A pre-trained language model for emergency department intervention prediction using routine physiological data and clinical narratives

• Published in: International journal of medical informatics (Shannon, Ireland) Ireland), 2024-11, Vol.191, p.105564, Article 105564
• Main Authors: Huang, Ting-Yun, Chong, Chee-Fah, Lin, Heng-Yu, Chen, Tzu-Ying, Chang, Yung-Chun, Lin, Ming-Chin
• Format: Article
• Language: English
• Subjects: Clinical decision support system; Clinical narratives; Critical interventions; Natural language processing; Pre-trained language model
• ISSN: 1386-5056; 1872-8243; 1872-8243
• DOI: 10.1016/j.ijmedinf.2024.105564

•Our study merges physiological data with narratives using BioClinicalBERT, boosting emergency intervention predictions.•BioClinicalBERT excels in ER environments, achieving an impressive AUROC score of 0.9, outperforming other models.•This research advances AI in emergency care, focusing to reducing diagnostic errors and improve patient outcomes.•The findings highlight the potential for creating a decision support system in emergency room. The urgency and complexity of emergency room (ER) settings require precise and swift decision-making processes for patient care. Ensuring the timely execution of critical examinations and interventions is vital for reducing diagnostic errors, but the literature highlights a need for innovative approaches to optimize diagnostic accuracy and patient outcomes. In response, our study endeavors to create predictive models for timely examinations and interventions by leveraging the patient’s symptoms and vital signs recorded during triage, and in so doing, augment traditional diagnostic methodologies. Focusing on four key areas—medication dispensing, vital interventions, laboratory testing, and emergency radiology exams, the study employed Natural Language Processing (NLP) and seven advanced machine learning techniques. The research was centered around the innovative use of BioClinicalBERT, a state-of-the-art NLP framework. BioClinicalBERT emerged as the superior model, outperforming others in predictive accuracy. The integration of physiological data with patient narrative symptoms demonstrated greater effectiveness compared to models based solely on textual data. The robustness of our approach was confirmed by an Area Under the Receiver Operating Characteristic curve (AUROC) score of 0.9. The findings of our study underscore the feasibility of establishing a decision support system for emergency patients, targeting timely interventions and examinations based on a nuanced analysis of symptoms. By using an advanced natural language processing technique, our approach shows promise for enhancing diagnostic accuracy. However, the current model is not yet fully mature for direct implementation into daily clinical practice. Recognizing the imperative nature of precision in the ER environment, future research endeavors must focus on refining and expanding predictive models to include detailed timely examinations and interventions. Although the progress achieved in this study represents an encouraging step towards a more innovative