- Study protocol
- Open Access
Healthcare system resilience in Bangladesh and Haiti in times of global changes (climate-related events, migration and Covid-19): an interdisciplinary mixed method research protocol
BMC Health Services Research volume 22, Article number: 340 (2022)
Since climate change, pandemics and population mobility are challenging healthcare systems, an empirical and integrative research to studying and help improving the health systems resilience is needed. We present an interdisciplinary and mixed-methods research protocol, ClimHB, focusing on vulnerable localities in Bangladesh and Haiti, two countries highly sensitive to global changes. We develop a protocol studying the resilience of the healthcare system at multiple levels in the context of climate change and variability, population mobility and the Covid-19 pandemic, both from an institutional and community perspective.
The conceptual framework designed is based on a combination of Levesque’s Health Access Framework and the Foreign, Commonwealth and Development Office’s Resilience Framework to address both outputs and the processes of resilience of healthcare systems. It uses a mixed-method sequential exploratory research design combining multi-sites and longitudinal approaches. Forty clusters spread over four sites will be studied to understand the importance of context, involving more than 40 healthcare service providers and 2000 households to be surveyed. We will collect primary data through questionnaires, in-depth and semi-structured interviews, focus groups and participatory filming. We will also use secondary data on environmental events sensitive to climate change and potential health risks, healthcare providers’ functioning and organisation. Statistical analyses will include event-history analyses, development of composite indices, multilevel modelling and spatial analyses.
This research will generate inter-disciplinary evidence and thus, through knowledge transfer activities, contribute to research on low and middle-income countries (LMIC) health systems and global changes and will better inform decision-makers and populations.
Contemporary pandemics, mobilities and climate change have highlighted more than ever the need to develop an empirical, integrative, applied approach for documenting the resilience of healthcare systems beyond concepts in global health . This situation is all the more dramatic for low and middle-income countries (LMIC) as the costs of these shocks and future, unforeseen emergencies will drain already scarce resources and accentuate existing inequities. Based on conceptual analysis and other empirical papers [2, 3], we define health systems resilience as: the constituents’ abilities of a health system facing destabilising experiences, events or shocks (contingent or expected, sudden or insidious, internal or external) to adapt and transform in order to maintain and/or improve access (for all) to comprehensive, relevant and quality healthcare and services without pushing patients into poverty.
Health systems resilience (HSR) is at the core of current preoccupations. The 2020 Lancet Countdown report on health and climate changes recommend to integrate climate change and the COVID-19 crisis responses, in a way that addresses inequality directly . Globally, 400 million people have no access to essential health services . Achieving universal health coverage (UHC), including financial risk protection, access to quality essential healthcare services and access to safe, effective, quality, and affordable essential medicines and vaccines for all is part of the sustainable Development Goal (SDG) 3.8 for 2030. “For all” is essential in this declaration, as a significant proportion of the population, such as migrants, do face distinctive vulnerabilities to poor health  and poor healthcare access .
Various kind of emergencies are expected to continue happening, but a more resilient and robust healthcare system can minimize their impacts, which is within reach of all countries, even the poorest . Because low- and middle-income countries (LMIC) and their populations face numerous social vulnerabilities and are already fragile in terms of health coverage and health structures; the necessity to document the resilience of national healthcare systems is even more compelling. There is an urgent need to build empirical evidence on the resilience of health systems and the strategies stakeholders organize to respond to these crises in order to thoroughly inform decision-makers to better anticipate.
Despite health system resilience being a widely researched topic, this concept lacks of conceptual maturity, needs clarification and empirical validation . In this context, we present the design of an interdisciplinary research project (ClimHB project) focusing on health systems resilience through the example of vulnerable localities in Haiti and Bangladesh. These two countries are highly sensitive to climate change and other global and local risks. These changes and risks are pushing out people from their communities, or trapping them when unable to move away  thus influencing populations mobility and immobility. Both countries are experiencing global changes through the exacerbation of disruptions caused by climate-related events, (im)mobility and the Covid-19 pandemic. In order to understand how local contexts, influence and interact with patterns of resilience and population health, we use a multi-sites approach, to integrate a variety of local contexts and vulnerabilities, because context matters [10,11,12]. This project aims to answer two interrelated questions in the context of climate-induced environmental change (CIEC), population (im)mobilities, Covid-19 pandemic and other possible changes, events and risks affecting the population and the healthcare systems:
What is the relative resilience of local health service providers in various settings vulnerable to climate change and (im)mobility in Haiti and Bangladesh?
What are the patterns and distribution of health status and access to healthcare services of the (im)mobile populations in these various settings in Bangladesh and Haiti?
In the first section of this paper, we present our conceptual framework, study sites and methodologies. In the second section, we discuss the challenges and innovations of this approach and protocol.
While the World Health Organization (WHO) is calling governments to improve health systems resilience, this concept is yet to be clarified . Our choice of definition includes all types of destabilising experiences affecting systems such as events (shocks, crisis), (everyday) changes and risks and encompasses both the institutional and the patient level. In this context, we propose our conceptual framework (Fig. 1) to systematically analyse the resilience processes at both provider and population levels and their effects on healthcare access and utilisation. In other words, we tried to develop a framework to understand both the resilience processes and the output of the resilience (health access). Various healthcare-use conceptual models have been proposed [13, 14]. Yet, they have in common that access to care is “the opportunity to reach and obtain appropriate healthcare services in situations of perceived need for care” . The Levesque’s Conceptual Framework of Access to Health  model stands out as it encompasses both healthcare service providers and population dimensions with five dimensions related to the institutional level (approachability, acceptability, availability and accommodation, affordability and appropriateness) and five capacities related to individuals or households (abilities to perceive, to seek, to reach to pay and to engage). We extend this framework with a conceptual resilience approach developed by the Foreign, Commonwealth and Development Office (FCDO, formerly UK Department for International Development- DFID). According to the FCDO, most resilience definitions share four common elements: i) context; ii) disturbance; iii) capacity to deal with disturbance; and iv) reaction to disturbance which appears relevant in our context of climate-related and other events and changes in Haiti and Bangladesh. FCDO has developed a framework for integrating resilience into their work on climate change and conflict prevention based on these four elements, because “it can be used to examine different kinds of resilience and help determine the level of resilience that exists” .
