Risk perception, sociodemographic factors, and knowledge of COVID-19 associated with mask use outside the home in Indonesia during the mandatory policy

Background

The COVID-19 pandemic, which began in late 2019, was marked by an incidence of symptoms such as severe tightness in the chest, similar to those observed in Middle East Respiratory Syndrome (MERS). The initial laboratory examinations of the coronavirus occurred in Wuhan City, China, leading to its identification as SARS-CoV-2. The World Health Organization (WHO) reported the outbreak and officially labeled the disease as COVID-19. By March 2020, due to its rapid global spread, the WHO declared COVID-19 a pandemic (World Health Organization (WHO), 2020a).

As of September 17, 2020, the pandemic had affected 213 countries, with approximately 30,340,825 reported cases worldwide. Of these, 22,031,183 individuals had recovered, while about 950,000 had succumbed to the virus, resulting in a relative mortality rate of 0.3%. By July 21, 2023, global cases had surged to 768,237,788, with Europe, the Western Pacific, and the Americas reporting the highest numbers. Total deaths reached around 6.9 million, with a relative mortality rate of 1.1%. In Indonesia there were 6,812,127 cases, which ranked 20th worldwide with 161,879 deaths in the year 2024 (COVID-19 Statistic Team, 2024).

COVID-19 is primarily transmitted through respiratory droplets, with an incubation period of 1 to 2 weeks. The Indonesian Ministry of Health has issued guidelines emphasizing preventive measures such as mask-wearing, regular hand hygiene with soap, maintaining physical distance, and avoiding crowded places (The Ministry of Health of The Republic of Indonesia, 2020; World Health Organization (WHO), 2021). Recommendations also included minimizing contact with the face, mouth, nose, or eyes in order to prevent droplet transmission, and disinfecting surfaces potentially exposed to respiratory droplets (World Health Organization (WHO), 2021).

Numerous studies have established the effectiveness of mask-wearing in preventing COVID-19 transmission (Wang et al., 2021; Hajmohammadi, Saki Malehi, and Maraghi, 2023). For instance, compliance was notably high in East Asian countries like Japan and Korea, with Hong Kong reporting a compliance rate of 96.6% (Fischer et al., 2021). This high adherence was linked to a marked reduction in COVID-19 cases compared to regions with lower compliance (Cheng et al., 2020). Moreover, a study in Singapore found a mask compliance rate of 84.5%, where factors such as gender and crowd density influenced behaviour (Xiang Ong et al., 2023). In contrast, European countries displayed more varied compliance rates, which did not consistently correlate with health outcomes (Spira, 2022). Germany’s mandatory mask policy improved compliance (Betsch et al., 2020), whereas other countries like France and Spain showed significant variation.

In Eastern Europe, mask use was more evenly distributed, but higher compliance did not necessarily lead to better health outcomes. Studies in Poland highlighted that despite regulations, compliance remained influenced by cultural and social factors (Matulewska et al., 2022). In the United States, compliance rates varied widely, with some regions experiencing low adherence alongside higher transmission rates.

Risk perception significantly influences mask usage in combating COVID-19 (World Health Organization (WHO), 2019, UNICEF, 2020). The Health Belief Model suggests that individual perceptions of risk, along with sociodemographic factors and knowledge, play crucial roles in determining mask use (Roth et al., 2024; Li et al., 2022; White et al., 2022).

During the COVID-19 pandemic, Indonesia implemented a mandatory mask policy for all residents (The Ministry of Health of The Republic of Indonesia, 2020), which led to increased compliance (Kar et al., 2023). However, following the WHO’s declaration of COVID-19 as endemic in early May 2023, and similarly in Indonesia, the strict enforcement of this policy was relaxed, even as transmission persisted. Given this context, this study aims to investigate the association between risk perception, sociodemographic factors, and knowledge of COVID-19 with mask-wearing behavior outside the home in Indonesia. By understanding these dynamics, this research aims to contribute valuable insights for future public health strategies.

