CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

clinic for free HIV Counseling and Testing (HCT) or, 2) choice of free HCT or HIV Self-testing (HIVST) kits. Young women choosing HIVST in the choice arm were provided with 5 HIVST kits (OraQuick); young women randomized to or choosing HCT were given 5 invitations to test for free at local clinics. Four kits/ invitations were intended for distribution to peers and partners plus one for themselves. Young women were invited to return 3 months after enrollment to assess testing uptake between the two arms and distribution to peers and partners. We examined differences in testing between arms using Wald crude risk differences and crude risk ratios. Results: We randomized 144 young women to the HCT arm and 140 to the HIVST/choice arm. Of those randomized to choice, 134 (96%) chose HIVST and 6 chose HCT. By September 21, 2017, 247 women had returned for their 3-month visit (121 in the HIVST/choice arm, 126 in the HCT arm). At the 3-month visit, 97% of women in the HIVST/choice arm reported testing compared to 48% of women in the HCT arm, a risk difference of 48% (Relative Risk 2.00 95% CI 1.66-2.40). These women reported inviting 465 peers (80% female) and 35 partners to test-- 170 (34%) by HCT arm participants and 330 (66%) by choice arm participants. Conclusion: We found that providing young women with a choice to self-test in addition to the option of clinic-based HCT led to 97% testing uptake within three months-virtually all through self-testing. In comparison, those offered HCT alone reported only half that amount of testing. In addition, we saw substantially more peer-referrals among women offered HIVST compared to the HCT arm. Many countries in sub-Saharan Africa are considering offering HIVST as another HIV testing option; we present strong evidence that this strategy will result in a substantial increase in HIV testing among young people compared with current practice. 993 FEMALE SEX WORKERS’ INTERPRETATIONS OF HIV SELF-TEST RESULTS: A PERFORMANCE STUDY Katrina F. Ortblad 1 , Daniel K. Musoke 2 , Thomson Ngabirano 3 , Aidah Nakitende 2 , Jessica E. Haberer 4 , Margaret McConnell 1 , Till Bärnighausen 5 , Catherine E. Oldenburg 6 1 Harvard University, Boston, MA, USA, 2 International Research Consortium, Kampala, Uganda, 3 Uganda Health Marketing Group, Kampala, Uganda, 4 Massachusetts General Hospital, Boston, MA, USA, 5 Heidelberg University, Heidelberg, Germany, 6 University of California San Francisco, San Francisco, CA, USA Background: HIV testing is essential for HIV treatment and prevention. Oral HIV self-testing increases HIV testing among female sex workers (FSWs) in Kampala, Uganda. However, for testers to learn their true HIV status through self-testing, they must correctly interpret HIV self-test results. Previous studies, not conducted among FSWs, may have overestimated HIV self-test performance because participants interpreted their own self-test results and these interpretations may have been biased by prior knowledge of HIV status. We evaluate the performance of FSWs’ interpretation of HIV self-test result images. Methods: Peer educators trained FSWs on how to use oral HIV self-tests and interpret the results; participants then had two opportunities to HIV self-test. One month after receiving the second test, we gave participants color images of oral HIVST results (strong HIV-negative, strong HIV-positive, inconclusive, and weak HIV-positive), identical to those in the manufacturer’s instruction guide, and asked them to interpret these. We calculated the sensitivity, specificity, negative predictive value (NPV) and positive predictive value (PPV) of FSWs’ interpretations. We estimated associations between participant characteristics and incorrect interpretation of HIV self-test results. Results: From October-November 2016, 544 FSWs completed HIV self-test training. Incorrect interpretation of strong HIV-negative, strong HIV-positive, inconclusive, and weak HIV-positive self-test results was 15%, 18%, 23% and 61%, respectively. FSWs’ HIV self-test interpretation had 82% sensitivity (95%CI: 79-86%), 85% specificity (95%CI: 82-88%); 92% NPV (95%CI: 89%-94%); and 68% PPV (95%CI: 64%-71%). Literacy and higher educational attainment were not associated with improved self-test result interpretation. Conclusion: FSW commonly misinterpreted HIV self-test results, although they received realistic pre-self-test training. Understanding HIV self-test performance in isolation from prior HIV status knowledge is important because HIV self-tests may move HIV testing outside healthcare facilities, where they may be used for first time testing or to test others in this unregulated environment. As HIV self-testing becomes a routine component of national HIV responses in sub-Saharan Africa, strong HIV self-test training and support interventions are urgently needed to ensure correct interpretation of self-test results.