We base our integrated analytical framework on both the FCDO conceptual proposal  and the importance of interactions between the health system and the population in achieving access to care, as Levesque et al. suggested  (Fig. 1). Consequently, we focus on both the process and outputs of local health system resilience in the context of global and local changes, including 1) environmental risks such as climatic-related events and pandemics such as Covid-19, 2) (im)mobilities and migration as socio-demographic changes and 3) any other global and local events, changes and risks that could affect healthcare systems at institutional and/or population levels.
According to the Global Climate Risk Index in date, Bangladesh was ranked 7th and Haiti 3rd on the Long-Term Climate Risk Index . Both countries are prone to disasters due to their geographical location and environmental characteristics. They are subject to hydro-meteorological disasters; especially cyclones, heavy rainfalls and flooding. Besides, Haiti is facing geophysical hazards. Vulnerability to climate change is also related to the combined consequences of other structural and institutional factors: deep social inequality, high poverty rate, poor public health infrastructure and accessibility, land use and access, high population density (Bangladesh), rapid urbanisation, civil unrest and violence (Haiti) [21,22,23]. The context in Haiti is that of severe socioeconomic deterioration, denunciation of corruption, long lasting political crisis and recent constitutional crisis. This has resulted in lockdowns, massive social unrest and demonstrations. Since 2018, Haiti has been experiencing a significant increase in human rights violations and abuses, crime and massacres, kidnappings, restrictions on freedom of expression and freedom of the press, continuing impunity and aggravated food insecurity [24, 25]. This serious socio-political instability has led to the closure of some care facilities . Mobility, whether climate-induced or not , could put an additional strain on healthcare systems. Because population mobility and migration have multiple causes and have a complex relationship with climate change, scientists are cautious about linking mobility and migration to climate change . However, some climate-related events have been shown to directly or indirectly influence mobility and migration decisions . This nuance being stated, mobility and migration affect the healthcare system at two levels: at the organisational level, for example with a structure, services and workforce that are not adapted to the patients and at the population level with an increased vulnerability of the population. Both health system performances are poor: Bangladesh is ranked 88th and Haiti 148th out of 191 countries in 2000 by the WHO. When considering the UHC effective coverage index in 2019,Bangladesh is in the 6–7 decile and Haiti is in the lowest decile .
Exploratory key informant interviews and reviewed secondary data resources were conducted to select two sites in each country, based on climate-induced vulnerabilities, mobility patterns and availability of health system secondary data for the past decade when possible. Sites boundaries are the ones defined by local health system units. We selected for their local context two rural coastal Haitian sites (Anse d’Hainault and Môle St Nicolas) and one rural (Tala) and one urban (Duaripara slum) Bangladeshi sites (Fig. 2, Table 1).
Regarding migration and mobility in Bangladesh in 2020, 88,000 people living in displacement (IDPs) were registered due to disasters and 427,000 IDPs due to conflict and violence, which, for most, are protracted cases from the partitioning and independence of the country, were recorded . As of 31 December 2019, 4,086,000 new displacements due to disasters were recorded, many in the form of life-saving pre-emptive evacuations, and 520 new displacements due to Intercommunal violence against Ahmadi Muslims in Rangpur division and political violence following elections .
Tala upazila (or Tala subdistrict) is located in the south-western coastal region of Bangladesh, in the Satkhira district of Khulna division, which is one of the most disaster-prone areas of the country, especially for flooding and waterlogging [32, 33]. Tala has at least 30,000 inhabitants, 15 to 20 community clinics and five sub-districts hospitals. The primary source of income is agriculture. Men, pushed out by lands waterlogging, migrate out for labour, while women, children and elderly, remain behind with poor access to sanitation and drinking water. Pregnant women often suffer from reproductive health issues because they cannot access healthcare .
Duaripara slum is located in ward 6 of Dhaka North City Corporation. Most households are extremely poor according to BRAC Urban Development Program (UDP). Its population increased recently when residents from the nearby Bhola slum moved here after they were evicted. Most of the Bhola slum residents migrated from Bhola Island after losing their land due to riverbank erosion or property loss and damage due to cyclones . Healthcare is mostly limited to over-the-counter treatment at pharmacies; residents go to government and private hospitals, which are far, only during serious illnesses. 24.7% of the residents work in a garments factory followed by 15.6% of residents who work as day labourers and other significant occupations are rickshaw-pulling and domestic work . All houses have electricity, although non officially, and piped water is shared by around ten households. The slum settlement area is built on wetlands. Flooding, standing water, fires, and storms, especially during the monsoon season, are frequent in the area.
In Haiti, many people displaced by the 2010 earthquake and hurricanes such as Matthew in 2016, are still living in displacement . As of 31 December 2019, the total number of people living in displacement (IDPs) includes 51,000 IDPs due to disasters and 2100 IDPs due to conflict or violence . For 2019, 1200 new displacements due to disasters and 2100 due to conflict and violence were recorded in Haiti .
