Factorial designs in clinical trials: options for combination treatment studies.

Abstract

This study reviews the use of factorial designs in clinical trials investigating combinations of therapies. Factorial designs may be used when (1) the factors are regarded as being independent or (2) the factors are thought to be complementary and a specific aim is to investigate these interactions. We describe what is meant by a factorial design and the issues that need to be addressed when using such a design. We discuss these issues in general and describe how they have been addressed in various prevention trials and in the COMBINE Study, which is a treatment trial of combinations of therapies for alcohol dependence. Trials of type (1) can provide substantial cost savings in conducting multiple unrelated prevention studies in the same group of participants. Such a factorial trial poses few design challenges beyond those of a standard parallel group trial. Trials of type (2) require consideration of aspects that are intrinsic to the factorial design. A factorial design is a useful way to examine the effects of combinations of therapies, but it poses challenges that need to be addressed in determining the appropriate sample size and in conducting interim and final statistical analyses.