Feinstein Institutes for Medical Research, the research arm of New Hyde Park, N.Y.-based Northwell Health, developed an artificial intelligence tool to predict which patients will remain stable overnight and don't need to be awoken for vital monitoring, according to a study published Nov. 13 in Nature Partner Journals Digital Medicine.
Researchers collected 24.3 million vital sign measurements, drawn from multiple Northwell Health hospitals between 2012 and 2019, to build an overnight stability prediction algorithm based on indicators such as respiratory rate, heart rate, temperature, systolic blood pressure and age.
The tool, which Northwell refers to as "Let Sleeping Patients Lie," cut in half the number of patients who were awoken during the night for vital sign checks, misclassifying less than two of 10,000 cases.
"Rest is a critical element to a patient's care, and it has been well-documented that disrupted sleep is a common complaint that could delay discharge and recovery," Theodoros Zanos, PhD, assistant professor at Feinstein Institutes' Institute of Bioelectronic Medicine, said in a news release. "Our findings highlight the safety and accuracy of machine learning-based solutions to pave the way for more peaceful and safe sleep in a hospital."
The predictive model also reduces nurses' workload, as they spend 20-35 percent of their time recording vital signs and about 10 percent collecting them.
The research team is working to launch the tool at several Northwell Health hospitals.