药物流行病学研究中处理偏倚的方法  

The Method of Dealing with Bias in Pharmacoepidemiology

作　　者：陈明珠 胥洋 张兴会 孙凤 王树月 张灵健 王俊彦 宗晓郁 李慧[1] 周锦 潘发明[5] 方世平[6] 宫建[1] 詹思延 Zhao Zhenying Zhu Xiaohong Chen Mingzhu;Xu Yang;Zhang Xinghui;Sun Feng;Wang Shuyue;Zhang Lingjian;Wang Junyan;Zong Xiaoyu;Li Hui;Zhou Jin;Pan Faming;Fang Shiping;Zhao Zhenying;Zhu Xiaohong;Gong Jian;Zhan Siyan(Research Group of Jian Gong on Pharmacoepidemiology and Clinical Drug Evaluation,Shenyang Pharmaceutical University,Shenyang 110016,China;Department of Medical Epidemiology and Biostatistics,Karolinska Institutet;Department of Epidemiology and Biostatistics,Peking University Health Science Center;Department of Pharmacy,School of Heping Hospital,General Hospital of Northern Theater Command of China Medical University;Department of Epidemiology and Health Statistics,School of Public Health,Anhui Medical University;Department of Pharmacy,Zhongnan Hospital of Wuhan University)

机构地区：[1]沈阳药科大学药物流行病与临床药物评价课题组,沈阳110016 [2]瑞典卡罗琳斯卡学院 [3]北京大学医学部公共卫生学院 [4]中国医科大学北部战区总医院和平院区学院药学部 [5]安徽医科大学公共卫生学院流行病与卫生统计学系 [6]武汉大学中南医院药学部 [7]Department of Pharmacy,Zhongnan Hospital of Wuhan University [8]Department of Medical Epidemiology and Biostatistics,Karolinska Institutet

出　　处：《药物流行病学杂志》2021年第7期431-434,495,共5页Chinese Journal of Pharmacoepidemiology

摘　　要：药物流行病学是在应对处理药源性疾病、药物不良反应过程中发展起来的一门医药学领域的综合应用性学科,是流行病学的分支之一。正确评价药物在人群中的利用及其效应主要取决于合理的研究设计和得当的实施和分析方法来控制系统误差。本文基于《中国药物流行病学研究方法学指南》,检索相关文献,并结合实例,对药物流行病学研究中处理偏倚的方法进行解读,系统理解处理偏倚的方法。当前国内药物流行病学研究对偏倚处理的方法仍需加强,需重点关注选择偏倚、错分偏倚、混杂偏倚、时间依赖性偏倚、系统评价,根据产生原理及过程,在试验设计阶段使用随机化、盲法、限制、匹配等方法消除或减少偏倚,系统评价时采用漏斗图、秩相关分析法、回归分析法、剪补法等评估发表偏倚,提高药物流行病学在医学中的价值,促进全民健康发展。Pharmacoepidemiology is one of the branches of epidemiology and a comprehensive applied subject in the field of medicine and pharmacy developed in the process of dealing with drug-induced diseases and adverse drug reactions. In this study, the correct evaluation of drug use and its effect on the population mainly depends on the reasonable research design, proper implementation, and analysis methods to control systemic errors. We review the domestic and foreign literature from PubMed and CNKI based on the Guide on Methodological Standards in Pharmacoepidemiology(hereinafter referred to as “the guideline”), closely follow the content of the guide, and combine with examples, to interpret and extend the methods of dealing with bias in pharmacoepidemiology research to systematically understand the methods of dealing with bias in pharmacoepidemiology research. The methods of bias treatment in domestic drug epidemiology research still need to be strengthened at present. The emphasis should be on selecting bias, misclassification bias, confounding bias, time-dependent bias, and systematic review according to their generation principle and process. Using randomization, blindness, restriction, matching and other methods to eliminate or reduce bias in the trial design phase, and systematically evaluate the use of funnel map, rank-related analysis method, regression analysis method, cutting method to assess publishing bias. It will provide a reference for further research to improve the application of pharmacoepidemiology in medicine so that promoting the healthy development of the whole people.

关 键 词：药物流行病学 偏倚 药源性疾病 药品不良反应/事件 

分 类 号：R181.35[医药卫生—流行病学]