Leaving no one behind: targeting mobile and migrant populations with health interventions for disease elimination—a descriptive systematic review

Our search yielded studies focusing on different types of MMPs—nomadic pastoralists, migrant laborers, IDPs, and refugees—each with various motivations for movement. Common migrant population definitions specify whether movement is cross-border or within single countries, being driven by the demands of livestock (nomadic pastoralists), in pursuit of economic opportunities (migrant laborers), or in response to natural disasters or conflict (IDPs and refugees) [18]. For example, two specific groups of interest, the Maasai and Turkana, move seasonally to access water sources for their cattle [19]. The literature describes their movement as independent of borders, contributing to persistent malaria transmission specifically [20, 22]. Of relevance for NTD epidemiology, motivation for movement is often associated with demographic characteristics that may also be associated with NTD risk. For example, economic migrants are often young and/or single adults, while IDPs or refugees moving in response to conflict or natural disaster are often comprised of family units that include women and children. Furthermore, barriers that women face due to their MMP status can also be exacerbated by their minority gender status [8]. Among pastoralist communities, movement patterns may be different for women and children than for male youth and adults who are moving livestock in certain seasons. Where the demographics of MMP groups and risk for NTDs overlap in endemic receiving or sending areas, programs can identify priority populations for NTD interventions.

We found three studies discussing the difficulty of reaching MMPs with health services in East Africa [17, 19, 23]. These papers described the issue specifically in the Maasai people in Kenya and Tanzania. Lawson et al. [19] found that the Maasai face barriers to healthcare access due to inadequate service provision in their remote areas. Therefore, when comparing Maasai health to other nomadic groups, the authors found that levels of child malnutrition and disease were very high. Mtuy et al. [23] observed that Maasai seemingly had limited health knowledge as interviews indicated that there is an erroneous belief that trachoma is caused by environmental allergens. An additional four studies illustrated low levels of engagement in healthcare services due to distrust of western medicine or misinformation about health service campaigns. Consequently, studies observed worse health outcomes in the MMPs, such as increased burden of malaria and trachoma [15, 20‐22]. In contrast, one study comparing health issues of settled and nomadic Turkana in Kenya reported that the settled Turkana suffered from higher rates of infections like eye infections, colds, coughs, and respiratory infections than the nomadic Turkana [14]. The majority of these studies’ findings were corroborated by our expert consultations, specifically about the types of MMP movement (e.g., cross border, seasonal, labor-specific travelers) in East Africa and how this mobility has affected focus disease service delivery like mass drug and vaccination campaigns in the region.

The studies describing East African MMPs, as well as additional records from domains two and three, also highlighted variables of population movement—such as timing, duration of movement, demographics of those moving, and border crossing—that affect how MMPs utilize healthcare services. Timing can affect health outcomes by exposing MMPs to environmental risks or impeding physical access to services (e.g., by increasing remoteness or by the degradation in roads, tracks or trails, or localized flooding in the wet season) [17, 23, 24, 119, 120]. Duration of migration can similarly affect exposure, disease risk, or access to services that are only available on a local or regional basis [20, 25‐27, 72]. Demographics are important when considering diseases that have outsize effects based on age or gender, and border crossing exposes migrants to the policies and practices of different government health care systems which might require international coordination and collaboration to prevent gaps in coverage [18, 28‐32, 73, 121]. Demographics are also an important consideration when migration flows are of large size and/or are unpredictable (such as when movement is dependent upon climate conditions, conflict, or natural disaster) as this can lead to situations where the health care infrastructure is not prepared to absorb or respond to the numbers requiring intervention [25, 33, 74, 75, 122].

