Understanding the pathways leading to socioeconomic inequalities in HIV testing uptake in 18 sub-Saharan African countries: a mediation analysis

Introduction Although socioeconomic inequalities in HIV prevention, testing and treatment services have been well documented, their drivers remain poorly understood. Understanding the different pathways between socioeconomic position and HIV testing across different countries could help designing tailored programs aimed at reducing such inequalities. Methods We analysed data from Demographic and Health Surveys conducted between 2010 and 2018 in 18 sub-Saharan African countries (Burkina Faso, Cameroon, Côte d’Ivoire, Congo DR, Ethiopia, Guinea, Kenya, Lesotho, Liberia, Malawi, Mali, Niger, Rwanda, Senegal, Sierra Leone, Tanzania, Zambia and Zimbabwe). Using a potential outcomes framework and the product method, we decomposed the total effect linking wealth and recent (< 12 months) HIV testing into i) direct effects, and ii) indirect effects, via demand-related (related to individual’s ability to perceive need for care and inclination to seek care) or supply-related (related to individual’s ability to reach, pay for and engage in health care) mediators. Multivariable gender-specific modified Poisson models were fitted to estimate proportions mediated, while accounting for exposure-mediator interaction when present. Results A total of 392,044 participants were included in the analysis. Pro-rich wealth-related inequalities were observed in a majority of countries, with nine countries with high levels of inequalities among women and 15 countries among men. The indirect effects of each mediator varied greatly across countries. The proportion mediated tended to be higher for demand-related than for supply-related mediators. For instance, among women, HIV-related knowledge was estimated to mediate up to 12.1% of inequalities in Côte d’Ivoire; this proportion was up to 31.5% for positive attitudes toward people living with HIV (PLHIV) in Senegal. For the four supply-related mediators, the proportion mediated was systematically below 7%. Similar conclusions were found when repeating analyses on men for the demand-related mediators, with higher proportions mediated by positive attitudes toward PLHIV (up to 39.9% in Senegal). Conclusions Our findings suggest that socioeconomic inequalities in HIV testing may be mediated by the demand-side more than supply-side characteristics, with important variability across countries. Overall, the important inter-country heterogeneity in pathways of socioeconomic inequalities in HIV testing illustrates that addressing inequalities requires tailored efforts as well as upstream interventions. A French version of the abstract is available upon request from the corresponding author.