key: cord-254944-9f3wkvxp authors: Pollak, Y.; Shoham, R.; Dayan, H.; Gabrieli Seri, O.; Berger, I. title: Background and concurrent factors predicting non-adherence to public health preventive measures during the chronic phase of the COVID-19 pandemic date: 2020-10-14 journal: nan DOI: 10.1101/2020.10.13.20211904 sha: doc_id: 254944 cord_uid: 9f3wkvxp To determine factors that predict non-adherence to preventive measures for COVID-19 during the chronic phase of the pandemic, a cross-sectional, general population survey was conducted in Israel. Sociodemographic, health-related, behavioral, and COVID-19-related characteristics were collected. Among 2055 participants, non-adherence was associated with male gender, young age, bachelorhood, being employed, lower decrease in income, low physical activity, psychological distress, ADHD symptoms, past risk-taking and anti-social behavior, low pro-sociality, perceived social norms favoring non-adherence, low perceived risk of COVID-19, low perceived efficacy of the preventive measures, and high perceived costs of adherence to the preventive measures. There appears to be a need for setting out and communicating preventive measures to specifically targeted at-risk populations. Novel Coronavirus 2019 (COVID- 19) outbreak has an enormous impact on public health and global economy. In response to the growing pandemic, most states took preventive measures to limit the spread of cases through community transmission of COVID-19. These measures included isolation of infected and suspected patients, use of personal protective equipment (face masks, gloves etc.), personal hygiene, restrictions on gathering and traveling, social distancing, as well as mandatory quarantine and lockdown (1) . Some of the preventive measures depend on how people cooperate, i.e. whether they adhere to the measures or not (2) . Despite the potentially harmful consequences for individuals and public health, non-adherence to the preventive measures (non-APM) for the COVID-19 pandemic, mainly at the acute phase, has been reported around the world (3) (4) (5) (6) (7) . For designing effective public health policy, it is mandatory to identify the factors that predict non-APM. Recently, several cross-sectional surveys were conducted, in an effort to identify predictors of APM during the early phases of the pandemic. For instance, immediately after the detection of the first COVID-19 patient in Hong-Kong, higher levels of adoption of social-distancing measures were associated with being female, living in the geographic regions in Hong Kong that share the border with mainland China, perceiving oneself as having a good understanding of COVID-19, and being more anxious (7) . In a survey conducted in Israel in April 2020, male gender, not having children, high levels of ADHD symptoms, smoking, past risk-taking behavior, and current psychological distress levels, all predicted non-APM. On the other hand, pro-sociality, understanding of the instructions, high perceived risk of COVID-19, and high perceived efficacy of the preventive measures predicted adherence (5) . As the COVID-19 pandemic continues, preventive measures become a constant part of our life. The objective of this study was to identify predictors of non-APM at the chronic phase of the pandemic. The literature suggests different conceptualizations of non-APM. Firstly, preventive measures are prescribed by health agencies and therefore can be considered medical instructions and healthy lifestyle. Secondly, non-APM might endanger the non-adherent and his/her vicinity, and consequently should be considered as a risk-taking behavior. Finally, preventive measures are often set as laws or regulations, implying that non-APM often means illicit behavior. Potential predictors of non-APM were chosen for this study based on the literatures regarding the risk factors for nonadherence to medical instructions (8) , engagement in risk-taking behavior (9, 10) , and engagement in anti-social behavior (11) . These included four groups of variables: 1. For the primary outcome measure, non-APM, respondents were asked to rate the extent of which they adhered to each of the 13 preventive measures that were released by the Israeli Ministry of Health at the corresponding period (e.g. social distancing, personal hygiene, facemask). A fivepoint Likert scale was used: 1="Not at all", 2="Somewhat", 3="Moderately", 4="Strictly", and 5="Very strictly". Individual mean response scores were calculated. The following scales were used to measure the independent variables: 1. Sociodemographic factors: Respondents completed a questionnaire consisting of items regarding age, gender, marital status, number of children, ethnicity, religious affiliation and level of observance, type of education, place of living (country region, and type of community), and background migration. In addition, respondent reported on