实验设计中样本含量估算方法探讨——以PI3K抑制剂对小鼠气道炎症影响的实验研究为例  

Discussion on different estimation methods of sample size in experiment design——taking an experimental study on influence of PI3K inhibitor on mouse respiratory tract inflammation as an example

作　　者：喻宁芳[1] 喻宁芬[2] 

机构地区：[1]广州市第一人民医院南沙医院检验科,广东广州511457 [2]广州市第一人民医院儿科,广东广州510180

出　　处：《国际检验医学杂志》2015年第9期1193-1194,1197,共3页International Journal of Laboratory Medicine

基　　金：广东省科技计划项目(73092);广州市医药卫生科技资助项目(2007-YB-026)

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。方法以磷酸酰肌醇-3激酶(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。结果公式法计算需12只,PASS软件Simple法计算需8只,Stata软件计算需10只,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 the different estimation methods of sample size in the medical experiment de- sign. Methods Taking the experimental study on the influence of PI3K inhibitor on respiratory tract inflammation in mouse. Results The formula method estimation needed 12 mice,while the PASS software Simple method needed 8 mice, and the stata software method needed 10 mice. The power of test was checked： 1-~0.9. Conclusion The sample size estimated by 3 different kinds of method is rational and effective. The experimental researchers can take the multiple calculated results as the basis to analyze the property of experimental study,comprehensively consider the influence of the research costs, feasibility and ethics requirements on the sample size and to determine the most appropriate sample number.

关 键 词：样本估算 实验设计 PASS软件 STATA软件 

分 类 号：R965[医药卫生—药理学]