Village health workers as health diplomats: negotiating health and study participation in a malaria elimination trial in The Gambia

Background: Although many success stories exist of Village Health Workers (VHWs) improving primary health care, critiques remain about the medicalisation of their roles in disease-specific interventions. VHWs are placed at the bottom of the health system hierarchy as cheap and low-skilled volunteers, irrespective of their highly valued social and political status within communities. In this paper, we shed light on the political role VHWs play and investigate how this shapes their social and medical roles, including their influence on community participation. Method: The study was carried out within the context of a malaria elimination trial implemented in rural villages in the North Bank of The Gambia between 2016 and 2018. The trial aimed to reduce malaria prevalence by treating malaria index cases and their potentially asymptomatic compound members, in which VHWs took an active role advocating their community and the intervention, mobilising the population, and distributing antimalarial drugs. Mixed-methods research was used to collect and analyse data through qualitative interviews, group discussions, observations, and quantitative surveys. Results and discussion: We explored the emic logic of participation in a malaria elimination trial and found that VHWs played a pivotal role in representing their community and negotiating with the Medical Research Council to bring benefits (e.g. biomedical care service) to the community. We highlight this representative role of VHWs as ‘health diplomats’, valued and appreciated by community members, and potentially increasing community participation in the trial. We argue that VHWs aspire to be politically present and be part of the key decision-makers in the community through their health diplomat role. Conclusion: It is thus likely that in the context of rural Gambia, supporting VHWs beyond medical roles, in their social and political roles, would contribute to the improved performance of VHWs and to enhanced community participation in activities the community perceive as beneficial.

