差异性、等效性、非劣效性和优效性设计中的t检验  

T-test in difference,equivalence,non-inferiority and superiority design

作　　者：陈卫[1] 徐利娜[1] 迭敏[1] 刘丽珍[1] 陈卫中[2] 

机构地区：[1]成都医学院医学检验系,医学检验专业2005级成都610083 [2]成都医学院公共卫生学教研室,成都610083

出　　处：《成都医学院学报》2009年第3期211-213,共3页Journal of Chengdu Medical College

基　　金：2008年成都医学院大学生创新性实验项目(SYSCX2008-46)

摘　　要：随着新药临床试验的开展越来越多,也越来越规范,等效性检验、非劣效性检验和优效性检验被广泛地提出,但也常遇到以传统显著性检验代替非劣效、等效和优效性检验的错误。因此,本文从样本均数的抽样分布出发,就它们的适用条件、检验目的、假设形式、公式的区别和联系进行了阐释。Along with more and more development and standard of clinical trials of new drugs, equivalence, non-inferiority test and superiority test are widely introduced. However,wrong replacements of non - inferiority, equivalence and superiority tests with traditional significance test are often encountered. Therefore, based on theoretical derivation of the sampling distribution of two-sample mean difference,this article explained the applicable condi tions, test purposes, hypothesis, and difference and association among the formulas of these three tests.

关 键 词：非劣效检验 等效性检验 优效性检验 显著性检验 T检验 

分 类 号：R195.1[医药卫生—卫生统计学]