key: cord-0962771-ddyk4jcn
authors: Moran, Chelsea; Campbell, David J T; Campbell, Tavis S; Roach, Pamela; Bourassa, Lyne; Collins, Zoe; Stasiewicz, Marysia; McLane, Patrick
title: Predictors of attitudes and adherence to COVID-19 public health guidelines in Western countries: a rapid review of the emerging literature
date: 2021-03-11
journal: J Public Health (Oxf)
DOI: 10.1093/pubmed/fdab070
sha: dbb6a7232db09281c7c282cfee368f6bd6c57983
doc_id: 962771
cord_uid: ddyk4jcn

BACKGROUND: Physical distancing, wearing face masks and hand hygiene are evidence-based methods to protect the public from coronavirus disease 2019 (COVID-19) infection. There has been a proliferation of research examining characteristics that can be targeted by public health interventions. This rapid review sought to identify predictors of attitudes toward and adherence to COVID-19 public health guidelines, and identify interventions aiming to improve adherence. METHODS: Articles were retrieved from multiple databases (e.g. MEDLINE, CINAHL and medRxiv) on 6 August 2020. Studies were limited to samples collected from Western countries. Studies were classified according to the types of factor (s) examined as independent variables. The consistency of evidence for each factor was scored by two reviewers. RESULTS: In total, 1323 unique articles were identified in the initial search, resulting in 29 studies in the final synthesis. The available evidence suggests individuals who are older, identify as women, trust governments, perceive COVID-19 as threatening and access information through traditional news media are more likely to adhere with COVID-19 public health guidelines. Interventions for improving adherence have not yet been investigated thoroughly, and this review identified only three experimental studies. CONCLUSIONS: This review has identified several characteristics that impact attitudes and adherence to COVID-19 public health guidelines.

The incidence of the coronavirus disease 2019 (COVID- 19) , the infection caused by the virus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has dramatically increased throughout the world. The World Health Organization (WHO) declared COVID-19 to be an international pandemic on 11 March 2020. In an effort to control the impact of COVID-19 on public health, national and local governments worldwide have recommended or mandated a variety of mitigation measures. Physical distancing, wearing face masks and hand hygiene are evidence-based non-pharmacological interventions designed to reduce transmission of SARS-CoV-2. Broad public uptake and longterm maintenance of these measures have been identified as essential to reduce transmission and minimize burden on health care systems. [1] [2] [3] Recent predictive modelling from Canada estimates that without the implementation of public health measures, 64.6% of the population would become infected with COVID-19, and ∼3.6% of those infected would die from COVID-19 related illness by January 2022. 4 Behavioral public health measures are crucial to curb infection rates as no curative treatment for COVID-19 is currently available and it is unclear in many jurisdictions when approved vaccines will be widely available to the general population. 5 As such, some form of these protective behaviors may be required into 2022, with risk of pandemic resurgence remaining elevated into 2024. 6 Behavioral mitigation procedures rely on public adherence to key health behaviors. However, adherence to these measures varies and there is interest in exploring individuallevel characteristics that predict adherence to COVID-19 guidelines, which can be targeted by public health messaging and interventions. Dozens of large national and international surveys have been conducted across the world to describe the relationship between various individual characteristics on attitudes and rates of adherence to COVID-19 public health guidelines. There is a need to synthesize the current state of knowledge in order to identify predictive factors that can be targeted by public health interventions, and to highlight gaps in this area.

The purpose of this rapid review is to summarize the emerging literature to provide insight into the following research questions:

(1) What factors impact attitudes toward COVID-19 public health guidelines, including physical distancing, wearing face masks and hand hygiene? (2) What factors impact adherence to COVID-19 public health guidelines, including physical distancing, wearing face masks and hand hygiene? (3) What interventions can create more positive attitudes toward following public health guidelines with the goal of increasing guideline adherence?

This study is a rapid review informed by the development protocol for the upcoming Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) rapid review guidelines. 7 A rapid review is a knowledge synthesis methodology that is designed to provide preliminary insight into an urgent research question. 7, 8 This methodology is appropriate to generate a preliminary summary COVID-19 behavioral research.

The results of this review were originally reported in the Alberta Health Services COVID-19 Scientific Advisory Group Rapid Evidence Report on Attitudes and Adherence to COVID-19, published on 25 September 2020. 9 The inclusion/exclusion criteria were selected to retrieve studies most applicable to the Canadian context.

