CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

reduce undiagnosed HIV. HIV self-testing is an attractive strategy enabling user autonomy in the timing, location and disclosure of testing as well as convenience. Self-testing also provides opportunities for providers to reach populations not engaged with conventional testing. We developed and evaluated a vending machine approach to target HIV self-testing to high-risk MSM in a UK community setting. Methods: A cross-sectional survey in a sex-on-premises venue (sauna) assessed feasibility and informed development of a vending machine interface. Co-design workshops between designers and LGBT community volunteers explored attitudes towards self-testing and a vending machine interface to deliver HIV self-test kits in community settings (bars/clubs, pharmacies, university campuses and train stations). A bespoke vending machine distributing BioSure© self-test kits was installed in a sauna in this high-HIV- prevalence city (Brighton). A cross-sectional mixed methods evaluation was conducted with 30 users. Demographics were collected via the machine’s user- friendly touch-pad screen. An online questionnaire and structured interviews gathered information on user-experience of the machine, and experience, acceptability and attitudes towards HIV self-tests accessed via a machine. Results: The survey and co-design workshops found that 32% of 281 sauna users had never tested for HIV, despite high infection risks. Acceptability of self-testing before installation of the vending machine was 93%. A total of 95 testing kits were accessed between 8th July and 25th Sept 2017: mean age 35 (Range 18-65); 7.4% (n=7) had never tested for HIV before; 15.8% (n=15) had tested within the last 2-5 years. Uptake of tests was higher via the vending machine compared to HIV testing conducted by community outreach workers in the same venue and study period (95 vs 12). Qualitative interview and online questionnaires demonstrated high acceptability and support for this intervention, which was considered accessible and appropriately targeted. Conclusion: Community co-design supported the development of an acceptable vending machine interface for the distribution of HIV self-testing kits. This delivered low-cost HIV self-tests to men with low levels of prior testing; and represents an acceptable targeted distribution method that could be applied in other settings. Nurilign Ahmed 1 , Lawrence Mwenge 2 , Linda Sande 3 , Collin Mangenah 4 , Sarah Kanema 2 , Mutinta Nalubamba 5 , Sepiso Libamba 5 , Euphemia Sibanda 4 , Hendramoorthy Maheswaran 6 , Cheryl Johnson 7 , Karin Hatzold 8 , Elizabeth L. Corbett 3 , Helen Ayles 2 , Fern Terris-Prestholt 1 1 London School of Hygiene & Tropical Medicine, London, UK, 2 Zambart, Lusaka, Zambia, 3 Malawi–Liverpool–Wellcome Trust Clinical Rsr Prog, Blantyre, Malawi, 4 Centre for Sexual Health and HIV/AIDS Research Zimbabwe, Harare, Zimbabwe, 5 Society for Family Health, Lusaka, Zambia, 6 University of Liverpool, Liverpool, UK, 7 WHO, Geneva, Switzerland, 8 Population Services International, Harare, Zimbabwe Background: HIV self-testing is a process whereby a person who wants to know his or her HIV status collects a specimen, performs the test and interprets the test result him/herself. The HIV Self-Testing Africa (STAR) project in Zambia utilised community-based distribution agents (CBDAs), voluntary medical male circumcision (VMMC) and health facility (HF) services to distribute HIVST kits. We present the cost of HIVST kits distribution strategies and examine the key cost drivers in Zambia. Methods: We analysed project financial expenditure data between July 2016 and May 2017. All costs are presented in 2016 US$. Total and unit costs per kit distributed were estimated for CBDA, VMMC, and HF distribution. Results: Over the lifespan of the project, 127,804 HIVST kits were distributed across 16 communities and distribution modes in Zambia, with CBDA distributing the vast majority of kits (81%), followed by HF (10%) and VMMC (9%). Personnel, HIVST prices and transport were the key cost drivers accounting for 49%, 24% and 12% of the total costs, respectively. The unit cost per HIVST kits distributed were $15.19 for CBDA, $14.71 for VMMC and $15.07 for the HF. Conclusion: Our findings show that HIVST can be distributed within communities at a reasonable cost. Though higher than our previous estimates for facility-based testing (~$4.24 (Mwenge 2017), it reduces users’ costs of testing, estimated in Malawi to be as high as $4 among men (Sande 2017) and therefore addresses, among others, financial barriers to testing. This lays the

