Unblinded Sample Size Reestimation: A Case Study
Silke Jörgens, Vladimir Dragalin · 2023
Unblinded sample size reestimation (SSR) based on conditional power has become one of the most popular adaptive design features in clinical trials. The example the authors present in this chapter is modeled after a completed phase 3 trial that had an unblinded SSR design based on combining p-values from two stages. They introduce the design of the study and provide statistical considerations and details of the SSR design. They then describe the conduct of the study including the interim results and the Independent Data Monitoring Committee recommendations, implementation issues, and a description and discussion of the final analysis results. As the assumptions of the expected treatment effect and variability may or may not be upheld, an interim analysis was planned with the purpose of stopping the study due to futility or reestimating the sample size.