Summary
Inefficiencies in screening patients for cancer clinical trials can slow down enrollment and hinder the progress of research. Presented at the ASCO Quality Care Symposium, the authors of this research evaluated a novel centralized, technology-enabled patient screening service using electronic health record-based structure variables, machine learning, and human abstraction to mitigate these issues.
Between January and May 2024, over 80,000 patients were screened across three sites. Preliminary analysis showed that this novel clinical screening approach improved the accuracy of finding potentially eligible patients by over 95% in comparison to using structured variables alone. Although only a small percentage of patients were ultimately deemed eligible (3.0% and 4.0%), likely in part due to the complex inclusion and exclusion criteria of the trials, the streamlined process led to 17% of these patients consenting to participate in the trials within 10 days of initiating the patient trial screening service at the participating sites.
Why this matters
This research tackles a critical issue in cancer clinical trials: the challenge of efficiently and accurately identifying eligible patients. By improving the accuracy and speed of patient screening, this research could help ensure that more patients who qualify for clinical trials are identified and enrolled promptly, reducing research timelines and improving the research experience across patients, providers, and research teams.