随机模拟法验证非劣效临床试验样本量计算公式  

Sample Size Calculation in Noninferiority Trial by Monte Carlo Method

作　　者：王杨[1] 李卫[1] 成小如[1] 贾宣[1] 

机构地区：[1]中国医学科学院阜外心血管病医院(卫生部心血管病防治研究中心),100037

出　　处：《中国卫生统计》2008年第1期26-28,共3页Chinese Journal of Health Statistics

摘　　要：目的探讨并验证非劣效临床试验样本量计算方法。方法通过理论公式的推导,得到非劣效临床试验样本量计算公式,并用随机模拟的方法,使用该公式计算出的样本量估计实际的检验效能,以验证公式的正确性。结果由概率论严格推导得到样本量计算公式,并通过SAS随机模拟宏程序验证了公式的正确性,即模拟出的检验效能与最初带入公式计算时设定的预期的检验效能一致。结论样本量计算与临床试验设计有机结合的方法,解决了现行临床试验样本量计算方法与研究设计脱节的问题。Objective In order to verify sample size calculation formulas in noninferiority clinical trial. Methods Hypothesis test theory was used to prove the current formula in noninferiority clinical trials. Then, the test power was estimated by Monte Carlo method. Results Sample size was obtained by validated formula. Then, a randomized simulation was constructed by the sample size result combined with significant level and other parameters. One SAS macro program was coded for the relevant power simulation. Conclusion The properly formula was proved and validated in this paper.

关 键 词：临床试验 样本量 非劣效 检验效能 随机模拟 

分 类 号：R311[医药卫生—基础医学]