A State-of-the-Art review published in JACC outlines the range of potential opportunities for the incorporation of artificial intelligence (AI) in the design and implementation of cardiovascular clinical trials, as well as the limitations and potential pitfalls.

The review, authored by Jonathan Cunningham, MD, FACC, Harlan M. Krumholz, MD, SM, FACC, et al., stems from a March 2024 special focus meeting of the Heart Failure Collaboratory evaluating the use of AI in all aspects of the clinical trial process: design, recruitment, informed consent, ascertaining physiological and clinical event outcomes, interpreting imaging, follow-up and endpoint adjudication, and analyzing and disseminating results. Sections on each discuss how AI may play a role.

The impetus behind integrating AI into clinical trials is the expense, length and insufficient diversity in current pivotal clinical trials, which are the gold standard for establishing the efficacy and safety of cardiovascular therapies. By streamlining and automating processes, trawling through enormous data sets to find additional candidates, screen potential participants and more, AI may have the potential to accelerate clinical trials throughout their life cycle.

Regarding publication and dissemination of results, they note that generative AI is already being used to speed up the preparation of academic manuscripts. Policies, such as those established by JACC, are needed that require disclosure of the use of AI in writing the manuscript and requiring the authors to take full responsibility for the final manuscript. "The gradual integration of AI with careful human oversight remains the most prudent path to improving efficiency while maintaining safety and refining best practices," they write, to guard against the potential for "biased, fraudulent, or simply incorrect interpretation of the trial results." Medical journals also have a role in evaluating AI methodology.

In looking at regulatory priorities and guidance on AI in clinical trials, the documents notes that the U.S. Food and Drug Administration has "accelerated efforts to create an agile regulatory ecosystem that can facilitate innovation and adoption while ensuring public safety and guarding against potential risks."
Among the risks of AI in cardiovascular trials that must be managed carefully are potentially less accurate data collection; susceptibility to data set shift; encoding or amplifying of biases against women or underserved groups through AI tools that have learned from completed trials, breach of confidential participant medical data; and the potential of clinical trial researchers not learning core competencies, because of relying on automated AI tools, creating vulnerabilities in the absence of such tools.

"Given the high-stakes role of randomized trials in medical decision making, AI methods must be integrated cautiously and thoughtfully to protect the validity of trial results," write the authors.