Methods

The study was observational and employed a cross-sectional design. Data collection was performed online using Google Forms. The recruitment targeted individuals with various educational and occupational backgrounds through online community groups. The questionnaire was distributed to the participants via WhatsApp (WA) to the following groups: 1) government offices, 2) private offices, 3) companies, 4) self-employed individuals, 5) professionals, 6) neighborhood association (RT), 7) university students, 8) other students, and 9) informal groups with their various ages and genders https://hdl.handle.net/20.500.12690/RIN/87OSKQ (Roosihermiatie 2025). The informal groups included fishermen, farmers, traditional market sellers, group of neighborhood associations, including private sector workers. This approach aimed to assess mask use within the broader community during the COVID-19 pandemic. Children were excluded because their preventive behavior, such as mask-wearing, is primarily influenced by their parents. Children were excluded because their preventive behavior, such as mask-wearing, is primarily influenced by their parents.

The sample size was calculated based on the estimated population proportion with relative precision of confidence interval (α) = 0.05, proportion (P) = ranges 0.25–0.50, and relative precision (ε) = 0.1. The sample size calculation was:

The sample size was 1153 people (Lwanga et al., 1991).

Data collection took place during the period from November 2020 to February 2021. The Google Form link was shared with contact persons of the above groups, who then distributed it to their WhatsApp (WA) groups. Prior to completing the questionnaire, participants were presented with a written informed consent form outlining the study’s objectives: “This online study aims to identify factors associated with preventive behaviors against COVID-19 within the community as part of efforts to mitigate the virus’s spread. I acknowledge that I have read and understood the purpose of the study. I voluntarily consent to participate without coercion. I understand I may withdraw at any time without penalties.” Participants who agreed selected the “Agree” button and were directed to the questionnaire. Those who declined selected “Reject” and were exited from the platform. A total of 1,317 people accessed the questionnaire. Of these individuals, 22 declined to participate and 142 provided incomplete responses. As a result, 164 respondents (12.4%) were excluded, and 1,153 participants were included in the study.

The study applied the Health Belief Model, comprising risk perception and the modifying factors of sociodemographic factors and knowledge of COVID-19, as determinants of mask use. The dependent variable was consistency of wearing mask outside the home, referring strictly to usage in outdoor or public spaces. The question was ‘Bagaimana kebiasaan Bapak/Ibu/Sdr menggunakan masker bila berada luar rumah?’ What is your habit of wearing a mask when you are outside the home? The answer choices were: never, seldom, sometimes, and always. Responses of “never”, “seldom”, and “sometimes” were categorized as inconsistent mask-wearing, while “always” was categorized as consistent mask-wearing outside the home.

The independent variables included risk perception, such as personal understanding of COVID-19, knowing someone who had the virus, ever having tested PCR positive, or recent close contact with a COVID-19-positive individual; the sociodemographic factors included location, age, gender, education, occupation, family income, number of family members, and ownership of a social health insurance scheme; and knowledge of COVID-19 included including knowledge of its causes, transmission, prevention (such as mask replacement frequency), post-outing hygiene practices, symptoms and severe symptoms, screening, confirmed diagnosis, management of mild and severe cases, comorbid conditions, and immune-boosting activities. Age groups were categorized based on social activity levels.

The questionnaire was developed based on the guidelines by the World Health Organization (2020b), Harvard Medical School (2020), and the Indonesian Ministry of Health (2020). It was pre-tested with individuals from various occupational groups to evaluate the clarity, relevance, and understanding of items related to sociodemographic factors, perception, and knowledge. In the pre-test, perception-related questions were open-ended questions and time-consuming to answer. These items were revised into close-ended questions with predefined response options based on the pre-test feedback.

Data were analyzed both using bivariate and multivariate logistic regression. Variables with a p-value ≤ 0.2 in the bivariate analysis were included in the multivariate model. Variables such as having tested PCR-positive (p = 0.804), close contact within two weeks (p = 0.580), and number of family members (p = 0.611) were excluded from the multivariate analysis. The multivariate analysis used backward likelihood ratio (LR) to identify independent associations among risk perception, where by interacting variables are deleted. It was to estimate each of risk perception, sociodemographic factors, and knowledge of Covid-19 variables, allowing for targeted public health interventions. Analyses were conducted using SPSS version 21 (Code 4-1EA96), with the significance was p < 0.05.

