Loading...
Electronic early warning systems (EWS) in hospitals use clinical data — often vital signs, labs, or other objective physiologic measurements — to attempt to predict patient deterioration and enable early intervention. However, EWS that use physiologic parameters have yielded mixed results (NEJM JW Gen Med Oct 1 2021 and JAMA Intern Med 2021; 181:1065). Investigators conducted a cluster-randomized trial in 74 hospital units within 2 healthcare systems (≈60,000 patient admissions) to test an EWS that used real-time nursing documentation processed by a machine-learning algorithm.
Patients in units randomized to the EWS algorithm had significantly lower in-hospital mortality (0.55% vs. 0.95%; number needed to treat, 250), shorter length of stay …