A literature search was conducted by a librarian from the Knowledge Management Department of Alberta Health Services on 6 August 2020. The search was designed to capture articles from the academic and grey literature, including preprints. The search was completed in OVID MEDLINE, PubMed, CINAHL, LitCovid, TRIP PRO, WHO Global research on coronavirus, COVID-19 Primer, National Collaborating Centre for Methods and Tools, medRxiv, bioRxiv, Google and Google Scholar. The MED-LINE search is reproduced in Supplementary Table 1.

Titles and abstracts identified in the search were reviewed by the librarian for an initial relevance screening, to exclude studies that were obviously not related to the purpose of the current review. One reviewer then screened the remaining titles and abstracts according to pre-specified inclusion and exclusion criteria (Table 1) .

Data extraction was completed by seven individual coders and was not conducted in duplicate due to time constraints. A standardized data extraction form, which was refined throughout the data extraction process, was used to collect information about study design, jurisdiction, sample size, study characteristics, sampling methods, independent variables (i.e. factors) and outcomes, mediating/moderating variables, reference groups used in statistical analyses and results (including effect sizes, confidence intervals and Pvalues).

Factors related to attitudes or adherence to COVID-19 public health guidelines were summarized in tabular format. Two independent raters assessed consistency of study results within each factor by examining studies that reported statistically significant results. Factors were labeled as high consistency (>80% of studies show an association of similar strength in the same direction), moderate consistency (50-79% of studies show an association of similar strength in the same direction), low consistency (≤50% of studies 

A total of 29 studies were included in the final synthesis. Database searches yielded 2562 results before deduplication, resulting in 1323 unique titles and abstracts to be screened. Initial screening resulted in 1101 articles excluded by the librarian, leaving 222 articles for full-text screening by the research team. After this selection process, 69 articles remained and initial data extraction was performed. In an effort to increase the quality of studies included in the synthesis, the study team decided to exclude purely descriptive studies and studies relying solely on convenience sampling methods if the sample size was <1000 participants, and where stratification, weighting or resampling analyses were not undertaken. After reviewing the 69 remaining articles according to the new criteria, 29 articles were retained for full data extraction.

Study characteristics for the 29 included studies are summarized in Table 2 . Studies originated from Europe, 10 17 experimental designs (k = 2) 21,32 and media analysis (k = 1). 37 Convenience sampling methods were used in almost all included studies (k = 14 convenience samples and k = 13 stratified convenience samples), with the exception of two studies that relied on random sampling 14 and quota sampling 34 methods. About one-third of the included studies (k = 11) were preprints 10, 12, 13, 15, 21, 23, 28, 32, 33, 36, 38 and one (Continued) 

Outcomes assessed by included studies are reported in Table 2 . Outcomes can be classified into three broad categories: (i) adherence to specific COVID-19 protective behaviors; (ii) overall adherence to COVID-19 public health guidelines and (iii) various types of attitudes related to COVID-19 (e.g. intention to adhere, misperceptions, resistance to public messaging, risk perception and belief in conspiracies). Outcomes were typically measured with selfreport items, with the exception of two studies using mobility data, 19, 38 and one study examining tweets. 37

Studies reported on a wide range of factors, summarized in Fig. 1 . Extracted data, including outcomes, effect size and statistical significance, are organized by factor in Supplementary Table 2 . Since most included studies primarily examined behavioral outcomes rather than attitudes, we decided to combine all outcome types in the final synthesis for ease of interpretation. The most frequently examined factors related to attitudes or adherence to COVID-19 public health guidelines were age (k = 14), sex or gender (k = 14), trust in government or authorities (k = 11) and education (k = 11). Results from these clusters of studies suggest that older age, being female/identifying as a woman, and having greater trust in government or health authorities are all factors that predict greater adherence to COVID-19 public health guidelines, whereas education was not related to adherence or attitudes. Other factors impacting attitudes toward and adherence to COVID-19 guidelines are summarized in Table 3 .

Only three studies investigating the effects of interventions on attitudes or adherence to COVID-19 public health recommendations were identified in this review. Yousuf et al. 17 conducted an uncontrolled experimental study using convenience samples (n = 16 072 [diagnostic survey] and n = 17 189 [post-campaign survey]) in the Netherlands. They report that exposure to both a targeted video campaign featuring a 22year-old male social media influencer and a related newspaper article with infographics improved handwashing duration and thoroughness. Everett et al. 21 conducted an experimental study exploring the effects of moralistic messaging and message source on intentions to adhere to public health guidelines using a stratified convenience sample (n = 1032). They found that messages stressing duty to wash one's hands (i.e. we are obliged to wash our hands for the sake of others) were more impactful than messages stressing that hand washing is virtuous (i.e. hand washing helps you be your best self). However, significant effects of message type were not observed for physical distancing behaviors.

