Efficiency of 6-month PrEP dispensing with HIV self-testing in Kenya: an open-label, randomised, non-inferiority, implementation trial

Abstract

Background: Oral pre-exposure prophylaxis (PrEP) for HIV prevention is highly effective and is being implemented at scale at health clinics throughout sub-Saharan Africa. However, barriers to clinic-based PrEP delivery remain. We aimed to establish the efficiency of semiannual PrEP clinic visits supplemented with interim home-based HIV self-testing (HIVST) versus standard of care for HIV testing, drug refilling, and adherence among PrEP users.

Methods: This was a randomised, open-label, non-inferiority trial done at the Partners in Health and Research Development clinic in Thika, Kenya. Eligible participants were HIV-negative adults (≥18 years) at risk of acquiring HIV who had started PrEP at least 1 month before enrolment. Participants were randomly assigned (1:1:1) to 6-month PrEP dispensing plus interim blood-based HIVST (with biannual clinic visits), 6-month PrEP dispensing plus interim oral fluid-based HIVST (with biannual clinic visits), or standard of care PrEP delivery (3-month PrEP dispensing with quarterly clinic visits). The three coprimary outcomes, measured at 6 months, were HIV testing (any testing between enrolment and the 6-month visit), PrEP refilling, and PrEP adherence (detectable tenofovir diphosphate concentration in dried blood spots). All analyses were done according to the intention-to-treat principle. We used binomial regression models to estimate risk differences and one-sided 95% CIs. 6-month PrEP dispensing was considered non-inferior to standard of care if the lower limit bound of the one-sided 95% CI was greater than or equal to -10%. This study is registered with ClinicalTrials.gov, NCT03593629.

Findings: Between May 28, 2018, and Feb 24, 2020, 495 participants were enrolled: 165 men and 130 women in HIV serodifferent couples and 200 singly enrolled women. 166 participants were randomly assigned to the standard of care group, 163 to the 6-month PrEP dispensing plus oral-fluid HIVST group, and 166 to the 6-month PrEP dispensing plus blood-based HIVST group. At 6 months, 274 (83%) of 329 participants in the combined 6-month PrEP dispensing group had tested for HIV compared with 140 (84%) of 166 participants in the standard of care group (risk difference -1·15%, 95% CI lower bound -6·89). Among participants in the combined 6-month PrEP dispensing group, 257 (78%) participants refilled PrEP compared with 134 (81%) participants in the standard of care group (-2·60%, -8·88), and 200 (61%) participants were adherent to PrEP compared with 95 (57%) participants in the standard of care group (2·37%, -5·05). No participants acquired HIV during the study.

Interpretation: 6-month PrEP dispensing with HIVST for interim testing reduced the number of PrEP clinic visits in half without compromising HIV testing, retention, or adherence.

Funding: US National Institute of Mental Health.

Figure 1.. CONSORT diagram

Abbreviations: HIV self-testing (HIVST); HIV serodifferent (SD) ¹Participants randomized to these intervention arms received six-month PrEP dispensing + interim HIVST (either oral-fluid or blood-based) and semiannual clinic visits.

Figure 2.. The effect of six-month PrEP dispensing + interim HIVST on recent HIV testing, PrEP refilling, and PrEP adherence at six months

These risk differences (RDs) compare the combined six-month PrEP dispensing + interim HIV self-testing (HIVST, either oral-fluid or blood-based) arm with the SOC PrEP dispensing arm at six months. The red lines indicate the marker of significance. The bolded RDs and 95% confidence interval (CI) lower bounds (LB) indicate the primary pre-specified outcomes: recent HIV testing (green bars), PrEP refilling (blue bars), and PrEP adherence (red bars) among all participants (N=495). We additionally pre-specified analysis of these outcomes among the following sub-groups: individuals taking PrEP who were members of HIV serodifferent couples (N=295), all women (N=330, including those enrolled in couples and those singly enrolled), and women singly enrolled (N=200).