A Meta-Synthesis of Policy Recommendations Regarding Human Mobility in the Context of Climate Change

Changing mobility patterns combined with changes in the climate present challenges and opportunities for global health, requiring effective, relevant, and humane policy responses. This study used data from a systematic literature review that examined the intersection between climate change, migration, and health. The study aimed to synthesize policy recommendations in the peer-reviewed literature, regarding this type of environmental migration with respect to health, to strengthen the evidence-base. Systematic searches were conducted in four academic databases (PubMed, Ovid Medline, Global Health and Scopus) and Google Scholar for empirical studies published between 1990–2020 that used any study design to investigate migration and health in the context of climate change. Studies underwent a two-stage protocol-based screening process and eligible studies were appraised for quality using a standardized mixed-methods tool. From the initial 2425 hits, 68 articles were appraised for quality and included in the synthesis. Among the policy recommendations, six themes were discernible: (1) avoid the universal promotion of migration as an adaptive response to climate risk; (2) preserve cultural and social ties of mobile populations; (3) enable the participation of migrants in decision-making in sites of relocation and resettlement; (4) strengthen health systems and reduce barriers for migrant access to health care; (5) support and promote optimization of social determinants of migrant health; (6) integrate health into loss and damage assessments related to climate change, and consider immobile and trapped populations. The results call for transformative policies that support the health and wellbeing of people engaging in or affected by mobility responses, including those whose migration decisions and experiences are influenced by climate change, and to establish and develop inclusive migrant healthcare.


Introduction
Globally, millions of people move in response to or in anticipation of environmental stress every year, and climate change is becoming more important in their decision to migrate.

Definitions
In this review, we used the World Health Organisation's (WHO) definition of 'health' as a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity [28] and extended this well-known definition to include the health determinants of food and water security.
We defined 'climate change' for this review as a change in the state of the climate that can be identified by changes in the mean and/or the variability of its properties, and that persists for an extended period, typically decades or longer [29].
We defined 'migration' as an overarching concept covering diverse human mobilities. More specifically, as the movement of persons away from their place of habitual place of residence, either across an international border or within a state regardless of legal status, degree of choice, causes of the movement, or length of stay [30]. Possible climate-related mobility responses comprise forced displacement, migration, and planned relocation [31] as well as immobility. Broadly, forced displacement refers to contexts of forced or involuntary movement of people within or across borders; migration refers to the movement of people within or across borders and has an element of choice; and planned relocation refers to the organised movement of people, typically with government support [19]. Population immobility in contexts of climate risks is increasingly discussed [31], yet lacks definitional clarity: it variously refers to 'trapped' populations that are unable to move away from sites of climate risk, and voluntary immobility where people are unwilling to leave their homes despite climate risks [32].

Search Strategy and Eligibility Criteria
This systematic literature view includes studies exploring climate variability and any climate hazard or extreme weather event that could be plausibly linked to climate change. Studies were eligible if health was measured or reported, including as a consideration in migration decision-making, or as an impact at origin, en route, or destination. The study populations include individuals, households, and whole communities who engaged in, or were affected by climate-related mobility: i.e., sending communities, mobile people and populations, and host communities. Importantly, studies explicitly addressing immobility in contexts of climate risk are also included. Taken together, this review includes empirical evidence about health and migration in the context of a changing climate. Given that this nexus is interdisciplinary, the included articles originate from diverse research fields such as public health, demography, policy studies, climatology, human geography, and international relations.
With this background and through consultation with a subject librarian, we selected four academic databases: CABI Direct-Global Health (1973 to present); Ovid Medline (1946 to present); PubMed (1966 to present), and Scopus (1970 to present). In addition, we searched Google Scholar applying the same search terms. The test searches we conducted demonstrated that these databases captured the most relevant articles to answer the review question. These databases were complementary, rather than duplicative and enabled an exhaustive search. Reference lists of selected articles were also searched for relevant articles. Grey literature (likely to cover conceptual papers and editorials) was not included in this systematic literature review, because it focused on peer-reviewed research including empirical data. The search strategy was designed to capture primary studies included in important grey literature reports. Four independent authors used the custom systematic literature review software Covidence (Covidence, Melbourne, Australia.) (PNS, JS, KB, CM) to complete the two-stage screening process.
Three concepts formed the basis of the systematic searches using a Population, Exposure, Outcome (PEO) approach (modified PICO). Specifically, Boolean operators "AND" and "OR" were applied to synonyms for the key search teams Climate Change, Migration, and Health as outlined in Table 1. We included Medical Subject Headings (MeSH) and free text in the searches. The timeframe was chosen because few relevant articles were identified prior to 1990 in numerous test searches and the review aimed to capture contemporary literature. The search for this manuscript was updated in August 2020. Inclusion and Exclusion criteria are outlined in Table 2. The population studied needed to be engaged in or affected by a mobility response and was not limited to 'mobile populations' but included sending and host communities. The mobility response could be any type of population movement at the individual, household, or community level including forced displacement, planned relocation, and, also immobility. The exposure had to be linkable to climate change or climate variability. Other environmental hazards (such as geophysical hazards) were excluded. Studies were not excluded due to poor quality nor study type. At least one health outcome (direct or indirect) needed to be included in the results. Social determinants of health (e.g., education) were not considered a health outcome. Only empirical peer-reviewed studies were included within the timeframe, published in English or German.

