An artificial intelligence (AI) algorithm developed by researchers at the National Institutes of Health (NIH) promises to significantly accelerate the matching of potential volunteers to clinical research trials. The AI, named TrialGPT, analyzes patient data and identifies relevant clinical trials on ClinicalTrials.gov, providing a clear explanation of how the individual meets the study's enrollment criteria. Published in Nature Communications, the study highlights TrialGPT's potential to aid clinicians in navigating the complex landscape of available clinical trials, potentially leading to improved enrollment rates and faster progress in medical research.
The innovative framework behind TrialGPT leverages large language models (LLMs). The algorithm processes patient summaries containing relevant medical and demographic information. It then identifies suitable clinical trials from ClinicalTrials.gov, excluding those for which the patient is ineligible. Finally, TrialGPT explains the rationale behind the eligibility assessment, generating an annotated list of clinical trials ranked by relevance. This allows clinicians to efficiently discuss potential clinical trial opportunities with their patients.
"Machine learning and AI technology have held promise in matching patients with clinical trials, but their practical application across diverse populations still needed exploration," said Stephen Sherry, PhD, Acting Director of the NLM. "This study shows we can responsibly leverage AI technology so physicians can connect their patients to a relevant clinical trial that may be of interest to them with even more speed and efficiency."
Accuracy and Efficiency Gains
To evaluate TrialGPT's accuracy, researchers compared its performance against that of three human clinicians in assessing over 1,000 patient-criterion pairs. The results demonstrated that TrialGPT achieved a level of accuracy comparable to the clinicians. Further, a pilot user study involving clinicians reviewing anonymous patient summaries and matching them to clinical trials revealed that using TrialGPT reduced patient screening time by 40% while maintaining the same level of accuracy.
Impact on Clinical Trial Recruitment
Finding suitable clinical trials for interested participants is often a time-consuming and resource-intensive process, potentially hindering medical research. TrialGPT offers a solution to this bottleneck by enabling clinicians to more efficiently connect patients with relevant clinical trial opportunities. According to Zhiyong Lu, PhD, NLM Senior Investigator and corresponding author of the study, TrialGPT can "help clinicians connect their patients to clinical trial opportunities more efficiently and save precious time that can be better spent on harder tasks that require human expertise."
Future Directions
Recognizing the promising results, the research team has been selected for The Director's Challenge Innovation Award to further evaluate the model's performance and fairness in real-world clinical settings. This ongoing work aims to enhance clinical trial recruitment effectiveness and reduce participation barriers for underrepresented populations. The study involved collaboration with researchers from Albert Einstein College of Medicine, University of Pittsburgh, University of Illinois Urbana-Champaign, and University of Maryland, College Park.
