HIV self-testing among female sex workers in Zambia: A cluster randomized controlled trial

BACKGROUND HIV self-testing (HIVST) may play a role in addressing gaps in HIV testing coverage and as an entry point for HIV prevention services. We conducted a cluster randomized trial of 2 HIVST distribution mechanisms compared to the standard of care among female sex workers (FSWs) in Zambia. METHODS AND FINDINGS Trained peer educators in Kapiri Mposhi, Chirundu, and Livingstone, Zambia, each recruited 6 FSW participants. Peer educator-FSW groups were randomized to 1 of 3 arms: (1) delivery (direct distribution of an oral HIVST from the peer educator), (2) coupon (a coupon for collection of an oral HIVST from a health clinic/pharmacy), or (3) standard-of-care HIV testing. Participants in the 2 HIVST arms received 2 kits: 1 at baseline and 1 at 10 weeks. The primary outcome was any self-reported HIV testing in the past month at the 1- and 4-month visits, as HIVST can replace other types of HIV testing. Secondary outcomes included linkage to care, HIVST use in the HIVST arms, and adverse events. Participants completed questionnaires at 1 and 4 months following peer educator interventions. In all, 965 participants were enrolled between September 16 and October 12, 2016 (delivery, N = 316; coupon, N = 329; standard of care, N = 320); 20% had never tested for HIV. Overall HIV testing at 1 month was 94.9% in the delivery arm, 84.4% in the coupon arm, and 88.5% in the standard-of-care arm (delivery versus standard of care risk ratio [RR] = 1.07, 95% CI 0.99-1.15, P = 0.10; coupon versus standard of care RR = 0.95, 95% CI 0.86-1.05, P = 0.29; delivery versus coupon RR = 1.13, 95% CI 1.04-1.22, P = 0.005). Four-month rates were 84.1% for the delivery arm, 79.8% for the coupon arm, and 75.1% for the standard-of-care arm (delivery versus standard of care RR = 1.11, 95% CI 0.98-1.27, P = 0.11; coupon versus standard of care RR = 1.06, 95% CI 0.92-1.22, P = 0.42; delivery versus coupon RR = 1.05, 95% CI 0.94-1.18, P = 0.40). At 1 month, the majority of HIV tests were self-tests (88.4%). HIV self-test use was higher in the delivery arm compared to the coupon arm (RR = 1.14, 95% CI 1.05-1.23, P = 0.001) at 1 month, but there was no difference at 4 months. Among participants reporting a positive HIV test at 1 (N = 144) and 4 months (N = 235), linkage to care was non-significantly lower in the 2 HIVST arms compared to the standard-of-care arm. There were 4 instances of intimate partner violence related to study participation, 3 of which were related to HIV self-test use. Limitations include the self-reported nature of study outcomes and overall high uptake of HIV testing. CONCLUSIONS In this study among FSWs in Zambia, we found that HIVST was acceptable and accessible. However, HIVST may not substantially increase HIV cascade progression in contexts where overall testing and linkage are already high. TRIAL REGISTRATION ClinicalTrials.gov NCT02827240.