Substance Use Classes Among YMSM in an HIV Digital Health Intervention Program: Implications for Acceptability, Engagement, and Health Outcomes.

Abstract

Young men who have sex with men (YMSM) have high rates of substance use, which increases their risk for HIV. Digital Health Interventions (DHI) have the potential to address HIV risk overall and reduce harms in the context of substance use. However, there is limited research on how YMSM with different substance use patterns respond to HIV DHIs and how these programs impact participant outcomes. In this study, we used latent class analysis (LCA) to identify subgroups of participants with specific substance use behaviors in the Keep It Up! (KIU! ) program, an HIV DHI with prior documented effectiveness. This study involved 2,124 participants in a Type III Hybrid trial, comparing two implementation strategies: one via 22 community-based organizations and another through direct-to-consumer recruitment. Evaluations were done at baseline and 12-week follow-up. This analysis identified four classes of substance use: Low Use (minimal alcohol risks, low illicit drug use, infrequent marijuana), Heavy Use (highest alcohol problems, frequent marijuana, elevated illicit drug use), Alcohol and Marijuana Use (high alcohol problems, frequent marijuana), and Methamphetamine and GBH Use (low alcohol risk, moderate marijuana frequency, high illicit drug use). Participants in the Meth & GHB User class were more likely to have an STI at baseline compared to those in the Low User class. Moreover, Heavy Use, Alcohol & Marijuana Use, and Meth & GHB Use reported a greater number of condomless anal sex partners compared to Low Use. Additionally, although Alcohol & Marijuana Use were more likely to use PrEP at the time of their most recent casual partner at baseline, they exhibited the smallest increase in PrEP use during follow-up compared to the other substance use classes. Our analysis did not unveil substantial differences in the success of implementation in terms of reach, suggesting that both implementation strategies effectively engaged YMSM with different levels of substance use. Our research showed a similar level of engagement, as evidenced by the completion rates of modules and time spent, across all substance use classes. However, those within the Meth & GHB category found the intervention highly acceptable, but less so compared to Low Use and Alcohol & Marijuana Use. Understanding how distinct substance use profiles influence intervention outcomes and exploring varied implementation methods can augment future prevention endeavors, broadening the scope and impact of public health initiatives.