Anse d’Hainault arrondissement is located in the Grand’Anse Department which has been affected by Hurricane Matthew in 2016. It has an estimated population of 98,522 inhabitants (IHSI estimates for 2015), with an extreme poverty rate of around 35% and counts 11 health facilities: six dispensaries, two health centres without beds, one with beds and two community hospitals. Coastal plains are subject to marine submersion as well as floods. Essential services like access to water or electricity are deficient. In rural areas, as in Mole St Nicolas, agriculture is the primary source of income.
Môle Saint-Nicolas arrondissement has an estimated population of 245,590 inhabitants (IHSI estimates for 2015) and is considered one of Haiti’s poorest departments, with an extreme poverty rate above 40%. It counts 36 health facilities: 30 dispensaries, two health centres without beds, three with beds and one community hospital.
Data collection methods
The first exploratory (contextual) phase consists of a contextual analysis to describe, political, social, economic, and health factors and their relationships, both at the national and local levels, through 1) a compilation of local monographs describing the four sites’ socio-political organisation, the history of the communities and their livelihoods, local healthcare system and demographic trends, 2) public health and public health reforms and migration policies at the national level, and any other experiences that have an impact on the local health system and population, through interviews, document analysis, and public service records and archives. In addition, a quantitative analysis of secondary data from routinely collected data (DHMIS) and SARA surveys, combined with the development of indicators, will be conducted to better understand the impacts of risks, events and climate change over the past decade on healthcare service providers (HSPs) at the four sites. These analyses are intended to allow for longitudinal and transversal analyses.
For the second phase, we use a qualitative methodology with a mixed inductive-deductive approach. This phase aims to map out the variety of disruptions and disturbances faced by HSPs and the populations and identify the diverse coping strategies implemented to understand the factors and stages in the process of resilience in the four sites. Mobility and migration will be part of the discussion on coping strategies of the population. In-depth and semi-directed interviews with national (n = 2) and regional/local (n = 1 + 2) health decisions makers and with the heads of the 10 selected healthcare service providers within each site, and focus group discussions with inhabitants of the community (n = 3), community-based groups (n = 3) and healthcare users (n = 3) will be organised within each site too. Those numbers will be adapted following a purposeful sampling strategy, based on criteria (age, sex, socio-economic status, being a patient, place of living) and according to the results of pilots and each previous phase . Finally, a mental health research arm, whose specific protocol will be described in another publication at a later stage, will be carried out to further investigate issues pertaining to mental health and trauma in the face of extreme environmental events. For this embedded study, qualitative data will be collected to complement our quantitative data, at a different timing than phase 2.
During the third phase, all HSP health staff and a sample of household members from the communities around the 10 selected HSPs will be interviewed to complete questionnaires (n = 500 households per site- see below for sampling strategy and eligibility criteria). This quantitative phase aims to quantify the (perceived) impacts of identified disruptive events, changes and risks on the healthcare providers, healthcare access and population health status, based on the results from phase 2. Specifically, the household questionnaire includes different modules in two sections, a household and a respondent section. The respondent section includes socio-demographic information, health status (acute and chronic diseases), reproductive health, maternal health and family planning, mental health and trauma, empowerment, history of health and history of mobility, social capital, livelihood security and community perception, and a module on healthcare access. The household section includes household socio-demographic and economic information. The HSP questionnaire includes questions on five supply-side determinants on healthcare access  in connection with the different SARA surveys available in both countries. It will describe the situation and assess, if possible, the impacts of disruptive events on local healthcare service providers. Interview guides (phase 2) and questionnaires (phase 3) are presented in the supplementary materials.
The fourth phase aims to document further and better understand the typologies and strategies that will have emerged from the two previous phases. During this qualitative phase, a nested sample of health professionals (n = 20) and household members (n = 30 households) will be interviewed.
Eligibility criteria and sampling strategy
We will select ten HSPs per site (20 per country) to represent the diversity of levels (first and second), geographical situations (urban and rural) and types (private, for-profit, private non-profit, public) of each site. We will consider only HSPs with a minimum attendance level: five consultations per day in Haiti and 30 in Bangladesh.
For the phase 2, we will identify key informants who will be able to provide specific information on our study topic  and will be selected to have a wide range of understanding of the phenomenon. Regarding the decision-makers (n = 5), they will be interviewed at the local (n = 2), regional (n = 1) and also national levels (n = 2).
Community inhabitants (mostly non-users of healthcare), community groups (women’s groups, trade unions, community health workers, etc.) and users of healthcare services (15 patients will be randomly selected from three of the 10 healthcare service providers selected above, while taking account of gender and residence distance from HSP), will be enrolled for focus groups participation according to their knowledge and experience. Community leaders, local NGOs and healthcare facility managers will assist us in selecting these individuals.
To implement phase 3, 500 households per site (1000/country), spread over 10 clusters corresponding to the 10 HSPS will be selected in each site. Clusters will coincide with circles with a kilometre radius in Bangladesh and a 1-to-3 km radius in Haiti around the previously selected HSPs. In each circle, we will identify households through photo interpretation. Then, we will select a spatialized random sample of 50 households.
The household survey will be conducted to select households that are
with recent (3 weeks) contact with an HSP for child immunization and/or
with a recent episode of illness (3 weeks) for children and adults or/and
with a birth in the last 12 months or/and
with a pregnant woman
Whenever possible, we will interview all adults who slept in the home the previous night. Questions about children and under 18 who also slept in the home the previous night will be asked of the child’s primary caregivers. A minimum of 100 adults per cluster will be interviewed. The number of households visited will be adjusted accordingly but should be around 50. Finally, during Phase 3, all health staff of the 10 local HSPs will be administered the HSP questionnaire.
For the qualitative phase 4, a nested sub-sample of households and HSP will be selected from among those who responded to the Healthcare Service Provider and Household Questionnaires in Phase 3, for in-depth qualitative interviews with HSPs (n = 16) and users and non-users (n = 40). These numbers are approximate as they will be determined based on the preliminary results of the quantitative analyses.