Literature Selection
Studies were selected through a two-stage blinded process requiring two independent votes to progress to the next stage. The custom software Covidence was used for study selection and captured all decisions and notes of the reviewers regarding how the studies met or did not meet the criterion. The first stage was title and abstract. The second was full-text screening. Disagreements regarding whether to include a study or not were resolved by a third independent reviewer. Further studies were identified by scanning reference lists of relevant articles. The screening process is documented in Figure 1. The main reasons for exclusion in both phases were that the study was not empirical (i.e.,-editorial) and the research did not include all three concepts but rather focused on just one or two. Interestingly, another reason for exclusion was that the health or migration was not human and related to a plant or animal indicating that migration and health issues in the context of climate change pertain to all living things.

Data Extraction
The extraction codes were the recommendations for policy, practice, and further research included in the selected studies. It became clear during the extraction process that the research recommendations were partly dated and had somewhat been filled. Whilst the recommendations for research were extracted and coded, they are not included in this meta-synthesis. We focussed on the recommendations for policy and practice to give practical insights and to meet the objectives of this review. Developing guidance for a potential research strategy on this topic was seen as beyond the

Data Extraction
The extraction codes were the recommendations for policy, practice, and further research included in the selected studies. It became clear during the extraction process that the research recommendations were partly dated and had somewhat been filled. Whilst the recommendations for research were extracted and coded, they are not included in this meta-synthesis. We focussed on the recommendations for policy and practice to give practical insights and to meet the objectives of this review. Developing guidance for a potential research strategy on this topic was seen as beyond the scope of this review.

Quality Assessment
This is the second manuscript from an updated systematic literature review, which presented different findings [33]. Therefore, the quality appraisal was complete for 50 of the included studies identified in the original search (December 2018). The additional 18 studies, identified in the updated search (August 2020), were appraised for quality using the same MMAT tool [34]. MMAT sets five quality appraisal questions for five study types: mixed methods, quantitative descriptive, quantitative non-randomised, quantitative randomised control trials, and qualitative studies. Since MMAT cannot be applied to modeling studies, five studies included in this review were not appraised. This did not make a difference to the results because studies were not excluded based on poor quality. The MMAT tool appraises both study quality itself and the reporting.

Data Analysis and Meta-Synthesis
Recommendations for policy and practice in the selected studies were presented in narrative form (text) and were heterogeneous. Therefore, thematic analysis for meta-synthesis was an appropriate analysis method through which to derive common meaning from the various studies. No selected studies were excluded from the analysis as all 68 studies included some form of recommendation. The extracted recommendations (the data set) were uploaded to N-VIVO (Version 12) and coded. We used Braun and Clarke (2006) to thematically analyse the recommendations and identify themes across the dataset [35]. This involved data familiarisation, code generation, theme searches (multiple), theme review, and defining themes. We used a meta-aggregative approach to the analysis that avoids re-interpreting the recommendations and inductive reasoning [36]. All authors were involved in and approved the thematic analysis that resulted in six themes.

