Abstract
INTRODUCTION: The Emergency Severity Index (ESI) triage tool is a commonly used triage tool in Emergency Departments (EDs). Use of the ESI is associated with challenges emanating from such variables as patients' presentations, chief complaints, disease severity and patient ages. This systematic review aims to examine the utility of the ESI triage tool in the prediction of patient's mortality rate and admission to critical care units. METHODS: A systematic search was conducted, covering the period from 1 April 2009 to 28 February 2023. Only studies published in English were included. The systematic review included primary research studies assessing and discussing the impact of using the ESI triage tool on mortality rates and admission to critical care units. Studies that reported the validity and reliability of the ESI triage tool were also included. Eligible studies specifically defined variable outcomes, such as death and admission to critical care units, in adult patients (aged 15 years and older). Only randomized controlled trials (RCTs) and observational studies were included, while retrospective studies were excluded. RESULTS: Of 2159 records screened, six studies were included (five cohort studies, one RCT), with sample sizes ranging from 580 to 9798 participants. ESI showed high inter-rater reliability (κ = 0.75), sensitivity (74.3 %), and specificity (94.4 %), with lower under-triage (10.7 %) and over-triage (6.2 %) compared with other triage systems. The predictive accuracy for mortality and ICU admission ranged from AUC 0.64-0.75, improving to AUC 0.91 when combined with additional clinical parameters. Studies consistently reported that ESI outperformed or equaled comparator triage tools. CONCLUSIONS: The ESI triage tool demonstrated strong validity and reliability in predicting mortality rates and critical care admissions across diverse adult populations. Studies showed that ESI performs comparably or better than other triage systems and can be enhanced when combined with clinical parameters. Consistent application and proper training were identified as crucial factors for optimizing its predictive accuracy and clinical utility in emergency settings.