Introduction

Transnational migrants play a crucial role in Thailand’s socio-economic development, representing over 10% of the total labor force [1]. Nationals from Myanmar constitute 68% of the migrant workforce in Thailand [1]. Migrant workers were particularly vulnerable during the COVID-19 pandemic in Thailand, with a high number of infections [2–6], particularly among those from Myanmar [7]. Migrant workers tend to live in overcrowded accommodations with insufficient protective measures, which can contribute to outbreaks of COVID-19 beyond the migrant community [8]. For example, asymptomatic infections among Myanmar migrant workers in Samut Sakhon Province triggered the second wave of COVID-19 pandemic in Thailand [9].

A study reported that only one-third of Myanmar migrant workers in southern Thailand regularly adhered to COVID-19 preventive measures, including mask-wearing, hand-washing, and physical distancing during the pandemic [10]. During the COVID-19 pandemic, a survey by the International Labour Organization reported that 57% of migrant workers in Thailand lacked facemasks and hand sanitizers [11]. The International Organization of Migration also reported that Myanmar migrants were in need of employment, food assistance, hygiene items, medicines, medical support, and personal protective equipment [12, 13]. The Migrant Working Group reported that migrant workers needed protective equipment and dried foods and faced challenges in implementing distancing measures due to overcrowded, unsanitary living conditions [14].

The World Health Organization reported that migrants often face significant obstacles in accessing health services in their host country, including language and communication barriers [15]. In the field of public health, language is often identified as a major healthcare barrier for migrants in Thailand [16] and other host countries [17]. Previous studies among migrant populations in Western countries found that the host country’s language proficiency was positively associated with infectious disease prevention, screening, and management [18–20].

In Thailand, migrant workers from Myanmar typically have low-skilled occupations [1], a low level of education [21–23], and a limited ability to communicate in Thai language [24, 25]. Although previous studies have shown that Thai language proficiency is associated with low COVID-19 vaccination [26] and a lack of COVID-19 healthcare information [27] among Myanmar migrant workers, several knowledge gaps still exist. Language proficiencies generally include four domains (speaking, listening, reading, and writing) [28], and a lower proficiency in one domain can be supplemented by higher proficiency in others. Thus, all of these domains need to be measured in other to accurately evaluate the extent to which language proficiencies are associated with COVID-19-related outcomes [28, 29]. However, previous studies tended to measure only speaking and listening skills among Myanmar migrants [25], which provided an incomplete picture of language proficiency. Migrants with moderate speaking skills who could read and write Thai would arguably be more capable of independently accessing Thailand’s healthcare system. A related issue was the measurement of language competency [29], which varies widely among different migrant populations [30–33] without standardization and may involve resources at a level not suitable for rapid population surveys.

Utilizing the Common European Framework of Reference (CEFR) [29], a globally recognized standard guideline for measuring foreign language ability, could enable a more comprehensive assessment of self-reported language proficiency. The CEFR provides a standardized framework for assessing foreign language proficiency comprising six levels, namely A1, A2, B1, B2, C1, and C2, each with an increasing level of proficiency [34]. The A1 level designates the ability to understand and use familiar everyday expressions, introduce oneself, and make simple interactions with users of the language [35]. The B1 level refers to the ability to understand the main points of conversations on familiar matters, produce simple texts on familiar topics, describe experiences and ideas, and explain opinions and plans [35]. The C1 level refers to the ability to understand complex clauses with implicit meaning, use the language for all purposes flexibly and effectively, and produce texts on complex subjects [35]. Adapting the CEFR’s definition for self-reported assessment can offer rapid measurement of language proficiency in all four domains and help us assess the extent to which language proficiency is associated with infectious disease prevention behaviors among vulnerable migrant populations. Stakeholders in migrant health and health promotion can use these study findings as baseline information on the extent to which the target population understand the host country’s language, and the extent to which such understanding is associated with COVID-19 protective behaviors. Similarly, stakeholders in infectious disease prevention and control can use the findings to inform and justify the need to provide interpreters and other language services, or lack thereof, when working with language-diverse populations. The objective of our study is to assess the extent to which self-reported command of the Thai language according to the CEFR is associated with adherence to COVID-19 protective behaviors among Myanmar migrant workers in Thailand.

