Behavioral Health Triage for Children in the Emergency Department

Emergency department (ED) triage plays a central role in sorting patients based on urgency, distinguishing those who require prompt intervention from those who can safely wait for care.1 When emergency systems are under pressure, such as during disasters, pandemics, or workforce shortages, effective triage becomes even more critical to ensure that limited resources are distributed fairly and efficiently. In the United States, the Emergency Severity Index (ESI) is the predominant triage framework, using a 5-level scale to categorize patient acuity.2 This system has been shown to perform reliably for children presenting with acute medical conditions and traumatic injuries.3 Behavioral health concerns, inclusive of mental health and substance use symptoms, account for a substantial proportion of emergency visits by children in the United States, yet the applicability of ESI for these presentations remains insufficiently characterized.4,5 Current ESI guidance classifies patients as high acuity (ESI level 2) when there is a concern for harm to self, others, or surroundings. As a result, children with behavioral health symptoms are frequently assigned this designation.6 Limited studies have examined how ESI scores are applied by triage nurses for children presenting with behavioral health concerns, including to what extent ESI may over- or underestimate clinical acuity. Optimal triage systems should strike a balance of avoiding excessive prioritization and dangerous delays in care. The Australian Mental Health Triage Scale (AMHTS; Figure 1) was specifically designed to assess urgency in patients with behavioral health presentations, and is routinely used across Australia.7 This five-tiered system links acuity categories to define time targets for clinical evaluation. Prior investigation of U.S. adults has demonstrated that AMHTS more consistently differentiates urgency and achieves recommended evaluation timelines compared to ESI.8 Despite its widespread use in Australia, evidence supporting AMHTS in children who seek emergency care is limited. This study, funded by the Pediatric Pandemic Network, aimed to evaluate interrater reliability of AMHTS and ESI for children presenting with behavioral health symptoms using written case scenarios rated by ED triage-trained nurses at two children's hospitals.9 To do this, triage-experienced nurse participants reviewed standardized written clinical vignettes and independently assigned acuity levels using ESI and AMHTS tools. This vignette-based approach has been previously applied to assess triage reliability in emergency care of children.3 A convenience group of 30 nurses from two urban, academic children's hospitals in the United States participated. Thirty brief case descriptions with a range of illness severity levels were developed by a multidisciplinary group including pediatric emergency medicine physicians, emergency medicine nurses, and child and adolescent psychiatrists. Reference materials were available to participants as they assigned acuity scores. Participants were predominantly female and White, with a median of 7 years of emergency department triage experience. Interrater reliability was higher for AMHTS (intraclass correlation 0.81; 95% confidence interval 0.72-0.88) than ESI (intraclass correlation 0.63, 95% confidence interval 0.50-0.76). Stratified analyses by triage experience (≤7 years and >7 years) demonstrated comparable interrater reliability within each subgroup. Although the most frequently assigned acuity level was 2 in both triage systems, the proportion assigned a 2 was considerably higher for ESI. Over 50% of cases designated as ESI level 2 received a different acuity assignment under AMHTS, demonstrating notable divergence in acuity classification between the two triage systems.Figure 1: Australian Mental Health Triage ScaleEmergency nurses play a central role in determining triage acuity for children with behavioral health presentations, a task that frequently requires complex risk assessment under conditions of uncertainty. Although ESI performs reliably, the lower interrater reliability for pediatric behavioral health presentations suggests nurses may apply acuity criteria more inconsistently, even with comparable triage training and experience. Similar interrater reliability among participants regardless of experience level indicates that the variability may reflect structural limitations within existing triage frameworks rather than individual proficiency. This may leave nurses to navigate complex assessments without behavioral health-specific guidance. Addressing these limitations will require prospective evaluation of AMHTS in U.S. emergency departments, while explicitly considering how such a tool can function within triage infrastructures anchored in the use of ESI. Initial studies should assess AMHTS performance in live clinical settings, including associations with time to evaluation, resource utilization, patient safety, and disposition. Additionally, nursing perspectives to evaluate usability, cognitive workload, and optimal workflow integration are needed. Prior U.S. experience demonstrates that behavioral health-specific triage principles can be operationalized in pediatric emergency care. Stricker and colleagues described implementation of an AMHTS-informed triage framework in a specialized pediatric emergency department supported by dedicated behavioral health spaces and on-site psychiatry residents supervised by pediatric psychiatry attending physicians.10 Although such resources are not available in many U.S. emergency departments, this work highlights an opportunity to evaluate how behavioral health-specific triage tools might function within ESI-based emergency departments that vary in available behavioral health staffing, physical space, and institutional support.