The advanced alert monitor uses algorithms created from machine learning, and data from 1.5 million patients to scan hosptital EHRs hourly and identify signs of patient decline, Gabriel Escobar, MD, lead study author, told Becker’s. The system was used across 21 Kaiser Permanente hospitals between August 2016 and February 2019 to compare outcomes between patients who reached the system’s alert threshold to similar risk patients who were not monitored with the system. The study included 15,487 patients in the alert system group, and 28,462 patients in the control group.
In addition to a lower mortality rate, researchers found patients monitored by the system had a lower chance of admission to an intensive care unit at nearly 18 percent compared to about 21 percent of patients who were not monitored, a shorter hospital stay (6.7 days versus 7.5 days) and overall lower mortality within 30 days of the initial alert (15.8 percent versus 20.4 percent.)
The system, homegrown at Kaiser hospitals, is automated and alerts are evaluated by off-site nurses, alleviating healthcare workers of additional responsibilities, according to a news release.
More articles on patient safety and outcomes:
Insurance type may affect quality of surgical care, study finds
Boston Children’s Hospital to end 2 intersex surgeries
COVID-19 viral load may predict patient outcomes, study suggests
