CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

were recruited through medical clinics, public health agencies, or community- based organizations. We obtained clinical information frommedical or case management records, and estimated location (U.S. vs. outside the U.S.) of HIV acquisition based on clinical data, testing history, dates of immigration and first positive/last negative HIV test, and sexual behaviors before/after immigration. We identified predictors of having >1 previous negative HIV test and U.S. HIV acquisition with log binomial regression. Results: Of 164 participants, 106 (65%) were women. Median age was 43 years (range: 18-72). Fifty-nine (59/159, 37%) reported HIV diagnosis prior to immigration, although this varied by year of arrival (25% before 2010 vs. 49% in 2010 or later, p=0.04). Among 98 persons diagnosed in the U.S., median time from arrival to diagnosis was 2.6 years (range 29 days-31 years). Among all APLWH, 43% (67/156) had >1 negative HIV test prior to diagnosis. Having a college degree compared to < a high school education was associated with a previous test (prevalence ratio [PR}:1.39, 95% CI:1.04-1.87), adjusting for sex, income, childhood family income, and year of diagnosis. Data were sufficient to estimate place of HIV acquisition for 132 (80%) participants (table); 23% definitely or probably acquired HIV in the U.S., while 72% probably or definitely acquired HIV in Africa. Among persons with definite or probable acquisition outside the U.S., median time from immigration to diagnosis was 274 days, and CD4 counts at diagnosis were lower than for others. APLWH who arrived before 2010 (PR: 5.26, CI:1.93, 14.4) and with high childhood family income (vs. low income, PR: 2.93, CI:1.06-8.07) were more likely to be categorized as probably or definitely acquiring HIV in the U.S., adjusted for sex and education. Conclusion: Most cases of HIV among APLWH in the US are acquired in Africa and are diagnosed shortly after arrival in the U.S. However, almost a quarter acquire HIV in the US, highlighting the need for prevention efforts focused on APLWH in the U.S.

33) of women were in an age disparate relationship, compared to 33% (95% CI 30, 37) in 2013. The prevalence of HIV was higher among women in an age disparate relationship (14.0%) than those with partners 0-4 years older (9.8%), but this was not statistically significant after adjustment (adjOR= 1.03, 95%CI 0.89, 1.19; see table). In young women <25 HIV prevalence was higher among women in an age disparate relationship (8.3%) than those with partners 0-4 years older (5.6%), but this difference was not statistically significant after adjustment (adjOR=1.14, 95% CI 0.91, 1.41). Young women had increased odds of HIV infection if they had multiple partners and reported inconsistent condom use. Young women had lower odds of HIV infection if they were more educated (secondary education adjOR=0.32, 95%CI 0.16, 0.62 or tertiary (adjOR=0.27, (95% 0.09, 0.77 vs. no education). Conclusion: The frequency of age disparate relationships among never married women has not changed over a 15-year period and was not associated with increased risk of HIV infection in Rakai, Uganda.

Poster Abstracts

934 HIGH INCIDENCE AND BURDEN OF HIV INFECTION IN EAST LONDON, SOUTH AFRICA Bhakti Hansoti 1 , Anna Eisenberg 2 , David Stead 3 , Nomzamo Mvandaba 4 , Eshan U. Patel 1 , Andy Parrish 3 , Madeleine Whalen 5 , Reinaldo Fernandez 1 , George Mwinnyaa 1 , Steven J. Reynolds 2 , Andrew D. Redd 2 , Stewart Dandorf 1 , Richard Rothman 1 , Oliver Laeyendecker 2 , Thomas C. Quinn 2 1 The Johns Hopkins University, Baltimore, MD, USA, 2 NIAID, Baltimore, MD, USA, 3 Frere Hospital, East London, South Africa, 4 Walter Sisulu University, Mthatha, South Africa, 5 Johns Hopkins Hospital, Baltimore, MD, USA Background: To define the scope of the current HIV epidemic in East London, South Africa, HIV incidence and burden of disease was evaluated in patients attending the Emergency Department (ED) at Frere Hospital. Methods: A cross-sectional, identity-unlinked serosurvey was conducted between September and November of 2016 to determine HIV seroprevalence and incidence at the Frere Hospital ED in East London, South Africa. HIV viral load (VL) testing was performed using the Abbott RealTime m2000sp-rt platform (limit of detection of 640 copies/mL). The CDC-validated multi-assay algorithm that is currently used by the Public Health Indicator Surveys was applied to estimate incidence. This algorithm includes a Limiting-Antigen avidity assay (<1.5 normalized optical density units) and a viral load <1000 copies/mL were applied to identify recently infected HIV positive individuals. Prevalence ratios (PR) and 95% confidence intervals (CI) of HIV prevalence and HIV viral suppression (VL<1000 copies/mL) were estimated by log-binomial regression. Results: HIV prevalence was 26.9% (565/2100) overall, 21.9% (230/1049) among males and 32.2% (321/998) among females. HIV infection significantly

933 NO CHANGE IN THE FREQUENCY OF AGE DISPARATE RELATIONSHIPS OR IMPACT ON HIV STATUS George Mwinnyaa 1 , Oliver Laeyendecker 2 , Kate Grabowski 3 , Joseph Ssekasanvu 3 , Anthony Ndyanabo 4 , Robert Ssekubugu 4 , Joseph Kagaayi 4 , Godfrey Kigozi 4 , Gertrude Nakigozi 4 , David Serwadda 4 , Ronald H. Gray 3 1 The Johns Hopkins University, Baltimore, MD, USA, 2 NIAID, Baltimore, MD, USA, 3 Johns Hopkins University, Baltimore, MD, USA, 4 Rakai Health Sciences Program, Kalisizo, Uganda Background: Age disparate relationships (≥5 years) are associated with increased HIV prevalence. We assessed age differences between never married women and their male sexual partners over time and the risk of HIV infection in Rakai, Uganda. Methods: 10,061 never married women, aged 15-49 in the Rakai Community Cohort Study (RCCS) provided information on the age of their male sexual partners from 1997 to 2013. Logistic regression was used to assess trends in age difference and associations with HIV prevalence. Results: 2 ,992 women (30%) had a male partner ≥ 5 years older with a median age difference 7 years (IQR 5, 12). Women in an age disparate relationship were older (median age 22 [IQR 15, 26]) compared to those not in such a relationship (median age 19 [IQR 15, 30]). There was no change in the frequency of age disparate relationships over time. In 1997, 29% (95% CI 25,

CROI 2018 356