig. 6 ResNet-50 architecture.
CNN is one of the systematic structures used in deep learning algorithms that is successful on huge data. A ResNet-50 model is adopted in this research, which is a pre-trained CNN network with 1000 categories and over one million images [6]. ResNet-50 will be trained to identify a new set of IPI photos and detect HDI using the IPI images collected in the previous section. An input layer, 50 deep layers (convolution layers, bulk normalization layers, and ReLU), classification layers (average pooling layers, fully connected layers, and SoftMax layers), and output layers comprise ResNet-50. Input images have been resized to 224x224 pixels to be compatible with ResNet-50 models.
During the learning phase, the number of data was expanded by using online image augmentation technologies to minimize imbalance and overfitting issues in the number of images for each class. A compelling image classifier was designed using MATLAB's image data augmentation module. This paper proposes a two-stage classification process: feature extraction using pre-trained deep network structures and labeling of IPI images using SVMC [28], [29], [30]. Deep CNN features were retrieved from ResNet-50 activations using modified fc1000 activations. The mean of all feature sets was adjusted to zero. In the classification phase, SVMC was employed with 75 percent training and 25 percent test data. The proposed method's validity and its impact on the evaluation criteria are presented concerning computing efficiency. All the experiments were performed in a MATLAB environment running on a PC with AMD Ryzen 5 2600 3.4 GHz CPU, 64 GB memory, and 12 GB NVIDIA GeForce RTX 2080 TI GPU.
3 Experimental results
The performance of the suggested design will be analyzed in this section using statistical assessment criteria. Accuracy (ACC), Sensitivity (Sens.), Specificity (Spec.), Precision (Prec.), F1 score (F1), and MCC were all calculated using the confusion matrix. The expected and actual classes are represented by rows and columns in the confusion matrix. Correctly categorized observations are denoted by crossover cells, whereas erroneously classified observations are represented by other cells. Each cell contains the number of observations as well as the percentage of the overall number of observations. The percentages of all samples projected to belong to each class categorized true (precision measures) and false are shown in the graph's far-right column (false discovery rate measures). The cell in the matrix's lower right corner represents total accuracy. The ROC curve, which shows the true positive rate as a function of the false positive rate, is also one of the most extensively used metrics for evaluating the effectiveness of machine learning algorithms. Equation 4–9 defines the evaluation metrics that were used.(4) Accuracy=NTP+NTNNTP+NTN+NFP+NFN
(5) Sensitivity=NTPNTP+NFN
(6) Specificity=NTNNTN+NFP
(7) Precision=NTPNTP+NFP
(8) F1=2NTP2NTP+NFP+NFN
(9) MCC=NTPNTN-NFPNFNNTP+NFPNTP+NFNNTN+NFPNTN+NFN
Here, NTP, NTN, NFP, and NFN define the number of true-positive classes, the number of true-negative classes, the number of false-positive classes, and the number of false-negative classes, respectively. The SVMC's multi-class confusion matrix, which was fed by the deep feature set, is shown in Fig. 7 a. There are three misclassified samples among the 46 validation samples, according to the confusion matrix. With two samples, the “L” class was determined to have the most misclassified samples. The ROC curves of the predicted classes with AUC values are given in Fig. 7b. The area of the ROC curve was obtained as 0.961 and 0.969 for “L” and “H” classes, respectively.Fig. 7 a) The confusion matrix of the SVMC with deep features. b) The ROC curves of all classes for deep features.
The SVMC's general classification scores are listed in Table 3 . The total metric values for the COVID-19 dataset were observed in the proposed design with 187x2 features, as shown in the table. For the COVID-19 dataset, the model utilizing deep features from the updated fc1000 layer had an overall accuracy of 93.4 percent. The remaining overall performance indicators, Spec., Sens., Prec., F1, and MCC, scored 88.2 %, 96.5 percent, 99 percent, 94.9 percent, and 85.9 %, respectively.Table 3 The general classification scores of the SVMC for the deep feature set.
Evaluation Metrics
Accuracy Sensitivity Specificity Precision F1 MCC
Results 0.934 0.965 0.882 0.990 0.949 0.859
4 Conclusion
Artificial intelligence has revolutionized diagnostic imaging systems and is one of the trendiest subjects in medical imaging. Deep learning approaches, particularly the usage of CNNs, have resulted in significant performance improvements over traditional machine learning techniques. The importance of non-medical data becoming clinical predictors has been extensively reviewed in the literature. This study is based on medical data and demonstrates the correlation between HDI calculated using socioeconomic inputs. To the best knowledge of the authors, this paper is the first that uses signal processing techniques with DCNN for the classification of HDI by using COVID-19 data. According to this study, HDI, which has integrated effects on health, education, and the economy, should be addressed in the context of pandemic causes. First, COVID-19 data from the countries were retrieved and structured into an acceptable structure using information obtained from a public and reliable source. Then, using statistical feature selection, we figured out which factors are closely associated with HDI, allowing the DCNN model to produce more accurate findings. The time-series data visualization was done using the CWT and scalogram methods. For the convenience of processing, three separate images of each country are merged into a single image to pierce each other. By going through different preprocessing steps, these images were made acceptable for the input of the ResNet-50 network, which is a pre-trained DCNN model. The feature vectors in the fc1000 layer of the network were created and given to the SVMC input after the training and validation operations. The complete performance parameters of specificity (88.2 %), sensitivity (96.5 %), accuracy (99 %), F1 Score (94.9 %), and MCC (85.9 %) were all achieved.
In this study, we found a significant correlation between the socioeconomic standing of a nation and the COVID-19 effects. Previous studies have demonstrated a substantial correlation between sociodemographic risk variables and COVID-19 incidence and mortality [31]. A lower percentage of COVID-19 vaccinations and fewer testing is linked to lower socioeconomic levels. It demonstrates how it has a negative impact on people's lives and increases the disease burden in low-income countries [32], [33]. On the other hand, prior research demonstrates that the COVID-19 epidemic has both expanded and generated new socioeconomic inequalities [34].
Key elements for speeding vaccination coverage at the national level include socioeconomic variables and the health policies of the governments. Daily test rates are one of these issues. One of our most crucial tools in the effort to prevent and stop the spread of the virus and minimize its effects during the pandemic is daily testing. The tests enable us to locate and confine infected peoples' contacts. It can also help allocate medical resources and staff more efficiently. Therefore, the use of daily test data as the determining factor in this study played an important role in the classification of countries in terms of HDI. Examining the graphs in Fig. 4, it becomes clear that the variance in daily test numbers accounts for the majority of the differences between nations with high and low HDI. This circumstance is believed to have a favorable impact on classification performance.
There are several types of research in the literature about the association between HDI and confirmed COVID-19 cases by country. The hierarchical multiple linear regression models are presented in [35] to identify factors associated with confirmed cases of COVID-19 with a focus on the HDI. Countries with a high HDI could conduct more tests per person, leading to more confirmed cases than those with a low HDI. The Pearson correlation test and univariate linear regression analysis were performed on the relationship between HDI, HDI-Education Level, HDI-Life Expectancy, and HDI-Gross National Income per capita and COVID-19 deaths in Brazil [36]. Palamim et al. showed the case fatality rate depends on at least 20–40 % of the HDI. Ameye et al. proposed a method to assess the Case Fatality Rate (CFR) of COVID-19 as an indicator to evaluate Nigeria's performance relative to other selected countries based on the HDI using hierarchical clustering [37]. Generally, statistical analysis was used in the methodology of these studies. To the best of the authors' knowledge, similar research using the methods suggested in this study has not been found in the literature. It is thought that this study, which structurally includes signal processing, statistical analysis, feature selection, and deep learning methods, will contribute to the literature. The study, which provides for the network analysis showing the relationship between the selected features and the dependent variable, will be on our next agenda.
Funding
This study was funded by Inonu University Scientific Research Projects Management Unit with the project number FYL-2021-2377.
CRediT authorship contribution statement
Gürkan Kavuran: Conceptualization, Methodology, Data curation, Validation, Software, Formal analysis, Visualization, Writing – original draft. Şeyma Gökhan: Conceptualization, Methodology, Data curation. Celaleddin Yeroğlu: Writing – review & editing, Supervision, Project administration, Funding acquisition.
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: ‘Celaleddin YEROGLU reports a relationship with Inonu University that includes: funding grants. This study was funded by Inonu University Scientific Research Projects Management Unit with the project number FYL-2021-2377.’
Data availability
Data will be made available on request.
==== Refs
References
1 WHO, 2021. Available from: <Https://Www.Who.Int/Emergencies/Diseases/Novel-Coronavirus-2019/>.
2 Department of Error: Nowcasting and forecasting the potential domestic and international spread of the 2019-nCoV outbreak originating in Wuhan, China: a modelling study (The Lancet (2020) 395(10225) (689–697), (S0140673620302609), (10.1016/S0140-6736(20)30260-9)), Lancet 395 (2020), doi: 10.1016/S0140-6736(20)30302-0.
3 who.covid19, 2021. Available from: <Https://Covid19.Who.Int/>.
4 www.un.org/en/coronavirus/UN-response, Https://Www.Un.Org/En/Coronavirus/UN-Response. (n.d.).
5 WorldBank, The Human Capital Index 2020 Update Human Capital in the Time of Covid-19, International Bank for Reconstruction and Development / The World Bank, 2020.
6 Hasnaoui J.A. Rizvi S.K.A. Reddy K. Mirza N. Naqvi B. Human capital efficiency, performance, market, and volatility timing of asian equity funds during COVID-19 outbreak J. Asset Manag. 22 2021 10.1057/s41260-021-00228-y
7 Kavuran G. SEM-Net: deep features selections with Binary Particle Swarm Optimization Method for classification of scanning electron microscope images Mater Today Commun. 27 2021 102198 10.1016/J.MTCOMM.2021.102198
8 Kavuran G. When machine learning meets fractional-order chaotic signals: detecting dynamical variations Chaos Solitons Fractals. 157 2022 111908 10.1016/J.CHAOS.2022.111908
9 Vaishya R. Javaid M. Haleem I. Haleem A. Artificial Intelligence (AI) applications for COVID-19 pandemic Diabetes & Metabolic Syndrome : Clinical Research & Reviews Arti fi cial Intelligence (AI) applications for COVID-19 pandemic Diabetes Metab. Syndr. 14 2020
10 Li L. Qin L. Xu Z. Yin Y. Wang X. Kong B. Bai J. Lu Y. Fang Z. Song Q. Cao K. Liu D. Wang G. Xu Q. Fang X. Zhang S. Xia J. Xia J. Using artificial intelligence to detect COVID-19 and community-acquired pneumonia based on pulmonary CT: evaluation of the diagnostic accuracy Radiology 296 2020 E65 E71 10.1148/radiol.2020200905 32191588
11 Zheng C. Deng X. Fu Q. Zhou Q. Feng J. Ma H. Liu W. Wang X. Deep learning-based detection for COVID-19 from chest CT using weak label MedRxiv. 2020 10.1101/2020.03.12.20027185
12 Loey M. Mirjalili S. COVID-19 cough sound symptoms classification from scalogram image representation using deep learning models Comput Biol Med. 139 2021 105020 10.1016/J.COMPBIOMED.2021.105020
13 Liao Z. Lan P. Fan X. Kelly B. Innes A. Liao Z. SIRVD-DL: A COVID-19 deep learning prediction model based on time-dependent SIRVD Comput Biol Med. 138 2021 104868 10.1016/J.COMPBIOMED.2021.104868
14 Pahar M. Klopper M. Warren R. Niesler T. COVID-19 detection in cough, breath and speech using deep transfer learning and bottleneck features Comput Biol Med. 141 2022 105153 10.1016/J.COMPBIOMED.2021.105153
15 X. Mei, H.C. Lee, K. yue Diao, M. Huang, B. Lin, C. Liu, Z. Xie, Y. Ma, P.M. Robson, M. Chung, A. Bernheim, V. Mani, C. Calcagno, K. Li, S. Li, H. Shan, J. Lv, T. Zhao, J. Xia, Q. Long, S. Steinberger, A. Jacobi, T. Deyer, M. Luksza, F. Liu, B.P. Little, Z.A. Fayad, Y. Yang, Artificial intelligence–enabled rapid diagnosis of patients with COVID-19, Nat. Med. 26 (2020), doi: 10.1038/s41591-020-0931-3.
16 Laguarta J. Hueto F. Subirana B. COVID-19 artificial intelligence diagnosis using only cough recordings IEEE Open J Eng Med Biol. 1 2020 10.1109/ojemb.2020.3026928
17 Sakib S. Tazrin T. Fouda M.M. Fadlullah Z.M. Guizani M. DL-CRC: Deep learning-based chest radiograph classification for covid-19 detection: a novel approach IEEE Access 8 2020 10.1109/ACCESS.2020.3025010
18 Sedik A. Hammad M. Abd El-Samie F.E. Gupta B.B. Abd El-Latif A.A. Efficient deep learning approach for augmented detection of Coronavirus disease Neural Comput Appl. 2021 10.1007/s00521-020-05410-8
19 Aslan M.F. Unlersen M.F. Sabanci K. Durdu A. CNN-based transfer learning–BiLSTM network: a novel approach for COVID-19 infection detection Appl Soft Comput. 98 2021 10.1016/j.asoc.2020.106912
20 Wu Y.H. Gao S.H. Mei J. Xu J. Fan D.P. Zhang R.G. Cheng M.M. JCS: an explainable COVID-19 diagnosis system by joint classification and segmentation IEEE Trans. Image Process. 30 2021 10.1109/TIP.2021.3058783
21 Ahamed K.U. Islam M. Uddin A. Akhter A. Paul B.K. Yousuf M.A. Uddin S. Quinn J.M.W. Moni M.A. A deep learning approach using effective preprocessing techniques to detect COVID-19 from chest CT-scan and X-ray images Comput Biol Med. 139 2021 105014 10.1016/J.COMPBIOMED.2021.105014
22 Kavuran G. İn E. Geçkil A.A. Şahin M. Berber N.K. MTU-COVNet: a hybrid methodology for diagnosing the COVID-19 pneumonia with optimized features from multi-net Clin Imaging. 81 2022 1 8 10.1016/J.CLINIMAG.2021.09.007 34592696
23 Dong E. Du H. Gardner L. An interactive web-based dashboard to track COVID-19 in real time Lancet Infect Dis. 20 2020 533 534 accessed February 9, 2022 32087114
24 Human Development Index (HDI) | Human Development Reports, (n.d.). Available from: <https://hdr.undp.org/en/content/human-development-index-hdi> (accessed February 9, 2022).
25 Ferri F.J. Pudil P. Hatef M. Kittler J. Comparative study of techniques for large-scale feature selection Mach. Intelligence Pattern Recogn. 1994 10.1016/B978-0-444-81892-8.50040-7
26 Hu X. Yuan S. Xu F. Leng Y. Yuan K. Yuan Q. Scalp EEG classification using deep Bi-LSTM network for seizure detection Comput Biol Med. 124 2020 103919 10.1016/j.compbiomed.2020.103919
27 Wang D. Deep learning reinvents the hearing aid IEEE Spectr. 54 2017 32 37 10.1109/MSPEC.2017.7864754 30636790
28 Cortes C. Vapnik V. Support-vector networks Mach Learn. 1995 10.1023/A:1022627411411
29 V. Vapnik, The support vector method of function estimation, in: Nonlinear Modeling, Springer US, Boston, MA, 1998, pp. 55–85, doi: 10.1007/978-1-4615-5703-6_3.
30 Burges C.J.C. A tutorial on support vector machines for pattern recognition Data Min Knowl Discov. 2 1998 121 167 10.1023/A:1009715923555
31 Roghani A. The relationship between macro-socioeconomics determinants and COVID-19 vaccine distribution AIMS Public Health. 8 2021 655 10.3934/PUBLICHEALTH.2021052 34786426
32 Ghosh A. Nundy S. Mallick T.K. How India is dealing with COVID-19 pandemic Sensors Int. 1 2020 100021 10.1016/J.SINTL.2020.100021
33 Karmakar M. Lantz P.M. Tipirneni R. Association of social and demographic factors With COVID-19 incidence and death rates in the US JAMA Netw Open. 4 2021 10.1001/JAMANETWORKOPEN.2020.36462
34 Qian Y. Fan W. Who loses income during the COVID-19 outbreak? Evidence from China Res. Soc. Stratif. Mobil. 68 2020 100522 10.1016/J.RSSM.2020.100522
35 Heo M.H. Kwon Y.D. Cheon J. Kim K.B. Noh J.W. Association between the human development index and confirmed COVID-19 cases by country Healthcare (Basel) 10 2022 10.3390/HEALTHCARE10081417
36 Palamim C.V.C. Boschiero M.N. Valencise F.E. Marson F.A.L. Human development index is associated with COVID-19 case fatality rate in Brazil: an ecological study Int. J. Environ. Res. Public Health 19 2022 5306 10.3390/IJERPH19095306/S1 35564707
37 Ameye S.A. Ojo T.O. Adetunji T.A. Awoleye M.O. Is there an association between COVID-19 mortality and Human development index? The case study of Nigeria and some selected countries BMC Res. Notes 15 2022 1 7 10.1186/S13104-022-06070-8/FIGURES/3 34983646
| 0 | PMC9742222 | NO-CC CODE | 2022-12-14 23:43:09 | no | Biomed Signal Process Control. 2023 Mar 12; 81:104499 | utf-8 | Biomed Signal Process Control | 2,022 | 10.1016/j.bspc.2022.104499 | oa_other |
==== Front
Intensive Crit Care Nurs
Intensive Crit Care Nurs
Intensive & Critical Care Nursing
0964-3397
1532-4036
Elsevier Ltd.
S0964-3397(22)00176-8
10.1016/j.iccn.2022.103373
103373
Research Article
Spiritual needs during COVID 19 pandemic in the perceptions of Spanish emergency critical care health professionals
De Diego-Cordero Rocío a1
Rey-Reyes Azahara b2
Vega-Escaño Juan c⁎
Lucchetti Giancarlo d
Badanta Bárbara e3
a Faculty of Nursing, Physiotherapy and Podiatry, Department of Nursing, University of Sevilla, c/ Avenzoar 6, 41009 Seville, Spain
b Faculty of Health Science, University of Malaga, c/ Arquitecto Francisco Peñalosa 3, 29071 Málaga, Spain
c Faculty of Nursing, Physiotherapy and Podiatry, Department of Nursing, University of Seville, c/ Avenzoar 6, 41009 Seville, Spain
d Department of Medicine, School of Medicine, Federal University of Juiz de Fora, Bandeirantes, Juiz de Fora - MG, 36047, Brazil
e Faculty of Nursing, Physiotherapy, and Podiatry, Department of Nursing, University of Sevilla, c/ Avenzoar 6, 41009 Seville, Spain
⁎ Corresponding author.
1 Research Group CTS 969 “Innovation in HealthCare and Social Determinants of Health”. School of Nursing, Physiotherapy and Podiatry. University of Sevilla, c/ Avenzoar 6, 41009, Seville, Spain.
2 ICU nursing in QuirónSalud Málaga Hospital, Av. Imperio Argentina 1, 29004 Málaga, Spain.
3 Research Group under the Andalusian Research CTS 1050 “Complex Care, Chronic and Health Outcomes”. School of Nursing, Physiotherapy and Podiatry. University of Sevilla, c/ Avenzoar 6, 41009, Seville, Spain.
12 12 2022
12 12 2022
10337313 9 2022
25 11 2022
6 12 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.
Objectives
To investigate the perceptions and attitudes of health professionals working in emergency services and critical care units in Spain about spiritual care provided during the COVID-19 pandemic.
Methods
A qualitative investigation was carried out using in-depth interviews with 48 emergency and emergency and ICU health professionals from different regions of Spain concerning their perceptions and opinions of spiritual needs and spiritual care during COVID-19 pandemic were done and thematic analysis were used.
Findings
The sample consisted of 48 health professionals. The qualitative analysis yielded four main themes that reflect the following categories: “the experience with spirituality in clinical practice”; “resources and barriers to provide spiritual care”; “the COVID pandemic and spiritual care” and “training in spiritual care”. In addition, two subdeliveries were also obtained: “ethical dilemma” and “rituals of death”.
Conclusions
The most emergency and critical care nurses believe spiritual care is important to their clinical practice, but there are still several barriers to address patients’ spiritual needs. During the COVID-19 pandemic in Spain, nurses felt that spiritual beliefs have emerged as important needs of patients and the restrictions imposed by the pandemic made health professionals more exposed to ethical dilemmas and end-of-life religious issues. The general impression of nurses is that more training and resources are needed on this topic.
Keywords
clinical practice
COVID-19
nurses
religion
spirituality
==== Body
pmc Implications for clinical practice • Emergency critical care health professionals should provide nursing care that meets the spiritual needs of their patients.
• Being aware of religious and spiritual symbols is a valuable approach to improve health care in crisis situations such as those suffered by the COVID-19 pandemic.
• Emergency services professionals should work and participate in the development of measures to overcome certain barriers present in emergency services, such as lack of time, lack of training, and misconceptions.
Introduction
As the definition of health and disease has evolved into a complex integrative and multidimensional construct; other overlooked dimensions, such as spiritual beliefs, have been incorporated into healthcare (Guirao Goris, 2013). Although there are different definitions, for the present study, spirituality will be considered “the aspect of humanity that refers to the way individuals seek and express meaning and purpose and the way they experience their connectedness to the moment, to self, to others, to nature, and to the significant or sacred“ (Puchalski & Larson, 1998).
In recent decades, spirituality research has been increasingly consolidating in the scientific community (Lucchetti & Lucchetti, 2014) spiritual beliefs have been recognised as a powerful coping mechanism for dealing with traumatic events (Koenig, 2012), allowing a more optimistic perspective and fostering faith and hope (Ortega Jiménez et al., 2016).
Despite this promising evidence, previous studies have shown that spirituality is not well valued in nursing education and poorly addressed in clinical practice (Lewinson et al., 2015). Nurses report that they feel a lack of knowledge, understanding, and skills in spiritual care and that addressing spirituality demands more knowledge and skills than they have (de Diego Cordero et al., 2019). Among the most common barriers reported by nurses, workload, lack of training, fear of imposing own beliefs, lack of time, lack of motivation and patient privacy were mentioned (Alch et al., 2021, Riahi et al., 2018, Veloza-Gómez et al., 2017).
Spiritual care is particularly important for critical care and emergency health providers, given that they work at stressful environments where patients are in critical and life-threatening conditions (Riahi et al., 2018). Previous research has supported that such patients consider important that health professionals address their spiritual needs in this situation (Santana Biondo et al., 2017). However, spiritual care does not seem to be perceived as a priority in emergency services, so it is often neglected and not present in acute care nursing, even though it can support patients and families to face adversity and existential issues such as death (Santana Biondo et al., 2017, Swinton et al., 2017). Likewise, spirituality seems to be important to health providers. Workload, ethical dilemmas, and stress also put the emotional well-being of healthcare professionals in critical care services, experiencing problems of anxiety and depression (Romero-García et al., 2022) and spiritual beliefs could be important coping mechanisms as well.
The pivotal role of spirituality in healthcare has been supported in the recent moment during the COVID-19 pandemic (de Diego-Cordero et al., 2022). Since the beginning of the pandemic, there have been more than 300,000,000 cases and 5,500,000 deaths from coronavirus (World Health Organization, 2021). This crisis has been a challenge for health systems around the world, testing contingency and public health plans (Sebastián Rocchetti et al., 2020). During the initial response to the surge of the pandemic, the spiritual and psychological needs of patients and their families lost priority and many critically ill COVID-19 patients died in isolation without appropriate support (Galbadage et al., 2020). This situation has meant an urgent need to address spiritual needs of these patients given the degree of isolation, loneliness, and vulnerability caused by this pandemic (Ferrell et al., 2020).
In Spain, the state of alarm was declared on March 14th 2020, which lasted until June 21st 2020 and led to home confinement, the suspension of educational, commercial, recreational and leisure activities, among others, limiting the exit of individuals only for the purchase of food and medicine, health care or going to work (Ministry of Presidency, 2020). The number of deaths in Spain was one of the highest in the World and the work overload of emergency personnel and the lack of hospital resources were definitely a problem (Dosil Santamaría et al., 2021).
In these difficult scenario, the use of spirituality and religiosity during isolation has been identified as a protective factor against anxiety, depression and suicide (Fountoulakis et al., 2021) and has been associated with better health outcomes (higher levels of hope and lower levels of fear, worry and sadness) and less suffering (Lucchetti et al., 2020).
Within hospitals, where feelings of vulnerability, stress, or helplessness are more frequent, spiritual care has emerged as an important healthcare tool (Badanta et al., 2022). Nevertheless, the lack of chaplains and certified religious leaders due to the risk of contagion, as well as, the unfamiliarity and lack of training of health providers with spiritual care, made several patients not having their spiritual needs met. (Ferrell et al., 2020, Santana Biondo et al., 2017).
A previous study aimed to analyse empirical evidence on the influence of Spirituality on critical care nursing (Ho et al., 2018). Furthermore, another article aimed to investigate the perceptions and attitudes of nurses working in intensive care units (ICU) in Spain about the spiritual care provided to patients and their families during the COVID-19 pandemic (de Diego-Cordero et al., 2021).
In this context, the present study aims to investigate the perceptions and attitudes of health professionals working in emergency services in Spain about spiritual care provided during the COVID 19 pandemic.
Methods
Study design
A qualitative, exploratory, and observational study was conducted using a phenomenological approach. This approach aims to describe the meaning of an experience by identifying essential subordinate and major themes (Moser & Korstjens, 2018).
Data collection consisted of in-depth interviews conducted by qualified investigators in two moments, from January to June 2020 (first wave of COVID-19 pandemic in Spain) and from February to May 2021 (third wave), totaling 10 months of data collection.
Setting, sample, and eligibility criteria
Phenomenology uses criterion sampling, being the most prominent criterion the participant’s experience with the phenomenon under study (Moser & Korstjens, 2018). Therefore, participants were included provided they were nurses working in intensive care units (ICUs) or emergency services from both public or private health institutions in Spain and treating critically ill patients with COVID-19. While the ICU nurses worked in hospitals, emergency care was provided in hospitals, primary care, and outpatient emergency units. Religious personnel from hospital centers, health professionals who were working outside ICUs or emergency services, as well as those not caring for patients (i.e., academic or management level) were excluded.
Procedures
A WhatsApp message that included a poster describing the study was widely distributed using professional and personal contacts (i.e., graduate students, registered nurses, and critical care and emergency services managers). Since this is a specific sample (emergency critical care personnel), very busy in the context of COVID-19 and usually difficult to be reached and to be willing to participate in interview studies, we have opted to add a snowball sampling procedure as well, in which participants interviewed by researchers were asked to suggest names of colleagues to take part in the study (Higginbottom, 2004). Interested participants directly contacted the researchers and eligibility criteria were applied. The eligible participants were invited by the main researcher, who is a nurse and anthropologist with expertise in spiritual care and has published several articles in the field of “Spirituality and Health”.
Since Spain was facing the implementation of a 'state of emergency' during the first wave of COVID-19 and there were restrictions for the free movement of people between Spanish cities in the third wave, no face-to-face meetings with health professionals were allowed. For this reason, interviews occurred individually at a time convenient to the participants (that is, respecting their preferences), using electronic devices. In this way, recordings of telephone calls and video calls were made using web meetings tools. The interviews were carried out by two researchers in the Spanish language and lasted approximately 50 to 60 minutes. Data collection continued until data saturation was reached. It was when no new analytical information was identified, and the study provided maximum information on the phenomenon (Urra et al., 2013).
Instrument
Since we opted to a phenomenological approach, semi-structured questions were used along with exploratory questions to clarify the lived experience of the participants. An interview script was developed, and the questions script was designed to encourage participants to tell their personal experiences, including feelings and emotions, and often focus on a particular experience or specific events (Moser & Korstjens, 2018).
The content of this interview script was developed using the Delphi method approach. The Delphi technique is an exercise in group communication that brings together and synthesises the knowledge of a group of geographically scattered participants (Boulkedid et al., 2011). In our case, we opted to an e-delphi approach, where experts were invited and fill in the forms online. A consensus of was considered for each question and a total of 15 experts in spiritual health and/or critical care were invited and agreed to participate. The characteristics of the expert panel are shown in the Table 1 . In order to achieve a consensus for all items of the delphi, two rounds of assessments were needed. Finally, the interview script was adapted according to the expert́s analysis (Table 2 ).Table 1 Characteristics of Experts for Delphi Panel
Code Age Gender Residence Highest academic level Occupation
Expert 1 - 1° wave 40 Woman Murcia PhD Medical anthropologist
Expert 2 - 1° wave 59 Woman Seville PhD University professor / ICU nurse
Expert 3 – 1° wave 63 Woman Seville PhD University professor / palliative care specialist
Expert 4 – 1° wave 57 Woman Seville MRcN University professor / midwife / Expert in community nursing
Expert 5 – 1° wave 43 Woman Seville PhD University professor in religious center
Expert 6 – 1° wave 42 Woman Seville PhD University professor in religious center
Expert 7 – 1° wave 33 Woman Seville PhD University professor in religious center
Expert 8 – 1° wave 59 Man Granada PhD Psychologist / EASP
Expert 9 – 1° wave 33 Woman Seville MRcN Research nurse
Expert 10 – 1° wave 47 Man Brazil PhD Doctor
Expert 1 – 3° wave 33 Woman Seville PhD Emergency nurse
Expert 2 – 3° wave 48 Man Seville Master Emergency nurse / University professor
Expert 3 – 3° wave 60 Man Seville Doctor Emergency nurse / University professor
Expert 4 – 3° wave 24 Woman Seville PhD Emergency nurse
Expert 5 – 3° wave 45 Hombre Seville PhD Emergency nurse / University professor
Table 2 Interview Guide.
1. Have you ever heard the term spiritual health? What do you mean by it?
2. Do you think that S / R influences in any way the health of patients, their coping with the disease? If so, how do you think it influence? Do you have any experiences or know some examples?
3. In your opinion, does the spirituality/religiosity of health professionals interfere with the professional-patient relationship? How does it influence?
4. Do you feel like discussing faith/spirituality with patients?
5. Have you ever asked your patients about religion/spirituality? * Yes (If the answer is “Yes”: How often do you usually do it? When or in what situations do you usually address this question?) Have your patients ever shown any religious or spiritual aspect that characterized them?
6. How do you think you can provide spiritual care in your daily activity?
7. To what extent, from 1 to 10, with 1 not being prepared at all and 10 being totally prepared, do you consider yourself prepared to address religious/spiritual issues with your patients? Why?
8. Do you feel difficulties or barriers that discourage you from discussing religion/spirituality with your patients? Which?
9. What is the role of your spiritual or religious belief at the time in dealing with this situation?
10. Have you seen aspects of spirituality/religiosity emerge in you, your colleagues, patients or family members to cope with the COVID 19 situation?
11. How do you perceive that the beliefs of the individual infected with COVID 19 influence their evolution of the disease?
12. Have you ever experienced a critical situation of a patient who, due to spiritual/religious convictions, wants or rejects some type of care or treatment for COVID 19 that confronts or conflicts with their beliefs (ethical dilemmas)? g How would you deal with it or what would be your position if it happened?
13. What part of the restrictions during the pandemic (due to covid-19) do you think may affect or have affected the mood or spirit of the patient or their family?
14. Regarding the restriction of visitors or companions at this time, how do you think it affects the patient or family?
15. Have you experienced the situation of a death from COVID 19? How have you and your colleagues experienced it? Have any spiritual or religious elements been present at the death of a patient?
16. How do you think spiritual care could be improved? Are there any resources or help that might be helpful?
17. Do you think that the health system has offered adequate, equitable and ethical care to all people during the pandemic? In what cases? Have you experienced it in your work?
18. What is dignified and ethical care for your patients for you? Characteristics or implications
19. Are you foreseen or have you experienced insufficient resources for all patients and the prioritization of some patients over others?
20. Do you know of any end-of-life support protocol for people affected by COVID-19 and their families? Have you experienced such a situation? What do you think about it?
21. Have effective means been used in your unit to transmit information to family members or to communicate patients with their families? If so, what resources have been used? (calls, video calls…); What effect does it have on patients?
22. Has the implementation of psychological services in general, for patients, relatives or professionals been considered in your service/unit? What do you think about it? Who do you think would benefit the most?
ACADEMIC TRAINING IN SPIRITUALITY
23. Do you consider it useful that these aspects and spiritual/religious care have more value within university education? Where: in undergraduate studies? In specific subjects? crosswise? in postgraduate studies: master or expert? Do you know something?
24. What level of importance could this training have for your work in the area of urgencies / emergencies? why?
Data Analysis
The data were analysed using a phenomenological approach, which followed the Amadeo Giorgi theory (Giorgi, 1997). Phenomenological analysis implies capturing and describing the “life world” of the study participant, and it is important for the researcher to avoid interpretations of the narrated experiences of the study participants and to present the life world as it appears to the respondent (Nyström & Dahlberg, 2001).
Thematic analysis, as described by (Braun et al., 2019) was also used. Qualitative analysis was carried out using the following steps: (1) familiarisation with the data; (2) generation of categories; (3-5) search, review, and definition of themes; and (6) final report, which was prepared with the statements of the informants, indicated by participant number, gender, age, place of work, wave 1-3, and transcription, literal reading and theoretical manual categorisation were performed. A Spanish-English translation of the manuscript was carried by a translation company, and for the quotations, an English-Spanish back-translation was performed by a Spanish native (n = 1).
Finally, the analysis and treatment of MAXQDA 12 qualitative data was used. MAXQDA is a software program designed for the use in qualitative, quantitative, and mixed methods research.
Trustworthy
This research followed The Consolidated Criteria for Reporting Qualitative Studies (COREQ) (Tong et al., 2007). The methods used for guaranteeing quality were data triangulation, including participants with different sociodemographic characteristics, and the triangulation of data analysis via different researchers (See the Supplementary Material Table S1).
Ethical Considerations
The study received approval from Andalusian Ethics and Research Committee. Participants were invited by the principal investigator to participate voluntarily, so they were incorporated into the study after accepting and signing the informed consent sent by email.
Findings
Description of the Sample
The sample consisted of 48 professionals who work in Spanish ICU and emergency services, 70.8% women with a mean age of 28.9 years (ranging from 22 to 52 years). Most of the participants were nurses (97,9%) and worked in emergency departments (60.4%), from public health institutions (75%) and lived in Seville (35.4%), Cataluña (22.9%) and Málaga (22.9%). Regarding their spiritual and religious beliefs, 41.7% of the sample defined themselves as spiritual and religious beings and 43.8% were Catholics. A total of 25% of participants had no religious affiliation and did not believe in God. The complete characteristics of the sample are shown in Table 3 .Table 3 Characteristics of the sample
Variable Participants (n=48)
Absolute frequency Relative frequency
Gender
Female 34 70,8 %
Male 14 29,2%
Middel ages 28,9 years old
Place
Málaga 11 22,9 %
Seville 17 35,4 %
Barcelona 11 22,9 %
Madrid 3 6,3 %
Huelva 2 4,2 %
Jaen 2 4,2 %
Granada 1 2,1 %
Cáceres (Jaraiz de la Vera) 1 2,1 %
Beliefs
Yes religious, Yes spiritual 20 41,7 %
Yes religious, No spiritual 4 8,3 %
No religious, Yes spiritual 8 16,7 %
No religious, No spiritual 12 25,0 %
I couldn’t answer 4 8,3 %
Religious orientation
None, but I believe in God 7 14,6 %
None and I don’t believe in God 18 37,5 %
Catholic 21 43,8 %
Buddhist 1 2,1 %
I believe in God and energy 1 2,1 %
Work unit
Emergency room 29 60,4 %
Ambulatory care 4 8,3 %
Ambulance 6 12,5 %
Intensive care unit 12 25 %
Family and community nursing home 1 2,1 %
Others 2 4,2 %
Financing the job you hold
Public 36 75,0 %
Public and private 8 16,7 %
Concerted/Subcontracted 3 6,3 %
Private 1 2,1 %
The qualitative analysis yielded four main themes that reflect the following categories: “the experience with spirituality in clinical practice”; “resources and barriers to provide spiritual care”; “the COVID pandemic and spiritual care” and “training in spiritual care”. In addition, two subdeliveries were also obtained: “ethical dilemma” and “rituals of death”.
These themes and sub-deliveries are described below.
Theme 1: The experience with Spirituality in clinical practice
First, we wanted to know the previous perceptions that the participants had about spirituality. A total of 39.6% of health professionals said they had never heard the term 'spiritual health', although they consider spirituality as something intangible that can improve hope and well-being, providing comfort and overcoming adverse situations.Participant 27, female, 23 years old, emergency care, 3ª wave: “It makes you feel good about yourself [spirituality], allowing you to have psychological and psychic health.”
Participant 3, female, 52 years old, ICU, 1ª wave: “Truly I dońt know exactly what it refers to, I have never heard it, but I imagine that it is the need that certain people have to feel wrapped up or think of something that gives them that comfort or always have something there that they can hold on to, to which they can pray.”
Second, the influence of religiosity/spirituality on the health of patients or to cope with the disease was explored. On this, there was agreement, 95,8 % of the participants believed that there was such an influence.
A common conclusion that was repeated among the participants was that S/R helped patients and their relatives to better cope with adverse situations such as suffering from a disease, life-threatening illnesses, bereavement, and the death of a relative. According to the participants, spirituality serves as support and, in their view, patients who did not have these beliefs used to have a worse outcome.Participant 47, female, 28 years old, ICU and ambulance, 3ª wave: “Having faith in something gives you a “light at the end of the tunnel” to try to evolve favorably in the disease.”
Participant 15, female, 26 years old, emergency care, 1ª wave: “In matters such as grief, I think it is more bearable in religious people because they are really convinced of a spiritual afterlife.”
After that, we wanted to know how they perceived that S / R influenced patient care. A common idea was that if the religious and/or spiritual beliefs of healthcare professionals coincided with those of patients, this could reinforce the relationship that was created between them, since the healthcare professional could empathize more with the patient and better cover their needs and could even improve the care and treatment they provided to their patients. Similarly, participants were observed to consider that they could have a negative relationship if beliefs did not coincide between the patient and the health professional.Participant 32, female 28 years old, emergency care, 3ª wave: “I think the more spiritual a professional or a person is, the more compassion, connection, and spiritual and emotional closeness he can have with someone else and I also think more healing ability in terms of helping the other person and feeling good about oneself.”
Participant 6, female, 22 years old, emergency care, 1° wave: “I think it may interfere if it does not match the patient's beliefs, that is, for a healthcare system with religious beliefs it may be difficult to understand a patient with different beliefs and vice versa.”
Theme 2: Resources and barriers to provide spiritual care
Concerning the provision of spiritual care, 56.3% of the participants reported that they did not address the spiritual needs of their patients, explaining that the workload, the professional neutrality, their own beliefs, the length of the patient’s admission or that this is out of the scope of their job as potential barriers.Participant 4, male, 31 years old, ICU, 1° wave: “I believe that by not considering myself a spiritual person of faith I have never noticed that need in my patients, possibly wrongly on my part.”
Participant 29, male, 22 years old, emergency care, 3° wave: “It is a topic that I don’t like to deal with patients because everyone is free to think what they want.”
Regarding available resources, most of the participants reported not having had access to spiritual or religious resources. Common ideas that emerged as necessary were the establishment of times and spaces and the incorporation of personnel specialised in spirituality. Similarly, participants agreed that they addressed this issue when patients suggested it or in situations of a poor prognosis. Despite this, only eight of the participants said they had referred their patients to chaplains or religious leaders.Participant 43, female 50 years old, emergency care, 3° wave: “One thing, for example, that I have demanded a lot is that if the person is sick, we can offer him to come to the priest to be there with him.”
Participant 7, female, 22 years old, emergency care, 1° wave: “In hospitals include more people who attend to the different religions that exist. In most hospitals there are only priests, but no imams for the Islamic religion, neither for the Jewish or Buddhist religion.”
We also tried to identify the barriers that health professionals found with respect to spiritual care and those highlighted were lack of training and ignorance about spirituality and/or religion, not feeling spiritual or religious, lack of time, and fear of discomfort or rejection.Participant 18, female, 25 years old, emergency care, 1° wave: “These aspects (spirituality/religiosity) are not usually shared with you.”
Participant 38, female, 25 years old, emergency care and ambulance, 3° wave: “I believe that ignorance is the main reason why maybe I don’t have that kind of conversation, because I don’t know how to approach it.”
It was also seen that lack of training and knowledge made health professionals feel insecure about spirituality and, for this reason, doctors did not open the door to patients to talk about it. Other barriers that were described by the participants were lack of time, perspective and resources, workload, lack of privacy, and interruptions.Participant 9, female, 23 years old, ICU, 1° wave: “Is not something recognised. You cannot talk to your partner about it (S/R) and he tells you 'that's fine' because it's not the same as saying 'I just took an arterial' or 'there's a complication'. All this is technical, there is nothing psychological you can say.”
Participant 36, male, 26 years old, emergency care, 3° wave: “It demotivates me (to offer spiritual care) because we don’t have time.”
Theme 3: The COVID pandemic and spiritual care
3.a. Ethical dilemma
A total of 81.25% of the participants denied having experienced an ethical dilemma during the pandemic, that is, patients who refused treatment or care for pneumonia by SARS-COV 2 due to religious or spiritual issues. Among the issues that caused ethical dilemmas in interviewees during the pandemic, they highlighted the limitation of therapeutic effort and avoidance of blood transfusions.Participant 43, female, 50 years old, emergency care, 3° wave: “Perhaps we have had many spiritual crises because there were patients who, not the famous intubate or not intubate, but we have wondered what we do with a patient who is really sick, is getting worse and has already had a bad quality of life before? So there have been many morals involved, many spiritualities involved or many religious values involved in healthcare as far as doing everything possible if where are we going? What are we doing?”
Regarding how health workers faced ethical dilemmas, most agreed that, despite posing a risk to patients, they should accept and respect their decisions as long as the patient had been informed and understood the risks and benefits of accepting or refusing treatment.Participant 5, female, 25 years old, ICU, 1° wave: “The first time I was quite surprised [the rejection of a treatment] because I thought “He will die for something that for me is as simple as its religion”, I have always respected it, but for me it is something irrational.”
3.b. Rituals of Death
One of the problems caused by COVID-19 pandemic was the isolation of patients and the restriction of family visits. Most of the interviewees agreed that this measure, although necessary, was hard and may have accentuated feelings of loneliness, sadness, anxiety, and/or suffering to patients, family members and even health professionals themselves.Participant 8, female, 24 years old, ICU, 1° wave: “I believe that it is one of the greatest handicaps of this pandemic, since the companionship of family members is very comforting to both; and the restriction of visits generates anguish and uncertainty; although we must understand that it is the best thing for all.”
Regarding the presence of spiritual or religious rituals, the most mentioned by the participants were the presence of personal objects, such as stamps or photographs, or rites in the death of relatives. This was limited because entering isolation was allowed only for restricted personnel to avoid contamination and contagion.Participant 28, female, 28 years old, emergency care and ambulance, 1° wave: “There have even been relatives who asked you when you went to put the family member in the shroud to leave the cross, he had or the bracelet he had.”
Participant 42, female, 26 years old, emergency care, 3° wave: “The moment they move into the COVID zone, all that changes. They go to an area where they don’t have their belongings, where what they have is the minimum possible, and maybe many don’t have a television or anything.”
On the other hand, isolation measures generally did not have any exception, even in end-of-life situations (more in the first wave than in the third). This had an important impact, as it prevented the accompaniment and separation between family and patient. Religious rites such as extreme nurses' union in Catholicism were not offered or were performed less frequently in covid positive patients compared to patients not infected with SARS-COV 2. This situation was described by health professionals as a very difficult experience to live with and they affected themselves in their mood.Participant 43, female, 50 years old, emergency care, 3° wave: “It’s the toughest part for me, having to let a person die alone, and that the relatives are out alone and that they cannot be accompanied in that trance, in that part of life that is death, because that has been the worst by far. It's the worst thing I’ve ever been through, I’ve gone home crying every time about the impotence of not being able to do anything about the deaths in the hospital.”
Theme 4: Training in Spiritual Care
Most of those interviewed (77%) considered that spiritual training should be given more value and be included in the healthcare schools as it would be beneficial to care for and meet all patient needs holistic and comprehensively. They also indicated that they preferred to receive theoretical and practical training. Additionally, health care said that they do not feel formed in spirituality and considered lack of training or ignorance an obstacle to dealing with spirituality with patients.Participant 7, female, 22 years old, emergency care, 1° wave: “Yes, I see it as really necessary [formation in spirituality] because we are caring for patients with many beliefs, and we are not able to cover that need for the patient.”
Finally, health professionals reflected on the importance of spiritual and/or religious care in their work. A total of 66.7% considered that spiritual care was of high or very high importance in critical or emergency care since such care could help to deal with bad news or borderline situations in their service.Participant 37, female, 23 years old, emergency care, 3° wave: “I think it is important because it is a service where there is a lot of stress, a lot of adrenaline and you see very shocking things. So, I think it is important to take care of spiritual health.”
On the other hand, there was another minority opinion current, 27.1% who considered that spiritual care was not relevant in critical and/or emergency services as biological patient care prevailed.Participant 46, female, 30 years old, ambulance, 3° wave: “If I have life or death care, I will not worry [about spirituality]. In case of emergency, there are other priorities.”
Discussion
Our findings revealed that most emergency and critical care nurses believe spiritual care is important to their clinical practice, but there are still several barriers to address patients’ spiritual needs, such as lack of training, misconceptions, and lack of time. During the COVID-19 pandemic in Spain, nurses felt that spiritual beliefs have emerged as important needs of patients and the restrictions imposed by the pandemic made health professionals more exposed to ethical dilemmas and end-of-life religious issues. The general impression of nurses is that training is necessary, and more resources should be allocated to this topic.
Regarding the previous conceptions of the participants, 39.6% of health professionals said they had never heard of the term 'spiritual health' and most of them have challenges to differentiate religion and spirituality. This result agrees with that of some authors in their work, Badanta et al. pointed out that previous studies had identified difficulties in differentiating between religiosity and spirituality among health professionals (Badanta et al., 2022). Despite the lack of conceptualization, most of the participants agreed that spirituality had an influence on the health of patients, helping them cope with the disease or thought that spirituality even influenced the professional-patient relationship. This is also in accordance with previous studies, showing that most health professionals believe spiritual issues are important for patients and should be considered in clinical practice (Kørup et al., 2021, Vasconcelos et al., 2020).
As necessary resources for approaching spirituality in emergency and critical care, the need for time and places that allowed for a spiritual approach and the availability of specialised personnel, such as spiritual guides covering all different religions, was pointed out. Despite this recognition that spirituality should be incorporated in their clinical practice, our study supported that this issue is sometimes underrecognised and not so frequently addressed in emergency and critical care settings. However, some nurses avoided the responsibility of spiritual care justifying that the emergency room was not the place to provide this care, since urgent acute care prevails and should be focused solely on physical and biological issues. This is particularly important if we consider that patients in these environments have acute and life-threatening conditions and tend to use religious and spiritual issues to cope with their conditions. A previous US study has found that 90% of patients believed that prayer may impact their recovery and 94% agreed that physicians should ask them whether they have such beliefs if they become gravely ill (Ehman et al., 1999).
In our study, the most common reasons for failing to address spiritual needs were related to lack of time, lack of training, misconceptions, lack of religious knowledge and fear of imposing beliefs. These factors are also fully supported by previous studies (Espinha et al., 2013, Gordon et al., 2018) and seems to be related to the lack of training for these professionals. Addressing spiritual issues do not demand too much time and when training is available misconceptions and the fear of imposing beliefs tend to diminish, as noted in an educational trial that assessed the efficacy of a spirituality training among healthcare students (Osório et al., 2017). Likewise, in our sample, participants alluded to the fact that the knowledge they had about religion was mainly from the Catholic religion, and in clinical practice they encountered several different religions. This problem could be reduced providing training to health professionals as well.
Finally, since our study took place during the pandemic, there were several assertions on the relationship between spiritual needs/beliefs and COVID-19. First, participants noted that there appeared religious dilemmas during the pandemic, mostly related to end-of-life issues and blood transfusions. End-of-life issues are strongly influenced by cultural and religious issues and previous studies have shown that more religious individuals tend to have more aggressive end-of-life preferences (Balboni et al., 2007). Nevertheless, spiritual care at the end-of-life has proven to be associated with less aggressive care at the end of life (Peteet & Balboni, 2013). Concerning blood transfusions, there are patients who refuse blood transfusions for religious reasons, such as the case of Jehovah’s Witnesses. For these patients, there are options such as optimisation of hemoglobin levels preoperatively, attention to blood-salvaging methods intraoperatively, minimization of blood draws postoperatively and transfusion alternatives (e.g. hemoglobin-based oxygen carriers) (Rashid et al., 2021).
Another important point reported by participants was the restriction caused by COVID 19, which affected the mood of families and make patients isolated due to the limitation of contact and visits to the family. The feelings that this aroused according to the interviewees were fear, fear, anxiety, frustration, impotence, uncertainty, or agony. The role of spirituality in coping with stressful situations is highlighted here again, since positive religious coping can lead to thinking positively about adverse events and help reduce depressive and/or stressful symptoms (Mahamid & Bdier, 2021). Furthermore, hope could act as a buffer against the anxiety and stress of the virus pandemic caused by the closure of COVID or social distancing (Al Eid et al., 2021). Since such limitations were imposed, religious support suffered important consequences and the religious rituals and needs during hospitalisation were impaired.
Throughout the interviews, feelings of emotional exhaustion were also observed in the nurses, who expressed tiredness, boredom, or detachment. In previous articles, the increase in poor mental health among nurses during the COVID-19 pandemic has already been described, (Kim et al., 2021) described an increase in anxiety and stress among the nurses interviewed during their research. These emotions could suggest the existence of Burnout Syndrome, which is known to be a phenomenon that causes physical, psychological and/or spiritual exhaustion when the worker, in this case the nurses, is subjected to an overload such as pain, death and lack of support from managers (De Diego-Cordero et al., 2021). The latter is compatible with the situation experienced during the pandemic by the interviewees: deaths from COVID, isolation, work overload, lack of resources, etc. Spirituality could have a strong weight in these situations, (De Diego-Cordero et al., 2021) concluded in a systematic review that spirituality or religiosity served as a coping strategy to alleviate burnout in clinical practice and could serve as a health promoter for nurses, helping them rediscover passion for their profession and life. Kim also noted that nurses with high levels of resilience, spirituality, and family functioning are two to six times less likely to have poor mental health (Kim et al., 2021).
Limitations and Strengths
The present study has some limitations that should be considered. First, this study was carried out in Spain and reflects the experiences of health professionals from Spanish health care facilities. It is difficult to guarantee that the same results would be observed in other countries, since cultural and religious backgrounds are different. Second, we used a purposive sample and generalisability should be made with caution. Third, this intervention was carried out during the first and third waves of the COVID-19 pandemic, and these findings could be different in other moments of the pandemic. Finally, our sample consisted predominantly of women (70.8%). Nevertheless, this feminisation of the sample coincides with what has been pointed out in previous studies that have shown that women are the majority gender in nursing (El Arnaout et al., 2019).
Future studies should evaluate the opinions and perceptions of emergency and critical care health professionals among different cultural contexts, since different societies could have different approaches for the addressing of spirituality. For instance, more secular societies could have more resistance to consider such issues and societies with a predominant religion could have more difficulties in accepting other religious traditions. Another interesting future direction for studies would be to compare the differences between addressing spirituality needs during moments of crisis and social isolation (such as the case of the pandemic) and compare to routine conditions (before or after the pandemic).
Conclusions
In conclusion, most emergency and critical care health professionals believe that spirituality is important to patients’ health and believe it could promote faith and hope. Nevertheless, spiritual needs are not frequently addressed in clinical practice due to several barriers such as lack of time, misconceptions and lack of training. According to these professionals, the resources necessary to provide targeted spiritual care are the creation of enabling spaces and the integration of religious personnel specialised in different cultures.
Spirituality plays an important role in people and even more so in patients during the COVID-19 pandemic, improving coping with adverse events and influencing end-of-life issues and ethical dilemmas. The need for spiritual care training is evident, although of low importance since emergency care is understood to be more focused on strictly biological components to overcome the acute and life-threatening emergency.
Uncited references
Cordero et al., 2018, de Diego-Cordero et al., 2022, de Diego-cordero et al., 2022, Lucchetti et al., 2021, Pub, 2020, Ronaldson et al., 2012.
CRediT authorship contribution statement
Rocío De Diego-Cordero: Conceptualization, Methodology, Software, Validation, Investigation, Resources, Writing – original draft. Azahara Rey-Reyes: Writing – review & editing. Juan Vega-Escaño: Methodology, Software, Validation, Formal analysis, Data curation, Writing – original draft, Supervision. Giancarlo Lucchetti: . Bárbara Badanta: Methodology, Software, Validation, Formal analysis, Data curation, Writing – original draft, Visualization.
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
Al Eid N.A. Arnout B.A. Alqahtani M.M.J. Fadhel F.H. Abdelmotelab A.s. The mediating role of religiosity and hope for the effect of self-stigma on psychological well-being among COVID-19 patients WOR 68 3 2021 525 541
Alch C.K. Wright C.L. Collier K.M. Choi P.J. Barriers to Addressing the Spiritual and Religious Needs of Patients and Families in the Intensive Care Unit: A Qualitative Study of Critical Care Physicians The American Journal of Hospice & Palliative Care 38 9 2021 1120 1125 10.1177/1049909120970903 33143446
Badanta B. Rivilla-García E. Lucchetti G. de Diego-Cordero R. The influence of spirituality and religion on critical care nursing: An integrative review Nursing in Critical Care 27 3 2022 348 366 10.1111/nicc.12645 33966310
Balboni T.A. Vanderwerker L.C. Block S.D. Paulk M.E. Lathan C.S. Peteet J.R. Prigerson H.G. Religiousness and spiritual support among advanced cancer patients and associations with end-of-life treatment preferences and quality of life Journal of Clinical Oncology : Official Journal of the American Society of Clinical Oncology 25 5 2007 555 560 10.1200/JCO.2006.07.9046 17290065
Boulkedid R. Abdoul H. Loustau M. Sibony O. Alberti C. Wright J.M. Using and reporting the Delphi method for selecting healthcare quality indicators: a systematic review PloS One 6 6 2011 e20476 21694759
Braun V. Clarke V. Hayfield N. Terry G. Handbook of Research Methods in Health Social Sciences 2019 Springer Singapore Singapore 843 860
Cordero R.d.D. Romero B.B. Matos F.A. Costa E. Espinha D.C.M. Tomasso C.d.S. Lucchetti A.L.G. Lucchetti G. Opinions and attitudes on the relationship between spirituality, religiosity and health: A comparison between nursing students from Brazil and Portugal Journal of Clinical Nursing 27 13-14 2018 2804 2813 29516571
de Diego-Cordero R. Ávila-Mantilla A. Vega-Escaño J. Lucchetti G. Badanta B. The Role of Spirituality and Religiosity in Healthcare During the COVID-19 Pandemic: An Integrative Review of the Scientific Literature Journal of Religion and Health 61 3 2022 2168 2197 10.1007/s10943-022-01549-x 35348988
De Diego-Cordero R. Iglesias-Romo M. Badanta B. Lucchetti G. Vega-Escaño J. Burnout and spirituality among nurses: A scoping review EXPLORE. 18 5 2022 612 620 34429263
de Diego-cordero R. López-Gómez L. Lucchetti G. Badanta B. Spiritual care in critically ill patients during COVID-19 pandemic Nursing Outlook 70 1 2022 64 77 34711420
de Diego Cordero R. Lucchetti G. Fernández-Vazquez A. Badanta-Romero B. Opinions, Knowledge and Attitudes Concerning “Spirituality, Religiosity and Health” Among Health Graduates in a Spanish University Journal of Religion and Health 58 5 2019 1592 1604 10.1007/s10943-019-00780-3 30771142
Dosil Santamaría M. Ozamiz-Etxebarria N. Redondo Rodríguez I. Jaureguizar Alboniga-Mayor J. Picaza Gorrotxategi M. Impacto psicológico de la COVID-19 en una muestra de profesionales sanitarios españoles Revista de Psiquiatría y Salud Mental 14 2 2021 106 112 10.1016/j.rpsm.2020.05.004
Ehman J.W. Ott B.B. Short T.H. Ciampa R.C. Hansen-Flaschen J. Do patients want physicians to inquire about their spiritual or religious beliefs if they become gravely ill? Archives of Internal Medicine 159 15 1999 1803 1806 10.1001/archinte.159.15.1803 10448785
El Arnaout N. Chehab R.F. Rafii B. Alameddine M. Gender equity in planning, development and management of human resources for health: a scoping review Human Resources for Health 17 1 2019 52 10.1186/s12960-019-0391-3 31296235
Espinha D.C.M. de Camargo S.M. Silva S.P.Z. Pavelqueires S. Lucchetti G. Nursing students’ opinions about health, spirituality and religiosity Revista gaucha de enfermagem 34 4 2013 98 106 10.1590/s1983-14472013000400013 25080706
Ferrell B.R. Handzo G. Picchi T. Puchalski C. Rosa W.E. The Urgency of Spiritual Care: COVID-19 and the Critical Need for Whole-Person Palliation Journal of Pain and Symptom Management 60 3 2020 e7 e11 10.1016/j.jpainsymman.2020.06.034
Fountoulakis K.N. Apostolidou M.K. Atsiova M.B. Filippidou A.K. Florou A.K. Gousiou D.S. Katsara A.R. Mantzari S.N. Padouva-Markoulaki M. Papatriantafyllou E.I. Sacharidi P.I. Tonia A.I. Tsagalidou E.G. Zymara V.P. Prezerakos P.E. Koupidis S.A. Fountoulakis N.K. Chrousos G.P. Self-reported changes in anxiety, depression and suicidality during the COVID-19 lockdown in Greece Journal of Affective Disorders 279 2021 624 629 10.1016/j.jad.2020.10.061 33190113
Galbadage T. Peterson B.M. Wang D.C. Wang J.S. Gunasekera R.S. Biopsychosocial and Spiritual Implications of Patients With COVID-19 Dying in Isolation Frontiers in Psychology 11 2020 588623 10.3389/fpsyg.2020.588623
Giorgi A. The Theory, Practice, and Evaluation of the Phenomenological Method as a Qualitative Research Procedure Journal of Phenomenological Psychology 28 2 1997 235 260 10.1163/156916297X00103
Gordon B.S. Keogh M. Davidson Z. Griffiths S. Sharma V. Marin D. Mayer S.A. Dangayach N.S. Addressing spirituality during critical illness: A review of current literature Journal of Critical Care 45 2018 76 81 10.1016/j.jcrc.2018.01.015 29413727
Guirao Goris J.A. La espiritualidad como dimensión de la concepción holística de salud ENE.Revista de Enfermería 7 7 2013 1 https://dialnet.unirioja.es/servlet/articulo?codigo=4422394&orden=417305&info=link%0Ahttps://dialnet.unirioja.es/servlet/extart?codigo=4422394
Higginbottom G.M.A. Sampling issues in qualitative research Nurse Researcher 12 1 2004 7 19 10.7748/nr2004.07.12.1.7.c5927
Ho J.Q. Nguyen C.D. Lopes R. Ezeji-Okoye S.C. Kuschner W.G. Spiritual Care in the Intensive Care Unit: A Narrative Review Journal of Intensive Care Medicine 33 5 2018 279 287 10.1177/0885066617712677 28604159
Kim S.C. Quiban C. Sloan C. Montejano A. Predictors of poor mental health among nurses during COVID-19 pandemic Nursing Open 8 2 2021 900 907 10.1002/nop2.697 33570266
Koenig H.G. Religion, Spirituality, and Health: The Research and Clinical Implications ISRN Psychiatry 2012 2012 1 33
Kørup A.K. Søndergaard J. Alyousefi N.A. Lucchetti G. Baumann K. Lee E. Karimah A. Ramakrishnan P. Frick E. Büssing A. Schouten E. Butcher W. Hefti R. Wermuth I. Hvidt N.C. The International NERSH Data Pool of Health Professionals’ Attitudes Toward Religiosity and Spirituality in 12 Countries Journal of Religion and Health 60 1 2021 596 619 10.1007/s10943-020-01077-6 32776266
Lewinson L.P. McSherry W. Kevern P. Spirituality in pre-registration nurse education and practice: A review of the literature Nurse Education Today 35 6 2015 806 814 10.1016/j.nedt.2015.01.011 25707759
Lucchetti G. Góes L.G. Amaral S.G. Ganadjian G.T. Andrade I. Almeida P.O.d.A. do Carmo V.M. Manso M.E.G. Spirituality, religiosity and the mental health consequences of social isolation during Covid-19 pandemic International Journal of Social Psychiatry 67 6 2021 672 679 33135559
Lucchetti G. Lucchetti A.L.G. Spirituality, Religion, and Health: Over the Last 15 Years of Field Research (1999–2013) The International Journal of Psychiatry in Medicine 48 3 2014 199 215 10.2190/PM.48.3.e 25492714
Mahamid F.A. Bdier D. The Association Between Positive Religious Coping, Perceived Stress, and Depressive Symptoms During the Spread of Coronavirus (COVID-19) Among a Sample of Adults in Palestine: Across Sectional Study Journal of Religion and Health 60 1 2021 34 49 10.1007/s10943-020-01121-5 33389439
Pub. L. No. 430/2020 2020 https://www.boe.es/eli/es/rd/2020/03/14/463/con
Moser A. Korstjens I. Series: Practical guidance to qualitative research. Part 3: Sampling, data collection and analysis The European Journal of General Practice 24 1 2018 9 18 10.1080/13814788.2017.1375091 29199486
Nyström M. Dahlberg K. Pre-understanding and openness – a relationship without hope? Scandinavian Journal of Caring Sciences 15 4 2001 339 346 10.1046/j.1471-6712.2001.00043.x 12453176
Ortega Jiménez M. Ojeda Vargas M.G. Ortiz Domínguez A. Guerrero Castañeda R.F. Spirituality and nursing care Espiritualidade e cuidados de enfermagem Revista Iberoamericana de Las Ciencias Sociales y Humanísticas 5 10 2016 231 https://www.ricsh.org.mx/index.php/RICSH/article/view/86
Osório I.H.S. Gonçalves L.M. Pozzobon P.M. Gaspar Júnior J.J. Miranda F.M. Lucchetti A.L.G. Lucchetti G. Effect of an educational intervention in “spirituality and health” on knowledge, attitudes, and skills of students in health-related areas: A controlled randomized trial Medical Teacher 39 10 2017 1057 1064 10.1080/0142159X.2017.1337878 28641027
Peteet J.R. Balboni M.J. Spirituality and religion in oncology CA: A Cancer Journal for Clinicians 63 4 2013 280 289 10.3322/caac.21187 23625473
Puchalski C.M. Larson D.B. Developing curricula in spirituality and medicine Academic Medicine : Journal of the Association of American Medical Colleges 73 9 1998 970 974 10.1097/00001888-199809000-00015 9759099
Rashid M. Kromah F. Cooper C. Blood transfusion and alternatives in Jehovah’s Witness patients Current Opinion in Anaesthesiology 34 2 2021 125 130 10.1097/ACO.0000000000000961 33577206
Riahi S. Goudarzi F. Hasanvand S. Abdollahzadeh H. Ebrahimzadeh F. Dadvari Z. Assessing the Effect of Spiritual Intelligence Training on Spiritual Care Competency in Critical Care Nurses Journal of Medicine and Life 11 4 2018 346 354 30894893
Romero-García M. Delgado-Hito P. Gálvez-Herrer M. Ángel-Sesmero J.A. Velasco-Sanz T.R. Benito-Aracil L. Heras-La Calle G. Moral distress, emotional impact and coping in intensive care unit staff during the outbreak of COVID-19 Intensive and Critical Care Nursing 70 2022 103206 10.1016/j.iccn.2022.103206
Ronaldson S. Hayes L. Aggar C. Green J. Carey M. Spirituality and spiritual caring: nurses’ perspectives and practice in palliative and acute care environments Journal of Clinical Nursing 21 15–16 2012 2126 2135 10.1111/j.1365-2702.2012.04180.x 22788554
Santana Biondo C. Oliveira M. Ferraz A. Lucia M. Silva M. Yarid S.D. Spirituality in urgent and emergency services Rev. Bioét. (Impr.) 25 3 2017 596 602 10.1590/1983-80422017253216
Sebastián Rocchetti N. Andrea Colautti M. Inés Amarilla D. Rovere M. Efectos colaterales de la pandemia por Covid-19 para la salud pública Revista Española de Salud Pública 94 1 2020 e1 e2 https://www.sanidad.gob.es/biblioPublic/publicaciones/recursos_propios/resp/revista_cdrom/VOL94/CARTA/RS94C_202010117.pdf
Swinton M. Giacomini M. Toledo F. Rose T. Hand-Breckenridge T. Boyle A. Woods A. Clarke F. Shears M. Sheppard R. Cook D. Experiences and Expressions of Spirituality at the End of Life in the Intensive Care Unit American Journal of Respiratory and Critical Care Medicine 195 2 2017 198 204 10.1164/rccm.201606-1102OC 27525361
Tong A. Sainsbury P. Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups International Journal for Quality in Health Care 19 6 2007 349 357 10.1093/intqhc/mzm042 17872937
Urra E. Muñoz A. Peña J. El análisis del discurso como perspectiva metodológica para investigadores de salud (Discourse analysis as a methodological perspective for health researchers) Enfermería Universitaria 10 2 2013 50 57 10.1016/S1665-7063(13)72629-0
Vasconcelos A.P.S.L. Lucchetti A.L.G. Cavalcanti A.P.R. da Silva Conde S.R.S. Gonçalves L.M. do Nascimento F.R. Chazan A.C.S. Tavares R.L.C. da Silva Ezequiel O. Lucchetti G. Religiosity and Spirituality of Resident Physicians and Implications for Clinical Practice-the SBRAMER Multicenter Study Journal of General Internal Medicine 35 12 2020 3613 3619 32815055
Veloza-Gómez M. Muñoz de Rodríguez L. Guevara-Armenta C. Mesa-Rodríguez S. The Importance of Spiritual Care in Nursing Practice Journal of Holistic Nursing : Official Journal of the American Holistic Nurses’ Association 35 2 2017 118 131 10.1177/0898010115626777 26873925
World Health Organization. (2021). WHO Coronavirus Disease (COVID-19) Dashboard. Coronavirus Disease (COVID-19) Pandemic. https://covid19.who.int
| 0 | PMC9742223 | NO-CC CODE | 2022-12-14 23:29:59 | no | Intensive Crit Care Nurs. 2022 Dec 12;:103373 | utf-8 | Intensive Crit Care Nurs | 2,022 | 10.1016/j.iccn.2022.103373 | oa_other |
==== Front
Vaccine
Vaccine
Vaccine
0264-410X
1873-2518
Elsevier Ltd.
S0264-410X(22)01526-2
10.1016/j.vaccine.2022.12.010
Article
Social cognitive predictors of vaccination status, uptake and mitigation behaviors in the Canadian COVID-19 Experiences survey
Hall Peter A. a⁎
Meng Gang b
Boudreau Christian c
Hudson Anna a
Quah Anne C.K. b
Agar Thomas b
Fong Geoffrey T. abd⁎
a School of Public Health Sciences, University of Waterloo, Waterloo, Ontario, Canada
b Department of Psychology, University of Waterloo, Waterloo, Ontario, Canada
c Department of Statistics and Actuarial Science, University of Waterloo, Waterloo, Ontario, Canada
d Ontario Institute for Cancer Research, Toronto, Ontario, Canada
⁎ Corresponding authors at: School of Public Health Sciences, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada. Department of Psychology, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada.
12 12 2022
12 12 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.
Emerging infectious diseases like COVID-19 will remain a concern for the foreseeable future, and determinants of vaccination and other mitigation behaviors are therefore critical to understand. Using data from the first two waves of the Canadian COVID-19 Experiences Survey (CCES; N = 1,958; 66.56 % female), we examined social cognitive predictors of vaccination status, transition to acceptance and mitigation behaviors in a population-representative sample. Findings indicated that all social cognitive variables were strong predictors of mitigation behavior performance at each wave, particularly among unvaccinated individuals. Among those who were vaccine hesitant at baseline, most social cognitive variables predicted transition to fully vaccinated status at follow-up. After controlling for demographic factors and geographic region, greater odds of transitioning from unvaccinated at CCES Wave 1 to fully vaccinated at CCES Wave 2 was predicted most strongly by a perception that one’s valued peers were taking up the vaccine (e.g., dynamic norms (OR = 2.13 (CI: 1.54,2.93)), perceived effectiveness of the vaccine (OR = 3.71 (CI: 2.43,5.66)), favorable attitudes toward the vaccine (OR = 2.80 (CI: 1.99,3.95)), greater perceived severity of COVID-19 (OR = 2.02 (CI: 1.42,2.86)), and stronger behavioral intention to become vaccinated (OR = 2.99 (CI: 2.16,4.14)). As a group, social cognitive variables improved prediction of COVID-19 mitigation behaviors (masking, distancing, hand hygiene) by a factor of 5 compared to demographic factors, and improved prediction of vaccination status by a factor of nearly 20. Social cognitive processes appear to be important leverage points for health communications to encourage COVID-19 vaccination and other mitigation behaviors, particularly among initially hesitant members of the general population.
==== Body
pmc1 Introduction.
During the first 18 months of the COVID-19 pandemic, public health officials recommended vaccination, social distancing, masking and enhanced hand hygiene throughout North America [1]. Despite the presence of relatively uniform public health recommendations, compliance remained variable even within geographic regions. Determinants of this variability are important to understand because of the long-term threat posed by COVID-19 and given the prospect of future infectious disease spillover events due to climate change [2]. Demographic factors, political orientation [3], [4], [5], [6], and social cognitive factors [7], [8], [9], [10], [11], [12], [13], [14], [15], [16] have all received significant attention as predictors of vaccine acceptance and other COVID-19 mitigation behaviors [17], [18], [19], [20], [21], [22], [23]. However, in relation to the latter, there is a possibility that the potency of any given predictor varies as a function of vaccination status: those who are vaccinated may have more positive attitudes and beliefs toward mitigation behaviors, see mitigation behaviors as more accepted by others around them, and perceive the risk of COVID-19 as more severe than their unvaccinated counterparts. It also may be the case that the amount of variability in each social cognitive measure (and mitigation behavior outcome) may be more limited in one group or the other, due to floor or ceiling effects within vaccination status groups. Such possibilities might render any given social cognitive variable more or less predictive of a target mitigation behavior across vaccinated or vaccine hesitant populations. Knowing the predictive power of social cognitive constructs within unvaccinated groups is particularly important, given that other COVID-19 mitigation behaviors may be important targets for public health messaging in this group.
The ability to examine the relative magnitude of any given predictor of COVID-19 mitigation behavior is hampered in most samples because of the smaller pool of unvaccinated individuals, making any parameter estimate more unreliable and subject to random fluctuation, in comparison to those in the vaccinated group. Any observed differences in the absolute value or statistical significance of observed predictive coefficients between the groups would be as likely to reflect the sample size difference as any true population difference (between vaccinated and unvaccinated individuals). The use of quota sampling addresses this problem [24], and allows for more straightforward exploration differences—or lack thereof—between vaccinated and unvaccinated subgroups, and also allows for maximally powered prediction of vaccination status itself from the focal predictors. The Canadian COVID-19 Experiences Project (CCEP) is comprised of two functionally interconnected studies, one of which is a Canadian national population cohort survey (Canadian COVID-19 Experiences Survey (CCES)), employing quota sampling to ensure equal numbers of vaccinated (50.2 %) and unvaccinated (43.3 % with no vaccines; 5.5 % with one vaccine with no intention for follow-up vaccinations) individuals [25]. The CCES is weighted to ensure representativeness of the larger population from which it is drawn [26], thereby ensuring that the coefficients and other parameter estimates can be inferred in relation to the Canadian population. The content of the CCES includes an extensive set of predictors of COVID-19 relevant behavior, based on a number of key theories in social and health psychology [27], [28].
The purpose of the current study was to examine the relationship between social cognitive factors and frequency of mitigation behaviors over the first two waves of the CCEP. A secondary purpose was to examine the extent to which any such relationships might vary as a function of vaccination status. Finally, for descriptive purposes, we attempted to quantify the total predictive power of theory-driven social cognitive constructs over and above sociodemographic variables as a group.
2 Methods
2.1 Participants and Procedure.
Wave 1 of the Canadian COVID-19 Experiences Survey (CCES) was conducted between September 28 and October 21, 2021; CCES Wave 2 took place approximiately 6 months later, between March 3 and March 21, 2022 (University of Waterloo, 2022). At Wave 1, 1,958 vaccine-hesitant (49.8 %) and fully vaccinated (50.2 %) members of the general population were recruited using the Leger survey panel. Quota sampling was employed among the Leger Opinion panel (an extensive Canadian national panel, whose recruitment procedures utilized probability sampling), to ensure representation of vaccinated and unvaccinated individuals. Of the initial cohort 17.0 % were aged 18–24 years, 40.2 % were aged 25–39 years, and 42.8 % were aged 40–54 years; 61.2 % were female and 25.2 % were of non-white ethnicity. At Wave 2, replenishment of the cohort was conducted from the same sampling plan to replace cohort members lost to attrition. The analyses presented below differ in the sample sizes in accordance with the respondents at Wave 1 and at Wave 2. During Wave 1 of the CCES, the dominant variant of concern in North America was Delta; during Wave 2, the dominant variant was Omicron [29].
Informed consent was obtained from all participants prior to the collection of data. This study complies with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008; the research protocol was approved by the Human Research Ethics Board at the University of Waterloo Office of Research Ethics.
2.2 Measures.
Measures were selected based on inclusion in one or more social cognitive models of health behavior, such as the Theory of Reasoned Action/Planned Behavior [28] or the Health Belief Model [33]. Social cognitive variables included three measures of social norms (descriptive, injunctive, dynamic), intention strength, perceived severity and threat, as well as attitudes. Additional detail pertaining to each of these measures can be found in the protocol paper [25] and at the following link: https://uwaterloo.ca/prevention-neuroscience-lab/.
2.3 Social norms
Social norms are perceptions of the respondent as to what individuals think or do around them, as well as perceptions about what others might desire of them. Three types of social norms were measured in the current study: dynamic norms, descriptive norms, and injunctive norms. These items were preceded in the survey with a request for each respondent to think of 5 influential people from their everyday life, and to list their names. Subsequent items below refer back to these individuals, when asking the respondent to estimate what others might be doing or thinking around them.
Descriptive norms. Descriptive norms are subjective estimates from the respondent of the number of people who have likely performed a given target behavior. Participants responded to a set of questions assessing vaccination (example, “Of your five most important people, how many of them have gotten FULLY VACCINATED?”, “Of your five most important people, how many of them have gotten just ONE vaccine shot?”, and “Of your five most important people, how many of them have gotten ZERO vaccine shots?”), social distancing (“Of your five most important people, how many consistently comply with social distancing when the government calls for it?”), mask wearing (“Of your five most important people, how many consistently comply with wearing masks when the government calls for it?”), and hand hygiene (“Of your five most important people, how many consistently engage in frequent and thorough handwashing?”). Participants responded using the following response options: 0=“None of them”, 1=“One”, 2=“Two”, 3=“Three”, 4=“Four”, and 5=“Five”. Cronbach’s alpha for the scale (based on 2 items) was 0.648 for social distancing, 0.412 for masking, and 0.623 for hand hygiene at Wave 1, indicating modest reliability across mitigation behaviors.
Injunctive norms. Injunctive norms are beliefs of the respondent about what others believe about the importance of a target behavior. Participants responded to questions about their five most important people assessing vaccination (“My five most important people believe that it is important to be fully vaccinated.”), social distancing (“My five most important people believe that it is important to practice social distancing when the government calls for it.”), mask wearing (“My five most important people believe that it is important to wear masks when the government calls for it.”), and hand hygiene (“My five most important people believe that it is important to wash hands frequently and thoroughly”). The scale was based on 3 items for each target behavior; Cronbach’s alpha for the scale (based on 3 items) was 0.877 for social distancing, 0.864 for masking, and 0.847 for hand hygiene at Wave 1, indicating good reliability across mitigation behaviors.
Dynamic norms. Dynamic norms refer to the respondent’s perceptions of how the behaviors of others are changing (or not) over time. This type of social norm is important to measure when the baseline level of the target behavior is below the desired level; it has been shown that dynamic norms can be more important than descriptive or injunctive norms in such cases, and generally are informative for the construction of norm-based behavioral interventions[30]. Participants responded to a pair of questions assessing vaccination (example: “Which statement best describes how the beliefs of YOUR 5 MOST IMPORTANT PEOPLE about VACCINES have changed in the PAST 4 MONTHS?”), social distancing (“Which statement best describes how the beliefs of your 5 most important people about social distancing have changed in the past 4 months?”), mask wearing (“Which statement best describes how the beliefs of your 5 most important people about wearing masks have changed in the past 4 months?”), and hand hygiene (“Which statement best describes how the beliefs of your five most important people about frequent and thorough handwashing have changed in the last 4 months?”). Participants responded using the following response options: 1 = Fewer of them believe that it is important to practice social distancing/mask wearing/hand washing to reduce the spread of COVID-19, 2 = There is no change in the number that believe that it is important to practice social distancing/mask wearing/hand washing to reduce the spread of COVID-19, and 3 = More of them believe that it is important to be fully vaccinated/practice social distancing/mask wearing/hand washing to reduce the spread of COVID-19. Cronbach’s alpha for the scale (based on 2 items) was 0.884 for social distancing, 0.877 for mask wearing, and 0.877 for hand hygiene, indicating good reliability across mitigation behaviors.
