A Pilot Machine Learning Study Using Trauma Admission Data to Identify Risk for High Length of Stay-Reference-Cited by-同舟云学术

A Pilot Machine Learning Study Using Trauma Admission Data to Identify Risk for High Length of Stay

Published:2022-11-17 Issue: Volume: Page:155335062211399
ISSN:1553-3506
Container-title:Surgical Innovation
language:en
Short-container-title:Surg Innov

Author:

Stonko David P.¹²^ORCID,Weller Jennine H.¹,Gonzalez Salazar Andres J.¹^ORCID,Abdou Hossam²^ORCID,Edwards Joseph²,Hinson Jeremiah³⁴,Levin Scott³⁴,Byrne James P.¹,Sakran Joseph V.¹,Hicks Caitlin W.⁵,Haut Elliott R.¹⁶³⁷⁸,Morrison Jonathan J²,Kent Alistair J.¹

Affiliation:

1. Division of Trauma and Acute Care Surgery, The Johns Hopkins Hospital, The Johns Hopkins Department of Surgery, Baltimore, MD, USA

2. R. Adams Cowley Shock Trauma Center, Baltimore, MD, USA

3. Department of Emergency Medicine, The Johns Hopkins University School of Medicine, Baltimore, MD, USA

4. Malone Center for Engineering in Healthcare, The Johns Hopkins University School of Medicine, Baltimore, MD, USA

5. Division of Vascular and Endovascular Therapy, The Johns Hopkins Hospital, Baltimore, MD, USA

6. Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore, MD, USA

7. The Armstrong Institute for Patient Safety and Quality, Johns Hopkins Medicine, Baltimore, MD, USA

8. Department of Health Policy and Management, Bloomberg School of Public Health, The Johns Hopkins Baltimore, MD, USA

Abstract

Introduction Trauma patients have diverse resource needs due to variable mechanisms and injury patterns. The aim of this study was to build a tool that uses only data available at time of admission to predict prolonged hospital length of stay (LOS). Methods Data was collected from the trauma registry at an urban level one adult trauma center and included patients from 1/1/2014 to 3/31/2019. Trauma patients with one or fewer days LOS were excluded. Single layer and deep artificial neural networks were trained to identify patients in the top quartile of LOS and optimized on area under the receiver operator characteristic curve (AUROC). The predictive performance of the model was assessed on a separate test set using binary classification measures of accuracy, precision, and error. Results 2953 admitted trauma patients with more than one-day LOS were included in this study. They were 70% male, 60% white, and averaged 47 years-old (SD: 21). 28% were penetrating trauma. Median length of stay was 5 days (IQR 3-9). For prediction of prolonged LOS, the deep neural network achieved an AUROC of 0.80 (95% CI: 0.786-0.814) specificity was 0.95, sensitivity was 0.32, with an overall accuracy of 0.79. Conclusion Machine learning can predict, with excellent specificity, trauma patients who will have prolonged length of stay with only physiologic and demographic data available at the time of admission. These patients may benefit from additional resources with respect to disposition planning at the time of admission.

Publisher

SAGE Publications

Subject

Surgery

Link

http://journals.sagepub.com/doi/pdf/10.1177/15533506221139965

Reference50 articles.

1. Trauma Systems

2. Regional collaboration across hospital systems to develop and implement trauma protocols saves lives within 2 years

3. Progress in the Development of Trauma Systems in the United States