临床试验缺失数据的统计学考量  

Statistical considerations on missing data in clinical trials

作　　者：王骏[1] 韩景静 黄钦[1] 

机构地区：[1]国家食品药品监督管理总局药品审评中心,北京100038

出　　处：《中国临床药理学杂志》2016年第5期469-472,共4页The Chinese Journal of Clinical Pharmacology

摘　　要：缺失数据是临床试验中常见的重要问题,可能引起试验结果难于解释,甚至影响整个试验的推断和结论,但目前在国内临床研究中尚未引起足够的重视。实际应用中,申请人对待缺失数据的处理存在盲目应用统计学方法的现象,给新药安全有效性的评价和确证带来诸多困难。本文针对三个实际案例进行深入地阐述和分析,探讨对缺失数据的预防、估计量的选择、缺失数据的统计分析方法以及敏感性分析等,以期有助于申请人在临床试验中加强对缺失数据预防和处理方法的重视。Missing data is a common and important issue in clinical trials. It could make the results difficult to be explained,even influence the inference and conclusion of the trial. Currently,not enough emphasis has been given to the problem in China. In the real world of statistical operation in dealing with clinical trial missing data,sponsors often applied various methods without any rational,which introduced many difficulties for evaluating and confirming the safety and efficacy of new drugs. In this paper,three real cases were specifically brought out for illustration and analysis,meanwhile,the prevention of missing data,the choice of esti-mator,statistical analysis methods and sensitivity analysis were also discussed. The aim is to attract the sponsors to focus attention on prevention and treatment of missing data in clinical trials.

关 键 词：临床试验 缺失数据 统计学 

分 类 号：R954[医药卫生—药学]