Results

Regarding mask usage, 2 (0.2%) respondents reported never using them, 37 (3.2%) seldom used them, 257 (22.3%) sometimes used masks, and 857 (74.3%) always used them. In terms of mask-wearing outside the home, 3 (0.3%) never used, 24 (2.1%) seldom used, 88 (7.6%) sometimes used, and 1038 (90.0%) always used.

Univariate analysis of determinants showed that, regarding perception of COVID-19, about a half of the respondents (585 or 50.1%) identified COVID-19 as a viral disease that can cause death, while other 554 (48.0%) individuals answered as viral disease with flu-like symptoms. There were 10 persons who answered “do not know” and 4 persons who described COVID-19 as non-communicable disease, with the total of only 14 (1.2%) respondents. Due to low frequencies and equal distribution across mask usage categories, these respondents were grouped into “common disease” category for the analysis, and not to COVID-19 was viral disease which could cause death. In terms of the feeling when neighbors were ill with COVID-19, most respondents (818 or 70.9%) reported anxiety, followed by 267 (23.2%) people who felt indifferent. Only 45 (3.9%) respondents had tested positive for COVID-19 by PCR, and 63 (5.5%) reported close contact with infected individuals in the previous two weeks.

Regarding the geographic areas, respondents were mostly from Greater Surabaya (502 or 43.5%), followed by areas outside Java, Sumatra, or Sulawesi (244 or 21.2%), and Greater Jakarta (191 or 16.6%). In terms of age distribution, 428 (37.1%) aged 15-24 (young adults), 438 (38.0%) aged 25-44 (adults), and 287 (24.9%) aged >44 years old (old adults). In terms of gender distribution, slightly above a half 603 (52.3%) were females. In terms of education background, the majority (517 or 44.8%) were academy/university graduates, followed by 288 (24.9%) with high school education. Regarding occupation, 348 (30.2%) were civil servants/police/soldier/government workers, followed by university/academic students of 198 (17.2%) and private employees of 193 (16.7%). In terms of family income, mostly 304 (26.4%) earned Rp >2,000,000-4,000,000, followed by Rp >4,000,000-7,000,000 (269 or 23.2%), and people up to Rp 2,000,000 244 (21.2%). Mostly respondents (796 or 69.0%) had ≤ 4 family members, and 961 (83.3%) were covered under National Health Scheme.

Regarding knowledge of COVID-19, most respondents (1,127 or 94.3%) mentioned the disease was caused by a virus; 1,087 (97.7%) understood the transmission was by droplets; 1,091 (94.6%) said the transmission can be from goods/surfaces; additionally, 787 (68.3%) said cloth mask should be replaced every 4-6 hours. Most respondents (1,017 or 88.2%) said the preventive measures of the disease were by all of wearing a mask, frequent hand washing with soap (using hand sanitizer), maintaining a minimum physical distance of 1 meter; 891 (77.3%) knew they should not directly take a rest immediately after going out.

For COVID-19 symptoms, 794 (68.9%) mentioned all four major symptoms (fever, flu, cough, and shortness of breath), while only 62 (5.4%) individuals knew just one symptom. Most respondents (1,032 or 89.5%) recognize that shortness of breath was a severe symptom. For screening, 1,018 (88.3%) mentioned the rapid test; similarly, 1,018 (88.3%) cited the swab test as a method for confirmed diagnosis.

For mild cases, almost all (1,080 or 93.7%) people mentioned self-isolation at home was the management, while 1,082 (93.8%) recognized that severe cases should be managed at referral hospitals. Regarding comorbid conditions, 769 (66.7%) respondents identified conditions that exacerbate COVID-19, excluding bone disease. In terms of immunity-boosting activities, 945 (82.0%) respondents identified all four recommended practices (light exercise, balanced nutrition, sufficient rest, and sun exposure between 09:00–11:00 for 30 minutes). Only 25 (2.2%) respondents mentioned just two of the activities.

