This is the authors' abstract. We don't add key points for this paper.
Background One of the great challenges in the emergency department (ED) is to accurately differentiate critically ill from stable patients. Although there are many triage systems in the world, these all rely to varying degrees on the subjective judgment of triage officers, which may contribute to mis-triage errors. To accurately differentiate critically ill from stable patients, we present a machine learning system (MLS) that can provide a second opinion to ED triage officers to help recognize critically ill patients. The system also provides individual explanations, visible to triage officers, as to why a patient scored the way they did. Methods To derive the MLS, an existing dataset of 22,272 patient encounters from 2012 to 2019 from our institution’s electronic emergency triage system (EETS) was used for algorithm training. Once the MLS was derived, we conducted a prospective randomized cohort study of ED patients between March and April 2020. The mis-triage rate (defined as a patient initially triaged to a EETS level 3 or 4 acuity that was then up-triaged or taken to the resuscitation room) was calculated with and without the input of the MLS. Results The MLS system achieved an area under the receiver operating characteristic curve (AUC) value of 0.86 in a retrospective validation dataset of 2,000 cases. In the prospective cohort study, compared to the traditional triage system’s 1.2% mis-triage rate in the control group, the mis-triage rate in the MLS-assisted group was 0.9%. Conclusion This MLS method with a real-time practical explanation for triage officers was able to lower the mis-triage rate of ED patients in need of critical care.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Liu et al. (2022) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: