倾向性评分法在中医复杂干预非随机化数据疗效评价中的应用  

Application of propensity score method in therapeutic evaluation of non-randomized data produced by TCM complex intervention

作　　者：叶晓勤[1,2] 谢雁鸣[1] 王利娜[3] 王永炎[1] 

机构地区：[1]中国中医科学院中医临床基础医学研究所,北京100700 [2]安徽中医药大学中医临床学院,合肥230038 [3]安徽中医药大学第二附属医院,合肥230061

出　　处：《中华中医药杂志》2015年第10期3414-3417,共4页China Journal of Traditional Chinese Medicine and Pharmacy

基　　金：国家高技术研究发展计划(863计划)项目(No.2007AA02Z4B2);国家自然科学基金项目(No.81173472);中国中医科学院自主选题项目(No.Z02108);安徽省自然科学基金项目(No.1408085QH184)~~

摘　　要：中医复杂或综合干预反映了中医临床实践过程中的实际情况,要准确界定中医复杂干预非随机化数据的有效性不是一件容易的事情,这使得验证其临床疗效变得极为艰难。倾向性分析方法是对控制混杂偏移的一种有效方法,该方法尽量将所有可能的混杂因素综合为一个变量,即倾向评分,从而使组间具有可比性。因此,该方法应用于中医复杂干预非随机化数据临床疗效评价是恰当和适合的。Complex or comprehensive intervention is a reflection of the actual situation during the clinical practice of TCM.It is not very easy to accurately define the efficacy of non-randomized data,which makes the evaluation of its clinical efficacy becomes extremely difficult.Propensity score analysis is an effective method for controlling confounding factors.All the possible confounding factors can be synthesized as a variable propensity score（PS） by using this method,so that there is comparability between groups.Therefore,it is appropriate and suitable to apply this method in the therapeutic evaluation of nonrandomized data produced byTCM complex intervention.

关 键 词：中医复杂干预 疗效评价 非随机化 混杂 倾向性分析 

分 类 号：R242[医药卫生—中医临床基础]