A care team at Aurora, Colo.-based UCHealth developed a sepsis-detecting AI tool that could save hundreds of lives per year — and with clinician feedback, they modified it to lessen the burden of a new technology, The Colorado Sun reported Nov. 10.
Using AI to detect sepsis has been investigated for years, but the results have been flimsy, according to the article. To counteract this, CT Lin, MD, chief medical information officer of UCHealth, and his colleagues worked to create an early-detection tool for sepsis. They piloted the program for three months, but no improvement happened.
Front-line clinical staff said they were too busy focusing on immediate tasks, and sometimes, they would be so busy they would write patients' vital signs on their arms to later put the data in the EHR at the end of the day, delaying the timeliness of sepsis alerts.
Back at the drawing board, Dr. Lin and his team decided to shift the alerts to the system's virtual health center in Aurora. There, intensive care nurses are stationed 24/7 to monitor patient information for as many as 500 patients at a time. When an alert sounds, a nurse can grab the patient's chart and enter the room remotely. They then determine if on-site action is needed and contact the bedside employees.
Amy Hassell, BSN, RN, director of UCHealth Virtual Health Center, called the sepsis tool "the bat signal."
"We've seen our time-to-identification for sepsis improved by over two hours," Ms. Hassell told The Sun. "For every hour you delay sepsis care, your mortality goes up by 10%. So to find sepsis earlier is a big deal for your mortality."