Modern Approaches to Clinical Trials Using SAS: Classical, Adaptive, and Bayesian Methods

7.6 Conclusions

Current dose-finding practices often are not satisfactory: Many dose-finding studies are not properly designed (too few doses, dose-ranges too narrow, sample sizes not large enough, etc.) and the inferences drawn from them are not adequately accurate, thus leading to major consequences in drug development that are undesirable (e.g., high failure rate in phase III, need for label changes after approval, etc.).

Here, we have described a unified strategy for analyzing dose-finding studies, including the testing for a dose-response signal and the selection of one or more doses to take into further development. The proposed methodology combines the advantages of the multiple comparison and modeling approaches, consisting of a multistage ...