We will recruit a sample of 500 households per site (N = 4 × 500). Sample size calculations are based on the main outcome variable, scale for self-rated health . Our targeted sample of 2000 households would yield a minimal detectable size of 12% at 80% power for a type I error (α) of 5%  using a linear regression. Such effect size is comparable or below to estimates reported in the literature [42, 43].
A research data management plan will be created to help researchers manage the data as part of their research activity or project as well as to facilitate the sharing of data at the end of the project at https://dataverse.ird.fr.
All interviews and focus group discussions will be digitally recorded, transcribed and coded using NVivo software. A code tree will be determined based on the conceptual framework. New codes may be added if new dimensions emerge during the process. We will perform two types of analyses: a framework analysis  to highlight empirical data and a more inductive analysis to allow a grounded understanding beyond our initial conceptual framework.
All data will be collected on tablets and immediately transmitted to a secure cloud. Descriptive statistical analyses and regression analyses will be used to describe the determinants of the resilience scores of HSPs. Furthermore, we will use propensity score matching methods  to assess the effect of (im)mobilities on health and healthcare access and to check for covariate balance between (im)mobilities settings groups, and (non) recent access to HSPs These methods are applicable for any generalized linear model form (logistic, probit, log linear, Poisson) depending on the outcome measure’s distribution and the desired effect measure scale. Life-event analyses (sequence analysis and cox model for example) will be also used to analyse changes in health and mobility during life path .
Concerning secondary data (SARA and DHMIS), the health systems resilience analysis will be carried out in relation to their capacity to maintain the use of care by populations. We will rely on DHMIS data to conduct interrupted time series analyses (ITSA) or difference-in-differences methods (when control groups will be available) - using the main events identified during the exploratory phase as interruptions – to examine changes in health service utilisation indicators over time . Like Odhiambo et al. , we will seek to collectively determine thresholds for each of the dimensions of our resilience scale (e.g. maintaining function as at least 80% of the 14 MNCH services maintained or improved their indicator values). We will rely on SARA survey data (both previously collected data and own data) to analyse changes in health service delivery preparedness over time and related those to the occurrence of CIEC. We will also use SARA survey data to generate explanatory variables that combine with our provider survey data to explain variation in resilience measures. These analyses will rely on the application of fixed-effects regression models. Statistical analyses will be performed with the R and RStudio software.
The analysis of qualitative and quantitative data will be articulated according to an integrated plan ; the data analyses from the qualitative phase (Phase 2) will enhance the questionnaires from the quantitative phase (Phase 3), and the data from the qualitative phase (Phase 4) will inform Phase 3 in more detail. The integration of quantitative and qualitative data will first help analyse the vulnerability of health systems and populations in the context of changes, events and risks, focusing on climate-induced environmental change and then provide a better understanding of the concept of resilience.
Ethics approvals have been granted from the Institutional Review Board (IRB) of the BRAC James P Grant School of Public Health, BRAC University (ref: IRB-19 November’20–050) in Bangladesh and from the Comité National de Bioéthique in Haiti (ref: 2021–10). We will not collect biological data. We will ensure usual considerations concerning fieldwork ethics and research involving humans. However, in Haiti, given the complex socio-political context, an advisory committee including national researchers, civil society representatives and health system focal point from both selected sites, will be set up to discuss issues of health ethics principles, climate change  and violence, in particular. Besides, this committee will provide advice and guidance in data collection in difficult areas. If conditions for quantitative data collection become too challenging, a new protocol, involving only qualitative research and a maximum of online/telephonic qualitative interviews, will be proposed. It will include the qualitative phase described here (phase 2), followed by another qualitative phase (new phase 3). This updated phase 3, the design of which will depend on the results of phase 2, will establish a typology of events and individual, household and community strategies when facing the events, since the phase 2. If necessary and possible, a third qualitative phase following similar patterns will follow.
Participation to the study will be voluntary. We will inform participants about the research results’ intended publications, information on data use and sharing. We will also inform the participants about their right to withdraw at any time without consequences and about their right to refuse to answer. For the quantitative data collection, the consent form will be completed for all participants or those legally allowed to consent on their behalf. Because being underaged, the consent forms will be asked to the adolescents and their caregivers. Questions about children will be asked to their caregivers. We will interview participants in an adequate setting allowing confidentiality. We will replace the respondents’ names by ID codes that will not appear in the public databases. Once all data are imputed and ready to use, the key, associating names and ID codes, will be destroyed. For the other data collection, depending on the people interviewed and the local issues at stake, written or verbal consent will be systematically requested from participants. The usual ethical criteria for qualitative health research will be respected. Methodology concerning consents was approved by both ethics’ committees in Bangladesh and Haiti. A pilot study will assess the cultural sensitivity of some questions and adjustments will be decided with the local teams. Most of the sensitive questions have already been tested for the SARA/DHS surveys. Enumerators are trained to appropriate policy and practice concerning confidentiality, anonymity or acknowledgement of research participants. Feedback to participants on the research results as appropriate will be planned.
Knowledge transfer will take place throughout the project, in close collaboration with stakeholders. Support will be offered in each of the two countries to develop and implement a knowledge transfer plan. These plans will be adjusted throughout the project according to its evolution. The plans will (at least) include the presentation of results through research snapshots and policy briefs  in English and national languages. We will also disseminate our finding at the national, regional and local Health Authorities through deliberative workshops  informal meetings, presentations, and policy briefs, which will have been reviewed by stakeholders for content and presentations  and also to other countries struggling with climate change, through policy briefs as well as through international and regional networks. Scientific communication is planned through publications and presentations at national and international conferences. Part of the material, such as research snapshots, policy briefs, evaluation protocols, videos and evaluation findings will be publicly available (after publication for any results dissemination) through the project website.
