Background: HIV self-tests (HIVST) may help increase HIV testing coverage to meet the 90-90-90 target in female sex worker (FSW) populations. We report the results of a randomized controlled trial of HIVST among FSW in Zambia.
Methods: Trained peer educators in Kapiri, Chirundu, and Livingstone, Zambia each recruited 6 FSW participants. Peer educator-FSW groups were randomized to: 1) direct distribution of an oral HIVST from the peer educator, 2) distribution of a coupon for an oral HIVST available from a health clinic/pharmacy, or 3) referral to standard HIV testing. HIVST-arm participants received one HIVST at baseline and another three months later. Participants completed baseline, month-1, and month-4 questionnaires.
Results: 965 participants were enrolled between September-October 2016; 20% had never tested for HIV. 98.3% of direct distribution arm participants reported using their HIVST at month-1, compared to 86.3% in the coupon arm (P=0.001); this difference had disappeared by month-4. There was no significant difference in reported past-month testing for HIV at month-1 or month-4, although rates were highest for the direct arm at both timepoints. At month-1, 94.9%, 84.4% and 88.5% of direct-, coupon- and standard-arm participants reported testing in the past month (P=0.10 direct vs. standard, P=0.29 coupon vs. standard). At month-4, past-month testing coverage was 84.1%, 79.8% and 75.1% (P=0.11 direct vs. standard, P=0.42 coupon vs. standard).

Of 144 participants reporting a positive HIV test at month-1, 51.0% and 52.8% in direct- and coupon-arm participants reported linking to care, compared to 74.6% in the standard arm (P=0.07 direct, P=0.12 coupon). At month-4, of 235 participants reporting a positive test, 71.6%, 76.6% and 85.7% of direct-, coupon-, and standard-arm participants reported linking (P=0.13 direct, P=0.17 coupon). Three cases of HIVST-related intimate-partner violence (IPV) were reported, despite 60% of participants reporting IPV in the previous year.
Conclusions: HIVST provision via peer educators to Zambian FSW led to high test uptake and rapid linkage to care, including amongst those who had never previously tested, without a significant increase in IPV. HIVST should be considered as part of an intervention package to maximize HIV protection for FSW populations.