994 HIV SELF-TESTING AMONG MEN WHO HAVE SEX WITH MEN AND TRANSGENDER WOMEN IN MYANMAR Andrea L. Wirtz 1 , Soe Naing 2 , Emily Clouse 1 , Hsu Hnin Mon 1 , Aung Zayar Paing 2 , Stefan Baral 1 , Chris Beyrer 1 1 Johns Hopkins University, Baltimore, MD, USA, 2 International HIV/AIDS Alliance Myanmar, Yangon, Myanmar Background: Efforts to improve HIV diagnosis rely on innovative interventions, particularly for key populations. The HIV epidemic in Myanmar is concentrated among men who have sex with men (MSM) and transgender women (TW) and national efforts now focus on improving engagement in HIV testing and care. This implementation science study tested the acceptability and use of HIV self-testing (HIVST) to address care continuum losses by increasing HIV testing uptake to aid early diagnosis of infection. Methods: We implemented a randomized trial in which HIV-uninfected MSM and TWwere recruited via respondent-driven sampling in Yangon. Participants completed a baseline survey and were randomized to standard, voluntary counseling and testing (VCT) or to HIVST. To mitigate stigma, VCT-assigned participants were referred for testing at community-based organizations (CBO) serving MSM and TW. Biologic specimens were collected for confirmatory testing. Participants were asked to return to the study to report their HIV test result and the acceptability of their assigned testing method. Results: A total of 577 MSM (84.7%) and TW (15.3%) participants were enrolled and randomized to VCT or HIVST between November 2015-July 2017. Self-reported HIV risk behavior was high: 29.8% had engaged in sex work (last 6 mo.); condom use at last sex was less than 30.0% for all partner types; and 32.8% had ever been tested for HIV. 342 (59.3%) returned for the second study visit to report test acceptability. VCT-assigned participants were marginally less likely to return, compared to HIVST participants (45.9% vs. 54.0%; p=0.055). HIVST participants were more likely to agree that, overall, their testing method was easy to implement and understand (98.4% vs. 95.4%; p=0.002). The majority (88.8%) of VCT-assigned participants to indicated they would test regularly if they could access HIVST. HIVST participants were more likely to report that HIVST would be the preferable testing modality for future HIV testing; HIVST was also favored for future testing by VCT participants (55.4%), followed by CBO clinics (36.3%) and government facilities (7.0%). HIVST identified 29 previously undiagnosed infections (9.9%) compared to 15 identified by VCT (5.3%; p<0.001). Conclusion: HIVST is an acceptable, alternative testing modality compared to community-based VCT for MSM and TW in Myanmar. Likely, HIVST may have greater acceptability and effectiveness compared to testing in government facilities, where stigmatization of key populations is common. 995 EVALUATION OF TARGETED HIV SELF-TEST KIT DISTRIBUTION VIA A DIGITAL VENDING MACHINE Alex Pollard 1 , Gillian Dean 2 , Suneeta Soni 2 , Carrie Llewellyn 1 , Carlos Peralta 3 , Liliana Rodriguez 4 , Jaime Vera-Rojas 2 1 Brighton and Sussex Medical School, Brighton, UK, 2 Brighton & Sussex University Hospitals NHS Trust, Brighton, UK, 3 University of Brighton, Brighton, UK, 4 Loughborough University, Loughborough, UK Background: Novel strategies are needed to increase HIV testing in high-risk groups including men who have sex with men (MSM), and to meet goals to

Poster Abstracts

CROI 2018 380