Geneva Health Forum Archive

Browse and download abstracts, posters, documents and videos from past editions of the GHF

How well do we know the “users” of health interventions? The example of (non-)enrolment in voluntary community-based health insurance schemes in Mali.

Author(s) Alexander Schulze1.
Affiliation(s) 1Global Health, Novartis Foundation for Sustainable Development, Basel, Switzerland.
Country - ies of focus Mali
Relevant to the conference tracks Social Determinants and Human Rights
Summary For health interventions to be successful, not only the health system needs to be considered, but also the patients understood. However, often in health research this understanding is reduced to their socio-economic status.
The present empirical study, focusing on the reasons for (non-) enrolment in voluntary health insurance schemes in rural Mali, expands the analysis of household characteristics to sociocultural orientations and intra-familial structures, including decision-making patterns.
The results reveal that the socio-economic status of households does not explain membership in a health insurance scheme, whereas the attitudes heads and family decision making patterns clearly do.
Background Systematic, quantitative evidence on user characteristics of health interventions is mostly limited to socio-demographic features and economic status. Yet, as research in medical anthropology has revealed, the acceptance of health related innovations is not only dependent on these factors.The example of (non-)enrolment in voluntary health community-based insurance schemes in sub-Saharan Africa mirrors this research gap. These schemes are characterized by modest enrolments rates and do not tap their potential in an already limited catchment area. In order to increase their coverage, the reasons for (non-)enrolment need to be better understood.However, to date, research has focused on the supply side, i.e. features of the insurance schemes and their contracted providers. On the demand side, comparative studies of enrolled and non-enrolled households have mainly concentrated on their socioeconomic status.Three research gaps remain which are not unique to the analysis of community-based health insurance: 1) sociocultural features are not analyzed in a quantifiable manner as socio-economic characteristics are. 2) household and family structures and their decision making patterns are not considered. Thirdly, significant factors are not explained and related to each other.
Objectives The present empirical study aims to fill the following research gaps:1) to expand the analysis of household characteristics beyond households’ socio-economic status by also taking account of their sociocultural orientations and intra-familial structures, including decision-making patterns. The main hypothesis underlying this approach is that socio-economic differences have been overestimated regarding their importance in determining the uptake of innovations, i.e. in the presented case membership in health insurance schemes, whereas attitudes and values, such as openness to or mistrust of social innovations, have been underestimated.2) A further goal is to systematically bring together and interrelate influencing factors that are meaningful beyond statistical significance, in order to ultimately reveal different lifestyle patterns associated with membership in, or refusal to join, a health insurance scheme.3) A further aim of this study is to present an analytical framework that links the socio-economic characteristics of households with their sociocultural orientations as well as with household and family structures.
Methodology A mixed methods approach, with different instruments employed sequentially, was developed. The central rationale behind a mixed-methods approach of this type was the idea to use qualitative methods of data collection not only in the exploratory phase but also at a later stage, to enhance interpretation of results from quantitative statistical data analysis and to develop explanatory patterns.The core element of the mixed methods approach chosen was household surveys on lifestyle patterns and social protection strategies. A total of 600 heads of households were interviewed in two localities. Of these, in each locality 200 had no insurance, while 100 did have insurance. In a second survey round, half of these household heads were again interviewed about their social protection strategies. In each study area this round covered 100 households with insurance and 50 households with no insurance.Households were the focus of research for two reasons: on the one hand, they are the unit of membership in the community-based health insurance schemes being studied. On the other hand, households are important repositories of lifestyle patterns within society and have a great degree of influence on the activities of their members. This is manifested, for example, in the transmission of values between generations and in the sharing of resources. Moreover, households are units of decision-making in different areas of life.Comparison of two rural areas in Mali had been considered useful insofar as community-based health insurance schemes are being advocated for peasant populations, primarily in French-speaking Africa. Furthermore, the two localities differ in terms of climate and in socio-economic and sociocultural terms but have a very similar health insurance scheme. Comparison of two areas also made it possible to examine to what extent social differences exist in rural Mali.

The household surveys were preceded by an exploratory phase including group and individual discussions, which primarily served to identify locally relevant categories of social differentiation. Following the household surveys, preliminary results were presented to representatives from the municipalities. These discussions brought out new points of view that were of use in further data collection and analysis. Following this, 20 group discussions and 24 individuals from selected families were interviewed in order to obtain insights into household and family decision-making patterns.

Results The results reveal that the socio-economic status of households does not explain membership in a community-based health insurance scheme, whereas the attitudes of household heads and family decision-making patterns clearly do. For instance, households of the lower socioeconomic tercile were even slightly more likely to be members than those of the middle tercile (p-value: 0.669; CI: .63-2.02; OR: 1.13).Households that are single households or are not part of a large family have greater freedom in decision-making. Moreover, heads of households with insurance appreciate different social changes (e.g. existence of new, formal organizations including community-based health insurance schemes), take on responsible roles in their communities and plan in a manner where the benefit is unsure or not immediately given, significantly more often than heads of uninsured households. For instance, those heads of household who only utter negative social changes are significantly less likely to be insured than those who also mention positive changes (p-value: 0.000; CI: .21-.57; OR: .35).Based on the results, a typology of households and their lifestyle patterns was empirically determined in both study areas. Each of the two study areas, two different household types were identified that are distinct not by their socio-economic status but by family structure and sociocultural orientation. The existence of different household types in both areas underscores the clear social differences that exist in rural Mali.
Conclusion The findings from this research on the reasons for (non-)enrolment of households in community-based health insurance schemes in rural Mali reveal that limiting the "users" of health interventions to their socio-economic status in many circumstances does not sufficiently explain why people "lack access to" or do not take up the services offered through health interventions.In order to better target potential users of health interventions and increase coverage, a more comprehensive but systematic analysis of their characteristics must be done. This includes socio-cultural features such as attitudes towards social innovations and change, leading values, consumption priorities as well as decision making patterns at household and family level.In other words, for health interventions to be successful, not only the heath system and its actors need to be considered, but also the patients. However, they must not be reduced to people living in poverty. Rather research must acknowledge that they are characterised by different livelihood patterns that expand beyond the issue of poverty.