Abstract

The design of clinical trials is an art on its own. There are numerous, partially conflicting aspects a clinical trial has to satisfy. First and most importantly, the trial, especially in phase III trials, should ideally demonstrate a statistically significant effect of the drug under investigation on a clinically relevant primary endpoint. To accomplish this task, several aspects need to be considered. A crucial aspect at this stage is the expected effect size of the drug and the known variation of the primary endpoint measure, which together govern the power calculation and the sample size needed. As any treatment effect will need time to develop, time is also an important component: if the study is too short important effects or relevant but rare adverse events may be missed; if the study is too long patient retention may be difficult and dropouts may lead to missing data. A realistic estimate of the expected treatment effect on the primary endpoint and a sound power calculation are the mainstay of a successful study.