Sneha S. Jain, Ashish Sarraju, Nigam H. Shah, Kevin A. Schulman, Euan A. Ashley, Robert A. Harrington, Kenneth W. Mahaffey
{"title":"The Coming AI Revolution in Clinical Trials","authors":"Sneha S. Jain, Ashish Sarraju, Nigam H. Shah, Kevin A. Schulman, Euan A. Ashley, Robert A. Harrington, Kenneth W. Mahaffey","doi":"10.1016/j.jacc.2024.10.093","DOIUrl":null,"url":null,"abstract":"<h2>Section snippets</h2><section><section><h2>Digital Strategy and Data Infrastructure</h2>Current systems for data storage and analysis are siloed, limiting the potential for AI to function efficiently across data networks. Without interoperable data systems, even the most advanced AI tools will struggle to deliver real impact. We need a unified data strategy—how we manage, integrate, and safeguard data across platforms—focused on a patient-centered data ecosystem that facilitates information flow across EHRs, health system applications, personal records, and clinical trial data</section></section><section><section><h2>Business Model Innovation</h2>Current clinical trial business models are designed around trial sites, not participants. They involve processes that are labor-intensive, duplicative, and burdened by high fixed costs for recruitment, site management, and data collection. Data are often collected redundantly, despite also being obtained for routine clinical care, with gaps in collection due to episodic follow-up instead of continuous monitoring. AI has the potential to streamline certain processes, shorten timelines, and lower</section></section><section><section><h2>Regulatory Modernization and AI Evaluation</h2>Regulatory bodies are crucial in ensuring the safety and efficacy of AI-augmented clinical trials, yet the current regulatory framework was not designed to accommodate the rapid advancements in AI technology or the unique risks of AI use in RCTs. Rather than relying on traditional linear approval pathways, regulators should embrace iterative approaches that incorporate regular feedback. Participants in RCTs that use AI tools for operations are a potentially vulnerable population in which there</section></section><section><section><h2>Funding Support and Author Disclosures</h2>Dr Jain has received consulting fees from Bristol Myers Squibb, ARTIS Ventures, and Broadview Ventures outside of the submitted work. Dr Sarraju holds stock in Medeloop. Dr Ashley is a founder of Personalis, DeepCell, and Svexa; is a founding advisor of Nuevocor; is a nonexecutive director at AstraZeneca; and has served as an advisor for SequenceBio, Novartis, Medical Excellence Capital, Foresite Capital, and Third Rock Ventures. Dr Harrington has served on the Board of Directors for the</section></section>","PeriodicalId":17187,"journal":{"name":"Journal of the American College of Cardiology","volume":"35 1 1","pages":""},"PeriodicalIF":21.7000,"publicationDate":"2024-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of the American College of Cardiology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1016/j.jacc.2024.10.093","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"CARDIAC & CARDIOVASCULAR SYSTEMS","Score":null,"Total":0}
引用次数: 0
Abstract
Section snippets
Digital Strategy and Data Infrastructure
Current systems for data storage and analysis are siloed, limiting the potential for AI to function efficiently across data networks. Without interoperable data systems, even the most advanced AI tools will struggle to deliver real impact. We need a unified data strategy—how we manage, integrate, and safeguard data across platforms—focused on a patient-centered data ecosystem that facilitates information flow across EHRs, health system applications, personal records, and clinical trial data
Business Model Innovation
Current clinical trial business models are designed around trial sites, not participants. They involve processes that are labor-intensive, duplicative, and burdened by high fixed costs for recruitment, site management, and data collection. Data are often collected redundantly, despite also being obtained for routine clinical care, with gaps in collection due to episodic follow-up instead of continuous monitoring. AI has the potential to streamline certain processes, shorten timelines, and lower
Regulatory Modernization and AI Evaluation
Regulatory bodies are crucial in ensuring the safety and efficacy of AI-augmented clinical trials, yet the current regulatory framework was not designed to accommodate the rapid advancements in AI technology or the unique risks of AI use in RCTs. Rather than relying on traditional linear approval pathways, regulators should embrace iterative approaches that incorporate regular feedback. Participants in RCTs that use AI tools for operations are a potentially vulnerable population in which there
Funding Support and Author Disclosures
Dr Jain has received consulting fees from Bristol Myers Squibb, ARTIS Ventures, and Broadview Ventures outside of the submitted work. Dr Sarraju holds stock in Medeloop. Dr Ashley is a founder of Personalis, DeepCell, and Svexa; is a founding advisor of Nuevocor; is a nonexecutive director at AstraZeneca; and has served as an advisor for SequenceBio, Novartis, Medical Excellence Capital, Foresite Capital, and Third Rock Ventures. Dr Harrington has served on the Board of Directors for the
期刊介绍:
The Journal of the American College of Cardiology (JACC) publishes peer-reviewed articles highlighting all aspects of cardiovascular disease, including original clinical studies, experimental investigations with clear clinical relevance, state-of-the-art papers and viewpoints.
Content Profile:
-Original Investigations
-JACC State-of-the-Art Reviews
-JACC Review Topics of the Week
-Guidelines & Clinical Documents
-JACC Guideline Comparisons
-JACC Scientific Expert Panels
-Cardiovascular Medicine & Society
-Editorial Comments (accompanying every Original Investigation)
-Research Letters
-Fellows-in-Training/Early Career Professional Pages
-Editor’s Pages from the Editor-in-Chief or other invited thought leaders