How AI Could Accelerate Clinical Trial Design and Patient Recruitment

Artificial intelligence (AI) could significantly reduce the time required for several traditionally labor-intensive steps in clinical trial planning, according to Eron Kelly, CEO of ConcertAI, and Shaalan Beg, MD, chief medical officer, oncology, ConcertAI, who discussed the technology’s potential in a recent video interview with BioPharm International.

Kelly explained that AI reasoning models can rapidly digitize inclusion and exclusion criteria, compare a proposed trial with similar studies listed on ClinicalTrials.gov, and identify relevant trial attributes. Tasks that previously required weeks of manual review could potentially be completed in minutes, he said.

How can AI improve clinical trial design and patient recruitment?

AI can also help sponsors assess whether an appropriate patient population exists within the current standard of care, particularly for second-line therapies. Kelly said these analyses can help sponsors identify potential clinical trial sites based on historical patient populations and ultimately accelerate enrollment and drug development.

The potential benefits extend to patients as well as sponsors, Kelly said, because faster trial execution could help patients access investigational treatments while enabling sponsors to bring therapies to market more quickly.

For Beg, however, trustworthiness is essential when AI is used to abstract and interpret clinical data. He described a three-step validation process designed to ensure that information extracted from medical records is both accurate and clinically meaningful.

The process begins with abstraction accuracy, determining whether information was correctly extracted from the patient record. The second step, record coherence, examines whether the information makes sense within the patient’s clinical timeline. Finally, cohort calibration evaluates whether the resulting patient cohort is biologically and clinically relevant to the questions being asked by practitioners, researchers, and R&D teams.

Beg emphasized that each stage must be satisfied before the information moves forward, underscoring the importance of validation as AI becomes more deeply integrated into clinical research.

Check out part one of this four-part interview here. Watch part two here.