Hepatitis B infection is a major cause of death and disability worldwide, especially in areas of poor infant vaccine coverage. This study was initiated to better quantify hepatitis B vaccine completion among migrants with high rates of hepatitis B seropositivity in two different contexts in Thailand, to assess the impact of the COVID-19 pandemic on vaccination, and identify factors associated with improved vaccine completion. Strict travel restrictions imposed on migrants during the first and second quarters of 2020 hindered vaccination in Tak Province, but the impact of the pandemic was dwarfed by dramatic differences between the two study sites. Pervasive socio-economic marginalization of migrants in Tak province (indicated by low wages, under-documentation, unstable residence, and lack of government insurance) appears to be the main driver of vaccine non-completion in this population.
By far the most important factor determining vaccine completion was entering the study at SH in Chiang Mai versus SMRU clinics in Tak province. The SH experience, with vaccination completion of 86%, shows that it is possible to achieve high infant vaccination rates in migrant communities in Thailand, and suggests that the 90% goal could be attainable in some settings. However, at 47%, the vaccine completion at SMRU fell far below the target [4]. This difference could theoretically be due to three variables: individual factors of the parents, facility-level factors of the services, or broader socio-economic and political factors of the different settings. Though it is impossible to control for all individual-level factors, it is unlikely that this dramatic difference can be explained by demographics: women at SMRU were slightly more educated than women at SH, and age, parity, and hepatitis B seropositivity were similar. Predominant ethnic groups differed between sites but, as women delivering at SMRU had demonstrated higher understanding of hepatitis B and reported similar intention to vaccinate, cultural differences are not likely to account for the disparity. Residence less than 6 months at the current address was more common at SMRU and this has been associated with other negative health outcomes in this community [29], but the number of affected individuals was small.
There are differences between the clinical services at the sites. Unlike SMRU, SH does active follow-up for vaccine visits using phone calls. On the other hand, SMRU provides services in the preferred language of the migrant workers, which might be expected to improve follow up. Reported satisfaction with health services was similar in recent focus group discussions about choice of birth facilities in these same migrant populations [30], and vaccine services at both locations are free of charge.
Considering the relatively small differences between the individual characteristics and service provision, explanation of this fourteen-fold difference in vaccine completion requires consideration of broader socio-economic differences between the study sites. Social scientists have documented social and legal structures in Tak province which perpetuate a workforce of inadequately documented migrants who are then paid far below minimum wage and risk arrest when they travel outside their workplace [14, 15]. The pattern of documentation in this cohort supports this conclusion, with 90% of women at SH holding immigration documents compared with only 10% of women at SMRU. However, vaccine completion was only 30% among documented women at SMRU, reflecting the phenomenon in Tak Province by which migrants are marginalized and fear exploitation by authorities regardless of their documentation status as suggested by Reddy et al. [15]. The fact that no parents of infants born at SMRU reported vaccination in the free Thai government community vaccination clinics suggests a perceived lack of access to government services regardless of cost and official availability.
These apparent socio-economic barriers are consistent with qualitative studies [13, 22] in this population where major reasons for missed appointments included inability to miss work for even a day to get a child vaccinated, cost of travel, and fear of being arrested or fined on the way to the clinic. Distance to the facility increases both risk of arrest and the financial burden of travel, which often amounts to more than a day’s wages in Tak Province. Distance is generally shorter for urban (i.e. SH) than rural (i.e. SMRU) workers, and distance to services has been shown to have a negative impact on health service utilization among women seeking pregnancy services at SMRU [31]. Some programs in Tak province have shown potential to overcome these barriers such as school-based vaccination [18], and vaccine outreach programs run by district hospitals.
