CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

analysis. More than 80% of women who reported exchange sex in the past 12 months were living below the federal poverty line, homelessness ranged from 41.9% in Detroit to 65.3% in Seattle, and injection drug use from 3.8% in Houston to 61.4% in Seattle. Aggregated HIV prevalence for women who exchanged sex was 4.9% (CI 2.7%-7.1%), for women of LSES who did not exchange sex was 1.6% (CI 0.3%-2.8%), for women in the general population was 0.56% (CI 0.55%-0.58%). HIV prevalence among women who exchange sex was 3.1 times (CI 1.6-5.9) as high as among women of LSES who do not exchange sex, and 8.7 times (CI 6.9-10.9) as high as among women in the general population. Conclusion: A high proportion of women who exchanged sex in this survey were homeless and lived in poverty. Their HIV prevalence was significantly higher compared to women in the general population and women of LSES who do not exchange sex. There is a need for further research of the role of exchange sex and LSES in the vulnerability for HIV among women and development of tailored interventions addressing these factors. 921 TRANSACTIONAL SEX MEASUREMENT AND ASSOCIATION WITH HIV INCIDENCE AMONG WOMEN Mary K. Grabowski 1 , Virginia Burke 1 , Gertrude Nakigozi 2 , Fred Nalugoda 2 , Robert Ssekubugu 1 , Larry W. Chang 3 , Godfrey Kigozi 2 , Deanne Kerrigan 1 , David Serwadda 4 , Betty Nantume 2 , Darix K. Ssebagala 2 , Ronald H. Gray 1 , Joseph Kagaayi 2 , Maria Wawer 1 , Caitlin E. Kennedy 1 1 Johns Hopkins University, Baltimore, MD, USA, 2 Rakai Health Sciences Program, Kalisizo, Uganda, 3 Johns Hopkins Hospital, Baltimore, MD, USA, 4 Makerere University, Kampala, Uganda Background: Transactional sex (TS) has been defined as “an exchange of money, favors, or gifts for sexual relations” and may overlap with sex work, but is often considered distinct and less formal. However, no population-based studies have documented the prevalence of TS among women in sub-Saharan Africa and its relationship with HIV incidence. We sought to compare reporting of TS using different survey measures, and assess the association between TS and HIV infection. Methods: We compared trends in reporting TS over time among sexually active women ages 15-49 in the Rakai Community Cohort Study, a longitudinal open cohort in Rakai, Uganda. From 1999-2013, women were asked if they had ever exchanged money or gifts for sex with any partner (global measure). Beginning in 2014, women were asked if they had ever exchanged money, gifts, or favors with each of their four most recent sexual partners (partner-specific measure). We then used 2014-16 data to assess the association between the partner- specific measure and HIV prevalence and incidence using Poisson regression with and without adjustment for demographics. Results: Reporting of TS among 19,155 women (58,414 woman-visits) in 30 continuously surveyed communities increased significantly with use of the partner-specific questions, from 0.5% to 8.9% of visits (p<0.001). Among 11,138 women surveyed in 2014-16, TS was reported at 13% (2,102/16,245) of visits. Only 4 women (0.03%) reported their occupation as sex workers. Women reported TS with husbands at 8% (952/11,667) of visits compared to 20% (1,150/5,631) for non-marital partners. There were 179 incident HIV cases identified, including 31 among women reporting any TS. Women reporting TS with marital partners had somewhat higher HIV prevalence (adjPRR=1.12, 95%CI:1.00-1.25) but not incidence compared to women not reporting TS with marital partners. Among women with non-marital partners, TS was also associated with higher prevalence regardless of community (adjPRR:1.10, 1.01- 1.21). HIV incidence increased only among women who reported TS with non- marital partners in fishing communities (11 vs. 6 per 100 woman-years), but this association was not statistically significant after adjustment (adjIRR=1.65, 0.92-2.98). Conclusion: Asking about TS using a partner-specific measure may elicit more accurate reporting than a global measure. TS may modestly correlate with HIV, but there is a need for more nuanced understanding of what dimensions of social relationships and HIV risk this measure captures.

Poster Abstracts

919 WITHDRAWN 920 HIV PREVALENCE AMONG WOMEN WHO EXCHANGE SEX RECRUITED IN 4 US CITIES, 2016 Senad Handanagic 1 , Brooke Hoots 1 , Kristen Leigh Hess 1 , Emily Higgins 2 , Paige Padgett 3 , Sarah L. Braunstein 4 , Sara N. Glick 5 , Dita Broz 1 , Gabriela Paz-Bailey 1 , Cyprian Wejnert 1 , Lina Nerlander 1 1 CDC, Atlanta, GA, USA, 2 Michigan Department of Health and Human Services, Lansing, MI, USA, 3 University of Texas at Houston, Houston, TX, USA, 4 New York City Department of Health and Mental Hygiene, Long Island City, NY, USA, 5 University of Washington, Seattle, WA, USA Background: Women who exchange sex have a high HIV prevalence in many countries, but limited data are available in the United States. In 2016, sampling for National HIV Behavioral Surveillance (NHBS) among heterosexuals focused on women who exchange sex for money or drugs in 4 of 22 cities (Chicago, Detroit, Houston and Seattle). We compared the HIV prevalence from this survey to women of low socioeconomic status (LSES) who did not exchange sex recruited as part of the 2013 NHBS cycle among heterosexuals, and to women in the general population of the included cities based on data from the National HIV Surveillance System (NHSS) for 2015. Methods: NHBS participants were recruited via respondent-driven sampling (RDS), interviewed and offered HIV testing. Women with valid HIV results were included in analysis. NHBS data in 2013 were limited to the 4 cities included in 2016. We applied RDS weights to estimate the aggregated HIV prevalence in 2016 among women who exchange sex, and in 2013 among women of LSES who do not exchange sex. HIV prevalence in the general population was calculated using census and NHSS data. We used prevalence ratios to compare the HIV prevalence in 2016 among women who exchanged sex to women of LSES who did not exchange sex and women in the general population. Results: In total, 1,440 women who reported exchange sex in 2016 and 671 women of LSES who did not report exchange sex in 2013 were included in the

CROI 2018 351