Vulnerable and risk groups amidst the COVID-19 pandemic: Preventive behaviours, preferences, and perceptions

Abstract

Objectives

Having an underlying medical condition or comorbidities, as well as being of advanced age, makes large groups of the population more vulnerable to detrimental effects of COVID-19. Among those effects are higher rate of hospitalization, increased mortality, and slower recovery. To protect these vulnerable groups and the population at large actions were taken by governments, particularly the implementations of preventive measures, both pharmaceutical and non-pharmaceutical.

Study Design

The data as whole stem from the Austrian Corona Panel Project (ACPP), which is a nationally representative panel-survey conducted in Austria.

Methods

This study makes use of panel-data from 2021 and 2022 in a longitudinal analysis through pooled regression models (n=2123). In addition, an in-depth analysis is provided with cross-sectional data from winter 2021 (n=1183) and by means or seemingly unrelated regressions (SUR). The predictor variables are age and two different measures of comorbidities. Dependent variables are preventive behaviours, preferences and perceptions. Robustness checks with control variables are carried out as well.

Results

Both in the longitudinal and the cross-sectional analysis age above 65 years is the clearest predictor for all investigated preventive behaviours, preferences and perceptions. People above 65 years of age show significant patterns to be more compliant, to support stricter pharmaceutical and non-pharmaceutical measures and to perceive others’ preventive behaviours and opinions on measures more favourably. Having one or several underlying diseases, disorders or injuries correlates positively with most preventive behaviours, as well as with preferences for more stringent public health measures. By contrast age, people with comorbidities perceive other people’s compliance and opinions on preventive measures as less pronounced.

Conclusions

People of advanced aged and those with underlying medical conditions exhibit stronger preventive behaviour, including vaccination readiness. They also support more stringent measures, even compulsory vaccination. These results can be interpreted as self-directed because vulnerable and risk groups have strong incentives to avoid becoming infected with COVID-19. It can be argued that such behaviours and preferences are directed at others to facilitate reciprocal actions by the public at large, and if necessary also to make preventive measures mandatory.

Introduction

Some groups within our societies are at heightened risk to be affected by COVID-19. They are more likely to become infected with the virus, to have more severe clinical courses (e.g. 1,2) and are also particularly vulnerable to other adverse effects in the spheres of the medical, the social and the economic (cf. 3).

Studies have shown that such risk groups are especially prone to suffer from impaired mental health throughout the COVID-19 pandemic (e.g. 4). They frequently develop inferior immune defence responses offered by the vaccines compared to the rest of the population (5). Such negative consequences, paired with already pre-existing vulnerabilities, often result in social exclusion, diminished mobility, substance abuse, disability to work, and less well-being overall (6, 7).

Equally, incisive life events, such as loss of employment, may lead to decreased mental health, engagement in unhealthy behaviour, or other factors that are adverse to health. Thus, poor or ill health often exists in a downward spiral or vicious circle with other social or economic hardships, irrespective if they arise from brute or option luck. So, there are significant groups in the population, who experience inequitable health care access, interventions and outcomes, due to medical, social and economic factors.

In this study health risk groups will be focused on because they are the most definitely at risk to suffer adverse consequences from the pandemic. Further, the investigated medical risk groups are largely more vulnerable due to brute luck. Specifically, the investigation will focus on heightened risk due to advanced age and due to pre-existing and often chronic diseases, disorders or injuries. Other relevant dimensions that exacerbate COVID-19 related risks, such as obesity and smoking of tobacco products, are not part of this study.

Existing research has concentrated on the effects of the COVID-19 pandemic on risk groups. This study will instead shed light on risk groups’ preventive behaviours, preferences for measures against the Corona-virus, and perceptions regarding personal and public risk, as well as perceived compliance with anti-COVID-19 measures. The research within the scope of this paper will explicitly consider pharmaceutical and non-pharmaceutical measures to contain the spread of the virus. This includes vaccine readiness, preferences for compulsory vaccination, as well as compliance with government measures.

Many underlying medical conditions - especially cardiovascular, respiratory and cancerous ones - and advanced age have been associated with more severe COVID-19 infections and higher mortality rates (8). Subsequently, people with such pre-existing conditions and also the elderly have been prioritised to receive vaccines, right after the initial immunisation of health care professionals. As lies in the nature of vaccine scarcity, this priority setting strategy is not without societal and ethical implications and trade-offs (cf. 9). However, it has been a key goal in many countries to avoid or minimise the need for ICU-triage through prioritisation on levels prior to intensive care (cf. 10). Underlying comorbidities were frequently included as relevant criteria in such prioritisation decisions (11).