Participatory visual anthropology will also be part of our knowledge transfer strategy. According to Baumann’s methodology , we will facilitate the creation of participatory videos in different communities. These videos will visually represent the complex issues of migrant access to healthcare, advocate for an embodied and bottom-up approach to healthcare issues and help communities to dialogue with public decision-makers. We will develop a Memorandum of Understanding with project participants to include project objectives based on stakeholders and consent to film and/or be filmed. In addition, a workshop will be organised between the viewing of the edited film and the broadcast to obtain consent. Consent will also be the subject of in-depth discussions on image rights, broadcasting rights and signing of a form by participants and researchers. A specific research protocol for these videos will be published.
Global changes, such as climate changes and variability, (im)mobilities and pandemics, are part of the most prominent contemporary health challenges. However, very little has been done to study health systems resilience and climate change . Due to these systems’ complex nature and mechanisms of these systems, an interdisciplinary approach is needed . Here, we present, an innovative and integrative conceptual framework associated with an interdisciplinary mixed-method research design.
We based our conceptual framework on two complementary frameworks: one on healthcare access and the other on the resilience process. The Levesque’s Conceptual Framework of Access to Health  advantages reside in its two-sides approach, which focuses on the interactions between healthcare service providers and populations, i.e. between demand and supply. More specifically, it illustrates the interactions between the five dimensions related to the health system (approachability, acceptability, availability and accommodation, affordability et appropriateness) and the five capacities related to individuals or households (abilities to perceive, to seek to reach to pay and to engage) during the care pathways. According to a scoping review on the use of Levesque’s framework, since its conceptualization in 2013, it has been used successfully in empirical researches in various healthcare services and settings. It allows “to comprehensively assess the complex and dynamic process of healthcare access” . In the context of universal health coverage (UHC), we consider that access to care is the proximal outcome of health systems’ resilience but more is needed considering the mechanisms of the resilience processes. For this reason, we decided to combine the Levesque’s framework  to the FCDO framework  to include a dynamic approach of healthcare access while considering the context, the disturbance, the capacity to deal with disturbance, and the reaction to it. This choice allows an innovative, integrative and systemic approach of resilience, while considering contexts and vulnerabilities [10,11,12]. We chose an interdisciplinary mixed-method exploratory protocol for the framework operationalisation . This multi-step protocol optimises primary and secondary data use, for a transversal and longitudinal approach within a 2-year timeframe in both countries. The multi-sites qualitative and quantitative data collection, with a sequential exploratory/explanatory design, permits an understanding and adjustments to local contexts. Spatial analysis, multivariable and multi-level analyses (individual, household, health provider, cluster, and country levels) will be conducted to understand local context variations.
Due to events uncertainty in the study sites in such a short time-frame and the long span of the Covid-19 pandemic, we ruled out a second data collection for the longitudinal approach. We preferred an approach based on secondary data, supplemented with our primary data.
During the elaboration of the conceptual framework, we encountered several difficulties. First, we focused on climate change, as it is often overlooked, yet non negligeable [52, 54]. The link between climate change and climate events is a complex and controversial topic [55,56,57]. However, there is growing evidence that extreme weather events are amplified by climate change, as shown by the development of the climate attribution field [58, 59]. For this reason, we decided to focus on climate-related events, such as drought, flood and cyclone. Besides, because other changes and events might be closely intertwined with climate, our methodology allows us to include other global and local events, changes and risks impacting local health systems and populations . We will list, for each site, events, changes and risks affecting locally populations and HSPs. As our research is taking place in the context of the COVID-19 pandemic, we will also be attentive to the relative perception of this shock compared to others for health systems.
Second, as a result of our preliminary literature review, no consensus amongst scholars on ‘climate-induced migration’ definitions, conceptual approaches or frameworks are to be found in the literature [60, 61]. This is mostly due to the complex and indirect causal links between climate change and mobility. Indeed, Homer-Dixon  states that “isolating individual causal agents of population movement as entirely environmental inappropriately emphasizes one cause amongst an array of overlapping and interconnected processes” and for Black  “conceptualizing environmental change as a primary cause of displacement is unhelpful, intellectually unsound, and unnecessary”. Rather than focusing on the nexus climate change-migration [64, 65] which would have been too complex to define and could be controversial, we have preferred to include all migration forms. Indeed, all types of migration might influence health systems - whether at the HSP level or at the level of individuals, families and communities.
Data collection in times of uncertainty is complicated to plan. We are working in two countries with different contexts in a time of pandemic. Currently, fieldwork in Haiti is extremely complicated and costly for security and access reasons. This is leading us to take into account the ethics of research in a context of repression and reduction of individual freedoms and human rights. We decide to put in place additional measures, such as the creation of a local advisory committee on health ethics principles in the context of climate change  and violence that will be active throughout the project, over and above what is usually required. Secondary data availability is a challenge for country comparison at the local level. Thus, we will focus mostly on longitudinal and transversal analyses inside each country, and country comparison will be made when possible.
Conclusion and expected outcomes
The CLIMHB project aims to provide an evidence base for research and policy development for health systems resilience in the context of global changes, in particular climate-related events, population mobility, pandemics and their intersections. This interdisciplinary project’s originality lies in its combined approach to the health system’s process and output in the context of global changes and its local impacts, i.e. access to care and resilience, at both institutional and population levels. This project will contribute to the developing field of research on LMIC health systems and climate change, as prescribed by Sustainable Development Goals 13 (climate change) and 3 (health) . Besides, this project will suggest operational solutions and lessons learned from decision-makers from the national level to the community level. Finally, it will support communities’ discussions around health services access, health practices and overall communities’ social and political organisation.
