Citation

Gorczyca MT, Toscano NC, Cheng JD. Comput. Biol. Med. 2019; 108: 9-19.

Affiliation

Department of Surgery, University of Rochester Medical Center, Rochester, NY, 14642, USA.

Copyright

(Copyright © 2019, Elsevier Publishing)

DOI

10.1016/j.compbiomed.2019.02.025

PMID

30965177

Abstract

Statistical theory indicates that a flexible model can attain a lower generalization error than an inflexible model, provided that the setting is appropriate. This is highly relevant for mortality risk prediction with trauma patients, as researchers have focused exclusively on the use of generalized linear models for trauma risk prediction, and generalized linear models may be too inflexible to capture the potentially complex relationships in trauma data. To improve trauma risk prediction, we propose a machine learning model, the Trauma Severity Model (TSM). In order to validate TSM's performance, this study compares TSM to three established risk prediction models: the Bayesian Logistic Injury Severity Score, the Harborview Assessment for Risk of Mortality, and the Trauma Mortality Prediction Model. Our results indicate that TSM has superior predictive performance on National Trauma Data Bank data and on Nationwide Readmission Database data.

Copyright © 2019 Elsevier Ltd. All rights reserved.

Language: en

Keywords

Machine learning in medicine; National trauma data bank; Nationwide readmission database; Risk prediction; Trauma quality improvement