Estimating the Prevalence of HIV and Sexual Behaviors Among the US Transgender Population: A Systematic Review and Meta-Analysis, 2006-2017.

Background. Transgender women (transwomen) in the United States have been shown to have high HIV risk with Black and Hispanic transwomen being particularly vulnerable. Growing research on transgender men (transmen) also shows increased HIV risk and burden, although not as much is known for this transgender population. Objectives. This systematic review estimates the prevalence of self-reported and laboratory-confirmed HIV infection, reported sexual and injection behaviors, and contextual factors associated with HIV risk of transgender persons living in the United States. Search Methods. We searched the HIV Prevention Research Synthesis database and MEDLINE, EMBASE, PsycINFO, CINAHL, and Sociological Abstracts databases from January 2006 to March 2017 and January 2006 to May 2017, respectively. Additional hand searches were conducted in December 2017 to obtain studies not found in the literature searches. Selection Criteria. Eligible reports were published US-based studies that included transgender persons and reported HIV status. Data collection and analysis. Data were double-coded and quality assessed. We used random-effects models employing the Der Simonian-Laird method to calculate overall prevalence of HIV infection, risk behaviors, and contextual factors for transwomen, transmen, and race/ethnicity subgroups. Main Results. We reviewed 88 studies, the majority of which were cross-sectional surveys. Overall laboratory-confirmed estimated prevalence of HIV infection was 9.2% (95% confidence interval [CI] = 6.0%, 13.7%; k = 24). Among transwomen and transmen, HIV infection prevalence estimates were 14.1% (95% CI = 8.7%, 22.2%; k = 13) and 3.2% (95% CI= 1.4%, 7.1%; k = 8), respectively. Self-reported HIV infection was 16.1% (95% CI = 12.0%, 21.2%; k = 44), 21.0% (95% CI = 15.9%, 27.2%; k = 30), and 1.2% (95% CI = 0.4%, 3.1%; k = 7) for overall, transwomen, and transmen, respectively. HIV infection estimates were highest among Blacks (44.2%; 95% CI = 23.2%, 67.5%; k = 4). Overall, participation in sex work was 31.0% (95% CI = 23.9%, 39.0%; k = 39). Transwomen (37.9%; 95% CI = 29.0%, 47.7%; k = 29) reported higher participation in sex work than transmen (13.1%; 95% CI= 6.6%, 24.3%; k = 10; P = .001). Most outcomes indicated high heterogeneity in the overall and subgroup analyses. Conclusions. The availability of more data allowed us to calculate estimates separately for transwomen and transmen. HIV prevalence estimates for US transwomen were lower than previous estimates, but estimates for HIV prevalence and participation in sex work were higher when compared with transmen. Evidence gaps remain for transmen and the syndemic relationship of HIV, risky behaviors, and contextual factors specific to the transgender experience. Public Health Implications. This study highlights gender disparities for HIV and risky sexual behavior, as well as evidence gaps that exist for transmen. Tailored programs and services for the transgender population need to be developed to encourage use of and access to HIV prevention services.