Effect of differentiated direct-to-pharmacy PrEP refill visits supported with client HIV self-testing on clinic visit time and early PrEP continuation.

INTRODUCTION: Delivery of oral pre-exposure prophylaxis (PrEP) is being scaled up in Africa, but clinic-level barriers including lengthy clinic visits may threaten client continuation on PrEP. METHODS: Between January 2020 and January 2022, we conducted a quasi-experimental evaluation of differentiated direct-to-pharmacy PrEP refill visits at four public health HIV clinics in Kenya. Two clinics implemented the intervention package, which included direct-to-pharmacy for PrEP refill, client HIV self-testing (HIVST), client navigator, and pharmacist-led rapid risk assessment and dispensing. Two other clinics with comparable size and client volume served as contemporaneous controls with the usual clinic flow. PrEP continuation was evaluated by visit attendance and pharmacy refill records, and time and motion studies were conducted to determine time spent in the clinics. Dried blood spots were collected to test for tenofovir-diphosphate (TFV-DP) at random visits. We used logistic regression to assess the intervention effect on PrEP continuation and the Wilcoxon rank sum test to assess the effect on clinic time. RESULTS: Overall, 746 clients were enrolled, 366 at control clinics (76 during pre-implementation and 290 during implementation phase), and 380 at direct-to-pharmacy clinics (116 during pre-implementation and 264 during implementation phase). Prior to implementation, the intervention and control clinics were comparable on client characteristics (female: 51% vs. 47%; median age: 33 vs. 33 years) and PrEP continuation (35% vs. 37% at 1 month, and 37% vs. 39% at 3 months). The intervention reduced total time spent at the clinic by 35% (median of 51 minutes at control vs. 33 minutes at intervention clinics; p<0.001), while time spent on HIV testing (20 vs. 20 minutes; p = 0.50) and pharmacy (8 vs. 8 minutes; p = 0.8) was unchanged. PrEP continuation was higher at intervention versus the control clinics: 45% versus 33% at month 1, 34% versus 25% at month 3 and 23% versus 16% at month 6. TFV-DP was detected in 85% (61/72) of samples, similar by the study group (83% vs. 85%). CONCLUSIONS: A client-centred PrEP delivery approach with direct-to-pharmacy PrEP refill visits plus client HIVST significantly reduced clinic visit time by more than one-third and improved PrEP continuation in public health HIV clinics in Kenya.

Population-level analysis of natural control of HIV infection in Zambia and South Africa: HPTN 071 (PopART).

INTRODUCTION: HIV controllers have low viral loads (VL) without antiretroviral treatment (ART). We evaluated viraemic control in a community-randomized trial conducted in Zambia and South Africa that evaluated the impact of a combination prevention intervention on HIV incidence (HPTN 071 [PopART]; 2013-2018). METHODS: VL and antiretroviral (ARV) drug testing were performed using plasma samples collected 2 years after enrolment for 4072 participants who were HIV positive at the start of the study intervention. ARV drug use was assessed using a qualitative laboratory assay that detects 22 ARV drugs in five drug classes. Participants were classified as non-controllers if they had a VL ≥2000 copies/ml with no ARV drugs detected at this visit. Additional VL and ARV drug testing was performed at a second annual study visit to confirm controller status. Participants were classified as controllers if they had VLs <2000 with no ARV drugs detected at both visits. Non-controllers who had ARV drugs detected at either visit were excluded from the analysis to minimize potential confounders associated with ARV drug access and uptake. RESULTS: The final cohort included 126 viraemic controllers and 766 non-controllers who had no ARV drugs detected. The prevalence of controllers among the 4072 persons assessed was 3.1% (95% confidence interval [CI]: 2.6%, 3.6%). This should be considered a minimum estimate, since high rates of ARV drug use in the parent study limited the ability to identify controllers. Among the 892 participants in the final cohort, controller status was associated with biological sex (female > male, p = 0.027). There was no significant association between controller status and age, study country or herpes simplex virus type 2 (HSV-2) status at study enrolment. CONCLUSIONS: To our knowledge, this report presents the first large-scale, population-level study evaluating the prevalence of viraemic control and associated factors in Africa. A key advantage of this study was that a biomedical assessment was used to assess ARV drug use (vs. self-reported data). This study identified a large cohort of HIV controllers and non-controllers not taking ARV drugs, providing a unique repository of longitudinal samples for additional research. This cohort may be useful for further studies investigating the mechanisms of virologic control.