298 – Advances in Bayesian Clinical Trial Design
Bayesian Modeling of Crossover Studies of Rare Diseases Including Escape Phase
Lili Ding
Cincinnati Children's Hospital Medical Center
Bin Huang
Cincinnati Children's Hospital Medical Center
Yongchao Liu
University of Cincinnati
Early-escape design (Amery and Dony 1975; Temple 1994) limits patient's exposure to ineffective therapy by allowing patients escape from the randomized treatment or taking rescue medications. Typically, the escape rate or time to escape or the amount of rescue mediation is used as the study outcome, and the treatment effect is evaluated using only data prior to escape. However, data after escape phase can provide useful information; ignoring data after escape costs power that are valuable for studies of rare diseases or studies constrained by small sample size. Motivated by a recently completed crossover study of rare disease with early escape design, we propose a Bayesian approach that utilizes data from both before and after escape for evaluating treatment effect.
