Chapter 14 . Inferential Statistics , Descriptive Statistics and The Analysis Plan

Provided that the data show that a compound is safe, it is the primary efficacy analysis that determines whether the compound is approved. The secondary analyses support, clarify and confirm, but only one prospectively planned efficacy analysis can be primary in support of an effectiveness claim for a given trial. Why should this be so? And how can we design the trial and set up the primary efficacy analysis to get the best chance of confirming efficacy? The object of this chapter is to explain, in a simple but concrete way, the theoretical ideas behind the primary efficacy analysis. Many people treat statistics as mumbo–jumbo, but they don’t have to. There are commonsense reasons behind statistical ideas. While one might be impatient to get beyond the concepts and quickly arrive at the bottom line, this hurry is a mistake. Using statistical concepts thoughtfully can mean major (possibly multi–million dollar in a large study) savings in the study design and, much more importantly, can make the difference in turning a clinically beneficial compound into an approved and marketable reality.