Abstract
Progress has been made to scale oral pre-exposure prophylaxis (PrEP) as part of combination HIV prevention, with the WHO recommending differentiated, simplified and demedicalized approaches. This study explored user preferences for components of a PrEP service delivery package, through a discrete choice experiment (DCE) among 307 people accessing primary healthcare services in South Africa between November 2022 and February 2023. Attributes included were: Source of information about HIV prevention and PrEP; Site for PrEP initiation and follow-up; Frequency of follow-up; PrEP pick-up point; HIV testing whilst using PrEP; Contact between appointments. A D-efficient DCE design was created with 16 choice sets in 2 blocks. DCE data were analysed using generalised multinomial logistic models. Compared to printed materials, participants preferred getting information about PrEP online (aOR = 7.73, 95% CI = 5.13-11.66) and through WhatsApp (aOR = 2.23, 95% CI = 0.98-5.55). PrEP initiation at a pharmacy or mobile clinic was valued equally to initiating PrEP at a clinic, but a community pop-up site was less preferred (aOR = 0.46, 95% CI = 0.33-0.64). There was a preference for 6-monthly over 3-monthly follow-up (aOR = 11.88, 95% CI = 5.44-25.94). Participants preferred collecting PrEP from a pharmacy (aOR = 5.02, 95% CI = 3.45-7.31), through home delivery (aOR = 2.18, 95% CI = 1.26-3.78) and from a vending machine (aOR = 1.43, 95% CI = 1.02-1.99) relative to where they initiated PrEP. Participants also preferred HIV self-testing over a healthcare provider test (aOR = 5.57, 95% CI = 3.72, 8.36). WhatsApp or Facebook groups (aOR = 4.12, 95% CI = 3.00-5.67), monthly phone calls (aOR = 2.84, 95% CI = 1.73-4.67) and weekly messages (aOR = 1.47, 95% CI = 1.10-1.97) were preferred contact between appointments, relative to no contact. To meet users' preferences, there is a need to expand decentralised and self-led HIV prevention services.