多个连续性共同终点非劣效临床试验的样本量估计

Sample Size Determination with Multiple Co-primary Continuous Endpoints in Non-inferiority Trials

目的针对多个以连续性正态分布变量作为共同终点的非劣效临床试验,探讨保护全局把握度的样本量估计方法。方法在阐明优效性临床试验多个连续性共同终点样本量估计的理论基础上,扩展提出了关于非劣效临床试验多个连续分布共同终点的样本量估计方法。结合临床麻醉学中一个重复测量终点非劣效临床试验的实际案例,基于其预试验的相关结果,在不同的I类错误水平、各重复测量间不同的相关系数、不同的非劣效界值设定下,估计了满足一定的全局把握度(例如80%)的样本量,并借助Monte-Carlo模拟方法对估计结果进行验证。结果本文介绍的非劣效临床试验多个连续性正态分布共同终点的样本量估计方法具有严密的统计理论基础,表明能较好的保护全局把握度,结合实际案例进行的样本量估计结果模拟研究验证了方法的正确性。结论本文为非劣效临床试验多个连续分布共同终点的样本量估计提供了方法学支持,具有较强的实用价值。

Objective To explore a sample size determination that protects overall power in non-inferiority clinical trials with multiple co-primary continuous endpoints.Methods Based on theoretical analysis of the sample size determination with co-primary endpoints in superiority clinical trials,the sample size determination with multiple co-primary continuous endpoints was proposed in non-inferiority clinical trials and was applied in clinical anaesthesiology to estimate the sample size which is repeated measurement.According to pre-experiment,the sample size was estimated to control the overall power(eg.80%) with different levels of the type I error rate,different correlation coefficients and different non-inferiority margins among continuous normal distribution variables.The Monte Carlo simulation method is used to verify the estimation result.Results The sample size determination in co-primary continuous endpoints in non-inferiority clinical trials described in this paper has a strict statistical theory basis and could protect the overall power better which is verified by Monte Carlo simulation with the case study.Conclusion The method in the paper provides methodological support for the sample size determination with co-primary continuous endpoints in a non-inferiority clinical trials which has strong practical value.