Included Studies
From 2425 studies identified by the search strategy, 68 eligible studies were included in the analysis ( Table 3). The number of relevant studies increased significantly from 2012 with 82% of studies published in the last 8 years (Figure 2). The main reason for exclusion was due to the study not focusing on the nexus between climate change, migration, and health, but rather limited to climate change and migration (without health) or climate change and health (without migration). Studies were also excluded because they did not present empirical evidence, or investigated health and migration of animal or plant species. The original systematic literature review search was first conducted in December 2018 and a publication focused on another data-set was published in Environmental Review Letters [33]. The search was updated in August 2020, capturing 18 additional studies using the same protocol. This publication focuses on the recommendation for policy, practice, and further research extracted from the included studies. Qualitative case studies **** Qualitative; Case study *****  Mixed Methods *** Studies were also excluded because they did not present empirical evidence, or investigated health and migration of animal or plant species. The original systematic literature review search was first conducted in December 2018 and a publication focused on another data-set was published in Environmental Review Letters [33]. The search was updated in August 2020, capturing 18 additional studies using the same protocol. This publication focuses on the recommendation for policy, practice, and further research extracted from the included studies.

Quality Appraisal
Overall, the quality of the included studies was high with the quantitative non-randomised studies (n: 12) achieving the highest quality assessment result, with 67% scoring between 'good' and 'very high' and no studies being appraised as poor (Figure 4). The quality of quantitative descriptive studies (n: 17) was also high overall, with 71% of studies scoring between 'good' and 'very high'. Overall, the qualitative studies (n: 17) were appraised as high quality with 65% scoring between

Quality Appraisal
Overall, the quality of the included studies was high with the quantitative non-randomised studies (n: 12) achieving the highest quality assessment result, with 67% scoring between 'good' and 'very high' and no studies being appraised as poor (Figure 4). The quality of quantitative descriptive studies (n: 17) was also high overall, with 71% of studies scoring between 'good' and 'very high'. Overall, the qualitative studies (n: 17) were appraised as high quality with 65% scoring between 'good' and 'very high'. The quality of mixed methods studies (n: 17) was also high overall with 65% attaining a 'good' or 'high' quality rating, yet no mixed methods studies achieved a 'very high' quality score according to MMAT.

The Links between Climate Mobility and Health
A range of climate hazards (n: 24) were associated with diverse mobility responses and health outcomes studied, without clear trends or patterns emerging. The climate hazards in the studies were half sudden-onset (n: 12) such as storms, floods, and cyclones, and half slow-onset (n: 12) such as drought, sea-level rise, and glacial retreat. The most studied hazards included floods (n: 19; 16%), rainfall variability (n: 18; 15%), drought (n: 17; 14%) and multiple climate hazards or general climate change (n: 14; 12%). Other common climate hazards included extreme heat, sea-level rise, and hurricanes (n: 6; 5%). The predominant mobility responses were forced displacement, relocation (planned and forced), seasonal migration, and rural-urban migration. The predominant health outcomes studied were food and water security, access to healthcare services, mental health issues, and infectious disease [33].

Thematic Analysis of Policy Recommendations
Six themes were identified in the thematic analysis of policy recommendations extracted from 68 studies. A synopsis of the themes with illustrative quotes are outlined in Table 4.

The Links between Climate Mobility and Health
A range of climate hazards (n: 24) were associated with diverse mobility responses and health outcomes studied, without clear trends or patterns emerging. The climate hazards in the studies were half sudden-onset (n: 12) such as storms, floods, and cyclones, and half slow-onset (n: 12) such as drought, sea-level rise, and glacial retreat. The most studied hazards included floods (n: 19; 16%), rainfall variability (n: 18; 15%), drought (n: 17; 14%) and multiple climate hazards or general climate change (n: 14; 12%). Other common climate hazards included extreme heat, sea-level rise, and hurricanes (n: 6; 5%). The predominant mobility responses were forced displacement, relocation (planned and forced), seasonal migration, and rural-urban migration. The predominant health outcomes studied were food and water security, access to healthcare services, mental health issues, and infectious disease [33].

