Beyond metrics to methods: a scoping review of transformers and large language models for detection of social drivers of health in clinical notes
Transformer and LLM performance for SDOH detection varied substantially across domains, with encoder-based models excelling at structured tasks and decoder-only models at linguistically complex tasks. Critical gaps in fairness assessment, external validation, and dataset diversity restrict generalizability and readiness for widespread clinical deployment.
Authors
- Amie J. Goodin (ORCID: https://orcid.org/0000-0002-0020-8720)
- Elham Hatef (ORCID: https://orcid.org/0000-0003-2535-8191)
- Ahmed N. Farrag (ORCID: https://orcid.org/0000-0001-9241-0127)
- Ahmed Soliman
Institutions
- Johns Hopkins University (US)
- Johns Hopkins Medicine (US)
- University of Florida Health (US)
- University of Florida (US)
- Florida College (US)
Publication Details
- Journal
- Journal of the American Medical Informatics Association
- Published
- 2025-11-11
- DOI
- https://doi.org/10.1093/jamia/ocaf201
- Primary Topic
- Food Security and Health in Diverse Populations
- Type
- article
- Field-Weighted Citation Impact
- 0.00