基于医院信息系统数据的中枢神经系统不良反应自动监测模块构建优化与实践

Establishment,optimization and practice of an automatic central nervoussystem adverse reactions monitoring module based on hospital informationsystem data

目的构建基于临床药品不良事件主动监测与智能评估警示系统-Ⅱ(ADE-ASAS-Ⅱ)的中枢神经系统不良反应(CNS-ADR)模块,开展亚胺培南西司他丁相关CNS-ADR大样本真实世界主动监测评价研究。方法以文献、自发报告、电子病历中CNS-ADR相关描述词作为初始词集,利用文本识别技术构建并优化CNS-ADR自动监测模块条件设置。回顾性监测2017—2021年使用注射用亚胺培南西司他丁钠的住院患者,对发生CNS-ADR的阳性病例的人口学特征、中枢神经系统症状、住院科室进行统计描述。结果基于1185例人工监测结果进行反复测试优化后,最终确定的模块最佳设置包括62个关键词集,阳性预测值(PPV)为13.63%,召回率为100%。利用该模块拓展监测8222例用药人群,报警2366例,PPV为11.88%,关联性阳性281例,发生率为3.42%,其中60岁以上患者占比50.17%,CNS-ADR表现主要为癫痫发作、头痛、躁狂、谵妄等,发生CNS-ADR主要分布在血液科、呼吸科、肿瘤内科等。结论基于ADE-ASAS-II建立的CNS-ADR自动监测模块,为开展CNS-ADR的真实世界研究提供了快捷可靠的文本数据挖掘支持。

Objective To construct a module for drug-induced central nervoussystem adverse reactions(CNS-ADR)within the Clinical Adverse Drug Event Active Monitoring and Intelligent Assessment Alert System-II(ADE-ASAS-II),and to conduct alarge-scale,real-world active monitoring and evaluation of CNS-ADR specifically related toimipenem/cilastatin.Methods Based on literature review,spontaneous report evaluation,andinitial word set of CNS-ADR related descriptions in electronic medical records,text recognitiontechnology was used to construct and optimize the condition settings of the CNS-ADR automaticmonitoring module.Hospitalized patients using imipenem/cilastatin were retrospectivelymonitored from 2017 to 2021,and the positive patients which had CNS-ADR were statisticallydescribed in terms of the demographic characteristics,CNS symptoms,and hospital departments.Results Based on a repeated testing optimization using 1185 manually monitored results,thebest setting for the determined module includes 62 sets of keywords,with a positive predictivevalue(PPV)of 13.63%and a recall rate of 100%.Expanding the monitoring to 8222 medicationusers using this module,281 cases of positive causality were identified,with an incidencerate of 3.42%.Among them,patients over 60 years old accounted for 50.17%,and the mainmanifestations of CNS-ADR were epileptic seizures,headaches,mania,and delirium.ConclusionThe CNS-ADR automatic monitoring module established based on ADE-ASAS-II provides fastand reliable text data mining support for conducting real-world research on CNS-ADR.