pre-treatment level of income (much above average, above average, average, below average, much below average), level of decrease in income since the onset of the coronavirus outbreak (on a 1 = not at all to 5 = extreme decrease Likert scale), and pre-treatment and current percent of position. Respondents were asked to report on their average number of daily hours of sleep, frequency of engaging in intensive physical activity, and smoking habits. They also reported whether they are chronically treated or followed up for any of the following reasons (that are All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20211904 doi: medRxiv preprint considered risk factors for COVID-19 (12) ): heart disease, lung disease, liver disease, AIDS, cancer, organ transplantation, diabetes, dialysis, steroid treatment, and prophylactic antibiotic treatment. Subjective health was probed by a single-item self-rated health (SRH) Likert scale describing their own health impression, ranging from 1 = poor to 10= excellent. This scale was found to reflect individuals" perceptions of their physical health and psychological well-being (13) . An adapted version of the Kessler Screening Scale for Psychological Distress (K6) was used to probe for non-specific psychological distress (14) . Difficulties Questionnaire (SDQ) (18, 19) was used for measuring pro-sociality. Respondents rated the extent to which a series of six attributes described them during six months reference period on a three-level response scale (0 = "not true", 1 = "somewhat true", or 2 = "certainly true"). A short form of the Adult Risk-Taking Inventory (ARTI) (20, 21) was used to measure past engagement in risky behavior. The short form consists of 14 items probing for the frequency of engagement in relatively frequent activities (e.g., sunbathing without sunscreen, smoking marijuana) with respect to their frequency during the preceding year on a rating scale, ranging from 1 (Not at all) to 7 (On a daily basis). Previous work has shown that the ARTI has good reliability and validity. Past anti-social behavior was assessed using 15-item 4-point frequency scale, ranging from 1 = Not at all to 4 = More than 5 times, adapted from Cho et al. (22) . For each of the three scales, the average continuous scores were converted to categorical scores by grouping values into four groups with quartiles as cutoff points. 4. COVID-19-related perception factors: Perceptions regarding the COVID-19 and the preventive measures were assessed using several five-point Likert scales: Perceived risk of COVID-19 was assessed by a nine-item self-report questionnaire that was designed for this study based on the risk perception literature (23) . For example, "How likely are you to get COVID-19?". In this sample, the scale had good internal consistency (Cronbach"s α = .80). Perceived efficacy of the prevent measure scale was measured by a self-report questionnaire designed for this study. The scale consists of five items was composed for measuring participants" perceived efficacy of the instructions. For example, "To what extent you think that adhering to the preventive measures will reduce the chances that you or your loved ones will get COVI-19?". In this sample, the scale had good internal consistency (Cronbach"s α = .83). Another scale composed for this study, consisted of seven items probing for the perceived costs of APM, including the perceived cost of APM on different domains of wellbeing (e.g., economic, social, spiritual). For example, "To what extent adhering to the preventive measures will impair your interpersonal relationship?". In this sample, the scale had good internal consistency (Cronbach"s α = .84). Perceived norms regarding APM were measured by four questions regarding the descriptive (i.e., the prevalence of non-APM) and the injunctive (i.e., the tolerance toward non-APM) norms of the family/friends and the community/workplace they are embedded in. Analytic approach First, unadjusted logistic regression analyses were used to calculate the associations between each of the independent variables and the primary outcome. Next, four adjusted models were examined using backward stepwise logistic regressions with probability of 0.05 for entry and 0.1 for removal. Multicolinearity was examined through Spearman's rank correlation analysis. In the first model, only the sociodemographic variables were included. In the second model, health-related variables were entered in a second block. Similarly, in the third and fourth models, the second block consisted of the behavioral and personality, and the COVID-19 related perceptions variables, respectively. P values were not corrected, a p-value < 0.05 was considered statistically significant. Table 1 summarizes the sociodemographic, health-related, behavioral, and COVID-19 perceptions related variables. