AI can identify at-risk sepsis patients, study finds

Listen
Text
  • Small
  • Medium
  • Large

A team of researchers from Philadelphia-based Hospital of the University of Pennsylvania developed a machine learning algorithm to predict patients at risk for severe sepsis.

Lead investigator Heather Giannini, MD, and her team trained the algorithm using EHR data — including lab results and physiological data — for 162,212 patients discharged from three University of Pennsylvania Health System acute care hospitals between 2011 and 2014.

The researchers validated the algorithm between October 2015 and December 2015 with 10,448 patients. The algorithm found roughly 3 percent of all acute-care patients were at risk for sepsis. The researchers used EHRs to alert these patients' care teams about the prediction.

The researchers on May 24 presented the study at the 2017 American Thoracic Society International Conference in Washington, D.C.

"We were hoping to identify severe sepsis or septic shock when it was early enough to intervene and before any deterioration started," said senior author Craig Umscheid, MD. "The algorithm was able to do this."

Copyright © 2021 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.

 

Featured Whitepapers

Featured Webinars