CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Conclusion: Despite efforts to reduce MCP behaviors , they were still common particularly among men. MCP was not associated with recent or prevalent HIV infection. Knowledge of positive HIV status was associated with lower MCP rates. Multiple socio-behavioral HIV prevention strategies are needed to reduce MCP. 928 ARE FISHERFOLK AT HIGHER RISK THAN THEIR NEIGHBORS: FINDINGS FROMWESTERN KENYA, 2015 Emily Zielinski-Gutirrez 1 , Sifunjo F. Odongo 2 , Daniel Kwaro 2 , Kennedy Mutai 2 1 US CDC Nairobi, Nairobi, Kenya, 2 Kenya Medical Research Institute, Kisumu, Kenya Background: Counties bordering Lake Victoria have the highest adult HIV prevalence in Kenya. Within the region, fisherfolk (FF) who catch, sell, or process fish and their spouses are considered a priority population in HIV transmission. We asked if FF differ from their immediate non-fishing neighbors (non-FF) in terms of HIV prevalence, risk behaviors and health service utilization in Siaya county, western Kenya. Methods: A cross-sectional bio-behavioral household survey was conducted at beaches and adjacent villages in the Kenya Medical Research Institute Health and Demographic Surveillance System from August 2014-March 2015. The survey collected demographics, HIV risk behavior and service utilization and offered HIV testing. Bivariate comparisons were used to examine factors of interest, evaluated by Pearson’s chi-square, with stratification by sex as appropriate. Results: Of 3462 participants aged 15-64 years, 940 (27.2%) were FF. Of 3344 respondents with HIV status, 17.1%were HIV positive; prevalence was higher among FF (24.1%) than non-FF (14.7%), p<.001. Most HIV-positive respondents (77.0%) self-reported their status. HIV prevalence was significantly higher among women (20.1%) than men (11.8%, p<.001), with a greater difference by sex among non-FF. HIV prevalence was highest among FF aged 30-49 years (34.4%) with a similar pattern among non-FF, peaking at 28.4% among the same age group. More non-FF men were circumcised than FF men (61.9% vs. 51.9%, p<.001). Among the sexually active, FF men were more likely than no-FF men to report two or more sexual partners in the past 12 months (34.9% vs. 25.9%, p<.006), and all FF were more likely to report no condom use with at least one sexual partner in past 12 mos. (77.2% FF vs. 46.8% non-FF, p<.001). Most respondents had previously tested for HIV (95.2% FFX vs. 92.2% non-FF, p<.001). Among 446 total respondents self-reporting as HIV positive, 78.0% reported taking anti-retroviral therapy (ART), with no significant difference by FF status. When accounting for all HIV positives, ART coverage was 56.9% among FFX and 62.9% among non-FF. Conclusion: HIV prevalence was higher among FF than among non-FF. FF reported higher HIV-related risk such as non-circumcision, more sexual partners and sex without condoms. While ever-testing rates were high, just over half of FF were on ART, suggesting that aggressive scale-up of testing, treatment and prevention interventions targeted for FF are required to meet the needs of this priority population. 929 NATIONAL TRENDS IN HIV PREVALENCE IN 3 KEY POPULATIONS IN UKRAINE Kostyantyn Dumchev 1 , Yana Sazonova 2 , Tetiana Salyuk 2 , Olga Varetska 2 1 Ukrainian Institute on Public Health Policy, Kyiv, Ukraine, 2 Alliance for Public Health, Kyiv, Ukraine Background: Once recognized as the most severe in Europe, the HIV epidemic in Ukraine is concentrated among people who inject drugs (PWID), men who have sex with men (MSM) and female sex workers (FSW). Integrated bio-behavioral surveys (IBBS), as a part of the second generation surveillance, are used to monitor HIV infection and risk behaviors in key populations since 2002. This analysis focused on HIV prevalence trends in four latest nationally- representative rounds of IBBS in Ukraine. Methods: Each of the four IBBS studies in 2008-9, 2011, 2013, and 2015 covered all 27 regions of Ukraine, and used respondent-driven sampling to recruit PWID and MSM, and time-location sampling for FSW. HIV seropositivity was determined by a single rapid test. Sociodemographic, HIV risk behavior and drug use characteristics were assessed using an anonymous questionnaire. Chi- square test was used to detect differences between rounds, Mantel-Haenszel test was used to assess significance of the overall trend. Results: Sample sizes and HIV prevalence in total samples, age group younger than 25, and MSM and FSW subgroups who ever injected drugs are presented in Table 1. Over the four rounds, HIV prevalence has decreased significantly

