Targeted HIV testing modalities in Zimbabwe: Achieving effectiveness through a revised model

Abstract

Introduction HIV testing remains central to epidemic control, but awareness of HIV status in Zimbabwe remains below UNAIDS targets. We evaluated a targeted HIV testing model using a risk-screening tool to triage clients for same-day provider-initiated HIV testing, while routing screen-negative clients to HIV self-testing. Methods We conducted a mixed-methods implementation evaluation across eight provinces from April 2021 to June 2022. Thirteen candidate screening questions were evaluated in 7825 clients and reduced to five domains for assessment of the final combined rule in 2116 complete-case assessment records. The rule screened clients last tested at least 3 months previously, or never tested, who reported at least one additional risk or symptom criterion. We also conducted 20 healthcare-worker interviews and reconstructed the index-testing cascade for 6308 index cases. Results The final combined rule achieved 94.1% sensitivity, 19.5% specificity, 9.2% positive predictive value, 97.4% negative predictive value, 18.0% testing volume reduction, and a number needed to test of 11. In the index-testing cascade, 66.4% of index cases were offered partner testing; 5080 contacts were elicited, 78.6% were tested, 43.5% tested HIV positive, and 91.8% initiated antiretroviral therapy. Positivity was associated with anonymous tracing (aOR 8.46, 95% CI 3.37–22.75), ≤7-day testing (2.78, 2.44–3.18), male sex (3.09, 2.74–3.49), and first-time testing (1.65, 1.43–1.91). Conclusion The targeted model maintained high screening sensitivity while modestly reducing same-day provider-initiated testing volume, and index testing achieved high positivity and ART linkage among tested contacts. The screening rule should be interpreted as a triage mechanism rather than a diagnostic discriminator, while implementation of the broader model requires timely contact tracing, differentiated testing options, and reliable linkage systems.