foundation for exploring economic efficiency in HIVST distribution modalities: it is expected that additional cost savings may be achieved through economies of scale by increasing the volume of the total population covered. Further research is needed to evaluate its cost-effectiveness and how distribution costs will change as programmes mature and scale up.

Poster Abstracts

997 COMPARING SELF-REPORTED VIRAL LOAD WITH HOME-COLLECTED HIV-RNA AMONG HIGH-RISK MSM Richard A. Teran 1 , Martin J. Downing 2 , Mary Ann Chiasson 2 , Hong Van Tieu 3 , Laura Dize 4 , Charlotte A. Gaydos 4 , Sabina Hirshfield 2 1 Columbia University Medical Center, New York, NY, USA, 2 Public Health Solutions, New York, NY, USA, 3 New York Blood Center, New York, NY, USA, 4 Johns Hopkins University, Baltimore, MD, USA Background: Among men who have sex with men (MSM), HIV transmission is attributed to both sexual risk behavior and unsuppressed HIV viral load (VL). In the US, an estimated 60% of HIV-positive MSM are virally unsuppressed. Studies show that self-perceived VL may influence sexual practices of HIV-positive MSM. Few US studies have used dried blood spot (DBS) collection outside of a clinical setting; however, improved DBS materials have streamlined biospecimen home collection fromMSM recruited online. The current study compared self-reported VL with a lab-based HIV VL test from an at-home DBS kit. Methods: From 09/2016-06/2017, US HIV-positive MSM (n=766) completing a 12-month online HIV prevention risk reduction intervention were invited to enroll in an at-home DBS collection study. Consenting participants were mailed a HemaSpot kit and instructed to provide a blood sample and return it by mail to a research laboratory. DBS samples were tested using an Open Mode protocol for DBS (1.0ml HIV-1 RNA DBS IUO US TT v11) on the m2000sp/m2000rt system. Results: Of 554 consenting participants, 418 (75%) returned a kit for lab testing; of those, 337 (81%) had sufficient blood for HIV VL quantification. Of MSM returning a quantifiable kit, 71%were White, 12% Black, and 17% Hispanic. Median age was 38. Among 314 MSM self-reporting current antiretroviral therapy (ART), 49% had <90% adherence in the past month. Of the 337 quantifiable kits, 53% had a detectable VL: 99 kits (29%) had <832 copies/ml; 9 kits (3%) had 832-999 copies/ml; and 69 kits (20%) had >1,000 copies/ml [range: 1,023-1,202,265 copies/ml; median: 5,129 copies/ml]. Among men self-reporting an undetectable HIV VL (n=284), 48% returned a DBS kit with detectable VL; 13% of kits had >1,000 copies/ml. Median time between discordant self-reported undetectable VL and a lab-quantified VL was 43 days. Men living with HIV for >1 year were more likely to have a discordant self- reported and lab-quantified VL (52% vs. 31%; p<0.01). Conclusion: Over half of home-collected DBS samples from HIV-positive MSM had a detectable VL despite most self-reporting ART. About half of men self- reporting an undetectable VL had a detectable lab-quantified VL, and were not treatment naïve, indicating a need for novel approaches to provide treatment services to high-risk HIV-positive MSM with a detectable VL who may be experiencing ART fatigue or not in care. Discordant self-reported and DBS-based VL in our study also highlights the need for validation of self-reported data.

996 COST ANALYSIS OF DIFFERENTIATED HIV SELF-TESTING KITS DISTRIBUTION IN ZAMBIA

CROI 2018 381