Abstract
Abstract Introduction: Identification of human subjects who qualify for a cancer trial can be challenging, particularly given narrow time frames for enrollment and highly specific criteria for eligibility. We sought to identify patients on a timely basis by matching trial criteria to data present in both the Pathology and clinical information systems. In a separate abstract submitted to the AACR meeting, we described the enrollment outcomes of our effort for two clinical trials. In this companion abstract, we describe the query methods that we employed. Methods: Method #1 involved querying structured data fields in the EMR, using dashboards or SQL queries of relational databases. Method #2 utilized natural language processing (NLP) queries of narrative text in pathology reports. We assessed the useability of each query approach, on a near-real-time basis, against eligibility criteria for open clinical trials at our institution. Results: We examined the first 25 open clinical trials for treatment of solid tumors listed publicly on our institutional website. For each trial, eligibility and query criteria were compiled by two reviewers (DC,YZ); discrepancies between reviewers were resolved by a third reviewer (NY). We found that of the 33 trials examined: 26 trials (79%) could be successfully analyzed by NLP in the pathology report. 2 trials (6%) required SQL query of discrete data fields in clinical databases. 2 trials (6%) required both methods to be used together. 3 trials (9%) required data elements outside the scope of either method. The data fields provided by NLP were: tumor morphology; tumor stage; specimen margin; and tumor receptor status. The data fields provided by SQL were: ICD 10 codes, CPT codes, Chemotherapy, and quantitative lab results. The data fields not accessible by either approach were: tumor resectability and risk calculation. Conclusion: There is no single method that is best for all trials, and availability of a broad set of tools is needed to craft a fitting query for each trial. However, 26 of 33 trials (79%) required NLP to parse eligibility criteria from narrative pathology reports. Cancer clinical trials enrollment may be substantially enhanced by NLP extraction of structured data from pathology reports as part of a near-real-time workflow for identifying eligible Human Subjects. Citation Format: Nalan Yurtsever, Yonah Ziemba, Daniel A. King, Dylan J. Cooper, Sergio Garza, Cheryl B. Schleicher, Sharon S. Fox, James M. Crawford, Joseph Herman. The need for structured pathology data in matching human subjects to cancer clinical trial criteria: A methodological approach [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3167.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.