Although studies on impact of climate change human lives, both health and non-health aspects, and migration as a coping mechanism has been studied around the world, less is known about utilization of health services among migrants especially in climate vulnerable coastal areas. Our study is the first to use strength of longitudinal data on migration and health care utilization to study the determinants of migration in climate vulnerable coastal area of Bangladesh and assess how intersection of coastal residence and migration create inequities in health care utilization. The findings from the study will be applicable to 2.5 million residents of several coastal regions of Bangladesh who are displaced every year due to frequent natural disasters and similar populations in low resource settings . Further the insights will help to inform policies and programs to reduce inequities in health service utilization among migrants who live in coastal areas.
According to our study findings, adults (≥15 years) were more likely to migrate compared to other age groups and more males migrated compared to females. Previous researches also reported that mostly adults were migrated in search of employment opportunities and in our cultural setting gender role assigned to men as breadwinners made them more likely to move in search of opportunities [32, 33]. On the other hand, we also observed that compared to those with little or no education, those with better educational status were more likely to migrate. Several studies suggested the similar findings which may indicate that education provide more job opportunities to people which in turn encourages migration [34–37]. In our study, we found that people engaged in agricultural professions were less likely to migrate compared to people in other professions. In previous studies, researchers explained that those engaged with agriculture often owned land and had better social status and power in the context of rural Bangladesh which probably discourages migration [38, 39]. Our study findings also demonstrated that members from poorest households were more likely to migrate than wealthier households. Similar findings were reported in several studies where researchers explained that the poor with limited resources are very vulnerable to natural disasters and lacking the resources to recover, often resort to migration for better employment opportunities [40, 41].
In this study, we found only 4.7%, 7.5% and 10.1% women of coastal areas accessed maternal healthcare services in terms of ANC, assisted delivery by SBA and PNC services respectively, significantly lower compared to those living in plains and the national average . Previous research conducted in geographically adverse rural areas of Bangladesh and other countries also reported the similar findings [43–49]. Researchers discussed a number of barriers of healthcare utilization at individual, community and health systems-level. At individual-level, lack of education, knowledge and awareness, poverty, limited mobility and autonomy of women; at community-level, difficulties in accessing health facilities due to lack of road and transport, cultural barriers related to using health facilities, and tradition of using local untrained traditional birth attendants; and at health systems-level, limited and expensive services and disrespectful behavior of health workers towards poor were associated with low service utilization [46, 49, 50]. The low service utilization was also apparent for child immunization, a primary care service with 95% national coverage  in the study population. The rate of child immunization coverage was only 23% among children of coastal areas, compared to that of the residents of the plains (67%). Previous studies conducted in rural or hard to reach areas of Bangladesh and other developing countries reported the lower childhood immunization rate than national average [52–58]. Although in Bangladesh the vaccination services are free of charge, Factors such as geographical and social inequalities, poor income, transportation cost, fragile communication system, limited knowledge and awareness about the benefit of vaccination, fear, stigma and lack of autonomy in decision making among mothers, have been associated with low rates of immunization among children [59–67].
Our research indicated that the intersection of living in disadvantaged areas and migration compounded the inequities in healthcare utilization. Our study findings showed that people who migrate in geographically vulnerable low-lying coastal areas have lower service utilization than residents. Researchers have described inequities in healthcare utilization among migrants compared to residents of the areas  . Several factors such as higher mobility, financial hardship, being poorly informed about health risk and available health facilities, high healthcare cost relative to income have shown to adversely affect the use of healthcare services among the migrants [46, 70–72].
Bangladesh has made remarkable progress in achieving several health indicators as indicated by the achievement of Millennium Development Goals (MDG 4 and 5) related to maternal mortality and child mortality . However, given that Bangladesh is one of the countries most affected by climate change and contain 2.5 million climate migrants, it is imperative that health of the migrant population is given priority [1, 17]. To achieve Sustainable Development Goal (SDG10) related to reducing inequality in the population, it is important that gaps in service provision are addressed and barriers to healthcare seeking among vulnerable and hard to reach populations such as the migrants, are removed . While climate migrants use migration as an adaptation strategy it is important that health systems have the provision to ensure that primary care is accessible and available to the migrants while they make efforts to overcome poverty. Multi-sector partnership, coordinated efforts, migration-friendly policies and strategies are needed to develop the migration-sensitive health systems.
The significant strength of this paper was that we analyzed the longitudinal data from multiple years which allowed us to look at trends over time and have sufficient numbers to look at determinants. This study also had few limitations that should be acknowledged. This study was cross sectional in nature, so it does not permit any cause and effect relationship to be inferred. As the analysis was done based on existing data we could not add any questions to refine our inquiry.