Apart from vaccination many governmental measures and instruments have been put in place to decrease the transmission of the Coronavirus. Among these are physical distancing, wearing of protective face-masks, remote work, closures of many kinds of facilities, rapid antigen and PCR tests, travel restrictions, as well as general lockdowns. Research on compliance with preventive measures during the COVID-19 pandemic has so far focused on the general population and thereby on socio-demographic characteristics and personality factors (e.g. 12, 13). Scarce prior evidence shows that risk groups and vulnerable populations exhibit more stringent preventive behaviour and more supportive preferences for restrictive policy responses than other parts of the population (14, 15).

This study provides a comprehensive picture of vulnerable and risk groups amidst the COVID-19 pandemic in terms of compliant behaviour, preferences for policy responses, and perceptions of others’ preventive behaviours and opinions on response policies. These include pharmaceutical and non-pharmaceutical measures.

Methods

Data

The data used in this study come from the Austrian Corona Panel Project (ACPP), which started data collection in March 2020. ACPP respondents are quota sampled to closely reflect the Austrian resident population in key demographic characteristics. Included are Austrian residents with the minimum age of 14. The ACPP participation amounts to approximately 1500 online interviews per panel wave. Details on the ACPP data and methodology are provided in Kittel et al. (16, 17).

Variables & analytical strategy

For this article pooled OLS and pooled Logistic regressions with clustered standard errors are employed to grasp the inter-variable relationships, using panel data. The five times the panels were fielded were mid-April 2021, late June/early July 2021, late September/early October 2021, late November/early December 2021, and mid/late February 2022. The first dependent variable is individually reported compliance with three COVID-19 counter-measures, which are transformed into one index-variable by means of principal component analysis (PCA). Another dependent is the uptake of COVID-19 tests (antigen, PCR or antibody) that have been done four weeks prior to each panel. Third, vaccine readiness is measured as whether one has received at least one dose of a vaccine against COVID-19 (dichotomous). A further dependent variable in the panel-design is the degree of support for compulsory COVID-19 vaccination. Finally, this study includes self-reported number of social contacts in the week prior to taking the survey and not counting persons in shared households. Predictor variables are having none, one or several medical conditions (cardiovascular diseases, diabetes, hepatitis C, chronic lung diseases, chronic kidney diseases, cancer), as well as age. Age is grouped as a dichotomous variable, with people who are 65 years or below and ones who are older than 65.^[1] Equal or similar cut-offs between age cohorts have been used in previous research (e.g. 18, 19, 20). After listwise deletion the panel sample size is 2123 panellists with 6633 individual observations.

To test and enhance the panel-study design a cross-section of ACPP data is analysed as well. The data for this cross-section stem from late November/early December 2021 (ACPP wave 27). The corresponding analyses are conducted by means of seemingly unrelated regressions (SUR). First dependent variable is an additive index (alpha=0.93) constructed from preferences for 13 different measures intended to contain the Coronavirus. Further dependent variables are perceptions on other Austrians’ compliance and on others’ opinions on the COVID-19 restrictions (both as PCA index). Preferences for compulsory vaccination are also included in the cross-sectional design, as well as number of social contacts. In the cross-section vaccine readiness is treated as ordinal dependent variable with four categories, i.e. having received 0 to 3 doses of vaccines. Grouped age remains as a predictor. A multiple correspondence analysis (MCA) is performed on the basis of 16 categories of diseases, disorders and injuries, which are structured as dummy variables. The MCA scores on the 1st dimension with 72.62% (principal inertia). The 1st dimension of the MCA is the second predictor variable. When a MCA value is low, a person has none or very few diseases. When a MCA value is high, a person suffers from several conditions, i.e. from comorbidities. To account for outliers the 1st and the 99th percentiles of the MCA’s first dimension are winsorised within both age groups (65 and below vs. above 65). After listwise deletion of missings the cross-sectional sample size amounts to 1183 respondents. All variables, except for the MCA 1st dimension, are normalised to have values lying between 0 and 1. The 1 dimension of the MCA is centred around 0 before winsorisation and has a post-winsorisation mean of  -0.011. The winsorised 1st dimension of the MCA correlates strongly with an additive measure of comorbidities (correlation=0.974). For more detailed information on each variable and index view the appendix. Control variables are included in the robustness checks (appendix). All statistical computations were performed with Stata 16.1.

