Factorial trials are common, but standards for their analysis and reporting have not been established. The most powerful analysis of a factorial trial is performed at the margins of the (in the simplest form 2 by 2) table. However, when an interaction between treatments exist, the analyses has to be performed inside the table, which is less efficient. The possibility of unrecognized interactions between therapies therefore causes concern.
McAlister et al. performed a systematic review to assess the magnitude of the problem. They searched MEDLINE, EMBASE, and the Cochrane Controlled Trials Register, and identified eventually 44 factorial trials. Of these 39 (89%) presented the results at the margin, and 29 (66%) specified that interactions were sought.
The authors conclude that investigators are appropriately restricting their use of the factorial design to those situations in which the treatments do not have the potential for substantive interaction.
1. McAlister FA, Straus SE, Sackett DL, Altman DG. Analysis and reporting of factorial trials – A systematic review. JAMA 2003;289:2545-2553.