In The Gambia, VHWs serve a village of a minimum of 400 people to provide maternal services, child health services, health education, promotion, and treatment of common illnesses, including malaria [40]. This paper presents the results of social science study which closely worked with 10 VHWs involved in the cluster-randomised trial, Reactive Household-based Self-administered Treatment against residual malaria transmission (RHOST). All of them were male farmers and/or herders, with similar socio-economic status as other community members. Most of them had not completed formal education, while a few had never been to a formal school (excepting Koranic school). Only the 3 VHWs who had (almost) completed primary education were able to communicate in English. RHOST trial (registered with ClinicalTrials.gov, {“type”:”clinical-trial”,”attrs”:{“text”:”NCT02878200″,”term_id”:”NCT02878200″}}NCT02878200) was conducted by the Medical Research Council unit in The Gambia (MRCG) between 2016 and 2018, in 34 villages (randomly split for intervention and control) in the North Bank region of The Gambia. The trial aimed to reduce malaria prevalence in the study villages by treating malaria cases and their potentially asymptomatic compound members [41]. The trial provided additional training and supply of malaria diagnostic and treatment tools to VHWs beyond the national health system scheme [42]. The VHWs role in the trial intervention arm was mainly to diagnose and treat malaria index cases, prescribe and distribute antimalarial drugs to compound members of the index case, and communicate with community members and with the trial team. This role of VHWs was shaped by community members participating in the Community Lab of Ideas for Health (CLIH) – a specific participatory approach developed and conducted within the trial [43]. In CLIH, the trial and communities co-developed implementation strategies in which VHWs became the important trial implementors because of their highly regarded social and political status and of trust by community members (see results). The trial provided each VHW with a monthly monetary incentive (1500Dalasi = ±25EUR, equivalent to a 50 kg bag of rice) during the implementation period (i.e. malaria season around June to December) for the VHWs to take up this additional role on top of their routine roles for primary health care. The population comprised mainly Mandinka, Fula, and Wolof ethnicities while also including Bambara, Turka, and Tilibonka minorities. There were almost no mono-ethnic group village but a combination of different ethnic groups residing together harmoniously. Inter-ethnic marriage was common [44]. These societies were polyglot while each has its own language. The population was Muslim and mostly farmers (for both self-consumption and cash-crop) and/or herders. Many villages were located far from the main road and health facilities, making access to health services difficult. In the villages, the social organisation was based on patrilineal kinship [45]. The head of the village was the Alkalo, a role traditionally inherited patriarchally from the village founder who was surrounded and advised by the council of elders consisting of an Imam, deputy-Imam, Marabout (religious teacher and healer), and elderly compound heads. Village administration was managed by the Village Development Committee (VDC) who were responsible for development activities in the community. The VDC comprised of VHW, sub-committees such as woman’s groups, youth groups, and a representative from each ethnic group. A compound head was responsible for his compound members consisting of his family, extended family members, and sometimes guests and/or seasonal workers, and was regarded as the role-model in the family [45–47]. A social science study was conducted within the trial, using a sequential exploratory mixed-method study design (QUAL-quan). Social science study aimed to provide contextual information to the trial and aid in co-creating trial implementation strategies with communities [43]. The field research team consisted of researchers and local fieldworkers with diverse backgrounds and experiences. The field research team conducted a total of 161 in-depth interviews, 93 focus group discussions (including exploratory participatory workshops, key-informants’ meetings, monitoring meetings), 160 monitoring calls, and observations with informal chats by visiting and staying in intervention villages (n = 17) between March 2016 and December 2017. We developed a topic guide prior to data collection and adjusted it accordingly to emerging findings and hypotheses. Interviews took place at respondents’ convenience mostly in informal settings such as one’s household. Discussions were held by appointment – we consulted the Alkalo and VHW for the best available date (e.g. Friday after prayer time) and invited pre-identified key-informants to join the discussions. All conversations were translated from local languages (Mandinka, Fula, and Wolof) to English and vice versa by the local fieldworkers. Semi-structured monitoring calls were made to VHWs by fieldworkers twice a month during the implementation. Additionally, as part of our observations, we carefully observed both everyday village life and VHW life. The first included hierarchies among household/compound members, neighbours, visitors, youth and the elderly, available malaria protective measures, day-to-day socio-economic activities, and mobility; while the latter included observations on VHWs’ relation to other community members, their performances during VHWs’ training (n = 2), performance drills (n = 10), and actual patient visits. We purposefully selected our respondents and carried out snowball sampling. Our respondents were mainly: VHWs, traditional birth attendants, Alkalos, Imams, traditional healers, compound heads, members of VDC including women and youth groups, teachers, farmers, and caretakers (often women) of malaria patients. Participants for the discussions were identified through the ongoing ethnographic study, which was then followed by a stakeholder analysis. We analysed data concurrently during data collection, through ongoing field analysis with the research team discussing and validating findings to minimise possible bias, as well as iteratively making and testing various hypotheses. Field notes were transferred into digital form immediately after data collection every day. Most audio-recorded conversations were transcribed verbatim in English by the fieldworkers. We used NVivo (ver.11) qualitative data analysis software to code transcribed interviews and notes. Two rounds of survey administration took place in June 2016 (baseline) and November–December 2017 (endline) in both intervention (n = 17) and control villages (n = 17) by the trained MRCG fieldworkers. The surveys were developed by the researchers based on the qualitative findings from the initial ethnographic study (March–May 2016). The surveys were paper-based containing both standardised closed- and open-ended questions. The surveys were first piloted to ensure clarity of questions and to avoid translation errors in local languages. Sample size was calculated by the trial epidemiologist ensuring that the number of households randomly selected was in proportion to the size of the village. A total of 324 baseline and 273 endline surveys were administered to adults above 16 years old (two samples of the same villages at different occasions and not among the same individuals). When any of the initially selected persons were unavailable, the next adult who shares similar characteristics (e.g. same-sex, closer in age) in the household was approached to complete the survey. When respondents were irreplaceable (no one to be found at a household), the non-response sheet was filled in. For this paper, we extracted the relevant questions related to VHWs (e.g. health-seeking behaviour, people’s perception towards VHWs) from the survey for intervention villages (baseline: n = 126, endline: n = 104) to focus on the baseline and endline differences in interventions villages. The size of the effective sample used in this study is thus 230 persons subjected to the trial intervention. Data were double entered by the trial data entry clerks. For this paper, a statistician calculated the frequencies, measures, and how likely these are compared with the expected distribution under randomness (the null hypothesis H0), with the Statistical Analysis System. The Chi-square test for two independent samples has been used, together with the Cramer’s V measure to have an idea of the strength of the relation between responses and baseline/endline variable for respondents in intervention villages [48]. Given the small number of cases (n = 230), we take 0.10 as the critical border (prob. < 0.10). For the analysis, we considered “NA (not applicable)” answer as a missing value and kept “DK (don’t know)” as a relevant response for analysis of people’s perception. All participants were provided an explanation, in their local language, of the study details and what it entailed prior to interviews, surveys, and discussions, as well as confidentiality and their rights to withdraw or not participate. Informed consent was obtained from all participants verbally, which was approved by The Gambia Government/MRC Joint Ethics Committee and the Institutional Review Board of the Institute of Tropical Medicine, Antwerp. Verbal consent was preferred due to the high rate of illiteracy among the study populations and to avoid sowing mistrust in communities by obliging signatures. We audio-recorded interviews and discussions with participant consent. All methods were carried out in accordance with relevant guidelines and regulations, including the Code of Ethics of the American Anthropological Association.

N/A

• masunaga, yoriko
• muela ribera, joan
• jaiteh, fatou
• de vries, daniel h.
• grietens, koen peeters