2.4 Intention strength
Behavioral intentions are the amount of personal effort or resources that the respondent is prepared to expend in order to engage in a behavior; it can be seen as a proxy for motivational strength. Participants who fell within the hesitant group of respondents were asked the following question about their intention to get fully vaccinated in the future (“What best describes your intention to get fully vaccinated in the next 4 months?”) using the following response options: 1= “No intention to get fully vaccinated in the next 4 months”, 2= “A very low intention”, 3= “A low intention”, 4= “A moderate intention”, 5= “A strong intention”, and 6= “A very strong intention”. Participants who had received their first vaccine dose responded to a question about their vaccination intention regarding a second dose (“What best describes your intention to get your next shot?”) using the following response options: 1= “I have received the one-shot only vaccine (Johnson & Johnson)”, 2=“I have NO plan to get a second shot”, 3=“I am unsure whether I will get the second shot”, 4=“I plan to get the second shot, but have NOT yet scheduled an appointment”, and 5=“I am planning to get the second shot and have scheduled an appointment”. No intention item was provided for those already having been classified as fully vaccinated (i.e., either 1 Johnson & Johnson or 2 mRNA shots), given that no additional shot was available—or believed to be on the horizon—at this early phase of the pandemic.
2.5 Perceived effectiveness (of target behavior).
Perceived effectiveness refers to an individual’s subjective appraisal of how effective a given precaution might be in offsetting an unwanted outcome. Participants responded to questions assessing the perceived effectiveness of vaccination (“Being fully vaccinated is an effective way of preventing serious infection and death from COVID-19.”), social distancing (“Social distancing is an effective way to prevent the spread of COVID-19.”), mask wearing (“If worn properly, masks can protect the wearer from getting infected by COVID-19.” and “If worn properly, masks can protect other people from getting infected by COVID-19.”), and hand washing (“Frequent and thorough handwashing is important to protect those who cannot be vaccinated (e.g., children)”). Participants responded using the following response options: 1=“strongly disagree” to 5=“strongly agree”, with 3=“neither agree nor disagree”. Cronbach’s alpha of the scale (based on 2 items) was 0.826 for social distancing, 0.845 for masking, and 0.778 for hand hygiene, indicating good reliability across mitigation behaviors.
2.6 Perceived severity (of COVID-19)
Perceived severity is the respondent’s subjective appraisal of the severity of the threat, in this case, of COVID-19. Participants responded to questions about perceived severity (“COVID-19 is an extremely serious threat to public health.” and “COVID-19 is no worse than the flu”) using the following response options: 1=“strongly disagree” to 5=“strongly agree”, with 3=“neither agree nor disagree”.
Participants responded to questions on perceived severity for people their age (“How severe do you think catching COVID-19 is for people your age who get infected?”) using the following response options: 1=“not at all severe” to 5=“extremely severe” and for them compared to others of the same age (“If you caught COVID-19, how severe do you think the illness would be for you, COMPARED TO others your age who caught it?” using the following response options: 1=“a lot less severe for me than others my age” to 7=“A lot more severe for me than for others my age”. Cronbach’s alpha for the perceived severity scale (based on 4 items) was 0.686 for social distancing, 0.686 for mask wearing, and 0.686 for hand hygiene, indicating good reliability across mitigation behaviors.
2.7 Attitudes
Attitudes are the extent to which behaviors are viewed positively or negatively. A single attitudes item was assessed using the following form, “What is your overall opinion of [target behavior] as a way of preventing COVID-19?” Responses were given on a 1 to 6 scale, wherein: 1 = Very positive. 2 = Positive, 3 = Neither positive nor negative, 4 = Negative, 5 = Very negative. Options were also offered for “refused” or “don’t know.”.
2.8 Worry
Worry refers to the presence of recurring negative thought about a perceived threat, normally accompanied by an anxious emotional state and physiological arousal. According to the Health Belief Model, worry is a proxy for perceived threat of a disease (in this case, COVID-19). As such, worry is hypothesized to be one of the critical motivators of protective behaviors, such as masking, social distancing and hand hygiene, as well as vaccinations. Six items were included, for example: “How worried are you that a family member or a close friend will get INFECTED by [SARS-CoV-2]?” and “How worried are you that you will get very sick from COVID-19?” Responses were given on the following scale, 1 = Not at all worried, 2 = Slightly worried, 3 = Moderately worried, 4 = Very worried, 5 = Extremely worried. Options were also offered for “refused” or “don’t know.” The Cronbach’s alpha for the worry measure (based on 6 items) was 0.960.
2.9 Mitigation behaviors
Participants responded to questions assessing social distancing (“How consistently do you follow the recommendations by your local or provincial public health officials about social distancing?” and “How often do you currently avoid INDOOR public places where you would expect a lot of people?”), mask wearing (“How consistently do you follow the recommendations by your local or provincial public health officials about mask wearing?”), and hand washing (“How consistently do you follow the recommendations by your local or provincial public health officials about handwashing?”). Participants responded using the following response options: 1=“I do not follow the recommendations at all”, 2=“I rarely follow the recommendations”, 3=“I sometimes follow the recommendations”, 4=“I follow the recommendations most of the time”, 5=“I follow the recommendations all the time or nearly all the time”, 6=“I go above and beyond the recommendations”.
Participants responded to questions assessing social distancing (“When outside your home, how consistently do you currently maintain a distance from others of at least 2 m?”), mask wearing (“How often do you currently wear a mask when you are in INDOOR public places?”), and hand washing (“How often when washing your hands during the day do you thoroughly wash to the standards recommended by your local or provincial health officials?”). Participants responded using the following response options: 1=“Not at all”, 2=“Rarely”, 3=“Sometimes”, 4=“Most of the time”, 5=“All of the time”, with option 6 (“I haven't had contact with others”) for social distancing and (“I am never in indoor public places”) for mask wearing.
2.10 Vaccination status
Vaccination status was was assessed using the following item, “Have you received any COVID-19 vaccine shots?” Responses were provided using the following response scales: “I have NOT received any vaccine shot,” “I have received ONE vaccine shot,” or “I have received TWO vaccine shots.” At the time of the survey, the second dose of the primary-two-shot primary series had been available for several months in all parts of Canada, and no booster shot (dose 3) had been yet approved. From this perspective, two shot vaccination would be considered “up-to-date” as of the timing of the study.
2.11 Demographic factors
Age in years, gender, combined household income (income), education, and ethnicity were self-reported by each respondent. Years of formal education was reduced into three finite categories as follows: Low=”Grade school/some high school”, “completed high school”; Moderate=”technical/trade school or community college”, “some university, no degree”; High=”completed university degree”, “post-graduate degree.”.
2.12 Sampling and sample weighting procedures:
Respondents were recruited via the Leger Opinion (LEO) Web panel. The Wave 1 sampling design consisted of an overall sample of 2,002 respondents, which comprised of two sub-samples of approximately 1,000 respondents each. The first sub-sample consisted of vaccine hesitant individuals – defined as having received no COVID-19 vaccination, or having received one shot of a two-shot COVID-19 vaccination with no plan, or being unsure, about whether a second shot will be received; whereas the second sub-sample consisted of vaccinated individuals – defined as having two shots of an approved COVID- 19 two-dose vaccine. At the time of Wave 1, the Johnson & Johnson’s one-dose Janssen vaccine was unavailable, therefore any respondents who claimed to be vaccinated with that vaccine were excluded from the analysis. Within each sub-sample, respondents were randomly selected and invited via email to complete the CCEP survey. The Wave 2 sampling design consisted in recontacting as many Wave 1 respondents as possible, and replenish those lost to follow-up as to maintain (as much as possible) the targeted/desired sub-sample size of 1,000 respondents each. Cross-sectional sampling weights were computed for the 1,958 respondents (983 fully vaccinated and 975 vaccine-hesitant) who completed Wave 1 of the CCEP survey; whereas, both cross-sectional and longitudinal weights were computed at Wave 2. The Wave 2 cross-sectional weights were calculated for the 1,783 respondents (980 fully vaccinated and 803 vaccine-hesitant) interviewed at Wave 2, and the Waves 1–2 longitudinal weights were calculated for the 1,109 respondents that were recruited at Wave 1 and successfully retained at Wave 2. For both the Waves 1 and 2 cross-sectional weights, respondents were first divided into two groups: fully vaccinated and vaccine-hesitant. Within each group, respondents were further subdivided based on gender × age and geographic region/language. To be precise, respondents were divided into 4 age groups (i.e., 18–24, 25–34, 35–44 and 45–54) and into 8 geographic region/language groups (i.e., the Maritimes, Québec French-speaking, Québec English speaking, Ontario, Manitoba & Saskatchewan, Alberta and British Columbia). Population totals from the 2016 Census were combined with the CCE disposition codes to obtain benchmark/calibration figures (e.g., estimated number of fully vaccinated 18–25 years old residing in the province of Ontario) for each subgroup. Separately for each of the 2 groups, a raking procedure was applied to calibrate the weights based on gender × age and geographic region/language. Finally, the weights were rescaled to sum to sample size. These cross-sectional weights are designed to make respondents of the CCEP survey representative of the Canadian population of fully vaccinated and vaccine-hesitant adults aged 18–54 at the time of data collection. This excludes those who received the single shot Janssen (Johnson & Johnson), as well as those who had only received their first shot, but were planning to get their second shot (regardless of whether or not they had already scheduled their appointment). Finally, the Wave 1–2 longitudinal weights were then rescaled Wave 1 cross-sectional weights adjusted for attrition between Waves 1 and 2. Consequently, computation of those weights essentially followed the same steps as described above for the cross-sectional weights.
2.13 Statistical analyses:
Survey linear/logistic regression models incorporating survey strata and weights were conducted to estimate the level of the three mitigation behaviors (social distancing, mask wearing, and hand-washing) and vaccination status (vaccinated vs non-vaccinated) and their associations with social cognitive predictors. Similar regression models were also conducted to examine the moderation effect of vaccination on the association of mitigation behaviors and social cognitive predictors at Wave 1 and Wave 2. The predictive power of the social cognitive predictors as a group, in comparison to socio-demographic factors, was examined using chi-square and adjusted chi-square tests. Longitudinal analyses (linear/logistic regressions) were performed to examine if the social cognitive predictors at Wave 1 were associated with the transition of vaccine hesitant participants to being fully vaccinated at Wave 2. All regression models controlled for respondents’ gender, age group (18–24, 25–39, and 40–54 years), income level (low, medium, high, and not stated), education level (low, medium, high, and not stated), ethnicity (White, non-White, and not stated), and geographic region (Canadian province). All analyses were conducted in SAS with SUDAAN V11. All confidence intervals and statistical significance were assessed at the 95 % confidence level.
3 Results
Demographic characteristics of the sample are presented in Table 1 ; zero order correlation coefficients are presented in Fig. 1 . Table 2, Table 3, Table 4 present the main effects of each social cognitive variable along with the corresponding interaction term with vaccination status, separately for each COVID-19 mitigation behavior (i.e., social distancing, masking, hand hygiene).Table 1 Demographic characteristics of the sample in Wave 1 and Wave 2 of the CCEP overall and as a function of vaccination status.
Wave 1 Wave 2
Hesitant Fully vaccinated Full Sample Hesitant Fully vaccinated Full Sample
freq % freq % freq % freq % freq % freq %
Vaccination Hesitant 975 49.80 803 44.71
Vaccinated 983 50.2 993 55.29
Gender Male 326 33.44 442 44.96 768 39.22 271 33.75 450 45.32 721 40.14
Female 649 66.56 541 55.04 1190 60.78 532 66.25 543 54.68 1075 59.86
Age 18–24 100 10.26 222 22.58 322 16.45 72 8.97 169 17.02 241 13.42
25–39 463 47.49 326 33.16 789 40.3 363 45.21 371 37.36 734 40.87
40–54 412 42.26 435 44.25 847 43.26 368 45.83 453 45.62 821 45.71
Income Low 185 18.97 120 12.21 305 15.58 155 19.3 117 11.78 272 15.14
Moderate 245 25.13 191 19.43 436 22.27 200 24.91 199 20.04 399 22.22
High 459 47.08 572 58.19 1031 52.66 365 45.45 602 60.62 967 53.84
No answer 86 8.82 100 10.17 186 9.5 83 10.34 75 7.55 158 8.8
Education Low 249 25.54 159 16.17 408 20.84 204 25.4 148 14.9 352 19.6
Moderate 361 37.03 349 35.5 710 36.26 358 44.58 349 35.15 707 39.37
High 344 35.28 466 47.41 810 41.37 238 29.64 493 49.65 731 40.7
No answer 21 2.15 9 0.92 30 1.53 3 0.37 3 0.3 6 0.33
Ethnicity White 753 77.23 678 68.97 1431 73.08 647 80.57 720 72.51 1367 76.11
Non-white 193 19.79 281 28.59 474 24.21 135 16.81 250 25.18 385 21.44
not stated 29 2.97 24 2.44 53 2.71 21 2.62 23 2.32 44 2.45
Region Alberta 147 15.08 98 9.97 245 12.51 97 12.08 108 10.88 205 11.41
BC 116 11.9 128 13.02 244 12.46 84 10.46 129 12.99 213 11.86
MB + SK 62 6.36 57 5.8 119 6.08 49 6.1 57 5.74 106 5.9
Maritimes 56 5.74 55 5.6 111 5.67 46 5.73 61 6.14 107 5.96
Ontario 377 38.67 360 36.62 737 37.64 303 37.73 376 37.87 679 37.81
QC-En 23 2.36 111 11.29 134 6.84 20 2.49 85 8.56 105 5.85
QC-Fr 194 19.9 174 17.7 368 18.79 204 25.4 177 17.82 381 21.21
Fig. 1 Scatterplots showing associations between social cognitive variables and mitigation behaviors at Wave 1. Regression line with 95 % CIs indicated. Sample sizes for vaccinated and unvaccinated are n = 975 and n = 983, respectively. Panel A shows the correlations separately for vaccinated (blue) and unvaccinated (red) groups separately; panel B shows the correlations for the total sample.
Table 2 Social cognitive predictors of distancing in Waves 1 and 2 of the CCEP.
Wave 1 Wave 2
β
[95 % CI] t p β
[95 % CI] t p
Descriptive Norms 0.22
[0.15, 0.29] 6.19 <0.001 0.30
[0.23, 0.37]
8.20 <0.001
Vaccine Hesitant* Descriptive Norms 0.32
[0.23, 0.41] 6.91 <0.001 0.21
[0.11, 0.31]
4.03 <0.001
Injunctive Norms 0.23
[0.16, 0.29] 7.10 <0.001 0.35
[0.29, 0.41]
11.25 <0.001
Vaccine Hesitant* Injunctive Norms 0.27
[0.19, 0.36] 6.23 <0.001 0.17
[0.09, 0.26]
4.06 <0.001
Dynamic Norms 0.09
[0.05, 0.14] 4.05 <0.001 0.20
[0.15, 0.25]
8.06 <0.001
Vaccine Hesitant* Dynamic Norms 0.28
[0.20, 0.36] 6.79 <0.001 0.24
[0.15, 0.33]
5.27 <0.001
Perceived Effectiveness 0.37
[0.31, 0.42] 13.30 <0.001 0.44
[0.38, 0.50]
15.28 <0.001
Vaccine Hesitant* Perceived Effectiveness 0.23
[0.16, 0.31] 5.95 <0.001 0.21
[0.14, 0.29]
5.40 <0.001
Attitudes 0.33
[0.27, 0.38] 11.98 <0.001 0.40
[0.35, 0.46] 14.96 <0.001
Vaccine Hesitant* Attitudes 0.23
[0.16, 0.31] 6.17 <0.001 0.18
[0.11, 0.26]
4.84 <0.001
Perceived Severity 0.47
[0.39, 0.54] 12.10 <0.001 0.53
[0.46, 0.60] 14.95 <0.001
Vaccine Hesitant* Perceived Severity 0.22
[0.11, 0.32] 4.01 <0.001 0.28
[0.18, 0.38]
5.41 <0.001
Worry 0.28
[0.23, 0.33] 12.10 <0.001 0.33
[0.29, 0.38]
14.67 <0.001
Vaccine Hesitant* Worry 0.30
[0.22, 0.38] 7.66 <0.001 0.32
[0.24, 0.41] 7.36 <0.001
Note: The main effect of each social cognitive variable is presented first followed by interactions with vaccination status. Covariates include vaccination status age, gender, non-white ethnicity, income, education and geographic region. All regression coefficients are standardized beta weights; main effects of each target variable are presented first along with interaction with vaccination status; reference category for interaction is vaccinated.
As can be seen in Table 2, all measures of normative influences were reliable predictors of distancing, though more so for descriptive norms and injunctive norms. Perceived severity of COVID-19 and perceived effectiveness of mitigation behaviors were the strongest predictors, however. For mask wearing a similar pattern emerged (Table 3 ), such that descriptive and injunctive norms predicted behavior more consistently than dynamic norms; although the latter was still statistically reliable. Finally, for hand hygiene, which was measured only at Wave 1, there was an identical pattern such that all social cognitive variables were significant predictors (Table 4 ); among the norm measures, descriptive and injunctive norms had the strongest associations with the frequency of hand hygiene. Again, perceived severity of COVID-19 and perceived effectiveness of hand hygiene behaviors were the strongest predictors of hand hygiene frequency.Table 3 Social cognitive predictors of mask wearing in Waves 1 and 2 of the CCEP.
Wave 1 Wave 2
β
[95 % CI] t p β
[95 % CI] t p
Descriptive Norms 0.17
[0.09, 0.24] 4.30 <0.001 0.35
[0.28, 0.42]
9.87 <0.001
Vaccine Hesitant* Descriptive Norms 0.36
[0.25, 0.46] 6.54 <0.001 0.20
[0.10, 0.30]
3.84 <0.001
Injunctive Norms 0.23
[0.16, 0.30] 6.44 <0.001 0.33
[0.27, 0.39]
10.61 <0.001
Vaccine Hesitant* Injunctive Norms 0.29
[0.19, 0.40] 5.52 <0.001 0.29
[0.20, 0.39] 6.19 <0.001
Dynamic Norms 0.07
[0.03, 0.12] 3.06 0.002 0.12
[0.08, 0.17]
5.18 <0.001
Vaccine Hesitant* Dynamic Norms 0.30
[0.22, 0.39] 6.71 <0.001 0.39
[0.29, 0.49]
7.92 <0.001
Perceived Effectiveness 0.38
[0.30, 0.46] 9.52 <0.001 0.43
[0.36, 0.49]
13.77 <0.001
Vaccine Hesitant* Perceived Effectiveness 0.27
[0.17, 0.38] 5.08 <0.001 0.36
[0.27, 0.45]
7.64 <0.001
Attitudes 0.35
[0.29, 0.41] 11.19 <0.001 0.42
[0.37, 0.47]
15.75 <0.001
Vaccine Hesitant* Attitudes 0.27
[0.18, 0.36] 6.06 <0.001 0.32
[0.24, 0.41]
7.42 <0.001
Perceived Severity 0.37
[0.28, 0.47] 7.75 <0.001 0.38
[0.31, 0.45]
11.34 <0.001
Vaccine Hesitant* Perceived Severity 0.26
[0.13, 0.39] 3.82 <0.001 0.44
[0.32, 0.56]
7.11 <0.001
Worry 0.19
[0.14, 0.23] 7.42 <0.001 0.22
[0.18, 0.27]
9.72 <0.001
Vaccine Hesitant* Worry 0.35
[0.26, 0.44] 7.82 <0.001 0.47
[0.37, 0.56] 9.37 <0.001
Note: The main effect of each social cognitive variable is presented first followed by interactions with vaccination status. Covariates include vaccination status age, gender, non-white ethnicity, income, education and geographic region. All regression coefficients are standardized beta weights; main effects of each target variable are presented first along with interaction with vaccination status; reference category for interaction is vaccinated.
Table 4 Social cognitive predictors of hand hygiene in Waves 1 and 2 of the CCEP.
Wave 1 Wave 2
β
[95 % CI] t p β
[95 % CI]
t p
Descriptive Norms 0.32
[0.25, 0.40] 8.42 <0.001
Vaccine Hesitant* Descriptive Norms 0.21
[0.11, 0.32]
4.04 <0.001
Injunctive Norms 0.28
[0.21, 0.35] 7.45 <0.001
Vaccine Hesitant* Injunctive Norms 0.16
[0.05, 0.28]
2.88 0.004
Dynamic Norms 0.15
[0.10, 0.21] 5.29 <0.001
Vaccine Hesitant* Dynamic Norms 0.19
[0.08, 0.29] 3.54 <0.001
Perceived Effectiveness 0.42
[0.35, 0.49] 11.92 <0.001
Vaccine Hesitant* Perceived Effectiveness 0.19
[0.08, 0.29] 3.60 <0.001
Attitudes 0.36
[0.29, 0.42] 11.33 <0.001
Vaccine Hesitant* Attitudes 0.11
[0.02, 0.21] 2.35 0.019
Perceived Severity 0.36
[0.26, 0.46] 7.07 <0.001 0.37
[0.29, 0.46]
8.23 <0.001
Vaccine Hesitant* Perceived Severity 0.10
[-0.04, 0.24] 1.39 0.166 0.27
[0.13, 0.40] 3.93 <0.001
Worry 0.28
[0.22, 0.34] 9.36 <0.001 0.27
[0.22, 0.33]
9.77 <0.001
Vaccine Hesitant* Worry 0.14
[0.01, 0.26] 2.16 0.031 0.28
[0.18, 0.38] 5.64 <0.001
Note: The main effect of each social cognitive variable is presented first followed by interactions with vaccination status. Covariates include vaccination status age, gender, non-white ethnicity, income, education and geographic region. All regression coefficients are standardized beta weights; the reference category for each interaction term is “vaccinated”. Norms, perceived effectiveness and attitudes were included only in Wave 1.
Without exception, coefficients were slightly larger when mitigation behavior measurement was performed at Wave 2 than at Wave 1. Likewise, in all cases, predictive coefficients were significantly greater in magnitude among unvaccinated than among vaccinated individuals, highlighting the potential utility of leveraging such factors in communications to enhance uptake in vaccine hesitant groups.
Finally, we used Wave 1 social cognitive variables to predict transition from vaccine hesitant at Wave 1 to fully vaccinated at Wave 2 (Table 5 ). Again, nearly all social cognitive variables were significant predictors of transition from vaccine hesitant to fully vaccinated. After controlling for demographic factors and geographic region, greater odds of transitioning from unvaccinated at CCES Wave 1 to fully vaccinated at CCES Wave 2 was predicted most strongly by a perception that one’s valued peers were taking up the vaccine (e.g., dynamic norms (OR = 2.13 (CI: 1.54,2.93)), perceived effectiveness of the vaccine (OR = 3.71 (CI: 2.43,5.66)), favorable attitudes toward the vaccine (OR = 2.80 (CI: 1.99,3.95)), greater perceived severity of COVID-19 (OR = 2.02 (CI: 1.42,2.86)), and stronger behavioral intention to become vaccinated (OR = 2.99 (CI: 2.16,4.14)). The only non-significant social cognitive predictor was injunctive norms.Table 5 Predictors of transitions from vaccine hesitant to fully vaccinated, from Wave 1 to Wave 2.
β
OR
(95 % CI) p
Descriptive Norms 0.44 1.56
(1.09, 2.23) 0.016
Injunctive Norms 0.25 1.28
(0.91, 1.80) 0.151
Dynamic Norms 0.75 2.13
(1.54, 2.93) <0.001
Perceived Effectiveness 1.31 3.71
(2.43, 5.66) <0.001
Attitudes 1.03 2.80
(1.99, 3.95) <0.001
Perceived Severity 0.70 2.02
(1.42, 2.86) <0.001
Worry 0.52 1.69
(1.26, 2.25) <0.001
Behavioral Intention 1.10 2.99
(2.16, 4.14) <0.001
Note. OR = odds ratio; covariates include vaccination status age, gender, non-white ethnicity, income education and geographic region.
Social cognitive variables as a group improved prediction of mask wearing, social distancing, and hand hygiene beyond demographic variables by a factor of 5 or more, and improved prediction of vaccination status by a factor of nearly 20 (Fig. 2 ).Fig. 2 Heat maps showing the increment in proportion of variance accounted for by demographic variables (A), social norm variables (B) and all social cognitive variables combined (C). Lighter regions correspond with higher proportions of variance explained for each target mitigation behavior. Top row (1) social distancing, (2) mask wearing, and (3) hand hygiene.
4 Discussion
The current findings indicate that social cognitive variables are important predictors of COVID-19 mitigation behaviors and vaccination. When explaining vaccination, this increment in predictive power was by a factor of nearly 20 beyond demographic factors alone; for mitigation behaviors, the increment was nearly 5-fold. When comparing those who were vaccinated and unvaccinated, social cognitive variables were more important predictors of mitigation behaviors among unvaccinated individuals than among vaccinated individuals. This may reflect a ceiling effect among vaccinated individuals, but also could be indicative of the true potency of thinking styles among those who choose to rely completely on behavior in order to mitigate disease threats, rather than make use of available vaccines.
The significant moderator effects involving vaccination stand in contrast to cognitive predictors, which are largely invariant in their predictive power across vaccinated and unvaccinated groups [31]. This argues against the ceiling effect interpretation, and suggests that higher cognition that is socially and semantically mediated--as opposed to neurobiologically mediated--maybe relatively more influenced by political orientation.
The trend for predictive values of social cognitive measures to be higher (invariantly so) in Wave 2 was unexpected. It remains to be understood fully why this is the case. However, there are a few possibilities. One is that in CCES Wave 1 (September 28, 2021 to October 21, 2021), many more behaviors were under the control of absolute rules (e.g., mandates), leaving little room for individual mental processes to influence decision-making. During Wave 2, which took place between March 3 and March 21, 2022, far fewer such mandates existed, leading to potentially more room for social cognitions to inform and influence decisions about behaviors. Indeed, it is possible to understand mandates as effectively pre-empting decision processes, or at least constraining them significantly, in the interests of required population-wide uniformity. Notably, however, even under such conditions, social cognition matters, as demonstrated in the Wave 1 data, which still shows the highly robust predictive value of nearly every social cognitive variable examined.
Among predictors of transitions from vaccine hesitant to fully vaccinated, the only non-significant effect was for injunctive norms. This is notable to the extent that is suggests that appealing to what others think a target “should” do may be futile. On the other hand, among the social norm variables, dynamic norms were particularly strong predictors of transitions. This suggests that when designing public health communications, the type of normative appeal may matter a lot for vaccine hesitant respondents: appeals to “should” and the threat of social disapproval have little sway, while making salient a normative shift in the behavior of valued peers (toward vaccination) may double the odds of becoming vaccinated.
This being said, perceived severity of COVID-19, perceived effectiveness of vaccines, favorable attitudes towards vaccines and strong intentions to become vaccinated were among the strongest predictors of transition from hesitant to vaccinated overall. For this reason, health communications that highlight effectiveness of vaccines in offsetting risk of morbidity and mortality, and the severity of COVID-19 are important. With respect to the latter, given the age gradient of adverse effects, severity for younger age groups may need to focus more on post-covid syndrome and other adverse outcomes that are more likely for such age groups.
Strengths of the current investigation include the use of a population representative sample, and approximately equal proportions of vaccinated and unvaccinated individuals, thereby maximizing the statistical power to predict vaccination status. As well, the prospective design allowed us to examine changes in vaccination status (from unvaccinated to vaccinated) across the two waves of data collection. Limitations include self-reporting of the outcomes and the reliance on quota sampling, which in some cases may be inferior to stratified random sampling when attempting to equalize sub-group size. Finally, the measure of descriptive norms was only modestly reliable, which may have reduced the observed association between it and other variables.
In conclusion, we examined the predictive power of theoretically-derived social cognitive factors for predicting three outcomes: COVID-19 mitigation behaviors, vaccination status and changes in vaccination status. In all cases, social cognitive predictors improved our ability to predict each outcome, often substantially. Across all predictors, social cognitive variables were most powerfully predictive among unvaccinated individuals, suggesting that theoretically driven communications designed to reach these individuals are likely to have a chance at succeeding. Specifically, for vaccination, communications that highlight shifts among peers toward vaccination (or completing vaccination), accurate communication of the adverse outcomes possible from severe COVID-19, and emphasizing the efficacy of vaccines—may be especially influential for those who are unvaccinated and considering becoming fully vaccinated, and may also serve to enhance adherence to masking, distancing and hand hygiene among unvaccinated individuals. The latter may be important, given that they may engage in lower levels of COVID-19 mitigation behaviors than vaccinated individuals do [32].
Author Contributions
PH and GF conceived the study, planned and oversaw the statistical analyses; PH wrote the final draft. GM planned and completed all statistical analyses and contributed to the writing of the final draft. GF, AH, AQ, TA, and CB contributed to the planning of the study and writing of the final draft.
Data sharing
The datasets used and analyzed during the current study available from the corresponding author on reasonable request.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
This research was supported by an operating grant to P. Hall, G. Fong and S. Hitchman by the Canadian Institutes for Health Research (CIHR), Institute for Population and Public Health (GA3-177733). The authors would like to acknowledge and thank all those that contributed to The Canadian COVID-19 Experiences Project: all study investigators and collaborators, and the project staff at their respective institutions, namely Alkarim Bilawalla, Jessica Lee, Marie Jolicoeur-Becotte, Jamie Wheeler, Kai Jiang, Shelly Jordan, Simon Thompson, Lindsey Webster of the University of Waterloo, Andrew Mattern and Samantha Rochon of Leger Opinion. We wish to also acknowledge the critical input of Dr. Sara Hitchman in the development of the measures of social norms.
==== Refs
References
1 Honein M.A. Christie A. Rose D.A. Brooks J.T. Meaney-Delman D. Cohn A. Summary of Guidance for Public Health Strategies to Address High Levels of Community Transmission of SARS-CoV-2 and Related Deaths, December 2020 Morb Mortal Wkly Rep 69 49 2020 Dec 11 1860 1867
2 Baker R.E. Mahmud A.S. Miller I.F. Rajeev M. Rasambainarivo F. Rice B.L. Infectious disease in an era of global change Nat Rev Microbiol 20 4 2022 Apr 193 205 34646006
3 Gollwitzer A. Martel C. Brady W.J. Pärnamets P. Freedman I.G. Knowles E.D. Partisan differences in physical distancing are linked to health outcomes during the COVID-19 pandemic Nat Hum Behav 4 11 2020 Nov 1186 1197 33139897
4 Hussain B. Latif A. Timmons S. Nkhoma K. Nellums L.B. Overcoming COVID-19 vaccine hesitancy among ethnic minorities: A systematic review of UK studies Vaccine 40 25 2022 May 3413 3432 35534309
5 Newman PA, Reid L, Tepjan S, Fantus S, Allan K, Nyoni T, et al. COVID-19 vaccine hesitancy among marginalized populations in the U.S. and Canada: Protocol for a scoping review. PLOS ONE. 2022 Mar 31;17(3):e0266120.
6 Nguyen L.H. Joshi A.D. Drew D.A. Merino J. Ma W. Lo C.H. Self-reported COVID-19 vaccine hesitancy and uptake among participants from different racial and ethnic groups in the United States and United Kingdom Nat Commun 13 1 2022 Dec 636 35105869
7 Fan C.W. Chen I.H. Ko N.Y. Yen C.F. Lin C.Y. Griffiths M.D. Extended theory of planned behavior in explaining the intention to COVID-19 vaccination uptake among mainland Chinese university students: an online survey study Hum Vaccines Immunother 17 10 2021 Oct 3 3413 3420
8 Gibson L.P. Magnan R.E. Kramer E.B. Bryan A.D. Theory of Planned Behavior Analysis of Social Distancing During the COVID-19 Pandemic: Focusing on the Intention-Behavior Gap Ann Behav Med 55 8 2021 Jul 22 805 812 34228112
9 Shmueli L. Predicting intention to receive COVID-19 vaccine among the general population using the health belief model and the theory of planned behavior model BMC Public Health 21 1 2021 Apr 26 804 33902501
10 Pan J.Y. Liu D. Mask-wearing intentions on airplanes during COVID-19 – Application of theory of planned behavior model Transp Policy 1 119 2022 Apr 32 44
11 Hayashi Y. Romanowich P. Hantula D.A. Predicting Intention to Take a COVID-19 Vaccine in the United States: Application and Extension of Theory of Planned Behavior Am J Health Promot 36 4 2022 May 1 710 713 35041541
12 Preis H, Mahaffey B, Lobel M. Factors related to COVID-19 vaccination intention and uptake among childbearing women. Health Psychol. 2022;No Pagination Specified-No Pagination Specified.
13 Yahaghi R. Ahmadizade S. Fotuhi R. Taherkhani E. Ranjbaran M. Buchali Z. Fear of COVID-19 and Perceived COVID-19 Infectability Supplement Theory of Planned Behavior to Explain Iranians’ Intention to Get COVID-19 Vaccinated Vaccines 9 7 2021 Jul 684 34206226
14 Murphy J. Vallières F. Bentall R.P. Shevlin M. McBride O. Hartman T.K. Psychological characteristics associated with COVID-19 vaccine hesitancy and resistance in Ireland and the United Kingdom Nat Commun 12 1 2021 Dec 29 33397962
15 Mertens G. Lodder P. Smeets T. Duijndam S. Fear of COVID-19 predicts vaccination willingness 14 months later J Anxiety Disord 88 2022 May 102574
16 Schumpe B.M. Van Lissa C.J. Bélanger J.J. Ruggeri K. Mierau J. Nisa C.F. Predictors of adherence to public health behaviors for fighting COVID-19 derived from longitudinal data Sci Rep 12 1 2022 Dec 3824 35264597
17 Ayre J. Cvejic E. McCaffery K. Copp T. Cornell S. Dodd R.H. Contextualising COVID-19 prevention behaviour over time in Australia: Patterns and long-term predictors from April to July 2020 in an online social media sample PLoS One 16 6 2021 Jun 29 e0253930 34185816
18 Bokemper S.E. Cucciniello M. Rotesi T. Pin P. Malik A.A. Willebrand K. Experimental evidence that changing beliefs about mask efficacy and social norms increase mask wearing for COVID-19 risk reduction: Results from the United States and Italy PLoS One 16 10 2021 Oct 11 e0258282 34634089
19 Bonell C, Michie S, Reicher S, West R, Bear L, Yardley L, et al. Harnessing behavioural science in public health campaigns to maintain ‘social distancing’ in response to the COVID-19 pandemic: key principles. 2020;3.
20 Fischer R. Karl J.A. Predicting Behavioral Intentions to Prevent or Mitigate COVID-19: A Cross-Cultural Meta-Analysis of Attitudes, Norms, and Perceived Behavioral Control Effects Soc Psychol Personal Sci 13 1 2022 Jan 1 264 276
21 Gouin J.P. MacNeil S. Switzer A. Carrese-Chacra E. Durif F. Knäuper B. Socio-demographic, social, cognitive, and emotional correlates of adherence to physical distancing during the COVID-19 pandemic: a cross-sectional study Can J Public Health 112 1 2021 Feb 17 28 33464556
22 Peterson L.M. Helweg-Larsen M. DiMuccio S. Descriptive Norms and Prototypes Predict COVID-19 Prevention Cognitions and Behaviors in the United States: Applying the Prototype Willingness Model to Pandemic Mitigation Ann Behav Med 55 11 2021 Nov 1 1089 1103 34487142
23 Prosser A.M.B. Judge M. Bolderdijk J.W. Blackwood L. Kurz T. ‘Distancers’ and ‘non-distancers’? The potential social psychological impact of moralizing COVID-19 mitigating practices on sustained behaviour change Br J Soc Psychol 59 3 2020 653 662 32584437
24 Moser C.A. Quota Sampling J R Stat Soc Ser Gen 115 3 1952 411 423
25 Hall P.A. Fong G.T. Hitchman S.C. Quah A.C.K. Agar T. Meng G. Brain and behavior in health communication: The Canadian COVID-19 Experiences Project Brain Behav Immun - Health 1 22 2022 Jul 100467
26 University of Waterloo Canadian COVID-19 Experiences Project Survey Waves 1 and 2 Technical Report 2022 Jul University of Waterloo Waterloo, Ontario, Canada
27 Sheeran P. Trafimow D. Armitage C.J. Predicting behaviour from perceived behavioural control: Tests of the accuracy assumption of the theory of planned behaviour Br J Soc Psychol 42 3 2003 393 410 14567844
28 Ajzen I. The theory of planned behavior Organ Behav Hum Decis Process 50 1991 179 211
29 Abbasi J. Omicron Has Reached the US—Here’s What Infectious Disease Experts Know About the Variant JAMA 326 24 2021 Dec 28 2460 2462 34870691
30 Sparkman G. Walton G.M. Dynamic Norms Promote Sustainable Behavior, Even if It Is Counternormative Psychol Sci 28 11 2017 Nov 1663 1674 28961062
31 Hudson A. Hall P.A. Hitchman S. Meng G. Fong G.T. Cognitive predictors of COVID-19 mitigation behaviours in vaccinated and unvaccinated general population members Vaccine 2023
32 Hall P.A. Meng G. Sakib M.N. Quah A.C. Agar T. Fong G.T. Do the vaccinated perform less distancing, mask wearing and hand hygiene? A test of the risk compensation hypothesis in a representative sample during the COVID-19 pandemic Vaccine 2023
33 Strecher VJ, Rosenstock IM. The health belief model. Cambridge handbook of psychology, health and medicine. 1997 Sep 25;113:117.
| 0 | PMC9742224 | NO-CC CODE | 2022-12-15 23:17:59 | no | Vaccine. 2022 Dec 12; doi: 10.1016/j.vaccine.2022.12.010 | utf-8 | Vaccine | 2,022 | 10.1016/j.vaccine.2022.12.010 | oa_other |
==== Front
Eur Neuropsychopharmacol
Eur Neuropsychopharmacol
European Neuropsychopharmacology
0924-977X
1873-7862
Elsevier B.V. and ECNP.
S0924-977X(22)00913-0
10.1016/j.euroneuro.2022.12.002
Article
Brain correlates of subjective cognitive complaints in COVID-19 survivors: a multimodal magnetic resonance imaging study
Paolini Marco abc
Palladini Mariagrazia abd⁎
Mazza Mario Gennaro abd
Colombo Federica abd
Vai Benedetta ab
Rovere-Querini Patrizia be
Falini Andrea bf
Poletti Sara ab
Benedetti Francesco ab
a Psychiatry & Clinical Psychobiology, Division of Neuroscience, IRCCS Scientific Institute Ospedale San Raffaele, Milan
b Vita-Salute San Raffaele University, Milan, Italy
c PhD Program in Molecular Medicine, University Vita-Salute San Raffaele, Milan, Italy
d PhD Program in Cognitive Neuroscience, University Vita-Salute San Raffaele, Milan, Italy
e Division of Immunology, Transplantation and Infectious Diseases, IRCCS Scientific Institute Ospedale San Raffaele, Milan, Italy
f Department of Neuroradiology, IRCCS Scientific Institute Ospedale San Raffaele, Milan, Italy
⁎ Corresponding author: Mrs Mariagrazia Palladini, Vita-Salute San Raffaele University, Istituto Scientifico IRCCS Ospedale San Raffaele, Dipartimento di Neuroscienze Cliniche, Via Stamira d'Ancona 20, 20127 Milano, Italy, Tel +39-02-26433156
12 12 2022
12 12 2022
© 2022 Elsevier B.V. and ECNP. 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.
Cognitive impairment represents a leading residual symptom of COVID-19 infection, which lasts for months after the virus clearance. Up-to-date scientific reports documented a wide spectrum of brain changes in COVID-19 survivors following the illness's resolution, mainly related to neurological and neuropsychiatric consequences. Preliminary insights suggest abnormal brain metabolism, microstructure, and functionality as neural under-layer of post-acute cognitive dysfunction. While previous works focused on brain correlates of impaired cognition as objectively assessed, herein we investigated long-term neural correlates of subjective cognitive decline in a sample of 58 COVID-19 survivors with a multimodal imaging approach.
Diffusion Tensor Imaging (DTI) analyses revealed widespread white matter disruption in the sub-group of cognitive complainers compared to the non-complainer one, as indexed by increased axial, radial, and mean diffusivity in several commissural, projection and associative fibres. Likewise, the Multivoxel Pattern Connectivity analysis (MVPA) revealed highly discriminant patterns of functional connectivity in resting-state among the two groups in the right frontal pole and in the middle temporal gyrus, suggestive of inefficient dynamic modulation of frontal brain activity and possible metacognitive dysfunction at rest.
Beyond COVID-19 actual pathophysiological brain processes, our findings point toward brain connectome disruption conceivably translating into clinical post-COVID cognitive symptomatology. Our results could pave the way for a potential brain signature of cognitive complaints experienced by COVID-19 survivors, possibly leading to identify early therapeutic targets and thus mitigating its detrimental long-term impact on quality of life in the post-COVID-19 stages.
Keywords
COVID-19
Magnetic resonance imaging
Subjective cognitive dysfunction
Diffusion Tensor Imaging
Resting-state
Functional connectivity
==== Body
pmc1 Introduction
As the world moves into the third year of COVID-19 pandemic, much effort is expended in further defining mental health clinical patterns exhibited by COVID-19 survivors, their endurance, and their putative pathophysiological mechanisms. Scientific and clinical data are evolving on the residual symptoms of the disease persisting in the long run, possibly linked to the COVID-19 multiple organ involvement (Jiang et al., 2020, Nalbandian et al., 2021). Although the direct neurotropism of the SARS-CoV-2 has not been confirmed to date (Spudich and Nath, 2022), there are heeds regarding COVID-19-induced brain demyelinating lesions (Zanin et al., 2020), limbic and subcortical hypo metabolism in extensive areas, along with abnormal functional intra- and inter-network connectivity in survivors experiencing neuropsychiatric sequelae (Zanin et al., 2020, Voruz et al., 2022, Zhang et al., 2022). According to the most favoured hypothesis, several factors entailing dysregulated peripheral immune system activation and prompted neuroinflammation, coagulopathy, and endothelial dysfunction may critically affect brain's morphology and functionality via indirect pathways (O'Shea et al., 2021, Tang et al., 2021, Tang et al., 2022). A key concern is the long-lasting cognitive consequences of the infection (Mazza et al., 2020, Mazza et al., 2021, Poletti et al., 2021), representing a critical and debilitating feature of the so-called Post-acute COVID-19 syndrome (PACS) (Nalbandian et al., 2021, Hampshire et al., 2021). Most strikingly, consistent evidence highlights an impressive rate of COVID-19 patients who still exhibit impaired cognitive performance six months after the virus clearance, showing a constant impaired cognitive profile over time (Poletti et al., 2021). Beyond cognitive deficits as evaluated through standardized batteries, recent investigations pinpoint a high prevalence of clinically relevant subjective cognitive failure among COVID-19 survivors in the months following the infection (Miskowiak et al., 2021). Notably, cognitive decline as reported by COVID-19 patients does not seem to be affected by the severity of acute pneumonia (Andersson et al., 2022); rather, existing data showed that subjective cognitive deficits in COVID-19 survivors (Ferrando et al., 2022) are associated with anxiety and depressive symptomatology and lower neuropsychological performance (Gouraud et al., 2021). These symptoms persist even several months after the acute infection, suggesting that the perception of cognitive complaints is a sign of long-term actual cognitive deficits. In keeping with this, it has been documented that most of COVID-19 patients exhibit intercorrelated objective and subjective cognitive impairments, altogether worsening quality of life (Miskowiak et al., 2021).
Given the well-documented altered brain patterns underlying neuropsychological deficits, it is crucial to further explore potential neural endophenotypes of cognitive impairments experienced by COVID-19 patients. Despite being also a hallmark of PACS, both when objectively measured or subjectively reported by the patients (Ceban et al., 2022, Schou et al., 2021), studies exploring brain correlates of post-COVID cognitive impairment are sparse.
Novel insights provide evidence of altered brain metabolism, microstructure, and functionality in COVID-19 survivors exhibiting cognitive deficiency, possibly underpinning acute and post-acute cognitive dysfunction in patients recovering from the illness (Voruz et al., 2022, Silva et al., 2021, Voruz et al., 2022, Yesilkaya et al., 2021). However, available investigations focused on standardized measures of cognitive functioning, and most of them, with cross-sectional design, cannot effectively disentangle between deficits acquired after COVID-19 and pre-infection cognitive status (Hampshire et al., 2021, Silva et al., 2021).
In our study we are now focusing on investigating long-term neural correlates of subjective cognitive decline in a sample of COVID-19 survivors studied with a multimodal brain imaging approach. Subjective perception of cognitive impairments may identify an early and possibly more sensitive sign of emerging deficits, although subjective difficulties could also reflect levels of depressive symptoms (Ott et al., 2016, Petersen et al., 2019). Specifically, we compared patients with and without subjective cognitive complaints on Diffusion Tensor Imaging (DTI) measures and Resting-State functional connectivity (rs-FC).
2 Material and methods
2.1 Participants and data collection
58 COVID-19 survivors were enrolled starting from January 2021 until January 2022 in the context of an ongoing prospective cohort study taking place at San Raffaele Hospital in Milan (De Lorenzo et al., 2020).
To keep a naturalistic study design, we enrolled inpatients aged between 18 and 70, who were hospitalized at San Raffaele Hospital for a SARS-CoV-2 infection.
Clinical and socio-demographic information - including age, sex, date of COVID symptoms’ onset, length of hospitalization, setting of care, presence of depressive symptomatology, presence of intellectual disability or organic illness - were gathered in the context of an unstructured clinical interview conducted by well-trained psychologists. Inclusion criteria were: I) diagnosis of COVID-19 infection as suggested from clinical and radiological findings obtained at the Emergency department and confirmed via reverse transcriptase polymerase chain reaction assays on the nasopharyngeal, throat, or lower respiratory tract swab; II) aged between 18 and 70. Exclusion criteria were: intellectual disabilities, history of drug or alcohol use disorder within the last six months, major neurological disorders, and pregnancy. After a complete description of the study procedure, written informed consent was obtained.
Cognitive complaints were assessed in the context of the clinical interview; they were considered present if participants answered ‘yes’ to at least one of the following clinical questions: ‘Did you develop regular cognitive difficulties in any of the following domains after the illness’ resolution?’ i) forgetfulness in activities of daily living; ii) difficulty with paying attention to external stimuli or maintaining concentration on a task. On the basis of the answer, we divided patients as cognitive complainer and non-complainer. A similar approach was previously validated in general population in a wide-cohort study and proven to be advantageous to investigate subjective cognitive complaints in COVID-19 survivors (Gouraud et al., 2021, Goldberg et al., 2017).
2.2 DTI images preprocessing and statistical analyses
All imaging was performed on a 3.0 T scanner (Ingenia CX, Philips, The Netherlands) with spin-echo echo-planar imaging (EPI) and the following parameters: TR/TE = 5900/78 ms, FoV (mm) 240 (ap), 129 (fh), 232 (rl); acquisition matrix 2.14 × 2.73 × 2.30; 56 contiguous, 2.3 mm thick axial slices reconstructed with in-plane pixel size 1.88 × 1.88 × 2.30 mm; SENSE acceleration factor = 2; 1 b0 and 40 non-collinear directions of the diffusion gradients; b value = 1000 s/mm2. Fat saturation was performed to avoid chemical shift artifacts.
Image analyses and tensor calculations were done using the “Oxford Center for Functional Magnetic Resonance Imaging of the Brain Software Library” (FSL 6.0; www.fmrib.ox.ac.uk/fsl/index.html) (Woolrich et al., 2009). Each DTI volume was affine registered to the T2-weighted b=0 volume using FLIRT (FMRIB's Linear Image Registration Tool) (Jenkinson and Smith, 2001). Correction for susceptibility-induced off-resonance field, eddy current-induced distortions, and subject movements was performed (Andersson et al., 2016) . Least-square fits were performed to estimate the fractional anisotropy (FA), eigenvector, and eigenvalue maps. Mean diffusivity (MD) was defined as the mean of all three eigenvalues (λ 1 + λ 2 + λ 3)/3, axial diffusivity (AD) as the principal diffusion eigenvalue (λ 1), and radial diffusivity (RD) as the mean of the second and third eigenvalues (λ 2 + λ 3)/2. Next, all individuals’ volumes were skeletonized and transformed into a common space as used in Tract-Based Spatial Statistics (Smith et al., 2006). Briefly, all volumes were nonlinearly warped to the FMRIB58_FA template supplied with FSL (http://www.fmrib.ox.ac.uk/fsl/tbss/FMRIB58_FA.html) and normalized to the Montreal Neurological Institute (MNI) space. Next, a mean FA volume of all subjects was generated and thinned to create a mean FA skeleton representing the centres of all common tracts. Individual FA values were warped onto this mean skeleton mask. The resulting tract invariant skeletons for each participant were fed into voxel-wise permutation-based cross-subject statistics. Similar warping and analyses were used on MD, AD, and RD data.
Voxel-wise DTI analyses were performed using nonparametric permutation-based testing (Nichols and Holmes, 2002) as implemented in Randomise in FSL. Within the GLM framework, a Two-Sample unpaired T-Test was performed between cognitive and non-cognitive complainers on FA, MD, AD, and RD across the WM skeleton; we entered age, sex, time from COVID-19 onset to MRI scan, and BMI, thus correcting the results for covariates known to affect the brain integrity. Furthermore, given its close association with subjective cognitive deficits, the presence of depressive symptomatology was also considered as a nuisance covariate in the analyses (Benedetti et al., 2021, Douaud et al., 2022, Stanek et al., 2011). Threshold-free cluster enhancement (TFCE) (Smith and Nichols, 2009) was used to avoid defining arbitrary cluster forming thresholds and smoothing levels. Voxel-wise levels of significance, corrected for multiple comparisons, were then calculated with standard permutation testing by building up the null distribution (across permutation of the input data) of the maximum (across voxels) TFCE scores, and then using the 95th percentile of the null distribution to threshold signals at corrected p < 0.05. The data were tested against an empirical null distribution generated by 5000 permutations for each contrast, thus providing statistical maps fully corrected for multiple comparisons across space. Corrected p < 0.05 in a minimum cluster size of k = 100 was considered significant.
2.3 fMRI data preprocessing and statistical analyses
Scanning sessions for resting-state fMRI images comprised 200 sequential T2*-weighted volumes (interleaved ascending transverse slices covering whole brain, tilted 30° downward with respect to bicommissural line to reduce susceptibility artifacts in orbitofrontal region), acquired using an EPI pulse sequence (TR = 2000 ms; TE = 30 ms; flip angle= 85°; field of view = 192 mm; number of slices = 38; slice thickness = 3.7 mm; matrix size = 64 × 62 reconstructed up to 96 × 96 pixels). Six dummy scans before fMRI acquisition allowed obtaining longitudinal magnetization equilibrium. Total time acquisition was 6 min and 56 s.T2*-weighted images were preprocessed using CONN toolbox (www.nitrc.org/projects/conn), running within Statistical Parametric Mapping (SPM 12). The standard preprocessing pipeline was implemented, which included the following procedures: i) realignment to a reference image in order to minimize variance due to head movements and unwarping; ii) slice timing correction was applied to mitigate temporal misalignment between different slices of the functional data; iii) detection of potential outlier scans (subject motion above 0.9mm and 0.02 rad or spikes in global signal intensity above 5 SD) by means of Artifact Detection Tool (ART, www.nitrc.org/projects/artifact_detect) – a threshold of 20% scans flagged as outliers was set to determine subject exclusion; iv) normalization to a standard MNI space and segmentation of the brain into GM, WM and CSF tissue classes; v) application of spatial smoothing using an 8-mm full-width at half-maximum isotropic Gaussian kernel to enhance the signal-to-noise ratio; vi) removal of confounding effects via an anatomical component-based noise correction procedure (aCompCor), which involves WM, CSF, pshysiological noise source reduction (e.g. six standard motion parameters and ART-based “scrubbed” signal artifacts) with relative derivatives, all taken as covariates in first-level analyses; vii) application of linear detrending to remove linear drift artifacts and high-frequency noise.
Whole-brain functional connectivity (FC) patterns were investigated through multivariate pattern connectivity analyses (MVPA), which allow computing the pairwise connectivity patterns between each voxel and all the other voxels in the brain by reducing the dimensionality of these multi-voxel patterns with principal component analysis (PCA) (Whitfield-Gabrieli et al., 2016). By estimating multivariate connectivity maps for each subject, MVPA provides a more sensitive and analytical approach to the study of the functional organization of the brain than traditional univariate analyses (Haxby, 2012). As a consequence, MVPA allows to decode perceptual stimuli and mental states from brain activation patterns, with evidence of efficacy in the prediction of diagnostic groups (Sundermann et al., 2014) and antidepressant treatment response (Wang et al., 2019). First-level MVPA was performed using 64 dimensions. MVPA-derived maps were then entered in the second-level analyses exploring differences in connectivity patterns between cognitive complainers and non-complainers. Analyses were designed in the context of GLM: the group was considered as a categorical predictor, whereas age, sex, time elapsed from COVID diagnosis to the MRI scan, BMI, and presence of depressive symptoms at the time of recovery as nuisance covariates. The resulting regions of significance indicate clusters of voxels that consistently share similar between-subject variance of their spatial connectivity associated with the presence or otherwise of subjective cognitive complaints. Considering the sample size, 4 components were kept for each voxel. Analyses were thresholded at peak level (p < 0.001, uncorrected; cluster level: p < 0.05 FWE-corrected). To further investigate the direction of the effects, the identified areas were used as seeds for the following seed-to-voxel analyses, aiming at determining whether their functional connectivity with the whole brain raise or decrease according to the group of belonging.
3 Results
Clinical and demographic characteristics of the sample can be seen in Table 1 .Table 1 clinical and demographic characteristics of the sample.
Table 1: Whole Sample (n=58) Non-cognitive complainers (n=29) Cognitive complainers (n=29) t-test or chi-squared significance (p)
Age 52,34 ± 11,73 54,41 ± 9,93 50,27 ± 13,13 0,181
Sex M=41, F=17 M=23, F=6 M=18, F=11 0,149
BMI 26,99 ± 4,86 28,04 ± 5,35 25,93 ± 4,15 0,100
Onset – MRI 173,12 ± 174,03 179,14 ± 173,52 167,10 ± 177,41 0,795
Days of Hospitalization 13,79 ± 13,83 16,20 ± 15,10 11,37 ± 12,23 0,186
ICU admission 9 (15,51%) 5 (17,24%) 4 (13,79%) 0,716
Presence of Depressive Symptomatology 30 (51,72%) 10 (34,48%) 20 (68,96%) 0,009*
⁎ p < 0.05
Twenty-nine participants (50%) presented with cognitive complaints one month after the virus clearance, and in the vast majority of them (twenty-four out of twenty-nine) they were still present at the time of MRI scan. All patients were hospitalized due to COVID-19 infection, and 9 patients required treatment in the ICU: no effect of ICU admission on the development of subjective cognitive impairment was found. No significant differences emerged for age, sex, and other variables, including day of hospitalization. However, 51,72% of the sample reported the presence of depressive symptomatology after discharge, with much higher rates in cognitive complainers (68,96%) than in non-complainers (34,48%).
Cognitive complaints associated with differences in WM microstructure, as revealed by significant differences in DTI measures between cognitive and non-cognitive complainers (Supplementary Table 1). COVID-19 survivors with subjective cognitive impairments showed greater MD in widespread portions of the WM skeleton, affecting bilaterally the inferior fronto-occipital fasciculus, uncinate fasciculus, corona radiata as well as several sections of corpus callosum. The increase in MD associated both, with increased RD in several WM tracts located in the left hemisphere (corona radiata, inferior longitudinal fasciculus, inferior fronto-occipital fasciculus, superior longitudinal fasciculus and uncinate fasciculus), and with higher AD in some inter-hemispheric associative tracts (Figure 1 ). Differences in FA were not significant at pFWE<0.05 but showed a marginal trend of lower values in cognitive complainers (pFWE= 0.09).Figure 1 Differences in DTI indexes between the cognitive-complainer group and the non-complainer counterpart. A) Axial Diffusivity; B) Radial Diffusivity; C) Mean Diffusivity. Only the voxels surviving the statistical threshold of corrected p < 0.05 at TFCE are shown.
Figure 1:
MVPA analysis revealed 2 significant clusters of different rs-FC between the two groups: the first cluster was located in the right frontal pole (peak MNI +20 +38 +32, Cluster Size 120, pFWE=0.0016) and the second in the middle temporal gyrus (peak MNI +68 -54 +06, Cluster Size 92, pFWE=0.0088) (Figure 2 , Supplementary Table 2).Figure 2 Differences in resting-state functional connectivity between the cognitive-complainers and the non-complainer group. First column: seeds of altered connectivity identified in the MVPA analysis. Second and third column: target clusters resulting from post-hoc seed to voxel analysis. Red: areas of increased rs-FC in cognitive complainers; Blue: areas of reduced rs-FC in cognitive complainers.
Figure 2:
Post-hoc seed-to-voxel analysis using the identified clusters as seeds highlighted several brain regions of increased or decreased FC between the two groups; concerning the first seed (right frontal pole), increased FC in subjects complaining deficits was found with eight clusters located in the bilateral Insular Cortex, bilateral Supramarginal and Opercular cortex, Anterior Cingulate gyrus, bilateral Precentral Gyrus and inferior Lateral Occipital cortex; reduced FC was instead found with five clusters located in bilateral superior Occipital cortex, posterior Cingulate gyrus, left middle temporal gyrus (MTG) and right Cerebellum. All the areas with increased FC were part of the Salience, Sensori-Motor or Dorsal Attention networks (except for the small cluster in the inferior LOC belonging to the Visual network); among clusters with decreased FC, the three with the highest statistical significance were found to be part of the Default Mode Network (DMN); no Network was identified for the left MTG cluster, while the last cluster belonged to cerebellar networks.
Concerning the second seed, in subjects reporting cognitive deficits it exhibited lower FC with two clusters located in the right Frontal Pole and in the left posterior MTG. No cluster of increased FC was identified. The Frontal Pole cluster was identified as part of the frontoparietal network, while no cluster was identified for the temporal cluster.
4 Discussion
This is the first study investigating MRI brain correlates of subjective cognitive deficits in COVID-19 survivors. The main finding of the present study is that WM microstructure and resting-state FC differ in patients reporting subjective cognitive impairment. When compared to non-cognitive complainers, patients with newly onset cognitive decline after COVID-19 exhibited far-reaching alterations of WM microstructure and FC. Specifically, patients experiencing cognitive impairment at one month follow-up showed increased widespread MD in several bilateral WM tracts; increased levels of RD and AD were also identified, while FA exhibited a trend towards statistical significance with lower values in subjects reporting cognitive deficits.
Higher rates of depressive symptomatology were reported in the complainer group, thus making it challenging to completely ascribe the connectivity abnormalities we found to the subjective cognitive status; however, the presence of depression was taken into account in all our analyses.
Alterations of DTI indexes have been widely described in patients exhibiting or reporting cognitive impairments; Increased MD and decreased FA are common findings in patients suffering from Alzheimer Disease and Mild Cognitive Impairment (Chandra et al., 2019); higher values of MD and lower FA were also found in preclinical subjects with subjective cognitive decline compared to healthy controls, in anterior and posterior WM tracts also identified in our analysis (Corona Radiata, superior and inferior longitudinal fasciculum, Corpus Callosum) (Brueggen et al., 2019). DTI alterations have been reported also in non-age –related forms of cognitive decline, such as post trauma (Oehr and Anderson, 2017) or iatrogenic post radio or chemotherapy ones (Durán-Gómez et al., 2022, Yahya and Manan, 2021).
As for COVID-19, abnormalities in brain microstructure and functionality linked to severity of neuropsychiatric sequelae are starting to be reported; a previous study by our group showed that indexes of post-COVID depressive and post-traumatic symptomatology significantly associates with disruption in WM integrity and abnormal functional connectivity at rest (Benedetti et al., 2021). Furthermore, alterations of WM microstructure were found to persist up to 1 year after COVID-19 recovery (Huang et al., 2022). In a large longitudinal study performed on 785 subjects from the UK Biobank, greater reductions in Cortical Thickness and increases in DTI Mean Diffusivity in parahippocampal, orbitofrontal anterior cingulate, and insular cortex were observed in subjects that contracted COVID-19 compared to controls. The same study also reported a greater decline of objectively measured cognitive abilities in post-COVID patients, revealing a close relationship between impaired cognitive performance and reduced cerebellar volume (Douaud et al., 2022).
Further supporting the notion of brain involvement in COVID-19 survivors, a higher prevalence and volume of White Matter Hyperintensities (WMH) seem to be a common finding among subjects who recover from the illness with residual cognitive impairments or depressive symptomatology (Andriuta et al., 2022, Cecchetti et al., 2022, Hellgren et al., 2021, Poletti et al., 2022). Finally, a diffusion microstructure imaging study comparing post-COVID subjects exhibiting neurological symptoms or cognitive impairment and healthy controls, found evidence of redistribution of water molecules with widespread decreasing intra-axonal and extra-axonal volume and increasing free water/CSF fractions in COVID-19 survivors; the magnitude of V-CSF-increase was found to correlate with the degree of cognitive impairment, and speculated by the authors to be reflective of vasogenic edema (Rau et al., 2022).
Our study also identified an increase of both AD and RD and a trend towards reduced FA values in subjects reporting deficits: higher values of RD and lower of FA could also be reflective of areas of WM integrity disruption (Winklewski et al., 2018). AD on the other hand reflects the diffusion coefficient along the principal eigenvector (λ 1), and has been proposed as a marker of axonal integrity (Winklewski et al., 2018); however when its increase is not associated with a corresponding increase of FA values, but is associated with higher MD, it is usually thought to be reflective of increased extracellular space and higher water diffusion (Alves et al., 2015), and has been repeatedly reported in patients with cognitive impairment and Alzheimer's disease (Mayo et al., 2019, O'Dwyer et al., 2011).
With regards to resting-state fMRI results, MVPA analyses exposed highly discriminant FC patterns among the two groups, with the complainer cohort showing abnormally high connectivity at rest of the frontal pole with networks critically involved in cognitive-demanding tasks, as well as lower functional pairing with the DMN. Reproducible negative correlations between DMN and the salience and dorsal attentive networks at rest have been widely reported in the literature (Zhou et al., 2018). Specifically, the DMN shows coordinated temporal activity at rest, whereas deactivates during attentional-demanding tasks. Conversely, the dorsal attention network increases its activity during goal-directed tasks, exerting an inhibitory control over the DMN (Di and Biswal, 2014, Smallwood et al., 2021). Notably, the switching processes from the DMN to the attentive networks are promoted by connections between the anterior prefrontal cortex and the insula within the salience network, suggesting that the anterior frontal lobe is fundamental for the transition between resting and cognitively-demanding states (Peng et al., 2018). Increased FC within the dorsal attentive network during resting-state has been previously associated with low cognitive performance in healthy elderly subjects (Charroud et al., 2016, Sala-Llonch et al., 2012). Similar results were found in HIV patients, where over activations in task-related regions are linked with poor cognitive performance at complex behavioural tasks (Hakkers et al., 2017). These results could be explained according to the brain reserve theory, which states that higher activations in task-positive networks are needed to counterweight inefficient cognitive functioning in low-performing subjects (Stern, 2009). Notably, we found that, albeit not involved in performing a cognitive-demanding task, subjective cognitive complainers exhibit increased connectivity of the frontal pole with the salience and dorsal attentive networks, which might indicate an inefficient dynamic modulation of frontal brain activity in these subjects that occurs even during rest.
Together with patterns of increased FC, negative correlations between the right frontal pole and regions encompassing the DMN, such as the superior occipital cortex and posterior cingulate gyrus, were observed in subjective cognitive complainers. Previous studies demonstrated that reduced functional coupling between frontal and posterior cingulate regions is predictive of poor performance during attentional and working memory tasks, both in healthy and pathological aging (Sala-Llonch et al., 2012, Damoiseaux et al., 2008, Greicius et al., 2004, Sambataro et al., 2010). In line with these results, we observed that COVID-19 survivors reporting subjective cognitive impairments exhibit reduced FC between the frontal pole and the DMN during rest, possibly reflecting altered attentional and memory processes which are independent from performance at cognitive tasks.
Finally, we found that subjective cognitive impairments are associated with reduced rs-FC of the right middle temporal gyrus with the right frontal pole and left middle temporal gyrus. Increased activations in the frontal pole, along with medial temporal regions and posterior cingulate cortices, have been associated with metacognitive processes, which represent the ability to monitor and reflect on one's cognition and experience (Chua et al., 2006, Moritz et al., 2006, Yokoyama et al., 2010). Previous evidence suggests that subjective cognitive complainers could be less confident in their cognitive performance compared to non-complainers, reflecting impairments in metacognition and memory self-efficacy (Ponds and Jolles, 1996, Reid and MacLullich, 2006). In our sample, the negative functional coupling between the middle temporal gyrus and frontal pole in subjective cognitive complainers may indicate that an underlying prefrontal hypo-connectivity might promote metacognitive dysfunction, representing a potential neural correlate of long-term cognitive impairments in these subjects.
It should be considered, though, that much higher rates of depressive symptomatology were found in the cognitive complainer group compared to the non-complainer counterpart. Besides being recognized as a marker of cognitive decline (Burmester et al., 2016), subjective cognitive complaints are a common feature of the depressive syndrome (Serra-Blasco et al., 2019, Srisurapanont et al., 2017). Extensive literature supports the notion of a close interplay between depressive symptoms and subjective cognitive failure, further corroborated by the treatment-induced parallel improvement of both affective and cognitive symptomatology in psychiatric populations (Allott et al., 2020). At the same time, since they have been both reported to be highly prevalent in post COVID-19 patients, we could speculate low mood and cognitive impairment to be closely related manifestations of the neuropsychiatric symptomatology of COVID-19 infection (Almeria et al., 2020). In any case, in a bid to investigate specifically the brain correlates of subjective cognitive impairment, we entered the presence of depressive symptomatology as a covariate in all our analyses.
Several pathophysiological mechanisms have been suggested for the development of post-COVID cognitive impairment: a direct neurotropism of the virus is supposed to play a secondary role, while a higher impact is attributed to neuroinflammation and abnormal immune response, endothelial and blood–brain barrier dysfunction and oxidative stress (Stefanou et al., 2022). Therefore, we can hypothesize the alterations of structural and functional connectivity we detected to be reflective of these pathophysiological brain processes, and to translate clinically to the post-COVID cognitive symptomatology. An alternative explanation is to imagine the brain connectome disruption to precede the infection and to confer a vulnerability towards the development of subjective cognitive deficits. Further studies with a longitudinal design will have to address this issue.
Strengths of the present study include a focused research question, state of the art MRI methods, and a real-world experimental setting, but our results must be viewed in light of some limitations. Patients were studied at variable intervals from symptoms’ onset. In any case, time from COVID-19 infection to MRI scan has been considered as nuisance covariate in all the analyses. Patients enrolled in this study were evaluated in a single clinical outpatient service, therefore it cannot be ruled out stratification issues. The drug treatments administered during the course of the illness could have influenced the clinical and biological picture.
The cross-sectional design of the current study limits the possibility to establish a clear causal link between the infection and cognitive deficits. However, subjective cognitive decline, even though difficult to disentangle from depressive symptomatology, may better mirror the dynamic changes of cognitive status before and after the illness than objective cognitive measures. Furthermore, subjective cognitive complaints have been previously found to be reflective of objective cognitive status in COVID-19 patients (Miskowiak et al., 2021). The onset of cognitive complaints could also be influenced by the severity of the infection and particularly by hypoxia; we checked the effect of setting of care (i.e., standard hospitalization or ICU) on the development of cognitive complaints, but this represents only a rough estimation of illness severity. The lack of a healthy control group prevents us from drawing definite conclusions about the pattern of brain alterations underlying subjective cognitive decline. Lastly, as already mentioned in the methods section, the assessment of cognitive status was not performed through a structured questionnaire. Nevertheless, the procedure we employed was adapted from previously validated protocols (Gouraud et al., 2021, Goldberg et al., 2017).
Considering widespread concern about cognitive complaints in the post-COVID stages, the above findings have high clinical relevance in driving early identification of brain patterns strictly associated with the experience of cognitive failure in everyday activities. Although further research is needed to replicate our results, hopefully a solid brain signature may lead to prompt interventions targeting cognitive dysfunction, allowing to early tackle neuropsychiatric sequelae possibly arising afterwards.
5 Contributors
FB, MP (Marco Paolini) and MP (Mariagrazia Palladini) designed the study and wrote the protocol. Authors FB, MP (Marco Paolini), MP (Mariagrazia Palladini), and FC managed the literature searches and analyses. Authors MP (Marco Paolini), MP (Mariagrazia Palladini) and FC undertook the statistical analysis, and authors MP (Marco Paolini), MP (Mariagrazia Palladini), MGM, and FC wrote the first draft of the manuscript. All authors contributed to and have approved the final manuscript.
Declaration of Competing Interest
All the authors declare that they have no conflicts of interest.
Appendix Supplementary materials
Image, application 1
Image, application 2
Funding source
None
Acknowledgements
Funding: MP salary: Italian Ministry of University, XXXVII PhD cycle, FSE REACT-EU 2021 PON projects, Action IV.5.
We also thank Miss Valentina Bettonagli, who kindly provided the data necessary for our analysis.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.euroneuro.2022.12.002.
==== Refs
References
Jiang L Tang K Levin M Irfan O Morris SK Wilson K COVID-19 and multisystem inflammatory syndrome in children and adolescents The Lancet Infectious Diseases 20 11 2020 e276 ee88 32818434
Nalbandian A Sehgal K Gupta A Madhavan MV McGroder C Stevens JS Post-acute COVID-19 syndrome Nature medicine 27 4 2021 601 615
Spudich S Nath A. Nervous system consequences of COVID-19 Science 375 6578 2022 267 269 35050660
Zanin L Saraceno G Panciani PP Renisi G Signorini L Migliorati K SARS-CoV-2 can induce brain and spine demyelinating lesions Acta Neurochirurgica 162 7 2020 1491 1494 32367205
Voruz P Cionca A Jacot de Alcântara I Nuber-Champier A Allali G Benzakour L Functional connectivity underlying cognitive and psychiatric symptoms in post-COVID-19 syndrome: is anosognosia a key determinant? Brain Communications 4 2 2022 fcac057 35350554
Zhang H Chung TW-H Wong FK-C Hung IF-N Mak HK-F. Changes in the Intranetwork and Internetwork Connectivity of the Default Mode Network and Olfactory Network in Patients with COVID-19 and Olfactory Dysfunction Brain sciences 12 4 2022 511 35448042
O'Shea A Parakh A Hedgire S Lee SI. Multisystem assessment of the imaging manifestations of coagulopathy in hospitalized patients with coronavirus disease (COVID-19) American Journal of Roentgenology 216 4 2021 1088 1098 32755217
Tang SW Helmeste D Leonard B. Inflammatory neuropsychiatric disorders and COVID-19 neuroinflammation Acta Neuropsychiatrica 33 4 2021 165 177 33926589
Tang SW Leonard BE Helmeste DM. Long COVID, neuropsychiatric disorders, psychotropics, present and future Acta Neuropsychiatrica 2022 1 18
Mazza MG De Lorenzo R Conte C Poletti S Vai B Bollettini I Anxiety and depression in COVID-19 survivors: Role of inflammatory and clinical predictors Brain, behavior, and immunity 89 2020 594 600 32738287
Mazza MG Palladini M De Lorenzo R Magnaghi C Poletti S Furlan R Persistent psychopathology and neurocognitive impairment in COVID-19 survivors: effect of inflammatory biomarkers at three-month follow-up Brain, behavior, and immunity 94 2021 138 147 33639239
Poletti S Palladini M Mazza MG De Lorenzo R Furlan R Ciceri F Long-term consequences of COVID-19 on cognitive functioning up to 6 months after discharge: role of depression and impact on quality of life European archives of psychiatry and clinical neuroscience 2021 1 10 33433677
Hampshire A Trender W Chamberlain SR Jolly AE Grant JE Patrick F Cognitive deficits in people who have recovered from COVID-19 EClinicalMedicine 39 2021 101044
Miskowiak K Johnsen S Sattler S Nielsen S Kunalan K Rungby J Cognitive impairments four months after COVID-19 hospital discharge: Pattern, severity and association with illness variables European Neuropsychopharmacology 46 2021 39 48 33823427
Andersson S Hestad K Johannessen D Okkenhaug I Gramstad A Andreassen O Subjective cognitive complaints and neuropsychological performance at six months post COVID-19 Journal of Psychosomatic Research 157 2022 110814
Ferrando SJ Dornbush R Lynch S Shahar S Klepacz L Karmen CL Neuropsychological, Medical, and Psychiatric Findings After Recovery from Acute COVID-19: A Cross-sectional Study Journal of the Academy of Consultation-liaison Psychiatry 2022
Gouraud C Bottemanne H Lahlou-Laforêt K Blanchard A Günther S Batti SE Association between psychological distress, cognitive complaints, and neuropsychological status after a severe Covid-19 infection: a cross-sectional study Frontiers in Psychiatry 2021 1530
Ceban F Ling S Lui LM Lee Y Gill H Teopiz KM Fatigue and cognitive impairment in Post-COVID-19 Syndrome: A systematic review and meta-analysis Brain, behavior, and immunity 101 2022 93 135 34973396
Schou TM Joca S Wegener G Bay-Richter C. Psychiatric and neuropsychiatric sequelae of COVID-19–A systematic review Brain, Behavior, and Immunity 97 2021 328 348 34339806
Silva LS Joao RB Nogueira MH Aventurato IK de Campos BM de Brito MR Functional and microstructural brain abnormalities, fatigue, and cognitive dysfunction after mild COVID-19 Medrxiv 2021
Voruz P, Cionca A, Jacot I, Nuber-Champier A, Allali G, Benzakour L, et al. Brain functional connectivity alterations associated with neuropsychological post-COVID syndrome. 2022.