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20211904 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Predictors of non-APM All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20211904 doi: medRxiv preprint Spearman's correlation coefficients were examined for all independent variables. All correlations were < 0.7. The correlations among age, marital status, and having children were in the 0.53-0.69 range. Table 2 presents the unadjusted and adjusted regression analysis results for non-APM. The following variables were found to predict non-APM after adjustment for sociodemographic variables. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Below cutoff Above cutoff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 14, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020. 10.13.20211904 doi: medRxiv preprint Sociodemographic variables: The following variables predicted non-APM on adjusted analyses: male gender, younger age (<=64), and being single. Being employed predicted more non-APM; conversely, decrease in income predicted more adherence. Health-related variables: Less regular physical activity and higher levels of ADHD symptoms and of psychological distress predicted non-APM. Behavioral and personality factors: High levels of past risk-taking behavior and anti-social behavior, as well as low levels of pro-sociality, predicted non-APM. COVID-19 perception variables: Non-APM was predicted by higher perceived non-adherence norms, lower perceived risk of COVID-19, lower perceived efficacy of the preventive measures, as well as higher perceived costs of adherence. This study aimed to identify risk factors for non-APM during the chronic phase of the COVID-19 outbreak. Several factors were found to predict non-adherence. Sociodemographic factors included male gender, young age, bachelorhood, being employed, and smaller decrease in income. Healthrelated factors included physical activity, psychological distress, and ADHD symptoms. Behavioral and personality factors included history of risk-taking and anti-social behavior, and low prosociality. Finally, COVID-19 perception factors included perceived social norms favoring nonadherence, low perceived risk of COVID-19, lower perceived efficacy of the preventive measures, and higher perceived costs of adherence to the preventive measures. Notably, the greatest predictors in terms of OR were lower age, past anti-social behavior, low perceived risk of COVID-19, the efficacy of the preventive measures, and the norms of adhering to the preventive measures. The variables that predicted non-APM at the chronic phase of the outbreak in Israel were similar to those that predicted non-APM during the first wave in Israel (5), suggesting that similar motivations drive the decision whether to adhere to preventive measures or not. Many of the non-APM predictors that were found in this study have also been reported by studies conducted in other states. For instance, male gender and young age were linked to non-adherence in the US, Somalia, Saudi Arabia, and Hong Kong during the COVID-19 outbreak (3, 4, 7, 24, 25) . The negative correlation between employment and adherence in the current study parallels the findings of Porten et al. during the SARS outbreak in Germany (26) . Several studies highlighted the association between All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020. 10.13.20211904 doi: medRxiv preprint adherence and perceptions about the infection and the preventive measures in a variety of states during the current pandemic (27) (28) (29) . The negative correlation between adherence and the perceived costs of adherence resembles the findings of DiGiovanni et al. (30) reporting that perceived economic costs of the quarantine in Canada during the 2004 SARS outbreak were related to nonadherence. The role of social norms has been demonstrated in a study concerning quarantine in Senegal during Ebola outbreak (31) and in Australia during H1N1 outbreak (32) . The negative correlation between adherence and past risk-taking behavior and unhealthy lifestyle is in line with a study reporting that among young adults with hazardous drinking, adherence to public policies is suboptimal (33) . Our study adds new predictors of non-adherence including ADHD symptoms, general risk-taking behavior, previous engagement in crime, as well as low pro-sociality, which contributed for better prediction of non-APM. Many of the above listed factors have been shown to predict non-adherence to medical treatment (8) , risk-taking behavior (9, 10) , and anti-social behavior (11) . Accordingly, adherence to preventive measures may be analyzed in all the corresponding theoretical frameworks. Notably, having medical risk factors for COVID-19 (i.e., background diseases) did not predict higher adherence to preventive measures. A similar independency between objective risk and adherence was found in a study reporting no effect of the total probable cases of