2.5. Statistical Analysis

Data were entered into EpiData version 3.1 (http://www.epidata.dk/) via double entry. After the data were cleaned and verified, statistical analysis was performed using SPSS version 23.0. For descriptive analyses, categorical variables are presented as frequencies and proportions, and continuous variables are presented as medians and interquartile ranges (IQRs). Differences in the general demographic characteristics were calculated using Student’s t-tests, chi-squared (χ²) tests, Fisher’s exact test, and Kruskal–Wallis tests. We determined the mean (standard deviation [SD]) and median (IQR) differences. Factors from univariate analysis with P-values < 0.10 and/or those previously shown to be associated with differences in socio-demographic characteristics were included in the multivariate regression models, which were used to calculate adjusted odds ratios (AORs) and their 95% confidence intervals (CIs). The primary outcomes of interest were HIV testing in the past 12 months, bisexual orientation, substance use to adjust psychiatric disorders, and having undergone AIDS interventions in the past 3 months. P-values < 0.05 and β = 0.1 were considered statistically significant.