EUVIMEDEUVIMED entdecken
ValidierungsstudieOpen Access · CC BY-NC 4.0

SBDH-Reader: a large language model-powered method for extracting social and behavioral determinants of health from clinical notes

Zifan Gu iD, Lesi He iD, Awais Naeem, Pui Man Chan, Asim Mohamed, Hafsa Khalil, Yujia Guo, Jingwei Huang iD, Ismael Villanueva-Miranda, Ying Ding, Wenqi Shi iD, Matthew E Dupre und 5 weitere

Journal of the American Medical Informatics Association · 2025

DOI: 10.1093/jamia/ocaf124 ↗

Kurz gefasst

Validierungsstudie in der Zeitschrift Journal of the American Medical Informatics Association, 2025. Verfasst von Zifan Gu und 16 weiteren Autor:innen. Beteiligte Einrichtung: Quantitative Biomedical Research Center, Department of Health Data Science and Biostatistics, Peter O'Donnell Jr. School of Public Health, The Univer…. Hauptthemen laut MEDLINE: Social Determinants of Health, Electronic Health Records, Natural Language Processing, Information Storage and Retrieval. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: NIAID NIH HHS, Texas Health Resources Clinical Scholars Program, Health Resources Clinical Scholars Program.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objective

Social and behavioral determinants of health (SBDH) are increasingly recognized as essential for prognostication and informing targeted interventions. Clinical notes often contain details about SBDH in unstructured format. Conventional extraction methods for these data tend to be labor intensive, inaccurate, and/or unscalable. In this study, we aim to develop and validate a large language model (LLM)-powered method to extract structured SBDH data from clinical notes through prompt engineering. Materials and Methods We developed SBDH-Reader to extract 6 categories of granular SBDH data by prompting GPT-4o, including employment, housing, marital status, and substance use including alcohol, tobacco, and drug use. SBDH-Reader was developed using 7225 notes from 6382 patients in the MIMIC-III database (2001–2012) and externally validated using 971 notes from 437 patients at The University of Texas Southwestern Medical Center (UTSW; 2022–2023). We evaluated SBDH-Reader’s performance against human-annotated ground truths based on precision, recall, F1, and confusion matrix.

Results

When tested on the UTSW validation set, SBDH-Reader achieved a macro-average F1 ranging from 0.94 to 0.98 across 6 SBDH categories. For clinically relevant adverse attributes, F1 ranged from 0.96 (employment; housing) to 0.99 (tobacco use). When extracting any adverse attributes across all SBDH categories, SBDH-Reader achieved an F1 of 0.97, recall of 0.97, and precision of 0.98 in the independent validation set.

Discussion

SBDH-Reader demonstrated strong performance in extracting structured SBDH data through effective prompt engineering of a general-purpose LLM, without the need for task-specific fine-tuning. Its modular design and adaptability to diverse datasets and documentation patterns support its applicability in real-world clinical settings.

Conclusion

SBDH-Reader has the potential to serve as a scalable and effective method for collecting real-time, patient-level SBDH data to support clinical research and care.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Social Determinants of HealthElectronic Health RecordsNatural Language ProcessingInformation Storage and RetrievalHumansLarge Language Models

Stichwörter der Autor:innen

EHRbehavioral determinants of healthlarge language modelnatural language processingsocial determinants of health

Autor:innen und Institutionen

  • Quantitative Biomedical Research Center, Department of Health Data Science and Biostatistics, Peter O'Donnell Jr. School of Public Health, The University of Texas Southwestern Medical Center, Dallas, TX 75390, United St…
  • Department of Health Data Science and Biostatistics, Peter O'Donnell Jr. School of Public Health, The University of Texas Southwestern Medical Center, Dallas, TX 75390, United States.
  • School of Information, University of Texas at Austin, Austin, TX 78712, United States.
  • The University of Texas Southwestern Medical Center, Dallas, TX 75390, United States.
  • Department of Population Health Sciences, Duke University, Durham, NC 27708, United States.
  • Department of Sociology, Duke University, Durham, TX 27708, United States.

ORCID-Profile: Zifan Gu, Lesi He, Jingwei Huang, Wenqi Shi, Guanghua Xiao, Ann Marie Navar, Donghan M Yang

Förderung

  • NIAID NIH HHS
  • Texas Health Resources Clinical Scholars Program
  • Health Resources Clinical Scholars Program
  • Cancer Prevention and Research Institute
  • NIDCR NIH HHS
  • NIGMS NIH HHS
  • NIH HHS
  • Cancer Prevention and Research Institute of Texas

Zitieren

APA

Gu, Z., He, L., Naeem, A., Chan, P. M., Mohamed, A., Khalil, H., Guo, Y., Huang, J., Villanueva-Miranda, I., Ding, Y., Shi, W., Dupre, M. E., Xiao, G., Peterson, E. D., Xie, Y., Navar, A. M., & Yang, D. M. (2025). SBDH-Reader: a large language model-powered method for extracting social and behavioral determinants of health from clinical notes. Journal of the American Medical Informatics Association. https://doi.org/10.1093/jamia/ocaf124

Vancouver

Gu Z, He L, Naeem A, Chan PM, Mohamed A, Khalil H, et al. SBDH-Reader: a large language model-powered method for extracting social and behavioral determinants of health from clinical notes. Journal of the American Medical Informatics Association. 2025. doi:10.1093/jamia/ocaf124

BibTeX
@article{Gu2025,
  author = {Gu, Zifan and He, Lesi and Naeem, Awais and Chan, Pui Man and Mohamed, Asim and Khalil, Hafsa and Guo, Yujia and Huang, Jingwei and Villanueva-Miranda, Ismael and Ding, Ying and Shi, Wenqi and Dupre, Matthew E and Xiao, Guanghua and Peterson, Eric D and Xie, Yang and Navar, Ann Marie and Yang, Donghan M},
  title = {SBDH-Reader: a large language model-powered method for extracting social and behavioral determinants of health from clinical notes},
  journal = {Journal of the American Medical Informatics Association},
  year = {2025},
  doi = {10.1093/jamia/ocaf124}
}
Kostenlos starten →