Background: HPTN 071 (PopART) is a community-randomised trial of the impact of a combination HIV prevention intervention on population-level HIV incidence, conducted in Zambia and South Africa (SA). The PopART intervention is delivered in three annual rounds (R1-R3) by Community HIV-care Providers (CHiPs) who provide home-based HIV testing services, referral for HIV care and antiretroviral therapy (ART), and follow up HIV+ clients to provide adherence and retention support. The aim of this analysis was to provide estimates of retention on ART among individuals in seven communities randomised to receive the PopART intervention, where routine viral load testing was not readily available.
Methods: The analysis included individuals aged 18+ from R3 (September 2016-December 2017) known to the CHiPs as HIV+ and had self-reported ever taking ART. A cross-sectional measure of retention during R3 was defined as the proportion who, on the date of the R3 visit, self-reported taking ART within the last month and had not missed pills in the last three days. Retention at 6/12 months was defined using an individual''s self-reported ART start date and the CHiP visit closest to 6/12 months after this date.
Results: In the four Zambian communities, on the date of the R3 visit 95.9% (9,491/9,902) adults were retained on ART. Restricting to those who initiated ART after the start of the PopART intervention in 2014, 94.1% (4,882/5,186) were retained on ART. Corresponding proportions in the three SA communities were 94.6% (4,339/4,651) and 93.5% (2,270/2,427)(table 1).
In Zambia retention in care 6/12 months following ART start was 93.9%/95.4% compared with 90.9%/91.1% in SA. A difference was observed whether an individual first participated in the PopART intervention in R3 or had participated previously in R1 and/or R2. Those participating previously had a lower proportion retained than those participating for the first time (e.g. 6 month retention in Zambia was 92.6% in previous participants compared to 98.5% in first time participants, with 89.4% vs. 96.0% in SA).
Conclusions: Overall self-reported retention was high. Estimates from individuals who participated in R1 and/or R2 are considered more reliable, as they use a history of CHiP follow-visits rather than a single self-report during R3.

ART initiation timingStarted ART anytimeStarted ART pre-2014Started ART in 2014 or later
Time at which retention was measuredAt time of R3 visitAt time of R3 visitAt time of R3 visit6 Months after ART start date12 months after ART start date
Zambia% Retained (All)95.9 (9,491/9,902)97.7% (4,609/4,716)94.1% (4,882/5,186)93.9% (3,209/3,419)95.4% (2,918/3,058)
 %Retained (First participated Rd3)94.3% (2,765/2,931)96.8% (1,185/1,224)92.6% (1,580/1,707)98.5% (712/723)99.5% (589/592)
 % Retained (Prev participated in Rd1/2)96.5% (6,726/6,971)98.1% (3,424/3,492)94.9% (3,302/3,479)92.6% (2,497/2,696)94.4% (2,329/2,466)
SA% Retained (All)94.6% (4,339/4,651)95.7% (2,129/2,224)93.5% (2,270/2,427)90.9% (1,365/1,502)91.1% (1,302/1,429)
 %Retained (First participated Rd3)95.1% (1,320/1,388)95.0% (575/605)95.2% (745/783)96.0% (316/329)95.6% (280/293)
 % Retained (Prev participated in Rd1/2)94.4% (3,079/3,263)96.0% (1,554/1,619)92.8% (1,525/1,644)89.4% (1,049/1,173)90.0% (1,022/1,136)
[Self-reported retention on ART in the seven communities in Zambia and S Africa in the HPTN 071 (PopART) trial]

Download the e-Poster