Atlanta-based Emory Healthcare developed an Epic application that alerts clinicians when patients with sepsis might benefit from a different treatment.
The health system created the artificial intelligence algorithm to suggest which patients might be good candidates to switch from traditional saline IV fluids to balanced crystalloids at the time of the fluid order.
So far, over 70% of Emory providers have accepted the recommendation in Epic and changed treatments.
"Our physicians always make the final call over any data and AI algorithm," Emory Chief Data and AI Officer Joe Depa told Becker's. "They can choose to opt in or opt out of the trial at any time."
Emory has been researching the tool as part of its PRECISE (Precision Resuscitation with Crystalloids in Sepsis) AI trial funded by Oakland, Calif.-based Kaiser Permanente's Augmented Intelligence in Medicine and Healthcare Initiative. Kaiser provided $750,000 over two years for the study.
"Besides the novel approach to testing an AI algorithm, what makes this trial unique is the speed of implementation," noted Sivasubramanium Bhavani, MD, the project's lead investigator and an assistant professor at Emory University School of Medicine. "Researchers from other institutions have been surprised at how fast we have taken an algorithm from development to validation to a systemwide clinical trial."
More than 200 patients have enrolled in the study to date. The AI algorithm runs in the background of the Epic EHR to identify potential candidates in Emory's emergency departments and intensive care units in near real time.
"We are firm believers that if done responsibly, AI can help improve patient outcomes, enable groundbreaking research, and ultimately help make our physicians' and clinical care teams' lives easier," Mr. Depa added.