Yesilkaya UH Sen M Balcioglu YH. COVID-19-related cognitive dysfunction may be associated with transient disruption in the DLPFC glutamatergic pathway Journal of Clinical Neuroscience 87 2021 153 155 33863524
Ott CV Bjertrup AJ Jensen JH Ullum H Sjælland R Purdon SE Screening for cognitive dysfunction in unipolar depression: validation and evaluation of objective and subjective tools Journal of Affective Disorders 190 2016 607 615 26583350
Petersen JZ Porter RJ Miskowiak KW. Clinical characteristics associated with the discrepancy between subjective and objective cognitive impairment in depression Journal of Affective Disorders 246 2019 763 774 30623822
De Lorenzo R Conte C Lanzani C Benedetti F Roveri L Mazza MG Residual clinical damage after COVID-19: A retrospective and prospective observational cohort study PLoS One 15 10 2020 e0239570
Goldberg M Carton M Descatha A Leclerc A Roquelaure Y Santin G CONSTANCES: a general prospective population-based cohort for occupational and environmental epidemiology: cohort profile Occupational and Environmental Medicine 74 1 2017 66 71 27884936
Woolrich MW Jbabdi S Patenaude B Chappell M Makni S Behrens T Bayesian analysis of neuroimaging data in FSL Neuroimage 45 1 2009 S173 SS86 19059349
Jenkinson M Smith S. A global optimisation method for robust affine registration of brain images Medical image analysis 5 2 2001 143 156 11516708
Andersson JL Graham MS Zsoldos E Sotiropoulos SN. Incorporating outlier detection and replacement into a non-parametric framework for movement and distortion correction of diffusion MR images Neuroimage 141 2016 556 572 27393418
Smith SM Jenkinson M Johansen-Berg H Rueckert D Nichols TE Mackay CE Tract-based spatial statistics: voxelwise analysis of multi-subject diffusion data Neuroimage 31 4 2006 1487 1505 16624579
Nichols TE Holmes AP. Nonparametric permutation tests for functional neuroimaging: a primer with examples Human brain mapping 15 1 2002 1 25 11747097
Benedetti F Palladini M Paolini M Melloni E Vai B De Lorenzo R Brain correlates of depression, post-traumatic distress, and inflammatory biomarkers in COVID-19 survivors: A multimodal magnetic resonance imaging study Brain, behavior, & immunity-health 18 2021 100387
Douaud G Lee S Alfaro-Almagro F Arthofer C Wang C McCarthy P SARS-CoV-2 is associated with changes in brain structure in UK Biobank Nature 604 7907 2022 697 707 35255491
Stanek KM Grieve SM Brickman AM Korgaonkar MS Paul RH Cohen RA Obesity is associated with reduced white matter integrity in otherwise healthy adults Obesity 19 3 2011 500 504 21183934
Smith SM Nichols TE. Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference Neuroimage 44 1 2009 83 98 18501637
Whitfield-Gabrieli S Ghosh S Nieto-Castanon A Saygin Z Doehrmann O Chai X Brain connectomics predict response to treatment in social anxiety disorder Molecular psychiatry 21 5 2016 680 685 26260493
Haxby JV. Multivariate pattern analysis of fMRI: the early beginnings Neuroimage 62 2 2012 852 855 22425670
Sundermann B Herr D Schwindt W Pfleiderer B. Multivariate classification of blood oxygen level–dependent fMRI data with diagnostic intention: a clinical perspective American Journal of neuroradiology 35 5 2014 848 855 24029388
Wang Y Bernanke J Peterson BS McGrath P Stewart J Chen Y The association between antidepressant treatment and brain connectivity in two double-blind, placebo-controlled clinical trials: a treatment mechanism study The Lancet Psychiatry 6 8 2019 667 674 31248841
Chandra A Dervenoulas G Politis M. Magnetic resonance imaging in Alzheimer's disease and mild cognitive impairment Journal of neurology 266 6 2019 1293 1302 30120563
Brueggen K Dyrba M Cardenas-Blanco A Schneider A Fliessbach K Buerger K Structural integrity in subjective cognitive decline, mild cognitive impairment and Alzheimer's disease based on multicenter diffusion tensor imaging Journal of neurology 266 10 2019 2465 2474 31227891
Oehr L Anderson J. Diffusion-tensor imaging findings and cognitive function following hospitalized mixed-mechanism mild traumatic brain injury: A systematic review and meta-analysis Archives of physical medicine and rehabilitation 98 11 2017 2308 2319 28433414
Durán-Gómez N López-Jurado CF Nadal-Delgado M Pérez-Civantos D Guerrero-Martín J Cáceres MC. Chemotherapy-Related Cognitive Impairment in Patients with Breast Cancer Based on Functional Assessment and NIRS Analysis Journal of Clinical Medicine 11 9 2022 2363 35566489
Yahya N Manan HA. Diffusion tensor imaging indices to predict cognitive changes following adult radiotherapy European journal of cancer care 30 1 2021 e13329 32909654
Huang S Zhou Z Yang D Zhao W Zeng M Xie X Persistent white matter changes in recovered COVID-19 patients at the 1-year follow-up Brain 145 5 2022 1830 1838 34918020
Andriuta D Si-Ahmed C Roussel M Constans J-M Makki M Aarabi A Clinical and Imaging Determinants of Neurocognitive Disorders in Post-Acute COVID-19 Patients with Cognitive Complaints Journal of Alzheimer's Disease Preprint 2022 1 12
Cecchetti G Agosta F Canu E Basaia S Barbieri A Cardamone R Cognitive, EEG, and MRI features of COVID-19 survivors: a 10-month study Journal of neurology 2022 1 13
Hellgren L Thornberg UB Samuelsson K Levi R Divanoglou A Blystad I. Brain MRI and neuropsychological findings at long-term follow-up after COVID-19 hospitalisation: an observational cohort study BMJ open 11 10 2021 e055164
Poletti S Paolini M Mazza MG Palladini M Furlan R Querini PR Lower levels of glutathione in the anterior cingulate cortex associate with depressive symptoms and white matter hyperintensities in COVID-19 survivors: Glutathione in COVID-19 European Neuropsychopharmacology 2022
Rau A Schroeter N Blazhenets G Dressing A Walter LI Kellner E Widespread white matter oedema in subacute COVID-19 patients with neurological symptoms Brain 2022
Winklewski PJ Sabisz A Naumczyk P Jodzio K Szurowska E Szarmach A. Understanding the physiopathology behind axial and radial diffusivity changes—what do we know? Frontiers in neurology 9 2018 92 29535676
Alves GS Oertel Knöchel V Knöchel C Carvalho AF Pantel J Engelhardt E Integrating retrogenesis theory to Alzheimer's disease pathology: insight from DTI-TBSS investigation of the white matter microstructural integrity BioMed research international 2015 2015
Mayo CD Garcia-Barrera MA Mazerolle EL Ritchie LJ Fisk JD Gawryluk JR Relationship between DTI metrics and cognitive function in Alzheimer's disease Frontiers in aging neuroscience 10 2019 436 30687081
O'Dwyer L Lamberton F Bokde AL Ewers M Faluyi YO Tanner C Multiple indices of diffusion identifies white matter damage in mild cognitive impairment and Alzheimer's disease PloS one 6 6 2011 e21745 21738785
Zhou Y Friston KJ Zeidman P Chen J Li S Razi A. The hierarchical organization of the default, dorsal attention and salience networks in adolescents and young adults Cerebral cortex 28 2 2018 726 737 29161362
Di X Biswal BB. Modulatory interactions between the default mode network and task positive networks in resting-state PeerJ 2 2014 e367 24860698
Smallwood J Bernhardt BC Leech R Bzdok D Jefferies E Margulies DS. The default mode network in cognition: a topographical perspective Nature reviews neuroscience 22 8 2021 503 513 34226715
Peng K Steele SC Becerra L Borsook D. Brodmann area 10: collating, integrating and high level processing of nociception and pain Progress in neurobiology 161 2018 1 22 29199137
Charroud C Le Bars E Deverdun J Steffener J Molino F Abdennour M Working memory performance is related to intrinsic resting state functional connectivity changes in community-dwelling elderly cohort Neurobiology of Learning and Memory 132 2016 57 66 27234057
Sala-Llonch R Arenaza-Urquijo EM Valls-Pedret C Vidal-Piñeiro D Bargalló N Junqué C Dynamic functional reorganizations and relationship with working memory performance in healthy aging Frontiers in Human Neuroscience 6 2012 152 22701409
Hakkers C Arends J Barth R Du Plessis S Hoepelman A Vink M Review of functional MRI in HIV: effects of aging and medication Journal of neurovirology 23 1 2017 20 32 27718211
Stern Y. Cognitive reserve Neuropsychologia 47 10 2009 2015 2028 19467352
Damoiseaux JS Beckmann C Arigita ES Barkhof F Scheltens P Stam C Reduced resting-state brain activity in the “default network” in normal aging Cerebral cortex 18 8 2008 1856 1864 18063564
Greicius MD Srivastava G Reiss AL Menon V. Default-mode network activity distinguishes Alzheimer's disease from healthy aging: evidence from functional MRI Proceedings of the National Academy of Sciences 101 13 2004 4637 4642
Sambataro F Murty VP Callicott JH Tan H-Y Das S Weinberger DR Age-related alterations in default mode network: impact on working memory performance Neurobiology of aging 31 5 2010 839 852 18674847
Chua EF Schacter DL Rand-Giovannetti E Sperling RA. Understanding metamemory: neural correlates of the cognitive process and subjective level of confidence in recognition memory Neuroimage 29 4 2006 1150 1160 16303318
Moritz S Gläscher J Sommer T Büchel C Braus DF. Neural correlates of memory confidence Neuroimage 33 4 2006 1188 1193 17029986
Yokoyama O Miura N Watanabe J Takemoto A Uchida S Sugiura M Right frontopolar cortex activity correlates with reliability of retrospective rating of confidence in short-term recognition memory performance Neuroscience research 68 3 2010 199 206 20688112
Ponds RW Jolles J. Memory complaints in elderly people: The role of memory abilities, metamemory, depression, and personality Educational Gerontology: An International Quarterly 22 4 1996 341 357
Reid LM MacLullich AM. Subjective memory complaints and cognitive impairment in older people Dementia and geriatric cognitive disorders 22 5-6 2006 471 485 17047326
Burmester B Leathem J Merrick P. Subjective cognitive complaints and objective cognitive function in aging: a systematic review and meta-analysis of recent cross-sectional findings Neuropsychology Review 26 4 2016 376 393 27714573
Serra-Blasco M Torres IJ Vicent-Gil M Goldberg X Navarra-Ventura G Aguilar E Discrepancy between objective and subjective cognition in major depressive disorder European Neuropsychopharmacology 29 1 2019 46 56 30503099
Srisurapanont M Suttajit S Eurviriyanukul K Varnado P. Discrepancy between objective and subjective cognition in adults with major depressive disorder Scientific Reports 7 1 2017 1 7 28127051
Allott K Gao C Hetrick SE Filia KM Menssink JM Fisher C Subjective cognitive functioning in relation to changes in levels of depression and anxiety in youth over 3 months of treatment BJPsych Open 6 5 2020
Almeria M Cejudo JC Sotoca J Deus J Krupinski J. Cognitive profile following COVID-19 infection: Clinical predictors leading to neuropsychological impairment Brain, behavior, & immunity-health 9 2020 100163
Stefanou M-I Palaiodimou L Bakola E Smyrnis N Papadopoulou M Paraskevas GP Neurological manifestations of long-COVID syndrome: A narrative review Therapeutic advances in chronic disease 13 2022 20406223221076890
| 0 | PMC9742225 | NO-CC CODE | 2022-12-14 23:30:06 | no | Eur Neuropsychopharmacol. 2022 Dec 12; doi: 10.1016/j.euroneuro.2022.12.002 | utf-8 | Eur Neuropsychopharmacol | 2,022 | 10.1016/j.euroneuro.2022.12.002 | oa_other |
==== Front
Lancet Reg Health Southeast Asia
Lancet Reg Health Southeast Asia
The Lancet Regional Health. Southeast Asia
2772-3682
2772-3682
The Author(s). Published by Elsevier Ltd.
S2772-3682(22)00147-0
10.1016/j.lansea.2022.100130
100130
Articles
Incidence of SARS-CoV-2 infection in hospital workers before and after vaccination programme in East Java, Indonesia - A retrospective cohort study
Soegiarto Gatot a∗∗
Purnomosari Dewajani b
Wulandari Laksmi c
Mahdi Bagus Aulia d
Fahmita Karin Dhia d
Hadmoko Satrio Tri d
Gautama Hendra Ikhwan d
Prasetyo Muhammad Edwin d
Prasetyaningtyas Dewi d
Negoro Pujo Prawiro d
Arafah Nur d
Prakoeswa Cita Rosita Sigit e
Endaryanto Anang f
Suprabawati Desak Gede Agung g
Tinduh Damayanti h
Rachmad Eka Basuki i
Triyono Erwin Astha j
Wahyuhadi Joni k
Keswardiono Catur Budi l
Wardani Feby Elyana l
Mayorita Fitriyah l
Kristiani Nunuk l
Baskoro Ari a
Fetarayani Deasy a
Nurani Wita Kartika a
Oceandy Delvac mn∗
a Division of Allergy and Clinical Immunology, Department of Internal Medicine, Faculty of Medicine, Universitas Airlangga - Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
b Department of Histology and Cell Biology, Faculty of Medicine Public Health and Nursing, Universitas Gadjah Mada, Yogyakarta, Indonesia
c Department of Pulmonology and Respiratory Medicine, Faculty of Medicine, Universitas Airlangga - Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
d Department of Internal Medicine, Faculty of Medicine, Universitas Airlangga - Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
e Department of Dermatology and Venereology, Faculty of Medicine, Universitas Airlangga - Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
f Department of Child Health, Faculty of Medicine, Universitas Airlangga - Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
g Division of Oncology, Department of Surgery, Faculty of Medicine, Universitas Airlangga - Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
h Department of Physical Medicine and Rehabilitation, Faculty of Medicine, Universitas Airlangga - Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
i Medical Service Bureau, Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
j Division of Tropical Disease and Infection, Department of Internal Medicine, Faculty of Medicine, Universitas Airlangga - Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
k Department of Neurosurgery, Faculty of Medicine, Universitas Airlangga – Dr. Soetomo General Academic Hospital, Surabaya, Indonesia
l Syarifah Ambami Rato Ebu Hospital, Bangkalan, Madura, East Java, Indonesia
m Division of Cardiovascular Sciences Faculty of Biology Medicine and Health, University of Manchester, Manchester Academic Health Science Centre, Manchester, United Kingdom
n Department of Biomedical Science, Faculty of Medicine, Universitas Airlangga, Surabaya, Indonesia
∗ Corresponding authors at: Division of Cardiovascular Sciences Faculty of Biology Medicine and Health, The University of Manchester, Oxford Rd, Manchester M13 9PL, UK (D. Oceandy),
∗∗ Corresponding author: Division of Allergy and Clinical Immunology, Department of Internal Medicine, Faculty of Medicine, Universitas Airlangga and Dr. Soetomo General Academic Hospital, Surabaya, Indonesia, Jl. Mayjen. Prof. Dr. Moestopo no. 6-8, Surabaya, 60286, East Java, Indonesia (G. Soegiarto). (G. Soegiarto)
12 12 2022
12 12 2022
10013017 8 2022
11 10 2022
1 12 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.
Background
The incidence of the Coronavirus Disease 2019 (COVID-19) among healthcare workers (HCWs) is widespread. It is important to understand COVID-19 characteristics among HCWs before and after vaccination. We evaluated the incidence of COVID-19 among HCWs in East Java, Indonesia comparing the characteristics of the disease between the pre- vs post-vaccination periods.
Methods
A retrospective observational study was conducted among HCWs in two major hospitals in East Java, Indonesia, between April 01, 2020, and Oct 31, 2021. All HCWs were offered vaccination with inactivated viral vaccine (CoronaVac) from Jan 15, 2021. Therefore, we divided the time of the study into the pre-vaccination period (between April 01, 2020, and Jan 14, 2021) and post-vaccination period (between Jan 15 and Oct 31, 2021). We then compared the pattern of COVID-19 infections, and hospitalisations between these periods.
Findings
A total of 434 (15.1%) and 649 (22.6%) SARS-CoV-2 infections were reported among study participants (n=2,878) during the pre-vaccination and post-vaccination periods, respectively. The vaccine effectiveness was 73.3% during the first 3-4 months after vaccination but this decreased to 17.6% at 6-7 months after vaccination, which coincided with the emergence of the delta variant. The overall hospitalisation rate was reduced from 23.5% in the pre-vaccination period to 14.3% in the post-vaccination period. Hypertension appeared to be the strongest risk factor affecting hospitalisation in the pre-vaccination period. However, the risk due to hypertension was reduced in the post-vaccination period.
Interpretation
The risk to contract COVID-19 remains high among HCWs in East Java, Indonesia. Vaccination is important to reduce infection and hospitalisation. It is essentially important to evaluate the characteristics of COVID-19 infection, hospitalisation, the impact of co-morbidities and vaccine effectiveness in order to improve the measures applied in protecting HCWs during the pandemic.
Funding
Mandate Research Grant No:1043/UN3.15/PT/2021, Universitas Airlangga, Indonesia
Keywords
COVID-19
healthcare workers
vaccine
hospitalisation
co-morbidity
==== Body
pmcIntroduction
In addition to therapeutic medications, vaccines and rapid diagnostics have been regarded as key tools to overcome the Coronavirus Disease 2019 (COVID-19) pandemic1. Several types of COVID-19 vaccines have successfully been developed and some of them have shown strong effectiveness in preventing SARS-CoV-2 infection or reducing the severity of the disease. One of the most widely used COVID-19 vaccines is the inactivated viral vaccine. The clinical trials and real-life data shows that inactivated viral vaccines provides significant level of protection against severe COVID-192 , 3. Although there were fewer reports regarding the effectiveness compared to the mRNA or adenoviral vaccine, the inactivated viral vaccine has been approved and used in as many as 91 countries4.
Indonesia is among the first countries in the world which started a mass vaccination programme at the beginning of 20215. Following a phase 3 clinical trial in the country6, the Indonesian authority for drug and food approval (BPOM) authorised the emergency use of inactivated SARS-CoV-2 vaccine (CoronaVac) in the national vaccination programme, which was started on Jan 13, 2021. Healthcare workers (HCWs) were the first group of people who received the COVID-19 vaccine during that period.
Despite the high protection of COVID-19 vaccines as reported in the phase 3 trials and in the initial real world data, there were concerns about reduction in vaccine effectiveness over time; possibly due to the waning of immunity7 , 8 and the emergence of new virus variants9 , 10. Therefore, it is important to understand whether COVID-19 vaccination programme affected the pattern and characteristics of COVID-19 incidence and hospitalisation in the population. Since HCWs have a greater risk for SARS-CoV-2 infection, the surveillance data to record COVID-19 incidence and hospitalisation are important to provide stakeholders and policy makers with information to consider the next strategies to protect HCWs from COVID-19. In this study a retrospective analysis was conducted on HCWs in East Java, Indonesia, who were offered inactivated SARS-CoV-2 vaccines at the early stage of the national COVID-19 vaccination programme. The main objective was to evaluate the incidence of COVID-19 in a cohort of HCWs during the pre-vaccination and post-vaccination periods. We also compared the rate of hospitalisation between these two periods and analysed whether demographic characteristics and the presence of co-morbidity affected the pattern of infections and hospitalisations in these two periods of the study.
Methods
Study participants
A cohort of 2,878 HCWs agreed to participate in this retrospective observational study. The participants were HCWs from two major hospitals in East Java, Indonesia: Dr. Soetomo General Academic Hospital in Surabaya (1,805 participants) and Syarifah Ambami Rato Ebu Hospital in Bangkalan (1,073 participants). The total number HCWs in both hospitals at the time of the study was 7,311. Thus, the participants were 39.4% of the total HCWs in both hospitals. All the HCWs in these hospitals were offered two doses of inactivated SARS-CoV-2 vaccine (CoronaVac). Based on the power calculation method described by Lwanga and Lemeshow11 a minimum of 1,537 samples was required to estimate incidence rate with relative precision of 5% and 95% CI. Our sample size exceeded this requirement.
Data collection
Data on demographic characteristics and the presence of comorbidities were collected using a computer-based questionnaire. The participants were recruited, and the questionnaires were completed between Sept 2 and Oct 31, 2021. We then conducted retrospective observations using hospital medical records to evaluate the incidence of SARS-CoV-2 infections in the study participants which occurred between April 01, 2020, and Oct 31, 2021. We also used the medical records to assess hospitalisation due to COVID-19 during the period above. We collected vaccination records of the participants from the P-CARE application, which was used by the health authority to record and monitor the implementation of the national vaccination programme in Indonesian. We combined data from the medical and vaccination records with the data from the questionnaire and tabulated them in SPSS software version 25 (IBM Corp., Armonk, NY). The workflow of this study is presented in Figure 1 .Figure 1 Study workflow. The study cohort involving health care workers (HCWs) in Dr Soetomo General Hospital, Surabaya and Syarifah Ambami Rato Ebu, Bangkalan. Both hospitals are located in East Java province, Indonesia. Participants were recruited and baseline demographic and comorbidities data were collected between Sept 02 and Oct 31, 2021. Medical records and vaccination data were collected retrospectively spanning the period between April 01, 2020 and Oct 31, 2021.
COVID-19 infection that occurred at more than 14 days after the second dose of vaccination was categorised as breakthrough infection (infection in a fully vaccinated participant). Otherwise, it was categorised as infection in incomplete or unvaccinated individuals.
Analysis of SARS-CoV-2 variants in East Java, Indonesia was performed using data from GISAID12. The sequence metadata of viruses isolated from East Java, Indonesia during the study period (between April 01, 2020, and Oct 31, 2021) were retrieved from the GISAID website and were analysed to evaluate the prevalence of SARS-CoV-2 variants in this region. The list of virus sequence entries which were used in this study can be accessed in the supplementary material.
Analysis of vaccine effectiveness
Analysis of vaccine effectiveness was conducted using the following equation as described in previous publication13:Vaccineeffectiveness=Incidenceinunvaccinatedparticipants−incidenceinvaccinatedparticipantsIncidenceinunvaccinatedparticipantsx100%
Ethics statement
We received ethical approval from the Local Health Research Ethics Committee of Dr Soetomo General Academic Hospital, Surabaya, Indonesia (No. 0145/KEPK/II/2021) to conduct this study. All study participants signed the informed consent form to confirm their agreement to participate in this study.
Data analysis
We used IBM SPSS Statistics software version 25 (IBM Corp., Armonk, NY) to analyse the data. To analyse categorical data, we used chi-squared or Fisher’s exact test if it is a 2 x 2 crosstab. The incidence of SARS-CoV-2 infection was analysed according to the period of study, i.e. pre-vaccination period (between April 01, 2020, and Jan 14, 2021) and post-vaccination period (between Jan 15 and Oct 31, 2021). We used chi-squared or Fisher’s exact tests to determine if the incidence of infection was correlated with demographic characteristics (i.e. sex, age, and work unit) and to determine if hospitalisation was associated with the presence of comorbidities. Multivariate logistic regression analysis was conducted to determine the odds ratio (OR) of being hospitalised by adjusting for the confounding effects of each variable. Age, sex, body mass index (BMI), history of hypertension, diabetes mellitus, cardiovascular diseases and lung diseases/asthma were included in the multivariate analysis. This was based on previous findings showing that these variables were significantly associated with the severity of COVID-1914, 15, 16. A P value < 0.05 was considered statistically significant.
Role of the funding source
This study was funded by the Mandate Research Grant No: 1043/UN3.15/PT/2021 from Universitas Airlangga, Surabaya. The funder did not have any role in the study design, data acquisition and analysis, data interpretation, or manuscript writing.
Results
Study participants
We conducted this retrospective study to assess the features of COVID-19 infection amongst HCWs in two major hospitals in East Java Indonesia between April 01, 2020, and Oct 31, 2021. The national vaccination programme in Indonesia was initiated on Jan 13, 2021 and the first HCWs in East Java received the COVID-19 vaccine on Jan 15, 2021, so we divided the study period into two categories: pre-vaccination period (between April 01, 2020, and Jan 14, 2021) and post-vaccination period (between Jan 15 and Oct 31, 2021).
Data were collected from a total of 2,878 participants. Most of the study participants were medical staffs (including physicians, residents, and medical students; 25.7%) and nurses (27.7%), whereas 19.5% of participants were administrative staff members. Technical staff (including radiographers, hospital information technology personnel, electromedical engineers or technician, hospital infrastructure and facilities maintenance personnel) accounted for 11.4% of participants and 2.5% of participants are pharmacists (Table 1 ).Table 1 Demographic data of study participants
Study participants (n = 2,878)
Gender
Male 1515 (52.6 %)
Female 1363 (47.4 %)
Age (mean + SD) (years) 36.15 + 9.88
Age group (years)
< 30 787 (27.3 %)
30 – 39 1202 (41.8 %)
40 – 49 504 (17.5%)
≥ 50 385 (13.4 %)
Work unit
Medical staff (physician, resident, medical student) 739 (25.7 %)
Nurse 796 (27.7 %)
Pharmacist 71 (2.5 %)
Technical staff 329 (11.4 %)
Administrative staff 560 (19.5 %)
Others (security, cleaning service, drivers, social care workers) 383 (13.3 %)
Demographic characteristics of the study participants are described in Table 1. The average age was 36.15 + 9.88 years with most of the participants (41.8%) were in the 30 – 39 years age group. The distribution of sex was slightly higher in males (52.6%).
The uptake of the COVID-19 vaccination amongst study participants is shown in Figure 2 . The first dose of vaccine was administered to HCWs starting on Jan 15, 2021, whereas the second dose of vaccine was given two weeks after the first dose. All participants received CoronaVac (inactivated SARS-CoV-2 vaccine) since this was the only type of vaccine used by the Indonesian Government at that time. As shown in Figure 2 by the start of March, 2021, about 76% of the participants had received the first dose of vaccine and around 66% of them had received the second dose. By the end of April, 2021, as many as 88% participants had received the first dose and 82% had received the second dose. At the end of the observation (Oct 31,2021), 93% of participants had received two doses of the vaccine. This indicates a very high uptake of COVID-19 vaccination amongst HCWs in major hospitals in East Java.Figure 2 COVID-19 weekly cases and vaccination coverage among the study participants. The weekly incidence of COVID-19 cases among healthcare workers in two major hospitals in East Java between April 01, 2020 and Oct 31, 2021. There were two surges of COVID-19 cases, first in December 2020 – January 2021 and second in June – July 2021. The incidence reflected the national and regional epidemic curve as described elsewhere23,24. The vaccination program started in mid-January 2021 and by April 2021, 82% of participants had received full doses of vaccination.
Incidence of SARS-CoV-2 infection
The epidemic curve of COVID-19 incidence amongst study participants during the period of this study is described in Figure 2, showing two waves of COVID-19 surges amongst HCWs with the weekly cases peaking for the first time at the end of the period between December, 2020 and January, 2021; and peaking for the second time during the period between June and July 2021.
It was likely that the dominant virus strain, which circulated among HCWs in these hospitals, was similar to the virus strain which circulated in the wider population in East Java at the respective time. Therefore, we examined virus sequence data from GISAID database to understand the dominant SARS-CoV-2 strains that circulated in East Java during the period of this study. As shown in Supplementary Figure 1 there were two different types of virus strains that were dominant in East Java. The first one was the lineage B.1.470. This variant was dominant during the first surge of COVID-19 infection between December 2020 and March 2021. The second dominant strain was the Delta variant (Pango lineage B.1.617.2). This variant was dominant during the second surge between June 2021 and August 2021. Please see Supplementary Table 1 for detailed information on virus data included in this analysis.
Characteristics of infected participants
The characteristics of infected participants are described in Table 2 . During the pre-vaccination period (289 days), the cumulative incidence of COVID-19 was 434 (15.1%) cases. This corresponded to a 30-day incidence of 1.57%. On the other hand, during the post vaccination period (290 days), the cumulative incidence was higher (649 cases, 22.6%) which was equal to a 30-day incidence of 2.34%. In the pre-vaccination period, the incidence of infection seemed to be higher among medical staffs and nurses compared to staff members of other work units, however, we observed an increased incidence of infection among staff of all work units during the post-vaccination period. There were also changes in terms of the age group of infected participants, in which the 30-39 years age group was the highest during the pre-vaccination period, whereas in the post-vaccination period it was the 40 – 49 years age group that had the highest incidence.Table 2 Incidence of SARS-Cov-2 infection
Pre-vaccination period (April 01, 2020 – Jan 14, 2021)
(n = 2,878) Post-vaccination period (Jan 15, 2021 – Oct 31, 2021)
(n = 2,878)
Cumulative incidence of SARS-CoV-2 infection 434/2878 (15.1 %) 649/2878 (22.6 %)
Gender
Male 253/1515 (16.7 %) 319/1515 (21.1 %)
Female 181/1363 (13.3 %) 330/1363 (24.2 %)
P value (Fisher’s exact) 0.011 0.044
Age group (years)
< 30 100/787 (12.7 %) 157/787 (19.9 %)
30 – 39 200/1202 (16.6 %) 283/1202 (23.5 %)
40 – 49 71/504 (14.1 %) 122/504 (24.2 %)
≥ 50 63/385 (16.4 %) 87/385 (22.6 %)
P value (Chi-square) 0.085 0.210
Work unit
Medical staff 122/739 (16.5 %) 145/739 (19.6 %)
Nurse 126/796 (15.8 %) 203/796 (25.5 %)
Pharmacist 6/71 (8.5 %) 24/71 (33.8 %)
Technician 39/329 (11.9 %) 74/329 (22.5 %)
Administrative staff 72/560 (12.9 %) 117/560 (20.9 %)
Others 69/383 (18.0 %) 86/383 (22.5 %)
P value (Chi-square) 0.044 0.018
Vaccine effectiveness in preventing infection
We then analysed the status of vaccination of participants at the time of infection. During the period of this study, 2,676 participants (93% of total subjects) had received two doses of inactivated viral vaccine. Of these, 572 participants (21.4%) had been infected with SARS-CoV-2 virus at >14 days after receiving their second dose of vaccination. In contrast, we observed 77 (38.1%) COVID-19 infections among participants with either incomplete (receiving only one dose) or no vaccination at all, out of a total of 202 subjects in this category. The difference was statistically significant (P<0.0001, Fisher’s exact test, Table 3 ). This finding indicated approximately 44% protection against infection in participants with two doses of inactivated viral vaccine between Jan 15 and Oct 31, 2021.Table 3 Incidence of SARS-CoV-2 infection according to vaccination status at post-vaccination period
Status of vaccination at the time of infection Fisher’s Exact test
Fully vaccinated Incomplete or unvaccinated
SARS-CoV-2 infection – Jan 15, 2021 to May 31, 2021 (pre-Delta period)
Yes 48 (1.94 %) 29 (7.25%)
No 2430 (98.06%) 371 (92.75%)
Total 2478 (100%) 400 (100%) P<0.0001
SARS-CoV-2 infection – June 1, 2021 to Oct 31, 2021 (Delta period)
Yes 524 (19.58%) 48 (23.76%)
No 2152 (80.42%) 154 (76.24%)
Total 2676 (100%) 202 (100%) P=0.1697
SARS-CoV-2 infection (combine)
Yes 572 (21.4%) 77 (38.1%)
No 2104 (78.6%) 125 (61.9%)
Total 2676 (100%) 202 (100%) P<0.0001
Since vaccine effectiveness might be reduced due to the presence of new virus variants, we compared infection rates between the period when the B.1.470 virus variant was still dominant (January – May 2021) and the period when the Delta variant (B.1.617.2) was dominant (June – October 2021). Our observations showed that the incidence of SARS-CoV-2 infections amongst fully vaccinated participants was lower than incomplete/unvaccinated participants during January-May 2021 and was statistically significant (1.93% vs 7.25%, P<0.0001), indicating a vaccine effectiveness of around 73.3%. In contrast, in the period between Jun – Oct 2021, there were 19.6% incidence of breakthrough infection compared to 23.8% infection amongst unvaccinated individuals (P=0.1697) with the vaccine effectiveness markedly reduced to 17.6% (Table 3).
Incidence and characteristics of hospitalised patients
We analysed the number of hospitalised subjects due to COVID-19. Although there was an increase in the total number of hospitalised participants during the post vaccination period, we observed a reduction in the proportion of hospitalisation/total infected participants from 23.5% during the pre-vaccination period to 14.3 % during the post-vaccination period (Table 4 ). We then assessed the demographic pattern and the presence of co-morbidities of hospitalised participants. As expected, the hospitalised subjects in both periods of study were significantly older and predominantly male (Table 4). We also found that there was higher incidence of hospitalisation in participants with body mass index (BMI) assessed obesity (BMI≥30). When we analysed specific comorbidities, we found a significant trend of higher hospitalisation in subjects with hypertension and diabetes mellitus, however, there was no significant correlation between the presence of cardiovascular and lung diseases with the incidence of hospitalisation (Table 4).Table 4 Demographic and clinical characteristics of hospitalised participants due to COVID-19
Hospitalisation (Pre-vaccination period) Hospitalisation (Post-vaccination period)
No (n=332)
(76.5%) Yes (n=102)
(23.5%) No (n=556)
(85.7%) Yes (n=93)
(14.3%)
Age 35.9 + 9.03 38.9 + 10.45 P=0.005a 36.12 + 9.1 38.5 + 11.47 P=0.028a
Sex
Male 187 (73.9 %) 66 (26.1 %) 262 (82.1 %) 57 (17.9%)
Female 145 (80.1 %) 36 (19.9 %) P=0.138b 294 (89.1 %) 36 (10.9 %) P=0.013b
Comorbidity
No comorbidity 240 (81.4 %) 56 (18.6 %) 395 (88 %) 54 (12 %)
1 comorbidity 76 (72.4 %) 29 (27.6 %) 121 (85.2 %) 21 (14.8 %)
>1 comorbidities 15 (45.5 %) 18 (54.5 %) P<0.0001c 40 (69 %) 18 (31 %) P=0.001c
BMI
BMI<30 280 (78.2 %) 78 (21.8 %) 485 (86.6 %) 75 (13.4 %)
BMI≥30 51 (68 %) 24 (32 %) P=0.072b 71 (79.8 %) 18 (20.2 %) P=0.102b
Hypertension
No 313 (79 %) 83 (21 %) 493 (86.8 %) 75 (13.2 %)
Yes 19 (50 %) 19 (50 %) P<0.0001b 63 (77.8 %) 18 (22.2 %) P=0.041b
History of DM
No 327 (77.1 %) 97 (22.9 %) 538 (86.5 %) 84 (13.5 %)
Yes 5 (50 %) 5 (50 %) P=0.060b 18 (66.7 %) 9 (33.3 %) P=0.009b
CVD History
No 326 (76.5 %) 100 (23.5 %) 552 (86 %) 90 (14 %)
Yes 6 (75 %) 2 (25 %) P=0.92b 4 (57.1 %) 3 (42.9 %) P=0.065b
Lung disease/asthma
No 310 (77.7 %) 89 (22.3%) 514 (86.4 %) 81 (13.6 %)
Yes 22 (62.9%) 13 (37.1 %) P=0.06b 42 (77.8 %) 12 (22.2 %) P=0.102b
at test; bFisher’s exact test; cChi-square test
BMI, body mass index; CVD, cardiovascular disease; DM, diabetes mellitus.
We next conducted multivariate analysis to further examine the association between demographic features, co-morbidities, and hospitalisation. As shown in Table 5 , history of hypertension appeared to have significant association with hospitalisation during the pre-vaccination period, whereas in the post-vaccination period male participants and those who had diabetes mellitus showed significant association with hospitalisation. Remarkably, when we evaluated the trend of the odds ratio (OR) of hospitalisation before and after the vaccination programme, we observed a marked reduction of OR in participants with history of hypertension whereas in other factors the OR of hospitalisation seemed comparable or only slightly changed (Fig. 3 ).Table 5 Multivariate analysis of the association between demographic features, co-morbidities and hospitalisation during pre-vaccination and post-vaccination period
Pre-vaccination period Post-vaccination period
Adjusted Odds ratio for hospitalisation (95% CI) P value Adjusted Odds ratio for hospitalisation (95% CI) P value
Age group (<30 as reference)
30 – 39 0.901 (0.486 – 1.669) 0.740 0.860 (0.474 – 1.560) 0.619
40 – 49 1.424 (0.685 – 2.961) 0.344 0.884 (0.440 – 1.775) 0.729
> 50 1.994 (0.927 – 4.292) 0.078 1.870 (0.932 – 3.751) 0.078
Sex male (female as reference) 1.425 (0.882 – 2.303) 0.148 1.773 (1.108 – 2.838) 0.017
BMI ≥ 30 (BMI<30 as reference) 1.596 (0.883 – 2.883) 0.121 1.701 (0.927 – 3.121) 0.086
HT history (no HT as reference) 2.998 (1.402 – 6.410) 0.005 1.226 (0.640 – 2.348) 0.539
DM history (no DM as reference) 3.319 (0.888 – 12.405) 0.075 2.455 (1.004 – 6.002) 0.049
CVD history (no CVD as reference) 0.400 (0.071 – 2.261) 0.300 2.711 (0.529 – 13.885) 0.232
Lung disease/asthma history (no lung disease/asthma as reference) 2.005 (0.939 – 4.284) 0.072 1.836 (0.894 – 3.767) 0.098
No complete vaccination (complete vaccination as reference) 1.224 (0.616 – 2.436) 0.564
CVD, cardiovascular disease; DM, diabetes mellitus.
Figure 3 The adjusted odds ratio of hospitalisation before and after vaccination. The adjusted odds ratio (aOR) and 95% CI to predict the risk of hospitalisation due to COVID-19 in the presence of the conditions above in comparison with the reference conditions. The reference conditions were: age < 30 years, female, BMI<30, and absence of each co-morbidity. The reduction in the risk for hospitalisation was markedly reduced in participants with hypertension after the vaccination programme. CVD: cardiovascular diseases, DM: diabetes mellitus.
Discussion
The present study described the pattern of COVID-19 incidence and hospitalisation among HCWs in two major hospitals in East Java, Indonesia, comparing the incidence between the pre-vaccination and post-vaccination period. We found that the cumulative incidence of COVID-19 infection in our cohort during the pre-vaccination period was 15.1%, which was equal to a 30-day incidence of 1.57%, whereas during the post-vaccination period the COVID-19 cumulative incidence was 22.6% (or 2.34% of 30-day incidence). This finding was comparable with previous reports, including those from Indonesia, showing that the 30-day incidence among HCWs during the first and second year of the COVID-19 pandemic was ranging around 0.5-4% 17, 18, 19, 20, 21, 22. These trends of COVID-19 cases in our cohort corroborated with the national (Indonesia) and regional (East Java) COVID-19 cases which also showed two peaks of COVID-19 at similar times during this study period23 , 24.
We found that during the first wave of the pandemic, the incidence of COVID-19 among medical staffs (physicians, residents, and medical students) and nurses were higher than in non-medical workers such as pharmacists, technicians and administrative staffs. This is obvious because nurses and medical staffs have more contact with infected patients. Furthermore, previous observations have indicated that HCWs with more direct contact with patients are at higher risk to be infected by the SARS-CoV-2 virus 18 , 19 , 21 , 22. Interestingly, during the post-vaccination period the incidence of COVID-19 in medical staff were comparable to incidence in non-medical staffs. The medical staffs, those who worked in the frontline, were prioritised to receive vaccinations. This might explain the finding that in non-medical staff, there was steeper increase in SARS-CoV-2 infection compared to medical staff who received vaccinations. In addition, it is possible that at the later stage of the pandemic, many HCWs had less compliance in using personal protective equipment (PPE). However, it is important to note that there was no change in the official guidelines for HCWs on the use of PPEs during work in hospitals between the pre-vaccination and post-vaccination periods. Thus, any reduction in compliance might be related to the decrease in anxiety level and less fear of contracting COVID-19 in HCWs following complete vaccination, as indicated by a study in South Korea25. Indeed, risk perception and distress are likely to influence the compliance in using correct PPE as suggested in previous observations 26 , 27.
Our study provides a remarkable finding that the cumulative incidence of COVID-19 was higher during the post-vaccination period compared to that of the pre-vaccination period. Detailed analysis of the data revealed that the incidence of COVID-19 declined during the first three months of the vaccination programme. However, by the end of May 2021 the infection rate was dramatically increased, which coincided with the emergence of the Delta variant of SARS-CoV-2 virus. The increase in SARS-CoV-2 infection during this period is in line with the regional (East Java) and national (Indonesian) epidemic curve23. It is very likely that the protective effect of the inactivated viral vaccine was significantly reduced against the Delta variant as indicated by the marked decline in vaccine effectiveness from 73.3% to 17.6% when Delta variant became the dominant variant in the province. This agrees with previous observations showing the reduction of vaccine effectiveness against the emerging Delta variant 28, 29, 30, 31. However, it is also important to note that although the infection rate was higher in the post-vaccination period, the COVID-19 incidence during this time was significantly lower in fully vaccinated participants compared to incomplete/unvaccinated individuals (21.4 % vs 38.1%, Table 3). This finding suggests that there is still some degree of protection of vaccination despite the emergence of the Delta variant.
Another possible explanation is the waning of the immune protection overtime. Our separate study revealed that the serum antibody levels in HCWs vaccinated with inactivated viral vaccine was at the highest at one month following vaccination and then gradually decreased at 3-5 months after vaccination32. This finding is in line with similar observations using different types of COVID-19 vaccines7 , 8 , 33. As the peak of COVID-19 infection in our cohort occurred around 3-5 months after the vaccination programme, it is possible that the reduction of the immune protection in combination with the emergence of Delta variant contributed to the surge of COVID-19 incidence between June – August 2021.
Another important observation in this study is the pattern of hospitalisation in the early phase of the pandemic compared to the post-vaccination period. It is not surprising that the total number of the hospitalised participants was higher in the second period of the study because the total number of infections was significantly higher in that period. However, the proportion of hospitalised subjects/total infected participants was markedly reduced from 23.5% in the pre-vaccination period to 14.3% in the post-vaccination period. This might be due to the combination of several factors. First, the severity of COVID-19 in most of the participants might be milder due to the protective effect of the vaccine. It is widely believed that the vaccination programme has successfully reduced hospitalisation as indicated in some real-world vaccine effectiveness reports34, 35, 36 and this might also be the case in our cohort. Second, the reduction in the rate of hospitalisations might be due to the change in the policy in determining the criteria for hospitalisation. During the second wave of COVID-19 in Indonesia, the Ministry of Health provided guidance with stricter criteria for hospitalisation37 , 38, in which only severe patients with hypoxia or those with severe comorbidities were admitted to the hospital. Patients who did not experience hypoxia or breathing difficulties were advised to self-isolate and be treated at their home or at a community quarantine facility. In contrast, in the early phase of the pandemic, the Indonesian authority was more focused on tracing efforts, identifying patients, and preventing transmission in the community, so that all the cases that met the definition criteria of COVID-19 (including those with only mild to moderate symptoms) had to be treated at several referral hospitals determined by the government in each province, district and city for treatment and also for isolation purposes. The change in policy might explain the reduction in the hospitalisation rate and underscore the importance of having a clear protocol from the national authority in providing the best strategy for tackling the surge of COVID-19 cases.
One of the most interesting findings of this study is the analysis of the association between demographic characteristics, the presence of co-morbidities and the likelihood of being hospitalised. Unsurprisingly, males and older subjects (≥ 50 years) showed higher odds ratio to be hospitalised which agrees with previous reports 14 , 39 , 40. The adjusted odds ratios (aORs) from the multivariate analysis showed that participants with history of hypertension, diabetes mellitus and lung disease as well as BMI≥30 (obesity) also displayed higher chances to be hospitalised. Hypertension appeared to be the strongest determinants in the pre-vaccination period with P value of 0.005. Interestingly, a marked reduction in the aORs between the pre- and post-vaccination periods was observed in subjects with hypertension. Of note, there were trends of significance (with P value between 0.05 – 0.1) of the aORs of being hospitalized in participants with history of diabetes mellitus, lung diseases and in older participants (age ≥ 50 years). However, unlike hypertension, the aORs of these comorbidities were comparable between the pre- and post-vaccination periods. Age, sex, and the presence of comorbidities such as hypertension, diabetes mellitus and lung diseases have been strongly associated with the severity of COVID-19 as indicated in previous publications14, 15, 16. Alteration of the immune system has been proposed as the key mechanism leading to severe COVID-19 in older people41and individuals with comorbidities such as hypertension42 and diabetes mellitus43. However, it remains unclear as to why participants with hypertension had reduced risk of being hospitalised after the vaccination programme while participants with other co-morbidities did not display such risk reduction. Further studies are needed to understand this phenomenon. Nevertheless, this finding underlines the importance of protecting vulnerable people with comorbidities by using vaccinations.
Despite the findings, we acknowledge some limitations of the study. First, our study was a retrospective study and some of the data were obtained from the completion of the questionnaire. We acknowledge that there might be some degree of bias in providing information regarding demographic characteristics and the presence of comorbidities. However, since most of the participants were HCWs, we believe that they had a good knowledge concerning health condition and so the possible inaccuracy of the information regarding their health condition should be minimum. Second, we were not able to assess COVID-19 death rates during the period of the study because this study was a retrospective study involving healthcare workers who were non-infected or survived from COVID-19 during the period of the study. Another limitation was regarding the analysis of the dominant SARS-CoV-2 variant during the period of the study. Ideally, the data of SARS-CoV-2 variants should be obtained directly from the participants of this study. However, due to the limitation of the DNA sequencing facilities in our centre, we could only analyse sequence data in the GISAID database. However, we believe that the data accurately represent the dominant strains that circulated in East Java in the respective times.
In conclusion, our analyses provide accurate information on the incidence, hospitalisation, and characteristics of COVID-19 cases among HCWs in East Java, Indonesia comparing between the pre- and post-vaccination periods. We observed that the risk to contract COVID-19 amongst HCWs in two major hospitals in East Java remains high despite the high coverage of vaccination. However, the rate of hospitalisation was reduced in the post-vaccination period. Information concerning the burden due to SARS-CoV-2 infection among HCWs is essential, first to ensure the health and safety of the HCWs, and second to provide representative data on the outbreak itself since the findings may represent what is happening in the wider population. The findings can be used to increase public awareness and to provide the authorities with recommendations related to COVID-19 management, the use PPE among HCWs, and vaccination including the need of booster doses.
Supplementary data
Research in context
Evidence before this study
Several studies have reported a relatively higher incidence of COVID-19 in healthcare workers (HCWs) compared to the wider population. However, there is only limited information comparing the characteristics of COVID-19 cases, hospitalisation, the effects of vaccination and the presence of co-morbidities between the periods before the vaccination programme started and after the vaccination programme was implemented in HCWs.
Added value of this study
The 30-day incidence of COVID-19 among HCWs in two major hospitals in East Java increased from 1.57% before the start of the vaccination to 2.34% after the vaccination programme. The increased incidence of COVID-19 could be attributed to the emergence of the delta variant. In comparison, the vaccine effectiveness in protecting against SARS-CoV-2 infection in HCWs was around 73% before the emergence of the delta variant but decreased to 17.6% at the period when the delta variant became dominant. However, there was a reduction in the rate of hospitalisation from 23.5% in the pre-vaccination period to 14.3% in the post-vaccination period. Our data adds new knowledge regarding the different pattern of COVID-19 cases and hospitalisation among HCWs before and after vaccination as well as the efficacy of vaccination using inactivated virus vaccine in HCWs.
Implications of the available evidence
Our data suggested that the risk of contracting COVID-19 among HCWs in East Java, Indonesia remains high, both before and after the vaccination programme. Although vaccination is important to reduce the rate of infection and hospitalisation, its effectiveness could decline overtime and during the emergence of new virus variants. Effective measures to protect HCWs from contracting COVID-19 always need to be reviewed and strictly implemented during the pandemic and beyond.
Acknowledgments
The authors would like to thank Anisa Octaviani and Astri Nur Amalia for administrative support.
Contributors
GS conceived the original idea, designed the study, collected and analysed all research data, wrote and edited manuscript; DP designed the study, supervised data analysis, edited manuscript; LW designed the study, managed funding, collected and analysed research data; BAM collected clinical data, performed additional data analysis; KDF, STH, HIG, MEP, DP, PPN, NA collected participants’ data; CRSP supervised, directed, and coordinated the project; AE supervised the project; DGAS supervised the vaccination programme and the project; DT supervised the ethical clearance; EBR supervised the vaccination programme and the project; EAT supervised the project; JW authorised the vaccination programme at Dr. Soetomo Hospital and supervised the project; CBK, FEW, FM collected participants’ and clinical data at Bangkalan Hospital; NK authorised the vaccination programme at Bangkalan Hospital and supervised the project; AB, DF supervised the project and provided suggestions during manuscript writing; WKN supervised the project; DO designed the study, performed data analysis, and wrote and edited the manuscript. All authors have read and approved the final manuscript.
Data sharing statement
The datasets created and analysed during this study are available from the corresponding authors upon reasonable request.
Declaration of interests
None.
==== Refs
References
1 Ramchandani R. Kazatchkine M. Liu J. Vaccines, therapeutics, and diagnostics for covid-19: redesigning systems to improve pandemic response BMJ 375 2021 e067488
2 Han B. Song Y. Li C. Safety, tolerability, and immunogenicity of an inactivated SARS-CoV-2 vaccine (CoronaVac) in healthy children and adolescents: a double-blind, randomised, controlled, phase 1/2 clinical trial Lancet Infect Dis 21 12 2021 1645 1653 34197764
3 Jara A. Undurraga E.A. Gonzalez C. Effectiveness of an inactivated SARS-CoV-2 vaccine in Chile N Engl J Med 385 10 2021 875 884 34233097
4 COVID-19 Vaccine Tracker Team. COVID-19 Vaccine Tracker: Approved Vaccines. 2022. https://covid19.trackvaccines.org/vaccines/approved/ (accessed 04/07/2022 2022).
5 Mathieu E. Ritchie H. Ortiz-Ospina E. A global database of COVID-19 vaccinations Nature Human Behaviour 5 7 2021 947 953
6 Fadlyana E. Rusmil K. Tarigan R. A phase III, observer-blind, randomized, placebo-controlled study of the efficacy, safety, and immunogenicity of SARS-CoV-2 inactivated vaccine in healthy adults aged 18-59 years: An interim analysis in Indonesia Vaccine 39 44 2021 6520 6528 34620531
7 Chemaitelly H. Tang P. Hasan M.R. Waning of BNT162b2 Vaccine Protection against SARS-CoV-2 Infection in Qatar N Engl J Med 385 24 2021 e83 34614327
8 Levin E.G. Lustig Y. Cohen C. Waning Immune Humoral Response to BNT162b2 Covid-19 Vaccine over 6 Months N Engl J Med 385 24 2021 e84 34614326
9 Pouwels K.B. Pritchard E. Matthews P.C. Effect of Delta variant on viral burden and vaccine effectiveness against new SARS-CoV-2 infections in the UK Nat Med 27 12 2021 2127 2135 34650248
10 Singanayagam A. Hakki S. Dunning J. Community transmission and viral load kinetics of the SARS-CoV-2 delta (B.1.617.2) variant in vaccinated and unvaccinated individuals in the UK: a prospective, longitudinal, cohort study Lancet Infect Dis 22 2 2022 183 195 34756186
11 Lwanga S, Lemeshow S. Sample size determination in health studies: A practical manual. Geneva: World Health Organization; 1991.
12 Khare S. Gurry C. Freitas L. GISAID's Role in Pandemic Response China CDC Wkly 3 49 2021 1049 1051 34934514
13 Cohen A.L. Taylor T. Jr. Farley M.M. An assessment of the screening method to evaluate vaccine effectiveness: the case of 7-valent pneumococcal conjugate vaccine in the United States PLoS One 7 8 2012 e41785
14 Docherty A.B. Harrison E.M. Green C.A. Features of 20 133 UK patients in hospital with covid-19 using the ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study BMJ 369 2020 m1985 32444460
15 Gao C. Cai Y. Zhang K. Association of hypertension and antihypertensive treatment with COVID-19 mortality: a retrospective observational study Eur Heart J 41 22 2020 2058 2066 32498076
16 Guan W.J. Liang W.H. Zhao Y. Comorbidity and its impact on 1590 patients with COVID-19 in China: a nationwide analysis Eur Respir J 55 5 2020
17 Soebandrio A. Kusumaningrum T. Yudhaputri F.A. COVID-19 prevalence among healthcare workers in Jakarta and neighbouring areas in Indonesia during early 2020 pandemic Ann Med 53 1 2021 1896 1904 34783269
18 Doernberg S.B. Holubar M. Jain V. Incidence and prevalence of COVID-19 within a healthcare worker cohort during the first year of the SARS-CoV-2 pandemic Clin Infect Dis 2022
19 Inghels M. Kane R. Lall P. Ethnicity and risk for SARS-CoV-2 infection among the healthcare workforce: Results of a retrospective cohort study in rural United Kingdom Int J Infect Dis 122 2022 115 122 35569751
20 Nguyen L.H. Drew D.A. Graham M.S. Risk of COVID-19 among front-line health-care workers and the general community: a prospective cohort study Lancet Public Health 5 9 2020 e475 e483 32745512
21 Platten M. Nienhaus A. Peters C. Cumulative Incidence of SARS-CoV-2 in Healthcare Workers at a General Hospital in Germany during the Pandemic-A Longitudinal Analysis Int J Environ Res Public Health 19 4 2022
22 Sabetian G. Moghadami M. Hashemizadeh Fard Haghighi L. COVID-19 infection among healthcare workers: a cross-sectional study in southwest Iran Virol J 18 1 2021 58 33731169
23 Satuan Tugas Penanganan COVID-19. Peta Sebaran COVID-19 - Jawa Timur. 2022. https://covid19.go.id/peta-sebaran (accessed 4/7/2022 2022).
24 World Health Organization. Indonesia: WHO Coronavirus (COVID-19) Dashboard. 2022. https://covid19.who.int/region/searo/country/id (accessed 04/07/2022 2022).
25 Park D.H. Lee E. Jung J. Changes in Anxiety Level and Personal Protective Equipment Use Among Healthcare Workers Exposed to COVID-19 J Korean Med Sci 37 16 2022 e126 35470600
26 Chia S.E. Koh D. Fones C. Appropriate use of personal protective equipment among healthcare workers in public sector hospitals and primary healthcare polyclinics during the SARS outbreak in Singapore Occup Environ Med 62 7 2005 473 477 15961624
27 Kim J.S. Choi J.S. Middle East respiratory syndrome-related knowledge, preventive behaviours and risk perception among nursing students during outbreak J Clin Nurs 25 17-18 2016 2542 2549 27273475
28 Bruxvoort K.J. Sy L.S. Qian L. Effectiveness of mRNA-1273 against delta, mu, and other emerging variants of SARS-CoV-2: test negative case-control study BMJ 375 2021 e068848
29 Lin D.Y. Gu Y. Wheeler B. Effectiveness of Covid-19 Vaccines over a 9-Month Period in North Carolina N Engl J Med 386 10 2022 933 941 35020982
30 McKeigue P.M. McAllister D.A. Hutchinson S.J. Robertson C. Stockton D. Colhoun H.M. Vaccine efficacy against severe COVID-19 in relation to delta variant (B.1.617.2) and time since second dose in patients in Scotland (REACT-SCOT): a case-control study Lancet Respir Med 10 6 2022 566 572 35227416
31 Mlcochova P. Kemp S.A. Dhar M.S. SARS-CoV-2 B.1.617.2 Delta variant replication and immune evasion Nature 599 7883 2021 114 119 34488225
32 Soegiarto G. Wulandari L. Purnomosari D. Hypertension is associated with antibody response and breakthrough infection in health care workers following vaccination with inactivated SARS-CoV-2 Vaccine 40 30 2022 4046 4056 35660034
33 Israel A. Shenhar Y. Green I. Large-Scale Study of Antibody Titer Decay following BNT162b2 mRNA Vaccine or SARS-CoV-2 Infection Vaccines (Basel) 10 1 2021
34 Ismail AlHosani F. Eduardo Stanciole A. Aden B. Impact of the Sinopharm's BBIBP-CorV vaccine in preventing hospital admissions and death in infected vaccinees: Results from a retrospective study in the emirate of Abu Dhabi, United Arab Emirates (UAE) Vaccine 40 13 2022 2003 2010 35193793
35 Mousa M. Albreiki M. Alshehhi F. Similar effectiveness of the inactivated vaccine BBIBP-CorV (Sinopharm) and the mRNA vaccine BNT162b2 (Pfizer-BioNTech) against COVID-19 related hospitalizations during the Delta outbreak in the United Arab Emirates J Travel Med 2022
36 Ye C. Lv Y. Kuang W. Inactivated vaccines prevent severe COVID-19 in patients infected with the Delta variant: A comparative study of the Delta and Alpha variants from China J Med Virol 94 8 2022 3613 3624 35365888
37 Kementerian Kesehatan Republik Indonesia. Keputusan Menteri Kesehatan Republik Indonesia Nomor HK.01.07/MENKES/4641/2021 Tentang Panduan Pelaksanaan Pemeriksaan, Pelacakan, Karantina, dan Isolasi Dalam Rangka Percepatan Pencegahan Dan Pengendalian Coronavirus Disease 2019 (COVID-19). Jakarta; 2021.
38 Kementerian Kesehatan Republik Indonesia. Keputusan Menteri Kesehatan Republik Indonesia Nomor HK.01.07/MENKES/5671/2021 Tentang Manajemen Klinis Tata Laksana Corona Virus Disease 2019 (COVID-19) di Fasilitas Pelayanan Kesehatan. Jakarta; 2021.
39 Garg S. Kim L. Whitaker M. Hospitalization Rates and Characteristics of Patients Hospitalized with Laboratory-Confirmed Coronavirus Disease 2019 - COVID-NET, 14 States, March 1-30, 2020 MMWR Morb Mortal Wkly Rep 69 15 2020 458 464 32298251
40 Petrilli C.M. Jones S.A. Yang J. Factors associated with hospital admission and critical illness among 5279 people with coronavirus disease 2019 in New York City: prospective cohort study BMJ 369 2020 m1966 32444366
41 Bartleson J.M. Radenkovic D. Covarrubias A.J. Furman D. Winer D.A. Verdin E. SARS-CoV-2, COVID-19 and the aging immune system Nature Aging 1 9 2021 769 782 34746804
42 Muhamad S.-A. Ugusman A. Kumar J. Skiba D. Hamid A.A. Aminuddin A. COVID-19 and Hypertension: The What, the Why, and the How Frontiers in Physiology 2021 12
43 Lim S. Bae J.H. Kwon H.-S. Nauck M.A. COVID-19 and diabetes mellitus: from pathophysiology to clinical management Nature Reviews Endocrinology 17 1 2021 11 30
| 0 | PMC9742226 | NO-CC CODE | 2022-12-14 23:29:59 | no | Lancet Reg Health Southeast Asia. 2022 Dec 12;:100130 | utf-8 | Lancet Reg Health Southeast Asia | 2,022 | 10.1016/j.lansea.2022.100130 | oa_other |
==== Front
RAE
sprae
Review & Expositor
0034-6373
2052-9449
SAGE Publications Sage UK: London, England
10.1177/00346373221132201
10.1177_00346373221132201
Original Research Article
Are Christians morally obligated to be vaccinated for COVID-19?
Eberl Jason T.
Saint Louis University, USA
Jason T. Eberl, Saint Louis University, St. Louis, MO 63103-2097, USA. Email: [email protected]
9 12 2022
9 12 2022
00346373221132201© The Author(s) 2022
2022
Review & Expositor, Inc
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
As the COVID-19 pandemic persists and new vaccine boosters targeting the latest subvariants have been approved, public debate concerning vaccines and vaccination mandates has not subsided. Such debate has been particularly acute among Roman Catholics and other Christians, with arguments having been put forth from scriptural and natural law bases in favor of vaccination against COVID-19, and counterarguments based on respecting individual conscience and concerns about moral complicity with abortion. In this article, I argue that principles of both secular public health and Christian social ethics justify vaccination mandates for COVID-19. I further show why certain objections Christians may have are ill-founded and conclude that no moral reason exists for a Christian to refuse to be vaccinated for COVID-19; rather, vaccination for COVID-19 is a moral obligation.
abortion
conscience
COVID-19
moral complicity
Thomas Aquinas
vaccination
edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
As the COVID-19 pandemic persists and new vaccine boosters targeting the BA.4 and BA.5 subvariants were approved,1 public debate concerning vaccines and vaccination mandates has not subsided. Such debate has been particularly acute among Christians. For example, one study in June 2021 showed that 24% of White Evangelical Christians in the United States refuse to be vaccinated for COVID-19.2 Another study showed that a significant number of White Evangelical Christians who were open to vaccination before the vaccines were approved became resistant to pro-vaccination messaging by the time the vaccines had become widely available.3 To counter hesitancy and resistance to being vaccinated among Christians, the National Council of Churches has argued from a scriptural basis in favor of vaccination against COVID-19:As Christians, our faith tradition is clear and consistent about how we are to treat others. We are taught that each person is made in the image and likeness of God and we are to be concerned about their well-being. Christ, as our example, reminds us in both his words and his actions that how we treat the most vulnerable matters to God. Indeed, there are different interpretations of scriptures and while some have tried to distort our sacred texts to justify discouraging people from taking the vaccine, the ones highlighted below amplify the importance of how taking a safe and proven vaccine that can save lives is not only an act of faith, it is also a moral choice and the right thing to do.
“You shall love your neighbor as yourself.” There is no other commandment greater than these. (Mark 12:31 NRSV)
Bear one another’s burdens, and in this way you will fulfill the law of Christ. (Galatians 6:2 NRSV)
Let each of you look not to your own interests, but to the interests of others. (Philippians 2:4 NRSV)
For we are what he has made us, created in Christ Jesus for good works, which God prepared beforehand to be our way of life. (Ephesians 2:10 NRSV)
Do not be deceived, my beloved. Every generous act of giving, with every perfect gift, is from above, coming down from the Father of lights, with whom there is no variation or shadow due to change. In fulfillment of his own purpose he gave us birth by the word of truth, so that we would become a kind of first fruits of his creatures. (James 1:13-18 NRSV)4
In addition to Evangelical and mainline Protestant Christians, among Roman Catholics debate concerning vaccination has been vociferous. On one side, the Catholic Medical Association (CMA), a professional organization of Catholic health care providers, and the National Catholic Bioethics Center (NCBC), an independent and influential “think tank,” argue for exemptions for those who object to receiving a COVID-19 vaccine for religious or moral reasons.5 Yet, several Catholic bishops have instructed their priests not to sign religious exemption requests or have mandated vaccinations for diocesan employees, and Vatican City instituted a mandate without allowing non-medical exemptions.6 In this article, I argue that principles of both secular public health and Christian social ethics justify vaccination mandates for COVID-19. I further show why certain objections Christians may have are ill-founded and conclude that no moral reason exists for a Christian to refuse to be vaccinated for COVID-19; rather, vaccination for COVID-19 is a moral obligation.7
Ethical principles supporting COVID-19 vaccination mandates8
The commonly held principles of biomedical ethics—respect for autonomy, nonmaleficence, beneficence, and justice—are primarily applicable in the clinical and research contexts in which the primary concern is the protection of individual patients or research participants.9 A complementary set of principles for the public health context primarily concerns the general welfare of society, which may be impacted by individuals’ choices.10 The first principle stipulates that one’s liberty may be restricted to prevent risk of harm to others,11 and ample evidence exists of the risks from lack of vaccination.12 If restriction of liberty is warranted, the least restrictive means should be used: one should start with education, then inducement, and, if such measures are not effective and the public health threat is sufficiently significant, coercive or punitive measure may be employed. Society has witnessed each of these steps since the vaccines became available. Reciprocity demands that mandated vaccines be made freely available and also informs efforts to minimize the penalties for those who choose not to be vaccinated under a mandate, such as requiring regular testing or mask-wearing13 or reconfiguring one’s job (e.g., reassigning a nurse to perform non-patient-facing functions within a hospital). Finally, transparency requires that all stakeholders have a voice in the public deliberation and ultimate determination of public policy, which does not entail that all stakeholders will get their way.
These secular public health principles cohere with key principles of the natural law and Christian social ethics. Beginning with Thomas Aquinas’s exhortation to exercise proper stewardship over one’s body: “It is prescribed that a human being sustains [their] body, for otherwise [they] murder [them]self. . . . Therefore, one is bound to nourish [their] body, and we are bound likewise with respect to all other things without which the body cannot live.”14 There is both a personal moral obligation to safeguard one’s health and an obligation for public authorities to help cultivate this and other virtuous dispositions: “Legislators make men virtuous by habituating them to virtuous works by means of statutes, rewards, and punishments.”15 Aquinas defines civil laws as those made by appropriate authorities, utilizing prudential reason to craft ordinances that serve the common good.16 Vaccination mandates that have been made by public authorities, and within various health care, educational, and other institutions, having reasoned through the relevant epidemiological evidence, are licit expressions of civil law serving the common good and fostering a more virtuous citizenry.
Although Catholic ecclesial leaders are not recognized as magisterial authorities by all Christians, elucidating how they have articulated the notion of the “common good” and the appropriate role of civil authorities in promoting it may be helpful. According to the Fathers of the Second Vatican Council (1962-5), the common good is “the sum total of social conditions which allow people, either as groups or as individuals, to reach their fulfillment more fully and more easily.”17 The Catechism of the Catholic Church further exhorts, “The dignity of the human person requires the pursuit of the common good. Everyone should be concerned to create and support institutions that improve the conditions of human life.”18 Concerning the proper function of public authorities, the Catechism concludes, “It is the role of the state to defend and promote the common good of civil society”19 and that “it is the proper function of authority to arbitrate, in the name of the common good, between various particular interests.”20
Insofar as vaccination mandates help to create a social condition, namely, herd immunity,21 that allows people to reach their fulfillment more fully and more easily and it is evident how the pandemic, particularly when lockdowns and social distancing have been required, has inhibited such fulfillment in terms of the economy, education, and mental health,22 the state is fulfilling its proper role to promote the common good and respect the dignity of the human person. Furthermore, the common good requires considering the “sum total” of relevant social conditions while also keeping in the forefront Christ’s exhortation in Matt 25:31-46 to care especially for the poor and vulnerable, that is, “the least of these.” Hence, the vulnerability of persons who cannot be vaccinated for medical reasons, the economic impact of lockdowns and quarantines, the amelioration of politically created health disparities,23 and the particular vulnerabilities experienced by persons with disabilities24 should be taken into account.
Conscience-based objections to COVID-19 vaccination
The primary basis for refusals to be vaccinated on the part of Christians who are pro-life is its remote material causal link to abortion. The Johnson & Johnson/Janssen vaccine was manufactured using an immortalized cell line (PER.C6) developed from the retina of a fetus aborted in 1985; the Moderna and Pfizer/BioNTech vaccines were tested using a cell line (HEK 293) developed from the kidney of a fetus that either was aborted or had naturally miscarried in 1972.25 Catholic theologians working in the Congregation for the Doctrine of the Faith (CDF), the chief doctrinal office in the Vatican, analyzed whether receiving a vaccine that was materially connected to abortion constitutes illicit cooperation with moral evil. The CDF concluded that “it is morally acceptable to receive Covid-19 vaccines that have used cell lines from aborted fetuses in their research and production process.”26
The starting point of the CDF’s analysis is the first precept of natural law, which demands that people do good and avoid evil.27 In actuality, however, many human actions have more than one effect, and a single action may have both good and evil effects, leading to the formulation of the principle of double-effect.28 Furthermore, a human being does not act in isolation from others and, in many instances, the actions of multiple agents are interconnected. When one agent’s action, which is good, intersects with another agent’s action, which is evil, one must carefully distinguish between licit and illicit cooperation in evil. The basis for this distinction rests in the intention of the first agent and the distance between the first agent’s good action and the other’s evil action.
Formal cooperation in evil occurs when an agent approves of another’s evil action and may be either explicit or implicit. In the former, an agent directly intends to cooperate in another’s evil action for the end of the act itself. In the latter, an agent intends to cooperate in evil, not for the end of the evil act, but rather for the end of some concurrent good. Both explicit and implicit formal cooperation are illicit because intending evil, either as means to an end or as an end in itself, is morally wrong.29
Material cooperation occurs when an agent is instrumental in another’s evil action without approving of the action. Material cooperation can be licit, but only if sufficiently removed from the evil action. In particular, one must look at the causal chain of mediating agents between the acting agent and the commission of the evil act. If the material cooperation is immediate, meaning the agent is causally proximate to the commission of the evil action, then the cooperation is illicit. If the material cooperation is mediate, meaning the agent is causally remote from the commission of the evil action, then cooperation may be licit, provided a proportionate reason exists for the agent to cooperate in the commission of the act. Finally, cooperation, of any of the above types, may be either active or passive depending on whether the agent performs an action that causally contributes toward another’s evil action or refrains from acting in a way that would prevent another’s evil action from occurring.
Applying these categories, the CDF concluded that using one of the vaccines in question constitutes remote, passive material cooperation, the lowest degree of cooperation, and is proportionately justified by the “grave danger” posed by COVID-19:The fundamental reason for considering the use of these vaccines morally licit is that the kind of cooperation in evil (passive material cooperation) in the procured abortion from which these cell lines originate is, on the part of those making use of the resulting vaccines, remote. The moral duty to avoid such passive material cooperation is not obligatory if there is a grave danger, such as the otherwise uncontainable spread of a serious pathological agent30—in this case, the pandemic spread of the SARS-CoV-2 virus that causes Covid-19. It must therefore be considered that, in such a case, all vaccinations recognized as clinically safe and effective can be used in good conscience with the certain knowledge that the use of such vaccines does not constitute formal cooperation with the abortion from which the cells used in production of the vaccines derive.31
Another analytical lens through which one should examine the question of using vaccines derived from aborted fetal tissue is whether it constitutes illicit appropriation of a past evil. Since the known or suspected abortions from which the immortalized cell lines (PER.C6 and HEK 293) were developed occurred in the past, analyzing one’s moral complicity in terms of “cooperation,” as the CDF does, may not be conceptually accurate. Catholic moral theologian M. Cathleen Kaveny has devised the concept of “appropriation,” the parameters of which closely align with those of cooperation, to characterize more accurately what is occurring when a moral agent benefits from a past, completed evil action.32
The first question to ask regarding appropriation of a past evil is whether doing so implies ratification of the action performed, analogous to formal cooperation. The next question is whether appropriation would have a corrosive effect on the character of the appropriating agent, which Kaveny identifies as potentially occurring via seepage and self-deception:Seepage and self-deception are no less vivid possibilities in appropriation cases than cases of cooperation. If another agent’s evil acts contribute in some way to our own objectives, particularly in an ongoing manner, it is difficult not to view them in a more positive light than we otherwise would. Moreover, it is tempting so to accustom ourselves to the benefits that flow from appropriation that we would be inclined to decide against taking steps to eliminate the wrongdoing, if the opportunity presented itself.33
A correlative question is whether the appropriation might have a similarly negative impact on the character of other moral agents, which would constitute scandal. Aquinas defines scandal as “something less rightly done or said, that occasions another’s spiritual [moral] downfall” and occurs when one person by “injunction, inducement or example, moves another to sin.”34 He further specifies two types of scandal. Active scandal occurs, when oneeither intends, by his evil word or deed, to lead [another] into sin, or, if [they do] not so intend, when [their] deed is of such a nature as to lead another into sin: for instance, when [one] publicly commits a sin or does something that has an appearance of sin.35
Passive scandal occurs “when [one] neither intends to lead [another] into sin, nor does what is of a nature to lead [another] into sin, and yet this other one, through being ill-disposed, is led into sin.”36 Aquinas contends that active scandal is always an occasion of moral wrongdoing on the part of the agent who scandalizes another, analogous to formal cooperation or ratification, while passive scandal may not entail moral wrongdoing on the agent’s part so long as the word or action that led to the other’s moral downfall was good in itself.37 A final condition of licit appropriation is that the risk of moral corrosion of either the appropriating agent or others must be proportionate to the benefits expected from the appropriating action.
Applying the appropriation framework to the question at hand, Melissa Moschella contends that the risk of seepage, self-deception, and scandal is low insofar as: (1) the COVID-19 vaccines developed from the PER.C6 and HEK 293 cell lines do not “rely on perpetuating the original injustice,” that is, no further abortions need be performed as these are immortalized (self-perpetuating) cell lines; (2) the vaccines “could have been developed without using any cell lines of illicit origin,” as demonstrated by the existence of COVID-19 vaccines that were not developed or tested using these cell lines; and (3) immortalized cell lines can be created “without any connection to abortion by using fetal tissue from a spontaneous miscarriage or nonviable preterm birth.”38 The latter two reasons, namely, the lack of need to use cell lines that have any material connection to abortion, create an expectation on the part of Christians and others who are morally opposed to abortion that they utilize vaccines not developed from these cell lines if possible. However, given the urgency of the COVID-19 crisis and the decisions made at various governmental and institutional levels regarding which vaccines to approve and make available, for most people choosing which vaccine they receive is not feasible.
Finally, some pro-life scholars have argued that not only ought the risk of scandal be avoided, but a positive witness to the value of unborn human life be exemplified by refusing to be vaccinated.39 At least one Christian scholar has gone so far as to call for civil disobedience to vaccination mandates.40 Moschella effectively rebuts these arguments by noting the “counter scandal” that would likely result:Indeed, the widespread (or just highly publicized) refusal of pro-lifers to take the vaccine could actually lead to a sort of counter scandal in which the pro-life movement is seen as unreasonably thwarting the crucial effort to achieve herd immunity even though taking the vaccine is not actually immoral. Moreover, this scandal could be increased by the recognition that the vaccines have no greater connection to abortion than do countless other products (like common medications and processed foods) that pro-lifers use on a regular basis without complaint.41
As Moschella concludes, Christians and others who are morally opposed to abortion may receive the available COVID-19 vaccines in good conscience insofar as “using HEK 293 [or PER.C6] or its products involves no cooperation with current evil, facilitation of future evil,42 or ratification of past evil, and because the dangers to moral character and the risk of scandal are weak.”43
Catholics are called to give “religious assent” to their Church’s teaching authority44 and thus, as I have argued elsewhere,45 do not have an appropriate basis to refuse vaccination for COVID-19 for non-medical reasons. Other Christians, however, form their consciences without relying on hierarchical authorities, though most Christian denominations have some sort of leadership structure or pastors at various levels influencing the consciences of members of their flock. The following are some examples of what the leadership or other representative bodies of various Christian churches have said about COVID-19 vaccination:
Executive Council of the Episcopal Church: “The proper and responsible use of vaccines is a duty not only to our own selves and families but to our communities. Choosing to not vaccinate, when it is medically safe, threatens the lives of others.”46
Christian Science: “For more than a century, our denomination has counseled respect for public health authorities and conscientious obedience to the laws of the land, including those requiring vaccination. Christian Scientists report suspected communicable disease, obey quarantines, and strive to cooperate with measures considered necessary by public health officials. We see this as a matter of basic Golden Rule ethics and New Testament love.”47
Global Anabaptist Health Network: “As members of the health care community and as Christians, we recognize that vaccines offer great hope for ending this pandemic. They offer personal protection and build resilience into our community and health systems. Many of our brothers and sisters in Christ have been working to bring about these interventions. They vigorously advocate for vaccination and accept it for themselves. Honour their work and example. Vaccination is a benefit that comes to us most powerfully if it is accepted broadly. We should also expose falsehoods about the harm vaccination could bring. Though the world may seek self-protection out of selfishness, we embrace vaccination as a way forward in love, accepting in our own bodies the chance to protect our neighbours, brothers and sisters (Philippians 2:3).”48
While the Southern Baptist Convention (SBC) has not formally passed any resolutions regarding vaccination for COVID-19, the SBC’s International Mission Board requires vaccination for COVID-19 for all missionaries.49 Furthermore, the Connectional Table of the United Methodist Church has declared increasing COVID-19 vaccination to be a denomination-wide “missional priority.”50 Finally, the Executive Committee of the World Council of Churches (WCC) has called on “congregations and communities to consider vaccination not only as a matter of self-protection but also as an act of care and compassion for the ‘neighbour’, for the whole community and for the future of our young people”;51 the WCC has further appointed nine church leaders as “Vaccine Champions” in partnership with UNICEF.52
Conscience and public health mandates
Returning to the intra-Catholic debate, the CMA and NCBC have centered their call for religious exemptions to COVID-19 vaccination mandates on the requirement for authorities to respect the right to express one’s conscience, citing the Catechism:[A human] has the right to act in conscience and in freedom so as personally to make moral decisions. “[They] must not be forced to act contrary to [their]conscience. Nor must [they] be prevented from acting according to [their] conscience, especially in religious matters.”53
This teaching on the nature of conscience, however, is nuanced in ways not explicitly appreciated by these organizations.54 For example, the above quotation is derived from Vatican II’s declaration on religious freedom, a document primarily concerned with totalitarian political regimes, primarily Nazism and Soviet-style communism, that inhibit the practice of religious faith, religious education, missionary outreach, and so on. The above assertion does not directly entail that public authorities have no role to play in restricting certain behaviors, conscientious though they may be, that threaten the common good:The right to religious freedom is exercised in human society: hence its exercise is subject to certain regulatory norms. In the use of all freedoms the moral principle of personal and social responsibility is to be observed. In the exercise of their rights, individual[s] and social groups are bound by the moral law to have respect both for the rights of others and for their own duties toward others and for the common welfare of all. [Individuals] are to deal with their fellows in justice and civility. . . . Furthermore, society has the right to defend itself against possible abuses committed on the pretext of freedom of religion. It is the special duty of government to provide this protection.55
While obedience to civil authorities is not an absolute moral duty, neither is freedom to act on one’s conscience, especially if one’s appeal to conscience or religious freedom is a mere pretext:Not a few can be found who seem inclined to use the name of freedom as the pretext for refusing to submit to authority and for making light of the duty of obedience. Wherefore this Vatican Council urges everyone, especially those who are charged with the task of educating others, to do their utmost to form [people] who, on the one hand, will respect the moral order and be obedient to lawful authority, and on the other hand, will be lovers of true freedom—[people], in other words, who will come to decisions on their own judgment and in the light of truth, govern their activities with a sense of responsibility, and strive after what is true and right, willing always to join with others in cooperative effort. Religious freedom therefore ought to have this further purpose and aim, namely, that [people] may come to act with greater responsibility in fulfilling their duties in community life.56
This carefully balanced view eschews the subjectivism entailed by defending a right to conscience as an absolute and denying that just civil laws bind one’s conscience.57 Pope John Paul II echoes this concern:The individual conscience is accorded the status of a supreme tribunal of moral judgment which hands down categorical and infallible decisions about good and evil. To the affirmation that one has a duty to follow one’s conscience is unduly added the affirmation that one’s moral judgment is true merely by the fact that it has its origin in the conscience. But in this way the inescapable claims of truth disappear, yielding their place to a criterion of sincerity, authenticity and “being at peace with oneself,” so much so that some have come to adopt a radically subjectivistic conception of moral judgment.58
The Vatican II Fathers connect this concern regarding “individualistic morality” with safeguarding the common good, including “the protection of health”:Profound and rapid changes make it more necessary that no one ignoring the trend of events or drugged by laziness, content [them]self with a merely individualistic morality. It grows increasingly true that the obligations of justice and love are fulfilled only if each person, contributing to the common good, according to [their] own abilities and the needs of others, also promotes and assists the public and private institutions dedicated to bettering the conditions of human life. Yet there are those who, while possessing grand and rather noble sentiments, nevertheless in reality live always as if they cared nothing for the needs of society. Many in various places even make light of social laws and precepts, and do not hesitate to resort to various frauds and deceptions in avoiding just taxes or other debts due to society. Others think little of certain norms of social life, for example those designed for the protection of health . . . they do not even avert to the fact that by such indifference they imperil their own life and that of others.59
Catholic lawyer and ethicist O. Carter Snead has rightly criticized the “expressive individualism” embodied by most public bioethics laws in the United States.60 Christians must also be careful not to utilize the same foundation of the expression of individual autonomy, recast in the name of “conscience,” above other moral concerns, such as public health insofar as it is partially constitutive of the common good.
Conclusion: Moral obligation to be vaccinated for COVID-19
The CDF states that, while receiving any of the currently approved COVID-19 vaccines is permissible, “vaccination is not, as a rule, a moral obligation and that, therefore, it must be voluntary.”61 On this basis, the CMA and NCBC defend exemptions to vaccination mandates. Some key ambiguities exist, however, in the language the CDF uses.62 First, something could be a “rule” in two ways: absolutely or prima facie. Understanding the CDF as asserting an absolute rule would contradict Pope Francis’s exhortation, “I believe that morally everyone must take the vaccine. It is the moral choice because it is about your life but also the lives of others.”63 The Pope and the CDF would be in alignment if one understands “as a rule” in a prima facie sense, meaning that, under ordinary circumstances, vaccination is not a moral obligation; the current pandemic has arguably created a “state of exception,” however, in which moral rules, though not abrogated, may be applied in different ways.64 In this case, another moral rule, the requirement to safeguard one’s health and promote the common good, overrides the prima facie rule against vaccination being a moral obligation. Second, the term “voluntary” is inherently ambiguous as it could mean either that one should not be coerced in any way to be vaccinated or that one should not be held down and jabbed against their will. While the currently implemented mandates could be construed as “coercive,” they are not forcing anyone to be vaccinated against their will; one could avoid vaccination by, for example, resigning from their job, refusing to dine in a restaurant, or traveling in a manner that does not require vaccination.
I conclude that a moral obligation to be vaccinated exists based on epidemiological evidence that vaccination, except for those with medical contraindications,65 is the most effective means of fulfilling one’s duty to safeguard their own health and promote the common good, which inherently respects the dignity of human persons, particularly those at higher risk of severe illness or death, including by attenuating the virus’s potential to mutate into more infectious and deadlier forms. Neither Christian nor secular critics of vaccination mandates have provided a countervailing moral reason of sufficient weight to forego vaccination.
Author biography
Jason T. Eberl, PhD, is Professor of Health Care Ethics and Philosophy and Director of the Albert Gnaegi Center for Health Care Ethics at Saint Louis University. His research interests include the philosophy of human nature and its application to issues at the margins of life; ethical issues related to end-of-life care, biotechnology, and healthcare allocation; and the thought of Thomas Aquinas. He is the author of Thomistic Principles and Bioethics, The Routledge Guidebook to Aquinas’ Summa Theologiae, and The Nature of Human Persons: Metaphysics and Bioethics, as well as the editor of Contemporary Controversies in Catholic Bioethics.