Table 1 shows the bivariate analysis that where the risk perception of COVID-19 as a viral disease that can cause death was positively associated with consistent mask-wearing outside the home (crude Odds Ratio: 3.102; 95% CI: 2.018–4.766). In contrast, perceiving a neighbor’s illness with COVID-19 as causing stress or depression was negatively associated (crude Odds Ratio: 0.478; 95% CI: 0.316–0.723), and feelings of fear were also negatively associated (crude Odds Ratio: 0.357; 95% CI: 0.039–3.238). However, feelings of anxiety or worry were positively associated (crude Odds Ratio: 1.035; 95% CI: 0.401–2.668) with consistent mask use, compared to those who felt indifferent.

The risk perceptions of having ever PCR-tested positive for COVID-19 (crude Odds Ratio: 1.141; 95% CI: 0.401–3.246) and having had close contact (in the same room) with an infected person during the past two weeks (crude Odds Ratio: 1.302; 95% CI: 0.511–3.315) were not significantly associated with consistent mask-wearing outside the home.

Table 2 presents the bivariate analysis which illustrates the association between various sociodemographic factors (i.e. age, gender, education, occupation, family income, family member and insurance ownership) and consistent mask-wearing outside the home. Among these factors, the number of family members was not significantly associated with consistent mask use (p = 0.611). In terms of geographic location, it showed varying correlations. Residents of outer Surabaya were positively associated with consistent mask use compared to those living in municipalities or districts outside Java, Sumatra, or Sulawesi (crude Odds Ratio: 2.235; 95% CI: 1.120–4.459; p = 0.023). In contrast, other geographic areas (Java, Sumatra, Sulawesi, outer Jakarta) did not show statistically significant associations.

Age was a significant factor. Individuals aged 25–44 years (crude Odds Ratio: 0.275; 95% CI: 0.148–0.510; p < 0.000) and over 44 years (crude Odds Ratio: 0.485; 95% CI: 0.254–0.926; p = 0.028) were less likely to consistently wear masks outside compared to those aged 15–24 years. Regarding gender, female respondents were more likely to consistently wear masks outside the home compared to males with the crude Odds Ratio 1.285 (95% CI: 1.173–2.572; p = 0.006).

Education levels were negatively correlated with consistent mask use outside the home. Compared to those with less than high school education, respondents with high school education and university/academy degrees, were significantly more likely to consistently wear masks. In terms of occupational status, it revealed varying trends. Compared to civil servants/police/soldier/government workers, students and private employees were less likely to consistently wear masks.

Regarding family income, families with higher income (Rp >10,000,000-15,000,000 and Rp >15,000,000) were positively correlated with consistent mask use outside the home. respondents earning: Rp >2,000,000–4,000,000 and Rp >4,000,000–7,000,000 were less likely to wear masks consistently compared to those earning Rp ≤2,000,000. More than 4 family members were correlated with consistent mask use outside the home (crude Odds Ratio: 1.112; 95% CI: 0.738–1.677). Finally, ownership of National Health Scheme was significantly associated with consistent mask use outside the home (crude Odds Ratio: 2.010; 95% CI: 1.288–3.137), indicating individuals with health insurance were more likely to adhere to mask-wearing guidelines.

Table 3 shows the knowledge about COVID-19, specifically regarding its causation, transmission, recommended time to change cloth mask, post-outing hygiene practices, symptoms, case management, comorbidities that cause severe COVID-19, screening procedures, confirmed diagnosis, and activities to boost immunity, was positively associated with consistent use of mask outside the home. However, in some cases, greater knowledge about prevention methods, recognition of symptoms, and immunity-enhancing activities was negatively associated with lower levels of mask-wearing outside the home.

