Evaluating the Utility of a Clinical Sepsis AI Tool in Emergency Waiting Rooms: A Preliminary Silent Trial

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Recognising sepsis in fast-paced Emergency Department (ED) environments is challenging. The Sepsis risk AI algorithm For Emergency department WAITing room (SAFE WAIT) model was developed to detect sepsis in the ED waiting room. We aimed to evaluate the performance and clinical utility of the SAFE-WAIT model across different patient subgroups in a silent trial. We reported standard evaluation metrics for model’s performance and calculated the model’s expected utility to complement the evaluation. SAFE-WAIT demonstrates an earlier prediction, better subgroup performance compared to the existing Sepsis Alert. Its expected utility highlighted a potential value in the future decision-making process. This silent trial is a first step of our multistage approach for a safe introduction and implementation of AI systems into clinical practice.

Article activity feed