Results

Panel analysis

Information on sample composition and intersectionality is located in the appendix. 

In the panel-analysis five dependent variables are put under investigation via pooled OLS regressions and one pooled Logit regression (vaccination readiness with min. 1 dose). Results regarding pre-existing conditions and comorbidities are listed, followed by findings regarding age. P-values and 95% confidence intervals are put in parentheses only when significant at the p<0.100 level. People who reported one underlying medical condition exhibited higher compliance with preventive measures (p=0.074, 95% CI  -0.003 to 0.061), whereas persons with comorbidities significantly complied less (p=0.039, 95% CI -0.108 to  -0.003). Looking at vaccination readiness, as at least one dose received vs. none-received, neither having one nor having several medical conditions produces significant results. People with one underlying condition took fewer Corona-tests (p=0.001, 95% CI  -0.079 to -0.019), while for people with comorbidities there is no significant result. Persons with one medical condition supported compulsory vaccination (p=0.031, 95% CI 0.005 to 0.101), and so did those with comorbidities (p=0.064, 95% CI  -0.004 to 0.131). Last, both people with one condition (p=0.000 95% CI -0.010 to -0.030) and with comorbidities (p=0.006, 95% CI -0.105 to -0.017) reported having met fewer social contacts.

Now turning to results for age, people above 65 years report higher compliance with preventive measures (p=0.000, 95% CI 0.069 to 0.125), higher vaccination readiness (p=0.000, 95% CI 0.576 to 1.088), a lower test uptake (p=0.000, 95% CI -0.121 to  -0.062), stronger support for compulsory vaccination (p=0.000, 95% CI 0.083 to 0.177), and fewer contacts (p=0.000, 95% CI -0.145 to -0.090). Regression tables are provided in the appendix.

Cross-sectional analysis

The cross-sectional analysis considers eight dependent variables with data from late November/early December 2021. Low values on the dimension 1 of the MCA correspond to having no or few indicated diseases, disorders or injuries. High values relate to suffering from several medical conditions, i.e. from comorbidities.

Turning to preferences for the governmental response to contain the Coronavirus, people that scored high on the MCA exhibited more support for stringent measures (p=0.069, 95% CI -0.001 to 0.031). Persons, who had high MCA scores, were more likely to be vaccinated against COVID-19 (p=0.002, 95% CI 0.012 to 0.050). By contrast, those with low MCA scores reported higher test uptakes than those with high MCA scores (p=0.004, 95% CI -0.051 to -0.001). 

High MCA scores correlate positively with self-reported compliance with preventive behaviour (p=0.031, 95% CI  0.002 to 0.040). People, who were high on the first dimension of the MCA, perceived others’ preventive behaviour to be less compliant (p=0.034, 95% CI -0.034 to  -0.001) and also perceived others’ opinions on the government responses to be less supportive (p=0.068, 95% CI -0.032 to 0.001). 

Finally, persons with comorbidities were more in favour of compulsory vaccination (p=0.016, 95% CI 0.006 to 0.056) and reported fewer social contacts (p=0.037, 95% CI -0.037 to -0.001). 

People above the age of 65 exhibited more supportive preferences for government measures (p=0.001, 95% CI 0.031 to 0.115), a higher vaccination readiness (p=0.000 , 95% CI   0.090 to 0.191), a lower test uptake (p=0.000, 95% CI -0.160 to -0.054), stronger compliance (p= 0.000, 95% CI  0.068 to 0.166), more supportive preferences for compulsory vaccination (p=0.000, 95% CI 0.080 to 0.210), and fewer social contacts (p=0.000, 95% CI -0.204 to -0.111). However, persons in the advanced age group perceived others’ compliant behaviour (p=0.016, 95% CI 0.010 to 0.093) and others’ opinions towards government measures (p=0.001, 95% CI  0.029 to 0.114) more favourably than the younger cohorts did. Regression tables are provided in the appendix.

Robustness checks with control variables

For the robustness checks four control variables are considered: gender (binary), trust in the public health care system, trust in the federal government, and perceived personal health risk from COVID-19. Gender has been previously identified as a significant variable in partially explaining preventive behaviour (21, 22) and so has trust in governments (23, 24), as well as risk perception (25). 

