Evaluation of protocols allowing emergency medical technicians to determine need for treatment and transport

Objective: To determine whether emergency medical technicians (EMTs) can sa fely apply protocols to assign transport options and to assess agreement be tween groups of providers on application of the protocols. Methods: Develop ed protocols categorized patients as: 1) needs ambulance; 2) go to the emer gency department (ED) by alternative means; 3) contact primary care provide r (PCP); or 4) treat and release. After education on the application of the protocols, first responders and ambulance EMTs categorized patients at the scene prior to transport but did not change current practice. Ambulance re ports were reviewed using a predetermined list of critical events that sign ified the need for an ambulance. Results: The EMTs categorized 1,300 study patients as follows: 1,023 (79%) needed ambulance transport, 200 (15%) coul d go to the ED by alternative means, 63 (5%) could contact a PCP, 14 (1%) c ould be treated and released. Categorizations by a first responder and the transporting EMT were compared for 209 patients. Collapsing categories to " need ambulance/do not need ambulance" showed fair concordance (kappa = 0.51 ). Initially, 30 of 277 (11%) patients determined not to need an ambulance appeared to experience a critical event. After review, 23 patients had even ts that may not warrant advanced life support transport. Seven (3%) had cri tical events in the ambulance warranting ambulance transport. Most were mis categorized by the EMT. Overall sensitivity and specificity for identifying patients needing ambulance transport were 94.5% and 32.8%, respectively. C onclusions: From 3% to 11% of patients determined on scene not to need an a mbulance had a critical event. Emergency medical services systems need to d etermine an acceptable rate of undertriage. Further study is needed to dete rmine whether better adherence to the protocols might increase safety.