Gutierrez et al. 32 investigated the effects of accurate or estimated COVID-19 death reports on adherence to physical distancing. They randomized 1022 participants to either receive accurate information about COVID-19 death toll (which accounts for delay in death reports) or estimates that do not account for delays in reporting and hence represent an underestimation of the COVID-19 death toll. Participants exposed to estimated death tolls were more likely to report lower intentions of complying with shelter-at-home recommendations and report a lower perceived risk of contagion when compared to participants who received accurate death toll data.

This rapid review identified 29 studies investigating predictors of attitudes and/or adherence to COVID-19 protective behaviors or reporting on effects of interventions to improve Notes:

The following factors were only examined by single studies include in this review, and therefore are not included in this table: COVID-19 related experiences (e.g. tested, diagnosed, etc.), 18 Media attention, 37 Prevalence and existing policies, 25 Provincial Residence and 28 Social networks (i.e. family, school and quality of social networks). 14 Statistical significance was determined based on the alpha level defined by the authors of each included study.

Two independent raters assessed consistency of study results within each factor by examining studies that reported statistically significant results. attitudes or adherence. According to these findings, the bulk of the current literature consists of cross-sectional surveys that use convenience sampling methods without correcting for sampling error. Although the initial purpose of this review was to identify factors that impact attitudes toward COVID-19 public health guidelines and factors that impact adherence to these behaviors, it was most productive to examine these outcomes together given the limited scope of evidence available.

To date, studies consistently show a positive association between attitudes/adherence and a number of individual characteristics: age, women/female sex, trust in governments and perceived threat of COVID-19. Less frequently mentioned factors positively related to adherence were higher socio-economic status, accessing traditional media sources, trust in science or medicine, perceived effectiveness of guidelines, ability to follow guidelines and larger households. Factors related to decreased adherence to COVID-19 public health guidelines were political conservativism and belief in conspiracy theories. Whereas, education, employment status, trust in others, race and health status were unrelated or inconsistently related to adherence.

This review identified a large gap in the COVID-19 literature: strategies for promoting adherence to public health COVID-19 guidelines have not been robustly investigated to date. Many recommendations for promoting guideline adherence from the literature are speculative since very few interventional studies or quasi-experimental studies have been published to date. Authors generally offer logical suggestions based on inferential findings based on results from convenience sample surveys, rather than evidence from tested interventions to change attitudes or behaviors. The most promising strategies appear to be communications to increase knowledge about the pandemic and perceived threat of the virus, and improve trust in government or authorities.

Evidence supporting specific messaging and content to enable behavior change in line with COVID-19 public health recommendations is very weak and limited. However, a robust field of literature exists in sociology and psychology regarding behavior change in multiple health and social contexts. This evidence would likely provide more helpful conclusions than the sparse literature currently available related to COVID-19. Reputable sources for guidance include the broader social psychology literature and established frameworks for influencing behavior change (e.g. Behavior Change Wheel 39 ), other related public health campaigns which have more rigorous evidence (i.e. hand hygiene) and local community and public engagement activities that engage minority groups, whose voices may be underrepresented in broad populationlevel surveys. Municipalities may also benefit from relying on their own jurisdictional data collection on public perceptions, which should be rigorously designed and follow guidelines for the appropriate conduct of survey-based research, 40, 41 and consider applying the recently released WHO methodology for conducting iterative behavioral insights research on COVID-19. 42, 43 What this study adds This review identified that those with limited knowledge of the pandemic, those who felt that COVID-19 posed a low risk, and those who were unconvinced of the efficacy of public health guidelines were more likely to exhibit consistently poor adherence. Public health messaging should therefore aim to improve general knowledge of the COVID-19 pandemic, and in particular, focus on the threat posed by the virus and the efficacy of public health guidelines to mitigate risk. Messaging should also be designed to target groups of individuals at higher risk of non-adherence or those with more negative attitudes about COVID-19 public health guidelines. This includes younger people, men, those who self-identify as politically conservative and those who are prone to lower levels of trust in government or science. Although the current review did not identify interventions targeting specific groups at higher risk for non-adherence to COVID-19 measures, an in-depth analysis of communication strategies used by nine democratic jurisdictions identified five broad strategies to enhance population-level adherence that could be applicable to both adherent and non-adherent groups. These include relying on supporting autonomy rather than placing broad orders, linking pandemic measures to existing sociopolitical values and positive emotions, receiving and incorporating feedback from citizens (especially from groups at high risk of nonadherence), communication frameworks emphasizing swift and transparent communication and framing COVID-19 as a democratic challenge requiring mass action. 44 Government and public health officials should attempt to create an environment that enables adherence to public health guidelines by addressing systemic and structural factors. This review highlighted three studies that consistently found that individuals' capacity to comply with public health guidelines was a significant driving factor in determining adherence levels. Interventions that promote behaviors to limit virus transmission require careful consideration of individual opportunity to adhere to COVID-19 preventive behaviors. 45 For instance, hand hygiene and mask wearing can be supported by providing widespread access to required materials (e.g. tissues, cleaning products, disposable and/or reusable masks) and appropriate facilities for safe disposal and/or decontamination of soiled products. Other behaviors, such as physical distancing and self-isolation when experiencing symptoms, require more complex systemic changes such as changes in spatial layouts of public spaces, access to home-based methods of work and financial support of individuals who do not have access to employment benefits that cover sick days or days taken off work to self-isolate.

