有缺失数据的生物等效性评价的考虑要点  

Essential considerations on the evaluation of bioequivalence study with missing data

作　　者：孙华[1] 李相鸿[1] 胡骅[1] 徐毛迪 谢海棠[1] SUN Hua;LI Xiang-hong;HU Hua;XU Mao-di;XIE Hai-tang(Center for Drug Clinical Evaluation,Yijishan Hospital Affiliated to Wannan Medical College,Wuhu 241000,Anhui Province,China)

机构地区：[1]皖南医学院弋矶山医院药物评价中心,安徽芜湖241000

出　　处：《中国临床药理学杂志》2020年第18期2891-2895,共5页The Chinese Journal of Clinical Pharmacology

摘　　要：生物等效性研究中受试者脱落或各种原因造成的数据剔除,会导致两周期生物等效性集不均衡或不完整,在不同的统计算法或不同版本统计软件计算的结果可能不完全一致,在试验中存在离群值或残留效应时,数据缺失还会增加统计分析的复杂性,给生物等效性结果的判定带来偏倚。本文系统阐述生物等效性研究中数据缺失的常见原因、对策、含缺失数据生物等效性研究的考虑要点与处理原则,包括数据集的划分、统计模型的选择,统计结果的敏感性分析等,为国内申请人开展生物等效性研究提供参考。In a bioequivalence trial, it is not unusual to observe a volunteer either withdrawing from the trial for an administrative reason or being dropped from the trial by the investigator because of one or more protocol violations. The results calculated by different statistical algorithms or different versions of statistical software may not be completely consistent when the dataset is imbalance or incomplete. When outliers or carryover effect exist in the experiment, the absence of data will also increase the complexity of statistical analysis and bring bias to the determination of bioequivalence results. This paper systematically expounds the common causes and countermeasures of missing data in bioequivalence research and the consideration points and treatment principles of missing data bioequivalence research, including the division of data sets, the selection of statistical models, and the sensitivity analysis of statistical results, etc., so as to provide references for sponsors to carry out bioequivalence research.

关 键 词：生物等效性 缺失数据 离群值 不完整数据集 不均衡数据集 敏感性分析 

分 类 号：R97[医药卫生—药品]