1. Brenda Goodman, “CDC Signs Off on Updated Covid-19 Boosters,” CNN, September 1, 2022, https://www.cnn.com/2022/09/01/health/acip-cdc-updated-covid-booster/index.html.
2. Ian Lovett, “White Evangelicals Resist Covid-19 Vaccine Most among Religious Groups,” The Wall Street Journal, July 28, 2021, https://www.wsj.com/articles/white-evangelicals-resist-covid-19-vaccine-most-among-religious-groups-11627464601.
3. Scott E. Bokemper, Alan S. Gerber, Saad B. Omer, and Gregory A. Huber, “Persuading US White Evangelicals to Vaccinate for COVID-19: Testing Message Effectiveness in Fall 2020 and Spring 2021,” Proceedings of the National Academy of Sciences 118.49 (2021): e2114762118.
4. National Council of Churches, “Fact Sheet: Christian Approach to Vaccine Hesitancy,” December 22, 2021, https://nationalcouncilofchurches.us/wp-content/uploads/2021/12/Vaccine-Hesitancy-NCC-Fact-Sheet-FINAL-2021-.pdf.
5. See “Catholic Medical Association Opposes Vaccine Mandates without Conscience and Religious Exemptions,” CMA, July 28, 2021, https://www.cathmed.org/catholic-medical-association-opposes-vaccine-mandates-without-conscience-and-religious-exemptions/, and “NCBC Statement on COVID-19 Vaccine Mandates,” NCBC, August 23, 2021, https://static1.squarespace.com/static/5e3ada1a6a2e8d6a131d1dcd/t/618c1ec220f0c56b14410bc6/1636572867052/NCBC_Statement_on_COVID-19_Vaccine_Mandates.pdf.
6. See Carol Zimmermann, “NY Priests Urged Not to Give Religious Exemptions to COVID-19 Vaccines,” National Catholic Reporter, August 5, 2021, https://www.ncronline.org/news/coronavirus/ny-priests-urged-not-give-religious-exemptions-covid-19-vaccines; Robert W. McElroy to Priests of the Diocese, August 11, 2021, The Roman Catholic Diocese of San Diego, https://www.sdcatholic.org/wp-content/uploads/news/documents/8.11.2021-Letter-from-Bishop-McElroy.pdf; Christine Rousselle, “Chicago Archdiocese Mandates COVID Vaccination for Clerics, Employees,” Catholic News Agency, August 25, 2021, https://www.catholicnewsagency.com/news/248775/archdiocese-of-chicago-mandates-covid-vaccination-for-clerics-employees; Elise Ann Allen, “Vatican Issues Vaccine Mandate for All Employees,” Crux, December 24, 2021, https://cruxnow.com/vatican/2021/12/vatican-issues-vaccine-mandate-for-all-employees. The Vatican City mandate does allow for proof of previous COVID-19 infection in place of vaccination. For more on this intra-Catholic debate, I recommend episodes 76-78 of “Bioethics on Air” podcast, sponsored by the NCBC, https://www.ncbcenter.org/bioethics-on-air-podcast-cms.
7. This article draws from a series of previous publications addressed principally to a Catholic audience: Jason T. Eberl, “Vaccine Mandates Are Coming. Catholics Have No Moral Reason to Oppose Them,” America: The Jesuit Review, August 10, 2021, https://www.americamagazine.org/faith/2021/08/10/covid-vaccine-mandate-exemptions-voluntary-ignorance-241196; Eberl and Tobias Winright, “Catholics Have No Grounds to Claim Exemption From COVID Vaccine Mandates,” National Catholic Reporter, August 17, 2021, https://www.ncronline.org/news/coronavirus/catholics-have-no-grounds-claim-exemption-covid-vaccine-mandates; Eberl, “Is There a Moral Obligation to Be Vaccinated for COVID-19?” Health Care Ethics USA, March 24, 2022, https://www.chausa.org/publications/health-care-ethics-usa/archives/issues/winter-spring-2022-articles/is-there-a-moral-obligation-to-be-vaccinated-for-covid-19; and James McTavish and Eberl, “Is COVID-19 Vaccination ‘Ordinary’ (Morally Obligatory) Treatment?” National Catholic Bioethics Quarterly 22.2 (2022): 319–33.
8. I set aside legal debates regarding the constitutionality of governmental mandates in the United States or whether mandates by employers are discriminatory, as well as the disputed question of whether previous COVID-19 infection provides sufficiently robust immunity, equivalent to or better than the currently available vaccines in efficacy and duration of protection, to warrant an exemption from vaccination. For a report on the latter, see Jennifer Block, “Vaccinating People Who Have Had COVID-19: Why Doesn’t Natural Immunity Count in the US?” British Medical Journal 374 (2021): n2101, https://doi.org/10.1136/bmj.n2101.
9. See Tom L. Beauchamp and James F. Childress, Principles of Biomedical Ethics, 8th ed. (New York: Oxford University Press, 2019).
10. See R. E. G. Upshur, “Principles for the Justification of Public Health Intervention,” Canadian Journal of Public Health 93.2 (2002): 101–103.
11. The historical root of this “harm principle” may be found in John Stuart Mill’s seminal essay On Liberty (Urbana, IL: Gutenberg Project, 2011), https://www.gutenberg.org/files/34901/34901-h/34901-h.htm.
12. Not only are the vast majority of COVID-related hospitalizations and deaths among the unvaccinated (Edouard Mathieu and Max Roser, “How Do Death Rates From COVID-19 Differ between People Who Are Vaccinated and Those Who Are Not?” Our World in Data, November 23, 2021, https://ourworldindata.org/covid-deaths-by-vaccination), the mutation of the SARS-CoV-2 virus into new variants is correlated with lower levels of vaccination (Saralyn Cruickshank, “How Viruses Mutate and What Can Be Done About It,” The Hub: Johns Hopkins University, July 19, 2021, https://hub.jhu.edu/2021/07/19/andrew-pekosz-delta-variants/).
13. Though such measures may be insufficiently effective against the newer variants. Vatican City’s vaccination mandate, for example, previously allowed for regular testing as an alternative, but it disallowed this alternative as the Omicron variant surged.
14. Thomas Aquinas, Super secundam Epistolam ad Thessalonicenses lectura, cap. III, lect. 2; my translation.
15. Thomas Aquinas, Commentary on Aristotle’s Nicomachean Ethics, bk. II, lect. 1, §251, trans. C. I. Litzinger (Notre Dame, IN: Dumb Ox Books, 1993).
16. Thomas Aquinas, Summa Theologiae, Ia, q. 95, trans. English Dominican Fathers (New York: Benziger, 1948).
17. Paul VI, Gaudium et spes, December 7, 1965, no. 26, https://www.vatican.va/archive/hist_councils/ii_vatican_council/documents/vat-ii_const_19651207_gaudium-et-spes_en.html.
18. Catechism of the Catholic Church (1997), no. 1926, https://www.vatican.va/archive/ENG0015/_INDEX.HTM.
19. Catechism, no. 1927.
20. Catechism, no. 1908.
21. Mayo Clinic Staff, “Herd Immunity and COVID-19: What You Need to Know,” Mayo Clinic, April 20, 2022, https://www.mayoclinic.org/diseases-conditions/coronavirus/in-depth/herd-immunity-and-coronavirus/art-20486808.
22. See ChaeWon Baek, Peter B. McCrory, Todd Messer, and Preston Mui, “Unemployment Effects of Stay-at-Home Orders: Evidence from High Frequency Claims Data,” IRLE Working Paper No. 101-20, July 10, 2020, https://irle.berkeley.edu/files/2020/07/Unemployment-Effects-of-Stay-at-Home-Orders.pdf; Office of Civil Rights, “Education in a Pandemic: The Disparate Impacts of COVID-19 on America’s Students,” U.S. Dept of Education, June 9, 2021, https://www2.ed.gov/about/offices/list/ocr/docs/20210608-impacts-of-covid19.pdf; Kevin Sikali, “The Dangers of Social Distancing: How COVID-19 Can Reshape Our Social Experience,” Journal of Community Psychology 48.8 (November 2020): 2435–38, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7461541/; Aaron Kheriaty, “The Other Pandemic: The Lockdown Mental Health Crisis,” Public Discourse, October 4, 2020, https://www.thepublicdiscourse.com/2020/10/71969/.
23. See Daniel E. Dawes, The Political Determinants of Health (Baltimore: Johns Hopkins University Press, 2020).
24. See Nicole Baumer, “The Pandemic Isn’t Over—Particularly for People with Disabilities,” Harvard Health Publishing, Harvard Medical School, May 25, 2021, https://www.health.harvard.edu/blog/the-pandemic-isnt-over-particularly-for-people-with-disabilities-202105252464; and Tom Shakespeare, Florence Ndagire, and Queen E. Seketi, “Triple Jeopardy: Disabled People and the COVID-19 Pandemic,” The Lancet 397.10282 (2021): 1331–33.
25. See https://theconversation.com/cells-from-human-foetuses-are-important-for-developing-vaccines-but-theyre-not-an-ingredient-157484. Worth noting is that these cell lines have been used to develop other pharmaceuticals, cosmetics, and processed food additives with no moral objections being voiced. Matthew P. Schneider, “Comparing COVID Vaccine to Other Vaccines,” Through Catholic Lenses, April 21, 2021, https://www.patheos.com/blogs/throughcatholiclenses/2021/04/comparing-covid-vaccine-to-other-vaccines/.
26. Congregation for the Doctrine of the Faith (CDF), “Note on the Morality of Using Some Anti-COVID-19 Vaccines,” (2020), no. 2, https://www.vatican.va/roman_curia/congregations/cfaith/documents/rc_con_cfaith_doc_20201221_nota-vaccini-anticovid_en.html (emphasis original). This position was recently reaffirmed by the Pontifical Academy for Life (PAL), an advisory body to the Vatican on pro-life issues. See Carol Glatz, “Vatican Academy for Life: COVID-19 Vaccines Present ‘No Ethical Problem,’” America: The Jesuit Review, December 22, 2021, https://www.americamagazine.org/politics-society/2021/12/22/covid-vaccine-ethics-papal-academy-242110. Further support of the CDF’s position has been provided by a group of prominent Catholic pro-life scholars: “Statement from Pro-life Catholic Scholars on the Moral Acceptability of Receiving COVID-19 Vaccines,” Ethics and Public Policy Center, March 5, 2021, https://eppc.org/news/statement-from-pro-life-catholic-scholars-on-the-moral-acceptability-of-receiving-covid-19-vaccines/.
27. Aquinas 1948, Ia-IIae, q. 94, a. 2.
28. Aquinas 1948, IIa-IIae, q. 64, a. 7. For further elucidation of the principle of double-effect, see P. A. Woodward, The Doctrine of Double Effect: Philosophers Debate a Controversial Moral Principle (Notre Dame, IN: University of Notre Dame Press, 2001); T. A. Cavanaugh, Double-Effect Reasoning: Doing Good and Avoiding Evil (New York: Oxford University Press, 2006); and Lawrence Masek, Character, Intention, and Double Effect (Notre Dame, IN: University of Notre Dame Press, 2018).
29. Catechism, nos. 1752 and 1755. The locus classicus for the distinction between formal and material cooperation with moral evil is Alphonsus Liguori, Theologia Moralis, ed. L. Gaudé, 4 vols. (Rome: Ex Typographia Vaticana, 1905–1912). For a more contemporary formulation, see T. A. Cavanaugh, “Cooperation: Material and Formal,” in Encyclopedia of Catholic Social Thought, Social Science, and Social Policy, ed. M. Coulter, S. M. Krason, R. S. Myers, and J. A. Varacalli (Lanham: Scarecrow Press, 2007); Kevin L. Flannery, Cooperation with Evil: Thomistic Tools of Analysis (Washington, DC: Catholic University of America Press, 2019); and Kevin L. Flannery, “Avoiding Illicit Cooperation with Evil: Alphonsus Liguori, Thomas Aquinas, and Contemporary Issues,” National Catholic Bioethics Quarterly 21.2 (2021): 231–46.
30. The CDF is drawing on conclusions reached by an earlier analysis of the moral liceity of using vaccines derived from aborted fetal tissue by the PAL: “Moral Reflections on Vaccines Prepared from Cells Derived from Aborted Human Foetuses,” (June 9, 2005), reprinted in The Linacre Quarterly 86.2–3 (2019): 182–87. These conclusions were affirmed in the CDF’s Instruction Dignitas personae on Certain Bioethical Questions (2008), nos. 34–5: https://www.vatican.va/roman_curia/congregations/cfaith/documents/rc_con_cfaith_doc_20081208_dignitas-personae_en.html.
31. CDF, “Note on the Morality,” no. 3 (emphasis original).
32. M. Cathleen Kaveny, “Appropriation of Evil: Cooperation’s Mirror Image,” Theological Studies 61 (2000): 280–313. The editor of this issue of Review & Expositor, Tobias Winright, was Kaveny’s research assistant for this article.
33. Kaveny, “Appropriation of Evil,” 307.
34. Aquinas 1948, IIa-IIae, q. 43, a. 1.
35. Aquinas 1948, IIa-IIae, q. 43, a. 1 ad 4.
36. Aquinas 1948, IIa-IIae, q. 43, a. 1 ad 4.
37. Aquinas 1948, IIa-IIae, q. 43, a. 2.
38. Melissa Moschella, “Dignitas Personae, HEK 293, and the COVID Vaccines,” National Catholic Bioethics Quarterly 21.1 (2021): 107–121 (115).
39. See Janet Smith, “The Morality of the COVID-19 Vaccines,” National Catholic Register, December 24, 2020, https://www.ncregister.com/commentaries/the-morality-of-the-covid-19-vaccines. Smith does contend that vaccination would be justified, perhaps even obligatory, for front-line workers and those at a high risk of dying from COVID-19 infection.
40. Douglas Farrow, “Whether There Is a Moral Obligation to Disobey the Coercive Mandates,” Catholic World Report, December 3, 2021, https://www.catholicworldreport.com/2021/12/03/opinion-whether-there-is-a-moral-obligation-to-disobey-the-coercive-mandates/. For a critical response to Farrow’s essay, see Jason T. Eberl, “What Would Aquinas Say in the Time of COVID-19?” Theopolis, March 3, 2022, https://theopolisinstitute.com/conversations/what-would-aquinas-say-in-the-time-of-covid-19%EF%BF%BC/.
41. Moschella, “Dignitas Personae,” 118–19.
42. In fact, as Moschella notes, “Continued use of HEK 293 actually disincentivizes the production of new fetal cell lines for general research purposes.” Moschella, “Dignitas Personae,” 111.
43. Moschella, “Dignitas Personae,” 119.
44. Lumen gentium (1964), no. 25, https://www.vatican.va/archive/hist_councils/ii_vatican_council/documents/vat-ii_const_19641121_lumen-gentium_en.html.
45. See n. 7.
46. “Advocacy for Stronger Governmental Vaccination Mandates,” Executive Council Minutes, June 10–13, 2019, The Archives of the Episcopal Church, https://www.episcopalarchives.org/cgi-bin/executive_council/EXCresolution.pl?exc_id=EXC062019.12.
47. “A Christian Science Perspective on Vaccination and Public Health,” press release, Christian Science, https://www.christianscience.com/press-room/a-christian-science-perspective-on-vaccination-and-public-health.
48. Steering Committee, “Caring for Our Brothers and Sisters,” Global Anabaptist Health Network, April 19, 2021, https://www.ccih.org/wp-content/uploads/2020/03/GAHN-letter-EN.pdf.
49. Julie McGowan, “IMB Updates Vaccination Policy to Maximize Access to Unreached Peoples,” International Mission Board, September 8, 2021, https://www.imb.org/2021/09/08/imb-updates-vaccination-policy-maximize-access-unreached-peoples/.
50. Heather Hahn, “COVID 19 Vaccination Named Missional Priority,” United Methodist Church, October 1, 2021, https://www.umc.org/en/content/covid-19-vaccination-named-missional-priority-gaf.
51. Executive Committee, “Call to Overcome Global Injustice and Inequity, to Defeat the Global COVID-19 Pandemic,” World Council of Churches, May 17–20, 2021, https://www.oikoumene.org/resources/documents/call-to-overcome-global-injustice-and-inequity-to-defeat-the-global-covid-19-pandemic.
52. “Vaccine Champions,” COVID-19 Resources, World Council of Churches, https://www.oikoumene.org/resources/covid-19-resources#vaccine-champions.
53. Catechism, no. 1782.
54. See M. Therese Lysaught, “Catholics Seeking ‘Religious’ Exemptions to Vaccines Must Follow True Church Teaching on Conscience,” National Catholic Reporter, September 21, 2021, https://www.ncronline.org/news/opinion/catholics-seeking-religious-exemptions-vaccines-must-follow-true-church-teaching.
55. Dignitatis humanae (1965), no. 7, https://www.vatican.va/archive/hist_councils/ii_vatican_council/documents/vat-ii_decl_19651207_dignitatis-humanae_en.html.
56. Dignitatis humanae, no. 8.
57. Aquinas 1948, Ia, q. 96, a. 4.
58. John Paul II, Veritatis splendor (1993), no. 32, https://www.vatican.va/content/john-paul-ii/en/encyclicals/documents/hf_jp-ii_enc_06081993_veritatis-splendor.html.
59. Paul VI, Gaudium et spes, no. 30.
60. O. Carter Snead, What It Means to Be Human: The Case for the Body in Public Bioethics (Cambridge: Harvard University Press, 2020).
61. CDF, “Note on the Morality,” no. 5.
62. For a similar interpretation of this statement, see Peter J. Cataldo, “Why the CDF ‘Note on the Morality of Using Some Anti-Covid-19 Vaccines’ Suggests a Moral Obligation to Receive SARS-CoV-2 Vaccines,” Health Care Ethics USA (Fall 2021), https://www.chausa.org/docs/default-source/hceusa/why-the-cdf-note-on-the-morality-of-using-some-anti-covid-19-vaccines-suggests-a-moral-obligation-to-receive-sars-cov-2-vaccines.pdf?sfvrsn=0. For an argument in favor of vaccination in general (prior to the outbreak of COVID-19) based on Catholic social teaching, see Paul J. Carson and Anthony T. Flood, “Catholic Social Teaching and the Duty to Vaccinate,” American Journal of Bioethics 17.4 (2017): 36–43.
63. Joshua J. McElwee, “Pope Francis Suggests People Have Moral Obligation to Take Coronavirus Vaccine,” National Catholic Reporter, January 11, 2021, https://www.ncronline.org/news/vatican/pope-francis-suggests-people-have-moral-obligation-take-coronavirus-vaccine. The Pope continues, “There is a suicidal denialism that I would not know how to explain but today people must take the vaccine.” Francis reiterated his call to be vaccinated in a Spanish-language PSA with several other prelates: “Unity Across the Americas | COVID-19 Vaccine Education Series,” August 17, 2021, Public Service Announcement, The Ad Council, https://www.youtube.com/watch?v=rU0prPIbfZ8&t=2s. One could view this PSA as a form of public fraternal correction for the sake of the common good, which is a proper function of prelates; see Aquinas 1948, IIa-IIae, q. 33, a. 3.
64. Other examples of “states of exception” include just war and circumstances of dire material scarcity; see Jason T. Eberl, “Unilateral Withdrawal of Life-Sustaining Treatment within Crisis Standards of Care,” Health Care Ethics USA (Winter 2021): 8–10, https://www.chausa.org/docs/default-source/hceusa/unilateral-withdrawal-of-life-sustaining-treatment-within-crisis-standards-of-care.pdf?sfvrsn=4.
65. See Jacqueline Stenson, “As Mandates Roll Out, Some May Ask for Medical Exemptions. What’s Really Valid?” NBC News, September 2, 2021, https://www.nbcnews.com/health/health-news/mandates-roll-out-some-may-ask-medical-exemptions-what-s-n1278264.
| 0 | PMC9742348 | NO-CC CODE | 2022-12-14 23:45:33 | no | 2022 Dec 9;:00346373221132201 | utf-8 | null | null | null | oa_other |
==== Front
RAE
sprae
Review & Expositor
0034-6373
2052-9449
SAGE Publications Sage UK: London, England
10.1177/00346373221132281
10.1177_00346373221132281
Original Research Article
“Those who are well”: Lessons from COVID for non-crisis times via Matthew 9:9-13
Kelly Conor M.
Marquette University, USA
Conor M. Kelly, Marquette University, Milwaukee, WI 53233, USA. Email: [email protected]
9 12 2022
9 12 2022
00346373221132281© The Author(s) 2022
2022
Review & Expositor, Inc
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
As the world shifts to the next phase of the pandemic, bioethicists need to consider anew what moral responsibility looks like during non-crisis times. This article turns to the calling of Matthew (Matt 9:9-13) to provide biblical insights Christians can use to contribute to this bioethical conversation. Drawing on the narrative context, which buries this pericope within a section of the gospel focusing on Jesus’s healing ministry, this article explains how the calling of Matthew underscores the holistic vision of health and well-being animating Jesus’s work as a healer and adds to Jesus’s primary emphasis on restoration for the marginalized. Examining Jesus’s claim that “those who are well have no need of a physician,” this article argues that Christians can best embrace this broad vision of healing by prioritizing public health so that the community will be better prepared to weather the next health crisis, should it emerge.
COVID-19 pandemic
mercy
public health
social determinants of health
solidarity
edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
My wife likes to say, only half-jokingly, that Matt 9:12 is her favorite scripture passage. The verse is toward the end of the Gospel’s recounting of the calling of Matthew and conveys Jesus’s response to the Pharisees who criticized his willingness to interact openly with sinners. Defending himself, Jesus offers a medical image to explain the priority he places on those who must still be called to conversion, reminding the crowds that “those who are well have no need of a physician, but those who are sick.” My wife, who is a pediatric nurse practitioner, loves to quote the first part of this line so she can remind her friends and colleagues of the valuable contributions nurses make to the health care system. In fact, when she worked in primary care, with its abundance of well child visits, she would often point to the passage as a theological endorsement of her own vocational trajectory. “Jesus was right,” she would playfully insist, “‘Those who are well have no need of a physician.’ They need a nurse practitioner!”
While my wife’s interpretation involves a slightly more dynamic translation of the original Greek than one typically finds, her commentary provides an important reminder of the fact that there is more to health and well-being, and thus health care, than the mere absence of disease. In theory, this insight should be a central component of a Christian approach to bioethics at all times, because Jesus’s own acts of healing were never oriented simply to the eradication of biological imperfections but also always included a dimension of social restoration for the sick, who were regularly isolated from their communities as a result of their disease.1 In this particular moment, however, as society slowly emerges from the worst of the COVID-19 pandemic, the deeply Christian insight that my wife’s quip served to highlight is even more pressing and ought to become a central element of the Christian contributions to the bioethical discourse about Christians’ moral responsibilities in a “post-pandemic” world. The point of this piece is therefore to discuss how Jesus’s words in the calling of Matthew (Matt 9:9-13) can help Christians learn from the tragedies of COVID to be able to build the kind of approach to health care in non-crisis times that will leave everyone better equipped to deal with the next health crisis whenever it emerges.
More precisely, the exposition builds on Jesus’s comments about “those who are well” to highlight the importance of turning from a health care system that prioritizes the treatment of diseases by physicians to an integrated public health infrastructure that promotes the well-being of all across the community. Then, this article shows how Jesus’s other reflections in this pericope, especially his closing admonition, “I desire mercy, not sacrifice,” can help Christians do the hard, collaborative work required to realize this shift. With these two parts, this article shows how to use biblical wisdom to learn the necessary lessons from the last pandemic if Christians want to avoid the next one.
No need of a physician: Expanding health care priorities
In the context of the calling of Matthew, Jesus’s comments about those who are well represent a straightforward response to the detractors who found his associations with tax collectors and sinners questionable. Scholars note that Jesus’s line has echoes of similar retorts found in the works of Greek authors that would have been circulating during his time. Diogenes, for instance, wrote that when the philosopher Antisthenes was challenged for consorting with “wicked men,” he retorted, “Physicians also live with those who are sick; and yet they do not catch fevers.”2 Jesus’s remark therefore builds on ancient parallels, but notably employs a different emphasis. Antisthenes’s comments, for instance, were meant to protect himself from the reputational damage that his critics believed he deserved as a result of his fraternization with less than savory characters. As the biblical scholar Daniel Harrington points out, in the Matthean text, “however, the focus is on those ‘who are sick’ (= tax collectors and sinners).”3 In contrast to the Greek parallels, then, Jesus’s justification called attention to the very people that led to his censure, focusing on their needs rather than his own social standing.
Christians ought to be similarly motivated by Jesus’s example, using his comments to ask how to be at the service of the needs of others instead of worrying about their own status. Of course, Jesus used the relationship between the well, the sick, and the physician as an analogy to make a statement about spiritual care and healing, but Christians can embrace the same spirit today in thinking about those who are literally the well and the sick. This way of thinking is particularly appropriate given that Matthew sandwiches this pericope between accounts of Jesus’s healing miracles, making it part of a broader segment (Matt 8–9) discussing the role of healing in Jesus’s ministry. Just as importantly, the narrative context of this segment reveals that its presentation of Jesus’s healing is not meant to be strictly descriptive but is also designed to serve a normative function, “provid[ing] a model for the disciples to follow in their own ministry.”4 Hence, although the immediate referent of Jesus’s words in the call of Matthew is spiritual sickness and health, the broader context in which these words appear indicates that the allusions to sickness and healing are not an accident, but a pertinent correlation that informs the interpretation of what it means for Jesus’s followers to imitate him in pursuit of a distinctive form of healing. Read in light of these connections, the calling of Matthew underscores the importance of attending to the marginalized in the ministry of healing, for this is the message Jesus sends by his example in this pericope and in his acts of healing more generally.5
The emphasis on the care for the marginalized that emerges from this story certainly supports the idea that Jesus’s followers have a duty to become the physician who cares for the sick, because this is a crucial way of providing healing and help to those in need. It also represents the most obvious application of Jesus’s reminder that the sick are the ones who need a physician. At the same time, the commitment to the marginalized invites an examination of how Christians can similarly be of service to those who are well, for it has become increasingly clear that the things that turn the well into the sick are often unequally distributed across the community and frequently affect the marginalized to a disproportionate extent. This disparity was clearly and painfully on display during the COVID pandemic, when the essential workers who were most likely to be exposed to the virus but who also had the least protection, at least during the early stages of the pandemic, were people who had some of the lowest paying jobs. They were also, not coincidentally, far more likely to be members of marginalized racial and ethnic groups in the United States.6 As a result of these conditions and other social trends, COVID death rates have been lower for White persons and higher for Latinos, Native Americans, and Black persons (when controlling for other demographic variables) throughout the pandemic.7
Significantly, the relationship between marginalization and worse health outcomes was not something COVID created; rather, it was a preexisting link that COVID simply exacerbated. Public health researchers have demonstrated that the “social determinants of health,” which include the non-medical dimensions of a person’s background (e.g., race and ethnicity, socioeconomic status, educational attainment, and the like), exert an extraordinary influence on an individual’s health status. Indeed, typical estimates attribute anywhere from one-third to more than one half of the root causes of health outcomes to the social determinants of health.8 By comparison, medical care is estimated to contribute only around 20% to health outcomes.9 Illustrating the link, a much-remarked 2016 research project found that something as seemingly innocuous as a zip code could change a newborn’s life expectancy by as much as 20 years.10 If Christians are indeed motivated by a concern for the marginalized and genuinely want the well to stay well, then Christian bioethics must do more to address the social determinants that first create marginalization and then harm health and well-being as a result.
The logical question is, of course, how to address these underlying social determinants of health. This issue is complex, eluding quick solutions and instead requiring a long-term commitment at the societal level. Jesus’s insistence that “those who are well have no need of a physician” can help with this challenge, though, by orientating action away from the traditional medical model, with its focus on treating individual patients after they have shown symptoms of a disease, toward a public health model, which prioritizes interventions that can promote health and well-being across entire populations before disease strikes.
Again, the COVID pandemic is instructive in appreciating the significance of this shift. Data show that variations in the strength of a country’s public health system were closely correlated with differences in COVID infection and death rates, such that nations with stronger public health systems had lower infection rates and fewer deaths while those with less investment in their public health infrastructure fared far worse. For instance, one comparative analysis of the United States and Cuba found that COVID death rates in the United States were more than 10 times higher than the rates in Cuba, not because the latter spent more money on health care but because Cuba already had a public health system oriented to population-level well-being in place before the pandemic and was thus able to implement more preventive measures earlier than the US’s decentralized health care system could manage.11 Another international analysis, meanwhile, found that the spread of COVID was linked more to the number of public health measures put in place than it was to the environmental factors that some experts initially thought might slow the spread of the disease.12 Perhaps most significantly, the comorbidities, such as diabetes and obesity, that left people most at-risk for serious consequences and even death from COVID are closely tied to one’s social context and socioeconomic status, pointing again to the importance of looking beyond the work of the physician to the issue of the social determinants of health.13
Using Jesus’s words to help see these trends, then one can quickly see the need to improve public health networks in the United States. Pursuing this outcome can start with a greater investment in primary care clinics so that more people, especially those on the margins, would be able to access preventive care and thus remain in better shape to weather any subsequent health emergencies that might arise. The success of such an endeavor, however, would require a dramatic shift in thinking. The prevailing mind-set in the United States views access to health care as an individual’s personal responsibility, leaving it subject to the vagaries of the marketplace and often contingent on an individual’s access to another major social determinant of health: gainful employment. Increasing access beyond the current ceiling will only happen if people accept that the task of broadening access is a collective responsibility. Happily, a few states have already started to adopt this perspective in the aftermath of the pandemic, so there is reason to hope that such a shift might actually be attainable.14
Notably, creating additional entry points into the existing health care system will only get the United States so far, because this strategy still treats prevention as part of the disease model that Jesus’s healing ministry rejects. A second part of the solution must therefore be an expansion of the public’s conception of what it means to support health and well-being. If those who are well truly do not need a physician, then one way in which Christians can encourage public health resilience and improve preparedness for the next crisis is by treating an investment in other public goods like education, safety, and even green space management as an investment in health and well-being. Implementing a change like this in the United States would be particularly powerful, for health policy scholars argue thatinadequate attention to and investment in services that address the broader determinants of health is the unnamed culprit behind why the United States spends so much on health care but continues to lag behind in health outcomes.15
By looking for ways to provide more social services to more people, communities could tackle this problem and reorient health care infrastructure toward the tools that meet basic needs. This change would have the effect of pulling people from the margins and, by extension, would help improve health care for all.
Christian communities should be spearheading this effort, not only because it represents a creative way of acknowledging Jesus’s assertion that those who are well need something other than a physician but also because it embodies the fundamental spirit at the heart of Jesus’s healing ministry. As Matthew’s healing narratives demonstrate, healing for Jesus was about far more than helping a sick individual get better physically, but instead encompassed a holistic transformation. While contemporary Christians should not expect to achieve the same degree of eschatological significance that Jesus’s healings entailed, they can nevertheless push the notion of healing beyond the narrow confines of the medical profession to include the activities that bring people back into community and sustain them all therein.16
As described above, implementing such an expansive view of healing will not be easy, because this task requires upending some well entrenched conventions in the realm of medicine. Fortunately, the calling of Matthew can do more than paint a prescriptive vision for what Christians should do to transform the reductionistic vision that tends to define healing in the contemporary social context. It can also provide some useful conceptual tools to facilitate the conversion of heart and transformation of mind-set that will be necessary to make this newfound commitment to public health a reality.
Mercy over sacrifice: Creating the foundation for change
The useful tools for conversion and transformation contained in the calling of Matthew can be found in a unique element that the evangelist added to Mark’s story about the calling of Levi (2:13-17), which appears to have been part of the source material for Matthew’s pericope. Whereas Mark’s account ends immediately after Jesus’s comments about how those who are well do not need a physician, Matthew includes a “substantive” update with the introduction of a quotation from Hos 6:6.17 The message is simple: “I desire mercy, not sacrifice.” The implications, however, are profound and provide a twofold rationale for a reinvigorated commitment to solidarity that can sustain the pursuit of a broader vision of health care.
The first way in which this exhortation to mercy can become a resource for embracing solidarity is in its juxtaposition with sacrifice. As a response to the critical challenge of the Pharisees, Jesus is both telling his disciples what to do (i.e., show mercy) and what not to do (i.e., impose demands on others that would require them to make sacrifices for the sake of demonstrating their dedication). The line is thus an invitation to adopt a compassionate disposition to others that allows followers to expand their circle of “fellowship.”18 This ability is essential if society is ever going to realize a shift toward public health, because the only way people can change their perspective to account for the well-being of the whole population is if they readily accept that everyone in that population is as worthy of their concern as they consider themselves to be. By calling his followers to compassion and turning them away from sacrifice, Jesus is reminding his followers that people do not have to prove their worth through a series of tests that establish their commitment. Instead, they deserve to have their worth honored because it belongs to them as children of God.
The second way that Jesus’s citation of Hosea can deepen a commitment to solidarity is related to the first. In part because the citation ends with Jesus’s addendum that “I have come to call not the righteous but sinners,” this passage serves as a stark reminder of how all human beings stand as equals before the eyes of God. To the extent that readers can see themselves as those who are called by Jesus, they are left with the logical conclusion that they are not “the righteous” but sinners, and this humble self-identity can make it easier to shed the Pelagian notion that some followers are better than others because they have done more work to merit a closer relationship with God. Pope Francis exemplifies the orientation to equality that this realization of one’s sinfulness can empower. When he became pope in 2013, Francis chose as his pontifical motto a line from Saint Bede’s (d. 735) description of the calling of Matthew, stressing the mercy Jesus showed to Matthew when he called the tax collector to follow him. Reflecting on this choice, Francis has remarked that it helps him remember who he is at his core, namely, “a sinner whom the Lord has looked upon.”19 Jesus’s summons to mercy, not sacrifice, encourages everyone to see themselves the same way, allowing people to set aside any haughtiness that might separate them from God and neighbor and to replace it with a sense of equality in imperfection. Upon adopting this perspective, humankind will find it easier to look upon others with a spirit of mercy and compassion, for people can recognize their own need for God to look upon them in this way. The natural result of this process is the profound awareness of solidarity that can emerge when a group of people finally recognizes that they are all in this together.
The call to solidarity is essential if the goal of comprehensive care for the marginalized is to be realized, because the changes needed to ensure everyone has a chance to be well and not need a physician will not be easy to implement. In fact, they will likely be costly, particularly for the people who have access to the health care system now or who find that the social determinants of health currently work in their favor. For example, the expanded access to primary care described above might translate into fewer opportunities for specialized treatments as the health care system tries to rebalance its resources to meet shifting demand. Likewise, the new investments in education, transportation, parks, and more that are required to address disparities in the social determinants of health will not be cheap if they are going to be effective. Consequently, these initiatives will impose a cost on communities and individuals, and this cost will, by necessity, almost certainly fall more heavily on the shoulders of those who have the most resources now. If those individuals think only about a narrowly defined self-interest, they will likely balk at this expectation, creating a profound obstacle for the success of this new pursuit of holistic healing. When they start from a place of solidarity, however, and operate with “a firm and persevering determination to commit oneself to the common good,” they are more apt to see the importance of making these changes, even if it threatens their financial standing.20 The reminder of mercy and the spotlight on one’s sinfulness found in the calling of Matthew are thus both essential pieces in the larger effort to use Jesus’s words about those who are well to pursue a holistic form of healing grounded in public health.
Conclusion
In reflecting on Christian responsibilities for the post-pandemic world that is best described in eschatological terms (i.e., already and not yet), Christians would do well to consider what the calling of Matthew, that peculiar pericope squeezed among healing narratives, has to teach about the best way to enact Jesus’s holistic healing ministry now. This consideration should pay particular attention to the idea that those who are well do not need a physician, because this idea can serve as a reminder that other needs, even those still related to healing, can and should be addressed by something more than the practice of medicine alone. Those attuned to this reality will ultimately be able to learn the valuable lessons COVID revealed about the importance of public health and the necessity of working together to achieve healthy communities and not just isolated pockets with some healthy people. When also using the calling of Matthew to recognize the equality of all before God and neighbor, Christians will be better positioned to put these lessons into practice, thereby contributing to health, healing, and wholeness as the pandemic subsides and society searches for a better way of living together in non-crisis times.
Author biography
Conor M. Kelly is Associate Professor in the Department of Theology at Marquette University. He regularly teaches a medical ethics course for undergraduates and has written about health care ethics at the end of life and in everyday contexts. He is the author of The Fullness of Free Time: A Theological Account of Leisure and Recreation in the Moral Life (Georgetown University Press, 2020) and Racism and Structural Sin: Confronting Injustice with the Eyes of Faith (Liturgical Press, forthcoming).
1. John Dominic Crossan, The Birth of Christianity: Discovering What Happened in the Years Immediately after the Execution of Jesus (San Francisco: Harper Collins, 1998), 293–96.
2. Diogenes Laërtius, The Lives and Opinions of Eminent Philosophers, trans. C. D. Yonge (London: George Bell and Sons, 1901), 219.
3. Daniel J. Harrington, The Gospel of Matthew (Collegeville: Michael Glazier, 1991), 126.
4. Walter T. Wilson, Healing in the Gospel of Matthew: Reflections on Method and Ministry (Minneapolis: Fortress, 2014), 15.
5. Francois P. Viljoen, “Hosea 6:6 and Identity Formation in Matthew,” Acta Theologica 34.1 (2014): 214–37 (232).
6. Bryan N. Massingale, “The Assumptions of White Privilege and What We Can Do about It,” National Catholic Reporter, 1 June 2020, www.ncronline.org/news/opinion/assumptions-white-privilege-and-what-we-can-do-about-it.
7. See Latoya Hill and Samantha Artiga, “COVID-19 Cases and Deaths by Race/Ethnicity: Current Data and Over Time,” Keiser Family Foundation, 22 February 2022, www.kff.org/coronavirus-covid-19/issue-brief/covid-19-cases-and-deaths-by-race-ethnicity-current-data-and-changes-over-time/.
8. “Social Determinants of Health,” World Health Organization, www.who.int/health-topics/social-determinants-of-health#tab=tab_1.
9. Carlyn M. Hood, Keith P. Gennuso, Geoggrey R. Swain, Bridget B. Catlin, “County Health Rankings: Relationships Between Determinant Factors and Health Outcomes,” American Journal of Preventative Medicine 50.2 (February 2016): 129–35.
10. Todd Bookman, “A Few Miles and a World of Difference as Study Finds Wide Disparity in Philly Life Expectancies,” WHYY, 7 April 2016, https://whyy.org/articles/a-few-miles-and-a-world-of-difference-as-study-finds-wide-disparity-in-philly-life-expectancies/.
11. Mary Anne Powell, Paul C. Erwin, and Pedro Mas Bermejo, “Comparing the COVID-19 Responses in Cuba and the United States,” American Journal of Public Health 111.12 (December 2021): 2191–92.
12. Peter Jüni, Maritna Rothenbühler, Pavlos Bobos, Kevin E. Thorpe, Bruno R. da Costa, David N. Fisman, Arthur S. Slutsky, Dionne Gesink, “Implication of Climate and Public Health Interventions on the COVID-19 Pandemic: A Prospective Cohort Study,” Canadian Medical Association Journal 192.21 (May 25, 2020): E566–E573.
13. On the impact of the social determinants of health on obesity, see Nicholas A. Christakis and James H. Fowler, “The Spread of Obesity in a Large Social Network over 32 Years,” New England Journal of Medicine 357.4 (July 26, 2007): 370–79 (377–78).
14. Noah Weiland, “In This Michigan County, Pandemic Stimulus Funds are Remaking Public Health Programs,” The New York Times (April 9, 2022), www.nytimes.com/2022/04/09/us/politics/michigan-pandemic-stimulus.html; Lisa Rab, “The Medical Crisis that Finally Convinced Republicans in North Carolina to Expand Medicaid,” Politico, 14 August 2022, www.politico.com/news/magazine/2022/08/14/new-moms-convinced-republicans-to-expand-medicaid-00049534.
15. Elizabeth Bradley and Lauren Taylor, The American Healthcare Paradox: Why Spending More is Getting Us Less (New York: Public Affairs, 2013), 3.
16. On the eschatological impact of Jesus’s healing ministry in Matthew, see Wilson, Healing in the Gospel of Matthew, 156–57.
17. Harrington, Gospel of Matthew, 127.
18. Viljoen, “Hosea 6:6 and Identity Formation,” 222–23.
19. Antonio Spadaro, “A Big Heart Open to God: The Exclusive Interview with Pope Francis,” America: The Jesuit Review, 30 September 2013, 15–38 (16).
20. John Paul II, Sollicitudo Rei Socialis, 30 December 1987, 38, www.vatican.va/content/john-paul-ii/en/encyclicals/documents/hf_jp-ii_enc_30121987_sollicitudo-rei-socialis.html.
| 0 | PMC9742349 | NO-CC CODE | 2022-12-14 23:45:33 | no | 2022 Dec 9;:00346373221132281 | utf-8 | null | null | null | oa_other |
==== Front
RAE
sprae
Review & Expositor
0034-6373
2052-9449
SAGE Publications Sage UK: London, England
10.1177/00346373221129647
10.1177_00346373221129647
Original Research Article
Practicing borderless Christianity: Challenges and opportunities of the Covid-19 pandemic
Oladipo Caleb O.
Campbell Divinity School, USA
Caleb O. Oladipo, Campbell Divinity School, Buies Creek, NC 27506-0567, USA. Email: [email protected]
9 12 2022
9 12 2022
00346373221129647© The Author(s) 2022
2022
Review & Expositor, Inc
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
The challenges and opportunities presented by Covid-19 are enormous, and Christians and non-Christians could take advantage of the pandemic to craft a borderless faith tradition. The Covid-19 pandemic has provided the opportunity to have a more comprehensive and positive image of every faith tradition, creating the best of all possible worlds for future generations.
borderless
borders
Covid-19
pandemic
transnationalism
edited-statecorrected-proof
typesetterts1
==== Body
pmcAll borders are artificial, and the national borders of our world are not only geographical, but also ideological. Borders are most often erected out of fear and mistrust. But they also inform and transform human identities. Rigid national borders are monotonous, and they have diminished cooperation among faith groups and have suffocated the life of the Church.
The occurrence of the Covid-19 pandemic colonized the world in an instant, exacerbating geographical boundaries and shedding new lights on national and international inequalities. Proposing a borderless world, this article concludes that the hope of realizing a peaceful world depends on collective efforts to redefine what kind of world people want, efforts which will require both determination and faith.
Introduction
In One Earth Many Religions, Paul F. Knitter wrote, at the turn of the century,As humanity steps into a new millennium, the religious traditions of the world find themselves at a turning point. Up until now . . . religious communities have understood themselves from within the circle of their own experiences and traditions; as this century slips into the next, they are being challenged to expand their ways of knowing who they are by allowing their circles to touch and overlap with others.1
He went on to say that “the nature of our intercommunicative world and of the crises this world faces offer and require such a dialogical, correlational manner of religious self-understanding.”2
The Church has entered a new era of self-awareness and introspection, and Christians now practice their faith in ways they have not done for over two thousand years. A Christian can travel to Lagos, Nigeria and attend his Sunday School class in Cary, North Carolina. A Christian in Buenos Aires, Argentina could worship in her church while on a business trip to London, England. A US Sunday School teacher could teach a class on Zoom while visiting relatives in Cape Town, South Africa.
The world is once again characterized by borderless-ness as it has always been naturally. Perhaps the age of technology is one of the greatest gifts to the modern church. In a way, Covid-19 has required Christians to function and become more creative due to the old tradition of “necessity being the mother of invention.” Christians in the contemporary world have become more creative in the production and dissemination of Christian knowledge. Perhaps in this moment of vulnerability one can say that the Covid-19 pandemic has been a valuable gift to the Church of the modern age.
In spite of technological advances, however, most traditional Christians still desire the rich fellowship of being face-to-face in congregational settings. Many continue to long for a return to the traditional mode of in-person worship because human beings are social and gregarious beings.
Throughout 2020 and into 2021, human beings have been colonized by the Covid-19 pandemic, presenting new challenges to national borders in every country.
The challenges and opportunities of Covid-19
The Covid-19 pandemic had shed new light on myriad human disparities worldwide. The Western world, for example, hoards and wastes vaccines or becomes stingy about making it available to developing and low-income nations. The vaccine-sharing pledges from Western nations to the nations in the Global South have been too slow and often made available when the vaccines are close to expiration dates. Approximately 70% of people in the European Union were vaccinated by the end of 2021; the United States vaccinated approximately 60% of its citizens by the end of the year. Fewer than 2% of the seven billion people who populate the world, however, are vaccinated, even as scientists and health care providers in the Western world warn that vaccinating 80% of the world is the necessary solution to slowing down the spread of the pandemic. The shortfall of assistance from Western nations is glaring, with the consequences of nationalism continuing.
Fatima Hassan, a human rights lawyer from South Africa and the founder and director of the Health Justice Initiative, maintains that “vaccine hesitancy hurts the global Covid-19 response.”3 Nothing was as discouraging about the disparity of the world magnified by Covid-19, however, as the protectionism of the Western nations. This reality became clear when the omicron variant of the virus began to surface in different parts of the world, especially in developing countries. The omicron variant first surfaced in The Netherlands 2 weeks before it was detected in Botswana, a country next to South Africa. When scientists in South Africa detected it among their citizens, the world took immediate and discriminatory action to lock out Southern Africa. Noubar Afeyan, the chairman and co-founder and chief executive officer of Moderna and the founder of Flagship Pioneering, stated that vaccines are still effective against omicron and urged everyone to keep up with vaccination because “we are going to get more variants.”4
The more immediate negative impact of Covid-19, however, has been how it has affected Christians worldwide. The pandemic response of Christians has been impacted by the fact that churches were closed and many clergy and church personnel were dispirited. Those who depended on the resources from parishioners and church members became anxious. Some took the drastic measures of promoting vaccine hesitancy to the extreme, preaching against mandatory and government impositions related to vaccinations. In December 2021, Covid-19 claimed the life of a popular televangelist, Marcus Lamb, who saw the promotion of vaccine protection against the virus as spiritual warfare in the Church. Many of Lamb’s followers portrayed his rejection of vaccination and his skepticism as a good fight, and his death did not curb vaccine hesitancy among them, many turning deaf ears to the call from scientists to get vaccinated.
That the world has been misled on many fronts is incontrovertibly true, with the death of such vaccine skeptics surprisingly not proving to be a wake-up call to promote vaccination. Often, vaccine-skeptics have been seen as heroes, putting up a good fight, like Lamb, against efforts to manage the disease.
As the battle to curb the Covid-19 pandemic continues, Christians often persist in making the mistake of flatly dismissing as unwise those who argue against vaccination. Critics fail to consider that some, from various parts of the world, sincerely believe that their perceptions of God’s revelation are more believable than scientific facts. Science, they claim, does not prove, but probes.
And now, the “omicron variant” of Covid-19 has entered as a new term in our lexicon of health consciousness and vocabulary. The omicron variant has been unpredictable, a fact that should not be a shock to anyone. The greatest problem the world faces in the ongoing battle is the disparity in vaccine distribution. The co-chair of the Africa’s Union Vaccine Distribution Alliance, Dr. Ayoade Alakija, pulls no punches in her November 2021 BBC interview.5 In her assessment, the disparity in the world system is catastrophic and has affected Africans disproportionately. She asserted that if, in 2019, the Covid-19 pandemic had first been detected in Africa instead of Wuhan, China, the continent of Africa would have been known to the world as the “continent of Covid,” with the world locking the continent away and throwing away the key. Alakija believes that no resulting sense of urgency would have arisen to develop a vaccine or make it available to the African population of 1.3 billion people. Her contention is that until everyone is vaccinated, no one is safe. She also challenged African leaders to be more assertive about the needs of their people and the discriminatory practices they suffer from the hands of more powerful leaders in the Western world. African political leaders need to share their own stories, because until the lion learns to write, the stories of the hunt will always glorify the hunter.
Today, as churches have become a bit more creative in how to worship technologically and what it means to congregate, Christians face a new world of global fluidity and now confront a world of transnationalism with the exciting challenges to dream new dreams that take seriously the Christian faith as a borderless faith tradition.
The coronavirus pandemic has been a gift to the Church in another sense. Christians have become more aware of their vulnerability to the rigidity and monotony of their own traditions. Christians have realized that traditional face-to-face fellowship is not necessarily the defining mode of true worship and that one of the characteristics of true Christian fellowship is flexibility. Traditions are, therefore, fallible, and Christians are not self-sufficient. The Covid-19 pandemic exposed this vulnerability and the insufficiency of traditionalism.6
Questions about borders in the era of the Covid-19 pandemic
The Covid-19 pandemic has also vividly demonstrated that no restrictive immigration policies can stop human beings in their justifiable determination to cross geographical borders. Bill Ashcroft, an expert on nationalism and geopolitical migration at the University of New South Wales, Australia, observes, “Nationalism is always complicated by ethnocentrism, racism, and populism, and as nationalism proliferates, violence increases. But at the same time, post nationalism and No Borders politics . . . are growing apace.”7 This observation is especially pertinent when one considers the borderless-ness and the nature of the history of Christian expansion through migration.
In 2021, Eerdmans published a book by Jehu J. Hanciles, an African scholar and the D. W. and Ruth Brooks Professor of World Christianity at Candler School of Theology at Emory University, titled Migration and the Making of Global Christianity. The book is not only a comprehensive chronicle of human nature as moving from one place to another throughout history, but it is also a detailed account of migration’s responsibility for making Christianity a global faith tradition. Hanciles states, “It requires no leap of imagination to grasp that human migration played a huge role in the spread of Christianity among the tribes and peoples well beyond the Roman Empire.”8 Throughout Christianity’s history, ordinary people such as merchants, soldiers, government agents, slaves, pilgrims, tourists, entertainers, artisans, sick and health care providers, as well as priests and planned missionaries, were responsible for the effective spread of the Christian faith as they migrated. Hanciles leaves no stone unturned in his argument, restating it again in chapter 7:To restate my central thesis, human migration has played an indispensable role in the cross-cultural spread of the Christian faith principally because migrants who are Christians inevitably fulfill a missionary function in their encounters with non-Christian peoples and societies.9
That Covid-19 has proved a wake-up call to the nature of humanity and the Christian faith is also an incontrovertible fact. The pandemic has shed new light on the artificiality of geographical borders. Not only human beings, but also other creatures migrate. That birds migrate is common knowledge, and seasonal migration is common among other animals. Those of the Christian faith worldwide are now asking many questions they were not used to asking in the past. Foremost among these questions are what the world would look like without its artificial borders. Corresponding to the “borderless-ness” of the Christian trajectory from the time of Jesus, what would the Christian world look like today without the geographical borders imposed by political nationalism?
The world has many kinds of borders, some more porous than others, but technologically the world has no national borders enforceable in the modern era. Politically speaking, national identities exist ideologically and with rigid borders still characterizing modern nations. In economic terms, there is only one world, fragile and fragmented in religious and governmental terms. In our newly globalized world, are attempts to enforce and protect national borders evil? Can rigid borders endure, considering the technological advancements that promote the porosity of geographical boundaries? Can human beings truly be free when classifications, such as rich and poor, democratic and non-democratic, Christian and non-Christian, remain?
Some reasons to consider for physical borders
Geographical borders have always existed out of the desire for domination by powerful and more prosperous or high-income nations. In the past, despite their power and prosperity, nations have not conquered other territories without fear of revenge from the powerless. Fear and uncertainties grasp and control people in the modern world because of bordering and nationalism.10 At least 77 border walls or fences exist in the world today, and most of them were erected since 2001.11 Walls are indicative of hostility, and building them is predictable. The erection of borders, however, is not consistent with the nature of Christianity, a borderless missionary religion.12 Other missionary religions include Islam and Buddhism, but no other religion is identified with missionary zeal and intensity more so than the Christian faith. Scholars have argued that the Bible itself is the blueprint of God’s sending in mission or Missio Dei in the world.13 Christians build bridges; they do not build walls.14
What the Covid-19 pandemic has demonstrated to Christians in the modern era is the increasing mobility of Christians worldwide. Navigating one aspect of this mobility, the Church is trying to embrace the modern world of technological borderless-ness without becoming lost in it. Building bridges is unpredictable and life-giving, evoking silent gratitude from its beneficiaries. The Covid-19 pandemic has issued the call to Christians to become Christians and followers of Jesus without religion, to see Christianity as a tool for the transformation of a borderless world without an overriding compulsion to convert those who are not Christians to their own faith traditions. The vision of a borderless world where the Christian faith could flourish as an active movement originated with Jesus of Nazareth when he was with his disciples.
Transnationalism and the borderless vision of Jesus of Nazareth
The Jesus of Nazareth that Christians encounter in the Gospels exemplified a transnational and borderless vision for the world under God. He constantly challenged religious leaders and his followers to cross social and religious borders. He saw only one world and one God, whom he called his Father. Jesus of Nazareth was open to “official outsiders” such as the Gentile centurion and Samaritans,15 praising them as examples of virtue.16
One of the most astounding claims of Jesus in the gospels is his transnationalism and borderless vision of God’s Kingdom. He took Gentiles seriously and was open to their unique contributions to the spiritual directions of the Kingdom of God that he inaugurated. Jesus established with his disciples a vision of the Kingdom of God as one that is borderless.
In Mark’s Gospel, John said to Jesus, “Teacher, we saw someone casting out demons in your name, and we tried to stop him, because he was not following us.” But Jesus said, “Do not stop him; for no one who does a deed of power in my name will be able soon afterward to speak evil of me. Whoever is not against us is for us” (9:38-40, NRSV). In the Gospel of John, Jesus said, “I have other sheep that do not belong to this fold. I must bring them also, and they will listen to my voice. So, there will be one flock, one shepherd” (10:16, NRSV). Thus, anyone who genuinely cares for the wellbeing of their fellow-human beings can also be valuable in the Kingdom of God. The contemporary world is interconnected. Our grueling experience with the Covid-19 pandemic has taught us that we will always need each other.
The challenges of bridging ideological gaps
True Christianity in its essence is not only a borderless faith, but also a faith that bridges the gap between God and humanity. One major difference in the interpretation of the relationship between Christianity and Hellenistic philosophy, for example, is the mediating principle or the bridge between God and the world. While Greek philosophers saw the Logos or Word (sometimes interpreted as reason or principle) as the necessary bridge because God (necessary being) cannot deal directly with the world that is contingent, early Christians viewed the Logos as necessary because of human alienation. The important thing to consider is how the Hellenistic philosophical tradition influenced the Christian faith and vice versa. The Greek concept of the Logos and the first chapter of the Gospel of John in the Greek New Testament is clear. Although it did not do so uncritically, Christianity subsequently assumed the Hellenistic vocabulary to justify and strengthen its claims. The early Christian Church fathers, especially in the traditions of Justin Martyr, learned that, as they came to grips with the Christian faith and presented it to their pluralistic world, they also had to make it more credible by interpreting it with the patterns of thought prevalent in their societies. Their societies and the religions associated with them varied from the Gospel the early Church fathers sought to proclaim. The strength of a bridging-of-the-gap approach was in accepting and taking seriously the differences between Hellenism and Christianity, affirming the unique contributions that each makes in ordering human lives and their spiritual destiny.17 Knitter states it this way: “Other religions are not only genuinely different, [but] they can also be genuinely valuable.”18 A careful relationship between one religion and another necessarily involves a careful inhaling and exhaling. Perhaps perilous for Christians is to cut off devotees of other religions in the age of a universal health crisis. The current era is one of interconnectedness and interdependency, and devotees of other religions will continue to need Christians as Christians need them.
Conclusion
It is not hyperbole to suggest that no other phenomenon in the modern era has demonstrated the interconnectedness of humanity more so than the Covid-19 pandemic. The pandemic has revealed human vulnerability and highlighted the fact that being human has no cure. It has blatantly exposed the weaknesses, as well as the strengths, of our humanity. Our natural rapacity and avarice, both personal and institutional, have been put under a magnifying glass. Our human nature has had no place to hide. The proposal in this article to dissolve the rigidity of borders can be realized only when humankind abandons and discards distinctions between rich and poor, black and white, Christians and non-Christians, East and West, and other divisions. Rigid borders are untrue and antithetical to the gospel of faith, hope, and love.
Author biography
Caleb O. Oladipo holds the inaugural Snellings Chair of Christian Evangelism and Mission at Campbell University Divinity School where he also serves as the Director of the Braswell World Religions and Global Cultures Center.
1. Paul F. Knitter, One Earth Many Religions: Multifaith Dialogue & Global Responsibility (Maryknoll NY: Orbis, 1995), 22.
2. Knitter, One Earth Many Religions, 22.
3. Fatima Hassan, “On GPS: A South African Perspective on Omicron,” interview by Fareed Zakaria, GPS, CNN, 5 December 2021, https://www.cnn.com/videos/tv/2021/12/05/exp-gps-1205-fatima-hassan-south-africa-omicron.cnn.
4. Noubar Afeyan, “On GPS: Moderna CEO on vaccines versus Omicron,” interview by Fareed Zakaria, GPS, CNN, 5 December 2021, https://www.cnn.com/videos/tv/2021/12/05/exp-gps-1205-noubar-afeyan-moderna-omicron.cnn.
5. Ayoade Alakija, interview by Phillipa Thomas, BBC World News, BBC, 27 November 2021, via Twitter, https://twitter.com/yodifiji/status/1464609837875085317?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1464609837875085317%7Ctwgr%5Eef8210f993b39697d399e57691766a85e8e20eb3%7Ctwcon%5Es1_&ref_url=https%3A%2F%2Fwww.timeslive.co.za%2Fnews%2Fworld%2F2021-11-29-watch–dr-ayoade-olatunbosun-alakija-praised-for-this-passionate-clap-back-at-travel-bans%2F.
6. See Leander E. Keck, The Church Confident: Christianity Can Repent, but It Must Not Whimper (Nashville: Abingdon, 1993), 45. Keck quotes Jaroslav Pelikan, saying, “Tradition is the living faith of the dead; traditionalism is the dead faith of the living.”
7. Bill Ashcroft, “Borders, Bordering, and the Transnation,” The English Academy Review: A Journal of English Studies 36.1 (May 2019): 6.
8. Jehu J. Hanciles, Migration and the Making of Global Christianity (Grand Rapids: Eerdmans, 2021), 189.
9. Hanciles, Migration, 269.
10. See Ashcroft, “Borders, Bordering, and the Transnation,” 5.
11. See Ashcroft, “Borders, Bordering, and the Transnation,” 6.
12. See Dyron B. Daughrity, The Changing World of Christianity: The Global History of a Borderless Religion (New York: Peter Lang, 2010). In the book, Daughrity argues that the Christian faith is the only faith tradition that has become truly globalized.
13. See Christopher J. H. Wright, The Mission of God: Unlocking the Bible’s Grand Narrative (Downers Grove IL: InterVarsity Press, 2006), in which the author argues that the Bible represents the activities of God in the world and chronicles God’s Mission for the transformation of the world. In other words, the writings that now comprise the Bible are themselves the product of and witness to the ultimate mission of God. The Bible renders to humanity the story of God’s mission through God’s people in their engagement with God’s world and for the sake of the whole of God’s creation. The Bible is the drama of this God of purpose engaged in the mission of achieving that purpose universally, embracing past, present, and future. Its center focus, climax, and completion come through Jesus Christ.
14. The statement that “Christians build bridges, not walls,” has been attributed to Pope Francis. In his statement, the Pope concluded that one of the challenges of the Church remains “to defend and preserve the dignity of your fellow citizens in the tireless and demanding pursuit of the common good.” “Address of the Holy Father,” US Capitol, Washington, DC, 24 September 2015, https://www.vatican.va/content/francesco/en/speeches/2015/september/documents/papa-francesco_20150924_usa-us-congress.html. See also John Carr, “The Role of Catholics in the time of Trump,” America: The Jesuit Review, 5–12 December 2016, https://www.americamagazine.org/politics-society/2016/11/14/role-catholics-time-trump.
15. See Luke 7: 1-10; 9:52.
16. See Luke 10:10-37; 17:11-19.
17. For a short but comprehensive treatment of the complementarity between Hellenistic philosophy and the Christian faith, see Caleb O. Oladipo, “Philosophy” in Encyclopedia of Mission and Missionaries, ed. Jonathan J. Bonk (New York: Routledge, 2007), 335–37.
18. Knitter, One Earth Many Religions, 32.
| 0 | PMC9742350 | NO-CC CODE | 2022-12-14 23:45:33 | no | 2022 Dec 9;:00346373221129647 | utf-8 | null | null | null | oa_other |
==== Front
Perfusion
Perfusion
spprf
PRF
Perfusion
0267-6591
1477-111X
SAGE Publications Sage UK: London, England
36484202
10.1177_02676591221144729
10.1177/02676591221144729
Case Report
Peripartum veno-venous extracorporeal membrane oxygenation in patients with severe CoViD-19-related-ARDS
https://orcid.org/0000-0001-6353-3283
Kakar Vivek 1
Ahmed Ihab 2
Ahmed Walid 2
https://orcid.org/0000-0001-9931-8905
Raposo Nuno 3
Kumar G Praveen 2
1 Cardiac Critical Care and ECMO, Critical Care Institute, 284697 Cleveland Clinic Abu Dhabi , Abu Dhabi, United Arab Emirates
2 Critical Care Institute, 284697 Cleveland Clinic Abu Dhabi , Abu Dhabi, United Arab Emirates
3 Vascular, and Thoracic Institute, Heart and Vascular Institute, 284697 Cleveland Clinic Abu Dhabi , Abu Dhabi, United Arab Emirates
Vivek Kakar, Critical Care Institute, Cleveland Clinic Abu Dhabi, Hamouda Bin Ali Al Dhaheri Street, Al Maryah Island, Abu Dhabi 112412, United Arab Emirates. Email: [email protected]
9 12 2022
9 12 2022
02676591221144729© The Author(s) 2022
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
We describe a case series of five pregnant or postpartum women with severe CoViD-19-related ARDS requiring VV ECMO at our centre between Jan 1 and Sep 30, 2021. All patients were cannulated at the referring hospitals by our team before transferring to our centre. None of the women were vaccinated against CoViD-19. All had severe ARDS with Murray’s Lung Injury Score of 3–4 and met the severity threshold for ECMO initiation that was used in the EOLIA study. All patients were discharged alive to home, acute rehabilitation, or lung transplant centre. One patient suffered intrauterine death before ECMO initiation and another while on ECMO. VV ECMO for refractory CoViD-19 related ARDS in the peripartum period is safe, and in this small series, it was associated with good maternal survival rates.
extracorporeal membrane oxygenation
severe acute respiratory syndrome coronavirus 2
acute respiratory distress syndrome
COVID-19
pregnancy
edited-statecorrected-proof
typesetterts10
==== Body
pmcIntroduction
Pregnant women infected with SARS-CoV-2 [severe acute respiratory syndrome coronavirus 2] are at significantly greater risk of needing critical care admission, invasive ventilation, extracorporeal membrane oxygenation [ECMO], pregnancy associated complications, and as a result, they have higher mortality.1,2 Maternal age ≥ 35-years, Body Mass Index [BMI] ≥ 30 kg/m2, socioeconomic deprivation, and pre-existing diabetes mellitus or hypertension have been recognized as risk factors for severe CoViD-19.3 The incidence of pre-term and stillbirth rates is significantly higher in pregnant women with CoViD-19 infection.3 Several, large case series have shown that patients receiving venovenous ECMO [VV ECMO] for CoViD-19 related severe acute respiratory distress syndrome [ARDS] have similar overall survival rates as those receiving VV ECMO for non-CoViD ARDS.4–6 Systematic reviews of peri-partum ECMO in non-CoViD ARDS have suggested excellent maternal and foetal outcomes with a low complication rate.7,8 However, no substantial data currently exists for ECMO use in pregnant women infected with CoViD-19. We would like to briefly share our experience of managing five peripartum women who received VV ECMO for refractory CoViD-19-related ARDS at a tertiary centre in the United Arab Emirates [UAE].
Case report
Twenty patients were placed on VV ECMO for CoViD-19-related ARDS between Jan 1 and Sep 31, 2021. Five patients (25%) were women who were either still pregnant (n = 1) or immediately post-partum (n = 4) at the time of ECMO initiation. Our Institutional Review Board granted a waiver as the case series involved less than 10 patients. However, a written informed consent was signed by all patients discussed in this report. All patients had severe ARDS with a Lung Injury Score of 3–4 and fulfilled the EOLIA [ECMO to Rescue Lung Injury in ARDS] trial criteria for severity9 despite optimal medical management that included neuromuscular paralysis, high positive end-expiratory pressure (PEEP) trials, restrictive fluid management, and prone positioning. All patients were cannulated by our ECMO team at the referring hospitals prior to transfer to our centre. The ECMO retrieval team comprised of 2 critical care physicians, a critical care nurse, a surgical nurse, and a perfusionist. Three out of five patients received femoral-jugular cannulation, our preferred configuration. One patient was referred late from a centre outside of the UAE after being on the ventilator for more than 2-weeks with evidence of advanced lung disease on chest imaging. This patient was accepted for VV ECMO as a bridge-to-transplant candidate based on informal discussion with the transplant team who insisted on a jugular bicaval [dual-lumen] cannulation to facilitate physical rehabilitation even though our centre has considerable experience in rehabilitating ECMO patients with femoral cannulation. Another patient received femoro-femoral cannulation initially. All patients underwent repeat CT scans of their head and chest after ECMO initiation, including a CT Pulmonary Angiogram. The CT Chest images for the patients at ECMO initiation are shown in Figure 1. None of the patients in our cohort had pulmonary embolism. All patients underwent transthoracic echocardiogram at the bedside on a regular basis to monitor the right ventricular function, pulmonary artery pressures, assess the volume status, and confirm access cannula position. One patient needed multiple transoesophageal echocardiograms [patient 1], to reposition the jugular bicaval cannula.Figure 1. CT chest at ECMO initiation [labels correspond to the patients].
None of the patients were vaccinated against SARS-CoV-2. The ability to sequence the CoViD variants was not available in house. Although the samples were sent out for sequencing, sadly, due to the lack of availability of testing kits at the height of pandemic, we only received results for three patients [patients 2, 3, and 4], all of whom were infected with the Delta variant [B.1.617.2]
Patient characteristics and an overview of the clinical management prior to and on ECMO initiation including complications is presented in Tables 1 and 2. We used the Maquet CardioHelp® System [Getinge Group] device along with the Maquet HLS Set Advanced 7.0® which incorporates a centrifugal pump into the oxygenator. Patients received therapeutic anticoagulation with an infusion of unfractionated Heparin, targeting an activated thromboplastin time [apTT] of 40–60s. Once patients were placed on VV ECMO, we switched to ultraprotective lung ventilation using the airway pressure-release ventilation [APRV] mode. Four out of five patients in this cohort had tidal volumes under 2 mL/kg for the first 96-h. The resulting hypercapnoea was managed exclusively by adjusting the sweep gas flow rates on ECMO. We targeted peripheral arterial saturations [SpO2] greater than 88%, partial pressure of oxygen in arterial blood [paO2] greater than 7 kPa, and a partial pressure of carbon dioxide in the arterial blood ≤ 8 kPa, provided that the arterial pH was ≥ 7.25.Table 1. Clinical management overview.
Patient 1 Patient 2 Patient 3 Patient 4 Patient 5
Demographic
Age [years] 35 22 30 37 41
Ethnicity Emirate Emirate Jordanian Emirate Emirate
BMI, Kg/m2 27 19.15 30 39 31.6
Comorbidities Sleeve gastrectomy Hodgkin’s lymphoma [on chemotherapy] None Gestational diabetes Obesity
Hypothyroidism
SLE
Mild asthma
Obesity
Pregnancy at CoViD infection 2nd pregnancy 33-weeks 1st pregnancy 21-weeks 2nd pregnancy 27-weeks 5th pregnancy 29-weeks 1st pregnancy 34-weeks
CoViD vaccine No No No No No
CoViD therapies Steroids Steroids Steroids Steroids Steroids
Remdesivir Remdesivir Remdesivir Remdesivir Remdesivir
Tocilizumab Tocilizumab
Pre-ECMO ARDS management
Pre-intubation support HFNC HFNC HFNC HFNC HFNC
NIV NIV NIV Intubation NIV
Intubation Intubation Intubation Intubation
Symptoms to intubation 14 13 4 11 6
Paralysis Yes Yes Yes Yes Yes
Proning No Yes, 120-h Yes Yes Yes
Inhaled NO Yes Yes No No No
Barotrauma Pneumothorax No No Pneumomediastinum, subcutaneous emphysema Pneumothorax
Variables at ECMO initiation
Aim of ECMO Bridge to transplant Bridge to recovery
PEEP, cmH2O 16 14 16 8 14
fiO2, % 100 100 100 100 100
PaO2/fiO2 [mmHg) 78 40 79 69 65
Intubation to ECMO, days 21 11 5 4 5
RESP score 3 0 4 5 2
SOFA score 4 13 5 4 4
Murray’s score 4 3.8 3.5 3.5 3.3
Vasopressors No Yes No No Yes
RRT No No No No No
Pregnancy status Post-partum Pregnant Post-partum Post-partum Post-partum
ECMO management
Initial Configuration Jugular bicaval Femoral-femoral Femoral-jugular Femoral-jugular Femoral-jugular
Need for VV-V Yes Yes Yes No No
Cannulae 31Fr avalon jugular & 25Fr femoral 25Fr/19Fr [plus 19Fr femoral] 25Fr/19Fr [plus 19Fr femoral] 25Fr/21Fr 25Fr/19Fr
Anticoagulation Unfractionated heparin [target apTT 40–60s]
Peak blood flows, LPM 5.6 5.5 6.2 3.7 4.9
Peak sweep flows, LPM 9 7 8 6 5
Prone positioning No No Yes, 2 sessions No No
Extubated No No 72-h only No No
Tracheostomy Yes, percutaneous Yes, percutaneous Yes, percutaneous Yes, percutaneous Yes, percutaneous
Ambulation Yes Yes Yes Yes Yes
apTT; Activated Partial Thromboplastin Time, ARDS; Acute Respiratory Distress Syndrome, BMI; Body Mass Index, ECMO; Extracorporeal Membrane Oxygenation, HFNC; High Flow Nasal Cannula, LPM; Litres Per Minute, NIV; Non-invasive Ventilation, NO; Nitric Oxide, PEEP; Positive End-Expiratory Pressure, RRT; Renal Replacement Therapy.
Table 2. Patient outcomes.
Patient 1 Patient 2 Patient 3 Patient 4 Patient 5
Status • Transferred on ECMO for lung transplant • Home • Acute rehabilitation • Home • Home
• CPC 1 • CPC 1 • CPC 1 • CPC 1
• Chemotherapy completed • Trach. Out • Trach. Out • Trach. Out
• Trach. Mask • Trach. Out • no oxygen • no oxygen • 1L oxygen
• Using exercise bike but not able to stand • no oxygen
ECMO days 122 58 32 7 76
Ventilation days 142 81 40 16 116
ICU LOS, days 142 88 41 26 121
Hospital LOS, days 156 185 62 43 193
Non-ECMO Complications • Empyema • AKI/CRRT • AKI/CRRT • Pneumomediastinum • Pneumothorax
• Critical illness myopathy • Pneumothorax • Haemothorax
• Pneumothorax • Ileus • Pneumothorax • Empyema
• Bronchopleural fistula
ECMO complications • Cardiac tamponade • DVT lower limb • Oxygenator failure None • Oxygenator failure
• Oxygenator failure • Oxygenator failure
• Accidental decannulation
Pregnancy outcome LSCS at 33/40 LSCS at 27/40 LSCS at 27/40 Induced at 29/40 LSCS at 34/40
Foetal outcome Alive, thriving IUD on ECMO Alive, thriving IUD pre-ECMO Alive, thriving
AKI; Acute Kidney Injury, CPC; Cerebral Performance Category, DVT; Deep Vein Thrombosis, LSCS; Lower Segment Caesarean Section, CRRT; Continuous Renal Replacement Therapy, ECMO; Extracorporeal Membrane Oxygenation, LOS; Length of Stay, Trach; Tracheostomy.
All patients underwent early intense physical rehabilitation, and to facilitate this, we performed early percutaneous tracheostomy, consent for which was taken along with the consent for ECMO. No complications occurred during the tracheostomy in any of the patients. Four out of five patients in this cohort were able to walk regularly with VV ECMO. One patient with severe critical illness neuromyopathy [patient 1] recovered considerable arm strength. Although the patient recovered enough strength in her legs to be able to exercise using a pedal bike and the leg press, she was unable to walk until the time of her transfer to the transplant centre.
Although we inserted reasonably large cannulae in most of our patients, three patients needed veno-venovenous [V-VV] ECMO to achieve higher flows. Significant recirculation was ruled out based on access blood oxygen saturations of less than 70% and adequate cannula separation [more than 10–15 cm] in case of femoral-jugular cannulation. Membrane lung [oxygenator] dysfunction was also ruled out in these patients by confirming return blood partial pressure of oxygen greater than 30 kPa. Patient 1 became severely hypoxic after 3-weeks on ECMO. Due to the inability to achieve higher flows with the jugular bicaval cannula, confirmed optimal positioning per transoesophageal echocardiography, and presence of pulmonary hypertension with severe tricuspid regurgitation, we decided to place additional access cannula which resolved the hypoxia. Given her significant critical illness neuromyopathy and to avoid hindering her ongoing rehabilitation for lung transplantation, we preferred not to sedate and paralyze her. Additionally, her repeat CT at this time showed diffuse bilateral consolidations with no aeration, so we did not expect any benefit from prone positioning. Patient 2 had newly diagnosed Hodgkin’s Lymphoma with large lymph nodes in the neck and mediastinum with concerns for superior vena cava obstruction. This patient was cannulated by our team at an outside facility. She had already been in prone position for 96-h and became critically hypoxic once turned supine for cannulation. We started with a femoro-femoral configuration, but the patient remained profoundly hypoxic. It was not deemed safe to transfer her without addressing the hypoxia. We then proceeded with a rather challenging cannulation of the right internal jugular vein and switched the configuration to V-VV ECMO and that resulted in improved oxygenation. Patient 3 became hypoxic after 3-weeks on VV ECMO during an episode of sepsis. Her cardiac output persistently remained greater than 10 L per minute despite physiological manipulations [cooling and Esmolol infusion]. She continued to be hypoxic despite blood transfusion to target higher haemoglobin; optimization of ECMO flows and mechanical ventilation; prone positioning; inhaled nitric oxide; and neuromuscular paralysis. Her echocardiogram revealed severe pulmonary hypertension with systolic pulmonary artery pressures greater than 90mmHg. We performed a repeat CT Pulmonary Angiogram that ruled out pulmonary embolism. Patient received an additional access cannula to switch to V-VV configuration resulting in a significant improvement in oxygenation as well as pulmonary artery pressures.
All but one patient could be weaned from ECMO and the mechanical ventilator. At the time of writing this report, all five patients have been either discharged (n = 4) or transferred to another centre for lung transplant (n = 1). All patients suffered barotrauma, prior to [patients 1, 4, and 5] or after [patients 2 and 3] ECMO initiation. None of the pneumothoraces were iatrogenic. Patient four needed VV ECMO for only 7-days despite significant parenchymal disease as evidenced by Murray’s Lung Injury Score of 3.3 and lack of response to prone positioning. Her cultures, including bronchoalveolar lavage ruled out secondary sepsis. We believe, pneumomediastinum and surgical emphysema were a significant factor in her clinical deterioration prior to ECMO. The problem returned once her tracheostomy was removed. A small tracheal laceration was subsequently found in her upper trachea and was managed conservatively with reinsertion of tracheostomy for an extended period. Oxygenator failure was the most common ECMO complication with four out of five patients needing at least one oxygenator change. One patient suffered accidental decannulation requiring emergent re-cannulation in a peri-arrest scenario.
Discussion
This small, single centre series suggests that VV ECMO in peripartum women with refractory CoViD-19-related-ARDS is associated with reasonable maternal outcomes. A recent multicenter case series of nine patients also reported excellent outcomes.10 Our small sample size prohibits any meaningful statistical analysis or comparisons, but we would like to discuss a few observations. First, all our patients were retrieved from referring hospitals. Delayed referrals [patients 3, 4 and 5], the need to retrieve from international location [patient 1], and lack of availability of ECMO machines [patient 2] resulted in longer duration between intubation and ECMO initiation in our series. While there is some evidence to suggest that intubation to ECMO duration of less than 3-days is associated with better outcomes,9 recent data on CoViD-19 ARDS patients found no such association.11
Second, systematic review of VV ECMO for CoViD-19 related ARDS have reported a mean ECMO duration of 15 days 12 but little information is available in published series regarding factors associated with longer duration of ECMO in these patients. Mean ECMO duration in our series was much longer, with a mean duration of 59-days and median 58-days. Three out of five patients in our series [patients 1, 2 and 5] developed advanced parenchymal disease warranting referral for lung transplant but only one patient was accepted for transfer [patent 1]. Patient 2 was declined due to active malignancy and patient 5 was declined due to intractable Candida Auris empyema. We persisted medically with these patients and were eventually able to wean them from ECMO. As a result, they had a much longer stay on ECMO.
Third, data on timing of tracheostomy and its impact on rehabilitation in VV ECMO patients is conflicting. A large international retrospective series in non-COViD-19 ARDS patients on VV ECMO, found no benefit of performing early tracheostomy in reducing the amount of sedation or achieving higher ambulation rates, and reported higher bleeding complications.13 A more recent review of the Extracorporeal Life Support [ELSO] registry data on tracheostomy practices in CoViD-19 patients receiving VV ECMO shows that early tracheostomy while patients were still on ECMO did help achieve higher ambulation rates albeit with higher surgical site bleeding rates.14 All our patients underwent percutaneous tracheostomy within 48-h of going on ECMO and managed to participate in physical rehabilitation soon after that. No patient had any minor or major complications from tracheostomy or ambulation.
