Compliance with the main preventive measures of COVID‐19 in Spain: The role of knowledge, attitudes, practices, and risk perception

Abstract In epidemics such as COVID‐19, major changes need to be made to the population's behavior to prevent infection and stop disease transmission. The three most recommended preventive measures are wearing a mask, washing hands with soap or hydroalcoholic gel, and watching an interpersonal distance of at least two meters (3W) from other people. This study aimed to assess adherence to these COVID‐19‐related three preventive measures and its association with knowledge, attitudes, risk perception, and practices in Spain. The COSMO‐Spain survey, based on the WHO Behavioral Insights questionnaire on COVID‐19, was conducted in the general Spanish population using an online questionnaire (n = 1,033). Sociodemographic, knowledge, attitudes, practices, and risk perception variables were included. A multivariable logistic regression model was carried out to evaluate the factors associated with compliance with the three preventive measures. Half of the respondents (49.8%) were women with a median age of 45 (Inter‐quartile Range, IR = 21) years. In the logistic regression, the factors associated with 3W compliance were being over 45 years; knowing about how COVID‐19 spreads and wearing masks properly; appropriate attitudes towards COVID‐19 (greater agreement with mandatory mask use); high risk perception (feeling that the coronavirus is spreading rapidly, being concerned about non‐mask wearers), and adherence to other preventive measures against COVID‐19, such as staying at home. Adequate knowledge, attitudes and risk perception are determinants of 3W compliance. Developing effective health education programs and frequent communication strategies are necessary, particularly for those who adhere less to preventive measures.


INTRODUCTION
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the cause of the coronavirus disease 2019 (COVID-19), a highly contagious disease that spreads rapidly through human-to-human transmission . The World Health Organization (WHO) declared the COVID-19 a global pandemic on 1 March 2020 (World Health Organization, 2020).
SARS-CoV-2 is transmitted human to human mainly by droplets inhalation in a direct contact with a case, and by indirect contact (airborne and contaminated surfaces or objects) .
Human-to-human transmission occurs through contact with infected people, through secretions such as infected saliva and respiratory secretions or their respiratory droplets, which are expelled when an infected person coughs, sneezes, talks, or sings when a person is in close contact (within 2 m). Indirect contact can occur if a person touches a surface contaminated with SARS-CoV-2, and then hands come into direct contact with mucous membranes such as the eyes, nose, or mouth (WHO, 2020).
In the absence of a treatment and until a higher vaccination rate is reached, the fight against COVID-19 has been based on behavioural preventive measures. The main individual preventive measures recommended were wearing a mask (regularly), washing your hands (with soap and water for 20 s or with an alcohol-based lotion) and watching your distance (keeping at least a 2-m distance from others), such measures being commonly known as the 3W (Centers for Disease Control and Prevention (CDC), 2021). Recently, ventilation of closed places was added (European Centre for Disease Prevention & Controls, 2020). While handwashing has always been one of the key cornerstones of infectious diseases prevention, the use of face masks has recently shown a significant reduction in infections caused by COVID-19 (Gallaway et al., 2020;Syed et al., 2003;Y. Wang, Ferro, et al., 2020). Also, studies have found that for every meter of distance that is increased, there is a significant reduction in the transmission of SARS-CoV-2 (Chu et al., 2020). Other non-pharmaceutical interventions such as movement restrictions, isolation of people arriving from affected areas or countries, restrictions on mass gathering, quarantine of confirmed cases, close contacts tracing, and testing were also implemented to reduce the spread of COVID-19 (Ayouni et al., 2021;Davies et al., 2020;Van Scoy et al., 2021;Yu, 2020). Since the COVID-19 pandemic began, many countries have conducted population behavioural studies to provide scientific evidence to promote prevention (Hosen et al., 2021;Van Nhu et al., 2020). Behavioural insight research is an important tool for understanding behavioural choices, barriers, and drivers within a population (European Centre for Disease Prevention & Control, 2021). Assessment of the adoption of preventive practices against COVID-19 among population is crucial. Studies about previous outbreaks suggest that knowledge, attitudes, and practices (KAP) survey is the best tool for assessing the population's acceptance and adherence to preventive measures (Ajilore et al., 2017;Person et al., 2004), identifying factors associated with these practices that could help to draw up effective prevention strategies and policies (Chakrawarty et al., 2020).
Population behaviour changes depend on the population's knowledge about the disease and individuals' ability to perceive risks associated with the virus and to adapt their behaviour accordingly (Wise et al., 2020). Previous studies suggested that self-efficacy in preventing disease and trust in health authorities are also significantly associated with the practice of preventive measures . Moreover, peoplet's beliefs about recommended preventive measures and reliance on relevant sources for information about COVID-19 play an important role in adherence to protective practices (Norman et al., 2020;Ozdemir et al., 2020;Rui et al., 2021;Van Scoy et al., 2021). The study aimed to assess, for first time in Spain, compliance of the 3W preventive measures in the general population and its association with knowledge, attitudes, risk perception, and practices to help understand prevention-measure adherence and promote the design of evidence-based information campaigns.

