Asylum Seekers’ Responses to Government COVID-19 Recommendations: A Cross-sectional Survey in a Swiss Canton

Asylum seekers face multiple language, cultural and administrative barriers that could result in the inappropriate implementation of COVID-19 measures. This study aimed to explore their knowledge and attitudes to recommendations about COVID-19. We conducted a cross-sectional survey among asylum seekers living in the canton of Vaud, Switzerland. We used logistic regressions to analyze associations between knowledge about health recommendations, the experience of the pandemic and belief to rumors, and participant sociodemographic characteristics. In total, 242 people participated in the survey, with 63% of men (n = 150) and a median age of 30 years old (IQR 23–40). Low knowledge was associated with linguistic barriers (aOR 0.36, 95% CI 0.14–0.94, p = 0.028) and living in a community center (aOR 0.43, 95% CI 0.22–0.85, p = 0.014). Rejected asylum seekers were more likely to believe COVID-19 rumors (aOR 2.81, 95% CI 1.24–6.36, p = 0.013). This survey underlines the importance of tailoring health recommendations and interventions to reach asylum seekers, particularly those living in community centers or facing language barriers. Supplementary Information The online version contains supplementary material available at 10.1007/s10903-022-01436-3.

COVID-19 health recommendations seems to play an important role. A recent report by the International Organization for Migration (IOM) confirmed that limited awareness of public health recommendations due to linguistic and cultural barriers was a vulnerability factor for asylum seekers [13][14][15]. Similarly, lower access to healthcare systems, including mental health, and the fear of legal repercussions increased the risk of health inequities [12,[14][15][16]. For example, asylum seekers living in community centers during the pandemic face somatic and mental health challenges which require specific public health recommendations [15]. Reception centers are indeed characterized by crowded living conditions, shared rooms and little or no privacy, which could increase both the risks and the fears of being infected [17][18][19][20]. Accordingly, a recent systematic review advocates for better consideration of asylum seekers living in reception centers during the pandemic [8] and urge the need for adapted public health measures [12,[21][22][23][24] .
The current literature suggests that linguistic and cultural barriers, poor health literacy, living conditions, and legal status could contribute to mistrust of authorities and increase the COVID-19 pandemic burden among asylum seekers and refugees [25][26][27]. However, there is little current data on asylum seekers' access to and understanding of health recommendations and their perception of the COVID-19 pandemic. Moreover, the experience of the pandemic and the understanding of health measures among asylum seekers might vary according to place of living (community center vs private apartment), legal status, level of proficiency in French (the official language of the Swiss Canton studied) or health literacy. We, therefore, aimed to explore asylum seekers' attitudes and knowledge concerning COVID-19 recommendations and to describe associations between these variables and participants' socio-demographic characteristics.

Participants and Data Collection
We conducted a self-administrated cross-sectional survey about participants' knowledge, attitudes and perceived adherence to recommendations about COVID-19. Most survey questions were adapted from an online survey of the general population of the Canton of Vaud [28]. We simplified the language of the questions to a lower readability level in English. Then we translated it into the nine most common languages among asylum seekers residing in the Canton of Vaud: French, Tigrinya, Dari, Arab, Somali, Georgian, Tamil, Albanian and Serbo-Croatian. We translated the English questionnaire into these nine languages with the help of bilingual medical and nursing students from a local NGO and community interpreters. Except for Tigrinya and Tamil, a second translator proofread each translation.
We included asylum seekers, defined as asylum applicants with a pending procedure (N permit in Switzerland), as temporarily admitted (F permit), fully admitted (B permit) or rejected (emergency aid) residing in the Canton of Vaud. We excluded children under 18 years old, individuals not living in the Canton of Vaud, and former asylum seekers with a settlement permit (C Permit). Asylum seekers who cannot read or write were also excluded. In October 2020, according to the cantonal administrative data, 744 asylum seekers lived in one of the ten cantonal asylum community centers.
We identified 29 NGOs helping asylum seekers in the Canton of Vaud. We contacted them by email and phone to present the study and the survey questionnaires. We also worked closely with the persons in charge of the community centers in the canton of Vaud. We organized visits to all the centers to present the study and questionnaires to the residents. Finally, the questionnaire was also available online with a link sent to all study partners, including the identified NGOs.
The first page of the survey provided information in the selected language explaining that the study would like to know how they feel about the COVID-19 public health recommendations to improve the canton response and help research in this area. We also informed participants that the survey was anonymous and voluntary, and that they would not be contacted again. No incentive was used to encourage participation. The questionnaire took 15-20 min to complete.
We distributed the questionnaires (online and paper form) and collected data between August and October 2020. Online questionnaires were developed using the RED-Cap web application. We added the paper form data to the REDCap database in a second step. At the time of the data collection, the following health measures were in force in Switzerland: wearing masks in public transport, respecting social distance of 1.5 m, encouragement of hand hygiene, and recommended home office work. In addition, quarantine and isolation measures were mandatory.
All procedures were conducted following the ethical standards of the Human Research Ethics Committee of Canton de Vaud and the Swiss Law on Human Research. As all data collected were anonymous, an ethics approval by the Ethics Committee was not required. (Article 2 of the Swiss Law on Human Research).

