(234) Antiretroviral Initiation and Adherence between Transgender and Cisgender Medicaid Beneficiaries with HIV

Background: The prevalence of HIV is 9.2% for transgender and gender diverse (TGD) people compared to 0.5% in the US overall. Studies exploring TGD people’s engagement with the care continuum have been limited by cross-sectional study design, single center studies, and self-reported outcomes.

Objectives: Using innovative data sources, we compare antiretroviral therapy initiation and adherence in TGD and cisgender Medicaid beneficiaries. We also explore ART use by Medicaid's coverage of gender-affirming surgery (GAS).

Methods: Our study leveraged Medicaid claims data from 14 states which constitute 75% of the HIV burden in the United States. We identified TGD enrollees using a validated algorithm drawn from diagnosis, drug, and procedure codes from inpatient, outpatient, and prescription claim data. We limited our study to people who were enrolled in Medicaid for at least 6 months during 2006-2015. We selected 10 nearest-neighbor cisgender matches for each TGD beneficiary based on propensity scores estimated from state, race/ethnicity, and age. We used logistic regression to estimate the incidence of ART initiation and persistence (≥ 90% days adherent to ART in an eligible month) for the TGD and cisgender cohorts controlling for age, race/ethnicity, and any outpatient utilization. We used generalized estimating equations to address repeated observations for beneficiaries, robust standard errors, and state fixed effects. We used a binary variable for any Medicaid coverage of GAS.

Results: In the 14 states, 16.6% of the TGD sample were living with HIV compared to 0.8% of the cisgender sample. TGD beneficiaries had 20.5 percentage points (pp) higher likelihood of initiating ART (76.4%, 95% CI: 74.7-78.1%) than cisgender beneficiaries (55.9%; 95% CI: 53.8-58.0%). Among beneficiaries who initiated ART, TGD beneficiaries were 9.1 pp more likely to adhere to ART (49.4%, 95% CI: 48.2-50.6%) than cisgender beneficiaries (40.3%, 95% CI:39.8-40.8%). In GAS coverage states, TGD people had a higher rate of ART adherence than in states without coverage (44.6% vs. 42.6%, difference 95% CI: 1.7-2.3pp). Conversely, cisgender people had a lower rate of ART adherence in GAS coverage states compared to non-coverage states (34.2% vs. 35.1%, difference 95% CI: -1.0- -0.8pp).

Conclusions: Our finding that 17% of TGD Medicaid beneficiaries are living with HIV indicates that this should be a priority Medicaid population. Despite TGD Medicaid beneficiaries having a higher prevalence of HIV than cisgender comparators, TGD beneficiaries in this study are more likely to initiate ART and adhere to ART after initiation, though overall initiation and adherence is well below clinical guidelines. For TGD beneficiaries, improving Medicaid coverage of GAS may be one lever to increase ART adherence.