In the cross-sectional analysis perceived personal health risk strongly predicts stronger preferences (p=0.000, 95% CI  0.289 to 0.386) for and better compliance (p=0.000, 95% CI 0.194 to 0.326) with preventive measures. It also correlates positively with perceived behaviour (p=0.005, 95% CI 0.025 to 0.140) and opinions (p=0.000, 95% CI 0.083 to 0.199) of others, and negatively with number of social contacts (p=0.000, 95% CI  -0.290 to -0.160). Trust in the public health care system seems to facilitate stronger compliance (p=0.000, 95% CI 0.109 to 0.247) and support for preventive measures (p=0.000, 95% CI 0.232 to 0.334. The same is true for trust in the federal government with compliance (p=0.002, 95% CI  0.040 to 0.175) and preferences for preventive measures (p=0.000, 95% CI 0.122 to 0.221). Finally, women appear to show stronger compliance (p=0.014, 95% CI  0.009 to 0.077) and report fewer social contacts (p=0.000, 95% CI -0.094 to -0.027). Detailed results of all robustness checks are located in the appendix.

As for the main predictors - underlying medical conditions, comorbidities and age - the robustness checks largely corroborate the initial main findings.

Discussion

This study delivers three key take-aways: 1) The presence of underlying medical conditions and of comorbidities matters. 2) Age matters a great deal. 3) Both of these do not eliminate each other in their effects.

The emphasis of this article on risk groups’ behaviours, preferences and perceptions regarding preventive measures, both pharmaceutical and non-pharmaceutical, amidst the COVID-19 pandemic offers some further implications.

Revisiting the fact that about one third of the panel-sample in each wave used for this study had one or several medical conditions and/or was over the age of 65, policy makers face a considerably sized minority, which has strong motives to avoid becoming infected with COVID-19. Besides being their own explicit advocates, these risk groups could play a crucial role in establishing social norms, related to compliance with preventive measures, vaccination uptake or other preemptive practices. As it will not be difficult to convince most vulnerable people of many necessary public health measures in crisis situations, they should still not be overlooked when it comes to setting and reinforcing best practices in public health. It will need a share of the public to initiate and promote pharmaceutical and non-pharmaceutical actions beyond certain thresholds.

Altogether, this article has shown that medically exposed risk groups support stringent government measures during a pandemic and also follow through by acting on said measures, with the exception of getting frequently tested. However, the risk groups perceive others’ compliance with preventive measures and others’ opinions on COVID-19 restrictions as lower than the remaining population, that is not at heightened risk, does. Thus, while risk groups in heterogeneous societies may be the vanguard in terms of collective pandemic response, they are not convinced by others’ reciprocal attitudes and actions.

Health policies to improve the situation of vulnerable and risk groups on a global scale would be the promotion of health literacy (26) and of universal health coverage, as well as of social protection (27). I argue that public health stakeholders and health policy makers should not solely rely on and appeal to peoples’ self-interest when implementing measures, such as to contain a highly transmissible and adaptable virus. Instead I pinpoint the role of social norms in facilitating and sustaining cooperative settings within society. While such norms may have a self-serving origin or base, they nevertheless have the potential to foster or prevent specific behaviours in groups of the public, who would otherwise not behave in the desired ways. Social norms can function as incentives or as punishments to attain behaviour supportive of public health goals. One common norm that should be named is the one of reciprocity in dyads, groups, networks or the public at large. Generally, reciprocity as nudge or as sanction can stabilise cooperative behaviour over a longer period of time.

Addressing this issue, previous research highlights the roles of group processes to foster adherence to public health behaviour (28) and of social norms to facilitate preventive behaviour (29). Such group dynamics and social norms become vital mechanisms, particularly when seeing public information and pandemic preparedness as public goods (cf. 30). Yet, as in any public good dilemma, pitfalls exist. Free-riders can undermine preventive behaviour, as for instance physical distancing (31). By contrast, vulnerable and risk groups in a pandemic scenario have strong incentives to comply and cooperate in order to protect themselves and to establish a social norm of reciprocity.

On a final note it should be said that those most vulnerable ought to be a core concern to public health theorists and practitioners alike. This holds particularly true when it comes to pandemic preparedness and to boosting the public’s resilience to crises. Therein, reciprocal behaviour assumes a central position in health care systems, including their accessibility, financing and provision on an equitable basis. Besides elevated health risk and advanced age, we also ought not to forget that a gradient in health care exists, which puts further pressure on the already economically and socially disadvantaged.