Table 4 shows that being aged 25 to 44 years was negatively associated with consistent mask-wearing outside the home (aOR: 0.486; 95% CI: 0.254–0.932), while perceiving COVID-19 as a viral disease that could cause death was positively associated (aOR: 2.502 (95% CI: 1.601-3.912). In addition, specific knowledge factors were significantly associated with consistent mask-wearing, including awareness of PCR testing as the method for COVID-19 confirmation (aOR: 2.238; 95% CI: 1.215–4.120), recognition of severe symptoms such as dyspnea (aOR: 1.981; 95% CI: 1.175–3.342), understanding of transmission modes (aOR: 1.974; 95% CI: 1.040–3.746), and knowledge of the recommended cloth mask replacement interval (aOR: 1.697; 95% CI: 1.118–2.576). In contrast, post-outing activities such as bathing and cleaning were not significantly associated with mask use outside the home.

Discussions

This online study investigated mask use during the COVID-19 pandemic under Indonesia’s mandatory mask policy. The rate of consistent mask-wearing outside the home was high (90.0%). Among mask types used, the majority reported using surgical masks (60%), followed by cloth masks (36.4%), N-19 masks (3.0%), and other types. Meanwhile, mask-wearing during the pandemic varied by region: 83% in the Western Pacific, 82% in Southeast Asia, 73% in the Eastern Mediterranean, 62% in Africa, 33% in Europe, and 32% in the Americas (Li et al., 2022). Initially, Indonesian people were less accustomed to wearing masks than other countries such as Japan, where mask-use has been culturally embedded (Ryall, 2020). Even in the early phase of the pandemic, many Indonesians did not adhere to mask use and they were met with social sanctions at the first wave of COVID-19, such as such as picking up trash, returning home to get a mask, buying one immediately, or doing push-ups. In several provinces like Jakarta and West Java, regional regulations even imposed fines for not wearing masks in public places.

Our study revealed a positive association between risk perception and consistent mask-wearing outside the home. In line with previous studies, viewing COVID-19 as a life-threatening viral disease was significantly associated with mask use (aOR: 2.502; 95% CI: 1.601–3.746) (Lo Moro et al., 2023; Shahnazi et al., 2020). However, feeling responses such as depression, fear, or anxiety when neighbors contracted COVID-19 were not significantly associated in the multivariate analysis (Yang et al., 2024). Additionally, personal experiences, such as having previous contracted COVID-19 or being in close contact with infected COVID-19 individuals during the two weeks prior to the study were not associated with the consistent mask use outside the home (White et al., 2022). These findings suggest that perceived risk of fatality, rather than direct exposure or feeling response, plays a stronger role in shaping mask-wearing behavior during a public health crisis.

According to the Health Belief Model, social behaviours are influenced by perceived risks, sociodemographic characteristics, and knowledge. In line with this model, the feelings of anxiety and fear, along with knowledge of disease significantly influence the risk perception (Sinicrope et al., 2021). Additionally, the anxiety and fear can shape social awareness (Xia et al., 2023), further impacting the adoption of preventive measures like mask-wearing.

Previous studies indicated sociodemographic factors such as older individuals, women, and government workers tend to adhere to mask use due to higher awareness of risks. Additionally, in urban areas, individuals with health insurance and larger households have also shown higher compliance and awareness of COVID-19 and subsequent mask use (Chu et al., 2020). This study found no multivariate association between consistent mask use outside the home and geographic areas, sex, education, occupation, family income, as well as ownership of social health scheme. Similarly, these findings align with a study conducted in South Africa, where age was negatively associated with consistent mask use during the early stages of the pandemic (Burger et al., 2022). Moreover, research in the United States showed younger individuals and those from lower socioeconomics were not associated with mask use (Willis et al., 2021). However, older individuals in the United States demonstrated a higher likelihood of mask-wearing, likely due to their greater awareness of personal health in relation to comorbid conditions (Liu & Arledge, 2022), The lack of association with other sociodemographic factors may be attributed to the mandatory mask policy implemented by the Indonesian government, which required all citizens to wear masks, particularly in public spaces. Additionally, the nationwide distribution of free masks, extended even to neighborhood levels, likely played a significant role in promoting widespread compliance across various demographic groups.

