Abstract Body

Testing key populations (KPs) for HIV is essential to achieving the first of the UNAIDS 90-90-90 goals. Identifying and testing partners of HIV-infected individuals, or assisted partner services (aPS), is an efficient tool for case-finding. aPS has not been used among people who inject drugs (PWID), one of the highest risk KPs. We determined whether aPS could find, test, and link to care the injecting and sexual partners of HIV-infected PWID in Nairobi, Kenya.

Recruitment of index participants (indexes) occurs at 3 needle and syringe exchanges in Nairobi. Indexes provide contact information for injection and sexual partners in the past 3 years. Blinded to the index’s identity, community-embedded peer educators (CEPEs) attempt to contact all named partners, first by phone, then community tracing. When partners are contacted, they are notified of their possible HIV exposure and are offered HIV counseling and testing. Participants also complete questionnaires and are offered rapid hepatitis C (HCV) testing. To examine aPS effectiveness, we determined the number of indexes needed to be interviewed (NNTI) to find a 1) first-time tester; 2) new HIV case; 3) known HIV-positive person not on treatment.

To date, 149 indexes have enrolled who have named 440 partners (Figure 1). Of named partners 332 (76%) have been traced and enrolled. Because partners could enroll multiple times if named by multiple indexes, the 332 enrolled partners represented 274 unique individuals, of whom 194 (71%) were injection partners (IPs), 35 (13%) sexual partners (SPs), and 45 (16%) injection and sexual partners (ISPs). Among partners 63 (22%) were HIV-infected, of whom 7 (11%) were unaware of their status and 8 (13%) were aware but not on ART. HIV prevalence was highest among IPs and ISPs (24%) and lower among SPs (11%). NNTI was 19 per first-time tester, 21 per new HIV case, and 10 per HIV-infected person not on ART. HCV Ab was found in 50 (33%) indexes and 57 (20%) partners. Confirmatory RNA tests are pending. Almost all partners required in-person tracing, as they could not be reached by phone. No adverse events have been reported related to aPS.

aPS using CEPEs is an effective tool for finding and testing high-risk partners of PWID. Nearly one quarter of partners reached were HIV-infected. Among these, one quarter did not know his/her status or was not on ART. We conclude that aPS is a novel testing strategy that may reduce HIV transmission and promote engagement in care among PWID.