Point estimation for adaptive trial designs.

摘要: Recent FDA guidance on adaptive clinical trial designs defines bias as "a systematic tendency for the estimate of treatment effect to deviate from its true value", and states that it is desirable obtain report estimates effects reduce or remove this bias. In many designs, conventional end-of-trial point are prone bias, because they do not take into account potential realised adaptations. While much methodological developments have tended focus control type I error rates power considerations, in contrast question biased estimation has received less attention. This article addresses issue by providing a comprehensive overview proposed approaches an design, well illustrating how implement them. We first discuss can affect standard estimators critically assess negative impact have. then describe compare unbiased bias-adjusted different types designs. Furthermore, we illustrate computation practice using real example. Finally, propose set guidelines researchers around choice reporting following design.