样本含量估算方法探讨  

To discuss different calculation methods of sample size

作　　者：喻宁芳[1] 

机构地区：[1]广州市第一人民医院南沙医院,广东广州511457

出　　处：《临床医药文献电子杂志》2014年第8期468-469,共2页Electronic Journal of Clinical Medical Literature

基　　金：广东省科技计划项目(项目编号:73092)

摘　　要：目的:介绍和比较医学实验设计中不同的样本含量估算方法。方法:以PI3K抑制剂对小鼠气道炎症影响的实验研究*为例运用不同方法计算样本含量。结果:1公式法计算需12例2PASS软件Simple法计算需10例3Stata软件计算需8例,验算其检验效能:1-β>0.9结论:3种不同方法估算的样本含量都是合理有效的,实验研究人员可以以多种计算结果为依据,分析实验研究性质,综合考虑研究成本、可行性与伦理学要求对样本含量的影响确定合适的样本数。Objective： To introduce and compare different calculation Methods of sample size in experiment design.Methods： As an example of PI3K inhibitor reduces respiratory tract inflammation in a murine model of Asthma.Results： In method of formula,12;in PASS software,8;in Stata software, 10.1-β〉0.9.Conclusion： Proper analysis of the nature of research design,setting the correct parameters,Based on a variety of calculations to estimate the sample size, and then considering the research costs, feasibility and ethics requirements impact on sample size, and ultimately determine the most appropriate number of samples.

关 键 词：样本估算 实验设计 PASS软件 STATA软件 

分 类 号：R-33[医药卫生]