Other factors significantly associated with missed vaccines were lack of insurance, short duration of residence, and poor knowledge about hepatitis B after counselling. Young age had borderline significance. Taken together, these factors shed light on counselling opportunities that may improve vaccine follow-up. Targeting younger and more mobile mothers with extra encouragement to vaccinate, addressing barriers to insurance uptake, and adjusting the content of antenatal counselling to maximize understanding all have the potential to improve vaccine completion. As previously reported, almost all women intended to vaccinate their children, but higher knowledge did improve actual vaccination, after controlling for other factors. Previous work has shown that whenever possible, counselling strategies should be done in the parent’s preferred language, using non-technical terms, and should avoid overloading parents with too much information [23]. Approaches that optimize adult learning should be utilized as much as possible.
The magnitude of the impact of the COVID-19 pandemic on vaccination completion is difficult to estimate. Completed vaccination at SMRU clinics was less than 30%, compared with 50% in the previous studies [7, 22] but overall vaccination completion in the SMRU cohort almost reached 50% when doses received in Myanmar were included. However, there were substantial delays for vaccinations that occurred after the border closed. Since 60% of parents whose children were missing from vaccination could not be contacted, it is possible that a high proportion of these missed doses were due to COVID-19. The shift in reasons for missed visits from onward mobility to lack of mobility (due to pandemic travel restrictions) suggests a scenario where these marginalized children are chronically left behind – whether by the need of their parents to migrate for income or by the restrictions that limit both mobility and livelihoods during the pandemic. The impact of the border closure was mitigated by the availability of Myanmar vaccination services accessed by 38 infants receiving 56 hepatitis B doses in Myanmar, collectively. These services have now largely collapsed following the 2021 coup d’etat so the current degree to which vaccination has been disrupted is likely significantly greater.
There are several limitations to this analysis. Designed to identify classic demographic factors typically used in epidemiological studies, this analysis stumbled into socio-economic and political territory that it was not designed to elucidate. However, the available data on education, knowledge, documentation and insurance status corroborated phenomena described in more detail in the social science literature. The adjusted OR for several of the independent variables had wide confidence intervals reflecting the small number of participants in some groups. Determining the reason for loss to follow up is a chronic challenge that was not overcome in this work, where a reason for loss to follow up could not be determined in 60% of cases. This challenge highlights the difficulty of using mobile-phone-based interventions in the Tak migrant population where only 20% of participants could be contacted by phone. This contrasts with earlier findings that suggested cell phones were potentially helpful for childhood immunization programs in the Thai border population [32]. Finally, the study may have accessed different types of migrants at each study location. There are some documented migrants in Tak province who do seek care at the local Thai hospitals and may resemble the population seeking care at SH more than the children vaccinated at SMRU. Likewise, there are undocumented migrants in Chiang Mai who may not be seeking birth or vaccination services in the government system, and they were not represented in this study.
Case studies of migration and health are inevitably grounded in a specific geographic context. However, the scenario of migrant workers from an unstable state existing with various levels of documentation and acceptance within host countries with stronger economies and health systems is a global phenomenon of urgent importance [9,10,11]. Lessons from this setting are likely to be applicable elsewhere, though similar research in other settings would be valuable. As the world still struggles to recover from wave after wave of COVID-19 infections driven largely by unvaccinated communities, determinants of vaccination uptake in marginalized populations are particularly relevant.
Despite our best efforts to trace patients, this study was only partially able to quantify vaccine completion rates for migrants and elucidate reasons for missed doses. It is apparent that good vaccination completion rates are possible in some migrant populations, but fall short of vaccination rates among Thais. Barriers remain significant in rural border communities. Improvements in vaccination rates may be achievable through improved antenatal care education and targeting interventions for young mothers and those without insurance. Active follow up would likely be beneficial, but requires funding and personnel and should not rely heavily on phone contact. Interventions that decrease the cost of vaccination for migrant workers by decreasing travel distance, risk of arrest, and lost time from work such as community-level vaccine outreach or school-based vaccine programs are likely to improve uptake. Overall, the data from this cohort suggest that by far the greatest gains in vaccination coverage can be expected by strengthening mechanisms to include all migrant workers in national health systems.