上市后不良反应评价中似然比检验方法的比较

doi:10.19803/j.1672-8629.20230132

中国药物警戒 ›› 2023, Vol. 20 ›› Issue (6): 655-660.
DOI: 10.19803/j.1672-8629.20230132

• 上市后药品不良反应监测定量统计学方法研究专栏 • 上一篇下一篇

上市后不良反应评价中似然比检验方法的比较

潘喆敏¹, 徐圣², 郑轶³, 陈晨鑫³, 刘永梅⁴, 张朋朋⁴, 叶小飞^3#, 贺佳^1,3,*

¹同济大学医学院,上海 200092;
²百济神州（上海）生物科技有限公司,上海 200040;
³海军军医大学卫生统计学教研室,上海 200433;
⁴联勤保障部队天津康复疗养中心,天津 300191

收稿日期:2022-03-09 出版日期:2023-06-15 发布日期:2023-06-15
通讯作者: ^* 贺佳,女,博士,教授,流行病学及新药评价。E-mail: hejia63@yeah.net。^#为共同通信作者。
作者简介:潘喆敏,女,在读博士,临床试验方法学及流行病学。
基金资助:
国家自然科学基金资助项目（82073671）; 中国药学会临床评价研究专业委员会研究课题（CPA-CDCER -2021-001）

Comparison of likelihood ratio tests in post-marketing safety surveillance

PAN Zhemin¹, XU Sheng², ZHENG Yi³, CHEN Chenxin³, LIU Yongmei⁴, ZHANG Pengpeng⁴, YE Xiaofei^3#, HE Jia^1,3,*

¹Tongji University School of Medicine, Shanghai 200092, China;
²BeiGene (Shanghai) Co., Ltd., Shanghai 200040, China;
³Department of Health Statistics, Naval Medical University, Shanghai 200433, China;
⁴Tianjin Rehabilitation Center of Joint Logistic Support Force, Tianjin 300191, China

Received:2022-03-09 Online:2023-06-15 Published:2023-06-15

摘要/Abstract

摘要： 目的对上市后安全性监测中各类似然比检验方法（likelihood ratio test, LRT）进行概述及比较。方法通过文献检索对上市后不良反应评价中LRT方法的原理进行综述并对进一步的探索方向进行展望。结果现有研究表明与报告比值比、信息成分方法相比LRT方法能更好控制I型错误发生率,并且有多种方法适用于不同的应用场景。结论 LRT方法是上市后不良反应评价可行的方法之一,尤其适用于需要严格控制伪发现率,或者需要权衡伪发现率/I型错误发生率与灵敏度/检验功效的情况,但仍有许多问题需进一步探索。

关键词: 似然比检验, 上市后不良反应评价, I型错误, 报告比值比法, 信息成分法

Abstract: Objective To survey and compare likelihood ratio test (LRT) approaches to post-marketing safety surveillance. Methods The fundamentals of likelihood ratio tests as well as the advantages and disadvantages of each method were summarized. Meanwhile, priorities of research in the future were predicted. Results Existing evidence suggested that LRT methods could better control type I errors than the reporting odds ratio and information component methods. Besides, LRT approaches employed a class of precise statistical tools which enabled objective evaluation under various scenarios. Conclusion LRT methods are feasible for post-marketing safety surveillance, especially when rigorous control of the false discovery rate is required or when a trade-off between the false discovery rate/type I error incidence and sensitivity/power is needed. However, there are still some problems that remain unsettled.

Key words: likelihood ratio test, post-marketing safety surveillance, type I error, reporting odds ratio, information component methods

中图分类号:

R911
R994

潘喆敏, 徐圣, 郑轶, 陈晨鑫, 刘永梅, 张朋朋, 叶小飞, 贺佳. 上市后不良反应评价中似然比检验方法的比较[J]. 中国药物警戒, 2023, 20(6): 655-660.

PAN Zhemin, XU Sheng, ZHENG Yi, CHEN Chenxin, LIU Yongmei, ZHANG Pengpeng, YE Xiaofei, HE Jia. Comparison of likelihood ratio tests in post-marketing safety surveillance[J]. Chinese Journal of Pharmacovigilance, 2023, 20(6): 655-660.