Thematic Analysis of Policy Recommendations
Six themes were identified in the thematic analysis of policy recommendations extracted from 68 studies. A synopsis of the themes with illustrative quotes are outlined in Table 4. Table 4. Synopsis of themes and illustrative quotes.

Theme
Illustrative Quote 1. Avoid the universal promotion of migration as an adaptive response to climate risks. Prevent forced migration by investing in climate change adaptation, disaster risk reduction and sustainable development.
Consider planned relocation as a last resort.
'The policy implication is that governments should not make assumptions a priori about whether a location is undesirable and promote migration as a blanket solution to the negative impacts of climate change'-Quantitative descriptive, Peru [12].
2. Preserve cultural and social ties of mobile populations. Strengthen governance of socio-ecological systems.
'Developing government frameworks that can draw on the strengths of the community-led approaches to relocation whilst also providing a mechanism for communities to stay intact will be an important step forwards for Small Island Developing States (SIDs) facing these climate pressures'-Qualitative case studies, Alaska & Solomon Islands [40].
3. Enable participation of migrants in their sites of relocation and resettlement. Support the self-sufficiency of both incoming and host communities by supporting new livelihoods, developing social networks, and integrating cultural considerations. 5. Provide migrants with the requirements and the determinants for good health.
'The rights of these displaced people, including the right to health, are often poorly protected in practice. More vigorous application of existing human instruments is needed, as well as clarification and possibly re-definition of the rights of those displaced'-Qualitative study, Bangladesh [89].
6. Integrate health into loss and damage assessments. Consider people at higher risk, including those that are immobile or trapped.
'The findings outlined a long line of climate-induced non-economic losses and damages that people faced through the rural-urban move from the island, and through the displacement in the slum. These included the loss of identity, honour, sense of belonging, physical and mental health or wellbeing'-Mixed Methods, Bangladesh [32].

Discussion
This review deepens our understanding around the complex mechanisms through which climate change impacts contribute to migration and health outcomes. This review reveals some useful guidance for migration and health policy and practice in the context of climate change. The geographical settings where this nexus research has taken place to date, the predominant study designs, their quality, and the relationships between climate hazards, mobility responses, and health outcomes are discussed below. Finally, a narrative synthesis of the policy recommendations extracted from the included studies is placed in the context of the current policy environment on this topic.

Study Settings
The predominant geographical foci of the studies in this review; Sub-Saharan Africa, South Asia, and Latin America. Bangladesh is of particular interest to climate change, migration, and health studies possibly, because it is highly exposed and vulnerable to climate hazards, densely populated.
These regions identified in the research align with existing evidence that indicates that currently, most climate-related migration is internal and takes place in developing countries [31,102]. Three regions in the world are projected to see more than 140 million internal climate-related migrants by 2050, without urgent global and national climate action and economic development [102].

Quality Appraisal and Study Design
Surveys and case reports were a common design of the included quantitative descriptive studies (n: 17). Some of these studies looked at the association between climate hazards, migratory behaviours, and health outcomes and sought to determine the extent to which migration was adaptive or maladaptive. Overall, the quality of quantitative descriptive studies was high with 71% scoring between 'good' and 'very high' (see Figure 4). Analysis ranged from simple descriptive statistics to regression analyses.
Cross-sectional analytical studies were a common quantitative non-randomised study design that compared groups to examine the impact of climate hazards on health (n: 12). These types of studies compared mobile and non-mobile households or communities. Others compared households at variable levels of climate risk concerning health. Overall, the quality of these quantitative non-randomized studies was high with 67% of studies scoring between 'good' and 'very high'. This was the only category of studies without any scoring 'poor' quality. Quality issues included mismatching of temporal-spatial scales of the hazard, mobility response, and health outcomes and inadequate demographic matching of comparison groups.
Case studies and narrative research were prevalent in the selected studies (n: 17). These studies used qualitative methods to unpack health concerns in settings where climate hazards were affecting people's mobility patterns. Overall, the quality was high with 65% of studies scoring between 'good' and 'very high', although these studies often incorporated small sample groups with limited ability to generalise.
Mixed methods (n: 17) studies included in this review commonly combined surveys and interviews or focus groups to explore the experiences of mobile/displaced people facing climate risks, and the consequences for health. The quality was high overall with 65% attaining a 'good' or 'high' rating with no studies achieving excellence according to Mixed Methods Assessment Tool (MMAT) criteria. Poorer scores were related to either the quantitative or qualitative component scoring higher and less transparent integration of results and interpretation of both methods. Overall, a more meaningful and transparent integration of climate data would have more clearly linked the mobility responses and health outcomes to the climate scenario [33].

