Abstract

Effect of an opt-out point-of-care HIV-1 nucleic acid testing intervention to detect acute and prevalent HIV infection in symptomatic adult outpatients and reduce HIV transmission in Kenya: a randomized controlled trial

Sanders EJ, Agutu C, van der Elst E, Hassan A, Gichuru E, Mugo P, Farquhar C, Babigumira JB, Goodreau SM, Hamilton DT, Ndung'u T, Sirengo M, Chege W, Graham SM
HIV Med. 2022;23

Permenent descriptor

BACKGROUND: In sub-Saharan Africa, adult outpatients with symptoms of acute infectious illness are not routinely tested for prevalent or acute HIV infection (AHI) when seeking healthcare. METHODS: Adult symptomatic outpatients aged 18-39 years were evaluated by a consensus AHI risk score. Patients with a risk score >/= 2 and no previous HIV diagnosis were enrolled in a stepped-wedge trial of opt-out delivery of point-of-care (POC) HIV-1 nucleic acid testing (NAAT), compared with standard provider-initiated HIV testing using rapid tests in the observation period. The primary outcome was the number of new diagnoses in each study period. Generalized estimating equations with a log-binomial link and robust variance estimates were used to account for clustering by health facility. The trial is registered with ClinicalTrials.gov NCT03508908. RESULTS: Between 2017 and 2020, 13 (0.9%) out of 1374 participants in the observation period and 37 (2.5%) out of 1500 participants in the intervention period were diagnosed with HIV infection. Of the 37 newly diagnosed cases in the intervention period, two (5.4%) had AHI. Participants in the opt-out intervention had a two-fold greater odds of being diagnosed with HIV (odds ratio = 2.2, 95% confidence interval: 1.39-3.51) after adjustment for factors imbalanced across study periods. CONCLUSIONS: Among symptomatic adults aged 18-39 years targeted by our POC NAAT intervention, we identified one chronic HIV infection for every 40 patients and one AHI patient for every 750 patients tested. Although AHI yield was low in this population, routinely offered opt-out testing could diagnose twice as many patients as an approach relying on provider discretion.