[Skip to Navigation]
June 1986

Power in Hypothesis Testing-Reply

Arch Intern Med. 1986;146(6):1231. doi:10.1001/archinte.1986.00360180251048

We agree with Dr Chester's comments and draw his attention to the forthcoming article by Drs Makuch and Johnson,1 our accompanying editorial,2 and an article that outlines a cost-effectiveness approach for defining the clinically significant difference in the planning stages of a clinical trial.3

The purpose of our previous article was to highlight the difference in approaching possible false-negative conclusions in the pretrial and posttrial stages. To the extent that readers are interested in risk reductions that are smaller than 25% (and we agree that there are many cases where smaller differences would be considered clinically significant), the tables presented in our article will be inadequate.

In establishing therapeutic equivalency, an investigator must come to grips with the question: "What is a clinically significant difference?" Most investigators and clinicians use consensus or arbitrary approaches to answer this question. We believe that a more formal analytical approach

Add or change institution