中国药物警戒 ›› 2023, Vol. 20 ›› Issue (5): 568-573.
DOI: 10.19803/j.1672-8629.20220447

• 安全与合理用药 • 上一篇    下一篇

药物致肝损伤数据集的比较研究

姚克宇1, 张舒琪1, 金锐2, 刘丽红1#, 朱彦1,*   

  1. 1中国中医科学院中医药信息研究所,北京, 100700;
    2中国中医科学院西苑医院药剂科,北京, 100010
  • 收稿日期:2022-07-28 出版日期:2023-05-15 发布日期:2023-05-16
  • 通讯作者: *朱彦,男,博士,副研究员,中医药信息学。E-mail: zhuyan166@126.com;#为共同通信作者。
  • 作者简介:姚克宇,女,博士,助理研究员,中医药信息学。
  • 基金资助:
    国家自然科学基金资助项目(82174534); 中国中医科学院科技创新工程(CI2021A05306、CI2021A05409); 中央级公益性科研院所基本科研业务费专项资金(ZZ13-YQ-126、ZZ13-YQ-128); 中国中医科学院基本科研业务费自主选题(ZZ150314、ZZ150316、ZZ150329JY)

Comparative study on data set of drug-induced liver injury

YAO Keyu1, ZHANG Shuqi1, JIN Rui2, LIU Lihong1#, ZHU Yan1,*   

  1. 1Institute of Information on Traditional Chinese Medicine, China Academy of Chinese Medical Sciences, Beijing 100700, China;
    2Xiyuan Hospital, China Academy of Chinese Medical Sciences, Beijing 100091
  • Received:2022-07-28 Online:2023-05-15 Published:2023-05-16

摘要: 目的 调研并对比药物性肝损伤(DILI)数据集的数据来源、DILI风险评估与药品分类方法,为DILI数据集构建及临床、科研、监管等部门提供参考。方法 选取包含药品数量在300以上且对DILI严重程度进行评估的5个数据集(DILIrank, DICIst, Liver Tox, Suzuki, Greene),比较分析其数据来源、数据量、DILI风险评估方法以及药品分类。结果 数据来源包括药品标签、文献、网络数据库、临床/注册中心病例、监管机构等多个渠道;LiverTox包含药品数量最多,DILIrank与其他数据集的重合度最高,DILIrank和LiverTox分类更详细;DILI风险评估按照肝损伤的严重程度及发生概率分类,并通过专家意见或因果关系评估量表确定DILI。结论 高质量数据集与其数据来源、数据数量、评估方法及药品分类有关;中国药物性肝损伤数据集尤其是中药DILI数据集需完善。

关键词: 药物性肝损伤, 数据集, 风险评估, DILI分类

Abstract: Objective To investigate and compare the data sources and methods for risk assessment and drug classification related to DILI so as to provide reference for the construction of DILI datasets and for clinical practice, research and regulation. Methods Five data sets (DILIrank, DICIst, Liver Tox, Suzuki, Greene) that involved more than 300 drugs and assessed the severity of human drug-induced liver injury were selected. The data sources, data volume, DILI risk assessment methods and drug classification were compared and analyzed. Results The data sources included drug labels, literature, network databases, clinical/registration center cases, regulators and other channels. LiverTox involved the largest number of drugs, DILIrank had the highest coincidence with other datasets, and the classification by DILIrank and LiverTox was more detailed. Risk assessment of DILI was classified according to the severity and odds of liver injury. DILI was determined by expert opinions or causal assessment. Conclusion A high-quality data set is related to the data source, data quantity, assessment methods and drug classification. Chinese drug-induced liver injury data sets, especially DILI data sets of Chinese medicines, need to be improved.

Key words: drug-induced liver injury, data set, risk assessment, DILI classification

中图分类号: