Epic Systems has revamped its widely criticized sepsis prediction model in a bid to improve its accuracy and make its alerts more meaningful to clinicians trying to snuff out the deadly condition.
Corporate documents obtained by STAT show that Epic is now recommending that its model be trained on a hospital’s own data before clinical use, a major shift aimed at ensuring its predictions are relevant to the actual patient population a hospital treats. The documents also indicate Epic is changing its definition of sepsis onset to a more commonly accepted standard and reducing its reliance on clinician orders for antibiotics as a way to flag the condition.
The changes follow the publication of a series of investigations by STAT that found an earlier version of Epic’s tool resulted in high rates of false alarms at some hospitals and failed to reliably flag sepsis in advance. One of the investigations found that the model’s use of antibiotics as a prediction variable was particularly problematic, resulting in late alarms to physicians who had already recognized the condition and taken action to treat it.
This article is exclusive to STAT+ subscribers
Unlock this article — plus in-depth analysis, newsletters, premium events, and networking platform access.
Already have an account? Log in
Already have an account? Log in
About the Author Reprints
To submit a correction request, please visit our Contact Us page.
STAT encourages you to share your voice. We welcome your commentary, criticism, and expertise on our subscriber-only platform, STAT+ Connect