Climate Mobilities and Health
The relationships between climate change, mobility, and health are complex and connections are population-specific and vary over space and time [33]. This review covers these relationships published in the peer-reviewed literature, which focuses on infectious diseases, access to healthcare, mental health, and food insecurity. While the global burden of disease has shifted to non-communicable diseases (NCD's), NCDs are an under-researched theme with a greater focus on infectious disease in climate-migrant populations. The review also reinforces the recognition that migration does not capture the diverse ways in which people do and do not move in response to a changing climate. Highlighting the need for researchers to challenge the 'climate change causes mass human migration' narrative and to shift attention from climate migration to more diverse forms of climate mobility and indeed immobility [103].

Thematic Analysis of Policy Recommendations
The effect of climate-related mobility on health depends on the policy decisions made by host, home, and transit states and involved organisations, rather than on the mobility itself [104] highlighting the value of evidence-based policy in migration and health governance. In light of this, we extracted and analysed the policy recommendations from all 68 included studies. Six themes were identified that are presented graphically with predominant codes in Figure 5. The themes range from overarching recommendations such as avoiding universal promotion of migration or supporting community participation in migrant health initiatives, to targeted recommendations such as specific climate-change adaptation activities or the importance of preserving social and cultural ties in contexts of climate mobility. Firstly, there were consistent recommendations to avoid universal promotion of migration as an adaptive response to climate risk, to prevent forced migration by investing in climate change adaptation at origin, and to consider relocation only as a last resort [12,43,55,58,70,76,85,97]. These recommendations echo the broader calls by, for example, the International Organization of Migration (IOM) to minimize forced climate-related migration [11] and also the need to avoid assumptions that mobility is inherently positive or negative [73]. Given widespread preferences to remain in sites of belonging, many studies reviewed here called for policy initiatives that enable people to cope with, avoid and prevent the impacts of climate change at origin to prevent forced migration [105].
The included studies made recommendations to support investment in climate change Firstly, there were consistent recommendations to avoid universal promotion of migration as an adaptive response to climate risk, to prevent forced migration by investing in climate change adaptation at origin, and to consider relocation only as a last resort [12,43,55,58,70,76,85,97].
These recommendations echo the broader calls by, for example, the International Organization of Migration (IOM) to minimize forced climate-related migration [11] and also the need to avoid assumptions that mobility is inherently positive or negative [73]. Given widespread preferences to remain in sites of belonging, many studies reviewed here called for policy initiatives that enable people to cope with, avoid and prevent the impacts of climate change at origin to prevent forced migration [105].
The included studies made recommendations to support investment in climate change adaptation and supporting the sustainable development of agriculture generally. Specific suggestions included providing credit facilities and building agricultural extension services-also known as agricultural advisory services-that build knowledge of agronomic techniques and skills to improve productivity, food security, and livelihoods. The focus was clearly on rural farming communities and adaptation with no studies exploring mitigation and few urban settings. This raises some questions about the extent to which the research is focusing on adaptation in a critical period when the mitigation window is closing. Alternatively, this gap may simply reflect that when migration is triggered, there is an urgent adaptation situation in play. In some contexts, existing migration flows and networks may provide an opportunity for investment in the economies of sending communities (via remittances) that can increase opportunities for in situ adaptation and resilience among those who remain.
The second theme highlighted the importance of the preservation of cultural and social ties for the health of mobile populations through preserving and revitalizing traditional solidarity measures [55,62,[74][75][76]94]. Selected articles recommended strengthening governance or socio-ecological systems, favouring community-led approaches, and in doing so, recognizing the agency and inherent resilience of communities and to promote collaborative, adaptive migration governance structures. This agency and resilience focus is predominant in research recommendations yet somewhat lacking in research questions whereby disease and risk factors are the focus, rather than positive health outcomes and protective factors.
The third theme pertained to policy recommendations that sought to enable migrants to participate socially and economically in destination sites. The focus was on employment and income that have clear advantages for the health and well-being of both migrant and host populations as well as sending communities if remittances are mobilized [38,[54][55][56][57]59,65,67,88,89,91,96]. These studies tended to focus on subsistence farmers in rural settings, although many other types of populations will have migration decisions influenced by climate change, and the causal pathways can appear over-simplified. This theme incorporated studies about relocation and resettlement and was the theme with the most focus on host and sending communities rather than purely mobile communities. A well-defined example of promoting self-sufficiency was planned relocation in the Carteret Islands, where there was a recognition that the resettled Carteret families may not have the skills necessary to cultivate kitchen gardens. New arrivals received training inappropriate agricultural techniques, enabling self-sufficiency through income generated from selling cash crops [55].