Methods

Results

We approached 1350 migrants, among whom 1050 agreed to participate (response = 77%). Most participants were males, Burmese, Buddhists, married, and factory workers (Table 1). Nearly one-third had a secondary level of education, with a median monthly personal income of 9500 Thai Baht (THB). Most respondents had lived in Thailand for less than five years, with their family or relatives. More than three-fourths of the respondents (79.1%) had neither migrated (those who did not move from one place or province to another) nor changed jobs (those who have not switched from one employer or occupation to another) within the past five years in Thailand. The majority of the respondents had health insurance and legal documents. More than half of the respondents had no Thai language proficiency [i.e. those who indicated that they had no ability to speak, understand, read, or write in Thai] in reading (93.0%), writing (93.5%), listening (52.7%), and speaking skills (52.6%), followed by A1 level in reading (3.0%), writing (4.3%), listening (38.3%), and speaking skills (35.3%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Background characteristics of the respondents (n = 1050).

https://doi.org/10.1371/journal.pone.0312571.t001

Although unadjusted linear regression models (Table 2) showed that workplace COVID-19 prevention behaviors adherence scores were significantly associated with Thai speaking and listening skills. However, multivariate linear regression models showed that the associations were not statistically significant (Speaking skill’s adjusted beta = 0.713, 95% CI: -0.011, 1.437; Listening skill’s adjusted beta = -0.367, 95% CI: -1.087, 0.353). The place of residence COVID-19 prevention behavior adherence scores had no significant association with any domain of Thai language proficiency (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Association between Thai language proficiency and adherence to workplace COVID-19 prevention behaviors (n = 1050).

https://doi.org/10.1371/journal.pone.0312571.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Association between Thai language proficiency and adherence to residence COVID-19 prevention behaviors (n = 1050).

https://doi.org/10.1371/journal.pone.0312571.t003

Discussion

In this cross-sectional study, we assessed the extent to which self-reported Thai language proficiency was associated with CPB adherence among Myanmar migrant workers in Southern Thailand. The findings revealed that slightly less than half of the participants reported CEFR A1 level or higher in Thai speaking (39.33%) and listening proficiencies (41.61%), and less than 10 percent did so for reading (3.24%) and writing skills (4.86%). Workplace COVID-19 prevention behavior adherence was associated with Thai speaking and listening proficiencies, but the associations were non-significant after adjusting for confounders. There was no association between Thai language proficiency and COVID-19 prevention behavior adherence at the place of residence. Our study findings provide basic information that is potentially useful for stakeholders in migrant services, migrant health, and infectious disease control.

Based on the CEFR framework, approximately half of our participants reported an A1 (beginner) level of speaking (53.23%) and listening abilities in Thai (54.95%), but literacy was very low for reading (3.24%) and writing (4.86%). Several issues should be considered in the interpretation of our study findings. Firstly, language skills in our study were all self-reported and subjected to various biases (e.g., social desirability, response acquiescence, and self-serving biases) [51, 52]. Future studies with adequate time and resources could consider using an abridged version of standardized language tests instead of self-reporting [30, 53]. Secondly, we measured general language proficiency rather than health-related language proficiency. Thus, there was an issue regarding construct validity. Our language proficiency measurement questions were on general language skills (e.g., ability to tell a story to the police when asking for help) and not on the language skills specific to navigating the public health system (e.g., ability to describe disease history and symptoms to healthcare providers). Future studies should consider measuring health-related language proficiency in addition to general proficiency. Thirdly, our study had a relatively limited scope on health-related behaviors and experiences of migrants. Navigating a foreign healthcare system generally requires additional skills [16, 54], such as understanding and negotiating the host country’s medical insurance and social welfare system. Future studies should consider measuring transcultural-transnational competencies related to health and social services in addition to language proficiency. Lastly, we did not measure several independent predictors of adherence to COVID-19 protective behaviors, including access to information about COVID-19 from the Ministry of Public Health (e.g., instructions, signs, and notifications in languages other than Thai), COVID-19 prevention guidelines and regulations at workplaces (e.g., the ’No mask = No entry’ sign in English or Burmese), overall awareness of COVID-19 related health information, and components of the health belief model (e.g., perceived susceptibility to COVID-19 infection, perceived severity of COVID-19 infection). These characteristics could have confounded the association between Thai language proficiency and adherence to COVID-19 protective behaviors. Future studies should consider measuring and controlling for these characteristics in a multivariable model, after which a more valid estimate of the association can be inferred.

