Background: HIV self-testing (HIVST) has the potential to help achieve the UNAIDS 90-90-90 targets. However, evidence is limited about how to ensure linkage to care (LTC) among individuals with HIV-positive results. We report LTC findings from a HIVST cluster-randomised trial nested within the HPTN071 (PopART) trial in Zambia. The PopART intervention was delivered in “annual rounds” from 1/12/2013-31/12/2017, during which community-HIV-care-providers (CHiPs) visited all households, and offered home-based HIV testing with a rapid diagnostic test using fingerprick blood (RDT), referral to routine clinic services, and support (including follow-up visits) for LTC.
Methods: In December 2016, 4 of the PopART intervention communities in Zambia, comprising 66 zones, were included in a cluster-randomised trial of adding oral HIVST to the standard intervention. Self-testing was offered in-person, supervised or unsupervised, and to absent partners via secondary distribution. We estimated the time from CHiP referral to LTC among individuals who were (re-)enumerated as a household member during 1/2/2017-30/4/2017, aged ≥16 years, and diagnosed HIV-positive based on initial or confirmatory (following HIVST) RDT. We used the Kaplan-Meier method for “time-to-event” analysis, and follow-up information to 30/9/2017.
Results: Among 13,267 individuals in 33 HIVST zones, 195 were diagnosed HIV-positive; additionally 20 tested HIV-positive with supervised/unsupervised self-testing but did not have confirmatory RDT, and 13 tested HIV-positive following secondary distribution but were not contacted in-person by CHiPs. Among 13,706 individuals in 33 non-HIVST zones, 204 were diagnosed HIV-positive. Among those diagnosed, 94% (184/195) in HIVST and 98% (199/204) in non-HIVST zones were referred to care. We estimated that 65% in HIVST, and 64% in non-HIVST, zones were LTC by 3 months after referral (hazard ratio 1.11, 95%CI 0.78-1.58; Figure 1,Table 1). In HIVST zones, there was a suggestion that LTC was slower for individuals who tested with unsupervised self-testing or via secondary distribution, compared with those who tested with RDT (Table 1).
Conclusions: LTC following an HIV-positive diagnosis and CHiP referral was not undermined by offering HIVST as a testing option, in the context of LTC support. Strategies to facilitate confirmatory RDT following an initial HIVST, and LTC following unsupervised self-testing and secondary distribution, may be important.

  Referred to HIV care (n/N, %)Linked to HIV care, by months after referral (%) 
   136Hazard ratio, 95% confidence interval
OverallNon-HIVST199/204, 98%50.363.876.61 (ref)
 HIVST184/195, 94%52.664.878.21.11 [0.78-1.58]
Overall, HIVST zonesRDT85/88, 97%51.271.478.91 (ref)
 Supervised80/86, 93%57.568.180.81.04 [0.66-1.63]
 Unsupervised14/16, 87%45.645.645.60.47 [0.19-1.14]
 Secondary distribution5/5, 100% [0.03-1.76]
"Minimum estimate" of linkage to HIV careNon-HIVST199/204, 98% 82/199 41.2%  
 HIVST184/195, 94% 82/184 44.6%  
[Table 1 - Time to link to HIV care after CHiP referral, by whether individual was referred from a HIVST or non-HIVST zone]

Figure 1 - Time to link to HIV care after CHiP referral, among individuals who were diagnosed HIV-positive by CHiPs
[Figure 1 - Time to link to HIV care after CHiP referral, among individuals who were diagnosed HIV-positive by CHiPs]

Download the e-Poster