SARS on likelihood of adherence (34) . In deriving implication for public health, it is important to differentiate between predictors that preceded the COVID-19 outbreak, and therefore can be considered risk factors for non-APM, and other variables that coincided with the outbreak and hence their causal relations with non-APM cannot be determined based on a cross-sectional study. The latter include the economic consequence of COVID-19, as well as the perceptions regarding the pandemic and the preventive measures. Nevertheless, these coinciding predictors may still be used for targeting populations atrisk to non-APM. The current findings of observable risk factors for non-APM suggest that the nature and the communication of the preventive measures should be targeted for different people. Policymakers may develop specific plans for populations at risk of non-adherence, focusing on messaging, fostering, and enforcing preventive measures, as well as on increased monitoring of infection rate. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020. 10.13.20211904 doi: medRxiv preprint Further research is warranted for identifying other risk factors for non-APM across longer periods and changing contexts and for examining the efficacy of public health policy in promoting APM. Supplementary material is enclosed. A Review of Current Interventions for COVID-19 Prevention How to improve adherence with quarantine: rapid review of the evidence. Public Health Attitude and Practice Toward COVID-19 Among the Public in the Kingdom of Saudi Arabia: A Cross-Sectional Study Social distancing in response to the novel coronavirus (COVID-19) in the United States Predictors of non-adherence to public health instructions during the COVID-19 pandemic Poor self-reported adherence to COVID-19-related quarantine/isolation requests Community Responses during Early Phase of COVID-19 Epidemic, Hong Kong. Emerg Infect Dis Adherence to long-term therapies: evidence for action . World Health Organization Risk-Taking Behavior in Attention Deficit/Hyperactivity Disorder (ADHD): a Review of Potential Underlying Mechanisms and of Interventions Who takes risks when and why? Determinants of risk taking Theoretical Foundations: Delinquency Risk Factors and Services Aimed at Reducing Ongoing Offending Covid-19: risk factors for severe disease and death Self-rated health Short screening scales to monitor population prevalences and trends in non-specific psychological distress Screening for serious mental illness in the general population with the K6 screening scale: results from the WHO World Mental Health (WMH) survey initiative Diagnosing ADHD in Israeli adults: the psychometric properties of the adult ADHD Self Report Scale (ASRS) in Hebrew The World Health Organization Adult ADHD Self-Report Scale (ASRS): a short screening scale for use in the general population The young adult Strengths and Difficulties Questionnaire (SDQ) in routine clinical practice Comparing the Strengths and Difficulties Questionnaire and the Child Behavior Checklist: is small beautiful? ADHD Is Associated With a Widespread Pattern of Risky Behavior Across Activity Domains What Drives Risky Behavior in ADHD: Insensitivity to its Risk or Fascination with its Potential Benefits? Differential Item Functioning on Antisocial Behavior Scale Items for Adolescents and Young Adults from Single-Parent and Two-Parent Families Factors in risk perception What Protective Health Measures Are Americans Taking in Response to COVID-19? Results from the COVID Impact Survey COVID-19 in Somalia: Adherence to Preventive Measures and Evolution of the Disease Burden SARS outbreak in Germany 2003: workload of local health departments and their compliance in quarantine measures--implications for outbreak modeling and surge capacity? J Public Health Manag Pract The Role of Illness Perceptions, Coping, and Self-Efficacy on Adherence to Precautionary Measures for COVID-19 Preventive behavior of Vietnamese people in response to the COVID-19 pandemic Factors associated with adherence to self-isolation and lockdown measures in the UK: a cross-sectional survey. Public Health Factors influencing compliance with quarantine in Toronto during the 2003 SARS outbreak Accepted monitoring or endured quarantine? Ebola contacts' perceptions in Senegal Understanding the school community's response to school closures during the H1N1 2009 influenza pandemic In-Person Contacts and Their Relationship With Alcohol Consumption Among Young Adults With Hazardous Drinking During a Pandemic Confidence in controlling a SARS outbreak: experiences of public health nurses in managing home quarantine measures in Taiwan Supplementary materials Figure 1: The distribution of the adherence to preventive measures scores (N = 2055) All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted This research was supported by the Ministry of Science & Technology, Israel.