Abstract Body

With UNAIDS’s 90-90-90 goal, high yield approaches to HIV testing and counseling (HTC) are critical. Partner tracing, where sexual partners of HIV-diagnosed clients are contacted by their partner or a health provider, is effective in identifying persons with undiagnosed HIV infection.  A Malawi study found a 64% infection rate among sexual partners of index clients; and in a Cameroon study half (50.1%) were infected.  Partner services for STI prevention have been utilized in the US and Europe, but this approach is underutilized in Sub-Saharan Africa (SSA). This study evaluated feasibility, acceptability and effectiveness of partner tracing in facility-based HTC in Tanzania.

The study was conducted in 3 hospitals in Njombe (Tanzania’s highest HIV prevalence region) from June to September 2015. Newly HIV-diagnosed men and women, tested through PITC or VCT, were enrolled as index clients, and offered the choice of referring or bringing in their partner for HTC, or having a health provider anonymously contact their partner with the recommendation to come for HTC. Partners presenting to the facility were offered HTC and referred into HIV care and treatment if found HIV positive.

4,832 individuals were tested for HIV. 765 (15.8%) tested positive; 643 consented to participate. Of these, 251 were ineligible for partner referral including for being underage (31); not having a sexual partner in the last 12 months (148); being mentally unstable (18); being at risk of intimate partner violence (8) or other (46). 387 (99.7%) eligible index clients participated in the listing and referral process. 392 partners were listed, of whom 242 (61.7%) came to the facility; 222 (91.7%) through index client referral and 20 (8.3%) after a health provider contacted them. Of the 228 (94.2%) partners who were tested for HIV, 144 (63.1%) tested positive, all newly diagnosed.  HIV positivity rate did not differ significantly (p=0.5) among male (56.9%) and female partners (61.4%) (See Table 1)

This study demonstrated feasibility, acceptability and effectiveness of partner notification/referral for HTC in facility-based HTC in Tanzania. Given the need for high yield of  identifying previously undiagnosed HIV infected individuals in the context of reaching 90-90-90 goals, and noting that notification was primarily by index clients, the findings present strong evidence for integrating partner notification and testing into facility-based HTC services in Tanzania and similar settings in SSA.