参考文献

[1] HUANG L, GUO T, ZALKIKAR J N, et al.A review of statistical methods for safety surveillance[J]. Therapeutic Innovation and Regulatory Science, 2014, 48(1): 98-108.
[2] HUANG L, ZALKIKAR J, TIWARI R C.A likelihood ratio test based method for signal detection with application to FDA’s drug safety data[J]. Journal of the American Statistical Association, 2011, 106(496): 1230-1241.
[3] DING Y, MARKATOU M, BALL R.An evaluation of statistical approaches to postmarketing surveillance[J]. Statistics in Medicine, 2020, 39(7): 845-874.
[4] HUANG L, ZALKIKAR J, TIWARI RC.Likelihood ratio test-based method for signal detection in drug classes using FDA’s AERS database[J]. Journal of Biopharmaceutical Statistics, 2013, 23(1): 178-200.
[5] SINGH SN.A note on inflated poisson distribution[J]. J Ind Stat Assoc, 1963, 1: 140-144.
[6] HUANG L, ZHENG D, ZALKIKAR J, et al.Zero-inflated poisson model based likelihood ratio test for drug safety signal detection[J]. Statistical Methods in Medical Research, 2017, 26(1): 471-488.
[7] LAMBERT D.Zero-inflated poisson regression, with an application to defects in manufacturing[J]. Technometrics, 1992, 34(1): 1.
[8] ZHAO Y, YI M, TIWARI RC.Extended likelihood ratio test-based methods for signal detection in a drug class with application to FDA’s adverse event reporting system database[J]. Statistical Methods in Medical Research, 2018, 27(3): 876-890.
[9] HUANG L, ZALKIKAR J, TIWARI R.Likelihood ratio based tests for longitudinal drug safety data[J]. Statistics in Medicine, 2014, 33(14): 2408-2424.
[10] HUANG L, ZALKIKAR J, TIWARI R.Likelihood-ratio-test methods for drug safety signal detection from multiple clinical datasets[J]. Computational and Mathematical Methods in Medicine, 2019(21):1-11.
[11] TIWARI R, ZALKIKAR J, HUANG L.Signal Detection for Medical Scientists Likelihood Ratio Testbased Methodology[M]. 1st ed. Boca Raton: CRC Press, 2021.
[12] NAM K, HENDERSON NC, ROHAN P, et al.Penalized logistic regression likelihood ratio test analysis to detect signals of adverse events from interactions in postmarket safety surveillance[J]. Statistics in Biopharmaceutical Research, 2021, 13(4): 405-414.
[13] NAM K, HENDERSON NC, ROHAN P, et al.Logistic regression likelihood ratio test analysis for detecting signals of adverse events in post-market safety surveillance[J]. Journal of Biopharmaceutical Statistics, 2017, 27(6): 990-1008.
[14] XU Z, XU J, YAO Z, et al.Evaluating medical device adverse event signals using a likelihood ratio test method[J]. Journal of Biopharmaceutical Statistics, 2021, 31(1): 37-46.
[15] TONG J, HUANG J, DU J, et al. Identification of rare adverse events with year-varying reporting rates for FLU4 vaccine in VAERS[J/OL]. AMIA Annu Symp Proc, (2018-12-05)[2023-03-01]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6371315/.
[16] FDA. LRT Signal Analysis for a Drug[EB/OL].[2023-03-01]. https://openfda.shinyapps.io/LRTest/.
[17] CHAKRABORTY S, MARKATOU M, LIU A. pvLRT: Likelihood Ratio Test-Based Approaches to Pharmacovigilance[EB/OL]. (2023-03-06)[2023-03-07]. https://CRAN.R-project.org/package=pvLRT.

上市后不良反应评价中似然比检验方法的比较

Comparison of likelihood ratio tests in post-marketing safety surveillance

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 3

编辑推荐

Metrics

[1]	王珊, 谢波, 刘慧敏, 李志浩. 基于FAERS数据库对塞来昔布不良事件信号的分析[J]. 中国药物警戒, 2024, 21(2): 190-194.
[2]	刘砚韬, 张伶俐, 林芸竹, 陈力, 王丽. 基于美国疫苗不良事件报告系统的肺炎疫苗不良反应信号挖掘与研究[J]. 中国药物警戒, 2023, 20(1): 96-100.
[3]	曹鑫, 孙玉泉, 黄贵民, 冉仲阳, 崔向丽. 他克莫司与伏立康唑联合使用不良反应信号挖掘与分析[J]. 中国药物警戒, 2021, 18(5): 483-488.