Fourth, it is not clear why four out of five patients deteriorated significantly, immediately after delivery. It has been reported that pregnant women can have a differential response to infections depending upon stage of pregnancy and the maternal immune system is deemed particularly pro-inflammatory during the 3rd trimester and around delivery, but this association has not been validated in large clinical studies.15 Perhaps pregnant women with CoViD-19 infection should be followed more closely and delivered at centres with ready access to tertiary critical care services.
Fifth, there is some reluctance to prone patients with late-stage pregnancy, even though it has been suggested that this can be safely accomplished.16,17 This may be why only one of our patients received prone positioning while they were still pregnant. This patient was proned at 21-weeks pregnancy for almost 120-h including a single 96-h session, given the continued severe hypoxia and delays in arranging ECMO retrieval. No immediate impact on the foetal wellbeing was noted on follow up assessments.
Sixth, only one of our patients was still pregnant at the time of ECMO initiation. Unfortunately, this foetus died 6-weeks later while the plans were underway to perform a semi-elective caesarean section given ongoing sepsis in the mother. Our nurses monitored foetal heart rate every shift, the physicians used the ultrasound once a day to confirm the foetal heart rate, and the obstetricians reviewed the patient at least once a week. However, the maternal and foetal heart rates were too close to each other [130–140 per minute] and we believe that this may have contributed to a delay in the diagnosis of intrauterine death. This raises questions about how best to monitor the foetal wellbeing in critically ill pregnant women away from a typical obstetric setup. One series reported 24-h electronic foetal monitoring for the first 24-h after going on ECMO and then for 1-h every 12-h18 However, there is currently no consensus or guidance on this topic and little information available on prevailing practices from systematic reviews.7
Conclusion
VV ECMO for refractory CoViD-19-related ARDS in the peripartum period showed good maternal survival rates and had low rates of complications in this small cohort and may be considered as a rescue option.
ORCID iDs
Vivek Kakar https://orcid.org/0000-0001-6353-3283
Nuno Raposo https://orcid.org/0000-0001-9931-8905
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
==== Refs
References
1 DeBolt CA Bianco A Limaye MA , et al. Pregnant women with severe or critical coronavirus disease 2019 have increased composite morbidity compared with nonpregnant matched controls. Am J Obstet Gynecol 2020; 224 : 510e1–510e12.
2 Metz TD Clifton RG Hughes BL , et al. Association of SARS-CoV-2 infection with serious maternal morbidity and mortality from obstetric complications. JAMA 2022; 327 : 748–759.35129581
3 Allotey J Stallings E Bonet M , et al. Clinical manifestations, risk factors, and maternal and perinatal outcomes of coronavirus disease 2019 in pregnancy: living systematic review and meta-analysis. BMJ 2020; 370 : 1–17.
4 Schmidt M Hajage D Lebreton G , et al. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome associated with COVID-19: a retrospective cohort study. Lancet Respir Med 2020; 8 : 1121–1131.32798468
5 Zhang J Merrick B Correa GL , et al. Veno-venous extracorporeal membrane oxygenation in coronavirus disease 2019: a case series. ERJ Open Res 2020; 6 : 1–8.
6 Barbaro RP MacLaren G Boonstra PS , et al. Extracorporeal membrane oxygenation support in COVID-19: an international cohort study of the extracorporeal life support organization registry. Lancet 2020; 396 : 1071–1078.32987008
7 Naoum EE Chalupka A Haft J , et al. Extracorporeal life support in pregnancy: a systematic review. J Am Heart Assoc 2020; 9 : 1–15.
8 Ramanathan K Tan CS Rycus P , et al. Extracorporeal membrane oxygenation in pregnancy: an analysis of the Extracorporeal Life Support Organization registry. Crit Care Med 2020; 48 : 696–703.32191415
9 Combes A Hajage D Capellier G , et al. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med 2018; 378 : 1965–1975.29791822
10 Barrantes JH Ortoleva J O'Neil ER , et al. Successful treatment of pregnant and postpartum women with severe COVID-19 associated acute respiratory distress syndrome with extracorporeal membrane oxygenation. ASAIO J 2021; 67 : 132–136.33229971
11 Hermann M Laxar D Krall C , et al. A duration of invasive mechanical ventilation prior to extracorporeal membrane oxygenation is not associated with survival in acute respiratory distress syndrome caused by coronavirus disease 2019. Ann Intensive Care 2022; 12 : 6.35024972
12 Bertini P Guarracino F Falcone M , et al. ECMO in COVID-19 patients: a systematic review and meta-analysis. J Cardiothorac Vasc Anesth 2022; 36 : 2700–2706.34906383
13 Schmidt M Fisser C Martucci G , et al. Tracheostomy management in patients with severe acute respiratory distress syndrome receiving extracorporeal membrane oxygenation: an international multicenter retrospective study. Crit Care 2021; 25 : 238.34233748
14 Kohne JG MacLaren G Cagino L , et al. Tracheostomy practices and outcomes in patients with COVID-19 supported by extracorporeal membrane oxygenation: an analysis of the extracorporeal life support organization registry. Crit Care Med 2022; 50 : 1360–1370.35607973
15 Mor G Cardenas I . The immune system in pregnancy: a unique complexity. Am J Reprod Immunol 2010; 63 : 425–433.20367629
16 Guérin C Albert RK Beitler J , et al. Prone position in ARDS patients: why, when, how and for whom. Intensive Care Med 2020; 46 : 2385–2396.33169218
17 Tolcher MC McKinney JR Eppes CS , et al. Prone positioning for pregnant women with hypoxemia due to coronavirus disease 2019 (COVID-19). Obstet Gynecol 2020; 136 : 259–261.32516274
18 Lankford A Chow JH Jackson A , et al. Clinical outcomes of pregnant and postpartum extracorporeal membrane oxygenation patients. Anesth Analgesia 2021; 132 : 777–787.
| 36484202 | PMC9742351 | NO-CC CODE | 2022-12-14 23:30:05 | no | Perfusion. 2022 Dec 9;:02676591221144729 | utf-8 | Perfusion | 2,022 | 10.1177/02676591221144729 | oa_other |
==== Front
Ann Otol Rhinol Laryngol
Ann Otol Rhinol Laryngol
AOR
spaor
The Annals of Otology, Rhinology, and Laryngology
0003-4894
1943-572X
SAGE Publications Sage CA: Los Angeles, CA
36482672
10.1177/00034894221138485
10.1177_00034894221138485
Original Article
Chemosensory Dysfunction 3-Months After COVID-19, Medications and Factors Associated with Complete Recovery
https://orcid.org/0000-0001-5213-2337
Fornazieri Marco Aurélio MD, PhD 1234
da Silva José Lucas Barbosa MD, PhD 1
Gameiro Juliana Gutschow MSc 1
Scussiato Henrique Ochoa MD, MSc 1
Ramos Rafael Antônio Matias Ribeiro 1
Cunha Bruno Machado 1
Figueiredo Alan Felipe 1
Takahashi Eduardo Hideki 1
Marin Gabrielli Algazal 2
Caetano Igor Ruan de Araújo 2
Meli Tainara Kawane 1
Higuchi Diego Issamu 2
dos Santos Rafael Rodrigues Pinheiro 2
Rampazzo Ana Carla Mondek 2
Pinna Fábio de Rezende MD, PhD 3
Voegels Richard Louis MD, PhD 3
Doty Richard L. PhD 4
1 Londrina State University, Londrina, Paraná, Brazil
2 Pontifical Catholic University of Paraná, Londrina, Brazil
3 University of São Paulo, São Paulo, State of São Paulo, Brazil
4 Smell and Taste Center, Department of Otorhinolaryngology, Head and Neck Surgery, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
Marco Aurélio Fornazieri, MD, PhD, Department of Clinical Surgery, Londrina State University, 60 Robert Koch Avenue, Londrina, State of Paraná 86038350, Brazil. Email: [email protected]
8 12 2022
8 12 2022
00034894221138485© The Author(s) 2022
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Objectives:
To examine the longitudinal prevalence and recovery of olfactory, gustatory, and oral chemesthetic deficits in a sizable cohort of SARS-CoV-2 infected persons using quantitative testing. To determine whether demographic and clinical factors, mainly the medications used after the COVID-19 diagnosis, influence the test measures.
Methods:
Prospective cohort in a hospital with primary, secondary, tertiary, and quaternary care. Patients with confirmed COVID-19 were tested during the acute infection phase (within 15 days of initial symptom, n = 187) and one (n = 113) and 3 months later (n = 73). The University of Pennsylvania Smell Identification Test, the Global Gustatory Test, and a novel test for chemesthesis were administered at all visits.
Results:
During the acute phase, 93% were anosmic or microsmic and 29.4% were hypogeusic. No one was ageusic. A deficit in oral chemesthesis was present in 13.4%. By 3 months, taste and chemesthesis had largely recovered, however, some degree of olfactory dysfunction remained in 54.8%. Remarkably, patients who had been treated with anticoagulants tended to have more olfactory improvement. Recovery was greater in men than in women, but was unrelated to disease severity, smoking behavior, or the use of various medications prior to, or during, COVID-19 infection.
Conclusions:
When using quantitative testing, olfactory disturbances were found in nearly all SARS-CoV-2 infected patients during the acute infection phase. Taste or chemesthetic deficits were low. Olfactory impairment persisted to some degree in over half of the patients at the 3-month follow-up evaluation, being more common in women and less common in those who had been treated earlier with anticoagulants.
Level of Evidence:
3
smell
taste
COVID-19
SARS-CoV-2
olfaction disorders
olfactory perception
sex
diagnostic tests
edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
Persons with chemosensory deficits due to SARS-CoV-2 frequently turn to physicians for prognosis and treatment. Unfortunately, informative data on these points are limited. Prevalence of dysfunction based on extant psychophysical tests range, over the course of a month since symptom onset, from 50.6% to 96% for olfaction,1-4 61.3% to 67.5% for taste.3,5 The chemesthesis is responsible for the burning, cooling, or tingling sensations of food and other stimuli and is reportedly distorted in some patients,6 and no studies have addressed the empirical frequency of these chemesthetic deficits in COVID-19.
A number of studies suggest that most patients recover their smell or taste capacities within 10 days from the onset of the COVID-19 infection, although nearly 40% does not achieve complete recovery.1 If the chemosensory dysfunction remains after 20 days, spontaneous resolution is believed to be less likely to occur.3 Although treatment options for these conditions are being explored, none have shown strong efficacy.7,8 It is unknown whether the drugs used for treatment during the acute phase, or prior to the infection, are related to the timing or degree to which recovery occurs.
In this study, we examined the prevalence of smell, taste, and chemesthetic disorders in patients with confirmed COVID-19 during its acute infection phase and 1 and 3 months later. We examined whether the degree of dysfunction noted at these time periods was related to age, sex, race, education, disease severity, and smoking behavior, as well as anticoagulant, antihypertensive, and diabetic medications administered before or after the COVID-19 diagnosis.
Methods
The initial study group was comprised of 187 patients with PCR-confirmed COVID-19 enrolled between September 2020 and October 2021. Of these, 113 were retested at the 1-month follow-up period and 73 at the 3-month follow-up period. The longitudinal attrition was caused by 1 death, 72 patients who could not be re-contacted or who did not wish to further participate, and 43 patients who were unable to complete a full test session. The severity of the COVID-19 was classified as either mild (ie, nonpneumonia and mild pneumonia) or severe/critical (dyspnea, respiratory frequency ≥30/min, blood oxygen saturation ≤93%, partial pressure of arterial oxygen to fraction of inspired oxygen ratio <300, and/or lung infiltrates >50% within 24 to 48 hours, respiratory failure, septic shock, and/or multiple organ dysfunction or failure).9 Demographic and clinical factors and all the medications used previously and after the COVID-19 diagnosis were collected.
Olfaction was tested using the Portuguese language version of the University of Pennsylvania Smell Identification Test (UPSIT®). The UPSIT® is a widely employed standardized 40-item 4-alternative forced-choice olfactory test known to correlate strongly with olfactory threshold test scores.10 Normal olfactory function was based on normative data from the Brazilian adaptation and validation study of this test.11
Taste was measured using the Global Gustatory Test (GGT).12 This test is a whole-mouth modified version of the regional taste test described by Soter et al.13 It employed aqueous solutions of 0.31 M NaCl (salty), 0.015 M citric acid (sour), 0.49 M sucrose (sweet), and 0.04 M caffeine (bitter). The stimuli were presented as oral spays, with 4 replications each presented in a pre-determined random order, that is, a total of 16 trials. Patients with more than 12 correct answers out of 16 were considered normogeusic, those with 4 to 12 were considered hypogeusic, and those with less than 4 were considered ageusic.12,13
Oral chemesthesis was tested using, for the first time, a novel carbonated water test. Carbonated water is known to activate lingual nociceptors as a result of the conversion of CO2 to carbonic acid which, in turn, activates oral trigeminal neurons (Dessirier et al., 2000). In this test, the subject sipped 50 ml of carbonated water (Água com Gás Crystal, The Coca Cola Company, São Paulo, Brasil) for 10 seconds and then either swallowed or expectorated the water into a waste container. The water bottle was opened immediately before presentation to the subject. The subject rated his or her chemesthetic sensation on a modified 10-point category scale, with 0 reflecting a completely distorted tingling sensation and 10 a completely normal tingling sensation. A score of 7 or above was classified as normal, a score between 3 and 7 as moderately distorted, and a score of 3 or less as severely altered. This test was also administered to 15 control individuals (age range: 22-61 years) without a history of COVID-19 infection. Fourteen reported a completely normal tingling sensation (score of 10) and only 1 scored 5 on the scale.
In addition to the empirical testing, subjects also rated their self-perceived taste and smell abilities on a 10-point category scale, with 0 being total lack of sensation and 10 normal sensation. Subjects were excluded from the study if they had a history of head trauma, a neuropsychiatric disorder, cerebrovascular disease, chemotherapy, radiotherapy, or self-reported loss of smell or taste that preceded COVID-19. Data from subjects who were unable to complete the testing were omitted from analysis. The Universidade Estadual de Londrina’s Ethics Committee approved the study, and all patients signed the informed consent.
Statistical Analyses
Prevalence is depicted in percentages and continuous data are reported as means and standard deviations or as indicated. Percentages of normosmics versus microsmics/anosmics among groups were compared by Fisher’s exact test. Logistic regression models were used to estimate the association between the variables and the rates of normosmia, full recovery criterium. Demographic and clinical data were assessed using a stepwise procedure based on the Akaike information criterion (AIC) for achieving the best model to predict the smell recovery.14P values at or below .05 were considered significant. When multiple comparisons were made, the Bonferroni correction was employed. The statistical software used was STATA 17 (StataCorp LP, College Station, TX).
Results
During the acute phase of COVID-19, 174 out of the 187 patients (93%) were found to be anosmic or microsmic and 55 (29.4%) hypogeusic. None were ageusic. Chemesthesis mouth perceptions were partially or severely altered in 25 volunteers (13.4%, Figure 1). The scores on these tests can be seen in Figure 2. At this time, 36 (19.4%) subjects reported experiencing parosmia and 29 (15.6%) phantosmia. By the 1- and 3-month follow-up periods, taste and chemesthesis functions were normal in 67 (91.8%) and 72 (98.6%) patients. However, fewer than half of the patients exhibited normal olfactory function at that time. At the 1-month follow-up, the parosmia and phantosmia values of the 113 subjects decreased to 4.8% and 6.9%, respectively. At the 3-month follow-up, similar rates of parosmia and phantosmia were noted in the 73 patients that were evaluated (7.5% and 6.7%, respectively). The most frequent qualitative sensation was described as smoke-like.
Figure 1. Flowchart of the study with major results.
Figure 2. UPSIT (A), Global Gustatory Test (B), and oral chemesthesis (C) scores for the acute phase (n = 187), 1 month (n = 113), and 3 month (n = 73) follow-up periods. The distribution of the patients’ scores in each group is depicted in a violin plot. Horizontal black lines: medians and interquartile ranges. Abbreviations: GGT, global gustatory test; UPSIT, University of Pennsylvania smell identification test.
The univariate unadjusted analyses comparing data between subjects with total olfactory recovery and subjects with partial or no recovery at the 3-month evaluation are shown in Table 1. Women had a worse total normosmia rate than men (28.2% vs 64.7%, P = .002). No differences between the 2 groups were found for COVID-19 severity, age, race, education, number of comorbidities, smoking behavior, or previous use of drugs for hypertension, diabetes, or depression. None of the patients who were normosmics during the acute phase developed olfactory dysfunction at the 3-month follow-up.
Table 1. Demographic and Clinical Characteristics of Patients Normosmic or Microsmic/Anosmic in the University of Pennsylvania Smell Identification Test after 3 Months SARS-CoV-2 Infection.
Characteristics Total (N = 73) Anosmia or Microsmia (N = 40) Normosmia (N = 33) P value
Age, mean (SD), y 41.5 (13.8) 42 (14.5) 40.9 (2.3) .73
Sex, No. (%) .002
Female 34 (46.6) 28 (71.8) 11 (28.2)
Male 39 (53.4) 12 (35.3) 22 (64.7)
Race, No. (%) .59
White 54 (74) 31 (57.4) 23 (42.6)
Non-white 19 (26) 9 (47.4) 10 (52.6)
Years of schooling, mean (SD), y 12.2 (3.2) 12.2 (3.3) 12.2 (3.1) .98
No. of comorbidities, No. (%) .43
0 116 (62) 27 (67.5) 25 (75.8)
1 44 (23.5) 9 (22.5) 3 (9.1)
2 16 (8.6) 1 (2.5) 5 (15.1)
3 6 (3.2) 1 (2.5) -
4 4 (2.2) 2 (5) -
5 1 (0.5) - -
Smoking, No. (%) .15
Smoker or former smoker 15 (20.5) 11 (73.3) 4 (36.7)
Non-smoker 58 (79.5) 29 (50) 29 (50)
Comorbidities, No. (%)
Yes 21 (28.8) 13 (32.5) 8 (24.2) .60
No 52 (71.2) 27 (67.5) 25 (75.8)
Hypertension .50
Yes 9 (12.3) 6 (15) 3 (9.1)
No 64 (87.7) 34 (85) 30 (90.9)
Diabetes 1.00
Yes 7 (9.6) 4 (10) 3 (9.1)
No 66 (90.4) 36 (90) 30 (90.9)
Hypothyroidism 1.00
Yes 3 (4.1) 2 (66.7) 1 (33.3)
No 70 (95.9) 38 (54.3) 32 (45.7)
Dyslipidemia 1.00
Yes 6 (8.2) 3 (50) 3 (50)
No 67 (91.8) 37 (55.2) 30 (44.8)
Asthma .59
Yes 3 (4.1) 1 (33.3) 2 (66.7)
No 70 (95.9) 39 (55.7) 31 (44.3)
Use of previous medications, No. (%)
Antihypertensives .72
Yes 8 (11) 5 (62.5) 3 (37.5)
No 65 (89) 35 (53.8) 30 (46.2)
Diabetes Mellitus medications .33
Yes 10 (13.7) 4 (40) 6 (60)
No 63 (86.3) 36 (57.1) 27 (42.9)
Antidepressants .28
Yes 8 (11) 6 (75) 2 (25)
No 65 (89) 34 (52.3) 31 (47.7)
COVID-19 severity, No. (%) .78
Mild 41 (66.1) 26 (63.4) 15 (36.6)
Severe/Critical 21 (33.9) 12 (57.1) 9 (42.9)
An analysis of the drugs administered during the acute phase found that anticoagulant use was associated with a higher percentage of subjects recovering their smell function by the 3-month time period (70.6% vs 37.5% of those who did not use, P = .02, Table 2). No such association was found for any of the other drugs used at this time, namely corticosteroids (P = .42), azithromycin (P = .14), ivermectin (P = 1.00), albendazole (P = − .32), or hydroxychloroquine (P = .12), although very few subjects were taking the latter 2 drugs, limiting statistical power to assess such an effect.
Table 2. Percentages of Tested Normal Smell Function and Unadjusted Comparisons Between Patients that Used or Did Not Use Different Medications After 3 Months of SARS-CoV-2 Infection.
Medication, No. (%) Total Complete recovery No or partial recovery P value
Anticoagulants .02
Yes 17 (23.3) 12 (70.6) 5 (29.4)
No 56 (76.7) 21 (37.5) 35 (62.5)
Systemic corticosteroid .41
Yes 18 (24.7) 10 (55.6) 8 (44.4)
No 55 (75.3) 23 (41.8) 22 (58.2)
Azithromycin .14
Yes 25 (34.3) 8 (32) 17 (38)
No 48 (65.7) 25 (52.1) 23 (47.9)
Hydroxychloroquine .12
Yes 4 (5.5) 0 (0) 4 (100)
No 69 (94.5) 36 (52.2) 33 (47.8)
Ivermectin 1
Yes 17 (23.3) 8 (47.1) 9 (52.9)
No 56 (76.7) 25 (44.6) 31 (55.4)
Albendazole .32
Yes 4 (5.5) 3 (75) 1 (25)
No 69 (94.5) 30 (43.5) 39 (56.5)
In the stepwise logistic regression model that including all clinical and demographic variables, the aforementioned effect of anticoagulant therapy failed to reach statistical significance (P = .23), only being a woman was associated with a worse smell outcome (OR: 0.32, 95% CI: 0.10-0.96, P = .04). As in the previous analysis, none of the medications used after SARS-CoV-2 infection was associated with a better prognosis on the olfactory outcome (ps > 0.10). Although a few patients failed to recover completely taste or chemesthesis function, the numbers were too small to provide adequate statistical power to determine whether such variables were meaningfully associated with lack of recovery.
Figure 3 shows the subjective assessments of olfactory and taste functions. An analysis of the self-reported olfactory function during the acute phase found that only 6 of the 41 patients (14.6%) who reported having a normal sense of smell were normal upon psychophysical testing. Of those tested 1 month later, even a higher number (23 out of 34, or 67.6%) exhibited this misperception. In the third-month follow-up, this percentage was 54.5% (12 out of 22 patients).
Figure 3. Self-reported smell (A) and taste (B) abilities on a 10-point category scale, with 0 being total lack of sensation and 10 normal sensation for the acute phase (n = 187), 1 month (n = 113), and 3 month (n = 73) follow-up periods.
Discussion
In this study, 93% of 187 Brazilian patients tested within the 15-day acute period of the SARS-CoV-12 infection exhibited olfactory dysfunction, as measured by the UPSIT®. The prevalence of the dysfunction decreased to 77.9% and 54.8% at the 1-month and 3-month follow-up periods, respectively. These findings are in general accord with earlier studies employing the UPSIT® in COVID-19 cases, although some differences are apparent. Moein et al15 reported a 98% of 60 Iranian patients exhibited some degree of smell dysfunction on the UPSIT® during the acute period. By extending the sample size to 100, a 96% prevalence rate was noted.1,15 The latter study retested 86 of the patients and found that the prevalence rate decreased to 39% 6 to 8 weeks after symptom onset, a value somewhat lower than the 3-month prevalence rate of the present study. González et al16 noted a 73% prevalence rate for 100 COVID-19 Chilean patients tested with the UPSIT® during the acute phase of the disorder which decreased to 41% for those tested a month later, the latter being similar to the 6 to 8 weeks prevalence observed by Moein et al.1 In a study of 56 Australian COVID-19 patients 6 months after their diagnosis, Leedman et al17 found 35.6% exhibited some degree of smell loss. Boscolo-Rizzo et al18 noted, using a modified version of the UPSIT®, that 60% of 87 Italian patients tested 6 months after the COVID-19 diagnosis still had some degree of smell dysfunction. Despite some variation, all of these studies are in agreement that the vast majority of persons, when empirically tested with a 40-odorant test, experience some measurable smell loss during the acute phase of the SARS-CoV-2 infection that continues in some patients for months after the onset of the infection.
While 19.4% of the patients reported experiencing parosmia and 15.6% phantosmia during the acute phase, the parosmia and phantosmia values of the 113 subjects decreased to 4.8% and 6.9%, respectively, at the 1-month follow-up. Similar rates of parosmia and phantosmia were noted in the 73 patients evaluated at the 3-month follow-up (7.5% and 6.7%). Although few studies are available in the literature concerning the prevalence of either parosmia or phantosmia in post-COVID-19 patients, Ercoli et al19 reported phantosmia in 11.8%, and parosmia in 23.5%, in their 17 patients who had their acute infection several months earlier.
Our study, like most others, demonstrates the disconnect between a patient’s self-report of olfactory function and measured function. On average, studies that have used psychophysical olfactory test measures have reported a higher prevalence of smell disorders than those who have relied on self-report.4,20 Such findings stress the importance of olfactory testing and making patients and physicians aware of the limitations of self-report. Early accurate identification of smell loss could lead to earlier treatment and better olfactory outcomes.21
In accord with most other quantitative studies, we found that only a minority of persons infected with SARS-CoV-2 exhibited abnormal taste test scores (13.4%). In a study of 111 COVID-19 patients, Niklassen et al22 found 26% to have demonstrable taste dysfunction during the infectious period, as measured by tests in which the tastants were either sprayed into the mouth or presented to the lingual surface on strips of paper embedded with tastants. After this period was over, only 6.5% continued to exhibit taste loss. In contrast, 90% of the patients initially exhibited some degree of smell loss which declined to 27% thereafter. Cao et al23 identified 11 of 250 healthcare workers who had experienced SARS-Cov-2 infection. All had been tested with the 27-item version of the Waterless Empirical Taste Test (WETT®), a test that also uses taste strips for stimulus presentation. While scores on a 12-item smell identification test were significantly lower in those who had a recent SARS-CoV-2 infection, this was not the case for the taste test scores that fell, on average, at the 50th percentile of those from a normative group.23 While the smell test scores were related to the time since the onset of chemosensory systems, this was not the case for the taste test scores. Somewhat higher taste dysfunction prevalence rates were noted by Vaira et al.24 These authors used a single-trial 4-solution taste test and reported that 47% of 72 patients exhibited either hypogeusia (n = 33) or ageusia (n = 1) early in the disease process. Petrocelli et al25, also using a homemade 1-trial 4-solution taste test, concluded that 38% of their sample of 300 patients exhibited ageusia.
An important observation of our study is that olfaction appeared to recover earlier in patients who had been treated with anticoagulants during the acute period of infection, although this effect was not strong and was not statistically significant at the .05 alpha level when the full model was employed. It is well established that anticoagulants can attenuate inflammation and that inflammation initially leads to coagulation. The presence of coagulant factors augment the inflammatory response, although subsequently the anticoagulant response diminishes inflammatory mediators.26 One way that COVID-19 affects the nervous system is its impact on cerebral microvasculature as manifested by vasculitis, microangiopathy, coagulopathy, and in-situ circulatory microthrombi.27-29 Hence, anticoagulants may lower vascular thrombi within the olfactory mucosa, notably the lamina propria, olfactory bulb regions, and possibly cerebral regions involved in olfactory transmission.29 Although other research has demonstrated that anticoagulation has benefits for COVID-19 patients independent of olfaction,30,31 the present study is the first to provide evidence that such therapy may be of value in sparing COVID-19-related olfactory dysfunction, providing one more potential therapeutic option for the treatment of such dysfunction. More research is clearly needed on this point.
Aside from the possible beneficial effects of anticoagulant therapy, our study found no other benefit from the medications that were evaluated, including corticosteroids. This is at variance with reports that such steroids may improve olfaction in COVID-19 patients.32,33 The reason for this discrepancy is not known. Our negative results for azithromycin, chloroquine, and hydroxychloroquine are consistent with the literature.34-37 However, the evidence of the use of these drugs on the olfactory function in COVID-19 patients is sparse in the literature and more studies are needed to investigate that relationship.
Confirming previous studies,38,39 the percentage of women having a complete recovery of their sense of smell at the third-month follow-up was lower than that of the men. Being a woman was associated with a 68% lower chance of total olfactory recovery. Interestingly, 1 study showed that women also have a more extended recovery period than men.40
As with the taste tests, the test of oral chemesthesis we employed found smaller prevalences of these symptoms compared to olfaction (13.4% vs 93%, 8% vs 76%, and 1.4% vs 54.8%, at the acute phase, 1 and 3-month follow up, respectively). Moreover, just 1 patient had altered chemesthetic perception at the third-month follow-up. Chemesthesis is well explored in the food industry.41 However, there are very few studies that have investigated chemesthetic distortion in clinical practice, with most simply asking about the symptoms via questionnaires.6 The present study is the first to use a novel carbonated oral water test for COVID-19 subjects. Another COVID-19 study has measured nasal chemesthesis. They employed a 70% acetic acid solution and rated the stinging sensation inside the nose after sniffing the vapor.42 Even after 1 year of COVID-19 infection, patients reported a reduced sensitivity at the nasal cavities with this substance compared to controls. The difference between their study and ours suggests that nasal chemesthesis may be more affected than oral chemesthesis during SARS-CoV-2 convalescence.
This study has both strengths and weaknesses. Strengths include (a) the longitudinal quantitative testing of smell, taste, and chemesthesis in a sizable number of COVID-19 patients during and 2 times following the acute phase and (b) the exploration of factors, including medications, that may be associated with faster recovery over time. Among the limitations of the study are (a) the lack of a concurrent control group and (b) the drop out of participants in the longitudinal arm of the study due to logistic and personal reasons. The latter not only decreases statistical power, but theoretically could also influence the conclusions of the study, as these patients may have declined to participate or could not be contacted because their sense of smell was normal. An additional potential limitation of the study is the use of a novel test of chemesthesis for which control data are limited. Although this test is modeled on tests found useful in the food industry, its clinical application and sensitivity to COVID-19 is limited so more validation of the test would seem useful.
Conclusions
We have verified that more than half of the COVID-19 patients persists with olfactory dysfunction after 3 months of the disease onset, this sequela being more prevalent in women. Taste and chemesthetic function returned to normal in almost all patients. No medications used during this period were associated with a better outcome on the smelling capacity, even though patients who used anticoagulants tended to have a high rate of total olfactory recovery.
Contributor’s Statements: Marco Aurélio Fornazieri: Prof. Fornazieri conceptualized and designed the study, conducted the data collection, conducted the analyses, drafted the initial manuscript, and revised the manuscript. José Lucas Barbosa da Silva, Juliana Gameiro, Henrique Ochoa Scussiato: coordinated data collection, conducted the analyses, drafted the initial manuscript, and revised the manuscript. Rafael Antônio Matias Ribeiro Ramos, Bruno Machado Cunha, Alan Felipe Figueiredo, Eduardo Hideki Takahashi, Gabrielli Algazal Marin, Igor Ruan de Araújo Caetano, Tainara Kawane Meli, Diego Issamu Higuchi, Rafael Rodrigues Pinheiro dos Santos, Ana Carla Mondek Rampazzo: Conducted the data collection, drafted the initial manuscript, and critically reviewed the manuscript. Fábio de Rezende Pinna and Richard Louis Voegels: Prof. Voegels and Prof. Pinna conceptualized and designed the study, and critically reviewed the manuscript. Richard L. Doty: Professor Doty provided statistical guidance and critically revised and reviewed the manuscript. All authors approved the final manuscript as submitted.
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: RLD is a consultant to Eisai Co, Ltd, Merck Pharmaceuticals, the Michael J. Fox Foundation for Parkinson’s Research, and Johnson & Johnson. He receives royalties from Cambridge University Press, Johns Hopkins University Press, and John Wiley & Sons, Inc. He is president of, and a major shareholder in, Sensonics International, a manufacturer and distributor of smell and taste tests, including the test used in this study. No other authors have disclosures.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Programa de Pesquisa para o Sistema Único de Saúde (PPSUS)
ORCID iD: Marco Aurélio Fornazieri https://orcid.org/0000-0001-5213-2337
==== Refs
References
1 Moein ST Hashemian SM Tabarsi P Doty RL. Prevalence and reversibility of smell dysfunction measured psychophysically in a cohort of COVID-19 patients. Int Forum Allergy Rhinol. 2020;10 (10 ):1127-1135.32761796
2 Lechien JR Chiesa-Estomba CM Beckers E , et al . Prevalence and 6-month recovery of olfactory dysfunction: a multicentre study of 1363 COVID-19 patients. J Intern Med. 2021;290 :451-461. doi:10.1111/joim.13209 33403772
3 Vaira LA Hopkins C Petrocelli M , et al . Smell and taste recovery in coronavirus disease 2019 patients: a 60-day objective and prospective study. J Laryngol Otol. 2020;134 :703-709.32782030
4 Hannum ME Ramirez VA Lipson SJ , et al . Objective sensory testing methods reveal a higher prevalence of olfactory loss in COVID-19–positive patients compared to subjective methods: a systematic review and meta-analysis. Chem Senses. 2020;45(9):865-874.
5 Petrocelli M Cutrupi S Salzano G , et al . Six-month smell and taste recovery rates in coronavirus disease 2019 patients: a prospective psychophysical study. J Laryngol Otol. 2021;135 (5 ):436-441.33888166
6 Parma V Ohla K Veldhuizen MG , et al . More than smell – COVID-19 is associated with severe impairment of smell, taste, and chemesthesis. Chem Senses. 2020;45 :609-622.32564071
7 Addison AB Wong B Ahmed T , et al . Clinical olfactory working group consensus statement on the treatment of postinfectious olfactory dysfunction. J Allergy Clin Immunol. 2021;147 (5 ):1704-1719.33453291
8 Doty RL. Treatments for smell and taste disorders: a critical review. Handb Clin Neurol. 2019;164 :455-479.31604562
9 Wu Z McGoogan JM. 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. 2020;323 (13 ):1239-1242.32091533
10 Doty RL Shaman P Dann M. Development of the University of Pennsylvania Smell Identification Test: a standardized microencapsulated test of olfactory function. Physiol Behav. 1984;32 (3 ):489-502.6463130
11 Fornazieri MA Doty RL Santos CA Pinna Fde R Bezerra TF Voegels RL. A new cultural adaptation of the University of Pennsylvania Smell Identification Test. Clinics. 2013;68 (1 ):65-68.23420159
12 Barbosa da Silva JL Doty RL Miyazaki JVMK , et al . Gustatory disturbances occur in patients with head and neck cancer who undergo radiotherapy not directed to the oral cavity. Oral Oncol. 2019;95 :115-119.31345378
13 Soter A Kim J Jackman A Tourbier I Kaul A Doty RL. Accuracy of self-report in detecting taste dysfunction. Laryngoscope. 2008;118 (4 ):611-617.18182967
14 Hilbe JM. Logistic Regression Models. CRC Press; 2017.
15 Moein ST Hashemian SM Mansourafshar B Khorram-Tousi A Tabarsi P Doty RL. Smell dysfunction: a biomarker for COVID-19. Int Forum Allergy Rhinol. 2020;10 :944-950. doi:10.1002/alr.22587 32301284
16 González C García-Huidobro FG Lagos AE , et al . Prospective assessment of smell and taste impairment in a South-American coronavirus disease 2019 (COVID-19) cohort: association with the need for hospitalization and reversibility of dysfunction. Int Forum Allergy Rhinol. 2021;11 (8 ):1273-1277.33848404
17 Leedman SR Sheeraz M Sanfilippo PG , et al . Olfactory dysfunction at six months after coronavirus disease 2019 infection. J Laryngol Otol. 2021;135 (9 ):839-843.34348821
18 Boscolo-Rizzo P Menegaldo A Fabbris C , et al . Six-month psychophysical evaluation of olfactory dysfunction in patients with COVID-19. Chem Senses. 2021;46 :bjab006.33575808
19 Ercoli T Masala C Pinna I , et al . Qualitative smell/taste disorders as sequelae of acute COVID-19. Neurol Sci. 2021;42 (12 ):4921-4926.34557966
20 Landis BN Hummel T Hugentobler M Giger R Lacroix JS. Ratings of overall olfactory function. Chem Senses. 2003;28 (8 ):691-694.14627537
21 London B Nabet B Fisher AR White B Sammel MD Doty RL. Predictors of prognosis in patients with olfactory disturbance. Ann Neurol. 2008;63 (2 ):159-166.18058814
22 Niklassen AS Draf J Huart C , et al . COVID-19: recovery from chemosensory dysfunction. A multicentre study on smell and taste. Laryngoscope. 2021;131 (5 ):1095-1100.33404079
23 Cao AC Nimmo ZM Mirza N Cohen NA Brody RM Doty RL. Objective screening for olfactory and gustatory dysfunction during the COVID-19 pandemic: a prospective study in healthcare workers using self-administered testing. World J Otorhinolaryngol Head Neck Surg. 2022;8 :249-256.
24 Vaira LA Deiana G Fois AG , et al . Objective evaluation of anosmia and ageusia in COVID-19 patients: single-center experience on 72 cases. Head Neck. 2020;42 (6 ):1252-1258.32342566
25 Petrocelli M Ruggiero F Baietti AM , et al . Remote psychophysical evaluation of olfactory and gustatory functions in early-stage coronavirus disease 2019 patients: the Bologna experience of 300 cases. J Laryngol Otol. 2020;134 (7 ):571-576.32605666
26 Hooper WC. The relationship between inflammation and the anticoagulant pathway: the emerging role of endothelial nitric oxide synthase (eNOS). Curr Pharm Des. 2004;10 (8 ):923-927.15032695
27 Bryce C Grimes Z Pujadas E , et al . Pathophysiology of SARS-CoV-2: the Mount Sinai COVID-19 autopsy experience. Mod Pathol. 2021;34 (8 ):1456-1467. doi:10.1038/s41379-021-00793-y 33795830
28 Lee M-H Perl DP Nair G , et al . Microvascular injury in the brains of patients with Covid-19. N Engl J Med. 2021;384 (5 ):481-483.33378608
29 Xydakis MS Albers MW Holbrook EH , et al . Post-viral effects of COVID-19 in the olfactory system and their implications. Lancet Neurol. 2021;20 (9 ):753-761.34339626
30 McBane RD Torres Roldan VD Niven AS , et al . Anticoagulation in COVID-19: a systematic review, meta-analysis, and rapid guidance from Mayo Clinic. Mayo Clin Proc. 2020;95 :2467-2486.33153635
31 Billett HH Reyes-Gil M Szymanski J , et al . Anticoagulation in COVID-19: effect of enoxaparin, heparin, and apixaban on mortality. Thromb Haemost. 2020;120 (12 ):1691-1699.33186991
32 Vaira LA Hopkins C Petrocelli M , et al . Efficacy of corticosteroid therapy in the treatment of long- lasting olfactory disorders in COVID-19 patients. Rhinology. 2021;59 :21-25.33290446
33 Daniel S Bon L Konopnicki D , et al . Efficacy and safety of oral corticosteroids and olfactory training in the management of COVID - 19 - related loss of smell. Eur Arch Otorhinolaryngol. 2020;278 :3113-3117.
34 Chivese T Musa OAH Hindy G , et al . Efficacy of chloroquine and hydroxychloroquine in treating COVID-19 infection: a meta-review of systematic reviews and an updated meta-analysis. Travel Med Infect Dis. 2021;43 :102135.34265436
35 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 (6 ):490-498.34269813
36 Rodrigues C Freitas-Santos RS Levi JE , et al . Hydroxychloroquine plus azithromycin early treatment of mild COVID-19 in an outpatient setting: a randomized, double-blinded, placebo-controlled clinical trial evaluating viral clearance. Int J Antimicrob Agents. 2021;58 :106428.34454044
37 Padhy BM Mohanty RR Das S Meher BR. Therapeutic potential of ivermectin as add on treatment in COVID 19: a systematic review and meta-analysis. J Pharm Pharm Sci. 2020;23 :462-469.33227231
38 Ugurlu BN Akdogan O Yilmaz YA , et al . Quantitative evaluation and progress of olfactory dysfunction in COVID-19. Eur Arch Otorhinolaryngol. 2021;278 (7 ):2363-2369. doi: 10.1007/s00405-020-06516-4 33385250
39 Shahrvini B Prajapati DP Said M , et al . Risk factors and characteristics associated with persistent smell loss in coronavirus disease 2019 (COVID-19) patients. Int Forum Allergy Rhinol. 2021;11 (8 ):1280-1282.33864717
40 Meini S Suardi LR Busoni M Roberts AT Fortini A. Olfactory and gustatory dysfunctions in 100 patients hospitalized for COVID-19: sex differences and recovery time in real-life. Eur Arch Otorhinolaryngol. 2020;277 (12 ):3519-3523.32500326
41 McDonald ST Bolliet DA Hayes JE. Chemesthesis: Chemical Touch in Food and Eating. Wiley-Blackwell; 2016.
42 Boscolo-Rizzo P Hummel T Hopkins C , et al . High prevalence of long-term olfactory, gustatory, and chemesthesis dysfunction in post-COVID-19 patients: a matched case-control study with one-year follow-up using a comprehensive psychophysical evaluation. Rhinology. 2021;59 (6 ):517-527.34553706
| 36482672 | PMC9742352 | NO-CC CODE | 2022-12-14 23:30:05 | no | Ann Otol Rhinol Laryngol. 2022 Dec 8;:00034894221138485 | utf-8 | Ann Otol Rhinol Laryngol | 2,022 | 10.1177/00034894221138485 | oa_other |
==== Front
RAE
sprae
Review & Expositor
0034-6373
2052-9449
SAGE Publications Sage UK: London, England
10.1177/00346373221136219
10.1177_00346373221136219
Thematic Words · · ·
Child health and COVID-19: How Mark 10 can inform a Christian ethic
Giunta-Stibb Hannah University of Rochester, USA
Stibb Joshua Evangelical Lutheran Church in America, USA
Hannah Giunta-Stibb, c/o Joshua Stibb, 3825 E. Henrietta Rd, Henrietta, Rochester, NY 14467, USA. Email: [email protected]
5 2022
5 2022
5 2022
119 1-2 7685
© The Author(s) 2022
2022
Review & Expositor, Inc
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
The COVID-19 pandemic created unprecedented challenges for children and families. While most of the public debate surrounding the pandemic naturally focused on mainstream concerns, vulnerable groups, including children, with unique concerns were pushed to the periphery. The fact that COVID-19 continues to impact these vulnerable groups gives Christians an opportunity to right past wrongs. In this article, we first describe the biblical priority Jesus gives to children as members of God’s kingdom by exploring Mark 10:13–16. We then highlight specific ways in which the consequences of public responses to COVID-19 disproportionately burdened children. Finally, we present two case studies through which we reimagine how Christians can respond to the collateral impacts of COVID-19 on children in a more biblically faithful manner.
children
COVID-19
kingdom of God
Mark 10
vulnerability
typesetterts1
==== Body
pmcIntroduction
The COVID-19 pandemic has caused a fundamental shifting of social priorities and a restructuring of social institutions. Nowhere is this shift more impactful than in the lives of children growing up amid this pandemic. Scripture is clear that Christians are to prioritize the needs and protect the rights of vulnerable populations, especially children. Christians are specifically called to think first of the needs of others rather than their own fears and needs. Unfortunately, some of the voices speaking loudest about children during the pandemic rarely considered children’s unique needs and conversations frequently devolved into petty political fights rather than thoughtful analyses of how to mitigate the disproportionate impact the pandemic might have. While some groups remained mindful of children, the overwhelming public attention was paid to the impacts the pandemic had on adults and their economic and social concerns. One does not have to look far for evidence. Consider a February 2022 newspaper article that described the increase in disruptions at Florida school board meetings. The article describes cases in which audience members had to be escorted off school premises en masse and in which some attendees were arrested.1 These behaviors disrupted school board meetings and prevented discussion of issues impacting children. Rather than showing up to discuss critical issues that affect children and families, the protesters showed up to fight for their own political objectives. As the newspaper article makes clear, these activities were certainly not confined to the state of Florida.
COVID-19 did not generally have immediate health impacts on children. For instance, a 2020 review noted that the average age of COVID patients early in the pandemic before the availability of a vaccine and the emergence of new therapies was 50, and most children did not have severe symptoms.2 COVID-19 had a variety of deleterious impacts on children, however, that will last far beyond the pandemic years. The purpose of this reflection is, first, to consider how the biblical call to protect children influences the modern understanding of how Christians ought to respond to threats to children’s health and well-being. Second, we consider specific threats to children’s health and well-being during the COVID-19 pandemic, including how COVID-19 disrupted routine pediatric health care, services for children with disabilities, educational progress, mental wellbeing, and the social safety net. Finally, we consider two case studies through which we imagine how Christians might respond to the needs of children in a more biblically faithful way. The pandemic is not behind us, and Christians have opportunities for more faithful action moving forward.
The value of children in Mark 10
From the first calls for careful stewardship in Genesis to the final promise of total salvation in Revelation, the biblical worldview of God’s people requires the faithful to love and support each other, especially vulnerable, marginalized, and oppressed communities. Among the most regularly identified communities are aliens/outsiders, widows, the poor, and children. When identifying the scriptural calls for the protection and welfare of children, the applicable verses are too numerous to list. For the sake of brevity, we restrict our view to Mark 10:13–16, one of Jesus’s foundational teachings about children. In this passage, Jesus takes the opportunity to teach his disciples and followers about the kingdom of God in a minor social interaction with some seemingly insignificant children.
In this text, Jesus welcomes children to him against the protestations of the disciples. In this act, Jesus is doing something new and unexpected.3 By making the children the paradigmatic recipient of the kingdom of God (v. 14), Jesus is lifting up everything for which childhood stands as something to be celebrated, emulated, and protected. An important distinction between the ancient world and contemporary postenlightenment thinking must also be made. When Jesus tells his followers to be like a child, Jesus is not identifying the qualities of purity, innocence, or sweetness as most contemporary listeners would likely assume. Instead, Jesus is highlighting their vulnerability, marginalization, and utter powerlessness.4
In the time and place of Jesus, honor and shame were the most important components for determining behavior. People would be excited to welcome someone of high status into their midst because it would increase their own honor and social equity, but children were of low status. Children were yet-to-be adults without true personhood, autonomy, and little societal value beyond their potential. No perceptible value existed in welcoming children in the midst of a growing ministry that would supposedly need honor to grow and thrive.5 So Jesus’s lifting up children as the exemplar of the kingdom of God flies in the face of that which society valued as most important (i.e., the fully matured and responsible adult man as head of family).6 This reversal of expectation is exactly what Jesus is trying to teach his disciples: to make the insignificant indispensable, to create a new center on the margins of society, and to value and love what society deems unimportant. In short, Jesus is saying children need to be loved, protected, and celebrated as important members of the community, if one is going to be his follower. Jesus’s saying that the kingdom of God belongs to the little children (and all they stand for) is nothing less than a complete reversal of ancient society and an imperative for all future Christian ethics. This imperative is so influential and world-altering that it would lead some scholars to assert that Christian philosophy basically invented the postenlightenment concept of children, including an understanding of development, autonomy, and personhood.7
To the contemporary reader, this shift in attitude is subtle, if not completely unremarkable. In the postmodern world, it is easy to look back at the antiquated views of the first-century families and sneer while taking pride in the more enlightened perspective through which children are supposedly respected, understood, and cherished. In this day and age, one can commonly hear parents declare “children are the center of my world,” while scheduling their entire life around the needs of their children, or hear churches assert that “children are the future” while catering nearly all their ministries toward family and children’s ministry. A sad irony is apparent if one considers how often these attitudes are purported, yet, much like the first-century culture, children are often reduced to objects and valued for their potential in achieving an adult’s goals. Anyone familiar with children’s ministry knows this struggle. A common example is children’s education programming: often some parents are not concerned in the least with their child’s religious education but are instead looking for a free babysitter. Of course, some parents need a short break, and many churches are happy to provide that support as a ministry to the parents; if, however, the true priority is the faithful education of the children, how different would that children’s ministry look? And what amazing benefits would that entire community receive if children were valued right now as full persons, instead of as adults-in-waiting?
The impact of the COVID-19 pandemic on children
Ideological commitments are often tested in times of crisis when people must act on the abstract values they profess. The COVID-19 pandemic required quick actions in response to evolving conditions. When confronted with uncertainty and risk, the decisions people make reveal the true condition of their hearts and highlight what they actually cling to as most important. This phenomenon is not new. In Matt 6:21 Jesus tells the crowds, “Where your treasure is, there your heart will be also.” The allotment of one’s resources reveals one’s true priorities. In an ideal world, the desire of one’s heart would correspond with what one intellectually professes as one’s values. But, if that were the case, Jesus would have said, “Where your heart is there your treasure will be also,” and many struggles caused by self-prioritization and greed would be gone in an instant. Alas, in this broken creation, people are faced with the harsh reality of the harmful status quo. With that observation in mind, the opportunity arises to look back and to see where the community’s professed values were misaligned with the actual outcomes. Here, we consider how children have been impacted by COVID-19 and the role decisions about resource allocation played in these outcomes.
COVID-19 had a negative impact on access to pediatric healthcare services. With clinics and hospitals closed or repurposed for treating COVID patients, children were unable to access routine preventive care or ongoing care for complex medical needs. The American Academy of Pediatrics recommends all children have frequent visits with healthcare providers to receive standard immunizations against serious infectious diseases and to identify medical and developmental problems as early as possible.8 These visits occur every few months in the first 2 years of life and then continue at least annually until young adulthood. Early identification of health problems provides an opportunity to mitigate negative impacts on future health and development. Given the fact that most children receive at least some routine care, measures of vaccination administration serve as an early surrogate marker of children’s access to health care. Early in the pandemic, researchers noticed a precipitous decline in the administration of routine childhood vaccinations. Using provider ordering data from 10 high-performing geographical jurisdictions, Murthy et al. found that vaccine administration was significantly lower across all vaccine categories from March to May 2020 compared to 2018 and 2019. Murthy and colleagues also noted that, after initial lockdown restrictions were lifted, vaccination rates increased slightly but were not sustained.9 A follow-up case–control study involving eight major health systems across the United States demonstrated declines in vaccine delivery from January to October 2020 compared to 2019. Vaccine rates were disproportionately low among Black, non-Hispanic children.10 In a systematic analysis of data from around the globe, SeyedAlinaghi et al. found evidence of reduced vaccination rates in many countries during the COVID-19 pandemic and suggest an urgent need for catch up vaccination.11 Reduced vaccination rates leave children vulnerable to preventable infections. Without access to appropriate preventive care, children are at risk in the coming years of excess morbidity and mortality.
Every child needs access to preventive care. Some children also need ongoing therapeutic services that have also been disrupted by the COVID-19 pandemic. The lack of access to timely care impacted the time to diagnosis of serious health conditions during the pandemic as well as the ongoing provision of treatment for illnesses identified before the pandemic. With so much focus on the needs of COVID patients, other groups of patients who could not speak for themselves, including children, received limited attention even when they had identified health needs. Children thankfully tend to be healthy generally, but delayed diagnosis of serious illnesses can be life-threatening. Pediatricians in New Zealand reported at least 55 instances in which acutely needed care was delayed or compromised and most children had moderately severe or severe health consequences as a result of this suboptimal care.12 Pediatricians in the United Kingdom noted that children presented to emergency departments with more severe symptoms, and this late presentation to care was a contributing factor in at least nine cases of pediatric deaths secondary to malignancy and sepsis.13 Pediatricians in Italy highlighted the fact that children presented later for emergency treatment despite having symptoms that concerned their caregivers and sometimes after receiving an incomplete evaluation via telehealth.14
Consider two feared diseases in children, cancer and Type I diabetes, that, if not detected early, can prove fatal. Physicians at the Children’s Hospital of Philadelphia and at Lucile Packard Children’s Hospital at Stanford reported five instances between March and May 2020 in which children presented in extremis before subsequently being diagnosed with cancer. These children required care in the ICU because they were so ill by the time they presented to the hospital.15 A review article prepared by the American Society of Clinical Oncology highlighted restrictions placed on pediatric cancer care during the pandemic and specifically pointed out barriers to continuing clinical trials. Many children with cancer receive some of their care through collaborative clinical trials that recruit subjects across the country.16 This decrease in availability in turn slowed the availability of new therapies and interrupted the treatment of current pediatric patients. In addition, an alarming increase occurred in the number of children with Type I diabetes presenting in diabetic ketoacidosis (DKA). This condition is a life-threatening complication of diabetes in which blood sugar levels rise precipitously and acid builds up in the bloodstream. If not promptly treated, diabetic ketoacidosis can result in coma, brain swelling, and death. An increase in pediatric Type I diabetes patients presenting in severe DKA was seen in many countries, including Poland,17 Israel,18 Italy,19 and the United Kingdom.20
Children with disabilities and complex medical needs have fared particularly poorly since the start of the pandemic. These children receive many of their supportive services through the healthcare and educational systems. Therefore, the closure of health care facilities and schools predictably had a disproportionate impact on these children and their families. As Houtrow et al. point out, children with disabilities were marginalized before the pandemic and were subsequently neglected during the pandemic. They liken the experience of children with disabilities and their families to groups weathering the same storm together but with differently outfitted boats. Children with disabilities immediately lost home health services, therapies, and access to supportive services at school and in their community. They then faced the harsh reality of possible discrimination due to their disabilities as local health leaders struggled to triage COVID-19 patients. Standard solutions for continued services during COVID-19, including online schooling and telehealth, left children with disabilities behind because the solutions were designed for typically developing children without consideration of those students with special needs.21
Routine care that children with disabilities receive is not as optional as it may first appear. Real harm was done to these children. A survey of over 300 parents of children with neurodevelopmental disabilities revealed that 64.5% of children had worsening of their neurological disorders or psychiatric comorbidities. Three-quarters had diminished health and well-being secondary to COVID-19 and associated restrictions.22 Many children with disabilities require daily therapies to maintain their comfort and mobility and to optimize their long-term developmental outcomes. For instance, a survey of caregivers for children with motor impairment found that, although 90% of these children received therapies prepandemic, only around 50% received regular therapies during the COVID pandemic. A lack of access to routine therapies negatively impacted children’s mobility and physical condition and placed stress on caregivers.23
Other children need time-sensitive services. Consider the situation of children with cochlear implants. Among these children, parents reported difficulty accessing speech and language services. Children with hearing loss who receive cochlear implants need a connection to speech processors that provide continuous auditory stimulation. The lack of continuous auditory stimulation can permanently alter speech and language development. Parents reported difficulty with remote therapies and equipment malfunction and breakdown that put children’s speech development in jeopardy.24
No issue was more divisive during the COVID-19 pandemic than school closure and reopening. School closures profoundly impacted all children’s mental and educational development. Even children in ideal circumstances demonstrated diminished educational achievement that will have continuing impacts on their eventual academic trajectories. Consider the Netherlands as one of the best-case scenarios. The Netherlands enjoys equitable school funding, and the nation shut down schools for a relatively brief period of time. The Dutch also have widespread broadband internet capabilities, and most children have access to the necessary technology for distance learning. Still, Dutch children lost about one-fifth of a school year in educational progress, the exact amount of time lockdowns forced schools to remain closed.25 In countries with more disparate baseline educational outcomes, disadvantaged students fared far more poorly than their advantaged peers. Children with learning issues were also more negatively impacted by school closures.26
Closing down major social health and educational institutions profoundly influenced mental and emotional health. The mental health of children and families was often neglected, and this occurred despite previous research during other epidemics that suggested social isolation and lockdown when combined with adverse childhood experiences was associated with an increased risk of mental health problems, including anxiety, post-traumatic stress disorder (PTSD), and depression, and longer-term risks of poor health in adulthood, substance abuse, cognitive impairment, and other noncommunicable diseases.27 During the COVID-19 pandemic children and adolescents experienced significant increases in anxiety, depression, disturbances in sleep and appetite, and impairment in social interactions.28 Newlove-Delgado et al. reported that youth in England were unable to access mental health care. They frequently lacked basic resources to maintain their normal routines, including access to places to study and access to the internet.29
The COVID-19 pandemic occurred against a backdrop of disadvantage and inequality. Children were more likely to face hunger and other adverse conditions during lockdown. For example, many large urban school districts had to scramble to deliver meals to hungry children. American children often receive the majority of their meals at school. The four large urban school districts of Houston, Los Angeles, Chicago, and New York City had to restructure their food service delivery systems completely. When shutdowns were imminent in 2020, Congress provided increased funding for school lunch programs, but these resources were not enough for equitable distribution of food to hungry children. School districts frequently had to change distribution plans and were unable to distribute food in all high-need neighborhoods. Families lacking access to transportation were limited in what they could bring home at any one visit. Nutritional quality also was suboptimal in some districts. Distribution strategies frequently changed, and districts struggled to provide accessible food. Children’s nutritional needs were not always met outside the school setting.30 Without usual social support services, children also were at increased risk of maltreatment. March through July 2020 saw an increase in maltreatment with a notable increase in emotional and mental neglect compared to baseline.31 While adults made decisions about institutions, children were forced to live with the consequences of these decisions.
In summary, responses to the COVID-19 pandemic frequently considered the needs of the dominant majority. Facing the prospect of an onslaught of COVID patients, health systems limited what they considered to be elective services. Yet, children, vulnerable individuals who had no say in these decisions, frequently had to go without necessary services. Adults often forget that pediatric services may be far more time-sensitive than similar adult services. While everyone had an opinion about school plans, few news outlets covered how to maintain essential services and supports for children. With children isolated at home, their needs were considered even less than they might usually be considered.
Case studies
Although no one can turn back the clock and change earlier responses to the pandemic, COVID-19 is likely here to stay. The pandemic has also highlighted previous inequalities (childhood hunger, inadequate services for children with disabilities, etc.). Christians will have future opportunities to right past wrongs. Christians must therefore consider how to respond to future circumstances in a more biblically faithful way. Here, we consider two case studies. First, we look at school closure and how Christians can advocate for the needs of vulnerable children. Second, we consider how Christians can support children with disabilities and their families.
Children are currently returning to in-person learning. Rising COVID numbers may, however, prompt a return to distance or hybrid learning strategies. As mentioned earlier, extreme behavior is becoming more common at school board meetings in which children desperately need a voice in decisions about educational policies. We first examine some current examples of how debates about COVID actually played out and then consider how Christians can participate in civic discourse while advocating for children and prioritizing their needs in a biblically faithful way.
As reported by ABC News, communities around the country struggled with how to mitigate the spread of COVID and educate students. Adult opinions about what students needed frequently dominated discussions, rather than viewpoints on practical measures to mitigate inequalities related to COVID. In Arizona, the state banned masks in classrooms and required a return to in-person learning, even as case numbers rose. Arizona districts were required, without consideration of local needs or desires, to adopt these practices to remain eligible for financial incentives. Florida and Texas enacted similar mask bans but were challenged as case numbers rose by parents and districts. On the other side, Louisiana enacted a mask requirement due to a rapidly rising case count, but legislators faced angry crowds at public meetings.32
The problem with these reactions to school closure was not that people had passionate but disparate views. The real issue was that neither side considered the true needs of children. Those advocating against mask mandates did not consider how rising infection rates might impact teacher availability or how it might keep children out of the classroom. Children who had to frequently quarantine would miss significant amounts of instructional time and risk falling behind. Passionate antimask advocates also did not consider how the COVID-19 infection differentially impacted certain groups. Children with serious health conditions faced different risks than their healthier counterparts. As previously discussed, any need to close school doors with rising infection rates would disproportionately impact the services offered for disabled students. On the other side of the debate, those advocating in favor of masks and other COVID mitigation strategies frequently did not consider the harm to children from closing school doors and limiting access to public services. They were concerned about COVID but neglected the downstream impacts. Fear led to neglect of other competing objectives. In all these debates, nuance was lacking.
So, how might Christians respond? First, Christians must refocus on the needs of children rather than on their own opinions about the state of politics today. There is a need for both mitigation of COVID and also advocacy for children who need and deserve an education right now. Both increases in COVID infections and school closures negatively impact children. A more biblical, child-centered approach would select available COVID mitigation efforts that maximize benefits and minimize burdens for children while rejecting the idea that mitigation is not needed. Christians can also respond by bringing the needs of those most disadvantaged to the forefront and filling crucial service gaps. If schools need to shut down again, imagine the good Christians could do by providing healthy food, increasing access to technology, or supporting students through tutoring and mentoring.
Finally, one must acknowledge that no absolute Christian answer solves all the problems of mask mandates, school closures, and politics. No panacea or supreme resolution exists to solve the problems of every community, because infection numbers, population differences, and myriad other factors change constantly, even within the same communities. Instead, the journey is as important as the destination. The process by which Christians, especially ones centered in Mark 10, come to the conclusions about solutions is as important as the solutions themselves. Uncertainties lie at the heart of the pandemic. There is no linear way through. Rather, humankind is wading together through undulating waves of fear, discovery, and change. The Christian path should endeavor to uplift the Christ-centered needs of the vulnerable children above those waves of the unknown. What makes these stories of school closures and mask mandates so unbearable to retell is that so many parents, legislators, and administrators did not even seem to try to address these needs. What is fatal to a faithful Christian process is not a lack of authority to the law (natural or otherwise), but a lack of communal pedagogy fixed upon the powerless and vulnerable. Changing the locus solely to the plight of the children would calm the anxieties in the process, put the political needs of the adults in the backseat, and make way for a just and equitable solution. In short, if science has offered one consistent answer throughout the pandemic, it has been, “I don’t know. Let’s work hard to find an answer together.” Why could Christians not similarly respond, “I don’t know. Let’s work faithfully to find the right answer”?
Turning now to the second case, we previously considered the negative impacts COVID-19 mitigation measures had on children with disabilities and their families. Children with disabilities were routinely put on the proverbial back burner to make way for the needs of adult patients, and services critical to their health and well-being were seen as elective rather than essential. COVID has highlighted a collective lack of attention to disabled children, and Christians have an opportunity to reconsider the needs of these children for the future.
Christians can advocate for children with disabilities in biblically centered ways in at least two areas. First, they can acknowledge the reality that resource allocation decisions cannot be all or none. Of course, patients who are severely ill with COVID-19 deserve access to medical treatment. Reserving most or all beds for COVID patients, however, reduces the availability of critical pediatric services. Christians know they are to be good stewards and to prioritize the vulnerable, and this stewardship may mean allocating more resources to children even amid a resource crunch in other areas. Second, with the introduction of effective vaccines, Christians must consider how personal decisions to defer vaccination may inadvertently take resources away from the vulnerable. This statement is not a universal call for vaccination of those with medical or religious objections but a demand for acknowledging that preventive actions are one way to protect the vulnerable. Each individual must make their own decision but also must acknowledge the impact that decision has on resource allocation for others. Resources deployed to fight COVID come at a cost to other groups, at least in the current health care environment. The situation is akin to the constant theological struggle of identity. Is one’s Christian identity found inside themselves, in their own aspirations, their own works, their own salvation? Or is it instead found in the other, in the needs and wants of the other, the imago Dei just below the surface waiting to be revealed in relationships with each other?
Concluding remarks
In this reflection piece, we have considered how COVID-19 presents unique obstacles to child health and well-being. A closer look at Mark 10 reframes the understanding of children and the inherent value Jesus places on them. Although COVID-19 negatively impacted access to pediatric health care, services for children with disabilities, education progress, mental well-being, and the social safety net for disadvantaged children, Christians will have many future opportunities to begin to right past wrongs as society continues to grapple with the pandemic’s continuing course. We have presented two case studies in an attempt to reconsider how Christians can, moving forward, live out their biblical commitments to children. We urge readers to reconsider how they respond to politically charged conflicts involving children by maintaining a truly biblical focus on children as valued members of the kingdom of God, and not as mere pawns in adult disputes.
Author biographies
Hannah Giunta-Stibb is currently completing her neonatology and pediatric pulmonology fellowships in Rochester, New York. She was raised in Peoria, Illinois, and graduated from Bradley University with a BS in cell and molecular biology and a BA in psychology. She received her osteopathic medical degree, her PhD in philosophy with a focus in bioethics, and her MPH from Michigan State University. She then completed a pediatrics residency at Mayo Clinic. Her research focuses on pediatric bioethics with significant interests in the ethics of invasive technologies, family decision-making, end-of-life care, and health equity. When not at work, Giunta-Stibb enjoys spending time with her husband and sheltie puppy as well as spending time outdoors, reading, and cooking.
Joshua Stibb is an ordained pastor of the Evangelical Lutheran Church in America. He attended Wartburg College in Waverly, Iowa and Luther Seminary in St. Paul, Minnesota. He currently serves at an Episcopal church in Henrietta, New York. Previously he served congregations in Las Vegas, Nevada, Nebraska, and Wisconsin. Stibb has also served as director or board member of organizations and nonprofits focusing on immigration and refugee services, domestic violence, and urban food security. He enjoys spending time with his wife Hannah Giunta-Stibb and their dog Zephaniah. When they have the time, they enjoy visiting the Finger Lakes to kayak and fish.
1. Staff, “Disrupted Florida School Board Meetings Now Common in Age of COVID and Mask Debates,” Ocala Star-Banner, February 25, 2022, https://www.ocala.com/story/news/education/2022/02/25/school-board-florida-parental-choice-mandatory-covid-mask-mandates/6922775001/.
2. Luis Escosa-García, David Aguilera-Alonso, Cristina Calvo, Maria José Mellado, and Fernando Baquero-Artigao, “Ten Key Points about COVID-19 in Children: The Shadows on the Wall,” Pediatric Pulmonology 55.10 (2020): 2576–86.
3. Richard Swanson, Provoking the Gospel of Mark (Cleveland, OH: The Pilgrim Press, 2005), 220.
4. For an in-depth discussion of familial relationships and the identity of the church as surrogate kin, see Joseph H. Hellerman, The Ancient Church as Family (Minneapolis: Fortress, 2001).
5. Adam Winn, “Resisting Honor: The Markan Secrecy Motif and Roman Political Ideology,” JBL 133.3 (2013): 583–601 (586).
6. Reider Aasgaard, “Children in Antiquity and Early Christianity: Research History and Central Issues,” Familia 33 (2006): 23–46.
7. For a full treatment of children in the early church and the notion of childhood as a Christian concept, see O. M. Bakke, When Children Became People: The Birth of Childhood in Early Christianity (Minneapolis: Fortress, 2005).
8. “AAP Schedule of Well-Child Care Visits,” HealthyChildren.org, American Academy of Pediatrics, August 30, 2022, https://www.healthychildren.org/English/family-life/health-management/Pages/Well-Child-Care-A-Check-Up-for-Success.aspx.
9. B. P. Murthy, E. Zell, K. Kirtland, N. Jones-Jack, L. T. Harris, C Sprague, J. Schultz, et al., “Impact of the COVID-19 Pandemic on Administration of Selected Routine Childhood and Adolescent Vaccinations—10 U.S. Jurisdictions, March–September 2020,” Morbidity and Mortality Weekly Report (MMWR) 70.23 (2021): 840–45, https://doi.org/10.15585/mmwr.mm7023a2.
10. M. B. Desilva, J. Haapala, G. Vazquez-Benitez, M.F. Daley, J. D. Nordin, N. P. Klein, M. L. Henninger, et al., “Association of the COVID-19 Pandemic with Routine Childhood Vaccination Rates and Proportion Up to Date with Vaccinations Across 8 US Health Systems in the Vaccine Safety Datalink,” JAMA Pediatrics 176.1 (2022): 68–77.
11. S. A. SeyedAlinaghi, A. Karimi, H. Mojdeganlou, S. Alilou, S. P. Mirghaderi, T. Noori, A. Shamsabadi, et al., “Impact of COVID-19 Pandemic on Routine Vaccination Coverage of Children and Adolescents: A Systematic Review,” Health Science Reports 5.2 (2022): e00516.
12. Mavis Duncanson, Benjamin J. Wheeler, Timothy Jelleyman, Stuart R. Dalziel, and Peter McIntyre, “Delayed Access to Care and Late Presentations in Children during the COVID-19 Pandemic New Zealand-Wide lockdown: A New Zealand Paediatric Surveillance Unit Study,” Journal of Paediatrics and Child Health 57.10 (2021): 1600.
13. Richard M. Lynn, Jacob L. Avis, Simon Lenton, Zahin Amin-Chowdhury, and Shamez N. Ladhani, “Delayed Access to Care and Late Presentations in Children during the COVID-19 Pandemic: A Snapshot Survey of 4075 Paediatricians in the UK and Ireland,” Archives of Disease in Childhood 106.2 (2021): e8.
14. M. Lazzerini, E. Barbi, A. Apicella, F. Marchetti, F. Cardinale, and G. Trobia, “Delayed Access or Provision of Care in Italy Resulting from Fear of COVID-19,” The Lancet Child & Adolescent Health 4.5 (2020): e10–e11.
15. Y. Y. Ding, S. Ramakrishna, A. H. Long, C. A. Phillips, R. Montiel-Esparza, C. J. Diorio, L. C. Bailey, et al., “Delayed Cancer Diagnoses and High Mortality in Children during the COVID-19 Pandemic,” Pediatric Blood & Cancer 67.9 (2020): e28427.
16. Daniel C. Moreira, Gerard C. Millen, Stephen Sands, Pamela R. Kearns, and Douglas S. Hawkins, “The Care of Children with Cancer during the COVID-19 Pandemic,” American Society of Clinical Oncology Educational Book 41 (2021): e305–e314.
17. Katarzyna Dżygało, Jedrzeg Nowaczyk, Alicja Szwilling, and Agnieszka Kowalska, “Increased Frequency of Severe Diabetic Ketoacidosis at Type 1 Diabetes Onset Among Children during COVID-19 Pandemic Lockdown: An Observational Cohort Study,” Pediatric Endocrinology, Diabetes, and Metabolism 26.4 (2020): 167–75.
18. S. Goldman, O. Pinhas-Hamiel, A. Weinberg, A. Auerbach, A. German, A. Haim, A. Zung, et al., “Alarming Increase in Ketoacidosis in Children and Adolescents with Newly Diagnosed Type 1 Diabetes during the First Wave of the COVID-19 Pandemic in Israel,” Pediatric Diabetes 23.1 (2022): 10–18.
19. Lazzerini et al., “Delayed Access.”
20. Lynn et al., “Delayed Access to Care.”
21. Amy Houtrow, Debbi Harris, Ashli Molinero, Tal Levin-Decanini, and Christopher Robichaud, “Children with Disabilities in the United States and the COVID-19 Pandemic,” Journal of Pediatric Rehabilitation Medicine 13.3 (2020): 415–24.
22. A. Masi, A. Mendoza Diaz, L. Tully, S. I. Azim, S. Woolfenden, D. Efron, and V. Fapen, “Impact of the COVID-19 Pandemic on the Well-Being of Children with Neurodevelopmental Disabilities and Their Parents,” Journal of Paediatrics and Child Health 57.5 (2021): 631–36.
23. E. N. Sutter, L. S. Francis, S. M. Francis, D. H. Lench, S. T. Nemanich, L. E. Krach, T. Sukal-Moulton, et al., “Disrupted Access to Therapies and Impact on Well-Being during the COVID-19 Pandemic for Children with Motor Impairment and Their Caregivers,” American Journal of Physical Medicine & Rehabilitation 100.9 (2021): 821.
24. Mohammed Ayas, Ahmad Mohd Haider Ali Al Amadi, Duaa Khaled, and Ahmad Munzer Alwaa, “Impact of COVID-19 on the Access to Hearing Health Care Services for Children with Cochlear Implants: A Survey of Parents,” F1000Research 9 (2020): 690.
25. Per Engzell, Arun Frey, and Mark D. Verhagen, “Learning Loss Due to School Closures during the COVID-19 Pandemic,” Proceedings of the National Academy of Sciences of the United States of America 118.17 (2021): e2022376118.
26. Svenja Hammerstein, Christoph König, Thomas Dreisörner, and Andreas Frey, “Effects of COVID-19-Related School Closures on Student Achievement—A Systematic Review,” Frontiers in Psychology 12 (2021): 4020.
27. Liubiana Arantes de Araújo, Cássio Frederico Veloso, Matheus de Campos Souza, João Marcos Coelho de Azevedo, and Giulio Tarro, “The Potential Impact of the COVID-19 Pandemic on Child Growth and Development: A Systematic Review,” Jornal De Pediatria 97.4 (2021): 369.
28. S. Meherali, N. Punjani, S. Louie-Poon, K. A. Rahim, J. K. Das, R. A. Salam, and Z. S. Lassi, “Mental Health of Children and Adolescents Amidst Covid-19 and Past Pandemics: A Rapid Systematic Review,” International Journal of Environmental Research and Public Health 18.7 (2021): 3432.
29. T. Newlove-Delgado, S. McManus, K. Sadler, S. Thandi, T. Vizard, C. Cartwright, and T. Ford, “Child Mental Health in England Before and during the COVID-19 Lockdown,” The Lancet Psychiatry 8.5 (2021): 353.
30. G. M. McLoughlin, J. A. McCarthy, J. T. McGuirt, C. R. Singleton, C. G. Dunn, and P. Gadhoke, “Addressing Food Insecurity through a Health Equity Lens: A Case Study of Large Urban School Districts during the COVID-19 Pandemic,” Journal of Urban Health: Bulletin of the New York Academy of Medicine 97.6 (2020): 759.
31. S. Sharma, D. Wong, J. Schomberg, C. Knudsen-Robbins, D. Gibbs, C. Berkowitz, and T. Heyming, “COVID-19: Differences in Sentinel Injury and Child Abuse Reporting during a Pandemic,” Child Abuse & Neglect 116 (2021): 104990.
32. Marlene Lenthang, “How School Board Meetings Have Become Emotional Battlegrounds for Debating Mask Mandates,” ABC News, August 29, 2022, https://abcnews.go.com/US/school-board-meetings-emotional-battlegrounds-debating-mask-mandates/story?id=79657733.
| 0 | PMC9742502 | NO-CC CODE | 2022-12-14 23:30:02 | no | 2022 May; 119(1-2):76-85 | utf-8 | null | null | null | oa_other |
==== Front
RAE
sprae
Review & Expositor
0034-6373
2052-9449
SAGE Publications Sage UK: London, England
10.1177/00346373221133718
10.1177_00346373221133718
Thematic Words · · ·
Unto the least of these: Caring for the vulnerable in the time of COVID
White Frederick J. III
Willis-Knighton Medical Center, USA
Frederick J. White III, Willis-Knighton Medical Center, Shreveport, LA 71103, USA. Email: [email protected]
5 2022
5 2022
5 2022
119 1-2 5063
© The Author(s) 2022
2022
Review & Expositor, Inc
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
As the COVID-19 pandemic initially unfolded in early 2020, medical systems were rapidly overwhelmed with critically ill patients. Intensive care resources were strained and, in some cases, insufficient. Concepts of triage and allocation of life-saving resources, once only hypothetical, were called into action. Vulnerable elderly, chronically ill, and disabled patients found themselves subject to protocols and guidelines that singled them out for disparate access to treatments. In this article, I overview the historical background of the early COVID-19 crisis, frontline triage guidelines in Italy and New York City, the conceptual nature of triage, the problematic practice of reallocation, the ethical principles that were challenged, how Judeo-Christian teachings inform these issues, and conflicts of physician duties with attendant moral distress. I close with a set of normative guideline statements that could help define a path through the extreme scarcities of a catastrophic pandemic crisis surge.
COVID-19
intensive care
persons with disability
reallocation
triage
ventilator
typesetterts1
==== Body
pmcIntroduction
The illness now simply called COVID has changed our world in a myriad of ways. Society has relearned the long-forgotten lesson that sometimes illness affects not only a person but an entire population. Society has seen that even the wealthiest nations and the most advanced medical systems may be driven to their knees by the powers of pestilence. Society has accepted that sometimes a crisis becomes not an event, but a way of life.
As the pandemic initially unfolded, medical systems were overwhelmed with critically ill patients. Intensive care resources were strained and, in some cases, insufficient. Concepts of triage and allocation of life-saving resources, once only hypothetical, were called into action. Vulnerable elderly, chronically ill, and disabled patients found themselves subject to protocols and guidelines that singled them out for disparate access to treatments. In this article, I overview the historical background of the early COVID-19 crisis, frontline triage guidelines in Italy and New York City, the conceptual nature of triage, the problematic practice of reallocation, the ethical principles that were challenged, how Judeo-Christian teachings inform these issues, and conflicts of physician duties with attendant moral distress. I close with a set of normative guideline statements that could help define a path through the extreme scarcities of a catastrophic pandemic crisis surge.
The historical background
In late December 2019, physicians in Wuhan, China, began to see patients hospitalized with a pneumonia of unknown origin, and by December 31 the outbreak prompted a bulletin from the Wuhan Municipal Health Commission, soon followed by involvement of the World Health Organization (WHO).1 The offending agent was quickly determined to be a virus of the coronavirus family and was named the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2).2 The highly communicable virus spread rapidly. The first case of COVID-19 illness in the United States was confirmed in Snohomish County, Washington, on January 20, 2020, in a traveler returning from Wuhan, China.3 The first case of COVID-19 in Italy was confirmed on February 20, 2020.4 By February 29, 2020, WHO recorded 6000 cases outside China in 53 countries, with significant localizations in the Republic of Korea and Italy.5 The global situation rapidly deteriorated, and, on March 11, 2020, with over 100,000 cases and 4200 deaths worldwide, the WHO declared COVID-19 to be a pandemic.6
At this point, the Lombardy region of northern Italy was in desperate straits, with physicians reporting an overwhelming surge in patients in need of intensive care and with regional resources at capacity. By mid-March, the Lombardy region, with a baseline intensive care unit (ICU) capacity of 724 ICU beds, had 1006 patients on advanced respiratory support.7 With daily ICU admissions growing exponentially, the physicians foresaw possible total collapse of the ICU system. Ominously, the Italian physicians warned, “Other health care systems should prepare for a massive increase in ICU demand during an uncontained outbreak of COVID-19.”8 Things were to become much worse.
