The Choice of Priorities in Contemporary Comparative Trials: Superiority vs non Inferiority. Causes and Effects

Abstract

The article compares two statistical approaches, which are commonly used in current comparative studies, a hypothesis that a drug is superior over another one (superiority) and a hypothesis that a drug is not inferior to another one in the efficacy and safety (non-inferiority). Using the example of specific studies, the difference between the methods and the tasks, for the solution of which one or another method should be applied, are shown. In order to prove the superiority in efficacy and safety of a new drug over an existing one, only a statistical approach that uses the "superiority" hypothesis is applicable. Studies using the "non-inferiority" hypothesis are generally used for comparing drugs, which are not considerably different in their efficacy, but the study drug has other advantages in the administration, storage, tolerability etc. The choice of statistical method is determined exclusively by the task of the study.