论文部分内容阅读
目的 介绍配对设计生存资料的统计分析方法及其在临床试验中的应用。方法 运用基于协方差矩阵稳健估计的Cox回归模型 ,通过调用STATA统计软件包中stcox命令中的cluster选项 ,对配对设计生存资料进行统计分析。并用治疗烧伤临床试验中的实例比较了运用和不运用配对设计生存资料的统计分析方法的两种结果。结果 实例分析表明运用配对设计生存资料的统计分析方法时 ,可得出治疗组的创面愈合时间小于安慰剂组的结论 ,而不运用配对设计生存资料的统计分析方法时 ,其结论为治疗组和安慰剂组创面愈合时间的差异无统计学意义。结论 在对配对设计的生存资料进行统计分析时 ,应选用配对设计生存资料的统计分析方法 ,否则将降低检验的效率 ,可能得出不正确的结论
Objective To introduce the statistical analysis method of paired design survival data and its application in clinical trials. Methods The Cox regression model based on the robust estimation of covariance matrix was used to analyze the data of survival of paired design by calling cluster option of stcox command in STATA statistical software package. Two examples of statistical analysis of survival data with and without paired design were compared with those in clinical trials of burn treatment. Results of case analysis showed that the use of paired design survival data of statistical analysis, the treatment group wound healing time can be drawn conclusions less than the placebo group, without the use of paired design survival data statistical analysis, the conclusions of the treatment group and There was no significant difference in wound healing time between placebo group and placebo group. Conclusions In the statistical analysis of survival data of paired designs, the statistical analysis method of paired design survival data should be used, otherwise the inspection efficiency will be reduced and the incorrect conclusions may be drawn