Abstract:
|
Dose finding studies are typically designed as mini-Phase 3 trials, with the hope that they may serve as one of the pivotal studies in the submission package. This has led to the prevalent use of pairwise comparison approaches when designing such studies, with emphasis placed on power and strong control of Type I error rate. A more efficient approach is to embrace the learning nature of dose finding and rely on model-based approaches to design studies aimed at understanding the dose-response relationship and estimating target doses of interest, such as the minimum effective dose. This talk will focus on the design of dose finding studies in the context of one of such model-based approaches, MCP-Mod, contrasting it to the traditional pairwise comparisons paradigm. Extensions to adaptive dose finding designs will be discussed.
|