Availability of data and materials
Secondary data are available on the DHS website: https://dhsprogram.com/data/available-datasets.cfm
Climate-induced environmental change
- CLIMHB project:
Climate Change, Migration and Health Systems Resilience in Haiti and Bangladesh project
Department for International Development
District Health Management Information Systems
(UK) Foreign, Commonwealth and Development Office
Health service provider/healthcare service provider
Health System Resilience/Healthcare System resilience
Interrupted Time Series Analyses
Low- and middle-income countries
Maternal, Newborn and Child Health
Service Availability and Readiness Assessment
Sustainable Development Goal
Universal Health Cover
World Health Organization
Saulnier DD, Blanchet K, Canila C, Cobos Muñoz D, Dal Zennaro L, de Savigny D, et al. A health systems resilience research agenda: moving from concept to practice. BMJ Glob Health. 2021;6(8):e006779.
Turenne CP, Gautier L, Degroote S, Guillard E, Chabrol F, Ridde V. Conceptual analysis of health systems resilience: a scoping review. Soc Sci Med. 2019 Jul;232:168–80.
Thomas S, Sagan A, Larkin J, Cylus J, Figueras J, Karanikolos M. Strengthening health systems resilience: key concepts and strategies [internet]. Copenhagen (Denmark): European Observatory on Health Systems and Policies; 2020 [cited 2021 Feb 10]. (European Observatory Policy Briefs). Available from: http://www.ncbi.nlm.nih.gov/books/NBK559803/
Watts N, Amann M, Arnell N, Ayeb-Karlsson S, Beagley J, Belesova K, et al. The 2020 report of the lancet countdown on health and climate change: responding to converging crises. Lancet. 2020;S014067362032290X.
WHO, World Bank. New report shows that 400 million do not have access to essential health services [Internet]. Geneva, New York City; 2015 Juin. Available from: https://www.who.int/mediacentre/news/releases/2015/uhc-report/en/
Tulloch O, Machingura F, Melaned C. Health, migration and the 2030 Agenda for Sustainable Development [Internet]. 2016 Jul. Available from: https://doc.rero.ch/record/308984/files/27-7._ODI_Health.pdf
Ridde V, Aho J, Ndao EM, Benoit M, Hanley J, Lagrange S, et al. Unmet healthcare needs among migrants without medical insurance in Montreal. Canada Global Public Health. 2020;15(11):1603–16.
Stenberg K, Hanssen O, Edejer T, Bertram M, Brindley C, Meshreky A, et al. Financing transformative health systems towards achievement of the health sustainable development goals: a model for projected resource needs in 67 low-income and middle-income countries. Lancet Global Health. 2017;5(9):E875–87.
Foresight. Foresight: Migration and Global Environmental Change - Future challenge and opportunity- Final Project Report [Internet]. London: The Government Office for Science; 2011 [cited 2021 Feb 15]. Available from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/287717/11-1116-migration-and-global-environmental-change.pdf
Pfadenhauer LM, Gerhardus A, Mozygemba K, Lysdahl KB, Booth A, Hofmann B, et al. Making sense of complexity in context and implementation: the context and implementation of complex interventions (CICI) framework. Implementation Sci. 2017;12(1):21.
Craig P, Di Ruggiero E, Frohlich KL, Mykhalovskiy E, White M, on behalf of the Canadian Institutes of Health Research (CIHR)–National Institute for Health Research (NIHR) Context Guidance Authors Group (listed alphabetically), et al. Taking account of context in population health intervention research: guidance for producers, users and funders of research [Internet]. 2018 Apr [cited 2021]. Available from: https://www.journalslibrary.nihr.ac.uk/nihr-research/canadian-institutes-of-health-research-cihr-and-nihr-collaboration.htm
Shoveller J, Viehbeck S, Di Ruggiero E, Greyson D, Thomson K, Knight R. A critical examination of representations of context within research on population health interventions. Crit Public Health. 2016;26(5):487–500.
Gilson L, Schneider H. Understanding health service access: concepts and experience. In: Global Forum Update on Research for Health; 2007. p. 28–32
Witter S, Ensor T, Jowett M, Thompson R. Health economics for developing countries: a practical guide. Amsterdam: KIT Publishers; 2000. p. 296.
Levesque J-F, Harris MF, Russell G. Patient-centred access to health care: conceptualising access at the interface of health systems and populations. Int J Equity Health. 2013;12(1):18.
DFID. Defining Disaster Resilience: A DFID Approach Paper [Internet]. 2011 [cited 2021 Feb 9] p. 20. Available from: https://www.fsnnetwork.org/sites/default/files/dfid_defining_disaster_resilience.pdf
Masson-Delmotte V, Zhai P, Pörtner H-O, Roberts D, Skea J, Shukla PR, et al. Global Warming of 1.5°C. An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty [Internet]. 2018 p. Annex I: Glossary. Available from: https://www.ipcc.ch/site/assets/uploads/sites/2/2019/06/SR15_AnnexI_Glossary.pdf
World Health Organization, editor. Monitoring the building blocks of health systems: a handbook of indicators and their measurement strategies. Geneva: World Health Organization; 2010. 92 p.
Van Olmen J, Criel B, Van Damme W, Marchal B, Van Belle S, Van Dormael M, et al. Analysing health systems to make them stronger. Antwerp, Belgium: ITGPress; 2010.
Kreft S, Eckstein D, Dorsch L, Fischer L, Germanwatch. Global Climate Risk Index 2016 Who Suffers Most From Extreme Weather Events? Weather-related Loss Events in 2014 and 1995 to 2014. 2015.
