Current applications of artificial intelligence in pharmacovigilance in the USA and implications

doi:10.19803/j.1672-8629.20230224

Abstract

Abstract: Objective To study the current applications of artificial intelligence in pharmacovigilance in the United States in order to provide reference for the supervision of China's postmarketing drugs. Methods Based on literature review, from the perspective of drug regulation, this article outlined the concepts and performance indicators related to artificial intelligence, the background of application, FDA's experience and challenges to pharmacovigilance from the perspective of monitoring of postmarketing drugs. Results In the past decade, FDA was extensively exploring the applications of artificial intelligence in pharmacovigilance. Currently, FDA focused on the applications of artificial intelligence in handling and evaluating individual safety reports submitted to the FDA adverse event reporting system, and major progress had been made in the retrieval and analysis of key information about ICSRs, as well as in causality assessment for ICSRs. Conclusion Based on the experience of FDA, it is recommended that we accelerate the introduction of artificial intelligence technologies, strengthen training of professionals, establish a social technology and management system for artificial intelligence and promote international exchanges and cooperation in this field.

Key words: artificial intelligence (AI), machine learning (ML), pharmacovigilance, adverse events (AE), individual case safety reports (ICSRs)

CLC Number:

R926
R994.11

WANG Tao, ZHENG Mingjie, LIU Hongliang, WANG Qing, SHEN Chuanyong. Current applications of artificial intelligence in pharmacovigilance in the USA and implications[J]. Chinese Journal of Pharmacovigilance, 2023, 20(10): 1129-1133.

References

[1] FDA. Guidance for industry-good pharmacovigilance practices and pharmacoepidemiologic assessment[EB/OL].[2023-04-03]. https://www.fda.gov/media/71546/download.
[2] FDA. Artiﬁcial Intelligence and Machine Learning (AI/ML) Software as a Medical Device Action Plan[EB/OL]. (2021-09-22) [2023-04-03]. https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-software-medical-device.
[3] ARONSON JK.Artificial intelligence in pharmacovigilance: an introduction to terms, concepts, applications, and limitations[J]. Drug Safety, 2022, 45(5): 407-418.
[4] Lexico.com.“Deﬁnition of natural language processing”. Oxford University Press.(2021)[EB/OL]. [2023-04-03]. https://www.lexico.com/deﬁnition/natural_language_processing.
[5] BALL R, DAL PAN G.“Artificial intelligence”for pharmacovigilance: ready for prime time?[J]. Drug Safety, 2022 , 45(5): 429-438.
[6] LIU Q, ZHU H, LIU C, et al.Application of machine learning in drug development and regulation: current status and future potential[J]. Clinical Pharmacology and Therapeutics(临床药理学和治疗学), 2020, 107(4): 726-729.
[7] BALL R, TOH S, NOLAN J, et al.Evaluating automated approaches to anaphylaxis case classiﬁcation using unstructured data from the FDA Sentinel System[J]. Pharmacoepidemiology and Drug Safety, 2018, 27(10): 1077-1084.
[8] GIBSON TB, NGUYEN M, BURRELL T, et al.Electronic phenotyping of health outcomes of interest using a linked claims-electronic health record database: ﬁndings from a machine learning pilot project[J]. Journal of the American Medical Informatics Association : JAMIA, 2021, 28(7): 1507-1517.
[9] PINNOW E, AMR S, BENTZEN SM, et al.Postmarket safety outcomes for new molecular entity (NME) drugs approved by the Food and Drug Administration between 2002 and 2014[J]. Clinical Pharmacology and Therapeutics, 2018, 104(2): 390-400.
[10] LESTER J, NEYARAPALLY GA, LIPOWSKI E, et al.Evaluation of FDA safety-related drug label changes in 2010[J]. Pharmacoepidemiology and Drug Safety, 2013, 22(3): 302-305.
[11] KULLDORFF M, DASHEVSKY I, AVERY TR, et al.Drug safety data mining with a tree-based scan statistic[J]. Pharmacoepidemiology and Drug Safety, 2013, 22(5): 517-523.
[12] BATE A, HORNBUCKLE K, JUHAERI J, et al.Hypothesis-free signal detection in healthcare databases: ﬁnding its value for pharmacovigilance[J]. Therapeutic Advances in Drug Safety, 2019, 10: 1-9.
[13] HUYSENTRUYT K, KJOERSVIK O, DOBRACKI P, et al.Validating intelligent automation systems in pharmacovigilance: insights from good manufacturing practices[J]. Drug Safety, 2021, 44(3): 261-272.
[14] ZHENG NN, LIU ZY, REN PJ, et al.Hybrid-augmented intelligence: collaboration and cognition[J]. Frontiers of Information Technology & Electronic Engineering, 2017, 18(2): 153-179.
[15] ICMRA.Horizon Scanning Assessment Report - Artificial Intelligence [EB/OL]. (2021-08-06)[2023-04-10]. https://www.icmra.info/drupal/sites/default/files/2021-08/horizon_scanning_report_artificial_intelligence.pdf.
[16] The State Council. Development Plan for the New Generation of Artificial Intelligence[EB/OL].(2017-07-20)[2023-04-03]. http://www.gov.cn/zhengce/content/2017-07/20/content_5211996.htm.
[17] National Center for Adverse Drug Reaction Monitoring. Annual Reports on Adverse Drug Reaction Monitoring (2022) [EB/OL]. (2023-03-24)[2023-04-03]. https://www.cdr-adr.org.cn/drug_1/aqjs_1/drug_aqjs_sjbg/202303/t20230324_50019.html.
[18] SITTIG DF, SINGH H.A new sociotechnical model for studying health information technology in complex adaptive healthcare systems[J]. Quality & Safety in Health Care, 2010, 19(Suppl3): i68-i74.
[19] CHEN F.Latest development of international pharmacovigilance and its implications[J]. Chinese Journal of Pharmacovigilance(中国药物警戒), 2020, 17(12): 867-870.