Reliability and Validity of the Modified Emergency Severity Index-2 as a Triage Tool

ABSTRACT

PURPOSE: We measured the inter-rater reliability and validity of the modified Emergency Severity Index-2 ((m)ESI-2) as a triage tool in the emergency department (ED). METHODS: We applied (m)ESI-2 to 2,724 patients who visited three EDs. In each ED, emergency medicine residents, interns, and nurses independently evaluated the severity of the patient with (m)ESI-2. Data on the triage result, the patient disposition, and the length of stay in the ED were collected prospectively. For patients who were admitted or who stayed more than 24 hours in the ED, the APACHE II score was calculated. RESULTS: The weighted kappa value was 0.602 between residents and interns, 0.541 between residents and nurses, and 0.451 between interns and nurses. Hospitalization (ED death, admission, transfer) rates were 100% in category 1 and 82.4%, 53.8%, 17.2%, 6.2% in categories 2 through 5, respectively. The mean length of stay was longest in category 3 (687 min) and shortest in category 5 (150 min). The mean APACHE II score was highest in category 1 (10.8) and 8.6, 6.2, 5.2, 2.9 in categories 2 through 5, respectively. CONCLUSION: The (m)ESI-2 demonstrated good to fair interrater reliability among residents, interns, and nurses, and the resultant categories were related with the hospitalization rates, the length of stay in the ED, and the APACHE II score. With further modification and refinement, (m)ESI-2 can be a reliable and a valid triage tool in the ED.