Abstract
BACKGROUND: As in much of sub-Saharan Africa, substantial HIV testing gaps remain in South Africa, particularly among adult men ages 20-35, young people ages 15-24 and key populations. Innovative strategies, such as HIV self-testing (HIVST), are needed to reach such under-served populations. We evaluated a range of HIV self-test kit distribution models' potential to reach adult men, young people and key populations in South Africa, to inform targeted approaches. METHODS: This cross-sectional study used data from community and facility-based HIV self-test kit distribution models implemented from October 2017 to April 2020. Self-test kits were distributed as part of the Unitaid-funded Self-Testing AfRica (STAR) programme. Data were collected from individuals who obtained self-test kits through five distribution models. Frequencies and proportions were used to describe the characteristics of the study populations and self-test kit distribution approaches. RESULTS: Over 2.5 years, 1 071 065 self-test kits were distributed across the five models. Community-based distribution accounted for 63% of total kits distributed, while the private sector (primarily workplace) accounted for 26%. Distribution at public sector health facilities accounted for 7% and distribution through the key population and secondary distribution models accounted for 2% each. Of those obtaining kits, and for whom we collected previous testing data (n = 771 612, 72%), 11% had never tested for HIV, 29% had not tested for at least a year, 41% had tested within the last 4-12 months and 19% had tested within the preceding three months. More men (64%) than women obtained self-test kits across all distribution models. The majority (80%) of men obtaining self-test kits were aged 20-40 years, and primarily received these at public transport terminals (36%), workplaces (18%) and hotspots (14%). A small proportion of men was reached through female sex workers. CONCLUSIONS: This analysis of programme data enabled us to identify HIV self-test kit distribution models that are best suited to reach specific priority and under-tested populations, particularly adult men and young people. Models/sub-models that reach self-test users where they live, work and spend time, are likely to result in higher HIVST uptake. Study findings can inform future HIVST scale-up in South Africa.