CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

initiation to be in the “control” group and all MSMwhose first clinic visit occurred after intervention initiation to be in the “treatment” group. We estimated a difference-in- differences linear regression model, comparing PrEP awareness and use at second visit among MSM in the treatment group relative to the control group. Results: Of 967 HIV-negative MSM receiving care at the STD clinic, 316 (33%) presented two or more times. Non-Hispanic Black MSM had statistically significantly lower PrEP awareness relative to non-Hispanic white MSM at their first visit (adjusted odds ratio: 0.36, p=0.03). Time trends in PrEP awareness (p=0.53, Figure 1) and use (p= 0.63) were equivalent between the treatment and control groups at the first clinic visit. At the second clinic visit, MSMwho had received the PrEP intervention during their first clinic visit were 19 percentage points (pp; p<0.01) more likely to be aware of PrEP (Figure 1) and 3.6 pp (p=0.04) more likely to use PrEP, relative increases of 47% for PrEP awareness and 133% for PrEP use relative to the period prior to the intervention. Conclusion: A brief, scalable PrEP education intervention at an STD clinic led to significantly increased PrEP awareness and use among MSM. Healthcare providers should consider implementing brief PrEP education interventions in sexual healthcare settings. It is particularly important for such interventions to reach non-Hispanic Black MSM due to lower levels of PrEP awareness in this population.

969 PREP GUIDELINES HAVE LIMITED ACCURACY IDENTIFYING YOUNG MSM PRIOR TO SEROCONVERSION Nicola Lancki, Ellen Almirol, Leigh Alon, Moira McNulty, John A. Schneider Univ of Chicago, Chicago, IL, USA

Background: Identification of clients at highest risk of acquiring human immunodeficiency virus (HIV) is a critical component to PrEP implementation. CDC published clinical practice guidelines for identifying individuals as PrEP candidates in 2014 and developed a risk-screening tool: HIV Incidence Risk Index for MSM (HIRI-MSM). Gilead also listed factors to identify individuals at high risk in the package insert. We examined the performance of CDC guidelines, HIRI-MSM and Gilead recommendations in identifying eligible PrEP candidates, including seroconverters, in a population-based sample of young Black men who have sex with men (YBMSM). Methods: We followed a population-based cohort of YBMSM aged 16-29 years during PrEP roll-out in Chicago from 2013-2016 (n=618). We computed the proportion of YBMSM with indications for PrEP using CDC guidelines, HIRI-MSM, and Gilead recommendations. We also calculated the sensitivity and specificity of guidelines in predicting HIV seroconversion. HIV seroconversion was measured using 4th generation and NAAT testing at three time points. Incidence Rate Ratios using Poisson regression were computed to compare sociodemographic and network factors associated with HIV incidence. Results: In the study cohort, 300 HIV uninfected YBMSM contributed 390.4 person-years (PY) of follow-up. The mean age at baseline was 22.3 years (SD=3.07), HIV incidence was 8.5 cases per 100 PY (95% CI, 6.0-11.9). Overall, 49% had an indication for PrEP using CDC guidelines; 72% using HIRI-MSM, and 86% using Gilead recommendations. HIV seroconverters (n=33) were identified as PrEP eligible prior to seroconversion with sensitivity/specificity of CDC, HIRI-MSM and Gilead guidelines of: 52%/52%; 85%/30%; and 94%/15%. HIV incidence did not differ significantly by individual risk behaviors that comprise indications for PrEP: condomless anal sex (IRR 1.2, 95% CI: 0.6-2.3); drug use (IRR 1.2, 0.6, 2.3); serodiscordant partnership (IRR 0.3, 95% CI 0.04, 1.9). Having a partner ≥ 10 years older was predictive of HIV incidence (IRR 2.1, 95% CI 1.0-4.5). Conclusion: Low sensitivity of CDC guidelines and limited specificity of HIRI-MSM and Gilead screening tools is of concern for PrEP implementation in most at risk populations. Consideration of local epidemiology and network factors may better guide identification of clients who could benefit the most from PrEP. 970 IDENTIFYING PREGNANT WOMEN FOR PREP USING ROUTINE ANTENATAL CARE INDICATORS IN KENYA Jillian Pintye 1 , Benson Singa 2 , Kennedy Wanyonyi 2 , Janet Itindi 2 , John Kinuthia 3 , Lucy Ng’ang’a 4 , Agnes Langat 4 , Abraham Katana 4 , Christine J. McGrath 5 , Grace John-Stewart 1 1 Univ of Washington, Seattle, WA, USA, 2 Kenya Med Rsr Inst, Nairobi, Kenya, 3 Kenyatta Natl Hosp, Nairobi, Kenya, 4 US CDC, Nairobi, Kenya, 5 Univ of Texas Med Branch, Galveston, TX, USA Background: PrEP could prevent HIV acquisition in pregnancy, however, implementation strategies have not been established for pregnant women in high HIV prevalence settings. Regional HIV prevalence and indicators assessed in antenatal care (ANC) could be used to prioritize PrEP. Data from a Kenyan national survey of 62 ANC facilities were used to estimate the proportion of women at potential risk for HIV who could be prioritized for PrEP. Methods: Facilities were selected using stratified random sampling by ANC volume. Data were abstracted from the first 10% of initial ANC visits per facility per year from 2011-2013. High HIV risk was defined as having syphilis and/or a male partner of unknown or positive HIV status. Survey weights and clinic-level clustering were applied. Kenya Demographic and Health Survey 2014 data were used for projected estimates. Results: Overall, 9250 records from first ANC visits of HIV-uninfected women were abstracted, of which 8634 (93%) met inclusion criteria (had syphilis or partner HIV status data); partner HIV status and syphilis data were available for 85% and 69% of records, respectively. Median age was 24 years and 18% of women were <20 years; 86%were married and 37%were primagravidas. Having a male partner of unknown HIV status was common (46%) and higher in Nyanza, a high HIV prevalence region, than in other regions (50% vs. 32%, p=0.04). Couples HIV counseling and testing was low (3%), without regional differences. Few women reported HIV-infected partners (1%) and 1% had syphilis infection. Overall, 39% of women had potential high HIV risk (as defined by syphilis and/or partner HIV status among women with data on either variable) with similar rates between 2011 and 2013; prevalence of high HIV risk was highest in Nyanza (51%) than other regions (Prevalence Ratio 1.5, 95% CI 1.1-2.2). In all regions combined, prioritizing PrEP to pregnant women with these ANC indicators would decrease the number of women offered PrEP by 61%while providing PrEP to all women at potential high HIV risk (Figure 1). An HIV prevalence-guided approach with PrEP provision only to all women in the highest prevalence region (Nyanza) would reduce the number of women exposed to PrEP by 74%, but exclude 63% of high-risk women nationally. Conclusion: A combination of prevalence and risk assessment strategies may be useful to strategically deliver PrEP in pregnancy. Many pregnant women remain unaware of partner HIV status; enhancing partner HIV testing could improve PrEP provision.

Poster and Themed Discussion Abstracts

CROI 2017 420