CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Methods: Medicaid claims do not specify gender identity inclusive of transgender status. In consultation with clinical experts on HIV and gender- affirming care, we developed an algorithm to identify transgender enrollees using diagnoses, prescriptions and sex at birth from claims records in 2013-2017. In order to identify those living with HIV, we matched Medicaid enrollees to individuals diagnosed with HIV before 2018 in the registry. Results: Our algorithm identified 6,043 unique transgender persons who accessed Medicaid in 2013-2017, with 1,472 (24%) reported to the HIV registry, 1,168 (79%) of whomwere identified as transgender in the registry. We found an additional 292 transgender individuals in the registry that had accessed Medicaid during this period but were not identified by our algorithm, for a total of 6,335 transgender individuals accessing Medicaid during this period (0.1% of the NYC Medicaid population) and 1,764 transgender PLWH (28% of transgender individuals accessing Medicaid). From 2013 to 2017, there was a 35% increase in transgender individuals accessing Medicaid. Conclusion: Using a novel method we identified a large sample of transgender individuals in Medicaid, many of whomwere PLWH. We were able to calculate the prevalence of HIV among transgender Medicaid beneficiaries and to improve ascertainment of transgender persons in the HIV registry. We also saw a sizeable increase in transgender individuals accessing Medicaid over the five-year period, likely due in part to expansion of Medicaid policy to cover transgender-related healthcare. Given the high coverage among transgender PLWH, Medicaid is a valuable source of health information for the transgender population, a group that is often difficult to identify due to issues of stigma, reduced access to appropriate care, and misgendering by healthcare personnel.

which we considered time-varying treatment status, the estimated difference in the four quantiles comparing before and after treatment were 25th: 0 [95% CI: -4.675, 3.889], 50th: -43.94 [95% CI: -156.7, 1.208], and 75th: -4360 [95% CI: -10930, -195.8]. Approximately 74% of viral loads were below 1500 copies/mL (a meaningful cutoff for the risk of HIV transmissibility) after the BUP treatment compared to 69% before treatment. Restriction to individuals who started BUP treatment after 2011 similarly suggested difference of viral loads in higher quantiles, but was limited in sample size (61). Conclusion: These data suggest that BUP treatment for OUD among PWH is likely to have beneficial effects on HIV RNA. By increasing the proportion of PWH below 1500 copies/mL, it would lower the overall risk for HIV transmission.

Poster Abstracts

886 AFFORDABLE CARE ACT’S IMPACT ON SUBSTANCE-USE TREATMENT IN PEOPLE WHO INJECT DRUGS Erik M. Hendrickson 1 , Tarik Benmarhnia 1 , Steffanie A.Strathdee 1 , Jazmine Cuevas-Mota 1 , Richard S. Garfein 1 1 University of California San Diego, La Jolla, CA, USA Background: Substance use treatment (SUT) for Persons Who Inject Drugs (PWID) can reduce the risk of HIV and HCV transmission, yet the lack of health insurance or insurance plans that cover these services is a major barrier to PWID entering SUT. Provisions in the U.S. Patient Protection and Affordable Care Act (ACA) were expected to increase the use of SUT in PWID by increasing access to health insurance and including these services as an essential health benefit. Methods: We analyzed SUT use before and after the implementation of the ACA in California on January 1, 2014 among participants enrolled in STAHR-II (2012-2016)—a longitudinal cohort study of PWID in San Diego, California that included a baseline and up to 4 semi-annual follow-up interviews. We examined changes in self-reported SUT within participants pre- and post-ACA implementation. We included participants who had both a baseline visit and a follow-up visit before and after the implementation of the ACA in California. We excluded visits with referent time periods that overlapped with the ACA implementation date. In bivariate analysis, we used McNemar’s test for paired comparisons to determine the association between the ACA and SUT, as well as potential confounders. We used multivariable logistic regression analysis with Generalized Estimation Equations (GEE) for repeated measures to assess the association between the ACA and SUT, adjusting for baseline covariates: age, sex, race, education, HIV, HCV, chronic disease, prior SUT use, past 6-month daily injection, past 6-month homelessness, perceived need for SUT. Insurance status was a time-updated covariate. Results: Of 170 participants who had both baseline visit and a follow-up visits before and after the implementation of the ACA in California, 71%were male, 50%were White and mean age was 45 years. There was an 11.8% increase in SUT use after the ACA, compared to before (52.4% vs. 40.6%, p=0.01) and a 10.6% increase in the proportion who had insurance after the ACA compared to before (81.2% vs. 70.6%, p<0.01). The positive impact of the ACA on SUT remained after adjusting for age, race, ever using SUT, perceived need for SUT, and insurance status (AOR: 1.85, 95%CI: 1.25-2.76). Conclusion: Implementation of the ACA in California was associated with an increase in SUT use among PWID in San Diego, suggesting that the ACA successfully achieved the objective of increasing access to SUT.

885 BUPRENORPHINE TREATMENT IS RELATED TO DECREASED HIV RNA LEVELS AMONG PEOPLE WITH HIV Jongyeon Kim 1 , Geetanjali Chander 2 , Catherine R. Lesko 1 , Anthony T. Fojo 2 , Richard D. Moore 2 , Bryan Lau 1 1 Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA, 2 Johns Hopkins University School of Medicine, Baltimore, MD, USA Background: Initiation of buprenorphine (BUP) for people with HIV (PWH) and opioid use disorder (OUD) may improve HIV clinical outcomes. We examined level of HIV RNA among PWH initiating BUP in an urban HIV clinic. Methods: In the Johns Hopkins Hospital HIV Clinic Cohort (JHHCC), we identified 207 PWH who started buprenorphine treatment between 2002 and 2018. We allowed multiple number of treatment episodes (defined as continuous buprenorphine prescription with gaps less than 30 days). A quantile linear model was used to assess the relationship between buprenorphine (BUP) and viral load considering skewed distribution and the large proportion of individuals who were suppressed. We estimated quantiles with cluster bootstraps to account for repeated observations within participants. We included CD4 counts, sex, race, age, injection drug use, and men who have sex with men as covariates. Each individual contributed viral loads one year before and one year after their BUP initiation. We present difference in the 25th, 50th, and 75th percentiles comparing prior and subsequent to any episode of BUP treatment. Results: The 207 PWH were primarily male (69%), black (88%), with median age of 49 (IQR: 44-53) at their initial BUP treatment. Individuals contributed a median of 1 (IQR:1-2) treatment episodes. HIV viral loads before and after initial treatment were a median of 80 (IQR: 50-6690) and 50 (IQR: 50-1721) copies/mL respectively. The figure shows a scatterplot and the unadjusted quantiles of HIV RNA as time prior to and subsequent to initial BUP treatment. In the model in

CROI 2020 331