Three-Stage Design for Drug Screening Trials Based on Continuous Endpoints

• Main Authors: Shan-YunHuang, 黃山耘
• Other Authors: Mi-Chia Ma
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/19133975740345468628

碩士 === 國立成功大學 === 統計學系碩博士班 === 100 === Due to the development of target therapy, the evaluation in cancer clinical trials will change. Different with drugs of chemotherapy, the drug of target therapy belong to the cytostatic drugs and the main goal is to stopping the growth of tumor for a period of time. Hence the traditional methods of evaluating the tumor by tumor response rate are meaningless. Tsou et al. (2008) proposed a two stage drug screening trials based on continuous endpoints to improve Simon’s two-stage design which used tumor response rate as primary endpoints. We extend the two-stage drug screening trials to a three-stage design based on the concept of Tsou et al. (2008) and change the normal distribution to exponential distribution and gamma distribution. To fix type I and type II error rate, the best solution can be obtained by minimized expected sample size and total sample size.