Most of our participants reported following COVID-19 protective behaviors in the workplace and in the residence. The scores were relatively high, suggesting good adherence to protective behaviors. However, we collected our data in late 2022, approximately 2 years after the start of the pandemic. The findings of the study may not be generalizable to other contexts. Our questionnaires were anonymously self-administered, which reduced the potential for social desirability bias. However, self-serving bias might have existed among participants who deemed COVID-19 protective behaviors to be signals of good virtues. Furthermore, individual behaviors tend to change over time, and our study instrument did not capture these changes. Future studies should consider longitudinal data collection to provide more insights into the adoption of and adherence to protection behaviors [55–58].

In our study, workplace and residential COVID-19 prevention behaviors were not significantly associated with any domain of Thai language proficiency after adjusting for confounders. COVID-19 spread widely among Myanmar migrants in Thailand during the 2^nd wave of the COVID-19 outbreak in December 2020 [7], approximately two years before our survey. These past experiences might have induced a community-wide awareness and vigilance against COVID-19, irrespective of language proficiency, which might have contributed to the lack of associations found in our findings. A prior study in late 2021 found that Myanmar migrants in Southern Thailand were compliant with wearing face masks, hand washing, and avoiding unnecessary travel [10], similar to another study conducted in the Thai-Myanmar border areas [59].

The main contribution of this study to the literature was the measurement of self-reported Thai language proficiency using the CEFR framework, a globally recognized standard for language proficiency assessment. The instrument of this study can serve as the basis for future studies in both linguistics and public health, considering the extent to which language proficiency intertwines with social support and the ability to navigate the social, economic, and health systems in a foreign country [18–20]. However, a few limitations should be considered in the interpretation of the study findings. Firstly, we did not make a systematic content validity assessment of our questionnaire due to lack of time and resources and relied mainly on internal deliberations for questionnaire development. This lack of internal validity assessment should be considered as an important limitation in the interpretation of the study findings. Secondly, the lack of back-translation of the Thai language proficiency skills measurement questions poses a substantial threat to the validity of our exposure measurement. Thirdly, self-reporting of information did not preclude exposure and outcome measurements from social desirability and other types of biases. Fourthly, our questionnaire was available only in the Burmese language, whereas our participants included Myanmar migrants who might not have been able to read Burmese fluently. The potential language barrier could have similarly introduced information bias to our study findings. Fifthly, our data collection staff were ethnic Burmese, and pre-existing mistrust and social dynamics between ethnic groups could have also influenced and introduced bias to our study findings. Lastly, the findings of our study may be generalizable only to the context of the study areas, where migrants work mostly in factories, construction sites, and fisheries. Thai proficiency and behaviors of Myanmar migrants in predominantly service sector areas, such as at business establishments in big cities, could have been markedly different from those of our participants. Future studies should consider revising the measurement questions to reduce biases, inclusion of additional languages of Myanmar (e.g., Karen, Mon, Rakhine, etc.), and increase the diversity of the data collection staff.

Conclusion

Myanmar migrants in low-paying occupations in Southern Thailand had generally low level of Thai language proficiency. However, language proficiencies were not associated with workplace and residential COVID-19 prevention behaviors after adjusting for confounders. Our study findings provide basic information that is potentially useful for stakeholders in migrant services, migrant health, and infectious disease control. However, information biases, language barriers among migrants who were not proficient in Burmese, and lack of generalizability should be considered as caveats in the interpretation of the study findings.

Supporting information

Acknowledgments

We would like to thank all the migrant workers who agreed to participate in our study, the managers and the administrative staff of the factories, construction, and fishery sites who facilitated our data collection, and the local key informants who provided valuable insights and assistance.