Brammer H. Floods, cyclones, drought and climate change in Bangladesh: a reality check. Int J Environ Stud. 2016 Nov;73(6):865–86.
Penning-Rowsell EC, Sultana P, Thompson PM. The ‘last resort’? Population movement in response to climate-related hazards in Bangladesh. Environ Sci Pol. 2013 Mar;27:S44–59.
USAID. Climate Risk Profile: Haiti [Internet]. 2017. Available from: https://www.climatelinks.org/sites/default/files/asset/document/2017_Cadmus_Climate-Risk-Profile_Haiti.pdf
United Nations. Le Conseil De Sécurité Prend Le Pouls Du Processus Électoral En Haïti Tout En S’inquiétant Du Climat De Tension Et De La Situation Humanitaire [Internet]. 2021 Feb [cited 2021 Feb 23]. Available from: https://www.un.org/press/fr/2021/sc14442.doc.htm
Bureau intégré des Nations Unies en Haiti. Rapport du Secrétaire général [Internet]. Conseil de Securité des Nations Unies; 2021 février [cited 2021 Feb 26] p. 16. Report No.: S/2021/133. Available from: https://binuh.unmissions.org/sites/default/files/s_2021_133_-_sg_report_on_haiti_-_french.pdf
The editorial board of NYTimes. Haiti Needs Help. This Is What the U.S. Sends Instead. New York Times [Internet]. 2021 Feb 10 [cited 2021 Feb 27]; Available from: https://www.nytimes.com/2021/02/10/opinion/haiti-us-deportations.html
Hoffmann R, Dimitrova A, Muttarak R, Crespo Cuaresma J, Peisker J. A meta-analysis of country-level studies on environmental change and migration. Nat Clim Chang. 2020;10(10):904–12.
Black R. Environmental refugees : myth or reality ? New issues in refugee research; 2001.
Perch-Nielsen SL, Bättig MB, Imboden D. Exploring the link between climate change and migration. Clim Chang. 2008;91(3):375–93.
Lozano R, Fullman N, Mumford JE, Knight M, Barthelemy CM, Abbafati C, et al. Measuring universal health coverage based on an index of effective coverage of health services in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2020 Oct;396(10258):1250–84.
IDMC. Bangladesh Country Information [Internet]. [cited 2021 Feb 9]. Available from: https://www.internal-displacement.org/countries/bangladesh
Dey NC, Parvez M, Saha R, Islam MR, Akter T, Rahman M, et al. Water quality and willingness to pay for safe drinking water in Tala Upazila in a Coastal District of Bangladesh. Expo Health. 2019;11(4):297–310.
Tareq SM, Tauhid M, Rahman MTU, Syed Z, Uddin D, Uddin J, et al. Waterlogging threaten biodiversity along the South-West coastal region of Bangladesh 2016;4:203–210.
Neelormi S, adri N, Ahmed AU. Gender dimensions of differential health effects of climate change induced water-logging: A case study from coastal Bangladesh. IOP Conf Ser: Earth Environ Sci. 2009 Feb 1;6(14):142026.
McNamara KE, Olson LL, Rahman MA. Insecure hope: the challenges faced by urban slum dwellers in Bhola slum. Bangladesh Migration and Development. 2016;5(1):1–15.
International Centre for Diarrhoeal Disease Research, Bangladesh, Arbeitsgemeinschaft Deutscher Betriebsingenieure, and Sweden. Slum Health in Bangladesh: Insights for Health and Demographic Surveillance [Internet]. Dhaka, Bangladesh: icddr,b; 2019 [cited 2021 Feb 10] p. 181. Report No.: 154. Available from: http://dspace.icddrb.org/jspui/bitstream/123456789/9298/1/icddrb-SP154.pdf
IDMC. Haiti Country Information [Internet]. Haiti Country Information. Available from: https://www.internal-displacement.org/countries/haiti
Creswell JW, Plano Clark VL. Designing and conducting mixed methods research. Third edition, international student edition. Los Angeles London New Delhi Singapore Washington DC Melbourne: Sage; 2018. 492 p.
Palinkas LA, Horwitz SM, Green CA, Wisdom JP, Duan N, Hoagwood K. Purposeful sampling for qualitative data collection and analysis in mixed method implementation research. Admin Pol Ment Health. 2015;42(5):533–44.
Sanchez-Santos MT, Zunzunegui MV, Otero-Puime A, Cañas R, Casado-Collado AJ. Self-rated health and mortality risk in relation to gender and education: a time-dependent covariate analysis. Eur J Ageing. 2011;8(4):281–9.
Hulley SB, editor. Designing clinical research: an epidemiologic approach. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2001. 336 p.
Nielsen SS, Krasnik A. Poorer self-perceived health among migrants and ethnic minorities versus the majority population in Europe: a systematic review. Int J Public Health. 2010 Oct;55(5):357–71.
Kondo N, Sembajwe G, Kawachi I, van Dam RM, Subramanian SV, Yamagata Z. Income inequality, mortality, and self rated health: meta-analysis of multilevel studies. BMJ. 2009 Nov 10;339(nov10 2):b4471–b4471.
Ritchie J, Spencer L. Qualitative Data Analysis for Applied Policy Research. In: The Qualitative Researcher’s Companion [Internet]. 2455 Teller Road, Thousand Oaks California 91320 United States of America: SAGE Publications, Inc.; 2002 [cited 2021 Feb 9]. p. 305–29. Available from: http://methods.sagepub.com/book/the-qualitative-researchers-companion/n12.xml
Austin PC. The relative ability of different propensity score methods to balance measured covariates between treated and untreated subjects in observational studies. Med Decis Mak. 2009 Nov;29(6):661–77.
