Artificial Intelligence-Powered Emergency Department Triage System Accelerates Flow
Research conducted in three emergency departments in Connecticut revealed that artificial intelligence-powered triage tools increased critical patient detection and shortened process times.
A quality improvement study conducted by Taylor and colleagues in three emergency departments across Connecticut showed that artificial intelligence-powered triage decision support systems increased critical care accuracy and reduced patient flow times.
Scope and Methodology of the Study
Taylor and his team evaluated the implementation of an artificial intelligence-powered triage decision support tool in emergency departments. The quality improvement study covered a total of 174,648 visits.
Working Principle of the Artificial Intelligence Tool
The system collected demographic information, mode of arrival, vital signs, chief complaint, and active problems. While triage nurses retained final decision-making authority, the tool generated an Emergency Severity Index recommendation with an individualized explanation.
Critical Care Detection and Processes
As a result of the research, the accuracy in defining critical care as high acuity increased from 78.8 percent to 83.1 percent. Low-acuity assignments rose from 23.9 percent to 35.4 percent.
Changes in Patient Flow Times
The time from patient arrival to the initial care area decreased from 12 minutes to 8 minutes. Additionally, discharge time dropped from 190 minutes to 182 minutes, and total length of stay decreased from 311 minutes to 292 minutes.