AUTHOR=Yu Sean C. , Gupta Aditi , Betthauser Kevin D. , Lyons Patrick G. , Lai Albert M. , Kollef Marin H. , Payne Philip R. O. , Michelson Andrew P. 

TITLE=Sepsis Prediction for the General Ward Setting

JOURNAL=Frontiers in Digital Health

VOLUME=Volume 4 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/digital-health/articles/10.3389/fdgth.2022.848599

DOI=10.3389/fdgth.2022.848599

ISSN=2673-253X

ABSTRACT=Objective: Develop and evaluate a sepsis prediction model for the general ward setting and extend the evaluation through a novel pseudo-prospective trial design.
Design: Retrospective analysis of data extracted from electronic health record (EHR).
Setting: Single, tertiary-care academic medical center in St. Louis, MO, USA.
Patients: Adult, non-surgical inpatients admitted between 1/1/2012 and 6/1/2019.
Interventions: None.
Measurements and Main Results: Of the 70,034 included patient encounters, 3.1% were septic based on the Sepsis-3 criteria. Features were generated from EHR data and used to develop a machine learning model to predict sepsis 6-hours ahead of onset. The best performing model had an AUROC (c-statistic) of 0.862 ± 0.011 and AUPRC of 0.294 ± 0.021 compared to that of Logistic Regression (0.857 ± 0.008 and 0.256 ± 0.024) and NEWS 2 (0.699 ± 0.012 and 0.092 ± 0.009). In the pseudo-prospective trial, 388 (69.7%) septic patients were alerted on with a specificity of 81.4%. Within 24 hours of crossing the alert threshold, 20.9% had a sepsis-related event occur.
Conclusions: A machine learning model capable of predicting sepsis in the general ward setting was developed using EHR data. The pseudo-prospective trial provided a more realistic estimation of implemented performance, and revealed that the model was able to recognize critical illness 52.6 hours ahead of practitioners.