Given the medical teams that medical teams can meet day to day, nursing equipment was designed to be as simple and straightforward as its name.
The team developed a tool called nursing hand that shows nurses the electronic health record on the one hand and the AI-generated output on the other. Using a hospital-delivered mobile device, nurses are able to easily review and add information throughout their shifts. Google’s MEDLM models running within the nurse can take, analyze and develop a coherent and brief overview of relevant patient information for the upcoming nurse for delivery.
Traditionally, nurses depend on their memory of events, conversations and data points, giving way to errors and inconsistency in the information provided during handing over. With the help of nursing handling during their shift, nurses can easily find the information they and their friends have collected through the automated shift report and add their own notes so that the system continues to build. This may include relevant patient data from notes, orders, tests and more. All of this happens within a very safe cloud environment to keep patient information confidential.
HCA Healthcare is currently piloting nurse in five of her hospitals, collecting feedback and continuing to fine -tune both its AI model and the user experience.
Nurses’ orders
An important part of the structure of an effective app that the Frontline nurses want to use worked directly with them on its development. It included KC Deshetler, a registered nurse at HCA Healthcare DT & I team, which became product owner for nurse handling.
“We fed our model, which the information nurses want to know, got the model in a number of ways, used retrieval competition to identify quotes for the generated content, provided templates to organize information that we want, etc.,” explains Derehelter.
The approach they settled on, nurses show a well-known electronic health record on one side of the screen and an AI-generated output on the other. Hall, the Nurse from Tennessee, was among a group of early testers who reviewed the model’s output to highlight what was unnecessary, repeated, inaccurate or missing.
“We reviewed this process three or four different times,” Hall remembers. “And each time it was a little more accurate, a little less full of lint that we don’t need. The more we worked with it and gave our feedback, the more useful it became for the delivery setting.”
As the HCA Healthcare Pilot with five hospital has passed, plans require a version to be rolled out for use by all 99,000 nurses working across health care. So far, the solution looks very promising: Nurses testing the nursing tool at a HCA health facility has rated it as 86% factual and 90% useful.
“When I talk to nurses and nurse leaders about technology and the emergence of AI in Nursing,” says Dr. Staub-Juergers, “I often challenge them to be bold, be brave, take the keys to the car, get in the driver’s seat and use their voices to drive the design for the solution that goes on.”