军事知识服务平台
基于湖北省自发呈报系统的联合用药不良反应信号挖掘研究
Research on Signal Mining of Adverse Reactions of Combined Medication Based onthe Self-Reporting System of Hubei Province
摘要 目的 利用湖北省药品不良反应(ADR)自发呈报系统数据,挖掘联合用药相关的不良反应风险信号,为临床安全用药提供参考.方法 提取 2014 年 1 月—2022 年 12 月湖北省 ADR 数据,筛选"化学药-化学药"联合用药报告.采用描述性统计分析患者基本情况、用药及 ADR 发生特征,综合运用加法模型、乘法模型、Ω 收缩测量法、卡方检验及组合风险比法5 种方法进行药物-药物相互作用信号检测.结果 共纳入ADR 报告550 556 份,涉及药品不良反应810 814例次.2 种药物联用导致 ADR 78 054 例次(9.63%).与单药治疗相比,联合用药的不良反应谱发生改变:血液及淋巴系统疾病(5.10%vs.0.60%)与各类检查异常(4.50%vs.1.70%)的占比显著升高.信号挖掘共获得 567 个阳性"药物-药物-ADR"风险信号.涉及药物以抗微生物药(490 频次)和抗肿瘤药(145 频次)为主,其中诸多信号(如"左氧氟沙星-奥硝唑-皮疹""奥沙利铂-紫杉醇-骨髓抑制")在 Micromedex 数据库中未被记载为已知相互作用.结论 联合用药不仅可能加重单药已知的毒性(风险叠加效应,如骨髓抑制),还可能诱发单药非特征性的新风险(风险漂移效应,如特定皮疹模式).抗微生物药、抗肿瘤药及抗精神病药是联合用药时安全性风险需重点关注的领域.基于真实世界数据的多方法信号挖掘,是发现与验证这些新的联合用药风险特征、补充现有药品安全知识体系的有效手段.
Abstract:
Objective To utilize the data from the Voluntary Reporting System of adverse drug reactions in Hubei Province to explore the risk signals of adverse reactions related to combined drug use,providing a reference for safe clinical drug use.Methods Extract the ADR data of Hubei Province from January 2014 to December 2022,and screen the combined drug use reports of"chemical drugs-chemical drugs".Descriptive statistical analysis was adopted to analyze the basic situation of patients,medication use and the characteristics of ADR occurrence.Five methods,namely the additive model,the multiplicative model,the Ω contraction measurement method,the chi-square test and the combined hazard ratio method,were comprehensively used to detect drug-drug interaction signals.Results A total of 550 556 ADR reports involving 810 814 drug adverse reaction incidents were included.Among these,78 054 incidents(9.63%)were caused by the combination of two drugs.Compared with monotherapy,the adverse reaction spectrum of drug combinations changed significantly:the proportions of diseases of the blood and lymphatic system(5.10%vs.0.60%)and various abnormalities in laboratory tests(4.50%vs.1.70%)were significantly higher.Signal mining identified 567 positive"drug-drug-ADR"risk signals.The involved drugs were primarily anti-microbial agents(490 frequencies)and anti-tumor agents(145 frequencies).Many of these signals(e.g.,"levofloxacin-ornidazole-rash","oxaliplatin-paclitaxel-bone marrow suppression")were not documented as known interactions in authoritative databases(Micromedex).Conclusions The drug combinations may not only exacerbate the known toxicity of individual drugs(risk additive effect,e.g.,bone marrow suppression)but may also induce new risks that are not characteristic of the individual drugs(risk drift effect,e.g.,specific rash patterns).Anti-microbial agents,anti-tumor agents,and antipsychotics are areas requiring focused attention for safety risks during combined use.Multi-method signal mining based on real-world data is an effective means to discover and verify these new risk characteristics of drug combinations and to supplement the existing drug safety knowledge system.
刊 名 医药导报 ISTICPKU
年,卷(期): 2026, 45(10)
分类号: R951
机标分类号: R95 R51 R181.3+5
在线出版日期: 2026年11月2日
基金项目: 72274065:国家自然科学基金CDR2024R01008:国家药品监督管理局药物警戒研究与评价重点实验室开放课题2025AFD713:湖北省自然科学基金创新发展联合基金项目,