Recruiting patients for clinical trials remains challenging in oncology research. New data from Cleveland Clinic show that AI-driven screening could transform this process.
At the 2025 American Society of Hematology Annual Meeting, researchers reported that Dyania Health’s Synapsis™ AI platform identified seven times more eligible patients for a polycythemia vera trial than traditional methods, achieving 100% positive predictive value after research staff verification. This medically trained, large language model-based system operates within electronic medical records.
Traditional recruitment involves manual chart reviews by clinicians and research nurses, often taking over 30 minutes per record. This is especially difficult for rare diseases, where eligible patients are hard to find and often treated outside major centers, causing slow enrollment and delayed therapeutic development.
Led by Aaron Gerds, MD, MS, at Cleveland Clinic’s Cancer Institute, the study assessed the platform’s effectiveness. The AI mirrored traditional prescreening by filtering large populations and narrowing to patients meeting detailed criteria using structured data and free-text clinical notes.
From 4.7 million electronic medical records in the Cleveland Clinic database, 28,200 oncology patients from the past three years were identified. The platform found 904 patients with polycythemia vera and completed eligibility assessments within one week, identifying 22 eligible patients. Human review confirmed 100% accuracy.
In contrast, traditional prescreening identified nine patients, enrolled four, and treated three in 12 months—about one enrolled patient every four months. The AI approach increased eligible patient identification seven-fold in much less time, reducing workload and accelerating timelines.
The study shows medically trained large language models can improve clinical trial recruitment by automating lengthy chart reviews. These efficiencies are vital in rare diseases, where manual recruitment often falls short. AI can speed therapy development by expanding patient identification while maintaining accuracy.
Cleveland Clinic and Dyania Health announced a collaboration to integrate Synapsis AI across the health system’s clinical research, aiming to expand AI-driven prescreening across therapeutic areas, improving trial efficiency and patient access.
Dr. Gerds emphasized AI as an essential tool to enhance clinical research and patient care. By automating routine tasks, AI allows clinicians to focus on strategy and patient interaction, potentially accelerating the development of life-saving therapies.
Leave a Reply