A Millian Look at the Logic of Clinical Trials

The use of a certain drug or treatment in the cure of a disease or condition should be based on appropriate tests of the hypothesis that said drug or treatment is efficacious in the cure of the disease or condition at hand. In this paper, we aim at elucidating how such “efficacy hypotheses” are tested and evaluated, especially in clinical trials. More precisely, we shall suggest that the principles governing the assessment of efficacy hypotheses, and, more generally, hypotheses of statistical causality, are provided by an appropriate statistical version of such a venerable procedure as the method of difference put forward by John Stuart Mill.