Automated detection of stigmatizing language in Electronic Health Records (EHRs) using a multi-stage transfer learning approach
Journal of the American Medical Informatics Association · 2026
DOI: 10.1093/jamia/ocaf193 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of the American Medical Informatics Association, 2026. Verfasst von Liyang Xue und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Library and Information Science, School of Communication & Information, Rutgers University, New Brunswick, NJ, 08901, United States. Hauptthemen laut MEDLINE: Electronic Health Records, Machine Learning, Language, Natural Language Processing. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 53 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.
Abstract (Originalsprache)
Objective
Stigmatizing language (SL) in Electronic Health Records (EHRs) can perpetuate biases and negatively impact patient care. This study introduces a novel method for automatically detecting such language to improve healthcare documentation practices. Materials and Methods We developed a multi-stage transfer learning framework integrating semantic, syntactic, and task adaptation using three datasets: hate speech, clinical phenotypes, and stigmatizing language. Experiments were conducted on stigmatizing language dataset which consists of 4,129 de-identified EHR notes (72.7% stigmatizing, 27.3% non-stigmatizing), split 80/20 for training and testing. Longformer, BERT, and ClinicalBERT models were evaluated, and model performance was assessed on 35 randomized subsets of the test set (each comprising 70% of test data). The Wilcoxon-Mann-Whitney test was used to evaluate statistical significance, with Bonferroni correction applied to control for multiple hypothesis testing. Baseline models included zero-shot and few-shot GPT-4o, Support Vector Machine, Random Forest, Logistic Regression, and Multinomial Naive Bayes.
Results
The proposed framework achieved the highest accuracy, with fully adapted Longformer reaching 89.83%. Performance improvements remained statistically significant after Bonferroni correction compared to all baselines (p < .05). The framework demonstrated robust gains across different stigmatizing language types.
Discussion
This study underscores the value of domain-adaptive NLP for detecting stigmatizing language in EHRs. The multi-stage transfer learning framework effectively captures subtle biases often missed by conventional models, enabling more objective and respectful clinical documentation.
Conclusion
This framework offers a statistically validated, high-performing framework for detecting stigmatizing language in EHRs, supporting responsible AI and promoting equity in clinical care.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Library and Information Science, School of Communication & Information, Rutgers University, New Brunswick, NJ, 08901, United States.
ORCID-Profile: Liyang Xue, Charles R Senteio, Vivek K Singh
Zitieren
APA
Xue, L., Rahman, A. M. M., Senteio, C. R., & Singh, V. K. (2026). Automated detection of stigmatizing language in Electronic Health Records (EHRs) using a multi-stage transfer learning approach. Journal of the American Medical Informatics Association. https://doi.org/10.1093/jamia/ocaf193
Vancouver
Xue L, Rahman AMM, Senteio CR, Singh VK. Automated detection of stigmatizing language in Electronic Health Records (EHRs) using a multi-stage transfer learning approach. Journal of the American Medical Informatics Association. 2026. doi:10.1093/jamia/ocaf193
BibTeX
@article{Xue2026,
author = {Xue, Liyang and Rahman, A M Muntasir and Senteio, Charles R and Singh, Vivek K},
title = {Automated detection of stigmatizing language in Electronic Health Records (EHRs) using a multi-stage transfer learning approach},
journal = {Journal of the American Medical Informatics Association},
year = {2026},
doi = {10.1093/jamia/ocaf193}
}
EUVIMED