Sociodemographic Characteristics
Sociodemographic characteristics collected included age, gender, level of education, French language proficiency, adapted and translated versions of a validated health literacy item [29], place of living (community centers vs private apartments) and legal status. The legal status variable is a dichotomization of the participant into two groups: the one with a permit (N permit, F permit or B permit) and the one with the Emergency aid status (rejected asylum seekers). This group represents indeed a particularly vulnerable category of asylum seekers and refugees as their legal status in Switzerland is highly insecure.

COVID-19 Data
The following questions asked participants whether they had been tested positive for COVID-19, were part of a group at risk (defined as people with comorbidities such as hypertension, diabetes, heart or lung problems or weaker immune system) and knew what to do if they had COVID-19 symptoms.
Then, the questions investigated the participants' understanding of the COVID-19 pandemic and public health recommendations. A knowledge score was developed with six true-false items about current government recommendations adapted from a previously published survey [28] (Fig. 1). Participants were also invited to answer six statements regarding COVID-19 rumors (Fig. 2). Visual analogue scores were used to measure self-reported adherence and perception of government measures.
The survey finally assessed the means of access to COVID-19 information and why participants stopped following the recommendations (Supplementary Information).

Statistical Analysis
We described study participant characteristics and answers to the questions using frequency (n) and relative frequency (percentage) for categorical variables and median and interquartile range (IQR) for non-normally distributed continuous variables (mean and standard deviation otherwise).
We used logistic regressions to explore associations between the outcomes of interest and participant characteristics, such as place of living (community center or private apartment), legal status (asylum seekers with a  permit vs rejected asylum seekers), health literacy (high vs low health literacy), education level (high vs low education level) or official language proficiency (high vs low French proficiency). Regression models were adjusted for age, gender and relevant confounders. Models' calibration was tested using the Hosmer-Lemeshow goodness-of-fit test. Associations with a p-value < 0.05 were considered statistically significant. Missing values were assumed to be missing at random. All analyzes were performed with STATA version 16.

Knowledge About COVID-19 Recommendations
In our study, only 43% (n = 104) of the participants had a high knowledge score (correctly answered at least 5 of the six questions of the knowledge score in Fig. 1), with a median score of 4/6. (See Table 2) After adjustment for age, gender and education, a lower knowledge score was associated with lower French language proficiency (aOR 0.34, 95% CI 0.13-0.89, p = 0.028) and living in a community center (aOR 0.41, 95% CI 0.20-0.84, p = 0.014), but not with health literacy level (aOR 1.52, 95% CI 0.81-2.84, p = 0.188).
Similarly, participants living in a community center were less confident about what to do if they got COVID-19 symptoms (naOR 0.30, 95% CI 0.15-0.60, p < 0.01), as well as participants with low health literacy (naOR 0.44, 95% CI 0.23-0.83, p = 0.01). (See Table 2) After adjustment, confidence remained associated with place of living and health literacy.

Access to Information About COVID-19 Recommendations
Most participants accessed information about COVID-19 recommendations on television (55%, n = 133), social media Table 1 Socio-demographic and COVID-19 related characteristics of participants (N = 242) a Dichotomized, "Often" and "Always" as high and "Never", "Rarely", "Sometimes" and "I don't know" as low health literacy b Dichotomized, "Very well" and "Well" as high, and "Not well", "Not at all" and "I don't know" as low French language proficiency  (49%, n = 119) and government websites (39%, n = 95) (Fig. 3). Participants living in community centers were statistically less likely to use television as a means of information. However, they were twice as likely to have accessed information via community interpreters than participants living in private apartments (22.3% vs 9.6%). These differences were statistically significant after adjusting for age, gender and education level for the use of television (aOR 0.42, 95% CI 0.23-0.75, p = 0.003) and community interpreters (aOR 2.99, 95% CI 1.29-6.91, p = 0.011).
About 51% of participants found that the COVID-19 measures were "about right", 11% found them not restrictive enough and 38% too restrictive. In a non-adjusted analysis, asylum seekers living in community centers considered the government COVID-19 measures as too restrictive (naOR 1.91, 95% CI 1.09-3.35, p = 0.03), whereas asylum seekers with low education level statistically significantly considered the measures as not strong enough (naOR 3.50, 95% CI 1.16-10.60, p = 0.03) (See Table 3).

Reasons to Stop Following COVID-19 Recommendations
The main reason for not following health recommendations was "the need to leave the house for food and essentials" (26.4%). In a non-adjusted analysis, this reason was statistically associated with a higher socioeconomic  (Fig. 4) . No association were found between this reason and the sociodemographic characteristics of participants (See Table 3).

