循证医学中统计指标的正确应用

Correct application of statistical measures in evi-dence-based medicine

循证医学实践包括提供证据(创造证据)和证据的应用,提供证据的研究人员(doer)应能正确选用有关的统计方法,证据的应用者(user)应能认识某研究证据中所使用的统计方法,以此推断其结果的可靠性。因此,理解有关的统计方法,对循证医学的研究者和应用者均是重要的内容。介绍循证医学中的常用统计指标,如OR(比值比,oddsratio)、RR(相对危险度,relativerisk)、RRR(相对危险度减少率,rela-tiveriskreduction)、ARR(绝对危险度减少率,absoluteriskreduction)、NNT(需要处理的患者数,numberneededtotreat)等的意义和用途,并着重介绍了这些指标的可信区间计算方法、用途和意义,可以供循证医学研究者参考。

Evidence-based medicine(EBM)includes the supply(or creation)and application of evidence.The doe r who provide evidence should accurately choose the relevant statistic methods,and the user of evidence should recognize the statistical measures used in a specific evidence research,and deduce the re liability of the conclusion.Therefore,the understanding of relevant statistical measures is vital to not only the researchers,but also the users of the EBM.In this paper,we introduce meaning and purp ose of statistical measures in EBM,for example,OR,RR,RRR,ARR,NNT.And we intro-duce calculating,meaning and purpose of confidence interval(CI ),It' s referance for user and doer of EBM in China.