This study identified key areas of COVID-19 knowledge that were positively associated with consistent mask-wearing: correct replacement intervals for cloth masks, awareness of severe symptoms, understanding of transmission, and confirmation through PCR testing. These results are consistent with studies showing that specific knowledge contributes to protective behavior (Duong et al., 2021; Lo Moro et al., 2023). Furthermore, high global prevalence of perceived effectiveness of mask use in public settings (Li et al., 2022), support the potential influence of Indonesia’s mask mandate (Kar et al., 2023; The Ministry of Health of the Republic of Indonesia, 2020; Wismans et al., 2022). These insights underscore the importance of risk communication and targeted health education in promoting effective public health and preventive behaviors.

While sociodemographic factors such as age is amenable to change, this study highlights risk perception and knowledge are modifiable as key components for intervention aimed at promoting mask-wearing behavior. Despite high initial vaccination rates in Indonesia (86.88% for the first dose and for 74.55% for the second dose), booster uptake has significantly declined (38.11% for the first booster and 1.92% for the second booster as of July 31, 2023), highlighting a critical gap that must be addressed (https://vaksin.kemkes.go.id/#/vaccines). COVID-19 vaccination in Indonesia began on January 14, 2021, was implemented in four stages. Stages 1 and 2 on January to April 2021 targeted prioritized groups. Stage 1 focused on for health workers and their assistants, supporting staffs and students who were undergoing professional medical education in health services. Stage 2 targeted public service officers, including military personnel and police, law enforcement officers, and other public service officers including officers at airports, ports, stations, and terminals. This stage also included workers in the banking sector, state electricity companies, regional drinking water providers, and other frontline officers directly engaged in community service delivery. Additionally, individuals aged 60 years and above were included in Stage 2. Meanwhile, Stages 3 and 4, conducted from April 2021 to March 2022, extended vaccination to broader populations. Stage 3 targeted vulnerable communities, particularly those affected by geospatial, social, and economic disadvantages. Stage 4 focused on the general public and economic actors, following a cluster-based approach aligned with vaccine availability.

While COVID-19 vaccination offers protection and immunity and reduces disease severity (Havard Medical School, 2020), risk perception and knowledge about the virus remains essential motivators of preventing behaviour, including mask use. Our findings underscore the need for strategic interventions (communication and education) that address risk perception and knowledge gaps, encouraging consistent mask-wearing, especially with the declining vaccination rates (Wismans et al., 2022). Additionally, mask use can help prevent other respiratory infections (Liang et al., 2020; Wismans et al., 2022). Integrating these insights into ongoing public health strategies will be essential for improving health outcomes and pandemic preparedness.

Conclusions

In conclusion, this study demonstrates that the risk perception of COVID-19 as a viral disease that can cause death was positively associated with consistent mask-wearing outside the home. Meanwhile, specific sociodemographic factor of being aged 25 to 44 years was found to be negatively influence mask-wearing behavior. In addition, possessing knowledge, such as proper intervals for changing mask clothes, severe symptoms, COVID-19 transmission, and the confirmed diagnosis of COVID-19 by PCR was positively associated with consistent mask-wearing. These findings underscore the importance of targeted public health interventions that address risk perception and enhance knowledge to promote mask use as effective COVID-19 prevention measures. Future research could explore additional behavioral and contextual factors influencing mask use, as well as the long-term public health impacts of these preventive practices.

Ethics and consent

Ethical clearance for this study was granted by the Ethics Committee of the National Institute of Health Research and Development, Ministry of Health of the Republic of Indonesia (Number LB.02.01/2/KE.415/2020), dated on November 5, 2020. The committee confirmed that the study protocol, titled Factors Related to Preventive Behaviors of COVID-19 in Society, Indonesia”, involving human participants, had been thoroughly reviewed and found to be in accordance with the principles of the Helsinki Declaration. The principal investigator, Betty Roosihermiatie, declared that the approved protocol was ready for implementation. The ethical approval is valid from November 5, 2020, to November 4, 2021. Any modifications (amendments) or requests to extend the study period would require re-submission of the revised protocol for further ethical review and approval.