论文部分内容阅读
In clinical trials, sequential monitoring procedures are often implemented to stop the trials earlier, therefore decrease the cost and reduce the sample size.Response adaptive randomization (a sequential procedure) has been developed and demon strated to be desirable randomization for many clinical trials.Is it possible to com bine these two sequential procedures together in a clinical trial? In this talk, we answer this question theoretically by showing that the joint distribution of some im portant sequential statistics is asymptotically Brownian process in response-adaptive randomized trials.Therefore, we can apply sequential monitoring procedures to response-adaptive randomized trials.Simulated studies also support our theoretical results.