我国药品不良反应自发呈报系统中信号强度分级阈值的探索  

Exploration of Signal Intensity Threshold in China Adverse Drug Reaction Spontaneous Reporting System

作　　者：史文涛[1] 侯永芳[2] 叶小飞[1] 朱田田[1] 张天一[1] 吴桂芝[2] 贺佳[2] 

机构地区：[1]第二军医大学卫生统计学教研室,上海200433 [2]国家药品不良反应监测中心

出　　处：《药物流行病学杂志》2015年第8期466-469,共4页Chinese Journal of Pharmacoepidemiology

基　　金：国家自然科学基金项目(编号:.81202285;编号:.81373105);上海市自然科学基金(编号:12ZR1453700);上海市公共卫生重点学科(编号:12GWZX0602);上海市领军人才计划(编号:022);上海市卫生和计划生育委员会科研课题(编号:201440379);罗氏(中国)投资有限公司提供经费资助

摘　　要：目的:我国药品不良反应自发呈报系统报告呈逐年增长趋势,由不相称测定分析得到的药品不良反应组合数也较多,不利于人工进行信号的判断以及强信号的及时发现。因此,为研究拟利用系统学知识构建一种快速识别不良反应强信号的方法。方法:本研究根据统计学原理,结合压缩估计(α=0.5),将ROR、PRR、IC三种信号检测方法不相称测定值进行强度分级,为之后进行人工筛选、药品风险管理、评价及决策提供依据。结果:ROR、PRR、IC三种方法压缩估计之后强信号阈值分别为12.84、11.67、3.83。结论:该结果能较好地用于检测强信号,将被运用至国家药品不良反应监测中心新版信号检测系统中。Objective： In recent years, the number of spontaneous reports about adverse drug reaction in China has grown rapidly. It is difficult to judge a signal by manual screening after disproportionality analysis because of a large number of potential signals. Therefore, our study aimed to identify the strong ADR signals quickly by statistical knowledge. Methods：Based on statistical theory, signals inentisy calculated by modified ROR, PRR and IC with shrinkage estimation （a =0.5 ） was classify. The classification of signal intensity could provide the evidence for artificial screening, risk man- agement and drug evaluation. Results ：The threshold of strong signal by modified ROR, PRR and IC with shrinkage estima- tion was 12.84, 11.67, 3.83, respectively. Conclusion：The results will be applied to the new version of the national ad- verse drug reaction monitoring system.

关 键 词：不相称性测定分析 压缩估计 强度分级 

分 类 号：R181.35[医药卫生—流行病学]