Confirm and Explore: A Stepwise Approach to Clinical Study Designs

Abstract

In general, the scientific learning process begins with exploration. Scientists learn from exploratory findings, generate a new hypothesis, and then perform experiments to confirm the hypothesis. After they do more exploration, and learn more, scientists continue with this process. The process is to explore first, then confirm, next explore and confirm again, and so on. The same is true for the entire drug development process.

However, in designing an individual phase 2 or 3 clinical trial, this thinking process needs to be reversed because each study is designed to address a specific clinical question. This question is formulated as the primary objective of the study. Clinical trials are designed to address this well-defined study objective. Hence this part is a confirmatory step. After the primary objective is achieved, scientists then explore all available data and see what additional information can be generated from the study.

In this article, we discuss the thinking process of clinical trials. It is clear that the clinical trial design follows a stepwise approach that first considers the confirmatory step by testing a pre-specified, well-defined statistical hypothesis, and then moves to the exploratory (or learning) step. We discuss a few examples to illustrate how this thinking process can be applied in designing some of the phase 2/3 studies.