The northern Italian epidemic peaked from mid-March to mid-April 2020, with data from the Italian government showing daily new cases in Lombardy during that time ranging from 1500 to 2500.9 Sixteen of each 100 cases required ICU admission for severe respiratory illness.10 Mortality among those hospitalized was 30%.11 By April 2020, Lombardy had over 62,000 cases with a case mortality rate of 18%.12 Hospitals ran out of beds, and patients, some on respiratory support devices, were in hallways. The number of mechanical ventilators to support respiration for the most critically ill patients was rapidly exhausted. In this dire situation, the physicians began to triage the patients, deciding who would and who would not get a ventilator, and in many of those circumstances who would live and who would die.13
In the United States, a similar catastrophe developed in New York City during the same time frame. There the first confirmed case of COVID-19 was on February 29, 2020.14 The outbreak exploded as cases grew exponentially. By the middle of March 2020, New York City had the highest case incidence rate in the United States, more than twice that of New Jersey and nearly five times that of Louisiana.15 The New York City experience rivaled that of Lombardy in intensity. By early April 2020, cumulative case rates were 915 per 100,000.16 In comparison, the July 2020 cumulative case incidence in Lombardy was 951 per 100,000.17
Parallel to the northern Italian experience, the New York City epidemic peaked from late-March to early-April 2020. During the last week of March, there were over 5100 new cases and 1500 new hospitalizations daily. By 1 June 2020, New York City had over 200,000 cases, with 27 of 100 requiring hospitalization and an overall case mortality rate of 9%.18 Mortality among those hospitalized was 36%.19 Hospital and ICU resources in New York City were quickly strained. On March 28, 2020, roughly 85% of both the 2011 ICU beds and the 20,330 hospital beds in the City were occupied.20 On April 5, at the peak of the epidemic, Mayor Bill DeBlasio indicated that New York City had 4000 patients on ventilators with a projected immediate need for 1000 additional ventilators, with fulfillment of most of that need expected from out-of-city sources.21 The ventilators arrived, the surge began to subside, and New York City narrowly skirted the breaking point.
Frontline triage guidelines
Despite actions of the Italian government to lock down affected regions, open additional hospital beds, and procure additional ventilators, the looming scarcity of intensive care resources prompted the Italian Society of Anesthesia, Analgesia, Resuscitation and Intensive Care (SIAARTI) to publish triage recommendations on March 6, 2020.22 To be applied under conditions of “extraordinary scarcity,” the recommendations rejected first-come, first-served triage and instead provided thatthe criteria for access to and discharge from the ICUs should include also principles of distributive justice and appropriate allocation of limited healthcare resources, in addition to clinical appropriateness and proportionality of care. As an extension of the principle of proportionality of care in a context of serious shortage of healthcare resources, we must aim at guaranteeing intensive treatments to patients with greater chances of therapeutic success favoring the “greatest life expectancy.”23
This structure placed the elderly, the weak, and the disabled under heightened scrutiny. In a related commentary, the authors noted that “frail elderly patients with severe comorbidities would likely have a more ‘resource consuming’ clinical course” and that the “age, comorbidities and functional status of any critically ill patient should be carefully evaluated.”24 Condoning a utilitarian life-years saved ethos, the recommendations explicitly permitted age limits:An age limit for the admission to the ICU may ultimately need to be set. The underlying principle would be to save limited resources which may become extremely scarce for those who have a much greater probability of survival and life expectancy, in order to maximize the benefits for the largest number of people.25
In addition to initial allocation of life-saving resources, the recommendations also endorsed withdrawal of previously allocated ventilators and ICU care to free up resources for other patients, a process sometimes termed “reallocation.” The language intimated that this decision was to be nonconsensual, to be communicated but not negotiated:Every admission to the ICU should be considered and communicated as an “ICU trial.” The appropriateness of life-sustaining treatments should be re-evaluated daily, considering the patient’s history, current clinical course, wishes, expected goals and proportionality of ICU care. When a patient is not responding to prolonged life-sustaining treatments, or severe clinical complications arise, a decision to withhold or withdraw further or ongoing therapies should not be postponed in a resource-limited setting during an epidemic.26
In the face of mounting public questions regarding triage, the Italian National Institute of Health (ISS), through its National Center for Clinical Excellence, published in January 2021 a “best clinical practices” document outlining triage principles for conditions of extreme scarcity of intensive care resources.27 This document was crafted by a workgroup of intensive care physicians and legal medicine experts and was heralded as the “reference standard.” It contained a series of statements pertaining to triage with a stated goal “to ensure life-sustaining treatments to as many patients as possible who may benefit from them.”28 These statements recommended use of objective clinical prognostic parameters for estimating short-term survival of the illness and rejected age-based cutoffs. They recommended insulation of the physician from the entirety of the decisions, instead relying on the collective judgment of a medical team. For patients not responding to treatment or worsening, the statements preserved proportionality-based decisions to withdraw “futile treatments.”29
In New York, planning for the scarcity of ventilators in a pandemic had been going on for some time. The New York State Task Force on Life and the Law, a gubernatorially appointed policy advisory body operating under the auspices of the Department of Health, published draft non-binding Ventilator Allocation Guidelines in 2007.30 Intended for use in a pandemic, the stated purpose of this guideline was to maximize the number of survivors of the acute illness, disavowing quality of life judgments and age-based exclusions. These guidelines applied a set of hard-stop medical exclusion criteria consisting of several conditions that objectively would have a high mortality even with ventilator use. Triage decisions would then rely upon a physiologic scoring system designed to assess short-term mortality, specifically including the Sequential Organ Failure Assessment (SOFA) scoring system.31 The guidelines provided for ventilator use as a “time trial” with patients reassessed at 48 and 120 hours and with patients subject to removal of the ventilator if they subsequently failed to meet “rationing standards for continued ventilator support.”32 The decision to withdraw a ventilator was removed from the treating physician and placed under the authority of a triage officer.
The New York guidelines underwent revision in 2015 but with no substantive change to the fundamental structural or operational provisions.33 The 2015 document again stressed the possibility of nonconsensual withdrawal of a ventilator:Public outreach will inform people about the goals and steps of the clinical ventilator allocation protocols. Information should emphasize that pandemic influenza is potentially fatal, that health care providers are doing their best with the limited resources, and the public must adjust to a different way of providing and receiving health care than is customary. Instead, a protocol based only on clinical factors will be used to determine whether a patient receives (or continues with) ventilator treatment to support the goal of saving the greatest number of lives in an influenza pandemic where there are a limited number of available ventilators. Patients and families should be informed that ventilator therapy represents a trial of therapy that may not improve a patient’s condition sufficiently and that the ventilator will be removed if this approach does not enable the patient to meet specific criteria.34
By late March 2020, as COVID-19 cases were exploding in New York City, the Governor of New York effectively disavowed the Ventilator Allocation Guidelines.35 During the height of the crisis, the State of New York did not endorse or authorize use of crisis standards of care, leaving individual facilities to variably develop and utilize crisis response practices.36 The Governor did authorize statewide redeployment of geographically dispersed ventilator inventory and endorsed repurposing anesthesia and bilevel positive airway pressure (BiPAP) machines as well as the splitting of a ventilator for more than one patient.37 As ventilator supply dwindled, the Mount Sinai Hospital developed and tested an experimental protocol for the controversial concept of ventilator splitting, though it ultimately was not required in clinical practice.38 On April 26, 2020, with all New York City hospitals having activated their incident command systems, the New York City Department of Health and Mental Hygiene issued a document recommending that all hospitals have a crisis care committee and providing detailed guidance on use of crisis standards of care and triage protocols.39 This document explicitly indicated that withdrawal of life support could be nonconsensual and was not subject to consent or agreement of the patient or family:When triage to “palliative care only” in disasters is not by patient choice but dependent on available resources, management of expectations and transitions is critical to the physical and mental well-being of patient, family and providers. Anticipating patient and family questions, explaining resource allocation and acknowledging their frustration and feeling of injustice may help them accept the protocol.40
Thus, in both northern Italy and New York City, the calamity of the pandemic brought to the fore allocation and reallocation triage concepts that had previously been theoretical and untested in the setting of extreme scarcity of intensive care resources. In particular, the protocols allowing for nonconsensual removal of a ventilator for reallocation purposes, with the presumptive certainty that the removed patient would die, brought forward unprecedented practical, ethical, and moral challenges.
The conceptual nature of triage
Formal mechanisms of triage allocation of medical resources in the face of demand/supply imbalance are a relatively recent phenomenon. Taking origin in the French concept of trier, or sorting, organized triage in medical scarcity was first espoused in the nineteenth century by British Naval Surgeon John Wilson, who sorted battle wounds into three groups: slight, which may be put off; serious, which call for immediate attention; and fatal, for which nothing can be done.41 This tripartite structure continues as foundational for modern concepts of triage. In 2009 the Institute of Medicine, in studying crisis standards of care, stated that “a triage program aims to rapidly screen, evaluate, and sort patients based on their medical status and likely outcome,” and included example triage protocols with exclusion of those unlikely to survive.42
Functionally, Ezekiel Emanuel and coauthors proposed early in the pandemic that the overarching ethical value in triage of life-sustaining resources should be maximization of value, understood as “saving the most individual lives or as saving the most life-years by giving priority to patients likely to survive longest after treatment,” with maximization of short-term survivorship as the priority aim.43 They held that long-term prognosis should be a tiebreaker.
Inherent in a most lives saved ethic in times of extreme scarcity is the idea that those with a very poor short-term prognosis may be excluded from receipt of scarce life-saving resources. In a 2014 consensus statement on care of the critically ill and injured during pandemics and disasters, The American College of Chest Physicians published exclusion criteria based on anaim to identify patients who are not candidates for ICU admission, including those (1) with a poor prognosis despite ICU care, (2) requiring resources that cannot be provided, and (3) whose underlying illness has a poor prognosis with a high likelihood of death.44
The last of these three exclusion categories has proved to be most controversial. Ezekiel and coauthors noted that “maximizing benefits requires consideration of prognosis—how long the patient is likely to live if treated—which may mean giving priority to younger patients and those with fewer coexisting conditions.”45 Persons with disabilities quickly realized that references to comorbidities and coexisting conditions could broadly be interpreted to establish a discriminatory regime that disadvantaged their access to scarce life-saving resources. On March 18, 2020, the National Council on Disability wrote Roger Severino, the Director of the Office for Civil Rights, U.S. Department of Health & Human Services, with a dire warning:OCR should rapidly head off what could be yet another time in US history when people with disabilities are left to die because medical decisions remain infused with disability bias or because physicians are not aware of their responsibilities under the Americans with Disabilities Act, the Rehabilitation Act, and the Affordable Care Act. More evidence of the need for immediate OCR action is found in a cursory review of State protocols for standards of medical care for times of crisis—like a pandemic—that show that people with existing disabilities will be, if the plans remain the same, discriminated against in the provision of COVID-19 care.46
The Director agreed, and on March 28, 2020 the Office for Civil Rights issued a Bulletin providing civil rights guidance concerning non-discrimination during the COVID-19 public health emergency. The Bulletin provided that disability discrimination laws for covered entities “remain in effect” during the crisis and warned thatAs such, persons with disabilities should not be denied medical care on the basis of stereotypes, assessments of quality of life, or judgments about a person’s relative “worth” based on the presence or absence of disabilities or age. Decisions by covered entities concerning whether an individual is a candidate for treatment should be based on an individualized assessment of the patient based on the best available objective medical evidence.47
Director Severino included within the Bulletin a stunningly direct emphasis of fundamental principles to govern the conceptual nature of triage:HHS is committed to leaving no one behind during an emergency, and this guidance is designed to help health care providers meet that goal. Persons with disabilities, with limited English skills, or needing religious accommodations should not be put at the end of the line for health services during emergencies. Our civil rights laws protect the equal dignity of every human life from ruthless utilitarianism.48
This Bulletin firmly refuted the primacy of a population-based utilitarian ethos of triage and forcibly guided these difficult decisions into a deontological path bound by principles of the fundamental worth and dignity of each human being.
Equal dignity and worth of each human being
The statements in the March 2020 Office for Civil Rights Bulletin are deeply resonant with Judeo-Christian moral teachings. At the throne of judgment, a characteristic of the righteous will be that they fed the hungry, gave drink to the thirsty, clothed the naked, visited the sick, and came to those in prison (see Matt 25:31–40). Here, the righteous cause is to recognize the equal dignity and worth of all in need and to provide for them as one may be able. The reality that needs may exceed resources is clearly understood but does not negate or diminish the moral imperative to do what may be done: “Since there will never cease to be some in need on the earth, I therefore command you, ‘Open your hand to the poor and needy neighbor in your land’” (Deut 15:11 NRSV). Likewise, the ability to repay, taken allegorically as socially defined worth of persons, must not cause favor:When you give a luncheon or dinner, do not invite your friends, your brothers or sisters, your relatives, or your rich neighbors; if you do, they may invite you back and so you will be repaid. But when you give a banquet, invite the poor, the crippled, the lame, the blind. (Luke 14:12–13 NRSV)
In the extreme scarcity of a pandemic surge, patients in need of life-saving resources will be rich and poor, strong and weak, young and old. Each possesses an equal claim to life-saving resources. Each must equally receive a just and objective assessment, fairly applying such standards as may sort those who will likely survive with help from those who likely will not. A utilitarian ethos favoring life-years saved, with a greater societal return on investment, violates this fundamental principle of equality.
Reallocation of life-saving resources
As difficult as triage at the time of initial allocation of life-saving resources may be, an even more troublesome question arises when decision makers seek to reallocate life-saving resources. Reallocation is a population-centered process during surge crisis by which scarce life-saving resources may be nonconsensually removed from patients to whom they previously have been allocated with the intent that those life-saving resources will then be assigned to other individuals thought to have a more favorable prognosis or a higher potential of benefit.49
Reallocation is founded in the idea that the initial allocation of a life-saving resource is a “time trial” of therapy. Therapeutic trials and cessation of non-beneficial treatments are common concepts in medicine. An antibiotic is stopped when an infection becomes resistant. Antihypertensive medications are changed if they do not produce the desired result. And life-saving resources may be consensually withdrawn, often on the patient’s initiative, when the patient has clearly become terminally ill and the therapy is only prolonging the dying process.
In the United States, these latter concepts were forged in the second half of the twentieth century through a series of difficult court cases. The parents of Karen Ann Quinlan gained recognition that the right to privacy includes the right to withdraw life-sustaining treatment when hope of emerging from a coma ends.50 In the case of Nancy Cruzan, the US Supreme Court recognized a 14th Amendment liberty interest in refusing unwanted medical treatment.51 These decisions were followed by the widespread adoption of laws authorizing advance directives known as living wills.52
In a pandemic surge, as in all other times, physicians should continue the usual practice of informing a patient or family when an illness has become terminal with no reasonable hope of recovery. That point is the entirely proper time to begin compassionate discussions regarding the consensual withdrawal of life-sustaining measures in order to allow the inevitable natural dying process to complete.
Nonconsensual removal of life-supportive measures is an entirely different matter. The process of taking a ventilator from a patient who is very sick but not hopelessly ill is included in many pandemic triage protocols. In a 2020 systematic review of publicly available state guidelines for ventilator allocation, 22 of 26 adult guidelines discussed removal of mechanical ventilation from one patient to give to another.53 Emanuel and coauthors noted that “many guidelines agree that the decision to withdraw a scarce resource to save others is not an act of killing and does not require the patient’s consent.”54
But it is clear that the patient from whom the life-saving resource is withdrawn is very ill and may well die upon removal. In fact, the grave severity of the illness forms the very basis of the utilitarian justification for removal. As the Institute of Medicine asserted in 2012, “The patient who is using the resource should, in the judgment of the triage team, have a substantially worse prognosis to justify withdrawal and reassignment of the resource.”55
If the patient dies as a direct consequence of nonconsensual removal of a ventilator, then little reasonable doubt exists that the reallocation was an act of killing. Thereby, some actions to reallocate ventilators may well be criminal acts. Cohen and coauthors clearly presented this case in April 2020:A clinician who intentionally withdraws a ventilator from a nonconsenting patient could conceivably be charged with criminal homicide. If the clinician knows that removing the ventilator will result in the death of the patient, the applicable charge would be murder. If the clinician knows there is a substantial risk the patient will die, and the patient does die, the applicable charge would be manslaughter. It does not matter whether the patient would have died soon regardless. Action that shortens a life, even if just by hours, can be prosecuted as a homicide, with charges potentially filed against any individual who participated in or directed the ventilator removal and against the hospital.56
Such actions are clearly violative of profound interests of the state in protecting life, particularly as to protection from the taking of innocent life by others. These state interests were explicitly enumerated by the US Supreme Court in its 1997 rejection of a constitutional right to assistance in suicide:These interests include prohibiting intentional killing and preserving human life; preventing the serious public-health problem of suicide, especially among the young, the elderly, and those suffering from untreated pain or from depression or other mental disorders; protecting the medical profession’s integrity and ethics and maintaining physicians’ role as their patients’ healers; protecting the poor, the elderly, disabled persons, the terminally ill, and persons in other vulnerable groups from indifference, prejudice, and psychological and financial pressure to end their lives; and avoiding a possible slide toward voluntary and perhaps even involuntary euthanasia.57
As the Christian Medical and Dental Association notes, “Non-consensual withdrawal of life-supportive resources (e.g., mechanical ventilation) involves an active, intentional, and direct taking from a vulnerable person incapable of resisting.”58
That some good may come of the reallocation of the ventilator to another patient has led some to invoke the principle of double effect as a justification for reallocation.59 In order to justify a moral harm as an unintended consequence of the pursuit of a morally good end, the first two conditions of the principle of double effect require that the act itself must be morally good or indifferent and that the agent permits but does not intend the bad effect.60 But the initial act, nonconsensual removal of the ventilator and its ongoing benefit with subsequent expected injury and harm to the removed patient, has no claim to be a morally good or neutral act per se. And as to the examination of intent, the American Law Institute notes:A person acts with the intent to produce a consequence if:
(a) the person acts with the purpose of producing that consequence;
or
(b) the person acts knowing that the consequence is substantially certain to result.61
Physicians clearly have substantial certainty of the outcome of their actions when nonconsensually removing a ventilator from a critically ill but not terminally ill patient on grounds that the patient is sicker than the next patient in need. With its first and second premises false, the double effect argument in justification of nonconsensual ventilator reallocation collapses as unsound. These physicians act with intent and with moral culpability.
Reallocation is a fundamentally different moral process than the sorting among patients with an equal claim to resources at the time of initial triage allocation. In the initial allocation, with both patients being found in their respective conditions, one is held to have a greater need; or, if both have equal need, then one has a better chance of short-term survival than the other; or, if both need and chances of short-term survival are equal, then lots are cast. But initially neither patient has possession of the ventilator or its benefits. In reallocation, a patient has a possessive claim to the benefit of the ventilator. With nonconsensual removal, that claim is derogated, and the patient is devalued by the taking process.
The immorality of taking
Judeo-Christian traditions condemn merciless takings from the weak and the vulnerable. The parable of the unforgiving debtor exemplifies this: one who had received mercy and the forgiveness of a great debt showed no mercy or forgiveness to another, and thereby came to judgment (see Matt 18:21–35). Likewise, Proverbs speaks of offense of the poor as an insult to God (see Prov 14:31).
True, in a pandemic surge both the ventilated patient and the patient in need of a ventilator may be allegorically considered as the poor. As such, both have a claim to relief. But the unjust taking distinguishes reallocation as immoral.
Suppose that in a time of famine a bag of grain is distributed to each family, first to smaller families with the idea that these families are more likely to survive. Two families of four remain in line as the last bag is reached. Lots are cast, and one of the families is given the grain. Just as they are leaving the forum, a family of three arrives. Is taking the bag of grain and reallocating it to the smaller family moral? That such an action is repulsive is a testament to the fact that there is justice in giving scarce resources but injustice in taking them.
Within the realm of medicine, condemning the unjust taking from the weak and the vulnerable is profoundly important. This principle guards against the exploitation of unwitting research subjects. This principle prevents economic line-breaking when vaccines are scarce. And this principle causes society to condemn the taking of transplantable organs from political prisoners.
Moral distress and the conflict of duties
Notably, although the SIAARTI recommendations sought to “share with clinicians the responsibility in the decisions making process,” they did not contain references to objective criteria to define prognosis and they left the ethical burden of the triage and withdrawal decisions squarely on the physician providing the care.62 This situation set the stage for intensified moral distress and moral injury in the frontline physicians.
Moral distress is “the psychological distress of a situation in which one is constrained from acting on what one knows to be right.”63 Persistent moral distress can produce moral injury, with erosion of self-worth and trust in social systems.64 In discussion with three frontline Italian physicians, Lisa Rosenbaum reported descriptions of ventilator constraints and triage as “unbearable,” “aversive,” and “exquisitely uncomfortable.”65 In a series of interviews with 15 emergency and critical-care physicians who served during the pandemic in Lombardy, 10 of whom were in triage roles, researchers found significant stressors, including limited health care resources, intensified patient triage, changeable selection criteria, limited therapeutic/clinical knowledge, and patient isolation.66
A substantial portion of the moral distress of pandemic triage is likely generated by the conflict of two fundamental duties of the physician: the duty to care and the duty to steward resources. These two duties come into direct conflict in the extreme scarcity of a pandemic surge. The principal duty of a physician is to care for the patient. This duty founds a trust-based relationship wherein the patient relies upon the representation of the physician that care will be given and decisions will be made in loyalty and prudence with the sole end being benefit to the health and well-being of the patient. The duty to care is found in the Hippocratic Oath, which affirms that, “Whatever houses I may visit, I will come for the benefit of the sick, remaining free of all intentional injustice.”67 The duty to care is more fully developed within the Physician’s Pledge that is the World Medical Association Declaration of Geneva:THE HEALTH AND WELL-BEING OF MY PATIENT will be my first consideration;
I WILL RESPECT the autonomy and dignity of my patient;
I WILL MAINTAIN the utmost respect for human life;
I WILL NOT PERMIT considerations of age, disease or disability, creed, ethnic origin, gender, nationality, political affiliation, race, sexual orientation, social standing or any other factor to intervene between my duty and my patient.68
The American Medical Association likewise enumerates the duty to care as a core ethical principle of the medical profession, stating in The AMA Principles of Medical Ethics that, “A physician shall, while caring for a patient, regard responsibility to the patient as paramount.”69
In the extreme scarcity of a pandemic surge, a secondary duty to steward resources places conflicting demands upon the physician. Under ordinary circumstances, this duty as espoused by the American Medical Association “requires physicians to be prudent stewards of the shared societal resources with which they are entrusted.”70 When caring for two patients in urgent need of one remaining ventilator, a treating physician bound by a duty to care is subject to a profound moral conflict, particularly if the projected utility is greater for one of the patients. When faced with this no-win situation, the optimal solution is to remove the treating physician from the allocation decision. This solution is best accomplished by assigning the duty to steward resources to an authorized triage officer or committee that has no explicit duty to care for the patient and that follows promulgated authoritative, objective, and just triage criteria and protocols.71 Another possible solution is the reconfiguration of existing resources. This reconfiguration may include repurposing of devices not primarily designed for long-term mechanical ventilation, such as limited function transport ventilators, anesthesia machines, and adapted continuous flow ventilators.72 Giving a patient a resource that offers suboptimal benefit in lieu of the disastrous course of no resource at all is certainly a Hobson’s choice, but it is a life-affirming path.
In either solution, the processes involved remove the treating physician from the no-win scenario, are not violative of the duty to care, and preserve principles of justice.
Principles addressing allocation, reallocation, and reconfiguration of life-sustaining resources as a response to crisis-induced scarcity in a pandemic
To conclude, I offer the proposition that given the ethical complexities of triage of life-sustaining resources in the extreme scarcity of a pandemic surge, the following normative principles should guide:
Each human being is of equal dignity and worth: those who are thought able and those who are not, those who are beset by illness and those who are not, and those who are thought likely to recover and those who are not.
Among the community of persons, all persons hold a fundamentally equal claim to the benefits of scarce life-sustaining resources.
The duty of the treating physician is to care for the person who, in becoming a patient, has entrusted that the physician will seek their well-being and best interests in matters of sickness and health and that duty to care is expressed by unfettered loyalty to those concerns. The duty to care does not include the duty to provide care that the patient refuses or that the treating physician judges as medically ineffective.
For the treating physician, the duty of stewardship of scarce resources is subordinate to the duty to care for the patient.
Under conditions of crisis-induced scarcity and while under declaration and dispositions specified by government, adjudication of competing claims for life-sustaining resources should be made by an authority other than that of the treating physician. Such authority may rest in a properly designated triage officer or team.
In adjudication of competing claims to life-sustaining resources, allocation of such resources is to be made by medical judgment informed by objective evidence and should be done by defined processes that are fair, just, and socially non-discriminatory. Such decisions should be clearly communicated to the patient or their decision-maker.
The claim to a life-sustaining resource, once allocated, persists.
When the treating physician concludes that a patient has reached a point of terminal and irreversible illness with no reasonable hope of recovery, the physician should counsel the patient or their decision-maker of that conclusion.
At the point of terminal and irreversible illness with no reasonable hope of recovery or upon proper request by the patient or their decision-maker, life-sustaining care and resources may be consensually withdrawn.
The reallocation of life-sustaining care or resources, here defined as the non-consensual taking or removal of life-sustaining care or resources from a patient, may not be done when accompanied by the intent that the patient will die or the knowledge of substantial risk that the patient will die.
Under the condition of crisis-induced scarcity, the reconfiguration of life-sustaining resources and care, here defined as repurposed or altered use of resources for life-sustaining care, may be performed with the goal of saving the life of a patient, even if such reconfiguration results in suboptimal benefit of the care or resource or increased risk to the patient, so long as the reconfiguration is done without the intent that the patient will die or the knowledge of substantial risk that the patient will die.
Author biography
Frederick J. White III practices cardiology in Shreveport, Louisiana. He completed his MD in 1982 at the LSU Health School of Medicine-Shreveport and completed residency training in Internal Medicine at the University of Arkansas for Medical Sciences and fellowship training in cardiology at the Vanderbilt Medical Center, where he served as an Instructor in Medicine and the Chief Fellow in Cardiology. White is certified by the American Board of Internal Medicine in both Internal Medicine and Cardiovascular Diseases and is a Fellow of the American College of Cardiology and the American College of Chest Physicians. He holds the Certificate in Health Care Ethics from the National Catholic Bioethics Center and is a Certified Healthcare Ethics Consultant by the HCEC Certification Commission of the American Society for Bioethics and Humanities. He is a member of the medical staff of the Willis-Knighton Medical Center and served as Chair of the Institutional Ethics Committee for 20 years.
1. Sudhvir Singh et al., “How an Outbreak Became a Pandemic: A Chronological Analysis of Crucial Junctures and International Obligations in the Early Months of the Covid-19 Pandemic,” Lancet 398.10316 (December 4–10, 2021): 2109–24.
2. Coronaviridae Study Group of the International Committee on Taxonomy of Viruses, “The Species Severe Acute Respiratory Syndrome-Related Coronavirus: Classifying 2019-Ncov and Naming It Sars-Cov-2,” Nature Microbiology 5.4 (April 2020): 536–44. The first SARS-CoV coronavirus caused a 2003 outbreak with roughly 8000 cases and 800 deaths worldwide. Robert A. Weinstein, “Planning for Epidemics—The Lessons of SARS,” New England Journal of Medicine 350.23 (June 3, 2004): 2332–34.
3. Michell L. Holshue et al., “First Case of 2019 Novel Coronavirus in the United States,” New England Journal of Medicine 382.10 (March 5, 2020): 929–36.
4. Danilo Cereda et al., “The Early Phase of the COVID-19 Epidemic in Lombardy, Italy,” Epidemics 37 (December 2021): 100528.
5. World Health Organization, “Coronavirus Disease 2019 (COVID-19): Situation Report – 40,” February 29, 2020, https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200229-sitrep-40-covid-19.pdf?sfvrsn=849d0665_2.
6. Tedros Adhanom Ghebreyesus, “WHO Director-General’s Opening Remarks at the Media Briefing on COVID-19,” March 11, 2020, World Health Organization, https://www.who.int/director-general/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19—11-march-2020.
7. Benedetta Armocida et al., “The Italian Health System and the COVID-19 Challenge,” Lancet Public Health 5.5 (May 2020): e253.
8. Giacomo Grasselli et al., “Critical Care Utilization for the COVID-19 Outbreak in Lombardy, Italy,” Journal of the American Medical Association 323.16 (April 28, 2020): 1545–46, doi.org/10.1001/jama.2020.4031.
9. Ministero della Salute, Istituto Superiore di Sanità, Lombardia Aggiornamento epidemiologico, Monitoraggio Fase 2 Report settimanale Report 3 / sintesi [Lombardy Epidemiological Update, Phase 2 Monitoring Weekly Report: Report 3 / Summary], June 3, 2020, https://www.salute.gov.it/imgs/C_17_monitoraggi_40_8_fileRegionale.pdf.
10. Grasselli et al., “Critical Care,” 1545.
11. Nicole Maurer et al., “The First 110,593 COVID-19 Patients Hospitalised in Lombardy: A Regionwide Analysis of Case Characteristics, Risk Factors and Clinical Outcomes,” International Journal of Public Health 67 (May 11, 2022): 1604427.
12. Anna Odone et al., “COVID-19 Deaths in Lombardy, Italy: Data in Context,” Lancet Public Health 5.6 (June 2020): e310.
13. Lisa Rosenbaum, “Facing Covid-19 in Italy—Ethics, Logistics, and Therapeutics on the Epidemic’s Front Line,” New England Journal of Medicine 382.20 (May 14, 2020): 1873–75.
14. Corinne N. Thompson et al., “COVID-19 Outbreak—New York City, February 29–June 1, 2020,” MMWR Morbidity and Mortality Weekly Report 69.46 (November 20, 2020): 1725–29.
15. CDC COVID-19 Response Team, “Geographic Differences in COVID-19 Cases, Deaths, and Incidence—United States, February 12–April 7, 2020,” MMWR Morbidity and Mortality Weekly Report 69.15 (April 17, 2020): 465–71.
16. CDC COVID-19 Response Team, “Geographic Differences,” 466.
17. A. Oliveri et al., “COVID-19 Cumulative Incidence, Intensive Care, and Mortality in Italian Regions Compared to Selected European Countries,” International Journal of Infectious Diseases 102 (January 2021): 363–68.
18. Thompson et al., “COVID-19 Outbreak,” 1726.
19. Thompson et al., “COVID-19 Outbreak,” 1727.
20. Josefa Velasquez et al., “Hospitals Nearing ICU Bed Limits as COVID-19 Surges in NYC,” The City, March 28, 2020, https://www.thecity.nyc/health/2020/3/28/21210412/hospitals-nearing-icu-bed-limits-as-covid-19-surges-in-nyc.
21. Jennifer Calfas and Talal Ansari, “New York Races to Get Coronavirus Supplies Before Cases Peak,” The Wall Street Journal, April 5, 2020, https://www.wsj.com/articles/new-york-races-to-get-coronavirus-supplies-before-cases-peak-11586123114
22. Marco Vergano et al., “SIAARTI Recommendations for the Allocation of Intensive Care Treatments in Exceptional, Resource-limited Circumstances,” Minerva Anestesiologica 86.5 (May 2020): 469–72.
23. Vegano et al., “SIAARTI Recommendations,” 470.
24. Marco Vergano et al., “Clinical Ethics Recommendations for the Allocation of Intensive Care Treatments in Exceptional, Resource-Limited Circumstances: The Italian Perspective During the COVID-19 Epidemic,” Critical Care 24 (2020): 165.
25. Vergano et al., “SIAARTI Recommendations,” 471.
26. Vergano et al., “SIAARTI Recommendations,” 471.
27. Luigi Ricconi et al., “The Italian Document: Decisions for Intensive Care When There Is an Imbalance Between Care Needs and Resources During the COVID-19 Pandemic,” Annals of Intensive Care 11.1 (June 29, 2021): 100.
28. Ricconi et al., “Italian Document,” 100.
29. Ricconi et al., “Italian Document,” 100.
30. New York State Task Force on Life and the Law, New York State Department of Health, “Allocation of Ventilators in an Influenza Pandemic: Planning Document, Draft for Public Comment March 15, 2007,” https://www.cidrap.umn.edu/sites/default/files/public/php/196/196_guidance.pdf.
31. J. L. Vincent et al., “The SOFA (Sepsis-related Organ Failure Assessment) Score to Describe Organ Dysfunction/Failure,” Intensive Care Medicine 22 (1996): 707–10.
32. New York State Task Force, “Allocation of Ventilators,” 35.
33. Susie A. Han and Valerie G. Koch, “Clinical and Ethical Considerations in Allocation of Ventilators in an Influenza Pandemic or Other Public Health Disaster: A Comparison of the 2007 and 2015 New York State Ventilator Allocation Guidelines,” Disaster Medicine and Public Health Preparedness 14.6 (December 2020): e35–e44.
34. New York State Task Force on Life and the Law, New York State Department of Health, Ventilator Allocation Guidelines, November 2015, https://www.health.ny.gov/regulations/task_force/reports_publications/docs/ventilator_guidelines.
35. Valerie G. Koch and Susie A. Han, “COVID in NYC: What New York Did, and Should Have Done,” American Journal of Bioethics 20.7 (July 2020): 153–55.
36. Tia Powell and Elizabeth Chuang, “COVID in NYC: What We Could Do Better,” The American Journal of Bioethics 20.7 (2020): 62–66.
37. New York State Adult Day Services Association, Inc., “4/2/2020 Daily Update: From Governor Cuomo’s Briefing in the Red Room of the NYS Capitol,” www.nysadultday.com/latest-information.
38. Garrett W. Burnett et al., “Managing COVID-19 from the Epicenter: Adaptations and Suggestions,” Journal of Anesthesia 35 (2021): 366–73.
39. NYC Department of Health and Mental Hygiene, “COVID-19 Pandemic Patient Surge: Preparing for Crisis Care,” April 26, 2020, www1.nyc.gov/assets/doh/downloads/pdf/imm/covid-19-patient-surge-crisis-care.pdf.
40. NYC Department of Health and Mental Hygiene, COVID-19 Pandemic Patient Surge, 21.
41. John Wilson, Outlines of Naval Surgery (Edinburgh: Maclachlan, Stewart, and Co., 1846).
42. Institute of Medicine, Guidance for Establishing Crisis Standards of Care for Use in Disaster Situations (Washington, DC: The National Academies Press, 2009), 17. In 2015, the Institute of Medicine was reconstituted as the National Academy of Medicine.
43. Ezekiel J. Emanuel et al., “Fair Allocation of Scarce Medical Resources in the Time of Covid-19,” New England Journal of Medicine 382.21 (May 21, 2020): 2049–55.
44. Michael D. Christian et al., “Triage: Care of the Critically Ill and Injured during Pandemics and Disasters: CHEST Consensus Statement,” Chest 146.4 Suppl (October 2014): e61S–74S.
45. Emanuel et al., “Fair Allocation of Scarce Medical Resources,” 2053.
46. Neil Romano, National Council on Disability, letter to HHS OCR Roger Severino, March 18, 2020, https://ncd.gov/publications/2020/ncd-covid-19-letter-hhs-ocr#_ftn3.
47. U. S. Department of Health and Human Services, Office for Civil Rights, “BULLETIN: Civil Rights, HIPAA, and the Coronavirus Disease 2019 (COVID-19),” March 28, 2020, www.hhs.gov/sites/default/files/ocr-bulletin-3-28-20.pdf.
48. U. S. Department of Health and Human Services, “BULLETIN.”
49. Bruce M. Altevogt, Clare Stroud, Sarah L. Hanson, Dan Hanfling, Lawrence O. Gostin, eds, Guidance for Establishing Crisis Standards of Care for Use in Disaster Situations (Washington, DC: National Academies Press, 2009), 34–35, https://nap.nationalacademies.org/read/12749/chapter/1.
50. In re Quinlan, 355 A.2d 647, 70 N.J. 10 (1976).
51. Cruzan v. Director, Missouri Department of Health (88-1503), 497 U.S. 261 (1990).
52. George J. Annas, “The Health Care Proxy and the Living Will,” New England Journal of Medicine 324 (1991): 1210–13.
53. Gina M. Piscitello et al., “Variation in Ventilator Allocation Guidelines by US State during the Coronavirus Disease 2019 Pandemic: A Systematic Review,” JAMA Network Open 3.6 (June 1, 2020): e2012606.
54. Emanuel et al., “Fair Allocation of Scarce Medical Resources,” 2053.
55. Committee on Guidance for Establishing Crisis Standards of Care for Use in Disaster Situations, Institute of Medicine, Crisis Standards of Care: A Systems Framework for Catastrophic Disaster Response: Volume 1: Introduction and CSC Framework (Washington, DC: National Academies Press, 2012), 4–31.
56. I. Glenn Cohen et al., “Withholding Ventilators during COVID-19: Assessing the Risks and Identifying Needed Reforms,” Journal of the American Medical Association 323.19 (2020): 1901–1902.
57. Washington v. Glucksberg, 521 U.S. 702, 117 S. Ct. 2258, 138 L. Ed. 2d 772 (1997).
58. Paul Hoener et al., “Triage and Resource Allocation during Crisis Medical Surge Conditions (Pandemics and Mass Casualty Situations): A Position Statement of the Christian Medical and Dental Associations Special Task Force,” Christian Journal for Global Health 7.1 (April 2020): 45–55.
59. Jason T. Eberl and G. Kevin Donovan, “Is It Ethical to Unilaterally Withdraw Life-Sustaining Treatment in Triage Circumstances?” Health Progress (April 1, 2020), https://www.chausa.org/publications/health-progress/archives/issues/pandemic-coverage/is-it-ethical-to-unilaterally-withdraw-life-sustaining-treatment-in-triage-circumstances.
60. Alison McIntyre, “Doctrine of Double Effect,” in The Stanford Encyclopedia of Philosophy (ed. Edward N. Zalta, Spring 2019), https://plato.stanford.edu/entries/double-effect/#Aca.
61. American Law Institute, Restatement of the Law, Third: Torts: Liability for Physical and Emotional Harm: Restatement Volume 2 (St. Paul, MN: American Law Institute Publishers, 2012), §1(b).
62. Vergano et al., “SIAARTI Recommendations,” 471.
63. F. Akram, “Moral Injury and the COVID-19 Pandemic: A Philosophical Viewpoint,” Ethics, Medicine and Public Health 18 (September 2021): 100661.
64. Akram, “Moral Injury and the COVID-19 Pandemic,” 100661.
65. Rosenbaum, “Facing COVID-19 in Italy.”
66. Guilia Lamiani et al., “Moral Distress Trajectories of Physicians 1 Year after the COVID-19 Outbreak: A Grounded Theory Study,” International Journal of Environmental Research and Public Health 18.24 (December 19, 2021): 13367.
67. Ludwig Edelstein, “The Hippocratic Oath: Text, Translation, and Interpretation,” in Ancient Medicine: Selected Papers of Ludwig Edelstein (ed. O. Temkin and C. L. Temkin; Baltimore, Md.: The Johns Hopkins University Press, 1967), 3–64.
68. World Medical Association, “WMA Declaration of Geneva,” World Medical Association Current Policies, July 9, 2018, https://www.wma.net/policies-post/wma-declaration-of-geneva/.
69. American Medical Association, “American Medical Association Principles of Medical Ethics,” in Code of Medical Ethics of the American Medical Association (Chicago: The American Medical Association, 2017), 1–2.
70. American Medical Association, “Physician Stewardship of Health Care Resources,” in Code of Medical Ethics of the American Medical Association (Chicago: The American Medical Association, 2017), 183–84.
71. Frederick J. White III, “The Prioritization of Life-Saving Resources in a Pandemic Surge Crisis,” Issues in Law & Medicine 35.1 (2020): 143–60.
72. Wen Ting Siow et al., “Managing COVID-19 in Resource-limited Settings: Critical Care Considerations,” Critical Care 24.1 (April 22, 2020): 167.
| 0 | PMC9742503 | NO-CC CODE | 2022-12-14 23:30:02 | no | 2022 May; 119(1-2):50-63 | utf-8 | null | null | null | oa_other |
==== Front
RAE
sprae
Review & Expositor
0034-6373
2052-9449
SAGE Publications Sage UK: London, England
10.1177/00346373221134615
10.1177_00346373221134615
Thematic Words · · ·
A shared humanity: COVID capitalism and the future of the health care ethics
Spaulding Hank W. III
Mount Vernon Nazarene University, USA
Hank W. Spaulding III, Mount Vernon Nazarene University, 800 Martinsburg Rd, Mount Vernon, OH 43050, USA. Email: [email protected]
5 2022
5 2022
5 2022
119 1-2 8699
© The Author(s) 2022
2022
Review & Expositor, Inc
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
The tension between the economy and health care in the United States was on full display during the COVID-19 pandemic and continues to raise uncomfortable questions for the medical and faith communities. Chief among the issues raised is the inequality that emerged between the healthy and vulnerable, which caused vocal proponents to encourage the vulnerable to sacrifice their lives in order for the economy to continue unfettered by the pandemic. This article explores how “COVID capitalism” constricted the ability of the health care community to execute its duties morally and promote the health and well-being of the nation’s elderly. It argues that the practices of vulnerability and dependence, viewed through the cardinal virtues, unseat the economic reason at the heart of COVID capitalism and promote health as a central good alongside economic well-being.
Christian ethics
COVID-19
health care ethics
neoliberalism
virtue ethics
typesetterts1
==== Body
pmc “. . . for the growing good of the world is partly dependent on unhistoric acts; and that things are not so ill with you and me as they might have been, is half owing to the number who lived faithfully a hidden life, and rest in unvisited tombs.”
George Elliot, Middlemarch
Introduction
“Medicine is a moral community,” Edmund Pellegrino and David Thomasma write, “because it is at heart a moral enterprise and its members are bound together by a common moral purpose.”1 The moral community that comprises medicine includes doctors and nurses who professionally administer treatments and curing remedies. As such, these individuals must align the ends of medicine with those of the common good. This community identified by Pellegrino and Thomasma must extend, however, beyond merely those who professionally consider matters of health. The moral community must also include the everyday citizens who participate in their neighbor’s health. Health care workers’ ability to align their actions with the ends of medicine (e.g., healing, caring, curing) is dependent to some extent upon the support of the community outside medical facilities to maintain such bonds of care.
The global pandemic of COVID-19 illustrated and continues to illustrate the fragility of community support for health care. Take, for example, the proponents who argued for the morally appropriate risk to the elderly to stimulate the economy. Chief among the proponents of this position was Texas Lt. Governor Dan Patrick, who famously proposed that the elderly should willingly sacrifice their lives to resume economic stimulation for their children and grandchildren. At the time of Patrick’s statements, medical professionals warned the elderly against exposure until a vaccine emerged. In this instance, however, the crisis created by economic concerns overrode the competing and more pressing goods in the medical sphere to protect the vulnerable from infection. Furthermore, supporters of Patrick’s position believed the various economic problems necessitated such a sacrifice and outweighed the good provided by the very lives of the elderly. This risk exposes the communal perception that health is subordinate to the economy. While the economic problems created by the shelter-in-place order represent important elements to consider, a dichotomy between physical and economic health presents a false choice to citizens. This false choice arises from the larger practical reason that funds what I term “COVID capitalism,” a form of capitalism that willingly sacrifices lives to deadly infection for economic profit. If the community works against the goods of health by exclusively privileging economic concerns, then the community’s health will necessarily suffer. Health cannot be an individual concern, as COVID capitalism would encourage, because health is a social concern.
This vicious economic formation should concern the Christian community. Though a large portion of US Christians evacuated their moral responsibility to care for their neighbors by refusing to wear masks and refusing to get vaccinated due to “religious convictions,” Christians must recognize their commitment to health care as an expression of neighbor love. The fact that many Christians choose to evacuate these concerns illustrates that COVID capitalism is but an expression of a larger economic formation that malforms communities away from the common good. Nevertheless, the subordination of health care to economic stimulation challenges health care ethics. A solitary gift of the pandemic was the clarity with which captivity to economic reason and practices came into view. As such, an opportunity to present virtues that clearly resist the moral formation of economic reason emerges. If COVID capitalism formed communities away from the virtues, practices, and postures necessary to promote health, new practices must emerge to resist it.
To this end, I argue that a virtue ethics approach to health care corrects the greater malformation of the economic reason that prohibits it. Furthermore, virtue ethics privileges the vulnerable as an essential task of health care ethics, which is not merely the responsibility of nurses and doctors but the entire Christian community. The virtues cultivate the postures of dependency and vulnerability in an age of COVID capitalism that only cherishes autonomy and competition. Dependency and vulnerability are key components of a shared vision of humanity, however, that provides an account of the good and goods necessary to sustain a virtuous community. In this time of crisis, an economic system can use the disorientation of a global pandemic to exploit the sick and place economic values above shared humanity. When shaped by COVID capitalism, competition and autonomy prevent the embodiment of communal virtues that prove important for the defeat of COVID-19. Only in this environment can a new way of thinking about humans outside the logic of economic reason alone become possible. In short, dependence and vulnerability cast a vision for common humanity in which individuals are not isolated in economic enclaves but rather are held deeply inside their shared humanity. A shared humanity enables a practical reason that can balance economic goods and health care not in competition but as a part of the singular vision of the good life.
COVID capitalism: Neoliberalism in the present crisis
COVID capitalism, as a discourse, is an isolated occurrence within the COVID-19 pandemic but emerges in the context of a larger form of economic reasoning already operative in the present age, namely neoliberalism. Scholars struggle to define neoliberalism, but for the sake of this work, I cite Wendy Brown, who defines neoliberalism as “a peculiar form of reason that configures all aspects of existence in economic terms.”2 Patrick’s call for the nation’s elderly to sacrifice their wellbeing for the economy arises out of this economic thinking that encompasses and surpasses moral thinking, which only secondarily considers the health of the elderly alongside the economy. In short, human dignity becomes a secondary virtue when the good becomes synonymous with the economically advantageous. Though, as I argue, this posture already existed below the surface, it was fully exposed during COVID-19.
COVID-19 is a once-in-a-lifetime global crisis. Certainly, other crises defined generations prior to COVID-19, but this present crisis changed life in a way not witnessed for a century. Furthermore, such a global crisis exposed economic rationality at the heart of everything done in the West, especially in the United States. Therefore, the response to COVID-19 occurred through the ideas available for economic reasoning. The infamous economist Milton Friedman writes, “Only a crisis—actual or perceived—produces real change. When that crisis occurs, the actions that are taken depend on the ideas that are lying around.”3 Friedman’s point is that humanity can only respond to a crisis based on the ideas “lying around” before it, even if those ideas were ill-equipped for the task. For example, the United States, one of the wealthiest nations in human history, relied heavily on the ideas central to free market capitalism as the means to respond to COVID-19.
However, free market capitalism is not a monolithic economic system; it has many iterations and expressions. Therefore, one must focus on the mode of free-market capitalism in its neoliberal form. This form of capitalism, which turns all thinking into economic thinking, contains a specific practical rationality inherent in its deployment. Neoliberalism, Adam Kotsko writes, “maintains the conditions necessary for vigorous market competition, trusting in the price mechanism to deliver more efficient outcomes than direct state planning ever could.”4 Neoliberalism then relies on the individual freedom of consumers and producers to operate independently of direct state interference to set the terms of the rules of the economy as a means to govern a peaceful order.
Kotsko recognizes that neoliberalism presents a compelling account of freedom that swallows other competing goods. The neoliberal account of freedom so vigorously defended by many relies on two elements: autonomy and competition. Autonomy arises as a virtue in direct relation to society’s preference against external coercion or intervention. This virtue arises, at least in part, from the Enlightenment doctrine of heteronomy, which affirms the will’s ability to choose independent of external coercion.5 Autonomous freedom from coercion also includes, however, one important feature when extended into the economic sphere: freedom from dependence. For example, one needs to look no further than the shape friendship takes in such neoliberal communities. In neoliberalism, friendship cannot constrain one’s moral life; rather, friends must serve as a utility to profit.6 In such friendship, central moral practices, such as accountability cannot exist because they cannot limit or infringe on autonomy. The prize of autonomy can find no use for others except they serve the interests of the autonomous self.
In addition to autonomy, neoliberalism does not allow friends in the full sense because one must compete. Theologian Kevin Hargaden writes, “We embrace the logic of competition because it promises us we can be winners.”7 Winning allows the individual to be recognized over and against their neighbor and prevents the type of care essential for true friendship. Skidelsky and Skidelsky write that competition becomes a “zero-sum game, because everyone, by definition, cannot have high status. As I spend more . . . I gain status but cause others to lose it. As they spend more to regain status they reduce my own.”8 Such a cycle never ends; more status equates to more success and less status less success. The goal of success creates greater freedoms and privileges. In short, it brings greater freedom. Under neoliberalism, one must make it to the top alone because, at least economically, one cannot share success. Competition encourages autonomy, and autonomy energizes competition. At their base, consumers do not want to depend on their neighbors for the items they desire. They simply want to own. Any place for the community would require a lack of autonomy. Therefore, consumers compete to win these items over their neighbors, so they do not need to be dependent upon them.
One must understand Patrick’s comments within this central emphasis on competition and autonomy. For better or worse, the economy secures one’s status and freedom. Theologian Philip Goodchild goes as far as to observe that in modernity, the economy replaces God as the objective guarantor of value.9 Therefore, Patrick’s comments present the internal logic of neoliberalism. Patrick continues,No one reached out to me and said, as a senior citizen, are you willing to take a chance on your survival in exchange for keeping the America that all America loves for your children and grandchildren . . . And if that’s the exchange, I’m all in.10
Patrick’s “America” is the neoliberal American economy because he wishes to preserve its central practices by his death. Patrick’s potential choice is not what stands out most, but rather why the choice seems necessary. Patrick continues, “And that doesn’t make me noble or brave or anything like that . . . I just think there are lots of grandparents out there in this country like me . . . that what we care about and what we love more than anything are those children.”11 The sentimental posture of this quote masks that the assumed prize of these grandchildren will be the economy. In other words, the flourishing of the nation’s grandchildren is less contingent on the presence and wellbeing of their grandparents and more dependent on their unhindered access and assimilation into the full potential of the economy.
Communities of health: The elderly and health care workers in COVID capitalism
Health is a community concern, and one cannot be healthy alone. Ideally, health should be a shared good toward which all strive. Even before the COVID-19 crisis, health care workers experienced long hours, heavy patient loads, and a lack of workers.12 COVID-19 increased these hardships on health care workers and even led to many early retirements in the field. Many nurses, for example, experienced high amounts of patient death during their long shifts. The impact of the loss of life under medical care that extended over many months of the pandemic left a psychological toll on nurses, leading to greater burnout and fewer health care workers. The relative instability of the health care system, thus, could not focus on general community health but only on disease prevention.
To this end, many governmental figures proposed a shelter-in-place order to alleviate the pandemic’s pressure on health care workers and decrease the spread of COVID-19. Almost from the beginning of the various lockdowns, however, opponents such as Patrick decried the infringement on autonomy and the loss of market competition. Workers’ and the general public’s shared moral commitment to health were not maintained. The libertarian desire to choose for oneself, however, is never solitary. As Patrick sees things, choice must come at the expense of the elderly and others who are uniquely vulnerable to the virus. Furthermore, individual choice further strains the medical community that cares for the patients impacted by this “autonomous” choice.
To honor the autonomous choice of some, others must die, as Patrick readily admits. The disposable nature of the elderly in the wake of COVID capitalism materially emerges through the various health problems they encountered during the pandemic. To be clear, not all health issues encountered by the elderly were solely related to contracting COVID-19. Other health concerns also increased in adults from 50 to 80 in the United States due to a reticence to seek medical care out of fear of exposure.13 Many adults already avoided care for chronic illness, but this fear increased among the elderly, who made up 80% of the mortality rate of the initial wave in 2020. The ableist and ageist mind-set at the heart of COVID capitalism can only envision the market’s needs and not the needs of a complex health system. In such a COVID capitalist frame, the market creates a social contract wherein the health of the elderly does not compare to the market’s needs. Thus, Patrick speaks logically within this frame when he insists that the elderly must pay the price for autonomy and competition.
The desire for autonomy and competition free from the restrictions imposed by shelter-in-place orders and the impact on the elderly further strained health care workers to exercise unethical oversight of patients. One cannot blame health care workers alone for this, but the high patient loads already encouraged in health care communities pushed hospitals to this limit. A classic tenet of health care ethics is the recognition of the patient as a person. During the exacerbated height of the COVID-19 pandemic, nurses were assigned higher patient loads due to a lack of health care workers and increased mortality rates.14 The health care workers serving their patients could not provide the intimate and necessary care to the patients created by increased exposure. Though the increased workload and shift obligations were an emerging issue prior to COVID-19, the pandemic again further exposed the weakness in the health care system to care for patients.
To be clear, the critique of COVID capitalism and its more sinister logic does not negate the serious economic concerns posed by the pandemic or the various other concerns, such as mental health or domestic abuse, often outside the scope of defense of the shelter-in-place. These concerns are also important. The purpose of this critique, however, is merely to highlight that a concern for the economy need not exclude public health, and choosing one over the other is a false choice. Ethicist Cathleen Kaveny argues that the challenge must envision the economy as a place where individuals can flourish rather than a “demigod we sacrifice human beings to.”15 The debate, then, is never the choice of one good over the other but the ability to weigh various goods. This exercise requires the development of practical reason, the ability to weigh, as Kyle Lambelet argues, “the goods or principles at play in a concrete situation . . . and to judge how they ought to act in such a situation to preserve those goods,” and make the most prudential choice that honors the myriad of goods at work in any situation.16 Thus, the ethics that must emerge in response to COVID capitalism must reclaim a thorough account of the practical reason necessary to weigh these goods in question.
Virtue ethics: The medical community
Within a COVID capitalism lens, humanity is supremely implicated in destructive systems that lack the ability to maintain health. To be able to weigh competing goods prudentially, the ethical framework must shift. Largely, goodness in COVID capitalism consists of the freedom to be free from external coercion to compete for a scarcity of resources. Practical reason in COVID capitalism thus weighs actions based on their relative profitability through maximizing autonomy and competition. A theological perspective, however, can and must challenge such forms of practical reason, namely that one must measure all activity to a solely economic calculus. In such a harsh calculus, measuring people’s lives exists merely in their usefulness or ability to generate profit, leaving some as expendable. Such economic reasoning supports health only insofar as it participates in profit-making, and health care only exists as a means for the economic good rather than as a good itself to be weighed alongside economic flourishing. Kaveny observes that the assumption that the economy must take a “planned moment of rest” during the pandemic seems ludicrous to many.17 Thus, health in general and the health of the elderly in particular can only play an instrumental purpose in society.
The moral formation of COVID capitalism only impedes the ends of medicine, namely caring, healing, and curing. The virtues necessary to engage in medicine must align and coordinate with these ends. Competition and autonomy cannot care for the needs of every neighbor; thus, certain individuals must face a greater risk for infection, greater isolation, or other health factors due to isolation. As such, this situation places greater pressure on health care workers to practice medicine in a way that troubles the worker’s ability to develop virtues that align with the ends of medicine. Higher patient loads and longer shifts, for example, lead to more mistakes and burnout, increasing patient loads and shifts for the remaining workers.18 The cycle created by COVID-19 and the acceleration of cases due to the opening economy expedited a significant exodus from health care.19 Therefore, one must recognize that the good of health and the good of the economy conflict with one another and thus require prudence, exercised not only by medical practitioners but also by the community as well, to support health care workers and community health itself. The intimate connection between the moral choices in the community and the medical field is ever present in COVID capitalism. The general health of a community depends largely on the desires and actions of the community. When discussing the medical community, therefore, one must not only think of doctors and nurses. The community at large, albeit differently, must also commit to the virtues that lead to the end of the practice of medicine.
Before considering the virtues necessary for the proper ends of medicine, one must consider the moral formation cultivated by virtues. Virtue in the ancient tradition was an attempt to find eudaimonia. The standard English translation of the word (“happiness”) tragically does not convey the original intention of the phrase. Ancient philosophers recognized eudaimonia not as a call to a general state of mind or pleasure, but rather as a shared understanding of a life well lived. Undoubtedly, eudaimonia is satisfaction, but satisfaction in perpetuating certain habits and acquiring desired virtues. To them, the moral life was an end, not a means to posture or to gain power. Seeking the latter ran counter to eudaimonia’s aim. The virtue ethics tradition broadly recognizes four cardinal virtues: prudence, fortitude, temperance, and justice. Pursuing practices and habits to be shaped into these virtues enables one to live well. Like an athlete perpetually training in the fundamentals of their sport to play the game well, these virtues help humans live their lives well. To be clear, living well is not necessarily a life of financial success but rather a life of flourishing, which should include health.
The cardinal virtues are essential to develop the practical reason necessary to hold in tension the health of the community and the economy. These cardinal virtues revolve around a gravitational matrix at the center of which exists the well-formed agent. In short, virtues travel in packs. For example, a good roommate, friend, or spouse all possess similar virtues. This assumes, then, that the people one desires to have in their community possess similar formations. The virtues that make up communities of character either directly or indirectly appeal to the cardinal virtues because they represent the good life worthy of emulation. Therefore, the virtues share a singular vision of the good life, for “life” defined appropriately.
Temperance forms the agent into desiring actions that do not lead to the performer’s destruction. It means avoiding that which might bring harm or shame and doing that which brings life. It encourages the practice of complementary virtues, such as humility, gentleness, and sobriety. Like temperance, prudence also affirms a kind of moderation.
The prudent person must weigh the competing goods in any given situation to make the most morally appropriate choice. In this way, prudence is the mother of all virtues, which aligns with the true practical reason that allows humans to develop the skills necessary to evaluate and perform the good simultaneously. The prudent person does not merely repeat a universal moral code abstract of context but can truly discern what is just and upright in any given situation and often seeks the wise counsel of others. The vices the prudent person resists are impulsivity, inconstancy, and negligence.
Justice is perhaps the most familiar virtue of all and involves recognizing and delivering what is due to another. The one who practices justice not only refrains from evil acts against one’s neighbor but also acts in such a way to affirm another’s dignity. Though justice has many subtle elements, justice concerns one’s ability to think communally. The community of justice acts in the interest of the common good.
Finally, fortitude is the resolution in the face of difficulty. One can recall Paul’s admonition to the Thessalonians as a call to fortitude, “hold fast to what is good” (1 Thess 5:21, NRSV). Fortitude, therefore, is the ability to do good in the face of uncertainty, even if a single individual is the only one doing it. A virtue closely related to fortitude is magnanimity, which encourages one to rejoice in the practice of virtue even if others do not laud one. Fortitude is a kind of resilience that finds joy in doing good only for the sake of doing good.
How do the virtues lead to eudaimonia? The answer lies in flourishing. In COVID capitalism, flourishing is primarily expressed as individual economic prosperity. Virtuous flourishing, however, is not a call to accumulate wealth but to grow content with oneself, one’s community, and the world. Flourishing, in the virtuous sense, is a communal concept even when expressed by an individual agent. As one can see from the cardinal virtues, virtue ethics encourages internal and external discernment. Temperance is a kind of self-care wherein one commits to loving oneself and neighbors as people worthy of love and affection. Justice is the recognition of that same love and respect due to another. In other words, virtue leads to a shared recognition of our own and one another’s dignity to shape our actions in such a way as to honor that dignity. As such, virtue must be not only a concern of the medical community but also the public at large.
Virtue ethics for medical workers is an instructive place to begin for thinking of an alternative to COVID capitalism. For the most part, medical ethics relies on a principle-duty approach, wherein workers are expected to apply certain principles to specific patients. Medical professionals know how to treat and ethically rely on principles to implement treatment. Yet, through the relationships forged in medical care, doctor and patient exist in a caring bond with one another. Being bound to the patient in care requires more than only the knowledge of treatment and the duty to administer it. Rather, there is a bond of trust and cooperation to promote the health and wellbeing of the patient toward their own physical flourishing. As such, recent medical ethics attempts to turn from a principled account of medical ethics to a virtue-centric approach to express this relationship better. This shift provides a promising way to engage in COVID capitalism.
To be clear, virtues rely on principles to some degree. Principles, such as those in the Hippocratic oath, form certain goods that the doctor or medical care worker must orient. The medical worker does not merely apply principles abstractly, however, but recognizes, as Pelligrino and Thomasma argue, that the principle is not simply “a duty in the Kantian sense, but [is] part of her character . . . and identity.”20 A virtue-centered ethics of medicine weighs principles, facts, and patient contexts to perform the actions necessary for an ethical account of medicine. Prudence, justice, fortitude, and temperance all appear in this lens. One must recognize the needs of the patient (justice), resolve to commit to a course of treatment even when difficult (fortitude), know the right amount of treatment needed (temperance), and weigh treatments alongside other principles, facts, and contexts in certain situations (prudence). The ends of health care shape characters, and the cardinal virtues provide the means to apply those ends in varying contexts.
The patient also plays a role in this context. In this analysis, the patient and doctor enter a unique relationship that binds them at the site of health. Each works to clarify, represent, and find the facts, principles, and contexts of the situation in need of health care to achieve “the good for the patient.”21 The quality of care given to the patient through this relationship “rests in the ordering of principles and concrete lived realities at the moment” of medical treatment.22
The unique role of virtue ethics for health care workers is an instructive point of departure from the practices of COVID capitalism and neoliberalism, namely autonomy and competition. First, the “caring bond” between patient and health care worker destabilizes the myth of autonomy. The patient requires the expertise and training of the doctor or health care worker, but the doctor also requires patient input to secure facts and contexts necessary for determining appropriate medical treatments. In short, no pure autonomy exists, only relative autonomy in a mutually cooperative relationship of medical practice. Furthermore, medicine’s unique role illustrates that its purpose must exist outside the logic of competition. Pelligrino and Thomasma argue that one should distinguish medicine from the practice of business in which “by contrast the competitor’s vulnerability is something to be exploited”; rather, medicine must treat and care for the vulnerabilities of others.23
Dependency and vulnerability: Virtue ethics and the general public against COVID capitalism
A virtue ethics approach to medical care enables a new moral vision for doctors and medical care workers. The roles inhabited by the characters of doctors are distinct, however, from the general public. The two groups’ virtues are connected but distinct. The virtue ethics approach charted for medical ethics must include a place for the public so the character formation essential for virtuous medicine may proceed unhindered. Medical care workers already possess practices and principles inherent in their profession. The general public, too, must rely on principles and practices that enable virtues for prudential weighing of communal health. These practices and principles must directly resist the formative power of autonomy and competition. I contend that two such practices, dependence and vulnerability, shift citizens away from autonomy and competition while concurrently opening them to realizing the possibility of health as a communal concern and care for the elderly specifically. These practices cultivate the virtues necessary for a community to weigh economic goods and health care.
Carol Gilligan illustrates the shift to dependence necessary for overcoming competition and autonomy. She says she has always “sought to represent the voices of contemporary American girls and women as they talked about moral conflict and choice, and to amplify and validate these voices by associating them with the voices in western literature.”24 Through focusing on women’s voices, Gilligan offers a tangential account of moral discernment free from many of the pitfalls of Western philosophy and neoliberalism.
Gilligan’s shift occurs within her proposed reframing of the moral question. Traditionally, moral discernment occurs through presenting an agent with a moral problem and asking, “How would you resolve it?”25 This method, Gilligan argues, arises in Piaget’s analysis of a child’s development. In the Western tradition, especially post-Enlightenment, moral development “consists of a system of rules.”26 Through learning how to deploy rules, a child can learn how to respect the rules. However, Gilligan proposes a new means of judging one’s moral sensibilities. She writes, “I did not begin by posing a moral problem and asking, ‘How would you resolve it?’ Rather, I asked people how they would define what a moral problem is.”27 Through experience and decisions made in the individual’s life, Gilligan finds a different account of the moral life than a system of rules.
Studying primarily young pregnant women considering an abortion, Gilligan found that these individuals did not first approach a system through a system of rules, but rather, these women made their decisions through their understanding of “responsibility.”28 This moral emphasis shifts the entire narrative of moral decision-making. The group’s moral quandary, namely abortion, did not match the system of rules approach sustained by the wider public. This recognition is significant in Gilligan’s approach because it offers a new way to think of moral discernment. She writes,The full view of choice, of the relationship between other and self, was fundamentally different. Choice, rather than being seen as an isolated moment, was a moment in an ongoing narrative of events, which in the abortion decision were specifically the events of the relationship.29
The events leading to the pregnancy, the events during the pregnancy, and the possible future events if the birth would or would not occur are all considered in the choice. Therefore, the decisions were made in the context of the relationship between the community and the agent rather than through rules. “There was,” Gilligan continues, “no way to separate self and other into a distinct opposition.”30 Any rules or norms are negotiated, not in isolation from one’s support community, but only within them.
Gilligan’s different voice in moral discernment stands in direct opposition to the moral thinking embedded within COVID capitalism. As previously stated, COVID capitalism relies heavily on autonomy. In that logic, relying on others is morally distasteful; indeed, lasting friendships hinder the ability to choose. Instead, one must possess economic mobility and detachment as two means to express autonomy. The resources one accumulates are for the express purpose of self-sufficiency, and individuals compete to become the most autonomous beings. Therefore, the rules of neoliberalism are a system that must be applied, even in a pandemic, to make moral decisions.
Against autonomy, Gilligan’s method of moral discernment uncovers another virtue: dependence. Gilligan recognizes that this term carries significant baggage in our culture because autonomy is regarded as the highest virtue and dependence is viewed as the greatest weakness.31 Yet, when Gilligan asked a different group of young women the meaning of dependence, they revealed an alternative account. She observes,The girls conveyed the assumption that dependence is positive, that the human condition is a condition of dependence, and that people need to rely on one another for understanding, comfort, and support . . . Dependence, rather, was created by choices to be there for others, to take care of them, to listen, to try to understand, and to help.32
Dependence, in the eyes of these young women, expresses a relationship of care. Furthermore, dependence is not the malformed “passivity” of humankind’s natural capacities in favor of overreliance on others.33 Rather, it means to these women that “someone would be there when you need them.”34 Therefore, dependence is an “active” choice that sustains relationships of moral community and care.35
Gilligan’s account of dependence provides an appropriate alternative to COVID capitalism because it leads to human flourishing during a pandemic. Evaluating Gilligan’s shift to dependence suggests that the good inherent in COVID capitalism is profoundly flawed and limited. Only through reliance on others does care emerge as a possibility. This move to dependence will necessarily lead to the formation of a practical reason that can be correctly discerned in a world of differing goods. When dependence becomes a central practice, ethics shifts from rules to responsibility. In a moral community, as the one in which the general public is bound in medicine, one must ask for whom we are responsible. To depend on others is not a moral weakness, but a recognition of the ones with whom we share the responsibility of care. Specifically, elderly persons serve as a reminder that the young once needed the care of the elderly. “Honor your father and mother, so that your days may be long in the land that the Lord your God is giving you” (Exod 20:12, NRSV). In this passage, God commands the Hebrew people to recognize their own needs as children for their parents, as well as their need as adults to treat their aging parents well. The mutuality in the command entails responsibility for the old to the young and the young to the old. With the recognition that humans flourish through mutual dependence, medicine emerges as a distinct possibility.
In addition to dependence, a shared vulnerability must also emerge as a distinct practice, contrary to autonomy and competition. In Alasdair MacIntyre’s work, one finds an account of virtue that undergirds all ethics.36 Still, MacIntyre recognizes that any virtue ethics that does not consider those impacted by “bodily illness and injury, inadequate nutrition, mental defect and disturbance, and human aggression and neglect” is inadequate for the task of moral reflection.37 Much like Gilligan, MacIntyre strays from traditional narratives of autonomy and competition toward dependence-based ethics. And, in addition to dependence, MacIntyre places equal weight on vulnerability. Practical reasoners are vulnerable at one stage or another and thus are dependent upon others. In our vulnerability, due to sickness or other deficiencies in our ability, we desire, as Gilligan already expressed, to know someone will be present.
This desire to know the presence of another illustrates the importance of vulnerability as it leads to dependence. Even though a communities’ most basic moral impulses lie in desire satisfaction, MacIntyre acknowledges that desire satisfaction is merely egocentric.38 This impulse easily maps over the competition/autonomy of COVID capitalism. The charitable actions of the economy’s so-called “winners” can be read as benevolent, but only insofar as they already possess these goods. The COVID capitalist society prefers the benevolent winner over goods held in common. MacIntyre writes that the divisionbetween self-interested market behavior . . . and altruistic, benevolent behavior on the other, obscures from view just those types of activity in which the goods to be achieved are neither mine-rather-than-others’ nor others’-rather-than-mine, but instead are goods that can only be mine insofar as they are also those of others, that are genuinely common goods, as the goods of networks of giving and receiving.39
Under the twofold emphases on competition and autonomy, the recognition of shared networks cannot be honored.
COVID capitalism might argue that shared goods are a fiction and that any goods available in the market are finite; thus, one must compete for them against one’s neighbors to obtain them, including in medicine, where the weak, the elderly, and vulnerable must be sacrificed so competition can continue. MacIntyre challenges such a view by placing shared recognition of vulnerability at the center of communal life. MacIntyre counters that “the basic political question is what resources each individual and group needs.”40 The question is, then, not which groups are worthy of salvation through the market, but how the market obscures vulnerability and our need for one another. The problem emerges when communal relationships exist in conflict with one another rather than participating in a shared vulnerability.41
Competition allows, to some extent, recognition of a shared vulnerability but ultimately requires that one exploit it. On the one hand, is the vulnerability associated with the economy and, on the other hand, the elderly. The COVID capitalism of the present pandemic pits these two groups against each other, adjudicating which group must be saved according to their productivity.42 For MacIntyre, this moral discernment is counterfeit because it fails to encourage the virtues of “giving and receiving” necessary for the awareness of the “common goods and common needs.”43 Vulnerability encourages these virtues instead through a shared recognition of our common needs and goods. MacIntyre writes, “Those who are [not yet old must] recognize in the old what they are moving towards becoming.”44 In the case of COVID capitalism, one must adjudicate between age groups based on which group can be eliminated and still maintain productivity. In vulnerability, there are shared needs, but these needs are not barriers to shared goods. In short, vulnerability prioritizes care for each other over any economic calculus. Without the virtues essential to giving and receiving, this “awareness [of shared vulnerability] cannot be achieved.”45
Vulnerability thus leads to an account of the good and of the goods preferable to COVID capitalism. It is not preferable because it enables a better way to decide who lives and dies, but because it rejects the false dichotomy altogether. In COVID capitalism, one cannot save those most susceptible to COVID-19 and the economy, and this either-or strategy does not exist in MacIntyre’s account of giving and receiving. To flourish despite COVID capitalism, one needs both MacIntyre’s vulnerability and Gilligan’s dependence. MacIntyre maintains that “we would need . . . to be able to receive from others what we need them to give to us and to give to others what we need to receive from them.”46 We recognize our vulnerability, namely our need and our dependence on others to fulfill that need, as they depend upon us to provide for them.
Dependence and vulnerability orient listeners rightly to the comments made by Lt. Governor Patrick. Regarding his first position, Patrick states, “There are more important things than living. And that’s saving this country for my children and my grandchildren . . . we’ve got to take some risks and get back in the game and get this country back up and running.”47 In this, he recognizes his shared vulnerability and that he does not want to die, but this recognition is not his error. Patrick’s error is the assumption that his children and grandchildren benefit more from economic advantage than from the presence of their father and grandfather. Of course, they too are vulnerable, but, Gilligan argues, he helps them negotiate their moral development. MacIntyre stipulates that the elderly cannot be neglected due to their lack of productivity, but instead that they offer wisdom extending beyond their economic utility. The value Patrick brings his children and grandchildren cannot be quantified. Furthermore, their lives cannot be quantified in relation to him. He proves this in his willingness to die for them. They both share their need for one another and, thus, are vulnerable. In turn, they depend on one another to fulfill their shared needs. This shift serves not only a need for presence but of moral discernment, as well as the virtues of giving and receiving that are capable of orienting the economy to society’s shared goods rather than goods to the economy. Only through a new moral formation can the weeds of economic reason be uprooted and a new way of relating begin to take root. Furthermore, only through this recognition can actions, such as a lockdown, emerge to make medicine a moral possibility.
In addition to the counter formation enabled by the practices of dependence and vulnerability for a fuller appreciation of health care and the elderly, these practices also allow for a deeper appreciation of an economic flourishing not tied to COVID capitalism. These practices cultivate the flourishing envisioned by the cardinal virtues as eudaimonia rather than as individual wealth. These practices present one’s neighbors not as barriers to true flourishing, as do competition and autonomy, but rather as essential to it. The virtues are communal in that one’s ability to exercise them requires the presence of others as the promise and fulfillment of the moral life. Flourishing, in the virtuous sense, requires this communal perception and shared human dignity. The ends of a well-formed agent, therefore, include the flourishing of others.
In COVID capitalism, however, the ends of the neighbor must be sacrificed so the individual can cultivate their autonomy and compete. The inability of the neighbor, through exposure, to compete is a weakness to be exploited. Thus, when the agent finds their ends exclusively in the logic of the market, exorbitant pressure weighs on the health care community.
So, in contrast, envisioned through virtue ethics and eudaimonia, the neighbor’s weaknesses must not be exploited. Rather, the needs and vulnerability of the neighbor reflect back on the individual’s needs and vulnerability. As in the doctor-patient relationship, a relationship of care emerges as a more pressing need than competition. A person’s flourishing diminishes in the face of one’s neighbor’s suffering rather than in one’s own ability to consume and produce. Thus, those with means will, in times of economic scarcity, surrender their material means to care for those who are sick and dying, because a crisis is a time to exercise justice, prudence, fortitude, and temperance. The ability of the virtuous to locate economics as a relative and not absolute good only comes from knowing their flourishing vis-à-vis another’s flourishing. The greater risk for the virtuous is not a lack of economic success but rather a lack of character in the face of crisis. One would hope that these values might be the resources available in times of disease rather than the acute formations of neoliberalism.
Conclusion
During the rise of COVID-19 cases in the United States, a young social worker in the Veteran Affairs Hospital in the Midwest maintained the cases of veterans nearing the end of their lives.48 One such elderly veteran was within a few days of death and wished to communicate with his wife, who was also confined to a nursing home since she was advanced in years and could not visit without fear of contracting the disease herself. The man was blind and hard of hearing and thus could not talk to her over the phone or easily write her letters. Therefore, this young social worker, compelled by a sense of eudaimonia derived from the cardinal virtues, offered to communicate on the veteran’s behalf with his wife and communicate her responses back to him. In the final days, the social worker communicated messages of love, gratitude for a lifetime of memories, and fidelity to the very end. In these deeply personal messages, one glimpses the humanity possible in a community of virtue, one not to be found in COVID capitalism. Not only are the elderly veteran and the wife dependent upon and vulnerable to the social worker for messages from their beloved one, but the social worker depends on the couple to illustrate and embody the virtue of fidelity. They share needs and remain dependent on one another to fulfill them. The couple and the social worker share something inexplicably sacred and, dare I say, good. This sacred value is not supported and guaranteed by the economy, solely because the sacred character of this deeply human experience cannot be bought. Rather, it can only be shared as the love between them. The social worker does not treat the elderly couple through utility but as sacred persons of shared vulnerability and dependence. The proper use of medicine thrives in the context of such care because this care does not merely treat symptoms of the disease but the whole complex, fragile person and their flourishing.
The lesson learned in this story is that virtues are necessary for dependence and vulnerability. The act of kindness embodied by the social worker illustrates her recognition of her vulnerability in this space and their dependency on her. The couple’s commitment to love shows their need to be vulnerable to and dependent on the social worker. Mutual dependence and vulnerability by both parties accommodate a shared humanity. Humans are not price-points on a spreadsheet but are deeply fragile fellow creatures held in God’s love and care. Unlike neoliberal rationality, the goal of medicine is not autonomy governed by competition. Rather, it requires the opposite. Society must envision vulnerability and dependence as character traits that define people’s actions in this crisis. MacIntyre writes:To identify an occurrence as an action is in the paradigmatic instances to identify it under a type of description that enables us to see it as flowing intelligibly from a human agent’s intentions, motives, passions, and purposes. It is, therefore, to understand an action as something for which someone is accountable, about which it is always appropriate to ask the agent for an intelligible account.49
Generations after the current one will look back to this time and ask for an account of society’s actions. One may hope that vulnerability and dependence will hold the key to such an account. Gilligan’s insight teaches that humans negotiate moral decisions in relationship to others. For better or worse, they will look to us for help. We are their tradition, and we are their ancestors. Like the social worker, these actions might be “unhistoric,” but I hope that people will contribute to the growing good of shared humanity for the world and that one day society might not be so ill as most individuals live these hidden lives.
Author biography
Hank W. Spaulding III is an Associate Campus Pastor and Associate Professor of Christian Ethics at Mount Vernon Nazarene University. He is also an Adjunct Professor of Christian Ethics at Ashland Theological Seminary, Ashland University, and George Fox University. He is the author of The Just and Loving Gaze of God with Us: Paul’s Apocalyptic Political Theology (Wipf & Stock, 2019).
1. Edmund D. Pellegrino and David C. Thomasma, The Virtues in Medical Practice (New York: Oxford University Press, 1993), 3.
2. Wendy Brown, Undoing the Demos: Neoliberalism’s Stealth Revolution (New York: Zone Books, 2015), 17.
3. Milton Friedman, Capitalism and Freedom, Fortieth Anniversary ed. (Chicago: University of Chicago Press, 2002), xiv.
4. Adam Kotsko, Neoliberalism’s Demons: On the Political Theology of Late Capital (Stanford: Stanford University Press, 2018), 12.
5. See, for example, Immanuel Kant’s discussion of heteronomy in Groundwork of the Metaphysics of Morals, The Cambridge Edition of the Works of Immanuel Kant, trans. and ed. Mary Gregor (New York: Cambridge University Press, 1996), 82–83.
6. Rover Skidelsky and Edward Skidelsky, How Much is Enough? Money and the Good Life (New York: Other Press, 2012), 165.
7. Kevin Hargaden, Theological Ethics in a Neoliberal Age: Confronting the Christian Problem of Wealth (Eugene, OR: Cascade, 2018), 23.
8. See Skidelsky and Skidelsky, How Much is Enough? 37.
9. Philip Goodchild, Theology of Money (Durham: Duke University Press, 2009), xiii.
10. Adrianna Rodriguez, “Texas’ Lieutenant Governor Suggests Grandparents are Willing to Die for U.S. Economy,” USA Today, March 24, 2020, https://www.usatoday.com/story/news/nation/2020/03/24/covid-19-texas-official-suggests-elderly-willing-die-economy/2905990001/.
11. Rodriguez, “Texas’ Lieutenant Governor.”
12. “Nurses on the Workforce,” Practice and Advocacy, American Nurses Association, https://www.nursingworld.org/practice-policy/workforce/.
