507 - Association of Emergency Severity Index and Resource Use for Children Presenting to Emergency Departments with Mental and Behavioral Health Conditions

Publication Number: 507.3588

Ashley Foster, University of California, San Francisco, School of Medicine, San Francisco, CA, United States; Cora H. Ormseth, University of California, San Francisco, School of Medicine, San Francisco, CA, United States; Jacqueline Grupp-Phelan, UCSF Benioff Children's Hospital San Francisco and Oakland, San Francisco, CA, United States; Sarah J. Lowry, Seattle Children's, Seattle, WA, United States; Elizabeth Alpern, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, United States; Jennifer Hoffmann, Northwestern University, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, United States

Background: Emergency Severity Index (ESI) is utilized in the emergency department (ED) to differentiate patients on arrival who require immediate attention from those who can safely wait for care. While ESI has demonstrated validity for triage of children with acute injuries and medical illnesses, its performance for mental and behavioral health (MBH) presentations in children has not been well studied.

Objective: We compared time-to-evaluation among children presenting with MBH symptoms versus non-MBH symptoms across ESI levels, and we examined the association between ESI level and number of ED resources used.

Design/Methods: This retrospective cross-sectional study examined ED visits by US children 5-17 years old using the 2018-2021 National Hospital Ambulatory Medical Care Survey. We identified MBH visits by reason for visit code categories including mood, behavioral, substance use, psychosis, and suicidal behavior. ESI was categorized as immediate=1, emergent=2, urgent=3, semi-urgent=4, nonurgent=5. We used the ESI handbook to define resource types: laboratory studies, electrocardiogram, radiograph, advanced imaging, intravenous fluids, nebulized medication, specialist consultation, simple and complex procedures. An adjusted negative binomial regression model was used to examine the association of ESI level and number of ED resources used.

Results: We identified 400 pediatric visits for MBH symptoms and 4,929 for non-MBH symptoms, representing 2,822,531 and 39,768,744 visits nationally. Of visits for MBH symptoms, the ESI distribution was 1% ESI 1, 45% ESI 2, 31% ESI 3, 18% ESI 4, and 5% ESI 5, with 75% resulting in ED discharge. Within each ESI category, there was no significant difference in time-to-evaluation by provider between visits with MBH versus non-MBH symptoms. During visits for MBH symptoms, greater than 1 resource type was used in 53% of ESI 1-2, 53% ESI 3, and 33% of ESI 4-5 visits (Table 1). After adjusting for visit-level characteristics, the estimated incidence rate ratio for number of resource types used was 1.55 (95% CI 1.16, 2.08) for ESI 1-2 and 1.41 (95% CI 1.08, 1.83) for ESI 3, compared to ESI 4-5 (Table 2).

Conclusion(s): An ESI level 2 is the most common designation for children seen in the ED for MBH symptoms. Although the number of resource types was associated with ESI acuity, a third of visits with ESI 4-5 required more than 1 resource, suggesting under-triage. Further study is needed to understand contributors to over- and under-triage of children with MBH symptoms presenting to the ED.

Markers of resource utilization during emergency department visits of children with mental and behavioral health symptoms


Association of Emergency Severity Index and number of resource types used for pediatric visits for a mental and behavioral health symptom, adjusted for visit-level characteristics


Markers of resource utilization during emergency department visits of children with mental and behavioral health symptoms


Association of Emergency Severity Index and number of resource types used for pediatric visits for a mental and behavioral health symptom, adjusted for visit-level characteristics