中医临床研究治疗前后数据比较存在的问题与对策  

Problems and countermeasures in the comparison of pre-and post-treatment data in clinical studies on traditional Chinese medicine

作　　者：毕京峰[1] 陈素红[2] 魏振满[1] 谭钧元[3] 朱真珍[1] 屠舒[1] 

机构地区：[1]解放军第三〇二医院新药临床试验中心,北京100039 [2]解放军第三〇二医院外科重症监护室,北京100039 [3]解放军第三〇二医院心医务部,北京100039

出　　处：《传染病信息》2011年第5期294-296,共3页Infectious Disease Information

基　　金：国家"十一五"科技重大专项(2009ZX10005-017);中国人民解放军第三〇二医院院内课题(YNKT2010012)

摘　　要：目的对当前中医药防治病毒性肝炎临床科研文献中治疗前后数据比较所采取的统计方法进行辨析。方法模拟某慢性乙型肝炎临床研究的前后治疗数据,分别应用两样本t检验和配对样本t检验进行统计分析,比较结果。结果由于两样本t检验和配对样本t检验主要针对独立样本之间的差异作出统计推断,而治疗前后数据比较属于来自同一样本的前后2次观察,前后之间相互影响,并非独立数据。因此,将两样本t检验或配对样本t检验用于治疗前后数据之间的比较是错误的。结论两样本t检验或配对样本t检验并不适合治疗前后数据之间的比较,应用重复测量分析才是正确的方法。Objective To differentiate statistical approaches adopted for the comparison of pre-and post-treatment data in clinical research literature on the prevention and treatment of viral hepatitis with traditional Chinese medicine. Methods Pre-and post-treatment data in a clinical study on chronic hepatitis B were stimulated, two sample t-test and paired t-test were used to analyze the data statistically, and the results of statistical analysis were compared. Results Two-sample t-test and paired t-test led to statistical inference that was based on the difference between independent samples. However, the pre-and post-treatment data were obtained from 2 separate observations of the same sample before and after treatment. The two groups of data affected each other and were not independent. Therefore, it was wrong to compare the pre-and post-treatment data with either two-sample t-test or paired t-test. Conclusions Neither two-sample t-test nor paired t-test is the appropriate approach to the comparison of the pre-and post- treatment data. Repeated measures analysis is the right approach.

关 键 词：中医药疗法 方差分析 配对 治疗结果 

分 类 号：R242[医药卫生—中医临床基础]