中医临床研究治疗前后数据比较存在的问题与对策

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...

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

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.