EUVIMEDEUVIMED entdecken
Open Access

Patient Stratification for Improving Acute Chest Pain Management and Mitigating Emergency Department Crowding: Machine Learning Model Development and Validation

Patientenstratifizierung zur Verbesserung des Managements akuter Brustschmerzen und zur Verringerung von Überfüllung in der Notaufnahme: Entwicklung und Validierung eines Machine-Learning-Modells

Chih-Chia Hsieh iD, Shao-Chung Chu iD, Jung-Ting Lee iD, Chih-Hao Lin iD, Chung-Yao Kao iD

JMIR Medical Informatics · 21.08.2026 · 14 · e83099-e83099

DOI: 10.2196/83099 ↗

Auf einen Blick

ANN-Modelle sagten bei triagegrad-2-Patient:innen mit akuten Brustschmerzen ACS und Niedrigrisiko-Gruppen gut voraus. Die sensitivitätspriorisierte Variante erreichte eine ACS-Sensitivität von 0,966.

Fragestellung

Können KI-Modelle bei triagegrad-2-Patient:innen mit akuten Brustschmerzen ACS und Niedrigrisiko-Patient:innen anhand eines einzelnen hochsensitiven Troponin-T-Tests identifizieren?

Methode

Retrospektive Single-Center-Studie; ANN-Modelle mit 5-facher Cross-Validation (2016-2020) und internem Test (2021-2022). Bewertung mit AUROC, AUPRC und subgruppenspezifischen Metriken.

Ergebnisse

17.935 Besuche; 1.209 GA, 3.873 GB1, 12.853 GB2. Test-AUROC: 0,942 (95 %-KI 0,920-0,965) für GA, 0,824 (0,808-0,841) für GB1, 0,893 (0,884-0,902) für GB2.

Einschränkungen

Einzelzentrum; nur interne Validierung; das sensitivitätspriorisierte Modell verfehlte die für einen Stand-alone-Rule-out üblichen Sensitivitätsanforderungen >0,99.

Schlussfolgerung der Autor:innen

Die Autor:innen schlussfolgern, dass ANN-Classifier die klinische Risikostratifizierung und Dispositionsentscheidung unterstützen können, aber nur als klinische Entscheidungshilfe und nicht als unabhängige Rule-out-Strategie.

Kernaussagen

  • Test-AUROC für ACS-Klassifikation: 0,942 (95 %-KI 0,920-0,965)
  • ANN-S3-L: ACS-Sensitivität 0,966; NPV 0,998
  • Geeignet als Entscheidungshilfe, nicht als Stand-alone-Rule-out

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Fachartikel in der Zeitschrift JMIR Medical Informatics, online erschienen am 21.08.2026. Verfasst von Chih-Chia Hsieh und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Emergency Medicine, National Cheng Kung University Hospital, Tainan, Taiwan. Hauptthemen laut MEDLINE: Chest Pain, Emergency Service, Hospital, Machine Learning, Crowding. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Acute chest pain (ACP) is one of the most common chief complaints in the emergency department (ED), accounting for approximately 8% of all ED visits. However, among patients presenting with chest pain suggestive of cardiac origin, fewer than 10% are ultimately diagnosed with acute coronary syndrome (ACS).

Objective

AI models were developed to support clinical decision-making for triage level-2 ED patients presenting with ACP. These models aim to accurately detect ACS and reliably identify low-risk patients based on a single high-sensitivity cardiac troponin T test result. By integrating this AI-assisted strategy into clinical workflows, we aim to reduce ED length of stay, alleviate crowding, and improve health care efficiency while maintaining high safety standards.

Methods

We conducted a retrospective study using single-center data from a tertiary teaching hospital between January 2016 and December 2022. Models based on artificial neural networks (ANNs) were trained to classify patients with ACP at triage level 2 into 3 clinical classes: critical patients with ACS (subgroup GA), critical patients without ACS (subgroup GB1), and low-risk patients (subgroup GB2). The models were trained and internally validated via a 5-fold cross-validation protocol using data from 2016 to 2020. Model performance was then evaluated on the hold-out testing data (2021‐2022) using area under the receiver operating characteristic curve, area under the precision-recall curve, and subgroup-specific metrics. Feature selection methods and the Shapley Additive Explanations value analysis were applied to identify features that drive the model’s predictive power. The study was approved by the institutional review board of the National Cheng Kung University Hospital, Tainan, Taiwan (A-ER-111‐199).