COVID-19 Rumors
First, 39% of participants agreed with at least one alternative theory (categorized here as COVID-19 rumors without evidence) about the origin of the new coronavirus or the origin of the pandemic ( Fig. 2; See Table 2). After adjustment for age, gender and education, rejected asylum seekers were statistically more likely to think that "The effects of the coronavirus have been intentionally exaggerated so that governments can better control their populations" (aOR 2.81, 95% CI 1.24-6.36, p = 0.013).
Similarly, rejected asylum seekers (aOR 0.32, 95% CI 0.14-0.75, p = 0.008) and participants with a lower French language proficiency (aOR 0.31, 95% CI 0.15-0.63, p = 0.001) were less likely to believe that "The new coronavirus occurred naturally due to mixing of human and animal viruses".

Discussion
In this cross-sectional survey of asylum seekers in Switzerland during the Covid-19 pandemic, almost half of the participants had low knowledge regarding COVID-19 measures, despite high self-reported adherence and satisfaction with the recommendations. The need to leave the house for food and essentials was the main reason for not following health recommendations. In addition, participants identified television, social media and government website as the primary sources of information about COVID-19 health recommendations. Moreover, living in a community center, being a rejected asylum seeker or having lower French language proficiency were significantly associated with lower knowledge and stronger beliefs in COVID-19 rumors.
A similar study conducted among the general population of the same region (Canton of Vaud, Switzerland) showed a high level of knowledge among 67% of participants (versus 43% in our study) [28]. This difference is most likely caused by differences in educational attainment, health literacy level, French language proficiency, and consequently access to and comprehension of information. Our results are also consistent with other surveys assessing COVID-19 knowledge, notably among Afghan and Syrian refugees in Germany [30], Somali, Karen and Latinx community members in the US [31] or Syrian refugee women in Jordan [32] . Second, although participants described a high adherence to recommendations, almost 40% believed at least one COVID-19 rumor. This result confirms previous data from male migrant workers in Singapore, where authors found a high rate of participants believing in COVID-19 rumors [33] . Social determinants such as housing conditions (community center vs private apartment), legal status and language barriers (low French language proficiency) were associated with lower knowledge and belief to rumors. These factors should be considered in health decisions related to the COVID-19 pandemic to mitigate health inequities [34,35] .
Our study also describes the means of communication used by asylum seekers and refugees to access to COVID-19 recommendations. It is interesting to note the critical role of community interpreters for participants living in community centers where access to other means of information such as television is limited.
Our findings suggest that linguistically and culturally adapted communication seems essential to improve asylum seekers' and refugees' knowledge and adherence. Participatory approaches through community engagement and co-production could be helpful to actively build trust and strengthen public health campaigns, such as COVID-19 vaccination [36][37][38].
Our study has some limitations. First, the observational cross-sectional design of our study precludes temporal or causal interpretation of the observed associations. That being said, the cross-sectional findings provide a basis for further research on equitable pandemic responses. Second, the survey translations were not back-translated or tested for concordance with the original French questionnaire, although proofreading by another translator was possible for most of the languages translated. Third, the self-reported questionnaire may be subject to desirability bias. This bias is, however, limited by the anonymous nature of this survey. Fourth, our study may have potential confounding biases. Even though we adjusted for a range of potential confounders, it is possible that other factors not considered may interfere with the results. Notably, we have not assessed the cultural backgrounds of the participants. Fifth, our study has potential selection bias. Participants may have a higher degree of integration in society than the overall population of asylum seekers in the canton of Vaud. However, through our recruitment method and the translation of the questionnaires into nine languages, we hoped to limit selection bias. The proportion of participants with a low French language proficiency or in a very precarious social situation (rejected asylum seekers) suggests that this bias is likely limited.
In conclusion, the burden of the pandemic on asylum seekers and refugees is partly related to issues of understanding health recommendations, access to information and the consequences of health restrictions on their daily lives. And this access to information about recommendations and the belief to rumors are associated with language barriers, socioeconomic living conditions and legal status. Therefore, better anticipation of asylum seekers' specific communication and information needs in future public health crises is required. More systematic use of community interpreters or the involvement of communities in disseminating public health messages are potential solutions to tackle those issues and limit the spread of misinformation. Similarly, identifying specific social networks used by asylum seekers could facilitate the dissemination of targeted public health messages. However, further studies, including studies in other countries and longitudinal analyzes, are required to understand better the issues of access to COVID-19 information among asylum seekers and refugees.

New Contribution to the Literature
Our study found that asylum seekers living in community centers or with language barriers were at risk of health inequities related to poor access to or understanding of COVID-19 public health recommendations. Our study underlined the importance of tailoring public health recommendations and interventions to reach vulnerable populations and considering social determinants of health such as living conditions or language barriers in managing the COVID-19 pandemic among asylum seekers. In addition, findings suggested that more systematic use of community interpreters could help spread public health recommendations more efficiently.

Supplementary Information
The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s10903-022-01436-3. access to Option Grid patient decision aids. She receives consulting income from EBSCO Health, and royalties. No other competing interests declared.
Ethical Approval According to article 2 paragraph 2c of the Swiss Federal Law on Human Research (LRH), no ethical approval is required for data collected anonymously. Therefore, the survey of this manuscript, whose data are strictly anonymous, does not require validation by the ethics committee.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.