药物警戒中不相称性测定理论应用问题分析  

Analysis of Application of Disproportional Measures in Pharmacovigilance

作　　者：孙亚林[1] 李永昌[2] 杜文民[3] 叶小飞[1] 贺佳[1] 

机构地区：[1]第二军医大学卫勤系卫生统计学教研室,上海200433 [2]南京军区总医院医务部 [3]上海市药品不良反应监测中心

出　　处：《药物流行病学杂志》2009年第3期147-150,共4页Chinese Journal of Pharmacoepidemiology

基　　金：国家自然科学基金(项目号:30872186);国家科技部重大专项(项目号:2008ZX10002-018;2008ZX09312-007)

摘　　要：目的:分析药品不良反应(ADR)信号检测中不相称性测定理论在实际应用中所存在的问题,阐明不相称性测定理论的使用条件,分析引起问题的原因。方法:通过数学计算、计算公式推导,配合实例进行分析和阐述。结果:不相称性测定理论在ADR信号检测中存在着测量值失衡、联合用药分析能力弱、产生假阳性和假阴性信号等问题,产生问题的根源在于自发呈报系统的数据特征,往往不符合不相称性测定理论的假设及相对数引起的分析偏差。结论:使用不相称性测定理论进行不良反应信号分析,需要明确方法的使用条件以及不良反应检测数据的特征;同时积极开发基于绝对数的新的信号检测方法。Objective ： To analyze the application of disproportional measures in drug-ADR signal detection, clarify its conditions and discover the reasons which caused problems. Method： Mathematic formulation deduction and examples were used to analyze it. Result： Its problems in applying disproportional measures in ADR signal detection included the unbalanced value of index, low efficiency in drug＆drug-ADR signal detection, false-negative and false-positive signal, etc. Their main reasons lied in that the characteristics of ADR monitoring data did not match the application conditions of disproportional measures, which led to biases in the calculation of the relative number index. Conclusion： When applying the disproportional measures in ADR signal detection, people should be first aware of their application conditions and data characteristics. Meanwhile algorithms based on absolute number index should be developed to make up these deficits.

关 键 词：药物警戒 药物不良反应监测 不相称测定 信号检测 

分 类 号：R181.35[医药卫生—流行病学]