Results

After excluding ED visits with missing triage data, incomplete medical histories, or unsuitable dispositions, 17,935 visits were included, with 1209 GA, 3873 GB1, and 12,853 GB2. Rendering prediction based on 24 feature variables (2 demographics, 6 vital signs, 4 blood test results, and 12 medical history), all ANN models demonstrated strong testing AUROC (95% CI) performance of 0.942 (0.920‐0.965) for GA classification, 0.824 (0.808‐0.841) for GB1, and 0.893 (0.884‐0.902) for GB2. Regarding multiclassification, ANN-S3 achieved balanced performance, with ACS sensitivity of 0.941 (95% CI 0.908‐0.973) and low-risk positive predictive value and sensitivity of 0.911 (95% CI 0.9‐0.921) and 0.837 (95% CI 0.824‐0.85), respectively. A sensitivity-prioritized variant, ANN-S3-L, increased ACS sensitivity to 0.966 (95% CI 0.94‐0.991) and negative predictive value to 0.998 (95% CI 0.996‐0.999), but at the cost of lower specificity and reduced low-risk sensitivity, indicating a safety-efficiency trade-off.

Conclusions

These findings suggest that ANN-based classifiers can effectively support clinical risk stratification and disposition decision-making in ACP care for patients presenting ≥3 hours after symptom onset. However, because even the sensitivity-prioritized ANN-S3-L variant falls short of the stringent sensitivity threshold (>0.99) typically required for a standalone ED rule-out tool, this system should be interpreted as a clinical decision-support aid rather than an independent rule-out strategy.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Chest PainEmergency Service, HospitalMachine LearningCrowdingHumansRetrospective StudiesTriageFemaleMaleMiddle AgedAgedROC CurveEmergency Room VisitsAcute Coronary SyndromePredictive Learning Models

Stichwörter der Autor:innen

AIacute chest painclinical decision supportemergency department crowdingpatient managementrisk stratificationtriage

Autor:innen und Institutionen

  • Department of Emergency Medicine, National Cheng Kung University Hospital, Tainan, Taiwan.
  • School of Medicine, National Sun Yat-Sen University, No.70 Lien-hai Road, Kaohsiung, 804, Taiwan, 886 75252000.
  • Department of Electrical Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan.

ORCID-Profile: Chih-Chia Hsieh, Shao-Chung Chu, Jung-Ting Lee, Chih-Hao Lin, Chung-Yao Kao

Zitieren

APA

Hsieh, C. C., Chu, S. C., Lee, J. T., Lin, C. H., & Kao, C. Y. (2026). Patient Stratification for Improving Acute Chest Pain Management and Mitigating Emergency Department Crowding: Machine Learning Model Development and Validation. JMIR Medical Informatics, 14, e83099-e83099. https://doi.org/10.2196/83099

Vancouver

Hsieh CC, Chu SC, Lee JT, Lin CH, Kao CY. Patient Stratification for Improving Acute Chest Pain Management and Mitigating Emergency Department Crowding: Machine Learning Model Development and Validation. JMIR Medical Informatics. 2026;14:e83099-e83099. doi:10.2196/83099

BibTeX
@article{Hsieh2026,
  author = {Hsieh, Chih-Chia and Chu, Shao-Chung and Lee, Jung-Ting and Lin, Chih-Hao and Kao, Chung-Yao},
  title = {Patient Stratification for Improving Acute Chest Pain Management and Mitigating Emergency Department Crowding: Machine Learning Model Development and Validation},
  journal = {JMIR Medical Informatics},
  year = {2026},
  volume = {14},
  pages = {e83099--e83099},
  doi = {10.2196/83099}
}
Kostenlos starten →