论文部分内容阅读
目前,在我国新药临床试验中,随机化是进行临床试验的规范程序,现在通用的随机化方法是1∶1均衡性随机,这样做的好处是在相同样本含量前提下可以获得较大的检验效能,并且从伦理学上来考虑,由于不能及时的根据已获得的处理结果,提早的判别何为较优疗效,重新调整分组概率,使得后续受试者面临额外的接受较劣疗效处理的风险。本文介绍一种基于贝叶斯的自适应随机化方法,以便解决上述问题。
At present, in clinical trials of new drugs in our country, randomization is the standard procedure for conducting clinical trials. The common randomization method is a 1: 1 balanced randomness. The advantage of doing so is that it can obtain a larger test under the same sample content Efficacy, and ethically, because of the inability to make early judgments about the superior outcome based on the treatment results obtained, the subgroup probabilities are readjusted so that subsequent subjects are at an additional risk of receiving less effective treatments. This paper presents a Bayesian adaptive randomization method to solve the above problems.