Background: Adult HIV prevalence remains high in Southern Africa. HIV self-testing (HIVST) is a new approach with potential to improve coverage. To inform HIVST implementation, we elicited user preferences to quantify the strength of preferences for alternative HIVST distribution and linkage options using discrete choice experiments (DCEs).
Methods: Four DCEs were implemented within representative household surveys in Malawi and Zambia. Adults (≥16 years) were randomly allocated to DCE on: HIVST distribution or linkage to confirmatory testing and care. We explored preference heterogeneity by country, age, gender and HIV testing experience.
Results: For distribution (n=1116), participants preferred to receive HIVST at home or at a distributor''s home to mobile clinics (and drugstores in Zambia only). Oral fluid tests were preferred to provider-administered or self-administered blood-based tests. HIVST distribution via sexual partner was viewed less favourably, especially among men. HIVST price had a strong negative impact on uptake.

Figure 1 - HIVST distribution DCE - Malawi
[Figure 1 - HIVST distribution DCE - Malawi]

For linkage to care (n=843), phone call was preferred to SMS, a personal visit or no follow-up. The home of the HIV testing services (HTS) provider was preferred for confirmatory testing after a reactive self-test in Malawi only. Facility user fee and waiting time for confirmatory testing were strong disincentives.

Figure 2 - Linkage to care DCE - Malawi
[Figure 2 - Linkage to care DCE - Malawi]

Conclusions: The DCE results were consistent across these countries with similar HIV context and suggest community-based distribution and linkage are critical aspects of HIVST delivery. A next step is to explore preferences of populations among whom HTS uptake and linkage to care are known to be low, such as men, youth and key populations.