Triage sorts patients into five levels—with one being the most critical—but often patients are clustered in level 3 or sorted incorrectly. Researchers have developed an electronic triage tool to improve patient care, helping improve physician decision-making with machine learning, according to the study published in the Annals of Emergency Medicine.
"With increases in annual visits to U.S. emergency departments (EDs), declines in capacity have led to unprecedented levels of crowding and consequential delays in care," said Scott Levin, PhD, associate professor of emergency medicine at the Johns Hopkins University School of Medicine. "So what emergency departments have to do is very quickly assess whether a patient is in need of real critical, time-sensitive treatment versus a patient who is safe to wait."
Using an algorithm based on machine learning methods, e-triage is able to identify the relationship between predictive data and patient outcomes. The study evaluated 173,000 ED visits to test the efficiency of the system and found significant differences in patient priority levels. E-triage found that of the over 65 percent of patients in level three, 10 percent of them would have benefited from being sorted up a level. The number of patients down-triaged also increased.
"Machine-based learning takes full advantage of electronic health records and allows a precision of outcomes not previously realizable," said Gabor Kelen, MD, director of the department of emergency medicine at the Johns Hopkins. "It is the wave of future healthcare, although some providers may be hesitant. Decision aids that take advantage of machine-learning are also highly customizable to meet the needs of an emergency department's patient population and local health care delivery systems."