Survey questionnaire
The survey, based on a WHO Behavioural Insights questionnaire and the COVID-19 Snapshot MOnitoring (COSMO) (Betsch et al., 2020), collected both quantitative and qualitative data to examine citizens' knowledge about the disease and risk perceptions, trust in health authorities and recommendations, adherence to recommended preventive measures, and trust in the sources of information and frequency of consultation.

Variables
An online questionnaire was prepared to collect information on basic The population's level of concern about the coronavirus/COVID-19 outbreak in general and in specific situations was queried using the question 'At the moment, how much do you worry about. . . ?' , and the answers were rated from 1 (do not worry at all) to 5 (worry a lot). Concerns were measured with the questions 'I have a feeling that the coronavirus/COVID-19 is spreading. . . ' using a scale from 1 (very slow) to 5 (very fast) and 'Do you believe that the worst of the epidemic. . . ?' using the answers 'has passed' , 'is happening now' , and 'has not arrived yet' .
Behaviour towards source of information and trust was measured on a scale from 1 (never) to 5 (very often) with the questions 'How often do you use the following sources to stay informed about COVID-19?' and trust in sources of information from 1 (none) to 5 (much) with the question 'How much do you trust the following sources of information in their reporting about COVID-19?' .

Data analyses
A descriptive analysis of participants' characteristics was performed, the categorical variables being described as frequency and percentages with 95% confidence intervals (CIs). The continuous variable was presented using median and inter-quartile range (IR).
First, a bivariate analysis of the associations between each of the 3W preventive measures with the KAP and risk perception variables was performed. Chi-squared test was used to study any statistically significant relationship between each of the 3W preventive measures and the KAP or risk perception variables, and p-values <.05 were considered significant.
Multivariable logistic regressions analyses were then performed to assess how COVID-19-related factors such as KAP, risk perception, and socio-demographic characteristics of study participants were associated with adherence to the 3W preventive measures together. The dependent variable was created including only those who answered that they had followed the three preventive measures (wearing facemasks, washing their hands, and ensuring physical distancing) during the last 7 days to avoid becoming infected with COVID-19. Independent variables were selected through forward stepwise procedure including only those significant at 5% level in a previous chi-squared test. The odds ratio (OR) and 95% CI were computed, and p-values <.05 were considered statistically significant. All statistical analyses were performed using IBM SPSS Statistics version 27.

Sample characteristics
The sample was composed of 1033 respondents, 49.8% of whom were women. The mean age was 45.7 (SD: 14.6) years, and 44.9% had attained vocational training, college degree, or higher qualification. The proportion of participants in work was 56.5%. Due to COVID-19, 5.4% of respondents were unemployed or had been laid off temporarily.
Around a third of the sample (33.8%) lived with someone over 60 years old. Most respondents (86.7%) lived in an urban area. Perceived health before COVID-19 was rated as good/very good by 59.9% participants.
A total of 34.2% participants had relatives infected with COVID-19 and 14.8% had family members who had died from COVID-19 (Table 1).

Disease knowledge and preventive measures
Implementing the 3W preventive measures at the same time was significantly associated with knowledge of how COVID-19 spreads: droplets released when coughing/talking, physical contact with someone infected and with contaminated surfaces (p <.001). 3W adherence was also associated with knowing that wearing a face mask is useful to avoid infecting others that a mask must cover your nose and mouth (p <.001) and other correct face mask usage (see Table 2). Table 3 shows how most people who followed the 3W preventive measures agreed with the mandatory use of masks (p <.001). In addition, people who trusted how the epidemic was being managed, mainly by scientists (p ≤.050) and health centres (p <.001), also complied with the 3W.

