CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

1 Vanderbilt Univ, Nashville, TN, USA, 2 Columbia Univ, New York, NY, USA, 3 Univ of North Carolina at Chapel Hill, Chapel Hill, NC, USA Background: HIV viral suppression (VS) is critical for reducing HIV disease progression and transmission. Age, racial, and regional disparities in VS, as well as health literacy, have been noted in the US. Studies of the influence of health literacy on VS are lacking from the Southern US, and those from other regions have yielded conflicting results. We therefore examined health literacy and demographic disparities in VS in a large southern HIV clinical cohort. Methods: Adults enrolled at Vanderbilt’s Comprehensive Care Clinic (Nashville, TN) from 1998-2012 contributed person-time until final visit, death, or end of study. Health literacy was assessed by Brief Health Literacy Screen (BHLS), which is scored 3-15. VS was defined as a final viral load (VL) of <200 copies/mL, among those with ≥1 VL, in each calendar year. Modified Poisson regression with BHLS score, age, sex, race, years of education, HIV acquisition risk, insurance, year of enrollment, baseline CD4+, and baseline log10 VL, was used to estimate adjusted relative risks (RR) and 95% confidence intervals (CI) for not having VS. Generalized estimating equations accounted for multiple individual outcomes; restricted cubic splines with 3-4 knots captured the association between BHLS score, age, education, year of enrollment and VS. Results: Among 568 individuals with BHLS score and ≥1 VL, median BHLS score was 13.5 (IQR: 11-15), median age was 40 (IQR: 33-46 years), 152 (27%) were female, 246 (43%) were black, and 65 (11%) had a history of injection drug use (IDU). Of 3,618 person-years contributed, 56%met criteria for VS. Older (mean difference=-1.74; CI=-2.89,-0.59 for 60 vs. 40-year-olds), black (mean difference=-0.59; CI=-1.09,-0.09 vs. white), and IDU risk (mean difference=-1.24; CI=-1.80,-0.69 vs. MSM) individuals had lower average BHLS scores. In the adjusted model, lower BHLS score (RR=1.16; CI=1.01,1.32 for 7 vs. 13.5), black race (RR=1.28; CI=1.14,1.44 vs. white), and younger age (RR=1.35; CI=1.09,1.66 for 20 vs. 40-year-olds) were associated with higher risk of not having VS (Table). Conclusion: In this Southern US HIV clinical population, lower health literacy significantly increased the likelihood of not achieving VS; further, age and racial disparities in VS persisted even after accounting for health literacy differences. Improving health literacy, though a worthy goal, may not be sufficient to narrow demographic disparities in VS outcomes in this population. 913 FOOD INSECURITY AND HIV TREATMENT OUTCOMES IN THE WOMEN’S INTERAGENCY HIV STUDY Sheri Weiser 1 , Lila A. Sheira 1 , Kartika Palar 1 , Tracey Wilson 2 , Lisa Metsch 3 , Adaora Adimora 4 , Igho Ofotokun 5 , Eryka Wentz 6 , Phyllis Tien 1 , Edward A. Frongillo 7 1 Univ of California San Francisco, San Francisco, CA, USA, 2 SUNY Downstate Med Cntr, Brooklyn, NY, USA, 3 Columbia Univ, New York, NY, USA, 4 Univ of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 5 Emory Univ, Atlanta, GA, USA, 6 The Johns Hopkins Univ, Baltimore, MD, USA, 7 Univ of South Carolina, Columbia, SC, USA Background: Food insecurity (the limited or uncertain availability of nutritionally adequate, safe foods), affects 18 million households in the U.S and is more prevalent among women. While food insecurity is a known contributor to increased HIV-related morbidity and mortality, there is little longitudinal research on this issue, particularly among women. This study assessed whether changes in food insecurity were associated with changes in HIV health outcomes among women in a nationally representative sample of women with or at risk for HIV. Methods: We analyzed longitudinal data from the Women’s Interagency HIV Study (WIHS), a multi-site prospective cohort study of women with or at risk for HIV. Data on 6778 observations from 2027 women with HIV were collected from spring 2013 to fall 2015 at 6-month intervals. Food insecurity was measured with the US Household Food Security Survey Module. Outcomes included 1) logarithm of viral load (copies/ml) and detectable versus undetectable viral load, 2) CD4 count (cells/mm3), and 3) physical health status (SF-36 scale). We used longitudinal multiple tobit, logistic, and linear regression models with random effects for examining associations between food insecurity and these outcomes, adjusting for age, race/ethnicity, income, education, child dependents, time on ART, CD4 nadir, current smoking, and hazardous drinking (>7 drinks/week). Results: Over one-third of the women (37%) were food-insecure. One-third (34%) had detectable viral loads; 8% had CD4<200 cells/mm3. In adjusted analyses, having very low food security was associated with 2.4 times higher viral loads (95% CI=1.4, 4.0, p<0.0001) and 