Design of human immunodeficiency virus intervention trials in developing countries

Abstract

Randomized controlled trials are recognized as the 'gold standard' for evaluating the effect of health interventions, yet few such trials of human immunodeficiency virus (HIV) preventive interventions have been conducted. We discuss the role of randomized trials in the evaluation of such interventions, and we review the strengths and weaknesses of this and other approaches. Randomization of clusters (groups of individuals) may sometimes be appropriate, and we discuss several issues in the design of such cluster-randomized trials, including sample size, the definition and size of clusters, matching and the role of base-line data. Finally we review some general issues in the design of HIV prevention trials, including the choice of the study population, trial end points and ethical issues. It is argued that randomized trials have an important role to play in the evolution of HIV control policies.