基于α调整的国际多中心临床试验目标区二分类终点有效性研究的桥接方法  

A Bridging Method Based on α Adjustment to Evaluate the Effectiveness of MRCT in Target Region for Dichotomous Endpoint

作　　者：虞成凯 王陵[1] 戚洋洋 李婵娟[1] 李晨[1] 蒋志伟[1] 夏结来[1] 

机构地区：[1]第四军医大学军事预防医学院卫生统计学教研室,710032

出　　处：《中国卫生统计》2016年第1期20-23,共4页Chinese Journal of Health Statistics

基　　金：国家自然科学基金资助(项目编号:81273176;81302509;81473069)

摘　　要：目的探索采用合理的调整后检验水准α'作为国际多中心临床试验(MRCT)在目标区的决策依据的可行性,为MRCT在目标区通过审批提供参考。方法利用Monte Carlo模拟构建二分类资料、优效性设计、包含目标区的MRCT数据模型,比较不同规模的MRCT在检验水准α=0.05水平显示试验组优效的前提下,目标区采用调整后检验水准α'作为决策依据的条件Ⅰ型错误率(CFPR)、条件检验效能(CP)随目标区样本量占总样本量比例K的变化情况。结果在不同规模的MRCT中,目标区的CFPR和CP均随着检验水准α'、样本比例K的增加而增加;当α'=0.5时,CFPR和CFPR'基本可以被控制在50%和5%左右。当样本比例K固定时,目标区CP相对稳定,尤其当K≥30%时随着α'的增加目标区CP几乎不受MRCT总样本量N的影响,且当α'不低于0.3时基本可以确保足够的CP(≥80%)。当K≥30%且设置α'=0.5时,即使目标区试验药物疗效略低于MRCT的疗效即f≥0.8,也可以确保目标区足够的CP和CP'(≥80%)。结论建议MRCT在方案设计时,目标区的样本比例不应低于30%,试验数据在目标区的检验水平不能超过0.5。该标准适用于目标区试验药物疗效略低于MRCT的疗效即f≥0.8的情形,若估计目标区试验药物疗效与MRCT的疗效相当,检验水平可适当降低,但不建议低于0.3。Objective To explore the feasibility of adopting rational adjusted significance level as the decision rule in the target region in MRCT and make reference for evaluating new drug application of MRCT in target region. Methods By Monte Carlo simulation, we studied the conditional false positive rate（CFPR） and conditional test power（CP） changing with the proportions of the sample size in target region by adopting different significance levels given that the treatment effect is shown to be significant under the significance level a = 0. 05 based on overall MRCT data for dichotomous data. Results The CFPR and CP are increasing with the significance level c~＇ and the ratio of sample size K increasing in target region. When the significance level a＇ =0. 5 the CFPR and CFPR＇ can be controlled under the 50% and 5% in basic. When the ratio of sample size is fixed, the CP in target region is relatively stable and can＇t be influenced by the sample size of local region. Especially when K≥30%, the CP in target region is almost not affected by the total sample size. When a＇ isn＇t lower than 0. 3 ,we can ensure enough CP （ ≥80% ）. Conclusion When protocol is designed,it is advisable that the ratio of sample size K shouldn＇t be less than 30%. The significance level a＇ in target region shouldn＇t exceed 0. 5. The decisional rule is suitable for the difference between the trial and the control group in target region is more than 0. 8 times the difference in MRCT. If the estimated treatment effect of the drug in target region is almost like it in MRCT, the significance level can be lower, however,less than 0. 3 is not recommended.

关 键 词：条件Ⅰ型错误率 条件检验效能 国际多中心临床试验 MONTE CARLO模拟 

分 类 号：R969.4[医药卫生—药理学]