Automatic Selection of Clinical Trials Based on A Semantic Web Approach

Area of Science:

• Medical Informatics
• Clinical Trial Management
• Semantic Web Technologies

Background:

• Low patient inclusion rates in clinical trials are a significant challenge.
• Factors include numerous trials, complex eligibility criteria, and physician workload.
• Automating the prescreening process is crucial for improving recruitment.

Purpose of the Study:

• To evaluate the ASTEC project's computerized recruitment support system (CRSS).
• To automate the prescreening phase of clinical trials during multidisciplinary meetings (MDM).
• To assess the system's performance using a semantic web approach.

Main Methods:

• Retrospective data collection over a 6-month period from Urology MDMs.
• Evaluation focused on 4 prostate cancer clinical trials.
• Assessed classification performance: precision, recall, and error rate.

Main Results:

• ASTEC system achieved 93% recall, 21% precision, and 37% error rate.
• Missing data was identified as the primary challenge.
• The system demonstrated scalability and usability in the MDM process.

Conclusions:

• The ASTEC CRSS shows potential for enhancing clinical trial prescreening and patient recruitment.
• Addressing missing data is key to improving system accuracy.
• The semantic web-based approach offers a scalable solution for trial automation.