The fourth theme brought out the recommendations around the need to strengthen health systems generally where migrants are (in both sending communities and destination areas) in terms of both primary health care and more specialized vertical programs such as for HIV and Maternal Child Health (MCH) services. The findings in the selected articles revealed financial, geographic, and cultural barriers for migrants accessing healthcare in the context of climate change and led to recommendations to reduce or remove these barriers to improve migrant health for example by including migrants in health insurance schemes [32,42,46,[59][60][61][62]64,70,73,74,79,[88][89][90]92,100]. There are clear benefits to broader population health from investing in health systems strengthening approach, so the recommendations within this theme would have substantial flow-on benefits to the community at large.
The fifth theme went upstream from quality accessible health services, to identify the importance of ensuring basic requirements for health such as food, water, and shelter, which are basic human rights and necessary to protect life, reduce suffering and preserve human dignity. These studies highlighted the need to focus on health equity in a range of settings including climate-vulnerable regions and sites of relocation and resettlement. This includes the need to integrate migrants into labour markets (see Theme 3) to support livelihoods and food security and to enable access to education [44,45,53,57,64,67,69,72,90,91,94].
This theme corroborates the demand for establishing migrant-inclusive health systems, as suggested by the Lancet Commission on Migration and Health [106]. They constitute the basis for the supply and utilization of patient-centered access to health and social protection. This theme reiterates WHO's Global Action Plan for promoting the health of refugees and migrants under Priority 4. Enhance capacity to tackle the social determinants of health and to accelerate progress towards achieving the Sustainable Development Goals, including universal health coverage.
Finally, the sixth theme focused on the need for policy to integrate health into the full range of loss and damage calculations [32,55,60,61,63,100]. Loss and damage refer to the negative effects of climate variability and change that people are not able to cope with or adapt to. Research and policy discussions of loss and damage recognize that climate change impacts are differential, with greater losses accruing to vulnerable populations and regions, thereby exacerbating inequities [107]. And it is important to note that vulnerability to climate change is fundamentally a matter of political economy, with those least responsible for climate change most at risk from adverse climate impacts. For example, immobile populations in sites of climate risk may be trapped and experience loss and damage including because they lack resources, assets, and networks to enable migration away from sites of risk [32].
There is critical need and value in initiatives that address 'vulnerability' and improve 'adaptive capacity' by investing in adaptation and human development in local sites of climate risk, thereby potentially limiting the need for out-migration. However, policy initiatives that focus on 'the vulnerable' and proximate social and environmental contexts are at risk of obscuring complex power relations and global inequities that create these vulnerabilities or limit adaptive capacity [108].
Policy recommendations referred to both economic and non-economic losses and damages and suggested widening our understanding of the linkages between mobility responses (including immobility) and wellbeing by looking at non-economic loss and damage and its links to mental health. It was argued that a lack of focus on this aspect might constitute a potentially costly public health inaction. The need identified was to provide culturally appropriate compensation for displaced and host populations with a range of populations at heightened risk requiring careful consideration including women and girls, elderly people, trapped populations, people living with disabilities, and people living with HIV/AIDS. Immobile populations living in such contexts may experience adverse health impacts that emerge from changes in water and food security, disease ecology, flooding and saltwater intrusion, and the psychosocial impacts of disrupted livelihoods [109]. These represent important aspects of loss and damage that emerge from climate change impacts [110]. Despite a widespread focus on livelihood security and damage to physical assets as key aspects of loss and damage, adverse health impacts among climate-affected populations-such as food and water insecurity-also represent aspects of loss and damage. These health costs and consequences, including among (im)mobile populations in contexts of climate risk, cause significant harm, and impede sustainable development.
The studies we review here all discuss the policy and practice significance of their findings. Most note that the scale of climate-related migration is projected to increase in the coming decades (while also noting that climatic factors rarely act alone in shaping human mobility) and that there are health risks and opportunities of climate-related mobility. Climate change is likely to act as an amplifier of health risks among mobile populations, rather than creating new vulnerabilities that require distinct policy and practice responses. Nevertheless, it is still an important and increasingly significant consideration in both migration and health policy. Therefore, it will remain of importance and value to strengthen the migrant sensitivity of health systems, ensure universal health coverage, and continue efforts to address the social determinants of health inequities for all including for mobile populations. Yet there may also be a need to respond to the increasing vulnerability of some migrants. There will likely be climate-related mobility across and within borders of countries in the Global South where health systems are generally weaker and have a lower capacity to deal with increased demand. And many people migrating in the context of climate change may move into areas of increased risk, such as in Bangladesh where informal settlements form in flood zones [65,79]. So, rather than creating new categories of vulnerable mobile populations, there will be an amplified need to address the health of mobile populations in often under-resourced and climate-vulnerable sites.

