S2 Data -

Objectives
To assess ecological, structural, community and individual level correlates of health services utilization along a continuum of HIV care, and sexual health and support services among gay and bisexual men worldwide.
Methods
Using a nonprobability internet sample of 6,135 gay and bisexual men, we assessed correlates of utilization of health services. Chi-Square Tests of Independence were performed to assess drop off along a continuum of HIV care. Multivariable logistic regression analyses using generalized estimating equation models were conducted adjusting for geographic region and clustering by country. In multivariable analyses, we determined the association between utilization outcomes, and ecologic, structural, community and individual correlates by fitting separate generalized estimating equation (GEE) logistic regression models for each of the outcomes, fitted with robust SEs, and accounting for clustering by country. Stratified by sexual identity, analyses adjusted for variables that could influence HIV-related health outcomes including racial/ethnic minority status, participant age, insurance, ability to make ends meet, as well as country-level income (income of country of residence defined by the World Bank).
Results
Among men living with HIV (n = 1001), being in HIV care (n = 867) was associated with being on ART (X2 = 191.17, p < .001), and viral load suppression (X2 = 14.03, p < .001); and using ART (n = 840) was associated viral load suppression (X2 = 21.66, p < .001). Overall, the pattern of utilization outcomes were similar for both gay and bisexual men. For example, utilization of PrEP, being in HIV care and utilization of most of the sexual health and support services were negatively associated with sexual stigma. Whereas, utilization of most HIV prevention, and sexual health and support services were positively associated with provider discrimination. Utilization of all HIV prevention and all sexual health services were positively associated with greater community engagement, and receiving services from LGBT-led organizations. Bisexual men had higher odds of reporting provider discrimination when utilizing condom services (gay: AOR = 1.14, [0.95–1.36]; bisexual: 1.58, [1.10–2.28]), PrEP (gay: AOR = 1.06, [0.77–1.45]; bisexual: AOR = 2.14, [1.18–3.89], mental health services (gay: AOR = 1.03, [0.86–1.23]; bisexual: AOR = 1.32, [1.07–1.64]), and community-based support (gay: AOR = 1.23, [1.05–1.45]; bisexual: AOR = 1.49, [1.14–1.93]) than gay men. Bisexual men also reported higher odds of accessing services from LGBT-led organizations when utilizing PrEP (gay: AOR = 5.26, [2.50–11.05]; bisexual: AOR = 7.12, [3.16–16.04]), and community-based support/self-help groups/individual counseling (gay: AOR = 2.63, [1.72–4.01]; bisexual: AOR = 3.35, [2.30–4.88].
Conclusions
It is essential that barriers to health services utilization be addressed at structural and community levels. Structural interventions should be designed to reduce sexual stigma, as well as train and sensitize healthcare providers; and strengthen community level interventions that bring gay and bisexual men together to lead comprehensive health services.