Gender and Racial/Ethnic Disparities in HIV Care and Viral Suppression Before Incarceration

Background: Studies examining the independent influence of gender and race/ethnicity on HIV care engagement among people living with HIV (PLWH) in correctional facilities have found that women and racial/ethnic minorities have a disproportionately low likelihood of accessing HIV treatment and achieving desirable viral load outcomes before incarceration. Few studies have examined the intersection of gender and race/ethnicity in this context, which may mask disparities among some population subsets. Objectives: 1) Jointly assess gender and racial/ethnic disparities in pre-incarceration HIV treatment utilization and viral suppression among jailed PLWH, and 2) evaluate modifiable characteristics accounting for disparities. Methods: The primary exposure was gender by race/ethnicity. Outcomes of interest included linkage to an HIV care provider, ART use, optimal ART adherence, and viral suppression. Potential mediators were classified into conceptual domains according to Andersen’s Behavioral Model of Health Care Utilization. We performed nested multivariable logistic regression analyses to assess differences in the outcomes between each gender-stratified racial/ethnic group and non-Hispanic White men, accounting for potential confounders and mediators. Model 1 accounted for potential confounders, Model 2 accounted for predisposing factors (e.g. homelessness), Model 3 further accounted for enabling resources (e.g. health insurance), and Model 4 further accounted for need factors (e.g. drug abuse severity). We determined whether disparities in the outcome existed based on the results of Model 1. We then assessed the proportion of the disparity explained by potential mediators in Models 2-4. Results: Overall, racial/ethnic minority women had the lowest odds of engaging in care relative to non-Hispanic White men. Non-Hispanic Black women had the greatest odds of lacking an HIV care provider, while Hispanic women had the highest odds of no ART use and the lowest odds of optimal ART adherence. NH Blacks and Hispanic men had the lowest odds of viral suppression. Factors accounting for disparities varied by gender-stratified racial/ethnic group and across outcomes. Conclusions: Study findings offer useful information to plan for the provision of community- and jail-based services for PLWH and the development of more effective, customized HIV interventions. Future research should aim to further clarify modifiable factors accounting for disparities in care utilization and viral suppression.