Limitations
In terms of study design, we used a priori protocol and a double-blinded approach for study selection and quality appraisal to limit bias. In addition to language and publication bias, there is always a risk in any systematic literature review-despite a thorough search strategy-that articles are missed. Studies were not omitted due to quality as per the MMAT tool recommendations [34], which may have led to findings of studies with weaker designs being included in the meta-synthesis. Five studies did not fit within the MMAT categorisation (modelling studies) and were therefore not assessed for quality, which remains a limitation of this review. While grey literature was not included, the focus on peer-reviewed empirical evidence was considered a strength.
There are limitations when considering the implications of this study. Considering health research is context and population-specific and climate-related migration is variable over space and time. Therefore, translating or generalising these findings to a range of diverse settings and populations may present risks. There is a need to understand the risks, exposures, vulnerabilities, and capacities of unique populations and settings to inform policy and practice decisions.
Finally, some included studies suggested (rather than demonstrated) potential (rather than actual) links to climate change. This may have led to an overestimation of the relationship between climate change, and health and mobility responses.

Future Research Directions
The recommendations extracted, analysed, and synthesized in this review pertained mainly to rural livelihoods revealing a potential gap in researching this nexus in urban settings. The research focused on mobile populations and less so on the entire migration ecosystem including sending and host communities echoing a call for research on climate mobilities to shift part of its focus from climate-sensitive sending areas to destination areas [103]. Further, there was a focus on adaptation with less attention paid to mitigation activities. It may be worthwhile looking at these concepts together, or on a spectrum whereby adaptation activities can play a role in mitigation and vice versa. There was very little to no reference to conflict in these studies although we know that conflict and migration are both related socially mediated, tertiary impacts of climate change and likely to be playing a role in some of these settings and influencing health and health resources. Finally, the included studies focused on agricultural climate change adaptation with less attention paid to infrastructure and health.

Conclusions
As climate change continues to shape patterns and scales of human migration, policymakers are challenged to make evidence-based decisions that enable competent governance of orderly, safe, regular, and humane migration, that safeguards human health and wellbeing. This review identified, analysed, and synthesized what is known to date from research investigating the climate change, migration, and health nexus. The findings largely confirm principles of good migration governance; that the universal promotion of migration should be avoided because it is not always adaptive, that forced migration should be prevented and that mobile populations should be supported in their decision-making. Some elements of our current understanding are reinforced in terms of the need to favour community-led approaches, provide durable solutions, preserve cultural ties and to enable migrant participation because it has the potential to maximise benefit for all those affected by mobility responses. This review also draws attention to more novel discussions such as non-economic loss and damages arising from climate change that include health impacts. The included studies focused on