Declarations

Ethical approval

Research ethics approval for this study was not required according to institutional and national guidelines.

Competing interests 

The authors declare that they have no competing interests.

Data sharing

The datasets generated during and/or analysed for the current study are available upon request: https://viecer.univie.ac.at/coronapanel/austrian-corona-panel-data/access-request/ and partially as open access in the AUSSDA repository, https://data.aussda.at/

References

1. Yang, H., Chen, W., Hu, Y., Chen, Y., Zeng, Y., Sun, Y., Ying, Z., He, J., Qu, Y., Lu, D. & Song, H. Pre-pandemic psychiatric disorders and risk of COVID-19: a UK Biobank cohort analysis. The Lancet Healthy Longevity, 2020, 1(2), e69-e79.
2. Ssentongo, P., Ssentongo, A. E., Heilbrunn, E. S., Ba, D. M., & Chinchilli, V. M. Association of cardiovascular disease and 10 other pre-existing comorbidities with COVID-19 mortality: a systematic review and meta-analysis. PloS one, 2020, 15(8), e0238215.
3. The Lancet, Editorial. Redefining vulnerability in the era of COVID-19. Lancet (London, England), 2020, 395(10230), 1089.
4. O'Connor, R.C., Wetherall, K., Cleare, S., McClelland, H., Melson, A.J., Niedzwiedz, C.L., O'Carroll, R.E., O'Connor, D.B., Platt, S., Scowcroft, E. & Watson, B. Mental health and well-being during the COVID-19 pandemic: longitudinal analyses of adults in the UK COVID-19 Mental Health & Wellbeing study. The British Journal of Psychiatry, 2021, 218(6), 326–333.
5. Juthani, P. V., Gupta, A., Borges, K. A., Price, C. C., Lee, A. I., Won, C. H., & Chun, H. J. Hospitalisation among vaccine breakthrough COVID-19 infections. The Lancet Infectious Diseases, 2021, 21(11), 1485–1486.
6. Mechanic, D., & Tanner, J. Vulnerable people, groups, and populations: societal view. Health Affairs, 2007, 26(5), 1220–1230.
7. Glover, R. E., van Schalkwyk, M. C., Akl, E. A., Kristjannson, E., Lotfi, T., Petkovic, J., Petticrew, M.P., Pottie, K., Tugwell, P. & Welch, V. A framework for identifying and mitigating the equity harms of COVID-19 policy interventions. Journal of clinical epidemiology, 2020, 128, 35–48.
8. Jordan, R. E., Adab, P., & Cheng, K. Covid-19: risk factors for severe disease and death. Bmj, 2020, 368.
9. Russell, F. M., & Greenwood, B. Who should be prioritised for COVID-19 vaccination?. Human vaccines & immunotherapeutics, 2021, 17(5), 1317–1321.
10. Wang, Q., Wang, X., & Lin, H. The role of triage in the prevention and control of COVID-19. Infection Control & Hospital Epidemiology, 2020, 41(7), 772–776.
11. Jöbges, S., Vinay, R., Luyckx, V. A., & Biller-Andorno, N. Recommendations on COVID‐19 triage: international comparison and ethical analysis. Bioethics, 2020, 34(9), 948–959.
12. Wright, L., Steptoe, A., & Fancourt, D. Patterns of compliance with COVID-19 preventive behaviours: a latent class analysis of 20 000 UK adults. J Epidemiol Community Health, 2022, 76(3), 247–253.
13. Liu, J., Tong, Y., Li, S., Tian, Z., He, L., & Zheng, J. Compliance with COVID-19-preventive behaviours among employees returning to work in the post-epidemic period. BMC Public Health, 2022, 22(1), 1–10.
14. Blayac, T., Dubois, D., Duchêne, S., Nguyen-Van, P., Ventelou, B., & Willinger, M. Population preferences for inclusive COVID-19 policy responses. The Lancet Public Health, 2021, 6(1), e9.
15. Yıldırım, M., Geçer, E., & Akgül, Ö. The impacts of vulnerability, perceived risk, and fear on preventive behaviours against COVID-19. Psychology, health & medicine, 2021, 26(1), 35–43.
