CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

1101LBIMPACT OF SERVICE INTEGRATION ON HIV TESTING UPTAKE AMONG KEY POPULATIONS IN INDIA Sunil S. Solomon 1 , Suniti Solomon 2 , Allison M. McFall 1 , Aylur K. Srikrishnan 2 , Santhanam Anand 2 , Pachamuthu Balakrishnan 2 , Elizabeth Ogburn 1 , Lawrence Moulton 1 , Muniratnam S. Kumar 2 , David D. Celentano 1 , Gregory M. Lucas 1 , Shruti H. Mehta 1 1 Johns Hopkins University, Baltimore, MD, USA, 2 YR Gaitonde Center for AIDS Research and Education, Chennai, India Background: Stigma and fragmented services impede HIV care access among people who inject drugs (PWID) and men who have sex with men (MSM). We evaluated the impact of integrated care centers (ICCs) in a cluster randomized trial in India. ICCs provided rapid HIV testing, linkage to care and treatment Methods: We randomized 22 cities (12 PWID, 10 MSM) to ICC or usual care (UC) at a 1:1 ratio. ICC services are supported by the Indian AIDS program and were available in UC sites, but not in integrated venues. We assessed outcomes with cross-sectional, respondent-driven sampling (RDS) surveys in each city, before (2012-13) and 2 years after ICC implementation (2016-17). Our primary outcome was self-reported recent (prior 12 months) HIV testing in ICC vs. UC cities at evaluation, adjusted for baseline testing and stratum. Because we used RDS, our trial evaluated the ability of ICCs to affect outcomes at the community level, irrespective of participants’ actual ICC exposure. We also collected biometric data to compare outcomes in RDS participants at ICC sites who did and did not visit the ICC. Population ICC exposure was calculated as the proportion in the evaluation RDS who visited the ICC based on biometric data. Results: During the intervention phase the 11 ICCs tested a median (range) of 1309 (829 - 2191) clients for HIV. 21,714 participated in the evaluation RDS. Compared with UC cities, ICC cities had 31% higher prevalence of recent HIV testing at evaluation (prevalence ratio [PR]: 1.31; 95% CI: 0.95, 1.81; Figure, Panel A). Moreover, those in ICC cities who had visited the ICC were significantly more likely to report recent HIV testing than those in UC cities (PR: 2.66; 95% CI: 2.19, 3.24). Among the ICC cities, ICC population exposure ranged from 7 to 55% and higher exposure was significantly associated with greater increase (from baseline to evaluation) in recent HIV testing (Figure, Panel B; p=0.002). Conclusion: While ICCs provided HIV testing to large numbers of PWID and MSM, they were not associated with a statistically significant increase in recent testing at the community level. Pre-specified analyses showed that exposure to ICCs within the target groups was limited within these populous cities and that the degree of population exposure was strongly correlated with testing rates, suggesting that increased ICC coverage could yield community-level impact. support for HIV and risk reduction services (e.g., opioid therapy, syringe exchange, condoms and sexually transmitted infection treatment).

1100 IMPROVEMENT IN HIV CARE INDICATORS AMONG THE HOMELESS IN SAN FRANCISCO Susan Scheer , Ling Hsu, Darpun Sachdev, Oliver Bacon, Stephanie E. Cohen, Albert Y. Liu, Jennie C. Chin, Susan P. Buchbinder San Francisco Department of Public Health, San Francisco, CA, USA Background: Despite significant declines in new HIV diagnoses and improvement in HIV-related care indicators in San Francisco (SF), health disparities persist particularly among homeless persons living with HIV (PLWH). City-wide linkage and case management programs seek to improve outcomes across the continuum of HIV care. We measured HIV care indicators comparing those who were homeless to those housed in SF. Methods: The SF HIV surveillance registry was used to determine linkage to care and viral suppression among PLWH by housing status. Homelessness was defined as a medical chart notation of homeless or living in a homeless shelter. Temporal trends from diagnosis to viral suppression by housing status for cases diagnosed in 2013-2016 were measured. Care was defined as having a CD4+ cell count or HIV viral load test. Viral suppression (VS) was defined as HIV RNA<200 c/mL. Results: In 2015, 29 (10%) of 296 newly HIV diagnosed cases were homeless. Linkage to care within one month and VS within 12 months of diagnosis were lower among homeless compared to housed persons; 66% vs 79% (p=.10) and 59% vs 79% (p=.02), respectively. Among 12,769 PLWH in 2015 with last known residence in SF, 301 (2%) were known to be homeless. Compared to all persons diagnosed in 2006-2016, homeless persons were more likely to be cis women (14% vs 7%, p<.0001), trans women (10% vs 3%, p<.0001), African American (27% vs 14%, p<.0001) or a person who injects drugs (58% vs 20%, p<.0001); 35%were <30 years old. In 2015, homeless PLWH were less likely than the housed to have had ≥1 care visit (52% vs 81%, p<.0001) or to have VS (31% vs 74%, p<.0001). In 2015, 30 homeless PLWH who were not-in-care enrolled in LINCS (a short-term intensive case management program); 27 (90%) were re-linked to care within 3 months and 77%were virally suppressed within 12 months. In 2013-2016, city-wide rapid linkage to care was scaled up; median days from diagnosis to viral suppression were greater for homeless than housed cases each year (Figure) but decreased significantly over time (p=.04). Conclusion: Although a small proportion of all SF PLWH, homeless persons had poorest linkage to care and VS. Time from diagnosis to VS has significantly improved over time for the homeless. Scale–up of city-wide rapid linkage to care and intensive case management programs are beginning to show progress in decreasing disparities among homeless PLWH, our most vulnerable population.

Poster Abstracts

CROI 2018 423