1.6 times higher odds of having a detectable viral load (95% CI=1.1, 2.1; p<0.01), compared to those with high food security (See Table 1). Those with very low food security had 24.0 cells/mm3 lower CD4 cells (p<0.01), and 2.4 points lower physical health composite scores (p<0.001), compared to those with high food security. Conclusion: Food insecurity was associated with worse HIV treatment outcomes among HIV-infected women in the US. These results provide strong evidence that food insecurity is associated with poor health among HIV-infected women, and should be addressed as part of comprehensive HIV care delivery. Studies are needed to determine the best intervention strategies to improve food security in different contexts and sub-populations in order to improve control of the US HIV epidemic. 914 HIV STATUS AWARENESS AND ART COVERAGE AMONG FEMALE SEX WORKERS IN JUBA, SOUTH SUDAN Avi Hakim 3 , Victoria Achut 1 , Alfred Okiria 2 , Alex Bolo 3 , Jennifer Wesson 4 , Joel Katoro 3 , Golda Caesar 1 , Acaga Taban 2 , Lee Hundley 5 1 South Sudan Ministry of Hlth, Juba, South Sudan, 2 IntraHlth Intl, Juba, South Sudan, 3 CDC–South Sudan, Juba, South Sudan, 4 IntraHlth Intl, Chapel Hill, NC, USA, 5 Emory Univ, Atlanta, USA Background: South Sudan’s HIV prevalence is estimated at 2.7% in the general population, but no biobehavioral data exist on female sex workers (FSW). We conducted a survey of FSW in Juba, South Sudan, to estimate HIV prevalence, awareness of HIV status, and treatment coverage. Methods: We used respondent-driven sampling to recruit 846 FSW in Juba from November 2015 to March 2016. Eligibility criteria included age ≥15 years; received money, goods, or services for sex in the last 6 months; lived or worked in Juba for at least 1 month; and able to speak English, Juba Arabic, or Swahili. Participants completed a face-to-face interview and were tested for HIV. Bivariate and multivariate analyses were conducted in RDS-A and SAS to assess factors associated with being unaware of one’s HIV infection. All results are weighted. Results: One-third (34.0%) of FSW in Juba were South Sudanese and 78.8% had ever tested for HIV. HIV prevalence was 37.9% (95% CI: 33.6-42.2). Among HIV-positive FSW, 37.1% (95% CI: 29.7-43.8) were unaware of their infection, 64.8% had a CD4< 500 and 46.4%were on treatment. Among FSWwho had accessed care, 54.2% had their last clinic visit in Juba. In bivariate analysis, the odds of being unaware were higher among FSWwho were South Sudanese (OR: 10.2, 95% CI: 4.3-24.3) or Congolese (OR: 4.1, 95% CI: 2.1-7.7), had not received condom information in the last 12 months (5.1, 95% CI: 3.0-8.9), and did not have a condom break in the last 6 months (2.8, 95% CI: 1.4-5.4). In multivariate analysis, a greater odds of being unaware of her HIV infection was observed among women <15 years old at first sex versus those who waited longer and women who had never spoken with a peer educator or outreach worker about HIV versus those who had (OR: 4.7, 95% CI: 1.4-16.3; 3.1, 95% CI: 1.3-7.5 respectively). Conclusion: HIV prevalence among FSW in Juba is very high compared to the national HIV prevalence of 2.7% and treatment coverage is far below the UNAIDS goal of 90%. More than one-third of FSW are unaware they have HIV and thus are not on treatment. Outreach activities should be strengthened to increase access to HIV testing and improve linkages to treatment for FSW. With the majority of FSW eligible for treatment with a CD4< 500, FSW should be prioritized for test and start to increase treatment coverage in Juba. Retention efforts should focus on FSWwho obtain care in other countries to mitigate possible treatment interruptions. 915 “TEST AND TREAT” ANTIRETROVIRAL THERAPY AMONG FEMALE SEX WORKERS IN KAMPALA, UGANDA Yunia Mayanja , Onesmus Kamacooko, Daniel Bagiire, Gertrude Namale, Janet Seeley MRC/UVRI Uganda Rsr Unit on AIDS, Entebbe, Uganda Background: Current guidelines on use of antiretroviral treatment (ART) recommend immediate treatment upon HIV positive diagnosis, regardless of CD4 count or WHO clinical stage (Test and Treat). Data on implementation of the guidelines among female sex workers (FSWs) in sub Saharan Africa are limited. We describe uptake of test and treat, and associated factors among FSWs attending a research clinic in Kampala, Uganda. Methods: A cohort of FSWs was established in 2008 and followed up at a research clinic (Good Health for Women Project) in Kampala. At quarterly visits, data are collected on HIV and other sexually transmitted infections, socio-demographics, reproductive health, substance use, and partner violence; CD4 tests are done 6 monthly. HIV-positive women were initiated on ART using CD4 and WHO clinical stage criteria until August 2014, when test and treat was implemented. Previously enrolled HIV-positive women who had not been

Poster and Themed Discussion Abstracts

CROI 2017 396