Research linking substance use with HIV risk behaviors and antiretroviral therapy adherence is typically based on self-reported data. However, studies in high-income countries show that men who have sex with men (MSM) often underreport substance use when compared to objective biological testing. Such comparisons have not been conducted among MSM in sub-Saharan Africa. We compared self-reported and objectively measured substance use among MSM participating in HPTN 075, a multi-site observational cohort study conducted in Kenya, Malawi, and South Africa. Urine samples (n = 734) from 382 participants were tested for the alcohol metabolite ethyl glucuronide and 43 other substances. These tests detect alcohol use in the prior 48-72 h and other substances up to 7 days earlier, depending on the drug. Of the 734 samples, 159 (21.7%) tested positive for ethyl glucuronide. Self-reported alcohol use was available for 97.5% of these cases and confirmed in 141 (91.0%) of them. Sixty samples (8.2%) tested positive for at least one of the 43 other substances. Self-report data were available for 95.0% of those, but use was acknowledged in only 19 (33.3%) cases. These findings suggest that alcohol use is generally reported accurately, while drug use is substantially underreported-likely due to legal prohibitions and social stigma. Incorporating objective substance testing alongside self-reports is recommended to improve the accuracy of substance use measurement in behavioral and clinical HIV research, especially in contexts where stigma or criminalization may inhibit disclosure.
Sandfort et al. (Sat,) studied this question.