16. Kittel, B., Kritzinger, S., Boomgaarden, H., Prainsack, B., Eberl, J. M., Kalleitner, F., Lebernegg, N. S.; Partheymüller, J., Plescia, C., Schiestl, D. W. & Schlogl, L. The Austrian Corona Panel Project: monitoring individual and societal dynamics amidst the COVID-19 crisis. European Political Science, 2021, 20(2), 318–344.
17. Kittel, B., Kritzinger, S., Boomgaarden, H., Prainsack, B., Eberl, J. M., Kalleitner, F., Lebernegg, N. S.; Partheymüller, J., Plescia, C., Schiestl, D. W. & Schlogl, L. Austrian Corona Panel Project (SUF Edition). AUSSDA, 2020.
18. Carethers, J. M. Insights into disparities observed with COVID-19. Journal of internal medicine, 2021, 289(4), 463–473.
19. Moline, H.L., Whitaker, M., Deng, L., Rhodes, J.C., Milucky, J., Pham, H., Patel, K., Anglin, O., Reingold, A., Chai, S.J. and Alden, N.B. Effectiveness of COVID-19 vaccines in preventing hospitalization among adults aged ≥ 65 years—COVID-NET, 13 states, February–April 2021. Morbidity and Mortality Weekly Report, 2021, 70(32), 1088.
20. Cohn, B. A., Cirillo, P. M., Murphy, C. C., Krigbaum, N. Y., & Wallace, A. W. SARS-CoV-2 vaccine protection and deaths among US veterans during 2021. Science, 2022, 375(6578), 331–336.
21. Pinchoff, J., Santhya, K. G., White, C., Rampal, S., Acharya, R., & Ngo, T. D. Gender specific differences in COVID-19 knowledge, behavior and health effects among adolescents and young adults in Uttar Pradesh and Bihar, India. PloS one, 2020, 15(12), e0244053.
22. Latkin, C. A., Dayton, L., Yi, G., Colon, B., & Kong, X. Mask usage, social distancing, racial, and gender correlates of COVID-19 vaccine intentions among adults in the US. PloS one, 2021, 16(2), e0246970.
23. Min, C., Shen, F., Yu, W., & Chu, Y. The relationship between government trust and preventive behaviors during the COVID-19 pandemic in China: exploring the roles of knowledge and negative emotion. Preventive medicine, 2020, 141, 106288.
24. Han, Q., Zheng, B., Cristea, M., Agostini, M., Bélanger, J. J., Gützkow, B., Kreienkamp, J., PsyCorona Collaboration & Leander, N. P. Trust in government regarding COVID-19 and its associations with preventive health behaviour and prosocial behaviour during the pandemic: a cross-sectional and longitudinal study. Psychological medicine, 2021, 1–11.
25. Qin, H., Sanders, C., Prasetyo, Y., Syukron, M., & Prentice, E. Exploring the dynamic relationships between risk perception and behavior in response to the Coronavirus Disease 2019 (COVID-19) outbreak. Social Science & Medicine, 2021, 285, 114267.
26. Paakkari, L., & Okan, O. COVID-19: health literacy is an underestimated problem. The Lancet. Public Health, 2020, 5(5), e249.
27. Barron, G.C., Laryea-Adjei, G., Vike-Freiberga, V., Abubakar, I., Dakkak, H., Devakumar, D., Johnsson, A., Karabey, S., Labonté, R., Legido-Quigley, H. & Lloyd-Sherlock, P. Safeguarding people living in vulnerable conditions in the COVID-19 era through universal health coverage and social protection. The Lancet Public Health, 2021, 7(1), e86-e92.
28. Drury, J., Carter, H., Ntontis, E., & Guven, S. T. Public behaviour in response to the COVID-19 pandemic: understanding the role of group processes. BJPsych open, 2021, 7(1).
29. Kittel, B., Kalleitner, F., & Schiestl, D. W. Peers for the fearless: Social norms facilitate preventive behaviour when individuals perceive low COVID-19 health risks. Plos one, 2021, 16(12), e0260171.
30. Stein, F., & Sridhar, D. Health as a “global public good”: creating a market for pandemic risk. bmj, 2017, 358.
31. Cato, S., Iida, T., Ishida, K., Ito, A., McElwain, K. M., & Shoji, M. Social distancing as a public good under the COVID-19 pandemic. Public health, 2020, 188, 51–53.