Purpose of review: This paper reviews cluster randomized trials for HIV prevention, focusing on the unique aspects of such trials and describes key cluster randomized trials for HIV prevention from the past decade. Relevant methodological issues are also reviewed.
Recent findings: Three cluster randomized trials evaluating sexually transmitted infection control as an HIV prevention tool have given mixed results. Recent modeling studies suggest that the differences are due to differences in the underlying populations rather than design differences. Two trials have shown reduced HIV incidence in intervention target groups without showing a community-wide effect. Areas of active methodological research include ensuring baseline comparability between arms in cluster randomized trials with relatively few clusters, and measuring incidence when the intervention may require time to be fully effective. Innovative approaches to measuring incidence based on so-called 'detuned assays' may be relevant to this problem.
Summary: Cluster randomized trials are qualitatively different from individually randomized trials. As the intervention population may differ from the assessment population, cluster randomized trials of infectious diseases can estimate both direct and indirect intervention effects. Most cluster randomized trial interventions involve behavioral or social components, so the generalizability of results to other settings is often difficult. Modeling studies can aid in the interpretation of results.