Most studies identified in this review consisted of crosssectional survey studies recruited using convenience sampling methods. Non-random sampling approaches compromise representativeness of the sample and produce results that are at high risk of bias, unless sampling error is accounted for through statistical correction. Further, as most studies are point-in-time studies, they do not account for change in drivers of attitudes and behaviors as the pandemic progressed. There were also issues with reporting of results, as some studies only report measures of effect size, frequently without information on statistical significance, while others presented only correlation or regression coefficients. In addition, few studies attended to health equity considerations or accounted for minority population groups' perspectives. A further weakness of the literature is that factors impacting guidelines and outcomes assessed are inconsistently defined and reported, making between-study comparison difficult.

The results of this review should be interpreted in the context of certain limitations. First, as this was a rapid review, our results may not include all published articles or preprints that meet inclusion criteria. It is also possible that information was missed since screening articles for inclusion and data extraction was not performed in duplicate. Second, this review did not include a formal quality assessment of the study design of the included studies. Third, inclusion and exclusion criteria were tailored to retrieve articles that were applicable to the Western context, and only articles written by authors in or including data from North America, Mexico, Europe and Australia were included in this review. Study eligibility criteria were further limited to attitudes and behaviors, which are more modifiable from a public health perspective. Studies that focused exclusively on the effects of personality characteristics (e.g. narcissism, impulsiveness and agreeableness), or on societal characteristics (e.g. individualism and collectivism), on uptake of public health guidelines were excluded. As such, results of this review do not speak to the effects of psychological or societal factors on adherence to COVID-19 guidelines. Furthermore, the review did not search out materials on systems factors (e.g. provision of isolation spaces) and societal factors (e.g. rates of poverty) which may have with a greater impact on public health guideline adherence than individual level factors. Although the results are preliminary, this presents the first effort to map the large volume of studies in this domain and provides direction for future empirical and knowledge synthesis efforts.

This rapid review highlights several factors that are related to attitudes toward and adherence to COVID-19 public health guidelines. The available evidence suggests individuals who are older, identify as women, trust in government, perceive COVID-19 as threatening and access information through traditional news media are more likely to report adherence to COVID-19 public health guidelines. Strategies for promoting adherence to public health guidelines have not yet been investigated thoroughly, but promising avenues for future research include promoting accurate knowledge of pandemic guidelines and highlighting the efficacy of public health guidelines to mitigate the threat posed by COVID-19. Evidence presented in this review is mostly based on cross-sectional survey research using convenience sampling, with most included studies using distinct methods to measure protective behaviors. Future research should utilize experimental designs and more robust sampling techniques to test the effects of public health interventions and messaging on attitudes and behaviors, and investigate targeted approaches for groups that are at increased risk for non-adherence to COVID-19 guidelines.

How will countrybased mitigation measures influence the course of the COVID-19 epidemic?

Reducing transmission of SARS-CoV-2

World Health Organization. Coronavirus Disease (COVID-19): Advice for the Public

Projected effects of nonpharmaceutical public health interventions to prevent resurgence of SARS-CoV-2 transmission in Canada

Narrative review of nonpharmaceutical behavioural measures for the prevention of COVID-19 (SARS-CoV-2) based on the health-EDRM framework

Projecting the transmission dynamics of SARS-CoV-2 through the postpandemic period

Developing PRISMA-RR, A Reporting Guideline for Rapid Reviews of Primary Studies (Protocol)

Evidence summaries: the evolution of a rapid review approach

COVID-19 Scientific Advisory Group Rapid Evidence Report: Attitudes and Adherence to COVID-19 Guidelines

Beliefs in COVID-19 conspiracy theories predict lower level of compliance with the preventive measures both directly and indirectly by lowering trust in government medical officials. Open Science Framework

Could attitudes toward COVID-19 in Spain render men more vulnerable than women?