One of the major stumbling blocks to understanding disease dynamics in MMPs is the difficulty in enumerating and surveying these groups. Commonly used means of constructing sampling frames, such as censuses or household enumeration, are not designed to capture and sample MMPs. We found twelve articles describing alternative methods to enumerating MMPs (see Table 4). Sampling strategies include the use of geospatial data to identify movements, tents, or settlements [86, 98, 101, 104, 111]. Study results suggest these alternative sampling frames can be comparable to standard methods. For example, a study in Cameroon on mapping nomadic pastoralist movement found that more than 75% of cattle camps identified as probable through satellite imagery were found to be camps upon manual, on-the-ground confirmation [101]. In addition to satellite imagery, tracking mobile phones or use of mobile phone apps has also been used to map the movement patterns of migrants and determine the length of stay and locations along a migration route [72, 77, 95, 107]. For example, Tomkins et al. [107] used mobile phone data to analyze Senegalese migration patterns and how they may affect malaria transmission. Their study found that 60% of people have recurring trips to the same location and most visits include an overnight stay which increases the risk of malaria infection. Albeit cell phone data can provide accurate information to determine travel routes, it is limited to those MMPs with phones and areas with good cellular reception, possibly excluding low-income populations and those in very rural settings. Additionally, cross-border migration may not be tracked if MMPs do not access different cell phone service providers networks operating on the other side of the border. Snowball or respondent-driven sampling has also been used to survey MMPs with success [102, 110]. Using focus group discussions and key informant interviews, Smith et al. [102] found that 54% of Nepalese malaria cases were imported from India due to work travel. Imported malaria cases were observed more in males (85%) than females and suggested that longer trips were more predictive of malaria infections. This study exemplifies that alternative sampling methods can be successfully used to enumerate MMPs when traditional methods cannot be applied.

An alternative sampling approach that has been successfully implemented is the engagement with MMPs themselves, such as participatory mapping and microplanning. This approach relies on local knowledge and local leaders or champions to identify barriers to access and points for intercepting target populations for services. In Nigeria, Uzoma and colleagues engaged with MMPs and completed route mapping to determine migratory routes and their potential contribution to polio transmission [108]. After successfully producing an accurate migratory map, a vaccination campaign was conducted which increased first dose vaccination coverage from 752 to 1155 nomadic children under 5 years of age over the 2-year campaign.

We found 11 papers detailing methods used to implement mass health programming in MMPs (see Table 5) [8, 38, 73, 76, 80, 82, 88, 92, 94, 96, 100]. Five of these detailed approaches and successful engagement with members of the MMP as community health workers to provide ongoing service delivery, identify disease transmission hotspots, and engage in ongoing health education in a community [8, 88, 92, 94, 96]. For example, Hu et al. [88] implemented an expanded vaccine program in China—including more frequent services, provided by migrant clinical attendants; expanded social mobilization; and widened screening to identify migrant demands for vaccines—and observed an increase (71.5 to 88.6%) in fully vaccinated migrant children. These studies emphasize the need for altering common implementation approaches to include MMPs. Some of these include targeting interventions to specific known MMP groups within target districts, adjusting intervention coverage to include remote areas, and offering interventions at multiple sites (often along borders or at work sites with known migrant workers) and times to account for seasonal movement [8].

Three articles highlighted One Health approaches that identify multiple entry points (such as veterinary care, agricultural extension, and other community points of entry) as a platform to engage the community in discrete health interventions [76, 80, 82]. For example, Bomoi et al. [82] reported successful use of joint animal and child vaccination, which improved childhood vaccination rates from 22.7 to 80.1% in Nigerian Fulani nomadic pastoralists over the course of their study; concurrently, animal vaccination rates also rose from 41 to 61%.

Lastly, two studies illustrate cross-border collaboration as another important strategy in engaging with MMPs [73, 100]. While the studies noted and described cross-border migration as a complicating factor that influences MMPs’ access to health care and disease risk (see outcome domains 1 and 2), Kleinschmidt et al. [73] and Haydarov et al. [100] suggest cross-border coordination as an approach towards lessening transmission risks and improving access to care. Means of cross-border collaboration ranged from formal efforts such as timing MDA on both sides of the border to coincide with one another, to informal WhatsApp groups that include communication between community health workers or other health service providers on both sides of the border. Informal communication channels have been used to communicate about movement, any increases in numbers of cases to assist in deploying rapid responses to community movement and/or increased morbidity.

A final observation that emerged under this outcome domain is the importance of planning and logistics. Emphasis is placed on the need to predict population movement and procure enough commodities, and providing services to MMPs will require greater resources—in time, human capital, and funding—than working with settled populations [22, 80, 97, 114].