Oral pre-exposure prophylaxis (PrEP) is highly effective and being scaled at health clinics throughout sub-Saharan Africa. Barriers to clinic-based PrEP delivery such as high costs for clients and clinics remain. Innovative models that improve the efficiency of PrEP delivery without jeopardizing client outcomes are needed in this setting.
In a randomized non-inferiority implementation trial, we tested a novel model for PrEP delivery in Kenya: 6-month drug dispensing supported with interim HIV self-testing (HIVST). All participants were ?18 years and had taken PrEP for 1 month. Participants were 2:1 randomized to: 1) 6-month PrEP dispensing (with semiannual clinic visits, supported by HIVST conducted at home after 3 months) or 2) standard-of-care (SOC) PrEP delivery (3-month supply with quarterly clinic visits). Our primary outcomes, measured at 6 months, were HIV testing, PrEP refilling, and PrEP adherence (detectable tenofovir-diphosphate in dried blood spots). We used binomial regression models to estimate risk differences (RDs) and interpreted 1-sided 95% confidence interval (CI) lower bounds (LB) ?-10% as non-inferior.
From May 2018 to February 2020, we enrolled and followed 495 participants: 165 men and 130 women in HIV serodifferent couples and 200 women not in known serodifferent couples. At 6 months, 83.3% (274/329) of those assigned the intervention tested for HIV compared to 84.3% (140/166) for SOC (RD -1.2%, 1-sided 95% CI LB -6.9%), Figure 1. Among intervention participants, 78.1% (257/329) refilled PrEP, compared to 80.7% (134/166) for SOC (RD -2.6%, 1-sided 95% CI LB -8.9%) and 60.8% (200/329) were adherent to PrEP compared to 57.2% (95/166) (RD 2.4%, 1-sided 95% CI LB -5.1%). In sub-group analyses among serodifferent couples, all women, and women singly-enrolled, findings were generally comparable to those among all participants. However, among women singly enrolled, PrEP adherence was 19.8% higher (2-sided 95% CI 5.8-33.8%) for those assigned the intervention compared to SOC, a superior result that was statistically significant. No participants acquired HIV.
Dispensing 6 months of PrEP with HIVST for interim testing at 3 months reduces the number of PrEP clinic visits in half without compromising HIV testing, PrEP refilling, or PrEP adherence. HIVST to support PrEP continuation can enable models of care that that require less frequent contact with the health system.