基于混合先验的MRCT中疗效一致性的评价

Evaluation of consistency for a specific region in multiregional clinical trials based on a Bayesian method

目的:在国际多中心临床试验(multiregional clinical trials,MRCT)研究中,对一种基于先验的疗效一致性评价准则作修改,并比较准则的特性。方法:除特定中心外,其他中心用极大似然法估计疗效;特定中心采用基于混合先验的贝叶斯方法。根据贝叶斯公式,得到该中心的估计疗效,以此制定评价该中心疗效一致性的准则。比较4种准则的确保概率(AP_1,AP_2,AP_3,AP_4)和错判率(FR)。结果:多个因素影响准则的确保概率。在特定中心疗效接近MRCT疗效时,除AP_3外的其他准则均能得到一致性的结果;在特定中心无效时,AP_2和AP_3的错判率较低。结论:综合确保概率和错判率的结果,推荐AP_2作为评价特定中心与MRCT疗效一致性的准则。

AIM： To modify an empirical Bayes approach to evaluation of results for a specific region in multiregional clinical trials（ MRCT）,and then compare the properties of rules. METHODS： Other than a certain region（ eg,the first region）,the maximum likelihood estimation（ MLE） is applied to compute efficacy for other regions. For the first region,we use mixed normal priors. Combined with the data from the first region,the efficacy could be gained by Bayesian formula to construct consistency methods. And we compare the four methods in terms of assurance probability（ AP） and false rate（ FR）.RESULTS： Several factors affect AP. All the AP except for AP3 can conclude consistency when the efficacy of the first region is close to that of MRCT.In the results of FR,AP2 and AP3 perfom well when the efficacy of the first region is close to zero.CONCLUSION： According to the results of simulation tests,we suggest AP2 as the method for consistency evaluation for a specific region in MRCT.