ABSTRACT
This chapter uses multiple Bayesian adaptive design examples to illustrate some essential elements of Complex Innovative Design (CID), while noting that CID is broader and not necessarily restricted to Bayesian design. The innovative design strategies including the consideration of multiple endpoints, the risk benefit tradeoff, adaptive randomization and adaptive decision rule. These strategies can be combined with other statistical and design methods to address a variety of clinical trial challenges. The independent approach, which evaluates the treatment effect in each cancer type independently, avoids these issues, but is less efficient and often lacks power to detect the treatment effect due to the limited sample size in each cancer type. The independent approach, which evaluates the treatment effect in each cancer type independently, avoids these issues, but is less efficient and often lacks power to detect the treatment effect due to the limited sample size in each cancer type.