Compliance in the 1.5 meter society: longitudinal analysis of citizens' adherence to COVID-19 mitigation measures in a representative sample in the Netherlands. Open Science Framework

The intelligent lockdown: compliance with COVID-19 mitigation measures in the Netherlands

Non-compliance with COVID-19-related public health measures among young adults in Switzerland: insights from a longitudinal cohort study

Scientific trust, risk assessment, and conspiracy beliefs about COVID-19 -four patterns of consensus and disagreement between scientific experts and the German public

Compliance with infection control rules among adolescents in Oslo during the COVID-19 pandemic. Tidsskr Den Nor Legeforening

Association of a public health campaign about coronavirus disease 2019 promoted by news media and a social influencer with self-reported personal hygiene and physical distancing in the Netherlands

Correlates of healthprotective behavior during the initial days of the COVID-19 outbreak in Norway

Stay-at-home orders, social distancing and trust. Global Labor Organization (GLO) 2020

Knowledge and behaviors toward COVID-19 among US residents during the early days of the pandemic: cross-sectional online questionnaire

The effectiveness of moral messages on public health behavioral intentions during the COVID-19 pandemic

Mask-wearing increased after a government recommendation: a natural experiment in the U.S. during the COVID-19 pandemic

Compliance without fear: individual-level predictors of protective behavior during the first wave of the COVID-19 pandemic

Nonpharmaceutical interventions for pandemic COVID-19: a cross-sectional investigation of US general public beliefs, attitudes, and actions

Consumers and COVID-19: survey results on mask-wearing behaviors and beliefs

Social distancing during the COVID-19 pandemic: who are the present and future noncompliers?

The causes and consequences of COVID-19 misperceptions: understanding the role of news and social media

What drives resistance to public health measures in Canada's COVID-19 pandemic? A rapid assessment of knowledge, attitudes, and practices

Health-protective behaviour, social media usage and conspiracy belief during the COVID-19 public health emergency

Coronavirus conspiracy beliefs, mistrust, and compliance with government guidelines in England

COVID-19 is rapidly changing: examining public perceptions and behaviors in response to this evolving pandemic

Information and behavioral responses during a pandemic: evidence from delays in COVID-19 death reports

Beliefs about COVID-19 in Canada, the

Novel coronavirus, old partisanship: COVID-19 attitudes and behaviours in the United States and Canada

Citizens' adherence to COVID-19 mitigation recommendations by the government: a 3-country comparative evaluation using web-based cross-sectional survey data

Moral outrage and social distancing: bad or badly informed citizens? PsyArXiv

Public perceptions and attitudes toward COVID-19 nonpharmaceutical interventions across six countries: a topic modeling analysis of twitter data

Social distancing around the globe: cultural correlates of reduced mobility

The behaviour change wheel: a new method for characterising and designing behaviour change interventions

Improving the quality of web surveys: the checklist for reporting results of internet e-surveys (CHERRIES)

Good practice in the conduct and reporting of survey research

Survey tool and guidance: rapid, simple, flexible behavioural insights on COVID-19

Risk Communication and Community Engagement (RCCE) action plan guidance COVID-19 preparedness and response

Democratic Health Communications during COVID-19: A RAPID Response. UBC Centre for the Study of Democratic Institutions

Applying principles of behaviour change to reduce SARS-CoV-2 transmission

We gratefully acknowledge Rachel Zhao for designing and running the search, as well as Alexandra Bennett, Carla Vetland, Kristal Turner and Armghan Ahmad for their assistance with data extraction. We would also like to thank the committee members who approved the Alberta Health Services COVID- 19 

Supplementary data are available at the Journal of Public Health online.

This study was funded through in-kind support from Alberta Health Services. CM was supported by doctoral awards from Vanier Canada, Killam Trusts, and Alberta Innovates and a Training in Research and Clinical Trials in Integrative Oncology (TRACTION) fellowship from the University of Calgary.

CM, PM and DC drafted initial manuscript and tables. PM, DC, PR, LB, ZC and MS contributed to study design. LB, ZC and MS completed data extraction. PM, DC, DR, PR, TC and CM contributed to interpretation of results. All authors reviewed and provided feedback on the final draft of the manuscript.

None to declare.