随机对照临床试验设计中的样本量估算方法

Methods of sample size estimation in randomized controlled clinical trial design

根据随机对照试验报告统一标准(CONSORT)声明和PICO设计指导原则,样本量估算是随机对照临床试验设计的重要组成部分。合理的样本量是获得高级别临床研究证据的重要保障,样本量过小不能保证得出可靠的研究结论;而样本量过大会造成不必要的人力、物力和财力的浪费,同时增加研究的难度。本文从影响样本量大小的参数入手,指出样本量应依据临床试验研究的主要疗效评价指标来计算,并明确“定量变量”和“定性变量”指标的获取途径。以举例方式分别介绍差异性临床研究设计、优效性临床研究设计、非劣性临床研究设计和等效性临床研究设计的样本量估算方法,以期为研究者今后开展最优尺度回归分析提供参考。

The methods of sample size estimation is an important part of the design of a randomized controlled clinical trial according to the Consolidated Standards of Reporting Trials(CONSORT) statement and PICO design guidelines.A reasonable sample size is an important safeguard for obtaining evidence from high-level clinical research.Too small a sample size cannot guarantee reliable research conclusions,while too large a sample size can result in unnecessary waste of human,material and financial resources,as well as increase the difficulty of the research.This article starts from the parameters affecting the sample size,points out that the sample size should be estimated based on the main efficacy evaluation indicators of clinical trial research,and clarifies the ways to obtain the indicators of “quantitative variables” and “qualitative variables”.The methods of sample size estimation for differential clinical research designs,superiority clinical research designs,non-inferiority clinical research designs and equivalence clinical research designs are introduced by way of examples so as to provide reference for researchers to conduct optimal scale regression analysis in the future.