years (PR: 0.59; 95% CI: 0.41-0.85), use alcohol around time of intercourse (PR: 0.59; 95% CI: 0.45-0.76), or have ≥2 partners in the last 12 months (PR: 0.65; 95% CI: 0.57-0.74). Conclusion: Self-reported risky sexual practices of adolescents and young adults in Botswana differed significantly between males and females. Economic stress was strongly associated with increased risk behavior in females and interventions targeting this vulnerability (income transfer, pre-exposure prophylaxis) have promise in Botswana. Programs targeting episodic risk, particularly around alcohol, could be more effective in young males.

927 HIGH PREVALENCE OF CONCURRENT SEXUAL PARTNERSHIPS IN A LARGE POPULATION SURVEY Etienne Kadima 1 , Kathleen Wirth 2 , Kara Bennett 3 , Tendani Gaolathe 1 , Jean Leidner 4 , Unoda A. Chakalisa 1 , Erik VanWidenfelt 1 , Pamela J. Bachanas 5 , Refeletswe Lebelonyane 6 , Mompati O. Mmalane 1 , Joseph Makhema 1 , Molly Pretorius Holme 2 , Shahin Lockman 2 , Max Essex 2 , Sikhulile Moyo 1 1 Botswana Harvard AIDS Inst Partnership, Gaborone, Botswana, 2 Harvard University, Cambridge, MA, USA, 3 Bennett Statistical Consulting, Inc, New York, NY, USA, 4 Goodtables Data Consulting, Norman, OK, USA, 5 CDC, Atlanta, GA, USA, 6 Botswana Ministry of Health, Gaborone, Botswana Background: Multiple and concurrent sexual partnerships (MCP) are considered a driver of new HIV infections in Sub-Saharan Africa. Understanding predictors of MCP could help design prevention strategies to reduce HIV incidence. We investigated predictors of MCP and the association between MCP and recent HIV infection in a large household-based cluster-randomized HIV prevention trial (BCPP) in 30 Botswana communities from Nov 2013–Nov 2015. Methods: In BCPP, we used structured questionnaires to evaluate for MCP over the prior 12 months via: a) standard UNAIDS-recommended method in which MCP was defined as having at least 2 active and overlapping sexual partners in the past 12 months, and b) a direct question regarding whether concurrent sexual relationships existed at the same time. MCP was defined as present if either method indicated MCP. Recent HIV infection was determined using an algorithm including the Limiting-Antigen Avidity Assay, as well a time of less than 2 years between the last documented negative HIV test to first documented positive HIV test. We estimated prevalence ratios and 95% confidence intervals for factors associated with MCP, adjusting for clustering, age and gender. Results: Data were available from 11,965 (94.9%) of 12,610 participants. Among 9,364 who were sexually active in the past 12 months, 2,878 (30.7%; 95%CI: 29.8–31.7) had multiple sexual partners. Among 9,363 with recent partner data, 2,770 (29.6%, 95%CI: 28.7–30.5) had MCP. Those reporting MCP were more likely to be male (prevalence ratio [PR]=1.6; 95%CI: 1.5-1.7), and of median age 32 years (Q1, Q3: 25,41). After adjusting for age and gender, MCP was significantly associated with being employed (PR=1.3; 95%CI: 1.3–1.5), transactional sex (PR=1.7;95%CI:1.5–2.1), intergenerational sex (PR=1.2; 95%CI:1.1–1.3), and high alcohol consumption (PR=1.5; 95%CI:1.4–1.6). MCP was not significantly associated with recent or prevalent HIV infection. Participants who knew their HIV-positive status were less likely to have MCP (PR=0.9;95%CI:0.8-0.96). Among those who believed their partners had other partners during the course of their relationships in the past year, 49.9%were in MCP (PR=2.0;95% CI:1.97-2.2).

Poster Abstracts

CROI 2018 354