13. Kara Gavin, “One-Third of Older Americans Delayed in Health Care Over COVID Concerns,” Michigan Health, June 17, 2021, https://healthblog.uofmhealth.org/wellness-prevention/one-third-of-older-americans-delayed-health-care-over-covid-concerns.
14. Theodore Lytras and Sotirios Tsiodras, “Total Patient Load, Regional Disparities and In-Hospital Mortality of Intubated COVID-19 Patients in Greece, from September 2020 to May 2021,” Scandinavian Journal of Public Health 50.6 (August 2022): 671–75, https://doi.org/10.1177/14034948211059968.
15. Cathleen Kaveny, quoted in Sarah Pulliam Bailey, “Should Older Americans Die to Save the Economy? Ethicists Call It a False Choice,” The Washington Post, March 24, 2020, https://www.washingtonpost.com/religion/2020/03/24/dan-patrick-economy-coronavirus-deaths-notdying4wallstreet/?fbclid=IwAR14CChE5gqEsgDgSQb4uhVR4jZgaRp28sKG0enr2hFFLt7VcYQ6rsrfBXY.
16. Kyle B. T. Lambelet, ¡Presente!: Nonviolent Politics and the Resurrection of the Dead (Washington, DC: Georgetown University Press, 2019), 6.
17. Kaveny, quoted in Baily, “Should Older Americans Die?”
18. For a review of the impact on health care workers’ retention, see Scottie Andrew, “Traumatized and Tired, Nurses Are Quitting Due to the Pandemic,” CNN, February 25, 2021, https://www.cnn.com/2021/02/25/us/nurses-quit-hospitals-covid-pandemic-trnd/index.html.
19. Office of the U.S. Surgeon General, Addressing Health Worker Burnout: The U.S. Surgeon General’s Advisory on Building a Thriving Health Workforce (Rockville, MD: U.S. Department of Health and Human Services, 2022), https://www.hhs.gov/sites/default/files/health-worker-wellbeing-advisory.pdf
20. Pellegrino and Thomasma, Virtues in Medical Practice, 22.
21. Pellegrino and Thomasma, Virtues in Medical Practice, 23.
22. Pellegrino and Thomasma, Virtues in Medical Practice, 23.
23. Pellegrino and Thomasma, Virtues in Medical Practice, 37.
24. Carol Gilligan, “A Different Voice in Moral Decisions,” in From Christ to the World: Introductory Readings in Christian Ethics, ed. Wayne G. Boulton et al. (Grand Rapids: Eerdmans, 1994), 172.
25. Gilligan, “Different Voice in Moral Decisions,” 173.
26. Gilligan, “Different Voice in Moral Decisions,” 173.
27. Gilligan, “Different Voice in Moral Decisions,” 173.
28. Gilligan, “Different Voice in Moral Decisions,” 173, emphasis original.
29. Gilligan, “Different Voice in Moral Decisions,” 173.
30. Gilligan, “Different Voice in Moral Decisions,” 173.
31. Gilligan, “Different Voice in Moral Decisions,” 175.
32. Gilligan, “Different Voice in Moral Decisions,” 175.
33. Gilligan, “Different Voice in Moral Decisions,” 176.
34. Gilligan, “Different Voice in Moral Decisions,” 175.
35. Gilligan, “Different Voice in Moral Decisions,” 176.
36. See Alasdair MacIntyre, After Virtue: A Study in Moral Theory, 3rd ed. (Notre Dame: University of Notre Dame Press, 2007).
37. Alasdair MacIntyre, Dependent Rational Animals: Why Human Beings Need the Virtues (Chicago: Open Court Press, 1999), 1.
38. MacIntyre, Dependent Rational Animals, 68–69.
39. MacIntyre, Dependent Rational Animals, 119.
40. MacIntyre, Dependent Rational Animals, 144.
41. MacIntyre, Dependent Rational Animals, 144.
42. MacIntyre, Dependent Rational Animals, 146.
43. MacIntyre, Dependent Rational Animals, 146.
44. MacIntyre, Dependent Rational Animals, 146.
45. MacIntyre, Dependent Rational Animals, 146.
46. MacIntyre, Dependent Rational Animals, 155.
47. Alex Samuels, “Dan Patrick Says ‘There Are More Important Things Than Living and That’s Saving This Country,’” The Texas Tribune, April 21, 2020, https://www.texastribune.org/2020/04/21/texas-dan-patrick-economy-coronavirus/?utm_campaign=trib-social&utm_content=1587474082&utm_medium=social&utm_source=facebook&fbclid=IwAR07Bl8EVz8AWURx5ZeXfMIcYHXQwanq0RIjKgsDFkGOZ5d3EQ5txCI3jg4.
48. I have stripped the name and details of the patient and worker to preserve anonymity and protect patient records.
49. MacIntyre, After Virtue, 209.
| 0 | PMC9742506 | NO-CC CODE | 2022-12-14 23:30:02 | no | 2022 May; 119(1-2):86-99 | utf-8 | null | null | null | oa_other |
==== Front
RAE
sprae
Review & Expositor
0034-6373
2052-9449
SAGE Publications Sage UK: London, England
10.1177/00346373221133843
10.1177_00346373221133843
Thematic Words · · ·
A theological and ethical analysis of the response of the Eastern Orthodox to the COVID-19 pandemic
LeMasters Philip
McMurry University, USA
Philip LeMasters, McMurry University, 1400 Sayles Blvd., Abilene, TX 79697, USA. Email: [email protected]
5 2022
5 2022
5 2022
119 1-2 110121
© The Author(s) 2022
2022
Review & Expositor, Inc
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
The response of the Eastern Orthodox Church to the COVID-19 pandemic reflects its distinctive theological and liturgical traditions as well as its decentralized system of governance. Foundational beliefs and practices inform Orthodoxy’s understanding of the imperative to care for the physical well-being of the sick. Points of disagreement arose in Orthodox communities concerning public health restrictions on attendance at the Divine Liturgy, the use of a common communion spoon, whether diseases may be transmitted through the Eucharist, and the appropriateness of receiving vaccinations tested or produced with cell lines derived from the tissue of aborted fetuses. Such contested matters reflect points of tension between characteristic beliefs and practices of Orthodoxy and its commitment to care for the health of neighbors during a global pandemic.
abortion
communion spoon
COVID-19
Eastern Orthodoxy
Orthodox bioethics
pandemic
typesetterts1
==== Body
pmcIntroduction
The response of the Eastern Orthodox Church to the COVID-19 pandemic reflects its distinctive theological and liturgical traditions as well as its decentralized system of governance. This article describes the rich foundational beliefs and practices that inform Orthodoxy’s understanding of the imperative to care for the physical well-being of the sick. It also analyzes points of disagreement in Orthodox communities concerning public health restrictions, communion practices, and the appropriateness of receiving vaccinations tested or produced with cell lines derived from the tissue of aborted fetuses. Such contested matters arise from points of tension between characteristic beliefs and practices of Orthodoxy and its commitment to care for the health of neighbors during a global pandemic.
Background and context
The Eastern Orthodox Church is a communion of churches with shared theological affirmations, liturgical practices, and spiritual disciplines. The Church understands itself to embody the fullness of the Church founded by Jesus Christ and to maintain the beliefs and practices of apostolic Christianity. The schism of Orthodoxy and Catholicism dates from 1054 CE and occurred due to different understandings of papal authority and other theological disagreements. In distinction from the dynamics of Catholic-Protestant relations in the West, the Eastern Roman or Byzantine experience shaped Orthodoxy decisively. The vast majority of Orthodox Christians live in historically Orthodox nations, such as Russia, Romania, and Greece, while adherents have been present in the West in noticeable numbers only since the end of the nineteenth century. Most of their communities outside of traditionally Orthodox regions stem from immigration, but Orthodoxy has received an increasing number of converts in recent decades.1
The word “Orthodox” has the literal meaning of both “right worship” and “right belief.” The service of eucharistic worship in the Orthodox Church is called the Divine Liturgy, the two most common forms of which are associated with saints Basil the Great and John Chrysostom. Orthodoxy was not shaped decisively by Medieval Scholasticism or the Enlightenment and appeals to natural law or other standards of morality universally known by reason are not as prominent as they are in many forms of Western Christianity. Consequently, discussions of both doctrinal and moral matters often draw heavily on the language and practices of sacramental worship, which manifest the corporate faith of the Church.2
The Divine Liturgy provides a shaping context for how Orthodoxy envisions God’s purposes for social interaction, including how people and societies should respond to persons in sickness or other types of need. After initial petitions “For the peace from above and for the salvation of our souls . . . [and] For the peace of the whole world,” prayers followFor favorable weather, for an abundance of the fruits of the earth, and for peaceful times . . . For those who travel by land, sea, and air, for the sick, the suffering, the captives and for their salvation . . . [and] For our deliverance from all affliction, wrath, danger, and necessity.
The petitions after the consecration of the eucharistic gifts in the liturgy of Basil the Great call for God toFree those who are held captive by unclean spirits; sail with those who sail; travel with those who travel; defend the widows; protect the orphans; liberate the captives; heal the sick. Remember, Lord, those who are in mines, in exile, in harsh labor, and those in every kind of affliction, necessity, or distress; those who entreat your loving kindness; those who love us and those who hate us; those who have asked us to pray for them, unworthy though we may be. Remember, Lord our God, all Your people, and pour out Your rich mercy upon them, granting them their petitions for salvation.3
These appeals for peace and salvation do not exclusively concern hope for eternal life in the eschatological consummation of the heavenly reign. They address tangible, mundane circumstances faced by persons and social groups in the present. These petitions occur at the very beginning of the service and immediately after the most solemn moment of the Divine Liturgy when the priest recalls the words of institution and prays for the Holy Spirit to descend upon the bread and wine such that they become the body and blood of Christ. That these concerns are placed in such positions of prominence indicates their great importance, as well as the obligation of communicants to care for the suffering physical bodies of their neighbors.4
Taking its sensibilities from the repudiation of Gnostic denigrations of the body in the early centuries of Christianity, Orthodoxy rejects escapist forms of spirituality that view physical realities as being spiritually irrelevant. An exalted view of materiality reflects beliefs in the goodness of creation, the full humanity of Jesus Christ, his bodily resurrection and ascension, and bread and wine becoming truly his body and blood in the Eucharist. The healing of the sick was a paradigmatic characteristic of Christ’s ministry as a sign of the blessedness of God’s reign, which also manifests the great importance of caring for suffering persons. Orthodoxy teaches that human persons are icons of God, as they bear the divine image and likeness. How people treat an image of something reveals their relationship to it. Consequently, those who disregard people in obvious need enact their rejection of God. As stated in 1 John 4:20-21 (NRSV),Those who say, “I love God,” and hate their brothers or sisters are liars; for those who do not love a brother or sister whom they have seen, cannot God whom they have seen. The commandment we have from him is this: those who love God must love their brothers and sisters also.5
From its origins as described in Acts, the Church continued healing the sick through the philanthropic ministries of the apostles. Even the pagans of Rome noted that the early Christians literally risked their lives to care for victims of plague, which demonstrates that a deep commitment to therapeutic practice remained characteristic of the believing community. The category of Orthodox saints known as “the holy unmercenary healers” includes physicians who cared for the ill without charging fees. Their veneration in the Orthodox Church stands as a reminder that the ministry of healing is a sign of Christ’s merciful deliverance of humankind from debilitation and death, consequences of the fall of humanity due to Adam and Eve’s primal disobedience.6
Given its affirmations of the goodness of the body and the imperative to care for the sick as a means of participating in Christ’s love for humankind, Orthodoxy honors the healing vocation and does not view the use of medicines or therapies with suspicion. To the contrary, it welcomes them as divine blessings that convey the love of God for humanity. As stated in the Wisdom of Sirach 38,Honor physicians for their services, for the Lord created them; for their gift of healing comes from the Most High, and they are rewarded by the king. . . . The Lord created medicines out of the earth, and the sensible will not despise them. . . . And he gave skill to human beings that he might be glorified in his marvelous works. By them the physician heals and takes away pain; the pharmacist makes a mixture from them. God’s works will never be finished; and from him health spreads over all the earth. (vv. 1-2, 4, 6-8 NRSV)
While the Orthodox Church prays for the sick and provides the sacrament of holy unction for the healing of soul and body to its ailing communicants, the use of conventional medical therapies is neither prohibited nor discouraged.7 The origins of the modern hospital are in the philanthropic endeavors of the early centuries of the Eastern Roman Empire, such as the Basiliad of St. Basil the Great, which employed the treatments and methods of health care used at that time.8 Likewise, the holy unmercenary healers availed themselves of medicines and therapeutic procedures, as well as prayer.9 In the twentieth century, Saint Luke of Simferopol, an Orthodox archbishop, surgeon, and political prisoner, combined steadfast piety with groundbreaking scholarly publications and lifesaving surgeries for Soviet soldiers in World War II.10
In order to ease pain, restore bodily function, and otherwise provide the best care possible for the living icons of Christ who present themselves to healthcare professionals, Orthodoxy endorses using medical science’s current therapeutic methods. While death is inevitable and the preservation of physical life is not the highest good, love for suffering neighbors requires caring for the embodied person as a living icon of God.11
Limiting attendance at liturgy
As of this writing, over 6 million people have died from the pandemic.12 The disease has no cure, even as highly contagious variants develop and marginalized populations suffer disproportionally from its effects.13 As the Church has not been immune from the political and social tension associated with the pandemic, factors other than the philanthropic vision of Orthodoxy have played important roles in shaping the responses of clergy and laity. In this light, the Assembly of Canonical Orthodox Bishops in the United States of America urged the faithful to embrace “an authentic life of patient obedience, sincere humility, genuine compassion, and sacrificial love even towards those with whom we differ.”14
Different ecclesiastical jurisdictions have responded in different ways to the many challenges of the pandemic. Orthodox churches around the world limited or forbade lay attendance at the Divine Liturgy, as well as other services, during the initial outbreak of the pandemic, which began during the season of Lent, the most spiritually intense and liturgically rich time of the church year. “A limited attendance” of Orthodox laity at services for Pascha (Orthodox Easter) in 2020 was apparently permitted only in Georgia and Bulgaria.15 In forums ranging from diocesan assemblies to parish councils and Internet discussion boards, debates ensued on whether restricting attendance by the laity at the Divine Liturgy was a necessity to be embraced out of loving concern for the safety of one’s neighbors or a sign of fear that is contrary to hope for eternal life.
In order to respond with moral integrity to the pandemic, Eastern Orthodoxy’s commitments to love the neighbor, provide care for sick, and treat even the lowliest people as those who bear God’s image should shape decisively how the Church responds to COVID-19. Some of the most basic characteristics of the sacramental life of the Church have presented challenges to doing so. The Orthodox Church is highly liturgical, teaching that the celebration of the Divine Liturgy is a mystical entrance to the heavenly kingdom. The service manifests the identity of the Church as the body of Christ with many members, a communion of persons united in love as an image of the Holy Trinity. Regular attendance at Liturgy is an expected spiritual discipline for all parishioners, though in practice often enacted by only the most pious. Such practices do not, however, eliminate the spiritual and moral import of how people’s behavior impacts the well-being of their neighbors who bear the image of God.
In the midst of a global pandemic, compassionate concern for others has required new and unprecedented sacrifices, including following the guidance of public health officials in order to slow the spread of the pandemic. Especially in the early months of the pandemic, bans on large gatherings and requirements for social distancing challenged the practices of liturgical worship and were often met with resistance. As Ioannis Kaminis notes,The many controversies over such restrictions raise the question of whether receiving Holy Communion and participating in the Eucharist has not been transformed into an ordinary ritual, a religious duty that forgets the word of Jesus: “The Sabbath was made for humankind, and not humankind for the Sabbath.” (Mk 2:27)16
Since the Eucharist manifests believers’ communion with Christ, insistence on celebrating religious services without regard for the health and well-being of the sick surely contradicts the demands of discipleship.
Important models within Orthodoxy are a guide for strengthening the spiritual life apart from corporate, sacramental worship. Metropolitan Joseph, the primate of the Antiochian Orthodox Christian Archdiocese of North America, appealed to the ancient example of Palestinian monks who spent each Lent alone in the Sinai desert and abstained from the Eucharist during this time in order to inspire parishioners who could no longer attend the Divine Liturgy due to bans on social gatherings. This example is especially fitting as restrictions on attendance began during the season of Lent. As well, one of the saints commemorated during this penitential season is Mary of Egypt, whom the monk Father Zosimas encountered in the Sinai desert when he was following his monastery’s practice in this regard. The Metropolitan wrote, “Although our time of social distancing is not quite the same type of asceticism, let us treat it as a sacrificial offering of love to God and our neighbor.”17
In refutation of criticisms concerning such restrictions, Metropolitan Joseph stated thatWe did not ask our faithful to worship from home out of a fear of death or a belief that our churches or sacraments are carriers of disease. Our world was confronted by a novel virus to which no one had yet been exposed and no doctor had yet learned to treat. In addition to those factors, the virus could be spread before the onset of symptoms by people unaware they were sick. We were asked to join our local communities in slowing the spread of the virus by not gathering in crowds. This was to prevent an overwhelming of the healthcare system, allowing the doctors and nurses to give adequate care to the sick and thus avoid unnecessary deaths.18
Similarly, Metropolitan Tikhon of the Orthodox Church in America appealed to both the norm of love for neighbor and the ascetical practices of Lent in writing that precautionary measures and restrictionshave been taken as our Christian response to protect our brothers and sisters. Our Lord tells us, “Greater love has no man than this, that a man lay down his life for his friends (John 15:13).” The life we “laying down” [sic] now is our normal life, because these are extraordinary times. We are making a sacrificial effort, which is in keeping with the present season of repentance and ascetical striving. Like the ascetics of old who would depart from their monasteries for the forty days of Lent in preparation for Holy Week, we should take this opportunity to prayerfully reflect on our life in Christ and increase our desire to be with Him.19
Not all ecclesial jurisdictions were led at the outbreak of the pandemic with such philanthropic concern. Kaminis speculates that most Orthodox churches around the world that followed the guidance of public health officials “decided to act due to government pressure and thus reluctantly accepted the government measures.”20 He notes that “some spiritual elders from Mt. Athos . . . supported and spread the conspiracy theory that a microchip could be slipped into the vaccines against Covid-19.”21 The Moldovan Orthodox Church also made statements supporting such fears.22 Lax adherence to public health guidelines or their repudiation led to the death of Patriarch Irinej of the Serbian Orthodox Church and of other bishops there. While the Russian Orthodox Church quickly endorsed the restrictions required by the government, many clergy and laity have opposed them.23 How widely clergy and laity are disregarding them is not clear, but “almost all large Russian monasteries have become hotbeds for spreading Covid-19, and in many of them the mortality rate of priests and monks is quite high.” Prominent bishops, priests, and monks from Russia, Ukraine, and Belarus have spoken publicly against public health restraints.24
Communion during COVID
In addition to tensions over restricting attendance and requiring social distancing and other public health practices, the manner of serving communion in the Orthodox Church has become a point of controversy. With degrees of frequency varying widely, during the Divine Liturgy, Orthodoxy laity receive the body and blood of Christ from a spoon that the priest or deacon uses to place communion in their mouths. Concern arose about the potential of spreading the virus through the use of the communion spoon. Early in the pandemic, some jurisdictions replaced the common spoon with disposable individual spoons or required disinfecting the spoon after its use by each communicant. Others instructed the laity to open their mouths as widely as possible so that the clergy could administer communion without the spoon touching the tongue or other parts of the mouth.25
With reference to icons, crosses, and other physical items that clergy and laity commonly kiss as a sign of honor, Archbishop Elpidophoros of the Greek Orthodox Archdiocese of America acknowledged that “material elements that can convey the blessings of God are also subject to the broken nature of our fallen world.” He prohibited kissing them at the outbreak of the pandemic, but he continued to require distribution of communion with a common spoon, for as “the sacrament of sacraments, the Holy Eucharist, is not simply a material element but the very body and blood of our Lord Jesus Christ.”26 Metropolitan Joseph of the Antiochian Orthodox Archdiocese of North America taught similarly:We have stated throughout these difficult days that Holy Communion is the “medicine of immortality,” not a vector of disease. We have also consistently stated that our method of distributing the Holy Gifts is not open to question.27
As Eugenia Constantinou wrote in support of maintaining this practice, “Whether it is received on a common spoon or not, the most sacred Mystery of the Church can never be the vehicle of illness.”28
Others perceived the method of administering the Eucharist as not being of decisive theological importance, especially since use of the spoon developed only in the eleventh century, and they affirmed the legitimacy of serving communion in a manner less likely to spread the virus. Sister Vassa Larin notes evidence that “communion in some areas was still given to the faithful in the hand, and in two separate species” until the twelfth century. The development of this practice seems to have reflected “possible abuses and irreverence on the part of the laity.”29 In the earlier history of the Church, the laity received the elements from the hand and the chalice, respectively, as Orthodox clergy continue to do. Rejecting the view that the use of a single spoon is a dogmatic matter, Larin argues that its development was for “practical and pastoral reasons” and has come to symbolize “the tongs, with which Isaiah received the heavenly coal” in Isaiah 6.30 In countries like Germany and Austria in which authorities have forbidden communion by a common spoon, bishops approved other means of serving the Eucharist to the laity early in the pandemic.31 The Russian and Romanian Orthodox churches have required or allowed practices such as disinfecting the common spoon or using individual disposable spoons.32
The Church does not attempt to define with rational precision how the eucharistic elements miraculously become Christ’s body and blood, and unsurprisingly the relationship between receiving the communion and contracting a potentially deadly disease is controversial within Orthodoxy. On the one hand, the Church embraces Ignatius of Antioch’s understanding of communion as “the medicine of immortality” and repeats Paul’s warning that judgment, disease, and death may result from communing in a spiritually unworthy way (1 Cor. 11:29-30).33 Receiving the Eucharist as true personal communion with Christ is at the very heart of the liturgical and spiritual life of the Church.34 Suggesting that communing sacramentally could be a means of contracting a deadly virus due to the natural process of infection remains anathema to most Orthodox commentators. On the other hand, the bodily experience of tasting and digesting the eucharistic elements is shared by all communicants. The eucharistic elements by all appearances retain the physical properties of bread and wine, including susceptibility to changes in temperature and to mold. The Church does not teach that the reality of the bread and wine disappear at their consecration, as though they were completely removed from the natural processes of the created world. As Father Alexander Schmemann puts it, their transformation is one “not of discontinuity, but of fulfillment and actualization.”35 In this light, forthcoming theological reflection is needed in order to respond to the charge that affirmation of the impossibility of spreading disease through the Eucharist amounts to “magical fundamentalism” or a denial of the full humanity of Christ, whose body and blood people receive in the Eucharist.36
Vaccination and abortion
The moral appropriateness of receiving vaccinations for COVID-19 has become controversial for some Orthodox Christians in light of the connection of the testing and production of vaccines to cells taken from aborted fetuses. As H. Tristram Engelhardt, Jr, comments, Orthodoxy forbids abortion, viewing the intentional destruction of fetal life “as a radical failure of love, as one of the worst of actions, whether or not the embryo is yet a person.”37 A statement from the Assembly of Canonical Orthodox Bishops in the United States does not address the relationship of the vaccines to abortion but places the question of whether its members should receive vaccinations and therapies for COVID in the context of biblical injunctions “to respect and protect the body as the temple of God (1 Cor. 6:19)” and to refrain “from either tempting or testing the Lord (Matt. 4:7).” In this light, the bishops teach that “it is neither wrong nor sinful to seek medical attention and advice. In fact, we welcome interventions that provide us more time for spiritual renewal and repentance.” The statement encourages people to consult their physicians on what medical treatments are most appropriate for them, noting that even as the clergy provide spiritual guidance, “your personal doctor will guide your individual medical decisions.”38
A statement from The Orthodox Theological Society of America notes that the Pfizer and Moderna vaccines were tested using cells “derived from the 1960s and 1970s from therapeutic abortions.” While the abortions did not occur for the sake of producing vaccines, “many of the other Covid-19 vaccines (e.g., AstraZeneca and Johnson and Johnson) are grown using the same fetal cell line.” These vaccines do not contain the cells of aborted fetuses, but cells descended from the tissue of aborted fetuses are used for producing and manufacturing these vaccines. The statement concludes that use of any of the vaccines is “the best ethical option to promote health and life.” It rejects the claim that the vaccines’ connections to cells taken from fetuses aborted for therapeutic reasons should forbid their use by Orthodox Christians:Most Church leaders have agreed that the many lives saved by vaccination are an important factor in permitting the use of these vaccines. While it is a sad reality that the origin of these cell lines is from these very few therapeutic abortions, the cell lines are already in existence, no new fetuses will be used, and as such it is far preferable to cure diseases as a result of the use of these cell lines than to totally forbid the use of these cell lines. The vaccines in no way legitimize or promote abortion; rather they combat disease and death, support health, and enable life—not death—to prevail, all of which are of the highest ethical value.39
Father Alexander Webster reaches the opposite conclusion, arguing that Orthodox ethics forbids the use of vaccines with “any connection to aborted preborn baby cells.” In support of his argument, he cites statements from the Holy Synod of the Patriarchate of Moscow against “so-called fetal therapy,” in which a fetus is aborted and “used in attempts to treat various diseases and to ‘rejuvenate’ an organism,” as well a statement from the Holy Synod of Greece against human cloning and experimentation on human embryonic cells. Webster also notes a statement on the official site of the Patriarchate of Romania that forbids extracting or transplanting tissues without the consent of the donor, which is obviously impossible for embryos or fetuses, as well as doing so in a fashion that causes the death of the donor. These statements provide a basis for condemning the actions of those who intentionally destroy embryos or fetuses for therapeutic benefit to others. They do not speak directly, however, to whether it is morally acceptable for people to receive potentially lifesaving vaccinations that were tested or produced through cell lines, the origins of which come from abortions performed decades ago.40
Webster seeks more direct support for his argument from a letter signed jointly by Archbishop Makarios of the Greek Orthodox Archdiocese of Australia together with local Catholic and Anglican archbishops, which states that vaccines “cultured on a fetal cell-line will raise serious issues of conscience for a proportion of our population.” The bishops “accept that the proposed vaccine may be sufficiently remote from the abortion that occasioned the derivation of the cell-line” but they call on the Australian government to respect the conscience of those who do not want “to benefit in any way from the death of the little girl whose cells were taken and cultivated, nor to be trivializing that death, and not to be encouraging the fetal tissue industry.”41 While the bishops hope for “an ethically uncontroversial alternative vaccine,” they do not, as Webster claims, reject “on moral grounds, any vaccines derived from ‘fetal cell lines.’”42
Webster argues that since Orthodoxy views abortions as “objectively, intrinsically evil,” to profit “from such abominations for any reason, even lifesaving in the present or future” is morally prohibited. He fears that benefiting from such evil deeds amounts to embracing a consequentialist ethic that could justify any action for the sake of the greater good. In contrast, Webster states that “Orthodox-informed consciences all testify to the uncontestable moral maxim that we may not do evil to achieve good. There is no ‘lesser evil’ that is tolerable to achieve, ostensibly, a ‘greater good.’”
Webster’s critique of the use of COVID vaccines overlooks the crucial factor of intention in discerning the moral import of actions. Surely, ethically relevant points of distinction exist between the actions of those who intentionally kill the innocent in order to achieve a desired end and of those who wish only to make use of potentially lifesaving vaccines that have a remote connection to abortion. Though Webster asserts that “time and distance are irrelevant to profiteering from such abominations for any reason,” he does not make a persuasive case that receiving a vaccine in this context equates to doing evil so good may come of it.43
Engelhardt’s analysis of the use of tissue from dead fetuses, written two decades before the pandemic, places these matters in an illuminating context. He does not prohibitusing for a good end something that has been acquired by heinous means, as long as one has not been involved in (1) employing these evil means, (2) encouraging their use, (3) avoiding their condemnation, or (4) giving scandal through their use.
He cautions, however, against endorsing or encouraging any immoral actions and notes, “Use of such materials must at the very least be approached penitentially as a concession to human weakness.”44
Webster criticizes Engelhardt for having a “sanguine, even cavalier attitude toward” aborted fetuses and for mounting “a strictly secular, un-Orthodox consequentialist argument” that ignores the “evil intent” and “violent destruction of human persons” intrinsic to intentional abortion. He accuses Engelhardt of employing “a consequentialist mantra” that absolves those who knowingly benefit from an abortion and claims that a person “who willfully benefits from such abortions provides a tacit post-factum encouragement of the original act and similar acts in the present and future.” He charges that “anyone who utilizes aborted baby cells even for an ostensibly good end,” by definition, gives scandal, as “the ‘scandal’ is inherent in the intrinsically evil act and the hands of anyone who exploits that act are also dirty.”45
Webster’s critique is unpersuasive, as he simply asserts that those who benefit from an abortion, even quite indirectly, by definition encourage it and give scandal. He gives fair consideration neither to Engelhardt’s clear condemnation of abortion nor to his view that a penitential attitude is necessary in such cases. A similar sentiment is present in the Orthodox Theological Society of America’s reference to “the sad reality” of the origins of the cell lines coming from the tissue of aborted fetuses. In addition to not attending critically to the crucial differences in intention between abortionists and those receiving vaccines for COVID-19, Webster also fails to grapple with the inevitably complex challenges of preserving the life and well-being of neighbors in a fallen world, especially in the midst of a deadly global pandemic. Orthodoxy does not teach that using resources connected to earlier evil deeds to save lives is intrinsically wrong. For example, people who would otherwise die of thirst may drink water from a well obtained by its owner through murder without condoning or sharing in the guilt of murder. People have no obligation to refrain from using the methods available to save the lives of others simply because earlier wrongdoing by others played a role in making those methods available. Employing those means, while regretting the brokenness of the circumstances by which they became available, does not necessarily make one culpable of, indifferent to, or an advocate of the prior wrongs.
Indeed, love of neighbor and concern for the physical well-being of others are so intrinsic to the Orthodox Christian faith that one could make a case that those who refuse vaccination risk are committing grave evil. Metropolitan Hilarion, when serving as chair of the department for External Church Relations for the Moscow Patriarchate, stated that those refusing to vaccinate were committinga sin for which they will have to atone throughout their lives. . . . I see situations every day where people visit a priest in order to confess that they had refused to vaccinate themselves or their close ones and unwillingly caused someone’s death. . . . The sin is thinking of oneself but not of another person.46
While a theological consultation of the Russian Orthodox Church “concluded that the decision to get vaccinated must remain an individual choice,” it also taught, “concerning the use of cells derived from aborted fetal tissue,” that people may receive the vaccine “in the absence of other alternatives.”47 Ecumenical Patriarch Bartholomew of Constantinople described his own decision to receive a vaccination as “not only a matter of necessity or choice; it is also a responsibility to fellow human beings.” Concerning those who reject scientifically based measures to contain the spread of the pandemic, the Patriarch stated,The New Testament affirms that whoever does not love [humankind], cannot love God. These people are indifferent to the protection of fellow human beings. The rejection of the mask and all precautionary measures does not arise simply from ignorance but from the necrosis of love within them.48
Conclusion
The response of the Orthodox Church to the pandemic not only reflects deep theological and moral sensibilities that demand caring for the sick but also provides examples of resistance against following public health measures and receiving vaccines by clergy and laity. Restrictions on some of the most characteristic practices of Orthodoxy, such as attendance at the Divine Liturgy and receiving communion from a common spoon, fueled controversies that exposed points of tension concerning the relationship between the sacramental life of the Church and the obligation to care for the sick out of love for neighbor. Debates within Orthodoxy on whether diseases may be contracted through communion remain especially controversial and indicate a need for further reflection that is both theologically sound and informed by contemporary medical science. An incarnational faith that views the embodied person as a living icon of God must not refuse to meet this challenge.
Author biography
Philip LeMasters is Professor of Religion at McMurry University and an archpriest of the Antiochian Orthodox Christian Archdiocese of North America. He is also a Senior Fellow in the Orthodoxy and Human Rights Project of the Orthodox Christian Studies Center of Fordham University. He has published several books and articles on moral theology and applied ethics in Eastern Orthodoxy on topics including marriage, peacemaking, politics, and bioethics.
1. See Timothy Ware, The Orthodox Church (London: Penguin Books, 1993); and Hilarion Alfeyev, The Mystery of Faith (London: Darton, Longman and Todd, 2002) for lucid introductions to Orthodox Christianity.
2. See John Breck, The Sacred Gift of Life: Orthodox Christianity and Bioethics (Crestwood: St. Vladimir’s Seminary Press, 2000).
3. N. Michael Vaporis, ed., “The Divine Liturgy of St. Basil the Great,” The Greek Orthodox Archdiocese of America, https://www.goarch.org/-/the-divine-liturgy-of-saint-basil-the-great.
4. Philip LeMasters, “Philanthropia in Liturgy and Life: The Anaphora of Basil the Great and Eastern Orthodox Social Ethics,” St Vladimir’s Theological Quarterly 59.2 (2015): 187–211.
5. See Daniel B. Hinshaw, Suffering and the Nature of Healing (Yonkers: St. Vladimir’s Seminary Press, 2013).
6. Jean-Claude Larchet, The Theology of Illness (Crestwood: St. Vladimir’s Seminary Press, 2002), 26–33.
7. See Jean-Claude Larchet, Mental Disorders and Spiritual Healing: Teachings from the Early Christian East (Hillsdale: Sophia Perennis, 2005), 40; and Stanley Samuel Harakas, “The Eastern Orthodox Church,” in Caring and Curing: Health and Medicine in Western Religious Traditions, ed. R. L. Numbers and D. W. Amundsen (New York: McMillan Publishing, 1986), 153.
8. Timothy S. Miller, The Birth of the Hospital in the Byzantine Empire (Baltimore: The Johns Hopkins University Press, 1997), 87.
9. Miller, Birth of the Hospital, 63–66.
10. Vasiliy Marushchak, The Blessed Surgeon: Life of Saint Luke Archbishop of Simferopol (Point Reyes Station: Divine Ascent Press, 2001), 122.
11. See Hinshaw, Suffering and the Nature of Healing, 33–36, for a discussion of traditions of Orthodox ambivalence toward medicine.
12. “WHO Coronavirus (COVID-19) Dashboard,” World Health Organization, https://covid19.who.int.
13. Katherine LeMasters, Lauren Brinkley-Rubinstein, Morgan Maner, Meghan Peterson, Kathryn Nowotny, and Zinzi Bailey, “Carceral Epidemiology: Mass Incarceration and Structural Racism During the COVID-19 Pandemic,” The Lancet Public Health 7.3 (2022): e287–90.
14. “A Message of Hope from the Assembly of Bishops,” Assembly of Canonical Orthodox Bishops in the United States of America, October 7, 2020, https://ocl.org/assembly-of-canonical-orthodox-bishops-of-the-usa-october-2020-meeting.
15. Emil M. Marginean, “The Institutional Reaction of the Orthodox Churches Faced with the Initial Covid-19 Crisis,” International Journal of Orthodox Theology 11.4 (2020): 169.
16. Ioannis Kaminis, “Eastern Orthodox Church and Covid-19 A Threat or an Opportunity?” Theologische Rundschau 86 (2021): 418–19.
17. Metropolitan Joseph, “Updated COVID-19 Policy From His Eminence Metropolitan Joseph,” Antiochian Orthodox Christian Archdiocese of North America, March 17, 2020, https://www.antiochian.org/regulararticle/623. As Alexander Agadjanian noted, Patriarch Kyril of Moscow made a similar appeal to Mary of Egypt; see “Pandemic, homo somatis, and Transformations of the Russian Orthodox Ethos,” Entangled Religions 12/13 (2021): 12 n. 40.
18. Metropolitan Joseph, “Letter to ‘Beloved Faithful in Christ,’” June 8, 2020, in author’s possession.
19. Metropolitan Tikhon, “Archpastoral Letter on the Coronavirus,” Orthodox Church of America, March 17, 2020, https://www.oca.org/holy-synod/statements/his-beatitude-metropolitan-tikhon/archpastoral-letter-on-the-coronavirus.
20. Kaminis, “Eastern Orthodox Church and Covid-19,” 422–23.
21. Ibid., 424.
22. Alexander Agadjanian and Scott Kenworthy, “Resistance or Submission? Reactions to the COVID-19 Pandemic in the Russian Orthodox Church,” Berkley Forum, Berkley Center for Religion, Peace & World Affairs, August 19, 2021, https://berkleycenter.georgetown.edu/responses/resistance-or-submission-reactions-to-the-covid-19-pandemic-in-the-russian-orthodox-church.
23. Kaminis, “Eastern Orthodox Church and Covid-19,” 424.
24. Kaminis, “Eastern Orthodox Church and Covid-19,” 424–25.
25. For a critique of Orthodox response to scientifically-based concerns about the pandemic, see Gayle E. Woloschak, “Life and Death in the Time of Covid-19,” The Wheel 23 (Fall 2020): 20–26. For a description of communion practices endorsed by the Russian Orthodox Church, see Agadjanian, “Pandemic,” 3 n. 6.
26. Archbishop Elpidophoros, “Encyclical of His Eminence Archbishop Elpidophoros of America and the Eparchial Synod on the Covid-19 Pandemic (Coronavirus),” Orthodox Observer, March 6, 2020, https://www.goarch.org/-/encyclical-covid-19-pandemic.
27. Metropolitan Joseph, “Letter.”
28. Eugenia Constantinou, “More Dangerous than Covid-19,” Orthodoxia Info, May 30, 2020, https://orthodoxia.info/news/more-dangerous-than-covid-19.
29. Vassa Larin, “The Communion Spoon as Icon,” The Wheel 23 (Fall 2020): 30.
30. Larin, “Communion Spoon as Icon,” 32.
31. Larin, “Communion Spoon as Icon,” 34. Marginean notes that the Metropolis of Austria chose to serve communion according to “a typicon of the Divine Liturgy of St. James in order to avoid using the communion spoon” (“Institutional Reaction,” 168).
32. See “Russian Orthodox Church Adopts Sweeping Anti-Coronavirus Rules” The Moscow Times, March 17, 2020, https://www.themoscowtimes.com/2020/03/17/russian-orthodox-church-anti-coronavirus-rules-a69649; and Ron Synovitz, “Coronavirus Vs. The Church: Orthodox Traditionalists Stand Behind the Holy Spoon,” Radio Free Europe/Radio Liberty, March 17, 2020, https://www.rferl.org/a/coronavirus-vs-the-church-orthodox-traditionalists-stand-behind-the-holy-spoon/30492749.html.
33. Ignatius of Antioch, “The Epistle to the Ephesians,” in Early Christian Writings: The Apostolic Fathers, ed. Andrew Louth (London: Penguin Books, 1987), 66.
34. See Alexander Schmemann, For the Life of the World (Crestwood: St. Vladimir’s Seminary Press, 1998), 23–46, for an interpretation of the centrality of the Eucharist to Orthodoxy.
35. Schmemann, For the Life of the World, 143.
36. See Kaminis, “Eastern Orthodox Church and Covid-19,” 426; and Cyril Hoverun, “‘Covid Theology,’ or the ‘Significant Storm’ of the Coronavirus Epidemic,” State, Religion, and Church 8.2 (2021): 23–25.
37. H. Tristram Engelhardt, Jr., The Foundations of Christian Bioethics (Lisse: Swets & Zeitlinger, 2000), 275.
38. “Statement Regarding Developments in Medicine: COVID-19 Vaccines & Immunizations,” Assembly of Canonical Orthodox Bishops of the United States of America, January 22, 2021, https://www.assemblyofbishops.org/news/news-archive/2021/statement-regarding-developments-in-medicine-covid-19-vaccines-and-immunizations.
39. “Covid-19 Vaccines: How They Are Made and How They Work to Prime the Immune System to Fight SARS-CoV2,” Orthodox Theological Society of America, March 8, 2021, https://www.otsamerica.net/wp-content/uploads/2021/03/Covid19-VaccineTech.pdf.
40. Alexander Webster, “The Moral Peril of Taking Most COVID-19 Vaccines,” Monomakhos, April 15, 2020, https://www.monomakhos.com/on-covid-vaccines-and-the-church. From the Patriarchate of Moscow, he cites Bases of the Social Concept of the Russian Orthodox Church, ch. 12, §7, http://orthodoxeurope.org/page/3/14.aspx. From the Holy Synod of the Church of Greece, he Cites “The Cloning of Embryonic Cells,” August 17, 2000, https://www.bioethics.org.gr/03_c.html#6. From the Patriarchate of Romania, he cites “Transplant of Organs,” https://patriarhia.ro/transplant-of-organs-6021-en.html.
41. Archbishop Glenn N. Davies, Archbishop Anthony Fisher, Metropolitan Makarios, “Letter to the Honorable Scott Morrison, MP,” August 20, 2020, d8231ace-cb4b-4c14-9d01-88aedb7b5ded (static9.net.au).
42. Webster, “Moral Peril of Taking Most COVID-19 Vaccines.”
43. Webster, “Moral Peril of Taking Most COVID-19 Vaccines.”
44. Engelhardt, Foundations of Christian Bioethics, 261–62.
45. Webster, “Moral Peril of Taking Most COVID-19 Vaccines.”
46. “Vaccinate or Repent, Russian Church Says Amid Hundreds of Daily COVID-19 deaths,” Reuters, July 5, 2021, https://www.reuters.com/world/europe/vaccinate-or-repent-russian-church-says-amid-hundreds-daily-covid-19-deaths-2021-07-05.
47. Agadjanian and Kenworthy (“Resistance or Submission?”) note, regarding the quote above in “Vaccinate or repent,” that “Metropolitan Hilarion’s remark concerned people who refused to be vaccinated and then passed on COVID to someone who died as a result—indicating that they were in some sense responsible, that they were thinking only about themselves in choosing not to get vaccinated and not thinking about others.”
48. Patricia Claus, “Ecumenical Patriarch Bartholomew Encourages Faithful to be Vaccinated,” Greek Reporter, January 12, 2021, https://greekreporter.com/2021/01/12/ecumenical-patriarch-bartholomew-encourages-faithful-to-be-vaccinated/.
| 0 | PMC9742507 | NO-CC CODE | 2022-12-14 23:30:02 | no | 2022 May; 119(1-2):110-121 | utf-8 | null | null | null | oa_other |
==== Front
J R Coll Physicians Edinb
J R Coll Physicians Edinb
RCP
sprcp
The Journal of the Royal College of Physicians of Edinburgh
1478-2715
2042-8189
SAGE Publications Sage UK: London, England
36476144
10.1177/14782715221142559
10.1177_14782715221142559
Case Report
Transverse erythronychia: A unique nail manifestation of COVID-19 infection and brief review of COVID-19 associated nail changes
Sil Abheek 1
Ghosh Shouvik 2
Das Anupam 2
https://orcid.org/0000-0002-3809-8926
Chandra Atanu 3
1 Department of Dermatology, Venereology, and Leprosy, RG Kar Medical College and Hospital, Kolkata, India
2 Department of Dermatology, Venereology, and Leprosy, KPC Medical College and Hospital, Kolkata, India
3 Department of Internal Medicine, RG Kar Medical College and Hospital, Kolkata, India
Atanu Chandra, Department of Internal Medicine, RG Kar Medical College and Hospital, Doctor’s Quarters (RG Kar Medical College Campus); 1, Khudiram Bose Sarani, Kolkata, West Bengal 700004, India. Email: [email protected]
8 12 2022
8 12 2022
14782715221142559© The Author(s) 2022
2022
Royal College of Physicians of Edinburgh
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Over the past 2 years, a plethora of mucocutaneous manifestations have been described to be associated with coronavirus 2019 (COVID-19) infection. Nail changes attributed to COVID-19 have rarely been documented in the literature. We describe here a unique nail finding ‘transverse erythronychia’ due to COVID-19 and review the literature on the diverse nail pathology attributed to the disease.
COVID-19
lunula
erythronychia
inflammation
hypercoagulability
edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
A gamut of mucocutaneous manifestations of COVID-19 disease has been described over the past year ranging from morbilliform, urticarial, vesicular, papulosquamous, petechial/purpuric eruptions, pernio-like lesions, livedo reticularis like rashes, and retiform purpura.1,2 Nail unit manifestations of COVID-19 are mostly non-specific possibly caused by the sensitive nature of the nail matrix when impacted by trauma, inflammation, and hypercoagulability. We report here a unique nail predicament ‘transverse erythronychia’ following COVID-19 infection and review the literature on nail changes associated with this raging viral infection.
Case presentation
An otherwise healthy 38-year-old Indian lady presented to our facility to seek dermatological consultation regarding sudden discolouration of her nails about a week after recovering COVID-19. Approximately 23 days since the onset of symptoms, she noticed a strange patterned discolouration involving the fingernails of the left hand. On examination, we noticed linear red bands of width 2–3 mm, which was slightly convex distally and traversed the distal nail plate margin at the level of isthmus, between the lateral nail fold margins (Figure 1). This finding was observed to involve all fingernails of the left hand, few digits of right hand while the toe nails were completely spared. Her other medical and dermatological history was unremarkable. This characteristic nail presentation was diagnosed as ‘transverse erythronychia’ associated with COVID-19.
Figure 1. Transverse erythronychia involving all fingernails of left hand and few of the right hand (a and b).
The more recognised ‘red lunula’ (reddish discolouration of the entire lunula) is associated with rheumatoid arthritis, systemic lupus erythematosus, alopecia areata, cardiac failure, hepatic cirrhosis, lymphogranuloma venereum, psoriasis, carbon monoxide poisoning, twenty-nail dystrophy, and reticulosarcoma. Red half-moon-shaped bands have been characteristically described as bordering the distal edge of lunula (‘red half-moon lunula’).
She was reassured about the benign and self-limiting nature of this condition. On follow-up after 3 months, the nail changes were not visualised.
Discussion
Erythronychia is a common and benign clinical entity characterised by red discolouration of the nails of one or multiple digits. Although the more commonly encountered ‘longitudinal erythronychia’ (extending from the proximal nail fold to the distal tip of the nail plate) is often idiopathic, it has also been associated with benign subungual tumours, malignant subungual tumours, and various cutaneous conditions and systemic diseases.3 The present description of transverse erythronychia has rarely been documented. This was first reported in four patients of Kawasaki disease.4 Chang et al.5 described a case of polydactylous transverse and longitudinal erythronychia in an elderly Caucasian male with a past of testicular carcinoma. Siragusa et al.6 had also previously documented similar transverse dyschromic changes in a patient of multiple system atrophy, where disturbances to microcirculatory homeostatic mechanisms secondary to multiple system atrophy were postulated to have a contributory role.
The red half-moon lunula is another novel and infrequently described manifestation of COVID-19 infection. The presentation includes appearance of a distally convex half-moon-shaped red band surrounding the distal margin of the lunula affecting all fingernails. Neri et al.7 first described this unique finding 2 weeks after onset of symptoms with confirmed COVID-19 diagnosis. On follow-up a month later, they noticed further widening of the bands. Méndez-Flores et al.8 reported the second demonstrating this sign, where the appearance was seen only 2 days after symptom onset and subsequently resolved within a week. These clinical findings are attributed to microvascular injury of the capillary network of the distal subungual arcade secondary to an inflammatory immune response and a procoagulant milieu associated with SARS-CoV-2 infection.7,8 A similar pathomechanistic reasoning explains the occurrence of transverse erythronychia in COVID-19.
Another reported nail finding includes transverse orange nail lesions.9 It was noted to occur 16 weeks after COVID-19 symptoms onset, demonstrating an orange discolouration of the distal nail plate with a sharp demarcation line from unaffected area. However, unlike our case, Tammaro et al.9 observed a much later onset (almost 4 months) and non-resolution of the nail lesions at 1-month follow-up. Transverse leukonychia, onychomadesis, transverse grooving (Beau’s lines), and leukonychia have also been reported in a few cases. A much delayed onset, ranging from 28 to 112 days, characterised these lesions.8,9 The clinical features of cases documenting COVID-19 associated nail changes are highlighted in Table 1.7–13
Table 1. Compilation of nail changes documented with COVID-19 infection.
Author/year Country No. of patients Age Sex Comorbidities Nail findings Onseta (days) Systemic symptoms of COVID-19 Hospitalisation COVID-19 treatment received Resolution
Neri et al./2020 Italy 1 60 F None Red half-moon lunula 14 Fever, cough, dyspnoea, anosmia, ageusia Yes HCQ, Lopinavir/Ritonavir, ceftriaxone, heparin, supplemental oxygen Widening of bands after 1 month
Méndez-Flores et al./2020 Mexico 1 37 F Not mentioned Red half-moon lunula 2 Anosmia, cough, fever No Not mentioned 1 week
Tammaro et al./2021 Italy 1 89 F Not mentioned Transverse orange discolouration 112 Cough, asthenia Not mentioned Not mentioned Unaltered at 1 month
Alobaida et al./2020 Saudi Arabia 1 45 M Not mentioned Beau’s lines 98 Diarrhoea, fever, and dyspnoea No Not mentioned Not mentioned
Ide et al./2020 Japan 1 68 M None Beau’s lines, leukonychia, periungual desquamation 28 Fever, dyspnoea Yes Hydroxychloroquine, methylprednisolone Not mentioned
Senturk et al./2021 Turkey 1 47 F Hypertension, diabetes mellitus Onychomadesis 84 Sore throat Yes Hydroxychloroquine, azithromycin, oseltamivir, ceftriaxone Not mentioned
Fernandez-Nieto/2020 Spain 1 47 M None Transverse leukonychia 45 Mild COVID-19 bilateral pneumonia Yes Lopinavir/Ritonavir Not mentioned
Present case India 1 37 F None Transverse erythronychia 23 Fever, sore throat, cough, myalgia No Azithromycin, Doxycycline, vitamin C, Zinc, paracetamol 3 months
M: male; F: female.
a Counted from onset of COVID-19 symptoms appearance.
Given the paucity of reports on nail changes in COVID-19 infection, we seek to add to the existing literature on this matter. Physicians involved in care of COVID-19 patients should continue to document observed mucocutaneous changes to better identify the diagnostic clues and understand the underlying pathophysiological mechanisms of this viral disease.
Conclusion
The evaluation of nails is often overlooked while carrying out general physical examination. Nail changes attributed to COVID-19 infection have been rarely documented in the literature. ‘Transverse erythronychia’ is a characteristic nail presentation, the presence of which should raise the suspicion of a retrospective diagnosis of COVID-19 infection.
Author contributions: AS and SG prepared the manuscript with adequate planning and execution. AC and AD contributed to patient management, review of literature, critical revision of content and final approval of manuscript. All authors are in agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
Patient’s consent: An informed written consent was obtained from the patient after full explanation regarding her images being published for academic interest. The patient did not have any objection regarding use of her images and gave due permission to use them.
ORCID iD: Atanu Chandra https://orcid.org/0000-0002-3809-8926
==== Refs
References
1 Panda M Dash S Behera B et al . Dermatological manifestations associated with COVID-19 infection. Indian J Dermatol 2021; 66 : 237–45.
2 Khan IA Karmakar S Chakraborty U et al . Purpura fulminans as the presenting manifestation of COVID-19. Postgrad Med J 2021; 97 : 473.33563711
3 Baran R. The red nail—always benign? ActasDermosifiliogr 2009; 100 : 106–13.
4 Lindsley CB. Nail-bed lines in Kawasaki disease. Am J Dis Child 1992; 146 : 659–60.
5 Chang C Beutler BD Cohen PR. Polydactylous transverse erythronychia: report of a patient with multiple horizontal red bands affecting the fingernails. DermatolTher 2017; 7 : 255–62.
6 Siragusa M Del Gracco S Elia M et al . Peculiar dyschromic changes of finger nails in a patient with multiple system atrophy. Int J Dermatol 1998; 37 : 156–60.
7 Neri I Guglielmo A Virdi A et al . The red half-moon nail sign: a novel manifestation of coronavirus infection. J Eur Acad Dermatol Venereol 2020; 34 : e663–5.
8 Méndez-Flores S Zaladonis A Valdes-Rodriguez R. COVID-19 and nail manifestation: be on the lookout for the red half-moon nail sign. Int J Dermatol 2020; 59 : 1414.32860426
9 Tammaro A Adebanjo GAR Erasmus HP et al . Transverse orange nail lesions following SARS-CoV-2 infection. Dermatol Ther 2021; 34 : 2–3.
10 Senturk N Ozdemir H. Onychomadesis following COVID-19 infection: is there a relationship? Dermatol Ther 2020; 33 : e14309.32935910
11 Ide S Morioka S Inada M et al . Beau’s lines and leukonychia in a COVID-19 patient. Intern Med 2020; 59 : 3259.33132338
12 Fernandez-Nieto D Jimenez-Cauhe J Ortega-Quijano D et al . Transverse leukonychia (Mees’ lines) nail alterations in a COVID-19 patient. Dermatol Ther 2020; 33 : e13863.
13 Alobaida S Lam JM. Beau lines associated with COVID-19. CMAJ 2020; 192 : E1040.
| 36476144 | PMC9742509 | NO-CC CODE | 2022-12-14 23:30:02 | no | J R Coll Physicians Edinb. 2022 Dec 8;:14782715221142559 | utf-8 | J R Coll Physicians Edinb | 2,022 | 10.1177/14782715221142559 | oa_other |
==== Front
RAE
sprae
Review & Expositor
0034-6373
2052-9449
SAGE Publications Sage UK: London, England
10.1177/00346373221133008
10.1177_00346373221133008
Thematic Words · · ·
COVID-19 and algorithmic medical ethics: A Christian perspective
Childs Brian H.
Mercer University, USA
Brian H. Childs, Department of Bioethics and Medical Humanities, Mercer University School of Medicine 1250 East 66th St. Savannah, GA 31404, USA. Email: [email protected]
5 2022
5 2022
5 2022
119 1-2 3340
© The Author(s) 2022
2022
Review & Expositor, Inc
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Triage plans which were largely developed in the face of the growing and lethal pandemic betrayed an underlying anthropology which unintentionally neglected to allow for the assignment of potentially limited interventions to underserved and less socially advantaged persons. This neglect is abetted by the structure of US medical delivery that treats medical care as a commercial commodity with an emphasis on high tech rescue medicine as opposed to preventive public health medicine. A Christian anthropology modeled by Karl Barth’s notion of analogia relationis would correct this neglect of the underserved and needy.
analogia relationis
COVID-19
Karl Barth
pandemic
theological anthropology
triage ethics
typesetterts1
==== Body
pmc Do not neglect to show hospitality to strangers, for by doing that some have entertained angels without knowing it. (Heb 13:2 NRSV)
For many years in the late twentieth century, Seward Hiltner, then Professor of Pastoral Theology at Princeton Theological Seminary, was a consultant at the renowned Menninger Psychiatric Foundation in Topeka, Kansas. One part of the expansive foundation’s facilities was a state institution for the developmentally disabled and challenged. Hiltner held weekly teaching rounds at this site that included staff physicians and psychiatrists, psychiatric residents, social workers, chaplains, and psychologists. In one of these sessions, a robust discussion about serious birth defects and disabilities developed. Such topics as abortion, eugenics, in vitro interventions, and euthanasia were debated vigorously among the participants with one exception: a hushed Hiltner. Near the end of the session, one of the attending psychiatrists noted that Hiltner was silent but clearly listening to the discussion, and he asked Hiltner for his point of view on the topic. Hiltner hesitated a few moments and then asked a question: “What does it mean to be a child of God?”
I was not in attendance at this meeting in Kansas, but three discussion participants, including Hiltner, confirmed the story. I worked with Hiltner during my graduate studies at Princeton. I am a Christian ethicist, specializing in medical ethics, and the impact of the story became relevant with what I saw and experienced in the early and subsequent days of the COVID-19 pandemic. Simply put, what medical ethics has done in response to the pandemic was demonstrate its inability to have moral authority and weight. Without its moral authority or a universal imperative, medical ethics relies on proceduralism and law or local hospital policy.1 Hiltner’s question is actually about Christians’ moral responsibilities and challenges. As creatures in God’s image, what is the Christian’s calling regarding how to engage one another and the rest of creation? Hiltner’s question is not a facile one. It is a profoundly critical one that asks how Christians may become more personal with those they know and those they need to know more fully.
When a worldwide pandemic was clearly developing in early 2020, The Guardian reported that people were ordering Albert Camus’s allegorical novel The Plague in such numbers that the publisher could not keep up with the demand.2 Camus’s daughter, Catherine, was quoted as saying that the novel brings to mind the question of human responsibility in the face of a crisis, regardless of whether it is political or natural. While Camus’s work is relevant both to the political and pandemic situation, Daniel Defoe’s historical novel A Journal of the Plague Year is equally relevant.3 Although Defoe’s work is fictional (he was merely 4 years old when the plague struck London in 1665), he used widely-available mortality data as well as reflections survivors of the pandemic gave to him. Camus’s and Defoe’s works both depict how persons behave under stress as well as offer an implicit anthropology (the study and the discerning of human life, including its meaning and intention): human survival in the face of mass illness and death is dependent on the need for community, and the good life is dependent upon lives lived collectively.
Christians can view the current COVID-19 pandemic as a stress test of how they understand the human condition and what it means to be human, even a child of God. The pandemic forces the Christian community to ask needed questions about human nature and obligations to one another. Distributive justice is at issue not only in health care but also in the economic and political realms.
When a pandemic was clearly threatening the world, medical ethicists began a process of considering how they might respond to the crisis. In the vast majority of cases of which I was aware, medical ethics concentrated its energies in dusting off and perhaps adjusting triage policies and algorithms developed during previous threatened pandemics such as SARS (severe acute respiratory syndrome, also caused by a coronavirus) in 2002 and H1N1 (swine flu) in 2009. As hospitals (but, importantly, not nursing homes or other non-acute facilities) across the United States anticipated reaching capacity and running out of resources, they began to become buildings of triage. Doctors and administrators made decisions about whom to see and treat, who should be placed in which location, or who would get which medication or other limited resources. All these decisions were to be made on the basis of elaborate algorithms. Just as critically, health care workers also had to anticipate where and how to focus their emotional energies and how to protect themselves in a faltering workforce.
A need will always exist to develop methods to maximize the benefit of sound medical interventions and allocate finite resources, including those people who work in a pandemic workforce. Yet something seems amiss in this focus on triage. As Sheri Fink has noted, triage plans have an implicit anthropology. They say as much about the human values assumed in their plans as they say about the emergency practice of medicine.4
Developing a triage plan to distribute potential limited resources fairly was clearly necessary. Apparent to me, however, first intuitively and later soundly, was that the triage plans had two assumptions: (1) medical care is a commodity in a free market and, relatedly, (2) medicine as practiced in the United States is essentially focused on rescue medicine, highly technical interventions for acute and life-threatening illnesses and injuries. As a result of these assumptions, medical ethics seemed unable to recognize and incorporate who is left out and what is unrecognized. It failed to address the underserved, and it failed to support public health preventive measures. Hiltner’s question resonated within me. What does it mean to be a child of God, and what are our obligations as such? In addition, what are our obligations to the environment which too is God’s creation that God’s children inhabit? Before offering some suggestions in answering these questions, I first outline what medical ethics neglected. I also point to how the implicit anthropology in algorithmic ethics failed to recognize significant portions of God’s creation, which includes humans and the rest of nature.
Below is a more detailed discussion of how algorithmic medicine has unintentionally limited the vision of medical ethics. By way of introduction to that discussion, algorithms are any computation, formula, or statistical way of predicting an illness outcome or what intervention one might use based on an aggregate of data collected from anonymous subjects. The problem is that the predictive or interventional value of the tool is dependent on what subjects are included in the database. For instance, it has been well established that research protocols and treatment algorithms for investigations and treatment of cardiac disease in women are based on data collected almost exclusively on men whose physiology and treatment parameters are different from those of women.5 So too is the situation for ethical deliberation on triage decisions using algorithms with a limited or focused database. While useful, algorithms need to be carefully constructed; yet, even with the best of constructions, the personal and the particular are minimized and, accordingly, result in a limited kind of ethics. But ethics ought not be impersonal.
Jeffrey Bishop, a practicing physician and philosopher, seems to agree with the dangers of objectifying the personal. In his important book, The Anticipatory Corpse: Medicine, Power, and the Care of the Dying, Bishop offers a compelling critique of contemporary medicine.6 He argues that medicine, in seeking to be a distinctly scientific discipline rather than a moral one incorporating technical expertise, is potentially impoverishing and anti-human. It treats the human more as an object than a subject. This objectification of life (both human and nonhuman lives) is a result of medicine’s reductionist tendency to see the body as a machine, following Claude Bernard’s nineteenth-century influential research on the dead body: the normative body is a machine such that the ill body is either fixed or abandoned or used for spare parts. Beginning in the late eighteenth century, medicine gave up on Aristotelian formal and final causation (what life is in its idealized form and what life is for) and elevated material and efficient causation (the material of out of which something is composed, how to build something). According to Bishop,the resulting metaphysics of efficient causation allows mastery over the living body as a machine, as dead matter in motion . . . in addition, this medical nominalism allows not only for an exhaustive description of the body in motion but also for an exhaustive description of the body politic.7
If one spends any time in a modern ICU, the notion of the human as a machine is clear to see, with machines mimicking bodily parts and organs. Allied with the mechanistic understanding of life, Bishop asserts that statistical medicine (evidence-based medicine) further impoverishes what it means to be alive (or ill). Statistics in medicine became a tool of control and became associated with the power of the state.8 Bishop ends his book with a provocative conclusion:It might be that we can learn once again from the places at the margins of contemporary life, at the margins created by liberalism and biopolitics. It might be that we can learn once again not from history—a static past—but from living traditions. It just might be that the practices of religious communities marginalized in modernity and laughed at as unscientific are the source of a humane medicine. Perhaps there, in living traditions informed by a different understanding of space and time, where location and story provide meaningful contexts to offer once again hospitality to the dying and both cura coporis and cura anime, we will find a unity of material function, form, and purpose.9
Indeed. Most people go into medicine to care for people who are ill and suffer. Most people who go into medicine will not abandon patients even in the face of inevitable death. Most people go into medicine to join a professional community of service. A Christian evaluation of the goals of medicine will offer a more robust and, if I may, a more humane anthropology. Karl Barth’s notion of analogia relationis is relevant and helpful here. For Barth, analogia relationis is the mode for the relationship between God and humanity known through the revelation of the life of Jesus Christ in history.10 I argue that Barth’s explication of the analogia relationis offers an anecdote to the depersonalization of algorithmic ethics and mechanistic understandings of the human and, moreover, all of nature. For Christians, having eyes to see means looking for those who are unseen. For Christians, having speech means speaking with those who have not been heard. Doing these acts reveals to Christians what has been given to them in creation, and their obligations are in response to the gift.
Benjamin Durheim provides a helpful insight to the importance of Barth’s theological anthropology as it applies to a vision of the common good:In order to know anything at all about God, or anything about how we humans may relate to God, we rely on the analogy between the way God relates to Godself on the one hand, and the way God in Christ relates to humanity on the other.11
Though humans are imperfect due to the stain of sin and cannot relate to God and humanity by their own power, through the revelation of Jesus Christ humanity is restored to the destiny that all will be covenant partners with God. With the restoration of the relationship with God through Christ, humanity is given the possibility of being persons in relationship with one another as well as for one another. To imagine a solitary human without relationships is impossible in Barth’s anthropology.12 Barth makes this point clearly: to be human is to be in encounter. Encounter is the most basic predicate of human existence. For Barth, relationship and encounter have four elements: looking eye to eye with each other, speaking and hearing with one another, mutually assisting one another, and doing all of this with gladness.13
The first element, looking eye to eye with each other, means that encounter must be personal and not at a distance. A statistical description is limited to what is encountered or numbered and is limited to the sample of which comprises that which is considered part of the statistics. For Barth, encounter must be comprised of a sort of intimacy: I and Thou:We give each other an insight into our being. And as we do this, I am not for myself, but for thee, and Thou for me, so that we have a share and interest in one another. This two-sided openness is the first element of humanity.14
As I suggest later in this essay, algorithmic ethics fails to see completely because its focus is unidirectional and fails to see the reciprocal, or I and Thou, aspect of caring.
The second element, speaking and hearing with one another, is a qualitative enhancement of the first element. Speaking and hearing intensifies the intimacy and the encounter; it takes one another beyond the seeing of one another:An openness between the I and the Thou, their reciprocal visibility, is only a preparatory stage to their mutual expression and address, so the latter cannot be an end, but only the means to something higher, to fellowship in which the one is not only knowable by the other, but is there for him. . . . We must see and be seen, speak and listen, because to be human we must be prepared to be there for the other.15
This element, too, is lacking in medical ethics.
The third element, mutually assisting one another, qualifies intimate encounter as moving from being with to being for. Here Barth makes the point that being in encounter is consistent with what is truly human in the covenantal relationship with God as exemplified in the life of Jesus Christ:If the man Jesus, even though He is Himself, is for us in the strictest sense, living for us, accepting responsibility for us, in this respect, acting as the Son of God in the power of the Creator, He differs from us . . . . Correspondence to His being and action consists in the more limited fact that we render mutual assistance. This correspondence is of course necessary.16
Barth alludes to the notion of mutual assistance in his explication of the Parable of the Good Samaritan. He makes the point that a proper reading of the parable is to see oneself not as the Samaritan but as the person in the ditch. The “I” is the one needing the compassion, and all are dependent on the other for assistance. It is always mutual assistance.17
The fourth and final element, doing all with gladness, is what makes the first three elements at all possible. Not hostility or resentment, but neutrality, for Barth, would be the opposite of gladness. “As Christ exhibits gladness in encounter, so must humankind, else it is not true encounter, and so not true humanity.”18 Gladness in mutual caring is not difficult to find in many day-to-day experiences of reciprocal assistance. Many psychologists studying altruistic behavior have noted gladness and wellbeing in persons who have practiced altruistic acts of kindness and aid.19
So how does the value of the Barthian analogia relationois help in understanding who was not seen, or spoken with, and who gave and received assistance in gladness? Although the methods used for triage of resources in short supply during the pandemic centered on ventilators and although anti-viral medications are based on well-established predictive procedures, these methods and procedures unintentionally failed to recognize those others who were not defined within the methodology.
A bit of explanation on how the methodology worked will clarify this caveat. Triage separates those who come to a medical facility in an emergency situation into three categories: those who can be saved but need immediate attention; those who can be saved but can be cared for with intermediate care and can even wait for that care; those who cannot be saved due to the severity of their condition. Those in the first or second category receive scores that predict a survival of their condition based on empirical studies of persons in a similar condition. The most common mechanism for this prediction is the SOFA Score (Sequential Organ Failure Assessment).20 The SOFA methodology has received criticism, however, that it is both color blind and socially blind to underserved populations. Because underserved populations have more health care problems, they are graded by the SOFA system at a disadvantage. When Crisis Standards of Care mechanisms were being prepared, using SOFA, at the beginning of the pandemic, Massachusetts Representative in Congress Ayanna Pressley stated:We know communities of color are more likely to have comorbidities not because of any genetic predisposition, but due to the legacy of structural racism and inequality that has resulted in unequal access to affordable healthcare, safe and stable housing, and quality schools and employment.21
Having these comorbidities will give these persons a lower score on SOFA, and they will, therefore, be at a disadvantage for receiving life-sustaining and potentially curative therapies. As Meires, McCulloch, and Wright argue, the way to fix the ailing US health care system in its neglect of the needy will be to reignite the human connection, the personal and empathic connection of the practice of medicine as a communal and relational undertaking such that persons see themselves as full-fledged partners in healing.22
The pandemic has disproportionately affected those who have been and are underserved, particularly African Americans, as well as Hispanic populations.23 The social determinants of health that create these disparities include the inability to work from home and the necessity to rely on public transportation. According to many epidemiologists, health care outcomes are associated 70% of the time on social and economic factors. Also, social disparities, such as lack of access to the internet, disadvantages children from attending school in the virtual environment.24 These populations are the unseen that need to be seen. They are also the many who, though unseen, are yet assisting even in the height of the pandemic. Many of these people are the rural farmers producing food, the persons cleaning the rooms of COVID patients in the hospitals, and the low-paid attendants in the nursing homes, as well as grocery store workers and those who collect refuse and garbage. As far as SOFA scores are concerned, the unseen need to be seen to account for the disparities within the triage algorithms.
A related problem is US health care delivery, which is not a unified system as it is in most countries with highly developed health care systems. The pandemic has underscored this weakness. One of the most glaring examples was the almost immediate shortage of personal protective equipment (PPE), such as simple and effective face masks and even sophisticated ventilating machines. Because health care is treated as a commodity for purchase, health care organizations must adopt a way to keep their expenses low by using just-in-time purchasing of materials. Just-in-time purchasing meant no ready supply or stockpile of material existed, including masks, gowns, and even oxygen, in part due to a concurrent general global shutdown or slowing of supply chains and the inability to provide the products for the supply chain. In addition, a hospital, regardless whether a profit or not for profit hospital, must deliver well-reimbursed elective services that come with predictable lengths of stay. When hospitals had to cancel these high revenue procedures, they faced significant economic shortfalls forcing furlough of lower salaried employees who were economically disadvantaged in the first place. Rural hospitals that serve the most disadvantaged are the most vulnerable to a drastic reduction of services, and many were actually forced to close entirely.25 Health care economics relies on highly sophisticated and financially lucrative reimbursement associated with what is called “rescue medicine.” Rescue medicine is acute treatment necessary to save lives. With the current mode of US health care delivery, this kind of medical care results in the diminishment of preventive medicine, such as public health interventions, that reduces the need to rescue. This situation further disadvantages many and the unseen, including those in nursing homes and long-term facilities in which the lack of PPE resulted in a disproportionate number of deaths compared to the rest of the population.26
Even God’s wider creation is involved in the current pandemic, as well as previous and future ones. For instance, a hypothesis is that COVID-19 began and spread from a wet market in Wuhan, China. This spread of the virus is an example of Zoonosis, defined as an infection transmitted from a vertebrate animal to the human animal.27 Urbanization and deforestation have increased the real threat of zoonotic disease transmission for which no understanding or treatment exists. Rescue medicine does not concern itself with how the disease came to be. That matter is the purview of disease prevention which, along with public health medicine, takes a minor role in US health care economic distribution.
Concerns about the exploitation and health of the greater environment certainly belongs within the purview of theology. Taking into consideration Barth’s notion of “looking eye to eye,” one might understand it to mean making recognition of the other to speak and mutually assist the other in gladness. Panu Pihkala makes this very point in an essay on eco-theology: “As a theological and Biblical basis, the command to ‘till and keep’ (Gen. 2:15) has been emphasized as the key to stewardship, and usually to explain what is actually meant by dominion (Gen. 1:26).”28 Pihkala asserts a position not unlike that of mid-twentieth-century theologian Joseph Sittler, who presciently made the argument that God created nature and humans with the ability to respond to each other in love. A failure to recognize the reciprocity or mutuality between humans and nature is a failure to follow the will and command of God. Following the command of God (as seen in the history of Jesus who is fully the God who commands as well as fully the human that obeys the command), Sittler claims, “According to its given ecological structure as a place for multiple forms of life . . . in a blunt and verifiable way we are ‘justified’ by grace even in our relation to things of nature.”29 Likewise, Sittler argues, if humans do not relate lovingly with nature, then the environment displays its condemnation through natural and human suffering.30 Both current and future pandemics are testimony to this relationship. Analogia relationis must include humankind’s relationship and reciprocity with the environment as well.
What I hope to have suggested in this essay is not only a critique of contemporary bioethics’ shortsightedness, but also a prolegomenon to a Christian bioethics within a dialectical theology as asserted by Karl Barth and others. Barth’s analogia relationis first requires not just humility on humanity’s part, but foremost a confession of humanity’s limitations, a confusing of a temporal thirst of power for God’s own justice. Barth has made it clear: there is but one savior and redeemer, and that is the man Jesus. We humans, though made in the image, are nonetheless through free will fallen and cannot participate in the role except by acknowledging the inheritance of the fruits of Jesus Christ’s works. Through participation in the life of the church, as the body of Christ, and the worship therein, we will participate in the living history of the life of Jesus in the reading and hearing of scripture and in the expression of the Word in preaching and teaching. Hearing the Word cannot be done by ears unopened due to a lack of confession, and doing the Word cannot be accomplished without following the command of God given anew in the Great Commandment as found in the Synoptic Gospels (Matt 22:35-40; Mark 12:28-31; Luke 25:25-28). Only then will the analogia relationis come alive. Repetition of the Gospel in worship and prayer is training in the faith and opens the Christian community to the activity of the analogia relationis. As Stanley Hauerwas has written:In With the Grain of the Universe: The Church’s Witness and Natural Theology, I began the penultimate chapter, which was entitled “The Witness of Barth’s Church Dogmatics,” with Barth’s statement, or better, his confession, “We can only repeat ourselves.” I suggested that Barth’s observation that he could only repeat himself reflected his discovery that the God who has found us in Christ makes possible finding ourselves within the confusions we call our life. Such a finding is only available through mediation, which requires, just as a musical score may require, repetition if we are to understand its truthful goodness and beauty.31
In the repetition of the Gospel in all its forms of media and in the worship of hearing and singing and service, we might come to have open eyes and mutual speech to recognize that which has not been recognized, who yet serve such that we all might have that transcendent felicity that resides in God and our communion with the Godhead. To answer Hiltner’s question with a resounding yes, we know, at first through a glass darkly but then face-to-face, what it means to be a child of God. Only then may we come to develop an alternative to the sour fruit of algorithmic ethics.32
Author biography
Brian H. Childs, PhD, MDiv, is Professor and Chair of the Department of Bioethics and Medical Humanities at the Mercer University School of Medicine. A Certified Clinical Ethics Consultant, he has research interests in organ transplant ethics, the medical humanities, and moral development. He studied under medical ethics and religious ethics pioneer Paul Ramsey at Princeton University, as well as Seward Hiltner at Princeton Theological Seminary.
1. H. Tristram Engelhardt, Jr., “Credentialing Strategically Ambiguous and Heterogeneous Social Skills: The Emperor Without Clothes,” HEC Forum 21.3 (2009): 293–306.
2. Kim Willsher, “Albert Camus Novel The Plague Leads Surge of Pestilence Fiction,” The Guardian, 28 March 2020, https://www.theguardian.com/books/2020/mar/28/albert-camus-novel-the-plague-la-peste-pestilence-fiction-coronavirus-lockdown. See Albert Camus, The Plague (London: Michael Joseph, 2010).
3. Daniel Defoe, A Journal of the Plague Year (New York: Dutton Adult, 1977).
4. Sheri Fink, “Ethical Dilemmas in Covid-19 Medical Care: Is a Problematic Triage Protocol Better or Worse Than No Protocol at All?” The American Journal of Bioethics 20.7 (2020): 1–5.
5. Mark Woodward, “Cardiovascular Disease and the Female Disadvantage,” International Journal of Environmental Research in Public Health 16.7 (2019): 1165.
6. Jeffrey P. Bishop, The Anticipatory Corpse: Medicine, Power, and the Care of the Dying (Notre Dame: University of Notre Dame Press, 2011).
7. Bishop, Anticipatory Corpse, 60.
8. Bishop, Anticipatory Corpse, 77–85.
9. Bishop, Anticipatory Corpse, 313.
10. Karl Barth, Church Dogmatics, III/2, ed. G. W. Bromiley and T. F. Torrance (London: T&T Clark, 2002), 220–22.
11. Benjamin Durheim, “The Human as Encounter: Karl Barth’s Theological Anthropology and a Barthian Vision of the Common Good,” Lumen Et Vita 1.1 (2011): 5.
12. Durheim, “Human as Encounter,” 7.
13. Barth, Church Dogmatics, III/2, 250–74.
14. Barth, Church Dogmatics, III/2, 251.
15. Barth, Church Dogmatics, III/2, 260.
16. Barth, Church Dogmatics, III/2, 261–62.
17. Karl Barth, Church Dogmatics, I/2, ed. G. W. Bromiley and T. F. Torrance (London: T&T Clark, 2002), 418–19.
18. Durheim, “Human as Encounter,” 10.
19. Stephen G. Post, ed., Altruism and Health (Oxford: Oxford University Press, 2007). In addition I have been involved in the organ transplant community for a number of years. I am often humbled and amazed by the gladness and even joy in the face of extreme loss and grief that a family might exhibit in agreeing with the donation of organs after the sudden and unexpected death of a loved one.
20. Fayed Mohamed, et al., “Sequential Organ Failure Assessment (SOFA) Score and Mortality Prediction in Patients with Severe Respiratory Distress Secondary to Covid-19,” Cureus (16 July 2022), https://doi.org/10.7759/cureus.26911.
21. Emily C. Manchanda, et al., “Inequity in Crisis Standards of Care,” New England Journal of Medicine 363.4 (23 July 2020): e16, https://doi.org/10.1056/nejmp2011359.
22. Jennifer Mieres, et al., Reigniting the Human Connection: A Pathway to Diversity, Inclusion, and Health Equity (Charleston: Forbes Books, 2022).
23. Shikha Garg, et al., “Hospitalization Rates and Characteristics of Patients Hospitalized with Laboratory-Confirmed Coronavirus Disease 2019—Covid-Net, 14 States, March 1–30, 2020,” Morbidity and Mortality Weekly Report 69.15 (2020): 458–64.
24. Allyson Kelley, Public Health Evaluation and the Social Determinants of Health (London: Routledge, 2020), 1–18.
25. Hoag Levens, “Already in Fiscal Crisis, Rural Hospitals Face COVID-19,” Leonard Davis Institute of Health Economics, University of Pennsylvania, 1 June 2020, https://ldi.upenn.edu/our-work/research-updates/already-in-fiscal-crisis-rural-hospitals-face-covid-19/.
26. Christopher J. Cronin and William N. Evans, “Nursing Home Quality, Covid-19 Deaths, and Excess Mortality,” Journal of Health Economics 82 (March 2022), https://doi.org/10.1016/j.jhealeco.2022.102592.
27. See “Zoonoses,” World Health Organization, 29 July 2020, https://www.who.int/news-room/fact-sheets/detail/zoonoses.
28. Panu Pihkala, “Recognition and Ecological Theology,” Open Theology 2.1 (2016): 940.
29. Joseph Sittler, Essays on Nature and Grace (Minneapolis: Fortress, 1972), 121.
30. Sittler, Essays, 122.
31. Stanley Hauerwas, Fully Alive: The Apocalyptic Humanism of Karl Barth (Charlottesville: University of Virginia Press, 2022), 27.
