中国食品药品监管2024,Issue(1) :58-75,中插19-中插21.DOI:10.3969/j.issn.1673-5390.2024.01.008

人工智能/机器学习在临床试验中的运用及监管挑战浅析——基于FDA讨论文件、EMA观点文件与利益攸关方视角

Application and Regulatory Challenges of Artificial Intelligence/Machine Learning in Clinical Trials:An Analysis Based on FDA Discussion Papers,EMA Reflection Papers,and Stakeholders'Comments

姚立新 吴天贺
中国食品药品监管2024,Issue(1) :58-75,中插19-中插21.DOI:10.3969/j.issn.1673-5390.2024.01.008

人工智能/机器学习在临床试验中的运用及监管挑战浅析——基于FDA讨论文件、EMA观点文件与利益攸关方视角

Application and Regulatory Challenges of Artificial Intelligence/Machine Learning in Clinical Trials:An Analysis Based on FDA Discussion Papers,EMA Reflection Papers,and Stakeholders'Comments

姚立新 1吴天贺2
扫码查看

作者信息

  • 1. 国家药品监督管理局南方医药经济研究所
  • 2. 埃默里大学
  • 折叠

摘要

药品开发,必须符合相关法规要求,需要以监管机构制定的行业指南为指针,参考监管机构制定的相关用例.新药获批必须基于临床试验生成的证据.随着生物制药行业的不断发展,临床试验复杂程度增加.单纯依靠增加投入,难以改善临床试验效率,解决管线产出率低、成本高企的问题.本文通过解析FDA讨论文件、EMA观点文件中涉及的临床试验中人工智能/机器学习(AI/ML)运用内容,以及介绍利益攸关方对FDA讨论文件的反馈意见,初步探讨AI/ML在临床试验中的运用以及监管挑战.

Abstract

Drug development must comply with relevant regulatory requirements,guided by industry guidelines developed by regulatory agencies,with reference to relevant use cases.The approval of new drugs must be based on evidence generated from clinical trials.With the continuous development of biopharmaceutical industry,the complexity of clinical trials has increased.Merely increasing investment is insufficient for improving clinical trial efficiency and addressing issues of low pipeline productivity and soaring costs.This paper explores the application and regulatory challenges of artificial intelligence/machine learning(AI/ML)in clinical trials by analyzing content from FDA discussion papers,EMA reflection papers,and stakeholders'comments.

关键词

人工智能/机器学习/临床试验/方案设计/试验效率/管线产出率/透明度

Key words

artificial intelligence/machine learning/clinical trials/protocol design/clinical trial efficiency/pipeline yield/transparency

引用本文复制引用

出版年

2024
中国食品药品监管
中国医药报社

中国食品药品监管

影响因子:0.099
ISSN:1673-5390
参考文献量26
段落导航相关论文