Machine learning model to predict mortality in patients with skin and soft tissue infection in emergency department
XGBoost outperformed LR, KNN, SVM, and RF in predicting SSTI mortality, offering both higher accuracy and operational efficiency. Its sequential tree-building, regularization, and robust handling of missing data enable superior discrimination in tabular clinical datasets. The simplified model, requiring only standard admission variables, provides a fast, cost-effective, and highly accurate tool for early identification of high-risk patients in the ED.
Authors
- Kai-Hsiang Wu (ORCID: https://orcid.org/0000-0002-7586-882X)
- Wen‐Chih Fann (ORCID: https://orcid.org/0000-0002-0571-0293)
- Cheng‐Ting Hsiao (ORCID: https://orcid.org/0000-0002-4173-4022)
- Chiao-Hsuan Hsieh
- Chia‐Peng Chang (ORCID: https://orcid.org/0000-0003-2856-3481)
- Yuwei Chen (ORCID: https://orcid.org/0000-0003-0148-3609)
- Po-Han Wu
- Leng-Chieh Lin (ORCID: https://orcid.org/0000-0001-6524-2324)
Institutions
- Chang Gung University of Science and Technology (TW)
- Chang Gung University (TW)
- Shu-Te University (TW)
- Chiayi Chang Gung Memorial Hospital (TW)
Publication Details
- Journal
- Scandinavian Journal of Trauma Resuscitation and Emergency Medicine
- Published
- 2025-09-24
- DOI
- https://doi.org/10.1186/s13049-025-01463-7
- Primary Topic
- Streptococcal Infections and Treatments
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Chang Gung Medical Foundation
- National Science and Technology Council
- Chiayi Chang Gung Memorial Hospital