32. I want to thank my colleague Caroline Anglim for her gracious reading of this essay, and I am thankful for her constructive criticism and suggestions. Of course, any shortcomings in this essay are my own.
| 0 | PMC9742513 | NO-CC CODE | 2022-12-14 23:30:02 | no | 2022 May; 119(1-2):33-40 | utf-8 | null | null | null | oa_other |
==== Front
Int J Sci Math Educ
Int J Sci Math Educ
International Journal of Science and Mathematics Education
1571-0068
1573-1774
Springer Nature Singapore Singapore
10344
10.1007/s10763-022-10344-9
Article
Shifting from Face-to-Face Instruction to Distance Learning of Science in China and Israel During COVID-19: Students’ Motivation and Teachers’ Motivational Practices
http://orcid.org/0000-0002-6157-4505
Fortus David [email protected]
1
http://orcid.org/0000-0003-3721-710X
Lin Jing 2
Passentin Shira 1
1 grid.13992.30 0000 0004 0604 7563 Weizmann Institute of Science, 234 Herzl Street, POB 26, 7610001 Rehovot, Israel
2 grid.20513.35 0000 0004 1789 9964 Beijing Normal University, 19 Xinwai Ave, Beitaipingzhuang, Haidian District, Beijing, 100875 China
12 12 2022
111
7 7 2022
24 11 2022
© National Science and Technology Council, Taiwan 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.
Science teachers in many countries were required to shift from face-to-face (F2F) instruction to distance learning (DL) during the COVID-19 pandemic. With the aim of helping science teachers learn how to support their students in negotiating such shifts in the future, we used an online motivation survey based on achievement goal theory to investigate the shifts to over two thousand 8th grade students’ perceptions of their science teachers’ motivational practices and their own goal orientations towards science that occurred during the transition from F2F instruction to DL in two very different countries, China and Israel. We hoped to identify issues common to both countries, assuming that these issues might be relevant to other countries as well. Factor analysis, t-tests, and multiple regression were used to identify key teacher motivational practices, changes to these practices and to students’ goal orientations, and relations between teacher practices and student goal orientations. The major predictor of students’ mastery orientation towards science in both F2F instruction and DL, teachers’ attentiveness to their students’ need to understand, declined for students in both countries during the shift from F2F to DL, and was associated with a decline in students’ mastery orientation, engagement, and enjoyment.
Keywords
Distance learning
Goal orientation
COVID-19
Teaching practices
==== Body
pmcIntroduction
With the outbreak of the COVID-19 pandemic, science teachers in many countries were required to shift from face-to-face (F2F) instruction to distance learning (DL). Distance learning was for many teachers a new instructional environment, placing new professional demands on them. There is no guarantee that teachers who were experts at facilitating F2F learning would be as successful at DL, as instructional practices that are effective in F2F instruction may not be feasible in DL, and even if they are, they may not be as effective.
One of the primary goals of middle school science instruction is to maintain and enhance students’ motivation to engage with science, a characteristic which has been repeatedly shown to decline during adolescence (Galton, 2009; Osborne et al., 2003; Vedder-Weiss & Fortus, 2011). Motivation is important because without it, little engagement can be expected, and without engagement, little learning will occur. Science teachers are significant adults in influencing adolescents’ motivation to engage with science, but their influence is mediated by students’ perceptions of their intentions. Therefore, it is important to consider students’ interpretations of their teachers’ practices and messages when investigating the relations between teachers’ practices and students’ affective stances (Meece et al., 2006; Vedder-Weiss & Fortus, 2012).
While the motivation to learn science was possibly impacted by the shift from F2F to DL, the ways in which these changes occurred were likely influenced by local conditions–cultural background of the students and teachers, differing access to technology, differing institutional expectations and constraints, and so on. The purpose of this study was to investigate and compare the shifts that occurred in two very different countries — China and Israel — in 8th grade students’ affective stances towards science, in their perceptions of their science teachers’ motivational practices, and in the relations between them, to give perspective on how local conditions may have shaped the transition from F2F instruction of science to DL of science, and from this to learn how to help teachers and students negotiate transitions from F2F to DL in the future, given their local context.
Theoretical Framework
Motivation — Achievement Goal Theory
This study draws on achievement goal theory, a motivation theory well-suited to the study of motivation in K-12 education (Linnenbrink & Pintrich, 2002; Utman, 1997). Achievement goal theory uses the construct of goal orientation to explain why and how students engage in academic undertakings. The theory identifies two primary goal orientations: mastery goals orientation and performance goals orientation. These different goal orientations are associated with different emotional experiences in relation to schooling and different ways of engaging in school-based activities (Schunk et al., 2008). A mastery goal oriented individual strives to develop understanding and competence, to attain a sense of mastery (Ames, 1992). Mastery goals have been shown to be associated with wide range of positive cognitive, emotional, and behavioral outcomes, such as self-efficacy (Kaplan & Maehr, 1999) and, in science, enhanced conceptual understanding (Patrick & Yoon, 2010). A performance goal oriented individual strives to demonstrate competence to others, and are concerned with others’ perceptions of their competence and with their ability relative to others (Ames, 1992), rather than with their own perception of their competence, as is typical of mastery oriented individuals.
Instructional Dimensions — TARGETS
According to achievement goal theory, different environments emphasize different achievement goals. Students’ perceptions of these different goal emphases lead to the adoption of different achievement orientations (Kaplan & Maehr, 2007). A study by Vedder-Weiss and Fortus (2013) showed that during late elementary school and middle school, the two main environmental factors perceived by students and influencing their goal orientations in science are their parents and their science teachers. Science teachers can draw upon seven different instructional dimensions, represented by the acronym TARGETS (task, authority/autonomy, recognition, grouping, evaluation, time, and social interactions) to convey different goal emphases (Anderman et al., 2002). Teachers and educational organizations can employ these dimensions to support their students’ adoption of the goals they wish to emphasize. Researchers can use these dimensions to identify the goals that underlie instruction. For example, a school that plans its schedule to allow students to work on projects continuously for several hours (time dimension) conveys a message that learning in this school is about developing mastery. On the other hand, a school that plans its schedule so that students can participate in marathon exercise-solving sessions in preparation for state or national high-stake tests conveys a message that learning in this school is about demonstrating performance. Similarly, science teachers that allow students to choose between several science projects (autonomy dimension) and make use of varied, challenging, high-order thinking assignments (task dimension) send a message that mastery is what is important to them, while other science teachers who publicly commend students who get the highest grades on exams (recognition dimension) send a message that performance is what counts (Maehr & Anderman, 1993).
Vedder-Weiss (2017) found that of all the TARGETS dimensions, instructional practices belonging to the task, authority/autonomy, and time dimensions are most likely to promote mastery goals. Mastery goals are desirable, as mentioned earlier, since they are associated with many positive educational outcomes. Thus, we focused in this study on these three TARGETS dimensions. Some of the features of instruction that are associated with these three dimensions are (Vedder-Weiss, 2017, p. 568): task — in which ways does the teacher organize learning activities in both psychological dimensions (e.g. high-order thinking, scaffolding, and situational interest) and structural dimensions (e.g. content, procedures, products, and materials)? Autonomy — how much autonomy and responsibility for their learning are students given by their teacher? What influence do students have over the classroom activities? Time — is there time for students to ask questions and to follow up on these questions? How flexible is the scheduling of activities?
Culture and Motivation
Culture plays an important role in shaping motivation (Chen et al., 2005). It has been shown to influence the personal and contextual determinants of motivation (Liem, 2016) and to lead to specific motivational forces that drive learning and achievement (Meissel & Rubie-Davies, 2016). Comparative multi-cultural studies are an important approach to identifying the role culture plays in shaping motivation (Lam et al., 2016).
Research Questions
The research questions that guided this study were: (1) How did teachers’ instructional practices, associated with 3 TARGETS dimensions, shift during the transition from F2F instruction to DL in the eyes of Chinese and Israeli students? (2) Which changes occurred to Chinese and Israeli students’ goal orientations towards science following the shift from F2F instruction to DL? (3) What are the relations between the changes to instructional practices identified in RQ1 and the personal motivational shifts found in RQ2? (4) What can the similarities and the differences between the results for the Chinese and Israeli students inform us of local culture and conditions may have shaped the identified shifts and relations identified the three former research questions?
Methods
Participants
The participants were 8th grade Chinese students (N = 1983) from Beijing and from the center of Israel (N = 308) whose science teachers (A) had a reputation of fostering their students’ motivation and learning of science, (B) who were teaching their classes in DL while the study was held, and (C) had taught the same classes in 7th grade, when instruction was still held F2F.
Instrument — Student Survey
An anonymous survey focusing on students’ experiences and attitudes towards science learning in F2F and DL environments, their goal orientations towards science, and their perceptions of their science teachers’ motivational practices was administered online during the COVID-19 pandemic, when instruction had moved to DL. The survey included 46 items describing various possible manifestations of several different constructs: mastery orientation in science, performance orientation in science, perceptions of teacher’s use of task-oriented practices, of autonomy/authority practices, and of time practices. Each construct was represented by at least four different items. All items in the survey were based on a 1–5 Likert scale and were drawn from existing validated scales that have been used several times before with this age group (e.g. Vedder-Weiss & Fortus, 2012), repeatedly demonstrating excellent reliability. Example items are given in Table 1.Table 1 Constructs assessed by the student survey and example items
The construct Example item
Mastery orientation I do my work in (F2F/DL) science classes because it is important to me to improve my knowledge in science
Performance orientation It is important to me that I look smart compared to others in my (F2F/DL) science classes
Task The teacher gives us book assignments in (F2F/DL) science classes
Autonomy Our science teacher lets us choose with whom to work during (F2F/DL) science classes
Time Our teacher moves on to new subjects before I fully understand what was just taught in (F2F/DL) science classes
Analyses
Exploratory factor analysis (EFA) was performed on the Chinese and Israeli dataset separately. Common factors to both datasets were identified, named, and a student score for each construct was calculated. All factors were normally distributed. Paired t-tests were used to identify changes to each factor in each country. Multiple linear regression was used for each country to identify significant relations between the students’ perceptions of their teachers’ practices and their goal orientations in the different learning environments.
Results
Four common personal constructs were identified for both countries: (A) mastery orientation towards science, (B) performance orientation towards science, (C) engagement with science studies, and (D) the need to look smart compared to others. Six additional common constructs relating to students’ perceptions of their teachers emerged: (A) teacher attentiveness with time (not moving on until students fully understand, allowing students to ask questions, etc.), (B) teacher use of various tasks, (C) teacher gives students autonomy, (D) teacher talks and the students need to listen, (E) teacher says there is not enough time, and (F) teachers’ use of bookwork. The loading of the items on these constructs and their reliabilities (Cronbach alpha) is provided in an online appendix. The last construct, teachers’ use of workbook, had a low reliability (alpha Cronbach = 0.49) and was therefore not used in further analyses.
In the shift from F2F instruction to DL, significant declines, all with p < 0.001, were identified for the Chinese students in their engagement (effect size (ES) = 0.23) and in the variety of tasks used by their teachers (ES = 0.22). Significant declines were identified for the Israeli students in their mastery orientation (ES = 0.35), in their need to appear smart compared to others (ES = 0.32), in the attentiveness of their teachers (ES = 0.23), and in the autonomy given to them by their teachers (ES = 0.23). See Table 2.Table 2 Constructs which declined in the shift between F2F instruction and DL, and in which country the decline occured
Construct which declined Where t-value df Effect size
Engagement China 8.92 1982 0.23
Variety of tasks used by teachers China 9.97 1982 0.22
Mastery orientation Israel 6.07 307 0.35
Need to appear smart compared to others Israel 5.63 307 0.32
Attentiveness of teachers Israel 3.94 307 0.23
Autonomy given by teachers Israel 3.94 307 0.23
p-value < 0.001 for all t-tests
In general, both before and after the transition to DL, Chinese students had significantly higher values on the following constructs than Israeli students: mastery orientation towards science (ES1 = 0.34 for F2F and ES = 0.47 for DL), engagement (ES = 0.42, only before the transition to DL), need to look smarter than others (0.49 for F2F and 0.60 for DL), and in their perceptions of the teachers using a variety of tasks (ES = 0.64 for F2F and ES = 0.37 for DL), giving autonomy (ES = 0.63 for F2F and ES = 0.77 for DL), and saying there is not enough time (ES = 0.87 for F2F and ES = 0.65 for DL). See Table 3.Table 3 Constructs which had higher values in China than in Israel
Construct F2F/DL t-value Effect size
Mastery orientation F2F 5.48 0.34
Engagement 8.91 0.42
Need to appear smart compared to others 8.04 0.49
Variety of tasks used by teachers 10.4 0.64
Autonomy given by teachers 10.3 0.63
Teachers say there’s not enough time 14.2 0.87
Mastery orientation DL 7.66 0.47
Need to appear smart compared to others 9.81 0.60
Variety of tasks used by teachers 6.10 0.37
Autonomy given by teachers 12.6 0.77
Teachers say there’s not enough time 10.6 0.65
p < 0.001 and df = 2289 for all t-tests
For both the Chinese and the Israeli students, mastery orientation was significantly predicted only by teacher attentiveness (for Chinese students: β = 0.47 for F2F and β = 0.49 for DL; for Israeli students: β = 0.40 for F2F and β = 0.39 for DL) and for the Chinese participants by the teacher relating things to daily life (β = 0.43 for F2F and β = 0.39 for DL), while mastery orientation predicted engagement (for Chinese students: β = 0.34 for F2F and β = 0.18 for DL; for Israeli students: β = 0.39 for F2F), all with p < 0.001. See Table 4.Table 4 The constructs predicting mastery orientation and engagement in Chinese and Israeli students
Country Predicting construct F2F/DL Mastery orientation Engagement
beta df t-value beta df t-value
China Teacher attentiveness F2F 0.47 1982 21.7 - - -
DL 0.49 1982 22.7 - - -
Teacher relating things to daily life F2F 0.17 1982 8.03 - - -
DL 0.18 1982 8.28 - - -
Mastery orientation F2F - - - 0.34 1982 16.3
DL - - - 0.39 1982 17.6
Israel Teacher attentiveness F2F 0.40 307 7.67 - - -
DL 0.39 307 7.33 - - -
Mastery orientation F2F - - - 0.39 307 7.43
Discussion and Implications
In general, the shift from F2F instruction to DL had a negative impact the Israeli participants’ mastery orientation towards science. As a mastery orientation is considered beneficial, was found in this study to be a main predictor of engagement (see also Vedder-Weiss & Fortus, 2013), and is positively associated with many other desirable learning characteristics, such as effort and persistence (Elliot et al., 1999), self-regulation (Pintrich, 2000), and transfer (Bereby-Meyer & Kaplan, 2005), this decline is worrisome. On the other hand, Israeli students’ mastery orientation towards science and towards academic learning in general tend to decrease during adolescence (Vedder-Weiss & Fortus, 2011), so it is not clear whether the decline identified in this study is exceptional. The Chinese participants’ mastery orientation also decreased during the transition to DL, but not in a statistically significantly manner.
On the other hand, there was a significant decline in the Chinese participants’ engagement with science. This is very concerning, since without engagement, little systematic learning will occur (Irvin et al., 2007).
Compared to the Israeli students, the Chinese students were more mastery oriented and felt the need to appear smarter than their peers. This may be due to the relative competitiveness of the Chinese educational system, where students compete for places at the top high schools and then at the universities (e.g. Yin & Buck, 2015). In Israel, the system is less competitive, students can typically choose in which discipline they want to major (Israel Ministry of Education, 2022), they tend to go to local schools, and the percent of the population that obtains a postsecondary degree is very high — 47% (Organization for Economic Cooperation and Development [OECD], 2017).
The Chinese participants perceived their teachers as providing more autonomy, using a wider variety of tasks, and saying that there is not enough time more than the Israeli participants did. Since we do not have an objective measure of what the teachers actually did in their classes, we cannot know whether these perceptions indeed reflected reality or whether they were an indication of students differing expectations from their teachers. This is an issue that faces all cross-cultural comparisons. If, for example, in one culture homework is typically never given, but then a few teachers occasionally give homework, their students may feel that they are being overburdened. If in another culture homework is typically given after each lesson, but then one teacher does not give homework a few times, the students may feel like they are on vacation. What matters often is what students see their teachers doing relative to what is perceived to be the norm in their culture. Thus, it is impossible to know if Israeli science teachers actually give their students less autonomy than their Chinese counterparts. It could be that providing autonomy to students in China is relatively rare in comparison with Israel, but when it is provided in China, it is subjectively highly rated by the students, unlike in Israel where they take it for granted. Note that the teachers in this study were reputed to be good motivators of their students, and since providing autonomy is an important way for teachers to motivate their students (Vedder-Weiss, 2017), we expect that the participating teachers all provided their students with significant autonomy.
Of the three teaching dimensions from the TARGETS framework (Anderman & Midgley, 2002) that were investigated in this study (task, autonomy/authority, and time), time proved to be the most significant predictor of student mastery orientation. Teachers’ willingness to provide their students with plenty of time to ask questions until they felt that they deeply understood the material being learned before moving on was a major predictor of students’ mastery orientation towards science in both F2F instruction and DL, for students in both countries. We understand these relations as follows: a student who is driven by mastery goals wants to develop a sense of deep understanding, of mastery. When the teacher moves on to a new topic before this sense is not reached, when the student does not have the opportunity to ask questions that may help clarify and organize ideas, the student is likely to feel frustrated. When this reoccurs several times, in order to prevent the repeated frustration, the students may change their expectations the development of the sense of mastery that comes with deep understanding, and with time become less mastery oriented. Indeed, Sørum et al. (2021) found that also college students find it more difficult to ask question in DL and then to be less active in DL than in F2F instruction. This finding is in line with past research that has indicated that science teachers’ time-related practices are central to shaping students’ goal orientations (Vedder-Weiss, 2017).
Interestingly, despite significant cultural differences between the two countries and differences in the behavior of several motivational constructs, two central relationships of great importance behaved similarly in both countries and maintained their significance during the shift from F2F instruction to DL: the relationship between teachers’ attentiveness to students’ need for time to understand and the students’ mastery orientation, and the relationship between students’ mastery orientation and their engagement with science learning. Teachers’ attentiveness to students’ questions and to their need to feel they fully understand before moving on is a crucial instructional practice that is a strong predictor of students’ mastery orientation. However, the emphasis on this practice, as perceived by both Chinese and Israeli students, declined during the shift from F2F instruction to DL. Either the teachers actually provided less time for students to ask questions and develop understanding or technological issue may have prevented the students from asking questions as they would have in F2F instruction. Or perhaps technological issues made it more difficult for the teachers to identify that their students still did not fully understand, and being misinformed of the actual situation, they moved on. Whatever the reason, the students perceived this as a decline in their teachers’ attentiveness to their need to understand, and this decline appears to have led to a decline in the students’ mastery orientation towards science. And when students’ mastery orientation declined, so did their engagement with science learning, though not always significantly. Since these declines occurred in both countries, despite significant cultural differences, they are possibly driven more by the inherent affordances and constraints of each learning environment than by cultural forces. If this is the case, we expect that this particular instructional practice, attentiveness to students’ need to ask question and develop understanding, declined in many different countries and cultures and had similar impact on students’ mastery orientation towards science and their engagement with science learning. Teachers and instructional designers everywhere should be aware of this and consider changes to lesson design and the technological environment supporting DL to minimize this negative impact.
During the COVID-19 pandemic, teachers were expected to be prepared for any scenario, F2F or DL. Hopefully, there will no need in the future to return to full DL in China, Israel, or any country, but in any case, the results of this study could be used to inform science teachers everywhere of the importance of being attentive to their students’ needs to ask questions and feel like they are developing understanding, both F2F instruction and DL.
Acknowledgements
Part of the data used in this study have been used previously in other studies (Fortus et al., 2022; Passentin & Fortus, under review).
Author Contribution
All authors contributed to the study conception and design. Data collection was performed by JL and SP. The first draft of the manuscript was written by DF, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding
This study was supported by Miel de Botton, through an internal fund at the Weizmann Institute of Science.
Data Availability
The dataset generated during and analyzed during the current study is available from the corresponding author on reasonable request.
Declarations
Conflict of Interest
The authors declare no competing interests.
1 The effect sizes given in the former paragraph and in Table 2 address the magnitudes of the changes that occurred in the transition for F2F instruction to DL. The effect sizes in the present paragraph address the magnitude of the differences between the Chinese and the Israeli students.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Ames C Classrooms: Goals, structures, and student motivation Journal of Educational Psychology 1992 84 261 271 10.1037/0022-0663.84.3.261
Anderman EM Midgley C Midgley C Methods for studying goals, goal structures, and patterns of adaptive learning Goals, goal structures, and patterns of adaptive learning 2002 Lawrence Erlbaum 1 20
Anderman LH Patrick H Hruda LZ Linnenbrink EA Midgley C Observing classroom goal structures to clarify and expand goal theory Goals, goal structures, and patterns of adaptive learning 2002 Lawrence Erlbaum 243 278
Bereby-Meyer Y Kaplan A Motivational influences on transfer of problem-solving strategies Contemporary Educational Psychology 2005 30 1 1 22 10.1016/j.cedpsych.2004.06.003
Chen JF Warden CA Chang H-T Motivators that do not motivate: The case of Chinese EFL learners and the influence of culture on motivation TESOL Quarterly 2005 39 4 609 633 10.2307/3588524
Elliot AJ McCregor HA Gable S Achievement goals, study strategies, and exam performance: A mediational analysis Journal of Educational Psychology 1999 91 3 549 563 10.1037/0022-0663.91.3.549
Galton, M. (2009). Moving to secondary school: Initial encounters and their effects. Perspectives on Education: Primary Secondary Transfer in Science, 2, 5–21.
Irvin, J. L., Meltzer, J., & Dukes, M. (2007). Taking action on adolescent literacy: An implementation guide for school leaders. Association for Supervision and Curriculum Development.
Israel Ministry of Education. (2022). Kol Zchut Noar. Retrieved May 1, 2022, from https://teen.kolzchut.org.il/he/
Kaplan A Maehr ML Achievement goals and student well-being Contemporary Educational Psychology 1999 24 4 330 358 10.1006/ceps.1999.0993 10508531
Kaplan A Maehr ML The contributions and prospects of goal orientation theory Educational Psychological Review 2007 19 2 141 184 10.1007/s10648-006-9012-5
Lam S-F Jimerson S Shin H Cefai C Veiga FH Hatzichristou C Polychroni F Kikas E Wong BPH Stanculescu E Basnett J Duck R Farrell P Liu Y Negovan V Nelson B Yang H Zollneritsch J Cultural universality and specificity of student engagement in school: The results of an international study from 12 countries British Journal of Educational Psychology 2016 86 1 137 153 10.1111/bjep.12079 26009931
Liem GAD Academic and social achievement goals: Their additive, interactive, and specialized effects on school functioning British Journal of Educational Psychology 2016 86 1 37 56 10.1111/bjep.12085 26146983
Linnenbrink EA Pintrich PR Achievement goal theory and affect: An asymmetrical bidirectional model Educational Psychologist 2002 37 2 69 78 10.1207/S15326985EP3702_2
Maehr ML Anderman EM Reinventing schools for early adolescents: Emphasizing task goals The Elementary School Journal 1993 93 5 593 610 10.1086/461742
Meece JL Anderman EM Anderman LH Classroom goal structure, student motivation, and academic achievement Annual Review of Psychology 2006 57 487 503 10.1146/annurev.psych.56.091103.070258
Meissel K Rubie-Davies CM Cultural invariance of goal orientation and self-efficacy in New Zealand: Relations with achievement British Journal of Educational Psychology 2016 86 1 92 111 10.1111/bjep.12103 26856404
Organization for Economic Cooperation and Development [OECD]. (2017). Israel. In Education at a Glance 2017: OECD Indicators. Author. 10.1787/eag-2017-53-en
Osborne JA Simon SB Collins S Attitudes towards science: A review of the literature and its implications International Journal of Science Education 2003 25 9 1049 1079 10.1080/0950069032000032199
Patrick H Yoon C Early adolescents’ motivation during science investigation The Journal of Educational Research 2010 97 6 319 328 10.3200/JOER.97.6.319-328
Pintrich PR Boekaerts M Zeidner M Pintrich PR The role of goal orientation in self-regulated learning Handbook of self-regulation 2000 Academic Press 451 502
Schunk, D. H., Pintrich, P. R., & Meece, J. L. (2008). Motivation in education: Theory, research and application. Pearson.
Sørum, H., Raaen, K., & Gonzalez, R. (2021). Can Zoom replace the classroom? Perceptions on digital learning in higher education within it. In I. C. Busch, R. Frieß, M. Steinicke, & T. Wendler (Eds.), ECEL 2021 20th European Conference on e-Learning (pp. 427–434). ACI.
Utman CH Performance effects of motivational state: A meta-analysis Personality and Social Psychology Review 1997 1 2 170 182 10.1207/s15327957pspr0102_4 15647124
Vedder-Weiss D Teaching higher and lower in mastery goal structure: The perspective of students The Elementary School Journal 2017 117 4 566 592 10.1086/691584
Vedder-Weiss, D., & Fortus, D. (2011). Adolescents’ declining motivation to learn science: Inevitable or not? Journal of Research in Science Teaching, 48(2), 199–216.
Vedder-Weiss, D., & Fortus, D. (2012). Students' declining motivation to learn science: A follow up study. Journal of Research in Science Teaching, 49(9), 1057–1095.
Vedder-Weiss, D., & Fortus, D. (2013). School, teacher, peers and parents’ goals emphases and adolescents' motivation to learn science in and out of school. Journal of Research in Science Teaching, 50(8), 952–988.
Yin X Buck GA There is another choice: An exploration of integrating formative assessment in a Chinese high school chemistry classroom through collaborative action research Cultural Studies of Science Education 2015 10 3 719 752 10.1007/s11422-014-9572-5
| 0 | PMC9742641 | NO-CC CODE | 2022-12-14 23:30:05 | no | Int J Sci Math Educ. 2022 Dec 12;:1-11 | utf-8 | Int J Sci Math Educ | 2,022 | 10.1007/s10763-022-10344-9 | oa_other |
==== Front
CCF Trans. Pervasive Comp. Interact.
CCF Transactions on Pervasive Computing and Interaction
2524-521X
2524-5228
Springer Nature Singapore Singapore
121
10.1007/s42486-022-00121-6
Regular Paper
CrowdTelescope: Wi-Fi-positioning-based multi-grained spatiotemporal crowd flow prediction for smart campus
Zhang Shiyu [email protected]
Shiyu Zhang
received her Bachelor of Engineering degree in Software Engineering from Xidian University, China, in 2021. She is currently a master student with the State Key Laboratory of Internet of Things for Smart City and majoring in Artificial Intelligence Application, University of Macau, Macao, China. Her research interests lie in Urban Computing and Spatiotemporal Data Mining.
Deng Bangchao [email protected]
Bangchao Deng
Bangchao Deng received his Bachelor of Engineering degree in Computer Science and Technology from Nanjing University of Aeronautics and Astronautics, China, in 2019. He is currently a master student with the State Key Laboratory of Internet of Things for Smart City and Department of Computer and Information Science, University of Macau, Macao, China. His research interests lie in Spatiotemporal Data Mining and Urban Computing.
http://orcid.org/0000-0002-6831-0422
Yang Dingqi [email protected]
Dingqi Yang
is an Associate Professor with the State Key Laboratory of Internet of Things for Smart City and Department of Computer and Information Science, University of Macau. He received his Ph.D. degree in Computer Science from Pierre and Marie Curie University and Institut Mines-TELECOM/TELECOM SudParis in France, where he won both the CNRS SAMOVAR Doctorate Award and the Press Mention in 2015. Before joining the University of Macau, he worked as a senior researcher at the University of Fribourg in Switzerland. His research interests include big data analytics, ubiquitous computing, and smart city.
grid.437123.0 0000 0004 1794 8068 State Key Laboratory of Internet of Things for Smart City and Department of Computer and Information Science, University of Macau, Macau SAR, Macau, China
12 12 2022
114
29 8 2022
25 11 2022
© China Computer Federation (CCF) 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.
Crowd flow prediction is one of the key problems in human mobility modeling, forecasting crowd flows of locations based on historical human mobility traces. Traditional human mobility traces (collected via telecommunication companies, online social media platforms, or field studies/experiments, etc.) suffer from severe data quality issues such as low precision, data sparsity, and insufficient coverage. In this paper, we investigate crowd flow prediction using Wi-Fi connection records on the campus of a university, which imply comprehensive, large-scale, high-coverage, and multi-grained (building/floor/room level) human mobility traces. However, we are facing not only non-trivial noises in the raw Wi-Fi connection data when extracting human mobility traces, but also the trade-off between location granularities and mobility patterns when modeling multi-grained crowd flow. Against this background, we propose CrowdTelescope, a Wi-Fi-positioning-based multi-grained spatiotemporal crowd flow prediction framework. We design a systematic approach for robust human mobility trace extraction from the noisy Wi-Fi connection records and adopt spatiotemporal Graph Neural Networks to model multi-grained crowd flow under a unified graph model for the three-level location hierarchy. We also develop a prototype system of CrowdTelescope, providing the interactive visualization of crowd flows on campus. We evaluate CrowdTelescope by collecting a Wi-Fi connection dataset on the campus of the University of Macau. Results show that CrowdTelescope can effectively extract informative human mobility traces from the noisy Wi-Fi connection records with an improvement of 3.3% over baselines, and also accurately predict on-campus crowd flow across different location granularities with 1.5%-24.1% improvements over baselines.
Keywords
Mobility
Crowd flow
Wi-Fi positioning
Smart campus
http://dx.doi.org/10.13039/501100006469 Fundo para o Desenvolvimento das Ciências e da Tecnologia 0038/2021/AGJ SKL-IOTSC(UM)-2021-2023 Yang Dingqi University of MacauMYRG2022-00048-IOTSC Yang Dingqi
==== Body
pmcIntroduction
Crowd flow prediction forecasts the crowd flows of locations based on historical human mobility traces (Luca et al. 2021), which can benefit both authorities and residents. For example, it can provide insight to authorities and organizations for decision-making in various aspects, such as risk assessment (Liang et al. 2021), resource management (Chen et al. 2020), predictive policing (Yang et al. 2018), etc. Meanwhile, it can also benefit residents by better scheduling their daily activities. In particular, facing the recent COVID-19 epidemic, social distancing (i.e., avoiding crowdedness) has been suggested as an effective measure by many countries worldwide (Chang et al. 2021); in this context, accurately forecasting crowd flow can significantly help the implementation of social distancing in practice (Swain et al. 2021).
Toward the goal of accurate crowd flow prediction, it is indispensable to collect human mobility traces. On one hand, Outdoor Positioning Systems (OPS) such as global positioning systems (GPS) or cell identification (CID)-based systems have been widely deployed, providing real-time positioning services to users in an outdoor environment. Due to the low positioning precision of CID and GPS in an indoor environment, existing work using OPS mobility traces usually focuses on macroscopic mobility, such as global level (Yang et al. 2016), country level (Fan et al. 2015), or urban level (Liang et al. 2021). On the other hand, Indoor Positioning Systems (IPS) such as Wi-Fi-based or Bluetooth-based position systems provide fine-grained indoor and outdoor (depending on the density of the deployed access points/hotspots) localization. Due to the implementation constraints of these Wireless Local Access Networks (WLAN) from their service providers, the scale of the publicly available mobility traces is often small, which is usually limited to hundreds of users (e.g., about 700 students in Copenhagen Networks Study (Stopczynski et al. 2014), about 200 users in Lausanne Mobile Data Challenge (Laurila et al. 2013)).
Against this background, in this study, we focus on Wi-Fi-positioning-based human mobility traces on the campus of the University of Macau. The uniqueness of on-campus Wi-Fi connection records implies comprehensive, large-scale, high-coverage, and multi-grained human mobility traces on campus. For example, on the campus of the University of Macau, over 7,000 Wi-Fi Access Points (APs) have been deployed, covering over 80% of the campus (both indoor and outdoor areas), providing Internet services to over 10,000 students and staff, as well as to guests. With the ubiquity of smartphones and wearable devices (e.g., smart watches or bracelets), individuals carrying mobile devices leave their spatiotemporal “digital footprints” when moving on campus, which are recorded by the (automatic) connection logs between the devices and Wi-Fi APs. For example, during a typical semester weekday on 1st March 2021, we observed 2,321,420 records from 28,551 devices and 7,096 APs. However, crowd flow prediction from such data sources faces the following two issues:How to extract informative human mobility traces from noisy Wi-Fi connection records. Although Wi-Fi connection records serve as a powerful crowdsensing paradigm for mobility traces, such a data source contains two types of non-trivial noises. First, connection records from some devices (such as Wi-Fi-equipped desktops and smart home equipment) that cannot reflect human mobility need to be filtered out. Second, connection records from multiple mobile devices carried by the same user (such as smartphones, tablets, and smartwatches of the same individual) cause over-sampled mobility traces from the user, which need to be merged to alleviate the over-sampling bias. In this context, it is thus indispensable to consider these intrinsic noises and design a robust method to extract informative human mobility traces.
How to model multi-grained crowd flow from on-campus mobility traces. Wi-Fi connection records reflect the user mobility traces at different levels of granularities, usually following a three-level hierarchy of “building-floor-AP”. Modeling such mobility traces faces the trade-off between location granularity and mobility patterns, where finer-grained crowd flow usually has weaker mobility patterns and verse vice. For example, the mobility transition between APs is usually less obvious than that between buildings. On the other hand, understanding such multi-grained mobility patterns is crucial for accurate crowd flow prediction, which can serve as a “crowd telescope” to analyze crowd flows at different granularities. It is challenging to model multi-grained crowd flow with varying mobility patterns.
To address these two issues, we propose CrowdTelescope, a Wi-Fi-positioning-based multi-grained spatiotemporal crowd flow prediction framework for smart campus. To address the first issue of noisy Wi-Fi connection records, we design a robust human mobility trace extraction method, which firstly uses a heuristic-based noisy data filter to remove those devices that cannot reflect human mobility and then learns to integrate mobility traces from devices carried by the same user using cross-grained features. To address the second issue of multi-grained crowd flow modeling, we adopt spatiotemporal Graph Neural Networks (GNNs) to model multi-grained crowd flow, by formulating the location graphs of different granularities under a unified graph model considering the three-level location hierarchy (“building-floor-AP”). Finally, we develop a Web-based prototype system visualizing both historical and predicted crowd flows via an interactive map. To evaluate our CrowdTelescope, we collect an in-house Wi-Fi connection dataset on the campus of the University of Macau and perform evaluation on two tasks, i.e., human mobility trace extraction and crowd flow prediction tasks. Results show that in the human mobility trace extraction task, CrowdTelescope can effectively integrate mobility traces from devices carried by the same user, with an improvement of 3.3% over baselines; in the crowd flow prediction task, it can also make accurate predictions of crowd flow across different location granularities, yielding 1.5%–24.1% improvements over baselines without using location graphs.
Related work
Human mobility data sources
In the early stage, human mobility study mainly relies on demographic data (Ravenstein 1885), which incurs significant human effort in data collection. With the recent advance of wireless sensing and communication technologies, various positioning systems have been used to monitor and collect human mobility traces, which mainly fall into two categories, i.e., outdoor and indoor positioning systems.
First, Outdoor Positioning Systems (OPS) such as global positioning systems (GPS) or cell identification (CID)-based systems have been widely deployed, providing real-time positioning services to users in an outdoor environment. However, these systems have their intrinsic limitation on the localization precision in an indoor environment. More precisely, while CID-based positioning systems (mapping to a nearby cell tower location) have an intrinsic drawback in localization precision (about 50 m) [14], GPS (using satellite signals and trilateration) is known to have poor indoor localization precision due to the signal attenuation caused by construction materials (del Peral-Rosado et al. 2017). Subsequently, existing work using OPS mobility traces usually focuses on macroscopic mobility, such as global level (Yang et al. 2016), country level (Fan et al. 2015), or urban level (Liang et al. 2021). Moreover, these data sources often have a sparsity issue of the collected mobility traces (Yang et al. 2020) and also a low and insufficient coverage of the population in the target area, as the data is usually collected by a telecommunication service provider (CID-based positioning (Blondel et al. 2012)) or an urban transportation company (e.g., Taxi mobility traces (Yuan et al. 2010)), or is crawled on an online social network platform (e.g., Foursquare/Twitter (Yang et al. 2015, 2019, 2020)), etc.
Second, Indoor Positioning Systems (IPS) such as Wi-Fi-based or Bluetooth-based position systems provide fine-grained indoor and outdoor (depending on the density of the deployed access points/hotspots) localization. However, due to the implementation of the Wireless Local Access Networks (WLAN) and its privacy sensitivity, the scale of the publicly available mobility traces are often small, which is usually limited to hundreds of users (e.g., about 700 students in Copenhagen Networks Study (Stopczynski et al. 2014), about 200 users in Lausanne Mobile Data Challenge (Laurila et al. 2013)). To the best of our knowledge, the only work using large-scale Wi-Fi-positioning-based mobility traces in this category is from Georgia Institute of Technology (Swain et al. 2021), which involves about 40K students and 7K Wi-Fi Access Points. The scale of this data is comparable to the mobility traces at the University of Macau. However, this large-scale dataset is not publicly accessible due to privacy protection regulations. Therefore, we collect an in-house Wi-Fi connection dataset on the campus of the University of Macau to study the on-campus human mobility.
Human mobility modeling methodology
According to the problem settings (Luca et al. 2021), human mobility modeling techniques generally fall into two types of tasks (i.e., predictive and generative tasks) with two types of data representation (i.e., mobility trajectory and flow). First, predictive tasks on mobility trajectories are known as next-location prediction problems, forecasting the location of an individual based on her historical mobility traces (Wu et al. 2018; Yang et al. 2020). Second, predictive tasks on crowd flow forecast the crowd flows (the number of individuals or vehicles) of locations based on historical human mobility traces (Lin et al. 2019; Lv et al. 2014). Third, generative tasks with mobility trajectories try to generate synthetic trajectories that are similar to real-world human mobility traces in terms of statistical patterns (Liu et al. 2018; Feng et al. 2020). Finally, generative tasks with mobility flow generate synthetic flows among locations, mimicking the real-world mobility flow (Shin et al. 2020; Simini et al. 2020). These mobility modeling tasks have been widely studied to support various smart city applications, such as urban event organization (Chen et al. 2016; Yu et al. 2018), location recommendation (Yang et al. 2013; Yu et al. 2015), crowdsensing (Yu et al. 2021, 2015), and urban resource allocation (Chen et al. 2016; Wang et al. 2022), etc.
This paper studies the crowd flow prediction problem using the on-campus Wi-Fi connection records, which implies comprehensive mobility traces on campus. Accurate crowd flow prediction requires subtly capturing spatiotemporal dynamics and dependencies of crowd flow. Traditional solutions to this problem use time series prediction algorithms based on autoregression, such as AutoRegressive Integrated Moving Average (ARIMA) (Shumway et al. 2000). However, the autoregression models often ignore spatial dependencies and also fail to capture complex temporal dynamics, resulting in unsatisfied results (Luca et al. 2021). Recently, Deep Learning models have been widely used for crowd flow prediction problems. Specifically, Recurrent Neural Networks (RNNs), such as vanilla RNN (Zhang et al. 2014), Long Short-Term Memory (LSTM) (Hochreiter and Schmidhuber 1997), and Gated Recurrent Unit (GRU) (Cho et al. 2014), are designed to capture sophisticated temporal dynamics over time series and sequences. On the other hand, spatial dependencies have been modeled by applying Convolutional Neural Networks (CNNs) (Gu et al. 2018) on a crowd flow matrix (where the matrix represents the targeted geographical region and each entry in the matrix represents the flow in a spatial grid), or adopting Graph Neural Networks (GNNs) (Wu et al. 2020) on a crowd flow graph (where the graph represents the spatial connection between locations and each node represents the flow at a specific location). Recent approaches to this problem combine RNNs and CNNs/GNNs into a unified model by jointly capturing spatiotemporal dynamics and dependencies. For example, STGCN (Yu et al. 2017) combines two temporal gated convolution layers and a spatial graph convolution layer as a “sandwich” structure; STTN (Xu et al. 2020) integrates a spatial and a temporal transformers to capture dynamical directed spatial dependencies and long-range temporal dependencies, respectively; GWNET (Wu et al. 2019) designs a learnable adaptive dependency matrix to capture the hidden spatial dependencies; MTGNN (Wu et al. 2020) learns to extract uni-directed relations among multi-variate variables to capture spatial dependencies. In this paper, we explore spatiotemporal GNNs for crowd flow prediction using Wi-Fi-positioning-based human mobility traces.
CrowdTelescope
Figure 1 shows the overview of our CrowdTelescope framework. First, from the large-scale and noisy Wi-Fi connection records, we design a robust human mobility trace extraction method to obtain informative human mobility traces. Second, based on the extracted human mobility traces, we adopt spatiotemporal GNNs to model multi-grained crowd flow, by formulating the location graphs of different granularities under a unified graph model considering the three-level location hierarchy. Finally, we develop a prototype system of CrowdTelescope visualizing both historical and predicted crowd flow via an interactive map on the Web.Fig. 1 CrowdTelescope overview with three steps: (1) Robust human mobility trace extraction, (2) Multi-grained crowd flow modeling and (3) Prototype development
Robust human mobility trace extraction
With the ubiquity of smartphones and wearable devices, the raw Wi-Fi connection records imply comprehensive human mobility traces on campus. However, these records contain two types of non-trivial noises: (1) records from some devices (such as Wi-Fi-equipped desktops and smart home equipment) that cannot reflect human mobility, and (2) records from multiple mobile devices carried by the same user (such as smartphones, tablets, smartwatches of the same individual) which cause over-sampled mobility traces from the user. To extract informative human mobility traces from such noisy data, we design a robust human mobility trace extraction pipeline as follows.
Wi-Fi connection log preprocessing
According to the configuration of the Wi-Fi AP provider (Aruba Networks), the Wi-Fi connection records have six types of connection events between devices and APs, i.e., Authentication request, Authentication success, Deauthentication from station, Association request, Association success and Disassociation from station. A device is uniquely identified by a MAC address when accessing Wi-Fi (here we do not have any device meta data due to the privacy protection regulation). When a device wants to connect to an AP for the first time, the log traces are “Authentication request—Authentication success—Association request—Association success”; the authentication process requires a valid user account registered at the university to reach a success state for the internal Wi-Fi services. When a connected device moves from one AP to another, if the device enables fast-roaming, the log traces are “Association request—Association success”; otherwise the system record a full set of logs, the same as the device connects to the AP for the first time. Subsequently, to extract the mobility traces of devices, we keep the “Association success” events only, which is the final step of all these log traces, indicating the presence of a device at an AP at a certain timestamp.
Note that we ignore “Deauthentication from station” and “Disassociation from station” events due to the fact that these two events are often significantly lagged in the log traces. For example, when a device moves from one AP to another, the “Disassociation from station” event at the former AP is often recorded with a timestamp later than the “Association success” event at the latter AP; subsequently, the “Disassociation from station” event cannot be used to record one’s presence at an AP.
Noisy device trace filtering
Based on the “Association success” events of a device, we can extract a trajectory of the device, represented as a sequence of AP-timestamp pairs. However, records from some devices usually fail to reflect human mobility. Through our empirical analysis, we identify three types of such noisy devices.Non-(or low-)mobile devices, such as Wi-Fi-equipped desktops, lab equipment or smart home equipment, cannot reflect human mobility. Such a device often attaches to a small number of different APs over a long period of time. We define these devices as those that have ever connected to less than 10 different APs during a week.
Publicly shared devices, such as shared handsets for campus management and security staff, do not reflect individual’s mobility traces. A shared device often connects to a large number of APs, such as a shared handset for security patrolling. We define these devices as those that have ever connected to over 500 different APs during a week.
Devices from irregular user accounts cannot reflect real human mobility traces. Specifically, the authentication process requires a valid user account. We observe a few user accounts that are associated with many mobile devices, implying that one user account is shared by many users. In this case, the device mobility traces cannot reflect a single user’s mobility traces. We empirically define irregular user accounts as those that have been used by over five different devices.
We filter out these noisy devices according to the above criteria. As shown in Fig. 2, noisy data defined above are mostly the outliers from the data distribution. Note that the above three types of noisy devices may overlap. For example, multiple publicly shared handsets for security patrolling may use the same user account for Wi-Fi authentication.Fig. 2 Data statistics in raw data. a The frequency of the number of unique APs connected by each device per week. (b) The frequency of the number of unique devices by each user
Integration of the mobility traces of the devices of the same user
Based on the filtered device mobility traces, we need to extract human mobility traces. Specifically, if a user has only one device, the device’s mobility traces represent the user’s mobility traces. However, when a user has multiple devices (such as a smartphone and a smart watch/bracelet), these devices’ traces cause over-sampled mobility traces from the users. Therefore, it is critical to overcome this bias by integrating the mobility traces of the multiple devices of the same user. To this end, we need to identify the devices of the same user. Note that the device ownership information is not always available due to various reasons. First, the authentication may not necessarily be conducted with UM registered account; devices can also access Wi-Fi via eduroam1 Wi-Fi networks or public Wi-Fi, where we cannot link these devices to any UM registered accounts. Second, user account information for accessing Wi-Fi may be protected due to privacy issues.
Against this background, we design a novel method to learn to identify whether a pair of devices belong to the same user. When two mobile devices are carried by the same user, their spatiotemporal mobility traces should be very similar (not necessarily identical due to the stochasticity of the wireless network connection). For example, from the spatial perspective, the two devices may connect to two neighboring APs, respectively; from the temporal perspective, the two devices may connect one after the other to APs. Subsequently, it is necessary to consider these aspects to tolerate such stochasticity when measuring the similarity between two devices’ trajectories. In this study, instead of manually defining a fixed spatiotemporal tolerance to accommodate such stochasticity, we define three levels of spatial granularities and four levels of temporal granularities and design a convolutional neural network to learn to classify whether two devices belong to the same user.Fig. 3 Spatiotemporal granularities tolerating the connection stochasticity
Fig. 4 Our proposed CNN model for the classification of device pairs
Figure 3 shows our defined spatiotemporal granularities for tolerating the connection stochasticity. We consider three levels of spatial granularity, i.e., building, floor, and AP, and four levels of temporal granularity, i.e., 10 mins, 5 mins, 1 min, and 1 sec. Subsequently, each device trajectory can be transformed into these 12 spatiotemporal granularities. To compute the similarity between two devices’ trajectories, we borrow ideas from Jaccard similarity between two sets. However, instead of using Jaccard similarity directly, we define the size of the intersection and the size of the union as two features. A toy example for feature extraction is shown on the top of Fig. 4 for one granularity (AP, 1 min), where we extract two features, i.e., size of intersection and size of union. Subsequently, we can extract 24 features under the 12 spatiotemporal granularities. Based on these features, we design a Convolutional Neural Network (CNN) model to learn to classify if two devices belong to the same user, as shown at the bottom of Fig. 4. Specifically, we reshape the extracted 24 features as an “image” of 4×3 (temporal granularities by spatial granularities) with 2 channels (size of the intersection and the size of the union). Afterward, the “image” is firstly fed to a convolutional layer with nt temporal filters of size 4×1, and then fed to another convolutional layer with ns temporal filters of size 1×3, followed by a fully connected layer to output the predicted score (the probability of the input pair of devices belonging to the same user). The key idea behind this design is to let the CNN model learn to tolerate the connection stochasticity across different spatiotemporal granularities for predicting devices belonging to the same user.
The model is trained using those devices of which we have the user ownership information. For each (positive) pair of devices belonging to the same user, we randomly sample a negative pair of devices not belonging to the same user. The negative pairs are randomly sampled in each epoch during the model training process. Finally, the trained model can be used to identify the devices belonging to the same user. To extract a user’s mobility trace from her devices, we aggregate the devices’ mobility traces by taking the most active device traces within each week, to form the user’s mobility traces.
Multi-grained crowd flow prediction
Based on the extracted user mobility traces, we adopt spatiotemporal GNN models for predicting multi-grained crowd flow. Specifically, Wi-Fi connection records reflect user mobility traces at different levels of granularities, following a three-level hierarchy of “building-floor-AP”. In this context, mobility modeling faces the trade-off between location granularity and mobility patterns, where finer-grained crowd flow usually has weaker mobility patterns, and verse vice (as evidenced by our experiments later). Therefore, we train three independent spatiotemporal GNNs for the respective location granularities, under a unified graph model capturing spatial dependencies between locations. In the following, we first present crowd flow estimation from extracted mobility traces, followed by our proposed crowd flow prediction models using spatiotemporal GNNs.
Crowd flow estimation from human mobility traces
Based on the filtered and aggregated human mobility traces, we estimate the crowd flow as follows. Considering the practical use cases (e.g., the frequency of campus loop shuttle is 10–15 min), we first define the targeted temporal granularity for crowd flow prediction as 10 min and assume that a user can contribute to only one AP’s flow in this period of time. The crowd flow of one AP is then estimated as the total number of users contributing to the AP in each time slot of 10 min. However, this crowd flow estimation method has to consider the following practical issues raised from our empirical analysis of the dataset.An extracted human mobility trace may connect to multiple APs in one time slot. In this case, we select the AP with the most connection records as the contributed AP. In contrast, if a user mobility trace has no connection in a time slot, we estimate her associated APs as follows.
When a user stays at the same place for a long time (e.g., attending a class for 45 min or sleeping in the dormitory), her devices may enter to a sleeping mode and do not have any connection records. In this case, if a user is associated with the same AP in a row, we assume the user always contributes to the AP’s flow during that time period.
If the two consecutively associated APs are different and the time interval is less than one hour (considering the campus size of 1.09 km2), we assume that the user is moving from the first AP to the second one. The user thus contributes to the flow of the first/second AP in the first/second half of the time interval, respectively.
If the two consecutively associated APs are different and the time interval is greater than one hour, we assume that the user is away from the campus and thus does not contribute to any APs’ flow during that time period.
Following these heuristics, we estimate the crowd flow for each AP in each time slot. Subsequently, we can further compute the crowd flow for other location granularities. Specifically, the locations of APs follow a three-level hierarchy of “building-floor-AP”. For example, an AP of ID “E11-GF-22” is located at the building E11, on the ground floor (“GF”). Using this information, the flow in one floor/building is computed as the sum of flows of all APs located in the floor/building, respectively.
Crowd flow modeling using spatiotemporal GNNs
Crowd flow modeling requires to capture the spatiotemporal dynamics and dependencies of the input flow. To this end, we adopt spatiotemporal GNNs which have been shown as a powerful technique for solving various crowd flow modeling problems (Luca et al. 2021; Wang et al. 2021). Specifically, spatiotemporal GNNs combines a GNN model that leverages a graph structure to encode spatial dependencies and a temporal component (mostly RNN models) that learns the temporal dynamics of the flow. In the following, we first present our unified graph modeling process for the three-level hierarchy of locations, followed by the spatiotemporal GNN models.Fig. 5 Graph modeling with the three-level hierarchy of locations
The graph topology uniquely defines the spatial dependencies between locations in GNNs. Figure 5 shows our graph model for the three-level location hierarchy of “building-floor-AP”; the hierarchy is represented by the red dashed edges on the right panel. First, for the building level, we adopt Delaunay triangulation (Delaunay 1934), which is a widely used method for surface morphology studies in Geographic Information Systems (GIS). Specifically, it treats each building as a node (with its GPS coordinates) and connects the nodes to form a triangular irregular network, ensuring that no node lies within the interior of any of the circumcircles of the triangles in the network. Note that Delaunay triangulation is the dual graph for Voronoi diagrams (Boots et al. 2009) which is also a popular spatial tessellation method in GIS. The left panel of Fig. 5 shows the Delaunay triangulation on buildings on the maps of the University of Macau, which is an unweighted and undirected graph (all edges have the same distance of one). Second, based on the building graph, we use the “building-floor” hierarchy to construct a floor graph, following three principles: (1) the floor nodes belonging to the same building are fully connected with a distance of one; (2) the floor nodes of two neighboring buildings (on the building graph) are connected with a distance computed by traversing the building graph, which is three in this case (floor→building→building→floor), where all the hierarchy edges also have a distance of one; (3) the floor nodes of non-neighboring buildings are not connected. Finally, following the similar logic of constructing a graph from the graph of the higher hierarchy, we construct an AP graph based on the floor graph. Note that the edges in the floor graph now have either a distance of one or three; the AP nodes of neighboring floors (in the floor graph) are now connected via (AP→floor→floor→AP) with a distance of either three (in the case of floor→floor having the distance of one) or five (in the case of floor→floor having the distance of three).
Based on the constructed graphs, we adopt spatiotemporal GNNs to model the spatiotemporal dynamics and dependencies of crowd flows. We train three independent spatiotemporal GNNs for the respective location granularities. Figure 6 shows the spatiotemporal GNNs with the building graph as an example. Specifically, we have a flow graph for each time slot, where each node (building) in this graph is associated with the computed crowd flow as its attribute. Given a sequence of such flow graphs in the past, the crowd flow prediction tries to forecast the attribute (crowd flow) of each node in this graph in a future time slot. Note that the topological structure of this graph is the same over time, while the node attribute (crowd flow) evolves. Following this problem formulation and settings, we will experiment with a sizeable collection of state-of-the-art spatiotemporal GNN models (see our experiments below) and adopt the best-performing one in our CrowdTelescope. Note that CrowdTelescope as a general framework can flexibly integrate with any spatiotemporal GNN models.Fig. 6 Crowd flow prediction using spatiotemporal GNNs. We use the building graph as a toy example
Prototype development
We develop a prototype system “CrowdTelescope” as a smart campus application. The prototype system is built with an interactive user interface visualizing both historical and forecasted crowd flows on campus, providing decision support to a wide range of users, including the campus management team, students, staff and visiting guests, etc. Specifically, Fig. 7 shows a snapshot of our Web user interface built on top of Mapbox.2 We use heat maps to visualize the crowd flow, where the hotspots can be easily identified by their color. A user-friendly interactive visualization interface is provided through a few control options. Users can switch between historical and forecasted crowd flow. For the historical crowd flow, users can specify a date and click the start/pause button to visualize the crowd flow of the selected date as a video. The progress bar also serves as an option to flexibly control (by sliding on the progress bar) the time of the crowd flow that users want to visualize. For the forecasted crowd flow, users can visualize the predicted crowd flow for the current day, through the same control penal as for the historical crowd flow.Fig. 7 The Web user interface of CrowdTelescope prototype https://pursue1221.github.io/CrowdTelescope/. The screenshot shows the visualization of the historical crowd flow on 01 Mar. 2021. We observe active crowd flow transition from residential colleges to central teaching buildings right before 10 am
Experiment
We evaluate our CrowdTelescope using a Wi-Fi connection record dataset on two tasks, i.e., human mobility trace extraction and crowd flow prediction. We present the experiment setup below, followed by the results and discussion.
Experiment setup
Dataset
We collect Wi-Fi connection records on the campus of the University of Macau for four consecutive weeks in March 2021. Table 1 shows the statistics of the dataset across different data processing steps. From the raw data, we first filter device traces using the criteria discussed in Sect. 3.1.2. We observe that 63% of the devices and 53% of the connection records are removed, which implies the raw data contains a large amount of noisy data, including non-(or low-)mobile devices, publicly shared devices, and devices from irregular user accounts. Afterward, we extract human mobility traces from the filtered device traces by identifying and integrating the mobility traces of the devices of the same user, using our proposed method in Sect. 3.1.3. We observe that in the final integrated user traces, the number of devices is slightly higher than the number of users, which implies that for a few users, the most active devices are different across different weeks. In other words, most of the users have a unique active device across the four weeks of the data collection period.Table 1 Dataset statistics in different data processing steps
Data processing steps Raw data Device traces (Sect. 3.1.2) User traces (Sect. 3.1.3)
#Device 48,565 18,136 13,621
#User 29,743 13,593 13,593
#Record 52,174,535 24,266,224 21,027,546
#Record per device 1,074 1,338 1,544
#Device per user 1.63 1.33 1.00
Evaluation protocol and baselines
We evaluate our CrowdTelescope in both human mobility trace extraction and crowd flow prediction tasks. We present our evaluation protocol and baselines for each task below.
For the human mobility trace extraction task, the key problem is formulated as a classification task to classify whether two devices belong to the same user, as discussed in Sect. 3.1.3. To evaluate our proposed method, we first consider a Single-Grained Feature, i.e., Jaccard similarity between the mobility traces of two devices under a single spatiotemporal granularity, and learn a threshold using a decision tree algorithm for classification. In addition, based on our Cross-Grained Features, i.e., 24 features with intersection and union features on each of the 12 spatiotemporal granularities as shown in Fig. 4, we consider several popular classification techniques as baselines, including Multi-Layer Perceptron, Logistic Regression, Naive Bayes, Decision Tree and Random Forest. For our proposed method CrowdTelescope (CNNs), we set the numbers of temporal and spatial filters as nt=16 and ns=64, respectively. To evaluate the classification performance, we collect a set of positive device pairs (belonging to the same user), and randomly sample the same amount of negative device pairs (not belonging to the same user). We split them into 80% training and 20% test datasets, with a balanced amount of positive and negative data in both. We report the accuracy for each method averaged over 10 repeated trails (randomly sample negative samples in each trail).
For the crowd flow prediction task, we follow the problem setting as specified in Fig. 6, and consider the following baselines. First, traditional time series prediction methods include Historical Average (HA), AutoRegressive Integrated Moving Average (ARIMA) (Shumway et al. 2000), Support Vector Regression (SVR) (Platt 1999). Second, deep sequence models include Recurrent Neural Networks (RNN) (Zhang et al. 2014), Long Short-Term Memory (LSTM) (Hochreiter and Schmidhuber 1997), and Gated Recurrent Unit (GRU) (Cho et al. 2014). For these two types of baselines, the flow of each node (building, floor or AP) is considered as an independent time series, and prediction is made only based on these time series without using the graph structure. Finally, we consider the following spatiotemporal GNNs which can all be integrated in our CrowdTelescope: DCRNN (Li et al. 2017) capturing the spatial dependency using bidirectional random walks on the graph and the temporal dependency using an encoder-decoder architecture; STGCN (Yu et al. 2017) combining two temporal gated convolution layers and a spatial graph convolution layer as a “sandwich” structure; STTN (Xu et al. 2020) combining a spatial and a temporal transformers to capture dynamical directed spatial dependencies and long-range temporal dependencies, respectively; HGCN (Guo et al. 2021) considering the hierarchical structure of location networks; GWNET (Wu et al. 2019) using a learnable adaptive dependency matrix to capture the hidden spatial dependencies; MTGNN (Wu et al. 2020) learning to extract uni-directed relations among multi-variate variables through a graph learning process for better capturing spatial dependencies. For each level of location granularities (building/floor/AP), we train each method using the first three week data and evaluate on the last week’s data. This experiment uses the algorithm implementation from LibCity (Wang et al. 2021).
Performance on human mobility trace extraction
Table 2 shows the results comparing different features and methods for device pair classification.Table 2 Accuracy of identifying devices of the same user. Best performing results in each category of the methods are highlighted in bold
Method Accuracy
Single Grained Feature Building, 10 mins 0.8811
Floor, 10 mins 0.8889
AP, 10 mins 0.8651
Building, 5 mins 0.8844
Floor, 5 mins 0.8901
AP, 5 mins 0.8668
Building, 1 min 0.8807
Floor, 1 min 0.8847
AP, 1 min 0.8764
Building, 1 sec 0.8055
Floor, 1 sec 0.8007
AP, 1 sec 0.7671
Cross Grained Features Multi-Layer Perceptron 0.8884
Logistic Regression 0.8895
Naive Bayes 0.8756
Decision Tree 0.8591
Random Forest 0.9019
CrowdTelescope (CNNs) 0.9037
First, comparing single-grained features across different spatiotemporal granularities, we observe the varying performance. In particular, the finest spatiotemporal granularity (AP, 1 sec) yields the worst performance, failing to accommodate the network connection stochasticity when identifying devices from the same user; under this granularity, even two devices always carried by the same user will not have similar traces due to connection stochasticity. When moving to coarser spatiotemporal granularities, the performance increases, while the best performing granularity is (Floor, 5mins). When further coarsening the granularity, the performance slightly drops, due to a high false positive rate; the devices of different users will have similar traces under the coarser spatiotemporal granularities.
Second, compared to single-grained features, we observe the cross-grained features achieve better performance on average with an improvement of 3.3% (average accuracy of 0.8864 and 0.8576 for cross-grained and single-grained features, respectively). This implies that cross-grained features are more informative than single-grained features for predicting devices belonging to the same user. Furthermore, compared to baseline classification techniques, our proposed method achieves the best performance, showing the superiority of our designed CNN architecture when learning to tolerate the connection stochasticity across different spatiotemporal granularities.
Performance on crowd flow prediction
To evaluate the crowd flow prediction performance, we report three metrics: Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Weighted Mean Absolute Percentage Error (WMAPE). Smaller values of these metrics imply better performance. The greater difference between MAE and RMSE, the greater the variance in the individual errors in the test set. Compared to MAE and RMSE, WMAPE discounts the absolute values of flow (e.g., different scales of flow values in building, floor and AP levels) and is robust against varying flows with very small values (e.g., zero flow values of some APs in some time slots); it can thus support the performance comparison across different levels of location granularities. Table 3 shows the results.Table 3 Crowd flow prediction performance. Best performing results are highlighted in bold for each metric. Note that “-” denotes the case where the method run out of GPU memory on our test PC with NVIDIA GeForce RTX 3090 of 24GB RAM
Method Building Floor AP
MAE RMSE WMAPE MAE RMSE WMAPE MAE RMSE WMAPE
ARIMA 10.585 18.173 0.253 2.920 4.909 0.677 0.483 1.135 1.499
HA 13.387 29.422 0.162 3.677 9.568 0.319 0.704 1.766 1.130
SVR 11.149 28.394 0.135 2.569 8.286 0.223 0.385 1.246 0.617
RNN 16.236 33.492 0.197 4.231 11.010 0.368 0.555 1.817 0.894
GRU 15.286 31.598 0.185 4.012 10.618 0.349 0.556 1.833 0.895
LSTM 14.941 31.464 0.181 3.858 10.380 0.336 0.556 1.826 0.895
DCRNN 8.800 19.164 0.107 2.502 7.849 0.218 – – –
STGCN 8.246 18.137 0.100 2.489 6.447 0.217 0.406 1.230 0.653
STTN 11.421 25.721 0.139 3.292 8.687 0.287 – – –
HGCN 8.876 16.876 0.109 2.414 5.466 0.212 0.317 1.096 0.515
GWNET 7.870 17.877 0.095 2.256 6.474 0.196 0.321 1.225 0.517
MTGNN 7.722 15.481 0.094 2.228 5.967 0.194 0.314 1.178 0.505
First, we observe that spatiotemporal GNNs achieve significantly better performance in general, compared to traditional time series prediction techniques and deep sequence models. This implies that our graph model captures crucial information on the spatial dependencies of locations of different granularities, which can significantly improve the crowd flow prediction performance. In particular, the best spatiotemporal GNN model, i.e., MTGNN, yields an improvement of 24.1%, 1.5% and 10.9% (on building, floor and AP levels, respectively) over the best-performing baselines without using location graphs. We thus adopt it in our prototype system.
Second, comparing the crowd flow performance across different levels of location granularities, we observe that finer-grained crowd flow usually has weaker mobility patterns. Specifically, comparing WMAPE of each method across different location granularities, we observe that finer granularities have a larger value of WMAPE, which is consistent for all methods. This implies that the finer-grained crowd flow shows weaker patterns and thus is more difficult to model. Note that MAE and RMSE are smaller for finer-grained locations, which is due to the smaller absolute values for flows of finer-grained locations; they are thus not appropriate for the performance comparison across different location granularities.
Discussion
Although we show that CrowdTelescope can achieve accurate crowd flow prediction, there are still inevitable data biases due to the data collection and preprocessing, causing the discrepancy between the mobility observed from Wi-Fi connection records and the actual mobility on campus. We discuss two major data biases below.The coverage of Wi-Fi connection records over the actual population on campus. The population of Wi-Fi users may not cover the actual population on campus, such as some users who prefer using cellular networks rather than Wi-Fi. However, according to the official statistics of the University of Macau3,4 there are 13,787 staff and students in 2021, which is close to the number of users processed by the User Traces in Table 2. We thus believe that the Wi-Fi connection records can well represent the human mobility on the whole campus. Note that the raw data include much more user accounts due to the fact that the accounts of the same user for internal Wi-Fi and eduroam are different, leading to the almost doubled number of users compared to the number of actual users. By extracting device traces, the number of users is already reduced by half, because students and staff mostly prefer internal Wi-Fi instead of eduroam, while the latter is mostly for guests from other educational institutions.
The bias of integrating the mobility traces of the devices of the same user. When integrating the mobility traces of the devices of the same user, we may have both false positives and false negatives. For example, if two devices of two classmates are together quite often, they may be treated as the same user; if a user has two devices that are carried alternatively, they may be treated as different users. However, as CrowdTelescope can achieve over 90% accuracy in classifying device pairs, we believe the integrated mobility traces are informative to represent the overall on-campus mobility.
Conclusion
In this paper, we propose CrowdTelescope, a Wi-Fi-positioning-based multi-grained spatiotemporal crowd flow prediction framework for smart campus. Specifically, crowd flow prediction using Wi-Fi connection records faces not only non-trivial noises in the raw connection records, but also the trade-off between location granularities and mobility patterns. To address the first issue, we design a robust human mobility trace extraction method, which firstly uses a heuristic-based noisy data filter to remove those devices that cannot reflect human mobility and then learns to integrate mobility traces from devices carried by the same user using cross-grained features. To address the second issue, we adopt spatiotemporal Graph Neural Networks (GNNs) to model multi-grained crowd flow, by formulating the location graphs of different granularities under a unified graph model considering the three-level location hierarchy (“building-floor-AP”). We also develop a prototype system of CrowdTelescope, providing the interactive visualization of crowd flows on campus. We evaluate CrowdTelescope by collecting a Wi-Fi connection dataset on the campus of the University of Macau. Results show that CrowdTelescope cannot only effectively extract informative human mobility traces from the noisy Wi-Fi connection records (outperforming baselines by 3.3%), but also accurately predict on-campus crowd flow across different location granularities (yielding 1.5%–24.1% improvements over baselines).
In the future, we plan to further investigate unified spatiotemporal GNNs to directly learn from the hierarchical location graphs, jointly modeling crowd flows across different location granularities.
Acknowledgements
This work is funded by the University of Macau (MYRG2022-00048-IOTSC) and the Science and Technology Development Fund, Macau SAR (0038/2021/AGJ and SKL-IOTSC(UM)-2021-2023). This work was performed in part at SICC which is supported by SKL-IOTSC, University of Macau. The authors also appreciate the data support from the Information and Communication Technology Office (ICTO) at the University of Macau.
Declarations
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
1 https://eduroam.org/.
2 https://www.mapbox.com/.
3 https://reg.um.edu.mo/qfacts/y2021/staff/.
4 https://reg.um.edu.mo/about-reg/facts-and-figures/students-figures/.
S. Zhang and B. Deng have contributed equally to this work.
==== Refs
References
Blondel, V.D., Esch, M., Chan, C., Clérot, F., Deville, P., Huens, E., Morlot, F., Smoreda, Z., Ziemlicki, C.: Data for development: the d4d challenge on mobile phone data. arXiv preprint arXiv:1210.0137 (2012)
Boots, B., Sugihara, K., Chiu, S.N., Okabe, A.: Spatial tessellations: concepts and applications of voronoi diagrams (2009)
Chang S Pierson E Koh PW Gerardin J Redbird B Grusky D Leskovec J Mobility network models of covid-19 explain inequities and inform reopening Nature 2021 589 7840 82 87 10.1038/s41586-020-2923-3 33171481
Chen, L., Zhang, D., Wang, L., Yang, D., Ma, X., Li, S., Wu, Z., Pan, G., Nguyen, T.-M.-T., Jakubowicz, J.: Dynamic cluster-based over-demand prediction in bike sharing systems. In: Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing, pp. 841–852 (2016)
Chen L Jakubowicz J Yang D Zhang D Pan G Fine-grained urban event detection and characterization based on tensor cofactorization IEEE Transactions on Human-Machine Systems 2016 47 3 380 391 10.1109/THMS.2016.2596103
Chen L Yang D Nogueira M Wang C Zhang D Data-driven c-ran optimization exploiting traffic and mobility dynamics of mobile users IEEE Trans. Mob. Comput. 2020 20 5 1773 1788 10.1109/TMC.2020.2971470
Cho, K., Van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., Bengio, Y.: Learning phrase representations using rnn encoder-decoder for statistical machine translation. arXiv:1406.1078 (2014)
del Peral-Rosado JA Raulefs R López-Salcedo JA Seco-Granados G Survey of cellular mobile radio localization methods: from 1g to 5g IEEE Commun. Surv. Tutor. 2017 20 2 1124 1148 10.1109/COMST.2017.2785181
Delaunay, B., : Sur la sphere vide. Izv. Akad. Nauk SSSR, Otdelenie Matematicheskii i Estestvennyka Nauk 7(793-800), 1–2 (1934)
Fan, Z., Song, X., Shibasaki, R., Adachi, R.: Citymomentum: an online approach for crowd behavior prediction at a citywide level. In: Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing, pp. 559–569 (2015)
Feng, J., Yang, Z., Xu, F., Yu, H., Wang, M., Li, Y.: Learning to simulate human mobility. In: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pp. 3426–3433 (2020)
Gu J Wang Z Kuen J Ma L Shahroudy A Shuai B Liu T Wang X Wang G Cai J Recent advances in convolutional neural networks Pattern Recogn. 2018 77 354 377 10.1016/j.patcog.2017.10.013
Guo, K., Hu, Y., Sun, Y., Qian, S., Gao, J., Yin, B.: Hierarchical graph convolution network for traffic forecasting. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, pp. 151–159 (2021)
Hochreiter S Schmidhuber J Long short-term memory Neural Comput. 1997 9 8 1735 1780 10.1162/neco.1997.9.8.1735 9377276
Laurila JK Gatica-Perez D Aad I Blom J Bornet O Do TMT Dousse O Eberle J Miettinen M From big smartphone data to worldwide research: the mobile data challenge Pervasive Mob. Comput. 2013 9 6 752 771 10.1016/j.pmcj.2013.07.014
Li, Y., Yu, R., Shahabi, C., Liu, Y.: Diffusion convolutional recurrent neural network: Data-driven traffic forecasting. arXiv preprint arXiv:1707.01926 (2017)
Liang, Y., Ouyang, K., Sun, J., Wang, Y., Zhang, J., Zheng, Y., Rosenblum, D., Zimmermann, R.: Fine-grained urban flow prediction. In: Proceedings of the Web Conference 2021, pp. 1833–1845 (2021)
Lin, Z., Feng, J., Lu, Z., Li, Y., Jin, D.: Deepstn+: Context-aware spatial-temporal neural network for crowd flow prediction in metropolis. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 33, pp. 1020–1027 (2019)
Liu, X., Chen, H., Andris, C.: trajgans: Using generative adversarial networks for geo-privacy protection of trajectory data (vision paper). In: Location Privacy and Security Workshop, pp. 1–7 (2018)
Luca M Barlacchi G Lepri B Pappalardo L A survey on deep learning for human mobility ACM Comput. Surv. (CSUR) 2021 55 1 1 44 10.1145/3485125
Lv Y Duan Y Kang W Li Z Wang F-Y Traffic flow prediction with big data: a deep learning approach IEEE Trans. Intell. Transp. Syst. 2014 16 2 865 873
Platt J Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods Advances in large margin classifiers 1999 10 3 61 74
Ravenstein EG The laws of migration J. Stat. Soc. Lond. 1885 48 2 167 235 10.2307/2979181
Shin, S., Jeon, H., Cho, C., Yoon, S., Kim, T.: User mobility synthesis based on generative adversarial networks: A survey. In: 2020 22nd International Conference on Advanced Communication Technology (ICACT), pp. 94–103 (2020). IEEE
Shumway, R.H., Stoffer, D.S., Stoffer, D.S.: Time Series Analysis and Its Applications vol. 3. Springer, ??? (2000)
Simini, F., Barlacchi, G., Luca, M., Pappalardo, L.: Deep gravity: enhancing mobility flows generation with deep neural networks and geographic information. arXiv preprint arXiv:2012.00489 (2020)
Stopczynski A Sekara V Sapiezynski P Cuttone A Madsen MM Larsen JE Lehmann S Measuring large-scale social networks with high resolution PLoS One 2014 9 4 95978 10.1371/journal.pone.0095978
Swain, V.D., Xie, J., Madan, M., Sargolzaei, S., Cai, J., De Choudhury, M., Abowd, G.D., Steimle, L.N., Prakash, B.A.: WiFi mobility models for COVID-19 enable less burdensome and more localized interventions for university campuses. medRxiv (2021)
Wang, L., Chai, D., Liu, X., Chen, L., Chen, K.: Exploring the generalizability of spatio-temporal traffic prediction: meta-modeling and an analytic framework. IEEE Transactions on Knowledge and Data Engineering (2021)
Wang, J., Jiang, J., Jiang, W., Li, C., Zhao, W.X.: Libcity: An open library for traffic prediction. In: Proceedings of the 29th International Conference on Advances in Geographic Information Systems. SIGSPATIAL ’21, pp. 145–148. Association for Computing Machinery, New York, NY, USA (2021). 10.1145/3474717.3483923
Wang L Yu Z Guo B Yang D Ma L Liu Z Xiong F Data-driven targeted advertising recommendation system for outdoor billboard ACM Transactions on Intelligent Systems and Technology (TIST) 2022 13 2 1 23
Wu, Z., Pan, S., Long, G., Jiang, J., Chang, X., Zhang, C.: Connecting the dots: Multivariate time series forecasting with graph neural networks. In: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pp. 753–763 (2020)
Wu, Z., Pan, S., Long, G., Jiang, J., Zhang, C.: Graph wavenet for deep spatial-temporal graph modeling. arXiv preprint arXiv:1906.00121 (2019)
Wu R Luo G Shao J Tian L Peng C Location prediction on trajectory data: A review Big data mining and analytics 2018 1 2 108 127 10.26599/BDMA.2018.9020010
Wu Z Pan S Chen F Long G Zhang C Philip SY A comprehensive survey on graph neural networks IEEE transactions on neural networks and learning systems 2020 32 1 4 24 10.1109/TNNLS.2020.2978386
Xu, M., Dai, W., Liu, C., Gao, X., Lin, W., Qi, G.-J., Xiong, H.: Spatial-temporal transformer networks for traffic flow forecasting. arXiv preprint arXiv:2001.02908 (2020)
Yang, D., Fankhauser, B., Rosso, P., Cudre-Mauroux, P.: Location prediction over sparse user mobility traces using rnns. In: Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, pp. 2184–2190 (2020)
Yang, D., Qu, B., Yang, J., Cudré-Mauroux, P.: Lbsn2vec++: Heterogeneous hypergraph embedding for location-based social networks. IEEE Transactions on Knowledge and Data Engineering (2020)
Yang, D., Qu, B., Yang, J., Cudre-Mauroux, P.: Revisiting user mobility and social relationships in lbsns: a hypergraph embedding approach. In: The World Wide Web Conference, pp. 2147–2157 (2019)
Yang, D., Zhang, D., Yu, Z., Wang, Z.: A sentiment-enhanced personalized location recommendation system. In: Proceedings of the 24th ACM Conference on Hypertext and Social Media, pp. 119–128 (2013)
Yang D Zhang D Chen L Qu B Nationtelescope: Monitoring and visualizing large-scale collective behavior in lbsns J. Netw. Comput. Appl. 2015 55 170 180 10.1016/j.jnca.2015.05.010
Yang D Zhang D Qu B Participatory cultural mapping based on collective behavior data in location-based social networks ACM Trans. Intell. Syst. Technol. (TIST) 2016 7 3 1 23 10.1145/2814575
Yang D Heaney T Tonon A Wang L Cudré-Mauroux P Crimetelescope: crime hotspot prediction based on urban and social media data fusion World Wide Web 2018 21 5 1323 1347 10.1007/s11280-017-0515-4
Yu, Z., Ma, H., Guo, B., Yang, Z.: Crowdsensing 2.0. Communications of the ACM 64(11), 76–80 (2021)
Yu, B., Yin, H., Zhu, Z.: Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting. arXiv preprint arXiv:1709.04875 (2017)
Yu Z Xu H Yang Z Guo B Personalized travel package with multi-point-of-interest recommendation based on crowdsourced user footprints IEEE Transactions on Human-Machine Systems 2015 46 1 151 158 10.1109/THMS.2015.2446953
Yu Z Zhang D Yu Z Yang D Participant selection for offline event marketing leveraging location-based social networks IEEE Transactions on Systems, Man, and Cybernetics: Systems 2015 45 6 853 864 10.1109/TSMC.2014.2383993
Yu Z Yi F Lv Q Guo B Identifying on-site users for social events: Mobility, content, and social relationship IEEE Trans. Mob. Comput. 2018 17 9 2055 2068 10.1109/TMC.2018.2794981
Yuan, J., Zheng, Y., Zhang, C., Xie, W., Xie, X., Sun, G., Huang, Y.: T-drive: driving directions based on taxi trajectories. In: Proceedings of the 18th SIGSPATIAL International Conference on Advances in Geographic Information Systems, pp. 99–108 (2010)
Zhang, Y., Dai, H., Xu, C., Feng, J., Wang, T., Bian, J., Wang, B., Liu, T.-Y.: Sequential click prediction for sponsored search with recurrent neural networks. In: AAAI (2014)
| 0 | PMC9742642 | NO-CC CODE | 2022-12-14 23:30:05 | no | CCF Trans. Pervasive Comp. Interact.. 2022 Dec 12;:1-14 | utf-8 | null | null | null | oa_other |
Subsets and Splits