Risk perception, concerns, and the 3W preventive measures
Compliance with the 3W preventive measures was significantly associated with the concern about losing a loved one, overstretched health services, people who did not wear a facemask, and about missing their holidays (p <.001) ( Table 4). In addition, adherence with the 3W preventive measures was associated with the feeling that COVID-19 was spreading very fast (p <.001) and believing that the worst of the pandemic was not over yet (p <.050). Regarding self-efficacy, thinking that it was easy/very easy to avoid being infected was also associated with compliance with the 3W preventive measures (p <.001).

Sources and confidence in the information
Most of respondents were very confident in the information provided by their health care professionals (717, 69.4%), and this confidence was significantly associated with the implementation of the 3W preventive measures. Figure 1 shows the frequency and confidence in other sources of information. Checking TV news frequently was significantly associated (p <.001) with following each of the 3W preventive measures, while checking the Spanish Ministry of Health's website was significantly associated with washing hands 92.5% (p = .017) and watching distance 87% (p = .048). Trust in the information delivered by social media (e.g. Facebook, Twitter, YouTube, and WhatsApp) was also associated with adherence to the 3W (p < .001).

Preventive practices
Implementing one of the 3W preventive measures was significantly associated with implementing either of the other two. In addition to high compliance with the 3W preventive measures, 86% of the interviewees said they had used hydroalcoholic gel and 64% had avoided using public transport. Also, 57% of the participants had avoided meeting relatives or friends, and 55% had disinfected surfaces to avoid COVID-19 infection (see Table 5).

Factors associated with adherence to all the 3W preventive measures regarding COVID-19
In the multiple logistic regression analysis, knowing at least one form of transmission (OR = 2.46) that masks are used to avoid infecting others (OR = 2.38) and their correct use (OR = 1.58) was significantly associated with adherence to the 3W preventive measures. Age was the only socio-demographic characteristic associated with adherence to the 3W measures (OR = 1.76).
Moreover, those who agreed with mandatory use of masks and who were worried about 'people who do not wear a mask' were almost twice as likely to comply with the 3W preventive measures. Feeling that coronavirus/COVID-19 was spreading fast was also associated with the 3W preventive measures adherence.
By comparison, people who did not trust in long-distance trains and buses had 1.60 probability to follow the 3W preventive measures. Finally, people who trusted the coronavirus/COVID-19 information given by healthcare professionals were almost twice as likely to comply with the 3W preventive measures, while people who trusted social media (e.g. Facebook, Twitter, YouTube, and WhatsApp) were 0.60 times likely to adhere to them (Table 6).

