A recent external validation study by Wong et al1  examined the Epic Sepsis Model (ESM), an early warning system in Epic electronic health records (EHRs) designed to assist clinicians in predicting the onset of sepsis.1  The researchers performed the retrospective cohort study on 38 455 hospitalizations at Michigan Medicine between December 6, 2018, and October 20, 2019.1 

Wong et al1  found that sepsis occurred in 2552 (7%) of the sample,1  meaning ESM failed to identify 1709 (67%) patients who were diagnosed with sepsis in the 2552 sample. The failure occurred despite alerts firing for an ESM score of 6 or greater in 6971 of all 38 455 hospitalized patients (18%).1  Of the 1709 patients with sepsis, 1030 (60%) still received timely antibiotics, indicating clinicians predicted patients who would develop sepsis better than ESM.1  The poor predictive ability of ESM combined with the widespread...

You do not currently have access to this content.