To the Editor.
—We agree with Drs Peterman and Aral1 regarding the need for more randomized controlled trials of behavior change interventions for human immunodeficiency virus (HIV) and sexually transmitted disease prevention. However, we would like to clarify a few points.Randomized controlled trials are not always appropriate for evaluating HIV or acquired immunodeficiency syndrome (AIDS) prevention programs. Although randomized controlled trials offer the most rigorous level of evidence, they are not always scientifically appropriate, logistically feasible, economically practical, or ethically defensible.Peterman and Aral performed a limited literature search with an algorithm of the key words "trial," "random," "control," "behavior," and "change," and used only one database of computerized abstracts (AIDSLINE) from international AIDS conferences of 1989 through 1992. This would not be expected to produce all significant references for this complex and dynamic field that cuts across medical and social sciences. Some relatively comprehensive reviews of the
Gerber AR, Campbell CH, Dillon BA, Holtgrave DR. Evaluating Behavioral Interventions: Need for Randomized Controlled Trials. JAMA. 1994;271(17):1317-1318. doi:10.1001/jama.1994.03510410029017