DISCUSSION
The adoption of behavioural preventive measures against  has been critical in controlling the pandemic. According to our study,  (Lennon et al., 2020;MacIntyre et al., 2021;Mouchtouri et al., 2020;Norman et al., 2020), and in line with findings from previous pandemics . Knowledge is a prerequisite for forming positive attitudes and  promoting positive behaviours in high transmissible pandemics such as COVID-19 (Norman et al., 2020). Several studies stated how good knowledge in the community about COVID-19 multiplied adherence to preventive practices against the disease, such as wearing a face mask, washing your hands with soap and water regularly, using a personal alcohol-based sanitizer, avoiding crowded areas and closed spaces, or not travelling out of state (Hatabu et al., 2020;Khasawneh et al., 2020;Nwonwu et al., 2020). In Spain, compliance with the all three preventive measures was associated with knowing the most mentioned COVID-19 modes of transmission and how to wear a mask properly.
These variables were also associated with greater compliance in other studies (Ferdous et al., 2020;Mouchtouri et al., 2020).
Attitudes such as agreeing with the mandatory use of masks in Spain, and being worried about people not wearing them also deter- transmission chain (Beesoon et al., 2020). In a study conducted in different English-speaking countries, almost all participants believed that when wearing a mask, they also needed to wash their hands and adhere to physical distancing (MacIntyre et al., 2021). Public health messages should emphasize the combination of the 3W instead of focusing on only one of them such as facemasks that can produce a false sense of protection.
Risk perceptions also influence individuals' protective behaviour.
Complying with the 3W preventive measures was associated with the feeling that coronavirus/COVID-19 was spreading fast, as well as with leaving the house as little as possible and telework. Risk perception was high in the population. Even so, at the time of the study, Spain had a relatively low accumulated incidence (ISCIII, 2020). Previous studies have shown similar results about how peoplet's concerns and riskperceptions contributed to preventive practices (Banerjee et al., 2021;Mouchtouri et al., 2020;Nazione et al., 2021;Ozdemir et al., 2020;Qian et al., 2020;Rui et al., 2021;Van Scoy et al., 2021), suggesting that evaluating these risk perceptions is critical for an effective and appropriate crisis response (Ajilore et al., 2017). On the contrary, preventive behaviour, including hand-washing and social distancing, decreased in people less concerned about COVID-19 or perceiving the risk to be exaggerated or disproportionate to the threat (Kantor & Kantor, 2020;Ozdemir et al., 2020).
Individuals with accurate health information are usually more motivated to engage in preventive health behaviour Qian et al., 2020). However, the volume of rumours about COVID-19 makes individuals vulnerable to misinformation (Cuan-Baltazar et al., 2020). One factor that could potentially affect information credibility is its source. In Spain, people who were more confident about coronavirus/COVID-19 information from healthcare professionals were more likely to adhere to the 3W preventive measures in line with studies where reliance on expert sources encouraged protective behaviour (Ali et al., 2020;Rui et al., 2021;Sakya et al., 2021) (Mboya et al., 2020). Local-level coordination of contextually appropriate strategies is likely to be the cornerstone of the national response (World Health Organization, 2020). Coordination of regional and district/council stakeholders such as the regional and district commissioners, local health management teams, religious leaders, political leaders, healthcare workers, the media, and non-governmental organizations is critical to improve the implementation of and compliance with preventive strategies (Betsch et al., 2020). Coordinated efforts among stakeholders in the country are necessary to contain COVID-19 through context-specific prevention strategies. Such interventions should also focus on ensuring compliance with the national and international recommended preventive measures. However, due to the evolving nature of this pandemic, it is clear that messaging needs to be dynamic and continuously updated in line with evidence gathered and changing recommendations (Mboya et al., 2020).
There is a complex interplay of changes in epidemiology data, media attention, pandemic control measures, risk perception, and public health behaviour (Reintjes et al., 2016). An effective risk communication in a pandemic such as COVID-19 must include information not just about the threat itself but also about how people perceive and respond to that threat (Abrams & Greenhawt, 2020;Varghese et al., 2021). Data gathered in our study could support effective interaction among authorities, health workers, journalists, and the public to design effective risk communication that encourages appropriate behavioural change, helps manage the crisis, and protects the most important asset in a crisis: public trust (Betsch et al., 2020).
The main limitation of this study is that it is based on an online-based survey, thereby restricted to only those with Internet access. However, the sample matched the Spanish general population in terms of age, education, sex, and area of residence. This is a cross-sectional study, and as the COVID-19 situation is changing rapidly, the results reported here represent the situation during the surveyed period.

Conclusions and recommendations
Our study provides information that could help authorities to monitor and improve adherence to the preventive measures in force.
Reinforcing the population's knowledge, positive attitudes, adequate practices, and risk-perception during epidemics such as COVID-19 is crucial to controlling the disease. Authorities have to develop effective, evidence-based health education programmes and risk communication campaigns to promote a change in behaviours aimed primarily at those who adhere less to recommended preventive measures. According to our study, it would be particularly important to target the subset of individuals who remain disengaged like young people, with low selfefficacy and low risk perception and who do not look for information early on the pandemic. Public health messages should emphasize the combination of the 3W, instead of focusing on only one of them such as facemasks and would reach more credibility if healthcare professionals transmit these messages. However, messaging needs to be dynamic and continuously updated in line with scientific evidence gathered and changing recommendations.

ACKNOWLEDGMENTS
The authors would like to thank the study participants for having understood the interest of this study and volunteering to answer the survey.

CONFLICT OF INTEREST
The authors declare no conflict of interest.

ETHICS STATEMENT
The authors confirm that the ethical policies of the journal, as noted on the journal's author guidelines page, have been adhered to. This study was approved by Ethics Committee of the Spanish National Health Institute, Carlos III (CEI PI 59_2020-v2).

DATA AVAILABILITY STATEMENT
All necessary data are included in the manuscript and can be obtained from the authors upon request.