Clinical trials: relevance of correlation between treatment modalities

Abstract

Background: Trials that do not allow rejection of the null hypothesis of no treatment effect may have had an inappropriate design. Trials are virtually never assessed for correlation between responses to different treatment modalities.

Methods: Using a model, a simple test and several published studies we studied the influence of correlation levels on the statistical power of clinical trials.

Results: The level of correlation between treatment modalities is a major determinant of the power of both self-controlled and parallel group clinical trials.

Conclusions: It is extremely relevant to assess correlation levels between treatment modalities of a trial a priori. With a presumably negative correlation a self-controlled design is likely to lack power. With a presumably positive correlation a parallel group design is likely to do so. Both designs are suitable for approximately zero correlations (e.g. comparison vs placebo).