At risk of undertriage? Testing the performance and accuracy of the emergency severity index in older emergency department patients

Abstract

Study objective: We test predictive validity, interrater reliability, and diagnostic accuracy of the Emergency Severity Index in older emergency department (ED) patients and identify reasons for inadequate triage.

Methods: We analyzed data of patients aged 65 years or older who were included in a prospective, single-center cohort study. Predictive validity was assessed by investigating associations of resources, disposition, length of stay, and mortality with Emergency Severity Index levels. Diagnostic accuracy was tested by calculating sensitivity and specificity of Emergency Severity Index level 1 for the prediction of a lifesaving intervention. For the assessment of interrater reliability, 2 experts independently reviewed the triage nurses' notes. Agreement was estimated as raw agreement and as Cohen's weighted κ.

Results: In total, 519 older patients were included. Emergency Severity Index level was associated with resource consumption (Spearman's ρ=-0.449; 95% confidence interval [CI] -0.519 to -0.379), disposition (Kendall's τ=-0.452; 95% CI -0.516 to -0.387), ED length of stay (Kruskal-Wallis χ(2)=92.5; df=4; P<.001), and mortality (log-rank χ(2)=37.04; df=3; P<.001). The sensitivity of the Emergency Severity Index to predict lifesaving interventions was 0.462 (95% CI 0.232 to 0.709), and the specificity was 0.998 (95% CI 0.989 to 1.000). Interrater reliability between experts was high (raw agreement 0.917, 95% CI 0.894 to 0.944; Cohen's weighted κ(w)=0.934, 95% CI 0.913 to 0.954). Undertriage occurred in 117 cases. Main reasons were neglect of high-risk situations and failure to appropriately interpret vital signs.

Conclusion: In our study, older patients were at risk for undertriage. However, our results suggest that the Emergency Severity Index is reliable and valid for triage of older patients.