Classifying Unstructured Text in Electronic Health Records for Mental Health Prediction Models: Large Language Model Evaluation Study
JMIR Medical Informatics2024Vol. 13, pp. e65454–e65454
Citations Over TimeTop 10% of 2024 papers
Nicholas C. Cardamone, Mark Olfson, Timothy Schmutte, Lyle Ungar, Tony Liu, Sara Cullen, Nathaniel J. Williams, Steven C. Marcus
Abstract
The LLM displayed high agreement with clinical experts when classifying EHR terms into certain mental health or physical health term categories. However, agreement with clinical experts varied considerably within both sets of mental and physical health term categories. Importantly, the use of LLMs presents an alternative to manual human coding, presenting great potential to create interpretable features for prediction models.
Related Papers
- → Changes in Electronic Health Record Use Time and Documentation over the Course of a Decade(2019)23 cited
- → Shared Decision-Making Tools Implemented in the Electronic Health Record: Scoping Review (Preprint)(2024)1 cited
- → Electronic Health Record(2011)18 cited
- → A Novel Electronic Record System for Documentation and Efficient Workflow for Community Health Workers: Development and Usability Study (Preprint)(2023)
- → Patients’ Use of Electronic Health Records Facilitates Patient-Centered Communication: Findings From the 2017 Health Information National Trends Survey (Preprint)(2023)