Recommending Clinical Trials for Online Patient Cases using Artificial Intelligence.

Area of Science:

• Medical Informatics
• Clinical Trial Management
• Artificial Intelligence in Healthcare

Background:

• Clinical trial recruitment faces significant hurdles, including limited patient awareness and complex eligibility criteria.
• The proliferation of online platforms (social media, health communities, PubMed) offers novel, untapped sources for identifying potential clinical trial participants.
• Traditional recruitment methods often fail to leverage these diverse, non-traditional data sources effectively.

Purpose of the Study:

• To evaluate the efficacy of TrialGPT, a large language model-based framework, in matching online patient cases to suitable clinical trials.
• To compare the performance of TrialGPT against conventional keyword-based search methods for clinical trial recruitment.
• To assess the potential of utilizing non-traditional patient case sources for clinical trial participant identification.

Main Methods:

• Collected 50 online patient cases from medical case reports and social media platforms.
• Employed the TrialGPT framework, utilizing a large language model, to match these patient cases with relevant clinical trials.
• Compared the matching performance of TrialGPT against traditional keyword-based search strategies.

Main Results:

• TrialGPT demonstrated a 46% superior performance compared to traditional keyword-based searches in matching patients to clinical trials.
• On average, patients identified through TrialGPT were eligible for 7 out of the top 10 recommended clinical trials.
• Positive feedback was received from both case authors and clinical trial organizers upon outreach, indicating feasibility and acceptance.

Conclusions:

• TrialGPT effectively leverages non-traditional online patient data to enhance clinical trial recruitment.
• The AI-driven approach significantly improves patient eligibility and access to clinical trials for specialized care.
• This framework presents a promising strategy for expanding patient access to innovative treatments through novel recruitment pathways.