Johannes Kepler University Linz, Kepler University Hospital, Department of Anesthesiology and Critical Care Medicine.
European Laboratory for Learning and Intelligent Systems Unit Linz, Linz Institute of Technology Artificial Intelligence Lab, Institute for Machine Learning, Johannes Kepler University.
Eur J Emerg Med. 2023 Dec 1;30(6):408-416. doi: 10.1097/MEJ.0000000000001068. Epub 2023 Aug 14.
Patient admission is a decision relying on sparsely available data. This study aims to provide prediction models for discharge versus admission for ward observation or intensive care, and 30 day-mortality for patients triaged with the Manchester Triage System.
This is a single-centre, observational, retrospective cohort study from data within ten minutes of patient presentation at the interdisciplinary emergency department of the Kepler University Hospital, Linz, Austria. We trained machine learning models including Random Forests and Neural Networks individually to predict discharge versus ward observation or intensive care admission, and 30 day-mortality. For analysis of the features' relevance, we used permutation feature importance.
A total of 58323 adult patients between 1 December 2015 and 31 August 2020 were included. Neural Networks and Random Forests predicted admission to ward observation with an AUC-ROC of 0.842 ± 0.00 with the most important features being age and chief complaint. For admission to intensive care, the models had an AUC-ROC of 0.819 ± 0.002 with the most important features being the Manchester Triage category and heart rate, and for the outcome 30 day-mortality an AUC-ROC of 0.925 ± 0.001. The most important features for the prediction of 30 day-mortality were age and general ward admission.
Machine learning can provide prediction on discharge versus admission to general wards and intensive care and inform about risk on 30 day-mortality for patients in the emergency department.
患者入院是一项依赖于有限数据的决策。本研究旨在为使用曼彻斯特分诊系统分诊的患者提供关于留观病房或重症监护室收治与出院、以及 30 天死亡率的预测模型。
这是一项单中心、观察性、回顾性队列研究，数据来自奥地利林茨的开普勒大学医院多学科急诊部患者就诊后十分钟内的资料。我们分别使用随机森林和神经网络训练机器学习模型，以预测留观病房或重症监护室收治与出院、以及 30 天死亡率。为了分析特征的相关性，我们使用了置换特征重要性。
共纳入 2015 年 12 月 1 日至 2020 年 8 月 31 日期间的 58323 名成年患者。神经网络和随机森林预测留观病房收治的 AUC-ROC 为 0.842±0.00，最重要的特征是年龄和主要诉求。对于重症监护室收治，模型的 AUC-ROC 为 0.819±0.002，最重要的特征是曼彻斯特分诊类别和心率，对于 30 天死亡率的预测，AUC-ROC 为 0.925±0.001。预测 30 天死亡率的最重要特征是年龄和普通病房收治。
机器学习可为急诊患者的普通病房和重症监护收治与出院以及 30 天死亡率风险提供预测。