Gosselin A, Desgrées du Loû A, Lelièvre E. How to use sequence analysis for life course epidemiology? An example on HIV-positive Sub-Saharan migrants in France. J Epidemiol Community Health. 2018 Jun 1;72(6):507.
Odhiambo J, Jeffery C, Lako R, Devkota B, Valadez JJ. Measuring health system resilience in a highly fragile nation during protracted conflict: South Sudan 2011–15. Health Policy Plan. 2020;35(3):313–22.
Singh JA. Why human health and health ethics must be central to climate change deliberations. PLoS Med. 2012;9(6):e1001229.
Dagenais C, Ridde V. Policy brief as a knowledge transfer tool: to “make a splash”, your policy brief must first be read. Gac Sanit. 2018 May;32(3):203–5.
Ridde V, Dagenais C. What we have learnt (so far) about deliberative dialogue for evidence-based policymaking in West Africa. BMJ Glob Health. 2017 Dec;2(4):e000432.
Baumann SE. Collaborative filmmaking: a participatory, visual research method - Sara E. Baumann, Pema Lhaki, Jessica G. Burke, 2020 [Internet]. 2020 [cited 2020 Nov 12]. Available from: https://journals.sagepub.com/doi/10.1177/1049732320941826
Ridde V, Benmarhnia T, Bonnet E, Bottger C, Cloos P, Dagenais C, et al. Climate change, migration and health systems resilience: Need for interdisciplinary research. F1000Res. 2019;8:22.
Cu A, Meister S, Lefebvre B, Ridde V. Assessing Healthcare Access Using the Levesque’s Conceptual Framework– A scoping Review. under review;
Marten R, Yangchen S, Campbell-Lendrum D, Prats EV, Neira MP, Ghaffar A. Climate change: an urgent priority for health policy and systems research. Health Policy and Planning. 2020 Dec 21;czaa165.
Jentsch A, Kreyling J, Beierkuhnlein C. A new generation of climate-change experiments: events, not trends. Front Ecol Environ. 2007 Sep;5(7):365–74.
Trenberth KE, Fasullo JT, Shepherd TG. Attribution of climate extreme events. Nature Clim Change. 2015 Aug;5(8):725–30.
Stott P. How climate change affects extreme weather events. Science. 2016 Jun 24;352(6293):1517–8.
Ornes S. Core concept: how does climate change influence extreme weather? Impact attribution research seeks answers. Proc Natl Acad Sci U S A. 2018;115(33):8232–5.
Swain DL, Singh D, Touma D, Diffenbaugh NS. Attributing extreme events to climate change: a new frontier in a warming world. One Earth. 2020 Jun;2(6):522–7.
Nicholson CTM. Climate change, forced migration, and international law. By Jane McAdam. J Refug Stud. 2013;26(2):311–3.
Nicholson CTM. Climate change and the politics of causal reasoning: the case of climate change and migration: climate change and the politics of causal reasoning. Geogr J. 2014;180(2):151–60.
Homer-Dixon TF. Environmental scarcities and violent conflict: evidence from cases. Int Secur. 1994;19(1):5.
Black R. Fifty years of refugee studies: from theory to policy. Int Migr Rev. 2001;35(1):57–78.
Ridde V. Climate migrants and health promotion. Glob Health Promot. 2018;25(1):3–5.
Cloos P, Ridde V. Research on climate change, health inequities, and migration in the Caribbean. The Lancet Planetary Health. 2018;2(1):e4–5.
Whitmee S, Haines A, Beyrer C, Boltz F, Capon AG, de Souza Dias BF, et al. Safeguarding human health in the Anthropocene epoch: report of the Rockefeller Foundation–Lancet commission on planetary health. Lancet. 2015;386(10007):1973–2028.
The co-authors of this manuscript would like to thank Anthony Cu for his review of the literature, Shabab Haider, Suhi Hanif and Rafee Tamjid for their support in developing the successive versions of this protocol and Emilie Dubois, from the Impakt Scientifik agency (Agence IMPAKT Scientifik, https://www.impaktsci.co) for the graphic design of the conceptual framework. Finally, we thank the reviewers for their questions and comments.
Funded by the French National Research Agency (ANR) as part of the presidential call “Make Our Planet Great Again” (MOPGA). The mental health study is co-financed by the Centre National de Ressources et de Resilience.
Ethics approval and consent to participate
Ethics approvals have been granted from the Institutional Review Board (IRB) of the BRAC James P Grant School of Public Health, BRAC University (ref: IRB-19 November’20–050) in Bangladesh and from the Comité National de Bioéthique in Haiti (ref: 2021–10). For the quantitative data collection, the consent form will be completed for all participants or those legally allowed to consent on their behalf. For the other data collection, depending on the people interviewed and the local issues at stake, written or verbal consent will be systematically requested from participants. The usual ethical criteria for qualitative health research will be respected. Methodology concerning consents was approved by both ethics’ committees in Bangladesh and Haiti.
Consent for publication
The authors declare no conflict of interest.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Quantitative tool: quantitative questionnaire for the population. QuantiPop.
Quantitative tool: questionnaire for the health service providers. QuantiHSP.
Qualitative tool: interview guide for the population. QualPop.
Qualitative tool: interview guide for the health service providers. QualHSP.
About this article
Cite this article
Clech, L., Meister, S., Belloiseau, M. et al. Healthcare system resilience in Bangladesh and Haiti in times of global changes (climate-related events, migration and Covid-19): an interdisciplinary mixed method research protocol. BMC Health Serv Res 22, 340 (2022). https://doi.org/10.1186/s12913-021-07294-3
- Healthcare systems resilience
- Global change
- Climate change