DPHARM: Community-based trials, patient-centered design and EHR AI proposed to ease enrollment crisis

At DPHARM, industry speakers described clinical-trial enrollment as a major operational and economic problem:
about 80% of trials miss enrollment timelines, while fewer than 5% of eligible adult cancer patients take part in trials, according to the report.

The proposed response is to move beyond reliance on major academic centers and conventional site-led recruitment. Speakers advocated expanding into community and rural care settings, where patients already receive routine care.

Patient input should shape protocol design and execution early—testing whether eligibility criteria, visit schedules and study demands are understandable, relevant and practical for intended participants and caregivers.

AI-enabled electronic-health-record mining was highlighted as a way to identify potentially eligible patients against complex criteria, reducing manual workload at overloaded research sites.

Eli Lilly’s Denise Sura outlined three priorities:
optimize sites for reliable randomized-patient volume, conduct more research in community-based settings, and make clinical trials a routine care option that patients can discuss with their clinicians.

The underlying goal is to embed research into normal healthcare pathways, improve representation in U.S. trial populations and reduce barriers that prevent otherwise eligible patients from participating.

The article cited Vision Lifesciences estimates that enrollment delays cost the industry roughly $50 billion annually; for a phase 3 oncology program nearing commercialization, a delay can mean an estimated $600,000 to $8 million in lost revenue per day.