Background: MSM remain at great risk for HIV infection. Program planners and policy makers need descriptions of interventions and quantitative estimates of intervention effects to make informed decisions concerning prevention efforts.
Objectives: 1. To locate and describe outcome studies evaluating the effects of behavioral and social interventions targeting MSM. 2. To summarize the effectiveness of these interventions among MSM. 3. To stratify results by characteristics of interventions and participants. 4. To identify gaps and indicate future research, policy, and practice needs.
Search strategy: We searched electronic databases (MedLine, PsycInfo, etc.); several current journals (e.g., AIDS, AIDS and Behavior, AIDS Education and Prevention, American Journal of Public Health, Journal of Acquired Immune Deficiency Syndromes, etc.); manuscripts submitted by researchers; bibliographies of relevant articles; and other published reviews, for published and unpublished reports from 1988 through 1997.
Selection criteria: Studies were considered in scope if they examined the effects of behavioral interventions to reduce risk for HIV or STD transmission. We reviewed studies in scope for criteria of outcome relevance (measurement of at least one of a list of behavioral or biologic outcomes, e.g., unprotected sex or incidence of HIV infections) and methodologic rigor (randomized controlled trials or certain strong quasi-experimental designs with comparison groups).
Data collection and analysis: As of June 1998 we had identified 13 eligible studies. Twelve studies (7 trials of small group interventions, 3 community-level interventions, and 2 individual level interventions) reported intervention effects on unprotected sex. Because few studies reported effects on condom use (3 studies), number of sex partners without regard to condom use (4 studies), or HIV or STD incidence (1 study in which no infections occurred) we do not address these outcomes at this time. We present those analyses which can be performed in the current Cochrane RevMan software, followed by more complete analyses that permit inclusion of community-level studies, adjustment for baseline conditions, calculation of effect sizes from a wider variety of statistics (e.g., an F-statistic from a one-way ANOVA), and simultaneous meta-analysis of continuous and dichotomous outcomes (Johnson 2002b). We translate the summary effect to reduction in risk behavior based on the background prevalence of unprotected sex. Finally we provide analyses stratified by intervention content (interpersonal skills addressed or not), intervention format (community vs small group or individual) and mean age of participants (23 to 31 vs. 32 to 36).
Main results: A summary measure of intervention effects on reducing unprotected sex was favorable (odds ratio = 0.73) and statistically significant (CI, 0.60 to 0.88), corresponding to a 23% reduction in the proportion of men engaging in unprotected sex. Effects were homogeneous among studies, but were slightly more favorable among community-level interventions, those that served populations in their 20s rather than their 30s, and those that promoted interpersonal skills.
Reviewer's conclusions: These studies demonstrate that interventions can promote risk reduction among MSM. Yet given the epidemiology of HIV in Pattern I countries, the small number of rigorous controlled intervention trials for this population is striking. Many more rigorous evaluations of HIV prevention efforts with MSM are needed to ascertain with confidence the effects of specific intervention components, population characteristics, and methodologic features.