Table 4.
Associations between hypothesized predictors and access to HIV care and treatmenta.
| Predictors | HIV provider access | Access to ARTb refills | ||||
|
|
Coefc | 95% CI | P value | Coef | 95% CI | P value |
| Community engagement | 0.81 | 0.75 to 0.86 | <.001 | 1.23 | 1.19 to 1.27 | <.001 |
| Comfort with provider | 0.54 | 0.53 to 0.55 | <.001 | 0.79 | 0.78 to 0.80 | <.001 |
| Sexual stigma | –0.87 | –0.95 to –0.78 | <.001 | –1.08 | –1.11 to –1.05 | <.001 |
| Provider discrimination | –0.77 | –0.81 to –0.73 | <.001 | –1.13 | –1.17 to –1.09 | <.001 |
aRegression models also adjusted for income, education, relationship status, urbanicity (urban vs rural), racial/ethnic minority status, health insurance, and region (Global North vs Global South) as covariates. In sensitivity analyses omitting urbanicity and region, results were similar with respect to magnitude and level of significance of estimates.
bART: antiretroviral therapy.
cCoef: regression coefficient.