In the United States, there continues to be high incidence of HIV infection among men who have sex with men (MSM), who represent 57% of new infections in 2009. While many studies report associations between non-injection substance use and sexual risk behavior among MSM, overall results are mixed. Summarizing these studies is difficult because researchers have used a variety of assessment periods for substance use and sexual behavior. We review the scientific literature on event-level measures, which assess substance use and sexual risk behavior immediately before or during a sexual encounter and provide the most precise link between these two behaviors. From January 2009 through March 2010, we searched four databases: Ovid (MEDLINE and PsycINFO), Web of Knowledge, and Sociofile. Across studies, results varied by substance with little within substance consistency or a lack of research except for two notable exceptions: methamphetamine and binge alcohol use. The findings underscore the importance of providing HIV risk-reduction interventions for substance-using MSM.

Despite several advantages to bringing couples together to learn how to protect themselves and new-born children from the risk of HIV infection, most interventions are designed for individuals or groups, not for dyads. This meta-analysis provides a direct test of whether couple-based interventions are more effective in promoting HIV protective behaviors than interventions delivered to individuals. We conducted systematic searches of five electronic databases and 60 journals. Eligible studies were controlled trials or prospective cohort designs; evaluated a couple-based intervention compared to an individual-level intervention; assessed at least one HIV prevention outcome (e.g., protective sex, drug use, HIV testing, medication adherence, and sexually transmitted infections [STI]); and were published between 1988 and 2014. Fifteen interventions, including 21,882 participants from China, Kenya, Rwanda, Tanzania, Trinidad, Zambia, and the USA, were evaluated. The results of random-effects models showed statistically significant intervention effects for protective sex (OR = 1.60, 95% CI = 1.21, 2.11), HIV testing (OR = 1.79, 95% CI = 1.31, 2.45), and Nevirapine uptake (OR = 1.51, 95% CI = 1.02, 2.24). The evidence demonstrates the usefulness of couple-based interventions in protecting individuals, partners, and new-born children from the risk of HIV transmission and infection.

Gay, bisexual, and other men who have sex with men (MSM) are disproportionately affected by HIV but few MSM-specific evidence-based interventions (EBIs) have been identified for this vulnerable group. We conducted a systematic review to identify reasons for the small number of EBIs for MSM. We also compared study, intervention and sample characteristics of EBIs versus non-EBIs to better understand the challenges of demonstrating efficacy evidence. Thirty-three MSM-specific studies were evaluated: Nine (27 %) were considered EBIs while 24 (73 %) were non-EBIs. Non-EBIs had multiple methodological limitations; the most common was not finding a significant positive effect. Compared to EBIs, non-EBIs were less likely to use peer intervention deliverers, include sexual communication in their interventions, and intervene at the community level. Incorporating characteristics associated with EBIs may strengthen behavioral interventions for MSM. More EBIs are needed for substance-using MSM, MSM of color, MSM residing in the south and MSM in couples.