Implementation preferences for the management of sexually transmitted infections in the South African health system: A discrete choice experiment

Abstract

Objectives Despite strengthening HIV prevention with the introduction of pre-exposure prophylaxis (PrEP), STI services have remained relatively unchanged and the standard of care remains syndromic management. We used a discrete choice experiment to investigate service users' preferences for the diagnosis and treatment of STIs in South Africa. Methods Between 1 March 2021 and 20 April 2021, a cross-sectional online questionnaire hosted on REDCap was administered through access links sent to WhatsApp support groups for HIV PrEP users and attendees of two primary healthcare clinics and two mobile facilities in the Eastern Cape and Gauteng provinces aged between 18 and 49 years. Participants either self-completed the questionnaire or received support from a research assistant. We used a conditional logit model for the initial analysis and latent class model (LCM) to establish class memberships, with results displayed as ORs and probabilities. Results We enrolled 496 individuals; the majority were female (69%) and <30 years (74%). The LCM showed two distinct groups. The first group, comprising 68% of the participants, showed a strong preference for self-sampling compared with no sampling (OR 2.16, 95% CI 1.62 to 2.88). A clinic follow-up appointment for treatment was less preferable to same-day treatment (OR 0.78, 95% CI 0.63 to 0.95). Contact slip from index patient (OR 0.86, 95% CI 0.76 to 0.96) and healthcare professional (HCP)-initiated partner notification (OR 0.63, 95% CI 0.55 to 0.73) were both less preferable than expedited partner treatment (EPT). The second group included 32% of participants with a lower preference for self-sampling compared with no sampling (OR 0.65, 95% CI 0.41 to 1.04). There was no treatment option that was significantly different from the others; however, there was a strong preference for HCP-initiated partner notification to EPT (OR 1.53, 95% CI 1.10 to 2.12). Conclusions Our results suggest that service users preferred STI testing prior to treatment, with the majority preferring self-taken samples and receiving aetiology-based treatment on the same day.