The aim of this study was to compare sensitivity and validity of the emergency severity index (ESI) using 3 vital signs vs. the modified ESI (mESI) with 7 vital signs.
This prospective, observational study comprised all patients without trauma (>18 years old), presenting between 1 September 2014 and 1 October 2014 from 08:00–16:00 h, and having ESI triage scores levels 3, 4, and 5. Different from the ESI, 7 vital signs for patients in levels 3, 4, and 5 were determined. When the result revealed an abnormality in at least one of the 7 vital signs, these patients were designated as level 2 and the mESI triage was applied to them.
A total of 4536 patients were included in the study. Comparing the hospitalized patient group and the patients treated as outpatients according to the ESI and mESI levels, the ESI and the mESI level median values were 4 (3–4) and 3 (2–4), respectively, and those of patients treated on an outpatient basis were 4 (4–5) and 4 (3–5). A significant difference was observed between the two groups with regard to both the ESI and the mESI scores (p < 0.001). Furthermore, when the ESI and the mESI were compared with regard to the sensitivity and the reliability in determining the patients for hospitalization, a significant difference was determined favoring ESI [ROC curve: area under the curve mESI: 0.690, 95 % confidence interval (Cl) 0.666–0.713; ESI 0.753, 95 % Cl 0.733–0.774; p < 0.001].
The ESI, in which 3 vital signs are measured in order to distinguish only level 2 and 3 patients, is an adequate and reliable triage system.