Kaiser researchers tap machine learning to identify subgroups of sepsis patients

Kaiser Permanente researchers created a machine learning algorithm to sort sepsis patients into subgroups, which could better inform treatment options, according to a study published in the Journal of the American Medical Informatics Association.

Researchers from the Kaiser Permanente Northern California Division of Research in Oakland, Calif., conducted the study. They analyzed EHR data on 29,253 adult sepsis patients hospitalized in Northern California between 2010 and 2013. Researchers used machine learning to develop a clinical signature for each patient based on each person's unique combination of 42 clinically recognizable treatment topics, such as diabetes or pneumonia. 

"When we used machine learning to examine these patterns closely, it turns out that there are very few sepsis patients that can be simply labeled as having 'pneumonia,'" senior author Vincent Liu, MD, a critical care physician and researcher at Kaiser, said in a news release. "We came up with [a] very … colorful array of subgroups. For the first time, this illustration vividly displays just how complex and heterogeneous sepsis patients are."

Based on these findings, Dr. Liu said sepsis should be viewed as an umbrella term —similar to cancer — in which patients are treated based on their specific subtypes.

"You wouldn't give the same treatment to everyone with cancer even if it's in the lung, and it's the same with sepsis," Dr. Liu said.

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.


Featured Whitepapers

Featured Webinars